THE CANON
1
There are extant in several languages large collections of Buddhist scriptures described by some European writers as the Canon. The name is convenient and not incorrect, but the various canons are not altogether similar and the standard for the inclusion or exclusion of particular works is not always clear. We know something of four or five canons.
(1) The Pali Canon, accepted by the Buddhists of Ceylon, Burma and Siam, and rendered accessible to European students by the Pali Text Society. It professes to contain the works recognized as canonical by the Council of Asoka and it is reasonably homogeneous, that is to say, although some ingenuity may be needed to harmonize the different strata of which it consists, it does not include works composed by several schools.
(2) The Sanskrit Canon or Canons.
(a) Nepalese scriptures. These do not correspond with any Pali texts and all belong to the Mahayana. There appears to be no standard for fixing the canonical character of Mahayanist works. Like the Upanishads they are held to be revealed from time to time.
(b) Buddhist texts discovered in Central Asia. Hitherto these have been merely fragments, but the number of manuscripts found and not yet published permits the hope that longer texts may be forthcoming. Those already made known are partly Mahayanist and partly similar to the Pali Canon though not a literal translation of it. It is not clear to what extent the Buddhists of Central Asia regarded the Hina and Mahayanist scriptures as separate and distinct. Probably each school selected for itself a small collection of texts as authoritative[599].
(3) The Chinese Canon. This is a gigantic collection of Buddhist works made and revised by order of various Emperors. The imperial imprimatur is the only standard of canonicity. The contents include translations of works belonging to all schools made from the first to the thirteenth century A.D. The originals were apparently all in Sanskrit and were probably the texts of which fragments have been found in Central Asia. This canon also includes some original Chinese works.
(4) There is a somewhat similar collection of translations into Tibetan. But whereas the Chinese Canon contains translations dated from 67 A.D. onwards, the Tibetan translations were made mainly in the ninth and eleventh centuries and represent the literature esteemed by the mediæval Buddhism of Bengal. Part at least of this Tibetan Canon has been translated into Mongol.
Renderings of various books into Uigur, Sogdian, Kuchanese, "Nordarisch" and other languages of Central Asia have been discovered by recent explorers. It is probable that they are all derived from the Sanskrit Canon and do not represent any independent tradition. The scriptures used in Japan and Korea are simply special editions of the Chinese Canon, not translations.
In the following pages I propose to consider the Pali Canon, postponing until later an account of the others. It will be necessary, however, to touch on the relations of Pali and Sanskrit texts.
The scriptures published by the Pali Text Society represent the canon of the ancient sect called Vibhajjavâdins and the particular recension of it used at the monastery in Anuradhapura called Mahâvihâra. It is therefore not incorrect to apply to this recension such epithets as southern or Sinhalese, provided we remember that in its origin it was neither one nor the other, for the major part of it was certainly composed in India[600]. It was probably introduced into Ceylon in the third century B.C. and it is also accepted in Burma, Siam and Camboja[601]. Thus in a considerable area it is the sole and undisputed version of the scriptures.
The canon is often known by the name of Tripiṭaka[602] or Three Baskets. When an excavation was made in ancient India it was the custom to pass up the earth in baskets along a line of workmen[603] and the metaphorical use of the word seems to be taken from this practice and to signify transmission by tradition.
The three Pitakas are known as Vinaya, Sutta, and Abhidhamma. Vinaya means discipline and the works included in this division treat chiefly of the rules to be observed by the members of the Sangha. The basis of these rules is the Pâtimokkha, the ancient confessional formula enumerating the offences which a monk can commit. It was read periodically to a congregation of the order and those guilty of any sin had to confess it. The text of the Pâtimokkha is in the Vinaya combined with a very ancient commentary called the Sutta-vibhanga. The Vinaya also contains two treatises known collectively as the Khandakas but more frequently cited by their separate names as Mahâvagga and Cullavagga. The first deals with such topics as the rules for admission to the order, and observance of fast days, and in treating of each rule it describes the occasion on which the Buddha made it and to some extent follows the order of chronology. For some parts of the master's life it is almost a biography. The Cullavagga is similar in construction but less connected in style[604]. The Vinaya contains several important and curious narratives and is a mine of information about the social conditions of ancient India, but much of it has the same literary value as the book of Leviticus. Of greater general interest is the Sutta Pitaka, in which the sermons and discourses of the Buddha are collected. Sutta is equivalent to the Sanskrit word Sûtra, literally a thread, which signifies among the Brahmans a brief rule or aphorism but in Pali a relatively short poem or narrative dealing with a single object. This Sutta Pitaka is divided into five collections called Nikâyas. The first four are mainly in prose and contain discourses attributed to Gotama or his disciples. The fifth is mostly in verse and more miscellaneous.
The four collections of discourses bear the names of Dîgha, Majjhima, Saṃyutta and Anguttara. The first, meaning long, consists of thirty-four narratives. They are not all sermons and are of varying character, antiquity and interest, the reason why they are grouped together being simply their length[605]. In some of them we may fancy that we catch an echo of Gotama's own words, but in others the legendary character is very marked. Thus the Mahâsamaya and Aṭânâṭiya suttas are epitomes of popular mythology tacked on to the history of the Buddha. But for all that they are interesting and ancient.
Many of the suttas, especially the first thirteen, are rearrangements of old materials put together by a considerable literary artist who lived many generations after the Buddha. The account of the Buddha's last days is an example of such a compilation which attains the proportions of a Gospel and shows some dramatic power though it is marred by the juxtaposition of passages composed in very different styles.
