Foreword to the third volume

In the Foreword to the first volume of this translation, and again in that to the second, the translator expressed his intention to give, with this third and final volume, a detailed study of the composition of the whole of the Mahāvastu. It is now seen, however, that such a study could not be contained within the limits of a foreword. A long article, or even a whole volume, would be needed to do anything like justice to the subject. Partial analysis, confined to the more prominent or extensive passages, would not suffice to give a fair conception of the manner and process of the composition of the whole text. The traditions which have gone to the making of the Mahāvastu are so numerous and so varied that a comparative study of them must needs make use not only of Pali texts, canonical and commentarial, but also of other Buddhist Sanskrit works, not to speak of Tibetan and Chinese.

Here, therefore, the translator must confine himself to a few general remarks, forgoing even the quotation of parallel or relevant texts which, in a full study, would be indispensable to the argument. These remarks may consequently seem to be based too much on impressions, but the translator is convinced, even from the little progress he has already made with his full critical and comparative study, that these general remarks give a fair idea of the conclusions to which such a study will lead. Besides, he may justly claim that any mere impressions he may seem to indulge in derive a certain degree of validity from the simple fact that they are the result of a long and close acquaintance with the text of the Mahāvastu.

Some scholars have spoken of an author of our text, others of its redactor or redactors, but a more accurate description of its creation would be gained by imagining a community of monks, over a more or less long period of time, busily engaged, out of motives of piety, in gathering traditions of the Buddha’s life and teaching from all sources accessible to them. Prose versions and metrical versions, and sometimes more than one of each, are often given of one and the same episode, and we can reasonably assume that these are from various traditions that came to the compilers’ notice at different times and from different sources. The narrative, indeed, would often be more coherent if many of these additional versions had been left out. They are not always introduced in the right place. When a narrative is given in mixed prose and verse it sometimes happens that the metrical version overlaps the prose, while at other times, on the contrary, it leaves a gap in the narrative. The addition of a single metrical version to the prose is perfectly natural. For when the two come together the metrical version may be seen to have been added because it was recognised as being more original, and, therefore, it serves as confirmation of the story as related in the prose. Anomalous forms are often, if not generally, common to prose and verse, which is proof that the verse tradition was present to the mind, if not actually in the hands, of the redactor of the prose paraphrase.

It sometimes happens, again, that a repetitive metrical version is not given in full. Parts are left out, and the context can only be understood by reference to the fuller prose version. But this does not necessarily mean that the latter is the earlier or original form of the tradition. Rather it may be taken to imply that the narrator, perhaps out of boredom, leaves out details which are assumed to be already known to the listener or reader.

Where the narrative is given in mixed prose and verse, the metrical version is inserted by way of confirmation of each stage of the story. It all looks very much as if the speaker or writer were paraphrasing in prose a story which he knew in an earlier verse form, and here and there sought to refresh his memory and keep his narrative straight by quoting the verse, which is often introduced with some such formula as tatredamiti ucyate, “here this is said.” The explanatory note on this expression given on page 15 of the first volume is quoted by E. Waldschmidt on page 6 of his edition of the Mahāvadānasūtra,^[1] with reference to the similar view expressed by E. Windisch in his Buddha’s Geburt.^[2] After pointing out that this theory does not hold true of the Mahāvadāna, where, he says, the verses give the impression of being secondary or derivative, he adds, in a footnote on the same page, “beim Mahāvastu liegt der Fall komplizierter. Hier gibt es in der Tat Verse, die sich als sehr altertūmlicher erweisen. Daneben stehen in dieser Zusammenwūrfeluṅg von Ūberlieferuṅgsgut sehr verschiedenen Alters viele ausge-sprochenen junge Verspartien.”

The case is different, of course, when more than one metrical version follows the prose, just as it is when various prose sūtras on the same episodes are given successively in whole or in part. These are but examples of what was described in the Foreword to the second volume^[3] as the proclivity of the compilers of the Mahāvastu to include in this canonical work of their sect every possible piece of tradition which bore on their subject. A good example of this proclivity may be seen in the present volume where as many as four versions are given of the tradition relating the Buddha’s hesitation to preach the doctrine.^[4] Each of the last three traditions is introduced by the formula etthametaṃ śrūyati, “this also is heard here or on this matter,” and if the use in them of the connective particles atha khalu^[5] be taken as a criterion, these traditions are all of them early rather than late. Similarly, three traditions of the temptation by Māra are given,^[6] each beginning in the conventional form with the description of the occasion, and ending with the words itthametaṃ śrūyati.

