Knowing Illusion: Bringing a Tibetan Debate into Contemporary Discourse, Volume I: A Philosophical History of the Debate, and Volume II: Translations

FEATURE REVIEW Illusions of Knowing Knowing Illusion: Bringing a Tibetan Debate into Contemporary Discourse, Volume I: A Philosophical History of the Debate, and Volume II: Translations. By The Yakherds (José Cabezón, Ryan Conlon, Thomas Doctor, Douglas Duckworth, Jed Forman, Jay Garfield, John Powers, Sonam Thakchöe, Tashi Tsering, and Geshé Yeshes Thabkhas). New York: Oxford University Press, 2020. Matthew T. Kapstein École Pratique des Hautes Études, PSL Research University, Paris matthew.kapstein@ephe.psl.eu Metaphysics is a subject much more curious than useful, the knowledge of which, like that of a sunken reef, serves chiefly to enable us to keep clear of it. Charles S. Peirce, “How to Make Our Ideas Clear” I. The Elusive Way to the Middle Knowing Illusion: Bringing a Tibetan Debate into Contemporary Discourse is the fruit of collaborative efforts by the Yakherds, a group of scholars stemming from the earlier Cowherds, in which two of the present herdsmen—Sonam Thakchöe and Jay Garfield—were active as well.1 In moving from the broad issues of Buddhist philosophy that occupied the Cowherds to a tightly focused study of a particular debate in Tibet, the group has parted from the gentle cattle of the Indian plains to tend more rugged Himalayan stock. The present work, in two volumes, offers an exceptionally lucid guide to some of the main lines of Madhyamaka thought in Tibet and the disputes that these engendered. It constitutes an essential contribution to our knowledge of the philosophical traditions of Tibetan Buddhism, but, beyond this, merits the attention of those focusing on their Indian background and on Buddhist philosophies more generally. Indeed, Know- ing Illusion is among the rare works in this area that may be profitably taken up by philosophers who are not specifically concerned with Buddhist or other Asian ways of thought, but who may wish to explore approaches beyond those of the Euro-American tradition to some of the central topics of philosophy overall: truth and falsehood, knowledge and ignorance, existence and illusion, belief and argument. Of the many interesting Philosophy East & West Volume 73, Number 4 October 2023 1023–1046 1023 © 2023 by University of Hawai‘i Press  questions raised in Knowing Illusion, whether from the perspective of specialized Tibetan Buddhist Studies or of philosophy more broadly, only a sample can be addressed in the space of this review. The unifying thread in Knowing Illusion is a series of written disputes that were provoked by a fifteenth-century teacher belonging to the Sakyapa order of Tibetan Buddhism, Taktsang Sherab Rinchen (1405–1477), in his treatise titled Freedom from Extremes Accomplished through Comprehensive Knowledge of Philosophy (hereinafter Comprehensive Knowledge of Philos- ophy).2 Central to these debates was the challenge of squaring the Madhyamaka philosophy of Nāgārjuna (second century C.E.), whose teaching of ultimate emptiness (sūnyatā) seemed to undermine most—perhaps all— claims to knowledge, with the positive theory of knowledge (pramāṇa) advanced by Dignāga (fifth century C.E.) and Dharmakīrti (sixth century C.E.). The problem had already preoccupied Indian Buddhist thinkers, including philosophers Bhāviveka (sixth century C.E.) and Sāntarakṣita (eighth century C.E.), who sought to harmonize the insights of Madhyamaka with those of the pramāṇa school, and Dharmakīrti’s contemporary Candrakīrti, whose trenchant interpretations of Nāgārjuna appeared to undermine the entire project of a ‘theory of knowledge’.3 In Tibet, thinkers of the former camp came to be known as proponents of Svātantrika-Mādhyamika (so-called because they affirmed that emptiness could be established through positive proof, svatantra), while those allied with Candrakīrti were said to follow Prāsaṅgika-Mādhyamika (who allowed only indirect proof, reductio ad absurdum, or prasaṅga).4 The debate reemerged forcefully in Tibet, and not merely as a historical exercise, in tandem with the expansion of monastic colleges there during the twelfth century, as Tibetan teachers increasingly elaborated original perspectives on the Indian classics. Many aspects of the Svātantrika/ Prāsaṅgika split were contested, for instance, as to whether the distinction applied to reasoning about ultimate reality (Skt. paramārtha-satya, Tib. don dam bden pa) alone, or if Candrakīrti’s critiques extended to our thinking on apparent, relative reality (Skt. saṃvr. ti-satya, Tib. kun rdzob bden pa) as well. To attempt to summarize in the simplest terms the conundrum that Tibetan thinkers found themselves to have inherited from their Indian forebears, we may consider the task of the Madhyamaka philosopher as an attempt to hold together two apparently incompatible propositions: (1) At bottom, all is error. (We are subject to deep and pervasive deception, vitiating the entire field of our ‘knowledge’).5 (2) Appearance, though vitiated by error, is sufficiently rule-like in its operations for us to have something like knowledge of it. (Despite the generally errant state of our ‘knowledge’, there is nonetheless enough trustworthiness in the appearance of things so as to permit us to rightly claim conventional, or relative knowledge.) 1024 Philosophy East & West If we affirm (1) in the strongest sense, (2) is ruled out and we are committed to an extreme skepticism entailing that all claims, even the claim of proposition (1) itself, must be disaffirmed. Silence alone is possible if we are to avoid self-referential contradiction (as exemplified in the famous ‘Cretan Liar’ paradox). If we weaken (1) just a little, opening a loophole that would permit us to embrace some version of (2), we may then accept some things to be relatively true, despite the ultimate truth of universal falsehood. But this may prove to be a slippery slope, leading to a more robust assertion of (2), so that (1) is to all intents and purposes abandoned in most contexts; the way is now clear for us to entertain varieties of foundationalism and realism,6 albeit weakened by the caveat, ‘but ultimately it’s all empty’. For, although all things are ultimately false and all knowledge is ultimately error, superficial appearance turns out to be real enough and our knowledge of it sound enough so that in most instances we really don’t need to worry about what ultimately may or may not be the case at all. We can merrily continue to affirm Madhyamaka’s claims of ultimate emptiness should we be moved to do so, but the affirmation is now largely vacuous, like asserting that, when we have a cubical object that measures 4 x 4 x 4, lo! we also have, in addition to it, a space of the same dimensions that it occupies. Everything, on this account, turns out to be a twofer.7 To untangle the Tibetans’ efforts to cut the Gordian Knot to which they were heirs, the Yakherds begin, following an introduction that usefully explains some of the key terms and concepts structuring their work, by setting out the approaches of three early and influential thinkers who were closely associated with the development of the Tibetan philosophical colleges: Patsab Nyima-drak (b. 1055) (see vol. I, pp. 29–31); Chapa Chöseng (1109–1169) (pp. 31–33); and Mabja Jangchup Tsöndrü (d. 1185) (pp. 33–35). The first, who played an essential role in the diffusion of the works of Candrakīrti in Tibet,8 favored a pronouncedly skeptical reading of Madhyamaka thought that undercut any moves in the direction of the foundationalist and realist leanings attributed to the Buddhist epistemolo- gists. The second, Chapa, was a major proponent of the pramāṇa tradition and with it embraced Svātantrika-Mādhyamika, clearly opting to affirm robustly rational credentials for Madhyamaka without countenancing a drift toward skepticism.9 Given this background, Mabja attempted to reconcile the two trends, reading the Prāsaṅgika approach as admitting, in its account of conventional truth, a place for elements of the epistemological tradition.10 It was