Pramāṇa · Search · About · API

The History of Buddhist Thought c28: Appendix I. The Scriptures

The History of Buddhist Thought c28

Language
English

Contents

Appendix I. The Scriptures

c28.2

The changes that have taken place in the conception of Buddhism as a religion and a philosophy are largely due to the piecemeal way in which the original documents have been brought to light. When Burnouf in 1844 issued his *Introduction à l'histoire du Buddhisme indien*, he mentioned as his chief predecessors Abel Rémusat,¹ the Chinese scholar, and the Moravian missionary Isaac Jacob Schmidt² as having thrown more light on the origin of Indian Buddhism than all those who up to that time had undertaken the study of the subject. Yet Rémusat's investigations were limited to Chinese works, and Schmidt, a scholar whose great merits have never been fully recognized, drew all his information from Mongolian and Tibetan translations. Burnouf was the first to investigate some of the Sanskrit originals, and his task was made possible through the enterprise of B. H. Hodgson, who was British Resident in Nepal from 1833 to 1843. During this time Hodgson sent from Nepal a large number of copies of Sanskrit manuscripts to various libraries in Europe, and it was chiefly on the basis of those sent to the Société Asiatique of Paris and to himself that Burnouf wrote his *Introduction*. But these documents were mainly representative of Mahāyāna, and their relation to the earlier Pāli and Sanskrit works was quite unknown.

c28.3

¹ Jean Pierre Abel Rémusat (1788–1832), professor of Chinese at the Collège de France.

c28.4

² I. J. Schmidt (1799–1847). The latest account of his work is by F. Babinger in *Festschrift für Friedrich Hirth*, Berlin, 1920.

c28.5

Burnouf was well aware of the importance of Pāli for the prosecution of his subject. In fact everything of a chronological nature that he had to say came from Pāli works. He was (with Lassen) the first to make the Pāli language known in Europe,³ and his unpublished papers show that he had copied Pāli manuscripts and inscriptions, and had made extensive studies in grammatical, doctrinal, and historical works.¹ But the whole of the Pāli Canon was still in manuscript, and although George Turnour of the Ceylon Civil Service (1799–1843) had issued translations of portions, and in 1837 had published the text and translation of the first thirty-eight chapters of the *Mahāvaṃsa*, it was not till 1855 that any portion of the Canon was published in Europe. Since then, through the labours of V. Fausböll, H. Oldenberg, and especially T. W. Rhys Davids, the whole of the Canon of the Theravāda school is now printed.²

c28.6

³ *Essai sur le pali*, par E. Burnouf et Chr. Lassen, Paris, 1826.

c28.7

¹ *Papiers d’Eugène Burnouf conservés à la Bibliothèque Nationale*, Paris, 1899.

c28.8

² There are now also editions printed in Ceylon and Burma, and especially in Siam. The last edition of the splendid text of the Canon, issued under the patronage of the King of Siam, was published at Bangkok in 1926–8 and the commentaries in 1920.

c28.9

Other forms of a complete Canon, those of the Sarvāstivāda schools, once existed in Sanskrit. These as a whole have been lost, but translations exist in Tibetan and Chinese, and the Sanskrit originals of portions have survived and are gradually being published.

c28.10

The remains of another important Canon, that of the Lokottaravādins, exist in the *Mahāvastu*. This is an Avadāna, a form of literature later than the suttas, based on the Vinaya of the Lokottaravādins, and it contains a large number of suttas and other portions of the Canon of that school. The great development known as Mahāyāna led to new schools and the production of a great number of new canonical works. These also exist in translation in Chinese and other languages, and those which survive in Sanskrit are being studied and made accessible.

c28.11

There have been disputes between scholars as to the relative value of the Sanskrit and Pāli sources. Burnouf’s documents belonged chiefly to late schools, and it was impossible in his time to distinguish those portions which really were comparatively early. Then the Pāli Canon was discovered. It was more complete than anything else that had been found. It claimed, like the Sanskrit works, but with more vraisemblance, to be the actual word of Buddha, and it was on the material of this Canon that expositions of Buddhism were written by Pāli scholars. The protests of the Sanskritists were largely unavailing; as a matter of fact

c28.12

they were unable to point to a Sanskrit Canon at all comparable to the Pāli. But the grounds of their objection were well put by L. de la Vallée Poussin.

c28.13

This scholar pointed out that the Pāli scholars relinquished the examination of Northern sources, and took no account of them. “They are passionately attached to the exegesis of the Southern [i.e. Pāli] Scriptures, which are in appearance more archaic and better documented. The results of these labours are of the highest importance, both for the history of religions in general, as well as for that of Buddhist and Indian ideas. Oldenberg’s book is a perfect exposition : Pāli Buddhism cannot be better described, or the intellectual and moral factors more artistically demonstrated, and the idea which a Sinhalese doctor makes of his religion and destiny more precisely set out. Oldenberg’s error was to entitle his book : *Buddha, his life, his doctrine, his community*. He should have added, ‘according to Pāli sources and the principles of the Sinhalese Church.’ Now by commenting on the Sinhalese traditions without exactly fixing their date and character there is a risk of falsifying the history of Buddhism in its general spirit and in its very signification. . . . Too easily persuaded that they know primitive Buddhism, European [Pāli] exegetes regard the different Canons of the Churches of the North as almost modern compositions, in which heretical, adventitious, and heterogeneous doctrines have become definite in the course of centuries.” ¹

c28.14

¹ *Bouddhisme, études et matériaux*, pp. 2 ff., London, 1898.

c28.15

To estimate this accusation fairly it is necessary to recognize to what extent the Pāli exegetes were right. It is a fact that there are numerous works in Sanskrit which are “almost modern compositions”. It is also true that in them there is much that is heretical, adventitious, and heterogeneous. This is not mere modern theory. It is admitted by the authors of these late works themselves. They taunted the older schools with clinging to the letter rather than the spirit. They even admitted that their opponents’ teaching really was the Buddha-word, but maintained that Buddha taught it to the simple-minded merely as a preliminary, and meant it to be superseded by the higher doctrine, the new goal of life and the new philosophical conceptions which they called Mahāyāna. Stcherbatsky declares that Buddhism “resulted, 500 years after the demise of the Master, in what may be called a quite new religion, reposing on a quite different philosophic foundation”.¹

c28.16

¹ *The conception of Buddhist Nirvāna*, p. 4, Leningrad, 1927. Nevertheless L. de Vallée Poussin, at least as late as 1902, continued to hold that the Mahāyāna was as old as the Sinhalese doctrine. “Nous ne pensons pas que les traditions, les vinayas et les dogmes du Grand Véhicule, ni même ceux du Véhicule tantrique soient moins archaïques ou moins anciens.” *Journ. As.*, ii, 1902, p. 238.

