JPTS 2007 Xxix

Journal of the Pali Text Society
Pali Text Society
JOURNAL
OF THE
PALI TEXT SOCIETY
VOLUME XXIX
Festschrift in honour of the 80th birthday of

K.R. Norman in 2005
and the 125th anniversary in 2006
of the founding of the Pali Text Society
EDITED BY
O. VON HINÜBER,
R.M.L. GETHIN
AND
MARK ALLON
Published by
THE PALI TEXT SOCIETY
BRISTOL
2007
COPYRIGHT NOTICE
All rights reserved. Subject to statutory exceptions, no part of this

publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, recording or
any information storage and retrieval system, without prior permission
in writing from the Pali Text Society, 36 Lake View, Calne, Wiltshire,
SN11 8JA, U.K.
© Pali Text Society 2007
isbn 0 86013 438 5

isbn-13 978 086013 438 1
issn 0958–8892
First published in 2007
Printed in Great Britain by

Antony Rowe Ltd, Chippenham, Wiltshire
K.R. Norman
(Photo courtesy of Geoffrey Wells)

The Journal of the Pali Text Society
Vol. XXIX (2007)
CONTENTS
Preface
by Oskar von Hinüber ix
Stretching the Vinaya Rules and Getting Away with It
by Petra Kieffer-Pülz 1
The Sus¥ma-sutta and the Wisdom-Liberated Arahant
by Bhikkhu Bodhi 51
A Note on the Heterodox Calendar and a Disputed
Reading in the Kålakåcåryakathå
by Paul Dundas 77
caveat lector
by Margaret Cone 95
The Buddhist Bhik!u’s Obligation to Support His
Parent in Two Vinaya Traditions
by Gregory Schopen 107
Commentaries, Translations, and Lexica : Some
Further Reflections on Buddhism and Philology
by E.G. Kahrs 137
A Note on micchådi††hi in Mahåvaµsa 25.110
by P.S. Jaini 153
Sa"khepasårasa"gaha : Abbreviation in Påli
by Kate Crosby 169
Recent Japanese Studies in the Påli Commentarial
Literature : Since 1984
by Sodo Mori 175
On MahåyånasËtrålaµkåra VII.1
by Lambert Schmithausen 191
vii
viii Contents
Sanskrit Ik!våku, Påli Okkåka, and Gåndhår¥ I!maho
by Richard Salomon and Stefan Baums 201
A Gåndhår¥ Version of the Simile of the Turtle
and the Hole in the Yoke
by Mark Allon 229
Remarks on the Third Precept : Adultery and
Prostitution in Påli Texts
by Steven Collins 263
A Note on vinaya
by Minoru Hara 285
Zombies and Half-Zombies : MahåsËtras and
Other Protective Measures
by Peter Skilling 313
Three Påli Works Revisited
by Nalini Balbir 331
What’s in a Repetition ? On Counting the Suttas
of the Saµyutta-nikåya
by Rupert Gethin 365
The Career of Women Disciple Bodhisattas
by William Pruitt 389
On the Correspondence of Helmer Smith
and Gunnar Jarring
by Siegfried Lienhard 407
An Index to JPTS Volumes IX–XXIX 421
Contributors to this volume 429
PTS Research Grants 431
Preface
Two particularly auspicious events in the history of the Pali Text

Society almost coincided in the years 2005/2006, when a pair of
memorable events in the Society’s life could be celebrated : The 125th
!
anniversary of the foundation of the Pali Text Society in 2006, and in

2005 the eightieth birthday of its long-serving President, editor of the
Journal and doyen of Påli studies, Kenneth Roy Norman. Therefore, the
council of the Pali Text Society decided during the meeting on
16!September 2005 to mark these birthdays by the publication of a
special number of the Journal as a joint Festschrift for both,
K.R.!Norman and the Society, which has been a centre of his life for
many years. The Pali Text Society gratefully acknowledges the
indebtedness to all scholars who without hesitation accepted the
invitation to contribute to the present volume.
When the foundation of the PTS was announced by Thomas
William Rhys Davids (1843–1922) during his Hibbert Lectures in May
18811 with the original plan to “render accessible to students the rich
stores of earliest Buddhist literature” (JPTS 1882, p. vii), printing of
major Påli text editions was already going on in England and continued
outside the PTS for some time. For, the Jåtaka edited by Viggo Fausbøll
(1821–1908) appeared from 1877 onwards, the Vinaya-pi†aka edited by
Hermann Oldenberg (1854–1920) from 1879 onwards followed in 1880
by the Milindapañha edited by Villem Trenckner (1824–1891). All
three editions were later integrated into the program of the PTS as
reprints.2
On the other hand, the PTS was originally, in spite of its name, not
closed to other Buddhist traditions. The very first report for the year
1882 lists the plan to publish titles such as the Jåtakamålå by Hendrik
1JPTS I (1882), p. 1. The first of the yearly Hibbert Lectures under the auspices
of the Hibbert Trust (founded in 1847 by Robbert Hibbert (1770–1849)) was
delivered by Friedrich Max Müller in 1878, “On the religions of India”. Rhys
Davids spoke about “Indian Buddhism” in the fourth Hibbert Lecture.
2The Milindapañha in 1928, Vol. I, of the Vinaya-pi†aka in 1929 and Vols.
II–V as late as 1964, and similarly the seven vols. of the Jåtaka from 1962 to
1964.
The Journal of the Pali Text Society, Vol. XXIX (2007), pp. ix–xiv
x Oskar von Hinüber
Kern (1833–1917),3 which later appeared as the first volume of the
Harvard Oriental Series in 1891, the Madhyamakav®tti by Cecil Bendall
(1856–1906) and the Bhagavat¥ by Ernst Leumann (1859–1931) —
neither of them ever materialised — and, without indication of a
prospective editor, also the Lalitavistara (JPTS 1882, p. 10). On the
other hand, one of the earliest, if not the very first publication of the
Society was the only Jaina text ever published by the PTS : Hermann !
Jacobi (1850–1937), “The Âyâraµga Sutta of the Çvetåmbara Jains”

(1882),4 which evoked some justification in the preface (p. vii) to avoid
“that Buddhist subscribers … might take umbrage at the intrusion, as it
were, of an heretical guest into the company of their sacred Suttas”.
This sin, however, was never repeated, and only translations such as the
one of the Mahåvastu (1949, 1952, 1956) by John James Jones
(1892–1957) and Ronald Eric Emmerick’s (1937–2001) The SËtra of
Golden Light (1970, 2nd ed. 1990, 3rd ed. 1996) are rare and
outrageous steps beyond the path of Theravåda orthodoxy.
From the very beginning, the PTS was run by a truly international
board of scholars presided over by T.W. Rhys Davids as longest serving
President ever from 1881 until his death in 1922, 5 followed by his wife
Caroline Augusta Foley Rhys Davids (1857–1942 ; first mentioned as a
!
3A translation of the Jåtakamålå by Jacob Samuel Speyer (1849–1913) “The

Jåtakamålå. Garland of Birth-Stories of ÓryaßËra” was published from London
1895 (reprinted Delhi, 1971) as Vol. I of the Sacred Books of the Buddhists
(SBB) originally edited by Friedrich Max Müller. This series was later
absorbed by the PTS when Rhys Davids took over as editor beginning with
SBB Vol. III!: Dialogues of the Buddha, Vol. II (Oxford 1910).
4The same year 1882 saw also the publication of the Buddhavaµsa together
with the Cariyåpi†aka. According to JPTS (1884), p. xiii, the first volume of
the A"guttara-nikåya was also published in the same year. However, the date
given on the title page is 1883 !; in fact, following objections from a number of
Sinhalese scholars to Richard Morris’s (1833–1894) practice of introducing
unwarranted abbreviations into the text (see JPTS (1883), p. xii), this volume
was withdrawn and a revised one was published in 1885.
5On the life of the founder see Ananda Wickremeratne, The Genesis of an
Orientalist!: Thomas William Rhys Davids and Buddhism in Sri Lanka (Delhi,
1985), reviewed by Charles Hallisey, JAOS 107 (1987), pp. 515–16 !; and
C.A.F. Rhys Davids, “The Passing of the Founder”, JPTS 1920–1923, pp.
1– 21, with a portrait of T.W. Rhys Davids and the signatures of the then
members of council (p. 21).
Preface xi
member of the board in J P T S 1 9 0 7 /8 ) from 1 9 2 3 until 1942.

Consequently, the Society was dominated for the first 61 years —
almost half of its existence — by the Rhys Davids family.
Besides T.W. Rhys Davids there were four board members in the
beginning : Viggo Fausbøll,6 Richard Morris (1833–1894), Hermann
!
Oldenberg and Émile Senart (1847–1928), who may have been the
longest-ever serving officer with forty-seven years of standing from
1881 until his death.
The early development of the PTS can be traced rather easily as
long as the Journal appeared more or less regularly until 1927, 7 because
annual reports were printed and so were the names of the members of
the board irregularly from time to time. This was not continued when
the new series of the Journal started to celebrate the hundredth
anniversary of the Society.8
The new series of the Journal began under the editorship of K.R.
Norman, who continued as editor until Vol. XX (1994) and served as
President of the Society until 30 September 1994, following his election
on 23 June 1981, when he succeeded Isaline Blew Horner (1896-1981),
President from 1959 to 1981 ; William Stede (1882–1952), President
!
from 1 9 5 0 to 1958 ; William Henry Denham Rouse (1863–1950),

!
President from 1942–1950 and immediate successor to the Rhys Davids

family. Successors to K.R. Norman are Richard Francis Gombrich from
6After his death, Fausbøll was succeeded by Dines Andersen (1861–1940).

7In the first series there are issues for the years 1882, 1883, 1884 (Vol. I of the
reprint)!; 1885, 1886, 1887 (Vol. II)!; 1888, 1889 (Vol. III)!; 1890, 1891/93,
1894/96 (Vol. IV) !; 1897/1901, 1902/3 , 1904/5 , 1906/7 (Vol. V)!; 1908, 1909,
1910/12 (Vol. VI) !; 1913/14, 1915/16, 1917/19, 1920/23 (Vol. VII) !; 1924/27
(Vol. VIII). The reasons for the interruption of the Journal until 1981, when
Vol. IX was published, are unknown. — An index to JPTS 1882-1927 by
P.D.!Ratnatunga was published in 1973 and is continued in JPTS XXVIII
2006, pp. 177– 83.
8On the first century of the Society, cf. three articles by K.R. Norman, “The
Pali Text Society 1881– 1981”, The Middle Way, 56/2 (1981), pp.!71–75 (=
Collected Papers II (1991), pp. 194– 99) !; cf. “The Pali Text Society!:
1981–1986”, Jagajjoti 1986, pp. 4– 8 [= Collected Papers III (1982), p.
108– 14] !; and “Påli Studies in the West!: Present State and Future Tasks”,
Religion 24 (1994), pp. 165–72!; “The present state of Påli studies, and future
tasks”, Memoirs of the ChËØ Academic Research Institute 23 (1994), pp. 1–19
(= Collected Papers VI (1996), pp. 68–87).
xii Oskar von Hinüber
30 September 1994 to 13 September 2002, Lance Selwyn Cousins from
13 September 2002 to 19 September 2003, and Rupert Mark Lovell
Gethin since 19 September 2003.
Looking back at the past twenty-five years the major change in the
life of the Society was effected by the unbelievably generous legacy by
I.B. Horner, which enabled the Pali Text Society overnight to promote
Påli studies in a completely new way by financing projects and by
giving grants to young Påli scholars. Biennial I.B. Horner Lectures have
been held since 1986 in memory of the donor.9
An important result of this new potential was the immediate plan of
a revision of the Påli–English Dictionary (PED) published by the Pali
Text Society in 1925, originally envisaged by K.R. Norman himself,
who, however, made only slow progress due to his many other commit-
ments. Therefore, Margaret Cone was employed from 1 October 1984
first as Research Assistant and then (from 1992) as Assistant Director of
Research in Pali Lexicography attached to the Faculty of Oriental
Studies at the University of Cambridge.10 The original plan of a mere
revision of the dictionary was soon abandoned and the first part of a
much larger work appeared in 2001 under the title A Dictionary of Påli
covering the entries a – khyåti.11
Moreover, the PTS added completely new areas to its activities
during the past quarter of a century. In 1994 the first medical text was
published, the BhesajjamañjËså (Chapters 1–18), followed in 2002 by a
9This was resolved during the council meeting of 18 March 1986. The
I.B.!Horner lectures are regularly listed in the Society’s Journal, cf. JPTS
XXVIII (2006), p. 175.
10According to the minutes of the council meetings on 25 September 1984 and
25 April 1985. — It had been planned earlier during the council meeting on
3!April 1984 “to employ an editorial secretary to assist the President by
working half-time on the Dictionary and half-time on copy-editing etc.”
11Reviewed by K.R. Norman, Buddhist Studies Review, 18/2 (2001),
pp.!252 –53 and Th. Oberlies, OLZ 94. 2004, columns 491– 95. — On the
history of Påli lexicography cf. K.R. Norman, “A Report on the Påli
Dictionaries”, Buddhist Studies / BukkyØ KenkyË 15 (1985), pp.!145–52 and
O.v. Hinüber, “The Critical Påli Dictionary!: History and Prospects”,
Lexicography in the Indian and Buddhist Cultural Field (Proceedings of the
Conference at the University of Strasburg, 25–27 April 1996) Studia Tibetica!:
Quellen und Studien zur tibetischen Lexikographie, Band IV (München, 1998),
pp. 65– 73.
Preface xiii
translation, both by Jinadasa Liyanaratne. Very recently in 2005, the

PTS ventured into another new field, the translation of Påli texts into
languages other than English by Danièle Masset Stances des Ther¥ and
by Nyånaponika Darlegung der Bedeutung (Atthasålin¥).
Since 1981 publications of texts and translations from South East
Asia have been incorporated into the program beginning with Heinz
Bechert (1932–2005) and Heinz Braun, Påli N¥ti Texts of Burma in
1981. In the same year Paññåsajåtaka (Zimmè Jåtaka), Vols. I and II
(1983), edited by Padmanabh Shrivarma Jaini, was published and later
accompanied by two volumes of translation in 1985/6. Other texts
followed such as the Lokaneyyappakaraâ (1986) and the Pa†hama-
saµbodhi (2003), edited from a manuscript prepared by George Cœdès
(1886–1969) and rediscovered in the Archives of the École française
d’Extrême-Orient by Jacqueline Filliozat. The Jinakålamålå Index
(1994) by Hans Penth, indispensable for the study of earlier northern
Thai history, was a major step in the publication of ancillary literature
for the understanding of Påli literature as was the huge catalogue on
Sinhalese Manuscripts in the Nevill Collection.12
Although very modestly listed under “ancillary works” the monu-
mental Collected Papers13 by K.R. Norman may be considered as one
of the outstanding publications of the PTS, as a standard work of
reference not only for Påli studies, linguistic or literary, but, at the same
time, also for Jainism. Together with the monograph A Philological
Approach to Buddhism (based on a series of lectures delivered at the
School of Oriental and African Studies, London, when he was the
BukkyØ DendØ KyØkai Visiting Professor in 1994, and reprinted by the
PTS only recently in 2006 in a corrected version), the Collected Papers
bear ample witness to the rich fruits of the scholarly life of
K.R.!Norman devoted to middle Indian studies which was honoured by
his election as a Fellow of the British Academy in 1985 and by a
felicitation volume at the time of his retirement from the professorship
in Indian Studies at the University of Cambridge on 30 September
12Published jointly with the British Library, where the collection is kept, in
seven volumes between 1987 and 1995.
13Vols. I (1990), II (1991), III (1992), IV (1993), V (1994), VI (1996), VII
(2001), VIII (2007) !; index I–VII by A. Rook, JPTS XXVI (2000), pp.
169– 231.
xiv Oskar von Hinüber
1992. 14 Almost simultaneously K.R. Norman withdrew from the
Critical Påli Dictionary, which he had edited very successfully from
1979 to 1990 thus covering the second half of volume II. It is certainly
not easy to find another set of contributions to Påli in size and
importance equal to all these lifelong efforts and achievements.
This induced the Pali Text Society to mark the eightieth birthday of
K.R. Norman by a token of recognition for his work as scholar and as
President of the Society, to whose life he contributed in many ways
over many decades, in particular after its hundredth anniversary, by
voluminous and rich publications, sound advice, and efficient
leadership.
Freiburg, Tuesday, 7 August 2007 Oskar von Hinüber
14IIJ35 (1992), parts 2/ 3 , pp. 81–272 with a bibliography only of books

published by K.R. Norman. Therefore, a complete bibliography remains an
urgent desideratum.
Stretching the Vinaya Rules and Getting Away with It*
Eleventh I.B. Horner Memorial Lecture, 2005
1. Introduction
One central point of interest in I.B. Horner’s fields of research was
Buddhist law. She was the fi rst to translate the Påli version of the
complete Buddhist law code (Vinaya-pi†aka) into a European
language. 1 In this eleventh I.B. Horner Memorial Lecture some ideas
about the perennial question of how to stretch the Vinaya rules and get
away with it are examined. The first part centres on the nature of
Buddhist law. It is followed by an overview of the legal literature of the
Theravåda tradition (as far as it is relevant to the final part), with special
attention to the question of how much authority is attributed to various
texts. The final part will deal with two methods for stretching the
Vinaya rules.2
* This article is an outcome of my work on “Die in der Vajirabuddhi-†¥kå

zitierten Gaˆ†hipadas!: ein annotierter Zitatenkatalog zur Geschichte der
Rechtssentwicklung bei den Theravådin” (The Gaˆ†hipadas quoted in the
Vajirabuddhi-†¥kå!: an annotated catalogue of quotations concerning the history
of the legal development of the Theravådins) at the Institute for Indology,
Martin Luther University Halle-Wittenberg, promoted by the Deutsche
Forschungsgemeinschaft. Reinhold Grünendahl read an earlier version of this
article and made many suggestions and improvements, Anne Peters supplied a
number of references to PTS editions not accessible to me, Peter Jackson in
proofreading the text made some further corrections and suggestions, and
William Pruitt corrected the remaining mistakes and completed the last
missing references. I wish to express my gratitude to them for their help.
1She only left out passages which seemed to her to be too rude for Westerners.
See Kieffer-Pülz 2001.

2I will not deal here with issues not covered by existing law. Such cases have to
be handled according to the guidelines (mahåpadesa) handed down in the

Khandhaka portion of the Vinaya, according to which new cases have to be
decided in analogy to, and avoiding conflict with, existing prescriptions (Vin I
250, 31–51, 6 !; BD IV 347). The commentarial tradition of the fourth or fifth
century developed this method systematically (Sp I 230, 21–33, 35 ad Vin III
The Journal of the Pali Text Society, Vol. XXIX (2007), pp. 1–49
2 Petra Kieffer-Pülz
1.1 The Character of Buddhist Law
A fully ordained Buddhist monk (bhikkhu) or nun (bhikkhun¥) has to
comply with an abundance of rules governing almost every aspect of
daily life. These rules are laid down in the Buddhist law code, the
Vinaya-pi†aka. Of the various Buddhist schools that developed during
the long history of Buddhism, many had a Vinaya of their own. I will
confi ne myself here to the Vinaya of the Theravåda, or, more
specifically, of the Mahåvihåra school, handed down in the Middle
Indic language Påli. Before this text was written down in Sri Lanka in
the first century B. C., it was transmitted orally. Thus we can say that the
Vinaya developed over a period of around four hundred years before it
took its final shape. It is divided into three parts : (1) the Suttavibha"ga
!
with the 227 rules constituting the Påtimokkha, to be recited every

fortnight, as the main part, (2) the Khandhakas containing the rules for
administrative affairs of the Buddhist community (sa!gha), and (3) the
Parivåra, a later systematization of the rules. This law code is still
authoritative for present-day Theravåda monks in South and Southeast
Asia.
During the Buddha’s lifetime and, in some respects, right up to the
time when the Vinaya-pi†aka was fi xed in writing, Buddhist law was
dynamic. There are various indications of this. To begin with, in some
cases the Vinaya provides several formulas for one and the same
ceremony, with layer added upon layer and the most recent formula
replacing the older ones.3 Then we have various Påtimokkha pre-
scriptions (paññatti) modified by several supplementary prescriptions
(anupaññatti), no less than seven in one case.4 Furthermore, we have
relaxing of restrictions for a number of rules for the borderlands.5
Finally, the youngest part of the Vinaya, the Parivåra, occasionally
23 ,37 , (BD I 42)). Vjb adds many examples for the different cases (Vjb
88,2–90,7).
3See for instance the case of higher ordination, n. 18.
4Påc 32 Mk, Vin IV 71,18–75 ,23 (BD II 306–14).
5Von Hinüber 2000, p. 144.
Stretching the Vinaya Rules and Getting Away with It 3
deals with subjects not spoken of in the rest of the Vinaya.6 However,
with the parinibbåna of the Buddha, and, at last, with the writing down
of the Vinaya, the dynamism of Buddhist law gradually came to an end,
with hardly any adaptations being made to new circumstances there-
after. Even though the Buddha himself had allowed for doing away with
minor rules, uncertainty as to what should be considered a minor rule
prevented the monks from changing the rules at all.7 Now, once the
wording of the law is considered fixed or even sacrosanct, the only way
left to adapt it to unforeseen circumstances is to interpret it in a different
manner.8
1.2 A sketch of the Vinaya commentaries
The practical relevance of Buddhist law for the Buddhist community led
to a multitude of commentaries, not only on the Vinaya, but also on the
Påtimokkha which, for practical reasons, was handed down as a
separate text alongside the Vinaya. The authority of these texts is also
reflected in the constant production of law handbooks and related
commentaries. More than twenty complete law commentaries written in
Påli up to the nineteenth century have come down to us. But the number
must have been much higher, as is evident from the many lost
commentaries quoted in the existing ones. Leaving aside the oldest
commentary, the Suttavibha"ga, the first commentary known to us is the
now lost S¥ha¬a††hakathå under which designation several commentaries
are subsumed, among them the Mahåpaccar¥ and Kurund¥, written
down, probably together with the canon, in Sri Lanka as early as the
6The mention of an atikhuddakå s¥må presupposes a definition of the smallest

measure of a s¥må, not given explicitly in the Vinaya (Kieffer-Pülz 1992,
p.!136, §!11.2.1)!; a khaˆ"animitta presupposes a definition of the marks first,
which also is not given (Kieffer-Pülz 1992, p. 137, §!11.2.3).
7Vin II 287 ,29ff. (BD V 398ff.). See von Hinüber 1995, p. 14.
8We find a very early example of this method in an old word-by-word com-
mentary on the rules of the Påtimokkha that has been incorporated in the
Suttavibha"ga.
first century B.C. 9 How far they date back we do not know, and we
probably never will.10 These early commentaries served as sources for
the great commentaries, the so-called a††hakathå literature of the fourth
and fifth centuries, i.e., the Ka"khåvitaraˆ¥, a commentary on the
Påtimokkha, and the Samantapåsådikå, which covers the entire Vinaya.
After the fifth century, another category of commentaries developed, the
so-called gaˆ†hipadas, written in Påli, Sinhalese, and possibly other
languages as well. Some of them still circulate in printed editions, many
others are preserved in manuscript form, but most are now lost, apart
from the passages quoted from them in other gaˆ†hipadas or in the sub-
commentaries, that is the †¥kås, written mainly in the twelfth to
thirteenth centuries. These are followed by Påli commentaries covering
the entire Vinaya or parts of it, and commentaries on Vinaya
handbooks.11
1.3 The authority of legal texts
Now what about the authority of these legal texts from the perspective
of the individual Buddhist monk? Every single monk has to make his
!
own decision as to the authority he attributes to a certain text. This

equally holds true for the authors of the legal texts just mentioned, who
were also monks. My work on the legal literature has led me to the
9Mhv 33,100–101!: pi†akattayapåliµ ca tasså a††hakathaµ pi ca mukhapå†hena

ånesuµ pubbe bhikkhË mahåmat¥!; håniµ disvåna sattånaµ tadå bhikkhË
samågatå cira††hitatthaµ dhammassa potthakesu likhåpayuµ. “The text of the
three pi†akas and the a††hakathå thereon did the most wise bhikkhus hand down
in former times orally, but since they saw that the people were falling away
[from religion] the bhikkhus came together, and in order that the true doctrine
might endure, they wrote them down in books.” [Translation by W. Geiger,
Mhv (transl.), p. 237].
10See von Hinüber 1996, §!210.
11One was written in Northern Thailand in the fifteenth century, and there are
two from Burma written in the seventeenth and nineteenth centuries

respectively. Furthermore, we have collections of judgements pronounced by
various sa!gharåjas and associated jurists on a range of legal topics, as well
as epistolary correspondence between monks from various countries
discussing questions of Buddhist law.
conclusion that the authority of the Vinaya is acknowledged by almost
all authors. By contrast, statements of the so-called S¥ha¬a††hakathå
were considered open for discussion in all later commentaries, which do
not hesitate to reject them or even declare them irrelevant on account of
their supposedly defective or missing argumention. 12 However, there
are also attempts to reconcile what are seen as inner contradictions of
the S¥ha¬a††hakathå. The teachings of the Vinaya and a††hakathås of the
fourth and fifth centuries are generally accepted as authoritative by the
gaˆ†hipada commentaries and the †¥kås, whereas the †¥kås frequently
reject opinions expressed in the gaˆ†hipadas, usually without even
considering it necessary to discuss them.
From more recent times we have some explicit statements of monks
regarding the authority they attach to certain law texts. Vajirañåâ
Maku†a, perhaps better known as King Mongkut, the founder of the
Dhammayuttika-Nikåya in nineteenth-century Thailand, explained in a
letter written in 1844 to a Sinhalese monk that a thorough investigation
of a topic has to start from the canonical writings, i.e., the Vinaya, and
that it should be possible to reach a solution on the basis of this
material. This refers to the Thai practice of the visuµgåmas¥må, but
nevertheless shows the author’s general attitude.13
At the beginning of the twentieth century, King Mongkut’s son, the
later sa!gharåja, Vajirañåâvarorasa, a member of the Dhammayuttika-
Nikåya, declared that the Vinaya showed clear signs of accretion over a
12For instance the Anugaˆ†hipada or the Vajirabuddhi-†¥kå on the opinion of

keci (Vjb 142,5 –10 ad Sp II 376, 19–20 ad Vin III 58,22– 24 [Pår 2 Mk]).
13Treatise (Se1 XXXIII!; S e2 IX)!: a††hakathå hi Pål¥naµ sa!g¥tikålato pacchå
katå. dhammasa!gåhakehi ca på†hasa!g¥tiµ ni††håpentehi sanni††hånaµ

kataµ!: ettakehi på†hehi pa†ipannakå sakkhissanti taµ taµ vinayalakkhaâµ
ñatvå anupa†ipajjitun ti. “For the a††hakathå was made after the period of the
[first] common recitation (i.e., council) of the [canonical] texts. And the
compilers of the Dhamma, who carried out that common recitation of the
texts, made the decision!: With so many texts [those] who have entered upon
the Path, knowing this and that definition of the Vinaya, will be able to follow
the practice.”
long period of time, and that therefore its words should not be followed
blindly.14
As these two more recent statements show, a monk’s opinion
regarding the authority of a given text is certainly influenced by his
adherence to a specific school, sub-school or local branch thereof, but
the decisive factor is his own judgement. The influence of the local
tradition — transmitted only orally in some cases15 — is difficult to
determine.
1.4 Interpreting the rules
The modern Buddhist monk is confronted with manifold interpretations
of Vinaya prescriptions and definitions of terms used therein, laid down
in the multitude of commentaries written over a period of more than
2,000 years, which produced an ever finer spun texture of regulations.
Hand in hand with the increasing density of regulations, the potential
for legal loopholes increased as well because each case or topic which
was not explicitly covered by these interpretations and definitions could
be seen as falling outside the scope of the respective prescription. This
opened up considerable possibilities for stretching the rules, which for
the most part resulted in a relaxation of the law. In the commentaries of
the fourth and fi fth centuries we observe the tendency to loosen the
rules by applying them only to those groups that visibly fall under the
14“In the Vinaya itself which was handed down for a long time both orally and
by writing, differences of understanding naturally have crept in at the time
when the Ócariyas who understood incorrectly, wrote it down” (Entrance to
the Vinaya I, p. xii). “My habit is not to believe all the words which are found
in the scriptures, but rather believing the reasonable words!; moreover, we
have learned the history of the sacred books, as outlined above, so that we
should not grasp them as our only source. The basis of my writing is that
which is found to be reasonable and this should be taken as credible evidence,
while what is defective should be opposed whether coming from the Påli or
from the A††hakathå” (Entrance to the Vinaya I, p. xiv).
15In the case of the Dhammayuttika-Nikåya, we know from Vajirañåâvarorasa
that the daily practice of this Nikåya, already in continuous use for sixty years
at that time, was handed down exclusively by oral transmission from teacher
to pupil (Entrance to the Vinaya I, p. x).
category explicitly defined in the respective prescription. For example,
the prohibition to ordain people with certain skin diseases was
interpreted in such a way that it applied only to those with increasing
and visible symptoms, while as long as the affected parts of the skin
were decreasing and hidden under the robe the candidate could be
ordained.16
In Buddhism there is no ecclesiastical high court whose decisions
are binding for the entire Buddhist community. Therefore, nobody can
be forced to accept a certain interpretation or doctrine. This leaves
ample space for conflicting doctrines developing and existing side by
side. What is a transgression of a Vinaya rule in the eyes of one group
may be considered legally acceptable by another.
2. Examples of stretching the rule

2.1 Ordination
The fi rst, and most common, method of stretching rules is to interpret a
term used in a Vinaya prescription in such a way that its area of
application is reduced to certain sections of the former defi nition — a
group of people or things, for example — while other sections are
conveniently counted out. The example I have chosen to illustrate this
method is the prohibition against ordaining a slave (dåsa) as a novice.
As is well known, in the beginning the Buddha himself performed
the ordination of new members to the Buddhist community. Later on, he
delegated the office of ordination to monks. At that time, no distinction
between novitiate and monkhood was made.17 Finally, with the
introduction of specific ceremonies for the ordination of novices
(pabbajjå), and the ordination of monks (upasampadå), the ordination
of a monk was performed in a legal procedure consisting of a motion,
16Sp V 995 ,15ff. ad Vin I 71,32–73 ,20 (BD IV 89ff.).

17The Buddha used the ehi-bhikkhu formula, Vin I 12 ,22– 25, 35–13, 1 (BD IV
18f.). With the delegation of the office of ordination to monks, the formula
used was modified. From then on, officiating monks had to recite the
threefold-refuge formula three times, Vin I 22,8– 23 (BD IV 30).
three proclamations and a resolution (ñatticatutthakamma). The
development now was by no means stopped. The Theravåda Vinaya
contains three formulas for the ñatticatuttha procedure. The second one
adds the formal request of the candidate to be given the higher
ordination ;18 the third one clears the candidate of all obstacles that
!
might have prevented his ordination as a monk. The list of possible

impediments contains fifteen obstacles.19 The candidate for ordination
as a monk has to be a human being, male, a free man (bhujissa), free
from debt (anaâ), and not in a king’s service (råjabha†a) ; he has to
!
have his parents’ permission ; he has to be at least twenty years old ; he

! !
has to own robes and a begging bowl ; he should not suffer from leprosy
!
(ku††ha), boils (gaˆ"a), eczema (kilåsa), consumption (sosa) or epilepsy

(apamåra); and he should know his own name and that of his preceptor
!
(upajjhåya).20
From the number of formulas handed down in the Theravåda
Vinaya we can infer that the definition of these impediments is a later
development. However, with its compilation the number of obstacles
was by no means fi xed. The Vinaya has a long chapter listing eleven
persons unqualified for ordination as a monk. 21
18Vin I 56,6 –9 !; 57 , 10–25!; 95,16– 34 (BD IV 72, 73, 123). Three formulas are
given, with each formula being more elaborate than the preceding one. For
the ordination of novices the threefold-refuge formula previously used for
ordaining monks was adapted.
19Other schools have much more (the MËlasarvåstivådins 80 !; Härtel 1956,
pp.!78ff.), which shows that these lists were constantly changing. For changes
within the Theravåda tradition, the Katikåvatas are instructive!; see Ratnapåla
1971, pp. 159f., §§!101 f.!; cf. pp. 255ff.
20Vin I 93,24– 32 (BD IV 120).
21These include the so-called eunuch (paˆ"aka), Vin I 85 ,27 – 86,9 (BD IV
108f.)!; one who gained access to the community by theft (theyyasaµvåsaka),

Vin I 86,10– 3 3 (B D IV 1 0 9 f.) !; one having gone over to another sect
(titthiyapakkantaka), Vin I 86,33– 35 (BD IV 110)!; an animal (tiracchånagata),
Vin I 86,36–88,3 (B D IV 110f.) !; a matricide (måtughåtaka), a patricide
(pitughåtaka), a murderer of a perfected one (arahantaghåtaka), a seducer of
nuns (bhikkhun¥dËsaka), one splitting the sa"gha (saµghabhedaka), one who
In addition to the impediments for higher ordination and to the
individuals unqualified for it, the Vinaya also lists impediments for
ordination as a novice, i.e., for pabbajjå. Some of these are identical
with those for higher ordination, i.e., suffering from one of the five
diseases,22 being in a king’s service (råjabha†a),23 being a debtor
(iˆåyika),24 and being a slave (dåsa).25 Other impediments, however,
are exclusively mentioned in the context of lower ordination, such as
falling under various categories of publicly known thieves.26 Another
section of the Vinaya lists thirty-two examples in which lower
ordination should not be given. This passage includes persons with
mutilations resulting from criminal activities (e.g., severed hands) and
persons with impairments due to diseases.27
sheds a Tathågata’s blood (lohituppådaka), and a hermaphrodite (ubhato-

vyañjanaka), Vin I 88,4–89,21 (BD IV 112ff.).
22Vin I 73 ,18– 20 (BD IV 91) na bhikkhave pañcahi åbådhehi phu††ho pabbåje-
tabbo. yo pabbåjeyya, åpatti dukka†asså ti. The five illnesses are listed Vin I
71,33– 34 (BD IV 89)!: ku††haµ gaˆ"o kilåso soso apamåro, Sp V 995,15– 18.
23Vin I 74 ,24– 2 5 (B D IV 92) !; Sp V 996 ,20– 97 . List of impediments for
upasampadå, Vin I 93,24–32 (BD IV 120)!; Sp does not comment on it.

24Vin I 76,18– 19 (BD IV 95)!; Sp V 999,9 –1000, 17.
25Vin I 76,26– 27 (BD IV 95f.)!; Sp V 1000,19–1002,16.
26On a thief wearing an emblem (dhajabaddha (° bandha) cora), Vin I 74,34– 35
(BD IV 93)!; Sp V 997,10ff. !; on a thief broken out of jail (kårabhedaka cora),

Vin I 75 , 15–17 (BD IV 94)!; Sp V 997 , 26–98, 17!; on a thief against whom a
warrant has been taken out (likhitaka cora), Vin I 75,27– 28 (BD IV 94)!; Sp V
998, 17–24!; on one having been scourged as punishment ( kasåhata katadaˆ"a-
kamma), Vin I 75,33–35 (BD IV 95), Sp V 998 ,24–99,1 !; and on one having
been branded as punishment (lakkhaˆåhata katadaˆ"akamma), Vin I 76,5 –7
(BD IV 95)!; Sp V 999,2– 9.
27Vin I 91 ,7 –11 (BD IV 115f.) !; Sp V 1026,11–31 ,24. Persons with severed or
mutilated hands, feet, ears, noses, fingers, nails, or tendons, with hands like a
snake’s hood (phaâhatthaka#; see BD IV 116, n. 2), a hunchback (khujja),
dwarfs (våmana), persons with a goitre (galagaˆ"i), again three types of
thieves (lakkhaˆåhata, kasåhata, and likhitaka, see n. 26)!; persons with
elephantiasis (s¥padi), with a serious illness (påparogi), persons who disgrace
an assembly (parisadËsaka, see BD IV 116 by some deformity)!; those who
are one-eyed (kåâ), crippled (kuî), lame (khañja), partly paralysed (pakkha-
Of the eight impediments the Vinaya lists for lower as well as
higher ordination, I would now like to take a closer look at the case of
slaves or, more generally, men whose freedom is confined in one way or
another. With regard to higher ordination, it is said that the candidate
has to be a free man (bhujissa),28 which categorically excludes slaves,
bondsmen, and others. By contrast, the restrictions imposed on lower
ordination are more explicit in that they exclude a slave (dåsa) from
pabbajjå, while other types of bondage are not mentioned.
Before I come to the rules themselves, allow me to say a few words
about the relationship between the Buddhist community and slaves.
2.1.1 Slaves and the Buddhist community
Although the possession, usage, and donation of slaves by kings,
merchants, and others seems to have been widespread in the society in
which the Mahåvihåra Vinaya took shape,29 the Vinaya mentions slaves
hata), whose movements are destroyed (chinna-iriyåpatha), who are weak of

age (jaradubbala), blind (andha), dumb (mËga) or deaf (badhira).
After the introduction of a novice’s ordination, it was obligatory to receive
the ordination as a novice before being ordained as a monk, it is therefore to
be supposed that the obstacles for novices were also valid for monks.
28Interestingly, the question in the MËlasarvåstivåda tradition is må asi dåsaß,
“You are not a slave!?”, and an additional question is må vikr¥takaß (Tib.

btso!s-pa ma yin nam), “You have not been sold!?” !; see Härtel 1956,
pp.!78– 79.
29See for instance the story of the householder Meˆ#aka who, in due succes-
sion, shows his own psychic power, that of his wife, son, daughter-in-law,
and that of his slave (Vin I 241,33, 34), or the story of J¥vaka Komårabhacca,
where a merchant’s wife inhaled ghee through her nose, spat it out through
her mouth and ordered a slave-woman (dås¥) to take it up with cotton (Vin I
271, 35). Later we are told that this ghee was used again for rubbing the feet of
slaves or labourers (dåsånaµ vå kammakarånaµ vå pådabbhañjanaµ) or for
pouring into a lamp (Vin I 272 ,7 –8 ). When the merchant’s wife was cured,
J¥vaka received money from several persons, but from the husband he
received in addition a male and a female slave and a horse chariot (Vin I
272, 16). Another merchant promises J¥vaka to become his slave himself if he
is cured (Vin I 274 ,9 !; 275, 17, 18 ). Many references are to be found in the
Jåtakas, see Ray 1986, pp. 96f.
only rarely in connection with the Buddhist sa"gha or its ordained
members. The term slave is defined as comprising three types : (1) one !
born as a slave, (2) one bought for money, and (3) a captive turned into
a slave.30 In other parts of the canon, four types are distinguished, the
three just mentioned and a person who decided to become a slave
himself.31
In one prescription the behaviour of nuns is criticized, when they,
in obvious imitation of the society around them, caused male and
female labourers (kammakara kammakar¥), and male and female slaves
(dåsa dås¥) to wait upon them (upa††håpeti).32 As a result, this
behaviour was prohibited. However, the respective rule does not forbid
the acceptance of slaves by the sa"gha, or an individual monk or nun.33
30Vin IV 224 ,25– 28 [Sgh 1 N] ( BD III 179) !; Vin IV 224, 33!; Geiger 1986, §!29,
p. 375, divides the third type, karamarån¥ta, into two groups, (1) those made
prisoners in war (karamara), and those carried off by force (ån¥ta), but see
DOP s.v. karamarån¥ta. Four types are listed in Nidd I 11,8 –11 (see n. 31).
Manusm¤ti (VIII.415) and Arthaßåstra (III.13) give seven and nine classes of
slaves respectively!: ( 1) those captured in war, (2 ) those who serve for their
food, (3) those born in the house, (4) those who are bought, (5) those who are
given, (6) those who are inherited from ancestors, and ( 7) those enslaved by
way of punishment. The Arthaßåstra adds two more !: those who have either
mortgaged or sold themselves.
31Nidd I 11 ,8 –10 !: dåså ti cattåro dåså!: antojåtako dåso, dhanakkitako dåso,
såmaµ vå dåsavisayaµ upeti, akåmako vå dåsavisayaµ upeti.

32For this meaning of upa††håpeti see CPD s.v. upa††håpeti, 1. Vin II 267 ,10
(see n. 33) (BD V 370!: “they kept slaves, they kept slave women”, etc., is
somewhat misleading!; see, however, BD V 370, n. 6).
33Vin II 267 ,5 –23 ( BD V 370) !: chabbaggiyå bhikkhuniyo .... dåsaµ upa-
††håpenti, dåsiµ upa††håpenti ... na dåso upa††håpetabbo, na dås¥ upa††håpe-

tabbå. “The six [bad] nuns caused a slave to wait upon [them], caused a
female slave to wait upon [them] ... a slave may not be made to wait upon
[oneself], a female slave may not be made to wait upon [oneself].” Sp VI
1293,28– 30!: dåsaµ upa††håpent¥ ti dåsaµ gahetvå tena attano veyyåvaccaµ
kårenti. dås¥-åd¥su pi es’ eva nayo. “They caused a slave to wait upon
[them means]!: Having taken a slave they made him carry out their own
housework. Also in the case of female slaves, etc., exactly this [is] the
method.”
This is confirmed by the chapter on the rains retreat, which mentions the
impending bestowal of a male or female slave and the allowance to
interrupt the rains retreat in avoidance of such offers if they were
considered incompatible with the religious life of a fully ordained
person.34 An unconsenting monk obviously did not have the possibility
to simply reject them.35
The Vinaya-pi†aka does not contain a prescription which explicitly
forbids the acceptance of slaves. 36 However, the Sutta-pi†aka states that
the Buddha himself did not accept male and female slaves.37 Based on
this regulation, the Vinaya †¥kås (twelfth and thirteenth centuries) finally
prohibit the acceptance of slaves. The Såratthad¥pan¥ interprets this as a
prohibition for monks to accept slaves for their own use,38 whereas the
The eighth Prak¥râka of the Mahåsåµghika-Lokottaravådins quoted by

Schopen as a proof that the personal possession of åråmikas was forbidden
by at least some Indian Vinayas (Schopen 1994B, p. 162!; taken for granted
by Yamagiwa 2002:365, n. 5), in fact does not deal with the possession of
åråmikas, but with their usage. What is forbidden in this rule is to cause an
åråmik¥, a ce†¥, a kalpiya-kår¥ to wait upon (upasthåpayati) [oneself]. This
could also be done when these persons belonged to the sa"gha, and thus the
transgression would be that one used åråmikas for one’s own affairs and not
for the sa"gha’s or the monastery’s. Therefore this Prak¥râka rule tallies
with the wording of the rule above given from the Mahåvihåra Vinaya,
except that it does not use the term slave.
34Vin I 150 ,6 –23 (BD IV 198)!: dåsaµ vå te demi, dåsiµ vå te demi.
35Such a rejection would deprive the donor of the merit which results from his
donation, and this presumably could not be an acceptable behaviour for a

monk.
36For further comments on slaves with respect to monasteries, see Geiger 1986,
§!187 , Gunawardana 1979, pp. 97ff.

37This attitude is codified in a set of rules called the minor s¥lå (cËlas¥la), found
in the Brahmajåla-sutta and elsewhere, D I 5 ,14f.!: dåsidåsapa†iggahaˆå pa†i-

virato samaô Gotamo!; D I 64 ,24!; M I 180, 12!; 268,24, etc. !: dåsidåsapa†i-
ggahaˆå pa†ivirato hoti.
38Sp-† II 330 ,22– 24 !: dåsaµ attano atthåya sådiyantassa pi dukka†am eva
dåsidåsapa†iggahaˆå pa†ivirato hot¥ ti (D I 5 ,14f.) vacanato. “Even for one

who accepts a slave for his own use only an [offence] of wrong doing [arises]
Vimativinodan¥-†¥kå declares that the a††hakathås reject the acceptance
of slaves based on this regulation, thus relating this statement to the
designation with which a slave may be accepted (see below).39 That this
still was a question in later times is shown by the Katikåvatas.40
The commentaries of the fourth and fifth centuries provide for the
acceptance of slaves by the Buddhist community, albeit on certain
conditions. For instance, slaves — even if designated as dåsa / dåsi — ! !
may be accepted by the sa"gha if they are part of the donation of a

palace, in which case they are counted among its inventory stock. 41
Furthermore, the sa"gha is explicitly allowed to accept a dyer-slave
(rajakadåsa) and a weaver-slave (pesakåradåsa), provided they are
on account of the [authoritative] statement [of the Sutta texts] !: ‘he abstains
from the acceptance of male and female slaves’.”
39Vmv I 272,9 –11 = Pålim-n† I 65,16– 19!: evaµ yåcato aññåtakaviññattidukka†añ
c’ eva dåsapa†iggahaâdukka†añ ca hoti dåsidåsapa†iggahaˆå pa†ivirato

hot¥ ti (D I 5 ,14f.) vacanaµ nissåya a††hakathåsu pa†ikkhittattå. “For one
begging in that way there arises [an offence of] wrong doing for asking
someone not related as well as [an offence of] wrong doing in case of the
acceptance of a male slave because [it] has been rejected in the a††hakathås
based on the [authoritative] statement [of the Sutta texts]!: ‘he abstains from
accepting male and female slaves’.”
40The Dambadenikatikåvata (twelfth or thirteenth century) states that in accept-
ing male and female slaves (däs-das, v.l. däsi däs) “a well-disciplined, wise
and modest bhikkhu should be [first] consulted and those [slaves , etc.] should
be accepted in the manner indicated by him” (Ratnapåla 1971, pp. 58, 153,
§!68). The K¥rtißr¥råjasiµha-Katikåvata I (eighteenth century) declares that
monks should not treat relatives or non-relatives with proper or improper
possessions, such as … [among others] male and female slaves (dåsi-dåsa,
v.l. däsi-das#; see Ratnapåla 1971, pp. 99, 169, §!103). In a similar way it is
expressed without the term dåsa being used in the K¥rtißr¥råjasiµha-
katikåvata II (eighteenth century) with respect to people living in villages
owned by the Vihåra (Ratnapåla 1971, pp. 109, 175 , §!11).
41Sp VI 1236 ,30 –37 ,1 [ad Vin II 169, 29]!: påsådassa dås¥dåsakhettavatthu-
gomahisaµ demå ti vadanti, på†ekkaµ gahaâkiccaµ natthi. påsåde

pa†iggahite pa†iggahitam eva hoti. “[If] they say!: ‘We give female and male
slaves, fields, grounds, cows and bulls for the påsåda’, there is not an
obligation of a separate acceptance. When the påsåda is accepted, [this] is in
fact accepted.” This was noted already by von Hinüber 2000, p. 147.
presented under the designation of åråmika. 42 In the commentaries on
the Sutta-pi†aka and the Vinaya-pi†aka the rule is extended to all slaves
(dåsa) labelled as åråmika, one who belongs to the åråma, i.e., the
monastery, as kappiyakåraka, legalizer, or as veyyåvaccakara, steward,
attendant.43 All three terms designate persons who carry out all sorts of
42Vin-vn, v. 665 !; Sp III 683 ,17– 18, see below, n. 43.

43Commentaries on the Sutta-pi†aka (Sv I 78,19!; Ps II 209 ,30ff. !; Spk III 304,32ff. !;
Mp III 192,1 –3 !; etc.!: dåsidåsapa†iggahaˆå ti ettha dåsidåsavasen’ eva tesaµ
pa†iggahaâµ na va††ati. kappiyakåraµ (v.l. kappiyakårakaµ) dammi,
åråmikaµ damm¥ ti evaµ vutte pana va††ati. Sp adds a third term,
veyyåvaccakara! : Sp III 6 8 3 , 6– 18 !: dåsaµ damm¥ti vadati, na va††ati.
åråmikaµ dammi, veyyåvaccakaraµ dammi, kappiyakårakaµ damm¥ti vutte
va††ati. sace so åråmiko purebhattam pi pacchåbhattam pi sa!ghass’ eva
kammaµ karoti, såmaêrassa viya sabbaµ bhesajjapa†ijagganam pi tassa
kåtabbaµ. sace purebhattam eva sa!ghassa kammaµ karoti, pacchåbhattaµ
attano kammaµ karoti, såyaµ nivåpo na dåtabbo. ye pi pañcadivasavårena
vå pakkhavårena vå sa!ghassa kammaµ katvå sesakåle attano kammaµ
karonti, tesam pi karaâkåle yeva bhattañ ca nivåpo ca dåtabbo. sace
sa!ghassa kammaµ natthi, attano yeva kammaµ katvå j¥vanti, te ce
hatthakammamËlaµ ånetvå denti, gahetabbaµ. no ce denti, na kiñci
vattabbå. yaµ kiñci rajakadåsam pi pesakåradåsam pi åråmikanåmena
sampa†icchituµ va††ati. “[If] one says!: ‘I give a slave’, it is not allowed!; if ‘I
give an åråmika, I give a veyyåvaccakara, I give a kappiyakåraka’ is said, it
is allowed. If an åråmika carries out work for the sa"gha before meals as well
as after meals, [then] even the whole care for the medicine has to be taken
over by him as by a novice. If he carries out work for the sa"gha only before
meals [and] after meals he carries out his own work, no ration is to be given
to him in the evening. Also to those who, having carried out work for the
sa"gha every five days or every fortnight, who during the rest of the time
carry out their own work, meals and ration are to be given only during the
time of [their] working [for the sa"gha]. If the sa"gha does not have work
[for them], they live carrying out only their own work!; if they procure money
from their manual labour [and] give it, it is to be taken. If they do not give it,
they are not to be spoken to at all. It is allowed to accept with the designation
åråmika whatever slave is a dyer and whatever slave is a weaver.” Khuddas-
p† 169, 25–70, 1 (ad Khuddas, v. 284!: donation of a dåsa is prohibited) allows
accepting slaves given with the terms åråmika, veyyåvaccakara, and
kappiyakåraka!; Vin-vn-p† I 308 ,24– 26 (ad v. 665) allows accepting slaves
given with the terms åråmika and veyyåvaccakara.
work in Buddhist monasteries, and they all seem to have the social
background of a slave. Before we proceed further, we, therefore, have
to take a brief look at the usage of these terms.
2.1.1.1 Óråmika
The term åråmika is only rarely used in the Sutta-pi†aka.44 Most
references are to be found in the Vinaya, where it is used in fiv e
contexts (alone and in compounds). First, most references are found in
the story of King Bimbisåra’s donation of five hundred åråmikas to
Venerable Pilindavaccha, which contributed considerably to the general
acceptance of monastery attendants (åråmika) for the sa"gha.45 Second,
the term appears in the regulations for establishing a monk as a
superintendent of monastery attendants (åråmikapesaka). 46 Third, the
åråmika is mentioned as a person to be asked for permission when a
monk wants to leave a monastery or when a nun wants to enter a
monks’ monastery, in case there is no monk or novice available to be
asked.47 Fourth, we come across the term in passages pondering the
44A II 78,31 (åråmikasamaûddesesu)!; III 109 ,31, 32 (a prophecy that in future

bhikkhus will be mingled with åråmikas and samaûddesas)!; 275 ,16
(determination of an åråmikapesaka)!; III 343, 2 = IV 343 ,25!; Ap I 39,6 !; 191 ,2 !;
205, 7!; 295, 5!; II 409, 14!; 447,24!; Bv 56, 28 (13.14)!; Ja I 251,2, 8!; M II 5, 21f. (see
n. 53).
45This story is told twice in the Vinaya, first as an introductory story to
Nissaggiya 23 Mk, which prescribes that medicines may be stored seven days
at most (Vin III 248,11– 50,29!; BD II 126–131 ), and second in the Mahåvagga
(Vin I 206,34–209 ,35!; BD IV 281ff.). One difference in wording is to be noted
taµ atikkåmayato nissaggiyaµ påcittiyan (Vin III 2 5 1 ,17– 18 ) against
Mahåvagga taµ atikkåmayato yathådhammo kåretabbo (Vin I 209 ,34– 35)!; for
this see von Hinüber 1999, pp. 54ff. The story has been investigated in detail
by Schopen 1994 B , pp. 145–173 , and more broadly by Yamagiwa 2002,
pp.!363–85.
46Vin II 177, 20–23 (BD V 248–249 ) !; 179, 31 (uddåna to the preceding). This is
taken up in the Parivåra, Vin V 204,32– 33!; 205 ,4 (uddåna to the preceding).
47The rule is to be found in the Cullavagga!: Vin II 211 ,24– 25 (BD V 296 f.)!; Vin
II 232 ,8 (BD V 322, uddåna to the preceding). It is hinted at in the word-by-

word commentaries to several Påcittiya rules!: Vin IV 40,20 [Påc 14 Mk] (BD
possibility that monks may want to leave monkhood to become
åråmikas or may ask to be considered as being åråmikas.48 And fifth,
the åråmika is mentioned in the function of a legalizer (kappiyakåraka),
without the word legalizer being used.49
Without exception, references to åråmika in the Vinaya are in its
later layers. 50 A definition of the term is not given anywhere in the text,
II 241)!; Vin IV 41,34–42,1 [Påc 15 Mk] (BD II 244)!; Vin IV 307,29–30 [Påc
51 N] (BD III 341f.).
48Vin III 24 ,27!; 25 ,8 [Pår 1 .8.2 Mk] (BD I 43ff.), word-by-word commentary,
where a monk declares his weakness in making known that he desires the
status of an åråmika or that he wants to be an åråmika!; Vin III 27,7 [Pår 1.8.3
Mk] ( BD I 45f.) disavowing the training in asking to be taken as an åråmika!;
Vin III 92,16 [Pår 4.3 Mk], (BD I 160), word-by-word commentary!: definition
of longing to be purified (visuddhåpekkha) as the wish to become an åråmika,
49All three references of this type belong to the Nissaggiya section!; it is used
twice in the word-by-word commentaries!: Niss 18 Mk (prohibition of the

acceptance of gold and silver!; Vin III 238,15 , BD II 103) and Niss 19 Mk
(engagement in transactions in which gold and silver are involved!; Vin III
240, 17, BD II 108). Once it is used in a Påtimokkha rule itself [Niss 10 Mk],
which, however, on account of its structure seems to be later (see n. 50).
There, an åråmika or a lay follower (upåsaka) should be indicated as a
monk’s personal attendant (veyyåvaccakara) who can function as a legalizer
in order to accept goods given by the king or people in the king’s service for
a certain monk (Vin III 221 ,26 [Niss 10.1.3 Mk], BD II 65f.).
50Those in the Suttavibha"ga (with one exception) come from the introductory
stories and from the word-by-word commentaries. The only reference from a
Påtimokkha rule, i.e., from Niss 10 Mk, may be relatively late. According to
von Hinüber (1999, p. 77), though the group of Nissaggiya prescriptions may
well contain old material, their existence as a separate group probably means
their inclusion was the last step in the development of the Påtimokkha with
150 rules. Thus it may well be that Nissaggiya 10, as we have it now, was
formulated only relatively late. All references in the Mahåvagga belong to the
story which also serves as an introductory story for Niss 23 Mk. According to
Schopen (1994B, pp. 151ff.) this story shows strong signs of a local origina-
tion in Sri Lanka, which implies that in the shape it has in the Theravåda
Vinaya it does not belong to the oldest layers of this text. The references from
the Cullavagga as well as those from the Parivåra refer to the superintendent
of åråmikas, which naturally could have come into being only after the
introduction of åråmikas.
which seems to imply that it was commonly known. From its use in the
Vinaya we can infer that åråmikas could marry, have children, and were
allowed to live together with their families in separate villages
(åråmikagåma) like slaves who also had their own villages. 51 They
could be presented to a single monk by the king. Explicit mention is
made of monks who decided to become åråmikas. The hierarchical
position of an åråmika is between a novice and a lay follower.52 He
may carry out physical or manual work (clearing caves or rock
overhangs). He has some authority with respect to the organization of
the monastery (he is asked for permission to leave [in a monk’s case] or
enter [in a nun’s case] a monastery if no monk and no novice is
present), or he acts as the personal attendant of a monk (veyyåva-
ccakara) in the function of a legalizer (kappiyakåraka). In the
Majjhima-nikåya53 åråmikas are classed with those following the five
rules for lay persons (sikkhåpadas).54
In the commentaries of the fourth or fifth century åråmika is used
as a comprehensive term for workers in a monastery, e.g., as a legalizer
(kappiyakåraka) ; an attendant (veyyåvaccakara) ; a distributor of rice
! !
51Cf. dåsagåmaka (Ap II 538 ,2 = Th¥-a [old edition] 151 ,27 !; [new edition]
148, 8) !; dåsagåmadvåra $ (v.l. dåsakammakaragåmadvåra) dåsagåmavasin
(Ap-a 263,1 –2 = Mp I 179 ,26f. = Spk II 195, 11f. = Th-a III 133,3– 4) and to the
statement that the town Anurådhapura had, among others, fourteen villages
for slaves (Spk II 194,5 f. with Spk-† [CSCD] II 167).
52This becomes evident from the possible order in which one might ask persons
for permission (bhikkhu, såmaêra, åråmika, see n. 47), and by the states a
bhikkhu might wish to revert to!: an upåsaka, åråmika, or såmaêra (see
n.!48).
53M II 5, 21f.!: åråmikabhËtå vå upåsakabhËtå vå pañcasikkhåpade samådåya
vattanti.
54In the Milindapañha (Mil 6,25f.) the god Sakka declares himself an åråmika of
the sa"gha. In Ap I 191, 2, Ap-a 464, 19f., a person declares to have been an
åråmika of the Buddha VessabhË!; in Bv-a 39 ,14 = It-a II 105, 12f. = Mp I
116, 29f., it is stated that Mahåbrahmå may serve as an åråmika or kappiya-
kåraka of the Buddha.
gruel, fruits, or hard food ; as one who clears an area of grass ; 55 as a !
mediator between king and monks ; 56 as one who guards the

!
possessions of the sa"gha;57 or as one who clears and levels the site at
!
the foot of a tree for the inferior tree ascetic, scattering sand on it,
making an enclosure and giving a door ; 58 and as one who has tasks that
!
are similar to those of a novice. 59 According to the Samantapåsådikå,

the monastery provides the åråmikas with food and a ration —
presumably of necessaries60 — equivalent to their work for the
community. For example, if they worked only half a day, the monastery
would not provide supper. They could also work every five days or
every fortnight only, or if the sa"gha had nothing to do for them, work
on their own account without subsidies from the sa"gha. If they earned
money by their own manual labour, they could give that money to the
monastery but obviously were not obliged to do so since they were not
to be spoken to at all in a case where they did not.61 This is remarkable
55See the explanation of how one gives up life as a monk with a synonym of
åråmika, where the synonyms given are kappiyakåraka, veyyåvaccakara,
appaharitakåraka, yågubhåjaka, khajjakabhåjaka, phalabhåjaka (Sp I
253, 29–33). Cf. Gunawardana 1979, p. 98, who adds some further functions
from more recent sources, for example a chief åråmika being responsible for
the decoration in a monastery (Sahassavatthupakaraâ) and åråmikas in
charge of the store of provisions and responsible for the preparation of meals
(S¥ha¬avatthupakaraâ).
56Spk III 23,27!; 24,6 .
57Vism 120 ,30–21,4 = Sp-† II 208, 14–20, where the åråmikas keep the cattle of
the families out of the fields of the monastery and shut off the floodgate so
that people do not obtain water for their fields, which causes trouble for the
monks, who are responsible for the åråmikas’ deeds. This passage is quoted
by Gunawardana 1979, p. 98 (from Sp-†) as a proof for åråmika being also
used as a designation for those who tilled the land of the monastery.
58Vism 74,14–16.
59Sp V 1121 ,22 !; VI 1161 ,23 . In that case åråmika is used in a similar way as
kappiyakåraka.
60For the explanation of nivåpa see Gunawardana 1979, p. 123 .
61See n. 43. Further references!: Spk III 34,3!; 40,3!; Sp II 380, 10ff.!; 474, 7– 11!; III
564, 16.!; 681,19, 21!; 692 ,3 .!; 733, 9!; IV 775 ,8 !; V 1099,26!; Ps I 122 ,23.
insofar as, according to the Hindu law books, slaves and the profit they
produced fell to their owner, which also seems to have been the regular
case in a worldly Buddhist context.62 At least in this respect the attitude
of Buddhist monasteries towards åråmikas differs from the attitude of
the normal population towards slaves. In the Såratthapakåsin¥ (fourth or
fifth century) åråmikas are addressed as lay followers (upåsaka) by
their interlocutors.63 Several donations of slaves to Buddhist monas-
teries and monks are recorded in the Sinhalese chronicles,64 and the
Sinhalese Katikåvatas from the eighteenth century recommend handing
62See Ja I 402 ,30 [no. 97], where a slave girl is beaten by her master and
mistress because she had not given them her wages (dåsim bhatiµ
adadamånaµ)
63Spk III 40,3 !; in Spk III 218 ,6 = Sv II 5 5 2 ,3 2 upåsakas are compared to
åråmikas. In the Vin-vn, v. 1059 åråmikam upåsakaµ, could be a lay

follower who is an åråmika, or it could mean åråmika and upåsaka,
describing two different persons. In other cases åråmika and upåsaka are
listed as separate groups (Ps II 152 ,10f.!: bhikkhu vå såmaêro vå åråmiko vå
vihårasåmiko vå).
64King Sirimeghavaˆâ (362– 409 !?) fixed the revenues of the åråmikas (Mhv
37.63) !; Aggabodhi I (568–601 ) granted one hundred åråmikas to the

Kurundavihåra (Mhv 42.15–16)!; King Silåmeghavaˆâ (617– 26) distributed
the Dami¬as he had overpowered and made slaves (dåsa) to various
monasteries (Mhv 44.70–73)!; King Aggabodhi IV (658–74) placed slaves
(dåsaka) as well as female slaves (dåsi) and åråmikas, which were his own
relatives, at the disposal of the Bhikkhu community (Mhv 46.10 ,14 )!; the
Dami¬a Pottaku††ha, in the service of Aggabodhi IV, assigned villages
together with slaves to the meditation hall (padhånaghara, Mhv 46.19–20)!;
Je††hå, the queen of Aggabodhi IV, granted a hundred åråmikas to the
Je††håråma (Mhv 46.27–28)!; Kassapa IV (896– 913) granted åråmikagåmas
to the hermitages he built (Mhv 52.26) !; Parakkamabåhu I (1153–1186)
assigned a male and a female slave (dåsa, dåsi) to each patient in the hospital
(Mhv 73.34–36)!; Queen Kalyåâvat¥ (thirteenth century) built a monastery
and granted it villages, etc., and slaves (dåsa, Mhv 80.35–36). Her general,
Óyasmanta, created a pariveâ and supplied it with male and female slaves
(dås¥dåsa, Mhv 80.40). King Kittisiriråjas¥ha (1747–1781) assigned relic
villages, etc., with many male and female slaves (dåsidåsa) to the holy Tooth
Relic (Mhv 100.11).
over donations to åråmikas or upåsakas, who are equated with kappiya-
kårakas. 65
2.1.1.2 Kappiyakåraka
The second designation enabling a monastery to accept the donation of
slaves is kappiyakåraka. In the canonical scriptures, this term is
confined to the Vinaya-pi†aka, more precisely to the sixth chapter of the
Mahåvagga on medicines, and to the anåpatti formulas of two Påcittiya
rules, which are even later than the word-by-word commentaries and
the introductory stories in the Suttavibha"ga.66 Obviously, the term
kappiyakåraka was even less common in the canonical texts than the
term åråmika. Likewise, kappiyakåraka is not defined, but used as if its
special meaning was commonly known. In contrast to åråmika, there
exists no prescription in the Vinaya explicitly allowing kappiyakårakas.
The function of a kappiyakåraka was to receive donations of items
forbidden for monks, such as fruit or money, and to make them
acceptable, or to exchange them with acceptable goods. The Vinaya’s
usage renders the impression that kappiyakåraka does not designate a
defined office in the monastery, but rather a function that could be
executed by any trustworthy person who was not an ordained member
of the Buddhist community. Consequently, an åråmika could act as a
kappiyakåraka, too, and according to three passages in the Vinaya, this
is one of the åråmika’s functions although the term kappiyakåraka is
not used there.
The commentarial literature distinguishes ten types of kappiya-
kårakas, depending on whether they are designated or not (niddi††ha / !
aniddi††ha), by whom they are designated, whether in presence or
65K¥rtißr¥råjasiµha-Katikåvata, Ratnapåla 1971, pp. 100, 171 , §!110.

66Vin I 206,12 (twice), BD IV 280 (same context as Vin IV 90,28 [Påc 40 Mk],
BD II 346, anåpatti formula) !; Vin I 211 , 3 7 , BD IV 2 8 8 (in a famine
kappiyakårakas take a greater part)!; Vin I 212 ,7 , 20, 23–25 , BD IV 289
(kappiyakårakas shall legalize fruits)!; 215 ,22 , BD IV 293 (similar to the
preceding)!; 245,2– 3 , BD IV 336 (kappiyakårakas may accept gold)!; Vin III
242, 11 [Niss 20 Mk], BD II 112 (anåpatti formula).
absence of their consignees, etc. 67 The Ka"khåvitaraˆ¥ states that any
individual not ordained in the Buddhist community could serve as a
legalizer.68 From the Samantapåsådikå we know that poor people
decided to become kappiyakårakas in order to earn their living based on
the sa"gha. 69 Therefore, in addition to slaves,70 free persons in need are
expressly mentioned as having become kappiyakårakas. In other cases
lay followers (upåsaka) function as kappiyakårakas.71 Sometimes the
functions of a kappiyakåraka have to be similar to the duties of a novice
(såmaêra) since both are listed alternatively.72 In another case one
who serves someone who is ill (gilånupa††håka) is compared to a
kappiyakåra and a såmaêra.73 In the Katikåvatas kappiyakårakas are
mentioned as those to whom one should hand over improper things.74
2.1.1.3 Veyyåvaccakara
The third designation, veyyåvaccakara, “attendant, steward”, is but
rarely used in the canonical scriptures, and except for two references in
67Sp III 675, 1ff . [Niss 10 Mk] !; Kkh 118, 11 [Niss 10 Mk]. Further references Sp
III 702,3 (son and !/!or brother are rendered into kappiyakårakas#; V 1070,30!;
VI 1228,23!; 1238,6 , 10.
68The Kkh (116, 27–28 ) equates veyyåvaccakara with kappiyakåraka, and
declares that anyone, aside from the five co-religionists (bhikkhu, bhikkhun¥,
sikkhamånå, såmaêra, såmaêr¥), may serve as a kappiyakåraka.
69Sp V 1001 , 18–19 !: duggatamanusså sa!ghaµ nissåya j¥vissåmå ti vihåre
kappiyakårakå honti.
70Buddhadåsa (362–409 ), for instance, granted kappiyakårakas to monks (Mhv
37.173), which indicates that they were not free men.

71Mp II 115, 2 $ Ps I 137, 6 $ Spk I 136 ,27 $ Sv I 236 ,12 $ Ud-a 288 ,18!; Ja IV
408, 16.
72Dhp-a II 182 ,20, 21!; IV 129,6 f.
73Dhp-a II 60,11.
74K¥rtißr¥råjasiµha-Katikåvata I (eighteenth century), Ratnapåla 1971, pp. 100 ,
171, § !110, where kappiyakåraka is equated with åråmika and upåsaka !;

K¥rtißr¥råjasiµha-Katikåvata II (eighteenth century), Ratnapåla 1971, pp.!110,
176, §!15 !; Råjådhiråjasiµha-Katikåvata (eighteenth century), Ratnapåla
1971, pp. 119f., 181 f., §§12, 13, 18.
the Jåtaka and the Apadåna, 75 we only fi nd it in two rules of the
Vinaya-pi†aka, namely in the Påtimokkha rule Nissaggiya 10 Mk
regulating the appointment of an åråmika or a lay follower as a monk’s
veyyåvaccakara, and in the anåpatti formula to Påcittiya 44 N, 76
according to which it is not an offence if a nun cooks for her personal
attendant. The fact that an åråmika or a lay follower may serve as a
monk’s veyyåvaccakara shows that, similar to kappiyakåraka, the term
veyyåvaccakara designates a certain function which may be executed
by different persons. It is obvious from the canonical literature that even
a monk may act as a veyyåvaccakara for other monks.77
Commentaries on the legal literature explain veyyåvaccakara with
the synonyms kappiyakåraka78 or kiccakara. 79
2.1.1.4 Summary
To sum up our findings : all three terms are used mainly in the later parts
!
of the Vinaya and rarely, if at all, in the Sutta-pi†aka. This implies that
they were alien to the early Buddhist texts. Óråmika is the technical
term for people belonging to, and working for, Buddhist monasteries.
Two types of åråmikas may be distinguished with regard to their social
status before they became åråmikas : (1) dependent persons, i.e., slaves,
!
and (2) free men. Obviously, in order to differentiate these two types of
åråmikas, the Samantapåsådikå introduces the term åråmikadåsa, a
slave who is an åråmika, to designate the first group. The terms
kappiyakåraka and veyyåvaccakara describe functions that could be
executed by åråmikas, but also by lay followers or other persons.
75Ja II 334 ,8 !; Ap I 138,8.

76Vin III 221 ,25– 28, 30, 32 (Påtimokkha rule)!; 222,23, 25, 27, 29 [Niss 10 Mk], BD
II 65f.!; and in the anåpatti formula to Vin IV 301 , 4 [Påc 44 N], BD III 329
(here the meaning is misunderstood by I.B. Horner).
77See the example of Dabba Mallaputta, who did the sa"gha’s work
(veyyåvaccaµ karoti#; DPPN s.v. Dabba Mallaputta), and the example of a

young bhikkhu who did not do the work of other bhikkhus (S II 277, 13!; Ee
veyyåccaµ).
78Kkh 116 ,23 [Niss 10 Mk]!; Sp III 672 ,22– 23 [Niss 10 Mk].
79Sp III 672 ,22– 23 [Niss 10 Mk].
Therefore, an åråmika could be a kappiyakåraka or a veyyåvaccakara,
and vice versa, but a kappiyakåraka and a veyyåvaccakara were not
necessarily åråmikas, at least not of the first type.
2.1.2 The lower ordination of slaves
The Vinaya rules that one should not confer lower ordination (pabbajjå)
on slaves.80 Commenting on that rule, the Samantapåsådikå — in
accordance with defi nitions given in the canonical writings —
distinguishes four types of slaves : ( 1) one born as a slave, (2) one
!
bought for money, (3) a captive turned into a slave, and (4) a person
gone into slavery on his own accord.81 The first two types of slaves may
receive lower ordination only after they are freed. 82 The third may not
receive lower ordination as long as he is held captive, but may be
ordained as a novice if he manages to escape or is released in the course
of a general amnesty. 83 The fourth may not be ordained. 84 Even a slave
without an owner had to be formally released before he could be
ordained.85 And if a slave who was unaware of his status had been
80Vin I 76 ,26– 27 !: na bhikkhave dåso pabbåjetabbo. yo pabbåjeyya, åpatti

dukka†asså ti. “Monks, a slave should not be let go forth. Whoever should let
[one such] go forth, there is an offence of wrong-doing.” (Translation by I.B.
Horner, BD IV 95f.)
81Sp V 1000 ,19– 2 0 ! : na bhikkhave dåso ti ettha cattåro dåså antojåto
dhanakk¥to karamarån¥to såmaµ dåsabyaµ upagato ti.

82Sp V 1000,23– 25 !: ete dve pi na pabbåjetabbå, pabbåjentena tattha tattha
cårittavasena adåsaµ katvå pabbåjetabbå. Cf. Dhp-a I 15,17f.!; Th-a I 73,13.

83Sp V 1000, 25–1001, 3.
84Sp V 1001 ,3– 6 !: såmaµ dåsabyaµ upagato (Sp 1000,20) nåma j¥vitahetu vå
årakkhahetu vå ahaµ te dåso ti sayam eva dåsabhåvaµ upagato . råjËnaµ

hatthi-assa-gomah¥sa-gopakådayo viya tådiso dåso na pabbajetabbo. “One
gone into slavery of his own accord means one who, for the sake of
livelihood or for the sake of protection, went himself into the state of a slave
[with the words] ‘I am your slave’. Like watchmen of kings’ elephants,
horses, cows, buffaloes, etc., is such a slave; they may not let him go forth.”
85Sp V 1001 ,27– 28 !: nissåmikadåso hoti so pi bhujisso kato va pabbåjetabbo.
“[If] one is an unowned slave, that one too may be ordained as a novice, only
having [first] been made a free man.”
ordained as a novice or as a monk and learned about his being a slave
only after the event, he had to be released retrospectively.86
As is obvious from Samantapåsådikå, Såratthad¥pan¥, Vimati-
vinodan¥-†¥kå and Påcityådiyojanå, persons who went into slavery held a
slave certificate 87 recording their name, and perhaps their status, their
owner, and possibly the place and time of their transfer.88 Practices of
86Sp V 1001,28– 29!: ajånanto pabbåjetvå vå upasampådetvå vå pacchå jånåti,

bhujissaµ kåtuµ eva va††ati. “[If] one not knowing [about his slave status]
learns [about it] after they have ordained him as a novice or as a monk, it is
allowed in fact to make him a free man.”
87Påc-y 244 ,12 !; Sp-† III 243 ,12, 14 !; Vmv II 111, 5 (dåsipaˆâ) #; Sp V 1001,9
(paˆâ). Paˆâ with forms of åropeti (not used in the canon but only in post-
canonical literature) for the most part means document (only once is it used
for letter, Ja VI 369 ,13– 14), and, depending on the context, stands for a slave
letter, a promissory note (also called iâpaˆâ!; Ja I 227 ,4 !; 230 ,2 !; Dhp-a II
128, 22!; 129, 19!; 133, 1!; 134 ,7 !; 135, 1– 2 !; III 12 ,19f.) , or an attestation of the
allotment of goods (Sp 387, 24 = Pålim 431, 12!; with Sp-† II 167, 12–13!; Vmv I
204, 10–11!; Pålim-n† II 328 ,6 –8 ). Óropeti in those cases does not mean “to
send”, as indicated by CPD (s.v. åropeti), as an idiomatic use of paˆâµ
åropeti, but “to post (up)” if it is used with the loc., and “to make out” if it is
used with the acc. Compare also the younger MËlasarvåstivåda tradition
where in Guâprabha’s VinayasËtra, the recording in a promissory note is
expressed by åropya patre (see Schopen 1994 A , p. 538). The compound
paˆˆåropana is used in the same meaning in the present context and in two
further places, Sv-p† I 423 ,16 !: sakkhikaraâpaˆˆåropanåni va""hiyå saha
vinå vå puna gahetukåmassa and, Sv-n†, CSCD, II p. 305 !: sakkhikaraâ-
paˆˆåropananibandhanaµ va""hiyå.
88Sp-† III 243,13!: sace sayam eva paˆâµ åropenti, na va††at¥ ti (Sp 1001,9 ) tå
bhujissitthiyo mayam pi dåsiyo homå ti sayam eva dåsipaˆâµ likhåpenti, na

va††ati. “If they themselves make out a certificate, it is not allowed [to
ordain their sons !: if] these free women themselves cause a slave certificate to
be written [with the words], ‘We too are female slaves’, it is not allowed [to
ordain their sons].” Vmv II 111, 3– 5 = Pålim-n† I 233, 6– 10!: sayam eva paˆâµ
åropenti, na va††at¥ ti (Sp 1001,9 ) tå bhujissitthiyo mayam pi vaˆâdåsiyo
homå ti attano rakkhaâtthåya sayam eva råjËnaµ dåsipaˆê attano nåmaµ
likhåpenti. “If they themselves make out a certificate, it is not allowed [to
ordain their sons!: if] these free women themselves for their own protection
cause their own name to be written in a slave certificate of kings [with the
releasing slaves varied from region to region. One way was to redeem
the slave by reimbursing his owner, possibly accompanied by a ritual
burning of the slave certificate. 89 Another method was to sprinkle
buttermilk on the slave’s head, or to wash (soak ?) it with buttermilk.
!
We do not know for certain whether in that case the slaves had to be
redeemed first. In any case, the respective references do not mention a
payment, which may be taken as an indication that the ablution with
buttermilk itself effected the release.90 The Mahåpaccar¥, one of the
early commentaries from around the first century B. C., already refers to
this last method. It is repeatedly mentioned in the commentaries of the
fourth or fifth century, and still known at the time of the †¥kås. 91
words], ‘We too are courtesans (lit. slaves of beauty)’, it is not allowed [to
ordain their sons].”
89Sv I 216, 20 = Ps II 321 ,5 –7 !: yathå puna (Sv B e pana) dåso kañcid eva mittam
upanissåya såmikånaµ dhanaµ datvå attånaµ bhujissaµ katvå tato pa††håya

yaµ icchati, taµ kareyya!; Vmv II 110 ,22–11,1 !: dåsacårittaµ åropetvå k¥to ti
(Sp 1000,23) iminå dåsabhåvaparimocanatthåya k¥takaµ nivatteti. tådiso hi
dhanakk¥to pi adåso eva. tattha tattha cårittavasenå ti (Sp 1000,24) tasmiµ
tasmiµ janapade dåsapaˆâjjhåpanådinå adåsakaraâniyåmena. Unowned
slaves (nissåmikadåsa) were allowed to free themselves, Vjb 424 ,10– 11!:
nissåmikaµ dåsaµ attanåpi bhujissaµ kåtuµ labhati. Sp-† III 243 ,19– 21!:
nissåmikadåso (Sp V 1001,27) nåma yassa såmikå saputtadårådayo (Pålim-
p† adds c a) matå honti, na koci tassa pariggåhako, so pi pabbåjetuµ na
va††ati, taµ pana attanåpi bhujissaµ kåtuµ va††ati.
90Sv I 2 6 6 ,24f.!: dh¥taraµ adås¥ti s¥saµ dhovitvå adåsaµ bhujissaµ katvå
dh¥taraµ adåsi. “He gave [him his] daughter [as a wife]!: Having washed
his head, [thus] having made [him] a non-slave (=) a free man, he gave [his]
daughter [to him].” Cf. Ap-a 263,5f . = Mp I 179 ,26f. = Spk II195,15f.= Th-a III
133, 7f .!: sace tumhesu ekekaµ bhujissaµ karoma, vassasatam pi na ppahoti.
tumh’ eva tumhåkaµ s¥saµ dhovitvå bhujisså hutvå j¥vathå ti. “If we make
each one among you a free man, even a hundred years will not suffice.
Having washed your head you indeed shall live as free men.” See also Vibh-
m† (CSCD) 182.
91Sp-† III 243,14– 17 = Pålim-n† I 233, 23–27!: takkaµ s¥se åsittakasadiså va hont¥
ti (Sp V 1001,14– 15) yathå adåse karontå takkena s¥saµ dhovitvå adåsaµ
karonti, evaµ åråmikavacanena dinnattå adåså va te ti adhippåyo.
takkåsiñcanaµ pana s¥ha¬ad¥pe cårittan ti vadanti. “They in fact resemble
[persons] on [whose] head buttermilk is sprinkled!: as [those] who make
According to the explanations of Dhammasiri’s Gaˆ†hipada and
Såriputta’s Sårattha d¥pan¥, this method was practised in Sri Lanka, 92
while the Vimativinodan¥-†¥kå declares that it was a usage in some
countries without specifying them.93
Among the various groups of slaves mentioned in the Samanta-
påsådikå, we find the speci fic group of åråmikadåsas, slaves who are
åråmikas. They represent the first of the two groups of åråmikas defined
before, i.e., those who are unfree. If these are given to the monastery
(v i h å r a ) by a king, they, according to the statement of the
Samantapåsådikå, may be ordained as novices only after their release. 94
Whether this is different if the donor was a commoner, we do not know.
In any case, it seems to be irrelevant which of the four categories of
slaves these åråmikadåsas belonged to.95
[slaves] into non-slaves, make [a slave] into a non-slave by washing his head
with buttermilk, so they, because of [their] having been given with the
designation åråmika, [are made] indeed non-slaves. [That is the] intention.
‘The sprinkling of buttermilk, however, is a usage in the S¥ha¬a island,’ they
say.” Vmv II 111 ,11–14 !: takkaµ s¥se åsittakasadiså va hont¥ ti kesuci
janapadesu adåse karontå takkaµ s¥se åsiñcanti, tena kira te adåså honti,
evam idam pi åråmikavacanena dånam p¥ti adhippåyo. “They in fact
resemble [persons] on [whose] head buttermilk is sprinkled!: in some
regions [those] who make [slaves] into non-slaves sprinkle buttermilk on
[their] head !; therewith, as is well known, they become non-slaves. In this
way also that donation with the statement åråmika is intended.” Påc-y
243,20–21!: åråmikaµ demå ti vacanaµ dåsånaµ bhujissavacanan ti vuttaµ
hoti. “It is said that the statement ‘we give an åråmika’ for slaves is the
statement [that one is] a free man.”
92Vjb 424 ,9 !: takkåsiñcanaµ S¥ha¬ad¥pe cårittaµ. Sp-† III 243,17, see n. 91.
93Vmv II 111,11– 13, see n. 91.
94Sp V 1001,11– 12 !: vihåresu råjËhi åråmikadåså nåma dinnå honti, te pi
pabbåjetuµ na va††ati. bhujisse katvå pana pabbåjetuµ va††ati. “Slaves who

belong to the åråma are given to the vihåras by kings !; these too may not be
ordained as novices. But having made them free men, [they] may be ordained
as novices.”
95Probably all four types of slaves were the property of kings.
Furthermore, if a monk receives a slave from his relatives or his
servants with the request to ordain him as a novice so that he may do
the monk’s work (veyyåvacca), or if the monk’s own slave is considered
for such a promotion, the Samantapåsådikå states that he may only be
ordained as a novice after he has been released.96 Thus in both cases —
(1) donation of slaves by a king to the sa"gha and (2) donation of a
slave by private persons to a monk — the slaves have to be released
first.
In this context, however, the Samantapåsådikå hands down a
quotation from the Mahåpaccar¥ (c. first century B.C.). There it is stated
that born and bought slaves are given to the community of monks with
the words “we give åråmikas”, that the status of these individuals then
resembles that of persons whose heads are sprinkled with buttermilk,
and that they are entitled to receive the lower ordination. 97
While the Samantapåsådikå, according to the initial statement,
would admit the ordination of the first two types of slaves only after
their release, the Mahåpaccar¥ attaches no further condition to their
lower ordination except that they are to be given to the community of
monks with the designation åråmika. The donor is not mentioned in this
case. Thus his identity, be it king or commoner, seems to be irrelevant.
If one extends that statement to cover born and bought slaves given by a
king, the Mahåpaccar¥ is in obvious disagreement with the Samanta-
påsådikå. However that may be, from the statement of the Mahåpaccar¥
96Sp V 1001,21– 23 !: bhikkhussa ñåtakå vå upa††håkå vå dåsaµ denti imaµ

pabbåjetha, tumhåkaµ veyyåvaccaµ karissat¥ti attano vå (Sp Ee v a) assa
dåso atthi, bhujisso kato ’va pabbåjetabbo. “[If] a monk’s relatives or
servants donate a slave [to him with the words!:] ‘Ordain that one as a novice,
he will do your work’, or [if] he himself (i.e., the monk) owns a slave, this
one may be ordained as a novice only after he has been made a free man.”
97Sp V 1001 ,13– 1 5 !: Mahåpaccariyaµ antojåtadhanakk¥take ånetvå bhikkhu-
sa!ghassa ‘åråmike demå’ ti denti. takkaµ s¥se åsittakasadiså ’va honti.

pabbåjetuµ va††at¥ ti vuttaµ. “In the Mahåpaccar¥ it is said, ‘They bring
persons born [as slaves] and [those] bought for money [and] give [them] to
the community of monks [with the words!:] “We give åråmikas”. [These]
become indeed similar to those on whose head buttermilk is sprinkled.’ ”
it follows that the statement “We give åråmikas” changes the social
status of the slaves and assimilates their status to that of free men.
The position of the Mahåpaccar¥, in turn, is contested by the
Kurund¥, another of the early commentaries quoted in the Samanta-
påsådikå. Without specifying the individuals given to the sa"gha, the
Kurund¥ agrees with the Mahåpaccar¥ as to the accompanying
designation (“We give an åråmika”), but not with regard to their
consequent entitlement to lower ordination.98 This document presents
not only a conflict of views with regard to the social status of åråmikas
given to the sa"gha, but also a difference of opinion concerning their
entitlement to ordination as novices. It shows us as well that this
conflict has a very long history, reaching back at least to the first
century B.C.
As for the Samantapåsådikå, there are indications that it agrees with
the Kurund¥ : firstly, because it expresses the same opinion with respect
!
to åråmikas given by a king ; and secondly, because it quotes the

!
Kurund¥ after the Mahåpaccar¥, which is a sign of acceptance.99

The next class of commentaries, the gaˆ†hipadas,100 contain
various statements on åråmikas. The first, Dhammasiri’s Gaˆ†hipada, is
undated and only survived in the passages quoted in the Vajirabuddhi-
98Sp V 1001,15– 17!: Kurundiyam pana ‘åråmikaµ demå’ ti kappiyavohårena

denti, yena kenaci vohårena dinno hotu, n’eva pabbåjetabbo ti vuttaµ. “But
in the Kurund¥ it is said, ‘They give with the [legally] acceptable designation
“we give åråmikas” !; with whatever designation one is given, he is by no
means to be ordained as a novice.’ ”
99Sp II 300 ,8 –9 !; cf. von Hinüber 1996, p. 107.
100Gaˆ†hipadavivaraâ or -vaˆânå, Gaˆ†hipadatthanicchaya, Gaˆ†hipad’-
atthavaˆânå, etc., or merely gaˆ†hipada is the name of a class of

commentaries commenting on words of the canonical texts and their
respective a††hakathås. The gaˆ†hipadas originated after the a††hakathå
literature and before the subcommentaries (†¥kå). They were written in Påli,
Sinhalese, and maybe other languages. Sometimes we only have the name of
the author to identify a certain gaˆ†hipada !; sometimes these gaˆ†hipadas
have names, for example Mahågaˆ†hipada. For further information, see Sv-p†
I xxxiff.
†¥kå, which suggests that it must have been written between the fifth and
twelfth centuries. Dhammasiri regards åråmikas as neither slaves nor
free men, 101 but nonetheless supports their ordination as novices. 102
This implies that the slave, with his presentation to the sa"gha as an
åråmika, achieves a social status between a slave and a free man, which
in turn enables his promotion to the status of a novice. Here
Dhammasiri clearly sides with the tradition of the Mahåpaccar¥ against
that of the Kurund¥ and the Samantapåsådikå.
Vajirabuddhi’s Anugaˆ†hipada, another undated commentary that
only survived in quotations by the Vajirabuddhi-†¥kå, was written after
Dhammasiri’s Gaˆ†hipada. Here the ordination of an åråmika is made
conditional upon the compensation of the communitiy with another
åråmika.103 Two interpretations are possible in that case : ( 1) The !
Anugaˆ†hipada considers the status of åråmikas as similar to that of free

men, and its primary concern is the question of compensation in order
to prevent the sa"gha from loss, or (2) if the åråmika is regarded as a
slave, his status can be transferred to the person presented as a
substitute. In the first case, the Anugaˆ†hipada would side with the
Mahåpaccar¥, in the second, with the Kurund¥.104
101Vjb 424 ,8 –9 !: åråmiko ca ‘n’ eva dåso na bhujisso’ ti vattabbato na då so ti

likhitaµ. “And an åråmika is not a slave, because it must be said that he is
neither a slave nor a free man, [thus] it is written [in Dhammasiri’s
Gaˆ†hipada].”
102Vjb 424, 10!: te ca pabbåjetabbå sa!ghassåråmikattå. “And these (referring
to the Mahåpaccar¥ quotation in Sp V 1001,14– 15, see n. 97) may be ordained

as novices, because [they] are åråmikas of the community.” This passage is
part of a larger quotation from Dhammasiri’s Gaˆ†hipada which refers to
several aspects of slaves’ ordination, starting at Vjb 424,9 and ending at Vjb
424, 12 with ti likhitaµ.
103Vjb 424, 5– 6!: åråmikaµ ce pabbåjetukåmo, aññam ekaµ datvå pabbåje-
tabban ti vuttaµ. “If one wishes to ordain an åråmika as a novice, the

[åråmika] may be ordained as a novice if another one is given for the one [to
be ordained].”
104Different from the Vimativinodan¥-†¥kå, which explicitly demands
redemption of value plus profit (see below), the Anugaˆ†hipada only provides
for the payment of the value, i.e., replacement of one åråmika by another one.
Coming to the †¥kås, the independent evidence provided by the
Vajirabuddhi-†¥kå (before the twelfth century A.D.) comes down to one
sentence that is not part of a quotation from one of the gaˆ†hipadas.
And here the Vajirabuddhi-†¥kå explains the position of the
Mahåpaccar¥, without, however, explicitly adopting it.105 In any case, I
find it quite remarkable that neither the Vajirabuddhi-†¥kå nor one of the
gaˆ†hipadas quoted in it shows any inclination to consider the contrary
position of the Kurund¥, although it must have been known to them.
This may be taken as an indication that the gaˆ†hipadas and the
Vajirabuddhi-†¥kå are in accord with the Mahåpaccar¥, against the
Kurund¥ and the Samantapåsådikå.
Såratthad¥pan¥ and Vimativinodan¥-†¥kå confirm the statement of
the Kurund¥, explaining that åråmikas may not be ordained as novices
because they are åråmikadåsas of the sa"gha. 106 Nonetheless, the
Såratthad¥pan¥ also comments on the Mahåpaccar¥ and it seems that it
does not take sides with any one of them.107 The Vimativinodan¥-†¥kå,
on the other hand, annotates the statement of the Samantapåsådikå that
105Vjb 4 2 4 ,6 –8 !: Mahåpaccarivådassa ayam idha adhippåyo!: “bhikkhu-

sa!ghassa åråmike demå” ti (Sp 1001 , 13–14) dinnattå na te tesaµ dåså.
“This is here the intention of the doctrine of the Mahåpaccar¥!: ‘because [they]
are given [with the words,] “We give åråmikas to the community of
monks”, they are not their (i.e., the monks’) slaves.’ ” This sentence probably
is a statement of the Vajirabuddhi-†¥kå!; however, it cannot be completely
excluded that it may be part of the quotation from Dhammasiri’s Gaˆ†hipada,
ending in Vjb 424 ,9 and starting here (Vjb 424,6) or in 424 ,7 .
106Sp-† III 243, 17–18 = Pålim-n† I 233 ,27–34,1 !: n’ eva pabbåjetabbo ti vuttan ti
(Sp V 1001,17– 18) kappiyavacanena dinne pi sa!ghassa åråmikadåsattå evaµ

vuttaµ. “It is said [in the Kurund¥,] that [someone given as an åråmika]
may by no means be ordained as a novice!: This is said in that way because
one, even if given with the legal statement [that he is given as an åråmika], is
a slave who is an åråmika of the community.” Vmv II 111 ,14– 15 = Pålim-n† I
234, 5– 7!: tathå dinne pi sa!ghassa åråmikadåso evå ti n’ eva pabbåjetabbo ti
(Sp V 1001,17) vuttaµ. “Even when given in that way he is only a slave who
is an åråmika of the community!; [therefore] it is said [in the Kurund¥ !:] ‘He is
by no means to be ordained.’ ”
107See n. 91.
åråmikas given to the sa"gha by a king may not be ordained. It explains
that the community is entitled to the value, plus profit, of an åråmika to
be redeemed with the aim of ordaining him as a novice.108 This makes
it perfectly clear that the Vimativinodan¥-†¥kå regards the åråmika
donated to the Buddhist community by a king as a slave. Furthermore,
from the way in which the Vimativinodan¥-†¥kå comments on the
statements of Mahåpaccar¥ and Kurund¥, it follows that it shares the
opinion of the Kurund¥,109 which is confi rmed by its position with
respect to the lower ordination of children of åråmikas (see below
2.1.3). The Påcityådiyojanå from nineteenth-century Burma adopts the
view of the Kurund¥.110
2.1.3 The lower ordination of children of åråmikas
Another question connected with the åråmikas is whether children of
åråmikas may be ordained as novices or not.
With regard to children of slaves, the Samantapåsådikå points out
that they are to be counted among the first of four categories of slaves,
namely those born [as slaves], or slaves by birth (antojåta, jåtidåsa).
Furthermore, the Samantapåsådikå states that if the mother or both
parents are slaves, children do not qualify for ordination as novices.
However, if the father is a slave and the mother is free, their children
108Vmv II 111 ,9 –11 = Pålim-n† I 233 ,12– 15!: bhujisse pana katvå (Pålim-n† katvå
pana) pabbåjetuµ va††at¥ ti (Sp V 1001 ,12– 13) yassa vihårassa te åråmikå
dinnå, tasmiµ vihåre sa!ghaµ ñåpetvå phåtikammena dhanåni datvå (Pålim-
n† dhanådiµ katvå) bhujisse katvå pabbåjetuµ va††ati. “Having made them,
however, free men, it is allowed to ordain [them] as novices!: having made
[the åråmikas] free men, by informing the community in that monastery to
which they are given as åråmikas [and] by giving the value [of the åråmika]
plus a profit [to the community], it is allowed to ordain [them] as novices.”
109Vmv II 111,11ff. (see n. 91), and 111 ,14– 15 (see n. 106).
110Påc-y 244,23–25!: dv¥su A††hakathåvådesu Kurundivådassa pacchå vuttattå so
yeva pamåân ti da††habbaµ. “It is to be shown that, because of the two

a††hakathå doctrines, the doctrine of the Kurund¥ is taught later!; only this one
is authoritative.”
are free, too, and therefore qualified. 111 This shows that children inherit
their status as slaves from the mother, not the father, which is in
agreement with Hindu tradition.
The majority of pertinent references is to the masculine form,
åråmika. As for its less common feminine counterparts, the Vinaya has
åråmikin¥ in the story of the donation of five hundered åråmikas to
Venerable Pilindavaccha,112 while åråmikå is documented in a passage
of the Vajirabuddhi-†¥kå (357, 8) introducing us to the niceties of
politically correct Buddhist speech. So the phrase “This is our male or
female slave” ( amhåkaµ eso dåso, dås¥) is prohibited, but it is perfectly
acceptable to say, “This is our male or female åråmika” (ayaµ
amhåkaµ åråmiko, åråmikå).
In the context of feminine terms, mention should also be made of
devadås¥s and the question of whether their children are qualifi ed to be
ordained as novices. Dhammasiri’s Gaˆ†hipada allows their ordina-
tion,113 and the same holds true for the three Sinhalese Gaˆ†hipadas
quoted in the Såratthad¥pan¥. 114 Only the Vimativinodan¥-†¥kå declares
that they are not qualified because even devadåsas are only slaves.115
Apart from these statements, only three more references for the
word devadås¥ / å are found in the Påli texts. In Dhammapåla’s
! !
111Sp V 1001,19– 21!: yassa måtåpitaro dåså, måtå eva vå dås¥, pitå adåso, taµ
pabbåjetuµ na va††ati. See also Sp V 1001, n. 9!: Bp inserts yassa pana måtå
adås¥ pitå dåso, taµ pabbåjetuµ va††ati.
112Vin I 208 ,10, 12, 17, 19 (BD IV 281ff.) = III 249 ,28, 30, 35, 37 (BD II 128ff.).
113Vjb 424, 5 = Pålim-n† I 233 ,15!: devadåsiputte va††at¥ti likhitaµ. “It is allowed
[to ordain] the sons of devadås¥s [as a novice !; this] is written [in
Dhammasiri’s Gaˆ†hipada].”
114Sp-† III 243, 22 = Pålim-n† I 234, 20!: devadåsiputtaµ pabbåjetuµ va††at¥ti t¥su
Gaˆ†hipadesu vuttaµ. “It is allowed to ordain the son of a devadås¥ as a

novice!; [this] is said in the three Gaˆ†hipadas.”
115Vmv II 111, 20 = Pålim-n† I 234 ,13!: devadåsåpi dåså eva. te hi katthaci dese
råjadåså honti, katthaci vihåradåså, tasmå pabbåjetuµ na va††ati. “Even

devadåsas [are] only slaves. For in one region they are slaves of kings, in
another [region] they are slaves of monasteries!; therefore, it is not allowed to
ordain [them] as novices.”
Suma"galavilåsin¥-poråâ†¥kå (Sv-p† I 477,5), and in the Suma"gala-
vilåsin¥-nava†¥kå (Sv-n†, CSCD II, p. 374 ; eighteenth century) devadås¥
!
is used to explain yakkhadås¥, “slave of a demon”, while the Nirutti-

d¥pan¥ (CSCD, p.!229 ; twentieth century) mentions devadås¥putta, “son
!
of a female slave of a deva (god, king, temple ?)”, and råjadås¥putta,

!
“son of a female slave of a king”, to exemplify a certain type of

compound.
The Vimativinodan¥-†¥kå explains that in some regions the word
devadåså means “slaves of a king”, and in other regions “slaves of a
monastery” (vihåra, see n. 115). Devadåså of Vmv might be a mascu-
line or feminine (?) pl. (though the regular feminine sg. form should end
in -¥).
Let us briefly return to the usage of the term devadås¥ in the gaˆ†hi-
padas. Assuming that it here designates female slaves of a king, we may
infer that their children had a special status exempting them from the
general prohibition against ordaining children of female slaves, which
would run against the intention of the Vinaya rule.116
However, if devadås¥ designates the female slave of a Buddhist
monastery, then these females must be åråmikås because otherwise the
monastery would not have been able to accept them. In that case
devadås¥ would be synonym ous with the term åråmikin¥ documented in
the Vinaya story of the gift of the five hundred åråmikås by King
Bimbisåra. As it happens, the story of their donation is also handed
down in the Tibetan version of the MËlasarvåstivåda Vinaya. The
Tibetan word used there, however, lha-’ba!s,117 corresponds to Skt
devadåsa, rather than to åråmika. 118 Since the context of the story is the
same, this may point to the synonymous use of devadås¥ and åråmikin¥,
116The prohibition to ordain slaves had the aim of not interfering with the rights
of the proper owner of the respective slave. Thus it would not make sense to
exempt the slaves of kings from this rule.
117Jäschke s.v. “slaves belonging to a temple”.
118Schopen 1994B , p. 158 (equates lha-’ba! with kalpikåra), 164 (here he
refers to devadåsa as the corresponding term).

with preferences probably varying according to region or tradition. All
four Gaˆ†hipadas — of which at least three, but probably all four, are of
Sri Lankan origin — would then advocate the legitimacy of ordaining
children of female åråmikås. Only one of them, Dhammasiri’s
Gaˆ†hipada, explicitly treats both the ordination of åråmikas and that of
their children. We can, however, safely assume that the three Sinhalese
Gaˆ†hipadas must have held the same view as Dhammasiri with respect
to the ordination of an åråmika, because otherwise, their attitude
towards the åråmika’s children would be difficult to account for.
Finally, the South Indian Vimativinodan¥-†¥kå would prohibit the
ordination of the children of åråmikas. From this it would result that the
Vimativinodan¥-†¥kå considers åråmikas, whether given by a king or by
someone else, as slaves. In that way, the Vimativinodan¥-†¥kå would
proceed with the tradition of the Kurund¥ and the Samantapåsådikå.
In summary, we can say that one branch of the Theravåda tradition,
represented at least in Sri Lanka, and stretching at least from the first
century B.C. to the time of the gaˆ†hipadas (sometime before the
twelfth century A.D.), excepts slaves belonging to a Buddhist monastery
(åråmikadåsa), as well as their children, from the general rule prohibit-
ing the pabbajjå of slaves. For this purpose the rules are stretched in
order to exclude åråmikas from the Vinaya’s definition of slaves. The
other branch of the Theravåda tradition, which can be traced from the
first century B. C. to the nineteenth century A.D., also represented in Sri
Lanka, but in later times adopted by the South Indian Vimativinodan¥-
†¥kå and by the Burmese Påcityådiyojanå, stuck to the Vinaya rule with-
out concessions regarding the status of slaves in Buddhist monasteries.
pabbajjå prohibited pabbajjå allowed for uncertain

for åråmikadåsas åråmikadåsas
Kurund¥ (first Mahåpaccar¥ (first
century B.C. or century B.C. or
earlier) earlier)
Dhammasiri’s Vajirabuddhi’s
Gaˆ†hipada (after the Anugaˆ†hipada
Samantapåsådikå, (after Dhammasiri’s
before Vajirabuddhi’s Gaˆ†hipada, before
Anugaˆ†hipada) the Vajirabuddhi-†¥kå)
Three Sinhalese Vajirabuddhi’s
Gaˆ†hipadas (Mahå-, Vajirabuddhi-†¥kå
Majjhima-, (before the twelfth
CË¬agaˆ†hipada!; century)
before the twelfth
century)
Såriputta’s
Såratthad¥pan¥
(twelfth century)
Vimativinodan¥-†¥kå
(twelfth!/!thirteenth
centuries)
Påcityådiyojanå
(nineteenth century)
2.1.4 The higher ordination of åråmikas

Let us finish this example with one last remark. Among those authoriz-
ing the pabbajjå for åråmikas, Dhammasiri’s Gaˆ†hipada explains that
they are neither slaves nor free men. This seems to imply that even
Dhammasiri excluded åråmikas from higher ordination because the
candidate for higher ordination has to be a free man. However, we have
to reckon with the possibility that, by being ordained as novices,
åråmikas lose their former status and therefore qualify for higher
ordination, too.
2.2 The tic¥vara
The second method for stretching the rules and getting away with it is
not to apply the wording originally provided for the respective case, but
to resort to another wording that allows a certain latitude.
As is well known, in the early days of Buddhism, monks had to
content themselves with robes made from rags from a dust heap
(paµsukËla). Very soon, however, they were also allowed to wear robes
donated by householders.119 The robe (c¥vara) every monk is obliged to
wear from the time of his higher ordination onward consists of the inner
garment (antaravåsaka), the upper garment (uttaråsa!ga) and the outer
cloak (sa!ghå†i).120 The inner garment covers the navel and the knees
and is fixed by a waistband.121 The upper garment reaches from the
neck to the ankles, thus covering the inner garment. The outer cloak had
the size of the upper garment and is made of two layers of fabric.122 A
monk was allowed to own no more than one set of three robes
119Vin I 280 ,35ff. (BD IV 397ff.).

