Deciphering the Hidden Meaning: Scripture and the Hermeneutics of Liberation in Early Advaita Vedanta

THE UNIVERSITY OF CHICAGO DECIPHERING THE HIDDEN MEANING: SCRIPTURE AND THE HERMENEUTICS OF LIBERATION IN EARLY ADVAITA VEDĀNTA A DISSERTATION SUBMITTED TO THE FACULTY OF THE DIVISION OF THE HUMANITIES IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF SOUTH ASIAN LANGUAGES AND CIVILIZATIONS BY ALEKSANDAR USKOKOV CHICAGO, ILLINOIS AUGUST 2018 Copyright © 2018 by Aleksandar Uskokov All Rights Reserved  Dedication To the memory of Steve Collins याञव्येति हाेवाच, यरायं पुुषाे रियिे । किमेनं न जहािीति । नामेति । अन्िं वै नामान्िा कववे देवाः । अन्िमेव स िेन लाेिं जयति ।  TABLE OF CONTENTS ACKNOWLEDGMENTS ........................................................................................................ vii INTRODUCTION .................................................................................................................... 1 LIBERATION AND THE HIGHEST GOOD ........................................................................................................ 1 MAHĀ-VĀKYA ............................................................................................................................................ 9 CHARACTERS, DATES, AND ARCHIVE ........................................................................................................ 15 VEDIC THEOLOGY ..................................................................................................................................... 17 MĪMĀṀSĀ AND VEDĀNTA AS PŪRVA- AND UTTARA-MĪMĀṀSĀ ............................................................... 33 THE HISTORY OF IDEAS ............................................................................................................................. 38 STRUCTURE AND CHAPTER OVERVIEW ..................................................................................................... 46 NOTE ON THE TITLE................................................................................................................................... 53 PART ONE: CIRCUMSCRIBING THE FIELD OF VEDIC THEOLOGY CHAPTER ONE: RETHINKING THE IDEA OF SCRIPTURE IN VEDIC THEOLOGY ..................... 55 INTRODUCTION .......................................................................................................................................... 55 THE CLASSICAL THEORY OF ŚRUTI ........................................................................................................... 56 ŚRUTI IN EARLY VEDIC THEOLOGY ........................................................................................................... 61 ŚABARA’S UNDERSTANDING OF ŚRUTI ...................................................................................................... 64 FROM ŚRUTI TO ŚĀSTRA ............................................................................................................................ 72 THE IMPERSONAL NATURE OF THE VEDA .................................................................................................. 82 ŚRUTI-ŚĀSTRA AS PERCEPTION ................................................................................................................. 86 ŚRUTI IN OTHER BRĀHMAṆICAL TRADITIONS ........................................................................................... 89 ŚAṄKARA’S UNDERSTANDING OF ŚRUTI .................................................................................................... 92 CHAPTER TWO: THE MĪMĀṀSĀ MODEL OF PURUṢĀRTHA AND THE ROLE OF SCRIPTURE .. 98 INTRODUCTION .......................................................................................................................................... 98 DHARMA AND PURUṢĀRTHA ..................................................................................................................... 99 DHARMA AND RITUAL CAUSALITY ......................................................................................................... 105 THE DOCTRINE OF APŪRVA AND THE TEMPORAL UNITY OF THE RITUAL................................................ 112 THE IDEA OF MEDIATE CAUSALITY ......................................................................................................... 118 BHĀVANĀ, ĀKĀṄKṢĀ, AND THE STRUCTURAL UNITY OF SACRIFICE AND TEXT ..................................... 119 MĪMĀṀSĀ CLASSIFICATION OF VEDIC TEXTS AND THE UPANIṢADS ....................................................... 125 LANGUAGE AND PRĀBHĀKARA-MĪMAṀSĀ ............................................................................................. 136 CONCLUSION ........................................................................................................................................... 143 PART TWO: LIBERATION AND THE HIGHEST GOOD IN PRE-ŚAṄKARA VEDIC THEOLOGY INTRODUCTION ................................................................................................................ 145 CHAPTER THREE: THE HIGHEST GOOD AND LIBERATION IN PRE-ŚAṄKARA MĪMĀṀSĀ ... 153 iv  HEAVEN AS LIBERATION ......................................................................................................................... 153 LIBERATION IN KUMĀRILA’S THOUGHT: INTRODUCTION ........................................................................ 156 KUMĀRILA’S FIRST ACCOUNT OF LIBERATION ....................................................................................... 158 KUMĀRILA’S SECOND ACCOUNT OF LIBERATION ................................................................................... 169 CONCLUDING REMARKS .......................................................................................................................... 178 CHAPTER FOUR: LIBERATION IN THE BRAHMA-SŪTRA .................................................... 184 INTRODUCTORY REMARKS: METHODOLOGY ........................................................................................... 184 THE DOCTRINE OF VIDYĀ/UPĀSANA ....................................................................................................... 188 BRAHMA-VIDYĀ ..................................................................................................................................... 197 THE PRACTICE OF MEDITATION ON BRAHMAN........................................................................................ 205 ATTAINING BRAHMAN ............................................................................................................................ 209 THE SELF IN LIBERATION ........................................................................................................................ 216 BRAHMA-SŪTRA, LIBERATION, AND THE TWO GREAT UPANIṢADS ......................................................... 221 CHAPTER FIVE: THE DOCTRINE OF PRASAṄKHYĀNA ....................................................... 231 INTRODUCTION ........................................................................................................................................ 231 THE SOTERIOLOGY OF BHARTṚPRAPAÑCA .............................................................................................. 236 PRASAṄKHYĀNA IN PĀTAÑJALA YOGA-ŚĀSTRA ..................................................................................... 251 THE VEDĀNTIC PRASAṄKHYĀNA ............................................................................................................ 255 THE “ANOTHER COGNITION” MEDITATION ............................................................................................. 264 CONCLUSION ........................................................................................................................................... 273 CONCLUSION.................................................................................................................... 276 PART THREE: DHARMA, SCRIPTURE, AND THE GOOD OF MAN IN EARLY ADVAITA VEDĀNTA INTRODUCTION ................................................................................................................ 280 CHAPTER SIX: GENERAL CHARACTERISTICS OF DHARMA AND THE PATH OF DISENGAGEMENT ......................................................................................................................................... 286 INTRODUCTION ........................................................................................................................................ 286 DHARMA AND THE VALIDITY OF THE VEDA ............................................................................................ 291 THE DHARMA OF ENGAGEMENT AND DISENGAGEMENT ......................................................................... 298 ŚAṄKARA’S PSYCHOLOGY AND THE HUMAN CONDITION........................................................................ 301 DESIRE AND QUALIFICATION................................................................................................................... 312 THE ATTAINMENTS OF DHARMA ............................................................................................................. 319 WINNING THE WORLD OF THE GODS ....................................................................................................... 327 BRAHMAN AS BRAHMĀ, THE ULTIMATE ATTAINMENT OF MEDITATION ................................................. 333 CHAPTER SEVEN: LIBERATION, RITUAL, AND THE ARISING OF KNOWLEDGE .................. 341 THE SELF AND THE NATURE OF LIBERATION ........................................................................................... 343 THE ROLE OF RITUAL AND VIVIDIṢĀ ....................................................................................................... 348 ARISING OF KNOWLEDGE ........................................................................................................................ 359 THE MODEL OF CAUSALITY .................................................................................................................... 366 CHAPTER EIGHT: YOU ARE THAT, ALL RIGHT, WE JUST NEED TO FIGURE OUT WHAT: VEDĀNTA- VĀKYA AND THE IDENTITY STATEMENTS ........................................................................ 375 THE UPANIṢADS AS PARA- AND APARA-VIDYĀ ....................................................................................... 375 THE SCOPE OF PARA-VIDYĀ TEXTS ......................................................................................................... 386 v  THE IDENTITY STATEMENTS OF THE UPANIṢADS ..................................................................................... 400 THE PROBLEM OF LANGUAGE ................................................................................................................. 405 THE CATEGORIES OF “THAT” AND “YOU” AND THE NOTIONS OF BRAHMAN AND THE SELF ................... 413 THE IDENTITY STATEMENT CONTEXT ..................................................................................................... 424 CHAPTER NINE: LIBERATION AND THE INQUIRY INTO BRAHMAN .................................... 438 THE DHARMA OF DISENGAGEMENT AND DESIRE .................................................................................... 438 THE PROCESS OF KNOWING BRAHMAN ................................................................................................... 443 ŚRAVAṆA AND MANANA IN THE UPADEŚA-SĀHASRĪ .............................................................................. 459 NIDIDHYĀSANA AND PARISAṄKHYĀNA ................................................................................................... 469 PARIṢAṄKHYĀNA: A SECOND AVENUE ................................................................................................... 477 THE PURPOSE OF PARISAṄKHYĀNA AND THE NATURE OF LIBERATION ................................................... 480 PART FOUR: FROM IDENTITY STATEMENTS TO MAHĀ-VĀKYA CHAPTER TEN: SARVAJÑĀTMAN AND THE DOCTRINE OF UPANIṢADIC MAHĀ-VĀKYAS ... 487 INTRODUCTION AND A HISTORICAL NOTE ............................................................................................... 487 MAHĀ-VĀKYA IN MĪMĀṀSĀ ................................................................................................................... 494 SARVAJÑĀTMAN AND THE PRELIMINARIES .............................................................................................. 499 THE VEDĀNTIC MAHĀ-VĀKYAS ............................................................................................................. 507 CONCLUSION ........................................................................................................................................... 517 EPILOGUE......................................................................................................................... 520 BIBLIOGRAPHY ................................................................................................................ 525 ABBREVIATIONS ...................................................................................................................................... 525 LIST OF PRIMARY SOURCES ..................................................................................................................... 528 EDITIONS, TRANSLATIONS, AND SECONDARY LITERATURE .................................................................... 533 vi  ACKNOWLEDGMENTS I wish to thank first the members of my dissertation committee. Once a Sanskrit student in Madison told me after SASLI was over, “So I guess my Sanskrit GPS will now always speak with a Slavic accent.” My own Sanskrit GPS is forever tuned to the channel of the present embodiment of Bhāratī, the sarva-tantra-svatantra Gary Tubb. Studying Sanskrit with him has been one of the unique experiences of my life, not the least because I have been able to retain incomparably more than anything else I have ever learned (something I do not claim credit for). I often felt, quite frankly, as if transposed into a different realm, one where folks habitually make everything into figures of speech, joke about rice and beans, infer mountain fires when there is no matchbox, entertain themselves with prakriyās, and bake atoms, all simultaneously and with ease. I don’t believe I have witnessed a mastery comparable to his: I consider him a living proof that Brahman is śabda-tattvam. He also stoically endured my countless requests for letters and signatures. I am very proud and thankful to call myself his student. From Wendy Doniger I learned that writing can (indeed, must) be both serious and entertaining, that one can both say important things, make bold arguments, and joke at the same time. That is, be both reverent and irreverent. I cannot express how liberating that was: hers is the writing style that I attempted to emulate. Additionally, being the closest approximation to an omniscient, omnibenevolent and, when the push comes to shove, omnipotent being that I am personally acquainted with, I sought her help on many occasions and she always delivered. For all of this, I remain ever grateful and indebted. I am, finally, thankful to Dan Arnold. As it will be obvious to the educated reader of this dissertation, my understanding of the character and nature of Indian philosophy, particularly with regard to what is characteristically philosophical in Indian systematic thought and in terms of appreciating the key non-sectarian notions and forms of argument, has taken shape in his vii Sanskrit classes and through reading his Buddhists, Brahmins, and Belief. In many ways, he remains the reader that I feel peeping at my screen as I struggle to state my arguments as clearly as I can. In addition to the dissertation committee, I wish to acknowledge two other influences that were formative for my intellectual development during doctoral studies. Reading Larry McCrea’s “The Teleology of Poetics in Medieval Kashmir” was, appropriately, a paradigm shift for me in how I approached the study of my archive. This work is the inspiration behind pursuing the Mīmāṁsā roots of the Advaita understanding of mahā-vākya, not at all obvious at first sight, and the significance of Larry’s oeuvre for the dissertation is incomparably larger than references and bibliography might suggest. To Steve Collins and his Pali classes, I owe learning the method of engaging with texts that I now identify with. More generally, just being able to absorb such an extraordinary intellect has been one of those rare blessings in life. The news of his early and sudden death felt almost as painful as another early and sudden death nineteen years ago, but now I would rather feel happy for knowing him and learning from him than sad that he is no longer with us. Were he still with us, I am sure he would have been horrified at defense time by the size of this thing, but I would have replied with one of his favorites: “I am sorry, but I did not have the time to make it short.” I also wish to thank my Sanskrit teacher Whitney Cox. He has been such a brilliant example of the kind of depth and breadth scholar I aspire to be: expert in several fields and highly competent in all. He also played an instrumental role in the first paper I published during doctoral studies. For all that, I remain a debtor. Speaking of Sanskrit, I wish to thank my partners in crime in SALC and the Divinity School for the countless happy hours we spent together with the language of the Gods in Foster 209 and the Reg: Jo Brill, Nell Hawley, Margherita Trento, Jahnabi Barooah, Talia Ariav, Victor viii D’Avella, Eric Gurevitch, Nabanjan Maitra, Stephen Walker, Pierre-Julien Harter, Sonam Kachru, Ted Good, Karl Schmid, Jetsun Deleplanque, Ishan Chakrabarti, Eduardo L. Acosta, Davey Tomlinson, Jamal Jones, Anil Mundra, Jackson Macor and others. Many other friends in SALC were always a good company, especially Ahona Panda, Malar Jayanth, Emma Kalb, Joya John, Jane Mikkelson, Ranu Roychoudhuri, and Taimoor Khan. Three individuals were crucial for my survival in Chicago. Abhishek Ghosh found me a place, fed me, and advised me on all kinds of practical issues, from course registration to taxes. He, Bambi and Rai are like a family here. Ilanit Loewy Shacham told me everything I needed to know about the dissertation proposal and remained the strongest support in that hell-leaning- limbo otherwise called “Advanced Residence.” They say that Alicia Czaplewski was the Foster mother to everyone in SALC. I don’t know about everyone—“one cannot think with someone else’s head,” wrote Meša Selimović—but she sure was every bit of a mother to me. I am endlessly thankful to all three of them. Tracy L. Davis was also very helpful with the dissertation defense. Finally, SALC-wise, I wish to thank my Bangla teachers Mandira Bhaduri and Thibaut D’Hubert. Mandira was kind to teach me what I was interested in learning, not what I would have been expected to be interested in. If Bādarāyaṇa and Plato were right about eternal names and ideas, I am sure that the model scholar (as a universal) has a beard, is nearly 7ft tall, knows most languages from Britain to Bangladesh, and answers to the name “Thibaut.” I wish to thank, further, several colleagues who were a pleasure to collaborate (and hang out) with on conference panels and otherwise: Barbara Holdrege, Travis Chilcott, Hrvoje Čargonja, Kenneth Valpey, Ravi Gupta, Anand Venkatkrishnan, and Kiyokazu Okita. During my time at the University of Chicago, I spent four very happy summers teaching Intermediate Sanskrit at the South Asia Summer Language Institute (SASLI), University of ix Wisconsin – Madison. Large chunks of the dissertation were written in Madison, where SASLI became my adoptive home and the Memorial Library my nurturing parent. I owe gratitude to the SASLI administrative and academic staff, particularly Anne Naparstek, Lalita du Perron, Laura Hammond, Sarah Beckham, John Burmaster, and Mark Kenoyer. Madison was otherwise very welcoming—as David Shulman once told me, it is the closest approximation to svarga—and to Katarzyna Pażucha, Mandira Bhaduri, Anya Golovkova, Anil Mundra, Mitilesh Mishra, Nikola Rajić and many others I remain indebted for the great company. Back in the old country, I am most thankful to Viktor Ilievski. It was he who lent me a copy of Michael Coulson’s “Teach Yourself Sanskrit,” my first and favorite primer, serendipitously found somewhere in Rijeka near the end of the last millennium. We started “teaching ourselves Sanskrit” at the same time in 1999, and throughout all tectonic shifts and changes in life, Sanskrit remained the one constant with both of us. While I was in Chicago and he in Budapest, we produced a joint translation of the Hitopadeśa and a volume of select Upaniṣads, and he remains my main interlocutor. I am also indebted to Dejan Jordanov, “ а давачот,” for the best jokes and laughs in life. If there is one thing that I missed here, it was the bālya with the two of them to supplement the pāṇḍitya. My research would not have been possible without the financial support of several institutions and bodies. I thank, first and foremost, the Division of the Humanities for covering my tuition and medical insurance, and for the annual and summer stipends. I also thank the Committee on Southern Asian Studies for several quarters of Dissertation Writing Support and Language Study Fellowships, as well as two Student Conference Travel Grants. I thank, finally, the Andrew W. Mellon Foundation for the Dissertation Write-up Fellowship. I am, at last, most thankful to those who were most affected by this long ordeal, my family. They remained loving and supportive throughout my many hours away from them, x physically and mentally. To Jasna (“oh well, maybe I will understand in the end just what it is that you are doing”) for being my soul mate for twenty-six years and counting, to Kalina (“so dad, what happens when you graduate, are you going to be like them philosophers, Plato?”) for turning my life around for the better and for drawing “Maṇḍana’s Fake Bird,” and to Angel for being an angel—I owe you everything. That birds figure so prominently in the dissertation is largely due to Birdie and Plato (named so by Kalina “so that he wouldn’t be dumb”). To my mother Rada and sister Gordana I am indebted in all kinds of ways, but most pertinently for remaining supportive despite having their family ripped away from them and cast some 5,000 miles to the West. And, to my father Dimitrija, soon to begin his twentieth year in Elysium, thank you for showing me how to be a decent human being. xi  INTRODUCTION “Where female parrots shut in cages at the door discuss intrinsic and extrinsic validity, debate whether action or the unborn Lord is the giver of results, and deliberate if the world is permanent or impermanent, know that place to be the abode of the learned Ma ḍana.”1 Liberation and the Highest Good There is an odd, curious textual structure in the commentary of Śa kara Bhagavatpāda (ca. 650- 800 C.E.) on the first chapter of the Taittir ya Upaniṣad. The commentary contains one of Śa kara’s most comprehensive accounts of the doctrine of liberation, mokṣa—we may for the time being understand it simply as freedom from the cycle of transmigration or embodiment— which is here explicitly called “the highest good,” paraṁ reyas. This account, however, is broken in two parts, one consisting of the full introduction to the commentary, and another one which is attached to a rather pedestrian gloss on the first chapter of the Upaniṣad. One almost gets the sense that Śa kara did not say all that he wanted to say on the topic of liberation and its attainment in the Introduction, and having performed duly his rather tedious commentarial duty on that part of the Upaniṣad whose topic would have hardly piqued his interest—cosmic relations between phonetics and ritual—he decides to conclude with an extemporaneous smash, as the “real Upaniṣadic deal” begins only in the second chapterŚ “We present now this deliberation for the purpose of distinguishing between knowledge and action: Is the highest good attained solely 1 svataḥ pramā aṁ parataḥ pramā aṁ kīrā ganā yatra giraṁ giranti | dvārastha-nīḍāntara-sanniruddhā jānīhi tan ma ḍana-pa ḍitaukaḥ || phala-pradaṁ karma phala-prado ‘jaḥ kīrā ganā yatra giraṁ giranti | dvārastha-nīḍāntara-sanniruddhā jānīhi tan ma ḍana-pa ḍitaukaḥ || jagad dhruvaṁ syāj jagad adhruvaṁ syāt kīrā ganā yatra giraṁ giranti | dvārastha-nīḍāntara-sanniruddhā jānīhi tan ma ḍana-pa ḍitaukaḥ. ŚDV 8.6.8. 1 through action; or, through action assisted by knowledge; or, rather, through knowledge and action togetherś or, through knowledge assisted by actionś or, through knowledge alone?”2 The two comments together, odd in their structure, can hardly be seen as odd in meaning. Looking carefully at Śa kara’s Upaniṣadic commentaries, one would inevitably notice that he generally begins all of them by discussing knowledge of Brahman or the Self as the means of liberation, which he commonly calls “the highest good,” not in isolation, but specifically contraposed to ritual action and its combination with “knowledge,” that is, meditation on Brahman. The introductions to his Upaniṣadic commentaries univocally announce that, to Śa kara’s mind, all that matters in the Upaniṣads, all that they are about, is liberation, the highest human good, and that it is necessary to clarify that the sole means of liberation is knowledge of Brahman. “Sole” is not an emphatic here nor a general negation, but specifically conveys that knowledge of Brahman is the means of liberation without the aid of ritual and meditation. What is, nevertheless, exceptional in this quirky introduction is Śa kara’s thoroughness on the topic of liberation, through which he does two things that I would like to emphasize here. First, he implicitly tells us what constitutes, to his mind and for his time, the pertinent and full scope of the discourse on liberation; to put it differently, he selects his interlocutors, and points out which accounts of liberation he finds deserving of attention and rebuttal. Second, through contrast with his interlocutors, he shows how his understanding of liberation is an absolute novelty for this pertinent scope at this point in history. Saving thorough analysis for chapters three through five, I now note that in the Introduction and Conclusion to the first chapter of Taittir ya-Bh ṣya, Śa kara presents and 2 atraitac cintyate vidyā-karma or vivekārtham – kiṁ karmabhya eva kevalebhyaḥ paraṁ śreyaḥ, uta vidyā- saṁvyapekṣebhyaḥ, āhosvid vidyā-karmabhyāṁ saṁhatābhyām, vidyāyā vā karmāpekṣāyāḥ, uta kevalāyā eva vidyāyā iti. TUBh 1.11.4, VI.46-7. 2 refutes several accounts of liberation, all of which concern the question of the combination of knowledge and action, and most of which can be directly associated with Vedic theologians who were his close predecessors or contemporaries: (1) liberation is just the state of being the Self, attained by the performance of ritual, a doctrine whose direct advocate was Kumārila Bha a and which was identified as such by Śa kara’s immediate student and commentator Sureśvaraś (2) liberation is “heaven,” svarga, a state of unexcelled felicity in the hereafter, attained by the performance of ritual, a doctrine that can be traced to the Mīmāṁsakas Śabara and Kumārilaś or alternatively by ritual aided by knowledge, which was in some way, shape or form the doctrine of various Vedāntins and Mīmāṁsakasś (3) liberation is attained by a “stream” of ritual and knowledge, that is, by the practice of continual meditation, a doctrine that was advocated by the proponents of the extremely influential prasaṅkhy na-v da (on which more later), prominent among whom were Bhart prapañca and Śa kara’s contemporary Ma ḍana Miśra; (4) liberation involved going to a different place in the hereafter, a doctrine found in many Upaniṣads, but, importantly, the doctrine of liberation of the Brahma-Sūtra itself. The accounts that Śa kara presents in the Taittir ya-Bh ṣya were, in fact, part of a single discourse, because their proponents had what I will later call a shared sphere of commitment. To anticipate briefly, these were all accounts of “Vedic theologians,” that is, of Mīmāṁsakas and Vedāntins who shared a very specific notional intersection that set them apart from their intellectual peers, namely the conviction that the Vedas were the sole authority on the questions of dharma—which we may for the time being translate fairly imprecisely as appropriate ritual and social behavior—and liberation; that, in other words, questions of liberation can be debated solely through recognizing scripture as a form of argument. To paraphrase one of the most orthodox among them, Kumārila Bha a: everything that pertains to dharma and liberation has its 3 origin in the Veda, and wherever it may be found, it must be adjudicated through the Veda.3 Indeed, a careful study would show that throughout his writings Śa kara generally does not debate liberation with Buddhists, Sā khyas, Naiyāyikas, and the like, for the simple reason that they do not belong to this shared sphere of commitment that accepts the Vedas not only as valid, but as conclusive argument. To his mind, the proper discourse on liberation was absolutely restricted to the Vedas. More generally, when Vedic theologians did debate others on liberation, as Kumārila did, for instance, debate Sā khyas, it was when these others made a “Vedic claim,” that is, tried to justify their accounts by an appeal to the Vedas. We should note not only the pertinent, but also the full scope of the discourse. Śa kara talks about liberation as the highest good, and the highest good in Vedic theology was a wider set of values that included not only liberation, but “heaven” or svarga as well. Modern scholars tend to see the two as strikingly different, indeed incommensurable values. Johannes Bronkhorst had in recent years, for instance, attacked the commonly accepted idea of an original unity of Mīmāṁsā and Vedānta through the assumption that from the very beginning Mīmāṁsā was all about heaven and none about liberation, whereas Vedānta was all about liberation and none about heaven.4 Nothing could be farther from the truth in Śa kara’s eyes: he read the Mīmāṁsā doctrine of heaven as a competing account of liberation that deserved due rebuttal. In fact, throughout the history of Mīmāṁsā and Vedānta, svarga and mokṣa were often defined in the same terms, nirati aya-pr tiṭsukhaṭ nanda, the highest happiness. The scope of the notion of the highest good spanned both unsurpassed happiness and total absence of suffering, and svarga and mokṣa were sides of a single coin: while everyone accepted the second, many were skeptical about the possibility of the first. 3 atra yāvad dharma-mokṣa-sambandhi tad veda-prabhavam. TV 1.3.2, I.166. 4 Bronkhorst 2007b. 4  The second thing that I want to extricate from the Taittir ya-Bh ṣya comment is that Śa kara in this discourse on liberation stood alone in claiming that liberation was achieved simply through knowledge of Brahman, knowledge qua knowledge, attained when the teachings of the Upaniṣads were fully understood. Such doctrine was a novel phenomenon in this shared sphere of commitment that I call Vedic theology. While already in the Brahma-Sūtra some Upaniṣadic texts were interpreted as deliberate fancy and classed under the rubric of symbolic meditations, but others as presenting factual ontological relations between Brahman on the one hand and the world and the Self on the other, there just wasn’t the notion that liberation could follow simply on understanding what those Upaniṣadic texts that present real ontological relations say. While such understanding was obviously necessary, it was just a prerequisite for proper meditation on Brahman, and it was not liberating knowledge. Even Śa kara’s contemporary and intellectual next-of-kin Ma ḍana Miśra affirmed that the propositional knowledge of the Upaniṣads had to be followed by meditation on Brahman, because mere intellectual understanding does not remove ignorance. So far as can be ascertained from the available textual evidence, all Vedāntins in Śa kara’s context would have held that meditation on Brahman that is facilitated by the Upaniṣads, and generally accompanied by Vedic ritual, was the characteristically Vedāntic soteriological practice. This is not to say that ideas about liberation that is a result just of intellectual understanding were not present in Śa kara’s wider intellectual context. Liberation was a “hot topic” of the day, and Śa kara’s doctrine contraposed to the rest of Vedic theology was, ultimately, derivative on the old divide between s ṅkhya and yoga, not the two philosophical schools that we associate with Īśvarak ṣ a and Patañjali, but the two general approaches to soteriology within the Hindu traditions. In his groundbreaking paper “The Meaning of Sānkhya 5 and Yoga,” Franklin Edgerton had shown that s ṅkhya and yoga in the early Indian history did not stand for any philosophical or metaphysical system, but for two distinct ways of conceptualizing salvation, s ṅkhya standing for the soteriological scheme which understands liberation as a result just of knowing a truth, knowing how things are, whereas yoga referring to the pursuit of liberation by some form of action or practice that is ultimately non-intellectual. This divide persisted in the philosophical Sā khya and Yoga, but Edgerton was right in concluding that Śa kara was as much an heir to s ṅkhya as was the S ṅkhya-K rik .5 In the Vedic theology of his time, however, the systematic exegesis of the Vedas through commonly accepted canons of interpretation, Śa kara was somewhat of a maverick in claiming that liberation was a result just of knowing BrahmanŚ whether or not the Mīmāṁsakas and Vedāntins who formed Śa kara’s intellectual context saw any value in knowing qua knowing, ultimately they all understood the pursuit of liberation as a form of yoga in which the soteriologically most significant element was non-intellectual. The first major argument of this dissertation, therefore, is that Śa kara’s interpretation of the Upaniṣads to the effect that liberation was a result just of knowing Brahman, knowing that is neither accompanied by ritual nor succeeded by meditation on Brahman, was a new thing in Vedic theology. By developing this argument, my intention partly is to illuminate one important question that had intrigued students of Advaita Vedānta and Indian philosophy. The question concerns the proper understanding of Śa kara’s significance in Indian intellectual history. The significance I have in mind is not that of the received Śa kara, the Śa kara of the hagiographies and the monasteriesś the royal Śa kara who rules India as a Śa karācārya from his seats at the four cardinal pointsś the Śiva born to join forces with Kumārila and banish Buddhism from Indiaś the 5 Edgerton 1924:34. 6 universalist Śa kara of Ram Mohan Roy, Swami Vivekananda, and Dr. Radhakrishnan; the Śa kara of the first feature film ever shot in Sanskrit; and that model Indian philosopher whose public image embodies and accommodates all things Sanskrit and, indeed, Hindu. I do not have in mind the received, but rather the historical Śa kara, although the first must have at least in some respect been derivative on the second. What I have in mind, then, is the significance of that Śa kara of whom, to use the image pained by Allen Thrasher, thinkers “as acute as Sureśvara,” his own immediate students, thought so highly as to consider themselves belonging to his school, rather than the school of some earlier teacher.6 The question I intend to illumine, then, concerns the significance of Śa kara in terms of some beginnings, some novelties, in his own intellectual context. What was he, really, about? The question of Śa kara’s significance has been asked by his most assiduous modern students. Sengaku Mayeda, for instance, proposed that Śa kara was not a “particularly original philosopher,” but rather a very bright commentator, as well as a “pre-eminent religious leader and a most successful religious teacher.”7 Mayeda was likely echoing Daniel Ingalls, who argued that what was original to Śa kara’s philosophy seems to have been the concept of Brahman without qualities; the other elements of his system were old, stated in various places, but the specific synthesis he made of them was “something quite new in the history of Indian philosophy.”8 For Allen Thrasher, on the other hand, there was no question that the philosophy of Śa kara’s Brahma-Sūtra-Bh ṣya was highly original, even if as a synthesis: the proper question of Śa kara’s significance was to ascertain whether this philosophy presupposed 6 Thrasher 1979:120. 7 Mayeda 2006b:6. 8 Ingalls 1952:12-3. 7 Ma ḍana Miśra’s Brahma-Siddhi, or it was, rather, the other way around. Whodunnit? Thrasher argued for the first, because, If the Brahma-Siddhi was written in ignorance of Śaṁkara’s works, and represents a current in pre-Śaṁkara Advaita, we must drastically reduce the usual estimate of Śaṁkara's originality. We must say that the commanding importance he has in the history of Vedānta, so great that all works before him except the Brahma-Sūtras themselves, the Gaudapāda-Kārikās, and possibly the BS, have disappeared, was due not to any newness in his ideas or his combination of ideas into a system, but to some other cause—perhaps to his zealous activity as an "evangelist" of Advaita Vedānta against the Buddhists, and the adherents of other systems within the Hindu fold, or to his activity in setting up maṭhas to carry on the tradition he followed.9 These accounts present a dichotomy of possibilitiesŚ Śa kara was either a great philosopher, even if not terribly original, or he was a religious leader and a teacher. I don’t doubt that he was most of this, although “pre-eminent religious leader” as an epithet for a man who was as elitist as they make them and who thought that there were very few deserving to be led directly by him sounds overly generous on the side of imagination. Be that as it may, seeing his major significance either as a philosopher or a religious teacher, or as both, disregards the overwhelming bulk of his writings, or what he himself was most concerned with. My argument here is that Śa kara’s significance for his own context was that of a theologian of liberation; that, in other words, what astute intellectuals like Sureśvara found so appealing about him as to see themselves as members of his school was a novel model of soteriological causality that he developed and its formidable defense. This novel model said that liberation the highest human good was attained solely by knowledge qua knowledge, and that the competent aspirant after liberation did not need to practice ritual and meditation, indeed, that he could not possibly practice them if he was really qualified. While, as I said, this was not a novelty in the wider intellectual context, it was very much so in Vedic theology. 9 Thrasher 1979:119. 8  In a sense, Daniel Ingalls had already recognized this, albeit implicitly, in his article “The Study of Śaṁkarācārya,” where his final word about Śa kara’s significance was that his novelty and original synthesis were “directed not so much against Buddhism, which is the traditional claim, as against the Mīmāṁsā and against schools of a more realistic Vedānta such as the Bhedābheda which flourished in Śaṁkara's time.”10 Ingalls still thought of contribution mostly in terms of metaphysics, ergo, philosophy, but he rightly identified Śa kara’s interlocutors. My study takes Ingalls’s insight and extends it over the highest good. The method through which I approach my study, on the other hand, on which more under the heading of “The History of Ideas,” turns illuminating this question into writing the history of the highest good in Vedic theology before and including Śa kara. Mah -V kya The principal objective of the dissertation, however, goes beyond this argument. When Śa kara says that liberation follows just by knowing Brahman, he thereby rejects not only meditation on Brahman, but also knowing Brahman as one would know any object. What he has in mind, rather, is knowledge of a fact or a state of affairs, specifically the fact of the Self—oneself, the agent of experience—being Brahman. Indeed, we will see that the doctrine of prasaṅkhy na or meditation on the non-dual Brahman argued that such meditation was necessary precisely because it approached Brahman as an object, the scriptural knowledge of which was deemed insufficient because it was mediate—it was knowledge by description or analysis, to use Henri Bergson’s phraseology—and argued instead that such knowledge must be followed by direct realization, or knowledge by intuition. While Śa kara agreed generally with the intuition claim, he argued that the requisite intuition could not pertain to Brahman as an object—indeed, such a 10 Ingalls 1952:13. 9 proposition was strictly impossible in a properly monistic ontology—but to a state of affairs that pertains to oneself, and, eo ipso, must be immediate, the knowledge of which was simply a matter of anamnesis, and would follow naturally once one had removed all possible identification points for the Self. Liberation, then, the highest good, followed not upon meditation on Brahman or by knowing Brahman, but upon knowing that oneself was Brahman. Liberation followed, in other words, upon fully understanding the Upaniṣadic identity statements, statements such as tat tvam asi, “You are that,” of the Ch ndogya Upaniṣad (6.8.7), and ahaṁ brahm smi, “I am Brahman,” of the Bṛhad- raṇyaka (1.4.10), that identify Brahman with the individual Self as the cognitive agent. These identity statements are commonly called mah -v kyas or “great statements” in scholarly literature, and the principal purpose of this dissertation is to investigate their origin as such, as mah -v kyas, and to understand just what was so “great” about them. Although mah -v kyas are a prominent feature of the conceptual range of most students of Hinduism, Vedānta, and Indian philosophy, there is generally no scholarly account that attempts to trace their origin or properly understand their meaning. In terms of what mah -v kya may mean, the scholarly use of the lexeme can be broken down into several varieties that represent a progressively shrinking scope. In Chapter Ten I will present this taxonomy in more detail, so it is sufficient here to simply lay it down and illustrate briefly. There is, first, a tendency to label any short and in some sense important statement from the wider range of the Hindu canon, including books such as the Bhagavad-G t , as a mah - v kya. Thus, Richard H. Davis in his “The Bhagavad Gita: A Biography,” says that “Indian commentators often highlighted especially powerful statements in the Gita for special attention 10 as mahavakyas (great utterances).”11 The second practice is identical with the first, except that it delimits the scriptural scope to the Upaniṣads: any short and important statement in the Upaniṣads seems liable to be called a mah -v kya. Thus, for instance, Chris Bartley calls the famous Taittir ya 2.1.1 statement that defines Brahman, satyaṁ jñ nam anantaṁ brahma, a mah -v kya. As we will see in Chapter Ten, this was one of the statements that were explicitly not mah -v kyas for Advaita Vedāntins.12 These two scholarly practices are, with some notable exceptions, unjustified, insofar as they do not reflect any indigenous theory about, and use of, the term, but seem to follow a simple, if uninformed, logic: if it is short and important, call it a mah -v kya. In this use, mah - v kya shares much of the semantic range of the English word “mantra,” which is at times defined as “a word or phrase that is repeated often and that expresses someone’s basic belief.”13 Such mah -v kyas do not concern us here. The third practice properly identifies the mah -v kyas with the Upaniṣadic identity statements in general, while the fourth is more specific, insofar as it concerns four such identity statements, associated with the four Vedas, the four monasteries allegedly established by Śa kara in the four cardinal points, and his four principal students. These four mah -v kyas, with their corresponding Veda, sacred place in a cardinal direction and Śa kara’s student, are: (1) prajñ naṁ brahma, “Consciousness is Brahman,” in AiU 3.3 of the Ṛgveda → Puri → Padmapādaś (2) ahaṁ brahm smi, “I am Brahman,” in BĀU 1.4.10 of the Yajurveda → Shringeri → Sureśvaraś (3) tat tvam asi, “You are that,” in the ChU 6.8.7 of the Sāmaveda → Dwaraka → Hastāmalakaś (4) ayam tm brahma, “This Self is Brahman,” in MāU 2 of the Atharvaveda → Badrinath → Tro aka. 11 Davis 2015: 99. 12 Bartley 1986:103, 105. 13 The definition is taken from Merriam-Webster’s Dictionary. I am thankful to Gary Tubb for pointing out this similarity of use to me. 11 We will see in Chapter Ten that the doctrine of four mah -v kyas was associated with monastic Advaita Vedānta, that it was relatively a latecomer, and that as a subject of inquiry it properly belongs to religious history rather than the history of ideas. As for the origin of the Upaniṣadic identity statements being called mah -v kyas, the predominant scholarly practice either does not ask the question, or simply assumes that it was Śa kara himself who began such practice. This will also be evident from the review in Chapter Ten, but here we may quote Andrew Nicholson’s statement as fairly commonplaceŚ “For instance, the eighth-century Advaita Vedāntin Śa kara dubbed four Upaniṣadic sentences as ‘great statements’ (mah v kyas)Ś ‘You are that’ (tat tvam asi), ‘I am Brahman’ (ahaṁ brahm smi), ‘This self is Brahman’ (ayam tm brahma), and ‘Brahman is consciousness (prajñ naṁ brahma).”14 Śa kara is credited not only with introducing the practice of calling the Upaniṣadic identity statements mah -v kya, but with the singling out of four of them as well. All things considered, very little is known about the Upaniṣadic identity statements in their being mah -v kyas, “great statements,” and particularly unclear is what is so “great” about them. The most “extensive” dedicated study available is K. Satchidananda Murty’s Revelation and Reason in Advaita Ved nta, which has a ten-page chapter titled “Interpretation of Mah v kyas,” a distillation that attempts to explain the details of the philosophy of language that are involved in the interpretation of the identity statements, but neither pursues the origin of the idea, nor attempts to understand its nuts and bolts.15 What is even less widely known16 is the existence of what appears to have been another, different idea of mah -v kya in Indian intellectual history, one that developed in the school of 14 Nicholson 2010:41. 15 Murty 1959:88-98. 16 And was so particularly before the groundbreaking work of Larry McCrea (2008). 12 Mīmāṁsā and was adopted as a term of art in the tradition of Sanskrit aesthetics that was influenced by Mīmāṁsā. In this use, mah -v kya literally meant a “great” statement, a long statement or sentence, an element of language whose range included anything from what we call a “paragraph” to a full book. All the three great pre-Śa kara Mīmāṁsakas knew about this idea, and while we will unravel its details later, it may be now worth our while for the purpose of contrast to look briefly at a relatively late definition of such mah -v kyas in the S hitya- Darpaṇa of the 14th-century aesthetician Viśvanātha Kavirāja. A “great sentence” is a collection of sentences that are mutually related through fitness (yogyat ), syntactic expectancy or cohesion ( k ṅkṣ ), and proximity ( satti), in other words, a text. Viśvanātha’s examples of such “great sentences” are tellingŚ the R m yaṇa, the Mah bh rata, and Kālidāsa’s Raghu-Vaṁ a.17 To put things in perspective, now, whereas the Advaita Vedānta mah -v kyas were sentences of two or three words, the theory of literary criticism that was influenced by Mīmāṁsā classed under mah -v kya one of the longest epic poems in world literature. The Vaiṣ ava Vedāntin Jīva Gosvāmin went even further in claiming that the full Vedic corpus, which in his reckoning included the itih sa-pur ṇa literature, was one grand mah -v kya.18 My argument with regard to mah -v kya is twofold, historical and conceptual. First, an explicit Advaita Vedānta notion of mah -v kya was not developed by Śa kara. Śa kara, in fact, used the lexeme only once in relation to the Upaniṣads, in his little-studied commentary on the Aitareya, and while it is clearly related to the knowledge that characteristically the Upaniṣadic identity statements provide—the understanding of oneself, the cognitive agent, being Brahman— and is used in the context of teaching, there is no explicit, underlying theory as to what constitutes a “great statement.” Śa kara otherwise did not have a preferred term for the identity 17 SD 2.1-2. 18 De 1961. 13 statements, and he tended to designate them metonymically, “tat tvam asi and the rest.” Śa kara’s prominent students Sureśvara and Padmapāda did not talk about mah -v kyas either, though one would have expected the first to do that in his Naiṣkarmya-Siddhi, which is a treatise solely concerned with the meaning of the Upaniṣadic identity statements. An explicit theory of mah -v kya, of that feature in virtue of which the Upaniṣadic identity statements were “great,” is for the first time laid down by the 10th-century Advaitin Sarvajñātman. And while it remains entirely possible, or perhaps even probable, that there was a notion of mah -v kya in monastic Advaita Vedānta before Sarvajñātman—in fact, with Sarvajñātman’s conceptual apparatus, one can fully reconstruct a mah -v kya doctrine in Śa kara’s own works, as we shall see eventually—it is Sarvajñātman who provides that threshold or transformation of concept that excites so profoundly historians of ideas. A transformation though it was, I will argue second, Sarvajñātman’s theory of mah - v kya was directly and quite explicitly modeled on the Mīmāṁsā blueprint of mah -v kyas as “long sentences.” What was great about the Upaniṣadic identity statements was that only through them the Upaniṣads as the jñ na-k ṇḍa section of the Veda, concerned solely with Brahman, could be read as a coherent, single, corpus. The proper and full understanding of tat tvam asi and ahaṁ brahm smi, on which was predicated the attainment of liberation the highest good, required the understanding of the two juxtaposed categories, Brahman and the Self, which were ellipses that were fully defined in various Upaniṣadic texts, and separately explained further in individual passages. The identity statements stood at the top of a textual hierarchy that was linked through relations of tight cohesion, a texture, that obtained finality of meaning solely through them, and would unravel without them. The Upaniṣadic identity statements, in other words, were only formally short, and could not be read and understood without reading and 14 understanding the Upaniṣadic corpus, which corpus, on its part, would be merely a sum of discontinuous passages without them at the top. Tat tvam asi and ahaṁ brahm smi were mah - v kyas, “great statements,” because they were the Upaniṣads. Characters, Dates, and Archive The network of this study through which several Vedāntins and Mīmāṁsakas intersect is Śa kara Bhagavatpāda. Like most dates in Indian history, we do not know when he was born or when he died, and all that can be said with some degree of certainty is that he lived sometime between 650 and 800 C.E.19 The precise dates are not, however, as important for the kind of intellectual history that I want to do here as is relative chronology. That is, it does not matter that much just when Śa kara was born and when he died, as much as it matters which philosophers and Vedic theologians were a significant part of his intellectual context, and which were positively and negatively influenced by him. What is known from the perspective of relative chronology, thus, is that Śa kara lived sometime between 650 and 800 C.E., but after Śabara, 19 The scholarship on Śa kara’s dates is vast, and good overviews of the major arguments are available in Nakamura (1983:48-89), Pande (1994:41-54), and Harimoto (2006). Several methods have been used for dating Śa kara. One is the explicit attribution of dates to him in later worksŚ in fact, the most common dates of Śa kara’s birth and death that we find in scholarly literature were proposed based on a short manuscript of an unknown title, which says that Śa kara was born in the year 710 of the Śaka era and died in 742, which is equivalent to 788 and 820 C.E. Several other works repeat these dates, none of which, however, is earlier than the 16th century. Another method is based on pursuing what is known from other sources about the flourishing of cities that Śa kara mentions in his works. Hajime Nakamura, however, had shown that in referring to names of specific cities, Śa kara was just following a customary practice, and that his referring to such places was not related to their contemporary significance (1983:59- 62). Yet another dating approach is through the attempt to locate historically three kings that Śa kara mentions in his commentary on the Brahma-Sūtra 4.3.5—Balavarman, Jayasiṁha, and K ṣ agupta—under the assumption that they were his contemporaries. From these names, and based on South Indian political history, Kengo Harimoto had proposed that Śa kara likely wrote his Brahma-Sūtra commentary sometime between 756 and 772 C.E. (Harimoto 2006). Finally, there is the method of relative chronology, which places him between philosophers and theologians whose dates are better known, and who were roughly contemporaneous with him. This method is still very imprecise and places him anywhere from 650 to 800 C.E. (Harimoto 2006:87-93) but for our purposes it is the only method relevant. 15 Bhart prapañca, Kumārila, and Prabhākara, and before his own students Sureśvara and Padmapāda, as well as Sarvajñātman.20 The only unclear detail relevant for this study in terms of intellectual history is Śa kara’s precise relationship with the other great Advaitin of his time, Ma ḍana Miśra, and the best available evidence suggests that the two were contemporaries, but that Ma ḍana’s Brahma- Siddhi presupposed Śa kara’s Brahma-Sūtra commentary.21 Given that Sarvajñātman can be dated to the second half of the tenth century C.E.22, the study examines scripture and the highest good in Vedic theology beginning with the canonical theological texts, the M m ṁs -Sūtra and the Brahma-Sūtra, and up to the end of the tenth and the beginning of the eleventh century C.E. The primary archive of the study consists ofŚ (1) Śa kara’s authentic works, that is, his independent treatise Upade a-S hasr and the commentaries on the Brahma-Sūtra, the Bhagavad-G t , and the principal Upaniṣads (Bṛhad- raṇyaka, Taittir ya, Ch ndogya, Aitareya, , Kaṭha, Muṇḍaka, Pra na, and two on Kena);23 (2) The works of the three great pre-Śa kara Mīmāṁsakas, that is, Śabara’s Bh ṣya on the M m ṁs -Sūtra, Kumārila’s loka-V rttika and Tantra-V rttika, and Prabhākara’s Bṛhat ; (3) Ma ḍana Miśra’s Brahma-Siddhi; (4) The works 20 On the dates of the three Mīmāṁsakas, see Kataoka 2011b:20-24, who gives 500-560 C.E. for Śabara, 600-650 C.E. for Kumārila, and 620-680 for Prabhākara. Krasser 2012 quite convincingly suggests earlier dates, the middle of the sixth century for Kumārila and, by implication, earlier for Śabara. Bhart prapañca’s date of 550 C.E. suggested by Nakamura (2004:131) is as good as any pre-Śa kara date. Padmapāda and Sureśvara, being Śa kara’s students, are his younger contemporaries, and their dates are tied to him. This study does not address the long- debated question of the identity of Sureśvara with Ma ḍana MiśraŚ as much as the present author would fancy this identity to have been real, the assumption here is that the two were different persons. 21 Thrasher 1979. 22 See Eswaran Nampoothiry’s Introduction to the Pram ṇa-Lakṣaṇa, p. ix-xxiii; Kocmarek 1985:7-11; Potter 435- 6. 23 The issue of the authenticity of Śa kara’s works has been a very productive question in Sanskrit studies, and the above list reflects the good work of Paul Hacker (1995:41-56), Daniel Ingalls (1952), and Sengaku Mayeda (1965a; 1965b; 1967). The authenticity of two commentaries is still uncertain, on the P tañjala-Yoga- stra (the Yoga-Sūtra with the commentary which is commonly attributed to Vyāsa), and on the gama- stra of Gauḍapāda that includes the M ṇḍukya Upaniṣad; the two are not consulted here. Although relatively recently arguments have been made in favor of the authenticity of the treatises Viveka-Cūḍamaṇi (Grimes 2004) and Pañc karaṇam (Sundaresan 2002), the two contain common Advaita concepts and expressions that are absent in Śa kara’s authentic works and, eo ipso, later. 16 of Śa kara’s immediate students, Padmapāda’s Pañca-P dik and Sureśvara’s Naiṣkarmya- Siddhi, Bṛhad- raṇyaka-Upaniṣad-Bh ṣya-V rttika, and Taittir ya-Upaniṣad-Bh ṣya-V rttika; and (5) Sarvajñātman’s Saṅkṣepa- r raka and Pañca-Prakriy . For Chapter Four dedicated to liberation in the Brahma-Sūtra, I will rely on several post-Śa kara commentaries insofar as they are useful for reconstructing the Brahma-Sūtra doctrine. For Chapter Five that deals with the doctrine of prasaṅkhy na meditation and Bhart prapañca’s soteriology, I will refer to sections of Ānandagiri’s commentaries on Śa kara’s Bṛhad- raṇyaka-Upaniṣad-Bh ṣya and Sureśvara’s Bṛhad- raṇyaka-Upaniṣad-Bh ṣya-V rttika. More on both in due course. Vedic Theology Throughout the dissertation I use the lexeme “Vedic theology,” and I speak about “Vedic theologians.” Lest the purpose of my project be misunderstood, under this and the next two headings I should like to clarify why I choose to talk about “theology,” what I mean under the term, and what kind of a project I am doing. There has been a growing trend in Hindu24 studies in recent years to make theology a legitimate field of inquiry, or simply to use the term without elaborate justifications, that involves two strands that should be untangled.25 On the one hand, 24 I use the contested terms “Hindu” and “Hinduism” in regard to the premodern period without a sense of need for a lengthy justification. David Lorenzen’s essay “Who Invented Hinduism?” (2006Ś1-36) should, for the intellectually honest reader, end the debate about the European construction of “Hinduism,” or about any construction, for that matter, that seeks to locate the origins of Hinduism, both the name and the game, to use Wendy Doniger’s witty turn of phrase (2014:3), in the colonial encounter. Particularly instructive is Lorenzen’s location of the debate in “the tendency of many historians of modern India—especially those associated with the subaltern school—to adopt a postcolonialist perspective that privileges the British colonial period as the period in which almost all the major institutions of Indian society and politics were invented or constructed.” (p.36) 25 The sustained discourse on “Hindu theology” seems to originate with Catholic theologians working in India, particularly Richard de Smet (1916-1997, on whose life see Malkovsky 2000a), who had, for better or worse, influenced many modern students of Śa kara and Vedānta. The recognition of the theological nature of Vedānta seems to have become mainstream in academia with studies about Rāmānuja rather than Śa kara, for instance van Buitenen’s edition and translation of Rāmānuja’s Ved rtha-Saṅgraha (1956), John Carman’s model study The Theology of R m nuja (1974), Eric J. Lott’s God and the Universe in the Ved ntic Theology of R m nuja (1976), and Julius Lipner’s The Face of Truth (1986). However, it is the work of Francis Clooney, with which I will engage more in this heading, that is largely responsible for Hindu theology becoming both a common mode of discourse and 17 there is a claim that understanding many traditions and individuals in the Hindu context is facilitated by treating them as theological and as theologians, respectively. This is commonly accompanied by an attempt to show that theology in such cases fits the bill better than philosophy; in other words, that many influential Hindu intellectuals, “such as Kumārilabha a, Śa kara, Rāmānuja, Madhva, Abhinava, or Vijñānabhikṣu,” to refer to Jonathan Edelmann’s list, were not philosophers and that it is misleading to call them that, given how the terms “philosophy” and “philosophers” are presently used.26 The justification for the claim is simple: philosophy is an inquiry that does not assume the authority of religious texts, whereas theology, although itself a reasoned inquiry, accepts the epistemic validity of texts and tradition; theology is scriptural interpretation, philosophy is not. Śa kara, Rāmānuja and their intellectual collocutors mostly argue about the correct meaning of texts, and use texts as arguments; ergo, they are not philosophers in the present sense of the term. Francis Clooney traces the discourse that treats the likes of Śa kara and Rāmānuja as philosophers to the colonial encounter and the position of subordination of theology relative to science and philosophy in post-enlightenment Europe: there was a need on the part of Indian intellectuals to present Indian thought in general as philosophical, thus, “respectable,” and avoid the stigma of dogmatism that theology carried with it. Today, “the views of theology in relation to science and philosophy are more nuanced and less heated,” and to continue avoiding Hindu theology is to continue reading “a problem indigenous to European history into an Indian context.”27 itself an object of deliberation in contemporary academia. Important engagements with Hindu theology as a discipline include Clooney (2003), Edelmann (2013), and Okita (2014). 26 Edelman 2013:430. 27 Clooney 2003:448-9. 18  On the other hand, there is a plea for “Hindu theology” as a way of doing theology today, commonly described by the proponents as “faith seeking understanding” and as “soteriological transformation.” Francis Clooney is again useful for articulating the presuppositions of such understanding of theology: it is an inquiry which is carried on by believers who allow their belief to remain an explicit and influential factor in their research, analysis and writing. Believing theologians are (usually) members of believing communities, and have those communities as their primary audiences … With their communities, they believe in some transcendent (perhaps supernatural) reality, the possibility of and (usually fact of) a normative revelation, and in the need to make practical decisions and life choices which have a bearing on salvation. Theologians do their work with an awareness of and concern for these beliefs, and with a desire to defend and preserve them, even if at one or another moment they may have to question, recontextualize and finally reformulate them in modes of discourse quite different from those already familiar to the community.28 Anantanad Rambachan describes such an undertaking as one which “articulates a personal interpretation and understanding of the tradition,” and Jonathan Edelmann says that modern Hindu theologians in the Euro-American academic context are those “for whom the academic study of Hinduism is part of their personal and religious development, and who believe their articulations of Hinduism are answerable to the academic community, as well as the specific Hindu communities and traditions of which they are part.”29 Such theology may also be practiced by scholars with commitments in one tradition who try to understand another, thus making themselves liable to a tension between “vulnerability to truth” found in the other tradition, and “loyalty to truth” of one's own tradition. This is the plea for “comparative theology.”30 There are, thus, several (non-exclusive) senses in which “Hindu theology” may be understood or practiced: an academic undertaking towards understanding, without commitments, that finds the categories of theology useful (and more so than other categories, particularly those 28 Clooney 1993:4. 29 Rambachan 2006:4; Edelmann 2013:428. 30 Clooney 1993:5-6. 19 of philosophy) for describing an intellectual, a tradition, a notion or practice, etc.; an undertaking towards understanding, with commitments, which may or may not be situated within the academic context. Edelmann and Kiyokazu Okita call the two “first- and second- order theology,” respectively.31 Both may be done by scholars or practitioners from within or without the tradition, or from within another tradition, with or without explicit comparative aspirations, if the field of inquiry is specifically theological. Under the heading on “The History of Ideas,” I will articulate why my project is not an essay in theology, although its main characters are throughout called “theologians.” Here I want to clarify in what sense I use “theology,” what method may be employed in delimiting the scope of “Hindu theology” relative to Indian philosophy, and why Vedic theology (which may be taken as a subset of Hindu) rather than Indian philosophy. Very briefly, my claim is that “Hindu theology” is a welcome intervention, but that the binaries theology or philosophy and theologian or philosopher are less helpful. I suggest, rather, that it is more useful to do two things: first, consider carefully modes of discourse and kinds of arguments that an individual or a tradition may develop and put forward; second, related to the first, be mindful of what I call the shared sphere of commitment, constituted by the context or scope in which the different kinds of arguments can be advanced in virtue of shared presuppositions. I do not discuss Hindu theology as a contemporary practice, because that serves little purpose for my project. Resisting essential definitions, there are several senses in which one may talk about Hindu theology as a mode of discourse, revolving around questions of subject and method. One is that of exegesis. Now, it is said that “theology” is obliquely applicable even in regard to religions or systems which are not focused on God (theos), but have a transcendent point of 31 Edelamann 2013; Okita 2014. 20 reference, such as Buddhism.32 Taking the “transcendent point of reference” in the Hindu context to stand for ontological reals (including causal relations) that are knowable solely from linguistic utterances ( abda), theology as exegesis is about ascertaining and understanding the explicit or implicit meaning of texts that concern such transcendent points of reference, by employing recognized canons of interpretation. Exegesis is different from general interpretation, not necessarily methodologically, but in terms of the character of its subject: although both may use the same tools, exegesis is concerned with ontological reals that are supersensible. As we know from Larry McCrea’s study of poetics in Medieval Kashmir, the Mīmāṁsā canons of sentence interpretation were thoroughly appropriated by Sanskrit alaṅk ra- stra; in other words, Mīmāṁsā and alaṅk ra- stra shared the methodology of interpretation.33 However, Mīmāṁsā was theology (that is, exegesis) in virtue of the character of its subject, real causal relations that are empirically unknowable, whereas literary criticism was not. Let us further clarify this through comparing the two Hindu traditions that are typically represented as atheistic, Mīmāṁsā and Sā khya, and see why the first qualifies as theology whereas the second doesn’t. Both Mīmāṁsā and Sā khya delineated their unique scope as consisting of that which is supersensible. The Mīmāṁsaka Śabara famously said in his Bh ṣya that dharma concerns things past, present and future that are minute, hidden, or remote, thus supersensible.34 The S ṅkhya-K rik likewise claimed that there are things which are real yet not immediately knowable, for a variety of reasons, such as excessive distance or proximity, sense impairment, inattention, minuteness, obstruction, covering, or mixture with other things in the 32 See, for instance, Heinrich von Stietencron’s entry on theology in the Brill Dictionary of Religion (2006Ś1879- 1883). 33 McCrea 2008. 34 MSŚBh 1.1.2. 21 same category.35 Mīmāṁsā set its scope of reals, specifically causal relations, as those which are knowable solely from the Veda (codan ), that is, are not traceable inferentially, but must be discerned “from hearing.” Sā khya, on the other hand, accepting that there are things which are knowable solely from scripture,36 claimed that its characteristic objects of inquiry, the “non- manifest” (avyakta) or prime matter (mūla-prakṛti), and the Self (puruṣa), were both knowable from inference that proceeds through analogical reasoning (s m nyato-dṛṣṭa), and did not bother with scriptural objects at all.37 Thus, while both Mīmāṁsā and Sā khya posit a thing with a “transcendent point of reference,” the scope of the first was theological because of being knowable solely from linguistic utterances, whereas the scope of the second was philosophical, or properly metaphysical. This distinction, however, obtains not through the content or the subject of the inquiry—in both cases it is a supersensible reality—but through the method: the first proceeds through interpretation of linguistic utterances, while the second through analogical reasoning that pursues causal relations and inferred reals from data that is empirically knowable. We can, thus, state the general principle that theology is exegesis, interpretation of texts about ontological reals, of whatever kind, if they involve a transcendent point of reference, that is, if they are understood as supersensible. This is an important restriction that is not commonly made, yet is required in order to properly circumscribe the field of Hindu theology. Clooney, for 35 SK 7. 36 SK 6Ś “That which is not immediately knowable and is not established even by that [analogical inference] is established through valid testimonyś” tasmād api cāsiddhaṁ parokṣam āptāgamāt siddham. Gauḍapāda’s instances of such thingsŚ “Indra is the king of the godsś there is the land of the Northern Kurusś there are nymphs in heavenś” tasmād api cāsiddhaṁ parokṣam āptāgamāt siddhaṁ yathendro deva-rājaḥ uttarāḥ kuravaḥ svarge 'psarasa iti parokṣam āpta-vacanāt siddham. 37 “The knowledge of supersensible things is got through inference from analogical reasoningś” sāmānyatas tu d ṣ ād atīndriyā āṁ prasiddhir anumānātś SK 6. Gauḍapāda thereonŚ “Matter and the Self, being supersensible, are established through inference from analogical reasoningś” pradhāna-puruṣāv atīndriyau sāmānyato d ṣ enānumānena sādhyete. 22 instance, argues that Advaita Vedānta is theology and not philosophy because it involves an extension of the Mīmāṁsā canons of interpretation over the Upaniṣads, but just from that it would not be clear how, for instance, Indian jurisprudence, which also developed as an extension of the same Mīmāṁsā principles, is not theological.38 While Clooney proceeds to characterize Advaita Vedānta as theology in the sense of “faith seeking understanding,” there is this more basic sense in which all forms of Vedānta, as well as the textual religious traditions of Vaiṣ avism and Śaivism, most of which were associated with Vedānta, had something in common with Mīmāṁsā: the commitment to a transcendent point of reference. We may call this exegetical or scriptural theology, a mode of discourse that is concerned with truths about reals knowable solely from linguistic utterances, whose understanding requires interpretation, and which uses characteristically theological arguments. While it is the method that is determinative here, the restriction of the subject is important, because without it the method would not amount to theology. Let me now give an instance of what I call “characteristically theological arguments,” drawn from a dispute between Bādarāya a and Jaimini as a paradigmatic Vedāntin and Mīmāṁsaka, respectively. In the fourth section of the third chapter of the Brahma-Sūtra, there is a long discussion in which Bādarāya a puts forward the claim that the Upaniṣadic meditations are the means of human good independent of Vedic sacrifices, and goes on to consider several Mīmāṁsā objections. The knockdown argument is given in sūtra 15: there is a Vedic text in the Bṛhad- raṇyaka from which it is evident that some Vedic folks do not marry, and thus cannot perform ritual because they have not lit the sacrificial fire that is presupposed on marrying, yet pursue the Self, that is, perform Upaniṣadic meditations; such could not be the case if meditation 38 Clooney 1993. 23 was supererogatory on ritual. Further, the practice of lifelong celibacy is justified in the Ch ndogya, which is an important point in the context because both Jaimini and Bādarāya a are traditionally represented as Sāma-Vedins and, thus, custodians of the Ch ndogya. Jaimini retorts: the Ch ndogya mention of lifelong ascetics is just that, a mere mention that recognizes the fact of there being such poor fellas, but not an injunction that justifies what they are doing. There is another text, in fact, which condemns such lifelong celibacy. Bādarāya a finally concludes: it is not just a mention, for two reasons: first, because there is “a direct statement of sameness,” that is, the lifelong celibate is listed along with the householder such that nothing really separates the two as good and bad; and second, there are cases of precedent in which Vedic existential statements are read as injunctions. The argument illustrates perfectly well the theological mode of discourse that involves a shared sphere of commitment constituted by the acceptance of Vedic statements as reliable epistemic warrants, of recognized canons of interpretation, and of categories such as “injunction,” “mention,” “direct statement,” “Vedic precedent,” etc. Vedānta of all walks was chock-full of them. Such arguments were possible and made sense because of this shared sphere of commitment, and would be useless against Sā khya, which had a technical term for the Vedic variety of bondage, d kṣiṇaka or the bondage respective to honoraria that one pays to Vedic priests,39 and against the Buddhists and Jains, who did not accept the Veda as a valid pram ṇa for any domain. Taking the subject of theology not in the oblique sense of having a transcendent point of reference, but properly God, one can define Hindu theology as a discourse about such ideas as Brahman, Īśvara, Antaryāmin, Paramātman, Bhagavān, etc. There is, still, an important 39 See SK 44 and commentaries. 24 distinction that must be drawn. While in virtue of the subject any kind of inquiry that is concerned with these notions and whatever is related to them—for instance, an appropriate aesthetics, cosmology, psychology, religious practice, a doctrine of the highest good—is theology, “God-talk,” this subject that is theos has historically in the West been a subject not only of theology, but of philosophy as well. That is, insofar as God was understood as the first principle, it was part of the discipline of metaphysics, quite independently from religious considerations or commitments. Such was the case from early on, in Aristotle’s determination of metaphysics as the inquiry into the first principles of reality, more specifically the supersensible substances, and most specifically God the first mover and pure actuality, which keeps the world rolling as its final cause but is itself not liable to change, and is essentially non-transitive consciousness.40 In the Catholic tradition, the inquiry into God as the subject of philosophy came to be called “natural theology,” a properly philosophical discipline that takes its data from the world and pursues factual and possible causal relations through the light of natural reason, in the hope of arriving at the first principle. To quote Frederic Copleston on St. Thomas’ understanding of the domain of the two disciplines, “the fundamental difference between theology and philosophy does not lie in the difference of objects concretely considered.”41 The prominent philosopher of German Enlightenment Christian Wolff characterized this natural theology as “special metaphysics,” that is, the terminus of the metaphysical inquiry into Being in the most general sense of any possible thing. Wolff set three purposes for such natural theology: (1) to prove the existence of God; (2) to ascertain the essential attributes of God; and (3) to determine the things 40 See Copleston 1993a:287-319, particularly 314-318. 41 Copleston 1993b:313. 25 that are possible given God's essential attributes.42 This is still the core of the study of philosophy of religion. How may one go about distinguishing scriptural from philosophical theology, and from philosophy more generally, in the Hindu context? I suggest that we need to look again at the mode of discourse and the shared sphere of commitment. A line is drawn sometimes between Nyāya as rational theology, thus, properly philosophy, and Vedānta as “revealed” theology: indeed, the Naiyāyikas developed inferential proofs for the existence of God, whereas Vedāntins stuck to their guns and claimed that Brahman was knowable solely from the Upaniṣads. It is more instructive, however, to look at Vedānta and Sā khya, since the second was not theological either through the subject or through the method (though, perhaps, it was so in origin): both traditions were primarily and originally concerned with first principles, the proper domain of metaphysics, whereas the philosophical theology of Nyāya did not really develop before the Buddhist challenge, and was, thus, more of an afterthought. Now, along with the concern with first principles, Sā khya and Vedānta also shared the general theory of causality, sat-k rya-v da, the doctrine that the effect was not a new thing, but a transformation of the cause. They parted ways on two questions: first, the material cause for Sā khya was prime matter, whereas for Vedānta it was Brahman, and the efficient cause for Sā khya was the proximity of prime matter with the Self, whereas in Vedānta it was still Brahmanś second, the first principles of Sā khya were knowable through inference, as we saw above, whereas Brahman was knowable from the Upaniṣads. Thus, Sā khya and Vedānta were at odds in terms of the pram ṇa appropriate for knowing the first principles, as well as the specifics of such first principles. Finally, Sā khya was historically Vedānta’s main rival at the time of its 42 Hettche 2016. 26 codification in the Brahma-Sūtra (BS), and the bulk of the first two chapters of the BS consists of clarification of Vedāntic doctrines primarily in view of a Sā khya challenge. Let us now pay some attention to the anti-Sā khya arguments in the BS and see what kinds of reasoning they involve. The BS opens with the definition of Brahman, “It is that from which come origination, etc.,” and immediately proceeds to affirm Brahman’s essential characteristics, consciousness and bliss, not directly, but through distinguishing Brahman from the first principles of Sā khya, prime matter and the Self. The arguments are theological: Brahman is known from scripture,43 and a proposed first principle will not fit the bill if it does not have the characteristics of Brahman that are known from the Upaniṣads. Prime matter cannot be the first principle, because it does not pass this scriptural test. That is, in the Upaniṣads, specifically the beginning of the 6th chapter of the Ch ndogya, it is said that the first principle which is Being, sat, reflected or visualized before creating the worldŚ “And it thought to itselfŚ ‘Let me become many. Let me propagate myself.’” The prime matter of Sā khya is an insentient principle, it cannot reflect, and so it fails the scriptural test: it is a abdam.44 Similar is the case with Brahman as bliss in sūtra 1.1.13, “That which is bliss abundant is Brahman, because of repetition.” This is a denial that the Sā khyan individual Self is the reference of the Taittir ya text that describes five successively higher layers of personhood, culminating with the Self of bliss, because the Upaniṣad proceeds to repeat “bliss” explicitly in association with Brahman throughout its 2nd and 3rd chapters, and says how Brahman “gladdens” the individual Self. Thus, the argument is not that the first principle must be conscious and blissful because causality demands that, but that Brahman is presented as 43 BS 1.1.3. 44 BS 1.1.5. 27 consciousness and bliss in the Upaniṣads, for which reason prime matter and the Self of Sā khya cannot be the reference of “Brahman.” Such arguments are what I consider characteristically theological arguments, driving home a point by an appeal to scripture, although the subject in this case—the first principle—was a proper subject of philosophy, that is, metaphysics. This specific argument was possible because Sā khya had a stake in the Brahman-talk, unlike other metaphysics, through the presence of its first principles in the scriptural corpus: prakṛti and puruṣa are common in the later Upaniṣads, and even more so in the smṛti literature. There was, in other words, a shared sphere of commitment between the two traditions in the “Vedic communityŚ” as Śa kara says, the good Vedic folks or iṣṭas accept many Sā khyan principles, and therefore it becomes imperative to state, for this community which obeys the force of scripture as arguments, just what in Sā khya is not acceptable, and if found in scripture, requires interpretation.45 But Sā khya was also an independent school with which Vedānta shared, as we just stated, the sat-k rya-v da. It is in this context that we see characteristically philosophical reasoning on the part of Bādarāya a. This reasoning is still theologically constrained, and expressly so: reasoning is inconclusive, and one never gets to avoid all undesirable consequences of a causal theory solely through reasoning.46 Yet, Bādarāya a goes on to engage precisely in such reasoning, with the general claims that Brahman fits best the requirements of a first principle in virtue of its characteristics,47 that the competing first principles make little sense under our common understanding of causality, and that, when no satisfying arguments from reason are forthcoming, the Sā khya notion of causality faces the same objections as Brahman.48 45 BSBh 2.1.3. 46 BS 2.1.11-12, 2.1.26. 47 BS 2.1.35. 48 BS 2.1.10, 2.1.28. 28  To illustrate, Bādarāya a takes exception to the Sā khya sat-k rya-v da claim that the effect must share the characteristics of the cause. The Sā khya opponent of the BS claimed that it was not possible for Brahman to be the material cause of the world, because the world the effect was radically different from Brahman the cause.49 The world is evidently insentient, impure, and full of suffering, whereas Brahman is defined as essentially sentient, pure, and bliss solid: it cannot be that the first is an effect of the second. Bādarāya a’s reply was that precisely such cases of causal relations where the effect was radically different from the cause were in evidence.50 The commentators give several instances of such cases, most of which fail to impress—worms produced from honey, dung-beetle from dung, etc.—but two have intuitive appeal: the insentient hair that grows from a sentient body, and the insentient cobweb that a sentient spider produces. Such cases of empirically knowable causal relations, then, are proof enough that there is no such requirement that the effect be of the same nature as the cause. It must be real, sat or Being, and insofar as such is the case, Vedānta endorses sat-k rya-v da equally with Sā khya, but the effect has a surplus of characteristics beyond sat that are not shared with the cause. Brahman, further, is such a cause which does not need the intervention of another agent for its transformation: it is constitutionally such a thing which, left to its own internal devices, would transform into its product, without the external intervention of another thing, like milk which left to its own internal structure would transform into curd without the addition of whey. Whether one finds the arguments compelling and sound or not, they are characteristically philosophical insofar as they endorse Sā khya’s own game, s m nyato-dṛṣṭa-anum na, analogical reasoning from known to unknown causal relations in the light not of scripture but of 49 BS 2.1.4. 50 BS 2.1.6. 29 reason. Such arguments were possible, again, because of a shared sphere of commitment, the doctrine of sat-k rya-v da. Bādarāya a was explicit about itŚ charged by the Sā khya opponent that his account of causality amounts to asat-k rya-v da, he replies that his contention is not against the doctrine of sat-k rya-v da, but just against the claim that the effect must be like the cause.51 Further, he was willing to engage the metaphysics outside sat-k rya-v da such as Nyāya-Vaiśeṣika and Buddhism, because he had a larger shared sphere of commitment to the discourse of causality as such. This is the part of the BS where there are no topical passages referenced by the individual sūtras, because the context is such that arguments from scripture won’t fly. It is strictly a philosophical mode of discourse. While its purpose was obviously not to advance original proofs about Brahman or ascertain Brahman’s peculiar characteristics—the Upaniṣads are the sole pram ṇa in that regard, and Vedāntins have never compromised with that—insofar as Brahman was sat, Being, early Vedāntins had a keen interest in engaging philosophically with other traditions. Śa kara himself bears witness to this fact: whenever the question of Being and non-Being presents itself in a text he is commenting on, he finds an occasion to advance his peculiar understanding of Being and argue against the several doctrines of asat-k rya-v da, specifically Nyāya, Vaiśeṣika, and the Buddhist philosophies, solely on the grounds of reason, and he expressly claims that it follows not only from scripture, but inference as well, that the world was Being in the beginning. Vedānta, thus, endorsed philosophical reasoning and advanced characteristically philosophical arguments because it had a shared sphere of commitment to the discourse of causality through non-sectarian forms of argument. 51 BS 2.1.7. 30  Thus, in adjudicating whether something is theology or philosophy, one should look at the kinds of arguments that are made, and that will be largely dependent on the shared sphere of commitment. No matter what pram ṇa one assigns to a domain, when it comes to defending one’s position in a context of diverging presuppositions, specifically regarding doctrinal authority—when it comes to arguing how one’s understanding, for instance, of the first principles makes more sense than competing doctrines—there is no avenue for arguing, so long as argument is wanted, but for non-sectarian, and thus non-theological, forms of reasoning. Śa kara himself, in fact, was perfectly aware of this: under the BS section that contains the Vedāntic arguments against the Buddhist schools, he says at one point that the reality of ether is established from scriptural statements, specifically the Taittir ya claim that ether arose from the Self, “but for those who are opposed to the authority of scripture, it must be presented as inferable from the quality of sound.”52 This is a typical case of the Sā khyan s m nyato-dṛṣṭa- anum na, inferring unknown causal relations and ontological reals from what is empirically available. On one occasion at least, Śa kara even exercised the classical Nyāya inferential argument for the existence of God, Īśvara, and its necessary attributes of omniscience and omnipotence.53 It makes little sense to describe such mode of discourse as theological under the aspect of “faith seeking understanding” either, because what is sought is not understanding, but vindication. One could belabor the point with many similar instances about other areas of philosophical inquiry, from Vedānta and other traditions, but that seems superfluous.54 Generally, apologetics was the origin of that mode of discourse in the Indian traditions which 52 vipratipannān prati tu ... anumeyatvaṁ vaktavyam; BSBh 2.2.24, II.388. 53 KUVBh 3.1. 54 See, for instance, the work of Arnold (2005), Taber (2005), and McCrea (2013). 31 was characteristically philosophical. It required the development of non-sectarian forms of argument, such as the theory of pram ṇa, philosophy of language, and the method of prasaṅga or unwanted consequences of a thesis, in which specific issues such as the validity of scripture or the being of first principles could be debated across traditions with various doctrinal commitments. This dissertation is concerned with that shared sphere of commitment constituted by the acceptance of Vedic statements as reliable epistemic warrants on all supersensible matters, and of the recognized canons of theological reasoning. I call this sphere “Vedic” rather than Hindu theology because it was restricted to the two Hindu schools whose specific concern were the Vedas: Mīmāṁsā and Vedānta. Thus, by “Vedic” I do not mean “in the Vedas,” but “pertaining to the Vedas.” The specific topics that I investigate are scripture and the highest good, specifically scripture as the instrument of attaining the highest good, as they were discussed in the shared sphere of commitment. This, perhaps, bears repeating one more time: while the highest good and the epistemic validity of scripture and knowing from linguistic utterances were debated fiercely across traditions, my essay is not concerned with such debates, but solely with the sphere in which theological arguments based on the Vedic canon were a valid form of reasoning. It is for this reason that I treat my characters as “Vedic theologians.” Someone with a different objective may be justified in calling them philosophers: generally, we do not debate whether a Sartre was a philosopher or a novelist or a playwright etc. We look at the mode of discourse. One final point. The fact that I talk about theology does not mean that we will not see many characteristically philosophical arguments. Śa kara, for instance, insisted that the attainment of liberation required both kinds of reasoning, theological and philosophical, stra 32 and yukti, the first taking its data from scripture and the second from experience. The two modes of reasoning—and I insist that both were modes of reasoning, such that the common distinction between faith or revelation vs. reason is misleading—were embodied in the processes of ravaṇa and manana and sharply distinguished. As we will see in Chapter Nine, liberation was downright impossible without philosophical reasoning, and such reasoning would be indistinguishable from what is properly Indian philosophy, except that it served an ultimately theological purpose: full understanding of the scriptural truths. But, in many cases the two were so intimately related that separating them would be at our own peril. While the theological mode of discourse helps us in identifying the shared sphere of commitment and facilitates nomenclature, I will approach this sphere holistically. Larry McCrea is fully justified in claiming that, at least in Mīmāṁsā and Vedānta, philosophical issues and matters of scriptural interpretation are “inextricably bound up together in manifold and complex ways, and any attempt to write the history of the discipline must strive to take account of the full range of internal and external factors that shape and constrain changes in the field.”55 Mm s and Ved nta as P rva- and Uttara-M m s My large thesis in the dissertation, as I stated in the beginning, is that the Vedāntic idea of mah - v kya was developed by Sarvajñātman on a Mīmāṁsā modelŚ indeed, that understanding what mah -v kya was about is facilitated by appreciating its Mīmāṁsā background. In the previous section, further, I claimed that there was such a thing as “Vedic theology,” in which the two schools of Mīmāṁsā and Vedānta formed a unique field, distinct from the other Hindu traditions. Since the question of the unity of the two schools has received some scholarly attention in recent 55 2013:141-2. 33 years, I should like to clarify here in what sense I take “Vedic theology” to have been a unique field. It is well-known that Mīmāṁsā and Vedānta are two systems of interpretation of the Vedas, concerned with ritual and Brahman respectively.56 They are, thus, commonly called karma-m m ṁs and brahma-m m ṁs , an inquiry into ritual and Brahman, codified in the M m ṁs -Sūtra and the Brahma-Sūtra. Likely already by the time of Śa kara, they were also known as the “prior” and the “subsequent” inquiry, pūrva-m m ṁs and uttara-m m ṁs .57 The precise nature of the pūrva-uttara relationship is open to some conjecture, but two possibilities are noteworthy. Hajime Nakamura made the sensible suggestion that Vedānta was posterior to Mīmāṁsā in the sense that “the Vedānta Mīmāṁsā presupposed the ritual Mīmāṁsā as a precondition. The ritual Mīmāṁsā can be set up without necessarily presupposing the Vedānta Mīmāṁsā, but the Vedānta Mīmāṁsā, on the contrary, from the first assumes the ritual Mīmāṁsā as a precondition.”58 Nakamura’s suggestion has an intuitive appeal, because understanding the BS is impossible without a good grip on principles of interpretation that can be learned only from the MS: the BS assumes a lot.59 Another valuable suggestion has been made by Asko Parpola, who proposed that the names of the two disciplines had come from the names, or rather headings, of the two parts of 56 There is no need to review here the full literature on the history of the two schools: an exhaustive overview and bibliography can be had from Nakamura 1983, particularly 369-447; Parpola 1981 and 1994; Bronkhorst 2007b; and Aklujkar 2010. Nakamura’s second volume (2004) also has a lot of relevant material. 57 See Nakamura 1983:409-412. Nakamura’s two volumes are still the most exhaustive and reliable one-stop shop source for the history of early Vedānta. 58 Nakamura 1983:412. 59 In his critique to Śa kara’s Brahma-Sūtra-Bh ṣya, Bhāskara says something along similar linesŚ taking stock of two claims of Śa kara’s, that the inquiry into Brahman is not consequent on the inquiry into ritual, and that some Upaniṣadic meditations combine with ritual but some are just about Brahman, he says that without the inquiry into ritual one could not possibly know which Upaniṣadic passages combine and which do not so as to make the distinction. Even in that negative sense, then, the inquiry into Brahman presupposes the prior m m ṁs . 34 one single work called M m ṁs -Sūtra. That is, initially the two sūtra compositions were two parts of a single M m ṁs -Sūtra, a first (pūrva) and a second (uttara) part respectively, and the present MS and BS as well as the two disciplines have evolved from the titles. Parpola makes a strong case for his claim, and his studies are the most thorough engagement with the early history of the two m m ṁs s, and the best historical explanation put forward so far.60 In any case, whether the MS and the BS were initially a single work or two distinct but closely related works, it is difficult to read them side by side without the impression that they belonged to a closely shared intellectual milieu. Johannes Bronkhorst, however, claimed relatively recently that the view according to which Vedānta was in the beginning inseparably linked to pūrva-m m ṁs contradicts some facts.61 Namely, the tradition of Mīmāṁsā up to and including Śabara and Prabhākara shows no awareness of liberation. “Śabara’s Bh ṣya deals with Vedic ritual, which as a rule leads to heaven.”62 Vedānta, on the other hand, “has, presumably from its beginning, been about liberation through knowledge of Brahma.”63 If the two were one 60 The outline of Parpola’s argument goes something like this. There are in Sanskrit literature several works that are divided in two parts, pūrva and uttara, and a systematic practice of doing so in the S ma-Veda; for instance, the Jaimin ya Gṛhya-Sūtra has a pūrva and an uttara/apara part. Further, there is clearly the possibility that Jaimini had written a Brahma-Sūtra himself, which would have been utilized and replaced by Bādarāya aŚ Sureśvara had, in fact, ascribed the BS to Jaimini. While the Sureśvara argument has been put forward by several scholars in support of an earlier Jaimini BS, Parpola to his credit shows how reworking someone else’s sūtra composition was a common practice in the early post-Vedic periodŚ if Bādarāya a had used and replaced Jaimini’s work, he would not have been the first to do such a thing. Next, it is well-known that the MS and BS quote several Vedic scholars or teachers, including Jaimini and Bādarāya a, numerous times, but an analysis of these quotations would show that Bādarāya a is later than Jaimini and had reworked his BS. The brevity of the present BS also supports its later date, as the rauta-Sūtras tend to abbreviate when taking material from another rauta-Sūtra. Finally, the BS clearly refers to the MS five times, using the phrase tad uktam, “that has already been explained.” The rest of Parpola’s work is a highly instructive analysis of the teacher quotations in the two Sūtras, which suggests that Jaimini was a central character in both, and that “Bādarāya a evidently is a teacher who has intruded into the Mīmāṁsāsūtra after its original composition.” See Parpola 1981 and 1994. Unfortunately, Parpola’s work is still unfinished. 61 Bronkhorst has presented his argument in several writings, but its fullest incarnation is in his edited volume M m ṁs and Ved ntaŚ Interaction and Continuity (2007b). 62 Bronkhorst 2007b:1. 63 Bronkhorst 2007b:4. 35 in the beginning, where did liberation in early Mīmāṁsā disappear? “It will be clear that the idea of an original unity of Pūrvamīmāṁsā and Uttaramīmāṁsā raises serious questions.”64 Having set the issue in these terms, the absence and presence of liberation from the beginning, Bronkhorst does not really tackle it: he reviews the arguments about the unity of the two schools that have been made in secondary literature and attempts to show that the evidence—which is, it bears mentioning, all circumstantial—does not support such unity. He does not bother examining just what heaven and liberation were in the two stras—were they really incommensurable—and he hardly engages with the sūtras at all. Unlike Parpola’s contextualization of Mīmāṁsā in the whole range of sūtra literature, Bronkhorst’s work is all “he said, she said,” inorganic. Bronkhorst’s challenge is, thus, weak, and even his reading of the circumstantial evidence is often faulty, as shown by Ashok Aklujkar.65 For my thesis here, however, the very question of the pūrva-uttara relationship in terms of origin is somewhat immaterial. The sense in which I take Vedānta to have been uttara to Mīmāṁsā so as to form a unique field with it concerns two presuppositions, both of which I have indicated under the previous heading. They both go back to the MS, and one of them was attributed to Bādarāya a himself. I will define the first by looking at Śa kara’s student Padmapāda’s accepting just one sense in which Vedānta as an inquiry into Brahman (but not the BS as a book) was uttara in relation to Mīmāṁsā. Padmapāda says that two rules stated in the M m ṁs -Sūtra were operative in the inquiry into Brahman as well, and it is the second that we are interested in here: it is MS 1.1.5, known as the autpattika-sūtra of Bādarāya a, which says that the relationship of words to their meanings is innate; that the Veda is not a production of a 64 Bronkhorst 2007b:3. 65 Aklujkar 2010. 36 personal agent of any kind, human or divine; and, that it is a reliable epistemic warrant that is not derivative on some other warrant of such kind.66 This rule, in Padmapāda’s words, was required with regard to Brahman just as it was required with regard to ritual.67 As we saw in the previous section, the characteristic nature of this reliable warrant, the Veda, was that it was the means of knowing supersensible things. The second presupposition was that the Veda in its full scope, from the mantras to the Upaniṣads, was essentially an instrument of human good, and that any doctrine one might develop must not make any part of the Veda meaningless by making it purposeless. The core of this presupposition was expressed in the MS 1.2.7, and two magnificent but very different testaments to it were Kumārila’s Tantra-V rttika 1.2.7 and Śa kara’s Bṛhad- raṇyaka- Upaniṣad-Bh ṣya 4.4.22.68 66 tatra yaḥ prathama-sūtre ‘tha-śabdāpādāna-sūcito nyāya svādhyāyasyārtha-bodhopayoga-pratipatti-hetuḥ yad apy autpattika-sūtre śabdārtayoḥ sambandha-nityatvena vedāntānāṁ cāpauruṣeyatvena kāra enānapekṣatvaṁ nāma prāmā ya-kāra am uktam. tad ubhayam ihāpy upayujyatām apekṣitatvāt. PP, p.59. 67 Padmapāda (and Advaitins generally) under brahma-jijñ s did not mean the Brahma-Sūtra that is the book. He recognized that large sections of the book absolutely required the rest of the “one thousand Mīmāṁsā rules,” specifically the sections on the formation of Upaniṣadic meditations, which Advaitins thought concerned the saguṇa Brahman and were part of the same dharma as ritual, and thus not part of the inquiry into Brahman. See PP, p.59. We will talk about this extensively throughout the dissertation, but even under this professedly thin account of unity, the BS as a book was absolutely uttara to the MS. 68 This principle that the Veda in all its scope must be purposeful was recognized even by those Vedāntins who were remote from the BS/MS complex, such as Gauḍapāda. To illustrate, the third prakaraṇa of Gauḍapāda’s gama- stra is an attempt to provide a frame of reading the Upaniṣads, but it is a frame very different from that of the BS. Gauḍapāda severs himself with a single incision from the MS/BS tradition in k rik s 3.1-2Ś “That dharma which resides in the Upaniṣadic meditations (up sana) is operative with regard to that Brahman which is originated. Before origination, everything is unborn. Therefore, such dharma is regarded as wretched. I will, thus, propound that state which is not wretched, without origination and uniform, so as to show that though things are originated wherever one looks, nothing, in fact, is originated at all.” upāsanāśrito dharmo jāte brahma i vartate | prāg utpatter ajaṁ sarvaṁ tenāsau k pa aḥ sm ta || ato vakṣyāmy akārpa yam ajāti samatāṁ gatam | yathā na jāyate kiñcid jayamānaṁ samantataḥ. In K rik 16 of the same prakaraṇa, however, he goes on to sayŚ “There are three stages of life, involving a vision which is low, middle, and high. The Upaniṣadic meditations are taught out of compassion for these three stages.” āśramās trividhā hīna-madhyamotk ṣ a-d ṣ ayaḥ | 37  We will unravel the first presupposition in the first chapter and the second in the rest of the dissertation, but I note here that I take these two as the core in the light of which the two m m ṁs s were a unique field of Vedic theology: the first accommodated the use of scripture as argument, and the second provided the ground for a discourse on the highest good in which the Mīmāṁsakas, in the eyes of Advaitins, were legitimate participants and contenders. Over and above these two presuppositions was the Advaita use of the Mīmāṁsā canons of text formation. The doctrine of mah -v kya itself emulated the formation of ritual idealities through hierarchy of scriptural statements, in which such idealities as units obtained the characteristic feature of finality of meaning. In that sense as well, Advaita Vedānta was uttara to pūrva-m m ṁs , a fleshing out of a Mīmāṁsā skeleton with Advaita meat. We will see this in Chapter Ten. The History of Ideas I said above that, although I talk about theology and theologians, my essay is not a work in theology. I should like to clarify that statement now. First, the dissertation it is not a work of a contemporary Hindu theologian of the kind that Francis Clooney had described: a believer who allows her “belief to remain an explicit and influential factor in their research, analysis and writing.” My object is, thus, not to articulate a personal interpretation of a tradition, and personal beliefs are intentionally bracketed. Second, my primary concern is not to reconstruct theologies so as to isolate them as phenomena, such as John Carman was doing in his remarkable work on the theology of Rāmānuja.69 Third, the dissertation it is not conceived as a work in comparative upāsanopadiṣ eyaṁ tad-artham anukampayā. We may take this as the origin of the Advaita principle that the meditations on the saguṇa Brahman, which can be combined with ritual, can be used for the purification of the agent. Thus, although dharma may be wretched, it must have a good use. 69 Carman 1974. 38 theology, either of the “soft” kind exemplified by Julius Lipner, a reading of Rāmānuja that relies on categories of Catholic theology, or of the “hard” kind exemplified by Francis Clooney, making one’s beliefs vulnerable to truths of other traditions in a personal hermeneutics project.70 Rather, I see this project as an essay in the history of ideas, with the additional clarification that the specific ideas whose history I attempt to track happen to be theological, or belong to a mode of discourse that is best characterized as theological. The history of ideas, of course, means different things to different historians, and here I have deliberately followed the method of historical interpretation that is concerned specifically with texts as speech acts, articulated by Quentin Skinner and inspired by the philosophy of language of J. L. Austin.71 Very briefly, Skinner’s thesis about the method of interpretation in intellectual history, which concerns itself with ideas expressed in texts, is that one ought to approach the study of ideas in a “properly historical style.” Such style attempts not to assume the perspective of our interest in an issue, so far as that is possible, but to “see things” the way that the authors of the texts we study have seen them.72 To bring home what this way of interpretation precisely involves, Skinner contrasts it with ways of reading that approach authors and texts with preconceived expectations of what they should discuss. One such way of reading is the querying of past great philosophers, political theorists, etc., on so-called “perennial issues,” questions universally relevant to man, in an attempt to reconstruct what a past master might teach us about things that are important to us, under the assumptions that, the questions being “perennial,” the past masters must have addressed them. A similar way of reading is through a focus on so-called “unit ideas,” for 70 Lipner 1986; Clooney 1993. 71 Skinner 2002. 72 Skinner 2002:vii. 39 instance, the idea of progress, social contract, equality, the problems of knowing, etc., that are from the start conceived as a sort of an ideal type—one might even describe them as Platonic— and are part of constituted disciplines. This way of pursuing the history of ideas often comes with a frame of evaluation that takes the vantage point of a formed notion, or of one’s own philosophy—think of Hegel—from which individual authors are judgedŚ they have “failed to develop” the idea, “anticipated” it, “contribute” to it, etc.73 In any case, the purpose is to look at and evaluate ideas and beliefs from our own perspective. The problem Skinner has with such readings is that they tend not to notice what was important to the authors of the texts: what precise concerns they had in writing them or in saying what they said. The “properly historical style,” on the other hand, attempts to situate texts in intellectual contexts and frameworks of discourse which would enable us to recognize not what such texts might mean to us, but what they have meant to their authors: this, in fact, is often indispensable even for the bare understanding of their meaning. To put it differently, to understand the meaning of texts as their authors have meant them, to see things their way, intellectual history as history should attempt to approach meaning not from our vantage point, but from the perspective of their authors. To appreciate this, it is necessary to disambiguate several senses of “meaning” that, in Skinner’s finding, are often conflated.74 One may talk about meaning in the sense of “sense” and “reference,” that is, in the sense of the denotative or signification functions of words and sentences in a text. Or, one may look at meaning from the reader-response approach to interpretation, such as that exemplified by the notion of “surplus meaning” of Paul Ricoeur, who recognized that a text might have initially had a “pristine” meaning, but claimed that such 73 Skinner 2002:1-26, 57-89. 74 Skinner 2002:90-102. 40 meaning over time, and through the polysemic and metaphorical features of language, assumes autonomous and acquired meanings not intended by its author. This way of looking at meaning answers the question, “what does this text mean to me?” Finally, “meaning” may be used in the sense of authorial intention, which answers the question, “what does a writer mean by what he or she says in a given text?” How does this third sense of “meaning” differ from the first? It is to distinguish the two that Skinner appeals to Austin’s theory of speech acts. Consider the following case. “A policeman sees a skater on a pond and says ‘The ice over there is very thin.’ The policeman says something and the words mean something. To understand the episode, we obviously need to know the meaning of the words. But we also need to know what the policeman was doing in saying what he said.”75 The policeman makes a declarative statement, which, nevertheless, has on top of its sense an “intended force” with which it is issued: it is a warning. The policeman not only says something, but does something in saying it. Austin called this feature of language “illocutionary force,” and we need not spend much time on it, except that we must note two things. The illocutionary force that turns speech into an act is a feature solely of speech: whether the policeman does succeed in warning the skater or not is, for the purpose of speech, irrelevant. The point is that the sentence as an act has such force, over and above its denotation. So, let us characterize this sense of meaning as speaker’s intention. Second, the fact that statements are a mode of communication makes this intention public, insofar as it is also intended to be understood. How does this apply to the interpretation of texts and the history of ideas? Well, to understand the meaning of a text, it is not enough just to understand how words are used or what 75 Skinner 2002:104. 41 sentences mean; it is, further, required to understand the authorial intention, the illocutionary force of statements in a text. In other words, it is required to understand what specific speech act an author was performing in writing the text or in saying what s/he says. To do that, it is required to identify the context in which something is said—the background, the intended audience, what is intended to be communicated—and once we approach interpretation form such standpoint, we begin asking, on top of what a text means, what specific speech act the author is performing. In the South Asian context, for instance, we may ask whether a text that we study is a case of justification through intertextuality, of overcoading, of an intervention in a preexisting debate or a discourse, etc. Identifying the kinds of speech acts in a text, then, not reconstruction of beliefs, becomes the “specifically historical mode of reading,” or intellectual history, because such reading attempts to track authorial intention. Two additional notes are apposite. First, Skinner insists that intentions are not motives: an intention does not answer to why an author wrote what s/he wrote, only to what s/he was doing in writing. Insofar as speech acts are forms of communication, they are intended to be understood, and, thus, are public. “[T]he intentions with which anyone performs a successful act of communication must, ex hypothesi, be publicly legible.”76 Such reading, then, does not aspire to “get in the head” of anyone, but simply to understand what the illocutionary force of a statement or a text is. In the case of the policeman, the public intention can be known from the tone with which the statement is given, from hand gestures, etc. In reading texts, “we need first of all to grasp the nature and range of things that could recognizably have been done by using that particular concept, in the treatment of that particular theme, at that particular time. We need, in short, to be ready to take as our province … the social imaginary, the complete range of the 76 Skinner 2002:97. 42 inherited symbols and representations that constitute the subjectivity of an age.”77 What one needs, really, is sort of a contextual omniscience. Second, such reading is often indispensable even if one’s express aim is not historical interpretation, but, say, properly philosophical engagement with a text. This is so because meaning as sense and reference is often impossible to ascertain without appreciating meaning as authorial intention. Let me illustrate this with a case that is closer to home. One of the most puzzling and discussed passages in Śa kara’s works is his comment on the statement satyaṁ jñ nam anantaṁ brahma of the Taittir ya Upaniṣad 2.1.1. This is, obviously, a very important passage because it contains the definition of Brahman and Śa kara’s attempt to explain how language and the Upaniṣads reveal Brahman that is outside the domain of language, but it is puzzling for various reasons, some of which involve Śa kara’s both affirming and denying that the pertinent statement is a case of co-referentiality such as the famous “blue lotus” illustration; both affirming and denying that the word jñ nam stands for the verbal action of “knowing,” bh va or dh tv-artha; and the use of lakṣaṇ ṭlakṣaṇa such that it is not clear if he talks about one or two features of language. (All of this I will discuss later.) Various interpretations have been given of the passage,78 none of which makes a significant effort to situate it in the wider context of the contemporary Indian philosophy of language, and all of which fail to see two crucial thingsŚ first, that Śa kara grapples with two problems relating to Brahman and language, first, that Brahman is not a sentential reference, v ky rtha, and, second, that Brahman is not expressed by the primary signification function of individual words, v cyam—the two are different problems; second, that midway in the comment, the argument changes because the context changes, namely Śa kara moves on to discussing the 77 Skinner 2002:102. 78 For instance, Hirst 2005:154-151; Lipner 1997; Bartley 1986. 43 further statement tasm d v etasm d tmanaḥ, which he takes to be identical in meaning to the paradigmatic identity statement tat tvam asi, and no longer views Brahman from the perspective of the category of tat, whose domain includes satyaṁ jñ nam anantaṁ brahma, but shifts to the identity statement perspective, where language operates through the secondary signification function. Śa kara, then, returns to satyaṁ jñ nam anantaṁ brahma from this higher perspective of the identity statement to modify the initial argument. The whole argument cannot be appreciated without some grasp of the contemporary Indian philosophy of language, particularly Bhart hari and Kumārila, and without consulting Śa kara’s comment on the BS 4.1.2, where he lays down the identity statement doctrine. Further, the argument can hardly be understood without figuring out what Śa kara is doing in writing the comment, which is addressing the doctrine of prasaṅkhy na meditation, in which Brahman of the Upaniṣads turned out to be a sentential reference, a definite description. Śa kara’s reply to the doctrine of prasaṅkhy na was the notion of the identity statements and of their meaning obtaining through the secondary signification function. Śa kara’s authorial intention, then, was to make an intervention in a preexisting discourse on Brahman, scripture, and liberation, that was very close to home for him, but from which it was absolutely necessary to make an exception. We will unfold the details of this in the later chapters, but here the point is that to fail to see these details is to fail to see things Śa kara’s way, but it is also to fail to understand the meaning of this all-important passage so that one could engage its philosophy of language, ontology, etc., philosophically, or its notions of scripture and Brahman theologically. With this in mind, I can now describe the character of my project. First, it is an interpretation of Śa kara’s understanding of a set of related ideas—scripture, dharma, the highest good, liberation—that does not attempt to reconstruct Śa kara’s understanding of the 44 notions in itself, but looks at his theology as a set of interventions in a preexisting discourse on these ideasś it reads Śa kara’s theology as consisting of various speech acts with publicly available authorial intentions. The discourse, then, that takes place in Vedic theology, provides the requisite context without which Śa kara’s interventions cannot be appreciated, and eo ipso his understanding cannot be reconstructed, and it is for this reason that half of the dissertation is concerned with laying out the context.79 Second, taking chapters one through nine as another, single context, the dissertation witnesses the appearance of the mah -v kya idea in the theology of Sarvajñātman and provides, against that context, an interpretation of what mah -v kya was for him and why he gave that name to the Upaniṣadic identity statements. In other words, the dissertation untangles from the context all the strands that are required to understand Sarvajñātman’s intention in calling the identity statements mah -v kya, and with that benefit it explains just what was “great” in them: it relies on the meaning of Sarvajñātman’s mah -v kyas as authorial intention to appreciate their 79 My essay, then, differs from other scholarly works that treat Advaita soteriology primarily in this method of interpretation, the “properly historical style.” To illustrate, the notion of “scripture” in Advaita Vedānta is prominent in Francis Clooney’s monograph Theology After Ved nta (Clooney 1993). Clooney’s book, however, is primarily concerned with reading texts as a soteriological practice and as a tradition; with committed or transformative reading; with writing commentaries through the need to keep the tradition of reading fresh, etc. It is a project in theology, and in comparative theology at that. Scripture is also important in J. G. Suthren Hirst’s book aṁkara’s Advaita Ved ntaŚ A Way of Teaching, but Hirst is concerned with reconstructing modes of teaching in scripture itself and in Śa kara, and whereas her reading of Śa kara is thorough, her engagement with Śa kara’s context is notŚ there is little in the way of solid intellectual history to be gleaned from the book (Hirst 2005). Anantanand Rambachan’s Accomplishing the Accomplished is, overall, a good intervention in the interpretation of Śa kara’s understanding of the respective role of scripture, personal experience, reasoning, and soteriological practice, but his engagement with Śa kara’s context is poor. Just to illustrate, he relies on the medieval Advaita Vedānta manuals the Ved nta-S ra of Sadānanda and Ved nta-Paribh ṣ of Dharmarāja to “explain” what “Advaita Vedānta” as a unit “thinks,” “accepts,” “has a position on”, “finds,” etc.ś indeed, the bulk of his book reads like a medieval manual (Rambachan 1991). A. G. Krishna Warrier’s The Concept of Mukti in Advaita Ved nta (Krishna Warrier 1961), Lance E. Nelson’s “Living Liberation in Śa kara and Classic Vedānta” (Nelson 1996), and Andrew J. Fort’s J vanmukti in Transformation (Fort 1998) are all valuable and insightful studies about liberation in Śa kara and Advaita Vedānta more broadly, some more historical than others, but generally exemplifying the “unit idea” approach. So far as I have been able to see, there are no dedicated studies of Śa kara’s understanding of dharma except insofar as it was the negative of Brahman as the proper domain of the pūrva-m m ṁs , although one could make a strong case that it is precisely the notion of dharma that is the connecting tissue which keeps scripture, liberation, and the highest good together, as we shall see. 45 meaning as sense. In doing the second, the essay subscribes to another principle in the study of ideas, expressed by one of the great intellectual historians of our time, Pierre HadotŚ “[I]n the words of Aristotle, if one wishes to understand things, one must watch them develop and must catch them at the time of their birth.”80 Structure and Chapter Overview The dissertation is divided in four parts and a total of ten chapters. The first part explores the two presuppositions of Vedic theology that I flagged as shared by Mīmāṁsā and Vedānta. In Chapter One, I investigate two related ideas, ruti and apauruṣeyatva, which together stand for the Vedas as books that were never written in time, but are coexistent with the world, represent its linguistic blueprint, and are a source of knowledge from linguistic utterances that is not testimonial in kind. Against all scholarly accounts, I show that the notion of scripture as ruti or “hearing” did not stand for the Vedas that have been heard at the beginning of creation by a few Vedic sages, but primarily for individual scriptural statements that must be taken as valid just as they are heard, a “hermeneutic data” for interpretation in the same way as perception supplies data for inference and other forms of reasoning. The commitment to these two ideas sets Mīmāṁsā and Vedānta from all other intellectual traditions in premodern South Asia. The second chapter investigates the notion of dharma as the central concern of Vedic theology, and its relation to the “good of man,” puruṣ rtha. I show that dharma in the tradition of Mīmāṁsā was understood as a means of attaining something desirable to men through harnessing causal relations that are knowable solely from the Veda. I analyze dharma through the teleology of “ritual causality,” and show how Mīmāṁsakas theorized the hierarchical organization of scripture around dharma such that everything that is in the Veda would be 80 Hadot 2002:2. 46 purposeful by serving ritual needs, and ultimately the telos of the “good of man.” The chapter, thus, uses the paradigm of ritual to present the basic categories of Vedic theology that we will see in the “soteriological causality” of Vedānta, and presents the idea of textual hierarchy that will be determinative for Sarvajñātman’s mah -v kya doctrine. In the second part, I delimit the focus of dharma further on the idea of the “highest good,” niḥ reyasa or paraṁ reyas, and I write the history of negotiation between two such ideals in pre-Śa kara Vedic theology, heaven (svarga) and liberation from the cycle of rebirth (mokṣa). As we saw under the heading on pūrva- and uttara-m m ṁs , the divide between the two is commonly taken by scholars as the line of demarcation that sets off Mīmāṁsā and Vedānta as two rival traditions. I, however, show that heaven and liberation throughout the two traditions were often defined in identical terms—a state of unexcelled felicity, nirati aya-sukha or pr ti—and that the real wedge of contention between Mīmāṁsā and early Vedānta was less about the highest good, and more about the means of attaining itŚ the Mīmāṁsakas were advocated of ritual, whereas Vedāntins promoted meditation on Brahman. Chapter Three focuses on heaven as the highest good in Mīmāṁsā, and on Kumārila’s accommodation of liberation as a human good that is also attainable solely through Vedic means. In Kumārila, we will see two accounts of liberationŚ in the first, the key role in terms of soteriological causality is played by ritual; in the second, by Upaniṣadic meditation. We will also see that the two accounts involved very different understandings of liberation as a state, and I will argue that Kumārila accepted the second as the highest good, but made provisions for the first as well, as sort of a liberation attained by a karma-yoga of the Bhagavad-G t kind. Kumārila’s second account was very similar to the doctrine of liberation in the Brahma- Sūtra, which is the topic of Chapter Four. There, I show that liberation in the BS meant attaining 47 the highest Vedic heaven, brahma-loka, through meditation on Brahman as one’s Self. This meditation was mental absorption by means of any one of the so-called brahma-vidy s, which were full Upaniṣadic counterparts to Vedic ritual, standardized through appeal to principles laid down in the M m ṁs -Sūtra. Liberation as going to brahma-loka consisted in independence or sovereignty, “having no other master but oneself,” which specifically referred to the ability to enjoy all desires that were commonly associated with Vedic ritual, and the ability to move through the heavens of the Vedic world without impediment. Thus, liberation meant becoming as similar to Brahman as possible, which included being able to have all desires fulfilled and resolves accomplished and excluded only the ability to create the world. In Chapter Five, we move to the doctrine of prasaṅkhy na meditation, which was much closer to home for Śa kara because its advocates were Advaitins, such as Ma ḍana Miśra, or otherwise Vedāntins in whose understanding individuality was lost in liberation, and ultimately there was just the single Brahman. This doctrine, which had come to Vedānta from the tradition of Yoga, said that the scriptural knowledge of Brahman from the Upaniṣads was insufficient for liberation because it was mediate, not a form of perceptual awareness, and sentential, dependent of concepts and mental constructs. By “scriptural knowledge,” we mean the propositional knowledge of Brahman in the Upaniṣads, presented in positive and negative descriptions, such as “Brahman is consciousness, bliss,” and “the Self is free from faults, from old age and death etc.” Liberation required this scriptural knowledge to be followed by meditation, whose purpose was to reconstitute the subject through purification. Only such meditation could provide direct experience of Brahman. The two kinds of meditation presented in Part Two, the BS and the prasaṅkhy na, were in a sense very different—we may call them assimilative and cathartic respectively, and the 48 attainments towards which they aspired were different—but as kinds of awareness both were forms of concentration or mental absorption. Two large issues will impress upon our understanding as we follow dharma, scripture, and the highest good from the second to the fifth chapter: first, that pre-Śa kara Vedic theologians generally held that the most important statements in the Veda were of the injunctive kind, such as those that enjoin performance of ritual or meditation on Brahman; in other words, that scripture were not as much about knowing something as they are about doing something; second, related to the first, that the paradigmatic Vedāntic process of liberation was meditation on Brahman in some combination with Vedic ritual; that, in any case, simply knowing Brahman as a form of understanding—knowing qua knowing—was insufficient for liberation; or that liberation was not a result of knowing something, but of doing something. This will partly serve a negative purposeŚ when we move to Śa kara’s brand of Advaita Vedānta, we will be in a position to understand Śa kara’s model of soteriological causality, and we will appreciate why some Upaniṣadic statements and passages that were immensely important in old Vedānta did not have a shot at becoming mah -v kyas, or at providing the basis for the doctrine of liberation that went with the mah -v kya notion. The doctrine of prasaṅkhy na meditation, however, will provide the sole background against which we will understand Śa kara’s more specific intervention in the discourse of liberation. In the third part, we will follow Śa kara’s engagement with scripture and the highest good through the lens of his understanding of dharma. In Chapter Six‚ we will see that Śa kara joined forces with the Mīmāṁsakas in taking the purpose of the Veda to be the good of man (puruṣ rtha), but fiercely opposed them in insisting that the characteristic feature of dharma in general was knowing, not doing. Such dharma could, further, transit into action, and whenever it 49 did, the purpose that such dharma served was that of prosperity or advancement, abhyudaya, the culmination of which was the attainment of brahma-loka, which, we will remember, in the Brahma-Sūtra was liberation itself. The highest good, however, niḥ reyasa, was solely a result of knowing Brahman, knowing which by its very nature could not transition into action. Scripture facilitates both these goals, abhyudaya and niḥ reyasa: its injunctive character is exhausted in ritual and Upaniṣadic meditations, but liberation is attained solely by means of the propositional or existential knowledge of the Upaniṣads. In Chapter Seven, we will begin mapping the path to liberation from inception to consummation. I will first argue that Śa kara, explicitly drawing on the Mīmāṁsā organization of ritual, replaced the traditional Vedāntic method of jñ na-karma-samuccaya or combination of “knowledge” (that is, meditation on Brahman) with action (that is, ritual) with a model of “mediate soteriological causality” or p ramparya, in which a means ceases being a direct means once it produces its characteristic result. The path to liberation begins with jñ na-karma- samuccaya, or alternatively just with meditation on Brahman, by a repurposing of meditation and ritual through giving up the specific desire that the respective meditation and ritual are said to bring. Such practice of disinterested ritual and meditation culminate in “purity of existence” or the “arising of knowledge,” whose ultimate point is the “desire after liberation,” mumukṣutva. At that point, ritual and meditation must be given up through formal renunciation, and be succeeded by another kind of jñ na, knowing not as meditation, but as understanding Brahman through theological and philosophical inquiry with a teacher, a form of dharma in which knowing does not transit into action. Although ritual and meditation are, thus, not the direct means of liberation, they remain so vicariously, through their causal contribution in the arising of knowledge. 50  In the center of such inquiry were the “identity statements” of the Upaniṣads, statements such as tat tvam asi, “You are that,” and ahaṁ brahm smi, “I am Brahman.” These form the topic of Chapter Eight, not in isolation, but as principles through which the Upaniṣadic propositional knowledge of Brahman and the Self could be teleologically organized for the attaining of liberation. I begin by elucidating Śa kara’s classification of general types of Upaniṣadic statements, before moving to considering his Brahma-Sūtra-Bh ṣya on 4.1.2 as the frame through which the doctrine of identity statements can be fully reconstructed. Traversing the two entities juxtaposed in the identity statements, Brahman and the Self, we see how the doctrine of the identity statements was, in fact, Śa kara’s direct intervention in the prasaṅkhy na discourse: that is, how a reading of the Upaniṣads was possible such that Brahman would not be a composite, relational entity; and, how placing the onus on the identity statements rather than the Upaniṣadic descriptions of Brahman and the Self would provide the requisite immediacy of understanding that, otherwise, meditation was required for. In the last chapter of Part Three, I move to considering the two kinds of inquiries into the meaning of the identity statements, theological ( ravaṇa) and philosophical (manana), as the direct means of liberation. I show that, to Śa kara’s mind, the purification that was supposed to happen through meditation in the doctrine of prasaṅkhy na had to be attained before one could inquire into Brahman: liberation was a result just of intellectual understanding, through the two inquiries, by means of which the scriptural cognition “I am Brahman” gotten from the identity statements obtains the characteristic of certainty. In other words, whereas the proponents of meditation claimed that even the intellectually indubitable but propositional cognition of Brahman was not sufficient for liberation because of personal impurity, Śa kara’s direct process of liberation began after such purity had been attained: thus, when one had fully understood the 51 meaning of the identity statement through the theological inquiry, and attained personal experience through the philosophical inquiry, liberation had to follow as a matter of course. I conclude the chapter by considering the question of meditation in Śa kara’s system, and show that although Śa kara affirmed its soteriological value as a process that could follow the philosophical inquiry (meditation, thus, as the third Vedāntic process, nididhy sana), as a kind of awareness this meditation (which he called parisaṅkhy na) was not mental absorption or concentration, but an intellectual or analytic process of reduction or dissociation from all possible points of identification for the Self, such that the Self would be known not as an object, but as the light of awareness that is the irreducible residue without which no reduction would be possible. In other words, I show that Śa kara replaced Yogic meditation (prasaṅkhy na) with characteristically Sā khyan reflection (parisaṅkhy na): not meditation on Brahman so as to obtain the insight into “I am Brahman,” but a removal of everything that I am not so that I could not but be Brahman. This was his characteristic contribution to Vedānta soteriology. Finally, in Part Four and Chapter Ten, we will see the Advaita doctrine of liberation which was focused on the Upaniṣadic identity statements morph into the theory of mah -v kya. I conclude the dissertation there by showing that: (1) a theory of Upaniṣadic mah -v kyas was for the first time explicated not by Śa kara, but by the 10-11th-century Vedāntin Sarvajñātmanś (2) an explicit but not theorized notion of mah -v kya was already present in Mīmāṁsā, where it did not stand for short Upaniṣadic statements, but for larger textual unitsś (3) Sarvajñātman modeled the notion of Upaniṣadic mah -v kyas on the Mīmāṁsā blueprint, fitting Vedānta building blocks into a Mīmāṁsā structureś (4) the Upaniṣadic identity statements were only formally short: they were cryptic ellipses whose elaboration required the full Upanishadic corpus, and through the principle of mediate causality could include even the ritual sections of the Veda; (5) 52 the key element in mah -v kya, both in the Mīmāṁsā blueprint and the Vedanta adaptation, was finality of meaning. Note on the Title I should like to end this Introduction with a note on the title. It is inspired by the definition of interpretation by the French philosopher Paul Ricoeur. Ricoeur claimed that interpretation of any kind, whether of dreams in psychoanalysis, of sacred texts in exegesis, of myths and rituals in the history of religions, even of life itself, had a certain “architecture of meaning,” which was that of “double” or “multiple” meanings. We may illustrate this with the interpretation of dreams. A dream has a direct meaning of whatever its content is, but for psychoanalysis that content is a symbol of some repressed desire, and the task of psychoanalysis as a form of interpretation is to understand what specific desire the symbol stands for. Ricoeur defines a symbol as “any structure of signification in which a direct, primary, literal meaning designates, in addition, another meaning which is indirect, secondary, and figurative and which can be apprehended only through the first.”81 Two elements are important in the definition. First, the symbol designates another meaning in addition to its direct signification, and second, the secondary signification can be apprehended only through the first. Such is the case because some meanings are necessarily hidden—otherwise they would not require interpretation—but they can be deciphered only because they are commonly associated with symbols. The role of interpretation, then, is to “translate” the symbol to an explicit meaning within the frame of reference of a specific discipline that concerns an aspect of human existence. “Interpretation, we will say, is 81 Ricoeur 1978:98. 53 the work of thought which consists in deciphering the hidden meaning in the apparent meaning, in unfolding the levels of implied in the literal meaning.”82 This, I believe, captures quite precisely Śa kara’s understanding of the Upaniṣadic identity statements. While grammarians and philosophers before him have dissected the modes of sentence meaning and the various signification functions of words, it was Śa kara who claimed that the identity statements could be meaningful only if the two collocated categories denoted their primary meaning, and in addition to it, designated another, secondary and figurative meaning. It was in that liminal space between the literal and the figurative meaning, fine as a “gnat’s wing,” that new knowledge appeared that made the attainment of liberation possible, when it dawned on one that I myself, known to me most intimately, in fact am that great ground of Being out there: that is, that this Being is not out there at all, but is me. Further, this surplus of meaning that appears when Brahman and the Self stand for the inner Brahman in addition to their individual direct meaning was the only means through which the inner Brahman could be known. It was so because this inner Brahman—but not Brahman as such or the Self as such—was a point “from which words return along with the mind, failing to reach it.” It was hidden to words, the identity statements were its symbol, and therefore it was essentially in the domain of interpretation, that is, accessible solely through words. The prize when this inner Brahman was properly interpreted was liberationś therefore, “deciphering the hidden meaning” and the “hermeneutics of liberation.” 82 Ibid. 54  PART ONE: CIRCUMSCRIBING THE FIELD OF VEDIC THEOLOGY CHAPTER ONE: RETHINKING THE IDEA OF SCRIPTURE IN VEDIC THEOLOGY Scripture should not be much doubted, being more reliable than the words of one’s mother and father, for one cognizes through scripture personally. It occupies the same rank as the senses.1 Introduction The purpose of this chapter for the global goals of the dissertation is to delimit the space of Vedic theology which Śa kara inhabited, and thus demarcate the field of his interlocutors. I said in the Introduction that the sense in which I take Vedānta to have been uttara or posterior to Mīmāṁsā was the commitment to the Veda as a source of knowing from linguistic utterances that are considered non-testimonial in character, a commitment which was not shared by other theologies in the same intellectual space. As is well known, the two branches of Vedic theology, the Pūrva and Uttara Mīmāṁsā, had two distinct concerns stated at the beginning of their respective canonical texts—dharma or ritual and Brahman or the great Being from which everything proceeds—yet they shared the conviction that the Veda in all its parts possesses epistemic validity and is the sole reliable warrant for all supersensible things. Whatever course of development the relation between the two brands of Vedic theology may have initially had, by the time of Śa kara and his near contemporary Kumārila, the idea of a single canon which includes the Upaniṣads was firmly established.2 1 śāstraṁ cānatiśa kyaṁ pit -māt -vacanād pramā ataram. svayaṁ hi tena pratyeti, indriya-sthānīyaṁ hi tat. MSŚBh 4.1.3, IV.1199. 2 TV on 1.2.7 is, for instance, an attempt to include everything in the Veda in a single canon, organized under the injunctive suffix of the optative. That includes the descriptive statements of the Upaniṣads, as I will show in the next chapter. 55  Additionally, Mīmāṁsakas and Vedāntins shared the commitment to the Veda not only as a reliable warrant, but as the sole means of attaining the highest good as well. This second commitment will take our attention in the rest of the dissertation. Here I want to focus on the epistemic validity of the Vedas constructed around three key ideas, ruti, stra and apauruṣeyatva, a complex which forms the connective tissue of Vedic theology. Understanding this complex is necessary not only to set the scope of the field, but also to grasp properly how the relation of scripture with the highest good was conceived. The key among these three ideas is that of ruti, which is so common in scholarly literature but so little understood that I will preface my account with a short review. The Classical Theory of Śruti ruti is one of those terms which students of Indian philosophy (and Indologists of all walks) learn early on in their carriers in a certain meaning or a set of related meanings—one which perpetuates itself through referencing, cross-referencing and no referencing—without seriously questioning how well this meaning maps on to the texts they are studying. “Revelation” or “that which is/was heard” are the most common meaning candidates, referring to the whole corpus of the Veda, from the Saṁhitās to the Upaniṣads. The following is a sample survey taken from studies which are dedicated specifically to ruti, studies where it plays an important role, or studies of Advaita Vedānta in general.3 Purusottama Bilimoria vacillates between “heard word,” “scriptural words,” “the revealed word” (“as it may be termed in western theology”), “heard word – of scripture,” insisting all the while in each case that this is what ruti “literally” means.4 Anantanand 3 Pollock 1997 gives a similar survey of its uses, along with those of smṛti, in more general Indological literature. 4 1988:5, 7, 17, 20. 56 Rambachan has just the bracketed “that which is heard” once, with an occasional addition of “Vedic revelation,” in a study which is all about Śa kara’s understanding of ruti.5 In another study he saysŚ “To emphasize the fact that the Vedas were transmitted orally from teacher to student, the texts are collectively referred to as ruti (that which is heard).”6 Ashok Aklujkar has “that which is heard or revealed,”7 and Nataliya Isayeva similarly has “[ ]ruti (lit.: heard), that is, eternally heard or communicated, actually, a revelation.”8 A Monograph by Kotta Satchidananda Murty is titled “Revelation and Reason in Advaita Vedānta,” and a festschrift for the famous neo-Vedāntin scholar T.R.V. Murti bears the title “Revelation in Indian Thought.”9 Wilhelm Halbfass’s important study “Tradition and Reflection” hasŚ “[ ]ruti Vedic-Upaniṣadic ‘revelation.’”10 David CarpenterŚ “Veda was also equated with ruti, literally "hearing" and referring to what has been heard, that is, revealed.”11 Some studies do not even bother to define the term12 or opt to use the simple, uninformative “scriptural texts.”13 Always paired with ruti is smṛti, standing broadly for the rest of canonical books of Hinduism, taken as “that which is remembered” or tradition. It also bears mentioning that quite often introductory textbooks of Hinduism treat ruti as revelation and as one of the key criteria of what being Hindu means. Gavin Flood’s widely used “An Introduction to Hinduism,” for instance, places ruti under “General features of Hinduism,” 5 1991:4, 46. 6 2006:117. 7 1991:3; this is, otherwise, a very rich and valuable study. 8 1993:31. 9 Murty 1959; Coward and Sivaram 1977. 10 1991:132. 11 1992:25. 12 For instance, Adluri 2015. 13 For instance, Hirst 2005:9. 57 describes it as “revelation,” identifies it with the Veda and associates it with the Vedic sages, to whom it was revealed and who passed it along in oral transmission.14 If we make a composite account of this, ruti would signify the Veda in some collective sense, revealed at some point in history and orally transmitted. A full-fledged version of this account was given by Barbara Holdrege in her important study “Veda and Torah.” Holdrege proposed that ruti refers to that which was heard by the Vedic ṛṣis, “’recorded’ through the vehicle of their speech and assumed a concrete form on earth as the recited texts of the mantras.”15 While the term may be interpreted to mean “that which is heard” in oral transmission, “it is also clear that the related term ruta was used as early as the Ṛg-Veda to refer to cognitions of the ṛṣis and that the term ruti itself still retains this association among contemporary Indian thinkers: Veda as ruti is ‘that which was heard’ by the ancient ṛṣis as part of a primordial cognition in the beginning of creation.”16 We should note that Holdrege proposes cognition instead of revelation. The ṛṣis intuit, see and hear, the mantras. Holdrege’s thesis was inspired by the work of the Vedic studies scholar Jan Gonda, who proposed that the meaning of the Sanskrit noun dh ḥ and its root siblings such as dh tiḥ and dhy na, in the Rig Veda but elsewhere as well, express the idea of seeing, vision, “exceptional and supranormal faculty, proper to ‘seers’ … the faculty of acquiring a sudden knowledge of the truth, of the functions and influence of the divine powers, of man’s relations to them etc. etc. It is this ‘vision’ which they [the seers] attempt to give shape, to put into words, to develop into intelligible speech, to ‘translate’ into stanzas and ‘hymns’ of liturgical value.”17 Gonda did not 14 Flood 1997:12. 15 1996:25. 16 Holdrege 1996:9. 17 1963:68-9. 58 associate these visions with ruti, but Holdrege claimed that even in the Rig Veda “the oral-aural dimensions of Vedic cognition are also emphasized by the ṛṣis,” although less than the visual.18 She offered, unfortunately, only one explicit reference in support of the claim.19 Holdrege’s account has the merit of spelling out clearly and with conviction what most scholars think when they read, hear or write “ ruti” and for trying to get to the bottom of it by pursuing what we may call “the classical theory of ruti.” The locus classicus of the theory is a famous passage from Yāska’s NiruktaŚ “The seers had directly seen dharma. Through instruction they handed down mantras to the later generations, who did not see it directly. The later generations, losing the power of instruction, handed down from memory this composition, and the Veda and the auxiliary Vedic sciences, for the purpose of keeping ahold of its likeness.”20 The complexities of understanding this passage aside, particularly what s kṣ t-kṛta and dharman mean,21 the ṛṣis in the passage are not described as hearing, but seeing dharma. Further in the Nirukta they are described as seers of the Vedic mantras, presumably referring to the Saṁhit portion of the VedaŚ “The seers have visions of the mantras.”22 Sheldon Pollock cites another 18 Holdrege 1996:229. 19 It is RV 8.59.6, quoted bellow with Jamison’s and Brereton’s translationŚ índrāvaru ā yád r̥ṣíbhyo manīṣā́ ṁ vācó matíṁ śrutám adattam ágre yā́ ni sthā́ nāni asr̥janta dhī́rā yajñáṁ tanvānā́ s tápasābhy àpaśyam “O Indra and Varu a, in that you in the beginning gave to the seers inspired thought, the thinking of speech, what is heard— the insightful (seers) launched these as poems as they stretched out the sacrifice. I looked upon them with fervor.” 20 sākṣāt-k ta-dharmā a ṣayo babhūvuḥ. te ‘varebhyo ‘sākṣāt-k ta-dharmabhya upadeśena mantrān samprāduḥ. upadeśāya glāyanto ‘vare bilma-graha āyemaṁ granthaṁ samāmnāsiṣur vedaṁ ca vedā gāni ca. 1.20.2, p.90. 21 They are exhaustively discussed in Aklujkar 2009. In translating bilma-graha āya with “for the purpose of keeping ahold of its likeness,” I take clue from Yāska’s alternative gloss of bilma as bhāsana. 22 evam … ṣī āṁ mantra-d ṣ ayo bhavanti. 7.3.8, p.549. 59 passage from Pā ini which talks about the s man chants being seen by ṛṣis, and concludesŚ “The ṛṣis are not normally said to have ‘heard’ mantras.”23 Obviously, such passages can be related to ruti only indirectly. They certainly support the idea that “the Veda” was early on seen as being revealed to or cognized by the seers, but there is no straightforward route of relating this understanding of revelation or cognition to ruti. Claiming that the idea of revelation or cognition must be the origin of the idea of ruti because such understanding “still persists today” is reading backwards. The conceptual difficulties of the classical theory aside, to which, as Pollock says, “long acquaintance and acquiescence may have inured us,”24 the textual evidence has just not been produced. Two notes before we proceed. First, what the Vedic rṣis see in this account are the hymns of the Vedic Saṁhit s, so we would expect that such hymns be the first referent of the word ruti in its use, at least in an important restricted sense. Second, as far as etymology is concerned, there is not much justification for translating ruti as “that which was heard.” The problem with this is that ruti is not a participle but, commonly, an action noun.25 ruti is, to use Bilimoria’s favorite word, “literally” hearing, an audition, just as dṛṣṭi is seeing or vision.26 Another meaning of the word is the instrument of hearing, either the sense in general or the ear as its seat. This was proposed by a v rttika of Kātyāyana under Aṣṭ dhy y 3.3.94 and is well attested.27 23 Pollock 1997:399. The sūtra is 4.2.7, dṛṣṭaṁ s ma, and is explaining the derivation of adjectives formed with the taddhita affix aṆ applied to ṛṣis who have seen s man chants, vasiṣ ha -> vāsiṣ haṁ sāma = vasiṣ hena d ṣ aṁ sāma. See Kartre 1989:403 and Sharma 1999:158-9. 24 Pollock 1997:400. 25 See Abhyankar 1986:130, under ktinŚ “k t affix ti added to roots to form nouns in the sense of verb-action; e.g. kṛtiḥ, sthitiḥ, matiḥ etc.” And 396, under rutiŚ “(1) lit. hearing, sound …ś perception, as a proof contrasted with inference …ś (2) the authoritative word.” Whitney 1993Ś432Ś “feminine nouns of action.” 26 Holdrege’s “cognition” is, obviously, much closer in form than “that which was heard.” 27 śruti-jiṣi-stubhyaḥ kara e ktin vaktavyaḥ. śruyate anayā śrutiḥ. See also Sharma 2002.3:528-9. 60  Śruti in Early Vedic Theology To Pollock goes the credit of realizing how ruti and the related smṛti float on thin air and how such a set of categories “basic to the formation and self-understanding of Sanskrit culture … has been misunderstood, or at least never clearly explained, in Western (and westernized) Indology.”28 While Pollock is primarily interested in the construction of the two categories as an elementary form of ideological power in classical Sanskrit culture, and I want to understand the epistemological significance of ruti, the starting point of his discussion and the origin to which he traces the divide provides the frame through which the connotations of ruti can be investigated: it is the association of ruti and smṛti with another pair, pratyakṣa and anum na or perception and inference. This association has roots in early Vedic theology and is evident already in the Dharma- Sūtras, which talk about express Vedic statements that trump in validity customs which are only inferably Vedic. The relevant distinction which the Dhārma-Sūtras draw is between practices based on evident statements, presumably those which are actually recited as part of Vedic study and known to the addressee from hearing the Veda, and practices that are customary but that have no clear warrant in extant injunctions, so that there are or were such injunctions must be inferred. The problem with the inferable rules is the intervening human factor, namely that the justification for the practices is not found in the Vedas as they are known, but only in the conduct of Vedic men, and men—as MS 4.1.2 claims—have pleasure as motive inherently and universally. Since there are no direct legitimizing statements for such practices, potentially at least it is always possible that they may turn out to be unjustified. 28 Pollock 1997:395-6. 61  The context allows for an equivalence to be drawn between ruti, pratyakṣa-vidh na, br hmaṇokt vidhayaḥ and pratyakṣa- rutiŚ they are all juxtaposed to rules whose sources are inferable, and to inference in general. rutis are, then, direct rules or injunctions found in the Brāhma as. Customs for which such ruti rules are in evidence are a safe bet and cannot be trumped by practices where the rules are only inferable.29 We should note that, whatever the full initial scope of ruti may have been, what the term covers here is practices based on scriptural rules. The distinction which is drawn is not between the Vedas on one hand and other scriptures on the other, but between practices legitimized by rules that are found somewhere in the Vedas and practices which lack such legitimacy. We should also note the privileged status of that which is expressed in rutiś it is this, it seems to me, that Pollock has in mind when he talks about the category as “an elementary form of ideological power.” It is an instrument of legitimizing what one does and believes, and it may at least provisionally point to something with a scope which is initially restricted. If this is, indeed, the first place where ruti appears in a context which allows for setting some theoretical frame, it is also mightily significant that the context is one of legitimizing action. Let us also note here that the Dharma-Sūtras do not use the term smṛti in this context, but talk about inferable rules and lost Brāhma a texts. David Brick has argued that originally smṛti 29 śrutir hi balīyasy ānumānikād ācārāt. drśyate cāpi prav tti-kāra am. prītir hy upalabhyate. “[F]or a Vedic text has a greater force than a practice for which a vedic text has to be inferred. We notice here, moreover, a motive for such a practice, for one derives pleasure from it.” ĀDhS 1.4.8-10 brāhma oktā vidhayas teṣāṁ utsannāḥ pā hāḥ prayogād anumīyante. yatra tu prīty-upalabdhitaḥ prav ttir na tatra śāstram asti. “All rites are described in the Brāhma as. The lost Brāhma a passages relating to some of them are inferred from usage. When a practice is undertaken because of the pleasure derived from it, it does not presuppose a vedic text.” ĀDhS 1.12.10-11 ekāśrayaṁ cācaryāḥ pratyakṣa-vidhānād gārhasthyasya gārhasthasyaŚ “There is, however, only a single order of life, the Teachers maintain, because the householder’s state alone is prescribed in express vedic texts.” GDhS 3.36 dharme ādhigato yeṣāṁ veda sa-parib ṁha aḥ. śiṣ ās tad-anumāna-jñāḥ śruti-pratyakṣa-hetavaḥ. “Cultured people are those who have studied the Veda together with its supplements in accordance with the Law, know how to draw inferences from them, and are able to adduce as proofs express vedic texts.” BDhS 1.6. All translations Olivelle. VDhS 6.43 is similar and seems to be based on this. See also Olivelle 2000:16. 62 did not refer to textual rules, but to traditional time-honored norms, comparable, for instance, to the modern rule that the father of the bride should pay for her wedding. As a distinct textual corpus, it initially denoted the Dharma-Śāstras, which develop as a result of a growing tendency towards versification in the Dharma-Sūtras, where verses mostly state traditional norms with which the listener would have been long familiar and in whose ears such norms would have been sacrosanct.30 A similar doctrine is found in the third p da of the first adhy ya of MS, particularly 1.3.1-4, where the question is what to do with that dharma or ritual action which is not based on the words of the Veda, but is practiced by the same good Vedic men that practice explicitly Vedic ritual. It should be accepted as valid, because the fact that the practitioners are the same allows for an inference. In case, however, it contradicts the Veda, it should be discarded. The section does not talk about the pursuit of pleasure as a possible reason for invalidity, but sūtra 1.3.4 is interpreted by Śabara in that way.31 Neither ruti nor smṛti are explicitly mentioned, and the terms used are abda and a abda, that dharma for which there is a word and that for which there isn’t, but the association of ruti with pratyakṣa is otherwise common in MS, and in sūtra 1.4.14 we find the same pratyakṣa-vidh na as in GDhS 3.36.32 We may conclude that an understanding must have been common to the intellectual space of early Vedic theology that there are direct Vedic injunctions, actually known because of being part of Vedic study, and that there are customs that whose origin cannot be related to an express Vedic statement; since both 30 Brick 2006. 31 (Pūrvapakṣa:) dharmasya śabda-mūlatvād aśabdam anapekṣaṁ syāt. “Since dharma is based on the word [of the Veda], that which is not based on the word of the Veda should be disregarded.”. (SiddhāntaŚ) api vā kart -sāmānyāt pramā am anumānaṁ syāt. “Rather, because the agents are the same, it [a word not based on the Veda] is valid, since there would be an inference.” virodhe tv anapekṣaṁ syāt, asati hy anumānam. “But in case of a conflict, it would be disregarded, for inference is there only when there isn’t [a conflict].” hetu-darśanāc ca. “And, because a reason is seen.” MS 1.3.1-4. 32 See also Clooney 1990:121. 