c28.17

If this were all, it would appear that the Pāli scholars were right. But to show that the Pāli sources are older does not prove that they are primitive. These scholars assumed that all the works in Sanskrit showing a connection with Pāli must have been translated from Pāli. But these related Sanskrit works are distinct both from the Mahāyāna compositions and the Pāli, and belong to the Canon of the Sarvāstivāda and its branches. These schools had translated them, not from the Pāli of the Theravādins, but from their own Scriptures in the Prakrit dialect in which they were at first composed. It is now possible to see how they all developed from one nucleus of doctrine preserved by memory, but were gradually differentiated owing to the precarious method of their preservation in widely separated communities.²

c28.18

² This is now recognized by de la Vallée Poussin. “Nous parlons, au singulier, du Canon. Il n’est pas douteux qu’un corps considérable d’Écritures a servi de base aux deux canons que nous connaissons, canons des sectes sthaviriennes, le canon de langue pâlie et le canon sanscrit des Sarvâstivâdins : on peut désigner ce corps d’Écritures sous le nom de Canon bouddhique.” *Le dogme et la philosophie du Bouddhisme*, p. 97.

c28.19

The language of the Pāli Canon is described by the commentators as Māgadhī (the language of Magadha). There can be no doubt that this was the original language, but it is now usually held that Pāli as we know it must have developed in the west of India, and the probability is that the present Pāli Canon comes from a Buddhist community of monks in the region of Ujjenī (Ujjain), who adapted it to their own dialect. It still retains traces of an earlier dialect in the verse passages. The Sarvāstivāda schools went further, and turned the whole of their Canon into Sanskrit, though in this case also traces of a popular dialect, a form of Prakrit, appear in the metrical passages. The Mahāvastu, which also aims at being Sanskrit, preserves many more traces of its original Prakrit.

The Tipiṭaka

c28.s1.1

Franke has stated “that the books of the Canon as a *whole* are not authentic; that the Canon was not composed and compiled in one and the same period of time . . . that even the first two Piṭakas (to say nothing of the Abhidhamma) cannot possibly have been presented as finished before either the ‘first’ or the ‘second’ Council, even if these events took place at the intervals assigned to them.”¹ This conclusion, in the sense that we do not possess a verbatim report of the discourses, is so obvious that it scarcely needed stating, but it is quite negative. We know that the Buddha-word—everything that Buddha was held to have given as a rule or taught as a doctrine—was preserved for centuries by memorizing. The statement in the Pāli Chronicles that the Scriptures were first written down in Ceylon in the reign of Vaṭṭagāmaṇi (29-17 B.C.) is of little significance. There may have then been such an official recension, but it is probable that much had been written down before. However that may be, the practice of learning it by heart was for long the only means of preserving it; there was thus no means of preserving a definite order, or even of being certain that any discourse was really the word of the Master.²

c28.s1.2

¹ *JPTS.*, 1908, p. 2.

c28.s1.3

² One of the earliest means of classifying them was to divide them into small groups (usually ten) and to record the titles of the suttas of each group in a verse called an *Uddāna*.

c28.s1.4

The disciples felt these difficulties, for we find in the Scriptures a discourse attributed to Buddha giving four rules for determining the genuineness of any doubtful passage. We can accept the fact that the monks found these rules necessary and made use of them, without supposing that Buddha, forseeing the difficulties to come, actually formulated them. The rules are, (1) if a monk says that he has heard directly from the Lord anything as being the Dhamma, the Vinaya, or the teaching of the Master, it is to be compared with the Sutta or shown to exist in the Vinaya. If it does not correspond, it is not the word of the Lord, and is to be rejected. The same method is to be used (2) if a monk says he has received it from an assembly (*sangha*) of monks with a leader, or (3) from a number of learned elders who have acquired the Āgamas, and who know by heart the Dhamma, the Vinaya, and the Mātikā, or (4) from such a single learned elder.¹

c28.s1.5

¹ Ang. ii, 167 ; Dīgha, ii, 123. This sutta is itself an example of how accretions could occur. The four rules at some period existed, and some pious monk, doubtless convinced that they were Buddha’s own rules, turned them into a sutta, which Buddha was supposed to have uttered a day or two before his death. Its unhistorical character is shown by the fact that it includes the Mātikā, and refers to the Āgamas. Āgama is the Sarvāstivāda term for the collections usually called Nikāyas in Pāli.

c28.s1.6

The division of the Scriptures into two classes, discipline and doctrine, was inevitable. But in these rules we find a third division, the Mātikā, a term now applied to the lists of subjects discussed in the class of Abhidhamma, but also used as a synonym of Abhidhamma. This last class, though not recognized by some schools and evidently later than the others, now forms the third part of the Triple Basket, the Tipiṭaka (Skt. Tripitaka). The term piṭaka as a division of the Canon is not in the Scriptures, but must be as old as the third century B.C., for the term peṭaki, “ reciter of the Piṭaka ”, occurs in inscriptions of that date.² The term pañca-nekāyika, “ reciter of the five Nikāyas,” also occurs, showing that the division of the Suttas into Nikāyas or Āgamas already existed. The arrangement of the whole Tipiṭaka is as follows :—

c28.s1.7

² This and other terms are discussed by Rhys Davids in Dial. i, p. xii. In Av. Śat. No. 84 tripiṭa as the name of a knower of the three Piṭakas occurs.

A. Vinaya-Piṭaka

c28.s1.t1.1

The rules of discipline, contained in two chief sections followed by a minor work.

c28.s1.t1.2

I. Sutta-vibhanga.³ Consisting of the Pātimokkha rules, each rule being followed (1) by a verbal commentary explaining each word of the rule, (2) an account of the incident which led to the promulgation of the rule, sometimes forming extensive legends, (3) special cases and exceptions. The portion of the Pātimokkha which deals with the summoning of the Uposatha is given in the Mahāvagga under the rules for holding Uposatha.

c28.s1.t1.3

³ The arrangement of this portion can be seen from the analysis of the Pātimokkha above, p. 15 ff.