120Vin I 289, 1– 3 !: anujånåmi bhikkhave tic¥varaµ diguâµ saµghå†iµ
ekacciyaµ uttaråsa!gaµ ekacciyaµ antaravåsakan ti. “I allow you, monks,
three robes!: a double outer cloak, a single upper robe, a single inner robe”
(B D IV 411). If the clothes were worn thin the antaravåsaka and the
uttaråsa!ga were allowed to be double, the sa!ghå†i fourfold, Vin I 290, 13–14
(BD IV 413).
121Vin II 135 ,34–36,5 (BD V 188f.).
122Following Sp III 643, 3– 8 = Kkh 94,18– 2 0 sa!ghå†i and uttaråsa!ga are,
according to the smallest size, in length five mu††hi (1.8 metres), in breadth
three mu††hi (1.08 metres)!; the antaravåsaka is in length the same, in breadth
two mu††hi (0.72 metres). For mu††hi as a measure of length, see Kieffer-Pülz
1993, p. 182, n. 46. The upper limit for all robes is given by the size of a
sugatac¥vara (nine vidatthi in length [1.98 metres] and six vidatthi in breadth
[1.32 metres]!; Vin IV 173, 28–29) which they must not exceed.
Six kinds of material were allowed!: Vin I 281 ,34– 36 (BD IV 398)!: anujånåmi
bhikkhave cha c¥varåni khomaµ kappåsikaµ koseyyaµ kambalaµ såâµ
bha!gan ti. “Monks, I allow six [kinds of] robe materials!: linen, cotton, silk,
wool, coarse hempen cloth, canvas.”
(tic¥vara).123 Any item in excess was regarded as an extra robe (ati-
rekac¥vara), and had to be assigned (vikappeti) to someone else after ten
days at the latest.124
Before a monk could use a cloth, he had to take formal possession
of it (adhiti††hati).125 This holds true for all nine clothes which serve as
requisites of a monk. These are (1–3) the three robes ( tic¥vara), (4) the
cloth to sit upon (nis¥dana), (5) a sheet (paccattharaâ), (6) a cloth for
wiping the face (mukhapuñchanaco¬a), (7) a requisite cloth (pari-
kkhåraco¬a), (8) the cloth for the rains (vassikaså†ikå), and (9) the itch-
cloth (kaˆ"upa†icchåd¥). Only two of them may be assigned (vikappeti)
to others after use, i.e., the cloth for the rains and the itch-cloth.126 For
most items a certain size and number are prescribed. 127
123Vin I 287 ,31–89,3 (BD IV 409 f.). The stories told in the Vinaya about monks
who entered a village with one set of three robes, remained in the monastery
in another set of three robes, and went down to bathe in another set, amply
show that such additional sets of three robes were regarded as extra robes
(atirekac¥vara) which could be kept for ten days at most (see Vin I 289, 3– 12,
BD IV 411).
124Vin I 289 ,29– 30 (BD IV 412)!; Vin III 196, 9– 11 [Niss 1 Mk] (BD II 4–5).
125E.g. Vin I 297 ,2 –10 (BD IV 423f.)!; 308 ,32– 35!; 309, 2, 3, 12, 13, 16, 19– 21 (BD IV
441ff.!; vissåsagåha!/ !adhi††håna, without the exact wording to be used)!; II

119, 6– 8 (BD V 163 !; with the wording)!; 123 ,32 (BD V 170f. referring to the
namataka)!; III 204, 36!; 246 ,25 (BD II 28 !; 121 !; referring to the patta) !; V
137, 29!; 140, 18, 37 (BD VI 222!; 227f.)!; 173,23, 25, 26!; 174,33!; 175, 13 (BD VI
281!; 283f. !; paccudd håra precedes the adhi††håna#; adhi††håna follows the
paccuddhåra)!; 176,26, 29, 32–33 (BD VI 286 !; ka†hina).
126Vin I 296 ,30–97,10 (BD IV 423).
127For the tic¥vara see above. The nis¥dana (Vin IV 170,29–31!; 171, 11–14 [Påc
89 Mk]!; BD III 96) was two vidatthi in length and one-and-a-half in breadth
according to the current vidatthi plus a border of one vidatthi breadth, thus
altogether 4 ! 3.5 vidatthi (c. 1 ! 0 .87 metres). The vassikaså†ikå, a cloth for
the rains retreat in the four months of the rains allowed for the monks (Vin I
294, 24, BD IV 420), was six vidatthi in length and two-and-a half in breadth
according to the current vidatthi (vassikaså†ikå!; Vin IV 172, 22–73, 3 [Påc 91
Mk] !; BD III 99), ca. 1.5 ! 0.62 metres. A kaˆ"upa†icchåd¥ was allowed in
case of certain skin diseases (Vin I 296,4– 5, BD IV 421)!; it spread from below
the navel to above the knees and was four vidatthi in length and two in
While the Vinaya describes the procedure of taking formal
possession with respect to a strainer only,128 the Samantapåsådikå is
more detailed and rules that a monk has to recite an accompanying
formula, for instance: “I take formal possession of this cloak” (imaµ
sa!ghå†iµ adhi††håmi).129
As already mentioned, a monk is allowed one set of the three robes.
If he wishes to accept a new set, he first has to formally abandon
(paccuddharati)130 the old one. Although the Vinaya is not very explicit
with respect to the formal abandonment of the three robes, it must be
presumed that it was common practice, at least during the final stage of
development of the Påtimokkha, because the Vinaya mentions the
breadth according to the current vidatthi (Vin IV 172,11–14 [Påc 90 Mk]!; BD

III 97f.), i.e., ca. 1 ! 0.5 metres.
Exceptions are the sheet, allowed in Vin I 295,27–29 (BD IV 421) in the size
one wants. Thus there is no limit as to its size. The size of the mukha-
puñchanaco¬a (Vin I 296,19–20, BD IV 422) seems to have not been de fined in
the Vinaya. However, it could be deduced from its function. In the com-
mentarial layer, according to some, two such garments were allowed, while
others declare that many may be used (Sp III 645,1–4). The requisite cloth is
allowed in Vin I 296,32–33 (BD IV 422). No limit with respect to the number
of requisite cloths is given, see Kkh 95 ,24–25 = Pålim 33 ,19–20 = Sp III
645 ,4–5 !: parikkhåraco¬e gaânå natthi. yattakaµ icchati tattakaµ adhi-
††håtabbam eva.
128Vin II 119 ,6 –8 !: sace na hoti parissåvanaµ vå dhammakarako vå saµghå†i-
kaˆô pi adhi††håtabbo iminå parissåvetvå pivissåm¥ ti. “If there is not a

strainer or a regulation water pot, then a corner of the outer cloak should be
determined upon with the words, ‘I will drink [water] having strained it with
this.’ ” (BD V 163). This example, though not general, shows that taking
formal possession of is an express statement in which the object and the fact
that it is taken possession of are mentioned.
129The two ways of making an adhi††håna are verbal and physical, Vin V
117, 37–38 (patta) !; 117, 38–18, 1 (c¥vara) !; Sp III 643 ,3 ff. (c¥vara)!; 705 ,16ff.
(patta).
130See BHSD s.v. pratyuddharati, “removes”. Horner, BD II 22, n. 3 , discusses
the term at length, but did not grasp the sense correctly. Here in the casuistry
it should mean, “if [the robe] is not formally given up, [but] he is of the
opinion that it has been formally given up” (apaccuddha†e paccuddha†asaññ¥,
Vin III 202 ,20).
formal abandonment of a robe in various sections.131 This indicates that
the knowledge of the practical details is taken for granted.
This said, there is little room left for a monk to own more than one
set of three robes at the same time without getting into conflict with the
law, one would think.
However, we have at least circumstantial evidence that already at
the time of the Vinaya monks had more than one set of three robes at
their disposal. (1) Firstly, there is a stereotype formula laying down the
duties of a pupil, etc., if his preceptor, etc., wishes to leave the
monastery to go to town. Here it becomes apparent that the preceptor
changes at least one of the three robes in preparation for the trip (he
receives a nivåsana and hands back a pa†inivåsana), and that he
changes it again on his return (he hands back the nivåsana and grasps
131In the Suttavibha"ga in the introductory story to Påcittiya 59 Mk and in the

Påtimokkha rule itself!: Vin IV 121, 17, 20, 23 (introductory story), 121,30 –33
(rule)!; 122, 17–18 (word-by-word commentary), 122,19 –21 (casuistry!; BD II
411ff.)!; referred to in Vin V 22,7– 14 (BD VI 34)!; furthermore in the casuistry
and in the anåpatti formulas of two Nissaggiya rules!: Vin III 2 0 2 ,20
(casuistry), 28 (anåpatti formula) [Niss 2 Mk] (BD II 22–23) !; Vin III
264, 21–22 (casuistry), 32 (anåpatti formula) [Niss 29 Mk] (BD II 159)!; in the
Parivåra (Vin V 176 ,24– 34 ) in the frame of the description of the ka†hina
ceremony, which — as is well known — represents a later stage than the
description of the ka†hina ceremony in the Mahåvagga. The rule Påc 59 Mk
has been misunderstood by Horner, BD II 411– 13, because she did not
recognize the technical meaning of paccuddharati. The translation in Påtim,
2001, p. 67!: “If any bhikkhu ... should use it without a formal taking back
[i.e., rescinding of the assignment]”, also does not fully grasp the sense, since
here the bhikkhu who uses the robe, and the one who formally takes it back,
i.e., rescinds his own assignment, are one and the same person. In fact the
bhikkhu who uses the [robe] (i.e., the one who had assigned the robe to a
second bhikkhu) is different from the bhikkhu whose robe he uses (i.e.,
whom he had assigned the robe to before), and who did not formally give it
up (apaccuddhårakaµ). For, if someone assigns an object to someone else,
that person has to take formal possession of it in order to be able to use it.
Before that person again may assign the robe to someone else, he first has to
formally give it up (paccuddharati) again.
the pa†inivåsana). 132 This clearly presupposes that the preceptor has
more than one set of three robes at his disposal.133
(2) Secondly, we have two instances in the Vinaya where the word
vihårac¥vara is used.134 The exact meaning of this word in the Vinaya
is not known, but it cannot be excluded that it refers to a robe to be used
by a monk in a vihåra. However, the Samantapåsådikå indicates that it
is a robe deposited as a requisite by the donors of the vihåra.135
(3) Thirdly, the ascetic practice (dhuta!ga), called the three-robe
wearer (tec¥varika), obliges a monk to wear only three robes, with only
one yellow shoulder cloth (aµsakasåva) allowed in addition. From the
fact that this is considered an austerity, we may safely assume that the
original confinement to a single set of three robes was no longer the
rule, but rather the exception. Since the three-robe wearer is mentioned
in the Sutta-pi†aka136 and in later layers of the Vinaya, i.e., in the report
132Vin I 4 6 ,1 2 – 1 3 = II 223, 14f. sace upajjhåyo gåmaµ pavisitukåmo hoti,

nivåsanaµ dåtabbaµ pa†inivåsanaµ pa†iggahetabbaµ (BD IV 60 !: “If the
preceptor wishes to enter a village, his inner clothing should be given [to
him], the inner clothing [that he is wearing] should be received [from him] in
return).” Vin I 46,25– 27!: paccuggantvå pattac¥varaµ pa†iggahetabbaµ, pa†i-
nivåsanaµ dåtabbaµ, ... nivåsanaµ pa†iggahetabbaµ. BD IV 60 !: “Having
gone to meet him, he should receive his bowl and robe, he should give back
the inner clothing [given] in return!; he should receive his inner clothing.”
133The robe is named nivåsana. Horner supposes that nivåsana is another word
for antaravåsaka (BD I 60, n. 1). She (BD I 60, n. 2) rejects the interpretation
of VinTexts I 155, where nivåsana is rendered as “under garment (i.e., his
house-dress !?)”, because in that case the monk would not be a tec¥varika.
Thus she rejects an interpretation because it does not fit her expectation.
Interestingly, the Mahåsåµghika-Lokottaravådins in their Abhisamåcårikå
in a parallel to our passage, differentiate between a gråmapraveßikac¥vara#/
nivåsana and an åråmacaraâkaµ c¥varaµ or a vihåracaraâkac¥vara!/
nivåsana. I owe this information to Seishi Karashima.
134Vin III 212, 20, 23 (BD II 46, with n. 2). This vihårac¥vara clearly is deposited
in a vihåra, and belongs to the sa"gha, i.e., it is not taken into formal
possession by some monk.
135Defined by the †¥kås, Sp-† II 403 ,1 = Pålim-n† II 309 ,10!: vihårac¥varan ti
senåsanac¥varaµ.
136A I 38,13!; M I 214 ,5 !; see BD IV 351, n. 3.
of the second council and in the Parivåra,137 this change of practice
must have taken effect at least by the end of the first century B.C. But
how could the new attitude be put into practice without transgressing
the rules ? There is a long passage which illuminates this point in the
!
Samantapåsådikå quoting early teachers and texts from at least the first
century B. C. The question discussed here is whether or not it is allowed
to take formal possession of the set of three robes as requisite cloth
(parikkhåraco¬a).138
The first authority quoted in this context is Thera Mahåpaduma, 139
a Vinaya specialist (vinayadhara) from Sri Lanka, a pupil of
Vinayadhara Thera Upatissa, who lived during the famine in the first
century B . C. 140 He declares that a monk may only take formal
possession of the set of three robes under precisely this designation (set
of three robes). 141 Interestingly enough, this literal interpretation turns
137In the description of the monks of Påvå (Vin I 253 ,6 !; II 299,6, 9 [report of the
second council])!; V 131,16!; 193 ,10.
138Sp III 643 , 31–44, 1 !: tic¥varaµ pana parikkhåraco¬aµ adhi††håtuµ va††ati na
va††at¥ ti#? “But is it allowed to take formal possession of the three robes as
requisite robes [or] is it not allowed!?”
139Mori 1989, p. 68 (130), no. 93.
140Sp I 263 ,24–64, 7. Thera Upatissa is mentioned together with Thera Phussa-
deva as one of those who protected the Vinaya when the great peril arose in
Sri Lanka (mahåbhaye uppanne, Sp I 263 ,25– 28). This famine is thought to
have taken place between 102 and 89 B.C.!; see Mori 1989, p. 61.
141Sp III 644 , 1– 4 ! : Mahåpadumatthero kiråha!: tic¥varaµ tic¥varam eva adh-
i††håtabbaµ. sace parikkhåraco¬ådhi††hånaµ labheyya udositasikkhåpade

parihåro niratthako bhaveyyå ti. evaµ vutte kira avaseså bhikkhË åhaµsu!:
parikkhåraco¬am pi bhagavatå va adhi††håtabban ti vuttaµ, tasmå va††at¥ ti.
“Thera Mahåpaduma, apparently, says that the set of three robes is to be
taken formal possession of only as a set of three robes. If the taking formal
possession of [the set of three robes] as a requisite cloth were allowed, the
protection in the storehouse rule (Niss 2 Mk!; i.e., to be allowed to wear fewer
than three robes during the ka†hina period, and after the ka†hina has been
closed, with the agreement of the bhikkhus) would become useless. When he
had spoken thus, then the remaining monks said, ‘Even the requisite cloth is
taught in fact by the Lord as one which has to be taken formal possession of,
out to be the minority. All the other monks hold that the tic¥vara
may also be taken into formal possession as a parikkhåraco¬a (Sp III
644,4–6).
Since no limit is given regarding the size and number of requisite
cloths (parikkhåraco¬a), there also is no need to formally give up
(paccuddharati) old parikkhåraco¬as before accepting new ones. In
theory, this leaves room for unlimited accumulation of such requisites
in all shapes and sizes.
The view of Thera Mahåpaduma’s opponents receives additional
support from the Mahåpaccar¥142 and also from Thera Mahåtissa, 143 an
inhabitant of Puˆâvå¬ika and a reciter of both Vibha"gas (ubhato-
vibha!gabhåâka), 144 who refers to it as an earlier practice of the forest
therefore it is allowed (i.e., it is allowed to take formal possession of the set

of three robes as a requisite cloth).’ ”
142Sp III 644 , 6– 10 !: Mahåpaccariyam pi vuttaµ “parikkhåraco¬aµ nåma
på†ekkaµ nidhånamukham etan ti tic¥varaµ parikkhåraco¬an ti adhi††hahitvå

paribhuñjituµ va††ati. udositasikkhåpade pana tic¥varaµ adhi††hahitvå pari-
harantassa parihåro vutto” ti. “Even in the Mahåpaccar¥ it is said, ‘Requisite
cloth means!: this [taking formal possession of as requisite cloth is] a distinct
one, mainly [serving] the storage, [thus] it is allowed to use the set of three
robes, having taken formal possession of them as requisite cloth. In the
storehouse rule (Niss 2 Mk), however, the protection is taught for him, who,
having taken formal possession of as the set of three robes, preserves [the
robes].’ ”
143Mori 1989, p. 67 (129), no. 90.
144Sp III 644 ,10– 1 7 ! : ubhato-Vibha!gabhåâko puˆâvålikavås¥ Mahåtissa-
tthero pi kira åha!: “mayaµ pubbe mahåtherånaµ assumha ‘araññavåsino

bhikkhË rukkhasusiråd¥su c¥varaµ †hapetvå padhånaµ padahanatthåya
gacchanti. såmantavihåre dhammasavanatthåya gatånañ ca nesaµ sËriye
u††hite såmaêrå vå daharabhikkhË vå pattac¥varaµ gahetvå gacchanti,
tasmå sukhaparibhogatthaµ tic¥varaµ parikkhåraco¬an ti adhi††håtuµ
va††at¥’ ” ti. “Even the reciter of the two Vibha"gas, the inhabitant of
Puˆâvålika, Thera Mahåtissa, as is well known, says, ‘We have heard from
the mahåtheras in earlier times that the monks living in the forest, having
deposited a robe in a hollow of a tree, etc., in order to exert [meditation] went
to [the place for] exertion, and that, when the sun arose, the novices and
young monks of these mahåtheras who had gone [there], having taken robe
and bowl, went to a neighbouring monastery in order to hear the dhamma.
monks. The Mahåpaccar¥ argues that forest monks had practised the
taking formal possession of the tic¥vara as a parikkhåraco¬a, because
within an undetermined monastic boundary (abaddhas¥må), as is the
case in a forest, there is no good protection for the set of the three
robes.145 Since the Mahåpaccar¥ dates from around the first century
B.C., if not earlier, it provides an impressive testimony to the practice of
taking formal possession of the set of three robes as a requisite cloth.
This is corroborated by the Samantapåsådikå, which quotes the
Mahåpaccar¥ as the final authority on this issue,146 as well as by the
Ka"khåvitaraˆ¥ 147 and Vajirabuddhi’s Anugaˆ†hipada.148 The Anu-
Therefore, it is allowed for the ease of use to take formal possession of the set
of three robes as requisite cloth.’ ”
145Sp III 644 ,17– 20 !: Mahåpaccariyam pi vuttaµ “pubbe åraññikå bhikkhË
abaddhas¥måyaµ dupparihåran ti tic¥varaµ parikkhåraco¬am eva adhi-

††hahitvå paribhuñjiµsË” ti. “Even in the Mahåpaccar¥ it is said, ‘In earlier
times the forest monks used the set of three robes having taken formal
possession of [them] only as requisite cloth, [owing to the fact] that in an
undetermined [monastic] boundary [there exists] poor protection [regarding
the three robes].’ ”
The monastic boundary consisting in seven abbhantara (sattabbhantara-
s¥må), which is the type of boundary (s¥må) valid in a forest, does not in fact
have the function of protecting monks from being separated from the three
robes (Sp V 1052,11). Since, however, this boundary comes into being only
for a legal procedure, for the remaining time the robe rules for the forest are
valid. Thus, a monk in the forest may not be more than seven abbhantara
distant from his robes (cf. Kieffer-Pülz 1992, B 15.2.3).
146See above n. 99.
147Kkh 95,11– 12 !: idañ ca pana tic¥varaµ sukhaparibhogatthaµ parikkhåra-
co¬aµ adhi††håtum pi va††ati. “And this set of three robes may even be taken
formal possession of as a requisite cloth for easy usage.” From the point of
view of content this statement reproduces parts of the opinion of Thera
Mahåtissa!; see above n. 144.
148Vjb 223 , 15–17 = Pålim-n† 93,27–94,1 ! : pa†hamaµ tic¥varaµ tic¥varådhi-
††hånena adhi††håtabbaµ, puna pariharituµ asakkontena paccuddharitvå

parikkhåraco¬aµ adhi††håtabbaµ. na tv eva ådito va idaµ vuttan ti vuttaµ.
“ ‘First, the set of three robes is to be taken formal possession of by [means
of] the taking formal possession of as a set of three robes!; by one not capable
of preserving [them], they, after having been given up formally, should again
gaˆ†hipada recommends that a monk unable to keep up the robes he has
taken formal possession of as three robes should formally give them up
(paccuddharati) and then take formal possession of them as
parikkhåraco¬a in order to avoid a transgression of Niss 2 Mk, which
forbids a monk to part with one of his robes even for one night after the
ka†hina period has been closed, except with the approval of the monks.
In case the set of three robes has not been taken formal possession of as
such, the rules for the tic¥vara do not apply to them.
The evidence of the Anugaˆ†hipada also shows that this rule
represents a later development, albeit of a considerable age. The Vajira-
buddhi-†¥kå’s lengthy pronouncements on this practice 149 are intended
to show that, although it does not belong to the earliest rules, it would
have been decided in exactly the same way by the Buddha and that it
could very well have been initiated by him. The practice was very
common in later times, as we can see from texts on monastic law dating
from the twelfth to seventeenth centuries.150
Thus taking formal possession of a set of three robes as a
parikkhåraco¬a was, and probably still is, an acceptable way of
circumventing the strict and complex rules applying to the tic¥vara.
be taken formal possession of as requisite cloths. But this has not been taught
from the very beginning’, [thus] it is said [in Vajirabuddhi’s Anugaˆ†hi-
pada].”
149Vjb 222 ,7 –23, 25 = Pålim-n† I 93,13–27 and 93,27–94,1.
150Khuddas-p† 96,14– 1 7 ! : kiµ pana tic¥varaµ parikkhåraco¬aµ adhi††håtuµ
va††at¥ ti!? åma va††ati, parikkhåraco¬aµ nåma på†ekkaµ nidhånamukham

etan ti#; tic¥varaµ parikkhåraco¬aµ adhi††hahitvå paribhuñjituµ va††ati (Sp
III 6 4 4 ,6 –8 ) !; Vin-vn-p† 282, 26–28 !: parikkhåraco¬asså ti parikkhåraco¬a-
nåmena adhi††hahitvå c¥varaµ paribhuñjitukåmassa parikkhåraco¬a-
nåmena adhi††håtuµ va††ati. Pålim-n† I 94,1 –4 $ Sp-† II 388, 29–31 = Pålim-p†
22,25– 29 $ Vmv I 313,17–18!: aparikkhåraco¬aµ adhi††håtun ti (Sp III 643 ,31)
parikkhåraco¬aµ katvå adhi††håtuµa (a–a Pålim-n†, Vmv omit). baddha-
s¥måyaµ (Pålim-n† s¥måya) avippavåsas¥måsammutisabbhåvato (Pålim-n†,
Vmv sambhavato) c¥varavippavåse pi (Pålim-n† omits) b n’ ev’ atthi doso ti b
(b–b Vmv omits) na tattha dupparihåratå (Pålim-n† dupparihåro) ti åha
abaddhas¥måyaµ dupparihåran ti (Sp III 644 ,18).
Even outside the ka†hina period, this practice enables a monk to travel
with fewer than three robes, it allows him to own more than just one set
of three robes, and it protects him from transgressing the rules applying
to the tic¥vara. This practice has met with general acceptance, except by
the early Thera Mahåpaduma, and it prepared the way for the legal
possession of multiple sets of three robes — which nowadays seems to
be the regular case in most instances.
These two examples should afford a glimpse of the methods used in
Theravåda legal literature for adapting the largely fi xed rules of
Buddhist law to changing circumstances or wishes. Though there may
exist still further forms of adaptation, it can safely be said that the first
of the two methods dealt with here is the most common and widespread
in the legal texts.
Petra Kieffer-Pülz
Weimar
REFERENCES AND ABBREVIATIONS

References to Påli texts are to the editions of the Pali Text Society (PTS), to the
editions of the Cha††hasa"gåyana Series (Chs), or to the databank on CD-ROM
(CSCD). For bibliographical details see CPD, Epilegomena, and the lists of
abbreviations in CPD, Vol. III.
ABBREVIATIONS
A A"guttaranikåya (PTS)
Ap Apadåna (PTS)
Ap-a Apadåna-a††hakathå (PTS)
BD I.B.Horner, The Book of the Discipline (Vinaya-Pi†aka), 6 Vols.,
London, 1938–1966 (Sacred Books of the Buddhists 10, 11, 13,
14, 20, 25).
BHSD F. Edgerton, Buddhist Hybrid Sanskrit Grammar and Dictionary,
Vol. 2 !: Dictionary, Delhi, repr. 1985.
Bv Buddhavaµsa (PTS)
Bv-a Buddhavaµsa-a††hakathå (PTS)
CPD A Critical Påli Dictionary, begun by V. Trenckner, ed.
D.!Andersen, H. Smith, H. Hendriksen, Vols. 1–2 , 3, fascs. 1ff.,
Copenhagen 1924 ff.
CSCD CD-ROM, Version 3, by the Vipassana Research Institute,
Dhammagiri, Igatpuri, India!: Databank of canonical and post-
canonical Påli texts according to the Burmese Cha††hasa"gåyana
Edition. [Quoted according to the pagination of the roman
editions if available, otherwise according to the Burmese
editions!; paginations sometimes deviate from the printed editions
by one page, depending on the script chosen].
Dhp-a Dhammapada-a††hakathå (PTS)
D D¥ghanikåya (PTS)
DOP Margaret Cone, A Dictionary of Påli, Vol. I (A–Kh), Oxford,
2001.
DPPN G.P. Malalasekera, Dictionary of Påli Proper Names, 2 Vols.,
London 1937–1938.
Jäschke H.A. Jäschke, A Tibetan–English Dictionary, London, repr. 1977.
Ja Jåtaka††hakathå (PTS)
Khuddas Khuddasikkhå (CSCD)
Khuddas-p† Khuddasikkhåpuråâ-†¥kå (Chs)
Kkh Ka"khåvitaraˆ¥ (PTS, 2003)
Mhv Mahåvaµsa (PTS)
M Majjhima Nikåya (PTS)
Mk monks’ rules
Mil Milindapañha (PTS reprint)
Mp ManorathapËraˆ¥ (PTS)
N nuns’ rules
Niss Nissaggiya (category of offences)
Påc Påcittiya (category of offences)
Påc-y Påcityådiyojanå (Chs)
Pålim Pålimuttakavinayavinicchaya (Chs)
Pålim-n† Pålimuttakavinayavinicchayanava†¥kå (Chs)
Pår Påråjika (category of offences)
Ps PapañcasËdan¥ (PTS)
PTS Pali Text Society
S Saµyuttanikåya (PTS)
Se Siamese edition
Sp Samantapåsådikå (PTS)
Spk Såratthappakåsin¥ (PTS)
Sp-† Såratthad¥pan¥ (Chs)
Sv Suma"galavilåsin¥ (PTS)
Sv-n† Suma"galavilåsin¥nava†¥kå (CSCD)
Th-a Theragåthå-a††hakathå (PTS)
Th¥-a Ther¥gåthå-a††hakathå (PTS)
Ud-a Udåna-a††hakathå (PTS)
v. verse
Vibh-m† Vibha"ga-mËla†¥kå (CSCD)
Vin Vinaya-pi†aka (PTS)
VinTexts T.W. Rhys Davids, Hermann Oldenberg, Vinaya Texts, Vol. 1–3,
Oxford 1881, 1882, 1885 (Sacred Books of the East, 13, 17, 20).
Vin-vn Vinayavinicchaya (CSCD)
Vin-vn-p† Vinayavinicchaya-poråâ†¥kå (Chs)
Vism Visuddhimagga (PTS)
Vjb Vajirabuddhi-†¥kå (Chs)
Vmv Vimativinodan¥-†¥kå (Chs)
SECONDARY LITERATURE
Entrance to the Vinaya. Vajirañåâvarorasa, The Entrance to the Vinaya.
Vinayamukha, 3 vols., Bangkok 1969, 1973, 1983 (original text in
Thai!: 2459 = 1916)
Geiger, W., 1986. Culture of Ceylon in Mediaeval Times, ed. H. Bechert,
Stuttgart, 2nd ed.
Gunawardana, R.A.L.H., 1979. Robe and Plough, Monasticism and Economic
Interest in Early Medieval Sri Lanka, Arizona (Monographs of the
Association for Asian Studies, no. 35)
Härtel, Herbert, 1956. Karmavåcanå. Formulare für den Gebrauch im
buddhistischen Gemeindeleben aus ostturkistanischen Sanskrit-Hand-
schriften, Berlin (Sanskrittexte aus den Turfanfunden, III)
von Hinüber, Oskar, 1995. “Buddhist Law According to the Theravåda-Vinaya.
A Survey of Theory and Practice”, Journal of the International
Association of Buddhist Studies 18.1, pp. 7–45
——— 1996. A Handbook of Påli Literature, Berlin. (Indian Philology and
South Asian Studies, 2)
——— 1999. Das Påtimokkhasutta der Theravådin. Studien zur Literatur des
Theravåda-Buddhismus II, Mainz (Akademie der Wissenschaften und
der Literatur, Abhandlungen der Geistes- und sozialwissenschaftlichen
Klasse, Jg. 1999, Nr. 6)
——— 2000. “Ordensregeln, Organisation und frühe Ausbreitung des
Buddhismus”, Buddhismus in Geschichte und Gegenwart (Weiter-
bildendes Studium), Vol. IV, Hamburg, pp. 137–150
Kieffer-Pülz, Petra, 1 9 9 2 . Die S¥må. Vorschriften zur Regelung der
buddhistischen Gemeindegrenze in älteren buddhistischen Texten,
Berlin (Monographien zur indischen Archäologie, Kunst und
Philologie, 8)
——— 1993. “Zitate aus der Andhaka-A††hakathå in der Samantapåsådikå”,
Studien zur Indologie und Buddhismuskunde, Festgabe des Seminars
für Indologie und Buddhismuskunde für Professor Dr. Heinz Bechert,
ed. by Reinhold Grünendahl and others, Bonn (Indica et Tibetica, 22),
pp.!171 –212
——— 2001. “Påråjika 1 and Sanghådisesa 1!: Hitherto Untranslated Passages
from the Vinayapitaka of the Theravådins”, Traditional South Asian
Medicine 6, pp. 62– 84
Mori, Sodo, 1989. “Chronology of the ‘S¥ha¬a Sources’ for the Påli Com-
mentaries (II)”, Studies of the Påli Commentaries. A Provisional
Collection of Articles, Japan, p. 61
Ratnapåla, Nandasena, 1971. The Katikåvatas !: Laws of the Buddhist Order of
Ceylon, from the Twelfth Century to the Eighteenth Century, critically
edited, translated and annotated, München. (Münchener Studien zur
Sprachwissenschaft, Beiheft N)
Ray, Himanshu, 1986. Monastery and Guild !: Trade in the Satavahana Period,
Delhi
Schopen, Gregory, 1994A. “Doing Buisiness for the Lord!: Lending on Interest
and Written Loan Contracts in the MËlasarvåstivådavinaya”, Journal
of the American Oriental Society 114, pp. 527– 54
——— 1994 B . “The Monastic Ownership of Servants or Slaves!: Local and
Legal Factors in the Redactional History of Two Vinayas”, Journal of
the International Association of Buddhist Studies 17,2, pp. 145–73
Treatise. A Treatise on Monastic Boundaries by King Råma IV of Siam, ed. by
Petra Kieffer-Pülz, (in press)
Yamagiwa, Nobuyuki, 2002. “Óråmika — Gardener or Park Keeper!? One of
the Marginals Around the Buddhist Saµgha,” Buddhist and Indian
Studies in Honour of Professor Sodo Mori, Nagoya, pp. 363– 85
The Sus¥ma-sutta and the Wisdom-Liberated Arahant
The Sus¥ma-sutta, a short discourse of the Saµyutta-nikåya, uses a

dramatic plot to shine a spotlight on a doctrinal topic that must have
been of vital interest to the ancient Sa!gha as it classified persons with
respect to their meditative expertise and paths of attainment.1 The topic
is the nature of the paññåvimutta arahant, the person who attains
liberation through the special efficacy of wisdom without reaching
extraordinary distinction in the sphere of samådhi or concentration. The
Sus¥ma-sutta merits special consideration because the Påli version has
three parallels preserved in Chinese translation, and thus a comparison
of the Påli discourse with its Chinese counterparts permits us to see
how, even in an early stage of textual transmission, the Buddhist
schools were already in subtle ways contemplating different solutions to
the doctrinal problem raised by the sutta. For ease of reference, I will
designate the Påli version S 12:70. Among the Chinese versions, one is
found in the Vinaya of the Mahåså!ghika school, which I will refer to
as M-Vin.2 Another version is sutta no. 347 in the Saµyukta-ågama,
commonly understood to be the Sarvåstivåda (or perhaps MËla-
sarvåstivåda) counterpart to the Saµyutta-nikåya.3 I will call this
version SÓ 347. The other is an incomplete citation in the Abhidharma-
1S 12:70; II 119–28.
2T22, 362b25–363b26. In my discussion, when I translate terms used in the
Chinese texts into their Indic equivalents, for the sake of consistency I will
generally use the Påli counterparts, even though these texts may have been
translated from Sanskrit or Buddhist Hybrid Sanskrit originals. For the same
reason, I will refer to all versions of the basic text as a sutta rather than use
sutta for the Påli version and sËtra for non-Påli versions
3SÓ 347 ; T2, 96b25–98a12. Richard Gombrich (How Buddhism Began : The
Conditioned Genesis of the Early Teaching (London : Athlone, 1996),

pp. 123–27) discusses the relations between S 12:70 and SÓ 347. My interpre-
tation of both versions differs considerably from Gombrich’s.
52 Bhikkhu Bodhi
vibhå"å-ßåstra.4 This citation terminates before we reach the end of the
sutta, but it covers most of the points relevant to our study. This version
will be referred to as Vibhå"å.
In this paper I will use S 12:70 as the primary basis for my
discussion and bring in the others later for purposes of comparison. I
will first present a summary of the “plot”. Then I will explore the theme
of the “arahant liberated by wisdom” based on the primary text,
followed by a discussion of its treatment in the several Chinese versions
of the sutta. At some later time, I hope to write a sequel to this paper to
explore the different versions of the second part of the discourse, which
deals with the two knowledges contributing to the status of one
liberated by wisdom.
1. The Plot
The sutta opens with the Buddha dwelling in the Bamboo Grove at
Råjagaha. At the time, he is respected and honored by the laity and
amply provided with all the requisites, as is the Bhikkhu Sa!gha.
Because of the Buddha’s rise to fame, the fortunes of the “wanderers of
other sects” have steeply declined. The wanderers resident at Råjagaha
therefore decide to assign a crucial mission to one of their members
named Sus¥ma. He is to go forth under “the Ascetic Gotama”, master
his doctrine, and then return and teach it to his own community. They
assume that the Buddha’s doctrine is the key to his success, and so, they
suppose, once they have learned his Dhamma and can teach it to the lay
folk, they will regain the support that they have lost to the Sakyan sage.
Sus¥ma agrees and heads off towards the Bamboo Grove. At the
4There are actually two parallel treatises that cite this version, with slight differ-
ences between them. The one I mostly draw upon is Abhidharma-vibhå"å-
ßåstra (no. 1546), which cites it at T28, 407c26–408b11. The larger version of
this treatise, Abhidharma-mahåvibhå"å-ßåstra (no. 1545), cites it at T27,
572b16–572c27. It is an open question whether this version is actually a sutta
with canonical or quasi-canonical status ; it may be, rather, merely the
treatise’s paraphrase of a sutta. For the sake of convenience, however, I will
refer to it as if it were another version of the sutta.
The Sus¥ma-sutta and the Wisdom-Liberated Arahant 53
entrance he meets the monk Ónanda and tells him he wants to lead the
spiritual life under the Buddha.5 Ónanda brings Sus¥ma to the Buddha,
who tells Ónanda to ordain him. Shortly thereafter, in the Buddha’s
presence, a number of monks declare final knowledge (aññå), that is,
arahantship, announcing, “We understand : Birth is finished, the holy
life has been lived, what had to be done has been done, there is no more
coming back to any state of being.”6 Sus¥ma hears about this and
approaches the monks to ask whether this report is true. When they
confirm it, he asks them whether they have attained the five mundane
types of super-knowledge : the modes of spiritual power, the divine ear,
the ability to read the minds of others, the recollection of past lives, and
the divine eye which sees how beings pass away and take rebirth in
accordance with their kamma.7 In each case, the monks deny possessing
these super-knowledges. Then Sus¥ma asks them whether they dwell in
the “peaceful emancipations, transcending forms, formless, having con-
tacted them with the body”.8 Again, they answer no. Now Sus¥ma is
puzzled. He tells the monks that he cannot understand how they could
declare arahantship yet deny that they attain these superhuman states.
They reply, “We are liberated by wisdom, friend Sus¥ma.”9
This answer does not satisfy Sus¥ma, but when he asks them to
elucidate they only repeat the same words, “Whether or not you under-
stand, we are liberated by wisdom.” So Sus¥ma goes to the Buddha in
5In all three Chinese versions, it is not Ónanda that he meets but a group of
monks. In SÓ 347 and Vibhå"å, the monks bring him to the Buddha, who tells
them to ordain him. In M-Vin, the monks tell Sus¥ma that, as a convert from
another sect, he must live on probation for four months, and then, if the
Sa!gha approves, they will give him the ordination.
6S II 120,30–32 : kh¥ˆå jåti vusitaµ brahmacariyaµ kataµ karaˆ¥yaµ nåparaµ
itthattåyåti pajånåma.
7S II 121–23. Briefly, in Påli : (1) iddhividha, (2) dibbasotadhåtu, (3) cetopari-
yañåâ, (4) pubbenivåsånussatiñåâ, (5) yathåkammacutËpapåtañåâ.

8S II 123,15–16 : ye te santå vimokkhå atikkamma rËpe åruppå te kåyena
phusitvå viharatha.
9S II 123,26 : paññåvimuttå kho mayaµ, åvuso Sus¥ma.
54 Bhikkhu Bodhi
quest of clarification and reports to him the entire conversation he had
with the monks. The Buddha too replies with an enigmatic one-sentence
answer, “First, Sus¥ma, there is knowledge of the persistence of
principles ; afterwards, knowledge of nibbåna.”10
Sus¥ma asks the Buddha to explain this concise statement in detail,
but the Buddha first responds simply by repeating his reply, “Whether
or not you understand, Sus¥ma, first there is knowledge of the
persistence of principles ; afterwards, knowledge of nibbåna.” However,
he then tries to guide Sus¥ma to an understanding of his words. He first
leads him through the catechism on the three characteristics — imper-
manence, suffering, and non-self — in relation to the five aggregates,
exactly as we find it in the second “argument” of the well-known
Anattalakkhaâ-sutta, the Discourse on the Characterstic of Non-Self (S
22:59).11 This culminates in the noble disciple becoming disenchanted
with the five aggregates ; through disenchantment, he becomes dispas-
sionate ; and through dispassion, his mind is liberated. With liberation
comes the knowledge of liberation and he understands : “Birth is
finished … there is no more coming back to any state of being.”
The Buddha next takes Sus¥ma through a catechism on dependent
origination (pa†icca-samuppåda), first with respect to arising : begin-
ning with “aging-and-death have birth as condition” and ending with
“volitional activities have ignorance as condition” ; and then with
respect to cessation, starting from “aging-and-death cease with the
cessation of birth” and ending with “volitional activities cease with the
cessation of ignorance”. At this point the Buddha asks Sus¥ma whether
“knowing and seeing thus” (evaµ jånanto evaµ passanto), he exercises
the five super-knowledges or attains the peaceful formless emancipa-
tions. When Sus¥ma says no, the Buddha asks him how he could answer
as he did while being unable to attain these states. The Buddha’s use of
the word “answer” (veyyåkaraâ) apparently refers back to his agreeing
10S II 124,10–11 : pubbe kho Sus¥ma dhamma††hitiñåâµ, pacchå nibbåne

ñåâµ.
11S III 67,22–68,25.
that he “knows and sees” each of the points the Buddha asked him
about in the chain of dependent origination.
Sus¥ma does not reply to the Buddha’s question. Instead, he
prostrates himself at the Buddha’s feet, confesses that he entered the
Buddhist order with thievish intent, and asks the Blessed One to pardon
him for this offence. The Buddha then tells him that he was indeed
foolish to have contemplated such a dangerous course of action. To
underscore the danger he relates a simile about a criminal who is
arrested by the king and beheaded to the south of the city. The
consequences of “going forth as a thief in the well-expounded Dhamma
and Discipline”, the Buddha says, are far graver than the punishment
suffered by the criminal ; but since Sus¥ma sees his transgression for
what it is, he pardons him for the sake of his future restraint.
2. The One Liberated by Wisdom

In S 12:70, as we have just seen, when Sus¥ma questions the monks
about their attainments, he asks about the five mundane super-
knowledges and the peaceful formless emancipations, and it is these
that the monks deny possessing. It is intriguing that Sus¥ma’s questions
do not pry into any attainments that the monks might possess below the
level of the formless emancipations. I assume that, whatever might have
been the historical basis for the origination of this sutta, the actual
dialogue, particularly in the first part, is partly the work of the compilers
of the texts. Once this assumption is granted, we may infer that the
compilers of the sutta had compelling doctrinal reasons for drawing the
cut-off point at the formless emancipations. For them to permit Sus¥ma
to ask the monks whether or not they had attained the jhånas, and then
to have the monks give negative answers to these questions, would have
been to directly contradict time-hallowed discourses and doctrinal
formulæ. It seems to me that the compilers of this sutta wish to
insinuate that the monks were actually not attainers of the jhånas, that
they subtly want to introduce into the canon the idea of the arahant who
lacks these distinguished states of concentration. At the same time,
56 Bhikkhu Bodhi
however, they did not want to force an ambiguity that was hovering
over the notion of the “wisdom-liberated arahant” to become resolved
too starkly in black-and-white terms. Hence they allowed the ambiguity
to linger in the canonical text while they resolved the issue in its
commentary, which in the earliest period must have been a teacher’s
oral explanation accompanying the sutta.
The Nikåyas distinguish among different classes of arahants, using
as the basis for the distinctions the attainments they possess ancillary to
their attainment of arahantship. In descending order, some arahants
possess the six “direct knowledges” (cha¬abhiñña) ; some have the three
“higher knowledges” (tevijja) ; some are “liberated in both ways”
(ubhatobhågavimutta) ; and some are “liberated by wisdom” (paññå-
vimutta).12 The main distinction that the Nikåyas draw is between those
arahants “liberated in both ways” and those “liberated by wisdom”. In
the K¥†ågiri-sutta (M 70), the arahant liberated in both ways is defined
as one who “contacts with the body and dwells in those peaceful
emancipations, transcending forms, that are formless, and whose
influxes are exhausted by his seeing with wisdom”.13 The arahant
liberated by wisdom, in contrast, is one who “does not contact with the
body and dwell in those peaceful emancipations, transcending forms,
that are formless, but whose influxes are exhausted by his seeing with
wisdom”.14 Questions can be raised about the exact meaning and
extension of these definitions : for example, to what degree must an
arahant possess the formless emancipations to qualify as “both-ways-
12See S I 191 (S 8:7). The six direct knowledges are the five mundane super-
knowledges enumerated above (see pp. 52–53) plus the knowledge of the
exhaustion of the influxes (åsavakkhayañåâ). The three higher knowledges
are the knowledge of the recollection of past lives, the knowledge of the
passing away and rebirth of beings, and the knowledge of the exhaustion of
the influxes.
13M I 477,26–28 : ekacco puggalo ye te santå vimokkhå atikkamma rËpe åruppå
te kåyena phusitvå viharati paññåya c’ assa disvå åsavå parikkh¥ˆå honti.

14M I 477,33–36 : ekacco puggalo ye te santå vimokkhå atikkamma rËpe åruppå
te na kåyena phusitvå viharati paññåya c’ assa disvå åsavå parikkh¥ˆå honti.

liberated” ? The Puggalapaññatti Commentary maintains that those who
attain the formless attainments but do not gain the attainment of
cessation (nirodhasamåpatti) are called “liberated in both ways” only in
a figurative sense (pariyåyena), while those who gain the eight
attainments (the four jhånas and the four formless attainments) as well
as the attainment of cessation are called “the best of those liberated in
both ways in the literal sense”.15 This same commentary, however, then
concedes that an arahant who attains just one among the formless-
sphere jhånas can still be called “liberated in both ways”.16
The arahant liberated by wisdom, it is clear, has the fourth jhåna as
the upper limit of achievement on the scale of concentration ; for some
reason, not explained in the suttas, an arahant of this type does not
proceed further to attain the formless emancipations. The more
interesting question, however, concerns the minimal attainment in
concentration possessed by an arahant liberated by wisdom. A number
of standard texts define the concentration included in several groups
among the thirty-seven “aids to enlightenment” (bodhipakkhiyå
dhammå) as the four jhånas. In particular, we find the faculty of
concentration (samådhindriya) among the five faculties and the right
concentration factor (sammå samådhi) of the noble eightfold path both
defined as the four jhånas.17 The four jhånas also enter into the standard
description of the progressive training of the monk, preceding the
attainment of the higher knowledges,18 and into the threefold higher
training, where they serve as the training in the higher mind
(adhicittasikkhå).19 If we rely upon these texts, taking them literally, it
would follow that any monk liberated by wisdom must have attained all
four jhånas.
15Pp-a 191. nippariyåyena ubhatobhågavimuttase††ho.

16Pp-a 191. arËpåvacarajjhånesu pana ekasmiµ sati ubhatobhågavimutto yeva
nåma hoti.
17S V 196,18–19, 198,24–32 ; D II 313,12–25 ; S V 10,5–18..
18E.g., at D I 73–76 ; M I 181–82, 276–78, etc.
19At A I 235,21–23, 235,38–36,2..
58 Bhikkhu Bodhi
Such a conclusion, however, would be extreme, for other texts
equally authoritative recognize the possibility of attaining arahantship
on the basis of any jhåna. For example, the Jhåna-sutta (A 9:36)
explains that one can gain any one of the four jhånas or lower three
formless attainments and then contemplate its constituents in eleven
ways : as impermanent, suffering, illness, a boil, an arrow, misery,
affliction, alien, disintegrating, empty, and non-self.20 If one is firm in
such insight, the text asserts, one will attain the exhaustion of the
influxes, that is, arahantship ; if one retains a subtle attachment to this
experience, one will emerge as a non-returner. Again, in the
A††hakanågara-sutta (M 52), Ónanda explains how it is possible to
attain the exhaustion of the influxes through any of eleven “doors to the
deathless”.21 One enters any of the four jhånas, the four divine abodes,
or the lower three formless attainments and contemplates it as
conditioned and constructed by volition. One then sees that anything
conditioned and constructed by volition is impermanent and subject to
cessation. This would imply that the texts that define the faculty of
concentration, the right concentration of the noble path, and the training
in the higher mind as the four jhånas should not be taken literally as
meaning that all four jhånas are needed to reach liberation ; rather, they
mean that to attain the final goal as a wisdom-liberated arahant, one
should be able to gain at least one jhåna as a basis for insight.
We might, however, ask whether even this much is indispensable.
In raising this question, we are pushing our line of inquiry further than
Sus¥ma dared to go in his conversation with the monks. Yet, in view of
the direction meditation theory has taken in the mainstream Buddhist
traditions, as we shall see, it is precisely this question that should pique
our curiosity. Now, if we read certain suttas at their face value it would
seem that the first jhåna is a minimum requirement for the attainment of
even the third fruition, the stage of non-returner. A text that lends strong
20A IV 422–26. The commentary explains that the fourth formless attainment is
too subtle to be contemplated with insight.
21M I 349–52.
support to this claim is the Mahåmålu!kya-sutta (M 64), where the
Buddha declares, “There is, Ónanda, a path to the abandoning of the
five lower fetters ; that anyone, without relying on that path, shall know
or see or abandon the five lower fetters, this is impossible.”22 As the
sutta unfolds, the “path to the abandoning of the five lower fetters” is
then shown to be the same course of practice described just above in the
Jhåna-sutta. One enters any of the four jhånas or three lower formless
attainments, and then contemplates its constituents from the same
eleven angles. If one can remain firm in this contemplation, one
exhausts the influxes and reaches arahantship ; if there is still a remnant
of attachment, one cuts off the five lower fetters and becomes a non-
returner.
If the above words — “that anyone, without relying on that path,
shall know or see or abandon the five lower fetters, this is impossible”
— are taken as categorical, there is indeed no possibility at all that an
arahant liberated by wisdom can be destitute of the first jhåna. It will
not suffice, either, to appeal to the Abhidhamma distinction between
form-sphere (rËpåvacara) and supramundane (lokuttara) jhånas and
then hold that while some arahants liberated by wisdom might be
destitute of mundane jhånas, they will still possess at least the first
supramundane jhåna. This claim could not be accepted in a discussion
based solely on the suttas, for the distinction between form-sphere and
supramundane jhånas is never explicitly drawn in the suttas nor is it
even discernible in them.23 If our analysis is to apply to the under-
standing of meditative attainments characteristic of the suttas, it must
use concepts intrinsic to the suttas themselves and not draw upon modes
22M I 434,25–28 : yo, Ónanda, maggo yå pa†ipadå pañcannaµ orambhågiyånaµ

saµyojanånaµ pahånåya taµ maggaµ taµ pa†ipadaµ anågamma
pañcorambhågiyåni saµyojanåni ñassati vå dakkhati vå pajahissati vå ti n’
etaµ †hånaµ vijjati.
23One possible exception to this is the Mahåcattår¥saka-sutta (M 117), which,
however, in the form it has come down, seems to be the reworking of an

archaic version under the influence of later ideas typical of the incipient
Abhidhamma.
60 Bhikkhu Bodhi
of analysis derived from a later phase of Theravåda Buddhist thought.
Although the words of the Mahåmålu!kya-sutta quoted above
might seem to rule out the possibility that those destitute of jhåna can
achieve arahantship, several texts scattered across the Nikåyas hint that
this conclusion would be a bit stern. We should remember that, while
the suttas are remarkably consistent with each other, they are not rigidly
so, and one can often find in some texts exceptions made to principles
apparently laid down as categorical in other texts. One discourse
relevant to our present discussion, the Asubha-sutta (A 4:163), speaks
about four modes of practice : two painful, with sluggish and quick
realization, and two pleasant, again with sluggish and quick realization.
The mode of practice that is painful, with sluggish realization, is
described thus :
Here, a monk dwells contemplating the unattractiveness of the body, per-
ceiving the repulsiveness of food, perceiving non-delight in the entire
world, contemplating impermanence in all formations ; and he has the
perception of death well established internally. He dwells depending upon
these five trainee powers : the powers of faith, moral shame, moral dread,
energy, and wisdom. These five faculties are manifest in him as weak : the
faculties of faith, energy, mindfulness, concentration, and wisdom. Because
these five faculties are weak, he sluggishly attains the immediacy condition
for the exhaustion of the influxes. This is called practice that is painful with
sluggish realization.24
What makes this meditator’s mode of practice “painful” (dukkha-

pa†ipadå) is the use of meditation subjects that focus upon the repulsive,
fearful aspects of human life, subjects that engender a mood of
disenchantment rather than of blissful absorption. It is true that the
definition ascribes to this practitioner the five faculties, among them the
faculty of concentration, sometimes defined by the jhåna formula. It is
likely, however, that this practitioner has merely a facile acquaintance
with jhåna or even none at all ; for the meditation subjects he uses are
taken up, not so much because they are conducive to the jhånas, but
24A II 150,32–51,5.
because they lead to disenchantment and detachment.25 In contrast, the
practitioner who takes the route described as “pleasant” (sukha-
pa†ipadå) is defined precisely as one who acquires the four jhånas. For
the contrast to be meaningful, one would have to conclude either that
the meditator on the “painful” path has no experience of jhåna or that
he assigns jhåna to a subordinate place in his practice. It could even be
that an alternative definition of the faculty of concentration found in the
Indriya-saµyutta is intended precisely for such kinds of practitioners.
This alternative definition defines the faculty of concentration, not as
the four jhånas, but as “the concentration or one-pointedness of mind
that arises having made release the object”.26
A similar contrast is drawn at A 4:169 between those persons who
attain nibbåna through strenuous practice (sasa!khåraparinibbåy¥) and
those who attain it through non-strenuous practice (asa!khårapari-
nibbåy¥).27 The strenuous practice is explained by way of the five
contemplations that constitute the painful path : the unattractiveness of
the body, the repulsiveness of food, perceiving non-delight in the world,
contemplating impermanence in all formations, and mindfulness of
death. The non-strenuous practice, for those fortunate ones, is nothing
other than the four jhånas. Again, a string of suttas in the A!guttara-
nikåya says of these five contemplations : (i) that they lead to complete
disenchantment, dispassion, cessation, peace, direct knowledge,
enlightenment, and nibbåna ; (ii) that they lead to the exhaustion of the
influxes ; and (iii) that they have liberation of mind and liberation by
25Itis true that Vism 265–66 explains how the meditation on bodily foulness
can give rise to the first jhåna, but the main emphasis of this meditation is on
the removal of sensual lust, not on mental absorption.
26S V 197,14–17, 198,23–24 : ariyasåvako vossaggårammaâµ karitvå labhati
samådhiµ labhati cittassa ekaggataµ, idaµ vuccati bhikkhave samådh’-

indriyaµ.
27A II 155–56.
62 Bhikkhu Bodhi
wisdom as their fruit and benefit.28
The Nikåyas never go so far as to say that those who choose such
meditation subjects as their vehicle of practice lack attainment of jhåna ;
and accordingly, when questioning the monks who claim to be arahants
liberated by wisdom, Sus¥ma does not pursue his inquiry below the
level of “the peaceful formless emancipations” and ask whether or not
they attained the jhånas. The issue is left daintily alone, as though it
were too sensitive to be touched upon. Perhaps the stock definition of
the path factor of right concentration in terms of the four jhånas, and the
role of the jhånas in the standard description of the gradual training of
the monk, occupied niches too hallowed within the canonical collection
for the Theravåda tradition to ever consider altering the received
heritage of suttas in a way that might explicitly state such attainments
are dispensable. Yet it is among those who use such subjects of
meditation as the unattractiveness of the body, mindfulness of death,
disenchantment with the world, and the impermanence of all formations
as their preferred vehicle that one might expect to find arahants
liberated by wisdom ; and because practitioners of these meditations are
contrasted with those who take the “pleasant” route of the four jhånas,
it is among the former that one might expect to find, by implication,
those who either attain jhåna with difficulty or opt instead for a mode of
practice that draws its primary strength from wisdom built upon the
minimum degree of serenity (samatha) needed to reach the destruction
of the defilements.
3. The Sukkhavipassaka Arahant and the Sus¥ma-sutta

In my reading of the Sus¥ma-sutta, the redactors of the text want to
suggest that the paññåvimutta arahants are in fact destitute of jhåna
attainments, but they dare not say this directly. That is why the
questions are not asked. The absence of the questions accomplishes two
28A III 83–84 (A 5:69–71). What is referred to here is no doubt the influx-free
liberation of mind and liberation by wisdom (anåsavå cetovimutti paññå-
vimutti) constituting arahantship.
things. It avoids the need to draw forth answers that would contradict
orthodox doctrine, which upheld the secure place of jhåna in the
structure of the Buddhist path ; and it deftly hints that these monks did
not have the jhånas. If the intention of the sutta were otherwise, Sus¥ma
could have asked about the jhånas, and the monks could have said,
“Some of us attain one jhåna, some attain two, some attain three, and
some attain all four.” But by passing over this issue in silence, they
discreetly imply that they do not attain the jhånas at all.
Where the redactors of suttas fear to tread, commentators step in
boldly. It is in the commentaries (including the Visuddhimagga) that we
first find explicit mention of the sukkhavipassaka or “dry-insight”
meditator, often in connection with passages that mention the
paññåvimutta or “wisdom-liberated” arahant. The dry-insight meditator
is defined as “one whose insight is dry, arid, because such insight is
unmoistened by the moisture of the jhånas”.29 Thus upon reaching
arahantship, such a practitioner becomes, of necessity, a wisdom-
liberated arahant. It must be borne in mind, however, that while the
dry-insight arahant is closely linked to the old canonical concept of the
wisdom-liberated arahant, a flat identity should not be drawn between
the two. Rather, the dry-insight arahant is technically only one subclass
within the broad class of wisdom-liberated arahants. The commentaries
consistently state, “The arahant liberated by wisdom is fivefold : the
dry-insight meditator together with those who attain arahantship after
emerging from any one among the four jhånas.”30 Thus the wisdom-
liberated arahants can also be those who attain the four jhånas. The
only attainments they do not achieve are the peaceful formless emanci-
pations, experience of which defines an arahant as “one liberated in
both ways”.
29Vism-mh† II 446 (VRI ed.) : so hi jhånasinehena vipassanåya asiniddha-

bhåvato sukkhå lËkhå vipassanå etassåti sukkhavipassako ti vuccati.
30Sv III 889 : so sukkhavipassako ca, catËhi jhånehi vu††håya arahattaµ pattå
cattåro cåti imesaµ vasena pañcavidho va hoti. See too Ps III 188 ; Mp IV 3 ;
Pp-a I 191.
64 Bhikkhu Bodhi
We might also note that even if the Nikåyas did envisage the
possibility of an arahant liberated by wisdom who does not attain the
jhånas, this would not mean that such a figure fulfilled the distinctive
criteria of the commentarial sukkhavipassaka arahant. For, as the name
suggests, the sukkhavipassaka is one who gives special emphasis to
vipassanå or insight meditation ; the commentaries and subcommen-
taries in fact often speak of this meditator, prior to attaining arahantship,
as the vipassanåyånika, “one who makes insight the vehicle”, or even as
the suddhavipassanåyånika, “one who makes bare insight the
vehicle”.31 These designations imply that at some point in the evolution
of Theravåda meditation theory, the practice of vipassanå came to be
regarded virtually as an autonomous means to realization that could be
undertaken quite independently of any supporting base of samatha. It is
quite conceivable that if the Nikåyas did see, even tacitly, the possibility
of wisdom-liberated arahants destitute of jhåna, they still would have
assumed these arahants had a minimal foundation of samatha. For such
arahants, it would just be the case that their practice of samatha did not
reach the level of the first jhåna.
Now while the concept of the dry-insight arahant is first introduced
in the commentaries, as often happens the commentators peer back into
the suttas to seek substantiation for their hermeneutical innovations.
And, sure enough, “seek and ye shall find”. Not to be left empty-
handed, the commentators find evidence for the dry-insight arahant in
several texts of the Nikåyas, and one of these that is given star billing is
the Sus¥ma-sutta. The Sus¥ma-sutta itself, as we saw, does not specify
where the monks liberated by wisdom stood in relation to the jhånas.
For all we know, based on the text alone, they could have been adepts in
all four jhånas. The commentary, however, apparently drawing upon
ancient oral tradition, fills in the gaps in the information we can derive
from the sutta itself with additional information apparently transmitted
in the lineage of teachers. Thus in the sutta, in reply to Sus¥ma’s
31See Vism-mh† II 351, 438, 474 (VRI ed.).

question, “Without the super-knowledges and formless attainments,
how can you claim to be arahants ?” the monks reply, “We are liberated
by wisdom.” In glossing these words, the commentary tersely says,
“Liberated by wisdom : ‘We are without jhåna, dry-insight meditators
liberated simply by wisdom only.’”32 Later, when interpreting the
Buddha’s exchange with Sus¥ma, the commentary says in regard to a
statement of the Buddha, “The purpose is to show the arising of
knowledge thus, even without concentration. This is meant : ‘Sus¥ma,
the path or fruit is not the outcome, benefit, and product of
concentration, but the outcome, benefit, and product of insight.’” 33And
the †¥kå or subcommentary to this passage, commenting on the words
“even without concentration” says, “This is said referring to the
meditator who makes insight the vehicle ; it means even without
previously achieved concentration that has reached the mark of
serenity.”34
4. The Chinese Parallels to the Sus¥ma-sutta

At this point it will be illuminating to turn to the parallels to S
12:70 preserved in Chinese translation. In M-Vin, Sus¥ma inquires from
the monks, not about all five super-knowledges, but only about the
divine eye that sees how beings pass away and take rebirth according to
their kamma, and about the recollection of past lives — the last two of
these super-knowledges, given here in inverse order from S 12:70 — as
well as about the peaceful formless emancipations. As in S 12:70, the
monks deny possessing these attainments. When Sus¥ma asks them how
they could declare final knowledge in the Buddha’s presence, they
32Spk II 127 : paññåvimuttå kho mayaµ, åvuso ti, åvuso, mayaµ nijjhånakå
sukkhavipassakå paññåmatten’ eva vimuttå ti dasseti.
33Spk II 127 : vinå pi samådhiµ evaµ ñåûppattidassanatthaµ. idañ hi vuttaµ
hoti : Sus¥ma, maggo vå phalaµ vå na samådhinissando, na samådhi-

ånisaµso, na samådhissa nipphatti ; vipassanåya pan’ eso nissando,
vipassanåya ånisaµso, vipassanåya nipphatti.
34Spk-p† II 107 (VRI ed.) : vinå pi samådhin ti samathalakkhaâppattaµ
purimasiddhaµ vinå pi samådhin ti vipassanåyånikaµ sandhåya vuttaµ.

66 Bhikkhu Bodhi
reply, “We are wisdom-liberated ones.”35 Thus M-Vin is fairly close to
S 12:70. It is virtually impossible to judge which is likely to be more
original, the five super-knowledges of S 12:70 or the two mentioned in
M-Vin. The twofold scheme has the advantage of economy, and greater
detail usually suggests lateness ; but if the sutta originates from a real
historical incident, it could well have been that the wanderers suspected
the Buddhist monks to be adepts in the spiritual powers and mind-
reading, the first and third super-knowledges, which ostentatious
ascetics would be most likely to use to impress gullible lay devotees
(particularly in a royal capital like Råjagaha).36 This would then better
explain Sus¥ma’s skepticism that there could be wisdom-liberated
arahants who lack such powers.
It is with SÓ 347 and the Vibhå"å version that the divergences from
S 12:70 become significant, for these versions straight away transform
the monks into Sarvåstivådin counterparts of the Påli commentarial dry-
insight arahants. Though these versions do not have a neat appellation
for this figure, it is evident that the text here wants the expression
“wisdom-liberated one” (慧解脫) to convey very much the same idea
that the expression paññåvimutta as used in the Sus¥ma-sutta conveys
for the Saµyutta Commentary : one liberated without jhåna (ni-
jjhånaka), entirely through wisdom. As in S 12:70 and M-Vin, so here
Sus¥ma enters the Sa!gha for the purpose of “stealing” the Dhamma. He
is ordained at the Enlightened One’s behest, but here it is stated that the
Buddha already knew his intention in going forth. A fortnight after his
ordination, one monk, at the head of a group of monks, tells him he
should be aware that they have all won the goal. Sus¥ma then asks him
whether he has attained the first, second, third, or fourth jhånas, or the
peaceful formless emancipations ; and, he adds to each question, “by the
35T22, 363a14 : 我是慧解脫人。

36See the Buddha’s explanation of the dangers in “the miracle of spiritual
powers” (iddhipå†ihåriya) and “the miracle of thought-reading” (ådesanå-
på†ihåriya) at D I 212–14.
non-arising of any influxes is your mind well liberated ?”37 In each case,
the monk answers no, the purport being that they have exhausted the
influxes and gained full liberation of mind without relying on any
jhånas or formless attainments. Finally Sus¥ma exclaims, “How could
this be ? What you have said is inconsistent ; your later [words]
contradict your earlier [words]. How is it that you don’t attain jhåna, yet
you make a declaration [of arahantship] ?” The monk then says, “I am
liberated by wisdom.”38 Thereupon the whole group of monks depart.
Realizing that he now needs clarification, Sus¥ma decides to seek help
from the Buddha.
The Vibhå"å version of the sutta is evidently later than the others,
at least in its final redaction, for it speaks of Sus¥ma, after taking full
ordination, as “having read and recited the Tripi†aka”,39 a phrase that
refers anachronistically to the classification of the sacred scriptures into
the “three baskets”, perhaps even in written form. In this version,
Sus¥ma asks the monks, “Was it on the basis of the first jhåna that you
attained the exhaustion of the influxes ?” They answer no. Then : “Was
it on the basis of the second, third, or fourth jhånas, or the peaceful
formless emancipations that you attained the exhaustion of the
influxes ?” They answer no. Sus¥ma then says, “Then without depending
on any jhåna you attained the exhaustion of the influxes ! Who can
believe that ?” The monks then say, “We are liberated by wisdom.”40
37T2, 97a7–18 : 不起諸漏。心善解脫耶. It must have been this phrase that led
Gombrich to suppose that the head monk “cannot even claim that they are free
of greed and hatred” and that the sËtra is “most uncomplimentary to a group
of monks” (How Buddhism Began, p. 124). The sequel to this passage,
however, leaves no doubt that the monks have terminated the åsavas and are
well liberated in mind. The readings in the Vibhå"å version, moreover,
corroborate this interpretation.
38T2, 97a19–21 : 云何。尊者所說不同。前後相違。云何不得禪定而復記說
比丘答言。我是慧解脫也。
39T28, 408a20 : 讀誦三藏
40T28, 408a29–b3 答曰。不也。依第二第三第四禪及過色無色寂靜解脫得
盡漏耶。答言不也。時蘇尸摩復作是言。汝等既不依禪定。而得盡

漏。誰當信耶。時諸比丘皆作是言。我等是慧解脫。 In the version at
68 Bhikkhu Bodhi
Then, as in the other versions, Sus¥ma turns to the Buddha for help. The
Buddha explains, “Those monks first exhausted the influxes based on
the access to the jhåna, and afterwards aroused the basic jhåna.”41
Though SÓ 347 and Vibhå"å differ in details, they are both
regarded as affiliated in some way with the Sarvåstivåda, which must
have been, not one unified entity, but an umbrella term for a range of
schools with collections of texts that showed considerable variation,
perhaps stemming from wide geographical distribution and long
duration through time. Since the Sarvåstivåda and the Theravåda are
widely recognized to be two branches of the old Sthaviravåda, and the
first major schism in the archaic Sa!gha resulted in its bifurcation into
the Sthaviravåda and the Mahåså!ghika, when the readings in one
Sthaviravåda version and a Mahåså!ghika version of a text agree and
the reading in another Sthaviravåda version diverges, it is likely that the
latter results from a later alteration or transmutation in the text. Of
course, we cannot always be absolutely certain that this is so, but the
above stipulation is generally a safe guideline to follow, and in the case
of SÓ 347 and Vibhå"å, quite apart from the passages with doctrinal
ramifications, a number of other “fingerprints” suggest that these
versions are less archaic than S 12:70 and M-Vin. One example is a
certain flair for detail in SÓ 347 ; another is the reference to the
Tripi†aka in Vibhå"å.
T27, 572c16–17, Sus¥ma asks the monks : “Was it on the basis of the first
jhåna up to the base of nothingness that the venerable ones attained
realization ?” (仁等所證依何定耶。為初靜慮為乃至無所有處耶). And to
this they answer no.
41T 28, 408b9-10 : 彼諸比丘。先依未至禪盡漏。後起根本禪。 The “access
to the jhåna” (未至禪, lit. “not-yet reaching jhåna”) is presumably a state
similar to upacåra-samådhi, the access concentration of the Påli com-
mentaries. This Vibhå"å passage does not altogether deny that these arahants
can possess jhåna, but the jhåna it allows them seems to correspond to the
minimal first lokuttara-jhåna that the Påli commentaries ascribe to the
sukkhavipassaka arahants.
5. Assessment
SÓ 347 and Vibhå"å thus present us with an interesting case where
the Sarvåstivåda recensions of a sutta give utterance to an idea that is
not found in the Theravåda version but was not unknown to the
Theravåda tradition, namely, the idea of an arahant who has reached
final liberation without attainment of the jhånas. In the Theravåda
tradition, however, this idea came to open expression only in the
commentaries, with the concept of the sukkhavipassaka or dry-insight
arahant. This idea must have gained such prestige that it presented
itself, either to the ancient anonymous authors of the lost Sinhala
commentaries, or to Ócariya Buddhaghosa, the compiler of the present
Påli commentaries, as the key to understanding the paññåvimutta
arahants of the Sus¥ma-sutta. In this way, the sukkhavipassaka arahant,
though hidden behind the text of the Sus¥ma-sutta itself, found a secure
lodging in its commentary.
Several canonical texts, however, suggest that even prior to the
commentarial period the archaic concept of the paññåvimutta was
already being reinterpreted in the direction of the dry-insight arahant.
We saw above that, according to the K¥†ågiri-sutta, the “arahant
liberated by wisdom” was distinguished from the “arahant liberated in
both ways” with respect to their relationship to the peaceful formless
emancipations. The latter can attain them ; the former cannot.42 In the
Puggalapaññatti, the fourth book of the Abhidhamma-pi†aka, these
definitions are subtly rephrased. The arahant liberated in both ways is
now “a person who dwells having contacted the eight emancipations
with the body, and having seen with wisdom, his influxes are
exhausted”.43 And, corresponding to this, the arahant liberated by
42See above, p. 55.

43Pp14 :idh’ ekacco puggalo a††ha vimokkhe kåyena phusitvå viharati ; paññåya
c’ assa disvå åsavå parikkh¥ˆå honti. The eight emancipations are not
identical with the four jhånas and the four formless attainments. The first
three emancipations are equivalent to the four jhånas, but they deal with the
state of jhåna in terms of its objects rather than in terms of its subjective
70 Bhikkhu Bodhi
wisdom is “a person who does not dwell having contacted the eight
emancipations with the body, but having seen with wisdom, his influxes
are exhausted”.44 The subtle change in wording between this definition
and that in the K¥†ågiri-sutta, an almost inconspicuous change from
“peaceful emancipations, transcending form, that are formless “ (santå
vimokkhå atikkamma rËpe åruppå) to “eight emancipations” (a††ha
vimokkhå), makes a world of difference with regard to meaning. The
new definition gracefully suggests that the arahant liberated by wisdom
need not possess any of the eight emancipations, including the lower
three, which comprise the four jhånas. Although the Puggalapaññatti
Commentary glosses these two definitions in the same way that it does
the older definitions, the new definition opens the door just a crack —
but does indeed open it — for admitting the dry-insight arahant into the
chamber of figures duly ordained by canonical authority.
Since the Puggalapaññatti is an Abhidhamma tract and thus of later
provenance than the Nikåyas, it may not be altogether surprising to find
a revised definition of the two types of arahants there. But it is a bit
astonishing to find the above definitions actually incorporated into a
sutta. A discourse in the A!guttara-nikåya called the Putta-sutta (A
4:87) distinguishes four types of ascetics, among them one known as a
red-lotus ascetic and another known as a white-lotus ascetic. The red-
lotus ascetic (samaâpaduma) is defined as a monk who has realized by
direct knowledge the taintless liberation of mind, liberation by wisdom,
and dwells in it ; and who also dwells having contacted with the body
the eight emancipations. The white-lotus ascetic (samaâpuˆ"ar¥ka) is
“a monk who, with the destruction of the taints, has realized for himself
with direct knowledge, in this very life, the taintless liberation of mind,
experience. Thus all four jhånas can be based on each of the first three
emancipations, since all four jhånas can experience their object in the way
defined by each of these emancipations. Emancipations 4–7 are the four
formless attainments, and the eighth emancipation is the cessation of
perception and feeling (saññåvedayitanirodha).
44Pp 14 : idh’ ekacco puggalo na h’ eva kho a††ha vimokkhe kåyena phusitvå
viharati paññåya c’ assa disvå åsavå parikkh¥ˆå honti.

liberation by wisdom ; and having entered upon it, dwells in it ; yet he
does not dwell having contacted with the body the eight emanci-
pations”.45 The commentary identifies the red-lotus ascetic with the
ubhatobhågavimutta, the arahant liberated in both ways, since the sutta
definition here matches the Puggalapaññatti’s definition of this type ;
but it does not simply identify the white-lotus ascetic point-blank with
the paññåvimutta, the arahant liberated by wisdom. It says, rather, “By
this he shows the dry-insight arahant.”46
It seems to me that there are two ways to account for the definitions
of the two types of lotus-ascetics in the Putta-sutta. Either the sutta
itself is a later composition that was inserted into the A!guttara-nikåya,
incorporating the new definitions of the ubhatobhågavimutta arahant
and the paññåvimutta arahant that were being framed during the period
of scholastic elaboration that brought the Puggalapaññatti into being.
Or, alternatively, the sutta itself is archaic, but an original version
employing definitions of the two types of arahants matching those in
the K¥†ågiri-sutta had been “updated” to accommodate the new
definitions that were canonized by the Puggalapaññatti. I would
suggest, too, that “behind the scenes” the mode of thought that
influenced the definitions of the two lotus-ascetics of the Putta-sutta
was also exerting its influence on the interpretation of the Sus¥ma-sutta.
So, while the wording of the Påli version of the Sus¥ma-sutta was not
altered and it could thus still be interpreted as simply denying that the
monks declaring arahantship possessed the super-knowledges and the
formless emancipations, among an influential body of early Påli
exegetes it was already being seen as a paradigmatic text for the figure
of the dry-insight arahant.
Within the Theravåda school, this interpretation of the sutta first
came to literary expression in its commentary. In contrast, among those
45A II 87,7–11 : idha, bhikkhave, bhikkhu åsavånaµ khayå anåsavaµ ceto-

vimuttiµ paññåvimuttiµ di††h’ eva dhamme sayaµ abhiññå sacchikatvå
upasampajja viharati, no ca kho a††ha vimokkhe kåyena phusitvå viharati.
46Mp III 113 : iminå sukkhavipassakakh¥ˆåsavaµ dasseti.
72 Bhikkhu Bodhi
in the broad Sarvåstivåda camp, a parallel commentarial stance towards
the sutta had already become strong enough to “burst the bonds” of
sacerdotal hesitancy and force its way into the primary text itself. In this
camp, the sutta must have become altered in such a way as to grant
canonical legitimacy to the figure of the arahant wholly bereft of the
jhånas. Not only do the wisdom-liberated monks of SÓ 347 and the
Vibhå"å version deny that they possess the four jhånas, but the
Sarvåstivåda commentaries even redefine the concept of the
paññåvimutta in such a way that any possession of a jhåna by a
paññåvimutta arahant “compromises” and “corrupts” the purity of his
possession of paññåvimutta arahantship. We thus find that the
Abhidharma-mahåvibhå"å-ßåstra, the great commentary of the Kashmiri
Sarvåstivådins, distinguishes two kinds of paññåvimutta arahant,
making specific reference to the Sus¥ma-sutta :
Question : It is said here several times that one liberated by wisdom [can]
arouse the knowledge of others’ minds. This necessarily depends on the
basic jhåna. But if the wisdom-liberated one can arouse the basic jhåna,
does this not contradict the Sus¥ma-sutta ? In that sutta it is said, “The
wisdom-liberated one cannot arouse the basic jhåna.”
Reply : There are two kinds of wisdom-liberated ones, the partial and the
complete. The one partially liberated by wisdom (少分慧解) is able to
arouse one, two, or three among the four jhånas. The one completely
liberated by wisdom (全分慧解脫) cannot arouse any of the four jhånas. …
The Sus¥ma-sutta speaks about the one completely liberated by wisdom,
who cannot arouse any of the four jhånas.47
The earliest version of the Sus¥ma-sutta received by the archaic

Sarvåstivåda school, we might suppose, was probably quite similar to S
12:70 and M-Vin regarding the qualities denied of the paññåvimutta
arahants. During the historical evolution of the sutta, however, these
came to be altered, transformed into a complete denial that they possess
the jhånas. While it is possible that such alterations could have occurred
by unconscious habit in the course of oral transmission, given that the
47 T 27, 564b5–13.
Sarvåstivåda came to place such a strong emphasis upon the absence of
jhånas in the ideal type of paññåvimutta arahant, we may suspect that
the alteration was deliberate, done under pressure from the evolving
Sarvåstivådin doctrinal system.
Unlike the Sarvåstivådin commentators, the Theravådins never
went so far as to distinguish degrees among the paññåvimutta arahants.
According to the Påli commentaries, all five types — those arahants
who attain any of the four jhånas and the dry-insight arahant — are
equally entitled to be called “liberated by wisdom,” as long as they do
not attain the peaceful formless emancipations. But despite this
“official” breadth recognized in the term paññåvimutta, one can detect
in certain texts a subtle shift taking place in its “weight” towards the
dry-insight arahant. This is manifest in the definition of the white-lotus
ascetic of the Putta-sutta as one without the eight emancipations,
defined by the A!guttara Commentary as the dry-insight arahant ;
again, it appears in a similar definition of the “one liberated by wisdom”
in the Puggalapaññatti ; and it crops up still again in the commentarial
gloss on the expression “liberated by wisdom” in the Sus¥ma-sutta as
“without jhåna, dry-insight meditators liberated simply by wisdom
only”. Still another example is seen in the commentarial gloss on the
word paññåvimutto occurring at A I 74. Here, the commentary succinct-
ly says, “Liberated by wisdom : the dry-insight influx-destroyer [i.e.,
arahant].”48 The †¥kå to this passage does not state that this is said
merely to exemplify the family of wisdom-liberated arahants, but
reinforces the idea that the wisdom-liberated arahant lacks attainments
in samatha : “Liberated by wisdom : liberated by the wisdom of the
supreme path [of arahantship] without a support of serenity.”49
We thus see that at the commentarial level, the Theravåda wound
up with an interpretative concept that closely matched an idea that the
Sarvåstivådins had already inserted into texts they regarded as sËtras
48MpII 147 : paññåvimutto ti paññåya vimutto sukkhavipassakakh¥ˆåsavo.

49Mp-†II 38 (VRI ed.) : paññåya vimutto ti samathasannissayena vinå agga-
maggapaññåya vimutto.
74 Bhikkhu Bodhi
coming directly from the Buddha’s own mouth, namely, the idea of an
arahant liberated without attainment of the jhånas. To what extent this
idea is already intended by the Påli suttas that speak about monks who
attain arahantship via the “painful” or “strenuous” path of meditation on
the unattractive nature of the body, the repulsiveness of food, the
inevitability of death, and other topics “tending to disenchantment” is
difficult to determine. What is certain, however, is that the Påli suttas
never explicitly admit the existence of arahants who altogether
dispense with the jhånas. It might also be important to note that the
Sarvåstivådins did not adopt the term sukkhavipassaka or any other
term that quite matches it. While this may be just a matter of
circumstance, simply because the term sukkhavipassaka arose in an
exegetical camp geographically far removed from their own centers of
activity, another more fundamental reason may also be involved. It is
possible that the Sarvåstivådins did not speak of a “dry-insight arahant”
because they never introduced the scission between samatha and
vipassanå as sharply as the Theravådin commentarial tradition did but
saw the path of any paññåvimutta arahant to involve an interplay of
these two meditative factors. To qualify as a “full paññåvimutta”
arahant, as one who is utterly bereft of jhåna, this practitioner must end
the development of samatha at a level called såmantakadhyåna,
“threshold meditation”, corresponding to access concentration (upa-
cårasamådhi) of the Theravåda commentaries. But, it seems, they never
conceived the idea of a meditator “who makes [bare] insight the
vehicle” (vipassanåyånika, suddhavipassanåyånika), the distinctly
Theravådin notion of the meditator who eventually reaches final fruition
as a dry-insight arahant.
Nevertheless, despite these differences, both these major Stha-
viravåda traditions, as well as the Mahåså!ghikas, have preserved
versions of the same story telling how an ascetic named Sus¥ma
infiltrated the Buddha’s Sa!gha and entered into dialogue with the
paññåvimutta monks and with the Tathågata himself. In the two
traditions that we know most about, the encounter became an important
canonical pillar for the belief that arahantship was possible without the
jhånas. In one tradition this idea was stated explicitly in the sutta itself ;
in the other it rested upon the explication of the text by the commentary.
But as I read it, even the older version of the sutta, S 12:70 and perhaps
too M-Vin, originally intended to establish the possibility of arahantship
without the jhånas. That is, I suspect that the questions Sus¥ma posed to
the paññåvimutta monks in this version, which confirm their lack of the
super-knowledges and the formless attainments but stop short at the
jhånas, were also intended to hint, by their very silence, that similar
questions could have been asked about the jhånas, and that the same
answer would have been given. If such is the case — and I must
emphasize that this is largely intuition on my part — we could then
understand that the compilers of the versions I have called SÓ 347 and
Vibhå"å did not so much add anything new as simply state explicitly
what the redactors of the older version had wanted the sutta to convey
from the start.
Though I say “this is largely intuition on my part”, I do have
reasons for this intuition. Apart from those I have brought forth above,
there is also the ensuing dialogue between Sus¥ma and the Buddha, on
the grounds for the possibility of paññåvimutta arahantship. This,
however, is a major topic in itself, which I intend to treat in the sequel
to the present paper.
Bhikkhu Bodhi
A Note on the Heterodox Calendar and a Disputed
Reading in the Kålakåcåryakathå
K.R. Norman is no doubt best known among Indologists for his

penetrating publications on the Påli Canon and the Theravåda Buddhist
textual tradition. It should, however, be emphasised that the particular
authority of his extensive philological work, whether editorial or
etymological in orientation, derives from its firm grounding upon a
detailed knowledge of a wide range of Middle Indo-Aryan dialects
whereby Påli is not to be studied in exclusively Buddhological terms
but rather as representing a component of a wider linguistic matrix in
which the insights gained from the phonology, grammar, and lexicon of
Ardhamågadh¥ and other Pråkrits play a vital elucidatory role. Further-
more, the editions and metrical analyses of short Jain canonical texts
and discussions of various linguistic and doctrinal parallels between
early Buddhism and Jainism to be found throughout the eight volumes
of his Collected Papers make amply clear that Mr Norman’s career-
long preoccupation with Påli philology has not led to Jain studies’ loss
being Buddhist studies’ gain.1 In this light I trust that my former teacher
will not think it inappropriate if in a congratulatory volume which
contains contributions dealing almost exclusively with Buddhist matters
I proceed to discuss a topic relating to Jainism which abuts on
Buddhism only tangentially but nonetheless represents a matter of no
little concern to the two heterodox, that is non-brahmanical, traditions,
namely the configuration of the ritual calendar.
As is well known, adherents of the early ßramaâ orders assembled
at various times of the month to preach their respective doctrines and
My thanks to John Cort for reading a draft of this paper.

1See Norman 1990–2001. As a postgraduate Mr Norman had originally con-
templated an edition of the Jain canonical text, the Sthånå!ga SËtra, a project
aborted because of lack of accessible early manuscripts at the time (personal
communication).
78 Paul Dundas
perform communal rituals of purification and solidarity. This is clearly
witnessed by early Buddhist and Jain sources. Thus the Mahåvagga of
the Theravåda Vinaya describes this custom as taking place on the
fourteenth, fifteenth, and eighth days of the half-month2 and goes on to
portray the Buddha as sanctioning the recitation of the På†imokkha on
the fifteenth day as an uposathakamma,3 while the Jain SËtrakÁta!ga
SËtra refers to the posaha being correctly observed on the fourteenth
and eighth days, on designated dates and on full-moon days.4
Notwithstanding the brahmanical origins5 of the term used for this
important day of observance,6 the Vedic ritual calendar proved
unacceptable to renunciatory groups such as the Jains and Buddhists
who wished to distance themselves from brahman customs and initially
a lunar calendar seems to have been adopted by them by way of
differentiation. However, the latter was in turn to be challenged by a
lunisolar calendar, of Greek origin and in use by around 380 C.E., which
gained ground at the expense of the former mode of reckoning through
its attempt to reconcile the 354 days of the lunar calendar with the 365"
2Vin I 101,4–6 #: tena kho pana samayena aññatitthiyå paribbåjakå cåtuddase

pannarase a††hamiyå ca pakkhassa sannipatitvå dhammaµ bhåsanti.
3Vin I 102,22–24.
4SËtrakÁtå!ga SËtra 2.2, pp. 188–89 #: cåuddasa††hamuddi††hapuˆâmåsiˆ¥su
pa"ipuˆâµ posahaµ sammaµ aûpålemåˆå. Cf. 2.7, p. 2 5 0 , where

JambËvijaya reads posadhaµ.
5The common source of the term variously rendered by the Buddhists and Jains
as uposatha, po#adha, and posaha (signifying both the day and the observance
connected with it) is upavasatha, used in Vedic texts of a particular form of
overnight fast associated with the full-moon sacrifice. Tieken (2000,
pp.#11–13) argues for the Buddhist uposatha as the counterpart of the secular
ak#apa†ala ceremony at which the king and his functionaries rendered account
of activities conducted during the previous eight months.
6In medieval Jainism, po#adha came to mean the day of the moon’s periodic
change and the fast carried out thereon, while today it designates a contem-
plative exercise structured over a half or whole day which is most generally
observed at the time of Paryu$aˆ (see below). See Cort 2001, p. 123, and
Williams 1963, p. 142. This note does not deal with the Jain po#adha ritual as
such.
A Note on the Heterodox Calendar 79
days of its solar equivalent. Adoption of one or the other of these
calendars was eventually to be among the strategies involved in the
formulation of sectarian identity amongst the Buddhists and so the
MËlasarvåstivådins came to organise their ritual calendar on the
lunisolar model, while the Theravådins used the older lunar model.7
MËlasarvåstivådin sources describe how that particular Buddhist nikåya
customarily performed the half-monthly po#adha ceremony involving
the recitation of the code of monastic law on the fifteenth day of the
fortnight, or, as a result of calendrical circumstances which necessitated
the omission of one day, exceptionally on the fourteenth day also, thus
ensuring that the observance always fell on a full-moon or new-moon
day.8 Although the Po$adhavastu of the MËlasarvåstivådin Vinaya
allows for special procedures to be permitted for monks observing
po#adha on the fourteenth day if they were visting a monastery where
the ceremony was normally held on the fifteenth day, the Pravrajyåvastu
asserts firmly that customary observance of po#adha on the fourteenth
day only was a practice of heretical sectarians (t¥rthya).9
It is most likely that the MËlasarvåstivådins regarded these heretics
as being the members of other Buddhist nikåyas, but we can assume that
the Jain community in the early common era was also caught up in the
7Vogel 1997, pp. 678, 680, and 686.

8“[A lunar] year is made up of twelve synodic months of about 29% days each,
this being the interval between two successive full (or new) moons styled a
lunation. If the Po$adha ceremony had been celebrated on the fifteenth
throughout, after two months already it would have taken place one day later
than [the] full moon, which would surely have been noticed by the monks. It is
for this reason that one day was dropped, and the function was held on the
fourteenth instead of the fifteenth every other month. Hence Po$adha always
fell on [the] full-moon or new-moon day” (Vogel 1997, p. 678). For a Jain
awareness of this situation, cf. the fourteenth-century Gurutattvaprad¥pa (see
note 36) 4.17. At an early date a rule was formulated by the Buddhist com-
munity to allow for three po#adha days for the laity every fortnight, namely
the eighth, fourteenth, and fifteenth. See Dietz 1997, p. 63, and cf. Hu-von
Hinüber 1994, p. 21, and Hureau 2006, p. 102.
9Vogel 1997, p. 678.
80 Paul Dundas
difficulties contingent upon calendrical innovation, even if detailed
contemporary evidence equivalent to that provided by the Buddhists is
lacking from this source. It is against this background that we can
proceed to discuss the main change in the dating system of the Jains,
introduced according to tradition by the teacher Kålaka in 466 C .E .
whose career is described in the cycle of hagiographies which can be
collectively and individually entitled the Kålakåcåryakathå, and the
possible implications of this for disciplinary observance held on the
fourteenth day of the month.10
The earliest account of Kålaka’s redating of Saµvatsar¥, the day
communally devoted to pratikramaâ, the general repentance of trans-
gressions (in this case, those committed in the previous year) which is
the culmination of the festival of Paryu$aˆ, occurs during the tenth
uddeßaka of the Niß¥tha CËrî by Jinadåsa (c . seventh century).11
Kålaka is portrayed there as authorising while in Prati$†håna the shift of
the date of Saµvatsar¥ from its traditional location on the fifth day of
the month of Bhådrapada to the fourth day at the behest of King
Íåtavåhana in order to avoid a clash with a Hindu festival. The portion
of the Niß¥tha CËrni version of the story most relevant to the present
discussion can be rendered as follows : “So Paryu$aˆ (i.e. the
#
Saµvatsar¥ pratikramaâ ceremony) must be performed on the fourth
10See Brown 1933. This Kålaka is no doubt not identical to the teacher of the
same name who according to the Kålakåcåryakathå invited the Íakas into
western India to revenge an insult to his sister. However, this is not relevant to
the present paper. The story of Kålaka became associated with the Kalpa
SËtra, the central text of the Ívetåmbara Jain festival of Paryu$aˆ. The
Gurutattvaprad¥pa 4.22 autocommentary states that there existed no early
(ådya) manuscript of the Kalpa SËtra which did not contain the story of Kålaka
and that the latter must have been included when the former was first written
down as an individual text (pÁthaglikhita). Both the Kalpa SËtra and the
Kålakåcåryakathå thus have equal antiquity and authority. The Gurutattvapra-
d¥pa also suggests (4.24 autocommentary) that the story of Kålaka originated
about two and a half centuries after its hero’s life.
11Niß¥tha CËrî on bhå#ya verse 3153, p. 131. For pratikramaâ in general, see
Cort 2001, pp. 123–24.

day [of Bhådrapada]. In this way the mighty teacher of this epoch
(Kålaka) caused the fourth day to be introduced [as the day for the
observance of Saµvatsar¥] since there was an appropriate motive for
this. That same date was approved by all the monks.”12 Later writers
(see below) also refer to the Paryu$aˆåkalpa CËrî, possibly approxi-
mately contemporary with the Niß¥tha CËrî, which gives as the
equivalent of its version, “[The Saµvatsar¥ day of] Paryu$aˆ was
performed on the fourth day of the month ; in this way the fourth day
#
became [a festival] inspired by an [appropriate] motive.”13 In other

words, the Niß¥tha CËrî and Paryu$aˆåkalpa CËrî record the fact that
a significant modification in the ritual calendar entered the realm of
Ívetåmbara Jain customary practice for a practical reason and with
general approval.
Versions of the story of Kålaka found in Ívetåmbara Jain narrative
collections in the later centuries of the first millennium C .E . such as
JayasiµhasËri’s Dharmopadeßamålåvivaraâ do not appear to concern
themselves with the finer issues of the account of the redating of
Saµvatsar¥. However, versions of the story produced at the beginning of
the second millennium introduce a statement describing the possible
implications of Kålaka’s redating which was to prove highly
controversial.14 An early example occurs in the Kålakåcåryakathånaka
found in the commentary of 1089–90 C . E . by DevacandrasËri (the
teacher of the celebrated Hemacandra) on PradyumnasËri’s MËla-
ßuddhiprakaraâ.15 After providing an approximate reproduction of the
12tåhe cautth¥e pajjosaviyaµ. evaµ jugappahåêhi cautth¥ kåraê pavattitå. sa

ccevåûmatå savvasåhËâµ
13cautth¥e katå pajjosavaˆå, evaµ cautth¥ vi jåyå kåraîyå. See Kulamaˆ&ana-
sËri (1353–99), VicåråmÁtasårasaµgraha, p. 34, and Gurutattvaprad¥pa, p. 66. I

have not had access to a published version of the Paryu$aˆåkalpa CËrî and
am not confident that one exists.
14Brown (1933, p. 2) suggests that the Kålakåcåryakathå cycle may have
emerged in the twelfth century.

15MËlaßuddhiprakaraâ, p. 130.
82 Paul Dundas
statement in the Niß¥tha CËrî given above16 and an epitomising verse
delineating the overall context of the redating as being prompted by
King Íåtavåhana,17 DevacandrasËri continues : “And because of that
#
(tavvaseâ) (i.e. the redating), the [regular] fortnightly observances [of

pratikramaâ] ( pakkhiyåiµ) were performed on the fourteenth day of
the month, which had otherwise been described in scriptural tradition
[as to be performed] on the full-moon day (i.e. the fifteenth day).”18
That is to say, Kålaka’s redating of Saµvatsar¥ is held to be responsible
for the relocation of another important observance, the fortnightly act of
repentance, which had previously (that is, back to the time of Mahåv¥ra)
been practised on the full-moon day,19 to the fourteenth day of the
month.20 The tavvaseâ statement also occurs in the first serious
16evaµ ca kåraêâ Kålagåyariehiµ cautth¥e pajjosavaâµ pavattiyaµ

samatthasaµgheâ ya aûmaˆîyaµ.
17kåraîyå ya cautth¥ ceiyajaisåhuvåsaânimittam |uddisiya Såtavåhaâ
påya††iyå Kåliya ’jjeâ.

18tavvaseâ ya pakkhiyå¥î vi cauddas¥e åyariyåî, aˆâhå ågamottåî
puˆîmåe. This will be designated hereafter as the “tavvaseâ statement”. The
spelling puˆîmåe/ punnimåe alternates in the various sources.
19See Renou and Filliozat 1953, p. 734, for the bright half (ßuklapak#a) of the
fifteenth day also being called pËrîmå (or some derivative of it).
20In his defence of the impossibility of any scriptural warranty for the fort-
nightly pratikramaâ being observed on the full-moon day, the fourteenth-

century Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, p. 24, specifically con-
nects the tavvaseâ statement with DevacandrasËri, albeit locating it in his
commentary on the Sthånå!ga SËtra, a work which seems to have been lost,
rather than his commentary on the MËlaßuddhiprakaraâ. However, Kula-
maˆ&anasËri further points out that in that same work there also occurs
confirmation of the general authority for the fortnightly observance taking
place on the fourteenth day, that is without any reference to a supposedly
scripturally sanctioned full-moon day observance, and he ascribes these two
differing judgements apparently made by DevacandrasËri to the influence of
the current time of decline ( janånåµ kålado#aviße#eâ). For a reference to the
Sthånå!ga SËtra commentary in conjunction with a version of the Kålak’-
åcåryakathå which possibly relates to the topic under discussion in this study,
see Catalogue 2006#: No. 668 (entry for Nånåvicåra-ratna-saµgraha), p. 472#:
iti ßr¥-DevacaµdrasËri-kÁtåyåµ Èhånå-vÁttau, tathå kasyåµcit Kålakåcårya-
modern edition of the Kålaka story, that of Jacobi, which was based on
a manuscript dating from 142821 and was to be drawn on by Brown in
his study of 1933 for an edition of what has come to represent the best-
known telling of the story, the so-called “Long Anonymous Version”.22
On the face of it, the assertion of the tavvaseâ statement that the
fortnightly pratikramaâ must be performed on the fourteenth day of
the month, when in fact scripture had originally stipulated that it should
take place on the full-moon day, might seem to represent a simple
acknowledgement of the necessity for a further adjustment of the ritual
calendar as a knock-on effect in the wake of the redating of Saµvatsar¥
to one day earlier than had been the ancient practice. However, the
authoritative ninth-century commentator Í¥lå!ka, elucidating SËtra-
kÁtå!ga SËtra 2.7.3 (~2.2.76) which describes the various days on which
po#adha could be celebrated (see above), connects only the three four-
monthly (caturmåsaka) pratikramaâ observances, in which repentance
is offered for transgressions in the previous four months, with full-moon
days, not the fortnightly pratikramaâ,23 and by the eleventh century
(that is, at the approximate time when developed versions of the
Kålakåcåryakathå were starting to appear) there can be found increasing
evidence that a connection of the fortnightly pratikramaâ with the
supposedly scripturally sanctioned full-moon day was difficult, or
indeed impossible, for elements of the Ívetåmbara Jain community to
accept. This situation is signalled by Bhojak, the editor of Deva-
kathåyåm api. Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, p. 26, describes

DevacandrasËri, the commentator on the Sthånå!ga SËtra, as being a
contemporary of King Jayasiµha Siddharåja and thus clearly the same teacher
as Hemacandra’s guru.
21Jacobi 1880, p.271.
22Brown 1933, pp. 36–52#; for the tavvaseâ statement, see p. 47.
23Í¥lå!ka, commentary on SËtrakÁtå!ga SËtra, p. 2 7 2c #: tathå caturdaßy-
a#†amyådi#u tithi#Ëpadi#†åsu mahåkalyåâkasambandhitayå puˆyatithitvena

prakhyåtåsu tathå paurâmås¥#u ca tisÁ#v api caturmåsakatithi#v ity artha˙,
evaµbhËte#u dharmadivase#u. This passage is referred to by the sixteenth-
century Dharmasågara (see below), Pravacanapar¥k$å 3.59.
84 Paul Dundas
candrasËri’s MËlaßuddhiprakaraâ, who notes that the tavvaseâ
statement is not found in two of the six manuscripts utilised by him.24
Another version of the Kålaka story approximately contemporary with
that of DevacandrasËri, that found in the Kahåval¥ of Bhadreßvara,25
does not contain the tavvaseâ statement at all, although the fourteenth
century Kulamaˆ&anasËri does in fact associate this version of the story,
albeit without quoting it directly, with the claim that the fortnightly
pratikramaâ was held on the full-moon day.26 Furthermore, two of the
manuscripts utilised by Brown for his edition of the Long Anonymous
Version of the Kålakåcåryakathå, dated repectively 1279 and 1287, read
ya caumåsåni/c a u m å s a y å ˆ i for pakkhiyå¥î (the equivalent of
pakkhiyåiµ in other versions), that is to say enjoining that the four-
monthly pratikramaâ observance be performed on the fourteenth day
of the month rather than the fortnightly observance.27 A version of the
24Bhojak also notes that the epitomising verse referred to above is only found in
two manuscripts.
25Although this significant Pråkrit narrative collective has unfortunately
remained in manuscript form to this day, Brown (1933, pp. 102–106) provided
an edition and summary of its version of the Kålaka story. Malvania (1983,
p.#81) argues that Bhadreßvara most likely flourished in the twelfth century.
For Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, p. 26, the author of the
Kahåval¥ was already of uncertain date (anirˆ¥tasaµbhavakåla).
26VicåråmÁtasårasaµgraha, p. 26. The Kahåval¥ describes King Íåtavåhana’s
wives being instructed by their lord to fast on amåvåsyå (the dark half of the
fifteenth day) for the sake of pratikramaâ and then to feed monks on the
pratipad, the first day of the fortnight. See Brown 1933, p. 104. The phrase
parikkhiya-pa"ikkamaˆ’atthaµ of Brown’s edition, while possibly meaning
something like “for the sake of repentance of faults which have been
examined”, is nonetheless odd, and we may conjecture that Kulamaˆ&anasËri
was referring to a manuscript of the Kahåval¥ which had a reading correspond-
ing to pakkhiya / pakkhiyåiµ. This is indeed the reading found in the quotation
of this passage at Gurutattvaprad¥pa 4.36 autocommentary p. 80.
27 Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, pp. 29–30, quotes the Niß¥tha
CËrî and other sources for the original observance of the four-monthly prati-
kramaâ on the full-moon day, pointing out that the fact that it had come to be
prescribed for the fourteenth day had been brought about by practice initiated
in ancient times (yac ca caturdaßyåµ vidh¥yate tatra pËrvapravÁttåcaraˆå
tavvaseâ statement contained in a Kålakåcåryakathå edited by
Leumann in 1883 also contains the reading caumåsiyaµ, although the
verse in which it occurs appears to have been added as a supplementary
amplification to the manuscript utilised by the Swiss scholar.28
There was a perfectly understandable calendrical rationale for the
redating of the four-monthly pratikramaâ to the fourteenth day. It
derived from the fact that the Kalpa SËtra, which in the form it exists
today most likely dates from around the fifth century C . E ., states
(p.#296) that Mahåv¥ra had commenced the Paryu$aˆ festival in which
Saµvatsar¥ occurs after a month and twenty days of the rainy season
retreat had elapsed. As the rainy reason for the Jains customarily started
on a full-moon day which was also an obvious date for one of the three
purificatory four-monthly pratikramaâs, a forward adjustment of
Saµvatsar¥ necessarily entailed a commensurate forward adjustment to
the fourteenth day for that particular caturmåsika observance and by
extension for the other two also.29 However, it seems clear that the story
of Kålaka’s redating of Saµvatsar¥ and the possible consequences of it,
involving as they did matters of authority and consensus as well as the
relocation of an ancient festival, was to become a highly charged issue
in the Ívetåmbara Jain community by around 1000–1100 C .E. and we
karaâm) i.e. not recently. MunicandrasËri (see below), Påk$ikasaptati, v. 66,

gives a negative and a positive justification for the four-monthly pratikramaâ
taking place one day earlier on the fourteenth day of the month#: firstly (here I
follow the commentator Maheßvara), because in the current debased time men
are generally incapable of enacting the appropriate disciplinary behaviour
associated with the observance, in this case a structure of austerity which
would have otherwise required a six-meal fast concluding on the full-moon
day (såmpratamanu#yåˆåµ båhulyena #a#†hatapasa˙ karaê ßakter abhåvåd
iti), and secondly, because it represents the customary behaviour of upright
Jains. For the first reason, cf. Gurutattvaprad¥pa 4.34.
28See Leumann 1998, p. 21 (III vv. 82–83)#: evaµ ceva cautth¥e kayaµ Kålaga-
sËriˆå | pajjosavaâ-pavvaµ, to savva-sangheâ manniyaµ. tav-vaseâ

mun’-indehiµ åîyaµ caumåsiyaµ | åyaraˆå coddas¥e annahå puna punnimå.
With reference to v. 83, Leumann notes, “Dieser Íloka ist im Texte selbst
vergessen, aber am Rande nachgetragen.”
29See Vaidya 1977, p. 271 and the helpful explanation in Cort n.d.
86 Paul Dundas
may regard the increasing production of extended versions of the
Kålakåcåryakathå as partly indicative of this. Furthermore, the
emergence of one particular disciplinary order, the Paurâm¥yaka
Gaccha, at the beginning of the twelfth century, reveals how this issue
exposed what must have appeared to some possible inconsistencies in
Ívetåmbara Jain customary practice.
The Paurâm¥yakas, or “Adherents of the Full-Moon Day”, are
credited with inspecting the logic of Kålaka’s redating and consequently
interpreting an ancient, that is to say pre-Kålaka, Saµvatsar¥ observed
on the fifth day of Bhådrapada as necessarily requiring the restoration of
the observance of the fortnightly pratikramaâ on the full-moon day
(the fifteenth) rather than the fourteenth, and they accordingly defended
the validity of this dating in what was among the first of many attempts
in Ívetåmbara tradition in the second millennium C.E. to reactivate the
context of the ancient scriptures, the word of the Jinas.30 Unfortunately,
no significant Paurâm¥yaka writings defending this position seem to
have survived, and their arguments have to be reconstructed from the
accounts of their opponents. So the twelfth-century Påk$ikasaptati,
“Seventy Verses on the Fortnightly Pratikramaâ” (also known as the
Óvaßyakasaptati, “Seventy Verses on the Obligatory Practices”) of
MunicandrasËri of the BÁhad Gaccha, with its commentary by
Maheßvara, prepares the ground for later polemicists in presenting the
Paurâm¥yakas as attempting to overthrow the consensual basis of
Kålaka’s redating. In this work MunicandrasËri affirms the
impossibility of the fortnightly pratikramaâ taking place on the full-
moon day31 on the grounds of general usage, textual authority, albeit
30The Paurâm¥yaka Gaccha seems to have ceased to exist as a significant

institutional component of Ívetåmbara Jainism in the eighteenth century.
However, vestigial traces of its continuity into more recent times can be found.
See Cort 2001, p. 45.
31See MunicandrasËri, Påk$ikasaptati, v. 40, for the terms caturdaß¥, “four-
teenth day”, and påk#ika, “fortnightly”, being identical in significance, a point

reiterated by all later anti-Paurâm¥yaka writers against the supposed claim
deriving from works apparently not predating the sixth century,32 and
customary behaviour since Kålaka’s time.33
The status of the readings pakkhiyå¥î and cåummåsiyåî (or some
close approximation to them) in the tavvaseâ statement was a topic not
dealt with in detail by MunicandrasËri, but their implications became
increasingly discussed from the thirteenth century by anti-Paurâm¥yaka
polemicists belonging to the Tapå Gaccha disciplinary order who
refused to accept that the fourteenth-day pratikramaâ observance was
originally observed on the full-moon day and subsequently conditioned
by Kålaka’s redating of Saµvatsar¥. So Kulamaˆ&anasËri claims in his
VicåråmÁtasårasaµgraha that there can be no reason to associate a
change to the fourteenth day with Kålaka, since no authoritative text
refers to this.34 He also asserts that there can be found no reference to
any teacher such as Kålaka performing the fortnightly observance on
the fourteenth day for the very first time, thereby confirming its
that the fortnightly observance could in fact overlap with the first part of the
fifteenth day of the month and thus coincide with the new moon.
32MunicandrasËri, Påk$ikasaptati, v. 12, with commentary (pp. 13–17) which
refers to cËrîs, the Mahåniß¥tha SËtra and the Samaråiccakahå.

33Maheßvara, commenting on Påk$ikasaptati, v. 67, p. 53, reiterates the reading
of the Niß¥tha CËrî, adding to it Kålagajjehiµ. MunicandrasËri, Påk$ika-

saptati, v. 59, addresses the argument that there can be found authoritative
textual evidence for lay people fasting on the fifteenth day. The example cited
occurs in the Kålakåcåryakathå when King Íåtavåhana informs his wives that
they must fast on the amåvåsyå, that is to say the fifteenth day when the new
moon rises. See Niß¥tha CËrî, uddeßaka 10, p. 131, and note 26. This is
interpreted as a special case by reference to a permissible act of worship on
this particular day of depictions of the temple on the uninhabited continent of
Nand¥ßvara. Cf. Påk$ikasaptati, vv. 60–62, which discusses the mention of
occasional cases of lay observance on the fifteenth day, stating that they do not
undermine the general authority of practice on the fourteenth day, and
Påk$ikasaptati, v. 65, which rejects endorsement of any popular leaning
towards the fifteenth day.
34Kulamaˆ&anasËri is here opposing the Paurâm¥yaka claim that the fourteenth
day pratikramaâ observance is purely conventional.

88 Paul Dundas
antiquity.35 On this basis and also on the authority of a central canonical
text like the SËtrakÁtå!ga SËtra, the fourteenth-day observance must be
regarded as having been promulgated by none other than the Jinas at the
beginning of each successive t¥rtha. It is therefore the Jain teacher
lineage as represented in Kulamaˆ&anasËri’s time by the leaders of the
Tapå Gaccha and those who have preceded them as far back as
Mahåv¥ra who guarantee the genuineness of this dating and the fact that
a challenge to its validity was not articulated in the Jain community
prior to the emergence of the Paurâm¥yaka Gaccha.36
The Gurutattvaprad¥pa, a polemical text written by a monk of the
B Áhatpoßålika sublineage of the Tapå Gaccha some time in the
fourteenth century and appoximately contemporary with Kulamaˆ&ana-
sËri, discusses the matter equally explicitly in its fourth chapter which is
directed against the Paurâm¥yaka Gaccha.37 It confirms that no
compelling textual connection can be established between Kålaka and
the institution of the fourteenth-day pratikramaâ observance. Detailed
mention in the Paryu$aˆåkalpa CËrî and what the Gurutattvaprad¥pa
styles the “short” (laghu) version of the Kålakåcåryakathå of a matter
not immediately relevant to the celebration of Paryu$aˆ and the dating
of Saµvatsar¥ would, it is claimed, be improbable because of the brevity
of these texts.38 Injunctions relating to both the fourth and fourteenth
days of the month are no doubt found in the “long” version of the story,
but even there no description is given of what observance is to be
35Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, p. 28, points out that even the

Íatapad¥, the foundational text describing the customary practice of the Añcala
Gaccha, an order which emerged from the Paurâm¥yaka Gaccha and still
observes Saµvatsar¥ on the fifth day of Bhådrapada (see Balbir 2003, p. 59),
admits that Kålaka was not responsible for initiating the observance of the
fortnightly pratikramaâ on the fourteenth day.
36Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, pp. 26–28.
37For the background to this text, see Catalogue 2006, p. 463 (entry no. 662),
and Dundas 2007, chapter four.

38Gurutattvaprad¥pa 4.20–21 with autocommentary.
carried out (caritånuvåda).39 In other words, the fourteenth-day prati-
kramaâ observance has its own authority without reference to the
Kålakåcåryakathå. The Gurutattvaprad¥pa further states that in the old
manuscripts of the Kålakåcåryakathå the reading in the tavvaseâ
statement is regularly tavvaseâ ya caummåsiåî vi cauddas¥e åyariyåî
and it attributes the existence of erroneous references to the fortnightly
observance found in other versions of the statement to an interpolation
by an ill-informed member of another sect (siddhåntånåbhijña
matåntar¥ya), who can no doubt be judged to be CandraprabhasËri, the
founder of the Paurâm¥yaka Gaccha, or one of his followers.40
The question of the correct reading in the tavvaseâ statement
remained an issue as late as the second half of the sixteenth century.
The Tapå Gaccha polemicist Dharmasågara claims that it was a member
of the Paurâm¥yaka Gaccha (råkårakta) who had altered the old and
genuine reading ( j¥râpå†ha) caummåsiåî to pakkhiåiµ in the portion
of DevacandrasËri’s commentary on the Sthånå!ga SËtra referred to by
Kulamaˆ&anasËri, supposedly thus giving a meaning making as much
sense as the phrase “water it with fire”. Dharmasågara refers to the
reading cåummåsiåiµ being found in many of the old palm-leaf
manuscripts in the famous libraries in På†aˆ and also to the absence of
the reading pakkhiyåiµ in old manuscripts in general, suggesting that
the presence of the phrase annaha ågamuttåî puˆîmåe is itself
indicative of this change of reading, since it could only convey meaning
39Gurutattvaprad¥pa 4. 22 autocommentary, p. 70.

40Gurutattvaprad¥pa 4.36 autocommentary. Cf. Gurutattvaprad¥pa 4.22 auto-
commentary, pp. 70–71, where those responsible for the change of reading in
the tavvaseâ statement are said to be recent (våkyata˙ påßcåtyånåm apy
ådeßo ’bhavat). Gurutattvaprad¥pa 4.36 autocommentary also refers to the
episode in the Kålaka story given in the Kahåval¥ (see note 32) in which King
Íåtavåhana instructs his wives to perform a fast on amåvåsyå, ascribing this
assertion either to the carelessness (anåbhoga) of the author Bhadreßvara or an
interpolation by a recent monk under the influence of the Paryu$aˆåkalpa
CËrî which describes a fast observed on this day.
90 Paul Dundas
in respect of the four-monthly observances.41 As Dharmasågara
emphasises, since the fourteenth-day observance must be regarded as
dating from the very founding of the Jain community by each Jina, it is
thus eternal ; historical agency in promoting it, as in the possible case of
#
Kålaka, has played no role.42 The immemorial (anådisiddha) nature of

the fortnightly observance and the heretical status of the Paurâm¥yaka
Gaccha are both confirmed by Dharmasågara by reference to the
description in Haribhadra’s commentary (c. sixth century C . E .) on
Óvaßyaka Niryukti, v. 468, of Jinadåsa, a lay follower of Pårßva, the
twenty-third Jina, fasting on the eighth and fourteenth days of the
month.43
Although what was perceived by later polemicists to be the original
legitimation for change in the calendar and the prime exemplification of
the introduction of customary practice by senior monks, namely
Kålaka’s redating of Saµvatsar¥ from the fifth to the fourth day of
Bhådrapada, had supposedly taken place in the fifth century C.E., it was
not until the beginning of the second millennium C .E . that disputes
about calendrical issues gained identifiable momentum in Jainism.44 It
can be seen from the foregoing that the version of the Kålakåcåryakathå
best known to western scholars, that edited by Jacobi and Brown,
enshrines a reading relating to the ritual calendar which became
markedly controversial within Ívetåmbara Jain tradition at that
particular time, and we can confidently assert on this basis that at least
one portion of the Kålakåcåryakathå was not an uncontested text but
subject to regular sectarian pressures.
Further conclusions are necessarily conjectural but worthy of more
detailed exploration. So it seems possible that the Kålaka story may not
41Pravacanapar¥k$å 3.59 autocommentary, p. 176, and 3.63 autocommentary,

p.#181. Cf. 5.36 autocommentary. For Dharmasågara, see Dundas 2007.
42Paryu$aˆådaßaßataka, v. 107 (evaµ cåuddasiaµ titthaµ titthappava††aˆå
êaµ), with autocommentary.

