3. The Dipavamsa



Having considered the epigraphical evidence, I would like to now turn to the later textual accounts. We have seen that important parts of the Pali tradition have been confirmed by the epigraphical findings. With the possible exception of the passage from the Mahāsaṅghika Vinaya discussed earlier, the northern traditions are entirely lacking in archaeological support for this period. But this does not mean that we should accept the Mahāvihāravāsin tradition in toto. I have already indicated that I have severe reservations about the Dīpavaṁsa’s account of the formation of the schools, and it is this that we now consider. The principle question is whether we can accept the Dīpavaṁsa’s identification of the Mahāsaṅghikas with the laxist Vajjiputtakas of the Second Council.

Recent scholarship applauds the death of the Dīpavaṁsa’s theory. But certain scholars, having attended the funeral in the sunny afternoon, return in the deep of night with a shovel. They dig the earth, still soft, and disturb the corpse from the sleep of eternity which it well deserved. With diverse wierdings and incantations they infuse it with a vitality that is unnatural, and set it to its awful task: to destroy the younglings that they should not grow to the fullness of new life. My mission is clear: to cut off the Dīpavaṁsa schism theory like a palm-tree stump, so that it is no longer subject to future arising; then chop the wood into chips, burn the chips, and disperse the ashes in the wind.

Obviously I do not wish to criticize the Dīpavaṁsa in general. Nor do I wish to criticize everything about the Dīpavaṁsa’s account of the sects: the sequence of arising of sects and their mutual interrelationships is, generally speaking, no less plausible than any other; and the fact that the text ascribes the root schism to a dispute on textual redaction has an element of plausibility.

Specifically, I wish to refute the Dīpavaṁsa’s assertion that the Mahāsaṅghikas originated from a reformed group of Vajjiputtakas who held a separate ‘Great Council’ after the Second Council. This is supported by no other source. It clearly contradicts the central message of the Second Council as recorded in all the Vinayas: the dispute was successfully resolved.

A close reading of the Dīpavaṁsa shows that the passage on the schisms is an interpolation into a separate passage dealing with the Second and Third Councils. Dīpavaṁsa 4.68 clearly expresses the conclusion of the Second Council: Aṭṭhamāsehi niṭṭhāsi dutiyo saṅgaho ayan’ti (‘In eight months the Second Council was completed.’) Here the word niṭṭhāsi conveys completion, telling us that the story was supposed to end here. This terminological hint is backed up with a syntactic feature:  the line ends with the particle –ti, which indicates the end of a section. Thus the Second Council as narrated in the Dīpavaṁsa (or its source) originally concluded with the successful resolution of the Council, in accord with all the Vinaya accounts.

These textual details may be ambiguous, but there’s more. Following this closure of the Second Council, the Dīpavaṁsa goes on to give the account of the emergence of the Mahāsaṅghika and the subsequent schisms leading to the formation of all eighteen schools. Obviously this must have been a process that took many years. But following all this Dīpavaṁsa 5.1 links back to the Second Council:


In the future, in a hundred years and eighteen,

Will arise that bhikkhu, a proper ascetic.[1]


The ‘proper ascetic’ is Moggaliputtatissa, and according to the Dīpavaṁsa’s chronology the date of ‘118 years in the future’ is the period between the Second and Third Councils.  In other words this phrase, though supposedly set after the entire schismatic process, is spoken from the point of view immediately following the Second Council. The entire story of the schisms has been interpolated here, leaving the ‘118 years in the future’ hanging without context. We could ask for no clearer indication that the entire account of the schisms and the formation of the Mahāsaṅghika is foreign to the account of the Councils.

Noting that the schisms account is entirely absent from the Vinaya commentaries, Cousins concludes that: ‘This strongly suggests that no account of the “eighteen schools” was preserved in the commentarial tradition of the Mahāvihāra.’[2] He further remarks: ‘Erich Frauwallner has presented evidence that the account of the formation of the eighteen schools in the Dīpavaṁsa does not derive from the old commentarial tradition of the Mahāvihāra and may in fact be from an Abhayagiri source…’.[3] In any case, the passage is closely related to Vasumitra, Bhavya I, and the Śāriputraparipṛcchā, and hence clearly derives from a ‘northern’ source. It is ironic that the same text that so strongly condemns all other schools itself contains a corrupt interpolation. The Mahāvihāra would have been better off sticking to their own more reliable commentarial traditions.

In accepting this northern source and attempting to reconcile it with their own quite different history, the Mahāvihāra inevitably ended up with an incoherent account. It is perfectly clear that the authors of the Second Council passages, both in the Vinayas and the Dīpavaṁsa, intended this to be read as the story of a significant trauma in Buddhist history, one which nevertheless was surmounted in harmony due to the diligent application of the principles of the Vinaya. Crucially, the Mahāsaṅghikas maintain exactly the same tradition in their own Vinaya. They have the same rules prohibiting the use of money as found in all other schools. Accordingly, they condemn the Vajjiputtakas, refute them in the Second Council, and conclude their Council passage by saying: ‘Thus all Elders should train together in harmony’.[4]

