Hinüber - Hoary Past And Hazy Memory (JIABS-29-2-193)

Hoary Past And Hazy Memory - On The History Of Early Buddhist Texts¹

Oskar V. Hinüber

Journal of the International Association of Buddhist Studies Volume 29 • Number 2 • 2006 (2008) pp. 193–210

On the occasion of the 215^th meeting of the American Oriental Society in Philadelphia in 2005, the Hittitist Gary Beckman from the University of Michigan read his presidential address, “The limits of credulity,” in which he sketched modern approaches to the art of writing history, presented a most useful overview or rather an extract of the flood of theoretical literature on this topic and, above all, discussed how far it is advisable and possible to trust sources and how to evaluate them. ² All this is exemplified by ancient Middle Eastern, first of all of course Hittite material. Although based on a culture with a strong written tradition, much can be learned from this article also for the thoroughly oral tradition of ancient India and early Buddhism in spite of some marked differences.

In contrast to Beckman’s after-dinner speech, the following deliberations do not concentrate on historiography, neither on modern historiography of the early Buddhist period, nor, and much less so, on an early Buddhist historiography, which is deplorably absent despite a remark made by Georg Bühler (1837-1898) to his friend, the renowned Arabist at the University of Strasbourg, Theodor Nöldeke (1836-1930), as early as 1877:

Mit Deiner Idee, dass die Inder keine historische Literatur haben, stehst Du auf einem veralteten Standpunkte. ³

This is certainly true, if we remember Kalhaṇa’s Rājataraṅgị̄̄ and the Nepalese Vaṃśāvalīs, or almost a millennium earlier in the area of Buddhism, the Dīpavaṃsa and the Mahāvaṃsa. Important as both vamsas are as sources, as underlined by the research of Wilhelm Geiger (1856-1943) and Erich Frauwallner (1898-1974), [^3] their value for the history of texts of the very early period of Buddhism is quite limited.

However, even if historiography begins too late for the period on which the following considerations are focussed, and if this Buddhist historiography contains little information on texts, we cannot conclude that this is due to a total lack of interest in history in general or in the history of texts in particular in ancient India. Early evidence proving the contrary is found in inscriptions. Already Aśoka vaguely, and it is true in a very general way, refers to the kings of yore at the beginning of his seventh pillar edict, and compares the successful propagation of his dhamma to the failure of those ancient kings to educate their peoples. In the well-known res gestae of his own reign, Khāravela looks back not only upon his own time. Khāravela also records a King Nanda, ruling either three hundred or, more likely, one hundred and three years before him, as having taken away a Jina image, which he, Khāravela, brought back to his capital.[^4] A similar memory is found in the Rudradāman inscription, where the Kṣatrapa Rudradāman commemorates in the year AD 150 that he repaired and embellished the Sudarśana tank after it was badly damaged by floods. This Sudarśana tank was originally built, as Rudradāman reminds the readers of his inscription, by Candragupta Maurya and was subsequently enlarged by Aśoka. [^5] This is indeed a rather long historical memory stretching over almost half a millennium. Interestingly, no traces of this memory are found in the text of the inscription of the Hindu ruler Skandagupta, where he records his repairs of the same tank three centuries later. ⁴

[^3] The relevant articles are quoted in O. v. Hinüber: A Handbook of Pāli Literature. Berlin 1996 (HPL), § 182, 183. [^4] Shashi Kant: The Hāthīgumphā Inscription of Khāravela and the Bhabru Edict of Aśoka. Delhi ²2000, p. 11, line 6 and p. 17, line 12 of the inscription. For tivasasata “103(?)” cf. terasavasasata “113”, line 11. [^5] Idaṃ tadākaṃ Sudarśanaṃ … mauryasya rājñah Candraguptasya rāstriyena Vaiśyena Puṣyaguptena kāritaṃ Aśokasya Mauryasya [kr]te Yavanarājena Tuṣāsphenādhiṣtāya pranālibhir alamkrtaṃ, F. Kielhorn: Junagadh rock inscription of Rudradāman; the year 72. EI 8. 1905-06, p. 36-49, esp. p. 43, 8.

With the notable exception of the Bhairāt edict by Aśoka, references to Buddhist texts are almost totally absent from early inscriptions. ⁵

However, historical memory is not only preserved in inscriptions, but in Buddhist texts as well, and, of course, this historical memory can be used with all due caution to date the texts that preserve them. This can be done only if the historical memory refers to a datable event in the political history, and this way of dating texts leads to approximations at best. Hardly ever was a text composed at the very time of the event being remembered, and never with the purpose to simply give a straightforward record of a certain event in ancient India. What we read is always an interpretation and a purposeful message of the authors to their audience. The information handed down by tradition thus depends on the intention and the will of the authors to select and to convey certain facts. This intention or will to shape the tradition being handed down is expressed both in the content and in the literary form of the texts, and both changed considerably during the transition from Vedic to early Buddhist literature.

