Chapter 1. Transmission to Sri Lanka and the conflict between the Mahāvihāra and Abhayagiri¶

From Essay II through Essay VIII, the discussion has followed developments within India. This part turns to Southern Buddhism, the tradition that takes the Pali scriptures as its canon and the Theravāda Vinaya as its institutional foundation. Its geographical center was Sri Lanka, from which it later spread to Myanmar, Thailand, Cambodia, and Laos.

This chapter begins with Buddhism's arrival in Sri Lanka and the prolonged internal dispute that followed. Sri Lanka was both the principal place of preservation of the Pali Tipiṭaka and the setting for a long schism within the Theravāda monastic tradition. Two monasteries facing each other across the city, the Mahāvihāra and Abhayagiri Vihāra, remained opposed for nearly a thousand years.¹

I. The historical sources are useful, but their provenance must be clear¶

An account of early Sri Lankan Buddhist history must draw on two Pali chronicles: the Dīpavaṃsa (Chronicle of the Island, written around the fourth century CE) and the Mahāvaṃsa (Great Chronicle, compiled by Mahānāma around the fifth to sixth centuries).² They provide the most detailed surviving continuous narrative of this history, and supply most of this chapter's chronological framework.

Both, however, were written within the Mahāvihāra tradition. They date from the fiercest period of rivalry between the two monasteries, and one of their purposes was to support the Mahāvihāra's claim to orthodoxy. In the Mahāvihāra's accounts, Abhayagiri is always the heterodox, lax party, contaminated by non-Buddhist teachings. Accounts of Abhayagiri tend toward interpretations favorable to the Mahāvihāra. The bias is not conjectural. The very organization of the Dīpavaṃsa's narrative serves the Mahāvihāra's lineage of legitimate transmission.

This book treats the two chronicles as important sources with a position to defend, rather than as neutral records. Accounts that can be corroborated through archaeology, inscriptions, and external texts are more credible. The external texts include, in particular, Chinese travel accounts such as Faxian's Record of Buddhist Kingdoms and Yijing's Record of Buddhist Practices Sent Home from the Southern Sea. Where a narrative appears only in the chronicles and plainly favors the Mahāvihāra, this book identifies its source and does not readily accept it as historical fact.

II. The arrival of Mahinda and Vibhajjavāda¶

The traditional narrative credits Mahinda (his Pali name), said to be Aśoka's son, with bringing Buddhism to Sri Lanka. The Mahāvaṃsa's version of the story is thoroughly mythologized. Its marvels include light appearing in the sky, the king encountering Mahinda on a mountaintop, and Mahinda testing the king's understanding with the famous questions about a mango tree. This book sets these marvels aside.

Once the marvels are set aside, the outline that can safely be stated is as follows:

• Around the middle of the third century BCE, during Aśoka's reign, monks who identified themselves with Vibhajjavāda (Pali) reached Sri Lanka. This designation meant that judgments must distinguish among the characteristics of dharmas, in contrast to the Sarvāstivāda's blanket assertion that all dharmas of the three times really exist, introduced in 2.4. Vibhajjavāda should not be treated as an identifiable institutional school. After comparing the accounts, Geiger judged it to be "a philosophical tendency rather than a particular school." The Theravādins associated it with their own conception of orthodoxy. Different traditions sometimes identify it with the Sthaviras, sometimes place it within the Sarvāstivāda, and sometimes say that it embraces the four schools of the Mahīśāsaka, Kāśyapīya, Dharmaguptaka, and Tambapaṇṇiya.² After entering Sri Lanka, this tradition came to be generally known as Theravāda, Pali for "the teaching of the elders," rendered in Chinese as Shangzuobu.
• Under the patronage of the king, traditionally identified as Devānampiyatissa, they established the institutional foundations of monasteries and the saṅgha in the capital, Anurādhapura.
• The Mahāvihāra was founded in a royal park south of the royal city. It became Sri Lanka's earliest principal monastery and subsequently its most authoritative.