63 are practiced by Vedic men, the customs which are not supported by express Vedic texts must have origin in textual rules and injunctions which are not accessible; such practices are good, as long as they do not contradict something in the express statements. abara’s Understanding of Śruti Śabara’s commentary on the mentioned p da provides a broad theoretical ground for conceptualizing smṛti, since smṛti it is taken precisely in the sense of memory of something previously experienced or seen, a prior cognition (pūrva-vijñ na), but then it is applied strictly in regard to memories the reliability of which is in question because the source of the prior cognition, in this case an explicit Vedic composition, is no longer evident. 33 The reliability of such memories is secured on inferential grounds: if good Vedic folks who do the Vedic stuff do this other thing as well, we can infer that those who had instituted the practice were aware of a Vedic text that is not known anymore. Śabara, in other words, still deals with particular actions of diverse types, some of which are rituals that are found in the Dharma-Śāstras, the paradigmatic case being the aṣṭak rite, while others are practices which haven’t found scriptural expression, for instance digging up wells and setting up drinking fountains.34 He does distinguish between the two, using strictly smṛti for the first and conduct or c ra for the second,35 but the underlying principle behind both is that of memory, smṛti or smaraṇa.36 33 The objection in the MSŚBh 1.3.1, I.161 is precisely to the effect of impossibility of memory of something not experienced or heard (na hy ananubhūto 'śruto vārthaḥ smaryate), which would cover all grounds of memory, worldly and Vedic. 34 This is also in the MSŚBh 1.3.1. 35 These two, of course, have a long history in the Dharma literature. See Brick 2006 for some discussion. 36 MSŚBh 1.3.1, I.159 in which he introduces five instances of such non-Vedic action readsŚ “Now, then, where we do not perceive a Vedic word and they remember, ‘This practice should be observed in this manner, for this purpose,’ would this be in the same way [valid like the Veda] or not?” athedānīṁ yatra na vaidikaṁ śabdam upalabhemahi atha ca smaranty evam ayam artho 'nuṣ hātavyaḥ etasmai ca prayojanāyeti. kim asau tathaiva syān na veti. It follows that memory is a common feature of all such acceptable Vedic practices, whether they have been smritified or not. Cf. also MSŚBh 1.3.15, I.243Ś “It is accepted that the validity of memory and conduct is from inferenceś” anumānāt sm ter ācārā āṁ ca prāmā yam iṣyate. 64  Although under 1.3.1 Śabara mentions the perceptibility of the composition as opposed to the general principle of memory,37 ruti is throughout this section paired with smṛti in the restricted sense and—we need to note this crucial point very carefully—Śabara does not have “the Veda” in mind, but single statements in the Veda, vaidika-vacana, almost invariantly in the Brāhma as, expressing rules or injunctions. This is particularly clear under the comment on MS 1.3.3, where ruti is for the first time introduced in the discussion, after the principle of Veda based on the word and smṛti on memory = anum na is discussed under the first adhikaraṇa, sūtras 1.3.1-2. ruti-virodha or contradiction to ruti is a contradiction to an express Vedic injunction. A specific rule needs a specific counter-rule to contradict, and it is here that ruti appears, not in the discussion of the principle. The principle of memory does not apply where it contradicts a specific ruti, or where some ulterior motive can be detected; in both these cases the source the memory of which the rule is would be something else, for instance greed for an honorarium. Śabara clearly does not presuppose just anything found in the Veda when he uses ruti in pair with smṛti, but—on the evidence we have here—specifically those sections/statements which regulate ritual and rites. Let us take stock of this and see how ruti is used elsewhere in the Bh ṣya.38 Apart from smṛti, ruti is paired in the MS and the Bh ṣya with several other counterparts or sets, invariantly as something specific in the Veda and revolving around the direct meaning of 37 pratyakṣe opalabdhatvād granthasya. MSŚBh 1.3.1, I.64. 38 Needless to say, there is no whiff of “tradition” in Śabara’s conceptualization of smṛti. The ground is fully epistemic and draws on the deliberations about the reliability of memory in general, which Śabara affirmsŚ “Memory is a reliable warrant, because it is a cognitionś” pramā aṁ sm tiḥ. vijñānaṁ hi tat. MSŚBh 1.3.2, I.163. There is no need for our purposes to consider Kumārila’s rethinking of the rutiṭsmṛti relationship (see McCrea 2010, Yoshimizu 2012.). Suffice it to say that Kumārila continues to use ruti in the sense of a single sentence in the Veda, of the nature of an injunction. In his canonization of smṛti as consisting of the Dharma-Śāstras and in establishing the validity of the other books and sciences that Vedic people accept by tracing their root in direct experience, mundane or Vedic, he is careful to be either vague and use veda or specific to use vidhiṭ ruti, mantra, arthav da etc. 65 what is said. This primarily means a full Vedic statement which enjoins something, either the principal sacrifice or some of the details related to it.39 Such statements are direct because the meaning of what is predicated in them—the ritual or its detail—involves the literal meaning of the word, one which imposes on one’s understanding just from hearing. That this is so is clear from the fact that this literal meaning of the word, or of a simpler linguistic element such as the contributory factors to action (k raka), for instance the instrument, expressed through the case endings, or the number also expressed through the ending, is also called ruti.40 We need to note this point well: a ruti or an express statement is one in which what is predicated, a single thing, is comprehended just from hearing, before any process of humanly contingent interpretation can begin. It is hearing a specific word in its direct, most common meaning. A sentence in which this most common meaning of the predicate is understood is rutiṬ Śabara’s definition isŚ “The denotation of a thing which is understood just from hearing the word is understood by rutiṬ ruti is hearing.”41 We note, then, this factor as crucial: direct meaning of what is said, got from hearing before interpretation. I should like to note here that it is the conscious, purposeful, textual interpretation that is meant, one which seeks to obtain a meaning different than the literal. Mīmāṁsakas did not, of course, mean that one is born with a full understanding of language and 39 The instances are too numerous to cite; an important comment is on MS 1.3.3, in the above discussed section, where Śabara consistently uses ruti as the term for the direct text to which inference is juxtaposed; other instances include 1.2.24, 1.3.18 (pratyakṣ rutiḥ), 1.4.2, 1.4.14 (pratyakṣa-vidh na glossed as pratyakṣa- ruti), 2.1.20, 2.3.13, 3.3.11. 40 This is obvious from Śabara’s comment on 3.1.12. I translate k raka as “contributory factor to action” following Matilal 1990. Cf. also MSŚBh 3.1.13Ś “Because of the use with an express statement. These categories are used in relation to hearing in the singular. There is only one thing here which is heard. And, in regard to action in which the word is the means of knowing, whatever the word says, it is that which the reliable warrant is.” śruti-samyogāt, ekatva-śruti-saṁyuktā ete padārthāḥ, ekaṁ hi dravyam eṣa śrūyate, śabda-lakṣa e ca hi karma i yac chabda āha, tad asmākaṁ pramā am. 41 yad arthasyābhidhānaṁ śabdasya śrava a-mātrād evāvagamyate, sa śrutyāvagamyate. śrava aṁ śrutiḥ. MSŚBh 3.3.14, I2.825. 66 can comprehend words “as they jump upon us.” They did believe, however, that words have common, single references, as we shall see a bit later, and when one says “stick,” this in the direct meaning, as heard, does not mean a gram of weed or something similar. Mīmāṁsakas, further, did not deny the linguistic determination of the data of experience which shapes any conceptualized perception one may have.42 It is in dealing with the text of the Veda where interpretation becomes important, as a process one performs to obtain a meaning from a sentence or a text which is not the one that jumps upon the competent reader upon hearing. This may be as innocent as determining how several sentences fit together, or as grave as understanding a particular word in a sense different than the one “got on hearing.” Crucially, Śabara identified the direct meaning of a Vedic text with pratyakṣa, while all forms of interpretation he marked as parokṣa, indirect, or num nika, inferential.43 One of such forms of interpretation is ascertaining the principal-subordinate relationship between parts of the ritual performance (more on which in the next chapter), for which six criteria (called pram ṇa) are delineated: (1) ruti or express meaning; (2) liṅga or implied meaning; (3) v kya or syntactical relation; (4) prakaraṇa or context; (5) krama or sequence; and (6) sam khy or technical term.44 There is no need here to analyze them, but we should note that they have successive validity, each of them more valid than the one following, but it is only ruti which has direct, unconditional validity, for two reasons. First, in ruti the sentence-reference relationship is direct (sannikṛṣṭaḥ ruty-arthaḥ); what is said is what is meant—“go get the 42 See Taber 2005, particularly pp.93-148. As we shall see later, the words of the Veda have been directly identified by Śa kara as creative, providing the natural categories under which things in the world are classified. 43 See on MS 2.3.18, II.605Ś “The sentential unity of the cooked rice with deposition is evident, while the relation with the sentence about the divinity is inferential, not evident. ‘He deposits the cooked rice’ is an express statement. ‘He deposits that which relates to B haspati’ is inferential.” upadadhātinā cāsya pratyakṣam ekavākyatvam, parokṣaṁ devatā-vacanenānumeyam. carum upadadhātīti hi pratyakṣaṁ vākyam. bārhaspatyam upadadhāti ity ānumānikam. The distinction is very common in the Bh ṣya. 44 These are given in the MS 3.3.14; this and the following paragraph are based on Śabara’s comment on this sūtra. See also Jha 1964:247-57; Patton 2005:69-70. 67 hammer” is taken to mean just that, not “I want to smack you with it,” a meaning possible but distant. In the Veda one does not have a license for deriving implicit meaning which is contrary to the direct meaning because, second, only ruti is properly Vedic, equivalent to the Vedic word. The other five are not so, ergo they are human, and are valid so long they are ruti-mūla, rooted in an express Vedic statement.45 They cannot undermine something expressly predicated by the Veda, because in that way they would be undermining themselves. Whatever concordance is achieved by the other five is valid, but—here it is again—inferential ( num nik ekav kyat ), and so it cannot be stronger than or optional to the express statement. To give one instance of this, there is the statement aindry g rhapatyam upatiṣṭhate, “One should attend to the domestic fire with the mantra to Indra.” There is also near to it a mantra in which Indra is celebrated. The proximity of the two texts raises the questionŚ “Should one attend to the domestic fire itself with the mantra, or to Indra who is somehow related to the fire?” In trying to answer the question, we note that the mantra only praises Indra and does not contain an express statement to the effect that Indra is the reference of the whole sentence. 46 In fact, only the first line of the mantra is quoted, where Indra’s name is not even mentioned (kad cana star r asi). That it is a mantra for Indra can only be arrived at by supposing that the mantra indicates that, perhaps through calling to mind the full mantra first, with the assumption 45 Cf. MSŚBh 3.3.14, II.844-45Ś “It has been saidŚ ‘Since dharma is rooted in the word, what is not the word should be disregarded’ (sūtra 1.3.1). Both of the context and implied meaning are not the word; and in regard to things beyond the range of perception, there is no means of cognition except the Veda.” uktam evaitad dharmasya śabda- mūlatvād aśabdam anapekṣaṁ syād iti. yad etat prakara aṁ li gaṁ ca ubhayam apy etad aśabdam. na cātikrānta- pratyakṣa-viṣaya evaṁ-lakṣa ake arthe śabdam antare a paricchedo 'vakalpate. 46 The two are put together in the Taittir ya-Saṁhit 1.5.8.3-4. The mantra itself is from the Rig Veda 8.51.7: kadā́ caná starī́r asi néndra saścasi dāśúṣe | úpopén nú maghavan bhū́ya ín nú te dā́naṁ devásya pr̥cyate || “Never are you a barren cow, nor, Indra, do you go dry for the pious man. Over and over, more and more, the gift coming from you, the god, becomes engorged.” (Jamison and Brereton 2014:1134) 68 that otherwise its recitation would be without a purpose. This having been obtained through indication, it does not have the power to override the express statement that the domestic fire should be worshipped. In fact, that Indra is addressed in the hymn is understood through the express statement, which mentions a mantra for Indra, and the fact that this specific mantra is in its proximity and is available for employment. This makes the inferential operation of indication ruti-mūla, rooted in something express. We hear the word aindry in the rutiś we see a mantra in proximity; we infer that the devat of the mantra is Indra, for a mantra without a devat is of no use.47 Another case of a set of six interpretative tools is the one determining the order of the ritual performance, again called pram ṇas, and ruti again has the precedence against the other five because it is direct.48 And, crucially, ruti is paired with and opposed to lakṣaṇ or figurative meaning, and it trumps here as well.49 In both cases, the counterparts to ruti are called inferential. We do not need to clutter the text with more instances, because a clear pattern emerges. ruti is that in the Veda which one gets “on hearing,” and what has direct efficacy and authority as a pram ṇa. The form of instantiations of ruti throughout the pairings is consistent, namely particular express Vedic statements, in the “smṛti-chapter” of the MS and elsewhere. That allows 47 “When mantra is near an act in the context which requires details of procedure, the mantra, in virtue of being in the text forms a verbal supplement to the act which requires details, ‘Sacrifice with this mantra.’ That is, a result being expected from the sacrifice, one should assist the sacrifice with this mantra. The mantra in question cannot assist unless Indra is its reference. So, it means that the mantra denotes Indra. This denotation is rooted in ruti.” itikartavyatārthinaḥ prakara avato 'rthasya sannidhāv upanipatito mantra āmnāna-sāmarthyād itikartavyatākā kṣasya vākya-śeṣatām abhyupetyaitena mantre a yajeteti. kim uktaṁ bhavati. yāgena abhīpsite sādhyamāne 'nena mantre opakuryād iti. na cāntare a indrābhidhānam ayaṁ mantra upakartuṁ śaknoti. tenaitad uktaṁ bhavati. anenendro 'bhidhātavya iti. ataḥ śruti-mūla evāyam arthaḥ. ... “This meaning that 'Indra should be worshipped with this mantra' is got from the direct statement, but relative to the inferential sentential unity and the force of indication (in the mantra).” yatas tu khalv ānumānikīm ekavākyatāṁ li ga-sāmarthyaṁ cāpekṣya śrauto 'yam artho yad indrasyopasthānam anena mantre ety avagamyate. MSŚBh 3.3.14, I.288-9. 48 See Jha 1964:263-8. 49 śrutiś ca lakṣa āyā jyāyasīti, MSŚBh 1.2.19, I.134ś śruti-lakṣa ā-viśaye ca śrutir jyāyasī, MSŚBh 1.4.2, I.324. 69 us to conclude that for Śabara, when the Veda states something directly and expressly, that is ruti and it is pratyakṣa, perceptible. All forms of human involvement requiring understanding which is arrived at by something more than the mere hearing of the express text is inferential, num nika. This is clearly an extension of the familiar express vs. inferred rules/statements distinction that we’ve found in the Dharma-Sūtras and that resulted in the rutiṭsmṛti classification. Whether it is historically later, we are not in a position to say. Most of what I found in the Bh ṣya of Śabara certainly goes back to the sūtras.50 We ought to see, though, a generalization of an idea: Veda, associated with hearing, associated with literal meaning, associated with direct authority, vs. memory/hearing with an afterthought, associated with human interpretation, associated with roots in hearing, associated with dependent authority. Because of the prominence of literal meaning, ruti constitutes the ideal form of Vedic evidence. It is a pram ṇa, existing in sets with other pram ṇas which operate strictly in the Vedas (vede vs. loke), where it trumps all others in all sets because it is direct. It is the perceptual in the Veda, on which all other hang, completely comparable to the status of sensory perception and its relation to the other reliable warrants in worldly matters (as we will see in the next section). In all the instances that I’ve inspected in the Bh ṣya, the passages so characterized are equivalent to Vedic injunctions (vidhiṭcodan ). The scope of the Veda was broader than this, and in some of its texts, called arthav da and containing mostly stories, Mīmāṁsakas allowed for 50 The six pram ṇas of ascertaining the principal-subsidiary relation are a matter of the third p da of the third adhy ya. Particularly sūtra 3.3.14 is significant. A matter concerning an application of a ritual modeled on a prototypical ritual is associated with anum na in sūtra 5.1.20, and this unable to trump what is expressly said, etc. As for the MS itself, “So we ought to conclude carefully that the MS might be of a rather high antiquity: 450—400 B.C. (? the age of Pā ini?), but the collection took its present form under the influence and the name of one or several Jaimini(s) in a later period. But when? Is it at the time when Kātyāyana the grammarian commented on Pā ini in his V rttika, about 250 B. C. (?), or when the Ved nta-sūtras were redacted, that is between 200 and 300 A. D. (?).” (Verpoorten 1987Ś5). 70 figurative meaning to have precedence over direct meaning so that such a narrative or commendatory passage may be related to an injunction through a common reference. Furthermore, in such passages, Śabara thought, the considerations of truth and falsity did not matter, and for this reason they could not be offered in evidence, which was the primary function of ruti.51 Notably, when such a passage is offered in evidence, Śabara declines to acknowledge it as having the force of ruti. While everything in the Veda has to be purposeful, not everything is evidence, and the arthav das were directly liable to interpretation so that they could be related to other direct statements.52 When such passages are called ruti, and that happens very rarely, they are invariantly further qualified: the arthav das are, for instance, sometimes called phala- ruti or sentences about results following from the use of specific substances in a sacrifice, or the accounts of creation are sṛṣṭi-pralaya- rutiṬ53 They are, furthermore, never offered in evidence. They are, thus, ruti (statements), which can be understood in the literal sense, but they are not really ruti because they should not be understood as they are heard. We may further speculate that the mantra section of the Veda would also have been given the similar treatment under this understanding, because in the ritual context mantras are not taken to express what they directly mean. Such surely would not have been the universal understanding even in Śabara’s time, and Prabhākara provides some internal evidence when he comments on the reasons for the investigation into dharma under MS 1.1.1Ś “’But, if this matter is understood just from the Veda, why is it inquired into here?’ – True, but the great seers are of 51 See MSŚBh 1.2.10. 52 See MSŚBh 1.3.13, I.242, where a sentence that looks like an injunction, c rya-v ca pram ṇam, adduced by the opponent in 1.3.11, is interpreted as arthav da because it hangs on another injunction and, thus, the word c rya must be taken in the figurative meaning, namely veda, which precludes direct validity. yat tu śrutir iti naitat. arthavādatvāt. katham arthavādaḥ. vidhy-antaraṁ hy asti, āgneyo 'ṣ ākapāla iti. atrācāryo vedo 'bhipretaḥ, ācinoty asya buddhim iti. 53 See MSŚBh 4.3.1 on the phala- rutis. 71 opposite opinions in this regard. Some explain it based on other reliable warrants, while others accept that even the mantras and arthav das are reliable warrants in regards to their meaning just as it is heard, saying that the Veda is a reliable warrant even in matters not pertaining to duty.”54 However, while a fuller history of ruti will require some more leisure (and we can be fairly certain that the idea does not go to the oldest Vedic texts), what happens in all likelihood is a progressive broadening of the scope of ruti (not the Veda!) parallel to the broadening of the scope of smṛti and likely driven by it— ruti does not appear to have been a thing before the tradition became aware that there are customs that are not directly evidenced in the Veda and must be justified as memory—from the ritual context to the whole of the Veda and from the Dharma literature to all the books that the Brāhma ical society came to accept. How far is the classical theory—the seers saw/heard the mantras, s mans—from the milieu in which ruti was likely first theorized! From Śruti to Śāstra Something more, however, is in evidence in the Bh ṣya of Śabara and the long fragment from the older commentary on the MS by the V ttikāra which Śabara quotes.55 ruti already being called pratyakṣa in the sense of express textual statements which one understands just from hearing becomes likened to pratyakṣa in the epistemological sense as perception and placed on a rank equal to it. Parallel to this, it is given primacy over anum na or inference in the specific, technical sense of inference, as well as all other reliable warrants. Related to this develops the 54 yadi punar vedād evāyam artho ‘vagamyate, kim atra jijñāsyate iti. satyamś vipratipannās tu maharṣayaḥ kecit pramā āntara-pūrvakam arthaṁ var ayantiś kecin mantrārtavādānām api yathā-śruta evārthe prāmā yam icchanty akārye ’py arthe vedaḥ pramā am iti vadantaḥ. Bṛhat , 1.1.1, I.19. 55 As reported by Taber 2005:176, Kumārila and Pārthasārathi Miśra identify the V ttikāra with Upavarṣa. Nakamura 2004:32-5 rejects this identification, but his reasoning seems to be based on the assumption that Śabara quotes the V ttikāra verbatim, and the V ttikāra would not have said “Upavarṣa says” in his own text if he were really him. This would not be a problem if Śabara, in fact, paraphrases the V ttikāra. 72 doctrine of the absence of an author of the Veda. This happens, it seems to me, by a transfer of meaning from ruti to a term related to it, stra, or Veda in its epistemological role as one of the reliable warrants, pram ṇas. It is possible, indeed, to talk about a set of related terms, terms with a single denotation but different connotations, which would, apart from ruti and stra, include veda, veda-vacana, gama, abda, v kyaṬ They all either directly refer to the Veda, or can be applied to the Veda in a more restricted sense, and their use is functional and relational—they are found in sets such as veda/loka or different classifications of pram ṇas. That they all have the same denotation allows for an occasional interchangeable use: the principle of synonymy. For instance, commenting under MS 1.3.14, Śabara uses stra where properly he should use rutiŚ In the prima facie view the words of the teacher [an author of a Kalpa-Sūtra] are, ‘The Amāvāsya ritual should be performed on all days.’ But, easily accessible is the scripture, ‘The Pūr amāsa ritual should be performed on the full moon day, and the Amāvāsya ritual on the new moon day.’ Therefore, because the sentence is opposed to the direct statement, these [statements of such kind, i.e. from the Kalpa-sūtras] are not veracious statements. Ergo, they are not reliable warrants.”56 Scripture or stra is used here clearly in the sense of ruti or express, particular statement. stra is otherwise the term used for the function of the Veda as a reliable warrant. It is the term already used by the V ttikāra in the classical formulation of the Veda as pram ṇa among the other pram ṇas, and Kumārila takes it as synonymous with codan and upade a, again the two terms used in the sūtras themselves to refer to the Veda as pram ṇa.57 But since 56 ācārya-vacanaṁ hi bhavati pūrvapakṣe sarvāsu tithiṣv amāvāsyā iti. sannihitaṁ ca śāstram paur amāsyāṁ paur amāsyā yajeta, amāvāsyāyām amāvasyayā yajeta iti. tena śruti-viruddha-vacanān na satya-vācaḥ, tasmād apramā am. MSŚBh 1.3.14, I.242-43. This is, of course, not to deny reliability to the Kalpa-Sūtras, but to deny them a ruti status. 57 sāmānya-rūpam apy etad adhikārād viśiṣyate | codanā copadeśaś ca śāstram evety udāh tam. ŚV abda 12. “The generic form [of verbal cognition] is here specified because of the context [that is, abda as the general way of learning from words is specified as scripture as learning from the words of the Veda]. It was already declared that codan (direction) and upade a (instruction) are what stra (scripture) is.” The two words are from sūtras 1.1.2 and 1.1.5. 73 both terms, ruti and stra, denote the Veda, the transfer of qualities is open and it is here that we witness how ruti that is called pratyakṣa for long centuries in the sense of express statements contributes to the idea of stra, the Veda in its epistemological role beyond that of mere interpretation, being likened to pratyakṣa precisely as perception. The idea that the knowledge of the Veda is equal in status and validity to perceptual knowledge is based on the Mīmāṁsā distinction between cognitions in which the object is immediately present and cognitions in which the object is remote on the one hand, and the notion that verbal knowledge as such does not presuppose a prior cognition in which a relation between an object and a word is established on the other. Going through the six reliable warrants that Mīmāṁsakas accept in the classical formulation by the V ttikāra, we witness how he draws a distinction between warrants in which the object is near and warrants in which the object is remote, that is, perceptible or otherwise (sannikṛṣṭa or asannikṛṣṭa), and we take note how scripture is classed in the second group, with the other four that are not perception.58 Normally that the object is not immediately given to perception yet otherwise knowable would mean that there had to have been a perceptual knowledge at some earlier point on which this new, non- perceptual cognition is based. Comparison, for instance, tells us something not about what is presently seen, but about what is otherwise known but not immediately present. “Comparison or similarity produces a cognition about a thing which is not present, like in the case of remembering the cow upon seeing a gavaya.”59 Seeing a gavaya or a wild cow in the forest is not knowing something about the gavaya, but about the domestic cow, because it is the cow which is not perceptually present. If comparison would be about the present object, it would be just a 58 They are perception (pratyakṣa), inference (anum na), scripture ( stra), comparison (upam na), postulation (arth patti) and absence (abh va). 59 upamānam api sād śyam asannik ṣ e 'rthe buddhim utpādayati. yathā, gavaya-darśanaṁ go-smara asya. MSŚBh 1.1.5. I.37, quoting the V ttikāra. 74 variation of perception. Comparison, then, proceeds by way of remembrance of prior perception of the cow and a present perception of the gavaya, giving rise to a new cognition about the non- present cow.60 The object is already perceptually known, but the specific new information about it, which a reliable warrant must provide, is not knowable through perception. Scriptural knowledge, however, is odd because its object is perceptually inaccessible, but it does not share with the other four the fact of having a prior cognition as the sine qua non through which it proceeds. An inference of fire on the hill requires actually seen smoke and a remembered relation between smoke and fire, and the cognized similarity of the cow with the gavaya requires a previously seen cow and a presently seen gavaya.61 Scriptural knowledge, on the other hand, is derived just from a verbal cognition,62 one which does not presuppose some other perceptual cognition for its very possibility.63 The V ttikāra insisted that this feature as essential to scriptural knowledge, which is had when “the word itself talks” and which, for this reason, cannot possibly be wrong because it does not report anything previously experienced. Later Mīmāṁsakas, however, universalized this feature of being inerrant to knowledge from verbal cognition as such, scriptural or testimonial, and it is convenient to begin with that point. Let us note first, though, that historically this happened as a response to the challenge of Dignāga and Dharmakīrti, who claimed that the ultimately existent things or unique particulars (svalakṣaṇas) are unnameable, since naming operates with concepts that are imposed over the unique particulars that are known in perception. Buddhist epistemology, in other words, privileged perception (really, sensation) and spurned language as inherently conceptual. As 60 This is an understanding of comparison different from that of Nyāya. See Taber 2005Ś198-9. 61 Śabara explicitly says in MSŚBh 1.1.4, I.22 that inference, comparison and postulation are preceded by perception, pratyakṣa-pūrvakatvāc cānumānopamānārthāpattīnām. 62 In terms of terminology I follow here Matilal 1990 as far as possible. 63 “Scriptural knowledge is a cognition about a thing which is not present, from cognition of a linguistic utterance.” śāstraṁ śabda-vijñānād asannik ṣ e 'rthe vijñānam. MSŚBh 1.1.5. I.10, quoting the V ttikāra. 75 Arnold puts it, “The basic idea is that a bare perceptual event is constitutively nonlinguistic, with the subsequent addition of linguistic interpretation representing, among other things, the point at which cognitive error creeps in.”64 The challenge was, in other words, that knowledge from words is errant as such.65 Take Prabhākara’s example, “There are hundreds of herds of elephants on the fingertip.” This is, of course, a claim which cannot be true, but the reason is not that it fails to communicate meaning. We do understand what “hundreds of herds of elephants” means and insofar we do understand it, we have a verbal cognition.66 Such knowledge from verbal cognition cannot be wrong as long as we grasp the meaning. Sensu stricto, words do not operate in regard to external, experiential matters, such as elephants on fingers, for—as Śālikanātha says—there just is no such rule. Errors can creep in only in sentences that pretend to report a factual state of affairs.67 If one 64 2005:25. 65 Here is how Prabhākara formulates the challengeŚ “Stating that words do not refer to things in, “But, is it not the case that an injunction may convey a falsehood,” he expresses a fault in the thesis. – And how is it that words do not refer to things? It is said: A reliable warrant which does not correspond to the thing is abandoned [as being such a warrant]; because, when [there is a faulty cognition, and] a defeating cognition appears, the opposite cognition based in that [object] does not reappear; but in regard to words, though there is a clear defeating cognition, the opposite cognition which is word-based appears again—for instance, in “there is a herd of elephants on the fingertip.” Therefore, the thing [as a referent to the word] is not even possible, as [it is possible] in the case of silver and mother-of-pearl. Therefore, since an injunction does not correspond to things, the thesis that it is through injunction that dharma is known does not make sense.” nanv atathā-bhūtam apy arthaṁ brūyāc codanā iti śabdasyārthāsaṁsparśitāṁ vadan pratijñā-doṣam āha. kathaṁ punar arthāsaṁsparśitā śabdasya? ucyate— arthāsaṁsparśi hīyate pramā amś tasya bādhaka-jñānotpattau satyāṁ na punaḥ pratipakṣa-vijñānaṁ tan-nimittam evotpadyate. śabde punaḥ spaṣ e ’pi bādhaka-jñāne punaḥ śabda-nimittam eva pratipakṣa-vijñānam udeti—a guly- agre hasti-yūtham asti – iti. tathā na śuktikā-rajatādiṣv asyārthasya sambhavaḥ. tasmād arthāsaṁsparśitvāc codanāyāḥ tal-lakṣa o dharmaḥ iti pratijñā nopapadyate. Bṛhat 1.1.2, I.23-4. For the challenge itself (an earth- shattering, defining event in Indian philosophy), see a short account in McCrea 2013 and a longer in Arnold 2005:13-31. 66 “You do not understand the defeated-defeater relation. Here what is defeated is the vision ‘there are hundreds of herds of elephants on the fingertip,’ which is [in the domain of] another reliable warrant. Not, however, ‘hundreds of herds of elephants.’” anabhijño bhavān bādhya-bādhaka-bhāvasya—a guly-agre hasti-yūtha-śatam—iti pramā āntara-darśa am atra bādhyate, na punar hasti-yūtha-śatam. Bṛhat 1.1.2, I.24-6. 67 Cf. Ṛju-Vimal 1.1.2, I.26-7Ś “Here another cognition, to be ascertained by inference, is defeated. But, [the cognition in] the form of ‘hundreds of herds of elephants’ is not an external thingś because, the sentence does not function in regards to such external things. A justified warrant can be defeated in regard to that sphere where it operates as such. Speech does not operate in regard to external matters, for there is no such rule.” anumānāvaseyaṁ pramā āntaram atra bādhyateś na punar hasti-yūta-śata-rūpaṁ bāhyaṁ vastu, tatra vacanasyāvyāpārāt. yatra hi 76 is a competent user of a language, one will have verbal cognitions, and if an utterance proves to convey an error, as in the case of elephants on fingers, we should look for the source of this error elsewhere. Mīmāṁsakas developed different theories about where precisely to locate such error. Following Prabhākara, Śālikanātha identified human speech with a form of inference in which what the speaker says is taken as an inferential mark (liṅga) of her cognitive state.68 For a good inference, two criteria would be necessary, namely that the speaker be trustworthy and that the reported knowledge be otherwise in the domain of what is perceptually knowable to the listener.69 In such an inference, an error can be made on two counts. We could, first, take something that is not an inferential mark (liṅga) of a reliable cognition of the speaker to be such a mark, and that could happen if we do not apply the two aforementioned criteria. In making such an inference we would have to guard ourselves against frauds, “who know one thing and say another.”70 In a good inference, in other words, we would correctly ascertain the intention of the speaker. The product of our inference, however, would be knowledge of the cognitive state of the speaker, not of the correspondence of what she is saying to the actual state of affairs, and even when we’ve ascertained it correctly there could still be an error if the speaker’s cognition has missed the reference. She might have been distracted, or the complex of conditions necessary for a correct perceptual cognition might have been compromised (bad eyesight, insufficient light etc.).71 She may, thus, claim that the “butler did it,” be trustworthy and convinced in her viṣaye yat pramā aṁ prav ttam, tatra tasya bādho bhavati. na ca vacanasya bāhye prav ttiḥ; tatra tasya prav tti- niyamābhāvād ity uktam. 68 Agreeing, by the way, with Dignāga and Dharmakīrtiś see Arnold 2005Ś40. This whole section is based on Ṛju- Vimal and Bṛhat on MS 1.1.2. 69 pratyayitasya hi vacanam avyabhicarāri pramā āntara-pūrvakam, yatropalabhyamāna-viṣayam; Bṛhat 1.1.2, I.26. 70 pratārakā anyathā jānanto ’py anyathā vākyāni prayuñjate. Ṛju-Vimal 1.1.2. I.32. 71 tathā pramādino duṣ a-sāmagrīkāś cānya-vivakṣāyām anyad eva vākyam uccārayanti. Ibid. 77 cognition, yet there could still be no correspondence between her words and the reference, the one who actually did it. Such an error would not be a product, however, of any of the pram ṇas that are involved in knowledge from linguistic utterance, namely perception, inference and verbal cognition. For Prābhākaras, an error in perception occurs when something representative, a form of memory, gets mixed with a present object.72 An image of the butler, in other words, interferes with what our lady perceptually cognizes in a generalized form as “this,” and she is unable to distinguish the two owing to distraction, because it is dark etc. The presentational for Prābhākaras was universally inerrant, the error was a matter of a mix-up. Further, Śālikanātha defined inference quite formally—a cognition of a cause from an effect is all what inference is73—and an error in the inference can happen, again, when one thing is mistaken for another, specifically when that which is not a mark is taken as such, say smog taken as smoke. This would not be an error of inference per se, but, again, a case of confusion of one thing for another, a case of wrong identification. In the case of testimony, such a confusion could happen if the words of one untrustworthy are taken as words of one who is trustworthy, because his true intention has not been properly ascertained. In any case, no error would come from the verbal cognition, for such cognition, like perception, is a form of awareness—knowledge of a thing got just from the word—and since awareness is presentational, “perceptual,” a reference would be immediately given to it.74 Whether the knowledge of the listener, however, will correspond to the reference actually intended by the speaker, when words are employed in sentences in other words, will 72 “There, it is not the perceptual knowledge which misses. One thing there is the visual awareness of “this is something,” but quite another “it is silver,” which is merely recollective awareness hanging on to silverness. Here the error consists in not distinguishing the two, but the perceptual knowledge in inerrant.” na tatraindriyakaṁ jñānaṁ vyabhicarati. ekaṁ hi tatra jñānam—idam—iti cākṣuṣam, aparam iti—rajatam—iti rajata-mātrāvalambi smara a-jñānamś tayor vivekāgraha a-nibandhano bhramo ‘yamś na punar aindriyakaṁ jñānaṁ vyabhicarati. Ibid. 73 kāryataś ca kāra āvagatir anumānam eva. Ibid. 74 Cf. Bṛhat 1.1.2, I.33: saṁvidaḥ pratyakṣatvāt. 78 depend on these other conditions. Prabhākara’s claim, thus, was that words expressing concepts are equally basic to experience as perception is, because they are given to awareness without mediation. They are not constructs or abstractions. Errors, however, are possible in testimonial, sentential, accounts. Kumārila’s response to the Buddhist challenge that conceptual thinking involving words is the point at which error enters cognition was a frontal and multipronged counterattack. It involved, first, a reaffirmation of conceptualized or verbalized perception. Kumārila argued at length that it is not only the bare percept of a unique particular, stripped of all concepts expressed in words, which qualifies as what is apprehended in perception.75 Conceptualized/verbalized perception is equally perceptual and revealing the object because if the object of the percept does not have the characteristics attributed to it, all ground for superimposing them by the mind would be lost. Why cognize something as a cow and not a horse if such conceptualized cognition is just a matter of construct? Who decides that, and how is intersubjectivity possible without it? Bhatt expresses the gist of Kumārila’s argument wellŚ “It should be noted that such features are not imagined but discovered by the assimilative and discriminative operations of the mind. They are as much objective as the individual (‘Vyakti’) revealed in the first stage [non-conceptualized perception], which is the concrete unity of universals and particulars.”76 The second point of Kumārila’s counterattack was the doctrine of intrinsic validity (svataḥ pr m ṇya) of all cognitions—bare or conceptualized perceptions, testimonials, comparisons, inferences etc.77 Validity, claimed Kumārila, is a faculty or a capacity ( akti) which 75 The whole argument is given in the Pratyakṣa-Pariccheda chapter of ŚV, which has been competently translated and interpreted by Taber 2005. 76 1994:16-17. 77 See Arnold 2005:59-114; Taber 1992; Kataoka 2011b:60-98, 228-308. 79 cognitions as kinds of things must have if they are to have it at all.78 Seeking justification for belief in a certain cognition outside the cognition itself, how it presents itself to our cognitive ability, is doomed to regressus ad infinitum because any further cognition would again seek justification in another cognition without end. All that one can offer, for instance, for the claim that what one saw was a ball is that it appeared as a round object, no doubt was present to one’s awareness etc. These are all cognitions like the first, ball-cognition, and pursuing validity for the first through them will always ask for a further point of validation. As Arnold puts itŚ “[I]f the initial cognition is not credited with the ‘capacity’ for validity, then no other cognition will be able to bestow that—unless, of course, the second-order cognition is intrinsically credited with that capacity, in which case, why not simply allow this with respect to the initial moment?”79 For there to be some form of validity to begin with, it has to be postulated that any cognition is prima facie valid.80 A cognition can be falsified if some other, subsequent cognition contradicts it, for instance when the cognition of mother-of-pearl replaces that of silver, or if there is a reason for doubt because the cognition does not present itself clearlyŚ “Is it a man or a post?” In any case, such invalidity would come from outside, from another cognition or from a defect in the factors contribution to the cognition, such as healthy senses, sufficient sunlight or what have you: a cognition is not independent in regard to its origin.81 But prima facie, any cognition is valid as it presents itself, unless and until proven wrong. This applies to testimonial knowledge as well. It 78 svataḥ sarva-pramā ānāṁ prāmā yam iti gamyatām | na hi svato 'satī śaktiḥ kartum anyena śakyate. ŚV Codan , 47. “All pram ṇas have intrinsic validity; a faculty that is in itself unreal cannot be made real by another.” 79 2005:69-70. 80 Validity as a akti, a category which is not among the existents in the world (vastu), but is a feature of such existents, having no other locus except for the existent, is known through postulation. Kumārila was otherwise very fond of the category of aktiṬ 81 ŚV Codan 48. 80 can prove wrong only by a subsequent cognition that the speaker is not trustworthy or could not have experienced what she reports.82 Kumārila was not ready to reduce even testimonial knowledge to inference, as Prābhākaras did, for several reasons, but ultimately because the cognitive state of the speaker, taken to be the inferential reason (hetu or liṅga) in virtue of the two criteria, the speaker being trustworthy and the object being in the experiential domain, is not really a mark of the reference, as Prabhākara himself admitted. It is a mark of an object present in the speaker’s awareness, but there is no invariant relation between the word the speaker uses—“butler”—and a necessary presence of the reference of the word, when, say, the maid did it. In plain language, “the butler did it” means just what the witness thinksŚ there may be no butler behind the witness’s thought, he was with the householder’s wife and cannot provide an alibi. For words not to miss the mark, they need to refer directly, human speech included, and the reduction of knowledge from verbal utterances to inference, as the Bauddhas proposed, would strip abda of validity entirely, because one cannot be certain in principle whether one’s words represent the object accurately. The sphere in which words are pram ṇa, therefore, has to be different than the sphere of inference. Now, it is true that in testimonial accounts we need to rely on the cognitive state of a specific speaker, and we also need to assume an invariant concomitance between “correspondence of cognition/words to experience” and “being a trustworthy speaker,” just as we need to know that smoke is universally related to damp firewood. Thus, though a cognition arises from a linguistic utterance as prima facie intrinsically valid,83 there is still an afterthought in regard to its validity, 82 See Kataoka 2011b:309-19. 83 purastād var itaṁ hy etat tasmāc chabdena yā matiḥ | tasyāḥ svataḥ pramā atvaṁ na cet syād doṣa-darśanam. ŚV abda, 53. “This [that the validity cannot rest on trustworthiness] was explained before. Therefore, the thought that is owing to an utterance is intrinsically valid, if no fault is seen.” 81 “Well, why should I believe him? Is he trustworthy, could he have actually experienced this?” Figuratively, we can call human speech inferential, since it shares with inference an element of invariant concomitance (albeit assumed), but the instrument which causes the cognition is the utterance, and the utterance is understood once we have understood what the individual words mean directly, not through the cognitive state of the speaker. We may, thus, call testimony inferential, but we cannot define it as inference.84 The Impersonal Nature of the Veda But these differing accounts aside—and there were many other related differences, some of which will appear in the next chapter—both camps agreed that knowledge from verbal cognition as such is inerrant; and, both the presupposition and the refined product of this was that words— as informative and corresponding to referents—do not denote individual things, but real universals, the shape or blueprint of things. Verbal knowledge as such is never about this specific pot in front of me or about historical events. It, rather, provides the natural categories through which particular experience is structured. Of course, in an actual testimony corresponding to actual experience, words composed in a sentence would, if so required by the sentence, refer to individuals, but again as instantiations of such universals. As Kumārila claimed, individual entities have that dual nature to be an indivisible unity of a particular and a universal.85 That words have universals as referents was in itself not controversial, and it was accepted even by Buddhist philosophers. However, whereas Buddhists thought that universals are constructs, Mīmāṁsakas, as we saw above, took them to be real on one hand and revealing 84 See ŚV abda 39-52, 109ś see also Pārthasārathi on 48. 85 See McCrea 2013:134-6. 82 rather than concealing the true nature of object. This Mīmāṁsā doctrine was formulated already in one of the crucial sūtras of the MS (1.1.5), commonly known as the autpattika-sūtraŚ “The relation of the word to its meaning is innate. Knowledge of such meaning is [had through] instruction, which is infallible in regard to imperceptible things. It is a reliable warrant, according to Bādarāya a, because it is independent.”86 The consideration of the sūtra leads us, finally, to the consideration of the nature of Vedic knowledge, stra as a specific form of abdaṬ Though knowledge from verbal cognition as such was theorized as inerrant, yet it was not accorded the status of pram ṇa because of the strict requirement that a pram ṇa should tell us something new. Once we’ve learned what words mean, when used in non-sentential context they could at best remind us of a previously known reference.87 And, as we have seen above, when sentences are formed from words, in testimonial accounts, there were two scenarios under which Mīmāṁsakas allowed an error to infect what is actually said: in intentional deception—“In all cases the thing is not cognized directly through personal speech, because it is contingent on personal desire, and persons use words even when the thing is not there”88—and in reports of erroneous cognitions even when intentions are aboveboard. Both of these scenarios, notably, are contingent on personal agency: they require a speaker who is absolutely trustworthy and omniscient. Therefore, both can be removed, thought Mīmāṁsakas, in a single stroke: remove the speaker and you remove both intention and prior cognition. Intention (vivakṣ , literally desire to say) as Mīmāṁsakas understood it is the key word here, for it covers not only cases of speech intending to convey a state of affairs, true of false, but 86 autpattikas tu śabdasyārthena sambandhas; tasya jñānam upadeśo 'vyatirekaś cārthe 'nupalabdhe, tat pramā aṁ bādarāya asya, anapekṣatvāt. My translation follows the V ttikāra. It is, perhaps, significant that this centralmost doctrine is attributed to Bādarāya a, the purported author of the Brahma-Sūtra. 87 See ŚV abda 99-111. 88 sarvatra pauruṣeye a vacasā na sākṣāt artha eva pratīyate, puruṣecchā-paratantratvāt. asaty apy arthe puruṣā vākyaṁ prayuñjānā d śyante. Ṛju-Vimal 1.1.2. I.25. 83 also the case where an individual or a group intends that a word should denote one thing rather than another. The second was arguably more dangerous for Mīmāṁsakas, since it could happen that the most cherished word of words, dharma, gets to mean what the Buddha intended it to mean. So, the first step to secure the inerrancy of Vedic knowledge was to propose that the relation between the word and its reference is innate, natural, never instituted into being by a personal (or contractual) whim, human or divine. The word-reference relation is non-personal, apauruṣeya, said the V ttikāra, and this is likely the first appearance of the idea that we have evidence of.89 There never happened in history an event when someone said, “let this thing be called ‘a ball,’ that thing ‘a cow’ and that yonder thing ‘dharma.’” The basic, non-technical meanings of words (in the Sanskrit language), which impose upon our understanding on hearing, are meanings which words have always had, since the (non)-beginning of time. (Plato should have consulted Jaimini to learn the names of his ideas in the topos hyperuranius.) But even such non-intentional words could be intentionally inflected by personal agency, as we have seen above. Therefore, at the last step it was claimed that there never happened in history an event where a person used the word-meaning relationship to compose the Vedas. The Vedas have always been there, transmitted in the same way as they are now.90 This is the last feat which should secure the inerrancy of the Vedas: they are, properly speaking, not accounts of anything and they do not presuppose a prior cognition of which they would be reporting. The absence of a prior cognition that is a matter of reporting should, thus, eliminate the second point 89 apauruṣeyaḥ śabdasyārthena sambandhas. MSŚBh 1.1.5, I.41. Śabara has aviyuktaḥ as a gloss of autpattikaḥ, inseparable for innate, and he does not use the term apauruṣeya except when quoting the V ttikāra, but does talk about pauruṣeyaṁ vacanam, human speech. 90 V ttikāra, “’Therefore, we think someone, a person, created a word-reference relation and composed the Veda in order to employ this relation.’ – On this, it is now said, it is proven, because the relationship is non-personal.” tasmān manyāmahe kenāpi puruṣe a śabdānām arthaiḥ saha sambandhaṁ k tvā samvyavahartuṁ vedāḥ pra ītā iti. tad idānīm ucyate – apauruṣeyatvāt sambandhasya siddham. MSŚBh 1.1.5, I.52-53. 84 at which error could creep in. There is no possibility for a mix-up or for the virtues or defects of the speaker to modulate testimony.91 Mīmāṁsakas have used many arguments to bolster the claim that the word-meaning relationship and the composition of the Veda are non-personal. There is no perceptual evidence of an author, and all the other reliable warrants operate with perceptual data.92 Were there one, we would have remembered him no matter how long ago it was, as that would have been a major event. Surely, we would have remembered if someone built the Himālaya; it is not our home garden after all, the memory of which could have faded in the family. Men, furthermore, do not have the power to compose on such supersensible matters as are treated in the Veda.93 One is tempted to interpret such understanding of the Veda as one of a naturally structured phenomenon: in the Vedas words are meaningfully ordered just as planets are in the solar system, as rivers flow down to the ocean, as water melts sugar and salt. Whereas for deists such ordered phenomena would betray personal intelligence behind it, Mīmāṁsakas would have agreed with Hume that this information is just not given in experience. Two arguments specifically seem to support such an interpretation. The first argues from the awareness of an author that we have in the use of technical language, such as of Pā ini when we talk Sanskrit grammar, and the absence of a similar impression in everyday language. While the claim that we are aware of a specific author is not particularly strong, we do have the sense of artificiality of technical lingo and notations, for instance in formal logic, which is strikingly different from the sense about the natural language acquired through observing how elders talk, 91 “In the Veda there is not even the possibility of a defeating cognition. Therefore, there is not even a doubt that words might not correspond to their reference.” vede punaḥ bādhaka-jñānābhāvāt arthāsaṁsparśitāśa kāpi nāsti. Bṛhat 1.1.2, I.24. See also Kataoka 2011b:272-92. 92 V ttikāraŚ puruṣasya sambandhur abhāvāt. kathaṁ sambandhā nāsti? pratyakṣasya pramā asya abhāvāt tat- pūrvakatvāc cetareṣām. MSŚBh 1.1.5, I.53. 93 Śabara 1.1.25, I.99Ś api caivaṁ-jātīyake 'rthe vākyāni saṁhartuṁ na kiñcana puruṣā aṁ bījam asti. 85 as Mīmāṁsakas would say, where we never stop to think, “someone made this up.” The second argument proposes that the claim that a personal author must have established the word-meaning relationship would involve a vicious circle, because such action of naming presupposes the use of language. “Let’s call this a ‘ball’” as a performative utterance depends on a preexisting verbal practice. Both arguments, however, refer just to the word-meaning relationship, not the composition of the Veda, and Mīmāṁsakas, one gets the impression, hope to prove both by proving the first. Be that as it may, such was in Mīmāṁsā understanding the intention and prior cognition that were lacking behind language in general and the Vedas in particular. Not only are the Vedas not like Marco Polo telling us about his travels: they do not even admit an original act of assigning names to things. Knowledge from the Veda is knowledge just from words and from sentences which do not depend on or convey a prior experience. It is knowledge, but not about what has already been seen or heard.94 Śruti-Śāstra as Perception We can see, now, why scriptural knowledge would be likened to perception and contraposed to inference. Perceptual knowledge does not require data from another reliable warrant. Human or personal speech does and therefore, though its operation is independent insofar as it provides new information in a way different from the other reliable warrants, it is ultimately dependent on 94 “But when the word itself talks, how can it be false? In that case there is no understanding from another person. When it is said ‘it talks’, it means that it makes known. It becomes the means of something being understood. Since the word is the means, it makes known by itself.” atha śabde bruvati kathaṁ mithyeti. na hi tadānīm anyataḥ puruṣād avagamaḥ. bravītīty ucyate bodhayati budhyamānasya nimittam bhavatīti. V ttikāra in MSŚBh 1.1.5, I.42. “That is, when a cognition has been brought about by means of words, there is no need for any other cognition (to corroborate it), or of any other person as having the same cognition.” na hy evaṁ sati pratyayāntaram apekṣitavyaṁ, puruṣāntaraṁ vāpi. ayaṁ pratyayo hy asau. MSŚBh 1.1.5, I.25. 86 perception.95 So are the other pram ṇas, as discussed above. But knowledge derived from the Veda is not of the testimonial kind; it is direct just like perception. “It occupies the same rank as the senses,” as Śabara says in the passage which serves as this chapter’s epigraph. The Vedas are, as it were, the “senses” for knowing that which is beyond the senses, the third eye of the Vedic theologian, and any form of knowledge which is structurally derivative cannot be applied in an argument against them. Inference proper, not just inference as a blanket term for scriptural interpretation as paired to ruti-pratyakṣa, cannot be offered against an express scriptural statement, just on the ground that it is inference, while the statement is perception: “Also, because of similarity to personal speech, the Vedic speech would be false.” – This is inference; the cognition in regard to Vedic speech, however, is perception. Inference which is against perceptual evidence is not a reliable warrant.96 The notion, “seeing falsity in the case of another cognition, this one will be false as well” is an inferential one; being contradicted by this perceptual cognition, it is defeated.97 This is, of course, a figurative use of the word pratyakṣa.98 Kumārila significantly also talks about inference in its nominal and definitional sense, saṁjñ and lakṣaṇa, as we have seen above with testimony being inferential but not inference. He also reinterprets the smṛti=anum na, which Śabara took literally, as smṛti=arth patti, a postulation of lost Vedic passages rather than a formal inference in virtue of the agents being the same, on the ground that any knowledge that comes after some other knowledge can be called anum na, inference.99 This allows us, it seems 95 See, for instance, ŚV Codan 71: pauruṣeye tu vacane pramā āntara-mūlatā | tad-abhāve hi tad duṣyed itaran na kadācana. “As for personal speech, it is dependent on another reliable warrant. Without such reliability it would be errant, but never so non-personal speech.” Kumārila proceeds to affirm that all pram ṇas are informative independently and even when they do provide the same knowledge, they do not do so in the same way (73-4). See also Kataoka 2011b:279-82. 96 api ca puruṣa-vacana-sādharmyāt veda-vacanaṁ vitatham ity anumānaṁ vyapadeśād avagamyate. pratyakṣas tu veda-vacanena pratyayaḥ. na cānumānaṁ pratyakṣa-virodhi pramā aṁ bhavati. MSŚBh 1.1.2, I.17. 97 yo 'py asya pratyaya-viparyāsaṁ d ṣ vātrāpi viparyasiṣyatīty ānumānikaḥ pratyaya utpadyate, so 'py anena pratyakṣe a pratyayena virudhyamāno bādhyate. V ttikāra, MSŚBh 1.1.5, I.42. 98 pratyakṣas tv iti … lakṣayet. Pārthasārathi thereonŚ aindriyaka-jñāna-vacana evāyaṁ pratyakṣa-śabdo vācanike jñāne gu a-v ttyā prayuktaḥ. ŚV Codan 187. 99 tasmād arthāpattir evātrāvyabhicārād upacārāt paścān mānād anumānatvenoktā. TV 1.3.2. 87 to me, to see the application of the concepts of pratyakṣa and anum na in a strict as well as a loose sense and then group items applying the second. Thus, the six reliable warrants are divided in perception on one hand, where the object is directly visible, and the rest, where the object is not present. Testimonial speech is grouped with the second. However, when the general abda is specified and stra is taken as separate, then we have perception and scriptural knowledge on the one side, in which no prior knowledge is presupposed for the operation of the pram ṇa, and abda and the other four pram ṇas, which depend for their operation of perceptual data. It is useful, at this point, to refer to the familiar loka/veda distinction, the two realms postulated by Vedic theology. The first refers to the world as we know it, with natural laws knowable directly through perception and the other three, four or five pram ṇas that harness perceptual data. The second refers to the scriptural domain, which has its own, “verbal” laws that one can harness to one’s benefit—more of which in the next chapter—and which is knowable directly through an express Vedic statement, ruti, and the different sets of pram ṇas that employ ruti-data. Each realm, in other words, has a basic pram ṇa, one that gives a direct access to the realm, and because the second shares such directness within its own realm with the first, in virtue of that quality it can be called perception. In both realms the other pram ṇas that do not have this direct access to the realm as a defining property are called anum na, not in the technical sense of inference but in the etymological. Mīmāṁskas, thus, dispensing with trustworthiness and yogic perception of the Buddha, Īśvara and any imaginable authority, replaced one perception with another, rutiṭ stra-pratyakṣa, “scriptural perception” on the part of the subject. In this light, to treat ruti as revelation would be justified only insofar as it affords a glimpse into otherwise concealed world to the subject, a means and not a product. 88  stra, properly speaking, the Veda paired with the other laukika or worldly pram ṇas, is an intruder in classification, a mixing of the realms. We may speculate that its inclusion in the list of pram ṇas of the V ttikāra, admittedly awkward even by Kumārila,100 and the toils to distinguish it from testimonial knowledge, came as a result of the duress of the Veda being under attack and requiring authority that cannot be questioned, by stretching the limits of some old ideas such as the likeness of ruti with pratyakṣa and the prominent role of the direct meaning associated with it. Śruti in Other Br hmaṇical Traditions But such deliberations will have to wait some other occasion. For our purpose it is necessary now to go back to the original question of delimiting the field of Vedic theology. While other Brāhma ical schools were recognized by the time of Kumārila and Śa kara as being part of the smṛti tradition, containing many teachings that the Vedic iṣṭas would approve of, they did not have any serious stake in the direct interpretation of the Veda. Larson’s evaluation of Sā khya’s attitude to knowledge from linguistic utterance is certainly true, if impreciseŚ “Sāṁkhya had never denied reliable verbal testimony ( ptavacana or ruti) as a legitimate and important means of knowing, but Sāṁkhya clearly gave pride of place in knowing to independent reasoning, even in the area of samyagdar ana and adhy tmavidy (that is to say, in the area of ultimate truth and the science of liberation).”101 Though Īśvarak ṣ a in SK 5 accepts knowledge from linguistic utterance as different from inference, his formulation did not recognize the Veda as apauruṣeya as Mīmāṁsakas would have wanted: pta- rutir pta- 100 pratyakṣādiṣu vaktavyaṁ śabda-mātrasya lakṣa am | tad atitvariteneha kiṁ śāstrād abhidhīyate. ŚV abda 1. “The definition of knowledge from linguistic utterance in general should be given in the discussion of perception and the rest. Why is, then, hurriedly the definition of scripture given?” This is the pūrva-pakṣa that opens the abda chapter of the ŚV. 101 Larson & Bhattacarya 1987:29. 89 vacanaṁ tu. The most straightforward translation of this would beŚ “But, hearing from a trustworthy source is a reliable statement.” The reconstructed text of the Suvarṇa-Saptati, which Paramārtha translated into Chinese and is likely the oldest available commentary on the S ṅkhya-K rik , supports such readingŚ “For instance, the four Vedas spoken by Brahmā and the Dharma- stra spoken by Manu.”102 pta was, otherwise, the common word for a trustworthy person, though Mīmāṁsakas preferred pratyayita. Other earlier commentaries, such as the Bh ṣya of Gauḍapāda, analyze pta- rutiḥ as a sam hara-dvandva and take pta to stand for teacher and ruti for the Veda.103 More important, however, is the fact that Sā khya developed in opposition to and rejecting the ritualism of the Veda, which is clear from the very outset of the S ṅkhya-K rik ,104 and however the compound is to be understood, it certainly is at best an early instance of “raising of the hat” to the Veda and moving on.105 Most important, the characteristically Sā khyan objects, puruṣa and prakṛti, were knowable through inference, not scripture. Early Yoga had several doctrines which ran counter the Mīmāṁsā account of the Veda. The understanding of knowledge from linguistic utterance was the familiar knowledge from a trustworthy source, with no special concession for the particular nature of the Veda.106 102 āpta-śrutir āpta-vacanam ucyate iti. yathā brahma ā manunā ca uktāś catvāro vedā dharma-śāstraś ca. SSS 1.5, p.9. 103 āpta-śrutir āpta-vacanaṁ ca. āptā ācāryyā brahmādayaḥ. śrutir vedaḥ. āptaś ca śrutiś ca āpta-śrutī tad-uktam āpta- vacanam iti. SKG 1.5, p.52. The M ṭhara-Vṛtti follows suit. 104 SK 2, which claims that Vedic ritualism as a means ending the threefold suffering is no better than the ordinary, for a variety of reasons. 105 Renou 1965:2. 106 See YS 1.7 and the Bh ṣya thereonŚ āptena d ṣ o 'numito vārthaḥ paratra svabodha-sa krāntaye śabdenopadiśyate, śabdāt tad-artha-viṣayā v ttiḥ śrotur āgamaḥ. yasyāśraddheyārtho vaktā na d ṣ ānumitārthaḥ sa āgamaḥ plavate. mūla-vaktari tu d ṣ ānumitārthe nirviplavaḥ syāt. “The mental modification arising from hearing the words of a reliable person who desires to convey his cognition to the hearer is gama-pram ṇa, i.e., authoritative testimony to the hearer. That testimony may be false, i.e., cannot at all be a pram ṇa, if the person communicating the knowledge is not trustworthy or is deceitful or is one who has neither seen nor experienced what he seeks to communicate. That transferred cognition which has its basis in the direct experience of the first authoritative 90 Furthermore, verbal cognition was understood as concealing the real object, and for such real object, “the thing in itself,” to be properly known, it would have to be first purged from all verbal and conceptual traces.107 Such knowing free from words and concepts constitutes “best perception and the basis and origin of inference and testimony.”108 Patañjali, in other words, was arguably privileging perception no less than his Buddhist peers. Early Yoga, furthermore, while accepting that words generally have a natural expressive power, took the imposition of specific words over specific things to be a matter of convention. In other words, it accepted, unlike Mīmāṁsā, an event of naming.109 The only competing tradition which developed a thorough account of the validity of the Veda was Nyāya, which, nevertheless, differed from Mīmāṁsā on all major counts. The Ny ya- Sūtra (NS) defined knowledge from linguistic utterance solely as testimony, “an instruction of a trustworthy person,”110 and Naiyāyikas were uncompromising in subsuming the Veda under this definition. Vātsyāyana, in fact, defined who a trustworthy person is in wording identical to Yāska’s definition of a ṛṣi, “one who has directly seen dharma,” and applied the definition to everyone—ṛṣis, Vedic nobles and barbarians alike—because they all do their business depending on trustworthy accounts.111 The specific nature of the Veda was only that it relates things that are humanly invisible, but were visible to the ṛṣis.112 Naiyāyikas, furthermore, claimed that the word-reference relationship is conventional, that is, instituted into being in an event of naming: exponent or in his correct inference is genuine and perfectly valid.” Translation in Hariharānanda Āra ya 2000:21. An excellent account of Yoga’s general attitude to scripture is available in Bryant 2009Ś35-38. 107 See YS 1.42-3 and the Bh ṣya thereon. 108 tat paraṁ pratyakṣaṁ tac ca śrutānumānayor bījam, tataḥ śrutānumāne prabhavataḥ. YSBh 1.43, p.93. 109 YSBh 1.27 and 3.17. 110 āptopadeśaḥ śabdaḥ. NS 1.7. 111 āptaḥ khalu sākṣāt-k ta-dharmā … ṣy-ārya-mlecchānāṁ samānaṁ lakṣa am. tathā ca sarveṣāṁ vyavahārāḥ pravarttanta iti. NSBh 1.7, p.24-5. 112 See sūtra 1.1.8, 2.1.67, and Bh ṣya thereon. 91 “Let this name apply to this object.”113 While in the Bh ṣya this convention was left unqualified, later Naiyāyikas attributed the act of naming, as well as of composing the Veda, to God, Īśvara.114 But most important, again, was the fact that Nyāya, much like Sā khya, did not have a major stake in the Veda, and developed all of its major arguments solely on the strength of reasoning. It was only the Vedāntins who joined Mīmāṁsakas in staking everything on the authority and validity of the Veda, because the Upaniṣads were part of the Vedic canon. If the Upaniṣads were to be valid as a pram ṇa, the whole Veda had to have independent validity. For this reason, even Advaitins had to have a stake in Vedic ritualism, as Mīmāṁsakas had to account for the validity of the Upaniṣads. For all of their opposition to Pūrva-Mīmāṁsā, Advaitins subscribed fully and unconditionally to the set of ideas which developed around the autpattika-sūtra, namely that words refer to universals, the word-reference relationship is natural, and the Vedas have no author, presuppose no prior cognition and no intention, and constitute a form of immediate knowledge. a kara’s Understanding of Śruti One cannot but be impressed with the level of sophistication of argument which Mīmāṁsakas produced to defend such a counterintuitive idea: there are books that were never composed by a god or a man, and they are immediately and unconditionally valid. Yet for all this sophistication, at the top of it all, Mīmāṁsakas insisted that it is only in regard to commands that the Vedas have such unconditional authority. Thus, when Śabara laid down a broad claim to anything otherwise 113 “And what, pray tell, is this convention? It is a rule which restricts the denotation of words, ‘this group of objects will be named by this word.’” kaḥ punar ayaṁ samayaḥ. asya śabdasyedam artha-jātam abhidheyam ity abhidhānābhidheya-niyama-niyogaḥ. NSBh 2.1.55, p.137. 114 See Kunjunni Raja 1963:23-4; Potter 1977:376-9; Chemparathy 1987. 92 unknowable being in the domain of the Veda, “the Vedic injunction only, and not any sense organ, can intimate a past, present and future thing that is subtle, concealed, remote or similar in kind,”115 Prabhākara, arguing from the standpoint of strictest Mīmāṁsā fundamentalism, was quick to “clarify” that these other things that the Veda may say have not their normal meaning, but only k rya, “obligation” that is constitutively future. To vulgarize the idea, when the Veda narrates stories, the reference behind such narration, the thing that the story is about, is nothing in the past or present, but just a ritual action that one needs to perform, one that impresses upon the addressee as requiring completion, and the story tells the addressee, “Get on with it.” Everything in the Veda, thus, is normative, in a very restricted sense. The Vedas are also not informative about things that are already constituted and complete (siddha), such as the existents in this world or behind it: cows, agriculture, colors red and blue, heaven as a real place worthy of description, or Brahman. Śa kara’s project of rethinking the Vedas in general and the Upaniṣads in particular took for granted most of what Mīmāṁsakas arrived at concerning the epistemic role and validity of the Veda: without the Veda, there is no means of perceiving that which is in the realm of the supersensible;116 the Veda is authorless and the relation between the word and its meaning is natural; words refer to universals or forms, not to individual substances, qualities or actions;117 115 codanā hi bhūtaṁ, bhavantaṁ, bhaviṣyantaṁ, sūkṣmaṁ, vyavahitaṁ, viprak ṣ am ity evaṁ-jātīyakam arthaṁ śaknoty avagamayitum, nānyat kiṁcanendriyam. MSŚBh 1.1.2, I.13. 116 na cātīndriyān arthān śrutim antare a kaścid upalabhata iti śakyaṁ sambhāvayitum, nimittābhāvāt. BSBh 2.1.1, II.284. 117 apauruṣeye vede vaktur abhāvāt, BSBh 1.2.2, I.98. The same affirmed in BSBh 1.3.28, I.190ś gavādi-śabdārtha- sambandha-nityatva-darśanāt. na hi gavādi-vyaktīnām utpattimattve tad-āk tīnām apy utpattimattvaṁ syāt. dravya- gu a-karma āṁ hi vyaktaya evotpadyante, nāk tayaḥ. āk tibhiś ca śabdānāṁ sambandhaḥ, na vyaktibhiḥ, vyaktīnām ānantyāt sambandha-graha ānupapatteḥś “The relationship between words such as 'cow' and meaning is seen to be eternal. For, it is not the case that when individual cows are born that their universal/shape is born as well. Only individual substances, qualities and action can have origin, not universals. Words are related to universals/shapes, not individuals, because the relation of words with individuals cannot be grasped, since individuals are innumerable.” BSBh 1.3.28. 93 the Veda is “perceptual,” direct, whereas smṛti is inferential, because the first does not depend on anything else for its validity while the second does.118 His project, however, focused on rethinking the nature of the Veda as pram ṇa as having first and foremost truth value and only then, if required, action value. In doing so, he hoped to open a space for some sentences of the Upaniṣads not to have an action value at all, and in this he argued against Prābhākaras on one end and fellow Vedāntins on the other. In most arguments he joined forces with Ma ḍana Miśra, but there were several crucial points at which he parted ways with him as well. Let us conclude with a short analysis where Śa kara differed from the Mīmāṁsā account presented above. The emphasis on the informative nature of the Veda as a canon opened up the possibility to treat the arthav da passages as having full and independent validity just like the Vedic injunctions.119 This in effect allowed Śa kara to affirm that the Veda, though apauruṣeya, comes from Brahman as an omniscient and omnipotent being, a doctrine having roots in the Upaniṣads and one which Mīmāṁsakas did not accept.120 As a Vedāntin he also upheld the doctrine that the world is created from and annihilated in Brahman. Mīmāṁsā did not share this doctrine either. Kumārila famously claimed that the simplest assumption is that the world now is as it has always been and that the doctrine of creation and annihilation is an arthav da which has 118 pratyakṣaṁ hi śrutiḥ, prāmā yaṁ praty anapekṣatvātś anumānaṁ sm tiḥ, prāmā yaṁ prati sāpekṣatvāt. BSBh 1.3.28, I.191. 119 See BSBh 1.3.31 and BĀUBh 1.3.1. I will treat this issue in detail later in the dissertation. 120 This is developed in BSBh 1.1.3, where BĀU 2.4.10 is given as referenceŚ asya mahato bhūtasya niḥśvāsitam etad g-vedo … Note also tasya mahato bhūtasya niratiśayaṁ sarvajñatvaṁ sarva-śaktimattvaṁ ceti, “That great being is absolutely omniscient and omnipotent.” Such understanding, in fact, places him dangerously close to the Naiyāyika account of the Veda, because, as I will show later, Brahman that has the properties of omniscience and omnipotence is precisely Īśvara in Śa kara’s theology. Kumārila argued against the doctrine that the Veda might originate from an omnipotent being in ŚV Codan . 94 given rise to creationist philosophies.121 This acceptance of periodical creation opened up the possibility for taking at face value those passages of the Veda which present the Veda as a form of revelation. The Veda is eternal, but it is intuited by Prajāpati at the beginning of the new cycle of creation. Because words in the Veda refer to universals, Prajāpati creates the manifold that is the world against the Vedas as the blueprint.122 Such doctrine does not require taking the creation account as historical events, because the individuals depicted are not really individuals, but posts. The Vedic gods are also accommodated in this scheme: they may be born and die with the creation and the destruction of the world, but they are really posts occupied by different individuals in different cycles of creation: Indra the king of heaven is like Milka the cow. The Vedas, then, enter humanity when they are intuited or cognized by ṛṣis through the power of their austerity.123 There were other differences concerning the role of yogic perception, the nature of smṛti etc., but they need not concern us here. We can, thus, legitimately describe the Veda as revelation in Śa kara’s Vedānta, unlike in Mīmāṁsā, but we ought to note how this idea is never related to that of ruti. Śa kara’s acceptance of the arthav da texts as potentially authoritative has made it possible to identify ruti with the whole Vedic canon and to consolidate the several terms such as stra, abda, ruti, such that we can hardly notice any distinction among them when they are used in regard to 121 “Mīmāṁsakas, however, in this and any other regard, do not accept anything more than what is evident.” mīmāṁsakaiḥ, punaḥ, idānīm iva sarvatra d ṣ ān nādhikam iṣyate. ŚV Codan 98-9. The doctrine that the accounts of creation and destruction in the arthav da passages, such as those in the Upaniṣads, are the origin of Brāhma ical creationist philosophies is proposed in the TV 1.3.2. For a similar understanding in Bhart hari, see Aklujkar 1991. 122 See Holdrege 1994 and 1996, Chapter 1, particularly pp. 43-62, for a collection and analysis of Brāhma a passages in which Prajāpati intuits the Veda through some interplay of mind and speech and creates the world with the Veda as its blueprint or by uttering vocalized speech. “The Vedic mantras, as the expressions of the divine speech of Prajāpati, are depicted in the Brāhma as as part of the very fabric or reality and as reflective of the structures of the cosmos.” (1996Ś56) This doctrine has some striking similarities with Plato’s demiurge in Timaeus creating the world by intuiting the eternal forms. 123 The whole account is developed in BSBh 1.3.28-31; we need not quote these but note that the support that he finds for most of this comes from smṛtis, mainly Manu, and Mah bh rata. 95 the Veda.124 But the central unifying feature persisting through the complex from Mīmāṁsā to Śa kara was that of pratyakṣa-like validity, and this comes out clearly from Śa kara’s insistence that there isn’t anything that can disprove the truth of things corresponding to notions conveyed by the words of scripture, got from hearing. Scripture relates to supersensible matters and it has a peculiar way of imposing upon our understanding.125 The Vedas are ruti not because they were heard by the ṛṣis, but because men learn from hearing them. In a sense, men through hearing the Veda recreate the vision that the ṛṣis had at the dawn of creation. This, in the words of the Bh gavata, is the rutekṣita-pathaḥ puṁs m, the way of seeing through the ears for men.126 Where Śa kara was strikingly different from his Mīmāṁsā predecessors was his relentless insistence on using ruti as the preferred term for Vedic evidence and the extent to which he depended on such Vedic quotations, incomparably more so than Mīmāṁsakas. Drawing on Walter Slaje’s thought, we may claim that it was a feature of Śa kara’s method to bombard the reader with endless quotes so as to create the impression of samanvaya, a unison of Vedic evidence supporting his point, strikingly different from the Mīmāṁsā context-sensitive quoting.127 Another reason for this surely was his agenda to drive home the point that the knowledge-texts of the Veda are equally ruti because they equally do not presuppose a prior experience and impose upon our understanding with the direct sense of their words: Also, because matters of knowledge are similar to matters pertaining to ritual. That the Darśa-paur amāsa ritual produces a particular result, has a specific procedure and subsidiaries which need to be performed in a specified sequence is a super-sensuous matter, not in the domain of perception, and we understand them as such from the sentences of the Veda. Likewise, that realities such as Paramātmā, Īśvara, devatās, are not 124 For instance, in BĀUBh 1.3.1, ruti, rauta- bda, stra all seem to refer to the Veda interchangeably. 125 See, again, his comment on BĀUBh 1.3.1. 126 Bh gṬ 3.9.11. 127 Slaje 2007Ś152Ś “Pūrva-Mīmāṁsakas applied the exegetical principle of 'context' (prakaraṇa) for interpreting Yājñavalkya's formulations. The correctness claimed by Advaita-Vedāntins for their interpretation of BĀU passages was based on a maximum of matching quotes associatively accumulated from as many different ruti and smṛti text- places as possible.” 96  gross and the like, are beyond hunger and thirst etc., is made known only by the sentences of the Veda, ‘because they are supersensible, they ought to be like that [as taught in the Veda.]’ And, there is no difference in the manner that texts about action and about knowledge impress themselves upon the understanding.128 128 kriyārthaiś cāviśeṣād vidyārthānāṁ yathā ca darśa-paur amāsādi-kriyedam-phalā viśiṣ aitikartavyatākā evaṁ- krama-prayuktā gā ca ity etad alaukikaṁ vastu pratyakṣādy-aviṣayaṁ tathā-bhūtaṁ ca veda-vākyair eva jñāpyate. tathā, paramātmeśvara-devatādi-vastu asthūlādi-dharmakam aśanāyādy-atītaṁ cetyevam-ādi-viśiṣ am iti veda- vākyair eva jñāpyate, ity alaukikatvāt tathā-bhūtam eva bhavitum arhatīti. na ca kriyārthair vākyair jñāna-vākyānāṁ buddhy-utpātakatve viśeṣo 'sti. BĀUBh 1.3.1, VIII.39. 97  CHAPTER TWO: THE M M S MODEL OF PURU RTHA AND THE ROLE OF SCRIPTURE No intelligent person would do an action which brings no felicity, even if told to do so by hundreds of sentences.1 Introduction We saw in the previous chapter how the broad field of Vedic theology shared some basic presuppositions about the validity of the Vedic canon. I also noted that Mīmāṁsakas ultimately took the validity of the Veda to pertain only to ritual action, and then argued that Śa kara’s project of reinterpreting the Veda had to begin by challenging this thesis. Advaita Vedānta, however, shared another important presupposition in Vedic theology, and that was the notion that the purpose of the Veda was to serve human needs and to provide for some good that is desirable to man, and Śa kara’s rethinking of the Veda as knowledge qua knowledge and not action was not an independent project in epistemology: it was meant to show that understanding was more basic than action to the good that the Veda can provide for man. This radical rethinking, moreover, depended thoroughly on general categories in the discourse of “the good of man” in Mīmāṁsā, such as goals and means (s dhya and s dhana). Śa kara’s reorganization of the Veda around the goal of liberation, furthermore, used some specific categories in what I call “Mīmāṁsā ritual causality,” such as the notions of r d- upak rakas and sannip tyopak rakas, which for the time being we will translate as direct helpers and aggregated helpers, and relied on novelties in the Mīmāṁsā discourse introduced by Kumārila, specifically the notion of p ramarya or mediate ritual causality. Next, Śa kara’s rethinking of the role of meditation as the Vedāntic means of liberation was directly influenced 1 na ca buddhi-pūrva-kārī puruṣaḥ puruṣārtha-rahitaṁ vyāpāraṁ vacana-śatenāpy ukto 'nutiṣ hati. TV 2.1.1, I.383. 98 by two types of Mīmāṁsā injunctions, niyama and parisaṅkhy or restriction and exclusion, and his core soteriology was predicated on Kumarila’s philosophy of language. Finally, the very notion of mah -v kya was modeled on what Mīmāṁsakas called prayoga-vidhi or ritual manual and depended on the Mīmāṁsā idea of completion of meaning. In other words, in his radical rejection of the Mīmāṁsā understanding of the Veda, Śa kara shared the basic presuppositions about the validity and use of the Veda, and constantly fell back on employing Mīmāṁsā categories. In this chapter, therefore, I will present a coherent account of the Mīmāṁsā model of puruṣ rtha or “the good of man” and its relation to scripture. I will focus on the brand of Mīmāṁsā that developed under the influence of Kumārila, who had a much more positive role in Advaita Vedānta, and conclude the chapter with the key differences in the school of Mīmāṁsā that was associated with the other great post-Śabara Mīmāṁsaka, Prabhākara. Throughout the chapter, we will negotiate between historical developments and permanent deep structures in Mīmāṁsā, focusing more on the second. The chapter will provide us with the basic challenge for Advaita Vedānta, one that we will see develop further through chapters three through five, but it will also serve as our indispensable resource to which we will often refer for understanding Śa kara in the later parts of the dissertation. The chapter does contain some important discoveries, particularly concerning Kumārila’s idea of mediate causality, and offers probably the first serious engagement with the Mīmāṁsā understanding of the Upaniṣads, but originality is not its goal. There is no “big argument” that is limited to the chapter, and its purpose is to serve the larger needs of the dissertation. Dharma and Puru rtha One of the key presuppositions in Mīmāṁsā was that the Veda was like a manual of “ritual technology” whose purpose was to secure something desirable to man. To Staal’s old question, 99 why Vedic people perform ritual, Mīmāṁsakas would have answered, for the same reason that they do anything at all: happiness.2 Ritual for Mīmāṁsakas was, then, something on which some human good depends, and both the ritual as the means and the good which it brings were felicitously covered by the Sanskrit term puruṣ rthaṬ Puruṣ rtha is a compounded Sanskrit word that can be interpreted in two ways, as a tat-puruṣa or a determinative compound, in which case the meaning would be a human need or purpose, and as a bahu-vr hi or a possessive compound, in which case the meaning would be that which is for some human need or purpose. The compound, in other words, could stand at the same time both for a goal or result (s dhya or phala) and for the means or instruments of achieving that goal (s dhana).3 Defining puruṣ rtha in the instrumental sense, Śabara saysŚ “That on which human happiness depends, i.e., that which when done man becomes happy, that is the category for the good of man (puruṣ rtha). How so? Because it exists on the account of the desire to obtain, not on account of scripture. … Since for the good of man is not separate from happiness, it in fact constitutes the means of obtaining happiness.”4 We will have occasion to unravel the details concerning the instrumental nature of puruṣ rtha in this chapter, but it is important to note here that Mīmāṁsakas—at least those of the later days—understood the Veda to be humanly 2 Staal 1979. As indicated above, by “Mīmāṁsā” here I have in mind the line of Śabara and Kumārila. As will be clear at the end of the chapter, Prābhākaras had a different understanding of dharma. 3 On these two types of compounds see Tubb and Boose 2013:96-125 and 127-37; Abhyankar 1961:179-80 and 283; Whitney 1992:489-494 and 501-11. 4 yasmin prītiḥ puruṣasya, yasmin k te padārthe puruṣasya prītir bhavati sa puruṣārthaḥ padārthaḥ. kutaḥ … tasya lipsārthena ca bhavati, na śāstre a ... avibhakto hi puruṣārthaḥ prītyā. yo yaḥ prīti-sādhanaḥ sa puruṣārthaḥ. MSŚBh 4.1.1, IV.1194. 100 centered:5 that men want happiness was objectively determined, through natural desires.6 The purpose of the Veda was human happiness, and the questions were, what kind of happiness can the Veda bring for man, and how. To answer this, we need to delve a little into Mīmāṁsā specifics and its technical language. Mīmāṁsā as a knowledge system, in fact, developed around the Veda and what early Vedic theologians saw as the central concern of the Veda, an idea which they called dharma and the definition of which was given right at the opening of the MS (1.1.1-2)Ś “Now, therefore, an inquiry into dharma. Dharma is that thing which is known from a Vedic injunction.”7 For our purpose here it is superfluous to investigate the development of this concept through its history, since the Bhā a Mīmāṁsā understanding of dharma was quite straightforward: it is what brings some good ( reyaḥ-s dhanat , reyas-kara), namely human happiness (puruṣa-pr ti). A lucid account that explicates how dharma was supposed to brings some good is found in Durga’s commentary on the s kṣ t-kṛta-dharm na ṛṣayaḥ from the Nirukta that we saw in the previous chapterŚ “Seers are those who see that from a certain action combined with a mantra referring to a particular thing there appears a transformation in the form of a result which has particular 5 Clooney had famously argued for “Jaimini’s decentering of the human” and claimed that “even if Jaimini takes seriously into account the human perspective … [he] locates the intelligibility of the sacrifice in the sacrificial element itself and not in the human person.” (1990Ś163) See also Clooney 1987. Clooney’s thesis certainly has much to recommend itself—Mīmāṁsā teleology can be likened to a system of merit where personal factors are excluded—but come Kumārila and the system is driven by the needs of the individual and not of the ritual action itself. The “decentering of the human” seems to me to be rather for the purpose of avoiding the pitfalls of human frailty and the contingencies of personal agency, not for the sheer joy of action, its obligatory nature or the securing of its continual performance (Clooney 1987:664), such that the expected result would inevitably follow for human good, because there would be no point at which the process could fail. This is not unlike the removal of personal agency from the Veda so that its validity could not be questioned at any point and its knowledge would be perfect. The “decentering of the human” is for securing the personal firmly in the center. In any case, whether Clooney is right about pre-Śabara Mīmāṁsā (his argument explicitly goes only that far) or not, Kumārila’s “centering of the human” can hardly be questioned. 6 Cf. Śabara on MS 3.1.7Ś “In the world, action is determined by needś in the Veda, by the word.” loke karmārthaṁ lakṣa am, śabda-lakṣa aṁ punar veda iti. 7 athāto dharma-jijñāsā. codanā-lakṣa o 'rtho dharmaḥ. 101 characteristics.”8 Dharma is, in other words, a means of human good which appears when things are used and arranged in particular ways, not knowable otherwise than from the Veda.9 This property of dharma belongs to a Vedic sacrifice as a unit organized around the central role of ritual action, but to its elements as well, such as the stuff that one offers or the sacrificial implements one uses, because under some circumstances they individually could produce a result which would constitute human happiness.10 While we do not need to consider details, we should note well the instrumental nature of dharma. For a Bhā a Mīmāṁsaka, dharma is never one of the goals of human life; it is a puruṣ rtha insofar as it an instrument of achieving such a goal.11 The specific nature of dharma as one means of human happiness is that it deals with laws of causality which are not known otherwise than from the Veda. This is, perhaps, best illustrated by the stock example of how individual items used in a sacrifice independently produce some form of human felicity. When during the performance of the Darśa-pūr amāsa rituals the sacrificer uses a milking vessel (go-dohana) rather than the regular vessel (camasa), this change brings added value to the sacrifice: the sacrifice generally performed for the attainment of heaven now brings cattle in the future as well, because the use of the regular vessel is sufficient for the attainment of the primary result. The milking vessel, thus, obtains independent instrumentality regarding human happiness: it is not anymore for the good of the ritual (on which more in a bit), 8 ṣyanti ye amuṣmā karma a evam-arthavatā mantre a saṁyuktād amunā prakāre aivaṁ-lakṣa aḥ phala- vipari āmo bhavatīty ṣayaḥ. Durgācārya’s commentary on Yāska’s Niruka 1.20, p.90-1. 9 Dharma is described as reyas-kara in Śabara’s Bh ṣya right under MS 1.1.2, while Kumārila’s term is reya- s dhanat (ŚV Codan 13-14ś Kataoka’s comments on verse 14 are most illuminating [2011bŚ208-9]). 10 ŚV Codan 13-14 and 190-1. 11 The occasional practice, therefore, to translate dharma as “duty” (as in Jha’s translation of loka-V rtika [1907], or Edgerton’s translation of the padev [1929]) in the Bhā a Mīmāṁsā context is misleading. 102 but is for the good of man.12 But, more important for us, this causal relation that obtains between the use of the milking vessel instead of the regular vessel and the future receiving of cattle is knowable from the Veda. The vessel is just an odd thing in the world with characteristics that we otherwise know—shape, material etc.—but that same thing has the additional characteristic of dharmat or procuring happiness in a way known only from the Veda. The Vedas, then, as far as they do inform about anything, do not inform so much about things as about causal relations. But these causal relations are not matter-of-fact relations, and there is nothing in our experience that would allow us to know about them. This, then, constitutes the specifica of dharma and the Vedas: dharma is a supersensible relation between a thing or an action and a future state of affairs agreeable to men, knowable from the Vedas. We ought to note that this is a future and producible state of affairs (bh vya and k rya). It is in the domain of becoming, not of Being. Furthermore, that these relations are knowable from the Veda is not what the Veda is really about. What matters is not so much that they are knowable, but that they are harnessable. Two further notes are in place before we go into the details of what I call here the “ritual technology” or “harnessability” of the Veda. First, while Kumārila explicitly attributes the property of dharma to ritual implements and material offerings individually, dharma is primarily associated with complex rituals in which the primary, good-producing factor is the ritual action of sacrifice, while the ritual implements, the material offerings and the other ritual elements are subordinate and only assisting in the production. 12 See Śabara on 4.1.2 for the godohana being puruṣ rthaṬ Cf. Jha 1964:260Ś “[T]he use of the Milking Vessel is Puruṣ rtha because, while the mere act of waterfetching could be done in other vessels also, the particular result, obtaining of cattle, could be obtained only if the water were fetched in the milking vessel.” 103  Second, the results that these sacrifices produce are of diverse kinds, such as wellbeing in general (bhūti), cattle (pa u), kingship (r jya), etc., but the common result that is usually presented as the result of a ritual performance and the end-factor in a Vedic injunction is “heaven” or svarga. As the Mīmāṁsā mantra goes, svarga-k mo yajetaŚ “He who wants heaven should sacrifice.”13 Now, by heaven Mīmāṁsakas did not mean a place one goes to after death, such as the one described in the Vedic corpus or in scriptures broadly. Śabara summarizes this popular view in the comment to MS 6.1.1Ś “Heaven is a specific place where there is neither heat nor cold, neither hunger nor thirst, no unpleasant stuff, no sorrow, to which go after death those who have led virtuous lives and not others.”14 While it is certain that Śabara rejects this view, it is not certain why; part of the reason, however, must be that heaven as a place is not a producible thing. If heaven is not a place one goes to, it is also not a thing of the world which is generally considered pleasurable and which people tend to describe as heavenŚ “For ascertaining the meaning of all words, we must depend on common usages; now, we find that in common usage, the word 'heaven' is used to refer to substance: 'Fine silken clothes are heaven; sandal-paste is heaven; sixteen-year-old girls are heaven.' Whatever pleasurable substance there is, that is heaven.”15 Appealing as it may be, this view is rejected by invoking a principle that has had wide acceptance in Indian philosophy: something is not what we think it is or what it appears to be unless it is that all the time. This is the principle of absence of vyabhic ra, or deviation. A substance is pleasurable in some cases but not in others, whereas happiness as a state is 13 Lest I be jumped at by expert Indologists and Mīmāṁsā specialists, it is necessary to point out here that it is the English word mantra that is intended. 14 nanu svarga-śabdo loke prasiddho viśiṣ e deśe. yāsmin noṣ aṁ, na śītaṁ, na kṣud, na t ṣ ā, nāratiḥ, na glāniḥ, pu ya-k ta eva pretya tatra gacchanti, nānye. MSŚBh 6.1.1, IV.1347-8. 15 sarveṣām eva śabdānām artha-jñāne laukikaḥ prayogo 'bhyupāyaḥ. tasmiṁś ca laukike prayoge dravya-vacanaḥ svarga-śabdo lakṣyate. kauśikāni sūkṣmā i vāsāṁsi svargaḥ, candanāni svargaḥ, dvy-aṣ a-varṣāḥ striyaḥ svarga iti. yad yat prītimad dravyam, tat tat svarga-śabdenocyate. MSŚBh 6.1.1, IV.1347. 104 happiness in all cases when it obtains. It is for this reason that the word “heaven” can only directly apply to happiness as a state, not to any substance that causes happiness. To summarize, if svarga is not an actual place in the hereafter, it is no Fred Astaire heaven either. By an extension of the above-stated principle of absence of vyabhic ra or deviation, it is eminently possible to infer that Mīmāṁsakas thought of heaven as a state of some uninterrupted happiness, happiness undetermined by fleeting causes, producible by the performance of sacrifice. We will revisit the idea of heaven in the next chapter, but now we turn to the manner of organization of the ritual—and its correspondence in the text of the Veda—which guarantees the attainment of the result, the “ritual technology” of Mīmāṁsā. Dharma and Ritual Causality I opened this chapter with presenting puruṣ rtha as a means of human happiness, and we saw that in its deepest structure, puruṣ rtha and dharma had a similar nature: dharma was a means of human happiness very much like wealth, but it was specifically a Vedic means, that is, a means that harnesses Vedic laws of causality. We also saw that dharma in the restricted sense referred to Vedic ritual. To be even more specific, dharma was the principal ritual in a complex ritual performance, the part in which the offering is made, not, however, in isolation, but in its feature of teleologically organizing the whole ritual and harnessing it for the obtainment of the expected result. In this most restricted sense, dharma was the instrumental puruṣ rtha as the principal act (pradh na) in a ritual to which everything else was subordinate ( eṣa/aṅga/guṇa, an auxiliary), but which was itself subordinate to the expected result such as heaven, which result in its turn was subordinate to man (the non-instrumental puruṣ rtha).16 Dharma organizes the ritual 16 I use the pradh na- eṣa pair here, but Mīmāṁsā uses several other terms, mukhya for the first and guṇa and aṅga for the second. See Clooney 1990-98-100. See also TV 3.1.2, II.653Ś “Because the word eṣa can express many 105 through this principle of subordination, and its success depends on properly understanding the principle and on working out its details so that the result would be achieved. The principle of subordination is quite straightforward, and it is stated in MS 3.1.2: one thing is auxiliary to another when it is for another.17 Śabara first explains being for another as doing service for or providing help to another, upak raka, but quickly proceeds to define it as complete dependence on another, such as that of a donkey or a slave on the master. The second was, really, the essential characteristic in Kumārila’s eyes. A man may drink water from a canal that has been dug up for irrigation, and that may help him every now and then, but it does not make the canal for drinking. The canal was dug for irrigation; irrigation is its telos, and the other helpful uses are merely incidental. The master-slave relation is, again, a useful illustration: while the master may do things for the slave, such as feeding him, he does not do them for the slave’s sake, but for his own benefit. An auxiliary is something that is for another and has no cause or ground or justification of being elsewhere than in the other.18 Let us see, now, how this principle of subordination works in a ritual and how the Veda corresponding to the ritual as its backbone is teleologically organized. A principal or superordinate element in a complex ritual performance must satisfy several criteria. First, as we saw in the beginning of the chapter, it must be directly related to human felicity: its performance or use must be objectively occasioned by man’s desire for happiness, not by the requirements of the ritual itself.19 Second, the principal is not principal absolutely; it is subordinate to the result— it is for the sake of the result—which in its turn is subordinate to man. However, third, since its meanings, such as ‘excess,’ etc., here it is used as a synonym for aṅga, guṇa, dharma etc.” śeṣa- śabdasyānekādhikādy-artha-vacanatvād ihā ga-gu a-dharmādi-paryāya-vācitva-parigrahārtham. 17 śeṣaḥ parārthatvāt. 18 na śeṣo ’nyaḥ parārthatvān na ca hetv-antare a saḥ. TV 2.1.5, p.533. 19 See MS 4.1.1 and 4.1.5 and Śabara thereon. 106 being for something else is objectively determined, in the context of the ritual performance it is not subordinate to anything, and everything else is subordinate to it. This is in virtue of the fact that only the principal is directly related to the expected result of the ritual, while everything else is vicariously so: it is the principal that is fruitful.20 Fourth, the principal is typically an action—it is what someone does in the ritual and it is not the sacrificer himself, the offering, or any odd ritual element; specifically, it is the action of sacrificing, pouring into fire or giving that constitutes the central element of the ritual. It is that part of the sacrifice in which, for instance, clarified butter ( jya) or the sacrificial cake (puroḍ a) are offered.21 Fifth, it is not necessary that there be a single principal. In some of the major rituals, in fact, there are several principal ritual performances; they all must be, however, fruitful, directly related to the result of the ritual as a complex.22 That an action is independently fruitful is determined by the application of the Mimāṁsā principles of scriptural interpretation.23 We will recall that the definition of an auxiliary was “being for another.” In the ritual context, an auxiliary is that which is for the good of the ritual, kratvartha in Sanskrit. Unlike the subordination of the principal, puruṣ rtha, which is objectively determined by the need of man, a kratvartha is subordinate on scriptural grounds. In other words, that something serves the needs of the ritual action is so because there is a Vedic sentence which institutes such a fact, or because 20 “That which is connected with the result is the principal, while that which is related to the principal is the subsidiary.” pradhānaṁ phala-saṁbandhi tat-saṁbandhy a gam iṣyate. TV 2.1.1, p.266. 21 MS 4.2.27 defines these actions as those that bring about a connection between the substance that is offered and the deity to which it is offered: yajati-codanā dravya-devatā-kriyaṁ samudāye k tārthatvāt. The general term is “to sacrifice,” but it refers to all forms of principal action, which Śabara lists as three: sacrificing, pouring and offering, yajati dadāti juhotīty evaṁ-lakṣa am. The next sūtra defines pouring or homa as identical with sacrificing, with the aditional element of pouring liquids, tad-ukte śrava āj juhotir āsecanādhikaḥ syāt. Śabara defines “giving” in the comment as giving up one’s ownership and establishing a relation to the ownership of another, ātmanaḥ svatva- vyāv ttiḥ parasya svatvena sambandhaḥ. Giving up (utsarga) is the common element of the three. 22 See, for instances, MS 4.4.34 and Śabara thereon, as well as the whole adhikara a, 4.4.29-38, in which the principle of fruitfulness is associated with a sacrifice being a primary and is ascribed to several distinct sacrifices within the complex Darśa-pūr amāsa because all of them are said to be fruitful. 23 See Śabara on MS 4.1.5. 107 it is otherwise possible to interpret a Vedic text to that effect. Given that the principal element of the ritual is the ritual action or kriy , the factors that most immediately participate in this action, called k rakas, perforce constitute auxiliary elements: the material offerings (dravya) such as the clarified butter and the sacrificial cake are auxiliaries in the sacrifice, as is the sacrificer himself. Though in an important sense the sacrificer is the ultimate principal of the sacrifice, since its result is for him while the sacrificial action is for the result, he is so as the enjoyer (bhoktṛ) of the results of action. As the agent (kartṛ), however, he is subordinate to the action. The divinities (devat ) to which the offerings are made are also auxiliaries to the ritual action, since they are not the beneficiaries of the ritual results but serve as mere nominal recipients of the offerings that the ritual as a sacrificial action must have in order to be a ritual.24 Chatterjee puts this rather nicelyŚ “[I]n a sacrifice, the deity is as important as a guest in the context of the act of hospitability.”25 No exhaustive account of this is necessary, however, and we need to note just two things. (1) The central elements of a sacrifice are of two kinds, namely puruṣ rtha and kratvartha; the first is directly related to and productive of the expected result, while the second directly serves the purpose of the first. (2) The principle of subordination is not exhausted with that, for the kratvartha elements have other things subordinate to them. More important than this—and we need to note this point very well—is that not all actions in a ritual are primary actions. In fact, Mīmāṁsakas have grouped all ritual actions broadly in two but more specifically in three categories, in virtue of what it is that they produce. The twofold division is that between principal and auxiliary actions, and the criterion of division 24 Clooney 1988 is a fine analysis of the role of devatās in Mīmāṁsā. They are purely textual figures, non-corporeal and not actually sitting in sacrifices, and their existence is “a strictly linguistic requirement, ensuring the intelligibility of the sacrifice through what is said …” (p.283). 25 1992:171. 108 is whether the action is concerned directly with ritual elements—the agent, the offerings, etc.— or not. Closely tied with this is the nature of their result: (1) is it something immediately given or visible (dṛṣṭa); or (2) is it something which is not visible (adṛṣṭa) but must be postulated so that a part or the whole ritual will make sense. Let us expand a bit on this. There are, to begin with, some ritual actions that produce very visible results (dṛṣṭa) and their shared characteristic is that they operate over ritual elements that are themselves already auxiliaries in the ritual complex. These actions have been grouped under four headings: (1) origination or utpatti, for instance when a rice paddy is made; (2) obtaining or pr pti, for instance when milk is got from a cow; (3) change or vikṛti, for instance when the solid clarified butter is melted; and (4) refinement or saṁskṛti, for instance when the rice paddy is sprinkled with water. We note, first, that they produce their result by operating on some of the ritual factors, for instance rice paddy the offertory. In doing so, they are auxiliaries to what is already subordinate to the principal ritual action of offering, so they themselves are auxiliary actions. We note, second, that the result of most of them is immediately evident: the action results in a rice paddy or milk. We note, third, that the fourth category, refinement, is the only one among the four which can produce an unseen result, an excellence added to the consecrated item in the form of suitability for ritual use, which is not necessarily empirically noticeable, as in the case of the consecration of the sacrificer. This fourth group also includes some intermediary actions with visible refinement as their result, such as threshing (avahanana) and grinding (peṣaṇa), and the blanket term that Śabara uses to refer to all forms of refinement, visible and invisible, is saṁsk ra, or “preparation.”26 On MS 2.1.6. Saṁskāra, of course, has felicitously a semantic range much broader than that or refinement. See 26 Kataoka 1999 for a most accessible account of saṁsk ras. 109  These four have come to be commonly called sannip tyopak rakas or aggregated helpers, in view of their fourth crucial characteristic, namely that their causal efficacy is absorbed by and terminated in the auxiliary ritual elements. This last feature of theirs is an instantiation of a general principle of Mīmāṁsā accounts of ritual causality, which is worth spelling out: causal efficacy terminates in the produced result and is vicariously carried over in the further process. Once this has happened, the action ceases being a means. I will have occasion to say a more about this below; as for the sannip tyopak rakas, to sum up, they are actions which produce visible or somehow palpable results and are subordinate to the ritual elements, which themselves are kratvartha or for the good of the ritual.27 Their relation to their superordinate element is, however, immediate, and they are absolutely required for the result to obtain. The second category of actions are the offertorial actions that we described as principal. Their differentia specifica is stated negatively: they are not meant to culminate in the preparation or the production of some substance.28 They are not, in other words, one of the four sannip tyopak rakas whose causal efficacy extends up to their immediate superordinate ritual element, and they are recognized by the fact that they do not produce anything visible or tangible. Put differently, if what they produce immediately is accepted as their actual result, that would render the sacrifice purposeless. The visible result of the offertorial actions is ashes left from the sacrifice, but no one would perform a sacrifice to end up with ashes. Therefore, an invisible future result (adṛṣṭa) is postulated as their telos, such as heaven, wellbeing, or cattle. 27 There are minor differences of opinion between the Bhā as and the Prābhākaras about them, concerning mainly how they are instituted by the Vedas, into which we need not go. See Prakaraṇa-Pañcik chapter 13 in Pandurangi 2004. My account of the nature of the sannip tyopak rakas is based on TV 1.2.7. 28 MS 2.1.7 and Śabara thereon. 110  However, in some complex Vedic rituals, several such offertorial actions are performed, and not all of them are said to bring a result. Therefore, a further division of these offertorial actions is introduced. Some are principal actions in the full sense, because they bring the ultimate result of the sacrifice.29 Other offertorial actions, however, are mentioned in the proximity to the fruitful actions, but are unrelated to a result. They are considered auxiliaries to the principal actions in the full sense.30 They are also principal in a way because they are full ritual acts, to which ritual elements such as the sacrificer or the offerings are auxiliary—in form they are hardly distinguishable from the principal proper—but they do not bring heaven or cattle. Therefore, by the application of the Mīmāṁsā principles of interpretation, they can be related to a principal proper and serve its purpose rather than the immediate good of man.31 They are, thus, for the good of the ritual (kratvartha), while only the principal proper is for the good of man (puruṣ rtha). Unlike the sannip tyopak rakas which are removed from the principal by the ritual element to which they are directly subordinate, such as the sacrificial cake or the sacrificer, they are directly related to the principal. For this reason, they have received the appellation of r d-upak rakas or direct helpers.32 However, they are related to the principal through postulation of contribution to its results, and not naturally and indispensably, like the sannip tyopak rakas are related to their superordinate element. Although the second are removed from the principal by their superordinate element whereas the first are not, they express a closer, essential relationship. We ought to note this concept very well, as it will play a crucial 29 MS 4.4.34: tat punar mukhya-lakṣa aṁ yat phalavattvam; that is characterized as the primary which is fruitful. 30 tat sannidhāv asaṁyuktaṁ tad-a gaṁ syātŚ that which is in its proximity but unrelated [to a result] is its subsidiary. Ibid. 31 Śabara on MS 4.1.5. 32 See, for instance, MSŚBh 2.2.3, II.483 on the distinction between the primary actions, which are directly related to the result, and their direct helpers, which are close to the primary which is fruitful. “Therefore, they [the primary actions], are related to the result; because the gh ra and other sacrifices are close to the [primary] which is fruitful, they are its direct helpers.” tata eṣāṁ phala-sambandhaḥ, phalavat-sannidhes tv āghārādīny ārād-upakārakā īti. See also TV 2.1.7. 111 role in Śa kara’a making sense of the role of ritual on the one hand and renunciation on the other in the pursuit of liberation. To sum up, then, all actions performed in a Vedic ritual are of broadly two and specifically three kinds: (1) some are auxiliary outright as they operate over ritual elements, which are themselves subordinate to the principal in the ritual, and are called sannip tyopak rakas; (2) others are principal outright, the offertorial fruitful actions, and are called puruṣ rthaś (3) finally, there are the offertorial non-fruitful actions, principal to ritual elements yet subordinate to the puruṣ rtha actions; they are called r d-upak rakas and are kratvartha. That the r d-upak rakas were not considered fruitful did not mean, of course, that they do not contribute anything to the ritual. That would render them purposeless, the worst nightmare of a Mīmāṁsaka. We will address this a little later in the chapter, but now I want to speak to a more pressing question. The Doctrine of Ap rva and the Temporal Unity of the Ritual I talked about the sacrifice as a ritual complex that consists of different types of elements, subordinate and superordinate actions, and a single ultimate result. We also discussed the principle of subordination that provides hierarchy to this ritual complex. However, we did not see how it is that these ritual elements enter a complex in the first place. What is it that aligns so many distinct actions and elements into a coherent whole, a whole which is not a sum of its parts as far as it is expected to end in a single result? There are, in fact, two aspects to this question. The first concerns the structural unity of the ritual complex: how and why is it the case that diverse and discrete actions, offertories and other ritual elements can become a whole in respect to a result? The ritual is an actual performance in history, so on the one hand this aspect concerns the facticity of the ritual: how are 112 the distinct actions, elements, and results in an actual performance constitutive of a whole? The ritual is also normatively laid down in the Veda, and so its structural unity concerns its textual ideality as well. The second aspect concerns the temporal unity: how is it the case that temporally circumscribed actions, many in number and different in kind, produce a single result, and one that is not immediately visible at the end of the complex ritual? Organizationally, it is more convenient for me to treat the second, temporal aspect first, and it is to this that we now turn. The ground of the temporal unity of the sacrifice was the doctrine of apūrva, a doctrine with a long history in Mīmāṁsā and a driving force behind many of Kumārila’s innovations. The unity of the sacrifice in the temporal sense was problematic on two counts. The first was because a sacrifice consists of several ritual actions organized on the principle of subordination and performed at different times. In a single sacrifice, it was possible to have several principal rituals, all of which would have had auxiliary rituals ( r d-upak rakas), some of which would be performed after the performance of the principal ritual. How is it that these temporally discrete and self-contained rituals produce a single result at the end of the sacrifice? Another (and more important) problem stemmed from the fact that heaven, cattle, or wellbeing obviously did not come at the very end of the ritual; the ritual action, generally described as the means in the ritual complex, would be completed, and it was not clear how it was related to the delayed result. How are the ritual and its result a unit? The traditional Mīmāṁsā answer to this was that a Vedic sentence that relates the ritual to a future result was a warrant enough for the causal relation to obtain. Let us expand on this. The general understanding of causality in Mīmāṁsā was quite down to earth and can be characterized as Humean. Śabara’s definition of a causal relation wasŚ “There is a causal relation 113 only there, where something is when another thing is and is not when that other thing is not.”33 For two things to be causally related, there had to be some contiguity between them. “It is only when an effect appears immediately after a cause that it is recognized as following from that cause.”34 Now, it was clear that Vedic causality cannot be quite like that, because the result is never seen together with the alleged cause. Vedic causal laws were different in kind: that a result follows the performance of a certain ritual is due to a Vedic fiat. An injunctive sentence of the Veda institutes into being a causal relationship between something ordinary, say the wood of khadira tree, and a result to which this ordinary thing is otherwise not related, say virility.35 It is no longer required that the cause and its effect be immediately related. “In the case of Agnihotra and similar acts, the causal relation is declared just by the words of the text; hence, even though the result is not seen at the time, there is the conviction that it will come in due course.”36 This principle was the rationale behind the notion of apūrva, “that which has no precedence.” The Vedic causal relationship is unprecedented, unique, empirically not known and therefore knowable only from the Vedas.37 33 On MS 4.3.2, IV.1246: kārya-kāra a-sambandho nāma sa bhavati, yasmin sati yad bhavati, yasmiṁś cāsati yan na bhavati, tatraiva kārya-kāra a-sambandhaḥ. 34 On MS 4.3.27, IV.1261: yac ca anantaram upalabhyate, tat tataḥ iti vijñāyate. 35 See MSŚBh 4.3.3 and the deliberations on the khādiraṁ vīrya-kāmasya yūpaṁ kuryāt injunction. 36 MSŚBh 4.3.2, IV.1246Ś agnihotrādiṣu tu śabdenaiva kārya-kāra a-sambandha ucyate. tasmāt tatra tat-kāle ‘d śyamāne ‘pi phale, kālāntare phalaṁ bhaviṣyatīti gamyate. 37 Clooney 1990:232-39 claims that apūrva for Jaimini refers to such “Vedic” elements in the sacrifice, the fact that it is sacrifice that brings one to heaven, not otherwise knowable, or that “one should husk the rice by beating it” for the sacrifice to be successful. We may say that whatever is determined by the Veda and introduced in the sacrifice is apūrva, unprecedented, at the instance of its introduction. Halbfass 1991Ś302 calls this the “prehistory of the classical Mīmāṁsā usage of apūrva,” and also mentions another, parallel line of development of the concept, in which apūrva is “an impersonal and substrateless (an rita) potentiality, a kind of cosmic principle or power to be manifested or actualized by the ritual acts.” A fuller history of this idea is available in Kataoka 2000, who calls it dharma-abhivyakti-v da or the doctrine of manifestation of dharma; see also Aklujkar 2004, particularly 281-5. The core of this doctrine is expressed by Bhart hariŚ “[D]harma-prayojano v refers to the view of the MīmāṁsakasŚ … dharma is already in place. It is only manifested (made operational) by such (rites) as the agni-hotra. Set in motion by them, it bestows the fruit, just as, in service, a master is moved by the servants … toward a result.” (See Aklujkar 2004:281 for the Sanskrit and reference.) 114  Kumārila, on the other hand, had different ideas about causality and apūrva. What was “unprecedented” about ritual was not so much that the relation between the ritual and its future result was not known “before the Veda,” but that ritual as a form of action had a fitness or faculty, capacity, that is absent before the action is performed and is, therefore, posterior to it.38 The novelty that gives its name to apūrva is not its sole knowability from the Veda, but the fact that prior to its performance the action is not capable of bringing about the result. Kumārila, further, claimed that apūrva was a general principle of cultivation and maturing that was common to processes of all kinds. Major actions in general, not only Vedic actions, do not produce results immediately, as is evident from cases such as agriculture and education. “Even ordinary worldly results, such as children, do not appear immediately, nor are they expected immediately.”39 Reflection on causality shows that observable contiguity is characteristic only of simple causal processes. Complex production requires maturation, in which the result is incipiently present immediately, but fully and visibly only after it has undergone several stages, and long after the cause has ceased operating in its initial form. Think about the The old Mīmāṁsā idea of apūrva being the novel thing introduced by and known through the Veda begins to change with Śabara and develops into the full-blown theory of Kumārila (see, again, Clooney 1990:221-53), where apūrva becomes the link that relates the sacrifice to heaven. Yoshimizu 2000 had controverted Clooney’s account and argued that Śabara did not introduce apūrva as meaning the link between the transient action and later heaven, but used the word in the old sense of that which is introduced by a Vedic injunction and knowable from it, organizing the sacrifice structurally but not temporally. This has been in turn disputed by Kataoka (2011b:18-9), but is of no concern for our account. Halbfass (1991:307, slightly reworked from Halbfass 1980) remains the most accessible account of Kumārila’s theory of apūrva. Kataoka 2011b:445-454 relates very successfully the short account of apūrva in ŚV Codan 196-200 with the much longer account in TV 2.1.5 and is certainly most accurate and thorough presentation of Kumārila’s understanding. 38 karmabhyaḥ prāg ayogyasya karma aḥ puruṣasya vā | yogyatā śāstra-gamyā yā parā sā’pūrvam iṣyate. TV 2.1.5, I.394. 39 aihikaṁ cāpi putrādi karma-janma-svabhāvakam | na karmānantaraṁ kaścid phalaṁ labdhuṁ vyavasyati. TV 2.1.5, I.397. 115 growth of a tree from the sprout or the production of yoghurt from milk. Heaven is not much different than these processes. We can describe it as the thickening of ritual action.40 It is, of course, Vedic causality in which things are linked in an uncommon, unobservable causal relation—kh dira wood and virility—and that is why the general process of maturation is not called apūrva, but only the Vedic one. Kumārila did not dispute the old, epistemological roots of the idea: it is from a Vedic injunction that one apprehends apūrva. But his understanding of apūrva was firmly situated in the ontological, and his point was that being Vedic was just the specific difference of the much more general nature of causality.41 All forms of complex causal relations involving maturation share also in epistemology: they are all knowable through postulation (arth patti) or positing of a fact to reconcile some other facts that do not sit well together on the evidence available. For instance, that a baby appears well after intercourse requires postulating a process of maturation that is not visible. The specific nature of apūrva is that we do not have precedence both on the part of seeing heaven, and of its causal link with the ritual action, which situates apūrva both in the realm of the unprecedented and of scriptural postulation. We know about heaven and sacrifice from the Veda, but attaining heaven by performing the sacrifice would be impossible without an invisible and trans-temporal link between the two.42 40 See TV 2.1.5, I.395. 41 Kataoka 2000Ś172 insists that apūrva for Kumārila had epistemological roots just like for the tradition following Prabhākara, and it did not mean something ontologically new. Prabhākara famously defined apūrva as m n ntar pūrva (Bṛhat …), and that much is not disputable. For Kumārila, Kataoka gives the negative evidence that he begins the pūrvapakṣa in TV 2.1.5 “presupposing the interpretation that apūrva is something epistemologically new or not known before.” (2000Ś180). Kumārila, however, does not have kind words for the theory of apūrva presented by the pūrvapakṣin—“Sure, this apūrva which you have concocted in your mind, as if it were something embodied, you can go on to refute, but that does not contradict anything on our part” (yad idaṁ svamati-parikalpitaṁ vigrahavad ivāpūrvaṁ bhavadbhir nirākriyate, na tenāsmākaṁ kiñcid virudhyate)—and he does not define apūrva epistemologically. While acknowledging that apūrva is knowable only from the Veda ( stra-gamy , veda-gamy ), the central element in his definition is that this apūrva does not exist before the ritual action is performed and is, thus, posterior to it. 42 Yoshimizu 1996 argues that the knowability of apūrva through postulation was also an innovation of Kumārila. 116  Kumārila described apūrva as a capacity of the ritual action, a akti, and it was important to emphasize that this capacity was substantially non-different from the action, for otherwise there was the danger that apūrva would replace dharma as the instrument of human happiness, or heaven as the goal. Apūrva was a mode under which the ritual action continues to exist after it has been completed, much as milk continues to be milk after acquiring the characteristics of sourness and thickness, and this feature of action made it both temporally circumscribed and somehow permanent. The category of akti allowed Kumārila to reap all the benefits of the claim that it is the ritual action which is dharma, without assuming the unwanted consequence of action not being evidently related to heaven as its result. The precise way in which apūrva operates is by leaving a mark or saṁsk ra on the Self of the sacrificer. The fact that the Self is eternal secures the permanency of the otherwise impermanent action. An action has been performed and completed, but the imperceptible gains from that action—and through that the action itself—reside in the eternal Self, because they have left a mark on it. This makes it possible to present the ritual as a trans-temporal unity of its separate constitutive actions that are performed at different times. The problem was, let us remember, that: (1) a ritual consists of a primary action to which several r d-upak rakas are subordinate, some performed before the main ritual but some after; and, (2) some of the large rituals have several primaries, often temporally quite removed from one another. The actual ritual actions are obviously non-contemporaneous, but the apūrva as the mark they leave— which, as Kumārila claimed, are the respective actions, in a different aspect—resides in the eternal Self, and in virtue of that they achieve simultaneity. All the main and subordinate rituals produce individual apūrvas, which combine to produce a final apūrva that matures and in time brings forth the intended result. We do not need to go into the details of how a final apūrva is 117 produced, but we do need to pay attention to one principle involved, and that is mediate causality. To this we now turn. The Idea of Mediate Causality An innovation of Kumārila that has gone unnoticed by scholarship but played a crucial role in Advaita hermeneutics and soteriology was the idea of p ramparya or mediate causality through close succession. The idea was directly related to apūrva, and is first introduced in TV 1.2.7, where Kumārila ponders how understanding the Veda serves human good (puruṣ rtha). Understanding the Veda was, of course, necessary for the performance of ritual, since ritual was laid down in the Veda, and the Mīmāṁsā assumption was that everything that a Vedic man does in a ritual context must be enjoined. There appears the question, therefore, how is understanding the Veda enjoined. Kumārila proposed that there is an injunction to that effect, sv dhy yo ‘dhyetavyaḥ, “One should do a recitation of the Veda.”43 The problem was, however, that the injunction pertains not to understanding the Veda, but to rote memorization and recitation. Kumārila here proposed that the injunction breaks down into consecutive goals and means, until it somehow terminates into being useful for human happiness, and that there was no need to seek an injunction for every step in this process. The injunction says that the Veda should be recited, and here we ask what is accomplished by such recitation. The obvious result was retention of the text. The retained text as a result becomes a means in its turn, and through it one proceeds to learning words as units. Once word boundaries are established, the words being the result assume the role of means for understanding word meaning. Word meaning in its turn engenders understanding sentence meaning, and one can then understand what the Veda says and perform ritual. Kumarila’s point 43 ŚB 11.5.6.3. 118 was that there is no requirement for separate injunctions for the steps following rote memorization, because they follow as consequences of each previous step. Puruṣ rtha is the performance of the ritual, but because understanding is necessary for such performance, it is at this final point that the injunction to recite the Veda becomes fruitful. It does that through a course of successive goals and means, p ramparya. The injunction itself undergoes progression, and its gains are further absorbed by another injunction for performing a ritual, in a causal succession terminating in human good. In such a course, the crucial takeaway is to see how a means ceases being a means once it produces its result, but through the series it mediately continues being a means and eventually reaches the ultimate result. In this case, the positing of such a series spared Kumārila the task of postulating an apūrva that would accrue from the recitation of the Veda. Through the series, the recitation of the Veda becomes for-the-good-of-the-ritual, and through the ritual it contributes to human felicity in a visible way. There is only one apūrva to postulate, between the ritual and heaven, while the rest is just a natural development of understanding through education. However, the principle of p ramparya applies when the result is an apūrva as well, there being no difference in causality in general, and Kumārila saw such series everywhere, for instance between the threshing of grain and the making of the sacrificial cake. There was no need for separate injunctions for every step between the threshing of grain and the making of the cake, because the first injunction could be broken down in a series of steps and terminate in the cake mediately. This idea of mediate causality, a principle which says that a cause terminates in its immediate result but is carried over through that result to the ultimate accomplishment, was likely a product of Kumārila’s rethinking of causality in general, prompted by the problem of apūrva, and had an 119 immense influence in Advaita soteriology and hermeneutics. It will come in sharp focus when we consider the progression of liberation in Śa kara’s Vedānta. Bh van , k k , and the Structural Unity of Sacrifice and Text The structural unity of the sacrifice is a result of a complex of related ideas. Any individual sacrifice is an action that combines present and prospective things in the world: there are offertorial substancesś actions and procedures of making themś agentsṅ expected results such as heaven; actions in which the agents offer the substances for the bringing about of heaven; etc. In Mīmāṁsā technical language, these are generally divided in two groupsŚ things that already are (siddha), already constituted, and things that still need to be (s dhya). We may think of them as actual and potential existents. Now, the structural unity of the sacrifice is centered on potential existents, because these are deficient and require actual existents for completion. Let us work through this by means of an illustration.44 In the classical Indian example of an action with its contributory factors, “Devadatta cooks rice in a pot by means of fire,” Devadatta the cook, rice the cooked result, the pot where the rice is cooked and fire the means of cooking are constituted existents: they are not essentially relational.45 The cooked rice is, of course, not yet a thing, but once it is cooked, its being is not contingent on other things. Once brought into being, it will be cooked rice. All the contributory factors, thus, can be seen in isolation, but the action of cooking itself is essentially relational, insofar as there could be no cooking unless there is a cook, ingredients, a cooking fire, a pot, a 44 This and the next section are largely indebted to McCrea 2000a. Useful are also Kataoka 2001, Ollett 2013, Edgerton 1928 and Freschi 2012:22-6. My account, however, is based on TV 2.1.1-4, which is relevant in its totality and, so, I will not include individual references. My presentation is less technical than a Mīmāṁsaka would want, as I am less interested here in the language aspect and more in the organization of action as a means of human good. 45 The example is taken from McCrea 2000a:434. I follow Matilal 1990 in translating k raka as “contributory factor for action.” 120 manner of cooking and a dish. This is the first distinction we need to note, in Mīmāṁsā jargon the distinction between that which is (siddha), and that which is not but could be, contingent on its relation to other things (s dhya). It is a distinction that involves Being and becoming on one hand, and relation on another. While the cooked dish is in becoming, it can achieve Being in which it will be non-relational. The sacrificial cake is a good instance of this: in relation to the action of making, it is s dhya, an expected result, but in relation to the action of offering, it is siddha, accomplished and available for taking up. The action of cooking, however, will always be relational. The fact that the action is essentially relational and deficient makes it the element which provides the structure to the sacrifice. The relationality of action, first, makes the action purposeful or intentional, and instrumental. When Devadatta cooks the rice, he does that to accomplish something: cook it and then eat it or serve it to others. He does it, in other words, to bring about a future state of affairs of there being cooked rice, to which future state of affairs the action has an instrumental value. To do that, he not only needs to cook, but he needs to cook in a certain way: he needs a recipe and some know-how. Thus, the action is not only (1) intentional or result oriented; and (2) instrumental; but also (3) procedural, i.e., unless it is performed in a certain manner it will fail to be an instrument and to bring about the desired future state of affairs. Finally, the action is (4) performable in the imperative sense, such that unless performed, nothing will come out of it. This performability is, really, the only fully relational thing in an action. Cooking in the instrumental and the procedural sense is substantiveŚ cooking is “a thing” and so is the manner of cooking, ready at hand for employment. For there to be pizza on my table tonight, however, some cooking must really happen. The action of cooking, then, can be analyzed into the fact that it is cooking among other forms of action, and that it is happening. 121  It is this performability that keeps the sacrificial action together. The fact that unless performed the action is non-productive and purposeless provides the force that harnesses disparate actions and things into a coherent whole. Mīmāṁsakas have termed this performability bh van , “bringing into being.” This bh van is the efficient force of the action, the one that gets the job done and secures the result, but the feature through which it holds the whole sacrifice together is its integrative force or k ṅkṣ , “need” or “curiosity” or a “charge.” The performability of the sacrificial action as a purposeful action requires certain things without which the action could not be performed. First, it requires that there be a result: the sacrifice should bring about heaven. Second, it needs that a sacrifice and not a soccer game take place. Third, it needs a set procedure in which the sacrifice happens. This procedure in its turn organizes all actions, substances, participants and what have you in the sacrifice in a coherent, teleologically organized whole, a whole of subordinate actions that result in substances that are offered in the principal action by the sacrificer, for the result. Without these three—result, means and procedure—a sacrifice will not be performed, and so the performability of the sacrifice seeks these out to “discharge its need.” These are, then, the three features of k ṅkṣ as the organizing structure of the sacrifice: the result or phala expressed by the accusative of the neuter interrogative pronoun, kim, “what” (svarga, heaven); the instrument or karaṇa, expressed by the instrumental of the interrogative pronoun, kena, “with what” (y ga, the sacrificial action); and the procedure of the sacrifice or itikartavyat , expressed by the interrogative adverb of manner, katham, “how.” Any subordinate action will have a performability and a need of its own, discharged in the production of the offertory but ultimately extending to the principal action through the offertory. Psychologically, 122 of course, this k ṅkṣ or ritual curiosity operates through the sacrificer and his intention to bring about heaven. Now, I said above that the sacrifice is not only a matter of facticity, but of textual ideality as well. In fact, the above account was somewhat my own reconstruction. When Mīmāṁsakas talk about bh van and k ṅkṣ , they talk about properties of textual, Vedic injunctions in relation to a prospective performer of a ritual, or more generally of any prescriptive sentence in relation to a listener. Despite their many differences, all Mīmāṁsakas understood sentences and texts as a collocation of words in which ultimately one thing is expressed, an action, in which the core role belonged to the verb. For the Bhā as, a Vedic sentence expresses an injunction as a unit, but its central element is the bh van or the performability of the action. Hearing a Vedic injunction, such as “he who wants heaven should sacrifice,” svarga-k mo yajeta, one cognizes that one should bring about heaven by the performance of sacrifice. The bh van here consists in “should bring about,” and is expressed by the optative ending in the finite verb yajeta. But this bh van is further divisible into “bring about” and “should.” The first of these is a property that all verbs have, a “general verbality” as Edgerton calls it,46 consisting in the fact that all verbal forms denote a productivity of some sort or effectuation of a state of affairs. Kumārila and his school of Mīmāṁsā called this feature rth or factual bh van . In our cooking example, “Devadatta cooks rice,” we can analyze the finite verb “cooks” into “s/he brings about” and “by means of cooking.” This “s/he brings about,” expressing the performability of action, is a feature of all verbal forms. Some finite verbs, however, specifically the optative and imperative forms (as well as gerundives), have the additional element of 46 1928:174. 123 “should” or the so-called verbal, abd bh van . What this abd bh van does is enjoining the subject, the one who desires the result, to perform the action so that the result may be attained. Let us work this through the paradigmatic injunction svarga-k mo yajetaṬ Its meaning is paraphrased as svarga-k maḥ svargaṁ y gena bh vayet, “he who wants heaven should bring about heaven by means of sacrificing.” We notice how the finite verb yajeta has been broken down into the specific instrumental action (by means of sacrificing) and the general performability or bh van (should bring about). In this injunction, heaven as the result of sacrifice is the object of the injunction, because its desirability, as we have demonstrated, is taken for granted—all men, in fact, want heaven. Further, the bh van here is of both types: “brings about” and “should.” The first is, of course, textual, but its reference is the actual sacrificial performance: it is interested in the text insofar as there is a sacrifice to be performed which requires a text corresponding to the sacrifice to be constituted. The second, also textual, looks towards the sacrificer and how it may constitute a text so that he may be induced to perform the sacrifice. Each of these two will have a set of the three parts of the cohesive force, the k ṅkṣ of the bh van . I will outline only the first here, and keep the second for later in the chapter. For there to be an actual bringing about of something, the bh van here expressed by the optative suffix in the finite verb, “should bring about,” requires from the text an object, a means, and a manner. Thus, in svarga-k mo yajeta, given that this is a central or puruṣ rtha injunction, the object feature of the k ṅkṣ is determined by the desire of the agent: what should be brought about is heaven. The object satisfies the purpose of the action. The means is expressed by the lexical meaning of the verbal root in the finite form yajeta, namely y ga or sacrifice, and this is the instrument feature of the k ṅkṣ . For these two we did not need to go out of the injunction, 124 but the sacrificial procedure—the recipe for cooking heaven—will be expressed in other parts of the Veda which describe what one does in a given ritual. For instance, if the injunction is jyotiṣṭomena svarga-k mo yajeta, then a Jyotiṣ oma sacrifice procedure is required. The suffix in the finite verb expressing bh van will attract the elements of this procedure—if one ought to bring about heaven, one will be prompted to ascertain how—first through textual proximity, but generally by applying the rules and methods of Vedic exegesis. The k ṅkṣ of the bh van will, in other words, organize sections of the Veda so that the purpose of the injunction may be realized: it will constitute a text describing a ritual. Mm s Classification of Vedic Texts and the Upani ads We saw, thus, how the Mīmāṁsakas understood ritual as a means of human felicity. By and large, they considered the Vedas to be only about this: the purpose of the Veda was to enjoin ritual and communicate in some way that its performance is a means of human happiness. For that purpose, a text had to be worked out which would cover all the details of the performance as well as the knowledge and know-how necessary for it. This text would need to contain knowledge about the agent, details about offering preparation, the use of implements and the role of the recipients, a script for the performance, an incentive for the sacrificer, all teleologically driven by the injunction which introduces the ritual. The text should be constituted from the Veda, and Mīmāṁsakas classified the entire Veda by types of passages that it contains in terms of their ritual applicability. In this section, we will outline this classification, and then we will try to pinpoint the role of the Upaniṣads in it. Mīmāṁsakas classified the Veda in four kinds of texts: (1) vidhi or injunctions; (2) mantra or sacrificial chants; (3) arthav da or descriptive passages; and (4) n madheya or names. It is no surprise that the injunctions were considered the central texts: only they were directly 125 related to dharma. The injunction group referred to the portions of the Brāhma as which enjoin the performance of sacrifices and included all kinds of injunctive texts, which Mīmāṁsakas meticulously classified. Vidhi covered not only the principal injunctions, but everything that is enjoined in the sacrifice, for instance the sannip tyopak rakas, the r d-upak rakas and all the individual actions they involved and auxiliaries they required. Crucially, they involve everything that is to be understood “just as it is heard.” Injunctions were classified primarily based on two criteria: (1) what they enjoin; and, (2) how they enjoin it. Under the first rubric, one common division is into utpatti-vidhi, an originative injunction in which the predicate is the principal element of the ritual performance; and viniyoga-vidhi, an applicative injunction which affirms that some auxiliary is related to the principal. The two are commonly referred to as a pair by Kumārila in the TV, and we can think of them as puruṣ rtha-vidhi and kratvartha-vidhi, enjoining that which is for the good of man or that which is for the good of the ritual, respectively. Good instances would be agnihotraṁ juhoti, “He performs the daily fire ritual,” and dadhn juhoti, “He uses yoghurt as the oblation.” To this pair, a third is commonly added, adhik ra-vidhi, a statement that introduces the ritual agent who is entitled to reap the fruits of the sacrifice, for instance the famous svarga-k mo yajeta, “He who wants heaven should sacrifice.” It is not necessary that these be stated in separate sentences. However, an injunction can affirm only one thing, and in complex injunctions the other elements are considered qualifiers of that one thing, producing thus a vi iṣṭa-vidhi, a qualified injunction. When these three injunctions in a ritual are put together through the three features of syntactic expectancy, k ṅkṣ , in a hierarchy ascertained through the Mīmāṁsā principles of interpretation, they produce a prayoga- vidhi, a whole manual that delineates the integral organization of the ritual and the manner of its 126 performance, as well as all the other details required. This is the text that I referred to above as the final product that covers the whole ritual procedure.47 This is commonly listed as the fourth type of injunction in this classification, and we should take a good note of it because it will play an important role in the mah -v kya idea. This fourfold classification is given in later manuals, but the individual types of injunctions are common currencies in Śabara’s and Kumārila’s works.48 The second classification of injunction asks the question, how is something enjoined. By this criterion, injunctions are commonly classified in three types: (1) apūrva-vidhi; (2) niyama- vidhi; and (3) parisaṅkhy -vidhi. The locus classicus on these is TV 1.2.34, although Kumārila there does not mention apūrva-vidhi, but rather talks about vidhi, niyama and parisaṅkhy .49 An apūrva-vidhi is a statement that enjoins by disclosing an otherwise unknown causal relationship. Take, for instance, the statement vr h n prokṣati, “he besprinkles the rice,” that is an action of the type of saṁskṛti or consecration, and adds an excellence of some kind to the substance over which it operates. That there is some causal relationship between the action of besprinkling and the excellence that obtains subsequently in the rice is not empirically knowable, and is solely due to the injunction. We can look at this from the point of view of the desired result. We need an element of excellence in the rice so that it can be used in preparing the sacrificial cake. Because this excellence is invisible, no action is empirically related to it. In Kumārila’s words, such relationship does not obtain “absolutely.” For all we know, it may be the 47 See Tachikawa, Bahulkar and Kolhatkar 2001 for a ritual modeled on the Darśa-pūr amāsa and a corresponding prayoga-v kya serving as a manual of its performance, the Pavitreṣṭi-prayoga. 48 A good overview is available in Pandurangi 2006:177-8. 49 Kumārila’s definition, which he then expands on, saysŚ “A sentence is an injunction when [some causal relationship] absolutely does not obtain [by other means of knowing]. A restriction happens when there is [general] optionality, whereas an exclusion when such optionality obtains in regard both to one and another.” vidhir atyantam aprāpte niyamaḥ pākṣike sati | tatra cānyatra ca prāpte parisa khyeti kīrtyate. TV 1.2.34, I.152. 127 action of arranging the rice into the image of LeBron James that will furnish the required excellence. Nothing of the kind is known to us “before the sentence,” and for this reason this type of injunction is “pure,” fully in the domain of the Veda.50 It is exclusively related to unseen results. It is eminently clear that this injunction is related to the original meaning of the notion of apūrva, unprecedented as knowable only from the Veda. There may be cases, however, when two ritual elements are commonly related, in multiple possible ways. Keeping with the rice example, once the rice has been consecrated, its husk needs to be removed so that it can be used in making the cake. We know how to do that, and we could imagine more than one appropriate ways—this is not empirically unavailable. An injunction in relation to this reads, vr h n avahanti, “he threshes the rice.” The predicate of this injunction is not the action of rice preparation, but its specific mode of threshing qualified by the natural consequence of excluding all other possibilities, never mind if they are all accounted for or not. The important thing is that optionality obtains in general. This type of injunction is called niyama, restriction. When, however, there is a similar situation but one in which the whole scope of what can be affirmed is known, and the point is not to affirm the stated element as intended but to exclude whatever is not stated, this injunction is called parisaṅkhy , exclusion. Take, for instance, the statement a v bhidh n m datte, “he takes the horse’s bridle.” While we need not go into involved details, the statement as it stands is problematic because the action of holding is supposed to be performed alongside the recital of a certain mantra, but the statement seems to reiterate something already affirmed in a related text, which puts its purpose in jeopardy: a pram ṇa cannot repeat something known. The solution is to take the statement as intending not 50 tatra yo ‘tyantam aprāpto na ca prāpsyati prāg vacanād ity avagamyate tatra niyogo śuddha eva vidhir. TV 1.2.42. 128 to enjoin the holding of the horse’s bridle while reciting the mantra, but as intending to exclude the holding of the donkey’s bridle which presents itself as an assumed alternative. The restriction still denotes the action of holding: holding as qualified not by what is said, but by what is not said when it could have been said. This, obviously, leaves a lot of leeway for dissent, and it is often a matter of disagreement whether a specific statement is a restriction or an exclusion. We need not worry about this, but we should note well how the three are defined, because they will play a formative role in Śa kara’s making sense of meditation vs. reflection in Vedānta. To summarize, the first classification provides for the structure of the ritual: it is set in motion by a sentence that presents the central ritual element; ritual details are related to the central element; the agent entitled to the results is pointed out; the structure itself is given. The second classification, on the other hand, is concerned with knowing causal relationships that govern the elements of the ritual. The second group of Vedic texts, mantras, referred to versified composition, generally in the Vedic Saṁhitās, which are recited in a sacrifice accompanying parts of the ritual and are intended as markers of these parts, the respective deities, the offertories, etc. “[M]antras allude to what is going on in the sacrifice as the priest executes it. Thus, recited in the proper sequence, they help the priest see what he is doing and remind him of what has yet to be done. They provide a running narrative of the rite.”51 Taber points to a crucial feature of theirsŚ “[T]heir meaning is usually evident as soon as they are pronounced. They are grammatical; they make sense of themselves. But, still, when a mantra is presented in the Veda as a formula to be uttered in the context of the ritual, one may take it to express what it means, or one may not.”52 Unlike the injunctions that must be taken as heard, mantras are more like a soundtrack. They do not say 51 Taber 1989:149. 52 Ibid, 145. 129 anything about Indra, Agni, the sacrificial fire, etc., but indicate what is happening at a given moment in the ritual and help the priest recall that detail, much as Wagner’s “Wedding March” played at a wedding to accompany the entrance of the bride is not about Elsa, but marks the entrance event. The reference of the waltz is the event, not Elsa. As for the arthav da or stories and descriptions found alongside the ritualistic sections of the Brāhma as, they were a problem for Mīmāṁsā because they evidently do not enjoin an action—all that the Vedas are valid for—but are part of the Vedas and cannot be discarded without compromising the validity of the corpus. An additional problem with these passages was that many of them were just contrary to sensory evidence. Think, for instance, of the many bandhus or correlations in the Brāhma as and the Upaniṣads, such as the famous identification of the sacrificial horse with the universe at the opening of the Bṛhad- raṇyaka Upaniṣad. Given the premise that the Vedas are valid only regarding (ritual) action and do not teach about things that we can see, what to do with such descriptions many of which are plain false and all of which are not about action? Śabara’s solution was to treat such passages as not being truth claims at all. Consider any story. A story can do two things: (1) it can give an account of past events; or (2) it can cause attraction or repulsion to something else, like an advertisement that makes you want the product no matter how accurate it is. This is what arthav das do: they advertise the ritual action. Their validity does not consist in whether what they say is true or not—though one can always interpret them to avoid contradiction with the evident—but in aiding the performance of the sacrifice by making it look good.53 To take the standard arthav da exampleŚ “One who wants prosperity should immolate a white animal to Vāyu. Vāyu is the swiftest deity. Vāyu comes with 53 See Śabara ad MS 1.2.10. 130 his own property and leads him [the sacrificer] to prosperity.” In Kumārila’s words, knowing that the cause and the effects are alike, one is made to believe that the sacrifice to the swiftest deity will make the result arrive without much delay.54 This makes the arthav das purposeful and, therefore, valid, not in terms of truth, but action. Kumārila brought his hallmark sophistication to the issue of arthav das.55 I mentioned that he distinguished two kinds of bh van or verbal productivity, arth or actual and abd or verbal bh van . In the paradigmatic injunction svarga-k mo yajeta, this bh van had the form of “bring about” for the arth bh van and “should” for the abd bh van . The arth bh van further had three points which it required for its completion: an object (heaven), an instrument (the sacrifice) and a procedure of sacrificial performance. Now, the abd bh van similarly needs to become complete in the same three points in order to accomplish its objective, which is to get the man perform the sacrifice. The abd bh van is all about the taking up of the sacrifice, not its accomplishment, so naturally its object, answerable to the kim-feature of the k ṅkṣ or the verbal need, is the taking up of the action, for which reason the sacrificer must be induced. Further, the abd bh van requires an instrument for effectuating this, corresponding to the kena-feature of the k ṅkṣ , and for this it must give rise to an understanding of the injunction on the part of the sacrificer, one that is contingent on experiencing that there is a causal relation between the action to which he is prompted and the result that he expects. Kumārila calls these two—the kim and kena features—puruṣa-pravṛtti and vidhi-jñ na respectively, engaging man in the sacrifice through understanding the injunction. Finally, the abd bh van needs to find a way to do that, corresponding to the katham- feature of the k ṅkṣ . The question, then, is how the optative suffix which expresses bh van — 54 sādhanānurūpa-sādhyotpatti-dvāre a kṣipra-devatā-sādhyaṁ karma kṣipram eva phalaṁ dāsyati. TV 1.2.7, I.115. 