By A. Berriedale Keith
Bulletin of the School of Oriental Studies
1932.09, pp. 859-866
p. 859
In a very interesting article, (l) Professor
Jacobi has arrived at the conclusion that, contrary
to the Buddhist tradition, we must hold that Mahavira
outlived the Buddha, probably by some seven years. In
point of fact, of course, it may seem of very little
consequence whether we accept this view or that of
Buddhist tradition, but the issue involves a very
important question affecting the value of our
authorities, and on this point it seems to me clear
that the position adopted by Professor Jacobi
involves serious difficulties.
Professor Jacobi treats as the assured
foundations for his investigations the dates of the
Nirvanas of the Buddha and of Mahavira, as 484 and
477 B.C. But it must be admitted that both these
dates rest on very unsatisfactory and late evidence.
The question of the date of the Buddha has been set
out, with his usual acumen and precision, recently by
Professor de La Vallee Poussin,(2) and he has shown
how utterly uncertain is the date 483 or 484 B.C. for
the Nirvana. From a very different point of view the
late Professor Rhys Davids confessed(3) that the date
was purely conjectural. We may readily believe that
the Buddha died sometime in the fifth century B.C.,
but to lay any stress on the exact date is completely
impossible with the evidence available. What is
perfectly clear is that knowledge of the early period
of Buddhism was imperfect,(4) and the same remark
applies even more strikingly to the traditions of
Jainism. In the case of Mahavira the earlier
tradition--of uncertain date--is emphatic in allowing
470 years between his Nirvana and the beginning of
the Vikrama era, which places the date in 528 or 527
B.C. The later tradition, given in Hemacandra's
Paricistaparvan, viii, 339, and somewhat earlier in
Bhadrecvara's Kahavali, ascribes 155 years as the
period between the death of Mahavira and
Candragupta's accession to the throne of Magadha,
which gives 477 B.C. as the probable date of
Mahavira's death. Here again we are on utterly
uncertain ground. We are obliged to treat the earlier
Jain tradition as of minimal value
______________________
1. SBA. 1930, pp. 557-68.
2. Indo-europeens et Indo-iraniens, pp.238--48; L'Inde
aux Temp des Mauryas, p.50
3. CHI. i, pp. 171, 172.
4. Keith, Buddhist Philosophy, chap. i.
p. 860
and there seems every ground for so doing; but the
tradition accepted by Hemacandra rests equally on no
assured foundation. The only possible conclusion
regarding it is that it cannot be trusted to be
accurate within a few years, and it seems wholly
impossible to base on two dates so acquired the view
that we must believe that the Buddha. predeceased
Mahavira. Nor is it irrelevant to note that Professor
Jacobi(1) himself has adopted slightly different
dates, namely 477 and 467 B.C. in other
contributions; but what is more important is that the
Jaina tradition contains one certain error which, if
rectified, destroys the value of its testimony for
477 B.C. By that tradition, apparently accepted by
Hemacandra as well as the rest of Jain opinion, the
date of the accession of Candragupta is placed at 255
years before the Vikrama era, i.e., in 313 or 312
B.C. This date is obviously too late; if we take 322,
as does Professor Jacobi, as a probable date(2) then
we must admit a clear error in the Jain tradition of
about ten years in respect of this interval;
admitting a like error regarding the earlier
interval, that between the accession of Candragupta
and the death of Mahavira, we would arrive at 487
B.C. for the death of the latter, and this would
place that event before the death of the Buddha, and
confirm the Buddhist tradition. This shows clearly
with what inadequate data we have to reckon, and
leaves the conviction that the supposed dates of the
deaths of the two great teachers are of too uncertain
character to afford_any conclusion as to the priority
of these events.
On the other hand, we have the clear and distinct
tradition of the Buddhist Canon which asserts that
Mahavira, died before the Buddha and does so, not
incidentally, but as giving rise to allocutions of
the Master regarding the tenets of his teaching,
recorded in the Pasadika Suttanta of the Digha Nikaya
and the Samagama Suttanta of the Majjhima Nikaya, and
of Sariputta, at the Master's bidding, in the this
definite tradition recorded in canonical texts?(3)
That these texts belong to the Period immediately
after the death of the Buddha. I confess I do not
believe, but they far outrank in age the tradi-
tions of the dates of the deaths of the Buddha and
Mahavira, and give us
______________________
1. Introduction to Kalpa Sutra, p. 9; Introduction to
Paricistaparvan, p.6.
2. In CHI. i, pp. 471-3, 321 is suggested as plausible.
For other dates see L. de La Vallee Poussin, L'Inde
aux Temps des mauryas, pp. 51, 52.
