In the last chapter we have seen that in the sixth century B.C. there was in India a very considerable amount of literature of a special sort. Hampered as it was by the absence of written books, by the necessity of learning by heart, and of constantly repeating, the treatises in which it was contained, the extent of the literature is evidence of a considerable degree both of intelligence and of earnestness in effort among the people of India in those days. A great deal of it, perhaps the larger portion of it, has absolutely perished. But a considerable part of the results of the literary activity of each of three different schools has survived. It is by a comparison of three sets of documents, each of them looking at things from a different point of view, that we have to reconstruct the history of the time.
Of these three the surviving books—if books they may be called which had never yet been written—composed and used by those of the brahmins who earned their livelihood by the sacrifices, have been now, for the most part, edited and translated; and a large part of the historical results to be won from them have been summarised and explained. But much remains to be done. The documents of the other two schools may be expected to throw fresh light on passages in the brahmin books now misunderstood. The unhappy system of taking these ancient records in the sense attributed to them by modern commentators with much local knowledge but no historical criticism, with great learning but also with considerable party bias, was very naturally adopted at first by European scholars who had everything to learn. The most practical, indeed the only then possible, course was to avail oneself of the assistance of those commentaries, or of the living pandits whose knowledge was entirely based upon them. In the interpretation of the Vedic hymns this method, followed in Wilson’s translation, has now been finally abandoned. But it still survives in many places in the interpretation of the documents nearest to the date of the rise of Buddhism. And we still find, for instance, in the most popular versions of the Upanishads, opinions that are really the outcome of centuries of philosophic or theosophic discussions, transplanted from the pages of Śankara in the ninth century A.D. into these ancient texts of the eighth or seventh century B.C.
This method of interpretation takes effect in two ways. A passage in the vague and naïve style of those old thinkers (or, rather, poets) is made more exact and precise, is given what is, no doubt, a clearer meaning, by putting into it the later ideas. And in the translation of single words, especially those of philosophic or ethical import, a connotation, which they had really acquired many centuries afterwards, is held applicable at the earlier date. In both these cases a better commentary could be drawn from the general views, and from the exact meaning of philosophic terms, preserved in documents much nearer in time to the Upanishads, though opposed to them on many essential points. As Professor Jacobi says[Footnote: Jaina Sutras, 2. xxvii.]:
“The records of the Buddhists and Jainas about the philosophic ideas current at the time of the Buddha and the Mahāvīra, meagre though they be [he is speaking of the incidental references to the ideas they did not accept], are of the greatest importance to the historian of that epoch.”
Of these records the Pāli ones (thanks, in great part, to the continuous efforts, during the past twenty years, of the Pāli Text Society), are very nearly all now available. We can say not only what they do, but (which is often of even more importance) what they do not, contain. The Jain records are unfortunately as yet known only in fragments. It is the greatest desideratum for the history of this period that they should be made accessible in full. The philosophical and religious speculations contained in them may not have the originality, or intrinsic value, either of the Vedānta or of Buddhism. But they are none the less historically important because they give evidence of a stage less cultured, more animistic, that is to say, earlier. And incidentally they will undoubtedly be found, as the portions accessible already show, to contain a large number of important references to the ancient geography, the political divisions, the social and economic conditions of India at a period hitherto very imperfectly understood.
It is difficult to appreciate the objections made to the authenticity and authority of these documents. The arguments advanced in 1884 by Professor Jacobi[Footnote: Jaina Sūtras, 1. xxxvii.–xlv.] seem quite incontrovertible, and indeed they have not been seriously disputed. The books purport to be substantially the ones put together in the fourth century B.C. when Bhadrabāhu was head of the community. The Jains themselves, of all divisions or schools, acknowledge that there had been older books (the Pūrvas, the Former Ones), now lost. Had they been inventing the story this is not the way in which they would have put it. They would have claimed that the existing books were the original literature of their Order. The linguistic and epigraphic evidence so far available confirms in many respects both the general reliability of the traditions current among the Jains, and the accuracy of this particular detail. Of course the name given in this tradition to the older books cannot have been the original name. They were only “former” as compared with the eleven Angas that are still preserved. And the existing books, if of the fourth century, can only be used with critical care as evidence of institutions, or events, of the sixth century B.C. Still, even so, we have here important materials for Indian history, at present only very imperfectly utilised.
