Norman - Pāli Literature - A History Of Indian Literature (1983)

Pāli Literature

A History Of Indian Literature

Edited By Jan Gonda

VOLUME VII
Fasc. 2

K. R. Norman

Including The Canonical Literature In Prakrit And Sanskrit Of All The Hīnayāna Schools Of Buddhism

Contents of Vol. VII

Vol. VII: Buddhist and Jaina Literature

• Fasc. 1: D. Seyfort Ruegg The Literature of the Madhyamaka School of Philosophy in India
• Fasc. 2: K. R. Norman Pâli Literature

CIP-Kurztitetsufnahme der Deutschen Bibliothe
A history of Indian literature / ed. by Jan Gonda.
Wiesbaden : Harrassowitz
NE: Gonda, Jan [Hrsg.
Vol. 7. Buddhist and Jaina literature. Fasc. 2. → Norman, Kenneth Roy : Pāli literatur
Norman, Kenneth Roy
Pâli literature : incl. the canon. literature in Prakrit and Sanskrit of all the Hinayāna schools of Buddhism
Kenneth Roy Norman. - Wiesbaden : Harrassowitz, 1983
(A history of Indian literature ; Vol. 7, Fasc. 2
ISBN 3-447-02285-X

© Otto Harrassowitz, Wiesbaden 1983. Alle Rechte vorbehalten. Photographische und photomechanische Wiedergabe nur mit ausdrücklicher Genehmigung des Verlages. Gesamtherstellung: Allgāuer Zeitungsverlag GmbH. Kempten. Printed in Germany. Sigel: HIL

CONTENTS

• Foreword … IX
• Chapter I The Pāli language and the Theravādin tradition
• 1. The name “Pāli” … 1
• 2. The history and development of the Pāli language … 2
• 3. The history of the Theravādin tradition … 7
• Chapter II The Pāli Canon
• 4. The divisions of the canon … 15
• 5. The Vinaya-piṭaka … 18
• 2.1. The Suttavibhanga … 18
• 2.2. The Khandhaka … 22
• 2.3. The Parivāra … 26
• 6. The Sutta-piṭaka … 30
• 3.1. The Dīgha-nikāya … 32
• 3.1.1. The Silakkhandhavagga … 32
• 3.1.2. The Mahāvagga … 36
• 3.1.3. The Pāṭikavagga … 41
• 3.2. The Majjhima-nikāya … 44
• 3.2.1. The Mūlapaṇnāsa … 45
• 3.2.2. The Majjhimapaṇnāsa … 46
• 3.2.3. The Uparipanṇāsa … 48
• 3.3. The Samyutta-nikāya … 49
• 3.3.1. The Sagāthavagga … 50
• 3.3.2. The Nidānavagga … 52
• 3.3.3. The Khandhavagga … 52
• 3.3.4. The Salāyatanavagga … 53
• 3.3.5. The Mahāvagga … 53
• 3.4. The Anguttara-nikāya … 54
• 3.5. The Khuddaka-nikāya … 57
• 3.5.1. Khuddakapāṭha … 57
• 3.5.2. Dhammapada … 58
• 3.5.3. Udāna … 60
• 3.5.4. Itivuttaka … 62
• 3.5.5. Suttanipāta … 63
• 3.5.6. Vimānavatthu … 70
• 3.5.7. Petavatthu … 71
• 3.5.8. Theragāthā … 72
• 3.5.9. Therīgāthā … 75
• 3.5.10. Jātaka … 77
• 3.5.11. Niddesa … 84
• 3.5.12. Paṭisambhidāmagga … 87
• 3.5.13. Apadāna … 89
• 3.5.14. Buddhavamsa … 92
• 3.5.15. Cariyāpiṭaka … 94
• 4. The Abhidhamma-piṭaka … 96
• 4.1. Dham masangaṇi … 99
• 4.2. Vibhanga … 100
• 4.3. Dhātukathā … 101
• 4.4. Puggalapaññatti … 102
• 4.5. Kathāvatthu … 103
• 4.6. Yamaka … 105
• 4.7. Paṭṭhāna … 106
• 4.8. The Abhidharma in Sanskrit … 107
• Chapter III Early post-canonical texts
• 1. Early prose texts … 108
• 2. The early chronicles … 114
• 3. The early commentators … 118
• 3.1. Buddhaghosa … 120
• 3.2. Buddhadatta … 130
• 3.3. Mahānāma … 132
• 3.4. Upasena … 133
• 3.5. Dhammapāla … 133
• Chapter IV Later post-canonical texts
• 4. The later chronicles … 138
• 5. The later commentators … 145
• 2.1. Aṭṭhakathās … 145
• 2.2. Tīkās … 148
• 6. Later Abhidhamma texts … 151
• 7. Collections of edifying tales … 153
• 5. Kāvya works … 156
• 6. Prophetic texts … 160
• 7. Medical texts … 162
• 8. Grammatical and lexicographical works … 163
• 9. Poetics and prosody … 167
• 10. Books on discipline … 168
• 11. Boundary texts … 171
• 12. Anthologies … 172
• 13. Cosmology … 174
• 14. Wisdom texts … 175
• 15. Jātaka collections … 177
• 16. Bibliographical works … 179
• Abbreviations … 185
• Glossary … 190
• Index of Authors … 198
• Index of Works … 199
• General Index … 206

FOREWORD

Despite the attempts made by A. K. Warder in his survey of Indian kāvya literature, and by various editors and translators in the introductions to their books, to assess the literary merit of certain Pāli works, it is nevertheless still true to say that for the Indologist Pāli literature means everything that is written in Pāli, irrespective of literary value in the accepted European sense of the word. In the space and time at my disposal, however, it was not practicable to include a mention of everything that has been published, much less include unpublished works which are at present only names in library catalogues and similar reference books. I have therefore omitted references to much of the later commentarial literature, but the biggest deficiency is in the field of South-East Asian Pāli literature. Comparatively little work has been done in this field, and although it would have been possible to extract names of works and authors secondhand from the studies that have been made, I have in general refrained from doing so. The resultant gaps in my work will need to be filled by specialist publications in due course.

Except where there is good reason to the contrary, I have dealt only with those texts which have been edited in the Roman script or been translated into a European language. I was asked to include Hīnayāna Buddhist canonical texts in Sanskrit and Prakrit, and here too I have dealt only with the works which have been published. Rather than deal with them separately, I have mentioned them under the relevant Pāli headings, since the Hīnayāna canons follow closely the pattern of the Theravādin canon, with the exception of the Abhidharma.

The bibliographical material included in the footnotes is intended only as a sample of the editions and translations which are available. With very few exceptions there is no mention of oriental editions, and it is sufficient to state here that all the Pāli canonical texts and some para-canonical ones, together with the atṭhakathās and most of the tīkās, have been published in the Chatṭhasañgāyana series which began to appear in Burma from 1956 onwards, while most of the same texts have also been published in Ceylon. The complete canon and most of the commentaries have been published in Thailand, and editions in the nāgarī script have appeared in India from Nalanda and elsewhere. The bibliographical information I have given can be augmented by reference to the other sources mentioned in the notes, and especially to the Epilegomena to Vol. I of the Critical Pāli Dictionary (Copenhagen 1948). Additional and up-todate information will be available in the new bibliography which is planned for the Epilegomena to Vol. II of that dictionary.

My debt to all others who have worked in this field will be obvious to readers of this book. When Geiger, Winternitz and Law have dealt so ably with the whole field of Pāli literature, and Malalasekera and Bode with Sinhalese and Burmese Pāli literature respectively, it is very difficult to write about the subject without repeating what one or other of these authorities has already said. Rather than do so, and to save space, I have on many occasions referred the reader to their books. In particular, Winternitz is to be recommended for the many quotations he includes, while Law gives lengthy summaries of the contents of many texts.

Although I have occasionally given information about the formation of individual texts, I have not done this consistently throughout, nor have I dealt with any one text at great length. Those who are interested in this aspect of Pāli literature will find references in the footnotes to Pande’s book, in which he deals with the first four nikāyas and some of the Khuddaka-nikāya. The discussion of chronological stratification is a complicated matter, and needs more space than I have at my disposal. It is also, in the absence of better criteria than we have at present, a very subjective matter, and while I would agree with much that Pande says, I cannot agree with him in everything.

Cambridge, 1 February 1982
K. R. Norman

CHAPTER I - THE PĀLI LANGUAGE AND THE THERAVĀDIN TRADITION

1. THE NAME “PĀLI”

The dialect of Middle Indo-Aryan which is found in the texts of the Theravādin Buddhists and usually called “Pāli” by European scholars ¹ is nowhere so called in the Theravādin canon. The word pāli is found in the chronicles and the commentaries upon the canon, but there it has the meaning “canon” and is used in the sense of a canonical text or phrase as opposed to the commentary (atthakathā ) upon it. ² This usage is made clear by the fact that the word pāli sometimes alternates with tanti. ³

It would seem that the name “Pāli” is based upon a misunderstanding of the compound pāli-bhāsā “language of the canon,” where the word pāli was taken to stand for the name of a particular bhāsā, as a result of which the word was applied to the language of both canon and commentaries. There is evidence that this misunderstanding occurred several centuries ago. Childers stated ⁴ that the English usage was taken from the Sinhalese, ⁵ who used the word in the same way. This probably accounts for Clough’s adoption of the name when he published his grammar in 1824. ⁶ Burnouf and Lassen also used the name “Pāli” in their essay on Pāli grammar which was published in 1826, ⁷ but in the survey of Pāli studies up to that year included in that work Burnouf pointed out that the first person to mention Pāli was Simon de La Loubère who visited Siam in 1687-88, and published a description of the kingdom of Siam in 1691, which was translated into English in 1693. ⁸

It is clear from this account that in Thailand in the late seventeenth century the name “Pāli” was already being used of the language of the Theravādin texts. La Loubère noted that in contrast to Thai, which was a monosyllabic language, “Balie” (or “Baly”) was inflected just like the languages of Europe. ⁹ He also drew attention to the fact that the names for the days of the week were similar in Pāli and Sanskrit, ¹⁰ and reported that he had been told that there were similarities between Pāli and the languages spoken near Coromandel. He commented that this was not surprising in view of the fact that the Buddha was reported to have been the son of a king of Ceylon. ¹¹

The Sāsanavaṃsa, written in Burma in 1861, uses the word pāli in a context where it seems to be the name of a language. ¹² Since the Sāsanavaṃsa is based upon an earlier Burmese text, ¹³ the usage of the name “Pāli” in Burma is probably earlier than would appear. It seems unlikely that the usage arose independently in all three countries, but in the present state of our knowledge it does not seem possible to determine where the misunderstanding first occurred.

2. THE HISTORY AND DEVELOPMENT OF THE PĀLI LANGUAGE

The commentaries state that the language spoken by the Buddha, which is the language of the canon, is Māgadhī. ¹⁴ This is referred to as the mūla-bhāsā¹⁵ the root language of all languages, and the language which a child would speak naturally if it heard no other language spoken. ¹⁶

An examination of the Pāli canon shows clearly that portions, at least, of it were either composed or transmitted through one or more other dialects of Middle Indo-Aryan, before being turned into the version which exists at present. It can be shown that these dialects included those where the voicing of intervocalic consonants took place, or their reduction to -y-, or -r - became -l -..¹⁷ Some of the Pāli material came from or through dialects where the absolutive was in -ttā, ¹⁸ the nominative singular in -e, or the locative plural in -ehi. ¹⁹ It is clear, therefore, that the statement that the canon is in one dialect, whether Māgadhī or anything else, cannot be true of all of it. What we know of Māgadhī as described by the grammarians in later times, however, enables us to say that Pāli is not Māgadhī, and although we have no direct evidence about the characteristics of Māgadhī in the centuries before Asoka, we can deduce with some certainty that Pāli does not agree with that either. ²⁰

It would seem likely that, because the texts tell of the Buddha at times preaching in Magadha, although none of the scenes of the great events in his life was situated within the boundaries of Magadha as we know it in historical times, ²¹ the tradition arose that all his sermons were preached in the dialect of that region of North India. It is also possible that the prestige attaching to Magadha, and by implication to Māgadhī, during the time of the Mauryan kings, and also the way in which the Māgadhī ²² of the original Asokan edicts was everywhere in India “translated” into the local dialect or language, led to the taking over by the Buddhists, at about the time of the council which the Theravādin tradition reports was held during the reign of Asoka, of the idea that their “ruler” too employed such a language.

Although there is some doubt about the interpretation of the phrase the Buddha used when asked if it was permissible to translate his sermons, ²³ it is generally agreed that he did not preach in Sanskrit, but employed the dialect or language of the area where he was preaching. We must assume that his sermons and utterances were remembered by his followers and his audiences as they heard them. In the course of time, during his lifetime and after his death, collections must have been made of his words, and translations or redactions of these must have been made as the need arose, either because the collections were being taken into an area where a different dialect or language was spoken, or because as time went by his words became less intelligible as their language became more archaic.

As Buddhism became established in various parts of North India, there must have been an attempt made to render all the holdings of any particular vihāra, which were probably still in various dialects as they had been remembered, roughly homogeneous in language, although we must bear in mind the fact that, as the dialects of North India had probably not diverged greatly from each other in the fourth and third centuries B.C., absolute perfection of “translation” was not essential. The anomalous forms in Pāli mentioned above probably represent the remnants of recensions in other dialects, which had not been completely translated.

The Theravādin tradition tells of councils (sangītis) being held to recite the canon, of which the third was held in the time of Asoka, ²⁴ and although the discrepancies with the Northern tradition cast doubt upon this, there must have been gatherings of some sort where recitations took place, and the “imprimatur” of the Sangha was bestowed. Such councils would inevitably have led to a normalisation of the language of the canon to a greater or less extent.

Since this normalised language was an “ecclesiastical” one, being recited by monks who probably spoke a variety of languages or dialects, there is no necessity to assume that it coincided exactly with any one particular spoken language. It has been claimed in the case of Pāli that as there are resemblances between it and the Girnār dialect of the Asokan inscriptions, ²⁵ and also between it and the language of the Hāthigumphā inscriptions, ²⁶ Pāli must have been the language of one or other of these two areas. A careful examination of the language of these inscriptions shows that Pāli is not identical with either of them, and there is, moreover, some doubt about the language of the Girnār version of the Asokan inscriptions, since it is possible that it represents, in part at least, the scribe’s attempt to convert the Eastern dialect he must have received from Pāṭaliputra into what he thought was appropriate to the region in which the edict was being promulgated, rather than the actual dialect of that region. ²⁷ The language of the Hāthigumphā inscription, although it agrees with Pāli in the retention of most intervocalic consonants and in the nominative singular in -o, nevertheless differs in that the absolutive ending is -(t)tā, and with two doubtful exceptions there are no consonant groups containing -r-.²⁸

While it is not impossible that there existed in India in the third century B.C. an unattested dialect of Middle Indo-Aryan which had all the features of Pāli, the fact that some of the consonant clusters found in Pāli are unhistoric and must therefore represent incorrect attempts at backformation, e.g. disvā (which cannot be from dṛṣṭvā) and atraja (which cannot be from ātmaja), makes it more likely that by the third century B.C. the dialect of the canonical texts of the Theravādins conformed to the general pattern of Middle Indo-Aryan dialects of that time, and all consonant clusters had either been assimilated or resolved. It is probable that this represented the form of the language of the Theravādin canon at the time of the reign of Asoka, which was perhaps the lingua franca of the Buddhists of Eastern India, and not very different from the language of the Hāthigumphā inscriptions.

At some unknown date, probably around the end of Asoka’s reign (c. 235 B.C.), the importance of Sanskrit which had been in eclipse began to rise again, and as we can see from the progressive Sanskritisation of the Mathurā inscriptions ²⁹ and the non-Pāli schools of Buddhism, an attempt was made to translate from the Middle Indo-Aryan dialects into Sanskrit. It is probable that this would have been done in a haphazard way at first, in the same way as the early texts had been remembered and translated from dialect to dialect. With the crystallisation of sects and schools, and with the increased use of writing from the time of Asoka onwards, it is probable that the use of Sanskritised forms by the Theravādins became more standardised.

The tradition recorded in the Sinhalese chronicles states that the Theravādin canon was written down ³⁰ during the first century B.C. as a result of threats to the Sangha from famine, war, and the growing power of the Abhayagiri vihāra, to which the king was more favourably disposed. There is no reason to reject this tradition, because there are indications that texts were already being written down before this date. ³¹ It seems probable that the Sanskritisation of Pāli was virtually fixed at the stage it had reached by the time of the commission to writing, and except for any changes made later by the scribes for the sake of consistency, no further progress was made with the restoration of consonant groups. Those changes which had been made, in a haphazard way as already stated, had probably occurred for specific religious or cultural reasons. It is, for example, probable that the authentic Middle Indo-Aryan form bamhana “brahman” had been replaced by the Sanskrit form brāhmana because of the strongly anti-brahmanical flavour of the Buddha’s teaching. It was perhaps of little value to condemn the brahmans or to maintain that a Buddhist “brahman” was better than a true-born brahman if the recipients of the attack did not recognise their name in its Middle Indo-Aryan guise. Once the br-group had been restored ³² in brāhmana, then the way was open for its “restoration” in other words, e.g. brüheti, even if it were not historically justified. On the other hand, there was no specific need to restore pr-, even though one might have expected the treatment of voiced and unvoiced labials in conjunction with -r to have been identical.

Although the chances of major changes being made to the language of the canon must have diminished once the texts had been written down, nevertheless there was room for a certain amount of minor emendation to take place in the course of the centuries-long scribal tradition. Once the Pāli grammarians had begun to classify and categorise the features of Pāli, it was inevitable that their writings would have some effect upon scribes. It is clear from some of the later commentaries that the commentators had a knowledge of Sanskrit. ³³ There is evidence that the opinions of the commentators sometimes had a bearing upon the readings which the scribes handed down, ³⁴ and it seems likely that their knowledge might well also affect the phonology and morphology they transmitted. It has been said, with some justification, that the Pāli of the canon as we have it now is a reflection of the Pāli of the twelfth century, ³⁵ when the influence of the Pāli grammarians was at its highest.

It was to be expected that Pāli would show the influence of Sanskrit, because it was copied by scribes who were influenced by the grammarians, who were themselves influenced by Sanskrit, but we must also recognise the fact that as Pāli became the religious language of Buddhism in Ceylon and South-East Asia, it was used by a variety of persons as a second language. This led inevitably to the introduction into Pāli of features of their first language, more particularly of vocabulary, but also of syntax. Studies of Pāli texts written in Ceylon show clearly the influence of Sinhalese Prakrit and Sinhalese, and also Dravidian, ³⁶ probably through the medium of Sinhalese. In the same way texts written in Burma ³⁷ or Thailand ³⁸ may reveal features of Burmese or Thai.

All Hinayāna canonical texts show evidence of being translated from a dialect of Middle Indo-Aryan. The North-Western Prakrit of the Gāndhārī Dharmapada has features, e.g. dental -s- where retroflex -s- is expected, which prove that it has been translated from some other Prakrit. ³⁹ The Sanskrit of the Sarvāstivādin texts from Chinese Turkestan and the Mūla-sarvāstivādin texts from Gilgit shows, in general, less Prakrit features than the language of the Mahāsānghika-Lokottaravādin texts, ⁴⁰ although the latter, mostly known only from a single manuscript, are not consistent in their degree of Sanskritisation. Their version of the Dharmapada, ⁴¹ for example, has restored some retroflex and palatal sibilants, many long vowels before consonant groups, and some consonant clusters containing -r-, but very few containing -v-, ⁴² e.g. absolutives have the ending -ttā. The nominative singular ending is -o. In the Bhikṣuṇī-Vinaya, ⁴³ on the other hand, the absolutive ending is - tvā, and the nominative ending -ah occurs.

3. THE HISTORY OF THE THERAVĀDIN TRADITION

All the schools of Buddhism agree that there was a council held soon after the death of the Buddha. According to the earliest version found in Pāli, ⁴⁴ there was a meeting of 500 bhikkhus, where Mahākassapa asked Upāli about the vinaya, questioning him about the rules of the Pātimokkha, where the offence was laid down, with respect to whom, on what subject, etc. He then questioned Ānanda about the dhamma, starting with the Brahmajāla-sutta of the Dīgha-nikāya, and then the Sāmaññaphala-sutta, and so on through the five nikāyas. As the two experts expounded these matters, the other theras repeated their words after them. Buddhaghosa, in his account of the same council, ⁴⁵ states that the fifth nikāya, the Khuddaka-nikāya, included whatever sayings of the Buddha were not included in the first four nikāyas. The Pāli account does not mention the Abhidhamma specifically, although Buddhaghosa’s interpretation of nikāya does not preclude the inclusion of the Abhidhamma in the Khuddaka-nikāya. The versions of the story included in the Dīpavamsa and the Mahāvaṃsa ⁴⁶ state that Ānanda repeated the dhamma, which could be taken to include the Abhidhamma. Some of the Northern schools, however, do specifically include the Abhidhamma, and state that this was expounded by Kassapa himself. ⁴⁷

The version in the Dipavaṃsa ⁴⁸ states that the theras divided up the ninefold dhamma into chapters, etc. This seems to be an attempt to reconcile the old nine-fold division with the arrangement of the canon as we have it now. Buddhaghosa attributes to the theras all the uddāna verses, noting of repetitions, etc., which exist in the canon. ⁴⁹ There are in commentaries by Buddhaghosa and others not infrequently statements that a portion of the text is due to the sangitikāras, and it is probable that these ascriptions are taken over from the earlier Sīhala atthakathās. Although there is no reason to doubt that these portions are indeed later additions to the canon, and were made by sangitikāras, there is no certainty that they were added at the time of the first council.

Although we may have reservations about the texts which were dealt with at the first council, there is no reason to doubt the general way in which it was held. The chosen expert in each section of the Buddha’s teachings expounded what he could remember, and when it had been approved as a genuine utterance of the Buddha the assembly as a whole confirmed their approval by repeating it together. The commentaries explain that the words “Thus have I heard” at the beginning of the suttas are the words which Ānanda used to introduce his recitation. ⁵⁰

Buddhaghosa tells us that after the council the Vinaya was entrusted to Upāli and his pupils. Similarly the Dīgha-, Majjhima-, Saṃyutta- and Anguttara-nikāyas were entrusted to Ānanda, Sāriputta, Mahākassapa and Anuruddha, respectively, and their pupils. ⁵¹ This was probably the beginning of the system of the bhānakas (“reciters”), who shared out the recitation of the dhamma among themselves. We find in the commentaries references to bhānakas of the first four nikāyas, and to Jātaka-bhānakas and Dhammapada-bhānakas. ⁵²

Although there seems to be only one reference to Khuddaka-bhānakas, ⁵³ we may assume that there were bhānakas of the other texts of the Khuddakanikāya. We may deduce from the fact that versions of one and the same sutta or utterance in different parts of the canon sometimes differ, that the bhānakas responsible for the transmission of each text were quite independent, and were not influenced by the traditions of the bhānakas of other nikayas. This would explain why the versions of Vangīsa’s stanzas in the Saṃyutta-nikāya and the Theragāthā do not entirely agree. ⁵⁴

The views of the bhānakas were not confined to the texts for whose recitation they were responsible. They also had views about the books which were to be regarded as canonical. We read that the Dīgha-bhānakas put the Khuddakagantha into the Abhidhammapitaka, classifying the texts in the following order: Jātaka Mahāniddesa Cūlaniddesa Paṭisambhidāmagga Suttanipāta Dhammapada Udāna Itivuttaka Vimānavatthu Petavatthu Theragāthā Therīgāthā. The Majjhima-bhānakas, however, put them in the Suttanta-pitaka together with the Cariyāpiṭaka, the Apadāna and the Buddhavaṃsa. ⁵⁵ They thus omitted the Khuddakapāṭha from the canon. We must assume that the Dīgha-bhānakas closed their list of the Khuddaka-nikāya before these texts were added to it, while the Majjhima-bhānakas closed their list before the Khuddakapāṭha was reckoned as being canonical.

The schools of bhānakas also had their own views about the history of early Buddhism, e.g. the Dīgha-bhānakas said that the Bodhisatta saw all four nimittas on the same day, ⁵⁶ while other bhāna kas had other views; the Majjhimabhānakas gave a different reason from the Digha-bhānakas for Ānanda’s late arrival at the Council. ⁵⁷ To judge from the way that Buddhaghosa speaks about them, the bhānakas were still active in his day, but they seem not to exist today, at least not in Ceylon. ⁵⁸ It is likely that the growth of the practice of writing, and the increasing use of MSS, made it unnecessary to keep up the schools which were dedicated to the recitation of the complete nikayas entrusted to their care.

The Pāli Vinaya-piṭaka goes on to relate how, following a dispute over certain points of discipline 100 years after the death of the Buddha, a second council was held. In the presence of 700 bhikkhus, Revata asked the thera Sabbakāmin about the discipline. This council is described in the Vinaya as a vinaya-sangiti. ⁵⁹ The Dīpavaṃsa calls it a dhammasangaha, and is followed in this by the Mahāvaṃsa. ⁶⁰ The latter source adds that the 700 were tipitakatta-yadharins. ⁶¹ Buddhaghosa expands the story, and says that 700 tipitakadharas were chosen to rehearse the dhamma and the vinaya, which they recited in their entireties. ⁶²

At some time after the second council, the dissident monks split off from the Theravādins and held their own “great council” (mahāsangiti). They consequently became known as Mahāsangītikas ⁶³ or Mahāsānghikas. ⁶⁴ During the following centuries further splits occurred, from both the Theravādins and the Mahāsānghikas, until by the time of the early chronicles eighteen sects were known and named. ⁶⁵

The earliest Pāli source for the account of the third council, held during the reign of Asoka and not mentioned as such in the Northern Buddhist sources, is the Dīpavaṃsa. ⁶⁶ There we read that the council was held after the expulsion of certain heretics from the Order. Moggaliputta presided over 1,000 bhikkhus, and during the course of the sangiti he recited the Kathāvatthu. The Mahāvaṃsa ⁶⁷ adds the detail that the bhikkhus were all tepitakas. Buddhaghosa states ⁶⁸ that they recited both the dhamma and the vinaya.

Buddhaghosa gives an account ⁶⁹ of an assembly held in Ceylon, soon after the introduction of Buddhism there, when the elder Mahā-ariṭtha preached the vinaya. The Mahāvaṃsa does not mention this assembly, but in the account of it given in the Saddhammasangaha, ⁷⁰ the recitation is said to have included both the vinaya and the dhamma, and the assembly is specifically called the fourth assembly, and verses attributed to the porannas are quoted which include the words catuttham sangaham katvā. ⁷¹ The Jinakālamālī ⁷² follows Buddhaghosa in stating that Mahā-ariṭtha recited the vinaya only. The verse found in the Saddhammasangaha is quoted by Buddhaghosa, ⁷³ without attribution to the porannas, but the words of the pāda in question are different, and make no mention of the catutthasangaha.

The Dīpavaṃsa states ⁷⁴ that during the reign of Vatṭagāmaṇī Abhaya (29-17 B.C.) the monks who had previously remembered the Tipitaka and its commentary orally now wrote them down in books, because of the threat posed by famine, war, and the growing power of the newly established Abhayagiri-vihāra, which enjoyed the king’s favour. The Mahāvaṃsa ⁷⁵ also refers briefly to the writing down of the canon and the commentaries at this time. The Pūjāvaliya and Nikāyasanggraha, [^1-76] however, written in Ceylon in the 13th and 14th centuries respectively, state that the writing down was a result of the holding of a council of 500 bhikkhus at the Aluvihāra (Āloka-vihāra), although neither source gives a number to the council. ⁷⁶ The Jinakālamālī, ⁷⁷ however, entitles this section of its narrative Catutthasangitikathā, and quotes a statement from the Sāra-sangaha, written in the 13th or 14th century, that the sangīti was catutthadhammasangītisadisā. The Saddhammasangaha does not specifically call this the fifth council, as it might have been expected to do, but states: pañcamam dhammasangitisadisam eva akāsi. ⁷⁸ Of the writing down of the Tipiṭaka and commentaries in the time of Vatṭagāmaṇī, the Sāsanavaṃsa, written in Burma in 1861 by Paññāsāmi, but following earlier sources, states ⁷⁹ that “this should be called even by the name of the fourth council.” It quotes as an authority for this statement the Vinaya-ṭikā called Sāratthadīpanī, written by Sāriputta ⁸⁰ probably in the 12th century, which states “The council for writing the books was indeed the fourth council.” The Sāsanavaṃsa also states ⁸¹ that the list of elders down to Siva which is found in the Vinaya ⁸² is the “succession of elders down to those who assembled in the fourth council as recorded in books.”

The Dīpavaṃsa, followed by Buddhaghosa and the Mahāvaṃsa, ⁸³ tells how one of the missions sent out by Moggaliputta after the third council consisted of Sona(ka) and Uttara who went to Suvaṇnabhūmi, which is usually identified as Burma. The Burmese tradition, however, has it that Buddhism was established in Burma even earlier than that. The Sāsanavaṃsa ⁸⁴ tells of the Buddha actually visiting that country, a story doubtless invented to match his alleged visits to Ceylon. The later Burmese tradition even claims Buddhaghosa as a native of Thatôn. ⁸⁵

There is evidence for the existence of Buddhism in both Burma and Siam by the seventh century A.D., and a number of Buddhist texts were certainly known in Burma in the eighth century, ⁸⁶ although it is not certain whether the whole Tipiṭaka was known there at that time. The Sāsanavaṃsa ⁸⁷ refers to inscriptions stating that King Anuruddha established the religion c. A. D. 1057 after conquering the town of Sudhamma (Thatōn). He had the Tipiṭaka brought from Ceylon and compared with that from Sudhamma. Communication with Ceylon at this time is confirmed by a report in the Mahāvamsa ⁸⁸ that King Vijayabāhu, finding that there were not enough ordained monks in Ceylon to hold the upasampadā ceremony, asked Anuruddha to send monks to hold the ceremony, and to recite the Tipiṭaka.

In the following century, when Parakkamabāhu I became king, he found that Buddhism in Ceylon was still rent by heresy. He accordingly called a council under Mahākassapa and had them settle the points in dispute, and thus managed to reconcile the three main sects in Ceylon at that time. ⁸⁹

We read in the Jinakālamālī ⁹⁰ that in A.D. 1430 the Sīhalasāsana was brought to Siam, and not long after, in 1475-77, a council was held in Siam ⁹¹ at which the three Piṭakas were “cleansed of scribes’ errors” by great elders appointed for the task, who were well versed in the Tipiṭaka. As in the case of Burma, the traffic was not all one way. In 1756 Buddhism had declined to such a state in Ceylon that it became necessary to re-establish the upasampadā, and monks were invited from Siam to do this. ⁹² They brought with them books of various sorts, on the dhamma and the vinaya, which were not extant in Ceylon. In 1802 Burmese monks were invited to Ceylon to introduce the Burmese upasampadā. ⁹³ Subsequently Sinhalese monks went to Burma and studied the Abhidhamma, and on their return they brought back Pāli books written by Burmese monks.

Towards the end of the 19th century King Min-dōn-min, whose tutor Paññāsāmi had written the Sāsanavaṃsa only a few years before, convened a fifth council (1868-71), where under the presidency of the king eminent monks and teachers read or recited the sacred texts to restore the best readings. The complete text of the Tipiṭaka was engraved on 729 stone slabs around the Kuthodaw Pagoda in Mandalay. The dates on the slabs indicate that the texts had probably already been carved before the council was held, and were then corrected in the light of any discussion. ⁹⁴

The Mandalay slabs were re-inked and copied for the sixth council which was held in Rangoon in 1954-56, to mark the 2,500th anniversary of the Buddha’s birth (according to the oriental tradition of chronology). We are informed ⁹⁵ that a draft edition of the Tipitaka, commentaries and sub-commentaries, based upon the fifth council edition which had been revised after comparison with texts from other countries, was prepared by a body of scholars. This was then checked and re-edited by a large number of Burmese mahātheras, and simultaneously by a smaller number of Sinhalese monks. The final version was decided upon (not without argument) by boards of reviewers composed of Burmese, Sinhalese, and Thai monks. The new editions were then ready for printing. The task of the council, spread over two years, was the ceremonial recitation and formal confirmation of the new editions. Although invited, no representative of Cambodia or Laos was able to attend the meetings for scrutinising the new editions, but the Cambodians and Laotians are reported to have given their assent to the decisions arrived at by the representatives of the other countries. While the vast majority of monks present at the sixth council were Burmese, recognition was given to the other countries by appointing their representatives as chairmen for the various sessions.

It is not inappropriate to talk of a Burmese or Siamese or Sinhalese tradition for the transmission of a particular text, and the differences which we find between the readings of the MSS belonging to the various traditions must go back to the councils which have been held from time to time in the different countries. The value of each tradition will depend upon the care with which evidence for variant readings was sifted, and the criteria which were adopted as the basis of the decisions which were made. We have, of course, no way of discovering this for the earlier councils. The way in which the preliminary work for the sixth council was carried out should have resulted in an eclectic edition of the canon and the commentaries, incorporating the best readings from all the oriental editions. It is probable, however, that the Chattthasañgāyana edition is based predominantly upon the Burmese tradition, but it is not possible to be certain about this without carrying out a detailed comparison between that edition and the one inscribed after the fifth council.

The way in which the upasampadā was re-introduced from one Buddhist country into another, and books were brought by visiting monks, has led to a situation where the traditions of each country have become to some extent interwoven. It is sometimes possible to detect the effect which this has had. When, for example, variant readings in a Sinhalese MS depend upon the similarity in shape of two aksaras which resemble each other in the Burmese script, but not in the Sinhalese syllabary, then we have a clear indication that at some time a Burmese MS has been transliterated into Sinhalese characters. ⁹⁶ The export of MSS sometimes results in the fact that a text which has been lost in the country of its origin may be found, safely preserved for centuries, in another country. ⁹⁷

In more modern times the ease of communication has led to a situation where it can become very easy for cross-fertilisation of traditions to take place. The editors of the second European edition of the Sutta-nipāta draw attention to the fact that they have ignored the Siamese edition of that text because it has been influenced by the first European edition. ⁹⁸ It is nevertheless possible that MSS are still extant in libraries in Ceylon, Burma, and Thailand which are based upon a tradition which pre-dates, and therefore perhaps preserves readings older than and rejected by, more recent councils and editions.

Although not so well documented as the Theravādin tradition, it must be assumed that the other traditions followed a similar pattern of recitation and validation as canonical by assemblies of monks. The council held by the Mahāsānghikas after the schism which followed the second council has already been mentioned. The alternative name of the sect suggests that the story of their holding a council is correct, although it is possible that its occurrence was merely deduced by the Theravādins, who observing differences between the canons of the two sects assumed that their rivals would, like themselves, need the authority of a council to authenticate their canon.

Hsüan-tsang records a story that under Kaniska 500 elders subjected the Buddhist texts to a revision and wrote commentaries upon them. ⁹⁹ These commentaries were then deposited in a stüpa. This is designated in the Kanjur as the third council. ¹⁰⁰ If the story is true, it presumably refers to the Sarvāstivādins, who were strong in Kashmir, where the council was held. It is, however, possible that the story was an invention, made up as a direct imitation of the accounts of the earlier councils. The fact that one of the commentaries alleged to have been composed at that council actually exists ¹⁰¹ does not prove the existence of the council.

CHAPTER II - THE PĀLI CANON

1. THE DIVISIONS OF THE CANON

Buddhaghosa gives several ways of classifying the Canon. He states ¹⁰² that it is uniform in sentiment, twofold as regards the dhamma and the vinaya, threefold as regards the first, middle and last words, and also as regards the pitakas, fivefold as regards the nikāyas, ninefold as regards the angas, and forming 84,000 divisions according to the units of the dhamma.

Except for two references to the five nikāyas in later passages in the Vinayapitaka, ¹⁰³ the usual classification of the canon found in canonical texts is that of nine angas, viz. suttam geyyam veyyākaranaṃ gātham udānam itivuttakam jātakam abbhutadhammam vedallam. ¹⁰⁴ Buddhaghosa explains these terms as follows: ¹⁰⁵

Herein, the twofold Vibhanga, the Niddesas, the Khandhaka, the Parivāra, Mañgala, Ratana, Nālaka, and Tuvaṭaka Suttas of the Sutta Nipāta and other sayings of the Tathāgata bearing the name sutta should be known as Sutta (Discourses).

All the suttas containing stanzas should be known as Geyya (Recitation), particularly the entire Sagātha-vagga in the Samyutta.

The whole of the Abhidhamma Piṭaka, suttas which contain no stanzas and any other (sayings from the) word of the Buddha not included in the other eight Angas should be known as the Veyyākarana (Expositions).

The Dhammapada, Theragāthā, Therīgāthā, and sections entirely in verse in the Sutta nipāta which are not designated as suttas should be known as Gāthā (Stanzas).

The eighty-two suttantas containing stanzas which were prompted by an awareness of joy should be known as Udāna (Utterances of Joy).

The 112 suttantas which have been handed down prefixed with the statement, “For this has been said by the Exalted One,” should be known as Itivuttaka (Thus Saids).

The 550 Birth Stories commencing with Apaṇnaka should be known as Jātaka (Birth Stories).

All the suttantas connected with wonderful and marvellous phenomena handed down with words to such effect as, “O monks, these four wonderful and marvellous qualities are seen in Ananda,” should be known as Abbhutadhamma (Marvellous Phenomena).

All the suttantas requested to be preached as a result of repeated attainment of wisdom and delight, such as Culla-vedalla, Mahā-vedalla, Sammādiṭṭhi, Sakkapañha, Saṅkhārabhājaniya, and Mahāpuṇnama Suttas and others should be known as Vedalla (Analyses). ¹⁰⁶

It is clear that all Buddhaghosa knew about some of these terms was the fact that there were in his time a number of texts which happened to have the same name as an anga. Despite the fact that books called Jātaka, Udāna and Itivuttaka actually exist in Pāli, it is probable that the list of nine angas did not originally refer to specific works in the canon, but was a description of various types of text. ¹⁰⁷ Some Sanskrit sources mention twelve angas, ¹⁰⁸ adding nidāna, avadāna (which under the form apadāna is the name of a specific text in Pāli) and upadeśa ¹⁰⁹ to the list. The Mahāvastu speaks of nine angas, ¹¹⁰ and the occurrence of a list of nine in the Saddharmapundarika shows that the smaller number was not confined to the Hīnayāna schools. The list in the latter text differs somewhat from that found in Pāli: sūtra gāthā itivrttaka jātaka adbhuta nidāna aupamya geya and upadeśa. ¹¹¹

According to the Dipavaṃsa, ¹¹² the change from the old ninefold classification into the later one took place at the first council. In its account of that council it states that the 500 theras who held it divided up the Buddha’s ninefold teachings into chapter (vagga), collection of 50 (paññāsaka), connected collection (samyutta), and section (nipātaka). “Thus they made the basket of āgamas (āgama-pitaka) which is known by the name of sutta (sutta-sammata).”

The division of the canon as it exists today is that into pitakas. The only reference to someone knowing the three pitakas which we find in the canon is in the later Parivāra section of the Vinaya-piṭaka, ¹¹³ but the occurrence of the word petakin “knowing the pitaka(s)” in inscriptions at Bharhut ¹¹⁴ indicates that in the centuries immediately after the time of Asoka the word pitaka had gained sufficiently wide usage for it to be used in this public way. Although Buddhaghosa, as stated above, uses the word nikāya in such a way that the term pañca nikāyā can be used to include the whole canon, the word is used in modern times only of the five nikāyas of the Sutta-piṭaka.

The Sarvāstivādins also divided their canon into three piṭakas. They had certainly done this by the time of Kaniṣka, for Hsüan-tsang reports ¹¹⁵ that at the council held during his reign commentaries were composed upon the three piṭakas. Their adoption of this division, however, must go back to an even earlier time, when the Sarvāstivādins and Theravādins were in close contact, for both give the pitakas the same names, even though the texts comprised in the Sarvāstivādin Abhidharma-piṭaka do not agree with those in the Theravādin Abhidhamma. ¹¹⁶

2. THE VINAYA-PITTAKA

The Pāli Vinaya-piṭaka ¹¹⁷ falls into three sections: the Suttavibhanga, which consists of the Mahāvibhanga (= Bhikkhuvibhanga) giving the regulations for the monks, and the Bhikkhunīvibhanga which gives the regulations for the nuns; the Khandhaka, which consists of the Mahāvagga and the Cullavagga; and the Parivāra.

In the oldest account which we possess of the first council, it is said that Mahākassapa asked Upāli about the twofold vinaya. ¹¹⁸ This suggests that he asked him only about the Pātimokkha rules for bhikkhus and bhikkhunīs. In the account which Buddhaghosa gives of that council, ¹¹⁹ he states that the theras classified the Mahāvibhanga, the Bhikkhunīvibhanga, the Khandhaka and the Parivāras, but he concedes that the Vinayapitaka as he knew it contained material which had not been recited at the first council. ¹²⁰ It is obvious, and presumably Buddhaghosa realised, that the final two sections of the Khandhaka, which deal with the first council and the second council which occurred 100 years later, could not have been recited on that occasion, but there is no indication of the portions of the Vinaya-piṭaka he had in mind when he wrote of texts not being recited.

2.1. The Suttavibhanga

The core of the Suttavibhanga is the Pātimokkha. This is a set of 227 rules for bhikkhus and 311 for bhikkhunīs. In the Pāli canon the Pātimokkha has no independent existence, as it does in the traditions of other Buddhist schools, but is imbedded in the Suttavibhanga. The rules fall into eight sections: Pārājika “Defeat” (4), Sanghādisesa “Formal meeting” (13), Aniyata “Undetermined” (2), Nissaggiya “Forfeiture” (30), Pācittiya “Expiation” (92), Pāṭidesanīya “Confession” (4), Sekhiya “Training” (75), Adhikaranasamathā dhammā “Legal questions” (7), giving a total of 227 rules for bhikkhus. The sections of rules for the bhikkhunīs follow the same pattern except that they have no Aniyata class. The numbers are: (8), (17), (-), (30), (166), (8), (75), (7) making a total of 311 .

Each section of the Suttavibhanga follows the same pattern, ¹²¹ except for the eighth section (Adhikaranasamathā dhammā ) which, since it has no explanatory stories and no Old Commentary as do the other sections, would seem to be an addition to the list. ¹²² The standard pattern consists of (a) a story which leads up to the formulation of the rule. Sometimes two stories can be placed in a chronological order, indicating that one rule was made after another. ¹²³ These stories sometimes occur elsewhere in the canon, which raises the question of which was the earlier position. ¹²⁴ It is possible that some of the stories were invented or borrowed from other sources to explain rules which already existed; ¹²⁵ (b) a Pātimokkha rule, which always states the penalty for breaking it. The nature of the Pācittiya rules suggests that this group as a whole was a later addition to the code. ¹²⁶ Although by far the greater number of rules are said to have been enunciated by the Buddha himself, many sub-rules are laid down without reference to him. ¹²⁷ This perhaps means that they were promulgated by some of his chief followers, or even added after his death; (c) the Old Commentary, which is really an analysis of words (Pada-bhājanīya). This defines the Pātimokkha rule word by word, ¹²⁸ giving, for the most part, an explanation in the form of a list of synonyms. The similarity between this method and that found in the Niddesa suggests that both belong to the same period; (d) more stories, telling of deviations from the rules, which were either not so grave as to entail the maximum penalty, or reasonable enough to warrant modification or relaxation of the rule. Occasionally they were of such a kind that they were not to be rendered permissible by any extenuating circumstances. Sometimes (c) and (d) are reversed in order. Sometimes there is no (d). The stories in (d) are comparable to those in (a). ¹²⁹

In the form in which the stories are told, it appears that every rule was made to fit a particular set of circumstances, and it seems likely that some, at least, of the stories are afterthoughts, added to explain the origin of the rule. ¹³⁰ Sometimes the stories are grouped together to form a set, with each story showing no more than a minute variation from the others. It seems very unlikely that the complete set of rules was formulated at once, ¹³¹ so some were undoubtedly proclaimed as the need arose. On the other hand, it is not unlikely that a general code of conduct for mendicant orders already existed in the time of the Buddha, from which some borrowing could take place. This view is supported by the general agreement between the basic rules of the Jain and Buddhist orders. We may conclude that it is most unlikely that the historical order of promulgation was as in the present form of the Vinaya-piṭaka. For the most part, the stories are told in the context of the biography of the Buddha, but a few stories of the Jātaka or Apadāna type are included. ¹³² It has been suggested that such inclusions are a sign of lateness of composition, and can be used as an indication of the relative dating of different versions of the Suttavibhanga. ¹³³

The Bhikkhunīvibhanga is somewhat misleading in appearance, in that it includes, in its present form, only those rules which are not already included in the rules for bhikkhus. Since the Sekhiya rules are the same for both, only the first and last of the 75 are given. At the end of the Pācittiya section, the number is said to be 166,¹³⁴ but only 96 have been recorded. Of the bhikkhus’ Pācittiya rules, however, 70 also apply to the bhikkhunis, making up the total of 166. The Old Commentary is included for those rules which are given, showing that the promulgation of the rules for bhikkhunis was early enough to pre-date the composition of that commentary.

Although the Pātimokkha does not have a separate existence as a literary text in Pāli, an independent Prātimokṣa is found in the canons of other Hīnayāna schools. The Bhikṣu-prātimokṣa of the Mūlasarvāstivādins was discovered in Gilgit, and has been published, ¹³⁵ as has the comparable text of the Mahāsānghika-Lokottaravādin school which was found in Tibet. ¹³⁶ Portions of the Bhikṣuprātimokṣa of the Sarvāstivādins have been found in Turkestan, ¹³⁷ and also fragments of their Bhikṣuṇīprātimokṣa. ¹³⁸ There are also fragments extant of a Dharmaguptaka Bhikṣuprātimokṣa, also found in Turkestan. ¹³⁹

For the most part, these Prātimokṣas of other schools do not differ greatly from the Pāli Pātimokkha. The main differences are in the Pācittiya and Sekhiya rules. ¹⁴⁰ Of the Pācittiya rules, the difference concerns rules which are similar to other rules, and might be considered unnecessary. Of the Sekhiya rules, it is probable that different monastic and lay conditions called for variations. The Mahāsānghika-Lokottaravādin Prātimokṣa includes in some cases, e.g. the Pārājika rules, ¹⁴¹ brief statements about the time and place at which the Buddha promulgated the rule. These are very similar to the stories included in the Pāli Suttavibhanga.

The other Hīnayāna schools also possessed texts equivalent to the Suttavibhanga. Fragments from Turkestan of the Sarvāstivādin Vinayavibhanga to both the Bhikṣu- ¹⁴² and Bhikṣunīprātimokṣas ¹⁴³ have been published, while fragments of manuscript found in Gilgit have been identified as coming from the Mūlasarvāstivādin Vinayavibhanga to the Bhikṣuprātimokṣa. ¹⁴⁴ The Mahā-sānghika-Lokottaravādin Bhikṣuṇī-vinaya ¹⁴⁵ includes the vibhanga to the Bhikṣuṇī-prātimokṣa as well as the astau guru-dharmāh, which in the Pāli Vinaya-piṭaka are found in the Bhikkhunī-kkhandhaka of the Cullavagga, ¹⁴⁶ and also the Bhikṣunī-prakīrṇaka and the uddāna to the Bhikṣu-prakīrṇaka. The latter text has also been published, under the title Abhisamācārikā. ¹⁴⁷ The Bhikṣuṇī-vinaya gives all the rules for bhikṣunīs, even those which duplicate the rules for bhikṣus. It gives stories to explain the promulgation of the rules, but they are not identical with those found in Pāli. There is also a word commentary included, but it is not the same as the Pāli pada-bhājanīya. There are more stories of the avadāna type in it than in the Pāli Bhikkhunīvibhanga, ¹⁴⁸ which possibly indicates that the Mahāsānghika-Lokottaravādin text in its present form is later than the Pāli equivalent.

2.2. The Khandhaka

The Khandhaka, in two portions (the Mahāvagga and the Cullavagga), is based upon the rules of conduct for the Buddhist Sangha, set in the framework of the biography of the Buddha. It comprises 22¹⁴⁹ sections: (a) a portion of a biography of the Buddha (Mahākhandhaka), starting from the enlightenment and extending as far as the entry of Sāriputta and Moggallāna into the Order. It includes an account of the first sermon and the four Noble Truths; (b) various rules for bhikkhus in 18 sections: ¹⁵⁰ “Observance” (Uposathakkhandhaka); “The rains” (Vassupanāyikakkhandhaka); “Invitation” (Pavāraṇakkhandhaka); “On hides” (Cammakkhandhaka); “On medicines” (Bhesajjakkhandhaka); “On kaṭhina” (Kaṭhinakkhandhaka); “On robe-material” (Cīvarakkhandhaka); “On (the monks at) Campā” (Campeyyakkhandhaka); “On (the monks of) Kosambī” (Kosambakkhandhaka); “Formal acts” (Kammakkhandhaka); “Probation” (Pārivāsikakkhandhaka); “Accumulation (of offences)” (Samuccayakkhandhaka); “Settlement” (Samathakkhandhaka); “Minor matters” (Khuddakavatthukkhandhaka); “Lodgings” (Senāsanakkhandhaka); “Schism” (Saṃghabhedakkhandhaka); “Observances” (Vattakkhandhaka); “Suspending the Pātimokkha” (Pātimokkhaṭhapanakkhandhaka); (c) rules for nuns (Bhikkhunikkhandhaka), (d) two supplements giving the history of the first ¹⁵¹ and second councils; (e) the introduction to the Pātimokkha, with a portion of the Old Commentary upon it, which has become detached from the rest of the Pātimokkha rules and inserted in an appropriate place in the Uposathakkhandhaka. ¹⁵²

As in the case of the Suttavibhanga, stories are told to illustrate the circumstances in which the rules were first promulgated. ¹⁵³ They are set in the context of a biography of the Buddha, but not in any chronological order, ¹⁵⁴ for they represent an attempt to classify the rules by subject matter, so that all the rules on any one subject, e.g. medicines, have been grouped together. Scattered throughout the khandhakas are various rules which were enunciated to the group of five. These must be the first five bhikkhus, in which case all these stories should have been together, if the material had been presented in chronological order. The arrangement by subject matter, so that the stories had to be inserted into the appropriate khandhakas, has led to their becoming separated. ¹⁵⁵

The stories are sometimes found in other texts, ¹⁵⁶ and it can sometimes be shown whether the Vinaya-piṭaka is the borrower or the lender. ¹⁵⁷ Sometimes the same story is told to illustrate different rules. Sometimes the same rule is promulgated in both the Khandhaka and the Vibhanga. ¹⁵⁸ In this case, the same story is told on both occasions. There is the occasional Apadāna story, ¹⁵⁹ which as we have seen is held to be a feature of lateness.

In the Uposathakkhandhaka, the Buddha’s permission to recite the Pātimokkha is followed by an exhortation to be used when beginning the recitation. This is, in fact, the exhortation from the Pātimokkha itself, as is shown by the fact that it is followed by the appropriate portion of the Old Commentary upon the Pātimokkha. ¹⁶⁰ Comparison with the Khandhaka section of the Vinayapiṭaka of other sects shows that there are gaps in the Pāli version. There is, for example, no equivalent to the stūpa-passages found in the Kṣudraka-vastu of the Vinaya of the Mūlasarvāstivādins. The reason is probably because the compilers of the Pāli canon thought that stüpas were the province of laymen, not of bhikkhus. ¹⁶¹ The Kaṭhinakkhandhaka is very unclear, and seems to have been corrupted to the state where it is barely intelligible without recourse to other versions or to the Parivāra. ¹⁶²

The section of rules for bhikkhunis (Bhikkhunī-kkhandhaka) is probably an addition, made after the rules for bhikkhus, and was therefore at one time the last section of the Khandhaka. As, however, it is now followed by the accounts of the two councils, and has parallels in the Khandhakas of other sects, it must be presumed that the Bhikkhunī-kkhandhaka predates the schism between the schools. The inclusion of the accounts of the councils at Rājagaha (after the Buddha’s death) and at Vesālī ( 100 years after) shows that the Vinaya-piṭaka in its present form is at least 100 years later than the death of the Buddha. Since these accounts are included in all versions of the Vinaya, it can be concluded that all sects had (approximately) the same Vinaya until the second council [^2-2-47] and the schism that followed it. The fact that part of the Vinaya is late does not, however, mean that all of it is late. The different arrangement of the khandhakas and the material in the various vastus shows that there was already the beginning of divergency before the time of the second council.

The greater part of the Mūlasarvāstivādin Skandhaka has been discovered in Gilgit and published. ¹⁶³ It lacks some of the khandhakas in the Pāli version, and some of the material is arranged differently, e.g. the historical information found in the Pāli Mahākhandhaka appears in the Samghabhedavastu. ¹⁶⁴ Fragments of the Sarvāstivādin Skandhaka have been found in Turkestan, ¹⁶⁵ and a fragment of the Mahāsānghika Skandhaka has been found in Afghanistan at Bamiyan. ¹⁶⁶ The Mahāvastu ¹⁶⁷ is based upon a recension of the Vinaya belonging to the Mahāsānghika-Lokottaravādins. ¹⁶⁸ It corresponds roughly to the Mahāvagga portion of the Pāli Khandhaka, but it contains very few monastic rules, ¹⁶⁹ and has many later additions of the Jātaka and Avadāna type.

The inclusion of many stories of the Avadāna type accounts for the fact that the work is frequently styled the Mahāvastu-Avadāna in the colophons to many of the chapters. The use of the word Avadāna in this context resembles that found in the Mahāpadānasutta (Sanskrit Mahāvadānasūtra) of the Dīghanikāya, ¹⁷⁰ where it is used of the Buddha, instead of theras and theris as in the late canonical Apadāna. These narrative legends are introduced, not as in the Mahāvagga to illustrate the promulgation of a rule but to illustrate the Buddha’s virtues in his earlier lives. To the material which is given in the Pāli Mahāvagga is added much which is found in the Nidānakathā of the Pāli Jātaka, so that the Mahāvastu gives an account of incidents in the remote past of the Buddha’s career, as well as his birth and childhood, his enlightenment and his activities as a teacher. ¹⁷¹ Although the Mahāsānghika-Lokottaravādins believed that the personality of the Buddha was docetic and he was really supramundane (lokottara), there is in fact only one short passage which deals with his transcendental nature. ¹⁷²

Not only are there close parallels between the Mahāvastu and the Pāli Mahāvagga, which suggest that both texts were dependent upon the same source material, ¹⁷³ but there are many other parallels too, sometimes so close verbally that one version may be corrected with the help of the other. Among the Pāli texts to which comparison can be made are the Jātaka, Suttanipāta, the Dhammapada, the Dīghanikāya, the Vimānavatthu and the Buddhavamsa. Even when all these insertions, which probably represent additions to the Mahāvastu, are disregarded, it can be shown that the Mahāvastu contains different strata, ¹⁷⁴ which differ in the language employed in them. ¹⁷⁵

It is not easy to understand why there is, except for a reference to the four types of ordination, an almost complete lack of monastic rules. It has already been noted ¹⁷⁶ that the Mahāsānghika-Lokottaravādin Bhikṣuṇī-Vibhanga includes a section which in the Pāli Vinaya-piṭaka occurs in the Bhikkhunīkkhandhaka, and it is possible that other Skandhaka material was transferred to texts, now lost.

When some of the formulae included in the Khandhaka came to be used for the transaction of business at meetings of the Sangha, e.g. for the ordination of bhikkhus, they gained an existence as independent texts. Although these usually follow the pattern laid down in the Vinaya-piṭaka, the fact that they have been used in different places and circumstances has led to changes in detail. ¹⁷⁷ A number of these separate kammavācās have been published. ¹⁷⁸

They also existed in the other Hīnayāna schools, and karmavākyas or karmavācanās for bhikṣus or bhikṣunīs have been discovered, belonging to the Sarvāstivādin and Mūlasarvāstivādin schools. ¹⁷⁹

2.3. The Parivāra

In Buddhist Hybrid Sanskrit ¹⁸⁰ the word parivāra is normally taken to mean “accessory, appendix, addendum,” ¹⁸¹ and such a meaning would suit the Pāli Parivāra, since it seems to be a supplement to the Vinaya. The place and time of its composition are not known, and some scholars favour the view that it was composed in Ceylon. ¹⁸² There is evidence for believing that portions of it, at least, were added in Ceylon, for it includes a long list of teachers who were responsible for teaching the Vinaya in Jambudīpa, and then in Tambapanṇi (Ceylon), ending with Sīvatthera. It has been claimed, ¹⁸³ on the basis of these names, that the present form of the Parivāra must have been fixed in the first century A.D., for the list contains no thera later than that century. The Sāsanavamsa, however, states ¹⁸⁴ that the list which is found in the Vinaya is the succession of elders down to those who assembled in the fourth council. If that is correct, then we should date Siva to the first century B.C.

The possibility of Siva being connected with the writing down of the canon is perhaps supported by the fact that there are two references to writing in the Parivāra. At the end of the first eight (of 16) sections of the first chapter, there is a statement that “these eight sections are written in a manner for recitation,” ¹⁸⁵ while at the end of the book, after the words Parivāro nitthito, there is a statement that Dīpanāma, having asked various questions about the ways of former teachers, thinking out this epitome of the details for the middle way of study, had it written down for the bringing of happiness to disciples. ¹⁸⁶ It seems, however, unlikely that if the whole of the Parivāra had been composed at the time it was written down, it would have had the references to writing inserted in the way they are, and there seems to be no good reason for doubting that these two statements are interpolations, ¹⁸⁷ added when the Parivāra, together with the rest of the canon, was written down in the first century B.C. They have, therefore, some bearing upon the present form of the Parivāra, but tell us little about the composition of the various chapters.

Since some of the chapters of the Parivāra are summaries of various sections of the Vinaya, it has been suggested ¹⁸⁸ that the Parivāra is, in fact, older than the Vinaya and originally performed the same function towards it as do the Mātikās for the Abhidhamma texts. While it is not impossible that the Parivāra contains a very early portion of this type which was later enlarged, it nevertheless shows all the signs of being an addition to the Vinaya. The Dipavamsa states ¹⁸⁹ that it was one of the texts which was not accepted by the Mahāsānghikas, which suggests either that it did not exist at the time of the schism, or if it existed it had not yet gained canonical status at that time. It seems more likely that a collection was made up of ancillary works of various dates which sprang up around the Vinaya, and this happened early enough for the collection to have attained canonical status, just as the Niddesa did, although neither claims to be Buddhavacana. ¹⁹⁰

As we have it, the Parivāra contains 19 chapters, whereas in his account of the first council Buddhaghosa states ¹⁹¹ that the Parivāra has 16 sections. After Chapter XIV, which deals with the Kathina regulations, the words parivāram nitthitam occur, ¹⁹² which suggests that at one time the Parivāra ended at that point, and the following chapters were added later. It seems possible to divide two of the earlier chapters into halves, which would give a total of 16 , and this may be the answer to the problem. ¹⁹³ In his commentary upon the Vinayapitaka, however, Buddhaghosa comments upon all 19 chapters, ¹⁹⁴ without any reference to the apparent inconsistency in his statement about the number of chapters, although he is sufficiently aware of some types of inconsistency to find it necessary to comment upon them. ¹⁹⁵

In the first chapter of the Parivāra (Mahāvibhanga), every rule in the Mahāvibhanga section of the Suttavibhanga is examined in the exact order in which it was laid down there, with special reference to the place in which it was promulgated. It is a kind of catechism, ¹⁹⁶ containing bare or condensed outlines of questions to be asked and answered, in the study of the sikkhāpadas. It makes a clear contribution to the task of learning and mastering the Vinaya, ¹⁹⁷ and by extracting material from the vast mass of the Vinaya it reduces it to manageable proportions. The next chapter (Bhikkhunīvibhanga) does the same for the bhikkhunīs’ rules. Other chapters are similarly devoted to questions and statements about the types and numbers of offences detailed in the Vinaya. The sixth chapter (Ekuttaraka) is arranged in the same way as the Sangītisutta and Dasuttara-sutta of the Dīgha-nikāya and the Añguttara-nikāya as a whole, with topics listed numerically, each section being one more than its predecessor. Although classifications of this type may be thought to be relatively late, the presence of the system in several places in the canon, and also in the canons of other sects, ¹⁹⁸ shows that it had nevertheless been established in Buddhism before the time of the schisms. The Ekuttaraka differs from the other texts of its type in that its enumerations are almost entirely restricted to items found in the Vinaya. ¹⁹⁹ It resembles the Añguttara-nikāya in that it lists items in groups up to the number eleven.

In common with the first two chapters, Chapter XIII “the great collection” (Mahāsamgāma) consists almost entirely of a commentary upon the phrases at the beginning of the chapter, which closely resembles the style of the Old Commentary found in the Sutta-vibhanga. Chapter XIV on the Kathina also consists of a commentary upon the opening phrases, and gives information about the proper way of making the kathina cloth. As already noted, ²⁰⁰ the relevant section of the Khandhaka is not entirely satisfactory, and it is to be observed that this section of the Parivāra is not merely a synopsis of what the Vinaya says, but includes material not found there. This must be old, since it corresponds with what is stated in other traditions about the kathina. ²⁰¹

The fifteenth chapter (Upālipañcaka) consists of a number of questions put by Upāli to the Buddha about the Vinaya, the answers to which were all in the form of sets of five. It is to be compared with the earlier chapter entitled Ekuttaraka, and even more with the section of the Añguttara-nikāya entitled Upālivagga, ²⁰² where, however, the Buddha’s answers are all in sets of ten. The eighteenth chapter (Sedamocakagāthā) is aptly named “Sweat-inducing,” for it consists of sets of puzzles, based upon apparent contradictions in the Vinaya, the answers to which call for a great deal of mental exertion. The style is reminiscent of the sets of dilemmas in the Milinda-pañha, ²⁰³ but in the Parivāra the problems are all drawn from the Vinaya.

The Vinaya-piṭakas of other sects also have supplements. In its Chinese translation, the Sarvāstivādin Vinaya is followed by the Vinayakṣudraka and the Uttaragrantha, ²⁰⁴ which includes an Upālipariprcchā, a series of questions on Vinaya put to the Buddha by Upāli. This would seem to correspond to the Pāli Upāli-pañcaka, but a Sanskrit portion of this found in Turkestan does not agree with the Pāli Parivāra. ²⁰⁵ One Chinese version of the Upālipariprcchā is said to be very similar to the Pāli Parivāra, although there seem to be abridgements and changes of order in the treatment of the various rules, and there is doubt about the sect to which this text belonged. ²⁰⁶ There also exists in Chinese an Ekottara section in the appendices to the shortened version of the Sarvāstivādin Vinaya called Daśādhyāya Vinaya. ²⁰⁷ The Dharmaguptaka Vinaya also has a section called Ekottara, ²⁰⁸ while the Mūlasarvāstivādins too had a supplement to their Vinaya called Uttara-grantha. ²⁰⁹ Without more investigation it is not possible to say what resemblances there are between these various works, but it is worthwhile referring to them here, despite the lack of Sanskrit versions of them, because the fact that other schools found it necessary to have supplements to their Vinayas supports the view that some, at least, of the Pāli Parivāra was composed in India before the introduction of Buddhism to Ceylon.

3. THE SUTTA-PITAKA

Buddhaghosa states ²¹⁰ that the Sutta- or Suttanta-pitaka consists of the five nikāyas: Dīgha-, Majjhima-, Saṃyutta-, Añguttara-, and Khuddaka-nikāya, and this is the classification generally accepted now although, as already noted, Buddhaghosa also uses ²¹¹ the term “five nikāyas” of the canon as a whole.

The first four nikāyas consist of discourses (suttas) ²¹² ascribed to the Buddha, or (rarely) to a disciple. These suttas are for the most part in prose, with some verse included. They usually have a short introduction giving the circumstances in which the sutta was delivered. In the case of the Dīgha- and Majjhi-ma-nikāyas, the names seem to reflect the length of the suttas, with the long suttas in the Dīgha- and those of middle length in the Majjhima-nikāya, ²¹³ although there is evidence that the size of some suttas has been increased by interpolation, so that their present length is not necessarily a guide to their original size, and we cannot be certain that the shortest sutta in the Dīgha- was longer than the longest sutta in the Majjhima-nikāya in the early collections. We can in any case deduce from the fact that the contents of the nikāyas do not agree exactly in the canons of the various sects that the division of suttas by length was, at best, a somewhat rough and ready method of classification.

By implication, the contents of the Saṃyutta- and Añguttara-nikāyas must be the short suttas. The name Saṃyutta- indicates that the suttas have been arranged together according to their contents, while the Añguttara-nikāya is made up numerically, with each nipāta exceeding its predecessor by one. There is inevitably an overlap between the contents of these two nikāyas, because some of the connected suttas deal with numerical matters.

In the Sanskrit canons the nikāyas are called āgamas, ²¹⁴ and the word āgama is occasionally used in Pāli in this sense. It is, however, more commonly used in Pāli in a general sense, ²¹⁵ and Buddhaghosa defines the word as the study of the Buddha’s words. ²¹⁶ Hence a thera can be described as āgatāgama or svāgama. It seems to be used in the Vinaya of the Pātimokkha, ²¹⁷ but it is sometimes used by Buddhaghosa in the sense of nikāya, and in his commentaries upon the Dīgha-, Saṃyutta- and Añguttara-nikāyas he writes of the Dīghāgama, the Saṃyuttāgama and the Anguttarāgama respectively. ²¹⁸ In the introductory verses to his commentaries upon the first four nikāyas he states that the Visuddhimagga is in the middle of the four āgamas. ²¹⁹

In the Sanskrit canons the āgamas are named Dirgha-, Madhyama-, Saṃ-yukta- and Ekottarika-, and the close similarity to the Pāli names indicates that the collections had begun to be formed while the schools were still in contact, i.e. before the schisms which started after the second council. The fact that one and the same sutta is sometimes found in more than one nikāya in the Pāli canon would seem to indicate that the bhānakas of the various nikāyas could not always agree about the allocation of suttas. The fact that the sütras in each Sanskrit āgama do not coincide with their Pāli equivalents would seem to indicate that each school had its own bhānakas who, while all agreeing in general with the other bhānakas of their own and other sects, nevertheless preferred to differ over the placing of some sütras. This suggests that there was in early times a large collection of suttas which were remembered by heart, and the task of allocating them to the various nikāyas/āgamas had not been finished, or the allocation completely agreed, by the time the schools began to separate. The reference in the Vinaya-piṭaka to an upāsaka inviting the bhikkhus to come and learn a sutta from him before it is lost ²²⁰ would seem to indicate that the collection of suttas had not yet been completed. The methods of authentication of suttas which are given in the Mahāparinibbānasutta ²²¹ would also seem to apply to the time when the collection of suttas was still going on.

The list usually followed for the fifth nikāya, the Khuddaka-nikāya, is that given by Buddhaghosa: ²²² Khuddakapāṭha, Dhammapada, Udāna, Itivuttaka, Suttanipāta, Vimānavatthu, Petavatthu, Theragāthā, Therīgāthā, Jātaka, Niddesa, Paṭisambhidāmagga, Apadāna, Buddhavaṃsa, Cariyāpiṭaka. In Burma the Peṭakopadesa, Nettippakaraṇa, Suttasangaha, and Milindapañha are also regarded as belonging to the Khuddaka-nikāya. ²²³

Buddhaghosa informs us ²²⁴ that the Dīgha-bhānakas did not accept the first and the last three in this list, but believed that the theras at the first council recited the other twelve (dividing the Niddesa into Mahā- and Cūla-niddesa), and calling them the Khuddakagantha included them in the Abhidhammapitaka. He goes on to state, however, that the Majjhima-bhānakas added the Cariyāpiṭaka, Apadāna and Buddhavamsa to the Dīgha-bhānakas’ list, also called the collection the Khuddakagantha, but included it in the Suttantapitaka. This presumably means that the Dīgha-bhānakas closed their list of the Khuddaka-nikāya before these four texts were regarded as canonical, while the Majjhima-bhānakas closed their list before the Khuddakapāṭha was accepted.

The Chinese translation of the Samantapāsādikā omits the Khuddakapāṭha from the Khuddaka-nikāya, and varies the order of the other 14 texts. ²²⁵ This suggests that either the Khuddakapāṭha was not included in the version of Buddhaghosa’s commentary which was being translated, or the translators were in contact with and influenced by someone who knew that a section of the Theravādin school did not accept it as canonical.

3.1. The Dīgha-nikāya

The Dīgha-nikāya ²²⁶ contains 34 suttas, in three groups: the Silakkhandhavag. ga (1-13) dealing with virtue (sīla); the Mahāvagga (14-23) containing suttas which mostly have the word Mahā- in their title; and the Pāṭikavagga (24-34), which gains its name from the fact that its first sutta is the Pāṭikasutta. The vaggas (chapters) differ in content and character, but they all contain a mixture of older and later material. ²²⁷ The earliest stratum is found mainly in the first vagga, and the latest in the last, while the longest suttas are in the second vagga. The form of the suttas varies. Those in the first vagga are in prose, as are some in the second and third. Many in the second and third vaggas are in prose interspersed with verse. Some (e.g. suttas 20 and 32) are almost entirely in verse.

3.1.1. The Silakkhandhavagga

Each of the suttas in this vagga has incorporated in it a list of the moral rules (silas). Most of the suttas describe the training of a bhikkhu in three stages, beginning with the silas, then proceeding to the practice of concentration (samādhi), and ending with the full knowledge (pañnaā ) of the arahant. The list occurs in its most complete form in the Sāmaññaphalasutta, from which portions of it are quoted in the other suttas, which suggests that either the Sāmaññaphalasutta existed before the other suttas were composed, or the list existed at first as an independent composition, ²²⁸ from which all the suttas, including the Sāmaññaphalasutta, borrowed it. Several of the suttas discuss the views of the brahmans on sacrifice and sacred knowledge, the doctrines of various religious schools, the value of caste and self-mortification, and expound important doctrines. The Buddha’s method of discussion in the suttas of the Silakkhandhavagga is much the same in each case. He takes as the starting point of the discussion the object put forward as desirable by his opponent, e.g. sacrifice, caste, or asceticism, and by inserting a higher meaning into the words being discussed, or by concentrating upon the ethical concepts involved, he leads his opponent up to his own conclusion, viz. the goal of arahant-ship.

The Brahmajālasutta (1) tells how the Buddha knows, like a good fisherman, how to catch in his net of views ²²⁹ all sophists and philosophers, and to prove their doctrines and speculations to be worthless and obstacles to true salvation. In the course of this, he enumerates 62 different philosophical views, from which the follower of the Buddha is to keep away. The sutta is important, not only for the information it gives about Buddhism, but also for the contribution it makes to the history of early religion and philosophy in India. There is also much dealing with social conditions, for as part of his rules for his followers the Buddha lists many activities and professions which they must avoid.

The Sāmaññaphalasutta (2) tells how King Ajātasattu of Magadha, after failing to get answers to his questions from other religious leaders, visits the Buddha and asks whether members of the Sangha who have given up the world gain any benefit from their action. The Buddha justifies the formation of the Sangha and the enunciation of the Vinaya rules, and gives a list of the advantages to be gained by his followers. The introduction to the sutta gives information about non-Buddhist sects at the time of the Buddha, and since some of it is in a non-Pāli dialect of MIA ²³⁰ it is likely that it is taken from the actual scriptures of the leaders being described. The advantages which his followers will gain are given in an ascending order of merit, starting with the honour and respect which are shown to those following the life of a samana, and continuing through a series which is not necessarily restricted to Buddhists, until the final stages, when the four noble truths are attained, the āsavas are destroyed, and arahant-ship is won.

In the Ambaṭthasutta (3) the young brahman Ambaṭtha visits the Buddha to see if he carries the 32 marks of a great man (mahā-purisa), and becomes involved in a discussion about caste, in which the Buddha proves that the khattiya Sākyas are superior to the brahmans. The sutta was clearly composed at a time when the 32 marks were already in existence. There is also mention of super-normal powers being employed to enable Ambaṭtha to see the two marks which are not normally visible.

The Sonadandasutta (4) is also concerned with the question of caste, and deals with the problem of what constitutes the essential quality which makes a man a brahman. The brahmanical view is that a man is a brahman only if he is born of brahmans on both his father’s and his mother’s side back through seven generations. ²³¹ By the Buddha’s argument, however, Sonadanda is led to accept that the one who attains arahant-ship is not only called a brahman in Buddhist terms, but is in fact declared to be the only true brahman.

The Kūṭadantasutta (5) is also anti-brahmanical, and tells of a brahman who is desirous of holding a great sacrifice, and comes to the Buddha to seek advice about the best way of doing it, and to enquire about the requisite utensils. In the guise of a legend about King Mahāvijita, the Buddha tells of the Buddhist form of sacrifice, where nothing, whether animal or vegetable, is harmed, and the sacrifice consists of the largesse of food, offered not on behalf of the king himself, but of all good people.

The Mahālisutta (6) falls into two quite separate parts. In the first the question of the ability to attain the supernormal eye and ear power is discussed, and the questioner is told that it is not to attain such powers that people join the Sangha. He is led on to consider arahant-ship as the aim, following the eightfold path. In the second part the questioner is told about Jāliya, who believed that the soul and the body were the same thing, but was led through the discussion set out in the Sāmañnaphalasutta up to the attainment of arahantship.

The Jāliyasutta (7) then follows as a separate sutta in its own right. It can be deduced that its inclusion in the Mahālisutta was early, and predated the formation of the Dīgha-nikāya as we have it, because the discussion of sīla occurs only in the Jāliya portion. Without it, the Mahālisutta would not have merited inclusion in the Silakkhandhavagga.

The Kassapasīhanādasutta (8) tells of the Buddha’s discussion with a naked ascetic about asceticism, and includes a general account of the practices adopted by various groups of ascetics. The Buddha refutes the charge of condemning all asceticism, which probably arises from his own rejection of such practices (although no reference is made to this in the sutta), ²³² and maintains that the practices laid down in the eight-fold path and the discipline laid down for bhikkhus are more difficult to perform than mere tapas. As in the Mahālisutta (6), there is a quotation from elsewhere in the Dīgha-nikāya, but it is not quoted in full here. The Buddha tells of a visit from a co-religionist named Nigodha, which is the subject of the Udumbarikasīhanādasutta (25).

The Poṭṭhapādasutta (9) begins with a discussion on the nature of the cessation of consciousness which occurs in trance. Through the path of training mentioned in the Sāmañnaphalasutta the discussion passes on to the jhānas, and in what is probably an addition to the sutta, to the first three arūpa-vimokkhas, which are not included in the Sāmaññaphalasutta. The idea of consciousness leads to a discussion of the question of soul, which Poṭṭhapāda regards as being a physical thing. The Buddha explains that there are certain things upon which he has expressed no opinion, viz. the ten indeterminates. ²³³ What he has expressed are the four noble truths.

In what appears to be an addition, the sutta goes on to tell how Poṭ̣̣hapāda, having been abused by the other mendicants, went back to the Buddha and told him of the mendicants’ actions. The Buddha’s reply is based upon two metaphors: the first is that of the man who loves a “beauty queen” ²³⁴ without knowing anything about her (which illustrates the folly of saying more than the evidence allows), and that of the transformation of milk into curds, etc. (which illustrates the existence of an unchanging identity behind changing appearances).

The Subhasutta (10) is attributed to Ānanda, and is specifically stated to have been delivered after the death of the Buddha. ²³⁵ It is almost identical with the Sāmaññaphalasutta, except that Ānanda arranges the states of mind enumerated there as the fruits of the life of a samana under the three headings of sīla, samādhi, and paññā.

In the Kevaddhasutta (11) the householder Kevaddha comes to the Buddha and asks him to tell one of his followers to perform a miracle. As in other suttas, the Buddha accepts the interviewer’s point of view. He accepts that miracles are possible, but says that it is not his practice to ask his followers to perform them. He states that of the three wonders which he himself has realised, ²³⁶ the wonder of education is the greatest, and he gives this in the words of the Sāmaññaphalasutta. To this is added a second, probably later, portion of the sutta which deals with the problem of the disappearance of the four elements. The answer is that they disappear in the state of arahant-ship, when all intelligence comes to an end. To this conclusion is prefixed a mythological preface which describes an ascending hierarchy of fourteen classes of gods. Its purpose is to make the gods confess their own inferiority to the Buddha, and to state the folly of seeking refuge in anyone but him.

In the Lohiccasutta (12) some points about the ethics of teaching are discussed, and blameworthy and blameless teachers are enumerated. This leads to a re-iteration of the exposition set out in the Sāmañnaphalasutta up to the state of arahant-ship.