Mabja’s line of thought that eventually paved the way for the work of Tsongkhapa Losang Trakpa (1357–1419), the much-revered founder of the Gelukpa order that later, under the Dalai Lamas, came to dominate the political and religious life of Tibet.11 Tsongkhapa, adhering to what was, by his time, a near consensus in Tibetan scholastic circles, affirmed the supremacy of Prāsaṅgika-Mādhya- maka among Buddhist philosophical systems and thus embraced Candrakīr- Matthew T. Kapstein 1025  ti’s interpretations of Nāgārjuna’s views of the universal emptiness of phenomena. But the consensus tended to end there; just how Candrakīrti regarded the relationship between relative and ultimate truths as delineated by Nāgārjuna in his rigorously formulated but paradoxical arguments, and the possibilities of genuine knowledge with respect to the two truths, were among the most sharply contested topics, together with many other specific areas of dispute.12 The Indian and Tibetan controversies around reading Candrakīrti that paved the way for Tsongkhapa’s proposed synthesis are surveyed by the Yakherds in chapter 2, “Reading Candrakīrti.” In chapter 3, “The Philosophical Issues at Stake,” the Yakherds turn to the main lines of the story they wish to tell, beginning with a broad account of the points Taktsang raises in his critique of Tsongkhapa in his Comprehensive Knowledge of Philosophy. Although Taktsang never actually names Tsongkhapa, he enumerates eighteen contradictions that he attributes to “some great scholars in later times,” where it is quite clear that he is referring to Tsongkhapa and his closest disciples.13 The common root of all the eighteen contradictions, he tells us, is taking the Buddha’s gnosis to be the ground for conventional as well as ultimate knowledge.14 As Taktsang suggests, enlightenment thereby becomes entangled in our groping in ignorance.15 The most prominent threads that emerge in the ensuing debates that are dissected by the Yakherds in chapters 4–11 are thus largely related to the problems of grounding our knowledge, that is, the foundations of epistemology, and of characterizing what is presumed to be the highest knowledge, the Buddha’s awakening. In short: how, if at all, are epistemol- ogy and gnoseology connected? Among the admirable accomplishments of Knowing Illusion is to have examined a wide range of Tibetan thinkers’ views about this and related matters, an accomplishment that required very considerable research into poorly studied or previously unstudied sources. The nature of a Buddha’s knowledge is central to these debates (vol. I, chap. 3, p. 82; chap. 4, pp. 89 ff.; chap. 5, p. 114; chap. 7, pp. 136–143; chap. 8, pp. 161–170; etc.). This is not to suggest, however, that what is at stake here should be read as belonging to the philosophy of religion alone, without broader philosophical interest. Consider, for instance, the role of such notions as an ‘ideal observer’ or a ‘view from nowhere’ in contempo- rary philosophy. Does the posited ideal perspective somehow serve to ground judgments and intuitions that appear otherwise to be merely matters of convention, or does it, rather, demonstrate the impossibility of our breaking outside the circle of conventional judgment? This might be the case if we insist, for instance, on our ideal observer’s being perfectly omniscient, a condition that we cannot actualize or even clearly conceive (even if it seems amenable to strictly formal definition). Taktsang thus argues, in effect, that appealing to the Buddha’s gnosis to resolve the problem of the criterion requires us to assume that it participates in the very errors from which it is supposed to be immune, errors that are indissociable from the limitations of 1026 Philosophy East & West conventional knowledge. The alternative is to remove awakening entirely from the realm of mundane appearance: “Thus gnosis of things as they are ends up being neither an ultimate nor a conventional epistemic instrument for knowledge of things in their diversity” (p. 141). Or, as Thomas Nagel once put it, “The objective standpoint can’t really be domesticated. Not only does it threaten to leave us behind, but it gives us more than we can take on in real life.”16 For Taktsang and like-minded thinkers, there is in the end nothing we can know or say of the absolute. We must accept conventional knowing for what it is—conventional—and recognize its fragility. But for Tsongkhapa and his Gelukpa followers, the Madhyamaka propels us toward a sort of stereoscopic vision, through which the awakened individual sees things at once as the ephemeral appearances of conditioned origination and as emptiness. Everything is a twofer. This is captured in a nutshell in one of the Gelukpas’ oft-repeated adages: “vase is not empty of vase, vase is empty of intrinsic nature” (bum pa bum pas mi stong, bum pa rang bzhin gyis stong). But, because in real life not many of us give much thought to ‘intrinsic nature’ before Buddhist philosophers teach us to do so, in this case it is not clear, pace Joni Mitchell, that we’ll ever know what we’ve got when it’s gone. Against this reading of emptiness, as a property somehow parasitic on conditioned things, Taktsang and his allies seem to be urging a type of quietism, abandoning all claims about how things really are. This is seen most forcefully in the writings, among those considered in Knowing Illusion, of the Karmapa hierarchs (vol. I, chap. 10; vol. II, chaps. 2 and 4), particularly those of the ninth Karmapa Wangchuk Dorjé (1556–1603), who goes to great lengths to make clear that he can only offer tentative statements if something must be said, but that he will assert no views himself: Some may ask: “Is Introduction to the Middle Way [by Candrakīrti] one of your school’s texts or not?” We don’t assert anything. “Is Introduction to the Middle Way acknowledged by others a Madhyamaka text according to others?” It is. “And who are those others?” They are mundane people who call Candrakīrti a Prāsaṅgika-Mādhyamika. . . (vol. II, p. 273). Despite his skeptical refusal to assert anything, Wangchuk Dorjé’s brand of quietism nonetheless involves notable attention to the details of method. He remarks, for instance, that he places the rule of double negation in brackets (vol. I, p. 190; vol. 2, p. 272), and he explicitly invokes what some contemporary logicians call ‘explosion’, the principle that, if an absurd conclusion follows from a given argument, then anything might follow from it. In Wangchuk Dorjé’s words, it “would entail the truth of any random idea” (vol. II, p. 272). Indeed, Wangchuk Dorjé is by no means alone, among the thinkers studied in Knowing Illusion, in his emphasis on Matthew T. Kapstein 1027  questions of method. The Panchen Lama, for instance, criticizes Taktsang’s bracketing of double negation (vol. II, p. 174). Part of the interest in these Tibetan philosophers is to be found in their minute concern for fine points of definition and argumentation, a characteristic that their work shares to a remarkable degree with Latin scholasticism and modern analytic philosophy in the West. That this continues to characterize Tibetan philosophical education today is in evidence in the two closing chapters of volume I, contributed by Tibetan scholars Geshe Yeshes Thabkhas and Tashi Tsering, of the Gelukpa and Sakyapa schools, respectively. II. Finding the Right Words The second volume of Knowing Illusion offers translations from the writings of five of the major authors involved in these debates—Taktsang; the eighth and ninth Karmapa hierarchs (Mikyö Dorjé and Wangchuk Dorjé), who were two of Taktsang’s major supporters within the Kagyüpa order; and two of his prominent Gelukpa critics, the Paṇchen Lama Losang Chöki Gyaltsen (1570–1662) and Purchok Ngawang Jampa (1682–1762). With two excep- tions, these works have not been previously available in English versions.17 As the second of these exceptions is not properly acknowledged, something must be said of it here. The fifth chapter of Taktsang’s Comprehensive Knowledge of Philosophy, which is the principle focus of Knowing Illusion and