43SËtravyåkhyånavidhißataka, v. 31, with autocommentary.
44Cf. Cort 1999, p.42.
in fact record a genuine historical event but may rather represent a later
narrative attempt to rationalise a period of calendrical complexity for
the Jain community as the implications of the novel style of lunisolar
reckoning became increasingly appreciated. Furthermore, the insistence
by anti-Paurâm¥yaka polemicists on the immemorial location of the
central disciplinary observance of the fortnightly pratikramaâ on the
fourteenth day of the month may reflect the fact that this date had been
in actuality influenced by wider calendrical conditions and possibly
maintained as a means of sectarian differentiation in the context of the
increasing prestige of the new lunisolar calendar.
As we saw at the beginning of this note, the MËlasarvåstivådins
viewed observance of po#adha on the fourteenth day as exceptional,
with the full-moon day being regarded as normative for this ceremony.
In legislating for potentially difficult situations arising from these
datings, Buddhist sources from the beginning of the first millennium
appear in general to have been much more articulate about the issues
involved in the various forms of calendrical reckoning in circulation in
north India at that time than their Jain counterparts. In that case, it
seems clear that if the later Jain controversy described in this note
concerning whether the fortnightly pratikramaâ be observed on the
fourteenth day or the full-moon day does actually bear witness to the
long-range influence of changes in calendrical systems introduced in the
opening centuries of the first millennium C.E., particularly in relation to
use of the lunar and lunisolar modes of reckoning, then the evidence of
Buddhist texts must be a necessary resource for students of Jainism
wishing to contextualise fully the early historical situation.
Paul Dundas
92 Paul Dundas
PRIMARY SOURCES
DevacandrasËri, Kålakåcåryakathånaka, in PradyumnasËri, MËlaßuddhipra-
karaâ with the commentary of DevacandrasËri, ed. A.M. Bhojak,
Ahmedabad#: Prakrit Text Society, 1971, story 22, pp. 123–34
Dharmasågara, Paryu$aˆådaßaßataka, Surat#: Â$abhdev Keßarmalj¥ Ívetåmbara
Saµsthå, 1936
Dharmasågara, Pravacanapar¥k$å, Surat#: Â$abhdev Keßarmalj¥ Ívetåmbara
Saµsthå Surat, 1937
Dharmasågara, SËtravyåkhyånavidhißataka, ed. Muni Låbhasågara,
Kapa&vaµj#: Ógamoddhåraka Granthamålå, 1961
Gurutattvaprad¥pa ed. Muni Låbhasågara, Kapa&vaµj#: M¥†håbhå¥ Kalyåˆcaµd
Pe&h¥, 1962
JayasiµhasËri, Dharmopadeßamålåvivaraâ, ed. L.Bh. Gåndh¥. Bombay#: Singhi
Jain Series, 1949
Jinadåsa, Niß¥tha CËrî, in Niß¥tha-SËtram, tÁt¥yo vibhåga˙, uddeßakå˙ 10–15,
ed. Amaramuni and Muni Kanhaiyålål, Ógrå/ RåjagÁha#: Sanmati Jñån P¥†h,
1982
Kalpa SËtra, translated by Hermann Jacobi, in Jaina Sutras Part 1, New York#:
Dover, 1968 (reprint)
Kulamaˆ&anasËri, VicåråmÁtasårasaµgraha, Surat#: Â$abhdevj¥ Keßar¥malj¥
Ívetåmbara Saµsthå, 1936
Mahåvagga, The Vinaya Pi†akaµ, Vol 1. The Mahåvagga, ed. H. Oldenberg,
London#: Williams and Norgate, 1879
MunicandrasËri, Påk$ikasaptati, ed. Muni Låbhasågara, Kapa&vaµj#:
Ógamoddhåraka Granthamålå, 1971
Í¥lå!ka, commentary on SËtrakÁtå!ga SËtra, ed. Muni JambËvijaya, in
Ócårå!gasËtraµ SËtrakÁtå!gasËtraµ ca, Delhi#: Motilal Banarsidass, 1978
SËtrakÁtå!ga SËtra. SËyaga&aµgasuttaµ, ed. Muni JambËvijaya, Bombay#: Ír¥
Mahåv¥ra Jain Vidyålaya, 1978
SECONDARY SOURCES
Balbir, Nalini, 2003. “The A(ñ)cala Gaccha Viewed from Inside and from
Outside”. In Jainism and Early Buddhism#: Essays in Honor of Padmanabh
S. Jaini. Edited by Olle Qvarnström, Fremont#: Asian Humanities Press,
pp.#47–77
Brown, W. Norman, 1933. The Story of Kålaka#: Texts, History, Legends, and
Miniature Painting of the Ívetåmbara Jain Hagiographical Work the
Kålakåcåryakathå, Washington#: Smithsonian Institution
Catalogue 2006. Catalogue of the Jain Manuscripts of the British Library,
Including the Holdings of the British Library and the Victoria & Albert
Museum, Vol. 2. Edited by Nalini Balbir, Kanhaiyalal V. Sheth, Kalpana K.
Sheth, and Candrabhal Bh. Tripathi, London#: The British Library and the
Institute of Jainology
Cort, John E., 1999. “Fistfights in the Monastery#: Calendars, Conflict and
Karma among the Jains”. In Approaches to Jaina Studies#: Philosophy,
Logic, Rituals and Symbols. Edited by N. Wagle and O. Qvarnström,
University of Toronto#: Centre for South Asian Studies
——— 2001. Jains in the World#: Religious Values and Ideology in India, New
York#: Oxford University Press
——— n.d. John E. Cort, “Fistfights in the Monastery. Respect and Authority
in the Jain Tradition#: Calendars, Karma, and Kriyå”, unpublished
manuscript
Dietz, Siglinde, 1997. “Anmerkungen zum buddhistischen Fastentag”. In
Bauddhavidyåsudhåkara #: Studies in Honour of Heinz Bechert on the
Occasion of His 65th Birthday, edited by Petra Kiefer-Pülz and Jens-Uwe
Hartmann, Swisttal-Oderndorf#: Indica et Tibetica Verlag, pp. 63–70
Dundas, Paul, 2007. History, Scripture and Controversy in a Medieval Jain
Sect, London#: Routledge
Hu-von Hinüber, Haiyan, 1994. Das Po#adhavastu# : Vorschriften für die
buddhistische Beichtfeier im Vinaya der Mëlasarvåstivådins, Reinbek#: Dr
Inge Wezler
Hureau, Sylvie, 2006. “Preaching and Translating po#adha Days#: Kumåraj¥va’s
Role in Adapting an Indian Ceremony to China”, Journal of the Inter-
national College for Postgraduate Buddhist Studies 10, pp. 87–119
Leumann, Ernst, 1998. Kleine Schriften. Stuttgart# : Franz Steiner
Jacobi, Hermann H., 1880. “Das Kålakåcårya-Kathånakam”, Zeitschrift der
Deutschen Morgenländischen Gesellschaft 34, pp. 247–318
Malvania, Dalsukh D., 1983. “On Bhadreßvara’s Kahåval¥”, Indologica
Taurinensia 11, pp. 77–95
Norman, K.R., 1990–2007. Collected Papers, Volumes I–VIII, Oxford#: Pali
Text Society
Renou, Louis, and Jean Filliozat, 1953. L’Inde classique#: manuel des études
indiennes, Volume 2, Hanoi#: École française d’Extrême-Orient
Tieken, Herman, 2000. “Aßoka and the Buddhist Saµgha#: A Study of Aßoka’s
Schism Edict and Minor Rock Edict 1”, Bulletin of the School of Oriental
and African Studies 63, pp. 1–30
Vaidya, F.L., 1977. Jain DÁ#†ie Tithidin ane Parvårådhan. Muµba¥#: Ír¥ Jain
Pravacan Pracårak Èras†
94 Paul Dundas
Vogel, Claus, 1997. “On the Date of the Po$adha Ceremony as Taught by the
MËlasarvåstivådins” In Bauddhavidyåsudhåkara #: Studies in Honour of
Heinz Bechert on the Occasion of His 65th Birthday. Edited by Petra
Kiefer-Pülz and Jens-Uwe Hartmann, Swisttal-Oderndorf #: Indica et
Tibetica Verlag, pp. 673–88
Williams, R.W., 1963. Jaina Yoga#: A Survey of the Medieval Íråvakåcåras,
London#: Oxford University Press
caveat lector
The 125th anniversary of the founding of the Pali Text Society is an

occasion for celebration — celebration of the fact of the continued
existence of the Society, and of the early work done by T.W. Rhys
Davids and his first committee of management (listed as Professor
Fausbøll, Dr Oldenberg, Dr Morris, M. Senart) and other contempo-
raries and successors. All of us who study Påli or Theravåda Buddhism
today stand on the shoulders of those early scholars. We have inherited
from them texts, commentaries, translations, dictionaries, grammatical
works. Where we are now depends on what they did.
I would like to consider here the next 125 years of the Pali Text
Society (PTS). It seems to me that it is time those of us whose main
concern is Påli should pause to think about the direction we should be
taking in the twenty-first century.
Let us look at our inheritance. The founder members of the PTS
wished “to render accessible to students the rich stores of the earliest
Buddhist literature” (JPTS 1882, p. vii). The PTS has indeed done that,
as the rows of its editions of the Canon and commentaries on my
shelves attest. These editions (and the Pali–English Dictionary of
T.W.´Rhys Davids and W. Stede as well as the Påli Grammar of
W.´Geiger) have been for most of us studying Påli, I imagine, the
gateway to our knowledge, and indeed the field in which we dig. The
great majority of those volumes was already there when I began to learn
about Påli forty years ago. This means that much of what I have on
those shelves of mine is the product of pioneering work, rather a first
draft than the culmination of centuries’ study (as are my Oxford
Classical Texts). The situation has been well described by Mr Norman,
whose eighty-two years of life and so many years of Påli scholarship are
another cause for celebration, in his article “Påli philology and the study
of Buddhism” (Norman 1990), and his lecture “Buddhism and
Philology” (Norman 2006). I can only reiterate his concerns, and try to
reinforce his warnings.
96 Margaret Cone
The transmission of the texts covers a very long period, but most of
our mss are comparatively recent. We can make no confident
assumption that what we have is anywhere near the actual Buddha-
vacana (or indeed Buddhaghosavacana) ; what we have is the product of
´
centuries of careful copying, careless copying, knowledge, incom-

petence, inspired emendation and bungling (see von Hinüber 1978).
And none of that stopped with the beginning of Western scholarship. Mr
Norman’s plea for more “philologists” capable of making good new
editions of the Canon seems so far to have gone largely unanswered.
In my writing of the PTS Dictionary of Påli (DOP), I search for
occurrences of each word on a CD-Rom of the Thai edition (Se) of the
Canon and commentaries. This, although I suspect not always accurate-
ly transcribed, gives me access to Thai readings, which I can check
against the PTS editions (Ee ), the Burmese Cha††hasa¤g¥tipi†aka
printed editions (Be ), and the Sinhalese Buddha Jayanti Tripi†aka
printed editions (Ce). What has become clear from this process is that
the PTS editions have a large number of, at best questionable, at worst,
plain wrong, readings.1 Often these editions are transcriptions of a very
few mss, or even of only one ; often they follow only one tradition,
´
usually the Sinhalese. Some editors made mistakes because they

misread the mss — various characters in the Burmese and Sinhalese
scripts can be very easily confused, especially in mss kha and ba, ta and
na, bha, ha and ga, pa, ya and sa, va and ca in the Sinhalese script
spring immediately to mind. Others made mistakes because they did not
know Påli or Sanskrit well enough (understandably so in the case of Påli
— they were pioneers). The ability to transcribe a Sinhalese or Burmese
ms is not a sufficient qualification to produce a reliable Påli text. I
suspect, also, that those who did know Sanskrit emended silently,
especially in matters of sandhi. The D¥gha-nikåya and its commentary
seem to me to have been tidied in this way. Even those editors, like
Fausbøll, whom one had always been taught to respect, produced texts
1Iwill give no examples in this article, but a glance at a few pages of the first
volume of DOP will show what I mean.
96
caveat lector 97
with many doubtful readings.2
I would count as wrong any reading in Ee which differs from a
unanimous reading in the other three editions, when (a) one can explain
the difference by a misreading of Burmese or, especially, Sinhalese
characters ; (b) the Ee reading is metrically incorrect (not just irregular,
´
but impossible) ; (c) we are faced with an inexplicable form, against

´
something which makes sense. In the case of (c), of course, the

argument against dismissing Ee as a wrong reading is a recourse to the
principle of lectio difficilior melior, and a suggestion that Be in
particular normalises readings and erases difficulties. I am not con-
vinced that the rule of lectio difficilior is valid for Påli texts, given the
uncertain and contaminated traditions of mss and editions. And to try to
defend a word, for which one can find no real etymological justification
or explanation, and reject an alternative, well attested and commented
on, which has an appropriate meaning, seems somewhat perverse. In
any case, Be itself exhibits some idiosyncratic readings, which a
normalising or rationalising tendency might have been expected to
change.3
On other occasions, where Ee differs from the other editions, I
would hesitate to say the reading was wrong. I accept that it may
represent another, valid, tradition, but I believe readers ought to be
made aware of the alternative. Often too there are varying agreements
and disagreements among the editions — Ee with Ce against Be and Se ; ´
B e and Ce against Ee and Se ; indeed all combinations are found. Nor

´
should one think that any of the editions or traditions is free from error.
As an extreme case, it is often hard to make sense of any of the editions
of the Pe†akopadesa.
2I have read the statement that Fausbøll’s “great edition of the Jåtakas … is still
unsurpassed”. Perhaps the writer meant “not yet superseded”. Otherwise it is a
meaningless assertion. Studies of individual Jåtakas have clearly improved on
Fausbøll’s text, and anyone who looked closely and carefully at his text would
find much to question.
3See e.g. the consistent reading in Be of dhamakaraâ, against the explicable
dhammakaraka found usually in the other editions.
97
98 Margaret Cone
If one looks at the readings of the other editions, one finds in all
traditions inconsistencies, incomprehensibility, more problems. Often,
however, even small differences from Ee give readings which are more
convincing, because more subtle, more elegant, more Påli. But the Påli
of the PTS editions is the only Påli seemingly used and depended upon
by Geiger in his Grammar (Geiger 1916), by Rhys Davids and Stede in
the Dictionary, by A.K. Warder in his Introduction to Pali (Warder
1963), and by most writers on Påli and Theravåda Buddhism even
today. It is as if those who studied and researched and published in the
past had produced works that somehow have also become canonical.
Yet every one of them was fallible, and was working with fallible
materials.
I have become convinced that we should take nothing on trust. We
should use all existing dictionaries and grammatical works with caution
and scepticism, checking statements and references wherever and
whenever possible. As Sir Monier Monier-Williams wrote in the
Preface to his Sanskrit–English Dictionary : ´
Nay, I am constrained to confess that as I advanced further on the

path of knowledge, my trustfulness in others … experienced a series
of disagreeable and unexpected shocks ; till now … I find myself left
´
with my faith in the accuracy of human beings generally — and

certainly not excepting myself — somewhat distressingly disturbed.
[My emphasis.]
When we open Geiger’s Grammar, we are presented with what appears
to be an exhaustive account of Påli grammatical forms. Statements are
made, rules devised, paradigms laid out. But what about the evidence ? ´
There are forms there I cannot find in the texts ; there are rules which
´
depend on one occurrence of a form, not supported in all editions ; there ´
is, inevitably, much missing, which could give a different picture of the
language.4
4The evidence for forms is often weak, e.g. Geiger §39´:1´: “k appears for g in´:
akalu Mil 338,13”. But only in Ce and Ee. Be has agalu-, Se aggalu-. And cf. Ja
98
caveat lector 99
Another inheritance, the Pali Text Society’s Pali–English
Dictionary (PED), remains useful, but the meanings it gives must be
checked against Sanskrit or Prakrit, and its references verified.
Especially with rare words or words with no obvious derivation or
Sanskrit equivalent, we should look at alternative forms in other
editions. When I told a Ph.D. student not to trust everything in PED, her
supervisor reproached me, but it is the first thing we should teach any
student of Påli. We should even on occasion question the Critical Påli
Dictionary (CPD). Its first editors, Dines Andersen and Helmer Smith,
were indeed giants of Påli scholarship, but their work too should not be
treated as canonical. They were not infallible, and for very good reason
they could not always be right (nor have all of their successors at CPD
had their wide knowledge and intelligent interpretative powers). And
IV 440,23* (Ee akalu´; Be, Ce aga¬u´; Se aggalu)´; VI 144,23* (Ee akalu-´;

Be, Ce aga¬u-´; Se aggalu-)´; elsewhere Ee has agaru or agalu. — “lakanaka,
‘anchor’ Mil 377,19, 23”. Again, only in Ce and Ee (nåvålakanaka … nåva
laketi). Be and Se have nåvålagganaka … nåva laggeti. — §39 ´:3 ´: “c
appears for j in´: påceti ‘drives’ Dhp 135 … beside påjeti Ja II 122,5”. At Dhp
135 Be and Se read påjeti. I am aware of course of the revision of Geiger by
K.R. Norman (PTS 1994), and of the Påli grammar of T. Oberlies (de Gruyter
2001). But both of these are vulnerable to the same criticism´: a reliance on Ee.
See e.g. Geiger 1994 §38 ´:1a (not in Geiger 1916)´: “kh is voiced to gh in
nighaññasi, ‘you will dig’, Ja VI 13,18*”´; Oberlies 2001 §51´: “nighaññati
‘strikes down’ ([denominative from] *nighañña)”. The form is puzzling, and
Oberlies may be correct, but it might be of interest to note the various
readings´: Ce, Ee ya kåsuyå nighaññasi´; Be, Se nihaññasi´; Ja VI 13,27´´: Ce
nighaññas¥ ti nihanissasi, ya tva ettha nihanissåm¥ ti saññåya kåsu
nikhaâti so ahan ti d¥peti´; Be, Se nihaññas¥ ti nihanissasi … ´; Ee nighaññas¥
ti nikhaîssasi, ya ma ettha …). The verse is quoted at Cp-a 225,15*, where
all the editions read nikhaññasi. We might also compare Ja IV 102,9* where
B e , Ee read´: sace adhammo hañchati dhammam ajja, while Ce , Se read
haññati, in all cases glossed (102 ,25´) with hanissati. See also åhañcha,
åhañchi(), with vll. of åhaññi. Oberlies, §14 ´:4, gives påceti, without
reference (see above), and without making clear that the more usual form in all
editions is påjeti. Re ajakara, also in §14´:4, note that this form appears only in
E e at Ja III 484 ,16* ´; in the other editions, and in Ee elsewhere, the form is
ajagara.
99
100 Margaret Cone
one of the weaknesses of Franklin Edgerton’s Buddhist Hybrid Sanskrit
Dictionary, useful and enjoyably personal though it is, is its trust in
PED.
Speaking here as a scholar of Påli, not of Buddhism, I think we are
sometimes limited and confined by our inheritance. We must bring to
our study of the texts our knowledge of language, of India, and of
religion, from outside those texts. The interpretation of Påli has looked
backwards (and inwards) for as long as we know. Buddhaghosa
followed the Sinhalese commentaries, which themselves presumably
were based on commentaries brought from India. Aggavasa in his
account of the Påli language, the Saddan¥ti, not only describes Påli as he
found it in the Buddhavacana — absolutely legitimately — but explains
and interprets it in a way limited by traditional beliefs about the
language and its status. Warder, in his Introduction to Pali, accepts the
same kind of restraints, and indeed often follows Aggavasa’s
interpretations. He teaches the language in a manner that is irritating and
bizarre to anyone who knows any Sanskrit. Some of the statements of
Aggavasa and Warder are true only in a very tenuous sense. One feels
all three of these scholars are deliberately ignoring or denying certain
things that they know in order to keep within the restraints of tradition
and the past.
The first Påli dictionary, the Abhidhånappad¥pikå, while including
some specifically Buddhist terms and connotations, largely looked to
Sanskrit and the Amarakoßa, and is not a true reflection of the Påli of
the texts, but it is probably the last work of Påli scholarship in which we
can complain of too heavy a reliance on Sanskrit. The Dictionary of the
Påli Language by R.C. Childers includes the material of the Abhidhåna-
ppad¥pikå, but depends much more on the interpretations of the Påli
commentaries and of the Sa¤gha. After Childers, in the work of the
early translators of Påli texts, we often find mere intuition and guess-
work, buttressed by a strong conviction of what a Buddhist context
required.
100
caveat lector 101
These translations are an influential inheritance, but in them, not
infrequently, the sense of the Sanskrit equivalent to a word was ignored
or rejected — I do not know whether this was due to a conscious
decision or to ignorance. Then the sense of English words was stretched
and indeed violated, or words were coined, invented. A prime exponent
of this method was Mrs Rhys Davids.5 Fortunately not many of her
coinages have survived, but other translators also preferred a rare word
to a common one, such as “fruition” for “fruit” ; or coined words,
´
perhaps by finding for each part of the Påli word the equivalent in
Latin.6 Many other neologisms or strained usages of existing words, for
example in the translations of Bhikkhu Ñåâmoli, seem to have
acquired that canonical status. There was apparently a desire to create
what one might call technical terms of Theravåda Buddhism. This is
convenient for translators and interpreters, of course, and means that
they do not have to consider what the texts are really saying in each
context. But I think such a practice obscures the meaning, and is
anyway false, since I cannot believe that the Buddha spoke, as for
example sociologists do, in a jargon no ordinary person would under-
stand without a glossary.
Another inheritance is the “literal” translation. A literal translation
is not a translation, because the meaning of a Påli word or passage has
not been expressed in English. For particular words, one English
equivalent is chosen as the basic meaning, and that English word is used
5Her oddness can perhaps be fully appreciated only by a native English speaker,
with such usages as “Norm” as a translation of dhamma, “the Well-farer” for
sugata, “clansman” for kulaputta. Her translations, especially of verse, have a
medieval air (e.g. “eke the dappled deer”) not really appropriate to the context,
or aim at poetry and attain only obscurity (“In grasping not O well is him”)
rendering anupådåya nibbuto´; note that in this Sayutta verse Ee reads
nibbËto.
6For example, for va††ati with the preverb å- an etymological equivalent would
be Latin advertere. And so we find used a verb “to advert”, Unfortunately, “to
advert” already exists in the English language, and the standard dictionaries do
not support a current meaning which is really equivalent to åva††ati. What was
wrong with “turn to”´?
101
102 Margaret Cone
in all contexts.7 Throughout a whole text, Miss Horner’s translations
furnish good examples of literalness (not always even accurate) which
produces at times incomprehensibility (e.g. “state of further-men” to
translate uttarimanussadhamma). Did such translators ever ask, “What
would an Indian hearer have understood from this passage ? What ´
indeed is the Buddha’s concern here, what problem is he addressing,

what is he saying ?” This type of pseudo-translation is also to be found
´
in PED, whose compilers seem sometimes perverse in their refusal to

take Sanskrit as evidence for the meaning of a word. It may seem that
my criticism is rather of style than of interpretation, but the influence of
PED and of past translators has been strong : one sees the same
´
translations and expressions, which often have very little justification,

appearing again and again in new translations and works on Buddhism,
perpetuating that strange and barbaric language, aptly called Buddhist
Hybrid English.8
It is no insult, it is not lèse-majesté to criticise the texts of Fausbøll
or Feer, to question the paradigms of Geiger, to disagree with the
translations of T.W. Rhys Davids and Bhikkhu Ñåâmoli as well as
with those of Mrs Rhys Davids and Miss Horner. On the contrary, it is
absolutely necessary if there is to be any progess in Påli.
If we are prepared to look to the next 125 years, I believe we should
be prepared also to re-do everything. First, we must re-edit the texts of
the Canon and the commentaries. Even before that, however, we need to
discuss and decide how this should be done. I myself would advocate a
middle way, between one extreme of considering and accepting isolated
readings, and the other of simply taking over e.g. the Burmese version.
But the task is not simple or straightforward. The question of whether to
try to consult as many mss as possible, and how to decide on their
7 Like Mr Norman (Norman 1997, pp. 17–18) I particularly dislike “skilful” as

a translation for kusala in contexts where it does not make sense. Who decided
“skilful” was the basic, literal meaning of Sanskrit kußala, rather than “good”,
or “healthy” or “wholesome”´?
8See Griffiths 1981.
102
caveat lector 103
usefulness, is a difficult one. There are many occasions when alternative
spellings and perhaps also alternative expressions are equally valid, and
when the editions show no consistency. I give twenty-one citations in
my dictionary article on the numeral 40 ; all editions have cattår¥sa-
´
sometimes, all have cattå¬¥sa- sometimes, and all but Ce have cattål¥sa-
sometimes. I expect a similar pattern when I reach the numeral 50.
There is alternation between -aka and -ika, e.g. -bhËmaka and -bhËmika,
and of course there are the alternatives of -ika, -iya, -ita. For a good
example of variation in readings, see the passages quoted in DOP I
s.v. åve†hikå.9 An editor will have to use his or her judgment and
provide an informative critical apparatus.10 We should make use of the
considerable amount of good scholarship, especially of the last forty
years or so, on the transmission of the texts and on individual words (as
long as they are not ghostwords) ; and of the study of the Sanskrit and
´
Chinese and Tibetan versions.11 There should also be agreement on

spelling and punctuation conventions, so that there can be consistency
and clarity. Let us then at least produce a meaningful text, for which we
have good evidence in more than one tradition.
9These and other variations can often be explained by reference to other Prakrits
or to scribal conventions, and can tell us something of the evolution of the
text as we have it. The search for an “original” text (a genuine Buddha-
vacana ´?) is interesting, but always speculative, and is probably not the
concern of an editor.
10It is on occasions hard to know what form of a word an editor should choose.
Often the present editions are not consistent, e.g. between ja¤gama and
ja¤gala, or between japa and jappa. Consider also the possibilities uppilåvita,
ubbilåvita, ubbillåvita, ubbillåpita (see DOP I s.v. uppilavati). Note also that
Geiger (§38 ´:6) cites only ubbillåvita and ubbillåpita. Or what should the
choice be for the name of Mahåv¥ra, always nå†aputta in Be and Se ,
sometimes nåtaputta, sometimes nåthaputta in Ce and Ee´?
11Compare the painstaking and meticulous work on tiny fragments, e.g. from
the Stein and Hoernle collections, with the way some writers on Påli do not
even consult the Burmese and Sinhalese editions before pronouncing on a
word.
103
104 Margaret Cone
On the basis of these new editions, we must then produce a new,
accurate, thorough grammar ;12 then, I am afraid, someone will have to
´
re-write my dictionary. And finally we can produce good, readable,

trustworthy translations.
I know, of course, that I am speaking of Utopia. The first, and
perhaps insurmountable, difficulty is to find people capable of editing
the texts, and willing to do so. Such people must have a solid
knowledge of Sanskrit and at least some Prakrit, and a firm grasp of Påli
metrics, as well as a thorough understanding of how Påli works, and of
the whole spread of Påli literature. And how can they, without reliable
grammars and dictionaries ? ´
But it may be that what I suggest is not necessary (or even

desirable). As Mr Norman said (Norman 1990, p. 33), “It may justifi-
ably be asked whether the errors which may remain in the editions of
Påli texts really matter, and whether they are likely to have resulted in
any misunderstanding of the basic and most important elements of
Buddhism.” The state of the texts, and of Påli scholarship generally,
probably does not matter to those, perhaps the majority of the readers of
the publications of the PTS, who are interested in Theravåda Buddhism,
not in Påli.13 Nothing any of us does, I suspect, will change the
understanding of the principal tenets of the religion, or give startling
new insights into the thoughts of the Buddha. We can go on, slightly
12 Ihope that DOP will be a foundation and provide material for this new
grammar. In my articles on verbs I aim to give examples of all tenses and
infinite forms´; on nouns and adjectives, examples of significant cases and
irregularities. I am also compiling lists of certain formations, e.g. feminines in
-n¥, such as åråmikin¥, isin¥.
13 cf. Zürcher 1959, p. 356, n. 152´: “The ideal of a cursory way of reading the
classics without detailed philological studies was much in vogue [in China] in
the fourth century´; it agreed with the prevailing hsüan-hsüeh opinion that the
written text is only an imperfect and expedient expression of the hidden
wisdom of the Sage, and that the student must try to grasp the general
principles underlying the words rather than indulge in a careful and
painstaking study of the letter of the text”— a practice taken over by some
Chinese converts to Buddhism.
104
caveat lector 105
improving the publications in a piecemeal manner,14 writing articles (in
learned journals) about individual words, continuing in a confining
circle of compromised accuracy, approximate truth.
Well, it does matter to me. For the few of us whose job is the Påli
language, I believe there is a responsibility to provide information and
material as accurate, as true, as we can possibly make them. To provide
them not only for those who are drawn to Theravåda Buddhism, but for
other scholars — those who study the texts of Buddhism in other
languages, who study other forms of Buddhism and other religions, who
study languages. Otherwise, what do we think we are doing ? ´
Margaret Cone
Cambridge
14Aswith the rather arbitrary corrections made when volumes are reprinted.
This is a problem for me and the dictionary, as I cannot check readings in
every reprint.
105
106 Margaret Cone
ABBREVIATIONS
CPD V. Trenckner, D. Andersen and H. Smith (et al.), A Critical Påli
Dictionary, Copenhagen, 1924–
DOP I M. Cone, A Dictionary of Påli, Part 1. Oxford´: PTS, 2001.
JPTS Journal of the Pali Text Society
PED T.W. Rhys Davids and W. Stede, The Pali Text Society’s Pali-English
Dictionary, London´: PTS, 1921–25
BIBLIOGRAPHY
Geiger, W., 1916. Påli Literatur und Sprache, Strassburg
——— 1994. A Påli Grammar, revised and edited by K.R. Norman, Oxford´:
PTS
Griffiths, Paul J., 1981. “Buddhist Hybrid English´: Some Notes on Philology
and Hermeneutics for Buddhologists”, The Journal of the
International Association of Buddhist Studies, Vol. 4, no.2 (1981),
pp. 17–32
von Hinüber, O., 1978. “On the Tradition of Påli Texts in India, Ceylon and
Burma”. In Buddhism in Ceylon and Studies on Religious Syncretism
in Buddhist Countries, edited by H. Bechert, Göttingen
Norman, K.R., 1990. “Påli Philology and the Study of Buddhism”. In The
Buddhist Forum, edited by T. Skorupski. London´: SOAS.
——— 2006. A Philological Approach to Buddhism, 2nd ed. The BukkyØ
DendØ KyØkai Lectures 1994, Lancaster´: PTS (1st ed. 1997)
Oberlies, T., 2001. Påli ´: A Grammar of the Language of the Theravåda
Tipi†aka, Berlin
Warder, A.K., 1963. Introduction to Pali, London´: PTS
Zürcher, E., 1959. The Buddhist Conquest of China, Leiden
106
The Buddhist Bhik¡u’s Obligation to Support His
Parents in Two Vinaya Traditions
One might get the impression from Buddhist literatures that the
Buddha always had something to say about whatever topic or issue
arose, even if — as in the well-known case of the avyåk¤tavastus or
“indeterminate questions” — it was only to say that that topic was not
worth discussion.1 Indeed, it is very, very rare to find the Buddha
presented as, in effect, throwing up his hands and declaring that it was
not in his power to say or do something about something. But while
very rare, such presentations do occur and they are always interesting,
one particularly so — they all, in one way or another, point to
immovable principles or established boundaries.
Some of these boundaries or principles are obvious enough, and the
texts then simply give them a striking clarity. In the Bhik¡uˆ¥vinaya of
the Mahåsåghika-Lokottaravådins, for example, the fixed principle
that Buddhist monastic rules apply only to ordained bhik¡us or bhik¡uˆ¥s
is expressed in one of the rare passages in which the Buddha is
presented as having nothing to say. The issue here is what is to be done
when a bhik¡uˆ¥ behaves badly (vipratipadyate) with an åråmika or
“monastery attendant”. The bhik¡uˆ¥’s misbehavior is clearly sexual,
and the text in fact begins with similar misbehavior involving a bhik¡uˆ¥
and a bhik¡u or “novice” (ßråmaêra). There is, of course, no mystery
about what is to be done in these cases. É. Nolot renders the first part of
the text :
´
Si un moine faute avec une nonne, d’un commun accord, les deux sont
exclus (ubhaye påråjikå bhavanti). Si une nonne faute avec un novice, la
nonne est exclue et le novice doit être expulsé (nåßayitavya).2
1 On the “indeterminate questions” see most recently Ruegg 2000 and the
sources cited.
2Nolot 1991, p. 63, §´117´; translating Roth 1970, pp. 78–79, §´117.
108 Gregory Schopen
Even though the full significance of both the terms påråjika and
nåßayitavya is far more complicated and unsure than common
translations or paraphrases might suggest,3 what is found here is by and
large what could have been expected : in the case of sexual misconduct
´
between a bhik¡u and bhik¡uˆ¥ both are said to be, in Edgerton’s words,
“deserving of expulsion” ;4 in the case of sexual misconduct between a
´
bhik¡uˆ¥ and a “novice”, the former, again, becomes påråjika, but the
latter, the text indicates without actually saying so, cannot — only a
bhik¡u can commit a påråjika offence, and a “novice”, obviously, is not
yet that. He can only be “expulsé”, and whatever that might actually
mean, at the very least the text would seem to be indicating that the
“novice’s” actions were subject to a further formal action of the Sa¤gha
or Community ; i.e., the Buddha had something to say about them. It is,
´
however, otherwise in the next case the text takes up, and here we have
a first instance where the Buddha is presented as expressing his limits,
and in this instance the limits of monastic rules.
The text then says, and here again the Buddha is speaking,
atha dåni bhik¡uˆ¥ a(å)råmike[na] saha vipratipadyate bhik¡uˆ¥ påråjikå
bhavati | åråmiko ag¤h¥ta-samvaratvåt kim vradyi¡yati5 | eva t¥rthikena |
Si une nonne faute avec l’intendant d’un monastère, la nonne est exclue´;
quant à l’intendant du monastère, puisqu’il n’a pas acquis la retenue [qui
découle de l’observance des règles disciplinaires], que peut-on lui dire´? —
de même avec un autre renonçant.
What we have in this instance, and in rather clipped diction, is a

first variant of a rhetorical question that will be met again ; k i m ´
3Itis becoming ever more clear, for example, that the commission of a påråjika
did not necessarily involve “exclusion” in the mainstream Indian Vinaya
traditions´; see Schopen 1998, pp. 157–79 (= Schopen 2004A , pp. 260–84´)´;
Clarke 2000. Dr Clarke will be treating the issue in much greater detail in, one
hopes, the reasonably near future. On nåßayitavya see Hüsken 1997A.
4BHSD s.v. påråjika.
5Read vucyi¡yati, with the ms according to Nolot. For Nolot’s correction of the
reading in Roth’s Bhik¡uˆ¥-Vinaya see Nolot 1991, p. 472, §´117.
The Buddhist Bhik¡u’s Obligation to Support His Parents 109
vucyi¡yati ; and it is very likely that Nolot has got the nuance right in her
´
translation “que peut-on lui dire ?” In effect, the Buddha first appears to
´
be saying that nothing will be, or can be, said about the åråmika. But in
this instance a specific reason is given ; nothing will be or can be said
´
about the åråmika because the åråmika is not subject to the authority of
the rule “from the fact that he has not accepted the [monastic]
restraint[s]”.6 The next sentence — equally clipped — can then be taken
in two ways. Eva t¥rthikena could be taken to mean “it is just as with a
member of another religious group”, i.e. the action of an åråmika is like
the action of a t¥rthika — neither is subject to Buddhist monastic rule,
so the Buddha declares he will have nothing to say about it. Here the
sentence is explanatory. But it could also be taken to mean : “it is just so
´
[when the bhik¡uˆ¥’s activity is] with a t¥rthika”, i.e. the sentence is
extending the judgement of the case involving an åråmika to a case
involving a member another religious group. Either way, the function of
the rhetorical question seems clear enough, and the same would seem to
hold, though it involves a different kind of principle, in a second text
that can be cited.
A second instance where this type of rhetorical question is put into
the mouth of the Buddha occurs in the K¡udrakavastu of the MËla-
sarvåstivåda-vinaya, in its account of the death of Mahåprajåpat¥ and the
five hundred young bhik¡uˆ¥s who accompany her.7 After Mahå-
prajåpat¥ was corrected or scolded for not honoring the Buddha in the
proper way, she determines to enter final nirvåâ. She declares her
intention to the Buddha, and to a series of Elders who are all related to
her, including Ónanda. To anyone familiar with the account of the death
of the Buddha himself, this would set up the expectation that her
6On the åråmika see most recently Yamagiwa 2002.

7 K¡udraka-vastu, Derge, ’dul ba Tha 110a.6–113b.3. Reference here and
throughout for Tibetan sources is to the Derge printing reprinted in Barber
1991, and will follow the same format. Unless otherwise noted they are to the
bka’ ’gyur, and will give the Indian title of the work, the section of the bka’
’gyur, the volume letter, and the original folio number(s).
110 Gregory Schopen
announcement would elicit — as it did in the case of the Buddha — a
request that she not pass away : she was, after all — again like the
´
Buddha — the “founder” and central pivot of the Order of Bhik¡uˆ¥s.

But in this case no such request is forthcoming. Instead, using a version
of the same rhetorical question we have seen, the Buddha says,
gau ta m¥ ’dus byas thams cad ni de lta bu’i chos can yin na smra ci dgos |8
Gautam¥, since all compounded things have such a nature, what can I say´?
He then repeats this to the five hundred young bhik¡uˆ¥s who

announce a similar intention and it is repeated a third time to Mahå-
prajåpat¥ and the bhik¡uˆ¥s by the various Elders.
There is very little doubt that ngas smra ci dgos here is rendering
something like ki vadeyam — the first person is certain, the tense or
mood a little less so. And here again the Buddha is presented as saying,
in effect, that he is powerless to say or do anything about the issue at
hand. But as in the Mahåsåghika-Lokottaravådin Bhik¡uˆ¥vinaya, here
too a specific reason is given for his inability : the Buddha can say
´
nothing about Mahåprajåpat¥ passing away because all compounded

things pass away — this is the way it is. The MËlasarvåstivådin account
of the death of Mahåprajåpat¥ does not, however, simply present us with
a second, similar instance of the Buddha verbally expressing —
however rhetorically — his limitations. It also provides a first instance
in which bhik¡us do, and there are other examples of this as well,
examples which might offer aid in understanding an otherwise curious
passage in the Påli Vinaya which will momentarily concern us.
There is, for example, another such instance in the K¡udrakavastu
itself. In this text — part of which has already been treated elsewhere9
— the bhik¡us of the Jetavana, in reaction to an attack by a group of
bhik¡uˆ¥s on another leading bhik¡uˆ¥, and out of fear that bhik¡uˆ¥s
might be carrying concealed weapons, make a kriyåkåra, or local
8K¡udraka-vastu, Derge ’dul ba Tha 111b.4.

9The text is found at K¡udraka-vastu, Derge ’dul ba Da 172b.2–174b.5´; the
treatment elsewhere at Schopen 1996 (= Schopen 2004A, pp. 329–59).
monastic ordinance, forbidding bhik¡uˆ¥s to enter the vihåra. The text
goes on to say that it was, prior to this, the customary practice of
Mahåprajåpat¥ to come daily and honor the feet of the Blessed One, but
after the kriyåkåra was put in place she was denied admittance. She
says, “Noble Ones, they (i.e. the other bhik¡uˆ¥s) have committed a
fault, but have I also done so ?”. But the bhik¡us respond by saying,
´
dge ’dun rnams de ltar bzhed pa la kho bo cag gis byar ci yod |10
In that the Communities have wished it so, what can we do´?
The biggest obstacle to immediately recognizing that this case is

essentially analogous to our first two is, probably, the fact that so little
is known about the force and function of local monastic ordinances. It
is, however, already known that — at least according to the MËla-
sarvåstivådin tradition — membership in a given monastic community
during the rain retreat was
determined not by acceptance of, or willingness to adhere to, a specific
Vinaya or monastic rule, but by the acceptance of, or willingness to adhere
to, these specific local ordinances.11
It is the local ordinances that must be announced at the ritual

preliminary to undertaking the rains retreat, not the Pråtimok¡a ; and a´
willingness to accept them — signaled by taking a counting stick

(ßalåka) — makes a bhik¡u a member of the community, not his
ordination. That kriyåkåras were even more binding than the canonical
monastic rule or Pråtimok¡a is also suggested from an unlikely source : ´
the comparatively late BodhisattvabhËmi says in a number of instances

that a bodhisattva might disregard, or act contrary to, a Pråtimok¡a rule,
but — again in several instances — it indicates that he must not act
contrary to “a local community ordinance” (såghika kriyåkåra),
even if adhering to a kriyåkåra results in what the authors of the
BodhisattvabhËmi would otherwise consider a fault (åpatti).12 This
10K¡udraka-vastu, Derge ’dul ba Da 174a.4.

11Schopen 2002, especially p. 361, for the quotation, and what follows here.
12 For numerous examples see Wogihara 1936, pp. 161.11ff. ´; Dutt 1966,
112 Gregory Schopen
would seem to be a tacit recognition by Mahåyåna authors that at least
local ordinances were strong enough to constrain some of their
ideological innovations, even if Pråtimok¡a rules were not. Indeed, the
strength of local ordinances is suggested as well in the same text when it
makes “keeping the local community ordinance’ (s å g h i k a
kriyåkåra anurak¡ata) parallel with “keeping royal dictate (råjå-
pathyam anurak¡ata).13
Seen in the light of such considerations, the bhik¡us’ response to
Mahåprajåpat¥ says, in effect, that she is asking them to do something
they cannot do. The kriyåkåra does not allow them to do otherwise — it
is beyond their control. In this regard the present case is all of a piece
with our first two. The bhik¡us can no more alter the terms of the
kriyåkåra than the Buddha can alter the fact that all compounded things
pass away, or make rules governing the behavior of non-monastics or
non-Buddhists. All of these are simply not possible, and while this is yet
another indication of the strength or reach of kriyåkåras, what is
important to keep in mind for our immediate purposes is the fact that it
is only in such situations that the Buddha or his bhik¡us use rhetorical
questions like “Que peut-on dire ?” “What can I say ?”, or “What can we
´ ´
do ?” This, in turn, would seem to make the construction of the main
´
Påli passage that we will be concerned with here that much more
curious.
What has already been noted in general terms holds, of course, for
the Påli Vinaya : instances where the Buddha expresses his inability to
´
say something about something are very rare in it. In fact there may
only be one clear case, and that alone renders this case notable, and
highlights the issue in regard to which it occurs. This case is also odd in
another respect as well : it may be one of the equally rare instances in
´
pp.´110.15ff.; for a translation, Tatz 1986, pp. 66ff., where kriyåkåra, khrims
su bca’ ba, is translated as “internal rule”. On the composition and date of the
BodhisattvabhËmi see most recently Deleanu 2006, Vol. I, pp. 162–67, 194–
96.
13Wogihara 1936, 178.2; Dutt 1966, 122.15; Tatz, 1986, p. 81.
this Vinaya where the Buddha expresses himself using the first person
plural. The case concerns a Buddhist bhikkhu giving material goods to,
or, in effect, supporting his parents, and it occurs in the C¥vara-
khandhaka. It reads :´
tena kho pana samayena aññatarassa bhikkhuno bahu c¥vara uppanna

hoti so ca ta c¥vara måtåpitunna dåtukåmo hoti. Bhagavato etam
attha årocesum. måtåpitaro hi kho bhikkhave dadamåne ki vadeyyåma.
anujånåmi bhikkhave måtåpitunna dåtu. na ca bhikkhave saddhådeyya
vinipåtetabba. yo vinipåteyya, åpatti dukka†asså ti.14
I.B. Horner has translated this as

Now at that time much robe-material accrued to a certain monk, and he was
desirous of giving that robe-material to his parents. They (i.e. the bhikkhus)
told this matter to the Lord. He said: “Because he is himself giving to his
parents, monks, what can we say´? I allow you, monks, to give to parents.
But, monks, a gift of faith should not be brought to ruin. Whoever should
bring [one] to ruin, there is an offence of wrong-doing.”15
Rhys Davids and Oldenberg’s earlier translation of the first part of the
Buddha’s statement here was somewhat looser and padded out.
Since they are his father and mother, what can we say, O Bhikkhus, though
he give them to them. I allow you, O Bhikkhus, to give [robes, in such a
case,] to your parents.16
Neither of these translations is, of course, a bad one, but, perhaps,

for one thing. And that thing affects the translation of far more than this
single passage. In both translations the Buddha says he allows bhikkhus
to give to their parents. The verb here is anujånåmi, and this verb occurs
hundreds, if not thousands of times in Buddhist Vinaya texts. But
H.´Bechert has rejected such a translation on more than one occasion.
He has said, for example, that “most Vinaya interpreters down to the
present day have translated the word anujånåmi as ‘I permit’, ‘I allow’,
14 Vin I 297–98. All references to Påli texts are to their Påli Text Society
editions.
15BD IV, pp. 424–25.
16Rhys Davids and Oldenberg 1882, p. 232.
114 Gregory Schopen
both of which are incorrect in this context [he is talking about a typical
passage in which the Buddha delivers a rule] : it means ‘I order’ here”.
´
In another place he also said, “Unfortunately, some authors still

translate anujånåmi in most passages in the Vinaya-pi†aka with ‘I
allow’. However, when the Buddha speaks, the appropriate translation
in most cases is ‘I prescribe’.”17 K.R. Norman has more matter-of-
factly — as is his wont — said, “It is well-known that in the common
Vinaya phrase anujånåti means ‘to ordain or prescribe’.”18
The implications of these remarks for understanding our present
passage are not difficult to see. Read in light of Bechert and Norman,
the Buddha of our passage does not “allow” bhikkhus to give to their
parents, he requires it ; he does not “permit” bhikkhus to do so, he
´
orders it ; and this, of course, is a very different thing. Note too that the
´
parenthetical padding supplied in the translation of Rhys Davids and

Oldenberg only makes more obvious the unrestricted range of the text
itself. They have : “I allow you, O Bhikkhus to give [robes, in such a
´
case,] to your parents.” The padding appears to be intended by the

translators to limit the prescribed (their “allowed”) giving to “robes”,
and to situations in which there is “much [bahu] robe-material” or cloth.
The text itself, however, says no such thing. There, the material to be
given is not specified and is therefore, unrestricted ; there, no specific
´
circumstances are stated and, therefore, no temporal limitations. And

this too is a very different thing.
Then there is the rhetorical expression of the speaker’s inability.
Here — as in our first examples — the speaker is the Buddha ; and here
´
too — as in all the previous examples — that expression is preceded by

an explanatory reason. But here the force of the explanatory reason is
not, perhaps, so immediately obvious. While it is easy enough to see
why the Buddha would have nothing to say about the actions of an
individual who was not a member of the group who recognized his
17Bechert1993, p. 7´; Bechert 1982, p. 63; see also Bechert 1968´; Bechert 1997,
p. 58.
18Norman 1992A (= Norman 1994).
authority — an åråmika or t¥rthaka — and it is easy enough to see how
he would ask how he could possibly have anything to say about a
situation governed by what was — for him — the inexorable fact that
all conditioned things must pass away, it is, however, more difficult, at
least for us, to understand why an act of one of his bhikkhus would
leave him speechless just because it was being done to benefit or
support that bhikkhu’s parents. This might be especially so since this
was ostensibly the same Buddha who had also said — to quote only one
possibly early example — that a follower of his “leaving behind son and
wife, and father and mother … should wander solitary as a rhinoceros
horn”, and should be “one who does not support another” (a n -
aññapos¥).19 This would seem to remain something of a mystery, even if
it be noted — and this rarely is — that Buddhist Vinaya and Buddhist
sËtra literature often do not say the same thing or express the same, or
even similar, values.20
It might be possible to explain the Buddha’s statement here as an
unwanted consequence of an already taken decision. Already by Aßoka
followers of the Buddha were publicly called bhikhus and bhikhunis,21
and whatever else this might have meant, it certainly identified them as
“beggars”. But to judge from Indian normative texts, this identification
would have in turn carried with it certain unavoidable expectations, at
least in a brahmanical or even brahmanized world. The Ópastamba-
dharmasËtra says, for example, in Olivelle’s recent translation,
The appropriate reasons for begging [bhik¡aê nimitta] are the following´:
to pay the teacher, to celebrate a marriage, to perform a sacrifice, trying to
support one’s parents [måtåpitror bubhËr¡a], and when a worthy person
19Sutta-nipåta, vv. 60, 65. The translation is from Norman 1992B. For the first
of these cf. Salomon, 2000, p. 108 (v. 18), pp. 144–46 (v. 18).
20See, from a somewhat different angle, Bronkhorst 2006, pp. 9–22, especially
pp.´21–22.
21Bloch 1950, pp. 152, 153, 155. Curiously, and as noted long ago by Lüders
(1963, p. 2, n.1) the terms “bhikhu or bhicchhu (bhikshu) for monks are never
used in Bhårhut inscriptions”, although bhikhun¥ or bhicchhun¥ are, and both
“occur very often” at Såñc¥.
116 Gregory Schopen
would have to suspend an obligatory act.22
And the Månavadharmaßåstra, again in Olivelle’s recent translation : ´
A man seeking to extend his line, a man preparing to perform a sacrifice, a

traveller , a man who has performed the sacrifice at which all his
possessions are given away, a man who begs for the sake of his teacher,
father or mother [pit¤måtrartha], a student of the Veda, and a sick man —
these nine should be known as “bath-graduates” [snåtaka], Brahmins who
are beggars pursuant to the Law [dharmabhik¡uka].23
It is possible that passages like these might explain something of

our Påli text. At the very least they suggest that any group that insisted
on calling its members bhikkhus or bhikkhun¥s — that is to say,
“beggars” — in an Indian setting in which dharmaßåstric values and
expectations were current might well have found it impossible to forbid
its followers to give material support to their parents, or to avoid the
expectation that its members would be engaged — at least in part — in
providing their parents with material goods. This, after all, would have
been, in such a place, one of “the appropriate reasons for begging
(bhik¡aâ)”, and begging is what a bhik¡u is supposed to do.
We might, then, have in these brahmanical sources on dharma the
elements of a possible explanation of why in the Påli Vinaya the
Buddha, when confronted by one of his bhikkhus wanting to give cloth
to his parents, is presented as being unable to say anything about it. But
while it might be a possible explanation, it is not a certain one, and there
are some further considerations which might render it even less so. To
have the heavy inexorability of, for example, the notion that all
conditioned things must pass away, these dharmaßåstric norms and
expectations would have to have had a cultural facticity that only would
have held in a thoroughly brahmanical environment. But we know that
the Påli Vinaya was certainly redacted and continuously edited in Sri
Lanka, and that there is little or no evidence of the presence of Indian
22Ópastamba-dharmasËtra 2.10.1 (Olivelle 2000, pp. 86–87).

23Månava-dharmaßåstra 11.1–2 (Olivelle 2005, pp. 215, 837).
dharmaßåstric literature or practices there.24 It could, of course, be
argued that such considerations would lose much of their force if our
passage could be taken as an old, established survival of the North
Indian tradition where brahmanical norms would have been, more
reasonably, a factor to be contended with, but this in turn would raise
the issue of how well our text was integrated into, and was
representative of, the Påli Vinaya as a whole, and here too there are
questions.
There is, indeed, much that is unusual about this passage. It may
well be the only place in the Påli Vinaya where the Buddha is presented
as unable to prohibit or alter a practice of one of his bhikkhus. It appears
to be the only place in the entire Påli Vinaya where the Buddha, in
referring to himself, uses the first person plural of the verb “to speak”. It
is the only passage there where there is any mention of a “gift of faith”
(saddhådeyya) being “brought to ruin” (vinipåtetabba, vinipåteyya).
In fact the compound saddhådeyya is itself very rare in the Påli Vinaya
and the structure of the text itself is not typical.25 From the point of
view of vocabulary, then, the text could hardly be called representative,
and if Rhys Davids and Oldenberg are right in assigning the term
dukka†a, or “wrong-doing”, to “the latest portion” or “final recension”
of the Påli Vinaya, then our passage also would not appear to be early.26
There are other factors as well that would seem to point in the same
direction.
The Påli text prescribing that a bhikkhu must give material goods to
24 In contrast with other Theravåda countries in Southeast Asia — see, for

example, Lingat 1949´; Schopen, Schopen 2004A, 186, and the sources cited
in n. 48´; 210 and n. 56.
25For vadeyyåma and saddhådeyya in the Påli Vinaya see Ousaka, Yamazaki,
Norman 1996.
26 Rhys Davids and Oldenberg 1881, p. xxv — they are responding here,
however, to a “trenchant attack upon Buddhist morality” based on the Påli
Vinaya by S. Coles, “the first ardent student of Pali among the missionaries
after the time of Gogerly”´; see Young and Somaratna 1996, pp. 171–72, and
n.´372.
118 Gregory Schopen
his parents is a little one of not much more than six lines, tucked away
in the middle of the C¥vara-khandhaka. It is also an isolated one. Neither
the prescription nor the practice are referred to elsewhere in the Påli
Vinaya in regard to bhikkhus, and there appears to be but a single,
curious or indirect reference to the practice in regard to bhikkhun¥s.
Påcittiya 28 of the Påli Bhikkhun¥-påtimokkha has recently been
translated by K.R. Norman as
If any bhikkhun¥ should give recluses’ robe material [samaâ-c¥vara
dadeyya] to a householder or to a male wanderer [paribbåjaka] or to a
female wanderer, there is an offence entailing expiation.27
But the exception clause (anåpatti) attached to this rule in the

Bhikkhun¥-vibha¤ga gives as the first exception : ´
anåpatti måtåpitunna deti.28

There is no offence if she gives [it] to [her] parents.
The first or most obvious thing about this reference to a member of

the Buddhist community giving material things to her parents is that it
does not occur in the Påtimokkha rule itself, but in the exception clause
attached to it in the Vibha¤ga and since von Hinüber, for example, has
more than once suggested that these exception clauses represent the
latest layer in the Vibha¤ga,29 this reference — if he is correct — would
appear to be, like the reference in the C¥vara-khandhaka, not an early
one. It is, moreover, worth noticing that the Bhikkhun¥-vibha¤ga does
not order the bhikkhun¥ to give to her parents. It does not even
recommend it. At the most it allows for it, and tacitly recognizes it as a
practice by citing it as an exception to the rule. This might seem curious
if the prescription now found in the C¥vara-khandhaka was already in
place or known. But the situation in post-canonical and Sri Lankan
sources — which we cannot follow here in any detail — is, if anything,
27Pruittand Norman 2001, pp. 172–73.

28Vin IV 286,3.
29See, for example, von Hinüber 1996, p. 14.
even more curious.
In regard to post-canonical Påli sources we can only note here, for
example, that there is a whole series of “stories of the present”
(paccuppannavatthu) in the Commentary to the Jåtaka dealing with
bhikkhus who are supporters (posaka) of their parents, but these
bhikkhus are repeatedly still being critiziced there by other bhikkhus for
doing so, and the criticism is expressed by citing a part — but only a
part — of what looks like the C¥vara-khandhaka prescription.30 The
critical bhikkhus say to the bhikkhu who gives to his parents, åvuso
satthå saddhådeyya vinipåtetu na deti. tva saddhådeyya gahetvå
gih¥na dadamåno ayutta karos¥ ti (“the Teacher does not allow a gift
of the faithful to be brought to ruin. In giving to householders after you
have taken a gift of the faithful you do what is improper”).31 Since the
critical bhikkhus know that the householders in question are the other
bhikkhu’s parents, this should have been an issue already resolved in the
C¥vara-khandhaka if the C¥vara-khandhaka passage looked at the time
of the Jåtaka Commentary as it does now. Equally odd, perhaps, is the
textual fact that in these stories when a bhikkhu decides to support his
parents while remaining a bhikkhu — and it is there his decision — he
cites as the justification not the prescription in the C¥vara-khandhaka,
but a statement that is attributed to the Buddha that appears not to be
found in the Påli Canon, but is not unlike a statement found in a very
different Vinaya : satthå pana pabbajitaputto va upakårako nåma ti
´
vadati (“But the Teacher says, ‘Even a son who has entered the
30The “stories of the present” in the following Jåtakas deal with bhikkhus who
support their parents: Nos. 164* (Gijjha-jåtaka), 385 (Nandiyamiga-), 398*
(Sutano-), 399 (Gijjha-), 455* (Måtiposaka-), 484* (Sålikedåra-), 513*
(Jayaddisa-), 532* (Sona-Nanda-), and 540 (Såma-). The last of these
presents the fullest account of such a bhikkhu, and all those marked here with
an asterisk refer to it for a full account´; it is the source for what follows here.
For the Påli Jåtaka Commentary see von Hinüber 1998, especially pp. 16–24,
for both the paccuppannavatthu and måtiposaka bhikkhus.
31Ja VI 71,15.
120 Gregory Schopen
religious life is, indeed, one who provides support’”).32 And finally,
although the Buddha in these stories repeatedly praises the bhikkhu who
supports his parents in very strong terms, and wants to strengthen his
resolve (tassa ussåha janetukåmo),33 he nowhere here makes giving to
one’s parents a rule for bhikkhus as he had (already ?) done in the
´
C¥vara-khandhaka. How best to account for all of this is, of course, far
from clear. Nor is it immediately obvious how this disparate material
fits — or if it fits at all — with a good deal of inscriptional and
historical material from Sri Lanka that would seem to indicate that the
support of one’s parents by Buddhist bhikkhus was there a recognized
and established practice throughout the medieval period and virtually up
to modern times. Two examples must suffice. A Sanskrit inscription
dated to the ninth century was discovered more than a hundred years
ago at Anuradhapura. It is almost certainly a kriyåkåra or “local
ordinance” of the monastery in association with which it was found —
such ordinances in Påli sources are called katikåvatas. It specifies —
among other things — what kind of bhikkhu can or cannot reside in the
monastery. It indicates, for example, that bhikkhus “ordained at another
vihåra” can only reside in this one if they have given up their privileges
and duties in their original vihåra ; that bhikkhus who own or receive
´
land may not reside there. As a part of this enumeration it then says,
mitthyåj¥vinå na va[stavya |] [str¥po¡a]kena na vastavyam | anyatra
måtåpit¤bhyåm |34
[A bhik¡u] getting his living in a wrong way must not reside here. One who
supports a woman must not reside here, except for [one supporting his]
parents.
What Ratnapala calls the Mahå-Paråkramabåhu Katikåvata — a

twelfth-century set of ordinances promulgated in the name of the king
32Ja VI 70,14f.
33Ja VI 71,23.
34 Wickremasinghe 1904–1912, especially p. 4, line 12. The fact that this
inscription is in Sanskrit obviously raises the question of North Indian
influence´; see below and sources in n. 46.
— has a similar, but even more elaborate exception clause in regard to a
different issue :
´
No permission should be given to any of these [bhikkhus] to enter the

village at improper times on any business other than on account of a journey
begging food for the unsupported parents who had given birth to them,
likewise for their consanguineous and widowed elder and younger sisters.35
In both of these ordinances, which were meant to govern the

activities of Buddhist bhikkhus in medieval Sri Lanka, the practice of
bhikkhus supporting their parents is not only recognized, but apparently
had such significance and currency that other regulations had to be built
around it and could not put constraints upon it — it overrode all sorts of
other considerations. Here, however, the Sri Lankan material cannot be
further pursued, and we can only return to our main focus and try to
sum up what is found in the canonical Påli Vinaya that bears on the
issue of a bhikkhu supporting his parents.
A first point that could be made about the canonical Vinaya is that
if one sticks strictly to its wording it, literally, says nothing about a
bhikkhu supporting his parents. Unlike the Jåtaka Commentary, it does
not use a word corresponding to the English word “support”. It uses
forms of the verb “to give” — the bhikkhu “desires to give” (dåtukåma),
is “giving” (dadamåna), and is ordered “to give” (anujånåmi … dåtu)
to his parents. This is ambiguous and, as we have seen, allows Rhys
Davids and Oldenberg to put restrictions on the giving by padding their
translation and limiting that giving to “robes” and to situations where
there is an abundance of them. But again, strictly speaking, the order is
simply to give, with no limits put on the objects that must be given, and
no temporal or situational qualifications put on the obligation. Then
there is the textual fact that the order itself is delivered almost
unwillingly — the Buddha is made to say ki vadeyyåma, as if he had
no choice. There is the additional textual fact that the language of the
text which contains the order is unusual — the Buddha using a first
35 Ratnapala 1971, p. 131 (§´9)´; for other examples see pp. 148 (§´78), 156
(§´87), 169 (§´103), 176 (§´14).
122 Gregory Schopen
person plural form of the verb ; the reference to the “gift of faith”
´
(saddhådeyya), etc. There is as well the fact that the prescription is an

isolated one and had no afterlife — it, like the practice it enjoins, is
referred to nowhere else in the Påli Vinaya except in an exception
clause in the Bhikkhun¥-vibha¤ga. Finally, and as already noted from a
different angle, the prescription in the Påli Vinaya is completely lacking
in detail and specificity — it is, for example, only indicated in later
sources like the Jåtaka Commentary, or the cited katikåvata, that
bhikkhus who give to or support their parents do so by begging. In short,
the situation in the Påli Vinaya is a thoroughly unsatisfactory one : a ´
bhikkhu who had to depend on it would have no clear guidance. In this

regard — if no other — his northern brothers would have been much
better off.
A second Buddhist Vinaya — one that more certainly circulated in

India — is slowly becoming better known. Like the Påli Vinaya or,
indeed, all the Vinayas that have come down to us, this Vinaya, the
MËlasarvåstivåda Vinaya, appears to have been redacted relatively late,
but its redaction almost certainly took place in North or Northwest India
in an environment where brahmanical norms and expectations were a
presence that had to be addressed. In this Vinaya the rules governing a
whole series of issues — contact with corpses, inheritance of property,
for example — were fairly obviously framed in such a way that
Buddhist practice would accommodate and incorporate larger
brahmanical values.36 In such an environment there is little doubt that
the kind of brahmanical expectations concerning beggars expressed in
the passages from Ópastamba and Manu already cited would have been
well fixed, and this, in turn, may account in good part for the fact that
the redactors of this Vinaya give no evidence that there was a felt need
to explain or justify the practice of Buddhist bhik¡us supporting their
36For contact with corpses and “pollution” see, for example, Schopen 1992,
Schopen 2006´; for inheritance, Schopen 1995 (= Schopen 2004A, pp. 170–
92), Schopen 2001 (= Schopen 2004A, pp. 122–69).
parents. Indeed, for Buddhist bhik¡us not to have done so might well
have required justification or explanation. In any case, it is clear that, in
regard to the practice of bhik¡us supporting their parents, the MËla-
sarvåstivåda-vinaya is everything the Påli Vinaya is not : unapologetic,
´
clear, and unambiguous ; detailed and specific ; the prescription it

´ ´
contains is well integrated, is referred to in places other than in the

passage in which it was originally delivered, and had a long life in
Vinaya handbooks. If nothing else, these factors make it much easier to
deal with the MËlasarvåstivådin rule in summary form, and that is all
that can be done here.
The text that delivers the MËlasarvåstivådin prescription needs little
commentary and is short enough to be translated in full. It occurs in
what is now called the Uttaragrantha, and although this is the least well
known section of the MËlasarvåstivåda-vinaya there are already clear
signs that it represents a particularly influential, and probably early, part
of this tradition.37 Since this portion of the MËlasarvåstivåda-vinaya is
not yet available in Sanskrit, the translation here is of the Tibetan
translation.38
The setting was in Íråvast¥.
A householder in Íråvast¥ took a wife from a suitable family and
made love with her. When he had made love with her, and when after that
she had become pregnant, then — eight or nine months having passed —
she gave birth to a son, a handsome boy who was a delight to see and
possessed of all his limbs. And having celebrated the festival of birth for the
newborn he was given a name.
The boy was nurtured and grew up, but then later — without asking
his father and mother — he entered the religious life (rab tu byung ba =
pravrajati) in the Order (bstan pa = ßåsana) of the Blessed One. In the
morning when he had dressed he, taking his bowl and robe, entered Íråvast¥
for alms, and when his old father saw him he said, “Son, since you have
37See Schopen 2001, pp. 101–105 (= Schopen 2004A, pp. 124–27).

38 The text occurs at Uttaragrantha, Derge ’dul ba Pa 112b.1–113a.1 — the
Sanskrit equivalents inserted into the English translation are all attested in one
form or another, but are, of course, only probable.
124 Gregory Schopen
entered into the religious life (pravrajita) who will support us´?” (nged cag
sus gso bar ’gyur = po¡i¡yati´; or “will be the supporter (po¡aka) of us´?”)´
The bhik¡u was crestfallen (spa gong nas = ma¤kubhËta) and said
nothing.
The bhik¡us reported to the Blessed One what had occurred, and the
Blessed One said, “Bhik¡us, one’s father and mother are the doers of what is
difficult (dka’ ba byed pa = du¡karaka) for a son. Therefore, I order (rjes su
gnang ngo = anujånåmi) that even a son who has entered the religious life
(bu rab tu byung yang = pravrajita-putra´?) must procure (sbyor ba = pra
yuj) food and clothing for both father and mother.”
When the Blessed One had said “even one who has entered the
religious life must procure food and clothing for both father and mother”,
and the bhik¡us did not know how it was to be provided, the Blessed One
said, “What there is beyond his bowl and robe — with that it must be
provided ´! If there is none, then begging from a donor (sbyin bdag =
dånapati), it must be provided´! If, as one who receives from the Com-
munity (saghalåbhin) he has a right to what belongs to the Community, he
must give half of that´! If he is one who begs his food (piˆ apåtika), he must
give half his alms´! If that is not done one comes to be guilty of an offence
(’das pa = atyaya).
There is — apart, perhaps, from the very idea of bhik¡us supporting

their parents — nothing very odd here. Certainly the language,
vocabulary, and conceptual world in this text are not at all unusual, but
rather typical, or even characteristic, of the MËlasarvåstivåda-vinaya as
a whole. The whole of the first paragraph, and most of the second, for
example, consists entirely of clichés or stenciled passages that occur
scores of times in the MËlasarvåstivåda-vinaya and the literature
dependent on it like the Avadånaßataka and Divyåvadåna.39 Indeed, the
construction of the text as a whole, the way in which the “action”
unfolds, is entirely typical of hundreds of texts in this Vinaya. The
assertion that parents are “the doers of what is difficult” (du¡karaka) is
repeatedly made in its Vibha¤ga, the Uttaragrantha again, its Bhai¡ajya-
and K¡udraka-vastus.40 The distinction in the text between bhik¡us
39Hiraoka 2002, p. 157 (3.A), p. 161 (3.H, I), etc.

40Vibha¤ga, Derge ’dul ba Ca 76a.7´; Uttaragrantha, Derge ’dul ba Pa 103a.4´;
supported by the Community (saghalåbhin) and those who beg
(piˆ apåtika) is found again in its Po¡adha-vastu and Vibha¤ga, for
example, and is so well established that it can occur in “stories of the
past” (bhËtapËrva), as it does in the Íayanåsana-vastu.41
Perhaps even more to the point, given the use of the expression
pabbajita-putta in the otherwise untraceable citation of the Buddha’s
words in the Jåtaka Commentary, in the MËlasarvåstivåda-vinaya the
Buddha himself and other bhik¡us are repeatedly seen using the term
pravrajita in making claims or assertions in regard to Buddhist bhik¡us
— easily available examples occur in its C¥vara- and K¡udraka-vastus,
Uttaragrantha, and Vibha¤ga.42 Even what appears to be an anomaly in
our text may not be one. The observant will have noticed that the boy in
our text entered the Order without asking his parents. This receives no
comment or correction, even though according to specific rule in this
Vinaya — and others — it should not have occurred.43 It is, of course,
possible to suggest that our text was set in a narrative time that preceded
that of the promulgation of the rule that required parental permission for
entrance into the Order, but the fact of the matter is that this kind of
situation — a narrative situation in which seemingly established rules
do not seem to be in place — occurs on a number of occasions in the
texts in the Uttaragrantha, and may be indicating that they represent an
early and partially independent strand of the MËlasarvåstivådin Vinaya
Bhai¡ajya-vastu, Derge ’dul ba Kha 5b.2 (= Divyåvadåna (Cowell and Neil)

51.20)´; K¡udraka-vastu, Derge ’dul ba Tha 253b.1. The same expression is
also found in early North Indian inscriptional sources´; see Salomon 1986,
p.´265 (8d.), 271 (8d.) — the inscription dates to the early first century C.E.
41 Po¡adha-vastu (Hu-von Hinüber) 292 (§20); Vibha¤ga, Derge ’dul ba Ca
147b.3; Íayanåsana-vastu (Gnoli) 41.4 — for the abbreviated references here
and below to Sanskrit MËlasarvåstivåda-vinaya sources see Schopen 2004A,
pp. xiii–xvii.
42For convenience see Schopen 2004 A , p. 115 (there “renouncers”) and 312
(“renouncers”); 181 (“renunciant”); 200 (“renunciant”)
43See, for example, the MËlasarvåstivådin ordination formulary translated in
Schopen 2004B, p. 236.
126 Gregory Schopen
tradition. Time will tell.
It seems, moreover, that unlike in the prescription in the Påli
C¥vara-khandhaka, there is little ambiguity in the MËlasarvåstivådin
text, even though it must be filtered through the Tibetan translation.
That the issue under discussion is the “support” of the bhik¡u’s parents
seems certain ; the father’s question uses the expression gso bar ’gyur,
´
which almost certainly is translating either the future form po¡i¡yati, or

a construction involving the substantive po¡aka plus a form of bhË, both
of which would have the general sense “support”, “nourish”, “foster”.44
The Buddha’s prescription uses the verb sbyor ba to describe what one
who has entered the religious life must do, and the sense of the Tibetan
verb in a context like this is “procure” or “provide”. The original was
very likely a form of pra-yuj which, of course, has a wide range of
meanings, among which are “prepare”, “produce”, or even “offer” or
“present” (a Sanskrit source we will shortly see uses the verb udvahet).
But if there is little ambiguity in the MËlasarvåstivådin text, there
appears to be even less ambivalence — certainly there is nothing like
the Buddha’s tacit statement in the Påli that he had no choice in
prescribing what he did. Finally, the MËlasarvåstivådin text — unlike
again in the Påli one — is detailed and specific ; it contains precise
´
instructions on how the prescription is to be fulfilled depending on a

range of different situations or circumstances.
But if there are distinct differences between the Påli and MËla-
sarvåstivådin texts in the form, language and delivery of the prescription
itself, there also appear to be distinct differences in regard to how well
their respective prescriptions are anchored or integrated into their
respective Vinayas, and in regard to its continuing influence. The Påli
prescription appears to have been, as already noted, an isolated one — it
is not, strictly speaking, referred to anywhere else in the Påli Vinaya.
This is not the case in regard to the MËlasarvåstivådin prescription.
Moreover, there is no doubt — as we will see in a moment — that the
44See Negi 2005, p. 7397, s.v. gso ba.

MËlasarvåstivådin prescription had a long shelf-life, whereas this has
yet to be demonstrated for the Påli tradition, and we have already seen
that in a place like the Jåtaka Commentary where bhikkhus are being
criticized for giving to or supporting their parents, and where one might
therefore legitimately expect to find reference to the prescription in
answer to the criticism, there is none. Here again, however, the
MËlasarvåstivådin material can be, and need only be, cited summarily,
using two particularly unequivocal examples.
One might, again, legitimately expect that if a rule like the
MËlasarvåstivådin prescription requiring bhik¡us to procure food and
clothing for their parents were fully integrated into their Vinaya, it
would be referred to elsewhere, and the MËlasarvåstivådin prescription
certainly is. A particularly striking example occurs in its Vibha¤ga, in
the longish section there devoted to the issue of bhik¡us paying tolls.
The text concerns a bhik¡u from Íråvast¥ and says,45
Once, while wandering through the countryside, when he obtained two great
pieces of cotton cloth he thought to himself, “Since it has been said by the
Blessed One, ‘Even one who has entered the religious life (rab tu byung ba
= pravrajita) must provide (bstabs par bya ba = pratipådyati) the means of
support (mkho ba = upakaraâ) to his father and mother,’ I, therefore, will
give one of these to my father, the other I will give to my mother.”
The text continues with the bhik¡u being stopped at a custom-house

on his way back to Íråvast¥, and being asked if he had anything “on
which duties are wanted”. The bhik¡u says, “No”, but the customs-agent
finds the cloth and accuses him of lying. Then
The bhik¡u said, “But, sir, these two are not mine.”
“Then whose are they, Noble One´?”
“Sir, one is my father’s´; the other my mother’s.”
But the customs-agent is not impressed and says,

“Noble one, since I do not find your father here, nor do I find your mother,
you must pay the tax, then begone from here´!”
45Vibha¤ga, Derge Ca 75b.5-76b.4.

128 Gregory Schopen
When the bhik¡us report to the Blessed One what had happened, the text
says,
The Blessed One said, “Though indeed, bhik¡us, for that bhik¡u there is no
offence, still a bhik¡u should not proceed in this way, but should first
declare in this way the praises of parents to the customs-agent, ‘Sir, the
Blessed One has said, “Bhik¡us, one’s father and mother are the doers of
what is difficult (dka’ ba byed pa = du¡karaka) for a son — they are
nourishers, supporters, fosterers …”46 If he declares in this way the praises
of his parents, and if he is let go, that is good. If he is not let go, then paying
the tax, he should proceed. If he does not proceed in this way he comes to
be guilty of an offence.”
There are a number of points worth briefly noting here, the first and
perhaps most obvious of which is that the first part of what the bhik¡u
thinks here in the Vibha¤ga is a loose quotation or close paraphrase of
the rule delivered by the Buddha in the Uttaragrantha, and is marked as
such. Internal quotations from one part of this Vinaya in another are not
infrequent and are always explicitly marked as such — as here — with
the phrase “it has been said by the Blessed One”. Such “quotations” are
also — again as here — almost never verbatim.47
A second and perhaps more surprising point has already been
alluded to : what the bhikkhu in the Jåtaka Commentary thinks when he
´
decides to support his parents as a bhikkhu is also marked as a

“quotation” (satthå pana pabbajitaputto va upakårako nåma ti vadati)
but while this quotation cannot, it seems, be traced in the Påli Vinaya, it
is remarkably similar to what the MËlasarvåstivådin bhik¡u thinks in the
Vibha¤ga when he makes the same decision (notice in particular the
Påli upakåraka and the MËlasarvåstivådin upakaraâ, and the
uncharacteristic use of pabbajita in the Påli). This situation might be
most easily explained as yet another instance of the influence of
“Northern”, particularly MËlasarvåstivådin, sources on the Påli
46For the stenciled passage praising parents that I have abbreviated here see the
reference cited in n. 38 above; for a translation, Schopen 2004A, p. 179.
47For some instances and remarks on these internal “quotations” see Schopen
2004A, pp. 103–04, 179–80, 183, 230 and n. 41, 311–12, and 355, n. 44.
commentaries — Frauwallner, for example, goes so far as to say that
they are “met with at every step when one scans the pages of the
Dhammapada-a††hakathå”.48 But this raises again the issue of why an
“external” source would be cited if the rule now found in the Påli
C¥vara-khandhaka were already in place.
A final point that might be noted here must be that this Vibha¤ga
text would seem to present an example of precisely the sort of thing that
one might expect to find if the rule regarding bhik¡us to provide for
their parents had been fully integrated into its Vinaya, if the rule had
become a practice. Here, as it were, the rule is narrativized and appears
in a context other than the one in which it was originally promulgated.
Here the rule — like so many other rules in Buddhist Vinaya — gives
rise to further rules. And the text would seem to suggest that the practice
of bhik¡us providing for their parents was established to the point that
the redactors of the MËlasarvåstivåda-vinaya, at least, thought it was in
need of further regulation and established guidelines here meant to
ensure that bhik¡us who engaged in it would not run afoul of the law or
create problems with the state.
Beyond, however, a text like that cited from the Vibha¤ga there are
still other indications that the rule requiring bhik¡us to provide for their
parents was, and remained for a very long time, an integral part of the
MËlasarvåstivådin Vinaya tradition, none clearer, perhaps, than that
provided by Guâprabha’s Vinaya-sËtra. The Vinaya-sËtra is a
remarkable digest of the MËlasarvåstivåda-vinaya, written in true sËtra
style, that has come down to us in Sanskrit. Guâprabha appears to have
written his Vinaya-sËtra sometime between the fifth and the seventh
century, but we know from a colophon that it was still being copied in
the eleventh–twelfth centuries at Vikramaß¥la ; that at least four
´
extensive commentaries were written on it in India ; and that both the

´
sËtra and these commentaries were translated into Tibetan and became
48Frauwallner 1956, p. 188, and sources cited in his n. 4´; Ruelius 1968, p. 175
and the sources cited there´; Hüsken 1997B, pp. 204–205 and n. 20´; cf. Pind
1996.
130 Gregory Schopen
— and remain — an important part of a bhik¡u’s training in the Tibetan
using world.49
Certainly, one of the more remarkable things about Guâprabha’s
Vinaya-sËtra is that it has reduced the nearly eight thousand pages of the
canonical Vinaya to an even hundred. It did this in part, at least, by
ignoring the enormous mass of narrative material in this Vinaya, but
also by an almost breathtaking economy of expression and a tight focus
on the bare essentials. Given these general characteristics it is of some
interest that the rule requiring bhik¡us to provide for their parents is
treated in some detail, even if it is in a very compact form. The text
says,
yoga bhaktåcchådanena pitror udvahet |
na cel låbhasya påtrac¥varåd atirekas samådåpya |
asapattau bhojanopanater upårdhasyådånam |
( pha ma la zas dang gos kyis gtang bar bya’o |
gal te lhung bzed dang chos las lhag pa’i rnyed pa med na blangs te’o | ma
grub na zan gyi skal ba las phyed sbyin no |)50
As with most of Guâprabha’s sËtras these three are barely intelligible

on their own, hence the four enormous Indian commentaries written on
it. But very often knowledge of the canonical passage that he is
digesting turns out to provide the best “commentary” on a given set of
sËtras, and that is the case here. Armed with a knowledge of the
canonical text these sËtras can be translated — with the necessary
padding — as
[A bhik¡u] should conscientiously provide his parents with food and
49 There is not yet anything like a good overview of the complexities of the
Vinaya-sËtra and its associated literature, and little is actually known about
Guâprabha’s life´; for the moment see Schopen 2004A, pp. 64–69´; 86, n. 55´;
126–28´; 257, n.78´; 312–18, but there is also important and on-going work
being done on the Sanskrit text of the Vinaya-sËtra and VinayasËtrav¤tti by
Masanori Nakagawa and the Study Group of Sanskrit Manuscripts in Tibetan
dbu med script at TaishØ University which cannot be enumerated here.
50 For convenience the Sanskrit text is cited here from Sankrityayana 1981,
89.27, and the Tibetan from Derge, bstan ’gyur, ’dul ba Wu 72b.6.
clothing.
If there is no surplus from the bowl and robes of his acquisition, [it must
be done] after having incited [a donor to provide him with them].
When that does not succeed [the parents are to be] receiving half of the
[bhik¡u’s] share of food.
Although I am less confident than I once was that Guâprabha’s

Vinaya-sËtra necessarily reflects the precise vocabulary of the canonical
sources he was digesting — in fact his vocabulary not infrequently
seems far more learned and recherché than that of the canonical texts —
these sËtras are no less important for that. At the very least they provide
a version of the MËlasarvåstivådin rule in an Indian language. But
beyond that they also indicate the continuing circulation of the rule and
— one might assume — its continuing relevance for the MËlasarvåsti-
vådin tradition over a very long period of time. The latter, indeed, may
also be underscored by the fact that Guâprabha not only digests the
rule itself, he also separately digests the Vibha¤ga passage requiring
bhik¡us to pay tolls on goods they are transporting that are meant for
their parents.51
In trying to summarize what might have been seen here it is

important not to allow the uncertainties and complexities of especially
the Påli material to conceal or confuse what might be the essential
point : we have here two Vinaya traditions — thought by some to be the
´
earliest and the latest — in which Buddhist bhikkhus/ bhik¡us are

required to give or provide (the exact wording differs) material goods or
support to their parents. Although the integration of, and the attitude
toward, the requirement differs in the two traditions — both have such a
requirement. In one (the Påli tradition) the rule appears not to be well
integrated and there are indications that it might even be additive ; the ´
attitude towards the requirement there may appear to be ambivalent or

conflicted, even though the actual practice of bhikkhus supporting their
parents in Sri Lanka seems to be strongly suggested. In the other (the
51Sankrityayana, VinayasËtra 16.29´; Derge, bstan ’gyur, ’dul ba Wu 15a.2.

132 Gregory Schopen
MËlasarvåstivådin) the requirement appears to be fully integrated,
detailed and specific, and its practice further regulated (there is even an
incidental reference to bhik¡us borrowing money for the sake of their
parents in this Vinaya52). The attitude toward the requirement in the
MËlasarvåstivådin tradition appears to be straightforward and positive.
We have, in short, two positions on a common rule or practice, but it is
this shared common rule or practice that is important and that raises the
very real question of whether or not the Buddhist bhik¡u or bhikkhu can
legitimately, or usefully, be called a “monk”. Although the term “monk”
or “moine” or “Mönch” or any other equivalent is itself a contested, and
even — as everything today seems to be — a political term, still it
seems that by virtually any definition a “monk” would not be allowed to
do what our rule requires a Buddhist bhik¡u to do. So even though the
issue of whether or not a Buddhist bhik¡u might be called a “monk” is,
of course, not a new one,53 it almost certainly will become a more
pressing one as Buddhist Vinaya literatures start to be better known and
more carefully studied — all of them — and examples like the one
treated briefly here begin to pile up. And although it certainly involves
and implicates translation, the issue here is far more than an issue of
translation. At issue is the very nature of the group of celibate men who
created and transmitted Buddhist traditions : it might very well be that
´
this is not best described as “monasticism”. In his recent translation of

the Påli Påtimokkha K.R. Norman has once again not used the word
“monk”, but let a bhikkhu be a bhikkhu until we know what that is. We
might do well to follow him here, as we must in so many other ways as
well.
Gregory Schopen
52See Schopen 2004A, pp. 137–39.

53See, for example, the discussion in BD, Vol. I, pp. xxxix–l.
ABBREVIATIONS
BD The Book of the Discipline (Vinaya-Pi†aka), translated by I.B.
Horner. Vol. I, 1938 (Sacred Books of the Buddhists X)´; Vol. IV,
1951 (Sacred Books of the Buddhists XIV)
BHSD F. Edgerton, Buddhist Hybrid Sanskrit Grammar and Dictionary´;
Vol.ÍI, Dictionary. New Haven, 1953
BIBLIOGRAPHY
Barber, A.W., ed., 1991. The Tibetan Tripitaka´: Taipei Edition, Taipei
Bechert, H., 1968. “Some Remarks on the Ka†hina Rite”, The Journal of the
Bihar Research Society 54, pp. 320–21
——— 1982. “The Importance of Aßoka’s So-Called Schism Edict” in
Indological and Buddhist Studies´: Volume in Honour of Professor J.W. de
Jong on His Sixtieth Birthday, edited by L.A. Hercus et al.. pp. 61–68,
Canberra
——— 1993. “The Laws of the Buddhist Sangha´: An Early Juridical System in
Indian Tradition”, Hokke-Bunka KenkyË 19, pp. 1–11
——— 1997. “Die Gesetze des buddhistischen Sangha als indisches
Rechtssystem” in Recht, Staat und Verwaltung im klassischen Indien, ed. B.
Kölver, pp. 53–64, Munich
Bloch, J., 1950. Les inscriptions d’Asoka. Collection Émile Senart 8, Paris
Bronkhorst, J., 2006. “The Context of Indian Philosophy” in Conflict between
Tradition and Creativity in Indian Philosophy: Text and Context, edited by
T. Wada. 21st Century COE Program. International Conference Series,
No.´7, pp. 9–22, Nagoya
Clarke, S., 2000. “The Existence of the Supposedly Non-Existent Íik¡ådattå-
ßråmaêr¥´: A New Perspective on Påråjika Penance”, Buddhist Studies
(BukkyØ KenkyË) 29, pp. 149–76
Deleanu, F., 2006. The Chapter on the Mundane Path (Laukikamårga) in the
ÍråvakabhËmi. Studia Philologica Buddhica. Monograph Series 20a, Tokyo
Dutt, N., 1966. BodhisattvabhËmi´: Being the XVth Section of Asa¤gapåda’s
YogåcårabhËmi, Pataliputra
Frauwallner, E., 1956. The Earliest Vinaya and the Beginnings of Buddhist
Literature. Serie Orientale Roma 8, Rome
von Hinüber, O., 1996. A Handbook of Påli Literature. Indian Philology and
South Asian Studies 2, Berlin/New York
——— 1998. Entstehung und Aufbau der Jåtaka-Sammlung. Studien zur
Literatur des Theravåda-Buddhismus I. Akademie der Wissenschaften und
134 Gregory Schopen
der Literatur. Abhandlungen der Geistes- und Sozialwissenschaftlichen
Klasse, Jg. 1998, Nr. 7, Stuttgart
Hiraoka, S., 2002. Setsuwa no kØkogaku: Indo BukkyØ setsuwa ni himerareta
shisØ, Tokyo
Hüsken, U., 1997A. “The Application of the Vinaya Term nåsanå”, Journal of
the International Association of Buddhist Studies 20.2, pp. 93–111
——— 1997B. “A Stock of Bowls Requires a Stock of Robes. Relations of the
Rules for Nuns in the Theravåda Vinaya and the Bhik¡uˆ¥-Vinaya of the
Mahåsåghika-Lokottaravådin” in Untersuchungen zur buddhistischen
Literatur II. Gustav Roth zum 80. Geburtstag gewidmet. Sanskrit-
Wörterbuch der buddhistischen texte aus den Turfan-funden, Beiheft 9,
pp.´201–37, Göttingen
Lingat, R., 1949. “The Buddhist Manu or the Propagation of Hindu Law in
Hinayanist Indochina”, Annals of the Bhandarkar Oriental Research
Institute 30, pp. 284–97
Lüders, H., 1963. Bharhut Inscriptions. Corpus Inscriptionum Indicarum 2.2,
Ootacamund
Negi, J.S., 2005. Bod skad dang legs sbyar gyi tshig mdzod chen mo (Tibetan–
Sanskrit Dictionary), Vol. 16, Sarnath
Nolot, É., 1991. Règles de discipline des nonnes bouddhistes. Publications de
l’Institut de Civilisation Indienne, fasc. 60, Paris
Norman, K.R., 1992A. “Påli Lexicographical Studies IX”, JPTS XVI, pp. 83–84
———, tr., 1992B. The Group of Discourses (Sutta-Nipåta), Vol. II. Pali Text
Society Translation Series 45, Oxford
——— 1994. Collected Papers, Vol. V, pp. 7–78, Oxford
Olivelle, P., 2000. DharmasËtras´: The Law Codes of Ópastamba, Gautama,
Baudhåyana and Vasi¡†ha, Delhi
——— 2005. Manu’s Code of Law´: A Critical Edition and Translation of the
Månava-Dharmaßåstra, Oxford
Ousaka, Y., M. Yamazaki, K.R. Norman, 1996. Index to the Vinaya-Pi†aka,
Oxford
Pind, O.H., 1996. “Saddavimala 12.1–11 and Its MËlasarvåstivåda Origin” in
La pureté par les mots, edited by F. Bizot and F. Lagirarde, pp. 67–72, Paris
Pruitt, W., and K.R. Norman, 2001. The Påtimokkha, pp. 67–72, Oxford
Ratnapala, N., 1971. The Katikåvatas´: Laws of the Buddhist Order of Ceylon
from the 12th Century to the 18th Century. Münchener Studien zur
Sprachwissenschaft. Beiheft N, Munich
Rhys Davids, T.W., and H. Oldenberg, 1881. Vinaya Texts, Part I. The Sacred
Books of the East XIII, Oxford
———1882. Vinaya Texts, Part II. The Sacred Books of the East XVII, Oxford
Roth, G., 1970. Bhik¡uˆ¥-Vinaya´: Including Bhik¡uˆ¥-Prak¥râka and a Sum-
mary of the Bhik¡u-Prak¥râka of the Órya-Mahåsåghika-Lokottaravådin.
Tibetan Sanskrit Works Series Vol. XII, Patna
Ruegg, D.S., 2000. Three Studies in the History of Indian and Tibetan
Madhyamaka Philosophy. Studies in Indian and Tibetan Madhyamaka
Thought. Part I, Wiener Studien zur Tibetologie und Buddhismuskunde,
Heft 50, Vienna, pp. 152–56
Ruelius, H., 1968. “Some Notes on Buddhist Iconometrical Texts”, Journal of
the Bihar Research Society 54, pp. 168–75
Salomon, R., 1986. “The Inscription of Senavarma, King of O i”, Indo-Iranian
Journal 29
———, tr., 2000. A Gåndhår¥ Version of the Rhinoceros SËtra. Gandhåran
Buddhist Texts 1, Seattle/London
Sankrityayana, R. 1981. VinayasËtra of Bhadanta Gunaprabha. Singhi Jain
Íåstra Íiksåp¥tha. Singhi Jain Series 74, Bombay
Schopen, G., 1992. “On Avoiding Ghosts and Social Censure. Monastic
Funerals in the MËlasarvåstivåda-vinaya”, Journal of Indian Philosophy 20,
pp. 1–39
——— 1995. “Monastic Law Meets the Real World: A Monk’s Continuing
Right to Inherit Family Property in Classical India”, History of Religions
35, pp. 101–23
———1996. “The Suppression of Nuns and the Ritual Murder of Their Special
Dead in Two Buddhist Monastic Codes”, Journal of Indian Philosophy 24,
pp. 536–92
———1998. “Marking Time in Buddhist Monasteries: On Calendars, Clocks,
and Some Liturgical Practices”, in SËryacandråya. Essays in Honour of
Akira Yuyama on the Occasion of His 65th Birthday, edited by P. Harrison
and G. Schopen. Indica et Tibetica 35, Swisttal-Odendorf, pp. 157–79
——— 2001. “Dead Monks and Bad Debts´: Some Provisions of a Buddhist
Monastic Inheritance Law”, Indo-Iranian Journal 44, pp. 99–148
——— 2002. “Counting the Buddha and the Local Spirits in a Monastic Ritual
of Inclusion for the Rain Retreat”, Journal of Indian Philosophy 30,
pp.´359–88
——— 2004A. Buddhist Monks and Business Matters. Honolulu
——— 2004B. “Making Men into Monks” in Buddhist Scriptures, edited by
D.S. Lopez, Jr., pp. 230–51, London
——— 2006. “A Well-Sanitized Shroud: Asceticism and Institutional Values
in the Middle Period of Buddhist Monasticism” in Between the Empires´:
136 Gregory Schopen
Society in India 300 BCE to 400 CE, edited by P. Olivelle, pp. 315–47,
Oxford
Tatz, M., 1986. Asanga’s Chapter on Ethics with the Commentary of Tsong-
Kha-Pa´: The Basic Path to Awakening, the Complete Bodhisattva, Studies
in Asian Thought and Religion 4. Lewiston
Wickremasinghe, D.M. de Zilva, 1904–1912. “J„tavanåråma Sanskrit Inscrip-
tion”, Epigraphia Zeylanica 1, pp. 1–9
Wogihara, U. 1936. BodhisattvabhËmi´: A Statement of the Whole Course of the
Bodhisattva (Being Fifteenth Section of YogåcårabhËmi), Tokyo, reprinted
1971
Yamagiwa, N., 2002. “Óråmika — Gardener or Park Keeper´? One of the
Marginals around the Buddhist Sagha” in Buddhist and Indian Studies in
Honour of Professor Sodo Mori, pp. 363–85, Hamamatsu
Young, R.F., and G.P.V. Somaratna, 1996. Vain Debates´: The Buddhist–
Christian Controversies of Nineteenth-Century Ceylon. Publications of the
De Nobili Research Library 22, Vienna
Commentaries, Translations, and Lexica: Some
In his series of lectures on A Philological Approach to Buddhism 1

K.R. Norman has made an excellent case for the importance of
philology in the study of Buddhism. In what follows I shall attempt a
very modest addition to this picture by looking at some instances that
highlight how knowledge of the specific techniques and conventions
applied by indigenous commentators and translators can be of
importance for lexicography and the interpretation of Buddhist texts.
Consider first of all the following entry in CPD : “udara-jivhå-
!
ma!sa, n., ‘the flesh of the stomach’s tongue’ ; description of the

!
spleen : pihakan ti ~a", Vism 257, 22 (jivhå-saˆ†håna" udarassa

!
matthaka-passe ti††hanaka-ma"sa", mh† Se II 29,7).” The expression

“the flesh of the stomach’s tongue”, even though in some sense
“literal”, makes no sense at all, and although the commentary or
mahå†¥kå is quoted, it has clearly not been consulted or understood. The
commentary’s interpretation of the compound comes a lot closer to
describing what the spleen actually is : “The piece of flesh that is
!
located at the upper side of the stomach and has the shape of a
tongue.”2 This particular CPD entry results from a simple failure to read
and take advantage of indigenous commentaries. Matters are not always
so straightforward, and it can sometimes be difficult to know exactly
when we are in a position to “remonter … à un pali d’intérêt
linguistique”, to use an expression from Helmer Smith.3
1The BukkyØ DendØ KyØkai Lectures delivered at SOAS in 1994!; Norman

1997.
2That said, I feel somewhat surprised at the expression matthaka-passe
ti††hanaka-maµsaµ in this passage. Is it possibly an error for matthaka-passe

ti††hanakaµ maµsaµ? I cannot see why the commentator would opt for a
såpek!asamåsa here.
3Smith1928, p. vi, “C’est donc dans la conviction que notre pali est une
fonction de celui du 12 e siècle — et que la connaissance de la philologie
138 E.G. Kahrs
The CPD entry for udånana runs : “udånana, n., vb. noun of
!
udåneti q.v., formed to explain udåna ; Sadd 382,21 : ken’atthena

! !
udåna" ? °’atthena. kim ida" ~a" nåma ? p¥ti-vega-samu††håpito

udåhåro (! Ud-a 2,11, reading udån’atthena, udåna" nåma).” Turning
to the CPD entry for udåna, one finds : “udåna, n. and m. [ts.], lit. ‘the
!
breathing upwards’ ; 1. (medic.) one of the five vital airs, rising up the
!
throat and entering the head ; 2. a solemn utterance, mostly, but not
!
necessarily, in metrical form, inspired by intense emotion and made

without regard to any listeners … ; 3. the fifth of the nine a#gas
!
(divisions) of the Buddhist scriptures … ; 4 . the third book of the

!
Khuddaka-Nikåya divided into 8 vaggas and consisting of 80 suttas

each of which contains an udåna introduced by the standing phrase
ima" udåna" udånesi ; … in very frequent standing phrase ~a"
!
udåneti (udånento, °nesi, °netvå).” The entry goes on to list

occurrences. BHSD, on the other hand, has “udåna, m. or nt. (= Pali
id. ; with acc. pron. usually imam, sometimes idam), a solemn but
!
joyous utterance (acc. to PTSD sometimes a sorrowful one in Pali),

usually but not always having religious bearings ; almost always in !
modulation of phrase imam … udånam udånayati (usually with sma

after verb), very common.” Later in the entry we find : “nt. udånam, as !
n. of a type or class of Buddh. literature, one of the 12 (Mvy) or 9

(Dharmas) pravacanåni, Mvy 1271 ; Dharmas 62 ; Udåna-varga, n. of a
! !
specific work (abbreviated Ud).” BHSD also has the entry “udånayati,
denom., utters an udåna : used virtually always with object udånam,
!
q.v. for forms and passages.”

In Påînian grammar the term kåraka (lit. : “doer ; accomplisher”)
! !
applies to direct participants in actions. Such a participant is a sådhana,
birmane et singalaise de ladite époque est indispensable à qui voudra remonter,

à travers la recension Buddhaghosa–Dhammapåla, à un pali d’intérêt linguis-
tique —, que j’ai entrepris l’étude de la norme palie enseignée par Aggavaµsa
dans les trois volumes qui forment la Saddan¥ti.” Quoted also by Caillat 1971,
p. 84!; Kahrs 1992, p. 5!; and referred to by Norman 1983, pp.!6, 165. See also
von Hinüber 1978.
Commentaries, Translations, and Lexica 139
a means of realising the action,4 and every participant is assigned to one
of a set of six kåraka categories. 5 The abstract syntactic level at which
kårakas are introduced in the grammar serves to mediate between the
levels of semantics and morphology. By this device Påîni is able to
account for the relationship between possible semantic choices on the
side of the speaker and some basic features of Sanskrit syntax and
morphology. Such participants in actions are, at the abstract level of
kåraka syntax, assigned to particular kåraka-categories. In a similar
way the set of lakåras, a set of ten suffixes marked with an L , are
introduced after verbal roots at the same abstract level. At this level all
verbal endings, except for the ones that denote bhåva (lit. : “being ; state
! !
of action”), can be said to signify agents and objects in relation to

activities. By A 3.4.69 la˙ karmaî ca bhåve cåkarmakebhya˙ an L-
suffix is added to a verbal root to denote — in addition to the agent
(kartari, A 3.4.67) — the object, or, in the case of verbal roots which
are objectless (akarmaka, that is, intransitive verbs), the mere activity
expressed by the verbal root (bhåva). When the relevant semantic
choices have been considered on the side of the speaker, the abstract
syntactic level of kårakas and lakåras is sorted out. The correct
distribution of case endings and finite verbal endings is then accounted
for in the syntax of a Sanskrit sentence by means of operational rules.
However, kårakas do not pertain to the derivation of sentences
alone. Any verbal noun derived by a k¤t-suffix (a primary suffix) is
considered to denote either a participant in an action, in which case it is
assigned to one or the other of the six kåraka categories, or it is
considered to denote the mere activity (bhåva) expressed by the verbal
root. By way of example, the suffix LyuÈ (-ana with guâ and
4I analyse the term sådhana as sådhyate ’nena, “[something] is realised/

accomplished through it”.
5In the A!†ådhyåy¥ these are defined in the following order!: apådåna “stable
point when there is movement away”, saµpradåna “recipient!; indirect goal”,

karaâ “instrument”, adhikaraâ “locus”, karman “object!; goal”, and kart¤
“agent”. Moreover, a subcategory of agent is defined, namely hetu, the causal
agent.
140 E.G. Kahrs
presuffixal accent) is introduced to form neuter action nouns (bhåve),
by A 3.3.115 lyu† ca, and also to form nouns expressing the instrument
or the locus of the action expressed by the root, by A 3.3.117
karaˆådhikaraâyoß ca. This means that when analysing a particular
word, alternative interpretations are often possible. Accordingly, the
word udåna quoted above can be interpreted as bhåvasådhana, that is to
say, as an action noun (with LyuÈ by A 3.3.115 lyu† ca) denoting the
mere activity of uttering or pronouncing, or, alternatively, as either adh-
ikaraâsådhana, that is to say, as a noun denoting a locus, “place of
utterance”, or karaâsådhana, as a noun denoting a means of uttering
(the latter two formed with LyuÈ by A 3. 3.117 karaˆådhikaraâyoß ca
which teaches that this suffix is added also to denote the instrument and
the locus). These rules do not permit its analysis as karmasådhana, that
is, as denoting the object of the action.
Let me now return to the passage from the Saddan¥ti quoted under
the CPD entry for udånana, Sadd 382, 21–22 : ken’ atthena udånaµ :
! !
udånanatthena, kim idam udånanaµ nåma: p¥tivegasamu††håpito

!
udåhåro, “In what sense udåna? In the sense of udånana. What is this
that one calls udånana? It is an utterance (or, rather, an act of uttering)
made to arise by the impetus of joy.” The whole point of analysing
udåna as udånana is simply to make it clear that it is interpreted as
bhåvasådhana, as the act of uttering itself, and not as karman, an
utterance in the form of an object, which would be the only reasonable
interpretation of udåna in expressions such as imaµ udånaµ udåneti or
imam udånam udånayati referred to above. Incidentally, the CPD entry
for udåhåra runs : “udåhåra, m. [ts.], utterance, pronouncement ; in
! !
definitions of udåna.” The entry goes on to list references. This is

clearly running in circles without bringing out the intentions behind the
words taken from the glosses or interpretations of the indigenous
sources.
Moreover, it certainly seems reasonable here to ask what
constitutes “un pali d’intérêt linguistique”, as it is natural to form a
verbal noun such as udånana from any verb. In Yåska’s Nirukta, more
than a hundred words are analysed by a construction that involves the
ablative form of a neuter verbal noun in -ana, to be interpreted as a
nomen actionis or bhåvasådhana if one adheres to Sanskrit termin-
ology. A representative example of this type of construction is (Nir
9.26) sindhu˙ syandanåt. I suggest the following interpretation :
! !
“sindhu˙ (the river Indus, or ‘river’ in general) [is so called] on account

of the streaming (syandanam)”. The name sindhu˙ is related to the
verbal noun by an ablative construction which explains why sindhu˙
came to signify the river of that name, or any river, and the construction
is thus a natural reply to the question kasmåt “why?” The most
commonly used Sanskrit dictionaries record that the neuter verbal nouns
in -ana employed in this type of construction quite frequently are
attested only in the Nirukta, a fact that indicates that these forms are in
principle derived by Yåska himself for the technical purpose of
nirvacana analysis. This type of analysis is also met with in Buddhist
texts, for example in the Abhidharmakoßabhå!ya when it explains the
word dharma at AK 1.2 : nirvacanaµ tu svalak!aâdhåraˆåd dharma˙,
!
“as for the nirvacana : on account of the holding/possessing

!
(dhåraâm)6 [its] unique particular, [it is called] a dharma”.

In his article “Sur quelques formations sanskrites en -ti-” Louis
Renou (1951) drew attention to the fact that much as nouns in -ti are, in
principle, “regular” formations, a number of them are nevertheless
listed in the uˆådisËtras, or, more precisely, in the commentaries on the
uˆådisËtras. As Renou points out (1951, p. 1), Hemacandra’s Dhåtu-
påråyaâ, for example, presents a large number of forms in -ti that are
lacking in standard Sanskrit dictionaries. “Plusieurs sont d’une
authenticité douteuse, mais toutes méritent d’être signalées dans un
Thesaurus, 7 étant donné l’intérêt qui s’attache à l’œuvre lexico-
6The noun dhåraâ is formed from the causative stem of the root dh¤, but this
root is commonly used in the causative stem with no change of its basic
meaning.
7Deccan College was planning a Sanskrit thesaurus at the time of the
publication of Renou’s article. The article appeared in the first issue of Våk,
published by Deccan College in 1951.
142 E.G. Kahrs
graphique de Hemacandra” (1951, p. 1). However, some formations in
-ti raise questions of a similar nature as did the neuter verbal nouns in
-ana above. That forms met with in the epigraphical record should be
included in dictionaries is obvious. As examples, Renou (1951, p. 2)
mentions aµhati (variant aµhiti) in the sense of “don”, and jñåti in the
sense of “information, connaissance”, among others.
Consider now the analysis of some forms in -ti and -ana met with
in the Prasannapadå, Candrak¥rti’s commentary on Någårjuna’s MËla-
madhyamakakårikå (edited by L. de La Vallée Poussin 1903–13, p.!4,
ll.!5 –6 ) :
!
tatra niruddhir nirodha˙ k!aâbha"go nirodha ity ucyate |

utpådanam utpåda˙ åtmabhåvonmajjana[m ity artha˙] |8
ucchittir uccheda˙ prabandhavicchittir ity artha˙ |
The first line one could render : “In this respect,9 nirodha is
!
niruddhi : it is k!aâbha"ga that is called nirodha.” In other words,

!
nirodha is explained by niruddhi, a verbal noun in -ti formed from the

same root with the same preverb. The form niruddhi is not met with in
extant lexica, including BHSD. It seems therefore likely that it was
derived by Candrak¥rti for the particular purpose of interpreting nirodha
as bhåvasådhana by glossing it with a form in -ti, that is to say as
meaning “a ceasing”. Candrak¥rti goes on to say that nirodha is
k!aâbha"ga. Now, it is not easy to provide an elegant translation for
the term k!aâbha"ga, nor is it unambiguous how best to analyse the
compound. Suffice it to say that the term refers to the fact that all
phenomena are of momentary existence, hence a vigraha of the
compound could be k!aˆåd bha"ga˙, the ceasing to exist after only one
moment, or, perhaps, k!aêâ bha"ga˙ or k!aê bha"ga˙, the ceasing
to exist every moment, that is to say, “continuous instant ceasing”.
8J.W. de Jong (1978, p. 29) prefers the reading åtmabhåvonmajjanam met with
in a manuscript acquired by G. Tucci which was not available to La Vallée
Poussin.
9The passage is introduced by the words avayavårthas tu vibhajyate, “but the
meaning of the various parts is explained in detail [as follows]”.

The second line of the passage is a little more problematic, but it
brings out the point I wish to make, namely that it is possible for the
purpose of interpretation to form verbal nouns in -ti or -ana from any
verb or verbal noun. The line could be rendered : “ utpåda is utpådana,
!
[that is to say,] the emerging of a bodily form”. Here utpåda is glossed

utpådana, which, as a causative formation, is somewhat difficult to
reconcile with the fact that the verb un-majj “to emerge” is intransitive.
Turning to standard lexica, we meet with further problems. For
utpådana as a neuter noun, MW has “the act of producing or causing,
generating, begetting”. A Dictionary of Påli (M. Cone 2001) has
“uppådana, n., ~å, f. [S. utpådana], producing, generating”. Should we
emend to utpadana, that is to say, to a non-causative form? Edgerton
lists the form utpadyana in BHSD: “utpadyana (nt. ; = Pali uppajjana :
! ! !
MIndic -ana formation to utpadyate), production, origination : Gv 48.5

!
(prose), read : har$a-utpadyana-saµtånåni (see s.v. saµtåna 2)”. “Pro-

!
duction” and “origination” are unlikely synonyms, since they are based
on transitive and intransitive verbs respectively. Under utpadyati,
however, BHSD has “(2) in mg. of Skt. caus. utpådayati, produces,
causes”. Turning to the entry for saµtåna that Edgerton referred to
above, one finds that he translates har!a-utpadyana-saµtånåni “their
mental conditions productive of joy”. Here he seems to take utpadyana
as transitive.
In any case, utpådana remains problematic. A form utpadana is not
met with in lexica. To emend to utpådyana is problematic. The solution
that emerges as the most plausible is therefore to conclude that
Candrak¥rti formed utpådana directly on utpåda, again to make it clear
that he interprets utpåda as bhåvasådhana, “an emerging”. This
conclusion is supported by the continuation of the explanation : åtma-
!
bhåvonmajjana, “the emerging of a bodily form”, where unmajjana is a

neuter action noun in -ana.
Finally, ucchittir uccheda˙ prabandhavicchittir ity artha˙ could be
rendered : “uccheda is ucchitti, that is to say, the ceasing of continuity”.
!
Once again, Candrak¥rti makes it clear that he takes uccheda “cutting

144 E.G. Kahrs
off, destruction, annihilation” to be bhåvasådhana by glossing it with a
feminine verbal noun in -ti, ucchitti, “a cutting off, a destroying”. Again
the gloss is formed from the same root with the same preverb. The form
ucchitti, however, is attested in lexica, as is the further gloss °vichitti,
another formation in -ti.
Now, one may argue that forms such as niruddhi and utpådana do
not merit the distinction of being “d’intérêt linguistique”, but what they
convey about the terms they are used to interpret certainly does.
It is of course not only in commentaries on Buddhist texts that
problems and issues of interpretation arise, but also in translations of
them into other languages such as Tibetan and Chinese. In his article
“La légende de Íåntideva”,10 J.W. de Jong (1975) reproduces the
Sanskrit text of the Íåntideva legend as edited by Haraprasåd Íåstri and
the Tibetan text of the Peking edition. Section X of the text reads as
follows :
!
ßåntidevanåmå praßåntatvåt pi†akatrayaµ ßrutvå dhyåyati sma |

bhuñjåno ’pi prabhåsvaraµ supto ’pi ku†iµ gato ’pi tad eveti
bhËsukusamådhisamåpannatvåt bhËsukunåmåkhyåtaµ |
#i-ba da"-ldan-pas #i-ba’i lha #es mi"-btags | der sde-snod gsum
mñan (P. mñam)pa’i rjes-la za-ru" ñal-ru" ’chags-ru" rgyun-tu
’od-gsal bsgom-pas bhu-su-ku #es ti"-"e-’dzin la gnas-pa’i phyir
bhu-su-ku #es mi" yo"s-su grags-so ||
In a note to the word prabhåsvaram, de Jong (1975, p. 173, n. 29)
remarks, “Il y a probablement une lacune dans le texte, cf. T.” (T. = la
traduction tibétaine). The Tibetan text de Jong translates (1975, p. 176),
“En raison de sa tranquillité on lui donna le nom Íåntideva. Ayant
écouté les trois pi†aka, il méditait sur la lumière sans interruption en
mangeant, en dormant et en marchant. Persistant ainsi dans le samådhi
appelé bhusuku, il fut connu sous le nom de Bhusuku.” Somehow, this
does not quite hit the mark. The Tibetan der indicates a tatra which is
not met with in the Sanskrit version, and rgyun-tu “always” probably
means the translator has read sadaiva for tad eva. However, there is no
10Apropos Pezzali 1968.

lacuna in the text. Bhusuku is the name of one of the eighty-four
Siddhas. The Tibetan translation has not picked up on or been unable to
render the nirvacana-based reasoning met with in the Sanskrit.
In the passage above, I take tad eva to mean tad eva prabhåsvaram.
iti … : “For this reason …”. In his translation, de Jong gives the name in
!
question as Bhusuku, and, for reasons that will be clear below, I too see
no reason for retaining the long Ë of the Sanskrit text. Moreover, I do
not feel comfortable with bhusukunåmåkhyåtam as Sanskrit. I would
expect a masculine °åkhyåta˙ here : “For this reason he was named
!
Bhusuku” (bhu-su-ku #es mi" yo"s-su grags-so). On this basis, one

could then translate the Sanskrit passage as follows: “Called Íåntideva
!
because of his tranquility ( praßåntatvåt), after studying the three Pi†akas

he meditated on the radiant [mind] even when he was eating
(bhuñjåna˙), even when he was asleep (supta˙), even when in the
[latrine] hut (ku†iµ gata˙),11 [and] on that alone. So because he was
immersed in Samådhi [even] when bhu[ñjåna˙], su[pta˙], and ku[tiµ
gata˙] he was called Bhusuku.” 12
11That the word ku†i here means “latrine” is indicated by the explanation
viˆmËtrotsargårtham met with in the Caryåmelåpakaprad¥pa in a section on
bhusukucaryå that contains a similar nirvacana analysis of bhusuku as the
one discussed above. Cf. CMP 99,4–11!: anenånupËrveâ yuktågamåbhyåm
adhigamarËpåµ sarvabuddhajanan¥µ nißcitya sarvårallir vi!ayåsaktiµ ca
prahåya bhusukucaryayå cared anena krameâ. tatråyaµ krama˙ – bhu iti
bhuktvå tanmåtram anusmarati saµgamam apaharati du!karair niyamair iti
kiµcin na cintan¥yam. su iti suptvå etad vijñåya na vidyopalak!itaµ
såk!åtkurv¥ta saivåvidyåµkußåkåraµ(råµ)kitavijñånaµ punar åvartayati
prabhåsvaram eva såk!åtkaroti nirmalasvabhåvam. ku iti ku†iµ gacchet
viˆmËtrotsargårthaµ tanmåtram anubhavati sa"gam apanayati kåyavedanå-
vi!ayendriyasvabhåvaµ ca na cintayed iti. The passage as it stands requires
some textual criticism, but that need not concern us here.
12However, Alexis Sanderson informs me that in the Grub thob brgyad bcu rtsa
bzhi’i lo rgyus, which the monk Smon grub ßes rab claims at its end to have
put into Tibetan after the stories had been narrated to him by an Indian guru
called *Abhayadattaßr¥ (Mi ’jigs sbyin pa dpal), we are told that Bhusuku
(identified with Íåntideva, as in the text above) was a notoriously ignorant
monk of Nålandå. Grub thob brgyad bcu rtsa bzhi’i chos skor (New Delhi!:
Chophel Legdan, 1973), p. 171, ll. 4– 5 !: mi" du ya" bhu su ku zhes grags la
146 E.G. Kahrs
Similar issues arise from sections XII and XIII. Section XII and the
first part of XIII run as follows :!
XII. pËrvak¤taµ sËtrasamuccayaµ ßik!åsamuccayaµ bodhicaryå-

vatåråkhyaµ granthatrayam ast¥ti cetasi k¤två siµhåsanagata˙
pråha kim år!aµ pa†håmi arthår!aµ vå ||
XII. mdo-sde kun-las btus-pa da" | bslab-pa kun-las btus-pa da" | bya"-
chub spyod-pa la ’jug-pa #es g#u"-gsum bdag-gis byas yod-do || de-
la spyod-pa la ’jug-pa gdon-par ’os #es bsams-nas gsu"s-te dra"-
so"-gis gsu"s-pa ’am | de’i-rjes las byu"-ba ga" gdon |
XIII. tatra ¤!i˙ paramårthajñånavån ¤!a gatåv ity atra auˆådika˙ kvi˙
¤!iˆå jinena proktam år!aµ nanu prajñåpåramitådau subhËtyådi-
deßitaµ katham år!aµ ity atrocyate yuvaråjåryamaitreyeâ ||
XIII. don-dam rtog-pa ni dra"-so"-"o || des mdzad-pa gsu"-rab-bo || de-la
brten-nas g#an-gyi bya-ba de rjes-las byu"-ba’o || ’phags-pa byams-
pas de gsu"s-pa |
Having compared the two versions of the text, de Jong (1975,

p.!177) states, “Il est évident que les textes sanskrit et tibétain doivent
remontrer au même texte original. Les différences entre les deux
versions dans les sections XIII et XIV sont dûes à des additions. Le
texte sanskrit a ajouté une phrase sur l’étymologie de ¤$i (¤$¥ gatåv ity
atra auˆådika˙ kin) et une référence à l’enseignement de SubhËti : nanu !
prajñåpåramitådau subhËtyådideßitaµ katham år$am ‘Comment ce qui a

été enseigné par SubhËti dans la Prajñåpåramitå, etc. peut-il être år!a?’
Le passage précédent explique qu’år!a est ce qui est dit par le ¤!i, i.e. le
jina.” However, these “additions” are integral to the two versions of the
text themselves. The Sanskrit version of section XIII from nanu through
katham år!am has been translated by de Jong above. As for the first
part, de Jong (1 9 7 5 , p. 1 7 4 , n. 28), acknowledging a note from
Professor Y. Ojihara, points out that one should read ¤!¥ gatåv ity atra
auˆådika˙ kin.
The Daßapådyuˆådiv¤tti (DPU) at 1.48 igupadhåt kit states ¤!¥
gatau tau° / ¤!at¥ti ¤!i˙ muni˙ / karttå, “the sixth-class [verbal root] ¤!
de ni za nyal chags gsum pa zhes bya’o. Cf. Bengali bhõs “fool”!; Kumaun¥
bhus “foolish, wild, uncivilized, rude” (CDIAL §!9545).
[occurs] when [the sense of] gati ‘going’ [is to be denoted] ; 13 ‘he !
moves (¤!ati)’, hence [he is called] ¤!i, [that is to say,] a sage, [in the
sense of the] agent [of the act of moving].”14 One is now in a position
to translate the first part of section XIII : “In this respect, a ¤!i is
!
someone who possesses knowledge (jñåna) of the supreme meaning,

[through the addition of] the uˆådi suffix kit to [the verbal root] ¤!
[which occurs] when [the sense of] gati ‘going’ [is to be denoted], [and]
år!am [is formed according to the analysis] : promulgated by a ¤!i, [that
!
is to say,] the Jina.”

The Tibetan version of section XIII begins : don-dam rtog-pa ni
!
dra"-so"-"o, “a ¤!i is someone who has knowledge of the supreme

meaning”. A ¤!i is thus said to possess knowledge of the supreme
meaning in both the Sanskrit and Tibetan versions. The explanation of
the term ¤!i as a k¤t or primary derivation from the root ¤! in the sense
of gati “a going” clearly draws on the traditional hermeneutic rule sarve
gatyarthå jñånårthå˙ which states that all words that mean “move” also
mean “know”. Just as the term år!a is explained in the Sanskrit version
of section XIII by the phrase ¤!iˆå jinena proktam år!am, it is explained
in the Tibetan version of section XII: dra"-so"-gis gsu"s-pa, “that
!
which has been proclaimed by a ¤!i”. At work here is a rule from

Påîni’s A!†ådhyåy¥. A 4.1.83 pråg d¥vyato ’ˆ teaches that the taddhita
suffix aÔ (-á with v¤ddhi strengthening of the first vowel) is added
under meaning conditions given in rules up to rule A 4.4.2 tena d¥vyati
khanati jayati jitam. That is to say, A 4.1.83 is a general rule (utsarga)
which teaches the addition of aÔ unless it is blocked by some other
suffix under conditions specified by a special rule (apavåda). A 4.3.101
tena proktam then teaches the addition of aÔ in the sense “promulgated
by him”, hence år!a in the sense of “promulgated by a ¤!i”.
13The Påîn¥ya dhåtupå†ha 6.7.