The Majjhima-Nikâya is a collection of 152 discourses of moderate (majjhima) length. Taken as a whole it is perhaps the most profound and impassioned of all the Nikâyas and also the oldest. The sermons which it contains, if not verbatim reports of Gotama's eloquence, have caught the spirit of one who urged with insistent earnestness the importance of certain difficult truths and the tremendous issues dependent on right conduct and right knowledge. The remaining collections, the Saṃyutta and Anguttara, classify the Buddha's utterances under various headings and presuppose older documents which they sometimes quote[606]. The Saṃyutta consists of a great number of suttas, mostly short, combined in groups treating of a single subject which may be either a person or a topic. The Anguttara, which is a still longer collection, is arranged in numerical groups, a method of classification dear to the Hindus who delight in such computations as the four meditations, the eightfold path, the ten fetters. It takes such religious topics as can be counted in this way and arranges them under the numbers from one to eleven. Thus under three, it treats of thought, word and deed and the applications of this division to morality; of the three messengers of the gods, old-age, sickness and death; of the three great evils, lust, ill-will and stupidity and so on.
The fifth or Khuddaka-Nikâya is perhaps the portion of the Pali scriptures which has found most favour with Europeans, for the treatises composing it are short and some of them of remarkable beauty. They are in great part composed of verses, sometimes disconnected couplets, sometimes short poems. The stanzas are only imperfectly intelligible without an explanation of the occasion to which they refer. This is generally forthcoming, but is sometimes a part of the accepted text and sometimes regarded as merely a commentary. To this division of the Pitaka belong the Dhammapada, a justly celebrated anthology of devotional verses, and the Sutta-Nipâta, a very ancient collection of suttas chiefly in metre. Other important works included in it are the Thera and Therî-gâthâ or poems written by monks and nuns respectively, and the Jâtaka or stories about the Buddha's previous births[607]. Some of the rather miscellaneous contents of this Nikâya are late and do not belong to the same epoch of thought as the discourses attributed to Gotama. Such are the Buddha-vaṃsa, or lives of Gotama and his twenty-four predecessors, the Cariyâ-Piṭaka, a selection of Jâtaka stories about Gotama's previous births and the Vimâna and Peta-vatthus, accounts of celestial mansions and of the distressful existence led by those who are condemned to be ghosts[608].
Though some works comprised in this Nikâya (e.g. the Suttanipâta) are very ancient, the collection, as it stands, is late and probably known only to the southern Church. The contents of it are not quite the same in Ceylon, Burma and Siam, and only a small portion of them has been identified in the Chinese Tripitaka. Nevertheless the word pañcanekâyika, one who knows the five Nikâyas, is found in the inscriptions of Sanchi and five Nikâyas are mentioned in the last books of the Cullavagga. Thus a fifth Nikâya of some kind must have been known fairly early.
The third Pitaka is known by the name of Abhidhamma. Dhamma is the usual designation for the doctrine of the Buddha and Buddhaghosa[609] explains the prefix abhi as signifying excess and distinction, so that this Pitaka is considered pre-eminent because it surpasses the others. This pre-eminence consists solely in method and scope, not in novelty of matter or charm of diction. The point of view of the Abhidhamma is certainly later than that of the Sutta Pitaka and in some ways marks an advance, for instead of professing to report the discourses of Gotama it takes the various topics on which he touched, especially psychological ethics, and treats them in a connected and systematic manner. The style shows some resemblance to Sanskrit sûtras for it is so technical both in vocabulary and arrangement that it can hardly be understood without a commentary[610]. According to tradition the Buddha recited the Abhidhamma when he went to heaven to preach to the gods, and this seems a polite way of hinting that it was more than any human congregation could tolerate or understand. Still throughout the long history of Buddhism it has always been respected as the most profound portion of the scriptures and has not failed to find students. This Pitaka includes the Kathâ-vatthu, attributed to Tissa Moggaliputta who is said to have composed it about 250 B.C. in Asoka's reign[611].
There is another division of the Buddhist scriptures into nine angas or members, namely: 1. Suttas. 2. Geyya: mixed prose and verse. 3. Gâthâ: verse. 4. Udâna: ecstatic utterances. 5. Veyyâkaraṇa: explanation. 6. Itivuttaka: sayings beginning with the phrase "Thus said the Buddha." 7. Jâtaka: stories of former births. 8. Abbhutadhamma: stories of wonders. 9. Vedalla: a word of doubtful meaning, but perhaps questions and answers. This enumeration is not to be understood as a statement of the sections into which the whole body of scripture was divided but as a description of the various styles of composition recognized as being religious, just as the Old Testament might be said to contain historical books, prophecies, canticles and so on. Compositions in these various styles must have been current before the work of collection began, as is proved by the fact that all the angas are enumerated in the Majjhima-Nikâya[612].