In attempting to fuse the account of the Council and the schisms, the Dīpavaṁsa obscures the plain fact that the problematic issues discussed in relation to the Vajjiputtakas in the Second Council have precisely nothing in common with the issues concerning the Mahāsaṅghikas of the ‘Great Council’. The Second Council accuses the Vajjiputtakas of the 10 points of laxity in Vinaya. But the story of the Mahāsaṅghika schism in the Dīpavaṁsa says nothing about Vinaya. There the crucial issue was a reshaping of the Buddhist scriptures. We must be clear about this: despite statements to the contrary by some modern scholars,[5] the Dīpavaṁsa does not ascribe the schism to the 10 points. Rather, it relates the Second Council narrative including the 10 points, then proceeds to describe how the defeated Vajjiputtakas reformed as the Mahāsaṅghikas[6] and revised the texts. The connection between the Mahāsaṅghikas and the 10 points is a narrative sleight-of-hand: it is the work of Māra. We are conditioned by the former passage to read the 10 points into the later passage; this is the narrative intent of the Dīpavaṁsa. But once we realize the two accounts have completely different origins, any connection between the Mahāsaṅghikas and the 10 points vanishes. Like a sky-flower, it was an illusion created by the mind.

The very idea that the Mahāsaṅghikas could have rejected the texts directly contradicts a crucial assumption of the whole Second Council story, that is, that the Sangha reached agreement regarding the 10 Vinaya issues by referring to their shared disciplinary code. All freely participated in the Council, and all agreed to solve the problem by appointing a committee of eight, whose verdict, since it was carefully justified point by point against the universally accepted Vinaya rules, was accepted by all.  If the Vajjiputtakas were interested in textual revision, they would surely have contested the textual references put forward by the committee.

A further difficulty with the Dīpavaṁsa’s position is that it assumes that the Vajjiputtakas could happily go about ignoring the Second Council and making their own schism without any response from the rest of the Sangha. This is patently absurd, since the events that triggered the Second Council itself were of less importance than a major schism, yet monks gathered from all over Buddhist India. Every other account we have of the root schism tells of a gathering of monks who disputed at length, and split only after failing to find a resolution.

This is particularly telling when compared with the Samantapāsādikā’s account of what happened after the Second Council. The Elders (unnamed) considered whether another disaster would afflict the sāsana, and saw that in 118 years in the time of Aśoka, many monks would enter the Sangha seeking gains and fame. They considered how to avert this, and saw that the only being capable was a Brahma named Tissa. They went to the Brahma world, and begged Tissa to descend to save Buddhism.[7] He agreed – how could it be otherwise? – whereupon the Elders returned to the human realm and organized a couple of young arahants, Siggava and Caṇḍavajji, to teach the Brahma when he was reborn as Moggaliputtatissa. This is a wonderfully dramatic scene-setter for the Third Council. But if we are to accept the Dīpavaṁsa’s account, then while the Elders were making such elaborate preparations for saving Buddhism in the future, under their very noses the Vajjiputtakas were destroying the unity of the Sangha forever. (Perhaps they were away in the Brahma world while this was going on.)

After describing the root schism, the Dīpavaṁsa tells us that the various schools split off from each other one by one. It doesn’t mention any reasons for why this multitude of schisms occurred, nor why they should happen so quickly. Nevertheless, the whole process was over and done with and the ‘eighteen’ schools were all formed before the time of Aśoka. The Second Council was in 100 an, and since the Dīpavaṁsa is a ‘long chronology’ text, this allows 118 years for the schools to form.[8] This is short enough, but if we follow the median chronology we have only 40 years or so. The process of forming a sect in a religion like Buddhism is not easy. It requires a charismatic leader, one who can articulate a convincing independent interpretation of the teachings, inspiring both monastics and lay followers. It requires a degree of geographical separation for the building of the requisite lay support. It requires building an institutional basis, i.e. at least one monastery, including shrines, meeting hall, residential quarters, and so on. All of this happened, according to the Dīpavaṁsa, within one or two generations, leaving not a single physical trace. This contrasts with other accounts like the Śāriputraparipṛcchā, which give the process several centuries to unfold.

Perhaps even more implausibly, this account implies that in the following centuries there were hardly any new sects. True, the commentaries do mention a few schools that arose subsequently, but we are expected to believe that ‘eighteen’ schools arose almost immediately, and in a thousand years after that only a small number of new schools gradually came to be.

A crucial consequence of the Dīpavaṁsa’s view would be that the Aśokan missions were ‘Theravādins’ in the narrow sense, meaning the same school as the Mahāvihāravāsins, rather than the Sthaviras or Vibhajjavādins in general. Thus the Theravādins alone were responsible for converting virtually the whole of India to Buddhism, a situation which blatantly contradicts all the available epigraphic and textual evidence.

It may seem ungenerous to impute to the Theravādins the idea that they themselves spread Buddhism over all of India, a perspective of breathtaking conceit. But the main epigraphic evidence for the school from the mainland confirms exactly that. Two inscriptions from the Sinhalese monastery in Nāgārjunikoṇḍa, dated to around 250 ce, refer to the teachers of the ‘Theriyas, Vibhajjavādas, Mahāvihāravāsins’, who have brought faith to various lands: Kaśmīr, Gandhāra, Yava[na] (= Yonaloka of the mission accounts = Greek Bactria), Vanavāsi, Cīna-Cilāta, Tosali, Avaraṁta, Vaṅga, Da[mila], [Pa]lura, and Tambapṁṇidīpa.[9] This evidence predates the Dīpavaṁsa and the mission accounts, but the similarity of the phrasing, as demonstrated by Cousins, shows that they must derive from a common source, presumably the old Sinhalese tradition.