It is really much the same with the existing records of the other school, of the men we now call Buddhists. They have as yet been only very imperfectly utilised, though they are better and more completely known than the last. This is partly, no doubt, because we call them Buddhists, and imagine them, therefore, to belong to a separate class, quite distinct from other Indians of that epoch. The Buddhists were, as a matter of fact, characteristically and distinctively Indian. They probably, at least during the fourth and third centuries B.C., formed the majority of the people. And the movement of thought out of which all these schools arose, so far from being a negligible quantity, as the priestly books suggest, was one of the most dominant factors the historian of the fourth, fifth, and sixth centuries B.C. has to consider.
As to the age of the Buddhist canonical books, the best evidence is the contents of the books themselves—the sort of words they use, the style in which they are composed, the ideas they express. Objection, it is true, has recently been raised against the use of such internal evidence. And the objection is valid if it be urged, not against the general principle of the use of such evidence, but against the wrong use of it. We find, for instance, that Phallus-worship is often mentioned, quite as a matter of course, in the Mahābhārata, as if it had always been common everywhere throughout Northern India. In the Nikāyas, though they mention all sorts of what the Buddhists regarded as foolish or superstitious forms of worship, this particular kind, Siva-worship under the form of the Linga, is not even once referred to. The Mahābhārata mentions the Atharva Veda, and takes it as a matter of course, as if it were an idea generally current, that it was a Veda, the fourth Veda. The Nikāyas constantly mention the three others, but never the Atharva. Both cases are interesting. But before drawing the conclusion that, therefore, the Nikāyas, as we have them, are older than the existing text of the Mahābhārata, we should want a very much larger number of such cases, all tending the same way, and also the certainty that there were no cases of an opposite tendency that could not otherwise be explained.
On the other hand, suppose a MS. were discovered containing, in the same handwriting, copies of Bacon’s Essays and of Hume’s Essay, with nothing to show when, or by whom, they were written; and that we knew nothing at all otherwise about the matter. Still we should know, with absolute certainty, which was relatively the older of the two; and should be able to determine, within a quite short period, the actual date of each of the two works. The evidence would be irresistible because it would consist of a very large number of minute points of language, of style, and, above all, of ideas expressed, all tending in the same direction.
This is the sort of internal evidence that we have before us in the Pali books. Any one who habitually reads Pali would know at once that the Nikāyas are older than the Dhamma Sangaṇi; that both are older than the Kathā Vatthu; that all three are older than the Milinda. And the Pali scholars most competent to judge are quite unanimous on the point, and on the general position of the Pali literature in the history of literature in India.
But this sort of evidence can appeal, of course, only to those familiar with the language and with the ideas. To those who are not, the following points may be suggestive:
On the monuments of the third century B.C. we find the names of donors—donors of different parts of the building—inscribed on those parts (pillars, rails, and bas-reliefs). When the names are common ones, certain epithets are added, to distinguish the donors from other persons bearing the same name. Such epithets are either local (as we might say, John of Winchester) or they specify an occupation (as we might say, John the carpenter, or John the clerk) or are otherwise distinctive. Among these epithets have been found the following:
1. Dhamma-kathika.—“Preacher of the System” (the Dhamma)—the “System” being a technical term in the Buddhist schools to signify the philosophical and ethical doctrine as distinguished from the Vinaya, the Rules of the Order.
2. Peṭakin.—“One who had (that is, knew by heart) the Piṭaka.” The Piṭaka is the traditional statements of Buddhist doctrine as contained in the Sutta Piṭaka. The word means basket, and as a technical term applied to a part of their literature: it is used exclusively by the Buddhists.
3. Suttantika.—“A man who knows a Suttanta by heart.”
4. Suttantakinī.—“A woman who knows a Suttanta by heart.” Suttanta is, again, a technical term used exclusively of certain portions of the Buddhist canonical books, more especially of the Dialogues. It means literally the “end of the Suttas.” In its technical sense it is the aim, object, outcome of them; and is applied to the Dialogues as giving, in a more complete and elaborate form, the general result of those shorter Suttas on which they are based.
The brahmins have an analogous term, Vedānta, applied, in post-Buddhistic writings, at first in the Śvetāśvatara and Muṇḍaka Upanishads and often afterwards, to the Upanishads, as being the highest outcome of the Vedas. Previously to this the word is only found in its literal sense, “the end of the Veda,” and the secondary sense is, therefore, probably adapted from the corresponding (and earlier) Buddhist term.