The intention why the collection later called Tipitaka was brought together is very clearly stated in the report on the first council held at Rājagaha. For we are explicitly told why the texts were assembled and formalized:

dhammañ ca vinayañ ca saṃgāyāma pure adhammo dippati dhammo paṭibāhīyati avinayo dippati, vinayo paṭibāhīyati pure adhammavādino balavanto honti, dhammavādino dubbalā honti …, Vin II 285,4-8

“Come, let us chant dhamma and discipline before what is not dhamma shines out and dhamma is withheld, before what is not discipline shines out and discipline is withheld, before those who speak what is not-dhamma become strong, and those who speak dhamma become feeble …” (I. B. Horner).

The purpose is obviously to preserve and to defend an orthodox tradition. This must have been something quite new in ancient India at that time, a new and considerable literary challenge to be confronted not only by early Buddhists, but also by the followers of other new systems created at that time in eastern India such as Jainas or Ājīvikas. This change in paradigm, the preservation of orthodoxy and no longer the continuation of the orthopraxy of the Veda, also called for new literary forms. For this purpose veyyākaranas and dhammapariyāyas, or suttantas as they were called later, were developed, perhaps after some experiments with the prose of the Brāhmaṇas, but certainly based on this model. The model of Vedic prose is easily detected in the Sagāthavagga of the Saṃyuttanikāya, whose form closely corresponds to the stories in the Brāhmaṇas and which even continues Vedic topics such as the fight between gods and asuras. The many short episodes telling the reasons for the rules in the Mahāvagga and the Cullavagga of the Vinaya recall the structure of Brāhmaṇas, as was observed by Erich Frauwallner long ago. ⁶ Moreover the story of the Buddha sneezing reads like an answer to a story in the Jaiminīya-Brāhmaṇa, as noticed by William Dwight Whitney (1827-1894) more than a century ago. ⁷

If the individual stories neatly connect Brāhmaṇa prose and early Buddhist literature, the Buddhists went far beyond their model and composed the first really long texts in ancient India, as shown by the overall structure of the Khandhaka in the Vinaya or the Mahāparinibbānasuttanta as an individual text. Moreover, the Mahāparinibbānasuttanta is the first text ever composed in ancient India, as far as we can see, with the explicit purpose of commemorating a historical event, the death of the Buddha and thus, at the same time, the first attempt to compose a long and coherent story.

The many difficulties encountered by those who shaped or made use of this new literary form, perhaps for the first time, can be traced easily in many details in the Mahāparinibbānasuttanta. It was, e. g., very obviously a matter of great effort to keep the story on track and not to get lost in numerous narrative side alleys: Once eight reasons for an earthquake are enumerated, which fits well into the story, a cascade of totally unrelated groups of “eights” follows suit, as if in the oral period of the early Buddhist tradition hearing or mentioning the figure “eight” immediately and almost unavoidably triggered the memory of the respective paragraphs from what we now call the Añguttaranikāya. ⁸

In spite of evident difficulties like these, which do not seem to have found much attention in research, those monks who created the Buddhist sūtras had a very clear idea about the formalization of the new texts. The idea of remembering the places where the Buddha was supposed to have delivered a certain sūtra at the beginning of each individual text was certainly an innovation. This happy decision to provide the texts with a geographical frame, quite in contrast to the earlier Vedic literature, where very little is found on topography, ⁹ not only preserved a large number of place names, both villages and towns, in the Buddhist literature. In addition, the particular wording introducing these place names can tell us much about the development of the literary form of early Buddhist texts and about the historical memory of the early authors.

The opening formula of a sūtra is almost too well known to be repeated here in the standard wording:

evaṃ me sutaṃ ekaṃ samayaṃ bhagavā Sāvatthiyaṃ viharati Jetavane Anāthapiṇ̣̣ikassa ārāme

Thus I have heard. At one time the Lord stayed at Sāvatthi in the Jetavana, the park of Anāthapiṇ̣̣ika.