II. The Khandhakas

c28.s1.t2.1

1. *Mahāvagga* (Great Series)

c28.s1.t2.2

(1) Rules for admission to the Order.
(2) Uposatha meeting and recital of the Pātimokkha.
(3) Residence during Retreat in the rainy season (*vassa*).
(4) The ceremony concluding Retreat (*pavāraṇā*).
(5) Rules for the use of leather for shoes, dress, and furniture.
(6) Medicine and food.
(7) The kaṭhina ceremonies for the annual making and distributing of robes.
(8) Material of robes, sleeping regulations, and rules for sick monks.
(9) The modes of executing official acts of the Order.
(10) Proceedings in case of dissensions.

c28.s1.t2.3

2. *Cullavagga* (Small Series)

c28.s1.t2.4

(1, 2) Rules for dealing with offences that come before the Order.
(3) Reinstatement of monks.
(4) Rules dealing with questions that arise.
(5) Miscellaneous rules for bathing, dress, etc.
(6) Dwellings, furniture, lodgings.
(7) Schism.
(8) Treatment of travelling monks, those living in the forest, etc., and the duties of teachers and novices.
(9) Exclusion from the Pātimokkha.
(10) The ordination and instruction of nuns.
(11) History of the first Council at Rājagaha.
(12) History of the second Council at Vesālī.

c28.s1.t2.5

III. *Parivāra*. A supplement containing summaries and classifications of the rules. It may be peculiar to the Theravādins, as the *Dīpavaṃsa* speaks of its rejection by the Mahāsanghikas, but other schools had a supplementary work.

c28.s1.t2.6

The Sarvāstivāda Vinaya follows the same general arrangement. The first portion, the *Vinaya-vibhaṅga*, corresponds to the *Sutta-vibhaṅga*; the second portion, the *Vinayavastu*, corresponds to the *Khandhakas*, and instead of the *Parivāra* it is followed by the *Vinayakshudraka*, “ minor Vinaya work,” and *Uttaragrantha*.¹ The last consists of a series of questions on Vinaya put to Buddha by Upāli.

c28.s1.t2.7

¹ This the order in *Mvyut*. 65. In the Tibetan the *Vinayavastu* comes first. It has seventeen divisions, which only partially correspond with the above. The two *Prātimokshas* and the *Bhikkhunī-vibhanga* are numbered separately.

c28.s1.t2.8

The *Vinayavastu* of the Mūla-Sarvāstivādins shows evident correspondences in arrangement and subject matter to the *Khandhakas* of the Pāli :—

c28.s1.t2.9

(1) *Pravrajyavastu* (admission to the Order).
(2) *Poshadhavastu* (Uposatha).
(3) *Varshavastu* (Retreat).
(4) *Pravāraṇavastu* (ceremony concluding Retreat).
(5) *Kaṭhinavastu* (material for robes).
(6) *Cīvaravastu* (rules for robes).
(7) *Carmavastu* (leather for shoes).
(8) *Bhaishajyavastu* (medicines).
(9) *Karmavastu* (official acts).
(10) *Pratikshayavastu* (possibly on dwellings and lodgings).
(11) *Kālākālasampadvastu* (on the right and wrong times).
(12) *Bhūmyantarasthacaraṇavastu* (rules for travelling monks).
(13) *Parikarmaṇavastu* (executing proceedings).
(14) *Karmabhedavastu* (probably corresponds to the Pāli section on different kinds of official acts, and the following section to different kinds of meetings).
(15) *Cakrabhedavastu*.
(16) *Adhikaraṇavastu* (rules dealing with cases that arise).
(17) *Śayanāsanavastu* (rules for beds).

B. Sutta-Piṭaka (Dhamma)

c28.s1.t3.1

The doctrinal portion, the Dhamma (Dharma), exists in four or five collections called in Pāli Nikāyas, in the other schools Āgamas. The separate discourses are usually known as suttas (sūtras) or suttantas (sūtrāntas), but an earlier term used in the suttas themselves is *dhammapariyāya*, “ section of doctrine.”² The separation of this portion into four or five collections is as early at least as the third century B.C., as the term *pañcanekāyika*, “knowing the five Nikāyas,” shows. This classification appears to have been a purely formal one due to convenience in committing to memory, the groupings being into (1) long suttas, (2) medium long, (3) grouped suttas, (4) suttas arranged on a numerical principle. It is not surprising that there should be other doctrinal works which did not fit into the fourfold scheme. The Theravādins formed them into a fifth Nikāya, but much of the material existed also in other schools, and among the Sarvāstivādins, a school with a long continuous growth in India, there was an enormous development, as will be seen in the discussion of the nine Angas. The Pāli order and arrangement is as follows :—

c28.s1.t3.2

² *Suttanta*, as has been shown by Mr. E. H. Johnston, is a collective term formed from *sutta*, but there is no real difference of meaning. The difference appears to be due to the usage of different classes of reciters of portions of the Scriptures. The term *sutta* (sūtra), lit. “ thread ”, refers in brahminical works to short aphoristic sentences intended to be committed to memory. With the Buddhists it is any part of this division of the Scriptures, which

c28.s1.t3.3

1. *Dīgha-Nikāya* (*Dīrghāgama*)

c28.s1.t3.4

The collection of long discourses, thirty-four suttas arranged in three *vaggas* or series.

c28.s1.t3.5

1–13. *Sīlakkhandha-vagga*.—Each of these has incorporated in it a list of moral rules known as the Silas. Most of the suttas describe the training of the monk in three stages, beginning with these moral rules, proceeding to the practice of concentration (*samādhi*), and ending with *paññā*, the full knowledge of the arahat. Several of them discuss the views of the brahmins on sacrifice and sacred knowledge, the doctrines of various religious schools, the value of caste and self-mortification, and expound leading doctrines.