5. Panca-nekāyika.—“One who knows the Five Nikāyas by heart.” The five Nikāyas, or “Collections,” as a technical term used of literary works, is applied to the canonical Buddhist texts, and to them only. Of the five, the first two contain the Suttantas, the next two are made up of Suttas arranged in two different ways, and the fifth is a supplementary collection, mostly of later works.[Footnote: See American Lectures, pp. 60–62.] As the word Nikāya also means a school, or sect, it is somewhat ambiguous, and was gradually replaced by the word Āgama, continually used in the later Sanskrit literature. The same remark holds good of the technical term Suttanta. That also was gradually replaced by the shorter and easier phrase Sutta.
The expressions here explained are used on Buddhist monuments and refer to Buddhist books. They are conclusive proof that some time before the date of the inscriptions (that is, roughly speaking, before the time of Asoka), there was a Buddhist literature in North India, where the inscriptions are found. And further, that that literature then had divisions known by the technical names of Piṭaka, Nikāya, and Suttanta, and that the number of Nikāyas then in existence was five.
But this is not all. Asoka, in his Bhabra Edict, addressed to the Buddhist Order (the Sangha), recommends to the Brethren and Sisters of the Order, and to the lay disciples of either sex, frequently to hear (that is to learn by heart), and to meditate upon, certain selected passages. And of these he, most fortunately, gives the names. They are as follows:
Ariya-vasāni (now found in the Dīgha Nikāya, in the portion called the Sangīti Suttanta).
Anāgata-bhayāni (now found in the Anguttara Nikāya, vol. iii. pp. 105–108).
Muni Gāthā (now found in the Sutta Nipāta, verses 206–220).
Moneyya Sutta (now found in the Iti-vuttaka, p. 67, and also in the Anguttara Nikāya, vol. i. p. 272).
Upatissa Pasina.—“The question put by Upatissa” (more commonly known as Sāriputta). There are so many such questions in the books that opinions differ as to which of them is the one most probably referred to.
There is a word at the commencement of this list which may either be an adjective applied to the whole list, or the name of another passage. However this may be, this Edict of Asoka’s gives the actual titles of some of the shorter passages included, in his time, in those books, the larger divisions of which are mentioned in the inscriptions just referred to.
Now the existing literature, divided into the same larger divisions, contains also the shorter passages. To suppose that it was composed in Ceylon is to suppose that, by an extraordinary series of chances, the Ceylon writers happened to hit upon just the identical technical terms, two of them then almost fallen out of use, that had been used in these old inscriptions (of which they knew nothing) for the names they gave to the larger divisions of the literature they made. And we must further suppose that, by another extraordinary series of chances, they happened to include in those divisions a number of shorter passages, each of them corresponding exactly to those mentioned by name, long before their time, in Asoka’s Edict, of which also they knew nothing. To adopt such a theory as the most probable explanation of the facts would be nothing less than absurd.
How is it, then, will be the immediate question, that this theory in almost, if not in all, the current books on Buddhism or on Indian history is taken for granted; that the Pali canonical literature is always called “the Southern Recension” or “the Singhalese Canon”?
The expression is ambiguous, and apt to be misleading. But though it is doubtless sometimes used in such a way as to suggest that these books were composed in Ceylon, this is not its real meaning, and it is never so used by careful writers. It simply means that of the few works known to the European scholars who first studied Buddhism, the MSS. of some came from Ceylon; and that such works were therefore called southern, to distinguish them from the others, known from MSS. which had come from Nepal, and therefore called northern.
It is very possible that Burnouf, to whom the popularity of this mode of speech is mainly due, leaned at first to the opinion that the canonical works had been actually written in Ceylon. He always spoke of them in his first work as “the Pali books of Ceylon,” not as “the Pali books of India.” But that phrase is also ambiguous. Very conscious how meagre, and for the most part how late, were the works he used, he was much too careful a scholar to express, at first, any clear opinion at all. At the end of his long labours, however, he certainly was quite clearly of the contrary opinion. For at the very close of his magnificent work, at p. 862 of the “Lotus,” he suggests that the Pali works “may have been popular among inferior castes, and the great mass of the people, in Magadha and Audh, while the Buddhist Sanskrit works were in use among the brahmins.” He at that time regarded them all, herefore, as North Indian works. And considering that he knew nothing of the inscriptions, and had only the internal evidence to guide him, this suggestion, though not exactly right, reflects the greatest credit on his literary judgment. Had he started with this view, we should probably have been saved the use of the ambiguous phrases, so suggestive of these works being written in Ceylon, which have had so great an influence in retarding the acceptance of the view that that great pioneer in Buddhist studies came at last, himself, to hold.