In the Tevijjasutta (13) the Buddha criticises the position of the brahmans who base their religious life on the system of the three Vedas. This is the only sutta in the Silakkhandhavagga which does not lead up to a discussion of arahant-ship. It goes only as far as the four brahmavihāras-the four states of mind which lead to rebirth in the heaven of Brahmā. The Buddha rejects the brahmanical idea of the three Vedas in favour of his own three vijjās. The brahmans with whom the Buddha was conversing had their own idea about union with Brahmā, and here, as commonly, the Buddha was using the brahmanical term Brahmā in a specifically Buddhist sense.

3.1.2. The Mahāvagga

A number of the suttas in this vagga include the word Mahā- in their title, and it is possible that they have been expanded from shorter suttas, but the corresponding suttas with the word Cūla- in their titles no longer exist. ²³⁷ It is, however, possible to deduce that in some cases addition or interpolation has occurred. Most of the suttas in this vagga are legends rather than discourses, and they have features which show them to be later than the suttas of the first vagga, where the Buddha is only a man, alive or recently dead.

In the Mahāpadānasutta (14) the lives of the seven Buddhas, i.e. Gotama and his six predecessors, are given as a preliminary to laying down the general conditions necessary for the arising of a Buddha, so that their whole course of actions is in accordance with the reign of law in the world (dhammatā). The life of Vipassin, the first of the Buddhas, is given in detail, including all the miracles of the conception and birth of a Buddha, and the 32 marks of a great man, which have already been mentioned in the Ambaṭthasutta (3). Although the inclusion of seven Buddhas and the other features suggest a late origin for this sutta, ²³⁸ it is noteworthy that the paṭiccasamuppāda which Vipassin realises has only ten links instead of the more usual twelve in the form which Gotama realised, ²³⁹ since it omits the two ultimate links avijjā and sañkhārā. The use of the word apadāna in the title of this sutta differs from that found in the late canonical text called by that name, ²⁴⁰ where it is used not of the Buddha but only of theras and therīs.

The Sarvāstivādin version of this sutta, entitled Mahāvadānasūtra, has been found in Turkestan and published. ²⁴¹

The Mahānidānasutta (15) gives the fullest exposition of the pațiccasamuppāda given in the piṭakas, although in the Paccayākāravibhanga of the Vibhanga in the Abhidhamma-piṭaka the formula is re-iterated and analysed with greater variety of presentation. ²⁴² The exposition is given as a rebuff to Ānanda, who describes the pațiccasamuppāda as “simple.” ²⁴³ As in the Mahāpadānasutta (14), the two final links, viz. avijjā and sañkhārā, are omitted, and it is possible that this represents the traditional view of the Dīgha-bhānakas who were responsible for preserving the Dīgha-nikāya.

Fragments of the Sarvāstivādin Sanskrit version of this sutta have been discovered in Turkestan, and published. ²⁴⁴

The Mahāparinibbānasutta (16) contains the story of the wanderings of the Buddha during the last few months of his life, his death, and the distribution of the relics after his cremation. It fits together so closely with the story, related in the Cullavagga of the Vinaya-piṭaka, of the first council held immediately after his death, that it seems clear that both stories are based upon what was originally one connected narrative. ²⁴⁵

There are clear indications of additions to the sutta, and detailed analyses have been made to distinguish older elements from later ones. ²⁴⁶ It has been shown that many of the episodes in the sutta occur in other canonical texts, ²⁴⁷ and the presence of the word Mahā- in the title implies that there existed at one time a Cūlaparinibbānasutta which lacked many of the later additions. ²⁴⁸ In some cases it can be shown how additions occurred. The earthquake which took place at the end of the speech in which the Buddha renounces his life occasions a statement about the eight causes of earthquakes, ²⁴⁹ which leads to the inclusion of a list of eight assemblies, eight positions of mastery, and eight stages of release, all of which also find their place in the Añguttara-nikāya. ²⁵⁰ A misunderstanding of a verse which seems to include a reference to crossing a river leads to an unnecessary insertion of a story about the Buddha crossing the Ganges by supernormal power. ²⁵¹ There is a reference to the future greatness of Pātaliputra which is almost certainly a prophecy made with the benefit of hindsight, and so must date from the second half of the 4th century B.C. ²⁵²

The Sarvāstivādin version of this sutta has been published. ²⁵³ This too contains some of the additions, including the list of eight earthquakes, which shows that the expansion of the original shorter form of the story must predate the schism between the Sarvāstivādins and the Theravādins. A fragment of a Dharmaguptaka version has also been published. ²⁵⁴

In the course of the wanderings leading up to the Buddha’s death recounted in the Mahāparinibbānasutta (16), the Buddha tells Ānanda of the former greatness of the city of Kusinārā, which the latter had said was an unworthy place for the Buddha to die in, and the story is told of King Sudassana, who had inhabited Kusinārā when it was a great city. The Mahāsudassanasutta (17) starts with the same words as are given at that point in the previous sutta, but goes on to tell the full story. It seems to be aimed towards the laity rather than the bhikkhus when it tells of the king’s greatness being due to generosity (dāna), self-conquest (dama), and self-control (samyama). ²⁵⁵

The Sanskrit version of this story is included in the Mahāparinirvāṇasūtra in the place where Ānanda first referred to the city. In both the Sanskrit and the Pāli versions it has the character of a Jātaka story, for the Buddha identifies himself with King Sudassana in a previous birth.

The Janavasabhasutta (18) is also an expansion of a portion of the Mahāparinibbānasutta (16), and deals with the rebirths of the Nādikas which are mentioned in the latter sutta. Like the Mahāsudassanasutta (17), which was probably an addition intended to explain the reference to Kusinārā, this sutta shows signs of being a later insertion, because it not only contains miraculous and mythical elements, but also treats doctrinal matters in a numerical way, e.g. the four ways to supernormal power (iddhipādā ) and the four inceptions of mindfulness (satipatthānā). Both features are typical of later texts.

The Mahāgovindasutta (19) also contains such miraculous elements, and tells of the joy in heaven at the appearance of new gods, who have arisen as such because of the good kamma performed by Gotama’s followers. Brahmā praises Gotama and gives his view of the ideal brahman. The inclusion of some mnemonic verses suggests that an older story has been adapted for Buddhist purposes, and the basis of the sutta seems to be the preaching of karunā-jhāna as a means of gaining direct realisation of Brahmā. The idea of realisation of Brahmā leads, as in the Tevijjasutta (13), to the enunciation of the four brahma-vihāras.

A Sanskrit version of this story, with some differences, is found in the Mahāsānghika-Lokottaravādin Mahāvastu. ²⁵⁶

As in the case of the two previous suttas, the Mahāsamayasutta (20) is concerned with the gods, this time the minor ones, who express their support for Buddhism. The greater part of the sutta is a list (in verse) of names of gods, of which a portion seems to be a reworking of a list which occurs in the Āṭānāṭiya-suttanta (32). The way in which it occurs seems clumsy, which suggests that it is an insertion, and had an independent existence before the sutta was composed. There is a prologue containing verses uttered by the four gods of the pure abode (Suddhāvāsakāyikā devā). There is also a framework, put into the mouth of the Buddha at the beginning (after the prologue) and the end. This has a separate existence elsewhere in the canon, ²⁵⁷ and confirms that the sutta is made up from parts that were originally independent.

There is a Sanskrit version of this sutta extant, which has the name Mahāsamājasūtra. ²⁵⁸ Since, according to the dictionaries, samaya and samāja can both mean “assembly,” ²⁵⁹ this may simply be a variation of vocabulary. It is, however, noteworthy that samāja occurs with the spelling samaya in the Shāhbāzgaṛhī version ²⁶⁰ of the Asokan inscriptions, and the possibility remains that the Sanskrit name is an incorrect back-formation from a Prakrit form. ²⁶¹

The differences between the Sanskrit and the Pāli versions of this sutta are not great. ²⁶² In the Sanskrit sūtra the gods in the prologue belong to the Brahmakāyika class, not the Suddhāvāsakāyika; the Buddha is said to be preaching to his disciples about nirvāna, while no sermon is mentioned in the Pāli; moreover, the Pāli version lacks the list of attributes conferred upon the arhats who make up the Buddha’s audience. These differences are precisely those which might be expected to arise in the course of an oral tradition, when a thera preaching a sermon might well change a number of unimportant matters each time he recited a sutta.

The Sakkapan̄hasutta (21) is the last of the series of mythological suttas, and culminates in the conversion of Sakka, the king of the gods, to Buddhism. It would seem to have been intended as a piece of propaganda, to persuade nonBuddhists not to be afraid of the new religion since even the gods of their own religion accepted it. The sutta tells how Sakka visits the Buddha, is allowed to ask him certain questions about ethics and psychology, and is so persuaded by his answers that he is converted. The sutta calls itself a veyyākaraṇa (explanation) at the end, ²⁶³ although Buddhaghosa includes it among the texts which he calls vedalla. ²⁶⁴ This perhaps means that in the early period of Buddhism it was included in that category of text in the nine-fold canon. It would seem to be an early sutta, since it is quoted by name in the Saṃyutta-nikāya. ²⁶⁵

Fragments of the Sarvāstivādin version have been discovered and published. ²⁶⁶

The doctrine expounded in the Mahāsatipaṭṭhānasutta (22) is perhaps the most important after the Aryan Path in early Buddhism, and this sutta is the oldest authoritative statement of the doctrine. The sutta is in two parts. The first part concerns mindfulness, and is the same as the Satipaṭṭhānasutta (10) of the Majjhima-nikāya. The second part is a term by term account of the four noble truths, and the addition of this second part perhaps explains why the sutta is called Mahā-, the other sutta in the Majjhima-nikāya being by implication the Cūla- sutta. The addition uses a style of definition which is similar to that found in the Abhidhamma-piṭaka, and its unusually detailed character has led to the suggestion that it is perhaps a fragment from an early commentary which has crept into the canon. ²⁶⁷

The Pāyāsisutta (23) refers to matters which took place after the death of the Buddha. It tells how the chieftain Pāyāsi visited the Buddha’s disciple Kassapa to question him about rebirth and kamma, neither of which Pāyāsi believed in. There is no mention of the Buddha, as either alive or dead, but in his commentary upon the Vimāna-vatthu Dhammapāla states that the sutta was recited after the erection of the thüpas over the relics. ²⁶⁸ During the course of the discussion, Pāyāsi’s arguments are all refuted, and he is converted to Buddhism. The climax of the sutta comes when a messenger arrives from the gods to teach the doctrine of generosity (dāna) by laymen. It is possible that the death of the Buddha had led to a falling-off in the gifts made to the Sangha, and the gods were employed as a fit means of reminding those who had followed them, before both gods and laymen became Buddhists, of the need to be generous.

3.1.3. The Pāṭikavagga

The third vagga is the most miscellaneous section, and its contents indicate that definite doctrinal beliefs about the nature of a Buddha had become much developed by the time it was collected together. This suggests a comparatively ²⁶⁹ late origin for some, at least, of the suttas it contains.

The Pāṭikasutta (24) falls into two parts. In the first, the Buddha discusses mystic wonders, while in the second he deals with the origin of things. Both subjects are treated elsewhere-the wonders in the Kev addhasutta (11) and the origin of things in the Aggaññasutta (27). The sutta begins with the mention of Sunakkhatta Licchaviputta, who had left the Buddha because he would not perform miracles or discuss origins, but the emphasis changes, and the Buddha becomes eager to show that he has worked wonders. He then begins to discuss origins, but suddenly with an abrupt change of subject he makes a comment upon the stage of release called “beautiful.” ²⁷⁰

The Udumbarikasīhanādasutta (25) deals with the same subject as the Kassapasīhanādasutta (8), i.e. the nature of asceticism both true and false, but it treats it in a more elaborate way. The Buddha discusses the different kinds of asceticism and their evil effects, and explains the life of a real liver of the the holy life (brahmacārin).

The Cakkavattisīhanādasutta (26) tells the story of the righteous monarch, of the corrupt conditions which followed after his time and led to a shortening of the human lifespan, and then a general improvement in morals which will lead to a lengthening of the lifespan again. When the lifespan of humans has grown to 80,000 years, the Cakkavattin Sankha and the future Buddha Metteyya will arise. Such an idea as future Buddhas would seem to represent a later stratum of Buddhist thought, ²⁷¹ which suggests that this is a later text.

The Aggaññasutta (27) represents another rejection of the brahmanical claim to be the best caste. To make this rejection, the Buddha tells a story about the beginning of the world, although his reluctance to talk about origins is made clear in the Pāṭikasutta (24). The Buddha gives etymological explanations for the arising of King Mahāsammata, the khattiyas, the rāja, brāhmanas, jhāyakas, ajjhāyakas, vessas and suddas. ²⁷² The sutta shows that righteousness is above lineage, although of those who do put their trust in lineage the khattiya is best. The mythological content of this sutta suggests that it is not early.

The Sampasādanīyasutta (28) is an elaboration of a passage in the Mahāparinibbānasutta (16), where Sāriputta is rebuked for saying that the Buddha was the best of all Buddhas. In this sutta, Sāriputta goes on to enumerate the various excellences (ānuttariyas) of the Buddha’s teaching. The sutta mentions in the main only such dogmata as relate to spiritual practice. Since no claim is made for the completeness of such beliefs, it would probably be unwise to try to draw conclusions from any omissions. The sutta ends with the Buddha telling Sāriputta to repeat his discourse frequently in the presence of any persons who may have doubts about him. Such a sentiment seems late, and suggests that the sutta is a late one.

The Pāsādikasutta (29) is said to have been delivered soon after the death of Nāthaputta, the Jain leader, as a result of the quarrelling which his death caused among his followers. In contrast to this turmoil, the Buddha tells of the conditions of a perfect religion, and gives information about the characte:istics of a Tathāgata and his powers. The sutta contains a collection of already existing doctrinal points, and seems to suggest nothing new.

The Lakkhaṇasutta (30) is an elaborate piece of Buddhology, describing in detail the 32 physical characteristics of the Buddha. In the history of Buddhism this iconographic development has often been supposed to be a late phenomenon and one which rather tends to Mahāyāna ideas. The lateness of the doctrine in this sutta is matched by the lateness of the metre of its verses. It shows a greater variety of metres ²⁷³ than any other canonical text, all of them being, moreover, either new classical types of metre or classical forms of old metres. ²⁷⁴

The orthodox tradition records that the verses are not as late and authentic as the bulk of the canon, and in his commentary Buddhaghosa attributes them to Ānanda. ²⁷⁵

The Sigālovādasutta (31) is really a poem with a prose commentary, and Buddhaghosa calls it a gihivinaya. ²⁷⁶ It tells of the duties of the Buddhist layman. The sutta relates how Sigāla practises the worship of the six directions (disās) as a result of his father’s death-bed wish. The Buddha sees him doing this, and interprets the six quarters in a Buddhist manner, in a way in which his doctrine of love and goodwill between man and man is set forth with more detail than elsewhere in the canon.

The Āṭānātiyasutta (32) is a saving chant (rakkhā-manta) to get rid of evil spirits, and is included among the list of protective utterances (parittās). ²⁷⁷ It deals with the sending away of gods, gandhabbas and yakkhas if they attack laymen and laywomen. It contains two lists of supernormal beings, one long and one short. As has already been noticed, ²⁷⁸ a reworking of a portion of the longer list, which begins with an invocation to the seven Buddhas, and then the gods of the four quarters, is found in the Mahāsamayasutta (20). Fragments of a Sarvāstivādin Sanskrit Āṭānāṭikasūtra have been found in Turkestan and published. ²⁷⁹

The occasion for the Sangitisutta (33) was the feuding that took place among the Jains after the death of their leader Nāthaputta at Pāvā, just as was the case for the Pāsādikasutta (29). The name of the sutta suggests that it represents a recitation of doctrinal matters, perhaps as an attempt to provide a summary of the doctrine as a precaution against a comparable confusion arising in the Buddhist Sangha. The title, the fact that the authorship is attributed to Sāriputta, and the nature of the text, which is numerical on the lines of the Anguttara-nikāya, all suggest that the sutta is a late one.

The nature of the text is reminiscent of the mātikās of the Abhidharmmapitaka and it is noteworthy that the Sarvāstivādins have in their Abhidharmapitaka a text called Sangitiparyāya, ²⁸⁰ which seems to be a commentary upon the Sangitisūtra, of which fragments have been found in Turkestan. ²⁸¹ The arrangement of the material in the Sangitisutta seems to be entirely haphazard. The aim appears to be the collection of all points of doctrine, and no account is taken of the fact that there is overlap and duplication, e.g. in the case of the five khandhas and the five upādāna-khandhas. ²⁸²

The Dasuttarasutta (34) is also ascribed to Sāriputta, and seems to be a systematic selection from the previous sutta. ²⁸³ Some of the wording is identical with the Sangitisutta. It follows the same pattern of sections dealing with subjects from one to ten, but has ten in each section, thus giving a total of one hundred. Fragments of the Sanskrit Daśottarasūtra have been found and published. ²⁸⁴

A comparison with the Chinese translation of the agamas shows that all the suttas of the Pāli Dīgha-nikāya have a parallel, although not necessarily a close parallel, in the āgamas. ²⁸⁵ They are not, however, all found in the Dīrghāgama, and their presence elsewhere, usually in the Madhyamāgama, ²⁸⁶ shows that the distinction between the two āgamas was not yet closely delineated at the time when the various sects of Buddhism separated. The reverse also holds true, and not all the sütras in the Dīrghāgama are found in the Pāli Dīgha-nikāya. The Sarvāstivādin version of the Dīrghāgama includes the Catuṣpariṣatsūtra, although this is found in the Mahākhandhaka of the Vinaya-piṭaka in the Pāli canon. ²⁸⁷

3.2. The Majjhima-nikāya

The Majjhima-nikāya ²⁸⁸ contains 152 suttas arranged in three groups of fifty (pannāsa). The pannāsas (1-50, 51-100, 101-52) are further divided into vaggas of ten suttas each, except for the last vagga but one, which has twelve suttas.

There is no obvious order followed in the arrangement of suttas, except that suttas with a similar name are sometimes grouped together. There are seventeen pairs of suttas with the same name, distinguished by the prefixes Mahā- amd Cūla-. ²⁸⁹ In most cases these pairs are together, with Mahā- sometimes coming first, and sometimes Cūla-, although in one case the pair are far apart in different vaggas. ²⁹⁰ One vagga (Mahāyamakavagga) consists of five such pairs, while the Cūlayamakavagga, on the other hand, contains only two pairs. There are several suttas entitled Mahā- to which there is no Cūla- counterpart.

Of such a pair, the Mahā- version is usually longer than the Cūla-, but not always. The Cūlahatthipadopamasutta (27) is longer than the Mahāhatthipadopamasutta (28), and the Cūlamāluṅkyasutta (63) is longer than the Mahāmāluñkyasutta (64). Sometimes such linked suttas have nothing in common but the reason for the name, e.g. the two Hatthipadopamasuttas (27 and 28) both make use of the simile of the elephant’s footprint, but do not have the same subject matter. On the other hand, the two Dhammasamādānasuttas ( 45 and 46) have much in common. Similarly, of the two Gopālakasuttas the Cūlaversion (34) makes use of one simile deriving from “cowherd” (gopālaka), while the Mahā- (33) version puts forward eleven.

The names of the vaggas to some extent reflects their contents, some being called after the first sutta in the vagga. The Gahapativagga consists of suttas in each of which, except for one sutta (57), the Buddha addresses householders. The third vagga surprisingly has no title. Since six of the ten suttas have the word upama in their titles, it would have been appropriate to call it Opammavagga. ²⁹¹

3.2.1. The Mūlapaṇṇāsa

A number of the suttas in this group are autobiographical, and give information about the Buddha’s career which is lacking elsewhere in the canon. The Ariyapariyesanasutta (26) contains an account of the Buddha’s activities between the time he left home and the conversion of the group of five bhikkhus, including his visits to the teachers Ālāra Kālāma and Uddaka Rāmaputta. ²⁹² The Mahāsaccakasutta (36) gives an account of the Buddha’s attempts to gain enlightenment by means of asceticism, and of his decision to give up such methods when they failed to gain the goal he desired. The Mahāsīhanādasutta (12) contains similar details.

Of the two Sāropamasuttas, the Cūla- version (30) is longer than the Mahāform (29). Both deal with the same subject (the simile of the pith) and both have the same summary at the end. It is possible that one is not an elaboration of the other, but that both are different versions of the same sutta. ²⁹³ An important addition in the Cūla- version of the sutta is an attempt to enumerate the states which are higher than knowledge and vision (ñāṇadassana). These are set out as the jhānas and vimokkhas.

The Satipaṭthānasutta (10) contains word for word the first part of the Māhāsatipaṭthānasutta of the Dīgha-nikāya (22), ²⁹⁴ and is possibly to be regarded as the Cūla- version of that sutta because it lacks the second portion dealing with the four truths. This has a separate existence in the Majjhimanikāya (141). ²⁹⁵

The Alagaddūpamasutta (22) takes its title from the simile of the watersnake which if handled properly is unable to harm the person who holds it. It also contains the parable of the raft, which is invaluable for crossing a river, but must be abandoned when it has served its purpose. In the sutta the Buddha attacks the view held by some that the world and the self (attā) are the same thing. This seems to be a reflection of the Upanisadic view of a World-Ātman. ²⁹⁶

Some of the suttas include ideas which have already been noted as comparatively late when they occurred in the Dīgha-nikāya. The Mahāsīhanādasutta (12) gives an account of the ten powers of a Tathāgata, while in the Cūlataṇhāsañkhayasutta (37) Moggallāna visits Sakka’s heaven and shakes the Vejayanta Palace with his big toe by means of supernormal power. The Brahmanimantaṇikasutta (49) shows the Buddha challenging a Brahmā, and proving that the Brahmā’s supernormal powers are inferior to his own.

A number of the suttas enumerate ascetic practices, e.g. the Mahāsīhanādasutta (12), or give information about the Jains, e.g. the Cūladukkhakkhandhasutta (14), while the Mahādukkhakkhandhasutta (13) in a series of similes explaining the consequences of the pleasures of sense gives a long list of punishments and tortures.

3.2.2. The Majjhimapaṇṇāsa

In this vagga too there are suttas dealing with other religious groups. The Upālisutta (56) tells how the householder Upāli, a Jain layman, visited the Buddha to refute him but became converted to Buddhism, and resisted an attempt by Nātaputta the Jain leader to persuade him to recant. The Kukkuravatikasutta (57) tells of bovine and canine ascetics whose future existence as cattle, dogs, or hell-dwellers is foretold. The uselessness and cruelty of asceticism are also dealt with in the Kandarakasutta (51), and the Apaṇṇakasutta (60).

Two suttas in this paṇnāsa tell of discussions which took place after the death of the Buddha. In the Madhurasutta (84) and the Ghotamukhasutta (94) an enquiry about the present whereabouts of the Buddha meets the reply that the Buddha has now attained final nibbāna (parinibbuto). ²⁹⁷ There are other indications that some suttas are later than others, even among those which are attributed to the Buddha himself. The Bhaddālisutta (65) discusses the reason for the growth of rules of training (sikkhāpadāni), and the Buddha points out that the need for an increase in rules only arises when the Sangha is great and of long standing.

In several suttas the Buddha continues his attack upon the brahmans. In the Assalāyanasutta (93) he examines the brahmans’ claim to be the best caste, and in the Madhurasutta (84) his disciple Mahākaccāna does the same. A number of the suttas in this section are narrative stories. The Raṭṭhapālasutta (82) tells of the young prince who became a wanderer, and of his father’s attempts to entice him away from religion when he visits the palace to beg years later. The verses with which he describes the uselessness of wealth and position also appear in the Theragāthā. ²⁹⁸ The Makhādevasutta (83) relates the story, also found in the Jātaka collection, ²⁹⁹ of the king who regards his first grey hairs as messengers of death, and so abandons his kingdom and becomes a wanderer. The Añgulimālasutta (86) tells the story of the robber who attempted to kill the Buddha but was instead converted by him. His verses also appear in the Theragāthā. ³⁰⁰

The Brahmāyusutta (91) resembles the Ambaṭṭhasutta of the Dīgha-nikāya (3), in that it tells of a brahman visiting the Buddha to see if he really possessed the 32 marks of a great man. As in the case of Ambaṭtha, supernormal powers are employed to enable him to see the two marks which are not normally visible. The brahman Brahmāyu then asks the Buddha how one becomes a possessor of the three Vedas (tevijja), and he is given a Buddhist interpretation of the brahmanical term. ³⁰¹

Two suttas are of great interest because they are occasions for the Buddha to be questioned about the existence of devas and adhidevas, i.e. beings superior to the devas (gods). In the Sañgāravasutta (100), the Buddha is asked if devas exist, and replies that adhidevas, i.e. super-gods, certainly do, so that by implication the existence of devas can be deduced. ³⁰²

“But now, Gotama, do devas exist?” “I certainly know for a fact, Bhāradvāja, that super-devas exist.” “But why, Gotama, when asked if devas exist, do you say that you know certainly for a fact that super-devas exist? Even if this were so, is your answer not useless and misleading?” “If anyone, when asked if devas exist, replies that super-devas exist, then someone with sense can deduce that devas must exist.” “Then why, Gotama, did you not say in the first place that devas exist?” “It is firmly accepted in the world, Bhāradvāja, that devas exist.” (II 212-13)

In the Kannakatthalasutta (90) he is asked if adhidevas exist. He concedes that they do, but his follower Ānanda makes it clear that the pre-eminent position of such adhidevas was of no importance. Elsewhere in the canon, however, the word atideva is used of the Buddha, and since this and adhideva are synonymous it seems likely that in the Sangāravasutta the Buddha was referring to himself as being superior to the devas, and in the Kann akatthalasutta it was the brahmanical adhidevas which were being disparaged. ³⁰³

3.2.3. The Uparipannnāsa

As already noted, this pannāsa, despite its name, contains 52 suttas, since there are twelve suttas in the fourth vagga. It has been suggested that suttas (132), (133) and (134), which are all three by followers of the Buddha, and are based upon (131), should perhaps be regarded as forming one sutta, which would make the number correct. ³⁰⁴ On the other hand, since the Saccavibhangasutta (141) seems to be no more than an early commentary on the Buddha’s first sermon about the four noble truths, and the Dakkhināvibhangasutta (142) repeats Ānanda’s pleading of Mahāpajāpatī’s cause found in the Vinaya, ³⁰⁵ it is possible that both suttas are interpolations. ³⁰⁶

It has been pointed out that this section contains more suttas for which a late origin can be postulated than the other two pannāsas. ³⁰⁷ In particular, the suttas in the Vibhangavagga (131-42) foreshadow the method of analysis and classification found in the Vibhanga in the Abhidhamma-pitaka. The Isigilisutta (116) gives a list of paccekabuddhas in prose and then a longer list in verse. The concept of the paccekabuddha seems to be a late borrowing into Buddhism from some other religion. ³⁰⁸ The Acchariyabbhutadhammasutta (123) describes the conception and birth of a Buddha in words very similar to those found in the Mahāpadānasutta of the Dīgha-nikāya (14), where they are applied to the Buddha Vipassin, and in the Nidāna-kathā of the Jātaka. ³⁰⁹

The Gopakamoggallānasutta (108) is said to have been delivered not long after the Buddha’s parinibbāna ³¹⁰ by Ānanda, who was asked whether the Buddha had designated any of his followers as a support (patisarana) to the Sangha after his death. Ānanda replied that no-one had been designated, but the Sangha was nevertheless united because the dhamma was the support. ³¹¹ The Sāmagāmasutta (104) tells of the quarrelling which took place among the Jains after the death of their leader Nātaputta. As in the case of the Sangitisutta of the Dīgha-nikāya (33), this led to the Buddha promulgating some sets of numerical categories.

A number of suttas show disciples preaching. In suttas (132) and (133) Ānanda and Mahākaccāna repeat and explain the Buddha’s words in the Bhaddekarattasutta (131). In the Dantabhūmisutta (125) and the Bhūmijasutta (126) prince Jayasena approaches the novice Aciravata and the venerable Bhūmija respectively, and asks to hear the dhamma. Both preachers subsequently visit the Buddha and are told how they could have improved their teaching.

The Cūlakammavibhangasutta (135) and the Mahākammavibhangasutta (136) both give an account of the working of kamma including a description othe suffering in hell (niraya) which awaits a wrong-doer. Another account of hell, with even greater detail, is found in the Bālapaṇḍitasutta (129). Comf parable descriptions of suffering in hell are found in Jain texts, ³¹² and presumably represent a more popular conception of the doctrine of kamma.

Most of the suttas in the Vibhangavagga (131-42) follow a distinctive pattern. The subject is first presented as a brief statement (uddesa) and is then followed by an exegesis (vibhanga), supplied in four suttas by leading theras, a detail which perhaps correctly reflects the Buddha’s way of teaching. ³¹³ The Saccavibhangasutta (141), however, after an initial statement about the setting of the wheel of the Dhamma in motion, provides the same exegesis of the four noble truths as is found in the Mahāsatipaṭṭhānasutta in the Dīgha-nikāya (22). ³¹⁴

It has been established that the Chinese translation of the Madhyamāgama is based upon a Sarvāstivādin original. ³¹⁵ An examination of the contents of the Chinese version shows that, of its 222 sūtras, 98 correspond to suttas in the Pāli Majjhima-nikāya, 79 in the Añguttara-nikāya, 10 in the Saṃyutta-nikāya, and 9 in the Dīgha-nikāya. ³¹⁶ Since the majority of texts found in Turkestan belong to the Sarvāstivādin school, it is remarkable that comparatively few portions of Madhyamāgama texts from that area have as yet been published. ³¹⁷

3.3. The Saṃyutta-nikāya

The Saṃyutta-nikāya ³¹⁸ gains its title from the fact that its suttas are grouped together (saṃyutta) according to their contents. There are 56 samyuttas, arranged in five vaggas. They contain 2,889 suttas in all, in the European edition, although Buddhaghosa states that there are 7,762 suttas. ³¹⁹

The arrangement is made by (1) subject or doctrine, (2) class of god, demon, or man, (3) some prominent person as speaker or hero, ³²⁰ e.g. the Sakkasamyutta (11) contains suttas where Sakka plays a part, while the Bojjhangasamyutta (46) is composed of those suttas in which the seven elements of enlightenment are discussed. This method does, however, lead to a great deal of repetition, e.g. the Salāyatanasamyutta (35) contains 207 suttas about the six senses. It is possible that the repetition arises from the fact that there were in existence many suttas on the subject, collected from various individuals, different vihāras, etc., all of which were of equal religious merit, and all of which deserved to be included in the collection. ³²¹ This method of arrangement does mean that the Samyutta-nikāya contains some of the most important Buddhist teaching, e.g. the Saccasamyutta (56) contains 131 suttas dealing with different aspects of the four truths.

The arrangement by samyuttas represents an obvious editorial practice, and the existence of a number of the suttas elsewhere in the canon probably indicates a conscious selection of material to group together in this way. There is inevitably an overlap with the Anguttara-nikāya, because some of the samyuttas deal with numerical subjects. Sometimes a comparison with the version of a sutta which occurs elsewhere shows quite substantial differences. A version of the Parinibbāna story is included in the Brahmasamyutta (6), because Brahmā gives a eulogy over the dead Buddha just as in the Mahāparinibbānasutta of the Dīgha-nikāya (16), but it reverses the order of the utterances by Ānanda and Anuruddha, omits the attainment of saññāvedayitanirodhasamāpatti, and also Ānanda’s mistaking of this for parinibbāna. ³²² These differences show that changes have been introduced into one or other, or both, of the narratives at some time, but the tradition has not changed since then, doubtlessly because the Dīgha-bhānakas and Samyutta-bhānakas preserved their texts independently.

3.3.1. The Sagāthavagga

This vagga contains the great majority, but by no means all, of the suttas which contain verses. ³²³ The layout indicates that a certain amount of editorial work had already been carried out before the vagga was fixed in its present form.

The first sutta of the Devatāsamyutta (1) has a short introduction describing how a devatā visited the Buddha and asked a question. The second sutta has an abbreviated introduction, while the third and subsequent suttas in the vagga ³²⁴ merely state that a devatā uttered the verse. Similarly, the first sutta of the second vagga has an introduction, while the second and third suttas have only a single sentence. The remainder of the vagga consists simply of verses, which, because they are in the Devatāsamyutta, must be presumed to have been originally uttered by a devatā. A similar pattern can be seen in later vaggas.

The subject matter of the Sagāthavagga is predominantly ethical, and most of the suttas deal with the Buddhist ideal of life. Many of the verses are of a question and answer nature. Some are riddles, while others are puns, e.g. sutta I.2.2. depends upon the two meanings, religious and secular, of upadhi: a man does not rejoice who is without upadhi (material possessions), but a man does not lament who is without upadhi (the clinging to existence which results from having material possessions). Sometimes both a religious and secular answer is given to the same question, e.g. in sutta I.8.4 there are two answers to the question of what is the best of things that rise and things that fall: seed and rain, and also knowledge and ignorance. In a number of suttas where only a religious answer to a question is given, it would seem probable that the secular answer has been omitted and forgotten, leaving only the paradoxical religious answer. The fact that some of the verses are also found in Jain literature prove their non-Buddhist nature. ³²⁵

The Bhikkhunīsamyutta (5) collects together stories about bhikkhunīs who were tempted by Māra. Many of the verses are also found in the Therīgāthā, ³²⁶ but some do not occur there, and others are ascribed to different therīs. ³²⁷ Once again, the solution to the problem probably lies in the fact that the early tradition was doubtful about the form of the verses and the speakers, and the Saṃyutta-bhānakas and Khuddaka-bhānakas transmitted independently a differing tradition. The uddāna to a Sanskrit version of this samyutta has been discovered in Turkestan. ³²⁸ It agrees with the Pāli Bhikkhunīsamyutta in its contents, showing that the differentiation of the bhāṇaka tradition pre-dates the schism between the Theravādins and Sarvāstivādins.

It is debatable whether the prose of the introductions to the verses in the Sagāthavagga is as old as the verses themselves, or whether it represents the work of the sangitikāras at the first or a subsequent council. Since only the verses of the Bhikkhunīsamyutta are found in the Therīgāthā, it is clear that the prose did not have canonical status for the Khuddaka-bhānakas. The same is true of Vangīsa’s verses, found in the Vangīsatherasamyutta with prose introductions, but without them, and with some differences, in the Theragāthā. ³²⁹ A fragment of the Vangisasamyukta has been found in Turkestan, ³³⁰ and portions of the Dhvajāgrasūtra ³³¹ and Candrasūtra ³³² have been discovered and published.

3.3.2. The Nidānavagga

The suttas in this and the subsequent vaggas of the Samyutta-nikāya deal mainly with epistemology, metaphysics and psychology. The vagga takes its name from its first samyutta, the Nidānasamyutta (12), which in 93 suttas deals with the subject of the twelve nidānas, i.e. the twelve links of the chain of cause and effect (paticcasamuppāda). A number of sütras from the Nidānasamyukta have been found in Turkestan and published. ³³³

The Anamataggasamyutta (15) in the same vagga contains 20 suttas which include the words “endless ³³⁴ is this samsāra,” with the exception of suttas (15.16-18), which are so abbreviated in the European edition that the words do not appear therein. ³³⁵

3.3.3. The Khandhavagga

This vagga takes its name from the Khandhasamyutta (22), which is its first samyutta. Despite its division into three groups of 50 (pañnāssa), the Samyutta actually contains 158 suttas. In them the subject of the khandhas is examined with great repetition, and the impression is given that every sutta which mentions the khandhas, and a few which do not, ³³⁶ has been collected together. Many can be found elsewhere in the canon. The Sāriputtasamyutta (28) contains 10 suttas which record ten sayings of Sāriputta, some of them a single sentence presented in a stereotyped setting. The Nāgasamyutta (29) contains 50 suttas, of which the last 30 are merely sketched out in a very abbreviated form, dealing with the different sorts of snake (nāga) and explaining the reasons for their birth and practices.

3.3.4. The Salāyatanavagga

The first samyutta of this vagga is the Salāyatanasamyutta (35), which despite the fact that it is divided into four groups of 50 (paññasakas) seems to contain 207 suttas. Many of the suttas differ from their neighbours in one word only, e.g. sutta (33) explains that everything is subject to birth (jāti-dhamma), while suttas (34-42) replace jāti by jarā, vyādhi, marana etc. The Upasenasutta (69) tells of Upasena being bitten by a snake, but remaining unchanged in appearance because he had no ideas about the eye or the tongue or the mind being “mine.” Fragments of the Sanskrit Upasenasūtra, used as a charm against snake-bite, have been found in Turkestan and published. ³³⁷ The Avassutasutta (202) of the same Samyutta tells of Moggallāna preaching to his fellow bhikkhus, at the request of the Buddha who was indisposed. Moggallāna chose for his subject the letting in (avassuta) and not letting in (anavassuta) of harmful influences through the eyes, etc. Two fragments of a Sanskrit version of this sutta have been found and published. ³³⁸

The Asańkhatasamyutta (43) deals with the uncompounded (asainkhata), i.e. nibbāna, ³³⁹ and the path leading to it. Each synonym for nibbāna, and each means of attaining it, is the subject of a separate sutta, and the whole is abbreviated by an instruction to treat each of the synonyms under the same 45 heads as asainkhata. ³⁴⁰

3.3.5. The Mahāvagga

A number of the Samyuttas in this vagga deal with matters of vital importance to Buddhism, including the Maggasamyutta (45) which deals with the eightfold path, the Bojjhangasamyutta (46), which treats the seven elements of enlightenment, the Satipatṭhānasamyutta (47), dealing with the bases of mindfulness, and the Jhānasamyutta (53), which deals with trance, but in so abbreviated a way that several of the vaggas consist of nothing more than uddāna verses with instructions to expand them. ³⁴¹

The final samyutta of the fifth vagga is the Saccasamyutta (56). Here among a number of very abbreviated suttas is found the Dhammacakkappavattanasutta (56.11), which also occurs in the Vinaya-piṭaka. ³⁴² Several suttas tell of the difficulty of hearing the truth, including the Chiggalasuttas (56.45, 47-48), in which comparison is made with shooting an arrow through a keyhole, or the possibility of a blind turtle pushing his head through the hole in a yoke drifting in the sea.

As already noted, there is an overlap between the Samyutta-nikāya and the Añguttara-nikāya in the Pāli canon, which arises from the possibility of classifying suttas in two different ways. An examination of the contents of the Chinese translation of the Saṃyuktāgama, which is based upon the Sarvāstivādin version, ³⁴³ shows that it includes a number of sütras which in the Pāli canon appear in the Añguttara-nikāya. ³⁴⁴ This would seem to indicate that at the time when the āgamas were collected the distinction between the “connected” and “numerical” classifications was not very clearly drawn.

3.4. The Añguttara-nikāya

The name of the nikāya ³⁴⁵ means literally “the by-one-limb-more collection,” and the text is of the type already seen in the Dasuttarasutta of the Dīghanikāya (34). Buddhaghosa mentions 9,557 suttas, ³⁴⁶ but the number in the text as we have it is approximately 2,308 (or 2,363 ). ³⁴⁷ It is difficult to decide the number accurately because the text and the commentary do not always agree on the way in which the suttas are to be divided. ³⁴⁸ The suttas are divided into eleven sections (nipātas), arranged in about 160 vaggas. Each vagga contains as a rule ten suttas, although the highest number in any one vagga is 262 and the lowest seven.

The Añguttara-nikāya is called the Ekuttara-nikāya in the Milindapañha, ³⁴⁹ and the Sanskrit version is called the Ekottarikāgama. Each nipāta contains suttas dealing with subjects in some way connected with the number of the section, e.g. the eka-nipāta begins with the single form which enslaves a man’s heart, i.e. that of a woman, while the ekādasaka-nipāta ends with the eleven qualities which must be cultivated to ensure the comprehension and destruction of lust.

In some of the higher-numbered nipātas, i.e. from the chakka-nipāta onwards, combinations of smaller numbers are sometimes resorted to to get the higher designations, e.g. sutta (9.28)³⁵⁰ contains the five bhayas and the four sotāpattiyañgas, while sutta (11.11) ³⁵¹ contains three different sets of three dhammas and a set of two dhammas. With a few exceptions, however, the component parts of such composite suttas are not mere repetitions of the material found in the appropriate places in earlier nipatas. ³⁵² Sometimes the complements have no obvious connection. Sometimes some connection can be deduced, e.g. the five cetokhilas and the four satipatthannas in sutta (9.71), ³⁵³ where the latter are to be practised for the destruction of the former. Sometimes the component parts are opposites, e.g. the five dukkhas and the five sukhas in sutta (10.65). ³⁵⁴

Many of the suttas occur elsewhere in the canon, and it seems reasonable to assume that in part, at least, the Añguttara-nikāya consists of compilations of such numerical matters as already existed. Sometimes, however, it is impossible to be certain whether the Anguttara-nikāya is the borrower, or has been borrowed from, e.g. the eight cardinal rules for bhikkhunis occur as an independent sutta in the Anguttara-nikāya, ³⁵⁵ but occur as part of the Bhikkhunī-kkhandhaka in the Vinaya-piṭaka ³⁵⁶ in an entirely appropriate place. On the other hand, the eight earthquakes which occur in the atthaka-nipāta ³⁵⁷ seem to be an insertion in the Mahāparinibbānasutta of the Dīgha-nikāya (16). ³⁵⁸ It is noteworthy that, while many of the enumerations in the Añguttara-nikāya occur without any introduction, in the case of the eight earthquakes the narrative story which leads up to their enumeration in the Mahāparinibbānasutta is also given in the Añguttara-nikāya, which seems to indicate that borrowing took place in both directions. The statement about eight types of assemblies which also occurs in the Mahāparinibbānasutta is extracted from there and given in the previous sutta in the Anguttara-nikāya. ³⁵⁹

There are certain noteworthy omissions in the Añguttara-nikāya, e.g. the three refuges (saranas) and the three characteristic properties (lakkhanas) are not found as such, nor are the four noble truths ³⁶⁰ nor the eight-fold path. ³⁶¹

With reference to the saranas and the lakkhanas it has been suggested that the omission arises from the fact that these doctrines had not yet become important by the time the tika-nipāta was closed. ³⁶² Such an explanation would not suit the four truths and the eight-fold path, and it seems likely that these doctrines were omitted because they were so important and so well-known that they were remembered in their original settings, and did not need to be transplanted to the Anguttara-nikāya. The fact that they had been included in the Samyutta-nikāya was perhaps also a factor contributing to their omission, although as has already been noted a certain amount of duplication has taken place. The fact that there is not more perhaps implies that there was a conscious “sharing” of suttas between the two bhāṇaka traditions, i.e. the Saṃyutta- and Anguttara-nikāyas are not entirely independent compilations. ³⁶³

One reason for the blurring of distinction between the Saṃyutta-nikāya and the Anguttara-nikāya can be seen in the way in which some vaggas contain suttas which all deal with one and the same subject, so that they resemble a samyutta. The ten suttas of vagga (1.1) deal with the relationship between husband and wife; vagga (1.14) has 80 suttas giving the names of the most prominent male and female disciples and their virtues; vagga (1.20) contains 262 suttas on different kinds of meditation leading to nibbāna; vagga (5.18) has ten suttas about upāsakas. There seem to be certain inconsistencies in the way in which the suttas and vaggas have been constituted. The first two suttas of vagga (1.6) are a continuation of vagga (1.5) and should have belonged to it. ³⁶⁴

Many of the suttas in the Anguttara-nikāya are of a non-narrative type, following a stereotyped formulaic presentation. This allows for interpolation into the appropriate nipāta at any time up to the final fixing of the canon. Some of the suttas at the end of the tenth nipāta seem to be additions, since they include groups of 20,30 and 40.³⁶⁵ It is probable that the whole of the eleventh nipāta is an addition. ³⁶⁶ It has hardly anything original in it, ³⁶⁷ and its three vaggas represent the smallest number of any nipāta. The Abhidharmakośavyākhyā states that the Ekottarikāgama has only ten sections, ³⁶⁸ and the same seems to be true of the Chinese translation. ³⁶⁹ It seems possible that some of the suttas at the very end of the tenth nipāta are, in fact, based upon the added eleventh nipata. After the suttas dealing with 20,30 and 40 , there are three suttas listing groups of ten. ³⁷⁰ These are, however, the ten qualities required to destroy lust, and are parallel to the eleven qualities which are found in the eleventh nipāta. ³⁷¹ It is likely that the compiler of the ekādasaka-nipāta realised that his material could easily be adapted to fit the earlier nipāta, and therefore compiled appropriate suttas and inserted them there.

It is likely that the Chinese version of the Ekottarikāgama belongs to the school of the Mahāsānghikas, and was translated not from Sanskrit but from some dialect of MIA or a mixed dialect of Prakrit with Sanskrit elements. ³⁷² An extract from the Saptamaithunasaṃyuktasūtra, quoted by Śāntideva in his Sikṣāsamuccaya, is in all probability to be traced to a Sanskrit counterpart to the Prakrit original of the Chinese translation. ³⁷³

Sanskrit fragments of the Ekottarikāgama have been found in Turkestan and published. ³⁷⁴ They all belong to the eka-nipāta, and correspond to the Nīvaraṇapahānavagga of the Añguttara-nikāya. A comparison with the Chinese version of the same sūtra shows a certain amount of deviation. Portions of a Sanskrit Ekottarikāgama, belonging to the first and second nipātas, have been found in Gilgit and published. ³⁷⁵ By chance there is an overlap with some of the Turkestan fragments, which makes it possible to state that there is a verbal correspondence between the two versions. This is perhaps not surprising, since they belong to the Sarvāstivādin (Turkestan) and Mūla-Sārvastivādin (Gilgit) schools.

3.5. The Khuddaka-nikāya

3.5.1. Khuddakapāṭha ³⁷⁶

As has already been noted, ³⁷⁷ both the Dīgha-bhānakas and the Majjhimabhānakas excluded this work from their lists of canonical texts, which probably means that it had not yet attained canonical status at the time when their lists were closed. The Khuddakapāṭha is also omitted from the Chinese translation ³⁷⁸ of Buddhaghosa’s Samantapāsādikā, which implies that either a recension of his work existed in which the Khuddakapāṭha was excluded, or the Chinese trans- lators were influenced by someone who knew that some schools did not accept the work as canonical. If we can be certain of our assumption that it had not yet gained canonical status when the schools of bhānakas closed their lists, then we can conclude that it is the latest text in the nikāya.

It consists of nine short texts. The first is the Buddhist confession of faith (the three refuges), the second a list of ten commandments for bhikkhus (veramañi-sikkhāpadāni), the third a list of 32 parts of the body, and the fourth a list of ten novice’s questions set out in numerical order-”What is one, etc. ?“including some of the most important Buddhist terms, such as the four noble truths and the eight-fold path. The other five pieces are short suttas, used for liturgical purposes: Mangalasutta, Ratanasutta, Tirokuḍdasutta, Nidhikanḍasutta, and Mettasutta. The first two and the last also occur in the Suttanipāta, ³⁷⁹ while the Tirokuḍdasutta occurs in the Petavatthu. ³⁸⁰ The first four compilations have parallels in the Vinaya-piṭaka, so that in effect all but the Nidhikanḍasutta are found elsewhere in the canon, and the whole work was probably compiled as an extract from the canon to serve as a handbook for novices. ³⁸¹ It probably owes its canonical nature to the fact that it is (nearly) all from the canon.

3.5.2. Dhammapada

The Dhammapada ³⁸² is a collection of 423 verses arranged in 26 vaggas, each containing between ten and 26 verses, except for the last, entitled Brāhmaṇavagga, which has 41 verses. The titles for the most part reflect the subject matter of the verses, e.g. the Pupphavagga (4) contains verses which with only two exceptions contain either the word puppha or the names of individual flowers, or include the key-word in a refrain, e.g. the Brāhmaṇavagga (26) includes the word brāhmaṇa in every verse except one (394), and has the refrain tam ahaṃ brümi brāhmaṇaṃ in 32 verses.

More than half the verses occur elsewhere in the canon, ³⁸³ although it is difficult to be certain whether they were collected together from those sources, or whether both took them from a third source. The existence of some of these verses in Jain or brahmanical texts ³⁸⁴ suggests that they were taken from the general store of floating verses which seems to have existed in Northern India in early times. The metres included in the Dhammapada are Tristubh (with some Jagatī), Vaitātīya (with some Aupacchandasaka), and Anustubh. Although small groups of verses, linked together by refrain, structure, or metre, clearly make small poems whose pre-existence is shown by the fact that they occur in the same form in other traditions as well, ³⁸⁵ for the most part the arrangement of verses, including their presence in particular vaggas, can be seen, by comparison with other texts, to be the choice of the compiler. In many cases verses are, by their content, appropriate to more than one vagga, ³⁸⁶ which accounts for the differences which can be observed between various recensions.

Although there are differences of opinion about the poetic quality of the verses in the Dhammapada, ³⁸⁷ the number of different versions we have of it are evidence for the popularity of this type of literature. From Turkestan there is a Sarvāstivādin version entitled Udānavarga, ³⁸⁸ which includes about 1050 verses, including uddānas, in 33 vargas. There is also an incomplete version in Gāndhārī Prakrit ³⁸⁹ which was acquired in Khotan. This perhaps belonged to the Dharmaguptaka school. ³⁹⁰ It contains, in its incomplete form, 342 verses of which many are fragmentary, and probably had 26 vargas, of which (4-5), (9), and (23-26) are lost. It is probable that both the Udānavarga and the Gāndhārī Dharmapada had originally about 360 verses in common with the Pāli Dhammapada, ³⁹¹ but although there are close parallels between these various texts, there are also considerable differences, e.g. the Sīlavarga of the Udānavarga (6) contains none of the verses found in the varga which probably had the same name in the Prakrit version, and the verses which the two texts have in common are spread over eight other vargas. The Pāli Dhammapada has no vagga entitled Sīla, and the verses it has in common with the (presumed) Sīlavarga of the Gāndhārī version are found in seven vaggas. ³⁹² Nevertheless, there is presumably some significance in the fact that the first three vargas of the Gāndhārī Dharmapada occur, in reversed order, as the last three of the Pāli version. ³⁹³

The Mahāsāngghika-Lokottaravādin Mahāvastu contains two vargas, one called the Sahasravarga, ³⁹⁴ the other unnamed but probably the Bhikṣuvarga, ³⁹⁵ and a few sporadic verses ascribed to the Dharmapada. ³⁹⁶ There has recently been published a complete Dharmapada belonging to the same school, which was found in Tibet. ³⁹⁷ It is in Prakrit, with some attempts made at Sanskritisation, and contains 415 verses in 22 vargas, of which only 13 have titles found in the Pāli version. ³⁹⁸ The relationship between these various versions and the four translations of the Dharmapada which exist in Chinese is not clear, although the oldest of these, dating from A.D. 244, seems to have some affinity with the Pāli Dhammapada, since it includes in its 39 sections all 26 of the Pāli version (9-32,34-35), with corresponding titles in the same order. The remaining verses seem to have come partly from the Udānavarga, and partly from another text or texts. Although the verses in the sections which correspond are not exactly the same as those in Pāli, the changes, additions, and omissions seem to be no greater than are usual in normal manuscript transmission. ³⁹⁹

3.5.3. Udāna

The Udāna ⁴⁰⁰ consists of eight vaggas, each containing ten suttas which end with a solemn utterance (udāna). This is usually a single verse, but it is occasionally more, and in a few cases is in prose.

To some extent, as in the case of other collections, some reasons can be seen for certain of the arrangements of suttas into vaggas, e.g, in vagga (1) all the verses contain the word brāhmana. In vagga (2) the udānas of nine of the ten suttas contain the word sukha, and the exception has the compound piyarūpam in common with the sutta following it, and is doubtless placed there by association of ideas. The same association of ideas probably accounts for other connections which can be seen in vaggas which do not have as obvious a theme, e.g. suttas (3.2) and (3.3) have a refrain in common; suttas (3.3) and (3.4) have the words pabbata and vedhati in common; suttas (3.8) and (3.9) have pādas in common.

The term udāna is one of the categories found in the nine-fold classification of the canon. Although Buddhaghosa states that in that classification the term actually refers to this text, ⁴⁰¹ it is more likely that it refers to a general type of literature, of which the Udāna is an example. In this connection it is of interest to note that Dhammapāla states ⁴⁰² that the majority of the Buddha’s udannas are in the Dhammapada. As has just been noted, the Sarvāstivādin version of the Dhammapada is entitled Udānavarga.

The commentator also states ⁴⁰³ that the verses of theras and therīs in the Theragāthā and the Therīgāthā are not udānas, but sīhanādas. This is probably an attempt to explain why the collection contains only 80 udannas. It is nevertheless clear that there are many other udānas in the canon, and it is uncertain why this collection should be so small. Some of the udānas which are found here occur elsewhere in the canon, with and without the narrative story which introduces them, e.g. the first four suttas of the first vagga occur at the beginning of the Mahākhandhaka of the Vinaya-piṭaka. ⁴⁰⁴ Sometimes the narrative portion of the sutta occurs elsewhere without the udāna, e.g. the first sutta of the fourth vagga. ⁴⁰⁵ As is usual, it is not always easy to see which text has borrowed from which, and in some cases it may be that both texts took the suttas directly from the great body of oral material which must have existed in the early days of Buddhism.

Although some of the udānas are in what was probably their true historical setting, it seems clear that some of the suttas have either been adapted, or even invented, to provide an occasion for the utterance. The first four suttas of the eighth vagga have the same story, which is merely a brief statement that the Buddha was preaching a sermon concerned with nibbāna. The fifth and sixth suttas of the sixth vagga also share an introductory story. It did not escape the commentator’s notice that the verses and narrative portions are separate, and he ascribes the narrative portion to Ānanda, ⁴⁰⁶ who at the first council introduced each of the suttas with the words evam me sutam.

The metres found in the udānas include Tristubh (and Jagatī), Vaitālīya (and Aupacchandasaka), Anuṣtubh, and Old Ārya.⁴⁰⁷ Some of them are prose, e.g. suttas (6.6-8,10) and (8.1,3-4). The udāna in sutta (3.10), which is found elsewhere, ⁴⁰⁸ is a mixture of prose and verse, some of the latter being prose which has been forced into verse form.

3.5.4. Itivuttaka ⁴⁰⁹

The Itivuttaka presumably takes its name from the opening words of each sutta:vuttaṃ h’ etam Bhagavatā. ⁴¹⁰ The text is stated by Buddhaghosa to be the work of that name included in the nine-fold classification of the canon. ⁴¹¹ It is, however, likely that as in the case of the Udāna, itivuttaka is simply a general term, and the text we possess is only one example of the type. The name appears in the form Itivrttaka in the Sanskrit lists, but this would appear to be an incorrect back-formation. ⁴¹²

The Itivuttaka consists of 112 suttas arranged in four nipātas, sub-divided into vaggas, and is written in a mixture of prose and verse. The prose is not intended as a narrative, giving the circumstances in which the verse was uttered, as in the case of the Udāna, but the prose and verse complement one another, so that the same idea, whether in the form of doctrine or admonition, is presented partly in prose and partly in verse. It seems likely that the prose passages are, in fact, explanations of the verses, despite the fact that they precede them. Sometimes only a single verse of a set has a prose counterpart, with nothing in the prose corresponding to the other verses. In many cases the prose and verse treat the same subject, but one is a supplement to the other. Occasionally the link between the two is very weak, e.g. in sutta (92=3.5.3) the verse states that the man who is full of lust (ejā) is far removed from one free from lust, whereas the prose tells of a man who might be near to the Buddha physically but far removed morally, and vice versa. The reason for the juxtaposition of prose and verse seems to be nothing more than the verbal similarity based upon “near” and “far.” ⁴¹³ The prose of sutta (93=3.5 .4) consists of the statement that there are three fires, i.e. those of lust, hate and delusion. The seven verses of the sutta constitute a short poem about those who by extinguishing the fires gain release from rebirth.

Like the Anguttara-nikāya, the Itivuttaka is arranged on the nipāta system, with items arranged numerically from one to four. It is not surprising that there are some parallels with both the Anguttara-nikāya and the Puggalapaññatti, which is also arranged on the same system. This is particularly so of the short fourth nipāta, which has only 13 suttas of which seven are identical with suttas in the catuttha-nipāta of the Anguttara-nikāya, ⁴¹⁴ and one is similar to a section of the Puggalapañnatti. ⁴¹⁵ Some of the sultas do not really fit into the nipāta pattern, e.g. the verse in sutta (78=3.3 .9) tells of the danger of the energetic consorting with the slothful, and has no reference to the number three. The prose is somewhat inappropriate, since it tells of like meeting with like, but its position in the tika-nipāta is justified by the inclusion of the triad past, present, and future. The first sutta of the catukka-nipāta (100=4.1) lists four groups of two (dānāni, samvibhāgā, anuggahā, and yāgā). The verse includes a mention of the dhamma-yāga.

The fact that a large proportion of the fourth nipāta seems to have been borrowed from elsewhere in the canon has led to the suggestion ⁴¹⁶ that the nipāta as a whole is a later addition to the text. This suggestion is supported by the Chinese version ⁴¹⁷ of the Itivuttaka, which was presumably translated from a Sanskrit or Prakrit original. ⁴¹⁸ In this, the fourth nipāta is completely lacking, while a number of suttas found in the third nipāta of the Pāli version do not occur. The origin of the Chinese version would seem to belong to the Sarvāstivādin school. ⁴¹⁹

3.5.5. Suttanipāta

The Suttanipāta ⁴²⁰ consists of 1,149 verses, with some prose passages, arranged in five vaggas. There is evidence for believing that the text is a compilation of material from different sources, and some of its contents can claim to be among the oldest Buddhist poetry known to us. ⁴²¹ The last two vaggas, the Aṭṭhakavagga and the Pārāyanavagga, are mentioned by name in other Pāli canonical texts ⁴²² and in Sanskrit texts, and the quotations which are given from them coincide with the text as we have it now. This does not, of course, prove that the whole text existed in its present form at the time of the Buddha. The great age of the two vaggas is also supported by the fact that a commentary upon them, the Niddesa, was composed at a sufficiently early date for it to be included in the canon. A number of individual suttas from other vaggas are found in Sanskrit texts of other traditions, suggesting that the suttas are old enough to pre-date the schisms between the sects.

The fact that the Niddesa deals with two vaggas and one independent sutta, and the separate existence of individual suttas in other traditions, suggest that at the time of the compilation of the Niddesa, and even later, not only were the vaggas still regarded as independent works, but so too were the suttas. It is, however, unwise to press this too far. At the time of the composition of the Chinese translations, which must have been long after the formation of the Theravādin canon, the vaggas were still separate in the tradition from which the Chinese made their translation, since it seems that only the Aṭṭhakavagga has a Chinese counterpart. ⁴²³

The Uragavagga begins with the Uragasutta, a short poem in the Aupacchandasaka metre, in which a bhikkhu ridding himself of worldly emotions is likened to a snake casting its slough, a simile which serves as a recurring refrain. ⁴²⁴ That this was originally a separate text is shown by the fact that in other traditions it is found as a part of the equivalent of the Dhammapada. ⁴²⁵ Linguistic differences between it and the other suttas of the vagga can also be seen. ⁴²⁶ The second sutta is a conversation between the herdsman Dhaniya and the Buddha, in which the two speakers, in alternate Vaitālīya/Aupacchandasaka verses, ⁴²⁷ contrast the domestic and ascetic life. Of considerable interest are hypermetric insertions into each verse (iti Dhaniyo gopo, iti Bhagavā) which make the identity of each speaker clear. These are probably to be regarded as reciter’s remarks, which have been incorporated into the text. Their age is shown by the fact that comparable instructions in the Aṭṭhakavagga and Pārāyanavagga are commented upon in the Niddesa, and must therefore pre-date the composition of that text. The later commentary upon the Suttanipāta, the Paramatthajotikā II, states that these insertions are due to the sangitikāras. ⁴²⁸

The third sutta of the vagga is the Khaggavisānasutta, a set of 41 verses extolling the life of a wanderer, of which all but one have the refrain “He should wander alone, like a rhinoceros horn.” ⁴²⁹ The sutta itself gives no indication of the speaker(s) of these verses, but the Niddesa, which also comments upon this sutta as well as the last two vaggas, states that they were uttered by paccekabuddhas, ⁴³⁰ and when the verses recur in the later canonical Apadāna they are attributed to the same class of speaker. ⁴³¹ Although the idea of the paccekabuddha is probably a late introduction into Buddhism, ⁴³² the early date of the composition of these verses is shown by the inclusion of a commentary upon them in the Niddesa, and by the fact that a version of them appears in the Mahāvastu, where they are said to have been uttered by pratyekabuddhas. ⁴³³

A number of suttas in this and other vaggas contain passages of prose, often as an introduction to the verses, setting the scene in which they were first spoken, but sometimes they are interspersed with the verses, and sometimes serve as a conclusion to the sutta. For the most part these prose passages give the impression of being later additions, ⁴³⁴ although in some cases the addition must have been made at an early date, for the Niddesa comments upon the portion of prose in the Pārāyanavagga. The Paramatthajotikā ascribes it to the sangitikāras. ⁴³⁵ In a few cases, however, some of the verses are scarely intelligible without the help given by the prose passages, and we may assume that from the very earliest days of Buddhism such verses had to be explained by their reciters, and the prose which we have now is doubtless a refinement of the impromptu comments which were made to audiences who needed help to understand them.

A number of the suttas in this vagga are dialogues in which someone questions the Buddha and is converted as a result of his replies. The Parābhavasutta (6) tells of a devāta asking about the things by which a man loses and by which he gains in the world; the Hemavatasutta (9) tells of two yakkhas who talk together about the qualities of the Buddha, and then visit him to put their questions personally; in the Ālavakasutta (10) another yakkha is converted as a result of the Buddha’s answers. In the Mettasutta (8) the virtues of a peaceful mind and goodwill are set forth in the Old Āryā metre, ⁴³⁶ which is a clear sign of antiquity. As already seen, this sutta is part of the Khuddakapāṭha. ⁴³⁷

The first sutta of the Cūlavagga is the Ratanasutta, which explains that salvation is to be obtained only through the Buddha, the Dhamma, and the Sangha. It too is in the Khuddakapāṭha, as is also the Mahāmangalasutta (2.4). ⁴³⁸ A number of suttas are attacks upon the brahman caste. The Brāhmanadhammikasutta (2.7) tells of an ideal society, along Buddhist lines, without wealth and sacrifices, and wins brahmans over to Buddhism by explaining that this ideal state once existed in the past, and was peopled by brahmans who followed the practices which the Buddha advocates. A number of suttas are composed in the Tristubh or mixed Tristubh/Jagatī metre, which is found predominantly in the oldest stratum of Pāli verse. ⁴³⁹ Among these suttas is the Nāvāsutta (2.8), in which the good teacher is likened to a skilful man who is able to help people across the stream (of samsāra) in his boat.

A man should honour one from whom he can learn the doctrine, just as the devatās honour Inda. Being well-disposed and learned, he reveals the doctrine when he is honoured.

If anyone cultivates such a man carefully, making it his aim and listening attentively, entering upon the doctrine and what conforms with the doctrine, he becomes wise, understanding, intelligent, and subtle.

But consorting with a poor fool, who has not learned the truth and is envious, one goes to one’s death, having failed to understand the doctrine clearly in this world, and not having overcome doubt.

How can a man who has gone down into a river, a swift-flowing stream in spate, and is carried along by the stream, help others to cross?

In the same way, how can one who has not learned the doctrine, and has not listened to the explanation given by the learned ones, who is ignorant of it himself and has not overcome doubt, help others to realize it?

Just as one embarking on a stout boat, provided with oar and rudder, could bring many others across there, being skilful, thoughtful, and knowing the means thereof,

In the same way, one who has knowledge and has developed himself, who is learned and unshakable, understanding it himself could make others realize it, if they have the ability to listen attentively.

Therefore one should cultivate a good man who is wise and learned. Understanding the meaning, and following (the path), knowing the doctrine, one attains happiness. (316-23)

The Mahāvagga includes a number of suttas which contain biographical data about the Buddha. The Pabbajjāsutta (3.1) praises the ascetic’s life, and gives a description of the way in which King Bimbisāra of Magadha tried to tempt the Buddha with wealth. The Padhānasutta (3.2) tells of the exertion which was needed to gain enlightenment on the banks of the Nerañjarā, and leads into a dialogue with Māra. Both of these suttas occur in Sanskrit in the Mahāvastu, ⁴⁴⁰ as does the Sabhiyasutta (3.6) in which the wanderer Sabhiya visits the Buddha ⁴⁴¹ and asks him to explain the meaning of a number of terms, including brāhmaṇa, samaṇa, nhātaka and bhikkhu. The Buddha answers his questions, in some cases by making use of “folk” etymologies. ⁴⁴² The Selasutta (3.7) relates the story of the brahman Sela who visited the Buddha to see if he possessed the 32 marks of a great man. In the way in which we have already seen in other texts, ⁴⁴³ the Buddha’s supernormal power permits Sela to see the two marks which are normally hidden. This sutta also occurs in the Majjhimanikāya ⁴⁴⁴ and, without the prose introduction, in the Theragāthā, ⁴⁴⁵ which suggests that in this example, at least, the prose is not an original part of the sutta. The prose introduction of the Kokāliyasutta (3.10), however, makes the meaning of some of the verses clear, and it is probable that some such commentarial passage was needed, and therefore probably existed, from the time of the composition of the sutta. ⁴⁴⁶

The Nālakasutta (3.11) tells of the descent of the sage Asita from the Tusita heaven to see the newly-born Buddha. He takes the child in his arms and prophesies its future greatness. This portion of the sutta is entitled the vatthugāthā⁴⁴⁷ The remainder of the sutta consists of a description of the highest state, that of wisdom, given by the Buddha to Nālaka, Asita’s nephew. As it stands, the sutta is a combination of two separate parts, and this view is supported by the fact that the episodes of Asita meeting the young Gotama and Nālaka questioning the Buddha are quite separate in the Mahāvastu. ⁴⁴⁸ The vatthugāthās in this sutta are composed partly in the Tristubh/Jagatī metre and partly in an extended version of the Tristubh with a redundant long syllable resolved into two short syllables. ⁴⁴⁹ This metre occurs in the Mahāvastu, but not in the Asita episode, which suggests that these biographical details are taken from a common store of history and legend concerning the Buddha, but are not closely related. The Dvayatānupassanāsutta (3.12) deals with a series of pairs (dvayatās), of which dukkha is always the second, and explains how the pairs are causally linked, the second element arising from the first. It has been suggested that the paticcasamuppāda system is only a recast of this “primitive fragment of Abhidhamma.” ⁴⁵⁰

The fourth vagga is the Aṭṭhakavagga. The fact that it is an early text is shown by its mention by name in other Pāli canonical texts, and in Sanskrit texts belonging to other traditions. ⁴⁵¹ The name is Sanskritised as either Artha-vargiya or Aṣtavarga, ⁴⁵² and although arguments can be produced for following either form, the existence of four suttas in the vagga with the term aṭthakasutta in their titles (3.2-5), each containing eight verses, suggests that the vagga was called after the suttas, as frequently happens, ⁴⁵³ and is correctly Sanskritised as Aṣtakavarga. In view of the fact that many of the suttas in other vaggas are found elsewhere in the canon, it is perhaps remarkable that none of the suttas in the Aṭthakavagga is found elsewhere in Pāli. The reason is, perhaps, that the vagga was regarded as a closed body of suttas very early in the history of Buddhism, as is shown by the fact that it is already in the Udāna ⁴⁵⁴ described as having sixteen suttas, which led to its obtaining a status of being original and indivisible. It seems that it is the only vagga to have been translated into Chinese, ⁴⁵⁵ which presumably means that it was the only vagga found in the canon from which the Chinese translation was made. The only fragments which have been published from Turkestan corresponding to any portion of the Suttanipāta also belong to this vagga. ⁴⁵⁶ They presumably belong to the Sarvāstivādin canon.

There are other indications that the vagga is old. As already noted, ⁴⁵⁷ a commentary upon it, entitled the Niddesa, is regarded as canonical. The Tuvaṭakasutta (4.14) is in the Old Ārya metre, ⁴⁵⁸ which is found only in a few very old texts. From the number of variant readings which are mentioned and commented upon in the Niddesa ⁴⁵⁹ it would seem that the vagga had already had a long history of tradition and development by the time of the Niddesa’s composition. The suttas have no prose introductions, although they are found in the Sanskrit and Chinese versions. ⁴⁶⁰ Those found there do not, however, agree with the uppattis which are given in the Paramatthajotikā, which indicates that such introductory stories are of later, independent, composition. Neither the Niddesa nor the Paramatthajotikā includes any comment upon verse 836, which suggests that it is a later interpolation.

Several suttas in the vagga warn the bhikkhu against sensual pleasures. The uselessness of philosophical systems is pointed out, and it is emphasised that purity is not gained by disputation. The ideal muni is described, ⁴⁶¹ and the duties of a bhikkhu are enumerated. The Tissametteyyasutta (4.7) warns of the dangers of sexual intercourse, and the Māgandiyasutta (4.9) tells how the Buddha rejected Māgandiya’s daughter when she was offered as his wife. The philosophical speculation which is attacked seems to be of a very general nature, and there is an absence of deep philosophical concepts, which tends to confirm the early date of the vagga.

The Pārāyanavagga begins with 56 vatthugāthās ⁴⁶² which tell of the brahman Bāvarī sending his sixteen followers to the Buddha. There follow the sixteen questions they put to the Buddha, and the final section of the vagga tells how the sixteen resolved to stay with the Buddha. The vatthugāthās are not commented upon by the Niddesa, which means that either they did not exist when the Niddesa was composed or, if they did exist, they were not yet regarded as canonical. The Paramatthajotikā ascribes them to Ānanda. ⁴⁶³ The concluding section, including a prose passage, is commented upon by the Niddesa. The Paramatthajotikā attributes it to the sangītikāras. ⁴⁶⁴

The sixteen questions (pucchās) are of a more metaphysical nature than the suttas of the Aṭthakavagga, and are concerned with the crossing of the stream, and the escape from birth and death, which accounts for the title of the vagga. ⁴⁶⁵ It is not surprising, considering the nature of the vagga, that individual portions of it do not occur elsewhere in the canon. Its antiquity is shown by the fact that five verses are quoted from the Pārāyanavagga in the Pāli canon, each time being specifically attributed to a particular question (pañha) in the vagga. ⁴⁶⁶ Despite its age, there are indications that it has been made up from earlier sections. If it had been composed as a single entity, one might have expected the metre to be standard throughout, but this is not so. Some sections are in the Tristubh/Jagatī metre, and some are written entirely in the Anustubh metre. Some are in a mixture of the two, and in the case of the Punnakamāṇavapucchā (5.4) interpolations have crept in to such an extent that it is not easy to decide what the original readings might have been. ⁴⁶⁷ The “reciter’s remarks” which have been inserted in each pucchā to make the identity of the speakers clear are commented upon by the Niddesa, which means that they must have been added at an early date.

Even in the vaggas and suttas which are not commented upon in the Niddesa there are metrical and linguistic features which are generally reckoned to be old. ⁴⁶⁸ Although it is not easy to date such phenomena, there seems little doubt that the Suttanipāta contains some of the oldest Pāli verse which we possess.

3.5.6. Vimānavatthu

The Vimānavatthu ⁴⁶⁹ consists of 83 stories in seven vaggas, describing the grandeur of the heavenly abodes (vimānas) enjoyed by those who have been reborn as devas as a reward for meritorious actions done during their births as human beings. The stories all follow the same pattern, in which a deva is asked by Moggallāna the reason for bis enjoying the vimāna in which he dwells, and the deva thereupon relates his previous good conduct. The stories are doubtless intended for laymen, to induce them to live a good life by promising rebirth as a deva as a reward. The text is probably a late addition to the canon, and represents a period when the doctrine of kamma had come to be represented as a simple deed-and-reward mechanism.

The metre ⁴⁷⁰ and style of these stories, in general, merit little attention, and most scholars agree that as we have it the Vimānavatthu must be one of the latest texts to be admitted to the canon. There is, however, no reason to doubt that the style of story is old. The existence of the Kanthakasya vyākaraṇa in the Mahāvastu ⁴⁷¹ which corresponds closely with the Kanṭhakavimānavatthu (7.7) indicates a date of origin prior to the split between the sects. There was also in the Sarvāstivādin Kṣudrakāgama a text known as the Vimānāvadāna, which may, on the strength of the fragment of that text which has been published, ⁴⁷² be assumed to have been a collection of the same character as the Vimānavatthu, and also based upon the same concept of punishment and reward, which must therefore have been older than the two works.

Despite the fact that the stories in their present form are not likely to be very old, there are nevertheless signs that some of the stories may be older than the rest, for three of them contain verses in the Āryā metre. ⁴⁷³ The fact that this metre went out of use in Pāli literature at an early date means that anything composed in it must be presumed to be comparatively early. There are, moreover, indications that the Vimānavatthu has been subject to a long history of development, during which time it was enlarged by late additions and the re-arrangement of material from other texts, especially the Jātaka collection, from which a number of verses have been borrowed. ⁴⁷⁴

3.5.7. Petavatthu

The Petavatthu ⁴⁷⁵ consists of 51 stories in four vaggas, which tell the stories of the departed ones (petas) who are suffering because of the bad actions they have committed during their previous existence. As in the case of the Vimānavatthu, the work seems to have been intended for laymen, to warn them against bad behaviour.

The pattern of most stories is similar. A bhikkhu sees someone suffering in some way and asks the reason, and the peta explains the wicked deed which has led to the present predicament. Some of the verses, however, seem to have been included in the collection simply because they include the word peta. The first vatthu (story) is the “Field-simile” (Khettūpamapetavatthu), which states that honouring arahants and petas is like planting seeds in a field. In due course one will receive the due reward (phala). The Saṃsāramocakapetavatthu (2.1) seems to be a mixture of a peta story with a vimāna story. A woman who was suffering because of her evil deeds then received the reward of a generous action by the thera Sāriputta which was transferred to her. She was thereby released from her suffering.

There can be little doubt that in its present form the Petavatthu is a late composition. The Nandakapetavatthu (4.3) mentions Pingalaka as the name of a king of Suraṭṭha. Dhammapāla informs us ⁴⁷⁶ that Pingala lived during the reign of Dhammāsoka 200 years after the Buddha’s nibbāna. The date of composition would, therefore, be comparable to that of some of the gāthās in the Theragāthā and Therīgāthā collections, whose authors were also said to have lived at that time.

As in the case of the Vimānavatthu, there are parallels with the Jātaka collection, ⁴⁷⁷ and there must have been a period of development, during which time enlargement of some of the stories took place. The existence of a text named Pretāvadāna in the Kṣudrakāgama of the Sarvāstivādin canon suggests that that sect had a text comparable to the Petavatthu, and fragments of it have been discovered. ⁴⁷⁸

As already noted, the Tirokuḍḍapetavatthu (I.5) has been borrowed by the Khuddakapāṭha.

3.5.8. Theragāthā

The Theragāthā ⁴⁷⁹ is a collection of 1279 verses, arranged in poems, ascribed to 264 theras who, for the most part, lived during the lifetime of the Buddha or shortly afterwards. The poems are arranged in nipātas according to the number of verses which they contain, e.g. all the poems of one verse are in the ekanipāta and so on. For the vīsati-nipāta and thereafter the nipāta number gives only a rough guide to the number of verses which the poems contain.

Within the nipātas there is no obvious arrangement, but certain patterns are apparent, ⁴⁸⁰ as verses are linked together by subject, or refrain, or by “catchword,” e.g. Th 55-60 have the common theme of kuti(kā). Sometimes verses are placed together because of some relationship between the authors, either of name or family, e.g. Th 11 is by Cūla-gavaccha and Th 12 by Mahā-gavaccha; Th 107 by Dhammasava and Th 108 by his father. Doubtless this is a reflection of the way in which the collection was made. Verses were recited by the sangitikāras as they remembered them, and the recitation of one verse prompted the memory of another.

