Chapter 1 · The Arms Race: Why Buddhism Needed Philosophy¶

Essay 4 closed on a scene: in the same monasteries, under the same Vinaya, lived two groups that condemned each other's writings as heretical. In other religions, such differences would usually lead to excommunication and war; in Buddhism, they had nowhere to go. The Vinaya did not deal with doctrinal disputes (the definition of schism in the saṅgha is procedural, a point already used in Essay 2, Chapter 3 and Essay 4, Chapter 1), the seat of a supreme religious leader stood empty, and there was no final adjudication. Essay 2, Chapter 1 counted three paths for a system without final adjudication: councils, royal authority, and separation. Indian civilization after the Buddha's passing happened to have a standing institution ready for a fourth path: the debate arena. It combined the first three: rules were organized by the participants, royal authority was invited to judge, and losers need not split off; what they lost was reputation, offerings, and followers. Differences now had somewhere to go, and the history of doctrine acquired a new environment of selection. This chapter examines how that environment armed Buddhism and reshaped it.

Before entering, see what Buddhism brought to stake, because an easily missed premise lies here: before one can afford to lose reputation and offerings, one must first have them; before one can fear losing them, one must care. Neither was invented by the debate arena. They were the accumulations described in the preceding essays. On the side of offerings, temporary settlements for the rainy-season retreat became permanent monasteries, and the bowl and three robes became estates and storehouses (Essay 3, Chapter 1). As early as a century after the Buddha's passing, accepting and keeping gold and silver had drawn the entire saṅgha into the dispute over the ten practices (Essay 2, Chapter 3). The withdrawal of support at Kauśāmbī and the purge at Pāṭaliputra had in turn shown the consequences of having offerings cut off and being cast out of orthodoxy (Essay 3, Chapter 2). Once the household had assets, it had something to fear losing. The transformation of reputation ran deeper: attainment could not be checked, and claims carried a cost (the interlude). A reputation that could be checked therefore became the only thing a monk could accumulate, pass on, and exchange for offerings. This is where the economy of visibility described in Essay 3, Chapter 1 came from. In other words, Buddhism did not learn to win and lose only after entering the debate arena. It entered with household assets accumulated over two or three centuries. The greater the assets, the higher the stakes.

Nor did Buddhism open the arena. By the first few centuries of the Common Era, debate across schools was already a mature public institution in India. The Nyāya school made debate itself a discipline. The Nyāya-sūtra divided dialogue into three forms: discussion seeks truth, disputation seeks victory, and destructive criticism refutes without establishing anything. It also classified twenty-two grounds of defeat, with specific provisions for silence and for changing the subject. Once victory and defeat had written rules, there were judges and stakes.¹ The Buddhist records reveal the situation more clearly still. The Yogācārabhūmi divides logic into seven topics. “Venues of Debate” lists suitable places to debate, with the royal court first. “Defeat in Debate” rehearses ways of losing. “Withdrawal from Debate” teaches one to assess three things before battle: each side's skill, the audience's sympathies, and the consequences of victory or defeat; after the assessment, one may choose not to debate. A discipline devoted to spiritual practice solemnly teaching its students how to avoid battle honorably tells us more honestly than any record of victory that the stakes were real.²

How real? Two monastic biographies leave specific figures and penalties. The Biography of the Dharma Master Vasubandhu records a Sāṃkhya teacher defeating a senior Buddhist at the royal court. The king rewarded him with gold, and Buddhism's reputation suffered heavily. Vasubandhu later composed the Seventy Verses on Truth in revenge, received three lakhs of gold, and used them to build three monasteries.³ Xuanzang's biography records a Lokāyata outsider posting forty theses at the gate of Nālandā, staking his head on the outcome. After defeating him, Xuanzang did not take his head but accepted him as an attendant, then soon released him.⁴ Under this volume's rules, such stories are read at half weight: who won, and how splendidly, is halved as victory narrative. But king as judge, gold as prize, head as rhetoric, and conversion as the measure of defeat form an institutional vocabulary consistent across sources that cannot be halved. The debate arena's wins and losses were real: propositions were tested here, schools won offerings here, and reputations were gained and lost here.