Not only ought such phrases to be dropt out of any works, on these subjects, claiming to be scholarly; but even the phrases “northern” and “southern” should be avoided. This seems a pity, for they look so convenient. But the convenience is delusive if they convey a wrong impression. And I venture to assert that most people draw the conclusion that we have two distinct Buddhisms to deal with, one made in Nepal, the other made in Ceylon. Every one now agrees that this is all wrong. What we have is not two, but very many different sorts of Buddhism; for almost every book gives us a different doctrine.
The more authoritative and ancient books, whether written in Pali or in Buddhist Sanskrit, are none of them either northern or southern. They all, without any exception,—if we disregard the absurdly unimportant detail of the place from which our modern copies of them are derived,—claim to belong, and do actually belong, to the Middle Country, as the Indians call it, that is, to the Ganges Valley. Each differs from the next (in point of date) by small gradations in doctrine. There are such differences even within the Nikāyas themselves. Many Sanskrit books, though they differ, by containing certain details of later opinion, from the oldest Pali ones, still, on the whole, have to be classed with the Pali rather than with the other Sanskrit works. The Sanskrit Mahā Vastu, for instance (“The Sublime Story”) is much nearer to the Pali Cariyā Pitaka (“The Tradition as to Conduct”) than it is to such Sanskrit books as the “Lotus of the Good Law.” All three alike had their origin in the Middle Country—where exactly, in that country, we cannot, with respect to any one of the three, determine. The only two ancient works we can specify as distinctly northern in origin, the Milinda and the Gosinga Anthology, are neither of them written in Sanskrit, and are identical in doctrine with what is called southern Buddhism. Is it not rather absurd to have to ticket as southern just the very two books we know to be the most northern in origin?
There is not now, and never has been, any unity either of opinion or of language in what is called northern, or in what is called southern Buddhism. There is a distinct disadvantage in continually suggesting a unity which has no existence in fact. In a word, the current division of Buddhist literature into northern and southern is entirely unscientific, and misleading. It contains a suggestio falsi in at least two important respects. It cuts across the only division that has a scientific basis, the division, not according to the locality whence we get our modern copies, but according to time, according to date of origin. Why then continue the use of an ambiguous phraseology which may be (and which we know, from experience, will be) misunderstood? The only way to avoid endless confusion is to drop the use of it altogether. And I take this opportunity of acknowledging my error in having used it so long myself. In my Buddhism, from the fifteenth edition onwards the mistake has been corrected. So slight is the change that no one is likely to have noticed it. The word “northern” has been replaced by “Tibetan,” “Japanese,” “Mahāyanist,” etc., according to the context. There has been no loss in clearness, or in conciseness, and much gain in precision.
We must take our Pali canonical books then to be North Indian, not Singhalese in origin; and the question as to whether they have suffered from their sometime sojourn under the palm groves of the mountain vihāras in the south must be decided by a critical study of them in their present condition. Toward such a study there are some points that can already be made.
The books make no mention of Asoka. Had they undergone any serious re-editing after the reign of the great Buddhist Emperor (of whom the Buddhist writers, whether rightly or wrongly, were so proud), is it probable that he would have been so completely ignored?
The books never mention any person, or any place, in Ceylon; or even in South India. They tell us a goodly number of anecdotes, usually as introductions to, or in illustration of, some ethical point. It would have been so easy to bring in a passing reference to some Ceylon worthy—in the same way as the brahmin Buddhaghosa does so often, in his Attha Sālinī, which was revised in Ceylon.[Footnote: See Mrs. Rhys Davids’s Buddhist Psychology, p. xxi.] If the Piṭaka books had been tampered with, would not opportunity have been taken to yield to this very natural impulse?