14Incidentally, the Tibetan rendering of ¤!i as dra"-so", or, more commonly,
dra"-sro" (dra"-po “straight”, so" “became, turned”!; sro"-pa “to make
straight, straighten [the body]”) is based on the Sanskrit nirvacana ¤ju˙ ßete
“he sits straight”.
148 E.G. Kahrs
Let me finally turn to the form arthår!a met with in the Sanskrit
version of section XII, which ends : pråha kim år!aµ pa†håmi
!
arthår!aµ vå, “He says, ‘Shall I recite år!am or arthår!am?’ ” Section

XIII quotes a verse from the Ratnagotravibhåga, which de Jong (1975,
p. 178) points out “ne fait pas de distinction entre år!a et arthår!a (Tib.
de-rjes las byu#-ba), mais dit que tout ce qui est dit en conformité avec
certaines conditions est år!am iva”. Pointing out that Edgerton (BHSD)
has identified the reading år!a in Wogihara’s edition of the Bodhi-
sattvabhËmi as a corruption of år!abha, de Jong goes on to say (1975,
p. 178), “Le mot år!a se rencontre dans le MahåyånasËtrålaµkåra
(XVIII.31) : år!aß ca deßanådharmo, mais le commentaire ne l’explique
!
pas. Il se peut très bien que le mot arthår!a soit corrumpu mais la
version tibétaine qui en donne une traduction libre ne permet pas de le
corriger. On ne retrouve la distinction entre år!a et arthår!a ni chez Bu-
ston ni chez Tåranåtha. … Pour conclure cette discussion signalons
encore que dans section XV, le texte sanskrit a arthår!am mais la
version tibétaine g#an-pa = anyad.”
The latter part of the Tibetan version of section XII runs as follows :
!
de-la spyod-pa la ’jug-pa gdon-par ’os #es bsams-nas gsu"s-te dra"-

so"-gis gsu"s-pa ’am | de’i-rjes las byu"-ba ga" gdon, “He says, ‘Shall
I recite that which has been proclaimed by a ¤!i or that which has come
after that [which has been proclaimed by a ¤!i]?’ ” The Tibetan trans-
lator has clearly had the reading anvår!am rather than arthår!am. That
is to say, anvår!am according to the analysis dra"-so"-gis gsu"s-pa’i-
rjes las byu"-ba, “that which has come after that which has been
proclaimed by a ¤!i”, rjes las byu"-ba rendering anugata, which is an
attested interpretation of anu (e.g. Sadd 883,14 : anusaddo anugate ; or,
! !
Sadd 883, 18 : tattha anugate anveti). The proposal of anvår!am for what
!
the Tibetan translates is appealing also because it provides a ready

explanation of the corruption through similarity of the conjuncts nvå
and rthå in post-Gupta scripts which indicate pre-consonantal r as a
horizontal stroke below the head-line added to the left side of the
following letter. This, of course, does not necessarily make anvår!am
the correct reading, and I am held back from accepting that it was by the
absence of citations of other occurrences of the word. The passage of
the Ratnagotravibhåga referred to above distinguishes between år!am,
what is not år!am (vipar¥tam anyathå ) and what is år!a-like (år!am iva)
and therefore acceptable teaching. An example of the last would be the
subhËtyådideßitam. To accept anvår!am in that sense one would need
examples of other anu-words with this of the same kind. Otherwise I
would be inclined to think that anvår!am might be a corruption of
anår!am. The latter is congruent with the Tibetan g#an (anyad =
anår!am) of XV.
Through the instances presented above, I have tried to highlight that
some of the specific techniques and conventions applied by indigenous
commentators and translators often consist of linguistic and hermen-
eutical devices rooted in the Sanskrit traditions of vyåkaraâ and
nirvacanaßåstra, and that a knowledge of these disciplines can be of
importance for a full understanding of Buddhist texts. These were the
disciplines Buddhist commentators and translators were versed in,
disciplines we might in the end simply call philology.
E.G. Kahrs
University of Cambridge
150 E.G. Kahrs
ABBREVIATIONS
A Påîni, A!†ådhyåy¥. Reference is to O. Böhtlingk, ed., Pâîni’s
Grammatik, herausgegeben, übersetzt, erläutert und mit verschiedene
Indices versehen. Leipzig : Haessel, 1887
!
AK Vasubandhu, Abhidharmakoßa(-bhå!ya), Vol. I. Edited with the

Sphu†årthå commentary of Yaßomitra by Swami Dwarikadas Shastri.
Bauddha Bharati Series 5. Varanasi : Bauddha Bharati, 1970
!
CDIAL R.L. Turner , A Comparative Dictionary of the Indo-Aryan Languages.

London : Oxford University Press, 1962–1969
!
CMP Ócårya Óryadeva, Caryåmelåpakaprad¥pam. Edited by Janardan

Shastri Pandey. Rare Buddhist Texts Series 22. Sarnath, Varanasi : !
Rare Buddhist Texts Research Project, Central Institute of Higher

Tibetan Studies, 2000
DPU UˆådisËtras in the daßapåd¥ recension. Reference is to Yudhi$†hira
M¥måµsaka, ed., Daßapådyuˆådiv¤tti. Princess of Wales Sarasvati
Bhavana Texts Series 81, Benares: Government Sanskrit College,
!
1943
MW Monier Monier-Williams, Sanskrit–English Dictionary. Oxford, 1899
Nir Yåska, Nirukta. Reference is to L. Sarup, ed., The Nighaˆ†u and the
Nirukta : The Oldest Indian Treatise on Etymology, Philology, and
!
Semantics. Sanskrit Text. Lahore : University of the Panjab, 1927

!
BIBLIOGRAPHY
Caillat, C., 1971. “Note préliminaire” in Y. Ojihara, “Un chapitre de la Sadda-
n¥ti comparé aux données pâînénnes”, JA 259, pp. 83–97
Cone, M., 2001. A Dictionary of Påli : Part I, Oxford: Pali Text Society
! !
von Hinüber, O., 1978. “On the tradition of Påli texts in India, Ceylon and
Burma”. In Buddhism in Ceylon and Studies on Religious Syncretism in
Buddhist Countries, edited by H. Bechert. Abhandlungen der Akademie der
Wissenschaften in Göttingen (Philologisch-historische Klasse, III.108),
pp.!55– 57
de Jong, J.W., 1975. “La légende de Íåntideva”, IIJ 16, pp. 161– 82
——— 1978. “Textcritical notes on the Prasannapadå”, IIJ 20, pp. 25–59
Kahrs, E.G., 1992. “Exploring the Saddan¥ti”, JPTS XVII, 1–212
La Vallée Poussin, L. de, 1903– 13. MËlamadhyamakakårikås (Mådhyamika-
sËtras) de Någårjuna avec la Prasannapadå commentaire de Candrak¥rti
(Bibliotheca Buddhica IV), Saint Petersburg
Norman, K.R., 1983. Påli Literature. A History of Indian Literature, Vol. VII,
fasc. 2, Wiesbaden : Harrassowitz
!
—— 1997. A Philological Approach to Buddhism. The BukkyØ DendØ KyØkai

Lectures 1994. The Buddhist Forum, Vol. V. London: School of Oriental
!
and African Studies!; reprinted, Pali Text Society, 2006

Pezzali, Amalia, 1968. Íåntideva, mystique bouddhiste des VIIe et VIIIe siècles.
Firenze : Vallecchi Editore
!
Renou, L., 1951. “Sur quelques formations sanskrites en -ti-”, Våk 1, 1– 4

Smith, H., ed., 1928–30. Saddan¥ti : la grammaire palie d’Aggavaµsa: I!Pada-
! !
målå, II Dhåtumålå, III Suttamålå, Lund: Gleerup!; reprinted, Pali Text

!
Society, 2001
A Note on micchådi††hi in Mahåvaµsa 25.110
In his pioneering work The Pali Literature of Ceylon (1928),

Professor G.P. Malalasekera dwells at length on the great Buddhist king
Du††hagåmaî Abhaya (101–77 B. C .E.) whom he hails as “the hero of
the epic Mahåvaµsa”. Based on the account of this king in Chapters 24
and 25 of the Mahåvaµsa, he details the career of this king, his
triumphant victory over the Dami¬a King E¬åra and his manifold deeds
of piety including the beginning of the construction of the MahåthËpa.
Malalasekera draws our attention to the magnanimity of the victorious
king Du††hagåmaî for his fallen adversary : The king constructed a
!
cetiya over the ashes of his dead enemy, and decreed that “no man,
prince or peasant, should pass the spot … riding in palanquin or litter or
with beating of drums.” Malalasekera says further that after his
coronation, the “king’s outlook on life had changed, the great and
glorious success for which he had lived and dreamed gave him no real
joy. He thought of the thousands of human lives on whom suffering had
been wrought to encompass this end, and he was filled with poignant
grief … he determined to start a new chapter in his life” (p. 35). He
devoted himself to the task of erecting several religious edifices.
What is conspicuously missing in this account is a major narrative
from Mahåvaµsa, Chapter 25, that tells us about an episode of the
king’s deep remorse over the death of a large number of warriors in his
victory. This particular incident raises a most problematic issue
regarding the way Theravådin Buddhists viewed death on a battlefield.
The passage in question, in seven verses, is given below from Geiger’s
edition (Mhv) and his translation (assisted by Mabel Bode ).1
103. sayito sirisaµpattiµ mahatiµ api pekkhiya
kataµ akkhohiˆâghåtaµ saranto na sukhaµ labhi.
1Geiger 1912.
154 P.S. Jaini
He, looking back upon his glorious victory, good though it was, knew no
joy, remembering that thereby was wrought the destruction of millions [of
beings].
104. Piya!gud¥pe arahanto ñatvå taµ tassa takkitaµ
påhesuµ a††ha arahante taµ assåsetum issaraµ.…
When the arahants of Piya"gud¥pa knew his thought, they sent eight
arahants to comfort the king.…
108. “kathaµ nu bhante assåso mama hessati, yena me
akkhohiˆ¥mahåsenåghåto kåråpito!?” iti.
Then the king said to them again “How shall there be any comfort for me,
O Venerable Sirs, since by me was caused the slaughter of a great host
numbering millions!?”
109. “Saggamaggantaråyo ca natthi te tena kammunå,
diya""hamanujå v’ ettha ghåtitå manujådhipa,
“From this deed arises no hindrance in the way to heaven. Only one and a
half human beings have been slain here by thee, O Lord of Men.
110. “saraêsu †hito eko, pañcas¥le pi cåparo,
micchådi††h¥ ca duss¥lå seså pasusamå matå.
“The one had come unto the [three] refuges, the other had taken on
himself the five precepts. Unbelievers and men of evil life were the rest,
not more to be esteemed than beasts.
111. “jotayissasi c’ eva tvaµ bahudhå buddhasåsanaµ,
manovilekhaµ tasmå tvaµ vinodaya narissara.”
“But as for thee, thou wilt bring glory to the doctrine of the Buddha in
manifold ways!; therefore cast away care from thy heart, O Ruler of
Men.”
112. iti vutto mahåråjå tehi assåsaµ ågato.
Thus exhorted by them, the great king took comfort.
The king’s remorse is quite in keeping with the Buddhist teachings.
One is reminded of the patricidal king Ajåtasattu’s visit to the Buddha
as described in the Såmaññaphalasutta of the D¥gha-nikåya.2 There the
2taggha tvaµ mahåråja, accayo accagamå …yaµ tvaµ pitaraµ … j¥vitå

voropesi. yato ca kho tvaµ … accayaµ accayato disvå yathådhammaµ
pa†ikarosi, taµ te mayaµ pa†igaˆhåma. vuddhi h’ eså ariyassa vinaye …
åyatiµ saµvaraµ åpajjat¥ ti. Såmaññaphalasutta, D I 100.
A Note on micchådi††hi in Mahåvaµsa 25.110 155
king confesses his guilt over the killing of his father. The Buddha does
not absolve him of his crime, but accepts his confession saying “Verily
O King it was sin that overcame you while acting thus. But in as much
as you look upon it as sin, and confess it according to what is right, we
accept your confession as to that. For that, O King, is custom in the
discipline of the Noble Ones, that whosoever looks upon his fault as a
fault, and rightfully confesses it, shall attain to self-restraint in the
future.”3 Another historical case is that of the Mauryan King Aßoka
who issued his famous Rock Edict after the subjugation of the people of
Kali"ga : “The Kali"ga country was conquered by King Piyadasi
!
Devånaµpiya, when he had been consecrated eight years. One hundred

and fifty thousand were carried away as captives and one hundred
thousand slain and many times that number died. … Devånaµpiya the
conqueror of Kali"ga has remorse now, because of the thought that the
conquest is no conquest, for there was killing. … That is keenly felt
with profound sorrow and regret. … Now even the loss of a hundredth
or even a thousandth part of all lives that were killed or died or carried
away captives is considered deplorable by Devånaµpiya.” 4 Aßoka’s
inscriptions do not show him seeking either consolation or absolution
from any religious establishment, nor does the Mahåvaµsa allude to his
war in the conquest of Kali"ga.
What is extraordinary about the account in the Mahåvaµsa is the
uncommon arrival of eight arahants representing the Buddhist sa"gha to
console Du††hagåmaî Abhaya and to assure him safe passage to
heaven.5 It is much to the credit of the king that he should anticipate
3DB, Vol. I, pp. 94–95.

4Murit and Aiyangar 1951, Rock Edict XII!: a†havasåbhisitayå Devånaµpiya#a
Piyadasine låjine Kaligyå vijitå | diya"hamåte påna#ata#ahaße ye taphå
apavu"he, ßata#aha#a måte tata hate, bahutåvaµtake vå ma†e | … .#e athi
anu#aye Devånaµpiya#å vijinitu Kaligyåni, avijitaµ hi vijinamane e tatå
vadha vå malane vå apavahe vå jana#å | #e bå"ha vedaniyamute gulumute cå
devånaµpiya#å |
5It may be noted that D¥p XIX, p. 101, is content in merely stating that the king
was reborn in the Tusita heaven!:
155
156 P.S. Jaini
severe obstruction to his rebirth in heaven (saggamaggantaråyo) as a
consequence of his act of warfare in which so many warriors perished
on the battlefield. The response of the arahants is truly astounding. They
not only say that there is no obstruction to the king’s rebirth in heaven
but also seek to legitimize their verdict by observing that out of the
“million lives” only one and a half men have been truly slain : one who
!
had taken refuge in the three saraâs (#) ; and another one who
!
additionally took the five precepts (1). The arahants declare that the
remaining dead were micchådi††his and duss¥las, and thus equal to
animals (pasusamå). They add further that the king will (because of this
victory) glorify the Buddhist faith and so he should overcome his
remorse.
Although Malalasekera saw fit to ignore this episode in his earlier
book, in the Dictionary of Påli Proper Names (1960), he allows a single
sentence : “From now onwards [after his final victory] consoled by the
!
arahants of Piya"gud¥pa, who absolved him from blame (italics added)

for the slaughter of his enemies …”. In contrast however, another
Sinhalese Buddhist scholar, the late Venerable Walpola Rahula in his
History of Buddhism in Ceylon (1956), duly notes this particular
episode. He reproduces the gist of the Mahåvaµsa and notes further that
it was the beginning of Buddhist nationalism. In observing the career of
King Du††hagåmaî Rahula says : “The entire Sinhalese race was united
!
under the banner of the young Gåmaî. This was the beginning of
nationalism amongst the Sinhalese. It was a new race with healthy
young blood organized under the new order of Buddhism. A kind of
religio-nationalism, which almost amounted to fanaticism, roused the
whole Sinhalese people. A non-Buddhist was not regarded as a human
katapuñño mahåpañño Abhayo Du††hagåmaî

kåyassa bhedå sappañño tusitaµ kåyaµ upågami.
This suggests the possibility that the authors of the Mahåvaµsa introduced
the episode of the king’s remorse. The Extended Mahåvaµsa makes further
elaboration, as below (n. 17).
being. Evidently, all Sinhalese without exception were Buddhists.” 6
It would not be wrong to assume that both Malalasekera and Rahula
were only following the lead given by the Vaµsatthappakåsin¥, a †¥kå on
the Mahåvaµsa : !
tena kammunå matå means by the act of your slaying a “million”. The
words diya""hamanujå v’ ettha means amongst these “millions”, only one
and a half men have been slain by you. seså pasusamå matå means the
remainder were truly not men because they were devoid of the virtues of a
human being!: they were devoid of proper views, and given to bad conduct.
And therefore they said they are pasusamå, equal to animals. Taking the
refuges and the five precepts are the virtues that make a human being, and
therefore the text says that one person had established himself in the refuges
and the other had the five precepts. For this reason, [O King,] you are free
from any obstruction in the way to heaven, and in the future you will glorify
the teaching of the Buddha.7
The arahants, it should be noted, only assured (assåsito) the king,

but the authors of the Mahåvaµsa were composing a chronicle of the
island and would be expected to glorify the deeds of a great king, even
to the extent of trying to “absolve” him of the karmic consequences of a
bloody war. But what is truly puzzling is the fact that the Theravådins
of La"kå over the centuries should accept the validity of the alleged
words of the arahants as understood by the author of the Mahåvaµsa-
†¥kå. This calls for a search of the canonical expositions on micchådi††hi,
given by the Buddha in the sermons specifically addressing the issues of
heaven and warfare. If this term is understood correctly, the words of
the arahants would appear to be credible and the statements of the
Mahåvaµsa to be consistent with the teachings of the Buddha.
6Rahula 1956, p. 79.

7tena kammunå ti tena tayå katena akkhohiˆ¥ghåtakammena$ ; … diya-
""hamanujå v’ etthå ti ettha akkhohiˆ¥senåya diya""h’ eva manusså tayå
ghåtitå$; seså pasusamå matå ti avaseså di††hivippanna††hena ca duss¥la††hena
ca naradhammavirahita††hena ca manusså nåma nåhun ti, sabbe pasusamå
matå ti avocun ti attho. saraâs¥låni hi manussakårakadhammåni, tena vuttaµ
saraêsu …cåparo ti. manovilekhaµ tasmå tvaµ ti yasmå tvaµ sagga-
maggantaråyavirahito va … iti vuttaµ hoti. Mhv-† II 491–92.
157
158 P.S. Jaini
Geiger and Bode’s translation of micchådi††hi as “unbelievers”, i.e.
non-Buddhists, is permissible since the context does convey that
meaning, intended or not, to a casual reader. Rahula’s translation as
“wrong-believers” is too general ; it does not identify a particular wrong
!
belief. Micchådi††hi and sammådi††hi are two oft-recurring technical

terms found in various places in the canon. The Mahåcattår¥saka-sutta
(M III 71–78) and the Apaˆâka-sutta (M I 400–13) of the Majjhima-
nikåya appear to be most relevant in this context.
In the first the Buddha defines the two di††his in the following words
(Lord Chalmers’ translation) :8 !
What are the wrong views (micchådi††hi)!? — They are views that —
there is no such thing as alms or sacrifice or oblations!; that there is no such
thing as the fruit and harvest of deeds good and bad!; that there are no such
things as this world or the next!; that there are no such things as either
parents or a spontaneous generation elsewhere!; that there are no such things
as recluses and brahmins who tread the right path and walk aright, who
have, of and by themselves, comprehended and realized this and other
worlds and make it all known to others.9
And what are the right views (sammådi††hi)!? — they are twofold. On
the one hand there are right views which are accompanied by Cankers
(såsavå), are mixed up with good works ( puññabhågiyå), and lead to
attachments. On the other hand there are Right Views which are Noble
(ariyå), freed from Cankers (anåsavå), transcending mundane things and
included in the Path.10
Those right views which are accompanied by Cankers … lead to attach-
8Chalmers 1927, Vol. II, pp. 194–95.

9Mahåcattår¥sakasuttta, M III 71f. !: katamå ca bhikkhave micchådi††hi$? natthi
dinnaµ, natthi yi††haµ, natthi hutaµ, natthi sukatadukka†ånaµ kammånaµ
phalaµ vipåko, natthi ayaµ loko, natthi paro loko, natthi måtå, natthi pitå,
natthi sattå opapåtikå, natthi loke samaâbråhmaˆå sammaggatå
sammåpa†ipannå, ye imañ ca lokaµ parañ ca lokaµ sayaµ abhiññå
sacchikatvå pavedent¥ ti.
10M III 72. katamå ca bhikkhave sammådi††hi$? sammådi††hiµ påhaµ,
bhikkhave, dvayaµ vadåmi. atthi bhikkhave sammådi††hi såsavå

puññabhågiyå upadhivepakkå !; atthi bhikkhave ariyå anåsavå lokuttarå
magga!gå.
ments, recognize that there are such things as alms and sacrifice and obla-
tions!; that there is indeed such a thing as the fruit and harvest of deeds good
and bad!; that there are really such things as this world and the next!; that
there are really such things as parents and spontaneous generation else-
where!; and that there are really such things as recluses and brahmins who
tread the right path and walk aright, who have, of and by themselves,
comprehended and realized this and other worlds and make it all known to
others.11
In the Apaˆâka-sutta, as in our Mahåvaµsa passage, the words
duss¥la and micchådi††hi appear together: duss¥lo purisapuggalo
!
micchådi††hi natthikavådo. The Apaˆâka-sutta further elaborates : “The !
next world (i.e. life after death) truly exists but this person denies it.
That constitutes his micchådi††hi.”12 As is well known this is a doctrine
11M III 72. By this rather wide definition anyone believing in a life after death
(and so forth) can be called a sammådi††hi$; the term is no longer restricted
only to a lay follower of the Buddha. The a††hakathå on the Sammådi††hi-
sutta of the Majjhima-nikåya (M I 46–55) anticipates such a possibility and
hence makes the following comments!:
The sammådi††hi is twofold, mundane (lokiyå) and supermundane
(lokuttarå). Of these the former consists of paññå, brought about by knowl-
edge of the doctrine of karma, and knowledge that conforms to the Four
Noble Truths.…
Human beings are also of three kinds !: an ordinary person, the disciple, and
the nondisciple. Of these the ordinary person is of two kinds!: The outsider
(båhiraka) and the follower of the Buddha (såsanika). The båhiraka is a
sammådi††hi by virtue of his view that affirms the doctrine of karma, but he
does not have faith in the Four Noble Truths, and he holds the view there is
an eternal self (attadi††hi), whereas the såsanika is sammådi††thi by having the
paññå of both kinds!:
så cåyaµ sammådi††hi duvidhå hoti–lokiyå lokuttarå ti. t a t t h a
kammassakatåñåâµ saccånulomikaññåâµ ca lokiyå sammådi††hi,
sa!khepato vå sabbå pi såsavå paññå. ariyamaggaphalasampayuttå paññå
lokuttarå sammådi††hi. puggalo pana tividho hoti!: puthujjano sekkho
asekkho ca. tattha puthujjano duvidho hoti$: båhirako såsaniko ca. tattha
båhirako kammavåd¥ kammassakatådi††hiyå sammådi††hi hoti, no
saccånulomikåya attadi††hiparåmåsakattå. såsaniko dv¥hi pi (Ps I 196).
12santaµ yeva kho pana paraµ lokaµ “natthi paro loko” ti ’ssa di††hi hoti!;
såssa hoti micchådi††hi. … ayaµ … purisapuggalo di††he va dhamme

viññËnaµ gårayho!: “duss¥lo purisapuggalo micchådi††hi natthikavådo” ti.…
159
160 P.S. Jaini
of uccheda (“annihilation”) originally attributed to a titthiya named
Ajita Kesakambali in the Såmaññaphalasutta of the D¥gha-nikåya.13
This micchådi††hi is truly the antithesis of the (såsavå or the first
variety of) sammådi††hi. A Buddhist is said to be a sammådi††hi because
he affirms the existence of the aforementioned ten items that are denied
by the “nihilist” (natthikavådo) or the “annihilationist”. Evidently such
a meaning of micchådi††hi is not appropriate to the same word in the
passage under discussion. Those who perished in the war were warriors
and it would be inconceivable that they would not seek heaven or some
such reward for their heroism on the battlefield. Fortunately there is a
whole section in the Saµyutta-nikåya, ironically called the Gåmaî-
saµyutta, which gives us a detailed description of the beliefs held by
the warriors during the Buddha’s time. It contains a remarkable
dialogue between a certain Yodhåj¥va (Fighting-man) and the Buddha,
which provides us with a different concept of micchådi††hi, one that is
not covered by the earlier usage. This unique dialogue explains both the
volitional aspect of the deed of killing (vadhakacetanå/duppaîhitaµ)
as well as the particular wrong view of the warrior concerning his death
and rebirth in heaven (F.L. Woodward’s translation of S IV 308f.) :14 !
Then Fighting-man (Yodhåj¥va),15 the trainer, came to see the Exalted

One.… As he sat at one side, Fighting-man, the trainer, said to the Exalted
One!:
“I have heard, Lord, this traditional saying of teachers of old who were
evaµ assåyaµ apaˆâko dhammo dussamatto samådiˆô ekaµsaµ pharitvå

ti††hati, riñcati kusalaµ †hånaµ (Apaˆâkasutta, M I 402–403).
13“Thus, Lord, did Ajita of the garment of hair (Kesakambali) … expound his
theory of annihilation.” The translators call this “the view of a typical

sophist” (DB I 73, n.1).
14KS IV 216–17.
15Bhikkhu Bodhi (CD II, p. 1334) translates Yodhåj¥va Gåmaî as “the head-
man Yodhåj¥va the Mercenary” and gives the following note (p. 1449, n.
339) !: “Spk explains the name as meaning ‘one who earns his living by
warfare (yuddhena j¥vikaµ kappento)!; this name, too, was assigned by the
redactors of the dhamma’. I take the occupation to be that of a mercenary or
professional soldier.”
fighting men!: ‘A fighting man who in battle exerts himself, puts forth
effort, thus exerting himself and putting forth effort, is tortured and put an
end to by others. Then, when body breaks up, after death he is reborn in the
company of the Devas of Passionate Delight.’ What says the Exalted One of
this!?”
“Enough, trainer!! Let be. Ask me not this question”.… Nevertheless I
will expound it to you.
“In the case of a fighting-man who in battle exerts himself, puts forth
effort, he must previously have had this low, mean, perverse idea!: ‘Let
those beings be tortured, be bound, be destroyed, be exterminated, so that
they may be thought never to have existed.’ Then, so exerting himself, so
putting forth effort, other men torture him and make an end of him. When
the body breaks up, after death he is reborn in the Purgatory of Quarrels (a
part of the Av¥ci niraya).
“Now if his view was this!: ‘A fighting-man who exerts himself, puts
forth effort in battle, thus exerting himself, thus putting forth effort, is
tormented and made an end of by others. When body breaks up, after death
he is reborn in the company of the Devas of Passionate Delight,’ — then I
say that view of his is perverted (micchådi††hi). Now, trainer, I declare that
for one who is guilty of perverted view one of two paths is open, either
purgatory or rebirth as an animal (nirayaµ vå tiracchånayoniµ vå).”16
16atha kho Yodhåj¥vo Gåmaî … etad avoca!: sutaµ me bhante, pubbakånaµ

åcar¥yapåcariyånaµ yodhåj¥vånaµ bhåsamånånaµ, yo so yodhåj¥vo sa!gåme
ussahati våyamati, taµ enam ussahantaµ våyamantaµ pare hananti
pariyådåpenti, so kåyassa bhedå paraµ maraˆå sarañjitånaµ devånaµ
sahavyataµ upapajjat¥ ti. … idha Bhagavå kiµ åhå ti!?
alaµ Gåmaî ti††hat’ etam, må mam etaµ pucch¥ ti… api ca tyåhaµ
vyåkarissåmi. yo so gåmaî yodhåj¥vo sa!gåme ussahati våyamati, tassa taµ
cittaµ pubbe h¥naµ duggataµ duppaîhitaµ!: ime sattå haññantu vå bajjhantu
vå ucchijjantu vå vinassantu vå må ahesuµ iti vå ti. taµ enam ussahantaµ
våyamantam pare hananti pariyådåpenti, so kåyassa bhedå paraµ maraˆå
sarañjitå nåma nirayå tatth’ upapajjati.
sace kho panassa evaµ di††hi hoti!: yo so yodhåj¥vo sa!gåme ussahati
våyamati tam enam ussahantam våyamantam pare hananti pariyådåpenti, so
kåyassa bhedå paraµ maraˆå sarañjitånam devånaµ sahavyatam upapajjat¥
ti, såssa hoti micchådi††hi.
micchådi††hikassa kho panåhaµ Gåmaî purisapuggalassa dvinnaµ
gat¥naµ aññataraµ gatiµ vadåmi, nirayaµ vå tiracchånayoniµ vå ti.
161
162 P.S. Jaini
In view of the Buddha’s emphatic words regarding the fate of those
who perish on the battlefield while entertaining such a view, there
should be no hesitation now in applying this definition of micchådi††hi
to the same word appearing in Mahåvaµsa (25, 110), instead of the
traditional canonical meaning of that term as natthikavåda or
ucchedavåda.
The word pasusamå (“equal to animals”) in the Mahåvaµsa is
undoubtedly used in a figurative manner. Even so, the declaration in the
Yodhåj¥va-sutta that such beings are destined to be reborn in niraya or
in the animal world lends support to the possibility that the figurative
expression was a kind of a prognostication of their destiny. The
Extended Mahåvaµsa (25, 256) makes it explicitly clear that the king’s
remorse was caused by a horrible sight of the countless dead Dami¬as : !
addakkhi … asaµkhiyånaµ maraâµ Daµi¬ånam. 17 While it is clear

that the Dami¬as are not Buddhists, the texts do not furnish us with any
information on their faith. Since they were coming from South India,
they may be considered as followers of some form of Íaivism or
Vai$âvism, similar to the one practised probably by the yodhåj¥vas in
the passage above. They may be open to the teachings such as given in
the Bhagavadg¥tå II, 37, where Lord K¤$â promises the warrior Arjuna
evaµ vutte Yodhåj¥v¥ Gåmaî parodi, assËni pavattesi. … nåham … api

cåhaµ bhante pubbakehi åcariyapåcariyehi yodhåj¥vehi d¥gharattaµ nikato
vañcito paluddo … devånaµ sahavyatam upapajjat¥ ti.
S IV 308–309
Similar answers are given with regard to the hatthårohå and assårohå,
those fighting while seated on elephants or riding horses (S IV 310–11).
17Extended Mahåvaµsa, 25, 256 –59!:
tassa hetuµ apekkhanto addakkhi manujådhipo

asaµkhiyånaµ maraâµ Dami¬ånaµ tadantare$:
vasuµdharåyaµ katvåna s¥saµ sabbadisåsu pi
akkh¥ni nikkhamitvåna gattåni uddhamåtakå,
kåkaka!kagijjhasoâsigålåd¥hi khåditå
hatthapåda!gapacca!gå chavånaµ chiddamånakaµ,
sattehi khådayantehi okiriµsu visuµ visuµ
saddå nesaµ sattånaµ mahantå bheravå ahu.
that if he is slain in battle he will attain heaven : hato vå pråpsyasi
!
svargaµ, jitvå vå bhok#yase mah¥m. The Theravådins of La"kå might

well have believed that the Dami¬as who perished in the war did aspire
to be reborn in heaven, and were for the most part born in the animal
world. Understood in this manner the arahants’ words can be said to be
consistent with the Buddha’s teachings on heaven and warfare as found
in the Yodhåj¥va-sutta.
The above interpretation, admittedly a little farfetched, is sup-
ported by a most remarkable corroboration from the Prakrit canonical
texts of the ancient samaâs called Nigaˆ†has (also known as Jainas),
datable to the same period as the Påli Saµyutta-nikåya. As is well
known from the Såmaññaphala-sutta, their teacher, a titthiya, Nigaˆ†ha
Nå†aputta (T¥rtha"kara Jñåt¤putra Mahåv¥ra), was a contemporary of
Gautama the Buddha and both flourished in Magadha. While the
Buddhist texts state that Ajåtasattu, the king of Magadha, embraced
Buddhism, the Jainas claim that his father Íreîka Bimbisåra was a
devotee of Mahåv¥ra. Both came from the warrior caste and had
witnessed many a battle raging in Magadha. Death on the battlefield
was considered honourable and questions were being raised regarding
the validity of the claim that such death was rewarded by rebirth in
heaven. It is not surprising therefore that the questions asked of the
Buddha by Yodhåj¥va and others find their close parallels in the Jaina
canon. The Book VII of the canonical text Viyåhapannatti (Vyåkhyå-
prajñapti) contains narratives about wars that were waged by the
Magadhan King KËîya (Ajåtasattu) in his fight against eighteen tribal
chiefs (gaâ-råya), that is to say, the nine Malla and the nine Lecchavi
kings of Kås¥ and Kosala, in which “millions” are said to have died. The
following dialogues between Mahåv¥ra and his chief mendicant disciple
IndabhËi Goyama, in the context of such wars, will further demonstrate
how close the two rival Íramaâ traditions were in their views on the
problem of death in battlefield and the karmic consequences following
such death.
The first narrative is about a war (saµgåma) called Mahå-
163
164 P.S. Jaini
silåkaˆ†aka :18
!
Venerable Sir!! How many people … were killed when the War of the
Big Stones took place!?
O Goyama!! In that war 8,400,000 were killed!!
Venerable Sir!! Among them there were men wounded in that war, who
were devoid of the good conduct (niss¥lå) … devoid of the holy practice of
observing the fasts, angry, malicious … who had not achieved peace. When
they died, what was their destiny, where were they reborn!?
O Goyama!! A great many of them were born in hells (naraga) and as
animals (tirikkha-joî).
As in the Mahåvaµsa, here too the dead are counted in tens of

millions, an exaggeration that may be ignored. The term s¥la stands for
the lay precepts (called aûvratas) that are similar to the five sikkhå-
padas of a Buddhist householder.19 The term niss¥lå thus agrees with
the word duss¥lå. The animal births declared here for the vast numbers
of the dead should enable us to understand the ambiguous Mahåvaµsa
expression pasusamå also to mean the same.
The next dialogue takes place in the context of another major war
initiated by King Ajåtasattu and is called the War of the Chariot with
the Mace (raha-musala-saµgåma). The monk Goyama asks Mahåv¥ra
the following question :20
!
18mahåsilåkaˆ†aye âµ bhaµte saµgåme va††amåne kai janasayasåhass¥o

vahiyåo!? Goyamå!! caurås¥iµ jaâsayasåhass¥o vahiyåo | te âµ bhaµte!!
maûyå niss¥lå jåva nippaccakkhåâ-posahovavåså ru††hå parikuviyå
samara-vahiyå aûvasaµtå kålamåse kålaµ kiccå kahiµ gayå kahiµ
uvavannå!? Goyamå!! osannaµ naraga-tirikkhajoîesu uvavannå. Suttågame
VII 9.
19The first four sikkhåpadas of a Buddhist upåsaka are identical with the first
four aûvratas (called “minor restraints” as against the mahåvratas of a

mendicant) of a Jaina upåsaka. Instead of surå-meraya-majja-pamåda††hånå-
veramaˆ¥, the fifth sikkhåpada, the Jainas have parigrahaparimåâ (“setting
limits to one’s property”). See Jaini 1979, pp. 170–78.
20bahujaê âµ bhante $! annamannassa evaµ åikkhai jåva parËvei$: evaµ
khalu bahave maûsså annayaresu uccåvaesu saµgåmesu abhimuhå ceva

pahayå samåˆå kålamåse kålaµ kiccå annayaresu devaloesu devattåe
uvavattåro bhavanti, se kahaµ eyaµ bhante$! evaµ$?
Venerable Sir !! Many people say to each other … and expound thus!:
“Indeed, men coming face to face in battles, large and small, wounded there
and dead, are reborn among the gods in various heavens.” Is this truly so,
Sir!?
O Goyama !! People who said such things to each other … and
expounded, truly have uttered a falsehood (micchaµ te åhaµsu). As for me,
O Goyama!! I say … and expound the following.
The word micchaµ used by Mahåv¥ra here to characterize the

disputed assertion is reminiscent of the term micchådi††hi employed by
the Buddha in the Yodhåj¥va-sutta. It is possible that the Jainas did not
wish to give the status of a dogma (di††hi) to the “idle talk” of the
people, but the term micchå is no less emphatic in conveying the falsity
of that talk. Indeed, Mahåv¥ra’s subsequent explanation lays down the
correct course of action, missing in the Buddhist literature, for a warrior
to attain heaven after death on a battlefield.
The Mahåvaµsa figuratively states that only “one and a half”
men (diya""hamanujå) — one with only the saraâs and another with
the lay precepts — were truly killed in that war. But there is no
narrative, in the Mahåvaµsa-†¥kå, the Extended Mahåvaµsa or even the
later work Rasavåhin¥,21 on these two pious men who were singled out
by the arahants out of the “millions” dead in the war. Fortunately, the
present Jaina narrative, which by a happy coincidence also speaks of
only two such men, illustrates the correct way for a layman to lay down
his life on the battlefield and be born in heaven or as a human being.
Mahåv¥ra gives an account of one of his lay disciples, an expert
archer named Varuâ of Vaißåli. He was a samaâ-uvåsaga and he had
taken the precepts of a layman, the first of which is ahiµså, refraining
from killing a human or animal being. At the time of taking his precepts
however, he had made an exception that would allow him to participate
Goyamå !! jaˆâµ se bahujaô annamannassa evaµ åikkhai jåva

uvavattåro bhavanti, je te evaµ åhaµsu micchaµ te evaµ åhaµsu, ahaµ
puâ Goyamå$! evaµ åikkhåmi jåva purËvemi — evaµ khalu Goyamå$!
Suttågame VII 9.
21The author of Ras simply quotes Mhv 25 108–11 (p. 277) without comment.
165
166 P.S. Jaini
in warfare if ordered by the king. Later when he was drafted by King
Ajåtasattu to fight in the raha-musala-saµgåma, Varuâ, armed with
bow and arrow, mounted his chariot and entered the war. He made a
further vow that he would not be the first one to shoot, and so he called
upon his adversary to shoot first. Only after his opponent’s arrow was
already on its deadly flight did he let fly his own arrow. His enemy was
killed instantly, while Varuâ himself lay mortally wounded. Realizing
that his death was imminent, Varuâ took his chariot off the battlefield,
sat down and held his hands in veneration to Mahåv¥ra, and said,22
Salutations to the Venerable Samaâ Mahåv¥ra, my teacher of dhamma.
I pay my respects to him wherever he may be.… Previously I have taken
from the Venerable Samaâ Mahåv¥ra the lifelong vow of refraining from
all forms of gross killing of life … up to … excess possessions. Now at this
time of my death, making the Venerable Samaâ Mahåv¥ra my witness, I
undertake the total renunciation of all forms of violence … and of all my
possessions … until my last breath.
Saying thus he pulled out the arrow and, with his mind at peace,
died instantly and was reborn in Saudharma, the first heaven.
The second man, a friend of Varuâ from childhood, fighting in the
22namo ’tthu âµ samaâssa bhagavao Mahåv¥rassa … mama dhammå-

yariyassa vaµdåmi âµ bhagavaµ tatthagayaµ ihagae. påsau me se
bhagavaµ tatthagae jåva vaµdai namaµsai. evaµ vayås¥$: pubbiµ pi âµ
mae samaâssa bhagavao Mahåv¥rassa antie thËlae påˆåivåe paccakkhåe
jåvajj¥våe evaµ jåva thËlae pariggahe paccakkhåe jåvajj¥våe. iyåî pi âµ
tasseva arihaµtassa bhagavao Mahåv¥rassa aµtiyaµ savvaµ påˆåivåyaµ …
paccakkhåmi jåvajj¥våe … caramehiµ Ësåsan¥såsehiµ vosiråmi tti ka††u …
samåhipa"ikkante samåhippatte åûpuvv¥e kålagae.
tassa âµ Varuâssa ege piyabålasaµvåsaye rahamusale saµgåme …
gå"hapahår¥ kae …Varuâµ påsai … evaµvayås¥!: jåiµ âµ … Varuâssa
s¥låiµ vayåiµ … veramaˆåiµ tåiµ âµ mamaµ pi bhavaµtu tti ka††u …
salluddharanaµ karei … kålagae.
Varuê âµ bhaµte … kålaµ kiccå kahiµ gae kahiµ uvavanne!? Goyamå$!
Sohamme kappe devattåe uvavanne …
Varuâsa piyabålavayaµsae kålaµ kiccå kahiµ uvavanne!? Goyamå$!
sukule paccåyåe.
Suttågame VII, 9, nos. 302– 303.
same battle, was also wounded ; but seeing his friend mortally wounded,
!
he helped him to sit comfortably. The text does not give his name or his
religion, but as he was helping Varuâ, he heard Varuâ’s words of
renunciation and said, “Whatever vows you have taken, let those be
mine too.” And so saying he also died and was reborn as a human being
in a noble family. 23
These stories of one person totally renouncing all violence at the
time of death, and the other person consenting to his renunciation in a
friendly way, and thus both dying a holy death on the battlefield, would
surely win the approval of the arahants who pointed to the one and a
half (diya""ha) good Buddhists in the story of King Du††hagåmaî’s
remorse.
This remarkable concordance between the two rival Íramaâ
traditions on the problem of heaven and warfare establishes the fact that
a study of one tradition sheds light on the other and helps us understand
both traditions at a deeper level. On this auspicious occasion of the
125th anniversary of the Pali Text Society, we do well to remember and
honour the name of Hermann Jacobi, the editor of the first volume of
the Pali Text Society published in 1882. Few now will even know that
this volume happened to be not of a Påli text, but the first book of the
Jaina canon, called the Óyåra"ga-sutta. We may recall today the words
he used in his introduction to the first volume in the series : “The
!
insertion of a Jaina text in the publication of the Pali Text Society will
require no justification in the eyes of European scholars. … But it is
possible that Buddhist subscribers … might take umbrage at the
intrusion, as it were, of an heretical guest into the company of their
sacred Suttas.” We should be grateful to Jacobi for showing us from the
beginning of the Pali Text Society that our studies of Påli and Buddhism
should go hand in hand with the studies of Prakrit and Jainism.
Padmanabh S. Jaini
Berkeley
23Foran abridged version, see Deleu 1996. This story also appears in Jaini
2000.
167
168 P.S. Jaini
ABBREVIATIONS
Abbreviations for Påli texts follow A Critical Påli Dictionary.
CD Bhikkhu Bodhi, tr., The Connected Discourses of the Buddha.
Wisdom/PTS, 2000
DB T.W. Rhys Davids and C.A.F. Rhys Davids, tr., Dialogues of the
Buddha
KS F.L. Woodward, The Book of the Kindred Sayings
Ras Rasavåhin¥. Transcribed from Sinhalese by Sharada Gamdhi.
Delhi!: Parimal Publications, 1988
Suttågame Pupphabhikkhu, ed. Suttågame, Viyåhapannatti (Bhagava¥), 1952
BIBLIOGRAPHY
Chalmers, Robert 1926–27. Further Dialogues of the Buddha
Deleu, Jozef, 1996. Viyåhapannatti (Bhagava¥), Motilal Banarsidass
Jaini, P.S., 1979. The Jaina Path of Purification, California
——— 2000. “Ahiµså: A Jaina Way of Spiritual Discipline” in P.S. Jaini,
Collected Papers on Jaina Studies, pp. 14–17, Delhi
Malalasekera, G.P., 1928. The Pali Literature of Ceylon, Colombo (reprinted
1958)
Murti, G.S., and A.N.K. Aiyangar, eds., 1951. Edicts of Aßoka (Priyadarßin)!:
Text and English Translation. Adyar Library
Geiger, W., tr. (assisted by Mabel H. Bode), 1912. The Great Chronicle of
Ceylon
Rahula, Walpola, 1956. History of Buddhism in Ceylon, Colombo
Sa¤khepasårasa¤gaha : Abbreviation in Påli
´
ratnattaya vanditvåna dh¥rassa bahussutassa

yassa savaˆânåyåya sa¤gaho’mhehi racito
yena bahulikhitena raññå uttaranarena
lekhåtid¥ghamattåya ninditå ’smi pubbakåle
tassa d¥ghåyulekhana nandant¥ payacchåm¥ma
mågadh¥sa¤khepanassa atisa¤khittavaˆâna
In celebrating the longevity and prolific contribution of our cause, I
seek to atone with the following brief synopsis of abridgement for
former length deemed by him excessive.
Reducing or replacing repetition that contains little or no variation
when recording texts in written form, skipping the chorus with a scribal
ditto for an aural fullness, is achieved through the term peyyåla
“formula, repetition” (PED s.v. and Norman 2006, pp. 113–14) reduced
further to pa, pe, pe … la. 1 Cf. Sanskrit peyyåla ( e . g .
1The characteristic repetition of some Påli literature is usually identfied as an

aid to the oral memory of a text, even though it does not assist memory of the
non-repetition (summary of theories to date´: Allon, pp. 354–57, his own 398).
An alternative avenue of exploration would be to consider the performance
function, drawing on textual anthropology´: The lead/expert monk(s) recite the
whole, resting their voice while the larger “chorus” pick up the refrains. This
would tie in with the observations made by Norman in his discussion of
“Buddhism and Oral Tradition” on the basis of anthropology by Tambiah who
in turn describes how the common and repeated formulae are those
remembered by most monks (Norman 2006, pp. 62–63). Current theories and
observation of the performance of Påli literature leave me with questions´: To
what extent has performance shaped the form of the text´? To what extent are
the D¥gha-nikåya texts more repetitive because important ceremonially´? Is the
performance function sometimes a factor in the difference between shorter and
longer versions of the same text´? For example, is a Mahåsatipa††hånasutta used
for a grander funeral´? Does the repetition really give the audience “an
opportunity to grasp” the content (Allon, p. 362) when the repetition is not of
the essence, or should we consider that at the time of its taking on that format
the text in Påli was, as today, already understood primarily as powerful sound
170 Kate Crosby
SamådhiråjasËtra) or alternatives such as pËrvavad yåvat (e.g.
Divyåvadåna).2 In a D¥gha-nikåya text peyyåla might replace thirty per
cent of unabridged content (Allon, pp. 275ff.), in Abhidhamma even
more.
Omission even in cases of variation is possible, where a sample
gives an impression of the whole, e.g. progressive intermittent numbers,
one verb where grammar requires more (Allon, pp. 354–57). A
compound conveys beyond itself, relationships unexpressed, linguistic
traces of an original context sometimes not fully erased : the samåsa,
´
plain, aluk, or syntactical (Norman 1991). Contractions, sometimes

contortions, also comprise external sandhi (Norman 1993).
Yåva(t) (i)ti ådi, etc., denote lists, whether numeric, specific, or
generic, giving only one or a few items.
Na-mo bu-ddhå-ya and a-ra-ha are examples of the parikamma
“aids” to practice in pre-reform Theravada, the microcosmic–macro-
cosmic identification that encapsulates the great within the tiny : five-
´
syllabled namo bu-ddhå-ya representing pentads such as khandha,

Buddhas ; trisyllabic araha representing triads – gems, robes, breaths,
´
Pi†aka (Crosby 2000, p. 147). They protect aurally or visually, perhaps

as a blue tattoo (Bizot 1981). Similarly, the full funerary works can be
performed on a budget : extracts of the seven Abhidhamma books
´
precede Pra Maleyya. Parallels are found in Sanskritic and Tibetan

Buddhism, where the budgetary and temporal restrictions on acquisition
of merit result in first-page recitation or simultaneity of all.
Overviews offer condensed coverage, comprehensive accessibility,
and decoctions of the essence. Title words : -sa¤khepa, -sa¤gaha,
´
-samåsa, -såra, e.g. Saccasa¤khepa, Abhidhammatthasa¤gaha, Nåma-

rËpasamåsa, Sårasangaha (von Hinüber 1996, Chapters VI, VIII). The
earliest is the Suttasa¤gaha (Norman 1983, pp. 172–73). Cognate
adverbs express authorial intent : sa¤khittena sa¤khepena the opposite
´
rather than through the verbatim meaning of its specific content´?

2My thanks to Andrew Skilton for these references to Sanskrit literature and to
the avadåna below.
´ 171
of vistarena. These can also refer to a familiar tale. Cf. Gåndhår¥
avadåna and pËrvayoga : “The whole [story] is to be done [i.e. recited]
´
in full … vistare janidave siyadi … sarva vistare yasayupamano siyadi”

(Salomon 1999, pp. 36, 38–39).
Ritual and regulatory reminders are generated by prompts that
provide the beginning but not the end, e.g. namo tassa for namo tassa
bhagavato arahato sammåsambuddhassa. Key words are used in the
elaborate abbreviation of yogåvacara manuals such as the Amatåkara-
vaˆânå to encapsulate an array of ritual and meditation instructions in
a single verse, fuller formulae to be drawn down from instruction given
earlier or elsewhere (Crosby 2005). The result was not recognised as a
list of keywords from sentences otherwise unrepresented and was
emended as if a set of single sentences with faulty grammar by
Ratanajoti and Ratanapali (1963), who then — not recognising the
import — in turn abridged the text further from 3818 to 1135 verses
(thus not as recorded Norman 1994, reprint, p. 268). A similar “drawing
down” familiar from Påîni along with the code letters triggering
treatment used therein is found in the Pali adaptations of the same, such
as the Kaccåyana-vyåkaraâ (Norman 1983. pp. 163–67).3
Måtikå are very productive as tables of content, key words to
summarize the whole, the mother who generates the teachings. Multiple
functions have been illuminated by Gethin : mnemonic (149), point of
´
access to whole (155), guide to structure (155), to composition (156), to

mindfulness (165), and adeptly summarised by Allon (7). In South-East
Asia the “mother” also generates the ritual fœtus (McDaniel 5), the
embryonic Buddha within (Crosby 2000).
Acronyms and acrostics encompass secret and powerful encapsula-
tions, such as the first syllables that form the “hearts” hadaya (Penth).
Compare dhåraˆ¥ such as the arapacana (Braarvig). Some are not so
secret : in South-East Asia the first letter of the seven Abhidhamma texts
´
(Swearer 1995 A ) ; in Thailand, the first syllables of each of the

´
3I have only seen manuscript versions. Norman (1983, p. 163) cites the printed
edition by E. Senart, Journal Asiatique 1871, pp. 193–544.
172 Kate Crosby
bodhisattas in the final ten Jåtakas (Shaw xxxiii).4
Numinous powers of the Buddha are harnessed through the poetic
synopses of biographic episodes to empower a statue (Swearer 1995B),
to heal or bring peace, or just to entertain (Somadasa: vii with examples
from the Nevill collection throughout).
Kate Crosby
School of Oriental and African Studies
BIBLIOGRAPHY
Allon, Mark, 1997. Style and Function´: A Study of the Dominant Stylistic
Features of the Prose Portions of Påli Canonical Sutta Texts and Their
Mnemonic Function, Tokyo´: The International Institute for Buddhist
Studies
Bizot, F., 1981, “Notes sur les yantra bouddhiques d’Indochine”, in Tantric and
Taoist Studies in Honour of R.A. Stein, ed. M.´Strickmann, Bruxelles´:
Mélanges Chinois et Bouddhiques, Vol. XX, pp. 155–91
Braarvig, Jens, 1985. “Dhåraˆ¥ and Pratibhåna´: Memory and Eloquence of the
Bodhisattvas”, Journal of the International Association of Buddhist
Studies 8 (1), pp. 17–29
Crosby, Kate, 2000. “Tantric Theravada´: A Bibliographic Essay on the Writings
of François Bizot and Other Literature on the Yogåvacara Tradition”,
Journal of Contemporary Buddhism, November issue 2, pp. 141–98
——— 2005. “Differences Between the Vimuttimagga-uddåna and the
Amatåkaravaˆânå”, Journal of Buddhist Studies (Sri Lanka´: Centre for
Buddhist Studies), Vol. 3, January, pp. 139–51
Gethin, Rupert, 1992. “The Måtikås´: Memorization, Mindfulness, and the List”
in In the Mirror of Memory ´: Reflections on Mindfulness and
Remembrance in Indian and Tibetan Buddhism, ed. Janet Gyatso,
pp.´149–72, Albany´: State University of New York Press´
4My thanks to Naomi Appleton for this reference.

´ 173
von Hinüber, O., 1996. A Handbook of Påli Literature, Berlin´: Walter de
Gruyter
Lopez, Donald S., ed., 1995. Buddhism in Practice, Princeton University Press
McDaniel, Justin Thomas, forthcoming 2008. “Philosophical Embryology´:
Buddhist Texts and the Ritual Construction of a Fetus” in Imagining
the Fetus, ed. Jane Marie Law and Vanessa Sasson, American
Academy of Religion Publications
Norman, K.R., 1983. Påli Literature´: Including the Canonical Literature in
Prakrit and Sanskrit of All the H¥nayåna Schools of Buddhism. A
History of Indian Literature, Vol. VII, fasc. 2, Wiesbaden ´: Otto
Harrassowitz
——— 1991. “Syntactical compounds in Middle Indo-Aryan” in Middle Indo-
Aryan and Jain Studies, ed. C. Caillat, Leiden´: Brill, pp.3–9
——— 1993. “External Sandhi in Påli (with Special Reference to the
Suttanipåta)”, Journal of the Pali Text Society XIX, pp. 203–13
——— 1994. “Påli Literature´: Appendix 1 (Additions and Corrections)”,
Journal of Påli and Buddhist Studies (Påligaku BukkyØ Bunkagaku 7,
1–22 = Collected Papers V, Oxford´: Påli Text Society, 1994 (no. 115),
pp. 262–83
——— 2006. A Philological Approach to Buddhism, 2nd edition (1st edition
1997), Lancaster´: Pali Text Society
Penth, Hans, 1997. “An Inscribed Buddha Image in Wat Ket S¥”, in Bauddha-
vidyåsudhåkara Studies in Honour of Heinz Bechert on the Occasion of
His 65th Birthday, ed. P.´Kieffer-Pülz and J.-U. Hartmann. Swisttal-
Odendorf´: Indica et Tibetica Band 30, pp. 495–507
Ratanajoti, Galkätiyagama, and Karalliyadd„ Ratanapåla, eds., 1963. Vimutti-
maggo. (Bhadantårahanta-Mahåri††ha-Upatissa-tthera-vara-ppan¥to),
Colombo´: Government Press of Ceylon
Rhys Davids, T. W., and W. Stede, eds., 1921–1925. The Pali Text Society’s
Pali–English Dictionary, London´: Pali Text Society, reprint 1986. (=
PED)
Salomon, Richard, 1999. Ancient Buddhist Scrolls from Gandhåra´: The British
Library Kharo¡†h¥ Fragments (with contributions by Raymond Allchin
and Mark Barnard), London´: British Library
Shaw, Sarah, 2006. The Jåtakas´: Birth Stories of the Bodhisatta, New Delhi´:
Penguin
174 Kate Crosby
Somadasa, K.D., 1990. Catalogue of the Hugh Nevill Collection of Sinhalese
Manuscripts in the British Library, Vol. 3, London and Henley on
Thames´: The British Library and Pali Text Society
Swearer, Donald K., 1995 A . “A Summary of the Seven Books of the
Abhidhamma”, in Lopez, pp. 336–42
——— 1995B. “Consecrating the Buddha”, in Lopez, pp. 50–58
Recent Japanese Studies in the Påli
Commentarial Literature: Since 1984
1. Introduction
After extensive research, my dissertation of nearly 750 pages, entitled A
Study of the Påli Commentaries : Theravådic Aspects of the A††hakathås
´
(in Japanese with an English summary), was published in 1984.

Although there had been a few works preceding it abroad,1 this work
(abbreviated as SPCJ hereafter) was really the first major publication in
this field of study, at least in Japan. In the twenty years or more since
then, Japanese studies related to the Påli commentarial literature
(A††hakathå texts)2 have improved remarkably, far beyond my own
expectations.
Based on SPCJ and other later works by me, many Japanese
scholars have done research in the Påli commentaries using various
points of view and lines of inquiry. This research, which dealt with the
commentaries not only as objects of research in and of themselves, but
also as primary material aiding the exploration of many issues in
Buddhist studies, can be classified here into the following six cate-
gories. I will subsequently discuss some of the outstanding achieve-
ments in each category.3
1 e.g.(1) E.W. Adikaram, Early History of Buddhism in Ceylon (Colombo´:

M.D.´Gunasena, 1946). His philological study on the commentaries in this
book, however, appeared only in 42 pages of Part I. (2) F. Lottermoser, Quoted
Verse Passages in the Works of Buddhaghosa (Göttingen´: author, 1982). Its
subject was very limited, not like a general discussion. Cf. Mori 1985 (in
English) as a review.
2 This literature is to be limited here to the Visuddhimagga and the direct
commentaries to the Påli Tipi†aka.
3As for the works published by foreign scholars, some of which are surely very
important, I shall discuss them in another article.
176 Sodo Mori
2. Japanese Translations of Some Commentaries
The publication of a Japanese translation series of the Påli Tipi†aka with
some other texts in Påli was completed in 1941, six years after it was
begun, as a result of the sincere cooperation of many scholars. It con-
tained seventy volumes altogether, and a useful general index was later
added by Kogen Mizuno.
As to the Japanese translations of the Påli commentaries, the
Visuddhimagga, Atthasålin¥, the Båhiranidåna of the Samantapåsådikå,
Kathåvatthu-a††hakathå, and the Nidånakathå of the Jåtaka††hakathå had
been published before 1984 when SPCJ was published. These transla-
tions were generally preceded by their English translations which had
been published mostly by the Pali Text Society.4
Since 1984, several translations of the commentaries into Japanese
have been published : some were preceded by their English versions and
´
others were not, meaning that the latter cases were the first translations
in the world. These are Murakami and Oikawa (1985–89) in four
volumes, the first translation of the Paramatthajotikå ; Naniwa (2004),
´
which consists of a full translation of the Vibha¤ga††hakathå and the

first translation of its MËla†¥kå ; Katsumoto (2007), as her dissertation
´
contains the initial full translation of the Cariyåpi†aka††hakathå, a text of

the Paramatthad¥pan¥. In addition, there is Fujimoto (2006 in Japanese),
a dissertation which also contains a new translation, i.e. the translation
of the major stories of the Petavatthu-a††hakathå with an abridged
translation of the rest of the stories. It can thus be expected that the
translation of works of the Påli commentaries into Japanese will
continue concurrently with future English translations.
3. Historical Studies of Buddhist Doctrine and Thought

Prior to the publication of SPCJ in 1984, the A††hakathå texts which
were referred to for the doctrinal studies were usually limited to a few
Abhidhamma works such as the Visuddhimagga, Atthasålin¥,
4 Regarding all the publications of the Pali Text Society including English
translations, see its web site (http://www.palitext.com).
Recent Japanese Studies in the Påli Commentarial Literature 177
Sammohavinodan¥, Kathåvatthu-a††hakathå, and so on. Since 1984,
however, many other commentaries have gradually been taken up as
important original texts, and now the achievements of this new
approach have progressed splendidly in both quality and quantity. I will
introduce here only the following five dissertations out of a great many
excellent examples.
Endo (1997), a work in English, discussed in detail the develop-
ment of the Buddha concept along with the Bodhisatta concept in
Theravåda Buddhism, referring to the Påli Canon, commentaries, and
some sub-commentaries. His work was highly esteemed in Sri Lanka,
where it was published, as well as here in Japan. Oikawa (1998, in
Japanese, unpublished) studied the Paramatthajotikå, the commentary
on the Khuddakapå†ha and Suttanipåta, for the first time, focusing on its
philological, historical, and social aspects, as well as its background.
This was written on the basis of his co-translation of the Påli original as
stated earlier. The greater part of his research appeared as Part II in the
work he co-authored with Murakami in 1990 (in Japanese). Fujimoto
(2006 in Japanese with an English summary) discussed the Buddhist
idea of merit transference with reference to the Peta stories as related in
the Påli Petavatthu and its commentary, a text of the Paramatthad¥pan¥.
His study added a great deal of new thought and knowledge to that
which was already prevalent in the Northern tradition, and contained
Japanese translations of many Peta stories in the commentary, related to
the above subject. Baba (2006, in Japanese, unpublished) is a very
valuable study which discusses the history of the ti-vijjå (three-
knowledge) tradition with special reference to changes in the biography
of the Buddha and to the formation of the meditation system in
Sectarian Buddhism of India. In his research, the Påli Canon, the
Visuddhimagga, and the commentaries on the first four Nikåyas were
primarily referred to in comparison with certain classical Chinese texts
of Northern Buddhism. Katsumoto (2006, in Japanese, unpublished),
already touched on in the previous section, is a very unique piece of
research which examines certain Mahåyåna elements depicted mainly in
178 Sodo Mori
the Cariyåpi†aka and its commentary, Buddhavasa and its
commentary, and also the Nidånakathå of the Jåtaka††hakathå. Her
research could clear up, as a result, some questions regarding the
relationship between the Påli commentaries and Mahåyåna texts, and
the influence of Mahåyåna, especially the Yogåcårin school, on the Påli
commentaries. It also raises many new questions as to the historical
interchange between Theravåda in Sri Lanka and Mahåyåna in India.
4. Philological Studies in the Source References

for the Commentaries
Of the source materials for the Påli commentaries, SPCJ initially
classified them according to the following six categories : (1) the Påli
´
Tipi†aka ; (2) three semi-canonical texts following the Tipi†aka ; (3) the
´ ´
Påli A††hakathås themselves ; (4) the so-called “lost” S¥ha¬a A††ha-

´
kathås ; (5) source references of other schools : views attributed to

´ ´
“some” (keci) and views attributed to “sophists” (Vitaˆ avådins) ; and

´
(6) others : D¥pavasa, Kaccåyanappakaraâ, Mahåniruttippakaraâ,

´
and so forth. Among the above source references, (1), (2), (3), and (6)
were already known, but (4) and (5) were entirely unknown sources.
Consequently, I investigated each of the altogether 35 categories of such
sources in SPCJ.5 However, according to subsequent research done by
me after SPCJ, their final number amounted to 40.6 Meanwhile, more
detailed research has been done on some sources. For instance, Endo
(1999, in English) studied thoroughly the Paramatthad¥pan¥ of
Dhammapåla, with a special reference to “some” (apare, keci, etc.) as
its source, and conclusively found certain important differences in
passages between the Paramtthad¥pan¥ and some works of
Buddhaghosa, and also between those in the Paramatthad¥pan¥ and
certain sub-commentaries, both of which have traditionally been
ascribed to Dhammapåla himself. These findings provided new
5 Incidentally, Adikaram (Early History of Buddhism, p. 10, see n. 1 in this

article) listed only 28 categories of such sources.
6Mori (1987D, 1989A in English´; 1989 in Japanese).
questions as to the authorship of some of the commentaries and sub-
commentaries together with biographical details pertaining to
Dhammapåla. His study exerted a great influence on Katsumoto’s study
as will be explained later. Endo (2002) investigated potthaka, a vague
source reference in the commentaries (other than those above), and
concluded that it was not a particular source reference which is now in
question. Then Endo (2003, in English) examined the quotations from
the Bhåâkas, as a sort of old source, and further Endo (2005, in
English) discussed the chronology of the “A††hakathå” which has
always been expressed in the singular form, and which, though now
lost, is a work regarded as the basic source material of Indian origin.
Concurrently, Hayashi (2005 in Japanese) analyzed the Vipåka-
kathå, another source material referred to in the Visuddhimagga, and
reached the conclusion that the section named the Vipåkuddhårakathå in
the Atthasålin¥, which contains a special reference to the Ussadakittana,
which is the same as in the Vipåkakathå, is none other than the
Vipåkakathå itself of the Visuddhimagga. Thus the study of the source
materials for the commentaries is still very much in progress.
5. Problems of Authorship Pertaining to Each Work

Regarding the traditional belief in the authorship of the A††hakathå
texts, there remain a great many problems yet to be solved. Some
scholars have worked on this quite difficult and complicated matter.
Hayashi (1997 in Japanese, 1999 in English) criticized the view that the
Atthasålin¥ is not the work of Buddhaghosa, which was once insisted on
in detail by P.V. Bapat7 and more recently by O.H. Pind.8 To the
contrary, Sasaki (1997, in Japanese, (1), pp. 57–58, n. 23) pointed out a
contradictory textual fact existing between the Atthasålin¥ and the
7 P.V. Bapat and R.D. Vadekan, eds., A††hasålin¥, Poona, 1942 ´: Bhandarkar
Oriental Series No. 3, pp. xxviii–xl.
8 O.H. Pind. 1992. “Buddhaghosa´: His Works and Scholarly Background”,
Buddhist Studies 21, pp. 135–56. Mori (1992 in Japanese) reviewed this
article.
180 Sodo Mori
Samantapåsådikå, both of which are attributed to Buddhaghosa himself
in the Theravåda tradition.
Incidentally, Sasaki and Yamagiwa (1997, in Japanese) started their
project of research on the Samantapåsådikå, the Vinaya commentary,
comparing it with the Påli Vinaya-pi†aka and other Vinaya-pi†akas,
some of which originally contain their respective commentaries from
the Northern tradition. As a part of their research, Sasaki (1997–99)
examined certain complicated relations among the three works, the
Visuddhimagga and the Samantapåsådikå, both equally ascribed to
Buddhaghosa, and the Gedatsu-dØ-ron, the Classical Chinese version of
the Påli Vimuttimagga authored by Upatissa, which is, in spite of the
non-Mahåvihåra fraternity text in ancient Sri Lanka, one of the most
fundamental source references for the Visuddhimagga.9 His conclusion
at present is as follows : it cannot be asserted that the author of the
´
Visuddhimagga was the same person as that of the Samantapåsådikå,

whereas there can be found a certain accord between the Samanta-
påsådikå and the Gedatsu-dØ-ron on some points. In fact the
triangulated relations among the Visuddhimagga, the other com-
mentaries and the Gedatsu-dØ-ron still remain unclear.
Concerning this, Mori (1982 in Japanese) had already given another
example as follows. Regarding the doctrine of the kamma††håna (the
object of meditation practice), the Gedatsu-dØ-ron states 38 kinds of
such objects, while the Visuddhimagga states 40, and since the latter
work was followed in this regard by such later Påli texts as the
9 The Vimuttimagga is still a very problematic text´: not only the school to
which it belonged, but also the words and passages in the Classical Chinese
version and so on are being seriously questioned, e.g., K.R. Norman, “The
Literary Works of the Abhayagirivihårins”, Collected Papers IV (Oxford´:
PTS, 1993), pp. 202–17´; Peter Skilling, “Vimuttimagga and Abhayagiri´: The
Form-Aggregate According to the Sask¤tåsask¤tavinißcaya” (JPTS XX
(1994)), pp. 171–210 ´; Kate Crosby, “History Versus Modern Myth´: The
Abhayagirivihåra, the Vimuttimagga and Yogåvacara Meditation” (Journal of
Indian Philosophy 27-6 (1999)), pp. 503–50´; Hayashi (2003, 2004, 2006 in
English). Cf. Mori (1988C in English).
Abhidhammåvatåra of Buddhadatta and the Abhidhammatthasa¤gaha of
Aruruddha, the system using 40 kinds as shown in the Visuddhimagga
has been widely recognized as the standard doctrinal system in the
Mahåvihåra tradition. With careful examination, however, it could be
found that the system using 38 kinds is also described in certain
commentaries such as the Samantapåsådikå, Såratthapakåsin¥,
Sammohavinodan¥, and Paramatthajotikå, just as in the Gedatsu-dØ-ron,
a non-Mahåvihåra text. The philological aspects as found among the
Visuddhimagga, the other A††hakathå texts and the Gedatsu-dØ-ron
show thus such a complicated situation that further research will be
needed for the final solution of the authorship problem of the commen-
tarial literature to be revealed.
6. Comparisons with the Texts of Northern Sects

The following has been taken as an effective method for the study of
early Buddhism : In a comparison of the Påli Vinaya- and Sutta-pi†akas
´
(with the exception of the Abhidhamma-pi†aka) with those of the

Northern sects in India, the corresponding passages and ideas among
them can be generally regarded as an older stratum which had been
formed before the initial schism of the Buddhist Order, and are therefore
more closely connected to the Buddha. The discordant passages and
ideas, on the other hand, are a newer stratum which was later altered or
added inside of each sect founded after the initial schism. Based on the
above methodological idea, studies on early Buddhism and sectarian
Buddhism were accomplished as a whole, searching the older and newer
strata of the two Pi†akas. For sectarian Buddhist studies in general, the
Abhidhamma-pi†akas and other Abhidhamma texts are of course to be
taken up as the essential material.
On the other hand, SPCJ pointed out the textual facts that the
present Påli commentaries consist of two fundamental strata : one is of
´
the older portions which were composed or cited mainly on the basis of
earlier source material of Indian origin, the contents of which can be
considered as closer to those of early Buddhism, and the other is of the
newer portions which were composed on the basis of the later sources of
182 Sodo Mori
Sri Lankan addition and alteration, the contents of which were accord-
ingly transformed into the Theravåda tradition. Since then a new
methodological tendency has been gradually prevailing which suggests
that the Påli commentaries should be properly utilized for the research
of Indian Buddhism including even early Buddhism in certain cases. As
a result, some portions which had not been found in the Påli Sutta- and
Vinaya-pi†akas, but were found only in the texts of the northern
tradition, could be newly discovered as being dormant in the Påli
commentaries.
For example, Yamagiwa (1996) and Sasaki (20002) respectively
searched the Samantapåsådikå in comparison with altogether six sorts of
Vinaya-pi†akas available today in Påli or Classical Chinese, and found
that certain corresponding passages or ideas are recorded not in the Påli
Vinaya, but in its commentary, i.e. the Samantapåsådikå under con-
sideration. Based on their findings, they expressed their view that the
Samantapåsådikå should be included as a necessary work for compara-
tive study of Vinaya texts within different traditions, which is definitely
useful to the study of the history of the Buddhist Order in India.
While on the other hand, Baba (2003, in Japanese) investigated
some Sutta texts preaching the theory of the “Chain of Dependent
Origination” (pa†iccasamuppåda) which is differently transmitted in
some sects. Regarding the Sutta-pi†aka he reached the same conclusion
as that of Yamagiwa and Sasaki concerning the Vinaya-pi†aka. Baba’s
dissertation, as touched on before, was a result of his further studies on
this subject. In any case, it should be noted that various studies which
sufficiently make use of the A††hakathå texts as indispensable references
can thus contribute not only to the historical studies of Theravåda
Buddhism in Sri Lanka, but also to Indian Buddhist studies in general.
7. Comparisons with Mahåyåna Texts and Studies

in Sri Lankan Mahåyåna
Although comparative study of the A††hakathås with Mahåyåna texts
had previously never been considered at all, my research after SPCJ
(Mori 1993 in Japanese ; 1997, 1999 in English) became a pioneering
´
study on this frontier. It can be summarized in the following manner. A
view of anonymous elders (ekacce therå) is negatively referred to as
deniable in some commentaries such as the Suma¤galavilåsin¥,
PapañcasËdan¥, ManorathapËraˆ¥, and Saddhammapajjotikå, but this
anonymous source reference can be known as the “Andhakas” and
“Vijñåâvådins” through the valuable comment on it recorded in their
sub-commentaries (†¥kås). In this regard, setting aside the case of the
Andhakas, whose trustworthy texts are not extant today, it is possible to
compare the commentaries in question with certain Vijñånavådin texts,
i.e. the Vißatikå Vijñaptimåtratåsiddhi in Sanskrit of Vasubandhu (4–
5c) and some Classical Chinese versions. As a result, the comment
under consideration could be proven as correct and this study gave us
the important insight that some of the compilers of the sub-
commentaries and perhaps of the commentaries possessed certain
adequate knowledge of the Vijñånavådins, as a Mahåyåna school, at
least in this topic.
Next to my study above, Shimoda (2000, in Japanese) tried to
examine a similar sort of topic. It was a discussion that made clear a
certain similarity between the A††hakathå and Mahåyåna texts : The ´
similarity in question is that the four kinds of classifications of Buddhist

preaching which were adopted in the later stratum, i.e. the commentarial
part, of the Mahåyåna Mahåparinirvåâ SËtra are actually indentical
with those explained about in the first four Nikåya-commentaries
ascribed to Buddhaghosa, although their terms themselves are not in
concord. 10 Shimoda discussed this matter, based on his detailed
research of the above Mahåyåna text. At any rate, we can be sure that
barely perceptible relations seem to be lying between the Theravåda
commentaries and the Mahåyåna SËtras.
10 In the Påli commentaries, it is called suttanikkhepa (attajjhåsaya, para-

jjhåsaya, pucchåvasika, a††huppattika), whereas in the Mahåyåna SËtra in
question, it is called dharmaparyåya. As for the Påli terms explained in the
commentaries, von Hinüber seems to have found them earlier´: Oskar von
Hinüber, A Handbook of Påli Literature (Berlin´: Walter de Gruyter, 1996),
pp.´114–15.
184 Sodo Mori
In the same year, i.e. in 2000, a continuing study by Katsumoto
began to be published. Among her already published articles, we may
draw special attention to Katsumoto (2005, 2006, in Japanese) for a
point of view on the relationship between Theravåda and Mahåyåna
texts. Furthermore, an abundant stock of more knowledge and
information obtainable from this research in the future will improve the
study of Sri Lankan Mahåyåna history, and then the study of Sri Lankan
Mahåyåna and that of Indian Mahåyåna should be compared and unified
into one theme, giving mutual influence and cooperating on research.
Mori (2006 in Japanese) can be a useful guide to further studies in this
field.
8. Further tasks and final goals of research

The above is a brief description of the present state of study in the six
classified subjects about the Påli commentaries which have been
accomplished by Japanese since SPCJ was published in 1984. Needless
to say, further studies in each of these six subjects along with some new
subjects, if any exist, have to be continuously made. Concurrently, how-
ever, we have to pay attention to a fundamental problem which remains.
That is the search concerning the earlier aspects on the older stratum of
the Påli commentaries, which is closely connected to early Buddhism
and sectarian Buddhism in India. Incidentally, this new search would
contribute to the whole introductory study or general remarks of the
commentarial literature, while SPCJ, with the sub-title, “Theravådic
Aspects of the A††hakathås” was none other than the other half. It
mainly discussed the later aspects on the newer stratum of the Påli
commentaries, which were surely added in the Theravåda tradition of
ancient Sri Lanka. In any case, in order to succeed in this task, the
methods of analysis to be adopted are the most essential : any suitable
´
problems which can be related to all commentary texts should be first

selected, and also more than one such problem should be independent,
with no mutual relationship which would indicate different phases of the
older stratum being necessary. I am now considering a few such
problems.
Thus, when some matters concerning the older stratum — such as
its chronology ; or its philological, doctrinal, or historical character-
´
istics ; and so forth — have been illuminated as a whole via the methods
´
stated above, more exact and trustworthy results of research will be

possible, and then the comparative study of the Påli commentaries with
the northern sectarian or Mahåyåna texts based upon the above results
can be further improved. This will surely contribute to the study of
Indian Buddhist history itself.
On the other hand, as a final task in this particular field, “the study
of the formational history of the Påli commentarial literature” should be
pursued in the future. In addition, another ideal goal of our research
should be considered in parallel, that is a publication of the Japanese
translation series of all the commentaries in question, as a sequel to the
Nanden DaizØkyØ, the Japanese translation series of the Påli Tipi†aka
plus other texts in Påli, published altogether in seventy volumes
between 1935 and 1941 as mentioned above.
I am grateful to Associate Professor Gregory Rohe at Aichi Gakuin

University for improving my English.
References to Påli texts refer to the Pali Text Society’s editions
unless otherwise stated.
Sodo Mori
186 Sodo Mori
ABBREVIATIONS
JPTS´ Journal of the Pali Text Society
SPCJ´ Mori 1984 in Japanese
SPCE´ Mori 1989 in English (This is not a translation of SPCJ above, but a
collection of my articles on A††hakathå study.)
Mori Festschrift Buddhist and Indian Studies in Honour of Professor Sodo
Mori. Hamamatsu´: Kokusai BukkyØto KyØkai, 2002, in English
SELECTED BIBLIOGRAPHY
(works by Japanese scholars since 1984)
A. IN JAPANESE
Baba, Norihisa, 2003. “Commentarial Elements as Found in the Northern
Buddhist Canon” (Hokuden-agon no ChËshakushoteki YØsØ´: engikanren-
kyØten), Buddhist Studies (BukkyØ KenkyË) 31, pp. 193–219
——— 2006. History of the “Three Knowledge” Tradition´: changes in the
biography of the Buddha and the formation of the meditation system in
Sectarian Buddhism, unpublished dissertation
Fujimoto, Akira, 2006. A Study of Merit Transference´: With Special Reference
to Påli Peta Stories (EkØshisØ no KenkyË), Hamamatsu´: Kokusai Bukkyoto
Kyokai (Ratna BukkyØ-sØsho Vol. 1)
Hayashi, Takatsugu, 1997. “On the Authorship of the A††hasålin¥´: Suttanta††ha-
kathå and Ógama††hakathå” (A††hasålin¥ no Chosha nitsuite), Indogaku
BukkyØgaku KenkyË 46–1, pp. (102)–(106)
——— 2005. “On the Vipåkuddhårakathå/ Vipåkakathå” (Ijyukugairon/
Ijyukuron ´: Påli ChËshaku-bunken no GensenshiryØ ni kanrenshite),
Indogaku BukkyØgaku KenkyË 53–2, pp. (107)–(113)
Katsumoto, Karen, 2005. “Relation between Påli Buddhism and Mahåyåna”
(Påli BukkyØ to DaijyØ no KyØkaisen). A Study of Mahåyåna Buddhism´:
essays in Honour of Professor YËshØ Muranaka on His 70th Birthday
(DaijyØ BukkyØ no KenkyË), Tokyo´: Sankibo, pp. (89)—(101)
——— 2006. “Cariyåpi†aka††hakathå and BodhisattvabhËmi” (Cariyå-
pi†aka††hakathå to BodhasattvabhËmi), Buddhist Studies (BukkyØ KenkyË)
34, pp. 173–92
——— 2007. The Practices of the Bodhisatta Way´: A Study of Påram¥s in the
Cariyåpi†aka and Its Commentary (Bosatsu no ButsudØ-ShugyØ), unpub-
lished dissertation
Mori, Sodo, 1982. “Some Aspects of the Kamma††håna Theory” (Gossho-setsu
no Shujyu-sØ), Studies in Buddhist Doctrine´: Essays in Honor of Dr.
YoshirØ Tamura on His 60th Birthday (BukkyØ-kyØri no KenkyË), Tokyo´:
ShunjËsha, pp. 127–40.
——— 1984. A Study of the Påli Commentaries´: Theravådic Aspects of the
A††hakathås (Påli BukkyØ ChËshaku-bunken no KenkyË), Tokyo´: Sankibo
——— 1989. “Ariyavasa and Ariyavasakathå “ (Ariyavasa to Ariya-
vasakathå), Indian Philosophy and Buddhism´: Essays in Honor of
Professor KØtatsu Fujita on His 60th Birthday (Indo-tetsugaku to BukkyØ),
Kyoto´: Heirakuji-shoten, pp. 225–44
——— 1992. “Problems Concerning the Works Ascribed to Buddhaghosa´:
With Special Reference to O.H. Pind’s Study” (Buddhaghosa Chosaku no
Mondaiten), Indogaku BukkyØgaku KenkyË 41–1, pp. 1–11
——— 1993. “The Vijñånavådin View as Depicted in the Påli Commentaries´:
With Special Reference to the Nirayapåla-kathå” (A††hakathå ni mirareru
Yuishiki-setsu), The Encounter of Wisdom between Buddhism and Science´:
Essays in honour of Professor KeishØ Tsukamoto on His 60th Birthday (Chi
no KaigØ´: BukkyØ to Kagaku), Tokyo´: Kosei Shuppansha, pp. 321–35
——— 2006. “An Introduction to the Study of Mahåyåna Buddhism in Sri
Lanka” (Sri Lanka DaijyØ BukkyØ KenkyË Jyosetsu), Taisho Daigaku SØgØ
BukkyØ KenkyËjo NenpØ 28, pp. (113)–(133)
Murakami, Shinkan & Shinkai Oikawa, 1985–89. Japanese Translation of the
Paramatthajotikå (Hotoke no Kotoba-chË), 4 vols., Tokyo´: ShunjËsha
——— 1990. A Study of the Paramatthajotikå (Hotoke to Seiten no DenshØ),
Tokyo´: ShunjËsha
Naniwa, Senmyo, 2004. Japanese Translation of the Vibha¤ga††hakathå and
Vibha¤gamËla†¥kå (Funbetsuron-chË), Kyoto´: Heirakuji-shoten
Oikawa, Shinkai, 1998. A Study of the Paramatthajotikå (Daiichigi-kaimei no
KenkyË), unpublished dissertation
Sasaki, Shizuka, 1997–99. “The Visuddhimagga and the Samantapåsådikå, (1),
(2), (3)” (Visuddhimagga to Samantapåsådikå), BukkyØ Daigaku SØgØ-
kenkyËjo KiyØ, 4, pp. 35–63´; 5, pp. 57–81´; 6, pp. 151–78
188 Sodo Mori
——— 2000. “The Samantapåsådikå and the Vinaya Texts´: The Second
påråjika (1)” (Samantapåsådikå to RitsuzØ), Buddhist Studies (BukkyØ
KenkyË) 29. pp. 69–89
Shimoda, Masahiro, 2000. “The MahåparinirvåâsËtra as a Mahåyåna
Commentary ´: A Comparison with Påli Commentaries” (ChËshakusho
toshiteno DaijyØnehangyØ) Abhidharma Buddhism and Indian Thought´:
Essays in Honour of Dr. JyunshØ KatØ on His 60th Birthday (Abhidhrma
BukkyØ to Indo ShisØ), Tokyo´: ShunjËsha, pp. 327–39
Yamagiwa, Nobuyuki, 1996. “The Importance of the Samantapåsådikå in the
Study of the Development of the Vinaya Rules”, Indogaku BukkyØgaku
KenkyË, 45–1, pp. (113)–(117)
Yamagiwa, Nobuyuki & Shizuka Sasaki, 1997. “An Introduction to the Study
of the Samantapåsådikå” (Samantapåsådikå KenkyË-jyosetsu), Påligaku
BukkyØ Bunkagaku KenkyË, 10, pp. 25–35
B. IN ENGLISH
Endo, Toshiichi, 1997. Buddha in Theravada Buddhism´: A Study of the Concept
of Buddha in the Påli Commentaries, Dehiwela, Sri Lanka´: Buddhist
Cultural Centre
——— 1999. “Keci and Apare in Dhammapåla’s Commentaries”, Journal of
the Postgraduate Institute of Pali and Buddhist Studies 1, pp. 36–68
——— 2002. “Potthaka (Book or Manuscript) in the Påli Commentaries”, Mori
Festschrift, pp. 79–102
——— 2003. “Views Attributed to Different Bhåâkas (Reciters) in the Påli
Commentaries”, Buddhist Studies (BukkyØ KenkyË) 31, pp. 1–41
——— 2005. “The ‘A††hakathå’ as Source-Material of the Påli Commentaries´:
An Inquiry into the Date of Their Compilation”, Dhamma-Vinaya´: Essays
in Honour of Venerable Professor Dhammavihara (Jotiya Dhirasekera),
Colombo´: Sri Lanka Association for Buddhist Studies, pp. 33–53
Fujimoto, Akira, 2003. “Meanings of Patti and Pattidåna´: They Mean Neither
Merit (puñña) nor Transference (pariˆåmanå)”, Buddhist Studies 31,
pp.´123–54
——— 2004. “Dåna and Dakkhiˆå in the Context of ‘Offer of Donation’´:
dakkhiâ ådis-, dåna uddis-, etc.”, Buddhist Studies 32, pp. 83–114
——— 2005. “Is the Law of Kammassakatå Violated in the Context of ‘Offer
of Donation’´?”, Buddhist Studies 33, pp. 1–31
Hayashi, Takatsugu. 1999. “On the Authorship of the A††hasålin¥”, Buddhist
Studies 28, pp. 31–71
——— 2003, 2004, 2006. “The Vimuttimagga and Early Post-Canonical
Literature I, II, III”, Buddhist Studies 31, pp. 91–122´; 32, 59–82´; 34, 5–33
Mori, Sodo, 1982. “The Vitaˆ avådins (Sophists) as Seen in the Påli
A††hakathås”, Påli BukkyØ Bunka KenkyË (Study of the Påli Buddhist
Culture)”, Tokyo´: Sankibo, pp. 1–18 (= SPCE)
——— 1983. “A††hakathåcariyas and A††hakathikas”, Indogaku BukkyØgaku
KenkyË 31–2, pp. 977–82 (= SPCE)
——— 1985. Review article´: “Friedgard Lottermoser, Quoted Verse Passages
in the Works of Buddhaghosa´: Contributions Towards the Study of the Lost
S¥ha¬a††hakathå Literature”, Buddhist Studies 15, pp. 12–43 (= SPCE)
——— 1987A. “Mahås¥vatthera as Seen in the Påli A††hakathås”, JØsai Jinbun
KenkyË 14, pp. 1–13
——— 1987B. “Chronology of the S¥ha¬a Sources for the Påli Commentaries
I”, Buddhist Studies 16, pp. 151–82. (= SPCE)
——— 1987C. “A Study of the S¥ha¬avatthuppakaraâ”, JØsai Jinbun KenkyË
15–2, pp. 1–23
——— 1987D. “Some Minor Sources for the Påli A††hakathås´: With Reference
to Lottermoser’s Study”, Indology and Buddhology´: Essays in Honor of Dr.
Jikido Takasaki on His Sixtieth Birthday (Indogaku BukkyØgaku RonshË),
Tokyo´: ShunjËsha (= SPCE)
——— 1987E. “Mahås¥vatthera as Seen in the Påli A††hakathås (revised)”, Sri
Lanka Journal of Buddhist Studies 1, pp. 117–27 (= SPCE)
——— 1987 F . “The Chronology of the S¥ha¬avatthuppakaraâ”, Bulletin
d’Études Indiennes 5, pp. 221–50 (= SPCE)
——— 1988A. “Chronology of the S¥ha¬a Sources for the Påli Commentaries
II”, Buddhist Studies 17, pp. 119–67.(=SPCE)
——— 1988 B . “S¥ha¬avatthuppakaraâ and Påli A††hakathå Literature”,
Påligaku BukkyØ Bunkagaku 1, pp. 47–72 (= SPCE)
——— 1988 C . “Uttaravihåra††hakathå and Sårasamåsa´: Some Unattributed
Non-Mahåvihåravåsin Sources for the Påli Commentaries”, Journal of the
Pali Text Society XII, pp. 1–47 (= SPCE)
——— 1989A. “Ariyavasa and Ariyavasakathå”, JØsai Daigaku KenkyË
NenpØ´: Jinbun Shakai Kagaku-hen 12, pp. 1–12 (= SPCE)
190 Sodo Mori
——— 1989B. “The Value of the Påli Commentaries as Research Material”,
JØsai Jinbun KenkyË 17–1, pp. 1–18. (= SPCE)
——— 1989C. Studies of the Påli Commentaries´: A Provisional Collection of
Articles, Niiza´: author
——— 1990. “The Origin and the History of the Bhåâka Tradition”, Ananda´:
Essays in Honour of Ananda W.P. Guruge, Colombo ´: The Felicitation
Volume Editorial Committee, pp. 123–29
——— 1991A. “Types of the Påli Commentarial Literature and Their Value as
Research Material”, Buddhist Studies 20, pp. 127–59
——— 1991 B . “Some Authorities as Sources for the Påli A††hakathås”.
Ótmajñåna´: Felicitation Volume Presented to Professor Sengaku Mayeda
on the Occasion of His Sixtieth Birthday (“Ga” no ShisØ), Tokyo´:
ShunjËsha, pp.´735–50
———, et al., 1994. Påli A††hakathå Correspondence Table, Oxford´: Pali Text
Society
——— 1997. “The Vijñånavådin View as Depicted in the Påli Commentaries´:
With Special Reference to the Nirayapåla-kathå”, Indica et Tibetica 30´:
Bauddhavidyåsudhåkara ´: Studies in Honour of Heinz Behert on the
Occasion of his 65th Birthday, Swisttal-Odendorf, Germany´: Indica et
Tibetica Verlag, pp. 453–64
——— 1998. “The Milindapañha and the Påli A††hakathå Literature”,
Indologica Taurinensia´: Official Organ of the International Association of
Sanskrit Studies, Vols. 23–24, Professor Gregory M. Bongard-Levin
Felicitation Volume (Torino), pp. 291–312
——— 1999. Mahåyåna Buddhism in Sri Lanka. Aichi ´: Aichi Gakuin
University
On Mah!y!nas"tr!laµk!ra VII.1*
1. The seventh chapter of the Mah!y!nas"tr!laµk!ra (edited by Sylvain

Lévi, Paris 1907, pp. 25–27) deals with the concept of supranormal
power (prabh!va) and thus corresponds to the fifth chapter of (the first
section of) the Bodhisattvabh"mi. Its structure is based on a kind of
standard pattern of six categories, viz. lak!a"a (or svabh!va, cf. 25,11),
hetu, phala, karman, yoga, and vÁtti,1 followed by a concluding verse in
the pu!pit!gr! metre extolling the greatness (m!h!tmya) of the
prabh!va of bodhisattvas.
2. The first verse (indravajr! metre : ¯ ¯ ˘ ¯ ¯ ˘ ˘ ¯ ˘ ¯ ¿ ), describing the
essential characteristic (lak!a"a) of the supranormal power of bodhi-
sattvas, runs thus :
utpattiv!kcitta#ubh!#ubh!dhitatsth!nani$s!rapad!parok!aµ |2
jñ!naµ hi sarvatragasaprabhede!v avy!hataµ dh¥ragata$ prabh!va$ ||
VII.1
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
*I am deeply indebted to Professor Y"sh# Wakahara and the Mah!y!nas"tr!-
laµk!ra study group at Ry"koku University for sharing their materials with
me (especially for a CD containing mss N2, N3 and NS, for which see n. 12),
and to Professor Oskar von Hinüber for valuable suggestions.!
1For this pattern, cf., e.g., also Mah!y!nas"tr!laµk!ra(-Bh!$ya) IX.56–59
(verse 57b should probably be read as °bh!van!-samud!gama$, and 57d

emended to °sarvath!-’k%ayat!-phala$!: bahuvr¥his as in 56, 58ab and 59ab) ;
Ratnagotravibh!ga (ed. E.H. Johnston, Patna, 1950) I.30, 35, 42, and 45 as
well as II.3, 8–9, 18–20, 29, and 38–41 (and the prose lines introducing these
verses ; read °phala$ in I.35 and °tathat!bhinnavÁttita$ or even °ka$ in I.45
[cf. WZKS 15/197, p. 147], and perhaps °yukta$ sva° in I.42) ; Yog!c!rabh"mi,
%rutamay¥ Bh"mi (T 30.1579) 361a17–20 (%r!vakabh"mi ms fol. 23a8–b1) ;
Abhidharmasamuccaya (ed. P. Pradhan, Santiniketan 1950) 103,1–8 (recon-
structed, but terminology confirmed by Abhidharmasamuccayabh!$ya (ed.
N. Tatia, Patna, 1976), p. 141).
2The edition by S. Bagchi (Darbhanga, 1970) reads °!dhi tat…pad! parok%am,
which does not make sense.
The!Journal!of!the!Pali!Text!Society,!Vol.!XXIX!(2007),!pp.!191–200!
192! Lambert Schmithausen! !
Sylvain Lévi3 translates :
La connaissance qui n’a pas en dehors de sa portée les Points
suivants : naissance, langage, pensée, dépôt de bien et de mal,
situation, Évasion, avec leurs subdivisions, qui est universelle, sans
entrave, c’est là le Pouvoir qui appartient au Sage.
In the translation edited by Robert Thurman,4 the verse runs as follows :
Direct knowledge of birth, speech, mind, the deposit of good and evil,
place, and escape is unobstructed toward these everywhere with all
varieties ; and it is the power of the brave.
3. None of the translators5 indicates any difficulties he may have had

with the syntax of the text, but it is obvious that sarvatraga° does not
construe well. Lévi translates it as an attribute of jñ!naµ, but this is
impossible if we keep to the printed text where it is compounded with
saprabhede!v. Nor is Thurman’s rendering convincing since it ignores
°ga° and translates as if there were only sarvatra, as a separate word, as
in the commentary (sarvatra lokadh!tau saprabhede!u …). But the
omission of °ga° in the verse would spoil the metre. For the same
reason, a reading sarvatragaµ, which would fit in with Lévi’s trans-
lation and make good sense, is excluded as well.
4. Now, there is a similar case at Mah!y!nas"tr!laµk!ra IX.9a (metre
m!lin¥ : ˘ ˘ ˘ ˘ ˘ ˘ ¯ ¯ ¯ ˘ ¯ ¯ ˘ ¯ ¯ ) : !
#ara"am anupamaµ tac chre!&habuddhatvam i!&aµ …

Thurman (p. 78) translates :
Supreme Buddhahood is the refuge without compare. ...

Yet, a karmadh!raya #re!&habuddhatva would seem to indicate a
specific form of Buddhahood that is superior to another one (e.g., better
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
3Mah!y!na-s"tr!laµk!ra, edited and translated by Sylvain Lévi, Vol. II (Paris,
1911), p. 55.
4Maitreyan!tha’s Ornament of the Scriptures of the Universal Vehicle,
Recorded by 'ry!sa(ga, Explained by Vasubandhu, English translation by

Lobsang Jamspal et al., edited by Robert A. F. Thurman (American Institute of
Buddhist Studies, 1979), p. 57.
5The unpublished translation by Peter Oldmeadow (Canberra, mentioned by
J.W. de Jong in IIJ 30 (1987), pp. 154 ff.) remains inaccessible.

! On Mah!y!nas)tr!laµk!ra VII.1 193!
than pratyekabuddhatva).6 This, however, would be somewhat surpris-
ing since in the preceding as well as in the following verses buddhatva
(and also buddhat!) is, without any qualification, consistently used for
Buddhahood proper, i.e., the state of highest, perfect Awakening.
Actually, in the following verse (IX.10) #re!&ha clearly qualifies not
buddhatva but #ara"a,7 and the same is true of a quasi-synonym of #re-
!&ha, viz. uttama, in the preceding pair of verses (IX.7–8)8 of which
IX.9 is a more artistic rephrasing.9 In both cases, the purport is that
Buddhahood is the supreme refuge. This doubtless makes better sense.
At any rate, the commentary on verse IX.9 does construe #re!&ha with
#ara"a and not with buddhatva :
By this third [verse the author] shows that … precisely this refuge
status [of Buddhahood] is unsurpassed because it (= the refuge status
of Buddhahood) is incomparable and supreme (tasyaiva #ara"a-
tvasyânupama#re!&hatvenânuttaryaµ 10 … dar#ayati).
Accordingly, Lévi (p. 71) translates the verse as follows :
Cette Bouddhaté est le Refuge excellent, incomparable. …
The Tibetan translation, too, supports this interpretation :
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
6Cf. XIX.62d bodhi$ #re%&h! (Bh!$ya!: #re%&h! bodhi$), but in contrast to
buddhatva the term bodhi is traditionally applied to %r!vakas (and Pratyeka-
buddhas) as well and hence requires specification when referring to a Buddha,
i.e., when used in the sense of anuttar! samyaksaµbodhi.
7“[T]his Buddhahood is regarded here as the best of [all] refuges” (… tad
buddhatvaµ #re%&ham ihe%&aµ #ara"!n!µ ; text follows Naoya Funahashi,

Mah!y!nas)tr!laµk!ra [Chapters I, II, III, IX, X] (Tokyo!: Kokushokankokai,
1985), p. 27).
8“Buddhahood protects from … ; therefore, [it] is the best refuge” ( paritr!"aµ
hi buddhatvaµ … tasm!c chara"am uttamaµ).

9The same pattern is also found in the preceding verses, the anu%&ubh lines
IX.1–2 and IX.4–5 being rephrased by IX.3 (#!rd)lavikr¥*ita) and IX.6 (srag-
dhar!), respectively. This pattern is, by the way, also found in the poetical
rephrasing of the Tath!gatagarbhas"tra at Ratnagotravibh!ga I, 96–126.
10Thus to be read with Tibetan mchog nyid kyis, against Lévi’s °#re%&hasya
cânu°. Among the mss accessible to me (see n. 12), mss B, N2, N3 and NS
read °%&hasvan!nu°, whereas ms A has °%&a·svan!nu°, with a dot between %&a
(sic) and sva. A misreading of tve as sa seems quite possible from a script
where the e-sign is a downward hook on the upper left side of the ak$ara. See
A 34b2 ; B 36b5 ; N2 37b4, N3 29b7 ; NS 31a6.
!
This Buddhahood is considered to be the incomparable, supreme
refuge (Peking Phi 10a1-2 : sangs rgyas nyid de skyabs ni dpe med
mchog tu ’dod).11
However, such an interpretation is clearly impossible if #re!&habuddha-
tvaµ is read as a compound. To conjecture a reading #re!&haµ is out of
the question because it would violate the metre. The only way out of the
difficulty occurring to me is to suggest that we should probably separate
#re!&ha from buddhatvaµ and take it as a BHS form of the nom. sg.
neuter (cf. F. Edgerton, BHSG § 8.31–34). Possibly what the mss12 write
as °a was actually pronounced °ã,13 i.e., a short nasal for which the
Br!hm¥ script has no sign, so that the scribes had only two options :
either to indicate the nasalization by means of an anusv!ra to the effect
of obscuring the metre, or to give precedence to the metre and leave the
nasalization unexpressed (as they actually do).14 If my argument is
correct, the line should be read (and was at any rate read by the
commentary) as
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
11Likewise the Tibetan translation of the p!da in Sthiramati’s commentary
(P Mi 125b6–7) : sangs rgyas skyabs ni dpe med mchog yin te. Cf. also P Mi
125b5 (skyabs de nyid dpe med pa dang / mchog tu gyur pa’i phyir) and 126a1
(skyabs ’di dam pa yin pas na mchog ces bya ste). The Chinese translation,
too, seems to take #re%&ha with #ara"a but construes anupama with Buddha-
(hood) when paraphrasing the commentary!: “Verse!: The Buddha is the
supreme refuge ; because [he] is incomparable, [it ?] is unsurpassed. …
Commentary!: This verse elucidates the supremeness of refuge. Because the
Buddha is incomparable, [as a refuge he] is unsurpassed” (T 31.1604: 602c4
and 6!: 偈曰焎佛為勝歸處無比故無上 ... 釋曰焎此偈顯歸依勝焎由佛無譬
喻故為無上焎).
12Five mss are accessible to me, viz. mss A and B published in Syôkô Takeuchi
et al. 1995 and mss N2 (NGMPP E 1923/5), N3 (NGMPP E 1367/11) and NS

(NGMPP A 114/1). According to Wakahara 2003, p. (34), NS is dated Nepal
saµvat 796 (= 1675/6 C.E.), N2 Vikrama saµvat 1957 (= 1900 C.E), and N3
Nepal saµvat 1025 (= 1904/5 C.E). Cf. also Wakahara’s articles in Journal of
Indian and Buddhist Studies 51.2 (2003), pp. (157)–(163) and 52.2 (2004),
pp. (157)–(162).
13For cases of aµ to be read as ã for metrical reasons in P!li verses, cf. Alsdorf
1967, p. 17, verses 7c = Sn 921c pa&ipadã (but cf. Norman 1992, 342 !) and
16b = Sn 930b payuttã ; p. 26 (J!taka no. 479) verse 2b K!li(gã ; p. 29
(J!taka no. 485) verse 6a imã mayhã ; etc.
14All mss available to me read °a, as does S. Lévi’s edition. See A 34b1 ; B
36b3 ; N2 37b2 ; N3 29b6 ; NS 31a5.

#ara"am anupamaµ tac chre!&ha (or chre!&hã) buddhatvam i!&aµ …
5. Returning now to the verse VII.1, an analogous solution would seem

to be possible : the difficulty sketched in § 3 would vanish if we assume
that the mss15 should be transcribed not as sarvatragasaprabhede!v but
as sarvatraga saprabhede!v, perhaps to be read as sarvatragã sapra-
bhede!v. In this case, it is no longer problematic to construe sarvatraga
(or °gã) as an attribute with jñ!nam, and Lévi’s translation can, in this
point, be accepted.
6. Still, there is yet another problem. At the end of the first line, Lévi’s
text reads °ni$s!rapad!parok!aµ, and in his translation he takes pada
to mean “points”, referring to the six items enumerated before with
regard to which the cognition of bodhisattvas is immediate or
perceptual (aparok!a). In the verse, this works fairly well. In the
commentary, however, such an understanding of pada appears to be
precluded. There, after the explanation of the six items we have the
following sentence :
e!u !a&sv arthe!u sarvatra lokadh!tau saprabhede!u pad!parok!am
avy!hataµ jñ!naµ sa prabh!vo bodhisattv!n!µ ... |.
This is translated by S. Lévi (p. 55) as follows :
Voilà les six catégories en question ; la connaissance qui porte sur
elles sans que nulle part, dans tous les mondes, avec toutes leurs
subdivisions, elles soient en dehors de sa portée, sans rien qui
l’entrave, c’est là le Pouvoir des Bodhisattvas. …
I cannot find an equivalent for pada in this translation, nor in that of
Thurman 16 who ignores it also in his translation of the verse (see § 2).
Actually, in the commentary I find it altogether impossible to construe
pada° as a prior member of a compound ending in °aparok!a, let alone
in any other way. Thurman may well have ignored it because he could
not find an equivalent in the Tibetan translation. But a closer inspection
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
15All the five mss at my disposal (see n. 12) read sarvatragasa°. See A 25b4 ; B
27a6 ; N2 27b6 ; N3 22a1 ; NS 23a3.
16“Such knowledge is directly present without impediment in all universes as
regards those six topics and their varieties, this knowledge is the bodhisattva’s
power ...” (Thurman [see n. 4] p. 57).
!
of the latter does show the way towards a reasonable solution, and
moreover suggests a different reading of the verse as well.
7. The Tibetan translation of the commentary passage runs like this :
What is, in this way, a direct, unobstructed cognition with regard to
these six items including their subdivisions in every world-system,
that is the [supranormal] power of bodhisattvas (P Phi 156b2–3 ; D Phi
147a3–4 : de ltar na ’jig rten gyi17 khams thams cad du don drug po
de dag rab tu dbye ba dang bcas pa la shes pa mngon sum du gyur pa
thogs pa med pa gang yin pa de ni byang chub sems dpa’ rnams kyi
mthu … yin no //).
It is obvious that the only word which has no equivalent in the Sanskrit
text as printed by Lévi is the relative pronoun gang ( yin pa),18 which in
connection which the subject jñ!nam would correspond to yad. Since
the ak!aras ya and pa are very similar in the mss, the conclusion
suggesting itself is that the disturbing pad!° is nothing but a misreading
of the relative pronoun yad followed by !parok!aµ or rather aparo-
k!aµ, at least according to the mss available to me.19 But even a reading
!parok!aµ could easily be explained as a metrical lengthening taken
over from the verse. For there, too, Tibetan, reading as it does, for p!da
b,
… de yi gnas dang ’byung ba mngon sum gang |,
shows that pad!parok!aµ is rather a miscopying of yad!parok!aµ, to

be resolved into yad !parok!aµ.20 This is anyway what one would
expect in view of Vasubandhu’s commentary, unless we suspect him of
having grossly misread the verse.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
17gyiD : gyis P.
18This reading is also confirmed by the prat¥ka in Sthiramati’s commentary (P
Mi 95b1!: mngon sum gang zhes bya ba ni …).
19Mss A and N2 pedapa°, ms N3 padapa°. But ms B clearly reads yadapa°, and
ms NS either yadapa° or yadaya°. See A 25b7 ; B 27b2 ; N2 28a2 ; N3 22a4 ;

NS 23a5.
20Mss A, N2 and N3 pad!parok%aµ ; ms B pad!rok%aµ, like ms NS where °d!
and ro° are, however, separated by a mark indicating the end of the preceding
chapter in the preceding line but extending into the line below. See A 25b4 ; B
27a6 ; N2 27b5–6 ; N3 22a1 ; NS 23a2.
8. However, if this is correct (and I fail to see how at least in Vasu-
bandhu’s commentary a reading pad!° or even pada° could be justified
syntactically), there arises another problem : how to construe the
compound immediately preceding the relative pronoun in the verse ? If
°ni$s!rapad!parok!aµ is, following the Tibetan and Vasubandhu’s
commentary, emended to °ni$s!ra yad !parok!aµ, the compound
preceding the relative pronoun would end with a stem form, which is of
course impossible in standard Sanskrit. What is required is rather a
locative dependent on jñ!nam, as is confirmed by the commentary
explicitly construing the six items of the first line as locatives (vi!aya-
saptam¥) depending on jñ!nam (viz. upapattau jñ!naµ, v!ci jñ!naµ,
citte jñ!naµ, °!dh!ne jñ!naµ, and ni$sara"e21 jñ!naµ). But emending
°ni$s!ra yad to °ni$s!re yad is, once again, precluded by the metre
requiring a short syllable.22 In view of the solution found for sarva-
traga, I suggest to interpret °ni$s!ra, in a similar way, as a BHS form
of the loc. sg.23 (BHSG § 8.11). I wonder if in this case a may not be
interpreted as a substitute writing for +, for which, once again, no sign is
available in the scripts derived from the Br!hm¥ alphabet. Among the
two available possibilities, viz. to write either e (normally long) or a
somehow similar short vowel like a or i,24 the metrically required
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
21But all the five mss at my disposal read ni$sara"ajñ!naµ ! In the preceding
item, mss B and NS read °gamanajñ!namÁddhi°, but mss A, N2 and N3 have
°gamanaÁddhi° (omitting °jñ!na°). See A 25b6 ; B 27a9 ; N2 28a1 ; N3 22a3 ;
NS 23a3.
22The reading °s!ra is confirmed by all the five mss available to me. See A
25b4 ; B 27a6 ; N2 27b5 ; N3 22a1 ; NS 23a2.

23Cf. Mah!y!nas"tr!laµk!ra XXI.48c, where ni$s!re occurs in a similar
context (six abhijñ!s), albeit as the object not of jñ!na but of avav!da.
24It has to be conceded that a for + is not usual, the normal representation pre-
serving the quantity being i (cf. Edgerton 1946, pp. 199 § 28 and 204 § 67 ; cf.
also, for Apabhraµ&a, Ludwig Alsdorf, Harivaµ#apur!"a (Hamburg, 1936)
[Alt- und Neu-Indische Studien Bd. 5], pp. 142–44). However, in the
analogous case of shortened o (i.e.!: ,), both u and a are attested (Edgerton
1946, pp. 199 § 28 and 204 § 68).
!
quantity of the vowel would, in this case too, have taken precedence
over the quality.
9. It has, however, to be admitted that the interpretation of °ni$s!ra in
the first line as a locative singular is odd in view of the fact that we
have, in the second line, the locative plural saprabhede!u, an adjective
which doubtless qualifies the six items to be supplied from the first line
(cf. the Bh!!ya : e!u !a&sv arthe!u … saprabhede!u). But since the
singular in the first line is collective (six items !), a reference to it in the
form of a plural ad sensum would not seem to be entirely inexplicable,
still less so in view of the constraints of the metre. Anyway, the only
alternative solution I for my part could imagine would be to interpret
the six-membered dvandva ending with °ni$s!ra as a virtual locative
plural, to be connected with jñ!naµ as a kind of split compound,
interrupted by yad !parok!aµ ; but I am unable to decide whether such
a construction is possible at all.25
10. My translation of the verse does not differ too much from S. Lévi’s :
A knowledge which is perceptual with regard to [the dying and
re]birth [of beings], to speech [even in other realms of existence],26 to
the thoughts [of others], to the deposit27 of good and bad [karma], to
[how to go to] the place where the [vineyas dwell],28 and to [the
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!
25Anyway, a similar case seems to occur at J!taka IV 384,14, where the metre
requires the reading app-eva naµ putta- labhemu -j¥vitaµ, on which cf.
Oberlies 1996, 119 (“compound in tmesis”).
26In the Bh!$ya (25,5–6) we should read, with ms A, °bhijñ! y!µ v!caµ tatra
tatropapann! bh!%ante ; cf. mss A and N2 v!cantatratatropa°, ms N3 v!caµ

tatratatropa° and Tibetan (Peking Phi 156a8) de dang de dag tu skyes pa
rnams. Mss B and NS read v!cantatragatropa°. In ms NS, some forms of ta
are not much different from ga. See A 25b5 ; B 27a7 ; N2 27b7 ; N3 22a2 ; NS
23a3.
27i.e. the residues (v!san!) accumulated in previous existences (cf. Sthiramati,
Peking Mi 95a3–5).
28For want of anything better, my interpretation of the telegraphic tat- (there is
no word in the line it might refer to) follows the Bh!$ya (25,7–8!: yatra viney!s
ti%&hanti tatsth!nagamanajñ!naµ Áddhivi%ay!bhijñ!) and Bh!$ya ad XX-
XXI.48 (185,13–14!: upetya vineyasak!#am Áddhyabhijñay!). According to
Sthiramati (Peking Mi 95a6–8), tatsth!na means the Buddha-fields where the
means for] escaping [from saµs!ra], and which is universal and
unobstructed with regard to [its aforementioned objects] along with
their subdivisions : [this] is the [supranormal] power of the bodhi-
sattvas.
11. The grammatical explanation of the verses VII.1 and IX.9a pro-
posed in the preceding paragraphs presupposes that the language of the
Mah!y!nas"tr!laµk!ra allows for non-standard grammatical features,
especially such as are known from Middle Indic, as in many other
Buddhist Sanskrit texts. Such features have indeed already been
registered by S. Lévi, e.g. in X.14 (janiya), XVII.14 (bahitas), 31
(t!yaka), 45 (arihat), or XIX.69–70 (dharama). Cf. also v! for iva at
IX.36. The most interesting case in connection with the present
investigation is the shortening of a long vowel at the end of a word at
XIX.75, where we find hetuna m.c. for hetun!. There is also a number
of non-standard compounds (cf. F. Edgerton, BHSG § 23.10) which
would deserve special investigation, especially at the beginning of
Chapter IX, and significantly in verses composed in fairly demanding
metres, viz. 3d : (ratn!n!m) prabh!va-mahat!m ;29 6b : dharma-ratna-
pratata-sumahata$ (Bh!!ya : sumahata$ pratatasya dharmaratnasya !) ;
6c : #ukla-sasya-prasava-sumahata$ ;30 6d : dharmâmbu-var!a-pratata-
suvihitasya (Bh!!ya : mahata$ suvihitasya … dharmâmbu-var!asya) ;
12d : vi!aya-sumahata$31 (jñ!nam!rg!t).
Lambert Schmithausen
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!! !

