Shizuka Sasaki has written a series of articles dealing with the Buddhist sects in the time of Aśoka. His work constitutes a major contribution to the field; particularly valuable are his translations of many otherwise little-known passages from the Chinese texts, especially the Vinayas. He takes a fresh and rigorous approach to the topic, and illuminates many points of interest regarding the nature of schism. In the early papers of the series he establishes his basic theses, then for the later papers he works out the implication of these in the context of some of the most important relevant texts; in addition, he addresses the criticisms that were made of his earlier papers by some other scholars. The whole body of work thus becomes quite complex, and I will only venture here to discuss a few of his more basic propositions.
Sasaki’s key passage is taken from the Mahāsaṅghika Vinaya (in Chinese translation). This tells of a ‘powerful upāsaka’ who may intervene if a schism seems imminent. Sasaki points out that this episode, which has no counterpart in any of the other Vinayas, bears a striking affinity with the Aśokan so-called ‘Schism Edict’ (more aptly called the ‘Unity Edict’). The key point of correspondence is that the powerful layperson, when the threat of schism is looming, should make the recalcitrant monks wear lay clothes. This act can be seen from two rather different perspectives: either it is a shameful punishment, or it is a benevolent guarantee of social support for ex-monks. Sasaki argues that it is the latter, but it seems to me that the act would have had a painful ambiguity from the start.
Sasaki argues that the ‘powerful upāsaka’ is none other than Aśoka himself. Personally, I find his arguments plausible, but it needs to be remembered that this is, at best, one hypothesis. It is quite possible that the ‘powerful upāsaka’ referred to has nothing to do with Aśoka. Perhaps it was some other layperson, or an earlier or later king. Of course, we have no archaeological records of any other people acting in this way, but that does not tell us much, since we have no such records of anything for this period. The kings, at least in theory, acted from a sense of Dharma, which meant that their choices were governed by timeless principles, one of which was the support of all religions. The acts threatened (or promised) by Aśoka in the Unity Edicts may well have sprung from his sense of the fit acts of a king in protection of religion, and hence he may have been following an ancient precedent. Similarly, once Aśoka had made his act so public, later kings may have taken this as a precedent and acted similarly; we know the Sri Lankan kings did so. We should also note that the Mahāsaṅghika Vinaya passage is purely hypothetical, and so we cannot be sure that it refers to an actual event (of course, the Edicts also phrase the expulsion of monks as a hypothetical; only the Sinhalese Commentaries and Chronicles refer to it as actually occurring).
Sasaki  downplays any connection between these passages and the Sinhalese accounts, following Bechert’s analysis (which I have discussed elsewhere) that ‘Schism Edict’ is not related to the Third Council. True, they do show matters from somewhat different perspectives: but then so do newspaper accounts of the current US election campaign, and we don’t doubt that they refer to the same thing. We have only a few fragmentary accounts, from widely differing sources, of events happening over 2000 years ago, and the remarkable thing is not that they disagree, but that we can find any coherence in them at all. Of course, we can never prove the various accounts refer to the same historical events, but it seems to me the compassionate way to treat our sources is to acknowledge points of correspondence and not expect total agreement. In fact, the existence of different perspectives enables us to develop a richer appreciation of the events.
One of the reasons Sasaki discounts the connection between the expulsion of monks described in the Edicts/Mahāsaṅghika Vinaya and the Sinhalese sources is that he adopts a most curious interpretive assumption.
For the life of me I cannot imagine why anyone would adopt such a perverse, crippling assumption. What on earth does the mention of the name of an occurrence have to do with the potential for modifying the account? Any account can be modified; or more realistically, every account has its own perspective, which will influence the presentation of the facts. This is a basic assumption of any serious textual work, and the pretence of ‘objectivity’ is a luxury we can no longer afford. Our task is rather to make conscious the kinds of perspectives that are colouring any particular presentation: the adoption of a simplistic assumption can never relieve us of this necessity. In any case, this means that Sasaki cannot, of course, rely on any of the traditional accounts of Aśoka for his reconstruction of the Aśokan period, and in fact must rely on the Mahāsaṅghika Vinaya account, not despite, but because it fails to mention Aśoka by name. It hardly needs commenting that anyone who does not share Sasaki’s assumption need not accept his conclusions.