Repeated editing did, as is only to be expected, produce some changes in the text, and in introducing these changes the editors were sometimes guilty of anachronisms. For example, in the dialogue between Gotama and his father,^[7] the former is made to say that he does not miss the adulation of the court, for he is celebrated instead in Discourse and Exposition, Suttanta and Veyyākaraṇa. Now these were two of the nine divisions of the completed corpus of Buddhist scriptures, which obviously was not in existence at the time. On the other hand, it may be interesting to note that the Mahāvastu nowhere explicitly mentions the Tripiṭaka.

The compilers of the Mahāvastu had, indeed, at one time and another a plethora of traditions to draw upon. When they thought, mistakenly or otherwise, that a narrative in the tradition they were following was not clear or full enough, they would have recourse to another tradition and insert an extract from it. Such an example of an interpolated passage can be seen in the bracketed part of the Gaṅgapāla Jātaka,^[8] which, incidentally, is nearer the Pali version of the story. Another passage bracketed for the same reason is found on page 302. The interpolation here actually does not form a complete sentence, but is interesting in itself as being clearly derived from a version nearer that of the Pali Vinaya than the rest of the narrative.

Another instance of the intermingling of traditions is found in the story of the conversion of Śreṇiya Bimbisāra.^[9] The introduction to the story is peculiar to the Mahāvastu; subsequently there is close verbal parallelism with the version in the Mahāvagga. But long before the end there is interpolated an episode from some other tradition, which seems, however, to be cut short, and the story resumes, with some variations, a parallel course with that in the Pali text.

Examples could be multiplied. Here only one more can be given. After the account of the ordination of the Thirty Monks is finished,^[10] our text begins a story about another group of Thirty Monks. But it does no more than begin it, and passes straight on to the ordination at the end. Evidently there is here a summary of the story as handed down in another tradition. There can be little doubt also that the story which follows it, that of Pūrṇa and his twenty-nine companions, forms yet another tradition of the same theme.

The theory that there are two main strata, each distinguished by its own peculiar style, in the composition of the Mahāvastu, first propounded and developed by Hermann Oldenberg,^[11] has received a large measure of acceptance from other scholars. It is now too well known to need a detailed account of it here. Briefly put, the theory is that the older style is characterised by the use of the connective particles atha khalu; while the newer style uses dāni instead. The earlier style, further, is marked by having its sentences formed with finite verbs, among them a copious supply of aorist tenses. The later style, on the contrary, shows a distinct fondness for nominal construction with participles or participial adjectives supplying the place of finite verbs. These distinguishing criteria have been applied to a fairly large number of passages, but by no means to the whole text, of the Mahāvastu. In the passages which have been subjected to examination on these lines, it has been generally assumed that the distinction in language and style is accompanied by a distinction in the nature of the subject-matter. That is to say, the earlier style is taken to be confined to passages which are more primordially canonical than those in the later style. The former passages generally deal with matters of doctrine, or at least with the historical or legendary occasions when the doctrine was formulated or preached. The later style is more particularly found in passages which are assumed to be accretions to the original tradition. A very characteristic example of accretion is said to be revealed in those passages of the Mahāvastu which purport to give the earlier histories of the lives of disciples or adherents who, in the assumed original tradition, appear on the scene only when they first meet the Buddha. Oldenberg, for example, instances the two versions, in the Mahāvastu and Mahāvagga respectively, of the story of Yaśoda. The former adds to the Pali narrative an account of Yaśoda’s former life, his birth and his upbringing, and this addition is marked by the characteristic features of the later style. Yet, we may note the sentence te dāni āhaṃsu in this so-called accretion, where if the expression te dāni is late, the verb āhaṃsu has every claim to be called early. According to Edgerton^[12] this verbal form, though common in the Mahāvastu, is not found in other Buddhist Sanskrit texts, and he concludes accordingly that “it was discarded early in the history of our dialect.”

Finally, this later style, called by Oldenberg style A, is taken to be marked by a fondness for elaboration, indulgence in detail, fantastic exaggeration of numbers and multiplication of marvels and miracles.