Buddhas live and the world systems where sentient beings live, and the
cognition referring to these means the knowledge how to go there by means of
supranormal accomplishments (Áddhi).
29Looks like a bahuvr¥hi with its members inverted (cf. Oberlies 1996, p. 119
[see n. 25] ; Oberlies 1989–1990, pp. 159–60, n. 7 ; Oberlies 2001, p. 123 ;

Norman 1992, p. 217 ad v. 370). But the compound could perhaps also be
understood as a tatpuru$a in the sense of “great as regards their power”.
30Probably in the sense of sumahata$ #uklasasyaprasavasya.
31See n. 29.
!
BIBLIOGRAPHY
Alsdorf, Ludwig, 1967. Die 'rya-Strophen des Pali-Kanons, Wiesbaden : Franz
Steiner
Edgerton, Franklin, 1946. “Meter, Phonology, and Orthography in Buddhist
Hybrid Sanskrit”, JAOS 66, pp. 197–206
Norman, K.R., 1992. The Group of Discourses, Oxford : PTS
Oberlies, Thomas, 1989–1990. “Miscellanea Palica I”, Bulletin d’études
indiennes 7–8, pp. 157–84
——— 1996, “Stray Remarks on Pali Phonology, Morphology, and Vocabu-
lary”, Münchener Studien zur Sprachwissenschaft 56, pp. 91–130
——— 2001, P!li : A Grammar of the Language of the Therav!da Tipi&aka,
Berlin : de Gruyter
Syôkô Takeuchi et al., eds., 1995. Sanskrit Manuscripts of the Mah!y!nas)tr!-
laµk!ra from Nepal, Kyoto : Ryukoku University
Wakahara, Y"sh#, 2003. “Remarks on Sanskrit Manuscripts in the Otani
Collection”, Bulletin of Institute of Buddhist Cultural Studies, Ryukoku
University 42
Sanskrit Ik!v!ku, P!li Okk!ka, and G!ndh!r" I!maho*
1. G!ndh!r" i!maho = Sanskrit ik!v!ku

Until now, the G!ndh!r" word i!maho has been known only from the
st!pa dedication inscription of Se#avarma (Bailey 1980, Fussman 1982,
Salomon 1986, von Hinüber 2003). This important document, written
on a gold leaf and dating from the early first century A.D ., is the longest
single inscriptional text known in G!ndh!r" language and Kharo!$h"
script. The word in question occurs three times in Se#avarma’s
inscription :
line 3a : utarase"aputre vasuse"e o#iraya i!mahokulade, “Vasuse#a, son
of Utarase#a, King of O%i, from the I!maho family”.
line 3c : se"avarme ayidase"aputre ate ceva i!mahorajakulasabhavade
o#iraja, “Se#avarma, son of Ayidase#a, and therefore, by virtue of
birth in the I!maho royal family, king of O%i”.
line 9e : bhadase"a raya upadae yava pravidamaha me di$ase"o o#iraya
sarva i(*!ma)horayakulasambhavo,1 “from King Bhadase#a up to
my paternal great-grandfather Di&ase#a, the kings of O%i, all born
in the I(*!ma)ho royal family”.
The word i!maho, whose meaning and etymology have been up to now
completely obscure, has usually been assumed to be a non-Indian name.
Thus, for example, Fussman (1982, p. 44) commented, “Ce mot semble
un nom propre, d’origine non-indienne”, and von Hinüber (2003, p. 34,
*We wish to express our gratitude to W. South Coblin (Iowa City), Max Deeg
(Cardiff), Zev Handel (Seattle), Oskar von Hinüber (Freiburg), Timothy Lenz
(Seattle), and Gary Tubb (Chicago), who provided assistance and advice in the
preparation of this paper. Tien-chang Shih (Seattle) in particular generously
assisted us in locating and interpreting relevant Chinese materials. Finally, we
thank the honoree of this volume, K.R. Norman, for encouraging, assisting and
inspiring us over many years and in many ways in our studies of areas of
common interest.
1Here the syllable !ma was apparently omitted by scribal error, as the normal
spelling is confirmed by the two other occurrences of the word in this

inscription. This error presumably does not have any linguistic significance.
202 Richard Salomon and Stefan Baums
n. 30) similarly remarked “Unarisch scheint der Name des Stammvaters
I!maho zu sein”.
Now, however, i!maho has been observed in another G!ndh!r" text
in a context which makes it clear that this name is not in fact non-
Indian, but rather is the G!ndh!r" equivalent of the name of the
renowned legendary king known in Sanskrit as Ik!v!ku and in P!li as
Okk!ka. The text in question is a Buddhist birch-bark scroll in G!ndh!r"
language and Kharo!$h" script in the Library of Congress (Washington,
D.C.), which appears to date from about the second century A.D. (figs.
1–2). This manuscript, which is only now beginning to be studied,
appears to consist of formulaic accounts of the lives of fifteen Buddhas,
from D"pa'kara to Maitreya, enumerating for each Buddha the kalpa in
which he lived, his life-span, his class (br%hma"a or k!atriya), the size
of his assembly (sa&nip%ta), the duration of his dharma, etc. Thus in its
format and contents this new text resembles biographical texts such as
the Mah!pad!na-sutta / Mah!vad!na-s(tra, Buddhava)sa, and Bhadra-
kalpika-s(tra, but it seems to have a particularly close similarity to
portions of the Bahubuddha-s(tra contained in the Mah!vastu (ed.
Senart, III 224.10–250.8).
The portion of the new text described above is preceded by a set of
fifteen verses containing a prediction (vy%kara"a) of the future Buddha-
hood of *!kyamuni, which are presumably being spoken by a previous
Buddha. The passage in question here is part of what appears to be the
third verse in this series. The surviving portion of the verse, comprising
part of the second and fourth quarters and all of the third, reads as
follows :
+ + + (*ka)///[p](*e) ido asakhae ·
i!mahovat&a#ara&akasiho ·
tari&asi devama#u[&a] ? /// +
Sanskrit Ik!v!ku, P%li Okk!ka, and G%ndh%r' I!maho 203
[An] incalculable world-age from now,2as the *!kya man-lion in the
I!maho lineage,3 you will cross over … gods and humans.4
2Cf. Mvu I 53.2, kalpasmi& ito asa&khyeye.

3The sense of this line is not completely certain. We propose to read the entire
quarter as a single compound, i!maho-vat$a-"ara-$aka-siho, although super-
ficially it might seem easier to divide it into two words, i!mahovat$a"ara
$akasiho, and translate “as a man of the I!maho lineage, the Lion of the
*!kyas”. But we provisionally reject this interpretation, mainly because -"ara
at the end of a compound i!mahovat$a"ara would be superfluous and
stylistically weak. We suspect that "ara$akasiho should rather be read as a
sub-compound, by way of a conflation of the two expressions $akasiho (= Skt
$%kyasi&ha) and "arasiho (= narasi&ha). Narasi&ha and equivalent epithets
of the Buddha such as puru!asi&ha and puru!avy%ghra, though rare in P!li,
are common in some Buddhist Sanskrit texts, especially in the Mah!vastu,
with which the new G!ndh!r" text under discussion here has many common
features of style and contents. For example, in narasi&hat%ye pra"idheti, “He
makes a vow to attain the state of a man-lion”, that is, “of Buddhahood” (Mvu
I 83.8), narasi&ha is used in a context of predictions of future Buddhahood, as
in our text. Similarly, the synonymous puru!asi&ho occurs in a context similar
to that of the passage in question in puru!asi&ho $%kyakul%nandajanano (Mvu
II 164.13).
But it must be conceded that in the proposed interpretation the construction
is still somewhat odd, with the sub-compound -"ara$akasiho instead of the
expected -$aka"arasiho. However, compounds with irregular word order are
not unknown in Buddhist usage (see Edgerton 1953, §23.10), and in this case
the peculiarity could be explained on metrical grounds, since the irregular
ordering of the words in -"ara$akasiho provides a normal ending for a tri!(ubh
line ˘ ˘ µ ˘ µ µ), whereas the normal compound order $aka"arasiho (µ ˘ ˘ ˘ µ
µ) would not fit the metre. Although ideal metrical patterns are often treated
rather loosely in G!ndh!r" texts (see, for example, Salomon 2000 : 49–51), a
preliminary analysis of the new text in question here seems to show that it
followed the standard metrical pattern of the tri!(ubh metre much more closely
than many other G!ndh!r" texts, perhaps because the text was originally
composed in G!ndh!r" rather than translated into G!ndh!r" from some other
Indo-Aryan language. For this reason, we take the metre of this text to be
phonetically and etymologically reliable, although we would not necessarily
do so for all G!ndh!r" texts.
4Possible reconstructions of the last quarter of this verse include deva-
ma"u$a[lo](*ga) (compare Buddhava)sa 2.55, sabbaññuta& p%pu"itv%

sant%ressa& sadevake) or devama"u$a[$a](*sta) (compare, e.g., Mvu I 239.9,
$%st% dev%n%& ca manu!y%"%& ca).
The key phrase for our purposes is the second quarter, i!maho-
vat$a"ara$akasiho. The reference to $akasiho = Skt $%kyasi&ha makes
it certain that the addressee here is indeed the (then) future “historical”
Buddha *!kyamuni, while the phrase i!mahovat$a indicates that he is
being associated with the I!maho lineage (va&$a).5 Since the *!kyas are
universally deemed in Buddhist tradition to be descended from the
lineage of the legendary cakravartin emperor Ik!v!ku, there can hardly
be any doubt that i!maho here is the equivalent of Sanskrit ik!v%ku,
despite the several unusual phonetic correspondences between the two
— correspondences which, however, are no more unusual, indeed
somewhat less so, than those between Skt ik!v%ku and P!li okk%ka, as
will be discussed in detail below (section 2).
Moreover, the association of the descendants of Ik!v!ku with the
*!kyas is expressed in similar terms to those of the new text in, for
example, Mah!vastu III 247.12–13, $uddhodanasya r%jño ik!v%kujasya
putro m%y%ya $%kyakulanandijanano $%kyo bh!t $%kyasukum%ro, “King
*uddhodana, the descendant of Ik!v!ku, had with M!y! a son, the
*!kya who brought delight to the *!kya clan, the tender *!kya youth”.
Similarly, the expression applied to the Buddha in Mah!vastu III
343.15, ik!v%kukulasa&bhave, “born in the Ik!v!ku clan”, is virtually
identical to i!mahorajakulasabhavade, the epithet adopted by Se#a-
varma in his inscription (line 3c). These parallels thus confirm that
G!ndh!r" i!maho does in fact correspond to Skt ik!v%ku / P!li okk%ka.
Though not previously attested as such, vat$a in the compound
i!maho-vat$a-"ara is a more or less normal G!ndh!r" correspondent to
Skt va&$a. Here the t has arisen as an excrescent consonant between the
underlying nasal (here left unwritten, as very often in G!ndh!r") and the
following sibilant : va&$a ([vã+a] or [va,+a]) > vat$a ([vant+a]). Parallel
developments (though involving the dental rather than the palatal
sibilant) are attested, for instance, in the G!ndh!r" Dharmapada from
Khotan, in matsa = Skt m%&sa and satsara = sa&s%ra (Brough 1962,
5The equivalence of G!ndh!r" vat$a with Sanskrit va&$a will be explained

below.
pp. 73–74 ; additional examples from Central Asian G!ndh!r" provided
in Burrow 1937, p. 19).
The ligature representing the consonant cluster in question in our
text, !, is nowadays usually transliterated as tsa, though t$a has also
been used for it. On purely visual grounds, it is difficult to distinguish
whether the second member is " $a or a simplified form of # sa. This
issue was discussed at some length by Brough (1962, pp. 73–77),6 who
preferred the transliteration tsa on both graphic and phonetic grounds,
since most of the examples available to him, such as the aforementioned
matsa = m%&sa and satsara = sa&s%ra, involved original dental sibi-
lants. However, the present case of vat$a = va&$a revives the question
of the correct transliteration, or perhaps rather transliterations, of !, and
suggests that it perhaps did double duty for both ts and t$. Whether this
represents an actual merger of the two, either in the writing system or in
the phonology of the language, is difficult to determine on the basis of
the data currently available. Although in general the three sibilants of
Old Indo-Aryan are retained as such in G!ndh!r", they tend to merge or
alternate graphically, if not phonetically, in consonant clusters ; for
example, the absolutive corresponding to Sanskrit dÁ!(v% is written in
different texts as di!pa and dhri$pa"a, and also, possibly, as dispa.7 In
any case, the equation between Sanskrit va&$a and G!ndh!r" vat$a is
supported on contextual grounds by a passage in A&vagho!a’s
Saundarananda (ed. Johnston, 1.24) : tasm%d ik!v%kuva&$y%s te bhuvi
$%ky% iti smÁt%), “Therefore those members of the lineage of Ik!v!ku
are known in the world as *!kyas”. Here the compound ik!v%ku-
va&$y%s mirrors i!maho-vat$a- in our new manuscript.
6See also the further discussion in Glass 2000, pp. 130–31.

7The last reading is however uncertain and largely reconstructed ; see Salomon
2000 : 143–44 and Allon 2001 : 93. For other citations, refer to the Early
Buddhist Manuscripts Project’s online G!ndh!r" dictionary (http://
depts.washington.edu/ebmp/dictionary.php).
2. Etymological problems
Thus there can be no reasonable doubt that i!maho is the G!ndh!r"
equivalent of Sanskrit ik!v%ku and P!li okk%ka. Although the form
i!maho cited here is a transliteration from Kharo!$h" script, in which
vowel quantity is not distinguished, we can safely assume that the
vowel of the second syllable was long. This is suggested first of all by
the corresponding long vowel of the Sanskrit and P!li forms, although
this alone is not conclusive in light of the several other problems in the
phonetic correspondences between these three words (as discussed
below). But it is confirmed by metrical considerations,8 since the word
in question appears at the beginning of a tri!(ubh line, where the
expected metrical pattern would be – – !.
Although this metrical pattern confirms the expected quantity of the
vowel of the second syllable, at the same time it suggests that the o
vowel of the third syllable is to be read as short. This is a bit surprising,
since we otherwise have no direct evidence of the existence of * as an
independent phoneme in G!ndh!r" or other MIA languages. Since u and
o alternate frequently in G!ndh!r" orthography, one might suppose that
i!maho is merely a graphic alternative for *i!mahu, with final u as
suggested by Sanskrit ik!v%ku. However, the fact that the name is
consistently written with -o in all four attestations speaks against this,
and we can therefore suppose that the pronunciation was i!m%h*,
although the phonological status and etymological significance of the
final vowel remain uncertain.
As noted above, although the functional equivalence of G!ndh!r"
i!maho to Sanskrit ik!v%ku and P!li okk%ka is clearly established, the
phonetic correspondences of the three forms of the name are anything
but normal :
For the initial vowel, P!li has o against Sanskrit and G!ndh!r" i.
8Compare n. 3 above.
For the consonant clusters in the second syllable, the three languages have
respectively kk, k!v, and !m, none of which are normally equivalents for
any of the others.
For the consonant of the third syllable, G!ndh!r" has, untypically, h
against k of P!li and Sanskrit.
For the final vowel, P!li, Sanskrit and G!ndh!r" have a, u, and o
respectively.
At first glance it therefore seems likely that in ik!v%ku / okk%ka /

i!maho we have an instance of the frequent pattern whereby proper
names in the various Indian Buddhist languages9 exhibit irregular
phonetic correspondences. This pattern was already well-established in
connection with P!li and Sanskrit, and recent discoveries of numerous
Buddhist literary texts in G!ndh!r"10 have shown that it applies there as
well. One example where the newly discovered G!ndh!r" form of a
proper name fails to correspond normally with either the P!li or the
Sanskrit forms — involving the name of the city of Taxila, namely
Sanskrit tak!a$il%, P!li takkasil%, and G!ndh!r" tak!aïla — is discussed
in detail in Salomon 2005B, and several other cases (some involving
material that has not yet been published) have also been observed. For
example, the G!ndh!r" equivalent of the name of the king known in P!li
as pasenadi and in Sanskrit as prasenajit — which, as usual, themselves
do not correspond normally — has now been revealed to be pras!eniga,
which again corresponds neither to the P!li nor the Sanskrit form (Allon
2001, p. 304 ; British Library Kharo!$h" fragments 12 + 14, line 75,
p[r]as!en[i]g[e]no). The overall problem of the relationship of the
aberrant manifestations of proper names in different Buddhist languages
has not yet been studied in any organized and comprehensive manner
9This is not to suggest that this phenomenon is unique or peculiar to Buddhist

languages, or even to Indian languages only. Similar inconsistencies between
dialectal forms of proper names, involving special etymological, phonological,
and/or orthographic patterns, could presumably be documented in other
language groups in India and elsewhere, although we are not aware of any
systematic studies of this phenomenon.
10For an up-to-date summary of these and related finds, see Allon, forthcoming.
(though we intend to address it in a future study with special reference
to proper names in G!ndh!r").
Buddhist scholars in ancient times, like us, wrestled with the
problem of the etymology of such proper names, and often had to resort
to explanations that from the modern perspective it is easy to dismiss as
“folk etymologies”, but that doubtless, once established, themselves
began to exert an influence on the transmitted forms of these names in
the Buddhist tradition. For example, the Sanskrit form of the name
ik!v%ku is typically derived from ik!u, “sugar-cane”, for which deriva-
tion a legend was created according to which the eponymous king
Ik!v!ku was born from a sugar-cane plant and named accordingly : e!o
kum%ro ik!uto j%to bhavatu imasya ik!v%kutti n%ma&, “This baby was
born from the sugar-cane (ik!u), so let his name be Ik!v!ku” (Mah!-
vastu II 422.19–20). In a variant of this legend recorded in the
M(lasarv!stiv!da-vinaya, the sage Suvar#advaip!yana found two new-
born boys in a sugar-cane field and named them after this findspot :
ik!uv%(%l labdhv% ik!v%k% ik!v%k% iti caturth' sa&jñ% sa&vÁtt%,
“Because they were taken from a sugar-cane field, their fourth name
became ‘Ik!v!ka, Ik!v!ka’ 11” (Sa'ghabhedavastu, ed. Gnoli, I 25–26).
After they grew up, both brothers in turn succeeded to the throne, and
the younger became the progenitor of the Ik!v!ku clan.
The corresponding P!li name, okk%ka, is differently but equally
fancifully derived by Buddhaghosa in his commentary on the D"gha-
nik!ya from ukk% “torch” (= Skt ulk%), on the grounds that when King
Okk!ka spoke it seemed as if the light from a torch (ukk%) came out of
his mouth : tassa kira rañño kathanak%le ukk% viya mukhato pabh%
niccharati, tasm% na& okk%ko ti sañj%ni&su, “They say that when that
king spoke, a light like [that of] a torch (ukk%) came forth from his
11Notethe final vowel -a, as in P!li okk%ka and Jaina Prakrit ikkh%ga (cited
below).
mouth, and therefore they named him Okk!ka” (Suma'galavil!sin" I
258.6–8).12
Yet despite these very different traditional etymologies for Sanskrit
ik!v%ku and P!li okk%ka, and despite the striking phonetic inconsis-
tencies between them, it is clearly not out of the question that they are
in fact etymologically related. No less an authority than Wilhelm Geiger
maintained that this was the case, supporting this correspondence with
the following three arguments :
(1) The initial o of the P!li name comes from the u- of an original
*ukkh%ka, according to the rule that “[n]ot infrequently i and u
become + and * before double-consonance” (Geiger 1943, p. 65).
(2) The form *ukkh%ka is justified on the grounds that “Sometimes in
P[!li] kkh and cch alternate in one and the same word” (Geiger 1943,
p. 100), so that a hypothetical *ukkhu could have coexisted in P!li or
related dialects with ucchu, which is the usual P!li equivalent of
Sanskrit ik!u “sugar cane” (Geiger 1943, p. 66, n.1).
(3) The deaspiration of the second syllable (*ukkh%ka > okk%ka) is
explained by comparison with other instances of “[m]issing aspiration
in sound-groups with the sibilant in second position” (Geiger 1943,
p. 105).
Each of these proposed changes is in and of itself plausible and

more or less well attested, but it is still noteworthy how much special
pleading is required to establish a regular etymological correspondence
between ik!v%ku and okk%ka, and it must also be pointed out that the
irregular contrast between the final vowels (u / a) remains unexplained.
The situation is further complicated by the corresponding name in the
Jaina Prakrits, which usually appears as ikkh%ga, although (teste Mehta
and Chandra 1970, p. 103) ikkh%gu is also attested in the compound
ikkh%gu-va&sa. Thus the usual Prakrit form, ikkh%ga, corresponds to
Sanskrit ik!v%ku except for the final vowel, which agrees with P!li
12Compare also the etymology of the name ik!v%ku found in the Brahmanical
tradition, where it is said that Ik!v!ku was born from the nose of his father
Manu when the latter sneezed (-k!u) ; e.g., k!uvatas tu manor jajñe ik!v%kur
ghr%"ata) suta) (Bh!gavata-pur!#a 9.6.4ab).
okk%ka (and with the Buddhist Sanskrit form ik!v%ka in the Sa'gha-
bhedavastu passage quoted above).13
Thus one can feel some sympathy for the opinion of E.J. Thomas,
diametrically opposite to that of Geiger, who declared that “P!li …
Okk!ka … cannot by any device be treated as a form of the name
Ikshv!ku” (1927, p. 6). Nonetheless, the peculiar phonetic correspond-
ences between certain proper names in Sanskrit and P!li, including
ik!v%ku / okk%ka as well as tak!a$il% / takkasil%, may yet prove to be
regular as our knowledge of their transmission improves. They may, for
instance, find a partial explanation in phonological features of the
Sinhala language which could have affected their rendition in P!li texts
as transmitted and canonized in Sri Lanka. This could explain the
otherwise anomalous deaspiration of expected kkh in both of the afore-
mentioned P!li forms (cf. Geiger 1938, pp. 39–40, and the third
argument from Geiger 1943 cited above).14
In balance, it may tentatively be concluded that, despite their rather
peculiar correspondence, Sanskrit ik!v%ku and P!li okk%ka probably are
etymologically related. The next question, then, is whether the same can
be said for the newly identified G!ndh!r" form of the name, i!maho. As
noted previously, there are two main problems in establishing a direct
parallelism between the consonants in i!maho and Sanskrit ik!v%ku.
Regarding the initial of the final syllable, the usual G!ndh!r" reflexes of
Sanskrit intervocalic -k- are g, gh or !, but not h. There is, however, at
least one fairly clear instance of -k- > -h-, namely tuspahu as the equiva-
lent of Sanskrit yu!m%kam, occurring eight times in scroll 5 of the
13The form of the name which appears in the Prakrit inscriptions of the Ik!v!ku
kings of the Deccan (see section 3a) is ikh%ku (graphic for ikkh%ku ; Vogel
1929 , p. 27). This relatively late form corresponds directly to the Sanskrit and
is presumably derived from it. It is therefore of no further significance for our
discussion.
14It is less clear whether Sinhala vowel harmony (Geiger 1938, pp. 22–25) can
be invoked to explain the variation in the final vowel of the name (P!li a,
Sanskrit u), since forms with final a also occur on the Indian mainland, as
noted above.
Senior G!ndh!r" manuscript collection (Glass 2007, §§ 5.2.1.1, 6.2.1),
and on the basis of this data it is at least plausible to equate the h of
i!maho with the k of ik!v%ku.15
It is more difficult, though not impossible, to establish a connection
between the clusters !m of i!maho and k!v of ik!v%ku. Two separate
problems are involved in this and will be discussed in turn : the apparent
reduction of OIA k! to G !, and the correspondence of OIA v to G m.
The reduction of k! to ! is initially puzzling, since in isolation the OIA
cluster k! is usually retained in G!ndh!r" as such, or rather is repre-
sented in writing by the Kharo!$h" character ! which is conventionally
transliterated as k! but which was probably a unitary consonantal
phoneme whose pronunciation cannot be precisely determined, but
which may have been [$!] or the like (Brough 1962, p. 72 and n. 4).
There are, to be sure, exceptions to this pattern. Thus, in certain cases
the equivalent of OIA k! is represented as kh in G!ndh!r", as in the
frequent bhikhu = bhik!u, but this and most other such cases are
explainable as borrowings of Buddhist technical terms into G!ndh!r"
from another MIA dialect. There is also at least one case, namely kuchie
= kuk!au “in the stomach”,16 where OIA k! is reflected by G!ndh!r" ch.
But there is no instance known to us where G!ndh!r" has ! for isolated
OIA k!.
In OIA ik!v%ku, however, special conditions obtain since here k! is
part of the rare three-consonant cluster k!v. No other parallel is
15There is also one instance where an intervocalic -h- in Sanskrit is represented

by -"- in G!ndh!r", namely sata"am = *sapt%ham, “for a week”, in an unpub-
lished fragment of a G!ndh!r" version of the Anavatapta-g!th! in the Senior
collection (fragment 14, line 20 ; Salomon 2003A : 79 ; Salomon, in progress).
This could be interpreted as a hypercorrection resulting from a (near-)merger
in the scribe’s dialect of the reflex of Skt intervocalic k and g with h. (The
Kharo!$h" letter ", a modified form of k, probably indicates the voiced
fricative [.].)
16In the British Library manuscript of another G!ndh!r" version of the
Anavatapta-g!th! (British Library Kharo!$h" fragment 1, line 38 ; Salomon

1999, pp. 30–33 ; Salomon, in progress).
available in G!ndh!r" for this particular cluster ; the only attested case of
a G!ndh!r" reflex of an OIA cluster of the type k!C involves k!y, which
is represented in G!ndh!r" as k! in dhrek!atu = *drak!yantu (Allon
2001, p. 89) and muk!asa = mok!yatha.17 We do, however, know that in
other forms of MIA three-consonant clusters could receive special
treatment, and in particular that sometimes the second consonant in
such clusters seems to have been articulated more strongly than the first,
outweighing it in assimilation :
Wenn sich jedoch die Silbengrenze in die Konsonantengruppe hinein-
verlagert, wird dadurch wie in der Kompositionsfuge die Hierarchie
scheinbar aufgehoben [...], da jetzt der zweite Konsonant stärker artikuliert
wird. (von Hinüber 2001, pp. 202f., based on Berger 1955, pp. 76f.)
Among the several examples listed by von Hinüber, two are especially
relevant for our discussion of OIA ik!v%ku and G i!maho : “Skt t'k!"a :
t'k-!"a > *t'!-!"a > mi. ti"ha neben mi. tikkha [...] und P tikhi"a <
*tikh"a” and “Skt abh'k!"a > *abhi!!"a > P abhi"ha neben P, Amg
abhikkha"a”. In the light of these parallels, G i!maho would represent
exactly the reconstructed middle stage in the development of such
clusters : k!v [k!v] has undergone assimilation to !m [!m] (see next
paragraph on the change from v to m), but sibilant and nasal have not
yet been metathesized. Indicating syllable boundaries by hyphens, as in
von Hinüber’s examples, the sequence of developments would then
have been : OIA ik-!v%-ku > *[i!-!v!-ku] > *[i!-!m!-ku] > G i!maho.
The apparent counter-examples of stable k! in G dhrek!atu and
muk!asa, cited above, have to be seen on the background of independ-
ent assimilation of OIA !y > G $ and the need for morphological clarity
at the boundary of verbal root and tense suffix.
The other problem in the correspondence of OIA k!v to G !m is the
apparent change of !v into !m. The normal outcome of OIA sibilant + v
in G!ndh!r" is $p : prabh(*a)[$p](*a)ra < prabh%svara (Allon 2001,
p. 96), pari$pei#a"a < parisvedit%ni (Glass 2006, p. 145), i$parasa <
17British Library, Anavatapta-g!th!, lines 95, 122.

'$varasya (von Hinüber 2003, p. 41). Alternative outcomes, especially
in the Khotan Dharmapada and the Niya documents, are preservation of
the original cluster and complete assimilation (e.g. svaga and saga <
svarga, Brough 1962, p. 103), but as far as the available data allows us
to judge, original sibilant + v never turns into sibilant + m in G!ndh!r"
(as already noted in Allon 2001, p. 96, n.8).18
OIA sibilant + m, on the other hand, has a broad range of G out-
comes, including besides sm, $p and s (cf. Allon 2001, pp. 95f.) also sv :
rasvi < OIA ra$mi or MIA *rasmi, svadi < smÁti (Brough 1962,
pp. 102f.), [s]v(*a)[d](*ima) < smÁtimant (Salomon 2000, p. 91).19
Thus, while it may be true that G !m itself cannot be considered a
regular outcome of OIA !v, in a more general sense sibilant + m and
sibilant + v seem to have functioned as phonetic variants in G!ndh!r". If
we further keep in mind that none of the currently attested G outcomes
of OIA sibilant + v involve an original retroflex sibilant and that none of
them involve an original three-consonant cluster, we may cautiously
suggest that !m in i!maho is at least a plausible G!ndh!r" phonetic
development of earlier !v. In conclusion, it appears possible to consider
the medial cluster !m in G i!maho not only a regular MIA development
of OIA k!v, but in fact an attestation of the type of reconstructed
intermediate form posited by Berger and von Hinüber for P ti"ha and
abhi"ha.
Having considered the relationship of the three main attested forms
of the proper name Ik!v!ku, we now turn to the question of its ultimate
origin and meaning. Since we have seen that G i!maho can plausibly be
derived from a form like OIA ik!v%ku whereas the inverse is not true
(expected back-formations would have been *i!m%(b)hu or even
18For a comprehensive discussion of the MIA development of stop or sibilant +

v see Sakamoto-Goto 1988.
19These examples also show that the G sound change sm > sv is of wider
application than the corresponding change in other dialects of MIA that is

usually explained as nasal dissimilation (Sakamoto-Goto 1988 : 96–100, von
Hinüber 2001 : 190).
*i!v%(b)hu, but not ik!v%ku), and since Skt ik!v%ku and P okk%ka agree
in having a k in their first cluster that G i!maho lacks, it appears that in
this case the G form of the name has no claim to greater antiquity than
either the Skt or the P form. In fact, in some respects it would seem to
be farther removed from the ultimate origin of the name than both the
Skt and the P form, and while any future investigation into this origin
will have to account for the newly-discovered G form, it would be
unwise, despite the early attestation of the G form, to base any ultimate
etymology on the form i!maho alone without giving equal consideration
to the other two forms.20
The traditional derivation from the word ik!u “sugar cane” is thus
neither confirmed nor contradicted by the new G evidence. It is clear,
however, that at the linguistic stage of G!ndh!r" itself any original
connection with ik!u would have ceased to be transparent and that a
secondary folk-etymological connection with ik!u (as in the Mah!vastu
and Sa'ghabhedavastu passages quoted above) would likewise have
been difficult to maintain : while the word for “sugar cane” has not yet
been found attested in G!ndh!r", its form would almost certainly have
been *ik!u (or maybe *uk!u) and thus clearly distinct from the word
i!maho. This is of course the same situation as obtains in P!li where, as
we have seen, Buddhaghosa did not attempt to establish a connection
between okk%ka and ucchu, but instead drew on the word ukk% “torch”
(pace Geiger’s attempt to connect okk%ka with ucchu).
The connection of the name Ik!v!ku with ik!u has independently
been cast into doubt by several modern authorities. Thus the derivation
from ik!u (“Augenwimper, Zuckerrohr”) plus a suffix %ku, as proposed
20Were it not for this, one could have speculated that i!maho might be related to
i!u ‘arrow’ or the rare Skt i!ma / '!ma / i!va / '!va ‘spring, name of the god
K!ma’ (comm. on U#!dis(tra 1.144 ; i!ma) k%mavasa&tayo) (P!#%eya
1985), p. 18 ; cf. also Monier-Williams, Sanskrit–English Dictionary, s.v.
'!ma). But in view of the preceding arguments, these two words could at most
have assumed a local Gandh!ran folk-etymological relationship to the name-
form i!maho and are highly unlikely to be the ultimate source of the attested
triplet of forms ik!v%ku / i!maho / okk%ka.
by Wackernagel and Debrunner (1954, p. 267), was dismissed by
Mayrhofer (1992, p. 186) as “nicht zielführend”, and Witzel (1999,
p. 357) characterizes the supposed suffix %ku as “strange”. It may be the
case that the true origin of this proper name, as of so many others in
Sanskrit and other languages, lies buried, probably irretrievably,
beneath the sands of time. That is to say, it may ultimately go back to
some long-lost word, whether Indo-Aryan or quite possibly belonging
to an indigenous substrate language. This in fact is the conclusion of
Kuiper (1991, pp. 6–7), who includes ik!v%ku among the “group of
persons who were on the side of the Aryan society but whose names
must, on morphological grounds, be considered non-Aryan”. This view
is also endorsed by Witzel (1999, pp. 356, 360), who classifies ik!v%ku
among the numerous proper names in the "gveda which he considers to
be “Non-IA or of doubtful etymology” (p. 356), and this conclusion
appears to be cautiously endorsed by Mayhofer (2003, p. 18), who lists
ik!v%ku as “Fremdname ?”. An attempt to trace such a pre-Indo-Aryan
etymology was in fact made by Berger (1959, p. 73), who explained
ik!v%ku “bitterer Kürbis, Citrillus Colocynthis” as a survival of an
Austroasiatic word for “pumpkin” (Kürbis), allegedly functioning as a
totemic clan name. This etymology is cited by Mayrhofer (1992,
pp. 185–86) without comment, but the justification provided by Berger
is sketchy at best and can hardly be considered definitive.
Of course, it is always possible that some future discovery or insight
may provide a more convincing solution to the problem of the ultimate
origin of the name Ik!v!ku, but at this point one hardly dares to hope
for this. For such a new source of information could have been hoped
for, if anywhere, in G!ndh!r" ; but in fact, we find that the G!ndh!r"
form does not do much to clarify this issue, at least for the time being.
This means, most likely, that the etymological issue is not one that is
definitively soluble, and the ultimate origin of the name may be lost in
the mists of prehistory.
3. Ramifications, historical and Buddhological
3a. The Ik!v%kus and the Kings of O#i
This, however, is by no means to say that the new G!ndh!r" data is of
no use to us. Quite to the contrary : although it does not solve the
etymological problem surrounding the name Ik!v!ku and its
equivalents, it does provide new insight into other issues. The first of
these involves the history of the I!maho kings of O%i, in one of whose
inscriptions, the st!pa dedication of Se#avarma, the G!ndh!r" form
i!maho was first noticed (section 1). The I!maho kings, who are known
only from three Buddhist reliquary inscriptions in G!ndh!r", ruled,
apparently, in lower Swat in or around the first century A.D.21 Like their
neighbours, the kings of Apraca,22 the O%i kings seem to have been
feudatory allies of the Saka and early Ku!!#a dynasties of Gandh!ra
and adjoining areas.
Now that it has become clear that their dynastic name I!maho is not
“non-Indian” or “non-Aryan” as once thought (see section 1), but rather
is the G!ndh!r" equivalent of the ancient and renowned name Ik!v!ku,
we can see that the nomenclature of the I!maho dynasty is part of a
recurrent historical pattern. For there are at least two other instances in
which Indian Buddhist dynasties of the historical period took on the
name Ik!v!ku in order to lay claim to an association with the lineage of
the Buddha himself, who, as a *!kya, was held to have belonged to the
venerable Ik!v!ku line. The first such case is the Ik!v!ku (= ikh%ku ; see
n. 13) dynasty of the eastern Deccan, which patronized the great
Buddhist monasteries at N!g!rjuniko#%a and elsewhere in the KÁ!#!
River Valley in the third century A.D . The second instance of this
pattern is documented in the Sri Lankan Buddhist historiographic
tradition, where the D"pava)sa “portrayed the Sri Lankan kings as the
true heirs to the Ik!v!ku legacy, a claim that the Ik!v!kus of Andhra had
21See Salomon 2003B : 39–51 for the most recent information on the
inscriptions and history of the O%i kings.
22For recently discovered inscriptions of and information on the Apraca kings,
see Salomon 2005A : 378–83.

earlier staked out for their imperial kingdom in which, at
N!g!rjuniko#%a, Sri Lankan Buddhists had received their first recorded
recognition” (Walters 2000, p. 118). Furthermore, in various Buddhist
literary traditions (as summarized in Lamotte 1988, pp. 218, 681–82)
the Mauryas and other dynasties are credited with a familial relation to
the *!kyas and thereby to the Ik!v!kus, and no doubt many further
examples could be cited.
Of course, the skeptical historian cannot fail to doubt the legitimacy
of these alleged descents from the Ik!v!ku line, and this skepticism
need not be restricted to the instances from the relatively later periods.
For even the claim of the *!kyas themselves to Ik!v!ku descent23 has,
to say the least, a legendary air about it. According to the account in the
Amba$$ha-sutta of the P!li D"gha-nik!ya (D I 92–93)24 and Buddha-
ghosa’s commentary thereon, the original King Okk!ka, under the
influence of his favorite wife, exiled his five eldest sons from his
kingdom, whereupon they settled near the slopes of the Him!laya and
became known as the Sakkas (= *!kyas). The legend of the exiled sons
seems a “likely story”, which could easily inspire one to question the
historicity of *!kyamuni’s Ik!v!ku descent. That is to say, one may
suspect that the association of the Buddhist lineage with the venerable
line of Ik!v!ku, who in Brahmanical tradition was the son of Manu, the
grandson of the Sun, and the progenitor of the royal line of R!ma, was a
device to establish legitimacy and nobility for the Buddhist line in the
eyes of the wider, non-Buddhist world of the time.
However this may be, we can be quite certain that the claims of the
kings of O%i to I!maho / Ik!v!ku lineage is, historically speaking, a
spurious one. For, although their dynastic name is now known to be an
Indian and not a foreign one, and although their personal names are all
(with one partial exception, Di&ase#a) “durchsichtige und gut deutbare
23As recorded, for example, in the Saundarananda (1.24), as quoted above

(section 1).
24This legend is also referred to in A&vagho!a’s Saundarananda 1.18–21 and in
Mah!vastu I 348.11–351.14.
Sanskritnamen” (von Hinüber 2003, p. 33, n. 30), this does not mean
that they were in fact of Indian ethnicity. A priori, given their historical
and geographical situation, one may expect that they were, like their
neighboring rulers and allies, Sakas or other Central Asian nomads who
had conquered territories in the northwestern borderlands of India
around the beginning of the Christian era and adopted the Buddhist
religion and Indian names. This suspicion is confirmed by the reference
in the inscription of Se#avarma (line 1c) to his identity as a “Kadama”
(tasa dayate"a me kadamasa, “of me, by descent from him a Kadama”).
This term is in all probability equivalent to the label k%rddamaka which
was applied to a member of the Saka dynasty of Western India in an
inscription at Ka#her", and also to kardamaga, the name of a king, very
likely also a Scythian, who is mentioned in one of the G!ndh!r"
avad%na texts among the British Library scrolls (Salomon 2003B,
pp. 48 ; 58, n. 9 ; Salomon 2005C, p. 318). Therefore it is very likely
that the I!maho kings of O%i were in fact Sakas or members of some
other Central Asian ethnic groups who claimed a spurious Indian
lineage in order to legitimize their Buddhist kingship.
3b. I!maho and the G%ndh%r' Hypothesis
Another point of interest regarding the name i!maho involves its
implications for the early history of Buddhism in China, and in
particular for the “G!ndh!r" hypothesis”, according to which some of
the earliest Chinese translations of Buddhist texts were prepared from
originals not in Sanskrit, but in G!ndh!r" or Sanskritized versions of
underlying G!ndh!r" texts.25 This theory was originally proposed on the
basis of the transcriptions of certain proper names in early Chinese
Buddhist translations which seemed to reflect G!ndh!r" rather than
Sanskrit pronunciations, or features of Kharo!$h" rather than Br!hm"
script, and the body of relevant evidence has grown and expanded in
recent years. The newly discovered G!ndh!r" word i!maho constitutes
25Fora general discussion of the “G!ndh!r" hypothesis”, see Boucher 1998 :

471–75.
another such case, in that it, rather than the Sanskrit form ik!v%ku, is
clearly reflected in certain Chinese renditions of this name.
The Chinese equivalents of ik!v%ku etc. are numerous, but they
appear to fall into three main groups. These are :
(1) 一叉鳩# y'ch%ji! (reconstructed Old Northwest Chinese [ONWC]
pronunciation, following Coblin 1994 : !iit-t!hä-ku), a transcription of
Sanskrit ik!v%ku. This appears, for example, in the 大般涅槃經# Dà
b%nnièpán j'ng (= Mah!parinirv!#a-s(tra ; T. vol. 12, no. 375,
p. 839c23).
(2) 甘蔗王# g%nzhèwáng “Sugar-Cane King”, a translation of the Sanskrit
ik!v%ku on the basis of the traditional etymology from ik!u “sugar-cane”
(as discussed in section 2). This form occurs, for example, in the
根本說一切有部毘奈耶藥事# G,nb-n shu. y'qièy/ubù pínàiy, yàoshì
(= M(lasarv!stiv!da-vinaya-bhai!ajyavastu ; T. vol. 24, no. 1448,
p. 33c23).
(3) 懿師摩# yìsh'mó (ONWC !i(s)-!i-ma ; also several related forms and
variants, discussed below), a very good phonetic approximation of
G!ndh!r" i!maho, which cannot be connected with Sanskrit ik!v%ku or
P!li okk%ka. This form of the name appears in the 四分律#Sìf,n l0 (=
Dharmaguptaka-vinaya ; T. vol. 22, no. 1428, p. 779b1, etc.).
It is particularly interesting that this third rendition of the name, the one
which clearly reflects a G!ndh!r" substrate, occurs in the vinaya of the
Dharmaguptaka school, because this concords with an already
established pattern of associations between the Dharmaguptakas and the
recently rediscovered remnants of Gandh!ran Buddhist texts. This
association is manifested in the following data :
(1) The British Library scrolls, the oldest and largest collection of G!ndh!r"
manuscripts known to date, were found in a pot bearing a dedication to
the Dharmaguptakas (Salomon 1999, pp. 166–67).
(2) A manuscript among the British Library scrolls containing the Sa'g"ti-
s(tra with commentary has a close relationship in its contents and
arrangement to the version of the Sa'g"ti-s(tra contained in the Chinese
translation of the D"rgh!gama (長阿含經# Cháng %hán j'ng), which is
almost certainly a Dharmaguptaka text (Salomon 1999, pp. 171–75).
(3) Fragments of a G!ndh!r" version of the Mah!parinirv!#a-s(tra in the
Schøyen manuscript collection similarly resemble the corresponding
version of this s!tra in the Chinese D"rgh!gama more closely than the
several other versions, although here the pattern is not as distinct as in
the case of the Sa'g"ti-s(tra (Allon and Salomon 2000, pp. 272–73).
(4) The G!ndh!r" version of the *r!ma#yaphala-s(tra contained in scroll 2
of the Senior collection of G!ndh!r" manuscripts (Salomon 2003A)
similarly seems, on the basis of a preliminary study, to resemble the
Chinese D"rgh!gama recension of this s!tra more than any of the
several other versions (Allon, in progress).
(5) Episodes from the life of the Buddha recorded in scroll 24 of the Senior
collection apparently resemble the corresponding versions of the same
stories in the Chinese Dharmaguptaka-vinaya more than those in other
vinayas (Allon, in progress).26
The correspondence of G!ndh!r" i!maho with 懿師摩 yìsh'mó of the

Dharmaguptaka-vinaya is thus consistent with the several other indica-
tions of connections between the newly rediscovered G!ndh!r" literary
corpus and the Dharmaguptaka tradition as it was transmitted to and
preserved in China. However, the matter becomes considerably more
complicated when we take into account the several other Chinese
renditions of the name in question, as follows :
聲摩# sh,ngmó (ONWC $e1-ma) : 長阿含經# Cháng %hán j'ng (=
D"rgh!gama), e.g., T. vol. 1, no. 1, p. 82c23 (s!tra no. 20) and
p. 149a20 (s!tra no. 30).
鼓摩#g2mó (ONWC ko-ma) : 佛開解梵志阿經#Fó k%iji- fànzhì %fú j'ng
(= Amb!!$has(tra), T. vol. 1, no. 20, p. 260a26.
鬱摩#yùmó (ONWC !ut-ma) : 彌沙塞部和醯五分律#Mísh%s%ibù héx' w2f,n
l0 (= Mah"&!saka-vinaya), T. vol. 22, no. 1421, p. 101a10.
26Also of interest in this context is a passage in the Dharmaguptaka-vinaya (T.

vol. 22, no. 1428, p. 639a14 ; discussed in Lévi 1915 : 440, Salomon 1990 :
255, and Boucher 1998 : 474) which refers to the recitation of the Arapacana
syllabary by monks. Since it is now established that the Arapacana was
originally the ordinary alphabetic order of the Kharo!$h" script (Salomon
1990 : 262, 265), this passage provides a further suggestion of an assocation
between the Dharmaguptaka school and G!ndh!r" textual traditions.
伊摩# y'mó (ONWC !ii-ma) : 大樓炭經# Dà lóutàn j'ng (= *Mah!-
parid!has(tra ?), T. vol.1, no. 23, p. 309a23.
The origins and relationships of these alternative forms of the name

are quite complex, but they all seem to be related to the G!ndh!r"-
derived 懿師摩 yìsh'mó as graphic and / or phonetic variants. For
example, in the D"rgh!gama (T. vol. 1, p. 149a20) 懿摩 yìmó is given as
a variant (Sòng and Míng editions) for 聲摩 sh,ngmó. This 懿摩 yìmó
is presumably a graphic variant, and since 聲摩 sh,ngmó is difficult to
explain as a phonetic equivalent of ik!v%ku etc., it is perhaps a
corruption of an original 懿摩 yìmó, the latter being in turn a shortened
transcription in place of the fuller form 懿師摩 yìsh'mó, of a sort that is
common in Chinese Buddhist translations (e.g. 目連 mùlián =
Maudgaly%yana / Moggall%na).
鼓摩 g2mó, occurring in an early independent s!tra translation by
Zh"qi!n 支謙 (A .D. 222–253), can similarly be explained as a graphic
variant for the aforementioned 懿摩 yìmó. 鬱摩 yùmó in the
Mah"&!saka-vinaya is conceivably also a graphic variant for 懿摩 yìmó,
while 伊摩 y'mó in the independent s!tra translation 大樓炭經 Dà
lóutàn j'ng might be a sound variant for it or a similar form.
The association of all of these forms with each other as graphic or
phonetic variants of an original 懿(師)摩 yì(sh')mó is in fact endorsed
by the Liáng-dynasty scholar S/ngyòu 僧佑 (d. A.D. 518) in his treatise
釋迦譜 Shìji% p2 “Genealogy of the *!kya Clan” (T. vol. 50, no. 2040).
S/ngyòu notes (pp. 3c23–4a2) with regard to this name : “In ancient
times there was a king named Yìmó 懿摩. (The Lóutàn j'ng says Y"mó
一摩.) The Dharmaguptaka-vinaya says G0sh"mó 鼓師摩, but the
Mah"&!saka-vinaya says Yùmó 鬱摩. These three sounds, y' (一), yì
(懿), and yù (鬱), are close to one another. Considering their sounds, I
suppose that Yìmó 懿摩 is the original one. But as for the characters 鼓
g2 and 懿 yì, they resemble each other, and therefore in the copying [懿
yì] was just a mistake for 鼓 g2.”27
27乃往過去有王。名懿摩(樓炭經云一摩)。曇無德律云。鼓師摩。彌沙塞
律云。鬱摩一懿鬱。此三音相近。以音而推。竊謂懿摩是正。但鼓懿字
If the interpretation proposed above is correct, it would mean that
the G!ndh!r"-derived or G!ndh!r"-influenced forms of the name
Ik!v!ku occuring in Chinese translations are not limited to Dharma-
guptaka texts. They are, to be sure, prevalent there, both in the Dharma-
guptaka-vinaya which has the clearly G!ndh!r"-based 懿師摩 yìsh'mó
and in the D"rgh!gama, a probable Dharmaguptaka collection, whose
聲摩 sh,ngmó is, as noted above, probably a variant of the former. But
we also have 鬱摩 yùmó, again likely a variant of 懿師摩 yìsh'mó, in
the Mah"&!saka-vinaya, as well as several other variants in early
individual s(tra translations of uncertain sectarian affiliation. Therefore,
although the data derived from the Chinese forms of this name does
support an association between the textual tradition of Gandh!ra and
that of the Dharmaguptakas as reflected in early Chinese translations, it
also reminds us that this is no by means necessarily an exclusive
relationship. Indeed, we should rather expect that texts of other schools
would have existed in G!ndh!r" (whether or not they have survived or
will ever be found), and that Chinese texts affiliated with those other
schools also would reflect G!ndh!r" substrate forms.28
相似。故傳寫謬為鼓耳。Interestingly enough, S/ngyòu here gives 鼓師摩

g2sh'mó as the reading of the Dharmaguptaka-vinaya, rather than 懿師摩
yìsh'mó as given in the Taish1 text edition (cited above). These and similar
textual variations themselves confirm the author’s point that the various
readings are merely alternatives for the same name.
A somewhat different interpretation is offered by B2ochàng 寶唱, another
Liáng-dynasty scholar-monk, in his treatise 經律異相 J'ngl0 yìxiàng “S(tra
and Vinaya Miscellany” (T. vol. 53, no. 2121, p. 32a23). He gives the name
corresponding to ik!v%ku as 欝摩 yùmó and explains the alternative renditions
懿摩 yìmó and 鼓摩 g2mó as dialect approximations (方言之左右 f%ngyán
zh' zu/yòu) of 欝摩 yùmó (又云懿摩。長阿含經云鼓摩。蓋方言之左右
耳。).
28Although references to the Dharmaguptakas are particularly prominent among
G!ndh!r" inscriptions, several other schools, such as the Sarv!stiv!dins,
K!&yap"yas and Mah"&!sakas, are also mentioned in them (Salomon 1999 :
176–77). Thus we could reasonably expect that these schools, or at least their
Gandh!ran branches, would also have had textual corpora in G!ndh!r".
We can only hope that further studies of this and other words by
specialists in Chinese Buddhist translation literature will clarify both the
immediate problem raised here and the broader issues that it involves
and implies. But in the meantime, this new data does, on the one hand,
provide further evidence in favor of the “G!ndh!r" hypothesis” and, on
the other hand, confirm the significant role of the Dharmaguptaka
literature in it.
Richard Salomon and Stefan Baums
University of Washington
REFERENCES
Allon, Mark (with a contribution by Andrew Glass), 2001. Three G%ndh%r'
Ekottarik%gama-Type S!tras : British Library Kharo3(h' Fragments 12 and
14. Gandh!ran Buddhist Texts 2, Seattle : University of Washington Press
——— forthcoming (2007). “Recent Discoveries of Buddhist Manuscripts from
Afghanistan and Pakistan and their Significance”, in Ken Parry, ed., Art,
Architecture and Religion on the Silk Road and Across Inner-Asian History.
Silk Road Studies : Proceedings from the Fifth Conference of the Australian
Society for Inner Asian Studies (A.S.I.A.S.), Macquarie University,
November 27th to 28th, 2004, Turnhout : Brepols
——— in progress. Ancient Buddhist Scrolls from Gandh%ra II : The Senior
Kharo3(h' Fragments
——— and Richard Salomon, 2000. “Kharo3$h" Fragments of a G!ndh!r"
Version of the Mah!parinirv!#as(tra”, in Jens Braarvig, ed., Buddhist
Manuscripts in the Schøyen Collection, Vol. 1 (Manuscripts in the Schøyen
Collection 1 ; Oslo : Hermes Publishing), pp. 243–73
Bailey, H. W., 1980. “A Kharo3$hr" Inscription of Se#avarma, King of O%i”,
Journal of the Royal Asiatic Society of Great Britain and Ireland 1980,
pp. 21–29
Berger, Hermann, 1955. Zwei Probleme der mittelindischen Lautlehre.
Münchener indologische Studien, Heft 1, Munich : J. Kitzinger
——— 1959. “Deutung einiger alter Stammesnamen der Bhil aus der
vorarischen Mythologie des Epos und der Pur!#a”, Wiener Zeitschrift für
die Kunde Süd- und Ostasiens und Archiv für indische Philosophie 3,
pp. 34–82
Boucher, Daniel, 1998. “G!ndh!r" and the Early Chinese Buddhist Translations
Reconsidered : The Case of the Saddharmapu#%ar"kas(tra”, Journal of the
American Oriental Society 118, pp. 471–506
Brough, John, 1962. The G%ndh%r' Dharmapada. London Oriental Series 7,
London : Oxford University Press
Burrow, T., 1937. The Language of the Kharo3(hi Documents from Chinese
Turkestan, Cambridge : Cambridge University Press
Coblin, W. South, 1994. A Compendium of Phonetics in Northwest Chinese.
Journal of Chinese Linguistics Monograph Series 7, Berkeley : Project on
Linguistic Analysis
Edgerton, Franklin, 1953. Buddhist Hybrid Sanskrit Grammar. William Dwight
Whitney Linguistic Series, New Haven : Yale University Press
Fussman, Gérard, 1982. “Documents épigraphiques kouchans (III). L’inscrip-
tion kharo3$h" de Senavarma, roi d’O%i : une nouvelle lecture”, Bulletin de
l’École française d’Extrême-Orient 71, pp. 1–46
Geiger, Wilhelm, 1938. A Grammar of the Sinhalese Language, Colombo : The
Royal Asiatic Society, Ceylon Branch
——— (tr. Batakrishna Ghosh), 1943. Pali Literature and Language, Calcutta :
Calcutta University Press
Glass, Andrew, 2000. A Preliminary Study of Kharo3(h' Manuscript Paleo-
graphy. M.A. thesis, Department of Asian Languages and Literature,
University of Washington (http://depts.washington.edu/ebmp/downloads/
Glass_2000.pdf)
——— 2007. Four Gandh%ra Sa&yukt%gama S!tras : Senior Kharo3(h' Frag-
ment 5, Gandh!ran Buddhist Texts 4, Seattle : University of Washington
Press
von Hinüber, Oskar, 2001. Das ältere Mittelindisch im Überblick, 2nd ed.
Österreichische Akademie der Wissenschaften, philosophisch–historische
Klasse, Sitzungsberichte, 467. Band I, Veröffentlichungen der Kommission
für Sprachen und Kulturen Südasiens, Heft 20, Vienna : Verlag der
Österreichischen Akademie der Wissenschaften
——— 2003. Beiträge zur Erklärung der Senavarma-Inschrift. Akademie der
Wissenschaften und der Literatur, Abhandlungen der geistes- und sozial-
wissenschaftlichen Klasse Nr. 1, Stuttgart : Franz Steiner Verlag
Kuiper, F.B.J., 1991. Aryans in the Rigveda. Leiden Studies in Indo-European
1, Amsterdam : Rodopi
Lamotte, Étienne (tr. Sara Webb-Boin), 1988. History of Indian Buddhism from
the Origins to the 4aka Era, Publications de l’Institut orientaliste de
Louvain 36, Louvain-la-Neuve : Université Catholique de Louvain, Institut
Orientaliste
Lévi, Sylvain, 1915. “Sur la récitation primitive des textes bouddhiques”,
Journal asiatique sér. 11, v. 5, pp. 401–47
Mayrhofer, Manfred, 1992. Etymologisches Wörterbuch des Altindoarischen, I.
Band, Heidelberg : Carl Winter Universitätsverlag
——— 2003. Die Personennamen in der Âgveda-sa&hit% : Sicheres und
Zweifelhaftes. Bayerische Akademie der Wissenschaften, philosophisch–
historische Klasse, Sitzungsberichte, Jahrgang 2002, no. 3, Munich : Verlag
der Bayerischen Akademie der Wissenschaften
Mehta, Mohan Lal, and K. Rishabh Chandra, 1970. Prakrit Proper Names.
4gamic Index, Vol. 1 : Lalbhai Dalpatbhai Series 28, Ahmedabad : L.D.
Institute of Indology
P!#%eya, R!m Avadh, 1985. 4ivar%matrip%(hi-kÁta Lak3m'niv%sako$a (U"%di-
ko$a), V!r!#as" : Vi&vavidy!laya Prak!&ana
Sakamoto-Goto, Junko, 1988. “Die mittelindische Lautentwicklung von v in
Konsonantengruppen mit Verschlußlaut bzw. Zischlaut”, Indo-Iranian
Journal 31, pp. 87–109
Salomon, Richard, 1986. “The Inscription of Senavarma, King of O%i”, Indo-
Iranian Journal 29, pp. 261–93
——— 1990. “New Evidence for a G!ndh!r" origin of the Arapacana
Syllabary”, Journal of the American Oriental Society 110, pp. 255–73
——— 1999. Ancient Buddhist Scrolls from Gandh%ra : The British Library
Kharo3(h' Fragments, London : British Library / Seattle : University of
Washington Press
——— 2000. A G%ndh%r' Version of the Rhinoceros S!tra : British Library
Kharo3(h' Fragment 5B. Gandh!ran Buddhist Texts 1, Seattle : University
of Washington Press
——— 2003A. “The Senior Manuscripts : Another Collection of Gandh!ran
Buddhist Scrolls”, Journal of the American Oriental Society 123, pp. 73–92
——— 2003B. “Three Kharo3$h" Reliquary Inscriptions in the Institute of Silk
Road Studies”, Silk Road Art and Archaeology 9, pp. 39–69
——— 2005A. “The Indo-Greek Era of 186/5 B.C. in a Buddhist Reliquary
Inscription”, in Osmund Bopearachchi and Marie-Françoise Boussac, eds.,
Afghanistan : Ancien Carrefour entre l’Est et l’Ouest (Indicopleustoi :
Archaeologies of the Indian Ocean 3, Turnhout : Brepols Publishers),
pp. 359–401
——— 2005B. “The Name of Taxila : Greek 56789:6, G!ndh!r" Tak3aïla,
Sanskrit Tak3a$il%, Pali Takkasil%”, East and West 55, pp. 265–277
——— 2005C. Review of von Hinüber 2003. Journal of the American Oriental
Society 125, pp. 316–20
——— in progress. Two G%ndh%r' Versions of the Anavatapta-g%th% : British
Library Kharo3(h' Fragment 1 and Senior Scroll 14, Gandh!ran Buddhist
Texts, Seattle : University of Washington Press
Thomas, Edward J., 1927. The Life of Buddha as Legend and History, London :
K. Paul, Trench, Trubner & Co
Vogel, J. Ph., 1929. “Prakrit Inscriptions from a Buddhist Site at Nagar-
junikonda”, Epigraphia Indica 20, pp. 1–36
Wackernagel, Jacob, and Albert Debrunner, 1954. Altindische Grammatik, II.2 :
Die Nominalsuffixe, Göttingen : Vandenhoeck & Ruprecht
Walters, Jonathan S., 2000. “Buddhist History : The Sri Lankan P!li Va)sas
and Their Commentary”, in Ronald Inden, Jonathan Walters, and Daud Ali,
eds., Querying the Medieval : Texts and the History of Practices in South
Asia, New York : Oxford University Press, pp. 99–164
Witzel, Michael, 1999. “Aryan and Non-Aryan Names in Vedic India : Data for
the Linguistic Situation, c. 1900–500 B.C.”, in Johannes Bronkhorst and
Madhav M. Deshpande, eds., Aryan and Non-Aryan in South Asia :
Evidence, Interpretation, and Ideology (Harvard Oriental Series, Opera
Minora 3, Cambridge : Department of Sanskrit and Indian Studies, Harvard
University), pp. 337–404
FIGURES
(both courtesy of the Library of Congress)
#
Fig. 1: A fragment from the beginning of the Library of Congress Scroll.
#
Fig. 2: Detail of verse 3c on the fragment shown in fig. 1,
with the word i3maho highlighted.#
A Gåndhår¥ Version of the Simile of the Turtle
and the Hole in the Yoke*
1. Introduction
The simile of a blind (or one-eyed) turtle, which surfaces every hundred
years, inserting its neck into a single hole in a (wooden) yoke that is
floating on a vast ocean is well known in Buddhist, Jain, and even
Brahmanical literature, where it is used to illustrate the rarity of
something occurring, such as birth as a human being.
Among the numerous Gåndhår¥ texts preserved in the Senior
collection of Kharo!†h¥ Buddhist manuscripts is a short sËtra for which
this powerful image is central.1 Appearing as the third of six texts
written on scroll 22r (ll. 31–56), which is one of the longest scrolls in
the collection, the sËtra represents a Gåndhår¥ version of the second of
two Påli suttas found in the Sacca-saµyutta of the Saµyutta-nikåya
which utilise this simile (nos. 56.47–48 ; V 455–57). The uddåna entry
for these two Påli suttas is chigga¬ena ca dve vuttå (S V 459,11),2 where
chigga¬a- “hole” is a reference to ekacchigga¬aµ yugaµ “yoke with a
single hole” of the simile. Based on this uddåna entry, the Burmese
edition (Be), for example, gives Dutiyacchigga¬ayuga-suttaµ as the title
of the second sutta (S no. 56.48 ; V 456,18–457,16).3 Although the
Gåndhår¥ sËtra lacks a title, the uddåna-like reference to it in the
*I would like to thank members of the Early Buddhist Manuscript Project

(Seattle), including Richard Salomon, Collett Cox, Timothy Lenz, and Stefan
Baums, for their comments on my reading of the Gåndhår¥ passage. I am also
indebted to Richard Salomon, Stefan Baums, Arlo Griffiths, and Oskar von
Hinüber for their remarks on several of my interpretations.
1For the Senior collection, see Salomon 2003 and my introductory chapter to
Glass 2007. I am currently writing a detailed catalogue of this collection

(Allon [forthcoming]) with financial support from the Australian Research
Council.
2Ee reads chiggalena against chigga¬a- of the text (see below).
3A title for this sutta is not recorded in the Påli commentaries.
230 Mark Allon
“index” scrolls nos. 7+8 (l. 6) is ekatarmao yuo “yoke with a single
hole” (see below for further discussion).4
There is apparently no Sanskrit parallel to this sËtra. Where the Påli
Saµyutta-nikåya has two chigga¬a suttas (nos. 56.47–48), the Chinese
Saµyuktågama (SÓ), Zá åhán j¥ng 雜阿含經, has only one, no. 406
(T 2 no. 99 108c6–20).5 The Chinese sËtra is a closer parallel to the first
of the two Påli suttas (no. 56.47). However, as it shares many elements
in common with our Gåndhår¥ sËtra and with the second Påli chigga¬a
sutta, it will be utilized in the following study. Like the Påli suttas, the
Chinese sËtra forms part of the saµyukta on the (four) truths, the
Dìxiångyìng 諦相應 (= Påli Sacca-saµyutta).6
In the Gåndhår¥ sËtra and in the second Påli chigga¬a sutta, the
simile illustrates the rarity of the occurrence of the optimal conditions
under which one may attain enlightenment, those conditions being the
presence of a Tathågata, his teaching the Dharma, and one’s birth as a
human being (according to the order of the Gåndhår¥). Both sËtras
advance the Four Noble Truths as the subject most worthy of attention
when these conditions are in place (this being the factor that qualifies
the Påli sutta for inclusion in the Sacca-saµyutta).
In the first Påli chigga¬a sutta and in the Chinese sËtra, the simile
illustrates the rarity of human birth only, as it does in many of the
occurrences discussed below. These two sËtras also refer to the Four
Noble Truths.
In an interesting article entitled “Middle Indo-Aryan Studies IX :
The Blind Turtle and the Hole in the Yoke” published in 1972,
Mr Norman discussed occurrences of this simile in Påli and Jain
literature. Space does not permit me to publish the full Gåndhår¥ sËtra
here, but as this Gåndhår¥ version of the simile contains several very
4For a briefdiscussion of the two “index” scrolls, see Salomon 2003, pp. 80–83,
and § 5 of my introductory chapter to Glass 2007.
5In Yìn Shùn’s reordering of the SÓ this sËtra is no. 598 (1983, Vol. 2, p. 130),
while according to the Fóguång (1983, Vol. 1, p. 22) it is no. 405.

6There are no uddånas in this section of the SÓ.
A Gåndhår¥ Version of the Simile of the Turtle 231
interesting features, the most notable among them being the preserva-
tion of an archaic word for the hole in the yoke, I present here an
edition, translation, and analysis of this section of the text in honour of
Mr Norman’s eightieth birthday and as a token of my admiration for his
scholarship and in gratitude for all I have learnt from him.7
2. The Gåndhår¥, Påli, and Chinese versions of the simile

The edition of the Gåndhår¥ text (RS 22r, ll. 33–43) presented here is
based on a reading of the digital colour and infrared images. Further
work on the manuscript is unlikely to significantly improve the reading
of this section of the text, unless some of the small, currently unplaced
fragments are found to belong to this section of the manuscript. In order
to save space, I have combined the edition and reconstruction. Text that
is difficult to read is contained within square brackets [ ] ; ak!aras of
uncertain reading are marked by a question mark ? ; reconstructions are
marked by an asterisk within parentheses (*).
The reading of the Påli parallel (S V 456,18–457,5) is based on the
four main published editions : European (Ee), Burmese (Be), Sinhalese
(Ce), and Thai (Se).8 The reading of the Chinese version follows the
TaishØ edition (T 2 no. 99 p. 108c7–14).
The numbering of the major divisions of the text (§§ 1, 2, 3) follows
that employed by the European edition (Ee) of the Påli. To facilitate
analysis of the text, I have subdivided § 3 into two subsections (§ 3a and
§ 3b) in the case of the Gåndhår¥ and Påli, and into three (§§ 3a–c) in the
case of the Chinese.
7The full text will be published in the near future along with the other texts on
this scroll in the series Gandhåran Buddhist Texts, Seattle : University of
Washington Press.
8The readings of Be and Ce are based on the electronic versions as the printed
editions were not available to me. See abbreviations for details.

232 Mark Allon
§ 1.
Gåndhår¥ 9bhayava ?10 eêa[ê aya] : (l. 33)
The Bhagavat said this :
Påli (bhagavå etad avoca) :11
The Bhagavat said this :
Chinese 爾時世尊告諸比丘。(p. 108c7)
Then the Bhagavat said to the monks :12
§ 2.
Gåndhår¥ sayaâavi bhik!ava aya mahapa"avi [34] [eko]êia asa. tatra
puru![e] ekatarmao yuo pak!ivea. [35] tam eâ purime vaêo paàimo
saharea paàimo vaêa 13puri14me [36] saharea utare vaê[a] dak!iô
saharea dak!iâ 15va[êe 16u17t](*a)18[r](*o)19 s(*a)h(*a)re(*a). [37]
20
atra hasa kaâ kachavo va!aßaêa umi[ jo] va!aßaêa[sa 21acaeˆ](*a)
[38] 22saha samiêa umic[e]a.23 (ll. 33–38)
“Monks, suppose that this great earth were one mass of water, and a
man were to throw a yoke with a single hole into it. An easterly wind
would carry it west ; a westerly wind would carry it east ; a northerly
wind would carry it south ; a southerly wind would carry it north. In it
9 Frag. A.
10The reading could be !u.
11Missing in Ee, Be, Ce, and Se (see commentary below).
12I am indebted to Lily Lee and Rod Bucknell for their comments on my trans-
lation of the Chinese text.

13Frags. A+Bb-2.
14Frag. A.
15Frags. A+Am-5.
16Frags. A+Am-2.
17Frag. A.
18Frag. Bc-2.
19The original reading could have been -ro or -ra, but not -re.
20 Frag. A.
21Frags. A+B.
22Frag. A.
23The scribe discontinued writing this line to avoid writing across the raised join
between two sections of bark.

there were a blind turtle which, emerging after a hundred years, with
the passing of a hundred years, would emerge over and over.
Påli seyyathåpi bhikkhave ayaµ mahåpathav¥24 ekodakå 25 assa. tatra
puriso ekacchigga¬aµ26 yugaµ pakkhipeyya. tam enaµ puratthimo27
våto pacchimena saµhareyya pacchimo våto puratthimena
saµhareyya uttaro våto dakkhiêna saµhareyya dakkhiô våto
uttarena28 saµhareyya. tatrassa kåô kacchapo so29 vassasatassa
vassasatassa accayena sakiµ sakiµ ummujjeyya (p. 456,18–24).
“Monks, suppose that this great earth were one mass of water, and a
man were to throw a yoke with a single hole into it. An easterly wind
would carry it west ; a westerly wind would carry it east ; a northerly
wind would carry it south ; a southerly wind would carry it north. In it
there were a blind turtle which would emerge once each time with the
passing of each hundred years.
Chinese 譬如大地悉成大海。有一盲龜壽無量劫。百年一出其頭。海中
有浮木。止有一孔。漂流海浪。隨風東西。(p. 108c7–9)
“Suppose the great earth were completely covered by a great ocean,
and there were a blind turtle of long life, an immeasurable kalpa [in
duration], which poked his head out every hundred years. In the ocean
there was a floating piece of wood with only one hole, floating on the
ocean waves to the east and west according to the wind.”
§ 3a
Gåndhår¥ 30[39] ta ki mañasa avi â ’!e kaô kachava va!i[aê](*a)sa ajaeâ
saha â(*am)[i][40]êa umi[jata] am[a]spi ekatarmao yuo grive
pak!ivea. (ll. 39–40)
What do you think, would this blind turtle, emerging over and over,
with the passing of a hundred years, insert its neck into that yoke with
the single hole ?
24~pa†hav¥ in Ce and Se.

25ekodikå in Se, Ce and in the two Sinhalese mss (S1, 3) used for Ee (S2 was not
used by the editor for this volume of S).
26~chiggalaµ throughout Ce.
27purimo in the Sinhalese mss (S1, 3) used for Ee.
28uttare in the Sinhalese mss (S1, 3) used for Ee .
29so is missing in the Sinhalese ms S3 used for Ee.
30Frag. B.
234 Mark Allon
Påli taµ kiµ maññatha bhikkhave api nu so31 kåô kacchapo vassasatassa
vassasatassa accayena sakiµ sakiµ ummujjanto 32 amusmiµ 33 eka-
cchigga¬e yuge g¥vaµ paveseyyå ti (p. 456,25–457,2).
What do you think, monks, would that blind turtle, emerging once
each time with the passing of each hundred years, insert its neck into
that yoke with the single hole ?
Chinese 盲龜百年一出其頭。當得遇此孔不。(p. 108c10)
Would the blind turtle, poking his head out every hundred years, meet
this hole or not ?
§ 3b
Gåndhår¥ aêicam eêa bhayava [41] suêalavam eva ya e!e kaô 34 kachavo
va!ihaêa omica va!ihaêa35sa aca[e]36[42]â saha samiêa37 [43]
umijata amaspi ekatarmao yuo grive pa"i[mu](*ce)a 38â [va]
pa"imu[ce](*a). (ll. 40–43)
It would be by chance, Bhagavat, it is very rare indeed, that this blind
turtle, emerging after a hundred years, with the passing of a hundred
years, emerging over and over, would put that yoke with the single
hole on his neck, or he may not put it on.
Påli adhiccam idaµ bhante yaµ so kåô kacchapo vassasatassa vassa-
satassa accayena sakiµ sakiµ ummujjanto amusmiµ ekacchigga¬e
yuge g¥vaµ paveseyyå ti (p. 457,3–5).
It would be by chance, venerable sir, that that blind turtle, emerging
once each time with the passing of each hundred years, would insert
its neck in that yoke with the single hole.
31kho in Se, Be and in the two Burmese mss (B1, 2) used for Ee.
32ca is added in the Sinhalese mss (S1, 3) used for Ee.
33amukasmiµ in the Burmese ms B2 used for Ee.
34 The scribe has marked both the e and o måtrås on ˆ. He probably wrote the o
måtrå second in accordance with the spelling elsewhere.

35êa could also be read as ha.
36The bottom of the e ak!ara is preserved on frag. Bb-2.
37The scribe has left the line short to avoid writing across the join between two
sections of bark.
38Frags. B+Bd-1.
Chinese 阿難白佛。不能。世尊。所以者何。此盲龜若至海東。浮木隨
風。或至海西。南．北四維圍遶亦爾。不必相得。(p. 108c10–13)
Ónanda said to the Buddha : “It would not be able to, Bhagavat. Why
is that ? If this blind turtle should arrive in the eastern part of the
ocean, the floating piece of wood might, according to the wind, arrive
in the west, south, or north of the ocean. Going around the four
directions in this way, they would certainly not meet each other.”
§ 3c
Chinese 佛告阿難。盲龜浮木。雖復差違。或復相得。(p. 108c13–14)
The Buddha said to Ónanda, “Although the blind turtle and the
floating piece of wood may miss [each other], perhaps they may also
meet each other.”
The wording of this section, which is missing in the Gåndhår¥ and

Påli, seems odd, unless the Buddha is merely confirming that they may
or may not meet each other.
3. Commentary
§ 1. The Buddha addresses the monks
The Såvatthi–Jetavana nidåna and initial interchange between the
Buddha and the monks are missing in the Påli editions and in the
manuscripts used for them. However, as this passage is given in full in
the first sutta of the Mahåvagga of the Saµyutta-nikåya in Ee (V 1), Be,
and Ce (and elsewhere throughout the Mahåvagga) and in the first sutta
of the Sacca-saµyutta in Ce and the Sinhalese manuscripts (S1, 3) used
for Ee, it has clearly been omitted through scribal abbreviation.
Undoubtedly, this passage would have been included when this sutta
was recited, as would the conclusion to this sutta which is also
abbreviated in the Påli editions and manuscripts (see Allon 2001,
pp. 253–55).
The nidåna of this Gåndhår¥ sËtra is repeated in the next sËtra on this
scroll (RS 22r, ll. 57–59). It also occurs twice in a collection of three
Ekottarikågama-type sËtras (EÓ-G) preserved in the British Library
Kharo!†h¥ collection (Allon 2001, pp. 225–32, 253–55).
bhayava ? eêa[ê aya] (l. 33) : The reading in the line 57–59 example
of this phrase is bhayava [!u] eêaê aya. The ak!ara of uncertain reading
236 Mark Allon
(?) in line 33 could be read as [!u], as in this second example. The
corresponding Påli expression is bhagavå etad avoca. The nominative
singular corresponding to Påli bhagavå / Skt bhagavån is bhayava or
bhaãava throughout the Senior manuscripts and bhayavadu in the EÓ-G
(Allon 2001, pp. 113–14). However, in the one complete example of the
equivalent of Påli bhagavå etad avoca in EÓ-G (l. 28) the reading is
bha[ya]va[!u] rather than bhayavadu as expected. In my edition of that
text I took this to be a scribal error and amended it to bhayava<*du>
(Allon 2001, pp. 225–26, 232), but these two examples in the Senior
manuscripts suggest that this amendment may not have been justified.
Given that bhayava / bhaãava is the nominative singular in all other
contexts in the Senior manuscripts, it is possible that we should take
!u39 as a separate word rather than as the termination of bhayava. The
most likely explanation is that !u is the equivalent of the Sanskrit
particle sma (or possibly su), which is attested elsewhere in Gåndhår¥ as
sa, !a, su, and possibly !u,40 although the usage of this particle in such a
context (e.g. Påli bhagavå (s)su etad avoca or the like) is not attested in
Buddhist literature to my knowledge.
Brough (1962, pp. 228–29) noted that hi !a in the Khotan Dharma-
pada (160d) corresponds to Âgvedic hi !ma, in contrast to ha ve of the
Påli version, which corresponds to ha vai of the Bråhmaâs, and
concluded that “[w]e have thus the interesting situation that the Prakrit,
from the North-west, appears to represent the survival of a common
Âigvedic usage, while the Pali, from a more central region, has instead a
group which is most familiar from the Bråhmaâs”. If !u in bhayava !u
in the Gåndhår¥ manuscripts discussed here does correspond to Skt sma,
then this would represent another instance of the archaic usage of this
39Although ! and k are indistinguishable in the Senior manuscripts, in EÓ-G

they are not. The reading is therefore unlikely to be ku in the Senior examples.
40See Norman (2004, p. 128) for the Påli, BHS, and Gåndhår¥ forms, and for
references to previous discussions. For the Gåndhår¥ spelling !u, see Norman
1971B, p. 218 = Collected Papers, Vol. 1, p. 118.
particle being attested in Gåndhår¥ (for which compare the appearance
of the archaic tardman- in this text discussed below).
For eêaê aya, which corresponds to the Påli expression etad avoca,
see Allon 2001, pp. 163–65 where it was transcribed as e!a! aya.
§§ 2–3. The simile of the blind (or one-eyed) turtle
As noted by Mr Norman (1972) in his article on this topic, the simile of
the blind turtle inserting its neck into a single hole in a yoke floating in
the ocean is referred to in Ther¥gåthå 500 as illustration of the rarity of
being born a human being :
sara kåâkacchapaµ pubbasamudde, aparato ca yugachiddaµ
sirã41 tassa ca pa†imukkaµ, manussalåbhamhi opammaµ.
In his translation of the Ther¥gåthå, Mr Norman (1971A , p. 49) trans-

lated this verse as “Remember the blind turtle in the sea in former times,
and the hole in the yoke floating [there] ; remember the putting on of it
(= the yoke) as a comparison with the obtaining of human birth.” But in
view of his later comments (Norman 1972, pp. 157–58), the first line
would be better translated “remember the blind turtle in the eastern
ocean, and the hole in the yoke [floating] from the western [ocean]”.42
In his 1972 article, Mr Norman also quotes several other Påli references
for this simile (M III 169,9–16 ; S V 455,24–29 [the first of the two
chigga¬a suttas currently under view] ; Mil 204,11–13 ; As 60,17–18) and
examples from Jain literature.
In a much earlier article on this topic, Harinåth De (1906–1907,
pp. 173–75) refers to two similar passages in Sanskrit Buddhist
literature. The first appears in Íåntideva’s Bodhicaryåvatåra (4.20) :
41The reading sirã (= sira[µ]) was proposed by Alsdorf in the European edition
the Th¥ (p. 248). Ce has sara (cf. Mr Norman’s translation).
42Winternitz (1913, p. 44 = Kleine Schriften, 1991, p. 547) translates the verse
as “Denke an das Gleichnis für die Erlangung einer Wiedergeburt als

Mensch : an die einäugige Schildkröte und das im Ozean nach Osten und
Westen herumgetriebene Loch des Joches und daran, ob der Kopf dieser
(Schildkröte in jenem Loch) stecken bleibe.”
238 Mark Allon
ata evåha bhagavån, månu!yam atidurlabham
mahårâvayugacchidrakËrmagr¥vårpaôpamam
translated by Crosby and Skilton (1995, p. 26) as “That is why the

Fortunate One declared that the human state is so hard to attain ; as
likely as the turtle poking its neck through the hole of a yoke floating on
the mighty ocean.”
The second is found in a prose passage in the Saddharmapuˆ"ar¥ka-
sËtra (463.4–5) :
durlabho hy amba tåta buddhotpåda udumbarapu!pasadÁßo mahårâ-
vayugacchidrakËrmagr¥våpraveßavat. durlabhaprådurbhåvå amba tåta
buddhå bhagavanta˙.43
For, father and mother, the appearance of a Buddha is rare as a flower on
a fig tree, as the likelihood of the turtle inserting its neck into the hole of
the yoke floating on the great ocean. Father and mother, Buddhas,
Bhagavats are ones whose appearance is rare.
Kumåraj¥va’s Chinese translation has 又如一眼之龜, 值浮木孔 “and as

the one-eyed turtle meeting the hole in the floating log” (T 9 no. 262
p. 60a29–b1).
In an article on the same topic, A.N. Upadhye (1972) quotes many
examples from Jain literature and the following Buddhist verse from
MåtÁce†a’s Adhyardha-ßataka (5) :44
so ’haµ pråpya manu!yatvaµ, sasaddharmamahotsavaµ,
mahårâvayugacchidrakËrmagr¥vårpaôpamam
I, attaining the human state, accompanied by the good Dharma which is
the great festival, which is as likely as the turtle poking its neck through
the hole of a yoke floating on the mighty ocean …
Not surprisingly, the image is also found in Brahmanical texts. An
example is the Laghu-Yogavåsi!†ha 6.15.14 (= YV 6a.126.4), which is
43Ed. Kern and Nanjio. One of the Central Asian manuscripts refers to the
single hole : (*uduµ)barapu!pasadÁßa(s) tåta tathågata yugamitaikacchi ///
(437a6) (ed. Toda 1983, p. 214).
44Cf. Winternitz (1913, pp. 46–47 = Kleine Schriften, 1991, pp. 549–50) for the
Chinese version.
quoted verbatim in the eighteenth century Bodhasåra by Narahari
(14.2.17) :45
calårâvayugacchidrakËrmagr¥våpraveßavat
anekajanmanåm ante, vivek¥ jåyate pumån
Like the turtle inserting its neck into the hole in the yoke floating on the
agitated ocean, a person becomes discerning at the end of many births.
There are numerous examples in Chinese and Tibetan Buddhist

texts, but my primary concern here is with the wording of the Indian
versions.46
sayaâavi bhik!ava aya mahapa"avi [eko]êia asa (ll. 33–34) : Påli
seyyathåpi bhikkhave ayaµ mahåpathav¥ ekodakå assa.
For sayaâavi = Påli seyyathåpi, see Allon 2001, p. 209.
The spelling of pa"avi in maha-pa"avi, which corresponds to Skt
pÁthiv¥ / Påli pathav¥, pa†hav¥,47 is a further example of “[s]poradic non-
etymological alternations between unaspirated and aspirated con-
sonants” in the Senior manuscripts noted by Salomon (2003, p. 86), for
which cf. also aêica = Påli adhiccaµ in § 3b (l. 40) below and pa"ama
= Skt prathamam / Påli pa†hamaµ in another Senior manuscript
(12.31). Previously attested spellings of this word in Gåndhår¥ are
pra†havi and pa"havi found in inscriptions.48
[eko]êia : The reading is ekodakå in Ee and Be of the Påli parallel,
but ekodikå in Se and Ce, and in the two Sinhalese manuscripts (S1, 3)
45For the commentary on the Bodhasåra verse, see Jacob 1909. I would like to
thank Jenni Cover for bringing the Bodhasåra reference to my attention and
for providing the context to it and Walter Slaje for drawing my attention to the
Jacob article and for verifying the details of the Yogavåsi!†ha reference.
46The occurrences of the simile in the Chinese translation of the SËtrålaµkåra
are discussed by Winternitz (1913, pp. 45–46 = Kleine Schriften, 1991,

pp. 548–49), who quotes Huber’s French translation of these. The Central
Asian manuscript of this text (Kalpanåmaˆ"itikå) edited by Lüders (1926,
p. 156 ; frag. 123 R2) preserves the words tadyugacchidram.
47See PED s.v. pa†hav¥. For references to †h < th when preceded by Á or r in
Påli, see Oberlies 2001, p. 80.

48pra†havi : Månsehrå inscription of the year 68, line 9 (Konow 1929, p. 20) ;
pa"havi : Ajitasena inscription, line 5 (Fussman 1986).

240 Mark Allon
e 2
used for E (S was not used by the editor for this volume of S). The
CPD (s.v. ekodaka) questions the variant reading ekodikå as a possible
feminine form. However, the SWTF (s.v. ekodaka) accepts Sanskrit
ekodikå on the basis of its occurrence in the similar phrase ekodikåyåµ
mahåpÁthivyåµ ekårâvåyåµ found in a fragment of the Sanskrit
version of the Aggañña-sutta from Central Asia ;49 for which also cf.
tadyatheyaµ mahåpïthivy ekodakajåtå bhavet in the Larger Sukhå-
vat¥vyËha (56.4).50 Gåndhår¥ ekoêia could go back to a form such as
ekodikå in the MIA version from which this text was translated or have
resulted from palatalisation of a under the influence of secondary -y-
(ekodakå > *ekodayå > ekoêia), which is common in Gåndhår¥. It is, of
course, possible that an MIA palatalised form lay behind ekodikå in a
Central Asian Sanskrit manuscript.
ekatarmao yuo (l. 34) “a yoke with a single hole” : Påli eka-
cchigga¬a- yuga-. This occurs three times in this sËtra : once in the
accusative singular tatra puru![e] ekatarmao yuo pak!ivea (l. 34) “a
man were to throw a yoke with a single hole into it [the ocean]”, where
the Påli has tatra puriso ekacchigga¬aµ yugaµ pakkhipeyya, and twice
in the locative singular am[a]spi ekatarmao yuo grive pak!ivea (l. 40)
“would insert its neck into that yoke with the single hole” and amaspi
ekatarmao yuo grive pa"i[mu](*ce)a (l. 43) “would put that yoke with
the single hole on his neck” (see below), where the Påli has amusmiµ
ekacchigga¬e yuge g¥vaµ paveseyya. As stated above, the words eka-
tarmao yuo also occur on scrolls 7+8 (l. 6) as the uddåna-like reference
to this sËtra.
An etymology for Påli chigga¬a, chiggala “hole”, which is recorded
in the Påli canon in eka-cchigga¬a-yuga- “a yoke with one hole” of the
current context and in tå¬a-cchigga¬a, tåla~ “key hole”, is not given by
the dictionaries and grammars (e.g. PED s.v.),51 where it is often
49See SWTF s.v. ekodaka for references. The Mahåvastu version (I 339.7, ed.
Senart) has ayam api mahåpÁthiv¥ udakahradaµ viya samudågacchet.
50Ed. Ashikaga 1965.
51For tå¬a-cchigga¬a, ~cchidda, see von Hinüber 1992, pp. 17, 31.
compared with Påli chidda- / Skt chidra- “hole”.52 Nor is it listed in the
CDIAL. As noted by O. von Hinüber (1992, pp. 17, 31) the word is
certainly non-Aryan in origin,53 although a comparable word (Påli
chigga¬a-) is not listed in A Dravidian Etymological Dictionary
(Burrow and Emeneau 1961 and 1968).54
The Gåndhår¥ word ekatarmao is to be taken as a compound of eka-
and -tarmao,55 where tarmao must be the equivalent of Skt tardman-56
n. “hole” (< !tÁd “to cleave, pierce” ; MW and CDIAL s.v.), with -ka
suffix. The word has not been recorded previously in Gåndhår¥ and
there appears to be no Påli or other MIA equivalent.57 Based on the
model of the development of Skt vartman- “road” to va††a- in Prakrit or
va†uma- (cf. vattan¥) in Påli, va††amaya-, va"ömaga- in Prakrit (CDIAL
s.v. tárdman-), the expected MIA forms for Skt tardman- would be
*ta""a- with -rd- > -""- or possibly *tadda- with rd > dd,58 both
involving the last member of the cluster -rdm- not being taken into
consideration in the assimilation of the cluster,59 or *ta"uma- or
52The Abhidhånappad¥pikå-†¥kå (Nandawansa 2001, p. 239) gives chinditvå

gacchat¥ ti chiggalaµ ; yadådi.
53The similarity between Pkt ga"a- “hole” (CDIAL s.v. ga"a-1), corresponding
to Skt garta- “hole” (cf. ßamyå-garta- “hole (garta) for the yoke pin (ßamyå)”,
CDIAL s.v.), and the latter part of chigga¬a- is therefore merely coincidental.
54If the word does have a Dravidian origin, then the latter part of the word may
be connected with Malayalam a¬a “hole” listed in this dictionary (1961, § 261).
55In this scribe’s hand t and d are indistinguishable. However, in view of the
interpretation given below, the reading -darmao and a connection with dåra-
“hole” (MW s.v. ; see also CDIAL s.v. dåra-1) is unlikely.
56Although tardman- is only found in the Vedic corpus, most Sanskrit words
quoted in this paper are not (e.g. kha-, chidra-). For the sake of consistency, I
have therefore omitted the accent from all Sanskrit words in my discussion,
including tardman-.
57None are listed in CDIAL s.v. I am indebted to Mr Norman for verifying this
absence.
58Pischel 1965, §§ 288–91 ; von Hinüber 2001, § 256.
59Pischel 1965, § 334 ; von Hinüber 2001, § 260 ; Oberlies 2001, § 17.
242 Mark Allon
*ta""uma- with the cluster split by epenthesis.60 The Gåndhår¥ form,
with its apparent omission of the dental stop, is unexpected. As the
cluster rd is regularly retained in Gåndhår¥, we would have expected
*tardao or *tardumao for this word. Alternatively, the development
could have been -rdm- > *-rmm- > -rm- following von Hinüber 2001,
§ 261. Noteworthy is the spelling tarman- throughout Kaul’s edition of
the Laugåk!igÁhyasËtra mentioned below, which is transmitted only in
Kashmiri manuscripts.
In the Ther¥gåthå verse (500) quoted above, the word for hole is
chidda- (yuga-[c]chiddaµ) = Skt chidra- in contrast to chigga¬a- of the
Saµyutta-nikåya sutta under discussion, for which compare the example
of the older tå¬a-cchigga¬a of the Saµyutta-nikåya being replaced by
tå¬a-cchidda in the Vinaya discussed by von Hinüber (1992, pp. 17, 31).
The interchangeability of chigga¬a- and chidda- in the context of a yoke
in Påli texts is also seen in the various commentaries on sammå- (Skt
ßamyå) “yoke pin” of the Brahmanical sacrifice sammåpåsa- (Skt
ßamyåpråsa-) mentioned in a verse found in several places in the Påli
canon (e.g. Sn 303 ; S I 76,20 ; A II 43,30) : samman ti yugacchidde
pakkhipitabbadaˆ"akaµ (e.g. Spk-† I 180 Be ; Mp-† II 299) and …
sammåpåso, yugacchigga¬e pavesanadaˆ"akasa#khåtaµ sammaµ
khipitvå … (It-a I 94,20–21 ; Mp IV 70,11–13).61
In the Buddhist Sanskrit examples mentioned above, the word is
yuga-cchidra-, as it is in the Jain examples quoted by Norman (1972)
and Upadhye (1972) : Skt yuga-cchidra- or Pkt juga-chi""a-.
As already noted, the equivalent of Skt tardman- is not found in
Påli texts. The word does not appear, for example, in any commentarial
gloss or in the list of words for hole in the Abhidhånappad¥pikå,
namely, randhaµ tu vivaraµ chiddaµ kuharaµ susiraµ bilaµ susi ’tth¥
chiggalaµ sobbhaµ (649–50),62 which is clearly based on the Amara-
60Pischel 1965, § 139 ; von Hinüber 2001, §§ 152–56.

61Cf. also those reported in Mr Norman’s notes on Th¥ 500 (Norman 1971A,
pp. 174 f.).
62Nandawansa 2001, p. 239.
koßa,63 where the word is similarly absent : … kuharaµ su!iraµ vivaraµ
bilam | chidraµ nirvyathanaµ rokaµ randhraµ ßvabhraµ vapå su!i˙
… (1.8.486–89).64 Again, the word does not appear to be recorded in
Buddhist literature in Sanskrit (e.g. the word is not listed in SWTF).
Although apparently not found in Buddhist literature, tardman- in
conjunction with yuga-, meaning hole in the yoke, does occur in early
Brahmanical literature.65 The earliest attested example is the Atharva-
veda (Íaunaka) 14.1.40 (= AVP 18.4.9) :66
ßáµ te híraˆyaµ ßám u santv !õpa˙ ßáµ methír bhavatu ßáµ yugásya
tárdma
ßáµ ta õpa˙ ßatápavitrå bhavantu ßám u pátyå tanvà1µ sáµ spÁßasva
translated by Whitney (1905, p. 747) as

Weal be to thee gold, and weal be waters ; weal be the post (methí), weal
the perforation (tárdman) of the yoke ; weal be for thee the waters
having a hundred cleaners (-pavítra) ; for weal, too, mingle thyself with
thy husband.
The following verse (AVÍ 14.1.41 = AVP 4.26.7), which also refers
to the hole in the yoke, is of some interest since it is taken from the
Âgveda (8.91.7) :
khé ráthasya khé ’nasa˙ khé yugásya ßatakrato
apålõm indra trí! pËtvõkÁô˙ sùryatvacam
63For the relationship between the Abhidhånappad¥pikå and the Amarakoßa, see
Nandawansa 2001, pp. xxvii–xxxi.
64Ed. Ramanathan n.d., Vol. 1, p. 146. The word does not appear in the com-
mentaries on the Amarakoßa edited by Ramanathan (pp. 146–47) or in the

Abhidhånacintåmaînåmamåla (1363–64), a similar lexigraphical work by
Hemacandra (eleventh–twelfth centuries), or its commentaries (ed. 2003,
pp. 624–25).
65Several of the following references are mentioned in Winternitz (1913,
pp. 43–44 = Kleine Schriften, 1991, pp. 546–47). I would like to thank Arlo
Griffiths for his responses to my questions on this Brahmanical material.
66Ed. Roth and Whitney 1856.
244 Mark Allon
In the hole of the chariot, in the hole of the cart, in the hole of the yoke,
O thou of a hundred activities, having thrice purified Apålå, O Indra, thou
didst make her sun-skinned (tr. Whitney 1905 , p. 748).
In this Âgveda verse the word for hole is kha- (khé yugásya “in the
hole of the yoke”) rather than tardman-, which does not occur in the
Âgveda, or chidra-, which occurs only once in the Âgveda (1.162.20) as
an adjective meaning “pierced”, “torn asunder” (MW s.v. chidrá).
These two verses of the Atharvaveda are not commented on by
Såyaâ,67 but in his commentary on the second verse as it occurs in the
Âgveda, kha- is glossed with chidra-.68
The compound yuga-tardman occurs several times in the
KaußikasËtra of the Atharvaveda, which the commentaries consistently
gloss with yuga-cchidra-.69 The three occurrences of tardman- in the
ÍrautasËtra of Kåtyåyana (6.1.30 ; 7.3.20 ; 15.5.27), which do not occur
in conjunction with yuga-, are similarly all glossed with chidra- by the
commentators (Karka and Yåjñikadeva).70 The Kå†hakagÁhyasËtra of
the Black Yajurveda (ed. Caland 1925), also known as the Laugåk!i-
gÁhyasËtra, contains two relevant sËtras. The first (25.9) quotes the
Âgveda verse (8.91.7) referred to above, which is also found in the
Atharvaveda (14.1.41), glossing khé yugásya with yugatardmani,71 then
67Ed. Vißva Bandhu et al. 1961, part 3, p. 1542.

68Ed. Rajwade et al., 1978, Vol. 3, p. 906.
69Extracts from the commentaries are given by Bloomfield (1890) in footnotes
to his edition of the KaußikasËtra and in an appendix (Paddhati of Keßava).

The references are 35.6 (p. 94 ; see also p. 336) ; 50.18 (p. 146 ; see also
p. 355) ; 76.12 (p. 203) ; cf. 76.13. For Dårila’s commentary, see the edition by
Diwekar et al. 1972, which only covers the first occurrence. Cf. also Caland’s
notes to his translation of the KaußikasËtra (1900, pp. 31, n. 5 ; 115, n. 5 ; 175,
n. 10).
70Ed. Weber 1972. Ranade ([1978]) translates tardman in these three occur-
rences as “cavity”, “holes”, and “perforations”, respectively (his numbering is

6.1.28 ; 7.3.17 ; 15.5.25).
71hiraˆyaµ ni!tarkyaµ baddhvådhy adhi mËrdhani dak!iâsmin yugatardmany
adbhir avak!årayate ßaµ te hiraˆyam iti.

quotes the preceding Atharvaveda verse (14.1.40), also quoted above.72
The second sËtra (26.3) is of similar wording, mixing kha- and
tardman-.73 The commentaries of Devapåla, Bråhmaâbala, and Óditya-
darßana presented in Caland’s edition of the Kå†hakagÁhyasËtra all gloss
khe and yugasya tardma of these two sËtras with yugacchidra- (Deva-
påla gives randhra- for kha- of the first sËtra). Finally, tardman- also
occurs in the Íatapathabråhmaâ (Mådhyandina 3.2.1.2), but the com-
mentaries of Såyaâ and Harisvåmin (published Delhi 1987) lack such a
gloss.
There are, no doubt, other instances of tardman- and yuga-tardman-
in Brahmanical literature, but the above will suffice for the current
purposes.74
Apart from yuga-tarmao of the Gåndhår¥ text under review, there is
only one other instance of a word meaning hole recorded in Gåndhår¥ to
my knowledge. This is chidra-, which appears in the Senior manuscripts
in the expression pa!aââa chid<*r>a 75 (20.6) “a hole in the stone (of
the city wall)” (Gåndhår¥ pa!aâ- = Skt på!åâ- “stone”). The Påli
parallel has påkåra-sandhiµ “a hole in the [city] wall” (e.g. S V 160,22–
23 ; A V 195,2) with v.l. påkåra-cchiddaµ recorded in the European
edition of the A#guttara-nikåya occurrence.76
72The reading throughout M. Kaul’s edition of the Laugåk!igÁhyasËtra is

tarman-.
73khe rathasya khe ’nasa˙ khe yugasya ca tardmasu khe ak!asya khe ava-
dadhåm¥ti yugatardmasu ßam¥ßåkhåm avadadhåti.

74The ÓpastambagÁhyasËtra (4.2.8, ed. U.C. Pandey) uses the word yuga-
cchidraµ.
75The reading appears to be chidva, but the context demands chidra. It appears
that the scribe accidentally overwrote the upward stroke of the post-
consonantal r making it appears like a post-consonantal v.
76chidra- also appears in the unrelated expression achidra-vuti in the Khotan
Dharmapada, where the corresponding Påli Dhammapada verse has the

equivalent acchidda-vutti- (= Skt acchidra-vÁtti-) “impeccable conduct”
(Dhp-GK 241c ; cf. achidra-vurti in Khvs-G 23a).
246 Mark Allon
The occurrence of yuga-tarmao in our Gåndhår¥ text suggests that
the equivalent of Skt tardman- was the word current in Gandhåra for the
hole in a yoke, in contrast to chidra- (or chigga¬a- in the case of some
Påli texts), which is the term found in Buddhist literature from other
regions and in Jain and later Brahmanical Sanskrit literature. Given that
tardman- is found in some early Brahmanical Sanskrit texts, but is
commonly replaced by chidra- in later texts, particularly commentaries,
it would appear that an old lexical item was preserved in usage in
Gandhåra. Of course, although the Senior manuscripts date to the
second or third century A.D. (Salomon 2003 , pp. 74–78 ; Allon et al.
2007) and yuga-tarmao may therefore have been the expression for the
hole in the yoke current in Gandhåra at that time, its usage cannot be
fixed more accurately in time and space. For instance, it remains
possible that this term was not current when these manuscripts were
written, but is rather a vestige of a much earlier period when this sËtra
was first translated (or transposed) into Gåndhår¥ from another MIA
dialect.
Finally, it is impossible to tell what word for hole the translators of
the Chinese texts had before them in their Indian originals. For example,
throughout the Saµyuktågama parallel and in Kumåraj¥va’s translation
of the Saddharmapuˆ"ar¥kasËtra simile, the term is kÌng 孔, “hole”.77
But what is interesting is that these two Chinese translations refer to a
piece of wood with (one) hole, rather than a yoke : the Saµyuktågama
parallel has 浮木止有一孔 “a floating piece of wood with only one
hole” (p. 108c8–9), while Kumåraj¥va has 浮木孔 “hole in the floating
log” (T 9 no. 262 p. 60b1).
tam eâ purime vaêo paàimo saharea (l. 35) : Påli tam enaµ
puratthimo våto pacchimena saµhareyya. The propensity for purima- to
replace puratthima- is witnessed in the Sinhalese manuscripts used for
Ee of the Påli, which read purimo for puratthimo, and in the Be reading
77Surprising is Wogihara’s (1979) actual listing of tardman in his Sanskrit–

Chinese–Japanese dictionary (s.v. tardman) ; the first of the two entries, xuè
穴, is also listed under chidra.
of the commentary on this word in another text : puratthimaµ disaµ
dhåvat¥ ti purimaµ (Ee puratthimaµ) disaµ gacchati (Nidd-a II 432,5–
6). For purima- as “eastern” in Påli, see BHSD s.v. purima.
The grammatical case of paàimo “western” in paàimo saharea, and
of the word which replaces it in each of the following parallel clauses
(purime, dak!iô, and ut(*a)r(*o)), must be accusative, in contrast to
the instrumental of the Påli pacchimena saµhareyya, etc.
atra hasa (l. 37) : The Påli has tatrassa = tatra assa “in it there
would be”. The Senior manuscripts contain several further examples of
the occasional writing of h plus vowel måtrå where a word-initial vowel
is expected : hiêam e[yi] (5.27) besides iêam eyi (5.21) = Påli idam
avoca ; himaspi (15.7) = Påli imasmiµ ; hidriaâ (5.34) besides idriaâ
(5.39) = Påli indriyånaµ ; and hirdhaüpaêaâ (5.34,38) = Påli iddhi-
pådånaµ (see Glass 2007 : § 5.1.1.). The expected initial vowel in each
of these latter Senior examples is the palatal vowel i. Several
interpretations of this phenomenon were presented in my discussion of
the word hasavaro = Skt / Påli athåparaµ (which is preceded by
sughadu = Påli sugato) found in the EÓ-G (l. 16), including that it
represents sandhi h-, an “easternism”, or “emphatic” h-.78 Although I
considered sandhi h- to be the most likely of these interpretations, the
rarity of the phenomenon led me to dismiss it. Rather, I concluded that
h- in these words represented the “sporadic, and as yet to be fully
understood, appearance of prothetic h- in Gåndhår¥” (Allon 2001 :
181).79
It is interesting that the majority of examples of prothetic h- in the
Senior manuscripts appear in palatal environment. This parallels the
occasional appearance of glide h in internal position, most examples of
78Allon 2001, pp. 180–81 ; cf. 102 ; add Oberlies 2001, p. 75 ; Norman 2002,
p. 227.
79It is tempting to see this phenomenon as purely graphic, since the only
difference between a word-initial vowel (e.g. i-) and h plus that vowel (e.g.
hi-) in Kharo!†h¥ is that the latter has a short horizontal stroke to the right at
the bottom of the ak!ara.
248 Mark Allon
which also appear in palatal environment.80 Examples from the Senior
manuscripts are uêahivadre (2.65[v29]), besides uêaïvadra (2.9) = Skt
udåyibhadra- / Påli udåyibhadda- “[prince] Udåyibhadra” ; bramahia
(19.13,30), besides bhamaïo (17.10) = Påli brahmacariya- “the holy
life” ; sahiâ- (12.10) = Påli såyaˆha- “evening”. For examples in the
Khotan Dharmapada, see Brough 1962 : § 39. In other words, when they
do occur there is a marked tendency for prothetic h- and glide -h- to
appear in the neighbourhood of palatal vowels, which may be due to the
palatal character of h (Wackernagel 1957 : §§ 213–16 ; Burrow 1973 :
77ff. ; von Hinüber 2001 : § 223).
kaâ kachavo (l. 37), kaô kachava (l. 39), kaô kachavo (l. 41) :
Påli kåô kacchapo. Mr Norman (1971A : 49) and Bhikkhu Bodhi
(2000 : 1871–72) translate the Påli as “blind turtle”, while Winternitz
(1913 : 44 = Kleine Schriften 1991 : 547) gives “one-eyed turtle
(einäugige Schildkröte)”. The Chinese Saµyuktågama parallel has
“blind turtle”, máng gu¥ 盲龜, while Kumåraj¥va’s translation of the
Saddharmapuˆ"ar¥kasËtra simile quoted above has “one-eyed turtle”
y¥yƒn zh¥ gu¥ 一眼之龜. These translations reflect the dual meanings of
one-eyed and blind (in both eyes) for kåâ-, which are attested in the
Påli commentarial glosses (see DOP and CPD s.v. kåâ-), such as kåô
ti ekakkhikåô vå ubhayakkhikåô vå (Ps IV 231,21).81
The phrases expressing the number of years after which the turtle
would surface in §§ 2, 3a, and 3b and their Påli counterparts are best
discussed together. They are
§2
Gåndhår¥ va!aßaêa umi[jo] va!aßaêa[sa acaeˆ](*a) saha samiêa umic[e]a
(ll. 37–38)
Påli vassasatassa vassasatassa accayena sakiµ sakiµ ummujjeyya
80For h as glide, see Brough 1962, § 39 ; Norman 1979, pp. 323–24 (= Collected
Papers, Vol. 2, pp. 75–76) ; von Hinüber 2001, § 274 ; Allon 2001, p. 102.
81The Be of the commentary on the Dutiyacchigga¬asutta (Spk III 302,17) and a
Burmese and a Sinhalese manuscript used for Ee read andha-kacchapassa

against kåâ-kacchapassa of Ee.
§ 3a
Gåndhår¥ va!i[aê](*a)sa ajaeâ saha â(*am)[i]êa umi[jata] (ll. 39–40)
Påli vassasatassa vassasatassa accayena sakiµ sakiµ ummujjanto
§ 3b
Gåndhår¥ va!ihaêa omica va!ihaêasa aca[e]â saha samiêa umijata (ll.
41–43)
Påli vassasatassa vassasatassa accayena sakiµ sakiµ ummujjanto
The Gåndhår¥ and Påli differ in several ways. Where the Påli has
vassasatassa vassasatassa accayena “with the passing of each hundred
years” in each section (§§ 2, 3a, 3b),82 the Gåndhår¥ has va!aßaêa
(/va!ihaêa) umi[jo] (/omica) va!aßaêasa (/va!ihaêasa) acaeâ saha83
“emerging after a hundred years, with the passing of a hundred years” in
§ 2 and § 3b, but va!i[aê](*a)sa ajaeâ saha “with the passing of a
hundred years” in § 3a. I take saha of the Gåndhår¥ to be the equivalent
of Påli/Skt saha “with”, rather than sakiµ of the Påli parallel, although
the expression accayena saha is not attested in Påli. The Chinese
parallels the Påli with bƒi nián y¥ 百年一 “every hundred years”.
The interpretation of umi[ jo], omica in va!aßaêa umi[ jo] (l. 37),
va!ihaêa omica (l. 41) is problematic. Although faint, the final o vowel
in umi[jo] is certain. This spelling suggests that umi[jo], omica is the
present participle nominative singular masculine of the verb cor-
responding to Skt ut-!majj construed with the ablative (or accusative ?),
the phrase meaning “emerging after a hundred years”. The Påli equiva-
lent would be vassasatå ummujjaµ, which is not recorded. However,
the nominative singular of this participle appears as umijata in the
following lines (ll. 40, 43), where the Påli has ummujjanto. It may
therefore represent the gerund of this verb, which appears in Sanskrit as
82The reading in Ee of the Majjhima-nikåya occurrence (M III 169,13–14) is

vassasatassa accayena sakiµ ummujjeyya, with the v.l. in two Sinhalese
manuscripts of vassasatassa vassasahassassa vassasatasahassassa accayena
…, which is also recorded in Be as the Sinhalese reading.
83I have removed the square brackets where the reading is verified by the
repetitions.
250 Mark Allon
unmajya or unmajjya (MW s.v. un-majj), but in Påli as ummujjitvå.
However, the final o in umi[jo] is unexpected in a gerund. Both
interpretations would give more or less the same meaning. I translate
this phrase as “emerging after a hundred years”, which covers both
possibilities.84
The optative third singular of the same verb appears as umic[e]a in
line 38, where the Påli has ummujjeyya. In contrast to the Påli verb,
which shows labialisation of the root vowel a under the influence of the
preceding labial consonant cluster mm,85 the Gåndhår¥ shows palatali-
sation of the vowel under the influence of the following palatal
consonant cluster (original jj).86
The alternation between j and c, as witnessed in the Gåndhår¥
spellings of these words, is common in the Senior manuscripts. The
reflexes of original intervocalic -jj-, -cc-, -j- and -c-, and of initial j- and
c- may appear as j or c in this scribe’s orthography. As noted by
Salomon (2003 : 87), this suggests “that this scribe, and presumably at
least some other contemporary speakers of Gåndhår¥ as well, did not
distinguish between c and j in their dialect”.
The Gåndhår¥ equivalent of Skt var!aßata- / Påli vassasata- “hundred
years” shows three spellings : (i) va!aßaêa- (l. 37 [Ÿ 2]), which could be
read as va!ayaêa- since ß and y are indistinguishable in this scribe’s
hand ; (ii) va!i[aê.] (l. 39) ; and iii) va!ihaêa- (l. 41 [Ÿ 2]). I transcribe
the two line 37 examples as va!aßaêa- rather than va!ayaêa- on the
basis of the spelling of this word in Gåndhår¥ inscriptions and in the
Khotan Dharmapada. Examples from the latter document are var!a-
ßada- (141a) and va!a-ßada- (316a). The above spellings va!i[aê.] and
va!ihaêa- show the palatalisation of final a of Skt var!a- under the
84A neuter noun ummujja- “emerging” is recorded in Påli (CPD s.v.). But this is
not likely here.
85Berger 1955, p. 60 ; Norman 1976B, p. 45 (= Collected Papers, Vol. 1,
p. 250) ; cf. CDIAL s.v. únmajjati.

