E Bruce Brooks
University of Massachusetts at Amherst
The Date of the Buddha
Ethnogenesis IV Conference, Harvard, 12 May 2002

This paper was was delivered in two minutes, as a 12-sentence summary, together with a two-page handout. A fuller version of the handout is given here..

Textual Argument

1. Preface and Abstract. It is agreed that the date of the death of the Buddha is important for early Indian chronology, and also that the former wide consensus about a c0480 value for that date no longer exists; many scholars now incline toward a date a century or more later. I have analyzed (in translation) one important Pâli text bearing on this question, and have also extensively studied (in the original) the contemporary Chinese text record as a whole. This double approach has suggested that, however schematic it may appear, a date in the neighborhood of the the previously accepted c0480 is after all correct. On the way to this conclusion, I detach from the Buddha's death two things which, in recent discussions, have been strongly linked with it. These are:

• (1) The degree of Ganga Urbanization (and/or the degree of development of Indian coinage) implied by the Buddhist suttas collectively. These suttas are nearly all retrospective (their purpose is precisely to claim the authority of the Buddha for rules made much later) and thus anachronistic. They often reflect a condition of society and economy later than that obtaining during the lifetime of the Buddha. Crediting them as coming from the time of the Buddha merely confuses the evidence.
• (2) The beginnings of Buddhist monasticism and the compilation of the Vinaya rules. The latter are conventionally attributed to Buddha himself, with his supposed direct disciple Upâli as the chief transmitter of those rules to posterity. Instead, I propose that Upâli is a doubtless important but later figure, and that the tradition linking him with the lifetime of the Buddha is a standard device of retrospective certification within a growing tradition, and deserves no credit as literal history.

2. The Mahâ-Parinibbâna Sutta (MpnS). This text is unique in the Pâli canon as a description of the last days of the Buddha; there are no competing accounts. The text is obviously composite. Other Pâli texts do not have this manifest composite character, but appear to be unitary works, whatever the probable date of their actual composition. If the MpnS was seen by later Buddhists as having unique authority to confer certification, the composite character of the text is readily explained: a core text was updated repeatedly, over time, to introduce into it more developed versions of early ideas, or to mention later practices on which it was desired to have the Buddha express an opinion, or to introduce, as early, institutional innovations for which a precedent was wanted. The composite character of the text implies its use as a certification device, and is best explained if the MpnS core was in fact of early date.

3. Types of Later Material. Many suggestions have been made as to how to distinguish early from late ideas within the Buddhist tradition. Impressions of centrality in mature Buddhist doctrine, or opinions about the beauty of certain passages, have no value as evidence of age. On the contrary, they would tend to point precisely to developed rather than to primitive features within the tradition. I suggest the following principles of decision as generally relevant to the distinction of layers in MpnS:

• (1) As between simple and elaborate versions of the same thing, the simpler is likely to be earlier. Thus, the Fourfold Truth of MpnS 2:2 probably precedes the devotional Eightfold Way of 2:9, and the 500 disciples of 6:9 are probably a later stylized exaggeration of the unstated but seemingly moderate number of disciples implied at most points in the MpnS narrative.
• (2) Display of magic powers, like the miraculous crossing of a river in 1:33 or the knowledge of the cosmic fate of various deceased persons in 2:7, is probably late.
• (3) Deities or supernatural beings, like those who mourn the Buddha in 6:15-16, are probably late.
• (4) Predictions of future events, such as the greatness of Pâtaliputta in 1:28, are probably late, and indeed, are probably close in time to the events which are thus predicted.
• (5) Explicit provision for the needs of more developed Buddhism fall in the same category as the preceding. Clear examples are the warning about future conflicting traditions in 4:7-11 and the prescriptions about sites of pilgrimage in 5:16-22.
• (6) Against the opinion of Przluski, it would seem that the verse portions of the text are not old, but on the contrary, have a decorative and elaborative function, and are thus later in date than the prose passages preceding them. A similar situation exists with the Arthashastra, and (as Olivelle Dharma xxix has noted) also within the series of Sanskrit Dharma Sûtras, in which later texts make increasingly frequent use of this type of decorative verse.