The results which have been reached by the method of criticism here sketchily outlined are, no doubt, suggestive and instructive, and, perhaps, should be accepted as established for the passages which have been put under examination. But it may be doubted whether, in the face of the variety and diversity of the traditions which are collected to form the Mahāvastu, one criterion alone is sufficient to distinguish the late from the early. Even in the passages which have actually been analysed a considerable intermingling of the two assumed styles has often to be dealt with, and attempts made to account for it. To the earlier style, which Oldenberg called style B, a subsidiary style, B, has had to be added. This subsidiary style is apparently to be regarded as style A affected by reminiscences of style B. In other words, the redactor or compiler of a certain episode may have known it or heard it in the language of style B, but, for some reason or other, wished to preserve it for his school or for posterity in the newer style A. But, according to the theory, he could not entirely shake himself free from the influence of the older style, so that now and again he unwittingly falls into its ways and adopts its phraseology. A very cursory examination of our text, however, would seem to indicate that the matter is not so simple as that. For, even if we limit our criterion to the use of atha khalu on the one hand and of dāni on the other, we are still face to face with such a frequent and intricate intermingling of styles that it would seem impossible for a critic with even the highest degree of analytic acumen to distinguish them.

This diversity of styles in the Mahāvastu is matched by a dialectal diversity in its language. There have been many theories about the identity of the Middle Indie dialect on which Buddhist Sanskrit was based, just as there have been about the identity of the language in which the Buddha preached his doctrine. But if we are right in holding that the Mahāvastu is a compilation of traditions from various sources, need we look in it for a single homogeneous language or dialect? The retailers of these various traditions could pass them on in their own vernacular and still be understood by their listeners in any part of the India in which Buddhism grew up. If the northern Buddhists increasingly in course of time tended to replace dialectal variations with normal Sanskrit forms, that was not due to their failure to understand or recite the traditions in their original form, but simply to the prestige of the classical language. It is nowadays generally admitted that Buddhism from the very start was preached in a variety of dialects. For thus only could the new faith be disseminated. Proof of this is found in the Pali Canon. In the Cullavagga (5.33) we read of certain brahmans complaining to the Buddha that the monks were using the vulgar language or dialects, whereas they themselves used, or wished to use, Vedic or Sanskrit. The Buddha’s reply is thus translated by Edgerton:^[13] “You are not to put the Buddha’s words into Vedic; who does so would commit sin. I authorize you, monks, to learn the Buddha’s words each in his own language”. Edgerton goes on to add, “Buddhaghosa and some moderns (notably Geiger) would take the last clause to mean, ‘in the Buddha’s own dialect.’ But Chinese versions make it abundantly clear that the correct interpretation is ‘each in his own dialect.’ The net result seems to me to show, not only that the Buddha authorized and commanded the recitation of his teachings in local dialects everywhere, but also that from the very start there were clearly perceptible differences in the speech of different members of the Buddhist Order, which showed in their ways of reciting the sacred texts.” And, it may be added, once given the Buddha legend, different communities, self-contained as they were, with their own vernaculars and with little inter-communication, would tend to develop the legend each in its own way, a way most likely in accord with, or influenced by, the forms of belief already prevailing among them. Hence, sooner or later, arose a crop of variant traditions.

Not all scholars interpret the Cullavagga passage just cited in the same way. E. J. Thomas,^[14] For example, gives it a different interpretation. According to him, what the brāhmans wished the Buddha to do was to authorise a standard versification of the texts to facilitate the memorising of them. But the brahmans’ expressed complaint was that monks of various clans were corrupting the texts by repeating them in their own grammar (nirutti), that is to say, using the grammatical forms of their respective dialects.