c28.s1.t3.6

14–23. *Mahā-vagga*.—Most of these suttas are not properly discourses but legends. The most important is the *Mahā-parinibbāna-sutta*, an account of the last days and death of Buddha and the distribution of his relics. Two others are really portions or extensions of this, the *Mahāsudassana-sutta*, in which Buddha on his death-bed tells of his former existence as king Sudassana, and the *Janavasabha-sutta*, an extension of another discourse which Buddha delivered on his last

c28.s1.t3.7

may be a discourse or poem, but which is always given as a separate utterance of Buddha or a disciple. Its earlier sense is seen in the *Pātimokkha*, where it is applied to each Vinaya rule, and in the title *Sutta-vibhanga*. The term *dhammapariyāya* was used by Asoka in the form *dhammapaliyāya* in the edict in which he recommends portions of Scripture to the monks and nuns for study. *Inscriptions of Asoka*, ed. Hultzsch, p. 173. Dr. Walleser holds that *sutta* is from Skt. *sūkta*, “well spoken, a hymn,” and this certainly fits the discourses very well, but when the Canon was turned into Sanskrit nothing was known of this derivation, nor do the Jains recognize it.

c28.s1.t3.8

journey. Other legends are the *Mahāpadāna-sutta*, an account of the last seven Buddhas with the life of Vipassin Buddha down to the beginning of his preaching; *Mahā-govinda-sutta*, a previous life of Buddha as Mahā-govinda, recorded by a heavenly musician, who went up to the heaven of Brahmā to hear it; *Mahā-Samaya-sutta*, the “great Assembly” of gods who visit Buddha, and who are described by him in a long poem. The most important doctrinal discourses are the *Mahānidāna-sutta* on the Chain of Causation, the *Mahā-Satipaṭṭhāna-sutta*, on the four contemplations with commentary on the Four Truths, and *Pāyāsi-sutta* given by the elder Kumārakassapa, on the doctrine of a future life and reward of actions.

c28.s1.t3.9

24-34. *Pāṭika-vagga*.—This is the most miscellaneous section, and it indicates that definite doctrinal beliefs about the nature of a Buddha had become much developed by the time the collection was made. Two suttas show Buddha explaining the origin and development of the universe (after its periodical destruction) and the differentiation of castes. There is the legend of the universal king (*cakkavattin*), a prophecy about the next Buddha (Metteya or Maitreya), a discourse on the thirty-two bodily marks of a universal king (which are also the marks of a Buddha), and a spell to serve as protection against evil spirits, which was given to Buddha by the Four Great Kings. It became the Buddha-word when Buddha repeated it to the monks. The *Sigālovāda-sutta*, the one discourse addressed to a layman, is really a poem with a prose commentary. The last two, *Sangīti* and *Dasuttara*, are long lists of classified doctrinal terms, and are ascribed to Sāriputta.

c28.s1.t3.10

2. *Majjhima-Nikaya* (*Madhyamāgama*)

c28.s1.t3.11

The division of discourses of medium length, 152 suttas in fifteen vaggas. There is the same description as in the *Digha* of the life of the monk in the three stages of morality, concentration, and enlightenment. This in several suttas is brought into connection with the story of Buddha’s enlightenment and first preaching. There are several other legends of Buddha’s life, his descent from heaven and his birth (No. 123), his debates with the Jains (Nos. 35, 56, 58, 79, 101, 104), his foster-mother as a nun (No. 146), the schism of Devadatta (Nos. 29, 58), and several discussions on caste (Nos. 84, 93, 96). Some are in the form of commentaries, and several show a tendency to the Abhidhamma style of exposition with classified lists of terms. Twenty-one of them are attributed to disciples.

c28.s1.t3.12

3. *Saṃyutta-Nikāya* (*Saṃyuktāgama*)

c28.s1.t3.13

The division of “ connected ” suttas in five vaggas divided into groups (saṃyuttas) and these again into smaller vaggas of separate suttas. Many short poems and scraps of verse are collected in the first, called :—

c28.s1.t3.14

(1) *Sagātha-vagga*, “ series with verses.” They are grouped in saṃyuttas according to the characters appearing in them, the gods, the king of the Kosalas, Māra, etc.

c28.s1.t3.15

(2) *Nidāna-vagga*, this, like the following, does not indicate the subject of the whole, but only of the first ten saṃyuttas on the Nidānas, the twelve links of the Chain of Causation.

c28.s1.t3.16

(3) *Khandha-vagga*, beginning with suttas on the five khandhas.

c28.s1.t3.17

(4) *Saḷāyatana-vagga*, beginning with suttas on the six senses.

c28.s1.t3.18

(5) *Mahā-vagga*, “ the great series,” beginning with suttas on the Eightfold Path.

c28.s1.t3.19

4. *Aṅguttara-Nikāya* (*Ekottarikāgama*)

c28.s1.t3.20

*Aṅguttara* means “ (one) member beyond ”, and *ekottarika* “ having one (member) beyond ”. The division is purely numerical, beginning with suttas discussing lists of one thing, lists of two, three, etc., up to eleven. The first gives a list of the one sight, the one sound, the one scent, etc., which occupies the thought of a man or woman. The last is a list of the eleven good and eleven bad qualities of a herdsman, and the corresponding qualities of a monk. It is in eleven *nipātas* (groups), each divided into vaggas, which usually contain ten suttas.

c28.s1.t3.21

5. *Khuddaka-Nikāya*

c28.s1.t3.22

The division of minor works. The contents of this Nikāya have varied at different times. The following fifteen are at present recognized in Ceylon :—

c28.s1.t3.23

and the latter on the fifth vagga (*Pārāyaṇa*) and on the *Khaggavisāṇa-sutta* in the first vagga. They contain much Abhidhamma material.

c28.s1.t3.24

(12) *Paṭisambhidā-magga*. “The way of analysis,” an analysis of various concepts, knowledge, heresy, the practice of breathing while meditating, etc. In method and material it is really an Abhidhamma work.

c28.s1.t3.25

(13) *Apadāna*. Accounts in verse of the lives and previous lives of monks and nuns.

c28.s1.t3.26

(14) *Buddhavamsa*. “History of the Buddhas,” in which, in response to a question by Sāriputta, Buddha gives an account in verse of his first forming in the presence of Dīpankara Buddha the resolve to become a Buddha, with the life of Dīpankara and the succeeding twenty-four Buddhas, including an account of himself.

c28.s1.t3.27

(15) *Cariyā-piṭaka*. “The basket of conduct,” i.e. the conduct of the future Buddha in order to attain the ten Perfections (*pāramī*), forming thirty-five tales from the *Jātaka* versified. Only seven of the ten Perfections are illustrated.

