The Mahavastu (great story)

by J. J. Jones | 1949 | 502,133 words | ISBN-10: 086013041X

The English translation of the Mahavastu (“great story”): one of the important and monumental canonical works of the Mahasanghika school of early Buddhism, a precursor of the Mahayana tradition. The Mahavastu contains three sections narrating the former lives of the Buddha, full of instructive stories, Jatakas and Avadanas. The core of the text ca...

Foreword to the first volume

The following translation of the Mahāvastu was undertaken at the request of the late Mrs. Rhys Davids. As is well known, it was her inspired aim to have all Buddhist scriptures made available for students in translation as well as in the original languages. She worked with such zeal and industry to this end that she was fast approaching the realisation of her aim when she passed away.

With regard to the translation of the Mahāvastu more than one scholar tried to dissuade her from the project, urging chiefly the unsatisfactory state of the text. Senart himself, the editor of the only printed text, had in the introduction to his work expressed the opinion that a complete translation would be ‘à la fois longue, fastidieuse et insuffisante’, and would involve lengthy discussions on linguistic and textual matters. But in reply to such objections Mrs. Rhys Davids would argue from the standpoint of a student of religion. The text, she would say, must be coherent enough and intelligible enough in its broad outline to admit of an English rendering which would be sufficiently correct to give the reader an adequate comprehension of yet another of the books in which the ancient Buddhists had expressed their faith and belief. And if subsequent work on linguistic and textual criticism wrought so many changes in the text that a fresh translation would become necessary, this pioneer effort at a first translation would not be wasted.

The translator, who took up the study of Sanskrit and Pali primarily in order to acquire first-hand acquaintance with the religious literature of India, readily concurred with this view. And his belief in the utility of a translation of this in many ways perplexing text has been strengthened in the course of his work in translating. He is firmly of opinion that no summary of the text and no treatise on it, however lengthy and detailed, can compare in utility to the student of Buddhism with a complete translation. This is not to say that the translator is unaware or inappreciative of the linguistic and textual difficulties of the Mahāvastu. He has throughout endeavoured to solve these to the best of his ability. In view of the immediate purpose of the translation the footnotes are as a rule confined to the elucidation of those difficulties which bear on the interpretation of the text. To go beyond this would involve either much repetition of Senart’s long notes or equally long criticisms of them. At the same time, as much use as possible has been made of Pali texts published or otherwise made known since Senart’s time. As the notes will show, Senart’s conjectures when he was faced with doubtful or unintelligible manuscript readings have in many instances been startlingly successful, being confirmed by parallel passages in Pali texts unknown to him. But in many other instances it will be seen that a manuscript reading rejected by Senart needs to be restored into correspondence with the tradition preserved in Pali texts, or, it may be, in other Buddhist Sanskrit texts.[1]

The exact linguistic or inflexional form of these emendations cannot be readily decided. As is well known the language of the Mahāvastu and other Buddhist Sanskrit texts presents a problem of the first importance. It is usual to term this language Buddhist Sanskrit, but this term conveys nothing as to its origin and its relation to other Indian dialects. While in Mahāyāna texts this Buddhist Sanskrit alternates with more or less classical Sanskrit, the Mahāvastu uses this dialect throughout, though with some degree of Sanskritisation here and there, especially in the prose. As compared with the Buddhist Sanskrit of other texts also, that of the Mahāvastu is decidedly closer to Pali, although it is not easy to say how much of this approximation is due to later copyists of the manuscripts. In not a few instances one manuscript will have a Sanskrit, Buddhist or classical, form where another has a pure Pali one.

We may expect some definite conclusion as to the real origin of Buddhist Sanskrit when Professor Franklin Edgerton, of Harvard University, who has been engaged on a study of this dialect for some time, publishes the result of his researches. Meanwhile we may quote an opinion which he expressed in 1936. ‘The proto-canonical Prakrit on which Buddhist hybrid Sanskrit was based, was a dialect closely related to both Ardhamāgadhi and Apabhraṃśa, but not identical with either.’ (Bulletin of the School of Oriental Studies, Vol. VIII, p. 516.) Perhaps some light may also be thrown on this question by the researches of Professor H. W. Bailey and others into the Buddhistic literatures discovered in recent years and written in Central Asian languages. Certainly, some of these languages would seem to provide some evidence for the phonology, if not the orthography, of some Buddhist terms at the time that Buddhism spread to the north of the Himalayas.