But whatever be the correct interpretation of this passage, the very fact of the brāhmans’ complaint undoubtedly shows that even in the Buddha’s lifetime his doctrine and the legends about him were already current in dialects which were more or less different from the dialect in which they had been first proclaimed. This is the conclusion which Edgerton himself has reached.^[15] He does not agree with H. Lüders and Hiän-lin Dschi, who would make Old Ardha-Māgadhi the language of the original Buddhist Canon, or even the language of the Buddha himself. He is definitely in favour of the contrary opinion of S. Lévi and Lin Li-Kouang, who maintain that the Buddhist teaching and traditions were from the very beginning related in a variety of dialects. Edgerton goes further and says that Buddhist Hybrid Sanskrit, as he calls the language of the Mahāvastu and other Buddhist Sanskrit texts, was never spoken as a vernacular dialect. But, with all due respect to a scholar of his eminence, it is difficult to see what grounds he has for maintaining that a language which was thus never spoken can “have existed for centuries as a religious language” and that it “seems to have become the prevalent language used by north-Indian Buddhists.” This apparently implies that there was a set policy among these Buddhists to frame or concoct a purely conventional language in which to preserve their various canons. It is evident, of course, and only natural that the language of these various texts did, with the passage of time, come to approximate more and more to a certain degree of uniformity. But that uniformity should rather, apart, that is, from what was due to late Sanskritising, be understood as the result of several centuries of re-telling and re-writing the traditions, in the course of which there would be a progressive abandonment of dialectal differences. Besides, as has been already pointed out, even were all the dialects preserved in their pristine purity, they were never so distinct or disparate that they could not coalesce to form the language of a single work.

When we thus relate the diversity of the traditional sources from which the Mahāvastu was compiled to the diversity of dialects in its language, it is not suggested that the dialect of any one particular tradition can be distinguished and isolated, at least readily and certainly. In the first place there was evidently an admixture of dialects at the very beginning. The missionary might not know the language of his converts well enough to translate his message into it with perfection, though there never was more than a dialectal difference between the speech of the two. Secondly, it cannot be assumed that all the Prakrit or Middle Indic dialects of the Buddha’s time have left records or traces behind them, and difficulties of vocabulary or grammatical forms may sometimes be due to our ignorance of a lost dialect to which they belonged. There are in the Mahāvastu, for example, many words and forms which are not found elsewhere, and so cannot be assigned to any particular dialect which is otherwise known. The number of such words and forms which are met with in the Mahāvastu is peculiarly and strikingly large. They cannot all belong to any one dialect or language, for they are not uniformly distributed throughout the text, as should be expected were the work composed in a single homogeneous language. Too many of them are found isolated in one or another particular narrative. Instances of this are the names of some of the articles mentioned in the prose version of the story of Kusa as having been made by that skilled and versatile craftsman.^[17] In the two long lists of trades and occupations given in the present volume,^[17] not only are many of the names of them absolutely unknown, but also only a few of them are identical with those found in similar passages elsewhere, for example, in the Milindapañha.^[18] Corruption of the original form of these words is not sufficient to account for the complete obscurity of their meaning. They must have had meaning somewhere and have been everyday words in the vernacular of the community which developed and preserved this particular form of the story. It may, of course, be argued that there was no other occasion in the Mahāvastu calling for the use of these words. But why are they not found elsewhere, except for the reason that we have no remains of the dialect to which they were native? It cannot be supposed that the articles and trades referred to were confined within the limits of a single community, any more than it can be supposed that the construction of a chariot varied essentially in different parts of India. And yet in the Mañjarī Jātaka we find two unidentified component parts of a chariot described by the entirely unknown and unparalleled terms manesī and kupsara. The same consideration applies to the names of musical instruments which are so frequently mentioned in our text, but several of which cannot be now identified. Here again the distribution of strange terms is not uniform. For example, in the incident where Dīpaṃkara wakes up to find the women of his harem asleep and, some of them, clasping various musical instruments,^[19] the names of these are well-known and fairly easily identifiable. But in the corresponding incident in the life of Gotama is found a much longer list of instruments. Some of these are readily identifiable, others less so, while the few remaining ones are totally strange. This episode was part of the stereotyped tradition of the Buddha’s home life, but it looks as though the form of the tradition relating to Gotama which was incorporated in the Mahāvastu was peculiar to a community with its own particular dialect.

Passing from the names of concrete things to abstract terms, we may notice the unique vyākutsanā, “disgust”, which is unknown outside the Mahāvastu, and even there is found only twice, namely in the story of Śreṇiya Bimbisāra and in what may be termed its corollary.^[20] The latter story, by the way, also contains occurrences of the very rare word pāri, “vessel”, found only here in the Mahāvastu, and related by Edgerton^[21] to Ardha-Māgadhi pārī.