3. The Upali Suttanta clearly asserts an illness, if
not the death, of Mahavira; Chalmers SBB. v, p. 278,
n.2.
p. 861
authentic views of the belief held in Buddhist
circles at some period considerably before the
Christian era. If we are to discredit their account,
we must be prepared to accept the consequences, which
involve acceptance of a scepticism as to the value of
the Buddhist and Indian traditions in general, which
is quite inconsistent with the faith placed by
Professor Jacobi in the tradition as to the dates of
the Nirvanasa or his acceptance of the view that the
Kautill;ya Arthacastra is the work of a minister of
the Emperor Candragupta. If we are on any logical
ground to discredit the Buddhist tradition, very
strong arguments are necessary, and those adduced
seem quite inadequate.
It is contended by Professor Jacobi that the
evidence of the three Suttantas is destroyed by the
fact that, while all agree in making the occasion of
Mahavira's death and consequent unrest in his
community the cause of the dissertations on the
Buddhist tenets, the divergence of the form of
argument in the three Suttas shows that cannot
represent what the Buddha actually said. This may, of
course, be conceded at once by those who believe(1)
that we have little or nothing of the ipsissima verba
of the Master. The view which seems natural is that
the Buddhists believed that there was difficulty in
the Jain community on the death of their leader, and
that this took place before the Buddha's death,
eliciting from him comments, which were probably not
preserved in any authentic form, leaving it open for
the composers of the Suttantas to present the
teachings each in his own way. The essential point is
really that different Buddhist authors held the same
tradition, which shows that it was a belief handed
down by tradition and widely spread in Buddhist
circles.
In the second place, Professor Jacobi argues that
the account in these Suttantas is contradicted by the
account in the Mahaparinibbana Suttanta, the oldest
account of the proceedings of the Buddha's last year
up to his Nirvana. This text does not refer to any
special anxiety of the Buddha as to the fate of his
community after his death as having been elicited by
the report of the dissensions in the community of
Mahavira, whence it is deduced that this report is a
later invention. But this reasoning rests on several
unproved assumptions. (1) That the Mahaparinibbana
Suttanta is older than the other three Suttantas is
assumed without any arguments being adduced, and its
age certainly is far from obvious. On the contrary,
______________________
1. See Winternitz, Geschichte der Indischen Litteratur,
ii, pp. 360 f.
p. 862
it appears to be a very sophisticated and worked up
account of the last days of the Buddha, and in fact
it is not open to Professor Jacobi to contend for its
early date. He himself shortly afterwards (p.562)
refers to the account given in that text of the plans
of Ajatacatru for the subjection of the Vrjis, and
points out that the undertaking was one demanding
careful planning. He adds: "Uber die von ihm
getroffenen Massnahmen enthalt das M.P.S. Angaben,
die aber in viel spaterer Zeit entstanden und darum
so gut wie wertlos sind." Very probably Professor
Jacobi's view of the statements of the Suttanta is
correct; but it is quite impossible to hold this view
of it, and then to ask us to accept the silence of
the Suttanta as entitling us to negate the evidence
of three Suttantas, two of which at least may well be
older than the Mahaparinibbana Suttanta. (2)
Moreover, the argument is essentially one ex silentio
and there is no form of contention more dangerous. It
would be necessary, in order to give it weight, to
show that the omission of the episode of the Buddha's
views on hearing of Mahavira's death is inexplicable,
if its occurrence were widely believed in Buddhist
circles. No such proof, however, is possible.
Professor Jacobi's view appears to be that the
episode of the hearing of the death of Mahavira took
place during the last journey of the Buddha en route
to Kusinara, and that, therefore, any full account of
his last days must necessarily include the episode in
question. If this view were sound, there might be
something to say for his contention, though the
argument would be far from conclusive. But there
seems no ground whatever to assume that the Buddhists
thought that the news of Mahavira's death came to the
Buddha just before his own
The Samagama Suttanta has nothing to suggest such
a conclusion. On the contrary the Buddha is at
Samagama when he hears of the death of Mahavira at
Pava,(l) and equally in the other two Suttantas the
Buddha's utterances are not connected with his own
last stay at Pava.(2) The fact that the death of
Mahavira evokes the mention Nirvana.
______________________
1. Cunda here appears as a novice, and so also in the
Pasadika Suttanta, which marks him out from his
description in the Mahaparinibbana Suttanta. The
Samgiti Suttanta does not use this term of him,
and seems to have been influenced by the
Mahaparinibbana in this point; compare Franke,
Digha Nikaya, p. 229. Two Cundas can hardly be
admitted, though the Mahaparinibbana is certainly
confused.
2. The Samgiti sets the scene in Pava, but under quite
other circumstances than those of the
Mahaparinibbana, namely the consecration of the
new Mote-Hall of the Mallas. This indicates that
the author had on desire to connect the episode
recorded with the death of the Buddha also. The
location at Samagama seems the more accurate
account. The fact that Cunda of Pava brought the
news to Ananda no doubt encouraged the idea that
the declaration of views took place at that town.
p. 863
of the possibility of the effect on the order of the
Buddha's death does not indicate that that death was
then imminent. It may be noted also that in the Upali
Suttanta the Buddha was at Nalanda when the episode
of the defection of Upali had so evil an effect on
Mahavira that it brought about, according to the
tradition followed by Buddhaghosa, his death at Pava.