c28.s1.t3.28

There has never been any general agreement even among the Theravādins about the number of works in this Piṭaka. The whole fifteen are mentioned in Buddhaghosa’s commentary on the Vinaya, but in the Chinese translation of this commentary No. 1 is omitted, showing that the original text of Buddhaghosa did not admit it.¹

c28.s1.t3.29

¹ No. 1, consisting of a series of short texts mostly from the suttas, was intended for novices. When they were once collected into a separate book it would be almost inevitable that it should be treated as Scripture.

c28.s1.t3.30

He also records two earlier traditions, one that the reciters of the Dīgha omitted the first and the last three, calling the rest the “Minor book” (*khuddaka-gantha*) which they placed in the Abhidhamma-piṭaka; the other that the reciters of the Majjhima included all but the first, and placed them in the Sutta-piṭaka. The Siam edition omits 6–10 as well as the last three.

c28.s1.t3.31

The four Āgamas of the Sarvāstivādins have been compared with the corresponding Nikāyas by Professor Anesaki,² and in more detail by Professor C. Akanuma.¹ The result is to show a close correspondence in the case of the Dīgha. The Chinese has thirty suttas, two of which are not found in the Pāli. Several of those in the Pāli Dīgha are in other books of the Chinese. The order of the suttas is different, but the same vaggas were known, for the ten in the Chinese which come last correspond to the Sīla-vagga. In the Majjhima there are 222 suttas as against 152 in the Pāli, and nineteen of them are not in the Chinese. The material also varies more than in the Dīgha, for many suttas of the Chinese Majjhima occur in the Pāli Saṃyutta and Anguttara, and *vice versa*. The differences are greatest in these two last. We may conclude that there was once a common stock of suttas learnt by heart. Probably differences in what was remembered had already begun between the two schools. The beginnings of the classification must have originated while the schools were in close connection, but it evidently went on and was completed in each independently.

c28.s1.t3.32

² *Some problems of the textual history of the Buddhist Scriptures*; and *The four Buddhist Āgamas in Chinese*. Trans. As. Soc. Japan, xxxv, parts 2, 3, 1908; *Le sagātha-vagga du Saṃyutta-nikaya et ses versions chinoises*. Muséon, 1905, pp. 23 ff.

c28.s1.t3.33

¹ The comparative catalogue of Chinese Āgamas and Pāli Nikāyas. (In Japanese.) Nagoya, Japan, 1929.

C. Abhidhamma-Piṭaka

c28.s1.t4.1

Unlike the two other Piṭakas the Abhidhamma of the Theravādins is quite distinct from that of the Sarvāstivādins. That of the Theravādins consists of seven independent works. The tradition is that they were thought out by Buddha in the fourth week after his enlightenment :—

c28.s1.t4.2

(1) *Dhammasaṅgaṇi*. “Enumeration of dhammas,” i.e. mental elements or processes.

c28.s1.t4.3

(2) *Vibhaṅga*. “Distinction or determination.” Further analysis of the matter of the foregoing.

c28.s1.t4.4

(3) *Dhātukathā*. “Discussion of elements.” On the mental elements and their relations to other categories.

c28.s1.t4.5

(4) *Puggalapaññatti*. “Description of individuals,” especially according to their stages along the Noble Path.

c28.s1.t4.6

(5) *Kathāvatthu*. “Subjects of discussion,” discussions and refutations of doctrines held by other schools.

c28.s1.t4.7

(6) *Yamaka*. “Book of pairs,” called by Geiger an applied logic. The subject matter is psychological, and the analysis is arranged in pairs of questions.

c28.s1.t4.8

(7) *Paṭṭhāna*. “Book of relations,” an analysis of the relations (causality, etc.) of things in twenty-four groups.

c28.s1.t4.9

The Abhidhamma of the Sarvāstivādins also consists of seven works,¹ one main work followed by six supplements, and unlike the Theravādin Abhidhamma they are not attributed directly to Buddha, but to seven disciples, four of whom are elders recognized by the Theravādins.

c28.s1.t4.10

¹ They now exist only in Chinese translations, and have been summarized and analysed by Professor Takakusu (*JPTS*. 1904–5, pp. 67 ff.). The Sanskrit names of six of the works omitting Vijñāna-kāya are given in *Mvyut*. 65. Bu-ston says that the Vaibhāshikas of Kashmir regard them as belonging to the word of Buddha (trans. by E. Obermiller, p. 49).

c28.s1.t4.11

(1) *Jñāna-prasthāna*, by Kātyāyanīputra. “The course of knowledge.” It is a long work, and discusses most of the philosophical concepts (bonds, karma, material elements, organs, concentration, views) involved in attaining knowledge.

c28.s1.t4.12

(2) *Saṅgīti-paryāya*, by Mahā-Kausṭhila or Śāriputra. “Section for recitation.” It consists of lists of doctrinal terms arranged numerically like the *Saṅgīti-sutta* in the *Dīgha*, on which Takakusu thinks it is modelled.

c28.s1.t4.13

(3) *Prakaraṇa-pāda*, by Vasumitra. “Work (on Abhidhamma).” A discussion on dhammas (doctrinal concepts). Vasumitra is the name of a well-known Sarvāstivādin author.

c28.s1.t4.14

(4) *Vijñāna-kāya*, by Devaśarman. “Group (of subjects) on consciousness.” Chinese tradition puts the author a century after Buddha’s death.

c28.s1.t4.15

(5) *Dhātu-kāya*, by Pūrṇa or Vasumitra. “Group (discussing) mental elements.”

c28.s1.t4.16

(6) *Dharma-skandha*, by Śāriputra or Maudgalyāyana. “Aggregate of dharmas,” discussing the chief philosophical concepts as in No. 1.

c28.s1.t4.17

(7) *Prajñapti-śāstra*, by Maudgalyāyana. “Book of instruction.” The instruction is on cosmological and secular matters, the qualities of a Bodhisattva and a Buddha, etc.

c28.s1.t4.18

Although there is some similarity in these titles to the Abhidhamma works of the Theravādins, the works are quite independent. When we consider that the other two Piṭakas have a common basis, it is possible to draw the important conclusion that the existing Abhidhamma works are much later than the contents of the first two Piṭakas, and arose when the two schools were definitely separated. Their sectarian character is also indicated by a description of the Abhidhamma Piṭaka quite different from these, which is given in the account of the first Council in the Dharmagupta Vinaya.¹ This account lets us know what the author of it imagined the primitive Abhidhamma to be, and he must have relied on his knowledge of the Dharmagupta Canon of his time. But although this school was a branch of the Sarvāstivādins, there is no recognition of the Abhidhamma as understood either by the Theravādins or Sarvāstivādins.