These examples are taken at random from among the more obvious. A minute search could no doubt multiply them indefinitely. The conclusion which, it is here suggested, may be drawn from these lexicographical peculiarities, would seem to be reinforced by a consideration of grammatical forms which are unknown outside the Mahāvastu. Like the strange words which we have found isolated in particular narratives, these grammatical forms, too, are often found similarly isolated. It cannot, for example, be mere accident, nor due to the compilers’ whim, that most of the examples of the anomalous ending matha for the first person plural of the verb should be found grouped together in the metrical version of the Kuśa Jātaka.^[22] It may reasonably be suggested that the metrical version, also, came from a community with its own proper dialect, whether the same community whence came the prose version, or not, may possibly be a question to which close linguistic study might provide an answer.

In phonology, again, we seem to have the same interesting and suggestive phenomena. For instance in the third volume of our text are found several examples of the use of the consonant “d” as a hiatus-bridger, all, with one exception, grouped together on page 54. It looks very much as if the tradition of the ordination of Mahā-Kāśyapa, which is here related, was either composed in a dialect using this mode of linking words, or at least was at some time or other subjected to the influence of such a dialect.

Finally, there is ground for suggesting that the distinction presented by dāni passages on the one hand and atha khalu passages on the other, is not necessarily a distinction between later and earlier styles. We cannot know for certain that these two contrasting styles, in so far as they can be isolated, were not a matter of dialectical differences. It seems to be established that the use of finite verbs which normally went with the use of atha khalu as a connective particle, was not maintained to the same extent in all the dialects of which we have record. For all that can be known some dialects may have abandoned the use of finite verbs entirely at the same time that others were still using them. It may be indeed, that the use of the nominal construction and the connective particle dāni was a mark of less culturally advanced communities. Such a construction would certainly come much more naturally to an untutored narrator than a construction requiring acquaintance with a complete system of verbal conjugation.

In the same way, it need not be supposed that the fondness for the fantastic, the miraculous and the supernatural, which is taken to be characteristic of the style using dāni and the nominal construction, is necessarily a sign of a late tradition. The area in which Buddhism arose does not seem at the time to have felt much of the influence of Brahmanism. In any case, the first converts to the new faith, even though they were brahmans, would be well-acquainted with, even believers in, a rich and luxuriant mythology based upon an extensive polytheism or polydemonism. The world around them was peopled with all kinds of supernatural beings capable of working all kinds of marvels and magic. The Buddha himself never seems to have gone out of his way to deny the existence of these beings, but rather adopted them into his scheme of things. Supernatural or magical events were therefore a matter of everyday belief and experience to those who heard the Buddha and his disciples, and who later went about retailing his doctrine and cultivating the traditions about him. The development of doctrine itself, of course, was controlled and stabilised by the Saṅgha or Order. But no authority could limit the growth of apocryphal stories whether among the monks or among the laity. And it has to be remembered that the Mahāvastu is primarily a collection of stories. Although it is said to be the Vinaya of its school or sect, the doctrinal part could be assumed to be well known enough to be taken for granted and left out, except in so far as the stories themselves might illustrate it. Besides, not all the converts would be chiefly interested in the finer points of Buddhist philosophy and ethics. A large proportion of them, as is the way of uneducated religious converts everywhere, would be more interested in edifying tales and fables, and in correlating what they could of the Buddha legend with the mythology already known to them. For they must already have had their own legends or folk-tales which they could adapt more or less fully to the new faith. Much of the poetry in the Mahāvastu has all the air of being traditional ballads originally quite independent of Buddhism, The verses of good-luck—we may even describe them as protective spells—which the Buddha is made to recite to the merchants Trapuśa and Bhallika^[23] have little or nothing in them that is peculiarly Buddhistic. The metrical Śara-bhaṅga Jātaka,^[24] again, has much in it which leads one to think that it was of non-Buddhist origin; it is so full of commendation of the gentler and more tender feelings which make for social comfort and the pleasure of living. It is also old-fashioned enough to prefer the once popular name of Indra to that of Śakra, his Buddhist supplanter. The story of Elapatra^[25] is another good example of the intrusion of an old fairy-tale into the legend of the Buddha. Finally, how many of the hundreds of Buddhas who are enumerated in the Mahāvastu may not have been originally minor local deities? These Buddhas are not necessarily the product of a dogma developed late in the history of Buddhism. At least, much of the language in which they are enumerated and eulogised looks to be as old as that of any part of our text. The belief in other or former Buddhas must have existed long before it was officially recognised by certain edicts of King Aśoka, some three centuries after the death of Gotama.