The Mahāvihāravāsins wanted to portray themselves at the centre of Buddhism. The unique creative genius of the Dīpavaṁsa is to enshrine this world-view within the fundamental myth of Buddhism. Right from the outset it declares that the Buddha, during the seven days after his awakening, surveyed the world, saw Sri Lanka, and predicted the advent of his Dhamma there after the Third Council.[10] The unified Sangha is referred to as the ‘Theravāda’ from the time of the First Council on.[11] There is no doubt, given the opening passages, that by this the Dīpavaṁsa, with a magisterial disregard for chronology, means the Theravādins in the narrow sense (= Mahāvihāravāsin).

In this context the motive for placing the root schism before Aśoka is clear. If the schisms happened after Aśoka, then it would be impossible to assert that Aśoka was the specific patron of the Theravāda. He would have to be seen as the supporter of Buddhism in general. If the schism was in the time of Aśoka, this would contradict the triumphant message of Moggaliputtatissa’s successful Third Council. The only solution is to put the schisms before Aśoka. Then the other schools are implicitly excluded from the narrative, and Aśoka becomes by default the special patron of the Theravāda.


The Heresy of Grammar

If we agree that the Dīpavaṁsa account of the schism cannot refer to the period immediately after the Second Council, can we establish when and in what context it really did originate? I think we can. To do this, we need to look more closely at the way the schism is actually described in the Dīpavaṁsa. It emphasizes the interpretative principles used at the Council:


Teachings that are metaphorical, and those that are definitive

Those with meaning drawn out and with meaning to be drawn out

were elucidated by the Sutta experts.[12]


This verse is mockingly echoed in the account of the ‘Mahāsaṅgīti’ (Great Council) of the Vajjiputtakas:


Teachings that are metaphorical, and those that are definitive

Those with meaning drawn out and with meaning to be drawn out;

without understanding, those bhikkhus [confused].[13]


The Dīpavaṁsa further explains (4.77) that the Vajjiputtakas (=Mahāsaṅghikas) confused the nouns, the genders, and so on. In short, they were grammatical heretics, whose foremost crime was bad textuality. It would be unkind to linger on this point, but it is ironic that this accusation is made by the Dīpavaṁsa, perhaps the most badly written book in the Pali language.

Another crucial accusation is that the Vajjiputtaka/Mahāsaṅghikas revised the ancient texts, rejecting the Parivāra, the six books of the Abhidhamma,[14] the  Paṭisambhidā, the Niddesa, some of the Jātakas, and some of the verses, and went on to compose others.[15] These works are all found in the Pali canon. Without exception, modern scholars are agreed that these works are in fact late and cannot be considered authentic buddhavacana. Thus the Mahāsaṅghikas may rightly claim to be the forerunners of an accurate historical-critical approach to Buddhist texts.

The Dīpavaṁsa’s description of the rejected texts is a projection of the Mahāvihāra’s dark side. Subconsciously, they know full well that these texts are late. The virulence of their attack – echoed elsewhere – demonstrates their fear of admitting this, and the concomitant need to externalize the problem. Why are they so afraid? Why not simply admit, as all the evidence would have it, that some of their texts are not buddhavacana? Admitting the unauthenticity of their own texts would destroy their own self-image as the true bastion of original, pure Buddhism. This in turn would make nonsense of the ideology of Sri Lanka as the ‘Dhammadīpa’, and would ruin the Mahāvihāra’s credibility in the competition for royal favours with the Abhayagiri. The fear is quite real: we need not doubt that at certain times the Mahāvihāra had to stand face to face with its own destruction. But the reality of the threat should not blind us to the illusions conjured in response to that threat.

The list of texts rejected is quite precise: ‘some of the Jātakas’, ‘some of the verses’. As is well known, certain Jātakas form part of the early corpus of scriptures, while others were added continuously over many years. Similarly, many of the verses of the Khuddakanikāya are early, but many more are among the latest strata of additions to the canon.

In their current form, all these rejected texts are post-Aśokan. While the Abhidhamma project must have been underway in the time of Aśoka – as suggested by Moggaliputtatissa’s Abhidhamma connections and confirmed by substantial similarities among existing Abhidhamma texts – the texts as we know them must have been finalized later. Similarly, the Paṭisambhidāmagga is dated around 100 bce.[16] The Niddesa is an application of Abhidhamma methodology to some early poems, and must stem from a similar period. Thus we are firmly in the ‘late canonical’ period of the Mahāvihāra literature, and accordingly should look for the dispute in this period.

If we want to know who the Mahāvihāravāsins were arguing with, the Kathāvatthu commentary, though redacted later, is our main source of information. Overwhelmingly, this mentions disagreements with the Andhakas,[17] a group of Mahāsaṅghika schools in the Andhra region, including Amarāvati, Nāgārjunikoṇḍa, etc. Thus we know that the Mahāvihāravāsins debated Abhidhamma extensively with the Andhakas, and it must surely follow that the Andhakas rejected the Mahāvihāra’s Abhidhamma and related literature. But this is perhaps not of such great importance in itself, for it is probable that most of the Indic schools did not accept the Mahāvihāra Abhidhamma – in fact, they had probably hardly even heard of it. What matters is not so much that the Andhakas rejected these texts, but that the Mahāvihāravāsins knew they rejected them, and it hurt.