2
This Tripitaka is written in Pali[613] which is regarded by Buddhist tradition as the language spoken by the Master. In the time of Asoka the dialect of Magadha must have been understood over the greater part of India, like Hindustani in modern times, but in some details of grammar and phonetics Pali differs from Mâgadhî Prakrit and seems to have been influenced by Sanskrit and by western dialects. Being a literary rather than a popular language it was probably a mixed form of speech and it has been conjectured that it was elaborated in Avanti or in Gândhâra where was the great Buddhist University of Takshaśîlâ. Subsequently it died out as a literary language in India[614] but in Ceylon, Burma, Siam and Camboja it became the vehicle of a considerable religious and scholastic literature. The language of Asoka's inscriptions in the third century B.C. is a parallel dialect, but only half stereotyped. The language of the Mahâvastu and some Mahayanist texts, often called the language of the Gâthâs, seems to be another vernacular brought more or less into conformity with Sanskrit. It is probable that in preaching the Buddha used not Pali in the strict sense but the spoken dialect of Magadha[615], and that this dialect did not differ from Pali more than Scotch or Yorkshire from standard English, and if for other reasons we are satisfied that some of the suttas have preserved the phrases which he employed, we may consider that apart from possible deviations in pronunciation or inflexion they are his ipsissima verba. Even as we have it, the text of the canon contains some anomalous forms which are generally considered to be Magadhisms[616].
The Cullavagga relates how two monks who were Brahmans represented to the Buddha that "monks of different lineage ... corrupt the word of the Buddha by repeating it in their own dialect. Let us put the word of the Buddhas into chandas[617]." No doubt Sanskrit verse is meant, chandas being a name applied to the language of the Vedic verses. Gotama refused: "You are not to put the word of the Buddhas into chandas. Whoever does so shall be guilty of an offence. I allow you to learn the word of the Buddhas each in his own dialect." Subsequent generations forgot this prohibition, but it probably has a historical basis and it indicates the Buddha's desire to make his teaching popular. It is not likely that he contemplated the composition of a body of scriptures. He would have been afraid that it might resemble the hymns of the Brahmans which he valued so little and he wished all men to hear his teaching in the language they understood best. But when after his death his disciples collected his sayings it was natural that they should make at least one version of them in the dialect most widely spoken and that this version should be gradually elaborated in what was considered the best literary form of that dialect[618]. It is probable that the text underwent several linguistic revisions before it reached its present state.
Pali is a sonorous and harmonious language which avoids combinations of consonants and several difficult sounds found in Sanskrit. Its excellence lies chiefly in its vocabulary and its weakness in its syntax. Its inflexions are heavy and monotonous and the sentences lack concentration and variety. Compound words do not assume such monstrous proportions as in later Sanskrit, but there is the same tendency to make the process of composition do duty for syntax. These faults have been intensified by the fact that the language has been used chiefly for theological discussion. The vocabulary on the other hand is copious and for special purposes admirable. The translator has to struggle continually with the difficulty of finding equivalents for words which, though apparently synonymous, really involve nice distinctions and much misunderstanding has arisen from the impossibility of adequately rendering philosophical terms, which, though their European equivalents sound vague, have themselves a precise significance. On the other hand some words (e.g. dhamma and attho) show an inconveniently wide range of meaning. But the force of the language is best seen in its power of gathering up in a single word, generally a short compound, an idea which though possessing a real unity requires in European languages a whole phrase for its expression. Thus the Buddha bids his disciples be attadîpâ atta-saraṇâ, anañña-saraṇâ: dhammadîpâ dhammasaraṇâ[619]. "Be ye lamps unto yourselves. Be ye a refuge unto yourselves. Betake yourselves to no external refuge. Hold fast to the truth as a lamp. Hold fast to the truth as a refuge." This is Rhys Davids' translation and excellent both as English and as giving the meaning. But the five Pali words compel attention and inscribe themselves on the memory in virtue of a monumental simplicity which the five English sentences do not possess.
But the feature in the Pali scriptures which is most prominent and most tiresome to the unsympathetic reader is the repetition of words, sentences and whole paragraphs. This is partly the result of grammar or at least of style. The simplicity of Pali syntax and the small use made of dependent sentences, lead to the regular alignment of similar phrases side by side like boards in a floor. When anything is predicated of several subjects, for instance the five Skandhas, it is rare to find a single sentence containing a combined statement. As a rule what has to be said is predicated first of the first Skandha and then repeated totidem verbis of the others. But there is another cause for this tedious peculiarity, namely that for a long period the Pitakas were handed down by oral tradition only. They were first reduced to writing in Ceylon about 20 B.C. in the reign of Vaṭṭagâmani, more than a century and a half after their first importation in an oral form. This circumstance need not throw doubt on the authenticity of the text, for the whole ancient literature of India, prose as well as verse, was handed down by word of mouth and even in the present day most of it could be recovered if all manuscripts and books were lost. The Buddhists did not, like the Brahmans, make minute regulations for preserving and memorizing their sacred texts, and in the early ages of the faith were impressed with the idea that their teaching was not a charm to be learnt by heart but something to be understood and practised. They nevertheless endeavoured, and probably with success, to learn by heart the words of the Buddha, converting them into the dialect most widely understood. It was then a common thing (and the phenomenon may still be seen in India) for a man of learning to commit to memory a whole Veda together with subsidiary treatises on ritual, metre, grammar and genealogy. For such memories it was not difficult to retain the principal points in a series of sermons. The Buddha had preached day by day for about forty-five years. Though he sometimes spoke with reference to special events he no doubt had a set of discourses which he regularly repeated. There was the less objection to such repetition because he was continually moving about and addressing new audiences. There were trained Brahman students among his disciples, and at his death many persons, probably hundreds, must have had by heart summaries of his principal sermons.
But a sermon is less easy to remember than a poem or matter arranged by some method of memoria technica. An obvious aid to recollection is to divide the discourse into numbered heads and attach to each certain striking phrases. If the phrases can be made to recur, so much the better, for there is a guarantee of correctness when an expected formula appears at appropriate points.