Sasaki assumes, as do all the scholars I have read, that at least some of the schools emerged before Aśoka, and therefore takes it for granted that we are dealing with a sectarian period (as the title of his articles makes plain). His analysis of the Mahāsaṅghika Vinaya account would seem to reinforce that hypothesis, for here we have an important statement on schism that is not shared among the schools. But closer examination reveals that this is not so. Firstly, as we have seen the actual links with the Aśokan period are hypothetical. Secondly, no actual schism is mentioned: the intervention of the powerful upāsaka is but one stage in the process leading towards schism, and if, as the Edicts assert, the intervention is successful, no schism occurs. Thirdly, the absence of this passage from the other (Sthavira) Vinayas is adequately explained on text-critical grounds: Sasaki, following Frauwallner (and contra Prebish), argues convincingly that the Mahāsaṅghika Vinaya was subject to a major later rewrite. Obviously any material referring to Aśoka would have to be part of such a rewrite, and this reminds us of the startling fact that the Sthavira Vinayas nowhere mention Aśoka. Hence the most plausible explanation for the inclusion of this passage is simply that the Mahāsaṅghika Vinaya was redacted later. And this leads on to the fourth, and crucial, point: all versions of these events (Edicts, Mahāsaṅghika Vinaya, Sinhalese sources) take the point of view of the Sangha that remains after the bad monks are expelled. These are the accounts of the survivors of a purge, not the opposing accounts of two groups who have split apart. In this, these accounts may be compared to the congruent accounts in all Vinayas of the Second Council events, which contradicts the assertion of the Dīpavaṁsa that the defeated party reformed and made the Mahāsaṅghikas. This contrasts strongly with the actual accounts of the schism, which clearly presents contradictory perspectives, as one would expect since they derive from different parties. If in fact, as I believe, all three of these sources are referring to the same events, this is strong evidence that there was no schism at this time.
A further central plank of Sasaki’s thesis is the differences between saṅghabheda as described in the various Vinayas. Sasaki  starts by comparing the Mahāsaṅghika Vinaya with the Sarvāstivāda Vinaya. In both cases, as in the Vinayas generally, the discussion of schism is centred around Devadatta.
The Mahāsaṅghika Vinaya characteristically omits the narrative background to Devadatta’s story and even gives instructions to look elsewhere for this. This is a typical sign of the removal of narrative portions from the Vinaya as part of its later redaction. The episode picks up with the Buddha instructing Ānanda to go to Devadatta and invite him for the uposatha. Devadatta refuses, renouncing the Buddha, Dhamma, and Sangha, and asserting that he will study the rules according to his will. He and his followers then perform uposatha separately, with the deliberate intention to cause a schism.
The Sarvāstivāda Vinaya describes Devadatta causing schism by stating his ‘5 points’ (a monk should wear discarded robes, live on alms, eat one meal only, live outside, and not eat fish or meat) three times in the midst of the Sangha, inviting those who wish to adhere to these rules joyfully to stand up and take the tally-stick. At the end he says: ‘Those who do not delight in these rules and those who cannot keep these codes should go far away from us. Those people must live separately and should not share words with us.’ Sasaki further quotes a definition in the Sarvāstivāda Vinaya which defines schism as caused by one of the 14 causes, such as describing what is not Dhamma as Dhamma [etc.], taking a tally-stick and making a speech.