Not all verses were in fact recited by the thera to whom they are attributed. Sometimes they were addressed to him, or are about him, but they became “his” by virtue of the fact that they were connected with his name in the tradition. Scmetimes verses ascribed to one and the same thera occur in two places; ⁴⁸¹ sometimes verses under his own name are in one place, and those uttered under a nickname are elsewhere. ⁴⁸² Sometimes the verses are collections in themselves, e.g. Ānanda’s verses are said by the commentator ⁴⁸³ to have been uttered on several occasions, and include verses spoken to him as well as verses recited by him. Some verses in the poems are clearly additions, in that they explain the circumstances in which the verses were uttered, or link various parts of the poems together. This fact was recognised by the commentator, who explains that these verses were added by the sangitikāras. ⁴⁸⁴ Certain defects in the collection are also probably to be attributed to the sangitikāras. It would seem that Sumanigala’s verse (43) should not have been in the eka-nipāta, since it is really two ganacchandas verses, ⁴⁸⁵ but the compilers seem not to have recognised this. Vadḍha’s verses ( 335-39) seem at one time to have been a whole with his mother’s verses in the Therīgāthā, ⁴⁸⁶ and the compilers would appear to have divided what was originally a single poem into two parts in an entirely arbitrary way.

There are also, as in the Suttanipāta, certain hyper-metrical phrases inserted into some narrative-style poems, e.g. iti Selo brāhmano, ⁴⁸⁷ which are perhaps also to be ascribed to the sangitikāras, or perhaps to early reciters, as remarks intended to make the identity of the characters in the narrative clear to the audience. It is possible, and Dhammapāla asserts in his commentary, ⁴⁸⁸ that the sangitikāras concerned were those at the first council, immediately after the death of the Buddha. The same commentator, however, states that Sambhūta uttered his verses at the time of the second council, ⁴⁸⁹ while Tekiccakāni’s verses were recited at the third council. ⁴⁹⁰ If Vitasoka and Ekavihāriya were Asoka’s brothers, as Dhammapāla states, ⁴⁹¹ then their verses must also be dated to the time of the third council.

Some of the verses found in the Theragāthā occur elsewhere, attributed to the same thera, e.g. Vangīsa, ⁴⁹² but in a slightly different form. This is perhaps due to the fact that the same thera might possibly utter his verses, in slightly different ways, on a number of occasions. Different schools of bhānakas might remember different versions of the verses, and the independence of the schools prevented one version being “corrected” in the light of another version. A number of the verses occur elsewhere in Pāli literature as the verses of other theras. This shows nothing more than the absence in ancient times of a law of copyright, and anyone could repeat any verse, which then became “his” verse if it was remembered in connection with his name. Even in the Theragāthā there are identical verses ascribed to two different theras, while the number of pādas, or even lines, which are held in common merely emphasises that anyone who wanted to become a poet could do so simply by stringing together verses from the storehouse of floating pādas. There are also verses found elsewhere, particularly in the Milindapañha, which are not included in the Theragāthā collection. ⁴⁹³ This shows that the original collectors were quite capable of failing to remember all the verses which existed, and also that, once the collection had been closed, no new additions were made to it. There is, however, a certain amount of confusion about the number of verses which the collection should contain, for the uddāna at the end of the text states that it contains 1,360 verses, ⁴⁹⁴ while the total obtained by adding together the numbers in the uddānas to each nirāta is 1,294 . The total in the text as we have it is, as already stated, 1,279. Whether these uddāna figures are based upon a different recension from that which we possess, or whether they are simply incorrect, is not yet clear. In favour of the latter suggestion is the fact that the uddāna to the catuk-ka-nipāta states that there are 13 theras’ verses in the nipāta, while in fact there are only 12 names listed in the uddāna, ⁴⁹⁵ unless one counts the word bhavati as a name, which agree completely with the 12 names given at the end of the 12 sets of verses in the nipāta.

There is a wide range of metres included in the collection, the majority being Anuṣtubh and Triṣtubh/Jagatī, but there is also a number of mātrācchandas and ganacchandas metres, ⁴⁹⁶ some of the latter being the Old Āryā type which went out of use early on in the history of Buddhism. Not only does the metre support the commentarial tradition in telling us that the composition of the Theragāthā covered a long period of time, perhaps three centuries, but there are also references to beliefs and practices of the type which in other texts has seemed to point to lateness. There are references to miracles, to great congregations of gods, and to great rewards coming from worshipping the Buddha with a single lotus flower, ⁴⁹⁷ which are all reminiscent of the later Mahāyāna.

Many of the verses in the Theragāthā are descriptions of the thera’s attempts to attain, and his attainment of, the peace of nibbāna. Some are merely sermons of a general nature, while in others the description of the attainment of peace is coupled with a statement of personal conditions and surroundings, frequently giving a poetic account of the scenery in the midst of which the thera is meditating. Quite often the poem consists of a secular portion, in which such descriptions of nature are given, followed by a religious section in which we learn of the thera’s attainment of his goal. The first part of the verse is similar to the style found in the later Māhārāstrīi Prakrit collection attributed to Hāla, ⁴⁹⁸ although Hāla’s material undoubtedly goes back to an earlier period. It seems likely that the genre was known to the theras who composed the verses found in the Theragāthā, but whereas the examples found in the Sattasaī have a description of some natural phenomenon, e.g. the beginning of the rains, which is then linked to a secular occurrence, e.g. the absence of the traveller, the Theragāthā verses begin with the natural phenomenon which is then linked with a religious experience. ⁴⁹⁹

Spread with garlands of kareri flowers, the regions of the earth are delightful. Resounding with elephants those lovely rocks delight me.

Those rocks delight me, the colour of blue clouds, beautiful, cool with water, having pure streams, covered with Indagopaka insects.

Like the ridge of a blue cloud, like an excellent gabled house, resounding with elephants those lovely rocks delight me.

The lovely surfaces are rained upon; the mountains are resorted to by seers. Made to resound by peacocks, those rocks delight me.

It is enough for me, desiring to meditate, resolute, mindful. It is enough for me, a resolute bhikkhu desirous of the goal. (1062-66)

Although it is probably true to say that the authors of these verses were not trying to be poets, but were merely aiming to give an account of their religious experiences in the way in which they felt was most appropriate for such a statement, i.e. in verse, nevertheless it is noteworthy that, together with some of the verse of the Dhammapada and the Suttanipāta, the Theragāthā and its companion collection the Therīgāthā form some of the more poetical, as opposed to more Buddhist, parts of the Pāli canon.

Fragments of the Sarvāstivādin Sthaviragāthā have been discovered in Turkestan and published. ⁵⁰⁰

3.5.9. Therīgāthā

The Therīgāthā ⁵⁰¹ consists of 522 verses forming 73 poems, arranged in nipātas. As with the Theragāthā, the final uddāna verse causes problems, since it refers to 494 verses ⁵⁰² uttered by 101 therīs. Since one poem is said to be by a group of 30 therīs, it would be possible to obtain a total of about 100 therīs, but this ignores the fact that another poem is said to have been uttered by 500 therīs.

The nipāta system is the same as in the Theragāthā, and the method of collection by association of ideas seems to have been very similar. ⁵⁰³ There are similar problems concerning the omission of verses and the attribution of some to other therīs in other texts. There is also the difficulty, already mentioned, of the verses found in the Bhikkhunīsamyutta of the Saṃyutta-nikāya (5), of which some do not appear in the Therīgāthā, and others are ascribed to different therīs. We may assume that there was uncertainty in early times about the authors and the circumstances in which they uttered their verses, ⁵⁰⁴ with the result that different bhānaka-schools followed different traditions about this.

Although some of the verses are of such a nature that it is impossible to tell the sex of the authors, many are indubitably by women. We read of mothers who are led to the Sangha by the death of a child or husband, and of courtesans who give up their way of life and become bhikkhunīs. One of the most beautiful poems in the whole collection is the lament, in Rathoddhatā metre, by the former courtesan Ambapālī for her lost beauty. ⁵⁰⁵

My hair was black, like the colour of bees, with curly ends; because of old age it is like bark-fibres of hemp; not otherwise is the utterance of the speaker of truth.

Covered with flowers my head was fragrant like a perfumed box; now because of old age it smells like a dog’s fur; not otherwise …

Formerly my breasts looked beautiful, swelling, round, close together, lofty; now they hang down like empty water-bags; not otherwise …

Formerly my body looked beautiful, like a well-polished sheet of gold; now it is covered with fine wrinkles; not otherwise …

Formerly my thighs looked beautiful, like an elephant’s trunk; because of old age they are like stalks of bamboo; not otherwise …

Formerly my calves looked beautiful, possessing delicate anklets, decorated with gold; because of old age they are like sticks of sesame; not otherwise …

Formerly my feet looked beautiful, like (shoes) full of cotton-wool; because of old age they are cracked, and wrinkled; not otherwise …

Such was this body; now it is decrepit, the abode of many pains; an old house, with its plaster fallen off; not otherwise is the utterance of the speaker of truth. (252-53,265-70)

The last two poems of the collection are long compositions in the Āryā metre. ⁵⁰⁶ Although they contain features which might be regarded as late, the fact that they are both in this metre, which went out of use quite early on in the history of Pāli literature, shows that they cannot be very late.

In the first of these two poems, ⁵⁰⁷ Isidāsi tells of unhappy marriages as a result of which she leaves home and becomes a wanderer. She gains the three knowledges and is able to relate the details of her last seven births and to explain how her unhappiness is the result of previous misconduct. Such explanations are a feature found more frequently in the Petavatthu and the Apadāna. A similar feature appears in the second of this pair, ⁵⁰⁸ in which Sumedhā renounces the world and goes forth on her wedding day. It is noteworthy that the similes she employs to illustrate the endlessness of samsāra are taken from a canonical text. ⁵⁰⁹ She too describes her former births, including one at the time of the former Buddha Konāgamana, which implies the growth of the cult of former Buddhas by the time when this poem was composed. This does not, however, necessarily imply a late date of composition. As we have already seen, the six Buddhas who lived before Gotama are mentioned in both the Vinaya-piṭaka and the Dīgha-nikāya, ⁵¹⁰ and since Asoka records that he enlarged an earlier stūpa dedicated to that same Buddha, ⁵¹¹ there is nothing in the poem which seems to put it outside the limit of about three centuries after the time of the Buddha which has already been established for the Theragāthā.

As in the case of the Theragāthā, the quality of the poetry found in these poems varies greatly. Some are complete literary compositions, while others are merely compilations of the same types as Ānanda’s verses in the Theragāthā. Uppalavaṇnā’s poem is made up of four separate episodes, as the commentary tells us, ⁵¹² and no attempt has been made to make an organic whole of verses uttered at different times.

3.5.10. The Jātakas

The Jātaka ⁵¹³ stories are a set of Indian fables which, as adapted by the Buddhists, purport to tell the previous births of Gotama Buddha before the final birth in which he gained enlightenment, i.e. while he was still a Bodhisatta (a being destined for enlightenment). In their Pāli version, the Jātaka stories consist of verses set in a prose setting. With a very few exceptions, only the verses are regarded as canonical. The stories are arranged in nipātas, according to the number of verses they contain, so that in the eka-nipāta the stories contain only one verse. In higher-numbered nipātas the number of verses does not always agree with the nipāta number, while the last nipāta contains ten very long verse stories, and is known merely as the mahā-nipāta.

The general form in which the Jātaka stories have come down to us is that of an incident in the present (paccuppannavatthu) which gives rise to the Buddha telling of an event from one of his previous births (atitavatthu); the gāthās are usually embedded in the atitavatthu, but are sometimes found in the paccuppannavatthu; then follows the word commentary to the verses (veyyākarana), and finally the samodhāna, where the characters in the atitavatthu are identified with those in the circle of the Buddha’s acquaintances (Buddhaparisā). The samodhāna therefore properly belongs to the paccuppannavatthu. ⁵¹⁴ The story of the present is often all but omitted, with the words “This story was told about …”, or with a reference to another story to be told later.

It has been shown that the word commentary, the prose of the framework, and the prose of the stories are all the work of the same author, ⁵¹⁵ and the question of his identity will be discussed later. Jātaka, however, is one of the classes in the nine-fold classification ⁵¹⁶ of the early Buddhist canon, and the presence elsewhere in the canon of a number of stories of the Jātaka type, some of them almost identical with stories found in the Jātaka collection, ⁵¹⁷ shows that the stories in which the verses are placed are sometimes of great antiquity. We may therefore assume that, for the most part, the commentator upon the Jātaka collection was making use of older material, some of it already associated with particular verses, when he composed his stories. This view is supported by the fact that Jātaka stories, some of them almost word for word identical with the Pāli version, are also found in Buddhist Sanskrit texts. ⁵¹⁸ The background story into which the Jātaka verses are inserted sometimes agrees with, and sometimes differs from, the corresponding Pāli story, showing that the association of verse and prose sometimes, but not always, pre-dates the schism between the sects. Although the prose stories in their present form are so much later than the verses, it is nevertheless convenient to deal with both verse and prose together.

To the Jātaka collection proper is prefixed the Nidānakathā, which gives what is in effect a biography of the Buddha. It is in three sections: 1. The story of the beginning in the remote past (dürenidāna), ⁵¹⁹ which tells the history of the Buddha-to-be from his existence as Sumedha during the time of the Buddha Dīpankara down to his re-birth in the heaven of the Tusita gods; 2. The story of the beginning in the not very remote past (avidūrenidāna), ⁵²⁰ beginning with the descent from the Tusita heaven, and ending with the attainment of enlightenment (bodhi); 3. The story of the beginning in the present (santikenidāna), ⁵²¹ relating the story from the enlightenment down to the donation of the Jetavanārāma by the merchant Anāthapindika.

The first section is directly connected with the Buddhavampa and Cariyāpitaka, from which many verses are quoted, and it is in effect a commentary upon extracts from these two texts. The story of Sumedha is taken directly from the Buddhavampa, ⁵²² and the striving for perfection in each birth is described in the Cariyāpitaka, which is quoted in what seems to be a different recension from the text we possess. ⁵²³ The miracles which occur at the birth of the Buddha are those which are described as occurring at the birth of the Buddha Vipassin in the Mahāpadānasutta of the Dīgha-nikāya (14) and of all Buddhas in the Acchariyabbhutadhammasutta of the Majjhima-nikāya (123). The agreement between the Nidānakathā and Buddhist Sanskrit sources ⁵²⁴ about these and other details suggests that both the Pāli and the Sanskrit traditions are based upon a common body of biographical material, both legendary and historical, which was in existence in North India before the time of the schisms.

An analysis of the more than 500⁵²⁵ Jātaka stories of the Pāli collection as we have it shows that it contains a wide variety of different types of verse composition. Many of them seem to be un-Buddhist in origin, and whatever Buddhist qualities they possess are found in the (later) prose story, and arise from the fact that the hero is the Bodhisatta. A number of verses, particularly those in the early nipātas, are intended to teach worldly wisdom (niti) and refer to animal fables and fairy stories of the type found in the Sanskrit Pañcatantra and Hitopadeśa. Some of these stories are also found in Aesop’s fables and other European literature and the problem of their inter-dependence has been much discussed. ⁵²⁶ Representations of some of them are found in the reliefs on the Bharhut stūpa, ⁵²⁷ showing that they had already become popular by the second century B.C. :

Other sets of Jātaka verses are ballads where the verses provide a dialogue and the prose identifies the speakers. Some are narrative stories, where the action may either be supplied in the verses or supplemented by prose insertions. A number of the longer Jātakas in the Mahānipāta are divided up into shorter sections (khandas), which suggests that they were at one time independent texts. If this was not so, then it is difficult to understand why the khandas of the Bhūridattajātaka (543) or the Vidhurapaṇ̣itajātaka (545) are not regarded as separate Jātakas.

Certain observations may be made about the relationship between the verses and the prose stories. The fact that the verse in the Bilārajātaka (128) includes a mention of a cat (bilāra), whereas the prose story is about a jackal (sigāla), indicates that the prose is a later, and somewhat careless, addition. In the Vessantarajātaka (547) the use of the word paccaya in an unusual sense of “proper” in connection with an elephant has led to the word being misinterpreted as the elephant’s name by the prose writer. ⁵²⁸ On the other hand, some of the verses are very difficult to underst and without a prose introduction.

One Jātaka presents a problem which from the commentator’s remark about it can be assumed to be very old. The Mahāsupinajātaka (77) has only one verse, as is to be expected as it is in the eka-nipāta. It seems to consist of seven pādas. The pratīka at the beginning of the Jātaka, ⁵²⁹ however, begins with the fourth pāda. Since the verse narrates the sixteen dreams dreamed by the King of Kosala, it is strange that the pratika seems to begin with the eleventh. The seven pādas are given in full in the course of the prose of the paccuppannavatthu. They are then repeated, in an abbreviated form, in the atitavatthu, but it is clear that, in a very unusual way, they are canonical in the story of the present because the commentary (upon the one word hi) ⁵³⁰ is there. With them, when repeated, is an eighth pāda,⁵³¹ which is commented upon in the atitavatthu. The commentator adds an explanatory note after the samodhāna, stating that the sangütikārakas after the death of the Buddha put the first three pādas into the commentary, ⁵³² presumably because they did not accept them as canonical, and made a gāthā of the remaining five pādas. ⁵³³

It has been suggested that the solution to this problem lies in the fact that the first three pādas are in fact prose, and could therefore not be included in a pratīka in a verse text. Although there is actually canonical prose in the Jātakas, it is possible that what seems to be a prose pratīka in the Kuṇālajātaka was thought to be verse. ⁵³⁴

Although many Jātakas can be regarded as being non-Buddhist, or even pre-Buddhist, there are some which are certainly Buddhist. The verses of the Kalingabodhijātaka (479) tell of King Kalinga visiting the bodhi tree and honouring with 60,000 cartloads of flowers the place where the Buddha had gained enlightenment. The whole poem is purely and devoutly Buddhist. The fact that it is composed in the Āryā metre, ⁵³⁵ which went out of general use in Pāli literature at an early date, indicates that it is not particularly young. This shows clearly that the Buddhists added material of their own to the Jātaka collection.

The Kuṇālajātaka (536) ⁵³⁶ is unique in the Jātaka collection in that it alone contains prose which is regarded as canonical, as is shown by the fact that there is a word commentary upon it. ⁵³⁷ This canonical prose is of two types: there is both narrative prose and also a few descriptive passages characterised by strings of compounds, some of them extremely long. ⁵³⁸ These latter passages are unique in that they are the only places in Pāli literature where we find the type of rhythmical prose called vedha. ⁵³⁹ This in itself would reveal the antiquity of the prose, even if we did not know it to be canonical, because this type of prose is elsewhere found only in the very oldest Jain and Buddhist Sanskrit texts. ⁵⁴⁰ It was doubtless the presence of this prose, which was at the same time verse, which confused the Jātaka-bhānakas who were responsible for the Jātaka tradition, so that they accepted the words evam akkhāyati, which are the first words of the canonical prose, as a pratīka, although as has just been suggested they rejected the prose “pādas” of the Mahāsupinajātaka.

The Kuṇālajātaka is a collection of stories and verses which deal with the wickedness of women. Many of the verses found in it are not Buddhist, but are common to the whole range of misogynous literature which is widespread in India. Such sentiments are found elsewhere in the Jātaka collection. Some of the verses and pādas of the Kuṇālajātaka are also found in the seventh vagga of the ekanipāta, which is entitled Itthivagga ⁵⁴¹ and contains a set of Jātaka stories which show how women are so iniquitous that they will stop at nothing to deceive their husbands.

Several Jātaka stories can be shown to be old because they have counterparts in Jain canonical texts, and sometimes the similarity between them is so close that one version can be used to restore the text of the other. The Mātañgajātaka (497) and the Cittasambhūtajātaka (498) both belong to the class of what has been called “ascetic poetry” ⁵⁴², in which the piety and self-control of the wandering ascetic is contrasted with the formal ceremonialism of the brahmans. These two Jātakas have direct parallels in Jain literature. ⁵⁴³ The Kumbhakārajātaka (408) also has parallels in Jain texts. The verses contain a memorial gāthā which gives the names and home cities of the four paccekabuddhas and four other gāthās ⁵⁴⁴ which describe the signs which led them to enlightenment. The verses do not state that they were paccekabuddhas, although the prose story makes this clear. The situation in Jain literature is the same. One verse gives the four kings and their cities; another gives the four signs. ⁵⁴⁵ The verses do not state that they were patteyabuddhas, but the (later) prose stories do. As in the case of the Mātañgajātaka and the Cittasambhūtajātaka, although the resemblances are great, there are sufficient differences to make it clear that, while it is obvious that the Buddhist and Jain texts must go back ultimately to a common ancestor, they had been so worked over by the redactors of the two traditions that it is not possible to restore in full whatever original lay beneath all the changes they had introduced.

In some cases it is possible to distinguish such insertions from the older material by means of metrical analysis. In the oldest stratum of Buddhist literature the Tristubh/Jagatī metre was used for narrative verse, while at a later date the Anustubh (Śloka) metre replaced it for this purpose. A Jātaka story which in its present form consists of a mixture of Tristubh and Anustubh verses can sometimes be restored to something nearer its original form by removing the verses in the latter metre. The Sivijātaka (499) tells the story of King Sivi who gave his eyes away as an act of extreme generosity. It consists of a mixture of Tristubh/Jagatī and Anustubh verses. It has been shown that the removal of the latter leaves a poem which is complete in itself. ⁵⁴⁶

The Bhūridattajātaka (543) is an anti-brahmanical Jātaka. In a long and complicated story it tells how the Bodhisatta was reborn as a nāga prince. He was captured by a treacherous brahman, who kept him in a basket and made him dance before audiences. The climax of the story is a long denunciation of the brahman caste and a rejection of their claim to be superior because of their knowledge of the Vedas and sacrificial ritual. It is written in a mixture of Tristubh and Anustubh verses, and by removing the latter it is possible to restore an earlier version of the Nāga-story in Tristubh metre. ⁵⁴⁷

In the Dasabrāhmaṇajātaka (495) a minister named Vidhūra tells King Yudhiṭthila of the ten classes of those who are brahman by name alone, having by their evil conduct forfeited their right to be so called. He then describes the only true brahmans and when instructed to show these genuine brahmans invites 500 paccekabuddhas to a meal. The Vidhurapaṇditajātaka (545) ) tells of the same minister being won in a game of dice by a yakkha and taken to the nāga world. There he delighted the nāga king by the answers he gave to questions put to him, and was released. An analysis of the metre shows that over half of the Jātaka is in the Anustubh metre. The removal of the verses in this metre leaves a narrative poem (ākhyāna) consisting of about 150 verses in the Tristubh and Vaitā̄̄̄ya metres. ⁵⁴⁸

The Vessantarajātaka (547)⁵⁴⁹ is the last, longest, and most important of the whole Pāli Jātaka collection. It tells of the Buddha’s earlier birth as Prince Vessantara, a man so generous that he gave away everything he possessed, including his two children and, finally, his wife. Because of the emphasis upon generosity, one of the greatest Buddhist virtues, this Jātaka story seems obviously Buddhist in character, and it has been regarded as a purely Buddhist legend, where the hero is already the Bodhisatta of the later dogma. ⁵⁵⁰ It has, however, been shown that when a small number of verses borrowed from the later Cariyāpiṭaka are removed from the text, the remaining verses are entirely un-Buddhist-in the sense that generosity might be expected to be a virtue in many Indian religions-and its hero is made a Bodhisatta only in the commentarial prose story. ⁵⁵¹

The text consists of over 700 Sloka (Anustubh) verses. There are clear signs that the author of the later prose story did not understand some of the details in the verses, and has consequently invented features to fit in with what he thought the verses meant. The way in which Vessantara’s elephant has acquired the name Paccaya has already been explained. ⁵⁵² Other misunderstandings have arisen from the fact that verses have been somehow transposed during the transmission of the story, and the author of the prose version has invented details to fit in with, and explain, the confusion which has resulted. ⁵⁵³

Despite the large number of stories found in the Jātaka collection, it is clear that it by no means includes all the Jātaka stories which existed in North India in the early days of Buddhism. There are stories found elsewhere in the Pāli canon which are technically Jātaka stories because they purport to tell of the Buddha in an earlier birth, and conclude with an identification of the characters, and are nevertheless not found in the collection. ⁵⁵⁴ There are also several Jātaka stories in the Mahāvastu which do not have a parallel in the Pāli collection. ⁵⁵⁵ The Dīpavaṃsa tells of the Mahāsānghikas making changes in the canon after their separation from the Theravādins, and states that they rejected a portion of the Jātaka. ⁵⁵⁶ This presumably means that the contents of the Jātaka collections of the two sects differed, doubtless because they both added to their collections after the schism.

3.5.11. Niddesa

The Niddesa consists of two parts. The Mahā-niddesa ⁵⁵⁷ is a commentary upon the Aṭṭhakavagga of the Suttanipāta, while the Culla-niddesa ⁵⁵⁸ is a commentary upon the Pārāyanavagga and the Khaggavisānasutta of the Uragavagga of the same text. The fact that these two commentaries are considered to be canonical shows that they must be very old, for the Theravādin tradition states that the last addition to the canon was the Kathāvatthu, which was recited at the time of the third council during the reign of Asoka. ⁵⁵⁹ The Niddesa must therefore be earlier than this. On the other hand, the Dipavaṃsa states that the Niddesa was one of the texts rejected by the Mahāsānghikas after the second council. ⁵⁶⁰ This is perhaps not to be taken literally, but means that the Niddesa was not included in the Mahāsānghika canon, either because it did not exist at the time of the schism, or because it had not yet attained canonical status.

According to the Saddhammappajotikā, ⁵⁶¹ the Niddesa was composed by the Buddha’s disciple Sāriputta, but it has been pointed out ⁵⁶² that if this was so, then it is surprising that he should give three different explanations of his own words as recorded in the Suttanipāta. ⁵⁶³ Although it is possible that some of the explanations given in the Niddesa go back to the time of the Buddha, and were perhaps compiled by Sāriputta, the whole work in its present form must have been produced at some later time. The attribution was probably made because the Niddesa is of the same type as the exegetical explanations attributed to Mahākaccāna and Sāriputta in various canonical texts. ⁵⁶⁴

As has already been noted, ⁵⁶⁵ the three texts on which the Niddesa comments can justly claim to be among the very oldest of Buddhist texts. The Atṭhakavagga and Pārāyanavagga are quoted from by name in other canonical texts, while the Khaggavisānasutta is not only taken over into another canonical Pāli text, ⁵⁶⁶ but is also found in the Mahāvastu. ⁵⁶⁷ It is therefore perhaps not surprising that commentaries should have been compiled upon these texts sufficiently early for the commentaries themselves to be accorded canonical status. Nevertheless there must have been a not inconsiderable period of time between the compilation of the texts and the commentaries upon them, for variant readings had already arisen which necessitated mention with alternative explanations. ⁵⁶⁸ That other commentarial traditions either existed at the time of the composition of the Niddesa, but were not employed, or were developed later, is shown by the fact that the Paramatthajotikā mentions explanations not included in the Niddesa. ⁵⁶⁹

The type of comment made in the Niddesa is very similar to that found in the Old Commentary in the Vinaya-piṭaka, ⁵⁷⁰ in that it consists of long lists of synonyms or near synonyms. It does, however, also include exegetical passages, which presumably means that it represents a slightly later type of commentary. There is no attempt made at abbreviation, and passages are given in full every time they are repeated. The beginning of stereotyped explanation can be seen, e.g. every time the word kappa, or a verb based upon the root kapp-, occurs, the Niddesa states that there are two types of kappa, i.e. tanhā-kappa and ditthi-kappa. [^2-3-362] This happens even if the words occur in successive verses. The result is not an organic structure of exegesis, but a series of disconnected phrases which serve as explanations of the individual words, not in the particular setting of the Suttanipāta, but in any setting. This state of affairs led the editor of the European edition of the Culla-niddesa to adopt a modified form of presentation instead of editing the text in the usual way. As a result of this arrangement, he hoped to facilitate the tracing of such expository material back to its source. ⁵⁷¹

Sometimes, however, the exposition is improved by giving explanations, e.g. kathamkathā “doubt” is explained as vicikicchĀ,⁵⁷² and examples are then given of the ways in which doubt can arise. ⁵⁷³ References to the Nikāyas are frequent, e.g. when explaining Sn 844 , which is quoted at S III 9,19 as occurring in the Māgandiyapañha of the Aṭṭhakavagga, the author of the Niddesa quotes S III 9,14-12,27 verbatim. ⁵⁷⁴

A comparison of the Niddesa with the portions of the Suttanipāta upon which it comments enables us to deduce certain information about the relative dating of both texts. The presence in the Niddesa of comments upon the reciters’ remarks ⁵⁷⁵ and the prose passages ⁵⁷⁶ enables us to state that these features, although assumed to be late, are nevertheless not as late as the vatthugāthās of the Pārāyanavagga which are not commented upon. These gāthās, however, are commented upon by the Paramatthajotikā, where they are attributed to Ānanda. ⁵⁷⁷ This presumably means that the author of the Paramatthajotikā found commentarial material upon the gāthās in the Mahā-aṭṭhakathā of the Mahāvihāra, or he would have made a comment to that effect as he did for other verses which lacked comment in the Sinhalese tradition. ⁵⁷⁸ The fact that neither the Niddesa nor the Paramatthajotikā comments upon Sn 836 presumably means that this verse is a very late interpolation.

The presence of certain geographical data, e.g. the mention of Java, Suvaṇṇabhūmi, Takkola, etc., ⁵⁷⁹ in the Niddesa has led to the suggestion of dates as late as the second century A.D. for its compilation. ⁵⁸⁰ It is true that the inclusion of the name Allasanda in the lists of places means that that particular portion of the lists is post-Alexander, but leaving aside the possibility of a name or names being inserted into an already-existing list, the beginning of the third century B.C. would seem to be quite suitable as the date of its composition.

Recent archaeological evidence ⁵⁸¹ has shown that there was trade between India and the East before the time of Asoka, so there is no reason to doubt that these names could have been known in India at a sufficiently early date for such geographical lists to have been compiled before the date of the third council.

Those who suggest late dates for the compilation of the Niddesa forget that the Sinhalese tradition recorded the fact, related to us by Buddhaghosa, that the Niddesa existed before the first century B.C., and had by that time been forgotten by all bhikkhus except one. From fear of its disappearing altogether the thera Mahārakkhita was persuaded to learn the Niddesa from this one bhikkhu, and others learned it from Mahārakkhita. ⁵⁸² It was the realisation that canonical texts could easily disappear if the oral tradition died out which precipitated the writing down of the canon during the reign of Vatṭa. gāmaṇī.

3.5.12. Paṭisambhidāmagga

The Paṭisambhidāmagga [^2-3-375] consists of three vaggas, each containing ten topics (kathās). It is intended to teach the path of discrimination (patisambhidā), and tries to show how understanding of the truth taught by the Buddha occurs. It is attributed by the Saddhammappakāsin ⁵⁸³ to the Buddha’s disciple Sāriputta, just as is the Niddesa. The reason for this may be that there seems to be a close relationship between the Paṭisambhidāmagga and the Dasuttarasutta of the Dīgha-nikāya (34), [^2-3-377] which is also attributed to Sāriputta.

The form of the text suggests that it is late, and this view is supported by the fact that the term patisambhida does not seem to occur in the older parts of the Vinaya-piṭaka or the first three Nikāyas. It occurs in some Abhidhamma texts, and in the Niddesa, Apadāna and Buddhavamsa of the Khuddakanikāya. ⁵⁸⁴ That the text is later than much of the canon is shown by the quotation of whole passages from the Vinaya-piṭaka and portions of the Suttapitaka. Nearly every section is given a sutta basis in this way. A general acquaintance with early Buddhist legends is assumed, e.g. in the section on supernormal power (iddhi) ⁵⁸⁵ the names of arahants who possessed such powers are given without any further comment, as if their stories were well known.

There is a list of contents (mātikā) at the beginning of the text, ⁵⁸⁶ but this gives only a list of the 73 knowledges discussed in the first section (Nānakathā), and does not apply to the whole work. Such a mātikā seems to be an Abhi-dhamma-type feature, and the various topics are treated in the form of questions and answers, as in the Abhidhamma, but in form it is a sutta text, since it includes the standard opening evam me sutam in some sections, and the vocative bhikkhave also occurs. ⁵⁸⁷ It is, however, possible that these words occur as parts of quotations from sutta texts, and are not fundamental features of the Paṭisambhidāmagga itself. ⁵⁸⁸

The text is included among the list of those which the Dipavamsa states were rejected by the Mahāsānghikas, ⁵⁸⁹ which probably means that, if it existed in any form at the time of the schism, it had not yet been accepted as canonical. The existence, however, of the technical term pratisamvid in Buddhist Sanskrit ⁵⁹⁰ shows that the concept was not an invention of the Theravādins but was in existence when the other sects separated from them. The Paṭisambhidāmagga is not included among those texts which Buddhaghosa informs us were not accepted by the Dīgha-bhānakas, ⁵⁹¹ so it may be deduced that it was recognised as canonical before the composition of the Apadāna, Buddhavaṃsa, Cariyāpiṭaka and Khuddakapāṭha.

The text presents a systematic exposition of certain important topics of Buddhism, and describes how insight is to be obtained. The way begins with proper knowledge, and the first kathaā, amounting to almost a third of the whole text, ⁵⁹² is consequently devoted to ñāṇa, which includes knowledge of the great pity of a Tathāgata. Other important topics to which kathās are devoted include the regulation of breathing (ānāpāna) as an aid to mindfulness (1.3); kamma (1.7); the noble truths (2.2); the factors (bojjhangas) of enlightenment (2.3); loving-kindness (mettā) towards creatures (2.4); the discriminations (patisambhidās) which give their name to the entire text (2.6); emptiness (suñña), i.e. the fact that all dhammas are devoid of self (2.10); supernormal powers (iddhi) (3.2); the foundation of mindfulness (satipatthāna) (3.8); and insight (vipassanā) (3.9) which together with serenity (samatha) is also treated in the section on the couple (yuganandha) (2.1).

Although the Paṭisambhidāmagga is not an Abhidhamma text, in the sense that it does not appear in the Abhidhamma-piṭaka, it has been suggested that at some early date it was in fact classed among the Abhidhamma texts. ⁵⁹³ It seems to have certain features in common with the Vibhanga, e.g. nearly a half of the section titles of the Vibhanga appear in it, and like the Vibhanga it contains an interwoven exegetical atthakathā. To some extent it seems to imply knowledge of the Vibhanga, e.g. when discussing the four satipatthānas the Paṭisambhidāmagga comments only upon bhāvanā, which is not included in the comment upon satipatthāna in the Vibhanga. ⁵⁹⁴ Sometimes the Paṭisambhidāmagga seems to go beyond what is found in Abhidhamma texts, e.g. in place of the seven powers (balāni) of an arahant ten are given; ⁵⁹⁵ similarly ten iddhis are listed in place of the usual four, ⁵⁹⁶ and ten disāpharanās ⁵⁹⁷ as well as the usual four disās.

3.5.13. Apadāna

The Apadāna ⁵⁹⁸ is a collection of legends in verse, in which the noble deeds (apadānas) done by Buddhist theras and therīs in earlier existences are glorified. It contains in 55 vaggas the biographies of 547 theras and in four vaggas the biographies of 40 therīs, all of whom lived at the time of the Buddha. Each of these stories gives first of all the life of the hero or heroine in one or more previous births. They all contain a story of the past and a story of the present, but they differ from a Jātaka story in that in the Jātaka collection the stories of the past and the present always refer to the Buddha, whereas the Apadāna deals usually, although not always, with an arahant. The stories generally follow the same pattern. The author tells of the adoration made to an earlier Buddha, and the prophecy made by that Buddha that the doctrine of Gotama Buddha would be heard. The story goes on to say how this came true, and how the author became an arahant. The purpose is to show that even the smallest meritorious act has the potentiality to give great results even after the passage of immense periods of time.

The Apadāna is almost an appendix to the Theragāthā and Therīgāthā, since it connects together the past and present lives of the theras and theris. The last two poems in the Therīgāthā are, in fact, of an Apadāna nature. ⁵⁹⁹ Many of the authors of Apadānas are also mentioned in the Theragāthā and Therīgāthā collections, and in his commentaries upon those texts Dhammapāla ⁶⁰⁰ includes the Apadāna section relevant to the hero or heroine. There are, however, certain discrepancies in his ascriptions, ⁶⁰¹ and one of the Apadānas he includes does not occur in the Apadāna collection in its present form. ⁶⁰² The collection, however, includes many theras who do not appear in the Theragāthā, and does not include all the therīs who are in the Therīgāthā. On the other hand, there is an Apadāna for Yasodharā, ⁶⁰³ although there is no poem for her in the Therīgāthā. A large number of the theras are not identifiable since they have names made up from the deed which is related in their Apadāna, e.g. “Mangogiver.”

Many of the stories which are told have a mythological nature, which suggests that the Apadāna is one of the latest books in the canon. ⁶⁰⁴ The type of act which is glorified also implies lateness, for there is mention of the worship of thüpas, shrines and relics, and there is an emphasis upon generosity and humanitarian deeds. Nevertheless, there is already mention of some of these activities in the Theragāthā and Therīgāthā, which can be dated with fair probability to the period between the fifth and the third centuries B.C. One of the Apadānas, however, refers to the Kathāvatthu as an Abhidhamma compilation, ⁶⁰⁵ which proves that that particular Apadāna must be later than the third council. The nature of the Apadāna collection is such that it was simple for more and more Apadānas to be added until the time when it was fixed as canonical, which was perhaps at the time of writing down, after which additions were probably not made.

A few Apadānas depart from the usual pattern. The first section of the text is entitled Buddhāpadāna, ⁶⁰⁶ and the use of the word apadāna in connection with the Buddha is reminiscent of its usage in the Mahāpadāna of the Dīghanikāya (14). ⁶⁰⁷ In this section ⁶⁰⁸ the Buddha himself tells of the Buddhakhettas, ideal lands of beauty where the Buddhas live. A picture is painted of Buddhas questioning each other, and there is mention of disciples questioning the Buddhas and vice versa. The whole idea would seem to be late, and it has been regarded as foreshadowing Mahāyānic beliefs. It has even been called a completely Mahāyānic passage, ⁶⁰⁹ but it does not seem to be necessary to go as far as that. A simpler and doubtless earlier idea of the Buddhakhettas is found elsewhere in the Apadāna, ⁶¹⁰ and also in the Theragāthā and the Milindapañha. ⁶¹¹ A more elaborate form of the idea is found in the Mahāvastu, ⁶¹² also a Hīnayāna text, although not free from some Mahāyānic influence. The form of classification which Buddhaghosa adopts for the Buddhakhettas ⁶¹³ shows that the concept was widespread by his time, although he gives no indication of the nature of these Buddha fields. Whether any Mahāyānic influence should be seen in the idea is doubtful. It has been rightly pointed out that many ideas in Buddhism follow from the dynamics of early Buddhist thought, which lead to the existence of one and the same idea in two forms in two different traditions. Linguistic evidence suggests an unusual origin for this Apadāna, as will be seen, but there is no reason to see here anything more than an idea which was carried further in this text than in any other Theravādin text, but which nevertheless follows on from the concept of a miraculous Buddha which occurs as early as the Dīgha-nikāya.

The second section of the Apadāna is the Paccekabuddhāpadāna, in which the Buddha is asked by Ānanda about the paccekabuddhas. In the course of his reply, the Buddha includes the verses in Tristubh metre which occur in the Khaggavisānasutta of the Suttanipāta. ⁶¹⁴ As has already been noticed, ⁶¹⁵ the Suttanipāta says nothing about the speakers of these verses, although the Niddesa attributes them to the paccekabuddhas, and so does the Mahāvastu.

In the Pubbakammapiloti-apadāna (387), ⁶¹⁶ the Buddha tells of his bad deeds, as a result of which he suffered numerous rebirths in hells. The remains of these deeds caused him many unpleasant experiences even in his last existence. This Apadāna is called Buddhāpadāna in the rubric, ⁶¹⁷ and the latter part of it is given and commented upon by the commentator at the end of the Buddhāpadāna. ⁶¹⁸ It is possible that in an earlier form of the tradition the two apadannas were combined in some way. This would explain why the Buddhāpadāna in its present lacks the essential feature of an apadāna: it says nothing about the Buddha’s previous existences, and seems more like an udāna than an apadāna.

The text seems to be composed entirely in the Sloka metre, except for the Paccekabuddhāpadāna which is in the Tristubh metre. Somewhat surprisingly, the Apadāna shows traces of features which are known in mainland Prakrits, but are not common, if found at all, in Pāli, e.g. masculine plural forms in -ānī and aorists in -e (the so-called optative used as an aorist). ⁶¹⁹ The majority of the latter occur in the Buddhāpadāna. Since there is no reason to assume that these features arose in Ceylon independently, we have to suppose that they are indeed old mainland features. This does not necessarily mean that the verses containing these features were all composed on the mainland of India at an early date, for we cannot rule out conscious archaising in some cases. What is, however, important is the fact that, whether genuine features or archaisms, they indicate knowledge of phenomena not common in Pāli. If they were at one time more common in Pāli, then they have been edited out of our texts, and their existence in the Apadāna presumably indicates that this text or portions of it, did not undergo the usual editing procedure. An alternative explanation demands that the composers of the verses containing the anomalous features had knowledge of them through a source no longer available to us, from which they borrowed them into Pāli.

The word apadāna occurs in Sanskrit ⁶²⁰ in the sense of a great or noble deed, but the more common spelling is avadāna, and this, as was noted ⁶²¹ in the case of the Sanskrit equivalent of the Mahāpādāna of the Dīgha-nikāya (14), is the usual form of the word in Buddhist Sanskrit texts. Just as the Apadāna goes with the Theragāthā and Therīgāthā, similarly in Sanskrit there are texts of a comparable type which go together. The Sanskrit Anavataptagāthā ⁶²² is a collection of Avadānas which goes with the Sthaviragāthā, and fragments found in Turkestan indicate that there are parallels with a number of the Pāli Apadānas. There are also parallels in the Vinaya-piṭaka ⁶²³ of the Mūla-Sarvāstivādins, found in Gilgit, and it seems that the Anavataptagāthā has been transferred to that text in the canon belonging to that sect. The fact that a portion of the Avadāna corresponding to the Pubbakammapiloti-apadāna has been discovered in Turkestan, and its existence in Tibetan and Chinese versions, indicate that the Apadāna was the common property of both Theravādins and Sarvāstivādins, and is therefore likely to be quite early.

3.5.14. Buddhavamsa ⁶²⁴

As has been noted, ⁶²⁵ in the Vinaya-piṭaka and the Dīgha-nikāya there is mention of the six Buddhas who preceded Gotama. The Buddha doctrine developed, and by the time that the canon was closed the number of Buddhas before Gotama had reached twenty-four.

The Buddhavamsa opens with the description of an exquisite Jewel Walk (ratana-cankama) which Gotama, just after he had become a Buddha, constructed in the sky and paced upon. He was visited by Sāriputta with 500 arahants, who asked him about the resolve and aspiration he had made for Buddhahood. In reply, Gotama relates the lives of 24 previous Buddhas, and tells what he, as a Bodhisatta, did under each Buddha. The account he gives of himself as the ascetic Sumedha in the first of these former births is the longest and most detailed.

The 24 chapters devoted to the previous Buddhas chronicle certain features, always substantially the same, although the details vary, ⁶²⁶ in their life on earth both as Bodhisatta and Buddha. As well as being a chronicle of those Buddhas, it is also a history of Gotama through the times of all the Buddhas who preceded him.

The Buddha doctrine shows in fact an even later development, because after the 26th chapter, in which the Buddha gives, very briefly, the details about himself, which probably served as a model for the history of the other Buddhas, he mentions in the 27 th chapter three Buddhas earlier than Dīpankara. ⁶²⁷ These three were perhaps not thought to be so important because tradition did not tell of Gotama making a resolve and aspiration under them. The commentator states that the 18 verses of this chapter were established by the sangitikāras and should be regarded as the envoi (nigamanagāthā). ⁶²⁸ It would seem that the Buddhavamsa, even as added to by the sangitikāras, should have ended there, but in the text as we have it there are two further verses, in which the Buddha refers to himself and to Metteyya, ⁶²⁹ who, as has been noted, ⁶³⁰ is referred to in the Cakkavattisīhanādasutta of the Dīgha-nikāya (26). A further chapter (28) gives an account of the distribution of the relics of the Buddha. which is related to and based upon that at the end of the Mahāparinibbānasutta of the Dīgha-nikāya (16). The commentary makes no mention of this chapter.

The reason for the production of the Buddhavamsa would seem to be the fact that the history of the one Bodhisatta who would become the Buddha Gotama needed an account of previous Buddhas to show that he did not suffer from the abnormality of uniqueness, and that his enlightenment was gained only after striving through many births for the fulfilment of the ten perfections. ⁶³¹ To complete them all is a necessary self-preparation for enlightenment. The Buddhavamsa is therefore a developed Bodhisatta doctrine, but it was not developed further, even in the Abhidhamma. Whether the Buddhavamsa with the idea in it was borrowed from another sect, as has been suggested, ⁶³² or whether it was an idea old enough to be common to both the Theravādins and the Sarvāstivādins, who developed it much further, is impossible to decide.

The whole text is composed in the Sloka metre, which shows no particularly early features, and in view of the state of development of both the Buddha and the Bodhisatta doctrines, it would seem that the Buddhavamsa is a relatively late addition to the canon. As has been noted, it was not accepted as canonical by the Dīgha-bhānakas. ⁶³³ It has been pointed out, however, that there is no reference to a tooth of the Buddha having been taken to Ceylon in the account of the distribution of the relics, although there is alleged to be a Sinhalese version with such a reference. This therefore implies that the text was in its present form including the additions at the end, before the arrival of the tooth in Ceylon. ⁶³⁴

3.5.15. Cariyāpiṭaka

The Cariyāpiṭaka ⁶³⁵ is the fifteenth and last book of the Khuddaka-nikāya. It consists of 35 Jātaka-type stories arranged in three vaggas, intended to illustrate the ten perfections (pāramīs). At the end it is called Buddhāpadāniya, ⁶³⁶ which again shows the unusual use of the word apadāna instead of Jātaka in connection with the Buddha. ⁶³⁷

According to the commentary, the Cariyāpiṭaka was related by the Buddha at the request of Sāriputta. Each story is told in the first person, of a rebirth in earlier time, in which the Buddha followed his set aim of obtaining enlightenment by mastering the ten perfections. These are mentioned in the Buddhavamsa, ⁶³⁸ and the Buddha expressed his wish to bring them to fulfilment. The first vagga has ten cariyās (lit. “proper conduct”), all concerning the dānapāramī. The second also has ten, concerning the sīla-pāramī. The third vagga has fifteen cariyās, five for the nekkhamma-, one for the adhiṭthāna-, six for the sacca-, two for the mettā-, and one for the upekkhā-pāramī. One of the niga-mana-gāthās at the end of the text implies that the other three pāramis have also been attained. ⁶³⁹ There are terms in various cariyās which can be interpreted as implying the three missing pāramis, ⁶⁴⁰ and it is perhaps not necessary to follow those scholars who believe that part of the text must be missing.

The Cariyāpiṭaka is largely based upon the Jātaka collection, from which stories illustrating the perfections have been adapted. Of the 35 cariyās 32 can be directly related to the Jātaka collection. Another can be related to the Dīgha-nikāya, ⁶⁴¹ and one to the Majjhima-nikāya. ⁶⁴² The remaining story consists of a single verse, and is too short for identification. ⁶⁴³ In the Nidānakathā of the Jātaka, the ten pāramitās are mentioned, with examples of each extracted from the Jātakas. ⁶⁴⁴ Reference is made to the Cariyāpiṭaka for a longer account. ⁶⁴⁵ There are only 21 Jātakas in common between the two, which has led to the suggestion that the reference may be to a different recension of the Cariyāpiṭaka from that available now. ⁶⁴⁶

The metre is Sloka throughout, except for two Tristubh verses taken over from the Jātaka. ⁶⁴⁷ Although the stories are, with the exceptions already noted, borrowed from the Jātaka, there is in some cases not a single verse in common with the Jātaka. The longest poem in the Cariyāpiṭaka is devoted to the story of Vessantara, and there are about 20 verses common to it and the Vessantarajātaka. Rather than these being borrowed from the Jātaka, however, it has been shown that the reverse is true. ⁶⁴⁸ The verses have been taken over into the prose story which amplifies the verse of the Vessantarajātaka, and are usually introduced with the words tena vuttam. Compared with the Jātaka verses those in the Cariyāpiṭaka are mediocre, and introduce sentimental and supernatural features lacking in the Jātaka. The form of presentation and the metre, and the fact that the Cariyāpiṭaka serves in effect as a supplement to the Buddhavamsa, all suggest that it is a late text. As already noted, it was not accepted as canonical by the Dīgha-bhānakas. ⁶⁴⁹

4. THE ABHIDHAMMA-PITTTAKA

It is clear that the Abhidhamma is later than the rest of the canon. There is no mention in the chronicles of reciting the Abhidhamma at the first or second councils, although the Mahāvaṃsa states that the arahants who held the second council knew the tipitaka. ⁶⁵⁰ The Mahāsānghikas who split from the Theravādins after the second council are said to have rejected the Abhidhamma. ⁶⁵¹ This presumably means that there was nothing corresponding exactly to the Pāli Abhidhamma in their canon, from which it may be deduced that the Abhidhamma did not exist at that time, or at least was not recognised as canonical.

In a number of places in the canon we find the triad Dhamma-dhara Vinayadhara Mātikā-dhara, ⁶⁵² from which it appears that Mātika and Abhidhamma are synonymous. The commentaries, however, explain that in this context Mātika means Pātimokkha, ⁶⁵³ and this interpretation probably arises from the fact that mātika means “summary,” and can therefore be used both of a summary of the Pātimokkha rules and of the type of summary of contents which is found at the beginning of the Paṭisambhidāmagga, ⁶⁵⁴ and in the majority of the texts of the Abhidhamma-piṭaka. Mātikās occur in two places in the Dhammasangani (one at the beginning, and one later), ⁶⁵⁵ in the Vibhanga (in four places, not at the beginning), [^2-4-7] the Dhātukathā, Puggala-paññatti, Yamaka, ⁶⁵⁶ and Paṭthāna. Although there is no mātika for the Kathāvatthu, nevertheless Buddhaghosa uses the word ⁶⁵⁷ when he maintains that the Buddha laid down the outline to be followed by Tissa when he recited that text at the third council. The Mohavicchedanī is a commentary upon the mātikās of all seven Abhidhamma texts: in the case of the Kathāvatthu the mātika is simply the index of contents. ⁶⁵⁸

It seems likely that by the time of the third council the mātikās were already in existence, with perhaps a certain amount of elaboration to show bhikkhus how each term of the mātikās should be expanded. In the case of the Dhammasangani it can be assumed that the text was already available at greater length, since the existence of what Buddhaghosa calls an atthakathā⁶⁵⁹ in the body of the text as we have it would seem to indicate that the chapter upon which it comments must be very old. The Buddhavamsa states that the Buddha Konāgamana preached seven treatises, ⁶⁶⁰ and the commentary states that these were the seven texts of the Abhidhamma-piṭaka. ⁶⁶¹ Although this is clearly an anachronism, it suggests that the Abhidhamma-piṭaka was in its present form by the time the Buddhavamsa was composed.

The word abhidhamma is found in the Vinaya-piṭaka and the Sutta-piṭaka, and seems in origin to be the preposition abhi together with the word dhamme, with the meaning “as regards the dhamma.” ⁶⁶² We find in the Majjhima-nikāya the phrase siyamsu dve bhikkhū abhidhamme nānāvādā⁶⁶³ and the word also occurs in a context with abhivinaye in the Vinaya-piṭaka: abhidhamme vinetum abhivinaye vinetum. ⁶⁶⁴ These were then taken as being the locative of abhidhamma and abhivinaya and were given a separate existence, e.g. ko tattha vinayo, ko tattha abhivinayo. ⁶⁶⁵

So the compound abhidhammakathā must have meant “discourse about dhamma” at first, and then “discourse about abhidhamma.” ⁶⁶⁶ When it is said of a bhikkhu that abhidhamme kho pana abhivinaye pañhaṃ puttho samsādeti no vissajjeti, ⁶⁶⁷ probably an elaboration and analysis of the doctrinal principles is intended, just as abhivinaye would mean a casuistic discussion of the rules of discipline. There is, however, no separate Abhivinaya, for the elaboration of the rules of discipline exists in the Vinaya-piṭaka itself. ⁶⁶⁸

The codification of such principles which require exposition would give rise to the mātikaā lists which form the basis of the Abhidhamma books, and now serve as tables of contents. The elaboration of these principles, their definition, and their proof by adding sutta passages would form the Abhidhamma proper. This is what we find in the Abhidhamma-piṭaka collection as we have it. These matters must have been discussed before the separation of the Sarvāstivādins from the Theravādins, but not long before. The two sects agree closely in their Vinaya-piṭakas and Sutta-piṭakas, but beyond agreeing about the number of Abhidhamma texts, and the title of one book, there is little agreement in their Abhidhamma-piṭakas. ⁶⁶⁹ The two collections are two sets of texts compiled by two sects who independently examined and elaborated the same subject matter.

Once the word abhidhamma had lost its original meaning, the commentaries interpreted it as “higher, special dhamma.” At the beginning of the Atthasālinī Buddhaghosa asks himself what abhidhamma is, and answers the question as follows: “Herein what is meant by ‘Abhidhamma’? That which exceeds and is distinguished from the Dhamma (the Suttas). The prefix abhi, like ati, is used in the sense of preponderance and distinction, as in such passages as, ‘Severe pains overwhelm (abhikkamanti) me, brother, and do not abate;’ and ‘of eminent (abhikkantā) beauty.’ Again: Hence when many sunshades and flags are uplifted, the sunshade which excels the rest in size and is of distinguished colour and form is called aticchattam ‘the pre-eminent sunshade,’ and the flag which is the largest and is replete with various distinguished paints and colours is called atidhaja ‘the pre-eminent flag’ … Even so this ‘dhamma’ is called Abhidhamma, because it excels and is distinguished by several qualities from the other Dhamma. In the Suttanta, the five aggregates are classified partially and not fully. In the Abhidhamma they are classified fully by the methods of Suttantaclassification, Abhidhamma-classification, and catechism … In the Suttantas the four applications in mindfulness are partially classified, not fully. But in the Abhidhamma they are classified in detail under the three methods … In the Suttantas knowledge is partially classified, not fully … But in the Abhidhamma there is a detailed classification of knowledge after the table of contents [mātikā] has been thus laid down: ‘Under the unitary method the basis of knowledge is …’ and so forth … Thus it is to be understood that the Abhidhamma exceeds and is distinguished from the Dhamma.” ⁶⁷⁰

The order of texts in the Abhidhamma-piṭaka is, according to Buddhaghosa: ⁶⁷¹ Dhammasañgani, Vibhañga, Dhātukathā, Puggala-paññatti, Kathāvatthu, Yamaka, and Paṭthāna, but in the Chinese translation of the Samantapāsādikā ⁶⁷² the Puggala-paññatti and Kathāvatthu come after Yamaka and Paṭthāna. Buddhaghosa states ⁶⁷³ that the textual order of the Abhidhamma originated with Sāriputta, who also determined the numerical order in the Paṭthāna. He is reported to have done this to make it easy to learn, remember, study and teach the dhamma. To guard against accusations that this would mean that the Abhidhamma was not Buddha-vacana, Buddhaghosa goes on to say that Sāriputta was not the first to understand the Abhidhamma, for the Buddha was the first ābhidhammika.

4.1. Dhammasangani

The mātikā of the Dhammasangani ⁶⁷⁴ consists of a list of 22 triplets (tikas), beginning with states that are good, bad, and indeterminate, and 100 couplets (dukas), beginning with states that are moral roots and not moral roots. ⁶⁷⁵ These various states (dhammā ) comprise all mental material conditions, and the mātikaā is an outline of a universal system of classification comprising the whole analytical teaching of the Buddha.

The Dhammasangani is a “Compilation of States”, dealing with the enumeration and definition under particular group headings of these states, in accordance with the Buddha’s different modes of analysis. The term dhamma is used in a broader sense than the word “state” in English, and comprises not only the discrete states of consciousness whose individual classification depends upon appropriately associated mental concomitants (cetasikā ), but also those mental concomitants themselves. Material quality (rūpa), the four great essentials (mahābhūtā), the dependent material qualities (upādāya rūpā), and even nibbāna (asañkhata dhamma) are included within this term. ⁶⁷⁶

The greater part of the Dhammasangani is an expansion of the mātika in accordance with the system of analysis. So, to the question “What are good states ?,” the answer is given that when a good thought concerning the phenomenal world has arisen there is contact (phassa), feeling (vedanā ), and 54 other states. These are then defined in detail, and then subjected to further analysis, with summaries given of the psychological functions involved in each. The whole analysis is then repeated for different types of thought, and then for the jhānas. The answer to the question consequently takes 66 pages. ⁶⁷⁷ The question “What are the bad states?” is answered more briefly, since it is discussed only in connection with twelve bad thoughts. ⁶⁷⁸

The final section of the Dhammasangani ⁶⁷⁹ serves as a commentary upon the text, although in the Dhammasangani itself no reference is made to it being anything other than a genuine portion of the work. ⁶⁸⁰ Buddhaghosa, however, calls it atthakathākanda, ⁶⁸¹ and gives it the name atthuddhāra.⁶⁸² He ascribes it to Sāriputta, and states that he uttered it for a pupil who was unable to remember the elucidation of the Dhammasangani. He gives, however, an alternative story, informing us that the Mahā-aṭ̣thakathā rejects Sāriputta’s authorship and ascribes it to the Buddha himself. ⁶⁸³ It is stated in the Dīpavamsa [^2-4-36] that the Atthuddhāra was one of the texts rejected by the Mahāsānghikas when they split from the Theravādins. This may mean that this portion of commentary did not exist at the time of the schism, or was not yet accepted as canonical. If the Sinhalese commentarial tradition accepted it as Buddhavacana, it is probable that it had already been included in the Dhammasangani by the time of the third council. [^2-4-37]

4.2. Vibhanga

There were a number of topics to which the Buddha devoted particular attention in some of his discourses, to which a special degree of further analysis could be afforded, to show that not only could a general statement be made about the subject, but further enquiry would show that the statement of general truth could also be a statement of particular truth. [^2-4-38] These topics included the five khandhas, the āyatanas, the dhātus, the four truths, the indriyas, etc. The beginnings of this type of analysis can be found in a number of vibhangasuttas in the Majjhima-nikāya. [^2-4-39]

The Vibhanga [^2-4-40] has eighteen chapters, most of which have three sections: Suttanta-bhājanīya, Abhidhamma-bhājanīya, and Pañhāpucchaka (catechism). In the Suttanta-bhājanīya portion the method of analysis and the definitions used by the Buddha in general discourses are shown, and some of these analyses are found elsewhere in the canon, e.g. that of the Saccavibhanga is found in the Saccavibhangasutta of the Majjhima-nikāya (141), [^2-4-41] with the difference that there it is a discourse put in the mouth of Sāriputta, whereas in the Vibhanga it is an impersonal set of questions and answers. In the Abhidhammabhājanīya we find the technical analysis and the definitions employed when the same matter is discussed from a strictly philosophical point of view. In the chapters which have a Pañhāpucchaka section there is an explanation in detail of how each of the special terms used are to be defined within the framework of triplets and couplets which is enumerated in the Dhammasangani.

The Vibhanga does not have a mātikā at the beginning, although the Mohavicchedanī, which is a commentary upon the mātikās, takes the titles of the titles of the sections as forming a mātika. [^2-4-42] There are, however, separate mātikās occurring in four of the vibhangas. [^2-4-43]

⁶⁸⁴

The last three vibhangas merit comment. The 16th and 17th differ from the earlier vibhangas in that they are numerical. The 16th deals withñāṇa, and collects together the very many groupings into which this subject may be classified in connection with the Buddha’s purpose. ⁶⁸⁵ The classification varies in extent from a single to a tenfold system of grouping. The first, second, and third depend very largely upon the triplet and couplet sections of the Dhammasangani, and are therefore similar in construction to the Pañhāpucchaka section of many other vibhangas. The 17th vibhanga deals with small items (Khuddakavatthuvibhanga), and approaches the matter of the bad states in a numerical system of classification similar to that found in the 16th vibhanga. The change of style found in these two vibhangas perhaps indicates that they are later than the first fifteen.

The same suggestion has been made about the 18th vibhanga (Dhammahadayavibhanga), ⁶⁸⁶ for it does not contain the three divisions found in most other vibhangas. Instead of dealing with a single topic, ⁶⁸⁷ it concerns itself with an analysis in terms of certain of the triplets and couplets of the Dhammasangani of twelve of the most important matters of the Buddha’s teaching, e.g. khandhas, āyatanas, dhātus, and truths. The first five to be discussed have already been examined in the first five chapters of the Vibhanga, but the method of approach is somewhat different. It seems to be intended not as a substitute or summary of what has already been given, but as a necessary supplement to it.

4.3. Dhātukathā

Although the Vibhanga devotes a whole chapter to the analysis of the elements (dhātus), the subject itself is of such an intricate and far-reaching nature that particular and detailed attention to its technicalities is required. ⁶⁸⁸ This great expansion of the analysis of the elements forms the contents of the Dhātukathā, ⁶⁸⁹ which forms the third volume of the Abhidhamma-pitaka, and also the third volume, with the Dhammasangani and the Vibhanga, of what may be regarded as a sort of trilogy.

It is a short text of 14 chapters, in the form of questions and answers about the elements of psychical phenomena and their mutual relations. The mātikā⁶⁹⁰ gives the 14 methods, viz. classification and unclassification (samgaho asamgaho), classified and unclassified (samgahitena asamgahitena), etc., under which the internal and external states will be considered, and then lists the 105 internal states, viz. the five khandhas, the twelve āyatanas, etc., and states that the external states are the 22 triplets and 100 couplets of the Dhammasangani mātikā. It also gives the principles of the methods and their characteristics.

The text itself is a working out of the mātikā, whereby all the states are classified under the aggregates (khandhas), bases (āyatanas) and elements (dhātus). Although the elements are also dealt with in the Dhātuvibhanga of the Vibhanga, and in the Dhātuyamaka of the Yamaka, the treatment there is in much less detail. The system employed in the Dhātukathā has been described as a most beautiful piece of work in the mathematical sense, a work of precision and analytical exactness. ⁶⁹¹

The text was expounded in order to dispel wrong ideas about individuality (attā ). ⁶⁹² It deals with the states (dhammas), which are the only things to be found, under the categories of khandhas, āyatanas, and dhātus, which are all anattā (non-self). Therefore there is no attā, but only the rising and passing away of states which are either aggregates, bases, or elements.

4.4. Puggala-paññatti

Despite its presence in the Abhidhamma-piṭaka, the Puggala-paññatti ⁶⁹³ owes much, in both form and content, to the Sutta-piṭaka. The non-metaphysical nature of the book is emphasised by the fact that in it puggala is not used in the sense of “underlying personality” (that is almost synonymous with attā), which is found in the Kathāvatthu and the Milindapañha, but simply in the sense of “person, individual”. This possibly supports the view that the Puggala-paññatti is the earliest of the Abhidhamma texts. ⁶⁹⁴

The text shows signs of having been earlier a work of the Vibhanga type, for the mātikā⁶⁹⁵ begins with the statement that there are six paññattis, viz. those of the khandhas, the āyatanas, the dhātus, the saccas, the indriyas and the puggalas. It then goes on to list the five khandhas, the 12 āyatanas, the 18 dhātus, the four truths, the 26 indriyas, and all the puggalas in classes from one to ten. The text itself, however, goes on to explain only the puggalas in detail. The form is very similar to that of the Sangitisutta of the Dīghanikāya (33), and a portion of the cattāro puggalā section is identical with the cattāro puggalā section of the Sangītisutta. ⁶⁹⁶ Similarly there are parallels with the satta and attha puggalā sections. ⁶⁹⁷

In the text itself of the Puggala-paññatti there are repetitions, inasmuch as the groups of nine and ten merely repeat a list of nine or ten already found in the list of ones. Much of the content of the text is taken from elsewhere, e.g. the tayo, cattāro and pañca puggalā sections are almost the same as passages which occur in the tika-, catukka-, and pañcaka-nipātas of the Añguttaranikāya. ⁶⁹⁸ The text gives the impression of being a collection of most of the statements made about “person” in the Sutta-piṭaka. [^2-4-58]

The book probably represents the results of the disputes which took place in early Buddhism about the nature of the “person”, and which are referred to in the first section of the Kathāvatthu. The Theravādins came to the conclusion that the puggala had no reality, was not one of the dhammas, but was a mere concept (pañnatti). It was therefore not correct for it to be included in the Vibhanga, which dealt with the real dhammas which existed, but it was made into a text by itself. The title “the concept of a person” shows that to the Theravādins the subject was unreal.

4.5. Kathāvatthu

The Kathāvatthu ⁶⁹⁹ is unique among canonical Pāli texts in that it is the only one for which tradition gives us both an author and a date of composition, because the early chronicles state ⁷⁰⁰ that at the time of the third council the elder Tissa set forth the Kathāvatthuppakaraṇa, to refute the heretical doctrines.

This led to a situation where not all the Theravādin schools accepted the text as part of the canon, and Buddhaghosa is at pains to assert its canonical nature in reply to the Vitandins’ objections ⁷⁰¹ that it was uttered by Tissa. He asserts that the table of contents was drawn up by the Buddha, foreseeing that it would be elaborated by Tissa. Consequently Tissa expounded the book not by his own knowledge, but according to the table of contents laid down, and the method given, by the Buddha. Hence the whole book became the word of the Buddha (Buddha-bhāsita). ⁷⁰² The justification is exactly the same as that given for the Madhupindakasutta of the Majjhima-nikāya (18) and other suttas, namely that (in that particular case) Mahākaccāna had answered in exactly the same way as the Buddha would have done, had he been asked. It is possible that Tissa made use of a structure employed by the Buddha and perhaps arguments put forward by him, e.g. in the first statement entitled Puggalakathā.

The Kathāvatthu consists of a discussion of certain statements, of which 500 were orthodox and 500 were unorthodox, according to Buddhaghosa, ⁷⁰³ although a modern authority states ⁷⁰⁴ that it contains the refutation of 252 heretical statements. The work is set out in 23 sections, each dealing with between five and twelve questions. The discussion is set in a framework of question and answer, although it is not always possible to be certain whether the protagonist in each discussion is the Theravādin opponent or the heretic. The commentary usually, but not always, indicates who is speaking.

There is no indication in the Kathāvatthu itself of the names of the propagators of the heresies which are being refuted. These are supplied by Buddhaghosa in his commentary, ⁷⁰⁵ and it has often been pointed out ⁷⁰⁶ that some of the heresies being confuted are ascribed by him to sects whose origins are usually dated many centuries later than Asoka. There is, however, no need in principle to assume that some of the heresies cannot be as early as the date normally given for the composition of the Kathāvatthu, simply because the schools to which they are attributed did not exist before the first century B.C. Very few of the heresies seem to be anything more than obvious quibbles, which the audience to any sermon could have raised. The date of their adoption by a recognised sect could well have been much later.

The framework of the dialogue is stereotyped and abbreviated, which means that it would have been a simple matter to add refutations of new heresies as they arose, merely by following the pattern of dialogue employed. The latest date for the addition of such refutations was probably the first century B.C., when the canon was written down, or the time of the closing of the commentary upon the Kathāvatthu, i.e. the Sinhalese commentary upon which Buddhaghosa’s was based, after which time any insertions would be obvious. There is, however, no need to doubt that the structure itself, and the first few heresies which are refuted, are old. The first section (Puggalakathā) deals ⁷⁰⁷ with the nonexistence of a permanent individuality (puggala), which was one of the Buddha’s most important teachings. The second section ⁷⁰⁸ deals with the refutation of a number of points about the nature of an arahant which so closely resemble what is stated in other traditions to have been the cause of the schism ⁷⁰⁹ which followed the second council that there can be no doubt that they are datable to a century or more before the time of Asoka.

The Kathāvatthu is noteworthy for the number of so-called “Māgadhisms” it contains, especially the nominative singular masculine form in -e. An analysis ⁷¹⁰ of these forms indicates that except for repeated occurrences of one stock phrase, all the statements, heretical and otherwise, are in Pāli. The -e forms are in the framework, where there is a consistent pattern of both -o and -e forms. This gives the impression that there must have been at an earlier time a structure in which the two dialects, viz. Pāli and Māgadhī, were completely differentiated, doubtless to make a clear distinction between the two speakers. This probably means that Tissa was using a structure which was already stereotyped in his time, and perhaps did in fact go back to the time of the Buddha, or at least to the second council.

The Kathāvatthu includes more canonical quotations than any other Pāli canonical text. It quotes from the Vinaya and the Sutta-piṭaka, including nine texts of the Khuddaka-nikāya. It also quotes from the Dhammasangani and the Vibhanga of the Abhidhamma-piṭaka. ⁷¹¹ There are, however, a few discrepancies in these quotations from the texts as they exist today, ⁷¹² which perhaps means that they were being quoted from another recension, or that the author was quoting from memory.

4.6. Yamaka

The Yamaka ⁷¹³ is a book on applied logic, and the name “book of pairs” is derived from the fact that all phenomena are considered in the light of a particular thesis and its opposite antithesis. It enables a student to check whether all examples of a phenomenon are covered by a statement, or only some of them, e.g. does the statement “All X is Y ” mean that all Y is X ?

The text is in ten chapters called Yamakas. Although the printed edition does not have a mātikā as such, the commentary treats the chapter titles as though they formed a mātikā,⁷¹⁴ and the Mohavicchedanī confirms that this is a ten-fold mātikā.⁷¹⁵ The title of each Yamaka explains the subjects to be discussed in it: Mūla, dealing with the roots of kusala and akusala dhammas; Khandha with the five khandhas; Āyatana with the twelve āyatanas; Dhātu with the 18 elements; Sacca with the four noble truths; Saṃkhāra with the three samkhāras (kāya-, vacī- and citta-saṃkhāra); Anusaya with the inclinations; Citta with the mind and mental states; Dhamma with kusala and akusala dhammas; and Indriya with the 22 indriyas.

With the exception of the Mūla-yamaka and the Anusaya-yamaka, the method of treatment in each chapter is threefold. ⁷¹⁶ There is first a Paṇnattivāra, in which the term and its concept are delimited, with a division into the uddesavāra, which states only the questions, and the niddesavāra, which repeats the questions with their various answers. The main part of the chapter is called in the commentary the Pavattivāra. This seems to refer not to procedure generally, but to living process, and considers the problem of where and as what the individual will be reborn. The Pariññāvāra deals with the extent to which a given individual, i.e. a class of beings, understands the category under consideration.

It has been suggested ⁷¹⁷ that such a text, which seems to be intended for someone who has already studied the system, but wishes to become fully competent in it, is perhaps the latest of the books in the Abhidhamma-piṭaka.

4.7. Paṭthāna

The Paṭthāna ⁷¹⁸ is the last and longest book of the Abhidhamma-piṭaka. It deals with the 22 triplets (tikas) and 100 couplets (dukas) of the Dhammasangani, i.e. all the realities, with reference to the 24 conditions (paccayas). It aims to show how the causes and their results are related. In the Paṭthāna, the dhammas in the form of triplets, couplets, couplet-triplet combinations, triplet-couplet combinations, triplet-triplet combinations, and couplet-couplet combinations are treated by four methods: positive, negative, positive-negative and negative-positive. The triplets, etc., are dealt with under seven headings: dependent, co-nascent, conditional, supported, conjoined, associated, and investigation. Under each heading the relations between the conditioning states and the conditioned states of the conditions involved in each state are treated.

The text consists of a mātikā,⁷¹⁹ listing the 24 conditions, viz. hetu, ārammana, adhipati, etc.; a questions chapter; and then the triplets followed by the couplets, etc. In the questions chapter, the questions are asked taking the conditions singly, and then by twos, by threes, etc., up to the maximum combination of 24 , referring to each of the seven ways of taking the triplet (i.e. the seven combinations of each), in relation to each way of taking the triplet, by the four methods, viz. positive, negative, etc., and under the seven headings, viz. dependent, co-nascent, etc. If all the possible combinations were taken, there would be about half a billion questions, ⁷²⁰ but not all can be answered, e.g. for the 49 questions obtained by combining the first (= hetu) condition with dependent heading, in the positive, only nine answers are obtained. In the classification (vibhanga) section these are dealt with at length, but in the enumeration (sankhyaā ) section only the numbers are given.

The text does not give all the questions and answers, but only those necessary for illustrating the types of the questions. So, only the 49 questions for the first condition, taken singly, are given in full, the remainder being abbreviated, although the numbers are given in the enumeration chapter. These chapters are very difficult to understand, since they consist mostly of numerals. Since the enumeration chapters of the Paṭthāna list arithmetically the numbers of answers to each question, the exposition of this subject is very susceptible to presentation by charts. ⁷²¹

4.8. The Abhidharma in Sanskrit

As already noted, ⁷²² the Sarvāstivādins possessed an Abhidharma-piṭaka, which like the Theravādin Abhidhamma-piṭaka contained seven texts. An examination of the titles of these texts, ⁷²³ however, shows that only one, viz. the Dhātukāya, resembles any of the Theravādin texts, and we must either assume that the number seven in each collection is a coincidence, or deduce that the Sarvāstivādins were aware of the number of texts in the Theravādin Abhidhamma-piṭaka and deliberately rejected their texts and replaced them by new ones after the schism. ⁷²⁴

One of the Sarvāstivādin texts is entitled Sangīti-paryāya, and is attributed to Śāriputra in the Chinese translations, but to Mahā-kauṣthila in Sanskrit and Tibetan sources. ⁷²⁵ This text seems to be a commentary upon the Sangitisutta of the Dīgha-nikāya (33), which is also attributed to Sāriputta. ⁷²⁶ A fragment has been found at Bamiyan and published, ⁷²⁷ and more fragments have been discovered in Turkestan. ⁷²⁸

CHAPTER III - EARLY POST-CANONICAL TEXTS

1. EARLY PROSE TEXTS

It seems clear that early in the history of Buddhism a need was felt for guidance on how to interpret suttas, so that those bhikkhus who wished to teach and explain the Dhamma to their followers might be able to do so. The result of this need was the production of two works which are in effect treatises upon hermeneutics, offering instructions for analysing, construing, investigating, and identifying common themes in the Dhamma, so that anyone wishing to interpret could derive from the texts a correct understanding of the meaning of the Dhamma. ⁷²⁹

The Petakopadesa ⁷³⁰ seems to be the earlier of these two works. Its name presupposes a pitaka of which it is the upadesa. ⁷³¹ As already noted, ⁷³² the word pitaka occurs very rarely in the canon in the metaphorical sense of “a basket of scriptures,” but it is not applied to the Buddha’s teachings. Since the word petakin occurs in inscriptions of the second century B.C., there would be no objection to dating the Petakopadesa to that century, or even earlier. There is nothing in the text to suggest any connection with Ceylon, and it was probably composed in India. ⁷³³

The work is ascribed to Kaccāna, ⁷³⁴ but if it was in fact composed by an author of that name, then the need to date the work some time after the completion of the canon means that it cannot be the Kaccāna who was a disciple of the Buddha. Since that Kaccāna was renowned for his analytical powers, ⁷³⁵ it is perhaps not surprising that a book entirely devoted to analysis should be attributed to him. It is probably this attribution which has led to the Petakopadesa being regarded as canonical in Burma. ⁷³⁶

Little is known about the history of the text. Buddhaghosa quotes a passage from a text he calls Petaka, ⁷³⁷ but the quotation he gives is not in our modern editions. He was perhaps quoting from a different text, with a similar name, or from a different recension of our text, ⁷³⁸ or it is possible that the sentence he quotes occurred in a passage which is now missing. The commentator Dhammapāla states that he has collated the Peṭakopadesa with the Nettippakaraṇa ⁷³⁹ before writing his commentary upon the latter text, but at some later date the Peṭakopadesa must have fallen into disuse, doubtless as a result of being eclipsed by its successor, and the manuscript tradition must have been very close to extinction. All available editions of the work must go back to a single manuscript, for they all show the same corruption, arising from the inclusion of a portion of the Sumañgalavilāsinī, from which a page of a manuscript must have intruded at some time without being noticed. ⁷⁴⁰

The method the Peṭakopadesa teaches is conceived simply for the purpose of the correct rewording of known ideas. It is intended to help those who already know intellectually the Buddha’s word and the ideas it promulgates. It is not intended to discover anything new, or to prove any conclusion at all. It is aimed at those who, knowing the Buddha’s Dhamma, wish to explain it and expand it for the benefit of those who do not know but wish to learn. Its purpose is to avoid wrong exegesis, which may lead to contradictions and departures from the teaching.