86Cf. Norman 1976a and 1983.
influence of the following initial ß of ßata-,87 with the development -ß- >
-h- in the case of va!ihaêa- and further weakening (-ß- > -h- > Ø) in the
case of va!i[aê.]. The same sound changes (-ß- > -h- and -ß- > -h- >
Ø),88 but this time in the context of a preceding or following original
palatal vowel, are seen in yoîho = BHS yonißa˙ “thoroughly” (l. 36)
and baïhodu (l. 82), baïhoda (l. 110) = Skt vaß¥bhËta˙ found in the
Gåndhår¥ Anavataptagåthå currently being edited by Richard Salomon.
A further example is the Gåndhår¥ spelling for the place name
tak!aß¥la-, which appears as tak!aïla- besides tak!aßila-.89 The above
can be compared with the development of original -h- > -ß- = [!] and -h-
> Ø in palatal environment. Examples of the former are (i) ißa = Skt iha
“here” found in the Niya documents, inscriptions, and in some of the
British Library manuscripts90 and iße found throughout the Senior
manuscripts (e.g. 2.37) and in some inscriptions, which shows palatali-
sation of final a ; and (ii) the Gåndhår¥ version of the epithet of king
Ajåtaßatru found in the Senior manuscripts : vedißaputra- (2.21, 31, 44,
etc.), veêißaputra- (2.13, 18, etc.) = Skt vaideh¥putra- / Påli vedehi-
putta-.91 Examples of the latter development (-h- > Ø in palatal
87Cf. Gåndhår¥ avißißadi, which probably = Skt avaßi!yate (Dhp-G K 200 ; see
Brough 1962, p. 243 ; Norman 1976A, p. 334 [= Collected Papers, Vol. 1,
p. 227]). For an example in Påli, see Norman 1983, p. 277 (= Collected
Papers, Vol. 3, p. 15).
88Cf. the examples of ß > h in Gray 1965, § 401, and -s- > -h- > Ø in the
Gåndhår¥ Dharmapadas discussed in Lenz 2003, p. 43 (see also von Hinüber

2001, § 221).
89Examples of tak!aïla- are found in the British Library manuscripts (e.g.
BL16+25, ll. 45–46 [see Lenz 2003, pp. 182–83] ; and BL2, l. 7) ; for
references to examples of tak!aßila- in the inscriptions, see Konow 1929
index.
90E.g. BL 16+25, ll. 21, 29, 32, 43 where the spelling alternates between ißa and
ißa (see Lenz 2003, pp. 155–56). For comments on this word, see Burrow
1937, §§ 17, 91.
91Note that the original final palatal vowel in vaideh¥- is not marked.
252 Mark Allon
environment) are Aßokan ia = iha found at Shåhbåzga®h¥92 ; sabaraka-
idaï = Skt samparåya-hitåya “for the benefit of the next life” and
vayari- = Skt vihårin- in the Khotan Dharmapada (Brough 1962 : § 39) ;
and in the Senior manuscripts amatrei (17.15) = Påli åmantehi,
pa"igae!u (12.19) = Påli pa†iggahesuµ, and priao (5.3) = BHS
pl¥hak- / Påli pihaka- “spleen”.93 This indicates that both -h- (< -ß-) and
-ß- (< -h-) in the above examples are an approximation at representing
[!], which tends to undergo further weakening (> Ø).
The palatalisation of a neighbouring vowel by ß and the weakening
of original -ß- and -h- in palatal environment as witnessed in the
examples listed here are, however, uncommon. The spellings for the
equivalent of Skt var!aßata- in Gåndhår¥ are more regularly va!aßaêa-
/ va!aßada- / var!aßada-. Similarly, original -ß- and -h- in palatal
environment normally remain, as they do in other contexts generally.94
Examples from the Senior manuscripts are kaßia-cadaâ (13.9) = Skt
kåßika-candanaµ “sandal from Kåßi” ; deßißama (13.12) = Skt
deßi!yåmi “I will teach” ; and suha-vihara (12.42) = Påli sukha-vihår¥
“living at ease”.
samiêa umic[e]a (l. 38), samiêa umijata (ll. 39–40, 42–43) : The two
Saµyutta-nikåya occurrences of the simile read sakiµ sakiµ
ummujjeyya “would emerge once each time” and sakiµ sakiµ
ummujjanto “emerging once each time”, while the European and
Burmese editions of the Majjhima-nikåya (M III 169,14) occurrence of
the simile do not repeat sakiµ.95 G samiêa must be the equivalent of
Påli/Skt samitaµ “continuously”, “over and over” (see PED s.v.
samita1). This appears in Påli texts in the expression satataµ samitaµ
“constantly and continuously”, an example being måro påpimå satataµ
92Cf. Norman 1962, p. 326 (= Collected Papers, Vol. 1, pp. 34–35) and von
Hinüber 2001, § 223.
93For priao, see Glass 2007, § 5.2.1.7.
94See Allon 2001, pp. 86–87 for references.
95Cf. tam enaµ puriso vassasatassa vassasatassa accayena kåsikena vatthena
sakiµ sakiµ parimajjeyya (S II 181,27–28).

samitaµ paccupa††hito (S IV 178,13–14) “Måra the Evil One is
constantly and continuously waiting by”, which occurs in conjunction
with a simile involving a jackal waiting for a turtle to extend its limbs.
avi â ’!e (l. 39) [§ 3a] : The Påli parallel has api nu so. The Påli
equivalent of the Gåndhår¥ would be api nu eso. Cf. ya e!e (= Påli yam
eso) in line 41, where the Påli parallel has yaµ so.
The phrase used to express the idea of the turtle inserting its neck
into the hole in the yoke is am[a]spi ekatarmao yuo grive pak!ivea in
§ 3a (l. 40) and amaspi ekatarmao yuo grive pa"i[mu](*ce)a â [va]
pa"imu[ce](*a) in § 3b, where the Påli parallel has amusmiµ eka-
cchigga¬e yuge g¥vaµ paveseyyå ti in both sections.
The verb in the Påli is paveseyya “would insert”. In contrast, the
verb in § 3a of the Gåndhår¥ is pak!ivea = Påli pakkhipeyya / Skt pra-
!k!ip, which can also mean “would insert”. This is the verb used to
express the idea of the man casting or throwing the yoke into the ocean
in § 2 : Gåndhår¥ tatra puru![e] ekatarmao yuo pak!ivea, Påli tatra
puriso ekacchigga¬aµ yugaµ pakkhipeyya. The verb in Gåndhår¥ § 3b is
pa"i[mu](*ce)a, which is repeated in the negative â [va] pa"i-
mu[ce](*a), where â va = Skt na vå “or not”. The Påli form of this
verb is pa†imuñcati (see PED s.v.), which, interestingly, occurs in the
Påli Jåtakas and commentaries in conjunction with g¥vå “neck” in the
sense of “put on”, “attach”. A particularly good example for our pur-
pose is påsaµ g¥våya pa†imuñcati “he puts the snare on his neck” (Ja IV
405,10), where the commentary (line 15) glosses pa†imuñcati with
paveseti, the verb found in the Påli sutta under discussion.96 Also of
interest is pa†imukkaµ in the Ther¥gåthå verse discussed above, a
derivative of pa†imuñcati.97 The active form of the verb (“would put
that yoke with the single hole on his neck”) does seem a little strange
96Other examples are (passive) kåkassa g¥våya pa†imucci (v.l. pa†imuñci ; As

272,32–33) and kacchaµ någånaµ bandhatha, g¥veyyaµ pa†imuñcatha (Ja IV
395,17).
97For Påli pa†imukka-, see Geiger 1994, § 197 ; Pischel 1965, § 566 ; von
Hinüber 2001, § 493 ; Norman 2002, p. 241.

254 Mark Allon
here, and it is possible that we are dealing with a passive (Påli
~muccati). If so, the translation would be “this blind turtle … would be
caught (or fastened) at the neck in that yoke”.
Finally, the Gåndhår¥ expression amaspi ekatarmao yuo grive
pa"i[mu](*ce)a â [va] pa"imu[ce](*a) “would put that yoke with the
single hole on his neck, or he may not put it on” is reminiscent of the
Buddha’s statement in § 3c of the Chinese : “Although the blind turtle
and the floating piece of wood may miss [each other], perhaps they may
also meet each other” (盲龜浮木。雖復差違。或復相得).
aêicam eêa bhayava suêalavam eva (ll. 40–41) : Påli adhiccam idaµ
bhante. For the non-aspirate consonant in aêicam = Påli adhiccam, see
the discussion of -pa"avi = Skt pÁthiv¥ above.
suêalavam = Skt sudurlabham. The Senior manuscripts contain
many examples of medial vowels not being marked, including u / o, as
here, and i / e. Further examples of u not being marked are caêamaâia
(2.11,14,16) = Påli cåtumåsin¥- “of four months” and ve"adal[e]a (13.5)
besides v[e]"udalae (13.2) = Påli ve¬udvåreyya(ka)- “belonging to
Ve¬udvåra”.
Summary
This Gåndhår¥ sËtra, for which the simile of the blind turtle and the
hole in the yoke is central, is as a whole quite close to the second of the
two Påli chigga¬a suttas preserved in the Saµyutta-nikåya (no. 56.48),
while the one similar sËtra found in the Chinese Saµyuktågama is
closer to the first of these two Påli Saµyutta-nikåya suttas (no. 56.47).
With regard to the portion of these suttas / sËtras analysed in detail
here, the prose simile, the Gåndhår¥ and Påli versions exhibit only minor
differences in terms of structure and wording. Both differ in several
important ways from the Chinese version.
The main differences between the Gåndhår¥ and Påli versions are
(i) different synonyms or near synonyms used : e.g. Gåndhår¥
-tarmao (= Skt tardman + ka), Påli -chigga¬a- “hole” ; Gåndhår¥
pak!ivea, Påli paveseyya “would insert” (§ 3a) and pa"imucea
“would put on” (§ 3b) ; cf. also Gåndhår¥ bhayava, Påli bhante
(§ 3b) ;
(ii) different pronouns or indeclinable used : Gåndhår¥ e!e “this”,
Påli so “that” (§§ 3a, 3b) ; Gåndhår¥ atra “here”, Påli tatra
“there” (§ 2) ;
(iii) differences in syntax : the accusative case of Gåndhår¥ paàimo
saharea in contrast to the instrumental of the Påli puratthimena
saµhareyya ;
(iv) a near synonym added : Gåndhår¥ aêicam … suêalavam, Påli
adhiccam (§ 3b).
Although still relatively minor, the greatest point of difference
between the Gåndhår¥ and Påli versions of the simile is in the phrase
used to express the period of time after which the turtle surfaces (§§ 2,
3a, 3b).
Greater differences are, however, evident in the sections of the
sutta / sËtra not discussed in this article (§§ 4–7). This includes the order
in which factors are listed (e.g. the three conditions that are most
conducive to attaining enlightenment) and in the wording used to
describe these, although much of the wording of the Gåndhår¥ text that
differs from the Påli parallel is in fact found elsewhere in the Påli
canon, a phenomenon already noted for this genre of text (see Allon
2001, e.g. pp. 178, 184, 256, 279).
The differences noted for this sutta / sËtra are of the same type as
those identified in a comparison of three Gåndhår¥ Ekottarikågama-type
sËtras with their Påli and Sanskrit parallels, for which the reader is
referred to Allon 2001 : 30–38.
Mark Allon
University of Sydney
256 Mark Allon
ABBREVIATIONS
The abbreviations used in this article are those of the CPD. Those not listed in
the CPD are
AVP Atharvaveda, Paippalåda recension
AVÍ Atharvaveda, Íaunaka recension
e
B Burmese (Cha††hasa#gåyana) edition(s) of Påli texts (= VRI-CD
unless otherwise stated ; page references are to the printed edition
as given by the VRI-CD)
CDIAL R.L. Turner, A Comparative Dictionary of the Indo-Aryan Lan-
guages (London, 1966)
Ce Sri Lankan edition(s) of Pali texts (= electronic version of Sri
Lanka Buddha Jåyant¥ Tripi†aka Series)
Dhp-GK Gåndhår¥ Khotan Dharmapada (= “Gåndhår¥ Dharmapada”, ed.
Brough 1962)
DOP Margaret Cone, A Dictionary of Påli, 1 vol. to date (Oxford, 2001)
EÓ-G Gåndhår¥ Ekottarikågama (ed. Allon 2001)
e
E European (Pali Text Society) edition(s) of Pali texts
frag(s). fragment(s)
l. / ll. line(s)
ms / mss manuscript(s)
RS The Robert Senior collection of Kharo!†h¥ manuscripts
SÓ Saµyuktågama
e
S Thai (King of Siam) edition(s) of Pali texts (= Mahidol Univer-
sity’s Budsir on CD-ROM : A Digital Edition of Buddhist
Scriptures [Bangkok : Mahidol University Computing Center,
1994])
SWTF Heinz Bechert, ed., Sanskrit-Wörterbuch der buddhistischen Texte
aus den Turfan-Funden, 2 vols. to date (Göttingen, 1994–)
T TaishØ ShinshË DaizØkyØ, eds. J. Takakusu and K. Watanabe. 100
vols. (Tokyo, 1924–34)
YV Yogavåsi!†ha of Vålm¥ki
BIBLIOGRAPHY
PRIMARY SOURCES
The Påli texts quoted are the editions of the PTS unless otherwise stated. Non-
Påli texts quoted in this paper are the following :
Abhidhånacintåmaînåmamålå by Hemacandråcårya ; DevasågaragaîkÁta-
vyutpattiratnåkarakalitå ; samßodhaka˙ sampådakaßca Gaˆ¥ Ír¥candra-
vijayah, Surata : Ír¥ Råndera Ro"a Jaina Sa#gha : Pråptisthåna Ír¥ Nemi-
Vijnåna-KastËrasËr¥ßvaraj¥ Jaina Jñånamandira, 2003
Amarakoßa, with the Unpublished South Indian Commentaries AmarapadavivÁti
of Li#gayasËrin, Amarapadapårijåta of Mallinåtha and the Amarapada-
vivaraâ of Appayårya, 3 vols., ed. A.A. Ramanathan, Adyar Library Series
101. [Madras] : Adyar Library and Research Centre, n.d. [c.1971–c.1983].
Ópastamba-gÁhya-sËtra, with the “Anåkula” Commentary of Ír¥ Hardatta
Mißra, The “Tåtparyadarßana” Commentary of Ír¥ Sudarßanåcårya and
Notes in Sanskrit by Mahåmahopådhyåya A. Chinnasvåm¥, ed., Umesh
Chandra Pandey, Kashi Sanskrit Series 59, Varanasi : Chowkhamba Sanskrit
Series Office, 1971
Atharvaveda (Íaunaka), with the Pada-på†ha and Såyaˆåcårya’s Commentary,
Part III (Kåˆ"as XI–XVIII), ed., Vißva Bandhu et al., Hoshairpur :
Vishveshvaranand Vedic Research Insititute, 1961
[Atharvavedasaµhitå] = Atharva Veda Sanhita. 3., unveränderte Auflage (nach
der von Max Lindenau besorgten zweiten Auflage), ed. R. Roth and W.D.
Whitney, Bonn : Ferd. Dümmler’s Verlag, 1856
Bodhasåra, with a Commentary by Divåkara, ed. Swami Dayånanda, Benares
Sanskrit Series, Benares : Chowkhamba Sanskrit Book Depot, 1905
Bodhicaryåvatåra of Íåntideva, with the Commentary Pañjikå of Prajñå-
karamati, ed. P.L. Vaidya, Buddhist Sanskrit Texts 12, Darbhanga : The
Mithila Institute, 1960
Fóguång Dàzàngj¥ng 佛光大藏經 [Fóguång Tripi†aka], Gåoxióng : Fóguång
ChËbƒnshè, 1983
Kå†hakagÁhyasËtra, with Extracts from Three Commentaries, Appendix and
Indexes, ed. Willem Caland, Lahore : D.A.V. College, 1925
The Kaußika SËtra of Atharva Veda, with Extracts from the Commentaries of
Dårila and Keßava, ed. Maurice Bloomfield, New Haven, Conn. : Journal of
the American Oriental Society 14, 1890 (reprinted Delhi : Motilal
Banarsidass, 1972)
258 Mark Allon
KaußikasËtra-Dårilabhå!ya, ed. H.R. Diwekar et al., Pune : Tilak Maharashtra
Vidyapitha, 1972
Laghuyogavåsi!†ha, ed. Våsudeva Íarman, Bombay : Nirâya Sagar Press, 1937
The Laugåkshi-gÁhya-sËtras, with Bhåshyam of Devapåla, Vol. 1, ed. M.K.
Kaul Shastri, Kashmir Series of Texts and Studies 49, Srinagar : Research
Department, Jammu and Kashmir Government, 1928
Mahåvastu avadånaµ : Le Mahåvastu, 3 vols., ed. Émile Senart, Paris : Société
Asiatique, 1882–1897
Âgveda-Saµhitå, with the Commentary of Såyaˆåchårya, ed. V.K. Rajwade et
al., Vol. 3, 6–8 Maˆ"alas, Pune : Tilak Mahåråshtra University, Vaidika
Saµshodhana Maˆ"ala, 1978
Saddharmapuˆ"ar¥ka, eds. H. Kern and Bunyiu Nanjio, Bibliotheca Buddhica
10, St. Petersburg : Académie impériale des sciences, 1908–1912
Saddharmapuˆ"ar¥kasËtra : Central Asian Manuscripts. Romanized Text, ed.
Hirofumi Toda, 2nd ed., Tokushima : Kyoiku Shuppan Center, 1983
The Íatapathabråhmaâ, According to the Mådhyandina Recension with the
Vedårthaprakåßabhå!ya of Såyaˆåcårya, supplemented by the Commentary
of Harisvåmin, 5 vols., Delhi : Gain Publishing House, 1987
[Íik!åsamuccaya] Çikshåsamuccaya : A Compendium of Buddhistic Teaching
Compiled by Çåntideva Chiefly from Earlier Mahåyåna-sËtras, 4 vols, ed.
Cecil Bendall, Bibliotheca Buddhica 1, St. Petersburg : Académie impériale
des sciences, 1897–1902
The ÍrautasËtra of Kåtyåyana, with Extracts from the Commentaries of Karka
and Yåjñikadeva, ed. Albrecht Weber, 2nd ed. Chowkhamba Sanskrit Series
104, Varanasi : Chowkhamba Sanskrit Series Office, 1972
Sukhåvat¥vyËha, ed. Atsuuji Ashikaga, Kyoto : Libraire HØzØkan, 1965
TaishØ ShinshË DaizØkyØ, 100 vols., ed. J. Takakusu and K. Watanabe, Tokyo,
1924–1934
The Yogavåsi!†ha of Vålm¥ki, with the Commentary Våsi!†hamahåråmayaâ-
tåtparyaprakåßa, ed. Wåsudeva Laxmaâ Íåstr¥ Paˆß¥kar, 3rd ed., Bombay,
1937
SECONDARY SOURCES
Allon, Mark, 2001. (With a contribution by Andrew Glass) Three Gåndhår¥
Ekottarikågama-Type SËtras : British Library Kharo!†h¥ Fragments 12 and
14. Gandhåran Buddhist Texts 2, Seattle : University of Washington Press
——— forthcoming. Ancient Buddhist Scrolls from Gandhåra II : The Senior
Kharo!†h¥ Fragments. Gandhåran Buddhist Texts, Seattle : University of
Washington Press
Allon, Mark and Richard Salomon, 2000. “Kharo!†h¥ Fragments of a Gåndhår¥
Version of the Mahåparinirvåâ-sËtra”. In Jens Braarvig, ed., Buddhist
Manuscripts in the Schøyen Collection I. Manuscripts in the Schøyen
Collection 1, pp. 243–73, Oslo : Hermes Publishing
Allon, Mark, Richard Salomon, Geraldine Jacobsen, Ugo Zoppi, 2007. “Radio-
carbon Dating of Kharo!†h¥ Fragments from the Schøyen and Senior
Manuscript Collections”. In Jens Braarvig, ed., Buddhist Manuscripts in the
Schøyen Collection III. Manuscripts in the Schøyen Collection 4, pp. 279–
91, Oslo : Hermes Publishing
Berger, Hermann, 1955. Zwei Probleme der mittelindischen Lautlehre.
Münchener Indologische Studien 1, Munich : J. Kitzinger
Bodhi, Bhikkhu, tr., 2000. The Connected Discourses of the Buddha : A New
Translation of the Saµyutta Nikåya. 2 vols. Pali Text Society Translation
Series 47, 48, Oxford : Pali Text Society
Brough, John, ed., 1962. The Gåndhår¥ Dharmapada, London Oriental Series 7.
Burrow, T., 1937. The Language of the Kharo!†hi Documents from Chinese
Turkesta,. Cambridge : Cambridge University Press
———. 1973. The Sanskrit Language, 3rd edition. London : Faber
Burrow, T., and M.B. Emeneau, 1961. A Dravidian Etymological Dictionary,
Oxford : Clarendon Press
——— 1968. A Dravidian Etymological Dictionary : Supplement, Oxford :
Clarendon Press
Caland, W., tr., 1900. Altindisches Zauberritual : Probe einer Uebersetzung der
wichtigsten Theile des Kaußika SËtra. Verhandelingen der Koninklijke
Nederlandse Akademie van Wetenschappen, Afdeeling Letterkunde,
Nieuwe reeks, deel 3, no. 2. Amsterdam : J. Müller
Crosby, Kate, and Andrew Skilton, tr., 1995. Íåntideva : The Bodhicaryåvatåra.
Oxford World’s Classics, Oxford : Oxford University Press
De, Harinåth. 1906–1907. “Notes by Harinåth De, M.A.”, JPTS V : 172–75
Fussman, Gérard, 1986. “Documents épigraphiques Kouchans IV : Ajitasena,
Père de Senavarma”, BEFEO 75, pp. 1–14
Geiger, Wilhelm, 1994. A Påli Grammar. Tr. Batakrishna Ghosh, rev. 2nd ed.
K. R. Norman, Oxford : Pali Text Society
260 Mark Allon
Glass, Andrew, 2007. Four Gåndhår¥ Saµyuktågama-Type SËtras : Senior
Kharo!†h¥ Fragment 5. Gandhåran Buddhist Texts 4, Seattle : University of
Washington Press
Gray, Louis H., 1965. Indo-Iranian Phonology, with Special Reference to the
Middle and New Indo-Iranian Languages, Columbia University Indo-
Iranian Series 2. New York : AMS Press
von Hinüber, Oskar, 1992. Sprachentwicklung und Kulturgeschichte : Ein
Beitrag zur materiellen Kultur des buddhistischen Klosterlebens. Akademie
der Wissenschaften und der Literatur ; Mainz, Abhandlungen der geistes-
und sozialwissenschaftlichen Klasse 1992, 6. Stuttgart : Franz Steiner
Verlag
——— 2001. Das ältere Mittelindisch im Überblick. 2nd ed. Österreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse,
Sitzungsberichte 467, Vienna : Verlag der Österreichischen Akademie der
Wissenschaften
Jacob, G.A., 1909. “A Sanskrit Simile”, JRAS 4, pp. 1120–21
Konow, Sten, ed., 1929. Kharosh†h¥ Inscriptions with the Exception of Those of
Aßoka. Corpus Inscriptionum Indicarum 2.1, Calcutta : Government of India
Lenz, Timothy, 2003. (With a contribution by Andrew Glass and Bhikshu
Dharmamitra.) A New Version of the Gåndhår¥ Dharmapada and a
Collection of Previous-Birth Stories : British Library Kharo!†h¥ Fragments
16 + 25. Gandhåran Buddhist Texts 3, Seattle : University of Washington
Press
Lüders, Heinrich, ed., 1926. Bruchstücke der Kalpanåmaˆ"itikå des Kumåra-
låta. Kleinere Sanskrittexte 2, Leipzig : Deutsche Morgenländische
Gesellschaft
Nandawansa, Bhikkhu Medagama, 2001. Abhidhånappad¥pikå : A Study of the
Text and Its Commentary, Pune : Bhandarkar Oriental Research Insititute
Norman, K.R., tr., 1971A. The Elders’ Verses, Vol. II, Ther¥gåthå. Pali Text
Society Translation Series 40, London : Pali Text Society
——— 1971B. “Notes on the Gåndhår¥ Dharmapada”, Indian Linguistics 32,
pp. 213–20
——— 1972. “Middle Indo-Aryan Studies IX : The Blind Turtle and the Hole in
the Yoke”, JOI(B) 21 : 331–35 (= Collected Papers, Vol. 1, 1990, pp. 156–
60)
——— 1976A. “Middle Indo-Aryan Studies XIII : The Palatalisation of Vowels
in Middle Indo-Aryan”, JOI(B) 25, pp. 328–42 (= Collected Papers, Vol. 1,
1990, pp. 220–37)
——— 1976B. “Labialisation of Vowels in Middle Indo-Aryan”, Studien zur
Indologie und Iranistik 2, pp. 41–58 (= Collected Papers, Vol. 1, 1990,
pp. 247–61)
——— 1979. “Two Pali Etymologies”, BSOAS 42, pp. 321–28 (= Collected
Papers, Vol. 2, 1991, pp. 71–83)
——— 1962. “Middle Indo-Aryan Studies III”, JOI(B) 11, pp. 322–27 (=
Collected Papers, Vol. 1, 1990, pp. 30–35)
——— 1983. “Middle Indo-Aryan Studies XVI : The Palatalisation and
Labialisation of Vowels in Middle Indo-Aryan”, JOI(B) 32, pp. 275–79 (=
Collected Papers, Vol. 3, 1992, pp. 12–18)
——— 1990–2001. Collected Papers. 7 vols., Oxford : Pali Text Society
——— 2002. “A Survey of the Grammar of Early Middle Indo-Aryan”, AO 63,
pp. 221–48
——— 2004. “A New Version of the Gåndhår¥ Dharmapada ?”, AO 65,
pp. 113–33
Oberlies, Thomas. 2001. Påli. A Grammar of the Language of the Theravåda
Tipi†aka : With a Concordance to Pischel’s Grammatik der Prakrit-
Sprachen. Indian Philology and South Asian Studies 3, Berlin : Walter de
Gruyter
Pischel, R., 1965. Comparative Grammar of the Pråkrit Languages. 2nd ed., tr.
Subhadra Jhå, Delhi : Motilal Banarsidass
Ranade, H.G., tr., [1978]. Kåtyåyana Írauta SËtra : Rules for the Vedic Sacri-
fices, Ranade Publication Series 1, Pune : R.H. Ranade
Salomon, Richard, 2001. “‘Gåndhår¥ Hybrid Sanskrit’ : New Sources for the
Study of the Sanskritization of Buddhist Literature”, IIJ 44, pp. 241–52
——— 2003. “The Senior Manuscripts : Another Collection of Gandhåran
Buddhist Scrolls”, JAOS 123.1, pp. 73–92
Turner, R.L., 1966. A Comparative Dictionary of the Indo-Aryan Languages,
Upadhye, A.N., 1972. “The Turtle and the Yoke-hole”, JOI(B) 22, pp. 323–26
Wackernagel, Jakob, 1957. Altindische Grammatik Vol. 1 : Lautlehre,
Göttingen : Vandenhoeck & Ruprecht
262 Mark Allon
Whitney, William Dwight, tr., 1905. Atharva-Veda Saµhitå, revised and
brought nearer to completion and edited by Charles Rockwell Lanman, 2
vols., Cambridge, Mass. : Harvard University
Winternitz, Moriz, 1913. “Beiträge zur buddhistischen Sanskritliteratur”, (Part
2) WZKM 27, pp. 33–47 (= Moriz Winternitz : Kleine Schriften. 2 vols.,
pp. 536–50, ed. Horst Brinkhaus, Stuttgart : Franz Steiner Verlag, 1991)
Wogiwara, Unrai (萩原雲來), 1979. Bonwa daijiten (梵和大辞典), 2 vols.
[Sanskrit–Chinese–Japanese dictionary], Tokyo : Kodansha International
Ltd
Yìn Shùn (印順) (Bhikkhu), 1983. Zá åhan j¥nglùn huìbiån 雜阿含經論會編, 3
vols., Taipei : Zhèngwén ChËbƒnshè
Remarks on the Third Precept : ´
Adultery and Prostitution in Påli Texts*
The Third Precept : Men, Women, and Wives

´
The Third Precept is to refrain from kåmesu micchåcåra,

“misbehavior in [matters of] kåma”, a rather general category. The word
kåma refers internally to the subjective emotions of desire and
experience of all sensual pleasures, and externally to the objects of those
emotions and experiences, so a broad construal of the Precept could be
broad indeed.1 However, commentaries usually elucidate it in relation to
sex : kåma is methuna-samåcåra, the act of intercourse, which is
´
twofold : contentment with one’s wife (or wives),2 or going to the wife
´
of another man, sadåra-santosa-paradåra-gamana. As one of the Six,

Eight or Ten Precepts, it requires chastity, so both kinds count as
misbehavior, but as one of the Five only the second does.
Many texts give two standard lists of ten kinds of women with
whom intercourse is forbidden, which include young women “under
protection” as well as “wives of other men”. They are agaman¥ya-
*It is an honour and a delight to contibute this small piece, intended as no more
than a preliminary and incomplete introduction to a much wider field of study,
in honor of K.R. Norman, from whose unfailing kindness I have benefited for
almost thirty years, up to and including this article.
1Thus Saddhatissa 1987, pp.88–92.
2The existence of polygyny is widely attested in Påli (see remarks on the word
dåra in text and the notes below). Many texts praise monogamy for man and
wife, in deed and thought, as a virtue; see, e.g., the Suruci Jåtaka (Ja IV
314ff.), which contains the very widespread motif that jealousy of one’s co-
wives (sapattiyo) is one of the sufferings particular to women. A man is urged
not to visit other men’s wives; women are encouraged not even to think of
other men (e.g. D III 190 with Sv 955). See also DPPN s.v. for the story of
Nakulapitå and his wife. The motif of couples being together over a series of
lifetimes is common in the Jåtakas. With the exception of the story of Kaˆhå in
the painfully misogynist Kuˆåla Jåtaka, which is modeled on Sanskrit literary
sources (see Bollée 1970, pp.132ff.), I know of no instance of polyandry.
264 Steven Collins
vatthu, literally “objects not to be gone to” : Forbidden Zones.3 The first
´
gives Ten Women (dasa itthiyo), all of whom are under some form of
protection, and the second gives Ten Kinds of Wife (bhariyå). In the
first list, of the Ten Women the first eight are protected by
1. mother (måturakkhitå)
2. father (pitu-)
3. mother and father (måtåpitu-)
4. brother (bhåtu-)
5. sister (bhagini-)
6. relatives (ñåti-)
7. clan (gotta)
8. fellow monastics (dhamma-, glossed as sahadhammika-)
The Protectors, in order to prevent their ward from having intercourse
with a man before she has come of age, do not allow her to go
anywhere, see other men, live by herself, and they tell her what to do
and what not to do. The final two are : ´
9. One who is under guard (sårakkhå), i.e. a girl who has been
promised to a man, from as early as when she was in the
womb.
10. One for whom a punishment has been set (saparidaˆ å) — i.e.
a girl, promised to someone, whose name has been put on a
public notice set up in a village, house, or street announcing a
penalty for anyone who “goes to her”.
The Ten Wives are : ´
1. “one bought for money” (dhanakk¥tå), i.e. through a bride

price or some such ; ´
2 . “one who lives [with her husband] through choice”
3See entries for itth¥ in PED, DOP, CPD. Searching the Cha††hasa¤gåyana CD
will reveal many more. There are some textual variations, of no importance
here. The term ajjhåcariya-vatthu, “object for transgressions” is also used (Pj I
31); ajjhåcariya can refer to the transgression of any Precept.
Remarks on the Third Precept 265
(chandavåsin¥) ; i.e. man and wife marry through mutual
´
affection ; ´
3. “one who lives [with her husband] because of possessions”

(bhoga-) ; a woman from the country who acquires tools such
´
as mortar and pestle, i.e. who marries for social advancement ; ´
4. “one who lives [with her husband] because of clothes”

(pa†a-) ; a poor woman who acquires even a small amount of
´
clothing, i.e., as in 3, who marries for social advancement ; ´
5. “one who lives [with her husband] because of [the ceremony

with] a bowl of water” (odapattakin¥) — oaths and vows were
often taken by two people putting their hands in a single bowl
of water ; here the officiant says “may you be joined together
´
unbreakingly as is this water” ; ´
6. “one who has taken off the head-pad” (obha†acumba†å) [a

head-pad is for carrying firewood, etc.], i.e. a former menial or
slave raised in status ; ´
7. “a wife who is also a slave” (dås¥) ; ´
8. “a servant-wife who works for wages” (kammakår¥) — a man

lives with her because his own wife is insufficient (anatthiko) ; ´
9. “one brought back under a flag” (dhajåha†å), i.e. a war-

captive ;´
10. “a temporary wife” (muhuttikå), i.e. a prostitute, used for a

shorter or longer period.4
For men all Ten Women and all Ten Wives are Forbidden Zones. In
the case of women, however, there is precise limitation : all Ten Wives
´
are guilty of wrongdoing if they have sex with a man, but only the last
two of the list of Ten Women are. This is because they have been
promised to a man, and are counted as “having a husband” (or “owner”,
4On the phenomenon of “temporary wives” in Southeast Asia see Reid 1988,
pp.154ff.´; Andaya 1998. Thanissaro 1994, p.119, interprets this term more
widely as “a date”, which is certainly possible linguistically, although it raises
intriguing historical and cultural questions.
266 Steven Collins
sassåmikå). The texts say (using Sp 555 with Sp-† Be II 329 [Sås 90–
91]5 :´
imåsu dasasu pacchimåna dvinnam eva purisantara gacchant¥na

micchåcåro hoti, na itaråsa,
Of these Ten women there is Misconduct in having sex with a man for the
last two, but not for the others.
pacchimåna dvinnan ti sårakkhasaparidaˆ åna micchåcåro hoti tåsa
sasåmikabhåvato. na itaråsan ti itaråsa måturakkhitåd¥na a††hanna
purisantaragamane natthi micchåcåro tåsa asåmikabhåvato … na ca
måtådayo tåsa phasse issarå. måtådayo hi na attanå phassånu-
bhavanattha tå rakkhanti, kevala anåcåra nisedhentå purisantara-
gamana tåsa vårenti. purisassa pana etåsu a††hasu pi hoti yeva
micchåcåro. måtåd¥hi yathå purisena saddhi savåsa na kappenti, tathå
rakkhitattå paresa rakkhitagopita phassa thenetvå phu††habhåvato.
For the last two´: for those who are under guard and for whom a punishment
has been set there is Misconduct because they have a husband/owner. But
for the others there is not´: for the other eight Women, those under the
protection of their mothers, etc., there is no Misconduct in having sex with a
man, because they do not have a husband/owner … Mothers, etc., do not
have authority over them in relation to [sexual] contact´; they do not guard
them for the sake of their own experience of [such] contact´; they [try to]
stop them from having sex merely to prevent misbehavior.6 But for a man
there is Misconduct in the case of these eight also. Because mothers, etc., do
not arrange for [their wards to have] intercourse with men, therefore [there
is Misconduct for a man] through the fact of Protection, through the state of
having stolen [sexual] contact which is protected and guarded by others.
The argument is not entirely clear to me, and more work will have
to be done, both text-critical and interpretative. The point seems to be
that mothers and other protectors of the eight kinds of women do not
have the kind of authority or ownership which a husband (actual or
promised) does. A husband owns the right to exclusive sexual access,
5 I am grateful to Thanissaro Bhikkhu for help with some of the passages

translated here.
6Presumably anåcåra here must refer to behavior contrary to social mores but
not the Third Precept.
and so adultery with any of the other twelve women constitutes a form
of theft. This seems to be the point of the sentence omitted from the
passage just cited :´
yå hi såmikassa santaka phassa thenetvå paresa abhirati uppådenti,

tåsa micchåcåro.
When women cause sexual desire to arise for [or in] other men, [thereby]
stealing the contact which belongs to their husbands, there is Misconduct on
their part.7
Perhaps light can be thrown on this by a passage from the

Upåsakajanåla¤kåra (nos. 2, 24, p.179) : ´
måturakkhitådayo … a††ha rakkhakåna anuññåya vinå v¥tikkamesu

purisassa micchåcåra bhajanti. tåsa pana natthi micchåcåro.
rakkhakåna anuññåya upagame ubhinna pi natthi micchåcåro.
The eight who are protected by mothers, etc., share in a man’s misconduct8
in cases of transgression [= sex] which occur without the permission of their
protectors. But there is no Misconduct (i.e. no breaking of the Precept) on
their part. When there is a [sexual] approach9 with the Protectors’
permission, there is no Misconduct for both [man and woman].
So a man’s breaking the Third Precept is connected to breaking the

Second, against theft. The eight women, not being the property of their
Protectors, do not steal anything by having sex with a man. But the
Protectors’ relationship to their wards is something akin to ownership,
since they can annul Misconduct for the man by giving their permission.
7It is possible to take the absolutive thenetvå here as going with paresa rather
than the subject of the sentence, in which case one would translate “when
women cause sexual desire to arise in other men, [making] them steal the
contact which belongs to their husbands …”. This would fit better with the
last sentence, which clearly uses thenetvå of men.
8 It is not clear to me what “share” means here, and I have not found other
passages which use the phrase.
9This sense of upagama is not, to my knowledge, found elsewhere in Påli. Påli
lexicographical texts (and cf. Sadd 883–84) relate it to upa-ni-¡ad, “to sit [next
to]”, and a sexual use of upa-gam is found in Sanskrit (MW s.v.). It would
seem here that it must be taken as a euphemism for sex.
268 Steven Collins
Thus, as is often the case worldwide, sexual transgressions are
committed by men not directly against a woman but against those who
either “own” her or are in some other way legally responsible for her.10
In all twenty cases wrongdoing is “adultery [which is] behavior in a
Forbidden Zone based on desire which transgresses conventional social
boundaries” (lokamariyåda atikkamitvå, Vv-a 72–73).
In a specifically Buddhist jurisprudential-ritual sense, therefore, lay
single women, of any kind (the unmarried [whether young or old],
divorcées, widows, and prostitutes, on which see more below), do not
break the Third Precept by having sex with a man. If they do, then what
is “wrong” about it is twofold, in quite different ways. First, it is
practically imprudent, given (male) marriage-expectations and social
disapproval.11 Second, from the ascetic–ultimate, karmic point of view
— augmented by misogynist attitudes toward the imagined promiscuity
of women — it is the expression of samsaric defilement.
Intercourse is defined very precisely in the Vinaya tradition (e.g. a
penis enters any of a woman’s three orifices as much as the length of a
mustard seed) ; it is not said whether this is to be taken as paradigmatic
´
for non-monastic cases. The word magga, literally “pathway”, which is

used there for “orifice” and where appropriate “sexual organ”, does
appear in a standard list of four sambhårå, “prerequisites”, or “consti-
tuents” necessary for there to be an infraction of the Third Precept (e.g.
Sv 1049, As 98) : ´
i. the existence of a Forbidden Zone (agaman¥ya- or ajjhåcariya-

vatthu)
ii. the intention to perform the act (sevanå-citta)
iii. an [appropriate] means (payoga) (transgressing the Precepts
10For this point in later Thai legal codes on marriage, adultery, rape, etc., see
Loos 2006.
11e.g. S I 6 komår¥ se††hå bhariyåna, “a virgin is the best of wives”, to which
Spk I 33 comments kumår¥-kåle gahitå, “taken [in marriage] at the time of
their youth”. A number of compounds with the word kumåra- refer to women
who marry as virgins or men who marry virgins´: cf. DOP s.v. kumår-/komår-.
involves one or more of a list of six such means, which
include, for example, getting someone else to break a Precept ; ´
in relation to sex there is but one means : såhatthika, literally

´
“with one’s own hand”, but better “personally”)

iv. consent to the physical interaction between the sexual organ
and an orifice (maggena magga-pa†ipatti-adhivåsana ; Pj I 31 ´
has simply sådiyati, “s/he agrees”).

The question of consent, and at what moments during an act of
intercourse consent can be given or refused, receives a fair amount of
discussion which I cannot go into here.12 Whether or not the woman is
willing, and in whatever senses that is understood, the man’s Mis-
conduct depends on the status of the woman. If she has not taken the
Precepts it is “not very blameworthy”, but great if she has,13 and the
offence gets worse as the status of the woman increases (e.g. Vibh-a
383) :
´
[T]he wrongdoing is not very blameworthy [when it involves] transgression

with a woman of bad Virtue, greater when the woman’s Virtue is like a
cow’s, greater [still, and incrementally] when she has gone for refuge, has
[also] taken the Five Precepts, is a novice nun, an ordinary nun [i.e. one
who has not attained any level of the Path], a Stream-Enterer, Once-
Returner, Never-Returner´; with an Enlightened Woman it is wholly and
completely blameworthy.
“Like a cow’s” is gorËpa-s¥laka. This is equivalent to what is called
12Many texts discuss volition, on both the man’s and the woman’s part´; this
needs much more research, but it does seem that women’s volition is
sometimes treated in misogynist ways´: e.g., as Thanissaro Bhikkhu puts it
(personal communication), in a discussion of rape at Sv-p† III 346 there seems
to be “the old excuse ‘The fact that she didn’t show any desire doesn’t mean
that she didn’t want it, for that’s the way women are’”. For a man, one precise
example is the case of ejaculation in dreams´: being unintentional it does not
break any Monastic Rule, but as a manifestation of desire it does have a
karmic result. See Collins 1997, p.190.
13 e.g. Ps I 199 ´: so pan’ esa micchåcåro s¥lådiguârahite agaman¥ya††håne
appasåvajjo, s¥lådiguâsampanne mahåsåvajjo
270 Steven Collins
elsewhere “an ordinary person’s Virtue” (puthujjana-s¥la)14 ; the image
´
may be that a cow is innocent, intending no harm : one sub-commentary

´
says “naturally good” (pakati-bhadda, Dhs-anu† Be 189). But another

says that such a person is “blind, of blundering intellect“ (mË¬ho
khalitapañño, Spk-p† (Be) I 160).
Adultery
Adultery in a general sense, when not tied to discussions of the Ten
Women and Ten Wives, is expressed by verbs such as aticarati or
atikkamati, “going too far, transgressing”, and also by nominal and
verbal forms derived from gam, to go, most commonly with the
compound paradåra-, “another man’s wife” (or wives, on which see
below) as their object ; the word paradårika is used for an adulterer. The
´
words jåra (masculine) and jårå/jår¥ (feminine), “lover”, are used for
partners in a sexual relationship outside normal marriage : the ´
relationship, real or alleged, can be between monks and nuns, or other

ascetics, monks and laywomen, and married men and women and their
lovers.15 It is hardly likely that any extensive story-literature, in any
culture, would not know of adultery ; in Påli, especially given the many
´
misogynistic Birth Stories which aim to demonstrate the untrust-

worthiness and moral depravity of women, naturally many such stories
are found.16 Admonitions against adultery in sermons by the Buddha
and others scarcely need documenting. But what is wrong with it, why
should one avoid it ?
´
If one or both partners are in one or more of the categories which

mean that their adultery breaks the Third Precept, the reasons are
obvious. But adultery is spoken of usually without reference to that
14 M III 255, glossed as go-s¥la-dhåtuko at Ps V 71, which adds that such a

person is honest, not deceitful, does not oppress others, and makes a living
rightly and properly through farming or trade.
15 Monks, nuns, e.g. Vin II 259, IV 91, monks and laywomen Vin IV 20,
married men, women and lovers Vin II 268, III 83, 138, 139, Ja II 292, III 92,
223.
16See Jones 1979´; Bollée 1970´; Amore and Shinn 1981.
particular jurisprudential manner of evaluation, and the arguments
offered against it are various. One story, the “Foot of a Mountain” Birth
Story (Ja II 125–27), has a remarkable mixture of disapproval and
acceptance. In both the Story of the Present, with the king of Kosala and
the Buddha, and that of the Past, with King Brahmadatta in Benares and
a wise councillor, one of the king’s ministers “does something wrong”
in the harem.17 In the Present, the king reflects that the minister is
useful, and the woman is dear (piyå) to him, so he cannot destroy them.
The Buddha agrees, saying that when servants are useful and women
dear one cannot do them harm, and he then tells the Birth Story, where
King Brahmadatta tells the minister what has happened in a riddling
verse, “At the pleasant foot of a mountain was an auspicious lake ; a ´
jackal drank from it, though he knew it was protected by a lion.” The
minister understands, and replies, “Great king, [whatever] animals drink
from a great lake, it is none the less a lake ; if she is dear to you, forgive
´
(khamassu) [them].” “None the less a lake” renders na tena anad¥ ; ´
a-nad¥ is literally “a non-lake”, and as usual the negative prefix can be
taken in the sense of a logical negation and/or in the sense of a negative
evaluation : “a not-X” and/or “a bad X”. The commentary says,
´
[A]ll creatures, two-footed, four-footed, snakes and fish, drink water from a
lake when they are thirsty, but it is not for that reason any less a lake´: it is
not a polluted lake. Why´? Because of its being common to everyone. Just as
a lake drunk by anyone and everyone is not corrupted, so a woman who
through defilement transgresses against her husband by having sex with
another man is none the less a woman. Why´? Because [of her, or all
women’s´?] being common (sådhåraâ) to everyone. She is not a polluted
woman. Why´? Because of becoming pure [again] through washing at the
end …. [So, the advice is] forgive both of them and preserve [your]
equanimity.18
17The verb is padussi, from pra-du¡, whence the word do¡a (Påli dosa), one of
the commonest and least specific words for a wrong, in some sense of that
word.
18“Polluted” is ucchi††ha, for which DOP has “left-over [of food], touched, spat
out, used, cast-off, polluted”, citing the compounds -odaka, “water that has
272 Steven Collins
The king does so, telling them not to commit such evil (påpakamma)
again, and they stop.
The most general argument against adultery is a version of the
golden rule given by the Buddha to dissuade male householders : ´
Again, householders, a noble disciple should reflect thus´: “If someone were
to commit adultery with my wives, that would not be pleasing and agreeable
to me. Now if I were to commit adultery with the wives of another, that
would not be pleasing and agreeable to the other either. What is displeasing
and disagreeable to me is displeasing and disagreeable to the other too.”
Having reflected thus, he abstains from sexual misconduct, exhorts others to
abstain from sexual misconduct, and speaks in praise of abstinence from
sexual misconduct. Thus his bodily conduct is purified in three respects.19
There are also more immediate disadvantages. A poem of the Sutta-

nipåta, “The Discourse on How Not to Thrive”,20 lists twelve kinds of
misfortune resulting from such things as being too fond of sleep or lazy,
not supporting one’s parents, and being rich but enjoying oneself alone.
The eighth (106) is “being a womanizer [which leads to] squandering
what one has acquired”. The tenth (110) is “being an old man who
brings home a [young] woman with breasts like timbaru fruit, [which
leads to his] not sleeping because of jealousy over her” ; the ´
commentary explains this as an eighty- or ninety-year-old man thinking

that his young wife will have no pleasure living with an old man and so
seek a younger one : burning with lust and jealousy he forgets to look
´
after his affairs and so comes to ruin. The ninth (108) is “being
unsatisfied with one’s wife and being seen with prostitutes and other
men’s wives” ; the commentary explains that visiting prostitutes means
´
giving away money, while adultery involves being punished by the

king.
been spat out”, and -geha, “a house which is not new´; a house already lived
in”. Odakantikatå, “ending with a [ritual] wash”, is a defining characteristic
of sex in the Monastic Code (Vin III 28).
19S V 354, tr. Bodhi 2000, p.1798.
20 Paråbhava-sutta, Sn 91–115. For the meaning of paråbhava see Norman
2001, p.186, ad Sn 92.
Other texts also suggest that adultery is a public crime, but not all.
The issue requires further research, and as with many other issues
discussed here, such research may reveal differences between different
Påli texts, which may reflect local variations in practice.21 Two verses
in the Dhammapada (309–10), in a chapter entitled “Hell”, state : ´
A careless man who courts another’s wife gains four things´: acquisition
of demerit, an uncomfortable bed, third blame and fourth hell.
Acquisition of demerit and an evil state of rebirth, and a brief delight for
the frightened man with the frightened woman, and the king imposes a
heavy punishment. Therefore a man should not court another’s wife.22
The word-commentary says that the king’s punishment involves

such things as cutting off hands and feet, but the story attached to the
verses in the same text — which might well be of different provenance
— is quite different (Dhp-a III 479–81). It concerns a handsome young
man called Khema, a nephew of the rich banker Anåthapiˆ ika and an
habitual adulterer. Women have merely to look at him to lose control of
themselves. (He has this ability thanks to the fact that in a previous life,
at the time of the Buddha Kassapa, he had offered two colored flags at a
Buddha-shrine with the wish “May all women apart from my family and
relatives fall in love with me on sight.”) He is arrested three times, but
the king releases him each time, feeling ashamed for the banker.
21Sp 561, commenting on alavacan¥yå, a woman who is or can be divorced,

says at Vin III 144 ´: yå hi yathå yathå yesu yesu janapadesu pariccattå va
hoti, … aya alavacan¥yå ti vuccati, “a woman is called ‘One about whom
Enough´! is to be said’ when she has been abandoned by whatever means [is
current] in whatever region”´; the sub-commentary (Sp-† Be II 329) adds´:
alavacan¥yå hont¥ ti desacårittavasena paˆâdånådinå pariccattå honti,
such women “are abandoned by such means as giving a letter, according to
the custom of the region”.
22cattåri †hånåni naro pamatto, åpajjati paradårËpasev¥ | apuññalåbha na
nikåmaseyya, ninda tat¥ya niraya catuttha || apuññalåbho ca gat¥ ca
påpikå, bh¥tassa bh¥tåya rat¥ ca thokikå | råjå ca daˆ a garuka paêti,
tasmå naro paradåra na seve. The translation is from Norman 1997, p. 45.
The commentary explains “an uncomfortable bed” as meaning that he cannot
sleep when he wants to, and sleeps little.
274 Steven Collins
Anåthapiˆ ika tells all this to the Buddha, who speaks the verses to
Khema “to show the fault (dosa) in going after other men’s wives”.
Leniency for adulterers is also suggested by another remark of the
Buddha in the same text, in which he “does not make any one Precept
lesser [than another]” because “they are all difficult to keep”.23 This
precedes a verse which says that “whoever … goes to another’s wife
digs up his own root here in this very world”. The word-commentary
specifies this as not paying attention to his business affairs and wasting
money, without mentioning public punishment.
In the long term, the karmic punishment for adultery is bad rebirth.
Male and female adulterers go to hell : an example very common in
´
modern Thai temple wall-paintings is one where they are forced by

armed guards repeatedly to climb a tree with sixteen-finger-long iron
thorns (see, e.g., Ja V 269, explained at 275). Various other karmic
effects are described : male adulterers are reborn as human women ;
´ ´
women who avoid adultery are reborn as men (e.g. Dhp-a I 327, where
the text remarks that “there are no men who have not previously been
women, nor women who have not previously been men”) ; one male ´
adulterer is reborn submerged head-deep in a pit of excrement, and an

adulteress flies through the air with flayed skin, attacked by vultures (S
II 259). In both cases, he/she “as a result of that deed cooked in hell for
many hundreds of years, many thousands of years, many hundred
thousands of years … through the power of the ripening of that same
deed”.
Single Women (Young and Old),

Divorcées, Widows, and Prostitution
It would seem to follow from the logic of the remarks about the last
two of the Ten Women and the Ten Wives that no other woman breaks
the Third Precept in having sex. This is said in some texts about
23Dhp-a III 355, ekas¥lam pi kaîtthaka akatvå sabbån’ eva durakkhån¥ ti,
preceding Dhp 246. The particle eva here could be read as meaning “equally”,
though that might be an exaggeration.
prostitutes (see below), but I do not know of it being said specifically of
other single women, who are usually referred to disparagingly. Young
girls, starting at around sixteen (the usual age for marriage) “wish for
men, lust for men” ; “the madness of youth” can make them enter into
´
inappropriate sexual liaisons.24 Older women who do not get married or

enter the Monastic Order are called thullakumårikå, which is best
rendered simply as “spinster”. The word thËla / thulla can mean
physically big : Horner has “grown girl”, referring to the commentarial
´
gloss mahallikå, “old”. It can mean “gross” in an evaluative sense : ´
Rouse has “coarse”, translating an explanation in the Cullanårada

Jåtaka : “You must understand that a ‘coarse’ girl does not mean one
´
whose body is fat, but be she fat or thin, by the power of the five sensual
passions she is called ‘coarse’.”25 Spinsters are one place to which a
monk should not go for alms — they are agocara : “such girls have
´
grown up, and are past their prime — they go about desiring men,
looking for intimacy with anyone”.26 The usual word for widow is
vidhavå (possibly simply vi-2dhava, “without a man”) ; widows, like
´
spinsters, are also said to be a place monks should not go for alms, since
they also “are on the lookout for intimacy with anyone”.27 On the other
hand, widows, like young girls, could be victims, as the terms kaññero
and vidhavero, “preying on virgins [and] widows” suggest.28 The
24 e.g. Dhp-a II 217 ´: tasmiñ ca vaye †hitå nåriyo purisajjhåsayå honti

purisalolå´; Dhp-a I 239–40´: yobbanamadamattatåya purisalolå (purisa-lola
is said to be one of five kinds of lust or greed afflicting women, Pj II 35-6,
Sås 220)
25Ja IV 219–20 (cf. Ja III 147) thullakumårikå ti na ca thËlasar¥rå da††habbå,
thËlå vå hotu kiså vå, pañcakåmaguîkarågena pana thËlatåya thulla-
kumårikå ti vuccati, tr. Rouse 1895, p.137. Horner’s version is at Horner 1982
[1951], p.87.
26 Sp 991, yobbanappattå yobbanåt¥tå vå kumåriyo´; tå purisådhippåyå va
vicaranti, yena kenaci saddhi mittabhåva patthenti´; Nidd-a 451, Vibh-a
339–40 have mahallikå anivi††hakumåriyo.
27e.g. Sp 991–92 tå yena kenaci saddhi mittabhåva pa††henti.
28Norman 1992, pp.88–90, referring to Ja IV 184, VI 508, and discussions in
grammatical texts.
276 Steven Collins
difficulty of life as a widow is a familiar topos of South Asian
literature ; widowhood is one of ten things “looked down on by people”
´
(Mil 288). I have not yet examined texts referring to divorcées, but the
fact that commentaries define a widow as a woman whose husband is
dead or living somewhere else would suggest that their moral–legal
status might be comparable.29
Limitations of space prevent further exploration of these issues. But
both for its own sake, and because of the contemporary significance of
prostitution in countries where Påli texts are seen as “the Buddhist
tradition”, where some people connect its growing prevalence with the
misogyny which is certainly found in some Påli texts, it is perhaps
worth while looking more closely at the issue. The Påli imaginaire as a
whole is uneven : on the one hand, prostitution is called a “defiled form
´
of action” which results in blame in this life and bad destinies in the
future ; words for “prostitute” are used as insults ; prostitutes are an
´ ´
unsuitable source of alms for monks ; and their alleged obsession with
´
sex and availability to all is used as a denigratory figure in misogynist

characterizations of women in general. On the other hand, prostitutes
are capable of virtue (s¥la) ; to be a wealthy and cultured prostitute can
´
be a reward for good karma ; they can give alms to monks ; and they
´ ´
may be ordained as nuns, and go on to attain enlightenment.

The most common word for prostitute is gaîkå. The word is from
gaâ, an amount, a number, or a crowd, but the exact etymology of the
term is uncertain : it may mean “one who belongs to a crowd”, or “one
´
who [is had ] for a [specific] amount”. It is sometimes said in secondary

sources that gaîkå denotes a high-class “courtesan”30 — that is, a
29e.g. Vibh-a 339 vidhavå vuccanti matapatikå vå pavutthapatikå vå. Although

they may not technically speaking break the Third Precept, one should note
such texts as Mil 205ff., which describes a woman whose husband was living
away but who nonetheless did not do wrong (påpa nåkåsi) with any man
even though she was offered large sums of money to do so.
30This word is often used simply as a euphemism. I use “prostitute” with no
pejorative sense intended.
woman who, like an ancient Greek hetaira or Japanese geisha, is
cultured and accomplished in the arts (especially dancing) as well as a
sexual partner — in contradistinction to ves¥, which denotes a lower-
class harlot.31 This is not consistently borne out by the use of the words,
however. They are given as synonyms32 ; in one story (Vin III 138–39),
´
a group of womanizers33 send a messenger to summon a ves¥ to a park

where they are enjoying themselves. She refuses, saying that she is rich
and prosperous and will not leave the city. They engage the services of
the monk Udåyi, who acts as a go-between with her, thus causing the
Buddha to promulgate Sa¤ghådisesa Rule no. 5, prohibiting monks from
acting as go-betweens to arrange for a marriage, a lover or a “temporary
woman”. The word ves¥, or vesiyå/vesikå, has been connected with
vessa, Sanskrit vaißya, the third of the four Brahmanical social
groupings (thus PED s.v. “a woman of low caste, a harlot”), but it is
probably from Sanskrit veßa, “a house (sc. of ill-repute)”, from viß, “to
enter or settle down”. All prostitutes are rËpËpaj¥vin¥, “women who live
off their rËpa”, which here may mean “[good] looks” or simply “body”.
Some higher-class ones, especially those who seem to have been
established by a city or township, are called nagarasobhin¥, “women
who beautify the city”. Another word is vaˆâdås¥, “slave of beauty”.34
31Thus Perera 1993, p.215, no. 341´: “The gaîkå, though serving the needs of
sex, is not the despicable creature that the prostitute is.” Cf. Murcott 1991,
pp.119–20.
32e.g. Sp 1293 on Vin II 267, Abh 233.
33They are called simply dhuttå, “rogues”, “abandoned” to one or more of three
things ´: women, alcohol and gambling. The commentary here (Sp 553)
naturally specifies them as womanizers, itthi-dhuttå.
34 Abh-† B e 169 explains the term at Abh 233 as vaˆâsampannå dås¥
vaˆâdås¥. dåsim pi hi vaˆâsampanna keci såmikå dhanalobhena gaîka
karonti, “a slave endowed with beauty is called a slave of beauty. Some
owners make a slave-woman a prostitute because of their greed for money.”
Some mss of Th¥ 442 + Th¥-a 248 use the word of someone said to be “neither
man nor woman” and “neuter” (napusaka) — presumably intending to refer
to a male prostitute.
278 Steven Collins
Prostitution is described, sometimes by prostitutes themselves, as a
“defiled form of action” (kili††ha-kamma) ;35 kili††ha is from kliß, to be
´
troubled, stained, defiled, whence the term kilesa, an ubiquitous

Buddhist term : kilesa-nibbåna, “the nirvana of the Defilements”, is a
´
defining characteristic of final nirvana (Collins 1998 : 148, 151). Texts

´
list various kinds and numbers of kilesa ; 36 those most relevant to

´
prostitution would appear to be no. 1 lobha, “greed” ; no. 9 ahirika,

´
shamelessness ; and no. 10 anottappa, not fearing blame. Vimalå, a

´
prostitute who became a nun and Arhat, falls in love with the monk
Moggallåna ; she goes to him and does palobhana-kamma in his
´
direction, which Pruitt (1998 : 101) translates “make seductive

´
action”.37 He repels her with verses on the foulness of the body and so
causes her to establish Shame and Fear of Blame (hiri-ottappa). Later
she recalls how, intoxicated with her youthful beauty, she used to stand
at the brothel door like a hunter, “revealing many secret places”
(specified as thighs, hips, and breasts, Th¥ 72ff. and Th¥-a 76–77). Yet
worse, some prostitutes abandon baby sons, preferring daughters they
can train in their own métier.38 A hakås¥ (e.g. Th¥-a 29–31) and
Ambapål¥ (e.g. Th¥-a 198–204), both of them wealthy, and who both
became nuns and Arhats, are said to have used the word gaîkå as an
insult to Buddhist nuns in previous lives, and as a result to have been
35 e.g. the term is used of Sirimå, who abandons it and attains the Fruit of
Stream-Entry (Vv-a 74–75)´; it is said by a nagarasobhin¥ of herself at Ja III
435ff., and of a gaîkå at Pv-a 195). It is used for other misdeeds, e.g. a
proposed act of incest (Ja IV 190), pork butchery (Dhp-a I 125–28), and
refuse-sweeping (Vbh-a 440-1). At Ja III 60 a gaîkå calls her trade n¥ca-
kamma, “inferior work”.
36e.g. ten (Vbh 341, Vism 683 = XXII 49), five hundred (Spk I 187), fifteen
hundred (Ud-a 138–39, 335f.)
37Palobhana, I think, has both simple and causative senses´: action based on
and intended to incite greed.
38 e.g. Sålavat¥ (Vin I 269), whose son survived and went on to become the
prosperous physician and Buddhist lay-supporter J¥vaka (cf. also Pv-a 195)´;
cf. Dhp-a I 174, and see Horner 1930, pp.87ff.. Their métier is called a
“tradition” (paveî).
reborn in hell and as prostitutes. Prostitutes are the first of the list of five
places to which a monk should not normally or regularly go for alms
(agocara), the others being widows, spinsters, nuns and bars. The
reasons given for this are that monks are likely to develop a fondness
for going there often, and that in any case their going there would be a
cause for reproach from others. (But if prostitutes wish to make merit to
transfer to dead relatives or to give monks “ticket-food”, monks may go
there as long as they establish mindfulness.39) Just one example of the
image of prostitutes in characterizations of women as a whole will
suffice, from the Kuˆåla Jåtaka.40 A verse and its commentary have : ´
“Like a lion eating blood and meat, a beast of prey, grabbing with its
paws and jaws, greedy, obtaining his food by force, ready to hurt others,
so are women : a man should not confide in them … Not only … are
´
[women] whores, harlots and prostitutes, not only strumpets : ´
murderesses are they !” “Murderesses” (vadhikåyo) is explained as

´
“husband killers” (431), where reference is made to another Birth Story

(Ja V 367), where “many women” are said to be common property like
a bar to drunks, and (a common trope) “the snare of Death”.
Nonetheless, other texts describe prostitutes as capable of s¥la. The
Kurudhamma Jåtaka (no. 276, Ja II 365–81) tells a utopian story of the
Kuru kingdom, where everyone, including prostitutes, keeps the Five
Precepts so assiduously that they worry that they may have broken them
because of “a trifle”. They are all “sages of old”, even though “they
were living the defiled life in a household” (agåramajjhe sakili††ha-
bhåva). Eleven examples are given, to messengers who come from
another kingdom where no rain falls in order to learn what it is about the
Kurus’ Virtue which causes rain to fall there. Each person doubts that
they have kept one or other Precept : two concern the Third. The queen
´
saw her husband’s brother, the viceroy, riding on an elephant one day,
39Vin I 70 + Sp 991–92, A III 128 + Mp III 278 + Mp-n† Be III 39, Nidd I 473
+ Nidd-a 451, Vbh 247 + Vbh-a 339–40, Vism 17 = I 18 + Vism-mh† Be I 42.
40No. 536, Ja V 412–56´; tr. Francis 1895, pp.219–45´; ed. and tr. Bollée 1970.
Text cited Ja V 425 = Bollée, p. 23, translation from Bollée, pp. 132–33.
280 Steven Collins
felt greed for him and fantasized that her husband would die, the
viceroy would become king and marry her ; she then doubted her virtue
´
because she had looked at another man “in a defiled manner” ; the ´
messengers assure her that “there is no adultery in the mere occurrence

of a thought” and pass on. The last is the prostitute. She doubts her s¥la
because in the past Sakka, in the form of a young man, gave her money
in advance of an assignation, but then returned to heaven for three years.
The prostitute, “fearing to break her Virtue”, refuses to accept anything
from any other man ; she thus falls on hard times and goes to the Chief
´
Justices to ask permission to start earning her wages as before. They

give it, but as she is about to take money from another man, Sakka
reappears and she refuses the money. He reveals his true identity,
admonishes the crowd to preserve Virtue as she has done, and leaves.
She nonetheless thinks that her virtue is faulty because she stretched out
her hand to take money from another man. The messengers insist that
her s¥la is in a state of “perfect purity” (paramå pårisuddhi).
In another story a young woman and a prostitute are among
fourteen cases of people who have fallen on hard times. A wise king
explains how they must mend their ways : the woman has a lover living
´
between her husband’s and her parents’ villages ; she pretends to visit
´
her parents but stays with her lover. The king says she should stay with
her husband, otherwise he (the king) might seize her and put her to
death. The prostitute used not to take money from another man until she
had fulfilled her contract with whoever had given her money, and so she
earned a lot ; but now, giving up that practice (or : form of propriety,
´ ´
dhammatå), she takes money from one man, but gives an opportunity to
another man instead of him, and so no one comes to her. She should
keep to her old dhamma (Ja II 308–309).
One text argues explicitly that prostitutes do not break the Third
Precept. It was edited by Jaini — who says that “this passage is
probably the only place in Buddhist literature where the problem of the
application of the lay discipline to a courtesan has been raised” — under
the title Lokaneyyapakaraâ, on the basis of one nineteenth-century
Thai manuscript in Khmer script, and dated by him tentatively “not later
than the fourteenth century A .D .” (1986 : xliii, xlvii). The relevant
´
section is found in the Kurudhamma Chapter (based on the Kuru-

dhamma Jåtaka), and it is not easy to interpret.41 A series of arguments
and analogies begins with the statement that a gaîkå has a fourfold
duty (kicca) : she is to (i) preserve s¥la by taking money from anyone,
´
whatever their social level ; (ii) remain calm (niccalå) throughout her
´
sexual encounters ; (iii) after taking money for a later assignation, not go
´
with anyone else even if they offer more money ; and (iv) remain
´
equanimous during encounters and not afterwards show personal

preferences for any customers, whatever their social level. It then — in
a style typical of Southern Asian philosophical texts — refutes an
imagined objector who claims that a prostitute breaks the Third Precept
because she goes with other women’s husbands. First, it argues that just
as a person whose retinue or slaves or relatives42 go on board a boat in
order to trade is competent (or : has the right, samattha) to rebuke or
´
strike them, but cannot impute blame to the ferryman, so the wives of
the men who have sex with a prostitute cannot impute blame to her.
Second (Lkn 194) : ´
yasmå poråˆå råjåno tam ånetvå tasså yattaka kåla bhati denti tesa
tåya saddhi methunasavåso tattaka kåla hoti, te pi sakasaka-
bhariyåyo må tasså dosam åropetha, ida råjadhana va hanatthåya
savattat¥ ti saññåpesu, tesa pi bhariyåyo ayañ ca ayañ ca me me
såmiko ti paggahesu, tasmå tasså majjhattacittena kåmesu micchåcårå
veråmaˆ¥ hoti n' eva nindå hoti.
Just as when kings in the past, bringing a prostitute [to their realm] had sex
with her for however much time they had paid her for, and conciliated their
respective wives, [saying,] “Do not impute blame to her, this is conducive
to increasing the royal wealth”, [while] the wives on the other hand
accepted it [each saying] “This is my såmika”´; therefore because of her
41 Jaini’s summary (1986, p.xlii) appears to be studiedly vague, and may be

mistaken in some details. I thank K.R. Norman (personal communication) for
help with the sentences I have been prepared to translate here.
42Reading (as suggested by K.R. Norman) parijanå dåså vå ñåtisålohitå vå.
282 Steven Collins
psychological equanimity there is abstinence from sexual misconduct, [and
so] there is no blame.
(Presumably the increase to royal wealth came from taxation of the

prostitute’s earnings from other men.)
A number of texts describe prostitutes who were expensive : they ´
become rich and those who tax them profit also. There are many
examples of ex-prostitutes who became nuns and even Arhats : see ´
DPPN, for example, for the stories of A hakås¥, Sirimå, Abhayamåtå,

Vimalå, and others, especially the doyenne of prostitutes in Påli texts,
Ambapål¥, who receives special attention and privilege from the Buddha
while still a prostitute.
In general, therefore, the attitude to prostitution in Påli texts seems
to be this : from the ascetic–ultimate perspective, prostitutes’ behavior is
´
a prime example of the greed, attachment, and defilement which tie all
those who live the household, married life to rebirth. Some can,
however, reform and attain enlightenment in the same life. From within
a karmic perspective prostitutes do not, or at least do not necessarily do,
wrong, and do not break the Third Precept. Men who visit them likewise
do not break the Precept (they are not a Forbidden Zone, as are the Ten
Women and Ten Wives), although the psychological and interpersonal
ideal of monogamous fidelity would seem to tell against the habit.
Steven Collins
University of Chicago
ABBREVIATIONS AND TEXTS CITED

All abbreviations for Påli texts follow CPD
DPPN = G.P. Malalasekera, Dictionary of Påli Proper Names
CPD = Critical Påli Dictionary
DOP = M. Cone, A Dictionary of Påli, Part 1
MW = Monier-Williams’ Sanskrit–English Dictionary
PED = Pali Text Society’s Pali–English Dictionary
The Cha††hasangåyana CD is available from www.vri.dhamma.org
Amore, R.C. and L.D. Shinn, 1981. Lustful Maidens and Ascetic Kings´:
Buddhist and Hindu Stories of Life, Waterloo, Ont.´: Wilfrid Laurier
University Press
Andaya, B., 1998. “From Temporary Wife to Prostitute´: Sexuality and
Economic Change in Early Modern Southeast Asia”, Journal of Women’s
History, 9 (4), pp. 11–34
Bodhi, Bhikkhu, 2000. The Connected Discourses of the Buddha´: A New
Translation of the Sayutta Nikåya, Vol. II, Boston, Mass.´: Wisdom
Publications
Bollée, W.B., 1970. Kuˆålajåtaka, London´: Luzac
Collins, S., 1997. “The Body in Theravåda Buddhist Monasticism”, in The Body
and Religion, ed. Sarah Coakley, Cambridge´: Cambridge University Press
——— 1998. Nirvana and Other Buddhist Felicities´: Utopias of the Pali
Imaginaire, Cambridge´: Cambridge University Press
Francis, H.T., 1981 [1895]. The Jåtaka or Stories of the Buddha’s Former
Births, Vol. V, London´: Pali Text Society
Horner, I.B., 1930. Women Under Primitive Buddhism ´: Laywomen and
Almswomen, London´: Routledge
——— 1982 [1951]. The Book of the Discipline (Vinaya-Pi†aka), Vol. IV,
London´: Pali Text Society
Jaini, P.S., 1986. Lokaneyyapakaraâ, London´: Pali Text Society
Jones, J.G., 1979. Tales and Teachings of the Buddha´: The Jåtaka Stories in
Relation to the Påli Canon, London´: Allen & Unwin
Loos, T.L., 2005. Subject Siam´: Family, Law, and Colonial Modernity in
284 Steven Collins
Thailand, Ithaca, N.Y.´: Cornell University Press
Murcott, S., 1991. The First Buddhist Women´: Translations and Commentaries
on the Therigatha, Berkeley, Calif.´: Parallax Press
Norman, K.R., 1992 [1985]. Collected Papers, Vol. III, Oxford´: Pali Text
Society
——— 1997. The Word of the Doctrine (Dhammapada), Oxford´: Pali Text
Society
——— 2001. The Group of Discourses (Suttanipåta), 2nd ed., Oxford´: Pali
Text Society
Perera, L.P.N., 1993. Sexuality in Ancient India´: A Study Based on the Pali
Vinayapi†aka, Kelaniya, Sri Lanka´: Postgraduate Institute of Pali and
Buddhist Studies, University of Kelaniya
Pruitt, W., 1998. The Commentary on the Verses of the Ther¥s, Oxford´: Pali
Text Society
Reid, A., 1988. Southeast Asia in the Age of Commerce 1450–1680, Vol. I´: The
Lands below the Winds, New Haven, Conn.´: Yale University Press
Rouse, W.H.D., 1895. The Jåtaka or Stories of the Buddha’s Former Births,
Vol. IV
Saddhatissa, H., 1987. Buddhist Ethics, London´: Wisdom Publications
Thanissaro Bhikkhu (Geoffrey deGraff), 1994. The Buddhist Monastic Code,
Valley Center, Calif.´: Metta Forest Monastery
A Note on vinaya*
The word vinaya is well known in Buddhist circle as the “norm of

conduct” in general and as a name of the Tripi†akas in particular. The
word is originally a verbal action noun (nomina actionis) formulated out
of the verbal root vi-n¥- which means “drive out”, or “remove”. Thus,
the original meaning of the word vinaya is the act of driving out or
removal. This meaning is preserved in Pali, as is seen in such
compounds as pipåsa-vinaya (removal, or quenching of thirst),1 kodha-
vinaya (removal, or repression of anger), etc. Apparently, its application
to the disciplinary meaning in the Vinaya texts is a later development
through the “removal” of immoral thought and evil actions. Yet, on the
other hand, in Classical Sanskrit literature the meaning of “removal” is
also developed in another direction and appears in an erotic context,
which is hardly compatible with the moral and disciplinary one. Under
such circumstances, it might be interesting to investigate how far its
semantic field is extended in Pali as well as in Sanskrit, and to locate the
Buddhist meaning within it.
However, according to the nature of the material which the present
writer has been able to collect so far, it is convenient to deal with the
problem in two sections. In the first section, we shall examine its wide
range of usage, extending from erotics to ethics in Sanskrit as well as in
Pali, and in the second, we shall discuss the meaning peculiar to
Sanskrit.
It is out of great respect for the scholarship of Mr K.R. Norman that
the present writer takes up the word vinaya and dedicates it to his
Festschrift.2
*This is an abridged, revised, English version of my paper in Japanese entitled

“Vinaya Study”, published in The Journal of the International College for
Advanced Buddhist Studies 7 (2004) pp. 217–70.
1A II 34,25, Ja III 290,26. Cf. also pipåså-vinayana in Mil 318,25–27.
2In fact, the present writer was inspired by an article by Mr Norman (1993).
286 Minoru Hara
1. vinaya in Påli and Sanskrit
For clarity’s sake, let us start with its usage in the erotic context.
1.1. Removal of a Garment (in an erotic context)
In sharp contrast to its moral meaning of “the rule of conduct”, the
word appears in an erotic context. Two examples suffice to illustrate the
situation.
uttar¥ya-vinayåt trapamåˆå rundhat¥ kila tad-¥k¡aâ-mårgam
åvari¡†a vika†ena vivo hur vak¡asaiva kuca-maˆ alam anyå.
Íißupålavadha 10.42
Another woman, ashamed of the removal of her upper garment
(uttar¥ya-vinaya) [by the hand of her husband, with the intention
of] obstructing the line of his sight, covered her plump breasts by
means of the broad chest of her husband.3
It is because of shyness to disclose her breasts that a woman wishes to
have the tight embrace of her husband.
A similar use is also seen in its verbal usage (vinayat-).
ambara vinayata priya-påêr yo¡itaß ca karayo kalahasya
våraˆåm iva vidhåtum abh¥k¡â kak¡yayå ca valayaiß ca
ßißiñje.
Girdle and bracelets twanged incessantly to ward off, so to
speak, a quarrel between the beloved’s hand [which tries to]
remove her garment.
It is remarkable that the word vinaya is used in the sense of the

removal of a garment (uttar¥ya-, ambara-) of a woman in love-making
by the hand of her lover.4
3 kucåßukåkar¡aâ ´: (Mallinåtha) “slipping off of garment covering her

breasts”.
4Another example is taken from a romantic context, though not so erotic. Here
the etymological meaning of “removal” (vi- and n¥-) can be observed.
vinayati sud¤ßo d¤ßa paråga praâyini kausumam ånanånilena
tad ahita-yuvater abh¥k¡âm ak¡ôr dvayam api ro¡a-rajobhir åpupËre.
A Note on vinaya 287
1.2. Removal of physical difficulties
As we have the Pali compound pipåsa-vinaya in A II 34 ,25 or
pipåså-vinayana in Mil 318,25-27, so we have t¤¡ˆå-vinayana in MBh.
1.2.1. t¤¡ˆå- (thirst)
Regretting what he has done, the old king Dh¤tarå¡†ra says as
follows:
caturthe niyate kåle kadåcid api cå¡†ame
t¤¡ˆå-vinayana bhuñje gåndhår¥ veda tan mama.
MBh.15.5.10
Now at the fixed time, that is, at the fourth division of the day or
sometimes at the eighth division, I take a little for quenching my
thirst (t¤¡ˆå-vinayana). [My wife] Gåndhår¥ knows this.
1.2.2. adhva-ßrama (fatigue)

ås¥nånå surabhita-ßila nåbhi-gandhair m¤gåˆåm
tasyå eva prabhavam acala pråpya gaura tu¡årai
vak¡yasy adhva-ßrama-vinayane tasya ß¤¤ge ni¡aˆâ
ßobhå ßubhra-trinayana-v¤¡otkhåta-pa¤kopameyåm.
MeghadËta 52
When thou hast come to the source of that river, the mountain
white with hoar-frost, where seated deer perfume the rocks with
must, settle on its peak to relieve the weariness of thy journey
(ßrama-vinayana), and thou shalt assume beauty that matches
clay cast up on himself by Shiva’s lustrous bull.
Tr. Edgerton
1.2.3. kapola-kaˆ u (the itch of the temple [of an elephant])

In describing the Mt. Himalaya, it is said,
kapola-kaˆ Ë karibhir vinetu
vigha††itånå sarala-drumåˆåm
While a lover was removing (vinayati = apanayati (Mallinåtha)) the pollen of

a flower from the eye of the charming-eyed one by means of his mouth-
breath, both eyes of her rival-lady were immediately filled with the dusts
(passion: rajas) of [jealous] anger.
288 Minoru Hara
yatra sruta-k¡¥ratayå prasËta
sånËni gandha surabh¥-karoti.
Kumårasabhava 1.9
Where the perfume arising from the milk-emitting nature of the
pine trees, rubbed by elephants for allaying (vinetum) the itch of
[their] temples, renders the summits fragrant.
1.3. Removal of mental difficulties

As we have the compound parissaya-vinaya (dispelling dangers) in
Sn 92, the phrase ka¤kha vinaya no ise (remove our doubt, O great
one) in M II 143,125 and its verbal form hadaya-pari¬åha vinaya in
Mil 318,4, the word is also construed with mental difficulty in Hindu
texts.
1.3.1. dukha (grief)
Seeing the increasing number of creatures and intending to lighten
the burden of the earth, Prajåpati created a girl m¤tyu (death) and asked
her to kill the creatures. She was frightened and entreated him to
relinquish his design.
vin¥ya dukham abalå så tv at¥våyatek¡aˆå
uvåca pråñjalir bhËtvå latevåvarjitå tadå.
MBh.12.250.1
Having driven off her grief, the large-eyed woman said with
joined hands and bending [her body] like a creeper …6
1.3.2. bhaya (fear)

Toward the end of the great war Yudhi¡†hira addressed Duryodhana,
who hid himself in a pond, as follows,
sa tvam utti¡†ha yudhyasva vin¥ya bhayam åtmana
5Cf. Sn 58, 559, 1025, Ja V 501,12, VI 375,20 and VI 222,15, 19, 21.
6 vin¥ya khalu tad dukham åga vaimanasya-jam
dhyåtavya manaså h¤dya kalyåâ savijånatå.
MBh.12.219.6
ramasva råjan piba cådya våruˆ¥ kuru¡va k¤tyåni vin¥ya dukham
mayådya råme gamite yama-k¡aya ciråya s¥tå vaßagå bhavi¡yati.
R.6.63.56 Bombay
ghåtayitvå sarva-sainya bhråt•ß caiva suyodhana (27)
nedån¥ j¥vite buddhi kåryå dharma-cik¥r¡ayå.
MBh.9.30.28ab
Arise and fight, casting off fear for yourself! Having caused all
your troops and brothers to be slain, O Suyodhana, now you
should not think of [saving your] life, if you wish to do justice!
1.3.3. jvara (affliction)

Kumbhakarâ encouraged his brother Råvaâ in distress, saying,
ramasva kåma piba cågrya-våruˆ¥
kuru¡va k¤tyåni vin¥yatå jvara
mayådya råme gamite yama-k¡aya
ciråya s¥tå vaßagå bhavi¡yati.
R.6.51.47
Make love, drink wine, do what you have to do and banish
affliction!
Today, when I send Råma to the abode of death, S¥tå will
become yours for ever.
1.3.4. åyåsa (distress)

When Bharata was summoned to Ayodhyå by Kaikey¥, he saw a terrible
dream and was greatly distressed.
tapyamåna samåjñåya vayasyå priya-vådina
åyåsa hi vine¡yanta sabhåyå cakrire kathå.
R.2.63.3
Observing how troubled he was, his affable companions tried to
ease his distress by engaging him in conversation in the
assembly hall.
Tr. Pollock
1.3.5. h¤daya-granthi (knot in the heart)
As we have hadaya-pari¬åha vinaya (remove my heart-burning
pain) in Mil 318 ,4 , so we have the expression granthi vin¥ya
h¤dayasya. The sage Ótreya in the disguise of a hasa bird encouraged
the Sådhyas as follows:
etat kåryam amarå saßruta me
dh¤ti ßama satya-dharmånuv¤tti
290 Minoru Hara
granthi vin¥ya h¤dayasya sarva
priyåpriye cåtma-vaßa nay¥ta.
MBh.5.36.4
I have learnt, Immortals, that this is one’s task; to be steady and
serene and to pursue truth and Law; having undone all the knots
of the heart, one should bring both the pleasant and the
unpleasant under control.
Tr. van Buitenen
1.3.6. asËya (jealousy)

Furthermore, its verbal form takes asËya in the accusative case.
Despite the request of Dh¤tarå¡†ra, Sajaya refused to speak in secret
and promised to speak out only in the presence of Vyåsa and Gåndhår¥.
tau te asËyå vinayetå narendra
dharmajñau tau nipuâu nißcayajñau
tayos tu två sanidhau tad vadeya
k¤tsna mata våsudevårjunåbhyåm.
MBh.5.65.7
For both of them, clever, knowing justice and resolute, can
dispel any ill-feeling you might cherish [against me].
In their presence I shall tell you all that K¤¡â and Arjuna have
in their mind.
1.4. Removal of yuddha-ßraddhå7

In the epic battle scene, we often meet warriors’ determination to
mar the warlike spirit (yuddha-ßraddhå)8 of their adversary. It is natural
in these contexts for vi-n¥- to appear in the future tense, either in the
simple form or in the periphrastic one.
ti¡†ha ti¡†ha na me j¥van droâ-putra gami¡yati
yuddha-ßraddhåm aha te ’dya vine¡yåmi raˆåjire.
MBh.7.131.62
7Cf. Hara 1992.

8 The literary meaning would be “the conviction (ßraddhå) [of victory] in
battle”.
Wait, wait, O son of Droâ, you shall not go [escaping] from me
with your life.9 I shall today dispel your eagerness for fighting.
1.5. darpa (arrogance)

In a similar context of battle we also meet darpa.
anuktvå samare tåta ßËrå yudhyanti ßaktita
sa yudhyasva mayå ßaktyå vine¡ye darpam adya te.
MBh.8.17.54
Heroic men fight their utmost in the battlefield without saying
anything. Thus, fight with me to the utmost. Today, I shall
destroy your arrogance.10
1.6. Removal of vices (= restraint or control)

The removal of jealousy (asËya) and arrogance (darpa) leads us to
that of vices in general. It is from this “removal of vices” onward that
the word vinaya is imbued with the tinge of moral and disciplinary
meaning.
9Hara 1996.
10v.l.
hani¡ye, vinaßyed for vine¡ye.
Cf. also,
ekaikaßa samarthå smo vijetu sarva-pårthivån
ågacchantu vine¡yåmi darpam e¡å ßitai ßarai.
MBh.5.54.19
e¡o yotsyåmi va sarvån nivårya ßara-våguråm
ti¡†hadhva yuddha-manaso darpa vinayitåsmi va.
MBh.14.77.5
ß¥ghram eva hi råk¡asyo vik¤tå ghora-darßanå
darpam asya hi vine¡yantu måsa-ßoîta-bhojanå.
R.3.54.24
The contrast between darpa and vinaya is also discerned in the following
passages´:
tad yuddham abhavad ghora deva-dånava-sakulam
k¡amå-paråkrama-maya darpasya vinayasya ca.
H.37.21
tat suråsura-sayukta yuddham atyadbhuta babhau
dharmådharma-samåyukta darpeâ vinayena ca.
H.35.3
292 Minoru Hara
1.6.1. In Pali text, the word vinaya is compounded with various kinds of
vice, such as
asmi-måna (the sense of ego) (Vin, I 3,30),
icchå (desire) (D III 252,19–20, A IV 15,12–13, V 165,7–8),
kodha (anger) (A I 91,20, A V165,12–13, S II 282,20),
upanåha (grudge) (A I 91,20),
gedha (greed) (Sn 152, 1098),
makkha (hypocrisy) (A V 165,17–18, S II 282,20),
måna (pride) (S II 282,20),
så†heyyå (treachery) (A V 165,22-23),
måyå (fraud) (A V 165,26-28),
chanda-råga (exciting desire) (S IV 7,9, 13–14, 19)
råga (lust) (S V 137,25, 241,24),
dosa (hatred) (S V 37,25, 241,24),
moha (delusion) (S V 137,26, 241,24)
bhaku†i (superciliousness) (Sn 485).
Yet, the commonest construction of vinaya with vices is sum-

marized in that of akusala dhamma in the plural, where these three
(råga, dosa, moha) are also included. For example : ´
aha hi S¥ha vinayåya dhamma desemi rågassa dosassa

mohassa aneka-vihitåna påpakåna akusalåna dhammåna
vinayåya dhamma desemi.
Vin I 235,26–28 III 3,9–11 = A IV175,7–9
O S¥ha, I am teaching this dhamma in order to remove lust,
hatred and delusion. [That is to say,] I am teaching this dhamma
in order to remove evil and unfit qualities of various sorts.11
1.6.2. However, in classical Sanskrit literature, these vices are mostly

preceded by vin¥ta- as is seen in such compounds as vin¥ta-råga
(MBh.12.172.37), vin¥ta-moha (MBh.12.237.35), vin¥ta-krodha-har¡a
(MBh.5.88.6), vin¥ta-ro¡a-t¤¡ˆå (MBh.12.172.36), vin¥ta-kilbi¡a
(MBh.5.193.29). Of these, the commonest one is vin¥tåtman.
11For samukhå-vinaya, sati-vinaya, amË¬ha-vinaya, cf. D III 254,12.

ta sa-dåro vinitåtmå sugr¥va plavagådhipa
pËjayå pratijagråha pr¥yamåâs tad arhayå.
MBh.3.266.13
With his wife, the courteous (vin¥tåtman) king of apes, Sugr¥va,
kindly received him with the honor due to him.
Tr. van Buitenen
1.7. Training
From the “removal” of vices, it is an easy transition of meaning to
that of shortcomings in general, that is “training”. Yet, this “training” is
exercised not only to human beings, but also to animals.
1.7.1. Taming of wild animals
In praise of Råma we read,
årohe vinaye caiva yukto våraâ-våjinåm
dhanurveda-vidå ßre¡†ho loke ‘tiratha-samata.
R.2.1.23
He was proficient in riding (åroha) and the training (vinaya) of
horses and elephants, and was regarded as the best expert among
masters of martial arts.12
12For vin¥ta-sattva, cf.

tapasvi-sasarga-vin¥ta-sattve tapo-vane v¥ta-bhayå vasåsmin.
Raghuvaßa 14.75ab
For vin¥ta-m¤ga, cf.
agastyasyåßrama ßr¥mån vin¥ta-m¤ga-sevita.
R.3.10.84
For elephants, cf.
vinaya-vidhåyini bhagne ‘pi cå¤kuße vidyata eva vyåla-våraâsya
vinayåya sakala-matta-måta¤ga-kumbha-sthala-sthira-ßirobhåga-bhidura
kharatara kesari-nakhara.
Har¡acarita 188.14-6
prabhinnåß ca mahånågå vin¥tå hasti-sådibhi
paraspara samåsådya sanipetus abh¥tavat.
MBh.6.91.26
ye tv ete sumahå-någå añjanasya kulodbhavå
idhodaka-pradåtåra ßËnya-pålakam åßrame
294 Minoru Hara
1.7.2. Training of youths
The “taming of animals” is akin to “the training” of children.
Human beings in the prime of their youth should be trained and
cultivated. Hence its construction with ßaißava, yauvana, etc.
1.7.2.1. ßaißava
The sage Divåkaramitra praises the inborn courtesy of Har¡a as
follows :´
asya tv ¥d¤ße ßaißave vinayasyopådhyåya dhyåyann api na

sabhåvayåmi bhuvi.
Har¡acarita 239.25–26
In my pondering, I cannot imagine any instructor in decorum on
the earth (vinayasyopådhyåya) in his childhood.
1.7.2.2. ßißutva
Mischief (a-vinaya) is natural to childhood. Hence its construction
with ßißutva and båla-bhåva.
janaka iva ßißutve supriyasyaika-sËnor
vin¥tam åcårya-kule suyukta guru-karmaî.

MBh.13.105.9
For horses, cf.
aßvånå prak¤ti vedmi vinaya cåpi sarvaßa
du¡†ånå pratipatti ca k¤tsna caiva cikitsitam.
MBh.4.11.7
vinayanta javenåßvån mahåråjasya paßyata.
MBh.4.18.32
karkåßåß ca vin¥tåß ca prabhinna-kara†åmukhå.
MBh.7.87.33
For bulls, cf.
tasmai pradeya pråyacchat pr¥to råjå dhana bahu
vin¥tån v¤¡abhån d¤¡†vå sahadevasya cåbhibho.
MBh.4.12.31
tathåna våha bråhmaˆåyåtha dhurya
dattvå yuvåna balina vin¥tam
halasya vo håram ananta-v¥rya
pråpnoti lokån daßadhenudasya.
MBh.13.72.43
avinayam api sehe påˆ avasya smaråri.
Kiråtårjun¥ya 17.64cd
Íiva put up with Arjuna’s indecorum (avinaya), as a father puts
up with his loving single son in his childhood.13
1.7.2.3. yauvana
Youth and decorum are often incompatible. In the praise of
Mådhavagupta we read,
paraspara-viruddhayor vinaya-yauvanayoß ciråt prathama-
sagama-cihnam iva bhrË-sagatakena kathayantam …
Har¡acarita 139.20
His meeting brows seemed to suggest the reconciliation after a
long time of those irreconcilables, youth and decorum …
1.7.2.4. abhinava-yauvana
avinaya-bahulatayå cåbhinava-yauvanasya …
Kådambar¥ 270.5
Since the prime of youth (abhinava-yauvana) is full of ill-
behaviour or mischief (avinaya) …
1.8. Education
Cultivation of youths is nothing but “education”. As we have abhi-
dhamme vinesi (“taught in the Abhidhamma”) in Mil 12,19, 13,19–20, the
word has the meaning of education.
1.8.1. Martial arts
pËrvam ahar-bhåga hasty-aßva-ratha-praharaâ-vidyåsu
13 sa evai¡a puna svayak¤tenåvinayena …

Kådambar¥ 578.9
avinaya-nißcetano nakha-pratibimbitam åtmåna bahu manyate.
Kådambar¥ 410.7
As for båla-bhåva, see H.2.51.1-2 and 4 (Bombay) which is paraphrased
by putra-durnaya in H.2.51.3. Cf. also Har¡acarita 78.11 where an elephant-
charmer gives instruction to a youth while scolding a young elephant “give up
fickleness (lolatå) and practise courtesy (vinaya-vrata)”.
kari-kalabha vimuñca lolatå cara vinaya-vratam ånatånana.
Har¡acarita 78.11
296 Minoru Hara
vinaya gacchet.
KAS.1.5.12
During the first part of the day, he (the prince) should undergo
training in the arts of [using] elephants, horses, chariots and
weapons.
Tr. Kangle14
1.8.2. Arts (kalå)

In enumerating the education of the courtesans, we read,
n¤tya-g¥ta-vådiya-nå†ya-citråsvådya-gandha-pu¡pakalåsu lipi-
jñåna-vacas-kaußalådi¡u ca sayag-vinayanam.
Daßakumåracarita 81.2-3
Proper training in dancing, singing, musical instruments, theatre,
painting, cooking, perfume, flower-arrangement …15
1.9. Courtesy and Decorum

As the result of training and education, a youth learns how to
behave, and becomes modest and courteous. The meaning of “moral
decipline” is now well-established in Hinduism. The N¥tivåkyåm¤ta
defines it as follows :
´
vrata-vidyå-vayo ’dhike¡u n¥cair åcaraâ vinayam (6)
14Cf. KAS.9.2.24 (praharaâ-vidyå-vin¥ta tu k¡atriya-bala ßreya). Cf. also

MBh.6.15.41 (sarvåstra-vinayopeta) and MBh.1.181.15 (ßastråstra-vinaya).
Furthermore,
viße¡årth¥ tato bh¥¡ma pautråˆå vinayepsayå
i¡v-astra-jñån paryap¤cchad åcåryån v¥rya-samatån (1)
nålpa-dh¥r nåmahåabhågas tathå-nånåstra-kovida
nådeva-sattvo vinayet kurËn astre mahå-balån.
MBh.1.121.2
gadåsi-carma-grahaê¡u ßËrån astre¡u ßik¡åsu rathåßva-yåne
sayag vinetå vinayaty atandr¥s tåß cåbhimanyu satata
kumåra
MBh.3.180.28
15Cf.
så råjå-hasair iva sanatå¤g¥ gate¡u l¥låñcita-vikrame¡u
vyan¥yata pratyupadeßa-lubdhair åditsubhir nËpura-siñjitåni.
Kumåra-sabhava 1.34
puˆyåvåpti ßåstra-rahasya-parijñåna sat-puru¡åbhigamya
ca vinaya-phalam
N¥tivåkyåm¤ta 11.7
Vinaya is to behave humbly toward those who are advanced in
religious observance (vrata), knowledge (vidyå) 16 and age
(vayas).17
The fruits thereof are the attainment of religious merit, insight
into the secrets of scriptures and association with good people.18
But in Hindu political literature, this quality is particularly desirable

for the education of the young prince.19
tebhyo ‘dhigacched vinaya vin¥tåtmåpi nityaßa
vin¥tåtmå hi n¤patir na vinaßyati karhi cit.
MS.7.39
Let him, though he may already be modest (vin¥tåtman), con-
stantly learn modesty (vinaya) from them (the elders); for a king
who is modest never perishes.
Tr. Bühler20
In describing Råma and his brothers, it is said,

svåbhåvika vin¥tatva te¡å vinaya-karmaˆå
mumËrcha sahaja tejo havi¡eva havir-bhujåm.
Raghuvaßa 10.79
Their inborn courtesy21 became stabilized22 by education,23 as
16Cf. KAS.1.5.11´: nityaß ca vidyå-v¤ddha-sayogo vinaya-v¤ddhy-artham, tan-

mËlatvåd vinayasys.
17 Though the first chapter of KAS is called vinayådhikårika “the topic of
training” (Kangle), “von Sachen der Erziehung und des Wohlverhaltens”
(Meyer), there is no definition of the word vinaya as such.
18Cf. Kane 51–52 and Botto 82–83.
19For king and vinaya, cf. Kane III 51–53.
20For its opposite avin¥taß ca du¡†åtmå, cf. MBh.4.20.25.
21 The youth of a noble-family is courteous by birth. Hence the expressions
svåbhåvika vin¥tatva here, and svabhåva-vin¥ta in R.2.17.13 and vinaya
sahaja in Uttararåmacarita 4.22.
22Cf. Hara 2000C.
23 For the inborn courtesy and later education, cf. k¤taka and svåbhåvika in
298 Minoru Hara
the innate splendour of fire becomes strengthened by clarified
butter.24
1.10. a-vinaya (insolence, mischief)
In the story of retribution we meet often the word avinaya. Ugly-
looking Kabandha tells Lak¡maâ the story of his previous life.
virËpa yac ca me rËpa pråpta hy avinayåd yathå
tan me ß¤û naravyåghra tattvata ßasatas tava.
R.3.66.15
Listen, tiger among men, truly I shall tell you why my form was
deformed through an act of insolence.
Also in Kådambar¥ we read,

tad ya sa kåmopahata-cetå svaya-k¤tåd evåvinayåd divya-
lokata paribhraßyan martya-loke vaißampåyana-nåmå
ßukanåsa-sËnur abhavat.
Kådambar¥ 578.8–9
Stupefied by love, he fell from heaven to the mortal world
because of his own mischief, and was born as the son of
Íukanåsa with the name of Vaißapåyana.25
1.11. In the above, we have surveyed step by step the various aspects of
vinaya, whose original meaning is “removal”. The original meaning is
apparently characterized by the “disjunctive” function of the prefix vi-,
and the usual meanings of “modesty” and “moral discipline” are later
developed in the course of its association with vices in general. Next we
shall proceed to its special meaning in Sanskrit literature.
KAS.1.5.3-5.
24 Forother expressions “modest” and “courteous”, cf. vinayånvita (MBh.
13.76.1), vinayopeta (MBh.12.285.38, 14.35.18) and vinaya-sapanna
(MBh.1.106.14, 2.5.29, 6.27.17). For vin¥ta-ve¡a (soberly dressed), cf.
MS.8.2.
25For the adjectives avin¥ta (-putra), cf. MBh.5.133.9, KAS.1.17.51 and for
durvin¥ta, cf. R.3.18.9, 7.53.18, 7.30.34.
2. Its particular use in Sanskrit
2.1. vi-naya in the sense of the absence of naya.
Beside its disjunctive function, the prefix vi- indicates absence
(yoga : viyoga) or reverse (kraya : vikraya). As a result, it is possible for
´ ´
vi-naya to mean the absence of naya. It is this possibility of which the

skilled writers in Kåvya took advantage and succeeded in composing a
verse with the double-entendre. We shall see the skill of Mågha in his
Íißupålavadha, where a verse can be read in the two ways of praise and
blame.
ahitåd anapatrapas trasann atimåtrojjhita-bh¥r anåstika
vinayopahitas tvayå kuta sad¤ßo ‘nyo guâvån avismaya.
2.1.1. The first meaning, in the good sense : ´
Where is someone else virtuous (guâvat) equal to you, pos-

sessed of decorum (vinayopahita), afraid of (trasan) evils
(ahita), prudent (an-apa-trapa), yet tremendously brave (ujjhita-
bh¥), pious (a-nåstika) and without arrogance (avismaya)´?
2.2.2. The second meaning, in the bad sense (paru¡a) : ´
Where is someone else unvirtuous (aguâvat) equal to you,

afraid of the enemy (ahita = ßatru), shameless (an-apatrapa =
nirlajja). cowardly (nati-måtrojjhita-bh¥) (literally, “escaping
fear only by obeisance” = “without fighting bravely”), an atheist
(an-åstika), without policy, and yet arrogant?
According to Mallinåtha, here the compound vinayopahita in the

first reading is vinayenånauddhatyenopahito vißi¡†a, taking vinaya in
the sense of anauddhatya (freedom from pride, modesty).
In the second reading, the compound is divided as vinayo ’pahita,
and vinaya is used in the sense of nayåt¥to (gone beyond good policy
(naya), that is, neglected policy) and apahita is taken in the sense of
hitåd apeta (deviating from the beneficial).26
26Forother word-play of a-naya, vi-naya, cf.

tasya tat pråpya du¡pråpyam aißvarya muni-satk¤tam
300 Minoru Hara
2.2. vinaya as a repetition of naya
The repetition of a word with its prefix is not uncommon in Pali,27
but it appears also in Epic Sanskrit.28 For example, the prefix upa- is
used in vana, upa-vana, diß, upa-diß and also in ni¡ad, upa-î¡ad.29
People often translate upa-vana as “small wood”, and upa-diß as
“intermediate quarter”. The same is also the case with the prefix vi-, for
we have such examples as diß, vi-diß, dhåtå, vi-dhåtå, jñåna, vi-jñåna
and, here including naya, vi-naya. We notice that in addition to the
“disjunctive” function, the prefix vi- here has a function of
“differentiation”. We shall see examples below.
2.2.1. diß and vi-diß
så råja-bhuja-nirmuktå nirmuktoraga-sanibhå
prajvålayant¥ gagana dißaß ca vidißas tathå
droˆåntikam anupråptå d¥ptåsyå pannag¥ yathå.
MBh.7.81.31
The spear (ßakti), hurled from the king’s arm, reached close to
Droâ, burning the sky and various quarters, like a female snake
with gleaming mouth which has just cast off her skin.30
d¥po yathå nirv¤tim abhyupeto naivåvani gacchati nåntarik¡am
dißa na kåcid vidißa na kåcit sneha-k¡ayåt kevalam eti
ßåntim.
Saundarananda 16.28
vibabhråma matis tåta vinayåd anayåhatå.

H.20.28
For naya, apa-naya, vi-naya, cf.
vinaya guˆå iva vivekam apanaya-bhida nayå iva
nyåyam avadhaya ivåßaraˆå ßaraâ yayu ßivam atho mahr¡aya.
Kiråtårjun¥ya 12.17
27 Cf. Allon 199 note and 248 (kampati, sakampati, sapakampati) and
Dhadphale 217 (neti, vineti, anuneti), 222 (kampi, sakampi, sampakampi),
225 (jhåyanti, pajjhåyanti, nijjhåyanti, apajjhåyanti).
28Hara 2000A.
29Hara 2000B.
30Cf. MBh.13.151.27, H.31.37, R.6.66.27.
Just as a lamp, which has reached the stage of extinction, does
not depart to the earth or the sky or any of the quarters or inter-
mediate quarters but from exhaustion of the oil merely goes out.
Tr. Johnston31
As Johnston takes it, the word vi-diß means the intermediate
quarters, differentiating the preceding word diß.
2.2.2. ße¡a-, viße¡a
råjño ’pi våso-yugam ekam eva k¡ut-sanirodhåya tathånna-
måtrå
ßayyå tathaikåsanam ekam eva ße¡å viße¡å n¤pater madåya.
Buddhacarita 11.48
A king too can only wear one pair of garments and similarly take
only a certain measure of food to still his hunger: so he can only
use one bed, only one seat. The other luxuries of a king lead only
to the intoxication of pride.
Tr. Johnston
Here Johnston takes ße¡å viße¡å as “the other (ße¡a) luxuries

(viße¡a) (of a king)”, but one may interpret the second word vi-ße¡a as a
specification or differentiation of ße¡a. Then ßeßa, viße¡a means “the
rest”, or “the various remainings, large and small”.
2.2.3. dhåtå-, vidhåtå32
These two are often rendered into “creator” and “distributor”, but
we may take them in the sense of “various gods”. In enumerating the
gods who attended the rite of the royal consecration of Skanda
Kårtikeya, the text says,
indrå-vi¡ˆË mahå-v¥ryau sËryå-candramasau tathå
dhåtå caiva vidhåtå ca tathå caivånilånalau.
MBh.9.44.4
Indra and Vi¡û of great energy, similarly the sun and moon, and
dhåt¤ and vidhåt¤, wind and fire33 …
31Cf. Saundarananda 16.29.

32Cf. Durga ad Nirukta 11.11 (dhåtaiva vidhåtå) as quoted in Dhadphale 223.
33Cf. MBh.7.69.46, 13.15.31, 13.145.39 (sa dhåtå vidhåtå …), 3.249.4 (dhåtur
302 Minoru Hara
The gods here enumerated (the sun and moon, wind and fire)
compose the typical pairs, and we do not need to take the second vi-
dhåt¤ in the sense of “distributor”. Here dhåtå, vidhåtå simply means
“various gods”.
2.2.4. jñåna-, vijñåna
As is well-known, F. Edgerton proposed to translate jñåna as
“theoretical knowledge” and vijñåna as “practical knowledge”.34 Thus
he translated, for example,
tasmåt tvam indriyåˆy ådau niyamya bharatar¡abha
påpmåna prajahi hy ena jñåna-vijñåna-nåßanam.
BhG.3.41
Thou, therefore, the senses first controlling, O bull of Bharatas,
smite down this evil one, that destroys theoretical and practical
knowledge.
Tr. Edgerton35
But we may take them in a similar way, taking them simply as “various
sorts of knowledge”.
2.2.5. naya-, vi-naya
Now we come to naya vinaya.36 As diß-, vi-diß mean “quarters and
vidhåtu), 12.224.49 (dhåtaiva vidadhåty uta), R.7.20.24 (yo vidhåtå ca dhåtå

ca suk¤te du¡k¤te tathå).
34Edgerton 1933.
35Cf. BhG 6.8, MS.9.41: tat-pråjñena vin¥tena jñåna-vijñåna-vedinå.
36 Of course, we would not entirely exclude the meaning of “modesty” for
vinaya, even when it appears in conjunction with naya. For example,
vipanne ca samårambhe satåpa må sma vai k¤thå
gha†ate vinayas tåta råjñåm e¡a naya para.
MBh.12.56.16
Even when some undertaking has failed, do not grieve! [In such a case]
modesty (vinaya) is proper (to be followed), for modesty is the highest
policy (naya) of kings.
We have mentioned above that courtesy (vinaya) is indispensable for the
education of princes.
intermediate quarters”, that is “various quarters” as a whole,37 so naya-,
vi-naya may mean “various naya (policy, stratagems)”, instead of
“policy and modesty”. Below we shall list some examples where vinaya
is used in this sense.
2.2.5.1. In the self-praise of Kasa, we read,
aha balena v¥ryena nayena vinayena ca
prabhåveâiva ßauryeâ tejaså vikrameâ ca
satyena caiva dånena nånyo ‘sti sad¤ßa pumån.
H.2.28.113 Bombay = H.73.822*7-9
Nobody is equal to me with respect to power, energy, majesty,
[using] various stratagems, heroism, splendour, courage,
truthfulness and giving.38
2.2.5.2. In describing K¤¡â, we read,

manu¡yåˆå mano-bhËtas tapo-bhËtas tapasvinåm
vinayo naya-v¤ttånå tejas tejasvinåm api.
H.30.36
He is the [true]39 mind of men, and the [true] asceticism of the
ascetics, vinaya of naya-v¤ttas, and the splendour of the
splendorous.40
Though we have here vinaya metri causa, otherwise we might

expect naya-bhËta in parallel with mano-bhËta, tapo-bhËta in the first
line, or naya with tejas in pada d. In this context, it is not necessary to
take vinaya as being independent of naya in the sense of “modesty”, or
“moral discipline”, but “the [true] naya of naya-v¤ttas (the true [=
distinctive] policy among politicians)”.
37We have similar constructions in diß-, upa-diß-, pra-diß-, pratidiß- also.

38Except for satya and dåna in the last line, all the items enumerated in the first
two lines are concepts of a military and heroic nature. If so, vinaya in the first
line is not necessarily to be taken in the sense of “modesty”, but goes together
with naya in the sense of “various” nayas.
39Here I take -bhËta in the sense of “true”, as is the case with caitya-bhËta and
kumåra-bhËta. Cf. Schopen.
40Cf. Våyu-puråâ 97.42 which has vinayo naya-t¤ptånå in c.
304 Minoru Hara
2.2.5.3. In his monologue, Yaugandharåyaâ says as follows : ´
vaira bhaya paribhava ca sama vihåya

k¤två nayaiß ca vinayaiß ca ßaraiß ca karma
ßatro ßriya ca suh¤dåm ayaßaß ca hitvå
pråpto jayaß ca n¤-patiß ca mahåß ca ßabda.
Pratijñåyaugandharåyaâ 4.6
[For I,] disregarding enmity, fear, and insult alike, have
accomplished my work with my designs, by self-control and
arrows, ending the glory of the foe and the disgrace of my
friends. Thereby have I won victory, the king himself, and great
renown.
Tr. Woolner
Though Woolner takes naya in the sense of “design” and vinaya

“self-control”, we may take them in the sense of various sorts of naya
(stratagem, tactic), the diplomatic means, both of which are contrasted
to ßara (arrow), the military means.41
2.2.5.4. Similarly,
ß¥lena såmnå vinayena s¥tå
nayena na pråpsyasi cen narendra
tata samutsådaya hema-pu¤khair
mahendra-vajra-pratimai ßaraughai.
R.3.61.16
If you cannot recover S¥tå by peaceful means (ß¥la), by concilia-
tion (såman), tact (vinaya), or diplomacy (naya), lord of men,
then unleash the flood of your gold-feathered arrows, as
devastating as great Indra’s thunder-bolts.
Tr. Pollock
41Apparently,the second line refers to the various means of defeating an enemy.