It is but natural, of course, that the Buddha legend would be embellished and elaborated in the course of time. But the fact remains that even the earliest recorders of the tradition had the same wealth of legendary or mythological parallels to draw upon as the later. Once the knowledge of the Buddha passed beyond the circle of his intimate friends and acquaintances, such was the force of his personality and the grandeur of his moral character, that ordinary men and women could conceive of him only in terms of the divine. And as soon as that happened there would be no bounds to the play of the mythopoeic fancy of those who had heard of him. If extravagant legends are not so frequent and elaborate in the earliest body of canonical texts that have come down to us, as in the Mahāvastu, that may well be due to editing and pruning by the Theravādin editors. Even so the difference between the Pali texts and the Mahāvastu in this respect is one of degree only, not of kind.

As the very great importance of Edgerton’s work for the study of Buddhist Sanskrit has been referred to, it may not be impertinent, in conclusion, to touch on the question whether the new light thrown by him on so many difficulties, affects in any way the correctness of the translation presented in the first two volumes of this work. The translator is glad to say, that as far as he has been able to discover, no part of his translation calls for any serious correction. Although, before Edgerton’s Grammar became available, the grammatical rationale of many forms may have been obscure, the context in the large majority of cases made the construction of the sentences as a whole self-evident. When this was not the case, comparison with parallel versions in other texts helped out, even if it were only by suggesting emendation of an inexplicable form with a more explicable one. All these doubtful forms should, perhaps, have been discussed in the footnotes. But, as was pointed out in the Foreword to the first volume, Senart, the editor of the text, had himself contributed long notes on most of these anomalous forms, and there was at the time of translating the first two volumes no reliable work to hand which could be used to test the validity of Senart’s explanations. No useful purpose would be served by merely repeating what he had said. If criticism of Senart’s explanations and emendations could be made, as it was on occasion, it was only on the score of the requirements of the context, and often with the support of parallel passages from other texts. Later on, the translator hopes, as part of a critical work on the Mahāvastu, to study his translation in detail, and, where necessary, to revise it, with the help of Edgerton’s dictionary and grammar.

Meanwhile, it is gratifying to record one or two instances where an emendation proposed by the translator has been found subsequently to be confirmed by Edgerton. On page 86 of the second volume a footnote calls attention to the need of emending Senart’s text, svayam va me pravāreya into tvayā ca me pravāreyā, an emendation which was not only called for by the context, but also brought the text of the Mahāvastu into accord with the parallel Pali passage, Jātaka, 3. 284. This emendation is exactly that proposed by Edgerton in his Grammar,^[26] as is also that of the difficult prattam in the same line into yaṃ ca. Again, on page 89 of the same volume there is a footnote on the word anantapāṇī, which Senart in his notes renders “des êtres vivants sans nombre.” The note in the translation, however, suggests that a sense more appropriate to the context would be obtained by reading for pāṇī, “beings”, the word pāna,^[27] “drink” or “water”. In his Dictionary (s.v.) Edgerton makes what is practically the same suggestion, although he is able to improve on it by adducing, instead of the Sanskrit pānīya, the Ardha-Māgadhi pāṇī, likewise meaning “water”.

The translator was also glad to see that the explanation which he offered in the second volume^[28] of the strange word hastokta, which is unknown outside the Mahāvastu, is practically the same as that which Edgerton gives in his Dictionary.