The Paṭisambhidāmagga and the Niddesa are also crucial here, though in a different way. They are both included in the Khuddakanikāya, but each has strong affinities with the Abhidhamma. The paṭisambhidās were a minor doctrinal set for the early Suttas. The primary meaning relates skill at textual exegesis with penetration to the Dhamma: dhamma (text); attha (meaning); nirutti (language); paṭibhāṇa (eloquence, i.e. the ability of one who, knowing the text and its meaning, and being fluent in the ways of expression, to spontaneously give an accurate and inspiring teaching). The Paṭisambhidāmagga takes this occasional group and, stretching their application almost beyond recognition, develops the first distinctive Mahāvihāra ‘Book of the Way’. As with all canonical Abhidhamma, the emphasis is on precise, clear-cut doctrinal definition. Warder shows that the emphasis on this particular doctrinal category is peculiar to the Mahāvihāra.[18]

The Niddesas are similarly about textual exegesis. They are a pair of Abhidhamma-style commentaries on the Khaggavisāṇa Sutta, Aṭṭhakavagga, and Pārāyanavagga, early poems subsequently compiled in the Sutta Nipāta. Their style is curiously Abhidhammic, in stark contrast with the casual, natural language of the texts on which they comment. In fact, they come across as an attempt to ‘tame’ some early texts which express doctrinal positions not easy to reconcile with the Mahāvihāra’s developing stance.

As for the late Jātakas and verses, it would seem as if these were not so likely to be doctrinally controversial. They mainly deal with the emerging Bodhisattva doctrine, which was prevalent throughout all Buddhist schools, and if anything we would expect Mahāsaṅghika schools, such as the Andhakas, to be the forerunners in this movement. Nevertheless, the Kathāvatthu does record several controversies regarding the Bodhisattva and his career. The Andhakas evidently asserted that the Bodhisattva was born as an animal or in hell of his free will (issariyakāmakārikāhetu),[19] which for them was an expression of his transcendent (lokuttara) nature, but which the Mahāvihāravāsins saw as a denial of the law of kamma. It is not sure whether the Mahāsaṅghikas rejected certain Jātakas and verses because of doctrinal implications such as these, or simply because they were extra-canonical.

Recalling the Dīpavaṁsa’s accusations of bad textuality, I am struck by the aptness of a remark by Franklin Edgerton. Previously, Emile Senart had edited one of the most important and difficult works in the Mahāsaṅghika literature, the Mahāvastu, in the light of traditional Sanskrit and Pali forms. Edgerton commented that: ‘Senart’s extensive notes often let the reader perceive the despair which constantly threatened to overwhelm him.’[20] Following Edgerton’s work, it is now generally acknowledged that the Mahāsaṅghika texts are written in a distinctively Mahāsaṅghika ‘Hybrid Sanskrit’, and are not just bad Sanskrit. But Senart’s despair would echo the reaction of any Mahāvihāravāsin scholars, brought up on the simpler, cleaner Pali tradition, who confronted the Mahāsaṅghika texts. We therefore suggest that the Dīpavaṁsa’s accusations of textual rejection and bad grammar were levelled specifically at the Mahāsaṅghika schools of Andhra, and by extension Sanskritic or ‘modernized’ Buddhism generally, such as the Abhayagiri.[21] In the usual mythic style, contemporary debates were backdated to give them a universal relevance.

There are certain other sources that likewise attribute the schisms to linguistic variation. For example, Vinītadeva gives this cause, and mentions the following language usages: Sarvāstivādins used Sanskrit; Mahāsaṅghikas used Prākrit; Saṁmitiyas used Apabhraṁśa; the Sthaviras used Paiśacī.[22] The Dīpavaṁsa’s account must be seen in this light, that is, it highlights a primarily linguistic dispute. But the linguistic differences are merely a consequence of geographical dispersal. It is hardly possible that communities living in the same location would dispute over what language to use. The languages must have diverged as the schools spread over India and followed the Buddha’s advice to teach the Dhamma in the local dialect.[23] Notice that the Sri Lankans did not follow this advice, and preserved the Dhamma in a foreign tongue, which they strenuously believed to be literally the language spoken by the Buddha.

The fact that the texts were not translated into Sinhalese indicates that they had attained a high degree of ‘canonization’ even before reaching the island. This tendency culminated in the later ideology of linguistic essentialism, where Pali was regarded as the ‘root language of all beings’.[24] This means that one who had attained the paṭisambhidās would through their own insight understand that phassā or vedano are incorrect nominative forms and would know that in the ‘essence-language’ (Pali) these should be phasso and vedanā. For the Pali school, the Mahāsaṅghika Hybrid Sanskrit was not merely a variant dialect, but was a fundamental subversion of the Dhamma.[25]


The Sri Lankan Context

All this makes more sense when we consider the climate in which the Dīpavaṁsa and subsequent chronicles were composed. The events described close with the death of king Mahāsena about 304 ce, which follows the the triumph of the Mahāvihāra over their bitter rivals the Abhayagiri monastery. This rivalry had started about 400 years earlier, when the Abhayagiri monastery, having been established by king Vaṭṭagāminī, became the home of Bahalamassutissa, the follower of a certain Mahātissa, who was expelled from the Mahāvihāra for unbecoming familiarity with lay-folk. This monastery was subsequently regarded as schismatic from the Theravāda.[26] The Abhayagiri became associated with suspect teachings imported from the mainland. Since little if any of their literature survives, it is unclear exactly how their doctrinal position evolved.[27]

Both monasteries received royal support until the time of Vohārika Tissa, around 230 ce, when the Abhayagirivāsins were accused of importing ‘Vetulya’ scriptures. It is usually presumed that these have something to do with Mahāyāna, though there is little direct evidence. In any case, these scriptures were suppressed. There is no discussion of the doctrines taught or why they are so dangerous. We might even be forgiven for wondering whether the actual contents of these texts were at all relevant.[28]