55 This section is based on TV on MS 1.2.7 131 and Kumārila is quick to point that the insentient suffix operates through the sacrificer’s awareness—can convince the sacrificer that the ritual action can furnish the result. “For, a man acts led by reason, and as long as he does not understand something as good, he will not act upon it.”56 We shift perspective now, from the optative suffix to the sacrificer, because the sacrificer must see the desirability of the sacrifice. This, Kumārila claims, can happen in two ways. The sacrificer can see, first, the excellence of the ritual action; it is an action laid down by a Vedic injunction, and the Veda is faultless. In this case, the abd bh van operates solely through the optative suffix. Or, he can realize how the sacrifice is good because some of the deities or substances that are its part are excellent in some way. This is accomplished by the arthav da sectionsŚ hearing, as I already said, how Vāyu is the swiftest deity, one’s understanding that the cause and the effects are alike is activated and a conviction that the sacrifice to the swiftest deity will make the result arrive quickly is born. The fact that an arthav da is juxtaposed to an injunction makes these two seek each other for completion, and while the suffix could perform the same function alone, the presence of the arthav da suspends that. Kumārila calls this feature of the abd bh van corresponding to the itikartavyat , the katham feature, the knowledge of excellence or pr astya-jñ na.57 The question now presents itselfŚ where is the place of the Upaniṣads and the knowledge of the Self as its domain in this classification? The seemingly easy answer isŚ the Upaniṣads are part of the Brāhma as and they do not enjoin action (or so it seems)ś ergo, they must be arthav das.58 This is how scholars tend to present the Mīmāṁsā understanding of the Upaniṣads. 56 buddhi-pūrva-kāri o hi puruṣā yāvat praśasto ‘yam iti nābudhyante, tāvan na pravartante. TV 1.2.7, I.114. 57 The fourth part of the Veda, N madheya or names, refers to texts which give references to particular sacrifices through their names, and seems to have been posited as a category just to avoid double injunctions. See Jha 1964:182-6. 58 For a reliable study of arthav da in Mīmāṁsā, see Harikai 1994. For a short overview, see Jha 1907:xxxv; for a longer overview, Jha 1964:177-182. 132 For instance, Halbfass writesŚ “Kumārila mentions the Upaniṣads side by side with arthavādas, and he tends to see the Upaniṣadic teaching about the Self as being auxiliary to dharma, that is, to the performance of ritual actions, insofar as the notion of a noncorporeal permanent self is a condition and an incentive for performing such acts which are supposed to bear fruit in another life or birth.”59 Hirst explicitly identifies the Upaniṣadic statements about the as self as arthav da in MīmāṁsāŚ “These [non-injunctive statements] last were classified as arthav da, secondary statements whose real function was to encourage a person to undertake ritual action. So, for example, all statements about the self were seen, not primarily as descriptions of the self, but as motivators to action, the self being the one who would accrue the result of the sacrifice performed.”60 “The application of the category of arthav da (secondary passages) (v) is slightly more complicated. The Ritualists developed this notion to account for apparently descriptive passages, particularly those found in the Upaniṣads.”61 Rambachan gives a similar explanation: The Pūrva Mīmāṁsā contention that the Upaniṣads have no independent purpose but are merely an appendage to the main body of injunctive text was a formidable challenge to Śa kara. … Many Vedic texts, for example, including the sentences of the Upaniṣads (ved nta-v kyas) are seen as having their purpose only in praising what has been enjoined in the injunctions (PSM 1.2.7). … According to Pūrva Mīmāṁsā, the Upaniṣads are merely an appendage to the main body of injunctive statements. The utility of the Upaniṣads lies only in praising the prescribed action or in providing some useful information, such as knowledge of the deity or agent for performance of a particular rite.62 The fact of the matter is more complex than this simple identification of “being subordinate to dharma” with “being arthav da,” and its corollary “the Upaniṣads as a unit are subordinate, ergo they are arthav da.” Let us examine carefully what Kumārila says about the Upaniṣads. 59 1991:150. 60 2005:38. 61 Ibid., p.63. 62 1992:34. 133  To begin with, Kumārila’s understanding of the complex ritual causality found a place for the Upaniṣads as providing knowledge about the agent in the sacrifice, ultimately serving the purpose of action but having truth value. We should recall here the puruṣ rthaṭkratvartha and principal/auxiliary organization of the sacrifice. The ritual agent in the sacrifice was an auxiliary factor, a kratvartha, and the Upaniṣadic texts which present knowledge of this agent were ultimately absorbed in the principal ritual action through the agent. Unlike the arthav das, which for Kumārila were strictly in the realm of abd bh van where truth values do not matter, the Upaniṣadic passages about the Self were absorbed in the arth bh van where accurate knowledge was important, although ultimately made use of in action. Ascertaining the details of procedure that involved the ritual agent was not related to like or dislike, and the success of the sacrifice was predicated on knowing such details. It is significant that Kumārila placed the Upaniṣads right there. “The Upaniṣads discharge their need ( k ṅkṣ ) through presenting the agent that is subordinate to the ritual action.”63 Thus, Kumārila included the Upaniṣadic description of the Self in the arth bh van , before he had the occasion to introduce the arthav das and the abd bh van as their domain. The issue of arthav da appears with passages which are fanciful or do not contribute anything obvious to the action. They are not required as part of the sacrificial procedure, but are present in the text and must be accommodated because of that. Vāyu may be the fastest deity for all we know, but the point is that this does not matter in the sacrificial procedure. The situation with the Self is different, and in one sense can be compared to the sacrificial cake: the cake is subordinate to the action of offering, yet the passages which enjoin how to prepare it are not arthav da, because they are predicated on having truth value. They enter the itikartavyat Ṭ With the Self, of 63 etena kratvartha-kart -pratipādana-dvāre opaniṣadāṁ nairākā kṣyaṁ vyākhyātam. TV 1.2.7, I.114. 134 course, the issue was somewhat more crucial, since without a permanent Self that enjoys the results of the sacrifice the authority of the Veda would be compromised. Kumārila thought that such knowledge of the Self as an eternal agent and enjoyer of ritual action follows even from the bare injunctive statements through scriptural postulation—there must be an eternal Self that will enjoy the results, or otherwise what the injunction says would be false—but such knowledge becomes firm through the study of the Upaniṣads.64 However, in other places Kumārila does treat Upaniṣadic texts as arthav da. For instance, in his account of the origin of smṛti, he attributed the various theories of creation and dissolution common among Vedic folks to ideas that originate in the mantras and arthav das.65 This certainly includes sections of the Upaniṣads. He traced even the origin of some Buddhist ideas—idealism, momentariness, the doctrine of no-Self—to the Upaniṣads and arthav da, explicitly paired, and meaning to prevent excessive attachment to sensual matters.66 Reasoning also had origin in the Upaniṣads and arthav da in pair.67 These ideas of his were not particularly revolutionary either: for most of them he had a precedent in Bhart hari.68 We may, further, venture to guess that he would have classified the Self-Brahman identification in the Upaniṣads as arthav da, since such a doctrine, as noted by Nakamura, makes the Mīmāṁsā doctrine, predicated on a plurality of Selves, fundamentally impossible. “The eternal existence of the individual tman, from the standpoint of the highest truth, is absolutely necessary and indispensable as the presupposition on which the M m ṁs philosophy can establish their 64 ŚV tmav da 141, 148. 65 yāś caitāḥ pradhāna-puruṣeśvara-paramā u-kāra ādi-prakriyāḥ s ṣ i-pralayādi-rūpe a pratītās tāḥ sarvā mantrārthavāda-jñānād eva. TV 1.3.2, I.168. 66 sarvatra hi tad-balena pravartate, tad-uparame coparamatīti vijñāna-mātra-kṣa a-bha ga-nairātmyādi-vādānām apy upaniṣad-arthavāda-prabhavatvaṁ viṣayeṣv ātyantikaṁ rāgaṁ nivartayitum ity upapannaṁ sarveṣāṁ prāmā yam. Ibid. 67 tatra lokārthavādopaniṣat-prasūtais tarka-śāstraiḥ sarva-vipratipatti-mukha-pradarśanam. Ibid. 68 Aklujkar 1991. 135 rites.”69 Finally, Kumārila refuted Vedāntic theories of the origin of the world and rejected the very possibility of a creator, which is so prominent in Vedānta.70 Therefore, it seems to me that it is a notional mistake to talk about a general Mīmāṁsā attitude to the Upaniṣads as a single corpus, as it is commonly done. Kumārila clearly had an idea that they are distinct, focused on knowledge of the Self, but they contained the same types of sentences as the Brāhma as: injunctions and their auxiliaries, and arthav das. The fact that both are treated as auxiliary to dharma is not the characteristic that calls for putting an equation sign; there is, rather, a crucial difference, insofar as one give information that must be taken literally and the other is for the purpose of inspiration. The key distinction in attitude between Mīmāṁsā and Advaita Vedānta was not about the Upaniṣads being arthav da, but about what kind of Self they presented: A Self that is essentially a ritual agent and an enjoyer, or a single aloof Self, one for all. A corollary to this concerned the status of passages that talk about liberation from saṁs raŚ are they true statements of result, or just statements of praise? We will see this conflict already in the next chapter. Language and Pr bh kara M m s Kumārila’s philosophy of language was fully consistent with the model of the sacrifice and its ritual causality. Words are meaningful intrinsically and before they combine in a sentence. In a sentence, word meanings are harnessed by the verb, absorbed, and finalized in a new, sentential meaning. Cognition also proceeds from recognizing individual word, to recognizing their meaning, to understanding what the whole sentence or passage stands for. That much was 69 1983:363. ŚV Sambandh kṣepa-Parih ra 83-4. The whole chapter is one massive statement against a first principle of the 70 world. 136 already said by ŚabaraŚ “Words, denoting their own objects, cease their [individual] function, and then the understood word meanings give rise to sentence meaning.”71 Word meanings are employed by the bh van of the verb to form a coherent sentence meaning, but they mean whatever they mean before that. This has become known as the abhihit nvaya-v da, or the doctrine of association of word denotations in a sentence. It is important to see how this understanding corresponds to the theory of mediate ritual causality which Kumārila developed, in which individual things and small and large actions of different kinds are gradually absorbed in the final action in a process of refinement and maturation, giving up their individual causality only to keep it.72 This idea will have a massive significance for Advaita soteriology and the notion of mah -v kya. The rival school of Mīmāṁsā, the school of Prabhākara, had a different approach to language, tied to a different approach to ritual as well. To begin with, whereas for Kumārila the verbal suffix in the sentence ultimately expressed an urging for the performance of action by presenting it as a means to a desirable end,73 for the Prābhākaras Vedic injunctions issue a mandate or niyoga to an individual to perform a ritual, having their own purpose in view, namely perpetuating ritual performance.74 To secure such performance, an injunction must have an object, something real that it produces as an outcome, so that the agent of the sacrifice, coming with his own purpose such as happiness, would take up the enjoined action. The injunction, in other words, issues a mandate, but it needs to provide all the requirements so that the mandate 71 padāni hi svaṁ svaṁ padārtham abhidhāya niv ttavyāpārā i. athedānīṁ padārthā avagatāḥ santo vākyārthaṁ gamayanti. MSŚBh 1.1.25, I.96. 72 Kumārila’s doctrine of sentence meaning is developed in the V kya chapter of the ŚV. 73 The theory was fully worked out by Ma ḍana Miśra, who argued that the abda-bh van expresses that the action has the property of being a means of something desirable, iṣṭa-s dhanat . See David 2013 for a very lucid account of this. 74 This account is indebted most of all to Yoshimizu 1994 and Hiriyanna 1972:85-96; I have also profited from McCrea 2000b. 137 would be carried out. This mandate is not unconditional or categorical, as it was characterized by early scholarship,75 but pertains to someone for whom the performance is relevant in virtue of the desire for the result, or some other criterion, i.e., someone who has an adhik ra for the ritual action.76 The injunction cannot select just anyone, say a Śūdra, and present him with an unconditional obligation, as Kant’s categorical imperative would, because the successful performance of the mandate requires that one understands the mandate as pertaining to oneself, and that happens through the object of the agent’s desire, such as heaven. But, that is how far the injunction goes. Whereas Kumārila expected the injunction to convey somehow that the performance of the ritual action can bring about the expected result, Prabhākara saw that as already covered by the desire for the result. To understand this clearly, think of the sentence, “She who wants a good life should pursue education.” The central message which the sentence wants to communicate for Kumārila would be how good education is, because no thinking person would take up a course of action unless first convinced that it is for one’s good. Prabhākara, on the other hand, would take that requirement satisfied by the desire for a good life, in which case the core message of the sentence would be that education must be pursued if one wants a good life; i.e., one must take up that course of action. To be, now, specific, the taking up of the course of action is the mandate issued by a Vedic injunction (niyoga).77 The specific form of action, in our case education, is the instrument by which one can accomplish the objective, a good life, but it is also the direct object 75 See Nakamura 1994 for a comparison of niyoga with Kant’s categorical imperative. While Prabhākara’s account of obligation is deontic, its characterization in Kantian terms is wrong on several important counts. 76 The desire would pertain to the rituals known as k mya-karma, which were meant for a specific result such as heaven or cattle. The very belonging to one of the three classes would constitute the criterion for the performance of the mandatory rites, nitya-karma, which such a member of the three classes would need to perform till the end of life. 77 atrābhidhīyate—niyogaḥ karma i niyu kta iti na samyag avadh taṁ bhavatā. ārambhe hi puruṣaṁ niyu kte, na karma i. Bṛhat 2.1.5, III.321. 138 of the urging expressed by the injunction. For Prābhākaras, contrary to the Bhā as, the specific action and not the result is the one element that needs to be produced (k rya), although the result remains superordinate to the action. Ultimately the educational system has its own purpose, to perpetuate itself. The good life which it promises is not superordinate to it, but it is superordinate to the performance of the action of study on the part of the student. Now, there is one difference between ordinary, worldly mandates, such as the one of education and a good life, and Vedic mandates, and that is that the result does not follow immediately upon the completion of the mandate—one is not seen going to heaven after the ritual—owing to which something permanent must be postulated over and above the action expressed by the mandate, that remains after the performance of the action and eventually brings about the result.78 This additional element is knowable solely from the Veda—it is, in fact, what makes the Veda a pram ṇa—and is why the Vedic mandate is called apūrva, unprecedented, unknowable otherwise.79 Prābhākaras clearly followed the old Mīmāṁsā understanding of ritual causality, where what matters is to get the ritual structure right and leave the rest to the Veda.80 If the Veda says that he who performs the sacrifice goes to heaven, then one had every right to believe so. There is no need for grand theories of ritual causality or general theories of maturation. There is something permanent which remains after the sacrifice, and nothing more than this is required. This central role of the structural unity of the sacrifice influenced the Prābhākara philosophy of language, in which words do not initially mean anything individually, but collectively first produce a single sentence meaning and only then acquire individual word 78 Hiriyanna 1972:92-3. 79 na pūrvam apūrvam. pūrvam tad ucyate yad avagatam prāk. yat tu kenacid prāg anavagataṁ pratīyate tad apūrvam. tathā-bhūtaś ca vidhy-arthaḥ. Ṛju-Vimal 2.1.5, III.312. 80 Clooney is, therefore, right in claiming as much. 1990:245-53. 139 meanings. Words are individually meaningful only in sentences, when they are syntactically related to other words. This understanding of language has become known as the anvit bhidh na-v da or the doctrine of denotation as single correlated meaning. This unified and unique sentence meaning is action, of the obligation kind (k rya), qualified by its contributory factors (k rakas), and a sentence without a verb expressing such an obligation cannot be construed as unitary and cannot express meaning. If a sentence does not obtain a meaning, neither can the individual words. It is, in fact, impossible to even learn what individual words mean unless they form part of such a sentence expressing obligation, because learning language happens through observing how elders deal with one another through commands.81 Such an attitude to ritual action and language meant that Prābhākaras did not care much for the arthav das, except for the fact that they happened to be part of the Veda. The k ṅkṣ or the syntactic expectancy of a sentence expressing action is discharged once all the contributory factors are supplied for the structural unity of the ritual to obtain. If some of the factors are missing, they must be supplied (a process called adhy h ra), but once that is done, there is no natural k ṅkṣ anymore and for any other words that happen to be in the vicinity. For instance, for the adjective “white” for a “cow,” an k ṅkṣ must be assumed just because the word happens to be there. Such was the case with the arthav das as well: they do not bring anything real to the ritual action, but are in proximity and need to be accounted for. Since verbal forms that do not express obligation, such as those in the present tense, do not have k ṅkṣ either, the arthav das are construed with their proximate injunctions. This attitude is well exemplified by an objection (pūrva-pakṣa) which Śālikanātha characterized as “half-conclusion”: there is no real loss if the 81 Śālikanātha’s V ky rtha-M tṛk is one grand statement of this philosophy. See Sarma 1987, 1988, and 1990. 140 arthav da portion of the Veda has no validity, but there would be an enormous loss if the injunctions, by association with something invalid, would lose validity too.82 This is a “half” conclusion because there is nothing wrong with it essentially, except that it puts the whole of the Veda in jeopardy and through that the injunctions themselves, and so it is necessary that the arthav das do get some validity, though nothing would be lost without them. They are, therefore, construed along with an injunction, in sheer view of the fact that they happen to be around. Śālikanātha applied the same reasoning regarding the Upaniṣadic descriptions of Brahman.83 An Upaniṣadic statement (ved nta-v kya) such as “Brahman is awareness, bliss” must be completed with some injunction expressing an obligation, for otherwise its words will not even obtain reference. If it is possible to supply them with some such obligation, all good and well. Otherwise, the disassociation of the Upaniṣadic statements from injunctions will simply mean that they remain inexpressive and will bring no real harm.84 Equally radical was the Prābhākara understanding of the knowledge of the Self. Sureśvara paraphrased eloquently this understanding in his Sambandha-V rttikaṬ85 As a pad rtha or a common category, the Self was not at all in the domain of knowledge from 82 tatra rāddhāntaikadeśyāha – mā bhūt tasya prāmā yaṁ, tasya kevalasyārthāvada-bhāgasya bhavatu apramā atā, na kācit kṣatiḥ. nanu tadāpramā ye vidhy-uddeśasyāpi jñāta-prāmā yasya aprāmā yāpattir mahatī kṣatiḥ. Ṛju-Vimal 1.2.1, II.2.) 83 Śālika developed his account as a response to Ma ḍana Miśra’s Brahma-Siddhi, but he was clearly spelling out an understanding which must have been much older than this. Śa kara and Sureśvara are both aware of this construal of the Upaniṣadic statements and argue against it (see, for instance BĀUBh 1.3.1 and NaiS 1.14-19). All we know with certainty about Śālika is that he was later than Ma ḍana, whom he directly quotes and attacks, and earlier than Vācaspati (see, for instance, Acharya 2006Śxxi). Śālika’s critique of the descriptive sentences in the Upaniṣads concerning Brahman is found in his Ṛju-Vimal on Prabhākara’s Bṛhat on 1.1.2. 84 evaṁ sarva-padānāṁ kāryānvitārthābhidhāyakatvād yadi vedānteṣu kāryaṁ yogyam adhyāhārādibhir labhyate, tadā kāryārthataiva vedānta-vākyānām apiś atha na, tato ’nabhidhāyakataiva vyutpatti-virahāt. ato vidhi-nirākara am api vedānteṣu na kṣatim āvahati. Ibid. 85 SV 440-454. The doctrine is presented as of “those who are enamored with mandate as the sole reference,” niyog rthaika-r ginaḥ (454). 141 linguistic utterances, abda. The Self was like rice, a thing whose properties we already know, but which we use in ritual through the fact that it becomes an auxiliary to the action of offering. 86 The Veda is a pram ṇa strictly on the mandate or niyoga that an action of some kind be performed. Through that mandate, the only new thing that the Veda says about rice qua rice is that it is usable in the relevant ritual, serviceable to the mandated action; in other words, that it is an auxiliary.87 The Self is, like rice, otherwise known, and its true nature is recognized by the aid of reasoning that proceeds by examination of what is permanent and what changes (anvaya- vyatireka) in its three states—waking, sleep and deep sleep.88 The proper pram ṇa for knowing the Self is pratyabhijñ na, recognition, which was a mode of perceptual awareness. The idea of pratyabhijñ na as a mode of perception was developed in the school of Nyāya, and we may quote Chatterjee on this with profit: To recognize thus means to cognize that which we are aware of having cognized before. Pratyabhijñ is recognition in this sense. It consists in knowing not only that a thing is such-and-such but that it is the same thing that we saw before.89 Pratyabhijñ na was a mode of perception in which the percept is qualified by traces of former percepts, and it was the pram ṇa for knowing the Self.90 Through reasoning, these traces of former percepts should be removed, at which point the pure Self alone remains. The Self is known through recognition assisted by reasoning, and the Veda has no say in this: if it does say 86 “Others sayŚ because the Self is a category, it is knowable by other means, not from scripture, like other categories such as rice.” anya āhuḥ padārthatvāt pramā āntara-gamyatām | ātmano nāgamāt siddhir vrīhyādy-anya-padārtha-vat. SV 440. 87 SV 451-454. 88 SV 441. Śabara, in fact, argued in MS 1.1.5 against the Vijñānavādin opponent that the Self was self-evident, known to oneself through the recognition of one’s persistence through time, but eminently private and not available to intersubjective perception. 89 Chatterjee 2008:188. 90 Cf. Jhalakikar and Abhyankar, 1928:543: tal-lakṣa aṁ tu indriya-sahak ta-saṁskāra-janya-jñānatvam. … pratyabhijñāyām ātmā viṣayaḥ iti siddhāntaḥ. 142 things about the Self, as it does, the corresponding scriptural cognition is a form of appearance present to our consciousness, pratibh , but it is not a veridical cognition produced by a pram ṇa.91 This does not mean that it is false awareness, but it is repeating something that is already known. The only new thing that the Veda can say about the Self is that it is auxiliary to the action of ritual or meditation.92 Conclusion This chapter provided us with several tools necessary for proceeding with the dissertation. First, Śa kara’s fundamental rethinking of the nature of the Upaniṣads in specific and the Veda in general, culminating in the notion of mah -v kyas, had as its starting point the formidable task of challenging the claim that the Veda was an authority just on action, primarily of the ritual kind, and now we can appreciate what this really meant. Second, Advaita Vedānta as a systematization of the Upaniṣads developed as a soteriological enterprise, and ventured into theology and philosophy out of apologetic concerns. Although Vedic theology was fully dominated by action and meditation, as we shall further see in the next chapters, it also provided the categories in which the Advaita form of soteriology was expressed—results, means, procedures, forms of causality, the central role of desire and the suitable candidate (adhik rin)—all of which Advaitins both challenged and appropriated. In other words, in organizing the Veda and the Upaniṣads around the ideal of liberation, Advaitins used Mīmāṁsā modes of thinking. This concerned particularly the rethinking of the role of meditation, crucial in which were the nature of the so-called niyama- and parisaṅkhy -vidhi, restriction and exclusion, and the formulation of the notion of mah -v kya on the model of the prayoga-vidhi or ritual manual. 91 “It is from linguistic utterance that a mere phenomenon arises, not from a pram ṇaś” tataś ca pratibhā-mātraṁ śabdād iti na mānataḥ. SV 450. 92 Ibid, 452-453. 143  Central was the role of Kumārila, because it was his model of dharma that Advaitins appropriated. First, Kumārila understood dharma as humanly centered, a means of some good ( reyas-kara), and Śa kara fully agreed with it. Second, Kumārila’s approach to language, in which individual elements gradually combine to obtain final reference, opened the door for rethinking the Veda as primarily being about knowledge and only then about action. This was best reflected in his rethinking of the nature of arthav das, in which they had to be meaningful before they become absorbed in the action. They had to convince a thinking man that the ritual action was good, and they had to engender understanding first, which only later results in taking up action. With this Kumārila accommodated at least the possibility for the thinking man to say, “I am not convinced. This is not for me.” Kumārila’s philosophy of language, further, had a direct influence on Śa kara’s soteriology, in which liberation followed upon understanding the identity statements of the Upaniṣads, which was contingent on understanding the meaning of the correlated categories. Finally, although the Mīmāṁsā categories of ritual action which Śa kara used in rethinking the soteriological role of the Veda were common in Mīmāṁsā long before Kumārila, it was their reorganization in terms of maturation through mediate causality (p ramparya) which Kumārila introduced that would play the crucial role in this rethinking. 144  PART TWO: LIBERATION AND THE HIGHEST GOOD IN PRE- A KARA VEDIC THEOLOGY INTRODUCTION We saw in the previous chapter that the Vedic theologians of the Pūrva-Mīmāṁsā school associated the authority of the Veda with dharma, defined as a capacity of existents to bring about some future good desirable to man ( reyas-kara). By and large the category of dharma was identified with ritual action, and the results which dharma was thought to bring spanned from worldly desirables of the Vedic variety—good progeny, cattle, wealth, virility, royal sovereignty—to a state of felicity in the hereafter, happiness pure and simple called “heaven” or svarga. In the following three chapters, I want to look at deliberations in pre-Śa kara Vedic theology on “the highest good,” niḥ reyasa. This is a term with some history in the corpus of dharma literature,1 where it initially meant any general good which one may attain by observing dharma in the broad sense, primarily one’s social duties, and which was commensurate with and appropriate to one’s social standing.2 The late Vasiṣṭha Dharma-Sūtra was the first to associate niḥ reyasa with the attainment of heavenŚ “Next comes the desire to know the Law for the sake of attaining the highest goal of man. Now, someone who knows the Law and follows it is a 1 The term itself is not all that common—three occurrences in the ĀDhS and one in GDhS and VDhS each—but it is prominently placed right at the beginning of ĀDhS and VDhS. 2 See, for instanceŚ ĀDhS 1.1.4-8: “There are four classesŚ Brahmin, Kṣatriya, Vaiśya, and Śūdra. Among these, each preceding class is superior by birth to each subsequent. Those who are not Śūdras and are not guilty of evil deeds may undergo initiation, undertake vedic study, and set up the sacred firesś and their rites bear fruit. Śūdras are to serve the other classes; the higher the class they serve, the greater their prosperity;” catvāro var ā brāhma a- kṣatriya-vaiśya-śūdrāḥ. teṣāṁ pūrvaḥ pūrvo janmataḥ śreyān. aśūdrā ām aduṣ a-karma ām upāyanaṁ vedādhyayanam agnyādheyṁ phalavanti ca karmā i. śuśrūṣā śūdrasyetareṣāṁ var ānām. pūrvasmin pūrvasmin var e niḥśreyasam bhūyaḥ. GDhS 11.23-26, in the section of the king and how he should judge: “Reasoning is the means of reaching a correct judgement. Having reached a conclusion in this manner, he should decide the case equitably. If there is conflicting evidence, he should consult those who are deeply learned in the triple Veda and reach a decision, for, it is said, acting in that way, he will attain prosperity;” nyāyādhigame tarko 'bhyupāyaḥ. tenābhyūhya yathā-sthānaṁ gamayet. vipratipattau traividya-v ddhebhyaḥ pratyavah tya niṣ hāṁ gamayet. tathā hy asya niḥśreyasaṁ bhavati. Both translations Olivelle 2000. 145 righteous man (dh rmika). Such a man becomes pre-eminent in this world and wins heaven after death.”3 We will see shortly that Mīmāṁsā in general associated this highest good or niḥ reyasa with the attainment of heaven. However, when we come to the Manu Smṛti, we witness a new, neat theory in Vedic theology of what constitutes the highest good. All attainments that were associated with dharma have now been grouped in a single category, and something quite different has become “the highest good”: Acts prescribed in the Veda are of who kinds: advancing, which procures the enhancement of happiness [ bhyudayika]; and arresting, which procures the supreme good [naiḥ reyasika]. An action performed to obtain a desire here or in the hereafter is called an ‘advancing act,’ whereas an action performed without desire and prompted by knowledge is said to be an ‘arresting act.’ By engaging in advancing acts, a man attains equality with the gods; by engaging in arresting acts, on the other hand, he transcends the five elements.4 All actions which are prompted by desire bring results that are characterized as furthering happiness both here and in the hereafter, and culminating in becoming one of the gods. This category of attainment from this point on gets the appellation of abhyudaya, prosperity or promotion. The highest good, on the other hand, to be achieved by action which is “arresting,” nivṛtta or disengaging, and not prompted by desire, is defined as “going beyond the five elements,” which is a common idiom of liberation. These “arresting” actions are identified in Manu as internal sacrifices, giving up the common Vedic rituals, cultivating knowledge of the Self, self-control, austerity.5 3 athātaḥ puruṣa-niḥśreyasārthaṁ dharma-jijñāsā. jñātvā cānutiṣ han dhārmikaḥ. praśasyatamo bhavati loke pretya ca svarga-lokaṁ samaśnute. VDhS 1.1-3. Translation Olivelle 2000. 4 sukhābhyudayikaṁ caiva naiḥśreyasikam eva ca | prav ttaṁ ca niv ttaṁ ca dvividhaṁ karma vaidikam || 12.88 || iha cāmutra vā kāmyaṁ prav ttaṁ karma kīrtyate | niṣkāmaṁ jñāta-pūrvaṁ tu niv ttaṁ upadiśyate || 12.89 || prav ttaṁ karma saṁsevyaṁ devānām eti sāmyatām | niv ttaṁ sevamānas tu bhūtāny atyeti pañca vai. 12.88-90. Translation Olivelle 2005. 5 Manu 12.93-93, 104. 146  While the doctrine presented in Manu can be taken as merely illustrative of a wider but not at all uniform worldview, it does conveniently point to a new ideal occupying the post of the highest good of human life, one which is essentially negative: niḥ reyasa here is not an attainment of something new and conducive to human happiness, but is a form of absence. The formulation appears at the very end of Manu, in the 12th chapter which talks about action and the process of rebirth closely related to it, where the goal is to “go beyond the five elements,” the sphere of action and matter, to destroy one’s “impurities” and “taints resulting from action” and attain the old Vedic ideal of immortality by means of knowing, renunciation and asceticism.6 This is a new classification of what the Veda can provide for men: (1) all the happiness here and in the hereafter, abhyudaya, and, (2) final liberation, niḥ reyasa. The two will become a comprehensive pair of ideals and an all-important rubric in Advaita Vedānta. It is not quite possible to give a single, uniform definition of liberation, and pursuing the early history of the idea is beyond the scope of this undertaking,7 but for our purposes a convenient way to begin thinking about liberation is through what Halbfass calls the “therapeutic paradigm,” a worldview in which liberation is likened to the state of medical health understood as the absence of disease.8 The majority of terms for liberation in Indian intellectual history are, in fact, negative expressions pointing to getting rid of something undesirable: mokṣa, mukti, nirv ṇa, nivṛtti, nirvṛtti, nirodha, kaivalya, apavarga, h na, an vṛtti, apunar- vṛtti, apunar- bhava, apunar-janma. The “disease” one needs to get rid of is generally identified as suffering, duḥkha, and the process of rebirth, saṁs ra, commonly understood not as being but as 6 Manu 12.101-4. 7 A sketch of such history can be drawn from Bronkhorst 2007a:15-54. 8 Halbfass 1991:243-263. 147 becoming.9 Suffering and rebirth are immediately caused by action (karma), and action as we commonly think of it, good and bad acts in general, bodily, mental, and vocal. Action, thus, constitutes a problem, and on its part, it is commonly a symptom of a more persistent condition, characterized by mental states or psychological torments such as desire, anger, greed (k ma, krodha, lobha). These can further be traced to ignorance (avidy ), which in the Brahmanical systems generally stands for the misapprehension of one thing as another, the Self as the body, but is often equated with a cosmic principle, “primordial matter.” Liberation is, therefore, at its barest breaking away from rebirth (saṁs ra) which is fueled by action (karma), and remaining without a body. This is necessarily a crude depiction of the bare contours, but if we disregard for the moment the early Upaniṣads and early Vedānta, it is a depiction of a worldview shared by all Brahmanical systems except for Mīmāṁsā. If we generalize even further the characterization of ignorance, it is a worldview shared by Buddhists as well. Significantly, it is a worldview prominent in compositions which became the most important smṛtis for Vedic theology, namely Manu, the Bhagavad-G t , and the Mah bh rata. It is also quite commonly taken as the single feature in virtue of which it is possible to think about Indian philosophy as a unique and unified field.10 9 The Bh gavata 10.1.4, for instance, talks about becoming or bhava as a form disease, and bhava-roga or bh va- roga is otherwise a popular and often used term. 10 For instance, the great historian of Indian philosophy Surendranath Dasgupta in his Yoga Philosophy in Relation to Other Systems of Indian Thought (1930:8-11) talks about the doctrine of karma and mukti as “the two fixed postulates which Hindu philosophy could not disavow even in its highest soarings,” in all its systems except for the Cārvākas. Karl Potter (1965) had organized the different systems of India’s philosophies around the doctrine of liberation, in an idiosyncratic but quite illuminating approach. We need not bother with this, nor with the attempts to break away from that paradigm and associate Indian philosophy as philosophy strictly with reason and argument, in which soteriological concerns are commonly absent (see, for instance, Ganeri 2001), so long we appreciate how central a place liberation has been accorded in defining the field of Indian philosophy. 148  Now, we saw in the previous chapter that Mīmāṁsakas took it for granted that “all men strive after happiness.” The worldview which was the result of such a premise was fundamentally positive: life is good, happiness is possible both here and in the hereafter, and ritual action is the way to secure it. At the center of the worldview were desire and ritual. The doctrine of liberation, on the other hand, started with the opposite premiseŚ “All men strive after the destruction of the chain of rebirth, whether they know it or not.”11 Happiness is not possible, life is not good, and action and desire are part of the problem, because the second prompts one to act and action leads to rebirth. The two had also essentially different understanding of what desire and significant action were: Vedic theologians by karma understood ritual, heteronomous action, one which is enjoined in the Veda. The traditions of liberation took karma as action in its broader and more autonomous sense, action in which one is responsible for one’s moral choices.12 Mīmāṁsakas saw desire as that which prompts one to act for one’s goodś liberationists saw desire as what prompts one to act, but commonly despite better judgement and against one’s benefit.13 There was also a plethora of ways proposed for preventing action that fuels saṁs ra, from the literal stopping of action in extreme asceticism, to intellectually understanding the eternal Self as inactive, and to acting with understanding in between.14 11 yataḥ sarvo vidvān avidvāṁś ca saṁsāra-kṣayam icchati. SKG 17, p.91. 12 The SK, for instance, rejected Vedic ritual as “impure” because of involving violence and excessŚ one must kill in sacrifices and one ends up better off than others (SK 2). Sā khya also had a favorite name for ritual action: it was bondage of the honorarium type (d kṣiṇaka bandha), in which one tries to buy one’s place in eternity by paying priests to do ritual for him (Vācaspati on SK 44, p.60Ś iṣ āpūrtena dākṣi akaḥ. puruṣa-tattvānabhijño hīṣtāpūrtakārī kāmopahata-manā badhyate iti). Another new development, the doctrine of the three guṇas or modalities of matter conditioning one’s awareness—a doctrine absent in the Vedic corpus including the old Upaniṣads and absent in Mīmāṁsā—took over the place of what determines the destiny in the hereafter. Winning heaven was no longer contingent on performing ritual, but on cultivating the sattva-guṇa. See, for instance, the last chapter of Manu and the 17th chapter of the Bhagavad-G t . 13 Cf., for instance, the famous exchange from the Bhagavad-G t 3.36-7Ś “Arjuna said: What is it that drives a man to commit evil, Vārṣ eya, however reluctantly, as though propelled by force? The Lord said: It is desire, it is anger, which springs from the force of rajas, the great devourer, the great evil: know that that is the enemy here.” Translation van Buitenen 1981:85. See also K ṣ a’s famous diatribe against ritual in 2.42-5, in which desire is not spared either. 14 See, again, Bronkhorst 2007a:15-54. 149  Both doctrines, however, had called their respective achievements niḥ reyasa, “the highest good,” and this was one of the reasons why, once the term was coopted fully by liberation, svarga was on the one hand treated as a competing doctrine of liberation by Advaitins and others, and Mīmāṁsakas had the occasion to address themselves to the doctrine of liberation on the other. What happened further in Indian intellectual history was a negotiation back and forth between the two, svarga and mokṣa, on what liberation or the highest good really is. In this part of the dissertation, I want to consider ideas of liberation and its status as the highest good of human life in pre-Śa kara Vedic theology. In doing so, I want to pay attention to the following questions: What is liberation—is it merely an absence of rebirth, or rather something positive? If it is the second, what does its positive character consist in? Is it some sort of an attainment, like heaven, and if so, of what kind? More broadly, how can liberation be effected: what is/are the means of liberation? As it is, hopefully, immediately apparent, niḥ reyasa and reyas-kara are etymologically related: if dharma is a means of some human good, can its complex causal structure be applied for achieving the highest good? If so, how? What are the factors involved in liberation, and how are they to be organized to achieve that highest good of human life? What is the role in this of ritual action, what of knowledge of the Self as different from the body? Is there a place for desire in the pursuit of liberation, and if so, of what kind? Does the Veda enjoin liberation in the way it enjoins the winning of virility and heaven? What is the role and nature of the Upaniṣadic knowledge in the pursuit of liberation? As we move through the three chapters of this part of the dissertation, we will see what the prominent answers to these questions were in pre-Śa kara Vedic theology. Our biggest gain will be to appreciate that early Vedic theology had little understanding for and even less interest in knowledge qua knowledge in the pursuit of liberation. The one theme that will emerge in 150 broad contours as the paradigmatic Vedāntic means of liberation will be meditation. We will see essentially two types of Vedāntic meditation, which we can describe as meditation on saguṇa and nirguṇa-brahman respectively, aiming at two different attainments and related to the two doctrines of niḥ reyasa. The significance of the first doctrine of meditation for our ultimate object of understanding the notion of mah -v kya will be largely negative: when we move to discussing Śa kara’s understanding of dharma and niḥ reyasa in the next part of the dissertation, we will appreciate why some Upaniṣadic statements and passages that were immensely important in old Vedānta, for instance injunctions of meditation such as ya tm pahata-p pm vijaro vimṛtyur vi oko vijighatso 'pip saḥ satya-k maḥ satya-saṅkalpaḥ so 'nveṣṭavyaḥ sa vijijñ sitavyaḥ from ChU 8.7.1,15 and the eighth chapter of Chandogya as a unit, did not have a shot at becoming mah -v kyas or at providing the basis for the doctrine of liberation that went along with the mah -v kya notion. On the other hand, the problems that will emerge when we move to discussing the second type of Vedāntic meditation, specifically the question of mediacy of scriptural knowledge against the immediacy of the alleged direct vision of Brahman attained through meditation, as well as the relational character of language and the issue of how words and thought can grasp and express a non-dual entity such as Brahman, will have an immense positive contribution to mah -v kya. Reflection on the identity statements of the Upaniṣads, the future maha-v kyas, as the direct instrument of liberation, de facto replaced this second type of meditation, and with that it inherited these two specific problems. Some of the key characteristics of and questions about mah -v kya developed in dialogue with and as a direct response to this doctrine of meditation. 15 “The self that is free from evil, from old age and death, free from sorrow, free from hunger and thirst, whose desires and intentions are real—that is the self that you should try to discover, that is the self that should be investigated and known distinctly.” 151  The three chapters are organized not systematically—that is, not by discussing the pertinent questions and issues independently—but by investigating the notions of the highest good and liberation in the thought of the most prominent pre-Śa kara Vedic theologians individually. Chapter Three discusses niḥ reyasa and liberation in pre-Śa kara Mīmāṁsā, focusing on Kumārila Bha a. In Chapter Four, I move to early Vedānta and the doctrine of liberation in the Brahma-Sūtra of Bādarāya a. I conclude this part with the doctrine of meditation on nirguṇa-brahman, otherwise known as the prasaṅkhy na-v da, in the theology of Bhart prapañca and Ma ḍana Miśra. The three chapters are conceived as partially independent units: they can be read individually, but each subsequent chapter looks back at the preceding ones. 152  CHAPTER THREE: THE HIGHEST GOOD AND LIBERATION IN PRE- A KARA M M S On this point some blabbermouths, ignorant of the specific meaning of what the ruti text says, proclaim that sons and similar things are means of liberation.1 Heaven as Liberation Scholars customarily say that the idea of liberation was absent in early Mīmāṁsā and that it was inorganic to the system. We may quote Halbfass’s assessment as quite representativeŚ “Final liberation (mokṣa), commonly accepted as a leading theme or even as the basic concern of philosophical thought, does not play any role in the older literature of the systemś Mīmāṁsā deals with dharma, not with mokṣa.”2 Halbfass goes on to argue that this attitude towards liberation changes with Kumārila, at whose hands Mīmāṁsā is transformed into a rounded and comprehensive philosophical system, in which “karma and saṁsāra, as well as mokṣa, become more significant and manifest in thought and argumentation, not so much as explicit themes, but as tacitly accepted presuppositions or as points of reference and orientation.”3 They become reference points, but not determinants of the system, and “the theme of final liberation … is not really his [Kumārila's] own concern.”4 This characterization certainly seems correct in view of Śabara’s silence on liberation, yet curiously if we look at Mīmāṁsā reception by Śabara’s and Kumārila’s contemporaries, it does not come as rightŚ Mīmāṁsakas are read as advocates of heaven as a state of liberation. 1 atra kecid vāvadukāḥ śruty-ukta-viśeṣārthānabhijñāḥ santaḥ putrādi-sādhanānāṁ mokṣārthatāṁ vadanti. BĀUBh 1.5.18, VIII.196. 2 Halbfass 1991:300-1. 3 Ibid. 4 Op.cit, 306. 153  We saw in the previous chapter that Śabara theorized the human good which dharma brings as “heaven” or svarga, not to be understood as a place one goes to in the hereafter or as some pleasurable substance of the more mundane kind, but just as a state of happiness. Śabara developed his account under some constrictions of the Mīmāṁsā ritual technology: it was necessary to define heaven as a state rather than a place or a substance, because only in that case could heaven be superordinate to the ritual performance: by the requirements of Vedic ritual, substances are generally for the purpose of the ritual (kartvartha), not for the purpose of man (puruṣ rtha), and places are not things which one can really produce. We also saw that Śabara defined heaven as a state of felicity which is desirable equally to all men, and a state under which all attainments desirable to men could be classified.5 This made the category quite open for different associations and determinations. While only the state of felicity would qualify as heaven, Śabara did not specify what causes the appearance of such happiness or whence it would derive, or even what this happiness precisely is. And, while a specific place to which the deceased go after death might not be what heaven is, that did not preclude the possibility that there is such a place in which the state of happiness called heaven can be experienced. Kumārila, in fact, refining Śabara’s definition, added that heaven is to be experienced in a specific place, other than the one in which the ritual action was performed: heaven is unsurpassed happiness, nirati aya-pr ti, to be enjoyed in another place and a future life.6 We should note that Vedāntins also commonly associated svarga with bliss. Śa kara, for one, commenting on the last paragraph of the Kena Upaniṣad, saysŚ “’In the heavenly world’ 5 prītir hi svargaḥ. sarvaś ca prītiṁ prārthyate. MSŚBh 4.3.15, IV.1256.ś sarva-puruṣārthābhidhāyī sāmānya-vacanaḥ śabdaḥ na viśeṣe avasthāpito bhavati. MSŚBh 4.3.20, IV.1258. 6 “We will explain in the sixth book that heaven and hell can only be experienced in another place and life, owing to being unsurpassable felicity and suffering in nature, and not immediately after the performance of the ritual action.” svarga-narakau ca niratiśaya-sukha-duḥkhātmakatvād deśāntara-janmāntarānubhavanīyau na karmāntaraṁ sambhavata iti ṣaṣ hādye vakṣyāmaḥ. TV 2.1.5, I.397. 154 means in Brahman that is bliss in nature. Because of the qualification ‘endless,’ the word ‘heaven’ does not refer to the place of the gods (triviṣṭapa).”7 On the other hand, there were, it appears, Mīmāṁsakas who claimed that heaven and hell can refer only to the happiness and suffering that are experienced in this very lifetime as a consequence of the performance of sacrifice and of engaging in prohibited action respectively. 8 The category of heaven was open. What Śabara did claim was that the attainment of heaven through ritual action is the highest good of human life, niḥ reyasa.9 So, if heaven was the term for all human attainments and if it was the term for the highest good of human life, a term which others by that time used for liberation,10 someone could have easily drawn the conclusion that Mīmāṁsakas hope to attain liberation by the performance of ritual action. And some sure enough did. Bhavya, for one, prefaced his critique of Mīmāṁsā precisely with such a viewŚ Without any sense of shame, some deny that meditation and insight [constitute] the true way to deliverance. They insist that deliverance can only be achieved by rituals. They say that according to tradition there is no other correct way to deliverance than the rituals prescribed in the sacred texts., i.e. [rituals that involve] rice, cattle, butter and intercourse with one's spouse.11 7 KUPBh 4.9, IV.76: anante aparyante svarge loke sukhātmake brahma īty etat. anante iti viśeṣān na triviṣ ape ananta-śabdaḥ, Cf. KUVBh on the same (p.121)Ś anante apāre avidyamānānte svarge loke sukha-prāye nirduḥkhātmani pare brahma i. See also Gonda 1966:73-107. On Triviṣ apa as the heaven of Indra, see Hopkins 1915:58-61. 8 We know this from Kumārila’s opponent in TV 2.1.5, whom he describes as nipunaṁ-many ḥ, “fancying himself clever.” “Because heaven and hell are happiness and suffering in nature, they are experienced in this very lifetime, right after the performance of the ritual.” sukha-duḥkhātmakatvena samāneṣv eva janmasu. kriyānantaram eveha staḥ svarga-narakāv api. (IV.392) Kataoka (2011b) seems quite justified in proposing that this is Bhar rmitra talking, Kumārila’s arch enemy if we believe Pārthasārathi under ŚV Pratijña 10, where Kumārila says that Mīmāṁsā has been largely made mundane in the world and Pārthasārathi addsŚ bhart mitrādibhiḥ. On Bhart mitra see Nakamura (2004:170-73) and Pandey (1983:229-36). 9 Under MS 1.1.2 (I.11, 13), Śabara described dharma as that category, consisting of sacrifices such as Jyotiṣ oma, which relates man to the highest attainment, niḥ reyasa. so 'rthaḥ puruṣaṁ niḥśreyasena saṁyunaktiś … ko 'rthaḥ? yo niḥśreyasāya jyotiṣ omādiḥ. 10 We saw it used in Manu, but along with apavarga it was the favorite term of Naiyāyikas as well, as is evident from Vātsyāyana’s commentary on NS 1.1.1-2. See also Potter 1977:28-34. 11 eke ‘pavarga-san-mārga-dhyāna-jñānāpavādinaḥ | kriyā-mātre a tat-prāptiṁ pratipādyāna-patrapāḥ || śāstrokta-vrīhi-paśv-ājya-patnī-sambandha-karma aḥ | 155 Furthermore, Śa kara, clearly pointing to Kumārila, was even more explicit in understanding svarga as a competing account of liberationŚ “Or rather, liberation is a result just of ritual actions, since such ritual actions are the cause of unsurpassed happiness which is denoted by the word ‘heaven.’”12 Yet Kumārila himself, though read by his contemporaries and close successors as identifying heaven with liberation—Bhavya used the term apavarga and Śa kara mokṣa— explicitly declined to make that identification in the context of discussing mokṣa. “If you propose that liberation consists in enjoying happiness, then it would be equivalent to heaven and it would be perishable.”13 “He who is after liberation and does not desire heaven, because heaven is bondage in nature…”14 The identification also assails our sensibility to the idea of liberation: how can there be liberation when heaven is to be experienced in another body? Is not liberation precisely freedom from rebirth and embodiment? Let us leave this question for now and revisit it in the conclusion. Liberation in Kum rila’s ThoughtŚ Introduction We opened this part with the scholarly recognition that liberation does not play any role in Śabara’s commentary to the MS, while it is an implicit presupposition in Kumārila’s thought. In fact, liberation was not just an implicit presupposition, but something Kumārila was quite explicit about: of the four traditional goals of human life or puruṣ rthas, all knowledge about matters of dharma and mokṣa originates in the Veda or in some lost portions of the Veda that nānyo mārgo ‘pavargāya yukta ity āhur āgamāt. MH 9.1-2. Translation Lindtner 2001b, with slight modification. 12 athavā, niratiśayāyāḥ prīteḥ svarga-śabda-vācyāyā karma-hetutvāt karmabhya eva mokṣaḥ. TUBh Introduction. 13 sukhopabhoga-rūpaś ca yadi mokṣaḥ prakalpyate | svarga eva bhaved eṣa paryāye a kṣayo ca saḥ. ŚV Sambandh kṣepa-Parih ra 105. 14 yaḥ svargaṁ na kāmayate bandhātmakatvān mokṣārthī. 6.3.2, IV.1408. 156 were kept in a form of memory, smṛti, whereas the other two are in the domain of worldly dealings.15 This assertion was formulated as part of Kumārila’s attempt to account for all practices, doctrines, disciplines of knowledge and texts that have taken hold in contemporary Hindu society or even beyond it as having an origin in one of the two basic pram ṇas corresponding to the two spheres, perception for the empirical and the Veda for the trans- empirical.16 Kumārila, in other words, did not just presuppose the worldview of karma, saṁs ra and liberation, but claimed that knowledge about them is the sole province of the Vedas. Kumārila addressed questions pertaining to liberation directly or indirectly in several places in his works and presented accounts which seem different if not mutually exclusive.17 This prompted modern scholars to conclude that he was changing his mind about the issue and reevaluating the role of sacrifice and the Veda, and in some cases to use the several accounts as good material for one of the favorite sports of Teutonic Indology, called “ascertaining the chronological sequence of Kumārila’s works.” Rogue Mesquita and Kiyotaka Yoshimizu, for instance, think that Kumārila’s initial understanding of liberation was similar to that of Nyāya, a negative state without either pain or pleasure, but that he later moved towards a Vedāntic notion. John Taber also recognizes two basic accounts but does not find them sufficiently different to constitute a real shift in understanding: the differences can be attributed to different attire Kumārila dons, a philosopher in the ŚV and a theologian in the TV. Larry McCrea doubts 15 tatra yāvad dharma-mokṣa-sambandhi tad veda-prabhavam. yat tv artha-sukha-viṣayaṁ tal loka-vyavahāra- pūrvakam iti vivektavyam. TV 1.3.2, I.166. 16 See TV 1.3.2. Kumārila’s project of pursuing the origins did not mean that he had committed himself to endorsing just anything that “good Vedic men ( iṣṭas) might accept,” to use Śa kara’s turn of phrase (BSBh 2.1.3), as having a truth value. On the contrary, he traced many of the doctrines against which he argued vehemently, such as those of Sā khya, Nyāya and Vedānta, to misinterpreting arthav das or inspirational stories in the Veda and the smṛti literature as real ontological accounts. 17 The two central arguments are found in the ŚV Sambandh kṣepa-Parih ra chapter and in TV 1.3.29. Liberation is also treated in the lost Bṛhat-ṭ k , some fragments of which have been found in later works of his followers and opponents (here I will use what has been included in Taber 2007); relevant is also the Ṭup-Ṭ k 6.3.1-2. 157 whether the ŚV account amounts to a real acceptance of liberationś the TV account certainly does and it represents a change in Kumārila’s understanding, but the change is a result of “a more sophisticated application of pre-existing Mīmāṁsā hermeneutical rules to the injunctions of self-knowledge.” 18 I concur with the view that there are two basic accounts, and with Taber against Mesquita that the B account is a refinement of the ŚV account. Contra Taber, I see the two as quite different, and contra McCrea it does not seem to me necessary to conclude that they constitute a change in Kumārila’s position. It seems possible, rather, to argue that Kumārila talks about two related but different things in the two accounts: there are two different kinds of liberation after which men may strive, and the Veda can provide the means for both. Kum rila’s First Account of Liberation In both accounts of liberation, the important questions are: what precisely the direct means of liberation is; what is the role of the Upaniṣadic knowledge of the Self in the Veda generally; and, is such knowledge a means of liberation. The first account is part of the Sambandh kṣepa- Parih ra chapter of the ŚV, and it may be worthwhile to be mindful of the context in which it appears. The chapter treats objections to the Mīmāṁsā view that the word-reference relation is natural, not brought into being either by a convention of men or by some first principle such as a creator of the world. Kumārila tackles the second by proving that there could not be such a creator or a first principle to begin with, in the course of which he refutes theories of creation in the Veda and the smṛti literature and in Nyāya-Vaiśeṣika, Vedānta and Sā khya. His account of liberation is inserted in a general course of refutation, in a context where it is not always straightforward to determine what he is willing to allow just for the sake of argument, but does not really accept. We should be careful, therefore, not to read the account out of its context. 18 Mesquita 1994, Taber 2007, Yoshimizu 2007, McCrea 2012. Compare also Ram-Prasad 2000a and 2000b. 158  And the context, to be specific, concerns the role of discriminative knowledge which Sā khyas and others claim to be a means of liberation, that is, knowledge from the Upaniṣads which presents the Self as different from the body.19 The Sā khyas are likely the philosophical school that bears that name, but the term could also stand for anyone whose way of liberation was through cultivation of knowledge as understanding, and specifically “knowledge of the absolute separateness of the soul and body, and independence of the soul from all acts and qualities.”20 This was in any case the soteriological significance of the term s ṅkhya, as demonstrated by Edgerton almost a century ago, and the systematic Sā khya would have been a prominent part of that wider complex. I will base my reading of the context on the systematic Sā khya, but we should bear in mind the broader context. Kumārila is less concerned with the details of what liberation is and more with how to achieve it, but the second is contingent on the first. Liberation here is defined in negative terms. Throughout it is referred to solely as mokṣa21 (something we should bear in mind, as it will be important when we move to the second account) and it is explicitly distinguished from heaven: liberation does not consist in the experience of happiness, because in that case it would be equal to heaven, a state produced by ritual action, and consequently it would not be eternal. “A thing which is caused cannot be imperishable.”22 19 Cf. verse 102Ś na ca sā khyādi-vijñānān mokṣo vedena codyate. PārthasārathiŚ dehād viveka-jñānam – avin v are ‘yam tm ity-ādi. … tatra yat sā khyādibhir mokṣa-sādhanatvenāśritaṁ viveka-jñānaṁ tad vedena mokṣarthaṁ na coditam iti. 20 Edgerton 1924:2-3. 21 In the ten core verses of the ŚV account mokṣa/mucyate are the sole terms used to refer to liberation (seven times in all), while in the B account the Sā khya-Yoga term kaivalya is used. 22 “If it be postulated that liberation consists in enjoying happiness, then it would correspond to heaven and it would be perishable. A thing which is caused cannot be imperishable.” sukhopabhoga-rūpaś ca yadi mokṣaḥ prakalpyate | svarga eva bhaved eṣa paryāye a kṣayo ca saḥ || 105 || na hi kāra avat kiñcid akṣayitvena gamyate | 106ab | 159  Liberation is, thus, essentially negative in nature. It is not a form of fashioning that should produce a novel and positive state of affairs, but it is an absence, abh va, getting rid of something undesirable, specifically the reason which keeps the Self in bondage. Liberation is, in other words, absence of bondage, absence of the relation of the Self to a body.23 With this Kumārila is treading a fine lineŚ he wants liberation to be eternal, which entails that it cannot be produced in the manner of heaven, but unlike Advaita Vedāntins he does not understand liberation as a state which the Self enjoys for all eternity without being aware of it. For Kumārila bondage is real and liberation is a future state of affairs no less so than heaven, yet one not produced, such that once achieved it will be there for good. Pārthasārathi interpreted the absence that is liberation as a form of permanent posterior non-existence through destruction (pradhvaṁs bh va), akin to the non-existence of the broken pot, and an anterior non-existence (pr g-abh va) of future bodies that will never be.24 Śa kara read Kumārila’s argument along similar lines. The means of achieving liberation should correspond to the expected result, and given that the result is an absence, its achieving must proceed not through production which depends on action, but through elimination.25 “One is released through the absence of the cause.”26 The question, then, becomes, what is the cause of bondage—what is it that requires eliminating—and how to put an end to it. Throughout the account, it is clear that what keeps the Self in bondage in 23 See his comments on 106Ś bhāva-rūpaṁ sarvam utpatti-dharmakaṁ gha ādi kṣaya-dharmakam eva, ato na sukhātmikā muktir ātma-jñānena kriyate iti … śarīra-sambandho bandhaḥ, tad-abhāvo mokṣaḥ. 24 niṣpannānāṁ dehānāṁ yaḥ pradhvaṁsābhāvaḥ yaś cānutpannānāṁ prāg-abhāvaḥ sa mokṣaḥ. Ibid. 25 “Except for its negative nature, there is no other ground for the eternality of liberation. And, no absence can be accepted as a result of any action.” na hy abhāvātmakaṁ muktvā mokṣa-nityatva-kāra am | na ca kriyāyāḥ kasyāścid abhāvaḥ phalam iṣyate || 107 || 26 hetv-abhāvena mucyate. ŚV Sambandh kṣepa-Parih ra 106. 160 Kumārila’s understanding is action and its results, well-known to us as the law of karma or action that produces results that perpetuate embodiment.27 We should make it abundantly clear, however, that “action” that is karma which gives rise to embodiment does not refer to action of just any kind (kriy ): involuntary or unconscious behavior such as blinking or sneezing, basic voluntary actions such as moving one’s arm and purposeful activity such as walking or jogging are largely inconsequential for karma, because such actions do not produce results that perpetuate embodiment. Furthermore, although karma is customarily defined as “moral action,”28 and in the traditions of liberation it did stand for acts done with the body, mind or speech in which moral choices matter,29 karma for Kumārila had nothing to do with moral concerns: what makes for good or bad karma are not acts which might strike our moral sensibility as “good” or “bad,” nor the purity of one’s life and awareness, which the Brahmanical traditions of liberation had associated with the three guṇas and made crucial for rebirth. What matters is how such acts are treated in the Veda.30 What Kumārila meant by “action” was something like an intersection of (1) intentional, future-oriented action, and (2) forms of voluntary personal and social behavior, both regulated by the Veda. That was the scope of karma-producing action, and Mīmāṁsakas classified it in two primary groups: (1) Vedic rituals which produce good karma in general (svarga or heaven) or of some particular form (such as good progeny, cattle). These are the rituals that we were concerned with in the previous 27 “Therefore, one is released through the exhaustion of karma alone, when the cause [of embodiment] is no more.” tasmān karma-kṣayād eva hetv-abhāvena mucyate, 106cd. Cf. Pārthasārathi thereonŚ karma-nimittaś ca bandhaḥ karma-kṣayād eva na bhavatīti. 28 See Doniger’s edited volume on karma (1980b), where the theme of morality runs as a constant throughout the contributions. See also Collins 1982Ś29Ś “karma – ‘action’, moral retribution’ … the moral quality of actions performed previously – usually but not necessarily in past lives – determines the happiness or suffering experienced thereafter.” 29 See the 12th chapter of Manu, 16th and 17th chapters of the Bhagavad-G t , S ṅkhya-K rik 2-3 for starters. 30 The most lucid treatment of karma in general and its relations to moral concerns in Indian philosophy (with a significant focus on Kumārila) remains Halbfass 1991, chapters four and nine. 161 chapters, and Mīmāṁsakas called them ritual actions of the optional variety (k mya-karma), i.e., rituals whose performance is occasioned by a desire for some good result on the part of the performer; (2) Action prohibited by the Veda (niṣiddha-karma), producing bad karma (pratyav ya, “the reverse,” naraka, hell; durita, bad lot; doṣaṭp pa, fault). The second consisted of social behavior proscribed by the Veda on the one hand, and malicious rituals enjoined in the Veda, such as the notorious yena, a ritual intended for effectuating the death of one’s enemy (and, it would appear, for giving Mīmāṁsakas sleepless nights of intense exegesis), on the other. Let us note for now that it is the action regulated by the Veda which matters for the purposes of bondage and liberation, and that such action can produce good and bad karma. There is this new karma on the one hand, referring directly to actions over which the Veda has a say and which produce good or bad results in the future, and there is a karmic stock consisting of the results of actions accumulated through countless lives and culminating in the present body on the other. These two together constitute what ultimately keeps the Self in bondage. We should remember at this point that the context in which Kumārila discusses liberation was, first, the role of the discriminative knowledge of the Self, which Sā khyas and others consider to be the means of liberation, and, second, the claim that such discriminative knowledge is enjoined in the Upaniṣads as a means of liberation. The Sā khya account amounted to claiming that discriminative knowledge of the Self independently stops karma as the link that keeps the Self bound to a body. For Sā khyas, karma was just a symptom of bondage, not its root cause. The root cause of bondage was ignorance (ajñ na), consisting in confusing the Self with the body or more generally with matter (prakṛtiṭpradh na) or some of its evolutes, and liberation was a state in which the Self has finally got rid of ignorance and through perfect discriminative 162 knowledge remains an uninvolved spectator.31 After this, Sā khyas thought, there still remains some karmic inertia: knowledge has for all purposes put an end to the store of karmic residue, and one has achieved what may be described as j van-mukti or liberation while living, but the Self continues to be embodied till death, “through the force of the past impressions, like the potter’s wheel that continues to whirl because of the impulse which it previously received.”32 With death there comes liberation which is absolute and final: full separation from the body.33 This meant that the instrumental cause of liberation is knowledge that distinguishes the Self from matter, and that karma, being just a symptom of bondage, disappears shortly after ignorance has been stopped. Such was the account of liberation which Kumārila argued against. Now, Kumārila conceded that knowledge which distinguishes the Self from the body does put an end to its antipode, ajñ na or ignorance; however, he claimed, all that could possibly follow from this is that one would stop creating new karma. “If ignorance be proposed as the cause of the rise of actions, from the destruction of ignorance no rise of [new] actions would follow, but one would still not get rid of the results of [already performed] actions.”34 Kumārila’s account is contingent on this concession, real or provisional. Further, it is inferable that this absence of the rise of new actions is primarily tied to good karma, one which is obtained by the 31 There hardly is a need of substantiating this from the Sā khya literatureŚ these themes are basic to Sā khya and they abound in the SK and its commentaries. A few references from Gauḍapāda’s Bh ṣya may suffice. “Bondage is caused by ignorance” (bandhaḥ ajñānaṁ nimittam, SKGBh 44, p.191)ś “Liberation is obtained by discriminative knowledge (of the Self) from of the manifest and the non-manifest (matter and the evolutes)” (vyaktāvyakta- vijñānān mokṣaḥ prāpyate, SKG 17, p.90)ś “When there is knowledge of the 25 principles, which is characterized by the awareness that the soul is different from the body, and one knows: this is pradḥāna, this is buddhi, this aha kāra, these are the five elements, from which the soul is different—from this knowledge the subtle body is destroyed, wherefrom liberation follows” (yadā pañca-viṁśati-tattva-jñānaṁ syāt sattva-puruṣānyathā-khyāti-lakṣa am—idaṁ pradhānam, iyaṁ buddhiḥ, ayam ahaṁkāraḥ, imāni pañca mahā-bhūtāni yebhyo 'nyaḥ puruṣo visad śa iti. evaṁ jñānāl li ga-niv ttis tato mokṣa iti; SKGBh 55, p.180). 32 tiṣ hati saṁskāra-vaśāc cakra-bhrama-vad dh ta-śarīraḥ. SK 67. The SK does not use the term j van-mukti, but the state described surely admits of such a designation. See Jakubczack 2004. 33 SK 65-7. 34 utpattau karma āṁ ceṣ am ajñānaṁ kāra aṁ yadi | tan-nāśāt syād anutpattis teṣāṁ na phala-varjanam || 101 || 163 performance of the optional Vedic rituals, k mya-karma. These are the new actions that do not arise, the optional Vedic rituals, and therefore Kumārila’s presupposition is that the destruction of ignorance does not give rise to new actions and their consequent karma because one has lost the desire for good karma. Discriminative knowledge of the Self engenders dispassion, which consists in the absence of desire for good karma. Kumārila does not say as much, but it can be inferred that such is his presupposition from verse 95— knowledge deconstructs attachment—as well as verse 111.35 Elsewhere he explicitly associates the absence of the desire for heaven (which, we will remember, provides the occasion for k mya-karma) with the pursuit of liberation.36 In the B refinement of the argument it is also clear that knowledge of the Self results in dispassion, which further reinforces knowledge.37 Let us note this well: knowledge removes ignorance consisting in confusing the Self with the body, which gives rise to dispassion/absence of desire, which in its turn does not prompt one to perform optional Vedic rituals, which stops the creation of good karma. This is a crucial point for Kumārila’s argument—the pursuit of liberation precludes desires for good results—as it also accounts for his insistence that knowledge of the Self presented in the Upaniṣads is not a means of liberation because it cannot be enjoined for some good of man (puruṣ rtha). We should remember at this point that for Mīmāṁsakas the category of for-the-good-of-man (puruṣ rtha) was for the most part restricted to the primary ritual action, to which all other elements of the ritual were subordinate, because it is from the ritual action that 35 “A result (of a sacrifice) is known (to follow) when it is desired, and it could not come to the one who does not want it. Since this (dispassion) is present in the knower of the Self, knowledge is useful (through that).” prārthyamānaṁ phalaṁ jñātaṁ na cānicchor bhaviṣyati | ātma-jñe caitad astīti taj jñānam upayujyate || 111 || 36 yaḥ svargaṁ na kāmayate bandhātmakatvān mokṣārthī… 6.3.2, IV.1408. 37 nitya-naimittikair eva kurvā o durita-kṣayam | jñānaṁ ca vimalī-kurvann abhyāsena ca pācayan | vairāgyāt pakva-vijñānaḥ kaivalyaṁ bhajate naraḥ. Verse 6 in Taber’s selection. 164 man gets the desired future result. The result itself, for which the action serves as a means, was subordinate to man, and the whole pursuit was provoked by man’s desire. In other words, no k ma, no puruṣ rthaṬ For knowledge of the Self to be enjoined, it must cater to some desire, and it is enjoined and is useful for providing the information that there is an eternal Self which will enjoy the results of the ritual in the future, because without that piece of information one would not take to the ritual performance.38 We should not confuse, however, desire as Mīmāṁsakas understood it with motivation in general. Kumārila understands that liberationists have a purpose, he calls them mokṣ rthins, but k ma is tied to some form of felicity, pr ti, and to productive undertakings. As for the forbidden action that produces bad karma, Kumārila did not say explicitly how such action stops: it may be that the ability to distinguish the Self from the body is sufficient to prevent one from transgression. However, Mīmāṁsakas claimed that bad karma can be amassed not only by prohibited action, but additionally through the failure to perform certain rituals, which belonged to a third, much murkier category called obligatory action, consisting of rites that had to be performed daily or on certain occasions (nitya-naimittika-karma) but without quite a clear understanding why. These rituals included, for instance, the daily Agnihotra and the fortnightly Darśa-pūr amāsa, rituals which were also classified as optional, to be performed for heaven, but which become obligatory when the desire for heaven has been lost in the pursuit of liberation. They become obligatory under a different provision, that one should perform them as long as one lives, and with the concession that their procedure may be simplified provided the 38 na ca sā khyādi-vijñānān mokṣo vedena codyate || 102cd || ātmā jñātavya ity etan mokṣārthaṁ na ca coditam | karma-prav tti-hetutvam ātma-jñānasya lakṣyate || 103 || 165 primary offertorial action is performed. Optional for those who will perform them no matter what, obligatory for those who might not bother at all.39 Mīmāṁsakas were not clear why these rituals must be performed. The MS just says that the non-performance of the obligatory rites results in a fault, which Śabara interpreted as “loss of heaven” (and which is tantamount to their being “for the sake of heaven,” as in their optional version);40 Kumārila argued that they have to be performed because otherwise bad karma would be created—“they bind the one who does not perform them”—while later adding that they can exhaust some old karma as well. 41 It seems clear, nevertheless, that they were obligatory precisely so that they would apply even after the desire for heaven has been replaced by the pursuit of liberation.42 If one was bent on achieving liberation, one would have to continue performing these obligatory rituals and thus prevent future or exhaust present bad karma. The abstention from optional rituals and acts proscribed in the Veda on the one hand and the performance of the obligatory rituals on the other should, thus, stop the replenishing of the 39 On these, see MS 6.3.1-7 and the commentaries. Dharmadhikari 2006:329-36 provides a good but not exhaustive overview of how nitya-karmas have been conceptualized in history (but not through history); it is quite clear from his account that the tradition struggled with how to precisely understand them, particularly in their relationship with the optional rituals. See also Krishan 1994. 40 “The ruti text says that a fault follows from the omission of the principal ritual actŚ ‘He who is a performer of Darśa-pūr amāsa is certainly cut from heaven who fails to perform the ritual on a full moon or a new moon. Speaking about a fault when the principal act is not performed, the ruti text shows that the act is obligatory.” pradhānātikrame doṣaḥ śrūyate, apa vā eṣa svargāl lokāc chidyate yo darśa-pūr amāsa-yājī san paur amāsīm amāvāsyāṁ vātipātayed iti pradhānātikrame doṣaṁ bruvaṁs tasya nityatāṁ darśayati. MSŚBh 6.3.3, IV.1411. 41 “One should perform the obligatory rites with the intention of warding off diminution (nitya-naimittike kuryāt pratyavāya-jihāsayā, ŚV Sambandh kṣepa-Parih ra 110cd)ś” “The regular and conditional rites, relating to the different varṇas and ramas, must be performed for the purpose of exhausting previous sins and the prevention of future sin on the account of their non-performance (praty-āśrama-var a-niyatāni nitya-naimittika-karmā y api pūrva-k ta-durita-kṣayārtham akara a-nimittānāgata-pratyavāya-parihārārthaṁ ca kartavyāni, TV 1.3.29, I.288)ś” “The obligatory rituals also bind the one who does not perform them (badhnanty akurvantaṁ nitya-naimittikāny api, B v.1)ś” “Exhausting bad karma by means of the obligatory rites (nitya-naimittikair eva kurvā o durita-kṣayam B v.6)ś” See also Yoshimizu 2007Ś203-12. 42 Cf. Dharmadhikari 2006Ś331 (in regard to Pārthasārathi)Ś “Now the question arises why svarga is not to be regarded as the result of the nimitta here. The answer is that svarga is not expected by those who are mumukṣus, i.e. desirous of liberation from the bondage of Karman, while pratyav ya-parih ra, i.e. keeping away the sin is expected by all those who perform the Nitya sacrifice, like agnihotra, etc.” 166 karmic stock. As for the old stock, some bad karma may be annulled by the performance of the obligatory rituals, but in general karma has to be exhausted through time by experiencing the karmic consequences; once that happens—the B says “in 10,000,000,000 age cycles”—one will become liberated.43 Kumārila, thus, did not allow knowledge and dispassion to exhaust the accumulated karma altogether by reducing it to inertia, and through this he denied any direct efficacy to knowledge in the pursuit of liberation.44 Such knowledge was reduced to a precondition for the giving up of optional rituals, and at best we could hazard a guess that its causal contribution would have been absorbed in the non-performance of karma through the principle of mediate causality (p ramparya) that we discussed in the previous chapter. Kumārila’s thinking in this was characteristically m m ṁsic: pursuing the factor which plays the most direct role. Actions are most immediately related to a future state of affairs of any kind, and it is because of actions that one gets another body after death. If action is the cause of embodiment, then a way must be found to stop action. I take the B argument to be a refinement of the ŚV argument because the two share features which distinguish them from the TV argument. First, in both accounts the non- performance of optional and forbidden action on the one hand and the performance of obligatory action on the other play the central role in achieving liberation. Second, both are predicated on 43 yuga-ko i-sahasre a kaścid eko vimucyate, v.4. 44 “Knowledge is not a means of liberation, because it is not a counterpart to the potential state of action. To be sure, it could not be antithetical to it. Even if we accept that actions, like attachment, etc., are brought about by ignorance, actions in their state of potentiality are not made inoperative through knowledge. There is no reliable warrant that there occurs a destruction of actions through knowledge, nor that only a slight result is experienced, like in the case of a transgression committed by a prince.” tac-chakty-apratiyogitvān na jñānaṁ mokṣa-kāra am | karma-śaktyā na hi jñānaṁ virodham upagacchati || yady apy ajñāna-janyatvaṁ karma ām avagamyate | rāgādi-vat tathāpy eṣāṁ na jñānena nirākriyā || karma-kṣayo hi vijñānād ity etac cāpramā avat | phalasyālpasya vā dānaṁ rāja-putrāparādha-vat. ŚV Sambandh kṣepa-Parih ra 94-6. 167 the absence of desire or k ma. In the B the role of knowledge of the Self is explicitly more positive. It is a precondition for taking up the process of liberation, but also something which undergoes development: the performance of obligatory rituals assists the perfecting and maturation of knowledge of the Self through exhausting the karmic stock, and such knowledge finally culminates in solitude of the Self. Let us note, though, that even in this case knowledge of the Self does not have an instrumental role in achieving liberation. It could not possibly have such role, for two reasons: first, it is not a form of action, which is generally the instrumental factor that brings the good of man (puruṣ rtha); it is, rather, a form of understanding, which, second, could become instrumental only in virtue of a specific desire (as I shall show under the next heading). This knowledge is, therefore, not a means of liberation, but a state that undergoes an evolution from being a precondition (absence of desire through disappearance of ignorance), turning into a patient (the perfecting and maturation of knowledge, intensification of dispassion) and maturing in a result (solitude). There is, unfortunately, some paucity in Kumārila’s account, but we could surmise that the disappearance of ignorance is also gradual, and that it initially consists in disillusionment with the world of sacrifice and desire, but that it eventually matures as understanding becomes clearer. Answering some of the questions we posed in the introduction, we can say that the means of liberation under this account is a combination of performing some actions while abstaining from others. Knowledge of the Self is required, but its causal efficacy is remote and absorbed in dispassion. Liberation is a state of isolation of the Self and it does not involve positive experience. Let us finish with how Śa kara read this account: liberation is just a state of remaining a pure Self without striving for anything, achieved by abstaining from the optional and prohibited 168 acts, by exhausting the karmic stock through enjoyment, and by preventing bad karma by performing the obligatory rituals. The obligatory rituals could also exhaust some old karma, both good and bad, as long as it hasn’t started bearing fruits. Liberation is both a future state and eternal, in the manner of the posterior absence through destruction.45 Although Śa kara does exaggerate a tad Kumārila’s argument—as far as I have seen, Kumārila does not say that obligatory rites could exhaust good karma—that is probably just for covering all bases. In commenting on this, Sureśvara leaves little doubt whom the master was addressing. He quotes directly the ŚV, followed by the sassy remark “Those who fancy themselves Mīmāṁsakas thus reject knowledge of the Self and claim that ritual is a means of liberation.”46 Śa kara’s interest is, in fact, piqued because he likes what the account amounts to, “remaining a Self without striving,” which is his own definition of liberation—and shared, we should mention, with the YS47—with the inevitable tm ca brahma, but is scandalized by everything else. Kum rila’s Second Account of Liberation The second account, to which we turn presently, is an account of liberation which is not quite that. As I already noted, the question in virtue of which the two accounts are about the same thing concerns not precisely liberation, but generally the nature of the knowledge of the Self in the Veda and specifically its use in the pursuit of liberation or other goals. In fact, Kumārila in the TV 1.3.29 does not talk about mokṣa, kaivalya, apavarga at all, but about niḥ reyasa and its paired counterpart abhyudaya. This is an important point, for two reasons. First, niḥ reyasa, as I 45 kāmya-pratiṣiddhayor anārambhāt ārabdhasya ca upabhogena kṣayāt nityānuṣ hānena ca pratyavāyābhāvāt ayatnata eva svātmany avasthānaṁ mokṣaḥ ... iṣ āniṣ a-phalānām anārabdhānāṁ kṣayārthāni nityāni iti cet ... yad dhi naṣ am, tad eva notpadyata iti pradhvaṁsābhāva-van nityo 'pi mokṣa ārabhya eveti cet. TUBh Introduction, VI.8. Similar discussion in BSBh 4.3.14. 46 iti mīmāṁsakaṁ-manyaiḥ karmoktaṁ mokṣa-sādhanam | pratyākhyāyātma-vijñānaṁ tatra nyāyena nir ayaḥ. TUBhV 1.10ś the verse quoted in 1.9 is ŚV Sambandh kṣepa- Parih ra 110. 47 tadā draṣṭuḥ svarūpe ‘vasthānam. YS 1.3. 169 noted in the introduction, was a term commonly used for liberation in Indian intellectual history, but not exclusively so. As an axiological term, referring to “that good which has no higher to itself,” it could and did stand for any good one might have considered the highest, and we saw that the VDhS and Śabara used it for heaven. Second, because of being a value term, niḥ reyasa connoted a commitment to the designated good, a commitment not necessarily expressed by the negative terms. We also saw that the ŚV account was inserted in a course of refutation of competing ideologies, and McCrea had alerted us that Kumārila’s endorsement of liberation seemed quite provisional. It is difficult, however, to talk about the highest good without some commitment to it. Before we can appreciate fully Kumārila’s reasoning in the TV, we need to acquaint ourselves with the ideology to which the account is tied. Here knowledge of the Self is not associated to Sā khya and similar doctrines, which wanted liberation to follow just from understanding that the Self is different from the body, but to chapter eight of the Ch ndogya Upaniṣad, which talks about knowledge of the Self as a means of fulfilling desires. The chapter consists of two parts which are related in structure: both begin with an injunction which says that the Self should be investigated; the first part talks about various worlds and desires that one may win and fulfill by knowing the Selfś the second part is story about Prajāpati teaching the gods and the titans whose purpose is to illustrate how this investigation of the Self might proceed; both parts talk about what may be described as the final attainment accomplished by means of knowledge of the Self. For our purposes, it is not necessary to penetrate fully the logic of the chapter, but we need to note a couple of things. First, the injunction runsŚ “One should investigate, clearly understand the Self that is free from faults, ageless and immortal, free from sorrow, not liable to 170 hunger and thirst, the Self whose desires and intentions are true. He who investigates and understands this Self obtains all the worlds, and all his desires are fulfilled.”48 Second, the results which are promised to follow from knowledge of the Self in the injunction indicate immediately that the Upaniṣad is not interested in knowledge of the Self as such, but as a means of fulfilling desires and winning the heavenly spheres. It also quickly becomes obvious that the “true” desires and intentions from the injunction are not desires and intentions for the Self, as we may be prone to think due to long acquaintance with readings of the Upaniṣads through the lens of Advaita Vedāntaś49 they are desires and intentions which, through knowledge of the Self, become immediately fulfilled and realized. These are the same familiar desires for the heavenly world of ancestors containing the common choice delights, but the issue is how to obtain freedom of motion to visit them at will, without depending on the exigencies of ritual action.50 And then, there is the familiar concern with re-death: how does one make sure that the highest heaven which one may obtain is not a result that will eventually perish? “[A]s here in this world the possession of a territory won by action comes to an end, so in the hereafter a world won by merit comes to an end.”51 In other words, the Vedic worldview of desires which 48 ya ātmāpahata-pāpmā vijaro vim tyur viśoko vijighatso 'pipāsaḥ satya-kāmaḥ satya-sa kalpaḥ so 'nveṣ avyaḥ sa vijijñāsitavyaḥ. sa sarvāṁś ca lokān āpnoti sarvāṁś ca kāmān yas tam ātmānam anuvidya vijānāti. This is the version of the injunction which opens the second part of the chapter. The first version, right at the beginning of the chapter, runsŚ “Now, here in this fort of brahman there is a small lotus, a dwelling place, and within it, a small space. In that space there is something—and that’s what you should try to investigate, that’s what you should try to know distinctly.” Translation Olivelle, with a slight modification. 49 Cf. for instance Deussen 1980Ś190Ś “These ‘true wishes’ are here delineated in a rather odd, clumsy manner. … This subsection [8.2] from its spirit and tone, stands off so much from the whole and interrupts and disturbs the whole context … that we conjecture in it a perceptible delineation of ideas [8.]3.2 by another hand … perhaps by the same hand which, at the conclusion of the previous chapter (Prap ṭhaka), explained the fulfillment of all things in the sense of a magical unfoldment of the person into many manifold individuals.” 50 See 8.2. The worlds attained by discovering the Self include those of the fathers, mothers, brothers, sisters, friends, perfumes and garlands, food and drink, music and women. “[T]hose here in this world who depart after discovering the self and these real desires obtain complete freedom of movement in all the worlds (8.1.8).” “Anyone who knows this goes to the heavenly world every day (8.3.5).” Olivelle’s translation. 51 8.1.6, Olivelle’s translation. 171 we discussed in the previous chapter and in the Introduction is not under question, but is affirmed throughout. The pressing concern is not how to get free from desires, but how to make sure they are not thwarted. The chapter also presents what may be called the highest attainment, referred to as “the highest light” (paraṁ jyotiḥ) and associated with the world of Brahman (brahma-loka), described very graphically: it is the third heaven from here, having two seas, Ara and Ṇya, a lake by the name of Airaṁmadīya, a banyan tree known as Somasavana, a fort called Aparājita and Brahman’s golden hall by the name of Prabhu.52 There is a path leading to it that goes from the human heart, where the Self resides, through one of the channels issuing from it—the central one leading to the top of the head—and the sun rays which form a continuum with these channels and culminate in the sun itself: the door of heaven. This is the stairway to the heaven of Brahman, and the password at the door is Om.53 This highest attainment is a place from which one does not return again to this “gray and toothless state, to the toothless, gray and slobbery state.”54 Attaining the world of Brahman presupposes liberationŚ “Shaking off evil, like a horse its hair, and freeing myself, like the moon from Rāhu’s jaws, I, the perfected self ( tman), cast off the body, the imperfect, and attain the world of brahman.”55 Wining this highest attainment is absolutely predicated on knowing the true Self, and a crucial role is played by the character of the Self in deep sleep, a common Upaniṣadic theme which presents the Self as persisting but remaining in a state of non-transitive awareness, not liable to the faults of waking and dream. This is a crucial role, because here the state of the Self 52 8.3.4; 8.4.3; 8.5.3; 8.12.2-5; 8.13. 53 8.6. 54 8.14-15. 55 8.13.1: aśva iva romā i vidhūya papaṁ candra iva rāhor mukhāt pramucya dhūtvā śarīram ak taṁ k tātmā brahma-lokam abhisaṁbhavāmīty abhisaṁbhavāmīti. Olivelle’s translation. 172 in deep sleep is a border state, a state in which the Self maintains its existence but is not liable to faults, death, old age, grief, hunger and thirst, simply by not being aware of its identities in waking and dream.56 However, it is a border that has to be crossed too, because there the Self “does not perceive itself fully as 'I am this,' ... It has become completely annihilated.”57 One, in other words, has to reach this state of the Self in deep sleep because it is a state of full separation from the body, but then one has to re-emerge in a positive situation of one’s “own true form,” svena rūpeṇa.58 This true form is one in which the Self is no longer embodied, but keeps the innate faculties of sight, smell, speech, hearing, thought, which function through its innate mind, its “divine sight.”59 These have, obviously, been suspended in the state of deep sleep, in virtue of which disembodiment and freedom from all faults is attained, but once that happens, they re- emerge to function autonomously, without their bodily seats. This state of the Self, “the highest person” (uttama-puruṣa) which is disembodied but positive—“like the wind, rain-cloud, lightning and thunder”—goes on to enjoy in the world of Brahman “with women, chariots and relatives,”60 with all desires fulfilled and all worlds won. Back to Kumārila now, we should remember that dharma as a means of some good was for the most part restricted to the central ritual action, because it was the ritual action which ultimately brings the desired good. This made the ritual action puruṣ rtha, subservient to the purpose of man, while all else in the sacrifice was deemed kratvartha, subservient to the purpose of the ritual itself. There was, however, a scenario under which a ritual meant for a particular result could also bring an added value, if instead of a standard ritual item a substitute was used. 56 That this is a crucial state is quite evident from 8.6.3Ś “So, when someone is sound asleep here, totally collected and serene, and sees no dreams, he has slipped into these veins. No evil thing can touch him, for he is then linked with radiance.” Olivelle’s translation. 57 8.12. 58 8.3.4; 8.12.3 59 8.12.4-5. 60 8.12.2-3. 173 Take, for instance, the daily Agnihotra ritual, an offering of milk into the sacrificial fire, which, as we saw above, was classified as an obligatory ritual whose purpose was to prevent the creation of bad karma. Now, if the sacrificer were to offer yoghurt instead of milk, this change would transform the nature of the sacrifice from obligatory to optional, and the sacrifice would not only prevent bad karma as its common result, but would also bring heaven as an added value. Two points are important to note here. First, what brings about the change in the nature of a sacrifice from nitya to k mya is the desire on the part of the sacrificer: the sacrificer is prompted by the desire for heaven to use yoghurt instead of milk. Second, yoghurt which, being a substance, is naturally subservient to the ritual, undergoes trans-instrumentalization and becomes subservient to the purpose of man (puruṣ rtha), the element which brings about the desired value, all the while remaining subservient to the ritual in other sacrifices. By this principle, yoghurt could be theorized both as kratvartha and puruṣ rtha. It would remain subservient to the central ritual action of offering in general, but it would maintain autonomy where there is an injunction which establishes a direct causal relation between it and, say, heaven, through an appropriate desire. This trans-instrumentalization would not work if the statement which says that the use of yoghurt brings particular results belongs to a context of another sacrifice where yoghurt is a common offertory, in which case the statement of results relating to its offering would have to be interpreted as arthav da advertising the ritual. For our purposes, we ought to note well again that a category cannot be puruṣ rtha without there being a desire for a specific result on the part of the agent.61 61 This is worked out at the beginning of the third p da of the fourth adhy ya of MS, where particularly important are sūtra 1, which states the general principle that statements about results associated with substances are arthav da, and sūtra 5, which states the exception. Some useful information on this can be gleaned from Yoshimizu 2004. 174  Kumārila applies this principle over the knowledge of the Self to theorize how it can both be kratvartha and puruṣ rthaṬ We saw in the previous chapter and in Kumārila’s first account of liberation that the knowledge of the Self from the Upaniṣads served as an impetus for taking up ritual action, as well as the principle which secures the relation between engagement and disengagement in actions and the future results which they bring: there is a permanent Self in which the ritual continues to exist and mature until blossoming in heaven.62 This was no mean role to play, since without such knowledge a ritual would not happen: An intelligent person will not do an action unless s/he knows that such action is for her good. Nevertheless, such knowledge was subordinate to the ritual, and it was not the element which directly procures the expected human good. But, imagine a scenario in which knowing the Self as different from the body is: (1) enjoined by a Vedic injunction; (2) associated with attainments of the kind which dharma brings in general; (3) which attainments are related to specific human desires; and (4) such knowing cannot in any straightforward manner be related to another ritual through the context. Under such a scenario, knowledge of the Self would undergo trans-instrumentalization in the manner of yoghurt and it would become a direct means of achieving a specific desired human good, puruṣ rtha. Precisely such a scenario is in play in the eighth prap ṭhaka of the Chandogya Upaniṣad that we touched upon above. First, there is an injunction that the Self should be investigated and known distinctly. Second, there are specific results of the kind which dharma produces that are accomplished by this knowledge, which, as I said above, Kumārila classified under the two groups of abhyudaya and niḥ reyasa. Kumārila defines abhyudaya here as the obtaining of 62 “Because, without it (knowledge of the Self) there would be no relation to engagement or disengagement in actions which bring results in the hereafter.” tena vinā paraloka-phaleṣu karmasu prav tti-niv tty-asambandhāt. TV 1.3.29, I.288. 175 supernatural powers of the kind produced by the practice of Yoga, such as the ability to become atomic in size.63 In the context of chapters seven and eight of the Ch ndogya this clearly refers to the freedom of motion (k ma-c ra) and the ability to visit the heavenly spheres.64 Niḥ reyasa, on the other hand, standing for the highest good, is defined as “a result which is a state of attaining the supreme Self, a state consisting in no-return.”65 Kumārila does not say what the “supreme Self” or param tman refers to here, but it likely is a gloss of the paraṁ jyotiḥ or “the highest light” from the Ch ndogya, or perhaps the “highest person,” uttama-puruṣa, the last and highest state of the Self attained after it had emerged from deep sleep and attained the world of Brahman, with its cognitive powers restored, unembodied but “in its own form.” As we shall see later, by his time the two were in any case seen as related. Third, these two attainments got through knowledge of the Self are clearly related to desires, and Kumārila explicitly treats them as a pairŚ the attainments consist in fulfilling desires, and the two form sentential supplements to the injunction for knowing the Self. Finally, it is not possible unambiguously to relate these attainments through context to some other ritual in which knowledge of the Self would be serving the purpose of the ritual (kratvartha), and therefore they cannot be explained away as being arthav da.66 From all of this it must follow that the results in the form of abhyudaya and niḥ reyasa are real, and that knowledge of the Self as the thing enjoined for their achieving is puruṣ rtha. Knowledge of the Self has, thus, become a crucial item for achieving prosperity and liberation, joining a select club consisting of yoghurt, the 63 yoga-janyā imādy-aṣ a-gu aiśvarya-phalāni var itāni, Ibid. 64 From the seventh chapter he cites 7.1.3, tarati okam tmavit, and from the eight 8.2 and 8.7.1, both of which talk about attaining all worlds, fulfilling all desires of the heavenly kind. 65 apunar-āv tty-ātmaka-paramātma-prāpty-avasthā-phala-vacanam. TV 1.3.29, I.288. 66 “Since an unambiguous connection or disconnection to some ritual cannot be established through their forming a part of some context, these statements about results are not arthav da like those about the ointment, the kh dira wood, the sruva etc.” aprakara a-gatatvenānaikāntika-kratu-sambandhāsambandhāc ca nāñjana-khādira-sruva- vākyādi-phala-śruti-vad arthavādatvam. Ibid. 176 wooden ladle, the milking vessel, the post made of khadira wood. We can almost visualize how scandalized Śa kara must have been by this. Just as in the first account, the attainment of liberation which is related to the injunctions of knowledge of the Self must be accompanied by the performance of the rituals and actions obligatory for all classes in general or respectively. The two are meant for accomplishing different results in this pursuit: action exhausts previous bad karma and prevents the creation of new karma, and knowledge of the Self brings abhyudaya and niḥ reyasaṬ Since they both produce results independently of one another, serve different purposes and proceed through different courses, they are not mutually optional, exclusive, or subordinate to one another.67 Since they both contribute to liberation as the final result, we may venture that Kumārila saw their relationship akin to that of two or more equally principal, apūrva-producing actions in the complex sacrifices such as the Darśa-pūr amāsa which we discussed in the previous chapter. This is rather an important point, because this second scenario of Kumārila does amount to a form of what has been theorized as a combination of knowledge and action as both exercising causal efficacy in the pursuit of liberation (jñ na-karma-samuccaya). Ideally this was a combination between two independent and principal causal elements, pradh na, but it was also understood as a relationship between a primary means and its direct subordinate, if the subordinate remained causally efficacious.68 What Kumārila presented in the TV amounts to a 67 “And, the injunction of knowledge does not preclude a relation to action. The obligatory and optional actions, relating to the different classes and ramas, have to be performed for the purpose of exhausting previous sins and the prevention of future sin on the account of their non-performance. Because these [knowledge on the Self one hand and action of the two types on the other] have different purposes and proceed through different paths, they do not cancel one another, become options or subordinate parts to one another.” na ca jñāna-vidhānena karma- sambandha-vāra am. praty-āśrama-var a-niyatāni nitya-naimittika-karmā y api pūrva-k ta-durita-kṣayārtham akara a-nimittānāgata-pratyavāya-parihārārthaṁ ca kartavyāni. na ca teṣāṁ bhinna-prayojanatvād bhinna-mārgatvāc ca bādha-vikalpa-parasparā gā gi-bhāvāḥ sambhavanti. Ibid. 68 Samuccaya in the MS and Śabara thereon seems to refer to combining things without any relationship of subordination, and surely such is the sense in the BS 3.3.58, which says that meditations which bring different 177 combination of action and knowledge as mutually independent (and therefore equally principal) means in the attainment of liberation. We will continue with Kumārila for a tad longer in the concluding remarks, but let us finish this section by reemphasizing that what makes the two accounts so different is the absence of desires in the first and their presence in the second. McCrea had suggested that Kumārila did change his ideas about the role of the knowledge of the Self from the ŚV to the TV, in virtue of which his understanding of liberation also must have changed, and he traced this change to a better grasp of pre-existing Mīmāṁsā rules of interpretation. However, it is quite evident that Kumārila was aware of this specific rule—how a thing can serve both the need of the ritual and the need of man—already in the ŚV, and it was precisely in consideration of this rule that he extended the definition of dharma as a means of some good from the limited scope of action to offertories and ritual details as well. I find it preferable, therefore, to suggest that he was acting in the role of a hermeneut to accommodate two very different accounts of liberation, both of which had currency in the Vedic theology of his time. We will touch upon this question in the conclusion of this part of the dissertation. Concluding Remarks It would be worthwhile here to take stock of Kumārila’s accounts, to put things in perspective and see clearly just what concerns around liberation were at stake. Kumārila defined dharma, the province of the Veda, as a means of some good, reyas-kara, a means that is of instrumental nature, a s dhana, which should bring about a state (avasth ) of felicity (pr ti), which is of the nature of an accomplishment and is necessarily future (s dhya). A central role in this was played results may be combined indiscriminately, since they produce their own results independently. Advaitins have understood the relationship in broader terms, as is evident, for instance, from the first chapter of Sureśvara’s NaiS. 178 by k ma, human desire for some form of happiness that is the motive for taking up the means of achieving the respective good. Sensu stricto, the Veda was the reliable warrant on things of the s dhya character, and authoritative on facts about already constituted existents (siddha) only insofar as such existents are employable and required for the accomplishment of the future state. All these categories were crucial in the discussions of liberation in Vedānta as wellŚ is liberation a new state that should be attained, or an already constituted existent, or something else? If it is not a new state, can the means for its accomplishment be puruṣ rtha? And, what is character of such means? The common Mīmāṁsā means, ritual action, was necessarily s dhya, and even when things of the siddha character were recognized as directly instrumental, they were subordinate to an action. Yoghurt may bring heaven for all we know, but not unless it becomes a part of a sacrifice, a part subordinate to the central action of offering, to be specific. And what about desires, so intimately related to the idea of puruṣ rtha? Does not liberation presuppose freedom from desire? Or, are there, perhaps, some other desires different from the common Vedic desires for sons, worlds, heavenly delights, but satisfying the criterion of puruṣ rtha? If there are, what are they precisely? Furthermore, what is happening with the descriptions of enjoying the same old Vedic objects of desires in the state of liberation in the Upaniṣads? Given that heaven was already associated with the highest possible bliss, a state created by action, a perishable product, could one possibly enjoy the same heavenly delights in liberation without compromising eternity? Finally, if liberation is solely in the province of the Veda, is it enjoined by a Vedic injunction in the same manner as ritual is? These questions occupied the center stage in early Advaita Vedānta, but they were explicitly or implicitly negotiated much before that. In Kumārila’s account, knowledge of the Self was of the siddha character: it is about an existent, and as such it naturally serves the needs of the ritual and not the good of man. In his 179 first account, liberation was presented in negative terms, a state of future absence that can be accomplished only through negative means, namely absence of karma as the most direct cause of embodiment. He denied that knowledge of the Self is enjoined for liberation, and he described liberation as a state with a single word: kaivalya, a solitude of the Self. The pursuit of liberation was predicated on absence of desire. This scenario was not a common case of Mīmāṁsā procedure. It could, however, be accommodated under Kumārila’s understanding of the nature of the Self: The Self is essentially pure awareness, an eternal omnipresent sentient substance, which experiences happiness, distress and the like directly and not through an assumed identity with the intellect as in Vedānta, and which for that reason undergoes change. Purify it of all karma, starve it from all experiences and it will remain a pure subject.69 Liberation was possible (and achievable only by means of action), but nowhere did he say that it was desirable. His second account, on the other hand, had all the hallmarks of a proper Mīmāṁsā procedure and result. Knowledge of the Self was enjoined for liberation. As we shall see shortly, though being of the siddha character, Kumārila’s commentators thought that it becomes a part of a central process, one which can be understood as s dhya. The pursuit of liberation was predicated on desire of some kind, and liberation as a result was an attainment of a state (avasth -pr pti). In fact, as I said in the beginning of the previous section, this was no mere liberation: the Upaniṣadic text to which Kumārila tied his account made liberation a precondition which is surpassed by and absorbed in the final attainment, reaching the world of Brahman, and Kumārila was clear about the desirability of this attainmentŚ it is niḥ reyasa, the highest good. For all we can see, Kumārila’s second account presents liberation as an attainment which is very similar to heaven. What Kumārila was, in fact, doing is a reaffirmation of the old Vedic 69 The two central places in which Kumārila discusses the Self are the tma-V da chapter of the ŚV, and TV 2.1.5. 180 ideal through absorbing the “new discovery” that is liberation: through knowledge of the Self as eternal, the attainment of heaven can be made permanent. Heaven is still a new state, one which the Self does not experience presently, but because the Self is eternal, one who knows this Self can attain heaven as a permanent state. This seems to be Kumārila’s way of tackling the problem which assumed central importance in his first account, but is completely ignored in the second: heaven is happiness and bondage, and if liberation is a new, positive state, consisting in experience of happiness, it would be “tantamount to heaven and perishable.” Well, if the element which is central in its attainment is eternal, perhaps the new state can be imperishable as well. It is, possibly, also important that this account did not present ritual action as directly productive of the future state. Ritual is causally efficacious, but its results are exhausting the stock of bad karma and preventing its replenishing; ritual is responsible for liberation in the strict sense, what the Upaniṣad describes as the freedom from fault, but not for the attainment of the new state. Finally, in the two accounts Kumārila applied two models of instrumental causal efficacy over the role of the knowledge of the Self in liberation, without explicitly mentioning either. In the first, knowledge of the Self gives rise to dispassion; its contribution is absorbed in dispassion, which becomes the cause for abstaining from performing optional Vedic rituals, which absence in its turn is the immediate cause of depleting the stock of good karma. A fitting model: negative means for a negative result. Knowledge of the Self was thrice remote and not directly related to the ultimate result. This was the p ramparya model of mediate causality that we introduced in the previous chapter. In the second model, jñ na-karma-samuccaya, both ritual and knowledge were directly contributing something to liberation and were to be performed side by side: ritual exhausts old karma whereby one becomes free, corresponding perhaps to the “shaking off evil, 181 like a horse its hair, and freeing myself” of the Upaniṣad, while knowledge of the Self brings the new state of affairs. We should bear in mind that these two were not mutually opposed models of causality. Applied over a complex ritual performance they work perfectly well together. Yet, they do amount to stratification of causal elements and hierarchy of means. Under the second model it is not possible to deny causal efficacy to what has been recognized as a means, even if that is just a helping element, because the principal means would fail without the assisting means. Under the first model, a means had ceased being a means once it has produced its result, and it is absorbed in a new production. Its causality has been mediated. All things considered, Kumārila presented two accounts of liberation, both of which are compatible with his understanding of the Self and which, therefore, need not constitute a change in his thinking. In doing so he was accommodating different understandings of liberation that had currency in Vedic theology of his time. The first was, to use Potter’s categories, an account of liberation from; a procedure for such liberation is not enjoined in the Veda, “but here is what I can do for you.” The second is not only liberation from, but liberation to as well, and not only is it enjoined in the Veda, but it is the highest attainment one can aspire for.70 Now, Kumārila’s commentators have related his two accounts to two competing models of liberation in early Vedānta, and more generally in Vedic theology. The Sā khyan knowledge of the Self as different from the body “which is not enjoined in the Veda for liberation” was identified both by Pārthasārathi and Someśvara as referring to the famous Yājñavalkya-Maitreyī dialogue from the Bṛhad- raṇyaka Upaniṣad.71 As we shall see in the next part, the whole fourth chapter of the BĀU containing Yājñavalkya’s teachings to Janaka and Maitreyī formed an early major account of liberation in Vedānta, one which was very different from the one in Ch ndogya 70 Potter 1965. 71 Pārthasārathi on ŚV Sambandh kṣepa-Parih ra 102 and Someśvara on TV 1.3.29. 182 and rectified to conform the Ch ndogya in the Brahma-Sūtra. We’ve already seen that Kumārila himself associated the second account to the 8th prap ṭhaka of the Ch ndogya Upaniṣad, and Pārthasārathi described this second reference to knowledge of the Self as meditational knowledge, one which forms a part of a larger meditational complex involving breath control and the other well-known constituents of yoga. This amounted to saying that discriminative knowledge of the Self, knowledge qua knowledge, is not enjoined for liberation, whereas meditational knowledge can be. This detail serves as a good transition point from Mīmāṁsā onto early Vedānta. 183  CHAPTER FOUR: LIBERATION IN THE BRAHMA-S TRA When one has attained Brahman through the meditation on Brahman that has the mentioned characteristics, one attains freedom of movement. – But, how could there be such freedom of movement everywhere without performing the requisite practice occasioned by the intention for attaining the respective world? – The attainment itself is the cause.1 Introductory Remarks: Methodology In this chapter, I will reconstruct the doctrine of liberation in Bādarāya a’s Brahma-Sūtra. I will show that the BS doctrine was based on the 8th chapter of the Ch ndogya Upaniṣad that we are already acquainted with from Kumārila’s second account of liberation. To be more specific, I will show that liberation in the BS meant attaining the highest Vedic heaven, brahma-loka, necessarily after one’s death, through ascending the universe via the “course of the gods” or deva-y na. The experience of liberation consisted in independence or sovereignty, “having no other master but oneself,” which specifically referred to the ability to enjoy all desires “of the Vedic variety” and the ability to move through the heavens of the Vedic world without impediment. This was a liberation characterized by enjoyment, bhoga. Additionally, liberation may have also included the bliss of experiencing Brahman’s “own nature.” The process of achieving liberation was meditation on Brahman as one’s Self. This meditation was absorption of one’s thought process in a single notion. It had to be accompanied by the performance of ritual by all practitioners except one group, lifelong celibates, and by other religious practices, as well as by cultivation of personal virtues. Both meditation and the other assisting practices had to 1 VPS 3.3.40, p.1030. 184 continue till the end of one’s life. My reconstruction will be based on the third and fourth books of the BS, where liberation and the process of achieving it are explicitly treated. The BS is, as is well-known, the tersest of all sūtra compositions and the meaning of the majority of individual sūtras is all but impossible to comprehend without a commentary. I should like to lay out, therefore, the methodology that I will employ in this reconstruction. It is based on four principles, ordered successively by importance. First, there are several sūtras that can be read without the aid of a commentary, and often they are of crucial importance. Their meaning is clear either directly or because they contain an unambiguous reference to what is customarily called viṣaya-v kya or topical text, typically but not exclusively a passage from the Upaniṣads. Let me give instances of both. (1) In the first p da of the fourth book, there is a series of sūtras, seven through eleven, from which it is immediately obvious that the process of liberation is meditative absorption of some kind. This meditative absorption must be practiced in a sitting posture and wherever concentration is possible. That much can be read straightforward and without recourse to a commentary. Nowhere else in the book is there another section which presents another or alternative means of liberation, and whenever the commentators present a different means or reinterpret meditation, they are supplying their own doctrine. (2) At the opening of the third p da of the third adhy ya, in the first sūtra where the general principle of how Upaniṣadic meditations are formed is stated, there is an unmistakable reference to M m ṁs -Sūtra 2.4.9. From this reference, it is immediately apparent that these meditations are formed after the model of the Vedic sacrifices that are given in the Brāhma as. Consulting what the MS says on this provides immense help in reading the sūtraś since this is an opening sūtra, it is important for our understanding of the entire book. Similarly, there are often 185 in the sūtras enough cues to point to the precise passage from the Upaniṣads that is under discussion. These also tend to be placed centrally, at the beginning of p das. Good examples are 3.4.1, puruṣ rtho 'taḥ abd d iti b dar yaṇaḥ, which points very likely at TU 2.1.1; and 4.4.1, sampady virbh vaḥ svena abd t, which is a certain reference to ChU 8.12.3. Other certain cases are the famous janm dy asya yataḥ (1.1.2), which is a direct reference to TU 3.1.1, and nandamayo ‘bhyas t (the second chapter of TU). In many cases, therefore, the precise Upaniṣadic passage can be identified, and then reading the passage in context facilitates understanding. Second, on many sūtras in the five commentaries that I decided to consult (on which more in a bit), very often there is a remarkable unanimity in interpretation. Daniel Ingalls had noticed that Śa kara and Bhāskara are often unanimous on the interpretation of many sūtras and that they often seem to follow an older commentator, a vṛttik ra, whose doctrine can be reconstructed by reading the two commentaries side by side.2 This method can be extended further: whenever all commentators agree on the meaning of a sūtra or adhikaraṇa, it is fair to assume that the meaning is unambiguous. In doing so, third, it will be found that the one commentary deviating the most is that of Śa kara, and in many cases these deviations can easily be seen and discarded. Let me give an example. I said that there is a series of sūtras in the first p da of the fourth adhy ya where it is clear that the process of liberation is meditative absorption, to be performed in a sitting posture. All the commentators interpret the sūtra in roughly the same way, and so does Śa kara. But then, Śa kara adds that this refers to “the knower of the lower Brahman,” and that the knower of real Brahman can sit or stand or do whatever he wants, because meditation is dependent on human 2 Ingalls 1954. 186 choice while knowledge is not: knowledge is how things are, and posture has no say in this. The distinction of higher and lower Brahman was, of course, Śa kara’s hallmark doctrine, and the difference between knowledge and meditation was his earth-shattering innovation in Vedānta. But, there is nothing in the BS to warrant such distinctions, so whenever he interprets a sūtra in the straightforward manner before cancelling out the direct meaning, it is fair to stick with the literal interpretation, “bring him back to the fold,” and assume that the meaning is unambiguous. Fourth, in the reconstruction, I will use five commentaries, those of Śa kara, Nimbārka, Śrīnivāsa, Bhāskara and Rāmānuja. Most useful among these for ascertaining Bādarāya a’s own doctrine is Nimbārka’s Ved nta-P rij ta-Saurabha. Nimbārka is still a very mysterious figure in Indian intellectual history, and old scholarship customarily proposed quite a late date for him. More recent studies suggest that the VPS might be the oldest BS commentary available.3 Whatever the case may be, more important than the date is the nature of the commentary. The colophon to every p da of the commentary identifies the work as r raka-m m ṁs -v ky rtha, an exposition of the literal meaning of the BS, and the commentary attempts to do just that: provide the minimum necessary to make the individual sūtras intelligible and identify the topical texts, without engaging in polemics of any kind. Ghate’s study of the BS, which is still the most valuable piece on the doctrine of Bādarāya a, found Nimbārka’s commentary be the closest to the intended meaning of the BS.4 The other commentaries need no introduction, except perhaps that of Śrīnivāsa, who was a follower of Nimbārka.5 He and Rāmānuja offer very similar interpretations, and recently it has 3 See Satyanand 1997, Agarwal 2000, Potter 2013:60-62. 4 Ghate 1981. 5 Sometimes it is claimed that this is a sub-commentary on the Ved nta-P rij ta-Saurabha, but that is not the case; it is a full, independent commentary. 187 been proposed that Śrīnivāsa’s Ved nta-Kaustubha preceded Rāmānuja’s r -Bh ṣya.6 We do not need to worry over such questions, but we do need to note that on a number of sūtras and adhikaraṇas, the commentaries are often divided on readings and interpretation, typically Nimbārka, Rāmānuja and Śrīnivāsa against Śa kara and Bhāskara. In such cases my policy is either to exclude these places from my account, or, if they proved too intriguing, to rely on Ghate’s study and to seek internal coherence in the text.7 Another very valuable and accurate study of liberation in the BS is that of Nakamura, but it is relatively short and no more than an overview.8 I should finally note that by “Bādarāya a” I mean whoever wrote or edited the BS in the form in which it was commented upon, whether it was one person or more of them. Unearthing layers or strands, synchronic or diachronic, is not an object of this study. The Doctrine of Vidy /Up sana The cornerstone of Kumārila’s first account of liberation, the claim that knowledge of the Self as different from the body is just for the good of the ritual and not for the good of man, was in fact a long-standing Mīmāṁsā position. As far as we can infer from the Brahma-Sūtra, the general Mīmāṁsā attitude towards the Upaniṣads was that they provide what is called vidy , some form of esoteric knowledge, whose purpose is to accomplish a refinement or embellishment of certain items in the sacrifice that are themselves for the good of the ritual. This was saṁsk ra or saṁskṛti, one of the four sannip tyopak rakas that we introduced in the Second Chapter. The scriptural justification of this principle was found in a statement from the Ch ndogya Upaniṣad: 6 A good summary of these arguments can be found in Potter 2013:95-7. 7 I have not included here the commentary of Madhva, as Ghate has already shown that it is not of much use in understanding the BS itself. 8 Nakamura 1983:519-32. 188 “Only that which is performed with knowledge, with faith, and with an awareness of the hidden connections becomes truly potent.”9 Two words are important here, vidy and upaniṣad. The context of the passage is the udg tha, the central element of a S ma Veda song, which is here identified with the syllable Om, the understanding of which identity fulfills all kinds of desire. This understanding “embellishes” the udg tha such that if a ritual in which the udg tha is chanted is performed with such esoteric knowledge, the results of the ritual are enhanced. Early Mīmāṁsakas applied this as a universal principle:10 knowing the hidden connections of the udg tha improves the udg tha and enhances the ritual, and likewise knowing the Self as distinct from matter embellishes the Self which is the ritual agent. If some independent result is said to follow from this knowledge of the Self, such as liberation, that is really the commercial break time of the Veda. This is the doctrine with which we are already very well familiar, and the BS puts it directly into Jaimini’s mouth.11 The principle that all kinds of Upaniṣadic vidy s are for refinement was rejected by Bādarāya a as being specific to the udg tha and not general,12 but Jaimini’s particular assertion about knowledge of the Self was a much more complex story. The claim that the Self as an agent and the knowledge of the Self as distinct from the body are subservient to the ritual was never controverted. What Bādarāya a denied was that the Upaniṣads as a scriptural corpus are really, or only, about that. Śa kara was typically eloquent on this pointŚ “Had the transmigrating soul alone, the embodied agent and experiencer, been taught in the Upaniṣads as merely distinct from 9 yad eva vidyayā karoti śraddhayopaniṣadā tad eva vīryavattaraṁ bhavati. ChU 1.1.10, translation Olivelle. 10 All the BS commentators take the Ch ndogya text to be the topical text of BS 3.4.4, tac-chruteḥ, and take it as a statement of a general Mīmāṁsā attitude towards all Upaniṣadic vidy s, and Someśvara under TV 1.3.29 invokes the same text in justifying Kumārila’s claim that discriminative knowledge of the Self serves the purpose of ritual. Note: the sūtras are quoted from and referenced to the edition of VPS (see Bibliography). 11 BS 4.1.2Ś śeṣatvāt puruṣārtha-vādo, yathānyeṣv iti jaiminiḥś “Because the agent is subsidiary to the action, the statement of results is just talk of the good of man, as in other casesś thus Jaimini.” 12 BS 3.4.10Ś asārvatrikī, “non-universal.” Nimbārka, p.1102Ś yad eva vidyay iti śrutir na sarva-vidyā-viṣayā. 189 the body, then the Upaniṣadic statement of results could have been an arthav da in the aforementioned manner.”13 That there is such an entity as a permanent Self which survives death and can enjoy the results of ritual in the future is knowable from the Veda, as the Veda is the reliable warrant on all matters supersensuous, and the Upaniṣads can provide that specific knowledge, which would make them subservient to ritual insofar as they act in that capacity. But, claimed Bādarāya a, the Upaniṣads primarily teach a principle higher than this transmigrating Self, in virtue of which they are the means of another, independent goal of man, liberation. Because of this higher teaching, liberation cannot be explained away as mere talk. This brings us to considering how the BS presented the Upaniṣads as a distinct canon. One of the central presuppositions of Vedic theology presented in the M m ṁs -Sūtra was that the purpose of the Veda was to provide for some good of men that is not available by natural means. In the M m ṁs -Sūtras, this was the role of dharma: dharma was that on which human good depended, and for the most part dharma was limited to ritual. The challenge for Bādarāya a in defining the Upaniṣads as a canon distinct from the Brāhma as of Mīmāṁsā was to show how the Upaniṣads serve human goals in a way different from or independent of ritual. This specific difference of the Upaniṣad was found in the doctrine of vidy and up sana, both of which I will translate as meditation, for reasons which will hopefully become clear soon. Strictly speaking, vidy referred to specific Upaniṣadic sections engaging what Olivelle describes as particular “hidden connections” between two distinct things which are in some way understood as identical.14 These vidy s were either one-off, forming one distinct section in one Upaniṣad; or repeated throughout the canon, sometimes even in the same Vedic branch ( kh ). 13 yadi saṁsāry eva ātmā śārīraḥ kartā bhoktā ca śarīra-mātra-vyatireke a vedānteṣu upadiṣ aḥ syāt, tato var itena prakāre a phala-śruter artha-vādatvaṁ syāt. BSBh 3.4.8, III.721. 14 Olivelle 1998:24-7. 190 By way of illustration of the first we may point to the famous identification of the sacrificial horse and the Universe at the opening of the Bṛhad- raṇyaka Upaniṣad. An example of the second would be the ṇḍilya-vidy , Śā ḍilya’s teaching about how this whole world is Brahman, which is most prominent in the ChU but also repeated in the atapatha Br hmaṇa (ŚB) and in the BĀU,15 or the Vai v nara-vidy , the teaching about the Self which is common to all, forming a part both of the ChU and the BĀU.16 Bādarāya a says that these vidy s constitute single, unique units, no matter whether they are one-off, restricted to one kh or found throughout the canon, in the same way that sacrifices described in different Vedic texts are single ritual models. They aim at the same result, for instance attaining Brahman; they have the same form, for instance they are both about Vaiśvānara the universal Self and involve the same detailsś they start with the same injunction, for instance that one should meditate on this universal Self; and they share the same name.17 They are constituted as units by way of combining the details mentioned in the different texts. Whatever these vidy s turn out to be in the end, let us note that they were understood from early on as the Upaniṣadic counterpart to Vedic ritual. They have an injunction, they aim at a result, and they have a procedure. Vedāntins of different backgrounds quite unanimously used the term up sana as a synonym for vidy , and it is here that we get an abundance of definitions from which it is 15 Of course, the BĀU itself is a part of the ŚB, which makes the vidy repeated twice in the Brāhma a of the Vājasaneyins. 16 ChU 5.11-18, BĀU 5.9. 17 BS 3.3.1: sarva-vedānta-pratyayaṁ codanādy-aviśeṣātś Nimbārka thereon, p.920Ś anekatra proktam apy upāsanam ekam eva, codanādy-aviśeṣāt, “Although a meditation is described in several places, it is one only, because of uniformity of injunction and the rest.” The explicit reference is to MS 2.4.9Ś ekaṁ vā samyoga-rūpa- codanākhyāviśeṣāt. Śabara thereon, II.635-6: sarva-śākhā-pratyayaṁ sarva-brāhma a-pratyayaṁ caikaṁ karma, artha-saṁyogasyāviśeṣāt, tad eva prayojanam uddiśya tad eva vidhīyamānaṁ pratyabhijānīmaḥ. rūpam apy asya tad eva dravya-devatam, puruṣa-prayatnaś ca tād śa eva codyate, nāmadheyaṁ cāviśiṣ am, tena tad eva karma sarva- śākhādiṣu, “The act in all branches and Brāhma a texts is one, because of uniformity of relation to purpose—we cognize the same act enjoined for the same purpose; because of uniformity of form in terms of offertories and deities; because the same human effort is enjoined; because of uniformity of name.” 191 immediately apparent that the intended meaning of both was that of meditation.18 Here is a short selection of definitions taken from works of prominent VedāntinsŚ A uniform stream of thought called contemplation (dhy na), cultivation (bh van ), meditation (up sana)… (Ma ḍana Miśra)19 Since meditation (up sana) if of the nature of contemplation (dhy na)… (Nimbārka)20 Meditation (up sana) is a current of uniform thoughts, not mixed with dissimilar notions, concerning a scriptural object and in a scriptural manner. (Śa kara)21 Meditation (dhy na), to define it, is a stream of awareness fixed on objects such as Deities described in scriptures and unbroken by [thoughts on] things of different kind. They call it concentration. (Śa kara)22 By meditation (up sana), a direct perception of the object of meditation, such as the Lord, is achieved. (Śa kara)23 Viewing an object as taught in Scripture and prolonged dwelling on that till one gets identified with that, is, indeed, said to be meditation. (Sureśvara)24 [T]he teaching of Scripture is conveyed by means of the term 'knowing' (vedana), which is synonymous with meditating (dhy na, up sana). That these terms are so synonymous appears from the fact that the verbs vid, up s, dhyai are in one and the same text used with reference to one and the same object of knowledge. (Rāmānuja)25 Now, dhyai means to think of something not in the way of mere representation (smṛti), but in the way of continued representation. And up s has the same meaning; for we see it used in the sense of thinking with uninterrupted concentration of the mind on the object. We therefore conclude that as the verb 'vid' is used interchangeably with dhyai and up s, the mental activity referred to in texts such as 'he knows Brahman' and the like is an often-repeated continuous representation. (Rāmānuja)26 18 Śa kara, for one, consistently uses vidy and up sana as synonyms, but for him the first was much broader in scope than the second, as we shall see in the next chapter. There is no indication that Vedāntins preceding him made any such distinction between the two. Bādarāya a does not use the term up sana, nor the all-important lexeme brahma-vidy , but he uses the term vidy throughout in the sense of distinct Upaniṣadic units, and names one of them, puruṣa-vidy , in 3.3.24. That he understood these vidy s as meditations is clear from the beginning of the first p da of the fourth adhy ya, where their application is described as dhy na, to be practiced in a sitting position, etc. 19 BrS p.74. 20 upāsanasya dhyāna-rūpatvād. VPS 4.1.8, p.1199. 21 upāsanaṁ ca yathā-śāstraṁ tulya-pratyaya-santatir asa kīr ā ca atat-pratyayaiḥ śāstroktālambana-viṣayā ca. TUBh 1.3.2-4, VI.17. 22 dhyānaṁ nāma śāsroktadevatādy-ālambaneṣv acalo bhinna-jātīyair anantaritaḥ pratyaya-santānaḥ. ekāgrateti yam āhuḥ. ChUBh 7.6.2, VII.431. 23 upāsanena sākṣāt-k te upāsye viṣaye īśvarādau. BSBh 3.3.59, III.710. 24 śāstrārpita-dhiyopetya hy ā tādātmyābhimānataḥ || cirāsanam bhaved arthe tad upāsanam ucyate. TUBhV 1.66, translation Balasubramanian 1984Ś237. 25 dhyānopāsana-paryāye a vedana-śabdenopadeśātś tat-paryāyatvaṁ ca vidy-upāsti-dhyāyatīnām ekasmin viṣaye vedanopadeśa-para-vākyeṣu prayogād avagamyate. ŚBh 4.1.1, I.713-4. Translation Thibaut 1904:714. 26 dhyānaṁ ca cintanam. cintanaṁ ca sm ti-santati-rūpaṁ na sm ti-mātram. upāstir api tad-ekārthaḥ. ekāgra-cinta- v tti-nairantarye prayoga-darśanāt. tad-ubhayaikārthyād asak t-āv tta-santata-sm tir iha brahma veda brahmaiva bhavati (MU 3.2.1); ibid. Translation Thibaut, ibid. 192  For by meditation is understood thought directed upon one object and not disturbed by the ideas of other things. (Rāmānuja)27 Contemplation, consisting of a continuous stream of thoughts having the form of the object contemplated, the synonyms of which are mentation (vedana) and meditation (up sana)… (Śrīnivāsa)28 If any distinction at all should be drawn between the two, vidy seems to stand more generally for the constituted textual ideality of an Upaniṣadic meditation, whereas up sana indicates its facticity in practice. This, however, is a tenuous distinction and it should not be pursued consistently. Besides, several other words were used as full synonyms for vidy and up sana: dhy na, vedana, bh van , dar ana, dṛṣṭi, vijñ na. Bādarāya a exclusively used vidy , but in the commentarial corpus up sana became the term of art, and for a good reason: we will not fail to notice the related etymology of up sana with upaniṣad. The unique feature of the Upaniṣads, then, was expressed in their title: they were meditational texts. Whatever ontological differences Vedāntins have had, before the times of Ma ḍana and Śa kara the Upaniṣads were texts of meditation. Let us briefly illustrate how an Upaniṣadic vidy ṭup sana would have looked like through the aforementioned identification of the sacrificial horse and the Universe and with the help of Śa kara and Sureśvara. In the br hmaṇa the different limbs of the horse which is to be sacrificed in an Aśvamedha are identified with significant spatiotemporal elements and categories of the Vedic worldŚ the horse’s head is dawn, its torso the year, its limbs and joints the seasons, months and fortnights, its feet—days and nights; its sight is the sun, its breath the wind and its gaping mouth the fire common to all men; its underbelly is the earth, its abdomen the intermediate space and its flanks the quarters; its bones are the stars, its flesh the clouds, its 27 ŚBh 4.1.8, I.719Ś dhyānaṁ hi vijātīya-pratyayāntarāvyavahitam eka-cintanam ity uktam. Translation Thibaut 1904:721. 28 dhyātir dhyeyākāra-pratyaya-pravāha-rūpasya dhyāyaty-arthasyopāsana-vedanādy-apara-paryāyasya. VK 4.1.8, p.1200. 193 intestines the rivers. The description goes on, but it is already apparent that the horse is likened to categories of time and space on the one hand, and to certain elements of significance in the Vedic worldview. Note, for instance, the complex of the sacrificial fire and the sun—the earthly and heavenly fire—which are related by the wind that carries sacrificial oblations from earth to heaven. What the br hmaṇa presented for Śa kara and Sureśvara was an identification of the sacrificial horse with the highest divinity of the Vedic worldview, Prajāpati, who was commonly seen as an embodiment of the totality of both the natural world and the world of the Veda, and variously called Hira yagarbha, Virā , Sūtra, k rya-brahma, M tyu or death, etc. The meditation consists in visualizing these correlations: to be specific, it consists in seeing or mentally assigning these spatiotemporal and Vedic categories in or to a specific horse, one which is to be sacrificed in an actual ritual performance. Two criteria of classifying the Upaniṣadic vidy s can be inferred from the BS. The vidy s are, first, instruments of procuring something desirable to menŚ “From this (vidy ) a human goal follows, because of the texts to that effectś so says Bārarāya a.”29 They can, therefore, be classified in terms of the intended result. A second and a more basic criterion is the nature of the correspondence between the two things correlated in the meditation. This correlation can be either real or merely based on a symbol (prat ka). The meditation on the Aśvamedha sacrificial horse is a good example of the second: the head of the horse is not really dawn, but dawn is mentally imposed over the horse’s head and meditated on as such in virtue of some resemblance between the two. Śa kara says, for instance, that primacy is a feature both of dawn and the horse’s head, and this is a ground enough for the one to be visualized as the other.30 In the 29 puruṣārtho ‘taḥ śabdād iti bādarāya aḥ. BS 3.4.1. 30 “Dawn, the period relating to Brahman, … is the horse’s head because of primacyś and, the head is the predominant the parts of the bodyś” uṣā iti, brāhmo muhurtaḥ uṣāḥ … śiraḥ, prādhānyātś śiraś ca pradhānaṁ śarīrāvayavānām. BĀUBh 1.1.1, VIII.9. 194 commentarial corpus this became known as prat kop sana, symbolic meditation, the symbolic resemblance being the important factor even if an argument could be made for some real ontological relation.31 For Bādarāya a, all Upaniṣadic meditations other than the symbolic belonged to a single class, which we may provisionally call meditations on Brahman, brahma- vidy or brahmop sana. The distinction between the two is drawn in two sūtras, 4.1.4 and 4.3.14, the upshot of both being that the symbol-based meditations do not have the attaining of Brahman as their result.32 Bārarāya a calls the symbolic meditations k mya, optional, with a clear allusion to the optional, desire-based rituals that are performed for specific results.33 Now, they are classified further as (1) either performed within a broader ritual, or as (2) performed independently. Bādarāya a’s term for the first was aṅg vabaddha meditations, meditations pertaining to subsidiary elements of a ritual, performed under the rubric of consecration or embellishment and meant to either enhance the result of the ritual or bring some added value.34 The meditation on the sacrificial horse can again be adduced as an example. In the Aśvamedha sacrifice, the sacrificial horse was classified under the category of a subsidiary part, subordinate to the principal element that was the ritual action, and the whole ritual was be performed for a specific result. If the ritual, however, was accompanied by the meditation on the horse as Prajāpati as delineated above, then the horse would be embellished through that meditation and the ritual would bring for the sacrificer attainment of the highest heaven, the world of Prajāpati. The insertion of this meditation was optional, contingent on the desire of the sacrificer for another 31 See, for instance, VK on BS 3.3.58. 32 Cf. also VPS on 3.3.58, who terms these meditations brahma-pr pti-vyatirikta-phala, having results other than the attainment of Brahman. 33 See BS 3.3.58-64 and the commentaries thereon. 34 BS 3.3.53. 195 result, in the manner of the milking vessel or go-dohana which brings added value, a different result, to a ritual where otherwise an ordinary vessel was to be used. This meditation would be, thus, both k mya and aṅg vabaddha. Although these meditations pertained to elements subsidiary to a ritual, Vedāntins univocally refused to accept that the meditations themselves were subsidiary to the ritual with which they were associated. They were taken as units unto themselves and when combined with the respective ritual, Vedāntins made the ritual subordinate to them, insofar as the result which was expected from the complex performance was the result associated with the meditation, not the ritual. Śa kara and Sureśvara even claimed that the same meditation could be performed in a non-ritual context as well, as an option to the combined performance. The meditation on the Aśvamedha horse, for instance, could be performed outside of a ritual and with no horse at all. If one did not have the requisite adhik ra for an Aśvamedha, which was, we should note, a royal sacrifice, our Advaitins claimed that one could perform the same meditation not on the horse but on oneself, on one’s own head as dawn and the rest, and still attain the world of Prajāpati. The second group of k mya meditations can be best defined negatively, through two characteristics: (1) they were strictly Upaniṣadic meditations, that is, meditations which were not tied to ritual subsidiaries and not to be performed in a ritual context; (2) their results were, however, of the variety which ritual was thought to bring, and not the attainment of Brahman. Illustrations for these may beŚ “He who knows thus the wind as the child of the quarters will not mourn the loss of a son.” “He who meditates on Brahman as name obtains freedom of movement as far as name reaches.”35 Such meditations are, in fact, interspersed in the Upaniṣads alongside 35 ChU 3.15.2Ś sa ya etam eva vāyuṁ viśāṁ vatsaṁ veda na putra-rodaṁ roditiś 7.1.5Ś yāvan nāmno gataṁ tatrāsya yahā-kāma-cāro bhavati yo nāma brahmopāste. Śa kara, Śrīnivāsa and Bhāskara uniformly cite both of these as instances, and it seems that the intention is to illustrate both types of counterparts, Brahman or something else. In 196 the meditations for the attainment of Brahman, and from the two instances it is clear that they could be meditations on something either as Brahman or as some other divine principle. In either case the meditation was based on symbolic likeness, prat ka. When Vedāntins talk about prat kop sanas, they generally have these independent Upaniṣadic meditations in mind, and even more restrictedly the meditations in which the symbolic counterpart is Brahman.36 It is clear, however, that the meditations on ritual subsidiaries were also understood as symbolic in nature. We may, thus, classify the k mya meditations that do not end in attaining Brahman as aṅg vabaddhop sana and prat kop sana and bear in mind that in the second the meditational counterpart was either Brahman or something else. We should not fail to note that these prat kop sanas are the same or similar textual loci which in Kumārila’s second account of liberation were associated with the attaining of prosperity (abhyudaya). Brahma-vidy I went into these details of classification of vidy s not to bother the specialist or deter the proverbial general audience from reading on, but to bring home the following point: for Bādarāya a, an Upaniṣadic meditation was (1) either symbolic and optional—related to a sacrificial element or independent—and resulting in an attainment other than Brahman; or (2) a meditation on Brahman proper. Once the first were properly identified and labeled, all the 4.1.4, which is strictly about symbolic meditations in which the counterpart is Brahman, all viṣaya-v kyas in all commentaries are taken from the ChU. 36 See BS 4.1.4 and the commentaries thereon, all along similar linesś VK, p.1194Ś pratīkopāsaneṣu saṁśayaḥ, kiṁ pratīkopāsaneṣv apy ātmatvenānusandhānaṁ kara īyam, ahosvin neti? kim tāvad yuktam? kara īyam eva, brahmopāsanatvāviśeṣād iti prāpte, “In regard to the symbolic meditations, there is the following doubtŚ should the meditation on the object as the Self be done in regard to the symbolic meditations as well or not? – What is the reasonable thing to do? – Surely it is to be done, because these are a species of meditations on Brahman; such is the prima facie view.” 197 remaining Upaniṣadic meditations were classified as meditations on Brahman proper, because they result in attaining Brahman. This was essentially a negative characterization, but in 4.1.3-4 it is combined with a positive one: a brahma-vidy is a meditation on Brahman as one’s Self. “As the Self, because that is what they admit and teachś but, not as a symbol, because the symbol is not the Self.”37 The commentaries, naturally, diverge in understanding the precise ontological relationship that undergirds the identification of Brahman with the Self, but we need not bother with that. It is sufficient to take what Nimbārka has to say on thisŚ “This is my Self” (ChU 3.14.3), thus the ancient admit. “This is your Self” (BĀU 3.4.1), that is how they instruct students. Therefore, the aspirant after liberation should meditate on the Supreme Self as one’s own Self. However, the Self is not to be intended in regard to a symbol, because the symbol is not the Self of the meditator.38 There are, in other words, texts in the Upaniṣads that identify one’s Self with Brahman, and they constitute brahma-vidy ; there are texts that identify something else with Brahman or something else, such as the mind as Brahman or the udg tha as the sun, and they are not brahma-vidy . The negative characterization, however, was more basic, and that was to accommodate one Upaniṣadic vidy which did not fit the Brahman-as-the-Self paradigm. That was the famed pañc gni-vidy or the knowledge of five fires from the ChU 5.3-10 and BĀU 6.2, the two textual loci which introduce the process of rebirth in the Vedic corpus. The pañc gni-vidy was somewhat of an oddball for the BS classification, because it does not relate two distinct things so that it could be a meditation of one thing as another. It was a depiction of saṁs ra, which was by some Vedāntins seen as a meditation on Brahman as an effect that is the world, but it promised the attainment of Brahman to those who know the process of rebirth through the same path 37 ātmeti tūpagachanti, grāhayanti ca. na pratīke, na hi saḥ. BS 4.1.3-4. 38 eṣa me ātmeti pūrve upagacchanti. eṣa te ātmeti ca śiṣyān upadiśanti. ato mumukṣu ā parama-puruṣaḥ svasyātmatvena dhyeyaḥ. pratīke tv ātmānusandhānaṁ na kāryaṁ na sa upāsitur ātmā. VPS 4.1.3-4, p.1190, 4. 198 which was associated with the common brahma-vidy s. Bādarāya a, therefore, emphasized the “not as a symbol” principleŚ if a meditation is not symbolic and it promises the attainment of Brahman, it is a brahma-vidy .39 The attainment of Brahman was, in fact, “the higher instruction,” the constituent in virtue of which vidy was a means of some human goodŚ “From this [vidy ] there follows the attainment of a human good, because there is scriptural evidence to that account—thus Bādarāya a.”40 The commentators have unanimously glossed the “higher instruction” as an instruction about the Supreme Self as opposed to the transmigrating enjoyer and ritual agent that the Mīmāṁsakas proposed as the domain of the Upaniṣads, and they have also unanimously selected the famous brahma-vid pnoti param from the Taittir ya as the topical sentence of the human good referred to in the quoted sūtra. This gives us the occasion to tackle now the question of brahma-vidy in some detail. In terms of scriptural theology, brahma-vidy is the textual ideality of a specific meditation on Brahman, to be reconstructed through combining the meditational details of its various iterations as well as some other elements common to all brahma-vidy s and to be applied optionally to the other brahma-vidy s in an outlined procedure, resulting eventually in the attainment of Brahman. We are already familiar with the combination of details, but let us see how all of it was supposed to work. A representative list of prominent brahma-vidy s and their respective Upaniṣadic loci reconstructed from the BS commentaries would look like this: ▪ ṇḍilya-vidy in ChU 3.14 and ŚPB 10.3, with a few details in BĀU 5.6, the teaching of Śā ḍilya about the innermost Self which is Brahman. 39 See BS 4.3.14-15 and the commentaries thereon. Later Vedāntins reworked the pañc gni-vidy as a meditation on “one’s imperishable nature as having Brahman as its Self” (so Nimbārka, Śrīnivāsa, Rāmānuja), a meditation in which the object is not Brahman but the unchanging Self, but in which Brahman is eventually inserted. 40 puruṣārtho ‘taḥ, śabdād, iti bādarāya aḥ. BS 3.4.1. 199  ▪ Bhūma-vidy in ChU 7, the teaching of Sanat-kumāra to Nārada about Brahman that is ▪ plenitude (bhūman). Sad-vidy in ChU 6, the famous instruction of Uddālaka Āru i to his son Śvetaketu on ▪ how Being (sat) is everything, including the individual Self. Upakosala-vidy in ChU 4.10-15, the teaching of Upakosala Kāmalāyana to Satyakāma Jābāla about the person in the sun and in the eye. ▪ nandamaya-vidy in TU 2, otherwise also known simply as Brahma-vidy , and ▪ discussing what became the essential positive nature of Brahman. Vai v nara-vidy in ChU 5.11-18 and BĀU 5.9, the teaching of the king Aśvapati to six ▪ householder Brahmins about the Self which is common to all. Akṣara-vidy in BĀU 3.8, Yājñavaklya’s teaching to Gārgī about the imperishable ▪ Brahman. Dahara-vidy in ChU 8.1-6, containing the teaching about the small space in the city of ▪ Brahman that is the heart. Madhu-vidy in BĀU 2.5, the teaching of Dadhyañc Ātharva a to the two Aśvins about ▪ the brilliant immortal person within everything. Pañc gni-vidy in ChU 5.3-10 and BĀU 6.2, delineating the process of rebirth. We should note that Bādarāya a does not treat these individually—even their names are culled from the commentarial corpus—but establishes the principles of unity of the separate vidy s, the optionality of the different brahma-vidy s, the possible aggregation of non-brahma-vidy s and different exceptions to these principles.41 Now, it will not escape the attention even of the resident Upaniṣadic expert that this is a bit of a medley of texts and topics. Some work had to be done not only to standardize the individual vidy s, but to normalize them across the board as well, so that they all would be equal meditations that bring one to Brahman. A template brahma-vidy had to be worked out to which they would all conform, yet keeping their individual details in virtue of which one of them could be practiced as per one’s preferences but the result would be the same in all cases. First of all, they would all have to aim at the attaining of Brahman through the so-called deva-y na or the course of the gods (on which more below).42 In fact, it was precisely because of 41 As I mentioned earlier, however, he does name one of the non-brahma-vidy s, puruṣa-vidy , in 3.3.24, so it is inferable that by his time the corpus was already standardized in different vidy s. 42 “The scholars of brahman who depart life by fire, by sunshine, by day, in the bright fortnight, and during the six months after the winter solstice go to brahman.” Translation van Buitenen 1981Ś103. 200 the deva-y na that the pañc gni-vidy , which does not even so much as mention Brahman as a counterpart to anything, made the brahma-vidy cut: it promised those who know the secret of rebirth and meditate in the wilderness to ascend to the world of Brahman through the course of the gods.43 The course, on the other hand, was not mentioned, for instance, in the ṇḍilya-vidy , madhu-vidy , vai v nara-vidy , so there it had to be inserted. It could be inserted because there are direct statements from ruti and smṛti which associate knowing Brahman with ascending to Brahman via the divine path that are taken as generally applicable whenever someone is a knower of Brahman, for instance Bhagavad-G t 8.24. Thus, ascending through the course of the gods becomes a part of all brahma-vidy s. By the principle of reciprocity, knowing Brahman is inserted in the pañc gni-vidy Ś if someone ascends through the deva-y na, surely, he must be a knower of Brahman.44 A second thing to normalize was Brahman itself, and that was necessary to make sure that the object of meditation and the attained result were the same. A single conception of Brahman was to permeate the vidy s, and so the notion of Brahman had to be standardized through inserting Brahman’s “essential characteristics,” culled from a few texts where Brahman is defined. First to be inserted were Brahman’s positive characteristics, which Bādarāya a calls “bliss and the rest.”45 This primarily referred to the well-known characterization of Brahman as Being, knowledge, limitless, Bliss, established on the basis of the Taittir ya (satyaṁ jñ nam anantam brahmaś nando brahma).46 That was, in any case, natural, because the Taittir ya account provided the paradigmatic brahma-vidy (called, in fact, simply brahma-vidy ), since it 43 ChU 5.10.1-2. 44 See on this BS 3.3.31: aniyamaḥ, sarveṣām, avirodhaḥ, śabdānumānābhyām, “No restriction, [the course belongs to] all [meditations]; there is no contradiction, through the evidence of scripture and inference (smṛti).” The commentaries are quite unanimous, again. 45 ānandādayaḥ pradhānasya, “Bliss and the rest of the principal.” BS 3.3.11. The commentaries are again mutually coherent. 46 TU 2.1 and 3.6.6. 201 gave the paradigmatic injunction—brahma-vid pnoti param—that justified all vidy -up sana as a means of human good, the essential definition of Brahman and most of the technical vocabulary (such as vidv n, one who has known Brahman, in effect the old term for what became j van-mukta.) A second set of characteristics of Brahman to be inserted universally in brahma-vidy s were Brahman’s negative characteristics taken explicitly from Yājñavalkya’s teachings to GārgīŚ That, Gārgī, is the imperishable, and Brahmins refer to it like this—it is neither coarse nor fine; it is neither short nor long; it has neither blood nor fat; it is without shadow or darkness; it is without air or space; it is without contact; it has no taste or smell; it is without sight or hearing; it is without speech or mind; it is without energy, breath or mouth; it is beyond measure; it has nothing within it or outside of it; it does not eat anything; and no one eats it.47 Such insertion should prevent mistaking Brahman for any of the finite beings that are its perishable products.48 Both insertions were justified by an appeal to a principle given in the MS, which stipulates that all characteristics essential to a primary element in a ritual follow that primary wherever it may appear.49 With these two additions, the concept of Brahman for the purposes of meditation would be complete, and other characteristics should be kept for the individual vidy s.50 Bādarāya a was, however, particularly alarmed by the absence of those positive characteristics that were prefaced by “true,” such as having true desires and resolves that we saw in the 8th Ch ndogya, in Yājñavalkya’s teaching to Janaka. Yājñavalkya tied the achieving of liberation to giving up all 47 BĀU 3.8.8. Translation Olivelle 1998Ś91. 