After so much has been said to show the translator’s appreciation of the value and helpfulness of Edgerton’s work, it may appear as rash presumption to dare criticise his interpretation of one or two passages. It may well be that a scholar mining a text like the Mahāvastu for linguistic examples is in danger at times of overlooking the demands of the context and of concentrating overmuch on particular forms and philological principles. It would certainly appear that in one case, at least, Edgerton has been misled by his philological ardour and acumen. In the metrical version of the Campaka Jātaka^[29] the Nāga king’s daughter, in reply to a question about her identity, answers Nāgakanyāhaṃ bhadrante avīcī iha āgatā. In the translation^[30] this has been rendered, “Sir, I am a Nāga maiden come hither on a quest.” But avīcī is a difficult word, and a footnote attempts to explain it as being compounded of avī, a participial form, though otherwise unknown, from av, “ to desire”, and cī “something” or “anything Alternatively, the suggestion is made, on the basis of one MS. reading acīnī, that the correct reading might be arthinī, “seeking”, and a comparison is made with the corresponding Pali passage^[31] which reads atthen(a) amhi idhāgatā, “I am come hither with a purpose.” The context certainly requires some such sense to the expression. Edgerton,^[33] however, translates, “I am a Nāga girl, bless you, come hither from the nether world.” That is, he takes avīcī(d) as an ablative case of avīcī, the name of one of the hells. This would make the Nāga king, Campaka, and his daughter, dwell in Avīci, a hell of torment and terror for the wicked. This is not only improbable in itself, but against what we are expressly told, both in the prose and metrical versions of the story, namely, that they dwelt in Kāśi, where their home had all the wealth and splendour of an abode of the devas.

Another instance where, perhaps, we may suspect that Edgerton has not paid due attention to the text is referred to in a footnote on page 15 of the present volume. In a paragraph of his Grammar,^[34] in which he gives examples of gerunds based on present passive stems, he quotes from our text,^[35] Yāvad... antarpurikāhi antaḥpuraṃ praveśīya krīḍāpanako bhaviṣyati. We are not concerned here with the grammatical point which is the subject of the paragraph, but with Edgerton’s translation of the passage. He translates, “until he (i.e. Kuśa) having been brought into the harem by the harem woman, became (lit. should become) their playfellow.” But yāvad here is better explained as an adverb meaning “as far as”. The four words, rājño priyo ti kṛtvā, which immediately follow yāvad are omitted in Edgerton’s quotation. Kṛtvā with the preceding ti means “quoting”, so that yāvad can be easily explained, as usually in such contexts, as meaning “as far as”. The point is that the story of Kuśa’s various occupations as related in the previous volume is here abridged to give only the first, that of garland-maker, and the last, that of cook in the royal kitchen. Then the events following are assumed to be known as far as the denouement of Kuśa’s becoming dear to the king. The future bhaviṣyati fits in better with the interpretation here suggested. Edgerton, as is seen, has to render it by the past, “became”, adding in parenthesis, “lit. should become.”

One other instance may be given where Edgerton’s preoccupation with philological problems may have led him a little astray. When the Buddha is contemplating the question as to how former Buddhas took their food, he asks himself bhājanapratigrāhakā utāho pāṇipratigrāhakā.^[35] Edgerton^[36] takes pāṇi here to be the Ardha-Māgadhi word for “water”, which has already been referred to, and he renders “did they take a bowl (of food) or water?” But, apart from the need of a parenthesis in such a rendering, the context would be much better fitted by our taking pāṇi in its Sanskrit sense of “hand”. For then we have the two natural alternatives of eating food from a bowl or directly with the hands without the use of any receptacle. The story is here concerned with how the Four Great Kings came to provide the Buddha with a bowl. There is no allusion as to what it was to contain.

These instances of agreement and of disagreement with Edgerton are few enough, and may appear trivial in themselves. But the translator hopes that they may be taken as an indication that he was throughout aware of the linguistic difficulties of the Mahāvastu, and that he took every care to be faithful to the text. To his knowledge, no difficulty, whether of vocabulary or of grammar has been ignored or slurred over. Nothing has been put in his translation which was not warranted by the text, and nothing has been left out on the score of difficulty and obscurity.

A word or two about the Index is necessary. It was thought advisable to group several classes of topics under general headings, as “animals,” “entertainers”, “musical instruments”, “trades and occupations”, and so on. The reason for this is obvious. The Mahāvastu only mentions these topics in passing; there is no discussion of them. But arranged under group-headings they may provide some guidance to the researcher into the natural and economic history of India. It is presumed, of course, that the researcher will not stop short with the translation. He must go back to the text himself. For the interpretation of many of the several terms in the various group-headings is very doubtful, even where they are found translated, and, as has been seen, many others are so obscure as to defy any attempt to render them into English. The exact definition of these awaits further research.