According to the n¥ti literature, the means (upåya) consists of peaceful tactics
(såman, dåna, and bheda) and violent means (daˆ a). In the above verse,
naya, vinaya correspond to peaceful means, while ßara to the violent one.
Pollock rightly takes vinaya as “tact”. Here again the second line
refers to arrows, a violent means, whereas the first line refers to
peaceful diplomatic means.42 Under such circumstances, we do not
necessarily take the words naya and vinaya literally, that is, in the sense
of “tactics” and “self-control”.
3. Other special meanings

Apart from vi- with the “disjunctive” and “differenciating”
function, we shall list below three special meanings of vinaya as
induced from context and commentary literature.
3.1. vinaya in the sense of niga a (fetter)
In the Pratijñåyaugandharåyaâ, a chamberlain tells Pradyota how
Vatsa is as a captive as follows :´
kåñcuk¥ya — åhita-vinayatvåt pådayor a¤ge tasya bahu-

prahåratvåc ca skandha-våhyena ßayan¥yena madhyama-g¤he
praveßita.
Pratijñåyaugandharåyaâ 2.13.10-11
Chamberlain´: He was so tractable and had so many wounds on
his feet and body, that he was carried into the Middle Palace on a
litter.
Tr. Woolner
Though Woolner translated åhita-vinaya as “tractable”, here the

compound should be taken in the sense of “having fetters placed”, as
Ganapatisastri commented (åhita-vinayatvåd vin¥yate ‘nenåparådh¥ti
vinaya iha nigala, sa åhito niveßito yasya sa åhita-vinaya). Here the
passage describes king Vatsa as a captive, having his feet and body
fettered as a criminal (aparådhin).
42Asis well known, såman (conciliation), dåna (bribery), bheda (sowing dis-
sension) and daˆ a (open attack) are means of success against an enemy
(MS.7.109, KAS.7.16.3). Here the first line corresponds to peaceful means
and ßara in the second line to violent means. Cf. also, MBh.12.223.8,
R.2.37.5, 4.17.28, 4.18.8, R.6.128.82 (Bombay), H.2.28.113 (Bombay).
306 Minoru Hara
3.2. vinaya as contrasted to visarga
Bh¥¡ma enumerates various aspects of d a ˆ a , the rod of
punishment.
aßakti ßaktir ity eva måna-stambhau vyayåvyayau
vinayaß ca visargaß ca kålåkålau ca bhårata.
MBh.12.121.28
[It is] power and impotence, arrogance, obstinacy, change and
stability, discipline and letting loose, the right time and the
wrong time.
Tr. Fitzgerald
Since positive and negative concepts are contrasted here, vinaya is

the reverse of visarga and thus pada c should be translated something
like “restraint and freedom”.
It is interesting to note that vinaya which means originally
“removal” here comes to mean “confinement [in fetters]” or “restraint”,
nearly the reverse of its original meaning of “taking away”.43
3.3. vin¥ta in the sense of prasårita (stretched, strewn)
Finally, we shall examine vin¥ta in the sense of prasårita. BR lists
under vi-n¥- (ausbreiten) an example from the Råmåyaâ. Prior to the
well-known scene of S¥tå’s abduction, she was curious about a golden
deer and asked Råma to capture it, saying:
nihatasyåsya sattvasya jåmbËnada-maya-tvaci
ßa¡pab¤syå vin¥tåyåm icchåmy aham upåsitum.
R.3.41.19
Were the creature to be killed, I should like his golden skin to be
stretched over a cushion of straw, to make a seat.
Tr. Pollock44
43This meaning of vinaya as confinement and restraint may be related to that

found in Pali ariyassa vinaye, which means “in the restricted sense of aryan”,
that is, “in the Buddhist sense of the term”. However, the discussion on this
subject needs another lengthy paper to be written.
44The commentary reads as follows´:
ßa¡pa-b¤syå båla-t¤â-parikalpita-tåpasåsane vin¥tåyå prasåritåyåm asya
Above, we have discussed vinaya used neither in the original sense
of “removal”, nor in the ordinary sense of “discipline” in classical
Sanskrit literature.45 Finally, we shall examine an allegorical story of
vinaya and its relationship to lajjå.
4. vinaya and lajjå

4.1. It is believed that vinaya is essential to the well-bred woman, as is
expressed by the Sanskrit compound sådhv¥-vinaya. When Hanumån
extended his hand to help her, S¥tå refused his offer, being afraid of
touching a man’s hand other than her husband. In his praise, we read:
yukta-rËpa tvayå devi bhå¡ita ßubha-darßane
sad¤ßa str¥-svabhåvasya sådhv¥nå vinayasya ca.
R.5.36.2
Oh queen, charming lady, what you said is appropriate. It is
suitable for womenhood (str¥-svabhåva) and the providence
(vinaya) of a chaste women´!
4.2. So bashfulness (lajjå) is proper for the well-bred young lady.

asatu¡†å dvijå na¡†å satu¡†åß ca mah¥bhuja
salajjå gaîkå na¡†å nirlajjåß ca kula-striya.
Hitopadeßa 3.64
The following [four] perish´: twice-borns unsatisfied, kings
satisfied, bashful courtesans, and unbashful women of a noble
family (kula-str¥).
Then, how are these two, vinaya and lajjå, related to each other ? ´
jåmbËnada-maya-tvacy upåsitum tvat-sam¥pe sthåtum icchåmi/yad vå

bhagavad-upåsanå kartum ity artha.
Upon a golden hide, strewn (vin¥ta = prasårita) over an ascetic seat, prepared
with young kußa grass, I wish to sit in your side, or to serve you. That is the
meaning.
45For the meaning of Anstandsbusse (fine due to indecorum (Geldstrafe bei Un-
gebührlichkeit) = solatium´?), cf. Meyer 291 11ff. Cf. also Nårada-sm¤ti 6.21d
(“fine”, Lariviere translation 120).
308 Minoru Hara
4.3. It is in the KËrma Puråâ 1.8 that vinaya is allegorically called the
son of dharma and lajjå. Twenty-four daughers of Dak¡a are mentioned
there, and thirteen of them are married to Dharma. Here, lajjå is
enumerated as one of them, and she brought to her husband a son called
vinaya. The relevant passage reads as follows : ´
buddhyå bodha sutas tadvad apramådo vyajåyata

lajjåyå vinaya putro vapu¡o vyavasåyaka.
KËrma-puråâ 1.8.23
Enlightenment (bodha) was born of intelligence (buddhi), so was
mindfulness (apramåda). Decorum (vinaya) is the son of shame
(lajjå), the determinant (vyavasåyaka) is the son of a handsome
figure (vapus).46
4.4. The mother-son relationship between lajjå and vinaya can be

illustrated more concretely by some romantic passages of Båâ’s
Kådambar¥.
4.4.1. We read : ´
pråyeâ prathama madanånalo lajjå dahati, tato h¤dayam/

ådau vinayådika kusume¡u-ßarå khaˆ ayanti paßcån
marmåî.
Generally, the love-fire first burns bashfulness (lajjå), and then
the heart. Kåma’s arrows first attack decorum (vinaya), and later
the vital parts.
Once lajjå is defeated, it is an easy step for Kåma to destroy her son
vinaya.
4.4.2. Similarly,
skhalite cetasi tal-lagnå pataty eva lajjå/trapåvaraâ-ßËnye h¤di
pravißya pada kurvan kena vå nivårito durnivåra
46Cf.
lajjåyå vinaya putro vyavasåyo vaso suta.
Li¤ga-puråâ 70.296
lajjåyå vinaya putro vyavasåyo vasyo suta.
Våyu-puråâ 10.36
sarvåvinaya-hetu kusuma-dhanvå.
When the mind stumbles, shame (lajjå) which clings to it falls.
Into the heart devoid of its protecting cover (of lajjå), the flower-
bannered one (Kåma) enters. Once he has entered, who can drive
him away, the god who is hard to drive out and causes all sorts
of indecorum (avinaya)´?
When bashfulness is taken away, all sorts of decorum are exposed

to danger.
4.4.3. In her self-reproach, the well-bred lady in love laments as
follows : ´
yadi tåvad itara-kanyakeva vihåya lajjåm, uts¤jya dhairyam,

avamucya vinaya, acintayitvå janåpavådam, atikramya
sadåcåram, ulla¤ghya ß¥lam, avagaâyya kulam... svayam
upagamya gråhayåmi påîm/eva guru-janåtikramåd adharmo
mahån.
Kådambar¥ 297.3
If I approach by myself and grasp [his] hand [for marriage], like
a common girl — abandoning shame (lajjå), giving up steadfast-
ness (dhairya), unharnessing decorum (vinaya), neglecting
people’s rumour (janåpavåda), transgressing good conduct (sad-
åcåra), traversing morality (ß¥la), neglecting the noble family[-
ness] (kula) … — then, through the offence to my respected
elders, there would be a great sin.47
The urge of love-passion deprives the young lady of all the

virtues.48 All these passages illustrate the close connection of lajjå and
vinaya.
Minoru Hara
Tokyo
47Cf. Kådamabar¥ 354.14–355.4.

48Besides kåma, mada (alcholic drink) also destroys vinaya.
paißunyena kula madena vinayo dußce¡†ayå pauru¡a
dåridryeâ janådaro mamatayå cåtma-prakåßo hata.
IS.1674
310 Minoru Hara
ABBREVIATIONS
A A¤guttara Nikåya (PTS)
BR O. Böhtlingk and R. Roth, Sanskrit Wörterbuch
D D¥gha-nikåya (PTS)
H The Harivaßa (Poona Critical Edition, unless otherwise indicated)
IS O. Böhtlingk, Indische Sprüche (Osnabrück Reprint 1966)
Ja The Jåtaka, ed., by V. Fausbøll (PTS)
KAS The Kau†alya Artha-Íåstra, ed., by R.P. Kangle (Bombay)
MBh The Mahåbhårata (Poona Critical Edition)
Mil Milinda-pañha, ed., by V. Trenckner (PTS)
M Majjhima-nikåya (PTS)
MS Manusm¤ti (NSP)
NSP Nirnaya-sagar Press (Bombay)
PTS The Pali Text Society
R The Vålm¥ki Råmåyaâ (Baroda Critical Edition, unless otherwise
indicated
S Sayutta-nikåya (PTS)
Sn Suttanipåta
Vin The Vinaya-pi†aka, ed. by H. Oldenberg (PTS)
YS Yåjñavalkya-sm¤ti (NSP)
BIBLIOGRAPHY
Allon, M., 1997. Style and Function´: A Study of the Dominant Stylistic
Features of the Prose Portions of Påli Canonical Sutta Texts and Their
Mnemonic Function, Tokyo
Botto, O., 1962. N¥tivåkyåm¤ta Somadeva SËri, Torino
Dhadphale, M.G., 1980. Synonymic Collocations in the Tipi†aka´: A Study,
Poona
Edgerton, F. 1933. “jñåna and vijñåna”, Festschrift M. Winternitz, Leipzig,
pp.´217–20
Hacker, P., 1978. Kleine Schriften, Wiesbaden, pp. 137–66
Hara, M., 1968. “A Note on the Sanskrit Word jana”, Festschrift F.J.B. Kuiper,
The Hague, pp. 256–69
——— 1992. “Íraddhå in the Sense of Desire”, Festschrift J. May, Asiatische
Studien 46/1, pp. 180–94
——— 1996. “A Note on the Epic-Phrase j¥van-mukta”, Adyar Library Bulletin
60, pp.181–97
——— 2000 A . “såra, asåra, sasåra”, Festschrift M.A. Mehendale,
Ahmedabad, pp. 139–59
——— 2000B. “Two Notes on the Word upani¡ad in the Mahåbhårata”, Studia
Indologiczne 7 (S. Schayer Volume), Warszawa, pp. 157–69
——— 2000C. “A Note on the Compound krodha-mËrchita”, Festschrift A.
Mette (Swisttal-Ordendorf 2000), pp. 343–57
——— 2002. “The Hindu Concept of Friendship, A Note on Sanskrit
praâya”, Rivista degli Studi Orientali 75 (2001), pp. 157–87
——— 2004. “ßåstra versus ßastra”, Gedenkschrift J.W. de Jong, Tokyo,
pp.´49–64
Kane, P.V., 1973. History of Dharma-ßåstra III, 2nd ed., Poona
Meyer, J.J., 1977. Das altindische Buch vom Welt- und Staatsleben, Graz
Reprint
Norman, K.R., 1993. “Theravåda Buddhism and Brahmanical Hinduism
(Brahmanical Terms in a Buddhist Guise)” in Collected Papers, Vol.
IV, Oxford, pp. 271–80
Schopen, G., 1975. “The Phrase ‘sa p¤thiv¥pradeßaß caityabhËto bhavet’ in the
Vajracchedikå´: Notes on the Cult of the Book in Mahåyåna”, Indo-
Iranian Journal 17, pp. 147–81
Zombies and Half-Zombies : ´
MahåsËtras and Other Protective Measures*
My work on the MahåsËtras, a set of MËlasarvåstivådin texts preserved

in Tibetan translation, was supported by the Pali Text Society during the
presidency of Mr K.R. Norman. As a result, the Society published the
rst two volumes of MahåsËtras : Great Discourses of the Buddha in
´
the series Sacred Books of the Buddhists.1 It is therefore with great

pleasure that I present further research pertaining to the MahåsËtras in
this volume dedicated to Mr Norman.
The Vinayavibha¤ga is a section of the MËlasarvåstivådin monastic
code, the Vinaya.2 Lost in the original Sanskrit, it is preserved in
Tibetan and Chinese translations. An important primary document for
the study of northern Indian Buddhism, it has not, so far, been edited,
systematically studied, or translated into any European language. In my
study of the MahåsËtras, I used a passage from the Vinayavibha¤ga for
two purposes : as an example of a MËlasarvåstivådin list of MahåsËtras,3
´
and as supporting evidence that the MahåsËtras were recited as

protective or rak¡å texts.4 The Vinayavibha¤ga is “supporting evidence”
*I am grateful to Shayne Clarke, Jan Nattier, and Mark Allon for their close
readings of this paper and for their valuable comments and corrections.
1 MahåsËtras ´: Great Discourses of the Buddha, Volume I, Texts ´: Critical
Editions of the Tibetan MahåsËtras with Påli and Sanskrit Counterparts as
Available, Sacred Books of the Buddhists XLIV, 1994 (reviewed by Helmut
Eimer, Zentralasiatische Studien 26, 1996, 235–39´; by J. W. de Jong, Indo-
Iranian Journal 40.3, July, 1997, 271–73)´; Volume II, Parts I & II, Sacred
Books of the Buddhists XLVI, 1997. Volume II, Parts 3 & 4, and Volume III
(translations), remain in a state of suspended animation. At the moment it is
impossible to determine which will come rst ´: the publication of the
remaining volumes of MahåsËtras or the end of the present æon.
2For a survey of this voluminous collection see Clarke 2002.
3 For MahåsËtra lists, see MahåsËtras II, Parts I & II, 3–61. Earlier studies
include Hôbôgirin I and Sasaki 1985.
4For the rak¡å status of the MahåsËtras, see MahåsËtras II, Parts I & II, 63–88
314 Peter Skilling
because several of the texts number among the great apotropaic classics
of early Buddhism — notably the Dhvajågra, the Ó†ånå†¥ya-, and the
Mahåsamåja-sËtras.
We still know very little about how the MahåsËtras were actually
used as a set, or to what degree the rituals may have corresponded to or
differed from the Paritta recitations of Sri Lanka and South-East Asia or
the Rak¡å rituals of Nepal. Certainly, several of the MahåsËtras have
parallels in the Paritta, and certainly, protection through recitation and
ritual was — and continues to be — one of the main functions or even
duties of Buddhist monastics.
An inscription on the “pedestal of a bronze image of the Buddha in
the bhËmisparßa-mudrå” from Bhagalpur District, Bihar, mentions
“MahåsËtradhåra Vahåkåya”, in “characters of about the twelfth
century”.5 Is this to be taken as published, with long “a” in “-dhåra”,
meaning architect, or perhaps stage-manager ? The reading remains to
´´
be conrmed. Even if the reading is correct, could “MahåsËtradhåra” be

an engraver’s error for “MahåsËtradhara”, with short “a” in “-dhara” ? ´
If that is the case, how do we read the compound ? Was Vahåkåya a

´
textual specialist, a “great SËtradhara”, a master of the SËtra literature,

or was he a ritual specialist, an “expert in or master of the MahåsËtras” ? ´
Could “MahåsËtradhara” be a title, a rank, for a “master of the Mahå-

sËtras” ? “SËtradhara” is a technical term of some antiquity, used widely
´
by all traditions, and attested in epigraphy, while “mahåsËtradhara” is

unattested in text or epigraphy. Perhaps the full inscription will help
determine the context ; at present the record is ambiguous, and it is
´
impossible to decide whether or not the inscription has any bearing on

the MahåsËtras.
In this paper, I give an English translation of the Vinayavibha¤ga
passage, extracted from the commentary on the third påråjika, followed
by editions of the Tibetan from the Vinayavibha¤ga, supplemented by
and Skilling 1992A, 125–29.

5Srinivasan 1986, p. 34. As far as I know the inscription has not been edited and
no photograph or rubbing has been published.
Zombies and Half-Zombies 315
its commentary, the Vinayavibha¤ga-padavyåkhyåna. Both texts were
translated by Jinamitra and Lui Gyaltsan (Klu’i rgyal mtshan), two of
the leading translators during the “rst diffusion” of Buddhism in the
Land of Snows, circa 800 C.E.6 Given the extraordinary prociency of
the two translators — and their teams, since they undoubtedly headed
translation committees — the passages are clearly and consistently
rendered. The author of the Vinayavibha¤ga-padavyåkhyåna is
Vin¥tadeva, about whom very little is known. He seems to have lived
and worked in northern India in the eighth century.
The passage itself is macabre. It concerns a monk who raises a
corpse — a vetå a or zombie — and orders it to kill someone, sending
it on its way in a two-wheeled cart, with two bells round its neck and a
double-bladed sword in hand.7 The primary concern of our text is not
the ethics of the matter as such, but what sort of infringements of the
monastic rules might be involved.8 The protective measures against
such an eventuality are interesting in their own right. I have divided
them into three groups. Group A lists protections that belong to the
folklore of the time, and are not as such Buddhist. In Group B, the
protection comes from the presence of a powerful and meritorious being
— a Buddha, a cakravartin, or a bodhisattva. The idea of the protective
presence of the Buddha is certainly ancient. In the Soâdaˆ a-sutta, for
example, it is said that “in whatever village or town Samaâ Gotama
6 For what little we know about Jinamitra, who along with Xuanzang ranks
among the great translators of all time, see MahåsËtras II, Parts I & II, 115–
125.
7 For the spelling vetå a see Skilling 1992 A , 111 n. 4 ´; the Påli equivalent is
vetåla/vetå¬a. For vetå a see Hôbôgirin I 68–69, s.v. “Bidara”. The creature
has become well-known as a “vampire”, for example in Burton (tr.) 1893. But
the habits of the “vampire” of Burton’s “Baital-Pachisi” are quite different
from those of the vetå a of our text, which seem closer to those of the
“zombie”. We therefore choose to translate the term with “zombie”, a name of
African origin, rather than with “vampire”, a term of Slavic origin.
8There is nothing remarkable in this, since the Vinayas do not deal with ethics
as such — they are monastic codes.
316 Peter Skilling
stays, non-humans do not harm the people of that village or town”.9
Group B has close parallels in the Sag¥ti-paryåya, which modern
scholarship describes as one of the earliest texts of the Sarvåstivådin
Abhidharma ;10 it is possible that the passage is old, dating back to at
´
least the rst century B.C.E. Group C names texts which if recited will
offer protection — the Pråtimok¡a and the MahåsËtras.
If the frustrated zombie turns back on the instigator and kills him,
the monk incurs a heavy fault (sthËlåtyaya). I do not know whether
there are any other cases of posthumous penalties in the monastic codes,
but here we have at least one. At the end the text notes that the trans-
gressions are the same in the case of a “half-zombie” (ardha-vetå a).
This curious creature is similar to the common or garden-variety
zombie : but in its case the monk installs it in a one-wheeled cart, ties a
´
single bell around its neck, and places in its hand a single-bladed
sword.11 The Sanskrit term ardha-vetå a is conrmed in the Sagha-
bhedavastu of the MËlasarvåstivådin Vinayavastu from Gilgit, and it
also occurs in other sections of the Vinayavastu preserved in Tibetan
translation but no longer extant in Sanskrit.12 That is, the “half-zombie”
belongs to the necromantic bestiary of the MËlasarvåstivådins. To the
best of my knowledge there is no equivalent Påli term.
The narrative runs smoothly, and is a good example of the style of
at least certain sections of the MËlasarvåstivådin Vinaya. It appears that
for the redactors the didactic function of narrative was paramount : good
´
stories, to be recited at least to the monastics within the walls of the

Mahåvihåras, were used to communicate the monastic rules. This
9D I 116,14´: samaô khalu bho gotamo yasmi gåme vå nigame vå pa†ivasati

na tasmi gåme vå nigame vå amanusså manusse vihe†henti. See Skilling
1992A, pp.´110–111.
10To the references given in the notes may be added Stache-Rosen 1968, p. 111,
last paragraph of translation of Sag¥ti–paryåya.
11The half-zombie is not well-known, and it is comforting to think that at least
the readers of this journal will know what to do in the event — the unlikely
event, I dare say´! — that they encounter one.
12See below, footnote 34.
editorial goal has, perhaps, confused modern scholarship, which has
tended to read Buddhist texts through the dim spectacles of historicism.
Translation13
1. With the intention to kill a man, a woman, or a hermaphrodite, a
monk goes to a charnel ground (ßmaßåna) on the night of the fourteenth
day of the waning moon (k®¡âpak¡a),14 and looks for a corpse that has
not been harmed (ak¡ata) or damaged (akhaˆ a) by any creature, even
by one as tiny as an ant (pip¥likå).
2. Finding one, he rubs it with white chalk (makkola) ; having rubbed it
´
with white chalk, he bathes it in scented water (gandhodaka).15 Having

bathed it in scented water, he dresses it in new cloth, anoints its feet,16
and utters a spell (mantra) : when it gets to its feet and stretches, he
´
places it on a two-wheeled cart, ties two bells round its neck, and places
in its hand a double-bladed sword.
3. When it gets up, it grunts17 and asks, “Whom should I slay ?18 Whom
´
should I kill ? Whose life should I take ?” Then the monk says to the
´ ´
zombie (vetå a), “Do you know such and such a man, woman, or
13 I am grateful to Fritz Grohmann (Taipei) for explaining the Chinese, for

which see also Hôbôgirin I, 69.
14The commentary points out that this is the twenty-ninth day of the month.
15For the use of white chalk and scented water, see Schopen 2004´: 288 (trans-
lating from Vinayavibha¤ga, D ña 65a2–66a4) and his remarks on termino-
logical problems, pp. 291–92.
16de’I rka¤ pa gñis kya¤ skud par byed´: skud pa translates forms of the roots
rlip and rmrak¡´: see e.g. Negi 1993, 182 (skud pa), 279 (bskus). The Chinese
has here “besmears its feet with ghee” (Shayne Clarke).
17 This is a guess for what the commentary helpfully describes as “utters the
blag blag sound”. I do not know what zombies do in such circumstances.
18From the context, the verb gto¤ ba (= rmuc) in its various forms here (gta¤
bar bya) and in the following (tho¤ ßig, gto¤ bar byed) can only mean kill,
although I have not found this meaning in any Tibetan lexicons or, for rmuc,
in Buddhistic Sanskrit or Påli usage. Cf. Monier-Williams 1976, p. 820c,
rmuc “with pråˆån, to deprive of life, kill … with kalevaram, deham,
pråˆån, or j¥vitam, to quit the body or give up the ghost, i.e. to die”.
318 Peter Skilling
hermaphrodite ?” When it replies, “I do”, he says, “Slay him ! Kill him !
´ ´ ´
Take his life !” If the zombie slays, kills, or takes [that person’s] life,
´
then that monk is defeated.

4. If protective measures are taken, such as :19 ´
4.A. (1) at the door a garland of forest-owers is strung up,20

(2) a vase full [of water] is set out,
(3) a cow and calf of matching [colour] are tethered,21
(4) a sheep is tethered,22
(5) a mortar and pestle are set out,
(6) an indrak¥la is laid at the door,23
(7) or a re is kept burning,24
4.B. (8) if the Conqueror (Jina) is staying there,
(9) or one appointed by the Conqueror [is staying there],25
19The list of protections in the Chinese Bhik¡u and Bhik¡uˆ¥ Vinayavibha¤gas

is very close in items listed, order, and number. Cf. Hôbôgirin I, 69.
20 Commentary ´: “a garland made from owers and fruits that grow in the
forest”´; Chinese´: “medicinal herbs made into garlands”. Cf. Íik¡åsamuccaya
in Bendall 1992, p. 139,11, vanakusumåni, in the context of rak¡å.
21The Chinese has “a cow together with a calf of the same colour is tethered at
the door”. The Commentary has “the offspring, both, [have] the same hair-
colour”(?). “Garland” (målå), “vase full of water” (pËrâkumbha), and “cow”
are included in lists of ma¤gala, “auspicious things”´: see Karunaratne 1971,
p. 48. “Cow with calf” (savacchakadhenu) is one of the ma¤gala on the feet
of the Buddha´: see Karunaratne 1976, p. 60 (item 79). The “full pot” is
important in the Theravådin paritta ceremony´: see de Silva 1981, pp. 79–86.
Cf. the list of auspicious symbols connected with the Buddha in Skilling
1992B.
22Chinese´: “a ewe together with a lamb of the same colour is tethered”.
23 ’khor gtan = indrak¥la, Mahåvyutpatti (Sakaki 1926) § 5582 (an alternate
translation, sgo’i them pa, also given at Mahåvyutpatti § 5582, is used at
MahåsËtra I 10.A and B, § 1.3). The indrak¥la is important in the Theravådin
paritta ceremony´: see de Silva 1981, pp. 57–79.
24For the last three, the Chinese has “(5) or in the house is a stone for pounding
medicine together with a grinding stone´; (6) or at the door is an indrak¥la´; (7)
or a never-extinguished re”.
25The Commentary interprets the phrase “one appointed by the Conqueror” as
(10) if a Wheel-turning Emperor (cakravartin) [is staying
there],26
(11) or a Wheel-turning Emperor is entering his mother’s
womb,27
(12) if a bodhisattva [is staying there],28
(13) or a bodhisattva is in the process of entering his mother’s
womb,29
4.C. (14) if one is about to recite30 the Pråtimok¡a-sËtra,31
(15) or recites it out loud in full (vistareâ svareâ
referring to “a messenger (dËta) of the Lord” or “one specied (ådi¡†a) by the
Lord”. Cf. Koßabh 2 ´:45ab, (P) 75 ,3 , and D¥pa 103 ,8 ´: a messenger of the
Conqueror (jinadËta) or one appointed by the Conqueror ( jinådi¡†a) cannot
be killed by either self or another´; Koßabh 3´:85a, (P) 176,4´: a jinadËta and a
jinådi¡†a cannot die before their time. Cf. La Vallée Poussin 1971, I 220, nn.
1, 2.
26Cf. Koßabh 2´:45ab, (P) 75,6, D¥pa 103,9´: a cakravartin is not killed by either
self or another.
27Cf. Koßabh 2´:45ab, (P) 75,6, D¥pa 103,10´: a cakravartin’s mother is not killed
by either self or another´; Koßabh 3´:85a, (P) 176,5´: a cakravartin’s mother
cannot die before her time.
28The Commentary glosses “a bodhisattva in his last rebirth” (caramabhavika).
Cf. Koßabh 2 ´:45ab, (P) 75 ,5 , D¥pa 103 ,9 ´: a bodhisattva in his last rebirth
(caramabhavika) is not killed by either self or another´; Koßabh 3´:85a, (P)
176,4´: a bodhisattva in his last rebirth cannot die before his time.
29Cf. Koßabh 2´:45ab, (P) 75,5, D¥pa 103,9´: a bodhisattva’s mother is not killed
by either self or another´; Koßabh 3´:85a, (P) 176,5´: a bodhisattva’s mother
cannot die before her time.
30 I am not certain of the meaning of ma bton pa ’don par byed pa. The
Commentary has kha ton du ma bslabs pa’o´: does this mean “in the reading
of which one is not trained” or “silently”´? There seems to be some contrast
with the following “reads out loud in full”, which the commentary glosses as
“with a voice heard by others”. The Chinese seems to interpret the rst phrase
as “is going to recite”, “is about to recite”. For now I follow this inter-
pretation, with the idea that the power of the text is sufcient to drive away
zombies and other nuisance-makers even when it is about to be read.
31In the Antaga a Dasåo, Chapter 6, the Jain ascetic Sudasaê is protected
from a dangerous Jakkha by making “full profession of the monastic vows”
(Coomaraswamy [1928–29] 1980, Part I, pp. 21–22).
320 Peter Skilling
svådhyåya karoti),32
(16) if one is about to recite any of the four classes of sËtras
(caturˆå sËtranikåyånå anyatamånyatama
sËtranikåya),33
(17) or recites them out loud in full (vistareâ svareâ
svådhyåya karoti),
(18) if one is about to recite the great and lofty sËtras : ´
1. CË aßËnyatå
2. MahåßËnyatå
3. Pañcatraya
4. Måyåjåla
5. Bimbisårapratyudgamana
6. Dhvajågra
7. Ó†ånå†¥ya
8. Mahåsamåja
(19) or recites them out loud in full (vistareâ svareâ
svådhyåya karoti) : ´
5. and, because of his failure [to kill his victim], the zombie decides to
kill the monk instead : if the zombie kills the monk, the monk incurs a
´
heavy fault (sthËlåtyaya).

6. If the monk kills the zombie, the monk incurs two heavy faults : the ´
rst from killing the zombie, the second from the previous stratagem
(pËrva-prayoga).
7. As for a zombie, so for a half-zombie (ardha-vetå a),34 but between a
32 For the Sanskrit see Carmavastu, in Gilgit Manuscripts (Dutt 1984) III–4
188,10.
33The commentary interprets the “four classes of sËtras” as “the D¥rghågama,
Madhyamågama, Sayuktågama, and Ekottarikågama”´: see MahåsËtras II,
20–22.
34 Cf. Sa¤ghabhedavastu (Gnoli 1978 238 ,24 ) vetå årdhavetå a (Tibetan in
Gnoli’s note h as here)´; Bhai¡ajyavastu (Tib.) ge 68a6´; Vinayak¡udraka (Tib.)
ne 200b3, Tib. idem. The “de nition” of vetå årdhavetå a in the
Ógamak¡udraka-vyåkhyåna, £u 197b4, resembles that of our text´: ro la¤s ni
zombie and a half-zombie there are these differences : [the monk] places
´
it in a one-wheeled cart ; he ties a single bell around its neck ; he places

´ ´
in its hand a single-bladed sword. This is a half-zombie. The establish-

ment of transgression (åpatti) should be described as entirely the same
as the preceding.
Peter Skilling
École française d’Extrême-Orient
Bangkok, Thailand
ABBREVIATIONS
D¥pa´: see Jaini 1977.
Koßabh (P)´: see Pradhan 1975
BIBLIOGRAPHY
Bendall, Cecil, ed., 1992. Çikshåsamuccaya´: A Compendium of Buddhistic
Teaching Compiled by Çåntideva Chiey from Earlier Mahåyåna-
sËtras, St. Petersburg 1897–1902 (Bibliotheca Buddhica I), reprinted
Delhi´: Motilal Banarsidass Publishers Pvt. Ltd., 1992
Burton, Richard F., tr., 1893. Vikram and the Vampire´: Tales of Hindu Devilry,
reprint 1989, Singapore´: Graham Brash
Clarke, Shayne, 2 0 0 2 . “The MËlasarvåstivådin Vinaya´: A Brief
Reconnaissance Report” in Early Buddhism and Abhidharma Thought
in Honor of Doctor Hajime Sakurabe on His Seventy-Seventh
Birthday, Kyoto´: Heirakuji Shoten, pp. 45–63
Coomaraswamy, Ananda K., 1980. Yak¡as. Part I originally published
Washington, 1928´; Part II originally published Washington, 1929,
reprinted in one volume, New Delhi´: Munshiram Manoharlal
Publishers Pvt. Ltd., 1980
Dutt, Nalinaksha, ed., 1984. Gilgit Manuscripts. Originally published Srinagar,
’khor lo gñis da¤ ldan pa’i ßi¤ rta la £on pa lag pa gñis kyis ral gri so sor
’dzin pa’o. ro la¤s (read ro la¤s phyed´?) ’khor lo gcig pa’i ßi¤ rta la £on ci¤
ral gri gcig ’dzin pas ñe bar mtshon pa’o.
322 Peter Skilling
1939–59, reprinted Delhi´: Sri Satguru Publications, 1984
Eimer, Helmut, ed., 1983. Rab tu ’byu¤ ba’i g£i´: Die tibetische Übersetzung des
Pravrajyåvastu im Vinaya der MËlasarvåstivådins, 2 vols., Wies-
baden´: Otto Harrassowitz
Gnoli, Raniero, ed., with the assistance of T. Venkatacharya, 1978. The Gilgit
Manuscript of the Sa¤ghabhedavastu. Part II, Roma´: Istituto italiano
per il Medio ed Estremo Oriente (Serie Orientale Rome XLIX, 2)
Hôbôgirin´: Dictionnaire encyclopédique du bouddhisme d’après les sources
chinoises et japonaises. Premier fascicule´: A–Bombai, Tôkyô´:
Maison Franco-japonaise, 1929–1930
Jaini, Padmanabh S., ed., 1977. Abhidharmad¥pa with Vibhå¡åprabhåv®tti,
Patna´: Kashi Prasad Jayaswal Research Institute (Tibetan Sanskrit
Works Series IV)
Karunaratne, T.B., 1971. “A¡†ama¤gala, the Eight Auspicious Symbols.”
Journal of the Ceylon Branch of the Asiatic Society, New Series, Vol.
XV, pp. 48–70
——— 1976. “The Signicance of the Signs and Symbols on the Footprints of
the Buddha”, Journal of the Royal Asiatic Society (Sri Lanka), Vol.
XX (New Series), pp. 47–60
La Vallée Poussin, Louis de, tr., 1971 L’Abhidharmakoßa de Vasubandhu, 6
vols., originally published Paris 1923–31, reprinted Brussels´: Institut
belge des Hautes Études Chinoises, 1971
Monier-Williams, Monier, 1976. A Sanskrit–English Dictionary, originally
published Oxford, 1899, reprinted Delhi, 1976
Negi, J.S., 1993. Tibetan–Sanskrit Dictionary, Vol. 1, Sarnath, Varanasi´: Dict-
ionary Unit, Central Institute of Higher Tibetan Studies
Pradhan, P., ed., 1975. Abhidharmakoßabhå¡yam of Vasubandhu, rev. 2nd ed.,
Patna´: K.P. Jayaswal Research Institute (Tibetan Sanskrit Works
Series VIII)
Sakaki, R., ed., 1926. Mahåvyutpatti, 2 vols., KyØto
Sasaki S., 1985. “The MahåsËtra of the MËlasarvåstivåda as listed in the lDan
dkar ma Catalogue”, Buddhist Studies (BukkyØ KenkyË), Vol. XV
(Dec. 1985) [in Japanese]
Schopen, Gregory, 2004. “Ritual Rights and Bones of Contention´: More on
Monastic Funerals and Relics in the MËlasarvåstivåda-vinaya” in
Gregory Schopen, Buddhist Monks and Business Matters´: Still More
Papers on Monastic Buddhism in India, Honolulu´: University of
Hawai’i Press, pp. 285–328. Originally published in Journal of Indian
Philosophy 22 (1994), pp. 31–80
de Silva, Lily, 1981. “Paritta´: A Historical and Religious Study of the Buddhist
Ceremony for Peace and Prosperity in Sri Lanka”, Spolia Zeylanica,
Vol. 36, Part I
Skilling, Peter, 1992A. “The Rak¡å Literature of the Íråvakayåna”, JPTS XVI,
pp. 109–82
——— 1992B. “Symbols on the Body, Feet, and Hands of a Buddha´: Part I.
Lists”, Journal of the Siam Society 80.2, pp.´67–79
——— 1997. “From bKa’ bstan bcos to bKa’ ’gyur and bsTan ’gyur.” In
Helmut Eimer, ed., Transmission of the Tibetan Canon´: Papers
Presented at a Panel of the Seventh Seminar of the International
Association for Tibetan Studies, Graz 1995, Vienna, pp. 87–111 (=
Vol. III of Ernst Steinkellner, general editor, Proceedings of the
Seventh Seminar of the International Association for Tibetan Studies,
Graz 1995´: Beiträge zur Kultur- und Geistesgeschichte Asiens,
No.´22, Österreichische Akademie der Wissenschaften, Philosophisch-
Historische Klasse Denkschriften, 257. Band)
Srinivasan, P.R., ed., 1986, Annual Report on Indian Epigraphy for 1975–76,
issued by Dr K.V. Ramesh, New Delhi´: Archæological Survey of
India
Stache-Rosen, Valentina, ed., tr., 1968. Dogmatische Begriffsreihen im älteren
Buddhismus II, Das Sa¤g¥tisËtra und sein Kommentar Sa¤g¥tiparyåya,
Parts 1 and 2, Berlin´: Akademie-Verlag (Sanskrittexte aus den
Turfanfunden IX)
324 Peter Skilling
OD=@O<I[email protected]
ÌO`so<Í^jhk\m`noìF\iepmq`mndjinja\i`s^`mkoamjhoc`Qdi\t\qd]c\´b\
jioc`np]e`ôjaqò°\)<ggq\md\ion'di^gp_dib^jiom\ôdjin\i_oc`pn`jaoc`
¡\_'\m`m`^jm_`_)Hdijmq\md\ion'hjnodiqjgqdiboc`¡\_#`nk`^d\ggt\ao`m_\´'
Ï\h'jm^d´$'\m`gdno`_n`k\m\o`gt\ooc`ì_]tk\m\bm\kciph]`m\i_mà`mì^`
gòo`m)Mò\di`_\nijo`n\m`bìpdi`q\md\ion\i_q\md\ionoc\oncjroc`h\di
gdi`nja\agd\odjijaoc`F\iepmn^jinpgo`_´;np](bmjpknnp^c\n=L'>E'jmGI
\m`kg\^`_rdococ`hdijmq\md\ion)Oc`k\oo`mija\agd\odji \bm``n rdoc oc\o
ncjriajmoc`Km\qm\etq\nopdi@dh`m,43.´:GINOm`km`nìoocÒc`hnk\ibn
h\gdi`\b`'\i_=>?ELocÒnc\gk\gdi`\b`)<nrdococ`Km\qm\etq\nop'oc`m`
\m` ij h\ejm m`^ìndji\g _daa`mì^`n #ja oc` otk` n`ì di' ajm `s\hkg`' oc`
?mph\fdii\m\me(k\mdk¢^^cjmoc`N\¯_cdidmhj^\i$nâom\$]òr`ìocòrj
gdi`\b`n)
ÌO`so =Í dn \i `s^`mko amjh Qdi®o\_`q\Ïn ^jhhìo\mt ji oc` Qdi\t\(
qd]c\´b$Do^jhk\m`nqÒ\iepmn'ijodib\ggq\md\ion)Oc`bì`m\g\agd\odjidn
>?\b\dinoBIL)Oc`mjjoo`sodnbdqìdi]jg_(a\^òtk`)
<)@som\ôamjhHâg\n\mqnodq_diQdi\t\qd]c\´b\
Ï?pg]\mi\hk\mÏ]t`_k\#Qdi\t\qd]c\´b\$'om\ing\o`_]tEdi\hdom\\i_FgpÏd
mbt\ghonc\i'.m_kmedf\#,+oc]\hkj$´:
=1)- Ï_pg]\ ^c\ ,4,], VQjg)1X
>,+.. Ï_pg]\ ^\ ,04\. VQjg)44X
?. Ï_pg]\ ^\ ,/-]. VQjg)0X.1
C. Ï_pg]\ ^\ -++\, VQjg)0X
E. Ï_pg]\ ^\ ,/.]. VQjg)0X
G Ï_pg]\ f\ ,2.]/ VQjg),X
I. Ï_pg]\ ^\ -,-\, VQjg)0X
L,+.- Ï_pg]\ ^c` ,-3]0 VQjg)43X.2
N. Ï_pg]\ ^\ -+0\. VQjg)0X
O, Ï_pg]\ f\ ,03\1 VQjg),X
,)_b`ngj´bdnnft`nk\_\´w#\$]p_h`__\´w#]$h\id´g\bn\_k\Ïdn`hnftdn
.0Ndbg\'\]]m`qd\odjin'^jiqìodjin\i_`_dojmd\gkmdi^dkg`n\m`\ndiH\cnâom\n
D)
.1=`mf`g`tm`kmdioQjg)-'k)--.).).)
.2Jo\idm`kmdioQjg)/-'k),4.)/)0)
w.3ug\]\h\mbtd´jÏdonc`n#^$]^p]³dg\#_$_pmfcmj_#`$_pnjńo`w.4¡d]\Ïd
mj#a$oc\i\nmjb^c\bnbmjbn]pm/+kcm\#b$hjnft\´h\nh\n#c$h\\hn#d$k\
oncjg]\m]t`_^d´vex
-)m`_i\nn\f\m#\$btdn_mdgkctd]t`_^d´wn\f\m#]$btdn_mdg#^$kctd]t\ni\n
nkjn^cpnÏfcmp]\m]t`_g\w#_$nkjn^cpn/,]fmpni\nm\nn\mk\nfjik\m]t`_
^d´w_`Ïdmf\´k\#`$bdnft\ńfp_k\m]t`_wbn\ń´\bnft\´mej_k\m]t`_
k\i\w/-_`g\ńk\m#O,03]$Ïbtpm/.o`]t\nht\ń#a$]t`_k\Ïdonc`w//_`n
_`#b$Ïfcjmgjbdn_\´g_\ik\Ïd¡d´mo\g\Ïejbk\m]t`_^d´w_`Ïdhbpg_p#c$
_mdg]p#d$bdnft\´Ï_jbnk\m]t`_wg\bk\mm\gbmd/0njbdnk\t\ń]tdik\m
]t`_k\i\w/1
.)_`g\ńo`#\$]g\b]g\b#]$u`m³d´#^$Ï_dnf\_^`n´\nb\´bo\´]\m/2]t\w´\nb\´
bn\_#_$k\m]t\w´\nb\ńmjb#N-+0]$_\´]m\g#`$]\m]t\³`nu`m]\Ïd#G,2/\$
onc`w/3_b`ngj´_`nmjg\ń_`g\Ï_dnf\_^`nfctj_ftdn#a$nft`nk\_\´w#b$
]p_h`_#L,-4\$_\´w#c$h\id´^c`b`hj³db¡`nn\hvdx³`n#e$nhm\nk\i\w
¡`nnj#f$³`nu`m]\Ïdonc`w/4_òcj´¡db0+w#g$_`nj_^dbw_`nmjb_\´]m\g]\m
#=,4-\$btdn¡db^`nnhm\ni\nw#h$#>,04]$b\g#i$o`mjg\ń_`n_`boj´]\m
]t`__\hw#?,/.\$bnj_k\m]t`__\hwnmjb_\´]m\g]\m]t`_i\w0,_b`
#I-,-]$ngj´kc\hk\mÏbtpmmjww#j$
/)b\go`_ì\nmp´]\m]t`_k\Ï_dgo\]pÏ_dgo\no`w0-
#,$ nbjmi\bnftdkcm`´]\]o\bnk\Ï\hw
.3wGINO´:=>?CELjh)w)
.4wCGINO´:=>?ELjh)w)
/+bmjbn]pm>?ELN´:bmjb]pm=CIO´:]pm#jigt$G)
/,^cpn=>?CELN´:^cpGIO)
/-wCGINO´:=>?ELjh)w)
/.Ïbtpm=>?EL´:btpmCGINO)
//wCGINO´:=>?ELjh)w)
/0m\gbmd=>?CEL ´:m\gbmdmG´:m\gbtdmIO ´:m\gbmdÏdN)Ajmoc`nk`ggdibm\g
btddiOpicp\ibHnn'n``Nfdggdib,44-='k)24'i)4/)
/1wCGINO´:=>?ELjh)w)
/2bo\´]\m>?ENO´:bo\´k\mL´:]o\´]\m=CGI)
0+¡db=>?EL´:³dbCGINO)
0,wCGINO´:=>?ELjh)w)
0-w=>?ELN´:CGIOjh)w)
326 Peter Skilling
#-$ ]phk\#\$b\´#C-++]$]\#E,//\$]³\bk\#]$Ï\hw
#.$ ]\_\´#^$]`ÏpÏ_m\]\]o\bn#_$k\Ï\hw
#/$ gpb]o\bnk\Ï\hw
#0$ h^cdbh^cdbbp#`$_\´]^\nk\]³\b#a$k\Ï\hw
#1$ nbjmÏfcjmbo\i]n\g0.]\Ï\hw
#2$ h`#b$oncpbnn]\m]\Ï\hw
#3$ mbt\g]\]³pbnk\Ï\hw
#4$ mbt\g]\n]f\Ïnon\gk\Ï\hw
#,+$ Ïfcjmgjnnbtpm]\Ï\hw
#,,$ Ïfcjmgjnnbtpm]\Ïdh\Ïdh´\g_pÏfcjmgjn#c$nbtpm]\³pbnk\Ï\hw
#,-$ ]t\´^cp]n`hn_k\ÏÏ\hw
#,.$ ]t\´^cp]n`hn_k\Ïd#d$h\Ïd#e$h´\g_p]t\´^cp]n`hn_k\Ï³pbnk\
Ï_pbk\Ï\hw#f$
#,/$ njnjmoc\mk\Ïdh_jh\]ojik\0/Ï_jik\m]t`_k\#g$Ï\hw
#,0$ ]ojik\00mbt\^c`m_]t\ńftdnfc\oji]t`_k\Ï\hw
#,1$ #O,04\$h_jn_`Ïdn_ònc\i]³dkjb\´t\´01mp´]\h\]oji02k\Ï_ji
k\m]t`_k\Ï\h#h$w
#,2$ ]ojik\mbt\^c`m_]t\ńftdn#i$fc\oji]t`_k\Ï\hw
#,3$ h_j#N-+1\$^cìkj^c`]\Ï_dgo\no`w^cp´´pnoj´k\d__\´w^cì
kjnoj´k\d__\´wg´\bnphk\_\´wnbtph\Ïd_m\]\_\´wbupbn^\i
nd´kjn]np]\_\´w#G,2/]$mbt\ghonc\i_\hk\_\´#j$wfpiop#k$
mbtp ]\ _\´ w#l$ fpi op#m$hdmbtp]\_\´ vnx Ïocpi03 k\Ïd h_j _\´ w
#=,4-]$Ï_pnk\^cìkjÏdh_j_`_\bh\]oji04k\Ï_jik\m]t`_k\
_\´w
#,4$ ]ojik\_\bmbt\^c`m_]t\ńftdnfc\oji]t`_k\#o$]³\bk\mbtpmi\w
0)bo\´`nk\tdik\nmjg\ń#I-,.\$ftdn_b`ngj´]_\bd_bnj_k\mn`hn
k\m1+Ïbtpm]\n1,wb\go`mjg\ńftdn_b`ngj´bnj_#\$k\m]t`_i\w1-_b`ngj´
0.]n\g=>?CEL´:n\gGINO)
0/k\=>?EL´:k\mCGINO)N``]`gjr'do`h#,1$)
00k\>?CEGINO´:g\=L)
01t\´=>?ELN´:CGIOjh)t\´)
02]oji>?CEGINO´:boji=L)
03Ïocpi>EGIO´:hocpi=?CLNO)
04]oji>?CEGINO´:boji=L)
1+k\m=>?ELN´:k\CGIO)
1,Ïbtpm]\n=?ELN´:btpmk\nCGIO)
1-wCGINO´:=>?ELjh)w)
g\#]$`nk\n]jhkjmÏbtpmmjww#^$
1)b\go`_b`ngj´bdn#\$mjg\ńbnj_k\m]t`_i\w1._b`ngj´g\#]$`nk\n]jh
kj#^$bdnnp#_$Ïbtpmo`wb^dbidmjg\ń]n\_#L,-4]$k\g\nnjww#`$bdnk\id
n]tjm]\n´\h\_`d_g\nnjww#a$
2)mjg\ń#>,1+\$g\edgo\]\#C-+,\$]³di_pmjg\ńkct`_g\1/t\´_`]³dio`w
mjg\ń_\´ v10xmjg\ńkct`_g\]t`]m\bidÏ_dtj__`wÏfcjmgjb^db_\´g_\i
k\Ïd¡d´mo\#\$g\Ïejbk\_\´whbpg_p_mdg]pb^dbÏ_jbnk\_\´wg\bk\mm\gbmd
nj#?,/.]$b^dbk\n]tdik\idmjg\ńkct`_^`n]t\Ïjww#]$gop´]\mi\hk\m
b³\bk\idoc\hn^\_n´\h\]³di_p]mej_k\m]t\Ïjww#^$
Hdijmq\md\ionoj@som\ôÌ<Í
,#\$)w=>?EGILN´:COjh)w)
,#]$)w=>?ELN´:CGIOjh)w)
,#^$)onc`n>?EGILNO´:onc`=)
,#_$)O\__ng\]`gjrgdi`)
,#`$)_pmfcmj_>?EGILNO´:_pfcmj_=)
,#a$)mj>?EGILNO´:ij=)
,#b$)kcm\>?EGILNO´:kcp=)
,#c$)nh\n>?EGILN´:nhm\n=O)
,#d$)\hn>?EGILNO´:`hn=)
,#e$)N\__nw´:ijodi=>?CEGILO)
-#\$)i\nn\f\m>?ELN´:i\nn\_f\mC´:i\n\f\mIO´:i\nf\m=G)
-#]$)n\f\m=>?EGILNO´:n\_f\mC)
-#^$)btdn_mdg=>?EGILN´:btd_mdO)
-#_$)w=CLN´:>?EGIOjh)w)
-#`$)O\__nk\]`gjrgdi`)
-#a$)nht\ń=>?CEGILO´:mht\ńN)
-#b$)_`=>?CEGILO´:_`ÏdN)
-#c$)_p=>?CEGILO´:Njh)_p)
-#d$)_mdg]p>?EGILNO´:_md]p=)
.#\$)o`>?EGILNO´:=jh)o`)
.#]$)]g\b>?EGILNO´:bg\b=)
1.wNO´:=>?EGILjh)w)
1/g\>?CEGINO´:k\=L)
10=>?EL\__w´:ijodiCGINO)
328 Peter Skilling
.#^$)O\__n³d´]`gjrgdi`)
.#_$)bn\_=?CEGILNO´:bnj_>)
.#`$)]m\g>?EGILNO´:ug\g#$=)
.#a$)ftdn>?EGINO´:ftd=L)
.#b$)w=>?ELN´:CGIOjh)w)
.#c$)w=>?ELN´:CGIOjh)w)
.#d$)CGIO\__w´:ijodi=>?ELN)
.#e$)³`n>?EGILNO´:³d´=)
.#f$)¡`nnj=>?ELN8¡`njGI´:Ojh)¡`nnj)
.#g$)w=>?CELN´:GIOjh)w)
.#h$)w=?CEGILNO´:>#ì_jagdi`$jh)w)
.#i$)b\g>?EGILNO´:g\b=)
.#j$)Ïbtpmmjww=>?ELN´:ÏbtpmjwwGI´:ÏbtpmmjwO)
/#\$)]phk\=>?CELN´:]p´k\GI´:]p´]\O)
/#]$)k\=>?CEGINO´:Ljh)k\)
/#^$)_\´>?EGILNO´:g\´=)
/#_$)]\]o\bn=>?EGILN´:]\mo\bnO)
/#`$)O\__nbp]`gjrgdi`)
/#a$)]³\b>?EGILNO´:]³\]=)
/#b$)h`=?GILN´:hd>E´:h\O)
/#c$)gjn>?EGILNO´:gjm=)
/#d$)_k\Ïd=>CEGILNO´:_k\Ï?)
/#e$)h\Ïd>?EGILNO´:hdÏd=)
/#f$)w=?CEGILNO´:>jh)w)
/#g$)k\=>?CEGINO´:h\L)
/#h$)=\__n_doojbm\kcd^k\m]t`_k\Ï\hw)
/#i$)ftdn=>?CEILNO´:ftdG)
/#j$)_\´=>?CEGILO´:_\hN)
/#k$)fpiop>?CEGINO´:fpi_p=L)
/#l$)w?CGINO´:=>ELjh)w)
/#m$)fpiop>?CGINO´:fpi_p=EL)
/#n$)?CGINO\__w´:ijodi=>EL)
/#o$)k\=>?EGILN´:k\mO)
0#\$)bnj_=>?CEGILO´:]nj_N)
0#]$)g\>?C#E$GINO´:=Ljh)g\)
0#^$)Ïbtpmmjww=>?EIN´:ÏbtpmjwwG´:ÏbtpmmjwLO)
1#\$)O\__nbdn]`gjrgdi`)
1#]$)g\>?EGILNO´:fc\#$=)
1#^$)kj=?CEGILO´:kjm>N)
1#_$)bdnnp=>?CELNO´:bdnpGI)
1#`$)g\nnjww=>?CENO´:g\nnjwLO´:g\njwwGI)
1#a$)g\nnjww>?CEN´:g\nnjwLO´:g\njwwGI´:g\g#Ä$njww=)
2#\$)¡d´mo\?GILNO´:¡dó\=>#E$)
2#]$)]t\Ïjww>?CEGIN´:]t\ÏjwO´:]t\]\Ïjww=´:]t\]\ÏjwL)
2#^$)ww=>?EGIN´:wLO)
=)@som\ôamjhQdi\t\qd]c\´b\(k\_\qtfcti\jaQdi®o\_`q\11
Ï?pg ]\ mi\h k\m Ï]t`_ k\Ïd oncdb mi\h k\m ]¡\_ k\#Qdi\t\qd]c\´b\(
k\_\qtfcti\$jaÏ?pg]\Ïdgc\#Qdi®o\_`q\$'om\ing\o`_]tEdi\hdom\\i_FgpÏd
mbt\ghonc\i´:
>/+3, Ï_pg]\ oncp 2,]/Ê2-\0
?/,,/ Ï_pg]\ oncp 1-]-Ê1.\-
B mi\hÏ]t`_ qp 3/].Ê30\/
I.1+2 h_j qp 14\1Ê14]0
L01,1 Ï_pg]\ÏdÏbm`gk\ qp 2/\/Ê2/]0
,)ug\]\h\mbtd´jÏdonc`n]^p]³d³`n]t\]\idonc`nd¡p_bpÏjww¡d]\Ïdmj
³`n]t\]\idhdÏdmjÏjwwh\nh\n³`n]t\]\id]pbph\]tp´]\Ïjwwh\\hn
k\³`n]t\]\idt\ig\b_ph]pmh\btpmk\Ïjww
-)]t\mht\ń12³`n]t\]\idgpnbtì_pÏ^cp13]\Ïjww
.)]g\b]g\bu`m³d´³`n]t\]\idnbm\]g\b]g\bu`m³d´´jwwmjg\ń³`n]t\]\
idÏ_m`v14xg\njbnk\hdh\tdik\Ïjww
/) #,$ i\bnftdkcm`´]\³`n]t\]\idi\bnonc\gbtdi\í\nnft`nk\Ïdh`
ojb_\´Ï]m\n]pg\]t\nk\Ïdkcm`´]\Ïjww
#-$ ]phk\b\´]\³`n]t\]\id2+^cpn]f\´]\Ïjww
11<ggq\md\ion\m`m`^jm_`_)Oc`k\m\bm\kciph]`mn^jmm`nkji_ojocjn`jaoc`
mjjo(o`so#O`so<$)>do\odjinamjhoc`mjjo(o`sodioc`>jhhìo\mt\m`kg\^`_
di]jg_(a\^òtk`)
12mht\ńBIL´:mht\´>?)
13Ï^cp>?IL´:^cpB)
14L\__n]\´:ijodi>?BI)
2+]\idBIL´:]\Ïd>?)
330 Peter Skilling
#.$ ]\_\´]`ÏpmÏ_m\]\³`n]t\]\idnh\_bdnnkpb\hocpik\Ïjww
#0$ h^cdbh^cdbp_\´]^\nk\³`n]t\]\idh^cdbnh\_kcmpbnnpg_\i
k\Ïjww
#2$ hòncpbnn]\m]\³`n]t\]\idh`]p_k\Ïjww
#3$ mbt\g]\³`n]t\]\id_`]³dib¡`bnk\Ïjww
#4$ mbt\g]\n]f\Ïnon\gk\³`n]t\]\id]^jhg_\iÏ_\nftdkcj\Ï\hw
]^jhg_\iÏ_\nftdn]no\ik\Ïjww
#,-$ ]t\´^cp]n`hn_k\Ï³`n]t\]\idnmd_k\oc\h\k\Ïj2,ww
#,/$ njnjmoc\mk\³`n]t\]\id_\´kjmoc\mk\Ïjwwh\]oji2-k\³`n]t\
]\idfc\oji2._ph\]ng\]nk\Ïjwwmbt\^c`m_]t\ńftdn³`n]t\]\
idb³\ibtdnocjnk\Ïdnbm\nnjww
#,1$ h_jn_`Ïdn_ònc\i2/]³dkj³`n]t\]\idgp´md´kj_\´wgp´]\mh\
_\´wt\´_\bk\mg_\ik\Ïdgp´_\´wb^dbg\nÏkcmjn20k\Ïdgp´_\bbjww
#,3$ h_j ^cì kj³`n]t\]\idkc\nftdmbjg]\g\nmbt\g]\m]t`_k\Ïd
kctdm_\´wbij_n]tdib_pbk\g\njbnk\g\nmbt\g]\m]t`_k\Ïdkctdm
wÏ]m\n]p^cìkj³`n]t\]\Ïdoc\oncdbbjww^c`]\³`n]t\]\id_ji
u\]k\d_ftdn21_ji^c`]\Ïjww_`_\bft\´b\´³ì\h_j^cì22
^c`]\]mbt\__`wÏ_dgo\no`³`n]t\]\n_´jnnp]no\ik\_\btdiijww
_`_\bbdgp´md´kjg\njbnk\_\bbd23mbtp__pbojbnk\tdiijww
0)bo\´³`n]t\]\idbi\nh\tdik\mbo\´]\Ïjww_b`ngj´]_\bd_bnj_k\m
n`hnk\mÏbtpm]\n³`n]t\]\idmjg\ń_`n_b`ngjń´\bnk\m\´d_bnj_k\m
]t`__j³`n]t\]\Ïdoc\oncdbbjww
2,h\k\ÏjBIL´:h\Ïj>?)
2-h\]ojiBI´:h\bojiL´:fc\oji>?)
2.fc\ojiBIL´:fc\_ji>?)
2/onc\i>?BL´:honc\iI)
20Ïkcmjn>?´:ÏkcjnBIL)
21>?IL´:ftdB)
22BIL´:>?jh)
23>B?I´:_b\bL)
Three Påli Works Revisited
Since it came into existence 125 years ago, the Påli Text Society has
kept editing works belonging to the Tipi†aka in its narrow sense as well
as other texts of all kinds. Some of those, however, have perhaps not
attracted as much attention as they deserve, partly because the interests
of Påli scholars and the fashion of scholarship have not been in their
favour. My modest purpose here is to awaken two of these somnolent
works which are fully entitled to have corresponding entries in any
history of Påli literature : (1) the Buddhaghosuppatti (Bu-up) and (2) the
´
Pa†hamasambodhi (Pa†h). 1 Finally, I would like to collect some

preliminary information on a third work, this time unpublished, (3) the
Vidaddhamukhamaˆ aâ (Vid), 2 with the hope that the Påli Text
Society could include it on its agenda, thus contributing to fullling one
of the desiderata for further Påli studies mentioned by K.R. Norman : ´
“The biggest deciencies in Påli publications in the West, however, are

in editions of †¥kås and of Påli texts composed in South-East Asia”
(1994 : 13–14 = 1996 : 80–81).
´ ´
1. The Buddhaghosuppatti or Buddhaghosanidåna

As the author of the Visuddhimagga and the famous commentator of the
Tipi†aka, Buddhaghosa is a highly venerated gure in the Buddhist
world, especially in South-East Asia. The recent reprint by the Påli Text
Society (in 2001) of the edition and translation of the so-called Buddha-
ghosuppatti by James Gray, originally published in 1892, is an occasion
to have a new look at the way the Påli tradition at some point, in some
1Both of them were read in toto or in part with students during classes held at
the Ecole Pratique des Hautes Etudes (“Philologie moyen-indienne”) in 2001–
2002 and 2004–2005. Brief preliminary remarks are available in the annual
reports (Livret-Annuaire) and, for Bu-up, in Balbir 2001. — I am grateful to
Dr Peter Skilling who provided Thai editions of these two works and a
Burmese edition of the Jinåla¤kåra (Jinål, see below section 2).
2See below, n. 21.
332 Nalini Balbir
place, looked at Buddhaghosa. Although Gray does not stand among the
most famous representatives of Påli philology, he did some useful work
in the eld during the years he spent in Burma where he taught Påli in
schools and translated some works which were of current use among
Buddhists of Burma at the end of the nineteenth century.3 His interest in
biographical and hagiographical works was materialised by his edition
and translation of two works : the Jinåla¤kåra (see below, section 2) and
´
the Buddhaghosuppatti, for which he expressed his strong liking in

unambiguous terms : “The narrative is intensely interesting.… It reads
´
in fact like an Arthurian romance.… The story in its entirety will be

found highly diverting as well as instructive” (Bu-up 1892, 1981 : 2, 9).
´
But he was aware of the historical limitations of the work : “Facts of

´
historical value cover only a limited space on the comparatively broad

canvas of the narrative, and will probably add very little to what is
already known of Buddhaghosa. The story, however, brings the per-
sonality of that eminent man vividly before our minds and enforces a
greater interest in him than ever ; and if it does this only, it can be safely
´
said that it was not written in vain” (Bu-up 1892, 1981 : 9). Indeed,
´
given the interests of Buddhist studies of those times in chronology, this

brief work of thirty pages had less to bring than the Sinhalese vasas
since it does not mention any king’s name or any date which could be
cross-checked. But it certainly contributes to constructing “la légende de
Buddhaghosa”, to quote the title of Louis Finot’s stimulating article
(1921), and could well have been included among the sources studied
by the contributors to the volume Sacred Biography in the Buddhist
Traditions of South and Southeast Asia (ed. Schober 1997).
Like other works dealing with Buddhaghosa, the present account is
organized around his pivotal rôle in the transmission and renewal of the
scriptures and their original language. It starts with his birth on earth as
a reincarnation of a god sent by Sakka for the special mission of
“translating” the teachings from their original “Sinhalese”, which could
3See Balbir 2001, n. 2.

Three Påli Works Revisited 333
no longer be understood, into “the language of Magadha”.4 This
narrative frame, which has been clearly drawn along the lines of
Någasena’s career as sketched in the Milindapañha (see Finot 1921 : ´
113 ; Bu-up 1892, 1981 : 69 ; Law 1923 : 25–47), sets Buddhaghosa in

´ ´ ´ ´
the broad perspective of an avatåra. The biography extends “before the

Cradle and past the Grave”, in the way traditional Burmese biographies
do (at-htok-pat-ti, Houtman in Schober 1997 : 311) : before his advent
´ ´
on earth Buddhaghosa is a god, after his death on earth he will be

Maitreya’s disciple. In between, the eight brief chapters which recount
his life are meant to show that in order to become a great man one
requires more than intellectual qualities.
If several episodes are viewed from this perspective, they do not
appear to be secondary or simply entertaining. On the contrary, they
play a part in providing the biography with meaning. The emphasis laid
on Buddhaghosa’s childhood is a part of the plan meant to show that the
intellectual brightness which is his outstanding characteristic has always
been there. The utterance of a paradoxical assertion which cannot be
understood by the audience rst provokes derision, then a respect which
leads to an inversion of the ordinary social rôles when Buddhaghosa’s
father admits, “You are my father and I am like your son” (40,27ff.). In
the fth chapter, which could seem a diversion, Buddhaghosa is a
witness to a dispute which has come up between two ladies and takes
note of the insults they throw at each other. These notes will have a
determining rôle to play in solving the matter at a later stage. They are
one of the several cases in this biography where the written document
appears to be of importance as a reserved, discrete testimony which can
be used when the situation arises. Such episodes underline both the
lucidity and modesty of the teacher and his connection with trans-
mission in general. The question of learning and the use of languages is
also dealt with in narrative disguise : apart from the initial replacement
´
of Sinhalese by Påli, the competition between Påli and Sanskrit, and the
4On this question see Granoff 1991.

334 Nalini Balbir
status of Sanskrit in the context of Buddhism are salient : as the son of a
´
Brahman working as a royal chaplain, the young Buddhaghosa is

probably conversant with Sanskrit. After his conversion to Buddhism,
he seems to leave it in the background and follows a purely Buddhist
curriculum sketched out in the story through selected technical terms.
But the idea that no education could be complete without a knowledge
of Sanskrit seems to be stressed when Buddhaghosa has to prove that he
masters this language in front of monks who thought he was ignorant of
it by reciting a few Sanskrit stanzas (rather badly treated in the manu-
scripts : Bu-up 1892, 1981 : 72–73).
´ ´
The origin and diffusion of the text need further investigation.

Gray’s edition is based on four manuscripts in Burmese script, for
which no details are given. On the other hand, on the basis of the
ascription of the text to a Thera Mahåma¤gala found at the end, Gray
was of the opinion that the text could have come from Ceylon or could
be dated “to the thirteenth century as the period when the Pi†akas and
their commentaries were taken to Ceylon from Burma” (Bu-up 1892,
1981 : 33). A little more can be said now that more documentation is
´
available. First, no manuscript seems to have emerged from Ceylon,

whereas a rather large number are to be found in South-East Asian
collections, whether they are kept in Burma, Thailand, or the West.5
The title Buddhaghosanidåna is largely prevalent over the title
°uppatti (which, anyway, is a sort of synonym, and could remind us of
the term used in Burma). The text is often provided with a nissaya and
was sometimes equated to a Jåtaka (in the broad acceptation of the
term). References to it or summaries are met with in late Påli historical
texts written in Burma (Jinakålamål¥, Gandhavasa, Såsanavasa),
where it seems to have become the standard for other works on
Buddhaghosa.6 As for Thailand, the Påli scholar Sammot Amarabandhu
(1860–1915), who wrote an introduction in Thai to an edition of the Påli
5SeeBalbir 2001, n. 28 for further details.

6See,respectively, PTS ed., p. 71, line 17´; Minayeff 1886, pp. 65 and 75´; B.C.
Law’s translation (London, 1923), pp. 32–33 (text, pp. 29–30).
text and its Thai translation, mentions the fact that people liked very
much to have sermons on this subject, which was called Thet Phra
Phuttakhosa, and that it was considered to convey benets (ånisasa).
He also indicates that the verses, which are interspersed at several
places in the text, are borrowed from the Vasamålin¥, that the style of
composition does not resemble older texts and that it is likely that the
work was composed in Thailand, Laos, or Burma, since it is not in the
style of old texts from Lanka.
In Burma, the Påli Bu-up has served as a basis for retellings in
Burmese.7 I would also suggest that Bu-up seems to exhibit a combina-
tion of canonical and local elements. Some technical details seem to be
in tune with what is known otherwise from South-East Asian traditions
and practices, although any interpretation should be done with great
caution. The narrative of Ghosa’s conversion to Buddhism, which could
be inspired by the Milindapañha, is not a mere reproduction of it. The
ordination ritual is different from canonical narratives as well. After the
rst stage, the removal of hair and beard expressed through the well-
known formula kesa-massu ohåretvå, the next one is to “take off the
layman’s smell through moist sandal powder” (alla-candana-cuˆêhi
gihi-gandha jhåpetvå, 44 ,27 ). This feature does not seem to be
mentioned before the commentary on the Vinaya and the texts based on
it. More relevant, the candidate wears white clothes and receives as the
ordination formula the ve topics of meditation (pañca kamma††håna),
i.e. the list of the rst ve body elements (keså, lomå, nakhå, dantå,
taco, 44,26ff.). This process recalls the traditions of the pabbbajjå as they
have been observed in South-East Asia.
Bu-up is a vivid example of the way religious instruction is
provided. Conversion of both Ghosa’s father (chapter 3) and Ghosa
himself is achieved through a teaching with practical or immediate
purpose, not through any elaborate doctrinal discussion, and the
Buddha’s teaching is called a manta (43,5, 44,4 ff.). Strikingly, the iti pi

7See Braun and Myint 1985, No. 222.

336 Nalini Balbir
so formula is one of the media in use.8 Finally, the way Buddhaghosa as
an enthusiastic new convert proceeds with his father, putting him for a
time in a gabbhaku†¥ (46,24) carefully locked (cf. yanta, 47,2) is reminis-
cent of the “embryogénie sacrée visant à fournir à l’adepte un modèle
dans sa quête du Nibbåna qui passe par une régression utérine”.9
On the other hand, it remains to be seen whether any conclusion can
be drawn from the analysis of the verses scattered in Bu-up and from
their identication or available parallels. In some cases, they are helpful
for a better establishment or understanding of the text. One of the most
striking cases is provided by the technically elaborated passage of
chapter 2 (43,16*–23*), the edition of which is unsatisfactory but can be
improved through recourse to the Abhidhammåvatåra and its
commentary (chapter 1, stanzas 29, 31, 62).10
2. THE PAÈHAMASAMBODHI
The Pa†hamasambodhi (Pa†h) can be described as a biography of the
Buddha coming from South-East Asia, and even more precisely from
Thailand, where the nineteenth-century version written by the prince-
monk Paramanuchit-chinorot (= Se ) is a well-known text : “[The ´
Pathamasambodhikatha] is a series of sermons intended for ritual

recitation at events such as the Wisakha Buuchaa, which are held all
night in commemoration of the Buddha’s birth, enlightenment, and
decease. It represents a Thai version of the standard biography of the
Buddha, which is based on canonical and Sinhalese commentarial works
and written in ornate prose style” (Taylor 1997 : 292). Although the Påli
´
Text Society edition was published only a few years ago, the interest in
this work is not new. It was brought to light by the French scholar
George Cœdès (1886–1969) who published two articles on this work
(1916 A and 1968) and had prepared its text using a large number of
8For bibliographical references regarding these two elements see Balbir 2001,
nn. 18 and 19.
9Bizot 1980, p. 222.
10See Balbir 2001, pp. 350–51 for further textual details.
manuscripts. This text is the basis of the PTS edition (= Ee), nalized
for publication by Dr Jacqueline Filliozat. Reading Pa†h through this
edition leaves the reader in a rather confused state, facing a large
number of variants which are not really helpful. On the other hand, the
tools which could be of use in understanding what Pa†h is or is meant to
be are missing. Given the form of the work where prose and verses
alternate, an index of stanzas, for instance, with a concordance would
have been appropriate ; instead the concordances to the verses of the
´
Nidånakathå (Nidåna-k),11 to which the Pa†hamasambodhi is obviously

connected, when mentioned, are buried among the variants. The stanzas
other than those examined below are either difcult to identify or come
from canonical works of wide circulation.12
Both George Cœdès and Jacqueline Filliozat were rightly puzzled by
the varying structure and contents of the work as evidenced by the
manuscripts. The number of chapters, for instance, is not always the
same. But one may go further into the textual history of the
Pa†hamasambodhi and its composition. In many ways the form of Pa†h
is reminiscent of a prose commentary to a verse text. Narrative prose
passages of varying lengths end with formulas of the type tena dassento
åha, accusative + dassento åha, or tena vutta, followed by one or
11Ja I 2–94.
12For instance Pa†h (references are to verse numbers) 70 = Nidåna-k (references
are to verse numbers) 271´; Pa†h 106 = Nidåna-k 272´; Pa†h 184–85 = Sn 544–
45´; Pa†h 190–91 = Dhp 153–54 and Nidåna-k 278–79´; Pa†h 160 = Dhp 179
and Nidåna-k 280 ´; Pa†h 198–99 = Vin I 3 ,27–30 and Ud 10,18*–21* (my
attention was drawn to these two stanzas by Th¥ Phumthapthim, Kånsukså
pr¥apth¥ap kamph¥ lalitwisatara ke khamph¥ pathomsomphØt. A comparative
study of the Lalitavistara and the Pathamasambodhi, Bangkok, Silpakorn
University, 2543 [2000], who on p. 6 draws a parallel between two Pa†h
stanzas and Lalitavistara, p. 380, lines 16–19 (Lefmann edition), p. 276
(Vaidya edition), but does not mention their old Påli occurrences. The parallel
is interesting but is it conclusive for any connection between the Påli and the
Sanskrit biographies of the Buddha?)´; Pa†h 224 = Nidåna-k 289´; Pa†h 225 =
Nidåna-k 290´; Pa†h 226 = Dhp 168 and Nidåna-k 292 ´; Pa†h 227, 228 = Dhp
169 and Nidåna-k 293.
338 Nalini Balbir
several stanzas. Their total number is 254 in Ee , but for a right
assessment of the situation it is better to take into account only the part
of the text going up to the end of the chapter Dhammacakkaparivatta
which is common to Ee and Se, i.e. the rst 223 stanzas of Ee, to which
nine stanzas wrongly not printed as verses (see n. 14) should be added.13
Fifty-nine of them, i.e. approximately one fourth, are similar to or
identical with stanzas found in the Jinåla¤kåra (Jinål ; see Table below).
´
This starts almost at the very beginning : ´
dasa-påram¥-dasa-upapåram¥-dasa-paramattha-påramiyo sabba-sama-
t¥sa-påramiyo pËresi. idåni eva bodhisatto påram¥ pËrento yathå amba-
rukkha-sadiså jana-chåyå phala-paribhoga-puñña-b¥ja-ropita-kkhaê yeva
ta upama sasandento imå gåthå åha (text quoted as in S e , p. 24;
compare Ee, p. 2)
tath’ eva sasåra-pathe janåna … (= Jinål 30)
yo sågare jalam adhika-rudhira adåsi … (= Jinål 31)
buddhå lokåloke loke jåto … (= Jinål 172)
Here, the verse concordance is not the only sign of the presence of
the Jinåla¤kåra. The preceding prose sentence (“like the shade for
people similar to the mango tree even at the time of sowing the seed of
merit for the enjoyment of the fruit”) is already a somewhat terse and
elliptic rewriting of verse 29 of this text preserving its important words
with a loose syntactic connection between them : ´
yo magga-passe madhur’-amba-b¥ja
chåyå-phal’-atthåya mahå-janåna
ropesi tasmi hi khaê va tena
chåyå-phale puññam aladdham uddha (Jinål 29)
He who has sown the seed of a sweet mango on the roadside with the object
of providing shade and fruit, even in the very moment of sowing it, in virtue
of the shade and the fruit [he intends to provide], there is acquired by him
whatever merit had not been obtained before (Gray’s translation, p. 85).
13Se represents an amplied version in 29 chapters (Cœdès 1968 in PTS edition´:

lvi–lvii) where, after this point, a great deal of additional material in prose and
verse is found.
In one case, the verses are precisely introduced with reference to
their literary form (yamakagåtham åha, Pa†h Ee 73,12), unmistakably
pointing to the Jinåla¤kåra, which is well known as a unique composi-
tion replete with tours de force, especially yamakas (see Pa†h 74 = Jinål
73). In another case, the sophisticated style of the stanza which makes
use of alliterations and paradoxical statements speaks for itself : ´
bhajita cajita pavana bhavana

jahita gahita samala amala
sugata agata sugati agati
namita (v.l. namåmi) amita namati sumati
Pa†h 53 [= 136,19–37,2 not printed as verse in Ee] = Jinål 173
In a single case, an author’s name is explicitly mentioned : ta ´
dassento Buddharakkhitåcariyo åha (Ee 114,12 ; Se p. 87), followed by a

´
verse (numbered 145 in Ee) which is identical to Jinåla¤kåra 115 (Ee).

This suggests that the connection between the two works was clear to
the redactor of the Pa†hamasambodhi himself and that the implied
grammatical subject of åha in many other cases is also the author of the
Jinåla¤kåra. This gives support to the identity of Buddharakkhita as the
author of the Jinåla¤kåra, a fact which was not unanimously admitted in
the tradition.14 The evidence of the Pa†hamasambodhi conrms what we
know for certain from the statements found at the end of the com-
mentary on the Jinåla¤kåra and from the colophons (Norman 1983 : ´
157 ; von Hinüber 1996 § 407), that the author of the Jinåla¤kåra was
´ ´
indeed Buddharakkhita, a Thera born in Rohaâ (Ceylon) who wrote it

in 1156 C.E.
On the other hand the distribution of the fty-nine stanzas common
to the Pa†hamasambodhi and the Jinåla¤kåra is not without signicance.
They are not spread over the fourteen chapters which build Pa†h in its
most complete form. They are found only in the part narrating the life of
the Buddha from his last incarnations in the Tusita heaven up to his
14See the conicting evidence of the Gandhavasa (Buddhadatta) and of the

Saddhammasa¤gaha (Buddharakkhita) quoted in I.B. Horner’s foreword to
the reprint of the Jinåla¤kåra.
340 Nalini Balbir
Enlightenment (Abhisambodhi, chapter VIII in Ee ; chapter XI in Se),
´
not in the later chapters. This fact could be additional evidence to

support the hypothesis made by Cœdès about the progressive develop-
ment of Pa†h in three stages starting around a core corresponding to this
period of the Buddha’s life :
´
The Pa†hamasambodhi originally may have included eight chapters which

traced the life of the Buddha from his life in the Tusita heaven to his
awakening.… A little later, the addition of two or three chapters continued
the story up to the sermon in Benares. This stage corresponds to the Yuon
translation in the manuscript of Copenhagen. Later still, the story was
continued up to the parinirvåâ. This stage included about fteen chapters
and is represented by the eighteenth-century manuscripts (Pa†h Ee lx).
As for the rst stage, one can now state that the Jinåla¤kåra stands
among its main sources. The table of correspondences (below) shows
that verses from the Jinåla¤kåra are often quoted in blocks so that some
sections of the Jinåla¤kåra are incorporated in toto or in part in the
Pa†hamasambodhi. Thus both works have a close intertextual relation.
This observation also gives weight to the chronological deductions
proposed by Cœdès on the basis of two other converging facts : (1) the
´
oldest sculptures that depict the Earth wringing out her hair in order to
inundate Måra’s army date from the twelfth century ; (2) a stanza of the
´
Pa†h found in an inscription from Nakhon Pathom in Thailand also

appears in the Såratthasamuccaya, which also dates from the twelfth
century. Since the Jinåla¤kåra also dates back to the same period, and
since the quotations from the Jinåla¤kåra appear precisely in the same
part of the work, we could be slightly more assertive than Cœdès, who
wrote, “We should not go as far as to imply that the Pa†hamasambodhi
itself dates from this period, even if the two chapters that include the
legend of the Earth and the stanza are part of the oldest part of the text”
(Pa†h Ee lxiv).
Trying to read the verses of Pa†h through the PTS edition is not an
easy task. First, there are passages which have been printed as prose
while they should have been printed as verses,15 or incorrect word
separations. Despite the considerable number of manuscripts used, the
text is often unsatisfactory, even in cases where it offers no special
difculty. Both the Thai edition of Pa†h reecting the full modern
version prepared by Prince Paramanujit in 1845 (= Se, pp. 278–79 for
the colophon verses already quoted in Cœdès 1916B : 4, n. 1) and the
´
editions of the Jinåla¤kåra have to be called on for help. On the other

hand, the establishment of the text of the latter would also benet from a
comparison of the two texts (see Table below).
It is not only in the verses that the connection between the
Pa†hamasambodhi and the Jinåla¤kåra is clearly seen. Comparing the
prose of the former with the commentary on the latter underlines their
mutual afnity. Although Pa†h is not strictly speaking a commentary, it
has some formal features of the genre : the style of the introductory
´
formulas preceding the quotation of stanzas is one of them (see above).

Another one is the typical device of singling out a word of a stanza just
quoted for explanation : ´
saya Nåråyanabalo abhiññåbalapåragË

jetu sabbassa lokassa Bodhimaˆ am upågam¥ ti (Pa†h Ee 148, Se p. 88,
v.l. påramibalapåragË = Jinål 118)
tattha “Nåråyanabalo” ti … tattha Nåråyanabalo nåma dassento åha´:
kå¬åvakañ ca Ga¤geyya
paˆ ara tåmba-pi¤gala
gandha ma¤gala hemañ ca
uposatha chaddan t’ ime daså ti (Pa†h Ee 149, Se p. 88´; Jinål-† Be p. 275)
tattha “daså” ti …
The wording of the commentary on Nåråyanabalo in Pa†h and in the

Jinåla¤kåra-†¥kå (Jinål-†, Be) are almost identical, and the stanza listing
the ten powers is also found at the same place in this commentary. A
15The following passages should be printed as verses´: Ee 23,8–10 (Se p. 40)´; Ee

79 ,11–13 (Måra … atthiko) and 79 ,13–14 (sabba … anuttaro) are two
anu¡†ubhs (Se p. 69)´; Ee 95,5–7 (Se p. 77)´; Ee 111,14–15 is the continuation of
the stanza numbered as 142 (Se p. 86)´; Ee 128,15–17 (Se p. 96)´; Ee 136,15–
137,2 (3 stanzas, see below´; Se pp. 100–101).
342 Nalini Balbir
full edged comparison of the prose of both texts, which cannot be
undertaken here, would easily show that this is just an instance out of
many where prose passages in both works have the same wording and
where stanzas incorporated in Pa†h are quoted at the same place in the
Jinål-†.16 The ultimate source of these common stanzas could well be a
third text : the ve stanzas listing the bad omens appearing before
´
Måra’s army in Pa†h (Ee 161–65) are also found (with variant readings)
in the Jinål-† (Be pp. 277–78), where they are introduced with the
sentence : vutta h’ eta poråê ti. Needless to say, great benet could
´
be taken from such a comparison for improving the often decient or

unclear text of Pa†h as given in Ee (despite the impressive critical
apparatus). The interrelation with Jinål, however, is rather complex. The
passage where the goddess Earth (Vasundharå vanitå) wrings out her
hair in order to inundate Måra’s army and cause his nal defeat (Pa†h Ee
134,17ff.), made famous by Cœdès’s article (1916 A ) is specic to the
Pa†hamasambodhi, and appears to be deliberately so, as the version of
the Jinåla¤kåra is in conformity with the classical depiction with the
earth shaking, the terrestrial noise, and the roaring noise in the sky
caused by a thunderbolt.17 Except for these few lines, the rest of the
prose of Pa†h is rather close to what can be found in the corresponding
Jinåla¤kåra-†¥kå (B e p. 285) : what comes before this episode is a
´
commentary on stanza 181 (= Jinål 138) similar to Jinål-† and what

comes after it (135,11ff.) is similar to Jinål-† on Jinål 140–41 (not quoted
in Pa†h).
16Compare, for instance, Pa†h Ee p. 137 and Jinål-† quoted in Jinål Ee p. 63 (Be
p.´289)´; Pa†h Ee 136 ,15–19 (not printed as verses´!) = Se p. 86 = Jinål-† Be
p.´286 as stanzas 161 and 162.
17The Earth as a beautiful lady who appeared in front of the Buddha is given at
an earlier stage of the narrative as told in Pa†h, at the time of the Great
Renunciation´: tadå Dharaˆ¥ varavanitå Bodhisattassa vitakka ñatvå, etc.
(Ee 80,8).
Table18
Pa†hamasambodhi verse number Jinåla¤kåra verse number in Ee
in Ee (chapter)19 (chapter)
3 (Tussita)´; b reads differently in Se´: 30 (Bodhisambhårad¥pan¥gåthå)
atthåya attånam achådayanto´; c
has been transmitted differently in
Pa†h and Jinål20
4 (Tussita)´; read jaladhikarudhira 31 (Bodhisambhårad¥pan¥gåthå)
in a Ee
5 (read ko nu mmatto´; bho prob. to 172 (Abhisambodhid¥pan¥gåthå)
be read as ko) = 54 (Tussita)
6 (Tussita) 32 (Gabbhokkantid¥pan¥gåthå)
7 (Tussita) 33 (Gabbhokkantid¥pan¥gåthå)´;
superuous ca in c
13 (Gabbhåbhinikkhamana)´; d is un- 34 (Gabbhokkantid¥pan¥gåthå)
metrical in all versions (dasa-
sahass¥ pakampitha Jinål).
14 (Gabbh.) 35 (Gabbhokkantid¥pan¥gåthå)
18 (Gabbh.)´; note Lumbali / Lum- 36 (Vijåyanama¤galad¥pan¥gåthå)
bani in b´; read vijåyi ta in d Ee
19 (Gabbh.) 39 (Vijåyanama¤galad¥pan¥gåthå)
20 (Gabbh.)´; c vivattanti´; d na 38 (Vijåyanama¤galad¥pan¥gåthå)´; c
dissare cåmarachattagåhakå vipatanti Ee´; vijenti Be´; d kha-
jjisu bher¥ ca nadisu sakhå
46 (Gabbh.)´; d te devå dåtå Ee is 178 (Navaguâd¥pan¥gåthå)´; d te
strange (no v.l.)´; Se te devå tadå devå brahmå
makes more sense.
48 (Lakkhaâpariggaha)´; b Ee 44 (Vijåyanama¤galad¥pan¥gåthå)´; b
subhattå, Se subhuttå subhuttå
53 (Lakkhaâ.) 173 (Abhisambodhid¥pan¥gåthå)
18Based on the comparison of Pa†hamasambodhi Ee with Jinåla¤kåra Ee.

19Chapter division as in Ee.
20Only the most signicant variants are recorded here´; incorrect word separa-
tions are not taken into account. For a full critical edition of the verses all
available documents would have to be taken up systematically and their
readings considered in view of the metrical constraints.
344 Nalini Balbir
54 (Lakkhaâ)´; not in Se at this 172 (Abhisambodhid¥pan¥gåthå)
place´; see above 5
55 (Lakkhaâ)´; wrong divisions of 180 (Navaguâd¥pan¥gåthå)
some words in Ee´; not in Se at
this place
56 (Råjåbhiseka)´; b Ee åbhåyika´; 45 (Vijåyanama¤galad¥pan¥gåthå)
Se abbhåyika, compare Gray
Jinål p. 56.
57 (Råjåbhiseka)´; d Ee Se deva- 43 (Vijåyanama¤galad¥pan¥gåthå)´; d
puttam upågami devaputtattam ågami
58 (Råjåbhiseka 47 (Agåriyasampattid¥pan¥gåthå)
59 (Råjåbhiseka)´; minor variants 48 (Agåriyasampattid¥pan¥gåthå)
73 (Mahåbhinikkhamana)´; d itthiyo 72 (Påduddhåravimhayad¥pan¥gåthå)
in Ee to be cancelled as indicated
in the note.
74 (Mahåbhinikkhamana) 73 (Påduddhåravimhayad¥pan¥gåthå)
75 (Mahåbhinikkhamana) 74 (Påduddhåravimhayad¥pan¥gåthå)
76 (Mahåbhinikkhamana)´; d Ee Se 75 (Påduddhåravimhayad¥pan¥-
saghu††hå gåthå)´; d ghu††hå.
79 (Mahåbhinikkhamana)´; b Ee 81 (Apunaråvattigamanad¥pan¥-
jiˆâvirËpåni raticchidåni but Se yamakagåthå)´; b th¥na virËpåni
th¥na virËpåni ratacchidåni ratacchidåni
81 (Mahåbhinikkhamana) 50 (Nekkhammajjhåsayad¥pan¥ya-
makagåthå)
82 (Mahåbhinikkhamana)´; c Ee 52 (Nekkhamma.)´; c maccumukhe
paccamukhe but Se maccumukhe
83 (Mahåbhinikkhamana) 53 (Nekkhamma.)
84 (Mahåbhinikkhamana)´; c Ee 54 (Nekkhamma.)´; c mayå te´; d
payåte´; Se mayå te´; d vinay(y)å vineyya
85 (Mahåbhinikkhamana)´; d Ee 59 (Nekkhamma.)´; d suta sutanta
suta sutan ´;Se suta sutanta
87 (Mahåbhinikkhamana)´; c Ee 88 (Dvipådabyåsayamakagåthå)´; c
GåyåbhirËpa pi Yasodhara- tåyåbhirËpa pi Yasodhara
vara´; Se tåyabhirËpa pi vara
Yasodhara vara
88 (Mahåbhinikkhamana)´; c to be 89 (Dvipåda.)
emended´: the pun on the verb
riñcati which is the main point of
the verse is lost in the text as it is
printed´; compare Jinål
91 (Mahåbhinikkhamana)´; c is 68 (Påduddhåravimhayad¥pan¥gåthå)
different in Pa†h and Jinål
92 (Mahåbhinikkhamana)´; b the 85 (Apunaråvattigamanad¥pan¥-
correct reading should be yamakagåthå)
bhåbhånibhåni
93 (Mahåbhinikkhamana)´; d EeSe 86 (Apunaråvatti.)´; d yåto tato hi
yåto tato hi nimittehi surissarehi mahito purisassarehi
94 (Mahåbhinikkhamana)´; minor 69 (Påduddhåra.)
variants
95 (Mahåbhinikkhamana)´; to be 70 (Påduddhåra.)
fully reconsidered
97 (Mahåbhinikkhamana) 78 (Apunaråvatti.)
100 (Mahåbhi.) 81 (Apunaråvatti.)
145 (BuddhapËjå)´; c Bodhi 115 (Måraparåjayad¥pan¥gåthå)´; c Ee
paråjitåsane ´; d Ee Se yuddhåya bodhimhi paråjitåsane ´; Be
Måre niccalo nis¥di bodhim uparåjitåsane´; d
yuddhåya Mårenacalo nis¥di
146 (BuddhapËjå) 116 (Måraparåjaya.)´; a read
månusa (not masa as printed
in Ee´; but the translation is right).
147 (BuddhapËjå) 117 (Måraparåjaya.)
148 (BuddhapËjå)´; Se påram¥bala- 118 (Måraparåjaya.)
påragË
166 (Måravijaya)´; b Ee Se chaddetha 120 (Måraparåjaya.)´; b cha††etha
chedaka ima ce†aka ima
170 (Måravijaya)´; a Ee åsinno ´; Se 123 (Måraparåjaya.)´; a kasmå åsi nu
åsi no Ee, no Be
346 Nalini Balbir
171 (Måravijaya) 124 (Måraparåjaya.)´; read dayåparo
instead of daya° in Ee
172 (Måravijaya)´; d Ee Se balo 125 (Måraparåjaya.)´; d kha¬o
173 (Måravijaya)´; c Ee ekapårimisså 127 (Måraparåjaya.)´; c ekapåramiyå
pi ´; Se ekapåramisså pi
174 (Måravijaya) 128 (Måraparåjaya.)
175 (Måravijaya)´; d Ee anumato 129 (Måraparåjaya.)´; d anummatto
sacetano´; Se anumatto sacetano
p. 136,19–37,2 (wrongly not printed 173 (Abhisambodhid¥pan¥gåthå)
as a verse in Ee) = 53
3. THE VIDADDHAMUKHAMAÔANA21
A solid hint as to the existence of this work in Burma is supplied in
Aggavasa’s Saddan¥ti (see Kraatz 1968 1 : xvi) :
´ ´
må vuccati sir¥´; tathå hi Vidaddhamukhamaˆ ana-†¥kåyå målin¥ ti

pådass’ attha vadatå “må vuccati Lakkh¥, alin¥ ti bhamar¥” ti vutta,
lakkh¥ saddo ca sir¥-saddena samån’-attho, tena “må vuccati sir¥” ti attho
amhehi anumato (244,19ff.).
må means “prosperity”. In fact, when giving the meaning of the word
målin¥, the commentary on the Vidaddhamukhamaˆ ana says, “må means
21Atthis stage I can only collect a few preliminary remarks. More details on the
text will follow on another occasion. I am grateful to all those who, in
addition to Dr Peter Skilling, helped me to progress in this research during my
stay in Bangkok (August 2007)´: Peter Nyunt, who is cataloguing the Fragile
Palm Leaves Manuscripts´; Venerable Mahathiab Malai of Wat Jetuphon (Wat
Pho), who granted permission to see the manuscripts kept there´; Jacqueline
Filliozat, who kindly sent the relevant information contained in her
unpublished catalogue of the manuscripts at Wat Jetuphon (Wat Pho) and
accompanied me there during our brief visit on 29 August´; Mr Dokrak
Payaksri and Mr Wisithisak Sattapan (EFEO, Bangkok), who kindly devoted
a few hours to the reading of parts of the two Tham manuscripts, photocopies
of which were kindly provided by Dr Peter Skilling (see below).
prosperity, alin¥ means bee.” The word lakkh¥ has the same meaning as the
word sir¥. This is why we have admitted the statement “må means
prosperity”.
This passage occurs within a section devoted to the discussion of

monosyllables (ekakkhara, 239 ,6 –46 ,8 ) in the context of nominal
declension and the establishment of grammatical gender of the words
considered. They are reviewed in alphabetical order, just as a
specialized lexicon of the class Ekakkharakosa would do. Starting with
ko meaning “Brahmå, wind, and body” (239,6ff.), the list ends with sa
(245,4ff.). Compounds formed with monosyllables are treated along the
way (such as vindo, “lord of the birds”, vi + indo, 240,4–5). As always
with Aggavasa, the discussion is substantiated by examples and
quotations taken from various texts.
The presence of this quotation in the Saddan¥ti implies that not only
the work itself but a corresponding commentary were known at the time
of Aggavasa, that is to say, in the second half of the twelfth century
C.E. The question of its origin and diffusion, however, have not yet been
solved. Vid is not specically a Påli work ; there is a Vid in Sanskrit,
´
which, in four chapters, presents both denitions and illustrations of

various types of riddles, and was widely disseminated in India It is a
sophisticated work which calls for knowledge of grammar and
vocabulary (especially monosyllable words or rare words) in all their
niceties. Therefore in addition to manuscripts and editions containing
the verses only, there are many where an elucidating commentary is also
provided. The religious afliation of the author, a certain Dharmadåsa,
about whom nothing reliable is known, has been debated : was he a´
Buddhist, a Jain (Vid is highly popular in Jain circles, where other

authors have also composed similar works), or neither ? His date is also
´
very uncertain : near the seventh century (Kraatz 1968: xviii) or much
´
later (the eleventh, thirteenth, or fifteenth century).

The passage quoted by Aggavasa refers to a stanza which reads as
follows in Sanskrit :
´
348 Nalini Balbir
urasi Mura-bhida kå gå ham åli¤gitåste´?
sarasija-makarandåmoditå nandane kå´?
giri-sama-laghu-varâir arâvåkhyåtisakhyair
gurubhir api k¤tå kå chandaså v¤tti-ramyå´?
målin¥. må = lak¡m¥´; alin¥ = bhramar¥´; målin¥ nåma chando v¤ttam
(2.36).
Which lady remains closely embraced to Mura’s murderer´?
In the Nandana who (fem.) is rejoiced in the pollen of the lotuses´?
Having a number of light syllables identical to [the number of]
mountains, and heavy syllables numbering the word “ocean”,
which among the metres is pleasant´?
The answer to the rst question is må, a monosyllabic designation

for Lak¡m¥, the answer to the second one is alin¥ “a bee”, whereas the
addition of both produces the answer to the third question, målin¥, as the
name of the famous metre having eight light and seven heavy syllables
(4 Ÿ       ¯ ¯, ¯  ¯ ¯  ¯ ¯´). The stanza is meant to illustrate the variety
of riddles known as v¤ttanåmajåti, where the answer to be found is the
name of a metre. It is is the second example of this variety : in the ´
preceding stanza, 2.35, the name of the metre to be guessed through a

similar method is ßikhariˆ¥. An additional nicety: although Dharma-
dåsa’s definition does not state it explicitly, both his examples show that
the riddle verse is written in the metre to be discovered.22
This parallel suggests that the Påli and the Sanskrit Vid are closely
interrelated. Further, Aggavasa’s quotation could make one expect
that manuscripts of a Påli Vid with commentary following the Sanskrit
model could be found in Burma. There are serious hints, indeed, to
suggest that the tradition relating to Vid was kept alive in Burma even
later than Aggavasa’s time. At a later period there are stray references
found in historical documents or lists of books. For instance, Vidagdha,
rightly understood by Bode (1909 : 108, No. 265) as the abbreviation of
´
Vidagdhamukhamaˆ ana, is mentioned among the non-canonical works

found in the Pagan inscription dated 1442 A.D. which gives the contents
22This
additional feature is made clear in the Jain reworking of the definition in
Mahåkavi Ajitasena’s Alakåracintåmaî, see Balbir 2004, p. 299.
of the library (Luce & Tin Htway 1976 : 246, No. 268). The Gandha-
´
vasa mentions Vid in a discussion relating to Vepulabuddhi’s works

(see below). The Pi†akat-tØ-samui¤ : (= Pi†akatthamain, Nos. 1065 and
´
1066) mentions a manuscript of the text of Vid (the language of which

is not given) and a manuscript of a commentary composed by Vimala-
buddhi, described as a monk from the Mahåvihåra in Anurådhapura, a
fact which suggests that the commentary could have been written before
1017 (Kraatz 1968 : xvi–xvii).
´
So far no manuscript evidence of Vid seems to be available from Sri

Lanka. The only manuscripts of Vid which have been traced come from
South-East Asia.
The only Burmese manuscript of Vid which could be traced and
consulted so far is not a Påli work (Ms 510 belonging to the Fragile
Palm Leaves collection, “Manuscript House”, Pakkret, Bangkok). It is a
very clearly written manuscript of the Sanskrit work by Dharmadåsa in
Burmese script, with the usual signs for noting Sanskrit phonemes.23
Vid occupies folios ka to khi (8 lines per page) and is at present in a
bundle containing the following works : Sandhikålåp på†h,
´
Paroparissabheda n¥ pat nisya, Abhidhammavibhåvan¥-†¥kå and È¥kå-

kyo.
Beginning´: nama sarvvajñåya || ||
siddhau¡adhåni bhava-dukha-mahågadånå, etc. (= Skt VMM 1.1)
End ´: iti Dharmmådåsa-kïte Vidaggamukhamaˆ ane caturttha pari-
ccheda || Vidagga-granthan ni††hita || || |||
akkharå ekkam ekañ ca Buddha-rËpa sama siyå
tasmå hi paˆ ito poso likheyyå pi†aka-ttaya ||
Marginal title on the last folio´: Vidag kwyam mrat.
This is a manuscript of the mËla only without any commentary. The

four chapters of the work as distributed as follows : 1 ends on k¥ verso,
´
line 7 ; 2 on kai recto, line 1 ; 3 on ka recto, line 1. The author’s name
´ ´
is consistently written throughout as Dharmmådåsa. The verses are

numbered, starting from 1 at the beginning of a new chapter, but not
23See Bechert 1979, p. xxi (“Table of Transliteration”).

350 Nalini Balbir
throughout. The verse to which Aggavasa refers is found on fol. kË
recto and verso :´
urasi muda (sic)-bhidå kå gå hyam åli¤gitåste´?

sarasija-makarandånanditå moditå kå´?
g¥ri-sama-laghu-varˆâir aˆâvåkhyåta-sa¤khye (sic)
garubhir api kïtå kå cchandaså (sic) vïttir agrå´? || målini || vïttanåma-jåti ||
|||
A few lines above (kË recto, line 5) the verse occurs which
successively gives the definitions of two varieties of riddles, the second
of which is the vïttanåma-jåti (see below). On the whole, the manuscript
is correct. Neither Vid nor the rest of the manuscript have any date or
place of copying. As is well known, the position of Sanskrit learning in
Burma was very different from that of Ceylon. Påli and Burmese were
the common languages in monastic education. Sanskrit, however, was
not absent and remained associated with specialized traditional
disciplines of knowledge (ßåstras). Vid, which combines knowledge of
grammar, lexicography, metrics, poetics, etc., belongs to such a sphere.
In particular, “King Bodawpaya (1781–1819) … sent a number of
missions to collect Sanskrit works in Varanasi and other places in India
and Ceylon. These books were transliterated into Burmese script and
many of them were translated into Burmese language or into Påli”
(Bechert and Braun 1981 : xxxix). The manuscript of Vid could date
´
from this period and could belong to this Sanskrit renaissance, although
the work does not appear in the rich list of “Sanskrit texts imported into
Burma between 1786 and 1818” (Than Tun 1960 : 132–41). Thus, this
´
idea is only a mere hypothesis for the time being. Given the small
number of Sanskrit works in Burmese script, it is certainly remarkable : ´
The scope of Sanskrit studies in Burma remained, however, a quite limited

one so that today not many Sanskrit works can be found in manuscripts
written in Burmese script (Bechert and Braun 1981´: xxxix).24
24See loc. cit. for examples´: “only eight Sanskrit manuscripts in Burmese script
with 14 different works, mostly grammatical and lexicographical texts, can be
traced” in the unpublished catalogue of the Mandalay collection.
As for the presence of Vid in other South-East Asian countries, the
situation is the following : no manuscript seems to be available today in
´
Cambodia.25 A manuscript from Laos has been reported long ago.26 The
existence of vernacular versions, however, has been reported (Skilling
and Pakdeekham 2002, 2004).
Nevertheless, the existence of a Påli Vid is not a myth. It is
attested in several manuscripts from Siam, all of which have not yet
been collected. 27 On the other hand, the list of works making an
extensive “painted Tipi†aka” found on the walls of the main hall in Wat
Thong Noppakhun (Thonburi ; end of the nineteenth century) shows that
´
Vid was known among works dealing with language (Saddåvisesa),

both in its Påli and in its Sanskrit versions until late : sixty titles are
´
listed in this category. No. 27 is På¬¥-Bidakdha and No. 55 is På¬¥-

Bidakdha-saka†a (Skilling, forthcoming).
My preliminary investigation of the Påli Vid is based on the
following material :´
One manuscript in Khom script kept at Wat Jetuphon (Wat Pho). No.
6/40. See Jacqueline Filliozat, “EFEO DATA Filliozat 2005, fichier 108”.
The whole bundle concerns Vid. The Påli version (Brah påli
vidagdhamukhamaˆ ana) is found on fol. ka to gË and was the only one I
could see briefly during my visit. The next ms (7 phËks) is the Vidagdha-
mukhamaˆ ana-d¥pan¥-†¥kå, followed by the Vidagdhamukhamaˆ ana-
yojanå (4 phËks) and the Mukhamaˆ anavidagdha-upadesa.28
Two photocopies made on the basis of the microfilms of two manuscripts
in Tham script from Wat Sung Men, Phrae Province. These manuscripts
25Information kindly given by Dr Olivier de Bernon (EFEO´; letter dated 7 May

2001). But see Coedès 1912´: 178 who saw a manuscript of the †¥kå.
26Finot 1917´: 214´: R 676 (= Luang Prabang Royal Library) containing 6 phËks.
27It would be important for a further study to have access to the ms kept in the
Royal National Library.
28For other manuscripts, including some containing vernacular renderings or
explanations, see Skilling & Pakdeekham 2002 under 4.49, 4.72 (Nissaya-
Vidagdhamukhamaˆ ana-Phadet), 4.99 (Yojanå-Vidagdhamukhamaˆ ana
“composed in Pukåm [Pagan] by Dhammakitti Thera Lokaråjamol¥”), 4.110
and 111´; Skilling & Pakdeekham 2004 under 5.101–104.
352 Nalini Balbir
were microfilmed under the “Preservation of Northern Thai Manuscripts
Project” (a Thai–German undertaking on which see Hundius 1990´: 15ff.).
Reference is to the phËk number and the Arabic numbers added on each page
of the microfilmed manuscripts.
WS 010408801 (= A), 6 phËks, 5 lines, complete, dated C .S . (=
CË¬asakaråja) 1198 = 1836 C.E.
WS 010409203 (= B), 6 phËks, 5 lines, complete.
Both manuscripts are additional documents attesting the brilliant

activity of the senior monk Venerable Gruu Paa Kañcana Araññavåsin
whose personality emerges from the colophons of the manuscripts he
had copied, and came to light through the superb study of the colophons
of Påli manuscripts from Northern Thailand conducted by Hundius
(1990, especially 34–36). In the 1830s this monk was greatly
instrumental in preserving and restoring Northern Thai culture in Lanna.
His home monastery, Wat Sung Men, “rose to become a centre of Påli
and Buddhist studies. Manuscripts were systematically collected and
numerous copying campaigns covering Phrae, Nan, Chiang Mai, Chiang
Saen, Rahaeng and Luang Prabang were pursued” (Hundius 1990 : 34). ´
Together with the ruler of Phrae he had ms A copied. His name also
appears in the colophon located at the end of each phËk of ms B. Under
his leadership, and with the cooperation of his disciple, this ms was
copied in Luang Prabang and brought to Lanna.
The Khom manuscript contains the root text of the Vid in Påli : ´
definition verses, illustrative verses followed by the answers to the

riddles. It is the work of Vipulabuddhi Thera, disciple of Sågarabuddhi
Thera :´
iti Sågarabuddhither’-antevåsika-Vipulabuddhithera-viracite Vidattha-

mukhamaˆ ane catuttho paricchedo … på¬¥ Vidatthamukhamaˆ aâ
ni††hita (fols. gË–ge).
This should be compared with the Gandhavasa of Nandapañña,

admittedly a modern work, where the number of works composed by
Vepullabuddhi Ócårya (either five or six as there seem to be conflicting
opinions) is discussed. Among them is one Vidadhimukhamaˆ ana†¥kå
(Minayeff 1886 : 64 and 74–75). The variant spelling vidadhi is no
´
hindrance to the identification. But the work available in the Khom

manuscript is not a commentary.
The two Lanna manuscripts are identical to each other and contain
the Vidagdhamukhamaˆ ana-d¥pan¥ (= Vid-d ; cf. Vidagdhamukha-
´
maˆ ana-d¥pani vakkhåmi, B I,1, line 3 ; A I,1, line 3). It is an

´
extensive Påli commentary on the Påli Vid. The verses of the mËla are
quoted påda by påda. They are identical to the work represented in the
Khom manuscript. Vid-d ends : iti varamati- se††hagaruna vajj¥rapañño
´
ti vihita-nåmadheyyena para-hitesinå uttama-dhamma-gavesi tena

nibbånålambaâ-cittena therena racitå Vidaggamukhamaˆ ana-d¥pan¥
nåmåya †¥kå anantaråyena samattå (B VI,30, line 5).29
The author of the Påli Vid explicitly considers himself to be a
translator of Dharmadåsa’s work at the outset :
´
… karissåmi sa-måtika aha Magadha-bhåsåya Vidagdhamukha-

maˆ ana.
The verses that follow, ending with the conclusion ti måtikå (Wat
Pho ms, fol. ki recto), list all the varieties of riddles which will be
treated in the work. These verses are Påli translations of the
corresponding verses found in Dharmadåsa’s work (1.9–18) with minor
adjustments in the use of particles. The technical designations are
identical. The verses supplying the definitions also conform to their
Sanskrit model. The definition of the first variety discussed in the Påli
Vid reads :
´
siyå pada-vibhågena kevalen’ eva pucchita

ya byattha ta samattha ya samudåyena pucchita (fol. ki recto,
line 2)
29The Vidagdhamukhamaˆ anad¥pan¥-†¥kå in Khom script (Wat Pho ms, see

above) has the same end and is the same work.
354 Nalini Balbir
cf. Dharmadåsa : ´
pï¡†a pada-vibhågena kevalenaiva yad bhavet

vidur vyasta samasta yat samudåyena pïcchyate (1.19).
This is just one instance to illustrate an overall conducted method.

The vïtta-nåma-jåti, which I focus on here because of Aggavasa’s
quotation, is defined along with the vi¡ama-jåti in the same verse. In
Dharmadåsa’s version : ´
yatra bha¤gasya vai¡amya vi¡ama tan nigadyate30

vïttanåmottara p¥¡†a bhavet tad vïttanåmakam (2.32)
variant : vï tta nåmottara yatra praßna tad vï tta-nåmakam

´
(Kraatz 1968 : 32).

´
… [Where] the question has as its answer the name of a metre it would be a
vïttanåmaka.
In Dharmadåsa as found in the Burmese manuscript (No. 510 see

above) it reads : ´
yatra bha¤gasya ve¡ama vi¡aman ti nigadyate

yatra praßnåsthita s tan nåmottaråda (?) vïtta-nåmaka.
In the Påli Vid (ms B III,25, line 2) : ´
yatra pabandhe bha¤gassa vesama atthi ta visaman ti nigadyate

yatra pabandha (for´: -e) nåmottara pañhå-††hita (sic) ta vutta-
nåmaka.
Thus, the general plan of both the Påli and the Sanskrit versions
goes along the same line. The fourth and last section, for example, also
deals with the same varieties as the Sanskrit model in the same
sequence. It relates to varieties where one has to discover a hidden verb,
a case form, a compound or a ending : kriyå-gutta, katta-gutta,
´
kamma-gutta, karaâ-gutta, sampadåna-gutta, apadåna-gutta,

adhikaraâ-gutta, sambandha-gutta, ålapana-gutta, samåsa-
gutta, etc.31
30Cadence of an even påda in a.

31Ms B phËk 6 fols. 4–15´; compare Vid in Skt. chap. 4 vv. 33ff.
If the matter stopped here, it would be very artificial to speak of a
Påli Vid. The originality of the Påli tradition regarding this work lies in
the illustrative verses. This is not surprising as the riddles are highly
dependent on linguistic constraints, which are partly different for
Sanskrit and Påli. No generalization is possible without a complete
reading of the work (not done so far). But the section concerning the
“metre name variety” shows that the examples are utterly different from
those in Dharmadåsa’s Vid. Therefore, up to now, Aggavasa’s
reference remains the only available trace of a Påli commentary on Vid
where the illustrative verse is supposed to be the same as in the Sanskrit
version.
Dharmadåsa had two examples for this variety (2.35 ßikhariˆ¥ and
2.36 målin¥). The Påli Vid as represented in our Khom and Tham
manuscripts also supplies two. But the metres they select are rucirå and
ketumat¥.
Example 132
(i)33 jinassa kå jalati varassa bhuvane ? ´
(ii) pahanti ’kena ’ghika-pajåya tena ke ? ´
(iii) abhiñña pañca garu lahu ’††ha sånikå

(iv) muni’-gga-vaˆâ-gha†ita-bandha-vutti kå ? — rucirå
´
Commentary : añña lakkhaâssa lakkhaâm åha jinass’ icc-ådinå.

´
tattha rucirå ti.

(i) varassa jinassa kå bhuvane jalati ? ruci. tattha bhavanti34 sattå
´
ettha bhuvana( ?)35 loko “bhË sattåyan” [= Dhåtupå†ha I.1] ti ti vå tu

´
32Ms A phËk III, 35–37´; ms B phËk III, 28–30. My aim is to give a sample of
the text because so far no discussion of the Påli Vid has been based on any
textual evidence. The present transliteration and translation, however, are
highly tentative and have gaps. Unfortunately, the relevant pages of the
photocopies are of rather poor quality and, at some places, hardly legible.
33These numbers refer to the question in the riddle. In Example 2 one of them
does not correspond to the påda boundary.
34B´: bhavanta.
35A´: etthå ti bhavana.
356 Nalini Balbir
yu-ssa 36 anatta u-kårassa 37 uvattañ ca.38 rucati attano guêna
virocati ti ruci rasi.
(ii) tena ekena39 aghika-pajåya ke pahanti ?40 ti. arå. arå. tattha
´
pahanti ’kenå ti pahiyyante ekena se††hena. 41 aghika-pajåyå ti

dukkhita-sattassa. arå ti kileså sasåra-cakkåro vå.
(iii–iv) abhiñña pañca garu la††ha sånikå42 muni-’gga-vaˆâ-
gha†ita-bandha-vutti kå ? rucirå. tattha abhi. la. sånikå ti abhiññå-
´
sakhåtehi pañca-garuhi ceva43 sånika jhåte samåpatti-sakhåtehi ca

a††hahi lahuhi ti samåno. muni. la. vutti ti agga-munino guêna
gha†ita-bandha-gåthå. kå ? ti, kå nåmå ? rucirå ti eva-nåmakå 44
´ ´
gåthå abhivisesena ra-guâ 45jånåti ti. abhiññå gåraviyate alahu-

karaâ ca sena bhaîyate ti garu, lahu …46 guîyate ti garu-nirutti-
nayena, 47 sånati 48 vå …49 karot¥ ti såna samåpatti tåya sånåya
sampannå sånikå gåthå. vaˆîyate sasiyate50 ti vaˆô,51 guˆå sa garu
36A´: yusså.
37A´: u-kårass’ uvattañ ca.
38Indigenous etymology of bhavana/bhuvana with reference to the root bhË and
grammatical formation of the word´: yu is the technical name of the suffix
-ana- (cf. Kaccåyana 549 nandåd¥hi yu and 624 or Sadd 859,23)´; -u-/ uv- in
words having this suffix.
39So A´; B´: te jinena na ekena aghika°.
40B´: panti.
41B´: written as sebbena.
42A´: samånikå (here, but later´: sånikå).
43B´: cava.
44A´: eva-nåmikå.
45So in both mss. Read´: °gaâ´?
46Very uncertain reading´: ke vyaddhi vya (??).
47Indigenous etymology of the word garu.
48Any connection with Sadd 398,5 såna tejane. tejana nisåna, sånati´?
49Too uncertain.
50So A´; B´: pasiyate.
51B´: vaô. Indigenous etymology of the word vaˆâ.
lahu hi rucati dippati ti52 rucirå.
(i) What is it (fem.) of the excellent Jina that shines in the world ? —
´
His brightness (ruci ; i.e. his rays, see cty rasi).

´
(ii) What are those (plural masc.) of a suffering creature that he

alone kills ? – The spokes (arå).
´
(iii–iv) Five higher knowledges [are] heavy, eight light …53 which
is the syllabic verse arrangement produced by the best of the sages ? — ´
The (metre) rucirå.

The metrical structure of this metre is as follows : 4 Ÿ  ¯  ¯     ¯ 
´
¯  ¯ (thus eight light and five heavy syllables ; cf. Sadd 8.3.2.4 and Vutt
´
89). As per Dharmadåsa’s model, the riddle verse itself is composed in

the metre to be discovered. It is a rucirå.
Example 2
(i) kissa vidhassa jantu muni bha¤go
(ii) loka-varo ’ssa dhamma-vidutå kå ? ´
(iii–iv) vutti lahu ccha panca garu bandha

nåtha-guâ¤ga-vaˆâ-racitå kå ? — ketumat¥.
´
Commentary : aññam åha : kiss’ icc-ådinå. tattha ketumati ti.

´ ´
(i) loka-varo muni jantu kissa vidhassa bha¤go ( ?) ti. ketu. tattha
´
kissa vidhasså ti kidisasså månassa, jantu ti jantuno. ketu hi unnati-

bhåvena dhaja-sadisassa månassa …54 måno kinåti55 unnamati ti ki
unnamati ketu56.
(ii) 57as(s)a dhamma-vidutå kå ? ti mati. tattha dhamma-vidutå ti
´
dhamma vijånana-bhåvo, mati ti.