translated in full in volume II (pp. 1–148), was previously translated in part by Jeffrey Hopkins and inserted in the midst of his sprawling rendition of Jamyang Shepa’s Great Exposition of Philosophical Systems.18 Though the Yakherds are clearly familiar with Hopkins’ work, and though they direct readers to it for its treatment of Jamyang Shepa’s critique of Taktsang, discussed in volume I, chapter 9, they nowhere indicate that Hopkins presents fifty pages of translated extracts from Taktsang’s text, or what relation, if any, their own effort has to this earlier publication. As will be seen in the brief sample presented in the appendix to this review, both translations are similarly reliable; in terms of style—both literary and philosophical—the Yakherds’ version seems to be a marked improvement, however. Nevertheless, academic convention might at least have warranted an explicit statement of clarification here. The original texts are written in a difficult register of philosophical Tibetan, involving many technical conventions and abundant quotations from and allusions to earlier Indian and sometimes Tibetan works. To translate such material with reasonable accuracy is a considerable challenge and to do so with a result that reads more or less like English, and not some ungainly variety of ‘Tibetlish’, is a rare achievement. The Yakherds have succeeded very well in both respects. Readers who wish to push beyond the accounts offered in the first volume are therefore urged to delve into the second. Those with a specialized interest in Tibetan philosophical writings 1028 Philosophy East & West will wish to thoroughly study the translations presented here, for they afford valuable illumination on numerous obscure points scattered throughout the primary texts. A brief example will illustrate the Yakherds’ approach to translation. In this passage, Taktsang briefly summarizes his understanding of Tsongkhapa’s view of Madhyamaka and states his principal objection to it (vol. II, p. 25): A person who has understood dependent arising necessarily ascertains both emptiness on the basis of appearance and appearance on the basis of emptiness—and comes to a point where the two are understood as mutually entailing. This is indeed the Madhyamaka approach, and according to the Svātantrikas, mere natures, or essences, exist objectively in conventional terms, and they supply qualifiers, such as absence of real natures. Here, however, without using qualifiers, our opponents claim on the one hand that objects have no nature or essence whatsoever, and on the other that such things as fire’s having the function of burning are not just mere mental imputations, but perform their functions in a nondeceptive, foundational way, albeit merely conventionally. For this reason, they say, the conventional and ultimate nature of phenomena must be used in tandem to induce ascertainment, because explaining the way things function merely in terms of the perspective of delusion does not provide sufficient grounding. They do not just say this once; they say it over and over: this is the principal contradiction to which they fall prey. This is because for the Great Madhyamaka there is absolutely no object of negation—viz., nature—beyond the idea that the functioning world has an objective, foundational nature. In this passage, the final phrase translated as “the idea that the functioning world has an objective, foundational nature” would be more literally “reliable causal efficacy from the side of the object” (yul ngos nas tshugs thub tu bya ba byed pa). Although the Yakherds’ rendition thus involves some degree of paraphrase—for instance, in making explicit that the ‘object of negation’ here is an idea—I believe that they have found a clear and elegant voice through which to express what the Tibetan text actually says. The Yakherds introduce some of the essential terms and argument forms used in their work in the introduction to volume I, with additional discussions of key terms and concepts throughout. A few points that aroused my attention follow, not so much to signal disagreements with the Yakherds’ approach, but rather to underscore questions that may merit further scrutiny. The Yakherds rightly remark that the common use of syllogism to describe probative arguments in Indian and Tibetan philosophical works is misleading, as that term “might suggest that Indian arguments are Aristote- lian in form” (vol. I, p. 12). In its stead, they propose employing the more neutral expression probative argument. Fair enough. But they then go on to redescribe certain of the key terms of Indian argument: for vyāpti (Tib. khyab pa), usually rendered ‘pervasion’, they adopt ‘entailment’ or ‘subcategoriza- tion’, and for liṅga (Tib. rtags), often ‘sign’, they employ ‘premise’. This Matthew T. Kapstein 1029  requires arguments to be reformulated to avoid the contours of the feature- placing logic that is in fact used in Indian and Tibetan sources, which will be unfamiliar to some readers.19 This reinterpretation leads the Yakherds to speak, at several points, of major premises and minor premises, which results in their redeployment of much of the paraphernalia of the Aristotelian syllogism that they apparently sought to avoid (e.g., in vol. II, pp. 270 ff.). None of this is to suggest that they misconstrue the arguments or present them in a misleading fashion—their constant effort is to make Tibetan forms of argument intelligible to anglophone readers and in this they largely succeed—but we do not find a clear model here for the close analysis of how arguments, in these texts, actually work. This has been an ongoing challenge for scholarship on Indian and Tibetan philosophies; the Yakherds illustrate that it has yet to be adequately overcome. The terminological conventions adopted by the Yakherds are presented in detail in useful glossaries at the end of each volume (vol. I, pp. 293–314; vol. II, pp. 407–424) listing the English terms employed in Knowing Illusion with their Tibetan and Sanskrit equivalents. Overall, their translation choices are judicious and clear; I find just a few points requiring reconsideration.20 ‘Nihilism’, for instance, is listed as the translation for both ucchedavāda (Tib. chad lta) and apavāda (Tib. skur ’debs), but, whereas it is a justifiable rendering of the first, the second is not an ‘-ism’ at all; apavāda is deprecation, disparagement, or denial. In some contexts, this may of course amount to an expression of nihilism (as when one denies that there is any basis for morality), but that need not invariably be the case. Elsewhere, we find the important doctrinal term nyon mongs (Skt. klesa) translated as pathology, pathological. This is misleading. The klesas, adverse states of mind, include ordinary desire, aversion, and stupidity, conditions that are not considered to be pathological except when they occur with abnormal intensity. For classical Buddhism, what we often think of as healthy sexual desire, for instance, is no pathology, but it is nevertheless a klesa. Both pathological lying and a garden-variety fib are, in Buddhist terms, similarly due to klesa.21 One of the few relevant terms that is not entered into the glossaries is ‘other emptiness’ (gzhan stong). In their discussion of this concept (vol. I, p. 201), the Yakherds erroneously characterize it as “the notion that that of which things are empty is entirely other than the things themselves.” But ‘other emptiness’ as used here is emphatically not a general doctrine pertaining to ‘things’. It applies, rather, to just one thing—as Taktsang puts it (vol. II, p. 134), “the ultimate truth is empty of everything other than itself” (emphasis added). Taktsang rejects this doctrine, which was the hallmark of the Jonangpa order of Tibetan Buddhism, as do the other authors treated by the Yakherds, so that, despite its importance in the history of Tibetan Buddhist thought overall, it does not figure very prominently in these pages. 