48 This is based on BS 3.3.33: akṣara-diyāṁ tv avarodhaḥ sāmānya-tad-bhāvābhyām aupasada-vat tad uktam, “Inclusion of the notions of imperishable, because of generality and its being that, like in the case of the Upasad sacrifice, that has been said.” The commentaries are again remarkably on the same line. 49 MS 3.3.9: gu a-mukhya-vyatikrame tad-arthatvān mukhyena veda-saṁyogaḥ, “When the primary and the subsibiary diverge [belong to a different Veda], because it [the subsidiary] is for the purpose of that [the primary], the relation to the Veda is through the primary.” 50 See BS 3.3.34. 202 desires, but Bādarāya a wanted the true desires to be inserted in that meditation on the pretext of its being the same meditation as the one from the Ch ndogya.51 We should note that for now, and I will have more to say about it later. Conspicuously absent from this concept of Brahman is an emphasis on its causal role in relation to the world, which was so prominently placed at the very opening of the BSŚ “Brahman is that from which proceeds the creation, sustenance and dissolution of the world.”52 This absence is a real giveaway of what brahma-vidy was about. Its aim was some attainment through self-assimilation. Through meditating on Brahman as one’s Self, one becomes Brahman in all respect, except for the ability to interfere with the creation of the world. So, Brahman’s agency in creation was not emphasized in the constructed meditational concept not because it was not deemed essential to Brahman’s nature, but because it was useless for the meditational aspiration. Of course, it would have been present in many vidy s that talk about it, but its role would have facilitated the correlation of the meditational counterparts, not Brahman’s role in creation. We will have more to say on this when we talk about the state of liberation. So, once the different vidy s have thus been normalized, whatever is left as characteristics of Brahman in the individual vidy s is peculiar to them, not to be combined further. Thus, given that a brahma-vidy correlates Brahman to the individual Self, its full- fledged formulation would have looked something like this: (P)(R,S)Brahman which is Being, knowledge, bliss, infinite, imperishable and thus different from its products, is my (Q)Self. [BV]53 51 kāmādītaratra tatra cāyatanādibhyaḥ, “(True) desires (should be added) elsewhere, and those there (to be added here), because of (sameness of the) abode.” BS 3.3.38. The commentators are again in agreement, and Śa kara follows suit, before drawing the qualified/supreme Brahman distinction as he typically does when he has a problem with the straightforward meaning. 52 BS 1.1.2. 53 I should like to emphasize that this formulation is not meant to be a logical notation expressing a relation, but a template containing variables. 203 The predicate notations would stand for elements which are peculiar to the specific brahma- vidy . P and Q would express the specific correlation. For instance, in the ṇḍilya-vidy the relation would be between Brahman that is “larger than the earth, larger than the intermediate region, larger than the sky, larger even than all these worlds put together,” and the Self “of mine that lies deep within my heart, smaller than a grain of rice or barley, smaller than a mustard seed, smaller even than a millet grain or a millet kernel.”54 R and S would signify features of Brahman characteristic to the individual vidy , which could be of several types. Some would be considered specific characteristics of Brahman, but restricted to the vidy s where they are mentioned. Again, in the ṇḍilya these would be “having true desires, true resolves, all actions, all smells and tastes” and the like. Other could be accidental properties that should facilitate concentration. For instance, in the brahma-vidy of the TU, Brahman whose essence is bliss is described as having a body “whose head is pleasure, right plank delight, left plank thrill and torso joy.”55 Because Brahman cannot be a compounded entity, these are not real properties, but are meant to facilitate concentration.56 There may be other details to work out in the individual vidy s, but the template would have looked something like that. Because the attainment as their integral part in all of them is the same—Brahman through the deva-y na—only one should be practiced by an individual practitioner: whereas the k mya meditations which bring attainments of the same kind as ritual can be combined as one desires 54 ChU 3.14.3, translation Olivelle 1998:209, slightly modified. 55 TU 2.5. 56 priya-śirastvādy-aprāptir upacayāpacayau hi bhede, “Non-obtainment (in the universal meditational concept of Brahman of qualities) such as ‘having pleasure as its head,’ because addition and subtraction (are possible) in (the context of) duality.” BS 3.3.12. 204 (samuccaya), the more the merrier, one brahma-vidy would bring the same attainment as any other, and therefore they were theorized as options to one another (vikalpa).57 The Practice of Meditation on Brahman Now the question is, how was a brahma-vidy or brahmop sana to be practiced? While Bādarāya a clearly talks only about meditation as the means of attaining Brahman, all Vedāntin commentators have understood brahma-vidy as consisting of three limbs. The topical text which became canonical for this division came from Yājñavalkya’s teachings to MaitreyīŚ tm v are draṣṭavyaḥ rotavyo mantavyo nididhy sitavyaḥ, “The Self, honey, is to be seenŚ it is to be heard about, pondered over and meditated upon.”58 Vedāntins have universally interpreted the first gerundive, draṣṭavyaḥ, as stating the goal, that one should eventually achieve a vision of Brahman, whereas the other three as expressing the procedure: that goal can be accomplished through instruction in scripture, presumably a specific text delineating a vidy ( ravaṇa); reflecting on the meaning of what was heard (manana); and meditation proper on Brahman as the constructed meditational object (nididhy sana).59 Two related points are important in this regard. First, these three were supposed to be practiced sequentially: one hears from scripture first, clarifies the meaning of what was heard, and finally meditates. In terms of soteriological causality, the contribution of each preceding limb is harnessed by the following. Second, the first two are in general not discussed in pre- Śa kara Vedānta at all or very vaguely, from what we know from Ma ḍana and Śa kara’s and Sureśvara’s engagement with opponentsŚ meditation proper was the means. Bādarāya a’s sole 57 vikalpo ‘viśiṣ a-phalatvāt. kāmyās tu yathā-kāmaṁ samuccīyeran na vā pūrva-hetv-abhāvāt, “There is option, because the result is the same. But, the optional-volitional meditations may be combined or not promiscuously, because the previous reason does not obtain.” BS 3.3.57-8. 58 BĀU 2.4.5 and 4.5.6. 59 The second is Ma dana’s preferred term. 205 concern was with meditation, and dar ana as the goal and dhy na as the means were essentially the same thing, a vision of Brahman achieved through practice. Srinivasa Chari’s observation may be profitably quoted on this pointŚ “Three stages are mentioned as preparatory to the vision of Brahman (dar ana). These are ravaṇa or hearing, manana or reflection and nididhy sana or meditation. … According to this teaching, nididhy sana or up san is the direct means to mokṣa, whereas ravaṇa and manana are subsidiary or aṅga to up san .”60 As we will see in the next chapter, Śa kara will have dramatically different ideas in this regard. We have, thus, zeroed in on meditation proper. Bādarāya a had several things to say on the practice of meditation. First, in terms of type of awareness, the meditation on Brahman was a fixed concentration on a notion or an ideaŚ “Because meditation is of the nature of concentration.”61 This “fixed concentration” was a persistent feature of Vedāntic characterizations of meditation. We saw some of the definitions of vidy ṭup sana in the beginning, but it may be worthwhile to revisit a few: it is a repetition of the same thought or notion (BhāskaraŚ sam na-pratyay vṛtti); it is a representational flow, focused mentation (RāmānujaŚ smṛti-santati-rūpamś ek gra-cinta-vṛtti); a continuous flow of a uniform notion/thought of the meditational object (ŚrīnivāsaŚ dhyey k ra-pratyaya-prav ha-rūpasya); uniform stream of thought called contemplation, cultivation, meditation (Ma ḍana: tat- sant navat dhy na-bh vanop san di- abda-v cy ). In terms of content, the meditational thought that one would have mulled over would have been a self-identification with Brahman through a variation of the [BV] proposition that I formulated above.62 As was generally characteristic of meditation in South Asia, meditation on 60 Chari 2002:283. 61 dhyānāc ca, “Because of concentration.” BS 4.1.8. Nimbārka’s glossŚ upāsanasya dhyāna-rūpatvāt. 62 BS 4.1.3, referenced above. 206 Brahman was to be practiced strictly in a sitting posture, but there was no restriction in terms of placeŚ it was to be practiced “wherever concentration is possible.”63 This is a clear giveaway that the paradigmatic meditator on Brahman would have been a householder. Along the same lines, this meditation was supposed to be accompanied by ritual and other religious practices, which included the daily Agnihotra and practices such as charity and austerity for the individual ramas, as well as cultivation of certain virtues for everyone. We saw how Mīmāṁsakas struggled to justify why ritual had to be performed by everyone, and the best they could offer was prevention of bad karma. Bādarāya a, on the other hand, proposed that ritual along with one’s duties could also foster meditation when performed in the pursuit of liberation.64 The individual ramas were supposed to continue performing their individual duties under the provision that they are conducive to liberation,65 which provision was traced to BĀU 4.4.22Ś “It is he that Brahmins seek to know by means of vedic recitation, sacrifice, gift- giving, austerity and fasting.”66 Charity was the duty of the householders and austerity and fasting of the renunciants, while sacrifice consisting of the daily Agnihotra was to be performed by everyone, the only exemption being “one Vedic branch some of whose members never light 63 asīnaḥ sambhavāt, “(Meditation should be practiced while) seated, because it is possible (only in that way).” BS 4.1.7. yatraikāgratā tatrāviśeṣāt, “Wherever concentration (is possible), there, since there is no specification.” 4.1.11. 64 agnihotrādi tu tat-kāryāyaiva tad-darśanāt, “But, Agnihotras and the rest are for that effect, because that is seen.” BS 4.1.16. Cf. VPS thereon, p.1214: vidyayāgnihotra-dāna-tapa-ādīnāṁ svāśrama-karma āṁ niv tti-śa kā nāsti vidyā-poṣakatvād anuṣ heyāny eva yajñādi-śrutau vidyotpādakatva-darśanāt, “There is no question of ceasing the duties of one’s rama, such as Agnihotra, charity, austerity, etc., through meditation. They must be observed because they nurture meditation, because we see in the ritual texts that they give rise to knowledge.” Also, BS 3.4.33: sahakāritvena ca, “And, as being assistants.” 65 sarvāpekṣā ca yajñādi-śruter aśva-vat, “(Meditation) depends on all ( rama duties), as per the text about sacrifice and the rest, in the manner of the horse.” BS 3.4.26. The horse is interpreted differently, but Nimbārka’s is the simplestŚ “as one depends on a horse for going.” 66 Translation Olivelle 1998:125. 207 up the fire.”67 The reference is to Vājasaneyin renunciants who would take up renunciation without ever marrying.68 Along with their rama duties, all who aspired after liberation were expected to cultivate certain virtues which were considered enjoined, in the BĀU immediately following the previous provisionŚ “A man who knows this, therefore, becomes calm, composed, cool, patient and collected.”69 Calm, self-control, tolerance, etc., thus, became mandatory virtues, and along with the rama duties they were subsidiaries to meditation.70 In fact, Bādarāya a and the commentators were typically Mīmāṁsic in turning the tables on MīmāṁsāŚ ritual and the rama duties were, really, primarily for achieving liberation by assisting meditation, but mandatory even for those who do not aspire after liberation, under the different provision that they are to be practiced as long as one lives, just as the kratvartha kh dira wood or yoghurt can become puruṣ rtha through the process which we discussed.71 “You want yoghurt? I’ll give you some yoghurt!” 67 kāma-kāre a caike, “One branch [say that knowers of Brahman give up action] voluntarily. BS 3.4.15. 68 “It was when they knew this that men of old did not desire offspring, reasoningŚ ‘Ours is this self, and it is our world. What then is the use of offspring for us?’ So they gave up the desire for sons, the desire for wealth, and the desire for worlds, and undertook the mendicant life. The desire for sons, after all, is the same as the desire for wealth, and the desire for wealth is the same as the desire for worlds—both are simply desires.” BĀU 4.4.22. Translation Olivelle 1998:125. 69 4.4.23, translation Olivelle 127. 70 śama-damādy-upetaḥ syāt tathāpi tu tad-vidhes tad-a gatayā teṣām avaśyānuṣ eyatvāt, “Still, he should be possessed of calm, self-control, etc., since they are to be practiced mandatorily on the strength of being subsidiary to meditation as per the injunction.” BS 3.4.27. 71 vihitatvāt cāśrama-karmāpi, “Because they are enjoined even as rama-duties.” BS 3.4.32. Cf. VPS on 32-3, p.1141-3: yad vidyā gādi tad amumukṣu ā cāśrama-karmatvenāpy anuṣ heyaṁ y vaj-j vam agnihotraṁ juhoti iti vihitatvāt. vidyā-sahakāritvenāpi vividiṣanti yajñena ity-ādinā yajñāder vihitatvān mumukṣu ām apy anuṣ heyaṁ samyoga-p thaktvenobhayārthatva-sambhavāt. “Sacrifice etc., which are subsidiaries to meditation, should also be performed be the one who does not aspire after liberation, through the injunction ‘He offers Agnihotra as long as he lives.’ Since they are enjoined as subsidiary to meditation in the text ‘Brahmins seek to know it through sacrifice,’ etc., aspirants after liberation should perform them as well, since that is possible by the rule of conjunction and separation [MS 4.3.5].” 208  Finally, meditation on Brahman was supposed to be practiced one’s whole life.72 This last stipulation is immediately relevant to considering the results of meditation, as it answered the question, what should one do when the meditational practice has borne fruit? The question was prompted by the assumption that there comes a point in time during one’s life when the meditation has become perfect, at which stage one becomes a vidv n, a knower of Brahman.73 The commentators do not have much to say about what this achievement was supposed to look likeŚ Brahman becomes manifest (NimbārkaŚ vyajyate; ŚrīnivāsaŚ brahma vyaktaṁ bhavati; RāmānujaŚ asya s kṣ t-k raḥ) in meditation which is of the nature of devotion (NimbārkaŚ bhakti-yoge dhy neś Śa kara: bhakti-dhy na-praṇidh n dy-anuṣṭh namś BhāskaraŚ bhaktiḥ, dhy n din paricary ś RāmānujaŚ samyak-pr ṇane bhakti-rūp panneś ŚrīnivāsaŚ nididhy sana- lakṣaṇe bhakti-yoge). A lexeme that is characteristically used is “steady recollection” (dhruv nusmṛti),74 which implies that once one has experienced Brahman, such awareness had to be maintained till the end of life—one is still in saṁs ra up until reaching brahma-loka—for which purpose the practice of meditation had to continue till one’s final breath, along with Agnihotra and religious duties that nurture it. Meditation, thus, became something like the principal nitya-karma for the aspirants after liberation. Attaining Brahman What happens after the vision of Brahman when one has become vidv n, on the other hand, is depicted in some detail. First of all, one becomes immediately free from the past bad karma 72 āv ttir asak d upadeśāt, “(There should be) repetition (of meditation) more than once, because such is the instruction.” BS 4.1.1. ā prayā āt tatrāpi d ṣ am, “Until death, for it is seen in scriptures (that it is done) even then.” BS 4.1.13. 73 api saṁrādhane pratyakṣānumānabhyām, “And, (Brahman is revealed) in perfect meditation, because of (the evidence) of perception and inference [that is, ruti and smṛti].” BS 3.2.24. 74 See, for instance, VK on 4.1.13. 209 which has not started bearing fruits, whereas the new karma which one would otherwise create does not stick.75 The past good karma is also gone, either immediately or at death, without one having to experience any of it, but one must live through the karma that has already started bearing fruits.76 The text justifying this is the famous passage from the ChUŚ “There is a delay for me here only until I am freedś but then I will arrive.”77 On this point Śa kara gives the potter’s wheel instance of karmic inertia from the SK that we saw in Kumārila’s first account and which famously posited the distinction between liberation while living and final liberation, but that is rejected head-on by Rāmānuja and rightly so.78 It is clear that for Bādarāya a there was no such thing as liberation before death, and even little after that, as we shall see in a bit. In terms of practice, everything was supposed to remain the same as well—one had to continue with meditation, with ritual and with one’s religious practices for life. After all, the topical text on the destruction of karma was ChU 5.24.3Ś “When someone offers the daily fire sacrifice with this knowledge, all the bad things in him are burnt up like a reed stuck into a fire.”79 Bādarāya a will have an additional intervention on this matter, but we will see that later. 75 tad-adhigame uttara-pūrvādyayor aśleṣa-vināśau tad-vyapadeśāt, “On the attaining of that, the anterior and posterior karma is destroyed and does not stick, because there is such a statement.” BS 4.1.13. 76 itarasyāpy evam-asaṁśleṣa pāte tu. anārabdha-kārye eva tu pūrve tad-avadheḥ, “The other kind [good karma] also does not stick, on the fall. Only [that karma] whose effects has not begun, because till that.” BS 4.1.14-15. There is a disagreement what p te tu meansŚ Nimbārka and ŚrīnivāsaŚ good karma does not stick equally as bad, and on the fall of the body one is liberatedś RāmānujaŚ good karma drops on the fall of the body, because it facilitates meditation. 77 ChU 6.14.2, translation Olivelle 1998:257. 78 See the comments on BS 4.1.15. On Rāmānuja, see also Fort 1998Ś77-83. A most informative discussion on the origins of the doctrine of j van-mukti is available in Slaje 2007:127-130. 79 Translation Olivelle 1998:245. 210  In fact, the concern was more how to keep in some way the reality of the good and bad karma which one is freed from at the attainment of Brahman: such karma is not really destroyed as previously claimed, but redistributed to one’s friends and enemies respectively.80 Finally, when death comes, one’s cognitive functions, life-breath and the subtle elements forming the subtle body progressively withdraw and gather around the Self, which at that point enters the heart and can exit through any of the channels that we saw in the 8th of ChU in Kumārila’s second account.81 He who performed solely ritual throughout life takes any of the lower channels and gradually attains the world of the forefathers, pitṛ-loka, through the course of the forefathers known also as the southern course (pitṛ-y na, dakṣiṇ yana), and eventually returns to earth when the good karma has been exhausted. For the vidv n, on the other hand, the top channel lightens up, the one forming a continuum with the sun-rays, at which point begins his ascension through the course of the gods, known also as the northern or upward course (deva- y na, uttar yana).82 He never returns. We don’t need to investigate the history of the idea of the two courses.83 However, since they are mentioned in several Upaniṣadic passages, for Bādarāya a it was important to standardize the deva-y na because it formed an integral part of the paradigmatic brahma-vidy 80 ato ’nyāpi hy ekeṣām ubhayoḥ, “For (there are is also karma) other than this (obligatory) of both (good and bad kind), (according to the text) of one (Vedic branch).” BS 4.1.17. The branch is not identified and the text which all commentators except Nimbārka quote is tasya putrā dāyam upayanti suh daḥ sādhu-k tyāṁ dviṣantaḥ pāpa-k tyām, which assigns the karma to one's progeny. NimbārkaŚ suh daḥ sādhu-k tyāṁ dviṣantaḥ pāpa-k tyām. I haven't found the reference for either, but the most famous instance of this theme is Kauṣ taki Upaniṣad 1.4: tat suk ta-duṣk te dhunute. tasya priyā jñātayaḥ suk tam upayanti. apriyā duṣk tam. “Then he shakes of his merits and demerits. His dear relatives obtain his merits, those who are not dear his demerits.” Translation Bodewitz 2002Ś16. Bodewitz gives few additional references (n.38). The theme of transfer of karma is, otherwise, a common feature of Hinduism, on which see Doniger O’Flaherty 1980a:10-13. 81 BS 4.2.1-7. 82 tac-cheṣa-gaty-anusm ti-yogāc ca hārdānug hītaḥ śatādhikyā, “Through the application of remembering the course that is a subordinate element to that [the brahma-vidy ] (and the knower) favored by the one dwelling in the heart (departs) through the hundred and first (channel).” BS 4.2.16. 83 The best historical account known to me is that of Karmarkar 1925. 211 as the course through which all knowers of Brahman achieve Brahman, lest someone conclude that there are multiple courses corresponding to different attainments.84 The course of the gods delineates the progress of the knower of Brahman from entering the channel that goes from the top of the head and all the way to brahma-loka, through a medley of intermediate stages that are of a very heterogeneous character, such as “flame,” “the waxing fortnight of the moon,” “lightning,” various divinities, the sun and the moon etc. Karmarkar suggests that originally the deva-y na referred strictly to multiple paths through which the gods were considered to travel to earth to attend sacrifices and then back to heaven. The paths of the gods were described in superlative attributes suggestive of light and increase of power (light, day, summer, the waxing moon, etc.) and leading to different divinities, but then the description assumed a literal sense in the Upaniṣads.85 For Bādarāya a they became guiding agents of some kind, a specification which was, according to the commentators, supposed to preclude the possibility that they be interpreted as road signs or rest areas where one could refresh, gas up or have a little fun.86 The manner of standardizing the course of ascending to the world of Brahman is, according to the commentators, through “combining the details mentioned in one place with all the rest,” in the same way in which “the details of meditations referring to the same object in different places are combined in one.” This can be done because “the course to the world of 84 Important loci in which the path is discussed, or which are otherwise relevant include: ChU 4.15-5; 5.10.1-2; ChU 8.5-6ś BĀU 5.10ś 6.2.15ś 4.3-4; KṣU 1.3ś MU 1.2.11. The path is also mentioned in the ŚB 2.1.3.1-3 and BhG 8.24- 5. Cf. Rāmānuja on 4.3.1, p.744, expressing the prima facie view: aniyama iti. kutaḥ? aneka-rūpatvān nairapekṣyatvāc ca, “There is no standardization, because they are different in form and independent on one another.” 85 Karmarkar 1925:461. 86 ātivāhikās tal-li gāt, “They are conductors, because there is such an indication.” BS 4.3.4. Cf. VK p.1278Ś tatra mārga-cihna-bhūtāḥ bhavantu, v kṣa-parvatādi-vat … bhoga-bhūmayo vā bhavantu, “They could be road signs, like trees or mountains, … or places of pleasure.” Likewise, Rāmānuja, Bhāskara and Śa kara, who is characteristically most elaborate. 212 Brahman is the same, and different texts simply refer to it by mentioning a few of its characteristics.”87 And, when all the details are worked out, the course should look like this. Once the knower of Brahman hits the top channel, he mounts the sun rays, which constitute the highway all the way to the sun. It does not matter if he dies by day or night, in summer or winter, because sunrays are there at night as evidenced by the fact that it is hot, while in winter they are just “overpowered by frost” but still there.88 The first conductor that takes the deva-y na itinerant is flame or light (arciḥ), which in the original accounts stood simply for the cremation fire. Flame hands him over to a series of conductors identified by temporal namesŚ “day,” “the waxing fortnight,” “the six months when the sun travels north” and “the year.” From the “year” he reaches the world of the gods, specifically of Vāyu the god of air, and from then on to the sun, moon, and lightning. From that point on, “a non-human person” takes him and leads him all the way to the world of Brahman,89 but successively assisted by Varu a, Indra and Prajāpati. When he reaches the world of Brahman, that is the end of saṁs ra.90 Now the question presents itself, what is this world of Brahman and which Brahman does the knower of Brahman attain? We saw in Kumārila’s second account how the ChU described 87 BS 4.3.1ś VK, p.1268Ś tasmāt sarvāsu śrutiṣu sarveṣām anyatroktānāṁ parva ām anyatropasaṁhare a sarva- viśeṣa a-viśiṣ o ‘rcir-ādi-mārga eka eva pratipadyata iti siddham. ŚBh, p.744Ś vidyā-gu opasaṁhāra-vad anyatroktānām anyatropasaṁhāraḥ kriyate. BSBh, III.822: ekaiva tv eṣā s tiḥ aneka-viśeṣa ā brahma-loka-prapadanī kvacit kenacit viśeṣa enopalakṣiteti vadāmaḥ. 88 raśmy-anusārī. niśi neti cen na sambandhasya yāvad-deha-bhāvitvād darśayati ca. ataś cāyane ‘pi dakṣi e, “Following the sunrays. If it be said, ‘but not at night,’ then no, because the relation (of the knower with his karma) lasts only as long as the body, as scripture shows. Therefore, also during the southern course of the sun.” BS 4.2.17- 19. VK, p.1256Ś rātrāv api raśmayo dehe auṣ opalambhān niścīyante, hemante tu tuṣāra-nikarābhibhavād anupalabdhiḥ. 89 tat puruṣo ‘mānavaḥ. sa etān brahma gamayati. ChU 5.10.2. 90 tad-apīteḥ saṁsāra-vyapadeśāt, “Because of the designation ‘saṁsāra’ until entering that.” BS 4.2.8. VK, p.1268Ś apītir brahma-bhāvāptiḥ, sārcirādikayā s tyā deśa-viśeṣaṁ gatvā bhavati, tad arvāg deha-sambandha-rūpa- saṁsārasya vyapadeśāt. "Entering means attaining the nature of Brahman, and this takes place when one has attained a particular region through the path beginning with light. Prior to that, the soul is subject to transmigratory existence," Likewise, Rāmānuja. 213 brahma-loka: it was a place of heavenly delights. It was the Kauṣ taki Upaniṣad, however, commonly referred to on this point by the commentators, which gave the most graphic description, and we may summarize what it had to say with profit.91 Once the knower of Brahman passes on from the world of Prajāpati towards brahma-loka, five hundred celestial nymphs dispatched by Brahman, a few of which are individually named, greet him with garlands, lotions, cosmetic powders, clothes, and fruits. His first stop is at a lake by the name of Āra, which he must cross with his mind, and if his knowledge is imperfect he drowns there. A watchman greets him next and he comes to a river by the name of Vijarā, which he also must cross with his mind: should he succeed, this is the exact point at which his saṁs ra ends. He then arrives at the tree Ilya, and the fragrance of brahman permeates him. Then he arrives at the plaza Sālajya, and the flavor of brahman permeates him. Then he arrives at the palace Aparājita, and the radiance of brahman permeates him. Then he arrives near the doorkeepers, Indra and Prajāpati, and they flee from him. Then he arrives at the hall Vibhu, and the glory of brahman permeates him.92 After some more heavenly adventures, he finally meets Brahman who sits on a throne, and presents himself before him. On Brahman’s questionŚ “Why are you?93” he repliesŚ “You are the Self of all beings, and I am who you are.”94 After some more chitchat, Brahman finally tells him: “You’ve truly attained my world, Mr. X, it is yours.”95 Now, already by the time of Bādarāya a such descriptions have not been agreeable to all Vedāntins, and so there appears the question, to which Brahman does the liberated Self go to via the path of the gods. A certain Bādari is reported to have advanced the view that the vidv n is lead to that Brahman which is the effect, k rya-brahma. This in later Vedānta became 91 This is found in the first book of the Upaniṣad, specifically chapters 3 through 7. 92 KṣU 1.5, translation Olivelle 1998:329-31. 93 ko sīti. KṣU 1.5. It is reassuring to learn that Brahman, when you meet him on his throne at liberation, asks questions in Serbian. 94 bhūtasya bhūtasya tvam ātmāsi. yas tvam asi so ‘ham asmi. KṣU 1.6. 95 tam āha āpo vai khalu me loko ‘yaṁ te ‘sāv iti. KṣU 1.7. 214 synonymous with the saguṇa Brahman or Hira yagarbha/Prajāpati, corresponding to Brahmā the demiurge of the universe of the Purā ic tradition, and at the same time the embodiment of the universe identified as saṁs ra and the highest deity of Vedic ritualism. I associate this development with the tradition of the Bṛhad- raṇyaka (on which more later). Since Brahman is really omnipresent, Bādari claimed that actions such as motion and attainment are not possible in relation to it, and therefore the ultimate endpoint of the path of the gods is Prajāpati, who can be named Brahman on the account of being the firstborn and closest to Brahman.96 The dominant Ch ndogya account involving going to brahma-loka was, however, accommodated by the invention of the doctrine of gradual liberation that was so influential in later Advaita and in forms of BhedābhedaŚ at the end of the universe, Hira yagarbha himself is liberated and along with him the deva-y na itinerant attains the supreme Brahman.97 This view was opposed by Jaimini the Ch ndogya master, who claimed that the supreme Brahman is attained by those who meditate on it, because that is the primary meaning of the word brahma.98 Bādarāya a disagreed with both, claiming that both accounts have problems. It was wrong to start from a supreme vs. effected Brahman distinction and then decide who goes where or at all: the significant distinction was between symbolic and non-symbolic meditations. If one’s meditation was not based on a symbolic representation, one would attain Brahman even if such meditation was on the effected Brahman or saṁs ra, as in the case of those who meditate by means of the pañc gni-vidy . On the other hand, claiming that going to the supreme Brahman makes no sense jeopardized those texts that do talk about attaining the supreme or the highest 96 kāryaṁ bādarir asya gaty-upapatteḥ. sāmīpyāt tu tad-vyadeśaḥ, “To (Brahman which is) the effect, says Bādari, because going makes sense in regard to him. Because of proximity he bears that name (Brahman).” BS 4.3.7, 8. 97 kāryātyaye tad-adhyakṣe a sahātaḥ param abhidhānāt. BS 4.3.9. 98 paraṁ jaiminir mukhyatvāt, “The supreme (Brahman, says) Jaimini, because of (its) being the primary meaning.” BS 4.3.11. 215 light.99 Bādarāya a was, in other words, uncompromisingly theological in his approach: look at the texts, gentlemen, we are in the sphere of the Veda!100 The Self in Liberation Liberation, thus, meant becoming Brahman in some sense, but there remained the task of specifying what that precisely meant, and of answering the questions that were peculiar to Vedic theology. Did liberation involve attaining a novel state of affairs desirable to men, as in Mīmāṁsā? If so, how could such a state be eternal, and what was the role of desire in its pursuit? What was the experience of liberation precisely like? Now, we saw that the principal element in a Vedic ritual was the action of sacrificing or offering, and the good which was desirable to man was ultimately the result of that principal element. The goal of the Mīmāṁsakas in advancing such a theory was to remove any contingency that could have occurred because of a personal whim. No human or divine factor ought to have a say whether one will get a result or not: if the action was done properly and all the contingencies were accounted for, the result had to follow just as in any mundane enterprise 99 apratīkālambanān nayatīti bādarāya a ubhayathā doṣāt, “He [the non-human person] leads those whose meditation does not depend on symbols, thus Bādarāya a. Because both views are faulty.” BS 4.3.14. 100 Śa kara wanted Bādari’s view to be the siddh nta, but that is a clearly not the case. Cf. Nakamura 1983:388: “Śa kara holds that the theory of Bādari in the Brahma-sūtra IV.3.7-11 is the established theory (siddh nta). Of course, this is not correct as an interpretation of the Sūtra, but the fact that Śa kara would go so far as to attempt an impossible interpretation and conclude that Bādari’s view was the final one, should perhaps be adequately noted.” A most useful and learned discussion on the three views is available in Ghate 1981:145-49, who thinks that sūtras 4.3.15-6, which present Bādarāya a’s view, form a separate adhikaraṇa that concerns the question “who is lead to Brahman,” whereas sūtras 4.3.7-14, presenting Bādari’s and Jaimini’s view, ask the question about the destination, in which case Jaimini’s view is the siddh nta. I don’t find this correction necessary, but it really amounts to semantics rather than any substantial distinction. It is clear that Bādarāya a wants the destination to be the highest Brahman, but his concern is to affirm that it is not the case that only those who meditate on such Brahman attain to it, because such a view would jeopardize the pañc gni-vidy which promises the same attainment as the regular brahma-vidy s. 216 such as agriculture.101 This brought with itself the problem of impermanency: the ritual action is an action, and the results produced by action are not permanent. As the Ch ndogya put it, “as here in this world the possession of a territory won by action comes to an end, so in the hereafter a world won by merit comes to an end.”102 Ritual was also problematic in the specifically Indian understanding of permanence: it proceeded in the manner of combining elements to produce a final result, and things that are got by compounding can be broken apart. The Muṇḍaka’s famous diatribe against ritual was along these linesŚ “Examining the world piled up through ritual, a Brahmin should become cognitively disengaged with it: what is not made cannot be got through what is made.”103 We also saw that vidy as a unit was constructed in the image of ritual, with an injunction, details of procedure, etc. Bādarāya a, however, thoroughly reevaluated the category of pradh na and replaced against Jaimini the principal factor in any Vedic undertaking, ritualistic or meditational, with Brahman.104 Brahman became the court of final jurisdiction at which the enterprise is judged. The reasoning behind this was simple: Brahman is the repository of all desires which one could possibly obtain through the performance of ritual, and all its intentions come to pass by necessity: satya-k ma and satya-saṅkalpa. Brahman is eternal, and it is also one’s Self, to be realized through meditation. If one could become Brahman, one would obtain both all desires and the requisite permanence at the same time. The bedrock of the idea of brahma-vidy , thus, was that Brahman is the Self—it is the higher Self in virtue of which Upaniṣadic meditations are instrumental to the bringing of a 101 Cf. VPS on BS 3.2.40, p.914.: dharmaṁ phala-hetuṁ jaiminir manyate, k ṣyādi-vat tasyaiva tad- dhetutvopapatteḥ; “Dharma is the cause of the result, thinks Jaimini, because only that makes sense, as in agriculture and the like.” 102 8.1.6, translation Olivelle 1998:275. 103 parīkṣya lokān karma-citān brāhma o nirvedyam āyān nāsty ak taḥ k tena. MU 1.2.12. 104 phalam ata upapatteḥ, “The result comes from it [Brahman] because that makes sense.” BS 3.2.38. 217 human good. The core of every brahma-vidy was that this Brahman should be meditated upon, and the idea was consistently applied throughout the Upaniṣadic corpus. It was also applied over the passage on which Kumārila locked horns with the Sā khyas, namely Yājñavalkya’s teaching to MaitreyīŚ “The Self, honey, is to be seenŚ it is to be heard about, pondered over and meditated upon.” Sā khyas wanted this passage to enjoin the pursuit of liberation through discriminative knowledge of the Self as different from the body, a pursuit predicated on dispassion and aiming at cognitive isolation. Kumārila acknowledged the pursuit as predicated on the absence of desire, and precisely for that reason denied that it could be enjoined. Vedāntins, however, tied the passage to Brahman the higher Self,105 and made good on Yājñavalkya’s claim that it is because of this Self that all objects of endearment are desirable.106 The pursuit of liberation became the pinnacle of all aspirations, predicated on the desire for Brahman as one’s Self, tma-k ma. This became the standard lexeme justifying the pursuit of liberation as being within the scope of regular Vedic theology. The issue of permanency also meant that the final attainment could not be quite a new or an adventitious state of affairs as Mīmāṁsakas wanted. Rāmānuja expressed beautifully what Mīmāṁsakas would have wanted the highest attainment to look like if it were to satisfy the criterion of being puruṣ rthaŚ “At the stage of attainment, it only makes sense that one becomes related to a producible personal character, for otherwise the scriptures of liberation would be concerned with something which is not a human good.”107 For Vedāntins this was a no-go for the reasons we have just stated: whatever is adventitious, gantuka, is bound to be lost. The saving grace was found in the idea of manifestation or virbh va of an essential personal character, but 105 See BS 1.4.19-22 and the commentaries thereon. 106 BĀU 2.4.5 and 4.5.6. 107 sādhyena rūpe a sambandha iti yuktam; anyathā hy apuruṣārthāvabodhitvaṁ mokṣa-śāstrasya syāt. ŚBh, p.758. 218 one that is presently not experiencedŚ “At the stage of attainment, there is a manifestation of the Self in its personal character, because the text says so.”108 The text was, of course, the famous Ch ndogya passageŚ “This deeply serene one, after he rises up from the body and reaches the highest light, immerges in his own true appearance.”109 Thus the attainment was neither novel nor quite not novel. It was becoming what one could essentially become, when liberated from all that was adventitious to one’s real natureŚ a sculpture carved out from the same omnipresent slab, not one constructed through addition.110 However, this was not the procedure of separation in the Sā khya manner, in which the Self eventually remains in isolation from matter, but a literal modeling, becoming a replica of the ideal model that is Brahman. Final liberation meant achieving “the highest similarity,” paramaṁ s myam, to the meditational model that was Brahman: When the seer [the meditator] sees that golden-colored Self, the creator and the Lord, the origin that is Brahman, then he, being a vidv n, rid of all merit and demerit and spotless, attains the highest similarity.111 Bādarāya a says that the Self at the point of liberation experiences itself as “not divided” from Brahman, which, as evidenced by the descriptions of what the state consists in, is clearly not a statement of absolute identity.112 One becomes of the same kind as Brahman, equally “awesome.” Two key ideas describe this state: independence and pleasure. The two were directly based on the 7th and 8th of the Ch ndogya, respectively. One becomes independent, without a master, a sovereign to oneself, which gets to mean that one can travel in all the heavenly spheres 108 sampadyāvirbhāvaḥ svena śabdāt. BS 4.4.1. 109 8.12.3. 110 BS 4.4.2. 111 MU 3.1.3. There is an unambiguous reference to this text in BS 4.4.21. 112 avibhāgena d ṣ atvāt. BS 4.4.4. 219 and enjoy all the desires positively affirmed in the 8th chapter of Ch ndogya—the worlds of one’s forefathers, perfumes and garlands, women, and chariots—by one’s mere will.113 This does not constitute a compromise to his liberation, because it happens in bodies—multiple bodies at the same time—which one creates by mere will and then pervades by one’s awareness, as a lamp pervades space which is contiguous with it.114 Liberation, thus, becomes ai varyaṭsv r jya in which one can do everything that Brahman can. Well, almost everything: one cannot interfere in the functioning of the world, in its creation, sustenance, and destruction, which remains the sole province of Brahman.115 “(The independence of the liberated) is limited to enjoyment, because such is the indication got from the ‘similarity.’”116 The topical text of this sūtra is TU 2.1, the paradigmatic brahma-vidy which was also the paradigmatic statement of ‘attaining the highest’ that made up sana useful: He who knows Brahman attains the highest. On this there is the following verse: he who knows Brahman as Being, knowledge, the infinite, hidden in the deepest cavity and in the highest heaven, attains all desires along with the wise Brahman. The knower of Brahman becomes as similar to Brahman as it is possible, and that seems good enough because all the good stuff is there. One of the final sūtras, a particularly obscure one, is interpreted by Nimbārka, Śrīnivāsa and Rāmānuja as positing a second and more essential characteristic of liberation, bliss or 113 sa kalpād eva tac-chruteḥ. ata evānanyādhipatiḥ, “(The liberated attains his forefathers etc.) simply through his resolve, because the text says so. Therefore, no one is his master.” BS 4.4.8-9. Cf. ChU 7.2.2, sa svar ḍ bhavati. 114 BS 4.4.10-15. Bādarāya a is trying to solve a theological dispute here between Jaimini, who says that there are bodies in liberation, and Auḍulomi, who says that there aren’t, by proposing that both views are possibleŚ one may be embodied and enjoy the aforementioned delights or be unembodied and still enjoy, like in the state of dreaming where experience happens in bodies that are created by the Lord. It is possible in the same way as the dvada ha ritual can both be a sattra, a ritual session not performed for anyone but for one's own prosperity, and an ah na, a ritual which a priest performs for another, for offspring, as one’s resolve may be. So much for the claim that the two Mīmāṁsas “could not have originally been one.” 115 jagad-vyāpāra-varjaṁ prakara ād asannihitatvāc ca, “(Such independence) does not include (interfering in) the functioning of the world, because the texts that are about that include no reference to the liberated Self.” BS 4.4.17. 116 bhoga-mātra-sāmya-li gāc ca. BS 4.4.21. 220  nanda, which consists in intuiting Brahman in its essential features. Nimbārka is concise enough to be quoted in fullŚ “The liberated Self intuits Brahman which is devoid of transformations such as birth, which is the totality of inherent, inconceivable, endless qualities, possessing all plenty.”117 It is hard to adjudicate whether Bādarāya a did mean this addition or not, but we should bear in mind that the most essential positive characteristic of Brahman to be inserted in all brahma-vidy s was precisely bliss, and Nimbārka’s formulation virtually mirrors the paradigmatic meditation, only that now the notion has become an intuition. The topical text quoted on this sūtra is right from the bliss-section of the Taittir yaŚ “He [Brahman] is flavor (rasa), and obtaining flavor one becomes blissful.”118 This would have also accounted for the Kauṣ taki description of brahma-loka where mister X is permeated by the fragrance and flavor, radiance, and glory of Brahman. Be that as it may, enjoyment was already posited as the feature of liberation through the attainment of all desires, and these two, ai varyaṭsv r jya and nanda, whatever the extent of the second was, became the two determinants of liberation. Finally, liberation undoubtedly involved going to a specific place, much like the experience of heaven in Mīmāṁsā. It could not be enjoyed right here right now. Brahma-S tra, Liberation and the Two Great Upani ads It has since long been recognized that the Brahma-Sūtra looks very much like a systematization of the Ch ndogya Upaniṣad. Paul Deussen was probably the first to notice that in the first adhy ya of the text, twelve of the twenty-eight topical passages were from the Ch ndogya, while no other Upaniṣad supplied more than four. Further, the passages from each Upaniṣad were discussed in the order as they appear in their texts, which prompted Deussen to suggest that 117 On BS 4.4.19, p.1354: janmādi-vikāra-śūnyaṁ svābhāvikācintyānanta-gu a-sāmagraṁ savibhūtikaṁ brahmaiva mukto ’nubhavati. 118 raso vai sa rasaṁ hy evāyaṁ labdhvānandī bhavati. TU 2.7. 221 Bādarāya a or a follower of his inserted sixteen passages from other Upaniṣads into an earlier work that systematized the Ch ndogya, keeping the principle that the original order of the extracts should be maintained.119 If we rely on commentarial concord, we would notice that the dominance of the Ch ndogya is even more striking in the other three adhy yas. For instance, the whole first p da of the third adhy ya is based on the doctrine of five fires as discussed in the fifth prap ṭhaka of the Ch ndogya. The third p da of the same adhy ya, further, deals with five sections of the Ch ndogya, again in the order in which they appear in the Upaniṣad.120 S.K. Belvalkar went farthest in proposing that there could have been a Ch ndogya- Brahma-Sūtra and a Bṛhad- raṇyaka-Brahma-Sūtra etc. The Brahma-Sūtra that became the normative was the Ch ndogya one, written by Jaimini, in which Bādarāya a or his students introduced passages from the other Upaniṣads as side illustrations. The principal goal in this was to secure the harmony within the Ch ndogya, but use materials from the other Upaniṣads as supporting evidence.121 As I said in the introduction, my purpose here is not to investigate such issues. I will take it as proven that the Brahma-Sūtra text known to us, the text that was commented upon by Vedāntins, was a systematization of the Upaniṣads based on the Ch ndogya. That much, I believe, cannot be doubted. I do want to address, however, another question, one which concerns the place of the other great Upaniṣad in this systematization, the Bṛhad- raṇyaka. I want to suggest that the BS does not attempt to illustrate the Ch ndogya doctrine with a few “side points.” Rather, another phenomenon comes in perspective in we read closely, and that is an attempt to create a pan-Upaniṣadic doctrine by way of rectification of other views. This 119 Deussen 1908:27-9; 1912:120-22. 120 Faddegon 1923. 121 Belvalkar 1918; 1927; 1929. 222 rectification primarily happens to what must be taken as a competing doctrine of liberation, one derived from Yājñavalkya’s teachings to king Janaka in the fourth book of the Bṛhad- raṇyaka. I will finish the section with this rectification. We cannot present Yājñavalkya’s account in detail, but at least its outline must be given. The account is presented in the part of the Upaniṣad where Yājñavalkya taches Janaka about the Self in its three states: waking, dream and deep sleep.122 The first state is identified with the visual sphere, while the second with the heart. The Self travels back and forth between the two through the same channels (n ḍ ) that we know about from the 8th chapter of Ch ndogya, and in the dream state it creates its own experiences “through its own radiance,” by rearranging impressions from the waking state. The Self is not really related to either, because the things from waking and dream stay where they belong—the dream chariots remain in dreams, he does not come back on them—and that is a proof that nothing really sticks to the Self. The third state of the Self, which is commonly known as suṣupti but which is here called sampras da, perfect calm, is one in which the Self comes in order to find real rest, “like an eagle, after flying around in the sky and getting tired, folds its wings and swoops down the nest.”123 In this state the Self sees no dreams, oblivious to anything “like a man embraced by a woman,”124 and it is freed from all the good and bad that it may experience in the two other states. Most significantly, because of the perfect calm and lack of transitive awareness, this is a state where “all desires are fulfilled, where the Self is the only desire, and which is free from desires.”125 122 Chapters 3 and 4 of Book 4. 123 BĀU 4.3.19, translation Olivelle 1998Ś115. 124 BĀU 4.3.21. 125 Ibid., tad vā asyaitad āpta-kāmam ātma-kāmam akāmaṁ rūpaṁ śokāntaram. 223  This is also a state of non-duality, in which the Self could cognize, being naturally the cognitive agent, but does not cognize anything because there is no second thing which could become an object to awareness: When there is some other thing, then one can see another, smell another, taste another, hear another, speak to another, think of another, touch another, distinctly know another. He becomes the sole ocean, the sole seer. This is the world of Brahman, … the highest goal, the highest attainment, the highest bliss.126 In other words, the Self in the state of non-duality characteristic of deep sleep is identified with brahma-loka, the highest attainment. The conversation naturally turns to liberation as the way to attain the state of brahma- loka, which Yājñavalkya contrasts to the process of rebirth that involves going to another place and back, rehearsing what we already know from the BS. When a man is on his deathbed, his cognitive functions are gradually lost as he has entered the heart. A channel lights up for him and the Self departs enveloped by the subtle body. What he turns out to be depends on action and behavior in this life. “A man resolves in accordance with his desire, acts in accordance with his resolve, and turns out to be in accordance with his action. And so people sayŚ ‘A person here consists simply of desire.’”127 What Yājñavalkya means by action here is still probably mostly ritual, as convincingly argued by Tull.128 Rebirth is ultimately related to desire, because the subtle body leads one to the next destination shaped by the actions performed through desire. Once the results accrued through action have been exhausted, one returns to Earth to do some 126 yatra vā anyad iva syāt tatrānyo 'nyat paśyed anyo 'nyaj jighred anyo 'nyad rasayed anyo 'nyad vaded anyo 'nyac ch uyād anyo 'nyan manvītānyo 'nyat sp śed anyo 'nyad vijānīyāt. salila eko draṣ ādvaito bhavati. eṣa brahma-lokaḥ samrā . … eṣāsya paramā gatiḥ. eṣāsya paramā saṁpat. eṣo 'sya paramo lokaḥ. eṣo 'sya parama ānandaḥ. BĀU 4.3.31-2. 127 BĀU 4.4.5. 128 Tull 1989. 224 more ritual; this is, so far, all well-known. “That is the course of the man who desires.” But then comes the twist: Now, a man who does not desire—who is without desire, who is freed from desire, whose desires are fulfilled, whose only desire is his self (ak ma, niṣk ma, pta-k ma, tma- k ma)—his vital functions [standing for the subtle body in the BS systematization] do not depart. Brahman he is and Brahman he attains. On this there is the following verse: When one has got rid of all desires that have taken shelter in the heart, a mortal becomes an immortal and attains Brahman right here.129 The Upaniṣad then quotes several verses which begin talking about the path by which the knowers of Brahman go to the heavenly world, svarga-loka, but predictably we hear none of the description that we are familiar with from the BS. The heavenly world turns out to be the Self itself, as already hinted, and if one knows this Self, one not only wins that world, but is the world. And crucially, this world is found, and immortality won while one is still here. The final attainment is nothing but getting into a state where one does not see any diversity. It is attaining sampras da or the perfect calm of deep sleep and non-transitive awareness while still living: One should see with the mind alone that there is no diversity whatsoever. From death to death goes he who sees any diversity here.130 The Self can be described only though negative attributes—this is the famous neti neti section— and its pursuit requires leading a celibate life and the cultivation of certain virtues. We have already seen these passages, but it is worthwhile repeating them: It was when they knew this that men of old did not desire offspring, reasoningŚ ‘Ours is this self, and it is our world. What then is the use of offspring for us?’ So they gave up the desire for sons, the desire for wealth, and the desire for worlds, and undertook the mendicant life. … A man who knows this, therefore, becomes calm, composed, cool, patient and collected.131 129 BĀU 4.4.7, translation Olivelle 1998Ś121, with some modification. 130 manasaivānudraṣ avyaṁ neha nānāsti kiñcana. m tyoḥ sa m tyuṁ āpnoti ya iha nāneva paśyati. BĀU 4.4.19. 131 BĀU 4.4.22-3, translation Olivelle 1998:125-7. 225 The way to attain liberation, thus, is to give up the desires of the Vedic variety, practice celibacy and cultivate virtues. Then one “becomes a Brahmin—free from evil, free from stain, free from doubt. He is the world of brahman, Your Majesty, and I have taken you to him.”132 In several ways, this account is in a stark opposition to the 8th chapter of Ch ndogya that formed the basis of the BS systematization. The ChU affirmed both desire and ritual: desires were good—ancestors, perfumes, garlands, chariots, women—and ritual was just inefficient in securing them. One had to continue performing ritual daily if intending to become free from the process of rebirth (ChU 5.24.3), but then discover the Self whose desires and intentions are effortlessly fulfilled and accomplished, satya-k ma, satya-saṅkalpa. Yājñavalkya, on the other hand, disparaged desire, and his man was ak ma, niṣk ma, pta-k ma, tma-k ma. He associated ritual strictly with the process of rebirth. Most problematic of all, however, was Yājñavalkya’s characterization of the Self and the promotion of deep sleep into its highest state, one in which there is no transitive awareness of any kind, what to say of enjoyment. This doctrine was attacked heads-on and rejected in no uncertain terms in the 8th chapter of Ch ndogya, where Prajāpati in his teaching to Indra promoted such a “deeply serene Self” as the real deal, before Indra realized how unappealing it is “even before he got back to the gods”: But this Self as you just explained it does not know itself distinctly as ‘I am this,’ nor does it know any of these beings here. It has become completely annihilated. I do not see anything enjoyable in this.133 Prajāpati conceded, and the Self had to reemerge back from the state of deep sleep, its cognitive faculties restored, and reach brahma-loka to enjoy desirable objects and see diversity aplenty. 132 Ibid. 133 ChU 8.11.1. Translation Olivelle 1998:285, with slight modification. 226  Related to this was the course of attaining liberation and its timing. In the ChU account, liberation involved going to another place through the deva-y na and it was attained necessarily after death, whereas Yājñavalkya rejected any kind of departure, a posthumous path or a postponed attainmentŚ “I’ve already taken you to brahma-loka. It is you!” The two accounts talked about the same thing, the Self in the heart in its three states and in liberation, but dissented on all essential points. Not only that, but the Ch ndogya openly attacked the Bṛhad- raṇyaka. Bādarāya a, on the other hand, in his systematization of the Upaniṣads started with the assumption that there is a pan-Upaniṣadic ideological coherence and that when Upaniṣads talk differently about the same thing, they ought to be saying the same thing. In effect, this meant that they ought to be repeating what the Ch ndogya says. Now, in this systematization, the 8th chapter of Ch ndogya and Yājñavālkya’s teaching to Janaka were classified under the same vidy , the dahara-vidy , since they both talk about the space-like Self that is in the heart. This Self is, really, Brahman, but Brahman that is to be meditated upon as one’s Self so that one may become like it. Because the two passages constitute the same vidy , the attributes mentioned in one are to be supplied in the other, and thus the true desires and resolves from the ChU are to be imported into the BĀU. In this way our ak ma, niṣk ma, pta- k ma, tma-k ma ascetic who does not see duality whatsoever and aspires for no progeny, to whom the Self is the world, gets to enjoy all the heavenly delights, forefathers and sons, garlands, perfumes and women.134 The standardized account, in which the sv r jya consisting in the ability to travel to any heavenly sphere in a self-created body and enjoy was the most general 134 kāmādītaratra tatra cāyatanādibhyaḥ, “The characteristics such as desire are to be supplied from the one to the other and the other way around, because the form (of the meditation) is the same.” BS 3.3.39. The commentators are in unison on this sūtra, and so is Śa kara, before claiming that the two Brahmans are different in the two passages, and only the BĀU is the higher oneś the desires in the BĀU are inserted just to show the majesty of Brahman, not for meditation, as this passage is not about meditation at all. Ghate, of course, notes that Śa kara’s interpretation is unjustified (1981Ś130). In Śa kara’s defense, the two passages are genuinely opposed to one another. 227 feature of liberation that all knowers of Brahman attain, made it possible to argue that one would get the heavenly delights even if such was not one’s aspiration. That is just what sovereignty is, you don’t get to be a king and not behave like one.135 Bādarāya a also rejected the idea that Yājñavalkya’s account amounted to achieving liberation while living and not taking the course of the gods. Yājñavalkya’s saying that a mortal becomes immortal when he has got rid of all desires and attains Brahman right here means no more than getting rid of the anterior and posterior karma in the manner in which the ChU described it. It clearly does not mean that one drops dead on the spot, which is a proof enough that one is still embodied, still in saṁs ra, and still with a subtle body, until final liberation is reached in brahma-loka.136 Furthermore, that his vital functions or subtle body do not depart does not mean that he does not go via the deva-y na. For Bādarāya a, a preferable reading of the same BĀU verse was the Mādhyandina recension, which says not tasya pr ṇ ḥ, but tasm t pr ṇ ḥ, amounting to a crucial difference. What the text says is not that the subtle body enveloping the Self of the knower of Brahman does not depart through the course of the gods, 135 See BS 3.3.40, upasthite ’tas tad-vacanāt, “Just on the account of approaching, because the text says so.” VPS, p.1030: ukta-lakṣa ayā brahmopāsanayā brahmopasampanne sarva-lokeṣu kāmacāro bhavati. nanu tat-tal-loka- prāpti-sa kalpa-pūrvakaṁ tat tat sādhanānuṣ hānaṁ vinā sarvatra kāmacāraḥ? tatrocyate ataḥ upasampatter eva hetoḥ paraṁ jyotir upasampadya svena rūpeṇ bhiniṣpadyate [ChU 8.3.4], sa svar ḍ bhavati [ChU 7.2.2] tasya sarveṣu lokeṣu k mac ro bhavati [ChU 7.25.2] iti vacanāt. “When one has attained Brahman through the meditation on Brahman that has the mentioned characteristics, one attains freedom of movement. – But, how could there be such freedom of movement everywhere without performing the requisite practice occasioned by the intention for attaining the respective world? – The attainment itself is the cause, as the texts make it clearŚ “Having attained the highest light, he is accomplished in his own character.” “He becomes independent.” “He attains freedom of motion in all worlds.” 136 samānā cās ty-upakramād am tatvaṁ cānupoṣya. tad-apīteḥ saṁsāra-vyapadeśāt. sūkṣmaṁ pramā ataś ca tathopalabdhaḥ. nopamardenātaḥ; “The same (course for the knower and non-knower) until the ascension to the path. The immortality (in the BĀU) is without having burned (the relation to the body). Because, saṁs ra is denoted until entering that (brahma-loka); and, the subtle body (still persists), because there is such an apprehension from proof. So, (the immortality) is not through destruction.” BS 4.2.7-10. 228 but that it does not depart from the Self: it sticks to the Self all the way until liberation is reached.137 Finally, one of the sūtras according to most of the commentators claimed that even the denial of plurality in Yājñavalkya’s part concerns only matters in which Brahman is not regarded as the Self.138 It would be, however, misleading to conclude that all that Bādarāya a was doing amounted to a normalization of Yājñavalkya’s “odd” teaching. Yājñavalkya’s dialogue with Janaka and Maitreyī contributed many positive elements in the final formulation of the doctrine of brahma-vidy . To begin with, the lifelong celibacy which the text promoted was used by Bādarāya a as an argument for the independence of meditation from ritual. If ruti talks about practitioners of brahma-vidy who do not light up the sacrificial fire—which follows as an inevitable consequence of not begetting sons, as one had to marry to perform ritual—then meditation cannot be a part of ritual, since they practice meditation but do not practice ritual. Second, the list of personal virtues that every knower of Brahman had to cultivate, such as tranquility and self-control, ama and dama, also came from Yājñavalkya. Third, Yājñavalkya’s negative description of the Self became a part of the paradigmatic brahma-vidy , which meant that Bādarāya a wanted every aspirant after liberation of the Upaniṣadic variety to study that text. Fourth, the tma-k ma theme became the justifying factor of the pursuit of liberation as a legitimate Vedic enterprise. Related to that, fifth, Yājñavalkya’s instruction to Maitreyī became 137 pratiṣedhād iti cen na śārirāt spaṣ o hy ekeṣām, “If it be objected, ‘There is a denial (of departure),’ then noś (The departure which is denied is) from the embodied Self, because it is clear in one recension.” BS 4.2.12. Śa kara divides this sūtra in two and manages to arrive exactly at the opposite conclusion, but this is not justified. See Ghate 1981:140-41. 138 See BS 3.3.39 and the commentaries. 229 the basis on which the three Vedāntic processes of liberation were formulated: ravaṇa, manana and nididhy sana. With these considerations in mind, we can still confidently claim that Yājñavalkya’s account of liberation was different from that of the BS, and that the Bṛhad- raṇyaka brought the development of Vedānta into a different direction. Liberation as the final attainment was understood as a state where no ontological difference of any kind obtains. Its influence was very noticeable in the later Upaniṣads, such as the Mu ḍaka, as well as in Gauḍapāda’s well-known doctrine of the four states of the Self. In terms of pre-Śa kara systematic Vedic theology, however, it was the commentary on the Upaniṣad written by Bhart prapañca that presented the normative account, and to him we turn next. 230  CHAPTER FIVE: THE DOCTRINE OF PRASA KHY NA Those who proclaim that liberation from saṁs ra comes through repeated meditation on it seem to be able to do just anything: they have been blessed by Agni Vaiśvānara.1 Introduction We saw in the previous chapter that meditation was the key soteriological practice in the unitary brahma-vidy doctrine of the Brahma-Sūtra. There was, however, another Vedāntic doctrine of meditation, one which directly concerned Śa kara’s favorite Upaniṣadic texts—the negative descriptions of the Self—and occasionally the future mah -v kyas. Except for the Brahma- Siddhi of Ma ḍana Miśra, we know about it only from Śa kara’s BĀUBh and Sureśvara’s V rttika thereon, as well as the Naiṣkarmya-Siddhi. In some cases, this type of meditation was called prasaṅkhy na, and it was not a Vedāntic innovationŚ it was appropriated from the tradition of Yoga. We will see in the case of one of its proponents, Bhart prapañca, just how the two kinds of meditation were different, but for now we note two key points of departure. First, whereas the Brahma-Sūtra doctrine promoted assimilative meditation that aimed at becoming Brahman in kind, prasaṅkhy na was generally reductive and aimed at full identity with Brahman that implied the loss of one’s separate existence. Second, and related to the first, the doctrine of prasaṅkhy na was thoroughly steeped in Yoga psychology and practice: whereas in the Brahma-Sūtra, ignorance that we identified as the root cause of transmigration in the Introduction had no prominent role—generally the BS system is focused on systematizing the Upaniṣadic meditations and takes embodiment for granted—the doctrine of prasaṅkhy na was a part of the therapeutic paradigm worldview. 1 saṁsāra-darśanābhyāsāt tan-muktiṁ ye pracakṣate | nākāryaṁ vidyate teṣāṁ vaiśvānara-varāśrayāt. BĀUBhV 1.4.700. 231  We may put this in the following way: even if ignorance was part of the BS system, it was at best a cover of the Brahman-characteristics that the Self innately possesses, and the BS meditation did not aim at removing such ignorance but at developing the innate characteristics, specifically the feature of experiential bliss consisting in the ability to enjoy all desires and accomplish all resolves.2 In the prasaṅkhy na doctrine, on the other hand, meditation was explicitly geared towards the removal of ignorance, which kept the Self separate from Brahman, by immersing the mind the product of ignorance in Brahman. Furthermore, although the two doctrines shared the same scriptural network, prasaṅkhy na put a recognizably yogic twist to it. Historically, it would appear that many pre-Śa kara Vedāntins were advocates of some form of prasaṅkhy na. A few indications suggest, for instance, that a prominent prasaṅkhy na- v din was Brahmadatta, a pre-Śa kara Vedāntin who is identified by Sureśvara’s commentator Jñānām ta as the character behind the following idea put forward in NaiS 1.67: Some, relying on the strength of their own tradition, sayŚ “The cognition ‘I am Brahman’ that is produced from the Vedāntic statement does not dispel ignorance on its mere appearance. – How then? – It drives away ignorance through the accumulation of meditation for him who meditates day after day for a long time. The evidence for this is the text ‘Becoming a god, he joins the gods.’”3 While Jñānām ta is dated to 1800 CE by Potter and is unlikely to be historically reliable, Ānandagiri in his comments on the section of the Sambandha-V rttika where Sureśvara discusses prasaṅkhy na says that it was Brahmadatta who relied on the doctrine of niyoga and 2 The only place in the BS where some doctrine of ignorance can be teased out is the beginning of the second p da of the third adhy ya, which discusses the creation of objects in the dream state. The commentators are sharply divided on the interpretation, but sūtra 5 and 6 seem to point out that the characteristics of the Self are hidden either through the wish of the Lord or through association with the body. See also Solomon 1969:116-124. 3 kecit svasampradāya-balāvaṣ ambhād āhuḥ, yad etad vedānta-vākyāt ahaṁ brahma iti vijñānaṁ samutpadyate, tan naiva svotpatti-mātre ājñānaṁ nirasyati. kiṁ tarhi? ahany ahani drāghīyasā kālenopāsīnasya sato bhāvanopacayāt niḥśeṣam ajñānam apagacchati, devo bhūtv dev n apyeti [BĀU 4.1.2ff] iti śruteḥ. 232 upheld the notion that knowledge of the unity of the Self required an injunction of meditation.4 This Brahmadatta is provisionally dated to 600-700 AD by Nakamura.5 Another, related doctrine is presented by Sureśvara in continuation of the same passageŚ Others sayŚ “By means of meditation one should bring about another, special cognition of the Self. By this cognition, the Self is known, and only this cognition dispels ignorance, not the cognition of the Self arisen from the Upaniṣads. The following statements have this meaningŚ ‘Having known, one should cultivate insightś’ ‘The Self should be seenŚ it should be heard about, pondered, meditated over,’ ‘One should search out, investigate that Self.’”6 This has commonly been identified as the doctrine of Śa kara’s contemporary Ma ḍana Miśra. Śa kara himself discusses a variant of this idea in BĀUBh 1.4.7, probably in some earlier formulation. It is not clear how much, and if at all, the two views presented by Sureśvara in the quoted passage are different. In fact, it would appear from the V rttika that the doctrine of prasaṅkhy na had an impressive variety of detail in what is oftentimes a fascinating interaction of Yoga, Vedānta, and the two forms of Mīmāṁsā. We will not go into all these details in the chapter, but the key differences concerned the nature of the injunction of meditation on the Self, whether meditation obtained through bh van or niyoga, what was the hierarchical structure of the injunction and the negative descriptions of the Self, including the identity statements of the Upaniṣads, etc. One of the prominent prasaṅkhy na-v dins, Ma ḍana Miśra, argued vehemently against the possibility of injunctions in the domain of the Upaniṣads, including meditation. 4 ĀG on SV 797. On Brahmadatta, read (cautiously) in Hiriyanna 1928; Pandey 1983:237-43; Nakamura 2004:181- 84; Balasubramanian 1988:65-6. Potter’s date for Jñānām ta is given in the online updated Bibliography of Indian Philosophies, http://faculty.washington.edu/kpotter/ckeyt/home.htm, accessed 2.2.2017. 5 Nakamura 2004:181. 6 apare var ayanti upāsanenātma-viṣayaṁ viśiṣ aṁ vijñānāntaraṁ bhāvayet, tenātmā jñāyate, avidyā-nivartakaṁ ca tad eva, nātma-viṣayaṁ veda-vākya-janitaṁ vijñānam iti. etasminn arthe vacanāny api—vijñ ya prajñ ṁ kurv ta [BAU 4.4.21], draṣṭavyaḥ rotavyo mantavyo nididhy sitavyaḥ [2.4.5], so 'nveṣṭavyaḥ sa vijijñ sitavyaḥ [ChU 8.7.1.3] ityādīni. BĀUBh 1.4.7. 233  Such differences notwithstanding, however, all forms of prasaṅkhy na addressed themselves to the same problem—yet another striking difference from the Brahma-Sūtra—which concerned the nature of language in its incarnation in scripture, and its capacity to convey knowledge of Brahman. In general terms, this problem said: knowing through scripture is knowing second-hand; it is a form of knowing in which Brahman is not directly experienced, and therefore scriptural knowledge must be followed by meditation that culminates in a vision of Brahman or the Self. We will unravel the details of this problem in the chapter, but for the time being we may think of it as the distinction between perceptual and book knowledge, say, of the parakeet who has found it convenient to observe the world from the top of my head and that I am immediately aware of as I write, and the parrots in Ma ḍana Miśra’s house that “were heard repeating sentences like 'Is validity intrinsic to knowledge, or extrinsic? Is Karma the giver of fruit, or is it God?’”7 about which I can read with some amusement but without certainty, not the least because of what kind of entities they are. A variety of the problem of language said that scripture presents Brahman as a relational entity, a definite description obtained through attribution of characteristics, which must culminate in the vision of Brahman as a non-relational, non-dual entity. Both these features of the problem were crucial for Śa kara’s notion of the identity statements that morphed into mah -v kyas, because the mah -v kya doctrine developed in direct response to these two concerns. It is important to emphasize that the advocates of this type of meditation were Advaitins and bhed bheda-v dins of the aup dhika type (Bhart prapañca), theologians with whom Śa kara would have had little or no disagreement on the conclusion that ultimately, in any case in the 7 Mahadevan 1968:27. 234 state of liberation, there was nothing but Brahman. In other words, this was a doctrine much closer to home than the Brahma-Sūtra account. The doctrine of prasaṅkhy na developed around a few key texts from the Bṛhad- raṇyaka. One was the statement that introduced the triple process of Vedāntic s dhana, Yājñavalkya’s tm v are draṣṭavyaḥ rotavyo mantavyo nididhy sitavyaḥ. The nididhy sitavya part was specifically related to texts that introduce meditation on the Self or Brahman, such as tmety evop s ta, “One should meditate on it as the Self.”8 The most important among these was vijñ ya prajñ ṁ kurv ta, “Having known, one should cultivate insight,” from BĀU 4.4.21, for two reasons. First, it seemed to promote just the sequence of knowing Brahman that the prasaṅkhy na doctrine wanted: having first known, learned from scripture, one should cultivate insight, meditate. Second, the text was followed by a statement that those who wish to know the Self practice sacrifice and other Vedic forms of action (4.4.22), and by the cultivation of certain virtues (4.4.23). These texts are known to us from the Brahma-Sūtra account, and just like the BS our prasaṅkhy na-v dins wanted ritual and the related rama practices to accompany meditation in the pursuit of liberation, but the set of virtues in 4.4.23 in prasaṅkhy na seems to have corresponded to the yama-niyama complex in Yoga. In this chapter, thus, we will introduce the doctrine of prasaṅkhy na. I will begin by offering an account of the soteriology of the prominent pre-Śa kara Vedāntin Bhart prapañca. There we will become acquainted, for one thing, with the form of jñ na-karma-samuccaya or the combination of action and knowledge that Śa kara was most explicitly arguing against, but more importantly we will have the chance to see forms of the two kinds of meditation side by side, or rather in progression. Then I will introduce prasaṅkhy na in the P tañjala-Yoga- stra, where 8 BĀU 1.4.7. 235 we will see the key issues that this kind of meditation was addressing. Next, I will reconstruct the contours of the Vedāntic notion of prasaṅkhy na from the works of Śa kara and Sureśvara. We will end the chapter with Ma ḍana Miśra and the doctrine of meditation on Brahman in its fullest expression. The Soteriology of Bhartṛprapañca Bhart prapañca was an old Vedāntin the only thing about whom we know is that he wrote a voluminous commentary on the BĀU, which is now lost but was profusely discussed by Śa kara and Sureśvara, and from which Ānandagiri quotes directly in his commentaries to the Bh ṣya and the V rttika. Although his commentary was lost, Bhart prapañca was immensely influential and important for the history of Vedānta. He was, along with Kumārila, Śa kara’s main foil, but he also influenced the latter in positive waysŚ Śa kara’s cosmological and psychological categories were taken from Bhart prapañca. Fortunately, his doctrine can be reconstructed in its broad contours from Śa kara, Sureśvara and Ānandagiri.9 I will begin with a short statement of his cosmology and psychology because their details are soteriologically significant. Bhart prapañca was an advocate of a version of the theology of transformation, brahma- pariṇ ma-v da. He held that there is a single, non-dual Brahman, but one which factually 9 Nakamura’s date of Bhart prapañca, 550 CE, is as good as any date preceding Śa kara. Important studies of Bhart prapañca include Hiriyanna 1957:79-94 and 1972:6-16, Nakamura 2004:128-52, Andrijanić 2015 and 2016, Dasgupta 1932:43-46 and Pandey 1983:209-28. My presentation of Bhart prapañca’s soteriology is based solely on my own reading of the three Advaitins, however, with some aid from Hiriyanna. I have, I should like to acknowledge, benefited greatly from Andrijanić 2015, who gives quite an exhaustive list of passages where Bhart prapañca’s views are discussed in the BĀU corpus. For reconstructing the doctrine of meditation on Brahman and the combination of ritual and meditation, I have relied on BĀUBh 1.4.15, BĀUBhV 1.4.1692-1779, BĀUBh 1.5.18 and BĀUBhV 4.4.706-740. For Bhart prapañca’s cosmology and psychology, I depend on BĀUBhV 1.6.46- 51, BĀBh 2.1.20, BĀUBhV 2.1.466-72, BĀUBh 2.3.6 and BĀUBhV 2.3.90-104ab and 112-24, BĀUBh 4.3.7, BĀUBhV 4.4.393-412. Ānandagiri’s commentaries on the Bh ṣya (Ṭ k ) and the V rttika ( stra-Prak ik ) likely contain direct quotes from Bhart prapañca, and these are generally marked with bhartṛprapañca-vy khy m utth payati or anuvadati, and are often repeated verbatim in the two commentaries. For this reason, I have relied on Ānandagiri without reservation. 236 transforms itself into the multiplicity of the world. This pure Brahman, properly called Paramātman or the Supreme Self, has two states, a causal state in which it is prone to transformations, and an effected state in which it exists as actual transformations. As itself, however, it is beyond both. The transformations of Brahman are real, and Brahman stands in relation to them as the ocean stands in relation to drops of water, waves, foam, bubbles etc.10 The transformation of Brahman proceeds along two lines. On the one hand, in terms of creation of the physical world, Brahman is a state of undifferentiated potentiality, avy kṛta, which transforms into the manifold of the world by way of evolution of the five common elements and into all the details of creation. Significantly, these details include the divinities that are commonly identified in the Upaniṣads with cognitive and active functions, such as speech, sight, life-breath, hearing, touch, etc.11 The second line of transformation is that of consciousness, and it proceeds from another causal state of Brahman which is called Antaryāmin or Īśvara, the inner controller, and which Bhart prapañca describes as “a slightly agitated state of the supreme Brahman.”12 This Antaryāmin becomes the distinct Selves which are generally called kṣetra-jña, j va or vijñ n tman, and which are transformations of Brahman’s own consciousness.13 These are further distinguished as cosmic—the Selves embodied into the different deities of the Vedic 10 See BĀUBh 5.1.1, p.687-8Ś “Thus in all three times—origin, maintenance and dissolution—the cause and the effect are both infinite. The single infinitude is spoken of as divided into a cause and an effect. So, Brahman is one, both dual and non-dual, as an ocean consists of water, waves, foam, bubbles etc. And, as water is real, its effects— waves, foam, bubbles etc.—are the ocean in nature and have the properties of appearance and disappearance are also really real. Thus, the whole world of duality is nothing but really real, corresponding to the waves to the water, while the supreme Brahman corresponds to the water of the ocean.” See also BĀUBhV 1.4.487. 11 See BĀUBhV 2.3.92 and Ānandagiri thereon. 12 “On this some say, ‘The supreme immutable Brahman which is like a great ocean and is naturally not shaken has the state of the inner controller, that in which it is slightly shaken.’” BĀUBh 3.8.12. 13 “Those who explain the word vijñ namaya as a modification of the consciousness that is the Supreme Self…ś” yeṣāṁ paramātma-vijñāpti-vikāra eva vyākhyānam… BĀUBh 4.3.7, IX.526. Ānandagiri thereon (identical comment on BĀUhV 4.3.318)Ś vijñānaṁ paraṁ brahma tat-prak tiko jīvo vijñānamaya iti bhart prapañcer uktam anuvadati. 237 pantheon—or individual. Among the Selves embodied in the divinities, the highest is that of Hira yagarbha, which is “the Self of the world,” the universal soul known also as sūtr tman or the Self that is the “thread” that keeps the world together. The Selves of the other divinities, such as the Sun, are the Selves behind the cognitive and active functions on the macrocosmic level (adhidaivam). They are replicated on the microcosmic level of every individual (adhy tmam). Thus, Brahman evolves the active and cognitive functions on a macrocosmic level through the first line of creation, animates them through becoming the individual Selves of the Vedic divinities, and then replicates the same functions on a microcosmic level. The two transformations come together in the embodied individual, where the first line makes for what is called “adjuncts” (up dhi) of the embodied Self, referring to the elements that form its gross body, sense objects, and the active and cognitive functions that form its subtle body.14 They are non-essential to the Self—hence adjuncts—but they do alter the Self, as we shall see in a bit. It may be worthwhile here to think of the standard example of an up dhi in Indian logic, dampness, which is non-essential to wood, but which alters fire as wood’s function by causing smoke. The transformation of the Supreme Self into individual Selves happens to a part of Brahman (param tmaikade a, par ṁ a), which is cut off from it as it were through the power of ignorance, avidy .15 Ignorance is a category that assumes several roles in Bhart prapañca’s doctrine, and the ignorance that cuts off a shard of Brahman is the cosmic principle that seems equivalent to Brahman’s causal state as the first point in its material transformation, the 14 See Ānandagiri on BĀUBhV 2.3.92Ś jīvopādhitvena yāvat kāryaṁ avyāk tāj jātaṁ tat sarvam etāvad eva. 15 See BĀUBhV on 3.2.55Ś saiṣāvidyā paricchetrī yayāyaṁ paramātmanaḥ | saṁsāritvaṁ paricchinno vijñnātmatvam aśnute. “It is but ignorance that is the delimiter, owing to which a piece cut from the Supreme Self becomes the individual Self and through which the Supreme Self becomes liable to transmigration.” 238 avy kṛtaŚ “Ignorance is a power of K ṣ a arisen from the Supreme Self directly. Deforming a part of the Supreme Self, it makes its residence in it.”16 The part cut off from the Supreme Self becomes like a desert, a blemish of a part of the earth, but not the whole of it.17 The cutting off is creation in the logical sense, not temporal. Ānandagiri alludes to this when he uses the present participle: the entity that is the individual Self obtains from the Supreme Self as it is being cut, not as a posterior effect of a prior and completed action.18 This cosmic ignorance, however, is just the factor of individuation that separates the individual from the Supreme Self: it is the category maker. The full individuation such that the Self cut off from Brahman becomes a j va requires a further complex of three intertwined factors that alter its Brahman-nature: these are impressions of previous experiences since time immemorial that have become habits (v san , scent); desires (k ma) formed through habituation; and karma, the results of previous action performed through desire. These three are kind of a no man’s land, because they are neither transformations of Brahman’s consciousness nor products of Brahman’s potential materiality, but what happens when the two get together. They do, however, directly reach the Self and alter it, “like flower scent that is put in a basket or distilled in oil that remains there even when the flower has been removed.”19 Through their medium, thus, 16 If that is indeed the case, it would correspond to Śa kara’s avy kṛte n ma-rūpe that is the adjunct of Brahman and the stuff of which everything is made. 17 k ṣ a-śaktir avidyāpi parasmād eva sotthitā | vik tya paramātmāṁśaṁ vijñānātmani tiṣ ḥanti || yathoṣarātmako doṣaḥ p thivyā eva jajñivān | kṣmaikadeśaṁ vik tyās te ’vidyā tadvat parātmanaḥ. BĀUBhV 2.3.122-3. 18 On BĀUBhV 2.3.103Ś paramātmanaḥ paricchidyamānasya yena viśeṣe a vijñānātmatva-lābhaḥ. 19 BĀUBh 2.3.6, VIII.288Ś sarvaḥ karmarāśiḥ—puṣpāśraya iva gandhaḥ puṣpa-viyoge 'pi pu a-tailāśrayo bhavati, tadvat—li ga-viyoge 'pi paramātmaikadeśam āśrayati gandhaḥ puṣpāśrayo yadvat pu am āśritya tiṣ hati. BĀUBhV 2.3.17-8Ś gandhaḥ puṣpāśrayo yadvat pu am āśritya tiṣ hati | kusumāpagame ’py evaṁ li gasthā vāsanātmani || vāsanā-kāma-karmā i li gasthāny eva nātmani | li gād ātmānam āyanti gandho gandha-pu aṁ yathā. 239 even external objects factually affect the Self. Yet they are not essential to the Self, because their relationship with it is that of contact, and a contact can be broken.20 Bhart prapañca also mentions a psychological form of ignorance, which he takes from the school of Yoga and which is the first of the five well-known psychological torments or kle as. Ignorance here refers to a persistent cognitive error consisting in ascribing to the Self properties that do not belong to it.21 It is not quite clear what its relation to the v san -k ma- karma complex is—I did not find any elaboration of this—but it is certain that it is not the starting point and the root of all saṁs ra, as with Śa kara. Saṁs ra is not just a cognitive error: it is factual cutting off from Brahman that sets off a cycle of rebirth that is a real cycle of habits, desire, action, a cycle in which the three factors constitute a chain with no starting point. Since the psychological ignorance is a habitual cognitive error, I surmise that it can be categorized as one of the v san s, perhaps the one with deepest roots. The dual nature of the individual Self—essentially Brahman since being its real product, but an agent and enjoyer of the results of action, liable to rebirth, insofar as alienated from Brahman—brings with itself a dual entitlement in Vedic terms: as an entity liable to transmigration, the individual Self is obliged to perform Vedic ritual; as Brahman, it is entitled to realize its nature as Brahman and pursue liberation.22 This is a curious doctrine that looks like a blend of Sā khya-Yoga and Nyāya-Vaiśeṣika. Brahman’s causal aspect of avy kṛta and the vijñ n tman are quite reminiscent of prakṛti and puruṣa, and the v san s were no doubt taken from Yoga. But then, as impressions they are not 20 See BĀUBh 4.3.22ś BĀUBhV 4.4.391-412. 21 See BĀUBhV 4.4.725. The Yoga doctrine of kle a is part of the second chapter of the YS, and avidy itself is defined in 2.5 22 yenānyas tena saṁsārī karmādhik ta iṣyate | ananya-pakṣe ‘haṁ brahmety ukter brahma prapatsyate. BĀUBhV 2.1.471. 240 just a feature of the psyche: they do color the Self, and through them the mind and the senses impact the Self and the Self becomes the agent and enjoyer, as in Nyāya-Vaiśeṣika. Śa kara and Sureśvara, in fact, ridicule Bhart prapañca as a self-styled Vedāntin who has made a pact with the logicians but is cheating on them with the Sā khyas and, afraid that they may tell on him, runs to the Vaśeṣikas for rescue.23 The key takeaway from this brief presentation of Bhart prapañca’s cosmology and psychology is the dual character of the individual Self: as an agent of action and experience, the Self is obliged to perform ritual, but as Brahman it is entitled to liberation. These two, claimed Bhart prapañca, are mutually exclusive at least at their face value, and the reason is that ritual presupposes agency that is in the domain of ignorance. The simultaneous performance of ritual and the pursuit of liberation wasn’t a problem for the old Vedānta, but it became a serious problem with the introduction of the category of ignorance. We saw that Vedic theologians had struggled with how precisely to define and classify the obligatory rituals, but in all cases, they pursued the strategy or reclassification through another provision according to the expected result. For early Vedāntins, ritual was not inherently badś it was inefficient in securing a permanent attainment, and rebirth was at its root a consequence of that inefficiency: what is made can be unmade. It was, therefore, easy to reclassify ritual as something that nurtures meditation when performed along with it. Vedic theologians have been doing such reclassifications often, as is hopefully obvious by now, and in a straightforward manner. But, the appearance of ignorance, in its dual cosmological and psychological role, made ritual a problem, and quite a specific one. In the traditions of liberation where the notions of ignorance developed, agency was a product of ignorance and a determining factor in rebirth: 23 BĀUBh 2.3.6 and BĀUBhV thereon. 241 agency was at the root of saṁs ra, and it formed a part of a causal chain of rebirth. Everyone had some variation of this chain,24 and so did Bhart prapañca, as we just saw: The Self is delimited from Brahman by cosmological ignorance, and conditioned by psychological ignorance as its deepest impression that prompts action; impressions of previous actions form desires, desires prompt one to act again, actions produce or reinforce impressions etc. This was agency in general, involving all aspects of one’s being. But then, Bhart prapañca was a Vedic theologian and for him ritual was the significant action. The issue with this was that ritual turned out problematic at the root, not at the fruit. This was a problem because on the one hand it presupposed a total commitment to the whole of the Vedic corpus, which comes quite obviously from the many quibbles of Śa kara with Bhart prapañca’s commentary on the question of whose doctrine really compromises the unity of the Veda, while on the other it meant that what the Veda enjoins is ignorance, and ignorance from which one could not possibly escape if one were to abide by the words of the Veda.25 Bhart prapañca brought out his best to impress how central a problem this was for the pursuit of liberation, and to do that he invoked historically the earliest justification of why obligatory ritual had to be performed. This was the doctrine of debts, specifically its Taittir ya Saṁhit formulation which says that every Brahmin is born a debtor to a number of creditors, particularly the gods, the Vedic sages and the forefathers. Olivelle’s account pursues the history of this idea in some detail,26 but for our purposes it is enough to say that the debt to the Vedic sages was to be discharged by living the life of a Vedic student, and the debt to the forefathers by begetting sons who would 24 Potter 1965:93-116 has quite an illuminating discussion of these causal chains of rebirth. 25 BĀUBh 4.4.22, 5.1.1. 26 Olivelle 1993:46-53. 242 continue performing the rituals for the good of the forefathers in heaven. These two one would discharge at a point in life, but the debt to the gods was a lifelong mortgage: one had to continue performing ritual for the gods as long as one lived, under the provision that is already well- known to usŚ “One shall perform the daily fire oblation as long as one lives.” To appreciate the logic behind this doctrine, we can quote here from the Bhagavad-G t with profit: Prajāpati, after creating creatures and sacrifice together, said in the beginningŚ “Ye shall multiply by it, it shall be the cow that yields your desires. Give ye the Gods being with it, and the Gods shall give ye being. And thus giving each other being ye shall attain to the highest good. Themselves enhanced in their being with sacrifice, the Gods shall give ye the pleasure ye desire: he who enjoys their gifts without return to them is but a thief.27 The G ta explicitly called this divine-human codependence karma-bandhana or the bond of ritual,28 while the BĀU had compared it to the relationship of men and livestockŚ “Just as an animal is for men, so is a man for the gods.”29 For Bhart prapañca, this meant that one was obliged to stay in ignorance, obliged to act as a ritual agent and revolve in the cycle of saṁs raŚ do sacrifice for the gods, go to heaven, come back to do more sacrifice to no end. This on its face value meant that the pursuit of liberation was impossible: if one had to perform ritual to the end of life, one had to see oneself as an agent, an element subordinate to the ritual action and “an animal to the gods,” reap the impermanent results and be reborn once they are exhausted. Ritual was literally “things gone south,” rebirth through the course of the forefathers, the southern course (dakṣiṇ yana). How could one, then, make use of the other part of the Veda that is geared towards liberation, to which one is entitled because of being a product of Brahman, if the Veda requires one to remain in ignorance? 27 BhG 3.10-13, translation van Buitenen 1981:83. 28 BhG 3.9. 29 BĀU 1.4.10. 243  Bhart prapañca’s solution to this conundrum was based on what I will mark here as the first of two texts crucial to his soteriology: tm nam eva lokam up s ta, “One should meditate on the world as the Self.”30 Because the world is ultimately a transformation of Brahman, it is dual in nature: it is ignorance and saṁs ra—the Upaniṣad identifies it with death itself, mṛtyu—but it is also Brahman. If one can, therefore, turn ritual from its state of effect to the state of cause, such ritual would not produce perishable results: On this, some say: the results of ritual of a knower of Brahman who meditates on the world as one’s Self does not decay because of the combination with meditation. There are two senses of the word “world” which is inseparably related to rites. One refers to the manifest state, which is the repository of ritual and called ‘that which pertains to Hira yagarbha.’ He who meditates on this manifest and limited world that is invariantly related to ritual obtain results of ritual that are exhausted because he identifies himself with limited ritual. However, he who has reduced the world that is invariantly related to ritual to its unmanifest state, its causal form, and then meditates on it, obtains results of ritual that are not exhausted, because he identifies with ritual that is unlimited.31 This is the formulation of the doctrine of jñ na-karma-samuccaya, combination of “knowledge” and ritual. The crucial element in this doctrine was meditation, because ritual was supposed to be transformed through meditation. Let us see, therefore, what Bhart prapañca meant under “meditation on Brahman.” Bhart prapañca’s definition of meditation was nothing out of the ordinary: meditation is a stream of awareness that is of the nature of becoming.32 By “becoming” he meant turning oneself into something else by assimilating to its nature through contemplative absorption in it.33 The “something else” is a scriptural object, a divinity, and because of that, meditation is dependent on 30 BĀU 1.4.15. 31 svātma-lokopāsakasya viduṣo vidyā-samyogāt karmaiva na kṣīyate ity apare var ayanti; loka-śabdārthaṁ ca karma-samavāyinaṁ dvidhā parikalpayanti kila—eko vyāk tāvasthaḥ karmāśrayo loko haira yagarbhākhyaḥ, taṁ karma-samavāyinaṁ lokaṁ vyāk taṁ paricchinnaṁ ya upāste, tasya kila paricchinna-karmātma-darśinaḥ karma kṣīyateś tam eva karma-samavāyinaṁ lokam avyāk tāvasthaṁ kāra a-rūpam āpādya yas tūpāste, tasyāparicchinna- karmātma-darśitvāt tasya karma na kṣīyata iti. BĀUBh 1.4.15, VIII.153-4. 32 bhāvanā-jñāna-santānaiḥ prajñā-kara am ucyate. BĀUBhV 4.4.708 33 tad-upāstyā tanmayatā. Ānandagiri on BĀUbhV 4.4.717. 244 hearing about some asserted identity from scripture and understanding properly what is being said. This is the same meditative complex that we saw in the BS account— ravaṇa, manana, nididhy sana. Let us see how this sequence precisely worked. First, one hears a passage in the Upaniṣads that contains some identity statement, and then ascertains correctly what the words are and what objects they refer to.34 In general, that happens through an investigation of the meaning of the passage so that it becomes clear what is the correct reference of the meditational counterpart, which is not always straightforward. Then, once the word-reference relationship has been ascertained, the mind obtains a clear cognition of the object. Now, the mind in this context is really the Sā khyan buddhi, which in any cognitive act is said to assume the shape of the cognized object, to mold itself into the object.35 Since the mind molds itself into the object, in a true cognition the mind is “not distinguished from the external knowable object,” it looks exactly like it.36 In our case, insofar as there is a correct cognition of a scriptural object from what has been heard, the mind has taken the shape of that scriptural object. Finally, any cognitive act is possible in virtue of there being awareness to begin with: the mind is just an instrument and is not itself sentient. The mind must reflect back to the Self the properly ascertained object, and because the individual Self is vijñ n tman, cognition in nature, the cognitive content (vijñeya) is not distinguishable from it: The Self takes the form of the cognized object (remember how the v san s reached directly the Self).37 What remains, then, is to mull over this cognitive content that is a mental formation until the Self turns into the cognized object in conformity to the cognition. It can do that because it is vijñ namaya, 34 See BĀUBhV 4.4.713ś Ānandagiri on BĀUBhV 4.4.721. 35 mano-buddhyor abhedena vyvahāra-śrutāv iha. BĀUBhV 4.4.717. 36 tatrārtha-pratyayo yo ‘sau pramātur mānaso bhavet | aviśiṣ aḥ sa bāhyena meyenārthena sa gateḥ. BĀUBhV 4.4.715. 37 na kevalaṁ pratyayasyārthatvaṁ kintv ātmano ‘pi. Ānandagiri on BĀUBhV 4.4.716. 245 consisting in cognition. This is a general principle of meditative absorption, and the procedure itself is justified by a passage from the BĀU, which I will mark as the second crucial text for Bhart prapañca’s soteriologyŚ Having known him [the Self,] a wise Brahmin should cultivate insight. He should not muse over many words, for that is just wearing of one’s voice.38 Vijñ ya prajñ ṁ kurv ta, “having known, one should cultivate insight,” meant for Bhart prapañca that one should hear from scripture until the meaning is clear, but then one should sit down and meditate and through absorption assume the state of the meditational object. Meditation is, then, cultivation of insight, prajñ -karaṇa.39 We can now go back to our first text, tm nam eva lokam up s ta, “one should meditate on the world as one’s Self.” Now, we saw in Bhart prapañca’s cosmology that “the world” was the place of saṁs ra, coextensive with the sphere of ritual and the divinities that constitute it. These divinities were represented on the microcosmic level as well, as the set of cognitive and active faculties: sight, hearing etc. Meditation on the world as the Self, then, was meditation on saṁs ra, and it was supposed to begin by meditation on these divinities as cognitive faculties of the world-Self and adjuncts to one’s own Self. The central text for this procedure was one which Yājñavalkya repeats six times in his teachings to Janaka. The two discuss successive meditations on Brahman as speech, life-breath, sight, hearing, mind and the heart, which correspond to Agni, Vāyu, Āditya etc., before Yājñavalkya concludes each meditation with devo bhūtv dev n apyeti, “becoming a god, one joins the gods.”40 The text obviously plays on the double meaning of deva as cognitive faculty and divinity, so Bhart prapañca says that one should successively meditate on oneself as these 38 BĀU 4.4.21. 39 BĀUBhV 4.4.708. 40 BĀU 4.1.2-7. 246 faculties—adjuncts to the individual Self as the senses on the one hand and to the world-Self as gods on the other—until one fully identifies with the respective divinity. Becoming one by one each of them, turning one’s adjuncts into the universal adjuncts, one fashions oneself into the world, in its gross feature: the vir ṭ. In doing so, one emulates a mythical, primordial event of creation in which the highest Vedic divinity, Puruṣa or Prajāpati, finds himself alone in the world, creates humans and animals, and then creates the gods as superior to himself—those for whom he must sacrifice— before realizing that there is really nothing except him: it is himself alone who has fashioned himself into the world, and he is everything. This is “the god [but also and the man, puruṣa] who forgot who he was”41 and invented other gods so that he could have someone to adore. So, the final step in one’s own refashioning into the world through meditation is becoming this very Self of the world through meditative absorption, in the manner described in the passage that contains our mah -v kya-to-be, ahaṁ brahm smiŚ Since people think that they will become the Whole by knowing brahman, what did brahman know that enabled it to become the Whole? In the beginning this world was only brahman, and it knew itself thinking, “I am brahman.” As a result, it became the Whole.42 Brahman here for Bhart prapañca was Hira yagarbha, the supreme vijñ n tman or individual Self, the “universal individual”43 embodied as Prajāpati, and through meditation on Hira yagarbha as the Self one completes the Self-refashioningŚ having turned one’s individual 41 Doniger 2005:117-124. 42 BĀU 1.4.9-10. 43 samasta-vyastātmakaṁ sautraṁ tattvam. ĀG on BĀUBhV 4.4.721. 247 faculties into the universal faculties, one finally transforms one’s Self, vijñ n tman, into the collective Self of the world.44 This is an important step for Bhart prapañca for the following reason: the mere performance of ritual brings only saṁs ra as its result, and so long one is an element subordinate to ritual (karma- eṣa) there is no question of liberation: it is all ignorance reinforcing ignorance. However, when one has become the universal Self, Prajāpati, one is no longer part of the primordial compact between the gods and the humans, but one who has forged the compact and stands outside of it. Karma-bandhana no longer applies, and the gods have lost one animal. One becomes not a sacrificer to the gods, but to oneself. By meditation on saṁs ra, one overcomes saṁs ra, the Vedic world of divinities and ritual. This, however, is not the end of saṁs ra as embodiment, for attaining the state of Hira yagarbha is attaining Brahman that is really Brahmā, Brahman that is both dual and non- dual in nature: There is a state of the supreme Self called Viriñca that is dual-nondual in nature and is revealed by a combination of meditation and ritual. Because it is dual, it is perishable.45 Hira yagarbha is still an individual Self, a shard of Brahman, a world-soul with the divinities as universal adjuncts. Attaining to the nature of Hira yagarbha is not liberation. So, the real gain of this Self-fashioning in meditation is arriving to a state where the pursuit of liberation becomes possible. One is no longer a factor subordinate to ritual, but not quite Brahman either. Or rather, one is Brahman, but that is the k rya-brahman that Bādari talked about in the BS. 44 See Ānandagiri on BĀUBhV 4.4.721Ś prāmā ika-vyavahāre śruta-sūtra-pratipādaka-śabdāt ahaṁ hira yagarbho ‘smi iti samasta-vyastātmakaṁ sautraṁ tattvam ātmā yadā paśyati tadā tasya buddhi-v tti-rūpaṁ vijñānam āv tti- yogaṁ bhavati sa caivam utpanna-jñānas tad-abhyasan vijñānamayaḥ san vijñānānusāre a vijñeya-sūtramayatvaṁ gacchati. 45 dvaitaikatvātmikāvasthā viriñcākhyā parātmanaḥ | vijñāna-karmābhivya gyā dvaitatvāt sā kṣayātmikā. BĀUBhV 1.4.1702. Viriñca is a common name of Brahmā. 248  Therefore, now a second type of meditation must commence, meditation on Brahman not as the effect, but as the cause, not on Hira yagarbha but on Brahman proper.46 This meditation also begins with a text that establishes an identity relation, now specifically tat tvam asi, the second mah -v kya-to-be. It is also preceded by an investigation of meaning to ascertain the exact meditational counterpart, but now the procedure is deconstructive. One must gradually remove from one’s constructed universal Self everything that constitutes an adjunct by discarding the divinities that are the universal faculties, until one eventually sees that the counterpart of the sentence, the tat, is Brahman which is “free from saṁs ra and its cause, non- dual, real, without beginning or end.” Once that has been accomplished, one must meditate on this Brahman incessantly.47 The raison d'être of meditation in general, that is, in both kinds, is that scriptural knowledge is mediate, lacking in direct experience, and must become immediate through constant absorption of the mind in that scriptural knowledge.48 We will investigate this later in the chapter. The key instrument in the performance of this meditation is the mind, which must ward off thoughts of duality and maintain its absorption in Brahman. This is a problem, for two reasons: keeping a mind just means maintaining a sense of duality in which no real vision of the single Brahman is possible, yet the mind is the only instrument of meditation at one’s disposalś and, the mind is a product of ignorance, or Brahman’s state of avy kṛta. The solution of this final conundrum is a gradual, slow reconversion of the mind from its being an effect back to Brahman 46 samuccayas tato ‘nyo ‘yam avyakta-brahma-vidyayā | vyakta-sūtrātma-vidyāyāḥ paro ‘py eṣa samuccayaḥ. BĀUBhV 1.4.1709. ĀnandagiriŚ sūtropāster nityādi-karma aś caikaḥ samuccayas tato ‘nyo dvitīyo yathoktopāsti-karma aḥ kāra opāstyā samuccayaḥ. 47 Ānandagiri on BĀUBhV 4.4.721Ś sūtrātmanā sthity-anantaram sa puruṣas tathā sthito yadā tat tvam ādi-vākyād ahaṁ brahmeti paraṁ tattvaṁ samyak paśyati tadā tad-anusandhānād dhiyo devatādi-rūpā i krame a tyaktvā sarva- saṁsāra-nirmuktaṁ tad-dhetunā ca hīnam advayaṁ vāstavam anādyanantaṁ brahma nairantarye a yo vīkṣate sa sad eva. 48 See Ānandagiri on BĀUBhV 4.4.709, which is an elaboration of Bhart prapañca’s interpretation of vijñ ya prajñ ṁ kurv ta. 249 the cause through continual absorption in Brahman, in the same way as metal which is deposited in the ground from which it initially emerged as ore would in time reconvert into earth. This really means that the mind through absorption in Brahman becomes nirvanized in Brahman and ceases being a mind. This is a key and final step in liberation, because—we shall remember—the complex of impressions, desires and karma that constituted being an individual Self, j vatva, had its residence in the mind and through the mind it colored the Self. Once full absorption in Brahman is achieved and the mind has been dissolved, there is nothing left that would keep the Self separate, and the vijñ n tman also dissolves into the Supreme Self. The dessert had been reclaimed, liberation achieved. This second meditation is also accompanied by the performance of ritual, but now ritual is also transubstantiated into its cause, through the proximity to Brahman, in the same manner as the mind. Meditation on Brahman transforms ritual just as a charm transforms poison into medicine, or like yoghurt with sugar mixed in it does not cause fever but has calming effects. 49 At the end, it all becomes offering of Brahman into Brahman, and ritual no longer brings perishable results, but is a direct means of attaining liberation. We can compare now the two meditations. The first was assimilative, like the Brahma- Sūtra meditation, and its aim was the attaining of Brahman, but this was the k rya-brahman of Bādari. It could be repurposed for liberation if one did not intend to attain brahma-loka, and in fact it had to be repurposed because it was one’s only shot at winning an entitlement to brahma- jñ na. We can understand it as a form of constructive meditation which involved fashioning oneself into a future state by means of meditative absorption. The second meditation was rather deconstructiveŚ ascertaining one’s true identity as Brahman through the Upaniṣads, but then 49 BĀUBh 4.4.22. 250 absorbing the mind and action into this Brahman for the gradual dissolution of mind as the seat of the individuating factors that keep the Self separate from Brahman. While the first was fashioning of oneself through identification, the second was, as prasaṅkhy na-v dins insisted, based on real identity, based on knowledge, one which was, however, deemed insufficient, because of lacking in direct experience. Our future mah -v kyas were tied to the respective meditations. Brahma in the first stood for Hira yagarbha and was the meditational counterpart in the final step of the constructive meditation, where one identifies with the universal soul but not the pure Self. Tat tvam asi, on the other hand, was part of the deconstructive meditation, where one first had to remove everything that was not pure Brahman, and then begin the absorption in the meditational object. Let us now tackle prasaṅkhy na-v da directly and see why the propositional knowledge of Brahman obtained from the Upaniṣads was insufficient and why it had to be followed by meditation. First, however, we must prepare the ground by investigating the origin of prasaṅkhy na in the P tañjala-Yoga- stra. Prasa khy na in P tañjala-Yoga- stra The term prasaṅkhy na is relatively common in the PYŚ,50 where it stands for that kind of meditational practice or dhy na which, when coupled with the practice of kriy -yoga that consists of austerity, recitation of Om, and dedication to Īśvara—tapas, sv dhy ya and vara- praṇidh na—purges the mind of the psychological torments, kle as, to the degree where the 50 I follow Philipp Maas (2010a, 2010b, 2013) in referring to a P tañjala-Yoga- stra, a lexeme standing for the Yoga-Sūtra and the Bh ṣya which is commonly attributed to Vyāsa as a single work. Maas presents a strong case that the attribution of the canonical commentary on the YS to a Vyāsa is a relatively late phenomenon, and that in earlier sources the Bh ṣya and the Sūtra text are treated as a single stra. Prasaṅkhy na occurs once in the Sūtra text (4.29) and seven times in the Bh ṣya. The following account is based on the PYS itself and on a short but most useful and reliable paper by Ko Endo (2000) entitled “Prasaṁkhy na in the Yogabh ṣya.” 251 mind becomes nothing but discriminative knowledge that illuminates the distinction of its own being from that of the real Self, puruṣa.51 It is particularly important to note that Yogic prasaṅkhy na needs to be accompanied by the practice of kriy -yoga, because its operation in destroying the psychological torments requires assistance: kriy -yoga attenuates the torments, and prasaṅkhy na makes them as good as burnt, such that even if objects which might activate them are present right in front of one’s eyes, the torment that could potentially be activated being as good as a burnt seed would fail to germinate.52 The cooperation of kriy -yoga and prasaṅkhy na meditation is compared to the cleaning of a piece of cloth from dirt: one first needs to shake off the gross dirt, and then to wipe away carefully the fine, intransigent dirt. The first step is comparable to the application of kriy - yoga, whereas the second to the diligent engagement in prasaṅkhy na meditation.53 We should also note that the three practices that constitute kriy -yoga are part of the wider yama-niyama complex of restrains and observances that a practitioner of Yoga must follow: they are the last three niyamas. This will be important when we move onto Vedāntic prasaṅkhy na. 51 “The being of the mind, its last trace of impurity purged, is established in its own nature which is but discrimination of the Self from its own [mind's] being, and tends towards contemplation of the rain-cloud of dharma. Meditators call this [being of the mind] the highest prasaṅkhy na.” tad eva rajo-leśa-malāpetaṁ svarūpa- pratiṣ haṁ sattva-puruṣānyatā-khyāti-mātraṁ dharma-megha-dhyānopagaṁ bhavati. tat paraṁ prasa khyānam ity ācakṣate dhyāyinaḥ. YSBh 1.2, p.6. “That [kriy -yoga,] when practiced, attenuates the psychological torments and thus nurtures concentration. That is, in that case the attenuated torments can be as good as burnt by the power of meditation, having the property of not being activated.” sa hy āsevyamānaḥ samādhiṁ bhāvayati kleśāṁś ca pratanū-karoti. pratanū-k tān kleśān prasa khyānāgninā dagdha-bīja-kalpān aprasava-dharmi aḥ kariṣyatīti. YSBh 2.2, p.114-5. Kriy -yoga is defined in YS 2.1 as consisting in tapas, sv dhy ya and vara-praṇidh na, which the Bh ṣya identifies as austerity, recitation of Om and the study of the scriptures of liberation, and dedication to the “supreme teacher of all practices.” 52 prasa khyānavato dagdha-kleśa-bījasya sammukhī-bhūte 'py ālambane nāsau punar asti, dagdha-bījasya kutaḥ praroha iti. YSBh 2.4, p.117. 53 “The gross mental formations of the torments, attenuated by kriy -yoga, should be eliminated by prasaṅkhy na until made fine and as good as burnt. Just as a gross impurity of a piece of cloth is first shaken off and then the fine dirt is removed by effort and a suitable means, likewise the mental formations of the torments that are gross are a minor obstacle, whereas the fine mental formations are a major obstacle.” kleśānāṁ yā v ttayaḥ sthūlās tāḥ kriyā- yogena tanū-k tāḥ satyaḥ prasa khyānena dhyānena hātavyā yāvat sūkṣmī-k tā yāvad dagdha-bīja-kalpā iti. yathā vastrā āṁ sthūlo malaḥ pūrvaṁ nirdhūyate paścāt sūkṣmo yatnenopāyena cāpanīyate tathā svalpa-pratipakṣāḥ sthūlā v ttayaḥ kleśānāṁ, sūkṣmās tu mahā-pratipakṣā iti. YSBh 2.11, p.130. 252  Thus, the prasaṅkhy na meditation as practice purifies the mind to the degree where it becomes made solely of discriminative knowledge, a state which the YSBh calls dharma-megha- dhy na or sam dhi. At the stage of such purification of the mind, prasaṅkhy na matures from a means into a result, from practice to a form of cognition. It is important to note, further, that the prasaṅkhy na meditation does not quite eliminate the psychological torments: they are as good as burnt seed, and the subtle impressions from past lives (saṁsk ras) that activate the torments may still overcome the mind such that its discriminative knowledge is lost. For this reason, prasaṅkhy na needs to be repeated over and again so that the mind may stay in the state of perfect discrimination and remain aloof from other ideas.54 In a sense, the perfect state of prasaṅkhy na as-a-result is always future. The key feature of prasaṅkhy na meditation, however, is that the insight or cognition which it provides is perceptual in kind, a result of yogic pratyakṣa. This is important for two reasons. First, perception is concerned with particulars, vi eṣa, unlike inference and scripture which are concerned with universals and convey knowledge of particulars only through knowing the relations of such universals. We know that there is fire on the hill because we know that smoke and fire as categories are universally related. We learn about a specific lotus through composite descriptions of universals that involve syntactic relationsŚ it is blue, in someone’s courtyard etc. We do not see the smoke or the lotus as genus in neither case. Inference and scripture provide knowledge that is mediate, parokṣa. The insight that prasaṅkhy na, on the other hand, provides is not how the Self in general or as a category is different from prakṛti, but how this Self of mine is like that. This affords the cognition obtained through prasaṅkhy na immediacy, pratyakṣa, that scriptural knowledge lacks. Second, in Yoga—as we saw in the First 54 Endo 2000:79. 253 Chapter—perception was epistemologically primitive and foundational: scriptural knowledge, on the other hand, was a second-hand account of something previously experienced, and we were justified in believing what scripture says because a personally experienced state of affairs was reported by someone trustworthy.55 This immediacy or the perceptual character of meditation, in which one is privileged to knowledge of truth produced by concentration, defeats the torment-activating percepts because it is equal in primacy to them: both are perceptual in kind, and the meditative absorption is not inherently weaker than the common perception. We can illustrate this with Kālidāsa’s wonderful description of Śiva’s meditation (prasaṅkhy na) which sheltered him from the temptations that Kāmadeva caused in his Himalayan hermitage: Though he heard the songs of the heavenly nymphs, at this time Shiva the Destroyer was deep in his yogic meditation— it is truly said that for those who are masters of themselves no interruption can break their concentration.56 Let us note these features of the prasaṅkhy na meditation well: (1) its function is cathartic, but operating on the fine impurities of the psychological torments; (2) it must be accompanied by kriy -yoga practices, which attenuate the torments to the degree where prasaṅkhy na can be functional; (3) the result of this cleansing is that the mind becomes so pure that it can discern between its own being and the Self, at which point prasaṅkhy na from a 55 “Something seen or inferred by a trustworthy person is reported verbally so that what one has personally experienced may pass to someone else. The mental formation of a hearer arising from a word concerning its reference is called scripture.” āptena d ṣ o 'numito vārthaḥ paratra svabodhasa krāntaye śabdenopadiśyate, śabdāt tadarthaviṣayā v ttiḥ śrotur āgamaḥ. YSBh 1.7, p.20. 56 Kum ra-Sambhava 3.40: śrutāpsaro-gītir api kṣa e 'smin haraḥ prasa khyāna-paro babhūva | ātmeśvarā āṁ na hi jātu vighnāḥ samādhi-bheda-prabhavo bhavanti. Translation Smith 2005:109. 254 means transits into a result; (4) the torments, however, are just “as good as” burnt seeds, dagdha- b ja-kalpa, where kalpa is the operative word, and subtle impressions may always activate them, for which reason prasaṅkhy na must be practiced continually; and, (5) the cognition afforded by this prasaṅkhy na has that characteristic of immediacy which makes it better than the mediate cognition that is characteristic of scriptural hearing. The Ved ntic Prasa khy na In their commentaries on the Bṛhad- raṇyaka Upaniṣad, Śa kara and Sureśvara discuss several doctrines of meditation that seem like curious cases of an intersection of Yoga and Mīmāṁsā in a Vedāntic settingś that is, accounts of liberation that are predicated on a thoroughly Yogic psychology and doctrine of embodiment, but seek to justify meditation as a form of action requiring a Vedic fiat through the means of Upaniṣadic injunctions. One such fascinating case is found in the Sambandha-V rttika 440-455, a Prābhākara appropriation of the basic Yoga doctrine of arresting the functioning of the mind as a way of interpreting the Upaniṣadic injunctions of meditation such as tmety evop s ta. The account says that the Self, being an actual something denotable by words, a pad rtha, is not in the domain of knowledge from linguistic utterances, or sentences. Think of rice, a thing whose properties we know but which we use in ritual through the fact that it becomes an auxiliary to the action of offering.57 The Veda is a pram ṇa strictly on the mandate or niyoga that an action be performed, and through that mandate the only new thing that the Veda says about rice qua rice is that it is usable in the relevant ritual, serviceable to the mandated 57 “Others sayŚ because the Self is a category, it is knowable by other means, not from scripture, like other categories such as rice.” anya āhuḥ padārthatvāt pramā āntara-gamyatām | ātmano nāgamāt siddhir vrīhyādy-anya-padārtha-vat. SV 440. 255 action. The Self is like rice: it is known through recognition or pratyabhijñ na, which is a mode of immediate perceptual awareness, assisted by reasoning. This is the proper pram ṇa for knowing the Self, not the Upaniṣads. When the Upaniṣads do say things about the Self, the corresponding scriptural cognition is a form of an appearance present to consciousness, pratibh , but not a veridical cognition produced by a pram ṇa.58 The only new thing that the Upaniṣads can say about the Self is that it is subsidiary to the action of meditation. That is how, in fact, the Upaniṣads facilitate liberation: since embodiment is a result of the relation of the Self with the mind that is the seat of impressions, v san s, the Upaniṣads issue a mandate of meditation, following which one can achieve a suppression of the familiar mental impressions or of the mind itself, nirodha. Once such suppression has been achieved, the self- luminous Self has no choice but to shine through.59 This is, evidently, a Prābhākaran take on the second sūtra of the PYŚ, yoga citta-vṛtti-nirodaḥ, appropriated to deny validity to the propositional knowledge of the Self from the Upaniṣads.60 58 “It is from linguistic utterance that a mere phenomenon arises, not from a pram ṇaś” tataś ca pratibhā-mātraṁ śabdād iti na mānataḥ. SV 450ab. 59 “At that point [when the veridical recollective awareness has obtained,] a man is mandated by scripture to suppress the mental impressions or the mind itself, but not to understand the constitution of a thing. The second proceeds through another pram ṇa, and speech has no use in that regard. When the impressions have been suppressed, the Self shines through on its own, not depending on another pram ṇa, being self-luminous in nature. … The Self encounters what is undesirable simply on the account of impressions, and liberation obtains just by suppression, whether there are other means present or not.” tad-vāsanā-nirodhe ‘taḥ pumāñ śrutyā niyujyate || manaso vā nirodhe ‘sau na tu vastv-avabodhane | mānāntare a tat-siddher nātra vyāpriyate vacaḥ || svayaṁ-jyotiḥ-svabhāvatvān niruddha-svānta-vāsanaḥ | pramāntarānapekṣo ‘pi svayam ātmā prakāśate. SV 442-444. vāsanā-mātra-hetutvād ātmano ‘nartha-saṁgateḥ | anyopaye saty asati nirodhād eva muktatā. SV 446. 60 That this doctrine was a real thing is further suggested by Śa kara’s pūrvapakṣin in the BĀU 1.4.7Ś “Since the suppression of the mental states is different from the scriptural knowledge of the Self, and since we know that it has been enjoined for practice in another system, let this be enjoinedś” athāpi syāc citta-v tti-nirodhasya veda-vākya- janitātma-vijñānād arthāntaratvāt, tantrāntareṣu ca kartavyatayāvagatatvād vidheyatvam iti cet. 256  Other related cases of Mīmāṁsic meditations abound in the Bh ṣya and the V rttika, but just one of them is explicitly styled prasaṅkhy na. It is discussed by Śa kara in the eighteenth verse chapter of the US as the pūrva-pakṣa for his promotion of the deliberation on tat tvam asi as the means to liberation, as well as by Sureśvara in the SV 756ff and NaiS 3.88ff. Furthermore, the key prasaṅkhy na contention, that scriptural knowledge of Brahman is not enough for liberation and that it must be followed by meditation, is often discussed by the two masters even when they do not mention prasaṅkhy na explicitly, particularly under BĀU 1.4.7 and 4.4.21, where the two key injunctions of meditation are stated, tmety evop s ta and vijñ ya prajñ ṁ kurv ta.61 Now, the proponents of prasaṅkhy na claimed that this kind of meditation was different from the Upaniṣadic meditations that were based on assimilative absorption facilitated by cosmic correlations. We will see much more of this knowledge vs. meditation issue in Śa kara’s distinctions between saguṇa- and nirguṇa-brahman and meditation and knowledge, but for the prasaṅkhy na-v dins the question of validity of cognitions was quite important. That comes out distinctly from Sureśvara’s opponent’s claim in the NaiS: prasaṅkhy na was not like the stereotypical Upaniṣadic fancy that requires concentration, but is otherwise not objective, that is, is without a true object corresponding to the notion, as in the case when a man thinks that a female body that is full of feces is, in fact, lovely, and keeps musing on it. Rather, prasaṅkhy na is based on a true cognition produced by the Upaniṣadic identity statements and clarified through reasoned inquiry. The specific role of meditation was to engender full understanding of that same scriptural cognition of a true state of affairs: prasaṅkhy na was based on true cognitions, Śa kara does not say much under BĀU 4.4.21, but Sureśvara has a long discussion against Bhart prapañca and 61 Ma ḍana Miśra. 257 and it produced even truer cognitions.62 Meditation of this kind was not just concentration: it was based on truth, and it aimed at truth. Thus, at the core of the Vedāntic prasaṅkhy na was the conviction that the identity statements of the Upaniṣads were informative about a factual state of affairs. They were knowledge qua knowledge, and although the epistemic status of meditation was problematic and ambiguous—Sureśvara’s opponent’s wording is significantŚ meditation perfects the scriptural cognition, paripūrṇ ṁ pramitiṁ janayati, which places prasaṅkhy na in the court of pram ṇa— the proponents of this doctrine seem to have been unambiguous that the propositional knowledge of the Upaniṣads and not meditation was the pram ṇa for knowing Brahman. This conviction was fortified with some theological arguments as well: scripture could not affirm both unity of the Self with Brahman as literally intended, and meditation as independently meaningful, because the second presupposed the subject-object distinction. Thus, meditation was an accessory to the propositional knowledge, not the other way around.63 Meditation, however, was required, because the propositional scriptural knowledge of the Self-Brahman identity was insufficient for liberation. Scripture deals with universals, categories, and when the Upaniṣad says tat tvam asi, that means that the category of the Self is identical with Brahman. Such cognition does not translate to the experience “I myself am Brahman.” I do not “see” non-duality. Sureśvara’s statement of prasaṅkhy na, in fact, explicitly places scripture in the same category with reasoning: the propositional knowledge of the Upaniṣads is to 62 prasa khyānaṁ nāma tat tvam asy-ādi-śabdārthānvaya-vyatireka-yukti-viṣaya-buddhy-āmreḍanam abhidhīyate. tac ca anuṣ hīyamānaṁ pramiti-vardhanayā paripūr āṁ pramitiṁ janayati, na punar aikāgrya-vardhanayeti, yathāśeṣāśuci-nīḍe strī-ku ape kāminīti nirvastukaḥ puruṣāyāsa-mātra-janitaḥ pratyaya iti. NaiS 3.90. 63 This is discussed in some technical detail in the SV 770-790. 258 Brahman what smoke is to fire: neither show directly their respective object.64 Perception, on the other hand, presents particulars that affirm duality moment after moment. Thus, no matter how much one may believe that there is nothing but Brahman, it is enough to open the eyes and see just the opposite. Perceptual duality which one encounters moment after moment is bound to defeat the scriptural cognition “I am Brahman.” In addition to that, percepts easily activate their corresponding impressions that have been formed through long habituation.65 For these two reasons—perception is immediate, and it activates impressions—the scriptural cognition “I am Brahman” should, first, itself become immediate in kind, and therefore it is better understood as the result that comes after the application of the process of meditation.66 64 “By their nature, linguistic utterances express their meaning mediately. They secure validity by soliciting the help of reasoning. Reasoning is also ineffective in regard to understanding things as they are, just as smoke is in regard to fire. Reasoning too reveals the object by recourse to meditation.” parokṣa-v ttyā śabdo hi vadan svārthaṁ svabhāvataḥ | sambhāvayan pramā atvaṁ yuktiṁ svīk tya vartate || yāthātmyāvagame ‘śaktā dhūmo ‘gnāv iva sā ’pi ca | svī-k tyaiva prasa khyānaṁ yuktir vastuni vartate. SV 777-8. 65 “One does not obtain permanent liberation on hearing ‘You are Being.’ Therefore, continuous meditation along with reasoning are required.” sad eva tvam astīty ukte nātmano muktatāṁ sthirām | prapadyate prasañcakṣām ato yuktyānucintayet. USP 18.9. “He who has realized the truth about the Self does not need meditation. The ignorant, on the other hand, does not attain that result [knowledge of the Self] even if he had heard about it.” pratipannātma-yāthātmyaḥ prasa khyānādi nekṣate | ajñas tu śrāvito ‘py asmād vinā nāpnoti tat phalam. SV 785. “The strong impression that is produced by percepts inevitably defeats the scriptural cognition ‘I am Being,’ and one is led to external things through faults. Since cognitions got from scripture and reasoning are concerned with universals, percepts of particulars inevitably defeat them.” sad asmīti ca vijñānam akṣajo bādhate dhruvam | śabdottaṁ d ḍha-saṁskāro doṣaiś cāk ṣyate bahiḥ || śrutānumāna-janmānau sāmānya-viṣayau yataḥ | pratyayāv akṣajo 'vaśyaṁ viśeṣārtho nivārayet. USP 18.13-14 66 “The means should be enjoined after the result, ‘You are Being,’ has been stated. Nothing else but prasaṅkhy na is constituted for that purpose.” sad asīti phalaṁ coktvā vidheyaṁ sādhanaṁ yataḥ | na tad anyat prasa khyānāt prasiddhārtham iheṣyate. USP 18.17. “Although it is known from scripture [that one is Brahman,] the mediacy in this regard does not depart. It is for a direct experience of this that prasaṅkhy na is enjoined. … Scripture enjoins prasaṅkhy na for the one who has heard the narrative about the Self from scripture, but wants to make it his own, personal attainment, although not experienced [previously].” api śāstrāt prapanne ‘smin pārokṣyānapahārataḥ | tat-sākṣāt-kara āyaiva prasa khyānaṁ vidhīyate || 259 Second, because perceptual awareness does not cease so long one lives, and it continues presenting impression-activating percepts, this immediate awareness “I am Brahman” must be actively maintained: prasaṅkhy na is a repeated meditation. The cognition “I am Brahman” is future to the application of meditation, and in an important sense it always remains future, as the cognition-producing meditation must be repeated over and again. The meditation, furthermore, must be accompanied by practices that are like the kriy -yoga of the PYŚ.67 The whole idea is, obviously, influenced by the PYŚ. A lexeme that is commonly used in this regard is am di or aman di, “having calm as its first.”68 This most certainly refers to the complex of personal virtues that are listed in BĀU 4.4.23, and which every practitioner of the standardized BS brahma-vidy had to cultivateŚ “A man who knows this, therefore, becomes calm, composed, cool, patient, and collected.”69 It is absolutely important, however, that this list is found in the section of the Upaniṣad just after the injunction of meditation that was one of the paradigmatic prasaṅkhy na texts, vijñ ya prajñ ṁ kurv ta in BĀU 4.4.21. This complex attaches to the Vedāntic prasaṅkhy na quite parallel to the kriy -yoga practice that must assist the yogic prasaṅkhy na, and even Śa kara customarily refers to it as a wider set that is yama-niyama in nature, as we shall see in the later chapters. As I said above, the proponents of prasaṅkhy na were adamant that the propositional knowledge of unity of the Self with Brahman was the specific cause of cognition, or pram ṇa. Such knowledge revealed a domain that was otherwise unavailable. Because of the problem of śruter jñātātma-v ttānto ‘nanubhūtam api svakam | sthānakaṁ vañchataḥ śāstrāt prasa khyānaṁ vidhīyate. SV 771, 773. 67 “Therefore, for attaining experience, one should practice prasaṅkhy na with effort, along with Self-control etc., giving up practices that are opposite to its result.” tasmād anubhavāyaiva prasañcakṣīta yatnataḥ | tyajan sādhana-tat-sādhya-viruddhaṁ śamanādimān. USP 18.18. Also, SV 761 and 763 describe prasaṅkhy na as sop yo vihitaḥ and am dy-aṅg nvitaḥ, enjoined along with or furnished with auxiliary means, self-control etc. 68 SV 764; USP 18.18. 69 tasmād evaṁ-vic chānto dānta uparatas titikṣuḥ samāhito bhūtvā. Translation Olivelle 1998Ś127. 260 mediacy, however, meditation was required, and such meditation had the status of a function or a vy p ra of this propositional knowledge, its efficient mode. Meditation was the propositional knowledge put to use. The prasaṅkhy na claim was, to remember, that the understanding of unity does not drive away ignorance on its mere arising, and the role of continuous meditation was to elevate pram ṇa as veridical cognition to its culmination point. Scriptural knowledge was not experience, and only as experience did pram ṇa serve the good of man, liberation.70 Finally, the doctrine of prasaṅkhy na, steeped in Yoga psychology and practice, was nevertheless a doctrine of Vedic theology, and it took the Upaniṣadic injunctions of meditation on the Self as statements with real injunctive force. The proponents of such meditation were aware that prasaṅkhy na was a Yogic doctrine, and by Yogic we mean the tradition of the PYŚ which is repudiated in the BS.71 Meditation seems to work well for yogis, all right, but for Vedic theologians Yoga was a laukika or worldly matter, and an important consideration was whether 70 “Prasaṅkhy na is enjoined for the one who has not experienced the Self although he has heard about it from scripture and wants to make it his own residence. Only through meditation as its own function does scripture reveal its own meaning. It cannot do so without such function. For this reason, there is no split of meaning of scripture [between the propositional knowledge and the injunction to meditation.] It is not right that agents [here, the propositional knowledge of the Upaniṣads] conceal their own function. Therefore, it is justified that meditation serve the meaning of unity [between Brahman and the Self]. Even after revealing such unity, scripture does not attain completion prior to understanding the human good, because of the danger that it may lose validity. … In this way, the linguistic utterance [the propositional knowledge of the Upaniṣads] personally establishes the object: by assuming primacy regarding the injunction, it becomes a reliable warrant.” śruter jñātātma-v ttānto ‘nanubhūtam api svakam | sthānakaṁ vañchataḥ śāstrāt prasa khyānaṁ vidhīyate || svavyāpāra-mukhenaiva śāstraṁ svārthāvabodhak t | na tu vyāpāra-virahān nāto dvikaratā śruteḥ || sva-vyāpāra-tirodhānaṁ kārakā āṁ na ca sthitam | tasmād aikātmya-tātparye prasa khyānādi susthitam || bodhayitvāpi caikātmyaṁ nāntarā paryavasyati | ā pum-arthābodheḥ śāstram aprāmā ya-bhayāt sphu am. SV 773-776. ity-ādi-vartmanā śāstraṁ sākṣād vastu prasādhayet | vidhiṁ prati pradhānatvaṁ svīkrtyābhyeti mānatām. SV 779. 71 David Gordon White in his recent book The Yoga Sutras of Patanjali: A Biography (2014) talks about Yoga with an upper-case Y as a philosophical system, “a theory of everything,” versus lower-case yoga for all other uses. There is no doubt that there is some yoga in the BS (we saw that a form of meditation was its soteriological practice), but Yoga with the doctrine of avidy , kle a etc. is foreign to it, and is presented as a form of Sā khya, which in general is the recipient of the fiercest and most sustained beating. 261 such meditation was approved of in the Veda. Given the specific problem that the Vedāntic notion of prasaṅkhy na was addressing, the mediacy of the propositional knowledge of the Upaniṣads, its proponents promoted a doctrine of Vedic injunctions that affirm meditation as a means of immediate knowledge of the Self. “It is known in the world that meditation is related to seeing. Lest it be doubted that it is likewise in the Veda, [meditation] is enjoined.”72 An injunction of meditation was required, in other words, for attaining immediate knowledge of the Self’s being Brahman as a form of direct experience.73 Śa kara identifies such injunctions as niyoga, which was the preferred terms of the PrābhākarasŚ the Veda not only discloses meditation as the means of the clear immediate knowledge of the Self: it also, and primarily, issues a mandate that such meditation be performed.74 Four texts are commonly discussed by Śa kara and Sureśvara as such injunctionsŚ Yājñavalkya’s statement to Maitreyī that introduced the Vedāntic method, tm v are draṣṭavyaḥ rotavyo mantavyo nididhy sitavyaḥ maitrey in BĀU 2.4.5 and 4.5.6ś tmety evop s ta in BĀU 1.4.7, in the text that contains the future mah -v kya ahaṁ brahm smi; vijñ ya prajñ ṁ kurv ta in BĀU 4.4.21 that is part of the text that introduces calm or ama and other virtues as practices that must accompany meditation; and ChU 8.7.1, ya tm pahata- p pm vijaro vimṛtyur vi oko vijighatso 'pip saḥ satya-k maḥ satya-saṅkalpaḥ so 'nveṣṭavyaḥ sa vijijñ sitavyaḥ, “One should investigate and know distinctly the Self which is free from faults, 72 loke darśana-sambaddhaṁ prasa khyānaṁ samīkṣitam | vede ‘pi kiṁ tathā tat syān na vā ‘tas tad vidhīyate. SV 765. 73 “Therefore, possessed of calm, etc., one should practice prasaṅkhy na with effort, for the purpose of direct experience, giving up whatever is contrary to this means and its result.” tasmād anubhavāyaiva prasañcakṣīta yatnataḥ | tyajan sādhana-tat-sādhya-viruddhaṁ śamanādimān. USP 18.18. 74 “Since [the Self's being Brahman] is not understood [just from the propositional knowledge of the Upaniṣads,] the mandate [to prasaṅkhy na] is not contradictory to it until such awareness becomes firm, as in the case of ritual.” niyogo 'pratipannatvāt karma āṁ sa yathā bhavet | aviruddho bhavet tāvad yāvat saṁvedyatād ḍhā. USP 18.11. 262 from old age, death, sorrow, hunger, thirst, whose desires and intentions are real.” More specifically, the proponents of prasaṅkhy na claimed that these statements were injunctions of the apūrva type that we discussed in Chapter Two, that is, injunctions which disclose a causal relation between a specific practice and a future result.75 This had two consequences, one concerning the nature of meditation, and the other, the nature of liberation. First, meditation was not a natural means of bringing the mediate propositional knowledge of the Upaniṣads from mediacy to immediacy. As I said above, meditation was a worldly thing, and a Vedic injunction was required to promote it as a means of immediacy: such immediacy of the Self was empirically unavailable, and without a Vedic fiat one could not be certain that meditation was adequate. This resulted in an interesting encounter of two doctrines of scriptural immediacy, that of Yoga (and the other traditions of liberation) and that of Vedic theology. In Yoga, scriptures were based on direct experience, and practice was supposed to emulate such experience so that one’s own experience of supersensible things would become likewise immediate. In Vedic theology, scriptures were a form of direct experience akin to perception, our only window to things supersensible, and there was no cognitive agent behind them. The proponents of prasaṅkhy na, thus, wanted a direct experience of Brahman for which scripture was necessary but insufficient, but needed scripture to sanction the practice of meditation. One kind of direct experience was supposed to sanction another. 75 This is most thoroughly discussed in BĀUBh 1.4.7 and the V rttika thereon, but Śa kara consistently introduces the possibility of the respective statements being apūrva injunctions before rejecting it, as we will see in Chapter Nine. From the SV 754 and Ānandagiri thereon, it would appear that these injunctions were further organized in specific groups that formed a coherent network: statements of the kind of tm draṣṭavyaḥ were uttpati-vidhis or statements that introduce meditation as a process, like the major ritual injunctions, whereas statements as vijñ ya prajñ ṁ kurv ta, interpreted by Ānandagiri as mokṣa-k maḥ prajñ ṁ kurv ta, were adhik ra-vidhis, presenting the practitioner to whom meditation pertains with the means to the desired result; through implication, this would be the one who had heard about the Self, acquired the propositional knowledge of the Upaniṣads. These are fascinating details for the history of Vedānta, but I will not pursue them, gaurava-bhay t. 263  Second, the vision of the Self that such meditation was supposed to bring about would terminate in liberation at a future time: liberation was not a visible result, and would necessarily be experienced postmortem. Even if immediacy of the propositional knowledge was attained in life, the prasaṅkhy na doctrine promoted a stream of awareness, like the stream of the Ganges, or an “accumulation of meditation” that had to be maintained for a long time, in any case till the end of life. This seems to have been the doctrine that Śa kara argues against in the BĀUBh 1.4.10 as well as in the introduction to the TUBh, namely that it was the stream of awareness concerning the Self that removes ignorance and brings about liberation, not the first clear cognition, with the final punch taking place at death.76 The “Another Cognition” Meditation Under the above prasaṅkhy na account, it was the same scriptural cognition, albeit transformed by the continuous meditation, that would drive away ignorance and bring about liberation. Meditation was like the salt to a dish: the substance was still the same, but with the added virtue of immediacy. There were, however, other Vedāntins who claimed that liberation required another cognition, different from the scriptural. Śa kara presents this idea in his BĀUBhŚ Others sayŚ “By means of meditation one should bring about another, special cognition of the Self. The Self is known by means of this cognition, and only this cognition dispels ignorance, not the cognition of the Self arisen from Vedic sentences. The following statements have this meaningŚ ‘Having known, one should cultivate insightś’ ‘The Self should be seenŚ it should be heard about, pondered over, meditated on,’ ‘One should search out, investigate that Self.’”77 Sureśvara expressed a similar ideaŚ 76 BĀUBh 1.4.10 should be read in detail regarding this, but Śa kara there argues against the doctrine that it is the last cognition of the Self, or alternatively the uninterrupted stream of awareness, that removes ignorance and leads to liberation. There is a little doubt that the two are sides of the same doctrine. 77 apare var ayanti upāsanenātma-viṣayaṁ viśiṣ aṁ vijñānāntaraṁ bhāvayet, tenātmā jñāyate, avidyā-nivartakaṁ ca tad eva, nātma-viṣayaṁ veda-vākya-janitaṁ vijñānam iti. etasminn arthe vacanāny api — vijñ ya prajñ ṁ kurv ta [BAU 4.4.21], draṣṭavyaḥ rotavyo mantavyo nididhy sitavyaḥ [2.4.5], so 'nveṣṭavyaḥ sa vijijñ sitavyaḥ [ChU 8.7.1.3] ityādīni. BĀUBh 1.4.7. 264  Others, however, sayŚ “Because the cognition ‘I am Brahman’ arisen from Upaniṣadic sentences is relational, it does not quite reach the real nature of the Self. – What then? – This same cognition brings about another, non-propositional cognition, for the one who meditates on it perpetually like the stream of the Ganges, and that cognition alone dissipates the darkness of ignorance without a remainder. The proof for this is the statement ‘Having known, a br hmaṇa should cultivate insight.’”78 It is not fully clear just how this view was different from the previous, since it is but a matter of perspective whether a transformed cognition is taken to be the same with or different from the original. However, this second view came along with an additional claim: the liberating cognition must be different from the scriptural, because the scriptural sentential cognition is relational, predicated on difference, and but an approximation of the non-dual Brahman. In one form or another, this doctrine was characteristic of the abd dvaita-v da of the grammarians and Ma ḍana Miśra.79 I will call it the doctrine of “another cognition” meditation. For the purpose of brevity, and having in view historical attributions in Advaita Vedānta, in reconstructing this doctrine I will focus on Ma ḍana Miśra.80 78 apare tu bruvate, vedānta-vākya-janitam ahaṁ brahma iti vijñānaṁ saṁsargātmakatvāt, ātma-vastu- yāthātmyāvagāhy eva na bhavati. kiṁ tarhi? etad eva ga gā-sroto-vat satatam abhyasyato ‘nyad eva avākyārthātmakaṁ vijñānāntaram utpadyate. tad evāśeṣājñāna-timirotsārīti, vijñ ya prajñ ṁ kurv ta br hmaṇaḥ iti śruter iti. 79 In the Vṛtti on the V kyapad ya, for instance, Hariv ṣabha (who was likely Bhart hari himself) says the following: “When there is knowledge of the non-duality of the nature of the word, preceded by [ascertaining the correct forms of] words, one attains unity through the termination of sequence [of verbal action]. Through the use of correct forms, a special form of merit appears, and becoming united with the great Word, one attains freedom from the senses. After reaching the undifferentiated state of speech, he comes to the source of all transformation of speech: the intuition. From that intuition which is but conformance to Being, and which looks towards its result through the repetition of cultivation of the union that is preceded by [the correct form of] words, one attains the highest source, which is no more than a state in which all description of transformation has been abolished.” śabda-pūrvaṁ hi śabda-svarūpasyābheda-tattva-jñāne krama-saṁhāre a yogaṁ labhate. sādhu-prayogāc cābhivyakta-dharma-viśeṣo mahāntaṁ śabdātmānam abhisambhavan vaikara yaṁ prāpnoti. so 'vyatikīr āṁ vāg-avasthām abhigamya vāg- vikārā āṁ prak tiṁ pratibhām anuparaiti. tasmāc ca sattānugu ya-mātrāt pratibhākhyāc chabda-pūrva-yoga- bhāvanābhyāsāksepāt pratyastamita-sarva-vikārollekha-mātrāṁ parāṁ prak tiṁ pratipadyate. VPV 1.14, p.47-8. While the Vṛtti is notoriously difficult to read, the import seems to be that the perfect cognition of Brahman that is speech in nature requires retracing the stages of transformation of speech, and the insight into the highest stage of speech requires the practice of meditation on the verbal intuition. 80 In the long commentary on the BĀUBhV under 4.4.21, Sureśvara presents two pūrva-pakṣa interpretations on vijñ ya prajñ ṁ kurv ta, and the second is identified by Ānandagiri as that of Ma ḍana Miśra (ma ḍanādīnāṁ tad- vyākhyām utthāpayati). The doctrine did not originate with Ma ḍana, however, as is evident from the BrS p.33-36, 265  Now, Ma ḍana famously saidŚ There are three forms of understanding regarding Brahman. The first is from linguistic utterance [scripture]. The second is variously called dhy na, bh van , up sana, and consists in continual meditation on what has been heard. The third is direct experience; it is the state of completion in which all mental constructs have been annulled.81 The Upaniṣads were, Ma ḍana affirmed, our only window into knowing Brahman,82 but the problem with them was that they presented Brahman as a relational entity. To be precise, they did present Brahman as the only real thing behind the apparent multiplicity, but in doing so they relied on words that stand for categories, whose use was predicated on ignorance. They presented Brahman in a way that had to rely on illusion, though not intending to affirm illusion: they could not present Brahman directly, but had to do so through composite, determinate description. Let us see how this was the case. Very briefly, Brahman for Ma ḍana Miśra was that most general category or pad rtha that apparently evolves into the multiplicity of creation, that is, into all the specific pad rthas of all kinds. This Brahman, further, was speech in nature, identical with the holy syllable Om: Ma ḍana was, like Bhart hari, a proponent of abda-vivarta-v da.83 Since creation apparently evolves from Brahman, Brahman can best be understood as Being and as the cause of everything. And, since this Being and cause is speech in nature, Brahman is that great genus, s m nya, that diversifies itself in all the categories of language. This general notion of Brahman as the ground of Being and the great genus is formed based on the creation passages of the and Śa kara likely knew about it from some older Vedāntins or abd dvaita-v dins. One could support Thrasher’s conclusion that Ma ḍana knew Śa kara’s Brahma-Sūtra-Bh ṣya rather than the other way around with additional evidence (1993:111-128), but that we will leave for another occasion. 81 tisraś ca pratipattayo brahma i. prathamā tāvac chabdāt, anyā śabdāt pratipadya tat-santānavatī dhyāna- bhāvanopāsanādi-śabda-vācyā, anyā tato labdha-niṣpattir vigalita-nikhila-vikalpā sākṣāt-kara a-rūpā. BrS p.74. 82 BrS 4.4, p.157. 83 On Bhart hari’s influence over Ma ḍana, see Timalsina 2009a and 2009b; on Ma ḍana’s abd dvaita-v da, see Thrasher 1993; Balasubramanian 1983:91-103 and 1976:134-145. 266 Upaniṣads, such as the famous yato v im ni bhūt ni j yante of the Taittir ya, but the crucial prerequisite for forming the notion of Brahman was that its formation was possible because causality and Being are categories that we are otherwise acquainted with.84 Brahman as the great cause and genus, however, was an insufficient characterization of Brahman: it was the positing of the category as a something denotable by words, describable in language, but it was not informative of Brahman’s characteristic nature. Now, in the Upaniṣads this general causality or Being is further associated with some positive qualities, such as consciousness and bliss, and more generally disassociated from the kinds of things that the apparent transformations of Brahman are, through negative attribution: Brahman is not gross nor fine, not liable to hunger and thirst, etc. The three modes of predication jointly manage to present Brahman in its true, specific nature: (1) the creation statements set the category to which Brahman belongsś (2) the positive characteristics present Brahman’s essential nature through qualities that are empirically available to us; (3) the negative statements deny any limitations and abolish the multiplicity of Brahman’s apparent transformation, such that only Brahman as a positive thing remains. The processes involved in this are association with some and exclusion from other qualities, saṁsarga and bheda, and generally this is how sentential references of determinate descriptions are formed.85 This Upanisadic knowledge of Brahman can be likened to descriptions of unknown things which we, nevertheless, understand. Take, for instance, a traveler who visits an island and 84 BrS, p.157. 85 The two processes were, in fact, two competing theories about sentence meaning formation in Indian philosophy of language, to be precise, about the relations of the things signified by the various terms in a sentence. Historically, they are traced to two pre-Kātyāyana grammarians, Vājapyāyana and Vyāḍi, and they stemmed from disagreement over the reference of individual wordsŚ while Vājapyāyana claimed that the reference was the universal, j ti or s manya, Vyāḍi claimed that it was the individual, dravya. Ma ḍana as a proper Mīmāṁsaka was a proponent of the first: words denote universals, and the sentence meaning is just the relation of the word meanings, their association; it seems, though, that he uses bheda with the negative attributes. On the two, typically lucid discussions are available in Hiriyanna 1972:73-77, and Kunjunni Raja 1977:191-93. 267 sees an uncommon bird. When she gives her descriptionŚ “On this island, there are these birds, they call them ‘Jewelfowlś’ they have feet made of emerald, beaks made of ruby, and wings made of gold and silver,” the description is informative enough because we are acquainted with the categories that the traveler had used. A specification of the “bird” genus is made through a predication of characteristics, such that through the testimony we are acquainted with a specific bird species that we have not seen before.86 It does not even matter if there really are such birds or not: we have meaningful verbal cognitions even when we know that there are no corresponding objects, as in the case of the expression “hare’s horn.”87 Through composite description, language as the ground of Being can convey the notion of almost anything, and reality in any case is not really any more real than Pegasus might be. The point is, the Upaniṣads must use these general categories if they are to be a source of knowledge about Brahman: they must convey the unknown through the known. More generally, their statements must be synthetic to be informative, because that is what sentences as pram ṇa do. It is not enough just to take Brahman as conveyed by the pad rtha of Being or causality, because that is not sufficiently informative about the specificity of Brahman, for which a sentence is required. Brahman in its specific nature is not in the domain of words, but a sentence can convey much about it. This poses the obvious problem that we have already discussed—such knowledge is mediate—but it poses another, specifically Advaita problem: it presents Brahman as a relational entity, a definite description that obtains as a sentential reference, a verbal composite as it were—saṁsarg tmaka as Sureśvara says—the knowledge of which depends on 86 The ‘Jewelfowl’ is my substitution for evaṁ-n m naḥ (with credits to Stephen Walker). yathā loke ‘dvīpa-viśeṣe evaṁ-nāmāno maratakamaya-pādāḥ padma-rāgamaya-cañcavaḥ sauvar a-rājata-pakṣāḥ pakṣi aḥ' iti pakṣi- sāmānyaṁ pramā āntarādhigatam ananya-sādhāra ena dharma-kalāpena saṁs jyamānaṁ pramā āntarānadhigataṁ viśeṣa-rūpam āsādayati. BrS p.157. 87 BrS p.18. 268 categories that we know of through duality. It is a form of analogical predication that depends on there being analogues with which we are acquainted through perception, and that presupposes duality and ignorance. The Upaniṣad may not affirm them as intended, but it cannot do without them. Therefore, this first sentential cognition of Brahman depends on ignorance, avidy . It provides knowledge of Brahman and it is the pram ṇa for knowing Brahman, but it depends on concepts, mental constructs or vikalpas, and for this reason it does not stop ignorance. One, therefore, must engage in the processes of repeated meditation on the Self from which all multiplicity has been denied, in the above manner, a meditation that is preceded by hearing and thinking, ravaṇa and manana, along with practices such as celibacy. This is, of course, the triple Vedāntic process of ravaṇa, manana and nididhy sana. Ma ḍana specifically calls these “means proclaimed in scripture” for dispelling ignorance, and it seems important to note that they are not a function of and an indispensable auxiliary to pram ṇa as was the case with the previous account of prasaṅkhy naŚ it is solely the Upaniṣads that disclose Brahman, but a further means is required to dispel ignorance and effectuate liberation.88 This is meditation on Brahman that is promoted as such a means in the Upaniṣads. Now, this is, again, an obvious problem, because meditation, along with the auxiliary means, is an instrument of some sort, and the exercise of its instrumental agency over a patient requires a distinction between the two. The patient in this case is the vision of duality, bheda- dar ana, and meditation should undo that on which it itself depends. Not an easy task. Ma ḍana, therefore, says that meditation is eventually deconstructed in the Self, deconstructing in the 88 kena punar upāyenāvidyā nivartate? śrava a-manana-dhyānābhyāsair brahmacaryādibhiś ca sādhana-bhedaiḥ śāstroktaiḥ. katham? yo 'yaṁ śrava a-manana-pūrvako dhyānābhyāsaḥ pratiṣiddhākhila-bheda-prapañce sa eṣa neti neti ātmani, sa vyaktam eva bheda-darśana-pratiyogī tan nivartayati. BrS p.12. 269 process the distinction between the agent, the action, and the object of meditation, in the manner of some powder that reacts with the dirt in water, eliminating both the dirt and itself, or of a poison that is an antidote to another poison.89 Such hearing and meditation are a form of prophylactic ignorance, then, a docta ignorantia of an Advaita sort, that is different from what the Upaniṣads call prapañca, bheda or mṛtyu. Ma ḍana refers to the famous Īśopaniṣad mantra which says that one overcomes death by ignorance, and then achieves immortality by knowledge: that is, by the processes of liberation that depend on distinctions and, thus, constitute ignorance, one gradually deconstructs grosser ignorance, multiplicity or “death.” Finally, and when the pure Self shines through, one achieves liberation, a state of remaining in the Self.90 This state is, of course, equivalent to the third form of understanding regarding Brahman, direct experience. This direct experience that is brought about by meditation is the “another, special, non-propositional cognition of Brahman” that Śa kara and Sureśvara have in mind. While the propositional knowledge reveals Brahman as a pram ṇa, the meditational complex that involves the practices of ravaṇa, manana and dhy n bhy saṭbh van ṭprasaṅkhy na, along with celibacy and the like, is the means that brings about the removal of ignorance, which is equivalent to the direct experience of Brahman, and with that it brings about liberation. 89 sa ca sāmānyena bheda-darśanaṁ pravilāpayann ātmanāpi pravilīyate. na ca śrot -śrava a-śrotavyādi-vibhāga- parihā yā vibhāgāntara-niv tti-viṣayāḥ śrava ādayaḥ, api tu sāmānyena. tathā ca tasminn api pravilīne svacchaḥ pariśuddho 'syātmā prakāśate. yathā rajaḥ-samparka-kaluṣitam udakaṁ dravya-viśeṣa-cūr a-rajaḥ prakṣiptaṁ rajo'ntarā i saṁharat svayam api saṁhriyamā aṁ svacchāṁ svarūpāvasthām upanayati, evam eva śrava ādibhir bheda-darśane pravilīyamāne viśeṣābhāvāt tad-gate ca bhede, svacche pariśuddhe svarūpe jīvo 'vatiṣ hate. BrS ibid. 90 vidyāṁ cāvidyāṁ ca yas tad vedobhayaṁ saha | avidyayā m tyuṁ tīrtvā vidyayām tam aśnute. ĪU 11. avidyayā m tyum iti. eṣo 'rthaḥ—nāvidyā vidyāyāḥ sādhanam. kiṁ tv avidyayā śrava ādi-lakṣa ayāpy avidyaiva nivartate. m tyur ity avidyaivocyate. tasyāṁ niv ttāyāṁ vidyā-rūpopalakṣitam am tam aśnute svarūpāvasthānaṁ spha ika-ma ir ivopādhyāśraya-nibandhanoparāga-tyāgāt. antare a prayatnāntaraṁ vidyā-svarūpe 'vatiṣ hata iti. BrS p.13. 