Interlude · Innumerable Saints: An Impossible Census¶

The fragmentation has been recounted, and the scholastic institution has yet to open its doors. Between the two essays, this volume pauses to ask a question both sides skirt: after the Buddha's passing, how many arhats remained in the world? In circles of practice, no one wants to touch it; doing so damages faith. In academia, no one can touch it; there is no evidence to examine. Noble attainment leaves no physical evidence, and the saṅgha had no population register. The “number of living arhats” at any moment cannot be verified. But the inability to answer can itself be studied. The tradition left five kinds of indirect evidence that can be roughly dated: numbers it reported, timetables it promulgated, legends of the “last one” and doctrines of “remaining in the world,” circumstantial evidence from inscriptions and travel accounts, and indirect readings from doctrine and literary form. Arrange these five kinds of evidence on a timeline, and what emerges is how Buddhism repeatedly revised its understanding of its own stock of saints, not the population of arhats.

Begin with the reported numbers. The formula opening discourses from the Buddha's lifetime is “together with twelve hundred and fifty, all arhats”: monk and arhat are almost synonyms. The First Council had five hundred participants, with arhatship as the qualification for entry; the scale of the meeting and the scale of the arhat population were of the same order. At the Second Council, the seven hundred were still regarded by the tradition as a noble assembly, but the focus had moved from attainment to the Vinaya. By the Third Council in the Southern narrative, the terms had changed: from sixty thousand monks in Pāṭaliputra, one thousand were selected to compile the canon. All these figures are formulaic—five hundred, seven hundred, one thousand—and cannot be read as population counts. But the shape of the series conveys information: even on the tradition's own reported numbers, “monk equals saint” retreats from an equation to a ratio of one selected in sixty. Nor need output have collapsed: Aśoka's patronage need only have enlarged the denominator (the Southern tradition itself records “infiltrators” who entered the order for food and clothing) for the concentration to collapse. The Northern tradition places the finale in Kaniṣka's era: the compilation of the Mahāvibhāṣā still claims five hundred arhats. The original line below the opening title reads “composed by five hundred great arhats and others.” Volume I, Part II, Essay 16, Chapter 1 also points out that this line compresses the council legend into an authorship attribution, judging that it is therefore a collective ruling and that the four Chinese characters “毗婆沙师” (“Vaibhāṣika teachers”) henceforth became an identity. That was the last echo of narratives of assemblies of living arhats. Thereafter, the protagonists of transmission stories became treatise authors, and arhats left the biographies of contemporary figures.

The second kind of evidence deserves closer scrutiny: the tradition promulgated a timetable for the decline of attainment. The earliest was the prophecy that admitting women to ordination would halve the duration: the true Dharma's thousand years would thereby be cut in half. The commentarial age completed the table. The Shanjian lü piposha records ten thousand years of the teaching's presence in the world, with attainment downgraded every thousand years: the first millennium produces arhats endowed with the unobstructed knowledges; the second, only arhats whose craving is exhausted; the following millennia descend through nonreturners, once-returners, and stream-enterers. In the last five thousand years there is study but no attainment, and after ten thousand years only the outward form of shaven heads and dyed robes remains.¹ What this timetable does must be made explicit: it makes “there are no arhats around us” require no further explanation. The calendar takes responsibility on everyone's behalf. The power of explanation drifts from “we have not practiced well” to “the age has run its course.” This is a substantial juncture of transformation: the absence of attainment becomes normal and no one's responsibility. The narrowing logic of “only this method is suitable in this age”—the role of the final age of the Dharma in Pure Land Buddhism and in Kamakura—rests on this ground.