This very wording continues after ārāme either by … tatra kho bhagavā bhikkhū āmantesi, MN I 6,27foll. (No. 2., Sabbāsavasuttanta) with the local adverb tatra, by the developed wording … atha kho bhagavā pubbaṇhasamayaṃ …, MN I 160,27foll. (No. 26., Ariyapariyesanasuttanta) or, finally, by … tena kho pana samayena $\bar{a}yasma\bar{A}A$ nando …, MN III 189,27 (No. 132., Ānanda-Bhaddekaratta suttanta 2). Whereas the very beginning evaṃ me sutaṃ ekaṃ samayaṃ has been discussed perhaps much more often than it really deserves ever since John Brough’s (1917-1984) article published almost sixty years ago, ¹⁰ little if any attention has been paid to the much more interesting place names and to the way in which they are introduced.

Besides this well-known introduction there are others, used much more rarely and phrased in a slightly different way, such as:

evaṃ me sutaṃ ekaṃ samayaṃ bhagavā kurūsu viharati - kammāsadhammaṃ nāma kurūnaṃ nigamo - tatra kho bhagavā bhikkhū āmantesi, MN I 55,28foll. (No. 10., Satipaṭṭhānasuttanta)

… the Lord stayed in the land of the Kurus - there is a market place in the land of the Kurus named Kammāsadhamma - there the Lord addressed the monks …

The phrase “there is a market place in the land of the Kurus named Kammāsadhamma” syntactically forms a parenthesis, which does not seem to be an exciting observation. However, almost half a century ago Karl Hoffmann (1915-1996) demonstrated that this particular way of introducing place names can be traced back to Indo-Iranian syntax. ¹¹ For the Pāli sentence just quoted is exactly parallel to the Old Persian phrasing found in the Achaemenian inscriptions of Dareios (521-486) at Behistun: “when he arrived in Media - a town Māru by name is in Media - there he joined battle with the Medes.” ¹² The use of phrases like this one continues through rare examples from the Vedic language only into the earliest layers of Pāli. Comparative syntax here allows the detection of a wording that is obviously a very early predecessor to the later common formula introducing suttantas by the names of towns like Sāvatthi, Rājagaha or others.

The place names mentioned in the older place name parenthesis are quite different from these towns. Hardly any of the famous Buddhist nagaras is mentioned, but only fifteen different market places nigamas occur such as kammāsadhamma and once a Brahmin village, a brāhmaṇagāma in:

evaṃ me sutaṃ ekaṃ samayaṃ bhagavā Magadhesu viharati - pācīnato Rājagahassa Āmbasaṇ̂ā nāma brāhmaṇagāmo - tass’uttarato Vediyake pabbate indasālaguhāyaṃ. tena kho pana samayena Sakkassa …, D II 263, 2ff. (No. 21., Sakkapañhasuttanta)

Thus I have heard. At one time the Lord stayed in Magadha - to the east of Rājagaha there is a Brahmin village named Ambasaṇ̣̄a - north of it … in the Indasāla cave …

This rather exceptional formulation is due to a seemingly exact description of the location of that particular cave.

These nigamas also occur in a slightly developed wording:

evaṃ me sutaṃ ekaṃ samayaṃ bhagavā Kurūsu cārikaṃ caramāno mahatā bhikkhusaṃghena saddhiṃ yena thullakotthitaṃ nāma kurūnaṃ nigamo tad avasari. assosuṃ kho Thullakoṭthitā brāhmaṇagahapatikā …, MN II 54,25foll. (No. 82, Raṭthappālasuttanta)

… at one time the Lord walking on tour among the Kurus … where there was the market place of the Kurus called Thullakoṭthitā, there he went …

In such phrases, describing the Buddha and his followers travelling, “the Lord walked in the land of …, where there was a place called so and so there he went”, not only nigamas are mentioned but, in addition to a very few nagaras, also more frequently again brāhmaṇagāmas, “Brahmin villages,” which occur almost only in the following formula:

evaṃ me sutaṃ ekaṃ samayaṃ bhagavā Kosalesu cārikaṃ caramāno mahatā bhikkhusaṃghena saddhiṃ yena nagaravindaṃ nāma kosalānaṃ brāhmaṇagāmo tad avasari. assosuṃ kho Nagaravindeyyakā brahmaṇagahapatikā … , M III 290,26foll. (No. 150., Nagaravindeyyasuttanta)

Thus I have heard. At one time, the Lord, walking on tour in Kosala together with a large group of monks, where there is the Brahmin village of Kosala named Nagaravinda, there he went.

Interestingly, nine of the altogether fourteen Brahmin villages mentioned in the Theravāda-Tipiṭaka are situated in Kosala, four in Magadha, and only one in the Malla country. ¹³ This compares well with the evidence gathered from Vedic literature on the history and geographical distribution of the Vedic schools. As research by M. Witzel has shown, Kosala was at the eastern fringe of later Vedic literature, and the Brahmins there used to study the Kāṇva Śākhā of the Śatapathabrāhmaṇa. ¹⁴ These then could well be the very Brahmins traced in ancient Buddhist literature.