270  This is, of course, redolent of the sab jaṭnirb ja-sam dhi of Yoga. In addition to that, in the few instances where Ma ḍana explicitly mentions prasaṅkhy na, the important thing is that such meditation is the only thing that can undo the psychological faults that have roots in ignorance and prevent one from experiencing Brahman and liberation.91 This suggests that meditation on Brahman is tantamount to purification that ends with the removal of ignorance. It is not a removal of faults in understanding or cognition, but a removal of personal faults. In this light, it makes perfect sense why Ma ḍana would consider ritual to be causally effective in the pursuit of liberation, alongside hearing, thinking and meditation and their auxiliaries such as celibacy. If meditation is a prophylactic agent, then its effect can be accelerated with a catalyst. This is mightily important to bear in mind, because for Śa kara the removal of personal faults will happen on a different stage, before one could begin the inquiry into Brahman, and the faults that the s dhaka has in the application of the three processes will be of a different kind, strictly concerning understanding. We should note here that such understanding of gradual purification after fully knowing Brahman from the Upaniṣads allowed Ma ḍana to claim how it was possible for ignorance to remain operative even after the propositional knowledge from the Upaniṣads was crystal clear and indubitable. Balasubramanian may be quoted on this with profit: In the same way, the illusion (mithy vabh sa) of the world continues even in the case of a person who has the knowledge of the non-dual Self conveyed by the Upaniṣadic texts. If the illusion continues in spite of the fact that the verbal cognition ( abda-jñ na) arising from the Upaniṣads is clear, certain, and indubitable, it is because of the power of deep- rooted impressions of the beginningless illusion. … In order to counter these impressions, something more than the verbal cognition arising from the Upaniṣads is required. Ma ḍana says that repeated contemplation (abhy sa) on the content of the verbal cognition generated by the Upaniṣadic texts is necessary in order to root out the impression of the beginningless ignorance. As a result of the repeated contemplation (abhy sa), the impressions of the knowledge of the non-dual Self obtained from the 91 yato na kāma-prāptyā kāma-pravilayaḥ, api tu doṣa-paribhāvanābhuvā prasa khyānena. BrS p.30. tasmāt prasa khyānam evaikaḥ kāma-nibarha opāyaḥ, karma-vidhayas tu viparyaya-hetavaḥ. Ibid. 271  Upaniṣads grow and develop, become powerful and get stabilized in such a way that they are able to remove the impressions of the beginningless illusion and thereby bring about the manifestation of the real nature of the Self ( tma-svarūpa- virbhava).92 We should, however, note carefully that Ma ḍana’s problem with the propositional knowledge of the Upaniṣads was not that it was verbal. Allen Thrasher seems justified in claiming that even the final knowledge of Brahman attained through meditation would still be somehow verbal, just because Brahman itself was speech in nature. “The possibility is left open that to Ma ḍana's mind even the final, non-dual knowledge of Brahman is still verbal, because its object is Brahman, which is still abda.”93 For Ma ḍana, reality itself at the core was verbal, and it was so even when devoid of all mental constructs: the world was a false appearance of the Word. The problem was, rather, that the knowledge of Brahman from the Upaniṣads obtained through the threefold predication in which Brahman in its specific nature was a determinate description, a v ky rtha or a saṁsarga-bheda, one that is meaningless without mental constructs. It was not just a verbal cognition that may not necessarily be verbalized, but a full-blown propositional knowledge. It is eminently significant that both Śa kara and Sureśvara framed the problem in terms of Upaniṣadic sentences, not just verbal cognition of any kind. Ma ḍana’s solution to the problem of Brahman as a sentential reference or a definite description was to promote meditation in which all mental constructs would be overcome through the repetition of the propositional knowledge of the Upaniṣads. Śa kara’s solution in promoting the identity statements of the Upaniṣads as the focal point of liberation was very different, as we shall see in Chapter Eight, and it involved a mode of predication in which the propositional knowledge of the Upaniṣads would not amount to a definite description of Brahman. 92 Balasubramanian 1983:333. 93 Thrasher 1993:98. 272  Conclusion In conclusion, it may be worth our while here to reiterate the central claim of the doctrine of prasaṅkhy na: scriptural knowledge of Brahman or the Self is not enough for liberation, because of its being mediate and relational. That claim is easy to appreciate, and in some variation, it has been commonly made by religious men of diverse affiliations: direct personal experience is different from scriptural knowledge. Its epistemic significance can be viewed, for instance, through the lens of the Bergsonian distinction between knowledge through analysis or description and knowledge through intuition. Henri Bergson famously claimed that any object that is not known empirically can be known in two ways: through analysis that involves description of the characteristics of the object in virtue of its having elements in common with other objects, thus through analogy; or through intuition that is a “kind of intellectual sympathy by which one places oneself within an object in order to coincide with what is unique in it.”94 Think of a character in a novel, for instance, who can be described by the author in various complexity of detail, but of whose traits and actions we can know just in virtue of the fact that we know them as concepts. Appreciating fully or absolutely such a character would require us on our part to “coincide with his person,” to attain a form of sympathy in which we would experience his emotions, to assume a vision of things from his perspective rather than from the outside. If such a form of knowing would be possible, it would be in the domain of intuition. This is roughly what the proponents of prasaṅkhy na were claiming. Knowing Brahman through scripture meant knowing Brahman from the outside and through description by means of abstract concepts, a description that was possible in virtue of Brahman’s having characteristics 94 Bergson 1913:7. 273 that were shared with its finite products. Such description did not quite work even on the conceptual level, however, because Brahman’s characteristics such as bliss and consciousness were not really like the characteristics that we are empirically acquainted with. Ma ḍana’s description of the uncommon bird was only an imperfect approximation, because Brahman did not have bliss like the bird had silver wings: the bliss of Brahman does not have a full analogue. The negative characteristics of Brahman in the Upaniṣads would additionally convey that Brahman was not a kind of an entity in the same order of being as its products were. This simultaneous application of cataphatic and apophatic theology made of Brahman an eminently composite entity, a sentential reference, a v ky rtha, the knowledge of which absolutely required concepts or mental constructs. That is just how sentences that involve definite descriptions work: a particular genus is individuated by the predication of unique characteristics. Knowing Brahman directly and as an entity that is absolute or simple, beyond mental constructs, would require becoming Brahman, experiencing Brahman’s bliss, and assuming Brahman’s kind of awareness. The prasaṅkhy na doctrine, however, was that such propositional knowledge of Brahman from the Upaniṣads was the only window into knowing Brahman. The prasaṅkhy na meditation was not supposed to start from scratch in the pursuit of Brahman, nor function independently as an alternative method. The notion of Brahman or the Self formed from the Upaniṣads was true as far as truth in the sphere of illusion was possible. It was the continual meditation on this same propositional knowledge formed through analogy with what was known, and the prasaṅkhy na- v dins as Vedic theologians were sure to emphasize that the Upaniṣadic statements of Brahman or the Self were the sole pram ṇa for knowing Brahman. For some of them, meditation was a mode of the propositional knowledge through which such knowledge became clearer. For 274 Ma ḍana Miśra, the scripturally formed cognition of Brahman, pram ṇa as a result or phala, became the specific cause of the final non-dual cognition, pram ṇa as an instrument or the unique cause, karaṇa. The route to knowing Brahman fully was solely through the Upaniṣads. It is, of course, difficult to find an epistemological justification why the same cognition that is repeated over and again, all other things being equal, would just by itself transition into an intuition or direct experience. Ma ḍana Miśra’a point, however, was that meditation on Brahman reconstitutes the subject through purification: it removes impressions of ignorance, it undoes desire, and it stamps its own impressions. Bhart prapañca’s idea was much along the same lineŚ the continual immersion of the mind, the product of ignorance, in Brahman its cause would gradually transform the mind, brahmanize it back to the cause, in the same manner as a metal piece made of ore would decompose back to ore and, eventually, to soil. This was, then, the crucial prasaṅkhy na claim: meditation effectuated purity of the subject, and without such purity obtaining, the cognition of Brahman would never become immediate. 275  CONCLUSION I said that Kumārila’s two accounts of liberation need not be interpreted as a change in his own understanding, but that it is better to see them rather as an attempt to accommodate two very different understandings of both what liberation is and how it can be achieved. As we saw, his first account was tied to “Sā khya,” and we have reasons to believe that this was the systematic Sā khya which was organized around treatises such as the S ṅkhya-K rik Ś the same arguments against it reappear in the BS commentaries. However, Kumārila’s account had wider application and it could refer to any variation of the doctrine that discriminative knowledge—knowledge of the Self as distinct from matter—is in itself sufficient for liberation. And, at this point it would be useful, it seems to me, to think of the famous Sā khya- Yoga distinction in the Bhagavad-G t , where Sā khya is introduced by K ṣ a in the second chapter just before the beginning of the fratricidal war, precisely in the sense of knowing the true Self as different from the body, not an agent nor a patient of action, neither a murderer nor liable to murder. Throughout the G t this Sā khya is affirmed as a means of liberation, alternative to Yoga, to yield the same results as Yoga but inherently more difficult. We also saw that the result which discriminative knowledge of the Self was supposed to bring was isolation of the cognitive agent and absence of transitive awareness. One problem with Sā khya for Vedic theology could have been that it is not easy to comprehend how discriminative knowledge can be a means that is constitutionally processual. How does one practice knowing to be a Self which is different from the body? It also seems fair to me to say that the texts that promote such knowledge rarely make it clear what is knowledge as knowledge, as a result—knowledge as content of awareness that has the quality of certainty— and what is knowledge as practice, a procedure to arrive at certainty. 276  Another problem must have been that an isolated Self with no transitive awareness for all eternity cannot be appealing to theologians who have spent their lives pondering over how rice, milk, and wood, when arranged properly, can bring about eternal and unsurpassed happiness. When early Vedic theology—both Mīmāṁsā and Vedānta—encountered “Sā khya” in this sense of discriminative knowledge, it treated it as information that one can do things with, use it in one way or another for one or another result, in the same manner as the repurposing of yoghurt or kh dira wood. Kumārila, however, had committed himself to claiming that both dharma and mokṣa were solely in the province of the Veda, so in his first account he tried to accommodate a theory of liberation that had taken hold in the Vedic worldview but was not quite to his liking. He did not share its presuppositions, but he could work something out. So, what he did in the end amounted to turning Sā khya into a form of niṣk ma-karma-yoga, where the information about an eternal Self had engendered disinterest in the attainments promised in the Veda, but ritual practice was repurposed for attaining liberation. Kumārila’s second account, on the other hand, should be grouped with the Brahma-Sūtra doctrine of liberation, with which it shared all important notions. This doctrine was centered on the role of meditation. Liberation was to be achieved by meditative absorption in Brahman as one’s Self—the higher Self that is not liable to transmigration, and not by any form of intellectual understanding or knowledge qua knowledge. Meditation on Brahman was the means of liberation even for Vedāntins who were much closer ontologically to Śa kara, such as Bhart prapañca, Ma ḍana Miśra, and the prasaṅkhy na-v dins, which distinguished mediate knowing through scriptural cognition and direct knowing through meditative absorption. In the BS, knowing as knowing was largely procedural, insofar as it was important for forming the 277 meditation or ascertaining the meditational counterpart of oneself, in just the way that sacrifices are formulated: the process of liberation was meditation. In the doctrine of prasaṅkhy na, on the other hand, the veridical cognition of Brahman or the Self obtained from the Upaniṣads was genuinely important, but even then, liberation did not follow by knowing Brahman or understanding with full certainty what the Upaniṣads say: there was no liberation without meditation on Brahman. This summarizes my overall argument in chapters three through five: in pre-Śa kara Vedic theology, liberation was never a result just of knowing as knowing, whether that be the discriminating knowledge of the Self or knowledge of Brahman from the Upaniṣads. Liberation was a result either of ritual or meditation on Brahman. Persistent throughout the chapter was the role of ritual. We saw that Vedic theologians were not quite sure what the role of ritual in the pursuit of liberation was. For Mīmāṁsakas, it was supposed to prevent the creation of bad karma, but Kumārila thought it could also exhaust some of the old karmic stock. In his first account ritual was, really, the sole means. The Brahma- Sūtra turned the table on MīmāṁsāŚ ritual’s primary role was to nurture meditation, but it was also mandatory for those who do not pursue liberation. Bhart prapañca, coming from the background of the therapeutic paradigm where ignorance caused agency and ritual perpetuated it, promoted ritual from an assistant to meditation to an equal partner. The role of desire, k ma, was another item of negotiation. Kumārila’s first account was predicated on the absence of desire for the common Vedic attainments, but his second account clearly presupposed it. In the related Brahma-Sūtra account the desires related to the Vedic attainments became essential characteristics of Brahman, accomplished by Brahman’s mere will, and to strive after fashioning oneself in Brahman’s image meant securing the accomplishment of 278 these desires. The pursuit of liberation became tma-k ma, and these true desires of Brahman were upheld even by the likes of Bhart prapañca, who was, at the end of the day, a monist. Finally, the idea of liberation was development not only of mokṣa, but equally if not more so of svarga. In fact, the definition of niḥ reyasa as unexcelled felicity, nirati aya-pr ti or nirati ay nanda, placed svarga and mokṣa in the same category, that of the ultimate human attainment. Bliss was, of course, different thing to different theologians, but so was heaven. Particularly in the early Vedānta that was focused on the Brahma-Sūtra, liberation was much closer to heaven than to the liberation in the schools of the therapeutic paradigm. To quote Nakamura, it was “almost unparalleled in the writings of any Indian school.”1 The “big discovery” of the Upaniṣads was that one could attain and keep the attainments that the ritualists aspired for, if one could just tap into their incorruptible source, Brahman, which is free from faults and whose desires and resolves are ever true. 1 Nakamura 1983:531. 279  PART THREE: DHARMA, SCRIPTURE, AND THE GOOD OF MAN IN EARLY ADVAITA VED NTA INTRODUCTION In this part of the dissertation I will consider the themes that were front and center so far— scripture, dharma, and the good of man—in the doctrine of the theologian who provides the network for my whole undertaking, the great 8th-century Advaita Vedāntin Śa kara Bhagavatpāda. This investigation will complete the history of rethinking the nature of the Vedic canon on Śa kara’s part, a history that we need to know so we can finally tackle the notion of ved nta-v kya in the restricted sense in which it was used by Śa kara and Sureśvara, and understand why in particular tat tvam asi and ahaṁ brahm smi became the two mah -v kyas in Sarvajñātman’s formulation of the doctrine, and not the injunctions of ritual as in Kumārila’s first account of liberation, or the injunctions of meditation around which the brahma-vidy s in the BS were formed, or Bhart prapañca’s favorites tm nam eva lokam up s ta and vijñ ya prajñ ṁ kurv ta, or the prasaṅkhy na injunctions of meditation such as tmety evop s ta, or even descriptive statements of Brahman such as satyaṁ jñ nam anantaṁ brahma. Completing this history will require that we investigate two of the most troubled questions in Śa kara’s thought, and a host of other difficult issues related to them as well. The two questions concern respectively the nature of and the soteriological efficacy of ritual and meditation. The first question is, how precisely did Śa kara conceive of the relationship of what is commonly called “knowledge” and “action,” jñ na and karma, and of their possible cooperation in the pursuit of liberation? As we saw in the previous part, Vedāntin theologians before Śa kara unanimously advocated for some form of jñ na-karma-samuccaya, combination of “knowledge” and “action,” whereas Śa kara, as we have heard so often, notoriously rejected the very possibility of combining the two. Patrick Olivelle’s assessment in this regard can be 280 cited as the standard and commonly accepted scholarly viewŚ “Advaita, for example, claims that works in general and ritual actions in particular contribute nothing toward the attainment of liberation, and that knowledge is the sole cause of liberation.”1 Olivelle also correctly notes that, while Śa kara’s arguments about the impossibility of the performance of ritual presuppose the attaining of liberation while living, j van-mukti, and the absence of the vision of duality, he expects even the seeker after liberation to give up ritual, a case in which the non-duality argument cannot be advanced.2 When we delve into Śa kara’s works, however, the image that emerges is not at all crisp and clear. While I do not wish to quote anything yet, there are places where Śa kara explicitly endorses jñ na-karma-samuccaya in a way that seems quite compatible with the BS account but not with Bhart prapañca, yet denies that such a combination could be productive of liberation. There are also places where he explicitly affirms that ritual does contribute something to the attaining of liberation, contra Olivelle’s assessment, and leading, for instance, Allen Thrasher to wonder whether his “complex doctrine … on works and knowledge” was fully consistent.3 I will show by the end of this part that Śa kara had a complex yet crystal clear doctrine about the mutual relationship of “works” and “knowledge,” and that “works” do contribute something to the attaining of liberation. Understanding just “what” this contribution is and “how” it stands in relation to knowledge requires an understanding of the complex Mīmāṁsā ritual technology, with which we are at this point well-acquainted. The second central question that we will have to investigate is, what was Śa kara’s understanding of the nature and the role of meditation? We saw that the principal Vedāntic 1 Olivelle 1986:18. 2 Ibid, 33. 3 Thrasher 1979:122. 281 means of liberation, in the standardized BS account, in Bhart prapañca’s doctrine, and in the theology of prasaṅkhyana, was brahma-vidy that was practiced as a form of meditative absorption, up sana or dhy na. Śa kara’s radical rethinking of the Vedic canon involved not only the rejection of ritual, but of meditation as well. On this we may quote with profit Hiriyanna’s almost century-old evaluation: In holding such a view [that meditation plays no role in the genesis of right knowledge], Śa kara stands alone; and practically all the other Ved ntins reject this distinction between jñ na and up san , and admit an injunction in one form or another in respect of the knowledge of the self. Thus they fall into line with the M m ṁsakas, who hold that the main purpose of the Veda as a whole is to inspire activity by prescribing something to be achieved, and not merely to state matters of fact. The only difference between the Pūrva- and Uttara-k ṇḍas according to these Ved ntins, is while in the former what is prescribed is generally a sacrificial act, in the latter, it is meditation which is purely a mental act. Thus it is injunctive statements found in the Upaniṣads like tm v are draṣṭavyaḥ that are of primary importance and not assertive propositions like Tat tvam asi which only subserve them by furnishing the theme for the meditation prescribed. The meditation, if it is to be practiced, presupposes a knowledge of certain details such as the nature of tman—the object to be meditated upon. The purpose of statements like Tat tvam asi is merely to impart this knowledge and not directly to lead to self-realisation. The tman therefore is, in M m ṁs phraseology, the eṣa of the up san -vidhi.4 In arriving eventually at tat tvam asi as the mah -v kya, we need to see why Śa kara had this problem with meditation and meditative texts. More than that, we need to ask the same question that we asked related to ritual: was it really the case that meditation played no role whatsoever in the pursuit of liberation? We also need to understand the distinction between knowledge and meditation, especially because both terms are commonly denoted by jñ na when the question of combination of “knowledge” and “action” is discussed. In pursuing these two questions, my thesis will be that, while liberation for Śa kara was a result of knowledge qua knowledge, or knowledge as “matters of fact,” to use Hiriyanna’s turn of phrase, both ritual and meditation had causal contribution to the attainment of liberation, not, 4 Hiriyanna 1928:3-4. 282 however, under the samuccaya or “combination” model, but under the model of mediate causality or p ramparya. Under this model, ritual and meditation as processes and the scriptural texts that present their idealities were both meaningfully subsumed under the identity statements of the Upaniṣads as the focal points of the striving after liberation, although they could not be practiced simultaneously with knowledge. I do not wish, however, to discuss these questions one by one or by explicitly aiming at them. I want, rather, to continue narrating the story that has been enfolding so far in the dissertation, the story of scripture, dharma, and niḥ reyasa, by keeping in mind the prior developments of the plot and moving steadily towards the denouement, the birth of mah -v kya. The best way to do that is, it seems to me, to pursue holistically and in context one of these three ideas in Śa kara’s system, one which is quite well-known because of being prominently placed in his works, at or near the beginning of the commentaries to the Brahma-Sūtra and the Bhagavad-G t . It is the idea of dharma. Following the course of Śa kara’s understanding of dharma as it unfolds when we track all its implications will naturally lead us to considering the pertinent questions about scripture and the highest good, about knowledge, ritual, and meditation, but in the manner of a coherent story. Once we complete the full course, we will understand the precise relationship between knowledge and action, the role of meditation, why Vedāntic injunctive texts could not be the central Upaniṣadic statements, and, eventually, how both ritual and meditation were soteriologically efficacious and formed integral parts of the future mah -v kyas. At the end of the course, we will see that Śa kara replaced meditation on Brahman with reflection on the Upaniṣadic identity statements as the proper Vedāntic method of liberation. 283  The examination of Śa kara’s understanding of dharma, then, is the theme of this part of the dissertation. Two notes are in order before I move on to that task. First, as I just said, I want to approach Śa kara’s dharma by reading it in context, and the context is constituted by Śa kara’s fellow theologians, Mīmāṁsakas and Vedāntins. I also want to study it holistically: that is, I want to investigate its underlying presuppositions and its bifurcation as a unified enterprise, as that seems to me to be the only way to comprehend the intricacies of the knowledge-action-meditation complex properly. For presentation purposes, this means that, first, our old friends Kumārila and Prabhākara, Bādarāya a, Bhart prapañca and Ma ḍana Miśra, will keep popping up occasionally; and, second, that the part is conceived as a unit that should be read as such. Nevertheless, someone once said “A big book is a big evil,”5 and adamant not to neglect this stylistic maxim entirely, I will break the part into four chapters. In the first, Chapter Six of the dissertation, we will concern ourselves with dharma in general and with the path of engagement that terminates in the attaining of brahma-loka, which is not coterminous with liberation as it was in the BS. In Chapter Seven, I will introduce Śa kara’s doctrine of liberation, and we will pursue the repurposing of the dharma of engagement as the first leg on the path to liberation and a preparation for the dharma of disengagement. In Chapter Eight, I will focus on Śa kara’s notion of the identity statements of the Upaniṣads, their place in the Upaniṣads as a corpus, and their relation to other kinds of Upaniṣadic passages. That will prepare the ground for the dharma of disengagement in Chapter Nine, and the inquiry into Brahman through the identity statements as the properly Vedāntic path to liberation. With that, I will complete one of the two major arguments of this essay, namely that Śa kara’s promotion of knowledge as knowledge and 5 Cassirer 1956:preface. 284 the reasoned reflection on the identity statements of the Upaniṣads as the means of liberation and the highest human good was a novelty in Vedic theology of his day, rather than the norm. Second, understanding Śa kara’s dharma requires familiarity with his psychology and cosmology, particularly the first. Both have been, of course, examined thoroughly and by many, yet the emphasis on Śa kara’s monism and illusionism have lead scholars to disregard some details that I find crucial for appreciating his soteriology and approach to dharma. For this reason, I will have a few things to say about both, not in the form of a self-contained study, however, but at a point where it seems convenient, and without aspirations to be exhaustive in either. 285  CHAPTER SIX: GENERAL CHARACTERISTICS OF DHARMA AND THE PATH OF ENGAGEMENT In view of propriety, it follows that those past mothers who were causes of his happiness would appear in front of him at his mere will. For, it is not right that a yogin of a pure mind should have a desire for or association with those mothers who were the causes of his misery and a means of his being born as a domestic swine.1 One should not think that the results of ritual and meditation are permanent, like relation is amongst the Dravidians.2 “The face of truth is covered with a golden dish. Open it, O Pūṣan, for me, a man faithful to the truth. Open it, O Pūṣan, for me to see.” — The man who performs a combination of “knowledge” and “action” is praying to the sun at the time of his death.3 Introduction In conceptualizing the Veda as a canon, Śa kara’s entry point was that scripture had to satisfy some need of man, it had to be useful and have practical value. “The purpose of ruti is to instruct about human needs.”4 As should be obvious by now, Śa kara inherited this conviction from Bhā a Mīmāṁsā, and he also fully shared the worst nightmare of Vedic theology, that 1 mātaro janayitryo 'tītāḥ sukha-hetu-bhūtāḥ sāmarthyāt. na hi duḥkha-hetu-bhūtāsu grāma-sūkarādi-janma-nimittāsu māt ṣu viśuddha-sattvasya yogina icchā tat-saṁbandho vā yuktaḥ. ChUBh 8.2.2-9, VII.483. 2 bhāvanājaṁ phalaṁ yat syād yac ca syāt karma aḥ phalam | na tat thāsnv iti mantavyaṁ draviḍeṣv iva sa gatam. NaiS 3.93. 3 hira mayena pātre a satyasyāpihitaṁ mukham | tat tvam pūṣann apāv u satya-dharmāya d ṣ aye. BĀU 5.15.1. Translation Olivelle 1998Ś141. yo jñāna-karma-samuccaya-kārī so 'nta-kāla ādityaṁ prārthayati. BĀUBh 5.15.1, X.737. 4 puruṣārthopadeśa-paratvāc chrutīnāmś TUBh 1.11.4, VI.50. 286 upholding certain doctrines may render parts of the Veda purposeless: The Veda in its entirety must be useful in satisfying some human need. Śa kara also shared the broadest significance of the Mīmāṁsā understanding that there are goals of different kinds that men strive after, and that there are means suitable for the achieving of such goals, both felicitously going under the appellation of puruṣ rtha: there are, in other words, human needs that are puruṣ rtha, such as getting comfort and avoiding discomfort, and then there are things and actions, such as houses and house construction, that satisfy these needs, which are also puruṣ rtha.5 He participated fully, thus, in the technical jargon of Vedic theology, where puruṣ rtha could be either a determinative or a possessive compound, and where there were goals and results (s dhya, phala) and means (s dhana) of the principal and subordinate kind (pradh na, guṇa), all teleologically organized in a unique whole. The goals that men aspire for could be classified in two broad groups: enjoyment, which can be generally defined as obtaining comfort and avoiding discomfort, and liberation from the cycle of rebirth.6 Pursuing enjoyment is, of course, not restricted to men, and the specific characteristic that constituted the human condition was that men were entitled to pursuing goals by Vedic means, that is, by dharma. This specifically human pursuit was supposed to be directed to the two goals that we introduced in the previous part, abhyudaya or promotion and niḥ reyasa or the highest good. “For, only men are specifically entitled to promotion and liberation.”7 As we saw in the previous part, these two were by Śa kara’s time a common currency in Vedic theology, standardized in the Manu-Smṛti around the category of k ma and used even by Kumārila. 5 See BSBh 2.2.1 and 2.2.6. 6 This twofold classification is reconstructed from his rebuttal of Sā khya in the first p da of the second adhy ya of the BSBh, but is so commonplace in his intellectual milieu that it can be safely said that he endorses it. 7 manuṣya-graha aṁ viśeṣato 'dhikāra-jñāpanārtham. manuṣyā eva hi viśeṣato 'bhyudaya-niḥśreyasa-sādhane 'dhik tā ity abhiprāyaḥ. BĀUBh 1.4.9, VIII.120. 287 Nevertheless, Śa kara would be the first to think about the two thoroughly and to organize the Veda in the most minute detail around them. Śa kara’s understanding of dharma was also shared with Bhā a MīmāṁsāŚ dharma is the specifically Vedic means of achieving a human good of the two afore-mentioned kinds.8 By “specifically Vedic,” we mean that knowledge of dharma was available only in the Veda and not through any other reliable warrant. These were all, in fact, shared presuppositions of the whole field of Vedic theology. Prābhākaras, of course, though that the essential characteristic of dharma was obligation and not instrumentality, but they did not deny that dharma was a means of achieving goodŚ Śālikanātha proposed that this feature of dharma as being a means to a desirable result was how one could recognize dharma among many other actions that were in themselves causes of suffering.9 An upalakṣaṇa of a sort. When we consider, however, how Śa kara understood the workings of dharma and the Veda in securing human good, a gulf of difference appears before us and we witness a comprehensive rethinking of the nature of the two. This rethinking, moreover, was eventually responsible for the specifically Advaita understanding of liberation and for the formulation of the mah -v kya idea. We saw in the Second Chapter that Mīmāṁsakas considered that the specific sphere in which the Veda is epistemically valid was that of things that are not empirically available. Śabara famously saidŚ “For, (only) an injunction and not any sense organ is able to make known a thing 8 prā ināṁ sākṣād abhyudaya-niḥśreyasa-hetur yaḥ sa dharmaḥ. BhGBh Introduction, XI.2. This understanding of dharma was extended at the hand of Ma ḍana Miśra and became the norm in Bhā a-Mīmāṁsā and Advaita Vedānta to refer to all human action, not only that regulated in the Veda, as iṣṭa-s dhanat , a means to something desirable. On this, see David 2013 for a most lucid account. 9 svabhāvena hi karmā i dukhotpāda-hetu-bhūtāni. teṣu kāryatvāvagamaḥ phala-sādhanatāvagama-nibandhanaḥ. VAM 2.9, p.37. 288 that is past, present or future and is subtle, hidden, remote or suchlike in kind.”10 This was a quite broad definition, but its implications for Mīmāṁsakas were not at all straightforward. The primary and independent validity of the Veda was restricted to ritual action, or, to be specific, to an otherwise unknown causal relationship between ritual action of a certain kind—offering, pouring, giving—and a future state of affairs desirable to men. This specific causal relationship was fully unknown from other sources. The teleology of the sacrifice and its expression in the Veda was, of course, predicated on the use of common things—clarified butter, rice, wooden posts, sacrificial animals, fire—but these things were empirically available, otherwise known, and the Veda was not a reliable warrant on their nature or characteristics. Its validity concerning them was exhausted in what Mīmāṁsakas called itikartavyat , details of sacrificial procedure, namely how these common things should be employed in sacrifice so that the ritual performance would bring the desired result. So, although Śabara said that matters of fact past and present are in the domain of the validity of the Veda, this was a very restricted and dependent validity. We also saw that related to the question of the validity of the Veda was the issue of language. Mīmāṁsakas claimed that only the Vedic sentences of the injunctive kind had independent validity. We may illustrate this with Śabara’s interpretation of the case that is now well-known to us, in which a substance assumes the role of the primary element in a ritual performance because its use brings a specific result: yoghurt for heaven, kh dira wood for virility. Śabara claimed that even in such cases, the character of the Vedic sentence was not existential or informativeŚ “The injunctive ending establishes the relation between kh dira wood and virility; it does not state anything existing in the present.”11 Even in such cases, then, there 10 codanā hi bhūtaṁ bhavantaṁ bhaviṣyantaṁ sūkṣmaṁ vyavahitaṁ viprak ṣ am ity evaṁjātīyakam arthaṁ śaknoty avagamayitum, nānyat kiṁcanendriyam. MSŚBh 1.1.2, I.13 11 vidhi-vibhaktiḥ kuryād iti vīrya-khādira-sambandhasya vidhātrī, na ca vartamānāpadeśinī. MSŚBh 4.3.3, IV.1247. 289 had to be an injunction that would establish the causal relationship, and a sentence without such an injunction was, strictly speaking, defective. An often-quoted proverbial verse saysŚ “In all the Vedas, the following five words are sure signs of an injunction: kury t (should make), kriyeta (should be made), kartavyam (ought to be made), bhavet (should become), sy t (should be).”12 This was all quite uncontroversial across the whole field of MīmāṁsāŚ The Veda commands and intimates some unknown causal relationship. However, the fault line in Mīmāṁsā appeared over the question of what has the epistemic priority in Vedic commands: is it the command itself comprehended first, or is it the causal relationship? Thus, Prābhākaras on the one hand argued that the Veda first commands and then informs. Ma ḍana Miśra expressed this both eloquently and succinctlyŚ “While conveying an obligation, the injunction intimates matters of fact as well.”13 The process of knowing from Vedic linguistic utterances was understanding an obligation as a unitary idea from the whole sentence or larger piece of text first; only then and through the command could one proceed to knowing the meaning of smaller linguistic units and any matter-of-fact information. Kumārila, on the other hand, started from knowing the meaning of words and smaller linguistic units first, where the awareness that one had to do something would appear only after one had understood some amount of individual information and comprehended the causal relationship between the ritual performance and the future state of affairs. In some cases, one also had to be convinced that the proposed ritual was good and beneficial solely through understanding non-injunctive texts, the arthav da passages from the Brāhma as. Kumārila’s insistence that a thinking man does not do anything without understanding that an action is beneficial opened in a sense a can of worms in Vedic theology. It 12 kuryāt kriyeta kartavyaṁ bhavet syād iti pañcamam | etat syāt sarva-vedeṣu niyataṁ vidhi-lakṣa am. In MSŚBh, ibidś BĀUBh 1.3.1, VIII.40. See Devasthali 1959Ś177. 13 kāryam artham avagamayantī codanaiva bhūtādikam apy artham avagamayati. BrS p.74. 290 became possible to askŚ “Well, what if I do not see this as beneficial?” It also provided the starting point for Śa kara’s rethinking of pram ṇa, dharma and the Veda. Dharma and the Validity of the Veda Śa kara’s epistemology started with the assumption that knowledge of any kind should result in some good of men. This is a principle that holds good across the board, both in ordinary life and in matters in the domain of scriptureŚ “The attainment of good depends on understanding things as they are, just as in normal life. He who understands things rightly in the world attains the desired and avoids what is not desired, not otherwise. Likewise, the attaining of the good is possible when the meaning of scriptural passages has been understood correctly, not otherwise.”14 What pram ṇa needs to accomplish, therefore, is correct understanding of the nature of things so that they can be used or avoided. Only after something has been known properly can there be considerations whether one should do something or not. Take, for instance, the practical need to distinguish between people as friends or enemies. The action of differential treatment that is directed towards attaining some good by associating with friends or avoiding a misfortune by staying away from adversaries is absolutely predicated on figuring out who is one’s friend and who is one’s enemy. Beneficial action normally does not happen before this is ascertained. Another of Śa kara’s favorite examples is poisonŚ the need to avoid poison and avert its harmful effect is predicated on the ability to recognize poison.15 Further, to understand if one is a friend or an enemy or if something is poison, one needs to know what these things are. 14 BĀUBh 1.3.1, VIII.37Ś aviparītārtha-pratipatteḥ śreyaḥ-prāpty-upapatter lokavat. yo hy aviparītam arthaṁ pratipadyate loke sa iṣ aṁ prāpnoty aniṣ ād vā nivartate, na viparītārtha-pratipattyāś tathehāpi śrauta-śabda- janitārtha-pratipattau śreyaḥ-prāptir upapannā, na viparyaye. 15 Both illustrations are used in BĀUBh 1.3.1. 291  This feature of pram ṇa is so basic that it is not even restricted to humans. Men proceed to act only after ascertaining if something is good or bad in the same way as animals to. Those of developed minds run away from fierce-looking, abusive and armed men when sense experience obtains because its causal conditions are met, just as an animal runs away from a man with a stick in hand, thinking “he wants to kill me,” but rushes towards someone carrying green grass.16 Śa kara, thus, agreed with Kumārila that man is buddhi-pūrva-k rin, someone who understands first and acts later,17 and was opposed to the Prābhākara doctrine of obligation. The criterion of any reliable warrant is whether it produces a cognition which is both certain and fruitful. If stra is to be a pram ṇa at all, neither the positive nor the negative criteria of the Prābhākaras are relevantŚ the cognition of something being one’s inevitable duty cannot be epistemologically primitive, nor are mere descriptions of things eo ipso disqualified from validity. For a scriptural statement to be a pram ṇa, it first needs to engender a cognition that passes the test of certainty, and then it needs to be conducive to some good.18 Both must obtain before the question of obligation can even be posed. As we shall see below, there are hosts of other factors that need to obtain as well before one would proceed with acting upon something that the Veda commands. 16 “Because, there is no difference from the animals. Just as animals have a cognition of something averse from sense data when their senses are related to sense object and turn away from it, or go towards it when it is appealing—when perceiving a man with a stick in his hand they start running away, thinking ‘he wants to kill me.’ but noticing one with green grass in hand go towards him—likewise even men with developed minds seeing fierce- looking, strong, abusive men, sword in hand, run away from them, but approach the opposite. Therefore, the function of objects of knowledge and reliable warrants is the same for man and animals.” paśv-ādibhiś cāviśeṣāt. yathā hi paśv-ādayaḥ śabdādibhiḥ śrotrādīnāṁ sambandhe sati śabdādi-vijñāne pratikūle jāte, tato nivartante, anukūle ca pravartanteś yathā da ḍodyata-karaṁ puruṣam abhimukham upalabhya ‘māṁ hantum ayam icchati’ iti palāyitum ārabhate, harita-t a-pūr a-pā im upalabhya taṁ pratyabhimukhī bhavanti; evaṁ puruṣā api vyutpanna- cittāḥ krūra-d ṣ īn ākrośataḥ khaḍgodyata-karān balavata upalabhya tato nivartante, tad-viparītān prati pravartante. ataḥ samānaḥ paśv-ādibhiḥ puruṣā āṁ pramā a-prameya-vyavahāraḥ. BSBh 1.1.1, I.3-4. 17 This lexeme so dear to Kumārila is used right at the opening of the BSBh. 18 “What constitutes the validity or invalidity of a sentence is not whether it talks about things or action, but if it produces certain and fruitful knowledge. If it does, then the sentence is valid; if it doesn’t, it is not.” na vākyasya vastv-anvākhyānaṁ kriyānvākhyānaṁ vā prāmā yāprāmā ya-kāra am. kiṁ tarhi? niścita-phalavad- vijñānotpādakatvamś tad yatrāsti tat pramā aṁ vākyamś yatra nāsti tad apramā am. BĀUBh 1.4.7, VIII.112. 292  It is for this reason that the Veda becomes a reliable warrant: the ability to produce certain and fruitful understanding. Although the Veda is divided per the kind of good that it provides for men—prosperity and liberation—as a pram ṇa it begins to function in both spheres equally: Matters of knowledge and meditation (vidy ) are similar to matters pertaining to ritual. That there is such a thing as a certain ritual, for instance Darśa-pūr amāsa, which has specific details of procedure and which is performed in a particular sequence, is a supersensible matter, not knowable empirically but communicated by Vedic texts. Likewise, things such as the Supreme Self, the Lord, and various divinities, as well as characteristics such as not being gross, being above hunger and thirst and the like, are also communicated by Vedic texts. Because they are supersensible, they must be as described [by the Veda.] There is no difference in the way texts about action and texts about knowledge impress themselves upon the understanding.19 Let us pause here and review again the Prābhākara account of Veda as pram ṇa. Mīmāṁsakas famously claimed that individual words in the Veda have the same meaning that they have in normal life.20 Words denote universals and in themselves stand for the categories of the world, pad rtha, things denoted by words. The Veda as a form of knowing from linguistic utterances has nothing to do with words (pada): the domain of the Veda are sentences or texts (v kya), because it is at the level of sentences that something new and empirically unknowable can be communicated. “Fire” is just fire and “butter” is butter, but a statement about the pouring of butter in fire has its own sentential denotation. While the Veda is not authoritative on fire as a category of the world knowable through perception, a sentence can express the empirically unknowable. Prābhākaras pushed this to the limit by claiming that, cognitively, the meaning of 19 kriyārthaiś cāviśeṣād vidyārthānām. yathā ca, darśa-paur amāsādi-kriyā idam-phalā viśiṣ aitikartavyatākā evaṁ- krama-prayuktā gā ca—ity etad alaukikaṁ vastu pratyakṣādy-aviṣayaṁ tathā-bhūtaṁ ca veda-vākyair eva jñāpyate. tathā, paramātmeśvara-devatādi-vastu asthūlādi-dharmakam aśanāyādy-atītaṁ cety evam-ādi-viśiṣ am iti veda- vākyair eva jñāpyate—iti alaukikatvāt tathā-bhūtam eva bhavitum arhatīti. na ca kriyārthair vākyair jñāna-vākyānāṁ buddhy-utpātakatve viśeṣo 'sti. BĀUBh 1.3.1, VIII.39. 20 prayoga-codanābhāvād arthaikatvam avibhāgāt, “(Words must have) one meaning (in the Veda as in the world) because (only in this way) injunctions of employment (of things in sacrifice) is possible, (and) because no distinction (between the two is seen).” MS 1.3.30. See also Devasthali 1959Ś2-3 and Clooney 1990:131-37. 293 Vedic sentences is prior to the meaning of their individual words: they have their own denotation before their smaller linguistic units obtain individual denotation. This denotation is an obligation: I must do something. Sentences are naturally injunctive, and when they are not, the natural injunctive force is blocked by an imposition of some other mode, for instance the indicative, such that the result is an accidental existential statement.21 Now, aside from the question of cognitive priority or posteriority of sentences and smaller meaning-bearing units, Śa kara and his followers agreed that the Veda operates in the sphere of sentences, not individual words. This is not always obvious, since Śa kara does talk about categories that are in the domain of the Veda; these are, however, insofar as the Veda has something to say about them, always ellipses that need to be elaborated in sentences, as we shall see somewhat later. However, for the Veda to be meaningful, it must inform about things as they are—the attaining of any good depends on that—failing which it could not be a reliable warrant. The sentential cognition produced by any Vedic text is, then, about some supersensible reality of a substantive kind: ritual, the Self, God, etc. What is meaningful in a Vedic text are not the individual words vara or dar a-pūrṇam sa, but how these are described. For the Veda’s being a pram ṇa, it is simpler to accept that Vedic texts denote things, not action, and that the natural verbal mode is not the injunctive but the indicative. It is the injunctive that is imposed independently of the sentence meaning, if there are circumstances conducive for that. Let us consider one of the favorite Mīmāṁsā sentences, dar a-pūrṇam s bhy ṁ svarga- k mo yajeta. The natural meaning of the sentence is not the obligation that one must perform the Darśa-pūr amāsa ritualŚ it could not possibly be, because the cognition of obligation as the sentential meaning would not obtain if one did not have the requisite qualification (adhik ra) for 21 See, for instance, Sarma 1990:25. 294 the performance of the ritual. In this case, the qualification would at the least involve having the desire for heaven, and in its absence not only would one not perform the ritual on the Prābhākara account: one would not understand that the sentence pertains to oneself, and it would fly right by like flatus vocis. What the sentence is about, then, is the Darśa-pūr amāsa ritual as a supersensible reality. The cognition that this ritual should be performed comes later, if the requisite qualification is present.22 Scriptural realities such as the Supreme Self are expressed in existential statements that denote qualified entities, and ritual is of that kind as well.23 Sureśvara, in fact, claimed that obligation as the sense that a ritual had to be performed was itself a category, pad rtha, a matter of words and not sentences, a personal element over and above the scriptural cognition of the supersensible reality: An injunction is fruitful when a category of observance is accepted as present over and above the understanding that is a result of the sentence, as in the case when something that has the nature of effort is understood over and above the cognition that corresponds to Agnihotra. In this particular case [ tmety evop s ta] there is no such thing.24 This is the rationale that made it possible for Śa kara to claim that there are two kinds of dharma, one proceeding by way of engagement and another by way of disengagement, and that the Veda is the reliable warrant on both because they are equally supersensible. To sum up, for 22 “An injunction is fruitful when human engagement is understood over and above the cognition produced simply by hearing the injunctive sentence. For instance, in ‘He who wants heaven should sacrifice by the Darśa-pūr amāsa ritual,’ the performance of this ritual is not equivalent to the cognition produced by the injunctive sentence about the ritual, because such performance depends on entitlement.” tatra hi vidheḥ sāphalyaṁ yatra vidhi-vākya-śrava a- mātra-janita-vijñāna-vyatireke a puruṣa-prav ttir gamyate. yathā darśa-pūr amāsābhyāṁ svarga-kāmo yajeta ity evam ādau. na hi darśa-pūr amāsa-vidhi-vākya-janita-vijñānam eva darśa-pūr amāsānuṣ hānam. tac cādhikārādy- apekṣānubhāvi. BĀUBh 1.4.7, VIII.108. 23 “Likewise, why wouldn’t there be a union of words that are parts of sentences that propound the Supreme Self, the Lord, etc. and include the existential verb, through the qualificandum-qualifier relationship?” tathā asti-pada- sahitānāṁ paramātmeśvarādi-pratipādaka-vākya-padānāṁ viśeṣa a-viśeṣya-bhāvena saṁhatiḥ kena vāryate. BĀUBh 1.3.1, VIII.40. 24 vidher hi tatra sāphalyaṁ yatra vākyottha-bodhataḥ | vyatirekād anuṣ heyaḥ padārthaḥ kaścid iṣyate || yathāgnihotra-yāthātmya-vijñāna-vyatirekataḥ | prayogātmā p thak tād neha kaścid apīkṣyate. BĀUBhV 1.4.810-11. 295 the Veda to be a reliable warrant, it must primarily inform. It must inform about that which is empirically unavailable, and the knowledge must be useful to men for getting what they want and avoiding what they don’t. What makes the being of knowledge is not that something has to be performed, because knowledge is about correspondence to things. – What is it then? – That it is known through a reliable warrant. Nor does the specific cognition obtain because it is about performance. – Why then? – Just because it is produced by a Vedic text. Since such is the being of realities that are understood from Vedic texts, if [the cognition] is qualified by performance, one will perform. If it is not qualified like that, s/he will not.25 The same theme is expressed a bit differently in Śa kara’s claim that knowledge is contingent on things, whereas action is contingent on man, or, more specifically, on the agent. This is a quite common theme in Śa kara’s writings and prominently placed near the beginning of the BSBh and in the first prakaraṇa of US.26 This distinction is commonly drawn in the context of knowledge of Brahman versus ritual action, but is evidently more basic and refers to knowledge of any kind. Knowledge of Brahman (brahma-vidy ) or scriptural knowledge (vidy ) is determined by the object to which it corresponds, in the same way as perceptual awareness is determined by the object. Knowledge is universally the result of a reliable warrant and is not a matter of human choice. If the Yugo car is in the field of my vision and the causal factors of perceptual awareness obtain, I don’t have the option to un-see the Yugo or fancy that it is a Ferrari. The instant I do that, my awareness is no longer knowledge and no longer caused by a 25 jñānasya tathā-bhūtārtha-viṣayatvāt na hy anuṣ heyatvāt tathātvamś kiṁ tarhi? pramā a-samadhigatatvāt. na ca tad-viṣayāyā buddher anuṣ heya-viṣayatvāt tathārthatvamś kiṁ tarhi? veda-vākya-janitatvād eva. veda- vākyādhigatasya vastunas tathātve saty anuṣ heyatva-viśiṣ haṁ ced anutiṣ hati; no ced anuṣ heyatva-viśiṣ haṁ nānutiṣ hati. BĀUBh 1.3.1, VIII.39-40. 26 “Therefore, knowledge of Brahman is not dependent on human exertion? – What is it dependent on, then? – It is dependent solely on the object, just like the knowledge of a thing that is in the sphere of what is perceptible.” ato na puruṣa-vyāpāra-tantrā brahma-vidyā. kiṁ tarhi, pratyakṣādi-pramā a-viṣaya-vastu-jñānavad vastu-tantraiva. BSBh 1.1.4, I.22-3. “’I am the agent, this is mine,’ this is how action functions. Knowledge is dependent on the thing, whereas injunction is dependent on the agent.” ahaṁ kartā memedaṁ syād iti karma pravartate | vastv-adhīnā bhaved vidyā kartr-adhīno bhaved vidhiḥ. USP 1.1.13 296 pram ṇa. Action, on the other hand, is not really a matter of pram ṇa, because it invariantly involves option: any odd action can be done, not done, or done otherwise than it should be done, because it depends on human effort, on what man thinks, or on the agent in general. So, if one can reply to svarga-k mo yajeta, “I am not going to do it,” yet has a clear cognition of the sentential meaning, the knowledge that the sentence provides cannot be that a ritual must be performed. The Veda as a pram ṇa, thus, cannot be injunctive at its core: no injunction will make my Yugo a Ferrari nor impose a sense of obligation solely on its own. All of this was echoed by Ma ḍana Miśra in his Brahma-Siddhi: That knowledge which is solely from linguistic utterance is not liable to injunction, because it obtains without it, as in the case of understanding ritual. Just as the understanding of the sentential meaning of the Vedic statement ‘He who wants heaven should sacrifice’ obtains solely from the validity of the statement, likewise an understanding of the Self obtains from the validity of texts that propound the essential nature of the Self, and it does not require an injunction. When the reliable warrant is complete with its causal conditions, not even desire for knowing is required on the part of man – even those who do not want to, do understand – how much less so an injunction.27 This insistence on knowing as being solely concerned with how things are and the distinction between knowledge as objectively determined and action as subjective is so basic to Śa kara’s system that it functions as a classificatory device for everything in the Veda. Crucially, as I said above, it determines what gets to be labeled as one kind of dharma and what as another. It will continue haunting us throughout the dissertation. 27 tatra yac chabdād eva jñānaṁ tad avidheyam, vidhim antare a bhāvāt karmāvabodha-vat. yathaiva svarga-k mo yajeta iti śruta-vākyasya vākyārthāvabodhas tad-vākya-prāmā yād eva bhavati, na vidhy-antaram apekṣate, tathātma-svarūpābhidhāyi-vākya-prāmā yād ātmāvabodho vidhy-anapekṣaḥ sañjāyate, sati pramā e tat- sāmarthyenaiva prameya-bodhotpatteḥ. na khalu sopakara e samagre pramā e satīcchāpi puruṣasyāpekṣyate jñānaṁ prati, aniṣ ānām apy avabodhāt, prāg eva vidhiḥ. BrS p.74-5. 297  The Dharma of Engagement and Disengagement The nature of the Veda as a whole, then, is to be informative, not injunctive.28 In general, the scope of the Veda is dharma, a means to obtains something desirable. We saw in the introduction that what the Veda can provide for men through teaching dharma is broadly divided into prosperity or promotion, and liberation. These two respectively determine the content of the Veda, what the Veda is informative about: for the attaining of prosperity, the Veda teaches action, whereas for the attaining of liberation, it teaches knowledge of Brahman and dispassion: The Lord, creating the world and intent on its upkeep, first projected Marīci and the other patriarchs and taught them the Vedic dharma that is characterized by engagement. Afterwards, he created the four Kumāras and taught them the dharma characterized by disengagement, that is, by knowledge and dispassion. For, there are two kinds of dharma that are taught in the Veda, one characterized by engagement and another one characterized by disengagement.29 Dharma, then, can be of two kinds: characterized by engagement, pravṛtti-lakṣaṇa, and characterized by disengagement, nivṛtti-lakṣaṇa. Two notes are apposite before we proceed. We should make it clear right now that the twofold division itself was not Śa kara’s own inventionŚ Mīmāṁsakas themselves often described Vedic prohibitions as deterrents from action (nivartaka) and the action described in them as something that should not be done (nivartayitavya).30 Furthermore, we saw in the previous chapter the same distinction in the Manu-Smṛti, although it was associated explicitly with action, karma, not with dharma. However, Śa kara’s presentation of what the dharma of disengagement involved was thoroughly 28 BĀUBh 1.4.10, VIII.127Ś jñāpakaṁ hi śāstraṁ, na kārakamś BĀUBh 2.1.20, VIII.257Ś śruter jñāpakatvātś BĀUBh 3.9.28.7, IX.488Ś jñāpakatvād vacanānāmś TUBh 1.11.4, VI.48Ś jñāpakatvād vacanasya. vacanaṁ nāma yathā- bhūtasyārthasya jñāpakam, nāvidyamānasya kart . 29 sa bhagavān s ṣ vedaṁ jagat, tasya ca sthitiṁ cikīrṣuḥ, marīcyādīn agre s ṣ vā prajāpatīn, prav tti-lakṣa aṁ dharmaṁ grāhayām āsa vedoktam. tato ’nyān ca sanaka-sanandanādīn utpādya, niv tti-lakṣa aṁ dharmaṁ jñāna- vairāgya-lakṣa aṁ grāhayām āsa. dvividho hi vedokto dharmaḥ, prav tti-lakṣa o niv tti-lakṣa aś ca. BhGBh Introduction, XI.1. 30 See, for instance, MSŚBh 1.2.10, 1.3.33. 298 original and amounted to claiming that it was, in fact, more basic to the Veda, as we shall in the in Chapter Nine. Second, we need to be aware that Śa kara talks about dharma in two ways.31 In one of them, dharma is strictly in the domain of Vedic ritual action in several of its kinds and invariantly related to prosperity, abhyudaya. It is common for Śa kara to juxtapose this use of dharma to Brahman as the two spheres of the two inquiries respectively, dharma-jijñ s and brahma-jijñ s , and argue how they are oh so very different. Such is, for instance, the case at the beginning of the BSBh, particularlyŚ “Knowledge of dharma results in prosperity and such knowledge is dependent on performance. On the other hand, knowledge of Brahman results in the highest good and it is not dependent on performance.”32 This distinction between dharma and Brahman hinges on the inherited difference in the two exegeses on the crucial question of Vedic theology: what is the element in the respective pursuit from which the final good comes, the pradh na factor. We saw that Mīmāṁsakas argued that this is generally ritual action, and exceptionally substances such as yoghurt, both of which were called dharma by Kumārila. In the previous chapter, we also saw that Bādarāya a replaced dharma with Brahman, the repository of all desires. In this sense, dharma and Brahman are comparable and mutually excluding categories, and they seem to have been so from the very start of the two exegeses. Brahman, however, was never theorized as an instrumental cause, a s dhana, in the attainment of the desired goal. It was more like the Aristotelian causa finalis, that thing which one strives to become. Dharma, on the other hand, was by and large coextensive with the ritual action (karma) that organizes teleologically different elements such as substances and mantras in 31 That is, in two relevant ways for the purposes of his soteriology, since he also participates in the discourse of dharma as essential properties of things throughout the BSBh. 32 abhyudaya-phalaṁ dharma-jñānam, tac cānuṣ hānāpekṣam; niḥśreyasa-phalaṁ tu brahma-jñānam, na cānuṣ hānāntarāpekṣam. BSBh 1.1.1, I.6. 299 a series of individual actions in the pursuit of the desired goal. In that sense, its Vedāntic counterpart was not Brahman, but vidy , and vidy was explicitly organized on the model of dharma, with a central injunction and the whole shebang. Both dharma/karma and vidy were instrumental causes, s dhanas. Śa kara’s take on vidy was a complicated matter, as we shall see, but it was still modeled after dharma, namely a process based on scripture, organized around a central text, and involving practices arranged in a coherent whole on a clear principle. In short, it was a means (s dhana) for a goal (s dhya). In this perspective, the same deep structure of dharma was operative throughout the scope of Vedic theology both on the Bhā a and the Advaita accounts, and it was but natural for Śa kara to rework the Manu-Smṛti classification: anything in the Veda that is a means to something desirable but not available by ordinary devices, that involves some form of practice and that has the same deep teleological structure, never mind how that structure is respectively achieved, is “dharma, itself the cause of prosperity or the highest good for living beings.”33 In general, the two types of dharma are relatable to two types of sentences in the Veda, indicative and injunctive, though we must bear in mind that whether a statement will be taken as an injunctive, and if so, what its precise injunctive force will consist in, is personally contingent. There are some more details in the first category and one significant gray area between the two, as we shall see shortly, but which of these two general types of dharma one will get to practice is contingent on what qualification or adhik ra one has, and this in its turn is related to what specific desire, k ma, prompts one’s striving. The pair of k ma-adhik ra is also responsible for how the knowledge obtained from the Veda becomes personally inflected. Thus, the category of desire comes in large focus with Śa kara and his followers. 33 prā ināṁ sākṣād abhyudaya-niḥśreyasa-hetur yaḥ sa dharmaḥ. BhGBh Introduction, XI.2 300  a kara’s Psychology and the Human Condition We saw throughout this dissertation that the notion of dharma in Vedic theology was tied to goods desirable to men, puruṣ rtha, and I opened this chapter with Śa kara saying that only men were specifically qualified for the two kinds of good which dharma typically brings, prosperity and liberation. At this point it becomes necessary for us to define precisely the human situation that Śa kara has in mind so that we can see what the practice of dharma involved, and to do that we must touch upon Śa kara’s psychology. Bearing in mind Śa kara’s well-known absolute monism, we must assume that there are certain cosmological categories of Being that somehow obtain—it is impossible to define the human situation without them—but we do not need to worry about their relation to Brahman. The individual Self in Śa kara’s system is a complex product that is built on an initial interaction between the real Self, one and only for everyone, and the so-called intellect, internal organ, or the mind (buddhi, antaḥ-karaṇa). The real Self is, essentially, nothing more than the category in virtue of which it is possible to have phenomenal consciousness of any kind. Śa kara quite often compares this Self to sunlight, the necessary factor of any perceptual awareness that is essentially formless, but assumes all kinds of forms contingent on the shapes that it illuminates. We will take the following definition from the Upade a-S hasr as exhaustive: tman is the self-effulgent perception, the seeing, internally existing and actionless. It is the witness which is directly cognized and interior of all, and the Observer which is constant, attributeless and non-dual.34 We need to bear in mind that this pure Self is not the subject of conscious experiences, because any cognitive act involves a distinction between a subject, an object, an instrument, and cognitive content, and cancels monism: the non-dual Self cannot participate in these and still 34 upalabdhiḥ svayaṁ-jyotir d śiḥ pratyak-sad-akriyaḥ | sākṣāt sarvāntaraḥ sākṣī cetā nityo 'gu o 'dvayaḥ. USP 1.18.26, translation Mayeda 2006b:174-5. 301 remain non-dual. The Self is what makes subjectivity possible. Insofar as Śa kara describes it as the witness, it is not itself what cognizes anything, but is the accommodating factor, the seeing behind the seer.35 The pure Self, thus, is the awareness that ever obtains but is never transitive. The subject properly speaking, the one that has cognitions, is a reflection of the pure Self in a set of adjuncts, up dhis, the crucial among which is the intellect or buddhi.36 This buddhi can be defined as the evolute of Brahman in which cognition (vijñ na) in general takes place. The pure Self is the only Self, but it is not one that can have transitive awareness of itself. The intellect, owing to its proximity to the pure Self in the evolution from Brahman, becomes the locus in which a sense of Self can obtain. Śa kara illustrates the relationship between the Self, the intellect, and the sense of Self with the reflection that appears when a face is placed in front of a mirror.37 The sense of Self that is like the image in the mirror is variously called ahaṅk ra, aham-pratyaya, asmat-pratyaya, tm bh sa etc, and becomes the basis on which the individual Self is eventually built.38 The Self is not its reflection, but becomes identified with it. It may also be figuratively said to be under illusion, thinking oneself something which it is not, if we understand that this thinking does not happen in the Self itself, but is accommodated by its light. The reflection is neither a property of the face nor of the mirror, but it is dependent on both, insofar as it can obtain only if both are present. It does not, however, obtain necessarily: it is accidental because the face must be in front of the mirror for one to think, “This is me.”39 In a different sense, it is a necessary relationship for there to be cognitive subjectivity at all, because the intellect is not a conscious principle—it is that thing which is the locus of cognition, but is 35 See BĀUBh 1.4.10. 36 See BĀUBh 1.4.7. 37 USP 1.18.43. 38 Throughout the USP, particularly in 1.4 and 1.18. The asmat-pratyaya is the term used famously right at the beginning of the BSBh. 39 Cf. USP 1.18.39: dvayor eveti cet tan na dvayor evāpy adarśanātś “If it be said that the reflection is a property of (a combination) of the two, we say no, because it is not seen even when the two are present.” 302 itself not conscious of anything—whereas the pure Self is not an agent. So, the properties of the one are placed over the other: the consciousness of the Self is superimposed over the intellect so that there can be a conscious experience, whereas the cognitive agency that involves the dualities of subject, object, instrument, and cognition that belong to the intellect are superimposed over the Self.40 Because the intellect is the place where the reflection of the Self obtains and is located, the first adjunct of the tman, the individual Self is commonly called the vijñ n tman, the Self of cognition, and is typically distinguished from the pure, inner Self called the pratyag- tman or param tman. The above account is based primarily on the Upade a-S hasr , and it seems to me that it is the necessary starting point as it provides the clearest idea of the focal point around which the individual Self is constructed, namely the sense of Self or aham/asmat-pratyaya: it is a reflection of the Self in the intellect, a reflection that relates the two. We can now broaden the presentation by drawing from the BSBh and the BĀUBh. What is superimposed over the pure Self is not just cognitive agency: it is agency in general, namely the complex that involves action (kriy ), its contributory factors (k rakas) such as the agent, specifically identified with the reflection of the Self around which the vijñ n tman is constructed, instruments, and results.41 This is most evidently instantiated in the case of cognition: there is an agent or cognitive subject, jñ tṛ, which is the sense of Self reflected in the mirror of the intellect ( tm bh sa); there is an object to this subject that includes anything that might become an object of awareness, from external things to internal states of any kind, called by Śa kara yuṣmat-pratyaya, ”the notion of Youś” there are the 40 US 1.18.65. Also, BSBh 1.1.1, I.5: evam aham-pratyayinam aśeṣa-sva-pracāra-sākṣi i pratyag-ātmany adhyasya taṁ ca pratya-ātmānaṁ sarva-sākṣi aṁ tad-viparyaye āntaḥ-kara ādiṣv adhyasyati: “Likewise, superimposing the internal organ that bears the sense of Self over the internal Self, the witness of the modifications of that bearer, one proceeds to reversely superimposing the internal Self over the internal organ.” 41 See BĀUBh 1.4.17, VIII.159Ś svābhāvikyā svātmani kartr-ādi-kāraka-kriyā-phalātmakatādhyāropa-lakṣa ayā avidyāvāsanayā vāsitaḥ so 'kāmayata kāmitavān. On the reflection of the Self being the agent, see USP 1.18.53a: ātmābhāsas tu ti -vācyaḥ. 303 mind and senses as instrumental causes; and there is the cognition itself, vijñ na, that take place in the intellect. All of this is superimposed on the Self either directly or indirectly. Agency, however, is general: it concerns any kind of agency. The complex of pram ṇa or reliable warrants is a restricted case of cognition—one that happens to be valid—and is equally superimposed over the Self: all reliable warrants, scripture in all its scope included, are superimposed over the Self and can operate because there is such a thing as the Self to illumine them: All forms of worldly and Vedic forms of behavior that involve knowable objects and reliable warrants become operational through the mutual superimposition of the Self and the non-Self, a superimposition that is called ignorance (avidy ), as do all scriptures that are concerned with injunctions, prohibitions, and liberation.42 The superimposition of agency brings with itself the superimposition of the enjoying the results that such agency implies, bhoktṛtva.43 Furthermore, the cognitive agency, of course, has the intellect as its location—it is there that cognition happens—but cognition is dependent on a set of other factors: it is dependent on the so-called manas, commonly translated as the mind but better understood as the faculty of attention; on the cognitive faculties that function in their respective sphere, commonly called senses, indriya; finally, on the body, which houses these senses. The light of awareness is, thus, further reflected in the rest of one’s personality, but it is also progressively restricted or dimmed because it is modulated by each previous reflection: 42 tam etam avidyākhyam ātmānātmanor itaretarādhyāsaṁ purask tya sarve pramā a-prameya-vyavahārā laukikā vaidikāś ca prav ttāḥ. sarvā i ca śāstrā i vidhi-pratiṣedha-mokṣa-parā i. BSBh 1.1.1, I.3. 43 See, for instance, the end of the Adhy sa-Bh ṣya of the BSBh 1.1.1, I.5, where the two are most explicitly paired in relation to superimposition: evam ayam anādir ananto naisargiko’dhyāso mithyā-pratyaya-rūpaḥ kart tva- bhokt tva-pravartakaḥ sarva-loka-pratyakṣaḥ; “Thus is this natural superimposition that is without a beginning or end, false notion in nature, the instigator of agency an enjoyment, evident to all.” Also, the BĀUBh Introduction, VIII.5: iṣ āniṣ a-prāpti-parihārecchā-kāra am ātma-viṣayam ajñānaṁ kart -bhokt -svarūpābhimāna-lakṣa amś “The cause of the desire to attain the good and avoid the evil, viz, ignorance regarding the Self, which expresses itself as the idea of one's being the agent and experiencer…” 304  The intellect, because of its transparency and proximity, becomes a reflection of the light of awareness of the Self. For this reason, even those who discriminate fancy themselves first as being the intellect. Next there is the reflection of awareness in the mind, due to proximity, by its comingling with the intellect; then in the senses, because they are in contact with the mind; and then in the body, because of its being in contact with the senses. Thus, in succession the Self with its own innate intelligence illumines the whole aggregate of body and organs.44 These are like mirrors within mirrors, and the Self could potentially identify—have the notion “This is who I am”—in regard to any of them, contingent on one’s discriminative ability. “It is for this reason that all people identify themselves with the body and organs and their functions in an unregulated way, as per their discrimination.”45 The buddhi/antaḥ-karaṇaṭvijñ na is the first adjunct of the Self, giving it the name vijñ n tman, but the rest become its adjuncts as well. This principle can be extended even to things that are merely related to oneself, considered “my,” and Śa kara calls the whole field of potential items of identification aham-mama-gocara, “the sphere of ‘I’ and ‘my.’”46 This field or sphere is concretized in relation to the sense of Self and becomes “the notion of ‘this’,” idam-dh , where idam is a variable that stands as a complement to the notion of “I” and forming a complex with it—“I am this”—whose value can be anything from the sphere of “I and mine,” any property of the non-Self that one can superimpose over the Self, as long as it is either reached by the light of awareness or is in relation to oneself. We can now appreciate one of the most striking passages written in the history of Indian philosophy: As we said, superimposition, to define it, is the notion of something in regard to something else. It is like when one superimposes external properties over the Self, thinking, “I myself am injured” or “I myself am whole” when one’s son or wife is injured 44 buddhis tāvat svacchatvād ānantaryāc ca ātma-caitanya-jyotiḥ-praticchāyā bhavatiś tena hi vivekinām api tatra ātmābhimāna-buddhiḥ prathamāś tato 'py ānantaryāt manasi caitanyāvabhāsatā, buddhi-samparkātś tata indriyeṣu, manaḥ-saṁyogātś tato 'nantaraṁ śarīre, indriya-samparkāt. evaṁ pāramparye a k tsnaṁ kārya-kara a-saṁghātam ātmā caitanya-svarūpa-jyotiṣā avabhāsayati. BĀUBh 4.3.7, IX.527-8. 45 IbidŚ tena hi sarvasya lokasya kārya-kara a-sa ghāte tad-v ttiṣu ca aniyatātmābhimāna-buddhir yathā-vivekaṁ jāyate. 46 See USP 1.18.27. 305  or whole; or when one superimposes properties of the body and thinks, “I am fat,” “I am lean,” “I am fair,” “I stand,” “I go,” or “I leapś” or when one superimposes properties of the senses, as in “I am dumb,” “I am blind in one eye,” “I am emasculated,” or “I am blindś” or when one superimposes properties of the internal organ, such as desire, resolve, doubt and certainty.47 This superimposition whose cause is false awareness, is, Śa kara claims, called ignorance or avidy by the learned.48 We should note here for the sake of being thorough that ignorance assumed an all-important role in post-Śa kara Advaita VedāntaŚ it became a cosmological category, standing for the primordial stuff of which the world is made or which operates on Brahman as it is about to don its causal garb. As Hacker and Mayeda have shown, ignorance was not a cosmological item in Śa kara’s thought.49 In fact, in how Śa kara talks about ignorance, the very possibility of ignorance presupposes that the cosmological diversification of Being had already taken place: buddhi and the rest of the adjuncts need to be present for the mutual superimposition of properties to take place. In general, we can say that ignorance is strictly a psychological category in Śa kara’s thought and looks at the tman-Brahman relationship on the side of the Self, whereas the counterpart cosmological category on the side of Brahman is the n ma-rūpe, name and form.50 47 adhyāso nāma atasmiṁs tad-buddhir ity avocāma. tad yathā—putra-bhāryādiṣu vikaleṣu sakaleṣu vā aham eva vikalaḥ sakalo veti bāhya-dharmān ātmany adhyasyatiś tathā deha-dharmān 'sthūlo ’haṁ k śo ’haṁ gauro ’haṁ tiṣ hāmi gacchāmi la ghayāmi ca' iti; tathendriya-dharmān—'mūkaḥ kā aḥ klībaḥ andho ’ham' iti. tathāntaḥkara a- dharmān kāma-sa kalpa-vicikitsādhyavasāyādīn. BSBh 1.1.1, I.4-5. 48 tam etam evaṁ-lakṣa am adhyāsaṁ pa ḍitā avidyeti manyante. BSBh 1.1.1, I.3. 49 Hacker 1995:57-100; Mayeda 2006b:22-26, 76-84. 50 A second issue related to ignorance developed in post-Śa kara Advaita Vedānta, growing into a dispute over which the school would divide in two camps. This was the question about the locus of this ignorance: one line of Advaitins, including his immediate students Sureśvara and Padmapāda, claimed that Brahman itself is the locus of ignorance, whereas another line, started by Vācaspati Miśra but continuing the tradition of Ma ḍana, claimed that the individual Self or j va is the locus. As shown by Ingalls (1953), Śa kara, while aware of the problem, chose not to deal with it because he considered that no solution within what is logically possible could be forthcoming. Śa kara discussed the question “Whose is ignorance?” in a couple of places (most notably BSBh 4.1.3, BĀUBh 4.1.6 and BhGBh 13.2), which all tend toward evasion and can be roughly characterized by the following dialogue: - Now, whose is this ignorance that you are talking about? - Well, of the ignorant, duh! - That would be me, I gather. 306  In the BhGBh this ignorance is said to be potentially of three kinds: (1) failure to grasp an object (agrahaṇa), as in the case of darkness; (2) seeing one thing as another (vipar ta-grahaṇa), under which most of Śa kara’s favorite examples would fit, such as seeing a snake in a rope, silver in the mother-of-pearl, or when the simple-minded see dirt and a flat surface in the sky; and (3) doubt (saṁ aya), the classical example of which in Indian philosophy is the uncertainty whether a silhouette in the distance is a man or a post, to which Śa kara also commonly refers.51 These are all cases of cognitive errors, and Śa kara’s object in using them is to show that they do not constitute an error on the part of the knower, but a flaw in the causal conditions of perception: the Self is not in illusion in essentia, but in actu. They must be taken as no more than illustrations, however, because the superimposition that Śa kara talks about is evidently of a very different kind: it is the mistake that makes all other mistakes possible—as well as all truths, inclusive of the final truth expressed in the eventual mah -v kyas. This form of ignorance is not just the common mistake of false recognition that brings embarrassment, or the uncertainty that a scarecrow may cause. Ignorance is the false awareness and the superimposition that is natural (naisargika) and without a beginning (an di), that is, it obtains as a normal state and not as a cognitive oddity. We may take Mahadevan’s lead, therefore, and distinguish this from of ignorance from the common as metaphysical.52 However—and this is a point not commonly discussed, yet crucial for Śa kara’s soteriology and dharma—the mere formation of the reflection of the Self, the consequent superimposition of the notions of agency and enjoyment, the complex centered around action, - There you have it, then. - But hold on a second, you say that I am Brahman, and that I, being Brahman, cannot be ignorant! - Good for you! If you understand that much, what is the problem? Ingalls sees in this the same strategy that was employed by the Buddha in answering metaphysical questions, such as those in the famous Cula-Malunkyaputta-Sutta. 51 BSBh 13.2. 52 Mahadevan 1985. 307 namely action itself, its contributing factors and its results (kriy , k raka, phala), and the potential of identification with anything that constitutes the field of “I and my,” fashion the category of the individual Self, the universal or pad rtha to which the word “Self” can be applied: this is not what makes the Self of any Devadatta or John Doe.53 Ignorance is the immediate factor of distinguishing the category of vijñ n tman or j va from the Supreme Self, but it is not the immediate factor of individuation. Two additional factors are required for there to be an individual Self. We may put this another way. What the image of the Self will look like is contingent on the mirror: the image of the face conforms to the mirror, and the mirror can be variously inflected.54 There are some contour points that need to be invariantly present in all images so that we could identify what kind of thing the image represents, and these are the sense of Self—“I am this”—and agency. What range of values “this” will take depends on two other factors: impressions that have the nature of habitual desire that prompts action (v san , bh van , saṁsk ra, k ma), and the results of previous action or karma. The three, really, form a circle that reinforces itself. The impressions are impressions of ignorance, results of past identifications involving agency—past actions—that color, or rather perfume, one’s awareness. They are in the form of volitional tendencies for something specific. Śa kara determines their scope as r g di, which clearly refers to the well-known set of psychological faults or kle as, namely attachment, aversion, and illusion (r ga, dveṣa, moha). Desires that are formed through impression are the medium that otherwise unproductive ignorance must take to become an instigator to action (bh van ). Action and its resultant karma 53 jāti-karmādimattvād dhi tasmiñ śabdās tv ahaṁk ti. “As this bearer of the "I"-notion has a universal, and is possessed of action, etc., it can be referred to by words.” USP 1.18.28. Translation Mayeda 2006bŚ175. 54 US 1.18.31. 308 on their part produce one’s future embodiment that is an instantiation of ignorance, as it involves a wrong identification in which one becomes naturally prone to specific desires and fit for the attaining of specific goals, requiring specific action. Because the superimposition that is ignorance is natural and without a beginning, this circle of avidy -> v san -> k ma -> karma - > avidy is a true circle: everything is logically predicated on ignorance, but ignorance historically or temporally requires an existing embodiment. Being perfectly stainless, tman is distinguished from, and broken by, nescience, residual impression, and actions.55 The living Self, individuated by ignorance, results of previous actions and past impressions, assisted by the chief breath and possessing a mind and senses … 56 Ignorance could be the cause of inequality through recourse to action that is set in motion by impressions that are torments, attachment, etc.57 There must be some force impelled by which one becomes averse to one's own world, the Self, as if he were helpless. – Is it not ignorance? For, he who is ignorant is averse and acts. – Ignorance is not an instigator to action, for it conceals the true nature of a thing. It obtains the state of being the seed of action like darkness that is the cause of the action of falling into a ditch. – Well say it then, what is the cause of a man's activity. – It is said here: it is desire.58 Suffused by the impression of ignorance that is natural to him and that consists in a superimposition of the notions of action, its factors such as the agent, and its results, over the Self, he desired.59 Desire is the cause of action, because of being an instigator.60 The individual Self, thus, is a work in progress, constituted by the three factors of ignorance, desires that are impressions in nature and prompt one to act, and the results of action that have shaped one’s present identity. While ignorance is the general factor of individuation, 55 avidyayā bhāvanayā ca karmabhir vivikta ātmāvyavadhiḥ sunirmalaḥ. USP 1.10.9ab. 56 jīvo mukhya-prā a-sacivaḥ sendriyaḥ samanasko ’vidyā-karma-pūrva-prajñā-parigrahaḥ. BSBh 3.1.1, II.527-8. 57 rāgādi-kleśa-vāsanākṣipta-karmāpekṣā tv avidyā vaiṣamya-karī syāt. BSBh 2.1.36, II.344. 58 tasmād bhavitavyaṁ tena, yena prerito 'vaśa iva bahirmukho bhavati svasmāl lokāt. nanv avidyā sāś avidvān hi bahirmukhī-bhūtaḥ pravartate—sāpi naiva pravartikāś vastu-svarūpāvar ātmikā hi sāś pravartaka-bījatvaṁ tu pratipadyate andhatvam iva gartādi-patana-prav tti-hetuḥ. evaṁ tarhy ucyatāṁ kiṁ tad yat prav tti-hetur iti; tad ihābhidhīyate—eṣa ā kāmaḥ saḥ. BĀUBh 1.4.16, VIII.157-8. 59 svābhāvikyā svātmani kartr-ādi-kāraka-kriyā-phalātmakatādhyāropa-lakṣa ayā avidyā-vāsanayā vāsitaḥ so 'kāmayata kāmitavān. BĀUBh 1.4.17, VIII.159. 60 karma-hetuḥ kāmaḥ syāt, pravartakatvāt. TUBh Introduction, VI.8. 309 the category maker, desire and karma are the two factors that make it possible for one to be born with a specific identity—that is, in a family belonging to a class—and have the fitness for specific desires and forms of attainment that are related to them. As is, hopefully, evident, Śa kara’s psychology was Bhart prapañca’s psychology from top to bottom. The individual Self was constituted by the same triplet of avidy -k ma-karma, with v san negotiating the transition between the first two. This was in both cases inspired by the psychology of Yoga, specifically the idea of five kle as expressed in the YS 2.3Ś “The torments are ignorance, the sense of Self, attachment, aversion and clinging to life.”61 If anything, Śa kara was more consistent in applying this Yoga classification, sticking to the psychological significance of ignorance that takes the Self to be its opposite, whereas Bhart prapañca’s leaned towards avidy as a cosmic power.62 The point of Śa kara’s departure from Bhart prapañca was his theory of reflection of the Self in the intellect as constituting the vijñ n tmanŚ the individual was not a chunk of Brahman that is cut off by a cosmic power of ignorance. It was a product of one big mistake, a succession of nested mirror images assuming substance because of being graced by the light of awareness.63 61 avidyāsmitā-rāga-dveṣābhiniveśāḥ kleśāḥ. 62 Cf. YS 2.5: anity uci-duḥkh n tmasu niyta- uci-sukh tma-khy tir avidy , “Ignorance is the notion that takes the self, which is joyful, pure, and eternal, to be the nonself, which is painful, unclean, and temporary.” Translation Bryant 2009. 63 The similarity of Śa kara’s psychology with that of Yoga, and the purported authorship of the Yoga-Sūtra- Bh sya-Vivaraṇa, prompted Paul Hacker to advance the thesis that Śa kara might have been a converted yogin. Cf. particularly the following statementŚ “For the time being it is not possible to decide to what extent Śa kara, with his point of agreement with the “ātmology” and psychology (“cittology”) of Yoga, follows an already existing (pre- monistic) Vedānta tradition, since no work of this literature except the enigmatic BS is any longer extant. But since he differs from other Advaitins on those points of ātmology as well as in avidyology which connect him with Yoga, we may assume for the time being that his relations with this system were particularly close as a result of his earlier allegiance to it.” (Hacker 1995Ś119) It should be evident from my previous chapter that the crucial categories of Śa kara’s “ātmology, cittology, avidyology and brahmology” were taken from Bhart prapañca, who was mocked for not being able to decide whom he wants to make alliance with. While Śa kara’s acquaintance with Yoga seems more thorough than what he could have gathered from Bhart prapañca, it was Bhart prapañca who, for all we know, used or introduced the terms avidy , v san , k ma as doṣa, vijñ n tman, avy kṛte and vy kṛte n ma-rūpe etc., in Vedānta. This, I think, makes Hacker’s thesis unnecessary. 310  We can now see what is the “human” condition in Śa kara’s eyes. Which dharma in any specific form will be pertinent to oneself is dependent on a set of specific categories that have been superimposed over the SelfŚ “Scriptural statements such as ‘a br hmaṇa should sacrifice’ function through superimposition of characteristics such as membership to class, stage of religious life, and age.”64 This is the specific superimposition that must take place for dharma to become pertinent, over and above the general identification with the mind, body and senses and their natural properties: membership in the varṇ rama system. One must have become a member of the social structure for which the Veda is relevant, and without such a state of affairs obtaining, dharma does not pertain to oneself in any way. The light of awareness that permeates the field of “I and mine” must illumine a specific area of social identity that is formed into a habitual nature through avidy -k ma-karma. Śa kara was, of course, aware that there was a world beyond Vedic society, but that world was of little interest to him. It was split between the “natural world,” in which everyone participates and which is comprised of natural actions such as breathing and eating when one is hungry, actions that have no consequence for the law of karma because of not being scripturally regulated,65 and the world of sheer desire and innate faults such as attachment and aversion, in which one acts against the regulations of the Vedic injunctions and prohibitions and glides down the scale of Being all the way to plant life.66 One must have been born in the three upper classes for starters: this is the Self for which dharma in general has pertinence: 64 tathā hi—brāhma o yajeta ity ādīni śāstrā y ātmani var āśrama-vayo’vasthādi-viśeṣādhyāsam āśritya pravartante. adhyāso nāma atasmiṁs tad-buddhir ity avocāma. BSBh 1.1.1, IV.4. 65 This is a recurring distinction in his comments. For instance, BĀUBh 1.3.1. 66 See, for instance, the introduction to the BĀUBh. This was, of course, a pan-Vedāntic attitude inspired by the “third state” of the Ch ndogya Upaniṣad, other than the southern and the northern course that we saw in the previous chapter: “Then there are those proceeding on neither of these paths—they become the tiny creatures revolving here ceaselessly. ‘Be born! Die!’—that is a third state.” ChU 5.10.8. Translation Olivelle 1998Ś237. 311  The ‘Self’ here refers to the natural person that is characterized by a complex of body and senses, a member of one of the castes.67 The regulated Vedic world provides the means for attaining the goals of prosperity and liberation, and the initial point of both pursuits requires one to be a member of Vedic society, born in it. Desire and Qualification It is deducible from Śa kara’s writings that dharma is organized around a set of four categories. These are: desire, k ma; eligibility, adhik ra; desired attainment, s dhya, which can be for the most part identified with attaining a world or sphere, loka, that contains an assortment of desirable things; and means, s dhana or hetu, adequate for the desired attainment.68 Dharma is, as we have seen above, identified with the means, characterized by action or its absence, but this is the means in its relationality to the three other factors, not in itself. Let us examine now these four key categories. We saw in the previous part that the category of desire played a role of paramount importance in the field of Vedic theology in its entirety, and the pursuit of liberation in early Vedānta was also tied to desires, namely to the notion of tma-k ma, desire for the Supreme Self whose desires are always fulfilled, satya-k ma. These in the Brahma-Sūtra were the desires from the eighth chapter of the Ch ndogya. With Bhart prapañca, we saw the therapeutic paradigm making inroads in Vedic theology and desires becoming a problem. With Śa kara, Advaita Vedānta became similarly a doctrine that belonged to both worlds—Vedic theology in which k ma had a crucial positive role, and the traditions of liberation in which it was at or near the root of the problem, the disease of transmigration, saṁs ra. This uneasy coexistence of two 67 ātmaiva—svābhāvikaḥ avidvān kārya-kara a-sa ghāta-lakṣa o var ī. BĀUBh 1.4.17. 68 This is clearest in his comment on BĀUBh 1.4.17. 312 divergent worldviews meant that k ma had to undergo some rethinking, so that it could simultaneously be spurned yet have its important theological role preserved. Śa kara’s way of doing this involved extending the scope of desire to refer to purpose that motivates any undertaking. The result of this rethinking was that the idea of k ma could be employed variously, to refer to desirable things or pleasurable attainments on the one hand, or to a motive of any kind that prompts one to undertake a certain course on the other. The Vedic use of k ma that Mīmāṁsakas theorized referred to the pleasurable attainments that ritual brings—heaven, prosperity, virility, wealth, sons, etc. Most of these, insofar as all pleasurable attainments could be subsumed under the notion of felicity, were an objective category; that is, the desire for their attainment was natural to men. Some had to be made desirable scripturally if they were required for the attaining of something else. There was, however, no underlying psychology to account for the desirability of heaven and the like. “Desires” were objects of desire, and people wanted them naturally. This was true across the board in Mīmāṁsā, and even Kumārila’s radical rejection of obligation sought justification for the adequacy of the means, and not the desirability of the attainment.69 One could draw the distinction between “Vedic” and “non-Vedic” “desires,” that is, desirables—wealth could be given as an instance of the second, being explicitly referred to as puruṣ rtha70—but the only meaningful difference between the two could be that the second were present and available naturally, while the first were future and the means of attaining them were knowable solely from the Veda. Being human meant wanting stuff, and this is all there is to desire; just make sure you do not want stuff prohibited in scripture. 69 Clooney 1990:193 perspicaciously notes that Jaimini had a very pragmatic view of human desire—people want stuff just because they are human—and that such a pragmatic view of desire was the limit to which his anthropology goes, “all he has to say about human beings.” 70 See MSŚBh 4.1.2ś see also Jha 1964Ś294-5. 313  Śa kara’s works are full of such “Vedic desires,” and he is explicit in identifying them with desirable objects: Under k ma, sons and the like are intended. That is, they are desired. – But, the word k ma stands for a type of desire, does it not? – No, because it is clear from the context that the word “k ma” is used in the sense of sons, etc.71 That much is k ma – that is the limit of objects of desire. Desirable objects are such things as are characterized by the wish for a wife, sons, wealth and ritual.72 However, as we saw above, desires were properly an individuating feature of the j va, and Śa kara described them in a Yoga-like manner: they are modes of the mind formed through impressions into habits, they are faults and torments that prompt one to act for the attaining of desirable objects or for avoiding something undesirable.73 And, coming from the background of the therapeutic paradigm, he identified all desires for objects, viṣaya, as products of ignorance that perpetuate embodiment. This obviously included the desires for ordinary, “non-scriptural” things, as well as desires that prompt one to act against scriptural injunctions and prohibitions, the road of degradation culminating in plant life. However, it also included “scriptural” desires, formed through “scriptural impressions” that prompt one to perform ritual and amass good karma: But, ignorance about the Self that is the cause of the desire for attaining what is desirable and avoiding what is undesirable and is characterized by fancy of one’s nature as being an agent of action and experience that has not been removed by its opposite, the knowledge of the nature of the Self as identical with Brahman. So long as it is not removed, one is impelled by the natural faults such as attachment and aversion for the results of action, and acting even in transgression to prohibitions laid down in scripture amasses with body, mind and words bad karma that leads to undesirable present and future results. This is because of preponderance of natural faults and it is the road down 71 putrādayaś ca tatra kāmā abhipreyante—kāmyanta iti. nanu kāma-śabdenecchā-viśeṣo evocyeran; na, at yuṣaḥ putra-pautr n vṛṇ ṣva [K U 1.1.25] iti prak tya ante k m n ṁ tv k ma-bh jaṁ karomi [K U 1.1.24] iti prak teṣu tatra tatra putrādiṣu kāma-śabdasya prayuktatvāt. BSBh 3.2.2, III.562. 72 etāvān vai kāma etāvad-viṣaya-paricchinna ity-arthaḥ. etāvān eva hi kāmayitavyo viṣayaḥ—yad uta jāyā-putra- vitta-karmā i. BĀUBh 1.4.17, VIII.160. 73 BSBh 1.1.1; 2.3.33. 314  to plant life. When there is a preponderance of impressions created through scripture, then one amasses good karma with one’s body, mind and words.74 The two forms of action and their corresponding desires were, obviously, very different in Śa kara’s eyes—one of them brings good karma of some sort and the other brings one down—but their deep structure was the same: both were predicated on ignorance, and both perpetuated ignorance, the superimposition of agency over the Self; both functioned through creating impressions that give rise to desires that are in the form of habits; and both aimed towards sense objects. Because of this, there was always a chance for one to slip from the scriptural striving and start gliding down towards plant life by forming unwholesome habits. Desire for sense objects of any kind was a problem. This psychology of desires as habits formed through past impressions that prompt action meant that there was no such thing as “objective desires,” things that all people want just by being human. One could say with Sarvajñātman that all men strive after unexcelled happiness and the eradication of all suffering,75 but unlike in Mīmāṁsā, that striving was modulated by what one apprehends as desirable or otherwise. Everyone is a product of some pattern of habits, a “work in progress,” and when people strive after some good through Vedic means, that is not invariantly the same good. There are, in other words, deep psychological desires, acquired through habit formation in long lifetimes, that are related to the Vedic desirable objects and 74 na tv ātmanaḥ iṣ āniṣ a-prāpti-parihārecchā-kāra am ātma-viṣayam ajñānaṁ kart -bhokt -svarūpābhimāna- lakṣa aṁ tad-viparīta-brahmātma-svarūpa-vijñānenāpanītam. yāvad dhi tan nāpanīyate, tāvad ayaṁ karma-phala- rāga-dveṣādi-svābhāvika-doṣa-prayuktaḥ śāstra-vihita-pratiṣiddhātikrame āpi vartamāno mano-vāk-kāyair d ṣ ād ṣ āniṣ a-sādhanāni adharma-saṁjñakāni karmā y upacinoti bāhulyena, svābhāvika-doṣa-balīyastvāt. tataḥ sthāvarantādho-gatiḥ. kadācic chāstra-k ta-saṁskāra-balīyastvam. tato mana-ādibhir iṣ a-sādhanaṁ bāhulyenopacinoti dharmākhyam. BĀUBh Introduction, VIII.5. 75 “In this world, all beings have the following aspirationŚ ‘May I have unexcelled happiness, and may all suffering born of sense objects and appearing as injury be gone.’” iha jagati sarva eva jantur niratiśayam sukham uttamaṁ mamāstu | uparamatu tathopaghāta-rūpaṁ viṣayaja-duḥkham iti sp hāṁ karoti. SŚ 1.66. 315 prompt one to perform ritual, just as there are habitual tendencies for the “natural” desirable objects. With this distinction in mind, we can appreciate how Śa kara could claim on one hand that ritual that is performed in the pursuit of liberation is desireless, niṣk ma, yet tie that same ritual to a specific desire on the other. Additionally, we can make sense of his charges against Kumārila’s first account of liberation, claiming that there is no such thing as obligatory action (nitya-karma) that is not prompted by desire: all forms of action are k mya, directed towards something. Ritual may not be performed for obtaining the common Vedic desirable things, but one still hopes to gain something through its performance. We will return to this point later. These “scriptural” desires are ultimately directed towards desirable attainments, s dhya, that can be broadly identified with four spheres, loka. The mere desire for the attaining of a specific sphere is, however, not a sufficient ground for one to take up the means or procedure adequate for the respective attainment. K ma is absorbed in the category of adhik ra, qualification, competence or entitlement for the attainment and for its adequate means. Adhik ra is a composite category, all whose constituents must be present for a successful enterprise. The closest Śa kara comes to delimiting the scope of adhik ra is BSBh 1.3.33, in the BS section that affirms contra Jaimini that the gods are qualified for knowledge of Brahman, even though they are not qualified for ritualŚ “Qualification is contingent on aspiration, ability, absence of prohibition, etc.”76 Let us reflect on these through the instance of the horse sacrifice ritual, the Aśvamedha, that is known to us from Chapter Four. For one to take up the performance of this ritual, one primarily must have the desire for the attainment that is related to it. For Śa kara, the ultimate 76 arthitva-sāmarthyāpratiṣedhādy-apekṣatvād adhikārasyaś I.204. 316 result of performing the Aśvamedha is attaining the state of identity with life-breath, pr ṇa, in its highest manifestation as Hira yagarbha the world soul; in other words, it is the attainment of brahma-loka. At the bare minimum, therefore, one must have the aspiration, arthitva or k ma, towards attaining brahma-loka. However, the performance of Aśvamedha is not open just to anyone with the adequate desire: it is specifically meant for royalty, and even br hmaṇas are forbidden from preforming it.77 One must not be, then, under a scriptural prohibition or pratiṣedha from performing the ritual that is directed towards the attainment related to one’s aspiration. One also must have the ability or s marthya to perform the ritual, which in this specific case would include the massive wealth that is required for an Aśvamedha. These three, and potentially other factors (as shall we see a bit later, being married is also part of the qualification for the performance of all Vedic rituals), must be present for there to obtain an entitlement for the performance of the ritual. This entitlement, adhik ra, is by no means uniform for all members of the varn rama system. As we saw above, everyone’s conditioning in terms of k maṭv san was different, and even if one did have the desire for some attainment through a ritual, a host of other factors were in play: Śūdras were, as is well known, prohibited both from ritual and the study of Brahman;78 very few kings could perform r jasūya (none in Śa kara’s time, per his own testimony)79; many 77 See, for instance, BĀUBh 1.1.1, VIII.6 (and the Vārttika thereon, 1.1.5-6), where Śa kara and Sureśvara justify the meditation on the sacrificial horse as useful for those who are, otherwise, not entitled to perform the a vamedhaŚ "The utility of this meditation related to the horse sacrifice is this: Those who are not entitled to this sacrifice will get the same result through this meditation itself." asya tv aśvamedha-karma-sambandhino vijñānasya prayojanam— yeṣām aśvamedhe nādhikāraḥ teṣām asmād eva vijñānāt phala-prāptiḥ. SureśvaraŚ akartāpy aśvamedhasya brāhma aś ceha buddhimān. Śa kara is even more explicit about the relative nature of qualification in reference to the royal consecration ritual, the r jasūyaŚ “Qualification should not be ruled out where it is possible because it is impossible somewhere else. Even men are not competent for all things. For instance, Brāhma as do not have competence for a r jasūyaś” na ca kvacid asambhava ity etāvatā yatra sambhavas tatrāpy adhikāro’podyeta. manuṣyā ām api na sarveṣāṁ brāhma ādīnāṁ sarveṣu rājasūyādiṣv adhikāraḥ sambhavati. BSBh 1.3.33, I.204. 78 See Clooney 1990:189-94 on the first, BS 1.3.34-8 on the second. 79 “’There never were kings who were world-sovereigns, just as there are none today,’ one would be obliged to say, and thus obstruct the injunction for a royal consecration ceremony.” idānīm iva ca nānyadāpi sārvabhaumaḥ kṣatriyo 317 of the Vedic rituals were caste specific, etc. We do not need an exhaustive account of all that an adhik ra involves, so long we appreciate how personalized it is. We are now in a position to make the full statement against the Prābhākara account of duty. For there to arise the cognition “this is my duty” regarding a Vedic injunctive statement, there first needs to obtain a superimposition of cognitive and active agency over the pure Self. One must, further, identify oneself as a member of the varṇ rama system through the set of v san -k ma-karma. This set also determines what kind of aspirations one has. One must also be free from scriptural prohibitions, as well as have the ability and means that are required for carrying out the specific duty. All of this must obtain for there to arise a sense of personal duty, over and above the cognition of the ideality of the scriptural object. This is the full import of Śa kara’s claim that we already discussedŚ An injunction becomes fruitful when human effort is understood independently, over and above the cognition produced merely by hearing the injunctive statement. It is like hearing that “He who wants heaven should perform the Darśa-pūr amāsa ritual.” The mere cognition produced by the injunctive statement relating to the Darśa-pūr amāsa is not in itself the performance of the Darśa-pūr amāsa. That will necessarily be contingent on qualification and the like.80 Drawing the consequences of this, we may say that the level at which the reliable warrants operate is the Self over which cognitive agency has been superimposed. This superimposition is, of course, ignorance, but it is ignorance that is constituting and affecting all equally—it is affecting the category of the individual Self, the vijñ n tman or the “Self of cognition.” For this reason, the reliable warrants are just about knowing something, and they are valid on intersubjective level. But the vijñ n tman is properly individuated by the further ’stīti brūyāt, tataś ca rājasūyādi-codanā uparundhyāt. This is an explicit jab at Kumārila's claim that the world now is as it has always been. BSBh 1.3.33, I.208. 80 tatra hi vidheḥ sāphalyaṁ yatra vidhi-vākya-śrava a-mātra-janita-vijñāna-vyatireke a puruṣa-prav ttir gamyate. yathā darśa-pūr amāsābhyāṁ svarga-kāmo yajeta ity evam ādau. na hi darśa-pūr amāsa-vidhi-vākya-janita- vijñānam eva darśa-pūr amāsānuṣ hānam. tac cādhikārādy-apekṣānubhāvi. BĀUBh 1.4.7, VIII.108. 318 categories of karma and k ma-v san Ṭ It is at this level that aspirations are operative and a cognition of duty possible. This level goes beyond even class membership, and is as individual as can be. The cognition of a sense of personal duty is facilitated by ruti, insofar as it hinges on the apprehension of a scriptural reality, but it is not in the domain of pram ṇa, which must be applicable universally. The Attainments of Dharma In general, there are four possible final attainments, s dhya, for which the means are knowable from the Veda, and they are all “worlds” or spheres, loka: the world of men, manuṣya-loka; the world of the forefathers, pitṛ-loka; the world of the gods, deva-loka; and the world that is the Self, tma-loka. I use the qualification “final” on purpose and in a specific senseŚ these are attainments that are not means to some other goal—they are at the top of a causal structure— although only the last is really the ultimate. There are a few other items that are also attainable through knowledge provided by the Veda or otherwise scripturally regulated, although they are part of the natural world, which are s dhya but also s dhana for these final attainments. The first three final attainments are all primarily in the domain of dharma that is characterized by engagement, pravṛtti, and their corresponding means are all forms of action or something immediately related to action. The fourth is in the domain of dharma that is characterized by disengagement, nivṛtti, and its respective means is solely knowledge. There are some details that muddle this neat classification, but let us first focus on the contours. I belabored above Śa kara’s claim that knowledge qua knowledge is something that is not humanly contingent and not in the domain of human choice or preference. We may say in the Cartesian spirit that matters of knowledge are such that, when seen clearly and distinctly, the human mind cannot but assent to them. This for our purposes meant that if one had a choice 319 regarding anything presented in the Veda as a means of some attainment, then insofar as one had such a choice, that was not a matter of knowledge, but of action. Whereas in pre-Śa kara Vedic theology action or karma referred strictly to ritual, bodily action, and vidy ṭup sana was its mental counterpart, unique in nature and procedure but diversified per the desired object and the relation of the meditational counterparts, Śa kara reclassified everything over which man had a choice as a form of action: Action is bodily, vocal, and mental, laid down in the ruti and smṛti literature and called dharma.81 – But hold on, isn’t knowledge, to define it, a mental action? – No, they are different. Action is that where something is enjoined without regard to the nature of the thing, and it is contingent on the operation of the human mind. … Although meditation, which is but mentation, is mental, man has the option to do it, not do it or do it otherwise, and so it is dependent on man. Knowledge, on the other hand, is produced by a valid cognition, and a valid cognition concerns a thing just as it is. Therefore, knowledge cannot be done, not done or done otherwise: it is fully dependent on the thing, not on an injunction or on man. Thus, although knowledge is mental, it is vastly different. It is like this: The notion about man and woman as fire that is expressed in the text “A man is surely fire, Gautama. A woman is surely fire, Gautama” is mental, but being produced solely by an injunction, it is nothing but action, and it is dependent on man. However, the notion of fire regarding a known fire is not dependent on an injunction or on man, but solely on the thing that is an object to perception, and so it is but knowledge, not action.82 Śa kara clearly takes the quoted passage as an instance of meditation, where man and woman can be seen as fire intentionally, although the meditator knows that no actual relation exists between the two: s/he has made the deliberate choice to see the one as the other. The enormous significance of Śa kara’s distinction between knowledge and action was that the doctrine of brahma-vidy was no longer a straightforward matter: the notion of a standard Upaniṣadic 81 śārīraṁ vācikaṁ mānasaṁ ca karma śruti-sm ti-siddhaṁ dharmākhyam. BSBh 1.1.4, I.19. 82 nanu jñānaṁ nāma mānasī kriyā, naś vailakṣa yāt. kriyā hi nāma sā, yatra vastu-svarūpa-nirapekṣaiva codyate, puruṣa-citta-vyāpārādhīnā ca ... dhyānaṁ cintanaṁ yady api mānasam, tathāpi puruṣe a kartum akartum anyathā vā kartuṁ śakyam, puruṣa-tantratvāt. jñānaṁ tu pramā a-janyam. pramā aṁ ca yathā-bhūta-vastu-viṣayam. ato jñānaṁ kartum akartum anyathā vā kartum na śakyam. kevalaṁ vastu-tantram eva tatś na codanā-tantram, nāpi puruṣa- tantramś tasmān mānasatve ’pi jñānasya mahad vailakṣa yam. yathā ca puruṣo v va gautam gniḥ yoṣ v va gautam gniḥ [ChU 5.7,8.1] ity atra yoṣit-puruṣayor agni-buddhir mānasī bhavatiś kevala-codanā-janyatvāt tu kriyaiva sā puruṣa-tantrā caś yā tu prasiddhe ’gnāv agni-buddhiḥ, na sā codanā-tantrā, nāpi puruṣa-tantrāś kiṁ tarhi? pratyakṣa-viṣaya-vastu-tantraiveti jñānam evaitat, na kriyā. Ibid, I.26. 320 brahma-vidy in which the object, the attainment, the procedure and the path of ascension are the same but details vary could not hold good, because in some of these vidy s one had to make the choice of seeing oneself as Brahman in meditative absorption, whereas in others one had no choice but to know oneself as Brahman, so long as one was capable of rightly understanding what was being said. We will say a little more on this in a bit, but for now we should note that anything in the Veda that is expressed by an injunction and dependent on human effort, be it bodily, vocal, or mental, is a form of action. The three action-related attainments were, of course, a commonplace in Vedic theology in general and in the Upaniṣads and Vedānta in particular, and two of them are already well- known to us from the accounts of liberation. The world of men is attained through giving birth to a son, but more specifically through the deathbed rite of entrusting, sampratti or samprad na, in which the father transfers the performance of his ritual duties to his son when it becomes clear to him that death is imminent. In this way, he vicariously continues his existence in the world of men and, thus, “wins” manuṣya-loka. The rite consists in the father saying to his sonŚ “You are Brahman, you are the sacrifice, you are the world,” and in the son replyingŚ “I am Brahman, I am the sacrifice, I am the world.”83 The son thus becomes the means or s dhana for winning the world of men. More specifically, however, since the son is not karma the preferred Vedic means, the means is the son as related to the rite of entrusting or transference, which by Śa kara’s reckoning is action of the vocal type.84 83 yadā praiṣan manyate ‘tha putram āha tvaṁ brahma tvaṁ yajñas tvaṁ loka iti. sa putraḥ pratyāhāhaṁ brahmāhaṁ yajño ‘haṁ loka iti. BĀU 1.5.17. Another important place for this idea is the second adhy ya of the Aitareya Upaniṣad, where the birth of a child is described as the second birth of one’s Self “for the continuance of the worlds.” 84 See the avataraṇa to the comment on the BĀUBh 1.5.17, VIII.192Ś putrasya tv akriyātmakatvāt kena prakāre a loka-jaya-hetutvam iti na jñāyate. atas tad vaktavyam ity athānantaram ārabhyate. 321  Getting a son is, of course, predicated on having a wife, j y , but marrying and obtaining a wife are much more than what the common categories might suggest. A wife becomes an essential part of the qualification, adhik ra, for the performance of ritual, which is the means or s dhana for the second final attainment, the world of the forefathers, pitṛ-loka. Marriage is, then, scripturally regulated not only in social terms, but as an essential element of Vedic ritual causality. After completing the course of studying the Vedas with a teacher and before commencing the life of daily ritual performance, a man must marry a wife. He must do that in imitation of a primordial mythic event, where the first man and sacrificer Prajāpati was overcome by desire and bored by being alone, grew double in size and split himself in two, a male and female, so that he could accept a wife and sacrifice to the Self-projected gods.85 A wife is, thus, a means that entitles one to perform ritual.86 The performance of ritual requires some wealth in the form of cattle and the like, and this wealth as an attainment or s dhya becomes the means, s dhana, for performing ritual.87 Ritual on its part is the means for attaining the world of the forefathers through the southern path that we have seen in the BS account. Pitṛ-loka is the final attainment, in the sense that it is not a means to anything further. It is not, however, ultimate, since one falls back to the world of men once the good karma has been exhausted, hopefully to take up ritual performance again. The rituals that bring one to the world of the forefathers are those that lead to heaven, such as the Darśa-pūr amāsa and the Agnihotra. We will remember at this point that these were rituals which Mīmāṁsakas claimed every member of the three upper classes had to perform, either through the desire for heaven (k mya) or under the scriptural provision that they had to be 85 This is narrated in the fourth br hmaṇa of the first adhy ya of the BĀU. 86 jāyā karmādhikāra-hetu-bhūtā me mama kartuḥ syātś tayā vinā aham anadhik ta eva karma i; ataḥ karmādhikāra- sampattaye bhavej jāyā. BĀUBh 1.4.17, VIII.159. 87 atha vittaṁ me syāt karma-sādhanaṁ gavādi-lakṣa am. Ibid. 322 performed as long as one lived (nitya). The other rituals, such as those for prosperity, virility, wealth, etc., were independent, and we saw that wealth was theorized as a form of final attainment insofar as it was directly conducive to human happiness, puruṣ rtha. Śa kara seems to think that these other rituals, or at least some of them, are absorbed in the causal structure that is ultimately geared towards the world of the forefathers or the world of men. For instance, the wealth-rituals are necessary because wealth is the means for the heaven-rituals, not a thing desirable independently. Furthermore, this whole structure, including the Agnihotra, etc., is in an important sense optional, k mya, because it is occasioned by the desire for the final attainment.88 To state this more emphatically: in the strict sense, there is no such thing as mandatory rituals that every member of Vedic society must perform. This is a point which Śa kara discusses often and in depth, and his arguments are directly aimed against Kumārila’s presentation of the obligatory rituals as immediately efficacious in the pursuit of liberation. We will remember that Kumārila’s first account of liberation presupposed that desire and ignorance have already been eradicated when one had to take up the performance of ritual solely under scriptural provision, to prevent future bad karma and exhaust some of the present karmic stock. Śa kara’s critique of this amounted to claiming that it was putting the cart before the horse: no one, in fact, does anything without being prompted by some urge and motive. As we saw while discussing Śa kara’s psychology, karma was third in line among the factors of individuation of the Self, a resultant immediately of k maṭv san and mediately of ignorance. Action was, in other words, predicated on desire just because any undertaking is prompted by a specific psychological setup, which consisted in forms of attachment and aversion. This had to be the case because desire was based on ignorance, and ignorance just 88 This seems clear from BĀUBh 1.4.17. 323 meant superimposing agency over the Self: so long as one acts, one affirms not knowing the Self, in which case some form of attachment and aversion are inevitable. It is not possible that there will be exhaustion of all action, because it is impossible that desires, which are the causes of action, would cease in the absence of knowledge. For, he who does not know the Self has desires, since desires have as their results whatever is not the Self.89 Furthermore, it did not follow that nitya-naimittika are all that different from the k mya actions even on the Mīmāṁsā account. Kumārila, as I just mentioned, argued that nitya- naimittika had to be performed because otherwise bad karma would be created, pratyav ya, but this presupposed a desire to avoid bad karma and was, eo ipso, a striving after something desirable. In a sense, desireless action was contradictio in adiecto, and action that was solely determined through scripture was an impossible notion. – It is not right not to perform that which is laid down in scripture, because such acts are not prohibited [like the prohibited eating of the meat of an animal killed by a poisonous weapon]. – Not so, because both are equally based on wrong ideas and produce bad results. Actions laid down in scripture are based on wrong notions and conducive to what is bad just as much as eating poisoned flesh is. Therefore, for the one who knows the real nature of the Supreme Self, it is but right not to perform actions laid down in scripture, since they are equally based on wrong notions and conducive to what is bad, when the false notion has been removed by knowledge of the Supreme Self. – It may be right in that case, but the obligatory rituals are solely based on scripture [that is, not on desire, which would have been removed by the removal of wrong notions] and are not conducive to what is bad, and so it is not proper that they be given up. – Not so, because they are enjoined for the one who has the faults of ignorance, attachment and aversion, etc. Just as the optional rituals such as Darśa-pūr amāsa are enjoined for the one who has the fault of desiring heaven, likewise the mandatory rituals are laid down for the one who has the fault of ignorance, the root of all evil, and the faults of attachment and aversion for attaining the desired and avoiding the undesired that are produced by this ignorance, and who seeks to attain the desirable and avoid the undesirable being equally impelled to act by the same ignorance. They are not based solely on scripture. Nor is there an intrinsic divide into optional and mandatory rituals of Agnihotra, Darśa-pūr amāsa, Paśubandha and Somayāga. They become optional owing to the fault of the agent who has the desire for heaven, etc. Likewise, the mandatory rituals are enjoined for the one who has the fault 89 na ca karma-hetūnāṁ kāmānāṁ jñānābhāve niv tty-asambhavād aśeṣa-karma-kṣayopapattiḥ. anātma-vido hi kāmaḥ, anātma-phala-viṣayatvāt. TUBh Introduction, VI.9. 324  of ignorance and who by his very nature wants to get the desirable and avoid the undesirable.90 Mīmāṁsakas, of course, meant that the k mya rituals are not meant for attaining any of the desirable objects promised in the Veda, and I suggested that Kumārila clearly did not mean to say that nitya-naimittika are undertaken without any purpose or motive. But we should remember that svarga was defined as felicity pure and simple, and Śabara explicitly subsumed under it all goals of man.91 It would be difficult to argue, therefore, that avoiding bad karma is not what is desirable to man or somehow constitutive of that, in other words, k myaṬ Thus, even if we disregard the psychological dimension of k ma, Śa kara’s critique is very much to the point. This meant that all rituals were prompted by desire, and that the traditional k mya rituals were tied to some specific desirable attainment—sons, wealth, heaven—whereas the nitya, when they were not for heaven, were driven by a general desire to get what is good and avoid what is bad. They were at their root both k mya. Nevertheless, in virtue of the twofold meaning of k ma—as desirable object and as psychological desire—it was still possible to describe the so- called obligatory rituals as niṣk ma, not tied to any specific desirable object. This will have a massive soteriological and theological significance for Śa kara, one which was directly 90 apratiṣedha-viṣayatvāc chāstra-vihita-prav tty-abhāvo na yukta iti cet, naś viparīta-jñāna- nimittatvānarthārthatvābhyāṁ tulyatvāt. kalañja-bhakṣa ādi-prav tter mithyā-jñāna-nimittatvam anarthārthatvaṁ ca yathā, tathā śāstra-vihita-prav ttīnām api. tasmāt paramātma-yāthātmya-vijñānavataḥ śāstra-vihita-prav ttīnām api, mithyā-jñāna-nimittatvena anarthārthatvena ca tulyatvāt, paramātma-jñānena viparīta-jñāne nivartite, yukta evābhāvaḥ. nanu tatra yuktaḥś nityānāṁ tu kevala-śāstra-nimittatvāt, anarthārthatvābhāvāc ca abhāvo na yukta iti cet, naś avidyā-rāga-dveṣādi-doṣa-vato vihitatvāt. yathā svarga-kāmādi-doṣavato darśa-pūr amāsādīni kāmyāni karmā i vihitāni, tathā sarvānartha-bījāvidyādi-doṣavatas taj-janiteṣ āniṣ a-prāpti-parihāra-rāga-dveṣādi-doṣavataś ca tat-preritāviśeṣa-prav tter iṣ āniṣ a-prāpti-parihārārthino nityāni karmā i vidhīyanteś na kevalaṁ śāstra-nimittāny eva. na cāgnihotra-darśa-pūr amāsa-cāturmāsya-paśubandha-somānāṁ karma āṁ svataḥ kāmya-nityatva-viveko ‘sti. kart -gatena hi svargādikāma-doṣe a kāmārthatāś tathā avidyādi-doṣavataḥ svabhāva-prāpteṣ āniṣ a-prāpti- parihārārthinaḥ tad-arthāny eva nityāni—iti yuktam; taṁ prati vihitatvāt. BĀUBh 1.3.1, VIII.42. 91 “It is the generic word, denotative of the ends of all men, not restricted to any particular end”; sarva- puruṣārthābhidhāyī sāmānya-vacanaḥ śabdaḥ na viśeṣe avasthāpito bhavati. MSŚBh 4.3.20, IV.1258. 325 facilitated by Kumārila’s refashioning of the obligatory rituals as meant for exhausting the karmic stock, not just for preventing fresh bad karma. We will deal with this in the next chapter, however, and for the present purposes we should just note that ritual was normally directed towards winning the world of the forefathers through the southern course. It was dependent on wealth as a means, which in its turn was dependent on marrying a wife as a qualification. These two, as well as a son as the means for the world of men, were desirable objects of the mixed s dhya-s dhana type, and were tied to respective rituals as well, or otherwise scripturally qualified. Together with the two final attainments, they constituted a causal ritual chain and formed the scope of k ma as desirable objects of the Vedic kind. These two hankerings after the ends and means are what desire is, prompted by which an ignorant man helplessly enmeshes himself like a silkworm, and through absorption in the path of rituals becomes externally directed and does not know his own world, the Self. As the Taittiriya Br hmaṇa says, “Infatuated with rites performed with the help of fire, and choked by smoke, they do not know their own world, the Self.” – But, how come you say that is the extent of desires? They are endless. – It does not matter if one wishes or not, one cannot get more than this, which consists of results and means. There is nothing in the world besides these results and means, visible or invisible, to be acquired. Desire is concerning things to be acquired, and since these extend no farther than the above, it is but proper to say, “That is the extent of desire.” That is, desire consists of the two hankerings after the ends and means, visible or invisible, which are the specific sphere of qualification of the ignorant man. Hence the wise man should renounce them.92 The two worlds, of course, contain an assortment of enjoyable objects which are eo ipso won and enjoyed by winning the worlds, and that prompts us to consider the nature of the third world, that of the gods. 92 te ete eṣa e sādhya-sādhana-lakṣa e kāmaḥ, yena prayuktaḥ avidvān avaśa eva kośa-kāra-vad ātmānaṁ veṣ ayati – karma-mārga evātmānaṁ pra idadhad bahirmukhī-bhūto na svaṁ lokaṁ pratijānātiś tathā ca taittirīyake – agni- mugdho haiva dhūmat ntaḥ svaṁ lokaṁ na pratij n ti [3.10.2.1] iti. kathaṁ punar etāvattvam avadhāryate kāmānām, anantatvātś anantā hi kāmāḥ — ity etad āśa kya hetum āha — yasmād na icchan ca na icchann api, ato 'smāt-phala-sādhana-lakṣa ād bhūyo 'dhikāraṁ na vinden na labheta. na hi loke phala-sādhana-vyatiriktaṁ d ṣ am ad ṣ aṁ vā labdhavyam astiś labdhavya-viṣayo hi kāmaḥ; tasya caitad-vyatireke ābhāvāt yuktaṁ vaktum — etāvān vai kāmaḥ iti. etad uktaṁ bhavati — d ṣ ārtham ad ṣ ārthaṁ vā sādhya-sādhana-lakṣa am avidyāvat-puruṣādhikāra- viṣayam eṣa ā-dvayaṁ kāmaḥ. BĀUBh 1.4.17, VIII.130-1. 326  Winning the World of the Gods The third attainment is the worlds of the gods, deva-loka, culminating in the world of Hira yagarbha or brahma-loka, achieved through the northern course known to us from the BS account. The means of attaining this world is meditation on the so-called lower Brahman, to which ritual may be optionally added to form a variety of jñ na-karma-samuccaya, a combination of “knowledge” and “action,” that is, meditation and ritual. This attainment is, really, the equivalent of the state of liberation in the Brahma-Sūtra. As I have already reminded us, the normative Vedānta account of liberation presented by Bādārāya a in the BS envisioned a standard brahma-vidy arranged around an injunctive text. The doctrine of vidy was explicitly modeled on the ritual, insofar as the individual brahma- vidy s were formed by combining details from different texts, had the same standardized object—Brahman—and involved the same attainment—brahma-loka—achieved by ascending through the same northern course. The attainment was a specific place, achieved by motion. The process was meditative absorption on Brahman through any of the thus standardized vidy s. The result involved enjoying the Vedic delectable objects: ancestors that manifest at will, garlands and perfumes, women, and carriages. Liberation itself consisted in winning sovereignty, sv r jya, and pleasure, bhoga, which were directly derived from the seventh and eighth prap ṭhaka of the Ch ndogya, respectively. I also said that for Śa kara this doctrine involved action of the mental kind because at the very least one had the choice to sit down and intentionally identify oneself with Brahman—or not. It was not a matter of knowing, but of doing, and doing was posterior to pram ṇa and involved a host of personal and other factors. A sure sign that a text was not about liberation was if the text was injunctive. “Knowledge arises from reliable warrants and concerns the real nature 327 of the thing. It can be neither produced nor blocked by a hundred injunctions.”93 Well, sure sign most of the time, as we shall see in Chapter Nine. Liberation, as we shall also see later, was just about knowing, a form of anamnesis, and had nothing to do with action of any kind. See, for instance, his comment on the one of the sūtras on meditation that I identified as crucial because of being pellucid, 4.1.7: The consideration about sitting and the like does not arise, to begin with, regarding meditations that are related to ritual subsidiaries, because in that case they are dependent on the ritual [and regulated by its provisions]. It does not arise regarding perfect seeing either, because knowledge is dependent on the thing. It is, however, with regard to other kinds of meditation that it should be deliberated whether one should meditate without a specific rule—standing, lying down or sitting—or strictly sitting.94 Knowledge qua knowledge cannot be scripturally regulated, because it follows the constitution of things, not some human, divine, or non-personal whim. Śa kara, thus, cancelled out what Bādarāya a affirmed, and his distinction of knowledge and meditation in effect divided Bādarāya a’s brahma-vidy into one leading to Brahman the effect, k rya-brahma or Hira yagarbha, as already suggested by Bādari, involving absorptive meditation and in which everything that we know from the BS account was fine and dandy, and another one concerning the pure Brahman, involving theological and philosophical reflection and a thoroughly reevaluated role of ritual. Śa kara continued using most of the same standard terms for both meditation and reflection: vidy , jñ na, vijñ na, dar ana, dṛṣṭi, even the one which was most clearly associated with meditation, up sana,95 but he explicitly drew the distinction when it had 93 jñānaṁ tu pramā a-janyaṁ yathā-bhūta-viṣayaṁ ca. na tat niyoga-śatenāpi kārayituṁ śakyate, na ca pratiṣedha- śatenāpi vārayituṁ śakyate. BSBh 3.2.21, III.595. 94 karmā ga-samvarddheṣu tāvat upāsaneṣu karma-tantratvāt na āsanādi-cintāś nāpi samyag-darśane, vastu- tantratvād vijñānasyaś itareṣu tu upāsaneṣu kim aniyamena tiṣ han āsīnaḥ śayāno vā pravarteta uta niyamena āsīna eveti cintayati. BSBh 4.1.7, III.782. 95 See, for instance, TUBh 1.11.4, where up sana is not meditation proper, but reflection. 328 to be drawn and used terms such as samyag-dar ana, “perfect seeing,” and pairs such as para- vidy and apara-vidy , saguṇa-vidy and nirguṇa-vidy , brahma-v kya and up san -v kya.96 There were a few other, equally important reasons why Śa kara denied that achieving brahma-loka was real liberation, most of which are commonly known and just what we would expect from him. Liberation is just being the Self, sv tmany avasth naṁ mokṣaḥ, and is, thus, equivalent to the Self.97 Ergo, liberation is omnipresent and non-different from the itinerant. The individual Selves are ultimately Brahman, so liberation cannot be something one could reach by going to another place, for it only makes sense to reach that which is separate from oneself.98 Further, the reaching of brahma-loka presupposes duality—an agent, action, object of attainment—and with that it presupposes and reinforces ignorance. It also jeopardizes texts that affirm non-duality: they could only make sense if they were related to liberation, whereas the meditational texts would still be meaningful with brahma-loka as their independent attainment.99 But for our purposes here the important thing was that going to brahma-loka involved attaining the same Vedic desirable objects, k ma—women and the like—that one wished to attain prompted by desire. The essential difference was just that one did not have to return to the human realm as in the case of ritual and the southern course. In the old Vedānta this was good enough a difference, since, as I argued, its disagreement with Mīmāṁsā was less on the attainment and more on the means. With Śa kara, however, we reach here the point where a wedge is inserted in the Veda as a canon: on the one hand are things k mya, geared towards attaining the common pleasurable objects in one of the three worlds through some form of 96 The first two pairs are omnipresent in his works. On the last, see BSBh 3.2.21. 97 TUBh Introduction. 98 gati-śruter āpya iti cet – sūrya-dv reṇa, tayordhvam yan ity evam-ādi-gati-śrutibhyaḥ prāpyo mokṣa iti cetś naś sarva-gatatvāt gant bhyaś cānanyatvāt. ākāśādi-kāra atvāt sarva-gataṁ brahma, brahmāvyatiriktāś ca sarve vijñānātmānaḥś ato nāpyo mokṣaḥ. gantur anyad vibhinna-deśaṁ ca bhavati gantavyam. na hi, yenaivāvyatiriktaṁ yat, tat tenaiva gamyate. TUBh 1.11.4, VI.48-9. Also, the Introduction to the eighth prap ṭhaka of the Ch ndogya. 99 TUBh 1.11.4 on both. On the second, BSBh 3.2.14-5 is a good delineation of the principle. 329 action, and on the other hand there is the world which is the Self, attained solely through knowledge. Brahma-loka, the northern course, meditation and sovereignty and enjoyment were part of the first group. One was prompted by the same desires as defilements to attain desirable objects of the same kind.100 This was the top of what could be achieved pursuing pravṛtti- dharma and promotion or abhyudaya as the human good toward which such dharma aimed. Ritual, with or without the accompaniment of meditation, which this ignorant man, for whom the divisions of caste, order of life and so forth exist, and who is bound to those rites, performs, leads to promotion beginning with human birth and ending with identity with Hira yagarbha.101 Introducing the path of the gods which is for obtaining the results in the lower meditations…102 The course of the gods must be common to all meditations “with qualities” that result in the attainment of promotion.103 Let us note this well: brahma-loka is at the top of the process of abhyudaya. What this meant was that the text which was front and center in the BS account of liberation, the dahara-vidy of the eighth chapter of the Ch ndogya, was not about liberation at all: it was about meditation that results in promotion. This was a profoundly troubling text for Śa kara, not only because the BS took its account of liberation as the normative in all brahma-vidy s, but because it infelicitously combined all kinds of things that Śa kara could not see hand in hand: the Self that is free from faults, old age, death, lamentation, hunger and thirst—in short, his own favorite description of the Self—with “true” desires and resolves, that meant the ability to have ancestors manifest themselves at one’s will, to have fun with women, carriages, garlands and perfumes, all predicated on the ability to move throughout the worlds without restriction. 100 See BĀUBh 1.4.17. 101 etasya hy aviduṣo var āśramādi-pravibhāgavato 'dhik tasya karma o vidyā-sahitasya kevalasya ca śāstroktasya kāryaṁ manuṣyatvādiko brahmānta utkarṣaḥ. BĀUBh 1.4.10, VIII.141. 102 aparāsu vidyāsu phala-prāptaye deva-yānaṁ panthānam avatārayiṣyan ... BSBh 4.2.1, III.699. 103 sarvāsām eva abhyudaya-prāpti-phalānāṁ sagu ānāṁ vidyānām aviśeṣe a eṣā deva-yānākhyā gatir bhavitum arhati. BSBh 3.3.31, III.666. 330  For our ultimate mah -v kya purposes, we should note that this separation of meditation that ultimately leads to brahma-loka meant that the central sentences in Bādarāya a’s brahma- vidy s, such as ya tm apahata-p pm vijaro vimṛtyur vi oko vijighatso 'pip saḥ satya-k maḥ satya-saṅkalpaḥ so 'nveṣṭavyaḥ sa vijijñ sitavyaḥ of the dahara-vidy , could not possibly be the most important Upaniṣadic statements, for the simple reason that their final attainment was not liberation. One gets the sense that Śa kara thinks such passages could be rectified for knowledge and liberation purposes, if read in the light of Yājñavalkya’s teaching of the Self, in which case the possession of unfailing desires and the like would not stand for what the meditator hopes to achieve, but for some form of praise of Brahman, and the injunctive force of the core statement would also fail to obtain.104 We could say that Śa kara turned the table on Bādarāya a: we can rectify passages for coherence, but not the way Bādarāya a wanted. This rejection of attaining brahma-loka as the equivalent to liberation meant going directly against the Brahma-Sūtra, so a saving grace was found in the idea of krama-mukti or gradual liberation, which was already there in the BS itself but was rejected by Bādarāya a. One could still achieve real liberation in brahma-loka if one became disillusioned with the idea of sovereignty and developed perfect knowledge of the Self, in which case one would still need to complete one’s term in brahma-loka—just as the liberated aspirant on earth would have to wait till death for full liberation—but would be then liberated along with Hira yagarbha. Śa kara was explicit that the sovereignty won through meditation had to end with the expiry of the kalpa, and 104 See BSBh on 3.3.39, III.681, where Bādarāya a affirms the unity of the “meditations on the heart” in the ChU and BĀU, and in effect makes the BĀU conform the ChUŚ “But, there is this difference between the twoŚ in the Ch ndogya account, the meditation on Brahman is of the qualified kind, because it is said that desires are to be known just as the Self isŚ ‘Those who depart from here without discovering the Self and these true desires.’ In the V jasaneyaka, however, we see just that Brahman without qualities being taught.” ayaṁ tu atra vidyate viśeṣaḥ— sagu ā hi brahma-vidyā chāndogye upadiśyate—atha ya ih tm nam anuvidya vrajanty et ṁ ca saty n k m n ity ātma-vat kāmānām api vedyatva-śrava āt, vājasaneyake tu nirgu am eva brahma upadiśyamānaṁ d śyate. 331 that the rise of perfect knowledge had to happen before one could be really liberated in brahma- loka. Great sages may have been attached to other meditations [lit., knowledges] that result in sovereignty and the like. It makes sense that after [attaining brahma-loka] they became disillusioned on seeing that such sovereignty depletes and, becoming fully absorbed in knowledge of the Supreme Self, attained liberation.105 Those who have attained the world of Brahman that is the effect reach the supreme and pure state of Viṣ u along with Hira yagarbha, the ruler of the world, after the destruction of the world, their perfect vision having arisen there [in brahma-loka]. In this way, the notion of gradual liberation should be accepted from the text that speaks about non- return.106 One gets a sense from this last comment as well as from his interpretation of the last sūtra in the BS that he trusted this sequence, from becoming disillusioned with sovereignty to developing perfect knowledge and getting liberated, was likely to happen in brahma-loka, but he was explicit that winning brahma-loka meant just avoiding rebirth till the next creation of the world: “It is understood that they return after this kalpa.”107 Sarvajñātman put the two eloquently togetherŚ If a man, his mind full of desiring, attains Brahmā who sits on a lotus [i.e., Hira yagarbha] by a combination of ritual and meditation, he, being ignorant, attains again the human condition without transmigrating [that is, in the next cycle of creation]. If, on the other hand, he comes to know the Supreme Self there [in brahma-loka], he is freed from all bondage. This is the gradual liberation presented in the texts of all Vedic branches, and one which is quite reasonable.108 Thus, attaining brahma-loka became a large gray area in Śa kara’s system, representing simultaneously the top attainment of the abhyudaya course and liberation, albeit gradual, getting 105 jñānāntareṣu ca aiśvaryādi-phaleṣv āsaktāḥ syur maharṣayaḥ. te paścād aiśvarya-kṣaya-darśanena nirvi āḥ paramātma-jñāne pariniṣ hāya kaivalyaṁ prāpur ity upapadyate. BSBh 3.3.32, III.671. 106 kārya-brahma-loka-pralaya-pratyupasthāne sati tatraiva utpanna-samyag-darśanāḥ santaḥ, tad-adhyakṣe a hira yagarbhe a saha, ataḥ paraṁ pariśuddhaṁ viṣ oḥ paramaṁ padaṁ pratipadyante—itītthaṁ krama-muktiḥ anāv tty-ādi-śruty-abhidhānebhyo’ bhyupagantavyā. BSBh 4.3.10, III.831. 107 tasmād asmāt kalpād ūrdhvam āv ttir gamyate. BĀUBh 6.2.15, X.780. 108 yadi vā samuccaya-vaśāt puruṣaḥ kamalāsanaṁ vrajati kāmuka-dhīḥ | punar eva mānavam imaṁ tu vinā parivartam āvrajati mūḍha-matiḥ || athavā sa tatra paramātma-matiḥ parimucyate sakala-bandhanataḥ | krama-yoga-muktir uditā śrutiṣu pratiśākham evam upapannatarā. SŚ 3.50-51. 332 one’s feet wet in niḥ reyasa. We will address the question of brahma-loka as being simultaneously abhyudaya and niḥ reyasa once again at the end of the chapter, relative to the different types of meditation. We should, finally, also note that the statement “liberation is attained solely through knowledge” should be taken with a grain of salt, for some of the k mya things were irredeemable—those explicitly tied to specific objects of desire—whereas other were dependent on the agent and, although constituting a form of action and not knowledge, played a crucial soteriological role as well. The wedge, thus, was not a line of demarcation, but an intersection. The crucial text here was one already known to us, BĀU 4.4.22. This will occupy us in the next chapter, but here is a foretaste: At this point it should be explained here how the whole Veda can be employed to the subject of the Self … By repeating what has been said in this chapter, along with the result, it is attempted to show that the whole of the Veda, sine the corpus dealing with optional rituals (k mya) is to be employed just for this purpose. Thus, a repetition is made starting with the words “That very.” … The whole of the ritual portion of the Veda, with the sole exception of the parts dealing with optional rituals (k mya), is employed through absorption in this knowledge of Brahman laid down in the present chapter starting with “What light” and having the delineated results.109 Brahman as Brahm , the Ultimate Attainment of Meditation It is apposite to end this chapter with a few comments on the ultimate point of the process of meditation and the course of the gods, the northern path. As I said above, it is the so-called lower Brahman or Brahman as the effect, apara-brahman, k rya-brahman. In cosmological terms, these two refer to Hira yagarbha, an individual Self or vijñ n tman whose body is the whole 109 etasminn ātma-viṣaye sarvo vedo yathopayukto bhavati tat tathā vaktavyam. … tac ca yathā asmin prapā hake abhihitaṁ saprayojanam anūdya atraivopayogaḥ k tsnasya vedasya kāmya-rāśi-varjitasya – ity evam-arthaṁ uktārthānuvādaḥ sa v eṣa ity-ādiḥ … kiṁ jyotir iyaṁ puruṣaḥ ity evam-ādi-ṣaṣ ha-prapā haka-vihitāyām etasyāṁ brahma-vidyāyām evaṁphalāyāṁ kāmyaika-deśa-varjitaṁ k tsnaṁ karma-kā ḍaṁ tādarthyena viniyujyate. BĀUBh 4.4.22, IX.641-4. 333 world. They are, however, a broader, complex notion that comprehends the totality of creation.110 As a vijñ n tman or j va and in virtue of possessing a relation to adjuncts, Hira yagarbha is an individual Self just like any other. However, his adjuncts happen to be “very pure.” By “pure adjuncts” Śa kara has in mind the same avidy -k ma-karma complex that individuates any individual Self, but in Hira yagarbha’s case they are attenuated to the degree that this will be his last birth. That principle accommodates the doctrine of Bādari in the Brahma-Sūtra according to which residents of brahma-loka who develop knowledge of Brahman through the process of gradual liberation are liberated along with Hira yagarbha at the expiry of the kalpa. This purity of adjuncts is what earns Hira yagarbha the right to be called Brahman: although he is Brahman to no higher degree than anyone else, it is relative to the purity of his adjuncts that scriptures attribute the appellation “Higher” to him, whereas they describe the (common) individual Selves as liable to transmigration because of preponderance of impurity of adjuncts.111 Hira yagarbha is the first entity that appears in the creation process, and there are several important textual loci that Śa kara relates to his appearance. For instance, he is what is called “death” at the beginning of the first adhy ya of the Bṛhad- raṇyaka: “In the beginning, there was nothing here at all. Death alone covered this completely, as did hunger; for what is hunger but death?”112 He is also that Brahman which appears as the first product in the creation process delineated in the Muṇḍaka UpaniṣadŚ “He is omniscient and knows it all, knowledge is his 110 A quite reliable and thorough presentation of Hira yagarbha in Śa kara’s system is found in a little-known essay by Anam Charan Swain (1971). Statements where Śa kara draws the direct equivalence between k ryaṭapara- brahman and Hira yagarbha are many: see, for instance, MUBh 1.1.8 and 1.1.9, AiUBh 3.3, PUBh 5.5. 111 “Because of the preponderance of purity of adjuncts, ruti and smṛti generally describe him as the higher [Brahman] … whereas in the case of the jīvas, because of preponderance of impurity of adjuncts only transmigration is generally affirmedś” hira yagarbhas tūpādhi-śuddhy-atiśayāpekṣayā prāyaśaḥ para eveti śruti-sm ti-vādāḥ prav ttāḥ. … jīvānāṁ upādhi-gatāśuddhi-bāhulyāt saṁsāritvam eva prāyaśo 'bhilapyateś BĀUBh 1.4.6, VIII.92. 112 BĀU 1.2.1. Śa kara thereonŚ “Death here refers to Hir yagarbha identified with the intellect, because the property of hunger belongs to the Self that is identified with the intellect (the vijñ n tman)ś” buddhy-ātmano 'śanāyā-dharma iti sa eṣa buddhy-avastho hira yagarbho m tyur ity ucyate. VIII.21. 334 creative power. From him are born Brahman, name, form, and food.”113 Finally, he is the second “full” or pūrṇa in the famous mantra of the Bṛhad- raṇyaka: “That over there is full, this over here is full: the full proceeds from the full. Taking the full from the full, only what is full remains.”114 He is also the recipient of the Vedas at the beginning of creation.115 Hira yagarbha is properly the universal soul that animates the whole world as the life- breath or pr ṇa that pulsates through the universe and keeps it together, just as it keeps individuals alive. Indeed, pr ṇa is how Śa kara most commonly identifies him in the BĀUBh. As the life force animating the world, he is commonly called the thread that runs through all things, “like the (invisible) pillars of a house,” for which reason he is also known as the sūtr tman.116 This sūtr tman is the innermost entity in brahma-loka, pervading it just as it pervades the whole world.117 Since he is the first creation of Brahman and himself the cause from which the creation of the five elements as well as the mind and the senses proceeds, he is sometimes described as the cause although himself an effect.118 He is also the collective vijñ n tman, the subtle body of the world, on which all individual Selves are strung through their own subtle bodies.119 Hira yagarbha has attained such a lofty status by being the best practitioner of the path of combined ritual and meditation in his past life, becoming so perfected in it as to be born as its very embodiment: 113 yaḥ sarvajñaḥ sarva-vid yasya jñāna-mayaṁ tapaḥ | tasmād etad brahma nāma rūpam annaṁ ca jāyate. MU 1.1.9. 114 pūr am ahaḥ pūr am idam pūr āt pūr am udacyate | pūr asya pūr am ādāya pūr am evāvaśiṣyate. BĀU 5.1.1. 115 BSBh 1.3.30. 116 Hira yagarbha is called the sūtr tman in BĀUBh 5.5.1. 117 BĀUBh 3.7.1-2. 118 See, for instance, MUBh 2.1.3. 119 sa hy antarātmā li ga-rūpe a sarva-bhūtānām. tasmin hi li gātmani saṁhatāḥ sarve jīvāḥ. tasmāt sa jīva-ghanaḥ. PUBh 5.5, IV.286. 335  Because Prajāpati in his past life, when he was a practitioner, was the first in virtue of practice of cultivation through perfect ritual and meditation among those who aspired for the status of Prajāpati through practicing ritual and meditation, he burnt all faults such as attachment and ignorance that prevent one from becoming Prajāpati before all the other aspirants for the position did.120 Hira yagarbha/pr ṇa also has a gross form, constituted by the world as a totality and by specific heavenly bodies and divinities related to them representing his faculties, for instance the sun and the moon as his sight. A verse from the Muṇḍaka may be given as an instance of what the being of this gross form involves: He is the inner Self of all whose head is fire [Śa kara—heaven], whose eyes are the moon and the sun, whose ears are the directions, whose speech are the revealed Vedas, whose breath is air, whose heart is the whole world and whose feet is the earth.121 Śa kara’s common appellation for this gross manifestation of Hira yagarbha/pr ṇa is piṇḍa, “the ball,” that is, the (general, collective) individual, and other common names for it are Virā — the preferred, standardized term in the commentaries on the MU and PU and in later Advaita Vedānta—Vaiśvānara or the Self common to all, Puruṣa-vidha or the Self that has the shape of a man, Ka the interrogative pronoun, Prajāpati, or even Hira yagarbha. The entire scope of ignorance that has been explained is of two kinds: internally, it is life- breath, the support, like pillars and the like of a house, that which gives light, the immortal. Externally it is characterized as the effect, non-luminous, subject to birth and death, like the grass, ku a, and plaster of the house, denoted by the word “real,” mortal. It was concluded, “By that, pr ṇa (denoted by “immortal”) is covered.” That pr ṇa is manifested variously in different media. Pr ṇa is said to be the one god. Its external persona is one, general, diversified as the sun, etc., designated by various words such as Virāt, Vaiśvānara, the Self in human form, Prajāpati, Ka, Hira yagarbha etc., of which piṇḍa is the main one.122 120 sa ca prajāpatiḥ, atikrānta-janmani samyak-karma-jñāna-bhāvanānuṣ hānaiḥ sādhakāvasthāyā, yad yasmāt, karma-jñāna-bhāvanānuṣ hānaiḥ prajāpatitvaṁ pratipitsūnāṁ pūrvaḥ prathamaḥ san, asmāt prajāpatitva-pratipitsu- samudāyāt sarvasmāt, ādau auṣat adahat. kim? āsa gājñāna-lakṣa ān sarvān pāpmanaḥ prajāpatitva pratibandha- kāra a-bhūtān. BĀUBh 1.4.1, VIII.80. See also BSBh 1.3.30. 121 agnir murdhā cakṣuṣī candra-sūryau diśaḥ śrotre vāg viv tāś ca vedāḥ | vāyuḥ prā o h dayaṁ viśvam asya padbhyāṁ p thivī hy eṣa sarva-bhūtāntarātmā. MU 2.1.4. Śa kara in the avataraṇa to the verse: yo hi prathamajāt prā ād dhira yagarbhāj jāyate ‘ ḍasyāntar virā etc. 122 sa ca vyākhyāto 'vidyā-viṣayaḥ sarva eva dvi-prakāraḥ—antaḥ prā a upaṣ ambhako g hasyeva stambhādi- lakṣa aḥ prakāśako 'm taḥś bāhyaś ca kārya-lakṣa o 'prakāśaka upajanāpāya-dharmakas t a-kuśa-m ttikā-samo 336  Through the process of creation, Hira yagarbha/pr ṇa and Virā /piṇḍa evolve into and comprehend the whole world: on the one side through the evolution of the elements, and on the other through the process of individuation from Virā to the petty creatures. This totality may be identified with what Śa kara calls the world as evolved name and form (n ma-rūpa-vy kṛtaṁ jagat), the world in all its bits and pieces, and it is this which is the broader sense of the term k rya-brahman, Brahman that is the effect. An important manifestation of Hira yagarbha is the sun, with which pr ṇa forms the ground that makes the northern course possible: the pr ṇa that flows through the human body from the channels of the heart forms a continuum with the rays of the sun, the sun itself is the gate that one must pass through at entering brahma-loka, “the golden dish” that has to be removed so that one could see and finally reach the highest divinity of the Vedic form of ignorance, Hira yagarbha himself at the top, the mental or subtle body of the world. The sun, of course, forms a continuum with all forms of fire, sacrificial or otherwise. This helps us understand the set of vidy s or meditations that Śa kara associates with k ryaṭapara-brahman and the attaining of brahma-loka: there is the pañc gni-vidy or the meditation of the five fires that presents the process of transmigration and the paths to the world of the forefathers and the world of the gods; the vai v nara-vidy or the meditation on the Self that is common to all and manifests as the fire of digestion; the meditation on Brahman as Satya, the three-syllabled Brahman that is the person in the orb of the sun; and in general, all meditations on the primacy of g hasyeva satya-śabda-vācyo martyaḥ; tena am ta-śabda-vācyaḥ prā aś channa iti copasaṁhratam. sa eva ca prā o bāhyādhāra-bhedeṣv anekadhā vist taḥ. prā a eko deva ity ucyate. tasyaiva bāhyaḥ pi ḍa ekaḥ sādhāra aḥ—virāḍ vaiśvānara ātmā puruṣavidhaḥ prajāpatiḥ ko hira yagarbha ity-ādibhiḥ pi ḍa-pradhānaiḥ śabdair ākhyāyate sūryādi- pravibhakta-kara aḥ. BĀUBh 2.1.1, VIII.216. 337 pr ṇa, or the life-breath.123 In fact, these are all meditations on pr ṇa, in which one identifies with the World Self and wins the world of Brahman: Now, by proceeding along the northern course, they attain that part of Prajāpati which is Prā a, the eater, and the sun. Through what? By knowing the pr ṇa, the sun, the Self of that which is moving and stationary, through austerity. Specifically, through control of the senses, faith, and meditation on oneself as Prajāpatiś in other words, by meditating “I am pr ṇa, the sun.”124 In religious terms, the lower Brahman is Brahmā the demiurge, who has absorbed in his persona a host of features and names of the central divinity of Vedic ritualism: Hira yagarbha, Ka, Puruṣa, Prajāpati. Śa kara even quotes the first verse of the famous “Who” hymn of the Rig Veda (10.121), where Hira yagarbha makes his grands appearance and which calls him the life- breath of the gods (dev n m asuḥ),125 and draws the equivalence with Brahmā explicitly.126 This k ryaṭapara Brahman is invariantly the highest attainment of all meditations, however, that proceed by way of absorption, even when their cosmological referent is not pr ṇa or the effected Brahman. Śa kara does seem to see a different set of meditations in which it is really the supreme Brahman that is being meditated on and not pr ṇa/Hira yagarbha, but with a 123 This is most obvious in his commentary on the fifth adhy ya of the BĀU. 124 atha uttare a ayanena prajāpateḥ aṁśaṁ prā am attāram ādityam abhijayante. kena? tapasā indriya-jayena. viśeṣato brahmacarye a śraddhayā vidyayā ca prajāpatyātma-viṣayā ātmānaṁ prā aṁ sūryaṁ jagataḥ tasthuṣaś ca anviṣya aham asmīti viditvā ādityam abhijayante abhiprāpnuvanti. PUBh 1.10, IV.245. 125 Under BSBh 1.2.23: hira yagarbháḥ sám avartatā́ gre bhūtásya jātáḥ pátir éka āsīt sá dādhāra pr̥thivī́ṁ dyā́ m utémā́ ṁ kásmai devā́ ya havíṣā vidhema “In the beginning the Golden Embryo arose. Once he was born, he was the one lord of creation. He held in place the earth and this sky. Who is the god whom we should worship with the oblation?” Translation Doniger 1981Ś For the hymn, see Doniger 1981:26-29; Brereton and Jamison 2014:1592-94. 126 Ibid., Śa kara quotes the following verseŚ “The first embodied Self is called puruṣa. It is he, Brahmā, the first creator of beings, who was born in the beginning.” sa vai śarīrī prathamaḥ sa vai puruṣa ucyate | ādi-kartā sa bhūtānāṁ brahmāgre samavartata. Note the echo of the hymn in this verse through the common verb, samavartata: it is he, Brahmā, who was born/arose in the beginning. The same verse is quoted by Sureśvara in BĀUBh 1.2.162. The theme of Brahmā being the first-born, appearing on a lotus stemming from the navel of Nārāya a as he lies on the waters of the causal ocean, is a commonplace in the Pur ṇas. See Dimmitt and van Buitenen 1978:16-29. 338 predication of certain qualities. These include, at the least, the dahara-vidy , the ṇḍilya-vidy in ChU 3.14, and meditations on Oṁ as a symbol of the supreme Brahman.127 The first two are singled out because the qualities that are predicated to Brahman concern directly his causal role. In fact, the second opens with the famous sarvam khalv idaṁ brahma tajjal n nta up s ta, which Śa kara relates directly with the creation, maintenance and destruction of the world, a role not associated with the lower Brahman, but Īśvara, the supreme Brahman in its garb as the cause of the world and its inner ruler (k raṇa-brahman and antary min).128 It is to Īśvara that certain qualities are predicated, not to Hira yagarbha. One does get the sense that these are more directly meditations meant for gradual liberation. For instance, the meditation on Oṁ as a symbol of the higher Brahman delineated in the Pra na-Upaniṣad is said to lead one to brahma-loka, where one would almost seamlessly “see the highest Puruṣa” and be liberated, because liberation depends on “seeing the thing as it is and not depending on fancy.”129 However, the general principle is that whenever absorptive meditation is the process, one that depends on distinctions and in which qualities are intended to be affirmed, the ultimate attainment is always the world of Brahmā because the respective meditations are predicated on duality, and the world of Brahmā is the highest one could get in that sphere. All such meditations, whether concerning pr ṇa/Hira yagarbha or Īśvara, are 127 See BSBh 1.2.1-2; ChU 3.14.4; BSBh 1.3.13. 