It is pleasant to realise that the translation of the Mahāvastu has, after many years of arduous toil, been at last completed. Some there have been who doubted that the text as a whole was worthy of a full-scale translation, possibly implying thereby that it was mainly material for a study of language, history, folk-tales and other accessory subjects. But how-over useful it may be for students of such subjects, the Mahāvastu must be regarded primarily as a religious work. It is true that it is not a religious classic, though we do well to remember that it is a very near relation of the great religious books of India. It is fully worthy to rank among the other sacred books which have been made accessible to English readers in the Sacred Books of the East and in the Series in which it now appears complete.

Once again, it is a pleasure, and a duty, to put on record the translator’s gratitude to Miss I. B. Horner for her never-failing help. What the work owes to her is to a large extent evident to anyone who looks at the footnotes. When it is added that she revised the whole manuscript, made most valuable corrections and improvements in it, read all the proofs, and was throughout always ready to give generously of her great learning whenever the translator wrote to pose his difficulties to her, the degree of his indebtedness will be still more fully realised.

Another helper, alas, has not lived to see the completion of the work. In the Foreword to the first volume the translator paid tribute to the patience of his wife. Some years have passed since then, but throughout them all, to see her husband devote his leisure hours to this work, to encourage and inspire him to complete it, and to rejoice in its progress had become a real part of her life. It is some consolation to remember that she was spared long enough to see the first proofs of this final volume. Whatever has been achieved in this work, she has a large share in it.

J. J. Jones.

Aberystwyth,
July, 1956.

Footnotes and references:

[1]:

Abhandl. d. deutsch. Akad. d. Wissenschaften. Klasse fūr Sprachen, Litteratur u. Kunst. Jhg. 1952, nr. 2.

[2]:

Pp. 100, 102.

[3]:

P. ix.

[4]:

Pp. 302 ff. Page references, unless otherwise stated, are to the present volume of the translation, where the corresponding pages of the text will be found numbered.

[5]:

See p. xiii below.

[6]:

Pp. 417ff.

[7]:

P. 120.

[8]:

p. 189.

[9]:

P. 439.

[10]:

P. 373.

[11]:

Studien zum Mahāvastu. (Nachrichten von d. k. Gesellschaft d. Wissen-schaften zu Göttingen. Philol.—Hist. Klasse, 1912).

[12]:

Grammar, p. 165. It has been the translator’s great misfortune that Prof. Franklin Edgerton’s truly monumental work, Buddhist Hybrid Sanskrit Grammar and Dictionary, was not available throughout. About two-thirds of the last volume had already been completed when Edgerton’s work appeared. But before the translation was sent to the printer, the whole of it was revised, so that the linguistic and lexicographical difficulties could be examined anew with the aid of his Grammar and Dictionary. The help which the translator has derived from these two books can be judged by the frequency with which Edgerton’s name, or sometimes, for brevity’s sake, only the title of his Dictionary, appears in the footnotes.

[13]:

Grammar, p. 1.

[14]:

The Life of Buddha, p. 254.

[15]:

Grammar, pp. 2 ff.

[16]:

Vol. 2, p. 414 ff.

[17]:

Pp. 111 ff., 443 f.

[18]:

P. 331.

[19]:

Vol. I, p. 183.

[20]:

Pp. 441. 452.

[21]:

B.H.S.D.

[22]:

P. 5.

[23]:

Pp. 293 ff.

[24]:

Pp. 362 ff.

[25]:

Pp. 381 ff.

[26]:

§20.13.

[27]:

Inadvertently for the fuller form pānīya. Here it may be pointed out that the word pāṇa which in vol. 2, p. 20 and on p. 433 of the present volume has been taken as a Sanskrit word for “trader”, is explained by Edgerton as an Ardha-Māgadhi word meaning “a cāṇḍāla or untouchable”. An intensive search would, no doubt, reveal other instances where Middle Indie words should replace words which in the translation have been taken to be Sanskrit.

[28]:

P. 66.

[29]:

Vol. 2, p. 181 (text).

[30]:

P. 175.

[31]:

J. 4.459.

[32]:

Grammar, §10, 67.

[33]:

§35.17.

[34]:

P. 16.

[35]:

Vol. 3, p. 304 (text).

[36]:

Grammar, §22.4.