At any rate, it is clear that we have no reason to
assert that Buddhist tradition placed the death of
Mahavira close to that of the Buddha, and it is then
obvious that the silence of the Mahaparinibbana
Suttanta is inevitable. If the tradition placed the
episode as to Mahavira before the short period
covered by that Suttanta, it could not possibly
include it in its narrative. So far, therefore, from
correcting the version of the other Suttantas, the
Mahaparinibbana Suttanta accords excellently with
them. Nor(3) can it be admitted that the Buddha,
according to tradition, shows no concern for the
future of his order after his death. This runs
counter to the fact, recorded in the Mahaparinibbana
Suttanta itself, that he assured Ananda that the
place of himself as teacher would be taken by his
doctrine. This assurance is significant of the
position. It accords exactly with the frame of mind
asserted in the other Suttantas to have been
engendered by the news of the dissensions in the Jain
community on Mahavira's death. In the three Suttantas
alike, the result of the news is to make the Buddha
insist that his doctrines provided a definite system
which would prevent schisms in the community. In the
Mahaparinibbana the Buddha gives the same advice; his
doctrine is to serve as the norm. So far, therefore,
from the Mahaparinibbana contradicting the testimony
of the three Suttantas, it is perfectly consistent
with it, while there is no evidence whatever that it
is earlier in date that the other three Suttantas, or
at least two of them.
Thirdly, to strengthen his view that the Buddha
could not have known of strain in the Jain community
on Mahavira's death, Professor Jacobi insists that
there is no record in the Jain tradition of such a
catastrophe in the Jain community at the death of
Mahavira as is suggested by the Buddhist tradition.
No schism, it can be asserted, was occasioned by the
death of Mahavira. Indeed sects among the Jains
developed relatively late, save in the case of the
division into Cvetambaras and Digambaras which was
not the result of a single period of conflict. The
Buddhists, on the other hand, knew of schisms in
their own community, arising soon after the Master's
death and resulting in the development of the new
religion of the Mahayana. They did not realize that
Mahavira was not the founder
p. 864
of a new religion, but merely the reformer of that of
Parcva, so that on Mahavira's death no catastrophe
was possible. The Buddhist account, therefore, in the
three Suttantas is based erroneous assumptions and
was evoked by dogmatic needs.
This interesting suggestion rests on a very
unsound basis. It assumes that the Buddhists believed
that a formal schism or a catastrophe afficted the
Jain congregation on the death of Mahavira. But this
is much more than we can justly deduce from the
Buddhist statements. All that is said is that there
arose disputes, division, and a wordy warfare in the
community and that the lay followers were disgusted
with the monks. Not a suggestion is made of a real
schism or catastrophe, and there seems no reason
whatever to suppose that the Suttantas intended to
assert that such a schism occurred. Moreover, it
seems hard to accept the view of the paucity and
lateness of schisms in the Jain community. The
evidence is that Mahavira was much troubled by the
rivalry of Gocala, whether we regard him as strictly
within the Jain community or not,(l) that in his
fourteenth year of power his son-in-law, Jamali,
raised opposition to him, and persisted in opposition
to his death, while two years after Jamali's revolt,
Tisagutta stood out in opposition.(2) Moreover, the
divergence between Cvetambara and Digambara is
fundamental, as is fully recognized by Jains at the
present day,(3) so that it was certainly unnecessary
for Buddhists to go to their own experience to find
justification for the belief in divergence within the
Jain community. There is, in fact, nothing whatever
to suggest that Buddhist tradition was wrong in
asserting that Mahavira's death caused commotions in
the Jain community. To judge from the bitter feud
between Mahavira and Gocala and from the revolts of
Jamali and Tisagutta, not to mention the defection of
Upali, we may take it as certain that the community
was far from being in ideal unity of heart. The
argument that there could be no schism, because (1)
Mahavira was the child of parents who were adherents
of Parcvanatha, as he perhaps also was, and (2) as a
Kevalin, Mahavira was above all worldly interests,
cannot be accepted. Apart from the fact that we are
not told of anything so serious as a definite schism
or catastrophe, it is clear that Mahavira was no mere
follower of Parcvanatha. The Jain tradition
______________________
1. Hoernle, ERE. i, pp. 267 ff., held that the Jain
division into Digambara and Cvetambara mav be
traced back to the beginning of Jainism, being due
to the antagonism of Mahavira and Gocala, the
representatives of two hostile sects.
2. See Chimanlal J. Shah, Jainism in Northern India,
pp. 60-5.