c28.s1.t4.19

¹ Quoted below, p. 278.

c28.s1.t4.20

The word abhidhamma is found in the suttas. It means “further dhamma” or “special dhamma”. In Majjh., i, 214, two monks are said to hold an abhidhamma discussion (abhidhammakathā). The same procedure is referred to in Ang., iv, 397, where a monk who duly knows the four Truths is asked a question about abhidhamma and abhivinaya, but leaves it alone and cannot expound it. Evidently an elaboration and analysis of the doctrinal principles is intended, just as abhivinaya would mean a casuistic discussion of the rules of discipline. There is no separate Abhivinaya, for the elaboration of the rules of discipline exists in the Vinaya itself. The codification of such principles requiring exposition would give rise to the mātikā (mātṛkā), the “lists”, which form the basis of the Abhidhamma books, and now form a kind of table of contents. The elaboration of these principles, their definition and their proof by adducing sutta passages would form Abhidhamma proper. This is in fact what we actually find in the two existing Abhidhamma collections. They are two sets of text-books compiled by two schools of teachers who independently undertook to expound the same subject matter.

The Nine or Twelve Angas

c28.s2.1

What we may conclude from the Angas throws an important light on the growth of the Sarvāstivāda Canon. In several places in the Pāli suttas² there is a list of nine divisions of

c28.s2.2

² Majjh. i, 130.

c28.s2.3

the Dhamma: *sutta*, *geyya*, *veyyākaraṇa*, *gāthā*, *udāna*, *itivuttaka*, *jātaka*, *abbhutadhamma*, *vedalla*. It does not appear to have been originally an actual classification of the works in the Scriptures, but a description of the respective types of *sutta*. Three of them, however, *udāna*, *itivuttaka*, and *jātaka* are the names of actual works. But the probability is that the terms were at first used generally like the rest to describe the character of the composition. There are still many *udānas* and *jātakas* in various parts of the Scriptures not included in the works of those names.

c28.s2.4

These nine are also included in an extended list of twelve *angas*, which have been quite arbitrarily called a *Mahāyāna* list. It is given in the *Mahāvyutpatti*, 62, and to the above list it adds *nidāna* and *avadāna* after *udāna*, and at the end *upadeśa*.¹ Also instead of *vedalla* it has *vaipulya*, which is placed after *jātaka*. As the *Mahāvyutpatti* contains much Sarvāstivāda material, it cannot be concluded that the list of twelve *angas* belongs to Mahāyāna. Several Mahāyāna works have lists of less than twelve. The *Lotus*, ii, 44 (p. 45), has a list of nine. So has the *Dharmasangraha* (62), and the *Kāraṇḍa-vyūha* has a list of ten and one of eleven. They all omit *avadāna*, which is a characteristic type of work of the Sarvāstivādins.

c28.s2.5

¹ It also reads *itivṛttaka* (thus-happened) for *itivuttaka*, doubtless a wrong sanskritization.

c28.s2.6

Buddhaghosa defines the *angas* more than once,² but it is evident that in the case of several of the items all he knows about them is a number of *suttas* which happen to bear the name of the *anga*. *Sutta* is said to include the Vinaya, certain *suttas* in verse, and “other utterance of the Tathāgata named *sutta*”. *Geyya* is *sutta* which contains *gāthās*, especially the first *vagga* of the Saṃyutta-nikāya. *Veyyākaraṇa* is the whole Abhidhamma-piṭaka and other utterances not among the other eight *angas*.³ *Gāthā*, “verses,” are the *Dhammapada*, *Thera-* and *Therī-gāthā* and those verses in the *Suttanipāta* not called *suttas*. *Udāna*, *itivuttaka*, and *jātaka* are the works so named. *Abbhutadhamma* consists of the *suttas* that describe “wonderful events”. *Vedalla* is described by giving a list of six *suttas*, two of which have the title *vedalla*. They are said to be in the form of question and answer and to give joy and satisfaction.¹

c28.s2.7

² Dīgha. com., p. 23.

c28.s2.8

³ For the Sarvāstivādins *vyākaraṇa* is a prophecy by Buddha about the future destiny of a person.

c28.s2.9

¹ The meaning is uncertain. In form vedalla (*vaidalya*) would be an abstract formed from vidala, like vaipulya from vipula. Skt. dala means a fragment, and it may be suggested that vidala, “ split open, expanded,” would mean fragmentary, and vedalla a sutta of fragments. This would suit the fact that those suttas are of a miscellaneous character and as it were made up of fragments. The term vaipulya is quite distinct, and appears to be a deliberate replacing of the earlier term. Although it is best known as a Mahāyāna term it was used, as will be seen, by the Dharmaguptas. It means abundance or fullness, and though the exact reason for its use cannot be proved, it was probably a term descriptive of the very extensive later suttas which were excluded from the four Āgamas.

c28.s2.10

The importance of the list of twelve angas lies in the fact that it shows the growth of other types of Scripture which could not be placed in the Piṭakas. That these types began not with Mahāyāna can be seen from the fact that they exist as Sarvāstivāda works. We cannot assume that the Sarvāstivādins borrowed a Mahāyāna form, for it is chiefly the Sarvāstivāda avadānas that were incorporated in the Mahāyāna Canon. We also find definite evidence for the Sarvāstivāda origin of avadāna and other additional terms in the Vinaya of the Dharmaguptas, a branch of the Sarvāstivādins. The passage is important enough to be quoted. It occurs in the account of the first Council, where Ānanda after reciting the Āgamas is said to have recited the Abhidharma Piṭaka.

c28.s2.11

So also he (Ānanda) replied concerning the Jātaka Sūtra, the “ good ” Nidāna Sūtra, the Vaipulya Sūtras, the Adbhuta (dharma) Sūtras, the Avadāna Sūtras, the Upadeśa Sūtras, the Ku-i Sūtra (Vyākaraṇa), the Dharmapada Sūtra, the Po-lo-yen Sūtra, the “ concourse of dangers ” Sūtra, the verses of the Holy One (Muni-gāthā), all these, composing the miscellaneous collection of Sūtras, he spoke of ; so also of others, in which difficulties and no difficulties in meaning were discussed, all these in their turns he spoke of, and so was collected the Abhidharma Piṭaka.²

c28.s2.12

² This is Beal’s version, *Abstract of four lectures*, p. 79, with the omission of some of his comments. The Po-lo-yen is probably the Pārāyaṇa of the Suttanipāta, and the “ Concourse of dangers ” (rather, “ dangers of intercourse ”) the Khaggavisāṇasutta.

c28.s2.13

It is not clear whether the “ miscellaneous collection of sūtras ” is a part of the Abhidhamma, or whether a fifth Āgama is implied, but the point is not here important. We find in it Nidāna, Avadāna, and Upadeśa, the three forms supposed to belong to Mahāyāna.