In the immediate predecessor of the later formula, which mentions a place name such as Sāvatthi together with a monastery such as the Jetavana of Anāthapindika, a wording without parenthesis and without naming a monastery occurs when a Brahmin village is referred to:

evaṃ me sutaṃ ekaṃ samayaṃ bhagavā Kosalesu viharati sālāyaṃ brāhmaṇagāme. tatra kho bhagavā bhikkhū āmantesi, S V 144,12foll. = 227,12foll.

Thus I have heard. At one time the Lord stayed in Kosala in the Brahmin village Sālā. There the Lord addressed the monks …

Interestingly, there were no vihāras in Brahmin villages, but, much more importantly, also not in the nigamas.

Taking together the very old place name parenthesis with the Brahmin villages and market places (nigamas), the missing nagaras and, above all, the missing vihāras, it is more than evident that these formulas belong to a very ancient layer of the formulation of Buddhist texts as preserved within the Theravāda-Tipitaka. Moreover, going back to the old parenthesis of place names, it is possible to trace the reason for the word order - the town preceding and the monastery following the verb viharati: Sāvatthiyaṃ viharati Jetavane … - in the sentence opening sütras, which is clearly conditioned by the stylistic prehistory of this formula.

Furthermore, the preponderance of Kosala as a location of Brahmin villages matches Vedic evidence. Consequently, we can be fairly confident of finding here really ancient village names preserved in the memory of the early Buddhists. This is confirmed by the simple fact that these early locations of the beginnings of Buddhism very soon faded into the background and were superseded by the five prominent cities enumerated at the beginning of the Mahāsudassanasuttanta, No. 17. in the Dīghanikāya: Campā, Rājagaha, Sāvatthi, Sāketa, Kosambi, Bārāṇasi, D II 169,11. First of all, Sāvatthi emerged as the prominent town, figuring at the beginning of 5 of 34 suttantas in the Dīghanikāya, but already in 67 of 152 in the Majjhimanikāya and in innumerable texts of the Saṃyutta- and Añguttaranikāyas. Still later texts were almost flooded by references to Sāvatthi, to such a degree that the Mūlasarvāstivinaya, as G. Schopen has shown, suggested inserting Śrāvastī whenever a monk forgot the place name when reciting a sūtra. ¹⁵

In the list of the five towns prominent in the Buddhist tradition, one place name is conspicuous by its absence, namely Pātaliputta, the later Maurya capital. The reason is obvious. Pātaliputta did not exist during the time of early Buddhism. Its foundation is described in a well-known paragraph at the beginning of the Mahāparinibb $\bar{a}$ nasuttanta, where the Buddha makes the following prediction during the reign of Ajātasattu: “As far as there are settlements of the Āryas, as far as there are trading routes, this will be the first city (agganagaram) Pātaliputta, a place where customs are collected (puṭabhedana),” D II 87,33-88,1. At the same time, the Buddha changes the name of the place from Pātaligāma to Pātaliputta and calls the new city a puta-bhedana, most likely in a word play with Pāṭali-putta. No mention is made of Pātaliputta as the capital (rājadhān̄̄, e.g., D II 7,29 ) of the Maurya empire, in spite of the fact that Mauryas are indeed referred to at the very end of the same text, when the relics are distributed.

The Moriyas of Pipphalivana ¹⁶ learn very late of the death of the Buddha, and when their envoy arrived at the site of the nibbbāna, all relics had been distributed and only charcoal was left, over which the Moriyas of Pipphalivana erected a stūpa. ¹⁷ This rather meagre result of the efforts made by the Moriyas to secure a share of the relics also points to a time long before the ascent of the Maurya-dynasty. Later this episode was obviously considered embarrassing and consequently cancelled by the redactors of the Sanskrit version of the Mahāparinirvāṇasūtra. ¹⁸

Of course the prophecy of the Buddha about the future of Pātaliputta caught the attention of scholars at an early date, at least since 1879, when H. Oldenberg (1854-1920) published the introduction to his edition of the Mahāvagga of the Vinaya, which contains a parallel to this part of the Mahāparinibbānasuttanta. One of the last in this line of scholars dealing with this reference seems to be K. R. Norman in his history of Pāli literature a century later in 1983. ¹⁹

Strangely, all modern scholars seem to follow the conclusion drawn by H. Oldenberg that the Mahāparinibbānasuttanta must have been composed during the time of the Maurya dynasty because the Buddha is assumed to be referring to the capital of the Maurya empire, which he is certainly not. On the contrary, the Buddha mentions very clearly a place where merchandise is exchanged, an important city certainly, but not a capital. This, however, was perhaps not too evident before B. Kölver (1938-2001) finally clarified the meaning of putabhedana, ²⁰ badly understood previously and still most strangely misunderstood as “scattering its seeds far and wide” (!?) instead of “market place” in a recent Dīghanikāya-translation published two years after B. Kölver’s article. ²¹