Sasaki claims that the Mahāsaṅghika definition revolves around performance of uposatha while the Sarvāstivāda definition ‘is not related to uposatha at all’. ‘If we accept the definition in the Sarvāstivāda Vinaya, it is impossible for monks who adhere to different doctrines to reside together. However, according to the definition of saṅghabheda in the Mahāsaṅghika Vinaya, all the monks can still reside together while holding different doctrines. Therefore, the difference in the definition in these two texts greatly influenced the management of the Sangha.’ [2.168-9]
Unfortunately, this conclusion in untenable on a number of grounds. The first and most obvious is that the 5 points in the Sarvāstivāda Vinaya have nothing to do with doctrine. They are five points of monastic conduct, i.e., Vinaya. Devadatta does not ask his followers to accept his doctrine, he asks them to follow his rules. Thus he is setting up a new community based around a separate disciplinary code, and hence asserts the impossibility of living together, i.e. in the same communion (samānasaṁvāsa). Crucially, he then says they should not ‘share words with us’. The phrase translated by Sasaki as ‘share words’ (共語) presumably does not mean the followers of Devadatta cannot speak with the Buddha’s followers at all. It must, in fact, be referring to the concept of ‘unified recital’ (ekuddesa), i.e., that all monks should recite their disciplinary code together. Thus the basic Sarvāstivāda Vinaya passage on Devadatta refers to a split on grounds of Vinaya, not Dhamma, and it clearly invokes the notion of sharing a common dwelling and a common recital, the key factors in a successful upostha.
Sasaki appears to want to derive further support for his notion that for the Sarvāstivādins schism is a matter of Dhamma by his manner of quoting the reference to the ‘14 causes’. This is a well known list (the exact details vary; Pali sources usually have 18), and includes referring to what is not Dhamma as Dhamma, what is not Vinaya as Vinaya, etc. The text here abbreviates the well-known long list and only presents the first pair, misrepresenting Dhamma as not-Dhamma, and not-Dhamma as Dhamma. Of course this is fine as far as an abbreviated translation goes, but it obviously does not in any way imply that schism is not also caused by Vinaya.
Sasaki himself quotes from the Sarvāstivāda Vinaya (or from what he calls a supplement to it) a passage that accepts the definition of schism as performing separate uposathas. His final paper adds a second passage in a supplementary part of the Sarvāstivāda Vinaya that also mentions the two parts. And of course this definition became standard in the school. This would seem to be simply making explicit the perspective that is embodied in the original passage on Devadatta, and hence we are not able to accept Sasaki’s thesis that the Vinayas themselves embody disparate takes on the notion of schism.
It is hard to draw conclusions from the Mahāsaṅghika Vinaya passage, since it itself claims to have abbreviated the relevant passage. Thus the absence of the story of the ‘5 points’, etc., may simply be an artefact of the redaction history of this text, which Sasaki persuasively argues was redacted later. And if we compare the abbreviated text of the Mahāsaṅghika Vinaya account with the fuller versions found elsewhere (as the text itself instructs us to do), we find that all the events in the story of Devadatta, including the intention to split the Sangha based on the ‘5 points’, leads up to Devadatta travelling with his schismatic party to Gayā, exactly as mentioned in the Mahāsaṅghika Vinaya account (e.g. Pali Vinaya 2.199, Dharmaguptaka Vinaya T. 1428, p909b; see translation Sasaki 3.179). We are thus perfectly justified to conclude that the Mahāsaṅghika Vinaya indeed should include this passage.
Sasaki continues to argue  that the Mahāsaṅghika Vinaya denies that a difference in the Dhamma is not a cause of disruption; but this is again not justified. The text only states that such disagreement over the Dhamma does not itself constitute saṅghabheda, not that it is not a cause of saṅghabheda; in this it agrees with the other accounts. The text then goes on to assert that the schism occurs when the different groups are residing within the same monastery and perform upostha separately. Again this agrees with the other accounts. The only difference is that the other accounts make it plain that the separate recital follows on from the disagreement over Dhamma, Vinaya, etc. The Mahāsaṅghika account does not make this explicit, but this appears to be merely a quirk in its phrasing. Why else would the two groups conduct a separate recital? Obviously they must disagree about some important matters, and such disagreement as to Dhamma, Vinaya, etc., must proceed the separate recital.