It may be too that the wearisome and mechanical iteration of the Pali Canon is partly due to the desire of the Sinhalese to lose nothing of the sacred word imparted to them by missionaries from a foreign country, for repetition to this extent is not characteristic of Indian compositions. It is less noticeable in Sanskrit Buddhist sûtras than in the Pali but is very marked in Jain literature. A moderate use of it is a feature of the Upanishads. In these we find recurring formulæ and also successive phrases constructed on one plan and varying only in a few words[620].
But still I suspect that repetition characterized not only the reports of the discourses but the discourses themselves. No doubt the versions which we have are the result of compressing a free discourse into numbered paragraphs and repetitions: the living word of the Buddha was surely more vivacious and plastic than these stiff tabulations. But the peculiarities of scholars can often be traced to the master and the Buddha had much the same need of mnemonics as his hearers. For he had excogitated complicated doctrines and he imparted them without the aid of notes and though his natural wit enabled him to adapt his words to the capacity of his hearers and to meet argument, still his wish was to formulate a consistent statement of his thoughts. In the earliest discourse ascribed to him, the sermon at Benares, we see these habits of numbering and repetition already fully developed. The next discourse, on the absence of a soul, consists in enumerating the five words, form, sensation, perception, sankhâras, and consciousness three times, and applying to each of them consecutively three statements or arguments, the whole concluding with a phrase which is used as a finale in many other places. Artificial as this arrangement sounds when analyzed, it is a natural procedure for one who wished to impress on his hearers a series of philosophic propositions without the aid of writing, and I can imagine that these rhythmical formulæ uttered in that grave and pleasant voice which the Buddha is said to have possessed, seemed to the leisurely yet eager groups who sat round him under some wayside banyan or in the monastery park, to be not tedious iteration but a gradual revelation of truth growing clearer with each repetition.
We gather from the Pitakas that writing was well known in the Buddha's time[621]. But though it was used for inscriptions, accounts and even letters, it was not used for books, partly because the Brahmans were prejudiced against it, and partly because no suitable material for inditing long compositions had been discovered. There were religious objections to parchment and leaves were not employed till later. The minute account of monastic life given in the Vinaya makes it certain that the monks did not use writing for religious purposes. Equally conclusive, though also negative, is the fact that in the accounts of the assemblies at Râjagaha and Vesâlî[622] when there is a dispute as to the correct ruling on a point, there is no appeal to writing but merely to the memory of the oldest and most authoritative monks. In the Vinaya we hear of people who know special books: of monks who are preachers of the Dhamma and others who know the Sutta: of laymen who have learnt a particular suttanta and are afraid it will fall into oblivion unless others learn it from them. Apprehensions are expressed that suttas will be lost if monks neglect to learn them by heart[623]. From inscriptions of the third century B.C.[624] are quoted words like Petakî, a reciter of the Pitakas or perhaps of one Pitaka: Suttântika and Suttântakinî, a man or woman who recites the suttantas: Pancanekâyika, one who recites the five Nikâyas. All this shows that from the early days of Buddhism onwards a succession of persons made it their business to learn and recite the doctrine and disciplinary rules and, considering the retentiveness of trained memories, we have no reason to doubt that the doctrine and rules have been preserved without much loss[625].
Not, however, without additions. The disadvantage of oral tradition is not that it forgets but that it proceeds snowball fashion, adding with every generation new edifying matter. The text of the Vedic hymns was preserved with such jealous care that every verse and syllable was counted. But in works of lesser sanctity interpolations and additions were made according to the reciters' taste. We cannot assign to the Mahâbhârata one date or author, and the title of Upanishad is no guarantee for the age or authenticity of the treatises that bear it. Already in the Anguttara-Nikâya[626], we hear of tables of contents and the expression is important, for though we cannot give any more precise explanation of it, it shows that care was taken to check the contents of the works accepted as scripture. But still there is little doubt that during the two or three centuries following the Buddha's death, there went on a process not only of collection and recension but also of composition.
An account of the formation of the canon is given in the last two chapters of the Cullavagga[627]. After the death of the Buddha his disciples met to decide what should be regarded as the correct doctrine and discipline. The only way to do that was to agree what had been the utterances of the master and this, in a country where the oral transmission of teaching was so well understood, amounted to laying the foundations of a canon. Kassapa cross-examined experts as to the Buddha's precepts. For the rules of discipline Upâli was the chief authority and we read how he was asked where such and such a rule—for instance, the commandment against stealing—was promulgated.
"At Râjagaha, sir."
"Concerning whom was it spoken?"
"Dhaniya, the potter's son."
"In regard to what matter?"
"The taking of that which had not been given."
For collecting the suttas they relied on the testimony of Ânanda and asked him where the Brahmajâla[628] was spoken. He replied "between Râjagaha and Nâlanda at the royal rest-house at Ambalatthika." "Concerning whom was it spoken?" "Suppiya, the wandering ascetic and Brahmadatta the young Brahman."
Then follows a similar account of the Sâmaññaphala sutta and we are told that Ânanda was "questioned through the five Nikâyas." That is no doubt an exaggeration as applied to the time immediately after the Buddha's death, but it is evidence that five Nikâyas were in existence when this chapter was written[629].