3. Chimanlal J. Shah, op. cit., p.78.
p. 865
does not even assert that he was an adherent, but, on
the contrary, tells us distinctly that he departed in
an essential from the doctrines of his predecessor,
as was long ago stressed by Professor Jacobi(1)
himself, who held that the innovation postulated a
decline in the morality of the community between
Parcva and Mahavira. Moreover, even if, as a Kevalin,
Mahavira was superior to worldly considerations, what
has that to do with the effect of his death on the
community? The disappearance of a great teacher is
always a time of trial for his adherents, and, so far
from doubting the truth of the assertions of the
Buddhist texts, we may treat them as representing the
normal result as in the case of Purana Kassapa, and
common sense invites us to believe that what is
normal really happens.
Still less satisfactory is the explanation
offered by Professor Jacobi of the cause of the
alleged Buddhist error. The Buddhists, he holds,
confused the place of Mahavira's death, which is now
identified with a village, Papapuri (Pavapuri) in the
Bihar part of the Patna district, with the town(2)
Pava in which the Buddha stayed in the house of Cunda
on the way to Kusinara. The correctness of the Jain
identification, Professor Jacobi holds, cannot be
doubted. This seems a strange assertion, for he holds
that the three Suttantas fall in the second or third
century after the Nirvana of the Buddha, and he does
not give any indication of the age of the Jain
identification.(3) To assert an error on the part of
the Buddhists demands support by adduction of proof
of the early date of the Jain view, which appears to
be lacking and, at any rate, is urgently required.
But, apart from this minor consideration, what ground
is there for holding that a mistake as to a place was
sufficient to cause the invention of an assertion of
the death of Mahavira in the lifetime of the Buddha?
It is perfectly legitimate to suppose that the
Buddhists were right in placing the death of the
rival teacher before that of Buddha, even if they
confused the two places. But that they were wrong in
their identification is so far quite unproved, though
possible.
It must be added that the tradition that the
Buddha died after Mahavira, thus asserted with
particularity in the Buddhist texts, recorded within
two or three centuries after his death, according to
______________________
1. IA. ix, p. 160.
2. Jacobi (p. 561) ascribes Pava to the Cakyas, but it
is clear that it was a Malla town.
3. The Kalpa Sutra ascribed to Bhadrabahu is clearly
not by that author, and is wholly uncertain in date;
see Winternitz, Geschichte der Indischen Litteratur,
ii, pp. 309 f.
p. 866
Professor Jacobi's own dating, is not contradicted by
anything expressed in the Jain tradition, and that
the contradiction rests on the strength of a
deduction from two late and unsatisfactory traditions
fixing the date of the deaths of the two teachers. If
the Jain tradition contradicted the Buddhist by
asserting that Mahavira died after the Buddha, the
case for Professor Jacobi's view would assume a
different aspect; but, though the Jains must for many
centuries have been aware of the Buddhist assertion,
there has been adduced no passage in which they
negatived it. The obvious conclusion is that no doubt
existed in either comunity on this point.
Professor Jacobi has endeavoured on the basis of
the Jain and Buddhist traditions to throw some light
on the political development of Magadha in the time
of the great teachers. but it may seriously be
doubted if we can make anything very satisfactory out
of these confused and obviously biased records. There
is no independent control available, and combinations
thus become subjective to the highest degree. But one
point with which he deals elsewhere(1) should be
noted, his belief that Parcva can be assigned
confidently to a period 250 years before Mahavira, a
view which is utilized by him as assigning to the
early part of the eighth century B.C. that influence
of popular religious belief on Indian philosophy,
which led to the innovations of the Yoga and Samkhya
systems, involving (1) belief in the personal
immortality of souls, and (2) the recognition of
moral principles, and thus advancing beyond the
monistic tendency of the older Upanisads with their
intellectual disdain for morals. We really cannot
accept, as in any sense valid, the date assigned to
Parcvanatha. If Jain tradition was wrong, as
Professor Jacobi holds it was. in dating the Nirvana
of Mahavira, how can we trust its assertions for a
period 250 years earlier? The mere figure is
suspicious, and why should We give it any greater
credence than we do to the figures equally afforded
by tradition(2) for the number of his adherents? All
that we can possibly rescue from the tradition is the
belief in the existence of Parcva at some time before
Mahavira; to claim more is misleading. There are
other objections to certain features of Professor
Jacobi's most interesting reconstruction of the early
Yoga, but these must be dealt with on another
occasion.(3)
______________________
1. SBA. 1930, pp. 326, 327.
2. See Kalpa Sutra, sections 161-4.
3. It is dubious if the Bhagavati vii, 9, 2, can be
understood, as by Professor Jacobi (p. 564), as
meaning that the Mallakis and Licchavis were the
chief of the Kacis and Kosalas.