In any case, the ‘Vetulya’ books were burned and the bhikkhus disgraced. Following this, the kings Vohārikatissa, Goṭhābhaya, and Jeṭṭhatissa supported the Mahāvihāra. But the Abhayagiri continued to cause trouble. 60 bhikkhus were expelled by Goṭhābhaya for upholding the Vetullavāda; these are described in the Mahāvaṁsa as ‘thorns in the conqueror’s religion’,[29] exactly as the Dīpavaṁsa called the Vajjiputtakas and other secessionists ‘thorns on the banyan tree’. Much later, the Nikāyasaṅgraha of Dharmakīrti (14th Century) was to turn this purely literary analogy into history, claiming that around 32 bce, shortly after the Abhayagiri was established, a group of Vajjiputtaka bhikkhus, under the leadership of a certain Dharmaruci, came to Sri Lanka and, being rejected by the Mahāvihāra, found support in the Abhayagiri. These were the laxist Vajjiputtakas/Mahāsaṅghikas.[30]

But soon the tables turned. A bhikkhu called Saṅghamitta arrived from India. Painted in the darkest colours by the Mahāvihāravāsins, this monk helped the Abhayagiri to regroup. He was rejected by king Jeṭṭhatissa and fled back to India; but on the accession of Mahāsena he returned and performed the consecration ceremony for the king. Under Saṅghamitta’s influence king Mahāsena persecuted the Mahāvihāra: the monks were driven from the monastery for nine years, and the Abhayagirivāsins, together with the evil minister Soṇa, stripped the Mahāvihāra of its treasures to adorn the Abhayagiri. Supporters of the Mahāvihāra were so appalled that a minister called Meghavaṇṇabhaya retreated to the Malaya region, where the Mahāvihāravāsins dwelt in exile, gathered an army and marched on the capital. But those were chivalrous days. The rebel minister reflected that he should not eat apart from his good friend the king, so on the eve of battle they shared a meal. The king asked why Meghavaṇṇabhaya was intent on war, and he answered that he could not bear to see the destruction of the Mahāvihāra. The king wisely asked forgiveness and pledged to rebuild the Mahāvihāra: an excellent example for those who would wage holy war today. But one of the king’s wives was so grieved she had Saṅghamitta and Soṇa assassinated. The Abhayagiri was then stripped to adorn the Mahāvihāra.

These events culminated with the death of Mahāsena. The Mahāvaṁsa, in Geiger’s translation, ends with the words: ‘Thus did he gather to himself much merit and much guilt,’ perfectly encapsulating the deeply ambiguous moral world of the Sri Lankan chronicles. Throughout we see a genuine devotion to the ideals of the Dhamma. While there is little evidence of the permeation of advanced teachings and practices through the culture, still the kings make persistent efforts to live up to the ideals of the righteous king as represented by Aśoka. But the demands of government inevitably compromise these lofty ideals. Having closely intertwined their conception of Buddhism with the Sri Lankan nation, the Sangha finds it impossible to retain an independence from the political arena. While we cannot approve of all we find within these bloodied pages, we must remember that history is like this, everywhere, all the time. On the whole Sri Lanka is no worse than any, and probably better than most. No doubt other Buddhist traditions have faced bitter choices and deadly struggles. The difference is that we know nothing about them, as the Sinhalese are the only Buddhists of ancient India to preserve a historical literature. That literature asserts that without sometimes violent support Buddhism would not have survived. While we must deplore the violence, we cannot deny that the tradition, including the texts that tell us this story, has in fact survived where all others failed.

The Dīpavaṁsa and Mahāvaṁsa were formed in a climate of desperate and vicious struggle. For the monks of the Mahāvihāra, the difference between sects was not a gentlemanly disagreement on points of Abhidhamma, but a deadly battle for survival. The formation of the ‘classical’ phase of Mahāvihāravāsin literature – the chronicles and commentaries – was the direct outcome of this struggle.

Of course this picture is one-sided and melodramatic. Fa-xian, who spent two years in Sri Lanka a little after the events we have described, sees the Abhayagiri as the main monastery; it had 5000 monks, while the Mahāvihāra could only muster 3000. Characteristically, Fa-xian does not speak of any tension, but praises the beauty and devotion he witnesses in both monasteries. The combative spirit of the chronicles is as much a symptom of a frame of mind as it is the record of actual disputes.

There is something in these stories of the past that filled an urgent need for the Sangha in the present. The Mahāvihāravāsins, in those violent and intensely politicized times, needed an ‘other’. This may be seen as an expression of the vibhajjavāda ideology, a need to separate oneself to create a sense of sacredness and purity. Throughout religious and magical thought, a ritualized physical separation is a source and a sustenance for holy power. The definition and identification of the ‘other’ is required in order to define and identify the ‘self’. The manifest need to demonize the ‘other’ hints at the dark side of the Mahāvihāravāsins: they are rejecting what they fear in themselves. We have already noted the ironies inherent within the Dīpavaṁsa: written atrociously, it accuses ‘them’ of bad textuality; and while one of its central theses is a badly grafted foreign import, it accuses ‘them’ of introducing alien elements. We shall see in our discussion of the Śāriputraparipṛcchā that the Dīpavaṁsa is not alone in focussing on the mote in its brother’s eye.