Now, if the Mahāparinibbānasuttanta, or more cautiously, this paragraph, was composed by the end of the fourth century, as K. R. Norman conjectures, it would be rather strange that neither the Mauryas nor Pātaliputta, as the capital of a state or at least as a city of any political importance, are referred to. This omission, on the other hand, makes a lot of sense if the text is pre-Mauryan, that is earlier than the accession of Candragupta Maurya in about 320 BC , and most likely quite a while earlier, because it can and has been demonstrated that the parallel to this part of the Mahāparinibbānasuttanta preserved in the Vinayapitaka is linguistically a slightly modernized version. ²²

Leaving aside the question of the date of this linguistic modernization of the story of the foundation of Pātaliputta, a more interesting question is whether the omission of the Mauryas can be used at all to determine a date. An answer to this question can be found only by investigating whether ancient Indian authors, given their presumed lack of historical interest, did pay enough attention to changing political situations to adapt their texts accordingly.

For this purpose it is necessary to look, at least very briefly, for references to historical events in older Buddhist literature.

Even a superficial inquiry limited only to those Buddhist texts preserved in Sanskrit yields some examples. A parallel story also concerning Pātaliputta and found in the Divyāvadāna, which is derived from the Mūlasārvāstivādavinaya, is particularly illuminating. When the Buddha makes a prediction referring to Pātaliputra in the Aśoka legend he says: ²³ varṣaśataparinirvrtasya tathāgatasya Pātaliputre nagare Aśoko nāmnā rājā bhavisyati caturbhāgacakravartī dhārmiko dharmarājā yo me śarīradhātūn vaistārikān kariṣyati, Divy 368,25foll.:

A century after the Tathāgata entered the Nirvāṇa there will be a king named Aśoka in the city of Pātaliputra, a conqueror of the four quarters of the world, a righteous king, who will spread my relics.

Here, long after the Maurya period, this dynasty is connected to Pātaliputra. A similar reference comes from the Bhaiṣajyavastu of the same Vinaya:

caturvarṣaśataparinirvrtasya mama … [Kuṣāṇavaṃśyaḥ] Kaniṣko nāma rājā bhaviṣyati. so ‘smin pradeśe (i.e. Kharjūrikā) stūpaṃ pratiṣthāpayati, tasya Kaniṣkastūpa iti saṃjñā bhaviṣyati, GM III 1,2,3 foll.

… there will be a king named Kaniṣka in the Kuṣāna family. He will establish a stūpa in this country (Kharjūrikā) which will be called KaniṣkaStūpa.

Now a third reference in an unclear fragmentary context found only recently by R. Salomon can be added. Most likely it is some Avadāna text, where it is said: [mahā]yānasamprasthito Huveṣko nā[ma rājā], “a king named Huveṣka, who has set forth on the Great Vehicle.” ²⁴

These three references demonstrate that Buddhist authors did indeed pay some attention to historical events, if only to honour rulers whose patronage was appreciated by the Buddhist samgha. The Mūlasarvāstivādavinaya, which gained its shape much later than that of the Theravādins, was evidently adapted to the politics of its time. And, as we can be fairly certain about the year 127 as the date of Kaniṣka since the research done by H. Falk, ²⁵ the redaction can be dated at the earliest to the first half of the second century.

Although it is true that the examples are few and far between, they show nevertheless that there was much more awareness of history than mostly assumed, not only in Buddhist literature. As M. Witzel has pointed out, the text of the Mahābhārata responds to political developments up to the Kuṣāna times, whereas peoples intruding into India as late as during the Gupta period are included. ²⁶ And as it seems, particularly the Buddhists always had an eye on history, as the tradition on the date of the Buddha and other indications also demonstrate.

Consequently, it may be meaningful that the Buddha is said to have predicted a brilliant future for Pātaliputra as a city of commerce and not of politics. Moreover, given the great affection and admiration for Aśoka found everywhere in Buddhist texts, it is indeed hard to conceive a date contemporaneous with, and still less likely after Aśoka. ²⁷

Given the importance of the rise of the Maurya empire even under Candragupta, who is better known for his inclination towards Jainism, one might conjecture that the latest date for the composition of the Mahāparinibbānasuttanta, at least for this part of it, is around 350 to 320 BC .

