Chapter 3 · The Cost of Debate: The Drift in Criteria¶

Chapter 6 of the latter part of Volume II's essay on Madhyamaka reached a judgment; first, set it out unchanged: as the tradition developed, the criteria of liberation underwent an identifiable drift. The experiential criterion of “the cessation of suffering” was gradually displaced by a set of epistemological standards: whether one could speak correctly about emptiness, demolish an opponent in debate, give a sharp Chan-style reply, or write a doctrinally precise commentary. The drift arose through a gradual structural process, not anyone's decision, and was inseparable from the institutions' transformation into academies.¹ That judgment was a philosophical analysis, concerned with logic. It still lacked a history: in which institutions did this substitution occur, at what pace, and through whose hands? The preceding two chapters have in fact supplied the conditions. This chapter puts them together.

Four conditions, each with its source. First, the vacuum of certification: attainment could not be adjudicated; there was no certifying institution, and therefore no institution for disproving claims (the interlude). A vacuum does not remain empty forever; anything open to public adjudication will be drawn in to fill it. Second, the economy of visibility: institutions and donors naturally prefer demonstrable achievements; the meditator loses to the lecturer (Essay 3, Chapter 1). Third, the debate arena: it happens to supply what the vacuum of certification most lacks—something public, adjudicable, equipped with judges and records of wins and losses (Chapter 1 of this essay). Fourth, the academic system: it institutionalizes debate results in entrance examinations, ranks, degrees, and promotion (Chapter 2). Once the four come together, replacing the criterion requires no resolution. No council announces “From now on, eloquence in debate is realization.” There are only a thousand small, concrete evaluations, each asking who should receive this teaching seat or whether this newcomer should be kept, and each with only one kind of usable evidence: what can be tested. The corpus ratchet described in Essay 4, Chapter 1 let scriptures in and never out; promotion here also moved in only one direction: those skilled in debate rise and do not fall for lacking skill in practice; those lacking skill in debate do not rise on the strength of their skill in practice.

The four conditions converge on one rising arrow. “Sitter” means meditator; the debate arena supplies publicly adjudicable wins and losses. “One direction only” expresses movement without reversal in promotion: those skilled in debate rise and do not fall for lacking skill in practice, while those lacking skill in debate do not rise on the strength of their practice. The claim concerns what the institution rewards, not merely whether someone chooses to practice.

The tool, independent, and institutional phases are three stages in the chapter’s account. The three scoreboards below them—debate, population, and survival—are separate comparisons, not three successive eras or criteria assigned one to each phase. The Gelug station must be read with Chapter 3’s explicit qualification: Gelug institutions preserved and transformed the mechanism of awarding status for scholastic performance; they are not an unchanged Nālandā system or a direct record of ancient Indian practice. The survival column refers to the chapter’s hereditary household and social ties, not a genetic measure of religious survival.

The final “thousand promotion reviews” is the author’s conjectural description of cumulative institutional decisions, not a newly counted dataset. The chapter retains the religious core within epistemology, continuing practical traditions, and internal protests as counterexamples to a total replacement of practice. Its conjecture would be weakened by a contemporary diagnosis of the drift followed by institutional action that shifted the criterion back.

The pace can be marked in three stages. First, the instrumental phase: in the Abhidharma movement, debate was still only a means of settling disputes and preserving doctrine. The criterion nominally remained cessation, but “what can be tested” was already firmly established within institutions (Essay 3, Chapter 1). Second, the phase of independence: in the age of the theory of valid cognition, establishing and refuting became an independent discipline, and debating skill a public identity. At the same time, under the exemption for overestimation of attainment, claims to attainment retreated into a private matter that could never be tested (Essay 4, Chapter 2). As one rose and the other fell, the socially recognized criterion had already changed. Third, the institutional phase: the degree system fixed the new criterion as a sequence of promotions. From the gatekeeper's questions to victory at the royal court, every promotion in a learned monk's life was decided by epistemological performance (Chapter 2). The institutions of the third stage continue today. Tibetan Gelug monasteries reorganized the Indian curriculum of treatises into a long course of teaching, memorization, debate, and examinations, with the five great treatises as the framework of geshe training. This is not an exact copy of the Nālandā system, but it preserves and transforms the mechanism of awarding status through scholastic performance.² After the institution disappeared in India, the criterion continued within another organization. Today's debate colleges can therefore provide comparative material, but cannot be taken directly as a record of ancient Indian reality.