The method follows the following pattern. ⁷⁴¹ Since any individual “thread” (sutta) of the Buddha’s teaching must, when reworded in explanation of it, be reworded in agreement with the teaching as a whole, and not in conflict with it, and since the teaching as a whole is enormously bulky, sixteen modes (hāras) of conveying a communication are set up, so as to represent the teaching as a whole. When an individual thread is expanded and explained in accordance with these sixteen hāras, the resulting new material should thereby avoid conflict with the teaching as a whole. The hāras deal only with phrasing. For guiding the phrases, when established in this acceptable way, to the meaning-as-aim, viz. liberation from suffering, which is prescribed by the teaching as a whole, five guide lines (nayas) are set out to guide the tested phrasing to the verbal expression of its meaning-as-aim.

Although early Western scholars thought that the Nettippakarana was merely a commentary, it is clear that it is, like the Peṭakopadesa, a guide to those who wished to write commentaries, or give explanations of canonical texts. ⁷⁴² A close comparison of the two texts shows that they are very similar in content. ⁷⁴³ The material which is found in the Peṭakopadesa, but not in the Nettippakarana, seems sufficiently unimportant to the method for its omission to be justified. The material found in both seems to be in an improved form in the Nettippakarana. It therefore seems correct to conclude that neither text is a continuation of the other, but that the Nettippakarana is a rewritten version of the Peṭakopadesa. Its superiority would explain why it has a commentary by Dhammapāla, while the Peṭakopadesa has no commentary. It seems inconceivable that anyone who has read the Nettippakarana could then go on to write the Peṭakopadesa, while the converse seems quite reasonable.

The fact that the Nettippakarana ⁷⁴⁴ contains verses composed in the Āryā metre ⁷⁴⁵ implies that these portions, at least, of the text were composed in North India ⁷⁴⁶ at some time prior to the introduction of the text into Ceylon, for it is clear that the knowledge of the Āryā metre was lost in Ceylon. ⁷⁴⁷ Since the use of the metre was lost in Pāli at an early date, it is likely that the verses must have been composed at some time before the beginning of the Christian era. This, of course, implies a date earlier than this for the composition of the Peṭakopadesa. Like the Peṭakopadesa, the Nettippakarana is regarded as canonical in Burma. ⁷⁴⁸ It is ascribed to Kaccāna, ⁷⁴⁹ who is identified with the Buddha’s disciple by the commentator Dhammapāla.

The influence of the Nettippakarana and, through it, of the Peṭakopadesa is great. Buddhaghosa makes use of its method, and many of the technical terms and normative details employed by him and other commentators are borrowed from it. ⁷⁵⁰ Buddhaghosa, on occasion, quotes from it. ⁷⁵¹

Another early prose text for which an Indian provenance has been surmised is the Milindapañha, ⁷⁵² which gives an account of a dialogue between the Greek king Milinda (= Menander) and the thera Nāgasena. Nothing is known certainly about the origin of this text. An earlier limit for its date of composition is given by the fact that Menander ruled in the middle of the second century B.C., ⁷⁵³ while it must have been completed by the fifth century A.D., since Buddhaghosa quotes from it, ⁷⁵⁴ and clearly regards it as authoritative. Since the words he quotes do not agree entirely with the text as we have it, it is not absolutely certain that the text he had available to him was identical with ours.

From the facts that Menander’s kingdom was in North-West India, ⁷⁵⁵ his ministers have names which can in some cases be related to genuine Greek names, ⁷⁵⁶ some knowledge of a typical Greek city is shown, ⁷⁵⁷ all geographical details relate to North India, ⁷⁵⁸ and there are one or two words and phrases in the text which do not follow the usual pattern of Pāli, ⁷⁵⁹ it has been assumed that the Milindapañha is a Pāli translation from a Sanskrit or Prakrit original composed in North India about the beginning of the Christian era. ⁷⁶⁰ There is also a Chinese version, which can be dated to a time earlier than the fourth century A.D. ⁷⁶¹ Since it differs somewhat from the Pāli version, it has been thought to be based upon the (hypothetical) Sanskrit version. It has also been suggested that the original of the Milindapañha was in fact written in Greek. ⁷⁶²

It is clear that the text is made up of material composed at different times. From the fact that the words Milindapañhānaṃ pucchāvissajjanā samattā occur at the end of Book III, ⁷⁶³ it has been deduced that this was the end of the original portion of the Milindapañha, but it is possible to see that even in this short portion there are insertions. There is a reference to the six teachers ⁷⁶⁴ which is entirely anachronistic, since it refers to conditions at the time of the Buddha, and is based upon the Sāmaññaphalasutta of the Dīgha-nikāya (2). There are also sections ⁷⁶⁵ dealing with the earlier births of Milinda and Nāgasena, and a biography of Nāgasena from his conception to the attainment of arahant-ship. xxiv. W. W. Tarn, The Greeks in Bactria and India, Cambridge 1938, pp. 225-26, suggests that he reigned c. 161-150 / 145 B.C.

In the remaining sections of this older part, the king puts forward problems about human identity ⁷⁶⁶ and the continuity of human existence, and Nāgasena solves them. Books IV-VII ⁷⁶⁷ are different in style, and are clearly later than the first portion and probably by a different author. Since Buddhaghosa quotes from this later part, it must have been composed before the fifth century A.D., but there is no way of telling how much before. This portion of the Milindapañha deals with the solution of puzzles which arise from apparently contradictory statements made by the Buddha. These are presented in the form of 82 dilemmas put to Nāgasena by Milinda. The texts from which extracts are quoted in these puzzles include the greater part of the canon, ⁷⁶⁸ and reveal the wide breadth of knowledge of the author of this portion of the text.

There are other indications in the Milindapañha that the tradition which lies behind it differs somewhat from that of other Pāli texts. One problem which it raises is that of the presence in it of a number of verses which are attributed to theras, and yet are not found in the Thera- or Therīgāthā, or elsewhere in the canon. ⁷⁶⁹ These are quoted by the author of the Milindapañha with as much or as little authority as the other theras’ verses he quotes. It seems likely that they were transmitted in some non-Theravādin, or at least some non-canonical Theravādin source, ⁷⁷⁰ perhaps in North India, which is no longer available to us. In support of the first suggestion is the fact that one feature included in the Milindapañha is very likely due to Sarvāstivādin influence. ⁷⁷¹

Although, as already noted, Buddhaghosa sometimes quotes from the Milindapañha by name, there are some verses in the Visuddhimagga and other commentaries which are identical with verses in the Milindapañha, but the latter text is not quoted as the source. On one occasion the verse is ascribed to the porannas, ⁷⁷² and this perhaps indicates that both the Milindapañha and Buddhaghosa took it from the mass of floating verses in the commentarial tradition. The author of the Milindapañha was also acquainted with a school of Khuddaka-bhānakas, ⁷⁷³ which seem not to be mentioned by any other source. ⁷⁷⁴

The relationship between these bhānakas and the Jātaka-bhānakas, whom he mentions at the same time, is not clear.

The fourth of these early prose texts which were probably composed in North India is the Vimuttimagga, and as will be noted below there is some doubt as to whether this work is still extant. Its importance lies in the fact that it was made use of by Buddhaghosa when he wrote the Visuddhimagga. Buddhaghosa does not quote the Vimuttimagga by name, but the author of the tikā upon the Visuddhimagga states that the word ekacce (at Vism 102, 31) refers to the thera Upatissa who was responsible for proposing a refuted method of classifying temperaments in the Vimuttimagga. ⁷⁷⁵ A Chinese translation of this text, made in A.D. 505 , exists, ⁷⁷⁶ and it includes the refuted suggestion. It also includes other points rejected by Buddhaghosa, which Dhammapāla ascribes to the Abhayagiri vihāra. ⁷⁷⁷ This has led to a belief that the Vimuttimagga was an Abhayagiri text, ⁷⁷⁸ which was regarded as lost with almost all other Abhayagiri texts when the Mahāvihāra finally triumphed over its main opponent. It has, however, been pointed out that all the points which Buddhaghosa refutes are non-Mahāyāna, and are well within the Hīnayāna field. ⁷⁷⁹ The Vimuttimagga has, therefore, been identified as a Theravāda text. ⁷⁸⁰

It has been suggested ⁷⁸¹ that the author Upatissa was the same as the Upatissa who is mentioned in the lists of theras in the Parivāra of the Vinaya-piṭaka. If this suggestion is correct, then the text would date from about the first century A.D., but there is no other evidence to support this suggestion. All that can be said is that, if Dhammapāla is correct, Buddhaghosa knew of the Vimuttimagga, which must therefore be earlier than the fifth century A.D., but not necessarily much earlier.

A comparison of the Visuddhimagga with the Vimuttimagga, in its Chinese form, shows that the latter is much shorter than the former. ⁷⁸² It follows the same three general divisions of sīla, samādhi, and paññā, but does not superimpose the pattern of the seven visuddhis. Less space is devoted to the paññā portions and there are no illustrative stories. Although the appearance in both works of what appear to be several nearly identical passages suggests that they both drew much from the same sources, the general style differs. The main differences between the two texts arise from the way of handling the four measureless states and the four immaterial states, while the way in which the Vimuttimagga describes the thirteen ascetic practices (dhutaṇgas) is quite different too. It has been suggested ⁷⁸³ that the Vimuttimagga was composed in North India, and the existence of a Tibetan version of the portion dealing with the dhutaṇgas possibly supports this view, ⁷⁸⁴ for translations from Pāli into Tibetan are very rare. ⁷⁸⁵ Another view, however, concludes that it was the product of either the Abhayagiri vihāra or the Jetavana vihāra. ⁷⁸⁶ The presence in it of some points accepted by the Abhayagiri vihāra does not, however, prove that it was a product of that centre. It could well have been composed elsewhere and then adopted by the Abhayagirivihārins.

An edition of a text entitled Vimuttimagga ⁷⁸⁷ was published in Ceylon in 1963. A reviewer of this edition has stated that in his opinion ⁷⁸⁸ it is a modern work, based upon a critical analysis of the Visuddhimagga and the Chinese translation of the Vimuttimagga, which he himself had published some years earlier.

2. THE EARLY CHRONICLES

Tradition, as recorded by Buddhaghosa, ⁷⁸⁹ states that the commentaries were brought to Ceylon by Mahinda. It can be deduced that included in this commentarial material was information about the early history of Buddhism in North India. To this in the course of time was added material concerning the introduction of Buddhism to Ceylon, and its subsequent progress there.

Nothing is known about the language or languages in which this historical material was composed. It seems probable that, like the canon itself, it was originally in a number of dialects, but it is uncertain whether it was reduced to a single dialect when the canon was rendered into Pāli. It is also probable that the additions made in Ceylon were in the current Sinhalese Prakrit. From the form of later texts which were based upon this historical tradition it has been deduced that it consisted of a mixture of prose and verse, and it has been claimed that the prose was in Sinhalese Prakrit while the verses were in Pāli. ⁷⁹⁰ The main reason for this belief appears to be the fact that later classical Sinhalese prose is often interspersed with Pāli and Sanskrit verse, but almost never with Sinhalese verse. ⁷⁹¹ As will be seen, there is some evidence that some, at least, of the verses were in a non-Pāli dialect of Middle Indo-Aryan. ⁷⁹²

The first extant attempt to put together history and legend in verse and to make a continuous chronicle (vamsa) about Ceylon is the Dīpavamsa. ⁷⁹³ This tells in 22 chapters the history of the island from the time of the Buddha, who is reported to have visited Ceylon three times, down to the end of the reign of Mahāsena in the fourth century A.D. ⁷⁹⁴ It must therefore have been written after the time of that king. The fact that it is quoted by name several times by Buddhaghosa ⁷⁹⁵ suggests that it is to be dated to the period between Mahāsena and Buddhaghosa, but some doubt remains because there are minor differences between Buddhaghosa’s quotations and the text which we possess, which may mean that Buddhaghosa was following a different recension of the text, ⁷⁹⁶ or that he was quoting from an earlier version of the text than the one we have, perhaps even a Dīpavamsa-atthakathā upon which our Dīpavamsa is based.

Before beginning the account of the history of Ceylon, the early history of Buddhism in India is related, down to the time of Asoka and the sending out of the missionaries, and including accounts of the three councils. The story of Ceylon proper begins in chapter nine. The text shows signs of being made up from a collection of material which frequently overlaps. The author makes no attempt to assimilate his material, but is content to place different versions of the same occurrence side by side. So we find two slightly divergent names of the same individual, ⁷⁹⁷ two descriptions of the first council ⁷⁹⁸ and of the second council, ⁷⁹⁹ with some verses identical in the two versions. There are also two versions of the third council, ⁸⁰⁰ while the story of the presents sent to Ceylon by Asoka is treated three times. ⁸⁰¹

Many verses in the Dīpavamsa are so sketchy that they are unintelligible without help. Where the relevant stories recur in the later Mahāvamsa the essential details are added, and we are able to use the later text as a commentary upon the earlier one. ⁸⁰² There are also verses which were clearly intended to be memory verses. ⁸⁰³ In the form in which they occur in the Dipavaṃsa they are sometimes meaningless, and it is only the fact that they are elaborated in the Mahāvaṃsa which enables us to understand them. ⁸⁰⁴ These facts are consistent with a belief that the Dipavaṃsa was dependent upon a written tradition which was in turn based upon an oral tradition, in which memory verses were essential.

The author’s name is unknown. His command of Pāli seems to have been poor. The syntax of some of his sentences is very difficult, and he is capable of producing grammatical monstrosities. ⁸⁰⁵ The metre is Sloka and Jagatī, except for a few passages which seem to be in prose, ⁸⁰⁶ although it has been claimed ⁸⁰⁷ that these are really verses in irregular metres.

Although the introductory Bāhiranidāna which Buddhaghosa prefixes to the Samantapāsādikā ⁸⁰⁸ is not a vamsa, it is convenient to deal with it here because it is to a large extent based upon the same materials as the Dipavaṃsa. Buddhaghosa wrote this introduction to explain the Nidāna, the “inception” or “origin” of the Vinaya, with the aim of establishing the authenticity of the Vinaya before starting to comment upon it. ⁸⁰⁹ To this end he gives an account of the history of Buddhism down to the proclamation of the Vinaya-piṭaka by Mahā-ariṭṭha in the presence of Devānaṃpiyatissa, under the presidency of Mahinda, after the introduction of Buddhism to the island.

In his account of the first three councils Buddhaghosa is dependent upon the historical material collected at the Mahāvihāra. As we have seen, it is clear from the use made of the same material by the author of the Dipavaṃsa that it consisted of two or three versions of some of the events, which had been transmitted in a quite disjointed fashion. Whereas these versions are simply included side by side in the Dipavaṃsa, Buddhaghosa makes a very successful attempt to integrate them.

Buddhaghosa twice quotes ⁸¹⁰ the Dipavaṃsa by name when quoting verses, but the fact that there are slight variations from the text of the Dipavaṃsa which we possess perhaps means that he was quoting from a different recension, or that he was quoting directly from the Sihala version of the Dipavaṃsa, upon which the text we have was based. Since some of the verses are memory verses, it is likely that they are old ones from the Sihala-atṭhakathā. On a further occasion Buddhaghosa quotes a verse found in the Dīpavamsa, but he merely introduces it with the words yathāha. ⁸¹¹ Another verse is given without any reference to the source, ⁸¹² and two further verses, almost identical with two in the Dipavaṃsa, are quoted on the authority of the porānas, ⁸¹³ which would suggest that Buddhaghosa was indeed quoting the sources from which the Dipavaṃsa itself drew its material. On the other hand, in his commentary upon the Kathāvatthu Buddhaghosa quotes a long passage from the Dīpavamsa ⁸¹⁴ which agrees so closely with the text which we have that it is certain that, for that portion at least, Buddhaghosa had access to a text virtually identical with ours.

The second of the great vamsa texts is the Mahāvaṃsa. ⁸¹⁵ The commentary upon the text states that the author’s name was Mahānāma. ⁸¹⁶ He gives as his reason for writing the Mahāvaṃsa the fact that the earlier work compiled by the porānas was sometimes too lengthy and sometimes too abbreviated, and contained many repetitions. ⁸¹⁷ This is so clearly a description of the Dipavaṃsa that it has been thought to apply to it, ⁸¹⁸ and it was assumed that the Mahāvaṃsa was intended simply as a replacement of the Dipavaṃsa. The author of the commentary upon the Mahāvaṃsa, however, states that the author of the Mahāvaṃsa was referring to a Sīhalaṭ̣̣hakathā-Mahāvaṃsa, ⁸¹⁹ which was full of faults. This was presumably a historical text, or an introduction to a work. The commentary further states that the Mahāvaṃsa is a translation of the porānatthakathā from the Sīhala language into Māgadhika-bhāsā, i.e. Pāli, and calls the text Padyapadoruvaṃsa because it is written in verse. ⁸²⁰

The Mahāvaṃsa begins and ends where the Dipavaṃsa does, ⁸²¹ but it both adds and subtracts. It omits, for example, the description of the changes made to the canon by the Mahāsānghikas after they had split from the Theravādins, ⁸²² but whereas the Dipavaṃsa devotes only 13 verses to Duṭṭhagāmaṇī, ⁸²³ the Mahāvaṃsa inserts what was doubtless in origin an independent poem about him, comprising eleven chapters. ⁸²⁴

Mahānāma may be said to have succeeded in his aim of writing a work which is free from faults. ⁸²⁵ The reason for the fact that the Mahāvamsa is so superior to the Dīpavamsa is perhaps to be seen in Buddhaghosa’s work. It has been suggested that Mahānāma educated himself for the task of writing his great chronicle by reading Buddhaghosa’s introduction to the Samantapāsādikā. ⁸²⁶ The text has many claims to be regarded as a kāvya. ⁸²⁷ The metre throughout is Sloka, apart from a prose quotation from the Yamaka, ⁸²⁸ except that the final verse of each chapter is in an ornate metre, e.g. Sragdharā (I), Rucirā (II), Vasantatilakā (III), etc.

Although the Mahāvamsa we possess is a very long work, taking the history of Ceylon right down to the nineteenth century, the older original composition, of which alone the commentary gives an explanation, ends in the middle of chapter XXXVII, where we find after verse 50 the words Mahāvamso nitthito. Verse 50 is in the Sloka metre, which indicates that it was not the original ending to the chapter, for which an ornate metre would have been expected. Presumably the original ending was removed when the chapter was coni i aued. ⁸²⁹

It would seem that the Dīpavamsa, Buddhaghosa’s Bāhiranidāna, and the Mahāvamsa are all based on the Sīhala atthakathā material which was kept at the Mahāvihāra, and the differences between these three works are due to the various authors making use of different strands of the material, and laying different emphasis upon the material which they have in common. It has been suggested ⁸³⁰ that some of the differences between the Bāhiranidāna and the Mahāvamsa arise from the fact that the former was translated directly from Sinhalese Prakrit prose into Pāli prose, while the author of the latter was under the restriction of metrical considerations. The differences between verses in the Dīpavamsa and the Bāhiranidāna have been put down to recension variation, ⁸³¹ but such differences could equally well be explained as arising from alternative interpretations of verses in non-Pāli dialects. ⁸³²

3. THE EARLY COMMENTARIES

At the beginning of several of his commentaries Buddhaghosa states ⁸³³ that he is basing his explanations upon the atthakathās which were first recited by the 500 theras, i.e. at the first council, and afterwards brought to Ceylon by Mahinda and translated into the Sinhalese language for the sake of the inhabitants of the island. From the many references to events and personages in Ceylon which we find in Buddhaghosa’s commentaries, and which were presumably taken from the Sīhala atthakathās, we can be certain that additions were made to those atthakathās after their arrival in Ceylon. A critical examination of the commentaries suggests that no additions were made to them after the first century A.D., ⁸³⁴ but since it is said that the Sīhala atthakathās were committed to writing at the same time as the canon in the first century B.C., it may be possible that little exegetical material was added after the latter date.

There is no direct evidence that any commentarial material was in fact recited at the first council, but there is clear evidence that some parts of the commentaries are very old, perhaps even going back to the time of the Buddha, because they afford parallels with texts which are regarded as canonical by other sects, and must therefore pre-date the schisms between the sects. ⁸³⁵ As has already been noted, ⁸³⁶ some canonical texts include commentarial passages, while the existence of the Old Commentary in the Vinaya-piṭaka and the canonical status of the Niddesa prove that some sort of exegesis was felt to be needed at a very early stage of Buddhism.

We cannot be certain whether Buddhaghosa was correct in stating that the early missionaries translated the commentaries they brought with them into Sinhalese Prakrit, but what we know of the development of that dialect between the first century B.C., when the commentaries were allegedly written down, and the fifth century A.D., when Buddhaghosa is believed to have written his commentaries, suggests that he was quite correct when he stated that the atthakathās in the Sīhalabhāsā were no longer intelligible outside Ceylon. ⁸³⁷ It may be suspected that they were only intelligible to a very small number of bhikkhus inside the island.

The fact that the Dīpavamsa sometimes gives two or three versions of the same event suggests that the tradition of the Mahāvihāra was based upon a mass of disjointed chronicle material, and there is no reason to doubt that the situation with regard to exegetical material was exactly the same, with comments gathered together from a number of sources. This would explain why we sometimes find commentators giving two or more, sometimes contradictory, explanations of the same word or phrase. Some commentators, especially Buddhaghosa, ⁸³⁸ give information about the commentaries they are following, and from the way in which some views are brought forward only to be rejected it is clear that those living at the Mahāvihāra also had access to the records of others sects which were regarded as hostile, if not actually heretical.

3.1. Buddhaghosa

The Mahāvamsa states ⁸³⁹ that after his conversion to Buddhism in North India Buddhaghosa wrote a book entitled Nāṇodaya and then a commentary upon the Dhammasañgani called the Atthasālinī. He then began to compile a commentary upon the Paritta, but was advised to go to Ceylon, for only there was the commentarial tradition preserved. In Ceylon he went to the Mahāvihāra, learned the commentaries, and wrote the Visuddhimagga, described as a summary of the Tipitaka, to persuade the Sangha that he was a fit person to be entrusted with their books.

The story in the Mahāvamsa states that to test his ability the theras of the Mahāvihāra gave him two verses to comment upon, and the account given at the beginning of the Visuddhimagga does not contradict this. ⁸⁴⁰ He begins by quoting two verses from the Samyutta-nikāya, ⁸⁴¹ and the whole of the Visuddhimagga is in effect a commentary upon the first of these. He states that the book was written at the request of the thera Sanghapāla, who is named as his teacher in the Mahāvamsa. ⁸⁴²

The Visuddhimagga ⁸⁴³ is not so much a commentary as a sort of encyclopaedia, a compendium of Buddhist doctrine and metaphysics presented in a logical and systematic manner under the three heads of sīla, samādhi, and paññā, as Buddhaghosa himself states. ⁸⁴⁴ These three headings have already been noted ⁸⁴⁵ in the Vimuttimagga, and there seems no doubt that Buddhaghosa made use of this earlier text when writing his own work, although it is going too far to say that the Visuddhimagga and Vimuttimagga are one and the same work appearing in different attires. ⁸⁴⁶

The Visuddhimagga extracts from the Piṭakas all the central doctrines which pivot upon the four truths, and presents them as a coherent systematic whole by way of quotation and explanation, interspersed with treatises on subjects of more or less relative importance. It illustrates by means of a large number of stories, set in either India or Ceylon, the whole being assimilated into an elaborate edifice. An examination of the stories shows that those referring to India are, where a date can be assigned, none of them later than the third century B.C., i.e. the date when Mahinda brought Buddhism to Ceylon. ⁸⁴⁷ None of the events in Ceylon can be dated later than the latest datable thera, and this, as has been noted, ⁸⁴⁸ is either the first century B.C. or the first century A.D.

The Visuddhimagga is a masterly summary of Buddhist teaching, and can be regarded as an exposition of the ninefold teaching of the Buddha. It is best regarded as a detailed manual for meditation masters, and as a work of reference. As such, it can stand alone, but since many of the passages Buddhaghosa quotes occur in texts upon which he has written independent commentaries, there is some overlap, and many passages commenting upon identical material occur word for word the same in each commentary.

The Gandhavaṃsa states ⁸⁴⁹ that Buddhaghosa was the author of twelve commentaries in addition to the Visuddhimagga: those on the Vinaya-piṭaka, the first four nikāyas, the Abhidhamma-piṭaka, the Pātimokkha, and the Dhammapada, the Jātaka, the Khuddakapāṭha, the Suttanipāta ⁸⁵⁰ and the Apadāna. Although there is some doubt about the last five, the remainder are generally accepted as being by Buddhaghosa.

At the beginning of the Samantapāsādikā ⁸⁵¹ Buddhaghosa states ⁸⁵² that he is basing his commentary upon the Mahā-aṭṭhakathā, the Mahāpaccariya and the Kurundī commentaries, and he is embodying the tradition of the elders (theravāda) in it. ⁸⁵³ He also refers to the opinions of individual elders, and clearly holds Mahāpaduma and Mahāsumma in high regard. ⁸⁵⁴ He also refers to the Sankhepa, an abridged form of the Mahāpaccarī, and to the Paccarī commentary, which is perhaps also an abridged version. ⁸⁵⁵ He also mentions the Andhaka-aṭṭhakathā, ⁸⁵⁶ although almost always to reject its views. ⁸⁵⁷ The fact that the Andhaka country in South India was a Dravidian-speaking area has led to the suggestion that this commentary came from that area and was in a Dravidian language. [^3-130] There is no evidence for this belief, either as a specific statement from Buddhaghosa or as a deduction from the views quoted from the Andhaka commentary. Although there are one or two references in other commentaries by Buddhaghosa to a Vinaya-aṭṭhakathā, ⁸⁵⁸ there are no references to it by name in the Samantapāsādikā. The references to it do not fit in well with the Samantapāsādikā itself, and it seems likely that it is a Sīhala atthakathā.

To the commentary upon the Vinaya-piṭaka proper, Buddhaghosa prefixes a Bāhiranidāna which has already been mentioned. ⁸⁵⁹ Besides the chronicle portions which can be compared with the other chronicles, there is a great deal of material about the classification and content of the canonical texts.

The Sumañgalavilāsinī, ⁸⁶⁰ Papañcasūdanī, ⁸⁶¹ Sāratthappakāsinī, ⁸⁶² and Manorathapūranin ⁸⁶³ are Buddhaghosa’s commentaries upon the Dīgha-, Majjhima-, Saṃyutta-, and Añguttara-nikāyas respectively. He states at the beginning of each of these that since certain things have already been dealt with in the Visuddhimagga, he will not deal with them again, for the Visuddhimagga stands between and in the midst of all four collections. ⁸⁶⁴ In the Samyutta commentary he inserts a stanza referring to the two previous collections, ⁸⁶⁵ i.e. the Dīgha-nikāya and the Majjhima-nikāya, for the explanation of the names of towns and illustrative stories. The Anguttara commentary replaces this by another stanza referring to the Dīgha and the Majjhima by name for the same purposes. ⁸⁶⁶ Since the Visuddhimagga is referred to in this way at the beginning of each of these four commentaries, we must assume that when he refers in the Visuddhimagga to the atthakathās upon the Anguttara-, Majjhima-, and Saṃ-yutta-nikāyas, ⁸⁶⁷ he is not referring to his own commentaries upon these texts, but to the Sīhala atthakathās thereon.

Buddhaghosa’s commentaries upon the Dhammasañgani and the Vibhanga are individually named as the Atthasālin ⁸⁶⁸ and the Sammohavinodanī, ⁸⁶⁹ whereas unlike them the commentaries upon the remaining five Abhidhamma texts are taken together under the title Pañcappakaraṇaṭthakathā or Paramatthadīpanī. ⁸⁷⁰ The Gandhavaṃsa uses the name Paramatthakathā ⁸⁷¹ for the commentary upon all seven Abhidhamma works. The name Paramatthadīpani is little used, perhaps because of the possible confusion with Dhammapāla’s commentary ⁸⁷² upon certain texts in the Khuddaka-nikāya.

For the most part, in the Abhidhamma commentaries their author avoids discussion of matters which have already been dealt with in the nikāya commentaries and the Visuddhimagga. That this is a definite policy, and not oversight, is clear from the fact that readers are referred specifically to the Visuddhimagga ⁸⁷³ or the nikāya commentaries, although he seems not to claim that these are his own work, whereas in the nikāya commentaries he does state that the Visuddhimagga is his own work. ⁸⁷⁴ There can, however, be no doubt that all the Abhidhamma commentaries are by one author, for at the end of the commentary upon the Paṭthāna he states that he has now concluded the commentary to the whole Abhidhamma-piṭaka. ⁸⁷⁵ The Atthasālinī, Sammohavinodanī, and the Pañcappakaraṇaṭthakathā all include the name Buddhaghosa in the epilogue, ⁸⁷⁶ but other than that there is no evidence that they were written by him. In fact, the first two state that they were written at the request of Buddhaghosa, ⁸⁷⁷ which has led to the suggestion that they were both written by a pupil of Buddhaghosa. ⁸⁷⁸ Examples of references to the Visuddhimagga and the nikāya commentaries can be found in all three works, where they are treated as authorative, but not as the author’s own work.

It has been suggested ⁸⁷⁹ that the Atthasālinī is not by the same author as the Visuddhimagga, for there are a number of points where the interpretation of the two texts differ somewhat, while authorities quoted in one text are not always included in the other, and on other occasions the views given in the Visuddhimagga are merely listed as alternatives. It seems strange that someone like Buddhaghosa should give his own interpretations as alternatives, and if he were really revising an earlier opinion it might have been expected that he would mention his change of mind. Strong though these arguments may be, they are not conclusive. The answer perhaps lies in the relative dating of the Visuddhimagga and the Atthasālinī. The Mahāvaṃsa states ⁸⁸⁰ that Buddhaghosa wrote the Atthasālinī in India, before going to Ceylon. If this was so, then he must have re-written it there, because he refers to commentarial material from the Mahāvihāra. ⁸⁸¹ If we accept that the Atthasālinī was an early work which was later re-written, ⁸⁸² then it does not seem unlikely that Buddhaghosa should incorporate later views from the Visuddhimagga while still retaining earlier views expressed in the “first edition” of the Atthasālinī. There are several references in the Sammohavinodanī to the Atthasālinī as the commentary which should take preference over it, ⁸⁸³ and since Buddhaghosa’s authorship of the former work has never been disputed it would seem reasonable to suppose that he wrote the latter work too. The suggestion that the Atthasālinī was written by a pupil of Buddhaghosa need not perhaps be taken very seriously. A study of Buddhadatta’s Abhidhammāvatāra shows parallels with the Atthasālinī, ⁸⁸⁴ and although dependence upon the same traditional commentarial material cannot be ruled out as the cause of this, the tradition ⁸⁸⁵ that Buddhadatta summarized Buddhaghosa’s commentaries suggests that the Abhidhammāvatāra is indeed based upon the Atthasālinī. Buddhadatta is not likely to have summarized something written by a pupil of Buddhaghosa.

Another possible solution for the problem of Buddhaghosa seemingly rejecting the views he had put forward in the Visuddhimagga is that these views may have been his own, ⁸⁸⁶ as opposed to traditional ones taken over from the Sihala atthakathās. Although the implication is that Buddhaghosa is at all times following the Mahāvihāra tradition, he, and other commentators, ⁸⁸⁷ do sometimes state that they are giving their own opinion about something, perhaps because their predecessors had given no information about it. It is possible that this happened more frequently than would appear. It is clear that on occasion the views put forward by the commentators lacked authority, since readers are sometimes allowed a choice about accepting the view or rejecting it. ⁸⁸⁸

Although, as stated, the Sammohavinodanī several times refers to the Atthasālinī, we also find in the Atthasālinī the statement, repeated, ⁸⁸⁹ that a point will be clarified in the Vibhanga-atṭ̣hakathā. This presumably means that Buddhaghosa was carefully planning ahead for one commentary while he was writing another, and was already making decisions about the proper place to deal with the traditional material he was sorting through. It may be assumed that a similar practice accounts for other cross-references which are found in his commentaries, ⁸⁹⁰ although extensive reworking and rewriting of his commentaries occurred, which would have enabled him to incorporate later views in books written earlier.

The five commentaries included in the Pañcappakaraṇaṭ̣̣hakathā ⁸⁹¹ are much shorter than the Atthasālinī and Sammohavinodanī. The most interesting of the set is the commentary upon the Kathāvatthu, ⁸⁹² which makes a considerable contribution to the study of the history of early Buddhism by identifying the greater part of the heretical views put forward in the Kathāvatthu as being the views of particular sects. It starts with a brief account of the origin of the 18 schools which arose before the third council, and then the six more which arose after the end of the second century of Buddhism.

The author of the commentary is clearly aware that some of the views referred to could not have arisen in Tissa’s time. He explains this by stating that Tissa spoke to prevent views which were going to arise, by analysing the topics of discourse (mātikās) established by the Tathāgata. ⁸⁹³ When identifying the holders of the views the commentator sometimes uses the word “now” (etarahi), implying that the schools still existed in his time and still had that view. ⁸⁹⁴ It is, however, possible that Buddhaghosa took over the word etarahi from the Sīhala atthakathās he was following, and “now” refers to the time when the latter were compiled. ⁸⁹⁵

Although, as already noted, ⁸⁹⁶ the Pātimokkha does not have a separate existence in the Pāli canon, being embedded in the Vinaya-piṭaka, Buddhaghosa wrote a commentary upon it, including the introductory portion which has become detached from the rest and is found, together with the Old Commentary, in the Uposathakkhandhaka. ⁸⁹⁷ He comments upon the Old Commentary as well as upon the Pātimokkha itself. He states that he is doing it briefly (sankhepato), and he refers to the Samantapāsādikā for a more detailed account. ⁸⁹⁸

The Kaṅkhāvitaranī ⁸⁹⁹ calls itself a mātikā-atthakathā,⁹⁰⁰ where mātika means the text of the Pātimokkha extracted from the Suttavibhanga and, in the case of the introductory portion, the Mahāvagga. This use of mātika is found elsewhere in works by Buddhaghosa. It has already been noted ⁹⁰¹ that he explains the word mātikā-dharā as dvepātimokkadharā, and in the Kañkhāvitaranī we find the Mahāvibhanga referred to as mātika ⁹⁰² and also a distinction drawn between the mātika and the Padabhājana. ⁹⁰³ In the Visuddhimagga mātika seems to be used of the Pātimokkha, ⁹⁰⁴ and in the Atthasālinī it refers to a part of the Vinaya-piṭaka. ⁹⁰⁵

The word mātika in this sense must be old, for the Sīhala atthakathā is called the Sīhala-mātikā-atṭhakathā. ⁹⁰⁶ Besides this, Buddhaghosa presumably also made use of the same commentaries he employed when writing the Samantapāsādikā, for he mentions the Kurundī and the Mahāpaccarī commentaries. ⁹⁰⁷ These authorities did not always agree, and Buddhaghosa notes that the Sīhala atthakathā differs from the explanation given at one point in the Samantapāsādikā. ⁹⁰⁸ References to the Samantapāsādikā are frequent, and there are more than a dozen passages which are very similar in content, ⁹⁰⁹ although it is impossible to say whether the Kañkhāvitaranī is actually quoting from the Samantapāsādikā, or whether both texts are quoting from the Sīhala atthakathā.

The exegesis given for all 227 rules of the Bhikkhu-pātimokkha ⁹¹⁰ follows the same pattern. ⁹¹¹ There is a parsing of the words of the rule (sikkhāpada), followed by a quasi-judicial estimation of defaults against it. Extenuating circumstances are then mentioned, and cases of guiltlessness, including wrong actions due to misconception or errors of judgment. A unique feature of the exegesis is the inclusion of a catechetic and hermeneutic summary of the main points at the end of each rule. These are called añgas because they form the links of the secondary rule (anuppaññatti) which follows from the main rule. They are probably intended to serve as aids to the memory for the trainee bhikkhu.

Of the other commentaries attributed to Buddhaghosa in the Gandhavamsa it is generally agreed that the Apadāna commentary was not written by him. ⁹¹² It is sometimes argued that the commentaries on the Dhammapada and the Jātaka may have been written by another Buddhaghosa because their style is so different from his other commentaries. ⁹¹³ The commentary upon the Dhammapada ⁹¹⁴ is said to have been written at the request of a thera named Kumārakassapa, ⁹¹⁵ and to be based upon commentarial material in the language of the island (dīpa-bhāsāya santhitā). ⁹¹⁶ At the end it is said that the work was composed by Buddhaghosa while residing in a residence built by king Sirikūṭa. ⁹¹⁷ The reason for the difference in style could be the difference in subject matter, for the work consists mainly of popular stories which play only a minor role in Buddhaghosa’s other works. ⁹¹⁸ There seems, however, to be no great objection to believing that Buddhaghosa collected together a set of traditional stories from commentarial sources and served as little more than an editor for this part of the work. His main commentarial duties were confined to the word for word commentary. In this connection it has been pointed out that while in his other commentaries Buddhaghosa is entirely consistent in his explanations of Dhamma, he is not always so consistent when giving different versions of the same story, e.g. the stories which he tells of the theras Tissabhūti⁹¹⁹ and Mahatissa. ⁹²⁰ It is possible that there was felt to be less need for strictness in such story material, although shortcomings would be much more obvious in a text consisting almost entirely of stories. ⁹²¹

The same would apply to the commentary upon the Jātakas. ⁹²² At the beginning the author states that he will give the commentary in accordance with the tradition of the Mahāvihārins, at the request of the three theras Atthadassin, Buddhamitta, and Buddhadeva, ⁹²³ the last of whom is said to belong to the Mahimsāsaka sect. [^3-197] There is no mention of the author’s name at the end. As has already been noted, ⁹²⁴ the stories, the framework, and the word for word commentary have been shown to be composed by the same person, but we may be certain that in the case of the greater part of the stories the author’s work, as in the case of the Dhammapada-atṭhakathā, consisted mainly of an editorial function, whereby a previously existing story was adapted, not always skilfully, for a new purpose.

Something can be seen of the way in which the editorial work was carried out. Although the story of the present is given in full for the first few Jātakas, it is thereafter very often all but omitted, being replaced by the statement that the Buddha told this story of the past. [^3-199] Sometimes the incident which causes the telling of the story is omitted, with a statement that it is the same as one already told or, in many cases, one which is to be told. ⁹²⁵

Among the commentarial sources which the author states he is following is a Jātakaṭṭhakathā, which was presumably a Sīhala atthakathā. Its disagreement with the remainder of the atthakathās is noted on one occasion. ⁹²⁶ He also refers to the Añguttaraṭṭhakathā ⁹²⁷ and the Saṃyuttaṭthakathā, ⁹²⁸ but we cannot be certain whether the commentaries by Buddhaghosa or the Sīhala atthakathās are intended. On a number of occasions he draws attention to the fact that a reading in the books (potthakesu) is not found in the atthakathās,⁹²⁹ or that the atthakathā has a different reading.⁹³⁰ He sometimes quotes, tacitly, material from an earlier commentary. ⁹³¹

To Buddhaghosa are also ascribed commentaries upon the Khuddakapāṭha and the Suttanipāta, both called Paramatthajotikā. Despite the fact that Buddhaghosa’s name appears at the end of both, doubt has been expressed about their authorship, and it has been suggested that the two works are perhaps by different authors. ⁹³² There is no reference in the Khuddakapāṭha commentary ⁹³³ to an earlier Sihala commentary, but there are references to former teachers (pubbācariyas) and teachers of the commentary (atthakathācariyas) as well as porānas and theriyas. ⁹³⁴ The author shows some acquaintance with Sanskrit and the Sanskrit grammarians, ⁹³⁵ which would be in keeping with the traditional view that Buddhaghosa was a brahman before he became a Buddhist, although it is easy to show from other commentaries attributed to Buddhaghosa that at times he showed ignorance of Sanskrit. ⁹³⁶ It seems on balance, however, that there are no very compelling reasons for believing that the Khuddakapātha commentary is not by the same author as the Visuddhimagga. ⁹³⁷

The commentary upon the Suttanipāta ⁹³⁸ also shows some knowledge of Sanskrit grammar, but no more than Buddhaghosa shows in the Visuddhimagga, ⁹³⁹ and this seems to be no reason for denying his authorship. The repetition which is found with the commentary upon the Khuddakapātha, e.g. the full commentaries in both texts upon the three suttas which the Khuddakapātha and the Suttanipāta have in common, ⁹⁴⁰ is certainly strange when his habit of referring from one commentary to another, as already noted in the case of his Sutta-piṭaka and Abhidhamma-piṭaka commentaries, is considered, but the fact that these repetitions are word for word the same can hardly be taken to prove that they are by different authors.

The Mahāvamsa states that Buddhaghosa went to Ceylon in the reign of king Mahānāma. ⁹⁴¹ While there is some doubt about the precise dating of Sinhalese kings, it is generally accepted that Mahānāma lived in the first half of the fifth century A.D. ⁹⁴² If the letter which is reported to have been sent to the Chinese emperor in A.D. 428 by Ma-ho-nan ⁹⁴³ was sent by the same Mahānāma, then the date is confirmed. At the end of the Samantapāsādikā Buddhaghosa states that he wrote the commentary between the 20th and 21st years of king Sirinivāsa. ⁹⁴⁴ It is assumed that the names Mahānāma and Sirinivāsa refer to the same person. If this is the same king who is called Sirikūta at the end of the commentary upon the Dhammapada, ⁹⁴⁵ then that commentary, whether or not by Buddhaghosa, can be dated to the same period. A lower limit to the possible date for Buddhaghosa is given by the fact that the Chinese translation of the Samantapāsādikā was made in A.D. 489.⁹⁴⁶

It is worth nothing that, although we are informed by Buddhaghosa himself that he translated his commentaries from the Sīhala language (and from Dravidian too, if we are wrong to reject this suggestion), there is no trace of this translation process to be seen. Although attention has been drawn to a few passages in Buddhaghosa’s commentaries where nominative singular forms in -e, which is a characteristic of early Sinhalese Prakrit, ⁹⁴⁷ are to be found, a close examination shows that these forms are rather to be regarded as Māgadhisms, or at least traces of some North Indian Prakrit. ⁹⁴⁸ There seems to be no evidence for the existence of untranslated Sinhalese Prakrit forms in any of Buddhaghosa’s commentaries.

3.2. Buddhadatta

There is a tradition, recorded in the Vinayasāratthadīpanī, [^3-224] a twelfth century commentary on the Vinayavinicchaya written by Vācissara, that Buddhadatta made summaries of Buddhaghosa’s works after the latter had translated them from the Sinhalese Prakrit atthakathās. The later Buddhaghosuppatti tells a different story. ⁹⁴⁹

The Gandhavaṃsa states ⁹⁵⁰ that Buddhadatta wrote the Vinayavinicchaya, the Uttaravinicchaya, the Abhidhammāvatāra and a commentary upon the Buddhavaṃsa entitled Madhuratthavilāsinī. The Rūpārūpavibhāga is also ascribed to Buddhadatta by the Sāsanavaṃsa, ⁹⁵¹ although the Gandhavaṃsa states that this was composed by Vācissara. ⁹⁵² Of these texts, the Vinayavinicchaya and the Uttaravinicchaya ⁹⁵³ are summaries of the Vinaya-piṭaka, and the Abhidhammāvatāra and Rūpārūpavibhāga ⁹⁵⁴ are summaries of the Abhidhamma. The reason for making such summaries would seem to be simply the need to make it easier for students to learn and remember, since verse is much more convenient than prose for such purposes.

Although there are verbal similarities with some of Buddhaghosa’s works, in particular the Atthasālinī, ⁹⁵⁵ it is by no means certain that Buddhadatta did, in fact, summarize Buddhaghosa’s works. He nowhere mentions Buddhaghosa’s name, but states that he wrote the Vinayavinicchaya at the request of his pupil Buddhasīha, ⁹⁵⁶ the Uttaravinicchaya at the request of Mahāthera Sañghapāla ⁹⁵⁷ (perhaps the same elder at whose request Buddhaghosa wrote the Visuddhimagga), and the Abhidhammāvatāra at the request of Sumati. ⁹⁵⁸ It is possible that the similarities between the two authors’ works led to the belief that one was dependent upon the other, and it is also likely that the name of Buddhaghosa was connected with that of Buddhadatta at a later date to give added authority to his works. ⁹⁵⁹

The similarities can, however, probably be explained by assuming that both authors based their works upon the same atthakathās in the Mahāvihāra. The fact that in his summaries of the Vinaya-piṭaka Buddhadatta frequently refers to the Kurundaṭṭhakathā, the Kurundī, the Mahā-aṭṭhakathā, and the Mahāpaccarī, ⁹⁶⁰ which were also used by Buddhaghosa, suggests that both authors referred to the same original Sīhala commentaries.

In the colophon to the Vinayavinicchaya ⁹⁶¹ Buddhadatta states that the text was composed during the reign of Acyutavikrama of the Kalamba (or Kalabbha or Kalambha) clan, said to be king of the Cola country. ⁹⁶² If, at some future date, it proves possible to give a firm date to this king on archaeological grounds, then we should be afforded a firm basis for the chronology of both Buddhadatta and Buddhaghosa and the contemporary kings of Ceylon.

An examination of the commentary to the Buddhavamsa, ⁹⁶³ which is also ascribed to Buddhadatta, shows that it must have been compiled at a later date, presumably by another Buddhadatta.

3.3. Mahānāma

There is a commentary ⁹⁶⁴ upon the Paṭisambhidāmagga ascribed to Mahānāma, ⁹⁶⁵ and entitled Saddhammappakāsinī. The Gandhavaṃsa states that it was written at the request of the upāsaka Mahānāma. ⁹⁶⁶ The colophon to the text says that it was written in the Mahāvihāra in the third year after the death of king Moggallāna, ⁹⁶⁷ who is generally dated to the first half of the sixth century A.D. ⁹⁶⁸

There is no evidence that a Sihala atthakathā upon the Paṭisambhidāmagga existed, which perhaps explains why Buddhaghosa did not write a commentary upon it. Buddhaghosa, however, quotes extensively from the Paṭisambhidāmagga in the Visuddhimagga, ⁹⁶⁹ and in turn Mahānāma borrows extensively from the Visuddhimagga in his explanations of the Paṭisambhidāmagga. ⁹⁷⁰ Buddhaghosa’s commentaries upon the four nikāyas all refer to the Visuddhimagga in their opening verses, implying that study of the latter was essential for understanding the nikāyas. It has been suggested that Mahānāma was doing much the same for the Khuddaka-nikāya, and by quoting where necessary from the Visuddhimagga, instead of merely referring to it, he was providing a basic commentary for the Khuddaka-bhānakas to refer to. ⁹⁷¹

The commentary begins with the statement that Sāriputta explained the Dhammacakkapavattanasutta by composing the Paṭisambhidāmagga, ⁹⁷² thus indicating the main theme of the attainment of enlightenment by understanding the four truths. It continues by seeking to present the Paṭisambhidāmagga as a systematic and orderly exposition of the way to arahant-ship, with each topic leading necessarily to the next. This is done by supplying lengthy expositions in the commentary to smooth out the rough statements and extend them with elaborations.

3.4. Upasena

There is a commentary upon the Niddesa entitled Saddhammappajotikā, ⁹⁷³ although the Gandhavaṃsa also calls it Saddhammaṭthitikā. ⁹⁷⁴ It is ascribed to Upasena, both in the Gandhavaṃsa ⁹⁷⁵ and in the colophon to the commentary itself. ⁹⁷⁶ It is said to have been written at the request of a thera named Deva. ⁹⁷⁷ Upasena himself states that he wrote the work at Anurādhapura in the 26 th year of the reign of king Siri-Sanghabodhi, ⁹⁷⁸ whose dates are disputed. He probably lived about the middle of the sixth century A.D. ⁹⁷⁹

Included in the commentary upon the Niddesa is comment where necessary on the relevant verses of the Suttanipāta about which the Niddesa comments. For this purpose the author has sometimes borrowed passages, without alteration, from the Paramatthajotikā. ⁹⁸⁰ It has been pointed out that he sometimes does this very carelessly, borrowing the later part of a story, without mentioning its beginning, ⁹⁸¹ so that it is unintelligible without reference to the Suttanipāta commentary.

Other explanatory material, e.g. on the pațiccasamuppāda and the noble truths, is borrowed from the Visuddhimagga and the Sammohavinodanī. Such wholesale borrowing from Buddhaghosa’s commentaries provides an upper limit for Upasena’s date.

3.5. Dhammapāla

The Gandhavaṃsa states ⁹⁸² that Dhammapāla wrote commentaries upon the Nettippakaraṇa, the Itivuttaka, the Udāna, the Cariyāpiṭaka, the Theragāthā, the Vimānavatthu and Petavatthu, and a tīkā upon the Visuddhimagga, as well as a number of other tīkās which are usually assumed to be by another Dhammapāla. ⁹⁸³

It seems likely that by the time he came to write his commentaries, those on the Khuddakapāṭha, Suttanipāta, Niddesa, and Paṭisambhidāmagga already existed as well as the Dhammapada-atthakathā and the Jātaka-atṭhakathā. He seems therefore to have set himself the task of completing the commentaries upon the Khuddaka-nikāya. His commentaries are called Paramatthadīpanī, ⁹⁸⁴ and comprise atthakathās upon (1) Udāna, (2) Itivuttaka, (3) Petavatthu, (4) Vimānavatthu, (5) Theragāthā, (6) Therīgāthā, and (7) Cariyāpiṭaka. At the end of each part of the Paramatthadīpanī, Dhammapāla states that he wrote it at Badaratitthavihāra, ⁹⁸⁵ which was in South India, after the style of the old commentaries (porānatthakathānayaṃ). ⁹⁸⁶ The commentary upon the Nettippakaraṇa also has a reference to Badaratitthavihāra. ⁹⁸⁷

Dhammapāla states that the exposition he is going to give is in accordance with the views of those dwelling at the Mahāvihāra, ⁹⁸⁸ from which it can be assumed that he visited Ceylon, and received help there with the writing of his commentaries. There is, however, no specific statement that he made use of Sīhala atthakathās in his work, nor does he refer by name to Buddhaghosa even when he quotes his commentaries, ⁹⁸⁹ but there are not infrequently identical explanations of certain words and phrases, ⁹⁹⁰ which suggests that Dhammapāla is either following Buddhaghosa tacitly, or he is dependent upon the same Mahāvihāra tradition which Buddhaghosa followed. It is clear that there were variant readings in the material available to him, for he sometimes states an alternative reading in his commentary, and gives alternative interpretations of the relevant verse. ⁹⁹¹ Sometimes in ambiguous circumstances the tradition was not certain of the correct interpretation, and Dhammapāla accordingly gives two explanations. ⁹⁹²

The style of the two commentaries upon the Petavatthu ⁹⁹³ and the Vimānavatthu ⁹⁹⁴ is similar. Each vatthu in the canonical texts is afforded a separate section. Each section falls into three parts, comprising an introductory story which explains the circumstances in which the verses which follow came to be delivered, then the verses, and finally a commentary upon the verses. Dhammapāla attributes the first two of these sections to the Buddha who, when explaining how it was that the verses came to be uttered, himself repeated the relevant verses. From this point of view, then, only the third section forms the commentary proper. The purpose of the commentary is to clarify the identity of the speakers of the verses, and to explain the meanings of the words which occur in the verses, by way of a series of linguistic equivalents and synonyms. There is little exposition of the doctrine supposedly lying behind the verses, and Dhammapāla is relatively silent on matters of doctrinal importance.

The situation with regard to the commentaries upon the Thera- ⁹⁹⁵ and Therīgāthā ⁹⁹⁶ is not greatly different. Dhammapāla gives an introduction to each set of verses, usually based upon the Apadāna tradition, although it is not always clear how the various Apadānas were linked with the various elders. In the case of the “great” elders, the stories given in the Manorathapūranī and the Dhammapada-atṭhakathā are closely followed, with some compression and omission of marvels. In each case the elder is stated to have attained arahantship, and the appropriate Apadāna is then quoted, with a small number of exceptions. Sometimes, however, the Apadāna insertion is made awkwardly; often it is attributed to the wrong elder; and once a non-existent Apadāna is quoted. It seems probable that Dhammapāla did not, in fact, include them in his commentary, but they have been added during the course of the scribal tradition. ⁹⁹⁷

Although many of the stories which Dhammapāla gives agree with information which we have about the elders from other sources, some of the stories seem to be deduced from the information given in the verses, or are pure invention, sometimes based upon misunderstanding of the verses. ⁹⁹⁸ In some cases he mentions conflicting traditions about the elders. ⁹⁹⁹ It is clear that the text of some of the verses had already been corrupted by the time that Dhammapāla wrote his commentary, and while some of the stories make the meaning of obscure verses clearer, in others it seems that the meaning had already been lost.

Even where no corruption has taken place, it is obvious that Dhammapāla, or the tradition ¹⁰⁰⁰ he was following, did not understand the meaning of some of the words he commented upon. ¹⁰⁰¹ Sometimes he contradicts himself, or gives different explanations of the same verse when it occurs twice, ¹⁰⁰² perhaps by oversight, or perhaps because there was a different tradition of explanation for the two occurrences. On the other hand, some of the explanations seem to be very old, and some of them seem to be based upon forms in dialects other than Pāli, ¹⁰⁰³ which probably means that they were brought from North India centuries earlier.

As in the case of the commentaries upon the Petavatthu and Vimānavatthu, there are sometimes alternative readings and explanations, ¹⁰⁰⁴ which suggests that there were already in the Mahāvihāra tradition collections of commentarial material from various sources with conflicting exegesis.

Since the Cariyāpițaka consists almost entirely of Jātaka stories in verse, the greater part of the commentary ¹⁰⁰⁵ is devoted to the narration of the birth stories in prose. The source for the commentary is several times stated to be the atthakathā or the Jātakaṭṭhakathā, ¹⁰⁰⁶ but whether this means the Jātakatthavaṇ̣anā which we possess today, or the atthakathās upon which that text is based, cannot be determined. The latter part of the commentary, however, consists of an explanation of the Nigamanagāthā of the Cariyāpiṭaka. This enables Dhammapāla to discuss the ten pāramitās at length. He points out that some Buddhists recognise only six pāramitās in place of the ten recognised by the Theravādins. As is well known, most schools except the Theravādins accept the smaller number, but it would seem more likely that ten represents an increase from six rather than that the rest reduced the number from ten to six, as Dhammapāla suggests. ¹⁰⁰⁷

While it has never been doubted that these commentaries, and those upon the Udāna ¹⁰⁰⁸ and the Itivuttaka, ¹⁰⁰⁹ are to be ascribed to Dhammapāla, there has been doubt expressed ¹⁰¹⁰ about the commentary upon the Nettippakaraṇa, ¹⁰¹¹ despite the fact that it too is said to have been composed by Dhammapāla, resident in the Badaratitthavihāra. Nevertheless it has been pointed out ¹⁰¹² that from the point of view of style and language the commentary is very similar to Dhammapāla’s other works. Like Dhammapāla, the author of the commentary upon the Nettippakarana refers to Buddhaghosa’s works without naming him. ¹⁰¹³

Nothing is known about the date of Dhammapāla, except that he must be later than Buddhaghosa. There has not, as yet, been found any trace of interdependence between Dhammapāla and Upasena or Mahānāma, which probably means that all three were writing at about the same time. We shall perhaps not be far out if we assume that Dhammapāla composed his works about the middle of the sixth century A.D.

Chapter IV - LATER POST-CANONICAL TEXTS

1. THE LATER CHRONICLES

The Buddhaghosuppatti states ¹⁰¹⁴ that the commentaries brought by Mahinda and the additions which had been made thereto were burned in a great bonfire when Buddhaghosa had finished his commentaries. It has long been recognised that this is merely an exaggerated way of accounting for the fact that the Sihala atthakathās fell into disuse after Buddhaghosa’s time. That they did not disappear immediately can be seen from the fact that they were available to the author of the Mahāvaṃsa, who probably wrote after Buddhaghosa’s time, ¹⁰¹⁵ and to the authors of other chronicles, and also to later commentators.

They were certainly still in existence when an unnamed ¹⁰¹⁶ author wrote a commentary upon the Mahāvaṃsa. ¹⁰¹⁷ It is classified as a tīkā in Burma, where manuscripts of the text were first discovered, but the author’s own name for it is the Vamsatthappakāsinī, ¹⁰¹⁸ and the work is stated to be the Padya-(or Pajja-) padoruvaṃsavaṇnanā. ¹⁰¹⁹ The name Padyapadoruvaṃsa for the Mahāvaṃsa has already been noted. ¹⁰²⁰

The tīkā adds little to the accounts of the councils, the rise of the sects, and the sending out of the missions. This perhaps indicates that the author of the Mahāvaṃsa had already extracted all the information which was available on these subjects from the Sihala atthakathās. The author of the tīkā names a number of sources ¹⁰²¹ for the information he gives, of which the atthakathā, presumably the Sihala atthakathā, is the one most frequently mentioned. He also refers to a Vinayaṭ̣thakathā, a Dīpavaṃsaṭ̣̣hakathā, and a Cetiyavaṃsaṭ̣̣hakathā, and quotes the porānas a number of times. It is noteworthy that he never quotes the Dipavaṃsa by name, ¹⁰²² and when he quotes verses which actually occur in that text he attributes them to the porannas. This perhaps means that he is quoting them from the same source as the author of the Dīpavaṃsa, rather than from the Dīpavaṃsa itself. Very slight variations which occur would perhaps support this suggestion. ¹⁰²³

It is very interesting to note that the tīkā also quotes the Uttaravihāravāsinam atthakathā¹⁰²⁴ as a source, for the Uttaravihāra is also known as the Abhayagirivihāra, ¹⁰²⁵ and the inhabitants of that vihāra were hostile to the Mahāvihārins. This presumably means that at the time of the separation of the two vihāras the Abhayagirivihārins had taken the Sīhala atthakathās with them, and had added to them material not found in the Mahāvihārin tradition. The fact that the author of the tīkā was able to refer to their atthakathā would seem to mean that the division between the two vihāras was not absolute. He does, however, make it clear that he is incorporating only such material as is not in conflict with orthodox tradition. ¹⁰²⁶ From this source he takes legends about Candagutta and Cāṇakka, ¹⁰²⁷ the nine Nandas who succeeded the sons of Kāḷāsoka, ¹⁰²⁸ and the story of Susunāga. ¹⁰²⁹

There seems to be no doubt that Buddhaghosa’s commentaries eventually made all others unnecessary, which led to their ultimate disappearance. Any text which shows acquaintance with the Sīhala atthakathās is unlikely, therefore, to be more than two or three centuries later than Buddhaghosa. The same is also likely to be true of the chronicles, especially since, as already noted, the Mahāvaṃsa seems to have made use of almost all the information available. It has consequently been suggested that the Vamsatthappakāsinī should be dated to the eighth or ninth century A.D. ¹⁰³⁰

The quality of the commentary afforded by the tīkā is good. The author shows an excellent knowledge of his subject, and while respecting his authorities is not afraid to point out errors.¹⁰³¹ He even points out a mistake in the Mahāvaṃsa itself.¹⁰³² He is not afraid to give his own opinion, although he makes it clear that it is his opinion. ¹⁰³³

There exists a Cambodian version of the Mahāvamsa ¹⁰³⁴ which is almost twice its length. ¹⁰³⁵ The redactor has enlarged the text by making a number of additions. ¹⁰³⁶ At the beginning he has inserted a poem of 677 verses about the earlier Buddhas, and a history of Gotama which has been added seems to be a versified version of the Mahāvagga of the Vinaya-piṭaka. Other insertions are passages from the Vamsatthappakāsinī turned into verse. One insertion concerns Asoka’s queen Asandhimittā. ¹⁰³⁷ The story is not known from other Sinhalese sources, and seems to be dependent upon a fund of stories about Asoka which were current in South-East Asia.

The author calls himself Moggallāna, ¹⁰³⁸ and since there are signs ¹⁰³⁹ that the text was copied from an earlier Sinhalese original, he may have been the Moggallāna mentioned in the Mahāvamsa as living in Ceylon in the twelfth century. ¹⁰⁴⁰

As already noted, ¹⁰⁴¹ the Mahāvamsa, in its original form, seems to come to an end halfway through a chapter. Its continuation is traditionally called Cūlavamsa. ¹⁰⁴² The first two verses of the continuation are identical with the last two verses of the Dīpavamsa, ¹⁰⁴³ which presumably represents a conscious effort to link the newer part to the old. The first addition ¹⁰⁴⁴ covers the period from the death of Mahāsena to Parakkamabāhu I (A.D. 1153-86), and is traditionally ascribed to a thera named Dhammakitti. ¹⁰⁴⁵ To the portion attributed to him is added a summarising verse and then a verse in an ornate metre, signifying the end of the chapter. ¹⁰⁴⁶ The second section ¹⁰⁴⁷ covers the period between Vijayabāhu II, the successor to Parakkamabāhu I, and Parakkamabāhu IV (c. A. D. 1300). This too ends with a summarising verse and a verse in an ornate metre. ¹⁰⁴⁸ The third portion ¹⁰⁴⁹ continues to the reign of Kittisirirājasīha (A. D. 1746-81). It is stated ¹⁰⁵⁰ that this king obtained copies of the Mahāvamsa from Siam in order to check the deficient version available in Ceylon, and the section is probably the work of Mahānayaka Tibbatuvava, who was entrusted with the task of comparing the two versions. ¹⁰⁵¹ The style of this portion differs considerably from that of earlier sections, with verses in ornate metres found within the chapters, and not just at the ends. The very last portion of the text as found in the European edition does not seem to occur in the manuscripts, and is taken from a Sinhalese printed edition. ¹⁰⁵² The final chapter is a supplement added by the editors of the Sinhalese edition. ¹⁰⁵³ It carries the story down to the arrival of the English (Inigirisi) in A.D. 1815.

The Mahābodhivamsa ¹⁰⁵⁴ or Bodhivamsa is written mainly in prose, with verses at the end of each chapter, taken from the Mahāvaṃsa, ¹⁰⁵⁵ and also at the end of the work. It begins with the Buddha Dipankara, then tells the story of Gotama and his enlightenment at the foot of the bodhi tree. It then relates the story of the three councils, Mahinda’s mission and the introduction of Buddhism into Ceylon, and the introduction of relics there and of a branch of the bodhi tree, followed by a description of the planting of the tree and the institution of the bodhipūjā ceremonies.

It is said in the introduction to be a translation from Sinhalese into Pāli. ¹⁰⁵⁶ The style of the Pāli is rather artificial and affected, with long compounds and use of Sanskrit, or Sanskrit-derived, words or words in a Sanskritised sense. The work has been described as signifying the beginning of Sanskritised Pāli, ¹⁰⁵⁷ although it has been pointed out ¹⁰⁵⁸ that some of these features can be found in the Middle Indo-Aryan languages before they were adopted into Sanskrit.

The Mahābodhivamsa is greatly dependent upon the Samantapāsādikā, Mahāvaṃsa, and the Jātaka-nidāna (or the althakathā upon which this is based) for its material. The Gandhavaṃsa and the Sāsanavamsa ¹⁰⁵⁹ do not name the author, but tradition ¹⁰⁶⁰ ascribes it to Upatissa, who wrote at the request of Dāṭhānāga. If this is the same person as the Dāṭhānāga named in the Mahāvaṃsa, ¹⁰⁶¹ then the text can be dated to the last quarter of the tenth century A.D.

The Dāṭhāvaṃsa ¹⁰⁶² is also a translation from Sinhalese according to the introduction to the text. ¹⁰⁶³ The work covers the story of early Buddhism from the time of Dīpankara onwards, and proceeds to the distribution of the relics after the Buddha’s parinibbāna. This begins the portion of the story dealing with the tooth relic. Its history is told until it eventually reaches Ceylon in the ninth year of the reign of Sirimeghavaṇna (A.D. 304-32). The work is also known as Dantadhātu. ¹⁰⁶⁴

The text is a work of no small literary merit. It has six chapters. The first is written in the Jagatī metre, the second in Slokas, the third in Tristubhs, the fourth in the Mālinī metre, the fifth and sixth in the Vasantatilakā metre, with the exception of the last verse or two of some of the chapters which are in such metres as Sragdharā, Mandākrānta, S̄ikhariṇī and Śardūlavikrīdita.

The author is named as Dhammakitti, ¹⁰⁶⁵ and he states that he wrote his work at the request of the general Parakkama who placed Līlāvatī, widow of Parakkamabāhu (called Parakkantibhuja), on the throne. ¹⁰⁶⁶ This incident is referred to in the Mahāvaṃsa, ¹⁰⁶⁷ and can be dated to A.D. 1211, so the work must have been composed shortly after this.

The Thūpavaṃsa ¹⁰⁶⁸ states that it was written by the thera Vācissara, ¹⁰⁶⁹ who translated it from Sinhalese because an earlier Thūpavaṃsa in Pāli was not satisfactory. ¹⁰⁷⁰ If this Vācissara is the thera of that name mentioned in the Mahāvaṃsa, ¹⁰⁷¹ then he lived between A.D. 1230 and 1276.

The central theme of the work is the building of the Mahāthūpa by Duṭṭhagāmaṇī, who was ruler of Ceylon 101-77 B.C. The work follows the accepted pattern for such texts by beginning with the resolve to become a Buddha which Gotama made in the time of Dīpankara. It recounts Gotama’s births under successive Buddhas, and then his life and parinibbāna in this era. The history of the relics in India and Ceylon is then related down to the time of the construction of the Mahāthūpa. The text ends with the death of Duṭṭhagāmaṇī.

The material in the Thūpavaṃsa is taken, with only little change, from earlier works, ¹⁰⁷² including the Buddhavaṃsa and its commentary, the Jātakanidāna, the Dīgha-nikāya and its commentary, the Samantapāsādikā, the Dīpavaṃsa and the Mahāvaṃsa with its tīkā. Some material, untraceable in earlier works available to us now, e.g. the story of Asoka obtaining relics for his thūpas from Ajātasattu’s underground thūpa, is presumably taken from atṭhakathās not utilised by other chronicle writers, and perhaps preserved only in South-East Asia.

Other vamsa (chronicle) texts deal with the history of a relic or a place of veneration. The Hatthavanagalla-vihāra-vamsa ¹⁰⁷³ tells the history of the temple of Attanagalla, which marked the spot where Sirisanghabodhi, after abdicating the kingship, gave his head (on which there was a price) to a poor man so that he could gain the money. It tells the history of the vihāra down to the time of Parakkamabāhu II (A.D. 1234-69), but is quoted in the Pūjāvaliya, which was written in A.D. 1266, so its date of composition must lie between the two. ¹⁰⁷⁴

The author is not named, but he states that he wrote his work at the request of Anomadassi. ¹⁰⁷⁵ The work is in eleven chapters and is written in a very ornate style of Pāli. The text is a campu, a kāvya in mixed prose and verse. The metres employed in the verses include Mālinī, Vasantatilakā, and Tristubh.

The continuing popularity of this type of chronicle text is shown by the existence of the Cha-kesa-dhātu-vamsa, ¹⁰⁷⁶ a modern Burmese work, which tells the history of the six hair-relic shrines (thūpas). It is in prose, with a few verses interspersed in it. The name and date of the author are unknown. ¹⁰⁷⁷

Although the Jinakālamālī ¹⁰⁷⁸ is not called a vamsa, it is a work very much in the vamsa tradition. It was written in Thailand by a thera called Ratanapan̄na, and the author states that he wrote it in a year corresponding to A.D. 1516,¹⁰⁷⁹ although it seems clear that events occurring up to A.D. 1527 are included. ¹⁰⁸⁰ The text is a study of the epochs of Buddhism, as it originated in India, spread to Ceylon, and thence to South-East Asia.

It falls into six parts: events up to the time of the parinibbāna of Gotama Buddha; the three councils; the history of Buddhism in Ceylon; the history of Haripuñjaya (Lampoon) until A.D. 1292; an account of the Lānnā kingdom, and the introduction of the Sinhalese form of Buddhism into Thailand; the spread and influence of this, an account of recent events known to the author, and a eulogy of King Tilok and his great-grandson. It has been pointed out that the later chapters which are devoted to the exaltation of the royal family are more a praśasti than a vamsa text. ¹⁰⁸¹

The sources of this work include Pāli canonical and non-canonical texts, some written in Ceylon as late as the thirteenth century. ¹⁰⁸² The text is important for the information it gives about the relationship between Ceylon and Thailand in medieval times. ¹⁰⁸³ The information it gives about religious buildings and shrines in the Lānnā and neighbouring kingdoms seems to be verified by archaeological evidence, ¹⁰⁸⁴ which suggests that the work is also trustworthy when it deals with less verifiable events, such as the council which, as already noted, ¹⁰⁸⁵ is reported to have been held in A.D. 1475-77 to cleanse the tipitaka of scribal errors. ¹⁰⁸⁶

The language in which the Jinakālamālī is written shows some interesting features, e.g. the unhistorical gemination of consonants and the converse, unusual retroflexion of dentals, and unusual spellings. These were regarded as errors and removed from the European edition, ¹⁰⁸⁷ but it is possible that they are genuine characteristics of Pāli as it was spoken in Northern Thailand in the sixteenth century, or actual features of the writing system used by the Thais when they gave up using the Sinhalese script for writing Pāli. ¹⁰⁸⁸

Another text dealing with the history of Thailand is the Sangitivamsa, ¹⁰⁸⁹ of which one chapter has been published in a European edition. It is attributed to Vimaladhamma, and was composed in the eighteenth century. ¹⁰⁹⁰

The Sandesakathā ¹⁰⁹¹ is another late historical text, written mainly in prose with a few verses inserted. It gives information about the writing of a number of texts, including Anuruddha’s Abhidhammatthasangaha, ¹⁰⁹² and throws some light on the relationship between Ceylon and Burma in medieval times. The editor gives no information about it, but it is clearly modern since it includes a date 2345 years after the parinibbāna of the Buddha, ¹⁰⁹³ which is equivalent to A.D. 1801 .

Another text which it is appropriate to deal with here, although it is a historical romance rather than a chronicle, is the Buddhaghosuppatti. ¹⁰⁹⁴ It was written by Mahāmangala, who lived in the fourteenth century in Burma, if he is the same person as the grammarian Mangala. ¹⁰⁹⁵ Although he states that he has made use of material left by the former teachers, ¹⁰⁹⁶ much of the information he gives differs from that available to us from other sources, as the author of the Sāsanavamsa has pointed out. ¹⁰⁹⁷ The Mahāvaṃsa ¹⁰⁹⁸ informs us that Buddhaghosa translated the commentaries into Māgadhī, i.e. Pāli, but the Buddhaghosuppatti states ¹⁰⁹⁹ that he translated the scriptures themselves. It has already been noted ¹¹⁰⁰ that the statement that Mahinda’s commentaries were burned when Buddhaghosa completed his task is clearly falsified by the fact that those commentaries were available to authors writing centuries later than Buddhaghosa.

2. THE LATER COMMENTARIES

2.1. Atṭhakathās

The task of writing commentaries upon the canonical texts for which none had as yet been made continued in the centuries after Dhammapāla. The fact that Dhammapāla wrote a commentary upon the Cariyāpiṭaka but not the Buddhavamsa suggests that there was already a commentary upon the latter text in existence, but it is strange that he never refers to it. ¹¹⁰¹ As has been noted, ¹¹⁰² the Gandhavamsa attributes such a commentary to Buddhadatta. A commentary ascribed to Buddhadatta exists, ¹¹⁰³ but although it states that it was written at the request of Buddhasīha, ¹¹⁰⁴ as was the Vinaya-vinicchaya, ¹¹⁰⁵ at Kāvīrapaṭṭana, ¹¹⁰⁶ which is where the Abhidhammāvatāra was written, ¹¹⁰⁷ there are doubts about this attribution.

First, unlike the other commentaries attributed to Buddhadatta, the Madhuratthavilāsinī includes a reference to one of Buddhaghosa’s works, ¹¹⁰⁸ without naming the author. More important is the fact that it states ¹¹⁰⁹ that the story about Kanaka’s rebirth as a devaputta is to be taken from the commentary upon the Vimānavatthu named Vimalatthavilāsinī. The story is to be found in Dhammapāla’s commentary upon the Vimānavatthu ¹¹¹⁰ but, as already noted, ¹¹¹¹ this is entitled Paramatthadīpanī. The Gandhavaṃsa, however, calls the same commentary Vimalavilāsinī, ¹¹¹² and it seems very likely that the reference is indeed to Dhammapāla’s commentary. If this is so, then it is clear that the Madhuratthavilāsinī is not by the Buddhadatta who was a contemporary of Buddhaghosa. An alternative explanation is that there was an early commentary upon the Vimānavatthu, entitled Vimalatthavilāsinī, from which both Buddhadatta and Dhammapāla drew their material. This would have become redundant after Dhammapāla wrote his Paramatthadīpanī, and then disappeared.

Although this second suggestion is quite possible, the style of the four summaries attributed to Buddhadatta is so different from that of the commentary upon the Buddhavamsa that, although this may be due in part to the difference in character of the works, it seems quite reasonable to suppose that there were, in fact, two Buddhadattas.

The last canonical text to gain a commentary was the Apadāna. ¹¹¹³ Although the Gandhavaṃsa ascribes a commentary upon this text to Buddhaghosa, ¹¹¹⁴ the fact that its stories are sometimes borrowed from the commentary upon the Theragāthā, and reference is specifically made to this on one occasion, ¹¹¹⁵ shows that the work is later than Dhammapāla. No author’s name is given either at the beginning or at the end.

Besides the commentary upon the Theragāthā, material is also taken from the Dhammapada-atṭhakathā. The Nidānakathā at the beginning is based upon the Jātakanidāna and the corresponding portions of the Buddhavamsaattthakathā. The stories in the commentary upon the Buddhāpadāna are taken from the Dhammapada-atṭhakathā, the Jātakatthavaṇṇanā, and the Cullavagga of the Vinaya-pitaka. The greater part of the comment upon the Paccekabuddhāpadāna is taken from the commentary upon the Khaggavisānasutta of the Suttanipāta found in the Paramatthajotikā. ¹¹¹⁶

Although the text comments upon the Buddhāpadāna, the Paccekabuddhāpadāna and the Therāpādāna portions, there is no comment upon the Therīapādāna portion, and no commentary seems to exist upon that section of the canon. ¹¹¹⁷

When a full commentary is given to an Apadāna, there is an introductory story, and the verses are then commented upon separately. The usual method of commenting ¹¹¹⁸ is to explain a word with a (later) synonym, e.g. absolutive forms in -tvāna are explained by -tvā forms. Some of these explanations are repeated over and over again, but as the commentary continues, only words which seem obscure are explained. After the commentary upon Apadāna No. 181, the stories are not repeated but the author merely states that they are the same as those already given, with some differences of merit done, the locality, names etc. Thereafter stories are only related in the case of well-known theras, or when there is some special significance.

The suggestion that the commentary is late ¹¹¹⁹ is confirmed by the fact that the author shows some knowledge of Sanskrit and Sanskrit grammarians and grammatical terms, e.g. veyyākaraṇa ti Canda-Pāṇinīya-Kalāpādibyākaraṇe chekā.¹¹²⁰ He mentions the Kaccāyanappakaraṇa ¹¹²¹ and quotes a sutta from the Kaccāyana-vutti. ¹¹²² He also includes a Pāli version of a Sanskrit stanza setting out five rules of nirutti. ¹¹²³

There is a commentary upon the mātikās of the seven texts of the Abhidham-ma-piṭaka entitled Mohavicchedanī. ¹¹²⁴ Despite the fact that it is called a mātikatthavannanā, it is not usually regarded as an atthakathā, but as an Abhidhamma work. ¹¹²⁵ It was written by the thera Kassapa who lived in the Cola country in South India between A.D. 1160 and 1230. ¹¹²⁶ He states that he is following the tradition of the Mahāvihāra, ¹¹²⁷ but this probably does not mean that he studied there, but only that he was following the teachings of Buddhaghosa and others. ¹¹²⁸

He does not restrict himself to explaining the mātikās, ¹¹²⁹ which in the case of the Kathāvatthu means a list of the topics treated in that text, but uses the commentary upon the mātikās as a framework upon which to expound the whole of the Abhidhamma system as current in his day. He succeeds in doing this in a very scholarly and comprehensive way.