Our texts are clearly distinguishing between the cause of schism and the point at which schism legally occurs. But this seems to have been lost on Sasaki. Intent on drawing major historical theses out of little variations in the letter of legal proceedings, he does not consider how these actually work in real life. Any schism must be the result of a long, complex process, as indeed depicted in the canonical accounts of both the problems with Devadatta and those at Kosambi. There is a gradual sense of alienation and separation, which may come about because of evil wishes (as with Devadatta), mere foolish stubbornness (as at Kosambi), or because of genuine differences in belief and practice (as, it would seem, in the emergence of the later schools). On a human and spiritual level, any community is negotiating such issues constantly, and they rarely become serious enough to threaten a schism. But in rare cases the conditions approach extremes, and the threat of schism looms. The warning signs appear long before, and there are usually plenty of chances to turn back. But they are ignored and the schismatics stubbornly persist. Up to this point the various causes of schism, such as serious disagreements on Dhamma and Vinaya, have been growing in strength. But at some point, the decision must be made. The new order is distinctively established in separation from the old. It is at this point, the distinct legal marker for the creation of a schismatic community, that the definition of schism as performance of separate uposathas becomes relevant.
Sasaki alleges a contradiction in the Pali account, since it defines schism in terms of the performance of separate uposathas, but in the story of Devadatta he says there is no mention of the schismatic group carrying out separate uposatha. [3.178] This is slightly surreal, since on the previous page Sasaki has translated the passage where Devadatta, ‘on the uposatha’ (Pali Vinaya 2.199: tadah’uposathe) pronounces his new disciplinary code – the 5 points – and has had his followers take the tally stick to announce their support. Obviously the account is straightforwardly a description of the uposatha ceremony carried out separately. The account is in the formalistic style of the typical saṅghakammas (e.g.: ‘yass’āyasmato imāni pañca vatthūni khamanti, so salākaṁ gaṇhatū’ti)
There is a similar interpretive error in the description of the Dharmaguptaka Vinaya, which Sasaki acknowledges in his final paper. The text gives two sets of things that can cause schism. The first is ‘false words and similar words’; this is cryptic, but perhaps should be connected with the 14 (or 18) causes, which include representing what was not said by the Buddha as said by the Buddha, etc. The next set is ‘performing a kamma and taking a tally stick’. This obviously agrees with the procedure that we have seen for the schismatic acts of Devadatta in each of the Vinayas, but Sasaki unaccountably dismisses this, saying without attempt at justification that: ‘This karman does not mean the rituals but declaring one’s opinion in front of a Sangha’ [3.178] But performing a kamma and taking tally sticks is exactly what Devadatta does in the passage translated immediately below by Sasaki. Thus the Pali and Dharmaguptaka are not exact opposites as Sasaki asserts, but in fact give perfectly compatible accounts, both acknowledging the importance of genuine differences in understanding and conduct, leading up to the decisive moment of performing the ritual separation.
The discussion of an apparent interpolation in the Pali Vinaya [3.186] is interesting, but in view of our previous observations, hardly convincing of Sasaki’s conclusions that it shows that kammabheda was a later development. The interpolated passage, Sasaki fails to note, is retained elsewhere in the Pali Udāna as a separate sutta (Udāna 5.8). The form of the sutta alone is enough to alert us to its separate provenance, although Sasaki’s information that the passage is not found in other Vinayas, and that it links two passages that are elsewhere separated, is useful. But once more it is hard to draw conclusions from this: the Udāna is a work that evidently harks back to an early period, but which is not found in similar form outside the Pali canon, so we cannot do a comparative study. It is therefore difficult to determine whether the passage was a late invention, or whether the textual situation in the Pali is simply a later stitching together of earlier patches. One difference between the Sutta and the Vinaya versions is that the Sutta explicitly states that it is the uposatha day, whereas in the Vinaya this is absent; however, the next section is the passage where Devadatta makes his announcement ‘on the uposatha’. In any case, as I have shown above, the acts of Devadatta obviously do consist of a separate uposatha.