Buddhism's first move on entering was to deny the procedures themselves. What Nāgārjuna's Vigrahavyāvartanī questions is not a particular proposition of the opponent's but the very basis for judging victory: you say that all cognition must be established by means of valid cognition; by what, then, are those means established? If they too require such means, an infinite regress follows; if they do not, “everything requires them” refutes itself. Āryadeva's works consist in refutation without establishing a thesis of his own. In the Prāsaṅgika line, “zero stance” became a complete methodology: using only premises you accept, I lead you into contradiction, while I have no thesis of my own for you to refute. Volume I, Essay 5 examined the doctrines' fortunes; here only the tactical form is taken up.⁵ This route is the same intellectual engine analyzed in Essay 4, Chapter 3: prajñā is the most powerful critical tool in polemic, deciding contests at a higher level. The route appears undefended: without a position, there is no position to lose. That statement must be narrowed in line with the precise meaning of having no thesis in Volume I, Part II, Essay 18, Chapter 5: what is absent is only a metaphysical thesis at the object level. At the metalevel there are other things to defend—definitions, logic, and the relational premise that is analytically true. Chapter 10 of the same essay also records this route's own occupational disease: the thing grasped is called having no thesis, the very technique for canceling views. It is harder to detect than holding a view because it looks empty. There is a position; it simply lies outside the object level and is therefore not easily counted by the debate arena's procedures.

But this tool cannot win a victory adjudicated by procedure. A royal court wants a decidable outcome, and adjudication requires procedures both sides share. To the judge, a debater who declares the procedures themselves invalid is hard to distinguish from a loser. Someone within Buddhism soon pointed this out. Bhāvaviveka argued that refuting an opponent required autonomous argument forms: an argument must be established under cognitive procedures both sides accept. Candrakīrti objected that one who holds a view of intrinsic nature and a Mādhyamika do not see the same world; the supposed common ground does not exist. Volume I, Part II, Essay 19, Chapter 2 reconstructs the step more precisely: for an autonomous inference to hold, both parties must accept the basis of the thesis and the property serving as the reason. But in the domain of the ultimate, what the realist accepts is precisely what is under examination, and what the Mādhyamika accepts is precisely what the opponent lacks. There is nowhere to establish common ground. Bhāvaviveka's response is a qualifier: he establishes the thesis with the qualification “in ultimate reality.” A dilemma follows: apply the qualification strictly, and the opponent does not recognize the terms in the thesis; omit it, and what is established is not one's own thesis. The statement of the cost is an especially apt footnote to this section: restriction to a single source closes the gateway to dialogue with the community of logicians, leaving no place for the public testability Bhāvaviveka sought. Chapters 1 and 7 of the same essay also judge that the two sides share more than they dispute. Their lack of doctrinal opposition must therefore be distinguished from the tactical forms described here. Volume I, Essay 5 adjudicated the dispute between autonomous argument and consequence on doctrinal grounds. This volume adds a reading from the history of transformation: it was also a dispute over the direction of armament. One side demanded weapons built by the battlefield's rules; the other refused to recognize that battlefield's sovereignty. In later history, the side that built weapons by the rules prevailed.

In the third round, Buddhism produced the best weapons in the arena. Dignāga established the three characteristics of a reason in inference, making the validity or invalidity of an argument testable against a checklist, point by point, and reduced cognition to two means of valid cognition, perception and inference. Dharmakīrti added precision at every point. For centuries afterward, these two effectively set the epistemological agenda for India's schools. Each generation of opposing thinkers had to answer the Buddhist challenge before establishing its own thesis. Volume I, Essay 8, Chapters 1 and 2 recorded the systems and their influence; this chapter adopts that account directly.⁶ Buddhism took four hundred years to move from denying the procedures to setting the common agenda of the schools. The iron law of the arms race appears in that course: those who refuse weapons withdraw from the battlefield, not from the weapons; to remain in the arena, one must arm oneself by its rules. By the time one is fully armed, one's shape has been given by the enemy.

Round 1 asks how the means of valid cognition themselves are established. It does not merely ask which means establish a cognition: the chapter’s regress challenges the justification of the means. “Zero thesis” is limited to an object-level metaphysical thesis. Chapter 1 explicitly retains commitments at the metalevel to definitions, logic, and the relational premise treated as analytically true. Round 2’s “consequential reductio” refers to the Prāsaṅgika route; the dispute concerns argumentative tactics and does not erase the doctrinal ground the two sides share. Round 3’s two means are perception and inference.

“Setting the agenda” means becoming a leading force in the debate agenda. It does not mean choosing every question independently of opponents: the canopy label and Chapter 1 expressly say that opponents help determine the subject matter. The three ascending bands illustrate the chapter’s three rounds, not quantities on a measured time scale. The final conjecture retains its stated test: early epistemological works organized around practical training, or an agenda shown to have developed independently of opponents, would defeat it.

Once debate bore on the whole monastery's offerings, the standing of scholastic teachers was immediately reordered. Nālandā's roster is the plainest evidence. Xuanzang's biography records ten thousand residents and guests, but its unit of counting is the number of texts, not attainment: more than a thousand could explain twenty sūtras and treatises; more than five hundred could explain thirty; ten could explain fifty, Xuanzang among them; Śīlabhadra alone was said to have mastered the whole. Compare Essay 2: the roll call at the First Council counted those who “were all arhats.” More than a thousand years later, the roster counted how many books one knew. The unit had changed; the interlude's curve found confirmation in a monastery's records. Treatment followed the unit: those with higher text counts received special daily provisions, rode elephants when going out, were exempt from monastic duties, and received teaching seats and rooms according to levels of learning.⁷ What of those who sat in meditation? The roster had no column for the strength of concentration. The standard by which a monastery measures people is the standard by which it rewards them; the standard by which it rewards them is the direction in which the young strive.