But, however fruitful they may prove to be, linguistic researches alone will not solve all the problems relative to the text of the Mahāvastu. They may here and there prove the greater probability of one inflexional form over the other or enable one to decide how to resolve an apparent metrical anomaly. But such linguistic criticism must take into account the fact that the Mahāvastu is not the composition of a single author written in a well-defined period of time. Rather, it is a compilation which may have been begun in the second century b.c., but which was not completed until the third or fourth century a.d. Even if, as Haraprasad Śāstri (Indian Historical Quarterly, 1, 1925, p. 205) claims, Buddhist Sanskrit was a spoken vernacular of the second century b.c., it would be unreasonable to expect that it could maintain its pristine purity in the Mahāvastu unaffected by the influence of the Pali texts from which so much was apparently taken up, not to speak of the influence of the Mahāyāna literature with its more radical departure from the proto-canonical Prakrit. Linguistic study of the Mahāvastu must, therefore, proceed hand in hand with a study of the various parts of which it is composed and an examination of the probable period in which they were incorporated, as well as of the sources from which they were taken.

For our text is not a homogeneous entity. Although it calls itself the Vinaya of the Lokottaravādins, a branch of the Mahāsāṅghikas, the earliest Buddhist schismatics, this title gives no adequate notion of the nature of its contents. Its peculiar dogma that the personality of the Buddha was docetic, that he was really supramundane (lokottara) and that he only apparently conformed to the habits of men, is, apart from two or three slight allusions, dismissed in one comparatively short passage (1. 168 f.). There is hardly anything about the rules of the Order or the history of their formation, as the title Vinaya would lead us to expect. There is early in Volume I (pp. 2-3) a description of the four kinds of ordination, but this is introduced abruptly and equally abruptly dismissed without being related to any other of the rules of the Order. Our text seems in a hurry to proceed to the more edifying story of the proclamation of Gotama Śākyamuni as a future Buddha by the former Buddha Dīpaṃkara.

As a matter of fact, the Mahāvastu is a collection of practically all the history, quasi-history and legends (avadānas) relating to the Buddha that passed current in the long period during which it was compiled. And if its claim to the title Vinaya is justified it can only be by the fact that the legends it records go back in their origin to the same biographical episodes which were used in the Mahāvagga of the Pali Vinaya to explain or illustrate the origin of the rules of the Order. That there is a very close relation between the Mahāvastu and the Mahāvagga is abundantly proved by the close, practically verbal parallelism between the last quarter or so of the former with the first twenty-four chapters of the latter. These parallels have been set out in detail by Windisch in his Die Komposition des Mahāvastu.[2] Yet in spite of the close resemblance between the two texts there are sufficient differences to warrant the possibility that the Mahāvastu was not copied from the Mahāvagga as we know it, but drew on the same fund of legends. If this is so, then it may be argued that this part of the Mahāvastu is early rather than late. Examination of the language of this part may, when the true origin of that language is settled, help to decide this question.

Here then is one source of the Mahāvastu. But this biographical part of the Vinaya has been enormously expanded after the fashion first set, perhaps, by the Nidānakathā, or introduction to the commentary on the Jātakas. And it is this mass of secondary or derived legends that forms the bulk of the Mahāvastu. This is not to say that it contains no sūtras setting forth Buddhist doctrine. But apart from the special tenet of the Lokottaravādins these show hardly any variation from recognised Theravādin teaching. Minor differences, as for example in the account of dhyāna (jhāna), are discussed in the footnotes, but they are trivial and not of any real significance. The section on the ten bhūmis or stages in the careers of Bodhisattvas may at first sight seem to represent an innovation in doctrine, and mark a late period of transition from Theravāda to Mahāyāna. But the multiplying of the numbers of Bodhisattvas and previous Buddhas is not of itself a Mahāyānist trait, though it prepared the way for the subsequent development of Mahāyāna. The Mahāvastu, also, is careful to stress the fact that the careers described are generalised from the career of the Bodhisattva par excellence, Gotama Śākyamuni. In fact, it would seem that the section on the ten bhūmis was inserted only because it was the policy of the compilers to include in the Mahāvastu every piece of Buddhistic lore that they came across. It is introduced abruptly, and certain inconsistencies in the recital show that it was not really understood. Or, perhaps, the inconsistencies are due to a deliberate attempt to amend or even suppress the Mahāyānist tendencies of other tracts on the same subject. There were, for example, the Mahāyānist Daśabhūmika and Bodhisattvabhūmi, the latter of which was claimed by the Yogācāras as upholding their own particular doctrine. But the Mahāvastu expressly condemns the teaching of this school, for it makes adherence to it on the part of Bodhisattvas one of the causes which prevent them rising from the fifth bhūmi to the sixth.