2.2. Tīkās

A tīkā is a secondary commentary, i.e. a commentary upon a commentary. ¹¹³⁰ The word is, however, also used occasionally of a commentary upon a noncanonical text. ¹¹³¹ The Saddhammasangaha tells ¹¹³² how the tīkās came to be written after a conference held under the presidency of the thera Mahā-Kassapa, with Parakkamabāhu I (A.D. 1153-86) as patron, but there is evidence for the existence of tīkās at a much earlier date than this. In his commentary upon the Udāna Dhammapāla refers to a tīkā upon the Kathāvatthu, ¹¹³³ and since the Sāsanavaṃsa states that Ānanda was the first to write tīkās,¹¹³⁴ this has been understood ¹¹³⁵ as a reference to Ānanda’s Mūla-ṭikā¹¹³⁶ upon the Abhi-dhamma-piṭaka. The fact that certain portions of the same commentary are very similar to passages in other tīkās may mean that they existed in Dhammapāla’s time and he was borrowing material from them, ¹¹³⁷ but the possibility of later writers of tīkās borrowing from Dhammapāla, or the sources he was following, cannot be ruled out. ¹¹³⁸

The Gandhavaṃsa states ¹¹³⁹ that Dhammapāla was the author of the tīkās upon the Visuddhimagga, the atthakathās upon the Dīgha-nikāya and three other nikāyas, the atthakathās on the Jātaka, the Nettippakaraṇa, and the Buddhavamssa. The Sāsanavaṃsa states ¹¹⁴⁰ that he was responsible for the tīkās upon the Visuddhimagga, and on the atthakathās upon the Dīgha-, Majjhima-, and Saṃyutta-nikāyas, and ascribes the other tīkās to other theras. ¹¹⁴¹ Scholars have long recognised that there are difficulties in assuming that the Dhammapāla who wrote the tīkās was the same as the author of the atthakathās, particularly as, in the case of the Netti, this would involve him in writing a tīkā upon his own atthakathā.¹¹⁴² The colophon to the tīkā upon the Visuddhimagga states that it was written at the request of the thera Dāṭhānāga who was residing at the Siddhagāma parivena. ¹¹⁴³ Since the Mahāvaṃsa states ¹¹⁴⁴ that this parivena was built by Sena IV, who ruled A.D. 973-76, and refers ¹¹⁴⁵ to a thera called Dāṭhānāga preaching the Abhidhamma during the reign of Sena’s successor Mahinda IV, it would seem very likely that this tīkā was written towards the end of the tenth century A.D. Since the Gandhavaṃsa lists four authors with the name Dhammapāla, ¹¹⁴⁶ it seems very probable that the atthakathās and this tīkā are by different Dhammapālas. There seems to be no good reason for identifying the tīkā-author with the Dharmapāla who lived at Nālandā in the seventh century A.D., and is mentioned by the Chinese pilgrims. ¹¹⁴⁷

In the absence of any evidence to the contrary, it seems reasonable to assume that the Gandhavaṃsa is correct in stating that the tīkās to the Visuddhimagga and the four nikāyas are by the same person. An examination of the tīkā upon the Dīgha-nikāya-atṭ̣hakathā ¹¹⁴⁸ shows that the author was acquainted with Sanskrit and the Sanskrit grammarians, and knew something about the main systems of Indian philosophy. He quotes Bhartṛhari’s Vākyapadīya and the Amarakośa, and seems especially interested in grammar and syntax. ¹¹⁴⁹

It is particularly interesting to note that the author of this tīkā sometimes identifies the holders of views attributed to “some” (keci) by Buddhaghosa in the Sumañgalavilāsinī. These are occasionally said to be the Uttaravihāravāsins or the Abhayagirivāsins, ¹¹⁵⁰ although the difference between the two is not entirely clear, ¹¹⁵¹ since the terms are elsewhere regarded as synonymous. This would seem to indicate that intercourse between the two vihāras, which has already been noted in the discussion about the sources of the Mahāvaṃsa-ṭīkā,¹¹⁵² was still going on.

A tīkā upon the Jātakaṭṭhavaṇṇanā exists, and has been utilised by the editors of the Critical Pāli Dictionary ¹¹⁵³ and by the editor of the European edition of the Kuṇāla-jātaka. It seems never to have been published, even in Burma, and it is consequently not possible to state whether or not it is the tīkā attributed to Dhammapāla in the Gandhavaṃsa. ¹¹⁵⁴

The Gandhavaṃsa mentions a Milindapañhavaṇṇanā, ¹¹⁵⁵ but gives no author’s name. An edition of a text entitled Milinda-ṭīkā,¹¹⁵⁶ also called Madhuratthapakāsinī, has been published. It is based upon a single manuscript in Cambodian characters. According to the colophon it was written by Mahātipiṭaka-Cūḷābhayatthera, ¹¹⁵⁷ and is probably to be dated to A.D. 1474. ¹¹⁵⁸

The text starts with a detailed description of the six-fold way of commenting upon the meaning of a word, and then proceeds to comment upon difficult words and phrases, drawing upon information in the Abhidhānappadīpikā and a number of, as yet unpublished, tīkās.¹¹⁵⁹ Amidst the information it gives is the complete list of the Buddha’s 80 anuvyañjanas (minor characteristics). ¹¹⁶⁰ It deals, however, with only a small number of the problems which occur in the Milindapañha, and much of the text is devoted to tracing the sources of the Jātaka stories mentioned therein. One portion of the tīkā is devoted to a collection of all the verses which occur in the Milindapañha, but since the list does not agree with the contents of the edition as we know it, this raises questions about the date and form of our modern editions. ¹¹⁶¹ In this connection it is to be noted that the author of the tīkā states that the five introductory verses of the Milindapañha, and other sentences in the prologue and epilogue, were composed by Buddhaghosa. ¹¹⁶² This presumably shows the existence of a tradition that the Milindapañha was revised or recast by Buddhaghosa, or during the time of Buddhaghosa, and this may account for some of the interpolations which have already been noted in this work. ¹¹⁶³ The last part of the tīkā is called Saṃkhyāsarūpa. Like the Sañgītisutta of the Dīgha-nikāya (33), it collects together a number of miscellaneous items and presents them in a numerical way.

Besides the early (purāna) tīkās, produced between the seventh and twelfth centuries, commentators continued in the following centuries to compile later (nava) tīkās, and anutīkās in turn upon the tīkās. ¹¹⁶⁴

3. LATER ABHIDHAMMA TEXTS

The study of the Abhidhamma, based at first upon atthakathās and summaries written in India and Ceylon, was continued in a number of manuals attributed to Sinhalese and Burmese authors. One of the most famous of these is the Abhidhammatthasañgaha, ¹¹⁶⁵ which is attributed to Anuruddha, who is said to have been a native of Ceylon. ¹¹⁶⁶ The fact that a paraphrase of this work was compiled by Sāriputta, who lived during the reign of Parakkamabāhu I (A.D. 1153-86), ¹¹⁶⁷ gives a lower limit for its date, and we shall probably not be far wrong if we assume that Anuruddha lived at the end of the eleventh century or the beginning of the twelfth.

The text does not aim to give a summary of the seven books of the Abhi-dhamma-piṭaka, and in fact mentions only one book by name, ¹¹⁶⁸ but covers the whole field of Abhidhamma in a very condensed way, with mnemonic summaries inserted after every section. It begins with an analysis of mind and consciousness and the objects of consciousness in the world of phenomena, and continues with a discussion of the planes of existence and a classification of kamma. It deals with dependent origination (paticca-samuppāda) and causal relations (patthāna-naya), and finishes with a statement about subjects for mental culture (kammatthāna), the path of purity, and the attainment of release.

The same author wrote the Nāmarūpapariccheda, ¹¹⁶⁹ containing 1845 verses in thirteen sections. He states that he is following the tradition of the Mahāvihāra, ¹¹⁷⁰ and he mentions Ceylon ¹¹⁷¹ in the colophon. The work does not aim to be as comprehensive as the Abhidhammatthasañgaha, and as its name suggests, it deals with general subjects such as nāma and rūpa, kamma, mindfulness (anussati), and insight (vipassanā ). The work is composed for the most part in the Sloka metre, with verses in other metres at the end of each section. Some of these verses show an elegance and feeling for style ¹¹⁷² which shows that Anuruddha was quite an accomplished poet.

The Paramatthavinicchaya ¹¹⁷³ is also ascribed to Anuruddha. It contains 1142 Sloka verses arranged in 29 sections, and in the colophon states that it was written in accordance with the views of the Mahāvihāra by someone born in the town called Kāvīra in the state of Kañcipura, while dwelling in the town of Tañja in the Tamba kingdom. Since Kañcipura is in South India, this has led to the suggestion ¹¹⁷⁴ that the author of the Paramatthavinicchaya is not the same Anuruddha who wrote the Abhidhammatthasañgaha and the Nāmarūpapariccheda. The subjects treated include mind, consciousness, form, and nibbāna.

The Saccasañkhepa ¹¹⁷⁵ consists of 387 verses arranged in five chapters. The colophon states that it was written by Dhammapāla while dwelling at the Mahāvihāra at Badaratittha, ¹¹⁷⁶ which doubtless accounts for the Sāsanavamsa ascribing it to the thera Dhammapāla, ¹¹⁷⁷ but the Gandhavaṃsa attributes it to a Culladhammapāla. ¹¹⁷⁸ The Saddhammasangaha states ¹¹⁷⁹ that the author was Ānanda, but this is probably a mistake, arising from the fact that Dhammapāla was Ānanda’s pupil. ¹¹⁸⁰ If this is the Ānanda who wrote the Mūla-ṭikā which is referred to by Dhammapāla in the atthakathā upon the Udāna, ¹¹⁸¹ then Culladhammapāla would, in any case, be approximately contemporary with the commentator Dhammapāla. The five chapters deal with rūpa, sensations (vedanā ), mind, miscellaneous subjects, e.g. pride and sloth, and nibbāna.

The Nāmarūpasamāsa, ¹¹⁸² also known as the Khemappakaraṇa, is a short prose work, with a few verses interspersed in it, and a final section of 28 mnemonic verses. The Gandhavamsa attributes it to the thera Khema of Ceylon. ¹¹⁸³ There is a commentary upon it by Vācissara, ¹¹⁸⁴ written in the twelfth century, which suggests that the work is to be assigned to the tenth or eleventh century. It consists of short descriptions of mind, and kusala and akusala cetasikas. The mnemonic verses include lists of the four elements, etc.

These five works, together with Buddhadatta’s Abhidhammāvatāra and Rūpārūpavibhāga, ¹¹⁸⁵ Kassapa’s Mohavicchedanī, ¹¹⁸⁶ and the Nāmācāradīpikā ¹¹⁸⁷ by Saddhammajotipāla, form a group known in Burma as the Little-finger Manuals (Let-than). ¹¹⁸⁸ Saddhammajotipāla, also known as Chapaṭa, was a native of Pagan in Burma. The Sāsanavamsa does not include the Nāmācāradīpikā among the texts attributed to Saddhammajotipāla, which perhaps supports the suggestion that he introduced it into Burma, rather than being the author. ¹¹⁸⁹ It is, however, reported ¹¹⁹⁰ that the colophon to the work, which consists of 299 verses, contains the information that it was written while its author was residing in an ārāma near the shrine named Tilokanayana-sabbaññudhātuuṇhīsacetiya, to the east of Arimaddana.

4. COLLECTIONS OF EDIFYING TALES

It is not surprising that in a situation where the well-being of the Sangha depended upon the generosity of laymen, a literature evolved which stressed the value of giving (dāna).

The Dasavatthuppakaraṇa ¹¹⁹¹ belongs to this type of literature. The name is shortened from Dasadānavatthuppakaraṇa, by which title it is sometimes known. The longer title better describes its nature, for it consists of 37 stories in mixed prose and verse, each relating the advantages of making offerings to the Sangha, Buddhas, Paccekabuddhas, and thüpas. To the collection are prefixed 16 Tristubh verses setting out the ten dānavatthus, the ten objects suitable as gifts. The stories are arranged in ten groups, each extolling one of the dānavatthus.

All the stories except one can be traced to accounts of dāna found in atṭhakathās to the canon, or in the Mahāvamsa. ¹¹⁹² The one exception is a story referring to Asoka’s queen Asandhimittā, ¹¹⁹³ which probably indicates an origin in South-East Asia, where a collection of stories about Asoka not available to us now seems to have been current in medieval times. Nothing is known about the date of the compilation of the work, except that, as it is mentioned in an inscription of A.D. 1442 found at Pagan, ¹¹⁹⁴ it must be earlier than that date.

The stories are written in a simple narrative style of Pāli, in some cases word for word the same as the sources from which they are borrowed.

The Sīhalavatthuppakarana ¹¹⁹⁵ is another set of edifying stories, aimed at glorifying the act of giving by recounting the benefits gained by donors in the past. The majority of the 77 stories it contains are set in Ceylon, which probably accounts for its title. The first 45 stories (the colophons to the various sections show that a further five are missing) ¹¹⁹⁶ are in mixed prose and verse. The last 32 are very short, almost entirely in prose, and seem to be intended as notes for reciters to use as the basis of further stories.

Some of the stories are not recorded elsewhere, but those which are have parallels with stories found in the Dīpavaṃsa, the Mahāvaṃsa and atthakathōs upon the canon. The resemblances are not close enough to be able to say whether they are direct borrowings from those texts, or whether they are all based upon some third source, e.g. the Sīhala atthakathās. ¹¹⁹⁷ The presence of hitherto unknown material in the Sihalavatthuppakarana probably means that the compiler or compilers had access to materials current in South-East Asia. ¹¹⁹⁸

The Burmese Piṭakattamain states that the work was composed in Ceylon, ¹¹⁹⁹ but does not name the author. In the colophons he is called Dhammanandi or Dhammadinna. His vihāra and town are variously spelled, ¹²⁰⁰ and have been doubtfully identified with a place in Āndhra. ¹²⁰¹ No information is available about the date of composition, but the fact that the work is mentioned at Pagan in the inscription of A.D. 1442 gives a lower limit. ¹²⁰² Some of the characters mentioned in the stories can be dated to early times, but this seems to have no bearing upon the date of compilation, and the dates of second century B.C. and fourth century A.D. which have been suggested for the first and second parts respectively seem quite unjustified. ¹²⁰³

Despite its name, the Sahassavatthuppakarana ¹²⁰⁴ contains 95 stories, and it has been very plausibly suggested that the title means not ” 1000 stories,” but “delightful stories.” ¹²⁰⁵ There is little system in its arrangement of material, and stories about India and about Ceylon are mixed up together. The language shows signs that the work is a translation from Sinhalese, since it abounds in direct translations of Sinhalese idioms and usages. ¹²⁰⁶

Nothing is known certainly about the date of compilation, although it must be earlier than A. D. 1442 since the Pagan inscription of that year mentions the Sahassavatthu. ¹²⁰⁷ If the text is identical with the Sahassavatthaṭthakathā mentioned three times in the Mahāvaṃsa-ṭīkā,¹²⁰⁸ then we should be able to date it earlier than the ninth century, but it seems unlikely that it can be, at least in its present form. The Mahāvaṃsa-ṭīkā states ¹²⁰⁹ that the story about Prince Sāli is to be found in the Sahassavatthaṭthakathā, but at this point in its narrative the Sahassavatthuppakarana gives only a single line, referring readers to the Mahāvaṃsa for the story. ¹²¹⁰ It seems more likely that the text referred to in the Mahāvaṃsa-ṭīkā was a Sīhala aṭthakathā, which is now lost, but upon which the Sahassavatthuppakarana was itself based.

The Rasavāhinī ¹²¹¹ is the fourth of this collection of edifying tales. It is divided into two parts: the first (Jambudīpuppattivatthu) contains 40 stories originating in India, and the second (Sihaladīpuppattivatthu) consists of 63 stories originating in Ceylon. Each part is divided into vaggas, with four in the first part and six in the second. Each vagga contains ten stories, and the tenth vagga is followed by three extra stories and the colophon. Each story concerns some aspect of the layman’s ethics, and the moral of each tale is clearly expressed in the concluding verse, giving (dāna) being the one which is most frequently extolled.

The colophon names the author as the same Vedeha who wrote the Samantakūṭavaṇnanā, ¹²¹² and the work can be dated to the thirteenth or fourteenth century. It is stated in the introduction that it is a revision of an earlier text, composed by the thera Raṭṭhapāla, which was found to be corrupt throughout with repetitions and other defects. ¹²¹³ There is some doubt about this earlier text. The Gandhavaṃsa states that Raṭṭhapāla’s work was entitled Madhurasavāhinī, ⁹²⁶ but no authority is known for this statement. Surprisingly, the Gandhavaṃsa does not mention Vedeha or the Rasavāhinī. Some verses in the Rasavāhinī agree almost exactly with those in the Sahassavatthuppakarana, which has led to the suggestion that Raṭthapāla’s work was the latter, but it seems more likely that Raṭthapāla was the author of the Sahassavatthaṭthakathā mentioned in the Mahāvamsa-ṭikā. If both the Rasavāhinī and the Sahassavatthuppakarana were based upon this atthakathā, it would account for the similarities between them.

Among the other sources named by the Rasavāhinī are the Mahāvamsa and the porānas. ¹²¹⁴ It is noteworthy that the story of Prince Sāli which does not appear in the Sahassavatthuppakarana does occur in the Rasavāhinī. ¹²¹⁵

5. KĀVYA WORKS

As has already been noted, ¹²¹⁶ the Mahāvamsa can in many ways be regarded as a kāvya work, but the beginning of the medieval period of kāvya literature, showing the influence of Sanskrit literature upon Ceylon, can be seen in the Mahābodhivamsa and the Telakaṭāhagāthā. The former, as has been seen, ¹²¹⁷ is assigned to the last quarter of the tenth century A. D. The latter ¹²¹⁸ is of unknown date and authorship, but must be earlier than the Rasavāhinī, where it is apparently referred to. The poem is an expansion of an incident referred to in the Mahāvamsa, ¹²¹⁹ where a thera was killed and thrown into the sea for intriguing with the king’s wife. In the Rasavāhinī the details are somewhat changed, and the thera is punished by being thrown into a cauldron of boiling oil.¹²²⁰ He becomes an arahant and recites 100 verses. The Telakaṭāhagāthā purports to be the verses he uttered, although it contains only 98¹²²¹ stanzas, divided into nine sections, each dealing with a fundamental doctrine of Buddhism, beginning with the three refuges and ending with dependent causation (paticcasamupppāda). The cause of suffering and death can be traced back to ignorance (avijjā). Escape from worldly misery can be obtained by following the noble path of the dhamma.

The Gandhavaṃsa states ¹²²² that the Jinālañkāra ¹²²³ was written by Buddhadatta at the request of the thera Sanghapāla. The same text states that a tīkā on the Jinālañkāra, ¹²²⁴ with the same name, was written by Buddharakkhita, and also mentions a navaṭikā on the Jinālañkāra, written by an unnamed Sinhalese scholar. ¹²²⁵ The Saddhammasañgaha ¹²²⁶ ascribes the Jinālañkāra to Buddharakkhita. The colophon to the text, omitted in the European edition, attributes the text to Buddharakkhita and states that it was composed 1700 years after the time of the Buddha. ¹²²⁷ The editor, however, has misunderstood the compound sattarasasate, and by giving it the meaning ” 117 ” instead of ” 1700 ” has dated the author’s birth to 426 B.C., ¹²²⁸ instead of dating the composition of the text to A.D. 1156.¹²²⁹ The statement in the colophon is quoted in the Jinakālamālī, ¹²³⁰ where it is said to occur in the Jinālañkāravaṇḍanā, and this suggests a solution to the problem. It is likely that Buddhadatta, whether the author of the summaries of the Vinaya and the Abhidhamma, or the author of the Madhuratthavilāsinī, or another Buddhadatta, wrote a work entitled Jinālañkāra. The work we possess is an embellishment (vaṇnanā) of this, written by Buddharakkhita in the twelfth century. ¹²³¹

Even if the colophon had not included a date, it would have been possible to deduce that the Jinālañkāra is a late work, for it is written in a very developed kāvya style. ¹²³² Some verses are palindromes, ¹²³³ while others show rhyme. ¹²³⁴ One verse has four identical pādas, ¹²³⁵ each to be interpreted in a different way; another verse begins and ends each pāda with the syllables rav(bh)e rave. ¹²³⁶ Other verses consist entirely of velar sounds, ¹²³⁷ or the consonant n.¹²³⁸ Such tricks are reminiscent of Daṇdin’s achievement of omitting all labial sounds in the seventh chapter of the Daśakumāracarita because the narrator, having been bitten on the lip, was unable to pronounce them. ¹²³⁹

The poem, without the colophon, consists of 250 verses in a number of metres, some of them highly ornate. ¹²⁴⁰ Divided into 30 sections, it tells the story of the Buddha from his first aspiration at the time of the Buddha Dīpankara to his parinibbāna. It ends with an exhortation to worship the Buddha’s relics as a means of attaining nibbāna, and a hope that the author will become a Buddha at some future time.

The Jinacarita ¹²⁴¹ is also a poem on the life of the Buddha. The colophon to the text ascribes it to the thera Medhankara, ¹²⁴² and this is confirmed by the Gandhavamna ¹²⁴³ and the Saddhammasangaha. ¹²⁴⁴ It is said to have been written in the parivena founded by Vijayabāhu ¹²⁴⁵ (probably Vijayabāhu III, who reigned A.D. 1232-36), and Medhankara is believed to have flourished under Bhuvanekabāhu I (A.D. 1273-84). He was also the author of a grammatical work entitled Payogasiddhi. ¹²⁴⁶

The poem consists of 472 verses, and begins with the aspiration made by Gotama in his birth as Sumedha at the time of Dīpankara. It omits the rebirths under the subsequent Buddhas, and referring only briefly to the attainment of the perfections, it proceeds to the birth as Gotama. The description of the contest with Māra and the enlightenment is full of poetic exaggeration, and the poem ends with a recitation of the places which the Buddha frequented during his 45 years as a wandering teacher. Like Buddharakkhita, Medhankara concludes with a wish to be reborn as a Buddha.

Estimations of the value of the work vary. Although it has been dismissed as a very mediocre poem, with clumsy versification, ¹²⁴⁷ others have put Medhankara in the foremost rank among poets when he writes from the depths of his own inspiration ¹²⁴⁸ instead of slavishly imitating other works, especially the Nidānakathā of the Jātaka, which seems to be the main source followed.

The Pajjamadhu ¹²⁴⁹ consists of 104 artistic stanzas in what is described as partly Sanskritised Pāli. ¹²⁵⁰ The first 69 verses describe the beauties of the Buddha’s person, and the remainder are in praise of his wisdom, concluding with a eulogy of the Sangha and nibbāna. The style is laboured and artificial, and some of the verses are intricate puzzles.

The author names himself as Buddhappiya, and states that he is a pupil of Ānanda. ¹²⁵¹ If this is the same Ānanda who was the teacher of Vedehathera, then Buddhappiya and Vedeha were contemporaries and probably lived at the end of the thirteenth or beginning of the fourteenth century. Buddhappiya was also the author of the grammar entitled Rūpasiddhi. ¹²⁵²

Vedeha was the author of a poem entitled Samantakūtavaṇnanā, ¹²⁵³ as well as the Rasavāhinī. ¹²⁵⁴ He lived in Ceylon in the late thirteenth or early fourteenth century, and belonged to the Araññāyatana fraternity of forest-dwelling monks. ¹²⁵⁵ The work is a life-story of the Bodhisatta Siddhattha up to the enlightenment, and thereafter the life of the Buddha up to the point which is the climax of the poem, the imprinting of his footprint on the Samanoḷa rock. It is possible that both Vedeha and Rāhula, at whose request he wrote the work, could actually see the mountain from their cells, if they resided at Puṭabhattasela in the vicinity of Samanola, as tradition records. ¹²⁵⁶

The story of the Buddha’s three visits to Ceylon and the planting of his footprint on Sumanakūṭa is found in the Mahāvaṃsa, ¹²⁵⁷ and this is followed closely for the account of the visits to Ceylon. Other material is based upon the Nidānakathā of the Jātaka, with some borrowings from the Samyutta- and Majjhimanikāyas and the Dhammapāda, with occasional verses being borrowed verbatim. ¹²⁵⁸ With a few exceptions the language is correct and easy-flowing. In the narrative portions, which are mainly in Sloka metre, the style follows that of the epics and ballads. In the more ornate portions there are lengthy compounds, and other kāvya features, e.g. puns, repetitions, alliteration, and assonance. Over half of the text’s 802 verses are Slokas, and nearly a quarter are Tristubhs. Vasantatilakā and Mālinī are next in order of frequency, with a handful of other ornate metres. ¹²⁵⁹

There are a number of examples of passages where the grammar is faulty by the normal standards of Pāli, and here the influence of the author’s own Sinhalese language is clearly to be seen. The work also includes several new words coined from Sanskrit, which indicates the prevalence of Sanskrit learning in Ceylon at this time. ¹²⁶⁰

The Saddhammopāyana ¹²⁶¹ deals in 621 verses, and eight closing verses, with the fundamental doctrines of Buddhism in general and the ethical doctrines in particular. It has 19 chapters and begins with a description of the eight akkhaṇas, the miseries of the ten akusalas, and the sufferings of the petas and animals, and the terrible evils (pāpādinavas), and then deals with the results of meritorious deeds and the benefits of generosity and virtuous conduct, etc., ending with appamāda. In one of the concluding verses, the author expresses the wish that he may himself, one day, become a Buddha.

The text itself says nothing about the author, but states that it is to be sent to a friend called Buddhasoma. ¹²⁶² A commentary upon it, entitled Saddhammo-pāyana-viggaha states that the author is called Abhayagiri Kavicakravarti Ānanda Mahāthera. ¹²⁶³ On the strength of this, the text has been identified as a work of the Abhayagirivāsins, ¹²⁶⁴ but there seems to be no evidence for this other than the author’s name. There is nothing in the contents which seems to depart in any way from standard Theravādin doctrine. ¹²⁶⁵

The language is simple and straightforward, and with very few exceptions the metre is Sloka throughout. The work probably dates from the thirteenth century. ¹²⁶⁶

The Pañcagatidīpana ¹²⁶⁷ is a poem in 114 verses, which describes the five conditions in which beings may be born: gods, men, petas, animals and hell-dwellers. It describes the actions which may lead to rebirth in these various categories, and goes into details about the various hells and the tortures to be suffered there. Nothing is known about the author or the date of its composition, but it has been regarded as a product of the fourteenth century. ¹²⁶⁸ The greater part of the text is based upon descriptions such as are found in the commentary upon the Petavatthu, and little can be considered as new or original. What literary merit it has has been attributed to its simple style of diction. ¹²⁶⁹

6. PROPHETIC TEXTS

In a prophecy about the future made by the Buddha in the Cakkavattisīhanādasutta of the Dīgha-nikāya (26), ¹²⁷⁰ there is mention of increased length of life, greater prosperity, a cakkavattin called Sankha and a future Buddha called Metteyya. This is the only future Buddha mentioned in the Pāli canon. He is also mentioned at the end of the Buddhavamsa ¹²⁷¹ as the fifth Buddha of this kappa.

Sankha and Metteyya are also dealt with in a poem of 142 verses entitled Anāgatavamsa, ¹²⁷² which the Gandhavaṃsa ¹²⁷³ ascribes to the Kassapa who was the author of the Mohavicchedani. ¹²⁷⁴ The poem begins with an account of how Sāriputta approached the Buddha and asked about the future Buddha, whereupon Gotama told him when and where Metteyya would be born. The European edition of this poem is accompanied by information about other recensions of the poem, some of which have prose interludes inserted. One of these gives a picture of the gradual decline of Buddhism, ¹²⁷⁵ with the scriptures disappearing piecemeal, beginning with the Paṭthāna, ¹²⁷⁶ the last book of the Abhidhamma-piṭaka. This version ends with a list of ten persons who will be future Buddhas, beginning with Metteyya. ¹²⁷⁷ It is interesting to note that in the Mohavicchedani Kassapa includes a short statement about the disappearance of the canon, ¹²⁷⁸ beginning with the Paṭthāna. ¹²⁷⁹

The Dasabodhisattuppattikathā ¹²⁸⁰ tells of the arising of ten future Buddhas, although the details do not agree entirely with the list just mentioned at the end of the Anāgatavamsa. The story of Metteyya is told in detail, and the stories of the other nine are then presented in a shorter form. A number of the stories have incidents involving bodily sacrifice, self-cremation, self-mutilation, and the offering of children. ¹²⁸¹ Although such deeds are very rarely recorded in Theravādin literature, they seem to be more common in Mahāyāna works. ¹²⁸² This has led to the suggestion that the Dasabodhisattuppattikathā, for which no author or date of composition is known, is probably a product of the late fourteenth century, ¹²⁸³ when Hindu and Mahāyāna influence in Ceylon was strong. The growth in the number of future Buddhas probably represents a result of the same influence, although except for Metteyya the ten Sinhalese Buddhas have nothing in common with their Mahāyāna counterparts. ¹²⁸⁴

Another text of the same type is the Dasabodhisattuddesa. ¹²⁸⁵ The author and the date of composition are unknown. There is a commentary upon the Anāgatavamsa called Samantabhaddikā, of which some extracts have been published. ¹²⁸⁶ If this is the atthakathā mentioned in the Gandhavaṃsa, ¹²⁸⁷ then it was written by Upatissa of Ceylon, possibly the author of the Mahābodhivaṃsa. ¹²⁸⁸ A short text, or portion of a longer text, called Dasabodhisattavidhi has also been published. ¹²⁸⁹ Some of the names given in this differ from those found in the Dasabodhisattuppattikathā, but it is clear that it belongs to the same category of text.

7. MEDICAL TEXTS

The beginnings of Pāli texts dealing with medicine can be seen in the Bhesajjakkhandaka ¹²⁹⁰ of the Vinaya-pitaka, where various rules promulgated by the Buddha about proper diet, hygiene, and medical treatment including the use of herbs and surgical instruments are collected together. Although the Buddha forbade bhikkhus to practise medicine as a livelihood, ¹²⁹¹ it is clear from the respect shown to physicians such as Jīvaka-Komārabhacca that medical knowledge was highly thought of, and medicine for the sick was one of a bhikkhu’s four necessities. ¹²⁹²

The chronicles give a certain amount of information about kings providing supplies of medicines to both bhikkhus and lay people, and the account of Asoka doing so ¹²⁹³ is confirmed by his own inscriptions. ¹²⁹⁴ The Mahāvamsa reports ¹²⁹⁵ that Buddhadāsa (late fourth century A.D.) made a summary of all the medical text books, although it is doubtful whether this is the work mentioned in the Pagan inscription of A.D. 1442.¹²⁹⁶ The Mahāvamsa also gives information about Parakkamabāhu I (A.D. 1153-86) and his keen interest in medicine, ¹²⁹⁷ but it was under his namesake Parakkamabāhu II (A.D. 1236-71) that the only extant Pāli medical text was composed, for there is a statement that during the reign of Narindasīha (A.D. 1707-39) Saranamkara ¹²⁹⁸ composed a commentary upon the Bhesajjamañjūsā, which had been written in the time of Parakkamabāhu.

The Mahāvaṃsa does not name the author, and he is referred to only as the Pañca-pariveña-adhipati, but the Yogaratnākara, a Sinhalese work of the late fourteenth century, mentions ¹²⁹⁹ a medical text called Mañjūsā, written by Atthadassi c. A. D. 1267, and this is presumably the Bhesajjamañjūsā. A lower date for its composition is afforded by the fact that it is mentioned in the Pagan inscription of A.D. 1442.¹³⁰⁰ The author’s reason for composing the work was that “those who have devoted themselves to the spiritual life should be spared illness.” ¹³⁰¹

8. GRAMMATICAL AND LEXICOGRAPHICAL WORKS

The grammatical terminology used by Buddhaghosa and Dhammapāla indicates that there was already before their time a fixed grammatical system, although there is not enough evidence to decide whether they were referring to an actual grammar, now lost. There are indications, as has already been seen, ¹³⁰² that Buddhaghosa was acquainted with parts, at least, of Pānini’s grammar.

The earliest Pāli grammar known to us is that of Kaccāyana, ¹³⁰³ who must be later than Buddhaghosa, since the latter does not refer to him. Kaccāyana may have used the Kāsikā, in which case he is later than the seventh century, ¹³⁰⁴ but the evidence for this is disputed. ¹³⁰⁵ Nothing is known about him except his name, and he seems to be unconnected with any other Kaccāyana known to us from Sanskrit or Pāli. His work is known as Kaccāyana-vyākaraṇa or Kaccāyanagandha, ¹³⁰⁶ and has been shown to owe much to Pāṇini and to the Kātantra. ¹³⁰⁷

The earliest commentary extant upon Kaccāyana is Vimalabuddhi’s Nyāsa, or Mukhamattadīpanī. ¹³⁰⁸ This is probably a work of the eleventh century, for it was itself commented upon by Chapaṭa at the end of the twelfth century in a work entitled Suttaniddesa or Nyāsapradīpa. ¹³⁰⁹ Chapaṭa or Saddhammajotipāla has already been mentioned as the author of the Nāmācāradīpikā. ¹³¹⁰ There is an abridgement of Kaccāyana, called Kaccāyana-sāra, which was written by Dhammānanda, according to the Gandhavaṃsa, ¹³¹¹ and a tīkā upon it written by the Burmese grammarian Mahāyasa. ¹³¹²

The Rūpasiddhi or Padarūpasiddhi ¹³¹³ is a re-arrangement of Kaccāyana’s grammar. The author’s name is given as Buddhappiya, or Dīpankara. ¹³¹⁴ If this is the same person as the author of the Pajjamadhu, ¹³¹⁵ then the text belongs to the second half of the thirteenth century. The Bālāvatāra, ¹³¹⁶ which is much used in Ceylon and Burma, is also a re-arrangement of Kaccāyana, giving the contents in a different, slightly more concise, form. The work is attributed to the same Dhammakitti ¹³¹⁷ who composed the Saddhammasangaha ¹³¹⁸ towards the end of the fourteenth century, although the Gandhavaṃsa ¹³¹⁹ attributes it to Vācissara. It is possible that Dhammakitti was only responsible for composing the Mahāsannaya upon the Bālāvatāra. ¹³²⁰ If Vācissara is the author, then the work is datable to the thirteenth century. ¹³²¹

The greatest of extant Pāli grammars is the Saddanīti, ¹³²² written by Aggavaṃsa from Arimaddana in Burma in A.D. 1154. ¹³²³ A copy of it was taken to Ceylon soon after. ¹³²⁴ On the whole it follows Kaccāyana, but Aggavaṃsa also made use of Pānini. It is said to be based upon the writings of the ācariyas as well as on canonical material. ¹³²⁵ The work consists of 28 chapters, 19 being called Mahāsaddanīti, and the last nine Cullasaddanīti. It is in three sections: the Padamālā (1-14) gives a complete morphology of Pāli, interspersed with nayas from the Buddhavacana, and criticisms of the precepts of Kaccāyana and various Sanskritisms found in Pāli from the works of Buddhaghosa down to the Mahābodhivaṃsa; the Dhātumālā is a Dhātupāṭha in eight ganas (chapters 15-19), with a commentary made up from canonical citations and lexicographical kārikās; the Suttamālā (chapters 20-28) resumes the “complete course” under the traditional form of a sūtrapartha, adding an important chapter on upasargas and nipātas.

The thera Moggallāna or Moggallāyana was the founder of a new school of grammar. He wrote the Moggallāyanavyākaraṇa, which is also known as Saddalakkhaṇa, together with the vutti upon it, and his own commentary upon it, entitled Moggallāyanapañcikā.¹³²⁶ He is reported to have lived in the Thūpārāma vihāra at Anurādhapura, and the colophon to the work states that it was written during the reign of Parakkamabāhu, after the king had purged the Sangha of all heretical bhikkhus, an event datable to A.D. 1165. ¹³²⁷ Moggallāna deals with the linguistic material more exhaustively and with greater understanding of the essence and character of Pāli, which with the influx of both Sanskrit and Sinhalese influence had changed somewhat since the earlier grammarians had written their works. As a result, there are considerable differences in the arrangement and grouping of the rules, as well as in the terminology, and he draws attention to such overlooked matters as the existence of short e and o. Despite his innovations, he draws heavily upon earlier Pāli grammars, and also the Kātantra, Pāṇini, and especially Candragomin. ¹³²⁸

Moggallāna too had his followers, and one of them was the Medhankara ¹³²⁹ who wrote the Jinacarita. ¹³³⁰ His Payogasiddhi bears the same relationship to the Moggallāyanavyākaraṇa as does Buddhappiya’s Rūpasiddhi to Kaccāyana’s Vyākaraṇa, and he makes good some of the deficiencies in Moggallāna’s work to which his opponents had drawn attention.

The writings of the grammarians at his period had a profound effect upon Pāli, and, as has already been noted, ¹³³¹ it has been suggested that the Pāli of the texts which we now possess has been greatly influenced by their theories. It is not overstating the case to say that a knowledge of the Burmese and Sinhalese philology of the period ¹³³² is essential if we wish to go past the recensions of Buddhaghosa and Dhammapāla and return to a Pāli of real linguistic interest. It seems very likely that the manuscripts consulted by Europeans editors go back to originals which have been revised in the spirit of Aggavamsa and his contemporaries. Although the grammarians were well acquainted with Sanskrit, they did not teach that Pāli was simply a derivation of Sanskrit, as the Prakrit grammarians did. ¹³³³ Nevertheless we are probably to see their influence in the occasional Sanskritised form found in Pāli texts.

The two schools of Kaccāyana and Moggallāna both had lists of roots, to match Aggavaṃsa’s Dhātumālā. The Dhātumañjūsā, ¹³³⁴ which is also called Kaccāyana-mañjūsā, ¹³³⁵ belongs to the Kaccāyana school. It is planned on the lines of Vopadeva’s Kavikalpadruma, and was compiled by Sīlavaṃsa about the fourteenth century. The arrangement of root classes follows Kaccāyana, and the author makes use of the Pāṇinian Dhātupāṭha. ¹³³⁶ It is also possible that he made use of the Dhātupāṭha, ¹³³⁷ which belongs to the Moggallāna school. Although its author and date are unknown, it seems probable that it is older than the Dhātumañjūsā. It follows the Moggallāna order of root classes, and is shorter than the other list. It differs from it in being in prose, not verse. ¹³³⁸

The beginnings of Pāli lexicography probably lie in the system of explanation by means of synonyms which we find in the Padabhājanīya section of the Old Commentary upon the Pātimokkha in the Vinaya-piṭaka, and in the Niddesa. ¹³³⁹ This tradition was carried on in the lists of words and glosses known as ganthipadas, which existed early enough to be used by the author of the Mahāvaṃsaṭikā. ¹³⁴⁰ The system was also developed in the Vevacana-hāra sections of the Peṭakopadesa and the Nettippakaraṇa. ¹³⁴¹

The earliest Pāli dictionary we possess is the Abhidhānappadīpikā, ¹³⁴² which seems to have been conceived upon the same lines as the Sanskrit Amarakośa. It is a dictionary of synonyms, and does not follow an alphabetical order. It is ascribed to Moggallāna, but since the Gandhavaṃsa ¹³⁴³ calls the author Navamoggallāna it is assumed that this Moggallāna was not the grammarian. The Abhidhānappadīpikā refers to Parakkamabāhu, which dates it to the late twelfth century, and therefore makes the two Moggallānas roughly contemporary. ¹³⁴⁴

The work consists of 1203 verses, and is in three sections: celestial (sagga), terrestrial (bhumma), and miscellaneous (sāmaññakaṇda). Each section is divided into sub-sections, and there are frequent overlaps between one division and another. The greater part of the work is devoted to synonyms, but the last two divisions deal with homonyms and indeclinables.

Large portions, particularly in the portion on synonyms, are taken in toto from the Amarakośa, and not infrequently the words listed are taken straight from Sanskrit and converted into Pāli according to the normal rules of phonological development. ¹³⁴⁵ A proportion of the vocabulary in the Abhidhānappadīpikā is therefore artificial, in the sense that it had no existence in Pāli until it had been specially coined for inclusion in the dictionary. This fact, however, hardly justifies its exclusion from modern dictionaries ¹³⁴⁶ of Pāli, for subsequent poets mined much material from the Abhidhānappadīpikā, as did their Sanskrit counterparts from the Amarakośa, as a result of which these artificial words acquired a life and existence in their own right. ¹³⁴⁷

A tīkā upon the work was composed about the middle of the fourteenth century, ¹³⁴⁸ and a commentary (samvannanā) was written on it in Burma during the reign of Kittisīhasūra in the same century. ¹³⁴⁹

A Burmese monk named Saddhammakitti wrote a work entitled Ekakkharakosa ¹³⁵⁰ in A. D. 1465. It is a metrical list of monosyllabic words on the lines of similar works in Sanskrit. ¹³⁵¹ His purpose was to ensure that Buddhism survived during the difficult period of persecution in the fifteenth century, and in the belief that survival depended upon correct understanding of the texts his dictionary was intended to help with the proper understanding of Pāli.

9. POETICS AND PROSODY

In the latter half of the twelfth century the thera Sangharakkhita, writing in Ceylon, composed a number of works on grammar, prosody, and rhetoric. ¹³⁵² He speaks of himself as a pupil of Moggallāna, whom he calls Sīlathera. ¹³⁵³

The Subodhālañkāra ¹³⁵⁴ is a work on poetics, and deals with the art of poetry under five heads: faults in composition; their avoidance; verbal ornamentation; rhetorical figures; elegance of sound (rasa) and the art of making verse sound pleasant. It consists of 370 verses, and the author states that his reason for writing ¹³⁵⁵ is that the earlier works by Rāmasamma and others are not suitable for Māgadhī, i.e. Pāli. Sañgharakkhita seems to be dependent upon Sanskrit models, from which the technical terms are borrowed and adapted.

There is a tīkā upon the Subodhālañkāra, ascribed to Vācissara. ¹³⁵⁶

Sañgharakkhita also wrote a book upon Pāli metre, entitled Vuttodaya. ¹³⁵⁷ He states specifically ¹³⁵⁸ that the books by Pingala and others were not suitable for those studying pure Māgadhī, i.e. Pāli, and he accordingly gives rules for writing Pāli verses. The Vuttodaya seems to be the only original work on Pāli metre extant. ¹³⁵⁹ In 136 stanzas, or portions of stanzas, arranged in six chapters, it deals first with general definitions and symbols, and then with all the metres known to Sañgharakkita, including mattāchandas, ganacchandas, and all the varieties of akkharacchandas, although he seems not to have been aware of the Old Āryā metre.

His work is based entirely upon Sanskrit prosody, from which terms are borrowed and adapted. There are occasionally sentences taken over from Pingala with only the minimum change made to adapt Sanskrit into Pāli. ¹³⁶⁰

There were several commentaries written upon the Vuttodaya, including a tīkā¹³⁶¹ by Vepullabuddhi, ¹³⁶² who lived at Pagan. ¹³⁶³

Sañgharakkhita also wrote a work on the Pāli verb and its use in syntax, entitled Sambandhacintā, ¹³⁶⁴ and a tīkā upon the Khuddasikkhā. ¹³⁶⁵

10. BOOKS ON DISCIPLINE

The need for bhikkhus to master the great mass of disciplinary rules contained in the Vinaya-piṭaka led at an early date in the history of Buddhism to the production of summaries designed to make learning easier. The existence of the Parivāra in the Vinaya-piṭaka itself, and of the summaries made by Buddhadatta, has already been noted. ¹³⁶⁶ The same need led to the production of two other texts of the same type.

The Gandhavamsa ¹³⁶⁷ states that the Khuddasikkhā ¹³⁶⁸ was composed by Dhammasiri. This was doubtless taken from the text itself, where Dhammasiri is said to be “a banner to Tambapanṇi” (Tambapaṇniyaketunā). ¹³⁶⁹ No author is named for the Mūlasikkhā ¹³⁷⁰ by the Gandhavaṃsa or the text itself, but a Burmese tradition states that Dhammasiri had a colleague named Mahāsāmi. The same tradition states that the two works were written about 920 years after the Buddha’s death, i.e. c. A.D. 375¹³⁷¹ While some modern scholars have accepted this date, ¹³⁷² others have suggested the sixth or seventh centuries A.D., ¹³⁷³ while one scholar has placed them as late as the eleventh century. ¹³⁷⁴ They cannot be later than that because they are named in an inscription by Parakkamabāhu I (A.D. 1153-86), ¹³⁷⁵ and Sangharakkhita, the author of the Vuttodaya and the Subodhālanikāra, who lived about the same time, ¹³⁷⁶ wrote a tīkā upon the Khuddasikkhā. ¹³⁷⁷

The two texts form a compendium of the Vinaya-piṭaka. They are mostly in verse, with some prose passages. Some passages seem to be parallels to portions of the Samantapāsādikā and the Kankhāvitaranī, rather than the Vinayapiṭaka itself, ¹³⁷⁸ and if this were provable, then a date later than Buddhaghosa would be certain. It is, however, possible that the summaries and Buddhaghosa were both drawing upon the same commentarial tradition, and the same dependence may be the explanation for the passages which do not seem to have parallels in the Vinaya-piṭaka. ¹³⁷⁹ Whether the differences between the summaries and the Vinaya-piṭaka, which have been said to be of considerable interest for the history of the Buddhist Sangha in Ceylon, ¹³⁸⁰ are indicative of an early or a late date of composition is something which still awaits investigation.

The two texts contain some strange forms, which are probably due in part to unskilled authors trying to adapt technical terms of discipline into metre. It seems, however, not to include the artificial features and Sanskritisms which would have proved a late date of composition. ¹³⁸¹

Although it is reported that there are more than 40 Jain śrāvakācāra texts, dealing with the proper conduct of a layman, ¹³⁸² it appears that there is only one systematic Pāli text dealing with this subject. This is the Upāsakajanālanikāra, ¹³⁸³ also known as Upāsakālañkarana or Upāsakālañkāra. ¹³⁸⁴ The author is stated in the colophon to be Ānanda, who was a Sihalācariya, and it is probable that the work was written during the reign of Parakkamabāhu I (A. D. 1153-86). ¹³⁸⁵ This suggestion is supported by the fact that the Pāli employed by the author shows certain Sanskritised features, e.g. some words have been used in a Sanskrit sense and Sanskrit words have been used in their Pāli forms. There are sometimes long compound words in the Sanskrit kāvya style, and the author shows himself to be a skilled poet in the verses he includes. ¹³⁸⁶

The author has based his work on both the canon and the commentaries, and has also made use of later works.¹³⁸⁷ He has surveyed all these works, collected together all the specific and general instructions to laymen which they contain, and then combined them into a single work. The first of its nine chapters deals with the three refuges, and Ānanda ends the section with illustrations from the canon of the benefits which the ti-sarana brings.

Subsequent chapters deal with the rules of morality (sila), ascetic practices (dhutaṇgas, of which only two of the usual list of thirteen are prescribed for a householder), the proper way of gaining one’s livelihood, ten types of meritorious deed and various harmful actions, and happiness in this world and other worlds, leading ultimately to enlightenment. The final chapter deals with the question of the person or individual who performs meritorious deeds, and the Buddha’s teachings about self (attā ) and individual (puggala) are re-iterated. Some of the statements made in this chapter seem to indicate a knowledge of Mahāyāna doctrines, and it seems likely that at the time when the text was written Mahāyāna texts were being studied in Ceylon. ¹³⁸⁸

The author states in the colophon that he wrote his book in accordance with the Mahāvihāra tradition, free from the views of other sects. ¹³⁸⁹ One text which he frequently quotes is the Saddhammopāyana which, as has been noted above, ¹³⁹⁰ is sometimes thought to have been an Abhayagiri text. The fact that Ānanda could quote from this text and at the same time state that he is following the Mahāvihāra tradition shows that there must have been very little fundamental difference between the views of the two vihāras. ¹³⁹¹

11. BOUNDARY TEXTS

In the Uposathakkhandhaka of the Vinaya-piṭaka ¹³⁹² we read of a number of questions which arose early in the history of Buddhism about boundaries between communities, and the Buddha gave various rules for establishing them. The relevant portion of the Samantapāsādikā is entitled Sīmākathā. ¹³⁹³

Although the original purpose of the sīmā was to limit the area of responsibility of a particular Buddhist community, in which they were allowed to perform pabbajjā, upasampadā, and other ceremonies, it is clear that before long the question of sīmā became an intensely legalistic matter, and texts were compiled to serve as the basis for settlement of any disputes which arose. The Mahāvaṃsa-ṭīkā¹³⁹⁴ states that the boundaries of the Mahāvihāra were fixed in the manner set forth in the Sīmākathā, and it is possible that the information found in the Samantapāsādikā is based upon this, or some similar text. ¹³⁹⁵ The Sinhalese manuscripts of the Mahāvaṃsa have an interpolation at this point which describes how the king himself guided the plough which was marking out the boundary. ¹³⁹⁶ It is likely that this passage is taken from a Sīmākathā belonging to the Mahāvihāra. ¹³⁹⁷

The Sīmālañkāra is attributed to Vācissara, ¹³⁹⁸ and a tīkā upon this is ascribed to Chapaṭa, also known as Saddhammajotipāla. ¹³⁹⁹ A summary upon the Sīmālañkāra, called either Sīmāsangaha or Sīmālañkārasañgaha is also attributed to a Vācissara, ¹⁴⁰⁰ who lived in the thirteenth century in Ceylon. At the beginning of his work it is stated that it is so called because it is based upon the Sīmālañkāraka. ¹⁴⁰¹ It consists of 100 verses, including the colophon, in which the author states that he has also written a commentary (vannanā) upon the Sīmālañkāra. ¹⁴⁰²

It is clear that the proper determination and erection of a sanctified boundary to a vihāra came to be considered as an effective means of purification of the Sangha, ¹⁴⁰³ and Vācissara expressed the hope that the unorthodox views of the Coliyan bhikkhus of South India, who apparently had dissentient views on the question of sīmās, would be rejected by all those who desired the well-being of the sāsana. ¹⁴⁰⁴

The Sīmāsañkarachedanī ¹⁴⁰⁵ is ascribed to Śrī Rāhula, who wrote in the fifteenth century, while the fact that the Sīmāvivādavinicchayakathā, ¹⁴⁰⁶ like the Sandesakathā, ¹⁴⁰⁷ includes a date equivalent to the very beginning of the nineteenth century ¹⁴⁰⁸ shows that it is a modern work, although it quotes extensively from canonical texts, atthakathās, and from later texts such as the Vimativinodanītikā which is ascribed to the thera Kassapa. ¹⁴⁰⁹ The continuing importance of the subject is indicated by the production of a work entitled Sīmālakkhaṇadīpanī, by the thera Vimalasāra in Ceylon in the second half of the nineteenth century. ¹⁴¹⁰

12. ANTHOLOGIES

In an edict addressed to the Sangha, ¹⁴¹¹ Asoka recommended to them for frequent study a list of seven suttas, ¹⁴¹² which doubtless represented his own selection of favourite sermons. Others too, no doubt, made selections of the suttas which seemed to them to contain the most important teachings of Buddhism, and certain manuals of this type exist.

The earliest is probably the Suttasangaha, ¹⁴¹³ which is regarded as canonical and forming part of the Khuddaka-nikāya in Burma. Nothing is known about its date or author, although the Burmese tradition states that it was compiled by a thera who lived in Ceylon at Anurādhapura. ¹⁴¹⁴ That it is post-canonical, at least in its present form, is shown by the fact that it includes certain passages from the atthakathās, including the atthakathā upon the Buddhavamsa, and in one case a tīkā upon an atthakathā.¹⁴¹⁵

The suttas are arranged under six headings. The first three consist of suttas concerned with accounts of giving (dāna), moral conduct (sīla), and heaven (sagga). Included in the last-mentioned are the Dhammahadayavibhanga, which gives information about the life-spans of various classes of devas, and several vatthus from the Vimānavatthu. The fourth section includes suttas which tell of the dangers to be found in sensual pleasures (kāmādinava); the fifth section contains suttas about householders or life as a human being; and the sixth section deals with ways of gaining thanks (anumodana). A number of the suttas, particularly those from the Vimānavatthu, are accompanied by their commentaries.

The Sārasangaha ¹⁴¹⁶ contains 40 chapters on various topics of Buddhist teaching, each illustrated by means of stories. ¹⁴¹⁷ The colophon to the text states that the author’s name is Siddhattha, and he is a pupil of Buddhappiya, the chief of the Dakkhipārāma at the end of the thirteenth and beginning of the fourteenth centuries. This was presumably the Buddhappiya who wrote the Rūpasiddhi and Pajjamadhu. ¹⁴¹⁸

The first chapter deals with the aspiration (abhinīhāra) for Buddhahood made by Bodhisattas, the second with the marvels (acchariyas) which occur in a Tathāgata’s life, and the third with the disappearance of the Buddha’s teaching. Many stories deal with the virtues of pious behaviour and generosity. There are also sections about nāgas, supannnas, petas, asuras and devas, about earthquakes, and storms. There is also a chapter on dreams, and the final section deals with cosmology (lokasanthiti).

Although most of the stories which are included can be traced to the canon or the commentaries, some sources have not yet been identified. ¹⁴¹⁹ The method of presentation is so uncritical that it would have been simple for additions to be made to the text during the centuries.

Another anthology, but of a special nature, is the Paritta or Mahāparitta, ¹⁴²⁰ a small collection of texts taken from the Suttapitaka. The word paritta means “protection, safeguard,” e.g. an amulet, ¹⁴²¹ but it is then used of a protective text which can be chanted as a charm or exorcism. There are instances in the canon of the Buddha teaching or allowing the use of such charms. In the Vinayapitaka there is an instance where, after a bhikkhu had died of snake-bite, the Buddha taught a charm to protect beings from snakes, by suffusing them with loving kindness (mettā). ¹⁴²² The Ātānātiyasutta of the Dīgha-nikāya (32) ¹⁴²³ is also a paritta, which was recited in the presence of the Buddha, who by his silence gave consent for its recitation. ¹⁴²⁴

The Milindapañha states that parittas were appointed by the Buddha, ¹⁴²⁵ and six are named there: Ratana, Khandha, Mora, Dhajagga, Āṭānāṭiya, and Añgulimāla. Buddhaghosa mentions four of these in the Papañcasūdanī ¹⁴²⁶ and five of them in the Visuddhimagga. ¹⁴²⁷ The story of the Ratanasutta being used by Ānanda in a great exorcism is told several times in the commentaries, ¹⁴²⁸ and the Mahāvamsa tells of the same paritta being used during a famine and plague during the reign of Upatissa (early fourth century A.D.). ¹⁴²⁹

The addition of the Mangalasutta, the Karanīyamettasutta, and the Bojjhangaparitta gives a set of nine parittas which is commonly found in Ceylon, ¹⁴³⁰ but a longer set, consisting of 28 parittas, and known as the Catubhānavāra, also exists. ¹⁴³¹ An examination of the list of 22 parittas quoted in the Critical Pāli Dictionary ¹⁴³² from the Catubhāṇavāratṭhakathā, which includes both the parittas and the commentaries upon them, shows that the set consists entirely of suttas from the canon, seven being from the Khuddaka-pāṭha, five from the Anguttara-nikāya, six from the Samyutta-nikāya, including the Candasutta ¹⁴³³ which is used as a spell against snakes, two from the Jātaka, and one each from the Majjhima- and Dīgha-nikāyas.

13. COSMOLOGY

As already noted, ¹⁴³⁴ the last section of the Sārasangaha deals with cosmology (lokasavthiti). The oldest extant Pāli text we have which deals exclusively with the subject of cosmology is the Lokapañnatti. ¹⁴³⁵ A lower limit to its date of compilation is provided by the fact that it was one of the sources employed by the author of the Traibhūmikathā, which was composed in Siamese c. A.D. 1345.¹⁴³⁶ It seems probable that the Lokapañnatti is the work of an unknown author, writing in Burma, probably at Thatôn, in the eleventh or twelfth century A.D. ¹⁴³⁷

Although the Northern Buddhists, writing in Sanskrit, produced descriptions of the world, the Theravādins seem to have been too busy escaping from the world to describe it. When they began to do so in medieval times, they were content to borrow from earlier writers on the subject. It is somewhat misleading to call the Lokapaññatti a treatise on cosmology, since it is not a systematic study of the subject, but a compilation of information about the nature of the world, collected from many sources, and put together in a haphazard way. Into the descriptions of the various parts of the world are inserted narrative stories, e.g. the section on the devaloka includes a long story ¹⁴³⁸ about Asoka and robots (bhūta-vāhana-yanta) from the kingdom of Roma, which is doubtless drawn from the stock of stories about Asoka current in South-East Asia, but presumably of North Indian origin.

The greater part of the Lokapañnatti is based upon the Lokaprajñapti, of which the Sanskrit original has been lost but a Chinese translation dating from A.D. 558¹⁴³⁹ is available. Presumably the Sanskrit original was still available in Burma in the twelfth century. It is also possible to deduce that a slightly different version of the Mahāvastu was available at the time when the Lokapaññatti was being compiled, for the text of the latter enables certain lacunae in the former to be filled in the description of the hells, from which borrowings have been made.

A passage on the origin of the world has also been taken from the Mahāvastu, while the Divyāvadāna is the source for some of the other stories, especially details about Asoka. ¹⁴⁴⁰ There are also many borrowings from the Pāli canon, the most important being from the description of the beginning of the world in the Aggaññasutta of the Dīgha-nikāya (27), the judgment of wrong-doers and their condemnation to hell by Yama in the Devadūtasutta of the Añguttaranikāya, and the dialogue between the yakkhas Hemavata and Sātāgira and the Buddha in the Hemavatasutta of the Sutta-nipāta. ¹⁴⁴¹

In A. D. 1520 a Thai mahāthera named Sirimangala, who lived in Chieng Mai, wrote a cosmological work entitled Cakkavāladīpanī, ¹⁴⁴² in six chapters. Like the Lokapaññatti, it is based upon material extracted from earlier texts.

14. WISDOM TEXTS

The learning of worldly wisdom (nīti) by means of collections of aphorisms is of great antiquity in India, and spread from there to all countries which were influenced by Indian culture. Although there is evidence that collections of subhāsitas found their way to Ceylon they seem not to have been imitated by Pāli authors, whereas collections introduced into Burma, probably from North East India, ¹⁴⁴³ were widely imitated there.

There is no mention of any Pāli nitti text in the Pagan inscription of A.D. 1442,¹⁴⁴⁴ which suggests that the genre had either not been introduced by the fifteenth century, or, if it had, it had not yet gained wide acclaim. The longest and probably the earliest of the Pāli nitti texts found in Burma is the Dhammaniti, ¹⁴⁴⁵ which probably dates from the late fourteenth or early fifteenth century. It contains 414 verses, while the later Lokaniti ¹⁴⁴⁶ and Mahārahaniti ¹⁴⁴⁷ contain 167 and 254 verses respectively. These three texts are closely related. The Rājanīti ¹⁴⁴⁸ follows a separate tradition, and could be as late as the eighteenth century.

The niti texts of Burma are not independent works, in the sense of being made up from verses composed in Pāli by Burmese authors. ¹⁴⁴⁹ They are simply compilations, i.e. collections and arrangements of verses which were already in existence in other Pāli texts, or were adaptations of verses available in Sanskrit, especially in the large corpus of Sanskrit verses attributed to Cānakya. ¹⁴⁵⁰ While Pāli parallels to many of the verses in the Dhammaniti and its related texts can be identified, ¹⁴⁵¹ all the verses in the Rājanīti are direct translations from Sanskrit. ¹⁴⁵² The largest Pāli source is the Jātaka collection, while other verses can be traced to the Dhammapada.

The Sutavadḍhananīti or Panditālañkāranīti¹⁴⁵³ was composed in the latter half of the eighteenth century. Its sources are mainly Buddhist canonical texts, with some borrowings from Sanskrit sources. The Pāli Cānakyanīti¹⁴⁵⁴ seems to be an exact translation into Pāli of the Sanskrit Cānakyanītiśāstra. It seems to have been made in the late nineteenth century to provide a Pāli translation for Burmese readers. Another modern compilation is the Gihivinayasangahanīti, ¹⁴⁵⁵ composed in A.D. 1830.

Even the oldest triad of Pāli nīti texts in Burma show a number of forms which differ from classical Pāli. ¹⁴⁵⁶ Many of these arise directly or indirectly from the fact that numerous verses are derived from Sanskrit originals, e.g. unassimilated consonant groups containing -r-,-y-, or -s - are reproduced in their Sanskrit forms on occasion, while the exigencies of metre and grammar have led to difficulties where Sanskrit and Pāli case endings are not identical in metrical length, or the introduction of svarabhakti vowels has increased the number of syllables in a word.

It has been pointed out ¹⁴⁵⁷ that since nīti literature is concerned with worldly wisdom, it is not surprising that the Burmese nīti texts do not contain a great deal of what could be regarded as essentially Buddhist thought. Many of the verses in the Jātaka collection deal with mundane matters, which made their borrowing by the writers of nitti texts all the more appropriate. Specifically Buddhist vocabulary is rare, although the words Buddha, jina and nibbāna do occur a few times.

15. JĀTAKA COLLECTIONS

It has already been noted that the Pāli Jātaka collection is by no means complete, and stories not included in that collection can be found elsewhere in the Pāli canon and also in Buddhist Sanskrit literature. There is also a collection of 50 Jātaka stories current in South-East Asia, generally referred to as “apocryphal” because they are not canonical.

The Paññāsa-jātaka exists in three recensions from Thailand, Cambodia and Burma, ¹⁴⁵⁸ although the last-named differs quite considerably in the selection and order of stories from the first two, which are very similar to each other. ¹⁴⁵⁹ Although these Jātaka stories are regarded as non-canonical, some of them are in fact based closely upon incidents and motifs found in the Pāli collection. Several of them include an episode where, illustrating generosity (dāna), the chief character gives away some valuable possession, including even his wife. Such episodes are clearly based upon the Pāli Vessantarajātaka, from which almost verbatim quotations are given in several stories. ¹⁴⁶⁰ A number of stories include an episode of a ship-wreck, where help is given by the water-sprite Manimekhalā, just as in the Pāli Mahājanakajātaka. ¹⁴⁶¹

Other stories have no Pāli equivalent, but can be compared with various Buddhist Sanskrit texts, showing that the latter were known in South-East Asia in medieval times, when these non-canonical studies may be supposed to have been composed. The Sudhanukumārajātaka has been shown to have a close relationship with stories related in the Mahāvastu and Divyāvadāna, ¹⁴⁶² but the many differences of detail between the three versions show that the Pāli Jātaka is in no way a translation of the other two, which in many ways provide little more than a theme for the Pāli redactor to elaborate. Some stories in the apocryphal collection seem to have no parallel in the extant Pāli or Sanskrit literature, and their origins are presumably to be sought in the mass of folk literature which is known from other sources to have been in existence in the area at the relevant time. Even when a story can be shown to have an external origin, a number of minor details are frequently innovations.

The Burmese recension of the apocryphal Jātakas is known as the Zimmè Paṇnāsa, where the word “Zimmè” is a corruption of the place-name Chieng Mai, ¹⁴⁶³ a town in North Thailand. This suggests that the Burmese believed that the collection came from Thailand, and in view of this belief the difference between the Burmese and the other recensions which has been mentioned above needs to be investigated carefully, for an examination of the parallel editions of the Samuddaghosajātaka shows quite extensive differences of detail, ¹⁴⁶⁴ and it is probable that other Jātakas show variations no less extensive.

The language of these Jātakas ¹⁴⁶⁵ shows the same kind of unhistoric gemination of consonants and the converse as has already been noted in the Jinakālamālī, ¹⁴⁶⁶ which helps to confirm the suggestion that these features are not merely scribal errors. There are also peculiarities of grammar and syntax, e.g. strange case usages and confusion of gender, which probably shows the influence of the local vernaculars.

An original feature of Buddhist literature in South-East Asia is the type of text illustrating advantage (ānisamsā ) which belongs to the Jātaka genre. The Brapaṃsukūlānisamsa ¹⁴⁶⁷ tells of the merits to be gained by the gift of robes to the Sangha, while the Pitakattayakārakānisamsa ¹⁴⁶⁸ contains details of the life of the Buddha, either in his existence as Gotama or in a previous life. One of these episodes concerns the Buddha’s last meal, and shows clearly that in medieval times in Thailand it was believed that the mysterious sūkaramaddava which Cunda gave to the Buddha was indeed pork, for in this version of the story the Buddha’s death was due to the fact that in an earlier existence he had killed a yakkha, who in the course of time was reborn as a pig which, in the form of poisoned meat, served as an instrument to punish the Buddha for his wrongdoing.

It is convenient to deal at this point with another text which, while not strictly a Jātaka story, is nevertheless told in the form of a Jātaka story. The Kosala-bimba-vaṇṇanā ¹⁴⁶⁹ is a medieval work composed in Ceylon, probably during the thirteenth or fourteenth century since it is referred to by name in the fourteenth century Saddhammasangaha. ¹⁴⁷⁰ It is intended to justify the practice of making images of the Buddha, although it is not clear why justification was needed at that time.

The story it tells is based upon an earlier story, preserved for us by the Chinese pilgrim Fa-hsien ¹⁴⁷¹ who heard at Srāvastī at the beginning of the fifth century A.D. the legend that Prasenajit had had a sandalwood image of the Buddha made which, at the arrival of the Buddha, rose to meet him. In the Pāli version of the story Pasenadi asks permission to make an image so that visitors arriving when the Buddha was absent could worship him in effigy. The image is made, and rises to greet the Buddha on arrival. Pasenadi asks the Buddha what merit he has acquired, and Ānanda asks about the merit to be gained by one who copies a Buddhist text. The answer is given that both will be reborn in good families, free from disease, and will live in prosperity until they ultimately gain spiritual distinction.

The quality of the Pāli employed is poor, with mistakes in gender and case usage. The verses included in the prose narrative are of little literary merit, and the exception to this-a kāvya style verse in elaborate metre-is possibly borrowed from elsewhere. ¹⁴⁷²

16. BIBLIOGRAPHICAL WORKS

It has already been noted that the chronicles give a certain amount of information about the composition of literary works, e.g. the recitation of the Kathāvatthu at the time of the third council, and Buddhaghosa’s translation of the Sinhalese atthakathās. ¹⁴⁷³ Other chronicles give much more information about books and authors, and it is not unreasonable to describe such books as bibliographical reference works.

According to the colophon, the Saddhammasangaha ¹⁴⁷⁴ was written by Dhammakitti Mahāsāmi, who came from India to Ceylon, was a pupil of Dhammakitti, and then returned to India and wrote the book. ¹⁴⁷⁵ The teacher Dhammakitti is probably one of the Sangharājas who lived towards the end of the fourteenth century. ¹⁴⁷⁶ The book contains an account of Buddhism in eleven chapters. The first three chapters give an account of the three councils, and include verses from the Samantapāsādikā and Mahāvaṃsa. The narrative then continues with an account of the introduction of Buddhism into Ceylon, and an assembly at which the elder Mahā-ariṭtha preached the vinaya is specifically called the fourth council, ¹⁴⁷⁷ and verses are quoted to this effect, on the authority of the porannas. Subsequent chapters deal with the accession of Duṭṭhagāmanī, the writing down of the tipitaka, Buddhaghosa’s work, and the compilation of the tīkās under the presidency of Mahākassapa during the reign of Parakkamabāhu I (A.D. 1153-86).

The ninth chapter gives an account of the principal works known to the author, of which the latest is datable to the thirteenth century. ¹⁴⁷⁸ They are not listed in chronological order, and there are a number of errors and differences from other sources, ¹⁴⁷⁹ but on the whole the information which the Saddhammasangaha gives is very helpful for establishing the history of the Buddhist religion and Pāli literature.

The last two chapters give an account of the merit to be gained from writing down the tipitaka and listening to the doctrine respectively. In the tenth chapter there is a reference to the statement about the merits to be gained from making statues, which is attributed to the Buddha in the Kosalabimbavaṇnanā. ¹⁴⁸⁰ The last chapter contains a number of illustrative stories on the theme of merit to be gained (ānisamsaā ), some concerning Ceylon. It has been described as a sort of anthology of odds and ends of old verses, of rare excellence and beauty. ¹⁴⁸¹

The Gandhavaṃsa ¹⁴⁸² was written by Nandapañña, who describes himself in the colophon as a forest-dwelling elder (arañ̃navāsinā). ¹⁴⁸³ It was composed in Burma, probably in the seventeenth century. ¹⁴⁸⁴ It consists of five sections, each of which contains the words Cullagandhavamsa, which suggests that at one time there was a longer Mahāgandhavaṃsa. ¹⁴⁸⁵ It aims to give a brief history of the canon and post-canonical books in Ceylon and India (which includes Burma).

The first section deals with the canon and its divisions, and the various ways of classifying it, e.g. as five nikāyas or nine anigas. This seems to follow the classifications given by Buddhaghosa in the Samantapāsādikā. ¹⁴⁸⁶ The second section deals with the teachers who wrote books (gandhakārakācariyas), the porānācariyas and the atthakathācariyas. The last two groups are the same, and are those who held the three councils. None of them is mentioned by name except for Mahākaccayana, who is also included in the third category because the grammar and other works including the Netti and the Petakopadesa are attributed to him. ¹⁴⁸⁷ The list of authors then follows, in approximate chronological order, beginning with the authors of the Kurundī and Mahāpaccarī commentaries, Buddhaghosa, Buddhadatta, Ānanda, and Dhammapāla. It is not known what sources of information the author of the Gandhavamsa used. In many cases he was merely taking information which was available in the colophons of manuscripts, but he often gives names even where the relevant text does not include its author’s name. Even when two works are attributed to the same author, he sometimes differentiates between two writers of the same name, e.g. he ascribes the Saccasaṃkhepa to Culladhammapāla, ¹⁴⁸⁸ and the Abhidhānappadīpikā to Navamoggallāna. ¹⁴⁸⁹ The list ends with Ariyavaṃsa and Udumbara. A number of anomymous works are then listed.

The third section arranges the authors according to whether they wrote in India (including Further India) or Ceylon, and information is occasionally given about the country in which they lived, its king, and the position they held before becoming Buddhists. The fourth chapter then states whether the authors wrote their works by request, or of their own volition (attano matiyā ). Once again, much of this information was doubtless taken from the introductions and colophons of the texts themselves. The fifth section tells of the advantages (ānisamsās) which come from writing books.

The Sāsanavaṃsa ¹⁴⁹⁰ was written in Burma in A.D. 1861 by Paññāsāmi, tutor of King Min-dōn-min who held the fifth council a few years later. ¹⁴⁹¹ It consists for the most part of a translation from an earlier work written in Burmese in 1831 with some details removed, and additions made to cover the period A.D. 1831-60.¹⁴⁹² After a brief survey of the first three councils based on the accounts found in earlier chronicles, it deals with the history of the doctrine in the nine places where it was taken by Asoka’s missionaries. Most of the nine are dismissed in a page or two, but the history of the Dhamma in Sîhala, Suvaṇnabhūmi, and especially Aparanta, which the author interprets not as the Western part of India but as part of Burma, are dealt with at greater length.

The author includes in his account of the history of the doctrine in these places a list of authors and the works they wrote. He frequently states the texts from which he has gained the information he is giving, and on occasion he discusses differences between them, e.g. the accounts about Buddhaghosa found in the Cūlavamsa and the Buddhaghosuppatti, and his own statements in the Visuddhimagga. ¹⁴⁹³

The style of Paññasami’s language in very similar to that of Buddhaghosa for the earlier periods of his history, ¹⁴⁹⁴ and this doubtless reflects the fact that these portions of his work are largely based upon Pāli originals. It seems likely however, that in the portions of his book which are based upon a Burmese source, some influence of the Burmese language remains. ¹⁴⁹⁵ In particular, the handling of Burmese personal and place names has been treated inconsistently. Some have been translated; some have been transliterated (with errors introduced by non-Burmese scribes); and some have been replaced by Pāli designations. ¹⁴⁹⁶

A comparable work, but devoted to the authors and books of Ceylon, was produced in 1880 by Ācariya Vimalasāra thera, who also wrote the Sīmālakkhaṇadīpanī. ¹⁴⁹⁷ This was the Sāsanavamsadīpa, ¹⁴⁹⁸ written in Pāli verse, and covering the history of Buddhism in Ceylon down to the time of the introduction of the Burmese upasampadā in A. D. 1802¹⁴⁹⁹

Although these bibliographical texts sometimes disagree among themselves, ¹⁵⁰⁰ and often give no information, or inadequate information, when we most need it, their importance lies in the fact that they are usually dependent upon traditional material which is no longer available to us. ¹⁵⁰¹ Their accounts frequently enable us to put authors and their works into a historical background in a way which would otherwise be difficult, if not impossible. Without these texts, any description of Pāli literature, and especially the later post-canonical literature, would be little more than a list of works and authors, as culled from manuscripts and printed editions, with a summary of their contents. This is, unfortunately, still the situation with regard to much of the Pāli literature written in Cambodia, Laos, and Thailand. Although valuable work has been done in this field, ¹⁵⁰² much remains to be done before we know as much about Pāli writers in these countries as we do about authors who worked in India, Ceylon, and Burma.