Now look outside the arena, for that is precisely where debate's greatest cost fell. The arena had its own scoreboard, on which Buddhism led for a long time: the leading force in the theory of valid cognition, setter of the agenda, repeated victor at the royal court. But there was a second scoreboard: population. Essay 4, Chapter 4 noted that in the seventh century Mahāyāna was still a minority in India; here look at the opponents outside the arena. What actually won the people of South India was the devotional wave beginning in the sixth and seventh centuries, unrelated to any particular thinker. The Āḻvārs and Nāyaṉmārs sang in Tamil, went on pilgrimage, and wept bitterly, bringing the deity directly into villages and markets. They did not enter the debate arena or accept the list of grounds of defeat, and so could not be defeated.³ Volume I, Essay 8, Chapter 6 has already settled the question of Śaṅkara: biographies of his complete victory in debate come from literature produced by his followers centuries after his death. His dating rests entirely on his references to the Mīmāṃsaka Kumārila and the Buddhist Dharmakīrti, while no Buddhist citation of him has been found. His school did not become uniquely supreme in one stroke because of him; the order's expansion was tied to the monasteries he established in the four directions and the Daśanāmi order. Brahmanical orthodoxy even mocked him as a “crypto-Buddhist,” showing how blurred the boundary between the two was.⁴ This volume adopts that chapter's conclusion, adding only one sentence about mechanism: displacement through absorption, described in the introduction's Chapter 2, does not need to win debates at all. It goes around them. Absorb your philosophy, restate it in the name of the Upaniṣads, then count your Buddha among the avatars: no step takes place inside the debate arena. Buddhism had prepared for debate as fully as it could, yet the opponents it met were of two other kinds: one that sang instead of debating, and one that burned instead of debating (Essay 8). Seven hundred years of accumulated skill in debate were of no use against either.

There is a third scoreboard, whose results take several centuries to appear: what became of the schools themselves. Go through the debate arena's old opponents one by one, and survival turns out to have almost nothing to do with debating results. Sāṃkhya, the school in the royal-court story in Chapter 1, ceased as an independent school after roughly the tenth century. Yet its concepts—prakṛti and puruṣa, the three guṇas, and the twenty-five principles—became vocabulary shared across India. Purāṇic cosmology, Yoga, Tantra, and Vedānta's world schemes all used them. The Sāṃkhya-sūtra, emerging in the fourteenth or fifteenth century, was a textual revival claiming ancient authority, not a revival of the school. The school disappeared; its vocabulary remained.⁵ Vaiśeṣika merged into Nyāya. Nyāya lived longest: after Buddhist theories of valid cognition died out, Navya-Nyāya took over the list of questions and continued working through it. What the debates had accumulated went to the school that survived. Mīmāṃsā gradually retreated to an auxiliary discipline of Vedānta. Vedānta joined with bhakti to become the central axis of orthodoxy. The lines of Rāmānuja and Madhva built their schools on temples and their temples on caste communities. They learned Buddhist forms of organizing an order while also having the bloodline Buddhism lacked. Jainism has survived to this day by relying on precisely the layer Buddhism lacked: endogamous communities of merchant castes. Its laypeople have their own identity and their own law (the comparison drawn in Essay 4, Chapter 3 is borne out in the matter of survival). Lokāyata had no institution and lives only in its opponents' quotations. One sentence sums up this scoreboard: those grounded in households, caste, and bloodline survived; those grounded only in institutions and debating results disappeared, or saw their concepts absorbed by others. The school that debated best disappeared. Among those who lost debates and those who did not debate, some have thrived. Debating results do not tell us who will survive.

The costs can now be listed one by one. First, the replacement of the criterion itself: truth's location moved from the meditation room to the teaching seat. Volume II has judged this; the present chapter supplies its mechanism. Second, where talent went: for centuries, the tradition's brightest minds devoted their lives to refining arguments. The opportunity cost cannot be measured, but its direction can be identified. Had those same talents been devoted to transmitting and testing methods of practice, the interlude's concentration curve might not have taken that course. This is a counterfactual offered as conjecture: it indicates a direction, not an inevitability. Third, an agenda set by opponents, recorded in Chapter 1 of this essay. Fourth, the blind spot: an institution that defines winning as victory in debate cannot see opponents who never enter the arena. When the devotional wave arose, the institutions that most needed to be alert were busy celebrating another complete victory at the royal court.