While these ironies may be quaint, even amusing, the same texts contain ironies of a far more dangerous sort. Most obvious is that, despite the tradition’s insistence on preserving ‘original’ Buddhism unchanged, in fact the burden of the chronicles is to legitimize the fusing of the Church and State, a revolutionary innovation without precedent on the mainland. This is why so much stress is laid on the mythic reinvention of Aśoka as champion of the Mahāvihāra’s brand of Buddhism. But going far beyond the example of Aśokan patronage of the Sangha or even interference in Sangha affairs, the chronicles pursue the politicization of Buddhism to its inevitable conclusion: the Buddhist justification of war. The Mahāvaṁsa depicts the guilt-ridden king Duṭṭhagāmini returning from the battlefield and seeking solace from the Sangha for killing thousands of people in battle, just as Aśoka sought solace from Moggaliputtatissa for the murder of the Aśokārāma monks, or Ajātasattu sought solace from the Buddha for his murder of his father king Bimbisāra. The arahants reassure the king that he need not feel so bad, since he has really only killed one and a half people: one was keeping the five precepts, the half had taken refuge in the Triple Gem. The rest don’t count.

Like all good myths, this passage is timeless; hence it has become central to the modern Sri Lankan Sangha’s justification of war against the Tamils. Theravāda, while maintaining a quality textual tradition, in practice preserved neither more nor less of true Buddhism than any other school. But the stark contrast between the ideal monk as depicted in the early Suttas and the reality of Buddhism as lived created a tension on a deep level, a tension which is not resolved, but is projected on the ‘other’.

It was king Parakkamabāhu I (1153-1186) who, in the midst of apparently endless military campaigns, finally reconciled the various Sangha fraternities. The Cūḷavaṁsa pointedly remarks that: ‘despite the vast efforts made in every way by former kings down to the present day, the [bhikkhus] turned away in their demeanour from one another and took delight in all kinds of strife’.[31] The analogy with the Aśokan Council is here made explicit: ‘Even as the Ruler of Men Dhammāsoka with Moggaliputtatissa, so he [Parakkamabāhu] entrusted the grand Elder Mahākassapa…’.[32] Following the Aśokan precedent, they gathered all the monks together, questioned them, solved the problems one by one, expelled the bad monks, and created a unified Sangha ‘as it had been in the Buddha’s time’.[33]

From these few examples – which could be expanded indefinitely – we can see how the Mahāvihāravāsin chronicles are built on a structure of repeating cycles, of recurring parallels. It becomes clear how the Dīpavaṁsa’s depiction of the Mahāsaṅghikas as bad Vajjiputtaka monks is a mythic back-reading from the situation in the time of the Dīpavaṁsa. In myth time is uroboric, perennially swallowing its own tail: it is like this now, so it must have been like this then. The names and the details display a glinting surface of ever-changing appearances, but the underlying patterns play themselves out with reassuring inevitability, like the changing of the seasons or the stars wheeling in the sky. The Sinhalese chronicles boldly meld the political and cultural history of their own people with the fundamental Buddhist myth, the life of the Buddha. Just as each ordination is a ritualized repetition of the Buddha’s renunciation, making that remote act real in the present, so each event in the mythic structure informs the eternal now, the immanent sense of history lived as destiny. Thus the scapegoating and expulsion of the Vajjiputtakas becomes a catharsis required whenever the purity of the Sangha is imperilled.


Was Buddhaghosa a Theravādin?

The notion of purity of lineage is an essential element in the strategy of establishing a school of Buddhism. This is despite the fact that the very notion of paramparā, a particular ordination lineage, is absent from the early texts. Of course, it is not unreasonable to infer from the early texts that they ascribe a certain value to the notion of a direct connection of ordinations from teacher to student. But this can hardly be construed as central.

 In the same way that Warder asked whether Nāgārjuna was a Mahāyānist, sometimes one may wonder to what extent Buddhaghosa, the 5th Century compiler of the definitive Mahāvihāravāsin commentarial tradition, was a Theravādin in terms of his ordination lineage.

There is nothing explicit to go on. The later tradition asserted that he was born in Magadha, but this is a transparent effort to affirm his orthodox background. Interestingly, the Burmese maintain that Buddhaghosa was born in Burma. While no-one but a Burmese would find this plausible, this tradition implies that his ordination would be traced by the Burmese to the mission of Soṇa and Uttara to Suvaṇṇabhūmi. In other words, he came from one of the other missions, not from the mission that established the Mahāvihāra. From the later Burmese perspective of course this is all ‘Theravāda’, but in Buddhaghosa’s day the notion of a unified form of Buddhism throughout south-east Asia did not exist, and there were in fact many schools in the region.

Since Buddhaghosa came from India, and given that the vast majority of Indian Buddhists were not affiliated with the Theravādins in the narrow sense required by the Dīpavaṁsa (= Mahāvihāravāsin), we may well wonder whether his ordination was really ‘Theravādin’. He does mention having stayed in a few places on the mainland, some of which have been tentatively identified in Southern India: ‘Mayūrasuttapaṭṭana’ (Mylapore near Chennai); Kañcipura (Conjevaram near Chennai); and the postscript to the Visuddhimagga describes him as ‘of Moraṇḍacetaka’ (Andhra?).[34]  However, the Mahāvaṁsa says he was born near Bodhgaya, although this is a much later tradition, attributed to  Dharmakīrti of the 14th Century. As far as his ordination goes, the Mahāvaṁsa could hardly be less specific: while wandering ‘around India’, he stayed at ‘a monastery’, where he met ‘a teacher’ called Revata, under whom he took ordination.[35] Revata is said to teach the pāḷi of the Abhidhamma, but pāḷi here is used in its general sense of text and need not imply the Pali canon we know. Buddhaghosa apparently prepared a treatise called Ñāṇodaya, of which nothing is known, and an Aṭṭhasālinī, a commentary on the Dhammasaṅgaṇī. The existing commentary by Buddhaghosa on the Dhammasaṅgaṇī is indeed called the Aṭṭhasālinī, but it is not known if this had any relation to the earlier work, if indeed it ever existed.