ABBREVIATIONS

I. Texts (original sources)

A	Añguttara-nikāya
Att	Hatthavanagallavihāravamsa
Anāg	Anāgatavamsa
Ap	Apadāna
Ap-a	Apadāna-atṭ̣hakathā (= Visuddhajanavilāsinī)
Abh	Abhidhānappadīpikā
Abhidh-av	Abhidhammāvatāra
Abhidh-s	Abhidhammatthasangaha
As	Atthasālinī
It	Itivuttaka
It-a	Itivuttaka-atṭ̣hakathā (= Paramatthadīpanī)
Utt	Uttarajjhayaṇasutta
Utt-vn	Uttaravinicchaya
Ud	Udāna
Ud-a	Udāna-atṭ̣hakathā (= Paramatthadīpanī)
Udāna-v	Udānavarga
Upās	Upāsakajanālañkāra
Kkh	Kankkhāvitaranī
KJ	Kuṇālajātaka
K Mhv	Kambodian (Extended) Mahāvamsa
Kv	Kathāvatthu
Kv-a	Kathāvatthu-atṭ̣hakathā
Khuddas	Khuddasikkhā
Khp	Khuddakapāṭha
G Dhp	Gāndhārī Dharmapada
Gv	Gandhavamsa
Cūl	Cūlavamsa (= Mahāvamsa XXXVII 51-CI 29)
Cp	Cariyāpiṭaka
Cp-a	Cariyāpiṭaka-atṭ̣hakathā (= Paramatthadīpanī)
Ja	Jātaka and Jātakatṭ̣hakathā
Jinak	Jinakālamālī
Jina-c	Jinacarita
Jināl	Jinālañkāra
Tikap	Tikapatṭ̣hāna
Tikap-a	Tikapatṭ̣hāna-atṭ̣hakathā
Tel	Telakatāahagāthā
Th	Theragāthā
Th-a	Theragāthā-atṭ̣hakathā (= Paramatthadīpanī)
Thī	Therigāthā
Thī-a	Therigāthā-atṭ̣hakathā (= Paramatthadīpanī)
Thūp	Thūpavamsa
D	Dīgha-nikāya
Dāṭh	Dāṭhāvamsa
Divy	Divyāvadāna
Dīp	Dīpavamsa
Dukap	Dukapaṭṭhāna
Dukap-a	Dukapaṭṭhāna-aṭṭhakathā
Dbk	Dasabodhisattuppattikathā
Dvp	Dasavatthuppakaraṇa
Dhātuk	Dhātukathā
Dhātuk-a	Dhātukathā-aṭṭhakathā
Dhātup	(Pāli) Dhātupāṭha
Dhātum	Dhātumañjūsā
Dhp	Dhammapada
Dhp-a	Dhammapada-aṭṭhakathā
Dhs	Dhammasangani
Nāmar-p	Nāmarūpapariccheda
Nidd I	Mahāniddesa
Nidd II	Cullaniddesa
Nidd-a I	Mahāniddesa-aṭṭhakathā ( = Saddhammapajjotikā)
Nidd-a II	Cullaniddesa-aṭṭhakathā ( = Saddhammapajjotikā)
Nett	Nettipakarana
Nett-a	Nettipakarana-aṭṭhakathā
Pajj	Pajjamadhu
Paṭis	Paṭisambhidāmagga
Paṭis-a	Paṭisambhidāmagga-aṭṭhakathā ( = Saddhammapakāsinī)
Paṭṭh	Paṭṭhāna ( = Mahāpakaraṇa)
Peṭ	Peṭakopadesa
Pj I	Paramatthajotikā on Khuddakapāṭha
Pj II	Paramatthajotikā on Suttanipāta
PJ	Paññāsa Jātaka
P Dhp	Patna Dharmapada
Pp	Puggalapañnatti
Pp-a	Puggalapañnatti-aṭṭhakathā
Ppk-a	Pañcappakarana-aṭṭhakathā ( = Dhātuk-a, Pp-a, Kv-a, Yam-a, Paṭṭh-a)
Pv	Petavatthu
Pv-a	Petavatthu-aṭṭhakathā ( = Paramatthadīpanī)
Ps	Papañcasūdanī
Bu-up	Buddhaghosuppatti
Bv	Buddhavamsa
Bv-a	Buddhavamsa-aṭṭhakathā ( = Madhuravilāsinī)
M	Majjhima-nikāya
Mil	Milindapañha
Mil-t	Milinda-ṭikā
Mūlas	Mūlasikkhā
Moh	Mohavicchedanī
Mp	Manorathapūraṇī
MPS	Mahāparinirvāṇasūtra
MBh	Mahābhārata
Mvu	Mahāvastu
Mhbv	Mahābodhivamsa
Mhv	Mahāvamsa
Mhv-t	Mahāvamsa-ṭikā ( = Vamsatthappakāsinī)
Yam	Yamaka-pakarana
Yam-a	Yamaka-aṭṭhakathā
Ras	Rasavāhinī
Rūpār	Rūpārūpavibhāga
Lal	Lalitavistara
Lp	Lokappaññatti
Vin	Vinaya-pițaka
Vin-vn	Vinayavinicchaya
Vibh	Vibhanga
Vibh-a	Vibhanga-aṭṭhakathā (= Sammohavinodanī)
Vism	Visuddhimagga
Vism-mhṭ	Visuddhimagga-mahāṭīkā (= Paramatthamañjūsā)
Vutt	Vuttodaya
Vv	Vimānavatthu
Vv-a	Vimānavatthu-aṭṭhakathā (= Paramatthadīpanī)
S	Samyutta-nikāya
Sacc	Saccasamkhepa
Sadd	Saddanīti
Saddh	Saddhammopāyana
Saddhamma-s	Saddhammasangaha
Saddharmap	Saddharmapundarīka
Sandes	Sandesakathā
Samantak	Samantakūṭavaṇ̣nanā
Sah	Sahassavatthuppakarana
Sās	Sāsanavamsa
Sīmāv	Sīmāvivādavinicchayakathā
Sīh	Sīhalavatthuppakarana
Subodh	Subodhālañkāra
Sn	Suttanipāta
Sp	Samantapāsādikā
Spk	Sāratthappakāsinī
Sp-ṭ	ṭikā on Samantapāsādikā (= Sāratthadīpanī)
Sv	Sumañgalavilāsinī
Sv-pṭ	purānaṭikā on Sumañgalavilāsinī (= Līnatthappakāsinī)

II. Abbreviated titles of books mentioned in the footnotes

Adikaram, EHBC	E. W. Adikaram, Early History of Buddhism in Ceylon, Migoda, Ceylon, 1946
Alsdorf, AS	L. Alsdorf, Die Aryā-strophen des Pali-Kanons, Mainz 1968.
Balasooriya, BSWR	Somaratna Balasooriya et al. (ed.), Buddhist Studies in honour of Walpola Rahula, London 1980
Bapat, VV	P. V. Bapat, Vimuttimagga and Visuddhimagga: a comparative study, Poona 1937
Bechert, BCSRS	H. Bechert (ed.), Buddhism in Ceylon and Studies on Religious Syncretism in Buddhist Countries, Göttingen 1978
Bechert, LEBT	H. Bechert (ed.), The language of the Earliest Buddhist Tradition, Göttingen 1980
Bechert, PNTB	H. Bechert and H. Braun, Pāli Nīti Texts from Burma, PTS London, 1981
Bode, PLB	M. H. Bode, Pali Literature of Burma, London 1909
Buddhadatta, Corr	A. P. Buddhadatta, Corrections of Geiger’s Mahāvaṃsa, etc., Ambalangoda 1957
Cousins, BSIBH	L. S. Cousins et al. (ed.), Buddhist Studies in honour of I. B. Horner, Dordrecht 1974
Franke, Gramm	R. Otto Franke, Geschichte und Kritik der einheimischen Pāli-Grammatik und -Lexicographie, Strassburg 1902
Frauwallner, EV	E. Frauwallner, The earliest Vinaya and the beginnings of Buddhist literature, Rome 1956
Geiger, DM	W. Geiger, The Dīpavamsa and Mahāvaṃsa and their historical development in Ceylon, Colombo 1908
Geiger, PLL	W. Geiger, Pāli Literature and Language, Calcutta 1956
Godakumbura, Cat	C. E. Godakumbura, Catalogue of Ceylonese Manuscripts, The Royal Library, Copenhagen 1980
Gray, Prov	J. Gray, Ancient proverbs and maxims from Burmese sources, the Niti literature of Burma, London 1886
Heesterman, Pratid	J. C. Heesterman et al. (ed.), Pratidānam, den Haag 1969
Horner, BD	I. B. Horner, The Book of the Discipline, 6 Vols., SBB London, 1938-66
Horner, MLS	I. B. Horner, Middle Length Sayings, 3 vols., PTS London, 1954-59
Law, HPL	B. C. Law, History of Pāli literature, 2 vols., London 1933
Lüders, BSU	H. Lüders, Beobachtungen über die Sprache des buddhistischen Urkanons, Berlin 1954
Malalasekera, DPPN	G. P. Malalasekera, Dictionary of Pāli Proper Names, 2 vols., London 1937-38
Malalasekera, PLC	G. P. Malalasekera, Pāli Literature of Ceylon, London 1928
Narain, SPB	A. K. Narain, Studies in Pali and Buddhism, Delhi 1979
Norman, EV I	K. R. Norman, Elders’ Verses I, PTS London 1969
Norman, EV II	K. R. Norman, Elders’ Verses II, PTS London 1971
Pande, SOB	G. C. Pande, Studies in the origins of Buddhism, Allahabad 1957
Rahula, HBC	W. Rahula, History of Buddhism in Ceylon, Colombo 1956
Rhys Davids, DB	T. W. Rhys Davids, Dialogues of the Buddha, 3 vols., SBB London, 1899-1921
Roth, BV	G. Roth, Bhikṣuṇī-Vinaya: manual of discipline for Buddhist nuns, Patna 1970
de Silva, DAT	L. de Silva, Dīghanikāyaṭṭhakathāṭīkā Līnatthavaṇṇanā, 3 vols., PTS London, 1970
Thomas, HBT	E. J. Thomas, History of Buddhist Thought, London 1933
Waldschmidt, CASF	E. Waldschmidt, Central Asian Sūtra Fragments, in Bechert, LEBT, pp. 136-74
Warder, IB	Warder, Indian Buddhism, Delhi 1970
Warder, IKL	Warder, Indian Kāvya Literature, Vol. II, Delhi 1974
Warder, PM	Warder, Pāli Metre, PTS London, 1967
Winternitz, HIL	M. Winternitz, History of Indian Literature, Vol. II, Calcutta 1933
Yuyama, VT	A. Yuyama, Vinaya-Texte, Wiesbaden 1979

III. Periodicals, etc.

ADAW	Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin, Klasse für Sprachen, Literatur und Kunst
AGAW	Abhandlungen der Akademie der Wissenschaften in Göttingen, Phil.-Hist. Kl.
AO	Acta Orientalia, Copenhagen
BEFEO	Bulletin de l’École Française d’Extrême Orient, Paris
BSOAS	Bulletin of the School of Oriental and African Studies, London
CPD	Critical Pāli Dictionary, Copenhagen
IC	Indian Culture, Calcutta
IHQ	Indian Historical Quarterly, Calcutta
IIJ	Indo-Iranian Journal, Dordrecht
IT	Indologica Taurinensia, Turin
JA	Journal Asiatique, Paris
JAS(B)	Journal of the Asiatic Society of Bengal, Calcutta
JBRS	Journal of the Bihar Research Society, Patna
JIABS	Journal of the International Association of Buddhist Studies, Madison
JOI(B)	Journal of the Oriental Institute, Baroda
JPTS	Journal of the Pali Text Society, London
JRAS	Journal of the Royal Asiatic Society, London
KST	Kleinere Sanskrit-Texte
MW	Sir Monier Monier-Williams, Sanskrit-English Dictionary, Oxford
NGAW	Nachrichten der Akademie der Wissenschaften in Göttingen, Phil.-Hist. Kl.
PBR	Pāli Buddhist Review, London
PED	T. W. Rhys Davids and W. Stede, Pali-English Dictionary, London
PTC	Pāli Tipiṭakaṃ Concordance, London
PTS	Pali Text Society
SBB	Sacred Books of the Buddhists, London
SBE	Sacred Books of the East, Oxford
StII	Studien zur Indologie und Iranistik, Reinbek
STT	Sanskrittexte aus den Turfanfund
WZKS(O)	Wiener Zeitschrift für die Kunde Süd(- und Ost)asiens, Vienna
ZDMG	Zeitschrift der Deutschen Morgenlāndischen Gesellschaft, Wiesbaden

IV. General

Be	Edition in Burmese characters
BHS	Buddhist Hybrid Sanskrit
Ce	Edition in Sinhalese characters
ch.	chapter
diss.	dissertation
ed.	edition
Ee	Edition in Ruman letters
ibid.	in same book/article as quoted in previous note
MIA	Middle Indo-Aryan
Ne	Edition in Nāgarī characters
n(n).	note(s)
op. cit.	the work quoted (in the specified note)
Pkt	Prakrit
p(p).	page(s)
repr.	reprinted
rev.	reviewer, reviewed by
Skt	Sanskrit
s.v(v).	under the heading(s)
tr.	translator, translated by
v(v).	verse(s)

GLOSSARY OF PĀLI, PRAKRIT AND SANSKRIT TERMS

akusala	unwholesome, not moral, evil
akkhana	inopportune time
aksara	syllable
aṅga	section, especially of nine-fold Canon
acchariya	marvellous; a marvel
ajjhāyaka	teacher
aṭṭha	eight
aṭṭhakathā	commentary
aṭṭhaka-nipāta	section of eights
atideva	super-god
atitavatthu	a tale of the past
attā	self; individuality
atthavannanā	commentary
adbhuta	= abbhuta
adhikaranasamathā dhammā	legal questions
adhitthāna	resolution
adhideva	super-god
adhipati	predominance (condition); overlord
anattā	non-self
anavassuta	not letting in
aniyata	undetermined
anuggaha	kindness
anutīkā	supplement to a tīkā
anuppaññatti	secondary rule
anomodana	thanks
anuvyañjana	minor marks, characteristics
anussati	mindfulness
apadāna	noble deed
appamāda	lack of carelessness
abbhutadhamma	marvellous phenomenon
abhi (prep.)	regarding
abhidhamma	regarding the dhamma
abhinīhāra	aspiration for Buddhahood
abhivinaya	regarding the vinaya
arañña-vāsin	forest dweller
arahant	perfected one, one who has gained nibbāna
arūpa-vimokkha	immaterial release
arhat	= arahant
avadāna	noble deed
avassuta	letting in
avijjā	ignorance
avidūre-nidāna	beginning in not remote past
astau	eight
asainkhata	uncompounded
asangaha	unclassification
asangahita	unclassified
asura	opponent of the gods
ākhyāna	narrative poem
āgatāgama	one who has mastered the agamas
āgama	= nikāya
ācariya	teacher
ātman	= attā
ānāpāna	breathing in and out
ānisamsā	benefit, merit, advantage
ānuttariya	excellence
ābhidhammika	one versed in the Abhidhamma
āyatana	base
ārammana	object (condition)
ārāma	monastery
āsava	influx
itivuttaka	one of the nine anigas
itivrttaka	= itivuttaka
iddhi	supernormal power
iddhi-pada	base of iddhi
indriya	sense-faculty
udāna	solemn utterance
uddāna	summary verse
uddesa	statement, explanation
upadeśa	one of the Sanskrit anigas
upadesa	instruction, teaching
upadhi	material possessions; desire for material possessions
upasampada	ordination
upasarga	prefix, preposition
upādāna-kkhandha	aggregates of grasping
upādāya-rūpa	dependent form, quality
upāsaka	lay-follower
upekhā	equanimity
uppatti	(story telling) the origin (of a sutta)
ekacce	some (authorities)
eka-nipāta	section of ones
ekādasa-nipāta	section of elevens
ejā	lust
etarahi	now
aupamya	one of the Sanskrit anigas
kathina	cotton cloth (for robe)
kathamkatha	doubt
katha	discourse, topic
kappa	mental figment; aeon
kamma	action
kamma-tṭ̣hāna	subject for meditation
kamma-vācā	official act of the sangha
karuṇaā	pity, compassion
karuṇā-jhāna	meditation on compassion
karma-vākya, -vācanā	= kamma-vācā
kāmādinava	dangers in sensual pleasures
kāya	body
kārikā	concise statement, especially in grammatical text
kāvya	ornate poetry
kuti(kā)	hut
kusala	wholesome, moral, good
keci	some (authorities)
kṣatriya	= khattiya
khanda	section
khattiya	member of second (warrior) caste
khandha	aggregate
khandhaka	section, chapter
gana	group, section
ganthi-pada	word-list
gandha-kāra	book-writer
gandhabba	heavenly musician, demi-god
gāthā	stanza
gihi-vinaya	householder’s discipline
guru	teacher
geya	= geyya
geyya	recitation
gopālaka	cowherd
cakkavattin	universal monarch
catukka-nipāta	section of fours
catuttha	fourth
cattāro	four
campu	a style of ornate poetry
cariyā	(story relating) conduct, action
citta	mind
cetasika	mental concomitant
ceto-khila	barrenness of mind
chakka-nipata	section of sixes
chatttha	sixth
jarā	old age
jātaka	birth story
jāti	birth
jina	conqueror
jhāna	meditation
jhāyaka	meditator
ñāṇa	knowledge
tikā	sub-commentary
tanhā	craving
Tathāgata	epithet of arahant
tanti	canon
tapas	ascetic practices
(t)taya	triad, group of three
tika	triplet
tika-nipāta	section of threes
tipitaka	three baskets (of the Canon)
ti-sarana	three refuges
tepitaka	versed in the three baskets (of the Canon)
tevijja	possessor of the three vedas or vijjās
thūpa	shrine, tope
thera	elder (m.)
thera-vāda	tradition of the elders
theriya	(one) belonging to the tradition of the elders
theri	elder (f.)
dama	self-conquest
dassana	vision, insight
dāna	giving, generosity
dāna-vatthu	object suitable as a gift
dittthi	(wrong) view
disā	direction
disā-pharana	pervasion of the directions
dīpa	island
duka	couplet
dukkha	misery
dūre-nidāna	beginning in the remote past
deva	god
devatā	deity
deva-putta	member of the divine race
deva-loka	world of the gods
dvayatā	pair
dvepātimokkha	two pātimokkhas
dhamma	(sg.) doctrine; (pl.) states; (ifc.) subject to …
dhammatā	natural law
dhamma-yāga	spiritual sacrifice
-dhara	carrier of = expert in
-dharin	= -dhara
dharma	= dhamma
dhātu	element
dhātu-pātha	list of roots
dhutanga	ascetic practice
naya	guide line; style, method; maxim
nava	new, modern, later
nava-ṭikā	modern tikā
nāga	snake
nāgarī	a North Indian script
nāma	immaterial factor
nikāya	one of the five sections of the sutta-pitaka
nigamana-gātha	concluding stanzas, envoi
niṭṭthita	finished
nidāna	inception, origin; one of the Sanskrit anigas; link in causal chain
niddesa	exposition, exegesis
nipāta	section (often numerical in nature); particle
nipātaka	section
nibbāna	extinction, quenching
nimitta	omen (especially the four seen by the Bodhisatta)
niraya	hell
nirutti	etymological explanation
nirvāna	= nibbāna
nissaggiya	forfeiture
niti	worldly wisdom
nekkhamma	renunciation
nhātaka	one who has washed
paccaya	(n.) condition; (adj.) proper, fitting
paccuppanna-vatthu	tale of the present
pacceka-buddha	one enlightened for himself, a non-teaching Buddha
pañca	five
pañcaka-nipāta	section of fives
pañcama	fifth
paññatti	concept
pañ̃̃a	knowledge
paññāsa	fifty
paññāsaka	group of fifty (suttas)
pañha	question
pațicca-samuppāda	dependent origination
pațisambhidā	discrimination
pațisarana	support
paṭthāna-naya	causal relations
paṇnāsa	group of fifty (suttas)
patteya-buddha	= pacceka-buddha
pada-bhājanīya	word analysis
pabbajjā	initiation
paritit	protection, safeguard; protective utterance
parinibbāna	final extinction, quenching
parinibbuta	finally extinct, quenched
parivāra	supplement
parivena	monastery
pācittiya	expiation
pātidesanīya	confession
pātimokkha	rules of discipline
pāda	foot (of verse)
pāpādīnava	terrible evil
pāramitā	perfection
pārami	perfection
pārājika	defeat
pāli	Canon
pāli-bhāsā	language of the Canon
pitaka	basket (of texts)
piya-rūpa	of pleasing appearance
pucchā	question
puggala	person, individual
puppha	flower
pubbācariya	former teacher
purāna	early, ancient
purāna-ṭikā	ancient tīkā
petakin	versed in the pitakas
peta	departed spirit
potthaka	book
porāna	ancient (authority)
pratisaṃvid	= paṭisambhidā
pratika	lemma (in commentary, or introduction)
pratyeka-buddha	= pacceka-buddha
prasasti	eulogy
prātimoksa	= pātimokkha
phala	fruit, reward
phassa	contact
bamhana	= brāhmaṇa
bala	power
bilāra	cat
Buddha	enlightened one
Buddha-khetta	Buddha field
Buddha-parisā	Buddha’s acquaintances
Buddha-bhāsita	spoken by the Buddha
Buddha-vacana	Buddha’s word
bojjhanga	factor of enlightenment
bodhi	enlightenment; enlightenment tree
bodhi-pūjā	honouring the bodhi tree
Bodhi-satta	being destined for enlightenment
brahma-cārin	one who lives the holy life
brahma-vihāra	sublime state of mind
brāhmaṇa	brahman, member of first (priestly) caste
bhaya	fear
bhāṇaka	reciter
bhāvanā	meditative practice
bhāsā	speech, language
bhikkhu	monk
bhikkhunī	nun
bhiksu	= bhikkhu
bhikṣunī	= bhikkhunī
bhumma	terrestrial
bhūta-vāhana-yanta	robot
mati	thought, volition
marana	death
mahā-thera	great i. e. chief elder
mahā-nipāta	section of long poems
mahápadesa	principal appeal to authority
mahā-purisa	great man
mahā-bhūta	great essential (element)
mahā-sangiti	great council
Māgadhika-bhāsā	language of Magadha
mātikā	topics of discourse; summary list of contents
muni	seer, sage
mūla-bhāsā	root language
mettā	loving kindness
yakkha	supernormal being, demi-god
yāga	sacrifice
yuga-nandha	association, coupling
rakkhā-manta	saving chant
ratana-cankama	jewel walk
rasa	elegance of sound
rāja	king
rüpa	form, material quality
lakkhana	characteristic property
loka-santhiti	cosmology
lokottara	supra-mundane
vamsa	chronicle
vagga	chapter of text
vācī	speech, voice
vannanā	embellishment; commentary
vatthu	subject matter; story; section
vatthu-gāthā	introductory stanzas (of a sutta)
varga	= vagga
vastu	= vatthu
vāra	portion, section
vicikicchā	doubt
vijjā	knowledge
vinaya	discipline
vipassanā	insight
vibhanga	exegesis, classification
vimāna	heavenly abode
vimokkha	release
visuddhi	purity
vissajjanā	answering
vihāra	monastery
vîsati-nipāta	section of twenties
vutti	commentary
veda	knowledge; brahmanical text
vedanā	feeling, sensation
vedalla	analysis
veyyākaraṇa	exposition, commentary; grammarian
veramanī	avoidance, abstinence
vessa	member of the third caste
vyādhi	disease
śrāvakācāra	conduct of a layman
samyama	self-control
samyutta	connected collection
saṃvaṇnanā	commentary
saṃvibhāga	sharing together
saṃsāra	journeying on, transmigration
sagga	heaven
sañkhāra	constituent element
sañkhepato	in brief
sañkhyā	enumeration
sañgaha	collection; classification
sañgahīta	classified
sañgāyana	council
sañgiti	council
sañgiti-kāra	one who holds a council
sañgha	Order of Buddhists
sañghādisesa	formal meeting of the Order
sacca	truth
saññāvedayitanirodhasamāpatti	attainment of extinction of sensation of perception
sati-patthāna	basis of mindfulness
satta	seven
santike-nidāna	beginning in the present
samana	ascetic
samatta	finished
samaya	assembly
samāja	assembly
samādhi	concentration
samodhāna	identification
sarana	refuge
sāmañña-kanda	miscellaneous section
sāsana	teaching
sikkhāpada	rules of training
sigāla	jackal
simā	boundary
sīla	morality, moral conduct
sīha-nāda	lion’s roar, shout of exultation
Sīhala-bhāsā	language of Ceylon
Sīhala-sāsana	Sinhalese tradition of teaching
sukha	happiness
suñña	empty; emptiness
sutta	thread; discourse
suttanta	discourse
sudda	member of the fourth caste
supanna	winged creature, the enemy of nāgas
subhāṣita	eloquent saying
sūkara-maddava	tender pig’s flesh
sūtra	= sutta
sūtra-pātha	text in aphorism form
sekhiya	training
sotāpatti-anga	phase of entry into the stream
stūpa	= thūpa
svarabhakti	epenthesis, insertion of epenthetic vowel
svāgama	versed in the āgamas
hāra	mode
hetu	root (condition)

INDEX OF AUTHORS

(authors of the same name are not separated; n after a page number = in the footnotes to that page)

• Aggavamsa 164-66
• Atthadassi 163
• Anuruddha 144, 151-52
• Ariyavamsa 172n, 181
• Ananda 148, 152, 159-60, 170, 181
• Udumbara 181
• Upatissa 113, 133n, 141, 162
• Upasena 133, 137
• Kaccāna 108, 110
• Kaccāyana 163-66
• Kassapa 147, 153, 161, 172
• Khema 152
• Canda 147
• Candragomin 165
• Cāṇakya 176
• Culladhammapāla 152, 181
• Cūḷābhaya(-tthera) 150
• Chapaṭa 153, 164, 171
• Tibbatuvava 141
• Tissa 96, 103, 105, 125
• Daṇdin 157
• Dīpankara 164
• Dhammakitti 140, 142, 164, 179-80
• Dhammadinna 154
• Dhammanandi 154
• Dhammānanda 164
• Dhammapāla 40,61,71,73,89,109-10, 113,123,133-37,145-46,148-50, 152,163,165,181
• Dhammasiri 169
• Dharmapāla 149
• Nandapañña 180
• Navamoggallāna 166, 181
• Paññāsāmi 11-12, 181-82
• Pāṇini 129n, 147, 163-66
• Pingala 168
• Buddhaghosa 8, 10-11, 15-18, 27, 30-32,42,50,54,57,61-62,87-88,91, 96-99,103-4,109-34,137-39,145- 47,149-50,163,165,169,179,181-82
• Buddhadatta 124, 130-32, 145-46, 153, 157,168,181
• Buddhadāsa 162
• Buddhappiya 159, 162n, 164-65, 173
• Buddharakkhita 157-58
• Bhartrhari 149
• Mangala 145
• Mahākaccāyana 181
• Mahākauṣṭhila 107
• Mahānāma 117-18, 132, 137
• Mahāmanigala 145
• Mahāyasa 164
• Mahāsāmi 169
• Medhahkara 158, 165
• Moggaliputta 10 (see Tissa)
• Moggallāna 140, 165-67
• Moggallāyana 165
• Raṭṭhapāla 155-56
• Ratanapañña 143
• Rāmasamma 167
• Rāhula 172
• Vācissara 130-31, 142, 152, 164, 168, 171
• Vimaladhamma 144
• Vimalabuddhi 164
• Vimalasāra 172, 182
• Vedcha 145n, 155, 159
• Vepullabuddhi 168
• Vopadeva 166
• Sāntideva 57
• Sāriputra 107
• Saṅgharakkhita 167-69
• Saddhammakitti 167
• Saddhammajotipāla 153, 164, 171
• Saranamkara 163
• Sāriputta 11, 43, 84-85, 87, 99, 107, 132,149n, 151
• Siddhattha(-thera) 162n, 173
• Sirimangala 175
• Sīlathera 167
• Sīlavamsa 166
• Hāla 75

INDEX OF WORKS

(Works of the same name are not separated. Abbreviations: -k= -kathā; -kkh= -kkhandha(ka); -g = -gāthā; -j = -jātaka; -ṭ = -ṭīkā; -n = -nikāya; -p-v = -petavatthu; -v = -vagga; -var = -varga; -v-v = -vimānavatthu; -s = -sutta(nta); -samy = -samyutta; -sū =-sūtra; n after a page number = in the footnotes to that page)

• Agāriya-v-v 71 n
• Aggañña-s 41, 175
• Añguttara-n 8, 28, 30-31, 37, 43, 49-50,54-57,62,103,122,174-75;
• -aṭṭhakathā 128, 149n; -āgama 31
• Añgulimāla-s 47, 174
• Acchariyabbhutadhamma-s 48, 79
• Ajitapañha 69 n
• Aṭṭhakathākaṇda 99
• Aṭṭhaka-v 63-64, 67-69, 84-86
• Aṭṭhaka-s 68
• Atta-v 59 n
• Atthasālinī 98, 120, 122-26, 131, 147n
• Atthuddhāra 99-100
• Anamatagga-samy 52
• Anavatapta-g 92
• Anāgata-vaṃsa 161-62
• Anusaya-yamaka 105
• Andhaka-aṭṭhakathā 121
• Apaṇnaka-j 15, 129n
• Apaṇnaka-s 46
• Apadāna 9, 20, 23-24, 31-32, 65, 77, 87-92,121,127,135,146-47
• Abhidhamma(-piṭaka) 8-9, 15, 17, 27, 32,37,40,43,48,87-90,94,96-107, 121,123,129,131,147-49,151,157, 161; -bhājanīya 100
• Abhidhammatthasañgaha 144, 151-52
• Abhidhammāvatāra 124, 131, 146, 153
• Abhidharmakośa 68 n ; -vyākhyā 56
• Abhidharma-piṭaka 17, 37, 43, 107
• Abhidhānappadīpikā 150, 166-67, 181; -ṭīkā 167; -saṃvaṇṇanā 167; -sūci 166 n
• Abhisamācārikā 21
• Amarakośa 149, 166-67
• Ambaṭṭha-s 33, 36, 47
• Ariyapariyesana-s 45
• Arthapada-sū 68 n
• Arthavārgīya 68
• Alagaddūpama-s 45
• Avadāna 24, 92
• Avalokita-sū 26 n
• Avassuta-s 53
• Avyākata-samy 35 n
• Aṣṭaka-var 68
• Aṣṭa-var 68
• Asaṅkhata-samy 53
• Assalāvana-s 46
• Āṭānāṭika-sū 43
• Āṭānāṭiya-s 39, 43, 173-74
• Ālavaka-s 65
• Itivuttaka 9, 15-16, 31, 62-63, 133-34, 136
• Itthi-v 82
• Isigili-s 48
• Uttaragrantha 28-29
• Uttaravinicchaya 131
• Udayapañha 69 n
• Udāna 9, 15-16, 31, 60-62, 68, 133-34, 136,148,152
• Udāna-var 59-61
• Udumbarikasīhanāda-s 35, 41
• Upadeśa-śāstra 16 n
• Uparipanṇāsa 48-49
• Upasena-s 53
• Upasena-sū 53
• Upālipañcaka 28-29
• Upālipariprcchā-sū 28-29
• Upāli-v 28
• Upāli-s 46
• Upāsakajanālañkāra 170
• Upāsakālañkaraṇa 170
• Upāsakālañkāra 170
• Uposatha-kkh 22-23, 126, 171
• Uraga-j 72 n
• Uraga-p-v 72n
• Uraga-v 64, 84
• Uraga-s 64
• Ussadakittana 124 n
• Ekakkharakosa 167
• Ekuttaraka 28
• Ekuttara-n 54
• Ekottara 29
• Ekottarāgama 57 n
• Ekottarikāgama 31, 54, 56-57
• *Opamma-v 45
• Kañkhāvitaranī 126-27, 169
• Kaccāyana-gandha 163, 181n
• Kaccāyana-ppakaraṇa 147
• Kaccāyana-mañjūsā 166
• Kaccāyana-vutti 147
• Kaccāyana-vyākaraṇa 163, 165
• Kaccāyana-sāra 164; -ṭīkā 164n
• Kaṭhina-kkh 22-23
• Kanṭhaka-v-v 70
• Kannakatthala-s 47
• Kanha-p-v 72n
• Kathāvatthu(-ppakaraṇa) 10, 84, 90, 96,98,102-5,112n, 117,125,148,179
• Kandaraka-s 46
• Kamma-kkh 22
• Kammavācā 25
• Karanīyametta-s 174
• Karmavākya 26
• Karmavācanā 26
• Kalāpa 147
• Kalingabodhi-j 81
• Kavikalpadruma 166
• Kassapasīhanāda-s 34, 41
• Kātantra 163, 165
• Kāmandakī(yanītisāra) 176n
• Kāśikā 163
• Kimchanda-j 72n
• Kukkuravaṭika-s 46
• Kuṇāla-j 81, 150
• Kumbhakāra-j 82
• Kurundī 121, 126, 131, 181; -aṭ̣̣hakathā 131
• Kūṭadanta-s 34
• Kūṭavinicchayika-p-v 72n
• Kevaddha-s 35, 41
• Kokāliya-s 67
• Kosamba-kkh 22
• Kosalabimbavaṇnanā 179-80
• Kṣudraka-vastu 23
• Kṣudrakāgama 70, 72
• Khaggavisāna-s 64, 84-85, 91, 147
• Khandha 174
• Khandhaka 15, 18, 22-25, 28
• Khandha-v 52
• Khandha-samy 52
• Khuddaka-gantha 9, 32
• Khuddaka-n 8-9, 30-32, 57-95, 105, 123,132,134,172
• Khuddaka-pāṭha 9, 22n, 31-32, 57-58, 65-66,72,88,121,129,134,174
• Khuddaka-vatthu-kkh 22
• Khuddaka-vatthu-vibhanga 100n, 101
• Khuddasikkhā 168-69; -ṭikā 168-69
• Khettūpama-p-v 71
• Khema-ppakaraṇa 152
• Gaṇ̣̣hipada 166
• Gandhavam̧sa 121, 123, 127, 131-33, 141,145-46,148-50,152,155,157- 58,161-62,164,166,169,180-81
• Gahapati-v 45
• Gihivinayasangahaniti 176
• Guttila-j 71n
• Guttila-v-v 71n
• Gulatthadīpanī 157 n
• Gona-p-v 72n
• Gopakamoggallāna-s 48
• Gopālaka-s 45
• Ghaṭa-j 72n
• Ghaṭikāra-s 84n
• Ghoṭamukha-s 46
• Cakkavattisīhanāda-s 41, 93, 160
• Cakkavāladīpanī 175
• Catutthasangiti-k 11
• Catubhānavāra 174; -aṭ̣̣hakathā 174
• Catuṣpariṣat-sū 44
• Canda-s 174
• Candra-sū 52
• Campeyya-kkh 22
• Camma-kkh 22
• Cariyā-piṭaka 9, 31-32, 79, 83, 88, 94-95,133-34,136,145
• Cāṇakya-nīti 176; -śāstra 176
• Cittasambhūta-j 82
• Civara-kkh 22
• Culla-gandhavamssa 180
• Culla-niddesa 79n, 84,86
• Culla-niruttigandha 181n
• Culla-v 18, 21-22, 37, 146
• Culla-vedalla-s 16
• Culla-saddaniti 164
• Cūla-kammavibhanga-s 49
• Cūla-gopālaka-s 45
• Cūla-taṇhāsañkhaya-s 46
• Cūla-dukkha-kkh-s 46
• Cūla-niddesa 9, 32
• *Cūla-parinibbāna-s 37
• Cūla-māluñkya-s 44
• Cūla-yamaka-v 44
• Cūla-rāhulovāda-s 44n
• Cūla-vaṃsa 140, 182
• Cūla-v 66
• Cūla-satipaṭṭhāna-s 45
• Cūla-sāropama-s 45
• Cūla-hatthipadopama-s 44
• Cetiyavamsatṭ̣hakathā 138
• Chakesadhātuvamsa 143
• Chavaka-j 20n
• Chiggaḷa-s 53
• Janavasabha-s 38
• Jambudīpuppattivatthu 155
• Jātaka 9, 15-16, 20, 24-25, 31, 47-48, 71-72,77-84,89,94-96,121,127- 28,136,148,150,158-59,174,176- 79 ; -atṭ̣hakathā 134, 136; -atthavaṇṇanā 136,146,150; -nidāna 141-42, 146,148,150
• Jāliya-s 34
• Jinakālamālī 10-12, 143-44, 157, 178
• Jinacarita 158,165
• Jinālañkāra 157; -ṭīkā 157; -navaṭīkā 157; -vaṇnanā 157
• Jhāna-vibhanga 100 n
• Jhāna-samy 53
• Nāṇa-k 88
• Nāṇa-vibhanga 100 n
• Nāṇodaya 120
• Thāṇamga 28 n
• Tika-paṭṭhāna 106 n
• Tirokuḍda-s 58; -p-v 72
• Tissametteyya-s 69
• Tuvaṭaka-s 15,68
• Telakaṭāha-g 156
• Tevijja-s 36, 39
• Traibhūmi-k 174
• Thūpavamsa 142
• Thera-g 9, 15, 31, 47, 52, 61, 67, 71-77, 89-92,112,133-35,146
• Therāpadāna 147
• Therī-apadāna 147
• Therī-g 9, 15, 31, 51, 61, 71, 73, 75-77, 89-90,92,112,134-35
• Dakkhiṇāvibhanga-s 48
• Dantadhātu(-vamsa) 142
• Dantabhūmi-s 48
• Daśakumāracarita 158
• Daśādhyāya-vinaya 29
• Daśottara-sū 43
• Dasadānavatthuppakaraṇa 153
• Dasabodhisattavidhi 162
• Dasabodhisattuddesa 162
• Dasabodhisattuppatti-k 161-62
• Dasabrāhmaṇa-j 83
• Dasavatthuppakaraṇa 153
• Dasuttara-s 28,43,54,87
• Dāṭhāvaṃsa 142
• Divyāvadāna 81n, 175,178
• Dīgha-n 7-8, 24-25, 28, 30-50, 54-55, 77,79,87,90-93,95,102,107,111, 122,142,148-50,160,173-75; -aṭṭhakathā-ṭīkā 149; -āgama 31
• Dīpavamsa 8-11, 16, 27, 84, 88, 100, 115-19,138-40,142,154; -atṭhakathā 115,138
• Dirghāgama 24n, 31,38n, 43n, 44
• Duka-paṭṭhāna 106 n
• Dutiya-agāriya-v-v 71n
• Devatā-saṃy 51
• Devadūta-s 175
• Dvayatānupassanā-s 67
• Dvāra-k 124 n
• Dvipādavyāsayamaka-g 157 n
• Dhajagga-s 174
• Dhammacakkappavattana-s 53, 132
• Dhammaniti 176
• Dhammapada 9, 15, 25, 31, 58-61, 64, 75,121,127-28,130,159,176; -atṭṭhakathā 134-35,146
• Dhammasangani 96-102, 105-6, 120, 122
• Dhammasamādāna-s 45
• Dhammahadayavibhanga 101, 173
• Dharmapada 7, 59-60
• Dhātu-k 96, 98, 101-2
• Dhātukāya 107
• Dhātupāṭha 165-66
• Dhātumañjūsā 166
• Dhātumālā 165-66
• Dhātuyamaka 102
• Dhātuvibhanga 102
• Dhutagunaniddesa 114 n
• Dhvajāgra-sū 52
• Nandaka-p-v 71
• Nandivisāla-j 20n
• Nāga-saṃy 52
• Nāmarūpapariccheda 151-52
• Nāmarūpasamāsa 152
• Nāmācāradīpikā 153, 164
• Nālaka-s 15,67
• Nāvā-s 66
• Nikāya-sangraha 11
• Nigamana-g 136
• Nidāna-k 25, 48, 78-79, 95, 146, 158-59
• Nidāna-v 52
• Nidāna-samyukta 52
• Nidāna-samy 52
• Niddesa 15, 19, 27, 31-32, 63-65, 68-70,84-87,91,119,133-34,166
• Nidhikaṇda-s 58
• Nibbāna-samy 53 n
• Nīvaranapahāna-v 57
• Nettigandha 181n
• Netti-ppakarana 31, 109-10, 133-34, 137, 148-49, 166, 181; -atṭhakathā 133-34,137,148-49; -ṭīkā 148-49
• Nyāsa 164; -pradīpa 164
• Paccayākāravibhanga 37, 100n
• Paccarī 121
• Paccekabuddhāpadāna 91, 147
• Pajjapadoruvamsa 117, 138; -vaṇṇanā 138
• Pajjamadhu 158, 164, 173
• Pañcagatidīpana 160
• Pañcatantra 79
• Pañcappakarana-atṭ̣hakathā 123, 125
• Paññāsa-j 177
• Pañhāpucchaka 100-1
• Patisambhidāmagga 9, 31, 87-89, 96, 132, 134
• Patṭ̣hāna 96, 98, 106-7, 123, 151n, 161
• Pandita-var 59 n
• Panditālañkāranīti 176
• Pannattivāra 105
• Pannāsa 178
• Pada-bhājana 126
• Pada-bhājanīya 19, 21, 166
• Pada-mālā 164
• Pada-rūpasiddhi 164
• Padya- see Pajja-
• Padhāna-s 66
• Papañcasūdanī 122, 174
• Pabbajjā-s 66
• Payogasiddhi 158, 165
• Paramattha-k 123
• Paramattha-jotikā 64-65, 68-69, 8586,121n, 129,133,147
• Paramattha-dīpanī 123, 134, 146
• Paramattha-vinicchaya 152
• Paramattha-vibhāvinī 134 n
• Parābhava-s 65
• Pariññāvāra 106
• Paritta 120, 173-74
• Parivāra 15-16, 18, 24, 26-29, 113, 168
• Pavattivāra 106
• Pavārana-kkh 22
• Pāṭika-v 32, 41-44
• Pāṭika-s 32, 41
• Pātimokkha 7, 18-23, 31, 96, 121, 126, 166; -ṭhapana-kkh 22
• Pāyāsi-s 40
• Pārāyana-v 63-65, 69, 84-86
• Pārāyana-sū 68 n
• Pārivāsika-kkh 22
• Pāsādika-s 42-43
• Pitakattamain 154
• Pitakattayakārakānisamsa 178
• Puggala-k 103-4
• Puggala-paññatti 62-63, 96, 98, 102-3
• Punṇakapañha 69 n
• Punṇakamāṇavapucchā 69
• Punyavanta-j 84 n
• Puppha-v 58
• Pubbakammapiloti-apadāna 91-92
• Puspa-var 64 n
• Pūjāvaliya 11, 143
• Peṭaka 109
• Peṭakopadesa(gandha) 14n, 31, 108-10, 166,181
• Petavatthu 9, 31, 58, 71-72, 77, 13334,136,160
• Potṭhapāda-s 35
• Prātimokṣa-sū 20-21
• Pretāvadāna 72
• Bālapaṇdita-s 49
• Bālāvatāra 164
• Bāhiranidāna 116, 118, 122
• Bimbisāra-sū 49n
• Bilāra-j 80
• Buddhaghosuppatti 130, 138, 145, 182
• Buddhavaṃsa 9, 25, 31-32, 79, 87-88, 92-95,97,131-32,142,145-46,148, 161, 172; -atṭ̣hakathā 146-47
• Buddhāpadāna 90-92, 146-47
• Buddhāpadāniya 94
• Brhat-k 86 n
• Bojjhangaparitta 174
• Bojjhanga-samy 50,53
• Bodhivamsa 141
• Bodhisattvabhūmi 68 n
• Brapaṃsukūlānisamsa 178
• Brahmajāla-s 7, 33
• Brahmanimantaṇika-s 46
• Brahma-samy 50
• Brahmāyu-s 47
• Brāhmaṇadhammika-s 66
• Brāhmaṇa-v 58
• Bhaddāli-s 46
• Bhaddekaratta-s 48
• Bhikkhunī-kkh 21-22, 24-25, 55
• Bhikkhunī-pātimokkha 127n
• Bhikkhunī-vibhanga 18, 20-21, 28
• Bhikkhunī-samy 51, 76
• Bhikkhu-pātimokkha 127
• Bhikkhu-vibhanga 18
• Bhikṣuṇī-prakīrṇaka 21
• Bhikṣuṇī-prātimokṣa 21
• Bhikṣuṇī-vinaya 7, 21, 25
• Bhikṣuṇī-vibhanga 25
• Bhikṣu-prakīrṇaka 21
• Bhikṣu-prātimokṣa 20-21
• Bhikṣu-var 60, 64n
• Bhūmija-s 48
• Bhūridatta-j 80, 82
• Bhesajja-kkh 22, 162
• Bhesajja-mañjūsā 163
• Makhādeva-s 47
• Magga-samy 53
• Mangala-s 15, 58, 174
• Majjhima-n 8, 30, 40, 44-49, 67, 79, 95, 97,100,103,122,149,159,174; -paṇnāsa 46-47
• Mañjūsā 163
• Madhupinḍaka-s 103
• Madhuratthappakāsinī 150
• Madhuratthavilāsinī 131, 146, 157
• Madhurasavāhinī 155
• Madhura-s 46
• Madhyamāgama 31, 44, 49
• Manorathapūranī 122, 135
• Mahā-aṭṭhakathā 86, 99, 121, 124n, 131
• Mahā-kammavibhanga-s 49
• Mahā-kkh 22, 24, 44, 61
• *Mahā-gandhavam̧sa 180
• Mahā-gopālaka-s 45
• Mahā-govinda-cariya 95 n
• Mahā-govinda-s 38, 95n
• Mahā-govinda-sū 39n
• Mahā-govindīya-sū 39n
• Mahā-janaka-j 177
• Mahā-dibbamanta 174 n
• Mahā-dukkha-kkh-s 46
• Mahā-nidāna-s 37
• Mahā-niddesa 9, 32, 84
• Mahā-nipāta 18, 80
• Mahā-niruttigandha 181n
• Mahā-pakaraṇa 161n
• Mahā-paccarī 121, 126, 131, 181
• Mahāpadāna-s 24, 36-37, 48, 79, 90, 92
• Mahā-paritta 173
• Mahā-parinibbāna-s 31, 37-38, 42, 50, 55,93
• Mahā-parinirvāṇa-sū 38
• Mahā-punṇama-s 16
• Mahā-bodhivamsa 141, 156, 162, 165
• Mahā-mangala-s 66
• Mahā-māluṅkya-s 44
• Mahā-yamaka-v 44
• Mahārahanīti 176
• Mahā-rāhulovāda-s 44n
• Mahāli-s 34
• Mahā-lomahamsacariya 95 n
• Mahā-vaṃsa 8-12, 96, 115-18, 120, 124,130,138-42,145,149,153-56, 159,162-63,171,174,180 ;-ṭīkā 150 , 155-56,166,171
• Mahā-v 18, 22, 24-25, 32, 36-41, 53-54,66,126,140
• Mahāvadāna-sū 24, 37
• Mahā-vastu(avadāna) 16, 24-25, 39, 60,64-67,70,84-85,91,175,178
• Mahā-vibhanga 18, 27, 126
• Mahā-vedalla-s 16
• Mahā-samgāma 28
• Mahā-saccaka-s 34n, 45
• Mahā-satipaṭṭhāna-s 40,45,49
• Mahā-saddanīti 164
• Mahā-sannaya 164
• Mahā-samaya-s 39, 43
• Mahā-samāja-sū 39
• Mahā-sāropama-s 45
• Mahā-sīhanāda-s 45-46,95 n
• Mahā-sudassana-s 38
• Mahā-supina-j 80-81
• Mahā-hatthipadopama-s 44
• Māgandiya-pañha 86
• Māgandiya-s 68n, 69
• Mātanga-j 82
• Mātikā-atṭ̣hakathā 126
• Milinda-ṭikā 150
• Milinda-pañha 26n, 28,31,54,74,91, 102,110-12,150,174; -vaṇṇanā 150
• Mukhamattadīpanī 164
• Muni-g 69n
• Muni-s 69n
• Mūla-ṭikā 148, 152
• Mūla-paṇnāsa 45
• Mūla-yamaka 105
• Mūla-sikkhā 169
• Metta-s 58, 65
• Metteyya-pañha 69n
• Maitreya-samiti 161 n
• Moggallāyana-pañcikā 165
• Moggallāyana-vyākaraṇa 165
• Mora-s 174
• Mohavicchedanī 96, 100, 105, 147, 153, 161
• Yamaka 96, 98, 102, 105-6, 118
• Yogaratnākara 163
• Raṭ̣̣hapāla-s 46
• Ratana-s 15, 58, 66, 174
• Rasavāhinī 145n, 155-56,159
• Rājanīti 176
• Rāṣ̣̣rapāla-sū 49n
• Rūpasiddhi 159, 164-65, 173
• Rūpārūpavibhāga 131, 153
• Lakkhana-s 42
• Lalitavistara 81n
• Līnatthavaṇnanā 149n
• Lokaniti 176
• Lokaneyyapakaraṇa 176 n
• Lokapaññatti 174-75
• Lokaprajñapti 175
• Lomahaṃsapariyāya 95 n
• Lohicca-s 36
• Vaṃsatthappakāsinī 138-40
• Vangisathera-samy 51
• Vaṇnanitigandha 181 n
• Vatta-kkh 22
• Varṇaśata 49n
• Vassupanāyika-kkh 22
• Vākyapadīya 149
• Vāgiśa-saṃyukta 52
• Vidhurapaṇdita-j 71n, 80,83
• Vinaya(-piṭaka) 8-9, 11, 15-16, 18, 20-21,23-28,31,37,44,48,53,55, 58,61,77,85,87,92,97-98,105,113, 116,119,121-22,126,131,140,147, 157,162,166,168-69,171,173; -aṭ̣̣hakathā 122, 138
• Vinaya-kṣudraka 28
• Vinaya-ṭikā 11
• Vinaya-nidāna 16n, 116n
• Vinaya-vinicchaya 130-31, 145
• Vinaya-vibhanga 21
• Vinaya-sāratthadīpanī 130
• Vipākuddhāra-k 124 n
• Vibhanga 15, 23, 37, 48, 89, 96, 98, 100-3,105,122; -aṭ̣̣hakathā 125 ; -v 48-49
• Vimativinodanī 172
• Vimalatthavilāsinī 146
• Vimalavilāsinī 146
• Vimānavatthu 9, 25, 31, 40, 70-72, 133-34,136,146,173
• Vimānāvadāna 70
• Vimuktimārga 114 n
• Vimuttimagga 113-14, 120
• Visuddhajanavilāsinī 146n
• Visuddhimagga 31, 112-14, 120-24, 126,129,131-33,148-49,174,182; -ṭikā 113, 133
• Vuttodaya 168-69; -ṭikā 168
• Vedanā-samy 50 n
• Vedalla-s 16,40
• Vevacanahāra 166
• Vessantara-j 80,83,95,177
• Vyākaraṇa 165
• Śikṣāsamuccaya 57
• Śila-var 59
• Saṃyuktāgama 31, 54
• Samyutta-n 8-9, 15, 30-31, 40, 49-54, 56,76,120,122,149,159,174; -aṭ̣̣hakathā 128; -āgama 31
• Saṃsāramocaka-p-v 71
• Sakkapañha-s 16,40
• Saka-samy 50
• Sagātha-v 15, 50-52
• Saṅkhārabhājaniya-s 16
• Saṅkhepa 121
• Saṅkhyāsarūpa 150
• Saṅgārava-s 47
• Saṅgiti-paryāya 43, 107
• Saṅgiti-vaṃsa 144
• Saṅgiti-s 28,43,48,102,107,150
• Saṅgiti-sū 43, 102
• Saṅghabheda-kkh 22
• Saṅghabheda-vastu 24, 44n
• Saṅgharājasādhucariyāva 1 n
• Saccavibhanga 100; -s 45n, 48-49,100
• Sacca-samy 50, 53
• Sacca-saṅkhepa 152, 181
• Sacca-savhayapaṇditacariya 95 n
• Satipaṭ̣̣āna-saṃy 53
• Satipaṭ̣̣āna-s 40,45
• Sattasai 75
• Saddaniti 164
• Saddalakkhaṇa 165
• Saddhamma-ṭ̣̣hitikā 133
• Saddhamma-ppakāsinī 87, 132
• Saddhamma-ppajotikā 84, 133
• Saddhamma-saṅgaha 10-11,124n, 148, 152,157-58,164,179-80
• Saddhammopāyana 159, 170; -viggaha 160
• Saddharma-puṇdarika 16
• Sandesa-k 144, 172
• Saptamaithunasaṃyukta-sū 57
• Saptaśataka 75 n
• Sabhiya-s 66
• Samatha-kkh 22
• Samantakūṭavaṇnanā 155, 159
• Samantapāsādikā 32,57,98,116,118, 121-22, 126-27, 130, 141-42, 169, 171,180-81
• Samantabhaddikā 162
• Samavāyampa 28 n
• Samuccaya-kkh 22
• Samuddaghosa-j 178
• Sampasādanīya-s 42
• Sambandhacintā 168
• Sammādiṭ̣̣hi-s 16
• Sammohavinodanī 122-25, 133
• Salāyatana-v 53
• Salāyatana-samy 50, 53
• Sahasra-var 60
• Sahassa-vatthu 155; -aṭ̣̣hakathā 15556; -ppakarana 154-56
• Sāmagāma-s 48
• Sāmaññaphala-s 7, 33-36, 111
• Sārattha-dīpanī 11
• Sārattha-ppakāsinī 122
• Sārattha-sangaha 162n, 173n
• Sārasangaha 11, 173-74
• Sāriputta-samy 52
• Sāropama-s 45
• Sāsana-vamsa 2, 11-12, 26, 131, 141, 145,148,152-53,181; -dīpa 141n, 182
• Sigālovāda-s 42
• Sivi-j 82
• Sīmā-k 171
• Sīmā-lakkhaṇadīpanī 172, 182
• Sīmālañkāra(ka) 171; -sangaha 171
• Sīmā-vivādavinicchaya-k 172
• Sīmā-sańkaracchedanī 172
• Sīmā-sangaha 171
• Sīla-kkh-v 32-36
• Sīhaladīpuppatti-vatthu 155
• Sīhalavatthuppakarana 154
• Sujāta-j 72 n
• Sutavaddhananīti 176
• Sutta-niddesa 164
• Sutta-nipāta 9, 14-15, 25, 31, 58, 6370,73,75,84-86,91,121,129,133- 34,147,175
• Suttanta-piṭaka 9, 30, 32
• Suttanta-bhājanīya 100
• Sutta-piṭaka 17, 30-95, 97-98, 102-3, 105,129,173
• Sutta-mālā 165
• Sutta-vibhanga 18, 20-22, 27-28, 126
• Sutta-sangaha 31, 172
• Sudhanukumāra-j 178
• Subodhālañkāra 167-69; -ṭīkā 168
• Subha-s 35
• Sumañgalavilāsinī 109, 122, 149
• Sūtra-piṭaka 16 n
• Sedamocaka-g 28
• Senāsana-kkh 22
• Serissaka-v-v 71n
• Sela-s 67
• Sonadanda-s 34
• Skandhaka 23n, 24-25
• Sthavira-g 75, 92
• Hatthavanagallavihāravamsa 143
• Hatthipadopama-s 44
• Hitopadeśa 79
• Hemavata-s 65, 175

GENERAL INDEX

(Persons or places with the same name are not separated)

• Abhayagiri(-vihāra) 5, 10, 29n, 11314, 139, 149, 160, 170
• Aciravata 48
• Acyutavikrama 131
• Aesop 79
• Afghanistan 24
• Aggabodhi 133n
• Ajātasattu 33, 143
• Akkharacchandas 168
• Āḷāra Kālāma 45
• Alokavihāra 11
• Alulena 11n
• Aluvihāra 11
• Ambapālī 76
• Ambaṭṭha 33-34, 47
• Ānanda 7-9, 35, 37-38, 42, 47-48, 50, 61,69,73,77,86,91,159,174,179
• Anāthapindika 79
• Andhaka 122
• Andhra 154
• Anomadassi 143
• Anurādhapura 133, 165, 172
• Anuruddha 8, 12, 50
• Anuṣṭubh 59, 61, 69, 74, 82-83
• Aparanta 182
• Ārāḍa Kālāma 45n
• araññāyatana 159
• Arimaddana 153, 164
• Āryā (including Old Āryā) 61, 65, 68, 71,74,76,81,110,168
• Aryan Path 40
• Asandhimittā 140, 153
• Asita 67
• Asoka(n) 3-5, 10, 16, 39, 73, 77, 84, 87, 104,115,140,142,153,162,172,175, 182
• Assaji 23n
• Attanagalla 143
• Atthadassin 128
• Aupacchandasaka 59, 61, 64
• Ayuthya 144n
• Badaratittha(-vihāra) 134, 137, 152
• Balie/Baly 2
• Bamiyan 24, 107
• Bāvarī 69
• Benares 3n
• bhāṇaka 8-9, 31-32, 50, 56, 74-76, 81, 90n, 94-95,112-113,132
• Bharhut 16, 80
• Bhātikatissa 115n
• Bhātutissa 115n
• Bhūmija 48
• Bhuvanekabāhu 158
• Bimbisāra 66
• Bodhisatta 9, 77, 79, 82-83, 93-94, 159,173
• Brahmā 36, 38-39, 46, 50
• Brahmakāyika 39
• Brahmāyu 47
• Buddha 2-3, 6-9, 11, 13, 19-25 et passim
• Buddhadeva 128
• Buddhamitta 128
• Buddhasīha 131, 145
• Buddhasoma 160
• Buddhism 4-7, 11-12, 48
• Buddhist(s) 1, 3, 5-6, 9-14, 18, 20 et passim
• Buddhist (Hybrid) Sanskrit 26, 30n, 78-79,81,88,92,177
• Burma 2, 7, 11-12, 14, 79n, 108, 110, 138,140n, 153,164,167,172,174-77, 180-83
• Burmese 2, 7, 11-13, 14n, 143, 151, 154,164-65,167,169,172,176-78, 181-82
• Burnouf 1
• Cakkavattin 41, 160
• Cambodia 13, 160n, 177, 183
• Cambodian 13, 140, 150
• Cāṇakka 139
• Candagutta 139
• Ceylon 2, 6, 10-12, 14, 26, 29, 92, 94, 108,110,114-15,118-21,124,130, 132,134,136n, 140-44,151-52, 154-56,159,161-62,164,167,169- 72,174-75,179-83
• Chaṭṭhasangāyana 13
• Chieng Mai 175
• Childers 1
• Chinese 7, 22n, 29, 32, 43n, 44, 49, 54, 56-57,60,63-64,68,92,98,107,111, 113-14,130,143n, 149,175
• Clough 1
• Cola 132
• Coliyan 171
• Coromandel 2
• Cūlagavaccha 72
• Cunda 178
• Dabba 19n
• Dakkhiṇārāma 173
• Dāṭhānāga 141, 149
• Deva 133
• Devānaṃpiyatissa 116
• Dhammakitti 180
• Dhammasava 72
• Dhammāsoka 71
• Dharmaguptaka 21, 29, 38, 43n, 59
• Dīpankara 78, 93, 141-42, 158
• Dravidian 6, 121, 130
• Duṭṭhagāmaṇī 117, 142, 180
• Ekavihāriya 73
• English 1, 141
• Fa-hsien 179
• Fan 63 n
• Gaṇacchandas 70n, 73-74,168
• Gāndhārī 7, 59
• Ganges 37
• Gilgit 7, 20-21, 24, 57, 92
• Girnār 4
• Gotama 36, 38, 45n, 67, 77, 89, 93, 140-43,158,161,178
• Govinda 39n
• Greek 110-11
• Haṇdakonḍa 154 n
• Haripuñjaya 143
• Hāthigumphā 4-5
• Hemavata 175
• Hīnayāna 7, 16, 20-21, 26, 29n, 91, 113
• Hindu 161
• Hsüan-tsang 14, 16n, 17
• Inda 66
• India(n) 3-5,14n, 29,59,65n, 77,79, 81,83-84,87,92,108,110-11,113- 14,120-21,124,130,134,136,142- 43,149,151-52,154-55,171,175-76, 179,181-83
• Iñgirisi 141
• Isidāsi 77
• Jagatī 59, 61, 66-67, 69, 74, 82, 142
• Jain(s) 20, 28n, 42-43, 46, 48-49, 51, 58,81-82,170
• Jāliya 34
• Jambudīpa 26
• Java 86
• Jayasena 48
• Jetavana 79, 114
• Jīvaka Komārabhacca 162
• Jotipāla 84 n
• Kaccāna 108, 110
• Kalabbha 132
• Kalamba 131
• Kalambha 132
• Kalasi 111n
• Kālāsoka 139
• Kalinga 81
• Kanaka 146
• Kañcipura 152
• Kaniṣka 14, 17
• Kanjur 14
• Kanṭakasolapatțana 154 n
• Kapilavastu 3n
• Kashmir 14
• Kasmīra 111n
• Kassapa 8, 40
• Kāśyapīya 59n
• Kavicakravarti 160
• Kāvīra 152
• Kāvīrapaṭṭana 146
• Kevaddha 35
• Khadiravaniya 73n
• Khemā 135n
• Khotan 59
• Kimbila 72n
• Kittisīhasūra 167
• Kittisirirājasīha 140
• Konāgamana 77, 97
• Konḍakasolapatṭana 154n
• Kosala 80
• Kumārakassapa 127
• Kusinagarī 3n
• Kusinārā 3n, 38
• Kuthodaw Pagoda 12
• La Loubère 1-2
• Lampoon 143
• Lānnā 143-44
• Laos 13, 183
• Laotian 13
• Lassen 1
• Let-than 153
• Licchaviputta 41
• Lîlăvatî 142
• Lokottaravādin 7, 21-22, 24-25, 39, 60
• Magadha 3,33,38n, 66
• Māgadhī 2-3,105,145,168
• Māgadhisms 104, 130
• Māgandiya 69
• Mahā-ariṭṭha 10,116,180
• Mahādeva 104 n
• Mahāgavaccha 72
• Mahākaccāna 46,48,85,103
• Mahākassapa 7-8,12,18,148,180
• Mahāmaudgalyāyana 53 n
• Mahānāma 130, 132
• Mahāpaduma 121
• Mahāpajāpati 48
• Mahārakkhita 87
• Māhārāṣṭrī 75
• Mahāsammata 41
• Mahāsānghika 7,10,14,21-22,24-25, 27,39,57,60,84,88,96,100,117
• Mahāsangiti 10
• Mahāsangītika 10
• Mahāsena 115,140
• Mahāsumma 121
• Mahāthūpa 142
• Mahātissa 127
• Mahāvihāra 29n, 113,116,118-20,124, 128,131-32,134,136,139,147,149n, 151-52,170-71
• Mahāvijita 34
• Mahāyāna 42,74,90-91,113,161,170
• Mahimsāsaka 128
• Mahinda 114,116,119,121,141,145
• Ma-ho-nan 130, 149
• Mālinī 142-43, 159
• Mallian 19 n
• Mandākrānta 142
• Mandalay 12-13
• Manimekhalā 117
• Māra 51,66,158
• Mātale 11 n
• Mathurā 5
• Mātrācchandas 70n, 74,168
• Mauryan 3
• Medhañkara 93 n
• Menander 110-11
• Metteyya 41, 93, 161
• Middle Indo-Aryan 1-2, 5-7, 33, 57, 115,141
• Milinda 110-12
• Min-dōn-min 12, 181
• Moggaliputta 10-11
• Moggallāna 19n, 22,46,53,70,115n, 132
• Mūla-sarvāstivādin 7, 20, 23-24, 26, 29, 44n, 57,92
• Nādikas 38
• nāga 83
• Nāgasena 110-12
• Nālaka 67
• Nandas 139
• Nandivisāla 20 n
• Narindasīha 163
• Nāt(h)aputta 42-43, 46, 48
• Nerañjarā 66
• Nigodha 35
• Paccaya 80,83
• Pagan 153-55, 162-63, 168, 176
• Pāli 1-10,12,16,18,21-26 et passim
• Parakkama 142
• Parakkamabāhu 12, 140, 142-43, 148, 151,163,165-66,169-70,180
• Parakkantibhuja 142
• Pasenadi 179
• Pāṭaliputra 4, 38
• Paṭṭakoṭṭi 154 n
• Pāvā 43
• Pāyāsi 40
• Pingala(ka) 71
• Poṭṭhapāda 35
• Prakrit 7, 29n, 39, 57, 59-60, 63, 91, 111,130,166
• Pramitākṣarā 42 n
• Prasenajit 179
• Punabbasu 23 n
• Puṣpitāgrā 42 n
• Puṭabhattasela 159
• Rāhula 159
• Rājagaha 24
• Rangoon 13
• Rathoddhatā 42n, 76
• Raṭṭhapāla 19n
• Revata 9,73 n
• Roma 175
• Rucirā 42n, 118
• Sabbakāmin 9
• Sabhiya 66
• Sāgalā 111 n
• Sakka 40,46,50
• Sākya 34
• Sāli 155-56
• Samanola 159
• Sambhūta 73
• Saṇdakonti 154 n
• Saṅghapāla 120, 131, 145n, 157
• Saṅgharāja 145n, 180
• Saṅkha 41, 160-61
• Sanskrit 2-3, 5-7, 16, 29-31, 38-39, 43,51,53-54,57,60,62-63,66-68, 79,92,107,111,114,129,141,147, 149,156,158-59,163,165-70,174- 78
• Saranañkara 93n
• Sardūlavikrīdita 142
• Sāriputta 8, 22, 42-43, 52, 71, 84-85, 87,93-94,98-100,107,132,151,161
• Sarvāstivādin 7, 14, 17, 21, 24, 26, 28-29,37-38,40,43-44,49,51,54,57, 59,61,63,68,70,72,75,92,94,97, 107, 112
• Sātāgira 175
• Sela 67
• Sena 149
• Shāhbāzgarhī 39
• Siālkot 111 n
• Siam 2, 11-12, 140
• Siamese 13-14, 174
• Siddhagāma 149
• Siddhattha 159
• Sigāla 42
• Sīha 23 n
• Sīhala 8, 116-19, 122, 124n, 125-32, 134,135 n, 138-39,144 n, 154-55,182
• Sīhalācariya 170
• Sikhariṇi 142
• Sinhalese 1, 5-6, 12-13, 14n, 86-87, 94,100,104,114,119,130,140-44, 151,155,157,161,163,165,171,179
• Sinhalese Prakrit 6, 114, 118-19, 130
• Sirikūṭa 127, 130
• Sirimeghavaṇna 142
• Sirinivāsa 130
• Sirisañghabodhi 133, 143
• Sīva 11,26
• Sivi 82
• Śloka 82-83, 91, 94-95, 116, 118, 142, 152,159-60
• Sona(ka) 11
• Sonadanda 34
• Sopāka 135n
• South-East Asia 6, 140, 143, 153-54, 175,177-78
• Sragdharā 118, 142
• Śrāvastī 3n, 179
• Sudassana 38
• Suddhāvāsakāyika 39
• Sudhamma 12
• Sudinna 19n
• Sumanakūṭa 159
• Sumangala 73
• Sumati 131
• Sumedha 78-79, 93, 158
• Sumedhā 77
• Sunakkhatta 41
• Suraṭṭha 71
• Susunāga 139
• Suvaṇṇabhūmi 11, 86, 182
• Takkola 86
• Tamba 152
• Tambapaṇni 26, 169
• Taṇhañkara 93n
• Tañja 152
• Tathāgata 46, 173
• Tekiceakāni 73
• Thai(s) 2, 7, 13, 144, 175, 180n
• Thailand 2, 7, 14, 143-44, 176n, 17778,183
• Thatôn 11-12, 174
• Theravādin(s) 1-5, 10, 14, 17, 32, 38, 51,64,84,88,91-92,94,96,98,100, 103-4,107,112-13,117,128n, 136, 160-61,174
• Thūpārāma 165
• Tibet(an) 21, 60, 92, 107, 114
• Tilok 143
• Tilokanayana 153
• Tissabhūti 127
• Triṣ̣ubh 59, 61, 66-67, 69, 74, 82-83, 91,95,142-43,153,159
• Turkestan 7, 21, 24, 29, 37, 43, 49, 5152,57,59,68,75,92,107
• Tusita 67, 78-79
• Uddaka Rāmaputta 45
• Udgatā 42 n
• Udraka Rāmaputra 45n
• Upāli 7-8, 18, 28-29, 46
• Upaniṣadic 46
• Upasena 53
• Upasthitapracupita 42 n
• Upatissa 113, 174
• Uppalavaṇnā 77
• Uttara 11
• Uttaravihāra 139, 149
• Vaddha 73
• Vaitālīya 59, 61, 64, 83
• Vamśastha 42 n
• Vangīsa 9, 51, 73
• Vasantatilakā 118, 142-43, 159
• Vaṭṭhagāmaṇī 10-11,87
• vedha 81
• Vejayanta 46
• Vesālī 24
• Vessantara 83, 95
• Vidhūra 83
• Vijayabāhu 12, 140, 158
• Vipassin 36, 48, 79
• Vitandins 103
• Vitasoka 73
• yakkha 83
• Yama 175
• Yasodharā 90
• Yudhiṭthila 83
• Zimmè 178

Footnotes

1. For a survey of earlier views about the name “Pāli,” see O. von Hinüber, “Zur Geschichte des Sprachnamens Pāli,” in Beitrāge zur Indienforschung: Ernst Waldschmidt zum 80. Geburtstag gewidmet, Berlin 1977, pp. 237-46. ↩

2. E.g. Pālimattam idh’ ānitam, n ’ atthi atthakathā idha (Mhv XXXVII 227); sabbākārena n’ eva Pāliyaṃ na atthakathāyaṃ āgataṃ (Vism 107, 15-16). ↩

3. Cf. imaṃ pālinayānurūpaṃ saṃvaṇnanaṃ dāni samārabhissaṃ (Sp 2, 11-12) with manoramaṃ bhāsam tantinayānucchavikaṃ āropento (Sv 1, 19-20). ↩

4. R. C. Childers, A dictionary of the Pāli language, London 1875, p.vii n. 1. ↩

5. The earliest example quotable seems to be from the Sañgharājasādhucariyāva (śaka 1701 = A.D. 1779), according to C. H. B. Reynolds (quoted by Bechert, BCSRS, p. 16). ↩

6. B. Clough, A compendious Pali grammar with a copious vocabulary in the same language, Colombo 1824. ↩

7. E. Burnouf et Chr. Lassen, Essai sur le Pali ou langue sacrée de la presqu’île au-delà du Gange, Paris 1826. ↩

8. Simon de La Loubère, The kingdom of Siam, London 1693 (repr. 1969). ↩

9. Ibid. p. 9. ↩

10. Ibid. p. 168. ↩

11. Ibid. p. 11. ↩

12. [Buddhaghoso] icc evam pālibhāsāya pariyattim parivattitvā pacchā … jinacakkam … atidibbati (Sās 31, 31); cf. parivattesi sabbā pi Sīhalatthakathā tadā sabbesam mūlabhāsāya Māgadhāya niruttiyā (Mhv XXXVII 244). ↩

13. See V. B. Liebermann, “A new look at the Sāsanavaṃsa,” in BSOAS 39 (1976), pp. 137-49. ↩

14. sammāsambuddho pi tepitakam Buddhavacanaṃ tantim āropento Māgadhabhāsāya eva āropesi (Vibh-a 388, 7-8). ↩

15. Buddhaghosa says: Māgadhikāya sabba-sattānaṃ mūlabhāsāya (Vism 441, 34); cf. sabbesaṃ mūlabhāsāya Māgadhāya niruttiyā (Mhv XXXVII 244). ↩

16. ubhinnam pi pana kathaṃ asunanto Māgadhabhāsam bhāsissati (Vibh-a 387, 32-33). ↩

17. See Lüders, BSU, §§ 58-86, 88-115. ↩

18. See O. von Hinüber, “Pāli as an artificial language,” to appear in IT 10 (1982), and K. R. Norman, “Four etymologies from the Sabhiya-sutta,” in Balasooriya, BSWR, p. 183 n. 21. ↩

19. See Lüders, BSU, §§ 1-21, 220-25. ↩

20. See K. R. Norman, “The dialects in which the Buddha preached,” in Bechert, LEBT, pp. 69-71. ↩

21. F. Edgerton (BHS Grammar, New Haven 1953, p. 3 n. 8) has pointed out that neither the Buddha’s home (Kapilavastu), nor one of his favourite dwelling places (Śrāvastī), nor the scene of his first sermon (Benares), nor the place of his death (Kuśinagarī, Pāli Kusinārā) was in Magadha. ↩

22. I use “Māgadhī” here in the sense of the language of Magadha at the time of Asoka, without thereby implying that it necessarily had the same features as the grammarians’ Māgadhī. ↩

23. When asked if his words could be translated chandaso, the Buddha forbade it, but added: anujānāmi bhikkhave sakāya niruttiyā buddhavacanaṃ pariyāpuṇitum (Vin II 139, 14-16). For a discussion, with references to earlier views, of the words chandaso and sakāya niruttiyā, see J. Brough, “Sakāya niruttiyā: Cauld kale het,” in Bechert, LEBT, pp. 35-42. ↩

24. If the Northern tradition that the second council took place in Asoka’s reign is correct (see Thomas, HBT, p. 34), then it is possible that the Theravādins held a third council later in the same reign, after ridding themselves of heretics. Perhaps the figure of 118 years connected with this third council refers to the number of years after the Buddha’s parinibbāna, not after the second council. ↩

25. J. Bloch, states: “Girnar fournit la langue la plus proche du pali des livres bouddhiques singhalais, qui nous servira de repère: c’est du reste en raison de cette ressemblance qu’on suppose le pali originaire d’une région voisine” (Les inscriptions d’Asoka, Paris 1950, pp. 44-45). ↩

26. See Oldenberg, Vin I, p. liv. ↩

27. See K. R. Norman, “The dialects in which the Buddha preached,” in Bechert, LEBT, p. 69. ↩

28. In his revised edition of the inscription, B. M. Barua read brahmanānam in line 8a and -prāci- in line 9 (“Hāthigumphā inscription of Khāravela,” in IHQ 14, p. 466). ↩

29. See Th. Damsteegt, Epigraphical Hybrid Sanskrit, Leiden 1978, p. 265. ↩

30. See p. 10 below. ↩

31. See Brough, G Dhp, p. 218. ↩

32. It would be possible to argue that the form brāhmana had been retained unchanged from the very oldest stage of MIA for this same reason. It can, however, be shown that in certain contexts, at least, the expected MIA form of the word occurred. The etymologies which connect brāhmana with the root bāh-, e.g. bāhitapāpo ti brāhmano (Dhp 388), make sense only in a dialect where initial br- had become b-. ↩

33. See de Silva, DAT, I, pp. lvi-lviii. ↩

34. See O. von Hinüber, “On the tradition of Pāli texts in India, Ceylon and Burma,” in Bechert, BCSRS, pp. 55-57. ↩

35. Helmer Smith (Sadd, p. vi) writes of “la conviction que notre pali est une fonction de celui du 12me siècle.” ↩

36. See K. de Vreese, “Dravidian idioms in later Pali,” in Orientalia Lovaniensia Periodica 11, pp. 179-222. ↩

37. See Victor B. Lieberman, “A new look at the Sāsanavamsa,” in BSOAS 39 (1976), p. 146. ↩

38. See N. A. Jayawickrama, The sheaf of garlands of the epochs of the conqueror, PTS London, 1968, pp. xxxvii-xxxix, and G. Terral, “Samuddaghosajātaka,” in BEFEO 48 (1956), pp. 263-65. ↩

39. See Brough, G Dhp, § 50. ↩

40. See F. Edgerton, BHS Grammar, New Haven 1953, p. xxv. Edgerton classes the Mahāsānghika-Lokottaravādin Mahāvastu as being the oldest BHS text we have. ↩

41. Ed. G. Roth, in “The language of the Arya-Mahāsānghika-Lokottaravādins,” in Bechert, LEBT, pp. 78-135, and also by N. S. Shukla, The Buddhist Hybrid Sanskrit Dharmapada, Patna 1979. ↩

42. From the Pkt form ditthā the redactor produced drsstā instead of drsttva. ↩

43. Ed. G. Roth, Bhikṣuṇī-vinaya: manual of discipline for Buddhist nuns, Patna 1970. ↩

44. Vin II 285, 9-287, 28. ↩

45. Sp 16, 14-15. ↩

46. Dip IV 11; Mhv III 35. ↩

47. See Thomas, HBT, p. 28 n. 1. ↩

48. Dip IV 18-20. See p. 16 below. ↩

49. Sp 30, 1-5. ↩

50. Brahmajālassāpi “evam me sutan” ti ādikam āyasmatā Ānandena pathama-mahāsamgiti-kāle vuttam nidānam ādi (Sv 2, 14-15). ↩

51. Sv 13, 23-24; 15, 2-13. ↩

52. See Adikaram, EHBC, pp. 24-32. ↩

53. Mil 342, 1. ↩

54. Compare especially Th 1253-62 with S I 196, 11-21. ↩

55. Sv 15, 22-29. ↩

56. Ja I 59, 31-32. ↩

57. Sv 11, 3. ↩

58. See Adikaram, EHBC, p. 32. ↩

59. Vin II 307, 35. ↩

60. Dip V 28; Mhv IV 63. ↩

61. Mhv IV 62. ↩

62. Sp 34, 16-22. ↩

63. Dīp V 32. ↩

64. Mhv V 4. ↩

65. Dīp V 51; Mhv V 10. ↩

66. Dīp VII 39-40; 57-59. ↩

67. Mhv V 275. ↩

68. Sp 61, 19. ↩

69. Sp 102, 23-103, 22. ↩

70. Saddhamma-s 43, 29-44, 29. ↩

71. Saddhamma-s 45, 11. ↩

72. Jinak 50, 17. ↩

73. Sp 104, 5-6. ↩

74. Dīp XX 20-21. ↩

75. Mhv XXXIII 100-101. 76 “During his: reign 500 Rahats resided in the Alulena cave and rehearsed the [sacred] books” (Pūjāvaliya, tr. B. Gunasekera, Colombo 1895, ch. 34). “At that time 500 Rahats who assembled at Alulena in the country of Mātale, under the patronage of a certain chief, recited and reduced to writing the text of the Three Pitakas” (Nikāyasanggraha, tr. C. M. Fernando, Colombo 1908). ↩

76. As I am informed by Mr C. H. B. Reynolds, Lecturer in Sinhalese at the School of Oriental and African Studies in the University of London. ↩

77. Jinak 61, 11-12. ↩

78. Saddhamma-s 49, 5-6. ↩

79. Sās 23,26-27. ↩

80. Gv 61, 30-31. ↩

81. Sās 20,5-6. ↩

82. Vin V 3, 1-28. ↩

83. Dip VIII 12; Sp 64, 3-4; Mhv XII 6. ↩

84. Sās 36,33-37,3. ↩

85. See Pe Maung Tin and G. H. Luce (tr.), The Glass Palace Chronicle of the Kings of Burma, London 1923, pp. 46-50. ↩

86. See G. H. Luce, “The advent of Buddhism to Burma,” in Cousins, BSIBH, pp. 119-38. ↩

87. silālekhanesu vuttam (Sās 64, 1-2). ↩

88. Mhv LX 5-7. ↩

89. Mhv LXXVIII 16-27. ↩

90. Jinak 93, 17-20. ↩

91. Jinak 115, 5-6. ↩

92. Mhv C 152-53. ↩

93. See Malalasekera, PLC, p. 309 and Bode, PLB, pp. 77-78 and 83-84. ↩

94. See W. B. Bollée, “Some lesser known Burmese Pāli texts,” in Heesterman, Pratid, pp. 493-99. ↩

95. See D. E. Smith, Religion and Politics in Burma, Princeton 1965, pp. 160-63. ↩

96. The anusvāra and the i-mātrā are similar in the Burmese script, but not in the Sinhalese, so that any Sinhalese manuscript which shows -am as a mistake for -i must be based upon a transcript from a Burmese manuscript. ↩

97. Nāyamoli suggests (Peṭ tr., p. xiii) that an original manuscript of the Petaakopadesa may have been imported into Burma from Ceylon or South India at an early date, and all other Indian and Sinhalese manuscripts have been subsequently destroyed by time and neglect, so that the text is now entirely dependent upon a single Burmese manuscript, now lost. ↩