1030 Philosophy East & West  One final terminological point that should be mentioned concerns the Yakherds’ treatment of ‘foundationalism’. It is evident that they read this topic, which came to occupy a central place in late-twentieth-century epistemology among analytic philosophers,22 as having particularly strong analogies with the Tibetan debates they cover. I agree that the similarities are often striking. However, it is less clear that the Tibetans have a word for it, if what we are looking for is an ‘-ism’ word (for there are plenty of Tibetan expressions that, depending on context, might be legitimately rendered ‘foundation’). A phrase that is sometimes used in the Tibetan texts, and translated here as meaning ‘nonfoundationalism’, is apratiṣt. hāna (Tib. rab tu mi gnas pa). It literally means ‘without fixed abode’ and appears to have been employed in some late Indian works and Tibetan writings of the early second millennium to refer roughly to what subsequently became known as Prāsaṅgika-Mādhyamika. The history of the term, in connection with Madhyamaka, is the subject of an excellent study by Orna Almogi to which the Yakherds duly refer.23 It is clear, based on the sources gathered by Almogi, why one might regard apratiṣt. hāna as at least entailing what we think of as nonfoundationalism. For instance, in a passage attributed to the eleventh-century master Atisa: “the true nature of all phenomena is [that they] have had no substratum since primordial times.”24 Nevertheless, although not entirely absent, questions of formal epistemology are not notably stressed in the texts Almogi studies.25 I would like to suggest that while apratiṣt. hāna may well entail nonfoundationalism, this is not quite what the term means. Almogi interprets it, accurately enough, as meaning ‘having no substratum’, but generally leaves it untranslated. One issue she does not address that I think is crucial here is just where the expression came from. One of the characteristic doctrines that emerged in early Mahāyāna Buddhism and that aroused the attention of pioneering scholars such as Sylvain Lévi, Louis de la Vallée Poussin, Theodore Stcherbatsky, and Étienne Lamotte, is the notion that, in contradistinction to the goal of the so-called ‘Hīnayāna’, nirvana is not presented in Mahāyāna sources as a static end conceived as final extinction. It is, rather, ‘nirvana without fixed abode’ (apratiṣt. hitanirvāṇa), that is, without adherence to the extremes of worldly existence or of nirvana-as-annihilation.26 It seems to me impossible that authors schooled in Mahāyāna Buddhist scriptures could have used a term like apratiṣt. hāna without reference to this background. The end of Madhyamaka thought, for the exponents of apratiṣt. hāna, was strictly analogous to buddhahood; the particular question of epistemological foundationalism was a concern when it arose, but it was not at the forefront of their thought and was not a banner about which they rallied. I will have more to say of the implications of this below. Matthew T. Kapstein 1031  III. Playing for Keeps What were these, and similar, Tibetan debates really about? What purpose did they serve? The easy answer is to say that Tibetan thinkers were interested in getting at the truth of things, in determining the nature of reality. But this is too easy. It ignores social and historical considerations that may prove to be inconvenient for ideal assumptions.27 This can be illustrated with reference to the reception history of Taktsang’s treatise. Although the Yakherds speak of the “publication” of the work (vol. I, p. 14), this requires some qualification. When Taktsang completed the verse part of his text in 1463, it appears to have been circulated in manuscript among close students and associates, who then requested that he clarify it by adding a commentary.28 This was composed sometime after the initial work, but there is no indication of just when the commentary was written. Be that as it may, it was, like the verse text, available only in manuscript and does not appear to have been widely distributed. It was only some years after Taktsang’s death that his disciples sought to honor their master by publishing his collected writings xylographically. As no copies of this editio princeps have yet become available to scholarship, if indeed any are preserved,29 we cannot say much more about this edition or its diffusion, besides the few telling details noted below. The Panchen Lama, in his refutation of Taktsang (translated in vol. II, chap. 3), makes it clear that, by his time, the critique of Tsongkhapa contained in the work was well known in Gelukpa circles, but he does not mention whether he was consulting a manuscript or a print. The printed xylographic edition of which we are now aware, which served as the basis for all subsequent modern editions, was published, however, during the seventeenth century by the leading disciple of the Panchen Lama, the Fifth Dalai Lama, and includes the latter’s colophon, which expresses nothing but the highest praise for Taktsang.30 As he tells us in his autobiography, he had studied this among other works by Taktsang when he was in his early twenties.31 The Yakherds, indeed, note that the Fifth Dalai Lama’s favorable opinion of Taktsang presents a puzzle (vol. 1, p. 124), but they have little to say about it. The problem is even more striking because the Fifth Dalai Lama and his government were notorious banners of books, including the writings of several of Tsongkha- pa’s critics, and including, it appears, the aforementioned editio princeps of Taktsang’s collected works.32 How are we to understand this deeply conflicted record? One fruitful way of looking at Tibetan philosophical disputes has been inspired by their similarities to competitive games.33 It is difficult, for example, when thinking of reports of Kalmyk Mongol scholars at Lhasa’s Drepung monastery, not to see some analogy between their passion for 1032 Philosophy East & West debate in the monastic court by day and their avid chess-playing in their quarters during the evening.34 This raises the question: are Tibetan works of philosophy like those studied in Knowing Illusion perhaps better compared with Irving Chernev’s classic, The Most Instructive Games of Chess Ever Played, than with the Western philosophical classics to which we often refer? (Here I would not wish to exclude the possibility of treating much of Western philosophy as a game as well, reminding us sometimes of Hesse’s imagined Glasperlenspiel.) Most pertinent here, as Chernev reminds us in the opening words of his introduction, is that “Chess masters play to win.”35 Tibetan monastic debates were also games played to win, and every great game requires great opposition. No champion shines against feeble rivals. So one way of looking at the Fifth Dalai Lama’s promotion of one of Tsongkhapa’s greatest critics is to imagine him saying to his disciples in the debate court, “this is the game you’ve got to beat.” Such an approach would help us to understand, too, why questions of philosophy in Tibet were seldom considered to be firmly settled, why we do not see a succession of philosophical movements supplanting one another, but rather an unceasing series of games, extending from the twelfth through the twentieth centuries, inviting myriad moves and endless fascination much in the manner of chess. This is very well exemplified in the work of the eighth Karmapa, Mikyö Dorjé (1507–1554), who appears to aim refutations not only at doctrinal opponents, but even at his own previous incarnations (vol. 1, pp. 201–203). The Yakherds express some consternation about this, but the situation becomes clearer when we consider that Mikyö Dorjé was known to have adopted positions in his writings with which he did not personally identify, and may even have gone so far that, after failing to receive a response to one of his own polemical works, he set about to refute it himself.36 He sometimes seems to resemble a prizefighter who sets up his ring in a public square to take on all comers. I will return to consider what he may have been up to, besides just bravado, below. The perspective I have been urging, which regards the debates studied by the Yakherds as games, finds some support, curiously