The Sarvāstivāda Avadānas

c28.s3.1

It was natural in a literature preserved so long by memory,

c28.s3.2

and one in which original composition continually went on, that there should be spurious and apocryphal works claiming a place in the Canon. It would not be easy to insert an additional sutta in the four Nikāyas after they had been classified and numbered, but there was always a number of unclassified works, and even when the Theravādins placed them in an additional Nikāya no agreement was ever reached as to what it contained. The literary activity of the Sarvāstivādins was even greater, for they continued to exist in India with a great system of education and centres of literary activity. The three forms of composition, nidāna, avadāna, and upadeśa became definite types of literature.¹ Of these, avadāna is the most important. Whatever the derivation of the word may be, there is no doubt that it means "heroic feat, glorious achievement", but this describes only the contents, and does not tell how as a class of composition Avadāna is to be distinguished from the rest. Yet the actual Avadānas have a very distinct character. They are either single works or a number of these works made into a collection. One of the best known is the Divyāvadāna, the divine Avadāna, and consists of thirty-eight tales. A number of them are evidently tales which illustrate the occasion leading to the formulation of a Vinaya rule, and S. Lévi has proved that most of the work is compiled from the tales in the Vinaya of the Mūla-Sarvāstivādins. Three other important Avadānas were evidently compiled in the same way, the Avadāna-śataka, the Karma-śataka, and the Damamūka.² It was easy for such works to be admitted as Scripture, for the tales of which they were composed mostly originated in canonical works, and any composition which contained sayings of Buddha might be reckoned as the Buddha-word.

c28.s3.3

¹ The names also occur in Pāli. But nidāna "cause, occasion", has remained the name of an introductory composition explaining the occasion or motive of a work. The Pāli word apadāna makes it probable that the form avadāna is a Prakrit form representing apa-, but the derivation is uncertain. As a type of literature the Pāli Apandāna has nothing to do with the Sarvāstivādin Avadāna. There is the Mahāpadāna-sutta in the Dīgha, and the Apadāna, a work in verse in the Khuddaka-nikāya, neither of them having any connection as a literary form with Avadāna. Upadeśa is "instruction", and it remained with that general meaning in the Pāli.

c28.s3.4

² For these, see Bibliography, p. 294.

c28.s3.5

The Avadāna differs from Jātaka first in being a later form of literature. Jātakas occur in the earlier parts of the Canon, and are tales about previous existences of Buddha.

c28.s3.6

It was by the Theravādins that this type was cultivated. They made a special collection, and were able to include secular tales and fables of any kind provided that one of the characters in the tale was the future Buddha. In the same way the Sarvāstivādins made collections of Avadānas. Their special feature is that an incident is told of a character in Buddha’s time, and it is then explained by Buddha as the result of some heroic act (*avadāna*) done in a previous existence. It is thus possible for Jātaka and Avadāna to overlap. The Avadānas have an important doctrinal significance, as they contain evidence for the rise of new dogmas during the centuries when the Canon as represented by the Vinaya and Āgamas was closed. This is seen also in the Avadāna of another school, the *Mahāvastu*.¹

c28.s3.7

¹ The Pāli Udāna shows evidence of having developed in the same way. There are Udānas, “fervent utterances” of Buddha, scattered throughout the Scriptures. The work of this name is a collection of such utterances with the legends explaining when they were uttered. A considerable number are from the Vinaya.

The Mahāvastu and Lives of Buddha

c28.s4.1

What has been said of the structure of the Avadāna applies also to the work of another school, the *Mahāvastu*. This is a work on a much larger scale than the others, but constructed in the same way. It has not always been recognized as an Avadāna, though it calls itself one. Even its introductory statement has been misunderstood, which is, “the beginning of the Great Story (*Mahāvastu*) of the Vinaya-piṭaka according to the text of the Mahāsanghikas, the Lokottaravādins of Madhyadeśa.”² This seems plain enough, and it is at least clear that this is what the work actually is. It is not the Vinaya but the “Great Story” of the career of Gautama Buddha, as conceived by an adherent of the Lokottaravāda school. In order to tell this he goes through the Vinaya and incorporates all the legends he can find. To what extent they were in his Vinaya cannot say, but some of the poems, eulogies, and ballads come from other parts of the Canon.³ In carrying out his plan, his respect for the Scripture has prevented him from welding the matter into a consistent unity, but the scheme of the whole can be seen. It is the career of the Bodhisattva conceived not as the life of one man, but as one instance of the career of the Bodhisattvas, who dedicate themselves to ages of training and acquiring the requisite qualities for one who is to be a saviour of the world. It begins with a statement of the stages of the bodhisattva’s career, and describes all the legendary events of Gautama’s life down to his enlightenment and first public preaching to king Bimbisāra. The thread of the story often seems to be missing, because the connecting link between two portions of narrative may not be there, and also because two or more parallel accounts of the same events may be given, so that there is often an apparent regression. The narrative is often broken by jātakas introduced to explain the cause of the events recorded. Apart from its importance as an instance of the later canonical literature, the *Mahāvastu* is also of great value in preserving much of the older canonical material which it quotes from the Scriptures of the Lokottaravādins, and it throws light especially on the development of dogmatic views about a Bodhisattva and Buddha.

c28.s4.2

² Āryamahāsamghikānām lokottaravādinām madhyadeśikānām pāṭhena vinayapiṭakasya mahāvastuye ādi. Vol. i, 2.