When Buddhaghosa wanted to do further work on a paritta commentary, Revata tells him that:


‘Here only the text [pāḷi] has been preserved,

there is no commentary here,

and similarly no Teacher’s Doctrine:

that has fallen apart and is not found.’[36]


Revata then praises the purity of the commentarial tradition of Sri Lanka and encourages Buddhaghosa to go there and learn. This story is a legendary construct to emphasize the superiority of the Sri Lankan tradition; it is doubtful whether the Indians saw things quite the same way. Polemics aside, this tradition gives us no credible basis on which to affirm that Buddhaghosa must have had an ordination in the Mahāvihāra tradition.

I take the example of Buddhaghosa only to make a rhetorical point. But it was normal for monks to travel around different monasteries, staying with different fraternities. This must have happened even more with the Abhayagiri monastery, who were said by the Mahāvihāravāsins to be accepting Indian monks of different traditions. But the Abhayagirivāsins were later fused with the Mahāvihāravāsins, despite this supposed impurity in their ordination lineage.

A similar situation must have obtained throughout south-east Asian Buddhism, for we know that the areas of Thailand, Burma, and Cambodia where Theravāda now flourishes were formerly dominated by Mahāyāna, or Sanskritic Śrāvakayāna Buddhism. We note the widespread occurrence of the cult of Upagupta throughout this region, which is totally absent from Sri Lanka, and wonder whether this gives a hint as to the kind of Buddhism prevalent before the Theravāda orthodoxy. According to I-Tsing, in the lands on the eastern boundaries of India all four major schools flourished, while in the island regions the Mūlasarvāstivāda predominated.[37]

When these areas ‘converted’ to Theravāda (which mainly occurred around the 11th-12th Centuries), it is impossible that all the monks took new ordinations. Of course, the official histories will assert that when the religion was reformed that all the monks conformed to the new system. But the practicalities of this are absurd: sending city administration monks wandering through 1000s of miles of tiger-stalked, bandit-infested, ghost-haunted jungle tracks seeking out countless little villages, trying to persuade senior monks that their ordination is invalid or improper and must be done again, all on the basis of some political compromise in a far-distant capital, in a region of ever-shifting borders and allegiances. As history this is sheer fantasy, and the reality must have been that the reforms would directly affect only certain central monasteries. Others maybe used an informal procedure like a daḷhikamma (strengthening act), which is just an ad hoc procedure invented in lieu of doing a genuine saṅghakamma. But for the majority the reforms would have been irrelevant, even if they heard of them. It is only rational to conclude that the current ‘Theravāda’ lineage, like all others, must be a blend of many different strands.

Bizot’s research in this area shows that the current situation in Theravāda in fact retains two distinct ordination styles.[38] One involves reciting the refuges once during the pabbajjā; in the other, the refuges are recited twice, once ending the words with the anusvāra –ṁ (pronounced -ng), and again with the labial nasal –m. The two statement pabbajjā has its roots in the ancient Mon Buddhism of the Dvāravatī period (7th 8th Centuries), which was possibly introduced into south-east Asia (‘Suvaṇṇabhūmi’) from southern India. Bizot believes that this two-statement pabbajjā was connected with certain esoteric meditation practices. The one-statement pabbajjā of the Mahāvihāra was introduced later, around the 14th – 15th Centuries, by monks who were in contact with Sri Lanka. But when the Sri Lankan lineage was re-established from Thailand, it was with the Mon two-statement pabbajjā. Meanwhile, the one-statement pabbajjā was progressively imposed on the Sangha in south-east Asia, especially following the modernist Dhammayuttika reforms of Prince Mongkut in the 19th Century. In one of those delicious ironies of history, the two-statement Mon pabbajjā now survives only in Sri Lanka, while the one-statement pabbajjā prevails throughout south-east Asia.

The complexity of the situation is acknowledged by Somdet Ñāṇasaṁvara, the current Saṅgharāja of Thailand, in an important work Buddha Sāsana Vaṁsa. This discusses the modern Thai ordination lineage and the reforms introduced in the 19th Century when the Dhammayuttika Nikāya was formed on the basis of the Burmese Mon tradition. It is believed that this tradition stems ultimately from the mission of Soṇa and Uttara to Suvaṇṇabhūmi in Aśoka’s time. Here are some of Somdet Ñāṇasaṁvara’s remarks:


‘From the Buddha’s Mahāparinibbāna until the present, more than 2000 years have passed, thus it is difficult to know whether the pure lineage has come down to us intact or not.’ (16)

‘If the lineage has faded away it is in no way harmful, just like Pukkusāti’s[39] dedication to homelessness was harmless.’ (18)

‘The sasana in both countries [Sri Lanka and Suvaṇṇabhūmi] merged as one in that their lineage came from the same sasana that king Aśoka had sent from the capital at Pāṭaliputta.’ (30)