The third kind of evidence consists of a vote and a set of saints who remain in the world. In the first century BCE, a great famine nearly severed Sri Lanka's oral transmission. After the disaster, the commentaries record a vote on direction: which was the foundation of the teaching, practice of the path or textual learning? The conclusion went to textual learning: “Even if a hundred thousand monks practice insight, without textual learning not one can attain the noble path.”² This had the same timing and cause as the writing of the Tripiṭaka at Aluvihāra. The saṅgha formally recast its mission from producing saints to preserving texts; from then on, the island tradition rested its orthodoxy on texts and the Vinaya. Essay 3 will examine it. Meanwhile, as living arhats gradually became unavailable, doctrine invented deathless arhats: Mahākāśyapa enters meditative absorption at Kukkuṭapāda, holding the robe and waiting for Maitreya; Piṇḍola is commanded not to pass away and continually receives offerings in the human world; the Record of the Dharma's Abiding establishes sixteen great arhats who remain in the world to protect the Dharma until the next buddha.³ The ingenuity of the doctrine of saints remaining in the world is that it never needs to be verified: sixteen ever-present saints make a list that is valid forever. After the Tang, images of arhats became widespread. The arhat shifted from a population category to an image category. To make offerings to a category with no living bodies left, images are the only medium.

The fourth kind is external circumstantial evidence, but its evidentiary weight must be kept low. The short inscriptions on the reliquaries from Sanchi Stupa 3 say only “of Śāriputra” and “of the great Maudgalyāyana,” referring to their relics; they do not inscribe the title “arhat.” They establish a relic cult for two deceased disciples, not that “the title of saint belonged only to the dead.” Contemporary donative inscriptions often refer to monks by functions such as reciter, Vinaya specialist, or learned one. But the absence of a particular title cannot conversely prove that living people lacked attainment. Records by Chinese monks seeking the Dharma show, at most, a shift in literary emphasis. Chinese materials from around the fifth century still describe Jibin as a land of meditation and lineages of saints. Xuanzang's arhats are mostly associated with ancient sites and legends, while eminent monks of his own time commonly excel in scholastic learning. By Yijing's time, attainment also rarely serves as a category of travel observation.⁴ These are circumstantial indications of how narratives are framed, not statistics on the stock of arhats.

The fifth kind of evidence comes from doctrine and literary form themselves. Mahādeva's five points are a contemporary complaint about the gap between actual and ideal arhats (examined in Chapter 3). The existence of the complaint demonstrates that there were still enough recognized arhats to question. Later generations no longer needed to question them, because there was nowhere to direct the challenge. Mahāyāna sūtras cast Śāriputra as a supporting character receiving instruction: the literary space for mocking an elite presupposes a loosening of that elite's actual prestige (this sentence is tentative). Abhidharma and stages-of-the-path literature chart the levels in ever finer detail. The causal sequence requires attention: subdivision was already under way in the four Nikāyas. The three discourses on trainees divide stream-enterers into those returning seven times, those going from family to family, and “one-seeders,” and nonreturners into five kinds beginning with those who attain nirvāṇa in between. Volume I, Part II, Essay 13, Chapter 6 judges this the fifth case of expanding and contracting lists. The primary impetus is mnemonic technique and cataloguing; the mechanism is pressure toward fixed form through repeated use. Categories used long enough cease to be containers and become structures. A shrinking territory can at most hasten the later stage; it cannot count as the cause. But the fifth kind of evidence also has another side that this chapter must state explicitly: the cost of claiming attainment is rising. Claiming what one has not attained is a pārājika, and after the five points such claims face closer scrutiny. Less reporting does not mean less production. Half of the “decline” in the stock may reflect tighter certification and reporting criteria. The distinction between the level of claims and the functional level applies here unchanged. This chapter counts discourse; it dare not pass itself off as a population census.