If this is not altogether too far off the mark, and if it is remembered that the date of the nirvāna can be assumed to be about 380 BC , this dating of the text certainly has also some consequences for the assessment of the content. For a distance in time of roughly thirty to sixty years from the event recorded to the text conceived allows for a fair chance to trace true historical memory.

Of course it is not intended to turn the Mahāparinibbānasuttanta as a whole into a full historical record now and to read it as a historical account instead of hagiography, which it is. However, while many references to Indian history found in the Tipitaka remain doubtful, one point should be beyond any reasonable doubt: the death of the Buddha occurred at some point in history and at a certain place. A second point is no less important. The death of the Buddha as the founder of the Buddhist samgha was an event of huge consequences for the then contemporary Buddhists, and an event witnessed by many monks and deeply penetrating into the collective memory of all Buddhists of the time. In contrast, the bodhi, certainly of prime importance for Buddhism and Buddhists, was not witnessed by any future monk and no collective memory could spring up from this event.

Although there were witnesses present at the nirvāna, mythological features abound in the description of the death of the Buddha because at that time no religious person could possibly die without accompanying miracles, and at the time after the Buddha’s death, no text describing the career of the founder of any religion could have possibly found acceptance without miraculous features.

Between fact and fiction are the earthquakes at the moment when the Buddha gives up his vital force and again at the moment of the nirvāna. ²⁸ Although geophysics does not necessarily rule out that they occurred, in all likelihood they did not; nobody would take them to be historical events. And the divine flowers showering on the deceased Buddha in such a quantity that they filled the whole small town of Kusinārā knee deep with heavenly mandārava flowers ²⁹ is just as evidently mythology as it is indispensable in any record of an event such as the Buddha’s nirvāna.

So, even if there had been the will or at least the intention of the early Buddhist authors of the Mahāparinibbānasuttanta to portray a more or less exact historical record of the nirvāna, miraculous and supernatural events were impossible to avoid in an environment in which the practice of all sorts of yogic achievements was common- place. Miracles simply were part of the world-view of the authors of the Mahāparinibbānasuttanta.

Consequently, if there are details that might be called historical, they are necessarily buried in what we call mythology. However, we can try to unearth a bit of history, as Gary Beckman has suggested in his lecture, by replacing the rather credulous question “Why might it be false?” by the more sceptical one: “Why should it be true?” if we venture to attempt to separate fact from fiction. ³⁰

For instance the route taken by the Buddha to Kusinārā might be such a detail based on remembered history, if the names of the many otherwise unknown villages are recalled. The same might be said perhaps in the case of the earliest record of what was possibly an epidemic at Nādikā, where twelve persons recently deceased are enumerated by name, among them strange names such as Sāḷha, Nikaṭa or Katissabha not mentioned in any other source. ³¹ The name of the very last monk ordained by the Buddha, Subhadda, ³² may be historical memory and, of course, the absence of nuns during the nirvāna. ³³

In contrast, another famous paragraph cannot be historical as it stands. These are the last words of the Buddha: handa dāni bhikkhave āmantayāmi vo: vayadhammā saṃkhārā appamādena sampādethā ti. ayaṃ tathāgatassa pacchimā vācā, D II 156,1foll.

Now, monks, I address you: Decay is inherent in all component things. Work out your salvation without indolence.

We can be sure that this is not exact historical memory, correct in spirit and content at best, but certainly not in wording, because the Buddha did not speak Pāli.

How ever this may have been in detail remains, to a considerable extent, a matter of conjecture, because our sources never allow us to go beyond more or less likely or probable conclusions about the roots of the texts that reach far back into the period of early Buddhism. On the other hand, determining the date of the closure of a text remains a still thornier problem. However, we can be sure to find quite a lot of very old material in the Theravāda tradition and we can, therefore, hope with a little confidence that the wonderfully solemn and unique conclusion of the Mahāparinibbānasuttanta may not be altogether wrong: ³⁴

evam etam bhütapubbam

“Thus it was in the days of yore.”

Abbreviations

ARIRIAB	Annual Report of The International Research Institute for Advanced Buddhology at Soka University
BSOAS	Bulletin of the School of Oriental and African Studies
D	Dīghanikāya
Divy	Divyāvadāna
EI	Epigraphia Indica
GM	Gilgit Manuscripts. Ed. by N. Dutt. Srinagar, 1939-1959.
HPL	O. v. Hinüber: A Handbook of Pāli Literature. Berlin 1996.
IIJ	Indo-Iranian Journal
JAOS	Journal of the American Oriental Society
JPTS	Journal of the Pali Text Society
MSS	Münchener Studien zur Sprachwissenschaft
S	Saṃyuttanikāya
SASt	South Asian Studies
Ud	Udāna
Vin	Vinayapitaka
ZDMG	Zeitschrift der Deutschen Morgenländischen Gesellschaft