3
Lines of growth are clearly discernible in the Vinaya and Sutta Pitakas. As already mentioned, the Khuddaka-Nikâya is, as a collection, later than the others although separate books of it, such as the Sutta-nipâta (especially the fourth and fifth books), are among the earliest documents which we possess. But other books such as the Peta-[630] and Vimâna-vatthu show a distinct difference in tone and are probably separated from the Buddha by several centuries. Of the other four Nikâyas the Saṃyutta and Anguttara are the more modern and the Anguttara mentions Munda, King of Magadha who began to reign about forty years after the Buddha's death. But even in the two older collections, the Dîgha and the Majjhima, we have not reached the lowest stratum. The first thirteen suttantas of the Dîgha all contain a very ancient tractate on morality, and the Sâmaññaphala and following sections of the Dîgha and also some suttas of the Majjhima contain either in whole or in part a treatise on progress in the holy life. These treatises were probably current as separate portions for recitation before the suttas in which they are now set were composed.
Similarly, the Vinaya clearly presupposes an old code in the form of a list of offences called the Pâtimokkha. The Mahâvagga contains a portion of an ancient word-for-word explanation of this code[631] and most of the Sutta-vibhanga is an amplification and exposition of it. The Pâtimokkha was already in existence when these books were composed, for we hear[632] that if in a company of Bhikkhus no one knows the Pâtimokkha, one of the younger brethren should be sent to some better instructed monastery to learn it. And further we hear[633] that a learned Bhikkhu was expected to know not merely the precepts of the Pâtimokkha but also the occasion when each was formulated. The place, the circumstances and the people concerned had been in each case handed down. There is here all the material for a narrative. The reciter of a sutta simply adopts the style of a village story-teller. "Thus have I heard. Once upon a time the Lord was dwelling at Râjagaha," or wherever it was, and such and such people came to see him. And then, after a more or less dramatic introduction, comes the Lord's discourse and at the end an epilogue saying how the hearers were edified and, if previously unconverted, took refuge in the true doctrine.
The Cullavagga states that the Vinaya (but not the other Pitakas) was recited and verified at the Council of Vesâlî. As I have mentioned elsewhere, Sinhalese and Chinese accounts speak of another Council, the Mahâsangha or Mahâsangîti. Though its date is uncertain, there is a consensus of tradition to the effect that it recognized a canon of its own, different from our Pali Canon and containing a larger amount of popular matter.
Sinhalese tradition states that the canon as we now have it was fixed at the third Council held at Pataliputra in the reign of Asoka (about 272-232 B.C.). The most precise statements about this Council are those of Buddhaghosa who says that an assembly of monks who knew the three Pitakas by heart recited the Vinaya and the Dhamma.
But the most important and interesting evidence as to the existence of Buddhist scriptures in the third century B.C. is afforded by the Bhâbrû (or Bhâbrâ) edict of Asoka. He recommends the clergy to study seven passages, of which nearly all can be identified in our present edition of the Pitakas[634]. This edict does not prove that Asoka had before him in the form which we know the Dîgha and other works cited. But the most cautious logic must admit that there was a collection of the Buddha's sayings to which he could appeal and that if most of his references to this collection can be identified in our Pitakas, then the major part of these Pitakas is probably identical in substance (not necessarily verbally) with the collection of sayings known to Asoka.
Neither Asoka nor the author of the Kathâ-vatthu cites books by name. The latter for instance quotes the well-known lines "anupubbena medhavi" not as coming from the Dhammapada but as "spoken by the Lord." But the author of the Questions of Milinda, who knew the canonical books by the names they bear now, also often adopts a similar method of citation. Although this author's probable date is not earlier than our era his evidence is important. He mentions all five Nikâyas by name, the titles of many suttas and also the Vibhanga, Dhâtu-kathâ, Puggala-Paññatti, Kathâ-vatthu, Yamaka and Paṭṭhâna.
Everything indicates and nothing discredits the conclusion that this canon of the Vibhajjavâdins was substantially fixed in the time of Asoka, so far as the Vinaya and Sutta Pitakas are concerned. Some works of minor importance may have had an uncertain position and subsequent revisions may have been made but the principal scriptures were already recognized and contained passages which occur in our versions. On the other hand this recension of the scriptures was not the only one in existence. If the patronage of Asoka gave it a special prestige in his lifetime, it may have lost it in India after his death and for many centuries the Buddhist Canon, like the list of the Upanishads, must have been susceptible of alteration. The Sarvâstivâdins compiled an Abhidhamma Pitaka of their own, apparently in the time of Kanishka, and the Dharmagupta school also seems to have had its own version of this Pitaka[635]. The date of the Pali Abhidhamma is very doubtful and I do not reject the hypothesis that it was composed in Ceylon, for the Sinhalese seem to have a special taste for such literature. But there is no proof of this Sinhalese origin.
According to Sinhalese tradition all three Pitakas were introduced into Ceylon by Mahinda in the reign of Asoka, but only as oral tradition and not in a written form. They received this latter about 20 B.C., as the result of a dispute between two monasteries[636]. The controversy is obscure but it appears that the ancient foundation called Mahâvihâra accepted as canonical the fifth book of the Vinaya called Parivâra, whereas it was rejected by the new monastery called Abhayagiri. The Sinhalese chronicle (Mahâvamsa XXXIII. 100-104) says somewhat abruptly "The wise monks had hitherto handed down the text of the three Pitakas (Piṭakattayapâlim) as well as the commentary by word of mouth. But seeing that mankind was becoming lost, they assembled together and wrote them in books in order that the faith might long endure." This brief account seems to mean that a council was held not by the whole clergy of Ceylon but by the monks of the Mahâvihâra at which they committed to writing their own version of the canon including the Parivâra. This book forms an appendix to the Vinaya Pitaka and in some verses printed at the conclusion is said to be the work of one Dîpa. It is generally accepted as a relatively late production, composed in Ceylon. If such a work was included in the canon of the Mahâvihâra, we must admit the possibility that other portions of it may be Sinhalese and not Indian.