enough, in a passage from a biographical account of Jamyang Shepa. As the Yakherds report it (vol. I, p. 179): “[Jamyang Shepa] placed that extraordinary book [Taktsang’s Comprehensive Knowledge of Philosophy] on a high bookshelf so that the book could not be roughly used. He worshipped it.”37 They comment that, “This may well be an apocryphal story, and it is clearly at odds with his denigration of Taktsang’s text in Great Exposition of Philosophical Systems . . . .” It is unfortunate, in this instance, that the Yakherds did not consult the original Tibetan account that they cite based on a summary translation. For clearly the anecdote relates to the period during which Jamyang Shepa was composing his Great Exposition. Most important, however, is the line just after their citation, which is mistranslated in their source; in fact, the Tibetan text states that “Jamyang Shepa Matthew T. Kapstein 1033  sometimes turned his attention [to Taktsang’s book] and, wrapping his upper robe about his waist [as was customary in the debate court], acted as if debating [with it].”38 This seems to me to be entirely plausible and by no means evidently apocryphal, given Jamyang Shepa’s concern, in the Great Exposition, to refute Taktsang. He honored Taktsang’s book in his private quarters not because he agreed with it, but because he knew that it was the one to beat.39 I do not wish to suggest, however, that we should reduce our understanding of philosophical dispute in Tibet to a notion of games, if games be thought of as lacking serious intent. For games are often deeply serious matters; depending upon the game and the culture in which it is played, games may determine honor or dishonor, the gain or loss of position or fortune, even life or death. Games are theaters for the cultivation of cunning, dexterity, endurance, humor, courage, magnanimity, and more. All of this was in play in the debate courts of Tibet. Sometimes, too, Tibetan philosophical debates had demonstrable political ramifications, despite the apparent distance of their subject matter from mundane concerns. For the Gelukpa authors studied by the Yakherds, for instance, the assertion of Tsongkhapa’s philosophical supremacy corresponded to a generalized Gelukpa triumphalism that was forcefully asserted, under the administration of the Fifth Dalai Lama and his successors, in the social and political spheres. And, as we have seen, when debate alone seemed insufficient, it was possible to turn to book-banning instead. IV. Beyond Nonfoundationalism The bête noire for Madhyamaka thought, as presented in Knowing Illusion, was foundationalism, whatever may be said of the nuances of the Sanskrit or Tibetan vocabulary used in this context. There appears to have been broad agreement in Tibet that the efforts of the Buddhist epistemological schools to determine incorrigible grounds for knowledge could not, in the end, be maintained. For Candrakīrti, this had meant that Dignāga and Dharmakīrti’s attempt to reduce the epistemically sure warrants to two—direct perception and inference—was to be jettisoned. Whatever warrants held sway in the world—perception, inference, testimony, and analogy were the ones Candrakīrti named (vol. I, p. 37 n. 47)—were to be accepted in the context of conventional discourse, but without taking them to be final grounds. Conventions are conventions and that is all. What seems to follow is an extreme epistemological modesty, with no expectation that there is a deeper, truer, better story to be told about knowing. Like Lévi-Strauss’s bricoleur, we address the projects that confront us with the tools at hand and can be pleased with our handiwork if it succeeds in fulfilling the needs of the day. From such a perspective, it seems inevitable to conclude, based on the materials presented in Knowing Illusion, that Tsongkhapa and his Gelukpa 1034 Philosophy East & West successors sought to have their cake and eat it, too. This emerges most clearly, I think, in connection with the arguments of Khedrupjé (vol. I, chap. 4, p. 99), to hold that the four warrants mentioned above in fact reduce to the two affirmed by the epistemologists. Prāsaṅgika-Mādhyamika philoso- phers, on this reading, can also be exponents of the pramāṇa system, so long as they reject ‘intrinsic natures’. They must just accept that some conventions (e.g., there being two warrants), like Orwell’s more equal animals, are somehow better than other conventions (there being four warrants or more), without discernible grounds for so doing.40 Try as I may, I cannot see my way clear to a genuinely ‘nonfoundationalist’ reading of this move. Their opponents, Taktsang and the Karmapa hierarchs, by contrast, do seek to pursue the demolition derby to its end. For this reader, at least, one of the most radical expressions of this emerges in the work of Karmapa Mikyö Dorjé. Among the many points at which the textual scholarship of the Yakherds shines is in their discovery of passages in Mikyö Dorjé’s sprawling œuvre (some 26 volumes totaling over 15,000 pages), in which he appears to address pointed criticism to his much venerated predecessor, the third Karmapa hierarch Rangjung Dorjé (1284–1339), setting his sights on two of the latter’s most esteemed works (vol. I, pp. 201–202), the Profound Inner Meaning (Zab mo nang don) and Aspiration of Mahāmudrā (Phyag chen smon lam). The Yakherds regard Mikyö Dorjé’s harsh com- ments about them as evidence of “an impressive disregard both for the party line and for political gain” (p. 203). Although this much seems true, I do not think that the Yakherds have fully grasped the eighth Karmapa’s intention. Curiously, their slight mis- construal of what I believe to be really at stake here mirrors the small misstep I detect in their treatment of apratiṣt. hāna as ‘nonfoundationalism’ (see above). To see what is involved here, we may consider the recent book of Swiss philosopher Michael Hampe, What Philosophy Is For. Here, he interrogates the value and use of assertions: [People] form groups around certain assertions, around which they gather as ‘adherents’ of a theory as if around a campfire. These campfires, then, are called ‘superstring theory’ and ‘panselectionism,’ ‘Keynesianism’ or ‘Spinozism.’ Adherents of this type consider certain systems of assertions to be true and are criticized by others who disagree. . . . [U]ttering a ‘valid’ assertion serves less the purpose of informing another person than ascertaining that through its use, one participates in the community of people who speak and opine in this or that way. He contrasts this with the attitude of Socrates: Socratic philosophical activity aims at learning the ability to conduct one’s life without having to cling to assertions rigidly. . . . Socratic philosophizing is not Matthew T. Kapstein 1035  concerned with the right as opposed to the wrong judgment. Rather, it is a liberating of the soul from the straightjacket of judgments. . . . Its aim, then, is the autonomy of individuals and not the truth of generalities.41 It is in the light of this Socratic freedom from the tyranny of assertion, I think, that we may be posed to appreciate Nāgārjuna’s much-discussed refusal of theses (vol. I, pp. 25–27, “What’s Wrong with Theses?”) or Wangchuk Dorjé’s sometimes convoluted way of saying and unsaying at once (see above), in which even asserting Candrakīrti’s textual authority for Madhyamaka is set aside. Here, too, we may begin to appreciate the true sense of apratiṣt. hāna, philosophizing without fixed abode, not even affirming ‘nonfoundationalism’ insofar as that may be taken, too, as a philosophical campfire around which we might gather.42 Finally, this allows us to see more clearly what Mikyö Dorjé was up to in heaping scorn on the words of his illustrious predecessor; for he saw that even the transcendent teaching of the Mahāmudrā might be dogmatized and thereby lose its true sense. That said, Knowing Illusion brilliantly illuminates materials that have often seemed obscure and succeeds in demonstrating, as the Yakherds affirm in their conclusions (vol. I, p. 229), that . . . by the fifteenth