Footnotes

1. This text was read as the Presidential Address on 23 June 2008 during the XV^th IABS conference held at Atlanta (Georgia) from 23 to 28 June 2008. The oral form of the presentation has been largely preserved. An enlarged and more detailed version dealing with the early history of Theraväda texts is under preparation. ↩

2. JAOS 125. 2005, 343-352. ↩

3. “Your idea that Indians do not possess literature on history is an outdated point of view,” quoted from Julius Jolly: Georg Bühler. Grundriss der Indo-Arischen Philologie und Altertumskunde. 1. Band, 1. Heft A. Strassburg 1899, p. 13. ↩

4. This inscription is published in Corpus Inscriptionum Indicarum III: Inscriptions of the early Gupta kings revised by Devadatta Ramakrishna Bhandarkar. Delhi 1981, no. 28, 296-305, cf. also O. v. Hinüber: Les documents épigraphiques indiens: Difficultés de leur interprétation - Examples concernant l’irrigation. Académie des Inscriptions et BellesLettres. Comptes Rendus des Séances de l’Année 2004. Avril-Juin. Paris 2004 [2006], p. 989-1011, esp. p. 989 foll. ↩

5. Only very general references such as trepitaka or vinayadhara are found occasionally in Mathurā or Amarāvatī. ↩

6. HPL § 32. ↩

7. Henry C. Warren: On superstitious customs connected with sneezing, JAOS 13. 1889, p. XVII-XX, esp. p. XX, where W. D. Whitney refers to Jaiminīya-Brāhmaṇa II 155 . ↩

8. HPL § 60. ↩

9. On this point, see K. Hoffmann, as note 13 below, p. 122. ↩

10. Thus have I heard …, BSOAS 13. 1950, p. 416-426, reprinted in J. Brough: Collected Papers. London 1996, p. 63-73; for further references see HPL § 53 and add: M. Tatz: Thus have I heard: At one time, IIJ 40. 1997, p. 117foll.; B. Galloway: A reply to Professor Mark Tatz, IIJ 40. 1997, p. 367-371; F. Tola, C. Dragonetti: Ekaṃ samayam, IIJ 42. 1999, p. 53-55. ↩

11. K. Hoffmann: Die Ortsnamen-Parenthese im Altpersischen und Vedischen (ZDMG 110. 1960), in: Aufsätze zur Indoiranistik ed. by J. Narten. Volume I. Wiesbaden 1975, p. 120-129, on Pāli S. 128f., cf. also G. E. Dunkel: Naming parenthesis in Indo-Iranian and Indo-European, MSS 41.1982, p. 11-21. ↩

12. R. Kent: Old Persian. Grammar, Text, Lexicon. New Haven 1953, p. 121: ya@ Mādam parārasa - Māruš nāma vardanam Mādaiy - avadā hamaranam akunauš, DB II, line 22 foll. § 22 . ↩

13. This is, at the same time, the only reference to the word brāhmaṇagāma used in the Vinaya in the definition of majjhimadesa at Vin I 197,27, but not in the usual formula as in Ud 78,5, the second of the only two references to the brāhmaṇagāma Thūṇa in the Thera-vāda-Tipiṭaka, cf. also IIJ 45.2002, p. 79. ↩

14. M. Witzel: The Development of the Vedic Canon and its Schools: The Social and Political Milieu, in: Inside the Texts, Beyond the Texts ed. by M. Witzel. Harvard Oriental Series. Opera Minora Vol. 2. Cambridge/Mass. 1997, p. 257-345, esp. p. 313foll. § 5.2 ↩

15. G. Schopen: If You Can’t Remember, How to Make It Up: Some Monastic Rules for Redacting Canonical Texts (1997), in: Buddhist Monks and Business Matter. Still More Papers on Monastic Buddhism in India. Honolulu 2004, p. 395-407. In later texts, the place names did not really matter, because there was no longer any immediate reference to an old tradition anyway. ↩

16. Although it is tempting to compare Pipphalivana to modern Piprāhvā, the difficulties involved are considerable. ↩

17. D II 166,21foll.; 167,17foll. ↩

18. A survey of the relevant material can be found in A. Bareau (1921-1993): Recherches sur la biographie du Buddha dans les Sūtrapiṭaka et les Vinayapiṭaka anciens II, 2: Les derniers mois, le parinirvāna et les funérailles. Publications de l’École Française d’Extrême-Orient 77. Paris 1971, p. 303. However, Bareau’s assumption that the Moriyas of Pipphalivana were only introduced at a late date and only by the Theravādins obviously turns the development of the text upside down. For, in addition to the reason given above, it is easy to see that the obscure pipphaliº was replaced by the much more common pippala in: pippalāyanah māṇavah pippalavatyām añgārastūpaṃ pratiṣthāpayati, Mahāparinirvānasūtra 51.21. On the other hand, going back from pippala to pipphali does not make any sense at all. ↩