The fact remains, therefore, that the chief interest of the Mahāvastu lies in its being a collection of Buddhist legends. Although it is styled a Vinaya it almost seems as if, in the course of the period of its compilation, all the elements characteristic of a Vinaya were deliberately omitted. The title Mahāvastu, ‘the great subject,’ no doubt corresponds to the title of the Mahāvagga, just as the Kṣudravastu of the Sarvāstivādins corresponds to the Cullavagga. But by the time the compilation was complete the emphasis had long been laid on the narrative parts of the subject. In almost all the colophons to the chapters the work is styled the Mahāvastu-Avadāna. The compilers indeed came very near achieving a mere collection of avadānas much resembling the collection made by the Sarvāstivādins and known as the Divydvadāna.

Although at first sight these legends seem to be arranged in a haphazard or arbitrary way, the purpose of their recital is in a general way the same as that of the biographical episodes in the Mahāvagga. That is to say they are more or less exegetical narratives. But whereas the narratives of the Mahāvagga explain the occasions of the institution of the rules of the Order, in the Mahāvastu they are introduced to illustrate the virtues of the Buddha in his various lives, and only rarely to explain a point of doctrine. We therefore find these tales, many of them Jātakas, interspersed throughout the whole work. In spite of the apparent incoherence in the order of the contents, there can be detected in the work as a whole something like the scheme of the Nidānakathā. The first volume may thus be seen to correspond to the Dūrenidāna, or incidents in the far past of the Buddha’s career; the second volume and part of the third to the Avidūrenidāna of his more recent history from his birth to his enlightenment, and the latter part of the third volume to the Santikenidāna or the history of the Buddha’s career as teacher and founder of his Order.

But this is not to say that the Jātakas in the Mahāvastu are necessarily reproductions of those we know from Pali texts. It is true that the text of the Mahāvastu tales can sometimes be rectified by reference to the Pali version. On the other hand, a few instances will be found, especially in the second volume, where the text of a Mahāvastu Jātaka will be seen to be superior because it gives a better constructed tale. While only a fraction of the Pali Jātakas are found in the Mahāvastu, there are many others in it which have no corresponding Pali versions, Some of these are obviously folk-tales adapted as Jātakas. Other tales are of the type known as Avadānas, which seem to have been the special creation of the Sarvāstivādins. They are not unlike in their nature to Jātakas which were first fashioned by Theravādins. For an avadāna is a tale in which the heroism or other virtue of a living character is explained by the Buddha as the result of a good deed performed in a previous existence.

The Mahāvagga and the Jātakas are far from being the only parts of Pali scriptures which are to be found incorporated in the Mahāvastu, or, we should more correctly say, which have their parallels in it. There are considerable quotations from other traditional Buddhist literature, as, for example, passages parallel to Pali ones in the Khuddakapāṭha, Vimāna-vatthu, Buddhavaṃsa, Suttanipāta, and the Dhammapada. And though the Pali version has generally the aspect of a more primary version, this is not always or necessarily so. The verses of the Khaḍgaviṣāna-sūtra (i. 357 ff) certainly seem to be more primitive than the corresponding Pali in the Suttanipāta, though the prose framework in which they are embedded is much later. Here, again, a close linguistic study will be necessary to confirm the findings of a study of the internal evidence.

If the three volumes of the translation of the Mahāvastu were being published together it would be possible as well as profitable to draw up a table of all passages in it to which parallels are found in Pali texts, not forgetting also the Buddhist Sanskrit texts. But as only one volume is now being published a comprehensive survey of the whole is impracticable. The present translation is primarily for the student of Buddhism who has no knowledge of Sanskrit, and arguments based on the contents of untranslated volumes would be profitless and even baffling to him. When the third and final volume comes to be published, the translator intends to include in it such a table as that referred to. This table may be rendered still more instructive as to the history of the contents of the Mahāvastu, and, therefore, of Buddhist belief, by the inclusion of evidence derived from Chinese and Tibetan sources. Also, the Central Asian literatures which have been already referred to, and which scholars are daily making more accessible to the general reader, are likely to provide useful material for comparative study.