century, Tibetans had articulated sophisticated critiques of foundationalism in epistemology, sophisticated skeptical attacks on the very project of epistemology, and sophisticated metatheories of epistemology that anticipate contemporary naturalized epistemology. Thus, these debates are not entirely antique; they continue today. The Yakherds abundantly demonstrate this to be so. Knowing Illusion will, I hope, be taken up widely by philosophical readers concerned to know what it might mean to know. Appendix I reproduce here a brief passage from Taktsang’s treatise, in the translations of Jeffrey Hopkins and of the Yakherds. In terms of the meaning, both translations are substantially similar; in my view, both convey the sense of Taktsang’s words to a high degree of accuracy. That there is a difference of translation style will be evident. For those who know Tibetan and might wish to compare, Taktsang’s text is given as well. Hopkins, p. 537: First Contradiction. That all objects are false and that [consciousnesses] having them as objects are non-deceptive are contradictory. [Tsongkhapa’s] explanation[,] that if one does not understand that [objects] are false, one does not realize the meaning of ‘obscurational’ (kun rdzob, saṃvr. ti) [,] is very good, but while realizing that conventionalities are false: 1036 Philosophy East & West  • his assertion that in the Consequence School’s own system all conventional objects are false and deceptive • and his assertion that awarenesses that have those as objects are non- deceptive are contradictory. For if an object is a deceptive phenomenon, it is not possible for an awareness of it to be a valid cognition, as is the case, for example, with an awareness to which falling hairs appear. Yakherds, vol. II, pp. 33–34: All objects being false contradicts their subjects being non-deceptive. Our opponents explain that unless one realizes that the object is false, one will fail to understand the meaning of relative truth. This is exactly right. In the Prāsaṅgikas’ own system, one indeed realizes that the relative truth is false. Therefore, it is contradictory to hold on the one hand that all relative objects are false and on the other hand that the cognitions that are their subjects can be nondeceptive and epistemically warranting. This is because when an object is a deceptive phenomenon, its cognition cannot possibly be an epistemic warrant, such as with a cognition to which illusory hairs appear. Stag tshang lo tsā ba shes rab rin chen gyi gsung ’bum pod dang po (Beijing: Krung go’i bod rig pa dpe skrun khang, 2007), p. 109 (root text) and pp. 273–274 (commentary): [109] yul kun rdzun dang yul can bslu med ’gal|| [273–274] rdzun par ma go na kun rdzob gyi don ma rtogs par bshad pa ni shin tu legs shing| thal ’gyur rang lugs la kun rdzob rdzun par rtogs bzhin pas kun rdzob gyi yul kun rdzun zhing slu bar ’dod pa dang de’i yul can gyi blo bslu med kyi tshad mar ’dod pa ’gal te| yul de slu chos yin na blo de tshad mar song mi srid pa’i phyir| dper na skra shad snang ba’i blo bzhin no|| Notes 1 – Refer to Cowherds 2011 and Cowherds 2016. Members were also involved in the following: D’Amato, Garfield, and Tillemans 2009; Tanaka, Deguchi, Garfield, and Priest 2015. Overlapping with the Cowherds and Yakherds, and thus perhaps meriting attribution to the Dzoherds (the dzo being the yak-cow hybrid), one may also note Duckworth, Eckel, and Garfield, et al. 2016. 2 – I have not usually bothered here to provide exact transcriptions of Tibetan proper names, which will be familiar to specialists but are simply an annoyance to others. I have, however, retained exact transcriptions for technical terms and book titles given parenthetically, Matthew T. Kapstein 1037  in notes, or in the appendix. The Yakherds include exact transcriptions for all Tibetan names and terms they use; in so doing, however, they incorrectly transcribe the syllable blo as bLo when it begins a proper name. It should be Blo in this context, even though it is the l that is in fact pronounced. 3 – A recently published reference work—Edelglass, Harter, and McClin- tock 2022—includes articles on the Indian thinkers mentioned here, with extensive references to the relevant scholarship about them. 4 – The Yakherds rightly state (p. 9) that Svātantrika-Mādhyamika and Prāsagika-Mādhyamika are not known from Indian sources but are modern Sanskrit renderings of the Tibetan names of these sub-schools: dbu ma rang rgyud pa and dbu ma thal ’gyur pa. It should be noted, however, that the terms svātantrika and prāsaṅgika, designating styles of argumentation rather than particular schools, are indeed current in Sanskrit. The philosophical issues that are pertinent in the present context are surveyed by the Yakherds, vol. I, pp. 52–60. 5 – The reading of Candrakīrti, following Tom Tillemans, as a “global error theorist,” is introduced by the Yakherds, vol. I, pp. 45–48. I note in passing that Buddhism is in general highly susceptible to forms of a global error theory. This stems from the fundamental—dare I say, ‘foundational’—role of avidyā, ignorance, in all conditioned experi- ence, as is affirmed throughout the Buddha’s teaching. A central problem for Buddhism, as similarly for Advaita-Vedānta, is to explain how, given our benighted condition as subjects of avidyā, we can bootstrap ourselves to a state of genuine knowledge. 6 – ‘Foundationalism’, as I understand it, applies primarily to epistemolog- ical theories asserting that there are incorrigible grounds upon which the edifice of our knowledge is securely built; ‘realism’ refers to metaphysical theories affirming there to be a true story about the reality of things that can be known. Thus, in this usage, ‘idealism’, taken as the view that external objects are not themselves real but are mind-dependent, entails a form of realism about minds. ‘Antirealism’ is currently used to refer to philosophies that dismiss the assumption that there is a determinate truth of the nature of reality. Advaita-Vedānta and Madhyamaka have thus both sometimes been characterized as antirealist in recent writing about them. It may be noted that the Yakherds sometimes seem to use ‘foundationalism’ to cover both foundationalism and realism as I define them here (as when speaking of ‘foundational objects’, vol. I, pp. 151 ff.), though in defining the concept (vol. I, p. 11), they stress the epistemological uses. 7 – Something like this is perhaps entailed by the recent work of one of the Cowherds’ collaborators, the logician Graham Priest, who in The 1038 Philosophy East & West  Fifth Corner of Four (Priest 2018) argues for treating emptiness as a distinct value, e, attached to states of affairs that may also be true, false, both, or neither. 8 – The Tibetan reception of Candrakīrti is studied in Vose 2009. 9 – On Chapa’s contributions, see Hugon and Stoltz 2019. 10 – For a detailed study of Mabja, refer to Doctor 2014. 11 – For an accessible overview of Tsongkhapa’s philosophical contribu- tions, see Jinpa 2002. Much has been written of the rise of the Gelukpa order and of the administration of the Dalai Lamas. Particularly pertinent in the present context is Sullivan 2020. 12 – For instance, whether or not an adherent of Prāsaṅgika-Mādhyamika could affirm the Yogācāra doctrines of the ‘ground consciousness’ [ālayavijñāna] and ‘reflexive awareness’ [svasaṃvedana] even in conventional terms. Tsongkhapa denied both. These issues, however, do not figure too prominently in the present work (though they are mentioned in vol. I, p. 64). Tsongkhapa’s unique interpretation of Candrakīrti’s philosophy was summarized in his approach to “eight major points of difficulty” (dka’ gnad chen po brgyad), the subject of a series of lectures recorded by his disciple Gyaltsapjé; see Ruegg 2002, pp. 137–280. 