128 See ChUBh 3.14.4, IV.177: manomaya-ity-ādinā jyāyānebhyo lokebhya ity-antena yathokta-gu a-lakṣa a īśvaro dhyeyaḥ. This is not the place to elaborate on Īśvara. See Hacker 1995 and Comans 2000 for some basic discussion. In general, Īśvara for Śa kara is para-brahman or k raṇa-brahman, which are correlative terms with the aparaṭk rya set. He is the higher Brahman because it is the cause of the lower, effected Brahman in its totality. His key attributes are omniscience and omnipotence, through which he is related to the creation of the world (BSBh 1.1.5), and being the internal ruler behind everything, including pr ṇa or the sūtr tman (BĀUBh 3.8.12), through which feature he is related to all individual Selves as their real Self. In religious terms, he is Nārāya a (īśvaro nārāya ākhyaḥ, BĀUBh 3.7.3, IX.432). 129 BSBh 1.3.13. 339 encompassed under the appellation of saguṇa-vidy , up sana, meditation on Brahman as possessing qualities. The adhikarins for meditation, to which ritual may be added to form a jñ na-karma- samuccaya, a combination of “knowledge” and “action,” are householders who have been instructed in some such meditation and practice it, as well as renunciants and hermits.130 This is, then, a legitimate form of jñ na-karma-samuccaya, but jñ na does not stand for knowledge of Brahman or the Self. Although it is based on some of the famous Upaniṣadic passages that are associated with Śa kara in the scholarly imaginaire, such as sarvaṁ khalv idam brahma, it is not knowledge but meditation, because in these texts Brahman is presented as an object of meditation, as if possessing qualities which it does not, in fact possess. I will elaborate on this in Chapter Eight, and our business with meditation is not nearly over. We may conclude this chapter with the observation that this process of meditation-cum- ritual is fully compatible with the BS account of combining jñ na and karma. The problem with it was, of course, that it did not lead to liberation. It brings one close, or, as Śa kara says, “its results are proximate to liberation,”131 but it was not the real thing. In any case, Śa kara’s beef with the jñ na-karma-samuccaya was not with this variation, whose value both he and Sureśvara affirm whenever possible. 130 ChUBh 5.10.12ś BĀUBh 6.2.15. 131 kaivalya-sannik ṣ a-phalāni, ChUBh Introduction, IV.9. 340  CHAPTER SEVEN: LIBERATION, RITUAL, AND THE ARISING OF KNOWLEDGE Sureśvara ✅ @realMa ḍanaMiśra FAKE NEWS about liberation in the @failing ŚV. Very un-Vedic. So- called Mīmāṁsaka the crooked Kumārila has been writing lies for 30 years. Bad (or sick) guy! Introduction We saw in the previous chapter that Śa kara divided the attainments which the Veda can provide for men in four spheres, loka, and tied to them a specific means or s dhana, an appropriate qualification or adhik ra, and a corresponding desire, k ma. From this point on, we will pursue in some length, in three chapters, the same scheme regarding liberation, the final and, in this case, ultimate attainment: the sphere, loka, that is the Self. A crucial role in the attainment of liberation, and in opposition to the accounts of liberation that we have examined in the second part of the dissertation, was played by the few sentences in the Upaniṣads which I will call “identity statements,” and which Śa kara most commonly referred to as “tat tvam asi and the rest.” Liberation was supposed to follow as a matter of course upon the full understanding of these identity statements, and they on their part organized whole sections of the Upaniṣads in a hierarchy geared towards facilitating such understanding. But, this understanding generally could not take place without some process of clarification of meaning, and one could not engage in such clarification without first satisfying some existential criteria. I will focus on the identity statements, the Upaniṣads, and the process of understanding in chapters Eight and Nine, and here I will deal first with the general nature of liberation and the necessary preliminaries. While scholars have written extensively on elements of this topic—and I will refer to the most important studies in my notes—I do have a rather major historical argument to make here, 341 one that is novel but, I contend, most useful for understanding Śa kara’s soteriologyŚ Śa kara developed the whole path to liberation generally, and the section to the state which he called “arising of knowledge” or jñ notpattiṭvidyotpattiṭjñ na-pr pti specifically, on the model of ritual causality of the Bhā a Mīmāṁsā system, drawing particularly on the ideas of mediate causality or p ramparya and the two kinds of non-major ritual actions, the r d-upak rakas and the sannip tyopak rakas, or the direct and aggregated helpers. A third important Mīmāṁsā principle which he also used with profit, but in conjunction with the aforementioned two and, thus, very differently than other contemporary Vedāntins, was the idea of trans-instrumentalization or repurposing of ritual and other rama practices for the needs of liberation. Reading Śa kara through the lens of the account that I will present has the benefit of seeing coherence in his system, particularly in his attitude toward ritual that I marked as puzzling in the introduction to this part of the dissertation. Early Advaitins have mapped the path to liberation in various degrees of detail. Most systematic in this was Sureśvara, whose scheme in the Naiṣkarmya-Siddhi 1.52 runs as follows: performance of ritual and rama practices (nitya-karma) → acquisition of merit (dharmotpatti) → destruction of bad karma (p pa-h ni) → purity of mind (citta- uddhi) → understanding the nature of bondage (saṁs ra-y th my vabodha) → dispassion (vair gya) → desire for liberation (mumukṣutva) → searching for a means (tad-up ya-paryeṣaṇam) → practice of Yoga (yog bhy sa), which his commentator Jñānottama rightly glosses as the practice of the three methods of ravaṇa, manana and nididhy sana, but which most certainly involved the Yogic practices of yama-niyama as well → inclination of the mind toward the inner Self (cittasya pratyak-pravaṇat ) → full understanding of the identity statements (tat tvam asy- di-v ky rtha- parijñ na) → destruction of ignorance (avidyoccheda) → remaining as the Self (sv tmany 342 avasth na). Sureśvara developed this scheme to show how ritual becomes a means of liberation “mediately,” through a succession in which every result becomes a means for another result and culminates vicariously in the ultimate attainment, exactly like Kumārila’s doctrine of how the understanding of the Veda or the threshing of the rice culminate in heaven mediately. Śa kara also offered an itinerary to liberation, which was, however, organized on important juncture points. In BhGBh 5.12, he says that one attains liberation through four successive stages: purity of existence (sattva- uddhi); acquisition of knowledge (jñ na-pr pti); renunciation of all action (sarva-karma-sanny sa); and steadfastness in knowledge (jñ na- nisṭh ).1 We will take these schemes as orientation points, but I will work out the path to liberation by looking at Śa kara’s wider corpus and by ascertaining what the stated juncture points involve. Along the way, I will also address points from the competing theories of liberation that still need addressing. The Self and the Nature of Liberation Śa kara defined liberation in a couple of related ways, depending on whom he was arguing against, but his most persistent and final definition was that liberation is just remaining in the state of being the Self: Therefore, liberation is remaining in one’s Self, [a state that follows] when ignorance that is the ground of desire and action has ceased.2 For, attaining the world that is the Self is just remaining in one’s Self at the cessation of ignorance.3 Liberation is the state of being the pure Self when the complex of the factors of individuation is removed. It comes about when the ground of individuation, ignorance or avidy that produces 1 Cf. Bader 1990:45-64. 2 tasmād avidyā-kāma-karmopādāna-hetu-niv ttau svātmany avasthānaṁ mokṣa iti. TUBh Introduction, VI.10. 3 ātma-loka-prāptir hy avidyā-niv ttau svātmany avasthānam eva. BĀUBh 4.4.22, IX.647. 343 the category of the individual Self, has been destroyed, at which point the other two factors that further give rise to the innumerable individual Selves—desire and action—fall off by themselves. In this sense, liberation is synonymous with the Self or with Brahman, and its definition is identical with the definition of Brahman: This one, on the other hand [as opposed to the permanently changing] is permanently changeless in the absolute sense, all-pervading like space, devoid of all transformation, ever content, partless, self-effulgent in nature. It is the state of being unembodied, called liberation, where good and bad karma along with their results, as well as the three periods of time, have no continuation.4 Liberation is, in fact, most directly said to be BrahmanŚ “For, in all Upaniṣadic texts liberation is ascertained as uniform. The state of liberation is, in fact, Brahman itself.”5 However, because liberation happens when ignorance has ceased, logically if not temporally, it was also possible to define it in negative terms: Liberation cannot be brought into being. For, it is nothing more than the destruction of bondage, and it is not producible. And, as we have just said, bondage is ignorance, and destroying ignorance by action is not possible.6 The most sustained presentation of liberation in such negative terms is in the USP 16 and 18. There bondage is defined as a cognitive error that consists in wrong superimposition, specifically the superimposition of the property of knowing to the intellect or buddhi in the manner of attributing the idea of a snake over a rope, and the failure to distinguish the intellect from the Self. Bondage is ignorance. Release happens when this cognitive error is undone. Such release does not constitute an attainment of a different state on the part of the Self; rather, it is a form of 4 idaṁ tu pāramārthikaṁ kū astha-nityaṁ vyoma-vat sarva-vyāpi sarva-vikriyā-rahitaṁ nitya-t ptaṁ niravayavaṁ svayaṁ-jyotiḥ-svabhāvam, yatra dharmādharmo saha kārye a kāla-trayaṁ ca nopāvarteteś tad etad aśarīratvaṁ mokṣākhyam. BSBh 1.1.4, I.20. 5 mukty-avasthā hi sarva-vedānteṣv eka-rūpaiva avadhāryateś brahmaiva hi mukty-avasthā. BSBh 3.4.52. Cf. Warrier 1961:469-75; Nelson 1996:19-20. 6 anārabhyatvān mokṣasya; bandha-nāśa eva hi mokṣaḥ, na kārya-bhūtaḥ; bandhanaṁ ca avidyety avocāmaś avidyāyāś ca na karma ā nāśa upapadyate. BĀUBh 3.3.1, IX.386. 344 anamnesis that can only figuratively be ascribed to the Self. The Self is neither an agent nor a patient, and the whole talk of attaining liberation as a goal, of means to attaining that goal and the like, is figurative in any case.7 Liberation is a s dhya or pr pya like the other three attainments, but only in a manner of speaking. The witty Sureśvara says that it is like getting the necklace that has been on one’s neck all along, or escaping the demon that is one’s own shadow. Neither of the two are real attainment or avoidance, yet something does happen on both occasions.8 One has forgotten a true state of affairs—logically again, not temporally—and needs to be reminded. To get the necklace, one simply has to get it. Defining liberation as the cessation of ignorance at knowing oneself as Brahman also meant promoting liberation as something that brings visible results, dṛṣṭa, that is, readily available and apparent, and not something one will experience in the hereafter, like heaven or brahma-loka. That knowledge had visible results was the accepted norm in Vedic theology, and the study of the Veda itself was commonly described as ending in visible results. Presenting liberation as dṛṣṭa had some significant theological consequences, since it immediately disqualifies the major forms of ritual action—the rituals as units—from being direct means to liberation. Rituals by hypothesis produce invisible, future results, and liberation is not of that kind. This was, in fact, one of the reasons why Advaita Vedānta stood so uncompromisingly by 7 “Bondage is a confusion of the intellect, and liberation is destruction of this confusion,” buddher bhrāntir iṣyate | bandho mokṣaś ca tan-nāśaḥ. USP 18.59 “The intellect, illuminated by the light of the sentient Self, thinks that it possesses sentience. That is its confusion. Because sentience is the nature of the Self, it is commonly applied to the intellect figuratively. This absence of discriminative knowledge is beginningless, and there is nothing more to transmigration than this. Liberation is the destruction of this error, it cannot be more than this.” bodhātma-jyotiṣā dīptā bodham ātmani manyate | buddhir nānyo ‘sti boddheti seyaṁ bhrāntir hi dhī-gatā || bodhasyātma-svarūpatvān nityaṁ tatropacaryate | aviveko ‘py anādyo ‘yaṁ saṁsāro nānya iṣyate || mokṣas tan-nāśa eva syān nānyathānupapattitaḥ. USP 16.60-62ab. See also USP 18.45-6, 107. 8 NaiS 1.31-4. 345 the problematic ideal of j van-mukti or liberation while living, and presented every attainment along the course as a visible result. That liberation is a visible result is often the unstated factor behind Śa kara’s claim that there are only four kinds of action—production, attainment, transformation, and refinement, the four sannip tyopak rakas—and that none of them can operate over the Self and bring about liberation. What he means is that only these types of ritual Vedic actions bring about visible results of the kind to which liberation belongs, not that Agnihotra or playing marbles are not action at all.9 In terms of strict soteriological causality, however, the cessation of ignorance was the final cause of liberation, not what liberation as a state was, and this formulation is better understood as a causal explanation rather than a proper intensional definition. Śa kara resorts to this causal definition when arguing against fellow theologians who would like liberation to be a novel, future state of affairs, producible in the manner of heaven, or against other competing doctrines, when it is important to bring home the idea that liberation is just what happens when bondage, its opposite, is no more.10 It was, nevertheless, important to affirm, specifically against Kumārila, that liberation was a state, not a negative one that consists merely in absence of the cause of embodiment, but a positive state equivalent to the Self. It wasn’t another or a future 9 See, for instance, BĀUBh 3.3.1Ś “And, it is not possible that ignorance be destroyed by action. The operation of action extends over the sphere of the visible, for the domains of the operation of action are production, attainment, transformation, and refinement. There is no domain of action other than the faculty of production, attainment, transformation, and refinement, as is well-known in the world. And, liberation is none of these four categories. That is why I just said that it is obstructed by ignoranceś” avidyāyāś ca na karma ā nāśa upapadyate, d ṣ a-viṣayatvāc ca karma-sāmarthyasyaś utpatty-āpti-vikāra-saṁskārā hi karma-sāmarthyasya viṣayāḥś utpādayituṁ prāpayituṁ vikartuṁ ca sāmarthyaṁ karma aḥ, na ato vyatirikta-viṣayo 'sti karma-sāmarthyasya, loke 'prasiddhatvātś na ca mokṣa eṣāṁ padārthānām anyatamaḥś avidyā-mātra-vyavahita ity avocāma. Also, in detail in BSBh 1.1.4. 10 For this reason, Mayeda’s account of Śa kara’s understanding of liberation (2006b:73-5), that is, liberation just as the cessation of ignorance, is incomplete and imprecise, as it relies solely on the US and completely misses the numerous passages where Śa kara defines liberation as a state that ensues upon the cessation of ignorance. Particularly problematic is the conclusion that “Śa kara's concept of final release is very similar to the Mahayana Buddhist view of nirv ṇa, characterized by Candrakīrti as 'being of the nature of destruction of all false assumptions' (sarvakalpan kṣayarūpa).” While this is procedurally true—liberation is achieved by undoing all concepts—to Śa kara’s mind it certainly was not true regarding liberation as a state. 346 state, but a present state of which one was not aware. The positive definition was important when arguing against Kumārila’s first account of liberation. We will remember here that the negative account of liberation said that freedom from rebirth was like the absence of a broken pot: the pot had a history but no future, and its absence, though brought into being, was subsequently eternal. To put it differently, its absence had no history, but had future. This was a form of liberation that was not acceptable to Vedāntins in general, and to Advaitins in particular, because it smacked of asatk rya-v da. The subsequent absence of the pot was just a figurative absence, because Being continued to be a positive remainder after the breaking of the pot. Being as a positive and unitary thing could assume distinguishing features, such as the shape of a pot and the action of a pot production, and then be separated from them, and the two states could respectively be described as coming into the being of a pot and its absence or destruction, but this was a form of vikalpa or mental construct. The important point, though, was that Being allowed such imagination to take place, whereas absence, a mere nothing, could assume no qualities or actions. It was absolutely non-relational. Properly speaking, there was no absolute future absence when the pot was broken. To define liberation, therefore, as absence of future embodiment—which was, we will remember, the key move on Kumārila’s part in the argument that liberation could not be enjoined as it does not admit of productive striving—was just not sound reasoning. Liberation was not absence: it was presence of the only thing that could be present, the Self.11 11 yad dhi naṣ am, tad eva notpadyata iti pradhvaṁsābhāva-van nityo 'pi mokṣa ārabhya eveti cet— na, mokṣasya bhāva-rūpatvāt. pradhvaṁsābhāvo ‘py ārabhyata iti na sambhavati, abhāvasya viśeṣābhāvād vikalpa- mātram etat. bhāva-pratiyogī hy abhāvaḥ. yathā hy abhinno 'pi bhāvo gha a-pa ādibhir viśeṣyate bhinna iva gha a- bhāvaḥ pa a-bhāva iti, evaṁ nirviśeṣo 'py abhāvaḥ kriyā-gu a-yogād dravyādi-vad vikalpyate. na hy abhāva utpalādi-vad viśeṣa a-saha-bhāvī. viśeṣa avattve bhāva eva syāt. TUBh Introduction, VI.10. 347  This was an important move, because the pursuit of liberation was an action and a form of striving, and such striving had to be prompted by a desire for something positive. Śa kara could now present the pursuit of liberation as striving for the Self, on the part of an tma-k ma or one with a sole desire for the positive Self. This Self, however, was not the satya-k ma, the Self that was the repository of all good desires. It was the Self all whose desires were fulfilled because it did not have any. It was Yājñavālkya’s Self. Because the Self did not have any desires, he who aspired just after that Self was eo ipso without the desires that were satisfied by ritual and similar means. One could at the same time have the desire for the Self, be prompted to action in the proper Vedic manner and with the adequate adhik ra, and be properly without desires as expected from an aspirant after liberation, simply because the Self did not have desires.12 Let us mark these three final steps that are in a logical sequence as the consummation of the path to liberation: full knowledge of Brahman, causing the destruction of ignorance, causing the state of being the Self. That there was a temporal break between them is explicitly denied in BSBh 1.1.4 with a most impressive arsenal of Upaniṣadic statements, but the sequence is affirmed consistently.13 The Role of Ritual and Vividi Kumārila’s first account promoted the obligatory rituals as the direct causal factor in the attainment of liberation. Ritual had to prevent the creation of future bad karma that would occur through non-performance, and after the karmic stock had been exhausted through experience, 12 karma-hetuḥ kāmaḥ syāt, pravartakatvāt. āpta-kāmānāṁ hi kāmābhāve svātmany avasthānāt prav tty-anupapattiḥ. ātma-kāmatve cāptakāmatā. ātmā ca brahma. tad-vido hi para-prāptiṁ vakṣyati. ataḥ avidyā-niv ttau svātmany avasthānaṁ para-prāptiḥ. Ibid, p.8. 13 api ca brahma veda brahmaiva bhavati [MU 3.2.9], kṣ yante c sya karm ṇi tasmin dṛṣṭe par vare [MU 2.2.8], nandaṁ brahmaṇo vidv n na bibheti kuta cana [MU 2.9], abhayaṁ vai janaka pr pto’si [BĀU 4.2.4], tad tm nam ev ved haṁ brahm sm ti, tasm t tat sarvam abhavat [BĀU 1.4.10], tatra ko mohaḥ kaḥ oka ekatvam anupa yataḥ [ĪU 7] ity evam-ādyāḥ śrutayo brahma-vidyānantaram eva mokṣaṁ darśayantyo madhye kāryāntaraṁ vārayanti. I.21. 348 one would be automatically liberated. In the Bṛhat-Ṭ k , the obligatory rituals were given even a more prominent role: they would not only prevent, but also exhaust some present bad karma, and thus speed up the process. Śa kara, as we saw in the previous chapter, rejected this account, arguing that ritual as a form of action was predicated on ignorance and desire—action was the third in line among the individuation factors. I said that Kumārila’s account presupposed that ignorance and desire had already ceased through the discriminative knowledge of the Self, but for Advaitins that just could not be the caseŚ if ignorance had ceased, so would have agency. Sureśvara’s causal chain of saṁs ra was particularly explicit about the place of any form of action, ritual included. The human condition was that of suffering, which was consequent on embodiment. The body was a result of good and bad karma, which was a result of deeds prohibited and enjoined in scripture. Action was predicated on attachment and aversion—the staple forms of desire—which were false ideas of what was pleasant and unpleasant. Such ideas had root in the uncritical acceptance of duality. At the bottom of this was the lack of understanding of the self-evident non-dual Self. Action that is not prompted by desire was just not possible. By performing the so-called obligatory rituals, one at the least hopes to attain the pleasant and avoid the unpleasant, and these are intrinsically related to r ga and dveṣa.14 Śa kara, further, rejected Kumārila’s claim that the failure to perform the nitya- naimittika rituals would create bad karma, pratyav ya. He argued that non-performance was a form of absence, nonexistence or abh va, and that as such it could not create positive effects.15 14 NaiS Introduction; TUBhV 1.6-8. 15 See, for instance, TUBh Introduction, VI.9Ś “The non-performance of the obligatory rituals is an absence, and therefore bad karma [as a result of that] does not make senseś” nityānāṁ ca akara am abhāvaḥ tataḥ pratyavāyānupapattiḥ. In all fairness, we should say that Kumārila had anticipated this objection—or perhaps it was already explicitly made, possibly by the Vājasaneyins who, as we have seen, argued that one could take to the life of 349 However, he took a shine to Kumārila’s claim that the so-called obligatory ritual could exhaust accumulated bad karma.16 To be sure, there was nothing obligatory about these rituals, and they were not really niṣk ma or performed without any expectations, but they had the good characteristic of being bound to the general good that a Vedic ritual can bring, namely heaven. They were not explicitly tied to specific desires, such as for cattle or virility, and so they could be repurposed through a Vedic fiat and a fitting desire on the part of the performer. We have already seen this repurposing procedure in Kumārila’s second account, with the knowledge of the Self, and in the BS, with ritual. I called it trans-instrumentalization. With Śa kara’s characteristic eloquence and clarity, this trans-instrumentalization comes into sharp focus. Ritual that is enriched with two elements—the Upaniṣadic meditations and the absence of interest in the objects that ritual otherwise brings, abhisandhi—produces a special effect, different from its common results, just as poison that otherwise causes death can be a cure when accompanied by a charm, or as thick sour milk that causes fever can have calming effect when renunciation immediately after the study of the Veda, without ever lighting the fire, which was endorsed by Bādarāya a against Jaimini—and replied to in the B : karmanāṁ prāg-abhāvo yo vihitākara ādiṣu | na cānartha-karatvena vastutvān nāpanīyate || sva-kāle yad akurvaṁs tad karoty anyad acetanaḥ | pratyavāyo ‘sya tenaiva nābhavena sa janyate. “The anterior absence of ritual actions, when such rituals are enjoined but not performed, does not lose substantiality because of causing something undesired. The unthinking man who does not perform them when he should, does [at that time] something else, and his bad karma is a result of that, not of absence.” Verses 2 and 3 in Taber’s excerpt (Taber 2007:182). Śa kara’s theory was that while not performing the obligatory rituals, one was creating fresh bad karma as a consequence of past accumulated bad karma, presumably by engaging in other things—thus following Kumārila’s polished argument—while the non-performance itself was just an indication that such thing was happening, “for otherwise there would be an origination of something positive out of mere absence, in contradiction to all reliable warrantsś” ataḥ pūrvopacita-duritebhyaḥ prāpyamā āyāḥ pratyavāya-kriyāyā nityākara aṁ lakṣa am iti śat - pratyayasya nānupapattiḥ – akurvan vihitaṁ karma [Manu 11.44] iti. anyathā hi abhāvād bhāvotpattir iti sarva- pramā a-vyākopa iti. TUBh Introduction, VI.9. Another prominent argument against nitya-naimittika was that they could not fully exhaust the bad karmic stock either, since it was endless. TUBh 1.11.4. 16 nityāny adhigatāni karmā i upātta-durita-kṣayārthāni. TUBh Introduction, VI.8. 350 sugar is mixed in it.17 A third element is mentioned additionally in the BGBh, and that is dedication of all action to the Lord.18 This special effect of ritual performed not for heaven but as a way of worship of the Lord is purification of one’s existence, called variously sattva- uddhi, tma-saṁsk ra, etc. This procedure also changes the type of ritual as action, and of the result that ritual brings. Ritual does not bring adṛṣṭa results, something necessarily experienced in a future life, but dṛṣṭa, results that are palpable here and now; it becomes a form of saṁskṛti, one of the four sannip tyopak rakas: And, I just said that actions are in the domain of one who does not know. For, actions of the type of production, obtainment, transformation, and refinement/purification are in the domain of ignorance. That is why I said that actions become a means of knowledge though purification of the Self.19 Śa kara will not call ritual a sannip tyopak raka for a different reason, however, as we will see by the end of the chapter. Let us note here that Śa kara expected those who are intent on liberation to take their ritual “with cream and sugar.” Such purification of existence or refinement of the Self consists in the removal of psychological faults that result from bad karma and that block the knowledge of the Self. In short, performance of ritual leads to what Śa kara calls the arising of knowledge, jñ notpatti or vidyotpatti. And, because of this, such obligatory ritual is properly a means of liberation.20 We 17 yat punar uktam, vidyā-mantra-śarkarādi-samyukta-viṣa-dadhy-ādi-van nityāni kāryāntaram ārabhanta iti – ārabhyatāṁ viśiṣ aṁ kāryaṁ, tad-iṣ atvād avirodhaḥ; nirabhisandheḥ karma o vidyā-saṁyuktasya viśiṣ a- kāryāntarārambhe na kaścid virodhaḥ … yeṣāṁ punar nityāni nirabhisandhīny ātma-saṁskārārthāni teṣāṁ jñānotpattyarthāni tāni. BĀUBh 3.3.1, IX.391, 393. See also BĀUBh 4.4.22. 18 Called variously īśvarārpa a, īśvara-samarpa a, īśvarārādhana, īśvarārtham karma etc., and omnipresent in the text. 19 karma āṁ ca avidvad-viṣayatvam avocāmaś avidyā-viṣaye ca utpaty-āpti-vikāra-saṁskārthāni karmā īty ataḥ – ātma-saṁskāra-dvaire a ātma-jñāna-sādhanatvam api karma ām avocāma – yajñādibhir vividaṣantīti. BĀUBh 4.5.15, IX.676. 20 evaṁ kāmya-varjitaṁ nityaṁ karma-jātaṁ sarvam ātma-jñānotpatti-dvāre a mokṣa-sādhanatvaṁ pratipadyate. BĀUBh 4.4.22, IX.646. 351 need to be mindful of several things in this regard, however, and we need to consider what this arising of knowledge is. The performance of ritual, consisting mostly of the daily Agnihotra, is now just one of the actions that are pertinent to the members of the varṇ rama social system. It is important to bear this in mind: when talking about karma, Śa kara commonly has ritual in mind, but ritual is just one of several possible practices a participant in the varṇ rama would engage in, and it is pertinent only for householders. Let us pay some attention to the following passage from the TUBh: – If that is the case, then the other ramas are irrelevant, since the arising of knowledge is caused by action [through removing the hindrances] and actions are enjoined for householders only, from which fact it follows that there can be only one rama. Thus, the statements that Agnihotra should be performed as long as one lives are all the more apposite. – No, because actions are many. It is not that only Agnihotra and the like are actions. There are actions that are associated with the other ramas such as chastity, austerity, truthfulness, calmness, self-control, non-violence, etc., as well as actions that are characterized by concentration, meditation, etc., that are the best for the arising of knowledge, because they are unadulterated.21 The nitya-karmas are now not the daily rituals that every Vedic man should perform, but whatever the members of the individual ramas do. Other practices are, in fact, better than disinterested ritual, because of being “unadulterated,” “unmixed.” Ānandagiri glosses this with “because they are not mixed with violence and the like,” obviously alluding to ritual slaughter. 22 This is well-corroborated in BĀUBh 4.5.15, where Śa kara offers a similar reasoning and claims that the obligatory actions of those who have gone forth are better suited for the rising of knowledge than ritual, which is mixed with violence, attachment, aversion, etc. Let us remember 21 evaṁ tarhi āśramāntarānupapattiḥ, karma-nimittatvād vidyotpatteḥ. g hasthasyaiva vihitāni karmā īty aikāśramyam eva. ataś ca yāvaj jīvādi-śrutayaḥ anukūlatarāḥ syuḥ. naś karmānekatvāt. na hy agnihotrādīny eva karmā iś brahmacaryaṁ tapaḥ satya-vacanaṁ śamaḥ damaḥ ahiṁsā ity-evam-ādīny api karmā i itarāśrama- prasiddhāni vidyotpattau sādhakatamāny asa kīr ā vidyante dhyāna-dhāra ādi-lakṣa āni ca. TUBh 1.11.4, VI.51-2. 22 asa kīr atvād hiṁsādy-amiśritatvād ity arthaḥ. Ānandagiri on TUBh 1.11.4, p.35. 352 the dual nature of ritual: it is commonly associated with desire, wealth, etc., and only exceptionally with the arising of knowledge. Thus, the arising of knowledge—and what that is we will discuss shortly—may happen through any of the obligatory practices of the respective ramas, so long they are scriptural, in addition to some factors that are not even related to the rama system, such as the grace of the Lord ( vara-pras da). Śa kara is not always clear whether all the obligatory practices of the respective ramas give rise to knowledge specifically through depleting the bad karmic stock. In the TUBh, for instance, he says that there is no such rule that knowledge arises just from the exhaustion of bad karma and not from practices such as non-violence, chastity, austerity, meditation, as well as from the grace of the Lord. All these are, in any case, only helpers to the triple Vedāntic process of hearing, thinking and meditation.23 In the BĀUBh 4.4.22, on the other hand, the most important comment on this matter, purity of existence that comes from the depletion of bad karma is associated specifically with the obligatory practices of the three ramas other than those who have gone forth, namely the recitation of the Veda, sacrifice, charity, and austerity. In the path to liberation, such purity of existence is an important, threshold step, and one of the cardinal points in Śa kara’s soteriology.24 We may, thus, surmise that the obligatory rama practices occasion the arising of knowledge through engendering purity of existence first, while 23 naś niyamābhāvāt. na hi, pratibandha-kṣayād eva vidyotpadyate, na tv īśvara-prasāda-tapo-dhyānādy-anuṣ hānāt iti niyamo 'sti; ahiṁsā-brahmacaryādīnāṁ ca vidyāṁ praty upakārakatvāt, sākṣād eva ca kāra atvāc chrava a-manana- nididhyāsanādīnām. TUBh 1.11.4, VI.56. 24 His thinking in this regard was most likely shaped by the Bhagavad-G t , specifically 18.5-6, which says that sacrifice, charity, and austerity should never be renounced, as they purify man, but should be performed with disregard for their result. 5.11 is also relevant, introducing the term tma- uddhi, and saying that yogis perform action with body, mind, words, and purified senses, but without attachment. The term sattva- uddhi or purity of existence was also likely taken from sattva-saṁ uddhi in 16.1, although it appears also in the ChU 7.26.2 353 the practices of those that have gone forth, as well as factors such as the grace of the Lord, do so directly.25 That these ritual and other rama practices give rise to knowledge of the Self can also be formulated in the strict terms of Vedic theology. The two key passages in this regard are already known to us from the BSŚ they are BĀU 4.4.22 and 23, and particularly significant is what Śa kara has to say about them under BS 3.4.26-7. The first passage said that Brahmins seek to know (vividiṣanti) the great Self through recitation of the Veda, sacrifice, charity, and austerity. Once they know the Self, they become sages and go forth as renunciants desiring that Self.26 For Śa kara, the statement presents two consecutive processes. The first is the striving after knowing the Self, vividiṣ , and that striving proceeds by performance of the rama practices. These are now canonized as the daily recitation of the Veda, sacrifices, charity, and austerity, to which Śa kara adds chastity.27 This list, now, corresponds fully to ChU 2.23.1: There are three types of persons whose torso is the Law (dharma). The first is one who pursues sacrifice, vedic recitation, and gift-giving. The second is one who is devoted solely to austerity. The third is a celibate student of the Veda living at his teacher's house—that is, a student who settles himself permanently at his teacher's house. All these gain worlds earned by merit. A person who is steadfast in brahman reaches immortality.28 25 Bradley Malkovsky has written a full book and an article (2000a, 2001) on the role of divine grace in Śa kara’s Vedānta, but other than pointing out that there is such a thing, and a prominent one, in Śa kara’s system, his work has little value for understanding the precise role of such divine grace. This is largely because of failing to comprehend that the arising of knowledge is relatively an early threshold in Śa kara’s soteriology, and not what directly brings down avidy . In terms of strict soteriological contribution, the role of divine grace is exercised before one can begin engaging in brahma-vidy . 26 tam etaṁ vedānuvacanena brāhma ā vividiṣanti yajñena dānena tapasānāśakena. etam eva viditvā munir bhavati | etam eva pravrājino lokam icchantaḥ pravrajanti. The pronouns refer to “sa vā eṣa mahān aja ātmā yo 'yaṁ vijñānamayaḥ prā eṣu” in the beginning of the passage. 27 Under BĀU 4.4.22, Śa kara reads an akena as an adjective for tapas , not a fifth independent action. That is, since tapas can refer to bodily mortification of any kind and an aka to starving that ultimately ends in death, the second should be read as an adjective of the first to give the meaning of not enjoying the objects of desire, k m n ana. In the BSBh 3.4.26, ChU 8.5.1 is quoted to the effect that chastity is equivalent to sacrifice. 28 Translation Olivelle 1998:197. 354 There are, then, five rama practices listed here, and from the comment on the cited passages (ChU 2.23.1) we can relate them to the respective ramas: (1) Chastity or brahmacarya for the student who lives with his teacher his whole life. Śa kara is clear that this does not include the regular student life, after which one must make the choice of an rama as a vocation, as Olivelle has shown, but for the life-long student whose vocation is to serve the teacher;29 (2) Sacrifice or yajña, consisting primarily of the daily Agnihotra; charity or d na; and Vedic recitation or adhy yana for the householder; (3) Austerity or tapas for an ascetic (t pasa) or a mendicant (parivr ṭ). The mendicant is, again, one who has made such vocational choice, to be distinguished from another category of mendicant whom we will soon see. The vocation consists in observing vows such as the c ndr yaṇa fast. All these rama practices eventually lead to the arising of knowledge, as we have seen, if they are not performed for other gain; that is, if they are repurposed. Otherwise, they all normally bring some form of good karma or puṇya that belongs to the sphere of promotion.30 Now, there is nothing inherently or naturally prophylactic about these practices. They are not like acetone for nail polish that removes stains just because such is its constitution or the way it reacts with other substances: we saw that nitya-naimittika-karma was generally for heaven. Rather, there is an injunction in the BĀU passage of the utpattiṭapūrva type, which, as we will remember from the Second Chapter, introduces a ritual undertaking for a specific purpose by disclosing a causal relation that is not knowable except from a Vedic statement. These actions 29 Olivelle 1993. From the BGBh, it appears that the so-called sm rta practices or the duties of the respective classes from the 30 Dharma- stras can be added to this class. 355 are for knowing the Self just because the Upaniṣad says so. To spell this out as clearly as we can, the injunction relates the rama practices to the state of appearance of knowledge of the Self as means to a result, just as an injunction for, say, Darśa-pūr amāsa relates the sacrifice to the appearance of heaven as means to a result. The practices of the ramas bring about vidyotpatti because there is an utpatti-v kya, an originative injunction to that effect. Performance of ritual thus becomes the starting point in the pursuit of liberation. The injunction, of course, informs: it does not command.31 And then, there comes a point where such knowledge of the Self has arisen, one has become an tma-vit, and the process is no longer vividiṣ . The absolutive in BĀU 4.4.22 marks that break clearly, etaṁ taṁ viditv , and introduces something new. This new thing is that one desires the Self and goes forth as a renunciant. This renunciant is different from the vocational renunciant. He corresponds to the brahma-saṁstha from the ChU, one who is steadfast in Brahman, and Śa kara clearly distinguishes him from the ordinary mendicant whose vocation is to do rites of austerity.32 In other words, Śa kara thinks that there comes a point, the arising of knowledge, where one gives up the rama practices that have made one a knower of the Self.33 – But, is it not contradictory to say that knowledge depends on the rama practices and that it does not depend on them at the same time? – No, we say! Knowledge, once arisen, does not depend on anything for fruition. However, it does depend for its arising.34 31 Cf. BSBh 3.4.27, III.739Ś “– But, I said that we do not see an injunction in the text ‘they aspire to know through sacrifice etc.’ [BĀU 4.4.22]. – You surely did say that. Nevertheless, because the relation is unprecedented, an injunction should be postulated. For, this relation of the desire to know to sacrifice and the other practices does not obtain before the statement, by which fact [of obtaining otherwise] it could be a restatement.” nanu uktam— yajñādibhir vividiṣantīty atra na vidhir upalabhyata iti—satyam uktamś tathāpi tu apūrvatvāt saṁyogasya vidhiḥ parikalpyate; na hi ayaṁ yajñādīnāṁ vividiṣā-saṁyogaḥ pūrvaṁ prāptaḥ, yenānūdyeta. 32 tapa eva dvitīyas tapa iti k cchra-cāndrāya ādi tadvāṁs tāpasaḥ parivrāḍ vā na brahma-saṁstha āśrama-dharma- mātra-saṁsthaḥ; brahma-saṁsthasya tv am tatva-śrava āt. ChU 2.23.1. 33 kiṁ caivam evātmānaṁ svaṁ lokam icchantaḥ prārthayantaḥ pravrājinaḥ pravrajana-śīlāḥ pravrajanti prakarṣe a vrajanti sarvā i karmā i saṁnyasyantīty arthaḥ. BĀUBh 4.2.22, IX.647. 34 nanu viruddham idaṁ vacanam—apekṣate ca āśrama-karmā i vidyā, nāpekṣate ceti. neti brūmaḥś utpannā hi vidyā phala-siddhiṁ prati na kiṁcid anyad apekṣate, utpattiṁ prati tu apekṣate. BSBh 3.4.26, III.737. 356 They have been useful for the arising of knowledge, but knowledge once arisen has a result of its own—Śa kara occasionally talks about the arising and the maturation of knowledge35—for which it does not require the help of the rama practices. We still need to figure out just what this “arising of knowledge” is, but let us first introduce the relevant part from BĀU 4.2.23. When the performance of the rama practices has borne fruit and one has become a knower of the Self, one further becomes nta, calm, “withdrawn from the action of the senses”ś d nta, self-controlled, “averse to the mental cravings”, uparata, tranquil, “free from all desires”ś titikṣu, tolerant, “bearing with the dualities”ś and sam hita, collected, “concentrated by disassociation from the spurs of the mind and the senses.”36 This is, in fact, the scriptural origin of the set of personal virtues— ama, dama, uparati, titikṣ and sam dh na—that we know from the BS as the virtues that all aspirants after liberation should cultivate. We have also seen them standing for the yama-niyama complex in the doctrine of prasaṅkhy na-v da, the counterpart of kriy -yoga in Yogic meditation. Now, we need to note two important things about them in the context of the arising of knowledge. First, while the BĀU passage says that one who had acquired these five has already become a knower of the Self—they appear as the result of the vividiṣ —they may also be practiced in lieu of the five rama practices, even when one does not have knowledge of the Self. Śa kara, in other words, treats them as rama practices like the other five, because they are the practices of those who are real knowers of Brahman. They have, thus, a dual nature of virtues that one has acquired, and of practices that one cultivates intentionally. Although they are 35 For instance, in the TUBh 1.11.4. 36 BĀU 4.2.23Ś tasmād evaṁvic chānto dānta uparatas titikṣuḥ samāhito bhūtvātmany evātmānaṁ paśyati. From the commentary: tasmād evaṁ-vic chānto bāhyendriya-vyāpārata upaśāntas tathā dānto 'ntaḥ-kara a-t ṣ āto niv tta uparataḥ sarvaiṣa ā-vinirmuktaḥ saṁnyāsī titikṣur dvandva-sahiṣ uḥ samāhita indriyāntaḥ-kara ācalana-rūpād vyāv tyaikāgrya-rūpe a samāhito bhūtvā. Ibid. 357 pertinent to someone who already knows the Self, manifest after the arising of knowledge—and Śa kara says that they are “directly related to knowledge,” for reasons we will see under the next heading—they can serve as substitutes for the rama practices that are just for the arising of knowledge, which comes to mean that one can practice just them instead of ritual and the like, renounce even before knowledge had arisen, and make knowledge arise through them.37 If they are “related to knowledge,” why couldn’t they give rise to knowledge as well? They are, in fact, better suited to give rise to knowledge, being “unmixed” with violence and the like, as we have seen. Second, like the prasaṅkhy na-v dins, Śa kara takes these five as quite synonymous with the Yogic yama-niyama complex—commonly referring to them either as ama-dam di or yama-niyama, where the respective items evidently correspond and mean the same thing, sense and mind control—and wider in scope than just the five. To put this differently, from some important loci such as TUBh 1.11.4, BĀUBh 4.5.15, and USG 1Ś4-5, it is evident that the virtues/practices that are “related to knowledge” are more than just five—they include non- violence, absence of anger, truthfulness, etc.—but they are essentially reducible to practices of sense and mind control, and to virtues such as humility. Under BĀU 4.5.15, Śa kara says that the rama practices that are directly related to knowledge are, first, yama-pradh n ni, predominantly sense control in nature; second, humility, etc.; and, third, mental practices characterized by meditation, knowledge, and dispassion.38 A similar classification can be made based on USG 1:4-5, where Śa kara says that the teacher should engage the student who 37 tasmāt viraktasya mumukṣoḥ vināpi jñānena brahmacary d eva pravrajed ityādi upapannam. BĀUBh 4.5.15, IX.677. 38 atha evaṁ sati avidvad-viṣayā ām āśrama-karma āṁ balābala-vicāra āyām, ātma-jñānotpādanaṁ prati yama- pradhānānām amānitvādīnāṁ mānasānāṁ ca dhyāna-jñāna-vairāgyādīnāṁ sa nipatyopakāratvamś hiṁsā-rāga- dveṣādi-bāhulyād bahu-kliṣ a-karma-vimiśritā itare – iti. BĀUBh 4.5.15, IX.676. 358 evidently has no grasp of knowledge in yamas such as non-violence and absence of anger; in niyamas that are not opposed to knowledge, which would likely correspond to dhy na, dh raṇa, jñ na, vair gya from BĀUBh 4.5.15 and TUBh 1.11.4ś and in humility and similar virtues. Nevertheless, the BĀU 4.4.23 text is important for theological purposes, because based on this text Śa kara claims that the ama-dam di complex is more proximate to knowledge than the practices of the common ramas. Let us summarize our findings. The path to liberation begins with the performance of the rama practices, along with Upaniṣadic meditation and no expectance of the results. This complex causes purity of the Self through exhausting bad karma, which removes the psychological faults and leads to the arising of knowledge. In theological terms, this is called vividiṣ and is based on BĀU 4.4.22. Alternatively, there are virtues that commonly appear when knowledge had arisen, but which can also be practiced instead of ritual and the like, for the arising of knowledge. They are essentially sense and mind control in nature, and theologically are traced to BĀU 4.4.23. The common rama practices are “comparatively external” because of their relation to the “desire to know,” vividiṣ , whereas the later are “proximate” because of their relation to “knowledge” or vidy itself.39 And, there is the grace of the Lord, which can make knowledge arise as well. Arising of Knowledge There is, unfortunately, no stipulative definition that I could find in Śa kara’s works that would reveal immediately what this arising of knowledge, jñ notpatti, vidyotpatti, or jñ na-pr pti, is. Nevertheless, there are several coordinates that can help us pinpoint it. The key to this is to 39 tasmād yajñādīni śama-damādīni ca yathāśramaṁ sarvā y eva āśrama-karmā i vidyotpattāv apekṣitavyāni. tatrāpi evaṁ-vit iti vidyā-saṁyogāt pratyāsannāni vidyā-sādhanāni śamādīni, vividiṣā-saṁyogāt tu bāhyatarā i yajñādīnīti vivektavyam. BSBh 3.4.27, III.739-40. 359 remember that there is a point at which knowledge is arisen, ritual and the other rama practices are of no use anymore, and knowledge itself brings the path of liberation to fruition. We should note the following important statement again: – But, is it not contradictory to say that knowledge depends on the rama practices and that it does not depend on them at the same time? – No, we say! Knowledge, once arisen, does not depend on anything for fruition. However, it does depend for its arising.40 Further, such fruition of knowledge is preceded by the practice of the triple Vedāntic methodŚ The result of knowledge is preceded by hearing, etc., in line with the statement that one should listen about, ponder over, and reflect on the Self.41 We will see in Chapter Nine that Śa kara presents renunciation of all rama practices as an aṅga, an essential requirement for the practice of the three methods. There are, then, three points of what we may call brahma-vidy , or knowledge of Brahman: a point at which knowledge is arisen, followed by knowledge as practice, and culminating in knowledge as a result. Let us also remember that after one had known the Self, one becomes endowed with the five virtues and seeks the Self, having renounced the rama practices, as per BĀU 4.4.23. Why seek the Self if one knows the Self? Furthermore, Śa kara in his delineation of the key juncture points on the path to liberation placed “attainment of knowledge” or jñ na-pr pti as the second stop, preceded by purity of existence achieved through the performing the rama practices, and followed by their renunciation. Clearly jñ notpatti does not stand for knowing oneself as Brahman, but is intimately related to the status of ritual and rama duties. To begin with, any student of Śa kara worthy of her salt would have by now recognized the set of virtues or practices from the BĀU 4.4.23 that I discussed above, ama, dama, uparati, 40 nanu viruddham idaṁ vacanam—apekṣate ca āśrama-karmā i vidyā, nāpekṣate ceti. neti brūmaḥ. utpannā hi vidyā phala-siddhiṁ prati na kiṁcid anyad apekṣate, utpattiṁ prati tu apekṣate. BSBh 3.4.26, III.737. 41 śrava ādi-pūrvakaṁ hi vidyā-phalam, rotavyo mantavyo nididhy sitavyaḥ [BĀU 2.4.5/4.5.6] ity-ādi-śruty- antarebhyaḥ. TUBh 2.1, VI.59. 360 titikṣ and sam dh naṬ They are the third member of what is sometimes called s dhana- catuṣṭaya or the fourfold means that should prepare one for direct engagement in the study of Brahman. The four appear at the opening of the BSBh as standing for the immediate sequence denoted by the first word atha. Therefore, something must be pointed out in the sequence of which the inquiry into Brahman follows. It is said, it is the discrimination between things eternal and transient; dispassion for the enjoyment of things here and in the hereafter; the acquisition of the means that are calmness, self-control, etc.; and striving after liberation.42 The third means here is sometimes called ṣaṭ-sampatti or the “acquisition of six,” because raddh or faith is commonly added to the list of five.43 This is a post-Śa kara development that is likely the result of merging the Kā va and the Mādhyandina recensions of the BĀU, since the second substitutes raddh vitaḥ for sam hitaḥ.44 Padmapāda in the PP follows the Kā va on which Śa kara’s commentary was based and lists the original five. The change likely appears with the Bh mat of Vācaspati, who quotes the Mādhyandina and lists raddh as the fifth practice. In Ānandagiri, we find the composite list of six, and they are usually given as such in scholarly accounts. As we saw under the previous heading, Śa kara himself generally refers to them as ama-dam di or yama-niyama, and for him the complex admitted other virtues as well. In the prose section of the US, compassion or day is added to the first two, standing likely for the set of personal virtues which he generally instantiates with humility, and rounding up the three kinds of practices and virtues. For convenience, we may from this point on refer to them as ṣaṭ-sampatti, but it should be clear that this is just a metonym. 42 tasmāt kimapi vaktavyam, yad-anantaraṁ brahma-jijñāsopadiśyata iti. ucyate—nityānitya-vastu-vivekaḥ, ihāmutrārtha-bhoga-virāgaḥ, śama-damādi-sādhana-sampat, mumukṣutvaṁ ca. BSBh 1.1.1, I.6. 43 See, for instance, Radhakrishnan 160:155. 44 The Mādhyandina text is atapatha Br hmaṇa 14.7.2.26Ś tasmāc chānto dānta uparatas titikṣuḥ śraddhāvitto bhūtvātmany evātmānaṁ paśyet sarvamātmani paśyati. 361  But we are getting sidetracked. We should remember that these were described as appearing once one has known the Self, after knowledge had arisen, yet could be cultivated for such arising in place of the rama practices and meditation. We also saw that there was a point at which knowledge was considered arisen and independently working towards fruition, at which point one would renounce the rama practices and meditation. We are, therefore, justified in understanding this arising of knowledge as an event that is not instantaneous, but takes some time. The midway placement of the ṣaṭ-sampatti, which were both consequent on and for the sake of knowledge, indicates that the first point of the arising of knowledge is the first in Śa kara’s list of four meansŚ discrimination between things that are eternal and transient, nity nitya-vastu-viveka. This is the big breakthrough that the disinterested performance of ritual and the other rama practices should bring about. The content of this discriminative knowledge is obviously not some full understanding of the Self on the one hand and matter on the other, because that would be tantamount or close to liberation. Rather, both Sureśvara and Padmapāda read this as understanding the true nature of transmigration, the fact that everything in the world, objectively and as the value of one’s religious aspirations culminating in the attainment of brahma-loka, is transient, perishable, subject to decay.45 Indeed, there is a motif recurring throughout Śa kara’s works that points to an event that must happen before one can engage in the properly Vedāntic or brahma-vidy practices, and that is understanding the world of dualities in general and of Vedic ritual in particular, and becoming disillusioned with them. Consider, for instance, the following statement that marks the beginning of inquiry into Brahman: 45 Sureśvara’s formulation in NaiS 1.51 is saṁsāra-yāthātmyāvabodha. See also PP p.62-3. 362  This chapter [of the Upaniṣad] is begun for understanding the truth of the nature of the Self for the one who has purified himself by observing a combination of meditation and ritual without expecting the results, who has removed the obstacles for the [appearance] of the knowledge of the Self, who sees the faults concerning duality because he has become cognitively averse to the sphere of the external, who is trying to uproot the seed of transmigration which is ignorance, and who is inquisitive about the inner Self.46 The passage presents a nice transition from the purification of existence accomplished by ritual and meditation to understanding saṁs ra and becoming averse to it. Similar statements are common throughout Śa kara’s works, and the same event of insight into the nature of saṁs ra followed by disillusion with the world must happen even in brahma-loka, in the case of those who continue their progress towards liberation on the krama-mukti track: Some great sages are attached to other meditations that bring about opulence, but later, through seeing how this opulence decays, they get disillusioned, become steadfast in knowledge of the Supreme Self, and reach liberation.47 The passage is about sages who have reached brahma-loka and are liberated with Hira yagarbha at the end of the universal cycle. Even in brahma-loka, then, there must happen an insight into the decay that is characteristic of saṁs ra, followed by becoming averse to it. That is the first of the four “means” required prior to the inquiry into Brahman, followed by the second: nity nitya- vastu-viveka and vair gya. A common characterization of the second is that one has become nirviṇṇa, disillusioned or disgusted with saṁs ra.48 There is a textual reason behind this use, as we shall see shortly. 46 ata ūrdhvaṁ phala-nirapekṣa-jñāna-karma-samuccayānuṣ hānāt k tātma-saṁskārasya uccinnātma-jñāna- pratibandhakasya dvaita-viṣaya-doṣa-darśinaḥ nirjñātāśeṣa-bāhya-viṣayatvāt saṁsāra-bījam ajñānam uccicchitsataḥ pratyag-ātma-viṣaya-jijñāsoḥ keneṣitam ity-ātma-svarūpa-tattva-vijñanāya ayam adhyāya ārabhyate. KUVBh Introduction, IV.83-4. 47 jñānāntareṣu ca aiśvaryādi-phaleṣv āsaktāḥ syur maharṣayaḥ. te paścād aiśvarya-kṣaya-darśanena nirvi āḥ paramātma-jñāne pariniṣ hāya kaivalyaṁ prāpur ity upapadyate. BSBh 3.3.32, III.671. 48 The following statement is, again, typical: “Brāhma as, seekers of Brahman, on getting a teacher who is like a boat on the boundless ocean which has for its water the suffering due to roaming in the cycle of birth, decay and death, desire to cross that ocean, and being disilusioned with the world of means and attainments consisting in good and bad karma and their respective means and results, long to attain the eternal, highest good which is entirely different from the above.” brāhma ā brahma vividiṣavaḥ janma-jarā-mara a-prabandha-cakra-bhrama a-k tāyāsa- duḥkhodakāpāra-mahodadhi-plava-bhūtaṁ gurum āsādya tat-tīram uttitīrṣavaḥ dharmādharma-sādhana-tat-phala- 363  These two—the understanding that everything that is won through ritual undergoes decay and the subsequent disgust with such ritual attainments—constitute the arising of knowledge or jñ notpattiṭvidyotpatti brought about by the disinterested performance of the rama practices. The arising is, plain and simple, an insight into the nature of saṁs ra, and dispassion. Once dispassion has been won, it leads to the appearance of the ṣaṭ-sampatti, which are now not only practices, as they were in relation to the desire to know the Self and practiced instead of ritual and meditation, but acquired virtues. We can now appreciate why Śa kara says that they are “related to knowledge.” Knowledge is dispassion, and for the one who is dispassionate it is natural to control the mind and senses and to develop the personal virtues. Paul Deussen in his classic “The System of the Vedānta” says the first of the four qualifications for the study of Brahman stands for “a general metaphysical disposition in virtue of which one has a consciousness of an unchanging being, in contrast with the changeableness of all worldly things and relations.”49 He says that the eternal substance here is Brahman, but this stage stands for some initial insight into its nature and not a full understanding. This is quite misleading, as we have seen above. The discrimination is an insight into the nature of transmigration as it pertains to the world of Vedic ritual, or to action more generally, including the mental action that is meditation. It stands for understanding that things obtained by ritual perish, an understanding that engenders dispassion towards the three lokas—this world, the world of the ancestors and of the gods—and sets one towards the Self. That one had become an tma-vit means that one had become disillusioned with what the Self is not.50 lakṣa āt sādhya-sādhana-rūpān nirvi āḥ tad-vilakṣa a-nitya-niratiśaya-śreyaḥ pratipitsavaḥ, BĀUBh 1.4.9, VIII.119. 49 Deussen 1912:80. 50 Scholars, of course, continue giving vague characterizations that cannot be supported with references. See, for instanceŚ “The first of these [nity nitya-vastu-viveka] involves primarily insight into the underlying reality that encompasses all change.” (Dubois 2013Ś10) But precision is, perhaps, too much to ask from a book that spells “Upāniṣad” throughout its 400 plus pages, including in its title. 364  The cultivation of ṣaṭ-sampatti should culminate in one’s becoming an aspirant after liberation. This is the high point of the arising of knowledge, the consummation of the “desire to know” and the transition to the desire for the Self, from vividiṣ to tma-k ma or mokṣa-k ma, from the discovery of the Self to “becoming” the Self. At this threshold, one must renounce ritual and the rama practices, since knowledge has arisen and is on its way to fruition. This was what those whose knowledge had arisen do, as we shall remember from the BĀU 4.4.22-3: those who have discovered the Self go forth as renunciants, striving after that Self as their world alone. Specifically, they renounce the desire for sons, wealth, and worlds, in effect their entitlement to ritual, the necessary means of ritual performance, and the attainment. The ṣaṭ-sampatti, although part of the wider set of rama practices, are not renounced, because they are just what knowledge is. They are not renounced even when one had understood Brahman, “become” the Self, for a reason which we will consider in Chapter Nine. Let us also note here that the event of arising of knowledge was doubtlessly patterned on one of the key textual determinants of Śa kara’s soteriology, a verse from the Muṇḍaka Upaniṣad that played the role of adhik ra- ruti in Advaita Vedānta, parallel to the adhik ra- vidhi of the Mīmāṁsakas that we introduced in the Second Chapter, and laying down the entitlement to the inquiry into Brahman and the results that it brings: Having examined the worlds piled up by ritual, a Brāhma a should become disillusioned, thinking “What is not made cannot be won by what is made.” For knowing that [Brahman,] he should, sacrificial fuel in hand, approach a teacher who is learned in the Veda and steadfast in Brahman.51 51 parīkṣya lokān karma-citān brāhma o nirvedam āyān nāsty ak taḥ k tena | tad-vijñānārthaṁ sa gurum evābhigacchet samit-pā iḥ śrotriyaṁ brahma-niṣ ham. MU 1.2.12. 365 The verse presents the sequence of the two in Śa kara’s “four means” from the BSBh, understanding followed by dispassion, where the word for dispassion is nirveda that is etymologically related to nirviṇṇa. In the next verse, we learn that the Brahmin who approaches a teacher in the described manner is pra nta-citta, of pure mind—the ultimate attainment of the rama practices—and am nvita, endowed with calmness. The last evidently corresponds to the first of the ṣat-sampatti, ama. This confirms our finding of what it is that Śa kara had in mind under nity nitya-vastu-viveka, the point at which knowledge appears. The transient refers to the worlds won by ritual, and it is this that must dawn on the agent of ritual and meditation.52 The Model of Causality Before moving to vidy and its functioning, let us complete the story of the soteriological role of ritual, meditation and the rama practices. Śa kara said, as we saw, that they were “comparatively external” to knowledge. On a couple of occasions, he called ritual that consists of the daily Agnihotra r d-upak raka, a term with which we got acquainted in the Second Chapter. To refresh our memory, r d-upak rakas were full-fledged rituals for which no separate results were directly stated, so by the application of the Mīmāṁsā principles of interpretation they were considered auxiliaries to other major rituals that were described in their 52 I do not wish to open here the theme of physical renunciation in Advaita Vedānta, on which many have written quite competently. It is transparent to me that Śa kara thought that such physical renunciation had to happen before one could properly engage in brahma-vidy . Relatively recently, however, Roger Marcaurelle claimed that what Śa kara had in mind under renouncing was primarily giving up the sense of agency, and that physical renunciation was not necessary in all cases, as a consequence of which full enlightenment, to Śa kara’s mind, was available to everyone, not only Brahmins. While engaging fully with Marcaurelle’s thesis would take us far afield, I wish to point out that he, like Malkowsky, misunderstands what jñ notpatti represents in Śa kara’s system. See, for instanceŚ “This rediscovery of the mind’s real nature is called the “emergence of knowledge” (jñ notpatti). As a spontaneous and immediate result of this emergence comes the destruction of ignorance (avidy nivṛtti) of the real nature of one’s Self. And with annihilation of ignorance also immediately ensues the eradication of its effect, that is, the erroneous superimposition on the Self of limiting adjuncts such as doership and experiencing.” (Marcaurelle 2000:25-6) As I have shown, jñ notpatti is not the final understanding of one’s being Brahman, but a relatively early threshold on the path to liberation. When that much is understood, engaging with Marcaurelle’s thesis becomes superfluous. 366 textual vicinity. “That which on its own does not have a result but is in the proximity of something that does is considered its auxiliary.”53 In practice, they were smaller independent rituals performed before and after each of the two main rituals in a Darśa-pūr amāsa. We saw that Mīmāṁsakas thought these r d-upak rakas produced some intermediate apūrva, whose causal contribution was absorbed in the apūrva of the respective principal fortnightly ritual that eventually mixed in the final apūrva. In terms of the attainment of heaven, the r d-upak rakas were indirect means, because they served directly their principal rituals. However, they got their name of r d-upak raka not through their indirect relation to heaven the result, but through their direct relation to the principal ritual they served, and in pair with the other type of non-principal actions, the sannip tyopak rakas, which I called “aggregated helpers” because their causal contribution was absorbed in a ritual auxiliary, namely the offertory. These were the four actions of production, acquisition, transformation, and refinement. They were all performed over a substance that was offered in a ritual—making the rice paddy, getting milk from the cow, melting solid butter, sprinkling the paddy—and their causal contribution to the principal action of offering was mediate, absorbed in the offertory. In Mīmāṁsā technical jargon, the r d-upak rakas were direct helpers to the principal ritual, because there was no intermediary between the two, whereas the sannipatyopakarakas were indirect helpers because their immediate relation was to the material used for offering. However, the sannip tyopak rakas expressed a closer relation to their superordinate material—there cannot be threshed rice without threshing—and one that is directly expressed in an injunction: vr hin avahanti. The auxiliary full rituals, on the other hand, were not really required for the principal rite in a Darśa-pūr amāsa, except for the fact that they are described in 53 phalavat sannidhāv aphalaṁ tad-a gaṁ bhavati. MSŚBh 4.4.19, IV.1277. 367 proximity and not related to a result. Their relation to the principal rites were not introduced into being by a statement to that effect either, but through scriptural postulation, arth patti: if they are not related to the principal rites, they would be useless, and this compromises the validity of the Veda and is unwanted. Let us pay attention to these complex hierarchical relations: the auxiliary rituals are directly helping the principal ritual, but the relation is distant, accidental; the sannip tyopak rakas are helping the principal ritual indirectly, through their superordinate material, but their relation to their material is close, essential; both are indirect in relation to heaven, since their causal contribution terminates in the primary ritual either directly or indirectly, and only the primary ritual is directly related to heaven. Now, Śa kara thought that ritual could be considered such a direct helper to knowledge because, under the circumstances we have discussed—accompanied by meditation and not related to its common result in the form of heaven—it would become fruitful by giving rise to knowledge through personal purification. In this scenario, knowledge was the principal means that eventually brought about liberation: – But liberation cannot be produced. How can you say that it is a product of ritual? – There is no such fault, because ritual is a direct helper. Ritual, bringing about knowledge, is figuratively called an indirect cause of liberation.54 I should like to note here that r d-upak raka is commonly but mistakenly translated as “indirect helper” in scholarly literature. Both Thibaut and Gambhirananda mistranslate r d- upak rakatv t karmaṇaḥ in the above sentence as “Works, we reply, may subserve final release mediately,” and “work helps from distance (i.e., indirectly) in producing the result,” because the 54 nanu anārabhyo mokṣaḥ, katham asya karma-kāryatvam ucyate? naiṣa doṣaḥ, ārād-upakārakatvāt karma aḥ. jñānasyaiva hi prāpakaṁ sat karma pra āḍyā mokṣa-kāra am ity upacaryate. BSBh 4.1.16, III.792. 368 adverb r d means both “directly” and “indirectly” and they wrongly relate r d-upak raka to liberation rather than knowledge, without appreciating the theological context of the argument.55 We can now conclusively answer the question that we posed at the beginning of this part of the dissertationŚ how did Śa kara precisely look at the possibility of combining ritual with knowledge? One available relationship was that between a principal and an auxiliary, pradh na and aṅgaṭ eṣa/guṇa, if the second was an essential element for the first insofar as without it, the whole complex would be impossible. Such was the case, for instance, with the relation of offertories to the action of offering: there cannot be an action of sacrifice without a sacrificial animal or an appropriate substitute. The sannip tyopak rakas expressed a similar close relationship to their superordinate material. Another available relationship, quite different, was that between two principal factors, in the manner of the two fortnightly rituals in a Darśa- pūra amāsa. These two scenarios exhausted the scope of what Vedāntins understood under the samuccaya relation. In either case, both elements were absolutely required for the success of the undertaking, and simultaneously so, either temporally or through the final combination of the respective apūrva in the transtemporal Self. This Śa kara rejected in no ambiguous terms, and for several related reasons. The arising of knowledge just meant understanding that the results of ritual were transient; that one cannot win immortality by wealth, the necessary means of ritual; that the unmade cannot be won by the made. To continue performing ritual in such circumstances would be kind of schizophrenic, affirming what one is trying to negate. Besides, liberation was a visible result, and only the four sannip tyopak rakas produced visible results. None of them could so much as touch the eternal Self. And so on. We should note very well that Śa kara under this model of samuccaya rejects 55 Equally wrong, it seems to me, are Alston and Balasubramanian in their respective translations of the Naiṣkarmya-Siddhi; understanding the Vedānta use of r d-upak raka requires appreciating its Mīmāṁsā context. 369 repeatedly the same analogy that we started with: ritual spiced with meditation is like sweet yoghurt or charmed poison. This analogy was, in fact, Bhart prapañca’s, who thought that it was the knowledge of the eternal non-dual Self, knowledge qua knowledge, that transubstantiates ritual from bondage to liberation.56 For Śa kara, however, samuccaya was the way in which meditation combines with ritual, not knowledge. It was meditation that transforms ritual, not knowledge of Brahman. Another possibility was to treat ritual as an r d-upak raka, a direct but distant helper to knowledge, direct insofar as its contribution is not absorbed by some other causal factor first, but distant insofar as it is not required after it had given rise to knowledge. Furthermore, because there was no natural relation between ritual and knowledge, but one established through a postulated originative injunction as we saw in BSBh 3.4.27, there was no reason that knowledge, dispassion, could not arise through those rama practices that were knowledge in nature—mind and sense control—or through the grace of God, etc. Although the relation between ritual and knowledge was direct, it was not essential. Knowledge had such an essential relationship with the opposite of ritual, that is, with renunciation of all varṇ rama duties except those that were renunciation in kind. Both ritual and renunciation of ritual had direct relationship to knowledge, but under different causal models. The first was an r d-upak raka, while Śa kara called the second, the ṣaṭ-sampatti complex, sannip tyopak rakas, essentially related to knowledge. This was so because without mind and sense control and a healthy dose of humility, knowledge— dispassion—was impossible.57 And, the first was related to liberation, the result of knowledge, mediately, through giving rise to knowledge which independently produces its own result, while 56 BĀUBh 4.4.22ś TUBh 1.11.4. 57 atha evaṁ sati avidvad-viṣayā ām āśrama-karma āṁ balābala-vicāra āyām, ātma-jñānotpādanaṁ prati yama- pradhānānām amānitvādīnāṁ mānasānāṁ ca dhyāna-jñāna-vairāgyādīnāṁ sannipatyopakāratvam. BĀUBh 4.5.15, IX.676. 370 the second, competent to give rise to knowledge, was necessary after such rise had taken place, and until the full understanding of unity (and even after, as we shall see in Chapter Nine). Although Śa kara does not say as much, he clearly saw the relation of ritual with knowledge through the p ramparya model of ritual causality that we considered in the Second Chapter. This was, to remember, the model of termination of the direct causal contribution of any ritual element in its immediate effect, but reaching the final result through a chain of successive intermediate results. It was, for instance, the model through which the threshing of rice was absorbed in the rice as its effect, but nevertheless reached heaven through the rice, the rice paddy, and the offering. I suggested in the Second Chapter that it was Kumārila who developed the model on a full scale. In any given ritual, both models of causality were involved, of course, and so were they in the pursuit of liberation. Knowledge had renunciation of all varṇ rama practices as its integral part, aṅga, and it could not be practiced without it. Ritual, meditation and the other varṇ rama practices, on the other hand, culminated in their contribution to the arising of knowledge, specifically in dispassion, and reached liberation mediately. On this lower lever, nevertheless, ritual and “knowledge,” that is, saguṇa meditation based the Upaniṣadic vidy s, were related in a samuccaya manner, had no problem combining, but had to be performed without desire for their common results. Although Śa kara referred elsewhere to the idea of p ramparya explicitly, the term which he used for ritual contributing indirectly to liberation was the adverbial praṇ ḍy , which we may read as its synonym.58 Sureśvara, on the other hand, as we saw in the introduction, laid out the full journey to liberation through a succession of stages in which the performance of ritual as the starting point and an r d-upak raka mediately culminates in liberation through a 58 BSBh 4.1.16. 371 p ramparya chain.59 From this point on, Advaitins begin talking about knowledge and ritual as means of liberation under the mode of s kṣ t-p rampary bhy m, directly and mediately, respectively.60 It is worthwhile pointing to Śureśvara’s inimitable wit in the way he described the transition from the renunciation of ritual and all varṇ rama practices to the sole engagement in knowledge. Renunciation was like the sampatti or samprad na ritual that we talked about in the previous chapter, the means of winning this world, when the dying father transferred his own ritual self to his son, to continue sacrificing vicariously through him.61 We can almost visualize ritual on its deathbed, whispering to knowledgeŚ “I have done all that I could. Now you carry on.” Of crucial importance was that this p ramparya chain of soteriological causality extended through lifetimes, and it was perfectly possible for one to have achieved the requisite mental purity by doing ritual and observing the rama practices in a previous life, in which case one was supposed to renounce and engage in brahma-vidy immediately after student life. The litmus test was dispassion. Once dispassion was achieved, ritual had nothing more to contribute.62 Padmapāda, in fact, claimed that this was why inquiry into Brahman did not have to be preceded by the inquiry into dharma. One may have done that in a previous life and gotten all that one could get.63 59 NaiS 1.45-52. 60 See, for instance, the commentaries on BSBh 4.1.16, particularly ĀnandagiriŚ “Treating the sūtra as being about nirguṇa-vidy , knowledge and action are a means of liberation directly and mediately, like the plow and eating are the means of livingś” nirgu a-vidyā-viṣayatvaṁ sūtrasyopetya lā gala-bhojanayor jīvana-hetutā-vad dhī-karma oḥ sākṣāt-pāramparyābhyāṁ mokṣa-hetutvam. 61 NaiS 1.49. 62 “Action characterized by Agnihotra and by celibacy and the like, performed in a previous life, facilitates the arising of knowledge such that some are evidently dispassionate from very birth.” TUBh 1.11.4. 63 na ca naiyogike phale kāla-niyamo ‘sti. tena pūrva-janmānuṣ hita-karma-saṁsk to dharma-jijñāsāṁ tad- anuṣ hānaṁ cāpratipadyamāna eva brahma-jijñāsāyāṁ pravartata iti na niyamena tad-apeṣko ‘tha-śabdo yujyate. “There is no rule when actions will bear fruit. One may have been purified by performed actions in a past life that bear fruit now; if so, inquiry into dharma is unnecessary, and so it cannot be the consequence that 'atha' would refer to.” PP, p.61. 372  We should also note that, as John Taber had recognized, the inquiry into Brahman to Śa kara’s mind could not be fruitful without the four prerequisites being satisfied, that is, without knowledge having arisen.64 Śa kara, in fact, said as much at the beginning of the prose part of the US. If the teacher recognizes signs that the student does not “grasp knowledge,” he should send him to remedial lessons on sense and mind control and humility. 65 Padmapāda was even more explicit about it: If somehow—by the will of providence or by curiosity or the desire for much learning— the inquiry is undertaken [without the four prerequisites satisfied in succession], one will not be able to understand without doubts that Brahman is the Self, because without obtaining the means enumerated, his mind, not tuned inward, will be engrossed only in the external.66 We will end this chapter by noting that the status of direct helpers given to ritual meant that the knowledge passages of the Upaniṣads and the meditation and ritual texts could form a unity of independent texts: they did not require one another syntactically, but there was a way to combine them through the vividiṣ , the desire to know the Self, in a unity of purpose. Kumārila called this v kyaikav kyat , unity of independent texts, distinct from the more common syntactic sentential unity or padaikav kyat .67 And, the result of such intertextual unity was that the integral Veda, without the explicitly k mya portions that could not be repurposed, was for liberation, as I hinted at the end of the last chapter. This is the big takeaway from this chapter for our ultimate, mah -v kya, purpose. 64 Taber 1983. 65 śiṣyasya jñānāgraha aṁ ca li gair buddhvā tad-agraha a-hetūn adharma-laukika-pramāda-nityānitya-vastu- viveka-viṣayāsaṁjāta-d ḍha-pūrva-śrutatva-loka-cintāvekṣa a-jātyādy-abhimānādīṁs tat-pratipakṣaiḥ śruti-sm ti- vihitair apanayed akroddhādibhir ahiṁsādibhiś ca yamair jñānāviruddhaiś ca niyamaiḥ. amānitvādi-gu aṁ ca jñānopāyaṁ samyag grāhayet. USP 1Ś4-5. 66 kathaṁcid vā daiva-vaśāt kutūhalād vā bahu-śrutatva-buddhyā vā prav tto ‘pi na nirvicikitsaṁ brahmātmatvenāvagantuṁ yathokta-sādhana-sampatti-virahāt anantar-mukha-cetā bahir evābhiniviśamānaḥ. PP, p.63. 67 See Kunjunni Raja 1977 161-2. 373  The whole of the ritual portion of the Veda, with the sole exception of the parts dealing with optional rituals (k mya), is employed through absorption in this knowledge of Brahman.68 As for the ritual portion of the Veda being auxiliary to the meaning of the knowledge portion, that is a case of unity of independent passages occasioned by a text that establishes a principal-auxiliary relation.69 68 BĀUBh 4.4.22. 69 jñāna-kā ḍārtha-śeṣatvaṁ karma-ka ḍasya yat punaḥ | viniyojaka-hetv etat tayor vākyaikavākyataḥ. SV 278. The viniyojaka text that establishes the principal-subordinate relation of independent items is the vividiṣ text, BĀU 4.4.22. 374 CHAPTER EIGHT: YOU ARE THAT, ALL RIGHT, WE JUST NEED TO FIGURE OUT WHAT: VED NTA-V KYA AND THE IDENTITY STATEMENTS As for those who ignore the surest warrant “You are Being” and seek to know by other means, well, they may as well taste through their eyes.1 The Upani ads and Para- and Apara-vidy As is commonly known, Śa kara understood the Upaniṣads as propounding the doctrine of the unity of the Self and Brahman, and as having liberation as their sole goal. It is quite common for Śa kara to say such things, particularly in the so-called Sambandha-Bh ṣyas, or introductions to his Upaniṣad commentaries, and at important juncture points. For instance: All the Upaniṣads, as well as the Bhagavad-G t and Mokṣa-Dharma, are exhausted just in determining the true nature of the Self.2 The intended meaning in this whole Upaniṣad is the knowledge that Brahman is without interior or exterior and homogenous like a lump of salt. This is ascertained based on the ending of the two portions [chapters 1 through 4], “this is the teaching,” “there is so much to immortality.” Likewise, in the Upaniṣads of all Vedic branches, knowledge of the unity of Brahman is the conclusive meaning.3 He did not shy away from being dramatic about it either: Those who, on the other hand, imagine a Self which is different from Brahman and reduce the science of bondage and liberation to arthav da, they dare to trace the footsteps of the birds or to pull the sky with their fist and cast it away as if it were skin. We are not able to do that. We understand that the conclusive meaning of all Upaniṣadic texts is that we are Brahman, which is always homogenous, non-dual, changeless, unborn, without old age and death, immortal, fearless, the Self.4 1 addhātamam anād tya pramā aṁ sad asīti ye | bubhutsante 'nyataḥ kuryus te 'kṣ āpi rasa-vedanam. NaiS 3.117. 2 sarvāsām upaniṣadām ātma-yathātmya-nirūpa aivopakṣayāt, gītānāṁ mokṣa-dharmā āṁ caivaṁ-paratvāt. ĪUBh Introduction, IV.5-6. 3 saindhava-ghana-vad anantaram abāhyam ekarasaṁ brahma—iti vijñānaṁ sarvasyām upaniṣadi pratipipādayiṣitārthaḥ—kā ḍa-dvaye 'py ante 'vadhāra ād—avagamyate—ity anu sanam, et vad are khalv amṛtatvam iti. tathā sarva-śākhopaniṣatsu ca brahmaikatva-vijñānaṁ niścito ‘rthaḥ. BĀUBh 1.4.10, VIII.127. 4 ye tu ato 'nyathā ātma-vastu parikalpya bandha-mokṣādi-śāstraṁ ca arthavādam āpādayanti, te utsahante—khe 'pi śākunaṁ padaṁ draṣ uṁ, khaṁ vā muṣ inā ākraṣ um, carmavad veṣ itum; vayaṁ tu tat kartum aśaktāḥś sarvadā samaikarasam advaitam avikriyam ajam ajaram amaram am tam abhayam ātma-tattvaṁ brahmaiva smaḥ—ity eṣa sarva-vedānta-niścito ‘rtha ity evaṁ pratipadyāmahe. BĀUBh 4.4.6, IX.623. 375 And, while we saw that upaniṣad was obviously related in etymology to up sana, the central idea of the Brahma-Sūtra, and meant meditation on some correlation, as Śa kara himself occasionally acknowledged in texts that were evidently not about liberation,5 he pursued several alternative etymologies, prominently placed in the introductions to his most important works, that were meant to bring home the same idea: upaniṣad is just about liberation: This very knowledge of Brahman is denoted by the word upaniṣad, because of completely exhausting (avas dan t) saṁs ra along with its cause for those who cultivate it; for, such is the meaning of the root sad when preceded by the preverbs upa and ni [i.e., upa+ni = ava].6 Upaniṣad is knowledge (vidy ). Because it strikes down or exhausts the states of being born in a womb and of old age for those who are devoted to it, or because it brings one near Brahman, or because the highest good is found, sat down (upaniṣaṇṇam), in it.7 He even called his own Upade a-S hasr an Upaniṣad: Therefore, this Upaniṣad is begun in order to destroy ignorance, put an end to transmigration and put forth knowledge of Brahman. The word upaniṣad is formed from the root sad preceded by upa and ni and followed by the zero suffix, because it weakens and destroys birth and the like.8 It was, after all, the knowledge that was directly signified by the word, and the composition was an Upaniṣad just because it contained such knowledge.9 We saw, nevertheless, in Chapter Six that Śa kara divided dharma into one characterized by action and another one by abstaining from action. The first was concerned with winning the 5 For instance, in TUBh 1.3, VI.14: saṁhitāyāḥ upaniṣadaṁ saṁhita-viṣayaṁ darśanam ity etat. 6 seyaṁ brahma-vidyā upaniṣac-chabdha-vācyā, tat-parā āṁ sahetoḥ saṁsārasyātyantāvasādanātś upa-ni-pūrvasya sades tad-arthatvāt. BĀUBh Introduction, VIII.3. 7 upaniṣad iti vidyocyate, tat-sevināṁ garbha-janma-jarādi-niśātanāt, tad-avasādanād vāś brahma a upanigamayit tvātś upaniṣa aṁ vā asyāṁ paraṁ śreya iti. TUBh Introduction, VI.10-11. This suggests that he is alternatively taking the root to be ṣad, 6th-class root that has the sense of motion and exhaustion/destruction, or that he thinks that √sad, preceded by upa and ni, has that sense. Cf. ṣadḷ viśara a-gati-avasādaneṣu (Katre 1989:1193). His repeated use of ni tana suggests that he also has the 6th-class √ ad in mind as well, although this could be only through similarity of sound. Cf. śadḷ śātane (Katre ibid.) 8 tasmād ajñāna-hānāya saṁsāra-viniv ttaye | brahma-vidyā-vidhānāya prārabdhopaniṣat tv iyam || sader upani-pūrvasya kvipi copaniṣad bhavet | mandī-kara a-bhāvāc ca garbhādeḥ śātanāt tathā. USP 1.25-6. 9 See the introductions to BĀU and K U. 376 three worlds as a form of promotion or prosperity, abhyudaya, while the second was concerned with liberation, niḥ reyasa. We also identified brahma-loka as the gray area where the two goals intersect. There was evidently a broader sense to what Śa kara called ved nta-v ky ni, Upaniṣadic passages, and while the general point of all the Upaniṣads may have been liberation, evidently not all Upaniṣadic texts were about liberation directly. Śa kara did believe that most of them could eventually be used in the service of liberation, but as units they had their individual goals. Some were directly about liberation, while some could be repurposed to serve the needs of liberation indirectly. Now, we also saw that all reliable warrants that are the means of true cognitions presupposed the operation of ignorance on what we may call an intersubjective level, before personal factors that are resultants of the k ma-v san -karma complex could interfere. This included the Veda as the reliable warrant on all supersensible matters, and with that the Upaniṣads as its subunit. Let us note this well: scripture as pram ṇa requires for its functioning such categories as an agent, an object, an instrument, and a result of knowing, and these are all staple items of ignorance. By being a reliable warrant, the Upaniṣads must indulge in ignorance, they cannot avoid it. What they can do, however, is intend to affirm this ignorance—or not. What this means is that although all Upaniṣadic passages must function in ignorance, some of them may not intend ignorance: by indulging in ignorance, they may intend to communicate such a state of affairs in which the distinctions of knowing ultimately do not obtain. In such a case, for Śa kara they were about knowledge, vidy , and serving the goals of knowledge, although they had to take their bearings in ignorance, avidy . If they intend to affirm distinctions, ignorance, then they were serving goals of ignorance, ergo, goals of desire. Such passages were not about 377 knowledge, but about meditation. In terms of scripture, this is where the wedge with which I ended Chapter Six is located. This distinction was also reflected in Brahman as the specific domain of the Upaniṣads. In the Upaniṣads, there are passages that deny multiplicity in the world, and these are commonly tied to creation accounts, or otherwise to Brahman’s relation to the world. Although these passages may present what Śa kara calls k rya-brahman and apara-brahman, which is ultimately equivalent to the totality of creation, they do not intend to affirm it, but use it as a way of bringing home the point that there is such a thing as Brahman without any distinctions whatsoever, nirguṇa or nirvi eṣa-brahman.10 A good instance of such an Upaniṣadic passage is Yājñavalkya’s dialogue with Gārgī during the Brahmin Super Cup (brahmodya) on the occasion of King Janaka’s sacrifice, the teaching about the imperishable Brahman in the eighth section of the third chapter of the Bṛhad- raṇyaka. In that section, Yājñavalkya presents Brahman in relation to the world—Brahman is that imperishable thing on which all created things “above the sky, below the earth, in between the two, past, present, and future,” are woven warp and woof— but the essential characteristics of this imperishable Brahman are all negative: It is the imperishable, and Brahmins refer to it like this—it is neither coarse nor fine; it is neither short nor long; it has neither blood nor fat; it is without shadow or darkness; it is without air or space; it is without contact; it has no taste or smell; it is without sight or hearing; it is without speech or mind; it is without energy, breath, or mouth; it is beyond measure; it has nothing within it or outside of it; it does not eat anything; and no one eats it.11 The purpose of such passages is not to present Brahman as a creator or otherwise related to creation. Rather, they are “texts whose point is to convey that Brahman, who is the Self, is beyond the phenomenal world.”12 Such passages do associate Brahman with certain adjuncts, but 10 See, for instance, BSBh 1.4.14, 2.1.14. 11 BĀU 3.8.8, translation Olivelle 1998Ś91. 12 vākyāni niṣprapañca-brahmātma-tattva-pradhānāni. BSBh 3.2.14, III.585. 378 their purpose, as I said above, is not to affirm such adjuncts as Brahman’s real characteristics, but to harness them so as to teach that Brahman is without them. Other characteristics that are not tied to creation accounts may also be affirmed of Brahman, for instance the identification of various words for bliss with parts of Brahman’s body in the Taittir ya, but these are commonly not intended as qualities at all, and serve the purpose of facilitating concentration or comprehension.13 As textual units, such passages are about what Śa kara calls para-vidy or nirguṇa-vidy , or passages where multiplicity and Brahman’s distinguishing characteristics are not intended, and which present “a uniform Brahman of which all multiplicity is denied.”14 Generally, thus, texts in which Brahman is presented in negative terms are texts of knowledge that deal with the real Brahman. We will focus on these texts under the next heading. On the other hand, there are Upaniṣadic texts that present Brahman solely with positive characteristics. Śa kara’s favorite instance of such texts was the ṇḍilya-vidy , the fourteenth part of the third chapter of the Ch ndogya that opens with the statement sarvaṁ khalv idaṁ brahma, “This whole world is Brahman,” of mah -v kya fame.15 The passage affirms a slew of positive characteristics of Brahman: Brahman is made of the mind and has the vital functions as its body, it is brilliant in form, contains all actions and desires, smells and tastes, etc. These characteristics are affirmed as intended of Brahman, for the sole reason that they are not denied. For Śa kara, this Brahman or Īśvara that is qualified by such characteristics is the Brahman of meditation. He does not really have these characteristics, but should be seen as if he did.16 Just in virtue of this, the passage could not be about knowledge: a falsehood is intended, not as an error 13 See BSBh 3.3.12. 14 pratyastamitāśeṣa-prapañcam ekākāraṁ brahma. BSBh 3.2.21, III.589. 15 See, for instance, Murty 1959:75. The passage is ChU 3.14. 16 yathokta-gu a-lakṣa a īśvaro dhyeyo na tu tad-gu a-viśiṣ a eva. ChUBh 3.14.4, VI.177. 379 but as a deliberate vision, which is the characteristic feature of meditation. The passage is, thus, about meditation, and it does not “consecrate one in sovereignty,” liberation, unlike the 6th and 7th chapter of the same Upaniṣad.17 Not in a million years could sarvaṁ khalv idaṁ brahma be a mah -v kya in Śa kara’s reckoningŚ it was in the wrong kind of text. Such texts are, really, useless for the purpose of knowing Brahman, because Brahman that is “all” is really Brahman with all desires, actions, tastes, smell, etc. Although the statement about Brahman being all is true in itself, its explication is not, and at the least, it must be first removed from its context.18 Such Upaniṣadic passages, then, which describe Brahman in superlative positive terms are passages of meditation in which the central statement is an injunction.19 Scriptural statements about Brahman are of two kinds: one present Brahman with distinguishing features, as in “It contains all actions, all desires, all smells, all tastes,” and the other present Brahman as devoid of distinguishing features, as in “Not gross nor fine, not short nor long.”20 These two kinds of statements are quite directly different. The first contain true, negative descriptions of Brahman; they also involve some duality insofar as Brahman is depicted as the creator or the support of the world, which however are not intended to be affirmed and are for the purpose of denying substance to multiplicity. These are non-duality statements and statements of knowledge. The second kind of statements contain positive descriptions of Brahman, such that he does not really have, which are, nevertheless, intended to be affirmed. 17 ata eva ṣaṣ ha-saptamayor iva tat tvam asi, tmaivedaṁ sarvam iti neha svārājye 'bhiṣiñcati. Ibid, p.178. 18 And, in fact, it was. Śa kara says that the sixth prap ṭhaka of the Chāndogya was an elaboration of sarvaṁ khalv idaṁ brahma. Thus, in its own context the statement was part of a meditation passage, but it branched out in the proper knowledge context as well. sarvaṁ khalv idaṁ brahma tajjal n ity uktaṁ [ChU 3.14.1], kathaṁ tasmāj jagad idaṁ jāyate, tasminn eva ca līyate 'niti ca tenaivety etad vaktavyam. ChUBh Introduction to Chapter Six, VII.335. 19 upāsana-vidhi-pradhānāni hi tāni. BSBh 3.2.14, III.585. 20 santy ubhaya-li gāḥ śrutayo brahma-viṣayāḥ—sarva-karm sarva-k maḥ sarva-gandhaḥ sarva-rasaḥ [ChU 3.14.3] ity evam-ādyāḥ saviśeṣa-li gāḥ; asthūlam anaṇv ahrasvam ad rgham [BĀU 3.8.8] ity evam-ādyāś ca nirviśeṣa-li gāḥ. BSBh 3.2.11, III.582. 380 These passages are statements of duality and are for meditation, and have an appropriate injunction of meditation.21 To illustrate the difference between these two kinds of texts, we may compare the aforementioned ṇḍilya-vidy with the madhu-vidy of the Bṛhad- raṇyaka 2.5.1-19. The texts are similar insofar as the same Upaniṣadic meditational structure of correspondence seems to be at play: in the ṇḍiyla-vidy the correspondence is between “the Self of mine within the heart” that is smaller than a grain of rice or barley, and the great Brahman that is larger than the earth, the sky and the intermediate space. This Self-Brahman is, further, identified with “all,” both in general and as a variable: this all is Brahman, and Brahman has all desires, actions, tastes, and smells. The identity between the two counterparts is that of assimilative identification, because the positive characteristics of Brahman are not denied: the text is one of meditation that is predicated on affirming the duality between the counterparts as intended. The madhu-vidy similarly draws the identity of the macrocosmic with the microcosmic, but in more explicit detail: there is a radiant and immortal person that resides both in the waters and in semen, in fire and in speech, in wind and in breath, in the sun and in sight, in the quarters and in hearing, in the moon and in the mind etc. Each macrocosmic counterpart is called “the honey of all beings,” and the radiant and immortal person on both sides of the counterpart is identified with the Self, with Brahman, and like in the ṇḍilya-vidy —with everything, sarva. In the madhu-vidy , however, the crucial statement is the final statement in 2.5.19, which says that “this Brahman is without a prior and without a posterior, without anything inner and outer; this Self is Brahman that directly experiences everything.”22 This everything, thus, is different from the everything of Śā ḍilya, because it serves the purpose of bringing home the point that there is only the non-dual Brahman 21 Cf. Hirst 2005:91-2. 22 tad etad brahmāpūrvam anaparam anantaram abāhyam. ayam ātmā brahma sarvānubhūḥ. 381 that is consciousness in nature. This is accomplished through the negative statement that cancels out the carefully drawn distinctions in the previous eighteen paragraphs. The text, thus, is not about duality and meditation, but about non-duality and knowledge. As for what some say, “even the Vedic texts that present forms of Brahman are just for the purpose of understanding the formless Brahman through the denial of multiplicity, and have no independent purpose,” that is not quite right either. How so? Some multiplicity is spoken of in the context of the higher knowledge, for instance in the text that begins in “for, to him are yoked ten organs, hundreds of them, he is the organs, the ten, the thousands, the many, the infinite.” [BĀU 2.5.19] Since the text ends in “this Brahman is without a prior and without a posterior, without anything inner and outer; this Self is Brahman that directly experiences everything” [ibid], that text may be for the purpose of denial. However, when multiplicity is spoken of in the context of an injunction to meditation, such as in “He consists of the mind, has life-breath as its body, effulgence in nature” etc. [ChU 3.14.2], it is not right that these be denied, because they are related to an injunction of meditation, such as in “he should make a resolve” [ibid.]. Since the text presents characteristics of such kind for the purpose of meditation, it is not right to postulate that they are for the purpose of denial through the secondary signification function.23 There are, on the other hand, Upaniṣadic passages in which it is not immediately clear just what they intend to affirm, because they present Brahman or the Self in mutually excluding terms. They talk of Brahman as having no distinguishing qualities, nirvi eṣa or nirguṇa, but then add further descriptions in which some positive qualities are predicated and, lo and behold, seem to be intended. Such is the case, for instance, in the dahara-vidy of the chapter eight of the Ch ndogya, where the Self that one should search out and know distinctly is described as free from faults, aging, death, sorrow, hunger, and thirst—all good—and then as having true desires and resolves, about which we quickly learn some juicy things. As Śa kara says, we hear not only 23 yad apy āhuḥ—ākāra-vādinyo ’pi śrutayaḥ prapañca-vilaya-mukhena anākāra-pratipatty-arthā eva, na p thag-arthā iti, tad api na samīcīnam iva lakṣyate; katham? ye hi para-vidyādhikāre kecit prapañcā ucyante, yathā—yukt hy asya harayaḥ at da eti, ayaṁ vai harayo’yaṁ vai da a ca sahasr ṇi bahūni c nant ni ca ity evam-ādayaḥ—te bhavantu pravilayārthāḥ. tad etad brahm pūrvam anaparam anantaram ab hyam ity upasaṁhārāt. ye punar upāsanā-vidhānādhikāre prapañcā ucyante, yathā—mano-mayaḥ pr ṇa- ar ro bh -rūpaḥ ity evam-ādayaḥ—na teṣāṁ pravilayārthatvaṁ nyāyyam. sa kratuṁ kurv ta ity evaṁ-jātīyakena prak tenaiva upāsana-vidhinā teṣāṁ sambandhāt. śrutyā ca evaṁ-jātīyakānāṁ gu ānām upāsanārthatve ’vakalpamāne na lakṣa ayā pravilayārthatvam avakalpate. BSBh 3.2.21, III.591. 382 about the Self, but about desires as well. As the passage unfolds, it turns out that it is the “true desires and resolves” statement that is intended to be affirmed, because the statement of result talks about attaining these true desires and resolves. The whole passage, though containing descriptions of Brahman pure and simple, belongs to what Śa kara calls saguṇa/apara-vidy , or passages in which the positive qualities predicated of Brahman are intended. In terms of statements about Brahman, the negative descriptions prevail because they present the true Brahman, but on the level of text the positive characteristics are not intended for denial, because the text is about gradual liberation or krama-mukti. Although “true desires and resolves” are characteristics of Brahman contingent on Brahman’s having limiting adjuncts,24 they are intended in the text, because the text is for dummies who cannot comprehend Brahman as a real thing if it is wholly without qualities, and are intent on reaching brahma-loka. “For, the absolutely real, non-dual Brahman that is devoid of distinctions such as direction, space, qualities, motion, and results appears as good as non-existent to the slowwitted. Therefore, scripture thinks, ‘Let me first set them on the right path, and I will slowly make them grasp the absolutely real as well.’”25 This text is different from the ṇḍilya-vidy because it does contain true descriptions of Brahman, which can be salvaged, so to speak, in composite accounts when Brahman is characterized by drawing from the Upaniṣadic corpus. But, satya-k ma and satya- saṅkalpa are never denied in the text itself, and the passage overall is for meditation. There is, thus, a restricted and a broader sense of vad nta-v kya or an Upaniṣadic statement or text. This is a significant point to keep in mind when one is reading Śa kara: it is important to be clear what sphere any given Upaniṣadic passage belongs to: para/nirguṇa-vidy 24 sa kalpāḥ kāmāś ca śuddha-sattvopādhi-nimittā īśvarasya, citragu-vat; na svataḥ neti neti [BĀU 2.3.6] ity uktatvāt. ChUBh 8.1.5, VII.479. 25 dig-deśa-gu a-gati-phala-bheda-śūnyaṁ hi paramārtha-sad advayaṁ brahma manda-buddhīnām asad iva pratibhāti. ChUBh 8.1.1, VII.471. 383 and knowing, or apara/saguṇa-vidy and action/meditation. So, we may say that in the Upaniṣads there are passages in which knowledge, non-duality, and liberation are intended, and then there are passages on meditation which are generally geared towards promotion, abhyudaya. More specifically, some are explicitly about promotion, and they correspond to the k mya or prat kop sana meditations that we know from the BS account, which identify Brahman with such things as the mind, the sun, name, etc., through deliberate fancy. There are, next, passages that delineate meditations on ritual subsidiaries, the aṅg vabaddha meditations from the BS, whose purpose is to improve the results that the ritual to which they are tied brings. Finally, there are passages on meditations that serve the purpose of gradual liberation, and these correspond to the brahma-vidy s of the BS, involving meditational absorption on Brahman as one’s Self, meditation on Oṁ, and the meditations on pr ṇa or life breath, the aparaṭk rya-brahman. Brahman is known in two forms, one that is qualified by adjuncts that are diversities such as name and form, and another one opposite to that that is devoid of all adjuncts. … There are Upaniṣadic passages by the thousands that show, through the difference concerning knowledge and ignorance, that Brahman has two forms. Among these, all practices that are characterized by a subject and object of meditation belong to Brahman in the state of ignorance. Among these, further, some are meant for promotion, some are meant for gradual liberation, and some are meant for improving the efficacy of ritual. They are different through the difference of the adjunct or specific quality. Although the object of meditation which is qualified by the individual qualities is the single Supreme Self, the Lord, still their results are different in accordance with the quality that is meditated on. … [From the Bhagavad-G t 10.41 it follows that] wherever there is preponderance of might, that is enjoined to be meditated with the idea, “It is the Lord.” … Likewise, although knowledge of the Self is the cause of immediate liberation, it nevertheless must be taught through certain adjuncts as well, but there the relation to adjuncts is not intended. Thus, it becomes important to examine the trend of the passages in order to determine whether a given passage belongs to the para or the apara sphere.26 26 dvi-rūpaṁ hi brahmāvagamyate—nāma-rūpa-vikāra-bhedopādhi-viśiṣ am, tad-viparītaṁ ca sarvopādhi-vivarjitam. ... sahasraśo vidyāvidyā-viṣaya-bhedena brahma o dvi-rūpatāṁ darśayanti vedānta-vākyāni. tatrāvidyāvasthāyāṁ brahma a upāsyopāsakādi-lakṣa aḥ sarvo vyavahāraḥ. tatra kānicid brahma a upāsanāny abhyudayārthāni, kānicit krama-mukty arthāni, kānicit karma-sam ddhy-arthāni. teṣāṁ gu a-viśeṣopādhi-bhedena bhedaḥ. eka eva tu para ātmeśvaras tais tair gu a-viśeṣair viśiṣ a upāsyo yady api bhavati, tathāpi yathā-gu opāsanam eva phalāni bhidyante. … yatra yatra vibhūty-ādy-atiśayaḥ, sa sa īśvara ity upāsyatayā codyate. … evaṁ sadyo-mukti-kāra am apy ātma- jñānam upādhi-viśeṣa-dvāre opadiśyamānam apy avivakṣitopādhi-sambandha-viśeṣaṁ parāpara-viṣayatvena sandihyamānaṁ vākya-gati-paryālocanayā nir etavyaṁ bhavati. BSBh 1.1.11, I.49-51. 384 The same classification of meditations into three kinds is given in the introduction to the Ch ndogya-Bh ṣya, with a rationale what is it that ties them with knowledge into the same corpus, the Upaniṣads: Such being the case, in this context of non-duality, there are described meditations that are means to promotion, meditations on a form of Brahman that is slightly modified from the non-dual and mentioned in texts such as “made of the mind and having pr ṇa as its body” and bringing results that are proximate to liberation, and meditations related to various ritual subsidiaries and augmenting the ritual results. They are mentioned in the context of non-duality because of the commonalities of secrecy and mental modification; for, these other meditations are mental modifications just as much as knowledge of non- duality is.27 In the Upaniṣads, then, in this broader sense, there are four kinds of passages, ved nta- v kya, only one of which are about knowledge. Let us call the last, in general, passages about propositional knowledge of Brahman and the Self. They contain facts about Brahman and the Self, and ways to illustrate and teach such facts. Propositional knowledge of the Self will also be found in texts that are not about knowing, but about meditation: there can be ved nta-v kyas in the strict sense within ved nta-v kya texts in the broader sense. Such knowledge can be salvaged from these passages and used in composite descriptions, but whatever positive qualities are attributed to the Self, such as having true resolves and desires, must not enter such descriptions. When we think about the Veda as a corpus, then, we can ask two questions. While we saw that the Veda as a pram ṇa must be about how things are, additionally an impetus for action can be served by its individual texts, and things that are not strictly true but are useful may also be stated—or not. Based on this, we can ask whether a Vedic passage is about knowledge or about action, in which case all the passages on meditation will fall into the second group. 27 tatraitasminn advaita-vidyā-prakara e abhyudaya-sādhanāny upāsanāny ucyante, kaivalya-sannik ṣ a-phalāni ca advaitād īṣad-vik ta-brahma-viṣayā i mano-mayaḥ pr ṇa- ar raḥ [ChU 3.14.2] ity-ādīni karma-sam ddhi-phalāni ca karmā ga-sambandhīni, rahasya-sāmānyāt mano-v tti-sāmānyāc ca—yathā advaita-jñānaṁ mano-v tti-mātraṁ, tathā anyāny apy upāsanāni mano-v tti-rūpā i—ity asti hi sāmānyam. ChUBh Introduction, VI.9. 385 Alternatively, we can ask whether something in the Veda is bodily or mental, in which case the meditation passages will be classed with the propositional knowledge of Brahman and the Self, against the ritual texts. It is the second criterion, to Śa kara’s mind, that sets the Upaniṣads apart as a corpus from the rest of the Veda. If we, however, apply the two criteria simultaneously, we would get texts about ritual, texts about meditation, and texts about Brahman and the Self, corresponding to the well-known threefold distinction of karma-k ṇḍa, up sana-k ṇḍa, and jñ na-k ṇḍa. In this and in the next chapter, we will concern ourselves first with the propositional knowledge of the Upaniṣads and then with the direct process of liberation focused on understanding this propositional knowledge. In other words, we will examine jñ na-k ṇda proper and the restricted sense of the Upanisads or ved nta-v kya. Once we have done that and we conclude this part of the dissertation, we will be able to see through contrast and “as clearly as a wood-apple in our palm” what precisely Śa kara was advocating for, as well as to understand better his significance in the history of Vedānta. The Scope of Para-vidy Texts In his BSBh, Śa kara says that the Upaniṣadic statements have two kinds of uses: in one, they ascertain the nature of the supreme Self; in the other, they present the identity of the individual Self with the supreme Self.28 For practical purposes, I will call the second kind “identity statements,” and whenever I use that lexeme I will refer to statements that identify the individual Self as the cognitive agent, the vijñ n tman, with the supreme or pure Self, the param tman. This twofold division is obviously the baseline classification of the ved nta-v kya in the 28 dvirūpā hi vedānta-vākyānāṁ prav ttiḥ—kvacit paramātma-svarūpa-nirūpa a-parāś kvacid vijñānātmanaḥ paramātmaikatvopadeśa-parā. BSBh 1.3.25, I.186. Cf. Murty 1959Ś74. 386 restricted sense of texts that are about knowledge and the higher Brahman. Another classification can be drawn through coordination of the first prose chapter of the US and BSBh 2.1.14. There are two kinds of Upaniṣadic statements: the first are statements of the characteristics of Brahman, and the second are “scriptural statements meant for presenting the non-duality of the Self.”29 The first categories in the two classifications evidently correspond to one another, and they further correspond fully to the “positive” and negative characteristics of Brahman that were part of the paradigmatic brahma-vidy in the BS. I write “positive” in quotation marks to distinguish these from the other positive characteristics, such as satya-k ma, that were restricted to the individual meditations. The “positive” characteristics were not positive for Śa kara, in any case. While we will deal with these in detail later, exemplary lists of such texts that “ascertain the nature of the supreme Self” of that “present the characteristics of Brahman” are drawn in BSBh 4.1.2 and USP 1.6, and they include the classical “positive” characteristics in satyaṁ jñ nam anantaṁ brahma, “Brahman is Being, consciousness, limitless,”30 and vijñ nam nandaṁ brahma, “Brahman is consciousness, bliss,”31 as well as a host of negative characteristics of Brahman drawn mostly, but not exclusively, from Yājñavalkya’s teachings at Janaka’s brahmodya. The second kind of statements in the second classification, those that “present the non- duality of the Self,” are evidently not equivalent to the identity statements, though they are directly related to them. A list of such passages drawn from the USP 1.6 and BSBh 2.1.14 shows 29 ātmaikya-pratipādana-parāḥ śrutīḥ in USG 1.6ś ātmaikatva-pratipādana-param vacana-jātam in BSBh 2.1.14, II.309; para-brahmaikatva-pratipādikāḥ śrutayaḥ in BĀUBh 2.1.20, VIII.250. 30 TU 2.1.1. 31 BĀU 3.9.28.7, 8. 387 that these are texts that present Brahman in a general way, as the great ground of Being and the cause of creation: ▪ yatra n nyat pa yati, “Where one does not see another,” from ChU 7.24.1, the section of ▪ the text on Brahman as plenitude. tm v idam eka ev gra s t, “In the beginning this world was the Self, one only,” from ▪ AiU 1.1.1. tmevedaṁ sarvam, “This world is the Self,” from the ChU 7.25.2. ▪ brahmevedaṁ sarvam, “This world is Brahman,” from BĀU 2.5.1.32 ▪ sarvaṁ khalv idaṁ brahma, “All this is Brahman,” from the ChU 3.14.1, and its ▪ elaboration in: sad eva somyedam agra s d ekam ev dvit yam, “In the beginning, my dear, this world was Being alone, one only without a second,” from ChU 6.2.1. ▪ aitad tmyam idam sarvam, tat satyaṁ sa tm tat tvam asi, “This whole world is a state of having that as its Selfś it is Reality, it is what you are,” of ChU 6.8.7 ff. ▪ idaṁ sarvaṁ yad ayam tm , “All these are what the Self is,” in BĀU 2.4.6. ▪ neha n n sti kiñcana, “There is no difference whatsoever here,” in BĀU 4.4.19. The context where these texts are discussed is the BS section on origination, against the Sā khya prakṛti and for the notion of satk rya-v da or the doctrine of the persistence of the cause in the effect. As kinds of texts, these passages belong to the same category as the canonical Upaniṣadic statement that introduces Brahman as the cause of origination, maintenance, and dissolution, TU 3.1.1Ś “That from which these beings are born, on which, once born, they live, and into which they return—know that distinctlyŚ it is Brahman.” These are, then, statements about the causal Brahman, or Brahman as the material and efficient cause of the world. However, Śa kara’s choice of these specific statements as paradigmatic of the causal Brahman is rather important, for several related reasons. First, while they are not equivalent to the identity statements, they provide the ground for them, insofar as Brahman is presented as the sole reality that does not admit of a second, including the individual Self. Invariantly, the two kinds of statements are 32 As discussed by Mayeda, the brahmevedaṁ sarvam in that form occurs in the Nṛsiṁhottara-T pan ya Upaniṣad 7.3, and it would be unlikely for Śa kara to quote that text. BĀU 2.5.1 is brahmedaṁ sarvam, which might have been assimilated to the tmaivedaṁ sarvam which occurs just before it in the Upade a-S hasr text. Mayeda 2006b:229. 388 related, and they are commonly related as opening and concluding statements of an Upaniṣadic text, upakrama and upasaṁh ra: first the causal Brahman is presented as the sole reality, and eventually this sole Brahman is identified with the individual Self.33 Second, to Śa kara’s mind, they present an account of causality that does not involve Brahman’s actual transformation into or even evolution of creation, but simply a way of bringing home the point that Brahman as the cause is the creation, that is, that Brahman is everything. Third, they are more specific as to what kind of a cause Brahman is, insofar as they relate well with the statements that “ascertain the nature of the supreme Self.” They are more determinate than the mere positing of Brahman as the cause of all created beings, and Śa kara, in fact, says as much: their purpose is to ascertain the nature of Brahman in its causal role.34 We will address the third point somewhere down the line. Let us now see in very broad terms why Śa kara claimed that these statements about the causal Brahman affirmed the sole reality of Brahman rather than that of real creation. His reasoning is predominantly theological, with a splash of metaphysics and epistemology, and the knockdown argument is the most hallowed principle of Vedic theology, namely that scripture must serve a human good. The texts that present Brahman as the cause commonly proceed to describe creation of entities, or to talk about them in a way that seems to acknowledge their reality: in the Ch ndogya, we read about the doctrine of triplication, where Being evolves into heat, water, and food, and further into 33 Cf. BĀUBh 2.1.20, VIII.258-9: upakramopasaṁhārābhyāṁ ca—sarvāsu hy upaniṣatsu pūrvam ekatvaṁ pratijñāya, d ṣ āntair hetubhiś ca paramātmano vikārāṁśāditvaṁ jagataḥ pratipādya, punar ekatvam upasaṁharati; tad yathā ihaiva tāvat idaṁ sarvaṁ yad ayam tm iti pratijñāya, utpattisthiti-laya-hetu-d ṣ āntaiḥ vikāra- vikāritvādy-ekatva-pratyaya-hetūn pratipādya, anantaram ab hyam, ayam tm brahma [BĀU 2.5.19] ity upasaṁhariṣyati. “Also, through the opening and concluding statements. For, in all the Upaniṣads, first unity is presented as a thesis; then by means of reasons and the illustrations of origination, maintenance and dissolution the arguments for the notion of unity of the transformation and transforming cause are presented, and the conclusion is stated in texts as ‘without interior or exterior, this Self is Brahman.’” 34 vedānta-vākyānāṁ jagat-kāra āvadhāra a-paratāś ... evaṁ-jātīyakasya kāra a-svarūpa-nirūpa a-parasya vākya- jātasya ... BSBh 1.4.14, I.250-51. 389 several kinds of living beings. In the Taittir ya, we read about the creation of the five elements from the SelfŚ space, air, fire, water, and earth. In Yājñavalkya’s teaching to Maitreyī, after the statement “All these are what the Self is,” there follows a description of how all the details of creation emerge from and are absorbed in the great Being. None of these descriptions, however, are related to the attainment to liberation or any other human good: specifically, there are no statements of the kind that relate the attainment of liberation, of Brahman, or of the Self, to knowledge of the creation of the world from Brahman. Therefore, the descriptions of creation that are attached to the statements of Brahman as the cause do not serve any good of man in some direct manner. “What is intended to be taught is not the manifestation of the world, for no human goal is seen or heard as related to such manifestation.”35 Here we should pause to list yet another kind of Upaniṣadic statements that belong to the para-vidy category, those that present the goal of human life that is liberation. Some of these include: ▪ brahmavid pnoti param, “The knower of Brahman attains the highest,” of TU 2.1.1. ▪ tarati okam tmavit, “The knower of the Self crosses over sorrow,” of ChU 7.1.3, ▪ tam eva viditv ti mṛtyum eti, “Knowing him alone, one crosses over death,” of ŚU 3.8. ▪ tatra ko mohaḥ kaḥ oka ekatvam anupa yataḥ, “What illusion or sorrow could there be when one sees unity?” of ĪU 7. ▪ abhayaṁ vai janaka pr pto ‘si, “You have attained fearlessness, Janaka,” of BĀU 4.2.4. ▪ nandaṁ brahmaṇo vidv n na bibheti kuta canaṬ etaṁ ha v va na tapatiṬ kim aham sadhu n karavamṬ kim ahaṁ p pam akaravam, “The knower of the bliss of Brahman does not fear anything. He does not agonize, thinking, ‘Why did I not do what is right? Why did I do wrong?’” of TU 2.4.36 They are statements of liberation that obtains when the Self or Brahman is known, and knowing the Self or Brahman really means knowing that Self which is not liable to transmigration, the 35 na hy ayaṁ s ṣ y-ādi-prapañcaḥ pratipipādayiṣitaḥ. na hi tat-pratibaddhaḥ kaścit puruṣārtho d śyate śrūyate vā. BSBh 1.4.14, I.252. 36 The list is based on BSBh 1.4.14 and 4.3.14. 390 “Higher Self” or the higher teaching of the BS 3.4.8, not, however, in itself, but as being the individual Self, taught in statements such as tat tvam asi.37 We should make a good note of this: liberation that obtains by knowing Brahman obtains by knowing Brahman as identical with the individual Self, the vijñ n tman. In other words, the statements of result or liberation are most directly related to the identity statements from the first twofold classification. Such statements identify the individual Self, a being which is by definition created, and a conscious entity, with the supreme Self, and affirm that no difference of the subject-object kind obtains between the two. These texts are on the one hand most directly related to the statements of liberation, and on the other they are the concluding passages or upasaṁh ra of those Upaniṣadic text in which Brahman is presented as the sole reality, through the affirmation that the conscious Self as the product of Brahman’s creation just is Brahman. The statements of liberation are related to the identity statements “scripturally,” insofar as liberation is explicitly said to be a result of knowing Brahman or the Self. They are also related “theologically,” because the inquiry into Brahman terminates in understanding one’s identity with Brahman, and once the identity statements have been fully grasped, there remains nothing further that one could possibly wish to knowŚ Śa kara says that there remains no k ṅkṣ once Brahman has been known in full. As we will remember from the Second Chapter, the notion of k ṅkṣ signified the syntactic expectancy of words and larger linguistic units that were not sufficient on their own for full understanding of whatever required understanding, but also the 37 See, for instance, BSBh 1.4.14, I.252-53: brahma-pratipatti-pratibaddhaṁ tu phalaṁ śrūyate—brahmavid pnoti param [TU 2.1], tarati okam tmavit [ChU 7.1.3], tam eva viditv ti mṛtyum eti [ŚU 3.8] iti. pratyakṣāvagamaṁ cedaṁ phalam, tat tvam asi ity asaṁsāryātmatva-pratipattau satyāṁ saṁsāry-ātmatva-vyāv tteḥ. “Scripture says that the result is contingent on understanding Brahman. And, this result is directly experienced, as per the statements ‘The knower of Brahman attains the highestś’ ‘The knower of the Self crosses over sorrowś’ ‘Knowing him alone, one crosses over death,’ for when the identity statement ‘You are that’ is understood, the transmigrating Self ceases.” 391 curiosity on the part of the cognitive subject. Such curiosity and syntactic expectancy were fully discharged when the result of a striving had been secured, and Śa kara says that once one had known that Self which is “single, pure, eternal, etc.,” no further curiosity was possible.38 On the other hand, the descriptions of creation involved an abundance of curiosity, until their understanding culminates in the identity statements. This can be illustrated by looking at the eighth section of the sixth chapter of the Ch ndogya, where hunger and thirst the paradigmatic characteristics of transmigration are sequentially reduced to food, water, and heat, and finally to Being, which Being is identified with the individual Self in tat satyaṁ sa tm tat tvam asi. Such reduction takes place because at each subsequent step there remains the curiosityŚ “What is hunger? What is food? What is water? What is heat?”39 But once the reduction terminates in Brahman that is the Self, there could not possibly be any curiosity remaining, at least not without running the risk of one’s head shattering. Liberation was related to the identity statements not only “scripturally” and “theologically,” but also “evidently.” Liberation, as we saw in Chapter Seven, was an “evident” result in Śa kara’s system, dṛṣṭa, and Śa kara certainly had axes to grind when he added to the paradigmatic text of liberation, “the knower of Brahman reaches the highest,” statements that 38 “Also, this evidence [of the Upaniṣads] is the final evidence, propounding the unity of the Self, after which no syntactic expectancy remails. In the world, when one says 'yajeta,’ the verb needs to be supplied with ‘who,’ 'with what,’ and 'how.' Such is not the case when it is said ‘You are that’ or ‘I am Brahman,’ No curiosity remains in such case, because such understanding extends over the Self which is everything. Curiosity is possible when there is some other thing remaining. But there is no other thing than the Self over which one could be curious about.” api cāntyam idaṁ pramā am ātmaikatvasya pratipādakam—nātaḥ paraṁ kiñcid ākā kṣyam astiś yathā hi loke yajetety ukte, kim? kena? katham? ity ākā kṣyate; naivaṁ tat tvam asi, ahaṁ brahm smi ity ukte, kiñcid anyad ākā kṣyam asti— sarvātmaikatva-viṣayatvād avagateḥś sati hy anyasminn avaśiṣyamā e ’rthe ākā kṣā syātś na tv ātmaikatva- vyatireke āvaśiṣyamā o ’nyo ’rtho’sti, ya ākā kṣyeta. BSBh 2.1.14, II.313. 39 “The origin statements cannot give such knowledge that leaves no room for curiosity, and it is evident that they aim at something outside of themselves. Thus, the statement 'Know this sprout, the body, to have come out of something, for it cannot be without a root,' culminates in the statement that Being is the source of the world that is to be known.” naivam utpatty-ādi-śrutīnāṁ nirākā kṣārtha-pratipādana-sāmarthyam astiś pratyakṣaṁ tu tāsām anyārthatvaṁ samanugamyateś tathā hi—tatraitac chuṅgam utpatitaṁ somya vij n hi nedam amūlaṁ bhaviṣyati [ChU 6.8.3] ity upanyasya udarke sata evaikasya jagan-mūlasya vijñeyatvaṁ darśayati. BSBh 4.3.14, III.835-36. 392 talk about the absence of the psychological faults such as illusion and grief: as we shall see in the next chapter, liberation was a result that was literally evident, through the absence of the psychological faults that were outgrowths of ignorance. Such evident liberation could possibly follow only when one had understood one’s true nature as Brahman, where ignorance does not obtain, and not by understanding creation. The texts that present Brahman as the sole reality, on their part, are commonly the opening statements of a text, upakrama, or statements that set the topic under discussion. It is a very important feature of theirs that they present this causal Brahman as a conscious entity. They do so by attributing to Brahman reflection or will, an instance of consciousness, in statements of the “let me be many, let me create progeny” kind. We may illustrate this with the opening statement of the sixth chapter of the Ch ndogya, “In the beginning, my dear, this world was Being, one without a second,” which proceeds to state that this Being “thought to itselfŚ ‘Let me become many, let me create progeny.’”40 We may also illustrate it with the Taittir ya account of creation, where the creation of progeny follows as a result of will rather than reflectionŚ “He desiredŚ ‘Let me become many, let me create progeny.’”41 Alternatively, they do so by explicitly denoting the causal Brahman “the Self,” tman, as in the Taittir ya 2.1.1Ś “From this Self, space came into being; from space, air; from air, fire; from fire, water; from water, earth; from earth, plants; from plants, food; from food—man.”42 The causal Brahman is, thus, a conscious entity, and that fact should provide the ground for establishing its identity with the cognitive agent that is the individual Self, the vijñ n tman, in the identity statement. Such, however, would be the 40 sad eva somyedam agra āsīd ekam evādvitīyam. … tad aikṣata. bahu syāṁ prajāyeyeti. ChU 6.2.1-3. 41 so ‘kāmayata. bahu syāṁ prajāyeya. TU 2.6.1. 42 tasmād vā etasmād ātmana ākāśaḥ saṁbhūtaḥ. ākāśād vāyuḥ. vāyor agniḥ. agner āpaḥ. adbhyaḥ p thivī. p thivyā oṣadhayaḥ. oṣadībhyo ‘nnam. annāt puruṣaḥ. TU 2.1.1. 393 case if this causal Brahman is the sole conscious entityŚ that is why Śa kara calls these “statements that present a single Self.”43 The creation of progeny and creation in general, however, does not sit well with the descriptions of Brahman as a sole reality, both in the passages that introduce the causal Brahman in the opening statements, and in the concluding statements that explicitly identify the created individual Self with the supreme Self. One way to deal with this issue would be to interpret the identity statements as a case of assimilative identification rather than a true identity, in which case the creation statements would be taken as statements of fact. The para-vidy texts, however, are not texts of meditation, but of knowledge, and the solution to this “sole reality” versus “factual creation” problem is to appeal to the negative characteristics of Brahman from the texts that “ascertain the nature of the supreme Self.” To refresh our memory, these were characteristics such as “neither short nor long, neither gross nor fine, without birth, aging and death, fearless,” etc., which deny of Brahman such characteristics that are liable to change. These are directly contradictory to the statements of creation because they present Brahman as a permanently changeless entity. Brahman is not an entity that can intrinsically accommodate mutually exclusive characteristics that things otherwise may have under different aspects, for instance when something is alternatively in motion or rest. Brahman cannot be both extended and not extended, with hunger and thirst and without them, both unborn and undergoing transformation 43 “By describing Brahman as knowledge and as the agent of will, it is ascertained that Brahman is awareness, and Īśvara is described as the cause in the sense of not serving the purpose of othersś by the word 'Self' used later as one that has entered all in the sequence of sheaths of the body, he has been ascertained as the inner Self within all. In 'may I become many, may I be born,' by teaching the becoming of all concerning the Self it was declared that the created transformations are non-different from the creator.” atra tāvaj jñāna-śabdena pare a ca tad-viṣaye a kāmayit tva-vacanena cetanaṁ brahma nyarūpayatś apara-prayojyatveneśvaraṁ kāra am abravītś tad-viṣaye aiva pare ātma-śabdena śarīrādi-kośa-paramparayā cāntar-anupraveśanena sarveṣām antaḥ pratyag-ātmānaṁ niradhārayatś bahu sy ṁ praj yeya iti cātma-viṣaye a bahu-bhavanānuśaṁsanena s jyamānānāṁ vikārā āṁ sraṣ ur abhedam abhāṣata. BSBh 1.4.14, I.251. 394 in which itself becomes born as many, etc. Permanently changeless means not liable to transformation under any description or aspect.44 So, the descriptions of creation are both uncomfortable with the statements that present Brahman as the sole reality, and contradictory to the negative descriptions of Brahman. In adjudicating their truth value, relative strength, and use, it is paramount to bear in mind that they are not directly relevant to the attainment of liberation: we saw that liberation was directly related to the identity statements. However, they must serve some purpose. Here the principle of textual coherence or ekav kyat kicks in. The meaning of a text as a unit is superordinate to the meaning of individual sentences. Since a result is affirmed to follow from understanding Brahman as a non-dual entity, and since no result attaches to texts that present Brahman as factually creating the world, these creation statements must be read along with the opening and concluding statements that present Brahman as the single reality, that is, in their light: It is not possible to postulate [a result attached to the creation statements], because we understand that they form sentential unity with the respective opening and concluding statements about Brahman.45 And, it cannot be that the understanding of Brahman’s transformation into the world is independently intended for some result, as the understanding of the unity of the Self is the means of liberation, for there is no evidence of that. Scripture shows that a result follows solely from understanding the sole reality of the permanently changeless Brahman, of the 44 “Is it not that by adducing the clay illustration, it is understood that the purport of scripture that Brahman has transformations? For, clay and the like are considered consensually in the world to be transforming. – No, there is evidence from the statements that deny all changes that Brahman is permanently changelessŚ ‘This is the great, unborn Self, unaging, undying, immortal, fearless Brahmanś’ ‘This Self is, ‘not thus, not thusś’’ ‘It is neither gross nor fine.’ And, it is not reasonable that the single Brahman have the properties of transformation and be devoid of them. – Can it not be like having motion and rest? – No, it has been qualified as permanently changeless. The permanently changeless Brahman cannot have different attributes, like rest and motion. As I’ve just said, Brahman is permanently changeless, because all transformations have been denied of it.” nanu m d-ādi-d ṣ ānta-pra ayanāt pari āmavad brahma śāstrasyābhimatam iti gamyateś pari āmino hi m d-ādayo ’rthā loke samadhi-gatā iti. nety ucyate—sa v eṣa mah n aja tm jaro ’maro ’mṛto ’bhayo brahma [BĀU 4.4.25], sa eṣa neti nety tm [BĀU 4.4.22], asthūlam anaṇu [BĀU 3.8.8] ity ādyābhyaḥ sarva-vikriyā-pratiṣedha-śrutibhyaḥ brahma aḥ kū asthatvāvagamātś na hy ekasya brahma aḥ pari āma-dharmatvaṁ tad-rahitatvaṁ ca śakyaṁ pratipattum; sthiti- gativat syād iti cet, naś kū asthasyeti viśeṣa ātś na hi kū asthasya brahma aḥ sthiti-gativad aneka-dharmāśrayatvaṁ sambhavatiś kū asthaṁ ca nityaṁ brahma sarva-vikriyā-pratiṣedhād ity avocāma. BSBh 2.1.14, II.313-14. 45 na ca kalpayituṁ śakyate, upakramopasaṁhārābhyāṁ tatra tatra brahma-viṣayair vākyaiḥ sākam eka-vākyatāyā gamyamānatvāt. BSBh 1.4.14, I.252. 395  kind that begins in “This Self is ‘not such,’ ‘not such,’” and ends in “You have become fearless, Janaka.” Such being the case, this is provedŚ since in the context where Brahman is discussed the result is attained from understanding Brahman that is devoid of all distinguishing characteristics, all the fruitless descriptions of Brahman such as its transformation into the world are employed for the purpose of understanding Brahman, through the principle “That which is fruitless in the proximity of something fruitful is its auxiliary.”46 This is the same principle through which the r d-upak raka rituals were attached to the principal rituals in a Darśa-pūr amāsa. The descriptions of creation are fruitless intrinsically, and become fruitful through the fact of their proximity to the statements that present Brahman as the cause and a conscious being. The obvious question now is, what does it mean to read the descriptions of creation in the light of the statements that present Brahman as the cause? They should be read as intending to bring home the point that Brahman is the cause, while not affirming that it factually transforms into the creation. To illustrate this, in the well-known case from the sixth chapter of the Ch ndogya, the statement of triplication of Being into heat, water, and food does not intend to affirm that these three principles and their subsequent products are transformations of Being, but that they are non-different from Being; thus, that they are Being. This is what the Upaniṣad itself affirms, to Śa kara’s mind, through using the examples of clay, iron, and copper as substances and clay things, nail-cutters and copper trinkets as their products, and the claim that only the respective substances are real while the individual products are unreal, mere name, a verbal handle. The use of examples is meant to be a form of analogical reasoning through which scripture conveys a particular notion of causality, in which the effect is absolutely non-different 46 na ca, yathā brahma a ātmaikatva-darśanaṁ mokṣa-sādhanam, evaṁ jagad-ākāra-pari āmitva-darśanam api svatantram eva kasmaicit phalāyābhipreyate, pramā ābhāvātś kū astha-brahmātmaikatva-vijñānād eva hi phalaṁ darśayati śāstram—sa eṣa neti nety tm ity upakramya abhayaṁ vai janaka pr pto ’si [BĀU 4.2.4] ity evaṁ- jātīyakam. tatraitat siddhaṁ bhavati—brahma-prakara e sarva-dharma-viśeṣa-rahita-brahma-darśanād eva phala- siddhau satyām, yat tatrāphalaṁ śrūyate brahma o jagad-ākāra-pari āmitvādi, tad brahma-darśanopāyatvenaiva viniyujyate, phalavat sannidhāv aphalaṁ tad-a gam itivat. BSBh 2.1.14, II.314. 396 from the cause, through employing instances of common cause-effect relationship. Śa kara quotes Gauḍapāda’s succinct statement on this point, which is worthwhile repeating hereŚ “Creation that is taught in various ways, such as through clay, iron, sparks etc., is only a means for introductionś in fact, there is no difference whatsoever.”47 Such analogical reasoning is required because knowledge of Brahman’s being the cause does not obtain otherwise than through analogical reasoning: we do not have experience with causes that do not transform, and thus common cases of causality must be employed as the closest approximation. We need not go into the details of such analogical reasoning, and a fine account is available in Suthren Hirst’s monograph,48 but we should make a note that once the principle that particular instantiations of a cause are false as effects but real as the cause has been grasped, the Upaniṣad can use the descriptions of creation, such as the Ch ndogya passage of triplication, as categories through which individual effects may conceptually be reduced back to Brahman the cause. This is, obviously, required, because we know the world of multiplicity perceptually, but for liberation one needs to come to know it as essentially non-different from Brahman; that is, one needs to know it as Brahman, “one alone, without a second.” This obviously makes the descriptions of creation texts with no truth value, but useful nevertheless, indeed necessary, serving the purpose of the statements about the causal Brahman as their sentential supplements. While Śa kara does not call them as such, this determination is identical with the definition of arthav da statements. We should note this very well and bear it in 47 tathā ca sampradāya-vido vadanti— m l-loha-visphuli gādyaiḥ s ṣ ir yā coditānyathā | upāyaḥ so’vatārāya nāsti bhedaḥ kathañcana. BSBh 1.4.14, I.252. The quote is from Gauḍapāda’s gama- stra 3.14. On this, most relevant is the long comment on BS 2.1.14 in its entirety. 48 Hirst 2005:76-80. 397 mind when we come to discuss mah -v kya in Chapter Ten. Sarvajñātman will explicitly call such texts “arthav da statements that facilitate reasoning.” There is evidently also another kind of Upaniṣadic texts that are part of the para-vidy corpus, and these are texts in which the discussion is not about the causal Brahman as the great ground of Being out there, but about the inner Self. Particularly important among these are TU 2.1.1-2.5.1, the teaching about the five sheaths of the Self, pañca-ko a, and BĀU 4.3.7-4.4.22, Yājñavalkya’s teaching to Janaka about the three states of the SelfŚ waking, dream, and deep sleep. The subject in these texts is the individual transmigrating Self, the vijñ n tman, but their intention is not to teach about it. Rather, by positing this transmigrating Self and its characteristics as the subject, the anuv da, they intend to predicate of that this Self its being the pure, supreme Self, through the gradual elimination of its characteristics.49 We will have much more to say about this later, but we should note that this is how Śa kara also interprets those texts that explicitly mention two Selves, the individual Self and the Lord, such as the famous statement about the two friendly birds from the Ṛg Veda and the Muṇḍaka and vet vatara Upaniṣad: Two birds, who are companions and friends, nestle on the very same tree. One of them eats a tasty fig; the other, not eating, looks on. Stuck on the very same tree, one person grieves, deluded by her who is not the Lord. But when he sees the other, the contented Lord—and the Lord’s majesty—his grief disappears.50 49 “Repeating the characteristics of the transmigrating Self, the text intends to affirm its unity with the supreme Brahman. For, the succeeding statement, ‘It meditates as it were, it runs as it were,’ is seen to be about the removal of the characteristics of the transmigrating Self. … We have proved, thus, that the transmigrating Self of cognition within the organs is the great, unborn Self, the supreme Self.” anūdya saṁsāri-svarūpaṁ pare a brahma āsyaikatāṁ vivakṣati; yataḥ dhy yat va lel yat va [BĀU 4.3.7] ity evam ādy uttara-grantha-prav ttiḥ saṁsāri-dharma- nirākara a-parā lakṣyateś ... yo ’yaṁ vijñāna-mayaḥ prā eṣu saṁsārī lakṣyate, sa vā eṣa mahān aja ātmā parameśvara evāsmābhiḥ pratipādita ity arthaḥ. BSBh 1.3.42, I.220. 50 ŚU 4.6-7; MU 3.1.1-2. Only the first verse is in the Ṛg Veda, 10.164.20. Translation Olivelle 1998:425. 398 Such texts do not intend to affirm the difference between the individual and the supreme Self, but the individual Self is introduced so that its identity with the supreme Self can be affirmed.51 We can formulate this as a general epistemological principle. Scripture is not an authority on the individual Self as such, or its characteristics. This Self is known perceptually, as Mīmāṁsakas have argued. Scripture does, however, say things about this Self that are unknown empirically, and when it does so, it first has to introduce the Self or its characteristics by setting them as the subject under discussion, the anuv da, before proceeding to affirm of it characteristics of the saguṇa-brahman in the texts of meditation, or deny its transmigrating characteristics in texts of knowledge. “When a sentence is found to be purposeful through its part that presents what does not obtain empirically, its part that presents what was already known should be held as a restatement.”52 To summarize, the scope of Upaniṣadic para-vidy texts is constituted by six kinds of statements: (1) the identity statements, or the statements that present the identity of the individual Self with the supreme Self; (2) statements that ascertain the nature of the supreme Self or present the characteristics of Brahman; (3) statements about the causal Brahman that present the sole reality of the Self or the supreme Brahmanś (4) “proliferation” passages or descriptions of the creation of the world that form sentential supplements to the statements about the causal Brahman; (5) statements of liberation that are directly related to the identity statements; and (6) texts that present and eliminate the characteristics of the individual Self so as to show that the individual is the pure, supreme Self. Central among these are the identity statements, because 51 “The individual Self is posited solely with the intention to affirm its identity with the supreme Self, not with any other intentionś” tādātmya-vivakṣayaiva jīvopādānam, nārthāntara-vivakṣayā. BSBh 3.3.34, III.674. 52 api ca aprāptāṁśopadeśena arthavati vākye sañjāte, prāptāṁśa-parāmarśasya nityānuvādatayāpi upapadyamānatvāt. BSBh 3.3.19, III.648. 399 liberation is related to them directly. They are full counterparts to the injunctions of ritual or meditation in the action section of the Veda, and to them we move next. The Identity Statements of the Upani ads The practice of knowledge for Śa kara was fully focused on the identity statements of the Upaniṣads, statements that correlate Brahman the ground of Being with the individual Self. This was in a sense very much like the standard process of liberation in the BS, meditation on Brahman as one’s Self. Śa kara, however, rejected the direct soteriological significance of meditation, for a variety of reasons, most of which we have already seen: some meditations were part of saguṇa-vidy s and were good either for promotion or for gradual liberation; meditation in general affirmed rather than negate the sense of agency. Meditation certainly had a positive role to play in the pursuit of liberation, but that was before one had taken to the life of renunciation: meditation was for purity of the mind and existence, and it was practiced within the scope of the apara-vidy s. Liberation, on the other hand, being directly consequent on the removal of ignorance, was a result of knowledge qua knowledge. The Upaniṣadic vidy s in the BS were built around injunctive statements, and even the triple pan-Vedāntic process was introduced by Yājñavalkya’s urging of Maitreyī. For Śa kara, however, all injunctive texts, even those that were part of nirguṇa-vidy s, presupposed duality, because they were based on a system of action and its contributory factors: they promoted the action of meditation, required an agent and an object of meditation, etc.53 Such statements could not occupy the textual locus around which the practice of brahma-vidy would be organized, because that would mean that the Upaniṣads meant to affirm non-duality as intended. We saw, in fact, this problem already in the doctrine of prasaṅkhy na, whose proponents accepted the 53 BSBh 2.1.22. 400 identity of the Self with Brahman disclosed by statements such as tat tvam asi, but wanted meditation to be enjoined as their subsidiary so that the textual intuition of Brahman could transit from mediacy to immediacy and beyond the subject-object duality. The prasaṅkhy na doctrine was also not acceptable to Śa kara, again for several reasons. First, it seemed to miss the root cause of saṁs raŚ ignorance. If one’s embodiment and transmigration were consequent on superimposing characteristics of one thing over another, the fallback to injunction and the practice of meditation just reaffirmed the same ignorant agency that had been, supposedly, annulled by the understanding of the negative descriptions of the Self in the Upaniṣads.54 Second, the essence of meditation was the repetition of the same notional action over and over. If the root problem of saṁs ra was ignorance, what new contribution would that same notion bring toward the removal of ignorance?55 Finally, meditation involved agency that was inherently antinomous: on the one hand, it reaffirmed ignorance such that so long meditation was practiced, there could be no liberation; on the other hand, it had to be practiced repeatedly, for there was no reason why one would not slip back into bondage when meditation was terminated, if understanding the identity statements was not enough. Was this stream of awareness what liberation was? In Śa kara’s system, thus, the ontological identity of Brahman and the Self replaced the meditative correlation of the two in the BS, and the statements which posit this identity replaced the injunctions of meditation as the core of the practice of brahma-vidy . Clarification of 54 Cf. for instance USP 18.21Ś “Once the superimposition has been negated through the neti neti text, no further injunction of superimposition is in any way possible.” so 'dhyāso neti netīti prāptavat pratiṣidhyate | bhūyo 'dhyāsa-vidhiḥ kaścit kutaścin nopapadyate. 55 See, for instance, NaiS 3.124Ś “If reasoning and hearing do not directly give rise to certain knowledge before, what new result could possibly appear by repetition?” yukti-śabdau purāpy asya na ced akurutāṁ pramām | sākṣād āvartanāt tābhyāṁ kim apūrvaṁ phaliṣyati. 401 meaning of the identity statements through the three processes of ravaṇa, manana and nididhy sana replaced meditative absorption in Brahman, and liberation followed once one had fully understood the identity statements. Playing with Śa kara’s language, we may call this stage of full understanding brahm nubhava, brahm tmatv nubhava, brahm tm vagati, v ky rth nubhava, kaivaly nubhava, that is, direct experience of Brahman, of Brahman’s being the Self, of the meaning of the identity statements, or of liberation.56 Sureśvara in the NaiS calls it tat-tvam-asy- di-v ky rtha-parijñ na, full comprehension of the meaning of the identity statements such as tat tvam asi, “you are that.”57 While in the Advaita chain of soteriological causality this stage was not the last link, being followed by the destruction of ignorance and the state of remaining in the Self, it was the culmination of one’s effort, since the last two links would follow by necessity, in a logical but not temporal sequence. We can, therefore, mark such full understanding of the identity statements as the point of consummation of the brahma-vidy practice. Now, what are these identity statements? We saw under the previous heading that they were Upaniṣadic statements that affirm the identity of the individual with the supreme Self, the vijñ n tman with the param tman. More formally, they are statements that affirm the being of the category or pad rtha of “that” to the category or pad rtha of “you,” which stand for Brahman and the Self respectively.58 The precise meaning of the two categories is ascertained through examination of Upaniṣadic passages, and we will see how that proceeds shortly. The identity that is affirmed of the two categories is not analytic, as noted by Rudolph Otto. That is, 56 These are adaptations of expressions from BSBh 1.1.4, 2.1.4, 3.3.32, 4.1.2. 57 NaiS 1.52. 58 Formulations of this kind abound in the BSBh, and the correlated entities are interchangeably called tat-pad rtha, brahma, vara, parame vara, and tvaṁ-pad rtha, j va, r ra etc. See for starters BSBh 2.1.21, 2.3.46, 3.2.6, and 4.1.2-3. 402 the second is not just another name for the first, although the ultimate reference of the two categories is the same. The identity statements are, rather, synthetic, that is, conveying new information that is unknown about the correlated categories before they are identified.59 That must be so, in any case, if the statements are to be pram ṇa at all. It is not necessary, furthermore, that their form be x = y, for Śa kara mentions among them statements of the “there is no other x than y” kind, where y stands for the Self directly or through some of its cognitive functions that an agent might identify with. This new knowledge contained in the identity statements cannot be had before one fully understands what the two categories ultimately stand for. In fact, the clarification of meaning as the process that replaced meditation was clarification of meaning of the respective categories. “A sentence such as tat tvam asi cannot give rise to certain knowledge regarding its meaning in the case of those for whom the two categories are blocked by ignorance, doubt, or confusion, because the meaning of the sentence depends on understanding the meaning of the individual words.”60 We will, therefore, first focus on what entities Śa kara thought were correlated in statements such as tat tvam asi. Before we do that, however, we should mention that there does not seem to be a technical term that is consistently used to refer to such identity statements as a category, nor do we find anywhere the full scope of their class. There are several designations that we can identify in the BSBh. One is abheda-nirde a, statement of non-difference, but that wording was influenced by the sūtra itself.61 The second is the one we have seen under the previous heading, vijñ n tmanaḥ 59 Otto 1970Ś102Ś “To judge by terms and speculative efforts, “Ātman is Brahman” is an analytical statement, or rather a verdict of identity. In secret, however, it remains a verdict of synthesis.” 60 tatra yeṣām etau padārthau ajñāna-saṁśaya-viparyaya-pratibaddhau, teṣāṁ tat tvam asi ity etad vākyaṁ svārthe pramāṁ notpādayituṁ śaknoti, padārtha-jñāna-pūrvakatvād vākyārtha-jñānasyeti. BSBh 4.1.2. 61 BSBh 2.1.22, II.328: nanv abheda-nirdeśo ’pi darśitaḥ—tat tvam asi ity evaṁ-jātīyakaḥ. The sūtra runs: adhikaṁ tu bheda-nirdeśāt. 403 param tmaikatvopade a-par (pravṛttiḥ), a mode of scriptural sentences that identify the individual with the supreme Self. The most common way to refer to them seems to be tat-tvam- asy- di, “the group in which the first member is tat tvam asi,” to which various additions can be attached, such as evaṁ-j t yaka and ity-evam- di-v kya, “of such kindś” v kyaś ved nta-v kyaś straś rutiś brahm tmaikatva-vastu-pratip dana-para, “a statement that affirms the unity of Brahman and the Self.”62 From passages in the BSBh, we can provide the following examples of identity statements as a representative but not an exhaustive list.63 Such a full list cannot be found in Śa kara’s works, and whenever he lists statements of this kind, his purpose is just to illustrate. ▪ ahaṁ brahm smi, “I am Brahman,” in BĀU 1.4.10. ▪ ayam tm brahma, “This Self is Brahman,” in BĀU 2.5.19 ▪ eṣa ta tm sarv ntaraḥ, “The self within all is this self of yours,” in BĀU 3.4.1. ▪ eṣa ta tm ntary my amṛtaḥ, “This Self of yours, the inner controller, the immortal,” in BĀU 3.7.3-23. ▪ n nyo ’to ’sti draṣṭ rot mant vijñ t , “There is no other seer, hearer, thinker, knower than him,” in BĀU 3.7.23 ▪ n nyad ato ’sti draṣṭṛ, “There is no seer other than this [imperishable],” in BĀU 3.8.11. ▪ [tat satyaṁ sa tm ] tat tvam asi, “That is Being, that is the Self, that is what you are,” in ChU 6.8.7-6.14.3.64 ▪ tvaṁ v aham asmi bhagavo devate ’haṁ vai tvam asi devate, “Blessed Lord, I am surely you and you are surely me.”65 Establishing an exhaustive list seems superfluous in any case, because the principle is that a statement which identifies Brahman with the Self in a text that has nirguṇa-vidy as its scope is 62 BSBh 1.1.4, 1.2.8, 1.3.19, 1.4.6, 3.3.32ś throughout Sureśvara’s NaiS as well. 63 References in the BSBh include 1.1.4, 1.2.8, 2.3.30, 3.2.27, 3.4.8, 4.1.3. 64 Joel Brereton (1986) had argued against the interpretation of tat tvam asi as an identity statement, claiming that the gender of the subject would have to follow the gender of the predicate, and that the sentence, if meant to be an identity statement, should read saḥ tvam asi. For my purposes this intervention is irrelevant, since Advaitins have universally read tat tvam asi as an identity statement. I do not wish to address the issue at length here, but it does not seem to me that Brereton is right. The personal pronoun tvam is not gendered, and it is but right for the demonstrative tat in such a case to follow the gender of the noun it stands for, which is sat, Being. 65 In BSBh 4.1.3, III.773, introduced by tathā hi parameśvara-prakriyāyāṁ jābālā ātmatvenaiva etam upagacchanti, “To illustrate, the Jābālas, in the chapter on the Supreme Lord, present him as the Self.” I have not been able to identify the reference. 404 “of that kind” as tat tvam asi. It is rather important to note that such a statement must be part of a nirguṇa-vidy : a glaring absence from this list is the statement from the Kauṣ taki, yas tvam asi so ‘ham asmi, that mister X addresses to Brahman on the couch in brahma-loka. One can hardly imagine a more explicit and fuller identity statement, and Śa kara’s aversion to this old Upaniṣad has been well noted by Signe Cohen.66 Śa kara, nevertheless, focused most of his attention on interpreting tat tvam asi and ahaṁ brahm smi, and there seems to be a good reason for that practice. Under BS 4.1.3, the sūtra that established meditation on Brahman as one’s Self because “they understand and teach it like that,”67 he divides the identity statements in two categories, related to the notions of understanding and teaching, respectively. Ahaṁ brahm smi is in the first group, while tat tvam asi in the second, providing thus neat blueprints for the two perspectives in the processes of brahma-vidy , that of the student and that of the teacher. Historically, however, it was another identity statement, one contained in the Taittir ya Upaniṣad, that anchored the text based on which the very notion of identity statement culminating in mah -v kya was developed: it was tasm d v etasm d tmanaḥ, “from this very [Brahman] which is the Self [creation proceeds].”68 Śa kara’s commentary on the second chapter of the Taittir ya was the primary source for Sarvajñātman’s Pañca-Prakriy , and therefore I will base my account there. I will also draw on BSBh 4.1.2 and 2.1.14-22, as well as on US. The Problem of Language The rejection of meditation as the means of knowing Brahman meant that Brahman could be known solely through scripture, that is, through language. To put this more technically, the 66 Cohen 2008:139-147. 67 ātmeti tūpagacchanti grāhayanti ca. 68 TU 2.1.1. 405 Upaniṣads were the only pram ṇa that could give rise to the cognition of Brahman. This pram ṇa could use aids, but not that of meditation. This was a tough proposition to defend, for reasons we have discussed in depth. Let us very briefly review them. The root problem with scripture was that it was supposed to be informative about supersensible things, but it had to do so by using categories that are known to us. Mīmāṁsakas claimed that the Veda was not an authority on categories, word meanings, but on unknown relations. By using known categories, the Veda could inform that there was such a relation between some future result and a ritual action of a kind. The Prābhākaran take on meditation was an uncompromising application of this doctrine: the only new information that the injunctions of meditation offered about the Self concerned its status of a subsidiary to the action of meditation. The Self was known perceptually, through the process of recognition. The pūrvapakṣa in BSBh 1.1.1 was along these lines: is Brahman a known or an unknown category? If it was known, then there would be no reason to inquire into it. If it was unknown, then it could not be inquired into, the unstated assumption being: because it was an existent, not a relation.69 Ma ḍana Miśra, on the other hand, argued that the Veda could be informative about Brahman if it was possible for Brahman to be a pad rtha, a category that is known to us and for which individual words can stand. Ultimately all words stood for Brahman, and we could start with the idea of Brahman as that great Being and the cause of everything and obtain a definite description from the Upaniṣads, a proper informative v ky rtha or sentential meaning that would reveal Brahman specifically, not just through the facts of its being the cause. Such v ky rtha in which the meaning was not a universal but a particular was achieved by relating attributes to a substantive, technically called vi eṣya-vi eṣaṇa-bh va. There was no reason for the Veda not to 69 tat punar brahma prasiddham aprasiddhaṁ vā syātś yadi prasiddhaṁ na jijñāsitavyam. athāprasiddhaṁ naiva śakyaṁ jijñāsitum iti. BSBh 1.1.1, I.8. 