In the meantime parallel Pali passages are as often as possible indicated in the footnotes. As will be seen from many examples the citation of the corresponding Pali has often been the means not only of restoring the right reading where the manuscript tradition was unintelligible, but also of making explicable many an obscure allusion.

With regard to the translation itself an effort has been made to make it as literal as possible. The reader should not look for a uniformly elevated style. That could not be achieved without departing too much from the form and manner of the original. No succinct literary judgment on the Mahāvastu, which will be true of the whole of it, is possible. Some passages do attain a degree of artistic charm which is worthy of comparison with the best in any literature. This is especially true of some of the verse passages, although these are not necessarily the work of any author associated with the compilation of the Mahāvastu. Many of them are traditional Buddhistic ballads, and owe their charm to the very nature of their origin and mode of dissemination. But however poetic the style, and however strong the temptation to be led by it to make a metrical rendering of the verse passages, it has been deemed more prudent to make a literal translation in prose form, and make them recognisable only by the visual aids of italics and indentation. The wisdom of this procedure was especially made evident whenever there was occasion to consult, for purposes of comparison, some verse passage in the Pali Jātakas. Too often was it found that in spite of the metrical ingenuity of the verse translation, it provided little or no help in the construing of the Pali original. Sometimes, indeed, the translation was seen to be inexact or even incorrect.

The style of the prose is not easy to describe, for there is such a variety of it. The form and manner of passages giving canonical doctrine would, of course, have to be fixed, and would give no scope for any literary ability on the part of the compilers. It is impossible to say whether they were incorporated at a time when they were still faithfully memorised as part of the training of Buddhist monks, or whether they were copied from already written scriptures. But many of the narrative passages have all the appearance of being written directly as they were recited in oral tradition. They are the unadorned tales of the primitive story-teller, for whom every word of the tradition as it had come down to him was sacrosanct. For example, there is the constant repetition of details in the narrative. Nothing is left open to the chance of being forgotten by a fickle memory. If a king decides to send a message he is made to speak out all the details of it as he thinks them out. The message is then given to the messenger and again we have it repeated in the exact words the king had formulated. The message is again repeated in full to the recipient, and if the first recipient is a door-keeper the message will yet again be repeated to his master. Again, a series of actions may be recounted as preliminary to a main action. When the main action comes to be recounted it can only be done by faithfully repeating in the same words all that had gone before. Our story-teller would have none of the adventitious aids of such phrases as ‘when he had done so This is not to forget the frequent occurrence of the phrase evam ukte, ‘when it had been thus said or spoken’. This expression does at first sight seem to serve the purpose of avoiding repetition. But in reality it is as much a feature of a primitive style as the Homeric ‘thus he spoke’, which became a trite conventionality in later epic. Readers of written literature, with leisure to consider the construction of a narrative, would not need to be expressly reminded at the end of it that what they had just read was a quoted speech. But the phrase would be a useful guide or signal to the hearer of an oral recital. Again, there are a few instances where aforementioned events are referred to collectively as artha or prakṛti (‘matter’ or ‘circumstances’), and here, no doubt, we definitely have a literary device for the avoidance of repetition. Whether or no this device occurs in passages which can be demonstrated to be comparatively late, the fact remains that a tendency to repetition is a striking and persistent feature of our text, as, indeed, it is of much of early Buddhist literature. It would, of course, be easy to paraphrase these repetitions, which may seem tedious and puerile to the English reader. But that would be to tamper unduly with what is so characteristic of the style of the Mahāvastu. So these repetitions are as a rule translated in full.

There are repetitions of another order in the Mahāvastu. The compilers were not always satisfied with giving only one version of a legend or episode. Two and more versions are often found, sometimes following one another, sometimes far apart. In the former case the first will generally be in prose, and the others in verse. The legend of the Buddha’s birth is given four times, although in different parts of the work and in connexion with different occasions. These different versions would, no doubt, on close study reveal a difference in dates of composition. For example, of the two accounts in the second volume of the Buddha’s departure from home, the first can readily be seen to be more primitive.