We know a great deal now of developed or corrupted doctrine current in Ceylon, of new technical terms invented, of new meanings put into the older phrases. Not one single instance has yet been found of any such later idea, any such later form of language, any such later technical term, in any one of the canonical books.
The philosophic ideas of the ancient Buddhism, and the psychological ideas on which they were based, were often curtly, naïvely, confusedly expressed. In Ceylon they had been much worked up, polished, elucidated, systematised. From several works now accessible we know fairly well the tone and manner of these later—and, as they must have seemed to Ceylon scholars, clearer, fuller—statements of the old ideas. In no single instance yet discovered has this later tone and manner found its way into the canonical books.
It would seem, then, that any change that may have been made in these North Indian books after they had been brought into Ceylon must have been insignificant. It would be a great advantage if we should be able to find even one or two instances of such changes. We should then be able to say what sort and degree of alteration the Ceylon scholars felt justified in making. But it is clear that they regarded the canon as closed.
While the books were in North India, on the other hand, and the canon was not considered closed, there is evidence of a very different tone. One whole book, the Kathā Vatthu, was added as late as the time of Asoka; and perhaps the Parivāra, a mere string of examination questions, is not much older. One story in the Peta Vatthu[Footnote: IV. 3.] is about a king Pingalaka, said in the commentary to have reigned over Surat two hundred years after the Buddha’s time; and another[Footnote: V. 2.] refers to an event fifty-six years after the Buddha’s death. The latter is certainly in its right place in this odd collection of legends. The former may (as the commentator thinks) have been added at Asoka’s Council. Even if it were, that would be proof that they then thought no harm of adding to the legendary matter in their texts. And the whole of this little book of verses, together with the Vimāna Vatthu (really only the other half of one and the same work), is certainly very late in tone as compared with the Nikāyas.
The same must be said of two other short collections of ballads. One is the Buddha Vaṃsa, containing a separate poem on each of twenty-five Buddhas, supposed to have followed one another in succession. The other is the Cariyā Piṭaka, containing thirty-four short Jātaka stories turned into verse. Both of these must also be late. For in the Nikāyas only seven Buddhas are known; and Jātakas, in the technical sense, are not yet thought of. This particular set of Jātakas is also arranged on the basis of the Pārāmitās, a doctrine that plays no part in the older books. The Ten Perfections (Pārāmitā) are qualities a Buddha is supposed to be obliged to have acquired in the countless series of his previous rebirths as a Bodhisatva. But this is a later notion, not found in the Nikāyas. It gradually grew up as the Bodhisatva idea began to appeal more to the Indian mind. And it is interesting to find already, in these latest of the canonical books, the germs of what afterwards developed into the later Mahāyāna doctrine, to which the decline of Buddhism, in the opinion of Professor Bhandarkar, was eventually so greatly due.[Footnote: J.R.A.S., Bombay Branch, 1900, p. 395.]
This question of the history of the Jātaka stories will be considered in greater detail in our next chapter. What has been here said (and other similar evidence will, no doubt, be hereafter discovered) is amply sufficient to show that some parts of the Canon are later than others; and that the books as we have them contain internal evidence from which conclusions may fairly be drawn as to their comparative age. Such conclusions, of course, are not always so plain as is the case in the four instances—the Peta and Vimāna Vatthus, the Buddha Vaṃsa, and the Cariyā Piṭaka—just considered. For example, let us take the case of the Sutta Nipāta.
This also is a short collection of poems. It contains fifty-four lyrics, each of them very short, arranged in four Cantos; and then sixteen others, as a fifth Canto, strung together by a framework of story. The last Canto (called the Pārāyana) had evidently once existed as a separate poem. It is so treated by the commentator, who calls it a Suttanta; and it is in fact about as long as one of those Suttantas in the Dīgha Nikāya which consist of verses strung together by a framework of story in prose. It is six times quoted or referred to by name, as a separate poem, in the Nikāyas.[Footnote: Saṃyutta, 2. 49; Anguttara, 1. 144; 2. 45; 3. 399; 4. 63.]
The preceding Canto, the fourth, is called “The Eights,” most of the lyrics in it containing eight stanzas apiece. This Canto is also referred to by name as a separate work, in other parts of the Canon.[Footnote: Saṃyutta, 3. 12; Vinaya, 1. 196; Udāna, 5. 6.] And it must in very earlier times have been already closely associated in thought with the fifth Canto, for the two together are the subject of a curious old commentary, the only work of the kind included in the Nikāyas. That this commentary, the Niddesa, takes no notice of the other three Cantos would seem to show that, when it was composed, the whole of the five Cantos had not yet been brought together into a single book.