13 – Taktsang’s exact words, as translated in Knowing Illusion, vol. II, p. 24, are: “Certain recent Tibetan holders of the teachings who were great scholars propounded Candrakīrti’s system accurately and faithfully early in their careers. Nonetheless, they later deviated, but not due to any improvement in their logical acumen.” This perhaps deserves to be unpacked more clearly than it is by the Yakherds. At issue is the fact that Tsongkhapa and a number of his associates and disciples were schooled in Candrakīrti’s Madhyamaka by the great Sakyapa teacher Remdawa (1349–1412), whom Taktsang greatly admired, but then later adopted Tsongkhapa’s novel understanding. This passage is directly parodied by the Panchen Lama in introducing his refutation of Taktsang: “In the later part of his life, the Dharma lord Drapa Sherab Rinchen, Taktsang Lotsawa, obtained unswerving faith in the lord Tsongkhapa, but in his midlife, motivated by devious worldly intrigues and by ignorance and misconceptions, he wrote a polemical work against the Lord Master Tsongkhapa . . . ” (vol. II, p. 168). As the Yakherds, however, show (vol. I, chap. 6), the Panchen’s assertion of Taktsang’s late life conversion is certainly a fiction. 14 – Knowing Illusion, vol. II, p. 30: “this idea that the gnosis of buddhas is the primary epistemic warrant for discerning the relative truth is a primary flaw of this system.” Matthew T. Kapstein 1039  15 – Op. cit., p. 31: “the relative truth is referred to as confusion, it is constituted by mundane and innate ego- grasping, and it is not seen by pure gnosis.” 16 – Nagel 1986, p. 231. 17 – The first exception is a short passage from the Eighth Karmapa’s commentary on Candrakīrti’s Madhyamakāvatāra, given in volume II, chapter 2, pp. 160–164, the text from which it has been drawn having been translated earlier in Goldfield et al. 2005. Other major works not included in volume II but considered by the Yakherds in volume I and translated elsewhere, of which readers may therefore wish to be aware, include: Cabezón 1992, discussed in Knowing Illusion, volume I, chapter 4; and Hopkins 2003, the subject of chapter 9. 18 – Hopkins 2003, pp. 527–575. 19 – To attempt an informal capsule statement of the system of inference at work here: if a feature (Skt. dharma, Tib. chos)—for instance the color red—occurs wholly within a specified domain (Skt. pakṣa, Tib. phyogs), such as that of color, it must exemplify whatever features have an extension equal to or greater than the domain in question, i.e., the features that ‘pervade’ (Skt. vyāp-, Tib. khyab) it; in this case, features such as being visible, being extended, etc. A given feature or a particular bearer of a feature may then be taken in argument as a sign, entailing the occurrence of the other features by which its domain is pervaded: being a red robin entails being a visible thing, because it has a color, because it is red. As the concern is with features and the loci in which they occur, the forms of argument developed on this basis are not formally identical to Western propositional or predicate logics and so must be ‘converted’ to be suitably expressed in them. Though the system of reasoning employed in Indian Buddhist sources was taken over by the Tibetans, the latter modified, to some extent, the way arguments were formulated. 20 – One very odd point to note, however, is the phrase mi za ba, for which the Sanskrit equivalent is given as abhuñjat and the English as ‘nonappearance’ (vol. I, p. 300; vol. II, p. 413). In this case, the Sanskrit and Tibetan alike mean ‘not eating’, ‘not consuming’, and the Yakherds’ reason for using ‘nonappearance’ remains unexplained. The mystery is compounded by the fact that ‘nonappearance’ seems to occur nowhere in volume I, and, where it does occur in volume II (p. 208 n. 119), it is in a footnote discussion of a text in which the Tibetan actually used is not mi za ba but rather mi snang ba, which is indeed correctly rendered ‘nonappearance’. 1040 Philosophy East & West 21 – Some may respond that Buddhists treat common faults as pathologies, but I doubt that this is really correct. In Tibetan medicine, for instance, the occurrence of the klesas is considered a universal condition in healthy and unhealthy bodies alike. See Gyatso 2015, pp. 202–203. 22 – The critique of foundationalism, notably as articulated by analytic philosopher Wilfrid Sellars (1912–1989), is indeed a leitmotif in the recent work of Garfield and several of his colleagues among the Cow- and Yakherds. See Garfield 2019, and my review, Kapstein 2021. 23 – Almogi 2010. 24 – Ibid., p. 162. 25 – In fact, epistemological issues appear more prominently in discussions of the position that is contrasted with the Apratiṣt. hānavāda, that is, the Māyopamavāda, the view of reality as ‘illusion-like’, that in some respects seems to anticipate the idea of Svātantrika-Mādhyamika. See the discussion of Jñānavajra’s account of the debate in Almogi 2010, pp. 146–152. 26 – I believe that the first modern scholar of Buddhism to broach the issue was Sylvain Lévi in his 1907–1911 editio princeps of the Mahāyāna- Sūtrālaṃkāra (Paris: Honoré Champion). 27 – I do not wish to suggest that the Yakherds ignore relevant historical background; on the contrary, they provide useful historical summaries of most of the figures they treat and indicate significant aspects of Tibetan thought that must be understood with reference to social context. Chapter 6 is particularly strong in this respect. While their historical sketches are mostly accurate, a detail requiring correction is found in their reference, in connection with Purchok Ngawang Jampa (chap. 11), to Tibet’s having been “invaded by an army of Khoshud Mongols in 1717 [during which invasion the] leader of the Mongol forces . . . ordered the beheading of the regent” (pp. 211–212). How- ever, the events involving the Khoshud Mongols, which they mention, occurred in 1705–1706; in 1717 there was a ferocious invasion of Tibet by a rival Mongol faction, the Zunghar, who had been allied with the deposed regent. Nevertheless, it is probably still true, as the Yakherds say (p. 212), that “Purchok largely distanced himself from Tibet’s complex political affairs.” 28 – The year is specified in the colophon of the verse text, in Stag tshang lo tsā ba shes rab rin chen gyi gsung ’bum pod dang po (Beijing: Krung go’i bod rig pa dpe skrun khang, 2007), p. 120; the brief colophon of the commentary, p. 360, adds that the author was urged by many to write it. Matthew T. Kapstein 1041  29 – The editors of Taktsang’s available writings, op. cit., p. 9, mention that many of the manuscripts on which their work was based contained copies of the print-colophons of these early editions, but that the original prints were evidently lost (nyams chag shor ba mngon). 30 – The Yakherds do not include this colophon in their translation, but they do allude to the Fifth Dalai Lama’s ordering the publication (vol. I, p. 124 n. 6). The Dalai Lama’s colophon, which makes clear that the volume also included Taktsang’s Comprehensive Knowledge of the Sciences (rig gnas kun shes), is reproduced in Stag tshang lo tsā ba, op. cit., p. 361. 31 – Karmay 2014, p. 149. 32 – On the banning of Taktsang’s works, together with the writings of many others, see Smith 2005. The Yakherds imprecisely state (p. 131) that “Geluk-controlled printing presses continued to copy his works,” but this was not true of his works as a whole; only the single volume commissioned by the Fifth Dalai Lama was available, though old copies of some of Taktsang’s other writings must have remained in circulation. The Dalai Lama himself sheds some light on his thinking in his commentary on Candrakīrti’s Madhyamakāvatāra, written in 1647. There, he tells us that he considers that, though Taktsang’s critique of the supposed contradictions in Tsongkhapa’s work merits a refutation, he demurs from undertaking this himself, in part owing to Taktsang’s important place in the lineages in which the Dalai Lama himself received instruction in Sanskrit grammar and related subjects. The Dalai Lama’s publication of Taktsang’s Comprehensive Knowledge of the Sciences (see note 30 above) further reflects this. 33 – The interpretation of Tibetan debate as a game was first clearly articulated by Tom Tillemanns (1999, p. 120), who tells us that it embodies “a system of moves, responses, rights and obligations in a rigidly structured game.” Most subsequent scholarship on Tibetan debate has been informed by this perspective. 