4. Reconstruction of the MpnS. A reconstruction of the text along these lines, and not involving internal inconsistencies, is possible, but the result cannot be presented in full on this occasion; it is currently underway. Some features of it may be mentioned here.

Layers in MpnS

5. The Core MpnS. This turns out to be a circumstantial narrative, without any magical elements, written entirely in prose, and of no great length. It recounts the events leading up to the death of the Buddha, and concludes with the death itself. This portion probably dates from a time shortly after the Buddha's death, before the Buddhist religion as such had been organized and elaborated. Exactly what that date was cannot be established directly from the contents of the core text.

6. The Framing Narrative. Two segments seem to have been designed as a final contexting frame for the MpnS as a whole; they need not have been added at the same time. They are:

• (1) The tale of the impregnable Vaggians at the beginning the work, who cannot be attacked with success by the Brahmins of Magadha because they display certain civic virtues, which implies an appeal for government patronage by offering advice to government, and
• (2) The argument between several governments over the Buddha's relics at the end, which may suggest the well-documented veneration paid by such later Magadha rulers as Asoka to scenes and objects associated with the Buddha. The implied historical context ranges from early to middle Maurya period, or late 04c to mid 03c.

7. Other Late Features. The use of numbered sets in some of the relatively late materials probably comes out of the same literary milieu as does the similar feature in some classical Sanskrit texts. That device appears for the first time in Chinese texts of the very late 04c. If the Indian and Chinese literary phenomena are connected, as the Chinese evidence suggests, it seems likely that the Chinese version is derivative, so that we should expect the first appearances of numbered sets in Sanskrit texts to be not later than, say, c0320. The classical Sanskrit texts are not directly datable, but these numbers are largely compatible with some relative chronologies which have been suggested by previous scholars.

8. Ânanda. One feature of the MpnS addenda is the narrative prominence of Ânanda, who is less dominant in the core MpnS narrative. Later tradition associates Ânanda with authority for transmission of doctrine, and gives Upâli similar authority for the vinaya or rules of monastic discipline. We would seem to have in these early (but not original) MpnS layers the beginning of the Ânanda emphasis. Of Upâli there is no trace in the MpnS core, and the MpnS makes no detailed or explicit provision for the later rise of Buddhist monasticism, under the authority of Upâli or anyone else. Its concern is with doctrine, not with residential arrangements.

9. Vinaya. There is a traditional sequence of Vinaya masters. The interpretation of Sarao (which is close to that of Gombrich) would give Upâli a life span of 0441-0367, and date his becomoing the Vinaya Master (Vinaya Pamokkha) to 0397, in the early 04c. This does not mean that the Buddha must be dated to the same period. Instead, it gives one version of the point at which the developing Buddhist religion had shifted from a mendicant to a full monastic mode, and had acquired enough experience of that new mode that a need arose to codify that experience and to supply an authority for it. Considerable time may well have elapsed between the death of the Buddha (who, let it be noted, is still portrayed in this material as the leader of an exclusively mendicant group) and this later development.

10. Date of the Earliest Expansions. What seem to be the earliest expansions of MpnS take the core narrative as given, but imply a concern that posterity will take a negative view of that narrative. Thus,

• (1) the original reason given for the Buddha's death (out of a total of three offered by various layers of MpnS) is his eating tainted pork presented by one of his hosts. In 4:57, the Buddha advises Ânanda how to deal with any feelings of remorse arising in the donor, or any blame which others are inclined to attach to him.
• (2) The MpnS core has Buddha dying in insignificant Kusinârâ, and in 5:41 Ânanda complains that he should instead die in some grander and more appropriate place. The Buddha replies in 5:42-43 that Kusinârâ had once been the capital of a great empire, so that it is actually a suitable location for this admittedly momentous event.