c28.s4.3

³ He also has quotations from the Vimānavatthu and the Buddhavamsa, showing that these works, though the latest in the Pāli Canon, came from India. The wildest views have been expressed about the date of the work. It is evident that the portions contributed by the compiler must be much later than the canonical passages which he quotes. For our purpose it is

c28.s4.4

There are two other works of Avadāna type, closely related to the *Mahāvastu* in that they give a biography of Buddha. The first, which exists only in Chinese, has been translated by S. Beal under the title, *The Romantic Legend of Sâkya Buddha* (London, 1875), and he declared it to be a version of the *Abhinishkramaṇa-sūtra*. The Chinese title, however, is *Fo-pan-hing-tsi-ching*, which Nanjio (No. 680) restores as *Buddha-pūrva-caryā-sangraha-sūtra*, “ Sūtra of the collection of the previous life of Buddha.” It forms a better arranged account than the *Mahāvastu*, and it is easy to see that it is constructed from Vinaya material. Much of it corresponds closely with the Pāli Vinaya, but a note at the end states that it belonged to the Dharmaguptas, and it was doubtless based on their Canon. The note also mentions the names of five lives of Buddha as current in different schools. Beal’s note seems to imply that they are different names of the same work. One of them is *Ta-sse*, which he correctly

c28.s4.5

only necessary to note that the canonical matter forming the bulk of the work belongs to a Hinayāna school parallel to the Theravāda and Sarvāstivāda, and that it contributed to the buddhology of Mahāyāna.

c28.s4.6

interprets as *Mahāvastu*, and which is doubtless our *Mahāvastu*, as it belonged to the Mahāsanghikas. It is clear that not one book in the different schools is meant, but that each school possessed its own life of Buddha.¹

c28.s4.7

¹ See Nanjio’s note, *Cat.* No. 680.

c28.s4.8

Another title of one of these lives, given by Beal as that of the book of the Sarvāstivādins, is *Ta-chong-yen*, “Great magnificence,” which Vasiliev identified with the *Lalita-vistara*. But the latter title means “extended account of the sports (of the Bodhisattva)”, and it is a Mahāyāna work. It is, however, easy to see that the *Lalita-vistara* is based largely on Sarvāstivāda material, and this material may even have been the work called “Great Magnificence”. The *Lalita-vistara*, besides containing much avadāna matter, still retains direct evidence of having once been a Sarvāstivāda avadāna, for it has two beginnings. After the usual introduction of a Mahāyāna sūtra, saying that the Lord once dwelt at Śrāvastī, it begins over again: “At that time the Lord was dwelling near the great city of Śrāvastī, honoured, reverenced, respected, worshipped.” If the words “at that time” were omitted, we should have the same abrupt beginning as we find in the avadānas, followed by the same stereotyped list of epithets, *satkṛto, gurukṛto*, etc. This looks like a survival of the original beginning of an avadāna. The work is of much greater literary ability than the *Mahāvastu*. The narrative is well ordered, and much of it, both in prose and verse, corresponds closely to the Pāli, but is most probably from the Sarvāstivāda Canon. Besides this there are long verbose passages of purely Mahāyāna origin, as well as poems very different from the canonical verses which it quotes. They are often in elaborate metres and in very corrupt Sanskrit due to their having been adapted and Sanskritized as far as possible from a Prakrit dialect.

c28.s4.9

Another work may be mentioned here, though it is not canonical, as being the chief Pāli authority for Buddha’s life. It is the introduction to the *Jātaka*, the *Nidāna-kathā*, “discussion of the causes or occasions (of Buddha’s deeds recorded in the *Jātaka*).” It was composed in Ceylon, and the author has compiled the first part by quoting largely from the *Buddhavamsa*. The rest, from the decision to be born to the first preaching, is based on the Canon and the old Sinhalese commentaries, to which the author directly refers. The reputation of the Pāli Canon has given a quite spurious authority to the *Nidāna-kathā*. It was treated as equally canonical, but whatever the age of that Canon, the *Nidāna-kathā* cannot be earlier than the fifth century A.D. The author probably knew the *Lalita-vistara* or one of the rival works, for he protests against the story that Buddha, when he was tempted by Māra’s daughters in the guise of old women, condemned them to remain in that state.

Mahāyāna Canonical Literature

c28.s5.1

The origin of the Mahāyāna Scriptures is even more obscure than that of the original collections. In form they are a continuation of the same type of literature—discourses purporting to have been delivered at various places by Buddha usually during his earthly career. But they are expressly termed Mahāyāna sūtras, and they often attack the doctrine of the older collections. About their real origin naturally nothing is expressed in the works themselves. Everything that can be said about their date, authorship, and provenance is mere supposition.

c28.s5.2

Some of their characteristic doctrines can already be traced in the Avadānas of the Sarvāstivādins and Mahāsanghikas. Such doctrines must have been taught in the monasteries, and we find the teaching about bodhisattvas, the Perfections, and the two kinds of truth infecting even Ceylon Buddhism. The doctrine that bodhisattvas utter true teaching owing to their being inspired by the direct power of Buddha (*buddhānubhāvena*) is frequent in the Mahāyāna sūtras, and it is possible that their composers may have been influenced by some such theory, which would make it unnecessary to class all the sūtras as conscious fabrications.

c28.s5.3

Mahāyāna doctrine was expounded by Nāgārjuna in the second century A.D., and we can certainly hold that some of the sūtras were in existence a century, perhaps two centuries, earlier.¹ But we cannot point to any one of them and feel sure that it belongs to the first century B.C. The question of division into schools is also obscure. The two earlier Mahāyāna schools, that of Nāgārjuna and that of Vasubandhu, both called themselves Mahāyāna, and both taught the doctrine of the void and the career of the bodhisattva. We do not know how far their peculiar doctrines brought them into distinct opposition as sects. Many of the Mahāyāna sūtras are not sectarian at all. Some of them emphasize special doctrines of buddhology, or present imaginative mythology, which may have been accepted by any Mahāyāna sect. The most we can say about the sūtras is that they fall into groups. So far as they have sectarian features their relationship is rather that they arose in certain literary centres. The most distinctive in doctrine and probably the oldest is the Prajñāpāramitā group. Another group is that which emphasizes the worship of Amitābha. The Laṅkāvatāra is distinct from both, not only in its Yogāchāra doctrine, but also in its distinct literary character, which suggests South India.

First 122 of 152 passages. Read on

Linked texts (72)