34 – I am grateful to the late Kalmyk Geshe Wangyal (1901–1983) for this recollection of his student days at Drepung during the 1920s and 1930s. 35 – Chernev 1965, p. 15. That the Yakherds’ view about this is perhaps not too distant from my own is suggested by their decision to translate the Tibetan expression ’khor gsum, exclaimed when a debater has succeeded in placing a clincher, as ‘checkmate!’ 36 – Mikyö Dorjé’s commentary on the Abhisamayālaṃkārasāstra, written defending an ‘other-emptiness’ (gzhan stong) position that he otherwise rejected, is a clear example of his dialectical versatility. His supposed 1042 Philosophy East & West  self-refutation is a polemical work, the Gsang sngags snga ’gyur las ’phros pa’i brgal lan, contained in the third volume of his collected works, written in response to an especially harsh critique of the teachings of the Nyingmapa order that had been diffused under his name. However, he begins by denying his authorship of it, despite its attribution to him. The case is therefore not entirely clear. 37 – The work cited is Sadhukhan 1991. The second element of the Tibetan name should be corrected to bzhad pa rather than bshad pa as given here. 38 – Dkon mchog bstan pa rab rgyas 1982, p. 363: ’di dus stag lo’i grub mtha’ kun shes spyan drangs| khong gzhan dang mi ’dra bas gang byung byed mi rung gsung stegs mthon po la bzhugs te mchod pa bshams bcug| thugs gtad nas bzhugs pa’i mtshams mtshams su sku gzan rked par bkris te rtsod pa’i rnam pa mdzad|. Sadhukhan (1991, p. 21) renders the last line: “Placing it attentively, he at times did religious discourses by wrapping the sacred cloth (used on upper body) round the waist.” The text, however, explicitly references feigned debate practice, rtsod pa’i rnam pa. 39 – We should note, too, that ‘denigration’ of one’s opponent was a common rhetorical technique in Tibetan debate to which one (or the members of one’s party) resorted in the absence of other valid means of refutation. Because the side that conceded lost, a concession won by (rhetorical) intimidation was fair play. Libermann (2004, pp. 77–78), relates this telling account: “I once witnessed a debate in which the challenger had the defender justly and fully trapped. At this point a rather large audience intervened to feign great surprise and criticism of the challenger’s attempt to put forward such a supposedly illogical position, thereby emboldening the defender. The challenger lost his confidence and gave up the attack.” It is in the light of such debate- court tactics that we must interpret gestures of derision in Tibetan philosophical writing. Jamyang Shepa’s ‘denigration’ of Taktsang, therefore, need not be seen as inconsistent with respect shown to the substance of his arguments, at least in the privacy of his own quarters. 40 – One might argue that conventional preferences do grant more weight to some conventions than to others. In this case, however, I suspect that testimony would emerge, perhaps together with perception, as the winner. 41 – Hampe 2018, pp. 48–49. 42 – For the use of apratiṣt. hā in Mahāyāna sources, consider especially Vimalakīrtinirdesasūtra, chap. VI, par. 6: “What is the root of errant conceptions?” The root of errant conceptions is groundless (apratiṣt. hā). Matthew T. Kapstein 1043  “What is the root of what’s groundless?” What root can there be, Mañjusrī, of what is groundless? All phenomena are rooted in what is groundless. References Almogi, Orna. 2010. “Māyopamādvayavāda versus Sarvadharmāpratiṣt. hāna- vāda: A Late Indian Subclassification of Madhyamaka and Its Reception in Tibet.” Journal of the International College for Postgraduate Buddhist Studies XIV:134–212. Cabezón, José Ignacio. 1992. A Dose of Emptiness: An Annotated Trans- lation of the Stong thun chen mo of Mkhas-grub Dge-legs-dpal-bzang. Albany: State University of New York Press. Chernev, Irving. 1965. The Most Instructive Games of Chess Ever Played. New York: Simon and Schuster. Cowherds, The. 2011. Moonshadows: Conventional Truth in Buddhist Philosophy. New York: Oxford University Press. ———. 2016. Moonpaths: Ethics and Emptiness. New York: Oxford University Press. D’Amato, Mario, Jay L. Garfield, and Tom J. F. Tillemans, eds. 2009. Pointing at the Moon: Buddhism, Logic, Analytic Philosophy. New York: Oxford University Press. Doctor, Thomas H. 2014. Reason and Experience in Tibetan Buddhism: Mabja Jangchub Tsöndrü and the Traditions of the Middle Way. Abingdon and Oxon: Routledge. Dkon mchog bstan pa rab rgyas, Mdo smad chos ’byung deb ther rgya mtsho. 1982. Beijing. Duckworth, Douglas S., Malcolm David Eckel, Jay L. Garfield, John Powers, Yeshes Thabkhas, and Sonam Thakchoe. 2016. Dignāga’s Investigation of the Percept: A Philosophical Legacy in India and Tibet. New York: Oxford University Press. Edelglass, William, Pierre-Julien Harter, and Sara McClintock, eds. 2022. The Routledge Handbook of Indian Buddhist Philosophy. London: Routledge. Garfield, Jay L., ed. 2019. Wilfrid Sellars and Buddhist Philosophy: Freedom from Foundations. London and New York: Routledge. Goldfield, Ari, Jules Levinson, Jim Scott, and Birgit Scott, trans. 2005. The Moon of Wisdom: Chapter Six of Chandrakirti’s Entering the Middle Way. Ithaca, NY: Snow Lion Publications. 1044 Philosophy East & West Gyatso, Janet. 2015. Being Human in a Buddhist World. New York: Columbia University Press. Hampe, Michael. 2018. What Philosophy Is For. Translated by Michael Winkler. Chicago: University of Chicago Press. Hopkins, Jeffrey. 2003. Maps of the Profound: Jam-yang-shay-ba’s Great Exposition of Buddhist and Non-Buddhist Views on the Nature of Reality. Ithaca, NY: Snow Lion. Hugon, Pascale, and Jonathan Stoltz. 2019. The Roar of a Tibetan Lion: Phya pa Chos kyi seng ge’s Theory of Mind in Philosophical and Historical Perspective. Vienna: Austrian Academy of Sciences Press. Jinpa, Thupten. 2002. Self, Reality and Reason in Tibetan Philosophy: Tsongkhapa’s Quest for the Middle Way. Richmond and New York: RoutledgeCurzon. Kapstein, Matthew. 2021. “The Fragility of Foundations and Frameworks: A Review Essay on Wilfrid Sellars and Buddhist Philosophy.” Journal of Buddhist Ethics 28:167–197. Karmay, Samten Gyaltsen, trans. 2014. The Illusive Play: The Autobiography of the Fifth Dalai Lama. Chicago: Serindia Publications. Libermann, Kenneth. 2004. Dialectical Practice in Tibetan Philosophical Culture. Lanham: Rowman & Littlefield. Mahāyāna-Sūtrālaṃkāra. Edited by Sylvain Lévi. 1907–1911. Paris: Honoré Champion. Nagel, Thomas. 1986. The View from Nowhere. New York: Oxford University Press. Priest, Graham. 2018. The Fifth Corner of Four: An Essay on Buddhist Metaphysics and the Catuṣkot. i. New York: Oxford University Press. Ruegg, David Seyfort. 2002. “Two Prolegomena to Madhyamaka Philoso- phy.” In David Seyfort Ruegg, Studies in Indian and Tibetan Madhya- maka Thought: Part 2, pp. 137–280. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien. Sadhukhan, Sanjit Kumar. 1991. . “Biography of the Eminent Tibetan Scholar ‘Jam-dbyaṅs bshad-pa Nag-dbaṅ brtson-’grus (A.D. 1648–1722).” Tibet Journal 16, no. 2:19–33. Smith, E. Gene. 2005. “Banned Books in the Tibetan Speaking Lands.” 21st Century Tibet Issue. Symposium on Contemporary Tibetan Studies. Collected Papers, pp. 368–381. Taipei: Mongolian and Tibetan Affairs Commission. Matthew T. Kapstein 1045  Stag tshang lo tsā ba shes rab rin chen gyi gsung ’bum pod dang po. 2007. Beijing: Krung go’i bod rig pa dpe skrun khang. Sullivan, Brenton. 2020. Building a Religious Empire: Tibetan Buddhism, Bureaucracy, and the Rise of the Gelukpa. Philadelphia: University of Pennsylvania Press. Tanaka, Koji, Yasuo Deguchi, Jay L. Garfield, and Graham Priest, eds. 2015. The Moon Points Back. New York: Oxford University Press. Tillemanns, Tom J. F. 1999. Scripture, Logic, Language: Essays on Dharma- kīrti and His Tibetan Successors. Boston: Wisdom Publications. Vose, Kevin. 2009. Resurrecting Candrakīrti: Disputes in the Tibetan Creation of Prāsaṅgika. Boston: Wisdom Publications. 1046 Philosophy East & West