Of the thirty-eight poems in the earlier three Cantos no less than six are found also in other parts of the Canon.[Footnote: Poem No. 4 = S. 1. 172; No. 8 = Kh. P. No. 9; No. 13 = Kh P. No. 5; No. 15 = Jāt. 3. 196; No. l6 = Kh. P. No. 6; No. 33 = M. No. 92.] They had existed as separate hymns, popular in the community, before they were incorporated into the several collections in which they are now found. When we find also that numerous isolated verses in these thirty-eight poems occur elsewhere in very ancient documents, the most probable explanation is that these were current as proverbs or as favourite sayings (either in the community, or perhaps among the people at large) before they were independently incorporated in the different poems in which they are now found.
We find, then, that single verses, single poems, and single Cantos, had all been in existence before the work assumed its present shape. This is very suggestive as to the manner of growth not only of this book, but of all the Indian literature of this period. It grew up in the schools; and was the result rather of communistic than of individual effort. No one dreamed of claiming the authorship of a volume. In the whole of the Buddhist canonical works one only, and that the very latest, has a personal name attached to it, the name of a leading member of the Order said to have lived in the time of Asoka. During the previous three centuries authorship is attributed not to treatises, or even poems, but only to verses; and to verses in two only out of the many collections of verses that have been preserved. Out of twenty-nine books in the Canon no less than twenty-six have no author at all, apart from the community.
This is decisive as to popular feeling on the point. And even in the priestly schools the then prevalent custom was not greatly different. Their works also were not produced by individuals, but grew up in the various schools of the priestly community. And no priestly work ascribed to an individual author can be dated much before the time of Asoka.
And yet another point, which will turn out, unless I am much mistaken, to be of striking importance for the history of Indian literature, arises in connection with the Sutta Nipāta. The fifth Canto regarded as a single poem, and about one-third of all the other poems in the collection, are of the nature of ballads. They describe some short incident, the speeches being always in verse, but the story itself usually in prose (though in a few instances this also is in verse). They resemble in this respect a very large number of Suttas found in other portions of the Canon. And even a few of the Suttantas— such as the “Riddles of Sakka,” for instance (certainly one of our oldest documents, for it is quoted by name in the Saṃyutta[Footnote: Saṃyutta, iii. 13.])—are characteristic specimens of this kind of composition. It is, in fact, next to the prose Sutta, the most popular style for literary effort during this period.
This manner of expressing one’s ideas is now quite unknown. But it has been known throughout the world as the forerunner of the epic. Professor Windisch has subjected those of these ballads that are based on the temptation legends to an exhaustive study in his masterly monograph, Mara und Buddha. He says, apropos of the two ballads on this subject in the Sutta Nipāta:
“These two Suttas might have been regarded as a fragment of an epic had we otherwise found any traces of an ancient Buddha Epic. But that is not to be thought of. Far rather are these Suttas to be looked upon as the early beginnings out of which, in certain circumstances, a Buddha Epic could eventually arise.
“We can mark with special ease how an Epic arises, and of what process an Epic, as a particular form of literature, is the consummation. Some years ago I drew attention to the historical points we have here to take into consideration in a lecture to the Congress of philologians at Gera on the Irish legends and the question of Ossian.[Footnote: Revue Celtique, 5. 70.] There I laid the chief stress on the old-Irish legends, but compared also the legends in ancient India. The latter subject was independently dealt with by Oldenberg in his well-known articles on the Ākhyāna hymns where the subject referred to (the relations of the Epic to previous literary forms) is dealt with in detail and thoroughly explained.[Footnote: Z. D. M. G., vols. 37 and 39.] Professor Geldner then considered the same subject, partly from new points of view, inasmuch as he followed them out also in the case of the Avesta, in his article in the ‘Vedische Studien.’[Footnote: 1. 284, foll.] Now we find also in the Buddhist literature, as Oldenberg was the first to point out, this epic narrative in mixed prose and verse.… The persons who act, the place where they act, and the action itself form the constituent elements of the narrative. But the latter only springs into life when the persons acting are also represented as speaking. Now the speeches are frequently what it is least possible to keep historically accurate, where, therefore, the fancy of the narrator and the art of the poet come most into play. Conversation (speech and rejoinder) is the first part of the narrative to be put into verse, and that especially at the crucial points of the story. Here the beginnings of epic and drama lie close together. That the more ancient epics in all countries contain many speeches and counter speeches can be seen too from the Iliad. It is only in the later epic form that this dramatic element is kept in the background. So in the old-Greek drama also we have an epic element in the speeches of the messengers. But a poem becomes completely epical only when to the speeches in verse is added also the framework of the story in metrical form. And the last stage is that the speeches grow shorter, or fall out, and only events are given in verse.”[Footnote: Windisch, Mara und Buddha, pp. 222, foll.]