98. See D. Andersen and H. Smith, Sn, p. vi n. 1. ↩

99. See S. Beal (tr.), Buddhist record of the western world, London 1884, I, pp. 152-56. ↩

100. See Thomas, HBT, p. 175. ↩

101. See ibid., p. 176. ↩

102. Sp 16, 18-22. ↩

103. pañca nikāye pucchi (Vin II 287, 27-28); nikāye pañca vācesuṃ (Vin V 3,5). ↩

104. E.g. M I 133, 24-25. ↩

105. Ps II 106, 8-28 = Sp 28, 8-29, 3. ↩

106. Tr. N. A. Jayawickrama, The inception of discipline and the Vinayanidāna, London 1962, pp. 25-26. ↩

107. So veyyākarana is commonly used of an exposition in a sermon, e.g. veyyākaranasmim bhaññamāne (D I 46, 27-28). ↩

108. See Thomas, HBT, pp. 276-78. ↩

109. This word too seems to be used of a type of text. Hsüan-tsang states that the commentary which was composed on the Sūtra-piṭaka was called Upadeśa-śāstra (S. Beal, Buddhist record of the western world, London 1884, I, p. 155). The equivalent of Pāli vedalla in the Sanskrit lists is vaipulya. This word seems in other contexts to be the equivalent of Pāli vetulla, which suggests a derivation from a pair of Prakrit words vevulla and veyulla, showing the -v-j-y- alternation. It is not clear whether these words are in fact connected with vedalla. ↩

110. Mvu I 300, 21. ↩

111. Saddharmap (ed. P. L. Vaidya, Darbhanga 1960), II. 45. ↩

112. Dip IV 18-20. ↩

113. Vin V 3, 14. ↩

114. See E. Hultzsch, “Über eine Sammlung indischer Handschriften und Inschriften,” in ZDMG 40, p. 58. ↩

115. See S. Beal (tr.), Buddhist record of the western world, London 1884, I, p. 155. ↩

116. See J. Takakusu, “On the Abhidharma literature of the Sarvāstivādins,” in JPTS 1904-1905, p. 73. ↩

117. Ed. H. Oldenberg, The Vinaya Piṭaka, 5 vols, London 1879-83. Tr. I. B. Horner, The book of the discipline, 6 vols, London 1938-66. The order in which Buddhaghosa comments upon the sections of the Vinaya-piṭaka makes it clear that the Suttavibhanga precedes the Khandhaka, i.e. Oldenberg’s Vols III-IV should rightly come before his Vols I-II. ↩

118. eten’ eva upāyena ubhato-vinaye pucchi (Vin II 289, 8-9). ↩

119. Sp 14, 19-15, 3. ↩

120. tattha paṭhamasamgītiyaṃ saṃgītañ ca asamgītañ ca sabbam pi samodhānetvā (Sp 18, 3-4). ↩

121. I follow the analysis given in Horner, BD I, p. xi. ↩

122. The section cannot, however, be a very late addition, as it occurs in the Prātimokṣas of the other schools. ↩

123. At Vin III 158, 24-26 Dabba, the Mallian, is appointed to the task of assigning lodgings and distributing meals. At Vin IV 38, 3-4 he is accused of acting out of favouritism, suggesting that Pācittiya rule XIV, which is associated with the latter incident, is later than Samghādisesa rule VIII, which is associated with the former. See Horner, BD II, p. xxxvii and also p. xl. ↩

124. E.g. the story of Moggallāna at Vin III 104-108 has a parallel at S II 254-62, where, however, there is no reference to Moggallāna being held by other bhikkhus to be claiming “a state of further-men” (uttarimanussadhamma), which seems to be the pivot of the Vinaya passage (see Horner, BD I, pp. xxv-xxvi). See also O. von Hinüber, “Sprachliche Beobachtungen zum Aufbau des Pāli-Kanons,” in StII 2, pp. 35-37, where he points out that a linguistic feature in the Sudinna/ Raṭṭhapāla episode in Vin III 11, 34-21, 17 is older than the corresponding feature in M II 55, 24-65, 6. ↩

125. See Horner, BD I, p. xxxiv. ↩

126. As suggested by Horner, BD II, p. xxxix. ↩

127. E.g. Saṃghādisesa rules IX, X, and XI (see Horner, BD I, p. xvi). ↩

128. Or almost word by word. Horner, BD II, pp. xxv-xxvi, points out a few omissions. ↩

129. See Horner, BD I, p. xxxiv. ↩

130. See Horner, BD I, p. xxxiv. ↩

131. The slight variations of order and number found in the Prātimokṣas of other schools would suggest an independent origin for some of the rules. ↩

132. E.g. the story of the ox Nandivisāla at Vin IV 5, 11-38 = Ja I 191-93 (Nandivisālajātaka). See also the comparison of the versions of the story about the impious brahman found at Vin IV 203-204 = Ja III 27-30 (Chavaka-jātaka) made by L. Alsdorf (“The impious brahman and the pious candāla,” in Cousins, BSIBH, pp. 9-13). ↩

133. See A. Hirakawa, A study of the Vinaya-piṭaka, Tokyo 1960 (English summary), p. 14. ↩

134. udditthā kho ayyāyo chasatthisatā pācittiyā dhammā (Vin IV 345, 24). ↩

135. Ankul Chandra Banerjee, Prātimokṣa-sūtram [Mūlasarvāstivādin], Calcutta 1954 . ↩

136. W. Pachow and Ramakanta Mishra, “The Prātimokṣa sūtra of the Mahāsāṃghikas,” in Journal of the Gangānāth Jhā Research Institute, 9, pp. 239-60; 10, App. pp. 1-48; 11-12, pp. 243-48, and N. TAtia, Prātimokṣasūtram, Patna 1976. ↩

137. For references see Yuyama, VT, p. 1. ↩

138. See Yuyama, VT, pp. 3-4. ↩

139. See Waldschmidt, CASF, pp. 164-67. ↩

140. See the concordance table given by Charles S. Prebish, in Buddhist Monastic Discipline, Pennsylvania State University Press 1975, pp. 140-48. ↩

141. See Prebish, ibid., pp. 50-52. ↩

142. See Yuyama, VT pp. 6-7. ↩

143. See Yuyama, VT p. 9. ↩

144. See Yuyama, VT pp. 19-20. ↩

145. Gustav Roth (ed.), Bhikṣuṇī-Vinaya, Patna 1970. ↩

146. Vin II 253-57. ↩

147. B. Jinananda (ed.), Abhisamācārikā (Bhikṣuprakīrṇaka), Patna 1969. Rev. J. W. de Jong, IIJ 16, pp. 150-52. ↩

148. The only Apadāna-type insertion in the Bhikkhunīvibhanga is at Vin IV 258-59. There are seven in the Bhikṣuṇī-Vinaya, according to Roth, BV, p. xxxviii. ↩

149. The number 22 agrees with Buddhaghosa’s statement (Sp 18, 5) about the contents of the Vinaya-piṭaka. The Chinese translation of Sp, however, states that there were 23 khandhakas. This may be a mistake, as the translator suggests (see P. V. Bapat and A. Hirakawa (tr.), Shan-Chien-P’i-P’o-Sha, Poona 1970, p. 10 n. 16), but it may indicate dependence upon a slightly different recension of Sp. The list of Sutta-piṭaka texts in the same translation (p. 11) does not include the Khuddaka-pāṭha (see p. 57 below). ↩

150. I follow the translations given in Horner, BD IV, p. xxvii and V, p. xxiii. ↩

151. This account seems to be a continuation of the Mahāparinibbānasutta, and in the canon of some schools is found in conjunction with it (see p. 37 below). ↩

152. Vin I 103, 12-104, 20. ↩

153. As in the case of the Suttavibhanga, the rules are not all promulgated by the Buddha, which allows the possibility that some were made by his disciples, perhaps even after his death. See Horner, BD IV, p. xxiii. ↩

154. Horner, BD IV, p. xiii. ↩

155. Frauwallner concludes (EV, p. 135) that the author of the Skandhaka work had available to him an old account of the way in which the Buddha gave fundamental instruction to his first followers. ↩

156. E.g. the uddissaka rule about meat eating (= Vin I 233-38) has the same story about the senāpati called Sīha as is found at A IV 179-88, but in the latter text there is no reference to the promulgation of the rule. See Chandra Shekhar Prasad, “Meat-eating and the rule of tikotiparisuddha,” in Narain, SPB, pp. 289-95. ↩

157. Frauwallner (EV, pp. 146-48) suggests that the account of the eight marvellous qualities in the sea (attha mahāsamudde acchariyā abbhutā dhammā) at Vin II 236-40 is derived from A IV 200-204. ↩

158. At Vin III 248-51 and Vin I 206-209 we find the same story about the ārāma attendant, leading up to the promulgation of the same rule. In the former context it is Nissaggiya (“Forfeiture”) rule XXIII, and in the latter context it is one of the rules concerning permissible medicines. Oldenberg has pointed out (Vin I, p. xxiii n. 1) that the disciplinary proceedings enunciated in connection with the story of the bhikkhus who are followers of Assaji and Punabbasu have a later form at Vin II 9-14 than they do at Vin III 179-84. See also Horner, BD V, p. vi. ↩

159. E.g. Vin II 201, 12-28. ↩

160. Vin I 103, 12-104, 20. ↩

161. As suggested by G. Roth in “Symbolism of the Buddhist stūpa,” in A. L. Dallapiccola (ed.), The stūpa: its religious, historical and architectural significance, Wiesbaden 1980, p. 186. ↩

162. See Frauwallner, EV p. 185, and p. 28 below. 47 See Frauwallner:, EV pp. 150-53. ↩

163. See Yuyama, VT pp. 24-28. ↩

164. See Yuyama, VT p. 52. In the Sarvāstivādin canon it is found in the Dīrghāgama (see p. 44 below). ↩

165. See Yuyama, VT pp. 9-11. ↩

166. See Yuyama, VT p. 39. S. Lévi (JA 1932, p. 5) identified it as dealing with the pratisāraṇiya karma (cf. Vin II 15-19). ↩

167. Ed. E. Senart, Le Mahāvastu, 3 vols., Paris, 1882, 1890, 1897. Tr. J. J. Jones, The Mahāvastu, 3 vols., SBB London, 1949, 1952, 1956. ↩

168. āryamahāsāmghikānām lokottaravādinām madhyadeśikānām pāthena vinayapitakasya Mahāvastuye (Mvu I 2, 13-14). ↩

169. Jones, Mvu tr., I, pp. xi-xii. ↩

170. See p. 36 below. ↩

171. See Jones, Mvu tr., I, p. xiv. ↩

172. Mvu I 167-68. ↩

173. See Jones, Mvu tr., I, p. xii. ↩

174. Jones (Mvu tr., I, p. xvii) points out that there are often two or more versions of the same story, usually one in prose followed by one in verse, suggesting a difference in date of composition. ↩

175. See H. Oldenberg, Studien zum Mahāvastu, Nachrichten von der k. Gesellschaft der Wissenschaften zu Göttingen, Philol.-Hist. Klasse, 1912 and G. Roth, “Particle dāni in the Vinaya texts of the Mahāsāmghika-Lokottaravādin and the inscriptions of Aśoka,” in Studies in Indo-Asian Art and Culture (Acharya Raghu Vira Commemoration Volume), 1973, pp. 211-18. ↩

176. See p. 21 above. ↩

177. See T. W. Rhys Davids, “A new Kammavācā,” in JPTS 1906-1907, pp. 1-3. ↩

178. See T. W. Rhys Davids and H. Oldenberg (tr.), Vinaya Texts Part I, Oxford 1881, p. xx n. 1, and G. L. M. Clauson, “A new Kammavācā,” in JPTS 19061907, pp. 4-7. ↩

179. See Yuyama, VT pp. 4-5, 17-18. ↩

180. See F. Edgerton, BHS Dictionary, New Haven 1953, s. v. parivāra. ↩

181. In the Mvu (II 397, 7) it is used of the colophon to the second version of the Avalokita Sūtra, and Jones comments that “in the present instance the supplement is foisted into the middle of the work” (Jones, Mvu tr., II, p. 35 n. 1). ↩

182. Including Rhys Davids, who noted that the Parivāra is not quoted in the Milindapañha (T. W. Rhys Davids (tr.), The questions of King Milinda, Oxford 1890, p. xxxix). See also Warder, PM, p. 4. ↩

183. See Adikaram, EHBC, pp. 86-87. ↩

184. Sās 20,5-6. ↩

185. Ime atṭ̂ha vārā sajjhāyamaggena likhitā (Vin V 48, 29). ↩

186. Imam vitthārasamkhepam sajjhāmaggena majjhime cintayitvā likhāpesi sissakānaṃ sukhāvaham (Vin V 226, 6-7). ↩

187. As suggested by N. A. Jayawickrama, The inception of discipline, London 1962, p. 107. ↩

188. By Malalasekera, DPPN, II, p. 162. ↩

189. Dip V 37. ↩

190. Horner, BD VI, p. ix. ↩

191. Solasa Parivārā (Sp 18, 5-6). ↩

192. Vin V 179, 19. ↩

193. As suggested by Horner, BD VI, p. xiii. ↩

194. Sp 1301-1414. ↩

195. E.g. tattha bhaddako te ummango ti bhaddakā te pañhā (Sp 1348, 6). ↩

196. See Horner, BD VI, pp. vii-xi. ↩

197. “I believe that in some Buddhist countries the monastic disciple has to learn it before he studies these other parts” (Horner, BD VI, p. vii). ↩

198. Pande points out that this system is also used in non-Buddhist writings, e.g. the Jaina Thāṇamga and Samavāyamga (Pande, SOB, p. 26 n. 44). ↩

199. Horner, BD VI, pp. xxi-xxii. ↩

200. See p. 23 above. ↩

201. See H. Bechert, “Some remarks on the kathina rite,” in JBRS 54 (1968), pp. 323 and 328. ↩

202. A V 70-79. ↩

203. See p. 112 below. ↩

204. See Thomas, HBT, p. 268 and Frauwallner, EV, p. 46. ↩

205. See Yuyama, VT, p. 11 and Valentina Stache-Rosen, “The Upālipariprcchāsūtra,” in Buddhism and Jainism, Cuttack (Orissa) 1976, p. 27. ↩

206. Taisho 1466. Stache-Rosen (ibid., p. 30) states: “The possibility that the Upālipariprcchāsūtra originated in any of these centres in India, where Pali was used, cannot be ruled out. It is, however, more likely that the text originated in Ceylon. If this is the case, the text cannot have belonged to the Mahāvihāra monastery, as the order of the rules is different. The Upālipariprcchāsūtra might have belonged to the Abhayagirivihāra.” It is difficult to follow this argument. It seems impossible to conclude that because certain passages in Chinese agree in form with certain Pāli passages, although not in order, the Chinese must have been translated from Pāli. Since the Prātimokṣas of all the Hīnayāna schools are very similar, it is not unreasonable to guess that their auxiliary texts were also similar, in which case the Chinese might have been translated from Sanskrit or Prakrit. ↩

207. See Stache-Rosen, ibid., p. 27. ↩

208. See Frauwallner, EV, p. 46. ↩

209. See Yuyama, VT, pp. 32-33. ↩

210. Sp 16, 12-14. ↩

211. See p. 8 above. ↩

212. Since the Buddhist usage of the word sutta differs so greatly from the brahmanical use of sūtra, it has been suggested that Pāli sutta is to be derived from Sanskrit sûkta, not from sūtra. If this is so, then the use of the word sūtra in BHS would be the result of an incorrect back-formation. The word sutta is frequently interchanged with suttanta. ↩

213. Kasmā pan’ esa Dīgha-nikāyo ti vuccati? dīghappamānānaṃ suttānaṃ samūhato nivāsato ca (Sp 26, 26-27). katamo Majjhima-nikāyo? majjhimappamāṇāni pañcadasavaggasamgahāni Mūlapariyāyasuttādīni diyaddhasatạ̣ dve ca suttāni (Sp 27, 6-8). ↩

214. See Thomas, HBT, p. 268. ↩

215. See PED, s.v. āgama. ↩

216. Agamo nāma antamaso opamma-vagga-mattassa pi buddhavacanassa pariyāpuṇanam (Vism 442, 30-31). ↩

217. E.g. Vin II 249, 16. ↩

218. Sv 2, 10; Spk I 2, 22; Mp I 3, 4. ↩

219. Majjhe Visuddhimaggo esa catunnam pi āgamānaṃ hi (Sv 2, 6= Ps I 2, 10= Spk I 2, 18= Mp I 2, 24). ↩

220. Āgacchantu bhaddantā imaṃ suttantaṃ pariyāpuṇissanti purāyaṃ suttanto palujjatī ti (Vin I 140, 37-141, 1). ↩

221. For the four mahā-padesas see D II 123, 30-126, 5= A II 167, 33-170, 19. ↩

222. Sp 18, 12-15. ↩

223. See Bode, PLB, pp. 4-5. ↩

224. Sv 15, 22-29. ↩

225. See P. V. Bapat and A. Hirakawa (tr.), Shan-Chien-P’i-P’o-Sha, Poona 1970, p. 11 . ↩

226. Ed. T. W. Rhys Davids and J. Estlin Carpenter, PTS London, Vol. I, 1890; Vol. II, 1903; J. Estlin Carpenter, Vol. III, 1911. Tr. T. W. Rhys Davids, Dialogues of the Buddha, SBB London, Vol. I, 1899; T. W. and C. A. F. Rhys Davids, Vol. II, 1910; Vol. III, 1921; (German) K. E. Neumann, Reden Gotamo Buddho’s aus der lāngeren Sammlung Dīgha-nikāyo des Pali-Kanons übers., Vols. I, II, III, München 1907, 1912, 1918. For an analysis of the text see Pande, SOB, pp. 77-115. ↩

227. See Pande, SOB, pp. 77-115. ↩

228. Rhys Davids, DB, Vol. I, p. 3 n. 1. ↩

229. Tr. (with commentaries) by Bhikkhu Bodhi, The Discourse on the allembracing net of views, Kandy 1978. ↩

230. See K. R. Norman, “Pāli and the language of the heretics,” in AO 37 (1976), pp. 117-26. ↩

231. Ubhato sujāto mātito ca pitito ca saṃsuddhagahaṇiko yāva sattamā pitāmahāyugā akkhitto anupakkuttho jātivādena (D I 113, 25-27). ↩

232. They are mentioned in the Mahāsaccakasutta = Majjhima-nikāya (36). See p. 45 below. ↩

233. The ten avyākatas are dealt with separately in the Avyākata-samyutta (S IV 374-403). ↩

234. Aham yā imasmim janapade janapada-kalyān̄̄ tam icchāmi tam kāmemi (D I 193, 4-6). ↩

235. Ekam samayam āyasmā Ānando Sāvatthiyam viharati … acira-parinibbute Bhagavati (D I 204, 2-4). ↩

236. Tīni kho imāni Kevaddha pātihāriyāni mayā sayam abhiñña sacchikatvā paveditāni. katamāni tīni? iddhi-pāṭihāriyam ādesanā-pāṭihāriyam anusāsanipātihāriyam (D I 212, 16-18). ↩

237. The same seems to be true of the Majjhima-nikāya. See p. 44 below. ↩

238. The six previous Buddhas are, however, already mentioned by name at Vin III 7, 34-37. ↩

239. As described at Vin I 1, 10-16. ↩

240. See p. 89 below. ↩

241. Ernst Waldschmidt, Das Mahāvadānasūtra, Teil I-II, Berlin 1953, 1956 (ADAW 1952 Nr. 8, 1954 Nr. 8). ↩

242. Vibh 135-92. ↩

243. Atha ca pana me uttānakuttānako viya khāyati (D I 55, 11). ↩

244. See Waldschmidt, CASF, p. 138 n. 9. ↩

245. See Frauwallner, EV, p. 43. ↩

246. See Pande, SOB, pp. 98-106. See also A. Bareau, “La composition et les étapes de la formation progressive du Mahāparinirvāṇasūtra ancien,” in BEFEO LXVI (1979), pp. 45-103. ↩

247. See Rhys Davids, DB, Vol. II, p. 72. ↩

248. See Winternitz, HIL, p. 39. ↩

249. See Frauwallner, EV, pp. 157-59. ↩

250. Cf. D II 107-12 with A IV 305-313. ↩

251. See Pande, SOB, p. 102. The error may be connected with the insertion of the words setum katvāna at D II 89, 26. As Rhys Davids pointed out (Buddhist Suttas, SBE XI, Oxford 1881, p. 22 n. 1), this is a gloss upon taranti anṇavam saraṃ which has crept into the text, as the metre shows. The mistake must be an old one, as the words setum krtva occur in the corresponding verse at Divy 56, 8 and MPS 7. 9. ↩

252. There is some doubt about the date when Pātaliputra became the capital of Magadha, but it seems certain that it was after the time of the Buddha. See Malalasekera, DPPN, Vol. II, p. 178. ↩

253. Ernst Waldschmidt, Das Parinirvāṇasūtra, Teil I-III, Berlin 1950-51 (ADAW 1949 Nr. 1, 1950 Nr. 2 and 3). ↩

254. Ernst Waldschmidt, Drei Fragmente buddhistischer Sūtras, in NGAW 1948, pp. 49-91. The discovery of this fragment is of great interest, since the Chinese translation of the Dirghăgama was made from the Dharmaguptaka version (see Waldschmidt, CASF, p. 136). ↩

255. D II 186, 4-5. ↩

256. Mahāgovindīya-sūtra = Mvu III 197-224. Rhys Davids, DB, Vol. II, pp. 256-58 comments on some of the differences. Waldschmidt, CASF, p. 136 n. 3 refers to Klaus Hahlweg, Das Mahāgovinda-Sūtra. Eine vergleichende Analyse der indischen und chinesischen Versionen, diss. Munich 1954. Jones (Mvu tr., III, p. 219 n. 1) points out that the purpose of the Sanskrit version is to emphasise and illustrate an incident in the Buddha’s career. In both versions the story has the character of a Jātaka, for the Buddha identifies himself with Govinda in a previous birth. ↩

257. It occurs at S I 27-28. ↩

258. Ed. and tr. by Waldschmidt, CASF, pp. 149-62. ↩

259. See MW, s. vv. samaya and samāja. ↩

260. See E. Hultzsch, Inscriptions of Asoka, Oxford 1925, p. 51. ↩

261. The alternative suggestion, that Pāli samaya is to be derived from Sanskrit samāja seems less likely, since it could only have been transmitted through a written North-Western Prakrit form. ↩

262. They are discussed by Waldschmidt, CASF, p. 149. ↩

263. D II 289, 3. ↩

264. See Sp 28, 27. ↩

265. Vuttam idam bhante Bhagavatā Sakkapan̄he (S III 13, 5). ↩

266. See Waldschmidt, CASF, p. 138 n. 10. ↩

267. See Rhys Davids, DB, Vol. II, p. 337 n. 1. ↩

268. Bhagavati parinibbute dhātuvibhāge ca kate tattha tattha thüpesu patiṭthāpiyamānesu (Vv-a 297, 16-17). ↩

269. The word “comparatively” must be emphasised. The fact that versions of the suttas of the Dīgha-nikāya are found in the canons of other sects shows that the nikāya, as a whole, must have been composed within a century or so of the Buddha’s death. ↩

270. Yasmiṃ samaye subham vimokkham upasampajja viharati, subhan t’ eva tasmim samaye sañjānāti (D III 34, 20-21). ↩

271. See p. 161 below. ↩

272. For a discussion of these etymologies see U. Schneider, “Acht Etymologien aus dem Aggañña-Sutta,” in Asiatica (Festschrift Weller), Leipzig 1954, pp. 575-83. See also K. R. Norman, “Four etymologies from the Sabhiya-sutta,” in Balasooriya, BSWR, p. 183 n. 23. ↩

273. The metres include Vamśastha, Puspitāgrā, Rathoddhatā, Upasthitapracupita, Rucirā, Udgatā, and Pramitāksarā. ↩

274. See Warder, PM, § 135. ↩

275. Etā pana gāthā porāṇaka-therā Ānanda-therena ṭhapitā vaṇnagāthā ti vatvā gatā (Sv 922, 35-36). ↩

276. See Sv 959, 5. ↩

277. See p. 173 below. ↩

278. See p. 39 above. ↩

279. Helmuth Hoffmann, Bruchstūcke des Ātānāṭikasūtra, Leipzig 1939 (KST, V). ↩

280. See p. 107 below. ↩

281. See Valentina Stache-Rosen, Das Sangitisūtra und sein Kommentar Sangitiparyāya, Berlin 1968 (STT, IX). ↩

282. See D III 233, 23-234, 2. ↩

283. Rhys Davids (DB, Vol. III, p. viii) writes of an “appendix” at the end of each nikāya, but Pande (SOB, p. 113) points out that although suttas 33 and 34 are also together in the Chinese translation of the Dirghāgama, they are not the last suttas there. The Chinese version is based upon the Dharmaguptaka version, as Waldschmidt (CASF, p. 136) notes. ↩

284. See Kusum Mittal, Dogmatische Begriffsreihen im ālteren Buddhismus, I: Fragmente des Daśottarasūtra, Berlin 1957 (STT IV); D. Schlingloff, Dogmatische Begriffsreihen im ālteren Buddhismus, Ia: Daśottarasūtra IX-X, Berlin 1962 (STT IV a) ; B. Pauly, Fragments Sanskrits de Haute Asie (Mission Pelliot), Paris 1958. See also Ch. Tripāṭhī, “Die Einleitung des Daśottarasūtra,” in Indianisme et Bouddhisme (Festschrift Lamotte), Louvain 1980, pp. 353-58. ↩

285. See Pande, SOB, p. 78. ↩

286. See ibid., p. 116 n. 1. ↩

287. It occurs in the Saṃghabhedavastu of the Mūla-Sarvāstivādin canon, as noted above (p. 24). ↩

288. Ed. V. Trenckner, PTS London, Vol. I, 1888; R. Chalmers, Vol. II, 1896; Vol. III, 1899. Tr. Lord Chalmers, Further Dialogues of the Buddha, SBB London, 1926-27; I. B. Horner, Middle Length Sayings, PTS London, Vol. I, 1954; Vol. II, 1957; Vol. III, 1959; (German) K. E. Neumann, Reden Gotamo Buddho’s aus der mittleren Sammlung (Majjhimanikāyo) des Pali-Kanons übers., 3 vols., Leipzig 1896, 1900, 1902. For an analysis of the text see Pande, SOB, pp. 116-79. ↩

289. See Horner, MLS, I, pp. x-xv. ↩

290. Mahā-Rāhulovādasutta (62) and Cūla-Rāhulovādasutta (147). ↩

291. As suggested by Horner (MLS, I, p. xi) following Chalmers. ↩

292. A comparable account of Gotama’s visits to Ārāda Kālāma and Udraka Rāmaputra is found at Mvu II 118, 1-120, 16. ↩

293. See Pande, SOB, p. 122. ↩

294. See p. 40 above. ↩

295. Saccavibhangasutta (M III 248-252). ↩

296. See K. R. Norman, “A note on attā in the Alagaddūpama-sutta,” in Studies in Indian Philosophy, Ahmedabad 1981, pp. 19-29. ↩

297. parinibbuto kho … etarahi so Bhagavā arahaṃ sammāsambuddho (M II 90, 16-17=162,24-25) ↩

298. Th 769-88. ↩

299. Ja (9) I 137-39. ↩

300. Th 866-91. ↩

301. See M I 144, 18-22. ↩

302. See K. R. Norman, “The Buddha’s view of devas,” in Beitrāge zur Indienforschung: Ernst Waldschmidt zum 80. Geburtstag gewidmet, Berlin 1977, pp. 32936. ↩

303. See K. R. Norman, “Devas and Adhidevas in Buddhism,” in JPTS 1981, pp. 145-55. The corruptions which have led to the loss of the word adhideva in both the Sangāravasutta and the Kannakatthalasutta must be very old. ↩

304. By Horner, MLS, I, p. ix n. 1. ↩

305. Cf. Vin II 254-55. ↩

306. As suggested by Mrs C. A. F. Rhys Davids in her review of M II in JRAS 1902, p. 475. ↩

307. Pande, SOB, p. 117. ↩

308. See K. R. Norman, “The pratyeka-buddha in Buddhism and Jainism,” in Papers on Buddhology, Curzon Press, London 1982. ↩

309. Cf. Ja I 52-53. ↩

310. acira-parinibbute Bhagavati (M III 7, 14-15). ↩

311. sappatisarañā mayam … dhammapatisaranā (M III 9, 23-24). ↩

312. Cf. Utt XIX 47-72. ↩

313. See Mrs C. A. F. Rhys Davids, rev. of R. Chalmers, Majjhima-nikāya, Vols II and III, in JRAS 1902, p. 475. ↩

314. See p. 40 above. ↩

315. See Waldschmidt, CASF, p. 136. ↩

316. See P. V. Bapat (ed.), The Majjhima Nikāya (1. Mūla-paṇṇāsakaṃ), Nalanda 1958, p. xxiv. ↩

317. For the Bimbasārasūtra, represented in the Chinese Madhyamāgama, see Waldschmidt, CASF, p. 138. See also E. Waldschmidt, “The Rāṣṭrapālasūtra in Sanskrit remnants from Central Asia,” in Indianisme et Bouddhisme (Festschrift Lamotte), Louvain 1980, pp. 359-74, and The Varnaśatam: an eulogy of one hundred epithets of Lord Buddha spoken by the Gṛhapati Upāli(n), in NGAW 1979 Nr. 1. A number of other Sarvāstivādin Madhyamāgama texts are listed in E. Waldschmidt, Sanskrithandschriften aus den Turfanfunden, IV, 1980, pp. 20-73. ↩

318. Ed. L. Feer, 5 vols., PTS London, 1884-98. Tr. C. A. F. Rhys Davids, Kindred Sayings, PTS London, Vol. I, 1917; Vol. II, 1922; F. L. Woodward, Vols. III-V, 1925, 1927, 1930; (German) W. Geiger, Saṃyutta-Nikāya, Vol. I, Munich 1930; Vol. II, Munich 1925. For an analysis of the text see Pande, SOB, pp. 180-229. ↩

319. Sp 18, 9-10. ↩

320. See Winternitz, HIL, p. 56. ↩

321. Winternitz, HIL, p. 57 n. 1, quoting Mrs Rhys Davids. ↩

322. See M. Przyluski, “Le Parinirvāṇa et les funérailles du Buddha,” in JA 1918, pp. 485-526, quoted by Pande, SOB, p. 189. ↩

323. Despite Feer’s statement (S I, p. xv) that all the verses have been gathered together in the Sagāthavagga, there are some in other vaggas. He later (S IV, p. 9) comments on the inconvenience of having another Sagāthavagga in the Vedanāsamyutta (36). ↩

324. The five vaggas are divided into samyuttas, which are in turn divided into vaggas. Feer (S I, p. viii) comments upon the inconvenience of this system. ↩

325. E.g. S I 40, 4 and 40, 6 have parallels at Utt XIV 22-23. ↩

326. Thi 182-203. ↩

327. See Norman, EV II, pp. xxvi-xxvii. ↩

328. See Waldschmidt, CASF, pp. 144-47. ↩

329. Th 1209-79. ↩

330. Waldschmidt, CASF, pp. 147-48. ↩

331. S I 218-20. See Waldschmidt, CASF, p. 138. ↩

332. S I 50. See Ernst Waldschmidt, “Buddha frees the disc of the moon (Candrasūtra),” in BSOAS 33, pp. 179-83. ↩

333. See S. Lévi, “Textes sanscrits de Touen-houang,” in JA 1910, pp. 438-40 (tr. J. M. Cooper, “A fragment of the Nidānasūtra,” in PBR, 5, 3 (1980), pp. 5358); E. Waldschmidt, “Sūtra 25 of the Nidānasamyukta,” in BSOAS 20, 1957, pp. 569-79; Chandrabhāl Tripāṭhī (ed.), Fünfundzwanzig Sūtras des Nidānasamyukta, Berlin 1962 (STT, VIII). ↩

334. For the meaning “endless” for anamatagga see T. Burrow, The problem of shwa in Sanskrit, Oxford 1979, p. 42. ↩

335. S II 189-90. ↩

336. As noted by Feer, S III, p. vii. ↩

337. S IV 40-41. See Ernst Waldschmidt, Das Upasenasūtra, Ein Zauber gegen Schlangenbiß aus dem Samyuktāgama, NGAW 1957, pp. 27-44. ↩

338. E. Waldschmidt, “Mahāmaudgalyāyana’s sermon on the letting-in and not letting-in (of sensitive influences),” in JIABS, 1, 1978, pp. 25-33. ↩

339. In the rubric to the first vagga it is called Nibbānasamyutta (S IV 361, 25). ↩

340. Yathā asañkhatam tathā vitthāretabbam (S IV 368, 26-27 = 373, 18). ↩

341. E.g. vitthāretabbam (S V 308, 27=309,13 ); yad api satipatthānam tad api vitthāretabbam (308, 13); yathā Maggasamyuttam evam Jhānasamyuttam vitthāretabbam (310,7-8). ↩

342. Vin I 10, 10-12, 18. ↩

343. See Waldschmidt, CASF, p. 136. ↩

344. Bunyiu Nanjio, Catalogue of the Chinese translation of the Buddhist Tripitaka, Oxford 1883, p. 135, quoted by Pande, SOB, p. 181. ↩

345. Ed. R. Morris, The Añguttara-Nikāya, PTS London, Vol. I, 1885 (Second edition, revised by A. K. Warder, 1961); Vol. II, 1888; E. Hardy, PTS London, Vols. III-V, 1897, 1899, 1900; M. Hunt, Vol. VI (Indexes), 1910. Tr. F. L. Woodward, The Gradual Sayings, PTS London, Vol. I, 1932; Vol. II, 1933; Vol. V, 1936; E. M. Hare, Vol. III, 1934; Vol. IV, 1935; (German) Bhikkhu Nāṇatiloka, Reden des Buddha aus dem Añguttara-Nikāya übers., 5 vols., Munich 1911-23. For an analysis of the text see Pande, SOB, pp. 230-47. ↩

346. Sp 18, 10-12. ↩

347. “There are, in fact, at most about 2,344 suttas in the Añguttara” according to Hardy (A V, p. vi). ↩

348. See Hardy’s comment at A IV, p. iii. ↩

349. Mil 362,17 ; 392,2. ↩

350. A IV 407. ↩

351. A V 326. ↩

352. See Hardy’s comment at A V, p. vii, and his list of composite suttas at A V 421-22. ↩

353. A IV 460. ↩

354. A V 120-21. ↩

355. A IV 274-79. ↩

356. Vin II 253-56. ↩

357. A IV 308-13. ↩

358. D II 107-109. ↩

359. Cf. D II 109-110 and A IV 307-308. ↩

360. Mrs Rhys Davids points out that the four truths occur in the catukka-nipāta, but under the titles of dukkha and loka, not of ariya, as elsewhere. See A VI, p. viii n. 2 . ↩

361. The eight-fold path, with the addition of sammāñannam and sammāvimutti, occurs in the dasaka-nipāta (A V 223). ↩

362. See Pande, SOB, p. 232. ↩

363. As suggested by Mrs Rhys Davids (A VI, p. viii). ↩

364. The vagga is entitled “Finger-snap” (accharā-saighāta), but this applies from the third sutta onwards. The first two suttas deal with the luminous mind (pabhassaram cittam), which is the subject of the last two suttas of the previous vagga. The mechanical division of suttas into vaggas of ten has led to connected suttas becoming separated. ↩

365. See A V 304-308. ↩

366. See Pande, SOB, p. 231. ↩

367. As noted by Hardy (A V, p. vii n. 1). ↩

368. Quoted (from Minayeff) by Pande, SOB, p. 231. ↩

369. Pande, SOB, p. 231. ↩

370. A V 309-10. ↩

371. A V 360-61. ↩

372. See Waldschmidt, CASF, p. 137. ↩

373. See Waldschmidt, CASF, p. 169. ↩

374. Discussed by Waldschmidt, CASF, pp. 169-74. ↩

375. Chandrabhāl Tripāṭhī, Ekottarāgama-Fragmente der Gilgit-Handschrift, Reinbek 1981. ↩

376. Ed. H. Smith, PTS London, 1915. Tr. Bhikkhu Nāyamoli, The Minor Readings, PTS London, 1960. ↩

377. See p. 9 above. ↩

378. P. V. Bapat and A. Hirakawa (tr.), Shan-Chien-P’i-P’o-Sha, Poona 1970, p. 11 n. 22. ↩

379. Sn pp. 46-47; vv. 222-38; vv. 143-52. ↩

380. Pv 14-25 (= I. 5). ↩

381. Some of the texts are used as parittas. See p. 174 below. ↩

382. Ed. V. Fausboll, Hauniae (Copenhagen) 1855; second ed., London 1900; S. Sumangala Thera, PTS London, 1914; ed. and tr. S. Radhakrishnan, London 1950. Tr. F. Max Müller, London 1881 (SBE X. 1) ; (Latin) V. Fausboll, Hauniae 1855; (German) L. von Schroeder, Worte der Wahrheit, Leipzig 1892; K. E. Neumann, Der Wahrheitpfad, Leipzig 1893; R. O. Franke, Dhamma-Worte, Jena 1923. ↩

383. Winternitz, HIL, p. 83. ↩

384. Cf. Dhp 404 with Utt XXV 28; and Dhp 223 with MBh V 39. 58. ↩

385. E.g. Dhp 228-30= Udāna-v XXIX 46-48= G Dhp 240-42. ↩

386. As Dhp 158 contains both the word attānam and the word pandito, it is appropriate to the Attavagga in the Dhp and to the Panditavarga in the G Dhp (227). ↩

387. See the assessment by Brough, G Dhp, p. xvii. ↩

388. F. Bernhard (ed.), Udānavarga, Vol. I, Göttingen 1965. See also L. Schmitthausen, ,,Zu den Rezensionen des Udānavargab”, in WZKSO 14, pp. 47-124. ↩

389. J. Brough (ed.), The Gāndhārī Dharmapada, London 1962. ↩

390. Brough (G Dhp, p. 45) considers the Dharmaguptakas and the Kāsyapīyas as eligible claimants to the text. ↩

391. See Brough, G Dhp, p. 23. ↩

392. See ibid., p. 25. ↩

393. See ibid., p. 28. This does not necessarily imply that either text was known to the redactors of the other. ↩

394. Mvu III 434-36. ↩

395. Mvu III 421-23. ↩

396. Mvu II 212, 19-20; III 156, 16-17. ↩

397. Ed. N. S. Shukla, The Buddhist Hybrid Sanskrit Dharmapada, Patna 1979: G. Roth, “Particular features of the language of the Arya-Mahāsāṃghika-Lokottaravādins and their importance for early Buddhist tradition,” in Bechert, LEBT, pp. 78-135. ↩

398. See Shukla, op. cit. (in n. 188), p. vi. ↩

399. See Brough, G Dhp, pp. 35-38. ↩

400. Ed. P. Steinthal, PTS London, 1885. Tr. D. M. Strong, London 1902; F. L. Woodward, Verses of Uplift, SBB London, 1948; (German) K. Seidenstücker, Augsberg 1920. ↩

401. Sp 28, 18-19. ↩

402. Ud-a 3, 23-28. ↩

403. Ud-a 3, 12. ↩

404. Vin I 1-3. ↩

405. It occurs at A IV 354-58. ↩

406. Ud-a 5, 16-17. ↩

407. Ud 15, 1-5. See Alsdorf, AS, pp. 18-19. ↩

408. = Pet 26,9-27,4 = Nett 156, 22-157, 14. For the better readings of these versions see Bhikkhu Nāyamoli, The Guide, PTS London, 1962, p. 207 n. 843/1-6, ↩

409. Ed. E. Windisch, PTS London, 1889. Tr. J. H. Moore, Iti-vuttaka or Sayings of Buddha, New York 1908; F. L. Woodward, Itivuttaka: as it was said, SBB London, 1948; (German) K. Seidenstücker, Leipzig 1922. ↩

410. The phrase appears at the beginning and end of every sutta until 79(=3.3 .10). Thereafter it appears only intermittently. That not all sayings were uttered by the Buddha can be deduced from the fact that the last sutta (112=4.13) refers to the Buddha in both the third and second persons. ↩

411. Sp 28, 20-21. ↩

412. See p. 16 above, and BHS Dictionary, s.v. itivyttaka. ↩

413. Such phenomena imply the work of a redactor, and not a very clever one. ↩

414. 4.2 = A II 27 ; 4.6= A II 10 ; 4.7= A I 132= II 70 ; 4.9= A II 26 ; 4.11 = A II 13 ; 4.12= A II 15 ; 4.13= A II 24. ↩

415. With 4.5 cf. Pp 54. ↩

416. See F. L. Woodward, It tr., p. ix. ↩

417. See K. Watanabe, “A Chinese collection of Itivuttakas,” in JPTS 19061907, pp. 44-49. ↩

418. Watanabe states, op. cit. (in n. 208), p. 45, that the original of the Chinese translation was a Sanskrit text, because the texts brought to China were alleged to be “in the Fan language.” “Fan,” however, means “Indian (language)” and may refer either to Sanskrit or Prakrit. See K. R. Norman, “The role of Pāli in early Sinhalese Buddhism,” in Bechert, BCSRS, p. 39. ↩

419. See Watanabe, op. cit. (in n. 208), p. 49. ↩

420. Ed. V. Fausbell, PTS London, 1884; Dines Andersen and Helmer Smith, PTS London, 1913. Tr. V. Fausbell, The Sutta-nipāta, London 1881 (SBE 10.2); Lord Chalmers, Buddha’s Teachings, Harvard 1932; E. M. Hare, Woven Cadences of early Buddhists, SBB London, 1945; (German) Nyanaponika, Konstanz 1955. ↩

421. See Warder, PM, § 303. ↩

422. See pp. 67 and 69 below. ↩

423. See M. Anesaki, “Sutta-nipāta in Chinese,” in JPTS 1906-1907, pp. 50-51. Winternitz (HIL, p. 92 n .3 ) states that the Pārāyanavagga also occurs in the Chinese Tipitaka, whereas Anesaki mentions only quotations from it there. ↩

424. So bhikkhu jahāti orapāram urago jinnam iva tacam purānam (Sn 1-17). ↩

425. It, or much of it, occurs as a separate varga in Udāna-v, G Dhp (in Bhikṣuvarga), and P Dhp (in an unnamed varga which is not the Bhikṣuvarga). In Udāna-v, part is in the Bhikṣuvarga, but the flower verses are in the Puṣpavarga. ↩

426. Brough has pointed out (G Dhp, p. 200) that the Uragasutta contains nine compound verbs forming their absolutives in -tvā. In the other eleven suttas of the vagga there is only one further example, viz. virājetvā in 139 and 171. ↩

427. The verses fall into pairs, with the Buddha making a punning reply to Dhaniya in the second verse of each pair. The fact that verses 20 and 21 seem unconnected suggests that two verses have been lost from the sutta. ↩

428. “iti Sātāgiro” ti ādi sanggītikārehi [vuttam] (Pj II 193, 27-28). ↩

429. Eko care khaggavisānakappo (Sn 35-44, 46-75). The Niddesa makes it clear that the comparison is with the horn of the rhinoceros: yathā khaggassa nāma visānam ekam hoti adutiyam, evam eva so paccekasambuddho takkappo tassadiso tappatibhāgo (Nidd II [Ne] 248, 6-8). Doubtlessly the singularity of the fact that the Indian rhinoceros, of all the horned animals known to the Indians, had only one horn, impressed them and resulted in its use as an example of “one-ness.” ↩

430. Ten’ āha so paccekasambuddho (Nidd II [Ne] 248, 15). ↩

431. Paccekabuddhehi jinehi vuttā gāthā (Ap 14, 1-2). ↩

432. See p. 48 above. ↩

433. Sarvā khaḍgavisānagāthā vistarena kartavyā. paṃcānāṃ pratyekabuddhaśatānām eka-ekā gāthā (Mvu I 359, 16-17). ↩

434. See the discussion of the prose passages in Pande, SOB, pp. 52-53. ↩

435. Ito param sangitikārā desanam thomentā “idam avoca Bhagavā” ti ādim āhamsu (Pj II 603, 28-29 ad Sn p. 218). ↩

436. See Alsdorf, AS, pp. 15-16. ↩

437. See p. 58 above. ↩

438. See p. 58 above. ↩

439. See Warder, IKL, II, pp. 28-29. ↩

440. Mvu II 198-99; II 238-40. ↩

441. Mvu III 394-401. ↩

442. See K. R. Norman, “Four etymologies from the Sabhiya-sutta,” in Balasoortya, BSWR, pp. 173-84. ↩

443. See pp. 34 and 47 above. ↩

444. M II 146, where the Ee merely refers to Sn. ↩

445. Th 818-41. ↩

446. It is interesting to note that there appears to be an interpolation in the verses, for the commentator notes that there was no old commentary upon Sn 677-78: avasāne gāthādvayaṃ eva pana Mahā-atṭhakathāyaṃ vinicchitapāṭhe n’ atthi (Pj II 477, 13-14). ↩

447. Sn 679-98. ↩

448. Mvu II 31-43; III 382-89. ↩

449. See Warder, PM, p. 213 n. 2. ↩

450. See L. de La Vallée Poussin, rev. of Mrs C. A. F. Rhys Davids, Dukap, I, in JRAS 1907, p. 453. ↩

451. Sabbān’ eva aṭthakavaggikāni sarena abhāsi (Vin I 196, 36); soḷasa aṭthakavaggikāni sabbān’ eva sarena abhani (Ud 59, 23). It is referred to by name at S III 9,19=12,20, where Sn 844 is quoted from the Māgandiyasutta. See also Kogen Mizuno, “Dharmapadas of various Buddhist schools,” in Narain, SPB, pp. 261-62. ↩

452. See Pande, SOB, pp. 53-54. ↩

453. Just as the Uragavagga is called after its first sutta, the Uragasutta. ↩

454. Ud 59, 23. ↩

455. See p. 64 above. ↩

456. See A. F. Rudolf Hoernle, “The Sutta Nipāta in a Sanskrit version from Eastern Turkestan,” in JRAS 1916, pp. 709-32 and JRAS 1917, p. 134. Fragments from the Sanskrit Pārāyaṇasūtra are, however, listed in E. Waldschmidt, Sanskrithandschriften aus den Turfanfunden, IV, 1980, pp. 236-38. Verses from a Sanskrit equivalent of the Suttanipāta are quoted in the Abhidharmakośa and the Bodhisattvabhūmi, as noted by S. LĒvi, “Sur la récitation primitive des textes bouddhiques,” in JA 1915, p. 414. ↩

457. See p. 63 above. ↩

458. See Alsdorf, AS, pp. 16-18. ↩

459. See p. 85 below. ↩

460. See Hoernle, JRAS 1916, pp. 718-19 and P. V. Bapat, The Arthapada Sūtra, Santiniketan 1951. ↩

461. The Munisutta (Sn 207-21) is probably the Munigāthā recommended for study by Asoka. See Winternitz, HIL, p. 607. ↩

462. Sn 976-1031. ↩

463. Tam attham gahetvā āyasmā Ānando sanḡītikāle Pārāyanavaggassa nidānaṃ āropento imā gāthāyo abhāsi (Pj II 580, 29-30). ↩

464. Ito param sanḡ̄tikārā desanaṃ thomentā “idam avoca Bhagavā” ti ādi āhamsu (Pj II 603, 28-29). ↩

465. Maggo so pārangamanāya, tasmā Pārāyanam (Sn 1130). ↩

466. Sn 1038 (Ajitapañha) at S II 47, 11; 1042 (Metteyyapañha) at A III 399, 22= 401,24 ; 1048 (Punṇakapañha) at A I 133, 6= II 45, 35; 1106-7 (Udayapañha) at A I 134,9 . ↩

467. W. B. Bollée, The pādas of the Suttanipāta (Studien zur Indologie und Iranistik, Monographie 7), Reinbeck 1980, p. 33, queries whether khattiyā brāhmaṇaā devatānaṃ yaññam at Sn 1043-45 is metrical. ↩

468. See N. A. Jayawickrama, A critical analysis of the Pāli Sutta Nipāta illustrating its gradual growth, Ph.D. dissertation London, published in parts in the University of Ceylon Review, January 1948-April 1951 (Vol. VI No. 1-Vol. IX No. 2). ↩

469. Ed. E. R. Gooneratne, PTS London 1886; N. A. Jayawickrama, PTS London, 1977; tr. J. Kennedy, SBB London, 1942; I. B. Horner, SBB London, 1974. ↩

470. There are no verses in ganacchandas metres, and those in mātrācchandas metres are probably late. See WARder, PM, § 193. ↩

471. Mvu II 191-95. ↩

472. See H. Bechert, “On a fragment of Vimānāvadāna, a canonical Buddhist Sanskrit work,” in Cousins, BSIBH, pp. 19-25. ↩

473. See Alsdorf, AS, pp. 80-88. ↩

474. E.g. Guttilavimāna (33) includes the two verses of the Guttilajātaka (Ja II 252), which in turn refers to the Vv in its prose story and quotes from it. The Agāriyavimānavatthu (65), the Dutiya-agāriyavimānavatthu (66), and the Serissakavimānavatthu (84) include verses which occur in the Vidhurapanditajātaka. ↩

475. Ed. J. Minayeff, PTS London, 1888; N. A. Jayawickrama, PTS London, 1977; tr. H. S. Gehman, SBB London, 1942. ↩

476. Satthu parinibbānato vassasatadvayassa accayena Suratthavisaye Pingalo nāma rājā ahosi (Pv-a 244, 15-16). ↩

477. The first and last half-verses of the Kūṭavinicchayikapetavatthu ( 34= III.9) are to be compared with verses 28 and 32 of the Kimchandajātaka ( 511= Ja V 9-10). The Kaṇhapetavatthu ( 18= II.6) closely resembles the Ghaṭajātaka ( 454= Ja IV 84-87). The verses of the Uragapetavatthu ( 12= I.12) are the same as those of the Uragajātaka ( 354= Ja III 164-67). The verses of the Gonapetavatthu (I.8) closely resemble those of the Sujātajātaka ( 352= Ja III 156-57). ↩

478. See H. Bechert, op. cit. (in n. 263), pp. 19-20 and E. Waldschmidt, Sanskrithandschriften aus den Turfanfunden, IV, 1980, pp. 227-33. ↩

479. Ed. H. Oldenberg, PTS London, 1883; 2nd edition with appendices by K. R. Norman and L. Alsdorf, PTS London, 1966. Tr. C. A. F. Rhys Davids, Psalms of the Brethren, PTS London, 1913; K. R. Norman, Elders’ Verses I, PTS London, 1969; (German) K. E. Neumann, Die Lieder der Mönche und Nonnen Gotamo Buddhos, Berlin 1899. ↩

480. See Norman, EV I, pp. xxiii-xxv. ↩

481. E.g. there are verses ascribed to Kimbila at Th 118 and 155-56. ↩

482. E.g. Revata’s verses are at Th 645-58; Th 42 is also ascribed to him under his nickname of Khadiravaniya. ↩

483. Tattha tattha bhikkhūnam ovāda-dāna-vasena attano patipatti-dīpanādi-vasena ca bhāsita-gāthā ekajjham katvā, anukkamena Khuddaka-nikāyam sangāyana-kāle Theragāthāsu sangitim āropento (Th-a III 113, 7-10). ↩

484. Ayaṃ sangitikārehi vuttagāthā (Th-a III 14, 6 ad Th 720). ↩

485. See Norman, EV I, p. 132. ↩

486. Thi 204-12. See Norman, EV I, p. 183. ↩

487. Th 825. ↩

488. Ayaṃ paṭhama-mahāsangīti-kāle āyasmatā Ānandena tesam therānam thomanattham bhāsitā gāthā ādi (Th-a I 4, 18-19). ↩

489. Vassa-sata-parinibbute Bhagavati, Vesālikesu Vajji-puttakesu dasa-vatthūni paggayha thitesu … thero … imā gāthā bhananto aññam vyākāsi (Th-a II 123, 5-19). ↩

490. Ettha ca Bindusārarañño kāle imassa therassa uppannattā tatiya-sangitiyam imā gāthā sanḡītā ti veditabbam (Th-a II 165, 32-33). ↩

491. [Vītasoka] Dhammāsoka-rañño kanitṭha-bhātā hutvā nibbatti (Th-a II 45, 9); [Ekavihāriya] Dhammāsoka-rañño kanitṭha-bhātā hutvā nibbatti (Th-a II 227, 2728). See Norman, EV I, p. xxvii. ↩

492. With Th 1209-79 cf. S I 185-96. ↩

493. See C. A. F. Rhys Davids, Psalms of the Brethren, PTS London 1913, pp. 423-24. ↩

494. Sahassam honti tā gāthā tīni saṭ̣̣hisatāni ca (Th p. 115). ↩

495. Th p. 36. ↩

496. For a survey of the metres in Th see Norman, EV I, pp. xxxiii-xxxiv. ↩

497. See Th 96. ↩

498. A. Weber (ed.), Über das Saptaśatakam des Hāla, Leipzig 1881. ↩

499. See S. Lienhard, “Sur la structure poétique des Theratherīgāthā,” in JA 263, 1975, pp. 375-96. ↩

500. H. Bechert, Bruchstücke buddhistischer Verssammlungen: 1. Die Anavataptagāthā und die Sthaviragāthā, Berlin 1961 (STT VI). ↩

501. Ed. R. Pischel, PTS London, 1883; 2nd edition with appendices by K. R. Norman and L. Alsdorf, PTS London, 1966. Tr. C. A. F. Rhys Davids, Psalms of the Sisters, PTS London, 1909; K. R. Norman, Elders’ Verses II, PTS London, 1971; (German) K. E. Neumann, Die Lieder der Mönche und Nonnen Gotamo Buddhos, Berlin 1899. ↩

502. Gāthā satāni cattāri asīti puna cuddasa (Thī, p. 174). ↩

503. See Norman, EV II, p. xxiv. ↩

504. See p. 51 above and Norman, EV II, pp. xxvi-xxvii. ↩

505. Thī 252-70. ↩

506. See Alsdorf, Thī (Appendix II), pp. 233-50. ↩

507. Thī 400-47. ↩

508. Thī 448-522. ↩

509. With Thī 496-99 cf. S II 178-93. ↩

510. See pp. 36 and 43 above. ↩

511. See E. Hultzsch, Inscriptions of Asoka, Oxford 1925, p. 165. ↩

512. See C. A. F. Rhys Davids, Psalms of the Sisters, PTS London, 1909, p. 115 n. ↩

513. Ed. V. Fausbell, 6 vols. London, 1877-96. Tr. T. W. Rhys Davids, Buddhist Birth Stories, London 1880; E. B. Cowell et al., The Jātaka, 6 vols., London, 1895-1907; (German) J. Dutoit, Leipzig, 1908-21. ↩

514. See Bollée, KJ, pp. v-vi. ↩

515. See H. Lüders, Bhārhut und die buddhistische Literatur, Leipzig 1941, p. 144 . ↩

516. See p. 15 above. ↩

517. See L. Alsdorf, “The impious brahman and the pious candāla,” in Cousins, BSIBH, pp. 9-13. ↩

518. See Jones, Mvu tr., II, pp. x-xi. ↩

519. Ja I 2-47. ↩

520. Ja I 47-77. ↩

521. Ja I 77-94. ↩

522. Cf. Bv II 1-188. ↩

523. See Winternitz, HIL, p. 164. ↩

524. See Lal 83-87. ↩

525. Fausbell’s edition contains 547 Jātakas, but some of those in the Mahānipāta contain more than one narrative, and the total number of Jātaka stories is difficult to establish. The reference to 500 Jātakas in the Culla-niddesa is probably nothing more than a “round number,” and should not be taken as evidence that at the time of the composition of that text the Jātaka collection was smaller than that we possess now: Bhagavā pañca jātakasatāni bhāsanto attano ca paresam ca atitam ādisati (Nidd II [Ne] 173, 23). There is, however, evidence that a collection of 550 Jātakas was known in Burma, and the names and order of the Jātakas found in some traditions differ from Fausbell’s edition. See G. H. Luce, “550 jātaka in old Burma,” in Artibus Asiae, 19, 3-4, pp. 291-307, and G. Martini, “Les titres des jātaka dans les manuscrits pāli de la Bibliothèque nationale de Paris,” in BEFEO 51 (1963), pp. 79-93. ↩

526. See Winternitz, HIL, p. 126. ↩

527. See H. Lüders, Bhārhut und die buddhistische Literatur, Leipzig 1941. ↩

528. See M. Cone and R. F. Gombrich, The perfect generosity of Prince Vessantara, Oxford 1977, pp. xxxiii-xxxiv. ↩

529. Ja I 334, 27. ↩

530. Hi-kāro pan’ ettha nipāta-mattam eva (Ja I 342, 17). This seems to be a comment upon the strange form bhayāhi (336,17), but is omitted by Be and Ne. ↩

531. Vipariyāso vattati, na idha-m-atthi (344, 23-24). This has been overlooked by Fausbell, who includes it in the prose. ↩

532. Parinibbute pana Bhagavati samgītikārakā usabhārukkhādīni tīni padāni atṭhakatham āropetvā lābünī ti ādīni pañca padāni ekam gātham katvā eka-nipāta-pālim āropesum (345, 14-16). ↩

533. See L. Alsdorf, “The ākhyāna theory reconsidered,” in JOI(B) 13, p. 200 n. 2. ↩

534. Evam akkhāyati ti (Ja V 412, 14). ↩

535. See Alsdorf, AS, pp. 25-28. ↩

536. W. B. Bollée (ed. and tr.), Kuṇālajātaka, SBB London, 1970. ↩

537. There are also prose passages at Ja II 147, 2-3** and 147, 10-11**, as Bollée (KJ, p. vi n.) points out, which are commented upon. These two short passages are, however, taken from elsewhere in the canon (Vin II 110, 16-18= A II 73, 6-8), and were therefore canonical prose before they were included in the Jātaka. ↩

538. As pointed out by Alsdorf, JOI(B) 13, p. 200 n. 2. ↩

539. See Bollée, KJ, pp. 166-72. ↩

540. See A. Mette, “Vedhas in Lalitavistara und Divyāvadāna,” in WZKS 17, 1973, pp. 21-42. ↩

541. Ja I 285-315. ↩

542. See Winternitz, HIL, p. 145 n. 2. ↩

543. See Utt XII and XIII. ↩

544. Ja III 381, 16-18 and 380, 6-21. ↩

545. See H. Jacobi, Ausgewāhlte Erzāhlungen in Māhārāṣṭrī, Leipzig 1886, p. 34. ↩

546. See L. Alsdorf, “Das Sivijātaka (499) : Ein Beitrag zu seiner Textgeschichte,” in Heesterman, Pratid., pp. 478-83. ↩

547. See L. Alsdorf, “Das Bhüridatta-Jātaka. Ein antibrahmanischer NāgaRoman,” in WZKS 21, 1977, pp. 25-55. ↩

548. See L. Alsdorf, “Das Jātaka vom weisen Vidhura,” in WZKS 15, 1971, pp. 23-56. ↩

549. M. Cone and R. F. Gombrich (tr.), The perfect generosity of Prince Vessantara, Oxford 1977. ↩

550. Winternitz, HIL, p. 152. ↩

551. See L. Alsdorf, “Bemerkungen zum Vessantara-Jātaka,” in WZKSO 1, 1957, pp. 1-70. ↩

552. See p. 80 above. ↩

553. See Cone and Gombrich, op. cit. (in n. 340), pp. xxxi-xxxii. ↩

554. At the end of the Ghaṭikārasutta of the Majjhima-nikāya (81), the Buddha identifies himself with the young brahman Jotipāla (aham tena samayena Jotipālo mānavo ahosim [M II 54, 18-19]) in the story of past time (bhütapubbam) he had told. This makes the story technically a Jātaka. ↩

555. E.g. the Punyavantajātaka (Mvu III 33-41). ↩

556. Dip V 37. ↩

557. L. de La Vallée Poussin and E. J. Thomas (ed.), Mahāniddesa, PTS London, Part I 1916, Part II 1917. ↩

558. W. Stede (ed.), Cullaniddesa, PTS London, 1918. ↩

559. Dip VII 41; Mhv V 278. ↩

560. Dip V 37. ↩

561. Nidd-a I 1, 14-15. ↩

562. By A. P. Buddhadatta, Nidd-a I, p. vii. ↩

563. Nidd I 446, 29-447, 6 (ad Sn 955). ↩

564. E.g. M I 111, 31-113, 7 and S III 7, 1-9, 8. ↩

565. See pp. 63 and 65 above. ↩

566. Ap 8-13. ↩

567. Mvu I 357-59. ↩

568. E.g. givanto ti givanto kujjanto nadanto saddam karonto. athavā kivanto katī kitthakā kivatakā kivabahukā (Nidd I 467, 5-7 ad Sn 959: givanto tattha bheravā). ↩

569. PJ II 572, 12-14 ad Sn 959 reads kivanto in the lemma and explains it as kittakā. It quotes khivanto (glossed as kūjanto) as a variant reading, but dismisses this as inappropriate in the context (na pana pubbenāparam sandhiyati). ↩

570. As noted by M. H. Bode, rev. of Mahāniddesa Part I, in JRAS 1918, p. 575. [^2-3-362:] Cf. Nidd I 97, 1-3 ad Sn 793 (vikappayeyya) with 97, 28-30 ad Sn 794 (kappayanti). ↩

571. Nidd II, p. x. ↩

572. Nidd I 268, 1 ad Sn 868. ↩

573. Nidd I 268, 10-24. ↩

574. Nidd I 197, 1-201, 2. ↩

575. E.g. Nidd I 139, 26-140, 7 ad Sn 814 (icc āyasmā Tisso Metteyyo). ↩

576. See Nidd II (Ne) 206-208 ad Sn p. 218. ↩

577. Tam attham gahetvā āyasmā Ānando sañgitikāle Pārāyanavaggassa nidānaṃ āropento imā gāthāyo abhāsi (Pj II 580, 29-30). ↩

578. Avasāne gāthādvayaṃ eva pana Mahā-ạthakathāyaṃ vinicchitapāṭhe n’ atthi (Pj II 477, 13-14 ad Sn 677-78). ↩

579. Nidd I 154-55=415. ↩

580. See S. Lévi, “Ptolemée, le Niddesa et la Bṛhatkathā,” in Études asiatiques, Vol. II, Paris 1925, pp. 1-55. ↩

581. See Himansu Bhusan Sarkar, “The date of the Pali Niddesa and its implication for the history of South-East Asia,” in JAS(B), 19, 1977, pp. 44-45. ↩

582. Sp 695-96. 375 A.: C. Taylor (ed.), PTS London, Vol. I 1905, Vol. II 1907. Tr. Bhikkhu Nāṇamoli, The path of discrimination, PTS London, 1982. There is an introduction to this translation by A. K. Warder which deals at length with the problems of the composition and dating of this text. ↩

583. Paṭis-a 1, 18. 377 See Warder:, Paṭis tr., p. xxxiii. ↩

584. See ibid., pp. vii-viii. ↩

585. Paṭis I 205-214. See Arnold, Paṭis II, p. iv. ↩

586. Paṭis I 1-3. ↩

587. Paṭis II p. iv. ↩

588. See Warder, Paṭis tr., p. xxxiii. ↩

589. Dip V 37. ↩

590. See BHS Dictionary, s.v. pratisaṃvid. ↩

591. See p. 9 above. ↩

592. Paṭis I 4-134. ↩

593. See Warder, Paṭis tr., p. xxxiii, and Arnold, Paṭis II, p. iv. ↩

594. See Arnold, Paṭis I, p. viii, and Mrs Rhys Davids’ rev. of Paṭis I, in JRAS 1908, p. 591. ↩

595. Paṭis II 168, 12. ↩

596. Paṭis II 205, 6. Cf. As 91-92, where the ten-fold iddhi is also found. ↩

597. Paṭis II 130, 17. ↩

598. Ed. M. E. Lilley, PTS London, Vol. I 1925, Vol. II 1927. ↩

599. See p. 77 above. ↩

600. See p. 135 below. ↩

601. See Müller, Thī-a p. ix. ↩

602. See Woodward, Th-a I p. vii. ↩

603. Ap 584-90. ↩

604. It is one of the texts rejected by the Dīgha-bhānakas (see p. 9 above). Malalasekera reports (DPPN, I, p. 116) that the Apadāna contains the names of 35 Buddhas. ↩

605. Abhidhammanayañño ‘ham Kathāvatthuvisuddhiyā
     sabbesam viñ̃̃āpetvāna viharāmi anāsavo (Ap 37, 1). ↩

606. Ap 1-6. ↩

607. See p. 36 above. ↩

608. It has been translated by Dwiendralal Barua. See “‘Buddhakhetta’ in the Apadāna,” in D. R. Bhandarkar et al. (ed.), B.C. Law Volume, Part II, Poona 1946, pp. 183-90. ↩

609. See H. Bechert, “Buddha-Feld und Verdienstübertragung: Mahāyāna-Ideen im Theravāda-Buddhismus Ceylons,” in Académie royale de Belgique, Bulletin de la Classe des Lettres et des Sciences morales et politiques, 5e série. Vol. 52, 1976, pp. 27-51. ↩

610. The word is used in a less extravagant way at Ap 44, 27, and in a different meaning at Ap 444, 19. ↩

611. Th 1087; Mil 176, 28. ↩

612. For references see BHS Dictionary, s.v. buddha-ksetra. ↩

613. See Vism 414, 19-27. ↩

614. Sn 35-75. ↩

615. See p. 64 above. ↩

616. Ap 299-301. ↩

617. Ap 301, 18-20. ↩

618. Ap-a 120-27. ↩

619. See H. Bechert, “Grammatisches aus dem Apadānabuch,” in ZDMG 108, 1958, pp. 308-16. ↩

620. See MW, s.vv. apadāna and avadāna, and BHS Dictionary, s.v. avadāna. ↩

621. See p. 37 above. ↩

622. H. Bechert, Bruchstücke buddhistischer Verssammlungen: I. Die Anavataptagāthā und die Sthaviragāthā, Berlin 1961 (STT VI). ↩

623. See Yuyama, VT, p. 51. ↩

624. Ed. R. Morris, PTS London, 1882; N. A. Jayawickrama, PTS London, 1974. Tr. B. C. Law, The Lineage of the Buddhas, SBB London, 1938; I. B. Horner, Chronicle of Buddhas, SBB London, 1975. ↩

625. See p. 77 above. ↩

626. See I. B. Horner, Bv tr., pp. xix-xlvi. ↩

627. Tanhankara, Medhankara and Saranañkara (Bv XXVII 1). ↩

628. Aparimeyye ito kappe caturo āsum vināyakā ti ādikā attharasagāthā sanḡ̄tikārakehi thapitā nigamanagāthā veditabbā (Bv-a 295, 31-32). ↩

629. Aham etarahi sambuddho Metteyyo cāpi hessati
     ete p’ ime pañca buddhā dhīrā lokānukampakā.
     etesam dhammarājūnam aññesam nekakotinam
     ācikkhitvāna tam maggam nibbutā te sasāvakā (Bv XXVII 19-20). ↩

630. See p. 41 above. ↩

631. See Horner, Bv tr., p..xiii. ↩

632. See Thomas, HBT, p. 204. ↩

633. See p. 9 above. ↩

634. See Rahula, HBC, p. 128. ↩

635. Ed. R. Morris, PTS London, 1882; N. A. Jayawickrama, PTS London, 1974. Tr. B. C. Law, The collection of the ways of Conduct, SBB London, 1938; I. B. Horner, Basket of conduct, SBB London, 1975. ↩

636. Buddhāpadāniyam nāma dhammapariyāyaṃ abhāsittha (Cp 103, 11-12). ↩

637. See p. 36 above. ↩

638. Bv I 76. ↩

639. Pandite paripucchitvā viriyam katvāna uttamam khantiyā pāramim gantvā patto sambodhim uttamam (Cp 368). ↩

640. See Horner, Cp tr., p. vi. ↩

641. With Mahāgovindacariya (Cp I.5) cf. Mahāgovindasutta of Dīgha-nikāya (19). ↩

642. With Mahā̄omahamsacariya (Cp III.15) cf. Mahāsīhanādasutta (also entitled Lomahamsapariyāya) of Majjhima-nikāya (12). See Horner, Cp tr., pp. viii-ix. ↩

643. Saccasavhayapaṇditacariya (Cp III.8). ↩

644. Ja I 45-47. ↩

645. Ayam ettha samkhepo, vitthārato pan’ esa attho Cariyāpitakato gahetabbo (Ja I 47, 16-17). ↩

646. Winternitz, HIL, p. 164. ↩

647. Cp 343-44 = Ja IV 31, 9-15. ↩

648. See L. Alsdorf, “Bemerkungen zum Vessantara-jātaka,” in WZKSO 1, 1957, pp. 2-14. ↩

649. See p. 9 above. ↩

650. Mhv IV 62. Buddhaghosa, however, says (As 27, 36) that it was recited. ↩

651. Dip V 37. ↩

652. See PTC, s.v. dhamma-dhara. ↩

653. Mātikādharā ti dvemātikādharā (Mp II 189, 23 ad A I 117, 29); mātikādharā ti dvepātimokkhadharā (Mp III 382, 11 ad A III 361, 25). See p. 126 below. ↩

654. Paṭis I 1-3. ↩

655. Dhs 1-8 ; 124-33. 7 Vibh :138-43; 244-45; 306-18; 345-49. See p. 100 below. ↩

656. There is no mātika as such printed at the beginning of the text. See p. 105 below. ↩

657. Tathāgatena thapita-mātikaṃ vibhajanto (Kv-a 7, 2-3). ↩

658. Moh 257-77. ↩

659. As 6, 38. ↩

660. Bv XXIV 6. ↩

661. Pakarane satta desento ti tattha devānaṃ sattappakaranasañkhātam Abhidhammapitakam desento vasi (Bv-a 259, 25-26). ↩

662. See CPD, s.v. abhidhamma. ↩

663. M II 239, 4 ( = abhivisitthe dhamme, imesu sattatimsabodhapakkhiyadhammess [Ps IV 29, 3]). ↩

664. Vin I 64, 28 ( = nāmarūpaparicchede vinetum na patibalo … sakale Vinayapitake vinetum na patibalo [Sp 990, 3-4]). See I. B. Horner, “Abhidhamma Abhivinaya,” in IHQ, 17, 3 (1941), pp. 291-310. See also Terry C. Muck, “The meaning of ‘Abhidhamma’ in the Pali canon’, in PBR 5, pp. 12-21. ↩

665. Vin V 1, 17. Cf. vinayo vibhatti abhivinayo (Vin V 2, 19). ↩

666. E.g. abhidhammakatham kathento (M I 214, 24). ↩

667. A IV 398, 1. ↩

668. Abhivinayo ti Khandhaka-Parivārā (Mp V 7, 2-3 ad A IV 398, 1). ↩

669. See Warder, IB, p. 219. ↩

670. As 2,13-3,20. ↩

671. As 3,21-23. ↩

672. P. V. Bapat and A. Hirakawa (tr.), Shan-Chien-P’i-P’o-Sha, Poona 1970, p. 11 . ↩

673. As 17,12-14. ↩

674. Ed. E. Müller, PTS London, 1885. Tr. Mrs C. A. F. Rhys Davids, Buddhist Psychological Ethics, London 1900. ↩

675. Dhs 1-8. ↩

676. See Iggleden, Vibh tr., p. xvii. ↩

677. Dhs 9-75. ↩

678. Dhs 75-87. ↩

679. Dhs 234-64. ↩

680. See Mrs Rhys Davids, Dhs tr., p. 334. ↩

681. As 409,22 . ↩

682. As 6,38 . ↩

683. As 410,1-10. ↩

684. [^2-4-36] Dip V 37. [^2-4-37] See K. R. Norman, “Four etymologies from the Sabhiya-sutta,” in Balasooriya, BSWR, p. 184 n. 33. [^2-4-38] See Iggleden, Vibh tr., p. xviii. [^2-4-39] See M III 187-257 (Vibhanga-vagga). [^2-4-40] Ed. Mrs C. A. F. Rhys Davids, PTS London, 1904. Tr. U Thitṭila, The Book of Analysis, PTS London, 1969, with a long introduction by R. E. Iggleden. [^2-4-41] M III 249-52. [^2-4-42] Moh 116-230. [^2-4-43] Paccayākāravibhanga (138-43); Jhānavibhanga (244-45); Nāṇavibhanga (306-18); Khuddakavatthuvibhanga (345-49). ↩

685. Iggleden, Vibh tr., p. lxiv. ↩

686. See H. Nakamura, Indian Buddhism: a survey with bibliographical notes, KUFS Publication Japan, 1980, p. 104 n. 10. ↩

687. Iggleden, Vibh tr., p. lxviii. ↩

688. Iggleden, Vibh tr., p. xxi. ↩

689. Ed. E. R. Gooneratne, PTS London, 1892. Tr. U Nārada, Discourse on Elements, PTS London, 1962, with a preface by Thein Nyun. ↩

690. Dhātuk 1. ↩

691. See rev. of Dhātuk tr., by R. E. Iggleden, JRAS 1964, pp. 78-79. ↩

692. See Thein Nyun, Dhātuk tr., pp. xxviii-xxix. ↩

693. Ed. R. Morris, PTS London, 1883. Tr. B. C. Law, Designation of Human Types, PTS London, 1924; (German) Nyānatiloka, Das Buch der Charaktere, Breslau 1911. ↩

694. See T. W. Rhys Davids, Buddhist India, London 1903, p. 188. ↩

695. Pp 1-10. ↩

696. D III 232-33. ↩

697. See Morris, Pp. pp. ix-x. ↩

698. See Morris, Pp, pp. x-xii. 58 Warder,: IB, p. 302. ↩

699. Ed. A. C. Taylor, PTS London, Vol. I, 1894; Vol. II, 1897. Tr. S. Z. Aung and Mrs C. A. F. Rhys Davids, Points of Controversy, PTS London, 1915. ↩

700. Dip VII 41; Mhv V 278. ↩

701. Vitandavād̄̄ pan’ āha: Kathāvatthum kasmā gahitam? (As 3, 25-26). ↩

702. As 5,32 . ↩

703. Sakavāde pañca suttasatāni paravāde pañca (As 6, 4). ↩

704. Geiger, PLL, p. 6. ↩

705. See p. 125 below. ↩

706. See Mrs Rhys Davids, Kv tr., p. xxxiii. ↩

707. For the methodology of Kv and the method of debate, see Jayawickrama, Kv-a, pp. xvii-xxi. ↩

708. Kv 163-204. ↩

709. See L. de La Vallée Poussin, “The ‘five points’ of Mahādeva and the Kathāvatthu,” in JRAS 1910, pp. 413-23. ↩

710. See K. R. Norman, “Māgadhisms in the Kathāvatthu,” in Narain, SPB, pp. 279-87. ↩

711. See Aung and Rhys Davids, Kv tr., pp. 401-404. ↩

712. See ibid., p. xxx n. 3, and cf. Kv 351, 18-21 with Khp IX 9. ↩

713. Ed. Mrs C. A. F. Rhys Davids, PTS London, Vol. I, 1911; Vol. II, 1913. ↩

714. Yam-a 52, 9-13. ↩

715. Moh 278, 2-4. ↩

716. Yam I, pp. xix-xxi. ↩

717. See Warder, IB, p. 302. ↩

718. Ed. Mrs C. A. F. Rhys Davids, Dukapatṭthāna, PTS London, 1906; Tikapaṭthāna, PTS London, Vol. I, 1921; Vol. II, 1922; Vol. III, 1923. Tr. U NĀrada, Conditional Relations, PTS London, Vol. I, 1969; Vol. II, 1981. See also U NĀrada, Guide to Conditional Relations, PTS London, Part I, 1979. ↩

719. See U NĀrada, Paṭṭh tr., Vol. I, p. 1. ↩

720. See ibid., p. xv. ↩

721. See ibid., between pp. lxiv and lxv. ↩

722. See p. 17 above. ↩

723. See J. Takakusu, “On the Abhidharma literature of the Sarvāstivādins,” in JPTS 1904-1905, pp. 66-146. ↩

724. See Warder, IB, p. 219. ↩

725. See Takakusu, JPTS 1904-1905, between pp. 74 and 75. ↩

726. See Warder, IB, p. 219. ↩

727. See S. Lévi, “Note sur des manuscrits sanscrits provenant de Bamiyan (Afghanistan), et de Gilgit (Cachemire),” in JA 1932, pp. 1-45. ↩