This book questions the claim that Mahinda was Aśoka's biological son. Stories of royal children transmitting the Dharma are extremely common in accounts of missionary activity in India and Southeast Asia. Their function is to give that activity the highest royal authority. Aśoka's rock inscriptions do repeatedly mention sending emissaries to neighboring regions to propagate the Dharma, but name neither Mahinda nor Sri Lanka. The inference drawn here is that Mahinda was a missionary monk of Aśoka's time whose precise identity and relationship to the royal family cannot presently be verified from external sources. A mission reaching Sri Lanka in the third century BCE is geographically plausible. Mahinda probably existed, but the mythologizing of his identity is a separate matter.

Tradition says that Saṅghamittā also came, bringing a branch said to derive from the tree beneath which the Buddha awakened. It was planted near the Mahāvihāra. This tree still stands at Anurādhapura, has a documented history, and is Sri Lankan Buddhism's most important sacred object.³

III. Committing the Pali Tipiṭaka to writing¶

For roughly two centuries after Buddhism's arrival, the Sri Lankan saṅgha remained broadly united. As the central monastery supported by the royal family, the Mahāvihāra accumulated an extensive oral transmission and tradition of learning.

In the first century BCE, Sri Lanka suffered a foreign invasion. King Vaṭṭagāmaṇī Abhaya was defeated, spent fourteen years in exile, and then regained power. Chapter 33 of the Mahāvaṃsa recounts these events in detail. The same chapter records the writing down of the Tipiṭaka, but does not make a causal connection between the two events. The Mahāvaṃsa and the Dīpavaṃsa give the same reason. Seeing "the decline of living beings," the monks assembled and committed the texts to books "so that the true Dhamma might endure." Neither chronicle mentions famine in this passage. The entire Mahāvaṃsa refers to hunger or famine only five times, none in this king's reign. Nor does either passage mention fear that all the reciters would die. Both explanations come from later commentaries and modern retellings.²⁴

The discussion in 1.4 covered the resilience of oral transmission and its fatal vulnerability. Oral transmission depends on reciters remaining alive, and once a generation of reciters has been lost, the transmission cannot be recovered. Writing did remove that vulnerability, but the two chronicles do not describe this as the motive for writing the canon down. The connection concerns the consequences, not a causal explanation recorded in the sources. This book's position is that the Pali Tipiṭaka took its final written form in Sri Lanka around the first century BCE, a decisive moment in the history of Southern Buddhist transmission.

Fixing the oral transmission in writing preserved the scriptures, but also froze one particular version as the authoritative text. Subsequent interpretation, additions, and deletions had to proceed on the basis of that fixed text. This provided the foundation for the Mahāvihāra commentarial tradition. The written canon became both Southern Buddhism's inheritance and this book's principal source for discussing early teachings. The next section first clarifies its status as historical evidence.

III.2. What can this canon establish?¶

Writing down the canon and claiming authority for it must be separated. The Mahāvihāra committed the Tipiṭaka to writing and fixed an unalterable list of texts. It also left an account claiming that this list alone constituted the Buddha's word. The first is an ascertainable historical fact, while the second is the participants' account of their own work. Accepting the first does not require accepting the second. Collins's lecture on the subject disentangles precisely this association. He examines three Pali terms that might correspond to "canon." Pāli originally means a line or path and, by extension, a text. When contrasted with commentary, it distinguishes the exact wording from its interpretation and implies no closure. Tipiṭaka first appears in first-century CE inscriptions, where it denotes three literary genres and three specializations among reciters, rather than three bibliographies. The discourses themselves already distinguish specialists in Vinaya, discourses, and doctrinal matrices. Only buddhavacana approaches the sense of canonical authority, but the term is ambiguous. It can mean the historical Gotama's own spoken words, or that whatever is well spoken is spoken by the Buddha. The element of exclusive closure was added by the Mahāvihāra during its struggle with Abhayagiri to become the sole legitimate custodian.⁵

The argument concerns the boundary, not the contents. It establishes only that not everything within the list is early, and not everything outside it is late. Collins explicitly states that arguments dating texts from their contents remain valid, giving the discourses' lack of knowledge of large polities like Aśoka's as an example. He labels his own conclusion a hypothesis, calling for empirical investigation case by case. Rejecting the equation "Pali canon = early Buddhism" is therefore distinct from judging that the Pali Nikāyas are the closest surviving texts to early teachings. The collapse of the former does not bring down the latter.