The concern of Ânanda is mirrored in the concern of Confucius's disciples in Analects (LY) 9:12, that Confucius is dying without having achieved high rank; they disguise themselves as the retainers of a mighty official. In a lucid interval in his final sickness, Confucius rebukes them for this imposture, remarking that at any rate he is not dying "by the roadside," as the Buddha had in fact done. The date of this passage, within the gradual accretion history of the Analects, can be fixed rather closely to c0405. If we see the Analects passage as an actual echo of the MpnS passage or its orally known equivalent, then the early expansion layers of MpnS itself must be at least somewhat earlier than c0405.

11. Date of the Core MpnS Narrative. Some details which may plausibly be ascribed to the core MpnS narrative also have Chinese counterparts in the Analects text, whose chronology has been clarified by my previous researches. These counterparts are found in the slightly earlier chapter LY 7 (c0450). This chapter is strangely at variance with the rest of the Analects, in ways which invite a hypothesis of an unusual influence, or an unusual compiler mindset, at work in LY 7, as nowhere else in the Analects, Samples:

• Confucius accepts humble pupils (LY 7:7, 7:29)
• Confucius leads a seemingly itinerant life as an impoverished learner and teacher of virtue (LY 7:2, 7:16, 7:17, 7:19, 7:20, 7:22)
• Confucius hopes to meet a sage, while discounting the idea that he is himself a sage (LY 7:26a, 7:34). This is the first, and for a long time the last, appearance of the "sage" concept in the Analects text.

If these correspondences are valid, then some notion of the core MpnS, or the tradition which it records, would seem to have reached the Chinese eastern states by the middle of the 05c, and by the previous argument, a slightly later version of that text or tradition would seem to have reached them by the end of the 05c. Without here going into further details, it would seem that the death of the Buddha would best be located somewhere in the first half of the 05c.

12. Chronological Outcome. We would then have something like the following scenario for the beginning and development of Buddhism:

• (1) The death of the Buddha and the core MpnS narrative (first half of the 05c).
• (2) The earliest expansions of that narrative, implying acceptance of Buddhism in urban centers but not yet a shift of Buddhism itself to monastic mode (second half of the 05c).
• (3) The onset of semi-monastic Buddhism, where the monks lived in separate, privately provided, nearby residences, and convened every two weeks for prayers and confessions. Those meetings were governed by their own text, the Patimokkha. Though now within the Vinaya text category, the Patimokkha is quite distinct from the Vinaya, for which see next.
• (4) The rise of full monasticism in the first half of the 04c, with Upâli featured as the prime authority for the Vinaya rules. A claim made then or later wrongly locates him in the time of the historical Buddha. He was instead a later personage; doubtless a leading figure in the Second Monastic Period.
• (5) Official state patronage under the Maurya ruler Asoka in the late middle 03c.

This timetable allows reasonable time for increasing popular acceptance of Buddhism, as well as increasing Ganga urbanization and the growing wealth of cities, to have an effect on the organization of the religion itself.

13. The Date of the Buddha. The traditional figure of 218 years between the death of the Buddha and the conversion of Asoka (with his resulting patronage and reform of Buddhism) is best taken as conventional. It amounts to the claim that between the death of the Buddha and the conversion of Asoka, there intervened

• 1. A first major event occurring after 100 years, this being the standard conventional interval of prediction in the later Buddhist literature
• 2. A second major event, occurring after another 100 years, this event being the rise of the ruler patron, or the coronation of Asoka.
• 3. A third event, occurring after a further 18 years. We may note that according to his own inscriptions, it was in the 18th year of his reign that Asoka was persuaded to accept Buddhism.