406 be able to tell more about something, so long that thing was somehow known to us. The problem with this was that it presented Brahman as a relational entity that is known only mediately, through description, and Ma ḍana’s solution wasŚ meditation. In terms of scriptural theology, the problem of language and conceptional thinking in relation to Brahman was expressed in the Upaniṣads as well. The most prominent textual locus was the Taittir ya 2.4.1: Whence words return along with the mind, not attaining it, he who knows that bliss of Brahman fears not at any time.70 Śa kara addresses this text quite consistently when talking about Brahman and language. Another important Upaniṣadic statement was Kena 1.3: There the eye goes not, speech goes not, nor the mind; we know not, we understand not how one can teach this.71 So, how could the Upaniṣads be the sole pram ṇa on Brahman, if language and conceptual thinking in which speech was couched had so much stacked against them? Before we tackle this, we need to address a couple of points concerning Śa kara’s take on a few categories of language that were important for him and were common in his intellectual universe. The first is related to the substantive-attribute or noun-adjective relation, vi eṣya-vi eṣaṇa-bh va, that I mentioned above. This relation was the way Sanskrit grammarians explained how individual words that otherwise stand for universals express a particular through combining. This is an involved topic that we not need consider in detail, and a few notes will suffice. In sentences and phrases such as kṛṣṇaḥ tilaḥ, “black sesame,” or n lam utpalam, “a blue lotus,” as well as in descriptive or coordinative compounds, karma-dh raya, such as n lotpalam, there obtains a relation between the two words with different meanings so that they both denote a particular, a 70 Translation Radhakrishnan 1992:545. 71 Translation Radhakrishnan 1992:582. 407 specific variety of lotus: not a lotus in general or a blue something in general, but specifically a blue lotus. To paraphrase Bhart hari, in kṛṣṇaḥ tilaḥ, the word “black” is used in the sense of some black substance whose genus is unknown, whereas the word “sesame” is used in the sense of a genus whose quality is unknown. Since their generalities do not relate, they first mutually specify their meaningŚ in the sentence, “black” gets to stand for the black color specified by the being of a sesame seed, and “sesame” stands for the being of a sesame seed specified by the black color. Their word denotations have changed, and now the two words are relatable. Finally, these two transformed denotations merge in one, and the whole phrase gets to denote a particular black sesame seed. The last was what Mīmāṁsaks call a v ky rtha, a sentential denotation.72 Ma ḍana’s definite description of Brahman was very much like that. Sanskrit grammarians identified this noun-adjective relation with the rubric of s m n dhikaraṇya or co-referentiality, the phenomenon where two words are coordinated both syntactically and semantically, that is, are in the same case and stand for the same reference. This was, further, part of the rubric of vṛtti, or complex formations. Śa kara, on the other hand, had a wider understanding of such co-referentiality, and for Vedāntins in general the idea of s m n dhikaraṇya had a much greater ontological, not merely linguistic, significance. It was important to understand, first, how Brahman is coordinated with its characteristics, such as in Ma ḍana’s definite description, and, second, how it is coordinated with other ontological reals, such as in the identity statements and the identifications of Brahman with the world. Thus, in addition to the problem of how speech in general can relate the supersensible Brahman, it was also important to figure out what kinds of ontological relations the descriptions of Brahman 72 Sastri 1964. See also Bhart hari’s V kyapad ya, Book 3, Vṛtti-Samude a, and Sharma’s translation and commentary on the Aṣṭ dhy yi 1.2.42 and 2.1.49 (Sharma 2000, 2002). 408 disclosed. Śa kara, thus, classed co-referentiality in four kinds, and his taxonomy is primarily ontological and only secondarily linguistic.73 One kind was the vi eṣya-vi eṣaṇa-bh va that we have seen, and it is common for Śa kara to give the classical “blue lotus” example as its illustration. In a sense, co-referentiality was synonymous with this relation, as any apposition would have had the form of one word playing the role of a qualifier to another substantive. However, there were several ways in which this could happen, and the noun-adjective relation was just one of them. A second type was the quite uninformative case of full synonymy, ekatvam, where two words have the same meaning and the same primary mode of signification, such as “Brahmin,” “best of the twice-born,” and “a god on Earth.”74 A third type was the case of superimposition, adhy sa, where one thing could be described as or identified with another through deliberate fancy; that is, when two things are identified but we know that they are different. This was the bread and butter of the symbolic meditations in the Upaniṣads, for instance the identification of name with Brahman, and more generally of all symbol-based modes of adoration, such as the worship of Viṣ u as the image or as the lagrama stone. Superimposition, of course, did not have to be deliberate, and we have seen previously that it was the very nature of metaphysical ignorance. Superimposition of the notion of one thing over another could have happened naturally, as in the case of the identification of the Self with the body or the cognitive faculties. A statement of co-referentiality of two categories could, thus, intend to negate such superimposition through pointing to an incommensurability. Śa kara called 73 The taxonomy is given in BSBh 3.3.8. 74 Uninformative, that is, for Brahmin grammarians and theologians, not for social historians. 409 this kind of co-referentiality apav da, negation.75 Advaitins commonly illustrate this with the right knowledge that has replaced the cognitive error of confusion in the statement “The man is a post,” and meaning, what has previously been wrongly seen as a man is now ascertained to be a post. Sureśvara called this form of co-referentiality a defeated-defeater relation, b dhya- b dhaka-bh va, and the term of art in later Advaita Vedānta became b dh y ṁ s m n dhikaraṇya, co-referentiality in the sense of a conflict.76 The second important idea is that of vṛtti, which is sometimes translated as “uses of words.”77 Kunjunni Raja calls them “the function of a word in its relation to the sense,”78 presumably having in mind Frege’s famous distinction between a sense and reference of expressionsŚ the phrases “morning star” and “evening star” or the names “Phosphorus” and “Hesperus” have different senses, but they both refer to the same object, the planet Venus.79 Thus, words have their sense, but they also have functions through which they do not have to stand for or refer to their sense. Deshpande’s term for vṛtti is “signification function,” and since vṛtti is commonly described as potency or akti, he defines this signification function as “a power or a special ability of a word to function in a certain capacity as a signifier of a certain meaning.”80 I find Deshpande’s formulation preferable. It does not seem to me that the sense- reference distinction has been posited in Indian philosophy of language quite as Frege did: the question that is front and center is, how can a word that otherwise stands for one thing get to stand for another. 75 Apavāda was otherwise a common technical term in Sanskrit grammar, standing for a special rule that is an exception to another, general rule, and which in effect cancels the operation of the general rule. See Abhyankar 1986:33. 76 NaiS 2.55-6, and otherwise throughout the treatise. 77 Such is the practice, for instance, by Kocmarek in his translation of Sarvajñātman’s Pañca-Prakriy (1985). 78 Kunjunni Raja 1977:24. 79 Frege 2007. 80 Deshpande 2007:53. 410  Śa kara’s understanding of vṛtti was shared with Bhā a Mīmāṁsā, and there seems to be little original that he brings to the idea.81 Words have their primary or “natural” signification, abhidh , in which they stand for universals. The signified in such a case is called v cya, directly expressed. In sentences, it is possible for this signification function to be blocked, in which case the word would exercise a secondary function, which is figurative and can neutrally be called aupac rika or bh kta.82 More commonly, this secondary signification function is called lakṣaṇ (and note well the long final vowel). Words can exercise such secondary signification function only in sentences, when the literal meaning is impossible in the context. Śabara, in fact, argued that words in sentences must exercise their secondary signification function, because in sentences they refer to class-members, not to universal class-properties.83 Śa kara most certainly accepted this line of reasoning and classified the secondary signification into one approximating to and another departing from the primary meaning. In the first case, the words signify class members, parts of the classes they signify in isolation, while in the second they signify something that is not included in their scope but is directly related to it, through metonymy, synecdoche etc.84 A classic example is “a village on the Ganges,” where Ganges does not stand for the river, but for the bank, which is, nevertheless, directly related to the river. The signified in such a case is called lakṣya, indirectly expressed or indicated. A second figurative signification function is one that proceeds specifically through metaphor, or qualitative similarity of the figurative denotation of the word to the natural 81 For a most lucid account of the Bhā a understanding of vṛtti, see McCrea 2008:91-7. 82 Instances of both are common in the BSBh: 1.1.6, 1.3.28, 1.4.9, 1.4.19, 2.3.5, 2.3.16, 2.3.29, 2.4.4, 2.4.20, 3.1.4, 3.1.7, 3.2.4, 3.3.24. 83 McCrea 2008:96. 84 lakṣa āyām api tu sannikarṣa-viprakarṣau bhavata evaś adhyāsa-pakṣe hi arthāntara-buddhir arthāntare nikṣipyata iti viprak ṣ ā lakṣa ā, viśeṣa a-pakṣe tu avayavi-vacanena śabdena avayavaḥ samarpyata iti sannik ṣ ā. BSBh 3.3.9, III.630-31. 411 meaning. The classic example of this which Śa kara also employs is “the man is a lion,” where the word that stands for lion as a class is first figuratively identified with valor as one of the leonine qualities, and then applied to an individual man in virtue of his possession of valor.85 This may be called a tertiary signification function insofar as it is twice remote from the primary meaningś that is, “lion” cannot be directly related to “man,” but only through the medium of valor. This is commonly called gauṇa-vṛtti, but the signified is also lakṣya. We will say more about signification in the last chapter, when we move to Sarvajñātman. The last thing we need to point out is that early Advaitins, and certainly Śa kara, Sureśvara and Sarvajñātman, endorsed the Śabara-Kumārila’s abhihit nvaya-v da theory of sentence meaning. To rehearse this very quickly, Bhā a Mīmāṁsakas thought that words in sentences first denote their word meanings, then cease their individual functions, combine, and denote the sentential reference. Such was also the order of understanding on the part of the competent user of a language: ascertain which word is said, recollect its meaning, and combine it to understand which individual existent the sentence intends to affirm. Not only did Advaitins endorse the theory, it was crucial to their soteriological project. Understanding the identity statements was predicated on properly understanding the categories of tat and tvam first.86 Word combination, in other words, was the process of liberation. What Advaitins did not accept, however, was that the meaning which the individual words produced in combination had to be a sentential reference, a v ky rtha. Other possibilities were open, and we will see what they were. 85 Śa kara refers to this metaphor in BSBh 1.1.4 and 3.1.6. 86 Karl Potter is, therefore, not correct when he claims in the first volume on Nyāya-Vaiśeṣika of his Encyclopedia of Indian Philosophies (1977Ś151) that Vedāntins espoused anvit bhidh na-v da along with Prābhākaras. His assessment is better in the volume on Advaita Vedānta (1981). 412  The Categories of “That” and “You” and the Notions of Brahman and the Self The problem of how language may express the supersensible Brahman did not concern the word brahma itself, or several other words that Śa kara would have considered full synonyms, such as sat or Being, kṣitṛ or Brahman that visualizes the world to set it in motion, or vara, God, all standing for Brahman’s causal function in regard creation.87 It is common for Śa kara to say that these words denote Brahman in their primary signification function. This also implied that Brahman was not quite unknown: the word had its direct meaning, derived from the root bṛṁh and standing for that which makes things grow, the cause of the world in the most general sense.88 In any case, Brahman had to be a known thing, linguistically expressible, for language to be meaningful, because words cannot convey that which is fully beyond experience. Brahman as such a cause of the world, that great ground of Being which makes things grow, was also the topic of an Upaniṣadic passage that posits its being, the textual basis of the Brahma-Sūtra definition of Brahman as the source of creation, sustenance and destruction of the world, Taittir ya 3.1.1Ś “That from which these beings are born, on which, once born, they live, and into which they return—know that distinctly: it is Brahman.” The problem with the denotation of words like brahma and with its corresponding description was that they were incomplete, “with a remainder” that still needs to be stated. They present Brahman in a vague, insufficient, imprecise way.89 They throw Brahman in and posit its relation to creation, but they require more. 87 tat-padena ca prak taṁ sat brahma īkṣit jagato janmādi-kāra am abhidhīyate. BSBh 4.1.2, III.769. 88 brahma-śabdasya hi vyutpādyamānasya nitya-śuddhatvād ayo ’rthāḥ pratīyante, b ṁhater dhātor arthānugamāt, BSBh 1.1.1, I.8. brahmavit brahmeti vakṣyamā a-lakṣa am, b hattamatvāt brahma, TUBh 2.1.1, VI.59. 89 annādi-brahma aḥ pratipattau dvāraṁ lakṣa aṁ ca yato v im ni ity-ādy uktavān. sāvaśeṣaṁ hi tat, sākṣād- brahma o ‘nirdeśāt. TUBh 3.1.1, VI.127. 413  So, while Brahman is directly signified by words such as brahma, vara, sat, they do not say much about its essential, peculiar nature. It may well be that the pradh na of Sā khya is denoted and intended in such passages. The specific nature of Brahman must be stated, and this nature comes in three sets of qualities presented in Upaniṣadic texts. The first are positive attributes from texts such as satyaṁ jñ nam anantaṁ brahma and vijñ nam nandaṁ brahma, TU 2.1.1 and BĀU 3.9.28.7, and Śa kara says that they present Brahman as the light of consciousness. The second and the third are negative attributes that deny change in Brahman on the one hand, and possession of the attributes of its gross products on the other.90 They are presented in texts such as adṛṣṭaṁ draṣṭṛ, avijñ taṁ vijñ tṛ, ajam ajaram amaram, asthūlam anaṇu, ahrasvam ad rgham, from Yājñavalkya’s teachings to Gārgī in BĀU 3.8.91 The predication of such qualities to Brahman obtains through the mode of co-referentiality, the substantive-attribute relation, and it is very much like the construction of the notion of Brahman the uniform meditational counterpart in the BS brahma-vidy s. These collectively are evidently the texts that ascertain the nature of the supreme Self from our classification.92 Now, such combination of words to obtain definite knowledge, not the mere positing of the general category, causes the apparent problem of presenting Brahman as a sentential reference, a v ky rtha, a denoted referent as something of a kind, a member of a class that is specified by the qualifying properties. Think of n lotpalam, the blue lotus, the flower in my garden belonging to the class of lotuses and made particular through the attribution of blue color 90 tatra ajādi-śabdair janmādayo bhāva-vikārā nivartitāḥ. asthūlādi-śabdaiś ca sthaulyādayo dravya-dharmāḥ. vijñānādi-śabdaiś ca caitanya-prakāśātmakatvam uktam. BSBh 4.1.2, III.770. 91 Adṛṣṭaṁ draṣṭṛ and avijñ taṁ vijñ tṛ are from 3.8.11, asthūlam anaṇu, ahrasvam ad rgham from 3.8.8. Ajam ajaram amaram seems to be based on the Madhyandina recension, in ŚB 14.6.8, but it does not correspond fully as ajam is not there. 92 Later Advaita Vedānta classified the two kinds of characteristics of Brahman, its general causality in TU 3.1.1 and its distinguishing characteristics, in taṭastha-lakṣaṇa and svarūpa-lakṣaṇa, non-essential and essential characteristics. See, for instance, in Deutsch 1969. 414 and other qualities. This could not work in the case of Brahman, because Brahman is one-off, it does not have a class from which it could be delimited through attribution. Thus, Brahman the pad rtha is too general, uninformative, and the substantive-attribute relation seems impossible because Brahman cannot have a class. Śa kara’s solution to this riddle was to propose that the vi eṣya-vi eṣaṇa-bh va does not need to be restricted to the substantive-attribute type that produces a sentential reference through definite description. It was possible to interpret it as a definiendum-definiens relation, a lakṣya-lakṣaṇa-bh va. How are the two different? Well, the second is a subtype of the first, but while the first delimits a member from its class, the second delimits a something from everything else. – But, a substantive is distinguished by shunning other (possible) attributes, as in the case of blue and red lotus. When there are several substances that belong to the same class and can have one attribute, only then is an attribute meaningful, not in regard to a unique thing, because there is no possibility of an alternative attribute. Just like the yonder sun is one, likewise is Brahman. There are no other Brahmans from which the one is to be singled out, as in the case of the blue lotus. – There is not that fault, since the attributes are for definition, not qualification. – And what is the difference between definiens-definiendum relation on the one hand and the qualifier-qualificandum relation on the other? – The qualifiers distinguish a substance from things of its class, while a definiens from everything, as in “space is that which provides room.”93 Compare the following two statements: ▪ ▪ The lotus is blue, large, and fragrant; Space is that which gives room. The first is a definite description obtained through attribution of color and other qualities to a class, while the second is a definition. The distinction is eminently clear in English because of 93 nanu, viśeṣyaṁ viśeṣa āntaraṁ vyabhicarad viśeṣyate, yathā nīlaṁ raktaṁ cotpalam itiś yadā hy anekāni dravyā i eka-jātīyāny eka-viśeṣa a-yogīni ca, tadā viśeṣa asyārthavattvamś na hy ekasminn eva vastuni, viśeṣa āntarāyogātś yathā asāv eka āditya iti, tathā ekam eva brahma, na brahmāntarā i, yebhyo viśeṣyeta nīlotpala-vat. na; lakṣa ārthatvād viśeṣa ānām. nāyaṁ doṣaḥ. kasmāt? lakṣa ārtha-pradhānāni viśeṣa āni, na viśeṣa a-pradhānāny eva. kaḥ punar lakṣa a-lakṣyor viśeṣa a-viśeṣyayor vā viśeṣaḥ? ucyate. sajātīyebhya eva nivartakāni viśeṣa āni viśeṣyasyaś lakṣa aṁ tu sarvata eva, yathā avakāśa-pradātr ākāśam iti. TUBh 2.1.1, VI.61-2. 415 the definite article, but it is not so in Sanskrit. Śa kara claimed that the predication of characteristics to Brahman in the Upaniṣads is an instance of a definiendum-definiens relation. In claiming this, which amounts to a crucial distinction from the account of Ma ḍana and other prasaṅkhy na-v dins in which Brahman ended being a sentential reference, Śa kara clearly took a cue from the tradition of Nyāya. Vātsyāyana, the author of the Ny ya-Bh ṣya, said that the method of Nyāya as a discipline had three parts. The first, called udde a, is the mere positing of a category by way of stating its name. The second is definition, lakṣaṇa, or stating the property of the posited category that differentiates it as a thing. The last is parikṣ , examination by means of reliable warrants whether the stated definition is applicable to the defined thing or not.94 Precisely the first two were happening at the beginning of the second chapter of the Taittir ya, where Brahman is posited as something to be known in the statement brahmavid pnoti param, “The knower of Brahman attains the highest,” but its specific nature is not stated. The immediately following statement satyaṁ jñ nam anantaṁ brahma is a definition of the posited category that distinguishes its nature from everything else by the attribution of the three characteristics of Being, consciousness, limitless. In doing so, the three characteristics that serve as definiens of Brahman the definiendum are predicated of Brahman serially, not simultaneously, which should further prevent the definition from being confused for a definite description. In a definite description, all attributes that are predicated of the substance need to be understood simultaneously for the sentential reference to obtain. Consider the expression, “a blue hardcover book of five hundred pages.” For this to be a single expression, it must be syntactically tight. There is no such requirement in 94 trividhā cāsya śāstrasya prav ttiḥ, uddeśo lakṣa aṁ parīkṣā ceti. tatra nāmadheyena padārtha-mātrasyābhidhānam uddeśaḥ. tatroddiṣ asya tattva-vyavacchedako dharmo lakṣa am. lakṣitasya yathā-lakṣa am upapadyate na veti pramā air avadhāra aṁ parīkṣā. Ny ya-Bh ṣya on 1.3. 416 satyaṁ jñ nam anantaṁ brahma. Śa kara says that these are, really, three separate statements, satyaṁ brahma, jñ naṁ brahma, anantaṁ brahma, that progressively refine the definition of Brahman. Brahman is, then, not a sentential reference, but is identified with individual words that stand for three separate categories: in effect, it is not identified with attributes at all, but with substantives that share the denotation with it. Brahman is not real or conscious: it is Being and consciousness.95 The three do eventually restrict one another, but there is no requirement that they do so simultaneously: the serial predication amounts to a gradual refinement of the notion of Brahman. Whereas for the unique lotus a full definite description is required, Brahman is already single. The profit from this mode of predication is obvious: Brahman becomes known in its specific nature, not just as a pad rtha, but is no longer a relational entity, a particular obtained by a collocation of a noun and adjectives, a sentential reference. Śa kara hoped in this way to avert what he saw as the pitfall of the prasaṅkhy na formulation of the notion of Brahman: it turned out a saṁsarg tmika entity, a verbal composite like the funny bird that we named Jewelfowl. Let us very briefly go through the three parts of the definiens and see what they contribute to defining Brahman in its specific nature. Satyam – that is true which does not alter its nature which is fixed by itself. If a thing alters its self-constituted nature (i.e., becomes what it is not), it is unreal. Therefore, a transformation is unrealś “The transformation is a verbal handle, as being it is only clayś” only being is real. Therefore, saying “Brahman is Being,” [the Upaniṣad] guards Brahman from transformation. From this it follows that Brahman is the cause.96 95 satyādi-śabdā na parasparaṁ sambadhyante, parārthatvātś viśeṣyārthā hi te. ata eva ekaiko viśeṣa a-śabdaḥ parasparaṁ nirapekṣo brahma-śabdena sambadhyate – satyaṁ brahma jñānaṁ brahma anantaṁ brahmeti. TUBh 2.1.1, VI.62. 96 satyam iti yad-rūpe a yan niścitaṁ tad rūpaṁ na vyabhicarati, tat satyam. yad-rūpe a yan niścitaṁ tad rūpaṁ vyabhicarati, tad an tam ity ucyate. ato vikāro 'n tam, v c rambhaṇaṁ vik ro n madheyaṁ mṛttiketyeva satyam [ChU 6.1.4], evaṁ sad eva satyam ity avadhāra āt. ataḥ satyaṁ brahma iti brahma vikārān nivartayati. ataḥ kāra atvaṁ prāptaṁ brahma aḥ. Ibid. 417 Now, we already knew that Brahman was the cause through the etymology of the word, “it is that which makes things grow,” so there must be more to the statement that satyam tells us how Brahman is the cause. The operative idea here is that Being is changeless, like clay that persists being clay throughout its different shapes. Our concept of Brahman is refined through understanding that Brahman as Being is changeless, but in effect that presents Brahman as the material cause of the world, up d na- k raṇa, and that brings two dangers. First, a material cause or the stuff that things are made of is, in our experience, an insentient thing, like clay the material cause of pots, pitchers, and the like. Second, an insentient cause requires a sentient agent, a separate nimitta-k raṇa or an efficient cause. The second attribution, that of jñ na or knowledge, prevents these two from obtaining. “Since a cause, being a thing, requires causal factors, as in the case of clay, its being unconscious may obtain. Therefore, it is said, Brahman is knowledge.”97 The further attribution of ananta, infinite, along with the prior fact of Brahman’s being satyam the singular Being, prevent jñ nam to be taken in the sense of a cognitive agent: Consciousness is knowing, awareness. The word “consciousness” refers to the verbal action, not the agent of knowledge, because consciousness is a qualification of Brahman along with truth and infinite. The two are impossible when there is an agent of knowing. How can that which is transformed as being the cognitive agent be Being and infinite? For, that is infinite which is not cut off from anything. If it is the agent of knowledge, it is separate from the knowable and from knowledge—how could it be infinite?98 First, that Brahman is changeless Being and limitless, satyam anantam, jointly prevent the common distinction of agent, object, and action of knowing to obtain. The agent-object 97 kāra asya ca kārakatvam, vastutvāt, m dvat, acid-rūpatā ca prāptāś ata idam ucyate—jñānaṁ brahmeti. TUBh 2.1.1, VI.62-3. 98 jñānaṁ jñaptiḥ avabodhaḥ, — bhava-sādhano jñāna-śabdaḥ — na tu jñāna-kart , brahma-viśeṣa atvāt satyānantābhyāṁ saha. na hi satyatā anantatā ca jñāna-kart tve saty upapadyete. jñāna-kart tvena hi vikriyamā aṁ kathaṁ satyaṁ bhavet, anantaṁ ca? yad dhi na kutaścit pravibhajyate, tad anantam. jñāna-kart tve ca jñeya- jñānābhyāṁ pravibhaktam ity anantatā na syāt. Ibid. 418 distinction presupposes limits, and is otherwise known: scripture would not be informative if it were to uphold it. Brahman is just knowing that does not involve agent and object. Second, the attribution of ananta prevents this knowledge that does not involve the agent-object distinction to be limited, as knowledge otherwise is. Brahman thus becomes omniscient, and we have arrived at Śa kara’s common definition of Brahman/Īśvara as the omnipotent and omniscient cause of the world. I should like to note that this is a very curious form of omniscience that a captious mind may be tempted to call omni-ignorance, were it not clear that in Śa kara’s eyes it is not an epistemological category at all. Now, in the passage that I have been following here, TUBh 2.1.1, there seems to be a shift in the argument when Śa kara points to the identification of this Brahman with the Self, which the Upaniṣad does by the typical anaphoric-cataphoric use of the pronouns, tasm d v etasm d tmanaḥ, which is followed by an account of creation. This is no longer just the “tat- pad rtha” context, and that is also evident from Sarvajñātman’s Pañca-Prakriy which we will consider in Chapter Ten: it is the tat = tvam context. I will come back to this passage later. For now, we should point out that the definition of Brahman as satyaṁ jñ nam anantam relates through the individual categories to other Upaniṣadic passages in the texts about the causal Brahman that present Brahman as the sole Self in our classification. Śa kara clearly identifies satyam with sat from the 6th chapter of the Ch ndogya, and that text is used to bring home the point that Brahman in its causal role does not undergo transformation, but just is that great thing out there, coordinated with everything. Elsewhere (BSBh 2.1.14), as we saw in the previous heading, he goes at great length to show how the creation passages are illustrations of how the world of multiplicity does not proceed from Brahman, but is Brahman; in other words, how Brahman the cause is changeless. Any textual locus that discusses satyam/sat can be used to 419 elaborate on Brahman’s nature as the unchanging Being from which creation proceeds, and these include, for example, the 1st of Aittareya and the 2nd of Taittir ya itself. Brahman’s feature of jñ nam anantam, on the other hand, is elaborated in passages such as the section of the bhūma- vidy where the Upaniṣad says, “where one does not see, hear or cognize another, that is plenitude.”99 Thus, the texts about Brahman as the sole Self, tmaikatva-pratip dana-paraṁ vacana- j tam, seem to be individually related to the categories that form the definiens of Brahman as their elaborations, and on their part, use the descriptions of creation to teach what kind of Being and consciousness Brahman is. The three are in a close synergy. We can begin to form some idea of hierarchy of Upaniṣadic passages on the side of the tat-pad rtha: ▪ brahma-vid pnoti param posits Brahman as that thing which should be known; it is known in general through the etymology, and its incomplete characterization is given in yato v im ni bhūt ni j yante, yena j t ni j vanti, yat prayanty abhisaṁvi anti, tad vijijñ sasva, tad brahma; ▪ satyaṁ jñ nam anantaṁ brahma defines this Brahman essentially; ▪ satyaṁ brahma is elaborated in texts such as sad eva somyedam agra s d ekam ev dvit yam of ChU, and tm v idam eka ev gra s t of AiU, as well as in the TU 2; ▪ the descriptions of creation in these texts are used to teach Brahman as satyam; jñ nam anantaṁ brahma is elaborated in yatra n nyat pa yati of the ChU. The next part of the definition of Brahman is the text vijñ nam nandam from BĀU 3.9.28.7, presenting bliss as the fourth characteristic that defines Brahman’s nature. The interpretation given of knowledge applies to bliss as well: Brahman is not an object of experiential happiness, because that would bring about the same basic cognitive duality of action, agent, and patient.100 This attribute also calls for an elaboration passage, but Śa kara’s immediate choice is not to point to the Taittir ya Brahm nanda-valli chapter that an Upaniṣadic scholar would immediately think of. The Taittir ya, all-important for the history of Vedānta as it 99 ChU 7.24.1. 100 BĀUBh 3.9.28.7. 420 gave the paradigmatic injunction of brahma-vidy , brahma-vid pnoti param, and the satyaṁ jñ nam anantam brahma definition of Brahman, was not a text without challenges for Śa kara, because the injunction and the definition were followed by a peculiar statement of resultŚ “A man who knows [Brahman as Being, knowledge, infinite] hidden in the cavity, the highest heaven, attains all desires together with the wise Brahman.” This was the paradigmatic statement of the nature of liberation in the BS, but from the perspective of para-vidy such passages for Śa kara were statements of praise, arthav da or Vedic propaganda meant to make liberation appealing.101 For Śa kara, the normative description of nanda was BĀU 4.3.32, which is a part of a section that shared something important with the 2nd of Taittir ya. Both texts contain descriptions that attempt to make a stratification of bliss, which goes roughly like this: the highest possible human pleasure is the basic measure of bliss. That highest human bliss times one hundred equals the bliss in the world of the Gandharvas, and thus progressively to the bliss of Prajāpati, which is one hundredth of the bliss of Brahman. While the Taittir ya culminates in claiming that this bliss of Brahman is beyond thought and words, evidently because there is no further standard of comparison, the BĀU proceeds to say that this is “the highest goal, the highest attainment, the highest world, the highest bliss,” but of the kind that one cannot experience because it pertains to the state of sampras da, deep calm characteristic of deep sleep, where Brahman is the sole entity and no cognition can obtain. Bliss is what Brahman is, thick, solid bliss, the ground that makes possible the experience of any form of pleasure, but which is itself not experienced.102 We can, now, appreciate Śa kara’s statement that this positive definition of Brahman presents Brahman as the light of consciousness, and as nothing more. It is, however, a bit more difficult to appreciate what new information could provide the other kinds of Upaniṣadic 101 Ibid. 102 TU 2.5, BĀUBh 3.9.28.7 421 statements about Brahman, those that deny change in Brahman on the one hand, and possession of the attributes of its gross products on the other. We saw their important hermeneutic role with regard to the statements about the causal Brahman and the descriptions of creations: Brahman cannot really transform into the creation because it is a permanently changeless entity. As for the definition of Brahman, however, they don’t seem to contribute much, since Śa kara’s positive characteristics were so “positive” that there was nothing left to deny. Or rather, there was, as we shall soon see, but only at the level of the identity statement, not on the tat-pad rtha side. Let us see what we have achieved through the definition of the tat-pad rtha. The indeterminate notion of that thing which makes everything grow that we are empirically acquainted through understanding causality and growth with has become essentially defined and known in its specific nature through identification with individual categories. The procedure made it possible to avoid presenting Brahman as a relational entity, which was one of the two major reasons why prasaṅkhy na-v dins wanted Upaniṣadic knowledge to be followed by meditation. We did not quite avoid the problem of supersensible Brahman and language. Brahman was still defined by words, and we have not seen any reason why such attribution could not happen through the direct signification function: we will have to come back to this problem. But for now, Brahman was not expressible in sentences in the manner that class members were, and this is an important point to absorbŚ Śa kara found a way out where prasaṅkhy na-v dins could not: the mode of predication was through lakṣaṇa. This was still a type of the vi eṣya- vi eṣaṇa-bh va, general co-referentiality through predication, but the prediction was definitional, not descriptive. The second category in the identity statements was the tvaṁ-pad rtha, the category of “You.” Here is Śa kara’s definitionŚ “The category of ‘You’ is the inner Self, the listener, 422 regarded as the inner Self starting with the body and ascertained as culminating in consciousness.”103 The definition is a bit cryptic, but it is clear nevertheless that the category refers to a student to whom the instruction in the identity statement is repeated, and who can potentially identify with anything that one may consider “the Self,” beginning with the body and culminating in pure consciousness. I will say much more about this in the context of the identity statement, but now let us note that just like the characteristics of Brahman that are elaborated in separate Upaniṣadic passages, the category of tvam is also taken up for deliberation in the Upaniṣads, such that its scope is gradually restricted so that it can become possible to present it as the pure light of consciousness. In the Taittir ya chapter that defined Brahman, this is done through the famous teaching of the five sheaths of the Self, pañca-ko a, namely man as the Self that is made of food (anna-rasamaya), life-breath (pr ṇamaya), mind (manomaya), cognition (vijñ namaya), and bliss ( nandamaya). In Śa kara’s reading, which is quite uncontroversial except for the fifth sheath, these are the physical human body consisting of the different limbs; the vital Self consisting of the various forms of pr ṇa; the mental body that is constituted by the Vedic mantras that one can recite internally; the cognitive Self that forms correct ideas from the Vedas as its pram ṇic field so that it could perform proper sacrifice; and the blissful Self, that is, the Self that enjoys “bliss,” the results of ritual and meditation. The last two obviously stand for the categories of kartṛ and bhoktṛ as specifically Vedic categories, the ritual agent and the one to whom the results of Vedic acts accrue.104 We should bear in mind that agency was the symptom of ignorance, the root of transmigration, and transmigration was suffering. Although I haven’t 103 tathā tvaṁ-padārtho ’pi pratyag-ātmā śrotā dehād ārabhya pratyag-ātmatayā sambhāvyamānaḥ caitanya- paryantatvenāvadhāritaḥ. BSBh 4.1.2, III.770. 104 See TUBh 2.1.1 through 2.5.1. 423 seen a statement directly spelling this out, the blissful Self that is in closest proximity to Brahman as bliss thick and solid is, really, the miserable Self, the transmigrating Self that thinks itself an agent. Sarvajñātman will explicitly identify nandaṁ brahma with nirduḥkha, absence of suffering that we need to understand in the wider sense of transmigration. The pañca-ko a, thus, are sheaths enveloping the Self with all the possible points of identification in the natural and the Vedic sphere, and Śa kara says that the Upaniṣad uses this teaching to unveil Brahman as the innermost Self, one’s own Self, by progressively removing the sheaths that are created by ignorance, in the manner of the gradual winnowing of grain of the kodrava species that has several layers of husk.105 The Identity Statement Context I said that it is was not easy to see what new information the negative characteristics add to the notion of Brahman in Śa kara’s account. It becomes, however, abundantly clear when we look at the third chapter of the verse portion of the US that these characteristics become meaningful on the assumption that Brahman is the Self, not before that. The reasoning in this four-verse composition is as follows. The question that is under discussion in, how should one understand the identity statement “I am he,” where “he” stands for Īśvara or Brahman? Two options are availableŚ Īśvara can be understood either as one’s own Self, such that this would be a full identity statement, meaning that I am literally Brahman, or it can be understood as something which is not the Self. Ānandagiri’s example clarifies what the second really meansŚ does “I am he” mean something like “I am Meru,” the great mountain and the greatest thing out there one 105 annamayādibhya ānandamayāntebhya ātmabhyaḥ abhyantaratamaṁ brahma vidyayā pratyag-ātmatvena didarśayiṣu śāstram, avidyā-k ta-pañca-kośāpanayanena aneka-tuṣa-kodrava-vituṣī-kara eneva ta ḍulān prastauti. TUBh 2.2.1, VI.75. 424 could possibly conceive of? If that were the case, the identity statement would intend to affirm assimilative meditation. The key argument for the first possibility is that characteristics such as “not gross” and “not fine” are predicated of Brahman, and if they are predicated to some external Meru-like thing, then such a thing would be unknowable—what could possibly such a “not gross, not fine” thing be—and the worst theological nightmare would follow: a scriptural statement would be either meaningless, or would intend to affirm the Buddhist emptiness. If, however, Brahman is the Self, myself, then such statements would be purposeful, because they would mean that I who am Brahman am not any of the products that may be described either as gross or as fine. Whatever Brahman is said to have created, I am not that, and by the principle of residue it would follow that I am Brahman pure and simple. The similar situation is with descriptions such as “without pr ṇa” and “without a mindŚ” if they, attributed otherwise to Brahman, are not taken to refer to the transmigrating Self, then they would be useless because of being not informative, as I also indicated at the end of the satyaṁ jñ nam anantam nandam brahma analysis. From the Brahman = Self standpoint, however, they become most informative, because now we can say that whatever one may identify with in the field of points of identification is meaningfully denied. Am I gross? No! Am I fine? No! Am I mortal? No! Do I have a mind? No! And so on. As Śa kara says, these characteristics are predicated to Brahman for negating the false superimposition of characteristics that do not belong to the Self. Whatever we may otherwise think of the argument, it helps us understand that Śa kara’s attribution of negative characteristics to Brahman was not just a theological requirement posed by the BS procedure of forming the notion of Brahman: it was quite meaningful, but not in isolation from the context of the identity statement. We are now, then, deep in this context, 425 where two things need to happen. First, the individual reference of the category of tvam must be fixed, if it has not been fixed yet: while the reference of the category of tat is quite independent, the reference of tvam seems to require procedurally the context of the identity statement. Second, some sense must be made of the identity statement itself. Let us go back, therefore, to this context, and the Taittir ya Upaniṣad. The Taittir ya text that we have been considering says that Brahman is Being, knowledge, infinite. Later the text says that one should know Brahman as bliss, and that completes the definition of Brahman. This Brahman is related to one’s Self in three important ways. First, creation proceeds from this Brahman and culminates with the birth of man who is the Self consisting of food, puruṣa anna-rasamaya; the Self comes from Brahman. Second, once this creation has been completed, Brahman enters created beings, tat sṛṣṭv tad evopr vi at, and is said to reside in the cavity that Śa kara interprets as the intellect.106 Third, this Brahman that creation proceeds from is most directly called the SelfŚ “From that [Brahman] which is this Self, space is born,” tasm d v etasm d tmana k aḥ sambhūtaḥ, etc. Thus, Brahman is presented as the source of creation in general and the five-layered Self in particular, is directly identified with the Self in tasm d v etasm d tmanaḥ, and is said to have entered creation as the Self: Brahman that creates is the Self, and as that same Self it enters creation and becomes nested in the man of five sheaths. Śa kara draws an explicit equivalence between tasm d v etasm d tmanaḥ and tat tvam asi, and in the context of the Taittir ya that we are following, this is the identity statement that juxtaposes the two categories. Since in the statement “from that [Brahman] which is the Self [space came about]” the word “Self” is used in the sense of Brahman itself, it follows that Brahman is the Self of the cognitive agent. Further, the text shows that Brahman is the Self in the text “He attains the Self of bliss.” The same follows from the entrance of Brahman, for the text 106 TUBh 2.1.1, VI.67. 426  “Having created it, it entered into it” shows that Brahman entered into the body as the individual Self.107 How is it that I am this great ground of Being, the causal Brahman? What does the Upaniṣad mean when it says, “You are that?” The statement is an instance of co-referentiality, s m n dhikaraṇya, and it would appear that this co-referentiallity can be taken in two ways, depending on where one’s point of identification is located. The general purpose of tat tvam asi is to negate whatever one may understand as the Self but is not so, and in that sense, it is an instance of negational co-referentiality, apav da-s m n dhikaraṇya, which Śa kara says explicitly under BSBh 3.3.9.108 As such, the identity statement is intimately related to the negative characteristics of Brahman that can be subsumed under another famous Upaniṣadic text, neti neti: while the individual negative characteristics of Brahman negate individually, neti neti negates generally and it negates everything that the Self is not. To be more specific, at this level what the identity statement says is that one is not anything that can be an object of consciousness and conceived as separate from oneself: the objective part of any propositional consciousness that the subject can identify with, from parts of one’s body to one’s most intimate thoughts, what Śa kara otherwise calls idam-aṁ a.109 This does not mean that one is no longer aware of any objects, of course, but simply that one does not identify with them. What remains when such negation of superimposition has been applied is pure subjectivity, jñ tṛtva. This is the intended 107 tasm d v etasm d tmanaḥ iti brahma y eva ātma-śabda-prayogāt veditur ātmaiva brahma. etam nandamayam tm nam upasaṅkr mati iti ca ātmatāṁ darśayati. tat-praveśāc caś tat sṛṣṭv tad ev nupr vi at iti ca tasyaiva jīva- rūpe a śarīra-praveśaṁ darśayati. ato vedituḥ svarūpaṁ brahma. TUBh 2.1.1, VI.65. 108 “A negation, to define it, is the posterior correct notion which removes a prior deep-rooted false notion, when there is such a deep-rooted false and settled notion in regard to a certain thing; for instance, the notion of the Self in regard to the psychophysical complex is subsequently driven away by the notion of the Self in regard to the Self only through the correct notion arisen from ‘You are that.’” apavādo nāma—yatra kasmiṁścid vastuni pūrva- niviṣ āyāṁ mithyā-buddhau niścitāyām, paścād upajāyamānā yathārthā buddhiḥ pūrva-niviṣ āyā mithyā-buddheḥ nivartikā bhavati—yathā dehendriya-saṁghāte ātma-buddhiḥ, ātmany eva ātma-buddhyā paścād-bhāvinyā tat tvam asi ity anayā yathārtha-buddhyā nivartyate. BSBh 3.3.9, III.629. 109 USP 6.5. 427 meaning of the tvaṁ-pad rtha in the identity statement, and when one understands it as such, the statement is more directly an instance of vi eṣya-vi eṣaṇa-bh va, a case where two words stand for the same reference. The statement literally says that the subject of knowing is Brahman, that is, is Being, consciousness, infinite, bliss. This is the ultimate point that words can reach and exercise their direct signification function. Let us expand on this. The relation of the Self to Brahman is established through Brahman’s feature of jñ nam or consciousness, but there is a problem with this. Brahman was defined as jñ nam anantam, unlimited consciousness and ontological omniscience, but its identification with the Self that is the cognitive agent imposes limitations on consciousness, since it presupposes the system of cognitive agency: agent, object, instrument, content, and action. The agent does not have to identify with any object that is presented to it, but objects are presented all the same, and they are required for the self-understanding as the pure agent. The distinction is, further, required for the very possibility of understanding the identity statement. So, if Brahman’s being consciousness means that Brahman is the cognitive agent, it would not be unlimited—an object is a second entity that eo ipso means limitation for the agent—and its being transient would follow because cognition implies change. That Brahman is just consciousness is no less problematic either, when we pause to consider what that really means. Śa kara said that consciousness or jñ na means the verbal action of knowing – jñ naṁ jñaptiḥ avabodhaḥ, bh va-s dhano jñ na- abdaḥ – and any form of action implies change, a transition from one state to another, not Being, but becoming. The final punch, therefore, is that Brahman that is defined by the word jñ nam and is identified with the cognitive agent cannot be directly signified by that word, but it can be indicated. 428  We will remember here that the empirical Self for Śa kara was at the core a reflection of the pure Self as the light of consciousness in the intellect, buddhi, which was for this reason the first and the closest point of Self-identification that happened through mutual superimposition. The Self as the agent of cognition, the one that is directly denoted by the personal affix tiṅ in the finite verbal form j n ti, “s/he knows,” and the one directly denoted by the personal pronouns tvam and aham, was the reflection of the Self in the intellect, the tm bhasa. Cognition as action that involves the syntactic relations or k rakas, such as the agent and instrument, takes place in the intellect—which was why the most direct appellation for the individual Self or j va was vijñ n tman, the Self of cognition—and the intellect itself that accommodates such cognition was its instrumental factor, karaṇa. The object of cognition was also a transformation of the intellect in the shape of the external object. Now, all of this was possible because the intellect, buddhi, was suffused by the light of consciousness that is the Self. The intellect is insentient, so it cannot have cognition on its own, whereas the Self is cognition pure and simple that cannot involve change, so neither of the two can be the reference of the finite verb “s/he knows,” and the saving grace is found in the sense of Self—the reflection of the Self—and the processual action of knowing that happens in the intellect. The agent is denoted by the personal ending, whereas the action of knowing is denoted by the root to which the ending is applied. Because all of this is possible through the fact that the Self illumines the intellect, the Self is indicated by these two—the action and the agent of knowing—through immediate proximity, but it is not denoted directly. The predication of the characteristic of jñ nam to Brahman, thus, happens through the secondary signification function of the word, lakṣaṇ (to be distinguished from lakṣaṇa, the word for definition).110 110 This is largely based on USP 18. 429  Because Brahman’s nature of being a knower cannot be separated from him and because it does not depend on instruments like senses, its being eternal is proven though Brahman’s nature of knowledge. Therefore, Brahman is not knowledge as the sense of the root, because [such knowledge that Brahman is] it is not action in nature. For this reason, Brahman is not the agent of knowing either, wherefore it is not denoted by the word “knowledge.” Still, it is indicated by the word “knowledge,” which is a specific attribute of the intellect and denotes the semblance of Brahman. However, it is not denoted, because it does not have the properties such as genus, which are the ground of uses of words.111 This is supported with the most favorite Advaita argument that is like the garam masala for every curry: it is the only way that scripture can be meaningful. Knowing presupposes cognitive agency, yet such agency is empirically known. If scripture were to affirm agency, if Brahman were the cognitive action or the agent that knows another or knows itself directly, it would not be a reliable warrant since it would have no sphere of operation. The Upaniṣads as a reliable warrant must have a reference to it, but they do not affirm it. It is like the necessary prima facie view that must be stated so that the true and conclusive view can be stated against it.112 Thus, Brahman is consciousness, but not of any kind that can be actually experienced and named. Now, this identification of Brahman with the inner Self in the feature of consciousness makes Brahman an inward category, myself, and this further poses a problem with the qualification of satyam, Being. We saw that this Being was what makes things grow and that it was coordinated with everything, but in such a way that the reality of all multiplicity was denied. This multiplicity is denied as real, but the assumption of multiplicity is necessary for the very possibility of affirming that Being is coordinated with everything, and for the possibility of 111 vijñāt -svarūpāvyatirekāt kara ādi-nimittānapekṣatvāc ca brahma o jñāna-svarūpatve 'pi nityatva-prasiddhiḥ. ato naiva dhātv-arthas tat, akriyā-rūpatvāt. ata eva ca na jñāna-kart ś tasmād eva ca na jñāna-śabda-vācyam api tad brahma. tathāpi tad-ābhāsa-vācakena buddhi-dharma-viśeṣe a jñāna-śabdena tal lakṣyateś na tu ucyate, śabda- prav tti-hetu-jāty-ādi-dharma-rahitatvāt. TUBh 2.1.1, VI.66. 112 USP 18.3-8. 430 denial, very much like the case of cognitive agency that requires the complex around knowing as action. This presents Brahman denoted by the word “Being” as an external object of knowing, and that cannot be reconciled with the sense of jñ nam, myself as knowledge simple. It turns out that Brahman cannot be denoted by satyam either, but must be indicated as well: Likewise, by the word ‘Being.’ Brahman, on the account of its essence being the state where all distinctions have been abolished, is indicated in ‘Brahman is Being’ through the word ‘Being’ whose sphere is the external universal of ‘Being.’ It is not, however, denoted by it.113 So, Brahman being that great Being out there that makes everything grow does not, in fact, make anything grow, because there is no second thing other than Brahman to begin with. All that causality is just a show whose only purpose is to make it possible to express how awesome the director is. Unsurprisingly, the same is true of Brahman’s characteristic of bliss. The blissful Self that was closest to Brahman serves as a mark through which Brahman as bliss thick and solid can be indicated. Bliss is experienced in the intellect, and the intellect is in proximity to the Self that is the light of consciousness. The mental experience of bliss, thus, points to Brahman that is the ground of any experience of bliss,114 “but when the subject-object distinction that is a product of ignorance has been set aside by knowledge, there remains only the essential, thick and solid, one, non-dual bliss.”115 Such Brahman as bliss must be understood as just the ultimate point of aspiration in which the pursuit of that nirati aya-pr ti that was the synonym of the highest human good culminates. Brahman as bliss is indicated through the experienced bliss, but it is not denoted. 113 tathā satya-śabdenāpi. sarva-viśeṣa-pratyastamita-svarūpatvād brahma aḥ bāhya-sattā-sāmānya-viṣaye a satya- śabdena lakṣyate satyaṁ brahma iti; na tu satya-śabda-vācyaṁ brahma. TUBh 2.1.1, VI.66-7. 114 TUBh 2.6.1. 115 niraste tu avidyā-k te viṣaya-viṣayi-bhāge, vidyayā svābhāvikaḥ paripūr aḥ ekaḥ ānandaḥ advaitaḥ bhavatīty etam arthaṁ vibhāvayiṣyann āha. TUBh 2.8.1-4, VI.109. 431  So, on the level of the identity statement, the two categories of tat and tvam, standing for Brahman the great Being out there described in superlative language, and the inner Self that is tinged by ignorance and is liable to suffering that is transmigration, restrict one another because of being co-referential, and there obtains a special meaning of the identity statement in which the reference is neither external nor liable to transmigration. Śa kara makes the point of emphasizing that this special meaning obtains without the respective categories giving up their individual meaning.116 This must be interpreted to mean that the respective categories do not directly obtain a secondary signification function in the sentence because the primary is blocked, as would be the case in “Devadatta is a lion,” where “lion” must be first reduced to “leonine.” In technical terms, the sentence does not require the words to exercise their gauṇa-vṛtti. This would be quite disastrous for the argument, in fact, because it would mean that either Brahman or the Self stands for something else, such as pradh na or the body. However, when combining in sentences through the vi eṣya-vi eṣaṇa-bh va, some restriction of meaning must obtain, as in Bhart hari’s black sesame or Śa kara’s black horse. This is not an equivalent case, as neither of the two pad rthas have a respective class from which it can be delimited, but it is quite like the definition of satyaṁ jñ nam anantaṁ brahma where something from the scope of the collocated categories had to drop. What drops in the meaning of the sentence is Brahman’s being external and the Self being liable to transmigration. In other words, clarifying the meaning of the tvaṁ- pad rtha, one had reduced the scope of the word through removing the sheaths covering the Self, only to realize that s/he had not considered the one point of identification that made the removal possible, Brahman the light of awareness. One learns from the sentence that this inner Self of mine is Brahman, that great unlimited Being that is not liable to change and 116 USP 18.171. 432 transmigration, but is not extraneous either.117 It is in this tiny space between the literal and indicated meaning, both of which are necessary, that liberation becomes possible, when it dawns on one that I myself, known to me most intimately, in fact am that great ground of Being out there: that is, that this Being is not out there at all. Elsewhere Śa kara says that what drops on the side of Brahman is its being the creator, which was, we should bear in mind, the key feature of Brahman as Īśvara.118 That will have tremendous consequences in Sarvajñātman’s formulation of the mah -v kya doctrine. But, non-difference is also stated in sentences such as “You are that.” How could both difference and non-difference be possible, given that they are contradictory? – There is not that fault, since we have established in the respective places the possibility of both on the analogy of the great space and the space in a pot. Moreover, when non-difference has dawned on one through statements such as “You are that,” the individual Self’s being liable to transmigration is lost, and so is Brahman’s being a creator, because full knowledge defeats the practical reality of difference that extends through false awareness.119 Brahman, thus, created the world, entered the cavity of the heart that is the intellect, enveloped itself with five sheaths to perform ritual and meditation and experience the “bliss” that is suffering that they bring, but really did nothing of the kind. I should like to point out here without going into details that this (and only this) is the demarcation line between the two truths or realities in Advaita Vedānta, the absolute, param rtha, and the practical, vyavah ra. However, that Brahman is literally the inner Self has tremendous consequences on the definition of Brahman as that inner, non-transmigrating Being. Precisely because the individual pad rthas in the identity statement do not give up their individual meanings, the definition of 117 USP 18.169-172, 194-5. 118 Julius Lipner (2000) is most certainly wrong in claiming that nothing is lost from the meaning of tat, as I hope is amply clear. What is crucially lost is the sense that Brahman is mediate, external, and that Brahman is the creator. 119 nanv abheda-nirdeśo ’pi darśitaḥ—tat tvam asi ity evaṁ-jātīyakaḥ; kathaṁ bhedābhedau viruddhau sambhaveyātām? naiṣa doṣaḥ, mahākāśa-gha ākāśa-nyāyenobhaya-sambhavasya tatra tatra pratiṣ hāpitatvāt. api ca yadā tat tvam asi ity evaṁ-jātīyakenābheda-nirdeśenābhedaḥ pratibodhito bhavati, apagataṁ bhavati tadā jīvasya saṁsāritvaṁ brahma aś ca sraṣ tvam, samastasya mithyā-jñāna-vij mbhitasya bheda-vyavahārasya samyag-jñānena bādhitatvāt. BSBh 2.1.22, II.328-9. 433 Brahman the entity behind the identity statement as satyaṁ jñ nam anantam nandam but identified with the inner Self now obtains through lakṣaṇ , the secondary signification function: words do not reach Brahman to express it directly, but because of the proximity of the pure Self and the intellect they can indicate it. Being cannot be that external Being that persists through changes, because it is my inner Self; knowledge cannot be the action or the agent, because they change; and, bliss cannot be the positive experience of pleasure, because that is transmigration. With this feat Śa kara avoided, to his mind, the second problem with Brahman and language. By treating the attribution of characteristics as definiendum-definiens relation, he avoided the prasaṅkhy na problem of Brahman being a v ky rtha, a sentential reference as a definite description; by claiming that words merely indicate Brahman, he avoided the problem of Brahman being designated directly by words, v cya. Therefore, it is proven that Brahman is not directly expressible by individual words, in keeping with the statements “Whence words return along with the mind without reaching it” and “inexpressible, non-supporting.” Furthermore, Brahman is not a sentential reference in the manner of the blue lotus.120 Finally, by putting the absolute onus on the identity statement rather than the definition of Brahman or the injunction, in opposition to the prasaṅkhy na-v dins, he could give the final punch and claim that such knowledge of Brahman, though obtained from verbal utterance, was not mediate in kind: – One cannot experience the satisfaction of eating by hearing a sentence. The sentence analysis is like making milk-rice from cow-dung. – True, the understanding from sentences that are not about the Self is mediate. However, it certainly is immediate in respect to the inner Self, like in the case of getting the number right. The inner Self is its own evidence, synonymous with “knowable to itself.” When the sense of Self ceases, there is an experience of one's own Self.121 120 ataḥ siddham yato v co nivartante apr pya manas saha (TU 2.4.1), anirukte 'nilayane (TU 2.7) iti ca avācyatvam, nīlotpala-vad avākyārthatvaṁ ca brahma aḥ. TUBh 2.1.1., VI.67. 121 yathānubhūyate t ptir bhujer vākyān na gamyate | vākyasya vidh tis tadvad gośak t-pāyasī-kriyā || 434 The case of getting the number right refers, of course, to Śa kara’s famous example of how knowledge of the inner Self is available to oneself directly, perceptually, but may be forgotten, in which case the anamnesis can happen only through a linguistic utterance. A boy was told that there are ten boys in total, but his perception presents only nine. He had forgotten to count himself, and someone must tell him, “Well, you are the tenth.” As soon as that happens and he is competent enough to understand it, no further perception is needed, as he is known to himself most intimately.122 Likewise, our student has been counting the sheaths enveloping the Self and got very close in figuring out what the Self could not possibly be, but must be told eventually that it is the very possibility of counting what he really is. Thus, tat tvam asi was a case of co-referentiality that was both a negation and qualification, apav da and vi eṣya-vi eṣaṇa-bh va. The definition of Brahman had been obtained from the Upaniṣads. Through the application of the negative characteristics of Brahman, all that is not the Self had been removed from the pure agent. Finally, tat and tvam qualified one another, where some negation again had to take place for a sentence reference to obtain through indication. This reference obtained through the denial of causality of Brahman and transitive awareness of the Subject, which was also the domain, we shall remember, of the negative characteristics. And this reference, the reference behind tat tvam asi, was neti neti. The word tat obtains the meaning of the inner Self, and the word tvam obtains the meaning of tat. The two conjointly remove suffering and mediacy. In this way, they mutually convey the meaning of neti neti.123 satyam evam anātmārtha-vākyāt pārokṣya-bodhanam | pratyagātmani na tv evaṁ sa khyā-prāpti-vad adhruvam || svayaṁ-vedyatva-paryāyaḥ svapramā aka iṣyatām | niv ttāv ahamaḥ siddhaḥ svātmano 'nubhavaś ca naḥ. USP 18.198-200. 122 The analogy is found in the TUBh 2.1.1 and USP 18.169ff. 123 tac-chabdaḥ pratyag-ātmārthas tac-chabdārthas tvamas tathā | duḥkhitvātpratyag-ātmatvaṁ vārayetām ubhāv api || evaṁ ca neti netyarthaṁ gamayetāṁ parasparam. USP 18.197-8. 435  Since there is no other more apposite description, Brahman is called neti neti. Because, apart from teaching by way of negating the phenomenal world of manifestations, there is no better description of Brahman.124 We still need to address the question of “another cognition of Brahman,” but we will do that after we return to knowledge as cultivation. Let us summarize now what we have discovered so far. The passages of the Upaniṣads that are concerned with para-vidy have as their goal to present the Self as single. They are governed by what we called the identity statements, because liberation follows as a result of understanding such identity statements. The identity statements correlate two categories, tat standing in general for Brahman as the cause of the world, and tvam standing for the individual who happens to listen to the Upaniṣads. The category of tat is defined in the TU as satyaṁ jñ nam anantam brahma, to which nanda is added as fourth. These characteristics, which Śa kara calls brahmaṇo lakṣanam or the defining characteristics of Brahman, are elaborated in texts within the category of Upaniṣadic passages that are about the single Self. Tat, further, has negative characteristics too, but it turns out that their purpose is to apply on the side of the individual Self: they relate the categories that are placed in identity. The category of tvam presents the individual Self in all its points of identification, and is also elaborated in the TU through the teaching of the five sheaths that envelop the Self that has lodged into the intellect. Initially the identity statement is an instance of negational co- referentiality, where Brahman’s negative characteristics are used to remove from the notion of the Self anything that it may objectively identify with. When tvam is purged of the sheaths so that only its relation to the intellect remains, the word stands for the cognitive agent that is a reflection of the pure Self/Brahman in the intellect. Once this reference has been fixed, the identity statement is an instance of vi eṣya-vi eṣana-bh va where two words that stand for a 124 na hi, etasmāt iti na, iti na, iti prapañca-pratiṣedha-rūpāt ādeśanāt, anyat paramādeśanaṁ brahma o’stīti. BSBh 3.2.22, III.600. 436 single reference mutually restrict their scope: Brahman is no longer external and the creator/cause, whereas the inner Self is not liable to suffering/transmigration. The identity statement literally says that the inner Self is Brahman, but the thus denoted single entity is now defined by satyaṁ jñ nam anantam nandaṁ brahma through the secondary signification function of the words. While Brahman in its essential nature is directly expressed neither through words nor though sentences, these are the only way that one can learn about it and attain liberation. 437  CHAPTER NINE: LIBERATION AND THE INQUIRY INTO BRAHMAN Since both experience of the results of action and knowledge of Brahman are results of the karma that had started bearing fruits, it is but proper that they not be mutually exclusive.1 The Dharma of Disengagement and Desire As we saw in the Second Chapter, for the Bhā a Mīmāṁsakas the driving force behind the constitution of a text as the blueprint of a sacrifice was the syntactic expectancy of the productive power of the verb, bh van . While we not need go into details, it appears from Śa kara’s comment on BĀU 1.4.7 that there were Vedāntins who wanted to apply the idea of bh van for forming meditations on Brahman on the ritual model. Now, we saw that Śa kara argued that the knowledge given in the Veda can transit into action only if the cognition produced by a Vedic statement becomes somehow qualified by the idea that one should do something. There were scriptural cases, he further claimed, where such transition into action could not happen in principle, yet they were for the good of man, puruṣ rtha. These were most evidently the Vedic prohibitions. When a Vedic statement does not enjoin a course of action but rather prohibits, the terminus of meaning of such a sentence would be just the cognition itself, because no action could really follow from it. Let us consider the sentence “Do not consume hemlock, it is poisonous.” Obeying what the sentence commands ideally does not require effort on the part of man, and the sentence becomes fruitful just by producing a clear cognition. Śa kara says that such sentences achieve completion or paryavas na just insofar as they are understood: they do not cross the threshold of 1 ārabdhasya phale hy ete bhogo jñānaṁ ca karma aḥ | avirodhas tayor yuktah. USP 4.4. 438 action.2 In fact, prohibitions act as checks to an impulse to action that otherwise obtains, by giving rise to some correct understanding. Seeing food, one is naturally prompted to eat it, and the prohibition is effective precisely by checking this natural ignorant tendency when food is poisonous. It does this by invoking recollection of the nature of poison. This is quite a significant argument, because it allows Śa kara to claim that there are types of Vedic sentences favored by Mīmāṁsakas over which their most prized possession, the theory of bh van , could not be applied in principle, precisely because there is nothing to do. The notion of paryavas na or completion is crucial here, because for Mīmāṁsakas it marked the point where the injunction had secured all the details necessary for the performance of the relevant ritual, the point where it was clear what should be done, for what purpose, following what procedure (kim, kena, katham), and where the ritual agent had been convinced that good would be obtained through the sacrifice.3 The idea of bh van could not really be applied in the case of prohibitions, because no production of anything was expected. While this may be stretching the argument a bit too far—it certainly is not immediately apparent in many prohibitions why something should not be done, and the arthav da texts attached to them may exercise some verbal bh van on Kumarila’s account—it is an ingenious critique because it ties the attaining of something good just to knowing. It particularly damages the procedural component of action, because it makes little sense to ask, “How should I not consume hemlock,” and it is a good starting point for Śa kara to claim that there are sentences in the Veda, even of the injunctive type, whose meaning terminates in sheer understanding (avagati-niṣṭh ) and which 2 na ca pratiṣiddha-viṣaye prav tta-kriyasya akara ād anyad anuṣ heyam asti. akartavyatā-jñāna-niṣ hataiva hi paramārthataḥ pratiṣedha-vidhīnāṁ syāt. BĀUBh 1.3.1, VIII.41. See also BSBh 1.1.5. 3 On paryavas na, see McCrea 2008:55-98. 439 serve dharma that is characterized by disengagement. Little wonder that Śa kara classified knowledge of Brahman as prohibitory in kind: Therefore, the prohibitory injunctions terminate just in knowledge that corresponds to the thing, and they have not even a whiff of reaching human effort. Likewise, the injunctions of knowledge that correspond to things such as the Supreme Self achieve completion in just that much. … Thus, the Vedic dicta inculcating the true nature of the Supreme Self, removing the erroneous notions about its being gross, dual etc., automatically assume the character of prohibitions of all actions, for in both cases there is equally an absence of action.”4 This is the reason Śa kara related the statement which introduces the triple process of brahma-vidy , Yājñavalkya’s tm v are draṣṭavyaḥ rotavyo mantavyo nididhy sitavyaḥ, to renunciation. The purpose of this injunction terminates in knowing—a Self needs to be seen— and the injunction discloses processes of knowing as its procedure. Precisely for this reason the injunction discloses for an adhik rin that ritual and the other rama practices must be renounced in the pursuit of the SelfŚ “The section on Maitreyī was commenced to indicate the means of immortality which is wholly independent of action. It is the knowledge of the Self, with renunciation of everything as its part.”5 Even more indicative is a statement in the introduction to the comment on the first run of the Yājñavalkya-Maitreyī dialogue, to the effect that going forth or p rivr jya is enjoined as an auxiliary to brahma-vidy .6 The status of aṅga accorded to renunciation meant that brahma-vidy could not proceed without it. In Sarvajñātman’s words, whereas in the ritual section of the Veda statements about 4 tasmāt pratiṣedha-vidhīnāṁ vastu-yāthātmya-jñāna-niṣ hataiva, na puruṣa-vyāpāra-niṣ hatā-gandho 'py asti. tathehāpi paramātmādi-yāthātmya-jñāna-vidhīnāṁ tāvan-mātra-paryavāsanataiva syāt. ... tasmāt, paramātmā- yāthātmya-jñāna-vidher api tad-viparīta-sthūla-dvaitādi-jñāna-nivartakatvāt sāmarthyāt sarva-karma-pratiṣedha- vidhy-arthatvaṁ sampadyate; karma-prav tty-abhāvasya tulyatvād, yathā pratiṣedha-viṣaye. BĀUBh 1.3.1, VIII.41- 3. 5 yat kevalaṁ karma-nirapekṣam am tatva-sādhanaṁ tad vaktavyam iti maitreyī-brāhma am ārabdham. tac cātma- jñānaṁ sarva-sannyāsā ga-viśiṣ am. BĀUBh 2.5.1. 6 tasmān na sādhanāntara-sahitā brahma-vidyā puruṣārtha-sādhanam, sarva-virodhāt, sādhana-nirapekṣaiva puruṣārtha-sādhanam—iti pārivrājyaṁ sarva-sādhana-sannyāsa-lakṣa am a gatvena vidhitsyate. BĀU 2.4.1, VIII.298. 440 engagement and disengagement, pravṛtti and nivṛtti or injunctions and prohibitions, were of equal validity, Yājñavalkya’s statement concerned solely disengagement pertaining to the one who renounces all action: brahma-vidy did not and could not involve action.7 This is how Advaitins have generally understood Yājñavalkya’s statement. We should also remember from the Seventh Chapter that knowledge was dispassion, predicated on understanding transmigration. Once knowledge had arisen, striving for something productive would be “non-productive,” and the further human good had to come just through understanding. One may, of course, object that knowing as a process rather than a result was also a form of action or engagement, but Śa kara’s point was that this action was not productive. It was a form of guided anamnesis, making a student remember something that he had always known, like Meno’s slave in Plato’s famous eponymous dialogue, and—here is the crucial claim—the requirements of this anamnesis intersected in the object one tried to know, Brahman. As we saw in Chapter Six, knowledge was dependent on the thing, not on human choice. In matters of human choice, one must follow the urge for the result; in matters of knowing, one must follow the requirements of the thing. Brahman is the “thing” or the pradh na factor in a brahma-vidy , and understanding proceeds through what Brahman requires: state the definition of Brahman; elaborate Brahman’s characteristics in the texts where they appearś reduce the scope of the Selfś draw the full identity; repeat until the cognition of unity is indubitable. Another thing that needed thorough reevaluation in this light was the role of desire. We saw in the Seventh Chapter that Śa kara associated liberation with the desire for the Self, which was the desire for liberation: tma-k ma, mokṣa-k ma. Advaitins have argued that this was a quantum jump of a sort, a desire that does not obtain naturally but must be brought about through 7 SŚ 1.71-2. 441 cultivation. Without it, inquiry into Brahman through the three processes could not be successful. However, if the desire was there, it did not matter how it got there. The inquiry into Brahman depended on the desire for liberation, not on understanding ritual, because when the desire for liberation had arisen, ritual had already become superfluous: it had accomplished its purpose, given rise to dispassion and the striving after liberation, and that could have just as well happened in a past life. Thus, for Advaitins the inquiry into Brahman that was otherwise the counterpart to the inquiry into dharma was radically different from the first. The inquiry into dharma was required so that a successful sacrifice could be performed, and the sacrifice itself had a corresponding injunction that stipulated who the sacrifice is for through its x-k ma statement. The jijñ s was just that, an inquiry that followed once the Veda had been studied. For Advaitins, however, brahma-jijñ s was not just an inquiry: it was the process itself, consisting in ravaṇa, manana, nididhy sana, that ends in liberation, it was the thing on which human good follows, and it had to express the appropriate desire through which the adhik rin would be ascertained. Ath to brahma-jijñ s was in a sense not a counterpart to ath to dharma-jijñ s , but to svarga-k mo yajeta. And, whereas the desire for heaven obtained naturally through the fact of embodiment, the desire for Brahman/liberation was so radically different that it had to be the starting point of the process. Padmapāda, thus, related Śa kara’s statement of the four prerequisites for the study of Brahman that we discussed in the Seventh Chapter to Śa kara’s choice to analyze jijñ s as jñ tum icch , desire to know, in the literal sense of the desiderative rather that the technical sense of vic ra or deliberation, where the object of deliberation is predicated rather than the desire itself. To put this more plainly, the immediate point of ath to brahma-jijñ s was not to 442 introduce the inquiry into Brahman as a matter of course that would follow in the Vedic curriculum after the study of the Veda and the inquiry into dharma, but to affirm that there is such a thing as the desire to know Brahman, radically different from our common notion of desire, which, when present, would lead to inquiry into Brahman through its own force, and when absent the inquiry would either not follow, or would fail miserably if it did. Such analysis of the desiderative justified the statement of the four prerequisites as that on which the inquiry was consequent, and they constituted what the desire to know Brahman was. In its consummation, the desire to know Brahman is mumukṣutva, the desire after liberation. In more theological terms, the presence of such desire to know Brahman was the transition point from vividiṣ to mumukṣutva/mokṣa-k ma.8 The Processes of Knowing Brahman and the Doctrine of Pram ṇa I mentioned in the Fourth Chapter that we can gather very little about the role of ravaṇa and manana in pre-Śa kara Vedānta. They are regularly discussed in the Bh ṣya and the V rttika on BĀU against opponents, but the onus is always on meditation, nididhy sana. The real Vedāntic s dhana was meditation, and meditation was the process that had to be repeated more than once, in accordance with BS 4.1.1, vṛttir asakṛd, upade t. Most informative were Bhart prapañca and Ma ḍana Miśra, in whose case it seemed as if ravaṇa was the first understanding of Brahman, which in its turn became the instrumental factor of meditation and the direct experience of Brahman. With Śa kara, however, the onus shifted to ravaṇa and manana, and they were to be repeated until full understanding. Sureśvara explicitly said that the principle of repetition involved in the notion of meditation could be meaningful only regarding ravaṇa and manana: 8 See PP, p.52-67. 443  Now, if it is the case that you could not possibly live without prasaṅkhy na, we will make that possible for you, but only concerning ravaṇa, etc. Let the principle of repetition in prasaṅkhy na operate with respect to hearing of the teaching. One understands after hearing perfectly that which was slightly or half-heard.9 This does not necessarily mean, it seems to me, that the third process could not be repeated, but that it was not to be taken as prasaṅkhy na. Sureśvara was, of course, echoing the master from the BSBh 4.1.2, who said that some may understand the identity statement on one hearing, but generally full understanding proceeded through incremental gains, removing one point of identification from the Self at a time. The repetition was of stra-yukti, obviously corresponding to ravaṇa and manana.10 Notably, there is no mention of meditation or a third step in general in this most important comment for Śa kara’s soteriology, based on which we reconstructed the doctrine of the identity statements. Now, Śa kara says that the three processes are “modes” of seeing the Self, which means that in Yājñavalkya’s statement, tm v are draṣṭavyaḥ presents the goal, vision of the Self, while the rest states the procedure.11 The three should be performed sequentially so that the first two may have a compounded effect in the third, where one would be able to “reflect with certainty,” and certainty in general seems to be the big attainmentŚ Therefore, the Self should be seen, it should be made an object of vision. It should be heard about first, from a teacher and from scripture; afterwards, it should be considered through reasoning (tarka); then, it should be reflected upon with certainty. Thus, the Self is seen through the execution of the three practices of hearing, consideration, and reflection. When they reach unity, then the perfect vision of the unity of Brahman is transparent, not otherwise, just by hearing.12 9 athaivam api prasa khyānam antare a prā ān dhārayatum na śaknoṣīti cet, śrava ādau eva sampādayiṣyāmaḥ. katham? prasa khyāne śrutāv asya nyāyo ’stv āmreḍanātmakaḥ | īśac-chrutaṁ sāmi-śrutaṁ samyak śrutvāvagacchati. NaiS 3.125. 10 BSBh 4.1.2. 11 tasmād ātmā draṣ avyaḥś sa ca śrotavyo mantavyo nididhyāstavya iti ca darśana-prakārā uktāḥ. BĀUBh 2.5.1, VIII.323. 12 tasmād ātmā vā are draṣ avyaḥ darśanārhaḥ, darśana-viṣayam āpādayitavyaḥś śrotavyaḥ pūrvam ācāryata āgamataś caś paścān mantavyaḥ tarkataḥś tato nididhyāsitavyo niścayena dhyātavyaḥ; evaṁ hy asau d ṣ o bhavati śrava a- 444  In all cases, however, consideration through reasoning must conform to what is ascertained through hearing. Likewise, reflection must conform to what is thought through reasoning, that is, what is known as certain through hearing and consideration.13 The three should, then, be performed in sequence, each subsequent step starting where the previous had left off and in conformance with its gains, and at the stage of reflection one should be able to cogitate with certainty on what had been arrived at through the first two. Now, as I said above, it is obvious that the three processes, but more restrictedly the first two which we may collectively name brahma-vidy , are equivalent with the brahma-jijñ s from the opening of the BS. Sarvajñātman, for instance, wrote his masterpiece the Saṅkṣepa- r raka, a versified summary of the BSBh, on the model of such brahma-vidy with a teacher. The same is also quite evident from Śa kara’s comment on the first two sūtras in the BS: the requirements of qualification on the part of the student are the same in BSBh 1.1.1 and in the opening of the prose portion of the US, and Śa kara talks about what amounts to ravaṇa and manana under 1.1.2. This is an important point to bear in mind for appreciating that the three processes do not amount to the study of the Upaniṣads in the strict sense. The Upaniṣads are studied prior to the three processes, and this is clear, for instance, from Śa kara’s rejection of the possibility that the inquiry into Brahman be consequent on the study of the Upaniṣads: the Upaniṣads are read with the rest of the Veda, before one begins with either dharma-jijñ s or brahma-jijñ s .14 Śa kara likewise says that one of the nitya-karmas in the BĀU 4.4.22, ved nuvacana or the daily recitation of the Veda, includes the recitation of the Upaniṣads in its manana-nididhyāsana-sādhanair nirvartitaiḥ; yadā ekatvam etāny upagatāni, tadā samyag-darśanaṁ brahmaikatva- viṣayaṁ prasidati, nānyathā śrava a-mātre a. BĀUBh 2.4.5, VIII.304. 13 sarvathāpi tu yathāgamenāvadhāritam, tarkatas tathaiva mantavyamś yathā tarkato matam, tasya tarkāgamābhyāṁ niścitasya tathaiva nididhyāsanaṁ kriyata iti. BĀUBh 2.5.1, VIII.324. 14 svādhyāyādhyayanānantaryaṁ tu samānam. BSBh 1.1.1, I.5. 445 scope.15 In other words, he who comes to an Advaita guru for instruction does not come a blank slate, Upaniṣad-wise. A student that comes to a teacher for an inquiry into Brahman had already memorized the Upaniṣads and knows what they say: he understands the sentences. He does not know, however, what they mean.16 It is important to appreciate this point, it seems to me, with regard to the doctrine of pram ṇa. The three methods are not part of the pram ṇa complex as we commonly think of it, or as what can be described as a causal account of cognition. That place is reserved for the Upaniṣadic statements as part of the Veda: the Veda is the sole pram ṇa on supersensible things. Yet, there are statements in Śa kara’s commentaries in which he juxtaposes scriptural knowledge from the Veda, that includes the Upaniṣads and is described in terms such as “heap of words,” to “realization” (vijñ na, glossed in BhGBh 7.3 as sv nubhava), “understanding” (avagati) or “experience” (anubhava) that seems to be over and above such scriptural knowledge or scriptural verbal cognition. Consider, for instance, the following statement: [The objection that the Upaniṣads, labeled as “higher” knowledge, would fail to be part of the Veda, called “lower” knowledge, does not hold good,] because the intendent meaning is “realization of an object of knowing.” What is primarily intended here as the “higher knowledge” is the realization of the immutable [Brahman] that is knowable solely from the Upaniṣads, not the heap of words that constitute the Upaniṣads. On the other hand, the word “Veda” in all cases is intended in the sense of “heaps of words.” Since understanding the immutable is not possible without additional effort, such as approaching a teacher, and dispassion, even when the heaps of words have been understood, knowledge of Brahman is mentioned separately and called “higher.”17 15 yadā vedānuvacana-śabdena nityaḥ svādhyāyo vidhīyate, tadā upaniṣad api g hītaiveti, vedānuvacana- śabdārthaika-deśo na parityakto bhavati. BĀUBh 4.4.22, IX.645. 16 One would expect this to be an obvious point, and it has been recognized since the time of Deussen (1912:17-8), yet strangely it is not obvious to everyone. See, for instance, Malkovsky 2001:86-7, who says that the “disciple begins by listening to scripture as expounded by the guru, memorizing the texts,” and otherwise manages to make a caricature of the three methods, claiming that they are a “preparation for brahma-vidy ,” by a process of copy-paste from sketchy sources. 17 na, vedya-viṣaya-vijñānasya vivakṣitatvāt. upaniṣad-vedya-viṣayaṁ hi vijñānam iha parā vidyeti prādhānyena vivakṣitam, nopaniṣac-chabda-rāśiḥ. veda-śabdena tu sarvatra śabda-rāśir vivakṣitaḥ. śabda-raśy-adhigame ‘pi yatnāntaram antare a gurv-abhigamana-lakṣa aṁ vairāgyaṁ ca nākṣarādhigamaḥ sambhavatīti p thak-kara aṁ brahma-vidyāyā atha parā vidyeti. MUBh 1.1.5, V.13-14. 446 The takeaway from this passage is that an “additional effort” is required to understand what an Upaniṣadic text means, over and above what it may say. This additional effort consists in dispassion, etc., which would correspond to the four prerequisites, and approaching a teacher, which is the starting point of brahma-vidy that involves engagement in the three processes. They are required for understanding, realization, personal experience of what the text says. If they are required for such understanding, what is their pram ṇa status? There is, further, an important, curious passage in BSBh 1.1.2, in which Śa kara apparently muddles the waters of what precisely the pram ṇa for knowing Brahman is: Unlike in the inquiry into dharma, in the inquiry into Brahman ruti etc. are not the only pram ṇa. Rather, here the pram ṇa are ruti, etc. as well as personal experience etc., as far as that is possible, because knowledge of Brahman terminates in personal experience and is in the sphere of existent things.18 Just prior to this statement, Śa kara also positively affirms the role of reasoning in brahma-vidy as some kind of a pram ṇa in the inquiry into Brahman. There are, of course, many statements where Śa kara says that the only pram ṇa for knowing Brahman are the Upaniṣadic statements, and such cases particularly concern the identity statements. One explicit example is BSBh 1.1.4: “Nor is Brahman, though a fully existent entity, in the domain of sensory perception or the like, because the fact that Brahman is the Self, given in the statement ‘You are that,’ is not knowable except from scripture.”19 This had given rise to some discussions about the respective role of scripture, reasoning, and personal experience in Śa kara’s theology, with regard to the doctrine of pram ṇa, particularly in scholarship influenced by neo-Vedanta, in which the upshot was that Śa kara 18 na dharma-jijñasāyām iva śruty-ādaya eva pramā aṁ brahma-jijñāsāyām. kintu śrutyādayo ’nubhavādayaś ca yathā-sambhavam iha pramā am, anubhavāvasānatvād bhūta-vastu-viṣayatvāc ca brahma-jñānasya. BSBh 1.1.2, I.11. 19 na ca pariniṣ hita-vastu-svarūpatve ’pi pratyakṣādi-viṣayatvaṁ brahma aḥ, tat tvam asi iti brahmātma-bhāvasya śāstram antare ānavagamyamānatvāt. BSBh 1.1.4, I.16. 447 apparently thought that what is known through scripture must be somehow confirmed in personal experience. Anantanand Rambachan had analyzed and rejected such neo-Vedantic interpretations, without, however, clearly spelling out the role of the three methods and of personal experience in their relation to the doctrine of pram ṇa. His contention was that “personal experience, etc.” were “only subordinate and supplementary to ruti,”20 and that manana involved not anum na as a pram ṇa, but a form of reasoning which “is termed s m nyato-dṛṣt num na and is equivalent in modern logic to analogical reasoning. It is also designated as yukti and tarka. This type of reasoning is not itself a pram ṇa, but operates as an ancillary to a pram ṇa. Its function is to produce a belief in the possibility of a thing. In relation to brahmajñ na, the aim of all such tarkas is to strengthen the teachings of the Upaniṣads.”21 Aside from the fact that Advaitins did not accept any form of reasoning as an ancillary to scripture the pram ṇa, and aside from the sweeping generalization in drawing an equivalence between s m nyato-dṛṣṭa and tarka/yukti,22 Rambachan is roughly right in identifying the nature of tarka. He is also very much justified in rejecting the neo-Vedanta interpretations that want scriptural knowledge to be “confirmed” in personal experience, but is wrong in believing that ravaṇa was somehow the real deal, to which manana was an auxiliary, whereas personal experience was there just to supplement. This is just not how Śa kara presents the three methods: 20 Rambachan 1991:114. 21 Ibid, 103. 22 S m nyato-dṛṣṭa was, of course, a common form of inference in the common sense of pram ṇa, an instrument or source of knowledge with its characteristic domain. See, for instance, Bhatt 1962:256-64, for a good overview focused on the Bhā a school of Mīmāṁsā, but also discussing the early Nyāya and Sā khya mentions of the term. Rambachan seems to be repeating what M. Hiriyanna (1995Ś172) had said decades agoŚ “But in the very nature of the case, the arguments based on such examples [taken from ordinary life] are only analogical (s m nyato-dṛṣṭa) for, while they are drawn from the realm of common experience, Brahman by hypothesis transcends it. They can thus only give support to, or indicate the probability of, Vedāntic truth, and cannot demonstrate it independently of revelation. In other words, this type of argument is utilized here not as a pram ṇa, but as only an accessory to it (yukti).” Both these statements smack of “medievalism,” that is, of representing a pan-Advaita doctrine based on the medieval manuals such as Ved nta-S ra or Ved nta-Paribh ṣ , on which Rambachan depends throughout in any case. 448 we saw that he wanted them to work in a sequence of incremental gains, that understanding was very much equivalent to “personal experience” of some kind, and that it was the goal in which the consideration of the identity statements terminates. The key insight to appreciate, however, is that none of the three, ravaṇa included, was a pram ṇa to begin with, in the common causal sense that associates a pram ṇa with a specific domain that is in its sole competence, and that a false opposition is at play. Let us walk through this slowly. Advaitins, to begin with, did not think that the inquiry into Brahman, consisting of ravaṇa and manana, had the status of a reliable warrant in the common causal sense, that is, that which gives rise to a veridical cognition. Sarvajñātman was, for instance, quite emphatic what was at stake if the inquiry into Brahman were to have any causal role in the rise of the cognition of Brahman: In giving rise to cognition of the Supreme Self, the inquiry serves neither as the efficient cause like the recitation of the Veda, nor as an ancillary to the Veda. For, otherwise, Vedic statements would be dependent in the production of veridical knowledge of the identity of Brahman and the Self, and that is not right. Inquiry has no role in the rise of cognition of the sentence meaning; such cognition evidently appears in the competent language user immediately on hearing a sentence. Therefore, if inquiry is regarded as the efficient or an auxiliary cause in giving rise to valid cognition of meaning through postulation, the Vedic statement would lose its independence.23 As I said above, the student who had begun the inquiry into Brahman knows the Upaniṣads and understands what they say: he has a verbal cognition. He does not understand what they mean, for several reasons, first of which is that they seem to be saying mutually opposing things that cannot be accommodated in a straightforward manner. It is also clear that the sentence meaning 23 svādhyāya-van na kara aṁ gha ate vicāro nāpy a gam asya paramātma-dhiyaḥ prasūtau | sāpekṣatāpatati veda-giras tathātve brahmātmanaḥ pramiti-janmani tan na yuktam || vyutpannasya hi buddhi-janma sahasā vākya-śrutau d śyate vākyārthe na tato ‘sti buddhi-janane mīmāṁsana-vyāp tiḥ | tenārthāt kara ādi-bhāva-bhajane mīmāṁsanasyāśrite vedārtha-pramitau tu veda-vacasaḥ sāpekṣatāyāsyati. SŚ 1.19. 449 that is at stake, the sentence pertinent to the veridical verbal cognition, is that of the identity statement: this is the point on which the inquiry is focused both in US and SŚ. However, it is not clear to the student that the identity statement does intend to affirm full identity, rather than an assimilative identification or the like. If we look at the first two chapters of the prose section of US, it is evident that in the first inquiry the point is to establish that scripture does affirm such identity against what appears as evidential difference between the Self and Brahman, and the exercise of ritual and similar practices that go well with this difference, whereas in the second the problem shifts to the very possibility of such identity on grounds of reason. In other words, the student cannot form the veridical notion or pramiti about what the statement amounts to if he happens to come with notions that cancel such understanding. ravaṇa and manana, then, aim at establishing what the Upaniṣads really intend to say, and how that might be possible, respectively. They are not a pram ṇa, but a theological and a philosophical inquiry that is geared toward dealing with two different problems that appear in a sequence, in a case where a cognition had appeared but is not yet a pramiti. Sarvajñātman hints at a pram ṇic analogy and his commentator Puruṣottama expands on it. Suppose a king had a favorite servant whom other royal servants led by some trickery to another place. They left him there, but told the king that he had died. Imagine, now, that the servant had somehow come back and remained hidden in the royal garden. The king sees him unexpectedly, mistakes him for a ghost, and runs away in fear. The point of the story is that there is nothing wrong in the king’s seeingś he does see his servant, but he had been so fully convinced that the servant was dead, that the sight of him did not give rise to a cognition that is taken as veridical.24 Imagine seeing your close friend for whom you gave the funeral eulogy, sipping her 24 SŚ 1.14, Puruṣottama thereon. 450 coffee just across from you in a coffee shop. Similar would be the shock at entertaining the literal meaning of the identity statements as the one intended: one would be left in utter disbelief whether that could possibly be true, just as seeing the friend would make one question everything about the friend or about one’s own sanity. Thus, the inquiry does not have a productive role in the rise of the cognition of the Brahman-Self unity, since otherwise the scope of scripture pertaining to supersensible things would be at stake: the operative factor in the inquiry is repetition, vṛtti, specifically mreḍana or reiteration of words, with a human teacher, not the Upaniṣadic statements themselves; they are already known. If the inquiry had a pram ṇic status, the impersonal nature of the Veda would be at stake. One hears “You are that” and understands what the statement saysŚ there is verbal cognition by just that much. What the inquiry amounts to, therefore, is removal of faults in understanding, and this is quite comparable to cognitions that are doubtful because of faults that prevent certainty to obtain. Sarvajñātman presents these faults as psychological in nature, puruṣ par dha,25 but Śa kara’s presentation is more epistemological. What prevents veridical knowledge to rise from the identity statement is when the respective pad rthas that form the statement are obstructed by ignorance, doubt, and inverse ascription. These are already familiar to us—the case of uncertainty whether what one sees is a man or a post, or forming the notion that a man is a post—but the point is that they arise as a form of obstruction to a cognition that would otherwise be veridical. Remove the faults, and the pram ṇa has all that it needs for a veridical cognition. This is the role of ravaṇa and manana, to remove obstructions or cognitive defects.26 25 SŚ 1.14-17. 26 ataḥ, tān pratyeṣ avyaḥ padārtha-viveka-prayojanaḥ śāstra-yukty-abhyāsaḥ. yady api ca pratipattavya ātmā niraṁśaḥ, tathāpi adhyāropitaṁ tasmin bahv-aṁśatvaṁ dehendriya-mano-buddhi-viṣaya-vedanādi-lakṣa am; tatra 451  Now, I will propose here that in the BSBh 1.1.2 statement that I quoted above, and otherwise in that comment where Śa kara talks about anum na and tarka/yukti as having a positive role in knowing Brahman, Śa kara uses the notion of pram ṇa in a different sense, or rather as pertaining to a different set of categories. We should first note that he does not say ruti, but ruty dayaḥ that stands for a set of pram ṇas, not a single pram ṇa. I discussed at length in the First Chapter the notion of ruti and claimed that initially it did not stand for scripture as a unit that forms a set of pram ṇas with perception, inference etc., but for single scriptural statements that must be taken in the literal sense, and that it participated in several sets of pram ṇas as canons of interpretation; in all such sets, all the counterparts of ruti were labeled “inference.” Therefore, in our case, ruty dayaḥ does not refer to what we commonly think of ruti as the pram ṇa alongside perception and the rest, but to principles of interpretation of cases where conflicting verbal cognitions arise and need to be variously accommodated. That is, they are tools to organize the “scriptural data” of verbal cognitions. This is, in any case, how Śa kara’s commentators have generally understood ruty dayaḥ. Ānandagiri identifies this with the six pram ṇas that establish the principal- auxiliary relation, viniyoga, in a text. For our purposes, this means ascertaining which statements should be taken as the principal and which as subordinate, for instance as illustrations, so that an order in the conflicting verbal cognitions may be achieved. Another such set would be the so- called ṣaṭ-t tparya-liṅga or the six indicators of meaning that are used to determine the single meaning of an integral, complex text. ekena avadhānena ekam aṁśam apohati, apare a aparam iti yujyate. tatra kramavatī pratipattiḥ. tat tu pūrva-rūpam eva ātma-pratipatteḥ. yeṣāṁ punaḥ nipu a-matīnāṁ na ajñāna-saṁśaya-viparyaya-lakṣa aḥ padārtha-viṣayaḥ pratibandho ’sti, te śaknuvanti sak d uktam eva tat tvam asi-vākyārtham anubhavitum iti, tān prati āv tty- ānarthakyam iṣ am eva. BSBh 4.1.2, III.770. 452  The point is that this is exclusively an enterprise in scriptural theology, where the idea is to ascertain a core statement that must be read literally, and interpret the rest of the pertinent text in its light. For instance, in the Sixth Chapter of the Ch ndogya that presents the paradigmatic identity statement, tat tvam asi, is the point that Śvetaketu is literally Brahman, and if so what does that involve, or is it the case that evolution of everything from Brahman, Śvetaketu included, is the point? This has further ramifications as to whether the identity is a case of real identity or of identification, and consequences for the choice of practice. The second goal of the process of scriptural inquiry, once it has been determined that the Upaniṣads do intend to affirm the identity in the literal sense, is to ascertain the references of the two categories: which tat and which tvam are the same thing? We will follow the two in the next section through the first prose chapter of the UŚ, but my point here is that such methods are not pram ṇic methods so long as we understand pram ṇa as instruments of veridical cognition. Such pram ṇa would be one’s veda-sv dhy ya, to echo Sarvajñātman again. They are pram ṇa, on the other hand, because they do assist in the process where a cognition that had otherwise appeared does not have a clear reference. Further, the inquiry into Brahman unlike the inquiry into dharma is itself the process, and the process, for there to be liberation in the end, must culminate in full understanding, tat tvam asy- di-v ky rtha-parijñ nam to use Sureśvara’s turn of phrase. Because both categories that stand in the relation of full identity are known, and liberation is a visible result, predicated on understanding that must happen as immediately available to one’s awareness, it is eminently possible if not inevitable for one to ask the question, “How could I possibly be such an entity as the defined Brahman?” This is so because difference is “naturally,” perceptually available to 453 oneself. Therefore, the second inquiry ensues, one that is evidently a proper philosophical inquiry. Clearest, in my assessment, about the purpose of manana was Padmapāda, who refers to it as tarka and yukti, reasoning, which Śa kara, we will remember, identified as the modus operandi of manana. The pertinent question of Padmapāda’s concern was the following: it may be the case that one had understood the identity statement, since it has been ascertained that the Upaniṣads do, indeed, affirm the identity of Brahman with the Self. A cognition of the Self- Brahman full identity had been formed, but the cognition, although clear, is not yet certain because its content appears impossible, like seeing something so out of place and time that one would not believe one’s eyes before examining how and if it was possible at all.27 The content of the cognition seems impossible because tvam denotes the individual Self, the j va, to which the being of Brahman the denotation of the tat-pad rtha is predicated—the first liable to transmigration etc., the second ever free, etc.—and no certainty is possible until the contradiction between the two is removed. Padmapāda says that this contradiction needs to be removed by tarka, which is defined as “a discriminating notion concerning reliable warrants, signification functions, and content of cognitions.”28 This, then, is in part a meta-pram ṇic inquiry that concerns the specific domain of the pram ṇas, aiming to establish that the unity of Brahman with the Self is solely knowable through scripture; in part philosophy of language and mind, which should clarify that Brahman as the Self is not denoted by words, but indicated, as we saw in the previous chapter, and represented whenever there is any phenomenal consciousness; and 27 prak te punar viṣaye vidyotpāditāpi na pratiṣ hāṁ labhate asambhavanā viparīta-bhāvanābhibhūta-viṣayatvāt. tathā ca loke yasmin deśe kāle cedaṁ vastu svarūpata eva na sambhavatīti d dha-bhāvitam yadi tat kaṁcid daiva- vaśāt upalabhyate tadā svayam īkṣamā o 'pi tāvan nādhyavasyati yāvan tat-sambhavaṁ nānusarati. PP p.39. 28 pramā a-śakti-viṣaya-tat-sambhavāsambhava-paricchedātmā pratyayaḥ. Ibid. 454 ontology proper, that should show by analogical reasoning how it makes sense that Brahman be everything, including one’s Self. And, because this Self-Brahman relation must be personally understood and experienced, given that it concerns oneself and the ground of Being as a known category, one’s own personal experience about the related categories as well as that of the teacher also play a positive role in the second inquiry that should clarify the scriptural understanding. This is the crucial difference, in Śa kara’s eyes, of the inquiry into Brahman from the inquiry into dharma, where the sole purpose is just to understand the hierarchical structure of the text, and the tarka-p da topics are treated through the demands of apologetics. That future virility is consequent on the use of kh dira wood cannot really be a matter of personal experience, but if the non-temporal unity of Brahman with the Self is not personally understood, liberation cannot follow. It is for these reasons, I will propose, that Śa kara treats reasoning and personal experience as pram ṇas, not as pramiti-karaṇam or an instrument of a veridical cognition in the characteristic causal sense, but as further means that must clarify and bring to consummation the verbal Upaniṣadic cognition of identity that rises from hearing the Upaniṣads and has been given a proper shape through the first, theological inquiry. The integral inquiry, thus, serves the purpose of removing the faults that prevent certainty in the cognition to obtain. As we saw, Śa kara claimed that in the third process one would be able to reflect with certainty on what had been determined through ravaṇa and manana. Certainty, thus, is the element that the inquiry should bring to the cognition, so that the cognition could transit to pramiti, and that does not play a role in the rise of the cognition itself. We can also look at this outside the strict frame of pram ṇa in the causal sense. The possibility of the inquiry comes from the following principle: that Brahman is the Self, the 455 identity statement, is knowable solely from the Upaniṣads. To quote the BSBh 1.1.4 againŚ “Nor is Brahman, though a fully existent entity, in the domain of sensory perception or the like, because the fact that Brahman is the Self, given in the statement ‘You are that,’ is not knowable except from scripture.”29 Both Brahman and the Self, on the other hand, are known entities, the first from the etymology of the word, as we saw in the formation of the notion of Brahman, and the second because of being immediately availableŚ “For, everyone cognizes the Self and does not think ‘I am not.’ Were it not the case that there is the Self, everyone would think, ‘I am not.’”30 Not only are they known, but must be so, for otherwise the unknown relation between them would be unknowable. This is very much like the apūrva-vidhi of the Mīmāṁsakas, which relates known categories—wood and virility—in an unknown causal relationship, just that in this case such relationship must become intimately known for there to be liberation, and can become known because it concerns oneself. That the two individual categories are known makes them available to individual inquiry, and because they are known—doubts and inverse ascription are possible in their regard. But, there is a catch. If they are known as common existing things, objects, then they would not be in the domain of the Upaniṣads, just as grain and threshing are not individually knowable from the Veda. The Veda is the pram ṇa on supersensible things. They are, therefore, not really known. As we saw in the deliberation on the identity statement, because Brahman is the Self, the Self is not really known even from the notion of the Self or the idea “I am.” This notion is the reflection of the Self in the intellect, and because of its being nearest in contact with the Self, it serves as its closest approximation point, a known through which the unknown can 29 na ca pariniṣ hita-vastu-svarūpatve’pi pratyakṣādi-viṣayatvaṁ brahma aḥ, tat tvam asi iti brahmātma-bhāvasya śāstram antare ānavagamyamānatvāt. BSBh 1.1.4. 30 sarvo hy ātmāstitvaṁ pratyeti, na nāham asmi iti. yadi hi nātmāstitva-prasiddhiḥ syāt, sarvo lokaḥ nāham asmi iti pratīyāt. BSBh 1.1.1, I.8. 456 become known. The category of the Self as it is known to us is a range of identification points which are not really the Self, but are witnessed by the Self, and need to be eliminated. The inquiry can examine these by showing how they are not the Self, so that eventually the real Self that is not an object would be left as the irreducible remainder that cannot be scrapped, because without its light of awareness the inquiry itself would be impossible. When the word [Self] is used with respect to the inner Self that possesses the body and is in the realm of difference, when it is denied that the Self is the body, etc., the word creates an apprehension of the remainder, Being, though that is inexpressible in words.31 Śa kara likens such knowledge to the case when an army is seen and one knows that the king is there, though he is hidden by the parasol, the flag, the banner, etc. When everything is examined and the king is still not seen, there arises the cognition that whoever is left unseen must be the king. He must be there because it was announced that the king was coming. This is what the inquiry should accomplish: considering all the candidates for the Self, all empirically knowable, it should zero in on the Self as a non-object. Its procedure is, therefore, negational, and Śa kara uses all words for negation common in Indian philosophy to refer to it: apoha, niṣedha, pratiṣedha, apav da.32 A similar procedure also applies on the side of the tat category. One of the purposes of the manana process is to show through analogical approximation how Brahman’s being the cause is reasonable, cause not in the sense of the evolution of the world from Brahman, but in the sense of the world’s being Brahman, as the effect just is the cause. Thus, general causality is knowable, but what is not knowable without the Upaniṣads is that Brahman is satyaṁ jñ nam anantam nandam. But, because causality is knowable in general, it can be used as an illustration 31 dehavati pratyag-ātmani bheda-viṣaye prayujyamānaḥ śabdaḥ, dehādīnām ātmatve pratyākhyāyamāne yat pariśiṣ aṁ sad avācyam api pratyāyayati. ChUBh 7.1.3, VII.418. 32 Instances of such uses are omnipresent in his works, and he most commonly applies the term pratiṣedha, otherwise the Mīmāṁsā term for Vedic prohibitions. 457 of Brahman’s causal role. Eventually, when full identity is known as obtaining through the secondary signification function, even Brahman’s being the cause is no longer the case. But without appreciating causality, that last intuition is impossible, for there is otherwise no prop that one can use to relate oneself to the great ground of Being. There is such a thing as Brahman, which is eternal, pure, aware, and free in nature, omniscient and omnipotent. From the etymological analysis of the word “Brahman,” the notions of eternality, purity and the like are cognized, for such is the meaning of the root bṛmh. The existence of Brahman is also well-known because of its being the Self of all; for, everyone is always aware of one's own existence, and it is never the case that one thinks “I am not.” If there were no awareness of self-existence, everyone would think “I am not.” And the Self is Brahman. – But, if it is known in the world that Brahman is the Self, then Brahman would be known and there would be no reason for inquiry. – No, because these are the disagreements about its specific nature. Ordinary men and lok yatas think it is simply the body that is the sentient self; others say it is the senses that are aware; still others it is the mind; others that it is just momentary awareness; yet others that it is void; others claim that there is a transmigrating agent and enjoyer, over and above the body; some claim it is an enjoyer, but not an agent; yet others claim there is a lord, over and above the self, omniscient and omnipotent; and still others say that this lord is the self of the enjoyer. And so, there are many of opposite views based on logic, scripture, and their semblance. If one were to accept any of them without examination, one would be deprived of the highest good and would get the undesirable. Therefore, starting with the introduction of the inquiry into Brahman, the meaning of texts will be examined here, with reasoning not opposed to them, so that the highest good may be attained.33 Excepting the Lord, as described, no other factor of origin, maintenance, and dissolution of the word, as described – for instance, the insentient pradh na, atoms, non-Being or some transmigrating self – can be imagined. Nor can this happen spontaneously, for there is the dependency on specific place, time and causes. The proponents of the doctrine that the Lord is the cause accept only this inference for proving the existence of the lord over and above the transmigrating self. … The understanding of Brahman comes about through certainty achieved by deliberation on the Upaniṣadic statements. It does not arise 33 asti tāvad brahma nitya-śuddha-buddha-mukta-svabhāvaṁ sarvajñaṁ sarva-śakti-samanvitam. brahma-śabdasya hi vyutpādyamānasya nitya-śuddhatvādayo ’rthāḥ pratīyante, b ṁhater dhātor arthānugamāt. sarvasyātmatvāc ca brahmāstitva-prasiddhiḥ. sarvo hy ātmāstitvaṁ pratyeti, na nāham asmi iti. yadi hi nātmāstitva-prasiddhiḥ syāt, sarvo lokaḥ nāham asmi iti pratīyāt. ātmā ca brahma. yadi tarhi loke brahma ātmatvena prasiddham asti, tato jñātam evety ajijñāsyatvaṁ punar āpannam. na. tad-viśeṣaṁ prati vipratipatteḥ. deha-mātraṁ caitanya-viśiṣ āmātmeti prāk tā janā laukāyatikāś ca pratipannāḥ. indriyā y eva cetanāny ātmety apare, mana ity anye. vijñāna-mātraṁ kṣa ikam ity eke. śūnyam ity apare. asti dehādi-vyatiriktaḥ saṁsārī kartā bhoktety apare. bhoktaiva kevalaṁ na kartety eke. asti tad-vyatirikta īśvaraḥ sarvajñaḥ sarva-śaktir iti kecit. ātmā sa bhoktur ity apare. evaṁ bahavo vipratipannā yukti-vākya-tad-ābhāsa-samāśrayāḥ santaḥ. tatrāvicārya yat kiṁcit pratipadyamāno niḥśreyasāt pratihanyeta, anarthaṁ ceyāt. tasmād brahma-jijñāsopanyāsa-mukhena vedānta-vākya-mīmāṁsā tad-avirodhi- tarkopakara ā niḥśreyasa-prayojanā prastūyate. BSBh 1.1.1, I.8-9. 458  through another pram ṇa, such as inference. Since it is the Upaniṣadic statements that present Brahman as the cause of the creation etc. of the world, this inference that is employed here – not contrary to the Upaniṣadic statements – is also a pram ṇa for reinforcing the understanding of the texts, and it is not excluded; for, the Upaniṣads themselves accept such reasoning as help.34 Wilhelm Halbfass had discussed the issue of reason and revelation in Śa kara’s Vedānta at some length, and he showed that Śa kara recognized a form of scriptural, guided reasoning and inference-making that is employed by the Upaniṣads themselves.35 For our purpose of understanding manana, this means that the Upaniṣads give a model of the use of reasoning that a human teacher can emulate in the process of instruction. ravaṇa and Manana in the Upade a-S hasr As it is well-known, the three chapters of the prose part of Śa kara’s US were written with the purpose of illustrating the three Vedāntic methods of liberation.36 Let us look now at the first chapter, conceived as a manual for conducting ravaṇa, and see in practice what the goal of ravaṇa was.37 The chapter opens with the qualifications of the student that would have been achieved through the desire for knowing the Self, vividiṣ : dispassion, and the possession of the required virtues. We learn that the student who had approached the teacher for instruction is a renunciant: the adhik ra text, MU 1.2.12-13, is explicitly mentioned—par kṣya lok n etc.—but 34 na ca yathokta-viśeṣa asya jagato yathokta-viśeṣa am īśvaraṁ muktvā, anyataḥ pradhānād acetanāt a ubhyo vā abhāvād vā saṁsāri o vā utpatty-ādi sambhāvayituṁ śakyam. na ca svabhāvataḥ, viśiṣ ādeśa-kāla-nimittānām ihopādānāt. etad evānumānaṁ saṁsāri-vyatirikteśvarāstitvādi-sādhanaṁ manyante īśvara-kāra a-vādinaḥ. ... vākyārtha-vicāra ādhyavasāna-nirv ttā hi brahmāvagatiḥ, nānumānādi-pramā āntara-nirv ttā. satsu tu vedānta- vākyeṣu jagato janmādi-kāra a-vādiṣu, tad-artha-graha a-dārḍhyāya anumānam api vedānta-vākyāvirodhi pramā aṁ bhavati, na nivāryate, śrutyaiva ca sahāyatvena tarkasyāpy abhyupetatvāt. BSBh 1.1.2, I.10-11. 35 Halbfass 1991:131-204. 36 “The three prakaraṇas of the Prose Part can, in content, be regarded as illustrating respectively the stage of hearing ( ravaṇa), the stage of thinking (manana) and the stage of meditation (nididhy sana), which constitute the three Ved ntic stages to attainment of final release (mokṣa).” Mayeda 2006aŚ66. 37 atha mokṣa-sādhanopadeśa-vidhiṁ vyākhyāsyāmaḥ. USG 1.1. The chapter is sometimes called iṣy nu sanam. See, for instance, the 2003 edition in the Sarasvatī-bhavana-grantha-mālā series, edited by Rama Kiśora Tripā hī. 459 BĀU 4.2.22-3 are obviously relevant as well.38 The teacher should first examine the student and see if he is indeed qualified, that is, if he has the required virtues and the preliminary grasp of transient and eternal things. If the opposite proves to be the case, he should send the student to remedial lessons in dispassion and the requisite virtues.39 He should then instruct the student in the body of texts that we identified as constituting the para-vidy , propounding that there is only one Self, and proceed to making the student grasp the characteristics of Brahman. These include the positive and negative characteristics that we saw in the definition of the tat-pad rtha, and there seems to be nothing different in the list of texts from the one in BSBh 4.1.2, except that they are not in the same order: here the list begins and ends with negative texts, and vijñ nam nandam and satyaṁ jñ nam anantaṁ brahma are stuck in the middle.40 This very obviously refers to establishing the reference of the tat-pad rtha, and is further corroborated with passages from the Bhagavad-G t that are “meant to propound that the Supreme Self is not liable to transmigration and that it is non-different from everything.”41 The link of tat with the individual Self is clearly established. And when that is done, if the student had grasped these characteristics of the Supreme Self, the teacher should ask him, “Who are you, my dear,” shifting the instruction to the tvaṁ- pad rtha. The goal of such instruction is to make the student understand his identity with Brahman by removing all that is incompatible with the notion of tat from his understanding of himself. Part of the strategy in doing this is to remind the student occasionally of the characteristics of the tat-pad rtha, but in such a way that the relation between the two, tat and 38 USG 1.2-3. 39 Ibid., 4-5. 40 Ibid., 6-7. 41 sm tibhiś ca ... ity-ādibhiḥ śruty-ukta-lakṣa āviruddhābhiḥ paramātmāsaṁsāritva-pratipādana-parābhis tasya sarve ānanyatva-pratipādana-parābhiś ca. USP 1.8. 460 tvam, becomes obvious. This is accomplished, first, by the description of creation through the process of quintiplication, pañc -karaṇa, or the evolution of name and form into the world and its minutest details, down to caste and social identity, and then by description of Brahman entering its own creation as the Self, in the manner that we saw in the Taittir ya Upaniṣad. The point of this description is to show how the student is different from his body, caste, and family identity, but also from his current rama position of a renunciant, and that he should rather identify with the inner Self that had entered creation after it evolved from the Self.42 At a point during this instruction, it has gotten across to the student that the identity statement as the teacher had explained it literally means that Brahman is his own self, himself, but such identity appears impossible to the studentŚ “But, we are different. I am ignorant, happy and miserable, bound and transmigrating. The Lord is different from me, not liable to transmigration.”43 The student presents an account of duality in which liberation is achieved by the performance of ritual for the pleasure of the Lord. There follows a new round of teaching in the identity statements of the Upaniṣads with the purpose of removing from the category of tvam the notion of being a ritual agent and, with that, of being liable to saṁs ra. The instruction has moved to this distinction, past bodily and social identification and to the liability to transmigration, and to what that might mean for the difference or non-difference of the two categories. This distinction is prompted by the fact that the student’s being liable to pain, hunger, thirst, and the like—the paradigmatic characteristics of saṁs ra—are perceptually evident to him, whereas scriptures say that the Supreme Self is not liable to them. It appears, therefore, that this is the root distinction between the two categories: the one is not liable to saṁs ra and the second is, from which follows that the Self is essentially an agent. In other 42 Ibid., 9-24. 43 anya evāham ajñaḥ sukhī duḥkhī baddhaḥ saṁsārī, anyo 'sau mad-vilakṣa o 'saṁsārī devaḥ. Ibid., 25. 461 words, the negative characteristics of the Supreme Self from the Upaniṣads that the teacher posited in the definition of tat and related to the Self are evidently not applicable to the Self as it is known to the student.44 The corollary of this is that the Upaniṣads intend to affirm difference between the Supreme and the transmigrating Self, and affirm ritual as the means both of prosperity and liberation. Here begins, then, an instruction that continues till the end of the chapter and aims to show that the Self is not the ritual agent. This is, clearly, the ultimate purpose of the ravaṇa process: show that the Upaniṣads do intend to affirm that the Self is not an agent that is liable to transmigration by maintaining that the Self is, indeed, Brahman. This is done by ascertaining the proper reference of the idea of the Self that, in the student’s understanding, undergoes transmigration because of evidently experiencing pain. The procedure is to distinguish the Self as the pure subject or witness of different states of propositional consciousness, all of which can be understood as objects and which include the sensations of pain, the psychological torments such as attachment and aversion, the objects that incite them, desire more generally etc. They are all located in the intellect: they are properties of buddhi, not of the Self.45 The Self is again affirmed as different from them through yet another barrage of Upaniṣadic quotations that consist in identity statements and negative descriptions of the Self.46 The distinction of the Self as the subject from the objective part of its states of propositional consciousness is the only significant excursion in more philosophical reasoning in this chapter, but the point is not really to indulge in reasoning pure and simple. Rather, it is an instance of application of the anvaya-vyatireka method, which was a common method with 44 Ibid., 33. 45 Ibid., 36. 46 Ibid., 37-8. 462 Indian thinkers in general, one of whose purposes was to ascertain words and word meanings by examining a complex and finding the element that persists through change. Grammarians used it, for instance, to establish uninflected nominal stems or pr tipadikas on the one hand and affixes on the other, by observing the concurrent occurrence and non-occurrence of elements in real, inflected, words, and by associating meaning with them. For instance, in the inflected words vṛkṣas and vṛkṣau, the element that occurs in both cases is vṛkṣa, whereas s and au differ. The two, however, may persist in other circumstances: for instance, au is concurrently occurrent in vṛkṣau and a vau, whereas vṛkṣa and a va are concurrently non-occurrent. This allows for associating vṛkṣa with “tree” or “something having roots, fruits, leaves etc.” as its meaning, and s and au with singularity and duality, respectively.47 Śa kara’s specific use of the method consisted in examining what idea persists when we use the personal pronouns, aham and tvam, and what changes. In any propositional consciousness of the “I am x” or “I see x” kind, the objective variable is different, but the subject of consciousness remains the same. Since that is the case, the subject of consciousness is the meaning of the personal pronoun, and whatever can become its object, including one’s most intimate thoughts, cannot refer to the Self, because it is the element that is different in each case. This method should be used both to ascertain this meaning and to bring it to mind when construing the identity statement.48 So, the teacher wants to tell the student that if he understands 47 See Deshpande 2007:7-9; Cardona 1967-8. 48 Śa kara elsewhere explicitly states that the anvaya-vyatireka method is used for ascertaining words and word meanings, very much in line with the grammarians (see USP 18.96, 176-180, 189). A typically lucid account of anvaya-vyatireka in early Advaita Vedānta is available in Cardona 1981. Mayeda in his introduction to the translation of US claims that the anvaya-vyatireka method was used by Śa kara to establish the meaning of the identity statement as a sentence, where what needs to drop from the meaning of tvam drops because of incompatibility with tat. Apparently, this method was later replaced by Sarvajñātman with jahad-ajahal-lakṣaṇ (Mayeda 2006b:53-8). Mayeda is most certainly wrong in claiming this. In USP 18.96, Śa kara explicitly says that anvaya-vyatireka applies to words and word meanings. The identity sentence meaning was established through s m ṇ dhikaraṇya or vi eṣa-vi eṣaṇa-bh va, where the single reference of the word obtains through secondary 463 what the notion of the Self refers to—and it must refer to that which persists in all cases when we use the word “Self”—suffering and the other states of saṁs ra could not belong to the Self, because they are invariantly the predicate in the propositional consciousness. The upshot of this is the realization that the Upaniṣads do intend to affirm the Self- Brahman identity literally, and not as a case of identification that would uphold the continuation of ritual and the perpetuation of the ritual Self and its accessories, such as the sacred thread. This is the big gain of the ravaṇa process. “From all these statements, it is proven that you are the sole Self, the Supreme Brahman, free from all the characteristics of transmigration.”49 In summary, ravaṇa as presented in the USG 1 should introduce the categories of tat and tvam and bring the student to understanding that the para-vidy passages of the Upaniṣads do not present the Self as a transmigrating ritual agent essentially different from Brahman; that is, they intend the identity literally. Next, ravaṇa, having first defined the notion of Brahman, should fix the reference of the tvaṁ-pad rtha, such that it becomes obvious that the two have the same reference. ravaṇa is, in other words, a heads-on, “classroom” exegesis, to be repeated until it is clear to the student that scripture does intend to affirm full identity. The US manual is a blueprint of how such pedagogical exegesis might proceed, and it surely provides its structure, but it would be modulated depending on what the student had already understood, the difficulty in understanding a certain point etc. Its form of reasoning is strictly theological, and the only significant excursion in philosophy consists in distinguishing the subject and object in cases of propositional consciousness where the Self is involved, but even that is just for establishing the reference of the personal pronouns that the Upaniṣad identifies with Brahman. signification. Cardona is, therefore, right in claiming against Mayeda that anvaya-vyatireka was not a method of meditation for Śa kara, but a general mode of reasoning. 49 ity-ādi-śrutibhyaḥ, sm tibhyaś ca ... ity-ādibhya eka evātmā paraṁ brahma sarva-saṁsāra-dharma-vinirmuktas tvam iti siddham. USG 1.37-8. 464  I do not wish to go into the details of the second prose chapter that discusses manana, because most of it would be a repetition of what I have already said. I need to add however, a few notes. The general theme of the chapter concerns determining the real nature of the Self: is it essentially an agent of action and experience, or is such agency accidental to it, in which case it would be the Self pure, one for all. This is prompted by the same problem that we saw in the first chapter: saṁs ra consisting in suffering painful sensations in the waking and dreaming states is a matter of presentational consciousness. Śa kara’s arguments follow the line of independence of consciousness that cannot be identified with its objective counterparts, nor can it participate in compositions of any kind: The Self is essentially different from the objects that are forms of painful sensations the marks of saṁs ra, and it is absolutely aloof from the body, mind, and senses. In other words, the chapter picks up where the previous chapter had left off, and its purpose is to tackle the doubts that the student will inevitably have in the possibility of the conclusion that was arrived at through the process of ravaṇa. Although this is a case of strictly philosophical reasoning, it is still permeated by the Upaniṣadic theme of the three states of awareness: waking, dream, and deep sleep. There are two key moments in this second process, really a variation of a single theme, and they concern language and cognition. First, if the Self is cognition in nature, as the teacher claims, then it cannot be permanently changeless, kūṭastha, as the teacher also claims, because cognition is verbal action, and action is change. The teacher’s reply is that the word “cognition” or upalabdhi is only figuratively applied to the Self, in virtue of the fact that cognition as a result is a reflection of the Self whose nature is cognition pure and unchanging, like the illumination of the sun. Second, because the Self is consciousness or cognition in nature, it is also understanding 465 or veracity in nature, avagati or pram . All veridical cognitions are, in fact, possible because of the light of the Self; they all terminate in the Self, and the Self is indicated through them: One may object to this: it is contradictory to say that understanding is the result of a reliable warrant, and that it is the eternally changeless light of the Self in nature. To the student saying this, the teacher replied: It is not contradictory. – How is it, then, the result of understanding? – It is so figuratively. Even though it is permanently changeless, it is seen in the results of cognitions that obtain through perception and the like, since it is their [ultimate] reference. Because such perceptual cognitions are impermanent, understanding also appears as if impermanent. For this reason, it is figuratively designated as the result of reliable warrants.50 The Self is, really, self-established in nature—it is what consciousness is—and it requires no proof by means of a pram ṇa. But, because such is the case, all veridical cognitions, though they are impermanent, are, really, indications of the Self that is unchanging understanding in nature. In its truest sense, pram or avagati is just the permanently changeless Self. I single out these two moments as crucial for the following reason. We saw in the previous chapter that understanding the identity statement was predicated on understanding that Brahman’s being the Self as knowledge or jñ nam, the crucial link between the two categories, obtains through the secondary signification function of the word. Without understanding this, as well as understanding what jñ nam ultimately refers to, full understanding of the identity statement is impossible, and this point is not discussed in the ravaṇa section, because it is not a theological topic. As a corollary of this, it follows that liberation is impossible without going the full length of manana, without stra-yukti together: since understanding the identity statement is the Vedāntic process of liberation, proper understanding is predicated on grasping the nuts and bolts of philosophy of language and mind. 50 tatrāha codakaḥ—avagatiḥ pramā ānāṁ phalaṁ kū astha-nityātma-jyotiḥ-svarūpeti ca vipratiṣiddham. ity uktavantam āha—na vipratiṣiddham. kathaṁ tarhy avagateḥ phalatvam? tattvopacārāt. kū asthā nityāpi satī pratyakṣādi-pratyayānte lakṣyate, tādarthyāt. pratyakṣādi-pratyayasya anityatve anityeva bhavati. tena pramā ānāṁ phalam ity upacaryate. USG 2, 108. 466  This conclusion is reinforced by the following point: we also saw that brahma-jñ na or tat tvam asy-adi-v ky rtha-parijñ na was supposed to culminate in “personal experience,” anubhava, or “understanding,” avagati. Avagati is the dominant theme in this second chapter—it is sometimes called just the avagati-prakaraṇam51—but we learn here that understanding is what the Self is: understanding the identity statement just is “becoming” the Self, because the Self is indicated by all veridical cognitions as the consciousness without which veracity is impossible, and which is veridical as consciousness even in cognitions that are not veridical.52 When full understanding obtains, one reaches the level of the Self, which is the absolute or param rtha perspective in which only consciousness is real, but not its content. The content does not necessarily disappear upon such understanding, but one knows from this perspective that what is seen in waking is as false as the content of dreams from the standpoint of waking. One has broken new ground, and it is, at last, the firm ground. There is no “confirmation” of the scriptural knowledge in personal experience, but personal experience that is understanding that is the Self itself, when the meaning of the identity statement had been fully clarified through the theological reflection, when it had been understood that in the identity statement the sentence reference obtains through the secondary signification function, and that one’s propositional awareness is but an indication of Brahman the light of consciousness. We return, thus, to the point with which we began in Chapter Seven: understanding the identity statement without the last shred of doubt is all that is required for liberation, for such understanding just is the Self. This understanding removes ignorance, as a consequence of which one remains the Self as the final step in the attainment of liberation. There is, in other words, no special intuition of brahma-jñ na, no mystical experience, no further “direct vision of 51 Such is the case both in Mayeda’s critical edition and in Jagadananda Swami’s translation (1949). 52 BSBh 2.1.14. 467 Brahman,” no second, non-scriptural cognition beyond the full clarification of doubts, just an intellectual understanding with full certainty, predicated on accepting the figurative nature of words that stand for consciousness, and of cognitions as mere indicators of “understanding.” It is arriving at avagati through stra-yukti over the plenitude of scriptural data. We can, in fact, pinpoint the exact moment when liberation takes place. After the last quoted statement, when the student realizes that all mental states are just reflections of the light of the Self, veridical as the Self but false as their individual content, he says the following, which is worthy of quoting in full: If that is the case, sir, then understanding (avagati) is permanently changeless, just the light of the Self in nature, known through itself because of not depending on a pram ṇa with regard to itself. Everything else is unconscious and for the purpose of another, being made to act through combination [with other things]. The non-Self exists for the sake of another just in virtue of the nature of understanding that appears as the cognitions that give rise to happiness, suffering, illusion, etc. Consequently, from the absolute standpoint (param rthataḥ), it does not exist at all. Just as it is seen in the world that the snake does not exist in the rope nor water in mirage except through awareness (avagati) of them, it is likewise but right that the dualities of waking and dream do not exist except through awareness of them. In the same way, sir, from the absolute perspective, the understanding that is the light of the Self is compact, and therefore it is permanently changeless and nondual, being present in all cognitions, whereas cognitions are not present in understanding itself. Just as various cognitions of things blue, yellow, etc., are not what understanding is in the dream state, likewise they are not so in the waking state. And, since cognitions of things blue, yellow, etc., are not what this understanding is, they ought to be unreal, and except for their understanding there is no separate agent of understanding, for which reason by its own nature it can neither be accepted nor rejected, given that there is no other thing.53 53 yady evaṁ bhagavan, kū astha-nityāvagatiḥ ātma-jyotiḥ-svarūpaiva svayaṁ-siddhā, ātmani pramā a- nirapekṣatvāt, tato ’nyat acetanaṁ saṁhatya-kāritvāt parārtham. yena ca sukha-duḥkha-moha-hetu-pratyayāvagati- rūpe a pārārthyam, tenaiva svarūpe a anātmanaḥ astitvam nānyena rūpāntare a. ato nāstitvam eva paramārthataḥ. yathā hi loke rajju-sarpa-marīcyūdakādīnāṁ tad-avagati-vyatireke a abhāvo d ṣ aḥ, evaṁ jāgrat-svapna-dvaita- bhāvasyāpi tad-avagati-vyatireke a abhāvo yuktaḥ. evam eva paramārthataḥ bhagavan, avagate ātma-jyotiṣaḥ nairantarya-bhāvāt kū astha-nityatā advaita-bhāvaś ca, sarva-pratyaya-bhedeṣu avyabhicārāt. pratyaya-bhedāś ca avagatiṁ vyabhicaranti. yathā svapne nīla-pītādy-ākāra-bheda-rūpāḥ pratyayāḥ tad-avagatiṁ vyabhicarantaḥ paramārthato na santīty ucyate, evaṁ jāgraty api. nīla-pītādi-pratyaya-bhedāḥ tām evāvagatiṁ vyabhicarantaḥ asatya-rūpāḥ bhavitum arhanti tasyāś ca avagater anya avagantā nāstīti na svena svarūpe a svayam upādātuṁ hātuṁ vā śakyate, anyasya ca abhāvāt. Ibid., 109. 468 The teacher approves of this, confirms that the student has gotten rid of ignorance by the power of knowledge, and is now liberatedŚ “You have thus attained fearlessness, and from now on you will not experience the suffering of waking and dream. You are freed from the misery of transmigration.”54 The student ascents with an Om. Needless to say, the mental state corresponding to the statement of liberation is a token of a mental state, just as those of the dream and waking states. It is, nevertheless, a shift in perspective, from vyavah ra to param rtha, where the world is still seen but “under the mode of mirage.” And, a mental state is all that it can be under the circumstances. It is, also, all that is required in Śa kara’s eyes, but we will talk about it after we deal with nididhy sana. Nididhy sana and Parisa khy na The last of the three processes, nididhy sana, had puzzled scholars for some time. As we saw in Chapter Five, in pre-Śa kara Advaita Vedānta this third process stood for meditation on Brahman that developed under the influence of the school of Yoga and was generally called prasaṅkhy na or dhy n bhy sa in Ma ḍana Miśra’s turn of phrase. Śa kara rejected the possibility of meditation on Brahman after Brahman had been known, and otherwise meditation had applicability in the saguṇa-vidy portion of the Upaniṣads, culminating in the arising of knowledge and in renunciation. There could not be meditation on Brahman as depicted in the nirguṇa-vidy passages. We also saw that the first two processes, ravaṇa and manana, could jointly culminate in full understanding and liberation. Yet, Śa kara affirmed this third process and, curiously, called it parisaṅkhy na in the third prose section of the US. In other words, he simply changed the preverb from pra to pari. What is this parisaṅkhy na and why call it that? 54 ity evaṁ tvam abhayaṁ prāpnoṣi, nātaḥ paraṁ jāgrat-svapna-duḥkham anubhaviṣyasi. saṁsāra-duḥkhān mukto ’sīti. Ibid., 110. 469  Mayeda approvingly refers to Paul Hacker’s translation of the word as “Rekapitulierende Betrachtung,” and says that parisaṅkhy na seems to be “a kind of meditation which consists in recapitulating the conclusion which had been arrived at through one’s previous study and discussion with a teacher,” pointing to Śa kara’s rejection of prasaṅkhy na but admitting that it is not clear how parisaṅkhy na was any different.55 In an important little monograph on meditation in Śa kara’s Vedānta, Jonathan Bader suggests, rightly, that Śa kara’s choice of the term parisaṅkhy na could have been influenced by the Mīmāṁsā notion of parisaṅkhy -vidhi.56 I had come to the same conclusion independently, before I had the occasion to read Bader’s monograph, and here I would like to expand upon what is just a hint in it. Prasaṅkhy na-v dins related meditation on the Self to injunctions such as tmety evop s ta, vijñ ya prajñ ṁ kurv ta and so ‘nveṣṭavyaḥ sa vijijñ sitavyaḥ. That is, the injunctions of meditation picked specifically the nididhy sitavya part of Yājñavalkya’s statement. It is quite instructive to see what Śa kara had to say about these injunctive statements. Under BĀU 1.4.7, Śa kara rejects what appears to be an adaptation of the Bhā a idea of bh van to meditation, the idea that the statement tmety evop s ta could be an apūrva-vidhi, an injunction that introduces meditation on the Self as a form of action that is a means to knowing the Self, but is itself not known as such a means without the injunction, as I discussed in the Second Chapter. The reason for his rejection is that there obtains optionality regarding the Self, which is evident from later passages in the Upaniṣad. Think, for instance, of Uṣasta Cākrāya a’s pestering of Yājñavalkya to tell him about Brahman that is direct and immediate. “Which one is it, Yājñavalkya, that is the Self within all?”57 There are so many points of identification of the Self, and one would have to 55 Mayeda 2006b:254. 56 Bader 1990:78-9. 57 atha hainam uṣastaś cākrāya aḥ papraccha. yat sākṣād aparokṣād brahma ya ātmā sarvāntaras taṁ me vyācākṣasveti. eṣa ta ātmā sarvāntaraḥ. katamo yājñavalkya sarvāntaraḥ. BĀU 3.4.1. 470 know which Self should be the object of meditation first, before the injunction could make sense as an apūrva-vidhi. Once one had ascertained the real Self, however, all other points of identification vanish because they are ignorance, and by the principle of residue only the option of thinking of the true Self remains, which makes the injunction dead and gone because one no longer sees oneself as an agent etc.58 Śa kara, however, says that the statement can be taken as an injunction, not of the apūrva, but of the niyama kind, a restriction. To refer again to our second, “resource” chapter, a restriction was an injunction that obtains when options are available, for instance in preparing the rice, and only one of them is enjoined. What Śa kara intends to say is that one should meditate on the true Self and not on any other point of identification available, for instance, the mind. The glaring problem with this is, as the pūrva-pakṣin is quick to point out, that Śa kara had just said that thinking on the true Self was inevitable by the principle of residue. Śa kara’s reply to this objection has some consequences both for understanding the role of nididhy sana, and the nature of liberation: But, how is it the case that there is optionality of meditation, when you said that the mental flow over the true Self would be permanent by the principle of residue? – That is right. Nevertheless, even though it is so, the results of actions that produced the body must bear fruit. And, even though full understanding had arisen, the functions of the body, mind and words are inevitable because the karma that had begun unfolding overpowers knowledge, like the case of the released arrow. For this reason, the functioning of knowledge is weaker and optionality obtains. Therefore, the mental flow of knowledge of the Self must be regulated by relying on methods such as renunciation, dispassion, etc.59 58 ātmety evopāsīta iti nāpūrva-vidhiḥ, pakṣe prāptatvāt. yat s kṣ d aparokṣ d brahma (BĀU 3.4.1), katama tmeti — yo 'yaṁ vijñ namayaḥ (BĀU 4.3.7) ity-evam-ādy-ātma-pratipādana-parābhiḥ śrutibhir ātma-viṣayaṁ vijñānam utpāditamś tatrātma-svarūpa-vijñānenaiva tad-viṣayānātmābhimāna-buddhiḥ kārakādi-kriyā-phalādhyāropa ātmikā avidyā nivartitāś tasyāṁ nivartitāyāṁ kāmādi-doṣānupapatteḥ anātmacintānupapattiḥ; pāriśeṣyād ātma-cintaiva. tasmāt tad-upāsanam asmin pakṣe na vidhātavyam, prāptatvāt. BĀUBh 1.4.7, VIII.106. 59 kathaṁ punar upāsanasya pakṣa-prāptiḥ, yāvatā pāriśeṣyād ātma-vijñāna-sm ti-santatiḥ nityaivety abhihitam? bāḍham, —yady apy evam, śarīrārambhakasya karma o niyata-phalatvāt, samyag-jñāna-prāptāv apy avaśyaṁ- bhāvinī prav ttir vā -manaḥ-kāyānām, labdha-v tteḥ karma o balīyastvāt—mukteṣv-ādi-prav ttivat; tena pakṣe prāptaṁ jñāna-prav tti-daurbalyam. tasmāt tyāga-vairāgyādi-sādhana-balāvalambena ātma-vijñāna-sm ti-santatir niyantavyā bhavati. BĀUBh 1.4.7, VIII.113. 471 I will discuss the consequences of this on the nature of liberation and knowledge of the Self in the next section, but let us note that even after full understanding it was possible to swerve and think on other things by the force of the already constituted conditioning. Because of this, it is required to focus on the true Self. Here I should point out that Śa kara referred to vijñ ya prajñ ṁ kurv ta of BĀU 4.4.21 as a statement of the same kind just before the quoted passage. He says the same thing about the third important statement commonly adduced as an injunction to meditation, so ‘nveṣṭavyaḥ sa vijijñ sitavya of ChU 8.7.1—it is not an apūrva-vidhi but niyama—emphasizing, however, the other important feature of apūrvaŚ it cannot be an injunction of such a kind because investigating and knowing have visible results, understanding, whereas injunctions of the apūrva kind are invariably related to invisible results, such as the presumed excellence in rice as a result of besprinkling, or the virility that comes about from using a post of kh dira wood.60 Third time’s a charm, and although even Sureśvara accepts this endorsement of niyama-vidhi begrudgingly or as hypothetical, there is no doubt that Śa kara was serious about it.61 Now, all advocates of prasaṅkhy na wanted ritual and the rama practices given in BĀU 4.4.22, as well as the six virtues in BĀU 4.4.23 that were identified with yama-niyama, to be either enjoined or practiced all the same alongside meditation: both these texts came just after the vijñ ya prajñ ṁ kurv ta. Śa kara rejected the first, as we saw, because the rama practices were related to the “desire to know the Self,” not to knowledge itself, and were “comparatively remote.” They culminated in understanding transmigration and engendering dispassion. The six virtues or the complex of yama-niyama, on the other hand, although having the status of 60 anveṣ avyaḥ vijijñāsitavya iti ca eṣa niyama-vidhir eva, na apūrva-vidhiḥ. evam anveṣtavyo vijijñāsitavya ity arthaḥ, d ṣ ārthatvād anveṣa a-vijijñāsanayoḥ. ChUBh 8.7.1, VII.507. 61 NaiS 1.88, and BĀUBhV 1.4.921-932. 472 practices of the rama of those who had gone forth, the monks, had to be observed not only after renunciation, but even after knowing Brahman, because it was possible for one to be distracted by hunger and similar natural discomforts. These were not only renunciation in nature—knowing that presupposes doing nothing productive—but also virtues that one had to cultivate even after full knowledge of Brahman had arisen, because such knowledge had to be guarded against the overpowering tendencies of embodiment. Consider, for instance, the following passage from ChUBh 2.23.1, where Śa kara argues at length that the monk who is established in Brahman, brahma-saṁstha, cannot have an obligation to perform ritual and the other rama practices because he has obtained the cognition of unity: – But, the rules and regulations of a monk also become impossible because the cognition of duality that is the ground for injunctions has been defeated by the cognition occasioned through the texts that propound seeing unity. – No, they are for the purpose of restraint, because it is possible for one to be diverted from the cognition of unity by hunger and the like.62 Such virtues were not only conducive to the rise of knowledge equally or better than the rama practices, but were, as Śa kara said, “related to knowledge itself,” indispensable. The statements of meditation were restrictions, niyama-vidhi, in this other important sense: they involved yama- niyama as auxiliary to meditation. Śa kara, of course, did not think that the statements he designated as niyama-vidhi exercised any commanding power. Under BSBh 3.2.21, he says that injunctive statements within the scope of para-vidy s are used for directing the attention, much as in ordinary language when 62 ekatva-pratyaya-vidhi-janitena pratyayena vidhi-nimitta-bheda-pratyayasyopamarditatvād yama-niyamādy- anupapattiḥ parivrājakasyeti cet, na, bubhukṣādinaikatva-pratyayāt prācyāvitasyopapatteḥ, niv tty-arthatvāt. ChUBh 2.23.1, VI.127. 473 one says, “see this” or “hear this.” This is required because without proper attention one may not form a cognition even when facing the object, so it is proper to direct the attention.63 Let us translate this now to the case of nididhy sana, the third method of knowing the Self. The question we must ask is, how does one restrict one’s mental flow on the Self when the Self is, as Śa kara repeatedly says, a “non-object,” and the very notion of unity is canceled by affirming such a mental flow that involves a subject, an object, an instrument, and an action? Now, I said in the Second Chapter that the third type of injunction according to manner was parisaṅkhy , exclusion, and that its essential characteristic was affirming one thing but intending to deny another. To use a more down to earth example, saying “read a book” would be such an exclusion, if two conditions obtain: first, the intended meaning is not what is explicitly said, but something like “do not watch a movie,” “do not play a video game,” etc.ś second, the intended counterpart or set must be known in full. In other words, “read a book” must exclude something specific and not just anything in general. Nididhy sana in Śa kara’s system, just like ravaṇa and manana, was meant for understanding the identity of the two pad rthas, its central text was tat tvam asi, and we already saw that knowing this identity was not knowing an object that one could keep one’s mind on. It was knowing a “non-object,” or what the reference behind the statement was not. Śa kara went as far as saying that the meaning that obtains when tat and tvam were identified was neti neti. Looking, now, at the third prose chapter of the US, we see something that looks very much like such a case of tat tvam asi unfolding. The parisaṅkhy na reflection sets the familiar 63 Mayeda (1998) fails to appreciate this when he rejects Bader’s suggestion that Śa kara took a cue from the parisaṅkhy -vidhi of the Mīmāṁsakas, because “it is no doubt an ‘injunction’ as the term vidhi clearly shows.” Aside not bearing in mind the general rethinking of the nature of injunctions as informative rather than commanding, Śa kara commonly uses the term vidhi for any kind of a positive statements, including the propositional statements of the Self-Brahman unity, for instance in the text from the ChUBh 2.23.1 that I just quoted. 474 principle of distinction between the knower and its objects, and then proceeds to enumerate the categories that may become such objects. This is the general principle of the procedureŚ “That by which the objects are known is different from them, being their knower.”64 The objects are, first, the sense objects—sound, touch, form, taste, and smell—and because they are perceived, they do not constitute the Self and should have no influence over the Self. The sense objects undergo the many kinds of transformation and cause the experience of happiness, suffering and the like, but their knowers are different from them. This is an important intuition, because it opens the avenue for the knower to disassociate himself from any of these experience-causing transformations. Consider the case of praise: it comes through words, and therefore it can be reduced to words as an instance of their transformation; words are sound perceived by the ear, and the agent of knowing is different from the object; ergo, praise has nothing to do with him. The same procedure can be applied to words that express falsehood, horror, humiliation etc. The tactile sensations such as heat and cold are likewise analyzed as transformations of touch that are different from the knower, and so are all kinds of form, taste, and smell. This reflection, thus, picks those Upaniṣadic texts that dissociate the Self from the objects of the senses, and Śa kara quotes the Kaṭha Upaniṣad 3.15Ś “It has no sound or touch, no form. It is free from decay, without tasteś eternal and without smell.”65 These are, of course, the negative characteristics of Brahman that relate the categories of tat and tvam. The sense objects, further, before they are experienced as a specific cognitive instance of happiness, suffering and the like, require the joint functioning of the cognitive faculties, the body where they are seated, the mind and the intellect. As perception, they are a composite product. The Self, on the other hand, does not participate in composition in general, and whatever may be 64 yena ca jñāyante, sa vijñāt tvāt ataj-jātīyaḥ. USG 113. 65 aśabdam asparśam arūpam avyayaṁ tathārasaṁ nityam agandhavac ca yat. Ibid. 475 presented to it in the manner of unfolding of karma is unrelated to it. Another set of negative Upaniṣadic texts affirm this non-relational status of the Self, before Śa kara states his knockdown argumentŚ “The most important reason is that things that are the non-Self do not exist.”66 The intention of the parisaṅkhy na reflection, in other words, is not to affirm “You are Brahman” as an object of meditation, but to serve the practical purpose of reminding the knower of Brahman what he is not. Tat tvam asi in the context of meditation should provide a procedure for the aspirant after liberation, when he is alone, without a teacher, to rememberŚ “Whatever you may identify with, no!” This is, in fact, the only way one can reflect on the Self or perform tm nucintanam, because the prasaṅkhy na procedure is both unproductive, doing what has already been done or piṣṭa-piṣṭi, grinding what is ground,67 but also further affirms ignorance by reintroducing the subject-object-instrument-action complex where it is, strictly, impossible. Sureśvara says a bit indiscriminately that the injunction of meditation can be taken either as niyama or parisaṅkhy : – But, from the apūrva injunctions ‘One should meditate on it as the Self,’ ‘The Self, honey, should be seen,’ we ascertain a mandate for the action of seeing the Self on the part of man. – No, we don’t, because the knowledge of things as they are, being the cause of liberation by removing the ignorance of the Self that is the seed of all bad things, is not humanly contingent. Even if we accept it as an injunction, it is not of the apūrva kind. The injunctive meaning can be either that of a restriction or an exclusion, because we meditate on the supreme Self solely through excluding the vision of the non-Self.68 It seems to me, however, that Śa kara saw a significant difference between the two. The injunction expressed the purpose of restricting the thought from going to the external objects 66 anātma-vastunaś cāsattvaṁ paramo hetuḥ. Ibid. 67 Ānandagiri on BĀUBhV 4.4.798. 68 nanu tmety evop s ta, tm v are draṣṭavyaḥ ity apūrva-vidhi-śruteḥ puruṣasyātma-darśana-kriyāyāṁ niyogo ‘vasīyata iti. naivam. apuruṣa-tantratvād vastu-yāthātmya-jñānasya sakalānartha-bījātmānavabodhotsāri o mukti- hetor iti. vidhy-abhyupagame ‘pi nāpūrva-vidhir ayam. ata āha — niyamaḥ parisa khyā vā vidhy-artho ‘pi bhaved yataḥ | anātmādarśanenaiva parātmānam upāsmahe. NaiS 1.88. 