This book's position, and what supports it. This book does treat the Pali Nikāyas as among the surviving texts closest to early teachings. That priority rests on their state of preservation, however, not their canonicity. The four Nikāyas are the only complete set of recensions transmitted by a single school and preserved throughout in an Indic language. The Chinese Āgamas are translations, and the Saṃyukta Āgama has the longstanding problem of disordered fascicles. The Sanskrit texts survive only in fragmentary leaves from Turfan and Gilgit. Gāndhārī texts survive in scattered scrolls, while Tibetan translations survive mostly in quotations. This is an accident of preservation, not a grant of authority. Once the basis is changed, the Mahāvihāra's account of itself ceases to be a premise of this book, and the association dismantled above does not bear on the use made of these texts here.

The replacement criterion is stricter than canonicity. Section VI of 2.2 established a criterion based on independent transmission of the same teachings by several schools that separated long ago. Material consistently preserved across their recensions very probably derives from a common source predating the division.⁶ Its character needs to be made explicit here. It is external and testable, and does not depend on any tradition's claims about its own boundaries. Its threshold is higher than canonicity, not lower. Under the criterion of canonicity, the whole Pali Tipiṭaka has equal standing. Under corroboration across recensions, the Jātaka, Apadāna, Buddhavaṃsa, and most of the Khuddaka Nikāya cannot enter the highest category. Passages within the four Nikāyas that lack Āgama parallels cannot enter it either. Abandoning the direct equation changes the criterion from "Who recognizes it?" to "How many independent traditions preserve it?" It also raises the threshold.

The argument provides no basis for accepting unofficial histories. The danger lies in the invalid inference that if the canon was constructed, then all constructions are equivalent. Collins explicitly aligns himself with Schopen, whose project is to turn to external evidence such as inscriptions and archaeology for adjudication, not to declare adjudication impossible. Under corroboration across recensions, a late text attributed to an earlier figure is in a worse position than under a canonical criterion, because it cannot produce a single independent recension. The documentary examination of the Four Alternatives in Chapter 3 of Essay XI follows precisely this approach.

"Closest" does not mean "close." First, comparison can establish relative ages among surviving texts, but cannot measure the distance between those survivors and the Buddha. That gap lies beyond this method. Moreover, comparison discourse by discourse shows that the Pali version is not invariably older. The Chinese Madhyama Āgama sometimes preserves an earlier reading of the same passage. Second, Schopen's critique operates on another axis. Inscriptions establish that monks actually owned property and actually worshipped relics, showing a divergence between the discourses' descriptions and actual practice. That concerns early Buddhism as a lived religion, not early teachings as a body of discourse. This book may therefore say that the Nikāyas are closest to early teachings, but cannot extend that statement to say that they are closest to early Buddhism.

The same criterion applies to the Abhidhamma Piṭaka. The discussions in 2.2 and 2.5 have already assessed its position. The core method is early, while the seven systematic treatises are later. That assessment uses precisely this criterion, with particularly clear evidence. Only two complete canonical Abhidharma collections survive: Southern Buddhism's seven Pali treatises and the Sarvāstivāda's seven treatises in Chinese translation. The two schools share a core of discourses and Vinaya, yet their seven-treatise collections have almost nothing in common.⁷ On this basis, Frauwallner places the dividing line before Aśoka's missions. The two Abhidharma traditions share an ancient basic stock that must predate those missions and have been carried abroad as a common inheritance. The developments built on that stock address the same problems but offer wholly different answers, belonging to each school's separate work.⁸ The discourses already contain examples of doctrinal matrices as a method. The seven treatises as books were written separately after the schools parted. Together, these two propositions state this book's judgment about the textual layers of the Abhidhamma Piṭaka.