But still the onus probandi lies with those who maintain the Sinhalese origin of any part of the Pali Canon and two strong arguments support the Indian origin of the major part. First, many suttas not only show an intimate knowledge of ancient Indian customs but discuss topics such as caste, sacrifice, ancient heresies, and the value of the Veda which would be of no interest to Sinhalese. Secondly, there is no Sinhalese local colour and no Sinhalese legends have been introduced. Contrast with this the Dîpa-and Mahâ-vaṃsa both of which open with accounts of mythical visits paid by the Buddha to Ceylon[637].
In Ceylon versions of the scriptures other than that of the Mahâvihâra were current until the twelfth century when uniformity was enforced by Parâkrama Bâhu. Some of these, for instance the Pitaka of the Vetulyakas, were decidedly heretical according to the standard of local orthodoxy but others probably presented variations of reading and arrangement rather than of doctrine. Anesaki[638] has compared with the received Pali text a portion of the Saṃyuktâgama translated by Guṇabhadra into Chinese. He thinks that the original was the text used by the Abhayagiri monastery and brought to China by Fa Hsien.
The Sinhalese ecclesiastical history, Nikâya-Sangrahawa, relates[639] that 235 years after the Buddha's death nine heretical fraternities were formed who proceeded to compose scriptures of their own such as the Varṇapiṭaka and Angulimâla-Piṭaka. Though this treatise is late (c. 1400 A.D.) its statements merit attention as showing that even in orthodox Ceylon tradition regarded the authorized Pitaka as one of several versions. But many of the works mentioned sound like late tantric texts rather than compositions of the early heretics to whom they are attributed.
Ecclesiastical opinion in Ceylon after centuries of discussion ended by accepting the edition of the Mahâvihâra as the best, and we have no grounds for rejecting or suspecting this opinion. According to tradition Buddhaghosa was well versed in Sanskrit but deliberately preferred the southern canon. The Mahayanist doctor Asanga cites texts found in the Pali version, but not in the Sanskrit[640]. The monks of the Mahâvihâra were probably too indulgent in admitting late scholastic treatises, such as the Parivâra. On the other hand they often showed a critical instinct in rejecting legendary matter. Thus the Sanskrit Vinayas contain many more miraculous narratives than the Pali Vinaya.
4
European critics have rarely occasion to discuss the credibility of Sanskrit literature, for most of it is so poetic or so speculative that no such question arises. But the Pitakas raise this question as directly as the Gospels, for they give the portrait of a man and the story of a life, in which an overgrowth of the miraculous has not hidden or destroyed the human substratum. How far can we accept them as a true picture of what Gotama was and taught?
Their credibility must be judged by the standard of Indian oral tradition. Its greatest fault comes from that deficiency in historic sense which we have repeatedly noticed. Hindu chroniclers ignore important events and what they record drifts by in a haze in which proportion, connection, and dates are lost. They frequently raise a structure of fiction on a slight basis of fact or on no basis at all. But the fiction is generally so obvious that the danger of historians in the past has been not to be misled by it but to ignore the elements of truth which it may contain. For the Hindus have a good verbal memory; their genealogies, lists of kings and places generally prove to be correct and they have a passion for catalogues of names. Also they take a real interest in describing doctrine. If the Buddha has been misrepresented, it is not for want of acumen or power of transmitting abstruse ideas. The danger rather is that he who takes an interest in theology is prone to interpret a master's teaching in the light of his own pet views.
The Pitakas illustrate the strong and weak points of Hindu tradition. The feebleness of the historical sense may be seen in the account of Devadatta's doings in the Cullavagga[641] where the compiler seems unable to give a clear account of what he must have regarded as momentous incidents. Yet the same treatise is copious and lucid in dealing with monastic rules, and the sayings recorded have an air of authenticity. In the suttas the strong side of Hindu memory is brought into play. Of consecutive history there is no question. We have only an introduction giving the names of some characters and localities followed by a discourse. We know from the Vinaya that the monks were expected to exercise themselves in remembering these things, and they are precisely the things that they would get rightly by heart. I see no reason to doubt that such discourses as the sermon preached at Benares[642] and the recurring passages in the first book of the Dîgha-Nikâya are a Pali version of what was accepted as the words of the Buddha soon after his death. And the change of dialect is not of great importance. Asoka's Bhâbrû Edict contains the saying: Thus the good law shall long endure, which is believed to be a quotation and certainly corresponds pretty closely with a passage in the Anguttara-Nikâya[643]. The King's version is Saddhamma cilathitike hasati: the Pali is Saddhammo cîratthitiko hoti. Somewhat similar may have been the differences between the Buddha's speech and the text which we possess. The importance of the change in language is diminished and the facility of transmission is increased by the fact that in Pali, Sanskrit and kindred Indian languages ideas are concentrated in single words rather than spread over sentences. Thus the principal words of the sermon at Benares give its purport with perfect clearness, if they are taken as a mere list without grammatical connection. Similarly I should imagine that the recurring paragraphs about progress in the holy life found in the early Suttas of the Dîgha-Nikâya are an echo of the Buddha's own words, for they bear an impress not only of antiquity but of eloquence and elevation. This does not mean that we have any sermon in the exact form in which Gotama uttered it. Such documents as the Sâmaññaphala-sutta and Ambaṭṭha-sutta probably give a good idea of his method and style in consecutive discourse and argument. But it would not be safe to regard them as more than the work of compilers who were acquainted with the surroundings in which he lived, the phrases he used, and the names and business of those who conversed with him. With these they made a picture of a day in his life, culminating in a sermon[644].