This foreword is not meant to serve as a complete introduction to the contents of the Mahāvastu. These will be found analysed by Senart in his introductions to the three volumes. For a general account of the Mahāvastu and its place in the history of Buddhist literature the reader may be referred to the second volume of Winternitz’ History of Indian Literature and to the article by L. de la Vallée Poussin in the eighth volume of Hastings’ Encyclopedia of Religion and Ethics. The latter quotes largely from Barth (Journal des Savants). The article on the Bodhisattvas by the same author in the second volume of the same encyclopedia should be read in conjunction with Har Dayal’s The Bodhisattva Doctrine in Buddhist Sanskrit Literature (1932). Further there is the work of B. C. Law entitled A Study of the Mahāvastu published in Calcutta in 1930. This contains translations of many extracts. References to the Mahāvastu are found in many modern treatises on Buddhism. In particular, E. J. Thomas in The History of Buddhist Thought (1933) makes a valuable contribution to the study of the Mahāvastu and its place in the development of Buddhist doctrine.

A few words are necessary to explain the treatment adopted for Buddhist terms. Even the general reader with no special knowledge of Buddhism does not require to have explained to him the meaning of terms like karma, dharma and nirvana. These words are therefore left untranslated and are not even commented on in the footnotes. Deva, too, although it was at one time translated ‘angel’, can be assumed to be by this time sufficiently well known to readers of Buddhist texts and translations. Devas were merely good men in the better after-world which they have merited by their goodness, although, as was natural in a system of ethics so highly organised and minutely systematised as that of the Buddhists, there were grades of them determined by the degree of their goodness. The highest classes can hardly be distinguished from gods, and they did actually include some of the deities of the Hindu pantheon like Brahmā and Indra. A late systematisation in the Pali Canon of the various classes of devas divided them into sammuti-devā, or conventional gods (kings, queens, princes), visuddhi-devā, devas by purity (Buddhas and Arhans), and upapatti-devā (the Four Great Lords and Indra, with their companies, etc.).[3] Even the vaguely conceived and still half-animistic supernatural beings of village, field and forest were admitted into the last of these classes, though their gati or sphere of existence was still the earth, and so they could be styled bhūmyā devā (bhumma-devā). A female deva (devī or devatā) figures in several episodes in Buddhist legend. Indeed, if the translator has not gained a wrong impression, the divinities of the lower culture play a rather more prominent part in the Mahāvastu than in other Buddhist works.

Other Buddhist terms for which it would be difficult to find a single English word as an equivalent are left untranslated, but are explained in a footnote when they first occur. For the same reason certain Indian expressions of number, space and time are also left untranslated.

It remains for the translator to make grateful acknowledgment of the ready help rendered him by various scholars. Mrs. Rhys Davids took the greatest interest in the progress of the work. Even more valuable than her constant and expert help was the zeal with which she inspired the translator to carry on when the difficulties seemed insurmountable. It is a matter of great regret that she is not here to see the completion of the work.

Dr. W. Stede read an early draft of part of the translation. He made many suggestions which were not only helpful with regard to the specific points concerned but also served to guide the translator in the rest of his work. He again read part of the manuscript in its final draft, and it was gratifying to have his commendation of the progress shown in this as compared with the first draft. Subsequently Mr. C. A. Rylands, of the School of Oriental and African Studies, read various portions of the manuscript and the translator is indebted to him for light on several difficult points of grammar and vocabulary. Professor H. W. Bailey read part of the manuscript just before it went to press and made some helpful criticisms.

To Miss I. B. Horner, the editor of the Series, the translator’s debt is greater than can be adequately expressed in words. She carefully read the whole manuscript and returned it with sheets full of suggestions for its improvement. The translator thus has had the benefit of Miss Horner’s wide knowledge of Pali literature. Many of the references to parallel passages in Pali, especially the more recondite ones, are due to her, and she has cleared up many a problem of Buddhist philosophy which was hitherto obscure to an inexperienced worker in the field. She has continued giving her generous help right up to the reading of the final proofs. She has earned the translator’s gratitude also by her indefatigable efforts to secure the publishing of the translation, and it is good to think that her efforts are being crowned with success.

But the translator himself must be held responsible for all the faults there may be in his work. If these are unduly many the critic is beseeched to be lenient with one whose enthusiasm for things oriental may have outrun his aptitude.

Lastly a tribute is due to the translator’s wife who has helped by bearing with patience the long and lonely hours of her husband’s withdrawal in his study.

J. J. Jones.

April, 1949.

Footnotes and references:


Pali works are cited by the abbreviated form of their titles used in the Pali Text Society’s Pali-English Dictionary.


Abhandlungen der philolog-hist. Klasse d. K. sächsischen Gesellschaft d. Wissenschaften. Bd. XXIV. No. XIV, 1909 pp. 469 ff.


See MA. 1.33, and other references in Pali Dictionary.