19. The Vinaya Piṭakaṃ ed. by H. Oldenberg. London 1879, p. XXXVII; K. R. Norman: Pāli Literature including the canonical literature in Prakrit and Sanskrit of all the Hīnayāna Schools of Buddhism. A History of Indian Literature VII,2. Wiesbaden 1983 p. 38. ↩

20. B. Kölver: Kauṭalyas Stadt als Handelszentrum: der Terminus putabhedanaZDMG 135. 1985, p. 299-311. ↩

21. The Long Discourses of the Buddha. A Translation of the Dīgha Nikāya by M. Walshe. Boston 1987, p. 238. ↩

22. This can be seen by comparing the place name Nādikā, which is preserved in its original eastern form only in the Dīghanikāya (D II 91,15), but changed to Nātikā in the Vinayapitaka (Vin I 232,31); the old historical gen. pl. rañhaṃ (D II 87,2 etc.) is regularly modernized as rājūnaṃ (Vin I 228,30 etc.), cf. O. v. Hinüber: Der Beginn der Schrift und frühe Schriftlichkeit in Indien. Akademie der Wissenschaften und der Literatur, Mainz. Abhandlungen der geistes- und sozialwissenschaftlichen Klasse, Jg 1989, Nr. 11, chapter X, p. 46foll. ↩

23. Cf. J. Strong: The Legend of King Aśoka. A Study and Translation of the Aśokāvadāna. Princeton 1983, p. 61. ↩

24. R. Salomon: A fragment of a Collection of Buddhist Legends, with Reference to King Huviṣka as a Follower of the Mahāyāna, in: Jens Braarvig [Ed.]: Buddhist Manuscripts Volume II. Manuscripts in the Schøyen Collection III. Oslo 2002, p. 255-267. There seems to be even a portrait of Huviṣka recently discovered on a relief from Gandhāra, cf. F. Grenet: Note additionnelle on B. Marshak: Une peinture kouchane sur toile, Académie des Inscriptions et Belles Lettres. Comptes rendus des séances de l’année 2006. 2006 [2008], p. 947-954 + 955-963, esp. p. 957. ↩

25. H. Falk: The yuga of Spujiddhvaja and the era of the Kuṣānas. Silk Road Art and Archaeology 7. 2001, p. 121-136. ↩

26. M. Witzel: The Vedas and the Epics: Some Comparative Notes on Persons, Lineages, Geography, and Grammar, in: Epics, Khilas, and Purānas: Continuities and Ruptures. Proceedings of the Third Dubrovnik International Conference on the Sanskrit Epics and Purānas. September 2002 ed. by P. Koskikallio. Zagreb 2005, p. 21-80, esp. p. 63. ↩

27. An extremely late date for the composition of the Mahāparinibbānasuttanta during “the 2^nd or 3^rd century of the Common Era” (!) was recently assumed without any reason given by C. Woodford Schmidt: Aristocratic Devotees in Early Buddhist Art from Greater Gandhāra, SASt 21. 2005, p. 25-45, esp. p. 25. This date, which is perhaps due to a misreading of “AD” for “BC” (??), can be safely and confidently ruled out. ↩

28. āyusaṃkhāraṃ ossaji, D II 106,22 and D II 156,36. ↩

29. D II 160,31. ↩

30. JAOS 125, 2005, p. 349. ↩

31. D II 91,26-92,11. ↩

32. so bhagavato pacchimo sakkhisāvako ahosi, D II 153,11. ↩

33. O. v. Hinüber: The Foundation of the Bhikkhunīsaṃgha. A contribution to the earliest history of Buddhism. Annual Report of the International Research Institute for Advanced Buddhology at Soka University for the Academic Year 2007 (ARIRIAB 11). Tokyo 2008, p. 2-29, esp. p. 22. - One of the most likely candidates for historical memory is perhaps the famous name of the last meal of the Buddha. In all likeliness the obscure sūkaramaddava (D II 127,5) is the name of a local dish, which was piously preserved, while the true meaning was soon forgotten: O. v. Hinüber: The Cause of the Buddha’s Death: The last Meal of the Buddha. A Note on sūkaramaddava. JPTS 26. 2000, p. 105-117. ↩

34. D II 167,20. ↩