Like the historical value of the Pitakas, their literary value can be justly estimated only if we remember that they are not books in our sense but treatises handed down by memory and that their form is determined primarily by the convenience of the memory. We must not compare them with Plato and find them wanting, for often, especially in the Abhidhamma, there is no intention of producing a work of art, but merely of subdividing a subject and supplying explanations. Frequently the exposition is thrown into the form of a catechism with questions and answers arranged so as to correspond to numbered categories. Thus a topic may be divided into twenty heads and six propositions may be applied to each with positive or negative results. The strong point of these Abhidhamma works---and of Buddhist philosophy generally—lies in careful division and acute analysis but the power of definition is weak. Rarely is a definition more than a collection of synonyms and very often the word to be defined is repeated in the definition. Thus in the Dhamma-sangaṇi the questions, what are good or bad states of mind? receive answers cast in the form: when a good or bad thought has arisen with certain accompaniments enumerated at length, then these are the states that are good or bad. No definition of good is given.
This mnemonic literature attains its highest excellence in poetry. The art of composing short poems in which a thought, emotion or spiritual experience is expressed with a few simple but pregnant words in the compass of a single couplet or short hymn, was carried by the early Buddhists to a perfection which has never been excelled. The Dhammapada[645] is the best known specimen of this literature. Being an anthology it is naturally more suited for quotation or recitation in sections than for continuous reading. But its twenty-five chapters are consecrated each to some special topic which receives fairly consecutive treatment, though each chapter is a mosaic of short poems consisting of one or more verses supposed to have been uttered by the Buddha or by arhats on various occasions. The whole work combines literary beauty, depth of thought and human feeling in a rare degree. Not only is it irradiated with the calm light of peace, faith and happiness but it glows with sympathy, with the desire to do good and help those who are struggling in the mire of passion and delusion. For this reason it has found more favour with European readers than the detached and philosophic texts which simply preach self-conquest and aloofness. Inferior in beauty but probably older is the Sutta-nipâta, a collection of short discourses or conversations with the Buddha mostly in verse. The rugged and popular language of these stanzas which reject speculation as much as luxury, takes us back to the life of the wanderers who followed the Buddha on his tours and we may imagine that poems like the Dhaniya sutta would be recited when they met together in a rest-house or grove set apart for their use on the outskirts of a village.
The Buddhist suttas, are interesting as being a special result of Gotama's activity; they are not analogous to the Brahmanic works called sûtras, and they have no close parallel in later Indian literature. There is little personal background in the Upanishads, none at all in the Sânkhya and Vedânta sûtras. But the Sutta Pitaka is an attempt to delineate a personality as well as to record a doctrine. Though the idea of writing biography has not yet been clearly conceived, yet almost every discourse brings before us the figure of the Lord: though the doctrine can be detached from the preacher, yet one feels that the hearers of the Pitaka hungered not merely for a knowledge of the four truths but for the very words of the great voice: did he really say this, and if so when, where and why? Most suttas begin by answering these questions. They describe a scene and report a discourse and in so doing they create a type of literature with an interest and individuality of its own. It is no exaggeration to say that the Buddha is the most living figure in Hindu literature. He stands before us more distinctly not only than Yâjñavalkya and Śankara, but than modern teachers like Nanak and Râmânuja and the reason of this distinctness can I think be nothing but the personal impression which he made on his age. The later Buddhists compose nothing in the style of the Nikâyas: they write about Gotama in new and fanciful ways, but no Acts of the Apostles succeed the Gospels.
Though the Buddhist suttas are sui generis and mark a new epoch in Indian literature, yet in style they are a natural development of the Upanishads. The Upanishads are less dogmatic and show much less interest in the personality of their sages, but they contain dialogues closely analogous to suttas. Thus about half of the Bṛihad-Âraṇyaka is a philosophic treatise unconnected with any particular name, but in this are set five dialogues in which Yâjñavalkya appears and two others in which Ajâtaśatru and Pravâhaṇa Jaivali are the protagonists.
Though many suttas are little more than an exposition of some doctrine arranged in mnemonic form, others show eloquence and dramatic skill. Thus the Sâmaññaphala-sutta opens with a vivid description of the visit paid one night by Ajâtasattu to the Buddha[646]. We see the royal procession of elephants and share the alarm of the suspicious king at the unearthly stillness of the monastery park, until he saw the Buddha sitting in a lighted pavilion surrounded by an assembly of twelve hundred and fifty brethren, calm and silent as a clear lake. The king's long account of his fruitless quest for truth would be tiresome if it were not of such great historic interest and the same may be said of the Buddha's enumeration of superstitious and reprehensible practices, but from this point onwards his discourse is a magnificent crescendo of thought and language, never halting and illustrated by metaphors of great effect and beauty. Equally forcible and surely resting on some tradition of the Buddha's own words is the solemn fervour which often marks the suttas of the Majjhima such as the descriptions of his struggle for truth, the admonitions to Râhula and the reproof administered to Sâti.