Sasaki’s discussion of āvāsakappa in the Second Council (pañcasatikakkhandhaka) is similarly confused [3.192]. He unclearly describes āvāsakappa as ‘monks in a Sangha perform uposatha separately’, while the Vinaya itself puts the matter much more clearly: ‘several monasteries within the same boundary (sīmā) perform uposatha separately.’ (Pali Vinaya 2.300) Sasaki then says: ‘In the pañcasatikakkhandhaka it is denied that āvāsakappa is as an evil act (dukkata) that is opposed to the Vinaya.’ Presumably this is merely a slip in wording, for the pañcasatikakkhandhaka, of course, says that āvāsakappa is a wrong-doing (dukkata) that oversteps Vinaya, the exact opposite of what Sasaki has written. Sasaki then says that such an act qualifies as ‘disruption’ (= saṅghabheda), which is a serious matter, but is dismissed here as a light transgression. Thus this passage contradicts the definition of saṅghabheda as performance of separate uposathas. Once again Sasaki insists on a narrow interpretation, ignoring the real life context. The Vinaya goes to great length to show that saṅghabheda emerges from a complex process and is not accomplished merely by a technicality. (e.g. Pali Vinaya 2.204) If the separate uposathas are performed as the outcome of a process where the dissenting group, deliberately misrepresenting the Dhamma and Vinaya and aiming at schism, schism results. What is referred to as āvāsakappa is merely a procedural flaw, where groups of monks in different monasteries, presumably because of laziness, don’t want to bother walking to a nearby monastery to do uposatha together. It has nothing to do with any attempt to cause schism.
We are thus left at the end of Sasaki’s immense and praiseworthy efforts feeling that we hadn’t quite got what we hoped. There is, of course, much that I have omitted that is of great value; I would particularly mention the detailed analysis of the Mahāsaṅghika Khandhakas. But I cannot accept some of his primary hypotheses. The episode describing involvement of a ‘powerful layperson’ in the schismatic process falls far short of establishing any connection between a schism and Aśoka, and in fact is perfectly in tune with the statements in the Edicts and the Sinhalese sources that schism did not take place at that time. Sasaki’s attempt to trace the influence of schism in the different versions of the Vinaya is an essential endeavour, and one which needs careful attention. But Sasaki’s mistakes, both in details and principles, undermine his attempt to stratify the accounts of schism into cakkabheda and kammabheda. This renders much of his subsequent analysis – a complicated working out of the implications of flawed hypotheses – of marginal value. It is still possible that the various sectarian accounts of schism may yield variations of significance, but this needs a complete re-examination.
The accounts of schism in the Vinaya present themselves as based on events that took place in the life of the Buddha. It is, however, quite plausible that these accounts should be revised at a later date to reflect the perspectives around the actual historical schisms. However, Sasaki’s account does not convince me that this has actually taken place. Rather it seems to me that the accounts, despite undeniable variations in the wording, present fairly similar perspectives on schism. If it is true that there are no great differences, this would seem to suggest that the accounts were reasonably settled before the schisms took place. That is, the accounts really are about the failed schism of Devadatta, and not about the historical separation of the schools. This agreement in the content contrasts with the obvious divergences in the literary forms of the texts. I would explain this in terms of the essentially geographical emergence of the schools. The different communities were separated in space before they were separated by doctrine. The relative geographical isolation allowed the texts to diverge in form, while not embodying any major disagreements in doctrine. Only at the very end of the process of the formation of the early texts (both Suttas and Vinayas) can a few hints of sectarian perspectives be discerned creeping in. In the case of the schisms, this is not the variations that Sasaki speaks of, but the divergent perspectives on wet dreams, the key schismatic issue as presented by Vasumitra, Bhavya III, etc.