Now, as required, an explicit label: what follows is this volume's conjecture, a concentration curve in three stages. The first, from the Buddha's passing to a century later: the first generation of direct pupils dies out, producing the first precipitous drop. The Chinese term for śrāvaka, shengwen, literally means one who hears the Buddha's voice. Once the generation that heard him personally is dead, the word's original meaning is left suspended. Of the three stages, only this one has biological certainty. The second, from a century after his passing to Aśoka, must be divided into three propositions of different strengths. A collapse in the proportion of arhats is almost certain: no change in output need be assumed; a surge in the denominator under royal patronage is enough. A decline in the total, by contrast, lacks evidence: traditional narratives still produce arhats abundantly in Aśoka's time, and leaders of the various missions are often recorded as arhats. A collapse in total numbers becomes the dominant narrative only later. What is truly settled is the reckoning at the level of discourse: the controversy over the five points brings the gap between ideal and actual arhats into the open, and assembly narratives change from “all arhats” to “one thousand selected from sixty thousand.” The third, from Aśoka to around the beginning of the Common Era: there is no supreme religious leader, the chain governs status but not attainment, and certification fails as verification drifts. The tradition changes its terms to save itself: it promulgates timetables, establishes the doctrine of saints remaining in the world, and revises its mission. Thereafter, the stock of arhats leaves the realm of what can be discussed and becomes calendar, image, and legend. If inscriptions or reliable early documents yield evidence, beyond formulaic narratives, that living arhats were still routinely certified in the first several centuries CE, the second and third stages must be rewritten.

“Names” means a roll call. In the second column, “their proportion” is the proportion of arhats among monks, not the proportion of the Buddha’s direct disciples. The three qualifications—certain, almost certain, and later narrative—retain the different strengths assigned in the chapter.

The Chinese curve rises slightly at the start before declining; the English curve declines from the start. Both are schematic illustrations of changes in discourse, without a statistical scale. They are not a census or a measured series of arhat numbers; the chapter distinguishes falling proportion from an unestablished decline in total numbers.

This question has a modern echo that usefully marks the far end of the curve. In the seventeenth and eighteenth centuries, during interruptions in Sri Lanka's ordination lineage, monks judged that “the paths and fruits can no longer be attained in the present age.” Volume I recorded this crisis of attainment. The meditation revivals in Burma and Thailand in the nineteenth and twentieth centuries went in the opposite direction, explicitly declaring that attainment remained possible in this life. In effect, they publicly tore up the ten-thousand-year timetable, and the roll call of living arhats began again: devotees wrote biographies retrospectively recognizing Ajahn Mun and Mahā Bua, the Sun Lun tradition gave dates for its founder's attainments, and some European and American practitioners even printed “arhat” on the covers of their own books.⁵ The boundary of this resurrection deserves attention: it occurred almost entirely in the Southern Buddhist world that held that timetable. Modern Chinese and Tibetan discourse on attainment uses the categories of awakening, emanations, and accomplished practitioners; “living arhat” is not among its titles for attainment. A finer feature of the present situation must also be recorded: today's Southern Buddhist world generally believes that saints of the first and even the second fruit still live among us, yet has developed a collective silence about who has attained and who has not. This silence has two sides. Its practical side: it does place a protective covering around the saṅgha's supramundane practice. When attainment is not publicly declared, fame and gain have one fewer route in. The Vinaya already prohibits monks from telling laypeople of their own attainments; the intent of that ancient legislation is still at work today. Its doubtful side: silence also breeds doubt, and devotees cannot escape the question: is this actually true? A stock reported only as a belief, with no instances supplied, is less guarded than suspended. Where discourse about the stock dies and revives traces precisely the territory in which “arhat” remains a living category.

The interlude ends here. This census cannot count heads; it has counted three changes in terms: from roll call, to calendar, to images. Two successors now divide the question of the stock between them: the scholastic institution draws its map ever more finely, as the next essay will show; the bodhisattva replaces the ideal altogether, as Essay 4 will show.

A cross-section of theory, recorded as a closing note to the volume. This essay's main text counts authority and rosters; this note adds the shape of doctrine itself at this moment: it has no shape yet. The Buddha left words spoken for particular occasions, not books written as a system. There are a thousand dialogues and not one outline. There are three vacuums, each later to become a major matter. What is a buddha: human or superhuman? No one yet needs to answer, because people who remember him are still alive. How can nonself and karma be compatible, and who receives the result? At this point analogies sustain the account; no one asks about the mechanism. Who determines realization? While the Buddha lives, the Buddha does. After he is gone, as the main text records, this becomes the first vacuum. The starting point of the history of theory is not a system but three questions no one yet considers questions.

References¶