476 under the overpowering influence of embodiment and keeping yama-niyama relevant even for the monk who does not, or rather should not, see duality. Parisaṅkhy , on the other hand, was tied to the identity statement and the manner in which such restricted reflection should proceed: affirming one thing but intending to negate another, much along the line of Śa kara’s claim that the real reference behind tat tvam asi was neti neti. Śa kara never says that the injunction itself might be a parisaṅkhy Ṭ If I am right, this point was lost on the later tradition. Sureśvara’s commentator Jñānottama, for instance, says that the injunction could be taken as a niyama if the vision of the non-Self overpowers the vision of the Self, or a parisaṅkhy if they are equal, which is a non-starter for Śa kara’s orthodox Brahma-Sūtra standpoint that the karma that had started bearing fruits always overpowers knowledge of the Self. Kumārila’s point that a parisaṅkhy affirms one thing intending to negate another was lost. Little wonder that this was all seen as hypothetical in any case. What Śa kara meant was that one should keep one’s thought on the Self by considering what the Self is not. Parisa khy naŚ A Second Avenue There is a second avenue through which we may approach the understanding of parisaṅkhy na: it is the Mah bh rata. The term itself or the absolutive form parisaṅkhy ya is used several times in the Mokṣa-Dharma section of Mah bh rata’s nti-Parvan, and in all cases, it is related to s ṅkhya as a discipline that is explicitly paired with and distinguished from yoga. We will focus on two such cases. First, the relation of parisaṅkhy na to s ṅkhya and its distinction from yoga is stated in verse 26 of chapter 306, in the dialogue between Vasiṣ ha and Karālajanaka that is nested in Yudhiṣ hira’s questioning of BhīṣmaŚ “So far I have accurately described the vison of yoga. Now I will speak the knowledge of s ṅkhya, which provides vision through 477 parisaṅkhy na.”69 Nīlaka ha glosses this parisaṅkhy na vision as “direct experience by means of full exclusion, dissolving every product into its immediate cause, in the manner of the snake and the serpent.”70 Vasiṣ ha proceeds to delineate a s ṅkhya system in which the twenty-four categories of creation, including the non-manifest prakṛti, emanate from and dissolve back into the inner Self.71 The Self is called the twenty-fifth principle, from which the other principles emanate but which, really, does not contain them at all (nistattva). Verse 42 says that the Sā khyas “perform a vision that is consequent on parisaṅkhy ” and is a s ṅkhya vision. Nīlaka ha’s glosses this as “the direct experience that is consequent on the dissolving of phenomenal manifestation in the conscious Self through pursuing the gross-to-subtle causal sequence.”72 Elsewhere he gives a similar gloss to the absolutive parisaṅkhy yaŚ “Defeating every preceding error with subsequent knowledge.”73 It seems fairly obvious that the idea involves making a full count of the categories of experience until one arrives at the Self, from which it all comes and to which it all resolves, not unlike the sixth chapter of the Ch ndogya. Second, in chapter 315 we read about another account of s ṅkhya, in which Yājñavalkya instructs Janaka in a teaching that is very much amenable to Śa kara’s doctrine of mutual superimposition of the Self and the non-Self—here puruṣa and prakṛti—through ignorance. The 69 yoga-darśanam etāvad uktaṁ te tattvato mayā | sā khya-jñānaṁ pravakṣyāmi parisa khyāna-darśanam. MBh 12.306.26, V.622 [12.294.26]. All references are to the “vulgate” text as commented by Nīlaka ha in his Bh rata-Bh va-D pa, with cross-references to the critical edition in the square brackets. 70 parisa khyānaṁ parivarjanaṁ rajjūraga-vat. uttarottarasya kāryasya pūrvasmin pūrvasmin pravilāpanaṁ tena darśanaṁ sākṣāt-kāro yasmiṁs tathā. V.622. 71 Cf., especially, verses 31-32: yasmād yad abhijāyeta tat tatraiva pralīyate | līyante pratilomāni s jyante cāntarātmanā || anulomena jāyante līyante pratilomataḥ | gu ā gu eṣu satataṁ sāgarasyormayo yathā. 72 parisa khyā sthūla-sūkṣma-krame a cid-ātmani prapañca-pravilāpanaṁ tāṁ anudarśanaṁ sākṣātkāraṁ sampādayantīty arthaḥ. V.623. 73 tatra uttarottara-jñānena pūrva-pūrva-bhramān parisakhyāya bādhitvā. On 12.301.5, V.602 [12.294.5]Ś jñānena parisa khyāya sadoṣān viṣayān n paḥ. 478 important knowledge here concerns the fact that qualities or characteristics (guṇa) that belong to prakṛti cannot be ascribed to the sentient puruṣa, which is, for this reason, nirguṇa, and does not come in touch with prakṛti. Yājñavalkya concludesŚ “This is the vision of s ṅkhya, the best full enumeration for you. Having thus fully enumerated, Sā khyas attain isolation.”74 Yājñavalkya proceeds in the next chapter with yoga that is the meditational, aṣṭ ṅga-yoga, and the two are clearly presented as alternatives for liberation. Parisaṅkhy na in the Mah bh rata, then, is s ṅkhya, whose purpose is to enumerate everything under the sun so as to facilitate, to refer to James Fitzgerald’s apt characterization of the context in which parisaṅkhy na is equated with s ṅkhya, disaffection from the world, a radical purging of egocentrism and desire from one’s taken for granted understanding of the world with oneself in it. This disaffection, called vairagya (a dissociation from life’s motivating stimuli at the visceral level of a person’s being), is effected by the systematic, enumerative contemplation of the entire system of the world. The word Sāṁkhya signifies comprehensive intuition,” or “all-gnosis,” and its cognates signify “enumerate, know the whole of some complex entity by itemizing and totaling every component of it.75 We should add: know it all so as to reduce it to what is further irreducible and know it as different from yourself. Back to the third prose chapter of UŚ, we cannot fail to notice that Śa kara is doing somewhat of that parisaṅkhy na that the Mah bh rata and Nīlaka ha are describing. The knower of Brahman who is troubled by things pertaining to the body, mind and words should cognitively reduce whatever is troubling him to the basic sense objects—sound and the rest— with the understanding that he as their knower is unrelated to them. Śa kara’s parisaṅkhy na, then, being a meditation, is a device that employs the pertinent categories of s ṅkhya to map all 74 sā khya-darśanaṁ etat te parisa khyānam uttamam | evaṁ hi parisa khyāya sā khyāḥ kevalatāṁ gatāḥ. 12.315.19, V.640 [12.303.19]. 75 Fitzgerald 2012:49. 479 possible identification points for the Self in such a way that they are reducible to the basic sense objects, from which he is different because they are not attributable to the Self. It is a meditation not on the Self, but on its difference from whatever is not the Self, such that when full dissociation is achieved, one cannot but be the Self and have that direct experience that makes all experience possible, but is essentially non-transitive. Parisaṅkhy na, this, is a structured reflection on neti neti, in which the full scope of identification points is mapped in a cause-effect chain. Both entry points, thus, lead to the same conclusion: parisaṅkhy na is s ṅkhya, in which the point is not to focus on the Self, but to eliminate everything that the Self is not. The two are, evidently, related through etymology, and the core meaning is “making the full count.” Kumārila’s account, however, is more informative, as he makes clear that the purpose behind the full enumeration is not to affirm what is stated, but to communicate what is not stated yet is intended. Be that as it may, we can now say the obvious: whereas prasaṅkhy na meditation was yoga, parisaṅkhy na reflection was s ṅkhya. The Purpose of Parisa khy na and the Nature of Liberation Liberation in Śa kara’s understanding was attained at the completion of the manana process. Did nididhy sana have, then, a causal contribution in the attainment of liberation? Śa kara’s gloss of Yājñavalkya’s nididhy sitavyaḥ was ni cayena dhy tavyaḥ, to be reflected with certainty, and certainty was the big common gain of ravaṇa and manana: the initial verbal cognition of unity produced by the identity statements of the Upaniṣad could finally terminate in pram ṭavagati by the processes of theological and philosophical reflection with a teacher. We saw under the previous heading that such vision of unity, however, was in permanent danger from the 480 overwhelming power of embodiment and the karma that had started bearing fruits, and nididhy sana had at the least the purpose of keeping this vision permanent. In the opening of the third prose chapter of US, Śa kara says that the parisaṅkhy na meditation is delineated for those aspirants after liberation who want to diminish (kṣapaṇa) the good and bad karma that had already been accumulated, as well as to prevent the replenishing of the karmic stock, with the following reasonŚ “The psychological faults are caused by ignorance, and they in their turn give rise to the vocal, mental, and bodily actions, through which karma of good, bad, and mixed nature is produced. (The parisaṅkhy na meditation is delineated) for the sake of liberation from them.”76 This obviously does not sit well with the end of the second chapter and the grand, almost pompous, announcement of liberation attained. Does it tell us something about the nature of liberation? Ānandagiri in his comment tries to grapple with the problem by introducing the notion of mediacy of the knowledge of the SelfŚ “It has been established that he who had attained the aforementioned knowledge that is devoid of injunctions and had accomplished his purpose does not need to observe anything prompted by an injunction. Now, the author shows that there is something that remains to be done on the part of those who have not attained immediate sentential knowledge because their understanding of the categories is not exceedingly clear.”77 Scholars in general also relate nididhy sana with the problem of mediacy of the knowledge of Brahman, but offer interpretations that are more in line with the prasaṅkhy na doctrine than with Śa kara’s own. M. Hiriyanna, for instance, saysŚ “The object of this stage [nididhy sana] is, as 76 mumukṣū ām upātta-pu yāpu ya-kṣapa a-parā ām apūrvānupacayārthināṁ parisa khyānam idam ucyate— avidyā-hetavo doṣāḥ vā -manaḥ-kāya-prav tti-hetavaḥ, prav tteś ca iṣ āniṣ a-miśra-phalāni karmā i upacīyante iti tan-mokṣārtham. USG 3.112. 77 vidhi-rahita-pūrvokta-vidyā-sampannasya k ta-k tyasya vidhito na kiñcid anuṣ heyam iti pratiṣ hāpitam. samprati yeṣām aparokṣaṁ vākyārtha-jñānaṁ sphu atara-padārtha-jñāna-virahatatvān na sampadyate teṣām anus heya-śeṣaṁ darśayati. Ānandagiri on USG 1.112. 481 often remarked before, to transform into direct experience the mediate knowledge of ultimate reality acquired by the study of the Upanishads and by reflection upon their teaching. It is accordingly vision that is sought now, and not mere knowledge.”78 Ānandagiri’s remark may not be fully on the point, because the two processes of ravaṇa and manana would have been repeated until the identity statements terminates in full understanding, as Śa kara had said under BS 4.1.2. Moreover, it was in later Advaita Vedānta that the notion of aparokṣa-jñ na took the central stage. It was crucial in the doctrine of prasaṅkhy na, and Śa kara was aware of it, but his reply to it was simple: because the identity statement is about myself, who am most intimately known to me, it does not require any additional measure of immediacy beyond pinpointing precisely the reference of the notion of the Self by eliminating all the other identification points. For Śa kara, the immediacy of the knowledge of the Self obtained as in the case of the boy who had forgotten to take himself into account, but is reminded of the fact when someone tells him, “You are the tenth person.” In general, although Śa kara clearly addressed the notion of the immediacy of brahma-jñ na, expressions such as aparokṣa-jñ na, aparokṣ nubhūti, or aparokṣ nubhava are absent from his works. Perhaps parisaṅkhy na was a contemplation that one would have practiced alongside ravaṇa and manana, on one’s own. There may be, however, something more to Ānandagiri’s comment. We will return to it in a bit, but let us now try to clarify a couple of things about the nature of liberation. Let us first finish our story with prasaṅkhy na-v da and the purported second understanding or cognition of Brahman. Ma ḍana Miśra, as we saw, thought that the scripturally formed notion of Brahman had to be meditated upon because it was relational, a verbal composite, and mediate. Śa kara solved the first issue by insisting that the scriptural notion of 78 Hiriyanna 1995:172. 482 Brahman was not a definite description, but a definition. He solved the second by placing the onus on the identity statements: they became the focal point of brahma-jñ na and no further immediacy was required because knowledge of Brahman just was knowledge of myself. And, while Ma ḍana thought that liberation would follow when ignorance had been dissolved through long meditation, for Śa kara ignorance was undone and liberation attained immediately on fully understanding the identity statement. He, in fact, explicitly argued against a doctrine according to which the very last cognition of the Self was the one responsible for undoing ignorance. Śa kara very much shared the prasaṅkhy na idea that knowledge of Brahman was prophylactic, but charged these fellow Vedāntins that they did not really understood what knowledge wasŚ We have already said that knowledge has visible results, the cessation of ignorance, grief, illusion, fear, and the like. Knowledge, in fact, terminates in the cessation of the psychological faults headed by ignorance. Whatever cognition is fruitful in bringing about the cessation of such psychological faults headed by ignorance, be it the first, the last, continuous, or discrete, only that is knowledge.79 For him, the crucial question was not how to transform scriptural knowledge into a direct vision of Brahman, but whether one felt this scriptural knowledge in one’s bones and one’s whole being so that one no longer craved after transient pleasures, feared transmigration and such. “I am Brahman” meant “I am none of these things.” As a corollary to this, he was serious and literal in insisting that knowledge had visible results: not a mystical intuition, but freedom from vices consequent on understanding that what prompts these vices has nothing to do with me. Knowledge was prophylactic in this sense of not only stopping the behavior that was predicated on ignorance, but undoing the psychological setup on which such behavior was predicated. This was the result that knowledge brings to consummation, from jñ notpati to phal vas na. And 79 avidyā-śoka-moha-bhayādi-doṣa-niv tteḥ pratyakṣatvād iti coktaḥ parihāraḥ. tasmād ādyo 'ntyaḥ santato 'santataś cety acodyam etat, avidyādi-doṣa-niv tti-phalāvasānatvād vidyāyāḥ—ya evāvidyādi-doṣa-niv tti-phala-k t-pratyayaḥ ādyaḥ antyaḥ santataḥ asantato vā, sa eva vidyā. BĀUBh 1.4.10, VIII.138. 483 once that has happened, all future “cognitions of Brahman” would be of the same kind—“I am Brahman as I am none of these things”—and it was but right that the first such cognition be the liberating cognition. We should also note that the final understanding of Brahman of the prasaṅkhy na kind was strictly impossible in Śa kara’s system. We saw in the previous chapter that the characteristic in virtue of which the Upaniṣads were a single corpus was that both meditation and knowledge were mental modifications, a mano-vṛttiŚ “Just as knowledge of non-duality is merely a mental modification, so are the other meditations mental modifications in nature.”80 We can try to understand this in consonance with Śa kara’s theory of cognitionŚ for there to be an ascertainment of an object, the mind must take the shape of the object and simultaneously be illuminated by the light of the Self.81 Because Brahman is a non-object, the mind in the cognition of non-duality cannot really assume any form, so we can surmise that the object would be just the reflection of the Self as understanding or pram pure and simple. Or, we can understand it preferably as the thought corresponding to the statement of the student in the Upade a-S hasr that we identified as the moment marking the attainment of liberation. In either case, however, it would be a mental modification that is predicated on the presence of the cognitive apparatus and the subject-object distinction. To put this differently, “experience of Brahman” in the truest possible sense in Śa kara’s system was a contradictio in adiecto. Ultimately, experience of Brahman, insofar as it was possible, was a product of ignorance. Śa kara, in fact, said as muchŚ “Both knowledge [of the Self-Brahman unity] and experience of karma are results of the actions that have started bearing fruits.”82 Karma was, of 80 yathā advaita-jñānaṁ mano-v tti-mātram, tathā anyāny apy upāsanāni mano-v tti-rūpā īti. ChUBh Introduction, VI.9. 81 See USP 18.118-120. 82 ārabdhasya phale hy ete bhogo jñānaṁ ca karma aḥ. USP 4.4. 484 course, one of the three factors of individuation that were consequent on ignorance, and thus knowledge of Brahman as one could possibly experience it, a mental modification, was just a product of ignorance. As such, it belonged to the same order of being, that of actuality, as any other experience of happiness, suffering or similar end products of manifest karma. The characteristic feature of this specific mental modification was that it had the ability to prevent the creation of new karma and the experience of the non-manifest karma, and this was at the least because the two belonged to a different order of being, that of potentiality. That which is not yet does not ever have to be. This was a mental modification that could in time put an end to all mental modifications, including itself. It had such power because it stopped the inverse attribution of qualities of the Self and the non-Self, yet itself was predicated on such inverse attribution, for which reason its nature was quite precarious. So long one was subject to the karma that had started bearing fruits, there had to be mental modifications of some kind, and there had to be a residue of karma that influenced one’s mental state. For this reason, it was of crucial importance to maintain this specific mental modification constantly alive to oneself, in face of the unfolding of the remaining karma. It was this, I would like to suggest, that was the main purpose of the parisaṅkhy na meditation: because one was liable to the karma that had started bearing fruits till final, post- mortem liberation in Brahman was achieved, it was possible to lose the immediacy of the knowledge of one’s being Brahman and to slip into identifying with any one thing from the sphere of the non-Self, whether words of praise or censure or sense objects. Parisaṅkhy na could in such a situation bring back the immediacy of knowing oneself as Brahman by recreating the awareness of what one is not, and with that stop the influence of the natural functioning of body, 485 mind, and words over oneself. In that sense, Ānandagiri seems justified in relating its use to immediacy, though perhaps not its first appearance. 486 PART FOUR: FROM IDENTITY STATEMENTS TO MAH -V KYA CHAPTER TEN: SARVAJÑ TMAN AND THE DOCTRINE OF UPANI ADIC MAH -V KYA Victorious is that immediate awareness, unaffected by illusion, residing in its natural eminence, the sole cause of the world, one whose plenitude is defiled by the differences fashioned by its own nescience, through which it is the world, the Supreme Self, and the individual Self.1 Introduction and a Historical Note In the previous two chapters, we focused our attention on the identity statements of the Upaniṣads and on their role as the final link in Śa kara’s chain of soteriological causality. They did not really have a name, though, and early Advaitins liked to call them metonymically “tat tvam asi and the rest.” Tat tvam asi was paradigmatic among them, since it explicitly related the two categories, but historically tasm d v etasm d tmanaḥ from the Taittir ya was more important for the formulation of the identity statement notion, because it was the second chapter of the Taittir ya that provided the core of the doctrine. In any case, however, one would not know that these identity statements did not have a name if one read scholarly accounts of Advaita Vedānta, and one would have heard countless times that these pithy identity statements were called mah -v kyas. One would have also read accounts of how Śa kara used and interpreted mah -v kyas. In this chapter, we will see that the doctrine of mah -v kya was not formulated by Śa kara himself, but by another early Vedāntin, namely Sarvajñātman. We will also see that mah -v kya had a prehistory in the fellow school of Vedic theology, Mīmāṁsā, and 1 svājñāna-kalpita-jagat-parameśvaratva- jīvatva-bheda-kaluṣīk ta-bhūma-bhāvā | svābhāvika-svamahima-sthitir asta-mohā pratyak-citir vijayate bhuvanaika-yoniḥ. SŚ 1.2. 487 that Sarvajñātman explicitly modeled the Vedāntic idea of mah -v kya on the Mīmāṁsā original, using a Mīmāṁsā structure to build an edifice of purely Śa karan blocks. We will in this way finally come to the resolution of this lengthy story. It will turn out that the identity statements of the Upaniṣads understood as mah -v kyas were only formally short statements, and from a fuller perspective they were massively long, a hierarchical organization of nearly the whole Vedic canon into a single text. But, before we do all that, let us first see how scholars commonly understand mah -v kyas and how much that is historically justified. The scholarly uses of mah -v kya can be classed in several loose groups of progressively reducing scope. To begin with, any short but somehow important statement from any foundational Hindu text can be called a mah -v kya. Edwin Bryant, thus, defines mah -v kya as “a ‘pivotal’, ‘most important’, or ‘representational statement’ for the theology of a sect.”2 Similarly, Richard H. Davis in his recent “The Bhagavad GitaŚ A Biography,” says that “Indian commentators often highlighted especially powerful statements in the Gita for special attention as mahavakyas (great utterances).”3 The late Tamal Krishna Goswami suggests that mah -v kyas that “are found across many traditions” are the “one good device” through the application of which “apparently contradictory passages can be brought into conformity with the canon.” He also seems to share Bryant’s idea that mah -v kyas are representational statementsŚ “Amid the vast firmament of a tradition’s authoritative texts, a mah v kya’s luminosity outshines that of all others, as it encapsulates the tradition’s inner core, the purport of its beliefs.”4 Goswami proceeds to identify an English mah -v kya, “Krishna is the Supreme Personality of Godhead,” as the cornerstone of the Gauḍīya theology of his own teacher, A.C. Bhaktivedānta Swami 2 Bryant 2003:xxi, 500. 3 Davis 2015:99. 4 Goswami 2012:128-9. 488 Prabhupāda. Such presentations evidently take mah -v kyas as some sort of an interpretative device, and they seem to be popular with scholars who come from a theological background, Catholic or otherwise. Most explicit in such theological approach was perhaps Julius Lipner, who says in his study of Rāmānuja that both Śa kara and Rāmānuja erected their respective systems of Vedānta on the foundation of select statements from the Upaniṣads, mah -v kyas, which served the roles of hermeneutic keys.5 More restricted in scope seems to be the practice of naming a mah -v kya any short and important statement, but strictly from the Upaniṣads, generally without providing justification for doing so. Thus, for instance, Chris Bartley calls the Taittir ya 2.1.1 statement that defines Brahman, satyaṁ jñ nam anantaṁ brahma, a mah -v kya, but he does not say why specifically that statement would be a mah -v kya, nor does he provide a reference in support.6 As we will see later, this was one of the statements that were explicitly not mah -v kyas in Sarvajñātman’s account. In a similar vein, Rudolph Otto in his classic “Mysticism East and West” labels the famous Ch ndogya 6.2.1 statement, sad eva somyedam agra sid ekam ev dvit yam, a “great saying.”7 Matthew Kapstein, on the other hand, talks about Śa kara’s careful deployment of mah -v kyas or passages derived from the Upaniṣads in the first prose chapter of the Upade a- S hasr as a way of structured teaching, suggesting, thus, that all the Upaniṣadic statements that Śa kara strings there in a textual cavalcade are mah -v kyas in some way.8 Finally, the identity statements are generally called mah -v kyas, typically tat tvam asi among them. This use is ubiquitous, and the following few examples are adduced merely for illustrative purpose. Bader: “Still more important here is the discussion of the sacred utterance 5 Lipner 1986: 82. 6 Bartley 1986:103, 105. 7 Otto 1970:20. 8 Kapstein 2015. 489 (mah v kya), 'Thou art that' (tat tvam asi). This mah v kya is the basis upon which Śa kara constructs his method of liberation.” “The focal point of the discipline leading to true knowledge is the mah v kya, a sacred utterance expressing only the highest truth.”9 MalkovskyŚ “Śaṃkara teaches that ignorance may be entirely destroyed upon hearing for the first time a ‘Great Saying’ (mah -v kya) from the Upaniṣads such as ‘That thou art’ (tat tvam asi) and fully gasping [sic] its import.”10 “The highest wisdom of the Upaniṣads, the essence of the revealed ruti, is most fully embodied in a small number of cryptic sentences called mahav kyas.”11 PandeŚ “A common view found among other Vedāntins was that the mere hearing of the Vedāntic Mah v kyas like Tattvamasi does not lead to the direct knowledge of the Self.”12 NarayanacharyaŚ “ hankara discriminates between the so-called Mah v kyas (The ‘great’ statements) and Laghuv kyas (the ‘ordinary’ ones), with no precedent before him and no sanction by the Vedas and the Mim msa discipline.”13 DeutschŚ “’Hearing’ ( ravaṇa), the first of the three, has to do with the initial (and sometimes immediately enlightening) acquaintance with the teachings of Advaita. The aspirant is encouraged to listen to the sages and study the Vedāntic texts. He must investigate the mah v kyas or great sayings of the Upaniṣads and determine their proper purport.”14 More specifically, sometimes a limited set of identity statements are called mah -v kyas. Mayeda says that eleven or twelve of them are enumerated, but the common set has four.15 These are tat tvam asi, ahaṁ brahm smi, ayam tm brahma, and prajñ naṁ brahma. Sometimes, 9 Bader 1990:49, 66. 10 Malkovsky 2000a:78. 11 Malkovsky 2001:84; and throughout. 12 Pande 232; and throughout. 13 Narayanacharya 1989:10. 14 Deutsch 1973:106. 15 Mayeda 2006b:117: “The Upaniṣadic sentence ‘ahaṁ brahm smi’ (I am Brahman) is one of the Mah v kyas (Great Sentences) which the Advaitins consider to be best indicative of the Advaita doctrine. There are said to be eleven or twelve Mah v kyas, but among them this sentence and ‘tat tvam asi’ (Thou art That) are best known.” 490 however, sarvaṁ khalv idaṁ brahma from the ṇḍilya-vidy is included in place of prajñ naṁ brahma.16 The doctrine of four mah -v kyas is commonly ascribed to Śa kara himself. Andrew Nicholson, for one, saysŚ “For instance, the eighth-century Advaita Vedāntin Śa kara dubbed four Upaniṣadic sentences as ‘great statements’ (mah v kyas)Ś ‘You are that’ (tat tvam asi), ‘I am Brahman’ (ahaṁ brahm smi), ‘This self is Brahman’ (ayam tm brahma), and ‘Brahman is consciousness (prajñ naṁ brahma).”17 These four mah -v kyas are sometimes described as one for each of the four Vedas: each mah -v kya expresses the essence of an Upaniṣad, which in its turn expresses the best and highest knowledge contained in one of the four Vedas; these four Vedas have been entrusted in care to the four great monasteries at the four cardinal points, coming in succession from one of the four principal students of Śa kara: (1) prajñ naṁ brahma in AiU 3.3 of the Ṛg Veda → Puri → Padmapādaś (2) ahaṁ brahm smi in BĀU 1.4.10 of the Yajur Veda → Shringeri → Sureśvaraś (3) tat tvam asi in the ChU 6.8.7 of the S ma Veda → Dwaraka → Hastāmalakaś (4) ayam tm brahma in MāU 2 of the Atharva Veda → Badrinath → Tro aka. In any case, the scholarly assumption is either that mah -v kyas were just always there, or that Śa kara invented the idea. Contrary to these reports, Śa kara himself was not the author of the mah -v kya doctrine. There is, in fact, only one mention of Upaniṣadic mah -v kyas in his authentic works. It is found in his Bh ṣya on the Aitareya Upaniṣad, and although it is clearly related to that knowledge which is available from the identity statements, no underlying theory can be reconstructed from that single comment: When by good fortune a teacher of supreme compassion beat near his ears the drum of the great sayings of the Upaniṣads whose notes were meant to wake up the knowledge of the Self, then the individual realized that Puruṣa, discussed as the Lord of creation— called Puruṣa because of taking a residence in the city (of the heart)—was Brahman, the 16 Murty 1959:74-5; Malkovsky 2001:84; Zaehner 1957:28-9; Potter 1981:59. 17 Nicholson 2010:41. 491  great. … How (did he realize it)? ‘Oh, I have seen this Brahman as the real nature of my Self!’18 There are, of course, numerous mentions of mah -v kyas in the many spurious prakaraṇa works that are attributed to Śa kara, all of which are nested in post-Śa kara developments of doctrine and style. If Śa kara did not explicate a mah -v kya doctrine, it follows by implication that the idea of four mah -v kyas was not Śa kara’s own either. While I have not pursued the formation of this specific notion, it appears that it must have been well-established by the 14th century and the great Advaitin Vidyāra ya. In his Pañcada , Vidyāra ya interprets the four mah -v kyas in Chapter Five, called Mah -V kya-Viveka, where he follows the common order of the four Vedas: Ṛg (prajñ naṁ brahma), Yajur (ahaṁ brahm smi), Sāma (tat tvam asi), Atharva (ayam tm brahma).19 The idea is present in the most influential hagiography of Śa kara, aṅkara- Dig-Vijaya of Mādhava, which says that Govindapāda personally instructed Śa kara in these four statementsŚ “That best of ascetics, exceedingly pleased by the worship that was done with devotion, instructed him in Brahman through the four sentences that are the summit of the [respective] Vedas.”20 The most prominent place of the four mah -v kyas idea is in the corpus of the so-called Maṭh mn ya texts. These are a few short compositions commonly attributed to Śa kara— Maṭh mn ya, Math mn ya-Stotra, Math mn ya-Setu, Maṭhetivṛtta—or in one case a 18 sa kadācit parama-kāru ikena ācārye ātma-jñāna-prabodha-k c-chabdikāyāṁ vedānta-mahā-vākya-bheryāṁ tat- kar a-mūle tāḍyamānāyām, etam eva s ṣ y-ādi-kart tvena prak taṁ puruṣaṁ puri śayānam ātmānaṁ brahma b hat … pratyabudhyata apaśyat. katham? idaṁ brahma mamātmanaḥ. svarūpam adarśaṁ d ṣ avān asmi. AiUBh 1.3.13, V.266-7. 19 Swami 1967:122-5. 20 bhakti-pūrva-k ta-tat-paricaryā-toṣito ‘dhikataraṁ yati-varyaḥ | brahmatām upadideśa caturbhir veda-śekhara-vacobhir amuṣmai. 5.103. On the aṅkara-Dig-Vijaya, see Bader 2000:53-62. Bader does not accept the common ascription of this work to Mādhava-Vidyāra ya and dates the work between 1650 and 1798. 492 Maṭh mn ya Upaniṣad.21 The most well-known of them, the Math mn ya-Setu, consists of 75 anuṣtubh verses. In Mathew Clark’s estimate, these texts are not more than four or five hundred years old.22 Maṭh mn ya literally means “the tradition of the Ma ha,” and these texts in fact provide the doctrinal frame of the Daśanāmī institution of Advaita renunciants by circumscribing the tradition. The assumption being that Śa kara founded four monastic institutions at the four cardinal points, these texts aim at filling out identification rubrics for each of the four traditions by providing the following information, taking Ś geri as an example: (1) Āmnāya (realm of one of the Vedas): Southern; (2) Ma haŚ Ś geri/Śāradā-pī ha; (3) Jurisdiction (maṇḍala)Ś Āndhra, Draviḍa, Kar ata, Kerala; (4) Orders (sanny sa names)Ś Sarasvatī, Bhāratī, Purīś (5) DeitiesŚ Ādi-Varāha, Kāmākṣīś (6) Holy place (t rtha): Tu gabhadra, Rāmeśvaramś (7) Veda: Yajur; (8) Mahā-vākyaŚ ahaṁ brahm smi; (9) GotraŚ Bhūrbhuvaś (10) Brahmacārī nameŚ Caitanyaś (11) SampradāyaŚ Bhūrivāraś (12) Śa kara’s student to whom the Ma ha was entrustedŚ Hastāmalaka (Sureśvara). There is no uniformity in these Maṭh mn ya texts regarding which Deity and which student of Śa kara is associated with which Ma ha, and only To aka is related to the Jyośima h in all accounts. The “Sampradāya” item is not very clear either, and it comes to stand as a collective term for the several sanny sa names that are given in the tradition. Mah -v kya here is the liberating mantra of the respective tradition.23 That is, it is the mantra which Daśanāmīs receive during the second stage of their initiation into renunciation (vidy -saṁsk r or viraj -havan).24 21 Clark 2006:114, 274, n.1. 22 Clark 2006:117. 23 On the Maṭh mn ya tradition, see Clark 2006:114-122; Mishra 2001:1-57. 24 Clark 2006:93-9. 493  This scheme of quartets does not seem to have much grounding in practice. In his study of the Ś geri tradition, Yoshitsugu Sawai notesŚ “According to tradition, Śa kara is said to have enjoined the Ś geri Ma ha especially to study the Yajur Veda and to meditate on the mah v kya, ahaṁ brahm smi, which occurs in the B hadāra yakopaniṣad of the Śukla Yajur Veda. The author of this work has observed, however, that some of the seventy brahmac rins at Ś geri Ma ha during his visit were studying not the Yajur Veda but the Ṛg Veda and to some extent the other Vedas as well.”25 It is, rather, a device of stamping authority over the realm, similar to the hagiographic notion of dig-vijaya or conquering the four quarters. In his commentary on the Maṭh mn ya-Setu, in fact, Mishra says that “for the stability of the world and protection of the Dharma,” Śa kara “divided entire Bharat into four Dharm Empires.”26 The four Śa karācāryas are the tutelary heads of the whole Daśanāmī institution that otherwise has little or no relation to the four Ma has. The instability of the quartets scheme is evident in sarvaṁ khalv idam brahma and Kāñcipuram perpetually lurking in the shadows as a fifth mah -v kya and Pī ha. Mah -v kya in M m s As I said in the Introduction, the idea of mah -v kya had its prehistory in Mīmāṁsā. This early side of the story is virtually unknown, however, and two scholarly accounts are known to me, those of Kunjunni Raja and Larry McCrea.27 Moreover, on the face of it, mah -v kya in Mīmāṁsā appears to be the exact opposite of mah -v kya in Advaita VedāntaŚ a Mīmāṁsā mah -v kya is a “great” sentences in the literal sense of the wordŚ it is long, not pithy. Kunjunni 25 Sawai 1992:64. 26 Mishra 2001:iv. 27 Kunjunni Raja 1977:161-2; McCrea 2008:55-98. 494 Raja calls it a “compound sentence,” which indicates that it is a larger textual unit than a single complete sentence. The idea of mah -v kya was a relatively minor, but still common currency in classical Mīmāṁsā. It is explicitly mentioned only rarely, it is never explicitly theorized—we must reconstruct its meaning from use—and it is never an object of contention, which suggests that it was an idea clear and acceptable to all. Śabara mentions it once under MS 6.4.25, and Kumārila twice, in ŚV V kya 140-143 and TV 1.2.7. Prabhākara also knows about it: he mentions mah - v kya in Bṛhat 1.2.1. The backbone of mah -v kya in Mīmāṁsā was a related notion or principle, one that we saw several times in the dissertation: paryavas na or completion, termination of meaning. The principle of paryavas na says that any larger sentential unity in which a smaller sentence has been absorbed, finalized and, if so required, altered, can be called a mah -v kyaṬ In principle, this works at any linguistic level, and it is rather in relation to short sentences that the term is used. Take, for instance, an injunctive sentence, such as “One should lie,” which, when negated, becomes a prohibitionŚ “One should not lie.” The second sentence becomes a mah -v kya in relation to the first, which is completed and altered in the second, yet still maintains its individual existence as a constitutive part of the final sentence. The two distinct units, in our example the injunction and the negative particle, restrict one another, such that the injunction becomes a prohibition and the negation becomes specified. The final sentence is a mah -v kya, while the two elements become av ntara, intermediate in relation to it. This is how, in fact, Śabara talks about mah -v kya in his Bh ṣya: And, when a mah -v kya has been formed, the intermediate sentence (av ntara-vakya) is no longer an evidence, because it has been defeated by the other word, as from the 495  intermediate sentence which enjoins looking a prohibition “One should not look at the rising sun” is understood in the mah -v kya.28 It is clear from this that mah -v kya in Mīmāṁsā was a “final” or superordinate sentence in relation to an intermediate sentence, av ntara, one that is finalized in the composite whole. Mah -v kya is, thus, most commonly paired with av ntara. Its important feature as such a final sentence is that it becomes pram ṇa or stronger evidence in relation to its constitutive av ntara- v kya, which has been defeated by it. While the av ntara-v kya remains a unit unto its own and a constitutive part of the mah -v kya, it has lost independence. Let us look now at Kumārila’s TV 1.2.7. This is part of a longer section that Kumārila presents as an alternative interpretation on part of the Bh ṣya, one which he eventually refutes, but the specific point about mah /av ntara-v kya is not controverted: In all cases, the intermediate sentences (av ntara-v kya) are not authoritative when there are final sentences (mah -v kya), just as when there is a final number the intermediate numbers don't count. But, when they are used alone, then they are authoritative, because they don't have expectancy, as in “there is a piece of cloth.” However, it is not that there is no expectancy in all cases just because we do not see expectancy in this case, given that [the sentence] is complete by this much. For, all sentences are completed when other meaningful words are not pronounced with them. Generally, those of skilled vision have the expectation over and above the heard words. But, when this “over and above” is not present, there is no expectancy either. When it is present, we understand that sentential unity obtains. Otherwise it would not be there even when words are pronounced. So, when there is no mah -v kya, the av ntara-v kya is authoritative, but when mah -v kya is possible, it is not.29 Kumārila’s example is based on Śabara’s Bh ṣya, who explains how sentences that are apparently complete can absorb additional words that are added to themŚ if the adjective “red” 28 na ca mahā-vākye sati avāntara-vākyaṁ pramā aṁ bhavati, padāntarasya bādhanāt, yathā, nodyantam dityam kṣeta—iti pratiṣedho gamyate mahā-vākyāt, avāntara-vākyād īkṣa a-vidhānam. MSŚBh 6.4.25, I.p.689. 29 sarvatrānāntara-vākyāni mahā-vākyeṣv apramā aṁ mahā-sa khyāsv ivāvāntara-sa khyā bhavanti. tāny eva tu yadā kevalāni prayujyante, tadā nirapekṣatvād bhavanti pramā am. yathā pa o bhavatīti. na ca kadācid etāvan- mātre a samāpter nirākā kṣya-darśanāt sarvatra nairākā kṣyam. yogya-padāntarānuccāra ena hi sarva-vākyāni paryavasyanti. itarathā punaḥ śruta-padātireke āpi nipu ad śāṁ bhavaty evāpekṣā. sā cānupalabdhyā nivartate. tathā ca yatropalapsyante, tatraikavākyatāṁ gamayiṣyante. anyathā hy asau tad-uccāra e 'pi na syāt. tad iha mahā- vākyena vināvāntara-vākyaṁ pramā am, tat-sambhāve tu na. TV 1.2.7, I.117. 496 happens to be near the sentence, it must become its part because words always have syntactic expectancy, openness to being modified by other words. “There is a piece of red cloth” is a mah -v kya in relation to “There is a piece of cloth,” its av ntara-v kya. While this may sound all too simple to be important, Śabara and Kumārila are discussing through the example how arthav da passages attach to injunctions to form sentential unity, and such unity can obtain on a level of a single sentence or between several sentences that form a text. It may be useful to think through the numbers example here in terms of arithmetic operations and rules of precedence. An expression such as 5 + 3 x 7 has two levels at which the respective operations will be performed. Multiplication has precedence over addition because it relates its terms more proximately, but precisely because of that the addition operation is the final and governs the full expression. The multiplication is intermediate in relation to it, since it simply gives one of the final terms that are related through addition. The multiplication can be seen in isolation, and unless it is subsumed under a further operation it will be final, but it is always potentially subsumable under addition or subtraction. At the core, a mah -v kya is not necessarily massively long: all that is required is that there is a longer sentence that subsumes another shorter sentence, which is av ntara in relation to it. In the ŚV, thus, Kumārila says that a four-word sentence is a mah -v kya, a long sentence, in relation to a laghu-v kya, a short sentence, of three words.30 However, we saw in the Second 30 “If it is assumed that an independent three-word sentence does not exist in the four-word sentence, then a tree would not exist in a forest. If you say that it is different because it is real, being separately formed, then words are also different from their sentences, and the same would apply to words and phonemes. Therefore, just as small sentences (laghu-v kya) continue existing in a large sentence (mah -v kya), likewise words and phonemes persist in sentence cognition. – But, these are never used alone. – The small sentence is also used for the purpose of the large. –But, it still denotes a small reference! – So do words.” p thak-prasiddha-sad-bhāvaṁ tripadaṁ ca catuṣpade | nāstīti yadi kalpyeta v kṣo na syāt tadā vane || 140 || p thak-prasiddhy-amithyatvāt syād vākyāntaratā yadi | vākyāc chabdāntaratvaṁ syāt tathaiva pada-var ayoḥ || 141 || 497 Chapter how the combination of individual words and sentences through the principle of subordination and the force of syntactic expectancy can go all the way to a point where a whole text is constituted, delineating a sacrifice organized around the centrality of the ritual action. Such description of a sacrifice is, in fact, the ultimate point of the paryavas na process, the “final” final meaning which terminates all possible individual meanings. In the truest sense, only this is a mah -v kya because only here everything is finalized and the syntactic expectancy of the bh van is fully saturated. In this, mah -v kya is synonymous with what Mīmāṁsakas call prayoga-v kya or prayoga-vidhi, a text delineating step by step the performance of a Vedic ritual and presupposing everything on which the success of the ritual depends. Such is the definition given in the M m ṁs -Ko a: “A large sentence is the pram ṇa that establishes the principal- auxiliary relationship. It has the form of an applicatory injunction (prayoga-vidhi). It is necessarily an inferable entity.” A mah -v kya is, thus, literally a “great,” long sentence, a whole book or a ritual manual, one in which the whole ritual has been delineated through the application of the principles of ascertaining the principal-subordinate relationship between the distinct ritual elements and in which a single sentential reference obtains: the ritual action qualified by all the ritual details.31 In such a mah -v kya, all the details have their individual, av ntara, context, and to them arthav da passages may be attached. All is finalized, however, in the mah context, which is governed by the injunction that institutes the sacrifice. That the mah -v kya is an inferable entity means, as far as I can see, that it needs to be worked out by the performer of the sacrifice, or someone who takes the trouble to do that by tasmāt yathā mahā-vākye laghūnām anirākriyā | tathaiva pada-var ānāṁ nāsattvaṁ vākya-buddhiṣu || 142 || kevalasyāprayogāc cet mahad-arthe laghor api | alpārthe tat-prayuktaṁ cet svārthe tadvat padāni naḥ. ŚV V kya 140-43. 31 MK VI.p.3132: mah -v kya (viniyojakaṁ pramā aṁ) prayoga-vidhi-rūpaṁ nityānumeyam eva. 498 producing a manual, through the application of the Mīmāṁsā principles of interpretation of the principal-subordinate relationship, viniyoga. In that sense, ritual manuals are human compositions in the form of organization of scriptural statements. Sarvajñ tman and the Preliminaries As I said in the introduction, Śa kara himself did not have much to say about mah -v kyas. Other than the AiUBh, the only other mention of mah -v kya is in the BSBh: For, when there is a mah -v kya which gives rise to a cognition of meaning, the intermediate sentence (av ntara) does not give rise to a cognition of meaning independently. For instance, in the negative sentence ‘One should not drink alcohol,’ only a prohibition of drinking alcohol is understood from the relation of the three words, and not over and above that an injunction of drinking alcohol from the relation of the two words, ‘One should drink alcohol.’32 His understanding is clearly identical to that of Śabara, and likely based on Śabara’s comment. Sureśvara and Padmapāda do not mention mah -v kya either,33 and for an explicit theory of the notion we must look at the great 10th-century Advaitin Sarvajñātman and his treatise Pañca- Prakriy .34 Before we move onto Vedāntic mah -v kyas, however, we need to introduce several ideas on which Sarvajñātman depends. 32 na hi mahā-vākye ’rtha-pratyāyake ’vāntara-vākyasya p thak-pratyāyakatvam astiś yathā na sur ṁ pibet iti nañvati-vākye pada-traya-sambandhāt surā-pāna-pratiṣedha evaiko ’rtho ’vagamyateś na punaḥ sur ṁ pibed iti pada- dvaya-sambandhāt surā-pāna-vidhir apīti. BSBh 1.3.33, I.p.206. Note that the word-count in the translation follows the Sanskrit, not the English. 33 I have read a good part of Sureśvara’s BĀUBhV and have not seen a single mention. Although it remains possible that it does appear somewhere, I am very skeptical of that. In any case, if Sureśvara had a mah -v kya theory, one would expect to find it in the Naiṣkarmya-Siddhi, which is all about tat tvam asi and ahaṁ brahm smi. 34 What is known about Sarvajñātman is that he wrote his masterpiece Saṅkṣepa- r raka during the reign of a king by the name of Manukulāditya. That much he says himself in the penultimate verse of the SŚŚ “The best of renunciants adorned by the name of “Sarvajñātman,” his mind purified by the touch of the dust from the lotus feet of Deveśvara, composed this Saṅkṣepa- r raka that magnifies the understanding of noble men when the glorious Manukulāditya of royal pedigree rules the Earth as a sovereign.” Based on inscriptional evidence, this Manukulāditya has been identified as the Chera king Bhaskara Ravi Varma, whose dates seem to be 962-1018 AD. Sarvajñātman was a Vaiṣ ava and likely a priest in the Padmanabhaswamy Temple in Trivandrum. There is a tradition based on commentaries on the Saṅkṣepa- riraka (including Madhusūdana Sarasvatī’s) that portrays him as a student of Sureśvara based on homonymy (deva = sura), but that is certainly wrong. In all his works Sarvajñātman refers to his teacher as “Deveśvara” and to Sureśvara as the Vārttikakāra and once as Sureśvara. In the colophon of his third preserved work, the epistemological treatise Pram ṇa-Lakṣaṇa, he gives the following 499  The first point concerns the signification function of words. We saw in Chapter Eight that the two categories of tat and tvam obtained a single reference when put in apposition, without giving up their natural denotation in the process. “Without giving up their individual meaning, the two words mingle to obtain a special meaning and terminate in understanding the inner Self. There is no other meaning opposed to this one.”35 This was required because on the one hand the statement had to be taken literally—it is not, for instance, that the four-headed demiurge Brahmā and/or my body would be intended to be affirmed as identical—and on the other the two entities were, in fact, different: tvam stood for the image of the Self in the mirror of the intellect, and tat for Brahman in its causal role. The identity statement had to presuppose a measure of difference in any case, to be meaningful or informativeŚ Śa kara had little interest in the so-called aikya- s m n dhikaraṇya or full synonymy. Furthermore, while standing for their individual referents in their primary signification function, the two words restricting one another obtained a single sentential reference not through the same primary, but the secondary signification function. Two conditions, thus, obtained for this lakṣaṇ : the categories had to express directly their individual references, and indicate a joint reference at the same time. Śa kara did not say much more about the secondary signification function. He did, however, specify that such secondary signification function could be “proximate” or “distant,” student-teacher lineageŚ Śreṣthānandapāda -> Devānandapāda -> Deveśvarapāda -> Sarvajñātmapāda. See Eswaran Nampoothiry’s Introduction to the Pram ṇa-Lakṣaṇa, p. ix-xxiii; Kocmarek 1985:7-11; Potter 435-6. Christopher Minkowski (2011) notes that Sarvajñātman became very important for Advaita Vedānta in early modernity (15 th to 18th century), and by implication for the status of Advaita as sort of an establishment philosophy for the 19 th and 20th century generation that achieved independence. He suggests that the reason for this might be the nature of Sarvajñātman’s Saṅkṣepa- riraka as a summary of Śa kara’s BSBh “in about 1250 accessible verses,” a proto- manual of the Ved nta-S ra sort. While there is obviously substance to Minkowski’s intuition—the SŚ is indeed very systematic—Sarvajñātman’s verses are not always accessibleś the sheer number of meters he uses is staggering for a work in philosophy, and he certainly has poetic aspirations. Furthermore, the polemical sections of the SŚ are with the old “enemy” the Mīmāṁsakas, and Advaita polemics in the early modern period had picked a different fight, with the movement of Madhva. Sarvajñātman’s place of honor needs further clarification. 35 svārthasya hy aprahā ena viśiṣ ārtha-samarpakau | pratyag-ātmāvagatyantau nānyo 'rtho 'rthād virodhy ataḥ. USP 18.171. 500 and he related them to vi eṣaṇa and adhy sa, qualification and superimposition, as in the blue lotus and the intentional superimposition of Viṣ u’s qualities on the image. These two, further, were a subset of the four types of co-referentiality, which, apart from the uninteresting synonymy, included negation, which was, in fact, the modus operandi of the identity statements up until the proper references of the two categories were fixed. At the last instance of tat tvam asi, when the references of the two individual words were settled, the relation was that of vi eṣya-vi eṣaṇa-bh va, since it was important for the words to keep their references, but even that was negational to a degree at the secondary signification function plain, which operated not on the level of the individual words, but of the sentence. Śa kara did not talk much about the signification functions and cannot, perhaps, be commended for absolute clarity. We should also be mindful that the secondary signification function could obtain only at the sentence level, not of words that stand alone. It obviously concerned the individual words, but for Kumārila it concerned the sentential reference as well. Think, for instance, of the sentence “The lion-boy is sleeping.” “Lion” here obviously must be understood metaphorically, through bravery as a shared quality, but for Kumārila the whole sentence, having the qualified boy as its reference, was also an instance of the secondary signification function: all sentences are about particulars, and words denote only universals. Let us now see what Sarvajñātman had to say about signification. There are three signification functions. The primary is the occurrence of the word in the sense of the universal, in the manner that the elders use it. The secondary signification function is of two kinds, one based on a direct relation of two things, and another on a relation through shared quality. The last is, thus, twice remote from the primary, because a word is used in a sense of another word not through a direct relation, but through some qualities that are characteristic for its denotation but 501 are predicated to another thing. Sarvajñātman’s example is “Devadatta is a lion,” where leonine qualities such as fierceness and bravery are predicated of Devadatta.36 We may call it metaphorical signification function. In Sanskrit, the three functions are called vṛttis, that is, mukhya- or prasiddhi-vṛtti, lakṣaṇ -vṛtti and gauṇa-vṛtti.37 I saved the second vṛtti for the end because it is more complicated than the first two: it is itself of three kinds. “The lakṣaṇ signification function is an occurrence of the word in the sense of another meaning through a relation with the primary meaning, when the understanding of the primary meaning is blocked by another reliable warrant.”38 The important feature is that this function obtains through a relation to the primary meaning of the word; that is, the use is still justified through the normal denotation. We may call it relational signification function. One kind of this lakṣaṇ is when the word occurs in the sense of something completely different from its common occurrence, yet immediately related to it, the classical example being “A hamlet in the Ganges,” where “Ganges” must stand for the bank of the riverŚ a different thing, yet contiguous with it. Sarvajñātman calls this secondary signification function jahal-lakṣaṇ , the relational signification function in which the word gives up its meaning.39 A second kind is when a word stands for a meaning that is different from its primary meaning, but includes this primary meaning in all its scope. Sarvajñātman’s example is “The red one is standing,” where the red one is a horse, specifically a red horse. “Red” here does not signify the universal “redness,” but does include it as a quality of the denotation. Sarvajñātman calls this function ajahal-lakṣaṇ , the 36 gu a-v ttis tu mukhyārtha-parigrahe pramā āntara-virodhe sati mukhyārtha-gu a-yogād arthāntare v ttiḥ itiś yathā siṁho devadattaḥ iti kraurya-śauryādi-siṁha-gu a-yogāt siṁha-śabdasya devadatte v ttiḥ. PPr 1, p.6-7. 37 tisraḥ śabdasya v ttayaḥ prasiddhi-lakṣa ā-gu a-v ttaya iti loke prasiddhāḥ. PPr 1, p.4-5. 38 lakṣa ā tu punaḥ mukhyārtha-parigrahe pramā āntara-virodhe sati mukhyārtha-sambandhād arthāntare v ttiḥ. PPr 1, p.6. 39 tatra jahal-lakṣa ā nāma śabdasya mukhyārtha-parityāgena arthāntare v ttiḥś yathā ga gā-śabdasya svārtha- parityāgena tīra-mātre v ttiḥ. PPr 1, p.7-8. 502 relational signification function where the word does not give up its meaning.40 Finally, a third form of this relational secondary signification function is when a word stands for a reference different from its own meaning, but not for its full scope. The example here is “This is that Devadatta,” where the proximate and the distal pronoun stand for an individual that is temporally and spatially qualified differently. The pronouns do not denote these qualifications, but the pure individual. Sarvajñātman calls this function jahad-ajahal-lakṣaṇ , the relational signification function where the word partially gives up and partially retains its meaning. In later Advaita Vedānta it was also known as bh ga-ty ga-lakṣaṇ .41 Now, there is a significant historical note to be made about Sarvajñātman’s presentation of the relational secondary signification function. The common Sanskrit terms for a signification function were akti and vṛtti, capacity and occurrence.42 Sarvajñātman used both, with vṛtti as the category term. The grammarians, however, used the term vṛtti in the sense of modes of complex language formations of several kinds, including compounded words.43 Discussing the lakṣaṇ - vṛtti, Sarvajñātman references the grammarian’s notion of vṛtti and the Aṣṭ dhy y 2.1.1, where specifically word compounding is discussed. It is instructive to look at Patañjali’s Mah -Bh ṣya comment on the sūtraṬ Patañjali talks about compound formation as one of the cases of vṛtti, and proposes two ways in which two words, one subordinate and one principal, combine to produce an integrated meaning. The first is called ajahat-sv rth vṛtti, in which the subordinate word does not give up its meanings in the compound formation, and the second is jahat-sv rth vṛtti, in which the 40 ajahal-lakṣa ā tu pu aḥ mukhyārtham aparityajya k tsnam eva svārtham upādāya arthāntare v ttiḥś yathā oṇas tiṣṭhati ity aśva-lakṣa āyāṁ śo a-śabdasya śo imānaṁ g hītvaivāśva-vyaktau v ttiḥ. PPr 1, p.8. 41 jahad-ajahal-lakṣa ā tu mukhyārtha-parigrahe sati mukhyārthaikadeśa-parityagena śabdasyaikadeśāntare v ttiḥ; yathā so ‘yaṁ devadattaḥ iti vākye so-‘yaṁ-padayoḥ deśa-kāla-śabala-vācinoḥ deśa-kāla-bhāga-parityāgena devadatta-vyaktau v ttiḥ. Ibid. 42 Deshpande 2007:53-6. 43 See Tubb and Boose 2007:35ff. 503 subordinate compound member does give up its meaning. The illustration being r jñaḥ puruṣa or “king's man,” when the compound r ja-puruṣa or “king-man” is formed, under the first view the word r jan stands not for a king, but for the man, employing its meaning as a qualifier of the man. Under the second view, both constitutive members give up their meaning and a single meaning obtains, namely a “king-man.” We may think of the difference through the phrase dar an yaḥ r ja-puruṣaḥ, “a handsome king-man.” Under the first view, the adjective qualifies the man, who also happens to be related to the king. Under the second view, it is the whole “king-man” that is qualified. Both views involve problems, which we not need go into, except that some prompt Patañjali to modify the presentation of jahat-sv rth to the effect that although the subordinate member gives up its own meaning, it does not do so entirely. Only that part of the meaning which is incompatible with the main member is abandoned.44 It is very much possible that Sarvajñātman put two and two together: if lakṣaṇ is the signification function or vṛtti that obtains when words are combined, and vṛtti in any case stands for complex formations, then the kinds of relations that obtain between a word and its meaning in compounds might just as well be the relations that obtain in lakṣaṇ . He might have reworked the jahat-sv rtha to apply in cases where the word does, indeed, give up its meaning in entirety, and took his cue from Patañjali’s refinement of jahat-sv rtha to propose the jahad-ajahal- lakṣaṇ , where the word gives up just a fraction of the scope of its meaning. Sarvajñātman himself, in fact, suggests as muchŚ “Accepting this signification function, those who are eminent knowers of the three Vedas made the threefold division of lakṣaṇ by formulating it in the technical terms of the language as jahat-sv rtha, ajahat-sv rtha, and jahad-ajahat-sv rtha, in 44 See Joshi 1968:9-10, 75ff. 504 the sūtra “An operation on inflected words happens when they are related [fit to be joined in a compound].”45 Be that as it may, let us note now that there were four kinds of secondary signification functions in his scheme, one gauṇa-vṛtti and three kinds of lakṣaṇ . We are justified in grouping them all under the category of “secondary signification function” in virtue of how they are designated through verbal formations: when a word is used in its literal meaning, the common practice is to say that something is ucyate, “said,” but more precisely “directly expressedś” when a word is used in a sense different from the literal meaning, regardless whether it is through the metaphorical or the relational signification function, the common practice is to say that something is lakṣyate, “indicated” or “figuratively expressed.”46 The same pair appears in the gerundive form, v cya and lakṣya. We move now to the second point in the preliminaries, and it concerns the reference of the categories of tat/brahman and tvam/aham. This reference is of two kinds, one which is directly denoted, primary or v cya, and one which is indicated, secondary or lakṣya. Sarvajñātman presents the direct references in both cases as “stained” or “dappled” entities, abala, consisting of a pure core and a surplus of stains or spots of the nature of adjuncts, up dhi.47 The pure core in the category of tat is Brahman that is a non-dual entity of the nature of bliss and consciousness. This Brahman, however, is in association with ignorance or avidy , in virtue of which, or rather through which as the immediate cause, it assumes the causal role in relation to the world. It is the Brahma-Sūtra Brahman, the entity from which proceed creation, 45 idam eva v ttir ayam a gīk tya samarthaḥ pada-vidhiḥ [2.1.1] ity atra sūtre jahat-svārthājahat-svārthā jahad- ajahad-svārtheti bhāṣā-pada-prakṣepe a traividya-v ddhāḥ lakṣa ā-vibhāgaṁ cakruḥ. PPr 1.13-4. 46 This is obvious, for instance, in the Second Chapter of PPr, where Sarvajñātman consistently uses the pair, having allowed in the First Chapter that Brahman be indicated either through gauṇa-vṛtti or through jahad-ajahal-lakṣaṇ . 47 pada-dvayārthau ca dvividhau, vācyau lakṣyau ca. tatra vācyau śabalauś lakṣyau śuddhau. PPr 2, p.18. 505 maintenance, and destruction of the world. From this Brahman dappled by ignorance the gross and subtle categories of creation proceed: from the five elements and down to the internal organ on the one hand, and Hira yagarbha and his existence as the various forms of pr ṇa on the other.48 This, of course, is the totality of k rya-brahman that we discussed in Chapter Six. The assumption of avidy on the part of Brahman through which Brahman becomes the creator and the product, the nimitta- and up d na-k raṇa, makes Brahman an external thing in relation to the Self that identifies with its products, and mediate, knowable in detail through scripture. Such mediacy, p rokṣya, is Brahman’s most stubborn stain. This is the entity denoted by the words tat and Brahman: a non-dual, mediate entity. The pure core, on the other hand, is not denoted, but indicated by the category.49 We should note that avidy by this time had replaced Śa kara’s n ma-rūpe as the cosmic power that provides the stuff for the world. The category of tvam/aham also has a pure core and a surplus of stains. The core is inner awareness or immediate consciousness, pratyak-caitanya, through which one most directly knows oneself. This inner awareness is the same in all entities, up to and including the gods, but it variously identifies with the gross and subtle categories of creation.50 It is the j va or the category of the individual Self. The very potentiality of identification with the categories of creation makes the j va a dual entity.51 We may think of this along the lines of the common viveka or discrimination procedures, whose very possibility is predicated on there being something extraneous to myself that I am not, yet of which I am necessarily aware for discrimination to make sense. This sense of duality is the principal stain of the category of 48 prā a-pi ḍa-kāra āvidyā-śabalaṁ advayānanda-caitanyaṁ brahma-śabda-vācyam. Ibid. 49 pārokṣya-sahitam advayānanda-caitanyaṁ brahma-śabda-vācyam. PPr 2, p.19. 50 prā a-pi ḍātmaka-kārya-śabalaṁ pratyak-caitanyam api daiva-paryantaṁ ahaṁ-śabda-vācyam. PPr 2, p.18. 51 sadvitīyaṁ pratyag-rūpam ahaṁ-śabda-vācyam. PPr 2, p.19. 506 tvam/aham. The primary denotation of the category expresses an entity that is a single inner awareness yet dual in character. As with the category of tat, the pure core is not directly expressed by tvam/aham, but indicated.52 I would like to note here briefly that the notion of Brahman being a spotted entity, abala, appears occasionally in prakaraṇa works that are attributed to Śa kara, most notably in the Pañc karaṇa. Śa kara never describes Brahman in this specific way, and so abala may be added as one of the minor criteria in deciding the authorship of works ascribed to Śa kara. The question now presents itself, through which of the four possible modes of secondary signification are the pure cores of the two categories indicated. Sarvajñātman proposes that both the metaphorical and the relational signification function are fit to play that role, but the acceptance of the first is certainly in deference to the authority of Sureśvara, whose Naiṣkarmya- Siddhi he quotesŚ “[The Self is indicated by the notion of “I”] because of being inner, very subtle, and conducive for the vision of the Self.”53 The innerness of the category of tvam is the important feature here, because it is the key shared quality with the pure core, but Sarvajñātman does not want just a procedure in which this pure core will be indicated: it is necessary that the surplus drops at the same time, and that is most clearly accomplished by the jahad-ajahal- lakṣaṇ . The construction of the mah -v kya notion is predicated on this function. The Ved ntic Mah -v kyas In Pañca-Prakriy , Sarvajñātman discusses ahaṁ brahm smi and tat tvam asi as, now explicitly, mah -v kyas, and there seems to be a reason why only these two. Ahaṁ brahm smi is the topic of Chapter Two of the treatise, which discusses more generally how knowledge of unity of 52 pratyak-caitanya-bhāgo lakṣyate. ibid. 53 pratyaktv d atisūkṣmatv d tma-dṛṣṭy-anu lan t [NaiS 2.55] iti gu a-yogād aham-ādi-śabdasya gau ī pratyag- ātmani v ttir a gī-k taiva. PPr 1, p.10. 507 Brahman with the Self, culminating in direct experience, arises in the qualified aspirant who had approached a teacher, one who turns out to be imagined through the student’s ignorance in the manner of dream objects, and who is surrounded by thousands of similarly imagined fellow brahmac rins. Issues such as the nature of and disagreements about the two types of liberation, absolute and living, are discussed. The mah -v kya, in other words, looks at the identity of the two categories from the side of the student and addresses questions that are pertinent to the student in whom knowledge arises and for whom liberation takes place. Chapters Three and Four discuss tat tvam asi and the notion of av ntara-v kya, and clearly the view of the mah -v kya is from the standpoint of the teacher: it is the teacher who should know how to organize the instruction around the two categories. The opening of Chapter Four is quite explicit about it, as it says that the av ntara-v kyas will be discussed for the benefit of the student who had duly approached the teacher, with the whole shebang of the adhik ra stuff that is familiar to us from the openings of the BSBh and the US.54 My object in emphasizing this is to bring home the point that there was no such thing as four Vedas cum four mah -v kyas doctrine in early Advaita Vedāntaś there wasn’t even a mah - v kya doctrine at all, in the sense we commonly think of it, namely as tat tvam asi from the 6th prap ṭhaka of the Ch ndogya or ahaṁ brahm smi from the 1st adhy ya of the Bṛhad- raṇyaka and situated strictly in their context. There was, rather, a mah -v kya context, a kind of Upaniṣadic statement that identifies Brahman the great cause and ground of Being with the inner Self of every creature that perceives itself as the cognitive agent. The mah -v kya context is a context of categories that is superordinate to any textual incarnation in which it may appear, and in which the only meaningful difference is that of perspective in a teacher-student relationship. 54 athātho ‘vantara-vākyārthaṁ vyākhyāsyāmaḥ, vidhivad upasannāya nityānitya-vastu-vivekādi-sādhana-catuṣ aya- sampannāya brāhma āya śrava a-manana-nididhyāsanādi-vidhi-preritāya yataye mukhyādhikāri e. PPr 4, p.38-9. 508 As I have been drumming for some time, in terms of this mah -v kya context, the tasm d v etasm d tmanaḥ of the Taittir ya provided the core of the doctrine, because it related the categories in the textual locus where the two were paradigmatically laid down. We will come back to this mah -v kya context. Now let us focus on what Sarvajñātman called the av ntara-v kyas, thus explicitly adopting the Mīmāṁsā mah ṭav ntara model. These are statements that pertain to the members of the mah -v kyas, the categories of tat and tvam, and refer to scriptural passages where they are individually described.55 Sarvajñātman calls them “sentential supplements of the mah -v kyas,” mah -v kya- eṣa, making it explicit that these statements are attached to the respective categories and serve their individual purposes in their own context, not the mah -v kya context.56 The av ntara-v kyas have a “scope” or an “extension,” parim ṇa, on each side of the two categories. On the side of the tat category, this scope is constituted by way of predication of “unrepeated qualities” to Brahman within nirguṇa passages. What Sarvajñātman means by “unrepeated qualities” is an idiosyncratic way to designate the positive and negative attributes that formed the notion of Brahman both in the Brahma-Sūtra and in Śa kara’s definition. They are “unrepeated” in the sense that they are not such qualities as satya-k ma and satya-saṅkalpa in the formation of Upaniṣadic saguṇa meditations by way of forging units out of the several passages where a respective meditation is described.57 They constitute the essential notion of Brahman. They are to be collected from the Upaniṣads of all Vedic branches, and special care should be made in the effort of gathering 55 tatrāvāntara-vākyas tāvad brahmātmaikatva-lakṣa a-mahā-vākyārthānvayi-tat-tvaṁ-padārtha-dvayam eva. PPr 4, p.40. 56 PPr 2, p.23. 57 nirgu a-brahma-para-samasta-veda-śākhopaniṣad-gatāpunar-ukta-samasta-padopasaṁhāre a. Ibid. 509 negative attributes, lest slacking may bring about an incomplete negation of the false points of identifications for the Self.58 In Pañca-Prakriy Sarvajñātman deals only with the av ntara-v kyas that present the positive attributes (vidhi) of Brahman, but in the Saṅkṣepa- r raka he also includes the negative attributes (niṣedha) and understands their purpose in the same way as Śa kara: they negate all possible points of identification for the Self within the sphere of fine and gross objects.59 Focusing on the positive attributes, the av ntara-v kyas are satyaṁ jñ nam anantam brahma and nando brahmeti vyaj n t from the Taittir ya 2.1.1 and 3.6.1, and there is nothing particularly different in this from Śa kara’s BSBh 4.1.2, except that Sarvajñātman is much more comfortable with nanda or bliss as an attribute of Brahman so as to stay within the Taittir ya setting, rather than go to the Bṛhad- raṇyaka. Although these are positive attributes, they present Brahman in terms of denial rather than affirmation: Brahman as a thing is opposed to unreal, insentient, limited things that are liable to suffering, which obviously constitutes the sphere of created things.60 This denial is accomplished through Brahman’s feature of being limitless, ananta, and that becomes very important for the next textual level, below the av ntara-v kyas. That is, for the purpose of teaching how this Brahman that is Being, consciousness, bliss is unlimited, different from the kinds of real, conscious, and pleasurable things that we are acquainted with, there are Upaniṣadic passages of five kinds. The first three are texts of creation, maintenance, and destruction of beings; the fourth are texts that describe Brahman entering its creation; the fifth are texts that describe how Brahman that had entered beings rules them from within. Sarvajñātman’s instances are all from the Taittir yaŚ “That from which these beings are 58 yadi punar na samāhara aṁ bhavet paramita-pariṣedhanam āpatet. SŚ 3.318. 59 SŚ 3.312-325. 60 an ta-jaḍa-paricchinna-duḥkha-viruddhaṁ vastu satya-jñānānantādi-śabdaiḥ brahma-śabdārthatvena nivedyate. PPr 4, p.40. 510 born, by which they live once they are born, and in which they return at death, that is Brahman, that is what you should strive to know distinctly”; “Once it had created them, it entered into them”; “It is out of fear from it that the wind blows and the Sun rises.”61 Now, Sarvajñātman calls such texts “arthav da passages that form supplements to the meaning of the affirmative and negational statements, like the statements of praise and censure.”62 His Mīmāṁsā discourse is at its peak here: he labels the av ntara-v kyas that contain the positive and negative attributes of Brahman vidhi and pratiṣedha, terms which Mīmāṁsakas used for injunctions and prohibitions, and he attaches to them Upaniṣadic passages that are arthav da, explicitly likened to the Mīmāṁsa arthav das that were passages that had no independent truth value, but served the purposes of the injunction or the prohibition that they latched on. The purpose that the Upaniṣadic arthav das serve is to facilitate reasoning into the possibility of Brahman’s being limitless, evidently through analogy, illustration, etc.63 There are obviously many points on which we could pause, examine what Sarvajñātman is saying in more detail or see how much his presentation follows Śa kara’sŚ the careful reader would have recognized all his building blocks in the previous chapters. However, that is irrelevant for our purpose here, which is to appreciate the relation between different kinds of Upaniṣadic passages and the hierarchy that obtains between them. There are texts that present Brahman’s characteristics, but these characteristics are not of the kind that is commonly known: they are unlimited, non-dual. The Upaniṣads facilitate the making sense of this through passages 61 tatra yato v im ni bhūt ni j yanteṬ yena j t ni j vantiṬ yat prayanty abhisaṁvi antiṬ tad vijijñ sasvaṬ tad brahmeti [TU 3.1.1]. iti s ṣ i-sthiti-pralaya-vākyāni. tat sṛṣṭv Ṭ tad ev nupravi at [TU 2.6.1] iti praveśa-vākyam. bh ṣ sm d v taḥ pavate, bh ṣodeti sūryaḥ [TU 2.8.1] ity-ādi-niyamana-vākyam. PPr 4, p.42. 62 tasyaiva brahma a tathā-nivedyamānasya ānantyopapādanāya pañca-vidhāni yukty-arthavāda-vākyāni vidhi- pratiṣedha-codanārtha-śeṣa-bhūta-stuti-nindārthavāda-vākya-vat s ṣ i-sthiti-pralaya-praveśa-niyamana-vādīni. PPr 4, 41-2. 63 etaiḥ pañca-vidhaiḥ sambhāvanārthavāda-vākyaiḥ ... PPr 4, p.42. 511 that, for instance, talk about creation, but do not intend to affirm creation. The av ntara-v kyas have arthav das under them. There are av ntara-v kyas on the side of the category of tvam as well, and their scope is, as we would expect, the pañca-ko a section of the Taittir ya.64 They serve the purpose of drawing out the inner Self that had entered the heart of beings by gradually reducing its scope, as we saw in Chapter Eight. Another procedure that can be applied for the same gain is the examination of the three states of awareness—waking, dream, and deep sleep—to find out that while the states change, the Self itself does not. Sarvajñātman’s reference point for this is Aitareya 1.3.13, tasya traya vasth s traya svapn ḥ, “These are its three dwellings, three sleeps,” but obviously the model would have been Yājñavalkya’s teachings to Janaka in the BĀU. The upshot of these two procedures is that the category of tvam stands for the inner Self, different from and unrelated to the three states, single yet ensouling all animate things in the world, from individuals and respective divinities to the world itself.65 At this point, at the completion of the understanding of tvam through its av ntara-v kyas, the principle that we have been pursuing through the course of the dissertation kicks in: all this is empty talk unless it serves some good of man. As Sarvajñātman puts it, “the purport of narrating about the three states of awareness, as in ‘These are its three states, three sleeps,’ does not serve a good of man in its own meaning.” Through the phala- ruti of Vedānta, the Taittir ya text “the knower of Brahman attains the highest,” it follows that such good of man is contingent of the unity of Brahman and the Self. Further, through the rule that we discussed in the Second Chapter in connection with the r d-upak rakas—a scriptural passage that is not related to a result on its 64 TU 2.2-2.5. 65 tathā jāgrat-svapna-suṣupti-vilakṣa o jāgrat-svapna-suṣupti-sambandha-rahito ‘dhyātmādhibhūtādhidaiva- śarīrastha ekas tvaṁ-padārthaḥ. PPr 4, 45. 512 own attaches to a proximate text that is—the passages that discuss tvam, such as those about the three states of awareness, seek the context of the mah -v kya, tat tvam asi as the statement through which the human good that is unity of Brahman with the Self is realized.66 We have, thus, gone a step above the av ntara context and reached the mah -v kya. Sarvajñātman had made several jumps, of course, some of the Urukrama kind, in coordinating texts from various places: his reference text about the three states of awareness is in the Aitareya, and the statement of result is the Taittir ya, and there is no real proximity as required by rule. Still, the pañca-ko a doctrine is in the Taittir ya, and that was a passage with an identical purpose, establishing the category of tvam. Tat tvam asi is, of course, in the Ch ndogya, far from the Taittir ya brahma-vid pnoti param, but Śa kara said, we will remember, that tat tvam asi and tasm d v etasm d tmanaḥ from the Taittir ya had the same meaning. Further, the Aitareya statement on the three states is relatively proximate to prajñ naṁ brahma, an identity statement like tat tvam asi. And, in culling the av ntara-v kyas for the category of tat, the procedure was to track down relevant passages from Upaniṣads of all Vedic branches. This provides a network in which everything is proximate to everything else, one way or another. In the mah -v kya context, now, the two categories are identified, that is, their primary references are identified. To be specific, Brahman which is characterized as limitless Being, consciousness, bliss, that is, which is opposed to unreal, insentient, limited things that are liable to suffering, but is a mediate, external entity known through scripture, is identified with the inner Self, which, because of its innerness, is “dual,” having things that are present to it as objects of awareness. The two categories obviously exclude one another in two respects: Brahman is non- 66 tasya traya vasath ḥ trayaḥ svapn ḥ ity-ādi-śrutau avasthā-trayasyopanyāsasya tātparyaṁ svārthe puruṣārthābhāvāt, brahmātmaikatva-jñāne ca brahmavid pnoti param [TU 2.1.1] iti puruṣārtha-śrava āt, phalavat- sannidh v aphalaṁ tad-aṅgam iti nyāyāt. PPr 4, p.45-6. 513 dual, whereas the Self is dual; and, Brahman is mediate, whereas the Self is immediate, known privately. Something’s gotta giveŚ for non-exclusion to obtain, mediacy drops from the scope of meaning of Brahman, and duality drops from the scope of meaning of the Self. This happens through the secondary relational signification function, jahad-ajahal-lakṣaṇ , in which the respective words indicate a single reference through eliminating the mutually exclusive parts. Thus, the two words, through directly expressing their respective meanings, indirectly denote a single entity that is defined as an inner Brahman, that is, Brahman whose nature is eternal, pure, conscious, free, real, supreme bliss, non-dual, and at the same time inner awareness.67 This is the Brahman that is the synonym for absolute liberation. I would like to draw now the attention to the end of Chapter Seven, where we concluded the preliminaries of liberation and saw how the whole Vedic corpus, without the explicitly k mya sections, was for liberation through mediate causality, p ramparya. We saw that unity obtained between the texts of knowledge and action through the vividiṣ statement in BĀU 4.4.22 and the principle of v kyaikav kyat or unity of purpose of distinct texts. Sarvajñātman does not talk about this in the PPr, but the theme does appear indirectly in his statement of adhik ra: before one should approach a teacher, one must be freed from all impurities, obtaining thus the results available through the karma-k ṇḍa section of the Veda, and then renounce all action.68 Introducing this as our final consideration in the Vedāntic mah -v kya doctrine, I offer 67 ato ‘sad-an ta-jaḍa-paricchinna-duḥkha-viruddhaṁ yat satya-jñānānantānanda-lakṣa aṁ brahma tat tvam asi iti mahā-vākye tac-chabdena svavācya-śabala-brahma-gata-pārokṣyāṁśa-parityāgena lakṣayitavyam. yathoktaś ca pratyag-ātmā tvaṁ-śabda-vācya-śabala-gata-sadvitīyāṁśa-parityāgena svavācya-sadvitīyatva-śabala-tvaṁ- padārthābhidhāna-dvāre a tvaṁ-śabdena lakṣayitavyaḥ jahad-ajahal-lakṣa ayāś anyathā sadvayam advayam, pratyakṣaṁ parokṣam—iti tat tvam asi mahā-vākyārtha-pratipattau virodha-prasa gāt. evaṁ padārtha-dvayaṁ śodhayitvā vyavasthitaṁ tad eva tat tvam asi ity ācāryo bodhayati. PPr 4. p.46-8ś idānīm antaḥ-kara ābhāvāt jāgrat- svapna-suṣuptīnām abhāve jīvatva-varjita-pratyagātma-caitanya-svabhāvo nitya-śuddha-buddha-mukta-satya- paramānandādvaya-svabhāvaṁ brahma tat-tvam-padārthau pariśiṣ au. PPr 3, p.36. 68 yajñādi-kṣapita-kalmaṣasya sarva-karma-kā ḍa-phala-bhūtasya sarva-karma-sannyāsinaḥ. PPr 2, p.20. 514 the following schematic representation of the hierarchy and structure of the notions and texts we covered: ______________________________________________________________________________ Scheme 1: Mahā-vākya Structure mah -v kya contextŚ tat tvam asi, ahaṁ brahm smi Reference [through jahad-ajahal-lakṣaṇ ]: Brahman as eternally pure, awake, free, real, supremely blissful, non-dual awareness. (nitya- uddha-buddha-mukta-satya-param nand dvaya-cit-pratyag-brahma) Statement of result: brahma-vid pnoti param (TU 2.1), relates the two categories through purpose av ntara-vakya context (mahā-vākya-śeṣa): tat-pad rtha tvam-pad rtha positive and negative (vidhi-pratiṣedha) pañca-kośa (TU 2.2-5) satyam jñ nam anantam brahma (TU 2.1.1) three states of awareness (BĀU) nando brahmeti vyaj n t (TU 3.6.1) reference (parokṣa, advaya): reference (pratyakṣa, sadvaya): Brahman that is opposed to the non-existent, unreal, The inner Self, without adjuncts, pure awareness, dull, limited, suffering, characterized as being, one in all creatures consciousness, unlimited, bliss (asad-an ta-jaḍa-paricchinna-duḥkha-virudhaṁ yat satya-jñānānantānanda-lakṣa aṁ brahma) Arthav da (avāntara-vākya-śeṣa), five types: (1-3) creation statements (s ṣ i-sthiti-pralaya-vākya) yato v im ni bhūt ni j yanteṬ yena j t ni j vantiṬ yat prayanty abhisaṁvi antiṬ tad vijijñ sasvaṬ tad brahmeti (TU 3.1.1) (4) entrance statements (praveśa-vākya) tat sṛṣṭv Ṭ tad ev nupravi at (TU 2.6.1) (5) regulation statement (niyamana-vākyam) bh ṣ sm d v taḥ pavate, bh ṣodeti sūryaḥ (TU 2.8.1) Qualifications context: Statement of adhik raŚ MU 1.2.12-13 Principles of causality and manner of achieving adhik ra (vividiṣ )Ś BĀU 4.4.22, 23 (1) P ramparyaŚ the performance of ritual and meditation without desire, for personal purity, culminating in dispassion (4.4.22) (2) Samuccaya: dispassion and the practice of yama-niyama (4.4.23) _____________________________________________________________________________ Looking at the scheme and the foregoing analysis in terms of textual structure and cohesion, it is immediately apparent that tat and tvam in the mah -v kya context are just a demonstrative and personal reference respectively to their av ntara-v kyas, insofar as their interpretation, not in the theological sense but purely on the level of meaning, is impossible without reference to texts where they are elaborated. The pronouns stand in a cohesive relation to their av ntara-v kyas through semantic identity, and the full sentences can be introduced in place of the pronouns by way of substitution. The av ntara-v kyas on their part are the 515 requirement for the resolution of the personal pronouns, but they are also presuppositions with their own requirements, insofar as the attributes of Brahman are incomprehensible without recourse to Upaniṣadic passages where they are discussed. In Sarvajñātman’s packed account, anantam was the ground for having recourse to such passages, but it was really the juxtaposition of anantam to the other three characteristics, as we saw in Śa kara’s account of the identity statements. “What do you mean, consciousness unlimited? – Look at the bhūma-vidy . – Being unlimited? The sad-vidy .” This was also possible through semantic identity, as satyam and sat from the Taittir ya and the Ch ndogya have the same meaning, as Śa kara said. The mah -v kya itself, then, is really an ellipsis for a much longer statement that joins the respective meanings of the two categories such that they are purged of their mutually exclusive elements, yet at the same time a mandatory ellipsis since the mutually exclusive elements need to be simultaneously meaningful. The final reference that obtains through the juxtaposition of the meaning of the counterparts, in which Brahman is no longer mediate and the Self is no longer dual, is not informative in the same way as when Brahman is both mediate and immediate and the Self is both dual and non-dual: it is in that liminal state that liberation takes place. Thus, the mah -v kyas are simultaneously pithy and massively long, through subsuming and requiring the av ntara and the arthav da context. Add to this the prerequisite context that provides for the very possibility of understanding the maha-v kyas, and they can potentially involve everything in the Veda that can be used in the pursuit of liberation. In the mah -v kya, the full finality of meaning of the Veda obtains. Before moving to the conclusion, I want to hazard the following idea. The mah -v kyas are, of course, scriptural statements, part of the Upaniṣadic corpus. Their expanded form, however, must be worked out by tracing the cohesion relations that obtain between the terms on 516 the mah , av ntara, and arthav da level, and through the vividiṣ between the jñ na and karma corpus, and by providing a structure that does not appear naturally, in the sense that there is no scriptural text as such that states the expanded form explicitly. As we learned from the M m ṁs -Ko a, a mah -v kya is an inferable thing. To go back to the Second Chapter, now, scripture is like a grand repository of empirically unavailable data that one can tap into contingent on one’s aspirations and abilities—“The Veda gives rise to veridical cognitions for a competent person,” as Sarvajñātman says69—but then do stuff with this data, explore how it can be structured for one’s needs. From such perspective, mah -v kyas are, like the ritual manuals or prayoga-vidhis, what that one must work out on one’s own or through a model. In this sense, a mah -v kya is something like the prose section of the US, a structure of scriptural data around the identification of the two categories, meant for attaining liberation. Conclusion To state now the obvious, in formulating the notion of Upaniṣadic mah -v kyas, Sarvajñātman took his bearings fully and explicitly from the Mīmāṁsā model. This was a model of hierarchy of scriptural statements on three levels, mah , av ntara, and arthav da, each lower level subsumed, terminated, and finalized in the level above. The fact needs emphasis, because we may be tempted to think that mah -v kya, even if Śa kara or Sureśvara and Padmapāda did not mention the lexeme, was just a change in nomenclature. This is what Murty’s account implicitly does: so long as Advaitins talk about tat tvam asi and such, they are talking about mah -v kyas, never mind the name. But, pursuing this line does not help us see why these statements are mah and not just identity statements. 69 adhik riṇaḥ pramiti-janako vedaḥ iti nyāyāt. PPr 2, p.21-2. 517  Sarvajñātman needed to pull strands from several sources to turn the identity statements into mah -v kyas. From Mīmāṁsā, he took the model organization of texts—the structure—and the key Mīmāṁsā mah -v kya characteristic: finality of meaning that does not cancel the independence and meaningfulness of the member statements in their context, but does so in its own context, like in the case of precedence of arithmetic operations. Sentences in general, mah - v kyas included, were, Kumārila said and Śa kara followed suit, like a cart whose wheels are distinct yet not independent, being unable to run on their own.70 Śa kara had already provided Sarvajñātman with the ideas that the categories are, really, full sentences or a distillate of scriptural passages, such as the pañca-ko a, and that the creation texts in the Upaniṣads are arthav das that do not have individual truth value but do have a purpose. These were the building blocks to fill in the structure. The glue might just have come from Patañjali’s Mah - Bh ṣya, which discussed how words in forming compounds keep part of their meaning but drop whatever is mutually exclusive. Patañjali talked not about lakṣaṇ but about vṛtti in a different sense, but it was easy to use his idea under the assumption that lakṣaṇ happens when words are juxtaposed. Historically, the doctrine of mah -v kya developed not around tat tvam asi and ahaṁ brahm smi, as we commonly think, but around a Taittir ya core. There are two reasons for this. The first was that the Taittir ya provided the normative definition of Brahman, the general category of the great cause from which creation proceeds, as Being, consciousness, bliss, limitless, and it was easy to seek the identity where Brahman was fully defined. One Taittir ya statement was important in this regard, tasm d v etasm d tmanaḥ, important because it looked anaphorically right back at the definition of Brahman, yet cataphorically at the creation of beings 70 ŚV V kya 148. 518 and the entrance of Brahman itself into them as the Self. This was not the paradigmatic mah - v kya, but it was historically the most fertile identity statement. The second reason is a bit more arcane. The identity statements doctrine that evolved into mah -v kya developed as a response to prasaṅkhy na-v da and the two problems of mediacy of scriptural knowledge and the relational nature of language. The scriptural basis of this doctrine was the BĀU 4.4.21 statement, “Having learned, one should cultivate insight, without pondering much over words, because that just tires the voice.” Prasaṅkhy na-v dins interpreted this to the effect that words did express Brahman, just not immediately, for which reason meditation was required. The same problem was, curiously, touched upon in the famous Taittir ya verse, yet to a very different effectŚ “Whence words return along with the mind, not attaining it, he who knows that bliss of Brahman, fears not at any time.”71 The Taittir ya verse rejected not only words, but also the capacity of the mind the instrument of meditation to reach Brahman. Śa kara based his account of language as indicting Brahman through the secondary signification function on this scriptural locus, and although the verse is oddly out of place at first sight, without any justification in its context, the Advaita doctrine of mah -v kya based on the Taittir ya core that occupies the space around the verse addressed precisely that issue. 71 TU 2.4.1, 2.9.1. 519  EPILOGUE “’But you realized it was only the effect of the green honey …’ ‘Yes, the vision was an illusion, but what I now felt inside was not; it was true desire. When you feel it, it’s not an illusion. It’s real.’”1 In the Introduction, I set as my goal to show that Śa kara’s significance for Indian intellectual history—not the great cultural hero Śa kara, but the hero of Padmapāda, Sureśvara, and their likes—can best be appreciated if we approach him as a theologian of liberation. Śa kara has been described as many things, but two epithets have stuck with him: that of a great philosopher, and that of a mystic. As for the first, his brilliant synthesis notwithstanding, the truth is that there was not a whole lot original to Śa kara’s philosophy. As Daniel Ingalls had claimed in his paper “The Study of Śaṁkarācārya,” Śa kara’s philosophy was a fix (of the Gauḍapāda or, perhaps, Bhart hari kind) on the doctrine of bhed bheda, and now we know that this bhed bheda was that of Bhart prapañca, whose categories of psychology and cosmology Śa kara inherited. Parallel to his being a great philosopher, Śa kara was habitually described as a mystic or a philosopher of experience, by Indian scholars who wanted to ground scripture and philosophy in personal experience, and in non-specialist accounts in the West, particularly in the studies of mysticism, where he has become one of the favorite characters. In the West, such depiction was partly the legacy of Rudolph Otto’s book Mysticism East and West, whose influence had persisted even in Indology through Paul Hacker, and partly a result of the more general approach to the study of religion focused on mystical experience or intuition as its source, the legacy of William James and Henri Bergson. To illustrate briefly this perspective of Śa kara, we may quote from Otto: But above all Śa kara holds that knowledge based on scripture is merely the finger which points to the object and which disappears when it is itself looked upon. The real knowledge is that which he calls “one's own vision” — darśanam. This vision for him, as knowledge for Eckhart, is not a matter of “having visions.” It is rather an awareness of identity with Brahman, and that as an “intuitus,” a dawning of insight, 1 Umberto Eco, Baudolino, 2002:91. 520  our own clear-sighted realization of that which the scriptures taught. This awareness cannot be “produced,” we cannot reason it out. It is not a “work.” It comes or does not come independently of our will. It must be seen. The way may be prepared by the words of the Vedas and by meditation (pratyaya) on them, but in the end it must be our own vision.2 While more recent scholarship such as the work of Anantanad Rambachan has done much to discredit this characterization of Śa kara as a mystic or a philosopher of experience, Śa kara’s ideas about scripture, intuition or personal experience, and liberation have never been fully studied in their context. Scholars have learned to read Śa kara carefully about the details of his system, but less so to appreciate his major participations and interventions in a preexisting discourse, to read him in a “properly historical style.” It is, however, the attribution of mysticism to Śa kara, examined in his context, rather than his being a great philosopher, that helps us pinpoint his significance, because the most prominent aspect of his theology was liberation, and mysticism was precisely what he sought to distinguish his soteriology from. The striving after personal experience that was over and above scriptural knowledge was very much a part of Śa kara’s intellectual universe, and certainly his own concern. It was expressed in the Yoga-Sūtra, 1.49Ś “It [seedless sam dhi] has a different focus from that of inference and scripture, because it has the particularity of things as its object.”3 Yoga was about direct experience, Yogic perception, an “intuitus,” and precisely this was its major distinguishing feature from Sā khya, whose Self was known not as an object, but as the light of consciousness known from analogical reasoning, and from the scriptural theologies of Mīmāṁsā and Vedānta. But such Yogic perception made an alliance with Vedic theology, most prominently in the doctrine of prasaṅkhy na, which claimed that scriptural 2 Otto 1970:51. 3 Translation Bryant (2009:159). 521 knowledge of Brahman was general, mediate, and had to culminate in direct experience through long, cathartic meditation as another, immediate cognition. This was the thesis that Śa kara repudiated. While the pursuit of liberation had to terminate in personal experience, this “intuitus” or “dawning of insight” or “vision” precisely had to be reasoned out, by means of the two forms of reasoning, theological or scriptural exegesis ( ravaṇa) and philosophical reflection (manana) based on analogy (s m nyato-dṛṣṭa), the characteristic methods of Mīmāṁsā and Sā khya, not of Yoga. While scripture was “merely the finger,” the finger was all that one hadŚ the knowledge of one’s being Brahman was beyond scripture and words, yet available solely through words. This was the “hermeneutics of liberation” that I mentioned in the Introduction, the deciphering of a hidden meaning with an expert in the techniques of anamnesis, upon which something additionally is known through symbols, but cannot be known otherwise because it is inexpressible except through symbols. There was no—and there could be no—insight that was different from the scriptural cognition once such cognition had been fully clarified through reasoning: one could not but have the personal experience, because this personal experience, avagati, was the Self, and the real fruit of the soteriological process was full, perfect dispassion, “being the Self” that was unrelated to anything that is not the Self and was known only negatively. That one had the insight was a “visible result,” visible in the perfect absence of identification and in appropriate behavior. And, it was visible to the teacher, much like the favorite joke of psychologists where a young practitioner of behavioral analysis asks his partner after a night of passionate loveŚ “You enjoyed it a lot. How about me?” It was the teacher in the Upade a-S hasr who had the last word on liberationŚ “You have thus attained fearlessness, and from now on you will not experience the suffering of waking and dream. You are freed from the 522 misery of transmigration,” and it was Yājñavalkya who had to tell the kingŚ “You have attained fearlessness, Janaka.” A special insight was impossible where one could know the Self only through what the Self was not, a non-object that was, nevertheless, the residue that was irreducible because without it no reduction could ever be possible. The final teaching was neti neti, the reference of the mah -v kyas. This, as we saw, was a characteristically Sā khyan mode of knowing, an intellectual reflection that was essentially opposed to performing ritual and that culminated in perfect disassociation from all possible identification points, but not in a different kind of experience. The full understanding, however, was predicated on personal purity, and this was another point on which Śa kara’s soteriology was different from the rest of Vedic theology. In the doctrine of prasaṅkhy na, meditation itself was cathartic, and the insight into one’s being Brahman that transformed mediate knowledge to immediate was a result of purification, which was accelerated by the performance of ritual. Śa kara turned this upside-down: such purification had to happen before one could possibly engage in the intellectual inquiry into Brahman, and in arguing this, Śa kara drew on the Mīmāṁsā intricacies of ritual technology to develop a new model of soteriological causality, p ramparya or successive causality. This model mapped the progress on the path to liberation by delineating attainments that turn into means upon their achievement, subsuming but also terminating their respective means. At the center of this novelty were the identity statements of the Upaniṣads that became the mah -v kyas. As realized by Hiriyanna almost a century ago, the central role of the identity statements was a new thing in Vedic theology, where meditation on Brahman, just like ritual, was organized around the Upaniṣadic injunctions such as vijñ ya prajñ ṁ kurv ta. By placing the onus on the identity statements and by interpreting them as affirming real rather than 523 meditative identity, Śa kara solved the key problem that was faced by the doctrine of prasaṅkhy na, the mediacy of scriptural knowledge. There was no question of the knowledge being mediate so long one recognized that it was about oneself. Sure enough, personal purity was required for such recognition, but the recognition itself was consequent on knowing qua knowing or qua clarification, not on repeated meditation on Brahman that reconstructs the subject and deconstructs itself in the process. To return to Bergson’s example of the character in a novel that I mentioned at the end of Part Two, for Śa kara the intuitive knowledge did not require for one to “coincide with the person” of the character through a form of sympathyŚ it had to dawn on one, rather, that “I am the character, the book is about me.” What was required, then, was not sympathy, but anamnesis. This, finally, meant that the mediate knowledge of Brahman from the Upaniṣads, expressed in statements such as satyaṁ jñ nam anantaṁ brahma, became part of what Sarvajñātman called, following Mīmāṁsā philosophy of language, av ntara-v kya, knowledge expressed in intermediate sentences that becomes finalized in the mah -v kya. Such was not the case in prasaṅkhy na-v da, as Hiriyanna recognized, and it is no wonder that the identity statements played no major role in Ma ḍana’s Brahma-Siddhi. The mah -v kyas gave the finality of meaning that, otherwise, the prasaṅkhy na meditation was supposed to provide. In short, the characteristically Sā khyan mode of knowing, the p ramparya model of soteriological causality, and the doctrine of the identity statements formed Śa kara’s theology of liberation that, I claim, was his major significance. Meditation on Brahman and direct, Yogic perception, came back to Advaita Vedānta through Vācaspati Miśra and his synthesis of Śa kara and Ma ḍana Miśra. It came back to haunt Śa kara’s legacyŚ he may have converted the man, all right, but he forgot to convert his ghost. 524  BIBLIOGRAPHY ABBREVIATIONS AiU – Aitareya Upaniṣad AiUBh – Śa kara’s Bh ṣya on the Aitareya Upaniṣad ĀDhS – pastamba Dharma-Sūtra ĪU – Upaniṣad ĪUBh – Śa kara’s Bh ṣya on the Upaniṣad US – Śa kara’s Upade a-S hasr USG – Śa kara’s Upade a-S hasr , the prose (gadya) portion USP – Śa kara’s Upade a-S hasr , the verse (padya) portion KUPBh – Śa kara’s Pada-Bh ṣya on the Kena Upaniṣad KUVBh – Śa kara’s V kya-Bh ṣya on the Kena Upaniṣad K U – Kaṭha Upaniṣad KṣU – Kauṣ taki Upaniṣad ChU – Ch ndogya Upaniṣad ChUBh – Śa kara’s Bh ṣya on the Ch ndogya Upaniṣad GDhS – Gautama Dharma-Sūtra - Kumārila’s Ṭup-Ṭ k on Śabara’s M m ṁs -Sūtra-Bh ṣya TU – Taittir ya Upaniṣad TUBh – Śa kara’s Bh ṣya on the Taittir ya Upaniṣad TUBhV – Sureśvara’s V rttika on Śa kara’s Bh ṣya on the Taittir ya Upaniṣad TV – Kumārila’s Tantra-V rttika on Śabara’s M m ṁs -Sūtra-Bh ṣya NaiS – Sureśvara’s Naiṣkarmya-Siddhi NS – Ny ya-Sūtra of Gautama 525 NSBh – Vātsyāyana’s Bh ṣya on the Ny ya-Sūtra PUBh – Śa kara’s Bh ṣya on the Pra na Upaniṣad PP – Padmapāda’s Pañca-P dik PPr – Sarvajñātman’s Pañca-Prakriy PYS – P tañjala-Yoga- stra, name for the Yoga-Sūtra along with the Bh ṣya traditionally attributed to Vyāsa BĀU – Bṛhad- raṇyaka Upaniṣad BĀUBh – Śa kara’s Bh ṣya on the Bṛhad- raṇyaka Upaniṣad BĀUBhV – Sureśvara’s V rttika on Śa kara’s Bh ṣya on the Bṛhad- raṇyaka Upaniṣad B - Kumārila’s Bṛhat-Ṭ k BDhS – Baudh yana Dharma-Sūtra BS – Brahma-Sūtra attributed to Bādarāya a BSBh – Śa kara’s Bh ṣya on the Brahma-Sūtra BrS – Ma ḍana Miśra’s Brahma-Siddhi BhG – Bhagavad-G t BhGBh – Śa kara’s Bh ṣya on the Bhagavad-G t Manu – Manu Smṛti MāU – M ṇḍukya Upaniṣad MU – Muṇḍaka Upaniṣad MUBh – Śa kara’s Bh ṣya on the Muṇḍaka Upaniṣad MK – Kevalānanda Sarasvatī’s M m ṁs -Ko a MS – M m ṁs -Sūtra attributed to Jaimini MSŚBh – Śabara’s Bh ṣya on the M m ṁs -Sūtra 526 MH – Bhavya’s Madhyamaka-Hṛdaya VDhS – Vasiṣṭha Dharma-Sūtra SK – S ṅkhya-K rik of Īśvarak ṣ a SKG – Gauḍapāda’s Bh ṣya on the S ṅkhya-K rik SŚ – Sarvajñātman’s Saṅkṣepa- r raka SSS – Suvarṇa-Saptati- stra commentary on the S ṅkhya-K rik SD – S hitya-Darpaṇa of Viśvanātha Kavirāja ŚU – vet vatara Upaniṣad ŚDV – aṅkara-Dig-Vijata, attributed to Vidyāra ya ŚB – atapatha Br hmaṇa ŚBh – Rāmānuja’s r bh ṣya commentary on the Brahma-Sūtra ŚV – Kumārila’s loka-V rttikaś it is always followed by the chapter title, for instance ŚV abda refers to the abda-Pariccheda chapter SV – Sureśvara’s Sambandha-V rttika (commentary on Śa kara’s introduction to the Bṛhad- raṇyaka-Bh ṣya and the introductory part of Sureśvara’s V rttika thereon) YS – Patañjali’s Yoga-Sūtra YSBh – The Bh ṣya on the Yoga-Sūtra attributed to Vyāsa VAM – Śālikanātha’s V ky rtha-M tṛk , eleventh chapter of his Prakaraṇa-Pañcik VK – Śrīnivāsa’s Ved nta-Kaustubha commentary on the Brahma-Sūtra VPS – Nimbārka’s Ved nta-P rij ta-Saurabha commentary on the Brahma-Sūtra VPV – Vṛtti on Bhart hari’s V kyapad ya 527 LIST OF PRIMARY SOURCES (1) Generally, prose sources, including commentaries, are quoted with text and page number. In multivolume editions, the volume and page are given together. For instance, MSŚBh 7.2.5, II.232 stands Śabara’s Bh ṣya on the M m ṁs -Sūtra, on sūtra 5 of the second p da of the seventh adhy ya, on page 232 of the second volume. (2) Sūtras and verses, as well as prose sections of the Upaniṣads and the prose part of Śa kara’s Upade a-S hasr , are quoted only by text number. (3) When more than one editions are consulted, the references are given according to the first edition listed. (4) Translations are listed after the editions. When multiple translations are given, they are listed in order of preference. In all cases, however, translations in the text are mine, unless specified otherwise. Ānandagiri. Upade a-S hasr -Vivṛti. Text in Tripā hī 2003. ———. Taittir ya-Upaniṣad-Bh ṣya-Ṭ k . Text in Pa ḍita 1889. ———. Ṭ k on Śa kara’s Bṛhad- raṇyaka-Upaniṣad-Bh ṣya. Text in Āgaśe 1891. ———. stra-Prak ik on Sureśvara’s Bṛhad- raṇyaka-Upaniṣad-Bh ṣya-V rttika. Text in Āpte 1892. ———. Brahma-Sūtra-Bh ṣya-Ṭ k . Text in Śāstri 1890. pastamba Dharma-SūtraṬ Text and Translation Olivelle 2000. Aitareya Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Īṣvarak ṣ a. S ṅkhya-K rik . Text and Translation in Jha 1896; Mainkar 1972. Kaṭha Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Kālidāsa. Kumara-Sambhava. Text and Translation in Smith 2005. Kauṣ taki Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Bodewitz 2002; Radhakrishnan 1992. 528 Kumārila Bha a. Bṛhat-Ṭ k Ṭ Selection in Taber 2007. ———. Tantra-V rttika. Text in Abhyankar and Joshi 1970ś Gosvāmī 1984. Translation Jha 1983. ———. Ṭupṭ k . Text in Abhyankar and Joshi 1970. ———. loka-V rttika. Text in Rai 1993. Translation in Jha 1907. Except: Pratyakṣa-Pariccheda. Text and Translation in Taber 2005; Codan -Sūtra. Text in Kataoka 2011a. Translation in Kataoka 2011b. Kena Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Gauḍapāda. gama- stra. Text and Translation in Bhattacharya 1989. Gauḍapāda. S ṅkhya-K rik -Bh ṣya. Text and Translation in Mainkar 1972. Gautama. Ny ya-Sūtra. Text in Āpte 1922. Gautama Dharma-SūtraṬ Text and Translation Olivelle 2000. Ch ndogya Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Jaimini. M m ṁs -Sūtra. Text in Abhyankar and Joshi 1970; Ramanatha Sastri and Subrahmanya Sastri 1934. Translation in Jha 1933. Jñānottama. Naiṣkarmya-Siddhi-Candrik . Text in Hiriyanna 1980. Taittir ya Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Nimbārka. Ved nta-P rij ta-Saurabha. Text in Brahmacārin 1904. Translation in Bose 2004. Nīlaka ha Caturdhara. Bh rata-Bh va-D p Ṭ Text in Mah bh rata 1929. Patañjali. P tañjala-Yoga- stra. Text and Translation in Hariharānanda Āra ya 2000; sūtra translation in Bryant 2009. Patañjali. Vy karaṇa-M h bh ṣya (on Aṣṭ dhy y 2.1.1). Text and Translation in Joshi 1968. Padmapāda. Pañca-P dik . Text in Bhāgavatāchārya 1891. Translation in Venkataramiah 1948. Pā ini. Aṣṭ dhy y . Text and Translation in Katre 1989. 529 Pārthasārathi Miśra. Ny ya-Ratn kara on Kumārila’s loka-V rttika. Text in Rai 1993. Prabhākara Miśra. Bṛhat . Text in Ramanatha Sastri and Subrahmanya Sastri 1934. Pra na Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Bādarāya a. Brahma-Sūtra. Text in Brahmacārin 1904. Baudh yana Dharma-SūtraṬ Text and Translation Olivelle 2000. Bṛhad- raṇyaka Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Bhagavad-G t . Text and Translation in van Buitenen 1981. Bhart hari. V kyapad yaŚ Kā ḍa 1 with Vṛtti attributed to Hariv ṣabha. Text in Subramania Iyer 1966. Kā ḍa 3. Text in Subramania Iyer 1973. Bhavya. Madhyamaka-Hṛdayam. Text in Lindtner 2001b. Translation of M m ṁs -Tattva- Nirṇay vat ra (chapter nine of Madhyamaka-Hṛdaya) in Lindtner 2001a. Bh gavata Pur ṇa. Text in Shastri 1983. Bhāskara. Brahma-Sūtra-Bh ṣya. Text in Kato 2011; Pandit 1915. Ma ḍana Miśra. Brahma-Siddhi. Text in Sastri 1984ś Tripā hī 1999. Manu Smṛti. Text and Translation Olivelle 2005. Mah bh rata. Text in Mah bharata 1929 and The Mah bh rata 1927. Mā harācārya. M thara-Vṛtti on the S ṅkhya-K rik . Text in Śarmā 1970. M ṇḍukya Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Muṇḍaka Upaniṣad. Text in Limaye and Vadekar 1958. Translation in Olivelle 1998; Radhakrishnan 1992. Yāska. Nirukta. Text in Śarma 1990. Rāmānuja. r bh ṣya. Text in Abhyankar 1914. Translation in Karmarkar 1959; Thibaut 1904. Rig Veda. Text in Nooten and Holand 1994. Translation in Brereton and Jamison 2014. 530 Vasiṣṭha Dharma-SūtraṬ Text and Translation Olivelle 2000. Vācaspati Miśra. Tattva-Kaumud . Text and Translation in Jha 1896. ———. Bh mat . Text in Śāstri 1938. Vātsyāyana. Ny ya-Sūtra-Bh ṣya. Text in Āpte 1992. Vidyāra ya. Pañcada Ṭ Text and Translation in Swami 1967. ———. aṅkara-Dig-Vijaya (attributed). Text in Āpte 1981. Translation in Swami 1978. Viśvanātha Kavirāja. S hitya-Darpaṇa. Text in Pandit 1936. Śa kara Bhagavatpāda. Aitareya-Upaniṣad-Bh ṣyaṬ Text in The Works of Sri Sankaracharya, 1910, volume 5ś Āpte 1931. Translation in Swami 1937b. ———. -Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 4. Translation in Swami 1937a. ———. Upade a-S hasr . Text in Mayeda 2006a; Tripā hī 2003. Translation in Mayeda 2006b; Alston 1990; Swami 1949. ———. Kaṭha-Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 4. Translation in 1937a. ———. Kena-Upaniṣad-Pada-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 4; Āgaśe 1909. Translation in Swami 1937a. ———. Kena-Upaniṣad-V kya-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 4; Āgaśe 1909. ———. Ch ndogya-Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volumes 6-7; Āgaśe 1934b. Translation in Swami 2003. ———. Taittir ya-Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 6; Āpte 1911. Translation in Swami 1937a. ———. Pra na-Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 4; Āpte 1932. Translation in 1937b. ———. Bṛhad- raṇyaka-Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volumes 8-10; Āgaśe 1891. Translation in Swami 1950. ———. Brahma-Sūtra-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volumes 1-3; Śāstri 1890ś Śāstri and Śāstrācārya 1938. Translation in Swami 1965ś Thibaut 1890. 531 ———. Bhagavad-G t -Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volumes 11-12; Āgaśe 1934a. Translation in Swami 2012. ———. Muṇḍaka-Upaniṣad-Bh ṣya. Text in The Works of Sri Sankaracharya, 1910, volume 5; Āpte 1935. Translation in Swami 1937b. atapatha Br hmaṇaṬ Text in Weber 1855. Śabara Svāmin. M m ṁs -Sūtra-Bh ṣya. Text in Abhyankar and Joshi 1970. Translation in Jha 1933. Śālikanātha Miśra. Ṛju-Vimal on Prabhākara’s Bṛhaṭ . Text in Ramanatha Sastri and Subrahmanya Sastri 1934. ———. Prakaraṇa-Pañcik . Text in Sastri 1961. Partial translation in Pandurangi 2004. ———. V ky rtha-M tṛka (chapter 11 of Prakaraṇa-Pañcik ). Text in Sarma 1990. Translation in Sarma 1987. Śrīnivāsa. Ved nta-Kaustubha. Text in Brahmacārin 1904. Translation in Bose 2004. Sarvajñātman. Pañca-Prakriy . Text in Chintamani 1946. Translation in Kocmarek 1985. ———. Pram ṇa-Lakṣaṇa. Text in Easwaran Nampoothiry 1973. ———. Saṅkṣepa- r raka. Text in Vaidya 1918. Translation in Veezhinathan 1972. Sureśvara. Taittir ya-Upaniṣad-Bh ṣya-V rttika. Text in Āpte 1911. Translation in Balasubramanian 1984; van Boetzelaer 1971. ———. Naiṣkarmya-Siddhi. Text in Hiriyanna 1980. Translation in Balasubramanian 1988; Alston 1971. ———. Bṛhad- raṇyaka-Upaniṣad-Bh ṣya-V rttika. Text in Āpte 1892. ———. Sambandha-V rttikaṬ Text in Āpte 1892. Translation in Mahadevan 1968. Suvarṇa-Saptati on the S ṅkhya-K rik . Text in Aiyaswami Sastri 1944. Someśvara Bha a. Ny ya-Sudh on Kumārila’s Tantra-V rttika. Text in Gosvāmī 1984. 532 EDITIONS, TRANSLATIONS, AND SECONDARY LITERATURE Abhyankar, Kashinath Vasudev. 1986. A Dictionary of Sanskrit Grammar. Gaekwad’s Oriental Series 134. Baroda: Oriental Institute. Abhyankar, K.V., and G.A. Joshi, eds. 1970. M m ṁs -Dar anam of Jaimini, with the abara- Bh ṣya of abara, Prabh of r Vaidyan tha stri, Tantra-V rttika and Ṭupṭ k of Kum rila BhaṭṭaṬ In 7 Volumes. Ānandāśrama Sanskrit Series. PoonaŚ Ānandāśrama. Abhyankar, Vasudev Shastri, ed. 1914. ri-Bh ṣya by R m nuj ch ryaṬ In 2 Parts. Bombay Sanskrit and Prakrit Series 68, 72. Bombay: The Department of Public Instruction. Acharya, Diwakar. 2006. V caspatimi ra’s Tattvasam kṣ , The Earliest Commentary on Maṇḍanami ra’s BrahmasiddhiŚ Critically Edited with an Introduction and Critical Notes. Stuttgart: Franz Steiner Verlag. Adluri, Sucharita. 2015. Textual Authority in Classical Indian ThoughtŚ R m nuja and the Viṣṇu Pur ṇa. London and New York: Routledge. Āgaśe, Kāśīnātha Śāstri, ed. 1891. Bṛhad- raṇyaka-Upaniṣad-Bh ṣya of aṅkara, with the Ṭ k of nandagiri. Ānandāśrama Sanskrit Series 15. PoonaŚ Ānandāśrama. ———, ed. 1909. Kenopaniṣad-Pada- and V kya-Bh ṣya of aṅkara, with nandagiri’s Ṭ k Ṭ Ānandāśrama Sanskrit Series 6. PoonaŚ Ānandāśrama. ———, ed. 1934a. Bhagavad-G t -Bh ṣya of aṅkara, with the Ṭ k of nandagiri. Ānandāśrama Sanskrit Series 34. PoonaŚ Ānandāśrama. ———, ed. 1934b. Ch ndogya-Upaniṣad-Bh ṣya of aṅkara, with the Ṭ k of nandagiri. Ānandāśrama Sanskrit Series 14. PoonaŚ Ānandāśrama. Agrawal, Madan Mohan. 1992. Essence of VaiṣṇavismŚ Philosophy of Bhed bheda. Delhi: Ajanta Publications. Aiyaswami Sastri, N. 1944. Suvarnasaptati SastraṬ S ṅkhya K rik Saptati of vara Kṛṣṇa with a Commentary, Reconstructed into Sanskrit from the Chinese Translation of Param rtha. Sri Venkatesvara Oriental Series 7. Tirupati: Tirumalai-Tirupati Devasthanams Press. Aklujkar, Ashok. 1991. “Bhart -Hari’s Concept of the Veda.” In P ṇini and the Veda, edited by Madhav M. Deshpande, V:1–18. Panels of the VIIth World Sanskrit Conference. Leiden: E.J. Brill. ———. 2004. “Can the Grammarian’s Dharma Be a Dharma for All?” In Dharma: Studies in Its Semantic, Cultural and Religious History, edited by Patrick Olivelle, 265–310. Delhi: Motilal Banarsidass. 533 ———. 2009. “Veda Revelation According to Bhart -Hari.” In BhartṛhariŚ Language, Thought and Reality, edited by M Chaturvedi. Delhi: Motilal Banarsidass. ———. 2010. “Unity of the MīmāṁsāsŚ How Historiography Hides History.” In V caspativaibhavamŚ A Volume in Felicitation of Professor Vacaspati Upadhyaya, 821– 900. New Delhi: D.K.Printworld (P) Ltd. Alston, A.J, trans. 1971. The Realization of the AbsoluteṬ The “Naiṣkarmya Siddhi” of Sure vara. London: Shanti Sadan. ———, trans. 1990. The Thousand Teachings of aṁkara. London: Shanti Sadan. Andrijanić, Ivan. 2015. “Quotations and (Lost) Commentaries in Advaita VedāntaŚ Some Philological Notes on Bhart prapañca’s ‘Fragments.’” Journal of Indian Philosophy 43: 257–76. ———. 2016. “Bhartrprapañca and the Eight States of Brahman.” Revista Guillermo de Ockham 14 (1): 57–67. Āpte, Hari Nārāya a, ed. 1911. Taittir ya-Upaniṣad-Bh ṣya-V rtika of Sure vara, with nandagiri’s Ṭ k . Ānandāśrama Sanskrit Series 13. PoonaŚ Ānandāśrama. Āpte, Mahādeva Cima ājī, ed. 1891. aṅkara-Dig-Vijaya of Vidy raṇya. Ānandāśrama Sanskrit Series 22. PoonaŚ Ānandāśrama. ———, ed. 1892. Bṛhad raṇyakopaniṣadbh syav rtikam of Sure varaṬ With nandagiri’s straprak ik Ṭ In 3 Volumes. Ānandāśrama Sanskrit Series 16. PoonaŚ Ānandāśrama. Āpte, Vināyaka Ga eśa, ed. 1922. Ny ya-Sūtra of Gautama, with the Bh ṣya of V tsy yana and the Vṛtti of Viṣvan tha. Ānandāśrama Sanskrit Series 91. PoonaŚ Ānandāśrama. ———, ed. 1931. Aitareya-Upaniṣad-Bh ṣya of aṅkara, with nandagiri’s Ṭ k and Vidy raṇya’s D pik . Ānandāśrama Sanskrit Series 11. PoonaŚ Ānandāśrama. ———, ed. 1932. Pra na-Upaniṣad-Bh ṣya of aṅkara, with nandagiri’s Ṭ k . Ānandāśrama Sanskrit Series 8. PoonaŚ Ānandāśrama. ———, ed. 1935. Muṇḍaka-Upaniṣad-Bh ṣya of aṅkara, with nandagiri’s Ṭ k . Ānandāśrama Sanskrit Series 9. PoonaŚ Ānandāśrama. Arnold, Dan. 2005. Buddhists, Brahmins, and Belief: Epistemology in South Asian Philosophy of Religion. New York: Columbia University Press. Bader, Jonathan. 1990. Meditation in aṅkara’s Ved nta. New Delhi: Aditya Prakashan. 534 ———. 2000. Conquest of the Four QuartersŚ Traditional Accounts of the Life of aṅkara. New Delhi: Aditya Prakashan. Balasubramanian, R. 1976. Advaita Ved nta. Madras: Centre for Advanced Study in Philosophy, University of Madras. ———. 1983. A Study of the Brahmasiddhi of Maṇḍana Mi ra. Varanasi: Chaukhamba Amarabharati Prakashan. ———, trans. 1984. The Taittir yopaniṣad-Bh ṣya-V rtika of Sure vara. Madras University Philosophical Series 20. Madras: Radhakrishnan Institute for Advanced Study in Philosophy, University of Madras. ———, trans. 1988. The Naiṣkarmyasiddhi of Sure vara. Madras: Radhakrishnan Institute for Advanced Study in Philosophy, University of Madras. Bartley, Chris. 1986. “Interpreting ‘Satyaṁ Jñānam Anantaṁ Brahma.’” Oxford University Papers on India 1 (1): 103–15. Belvalkar, Shripad Krishna. 1918. “The Multiple Authorships of the Vedānta Sūtras.” Indian Philosophical Review 2: 141–54. ———. 1927. “Jaimini’s Śārīraka-Sūtra.” In Aus Indiens Kultur Ś Festgabe Richard von Garbe Dem Forscher Und Lehrer Zu Seinem 70. Geburtstag Dargebracht von Seinen Freunden, Verehrern Und Schülern. Veröffentlichungen Des Indogermanischen Seminars Der Universität Erlangen 3. Erlangen: Palm & Enke. ———. 1929. Shree Gopal Basu Mallik Lectures on Ved nta Philosophy. Poona: Bilvakuñja Publishing House. Bergson, Henri. 1913. An Introduction to Metaphysics. Translated by T.E Hulme. London: Macmillan and Co. Bhāgavatāchārya, Rāmaśāstrī, ed. 1891. The Pañcap dik of Padmap da. Benares: E.J. Lazarus & Co. Bhatt, Govardhan P. 1962. Epistemology of the Bh ṭṭa School of Pūrva M m ṁs . The Chowkhamba Sanskrit Studies 17. Varanasi: The Chowkhamba Sanskrit Series Office. ———. 1994. “Mīmāṁsā as a Philosophical SystemŚ A Survey.” In Studies in M m ṁs Ś DrṬ Mandan Mishra Felicitation Volume, edited by R.C Dwivedi, 3–25. Delhi: Motilal Banarsidass. Bhattacharya, Vidhushekhara. 1989. The gama stra of Gauḍap da. Delhi: Motilal Banarsidass. 535 Bilimoria, Purushottama. 1988. abdapram ṇaŚ Word and Knowledge. Studies in Classical India 10. Dordrecht, Boston, London: Kluwer Academic Publishers. Bodewitz, Henk. 2002. Kauṣ taki UpaniṣadŚ Translation and Commentary. Groningen Oriental Studies. Groningen: Forsten. Boetzelaer, J.M van, trans. 1971. Sure vara’s Taittir yopaniṣad-Bh ṣyav rtikam. Orientalia Rheno-Traiectina 12. Leiden: E.J. Brill. Bose, Roma. 2004. Ved nta-P rij ta-Saurabha of Nimb rka and Ved nta-Kaustubha of r niv saŚ English Translation. Bibliotheca Indica 259. Delhi: Munshiram Manoharlal. Brahmacārin, Nityasvarūpa, ed. 1904. Brahma-Sūtra with Ved nta-P rij ta-Saurabha of Nimb rka, Ved nta-Kaustubha of r niv sa, and Ved nta-Kaustubha-Prabh of Ke ava K m r . Vrindavana. Brereton, Joel. 1986. “‘Tat Tvam Asi’ in Context.” Zeitschrift Der Deutschen Morgenländischen Gesellschaft 136 (1): 98–109. Brereton, Joel, and Stephanie Jamison. 2014. The Rigveda: The Earliest Religious Poetry of India. Oxford and New York: Oxford University Press. Brick, D. 2006. “Transforming Tradition into TextsŚ The Early Development of Sm ti.” Journal of Indian Philosophy 34: 287–302. Bronkhorst, Johannes. 2007a. Greater Magadha: Studies in the Culture of Early India. Leiden and Boston: Brill. ———, ed. 2007b. M m ṁs and Ved ntaŚ Interaction and Continuity. Papers of the 12th World Sanskrit Conference, 10.3. Delhi: Motilal Banarsidass. Bryant, Edwin. 2003. KrishnaŚ The Beautiful Legend of God ( r mad Bh gavata Pur ṇa Book X). London: Penguin Books. ———. 2009. The Yoga Sūtras of PatañjaliŚ A New Edition, Translation, and Commentary, with Insights from the Traditional Commentaries. New York: North Point Press. Buitenen, J.A.B van. 1956. Ved rthasaṁgraha. Deccan College Monograph Series 16. Poona: Deccan College. ———. 1981. The Bhagavad-G t in the Mah bh rata. Chicago and London: University of Chicago Press. Cardona, George. 1967. “Anvaya and Vyatireka in Indian Grammar.” The Adyar Library Bulletin 31–32 (August): 313–52. 536 ———. 1981. “On Reasoning from Anvaya and Vyatireka in Early Advaita.” In A Memorial Volume in Honour of Pandit Sukhlalji Sanghvi, 79–104. Ahmedabad: L.D. Institute of Indology. Carman, John Braisted. 1974. The Theology of R m nujaŚ An Essay in Interreligious Understanding. Yale Publications in Religion 18. New Haven and London: Yale University Press. Carpenter, David. 1992. “Bhrt hari and the Veda.” In Texts in Contexts: Traditional Hermeneutics in South Asia, edited by Jeffrey R. Timm, 17–32. Albany: State University of New York Press. Cassirer, Ernst. 1956. An Essay of Man: An Introduction to a Philosophy of Human Culture. New York: Doubleday Anchor Books. Chatterjee, K.N. 1992. “Deities in Mīmāṁsā.” In Corpus of Indological Studies. Prof. Ramaranjan Mukherji Felicitation Volume, edited by Anantlal Thakur, 171–73. Delhi: Sharada Publishing House. Chatterjee, S.C. 2008. Ny ya Theory of Knowledge. Delhi: Bharatiya Kala Prakashan. Chemparathy, George. 1987. “Nature and Role of Apta in the Nyaya-Vaisesika Thought.” In Kusum ñjaliŚ New Interpretation of Indian Art and CultureṬ ShṬ C. Sivaramamurti Commemoration Volume, edited by M.S Nagaraja Rao, 1:39–47. Delhi: Agam Kala Prakashan. Chintamani, T.R, ed. 1946. Pañcaprakriy of Sarvajñ tman, with the Commentaries of nandajñ na and Pūrṇavidyamuni. Bulletins of the Sanskrit Department 4. Madras: University of Madras. Clark, Matthew. 2006. The Da an m -Saṁny sisŚ The Integration of Ascetic Lineages into an Order. Brill’s Indological Library 25. Leiden and BostonŚ Brill. Clooney, Francis X., S.J. 1987. “Why the Veda Has No AuthorŚ Language as Ritual in Early Mīmāṃsā and Post-Modern Theology.” Journal of the American Academy of Religion 55 (4): 659–84. ———. 1988. “Devatādhikara aŚ A Theological Debate in the Mīmāṁsā-Vedānta Tradition.” Journal of Indian Philosophy 16 (3): 277–98. ———. 1990. Thinking RituallyŚ Rediscovering the Pūrva M m ṁs of Jaimini. Publications of the De Nobili Research Library 17. Vienna: Institute for Indology, University of Vienna & Commission Agents, E. J. Brill. ———. 1993. Theology After Ved ntaŚ An Experiment in Comparative Theology. Albany: State University of New York Press. 537 ———. 2003. “Restoring ‘Hindu Theology’ as a Category in Indian Intellectual Discourse.” In The Blackwell Companion to Hinduism, edited by Gavin Flood, 447–77. Oxford: Blackwell Publishing. Cohen, Signe. 2008. Text and Authority in the Older Upaniṣads. Brill’s Indological Library 30. Leiden and Boston: Brill. Collins, Steven. 1982. Selfless PersonsŚ Imagery and Thought in Therav da Buddhism. Cambridge and New York: Cambridge University Press. Comans, Michael. 2000. The Method of Early Advaita Ved ntaŚ A Study of Gauḍap da, aṅkara, Sure vara and Padmap da. Delhi: Motilal Banarsidass. Copleston, Frederick S.J. 1993a. A History of Philosophy. Volume I: Greece and Rome. New York: Image Books, Doubleday. ———. 1993b. A History of Philosophy. Volume II: Medieval Philosophy. New York: Image Books, Doubleday. Coward, Harold, and Krishna Sivaraman, eds. 1997. Revelation in Indian Thought: A Festschrift in Honour of Professor T. R. V. Murti. Emeryville, California: Dharma Publishing. Dasgupta, S.N. 1930. Yoga Philosophy in Relation to Other Systems of Indian Though. Delhi: Motilal Banarsidass. ———. 1932. History of Indian Philosophy. Vol. 2. London: Cambridge University Press. David, Hugo. 2013. “Action Theory and Scriptural Exegesis in Early Advaita-Vedānta (1)Ś Ma ḍana Miśra on Upadeśa and Iṣ asādhanatā.” In Scriptural Authority, Reason and Action. Proceedings of a Panel at the 14th World Sanskrit Conference, Kyoto. Sept. 1-5, 2009, edited by Vincent Eltschinger and Helmut Krasser, 271–318. Wien: Verlag der Österreichischen Akademie der Wiessenschaften. Davis, Richard H. 2015. The Bhagavad-Gita: A Biography. Lives of Great Religious Books. Princeton and Oxford: Princeton University Press. De, Sushil Kumar. 1961. Early History of the Vaiṣṇava Faith and Movement in BengalŚ From Sanskrit and Bengali Sources. Calcutta: Firma K.L. Mukhopadhyay. Deshpande, Madhav M. 2007. The Meaning of Nouns: Semantic Theory in Classical and Medieval India. New Delhi: D.K.Printworld (P) Ltd. Deussen, Paul. 1912. The System of the Vedanta. Chicago: The Open Court Publishing Company. 538 ———. 1997. Sixty Upanishads of the Veda. Translated by V.M Bedekar and Palsule. Delhi: Motilal Banarsidass. Deutsch, Eliot. 1969. Advaita Ved ntaŚ A Philosophical Reconstruction. Honolulu: East-West Center Press. Devasthali, G.V. 1959. M m ṁs Ś The V kya- stra of Ancient India. BombayŚ Booksellers’ Publishing Company. Dharmadhikari, T.N. 2006. “Kinds of Vedic Sacrifices.” In Pūrvam m ṁs from an Interdisciplinary Point of View, edited by K.T. Pandurangi, 329–36. History of Science, Philosophy and Culture in Indian Civilization, Volume II, Part 6. Delhi: Centre for Studies in Civilization. Dimmitt, Cornelia, and J.A.B van Buitenen. 1978. Classical Hindu Mythology: A Reader in the Sanskrit Pur ṇas. Philadelphia: Temple University Press. Doniger O’Flaherty, Wendy, ed. 1980a. Karma and Rebirth in Classical Indian Traditions. Berkeley, Los Angeles and London: University of California Press. ———. 1980b. “Karma and Rebirth in the Vedas and the Purā as.” In Karma and Rebirth in Classical Indian Traditions, edited by Wendy Doniger O’Flaherty, 3–37. Berkeley, Los Angeles and London: University of California Press. ———. 1981. Rig Veda. London: Penguin Books. Doniger, Wendy. 2005. The Woman Who Pretended to Be Who She Was: Myths of Self- Imitation. Oxford and New York: Oxford University Press. ———. 2014. On Hinduism. Oxford and New York: Oxford University Press. Dubois, Joёl André-Michel. 2013. The Hidden Lives of BrahmanŚ aṅkara’s Ved nta through His Up niṣad [Sic] Commentaries, in Light of Contemporary Practice. Albany: State University of New York Press. Easwaran Nampoothiry, E., ed. 1973. Pram ṇalakṣaṇam of Sarvajñ tmamuni. Kerala University Sanskrit Department Publication 8. Trivandrum: Department of Sanskrit, University of Kerala. Eco, Umberto. 2002. Baudolino. Orlando: Harcourt, Inc. Edelman, Jonathan. 2013. “Hindu Theology as Churning the Latent.” Journal of the American Academy of Religion 81 (2): 427–66. Edgerton, Franklin. 1924. “The Meaning of Sā khya and Yoga.” The American Journal of Philology 45 (1): 1–46. 539 ———. 1928. “Some Linguistic Notes on the Mīmā sā System.” Language 4 (3): 171–77. ———. 1929. The M m ṅs Ny ya Prak a or padev Ś A Treatise on the M m ṅs System by padeva. New Haven: Yale University Press. Endo, Ko. 2000. “Prasaṁkhyāna in the Yogabhāṣya.” In The Way to Liberation: Indological Studies in Japan, edited by Sengaku Mayeda, I:75–89. Japanese Studies on South Asia 3. Delhi: Manohar. Faddegon, Barend. 1923. “The Chapter (III, 3), the Method of Exegesis, in the Vedānta-Sūtra.” Acta Orientalia (Netherlands) 1: 105–13. Fitzgerald, James L. 2012. “A Prescription for Yoga and Power in the Mahābhārata.” In Yoga in Practice, edited by David Gordon White. Princeton and Oxford: Princeton University Press. Flood, Gavin. 1996. An Introduction to Hinduism. Cambridge: Cambridge University Press. Fort, Andrew O. 1998. J vanmukti in Transformation: Embodied Liberation in Advaita and Neo- Vedanta. Albany: State University of New York Press. Freschi, Elisa. 2012. Duty, Language and Exegesis in Pr bh kara M m ṁs . Leiden and Boston: Brill. Ganeri, Jonardon. 2001. Philosophy in Classical India: The Proper Work of Reason. London and New York: Routledge. Ghate, V.S. 1981. The Ved nta. Poona: Bhandarkar Oriental Research Institute. Gonda, Jan. 1963. The Vision of the Vedic Poets. Disputationes Rheno-Trajectinae 8. The Hague: Mouton. ———. 1966. Loka: World and Heaven in the Veda. Amsterdam: N. V. Noord-Hollandsche Uitgevers Maatschappij. Gosvāmī, Mahāprabhulāla, ed. 1984. The M m ṁs Dar ana of Maharṣi JaiminiṬ With barabh ṣya of abaramuni, Tantrav rtika by Kum rila-Bhaṭṭa with Its Commentary Ny yasudh of Some vara-Bhaṭṭa, Bh ṣyavivaraṇa of Govind mṛtamuni and Bh vaprak ik by DrṬ Mah prabhul la Gosv m Ṭ In 4 Volumes. Pracyabharati Series 16, 22, 23, 32. Varanasi: Tara Printing Works. Goswami, Tamal Krishna. 2012. A Living Theology of Krishna Bhakti: Essential Teachings of A.C. Bhaktivedanta Swami Prabhupada. New York: Oxford University Press. Grimes, John. 2004. The Vivekacūd maṇi of ankar c rya Bhagavatp daŚ An Introduction and Translation. Burlington, VT: Ashgate. 540 Hacker, Paul. 1995. Philology and Confrontation: Paul Hacker on Traditional and Modern Ved nta. Edited by Wilhelm Halbfass. Albany: State University of New York Press. Hadot, Pierre. 2002. What Is Ancient Philosophy? Translated by Michael Chase. Cambridge, Mass.: The Belknap Press of Harvard University Press. Halbfass, Wilhelm. 1980. “Karma, Apūrva and ‘Natural Causes’Ś Observations on the Growth and Limits of the Theory of Saṁsāra.” In Karma and Rebirth in Classical Indian Traditions, edited by Wendy Doniger O’Flaherty, 268–301. Berkeley, Los Angeles and London: University of California Press. ———. 1991. Tradition and Reflection: Explorations in Indian Thought. Albany: State University of New York Press. Hariharānanda Āra ya, Swami. 2000. Yoga Philosophy of Patañjali, with Bh svat . Calcutta: University of Calcutta. Harikai, Kunio. 1994. “On the Three-Fold Classification of the Arthavāda.” In Studies in M m ṁs Ś DrṬ Mandan Mishra Felicitation Volume, edited by R.C Dwivedi, 299–311. Delhi: Motilal Banarsidass. Harimoto, Kengo. 2006. “The Date of Śa karaŚ Between the Cāḷukyas and the Rāṣ rakū as.” Journal of Indological Studies 18: 85–111. Hettche, Matt. 2016. “Christian Wolff.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Winter 2016. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/archives/win2016/entries/wolff-christian/. Hiriyanna, M. 1928. “BrahmadattaŚ An Old Vedāntin.” The Journal of Oriental Research 2: 1–9. ———. 1957. Indian Philosophical Studies 1. Mysore: Kalyalaya Publishers. ———. 1972. Indian Philosophical Studies 2. Mysore: Kalyalaya Publishers. ———, ed. 1980. The Naiṣkarmya-Siddhi of Sure var c rya, with the Candrik of Jñ nottamaṬ Edited with Notes and Index by the Late Colonel G.A. Jacob. Revised Edition with Introduction and Explanatory Notes. Bombay Sanskrit and Prakrit Series 36. Bombay: Bhandarkar Oriental Research Institute. ———. 1995. The Essentials of Indian Philosophy. Delhi: Motilal Banarsidass. Holdrege, Barbara A. 1994. “Veda in the Brāhma asŚ Cosmogonic Paradigms and the Delimitation of Canon.” In Authority, Anxiety, and Canon: Essays in Vedic Interpretation, edited by Laurie L Patton, 35–66. Albany: State University of New York Press. 541 ———. 1996. Veda and Torah: Transcending the Textuality of Scripture. Albany: State University of New York Press. Hopkins, E. Washburn. 1915. Epic Mythology. Strassburg: Verlag von Karl J. Trübner. Ingalls, Daniel H.H. 1952. “The Study of Śaṁkarācārya.” Annals of the Bhandarkar Oriental Research Institute 33 (1/4): 1–14. ———. 1953. “Śaṁkara on the QuestionŚ Whose Is Avidyā?” Philosophy East and West 3 (1): 69–72. ———. 1954. “Śaṁkara’s Arguments Against the Buddhists.” Philosophy East and West 3 (4): 291–306. Isayeva, Natalia. 1993. Shankara and Indian Philosophy. Albany: State University of New York Press. Jakubczak, Marzenna. 2004. “Living Liberation (Jīvan-Mukti) in Sāṁkhya and Yoga.” In Essays in Indian Philosophy, Religion and Literature, edited by Piotr Balcerowicz and Marek Mejor, 363–71. Delhi: Motilal Banarsidass. Jha, Ganganatha, ed. 1896. Tattva-Kaumud (S ṅkhya) of V caspati Mi raŚ An English Translation with the Sanskrit Text. Bombay: Tookaram Tatya, F.T.S., for the Bombay Theosophical Publication Fund. ———, trans. 1907. lokav rttikaṬ Translated from the Original Sanskrit with Extracts from the Commentaries of Sucarita Mi ra (the K ik ) and P rtha S rath Mi ra. Bibliotheca Indica. Calcutta: Asiatic Society of Bengal. ———, trans. 1933. Shabara-Bh shyaṬ In 3 Volumes. Geakwad’s Oriental Series 66, 70, 73. Baroda: Oriental Institute. ———. 1964. Pūrva-M m ṁs in Its Sources. Varanasi: The Banaras Hindu University. ———. 1978. The Pr bh kara School of Pūrva M m ṁs . Delhi: Motilal Banarsidass. ———, trans. 1983. Tantrav rttika of Kum rila BhaṭṭaŚ A Commentary on abara’s Bh ṣya on the Pūrvam m ṁs Sūtras of JaiminiṬ In 2 Volumes. Sri Garib Das Oriental Series 9–10. Delhi: Sri Satguru Publications. Jhalakikar, Bhimacarya, and Vasudev Shastri Abhyankar. 1928. Ny yako a, or Dictionary of Technical Terms of Indian Philosophy. Poona: Bhandarkar Oriental Research Institute. Joshi, S.D. 1968. Patañjali’s Vy karaṇa-Mah bh ṣyaŚ Samarth hnika (PṬ 2Ṭ1Ṭ1), Edited with Translation and Explanatory Notes. Poona: University of Poona. 542 Kapstein, Matthew. 2015. “Interpreting Indian PhilosophyŚ Three Parables.” In The Oxford Handbook of Indian Philosophy, edited by Jonardon Ganeri. Oxford Handbooks Online. Oxford: Oxford University Press. Karmarkar, R.D. 1925. “The Devayāna and the Pit yā a.” Proceedings and Transactions of the Third Oriental Conference, 451–64. ———, ed. 1959. r bh ṣya of R m nujaṬ In 3 Volumes. University of Poona Sanskrit and Prakrit Series 1. Poona: Unversity of Poona. Kataoka, Kei. 1999. “The Mīmāṁsā Concept of Saṁskāra and the Saṁskāra in the Process of Cognizing a Word-MeaningŚ Pūrva-Var a-Janita-Saṁskāra.” Journal of the Japanese Association for South Asian Studies 11: 1–24. ———. 2000. “Reconstructing the Dharma-Abhivyakti-Vāda in the Mīmāṁsā Tradition.” In The Way to Liberation: Indological Studies in Japan, edited by Sengaku Mayeda, I:167–81. Japanese Studies on South Asia 3. Delhi: Manohar. ———. 2001. “Scripture, Men, and HeavenŚ Causal Structure in Kumārila’s Action-Theory of Bhāvanā.” Journal of Indian and Buddhist Studies 49 (2): 1031–1028. ———. 2011a. Kum rila on Truth, Omniscience, and KillingṬ Part 1Ś A Critical Edition of M m ṁs - lokav rttika Ad 1Ṭ1Ṭ2 (Codan sūtra). Wien: Verlag der Österreichischen Akademie der Wiessenschaften. ———. 2011b. Kum rila on Truth, Omniscience, and Killing. Part 2: An Annotated Translation of M m ṁs - lokav rttika Ad 1Ṭ1Ṭ2 (Codan sūtra). Wien: Verlag der Österreichischen Akademie der Wiessenschaften. Kato, Takahiro. 2011. “The First Two Chapters of Bhāskara’s ŚārīrakamīmāṁsābhāṣyaŚ Critically Edited with an Introduction, Notes and an Appendix.” Halle-Wittenberg: Philosophischen Fakultät der Martin-Luther-Universität Halle-Wittenberg. Katre, Sumitra. 1989. Aṣṭ dhyay of P ṇiniŚ Roman Transliteration and English Translation. Delhi: Motilal Banarsidass. Kevalānanda Sarasvatī. 1993. M m ṁs -Ko aṬ In 7 VolumesṬ Delhi: Sri Satguru Publications. Kocmarek, Ivan. 1985. Language and ReleaseŚ Sarvajñ tman’s Pañcaprakriy . Delhi: Motilal Banarsidass. Krasser, Helmut. 2012. “Bhāviveka, Dharmakīrti and Kumārila.” In Devadattiyam: Johannes Bronkhorst Felicitation Volume, edited by Francois Voegeli, Vincent Eltschinger, Feller Danielle, Maria Piera Candotti, Bogdan Diaconescu, and Malhar Kulkarni, 535–94. Bern: Peter Lang. 543 Krishna Warrier, A.G. 1961. The Concept of Mukti in Advaita Ved nta. Madras: University of Madras. Krishnan, Y. 1994. “Nitya and Naimittika Karma in the Pūrva Mīmāṁsā.” Annals of the Bhandarkar Oriental Research Institute 75 (1): 177–84. Kunjunni Raja, K. 1977. Indian Theories of Meaning. Madras: The Adyar Library and Research Centre. Larson, Gerald J., and Ram Shankar Bhattacarya. 1987. S ṁkhyaŚ A Dualist Tradition in Indian Philosophy. Encyclopedia of Indian Philosophies, Volume 4. Delhi: Motilal Banarsidass. Limaye, V.P., and R.D. Vadekar, eds. 1958. Eighteen Principal Upaniṣads. Poona: Vaidika Saṁśodhana Ma ḍala. Lindtner, Christian. 2001a. Bhavya on M m ṁs Ś M m ṁs tattvanirnay vat raḥ. Chennai: The Adyar Library and Research Centre. ———, ed. 2001b. Madhyamakahṛdayam of Bhavya. The Adyar Library Series 123. Chennai: The Adyar Library and Research Centre. Lipner, Julius. 1986. The Face of Truth. Albany: State University of New York Press. ———. 1997. “Śa kara on Satyaṁ Jñānam Anantaṁ Brahma.” In Relativism, Suffering and Beyond: Essays in Memory of Bimal K. Matilal, edited by P. Bilimoria and J.N Mohanty, 301–18. Delhi: Oxford University Press. ———. 2000. “The Self of Being and the Being of SelfŚ Śaṁkara on ‘That You Are’ (Tat Tvam Asi).” In New Perspectives on Advaita Ved ntaŚ Essays in Commemoration of Professor Richard De Smet, S.J., edited by Bradley J. Malkovsky, 51–69. Leiden and Boston: Brill. Lorenzen, David N. 2006. Who Invented Hinduism: Essays on Religion in History. New Delhi: Yoda Press. Lott, Eric J. 1976. God and the Universe in the Ved ntic Theology of R m nujaŚ A Study of His Use of the Self-Body Analogy. Madras: Ramanuja Research Society. Maas, Phillip. 2010a. “On the Written Transmission of the Pātañjalayogaśāstra.” In From Vasubandhu to Caitanya. Studies in Indian Philosophy and Its Textual History, edited by Johannes Bronkhorst and Karin Preisendanz, 157–72. Papers of the 12th World Sanskrit Conference, 10.1. Delhi: Motilal Banarsidass. ———. 2010b. “Valid Knowledge and Belief in Classical Sāṃkhya Yoga.” In Logic and Belief in Indian Philosophy, edited by Piotr Balcerowicz, 383–92. Warsaw Indological Studies 3. Delhi: Motilal Banarsidass. 544 ———. 2013. “A Consise Historiography of Classical Yoga Philosophy.” In Periodization and Historiography of Indian Philosophy, edited by Eli Franco, 53–90. Publications of the De Nobili Research Library 37. Vienna: Institut für Südasien-, Tibet- und Buddhismuskunde der Universität Wien. Mah bh rata, with the Commentary of N lakaṇṭhaṬ In 7 Volumes. 1929. Poona: Shankar Narhar Joshi. Mahadevan, T.M.P. 1968. Sankaracharya. New Delhi: National Book Trust, India. ———, trans. 1972. The Sambandha-V rtika of Sure var c rya. Madras University Philosophical Series 6. Madras: University of Madras. ———. 1985. Superimposition in Advaita Ved nta. New Delhi, Bangalore, and Jalandhar: Sterling Publishers Private Ltd. Mainkar, T.G. 1972. S ṁkhya-K rik of varakṛṣṇa, with the Commentary of Gauḍap daŚ Translated into English with Notes. Poona: Oriental Book Agency. Malkovsky, Bradley J. 2000a. “IntroductionŚ The Life and Work of Richard V. De Smet, S.J.” In New Perspectives on Advaita Ved ntaŚ Essays in Commemoration of Professor Richard De Smet, S.J., edited by Bradley J. Malkovsky, 1–17. Leiden and Boston: Brill. ———. 2000b. “Śaṃkara on Divine Grace.” In New Perspectives on Advaita Ved ntaŚ Essays in Commemoration of Professor Richard De Smet, S.J., edited by Bradley J. Malkovsky, 70– 83. Leiden and Boston: Brill. ———. 2001. The Role of Divine Grace in the Soteriology of aṁkar c rya. Numen Book Series: Studies in the History of Religions, XCI. Leiden: Brill. Marcaurelle, Roger. 2000. Freedom Through Inner RealizationŚ aṅkara’s Philosophy in a New Light. Albany: State University of New York Press. Matilal, Bimal Krishna. 1990. The Word and the WorldŚ India’s Contribution to the Study of Language. Delhi and New York: Oxford University Press. Mayeda, Sengaku. 1965a. “The Authenticity of the Bhagavadgītābhāṣya Ascribed to Śa kara.” Wiener Zeitschrift Für Die Kunde Süd- Und Ostasiens Und Archiv Für Indische Philosophie 9: 155–97. ———. 1965b. “The Authenticity of the Upadeśasāhasrī Ascribed to Śa kara.” Journal of the American Oriental Society 85 (2): 178–96. ———. 1967. “On Śa kara’s Authorship of the Kenopaniṣadbhāṣya.” Indo-Iranian Journal 10 (1): 33–55. 545 ———. 1998. “Reviewed WorkŚ Meditation in Śa kara’s Vedānta by Jonathan Bader.” Indo- Iranian Journal 41 (1): 64–69. ———. 2006a. aṅkara’s Upade as hasr Ś VṬ 1Ṭ Introduction, Text & Indices. Delhi: Motilal Banarsidass. ———. 2006b. aṅkara’s Upade as hasr Ś VṬ 2Ṭ Introduction and English Translation. Delhi: Motilal Banarsidass. McCrea, Lawrence. 2000a. “The Hierarchical Organization of Language in Mīmāṁsā Interprative Theory.” Journal of Indian Philosophy 28: 429–59. ———. 2000b. “Yoshimizu Kiyotaka, Der `Organismus’ Des Urheberlosen VedaŚ Eine Studie Der Niyoga-Lehre Prabhākaras Mit Ausgewählten Übersetzungen Der Brhatī.” Indo-Iranian Journal 43 (2): 177. ———. 2008. The Teleology of Poetics in Medieval Kashmir. Harvard Oriental Series 71. Cambridge, Mass.: Harvard University Press. ———. 2010. “Hindu Jurisprudence and Scriptural Hermeneutics.” In Hinduism and Law: An Introduction, edited by Timothy Lubin, Donald R. Davis, and Jayanth K. Krishnan. Cambridge and New York: Cambridge University Press. ———. 2012. “Rationalisation and Sincerity in Mīmāṁsā Hermeneutics.” In World View and Theory in Indian Philosophy, edited by Piotr Balcerowicz, 5:119–35. Warsaw Indological Studies. Delhi: Manohar. ———. 2013. “The Transformations of Mīmāṁsā in the Larger Context of Indian Philosophical Discourse.” In Periodization and Historiography of Indian Philosophy, edited by Eli Franco, 127–43. Wien: Institut für Südasien-, Tibet- und Buddhismuskunde der Universität Wien. Mesquita, Roque. 1994. “Die Idee Der Erlösung Bei Kumārilabha a.” Wiener Zeitschrift Für Die Kunde Südasiens 38: 451–84. Minkowski, Christopher. 2011. “Advaita Vedānta in Early Modern History.” In Religious Cultures in Early Modern India: New Perspectives, edited by Rosalind O’Hanlon and David Washbrook, 73–99. New Delhi: Routledge. Mishra, Parameshwar Nath. 2001. Maṭh mn ya Setu or Mah nu sanam of r mad Jagadguru di aṅkar c rya Along With ard Bh ṣyam of Parameshwar Nath Mishra. Dwarka: Shri Sadanand Brahmachari (Shankaracharya Memorial Trust). Murty, K.S. 1959. Revelation and Reason in Advaita Ved nta. Waltair and New York: Andhra University and Columbia University Press. 546 Nakamura, Hajime. 1983. A History of Early Ved nta Philosophy. Vol. 1. Delhi: Motilal Banarsidass. ———. 1994. “Problem of Categorical Imperative in the Philosophy of Prabhākara SchoolŚ A Brief Note.” In Studies in M m ṁs Ś DrṬ Mandan Mishra Felicitation Volume, edited by R.C Dwivedi, 169–83. Delhi: Motilal Banarsidass. ———. 2004. A History of Early Ved nta Philosophy. Vol. 2. Delhi: Motilal Banarsidass. Natarajan, Kanchana. 1995. The Vidhi-Viveka of Maṇḍana Mi raŚ Understanding Vedic Injunctions. Delhi: Sri Satguru Publications. Nelson, Lance E. 1996. “Living Liberation in Śa kara and Classical Advaita: Sharing the Holy Waiting in God.” In Living Liberation in Hindu Thought, edited by Andrew O. Fort and Patricia V. Mumme, 17–62. Albany: State University of New York Press. Nooten, Barend A. van, and Gary B. Holland, eds. 1994. Rig Veda: A Metrically Restored Text with an Introduction and Notes. Harvard Oriental Series 50. Cambridge, Mass.: Harvard University Press. Okita, Kiyokazu. 2014. “Hindu Theology and the Question of Qualification.” International Journal of Hindu Studies 18 (2): 153–79. Olivelle, Patrick. 1986. Renunciation in Hinduism: A Medieval Debate. Volume One: The Debate and the Advaita Argument. Publications of the De Nobili Research Library, XIII. Leiden, Vienna, and Delhi: E.J. Brill, Gerold & Co., Motilal Banarsidass. ———. 1993. The rama SystemŚ The History and Hermeneutics of a Religious Institution. Oxford and New York: Oxford University Press. ———. 1998. The Early Upaniṣads. Oxford and New York: Oxford University Press. ———. 1999. DharmasūtrasŚ The Law Codes of Ancient India. Oxford and New York: Oxford University Press. ———. 2000. DharmasūtrasŚ The Law Codes of pastamba, Gautama, Baudh yana, and VasiṣṭhaṬ Annotated Text and Translation. Delhi: Motilal Banarsidass. ———. 2005. Manu’s Code of LawŚ A Critical Edition and Translation of the M nava- Dharma stra. Oxford and New York: Oxford University Press. Ollett, Andrew. 2013. “What Is Bhāvanā?” Journal of Indian Philosophy 41: 221–62. Otto, Rudolph. 1970. Mysticism East and West. New York: The Macmillan Company. Pande, Govind Chandra. 1994. Life and Thought of aṅkar c rya. Delhi: Motilal Banarsidass. 547 Pandey, Sangam Lal. 1983. Pre- aṁkara Advaita Philosophy. Allahabad Philosophical Series 2. Allahabad: Darshan Peeth. Pandit, Durgāprasāda Dviveda. 1936. S hitya-Darpaṇa by Vi van tha Kavir ja, With the Commentary of R macharaṇa-Tarkav g a Bhatt ch rya. Bombay: Nirnaya Sagar Press. Pandit, Vindhyesvarī Prasāda Dvivedin, ed. 1915. Brahmasūtra with A Commentary by Bh skar ch rya. Chowkhamba Sanskrit Series 70, 185, 209. Benares: Chowkhamba Sanskrit Book Depot. Pa ḍita, Vāmana Śāstri Islāmpurkar. 1889. Taittir ya-Upaniṣad-Bh ṣya of aṅkara, with the Ṭ k of nandagiri. Ānandāśrama Sanskrit Series 12. PoonaŚ Ānandāśrama. Pandurangi, K.T. 2004. Prakaraṇapañcik of likan tha with an Exposition in English. New Delhi: Indian Council of Philosophical Research. ———. 2006. “Exposition of Dharma as the Central Theme of Pūrvamīmāṁsā.” In Pūrvam m ṁs from an Interdisciplinary Point of View, edited by K.T. Pandurangi, 175– 233. History of Science, Philosophy and Culture in Indian Civilization, Volume II, Part 6. Delhi: Centre for Studies in Civilization. Parpola, Asko. 1981. “On the Formation of Mīmāṁsā and the Problems Concerning JaiminiŚ With Particular Reference to the Teacher Quotations and the Vedic Schools.” Wiener Zeitschrift Für Die Kunde Süd- Und Ostasiens Und Archiv Für Indische Philosophie 25: 145–77. ———. 1994. “On the Formation of Mīmāṁsā and the Problems Concerning JaiminiŚ With Particular Reference to the Teacher Quotations and the Vedic Schools (Part II).” Wiener Zeitschrift Für Die Kunde Süd- Und Ostasiens Und Archiv Für Indische Philosophie 38: 293–308. Patton, Laurie L. 2005. Bringing the Gods to Mind: Mantra and Ritual in Early Indian Sacrifice. Berkeley, Los Angeles and London: University of California Press. Pollock, Sheldon. 1997. “The ‘Revelation of Tradition’Ś Śruti, Sm ti, and the Sanskrit Discourse of Power.” In Lex et Litterae: Studies in Honour of Professor Oscar Botto, edited by Siegfried Lienhard and Irma Piovano, 395–417. TorinoŚ Edicioni Dell’Orso. Potter, Karl. 1965. Presuppositions of India’s Philosophies. New Delhi: Prentice-Hall of India (Private) LTD. ———, ed. 1977. Indian Metaphysics and Epistemology: The Tradition of Ny ya-Vai eṣika up to Gaṅge a. Encyclopedia of Indian Philosophies 2. Princeton, N.J.: Princeton University Press. 548 ———, ed. 1981. Advaita Ved nta up to aṃkara and His Pupils. Encyclopedia of Indian Philosophies 3. Princeton, N.J.: Princeton University Press. ———, ed. 2006. Advaita Ved nta from 800-1200 AD. Encyclopedia of Indian Philosophies 11. Delhi: Motilal Banarsidass. ———, ed. 2013. Bhed bheda and Dvait dvaita Systems. Encyclopedia of Indian Philosophies 15. Delhi: Motilal Banarsidass. Radhakrishnan, S. 1960. The Brahma SūtraŚ The Philosophy of Spiritual Life. London: George Allen & Unwin LTD. ———. 1992. The Principal UpaniṣadsŚ Edited with Introduction, Text, Translation and Notes. New York: Humanity Books. Rai, Ganga Sagar, ed. 1993. lokav rttika of r Kum rila Bhaṭṭa, With The Commentary Ny yaratn kara of r P rthas rathi Mi ra. Varanasi: Ratna Publications. Ramanatha Sastri, S.K, and Subrahmanya Sastri, eds. 1934. M m ṁs -Sūtra-Bh ṣya of abara Sv min, Bṛhaṭ of Prabh kara Mi ra and Ṛjuvimal of likan thaṬ In 5 Volumes. Madras University Sanskrit Series 24–26. Madras: University of Madras. Rambachan, Anantanand. 1991. Accomplishing the Accomplished: The Vedas as a Source of Valid Knowledge in aṅkara. Honolulu: University of Hawaii Press. ———. 1992. “Where Words Can Set FreeŚ The Liberating Potency of Vedic Words in the Hermeneutics of Śa kara.” In Texts in Contexts: Traditional Hermeneutics in South Asia, edited by Jeffrey R. Timm, 33–46. Albany: State University of New York Press. ———. 2006. The Advaita Worldview: God, World, and Humanity. Albany: State University of New York Press. Ram-Prasad, Cakravarthi. 2000a. “Knowledge and Action IŚ Means to the Human End in Mīmāṁsā and Vedānta.” Journal of Indian Philosophy 28: 1–24. ———. 2000b. “Knowledge and Action IIŚ Means to the Human End in Mīmāṁsā and Vedānta.” Journal of Indian Philosophy 28: 25–48. Ricoeur, Paul. 1979. The Philosophy of Paul Ricoeur: An Anthology of His Work. Edited by Charles E Reagan and Stewart. Boston: Beacon Press. Sarma, Rajendra Nath, trans. 1987. The V ky rtha-M tṛk of likan tha Mi ra with His Own Vṛtti. Delhi: Sri Satguru Publications. ———. 1988. M m ṁs Theory of Meaning. Delhi: Sri Satguru Publications. 549 ———. 1990. Verbal Knowledge in Pr bh kara M m ṁs . Delhi: Sri Satguru Publications. Śarma, Śivadatta, ed. 1912. Nirukta, With the Commentary of Durga. BombayŚ Venkateśvara Press. Śarmā, Viś u Prasād, ed. 1970. S ṁkhya-K rik of r mad varakṛṣṇa with the M ṭharavṛtti of M ṭhar c rya. Chowkhamba Sanskrit Series, 296 (Work no. 56). Varanasi: The Chowkhamba Sanskrit Series Office. Śāstri, Anantk iṣ a, and Bhārgav Śāstri Śāstrācārya, eds. 1938. The Brahmasūtra ṅkara Bh ṣya, with the Commentaries Bh mat , Kalpataru and Parimala. Bombay: Nirnaya Sagar Press. Sastri, Kapila Deva. 1964. “Bhart hari’s Discussion on Sāmānādhikara ya.” The Adyar Library Bulletin 28 (1–2): 41–54. Sastri, Kuppuswami, ed. 1984. Brahmasiddhi, With Commentary by aṅkhap ni. Sri Garib Das Oriental Series 16. Delhi: Sri Satguru Publications. Śāstri, Nārāya a, ed. 1890. Brahma-Sūtra-Bh ṣya of aṅkara, with the Ṭ k of nandagiriṬ In 2 Volumes. Ānandāśrama Sanskrit Series 21. PoonaŚ Ānandāśrama. Sastri, Subrahmanya. 1961. Prakaraṇa Pañcik of r likan tha Mi ra with Ny ya-Siddhi. Banaras Hindu University Darśana Series 4. VaranasiŚ Banaras Hindu University. Satyanand, Joseph. 1997. Nimb rkaŚ A Pre- aṁkara Ved ntin and His Philosophy. New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. Sharma, Rama Nath. 1999. The Aṣṭ dhy y of P ṇiniṬ Volume IVŚ English Translation of Adhy yas Four and Five with Sanskrit Text, Transliteration, Word-Boundary, Anuvṛtti, Vṛtti, Explanatory Notes, Derivational History of Examples, and Indices. New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. ———. 2000. The Aṣṭ dhy y of P ṇiniṬ Volume IIŚ English Translation of Adhy ya One with Sanskrit Text, Transliteration, Word-Boundary, Anuvṛtti, Vṛtti, Explanatory Notes, Derivational History of Examples, and Indices. New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. ———. 2002. The Aṣṭ dhy y of P ṇiniṬ Volume IIIŚ English Translation of Adhy yas Two and Three with Sanskrit Text, Transliteration, Word-Boundary, Anuvṛtti, Vṛtti, Explanatory Notes, Derivational History of Examples, and Indices. New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. Shastri, J.L, ed. 1983. Bh gavata Pur ṇa of Kṛṣṇa Dvaip yana Vy sa with Sanskrit Commentary Bh v rthabodhin of r dhara Sv min. Delhi: Motilal Banarsidass. 550 Skinner, Quentin. 2002. Visions of Politics. Volume 1: Regarding Method. Cambridge: Cambridge University Press. Slaje, Walter. 2007. “Yājñavalkya-Brāhma as and the Early Mīmāṁsā.” In M m ṁs and Ved ntaŚ Interaction and Continuity, edited by Johannes Bronkhorst, 115–58. Delhi: Motilal Banarsidass. Smith, David, trans. 2005. The Birth of Kumara by Kali-Dasa. Clay Sanskrit Series. New York: New York University Press. Solomon, E.A. 1968. Avidy — A Problem of Truth and Reality. Ahmedabad: Gujarat University. Srinivasa Chari, S.M. 2002. The Philosophy of the Upanisads: A Study Based on the Comments of aṁkara, R m nuja and Madhva. New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. Staal, Frits. 1979. “The Meaninglessness of Ritual.” Numen 26 (1): 2–22. Stietencron, Heinrich von. 2006. “Theology.” The Brill Dictionary of Religion. Leiden and Boston: Brill. Subramania Iyer, K.A., ed. 1966. V kyapad ya of Bhartṛhari, with the Commentaries Vṛtti and PaddhatiṬ K ṇḍa 1. Deccan College Monograph Series 32. Poona: Deccan College. ———, ed. 1973. V kyapad ya of Bhartṛhari, with the Prak rṇakaprak a of Helar jaṬ K ṇḍa 3, Part 2. Deccan College Monograph Series 21. Poona: Deccan College. Sundaresan, Vidyasankar. 2002. “What Determines Sankara’s Authorship? The Case of the Pancikarana.” Philosophy East and West, no. 1: 1–35. Suthren Hirst, J.G. 2005. aṁkara’s Advaita Ved ntaŚ A Way of Teaching. London and New York: RoutledgeCurzon. Swain, Anam Charan. 1971. “Concept of Hira yagarbha in the Philosophy of Śaṁkara.” Ṛṣikalpany saḥ 1: 126–33. Swami, Gambhirananda, trans. 1937a. Eight UpaniṣadsŚ Volume One ( , Kena, Kaṭha, Taittir ya)Ṭ With the Commentary of aṅkar c rya. Calcutta: Advaita Ashrama. ———, trans. 1937b. Eight UpaniṣadsṬ Volume Two (Aitareya, Muṇḍaka, M ṇḍukya & K rik , Pra na)Ṭ With the Commentary of aṅkar c rya. Calcutta: Advaita Ashrama. ———, trans. 1965. Brahma-Sūtra-Bh ṣya of r aṅkar c rya. Kolkata: Advaita Ashrama. 551 ———, trans. 2003. Ch ndogya Upaniṣad With the Commentary of aṅkar c rya. Kolkata: Advaita Ashrama. ———, trans. 2012. Bhagavadg t With the Commentary of aṅkar c rya. Kolkata: Advaita Ashrama. Swami, Jagadananda, trans. 1949. A Thousand Teachings, in Two Parts—Prose and Poetry, of Sr Sankar ch rya. Mylapore, Madras: Sri Ramakrishna Math. Swami, Madhavananda, trans. 1950. The Bṛhad raṇyaka Upaniṣad with the Commentary of aṅkar c rya. Mayavati: Advaita Ashrama. Swami, Swahananda, trans. 1967. Pañcada of r Vidy raṇya Sw m Ś Edited with English Translation and Notes. Mylapore, Chennai: Sri Ramakrishna Math. Swami, Tapasyananda, trans. 1978. Sankara-Dig-Vijaya: The Traditional Life of Sri Sankaracharya by Madhava-Vidyaranya. Madras: Sri Ramakrishna Math. Taber, John. 1983. Transformative PhilosophyŚ A Study of aṅkara, Fichte and Heidegger. Honolulu: University of Hawaii Press. ———. 1989. “Are Mantras Speech Acts?” In Mantra, edited by Alper P. Harvey, 144–64. Albany: State University of New York Press. ———. 1992. “What Did Kumārila Bha a Mean by Svataḥ Prāmā ya?” Journal of the American Oriental Society 112 (2): 204–21. ———. 2005. A Hindu Critique of Buddhist EpistemologyŚ Kum rila on PerceptionŚ The “Determination of Perception” Chapter of Kum rila Bhaṭṭa’s lokav rttika. London and New York: RoutledgeCurzon. ———. 2007. “Kumārila the Vedāntin?” In M m ṁs and Ved ntaŚ Interaction and Continuity, edited by Johannes Bronkhorst, 159–84. Delhi: Motilal Banarsidass. Tachikawa, Musashi, Shrikant Bahulkar, and Madhavi Kolhatkar. 2001. Indian Fire Ritual. Delhi: Motilal Banarsidass. The Mah bh rataś for the First Time Critically Edited by Vishnu SṬ Sukthankar, with the Co- Operation of Balasaheb Pant Pratinidhi [and Others] and Illustrated from Ancient Models by Balasaheb Pant Pratinidhi. In 19 Volumes. 1927. Poona: Bhandarkar Oriental Research Institute. The Works of Sri Sankaracharya. In 20 Volumes. 1910. Srirangam: Sri Vani Vilas Press. Thibaut, George, trans. 1890. The Ved nta-Sūtras, with the Commentary of aṅkaraṬ In Two Parts. The Sacred Books of the East 34, 38. Oxford: Clarendon Press. 552 ———, trans. 1904. The Ved nta-Sūtras, with the Commentary of R m nuja. The Sacred Books of the East 48. Oxford: Clarendon Press. Thrasher, Alan. 1979. “The Dates of Ma ḍana Miśra and Śaṁkara.” Wiener Zeitschrift Für Die Kunde Südasiens Und Archiv Für Indische Philosophie 23: 117–39. ———. 1993. The Advaita Ved nta of Brahma-Siddhi. Delhi: Motilal Banarsidass. Timalsina, Sthaneshwar. 2009a. “Bhart hari and Ma ḍana on Avidyā.” Journal of Indian Philosophy 37: 367–82. ———. 2009b. “The Brahman and the Word Principle (Śabda)Ś Influence of the Philosophy of Bhart hari on Ma ḍana’s Brahmasiddhi.” Journal of Indian Philosophy 37: 189–206. Tripā hī, Kedāranātha. 1999. Brahmasiddhi of c rya r Maṇḍana Mi ra, With the Commentary Kal by c rya r Ked ran tha Trip ṭh . Vidyāpī ha-Granthamālā 4. New Delhi: Sri Lal Bahadur Shastri Rashtriya Sanskrit Vidyapeeth. Tripā hī, Rama Kiśora. 2003. Upade as hasr of r aṅkarabhagavatp daṬ With the Commentary S hasr vivṛti of r Bhagavad nandajñ n c rya. Sarasvatībhavana- Granthamālā 146. VaranasiŚ Sampurnanand Sanskrit University. Tubb, Gary, and Emery Boose. 2007. Scholastic Sanskrit: A Manual for Students. New York: American Institute of Buddhis Studies. Tull, Herman W. 1989. The Vedic Origins of Karma. Albany: State University of New York Press. Vaidya, Ra ganātha Śāstri, ed. 1918. Saṁkṣepa- r raka of Sarvajñ tman, with the Subodhin of Agnicitpuruṣottama Mi ra and Anvay rtha-Prak ik of R mat rthaṬ In 2 Parts. Ānandāśrama Sanskrit Series 83. PoonaŚ Ānandāśrama. Veezhinathan, N., trans. 1972. The Saṁkṣepa r raka of Sarvajñ tman. Madras University Philosophical Series 18. Madras: Centre for Advanced Study in Philosophy, University of Madras. Venkataramiah, D., trans. 1948. The Pañcap dik of Padmap da. Gaekwad’s Oriental Series, CVII. Baroda: Oriental Institute. Verpoorten, Jean-Marie. 1987. M m ṁs Literature. A History of Indian Literature (Ed. Jan Gonda), Volume VI, Fasc. 5. Wiesbaden: Otto Harrassowitz. Weber, Albrecht, ed. 1855. The atapatha-Br hmaṇa in the Madhyandina- kh with Extracts from the Commentaries of S yaṇa, Harisv min and Dvivedagaṅga. Berlin & London: Fred. Dümmler’s Verlags-Buchhandlung & Williams and Norgate. 553 White, David Gordon. 2014. The Yoga Sutra of Patanjali: A Biography. Lives of Great Religious Books. Princeton and Oxford: Princeton University Press. Whitney, William Dwight. 1993. Sanskrit Grammar. Delhi: Motilal Banarsidass. Yoshimizu, Kiyotaka. 1994. “Bādari and Prabhākara.” In Studies in M m ṁs Ś DrṬ Mandan Mishra Felicitation Volume, edited by R.C Dwivedi, 389–413. Delhi: Motilal Banarsidass. ———. 1996. “The Means of Knowing Apūrva in Śabarasvāmin’s Bhāṣya.” Journal of Indian and Buddhist Studies 45 (1): 513–509. ———. 2000. “Change of View on Apūrva from Śabarasvāmin to Kumārila.” In The Way to Liberation: Indological Studies in Japan, edited by Sengaku Mayeda. Japanese Studies on South Asia 3. New Delhi: Manohar. ———. 2004. “Notes on Kumārila’s Approach to the Ritual Scripture.” In Three Mountains and Seven RiversŚ ProfṬ Musashi Tachikawa’s Felicitation Volume, edited by Shoun Hino and Wada, 735–51. ———. 2007. “Kumārila’s Reevaluation of the Sacrifice and the Veda from a Vedānta Perspective.” In M m ṁs and Ved ntaŚ Interaction and Continuity, edited by Johannes Bronkhorst, 201–53. Delhi: Motilal Banarsidass. ———. 2012. “Kumārila and Medhātithi on the Authority of Codified Sources of Dharma.” In Devadattiyam: Johannes Bronkhorst Felicitation Volume, edited by Francois Voegeli, Vincent Eltschinger, Feller Danielle, Maria Piera Candotti, Bogdan Diaconescu, and Malhar Kulkarni, 643–81. Bern: Peter Lang. Zaehner, R.C. 1957. Mysticism Sacred and Profane. London, Oxford and New York: Oxford University Press. 554