Under this volume's rules, counterexamples come first, lest the judgment become an unfalsifiable accusation. First, the theory of valid cognition has a real religious core. In the chapter on establishing valid cognition of his Pramāṇavārttika, Dharmakīrti establishes the Blessed One as an authoritative person through great compassion cultivated over many lives. In his hands, the whole epistemology ultimately serves to argue that “the Buddha can be trusted.” Reading the entire theory as a product of the arms race erases precisely this layer.⁶ Second, the replacement of the criterion never became total. In the traditions where institutional debate flourished most, lineages of actual practice always continued alongside it (the Tibetan forms will be examined in Essay 12). Drift means a shift in the center of gravity, not the extinction of practice. Third, voices within the tradition protested. Songs of realization by accomplished practitioners made mocking scripture-reciting paṇḍitas into a genre. Saraha sang most plainly: “The paṇḍitas expound all the treatises, yet do not know that the Buddha dwells within the body.”⁷ The fate of these protests, however, confirms by contrast the position of the mainstream criterion: the protesters were not treated as grounds for correction, only assigned to a separate path of practice. That path is the subject of Essay 7.

Finally, as required, an explicit label: what follows is this volume's conjecture and this essay's conclusion. No one decided to abandon the criterion of cessation. It was silently replaced in a thousand promotion reviews. Every review was locally reasonable; every reviewer might have been sincere. The substitution could not be seen in any single review, appearing only across centuries, beyond any participant's lifetime. If the tradition explicitly identified this drift at the time and took institutional measures to shift the criterion back, this conjecture carries less weight. Volume II's philosophical judgment and this chapter's historical mechanism now meet. The costs this essay has recorded—replacement of the criterion, concentration in a few great monasteries, an agenda set by opponents—are also the pressures the following essays take up. The nearest pressure can already be named: the debate arena honed the absence of intrinsic nature into the most powerful critical tool, able to refute any position that asserts an intrinsic nature. Its scope is bounded as described in Essay 4, Chapter 3: it has no purchase on opponents who make no such assertion. This also explains why those who sang instead of debating were untouched by it; the devotional wave simply lay outside its domain of propositions. Yet many believers wanted precisely something that could not be refuted and could be relied on. The more powerful the tool, the stronger this need became. The next essay examines the return of true permanence.

A cross-section of theory, recorded as a closing note to the volume. The form the debate arena gave doctrine was epistemology. The theory of valid cognition became the common language: perception and inference as the two means of valid cognition, the three characteristics of a reason in logic. For the first time, Buddhist thought could defend itself in forms of argument shared by all the schools. The vacuum shifted too. Dharmakīrti established the Blessed One as an authoritative person, proving that the Buddha could be trusted, not that buddhahood could be reached. In theory, victory or defeat in debate has no connection with the presence or absence of liberation. The fissure between the two truths widened at the same time: the ultimate cannot be spoken, while the debate arena's entire livelihood lies in speech. Those who spoke of the unspeakable won the contest, yet the contest itself belonged to the speakable side. The main text recorded the drift in criteria; this note adds only that the drift had an accomplice within theory. The wording of that complicity must avoid a reading already ruled out. Placing the ultimate outside language amounts to reading the two truths as two domains. Volume I, Part II, Essay 18, Chapter 6 judges that the two truths distinguish levels of assertion, not two worlds, and cites Qingmu's commentary: the ultimate meaning always depends on speech, and speech is conventional. It then concludes that the metalevel is expressed through conventional language; this is by design, not a flaw. Essay 19, Chapter 6 is more explicit still: whenever the two truths are made into two domains, all residents of the second domain fall within the scope of application. The conclusion this section needs does not depend on dividing domains. Winning and losing in debate occur at the level of speech; whether cessation occurs is not determined by victory or defeat at that level. However great the victory or defeat, it cannot reach what judges it.

References¶