[After the time of king Parakkamabāhu of Sri Lanka] ‘Sri Lankan bhikkhus conferred with the Rāmañña [Mon] bhikkhus and were of the opinion that since the Sri Lankan bhikkhus were of the line of Soṇa and Uttara they were of the same communion.[40] The Elders thus invited one another to participate in saṅghakamma and together gave higher ordination.’ (31)

[The lineages entered Thailand] ‘many times through many periods… as Buddhism entered the country in different periods, sects, and forms, it is difficult to know how they merged and how they declined.’ (76)

[The Dhammayuttika Nikāya revitalized Thai Buddhism through] ‘re-establishing in Siam a direct lineage from Venerables Mahinda, Soṇa, and Uttara.’ (77)


So while there sometimes appears to be an almost mystical belief in the inviolability of ordination lineages, saner voices are still to be found. No monk alive can guarantee his own ordination lineage. In this situation it is safer and more reasonable to focus on the way the holy life is lived rather than on unverifiable claims of a largely undocumented past.



[1] Anāgate vassasate vassāna’ṭṭhārasāni ca, Uppajjissati so bhikkhu samaṇo paṭirūpako.

Here paṭirūpaka obviously does not mean ‘counterfeit’.

[2] Cousins, The 'Five Points' and the Origins of the Buddhist Schools, 56

[3] Cousins, On the Vibhajjavādins, 153

[4] 如是諸長老應當隨順學 (CBETA, T22, no. 1425, p. 493, c10)

[5] E.g. Nattier and Prebish, 200

[6] The Dīpavaṁsa usually uses the synonymous term Mahāsaṅgītikas.

[7] A mythic mirror-image of the ‘Entreaty by Brahma’ that motivated the Buddha to teach.

[8] See Appendix.

[9] EI, XX, 1929, pg. 22. See Lamotte, History of Indian Buddhism, 299; Cousins, On the Vibhajjavādins, 141.

[10] Dīpavaṁsa 1.14 ff.

[11] Dīpavaṁsa 4.11, 18, 31, 32, 33, 54, 84, 88, 90; 5.28; 6.24, 29, 39, 43, 54.

[12] Dīpavaṁsa 4.22: Pariyāyadesitañcāpi atho nippariyāya desitaṁ,

Nītathaññeva neyyathaṁ dīpiṁsu suttakovidā.

[13] Dīpavaṁsa 4.73: Pariyāya desitaṁ cāpi atho nippariyāya desitaṁ,

Nītathaṁ ce'va neyyathaṁ ajānivāna bhikkhavo.

[14] ‘Six’, because the seventh book, the Kathāvatthu, was not composed until the Third Council, which is later according to the Mahāvihāra’s chronology.

[15] Dīpavaṁsa 4.76, 82

[16] Ñāṇamoḷi, The Path of Discrimination, xxxvii ff.

[17] About half of the disputes are with the Andhakas or their sub-schools.

[18] Ñāṇamoḷi, The Path of Discrimination, introduction

[19] Kathāvatthu 622

[20] Quoted in Prebish, Śaikṣa-Dharmas Revisited, 191

[21] Cf. Roth, lv

[22] Pachow, 42

[23] See Edgerton, 1-2; Lamotte, History of Indian Buddhism, 552-556

[24] Ñāṇamoḷi, Path of Purification 486-487 (XIV 25)

[25] Collins (81) speaks of the ‘text-oriented self-definition’ of the Mahāvihāra.

[26] Mahāvaṁsa 33.99

[27] There is a record in Samantapāsādikā 3.582 of a dispute over a point of Vinaya, which, in a remarkable reminder of the influence of the Aśokan precedent, was resolved by the king’s minister. I cannot locate this passage in the Sudassanavinayavibhāsā, which may have an Abhayagiri connection.

[28] In the Cūḷavaṁsa (the later continuation of the Mahāvaṁsa) there is a story of a certain text called the ‘Dhammadhātu’, which was brought from India. (Cv 41.37ff.)  The king, unable to discern what was right and wrong, enshrined it and worshipped it. The doctrines taught in the text are entirely beside the point: we are told that the king did not understand them. What was at stake was the ritual worship of the physical manuscript.

[29] Mahāvaṁsa 33.111: vetullavādino bhikkhū, abhayagirinivāsino / gāhayitvāsaṭṭhimatte, jinasāsanakaṇṭake.

[30] Lamotte, History of Indian Buddhism, 371. Some modern writers (see Perera, 37) connect these with the Vātsīputrīyas (Puggalavādins). This may not be wholly unjustified, since by the time of the Nikāyasaṅgraha there was not a great deal of clarity regarding these sects.

[31] Cūḷavaṁsa 73.19  These events are also recorded in Parakkamabāhu’s Galvihara inscription. See Hallisey, 178

[32] Cūḷavaṁsa 78.6

[33] Cūḷavaṁsa 78.27

[34] Buddhaghosa, xvi

[35] Mahāvaṁsa 37.216ff

[36] Mahāvaṁsa 37.227

[37] I-Tsing, 9-10

[38] My thanks to Rupert Gethin for this information.

[39] This is in reference to the story of Pukkusāti in the Dhātuvibhaṅga Sutta, who went forth out of faith in the Buddha before formally receiving ordination. Ñāṇasaṁvara also mentions the going forth of Mahāpajāpati, the first nun, as a worthy precedent in this context.

[40] Samānasaṁvāsa, a technical Vinaya term meaning able to perform saṅghakamma together.