728. See Valentina Stache-Rosen, Das Sangitisūtra und sein Kommentar Sangitiparyāya, Berlin 1968 (STT, IX). ↩

729. See G. D. Bond, “The nature and meaning of the Netti-Pakaraṇa,” in Narain, SPB, p. 30. ↩

730. Ed. Arabinda Barua, PTS London, 1949. Tr. Bhikrhu Nāṇamoli, The Piṭaka-Disclosure, PTS London, 1964. ↩

731. N̄āṇamoli, Peṭ tr., p. xi. ↩

732. See p. 16 above. ↩

733. N̄āṇamoli, Peṭ tr., p. xi. ↩

734. See Peṭ, p. ii. ↩

735. Etad aggaṃ bhikkhave mama sāvakānaṃ bhikkhūnaṃ … sañkhittena bhāsitassa vitthārena attham vibhajantānaṃ yadidaṃ Mahā-Kaccāno (A I 23, 27-29). ↩

736. See Bode, PLB, p. 5. ↩

737. Vism 141, 13= As 165, 32= Sp 143, 29. See N̄̄̄̄amoli, Peṭ tr., p. 400. ↩

738. See N̄ānamoli, Nett tr., p. xiii n. 18 and p. xv n. 22. ↩

739. Peṭakenāpi samsandetvā (Nett-a quoted at Nett, p. x). ↩

740. See N̄ānamoli, Nett tr., pp. xiv-xvi. ↩

741. See ibid., p. viii. ↩

742. See ibid., p. vii. ↩

743. See ibid., pp. xvii-xviii. ↩

744. Ed. E. Hardy, PTS London, 1902. Tr. Bhikrhu Nāṇamoli, The Guide, PTS London, 1962. ↩

745. Nett 1, 4-13. ↩

746. A North Indian origin is also suggested by the number of unidentified verses it contains. See Horner, Mil tr., I, p. xv. ↩

747. See Alsdorf, Thi (Appendix II), p. 233. ↩

748. See Bode, PLB, p. 5. ↩

749. See Nett, pp. viii-ix, and Nāṇamoli, Nett tr., pp. x-xii. ↩

750. See Nāṇamoli, Nett tr., pp. liii-liv. ↩

751. E.g. Ps I 31, 8. ↩

752. Ed. V. Trenckner, London, 1880. Tr. T. W. Rhys Davids, The questions of King Milinda, London, Vol. I, 1890; Vol. II 1894 (= SBE Vols XXXV, XXXVI); I. B. Horner, Milinda’s Questions, SBB London, Vol. I, 1963; Vol. II, 1964; (French) L. Finot, (Books I-III), Paris 1923; (German) F. O. Schrader, (Books I-III), Berlin 1905; Nyāṇatiloka, München 1919. ↩

753. For suggestions about the date of Menander see Horner, Mil tr., I, pp. xxii- ↩

754. See the list of correspondences given by Horner, Mil tr., I, p. xx. ↩

755. Milinda is said to have approached Nāgasena at Sāgalā, which is perhaps the modern Siāllot. See Horner, Mil tr., I, p. 1 n. 3. ↩

756. See Rhys Davids, Mil tr., I, p. xix, and Tarn, op. cit. (in n. 25), pp. 422-23. ↩

757. See Tarn, op. cit. (in n. 25), pp. 419-20. ↩

758. There is mention of Alasanda, Kasmira and the village of Kalasi (Mil 82-83). ↩

759. Trenckner points out (Mil, p. vii) that the opening phrase tamyathā ‘nusūyate is found in no other Pāli writing, and it is only in Mil that quotations are introduced by bhavatiha. The word dātta “sickle” (Mil 33, 3) with an unexpected long vowel before a geminate consonant must also be a borrowing from a non-Pāli dialect. See R. L. Turner, Collected Papers: 1912-1973, pp. 432-34. ↩

760. See Rhys Davids, Mil tr., I, p. xi. ↩

761. See P. Demiéville, “Les versions chinoises du Milindapañha,” in BEFEO, 24 (1924), pp. 1-258. ↩

762. See Tarn, op. cit. (in n. 25), pp. 414-36. ↩

763. Mil 89, 17. See Winternitz, HIL, p. 176. ↩

764. Mil 4, 15-5, 30. ↩

765. See Winternitz, HIL, p. 178 n. 1. ↩

766. Nāgasena’s statement na h’ ettha puggalo upalabbhati (Mil 25, 12-13) is reminiscent of the first question of the Kathāvatthu: puggalo upalabbhati saccikatthaparamatthenā ti (Kv 1,4). ↩

767. Mil 90-420. ↩

768. See Rhys Davids, Mil tr., I, pp. xxvii-xxxvi, and Horner, Mil tr., p. xi. ↩

769. These are discussed by Horner, Mil tr., pp. xii-xv. ↩

770. See Horner, Mil tr., p. xv. ↩

771. Nāgasena states (Mil 268 and 271) that both nibbāna and ākāsa are akammaja. The usual Theravādin teaching is that nibbāna alone is asamkhata (see Horner, Mil tr., p. xviii). Horner (p. xlii) comments on the fact that this particular tenet was allowed to remain in Mil, and was not expunged as being contrary to orthodox teaching. The Milindapañha is, in fact, regarded as canonical in Burma (see Bode, PLB, p. 4). ↩

772. Buddhaghosa attributes Vism 270, 16-18 to the porannas, while the same verse at Mil 369, 22-24 is said to have been uttered by the theras who made a recension of the Dhamma (bhāsitam … therehi dhammasangāhakehi). ↩

773. Mil 342, 1. ↩

774. See Adikaram, EHBC, p. 25, and Horner, Mil tr., I, p. xxii. ↩

775. Ekacce ti Upatissattheram sandhāy’ āha: tena hi Vimuttimagge tathā vuttam (Vism-mhṭ [Ne 1969] I 221, 22). ↩

776. There is an English translation by N. R. M. Ehara, Soma Thera, and Kheminda Thera, The path of freedom, Colombo 1961. ↩

777. See N̄ānamoli, Vism tr., p. xxvii, and Bapat, VV, pp. xli-xlii. ↩

778. See Bapat, VV, p. xlix, and P. C. Bagchi, “On the original Buddhism, its canon and language,” in Sino-Indian Studies, II (1946-47), p. 113. ↩

779. See N̄ānamoli, Vism tr., p. xxvii. ↩

780. See A. Bareat, Les sectes bouddhiques du petit véhicule, Saigon 1955, p. 242. ↩

781. By M. Nagar, “The Vimutti-magga: the ‘way to deliverance,’” in JPTS 1917-1919, pp. 69-80. ↩

782. See N̄ānamoli, Vism tr., p. xxvii. ↩

783. By Bapat, VV, p. liv. ↩

784. P. V. Bapat, Vimuktimārga Dhutaguṇanirdeśa, Delhi, 1964. ↩

785. They do, however, occur. See Ehara, op. cit. (in n. 48), pp. xl-xli. ↩

786. See Bareau, op. cit. (in n. 52), p. 242. ↩

787. Ed. Shri Ratanajoti and Shri Ratanapala, printed at Government Press, Ceylon 1963. ↩

788. Rev. by P. V. Bapat, Journal of Vidyalankara University of Ceylon, pp. 17290 . ↩

789. Sv 1,17= Ps I 1,23= Spk I 1,19= Mp I 1,20 . ↩

790. See Geiger, DM, p. 45, and Oldenberg, Dip, p. 4. ↩

791. See Bechert, BCSRS, p. 15. ↩

792. See p. 118 below. ↩

793. Ed. and tr. H. Oldenberg, London 1879; B. C. Law, in Ceylon Historical Journal, 7 (1957-58), pp. 1-266. ↩

794. Mahāsena is dated A.D. 277-304 by Geiger in DM, p. 34, but A.D. 325-52 by the same scholar in Mhv tr., p. xxxviii. ↩

795. See p. 116 below. ↩

796. See Oldenberg, Dip, p. 9 n. 1. ↩

797. E.g. Bhātutissa at XXII 18, but Bhātikatissa at XXII 30. ↩

798. At IV 1-26 and V 1-15. ↩

799. At IV 47-53 and V 16-38. ↩

800. At VII 34-43 and VII 44-59. ↩

801. At XI 32-40, XII 1-4, and XVII 83-86. ↩

802. Oldenberg draws attention to the story of the conversion of Moggallāna at Dip V 55-68 and Mhv V 95-153. Cf. also Dip II 17-18 with Mhv I 52-57; Dip II 29 with Mhv I 48-51; Dip VI 13 with Mhv V 89-97; Dip VII 49 with Mhv V 238-42. ↩

803. E.g. Dip XVII 3-4, expanded at 5-25. ↩

804. E.g. Dip XIX 2-3 is expanded at Mhv XIX 2-12. ↩

805. E.g. therānaṃ occurs as an accusative singular form at Dip XII 83-84 and XIV 60, doubtless by analogy with rājānaṃ. ↩

806. Dip IV 47 and XII 30-33. ↩

807. By N. A. Jayawickrama, “Literary activity in Pāli,” in PBR 5 (1980), p. 80 n. 12 . ↩

808. Ed. and tr. N. A. Jayawickrama, Inception of Discipline and Vinayanidāna, SBB London, 1962. ↩

809. Jayawickrama, ibid., p. xi. ↩

810. Dip XI 15-17 quoted at Sp 74, 19-22 and Dip XII 1-6 quoted at Sp 75, 1576,11 . ↩

811. Dip XII 52 quoted at Sp 76,19-20. ↩

812. Dip IV 50-54 quoted at Sp 34,27-35,6. ↩

813. Dip XII 35-37 quoted at Sp 71, 18-23; Dip XII 38-39 quoted at Sp 70, 2471,1 . ↩

814. Dip V 30-53 quoted at Kv-a 3, 10-5, 9. ↩

815. Ed. and tr. G. Turnour, Ceylon 1837; Ed. W. Geiger, PTS London, 1908; Tr. W. Geiger, PTS London, 1912. ↩

816. Mhv-t 687, 4. ↩

817. Mhv I 2. ↩

818. See Geiger, DM, p. 17. ↩

819. Mhv-t 36, 5-7. ↩

820. Mhv-t, p. lvii. Cf. padyapadagāthābandhena kathesi (Mhv-t 42, 8). ↩

821. Doubtless because both works were based upon the same body of commentarial material. ↩

822. Dip V 32-38. ↩

823. Dip XVIII 53-XIX 11. ↩

824. Mhv XXII-XXXII. See Geiger, DM, pp. 19-20. ↩

825. See Mhv I 3. ↩

826. By Geiger, DM, p. 74. If this is so, then Mahānāma is later than Buddhaghosa. ↩

827. See Geiger, DM, pp. 16-17. ↩

828. Mhv V 145. ↩

829. See Geiger, DM, pp. 18-19. ↩

830. See Jayawickrama, Sp tr., p. xxxv. ↩

831. See p. 115 above. ↩

832. See K. R. Norman, “The role of Pāli in early Sinhalese Buddhism,” in Bechert, BCSRS, p. 36. ↩

833. Sv 1, 15-18 = Ps I 1, 21-24 = Spk I 1, 17-20 = Mp I 1, 18-21. ↩

834. See Adikaram, EHBC, p. 87. ↩

835. See K. R. Norman, “The dialects in which the Buddha preached,” in Bechert, LEBT, pp. 72-73. ↩

836. See pp. 89 and 99 above. ↩

837. Sp 2, 7-10. ↩

838. See p. 121 below. ↩

839. Mhv XXXVII 225-37. ↩

840. Vism 2, 15-24. ↩

841. S I 13,20-21 and 13,18-19. ↩

842. Vism 711, 25; Mhv XXXVII 232. ↩

843. Ed. C. A. F. Rhys Davids, PTS London, Vol. I, 1920; Vol. II, 1921; H. C. Warren and D. Kosambi, Cambridge Mass., 1950. Tr. Pe Maung Tin, The path of purity, PTS London, Part I, 1923; Part II, 1929; Part III, 1931; Bhikkhu Ñāṇamoli, The path of purification, Colombo 1956; (German) Nyanatiloka, Der Weg zur Reinheit, Konstanz 1952. ↩

844. Sīla-samādhi-paññā-mukhena desite Visuddhimagge(Vism 372, 21=710, 27-28). ↩

845. See p. 113 above. ↩

846. As is claimed by M. Nagar, “The Vimutti-magga: the ‘way to deliverance,” in JPTS 1917-1919, p. 80. ↩

847. See Nāṇamoli, Vism tr., p. xx. ↩

848. See p. 119 above. ↩

849. Gv 59, 18-30. ↩

850. Gv 59, 18-30 does not include Paramatthajotikā, but at 68, 34 it is stated that Buddhaghosa wrote the aṭthakathā on the Suttanipāta without being requested (attano matiyā). ↩

851. Ed. J. Takakusu and M. Nagar, PTS London; Vol. I, 1924; Vol. II, 1927; Vol. III, 1930; Vol. IV, 1934; Vol. V, 1938; Vol. VI, 1947; Vol. VII, 1947; Vol. VIII (Indexes, compiled by H. Kopp), 1977. ↩

852. Sp 2, 16-18. ↩

853. Sp 2, 21. ↩

854. See Adikaram, EHBC, p. 81. ↩

855. See ibid., pp. 12-13. ↩

856. See ibid., p. 12. ↩

857. Another authority, usually mentioned only to reject, is keci “some.” See I. B. Horner, “Keci, ‘some’ in the Pāli commentaries,” in JPTS 1981, pp. 87-95. 130 See Adikaram, EHBC, p. 12. ↩

858. E.g. Vism 72, 23 and 272, 2. See Jayawickrama, Sp tr., pp. xxii-xxiv. ↩

859. See p. 116 above. ↩

860. Ed. T. W. Rhys Davids and J. E. Carpenter, PTS London, Vol. I, 1886; W. Stede, Vol. II, 1931; Vol. III, 1932. ↩

861. Ed. J. H. Woods and D. Kosambi, PTS London, Vol. I, 1922; Vol. II, 1928; I. B. Horner, Vol. III, 1933; Vol. IV, 1937; Vol. V, 1938. ↩

862. Ed. F. L. Woodward, PTS London, Vol. I, 1929; Vol. II, 1932; Vol. III, 1937. ↩

863. Ed. M. Walleser, PTS London, Vol. I, 1924; M. Walleser and H. Kopp, Vol. II, 1930; H. Kopp, Vol. III, 1936; Vol. IV, 1940; Vol. V, 1957. ↩

864. Majjhe Visuddhimaggo esa catunnam pi āgamānaṃ hi (Sv 2, 6= Ps I 2,10= Spk I 2, 18= Mp I 2, 24). ↩

865. Spk I 2, 3-4. ↩

866. Mp I 2, 9-10. ↩

867. E.g. Vism 315, 24; 72, 24; 387, 8. ↩

868. Ed. E. Müller, PTS London, 1897; revised edition with indexes by L. S. Cousins, 1979. Tr. Pe Maung Tin and Mrs C. A. F. Rhys Davids, The Expositor, PTS London, Vol. I, 1920; Vol. II, 1921. ↩

869. Ed. A. P. Buddhadatta, PTS London, 1923. Tr. Bhikkhu Nāṇamoli, The Dispeller of Delusion, PTS London, to appear in 1983. ↩

870. See Jayawickrama, Kv-a, p. vii. ↩

871. Gv 59, 25-26. ↩

872. See p. 134 below. ↩

873. E.g. sabbaṃ aṭthakathānayena gahetvā Visuddhimagge vitthāritam (As 168, 21-22). ↩

874. Contrast iti pana sabbam yasmā Visuddhimagge mayā suparisuddham vuttam (Sv 2, 3-5) with Visuddhimagge pan’ idam yasmā sabbam pakāsitam (As 2, 8). ↩

875. Buddhaghoso ti gurūhi gahitanāmadheyyena therena katā ayam sakalassa pi Abhidhammapitakassa aṭthakathā (Ppk-a [Ne] III 245, 18-20). ↩

876. As 430, 19-21 ≠ Vibh-a 524, 7-8 ≠ Ppk-a [Ne] III 245, 18-20). ↩

877. Visuddhācārasilena nipunāmalabuddhinā bhikkhunā Buddhaghosena sakkaccam abhiyācito (As 1, 17-18); tassāham yācito thitagunena yatinā adandha-gatinā subuddhinā Buddhaghosena (Vibh-a 523, 9-10). The work bhikkhunā suggests rather that the ācariya Buddhaghosa was asked by a pupil to compose these works. This clearly is the view of the author of the Gandhavamsa: sattānam abhidhammagandhānam aṭthakathāgandho Culla-Buddhaghosanāmabhikkhunā āyācitena Buddhaghosācariyena kato (Gv 68, 24-26). ↩

878. See Mrs C. A. F. Rhys Davids, A manual of Buddhism for advanced students, London 1932, p. 30. ↩

879. See P. V. Bapat and R. D. Vadekar, Atthasālinī, Poona 1942 (Bhandarkar Oriental Series, 3), pp. xxxv-xl. ↩

880. Mhv XXXVII 225. ↩

881. As 2, 4. One passage (As 82-106), entitled Dvārakathā, is specifically attributed to the Mahā-aṭṭhakathā, while another, entitled Vipākuddhārakathā (As 267-87), seems to have been taken over almost verbatim or perhaps slightly condensed from a rather formalised earlier source, almost certainly a Sīhala atthakathā. It starts with what it calls a mātikā, in effect a table of contents, and gives a reference to the Ussadakittana, which is referred to at Vism 103-104 as an authorative decision in accordance with the views of the teachers of the atthakathā. See L. S. Cousins, “The Paṭthāna and the development of the Theravādin Abhidhamma,” in JPTS 1981, pp. 22-46. ↩

882. Re-writing is perhaps indicated by the fact that the Saddhammasangaha quotes Mhv XXXVII 225-41 (54, 5-55, 4) but later (56, 22) states that Buddhaghosa wrote the Atthasālinī in Ceylon after composing his commentaries upon the nikāyas. ↩

883. See Vibh-a 43, 14-15; 396, 14-15; 410, 4-5; 479, 13-14. ↩

884. Cf. Abhidh-av 2, 33-3, 14 with As 62, 1-17. ↩

885. See p. 130 below. The alternative explanation, that the Atthasālinī is based upon the Abhidhammāvatāra, seems untenable, because some of the latter is almost unintelligible without reference to the former. ↩

886. See U. Dhammaratana (ed.), Sammohavinodanī, Nalanda 1961, p. xxxix. ↩

887. E.g. ayam ettha amhākam khanti (Vism 434, 2); avicāritam etam porānehi. ayam pana me attano mati (Ps I 28, 30-31); amhākam khanti (Pj I 19, 34); ayam amhākam khanti (Pj II 387, 27-28); amhākam pan’ etam na kkhamati (Pj II 394, 6). Similarly Dhammapāla: ayam amhākam khanti (It-a II 126, 1). ↩

888. E.g. yathā ruccati tathā paṭhitabbaṃ (Pj I 78, 19); yathā vā tathā vā hoti (Pj I 165, 17); yam ruccati tam gahetabbam (Pj II 23, 2, 17; 43, 25; 136, 1; 378, 11). ↩

889. Vibhangatthakathāyam āvibhavissati (As 368, 2-3 = 407, 16-17). ↩

890. See Bapat and Vadekar, op. cit. (in n. 152), p. xxxiv. ↩

891. Dhātuk-a ed. E. R. Goonaratne (with Dhātuk), PTS London, 1892; Pp-a ed. G. Landsberg and Mrs C. A. F. Rhys Davids, in JPTS 1913-14, pp. 170-254 (reprinted with Pp, PTS London, 1972); Yam-a ed. Mrs C. A. F. Rhys Davids, in JPTS 1910-12, pp. 52-107; Paṭ̣h-a ed. Mrs C. A. F. Rhys Davids (with Dukap 1906, and Tikap 1921, 1922, 1923). ↩

892. Ed. J. P. Minayeff, in JPTS 1889, pp. 1-199; new ed. N. A. Jayawickrama, PTS London, 1979. Tr. B. C. Law, The Debates Commentary, PTS London, 1940. ↩

893. Sathārā dinnanayavasen’ eva Tathāgatena thapitamātikam vibhajanto (Kv-a 7,2-3). ↩

894. See Aung and Rhys Davids, Kv tr., pp. xxxiii-xxxiv. ↩

895. Just as yāva ajjatanā at Sp 62,10 probably refers to a date in the first century A.D. See Adikaram, EHBC, p. 87. ↩

896. See p. 18 above. ↩

897. Vin I 102, 33-103, 11; 103, 12-104, 20. ↩

898. E.g. Kkh 84, 37-38. ↩

899. Ed. Mrs D. Maskell, PTS London, 1956. ↩

900. Kañkhāvitaranī nāma Mātikaṭthakathā (Kkh 1, 1-3). ↩

901. See p. 96 above. ↩

902. Kkh 153, 40. ↩

903. N’ eva mātikāya na padabhājane vuttaṃ (Kkh 95, 23). ↩

904. Dve mātikā pagunaṃ katvā (Vism 312, 29). See Nāṇamoli, Vism tr., p. 17 n. 11 . ↩

905. As 19,5 . ↩

906. Ayam eva hi attho Sīhalamātikatṭhakathāyaṃ vutto (Kkh 159, 24-25). ↩

907. See Kkh 138, 11 and 110, 26. ↩

908. Samantapāsādikāya pan’ assa vicāraṇakathā (Kkh 159, 25). ↩

909. See Maskell, Kkh, p. x. ↩

910. The exegesis for the Bhikkhunī-pātimokkha is not given in full because of overlap with the Bhikkhu-pātimokkha. ↩

911. See Maskell, Kkh, pp. viii-ix. ↩

912. See p. 146 below. ↩

913. See Burlingame, Buddhist Legends (see n. 187), Vol. I, pp. 59-60. ↩

914. Ed. H. C. Norman, PTS London, Vol. I, Part 1, 1906; Part 2, 1909; Vol. II, 1911; Vol. III, 1912; Vol. IV, 1914; Vol. V (Index) compiled by L. S. Taitang, 1915; H. Smith, Vol. I, 1925. Tr. E. W. Burlingame, Buddhist Legends, Cambridge, Mass., Vols. I-III, 1921. ↩

915. Dhp-a I 1, 18. ↩

916. Dhp-a I 1, 14. ↩

917. Dhp-a IV 235, 6. ↩

918. See Nāṇamoli, Vism tr., p. xvii. ↩

919. Cf. Ps I 66 with Mp I 39-40. ↩

920. Cf. Ps I 185 with Mp I 42. See N̄̄̄̄amoli, Vism tr., pp. xvii-xviii. ↩

921. It is not impossible that Buddhaghosa entrusted some of the less important parts of his commentaries to assistants, and he was responsible only for their supervision. See Dhammaratana, op. cit. (in n. 159), pp. xxxviii-xxxix. ↩

922. Ed. V. Fausbell, London, Vols. I-VI, 1877-96; Vol. VII (Indexes, compiled by D. Andersen), 1897. Tr. E. B. Cowell et al., Vols. I-VI, 1895-1907; Vol. VII (Indexes), 1913. ↩

923. Ja I 1, 15-23. 197 Th:is suggests that the Theravādins and Mahimsāsakas had texts, especially the Jātakas, in common. Bareau suggests that the Mahimsāsakas were the mainland Theravādins. See Bareau, op. cit. (in n. 53), p. 183. ↩

924. See p. 78 above. 199 It:oparam pana bhikkhūnam yācanam bhavantarapaṭicchannatañ ca avatvā “atitam āharī ti” ettakam eva vakkhāma (Ja I 153, 28-154, 2). ↩

925. E.g. tam heṭ̣hā Nidānakathāyam kathitam eva (Ja I 137, 15-16); vatthum Dvādasanipāte Bhaddasālajātake āvibhavissati (Ja I 133, 23-24). ↩

926. Gv 63, 8. ↩ ↩²

927. Tesam vitthāro Añguttaraṭ̣̣hakathāya vutto (Ja I 131, 23’). ↩

928. Samyuttatthakathāya pana idāni vattamānasampatti yeva kathitā (Ja V 38, 4). Although this reference is found at Spk I 281, we cannot rule out the possibility that both Ja and Spk are following a Sīhala atthakathā here. ↩

929. E.g. potthakesu pana “tathā maṃ saccam pāleti” ti likhitam taṃ atthakathāyan n’ atthi (Ja V 95, 6-7). ↩

930. E.g. atthakathāyam pana “kasmā tuvan” ti pātho (Ja II 241, 17’). ↩

931. E.g. there is an exegesis on kantāra in the prose story of the Apanṇakajātaka at Ja I 99, 14-22. ↩

932. See Adikaram, EHBC, pp. 7-8, and N̄̄̄ṇamoli, Pj I tr., pp. x-xv. ↩

933. Ed. H. Smith, PTS London, 1915. Tr. Bhikkhu N̄̄̄ṇamoli, Minor Readings and Illustrator, PTS London, 1960. ↩

934. See N̄̄̄ṇamoli, Pj I tr., pp. viii-ix. ↩

935. See N̄̄̄ṇamoli, Pj I tr., p. 310 s.v. “Pānini” and p. 311 s.v. “Sanskrit allusions.” ↩

936. See Warren and Kosambi, op. cit. (in n. 115), pp. xii-xiv. ↩

937. See N̄̄̄ṇamoli, Pj I tr., p. xv. ↩

938. Ed. H. Smith, PTS London, Vol. I, 1916; Vol. II, 1917; Vol. III (Indexes), 1918. ↩

939. See K. R. Norman, “The role of Pāli in early Sinhalese Buddhism,” in Bechert, BCSRS, p. 42. ↩

940. See p. 58 above. ↩

941. Mahānāma’s accession is recorded at Mhv XXXVII 210 and his death at XXXVII 247. ↩

942. According to Geiger (Cūl tr., II, p. vi) Mahānāma reigned A.D. 409-31. ↩

943. See Geiger, Cūl tr., II, pp. v-vi. ↩

944. Sp 1415,18-20. ↩

945. See p. 127 above. ↩

946. See P. V. Bapat and A. Hirakawa (tr.), Shan-Chien-P’i-P’o-Sha, Poona 1970, p. xiii. ↩

947. See CPD, Vol. I, p. 471, s.v. avitakka². ↩

948. See K. R. Norman, “Pāli and the language of the heretics,” in AO 37 (1976), pp. 117-126. ↩

949. See Buddhadatta, Corr., pp. 158-59. ↩

950. Gv 59, 31-60, 2. ↩

951. Sās 33, 36-37. ↩

952. Gv 62, 14-15. ↩

953. Ed. A. P. Buddhadatta, Buddhadatta’s Manuals, Part II, PTS London, 1928: Vin-vn, pp. 1-230; Utt-vn, pp. 231-304. ↩

954. Ed. A. P. Buddhadatta, Buddhadatta’s Manuals, Part I, PTS London, 1915: Abhidh-av, pp. 1-138; Rūpār, pp. 149-59. R. Webb (PBR, 5, 3, 1980, p. 88) reports a tr. of Rūpār: R. Exell, “The classification of forms and formless things,” in Visakha Puja, Bangkok 1964. ↩

955. See p. 124 above. ↩

956. Vin-vn 3177. ↩

957. Utt-vn 965. ↩

958. Abhidh-av 137, 12. ↩

959. See Winternitz, HIL, p. 220. ↩

960. See Index to Buddhadatta’s Manuals, Part II, s.vv. Kurundaṭṭhakathā, etc. ↩

961. Vin-vn 3179. ↩

962. See Buddhadatta’s Manuals, Part I, p. xvi, and Buddhadatta, Corr., pp. 163-64. ↩

963. See p. 145 below. ↩

964. Ed. C. V. Joshi, PTS London, Vol. I, 1933; Vol. II, 1941; Vol. III, 1947. ↩

965. Gv 61, 3. ↩

966. Gv 70, 16. ↩

967. Paṭis-a 704, 2. ↩

968. The probable date for Moggallāna I is A.D. 496-513, which gives A.D. 516 as the date of completion for Paṭis-a. See Warder, Paṭis tr., p. xliv. ↩

969. See Vism, pp. 758-59. ↩

970. Warder, Paṭis tr., p. xlv. ↩

971. Warder, Paṭis tr., pp. xlii-xliii. ↩

972. Paṭis-a 1-2. ↩

973. Ed. A. P. Buddhadatta, PTS London, Vol. I, 1931; Vol. II, 1939; Vol. III, 1940. ↩

974. At Gv 61, 11, although at Gv 70, 23 it is called Saddhammappajotikā. ↩

975. Gv 61, 11 and 70, 24. ↩

976. Nidd-a II 150, 25. The editor’s reference to Upatissa (I, pp. viii-x) is an error, as he later points out (II, p. v). ↩

977. Gv 70, 24. ↩

978. Nidd-a II 152, 1-3. ↩

979. Buddhadatta (Nidd-a I, p. ix) concludes that Siri-Sanghabodhi is Aggabodhi I, who came to the throne c. A.D. 554. This would give a date around A.D. 580 for the completion of Nidd-a. ↩

980. See Nidd-a I, pp. vii-viii, where the editor ascribes Pj II to Dhammapāla. ↩

981. See ibid., p. viii. ↩

982. Gv 60, 5-17. ↩

983. See p. 148 below. ↩

984. The title Paramatthavibhāvinī also occurs at It-a 95, 24 and 180, 33. ↩

985. Ud-a 436, 32= It-a 194, 5 (Padara-) = Pv-a 287, 18 (Vada-) = Vv-a 355, 13 = Th-a III 210, 23= Thī-a 301, 3 (Padara-) = Cp-a 336, 7. ↩

986. Mrs Rhys Davids’ interpretation of this as “three commentaries” (Thī tr., p. 178) is based upon a misreading of -n- as -t-. See Mrs Rhys Davids, Th tr., p. xxiv n. 1. ↩

987. See Nett 249, 33. ↩

988. E.g. at Ud-a 2, 6. ↩

989. E.g. he refers to Vism at Th-a II 206, 1; III 54, 17; 198, 9. See also Mrs Rhys Davids, Th tr., pp. 426-27. ↩

990. Cf. the explanation of the word tathāgata given at Ps I 45, 10-15 with that at Th-a I 36, 7-12. ↩

991. E.g. besides the reading uppacca at Pv 236 ( = II 7.17) Dhammapāla records the v.l. upecca, and explains both (Pv-a 103, 29-104, 1). ↩

992. E.g. two explanations are given for the word subhe (Pv 348): subhe ti suddhe, subhe ti vā tassā ālapanam (Pv-a 158, 2). See Masefield, Pv-a tr., p. ix. ↩

993. Ed. E. Hardy, PTS London, 1894; Tr. U Ba Kyaw and P. Masefield, Peta Stories, PTS London, 1980. ↩

994. Ed. E. Hardy, PTS London 1901. ↩

995. Ed. F. L. Woodward, PTS London, Vol. I, 1940; Vol. II, 1952; Vol. III, 1959. ↩

996. Ed. E. Müller, PTS London, 1893. ↩

997. See Woodward, Th-a I pp. vi-vii. ↩

998. See Norman, EV I, pp. xxix-xxx, and EV II, pp. xxxi-xxxii. ↩

999. See Thi-a 128, 26-29, where he draws attention to the difference between the details about Khemā (Thī 139-44) in the atthakathās and the Apadāna. Similarly, two traditions are recorded about Sopāka (Th 480-86). See Norman, EV I, p. 199. ↩

1000. There seems to be no evidence for the existence of a Sīhala atthakathā upon Th and Thi, as such, although information about the elders would be available in the atthakathās upon the Añguttara- and other nikāyas. In many ways the Apadāna seems to be intended as a commentary upon Th and Thi. ↩

1001. E.g. he did not realise that phalesin (Th 527 and 1121) is a future active participle. ↩

1002. E.g. he gives different explanations for Th 909 and 1181, although they are identical. ↩

1003. Commenting upon cola (Th 170), Dhammapāla gives an explanation which must have been given originally in a dialect where the word for “thief” was pronounced as cola (Th-a II 46, 4-8). The gloss tosana given for abhibhāsana (Th 613) is more suitable for abhihāsana, and presumably goes back to a dialect where both words were pronounced in the same way (Th-a II 260, 2-3). ↩

1004. E.g. three explanations are given for the word indagopaka (Th 13) at Th-a I 62,13-17. ↩

1005. Ed. D. L. Barua, PTS London, 1939; 2nd edition (with indexes by H. Kopp), 1979. ↩

1006. E.g. Cp-a 206, 32; 229, 13; 3, 12; 16, 13. ↩

1007. Keci pana chabbidhā ti vadanti. tam etāsam sañgahavasena vuttam (Cp-a 278, 4-5 ). ↩

1008. Ed. F. L. Woodward, PTS London, 1926. ↩

1009. Ed. M. M. Bose, PTS London, Vol. I, 1934; Vol. II, 1936; Vol. III (Indexes, compiled by H. Kopp), 1977. ↩

1010. See H. C. Ray, University of Ceylon History of Ceylon, Vol. I, Colombo 1959, p. 391. ↩

1011. The PTS ed. of Nett includes extracts from Nett-a (= Nett 194-249). ↩

1012. See Nett, p. xv. ↩

1013. Atthasāliniyam (Nett 215, 17; 240, 30). ↩

1014. Bu-up 60, 24-26. ↩

1015. See p. 118 above. ↩

1016. There is a tradition in Ceylon that he was called Mahānāma. See Malalasekera, Mhv-t, p. civ. ↩

1017. Ed. G. P. Malalasekera, Vamsatthappakāsinī, PTS London, 2 Vols., 1935. ↩

1018. Mhv-t 687, 12. ↩

1019. Mhv-t 689,5 . ↩

1020. See p. 117 above. ↩

1021. See Malalasekera, Mhv-t, pp. lvi-lxxii. ↩

1022. To conclude from this that the Dipavaṃsa was written by nuns seems rather farfetched. See Malalasekera, PLC, pp. 135-37. ↩

1023. Cf. Dīp I 1 with Mhv-t 36, 1-3. ↩

1024. E.g. at Mhv-t 125, 7-8. ↩

1025. See Malalasekera, DPPN, I, p. 360. ↩

1026. Mayam pi sañkhepena tesaṃ uppattimattam samayāvirodhamattam kathayāma (Mhv-t 177, 31-32). ↩

1027. Mhv-t 187, 6-7. ↩

1028. Mhv-t 177, 30-31. ↩

1029. Mhv-t 155, 15-16. ↩

1030. See Malalasekera, Mhv-t, p. cix. ↩

1031. E.g. katthaci “duve satāni purise” likhanti; taṃ pamādalekhaṃ (Mhv-t 284, 20-21 ad Mhv X 27) ; see also 423, 24-424, 1; 659, 4-5. ↩

1032. Mhv XXIX 44 states that the porānas did not mention the number of monks of Ceylon who had come to the ceremony. The tīkā points out that the total number was given: samāgatānaṃ sabbesaṃ bhikkhūnaṃ visuddhippattā khīnāsavā eva channavutikoṭiyo ti porānehi bhāsitā ti vuttam hoti (Mhv-t 521, 5-7). ↩

1033. Ayaṃ amhākam khanti (Mhv-t 208, 20-21). ↩

1034. Ed. G. P. Malalasekera, Extended Mahāvamsa, Colombo 1937 (Aluvihāra Series, Vol. III). ↩

1035. Geiger, DM, p. 27. ↩

1036. See Malalasekera, K Mhv, pp. xl-l. ↩

1037. See Geiger, Mhv tr., p. 34 n. 3. ↩

1038. K Mhv XXXVIII 24. ↩

1039. See Geiger, DM, p. 31. ↩

1040. Mhv LXXVIII 9. ↩

1041. See p. 118 above. ↩

1042. Ed. W. Geiger, Cūlavamsa, PTS London, Vol. I, 1925; Vol. II, 1927. Tr. W. Geiger, Cūlavamsa, PTS London, 2 Vols. 1929. ↩

1043. Mhv XXXVII 51-52 = Dīp XXII 75-76. ↩

1044. Mhv XXXVII 51 - LXXIX 84. ↩

1045. According to Mhv LXXXIV 12 he came to Ceylon from Burma in the reign of Parakkamabāhu II. ↩

1046. Mhv LXXIX 85-86. ↩

1047. Mhv LXXX 1-XC 102. ↩

1048. Mhv XC 103-104. ↩

1049. Mhv XCI-C. ↩

1050. Mhv IC 76-80. ↩

1051. See Geiger, Cül tr., II, p. 263 n. 1. ↩

1052. Geiger stated (Cūl, p. v), that he had seen no manuscript which went beyond C 292, and the remainder of C and the whole of CI were taken from the Colombo edition. ↩

1053. H. Sumangala and DAdS. Batuwantudawa (ed.), The Mahawansa from the Thirty-seventh Chapter, Colombo 1877. ↩

1054. Ed. S. A. Strong, PTS London, 1891. ↩

1055. See Geiger, DM, p. 77. ↩

1056. Mhbv 1, 7-8. ↩

1057. See Malalasekera, PLC, p. 159. ↩

1058. See A. K. Warder, “Some problems of the later Pāli literature,” in JPTS 1981, pp. 203-205. ↩

1059. Gv 61,5 ; 70,18 ; Sās 34,29 . ↩

1060. Quoted from the Sāsanavamsadīpa by B. C. Law, On the chronicles of Ceylon, Calcutta 1947, p. 39. ↩

1061. Mhv LIV 36. A. K. Warder (JPTS 1981, p. 198) points out that there is no certainty that the two Dāṭhānāgas are the same person. ↩

1062. T. W. Rhys Davids and R. Morris (ed.), JPTS 1884, p. 109-51. Ed. and tr. B. C. Law, Lahore, 1925. ↩

1063. Dāṭh I 10. ↩

1064. Gv 72, 5. ↩

1065. Dāṭh VI 6. ↩

1066. Dāṭh I 6. ↩

1067. Mhv LXXX 49-50. ↩

1068. Ed. B. C. Law, PTS London, 1935. Tr. B. C. Law, The legend of the topes, Calcutta 1945. Ed. and tr. N. A. Jayawickrama, SBB London, 1971. ↩

1069. Thūp 105, 19-20. ↩

1070. Thūp 1, 17-20. ↩

1071. Mhv LXXXI 18. See Geiger, DM, p. 84. ↩

1072. See Geiger, DM, p. 85. ↩

1073. Ed. C. E. Godakumbura, PTS London, 1956. ↩

1074. See Godakumbura, Att, p. x. ↩

1075. Att 1, 13-14. For Anomadassi, see Malalasekera, DPPN, I, p. 102. ↩

1076. Ed. J. P. Minayeff, JPTS 1885, pp. 5-16. ↩

1077. JPTS 1885, p. 5. ↩

1078. Ed. A. P. Buddhadatta, PTS London, 1962. Tr. N. A. Jayawickrama, Epochs of the conqueror, PTS London, 1968. ↩

1079. Satthu nibbānato vassānaṃ saṭthiadhike dvisahassaparimāne, atthasattatiadhike ca atthasata-Sakarāje Mūsikasaññite samvacchare (Jinak 129, 9-11). The designation of years by animal names (mūsika) is borrowed from the Chinese “rat”, etc. See Jayawickrama, Jinak tr., pp. xii-xiii. ↩

1080. It seems probable that the author himself made these additions, either because important events occurred after the completion of the work which he thought should be recorded, or as the beginning of a supplement, on the lines of the Cūlavarpsa. See Jayawickrama, Jinak tr., pp. xxviii-xxix. ↩

1081. See ibid., p. xv. ↩

1082. See ibid., pp. xix-xxi. ↩

1083. E.g. the visit of Thai monks to Ceylon, their ordination there in A.D. 1424, and the subsequent introduction of the Sihala dispensation into Thailand (Jinak 92-95 ). ↩

1084. See Jayawickrama, Jinak tr., p. xvii. ↩

1085. See p. 12 above. ↩

1086. Although some of the dates the author gives seem to be doubtful. See Jayawickrama, Jinak tr., p. xiii. ↩

1087. The editor says nothing about this, but see Jayawickrama, Jinak tr., pp. 187-210. ↩

1088. See Jayawickrama, Jinak tr., pp. viii-ix. ↩

1089. See G. Cœdès, “Une recension Pālie des Annales d’Ayuthya,” in BEFEO 14 (1914), 3, pp. 1-31, where an ed. and tr. (French) of ch. 7 is given. ↩

1090. See ibid., p. 2. See also H. Saddhatissa, “Pāli literature of Thailand,” n Cousins, BSIBH, pp. 219-20. ↩

1091. Ed. J. P. Minayeff, JPTS 1885, pp. 17-28. ↩

1092. Sandes 28, 27-28. ↩

1093. Sammāsambuddhaparinibbānato pañcacattālisatisatadvisahasse jinasāsanavasse (Sandes 28, 20-21). ↩

1094. Ed. and tr. J. Gray, London 1892. ↩

1095. See Geiger, PLL, 40. 3. Gray (Bu-up tr., pp. 32-33) suggests that the author is to be identified with the tutor of the Vedeha who wrote the Rasavāhinī, in which case he lived in the thirteenth or fourteenth century. ↩

1096. Pubbācariyānaṃ santikā yathāpariyattiṃ (Bu-up 67, 2-3). ↩

1097. Sās 30,17-21 points out that in Vism (711, 5) it is stated that the work was composed at the request of Sanghapāla, whereas in Bu-up (55, 4-5) the thera is named Sangharāja. ↩

1098. Mhv XXXVII 244. ↩

1099. Bhagavato sāsanam dīpabhāsato parivattetvā Māgadhabhāsāya divase divase likhi (Bu-up 59, 16-17). ↩

1100. See p. 138 above. ↩

1101. Not even in Cp-a, where, as B. M. Barua points out (Buddhadatta, Corr., pp. 177-78), he might have been expected to do so. ↩

1102. See p. 131 above. ↩

1103. Ed. I. B. Horner, Madhuratthavilāsinī, PTS London, 1946. Tr. I. B. Horner, The clarifier of the sweet meaning, SBB London, 1978. ↩

1104. Bv-a 1, 26-27. ↩

1105. Vin-vn 3177. ↩

1106. Bv-a 299, 16. ↩

1107. Abhidh-av 138, 4. ↩

1108. Ayaṃ gāthā Atthasāliniyā Dhammasañgạ̣i-atṭ̂hakathānidānavạ̣̄anāya Dī. pañkarabuddhavamse likhitā (Bv-a 126, 29-30). ↩

1109. Bv-a 284, 27. ↩

1110. Vv-a 314. ↩

1111. See p. 134 above. ↩

1112. Gv 60, 8. ↩

1113. C. E. Godakumbura (ed.), Visuddhajanavilāsinī, PTS London, 1954. ↩

1114. See p. 121 above. ↩

1115. Gamanavaṇnanāgāthānaṃ attho Theragāthatṭhakathāyaṃ vutto yeva (Ap-a 540, 20-21). ↩

1116. See Godakumbura, Ap-a, p. xv. There are also references to the Atthasālinī at Ap-a 194,17 and 201, 10-11. ↩

1117. See Godakumbura, Ap-a, pp. xiv-xv. ↩

1118. See ibid., pp. xvii-xviii. ↩

1119. See L. S. Cousins, rev. of de Silva DAT, in Religion, II, p. 162. ↩

1120. Ap-a 220, 2-3. ↩

1121. Ap-a 491, 19. ↩

1122. Ap-a 102, 19. ↩

1123. Ap-a 102, 17-18. ↩

1124. Ed. A. P. Buddhadatta and A. K. Warder, PTS London, 1961. ↩

1125. See Buddhadatta, Moh, p. viii. ↩

1126. See ibid.. p. xi. ↩

1127. Moh 1, 21. ↩

1128. See Warder, Moh, p. xvii. ↩

1129. A detailed investigation into the mātikās is made by Warder (Moh, pp. xix-xxvii). ↩

1130. See Malalasekera, PLC, pp. 192-93. A supplement to a tīkā is called an anuṭ̂kā. ↩

1131. E.g. Mahāvam̧sa-ṭīkā (see p. 138 above) and Milinda-ṭīkā (see p. 150 below). ↩

1132. Saddhamma-s 58, 27-60, 24. ↩

1133. Vitthāro pana … antarābhavavicāro Kathāvatthupakarañassa tīkāya gahetabbo (Ud-a 94, 7-9). ↩

1134. Abhidhammat̄̄kam pana Ānandathero akāsi. sā ca sabbāsam tīkānam ādibhūtattā Mūlatīkā ti pākatā (Sās 33, 16-17). Cf. Ānando nāmācariyo sattābhidhammagandha-atṭhakathāya Mūlatīkaṃ nāma tīkaṃ akāsi (Gv 60, 3-4). ↩

1135. See Cousins, op. cit. (in n. 106), p. 161. ↩

1136. No Ee yet published. ↩

1137. E.g. Ud-a 22, 22-23, 16 ≠ Sv-pṭ I 59, 10-60, 6; Ud-a 250, 22 = Sp-t (Be) III 1960 360, 29; Cp-a 276-332 = Sv-pṭ I 86-130. CPD (Vol. I, p. xxviii) seems to imply that Dhammapāla was making use of tīkās, although de Silva (DAT I, pp. xlii-xliii) seems to regard these parallels as evidence for common authorship. ↩

1138. See A. K. Warder, “Some problems of the later Pāli literature,” in JPTS 1981, pp. 201. ↩

1139. Gv 60,10-15. ↩

1140. Sās 33,18-21. ↩

1141. It attributes the tīkā on the Añguttara-atṭ̣hakathā to Sāriputta. ↩

1142. See Hardy, Nett, p. ix n. 6. ↩

1143. Vism-mhṭ (Ne) III, 1691, 9-10. ↩

1144. Mhv LIV 6. ↩

1145. Mhv LIV 36. ↩

1146. Gv 66-67. ↩

1147. The suggestion was rejected by E. Hardy (“Ein Beitrag zur Frage, ob Dhammapāla im Nālandasañghārāma seine Kommentare geschrieben”, in ZDMG 51, 1897, pp. 105-27), but has recently been revived by de Silva (DAT I, p. 1). ↩

1148. Ed. L. de Silva, Dīghanikāyaṭṭhakathāṭīkā Līnatthavaṇṇanā, PTS London, 3 vols., 1970. ↩

1149. See de Silva, DAT, pp. 1-lviii. ↩

1150. See ibid., pp. lix-lxiii. ↩

1151. It has been suggested that Abhayagirivāsin means all non-Mahāvihāravāsins, whereas Uttaravihārin means only the headquarters of the Abhayagirivāsins at the Uttaravihāra. See de Silva, DAT, p. lxi. ↩

1152. See p. 139 above. ↩

1153. See CPD, Vol. I, p. xxviii. ↩

1154. See Bollée, KJ, p. xii, where it is accepted that it is Dhammapāla’s tīkā. ↩

1155. Gv 65, 29. ↩

1156. Ed. P. S. Jains, PTS London, 1961. ↩

1157. Mil-t 71, 23-24. ↩

1158. See Jaini, Mil-t, p. xiv. ↩

1159. See ibid., p. vii. ↩

1160. Mil-t 17, 12-18, 11. ↩

1161. See Jaini, Mil-t, p. xi. ↩

1162. Bhadanta-Buddhaghosācariyena katā ti parihāro. na kevalam pañca gāthā va, Thera-Rājavacane pi aññam pubbāparavacanaṃ pi tena vuttam (Mil-t 3, 16-18). ↩

1163. See p. 111 above. ↩

1164. It is not practicable to try to list them all. For details see Gv 60-66 and CPD, Vol. I, Epilegomena pp. 37*-60*. ↩

1165. Ed. T. W. Rhys Davids, JPTS 1884, pp. 1-46. Tr. S. Z. Aung and Mrs. C. A. F. Rhys Davids, Compendium of philosophy, PTS London, 1910. ↩

1166. Gv 67, 2. ↩

1167. See Malalasekera, PLC, pp. 168-69. For other commentaries upon Abhidh-s, see H. Saddhatissa, “The Abhidhammatthasañgaha and its tīkā,” in Studies in Indian Philosophy, Ahmedabad 1981, pp. 315-23. ↩

1168. Paccayā nāma Patthāne catuvīsati sabbathā (Abhidh-s 39, 11). ↩

1169. Ed. A. P. Buddhadatta, JPTS 1913-14, pp. 5-113. ↩

1170. Nāmar-p 2. ↩

1171. Nāmar-p 1854. ↩

1172. E.g. Nāmar-p 120; 208; and 478. ↩

1173. There is as yet no Ee of this, but the typescript of an edition by A. P. BudDhadatta is in the possession of the PTS London, awaiting publication. ↩

1174. By Buddhadatta in the typescript mentioned in the previous note. ↩

1175. Ed. P. Dhammārāma, JPTS 1917-19, pp. 3-25. ↩

1176. Sacc 25, 15-19. ↩

1177. Sās 34,2-3. ↩

1178. Gv 60, 30-31. ↩

1179. Saddhamma-s IX 16. ↩

1180. Ānandācariyassa jetthasisso Culladhammapālo (Gv 60, 30). ↩

1181. See p. 148 above. ↩

1182. Ed. P. Dhammārāma, JPTS 1915-16, pp. 3-19. ↩

1183. Gv 61, 28-29; 67, 2-3. ↩

1184. See Aung, Abhidh-s tr., p. ix n. 1. ↩

1185. See p. 131 above. ↩

1186. See p. 147 above. ↩

1187. No Ee has as yet been published. ↩

1188. See Aung, Abhidh-s tr., p. ix. ↩

1189. See Bode, PLB, p. 18. ↩

1190. See Buddhadatta, Corr., pp. 208-9. ↩

1191. Ed. and tr. (French) J. Ver Eecke, EFEO Paris, 1976. ↩

1192. See Ver Eecke, Dvp, p. xiv. ↩

1193. See ibid., p. ix. ↩

1194. See Bode, PLB, p. 104. ↩

1195. Ed. and tr. (French) J. Ver Eecke, EFEO Paris, 1980. ↩

1196. See Ver Eecke, Sīh, pp. iv-vii. ↩

1197. See ibid., p. ix. ↩

1198. See p. 140 above. ↩

1199. See Bode, PLB, p. 104 n. 9. ↩

1200. The vihāra is called Paṭṭakoṭti, Haṇ̣akonda or Saṇ̣akonti. The town is called Kaṇtakasolapaṭtana or Koṇḍakasolapaṭtana. ↩

1201. See Ver Eecke, Sīh, p. iii. ↩

1202. See Bode, PLB, p. 105. ↩

1203. By Sodi Mori, “A Japanese translation and notes on the Sihalavatthupakarana (II), Chapter 1, Stories 3-5,” in Studies in the Humanities of Josai University, Saitama, Japan, 1 (December 1973), p. 99. ↩

1204. Ed. A. P. Buddhadatta, Ambalangoda 1959. ↩

1205. By Malalasekera, quoted by Rahula, HBC, p. xxxiv. ↩

1206. See Buddhadatta, Svp, p. xxv and Rahula, HBC, pp. xxx-xxxi. ↩

1207. See Bode, PLB, p. 105. ↩

1208. At Mhv-t 451, 19; 452, 28; and 607, 8-9. ↩

1209. Mhv-t 607, 8-9. ↩

1210. Sālirājakumāravatthum Mahāvaṃse vuttanayena veditabbam (Sah 108, 5-6). ↩

1211. It is surprising that although F. Spiegel edited and translated selected stories in 1845 (Rasavāhinī eine buddhistische Legendensammlung, 1-4. Anecdota Pālica I, Leipzig 1845), no complete Ee has yet appeared. See also M. and W. Geiger (ed. and tr.), Die zweite Dekade der Rasavāhinī, Munich 1918. The sections (vaggas) and stories (vatthus) are analysed with the corresponding pages of a Ce (ed. Sarapatissa Thera, Colombo 1928) by Godakumbura, Cat. pp. 56-59. ↩

1212. Yo ca Samantakūṭassa vaṇṇanam vaṇnayī subhamtena Vedehatherena katāyam Rasavāhinī. ↩

1213. See Malalasekera, PLC, p. 225. ↩

1214. See Rahula, HBC, p. xxxix. ↩

1215. See ibid., p. xxix. ↩

1216. See p. 118 above. ↩

1217. See p. 141 above. ↩

1218. Ed. E. R. Gooneratne, JPTS 1884, pp. 49-68. In PBR 2, 3 (1977), pp. 127-40, R. Webb draws attention to a translation by B. C. Law in Indian Culture, 5 (1938-39), p. 25, and republishes one by C. Samaresingha; in PBR 5, 3 (1980), p. 88, he notes a third translation by S. K. Ramachandra Rao in the Quarterly Journal of the Mythic Society (Bangalore), 1957, pp. 214-27 and 260-81. ↩

1219. Mhv XXII 13-20. ↩

1220. See Gooneratne, JPTS 1884, p. 49. ↩

1221. The word sata is often not to be taken literally, but in the sense of “a large number” (see MW, s.v. śata). It is perhaps being used in Ras merely as a “round number” (see Malalasekera, PLC, p. 163). The fact that Tel has only 98 verses would seem to prove that it is earlier than Ras, otherwise the author would certainly have composed 100 verses. ↩

1222. Gv 69,15. ↩

1223. Ed. and tr. J. Gray, London 1894; repr. PTS London, 1981. ↩

1224. Gv 72, 9. ↩

1225. Gv 65, 19. Gray states (Jināl, p. 15) that he made use of a tīkā and a text entitled Gulatthadīpanī when making his edition. ↩

1226. Saddhamma-s 63, 9. ↩

1227. Sattarasasate vasse gate. ↩

1228. See Gray, Jināl, p. 7. ↩

1229. See Geiger, PLL, 34.3. ↩

1230. Jinak 71, 19-24. ↩

1231. Godakumbura, however, quotes (Samantak, p. 81) a passage from a Jinālañkāravaṇ̣anā which, since it is in prose, must be a different text of the same name, perhaps a commentary upon the Jinālañkāra. See also Godakumbura, Cat. pp. 50-52. ↩

1232. Some of the sections have headings which signify the particular feature to be exemplified in it, e.g. XIII. “Distributed rhyme in two pādas” (Dvipādavyāsayamakagāthā). ↩

1233. Jināl 99-100. ↩

1234. E.g. Jināl 173, where each successive pair of words rhyme. ↩

1235. Jināl 97, where each pāda is sakāmadātā vinayāmanantagū. ↩

1236. Jināl 98. ↩

1237. Jināl 101. ↩

1238. Jināl 105. ↩

1239. See A. B. Keith, A history of Sanskrit literature, London 1928, p. 306. ↩

1240. See Gray, Jināl, pp. 69-72. ↩

1241. Ed. and tr. W. H. D. Rouse, JPTS 1904-1905, pp. 1-65; Ch. Duroiselle, Rangoon, 1906. ↩

1242. Jina-c 471. ↩

1243. Gv 62, 24-25. ↩

1244. Saddhamma-s 63, 11-12. ↩

1245. Jina-c 469-70. ↩

1246. See Malalasekera, PLC, p. 230, and p. 165 below. ↩

1247. See Winternitz, HIL, p. 224. ↩

1248. Law, HPL, p. 614, quoting Duroiselle. ↩

1249. Ed. E. R. Gooneratne, JPTS 1887, pp. 1-16. ↩

1250. Gooneratne, JPTS 1887, p. 1. ↩

1251. Pajj 103. ↩

1252. See Malalasekera, PLC, p. 220 and p. 164 below. ↩

1253. Ed. C. E. Godakumbura, PTS London, 1958. ↩

1254. See p. 155 above. ↩

1255. See Godakumbura, Samantak, p. ix. ↩

1256. See ibid., p. xi. ↩

1257. Mhv I 77. ↩

1258. See Godakumbura, Samantak, pp. xiv-xv. ↩

1259. See ibid., p. xx. ↩

1260. See ibid., pp. xx-xxiv. ↩

1261. Ed. R. Morris, JPTS 1887, pp. 35-98. ↩

1262. Saddh 3. Gv 75, 15 says nothing about the author or the sponsor. ↩

1263. See Malalasekera, PLC, p. 212. ↩

1264. See H. Bechert, “Notes on the formation of Buddhist sects and the origins of Mahāyāna,” in German Scholars on India: contributions to Indian Studies, Varanasi 1973, Vol. I, p. 11. ↩

1265. Which presumably accounts for its survival, if it is an Abhayagiri text. See N. A. Jayawickrama, “Literary activity in Pāli,” in PBR, 5, 3 (1980), p. 85. ↩

1266. See Malalasekera, PLC, p. 212. ↩

1267. Ed. L. Feer, JPTS 1884, pp. 152-56. ↩

1268. See Geiger, PLL, 40.2. It is included in the list of Pāli texts originating in Cambodia by H. Saddhatissa, “Pali literature in Cambodia,” in JPTS 1981, pp. 186-87. ↩

1269. See Law, HPL, p. 629. ↩

1270. See p. 41 above. ↩

1271. Bv XXVII 19. ↩

1272. Ed. J. P. Minayeff, JPTS 1886, pp. 33-53; E. Leumann, Maitreya-samiti, Strassburg 1919, pp. 184-226. ↩

1273. Gv 61, 1. ↩

1274. Although Buddhaghosa mentions the names of Metteyya’s father and mother (Vism 434, 28-30 = As 415, 9-12), he does not quote Anāg 96, as Mrs Rhys Davids (Vism. p. 761) seems to imply, and Anāg is not, therefore, earlier than Buddhaghosa. ↩

1275. See Minayeff, JPTS 1886, p. 35. ↩

1276. Which is called Mahāpakaraṇa. ↩

1277. See Minayeff, JPTS 1886, p. 37. ↩

1278. See Moh 202, 3-22. ↩

1279. Again called Mahāpakaraṇa. ↩

1280. Ed. H. Saddhatissa, SBB London, 1975. ↩

1281. See Saddhatissa, Dbk, p. 9. ↩

1282. See ibid., p. 103. ↩

1283. See ibid., p. 19. ↩

1284. See ibid., p. 19. ↩

1285. Ed. and tr. (French) F. Martini, BEFEO 36, 2 (1936), pp. 287-390. The work named Dasabodhisattuppattikathā which is added in an appendix (pp. 391413) would seem to be the text labelled “D” by Minayeff (JPTS 1886, pp. 39-40). It would appear that Saddhatissa did not know of this publication, for he states (Dbk, pp. 3-4) that despite a search he was unable to find this manuscript. ↩

1286. See Minayefr, JPTS 1886, pp. 38-39. ↩

1287. Gv 72, 11. ↩

1288. See Malalasekera, DPPN, I, p. 66. ↩

1289. See Saddhatissa, Dbk, pp. 4-5. ↩

1290. Vin I 199-252. ↩

1291. See D I 12, 1-12. ↩

1292. Cf. civara-pindapāta-senāsana-gilānapaccayabhesajja-parikkhārā (Vin III 132, 10-11). ↩

1293. Mhv V 224-25. ↩

1294. Cikichā katā manusacikichā ca pasucikichā ca (Rock Edict II (A)). ↩

1295. Mhv XXXVII 146. See also J. Jolly, Medicin, Strassburg 1901, p. 15. ↩

1296. As Bode (PLB, p. 108 n. 7) suggests, following Turnour’s suggestion that sāratthasañgaha is the title of a book. Geiger (Cül tr., I, p. 13 n. 7) rejects the suggestion. The Sāratthasañgaha mentioned at Pagan is probably that by Buddhappiya or Siddhatthathera (see Malalasekera, DPPN, II, p. 1106). ↩

1297. Mhv LXXIII 34-54. ↩

1298. Mhv XCVII 58. Saranaṃkara is also reported to have extended the Bhesajjamañjūsā. See Malalasekera, PLC, pp. 215-16. ↩

1299. See Malalasekera, PLC, p. 215. ↩

1300. See Bode, PLB, p. 108. ↩

1301. Mhv XCVII 60. ↩

1302. See p. 129 above. ↩

1303. Ed. E. Senart, JA 1871, pp. 193-544. ↩

1304. See Franke, Gramm, pp. 17-19. ↩

1305. See H. Scharfe, Grammatical Literature, History of Indian Literature, V, 2, Wiesbaden 1977, p. 194 n. 8. ↩

1306. See Geiger, PLL, 30. ↩

1307. See Franke, Gramm, pp. 14-20. ↩

1308. See Malalasekera, PLC, pp. 179-80. ↩

1309. See Bode, PLB, pp. 17-18, and Franke, Gramm, p. 23. ↩

1310. See p. 153 above. ↩

1311. Gv 74, 1-3. ↩

1312. See Bode, PLB, p. 36, although Gv 74, 1-2 states that Dhammānanda wrote the tīkā on the Kaccāyana-sāra. ↩

1313. See Malalasekera, PLC, p. 221. ↩

1314. In the colophon to the text, quoted by Malalasekera, PLC, p. 220. ↩

1315. See p. 159 above. ↩

1316. Ed. and tr. L. Lee, The Orientalist, 2, 1892. ↩

1317. See Malalasekera, PLC, pp. 243-44. ↩

1318. See p. 179 below. ↩

1319. Gv 62, 9 and 71, 6. ↩

1320. See Godakumbura, Cat, p. 69. ↩

1321. See Malalasekera, PLC, p. 109 for the date of Vācissara. ↩

1322. Ed. H. Smith, 5 vols., Lund 1928-66. ↩

1323. See Bode, PLB, p. 16. ↩

1324. See ibid., p. 17. ↩

1325. Sadd 926, 9. ↩

1326. See Geiger, PLL 48, who states that the Moggallāyana-pañcikā is lost. He is followed in this statement by Scharfe (op. cit. [in n. 293], p. 195), but CPD, I, Epilegomena p. 55* lists an edition of it, as does Scharfe himself (p. 208). ↩

1327. See Malalasekera, PLC, p. 186. ↩

1328. See Franke, Gramm, pp. 34-43. ↩

1329. See Malalasekera, PLC, p. 231. ↩

1330. See p. 158 above. ↩

1331. See p. 6 above. ↩

1332. For a more detailed account of the Pāli grammarians, see Franke, Gramm, pp. 1-64, and for a bibliography see D. L. Barua, in IC 15, pp. 194-202. ↩

1333. See Scharfe, op. cit. (in n. 293), p. 195. ↩

1334. Ed. (with the Dhātupāṭha) D. Andersen and H. Smith, Copenhagen 1921. ↩

1335. Dhātum 52, 11. ↩

1336. See Geiger, PLL, 50. ↩

1337. Ed. (with Dhātumañjūsā) D. Andersen and H. Smith, Copenhagen 1921. ↩

1338. See Geiger, PLL, 52. ↩

1339. See pp. 19 and 85 above. ↩

1340. Cf. evam pi so gato ti Ganthipadatthavaṇnanāyaṃ vuttam (Mhv-t 148, 5). ↩

1341. See Pet 96, 4-23 and Nett 53, 1-56, 24. ↩

1342. Ed. W. Subhuti, Colombo (2nd ed.) 1883. Index compiled by W. Subhuti, Abhidhānappadīpikāsūci, Colombo 1893. ↩

1343. Gv 62, 3. ↩

1344. See Geiger, PLL, 51, and Malalasekera, PLC, p. 188. ↩

1345. See Geiger, PLL, 51. ↩

1346. See W. Stede, PED, Afterword, p. 734. ↩

1347. See A. K. Warder, “Some problems of the later Pāli literature,” in JPTS 1981, pp. 204-205. ↩

1348. See Geiger, PLL, 51. ↩

1349. See Malalasekera, PLC, p. 189. ↩

1350. Ed. W. Subhuti, as an appendix to his ed. of Abh (see n. 330). ↩

1351. See Geiger, PLL, 51. ↩

1352. Gv 61, 15-19, although the author of the Vuttodaya is called Vuttodayakāra. ↩

1353. Vutt 134. ↩

1354. Ed. and tr. G. E. Fryer, JAS 44 (1875), pp. 91-125. ↩

1355. Subodh 2. ↩

1356. See Malalasekera, PLC, p. 204, and Gv 62, 6. ↩

1357. Ed. and tr. G. E. Fryer, Calcutta 1877. See also R. Siddhartha, Sangharakkhita’s Vuttodaya: a study of Pali metre, Journal of the Department of Letters, University of Calcutta, Vol. 18 (1929), No. 7 (repr. Dehli 1981). ↩

1358. Vutt 2-3. ↩

1359. See Malalasekera, PLC, p. 198. ↩

1360. See ibid., p. 198. ↩

1361. Gv 64, 28. ↩

1362. Gv 64, 27. ↩

1363. See Fryer, Vutt, p. 4. The date given there for Vepulla (c. A.D. 1212) must be too early. ↩

1364. Gv 61, 17-18. ↩

1365. See p. 169 below. ↩

1366. See pp. 27 and 131 above. ↩

1367. Gv 61, 20-21. ↩

1368. Ed. E. Müller, JPTS 1883, pp. 88-121. ↩

1369. Khuddas L 5. ↩

1370. Ed. E. Müller, JPTS 1883, pp. 122-30. ↩

1371. Quoted by Müller (JPTS 1883, p. 87) where, perhaps by mistake, he reverses the authors’ names. ↩

1372. E.g. T. W. Rhys Davids in JPTS, 1883, p. xiv. ↩

1373. See Müller, JPTS, 1883, p. 87. ↩

1374. See Geiger, PLL, 27. ↩

1375. See Malalasekera (PLC, p. 76) who, however, dates Parakkamabāhu to A.D. 1065 . ↩

1376. See p. 167 above. ↩

1377. Gv 61, 19. ↩

1378. See Müller, JPTS 1883, pp. 131-32. ↩

1379. See ibid., p. 87. ↩

1380. See Rhys Davids, JPTS 1883, p. xiv. ↩

1381. See Müller, JPTS 1883, p. 86. ↩

1382. See P. S. Jaint, The Jaina path of purification, Berkeley 1979, p. 160 n. 5. ↩

1383. Ed. H. Saddhatissa, PTS London, 1965. ↩

1384. See Saddhatissa, Upās, p. 1 n. 1, and Upās 123, 8. ↩

1385. See Saddhatissa, Upās, p. 45. ↩

1386. See ibid., p. 3. ↩

1387. For a list of quotations and allusions, see Saddhatissa, Upās, pp. 115-22. ↩

1388. See Saddhatissa, Upās, pp. 104-11. ↩

1389. Upās 357, 3-4. ↩

1390. See p. 160 above. ↩

1391. See Saddhatissa, Upās, pp. 63-64. ↩

1392. Vin I 106-11. ↩

1393. Sp 1035, 23-1056, 30. See also Kkh 59, 23-60, 24. ↩

1394. Mhv-t 362,6 . ↩

1395. See Malalasekera, Mhv-t, pp. lxviii-lxix. ↩

1396. See Mhv, App. B, pp. 331-33. ↩

1397. See Geiger, Mhv tr., p. 111 n. 3. ↩

1398. Gv 62, 13-14. ↩

1399. See Bode, PLB, p. 18, although Geiger (PLL, 33) states that he wrote a tīkā upon the Sīmālañkārasaṅgaha. ↩

1400. See J. Dhirasekera, “A historical introduction to the Sīmālañkārasaṅgaha of Vācissara Thera”, in Buddhist Studies (Bukkyo Kenkyu), 1 (1970), pp. 73-76. ↩

1401. Ibid., p. 74 n. 1. ↩

1402. Ibid., p. 73 n. 11. ↩

1403. See Bode, PLB, pp. 38-39. ↩

1404. See Dhirasekera, op. cit. (in n. 388), pp. 74-75. ↩

1405. See Malalasekera, PLC, p. 251. ↩

1406. Ed. J. P. Minayeff, JPTS 1887, pp. 17-34. ↩

1407. See p. 144 above. ↩

1408. Sammāsambuddhassa parinibbānato samvaccharagananena catucattālisādhikam tisatadvisahassam sampatte (Sīmāv 18, 13-14) = A.D. 1800. ↩

1409. At Sās 33, 24-25 this is ascribed to Damilaraṭ̣thavāsī Kassapathero. It is not clear whether this is the same Kassapa who wrote the Mohavicchedanī, but since Gv 60, 33-61, 2 does not include it among the works of Kassapa, we are probably dealing with two authors of the same name. ↩

1410. See Malalasekera, PLC, p. 311. ↩

1411. See Hultzsch, Inscriptions of Asoka, Oxford 1925, pp. 172-74. ↩

1412. See Winternitz, HIL, pp. 606-609. ↩

1413. Ed. R. P. Chaudhuri and D. P. Gula, Calcutta 1957. See Geiger, PLL, 19. ↩

1414. See Bode, PLB, p. 5 n. 1. CPD Vol. I, Epilegomena p. 46* ascribes it to Ariyavamsa. ↩

1415. For the list of contents see CPD Vol. I, Epilegomena pp. 95*-96*. ↩

1416. See Geiger, PLL, 39. 1. Malalasekera, PLC, p. 229 calls the text Sāratthasangaha. ↩

1417. For a list of contents see Godakumbura, Cat, pp. 60-61. ↩

1418. See pp. 159 and 164 above. ↩

1419. See Godakumbura, Cat, p. 61. ↩

1420. See Malalasekera, PLC, pp. 75-76, and Bode, PLB, pp. 3-4. ↩

1421. E.g. parittasutta “protective thread” (Mhv VII 14). ↩

1422. Vin II 109-10. ↩

1423. D III 194-206. ↩

1424. Adhivāsesi Bhagavā tuṇhi-bhāvena (D III 195, 24). ↩

1425. Puna Bhagavatā parittā ca udditthā (Mil 150, 27). ↩

1426. Ps IV 114, 6-7. ↩

1427. Vism 414, 21-22. ↩

1428. See Dhp-a III 441; Pj I 164; Pj II 278. ↩

1429. Mhv XXXVII 195. ↩

1430. See Godakumbura, Cat, pp. 25-26. ↩

1431. See Geiger, PLL, 17. For a paritta text from South-East Asia see Padmanabh S. Jaini, “Mahādibbamanta: a Paritta manuscript from Cambodia”, in BSOAS 28 (1965), pp. 61-80. ↩

1432. See CPD, Vol. I, Epilegomena pp. 93*-95*. ↩

1433. See p. 52 above. ↩

1434. See p. 173 above. ↩

1435. Ed. and tr. (French) E. Denis, La Lokapañnatti et les idées cosmologiques du Bouddhisme ancien, 2 vols., Lille 1977. ↩

1436. See Denis, Lp, I, p. vii. ↩

1437. See ibid., p. ix. ↩

1438. Lp 157-59. ↩

1439. See Denis, Lp, I, p. ii. ↩

1440. See ibid., pp. xvi-xvii. ↩

1441. See ibid., p. xvii. ↩

1442. Ed. Bangkok, 1979. See also H. Saddhatissa, “Pāli literature of Thailand,” in Cousins, BSIBH, p. 217. ↩

1443. See L. Sternbach, Subhāṣita, gnomic and didactic literature, A history of Indian literature, IV, 1, Wiesbaden 1974, pp. 41-42. Niti verses are, however, found in the Lokaneyya-pakarana, which probably orginated in North Thailand. See Padmanabh S. Jaint, “Some niti verses of the Lokaneyya-pakarana”, in Hammadava Saddhatissa Felicitation Volume (forthcoming). ↩

1444. The Kāmandakī which is mentioned at Pagan (see Bode, PLB, p. 107) is the Sanskrit Kāmandakīyanītisāra. See Bechert, PNTB, p. lxvi. ↩

1445. Ed. J. Gray, Rangoon 1883; Bechert, PNTB, pp. 3-71. Tr. Gray, Prov. pp. 37-118. ↩

1446. Ed. Bechert, PNTB, pp. 72-98. Tr. Gray, Prov, pp. 1-36. ↩

1447. Ed. Bechert, PNTB, pp. 99-131. ↩

1448. Ed. ibid., pp. 132-60. Tr. Gray, Prov, pp. 119-41. ↩

1449. See Bechert, PNTB, p. lxxxii. ↩

1450. See ibid., p. lxxxiii. ↩

1451. See ibid., pp. lxxiii-lxxiv. ↩

1452. See ibid., p. lxxvii. ↩

1453. Tr. Gray, Prov, pp. 142-57. ↩

1454. See Bechert, PNTB, pp. lxv-lxvi. ↩

1455. See ibid., pp. lxvi-lxvii. ↩

1456. See ibid., p. lxxiv. ↩

1457. See ibid., p. lxxv. ↩

1458. Ed. P. S. Jains, PTS London, Vol. I, 1981; Vol. II, to appear in 1983. ↩

1459. See Jains, PJ, I, p. v. ↩

1460. Cf. PJ I 48, 7-20 with Ja VI 494, 14-495, 19. ↩

1461. Cf. PJ I 76 with Ja VI 35. ↩

1462. See P. S. Jaini, “The story of Sudhana and Manoharā: an analysis of the texts and the Borobudur reliefs,” in BSOAS 29 (1966), pp. 533-58. ↩

1463. See Jaini, PJ, I, p. v. ↩

1464. See G. Terral(-Martini), “Samuddaghosajātaka: conte pali tiré du Paññāsajātaka”, in BEFEO 48 (1956), pp. 249-351. ↩

1465. See ibid., pp. 310-38. ↩

1466. See p. 144 above. ↩

1467. See G. Martini, “Brapaṃsukūlānisamsam,” in BEFEO 60 (1973), pp. 55-78. ↩

1468. See G. Martini, “Un jātaka concernant le dernier repas de Buddha,” in BEFEO 59 (1972), pp. 251-55. ↩

1469. Ed. and tr. R. F. Gombrich, in Bechert, BCSRS, pp. 281-303. ↩

1470. Saddhamma-s 66, 5. ↩

1471. See J. Legge (tr.), A record of Buddhistic kingdoms, Oxford 1886, pp. 56-57. ↩

1472. See Gombrich, op. cit. (in n. 457), p. 302 n. 44. ↩

1473. See pp. 103 and 120 above. ↩

1474. Ed. N. SaddhĀnanda, JPTS 1890, pp. 21-90. Tr. B. C. Law, A manual of Buddhist historical traditions, Calcutta 1941. ↩

1475. Saddhamma-s 90, 7-10. It has been suggested that Dhammakitti was a Thai. See G. Ciedès, “Note sur les ouvrages pâlis composés en pays Thai,” in BEFEO 15,3 , p. 43. ↩

1476. See Malalasekera, PLC, p. 245. ↩

1477. Catuttham samgaham katvā (Saddhamma-s 45, 11). See p. 10 above. ↩

1478. See Geiger, PLL, 39.2. ↩

1479. See Malalasekera, PLC, p. 246 n. 2. ↩

1480. Saddhamma-s 66, 5-7. See p. 179 above. ↩

1481. See Malalasekera, PLC, p. 246. ↩

1482. Ed. J. P. Minayeff, JPTS 1886, pp. 54-80. Index compiled by M. Bode, JPTS 1896, pp. 53-86. ↩

1483. Iti pāmojjatthāyāraññavāsinā Nandapaññācariyena kato Cullagandhavamso (Gv 80, 5-6). ↩

1484. See Winternitz, HIL, p. 190. ↩

1485. See Malalasekera, DPPN, I, p. 748. It is, however, possible that culla goes with gandha, not vamsa, and the work is intended to be a history of the “small books” as opposed to the mahāgandhas, which would be the canonical texts. ↩

1486. See p. 15 above. ↩

1487. Katame tividhanāmakācariyā. Mahākaccāyano tividhanāmo. katame gandhā Kaccāyanena katā. Kaccāyanagandho Mahāniruttigandho Cullaniruttigandho Nettigandho Pefakopadesagandho Vaṇnanitigandho ti ime cha gandhā Mahākaccāyanena katā (Gv 59, 11-14). ↩

1488. Gv 60, 30-31. ↩

1489. Gv 62, 3-4. He thus separates this Moggallāna from the author of the grammar (see p. 166 above). ↩

1490. Ed. M. Bode, PTS London, 1897. Tr. B. C. Law, SBB London, 1952. ↩

1491. See p. 12 above. ↩

1492. See V. B. Liebermann, “A new look at the Sāsanavamsa,” in BSOAS 39 (1976), pp. 137-49. ↩

1493. Sās 30-31. ↩

1494. See Bode, Sās, p. 10. ↩

1495. See ibid., p. 10. ↩

1496. See Liebermann, op. cit. (in n. 480), pp. 148-49. ↩

1497. See p. 172 above. ↩

1498. See Malalasekera, PLC, p. 311. ↩

1499. See p. 12 above. ↩

1500. See p. 180 above. ↩

1501. See Geiger, PLL, 44.6. ↩

1502. See G. Cœdès, “Note sur les ouvrages pāli composés en pays Thai,” in BEFEO 15, 3 (1915). pp. 39-46 and H. Saddhatissa, “Pāli literature of Thailand,” in Cousins, BSIBH, pp. 211-25; “Pali literature in Cambodia”, in JPTS 1981, pp. 178-97; and “Pali literature from Laos,” in Narain, SPB, pp. 327-40. ↩