An alternative interval of 256 years, as Narain has argued, is based on counting backward from a later date in Asoka's reign, namely, the year of his abdication to pursue a life of virtue. These numbers, if literally interpreted, would give a Buddha death date of 0483. The notion that major events occur at 100-year intervals is overly schematic, but need not be integral to that chronology. Like the general development of Buddhism implied by these findings, it too is historically plausible.

If one prefers not to take these numbers as exact, we arrive at a range later than a year, for the death of the Buddha. The probable center of that range nevertheless corresponds generally to the date preferred by the majority of earlier investigators. We may say in conclusion that it appears, from this argument from both Chinese and Pâli sources, that two of Jaspers' "Axial Age" figures, the Buddha and Confucius (0549-0479), were not only contemporaries, but that the early traditions about them were in contact, and that the Buddhist tradition affected some details of the evolving Confucian tradition.

Notes

Important. The date of the Buddha's decease was long regarded as the earliest firmly known date in Indian history. Much depends on it: the chronology of Indian kings, the correlation of Buddhism with the archaeological record, and the pace of development in Buddhism itself. [Return]

Scholars. Skepticism about the previously accepted date of the Buddha's death (which ranged from 0486 to 0483), and preference for a later date, go back to Rhys Davids (SBE v11, xlvi). It has gained a wider acceptance with the work of Heinz Bechert and his associates (Dating, v1 1991). [Return]

Chinese. For a summary of this research, see in general the Home Page of the present site, and references available there. [Return]

Certification. The ascription of later developments to the founder of a movement, and the creation of intermediary figures who are then claimed to have a direct link to the founder, are common devices in the internalizing and rationalizing of later developments. Origin myths in all cultures have a similar dynamic. One parallel Chinese case is the supposed disciple Shang Jyw, who never appears in the Analects of Confucius (agreed to be the source closest to early Confucian tradition), and who functions in a later disciple list solely as the transmitter of Confucius's knowledge of the Yi or Divination Classic. The Analects itself, in its earlier layers, contains no direct or indirect evidence that Confucius knew or used the Yi, and as late as the 03c, the rival Confucians Sywndz and Mencius both shunned the Yi altogether. The Yi had begun to acquire commentaries already in the 03c, and during Han it became part of the recognized canon. Shang Jyw was invented at about this time to give a Confucian pedigree for this fundamentally non-Confucian work. [Return]

Text. I use the numbering of sections in the SBE version of the Rhys Davids translation, but have ignored the special conventions imposed on the spelling of Pâli terms by the SBE editors. [Return]

Composite. This was argued extensively by Rhys Davids (SBE v11 xii-xviii). Pande Studies (4ed 1995) 98-106 calls the text "a veritable mosaic," and summarizes various opinions about later material within MpnS. Von Hinüber Pâli Literature (1996), while acknowledging the existence of eg Bareau Composition (1976), notes that the text "has never really been investigated." The results of our investigation cannot be presented here, but it is hoped to publish them presently. [Return]

Magadha. The "Brahmin minister of a Magadha ruler" inevitably suggests Kautilya, minister of the first Maurya ruler Chandragupta, who is often credited as the architect of the Mauryan Empire. That he was a Brahmin is shown by the emphasis on Vedic traditional learning in the sayings which are attributed to him in the Arthashastra. Our reconstruction of the original Kautilya maxims is no longer available on this site, but it is hoped to publish it presently. [Return]

Analects. For the segmentation and chronology of this text, see Brooks and Brooks, The Original Analects (Columbia 1998). [Return]

Narain. See "The Date of Gotama Buddha's Parinirvâna," in Bechert Dating v1 (1992) 185-195. [Return]

Postscript. I am grateful to the Ethnogenesis IV audience for helpful questions and comments directly following the presentation, and in subsequent informal conversation during the remainder of the conference. The text above has been slightly expanded in the light of those questions and comments. Further comments from viewers of this page will be appreciated.

All lectures and abstracts posted on this site are Copyright © by their authors.

 

2 Sept 2000 / Contact The Project / Exit to Conferences Page