With the criterion established, the question after the canon's commitment to writing shifts from "What does this canon contain?" to "Who has the authority to interpret it?" In Sri Lanka, that question soon took the form of concrete institutional opposition.

IV. Where did the rift at Abhayagiri begin?¶

After defeating the foreign invaders and regaining his throne, King Vaṭṭagāmaṇī Abhaya granted Abhayagiri Vihāra to the monk Mahātissa around 29 BCE. Mahātissa was a different person from the missionary Mahinda. This grant later had far-reaching consequences.

Mahātissa had previously been expelled by the Mahāvihāra saṅgha on grounds of a disciplinary offense. In the Mahāvihāra's account, the king's act gave royal protection to someone whom the saṅgha had judged. This marked the beginning of a rift between the Mahāvihāra and the royal court.⁹

The Mahāvihāra immediately severed relations with Abhayagiri, refusing to hold the uposatha together with monks residing there. The uposatha is the central ceremony affirming the saṅgha's concord, and its joint observance is a crucial sign that two groups of monks belong to the same legitimate saṅgha. Breaking uposatha relations was, institutionally, a declaration that the two monasteries belonged to different lineages.

Yet it is necessary to ask whether the disciplinary case was only a disciplinary case.

Whether the charge was true and why Mahātissa was the one charged are separate questions, though they are easily conflated.

On the first question, the grant of the monastery actually counts against Mahātissa. The grant came after the disciplinary action. This is often read as the king's retaliation, but its evidentiary force is stronger than that. Building a monastery for an uncontroversial elder would have cost the king little. Building one for a man just judged by the country's most authoritative monastery carried the cost of an open breach with the Mahāvihāra. Willingness to pay that price itself shows the depth of the relationship. Moreover, the disciplinary charge of excessive association with lay households judges the actual state of the relationship, not the monk's subjective fault. The rule offers no exemption on the grounds that "I did not initiate it." There was therefore probably no discrepancy between the charge and the facts. Whether the relationship was opportunism or loyalty remains undetermined, since the chronicles permit both readings. On the reasoning above, however, whether that can be determined does not affect the conclusion.

This inference follows the sequence adopted in the discussion, in which the grant comes after the expulsion. The summary of Gombrich in the note on the two monasteries gives the grant before the expulsion.

The second question is the important one. In a saṅgha sustained by royal and aristocratic patronage, almost anyone might cross this boundary. The Mahāvihāra itself was a royal foundation and had received royal grants for two centuries. If close association with a king constituted an offense, every elder of the Mahāvihāra was implicated as well. When a rule that almost anyone might violate is invoked against just one person, what matters is the mechanism that activates it, rather than the rule itself. The claim that "it began as a disciplinary dispute and only later became political" therefore fails. In such a saṅgha, disciplinary affairs and the politics of patronage were two sides of the same matter from the outset. What could not be accommodated was a rival center of equal standing, separately established around one person with royal funding, rather than a particular private relationship. The palace had redrawn the boundaries of the saṅgha.

The confrontation later developed far beyond a single disciplinary action, as several pieces of external evidence show.

Faxian's external testimony, around 410 CE. On reaching Anurādhapura, Faxian left what remains the most important external record. The city had three great monasteries with a combined population of more than ten thousand monks. Abhayagiri, he wrote, had "more than five thousand monks." He stayed on the island for two years and obtained the Mahīśāsaka Vinaya, along with the Dīrgha Āgama, Saṃyukta Āgama, and other texts unavailable in China.¹⁰ This shows that texts from schools outside the Mahāvihāra tradition circulated within the Abhayagiri system, beyond the Tambapaṇṇiya (Theravāda) lineage transmitted by the Mahāvihāra. The monastery the chronicles described as heterodox and lax clearly exceeded the Mahāvihāra in size.

Yijing's indirect record, around 671–695 CE. Yijing did not visit Sri Lanka, but recorded that its monks fell into two groups, attached respectively to the Mahāvihāra and Abhayagiri. They "do not reside together, eat together, or observe the uposatha together" and formed two wholly independent traditions.