5
As mentioned above, our Pali Canon is the recension of the Vibhajjavâdins. We know from the records of the Chinese pilgrims that other schools also had recensions of their own, and several of these recensions—such as those of the Sarvâstivâdins, Mahâsanghikas, Mahisâsakas, Dhammaguttikas, and Sammitîyas—are still partly extant in Chinese and Tibetan translations. These appear to have been made from the Sanskrit and fragments of what was probably the original have been preserved in Central Asia. A recension of the text in Sanskrit probably implies less than what we understand by a translation. It may mean that texts handed down in some Indian dialect which was neither Sanskrit nor Pali were rewritten with Sanskrit orthography and inflexions while preserving much of the original vocabulary. The Buddha allowed all men to learn his teaching in their own language, and different schools are said to have written the scriptures in different dialects, e.g. the Mahâsanghikas in a kind of Prakrit not further specified and the Mahâsammatîyas in Apabhramsa. When Sanskrit became the recognized vehicle for literary composition there would naturally be in India (though not in Ceylon) a tendency to rewrite books composed in other dialects[647]. The idea that when any important matter is committed to writing it should be expressed in a literary dialect not too intelligible to the vulgar is prevalent from Morocco to China. The language of Bengal illustrates what may have happened to the Buddhist scriptures. It is said that at the beginning of the nineteenth century ninety per cent, of the vocabulary of Bengali was Sanskrit, and the grammatical construction sanskritized as well. Though the literary language now-a-days is less artificial, it still differs widely from the vernacular. Similarly the spoken word of the Buddha was forced into conformity with one literary standard or another and ecclesiastical Pali became as artificial as Sanskrit. The same incidents may be found worked up in both languages. Thus the Sanskrit version of the story of Pûrṇâ in the Divyâva-dâna repeats what is found in Pali in the Saṃyutta-Nikâya[648] and reappears in Sanskrit in the Vinaya of the Mûlasarvâstivâdin school.
The Chinese Tripitaka has been catalogued and we possess some information respecting the books which it contains, though none of them have been edited in Europe. Thus we know something[649] of the Sarvâstivâdin recension of the Abhidhamma. Like the Pali version it consists of seven books of which one, the Jñâna-prasthâna by Kâtyâyanîputra, is regarded as the principal, the rest being supplementary. All the books are attributed to human authors, and though some of these bear the names of the Buddha's immediate disciples, tradition connects Kâtyâyanîputra with Kanishka's council. This is not a very certain date, but still the inference is that about the time of the Christian era the contents of the Abhidhamma-Pitaka were not rigidly defined and a new recension was possible.
The Sanskrit manuscripts discovered in Central Asia include Sûtras from the Saṃyukta and Ekottara Âgamas (equivalent to the Saṃyutta and Anguttara Nikâyas), a considerable part of the Dharmapada, fragments of the Sutta-Nipâta and the Prâtimoksha of the Sarvâstivâdin school. These correspond fairly well with the Pali text but represent another recension and a somewhat different arrangement. We have therefore here fragments of a Sanskrit version which must have been imported to Central Asia from northern India and covers, so far as the fragments permit us to judge, the same ground as the Vinaya and Suttas of the Pali Canon. Far from displaying the diffuse and inflated style which characterizes the Mahâyâna texts it is sometimes shorter and simpler than our Pali version[650].
When was this version composed and what is its relation to the Pali? A definite reply would be premature, for other Sanskrit texts may be discovered in Central Asia, but two circumstances connect this early Buddhist literature in Sanskrit with the epoch of Kanishka. Firstly the Sanskrit Abhidharma of the Sarvâstivâdins seems to date from his council and secondly a Buddhist drama by Aśvaghosha[651] of about the same time represents the Buddha as speaking in Sanskrit whereas the inferior characters speak Prakrit. But these facts do not prove that Sanskrit was not the language of the canon at an earlier date[652] and it is not safe to conclude that because Asoka did not employ it for writing edicts it was not the sacred language of any section of Indian Buddhists. On the other hand some of the Sanskrit texts contain indications that they are a translation from Pali or some vernacular[653]. In others are found historical allusions which suggest that they must have received additions after our era[654].
I have already raised the question of the relative value attaching to Pali and Sanskrit texts as authorities for early history. Two instances will perhaps illustrate this better than a general discussion. As already mentioned, the Vinaya of the Mûlasarvâstivâdins makes the Buddha visit north-western India and Kashmir, whereas the Pali texts do not represent him as travelling further west than the country of the Kurus. The Sanskrit account is not known to be confirmed by more ancient evidence, but there is nothing impossible in it, particularly as there are periods in the Buddha's long life filled by no incidents. The narrative however contains a prediction about Kanishka and therefore cannot be earlier than his reign. Now there is no reason why the Pali texts should be silent about this journey, if the Buddha really made it, but one can easily imagine reasons for inventing it in the period of the Kushan kings. North-western India was then full of monasteries and sacred sites and the same spirit which makes uncritical Buddhists in Ceylon and Siam assert to-day that the master visited their country impelled the monks of Peshawar and Kashmir to imagine a not improbable extension of his wanderings[655].
On the other hand this same Vinaya of the Mûlasarvâstivâdins probably gives us a fragment of history when it tells us that the Buddha had three wives, perhaps too when it relates how Râhula's paternity was called in question and how Devadatta wanted to marry Yaśodharâ after the Buddha had abandoned worldly life[656]. The Pali Vinaya and also some Sanskrit Vinayas[657] mention only one wife or none at all. They do not attempt to describe Gotama's domestic life and if they make no allusion to it except to mention the mother of Râhula, this is not equivalent to an assertion that he had no other wife. But when one Vinaya composed in the north of India essays to give a biography of the Buddha and states that he had three wives, there is no reason for doubting that the compiler was in touch with good local tradition.