Both the general accuracy and the great importance of this far-reaching generalisation will be admitted by all. Now we have in the Nikāyas all sorts of the earliest forms of the evolution referred to. We find (in the Thera- and Therī-Gāthā, for instance) only the speeches in verse in the canonical books, and the framework of prose, without which they are often unintelligible, handed on, by tradition, in the Commentary. We find (as in the Suttantas in the second volume of the Dīgha, or in the Udāna) speeches in verse, and framework in prose, both preserved in the canonical book. And we find ballads (such as the two Suttas discussed by Professor Windisch) in which speeches and framework are both preserved in verse. But it is not till long afterwards, in the time of Kanishka, that we have a fully developed Buddha Epic.
Are we then to suppose that the Indians had a rnental constitution different from that of the other Aryan tribes (after all, their relatives in a certain degree) throughout the world? Or are we to suppose that the Buddhist community formed a section so completely cut off from the rest of the people that they were uninfluenced by the existence, in their immediate surroundings, of the great Indian Epics. The Rāmāyaṇa, as Professor Jacobi has shown, was composed in Kosala, on the basis of ballads popularly recited by rhapsodists throughout that district. But the very centre of the literary activity of the Buddhists was precisely Kosala. After the Rāmāyaṇa had become known there as a perfect epic, with the distinctive marks of the epic style, would such of the people in Kosala as had embraced the new doctrine have continued to use only the ancient method of composition? This would be quite without parallel. But we have to choose between this supposition (not a probable one) and the alternative proposition—that is to say, that whatever the date to be assigned to this ballad literature, in mixed prose and verse, preserved in the Nikāyas, the date of the Mahā-bhārata and of the Rāmāyaṇa, as Epics, must be later.
We may be pretty sure that if the Epics had existed at the period when this Buddhist literature was composed, they would have been referred to in it. But they are not. On the other hand, the ballads in prose and verse, such as those sung by the rhapsodists (the stage out of which the epics were evolved), are referred to under their technical name of akkhānas (Sanskrit ākhyānas) in one of the oldest documents.[Footnote: Dialogues of the Buddha, 1. 8.] Mention is there made of various sorts of public spectacles, and one of these is the reciting of such Ākhyānas. And when the commentator in the early part of the fifth century A.D. explains this as the reciting of the Bhārata, the Rāmāyaṇa, and so on, that is, as exegesis, perfectly right. This was the sort of thing referred to. But his remark is evidence of the existence of the perfect Epics, only at his own time, not at the time of the old text he is explaining.
This may seem, I am afraid, to have been a digression. But it is really very much to the purpose, when discussing Indian literature in this period, to bring out the importance of the wide prevalence of the versifying faculty, and to discuss the stage to which it had reached, the style of composition in which it was mostly used. We hear of four kinds of poets:—the poet of imagination (who makes original verses): the poet of tradition (the repeater of current verses); the poet of real life (or perhaps of worldly as distinct from religious topics); and the improvisatore.[Footnote: Anguttara, 2. 230; compare Sum. 95.] We have several instances in the books of such impromptu verses. Though they were probably not quite so impromptu as they are described to be, we need not doubt the fact that the art was then a recognised form of ability. And when a man is charged with being “drunk with poesy”[Footnote: Saṃyutta, 1. 110.] (kāvey-yamatto) the rapt and far-away look of the poet in the moment of inspiration cannot have been altogether unfamiliar.