Taken together, the evidence indicates that the Abhayagiri dispute began as a conflict over patronage politics and monastic boundaries. It subsequently also became a dispute about what counted as orthodox Buddhism. The Mahāvihāra upheld Vibhajjavāda's closed canon and Vinaya, while Abhayagiri remained open to a wider range of Buddhist traditions coming from India.

V. What was at stake between the two monasteries?¶

The division cannot simply be described as "orthodoxy versus heresy." That is the Mahāvihāra's account of itself, not a historical description. This book sees the monasteries as representing two different strategies within Southern Buddhism:

The Mahāvihāra took the Tipiṭaka and the old commentaries as its standard, and repeatedly used royal power to exclude teachings from outside. The Mahāvaṃsa records kings suppressing Vetulya teachings and using the minister Kapila to curb dissent; it also records the expulsion from the island of sixty Abhayagiri monks influenced by Vetulya. The Mahāvihāra grounded its standing in the authenticity of the Pali scriptures fixed in writing and in disciplinary continuity marked by joint observance of the uposatha.

Abhayagiri chose openness and assimilation. It did not embrace the Mahāyāna at the outset, but was inclined to accept different currents of thought from India. First came Sarvāstivāda and Mahāsāṃghika texts, then Mahāyāna scriptures, and possibly Vajrayāna influences later still. From Abhayagiri's perspective, this meant remaining attentive to the wider Indian Buddhist tradition, not abandoning its own. A specific expression of this openness was Abhayagiri's cult of Avalokiteśvara, locally called Nātha, which had no counterpart in the Mahāvihāra tradition. Xuanzang's seventh-century account likewise distinguishes the two monasteries, describing Abhayagiri as studying both the Greater and Lesser Vehicles, while the Mahāvihāra upheld the Lesser Vehicle alone and rejected the Mahāyāna. Xuanzang himself never visited Sri Lanka. His report was hearsay and belongs to a different category from Faxian's eyewitness account, but the two sources point to the same configuration.

VI. Three competing monasteries and a prolonged confrontation¶

In the third to fourth centuries CE, Jetavana Vihāra appeared at Anurādhapura as its third principal monastery. It initially separated from Abhayagiri over differences of position and established itself independently. It eventually formed a distinct tradition. Anurādhapura thus came to have three competing monasteries.

This was the situation Faxian encountered around 410. All three monasteries were substantial. Abhayagiri was the largest, followed by the Mahāvihāra and then Jetavana. Their shares of royal patronage rose and fell with successive kings' religious inclinations. Some kings generously supported Abhayagiri; others were closer to the Mahāvihāra.

In the fifth century CE, Buddhaghosa, from South India, resided at the Mahāvihāra, composed the Visuddhimagga, and systematically commented on the divisions of the Pali Tipiṭaka, as the next chapter discusses. This step advanced the Mahāvihāra's interpretive authority from guarding an authoritative text to producing authoritative commentary. Thereafter, even when Abhayagiri enjoyed a political advantage, it could hardly match the accumulated commentarial resources of the Mahāvihāra.

VII. The Mahāvihāra's ultimate victory¶

The confrontation among the three monasteries, lasting nearly a thousand years, finally ended in the twelfth century CE with a politically coercive merger.

King Parakkamabāhu I, who reigned around 1153–1186, made unifying the saṅgha and reforming the dispensation his political objectives. Under his direction, monks of Abhayagiri and Jetavana were required to accept reordination under the Mahāvihāra's authority or else return to lay life or leave. This was an enforced religious unification, not a voluntary doctrinal synthesis. Thereafter the Mahāvihāra's Theravāda tradition became Sri Lanka's sole institutional orthodoxy, continuing to the present.¹¹

This book's position is that the Mahāvihāra's ultimate victory was determined by politics, rather than by the natural outcome of doctrinal competition. The open approach represented by Abhayagiri was never refuted in argument. Its defeat was chiefly due to the failure of its relations with successive kings and the gap that developed between it and the Mahāvihāra in building a commentarial system. Accounts of Sri Lankan Buddhist history often describe the twelfth-century forced merger as "purification" and a "return to the right path." That account itself still comes from the Mahāvihāra's textual tradition. This book simply makes that provenance explicit.