Looking from the roster to the people on it brings us to a question that must be answered directly. Essay 4, Chapter 4 counted the population: as late as the seventh century, Mahāyāna remained a minority. Yet the thinkers whose names we know, from Nāgārjuna, Āryadeva, Asaṅga, and Vasubandhu to Dignāga, Dharmakīrti, Bhāvaviveka, and Candrakīrti, belong almost uniformly to the Mahāyāna line. How can a minority's dominance of the list of famous teachers and that population record both be true? The answer has three layers. First, who left the list? This list of famous teachers was transmitted mainly through two routes, Chinese and Tibetan, and what China and Tibet sought was the learning of the lecture platform (who controlled those routes will be examined in the next chapter). The mainstream schools did not lack people. Within the same millennium as Xuanzang, before or after him, Ceylon had Buddhaghosa, and the Saṃmitīyas had their own learned elders; there was simply no one to compile textbooks or biographies for them. The Gupta mentioned in Essay 4, “unrivaled in linguistic learning, esteemed by the king and the people alike,” was a celebrated non-Mahāyāna thinker. We know of him only because Xuanzang passed through. Second, respect depends on the setting. Among villages and ordinary followers, the mainstream schools had their own elders. In the elite debates of royal courts and great monasteries, Mahāyāna thinkers did increasingly prevail, because the arms race's two main weapons—prajñā's refutations and the theory of valid cognition—happened to be their own schools' learning. The third layer is deepest: these people carried two identities. According to the biography, the brothers Asaṅga and Vasubandhu entered the monastic life and received ordination in the Sarvāstivāda. Yijing lived and studied at Nālandā for ten years, where the Vinaya in use was the Mūlasarvāstivāda Vinaya. Yet he says, “Those who venerate bodhisattvas and read Mahāyāna sūtras are called followers of the Great Vehicle.” The distinction between great and small was an individual's orientation in study and practice, not a monastery's affiliation (cited in Essay 4, Chapter 4).⁸ In other words, the identity of a “Mahāyāna thinker” belonged to a school; only the lecture platform was Mahāyāna. Before asking whether these thinkers were respected, ask on which roster. On the Vinaya roster, they were monks properly ordained in a particular school, no different from those seated beside them. On the debate arena's roster, they were renowned masters. One person, two rosters: the paired positions established in Essay 2—rules and chain—embodied in living people.

The four questions are tallied as usual. What changed? The environment in which doctrines were produced and selected. Whether a proposition survived depended increasingly on its defensibility in debate: definitions had to withstand the list of grounds of defeat, argument forms had to pass the three characteristics of a reason, and positions had to hold under reductio. What forces drove it? The strand of competing religions set the agenda; royal authority supplied judges and prizes; the social composition of the community provided matching promotion within monasteries. The ladder for learned monks described in Essay 3, Chapter 1 thus led outward to the debate arena. Who benefited? The class of scholastic teachers and the monastery's reputation were the visible beneficiaries. Another must be recorded fairly: philosophy itself. This contest produced the most precise epistemology and logic in the history of Indian thought. Many tools with which this book's first two volumes analyze Buddhism took shape in those centuries of debate; critics of this arms race are also its beneficiaries. The contest also truly protected the order: winning debates meant temple property, endowed villages, and donors who converted. In a world where kings allocated resources by wins and losses, remaining unarmed meant waiting to be slaughtered.

The cost fell on the agenda. As required, an explicit label: what follows is this volume's conjecture. Half the precision of Buddhist philosophy was shaped by its enemies. Opponents decided which questions counted as questions; the agenda decided what counted as important; and the cessation of suffering was not on the agenda. For circumstantial evidence, look at the contents of works on valid cognition: particulars and universals, how exclusion yields reference, how momentariness is established, how another mind can be inferred. Each is crucial in external debate; none takes cessation as its organizing principle. Buddhism's most treasured experience of practice later entered the theory of valid cognition too, but the way it was filed reveals who controlled the agenda. When Dharmakīrti treated yogic perception, he asked whether it counted as valid cognition and under what conditions—its legitimacy as evidence, not its operating procedures. In its own epistemology, practice was the object under examination, not the standard of examination. If the earliest works on valid cognition are organized around procedures of practice, or if their list of questions can be shown to have arisen independently of opponents' agendas, this conjecture fails.

Armament needs an arsenal; a standing army needs pay. The next chapter turns to Nālandā.

References¶