It is interesting to notice that, just as we have evidence at this period of the first steps having been taken towards a future Epic, so we have evidence of the first steps towards a future drama—the production before a tribal concourse on fixed feast days of shows with scenery, music, and dancing. There is ample evidence in the Buddhist and Jain records, and in Asoka inscriptions, of the existence of these samajjas, as they were called, as a regular institution.[Footnote: See the passages quoted in Dialogues of the Buddha, 1. 9. 10, and Jacobi’s Jaina Sutras, 2. 303.] That they are not mentioned in the priestly books need inspire no doubt upon the point. This is only another instance of the priestly habit of persistently ignoring what they did not like. We see from the Sigālovada Suttanta[Footnote: In Grimblot’s Sept Suttas Palis, p. 300, where the reading must be corrected accordingly.] that recitations, or the telling of stories, in mixed prose and verse (akkhāna), also took place at these meetings. But this seems, from the evidence at present attainable, to have been distinct; and the interpretation of the word I have rendered “scenery” is open to doubt. We cannot talk, therefore, as yet, of drama. When we see, however, that these meetings took place at sacred places, on the hilltops, and that high officials were invited and had special seats provided for them, we find ourselves in presence, not of private undertakings, but of such religious and communal ceremonies as those to which the beginnings of drama have elsewhere also been traced back. It is true that the kind of religion which we have here to consider is not the religion of the brahmins. The general prohibition which forbade a brahmin to see or hear dancing or music[Footnote: See, for instance, the Pāraskara Gṛhya Sūtra, 2. 7. 3.] must have included such performances. But it was at that time none the less on that account, a very vital and popular part of the national faith.[Footnote: The oldest dramas mentioned by name (second century B.C.) are mystery plays based on episodes in the life of Krishṇa. From this time onward there is more frequent mention of actors. But the earliest dramas are all lost. The oldest extant ones are of the sixth or seventh century A.D.]
I have dealt in this chapter, not with the contents, which I have described elsewhere,[Footnote: American Lectures, chapter ii.] but only with the outward form and style of the literature. It shows a curious contrast between the value of the ideas to be expressed and the childlike incapacity to express them well. We have here, as to style, only the untrained adolescence of the Indian mind. But what vigour it has! The absence of writing materials seems naturally to have affected less the short poems than the style of the prose, and there is much rough and rugged beauty both in the ballads and in the lyrics. Now the style, and much of the thought, is not Buddhist but Indian; and is in some respects the only evidence we possess of the literary ability, at that time, of the Indian peoples. If only we had still some of the ballads out of which the Epics were subsequently formed, they would, I am convinced, show equal limitations, but also equal power. In after times we have evidence of more successful study of the arts and methods of rhetoric and poetry. But never do we find the same virility, the same curious compound of humour and irony and love of nature on the one hand, with a deadly earnestness, and really on the whole a surprisingly able grasp of the deepest problems of life, on the other. As we shall see presently in the case of the philosophy, so also is it true of the literature that it is in this period that India came nearest to having a Golden Age. And the learned, ornate poetry of later times is to the literature of this period what the systemisations and learned commentaries of Buddhaghosa and Śankara are to the daring speculations and vivid life of the early Upanishads and of the Four Nikāyas.
APPENDIX TO CHAPTER X
CHRONOLOGICAL TABLE OF BUDDHIST LITERATURE FROM THE BUDDHA’S TIME TO THE TIME OF ASOKA
1. The simple statements of Buddhist doctrine now found, in identical words, in paragraphs or verses recurring in all the books.
2. Episodes found, in identical words, in two or more of the existing books.
3. The Sīlas, the Pārāyana, the Octades, the Pātimokkha.
4. The Dīgha, Majjhima, Anguttara, and Saṃyutta Nikāyas.
5. The Sutta Nipāta, the Thera- and Therī-Gāthās, the Udānas, and the Khuddaka Pāṭha.
6. The Sutta Vibhanga and the Khandakas.
7. The Jātakas and the Dhammapadas.
8. The Niddesa, the Itivuttakas, and the Paṭisambhidā.
9. The Peta- and Vimāna-Vatthus, the Apadānas, the Cariyā Piṭaka, and the Buddha Vaṃsa.
10. The Abhidhamma books; the last of which is the Kathā Vatthu, and the earliest probably the Puggala Paññatti.
The above table represents the probable order in which the extant Buddhist documents of this period were composed. They were not yet written, and a great deal has no doubt been lost.