VIII. Its place in intellectual history¶

Sri Lanka's thousand-year dispute was foundational for the history of Southern Buddhism as a whole.

It determined what "Southern Buddhism" means. The Mahāvihāra's victory means that the institutional foundations, scriptural system, and commentarial tradition of what is now called Southern Theravāda Buddhism are largely its inheritance. Abhayagiri's more open path accepted diverse Indian traditions and remained flexible toward Mahāyāna texts. It ended in defeat in Sri Lanka. Some of its influence nevertheless reached mainland Southeast Asia through other channels, as discussed in 9.3.

It established the centrality of Pali literature. The Pali Tipiṭaka's commitment to writing in Sri Lanka and preservation by the Mahāvihāra made the island the most important source for the Theravāda traditions of Southeast Asia for more than a thousand years thereafter. The Theravāda traditions of Myanmar, Thailand, Cambodia, and Laos can all be traced back to this Sri Lankan line (9.3).

It provides a historical example of the strategy of a closed canon. The Mahāvihāra's closed adherence to orthodoxy secured the purity of its transmission in the short term, but in the long term depended on political protection to overcome competitors. Abhayagiri's openness was more responsive to living currents of thought, but left it structurally vulnerable through the loss of a core institutional identity. The observation is historical, and this book makes no value judgment between closure and openness.

The pattern recurs more widely. Extending the argument of this chapter reveals a recurring structure. The historical victories of Buddhist schools and traditions have mostly resulted from political choices, with doctrinal superiority serving merely as retrospective decoration. Aśoka's purification of the saṅgha was royal intervention to select which school was more orthodox. The Tibetan court's choice of Indian Madhyamaka and expulsion of the Chinese Chan tradition after the Samyé debate was a choice by courtiers and political alliances between two approaches, not the natural outcome of scriptural and doctrinal discussion. Essay XI examines this in detail. As this chapter has just shown, the Mahāvihāra prevailed over Abhayagiri through a king's decree, not by refuting its doctrines.

This book's position is that a still wider view concerns both which branch prevailed and the rise and decline of Buddhism itself. Its broad direction, too, was a function of political change. The rapid early growth of the Buddha's saṅgha depended on continuing support from political patrons such as Kings Bimbisāra and Prasenajit. Nālandā's almost overnight destruction in the late twelfth century was less a result of intellectual exhaustion than a direct consequence of the Pāla dynasty's collapse and Bakhtiyār Khaljī's military campaigns. The Buddhist persecutions under China's Three Wu Emperors and One Zong Emperor were driven by secular rulers' economic calculations and ideological consolidation, not by the refutation of a Buddhist philosophical proposition. The Meiji government's separation of kami and buddhas and its coercion of monks into lay life and marriage fundamentally concerned replacing Buddhism with Shintō in the public order during nation-state construction. It changed the institutional form of modern Japanese Buddhism as a whole, rather than rejecting a particular school's doctrines.

There were, of course, developments genuinely driven by internal doctrinal concerns, and this book identifies them where appropriate. In broad direction, however, Buddhism's historical expansion, contraction, division, and unification more closely resemble a history of responses to political conditions than an internal history of thought unfolding on its own. This background must remain in view when reading a history of Buddhist philosophy.

The next chapter turns to the core of the Mahāvihāra commentarial tradition, the fifth-century Buddhaghosa and his Visuddhimagga. The Visuddhimagga was Southern Theravāda Buddhism's most important project of systematizing philosophy and practice.

Further reading¶

• Peter Skilling. "Theravāda in History." Pacific World: Journal of the Institute of Buddhist Studies, 3rd series, no. 11 (2009): 61–93.
• Stephen C. Berkwitz. South Asian Buddhism: A Survey. London: Routledge, 2010.

References¶