(iii-iv) pañcåsanåthaguâ¤gavaˆâ-racitå lahu ccha pañca garu
52Indigenous etymology of rucirå as the name of the metre, meaning “pleasing,

shining, illuminating”.
53Despite the commentary I am at a loss to understand the word sånikå.
54Uncertain´: vidhati ettha na vidahati ti vivo (?) in B´; A is illegible.
55Compare Abhidhamma-avatåra 2´: kinåti vinåseti vå para-dukkhan ti karuˆå.
56Etymology of ketu connected with the root ki, kinåti.
57This part not in A.
358 Nalini Balbir
bandha-vutti kå ? ti sa ketumati. tattha akkharehi vutti ti gåthå. lahu
´
pañca (ga ?)ru bandha ti saha lahuhi, .. pañca garuhi ca bandha nåma
´
guâ¤ga-vaˆâ-racitå ti. nåthassa guâ a¤ga ..ehi vaˆêhi racitå

ketumat¥ ti, eva-nåmikå gåthå akitabbå lakkhitabbå ti akåte yeva
a¤gå, garu lahu li una. tatthå ketu viyå ti ketu. .. aså atthi ti ketumati.
vutta-nåma tassa jåti ti.58
(i) A being ( ?) of which type does the sage, the best in the world,
´
break ?59 — The banner (i.e. conceit) (ketu).

´
(ii) What is it (fem.) belonging to him that enables him to know the
Dhamma ? — The intellect (mati).
´
(iii–iv) A metre with six light and five heavy syllables. …60
According to Vutt 111 (and Sadd 8.7.2.16), the metrical structure of
the ketumat¥ metre, which belongs to the visama category where odd
and even quarters are different is as follows : 2 Ÿ   ¯  ¯    ¯ ¯ || ¯   ¯ 
´
61
¯    ¯ ¯. Five is the number of heavy syllables in the even ones. In our
verse, however, this pattern seems to be reversed. The quarters with five
heavy syllables are a and d.
The general pattern of the riddles is the same as in the Sanskrit
examples : the first two or three questions relate to any topic, but the last
´
one always gives an indication about the structural pattern of the metre
to be guessed (number of light and heavy syllables, indicated in an
indirect manner to make the matter more attractive!). The first two
58 This is the text as in B. A (III,36, line 4 to 37, line 2) reads (with some
repetitions)´: lahu ccha pañca garu bandha-vutti kå´? ketumati. tattha vutti ti
gåthå. lahu ccha pañca garu bandha ti va lahuhi ceva pañca garuhi bandhå
nåma guâ¤gavaˆâ racitå ti nåthassa guâ¤gavaˆâ racitå. la. pañca garu
bandha vutti kå´? ketumati. tattha vutti ti gåthå. lahu pañca garu bandha se
lahu ceva pañca garuhi bandha nåtha guâgavaˆâ racitå ti nåthassa
guâ aîtabbehi vaˆâhi vaˆêhi racitå ketumat¥ ti eva()-nåmakå gåthå
akatabba va .a.itabba va akåte yeva a¤gå va garu lahu ti ti una tatthå ketu
viyå ti ketu asså atthi ti ketumati. vutta-nåma-jåti vutta.
59The syntax is not clear to me.
60Not fully clear.
61visame sa-jå sa-guru-yuttå ketumat¥ same bha-ra-na-gå go (Vutt).
questions are a charade : Answer 1 + Answer 2 are components of the
´
complete word (ruci+arå > ruci’rå ; similarly ketu+mati > ketumati).

´
As could be expected, in contrast with Dharmadåsa’s riddles the

cultural references of which are Hindu mythology (and this could be a
clue to his religious affiliation), the universe of the Påli Vid is a
Buddhist universe. The questions concern the Buddha’s personality, his
physical and intellectual features. 62 The answers presuppose a
knowledge of the tradition, which is expanded in the commentary. Thus,
(ii) of Example 1, where hanti “to kill” occurs in the question and arå in
the answer (equated with kileså in the commentary), is a reference to
some of the etymologies of the word araha where two components are
distinguished : ´
arå sasåra-cakkassa hatå ñåˆåsinå yato

loka-nåthena ten’ esa arahan ti pavuccat¥ ti (quoted in Sadd 579,9-10) or
sasåra-cakkassa vå arå kileså hatå anenå ti arahå .
(Abhidhammatthasa¤gaha)
Similarly, the metrical structure of the rucirå with five heavy and
eight light syllables is also understood at a doctrinal level and connected
with the five abhiññås and the eight samåpattis. These qualities are
ascribed to the Buddha in several passages (e.g. Mahåvagga-a††hakathå
II 632 : mahåpuriso pana sabbå pi a††ha samåpattiyo, pañca abhiññåyo
´
ca nibbattetvå… ; Ja I 30,11). Finally the equation ketu/måna (Example

´
2, i) is common in traditional exegesis, where the two words are

synonyms (måno ahakåro unnati ketu paggaho avalepo ti pariyåyå,
Sadd 485 ,14 ) or where ketu-hå is explained as måna-ppahåy¥ in the
commentary on Th 64 (a stanza revolving around the manifold
meanings of ketu).
On the other hand, the genre of learned riddles such as those of
Dharmadåsa or his Påli counterpart implies a special usage of the
language where all its niceties and rarities are called for. Monosyllables
62A similar tendency can be observed in Jain riddles whether they are adapted
or not from Dharmadåsa’s work´: the personality of the Jinas is a source of the
questions asked.´See Balbir 2002.
360 Nalini Balbir
are one such extreme case. They are not used in our Påli illustrative
verses. Rare words or formations are, however, present. Påli aghika
(example 1, ii) seems to be based on agha or, at least, seems to be
understood in this way by the commentator when he equates aghika-
pajåya with dukkhita-pajåya. This equivalence is similar to
Aggavasa’s discussion of the word agha (Sadd 527,30ff.) : aghan ti
´
dukkha … agho ti kileso, tena aghena arahå anagho. The abstract

noun vidutå (in dhamma-vidutå, example 2, ii) is a secondary derivative
from a well-known compound and shows the productivity of the suffix
–tå.
In brief : for a correct appraisal of the diffusion of Påli literature, for

´
the understanding of its making and for the establishment of the texts,
the intertextual Påli (or Sanskrit) network to which a given work
belongs should not be put aside. For works combining prose with
verses, no edition should be published without the basic tools that make
it possible to assess the place and possible sources of these verses. This
is a necessary stage in the process of any critical edition, as relevant as
the consultation of a large number of manuscripts.
Nalini Balbir
REFERENCES
PRIMARY PÓLI SOURCES RELATING TO THE WORKS UNDER CONSIDERATION
Buddhaghosuppatti (Bu-up)
Buddhaghosuppatti or The Historical Romance of the Rise and Career of
Buddhaghosa, ed. and trans. by James Gray, Påli Text Society, 1892,
reprinted 2001
(The Burmese edition Buddhaghosuppatti på†h nisya kyam:, Rankun´: Praññ
kr¥: mit chve pi†akat Press, 1270 B.E./1908 A.D., mentioned in Braun
and Myint 1985´: no. 50 could not be consulted.)
Jinåla¤kåra (Jinål)
Jinâla¤kâra or “Embellishments of Buddha” by Buddharakkhita, ed., with
introduction, notes, and trans. by James Gray, Påli Text Society,
London, 1884, reprinted 1981 (Sacred Books of the Buddhists, Vol.
XXXVI)
Jinåla¤kåra†¥kå på†h, Rangoon´: Sudhammavat¥ Press, 1302 = 1940
(The Sinhalese edition mentioned in von Hinüber 1996 §´406 could not be
consulted.)
Pa†hamasambodhi (Pa†h, references are to verse numbers in Ee)
E e = The Pa†hamasambodhi, ed. by George Cœdès, edition prepared by
Jacqueline Filliozat, Oxford´: Påli Text Society, 2003
Se = Pa†hamasambodhi composed by Somdet Phra Mahåramaâ Chao Krom
Phra Paramånujita Jinarasa, transcribed from palm-leaf manuscript in
Khom script, rst printing, Bangkok, B.E. 2537
Transcription of the Pa†hamasambodhivitthara typed by Nai Dad Puñña
Bibhida in 1925 made for Sylvain Lévi by order of Son Altesse Royale le
Prince de Chandaburi, 408 pages (original kept in the Library of the
Institut d’Etudes Indiennes du Collège de France, Paris´; shelfmark
BUD.PAL.P I.1a and I.1b).
Vidaddhamukhamaˆ ana (Vid)
Manuscripts´: see above, section 3.
The Sanskrit version has been published very often. See for instance´:
Vidagdhamukhamaˆ anam of Dharmadåsa SËri, with ‘Chandrakalå’
Commentary and Hindi translation, ed. and tr. by Óchårya Íe¡aråja
Íarmå, Varanasi, Chaukhambha Orientalia, 1984 (Chaukhambha
Prachyavidya Series No. 7). See below, Kraatz 1968.
362 Nalini Balbir
SECONDARY SOURCES
Balbir, Nalini, 2001. “À propos d’une Vie de Buddhaghosa” in Bulletin
d’Etudes Indiennes 19, pp. 342–52
——— 2002. “Théorie et pratique de la devinette en milieu jaina´: I. Les Cent
soixante et une devinettes de Jinavallabha´; II. Devinettes en contexte”,
Bulletin d'Etudes Indiennes 20.2, pp. 83–243
——— 2004. “Grammatical Riddles from Jain works” in JambË-jyoti
(Munivara JambËvijaya Festschrift), edited by M.A. Dhaky,
J.B.´Shah, and Shresthi Kasturbhai Lalbhai Smarak Nidhi, Sharadaben
Chimanbhai Educational Research Centre, pp. 269–309
Bechert, Heinz, 1979. Burmese Manuscripts, Part 1, compiled by Heinz
Bechert, Daw Khin Khin Su, and Daw Tin Tin Myint, Wiesbaden´:
F.´Steiner (Verzeichnis der Or. Handschriften in Deutschland, Band
XXIII, 1)
——— and Heinz Braun, 1981. Påli N¥ti Texts of Burma, Critical Edition and
Study by Heinz Bechert and Heinz Braun, London´: Påli Text Society,
1981 (Text Series No. 171)
Bizot, F., 1980. “La grotte de la naissance”, Bulletin de l'Ecole française
d'Extrême-Orient 67, pp. 222–73
Bode, M., 1909. The Påli Literature of Burma, London
Braun, Heinz, and Daw Tin Tin Myint, 1985. Burmese Manuscripts, Part 2,
Stuttgart, 1985 (VOHD XXIII,2)
Cœdès, George, 1912. “Rapport sur une mission d’études philologiques et
archéologiques au Cambodge”, Bulletin de l’Ecole française
d’Extrême-Orient 12, 1912, pp. 176–79
——— 1916A. “À propos d’une stèle sculptée d’Angkor Vat”, Mémoires con-
cernant l’Asie orientale, II (1916), pp. 117–22
——— 1916B. “Note sur les ouvrages palis composés en pays thai”, Bulletin de
l’Ecole française d’Extrême-Orient 15, 3, 1916, pp. 1–21
——— 1968 . “Une vie indochinoise du Buddha´: la Pa†hamasambodhi”,
Mélanges d’indianisme à la mémoire de Louis Renou, Paris´: Publi-
cations de l’Institut de Civilisation Indienne 1968´: 217–27 (English
translation in the PTS ed. of the Pa†hamasambodhi, pp. lv–lxvii)
Finot, Louis, 1917. “Recherches sur la littérature laotienne”, Bulletin de l’Ecole
française d’Extrême-Orient 17, pp. 1–219
——— 1921. “La légende de Buddhaghosa”, in Cinquantenaire de l’Ecole
Pratique des Hautes Etudes, Paris, 1921, pp. 101–19
Gandhavasa, ed. by J. Minayeff, Journal of the Påli Text Society II (1886),
pp. 54–80
Granoff, Phyllis, 1991. “Buddhagho¡a’s Penance and Siddhasena’s Crime´:
Remarks on Some Buddhist and Jain Attitudes Towards the Language
of Religious Texts”, in From Benares to Beijing´: Essays on Buddhism
and Chinese Religion, ed. by Koichi Shinohara and Gregory Schopen,
Oakville, N.Y.´: Mosaic Press´: 17–33
von Hinüber, Oskar, 1996. A Handbook of Påli Literature, Berlin´: Walter de
Gruyter (Indian Philology and South Asian Studies, Vol. 2)
Hundius, H., 1990. “The colophons of 30 Påli manuscripts from Northern
Thailand”, Journal of the Påli Text Society XIV, pp. 1–173
Kraatz, Martin, 1968. Das Vidagdhamukhamaˆ ana des Dharmadåsa (ein
Lehrbuch der Rätselkunde 1. und 2. Kapitel), Teil 1 Einleitung, Text
und Übersetzung´; Teil 2 Kommentare, Anmerkungen. Inaugural-
dissertation zur Erlangung der Doktorwürde der Philosophischen
Fakultät der Philipps-Universität zu Marburg, Marburg/Lahn, 1968
Law, Bimala Charan, 1923. The Life and Work of Buddhaghosa, with a
Foreword by Mrs C.A.F. Rhys Davids, Calcutta and Simla (Calcutta
Oriental Series No. 9.E.3)
Luce, G.H., and Tin Htway, 1976. “A 15th Century Inscription and Library at
Pagán, Burma” in Malalasekera Commemoration Volume, ed. by O.H.
De A. Wijesekera, Colombo, 1976, pp. 203–56
Minayeff, I.P., 1886. See Gandhavasa
Norman, K.R., 1983. Påli Literature Including the Canonical Literature in
Prakrit and Sanskrit of All the H¥nayåna Schools of Buddhism, Otto
Harrassowitz, Wiesbaden
——— 1994. “ The Present State of Påli Studies, and Future Tasks”, in
Memoirs of the ChËØ Academic Research Institute, No. 23, Dec. 1994´:
1–19 = Collected Papers, Vol. VI, Oxford´: Påli Text Society, 1996,
pp. 68–87
Saddan¥ti ´: La grammaire palie d’Aggavasa, ed. by Helmer Smith, Lund,
1928–1966
364 Nalini Balbir
Schober, Julian, ed., 1997. Sacred Biography in the Buddhist Traditions of
South and Southeast Asia, University of Hawai’i Press 1997 ´; 1st
Indian edition, Delhi, Motilal Banarsidass, 2002
Skilling, Peter, forthcoming. “`Painted Tipi†aka’ from Wat Thong Noppakhun”
——— and Santi Pakdeekham, 2002. Påli Literature Transmitted in Central
Siam, Bangkok
——— and Santi Pakdeekham, 2004. Påli and Vernacular Literature Trans-
mitted in Central and Northern Siam, Bangkok
Taylor, James L., 1997. “The Textualization of a Monastic Tradition´: Forest
Monks, Lineage, and the Biographical Process in Thailand” in
Schober 1997, pp. 289–309
Than Tun, 1960. “The influence of occultism in Burmese history with special
reference to Bodawpaya’s reign 1782–1819”, in Bulletin of the Burma
Historical Commission Vol. I, Part I, pp. 117–45
Vutt = Sagharakkhita’s Vuttodaya. A Study of Påli Metre. Påli text and trans-
lated into English by R. Siddharatha, with a preface by J.W. de Jong,
Delhi´: Sri Satguru Publications
What’s in a Repetition ?
On Counting the Suttas of the Saµyutta-nikåya*
1. Introduction
One of the stylistic features of ancient Indian Buddhist texts is their
repetitiveness. Of course, other ancient Indian literatures display some
of the same repetitive devices, yet it seems that none develops the art of
repetition quite to the extent that Buddhist texts do (cf. Allon 1997,
p. 360). While this stylistic feature has been frequently noted, as Allon
comments, it “has never been satisfactorily analysed or quantified”
(1997, p. 273). Certainly Mark Allon’s own 1997 study of the function
of certain stylistic features in Påli texts (the product of doctoral research
carried out in Cambridge under the supervision of K.R. Norman) makes
an important contribution to our understanding of the nature of
repetition in early Buddhist literature, but his study was not intended as
exhaustive and more remains to be said.
In his analysis of repetition in the Udumbarikas¥hanåda-sutta (D III
36–57), Allon calculates that 30% of the full text can be classified as
“verbatim repetition”, while 86.8% can be classified as repetition of one
sort or another (pp. 358–59). He distinguishes five types of repetition :
verbatim, repetition with minor modifications, repetition with important
modifications, repetition of structure types 1 and 2 (p. 287). While the
five different types are important for his calculations, in the present
context I shall collapse Allon’s first three categories into what might be
called “narrative repetition” and his last two into “structural repetition”.
By “narrative repetition” I refer to repetition of blocks, with or
without modification, in the course of a narrative. Thus a text may
describe events relating to person A who then describes these events in
full to person B who then in turn relates to them to person C in full who
then meets person A and asks, describing the events in full yet again,
*I am grateful to Peter Jackson for his observations on a first draft of this paper.
366 Rupert Gethin
whether they are true.1 By “structural repetition” I refers to the practice
of providing a framework structure which can then be used as the basis
for a series of repetitions by substituting different items and/or
modifying the frame. For example, the Ga!gå-peyyåla of the Saµyutta-
nikåya uses the following frame : “Just as the river Ganges flows to the
east, so a bhikkhu who develops the noble eightfold path resorting to
seclusion flows to nirvana.” By substituting different rivers for the
Ganges, different items for the noble eightfold path, “great ocean” for
“east”, different expressions for “resorting to seclusion”, a whole series
of repetitions are achieved (S V 38–41). Such repetitions are especially
characteristic of the Saµyutta- and A!guttara-nikåyas and also the
canonical Abhidhamma texts.
Both kinds of repetition are routinely abbreviated in the
manuscripts and printed editions by the use of the term peyyåla, itself
usually abbreviated to pe or la. The use of abbreviation in this
connection poses something of a problem for the full analysis of
repetition in Påli texts, since it is not always clear precisely what is to
be repeated. In the present paper, offered on the occasion of the 125th
anniversary of the founding of the Pali Text Society in 1881 and K.R.
Norman’s 80th birthday in 2005, I should like to focus on the use of
structural repetition in the Saµyutta-nikåya, considering in the first
place its extent and in the concluding section its possible significance
and function.
2. Counting the suttas of the Saµyutta-nikåya
With reference to the 56 vaggas that make up the Saµyutta-nikåya,
K.R. Norman observes that “[t]hey contain 2,889 suttas in all, in the
European edition, although Buddhaghosa states there are 7,762 suttas”
1As K.R. Norman (2006, pp. 70–71) has pointed out, this kind of repetition is
well exemplified by the opening of the AlagaddËpama-sutta (M I 130–31); this
describes how Ari††ha is beset by a pernicious view, how bhikkhus hear of this
and proceed to ask Ari††ha if it is true, how Ari††ha confirms it is true, how the
monks reprimand Ari††ha and then report to the Buddha, relating everything in
full to him.
What’s in a Repetition ? 367
(1983, p. 50). This discrepancy between the European edition and
Buddhaghosa is worth pondering. Buddhaghosa also gives figures for
the number of suttas in the other Nikåyas : 34 for the D¥gha-nikåya, 152
for the Majjhima-nikåya and 9,557 for the A!guttara-nikåya.2 The fact
that the figures Buddhaghosa gives correspond to the number of suttas
found in modern European editions in the cases of the D¥gha-nikåya and
Majjhima-nikåya but are wildly out of line in the cases of the Saµyutta-
nikåya and A!guttara-nikåya (the European edition of the latter counts
between 2,308 and 2,363)3 should give us pause for thought. Buddha-
ghosa’s figures do not seem intended as vague big numbers — like, say,
84,000 — but as a precise count, so either the tradition he reports was
talking about a very different text from the one that has come down to
us, or it counted suttas in a very different way. In fact it is clear from the
introductions to their editions that both Feer and Hardy struggled with
how to present the Saµyutta-nikåya and A!guttara-nikåya and that a
significant issue was the problem of repetition and what to count as a
single sutta. Feer claims that by counting the suttas of the Saµyutta-
nikåya in a different way “the sum of 7,762 can be attained, but not be
got from the data of the MSS” (S V ix). Yet his claim that he “counted
the suttas according to the Uddånas” is problematic,4 because, as we
shall see, in the first place the uddånas are not always clear on numbers
and in the second place he seems on occasion to ignore — or at least
interpret in a conservative way — the uddånas’ instructions to expand.5
2Sp 18 = Sv I 17 = As 18 (cf. Spk I 2). The Chinese translation of Sp gives the

number of suttas for D as 44 (possible variant noted), for M as 252, but the
numbers for S and A are as in the Påli Sp. See Bapat and Hirakawa 1970,
pp. 10–11.
3See Norman 1983, p. 54.
4Elsewhere Feer seems in fact to favour counting larger numbers in certain
instances S IV xii : “But if we count 247 suttas in the Sa¬åyatana and 1,463 in
Asa!khata, — what the text seems to permit — if not require, — this total
would amount to 1,850 suttas.”
5For example the uddåna at S II 133 is explicit that 132 suttas should be
counted.
368 Rupert Gethin
The same problem has troubled these texts’ translators. For the most
part C.A.F. Rhys Davids and Woodward followed Feer’s lead, though
correcting some obvious slips. In the introduction to his recent
translation Bhikkhu Bodhi makes some attempt to address the problem
of the number of suttas in the Saµyutta-nikåya, providing tables of
Feer’s and his own count, and suggesting that since Buddhaghosa’s
Såratthappakåsin¥ comments on a text that seems to correspond to what
we have, “the difference in totals must certainly stem merely from the
different ways of expanding the vaggas treated elliptically in the text”,
although he still finds it “difficult to see how the commentator could
arrive at so large a figure” (2000, p. 26).
The “problem” of repetition seems to have two facets. The first is
that, as the editors point out, the manuscripts they had before them were
inconsistent, using different ways of presenting an abbreviated text,
though it is not exactly clear that this meant different numbers of
repetitions were evidenced in the manuscripts. The second facet of the
problem is that editors seem to have found the repetitions “tiresome”, so
much so that they were predisposed to play down the numbers of suttas
implied by the repetitions.6 Certainly it seems worth trying to establish
whether it is possible on the basis of the text of the Saµyutta-nikåya
that has come down to us to arrive at the number of suttas Buddhaghosa
counted. It also seems worth pondering further the question of why all
these “tiresome” repetitions.
Ideally the question of counting the suttas of the Saµyutta-nikåya
should be addressed by going back to representative manuscripts. In the
present context I shall confine myself to carrying out a preliminary
study on the basis of a selection of modern printed editions : the five
6So Feer at S V v–vi comments, “The tiresome repetitions, peculiar to the

buddhist scriptures, abound exceedingly in the Mahå-Vaggo, and form so great
a proportion in several of its Saµyuttas that important abridgments are
required. The singhalese and burmese MSS. differ so much in the manner and
quantity of their abbreviation that they seem to have nothing in common,
although they are dealing with the same subject.” On the issue of the early
European tendency to abbreviate Påli texts, see also Norman 2006, p. 113.
e
volumes of Feer’s PTS edition of 1884–1898 (E ) ; the five volumes of
the Syåmara††ha-tepi†aka of 1927 (BE 2470) (Se) ; the three volumes of
the Cha††hasa!g¥ti-pi†aka of 1957 (Be) ; the six volumes of the Buddha-
jayanti-tripi†aka of 1960–83 (Ce).7 Of course, this is not ideal since all
these Asian editions may have been influenced to some extent by Feer’s
European edition.8
As I have already indicated, Feer gives his count of the total
number of suttas in the introductions to each volume of his edition ;
unfortunately, for the most part Se counts paragraphs or sections rather
than suttas, so does not make explicit how many suttas it recognizes,
though the edition is still useful for comparing the number of repetitions
understood in the text. Both Be and Ce give a running count of suttas for
each of the five vaggas of Saµyutta-nikåya. Bhikkhu Bodhi also offers
a count in the introduction to his translation. The various enumerations
of suttas are set out in Table 1.
Table 1. Numbers of suttas counted in the Saµyutta-nikåya
vagga suttas
Ee Feer Be CS Ce BJT Bodhi
Sagåtha 271 271 271 271
Nidåna 286 246 406 286
Khandha 733 716 716 716
Sa¬åyåtana 391 420 2286 434
Mahå 1208 1201 3977 1197
2889 2854 7656 2904
7Ihave had access to Se and Be in both the printed editions and also the digital
editions in the form of the BUDSIR (Bangkok: Mahidol University, 1994,
1996) and “Cha††ha Sa!gåyana” (Igatpuri : Vipassana Research Institute, 1999)
CD-ROMs respectively. Unfortunately I have only had direct access to the
digital edition of Ce (Sri Lanka Tripitaka Project, Colombo ; www.
buddhistethics.org/palicanon.html), though I am grateful to Peter Jackson for
supplying me with some details directly from the printed edition.
8The Syåmara††ha edition has been reprinted with the addition of at least some
variants in 1956 (BE 2499), 1979 (BE 2522), 1995 (BE 2538). I have used the
1995 reprint; how far this differs from the original is unclear.
370 Rupert Gethin
In each vagga, except the Sagåtha-vagga where the counting of suttas
seems unproblematic, there is some variation ; particularly in the
Sa¬åyåtana- and Mahå-vaggas the discrepancies are considerable.
Tables 2–5 show the differences in detail for each vagga. The figures
which appear initially discrepant are highlighted in bold. These
discrepant figures allow us to identify places where it seems likely
different methods of counting are in operation. In tables 2–5 I have
added a column giving my own count of suttas.
The discrepancy in the nidåna-saµyutta turns out to be precisely
connected with a repetition section that closes the saµyutta, the antarå-
peyyåla (S II 130–33). This peyyåla applies a structure based on the
four truths to each of eleven links of the formula of dependent arising in
turn (avijjå is omitted) : someone who does not know or see old age and
death, etc., their arising, their ceasing, and the path leading to their
ceasing as they truly are should seek the Teacher in order to know them
as they truly are. This gives eleven suttas.9 The peyyåla section then
gives a further eleven alternatives to seeking the Teacher that someone
who does not know or see should do in order to know and see. This
gives a total of (11 Ÿ 12 =) 132 repetitions or suttas acknowledged in
the uddåna.10 This gives Ce’s total of 213 for the saµyutta.11 In fact, all
editions recognize the same number of repetitions, but in Be these are
counted as just one, and by Feer and Bodhi as 12. In the preceding
samaâbråhamaâ-vagga where Feer, Ce and Bodhi count 11, Be treats
9S II 130,28–29 makes it clear at the end of the initial treatment of jaråmaraâ

that someone at some point in the history of the texts regarded this as a sutta:
suttanto eko. sabbesaµ evaµ peyyålo.
10The uddåna has a number of variants in the manuscripts and printed editions :
S II 133,5 talks of suttå dvattiµsasatåni, presumably to be construed as “suttas

numbering thirty-two and a hundred”, while the variant Feer records from his
Sinhalese manuscripts has antara-peyyålassa suttantå ekasatañ ca dvattiµsa
bhavanti.
11Strictly Ce seems not to recognize a nidåna-saµyutta, but counts it as part of
the abhisamaya-saµyutta.
a similar application of a formula to each of the same eleven links as
two and thus reaches a total of only 73 for the nidåna-saµyutta.
In the låbhasakkåra-saµyutta and Råhula-saµyutta Be in fact
counts the same number of suttas in each of the four (10 + 10 + 10 + 13
= 43) and two vaggas (10 + 12 = 22) that make up these saµyuttas, but
the running total of suttas for the whole Nidåna-vagga anomalously
counts eight abbreviated suttas as one at S (Be) I 430,18–19, six as one at
S (Be) I 438,1–2, and a further eight as one at S (Be) I 443,14–15.
In the khandha-saµyutta Feer’s edition simply omits a sutta which
should have dukkhånupass¥ vihareyya for the aniccånupass¥ vihareyya
of sutta 147 (S II 179).
The arrangement of the di††hi-saµyutta is problematic ; see Feer at
S III ix–x and Bodhi 2000, pp. 1097–98 (n. 264). Since there are in toto
26 views and four different frames, one would expect 104 as the total
number of repetitions, but the initial frame appears to be only applied to
18 views, so we have 18 + (26 Ÿ 3) = 96. Feer suggests, somewhat
anomalously, counting 114.
In the sa¬åyatana-saµyutta the main problem is the sa††hi-peyyåla
(S IV 148–56). Since this peyyåla seems to upset an implied structure
for the whole saµyutta of four sets of fifty suttas (paññåsaka), each
comprising five vaggas, Feer asked : “Ought not this peyyåla to be
lessened ? I thought so.” (S IV viii) Notwithstanding its name, he
suggests reducing this peyyåla to 20 by not treating certain repetitions
as qualifying as suttas.
At S IV 126–28 Feer counts only one sutta, but Be, Ce, and Bodhi
count two : the first with verses, the second precisely the same without
verses. This seems unusual and Feer may well be right in counting only
11 suttas in this vagga rather than 12.12
In the final vagga of the vedanå-saµyutta Feer counts only 9 where
B , Ce and Bodhi count 11, understanding new suttas to begin at S IV
e
12Theuddåna as given by Feer at S IV 132 reads : agayha dve honti palåsinå,

and Feer presumably takes the dve as applying only to palåsinå.
372 Rupert Gethin
233,25 and at S IV 235,21. The uddåna at S IV 238 might be construed
in either way, but the latter seems more likely to me.
In the two peyyåla-vaggas of the måtugåma-saµyutta, Ce repeats
the formula pañcahi kho Anuruddha dhammehi samannågato …
nirayaµ upapajjat¥ ti containing kodhano ca hoti (S IV 240,25–241,2)
twice, thus creating an extra sutta. And later it counts what is clearly an
introductory paragraph (S IV 243,16–24) as a separate sutta. Its count of
36 for this saµyutta is thus a clear error.
The Såmaˆ"aka-saµyutta is a straightforward repetition of the 16
suttas of the immediately preceding Jambukhådaka-saµyutta sub-
stituting Såmaˆ"ako paribbåko for Jambukhådako paribbåko through-
out ; Be gives only the first and last sutta separated by the comment
yathå Jambukhådakasaµyuttaµ tathå vitthåretabbaµ (S (Be) II 455,20),
and counts only 2 suttas although it recognizes the repetition of all 16.
The difference in the count for the Moggallåna-saµyutta is more
complex and concerns what in Ee and Be are counted suttas 10 (S IV
269–280) and 11 (S IV 280). The former initially describes how Sakka
accompanied by 500 devas approaches Moggallåna and they both agree
that going for refuge in the Buddha, Dhamma and Sa!gha is a good
thing since as a result some beings are reborn in heaven (S IV 269,21–
270,24). What follows is abbreviated with pe but indicates that the
preceding section should be repeated a further four times in full with
Sakka approaching with, in turn, 600, 700, 800, and 80,000 devas (S IV
270,25–271,19).13 The second section repeats all this — in effect five
suttas — in full but this time Sakka and Moggallåna agree that the good
thing is having trust in the Buddha, Dhamma and Sa!gha that is based
in understanding (avecca-pasåda) (S IV 271,21–274,27). The third
section once again provides five further repetitions by returning to the
theme of going for refuge as the good thing, but adding that the beings
reborn in heaven surpass other devas in ten respects (S IV 274,29–
276,31). A fourth section gives five more repetitions by combining the
13AsBodhi 2000, p. 1440 (n. 282) notes, Ee in fact has as¥tiyå devatåsatehi but
other editions have as¥tiyå devatåsahassehi.
trust based in understanding (section two) with the ten respects in which
beings surpass devas (S IV 276,33–280,19). To this point we have thus
had twenty repetitions. What is counted as sutta 11 indicates that Sakka
is to be replaced by the names of five further devas (Candana, Suyåma,
Santusita, Sunimmita, Vasavatti) followed by the instruction ime pañca
peyyålå yathå Sakko devånam indo tathå vittharetabbån¥ ti. This gives
five further sets of 20 repetitions and a total for this saµyutta of 129
suttas — 9 + (20 Ÿ 6).
The asa!khata-saµyutta (S IV 359–73) begins with a sutta setting
out the “unconditioned” (asa!khata) and “the path leading to the
unconditioned” (asa!khatagåmi-magga). The latter is explained as
kåyagatå-sati. This is followed by a second sutta identical in every
respect expect that the path is this time explained as samatha and
vipassanå. The same structure is then repeated with a further nine
explanations of the path, and thus a total of eleven suttas (S IV 359–61).
This concludes the first vagga. Explanations 2–11 are in the form of
numerically increasing sets of items : samatha and vipassanå ; three
kinds of samådhi, a further three kinds of samådhi, four satipa††hånas,
four sammappadhånas, four iddhipådas, five indriyas, five balas, seven
bojjha!gas, the eightfold path. The second vagga now proceeds by
using the same framework but explaining “the path leading to the
unconditioned” as each individual item from each of these ten sets in
turn, giving a total of 45 suttas (2 + 3 + 3 + 4 + 4 + 4 + 5 + 5 + 7 + 8).
We have now had a total of 56 (11 + 45) suttas, although Feer
arbitrarily counts the second vagga as only a single sutta. The third
vagga replaces asa!khata and asa!khatagåmi-magga with anta and
antagåmi-magga. Feer’s PTS edition gives — or rather suggests — in
radically abbreviated form a further set of 45 suttas. These are followed
by 31 further sets of 45 suttas achieved by replacing the original
asa!khata by 31 different terms. Feer’s edition thus implicitly
recognizes a total of 1,496 suttas for the saµyutta — 11 + (45 Ÿ 33) —
374 Rupert Gethin
although he himself prefers to count only 44 (11 + 33).14 The oriental
editions of this saµyutta seem to understand things differently. The
Siamese Royal Edition states of anta and the final term paråyana : yathå
asa!khataµ vitthåritaµ tathå vitthåretabbaµ.15 This suggests that we
should in fact understand the saµyutta as containing a total of 1,848
suttas : (11 + 45) Ÿ 33. The Burmese Cha††hasa!g¥ti and Sinhalese
Buddha-jayanti-tripi†aka seem to understand the text similarly.16
The largest number of discrepancies in the counting of suttas in the
different editions is found in the Mahå-vagga. The first saµyutta — the
magga-saµyutta — ends with a series of nine vaggas, five of which are
explicitly referred to in the manuscripts as peyyålas, that almost entirely
consist of repetitions once more indicated by the term pe or la in the
manuscripts.
The aññatitthiya-peyyåla gives a series of eight items for the sake
of which the spiritual life is lived. In each case it is further explained
that the way or path to reach the aim of the spiritual life is the noble
eightfold path. This gives a total of eight radically abbreviated suttas.
The suriya-peyyåla gives a series of seven items which prelude the
arising of the noble eightfold path just as the dawn preludes the arising
of the sun. In each case it is further explained that it is to be expected of
a bhikkhu who is accomplished in the particular item that he will
14Woodward 1927 and Bodhi 2000 follow Feer in counting 44. Feer, however,
then seems to get misled by his own method of counting and so at S IV x–xi
claims the second vagga comprises only 44 alternative “paths leading to the
unconditioned” instead of the actual 45, which leads him to conclude that the
total number of suttas can be counted as either 44 or 1,463 (11 + (44 Ÿ 33)).
This error is repeated by Wynne (2004, p. 107, n. 24). Collins (1998, pp. 199–
200) suggests a different enumeration for this saµyutta: 1,485 (45 Ÿ 33 —
although he states 32) or 1518 (46 Ÿ 33).
15S (Se) IV 450, 453.
16S (Be) II 541, 543 : (Ce ) IV 656, 666: yathå asa!khataµ tathå vitthåretabbaµ.
The numbering in Ce also makes explicit that the editors understood the
repetition of a full set of 56 suttas for each of 33 items. Skilling (1994,
pp. 79–81) also concludes that this saµyutta comprises 1,848 suttas.
develop the noble eightfold path. This is followed by a statement of
how the bhikkhu develops the eightfold path : he develops each
constituent of the path with reference to two different formulas : the
vivekanissita and rågavinaya formulas. This then gives us a total of
fourteen (7 Ÿ 2) abbreviated suttas.
The ekadhamma-peyyåla I and ekadhamma-peyyåla II take the
same seven items used in the previous vagga and state how each
represents one quality in particular suited to the arising of the noble
eightfold path (ekadhamma-peyyåla I) or how the Buddha sees no other
single quality which leads to the arising and full development of the
noble eightfold path (ekadhamma-peyyåla II). The two vaggas then
follow the pattern of the suriya-peyyåla. This gives two further sets of
fourteen suttas.
The Ga!gå-peyyåla describes how just as five separate rivers and
then all five rivers together flow (1) to the east and (2) to the great
ocean so the bhikkhu who develops the noble eightfold path flows to
nibbåna. This gives an initial set of twelve suttas. But as in the Suriya-
and ekadhamma-peyyålas, each sutta incorporates a statement of how
the bhikkhu develops the eightfold path : but here he develops each
constituent of the path with reference to four (not two) different
formulas : the vivekanissita, rågavinaya, amatogadha and nibbånaninna
formulas. This then gives the peyyåla a total of 48 suttas (6 Ÿ 2 Ÿ 4).
The appamåda-vagga gives a set of ten different similes for the
way in which wholesome qualities are rooted in heedfulness
(appamåda). In each case it is further explained that it is to be expected
of a bhikkhu who is heedful that he will develop the noble eightfold
path. This is followed by a statement of how the bhikkhu develops the
eightfold path : he develops each constituent of the path with reference
to four (not two) different formulas : the vivekanissita, rågavinaya,
amatogadha and nibbånaninna formulas. This then gives the vagga a
total of 40 suttas (10 Ÿ 4).
The balakaraˆ¥ya-vagga gives a set of twelve different similes
relating to the way in a bhikkhu develops the noble eightfold path. As in
376 Rupert Gethin
the appamåda-vagga, this is followed by a statement of how the
bhikkhu develops the eightfold path : he develops each constituent of the
path with reference to the same four formulas : the vivekanissita,
rågavinaya, amatogadha and nibbånaninna formulas, though Feer,
mistakenly in my view, questions whether all four formulas should
apply here.17 So on the assumption that they should, this gives the
vagga a total of 48 suttas (12 Ÿ 4).
The esana-vagga gives 10 — or 11 if the final repetition based on
tasinå is treated as a distinct repetition from that based on the preceding
taˆhå, which I suspect it should not be18 — items for the direct
knowledge (abhiññå) of which the eightfold path is developed. Once
17At the end of the first sutta of this vagga Feer’s PTS edition states para-
ga!gåpeyyål¥vaˆîyato paripuˆâsuttan ti vitthåramagg¥. Feer notes (p. 46,
n. 3): “This phrase is to be found in the burmese MSS. which add, according
to the preceding case, the three statements referring to 1. råga-dosa-moha ; 2.
amata ; 3. nibbåna. — Nothing of this appears in the singhalese MSS.
Therefore I bound myself to this note upon this matter.” However the same
phrase appears in the Syåmara††ha edition at S (Se) V 68, which then proceeds
to repeat the sutta with the additional three formulas : the Cha††hasa!g¥ti does
the same at S (Be) III 42–43, while BJT simply gives all four formulas in full.
Woodward (1930) does not translate the concluding phrase and simply passes
over the question of whether the sutta is to be repeated with all four formulas ;
Bhikkhu Bodhi (2000, p. 1553), however, notes that each of the twelve suttas
of the vagga is to be expanded by way of the four formulas, though he does
not count each as a separate sutta in his numbering.
18This explains the extra sutta counted by Be for the magga-saµyutta when
compared with Ee and Bodhi’s translation ; while both the latter include the
tasinå repetition they do not number it separately (see Bodhi 2000, p. 1898,
n. 46). It also explains similar discrepancies in some of the other saµyuttas of
the Mahå-vagga. The word tasinå (or tasiˆå) is, of course, simply another
Prakrit form, alongside taˆhå, of Sanskrit t®#ˆå, showing svarabhakti rather
than assimilation of the consonant group (cf. Geiger & Norman 1994, § 30.3).
This alternative form is extremely rare, however, such that it would seem
appropriate to regard it as anomalous in Påli. In the present context tasinå is
not included in Se and Ce, while Ee (S V 58, n. 1) notes that it is not found in
the Sinhalese manuscripts. Electronic searches of Ee, Ce, Se and Be give no
other occurrences of the form tasinå, while the form tasiˆå appears at Dhp
342–43, Nidd I 488 (v.l. and other editions, tasitå), and Nidd II 221.
again it is explained that the bhikkhu develops each constituent of the
path with reference to the vivekanissita, rågavinaya, amatogadha and
nibbånaninna formulas. A further set of repetitions is then obtained by
substituting thorough knowledge (pariññå), destruction (parikkhaya)
and abandoning (pahåna) for abhiññå. This gives the vagga a total of
160 suttas (10 Ÿ 4 Ÿ 4).
The ogha-vagga exactly repeats the pattern of the esana-vagga by
giving a further set of 10 items for the direct knowledge, thorough
knowledge, destruction, and abandoning of which the eightfold path is
developed. The vagga thus again contains a total of 160 suttas (10 Ÿ 4 Ÿ
4).
These nine peyyålas/vaggas of the magga-saµyutta thus contain a
total of 506 suttas. The figure of 506 repetitions is not in doubt (apart
from the issues with the esana- and balakaraˆ¥ya-vaggas noted above) :
it is simply that Feer and the Mahå-vagga’s two English translators have
chosen somewhat arbitrarily not to count each repetition as a sutta in its
own right. The BJT Ce edition, however, makes its total number of
suttas for the magga-saµyutta explicit : 546. And while the Syåmara††ha
edition does not give a running total for suttas, it indicates the beginning
of repetitions with the expression Såvatth¥nidånaµ,19 making clear that
it is treating each as a sutta. Moreover, as we shall discuss presently, it
is only by counting such repetitions as suttas in their own right that we
can arrive at something like the figure Buddhaghosa gives for the
number of suttas contained in the Saµyutta-nikåya. In other words,
there must be a long tradition of treating such formulaic repetitions as
suttas.
The last five of the above nine peyyålas/vaggas (comprising 456
repetitions in the magga-saµyutta) occur again in a further seven
saµyuttas of the Mahå-vagga, substituting in each case for the eightfold
path the set of items that constitute the subject of the saµyutta : the
seven bojjha!gas, the four satipa††hånas, the five indriyas, the four
19Although this expression itself gets lost in the abbreviations and does not
occur 506 times.
378 Rupert Gethin
sammappadhånas, the five balas, the four iddhipådas, and the four
jhånas. In the case of the sammappadhånas, the balas and the jhånas,
this set of five peyyålas/vaggas in fact constitutes the entire saµyutta.
However, rather than allowing a full set of 456 repetitions in the
contexts of these seven saµyuttas, Feer’s edition (followed by the
English translations) seems to suggest a reduction in the number of
repetitions. That Feer wants to limit the number of repetitions is clear
from the figures he gives in the table in the introduction to his edition (S
V v). Yet it is not clear from the text presented by Feer himself that
such a reduction in repetitions is warranted.
Feer’s edition is based on rather limited materials, just four
manuscripts, two in Sinhala script and two in Burmese ; one of the
Sinhala manuscripts had three missing sheets, while one of the Burmese
he describes as “unfortunately very deficient in this part, as many sheets
are wanting” (S V vii). It is also difficult to follow in the abbreviated
sections, perhaps reflecting inconsistencies in the manner of
presentation of the abbreviations in his manuscripts.
In the case of the bojjha!gas, indriyas and balas, Feer concludes
that only the vivekanissita and rågavinaya formulas apply (omitting the
amatogadha and nibbånaninna formulas), which effectively reduces the
number of repetitions by half from 456 to 228. Feer’s conclusion is
apparently based on the fact that his manuscripts only make explicit that
these two formulas apply. In the case of the satipa††hånas, sammappa-
dhånas, iddhipådas, and jhånas, Feer’s text omits all four formulas
(vivekanissita, rågavinaya, amatogadha and nibbånaninna), which
effectively reduces the number of repetitions by three quarters to 114.
Feer’s conclusion is apparently based on the fact that his manuscripts
fail to make explicit that any of these formulas apply — if they do apply
they are lost in abbreviation.
Nevertheless, apparently following Burmese manuscripts, the
bojjha!ga-saµyutta ends in his edition with yad api maggasaµyuttaµ
vitthåretabbaµ tad api bojjha!gasaµyuttaµ vitthåretabbaµ (S V 140),
the satipa††håna-saµyutta with yathå maggasaµyuttaµ vitthåritaµ
evaµ satipa††hånasamyuttaµ vitthåretabbaµ (S V 192), and the jhåna-
saµyutta with yathå maggasamyuttaµ evaµ jhånaµ saµyuttaµ
vitthåretabbaµ (S V 310). The Ga!gå-peyyåla of the indriya-saµyutta
concluded again in his Burmese manuscripts with yathå maggasaµyutte
evaµ bhavati indriyasaµyutte (S V 240, cf. n. 1). Notes at the end of
the indriya- and bala-saµyuttas (S V 243, n. 1 ; 253, n. 3) record that in
fact his two Sinhalese manuscripts included a reference to the two
additional formulas (amatogadha and nibbånaninna), while the ogha-
vagga of the bala-saµyutta in his Sinhalese manuscripts also had yathå
pi maggasaµyuttaµ tathå pi indriyasaµyuttaµ vitthåretabbaµ (S V
251, n. 3). In the case of the remaining saµyuttas, which Feer presents
as limited to the vivekanissita formula, we have only phrases such as
Ga!gapeyyåla [sic] satipa††hånavasena vitthåretabbaµ (S V 190),
sammappadhånasaµyuttassa Ga!gåpeyyål¥ sammappadhånavasena
vitthåretabbå (S V 245), Ga!gåpeyyali iddhipådavåsena vitthåre-
tabbaµ (S V 291) — phrases which would seem to leave the question
of whether or not all four formulas apply at least open. These various
phrases are, incidentally, omitted by the Mahå-vagga’s English
translators.
In sum, the manuscript evidence as presented by Feer would seem
in fact capable of being interpreted differently, and might be taken as
suggesting that in every case the full 456 repetitions are to be
understood. Moreover, as a general rule in Påli texts, where we find
abbreviations, we would expect to refer back to the place where the
unabbreviated text first occurred in full, in this case the relevant
peyyålas/vaggas of the magga-saµyutta.
Turning to the modern Asian editions, however, there is some
confusion and inconsistency on this issue. Like Feer, both Se and Be
generally make only the application of the vivekanissita and rågavinaya
sets of repetitions explicit in the case of the bojjha!gas, indriyas and
balas. Yet they both contain anomalies. At the equivalent of S (Ee) V
137,8, both Se and Be seem to indicate that all four formulas should
380 Rupert Gethin
apply to the bojjha!gas.20 The numbering of suttas in BJT Ce makes
clear that it understands all four formulas should apply in all cases.
It is also worth noting that the amatogadha formula is anyway
applied to the indriyas at S V 220–23, 232–33, while the nibbånaninna-
nibbånapoâ-nibbånapabbhåra formula is already in effect applied in
each of these saµyuttas since it is imbedded in the Ga!gå-peyyåla
frame. This makes clear that we should not think in terms of there being
some sort of a priori doctrinal objection to applying these formulas to
items other than the eightfold path.
None the less, although BJT Ce wants to apply all four formulas in
all cases,21 it is not entirely clear how to apply any of the four formulas.
Usually they are inserted after bhåveti,22 but the exposition of the
satipa††hånas, sammappadhånas and jhånas does not follow the same
pattern ; the main verb is viharati or padahati rather than bhåveti, and it
is not clear how the formulas would fit into such sentences.23 In other
20S (Se) V 187,19–188,6 = (Be) III 120,18–25: idha, bhikkhave, bhikkhu sati-
sambojjha!gaµ bhåveti vivekanissitaµ viråganissitaµ nirodhanissitaµ
vossaggapariˆåmiµ || pa || upekkhåsambojjha!gaµ bhåveti rågavinayapari-
yosånaµ dosavinayapariyosånaµ mohavinayapariyosånaµ || amatogadhaµ
amataparåyanaµ amatapariyosånaµ || nibbånaninnaµ nibbånapoâµ
nibbånapabbhåraµ. imesaµ kho, bhikkhave, bhikkhu pañcannaµ uddham-
bhågiyånaµ saµyojanånaµ abhiññåya pariññåya parikkhayåya pahånåya
ime satta bojjha!gå bhåvetabbå. The above occurs at the conclusion of the
first rehearsal of the ogha-vagga, which begins by applying only the viveka-
nissita formula and is followed by further rehearsals of the Ga!gå-,
appamåda-, balakaraˆ¥ya-, esanå- and ogha-vaggas applying the råga-vinaya
formula.
21Thus, for example, S (Ce) V 340 states with reference to the Ga!gåpeyyåla in
the satipa††hånasaµyutta: vivekanissitådivasena rågavinayapariyosånådi-

vasena amatogadhådivasena nibbånaninnådivasena ca ekekasmiµ cattåro
cattåro katvå a††hacattål¥sasuttantå vitthåretabbå.
22bhikkhu sammådi††hiµ bhåveti vivekanissitaµ viråganissitaµ nirodhanissitaµ
vossaggapariˆåmiµ, etc.
23To apply the vivekanissita formula to the sentence idha bhikkhave bhikkhu
kåye kåyånupass¥ viharati åtåp¥ sampajåno satimå vineyya loke abhijjhå-

domanassaµ, the only option would seem to be to make vivekanissita qualify
bhikkhu which is hardly possible.
contexts in the Nikåyas we find the vivekanissita formula only applied
to the magga, indriyas, balas and bojjha!gas, though in the
Nettippakaraâ and some Buddhist Sanskrit sources it is applied to the
iddhipådas/Áddhipådas (Gethin 1992A, pp. 92, 162–68). On balance I
think Feer was probably right to exclude the application of all four
formulas from the satipa††håna-, sammappadhåna-, iddhipåda- and
jhåna-saµyuttas, but wrong to limit the application of these to the
vivekanissita and rågavinaya formulas in the case of the bojjha!ga-,
indriya- and bala-saµyuttas.
Finally in the sacca-saµyutta, Ce counts 15 instead of the 11 of the
other editions. The 4 extra suttas are found by taking the terms in the
compounds tulåkË†a-kaµsakË†a-månakË†a (S V 473,15–16) and
ukko†ana-vañcana-nikati (S V 473,20–21) as the basis of six separate
suttas rather than just two. This is possible though somewhat arbitrary
given the occurrence of dvandva compounds in other suttas of this
vagga which are not so treated.
3. Conclusions
1. Buddhaghosa’s total of 7,762 suttas for the Saµyutta-nikåya suggests
that the Påli tradition itself has long opted for the maximum number of
repetitions in considering this text. Moreover, in contrast to the text’s
European editors and translators, it has wanted to count these repetitions
as “suttas” in their own right.
2. But even taking the option of the maximum number of
repetitions, I have not succeeded in reaching Buddhaghosa’s total. The
figure I reach is 6,696, a figure which is still 1,066 short of his total. 24
This suggests that either I have made a mistake and overlooked some
section of repetitions or that the text of the Saµyutta-nikåya that has
24We might add 342 to the total for the iddhipåda-saµyutta on the grounds that
the vivekanissita, etc., formulas could conceivably be applied, but that still
leaves us 724 short, and if, against reason, we attempt to apply the
vivekanissita, etc., formulas and add 342 also in the case of the satipa††håna-,
sammappadhåna-, and jhåna-saµyuttas we have 8,064 — 302 over.
382 Rupert Gethin
come down to us is not as Buddhaghosa himself (or at least his source
for the figure 7,762) had it.
3. What then are we to make of these repetition sections of the
Saµyutta-nikåya ? Mark Allon (1997, pp. 360–63) has summed up some
of the suggestions that have been made concerning the significance and
function of repetitions generally in Buddhist texts. To paraphrase, these
include aiding memorization, getting the message across, cultivating
mindfulness, and enhancing the aesthetic appeal of the texts.
4. It is difficult to see how the structural repetitions of the kind we
have been considering have a straightforward mnemonic function in so
far as they themselves are what is to be remembered rather than an aid
to remembering it. But certainly we might see these kinds of repetition
as functioning as a way of getting the message across, cultivating
mindfulness, and enhancing the aesthetic appeal of the texts. The
doctrinal and practical importance of the items that are the subject of the
most repetitions — the unconditioned, and the seven sets of items that
come to be termed “dhammas that contribute to awakening”
(bodhipakkhiya-dhamma) — is clearly highlighted and enhanced by the
repetitions. Moreover this kind of structural repetition involving as it
does the substitution of various items in turn must require and develop a
certain mental alertness and agility that goes beyond mere rote
repetition, such that it might be considered a practice for developing the
Buddhist meditative virtues of mindfulness and concentration. But we
can perhaps go a little further in considering this function of repetition.
5. Although the items that are the subject of structural repetition
may be doctrinally important, it is hard to see how it could be doctrinal
considerations that are driving the repetitions. That is, in the Ga!gå-
peyyåla, it would seem it does not matter doctrinally whether it is the
river Ganges or the Yamunå ; or whether they are flowing to the “east”
or the “great ocean”. What is driving the repetition seems to be the very
requirement to repeat. This gives this kind of repetition something of
the quality of the kind of repetitive recitation that is found in various
religious traditions and often associated with the use of a rosary as a
means of counting off the repetitions. Of course, I am not suggesting
that a rosary was actually used in the recitation of the Saµyutta-nikåya,
merely that consideration of broader religious practices can help us
understand the possible functions of repetition in early Buddhist texts.
6. Given that what matters is not whether we are talking of the
Ganges or the Yamunå, but repetition for its own sake, why in the
Ga!gå-peyyåla stop at six rivers ? Why not throw in a few more ? Why
in the asaµkhata-saµyutta not add a few more substitute terms for the
unconditioned ? One response to such questions might be to say that one
cannot add any more rivers because this is buddhavacana and this is the
text and it cannot be changed. But such a response seems to me to miss
the point. Certainly the modern editions and the manuscripts on which
they are based each provide a fixed text, but when these different fixed
texts are considered collectively, although we can move some
considerable way towards determining a textual consensus, we are
confronted by the fact that in certain places the editions and manuscripts
indicate patterns of repetition that are by their very nature at least to
some extent open ended. My suggestion is that, although over time
these repetition sections have become more or less fixed, they originally
seem to have been composed in a manner that invites addition and
expansion — within certain parameters.25
7. The term peyyåla itself is rather curious. It appears to represent
Sanskrit paryåya in the sense of “repetition” : paryåya > payyåya >
25I made somewhat similar observations in Gethin 1992A (p. 252) and 1992B
(pp. 157–58) which have recently been the subject of criticism by Alexander
Wynne (2004, pp. 104–108) : while I would wish to tighten the use of the term
“improvisation” and exclude the implication of composition in performance,
on grounds that I hope are apparent in the present paper, I would wish to stand
by the claim that there are good reasons for thinking of different recensions of
Buddhist texts crystallizing after a period of somewhat freer composition and
adaptation. These are extremely complex issues and it seems to me that we
still lack a convincing model for the oral composition and transmission of
early Buddhist texts that can explain the kinds of difference and correspon-
dence that we find between versions of material in Påli, Sanskrit, and Chinese
and Tibetan translations.
384 Rupert Gethin
peyyåya > peyyåla (cf. Geiger 1994, §§ 52.5, 52.9, 46.3 ; Trenckner
1908, p. 117). But the technical sense of “repetition” seems to be
reserved for this particular form, which occurs alongside Påli pariyåya,
used in other senses. Similar Middle Indic forms such as peyåla and
piyåla are found used in the same way in Buddhist Sanskrit texts (q.v.
BHSD). Thus the term peyyåla in the sense of “repetition” seems to
have become frozen and is left unchanged when Buddhist texts are
transposed from one Middle Indian dialect to another. K.R. Norman
(2006, p. 114) has drawn attention to the fact that peyyåla seems to
represent an eastern dialect form. If we assume that peyyåla, pe, and la
were only used in abbreviating written texts, then as Norman points out,
the eastern form of the word might indicate that the texts began to be
written down before they were transposed into a western dialect ;
alternatively peyyåla in its technical usage is borrowed from some other
source at some later date. Another alternative, however, might be that
peyyåla was already used to abbreviate texts in oral recitation. It does
not seem to me implausible — pace Wynne 2004, p.107 — that reciters
and teachers of the texts might have resorted to the use of peyyåla to
establish the framework for patterns of repetition of the kind we have
been considering in the Saµyutta-nikåya ; these specific repetitions
might then have been recited in full as a religious exercise.
Table 2. Numbers of suttas counted in S II (Nidåna-vagga)
saµyutta suttas
e e
E Feer B CS Ce BJT Bodhi RMLG
nidåna 93 73 213 93 213
abhisamaya 11 11 11 11 11
dhåtu 39 39 39 39 39
anamatagga 20 20 20 20 20
Kassapa 13 13 13 13 13
låbhasakkåra 43 31 43 43 43
Råhula 22 14 22 22 22
lakkhaâ 21 21 21 21 21
opamma 12 12 12 12 12
bhikkhu 12 12 12 12 12
TOTAL 286 246 406 286 406
Table 3. Numbers of suttas counted in S III (Khandha-vagga)

saµyutta suttas
Ee Feer Be CS Ce BJT Bodhi RMLG
khandha 158 159 159 159 159
Rådha 46 46 46 46 46
di††hi 114 96 96 96 96
okkantika 10 10 10 10 10
uppåda 2610 10 10 10 10
kilesa 10 10 10 10 10
Såriputta 10 10 10 10 10
någa 50 50 50 50 50
supaˆâ 46 46 46 46 46
gandhabba 112 112 112 112 112
valåhaka 57 57 57 57 57
Vacchagotta 55 55 55 55 55
jhåna/samådhi 55 55 55 55 55
TOTAL 733 716 716 716 716
Table 4. Numbers of suttas counted in S IV (Sa¬åyatana-vagga)

saµyutta suttas
Ee Feer Be CS Ce BJT Bodhi RMLG
sa¬åyatana27 207 (247) 248 248 248 248
vedanå 29 31 31 31 31
måtugåma 34 34 36 34 34
jambukhådaka 16 16 16 16 16
såmaˆ"aka 16 2 16 16 16
Moggallåna 11 11 57 11 129
citta 10 10 10 10 10
gåmaî 13 13 13 13 13
asa!khata 44 (1463) 44 1848 44 1848
avyåkata 11 11 11 11 11
TOTAL 391 420 2286 434 2356
26Table at S III xi has “13” but this must be an error.

27Ce counts with next.
386 Rupert Gethin
Table 5. Numbers of suttas counted in S V (Mahå-vagga)
saµyutta suttas
e e
E Feer B CS Ce BJT Bodhi RMLG
magga 180 181 546 180 546
bojjhanga 187 185 632 184 632
satipa††håna 103 104 506 104 164
indriya 185 180 526 178 526
sammappadhåna 54 54 456 54 114
bala 110 108 456 108 456
iddhipåda 86 86 488 86 146
Anuruddha 24 24 24 24 24
jhåna 54 54 114 54 114
ånåpåna 20 20 20 20 20
sotåpatti 74 74 74 74 74
sacca 131 131 135 131 131
TOTAL 1208 1201 3977 1197 2951
Rupert Gethin
University of Bristol
BIBLIOGRAPHY
Allon, M., 1997. Style and Function : A Study of the Dominant Stylistic Features
of the Prose Portions of Påli Canonical Sutta Texts and Their
Mnemonic Function, Tokyo : The International Institute for Buddhist
Studies
Bapat, P.V., and Akira Hirakawa, 1970. Shan-chien-pî-pô-sha : A Chinese
Version by Sa!ghabhadra of Samantapåsådikå, Poona : Bhandarkar
Oriental Research Institute
Bodhi, Bhikkhu, 2000. The Connected Discourses of the Buddha : A New
Translation of the Saµyutta Nikåya, 2 vols, Oxford : Pali Text Society
in Association with Wisdom Publications
Collins, S., 1998. Nirvana and Other Buddhist Felicities : Utopias of the Pali
Imaginaire, Cambridge : Cambridge University Press
Geiger, W., 1994. A Påli Grammar, translated by Batakrishna Ghosh, revised
and edited by K.R. Norman, Oxford : Pali Text Society ; reprinted with
corrections 2000, 2005
Gethin, R.M.L., 1992A. The Buddhist Path to Awakening : A Study of the Bodhi-
Pakkhiyå Dhammå, Leiden : E.J. Brill
——— 1992B. “The Måtikås : Memorization, Mindfulness and the List”, in
J. Gyatso, ed., In The Mirror of Memory : Reflections on Mindfulness
and Remembrance in Indian and Tibetan Buddhism, pp. 149–72,
Albany : State University of New York
Norman, K.R., 1983. Påli Literature : Including the Canonical Literature in
Prakrit and Sanskrit of all the H¥nayåna Schools of Buddhism,
Wiesbaden : Otto Harrassowitz
——— 2006. A Philological Approach to Buddhism. 2nd ed., Lancaster : Pali
Text Society, 1st ed., London : School of Oriental and African Studies,
1997
Skilling, P., 1994. “The Synonyms of Nirvåâ According to Prajñåvarman,
Vasubandhu and Asa!ga”, Buddhist Studies Review 11, pp. 29–49
Trenckner, V., 1908. “Critical and Philological Notes to the First Chapter
(Båhirakathå) of the Milinda-pañha”, Journal of the Pali Text Society 6,
pp. 102–51
Woodward, F.L., 1927. The Book of Kindred Sayings, Vol. IV, London : Pali
Text Society
——— 1930. The Book of Kindred Sayings, Vol. V, London : Pali Text Society
Wynne, A., 2004. “The Oral Transmission of the Early Buddhist Literature”,
Journal of the International Association of Buddhist Studies 27, pp. 97–
127
The Career of Women Disciple Bodhisattas*
The whole purpose of conditioned existence is the attaining of

awakening, Nibbåna, according to the Theravåda Buddhist view of the
world. Eventually, individuals who develop their minds correctly will
all attain the state of being Noble Ones, arahats. The Påli Canon and
commentaries can be seen as being based on this concept. The teachings
and stories, the rules of conduct, and even the explanation of the
meaning of words all revolve around awakening. In the Canon, the
awakening of Buddha Gotama is the key event because he was able to
discover how to reach awakening and then teach others to do the same.
Like the Vinaya-pi†aka, many of the commentaries begin by
describing the career of the Great Bodhisatta (mahå-bodhisatta), or
Great Being (mahå-satta) as he is also called, who became the Teaching
Buddha Gotama (Sammå-sambuddha). The Vinaya-pi†aka account leads
up to the rules for the monks and nuns, the Påtimokkha. The intro-
duction to the Jåtaka commentary (SGB) leads up to Anåthapiˆ ika’s
gift of the Jetavana monastery to the Buddha and the Order of Monks,
presumably because this place was used by former Buddhas and was the
place most of the Jåtaka stories were told. The Ther¥gåthå commentary
leads up to the founding of the Order of Nuns.
The details of how an individual becomes a Teaching Buddha or a
Pacceka Buddha are readily available in translations into English.1
There is less information about what an individual does to become an
awakened disciple of a Teaching Buddha, so I would like to give an
overview of the career of those intent on awakening as women disciples,
taking most of my information from the Apadåna verses (many of
*Itis a privilege to make a contribution to this celebration of the hundred and

twenty-fifth anniversary of the Pali Text Society and the eightieth birthday of
K.R. Norman.
1Besides SGB, see also CSM, the commentary on the Buddhavasa ascribed to
Buddhadatta, and TP, from Dhammapåla’s commentary on the Cariyåpi†aka.
390 William Pruitt
which are included in Th¥-a), Buddhaghosa’s commentary on the
A¤guttara-nikåya (Mp),2 and Dhammapåla’s commentaries on the
Cariyåpi†aka and the Ther¥gåthå as well as his introduction and
conclusion to his commentary on the Theragåthå. This is mainly limited
to women disciples and should not be considered to be a comprehensive
view of the subject.
The teachings found in the Canon concentrate on attaining arahat-
ship, or at least one of the three lower states of awakening that mean an
individual is assured of becoming an arahat. The commentaries tell of
the countless number of human beings, Devas, and Brahmås who have
done the necessary preparation in the past to encounter a Teaching
Buddha, hear the Doctrine, practise it, and attain the highest goal. There
are, however, some indications of people who are merely started on the
path. For some lay people, the Buddha only taught the beginning steps
of generosity and moral conduct.3 Much more information about the
disciple’s path is found in the commentaries. This could be seen as
reecting the view that after the Buddha’s demise, fewer and fewer
people are born who have made the necessary preparations in past lives
to attain Nibbåna in this life. It is seen as crucial for these people, who
are unready to attain the nal goal, to make a maximum effort to be
generous and live moral lives. These actions lead to good lives in the
future, lives in which they can begin to put the Buddha’s Doctrine into
practice. If possible, a person in this life should also study the Buddha’s
Doctrine and practise training the mind and training in insight.
First, let us look at a few details of what must be done to become a
Teaching Buddha or a Pacceka Buddha. An individual (not necessarily a
human being) must encounter a Teaching Buddha, one of his disciples,
or a Pacceka Buddha and have faith in them. This faith profoundly stirs
that individual’s mind. That leads to doing a good deed, and eventually,
the individual is able to hear the Doctrine taught by the Buddhas and
make an aspiration to attain awakening. Then the individual begins to
2The passage on the foremost bhikkhun¥s is found in WL.

3See his discourses to two pairs of old brahmans, for example (A I 156).
The Career of Women Disciple Bodhisattas 391
put the Doctrine into practice. Over a number of lifetimes, the ten
perfections are cultivated,4 until the individual is ready to experience
Nibbåna for himself or herself. At this point, certain individuals aspire
to something higher : either becoming a Pacceka Buddha or a Teaching
´
Buddha. 5 From this point on, the individual will never be born on a
plane lower than the animal world and will never be female. Pacceka
Buddhas develop the ten perfections on another level, spoken of as the
twenty perfections. Teaching Buddhas develop a third level, or the thirty
perfections.6 The length of time is longer for Teaching Buddhas than for
Pacceka Buddhas, and even Teaching Buddhas can develop them for
three differing lengths of time. There are many events and
characteristics of Teaching Buddhas that are true for all of them and
others that vary from Buddha to Buddha.
Dhammapåla speaks of three categories of Buddhas : Sammå- ´
Sambuddhas (Teaching Buddhas), Pacceka Buddhas, and Såvaka

Buddhas (Disciple Buddhas) (Th-a I 10). Canonical texts use the title
arahat (or arahant) for all three types of Buddhas, and their pari-
nibbåna is said to be equal.7 Dhammapåla also uses the terms sammå-
sambodhi, pacceka-sambodhi, and såvaka-sambodhi (Th-a I 8, III
205f.). Disciple Buddhas are divided into three categories : chief ´
disciples (agga-såvaka), leading disciples (mahå-såvaka), and ordinary

disciples (pakati-såvaka) (Th-a III 206).
4A Burmese Buddhist explained to me once that an action done while aspiring

to the attainment of Nibbåna would come under the category of the perfections
and would continue to give results until a person became an arahat. An action
done while aspiring to something lower, such as wealth, would only work for
that effect and would be exhausted once the goal was reached.
5On the eight qualities that must be present for this aspiration to be effective,
see TP 262.
6See TP 312f. Dhammapåla gives several different explanations that have been
handed down concerning the interpretation of the ten, twenty, and thirty
perfections. One explanation is that ten perfections are necessary for
awakening for a disciple, twenty for a Pacceka Buddha, and thirty for a
Teaching Buddha.
7See TP 324.
392 William Pruitt
For Teaching Buddhas, the amount of time they must prepare
themselves as individuals intent on awakening (bodhisatta) falls into
three categories : (1) the minimum (four incalculables [asa¤kheyya] and
´
100,000 æons8), (2) the middle gure (eight incalculables and 100,000
æons), and (3) the maximum (sixteen incalculables and 100,000 æons)
(TP 325f.). Pacceka Buddhas must develop the perfections for two
incalculables and 100,000 æons (Th-a I 11). 9 The three types of
disciples who attain awakening must prepare themselves for the
following periods : (1) chief disciples, one incalculable and 100,000
´
æons ; (2) leading disciples, 100,000 æons (Th-a I 11) ; (3) ordinary
´ ´
disciples, 100 to 1,000 æons (?).

For the last category of ordinary disciples, no specic number
seems to be given. In the Visuddhimagga (XIII ¶ 16) Buddhaghosa says
´
other sectarians remember back 40 æons ; ordinary disciples remember

´
as far back as 100 to 1,000 æons (because their understanding is

strong) ; the eighty great disciples remember as far back as 100,000
´
æons ; the chief disciples remember as far back as an incalculable and

´
100,000 æons ; Pacceka Buddhas remember as far back as two incal-

´
culables and 100,000 æons ; but there is no limit to how far back Teach-
´
ing Buddhas can remember. In his commentary on the list of the fore-
most bhikkhun¥s (Mp I 376–77), Buddhaghosa says that Ther¥ Bhaddå-
Kaccånå, whom he identies with the wife of the Buddha, was one of
four disciples who possessed great supernatural knowledge (mahå-
bhiññå). The other three were the two Chief Disciples (Såriputta and
Mahå-Moggallåna) and Thera Bakkula. These four could remember
further back than any of the other disciples : “The rest of the disciples
´
can recall a hundred thousand æons, but, on the other hand, these four,
after attaining to great supernatural knowledge, can remember an
8TP 325 has “great æons” (mahåkappa), but “æon” (kappa) seems to be used
elsewhere. See the note on TP 325 concerning the length of time involved in
an asa¤kheyya and a mahåkappa.
9 According to Buddhadatta, it takes them one incalculable and more than
100,000 æons (CSM 88).
incalculable (asa¤kheyya)] plus a hundred thousand æons” (based on
the translation by Bode, WL, p. 789).
The wife of the Buddha is exceptional in that she is said to have
rst been associated with the future Buddha when he made his rst
resolve to become a Teaching Buddha. In the Apadåna her name is
given as Yasodharå,10 who says she gave eight handfuls of lotuses to the
Bodhisatta Sumedha, and he offered these to Buddha D¥pa¤kara. She
does not make an aspiration to become awakened in that life, however.
Dhammapåla says there are distinct differences between a Great
Bodhisatta and Pacceka and disciple bodhisattas. These differences are
seen in their faculties (indriyata), ways of practice (pa†ipattita), and
skilfulness (kosallata). The Great Bodhisatta has lucid faculties and
lucid knowledge, and he practises not for his own welfare, but for the
welfare of others. The other two types of bodhisattas do not. The Great
Bodhisatta applies skilfulness to his practice through his ingenuity in
creating opportunities (to benet others) and his skill in distinguishing
what is possible from what is not possible (TP 266f.). But for all of
them, the perfection of virtue is the foundation of their awakening (TP
276), with the difference that compassion and skilful means are the
forerunners for a Great Bodhisatta (TP 303). Dhammapåla points out
that the rst chapter of the Visuddhimagga discusses virtue as it should
be practised by those who seek to become awakened as disciples (TP
303).
We can also see how individuals aspiring to become Pacceka
Buddhas or disciples have lower aspirations than a future Teaching
Buddha. The Great Bodhisatta, Dhammapåla says, does not dedicate the
merits from his practice of virtue to his own release from afiction in
the unfortunate destinations or to his own achievement of kingship in
the fortunate destinations or to becoming a Universal Monarch, a Deva,
Sakka, Måra, or Brahmå, and he does not dedicate it to his own attain-
ment of the threefold knowledge, the six types of higher knowledge, the
10 ApII 592–96. For details of the different names used for her, see Bareau
1995. At the time of D¥pa¤kara, her name was Sumittå.
394 William Pruitt
four discriminations, the awakening of a disciple, or the awakening of a
Pacceka Buddha. He only dedicates it to becoming an omniscient
Buddha (TP 303). We can deduce that disciple bodhisattas aspire to
these lesser attainments.
Now let us look at the stories given in the Apadåna and the
commentaries about the past lives of the group of women who became
arahats and whose poems are included in the Ther¥gåthå. Tables 2 and 3
provide an overview.
A good story is one of the most useful teaching devices. A story not
only gets across a message in an entertaining way, it can also give us
information that does not appear in a general discussion of a topic. The
stories of the careers of women who attained awakening at the time of
Buddha Gotama or shortly after give us many details of the steps lead-
ing up to arahatship. For an individual who will become a Teaching
Buddha, it is at that point that he defers the attainment of arahatship and
resolves to work for the more difcult goal of becoming an omniscient
Buddha. Here, we will only look at the career for the disciples.
For many of the women there are two versions of their story in the
Ther¥gåthå commentary. One is in the prose text of the commentary
proper. The other is in the Apadåna verses that may have been included
in the commentary by Dhammapåla but were perhaps added later. There
are some problems as to whether the right verses are associated with the
right women. Since the names given with the Apadåna verses some-
times refer to the action done in a past life and are not the names of the
women at the time of Buddha Gotama, it is understandable that some
confusion could arise. I will not go into all the variants here, but the
tables show how the Apadåna verses tend to give more details ; when´
“Ap” is given, the information is only found in those verses. It is also

possible that some of the stories in the prose text are associated with the
wrong nuns, but that is not crucial to our discussion.
The rst step involves being born in a plane of existence where one
is sufciently intelligent and where one can perform meritorious deeds.
In the stories of the women elders, they were either human beings —
all of them but one being women in past lives — or they were kinnar¥,
depicted as half-human, half-bird. Such an existence assumes good
deeds through generosity and moral restraint in prior lives, but we will
begin the disciples’ careers with the life during which they rst had faith
in either a Teaching Buddha, a Pacceka Buddha, or a disciple of a
Teaching Buddha (a chief disciple, an arahat monk, or an arahat nun).
They are usually said to have been impressed with the demeanour of the
arahat they saw. In one case, an old woman visits a Bodhi tree and
thinks to herself, “If the Buddha, the Blessed One, is incomparable,
without equal, unrivalled, then may this [tree] of awakening show me a
marvel.” The tree shines forth, its limbs looking like gold, and the
woman is so impressed she sits there under the tree for seven days and
seven nights, then honours it with lighted lamps.11
At the very least, the believing individual pays respects to the
arahat, but she usually makes a gift of owers, food, robes, etc. Table 1
includes a list of the various gifts given by the women in past lives.12
The stories emphasize the material good results of such gifts, telling us
about the women’s lives in Deva worlds and the human world, about
their being beautiful and having large retinues, being wealthy, being the
chief queen of Deva kings, Universal Monarchs, or kings ruling over
large realms, etc. And this conrms the remark made above about the
motivation of disciples not being as lofty as that of future Teaching
Buddhas.
Other examples of such limited aspirations are found in the stories
when women wish for such things as many children or beauty. Uppala-
vaˆˆå, for example, gave a gift of ve hundred grains of fried rice and a
lotus ower to a Pacceka Buddha who had just come out of the
cessation state. She wished for as many children as there were grains of
11Th¥-a 60 (CVT 83).

12All of the foremost Bhikkhun¥s mentioned in Mp I 337–81 are included in
Th¥-a except for Bhaddå Kaccånå and Sigålakamåtå. No details are given con-
cerning their good deeds in past lives aside from aspiring under Buddha
Padumuttara to a foremost position.
396 William Pruitt
rice and for lotus owers to spring up at every footstep. In a later life,
she gave birth to ve hundred sons, all of whom become Pacceka
Buddhas. And lotuses sprang up with every footstep she took.13 These
lesser aspirations are not frequent in the stories, however.
The women who became chief disciples or great disciples under
Buddha Gotama saw Buddha Padumuttara praise a nun who was
foremost in a specic quality. This inspired them to make a gift, pay
respects, and aspire to become foremost in that same quality. Buddha
Padumuttara then looked into the future to see if their wish will come to
fruition and makes a prediction. They all make their aspiration under the
same Buddha because he lived one hundred thousand æons ago, and that
is the time required to become a leading disciple. This also seems to be
the period of time for the two chief disciples among the nuns (Khemå
and Uppalavaˆˆå), as only one of the nuns is said to have encountered a
Buddha earlier that Padumuttara. The exception is found in the Apadåna
(no. 28) under the name Yasodharå, one of several names for the former
wife of the Buddha.14 At the time of Buddha D¥pa¤kara, four incal-
culables and 100,000 æons ago, she gave eight handfuls of lotuses to the
Bodhisatta Sumedha, the future Buddha Gotama, who offered them to
Buddha D¥pa¤kara. This was the lifetime during which the Bodhisatta
received his rst prediction of Buddhahood.
For the chief disciples among the monks (Såriputta and Mahå-
Moggallåna), preparations took one incalculable and one hundred
thousand æons. They made their aspiration under Buddha Anoma-
dass¥. 15 We only nd mentioned eleven women in the Ther¥gåthå
commentary as aspiring to be foremost in some particular quality, thus
becoming leading disciples. Dhammapåla, in his concluding remarks,
simply says that the leading disciples among the nuns were Mahå-
Pajåpat¥ Gotam¥, etc. (Th¥-a 271, CVT 382). For the leading disciples
13Th¥-a 177, 179, 181 (CVI 233, 236, 239).

14For a discussion of the information found on the Buddha’s former wife, See
Bareau 1995.
15See Th-a III 90ff., CSM 255 (and the references in the note there).
among the monks, however, he gives all eighty names (Th-a III 205f.).
Ten laywomen who were foremost in some quality are listed in the
A¤guttara-nikåya (A I). The commentary (Mp I 401–404) says that each
of them made an aspiration to attain the quality in the presence of
Buddha Padumuttara, just as the foremost Ther¥s had done. They are
then said to have had many lives among devas and men up to the time
of Buddha Gotama. Only one is mentioned as having encountered
Buddha Kassapa, and that was the laywoman Visåkhå, who was one of
the seven sisters who were King Kik¥’s daughters and sisters of Buddha
Kassapa.
There is great variety in how the various women developed the
qualities necessary to attain arahatship. A few did bad deeds that led to
lives in hell. Two insulted an arahat nun, one committed adultery.
Others did mixed deeds that led to mixed results. One nun lled a
Pacceka Buddha’s bowl with mud, but she repented when a crowd of
people criticized her. She cleaned the bowl and lled it with food. As a
result, she had very bad breath in future lives, but was able to cure it.
Another woman gave a Pacceka Buddha a lotus, took it back because
she thought that he would have no use for a ower, then decided it
would be useful as a cover to his bowl and gave it again. She too
experienced pain for having taken back the ower, but eventually
overcame her problems in future lives. It is important to note the fact
that the mental attitude of the person doing an action is the main factor
in what the results will be. Also, the message is clearly that you cannot
wipe out a bad deed, but you can do a good deed to help mitigate the
results.
These stories, of course, were meant to serve as an inspiration to
Buddhists. The suffering caused by bad deeds should be a deterrent. The
rewards for good deeds should be encouraging. Even the smallest deed
done for an arahat — especially one who has just come out of the
cessation state — is of great reward in many lives. Large gifts give even
better results. The message here is that even a poor person can do
meritorious deeds that will establish him or her on the path to liberation.
398 William Pruitt
There is also the idea that being wealthy means one is able to make
large gifts, that possessing wealth carries the responsibility of being
generous (assuming one wishes to be happy in the future).
The standard phrase for women disciples for whom details of past
actions are not given is : she did meritorious deed(s) under previous
´
Buddhas and accumulated good (actions) as her basis for various lives
(ayam pi purimabuddhesu katådhikårå tattha tattha bhave viva††Ëpa-
nissaya kusala upacinant¥). In the stories given, the good actions
most frequently mentioned are paying respects, giving, and leading
virtuous lives. Exceptionally, some women ordained under former
Buddhas. Only one woman (Bhaddå Kåpilån¥) is said to have developed
the ability to go into absorption states (jhåna) in a past life. As a result,
she is the only woman said to have had a life in a Brahmå world.
Another important aspect of the disciples’ mental attitude is being
profoundly stirred (savega).16 In his introductory remarks to his com-
mentary on the Ther¥gåthå, Dhammapåla says,
Women of good family, daughters-in-law of good family, and young
women of good family heard of the full awakening of the Buddha, of the
Doctrine in accord with the [true] Doctrine, and of the proper establishment
of the Order. They had faith in the teaching and a profound stirring
concerning continued existence. Then they had their own husbands, mothers
and fathers, and relatives give them permission, and devoting themselves to
the teaching, they went forth. Having gone forth, they were of good and
virtuous conduct. They received instruction in the presence of the Teacher
and all the various theras. Then, striving and making effort, they realized
arahatship after a very short time.
Th¥-a 4 (CVT 8)
In the individual stories, the women are usually said to be pro-

foundly stirred after listening to a discourse on the Doctrine, and then
they are ordained. The only women for whom being profoundly stirred
is mentioned in past lives are those who ordained as nuns under former
Buddhas. So savega is a reaction to the conditioned world that
16Venerable Ñåâmoli (Path) translates as “sense of urgency”. Mr Norman uses

“religious excitement”.
motivates the individual to make a maximum effort to transcend the
misery (dukkha) that results from constant change (anicca) and the lack
of a permanent, controlling self (anattå). The usual order of events in
the attainment of arahatship during the time of Buddha Gotama is (1)
hearing the Doctrine, (2) being profoundly stirred, (3) ordaining in the
Community of Nuns, and (4) developing insight. For a number of
women, the Buddha, while seated in his PerfumeD Chamber, sends
forth a luminous image of himself that appears before the woman as she
is meditating. The image pronounces a verse that gives the woman just
what she needs to hear in order to understand correctly and attain
arahatship.
This brief look at the careers of the women elders does not give a
complete picture for the disciple arahats, of course. Other details or
variants in versions of stories are found in other commentaries, and
there are especially the number of occasions found in the Jåtaka
commentary when various women were born at the same time as the
Great Bodhisatta. As we saw, the women chief disciples do not prepare
as long as the men chief disciples. Other differences are to be expected,
so this essay should be seen as only a rst step in examining this
subject.
William Pruitt
400 William Pruitt
ABBREVIATIONS AND REFERENCES
Abbreviations of Påli texts follows A Critical Dictionary of Påli.
References to Th¥-a are to the second edition.
Bareau, André, 1995. “L’épouse du Buddha”, Recherches sur la biographie du
Buddha III, Paris´: École française d’Extrême-Orient, pp. 119–47.
CSM The Clarier of the Sweet Meaning (Madhuratthavilåsin¥), tr. by I.B.
Horner, Pali Text Society, 1978
CVT The Commentary on the Verses of the Ther¥s, tr. by William Pruitt, Pali
Text Society, 1998
Path The Path of Purification (Visuddhimagga), tr. by Venerable Ñaâmoli,
Buddhist Publication Society, 1956
SGB The Story of Gotama Buddha (Jåtaka-nidåna), tr. by N.A. Jaya-
wickrama, Pali Text Society, 1990
TP “A Treatise on the Påram¥s,” in The Discourse on the All-Embracing Net
of Views (pp. 254–330), tr. by Venerable Bhikkhu Bodhi, Buddhist
Publication Society, 1978
WL “Women Leaders of the Buddhist Reformation”, Journal of the Royal
Asiatic Society n.s. XXV (1893), pp. 517–66, 763–98´; with Påli text and
tr. by Mabel Bode from Mp I 337–81
Table I
Gd dds Gd dds/Bd Padu Pac Bu Ord. Bad dd Verse
1. Therikå Ÿ Ÿ
2. Muttå Ÿ Ÿ
3. Puˆˆå Ÿ 3 Ÿ
4. Tisså 1 Ÿ Ÿ
5. Tisså 2 Ÿ Ÿ
6. Dh¥rå 1, = 5 Ÿ Ÿ
7. Dh¥rå 2, 5 Ÿ
8. Mittå, = 5 Ÿ Ÿ
9. Bhadrå, = 5 Ÿ Ÿ
10. Upasamå, = 5 Ÿ Ÿ
11. Muttå Ÿ Ap
12. Dhammadinnå Ÿ
13. Visåkhå, = 6 Ÿ Ÿ
14. Sumanå, = 4 Ÿ Ÿ
15. Uttarå, = 4 Ÿ Ÿ
16. Sumanå 2 Ÿ
17. Dhammå Ÿ
18. Sa¤ghå Ÿ Ÿ
19. Nandå Ÿ
20. Jent¥, 19
21. Unknown 1 Ÿ
22. A hakås¥ Ÿ Ÿ
23. Cittå Ÿ Ÿ
24. Mettikå Ÿ
25. Mittå Ÿ
26. Abhayamåtå Ÿ
27. Abhayatther¥ Ÿ
28. Såmå Ÿ
29. Såmå 2 Ÿ
30. Uttamå Ÿ
31. Uttamå 2 Ÿ
32. Dantikå
33. Ubbir¥ Ÿ
34. Sukkå Ÿ
35. Selå
36. Somå, 27 Ÿ Ÿ
37. Bhaddå Kåpilån¥
38. Unknown 2 Ÿ Ÿ Ÿ (3)
39. Vimalå Ÿ
40. S¥hå Ÿ
41. Nandå, 19 Ÿ
42. Nanduttarå Ÿ
43. Mittåkål¥ Ÿ
44. Sakulå Ÿ Ÿ
45. Soˆå Ÿ Ÿ
46. Bhaddå, former Jain Ÿ
47. Pa†åcårå Ÿ
48. Thirty nuns Ÿ
49. Candå Ÿ
402 William Pruitt
Gd dds Gd dds/Bd Padu Pac Bu Ord. Bad dd Verse
50. Five hundred nuns Ÿ Ÿ
51. Våsi††h¥ Ÿ
52. Khemå Ÿ Ÿ
53. Sujåtå Ÿ
54. Anopamå Ÿ
55. Mahå-Pajåpat¥ Gotam¥ Ÿ Ÿ
56. Guttå Ÿ
57. Vijayå Ÿ
58. Uttarå Ÿ
59. Cålå Ÿ
60. Upacålå Ÿ = 59
61. S¥sËpacålå Ÿ = 59
62. Va ha’s
mother Ÿ
63. Kisågotam¥ Ÿ
64. Uppalavaˆˆå Ÿa
65. Puˆîkå Ÿ Ÿ Ÿ
66. Ambapål¥ Ÿ Ÿ
67. Rohiˆ¥ Ÿ
68. Cåpå Ÿ
69. Sundar¥ Ÿ
70. Subhå, Ÿ
the smith’s daughter Ÿ
71. Subhå J¥vakambavanikå
72. Isidås¥ Ÿ 3
73. Sumedhå Ÿ
Gd dds = good deeds, Gd dds/Bd = good and bad deeds, Padu = Padumuttara, Pac Bu =
Pacceka Buddha, Ord. = ordained, Bad dd = bad deeds, Verse = verse of radiance
Gifts, good acts´:

A. Unspecied (4–11, 13–18, 21, 28, 37–45, 48–51, 53, 54, 56–63, 65, 66, 68)
B. Good deeds´:
Paid respects (2, 69), ordained (22, 34 [expert in Doctrine], 52, 65, 66), kept
Observance Days (30), went forth as female ascetic and developed absorption states
(37), learned a discourse by heart (52), outstanding merit with regard to the Triple Gem
(73)
Gave two gifts when told to give one (12), great gift (41), good deeds like giving (55)
C. Food´:
Gift (Ap´: food) to a chief disciple after he rose from the state of cessation (12), food
(25, 33), ladle full of food (26), three cakes (31, 52), invited Buddha and Sa¤gha for
seven days (47, 52, 55 Ap, 64), cakes (67), spoonful of food (69 Ap)
D. Food and ower(s)´:
Fried rice (500 grains) and lotus (64)
E. Flowers
Flowers & perfumes (1), garland of reeds/reeds and owers (3, 23), seven lotuses (27,
36), sala¬a owers (29), sål branches (32 Ap), owers (32), seven owers (64 Ap), lotus
a Mixed.
owers (64 Ap)
F. Robes´:
Set of three robes (1, 47, 55, 64 Ap), two robes (25), yellow cloth (64 Ap)
G. Lodgings´:
Pavillion of branches (1), built monastary (46, 47, 52, 64), park (73)
H. Requisites´:
Requisites (45 Ap), huts, requisites (couches, seats, drink, food, bowls, etc.) (56)
J. Made or honored a shrine´:
Honored shrine (19, 20w?), jewelled belt (given to shrine) (24), made a shrine (24 Ap),
honored Bodhi tree with lamps (5 lamps, Ap) after sitting there seven days and nights
(35), golden umbrella with gems (19, 20w?), golden tile for shrine (37), honored shrine
with oil (44)
K. Participated in great offering´:
Presumably joined citizens in the following event´: “When the teacher approached, all
the citizens, happy, pleased, went to meet him and strewed sand [for the festivities].
They swept the road and prepared banners and pots full of plantains. The teacher was
honoured with smoke (of incense), powder, and owers. A hall was prepared, the guide
was invited. A great offering was given in hope of full awakening.” (11)
L. Shared in other’s merit´:
Rejoiced at husband’s gift (37)´: (a) robe, (b) meal, jewelled pavilion, bowl of gold, (c)
7,000 bowls with 7 jewels, lled with ghee´; lamp wicks by thousands, lit, (d) shawl, (e)
supported Pacceka Buddhas for 3 months´; gave sets of three robes
M. Gift of hair´:
Let hair fall down and gave it as a gift (52)
Bad deeds done in past
A. Insulted a nun (22, 66)
B. Unspecied (led to children dying) (50)
C. Adultery (72)
Mixed deed (partly bad)
A. Filled a Pacceka Buddha’s bowl with mud but repented and cleaned it (37)
B. Gave a lotus, took it back, then gave it again (64)
Gifts to´: Teaching Buddha, Pacceka Buddha, Chief disciple, disciple (arahat), nun
(arahat), shrine to the Buddha, Bodhi tree (which showed her a marvel [35])
404 William Pruitt
Table 2
Pa Si Ti Ph Vi Si Ve Kak Koˆ Kass
1. Therikå Ÿ Ÿ
2. Muttå 1 Ÿ
11. Muttå 2 Ÿ Ap
12. Dhammadinnå* Ÿ Ÿa
19. Nandå Ÿ
22. A hakås¥ Ÿ
24. Mettikå Ÿ
25. Mittå Ÿ
26. Abhayamåtå Ÿ
27. Abhayatther¥ Ÿ
29. Såmå 2 Ÿ
30. Uttamå Ÿ
31. Uttamå 2 Ÿ
33. Ubbir¥ Ÿ
34. Sukkå Ÿ Ÿ Ÿ Ÿ Ÿ Ÿ
35. Selå Ÿ
36. Somå
37. Bhaddå Kåpilån¥* Ÿ Ÿ
41. Nandå* Ÿ
44. Sakulå* Ÿ Ÿ
45. Soˆå* Ÿ
46. Bhaddå,
former Jain* Ÿ Ÿa
47. Pa†åcårå* Ÿ Ÿa
52. Khemå*, ** Ÿ Ÿa Ÿa
55. Mahå-Pajåpat¥
Gotam¥* Ÿ Ÿa
63. Kisågotam¥*
64. Uppalavaˆˆå*, ** Ÿa
65. Puˆîkå Ÿ Ÿ
66. Ambapål¥ Ÿ Ap Ÿa
67. Rohiˆ¥ Ÿ
69. Sundar¥
73. Sumedhå Ÿ Ÿ
Bhaddå Kaccånå* Ÿ
Mp I 376–77
Sigålakamåtå* Ÿ
MP I 381
*Foremost in some quality. **Two chief women disciples
Buddhas´: Pa = Padumuttara (100,000 æons ago), Si = Siddhattha (94 æons ago), Ti =
Tissa (92 æons ago), Ph = Phussa (92 æons ago), Vi = Vipass¥ (91 æons ago), Si = Sikh¥
(31 æons ago), Ve = VessabhË (31 æons ago), Kak = Kakusandha (the present æon), Koˆ
= Koˆågamana (the present æon), Kass = Kassaya (the present æon)
a One of seven sisters, daughters of King Kik¥. At Ja IV 481, The Buddha’s mother,
Mahåmåyå, is given as one of the seven in place of Bhaddå the former Jain. The seventh
sister is the laywoman Visåkhå
bOrdained.
Table 3
Gdq`ndi >cdàlp`ìja
C`gg ?`q\ Nk`^d^?`q\ =m\ch ?`q\ Pidq) fdibn
rjmg_n rjmg_n rjmg_ fdibn Hji)
1)Oc`mdfåa Ÿ
2)Muttå 4
3)Puˆˆå 2 36 10
11)Hpooå Ÿ
12)?c\hh\_diiå Ÿ 2 (twice)
19)I\i_å Ÿ
22)< hakås¥ Ÿ
23)>dooå Ÿ
24)Hòodfå Ÿ
25)Hdooå Ÿ 2 (Ap) 30 20
26)<]c\t\håtå Ÿ 2 (Ap) 36 50
27)<]c\t\ooc`m¥ 70 36
30)Poo\hå 2 64 63
31)Poo\hå 2 Ÿb
32)?\iodfå Ÿ 2 (Ap) 36 10
33)P]]dm¥ 2
34)Npffå Ÿ 2, 4
35)N`gå Ÿ 2 80 100
37)=c\__å Kåpilån¥ Ÿ Ap Ÿ
41)I\i_å 2–6 many
44)N\fpgå Ÿc 1
45)Njˆå Ÿ 2 (Ap)
46)=c\__å, former Jain Ÿd 2, 2–6 (Ap) many Ÿ Ÿ
47)K\†åcårå Ÿ 2 (Ap, twice)
52)Fc`hå Ÿ 2, 4, 2–6 (Ap) many many many
and all 6 (Ap)
55)H\cå-Pajåpat¥ Gotam¥ Ÿ 2 (Ap)
56)Bpooå Ÿ
63)Fdnågotam¥ Ÿ 2 (Ap, twice)
64)Pkk\g\q\ˆˆå Ÿ 2 (Ap, twice)
66)<h]\kål¥ Ÿ 2 (Ap)
67)Mjcdˆ¥ 2 (Ap)
69)Npi_\m¥ 2 (Ap)
72)Dnd_ås¥ Ÿe Ÿ
73)Nph`_cå Ÿ 2, 2–6 many many many
Hk=c\__F\^î Ÿ
HKNdbg\f\ho Ÿ
?`q\rjmg_n5ndsnìpnjpnrjmg_n#fh\(gjf\$51)>åtummahåråjikadeva, 2. Tåvatisa, 3. Yåma, 4.

Tusita, 5. Nimmåna-rat¥, 6.´Paranimmita-vasavatt¥ (or Vasavattipura). Some women are said to be born
in ve Deva worlds, presumably numbers 2–5.
a
Also only in Deva worlds for a Buddha interval.
b
No lower births for 91 æons.
c
Deva worlds only.
d
Good lives for a Buddha interval.
e
Because she committed adultery when she was a man, she cooked in hell for hundreds of years, was
born three times as an animal, then as a hermaphrodite, and nally as a woman in a poor family.
On the Correspondence of Helmer Smith
and Gunnar Jarring
During the years 1988 and 2000, when I was acting as the delegate
of the Royal Swedish Academy of Literature, History, and Antiquities
at the Union Académique Internationale, it was one of my regular duties
to give an annual report on the latest meeting of the Academic Union.
Whenever I did so, I frequently had the occasion of mentioning A
Critical Påli Dictionary, which, being one of the oldest projects, had
been placed under the auspices of the International Academic Union at
an early stage. After one of my autumnal reports at the Plenary Session
of our Academy in Stockholm, Gunnar Jarring 1 remarked to me that my
presentation had interested him, particularly my comments on the
progress of A Critical Påli Dictionary, since he had kept up a long-
lasting correspondence with Helmer Smith.2
Gunnar Jarring (1907–2002) was born in southern Sweden (Skåne).
He received his Ph.D. from Lund. In the very beginning he went in for
German and Scandinavian languages, but soon took up Sanskrit and
comparative Indo-European philology, with Helmer Smith as one of his
teachers, and made profound studies in Slavic languages, above all
Russian, which finally led him to his chief subject, Turkology. A very
industrious and competent lecturer, Gustaf Raquette (1871–1945),3
taught Turkish at that time at the University of Lund. Before his
academic career docent Raquette had spent twenty-five years (1896–
1
Staffan Rosén, “Gunnar Jarring”, in Kungl. Vitterhets Historie och Antikvitets
Akademiens Årsbok 2003 (Stockholm 2003) ; pp. 34–40 ; VEM ÄR DET ’99
(Stockholm 1998), p. 540.
2
Björn Collinder, “Helmer Smith”, in Kungl. Vitterhets Historie och Antikvitets
Akademiens Årsbok 1956 (Stockholm, Lund, 1956), pp. 55–59. Hans
Hendriksen, “Helmer Smith 26th April 1882–9th January 1956”, in A Critical
Påli Dictionary, Vol. II (Copenhagen 1960), pp. v–viii.
3
Cf. S. Rosén, 2003, p.35. Raquette is often mentioned in Helmer Smith’s
letters to Gunnar Jarring.
408 408 Siegfried Lienhard
1921) as a missionary in Kashgar in East Turkestan (today’s Xingjiang).
He was fluent in Turkish, especially East Turkish, and, naturally, had a
decisive influence on the young Jarring.4 In the same year in which
Jarring obtained his doctor’s degree (1933), he was employed as
“docent” and examiner in Turkish linguistics at Lund University. In
1940 Jarring’s lectureship expired, but thanks to his excellent
knowledge of Russian and Turkish, he was now placed in the Swedish
Security Service, first in Ankara and in 1941 in Teheran. From this year
onward he followed a diplomatic career.
He held many outstanding posts. To mention only his most
important assignments, he was chargé d’affaires in Addis Abbaba
(1946), ambassador in New Delhi (1948–1952), ambassador in the
Swedish Foreign Office (1952–1956), ambassador to the United
Nations (1956–1958), ambassador in Washington (1958–1963) and,
finally, ambassador in Moscow (1964–1973). In his spare time Jarring
continued to pursue his beloved oriental studies.5
4
After his return from Kashgar, Raquette taught two years at the Mission
School of the Swedish Missionsförbund (1922–1924) in Lidingö, a suburb of
Stockholm. He worked in Lund from 1924–1937. Here he was given the
degree of a doctor honoris causa in 1937, an honour probably suggested by
Jarring. “You awarded Raquette a doctorate ; this was well done” (“Ni
promoverade Raquette, det var bra gjort”) remarks Smith in a letter dated 2
June 1937. Raquette published a series of fundamental works on East Turkish,
as, for example, Eastern Turki Grammar, 3 vols., 1912–1914 ; English–Turki
Dictionary, 1927 ; The Accent Problem in Turkish, 1927 : and edited East
Turkish literature.
5
For his comprehensive scientific work see the publications listed in Gunnar
Jarring — en bibliografi redigerad av Christopher Toll & Ulla Ehrensvärd
(Stockholm 1977) ; The Published Writings of Gunnar Jarring 1977–1988 : A
Bibliography Compiled by Ulla Ehrensvärd. Turcica et Orientalia ; Studies in
Honour of Gunnar Jarring on His Eightieth Birthday 12 October 1987
(Stockholm 1988), pp. 192–204 ; and Gunnar Jarring : En bibliografi
redigerad av Ulla Ehrensvärd (Stockholm 1997). See also Roger Nilsson and
Johan Fresk, eds., A Bibliography of Literature on Journeys and Explorers in
Asia in the Gunnar Jarring Library at Stockholm University (Stockholm
2007).
On the Correspondence of Helmer Smith and Gunnar Jarring 409
Helmer Smith (1882–1956), born in Stockholm, studied at the
University of Uppsala, where he took the degree of Fil.lic. (M.A.) in the
year 1908. In 1925 he received the honorary doctorate from the
University of Lund.. After having spent many years abroad, mainly in
Berlin, Paris and Copenhagen, he obtained a lectureship (“docentur”) at
Lund (1921–1935). Thus both Jarring and Smith lived many years in
this old university town at times which partially coincided.
In a recently published article6 Olle Kvarnsträm narrates how
Gunnar Jarring met Helmer Smith the very first time. Kvarnsträm
writes, “The diplomat and orientalist Gunnar Jarring mentioned that
during his years as a student in Lund a man was living in the house
opposite (his own flat) in Nygatan. The gentleman in question seemed
to devote the greater part of the day to something that looked like
playing at patience. A fact that particularly struck Jarring was, however,
that at regular intervals a lady entered the room and collected the cards
which were then kept in an adjacent room. By and by, Jarring was
introduced to the gentleman on the other side of the road and got to
know that what had looked like patience cards were, as a matter of fact,
excerpt cards for the A Critical Påli Dictionary founded by Helmer
Smith and Dines Andersen.”7
In 1936, at the age of fifty-five, Helmer Smith was called to occupy
the professorship of Sanskrit and Comparative Indo-European
Linguistics at the University of Uppsala, a chair he had applied for
6
“Från Hampton Roads to Lundagård. Forskning om indiska religioner vid
Lunds universitet 1880–2005”, in Årsbok 2006, pp. 42–56.
7
“Diplomaten och orientalisten Gunnar Jarring har berättat från sin studietid i
Lund om en man som var bosatt i huset mitt emot honom på Nygatan. Mannen
ifråga tycktes ägna större delen av dygnets timmar åt vad som såg ut att vara
patiens. Vad som förbryllade Jarring var emellertid den kvinna som
regelbundet trädde in i rummet för att samla ihop kortlapparna vilka sedan
förvarades i ett angränsande rum. Jarring blev så småningom presenterad för
mannen på andra sidan gatan och fick då veta att vad som såg ut som
patienskort i själva verket var excerptlappar avsedda för den av Helmer Smith
och Dines Andersen grundlagda A Critical Påli Dicionary.”
somewhat hesitatingly. He held it until 1947. Though mainly devoted to
Påli and Sinhalese studies, he shared many interests with Gunnar
Jarring. His predecessor in Uppsala was Jarl Charpentier (1884–1935).
Both Smith and Jarring were ordinary members8 of the Royal
Swedish Academy of Literature, History, and Antiquities in Stockholm.
Smith was elected in 1940, Jarring in 1969. In his inaugural lecture
Smith discussed “En detalj i den indiska metriken” (“a detail in Indian
metrics”), while Jarring devoted his lecture to “Poltava och karolinskt
kulturarbete i Turkiet och Sibirien”.9 Jarring’s lecture was published in
the Academy’s yearbook, but Helmer Smith’s — due to the ongoing
World War II — unfortunately not. Only a handout of seven pages with
verses in Sanskrit, Persian, Greek, and Latin as well as metrical patterns
is supposed to be left of this inaugural lecture.10 Interestingly, a minor
incident occurred while Smith delivered his paper on Indian metrics, a
subject which certainly was of no great interest to most of his
colleagues in the learned audience. At that time His Royal Highness
Gustav Adolf, later King Gustav VI. Adolf (1950–1973), was the Patron
of the Academy and liked to act as chairman at the Academy meetings.
As he was only moderately attracted by the subject dealt with by Smith,
the Crown Prince nodded off for a while. Helmer Smith, who, as it
seems, was easily offended, felt so distressed by this lack of interest on
the part of His Royal Highness that he never attended another meeting
of the Academy.
The two gentlemen were rather different with regard to birth,
temperament, work, and lifestyle. Smith was the proud specialist of Påli
grammar and prosody which he had mastered brilliantly, while Jarring
stood out as the indefatigable explorer of Turkish and Central Asian
languages. Whereas Smith was one of those scholars who had neither
visited the Indian subcontinent nor any of the various countries where
Påli is still spoken and studied, Jarring made many journeys, sometimes
8
In Swedish called “arbetande ledamöter”, that is to say, “working members”.
9
“Poltava and Carolingian cultural work in Turkey and Siberia”.
10
B. Collinder, p. 58.
even painstaking land travels on horseback. On his paternal side Smith
was of British descent.11 His great-grandfather had been a sea captain
from Belfast. Jarring, however, came from the countryside. He was the
son of a couple of farmers from Brunnby in Malmöhus län named
Gottfrid Jönsson and Betty Svensson, but changed his family name in
early adulthood to Jarring. Already before he obtained his doctor’s
degree with a thesis on “Studien zu einer osttürkischen Lautlehre”, he
undertook research trips to Central Asia and East Turkestan, particularly
Kashgar, where his teacher Gustaf Raquette had spent so many years.
Some time after his death Gunnar Jarring’s personal library and
other documents of his legacy — among them also the letters written to
him by Helmer Smith — were transferred to the Institute of Oriental
Languages in Stockholm on the initiative of Staffan Rosén, professor of
Korean language and literature at Stockholm University. His books are
still kept in Stockholm ; his letters and papers, however, have been
deposited in Lund. I was, of course, curious about the letters and
postcards which Helmer Smith, twenty-five years older than Jarring,
had sent to his young colleague and, therefore, some time ago requested
the authorities in charge of Jarring’s legacy to kindly grant me access to
Smith’s letters, which are now preserved at the Manuscript Department
(“Handskriftsavdelning”) of the Library of the University of Lund.12
The “Collection Gunnar Jarring” contains altogether 181 letters,
letter-cards and postcards from Helmer Smith, most of which were sent
to Jarring during the first five years of the two scholars’ acquaintance—
1934 : 26 letters, 1 letter-card, and 1 postcard,
1935 : 50 letters and 7 letter-cards,
1936 : 43 letters, 6 letter-cards, and 1 postcard,
1937 : 22 letters and 2 postcards, and
11
B. Collinder, p. 56.
12
I would like to record my gratitude to Birgitta Lindholm, Chief Librarian at
the Manuscript Department of the University Library in Lund, for all the
generous help I received from her during a short stay in Lund in March 2007.
In this article Helmer Smith’s letters and other Swedish documents have been
translated into English. The Swedish originals are quoted in the footnotes.
1938 : 14 letters and 2 letter-cards —
while during the following seven years Jarring received relatively little
mail from his correspondent, namely —
1939 : 4 letters,
1940 : 1 letter,
1947 : 1 letter,
1949 : 4 letters,
1950 : 8 letters,
1951 : 7 letters, and
1952, four years before Helmer Smith’s death, again only 1 letter.
All the letters are handwritten, in an even and easily readable ductus
litterarum. Smith addresses Jarring always as “Broder”, that is to say,
“brother”, a formal address still used, particularly in academic circles.
Every single letter is fully signed with “Helmer Smith”, never with
simply “Helmer”, although the pronoun used is the familiar “du” and
the verb form that of the second person singular. The letters were
generally sent from Lund or Uppsala, some also from Stockholm. Most
of them were, however, posted in Kummelnäs, his favourite abode,
which he, being an only child, inherited from his well-to-do father’s
wife and was beautifully situated in the inner Archipelago of
Stockholm, at about fifteen kilometres distance from the centre and
opposite Vaxholm.
As can be seen from the table given above, between 1949 and 1951
Helmer Smith sent again a few more letters to his former disciple. It
was during these years that Jarring was Swedish ambassador in New
Delhi, a position that evidently made quite an impression on Helmer
Smith.13 Moreover, Smith was interested in certain books published in
India.14
Before I started going through his correspondence, I had hoped
13
Though Jarring was not yet ambassador when he stayed in Teheran, Smith
addressed his letter of the 27 November 1951 to “His Exc. Dr Gunnar Jarring,
Royal Swedish Legation, Teheran”.
14
A name that occurs several times is Suniti Kumar Chatterji.
above all to find in his letters many passages in which he discussed
problems connected with Påli and Sinhalese. To my great
disappointment, however, this is rarely the case. Most letters deal with
events in his daily academic life, his contacts with colleagues, their
ability and behaviour, their applications for university posts, their
successes and, more frequently, shortcomings and failures. Almost all
Scandinavian scholars of Sanskrit, Iranian and related studies of the
time pass review, in his correspondence, but need not be mentioned by
name in this paper. In particular, Smith often refers to Hannes Sköld
(1886–1930), then one of the most outstanding linguists of the
University of Lund, in whose edition of the “Materialien zu den
iranischen Pamirsprachen” (1936) Helmer Smith was deeply involved.15
“I have now been acting under the strict supervision of my Danish
employer,16 and thus ‘Sköld’ has rested…”,17 writes Smith on 8 August
1935. He also assisted Wilhelm Geiger in Munich in his A Dictionary of
the Sinhalese Language, which appeared in Colombo in 1941.18
Smith’s correspondence shows naturally a clear predilection for
lexicography and etymology. Almost every letter abounds in words or
phrases from the numerous languages he knew (in addition to Sanskrit,
Påli, and Prakrit especially Sinhalese, Khotanese, Hindi, Tamil,
Burmese, Tibetan, Persian, Urdu, Turkish, and other idioms). While
reading his letters, we must bear in mind that Smith’s correspondent
15
An internationally especially renowned work of this untimely deceased
scholar — “docent” at the University of Lund — is, of course, his “The
Nirukta, Its Place in Old Indian Literature, Its Etymologies” (1926) ; cf.
O. Kvarnström, p. 47.
16
This was, of course, Dines Andersen, with whom Helmer Smith together with
Hans Hendriksen had edited Vol. I of A Critical Påli Dictionary (1924–
1948). Hans Hendriksen (1913–1989) was Smith’s successor in Uppsala
(1947–1951). Dines Andersen (1861–1940) held the chair of Indian Philology
at the University of Copenhagen from 1903 to 1927.
17
“Jag har nu i sex veckor stått under omedelbar uppsikt av min danske
arbetsgivare, så ‘Sköld’ har vilat”.
18
See Wilhelm Geiger, Kleine Schriften zur Indologie und Buddhismuskunde,
Heinz Bechert, ed. (Wiesbaden 1973), p. *131.
was a specialist in Turkish and that, therefore, Central Asian languages
are in the foreground. Once, when Jarring was abroad on his first
journey to East Turkestan and, while undertaking various trips, he had
not given a more detailed indication of his whereabouts than just “next
address Kashgar”, Smith sent a letter to Mrs Jarring, dated 2 June 1935,
requesting her, “as a deputy guide for Turkish studies”,19 to inform him
about useful manuals of Osman Turkish presenting the new writing
system. “I shall never learn it,” he confesses, “but I need putting my
nose into one thing and the other”. He took a great interest also in
Khotanese, the ancient south-east Iranian language of the Sakas. This
got him in touch with Sir Harold Bailey (1899–1996) in Cambridge,
though Smith’s letters to Jarring do not reveal which of the two
contacted the other first.
Helmer Smith liked to associate with his friends. In a letter sent on
2 June 1937, he gladly informs his addressee of the good news that his
wife “Ellen has succeeded in getting a little housemaid for the summer
which means that it will be easier for us to receive our friends this
summer”.20 Colleagues with whom he entertained especially close
relations were his French “friends from the twenties”, such as Jules
Bloch, Louis Renou, Armand Minard, Pierre Meile, and Jean Filliozat,
who are all mentioned in his letter of 22 August 1937. There he also
reports to Gunnar Jarring that “Ellen and I could spend almost the
whole month of July in Paris (more correctly in Sèvres as parasites at
Jules Bloch’s), amidst a bustle of orientalism and orientals”.21 But
already in a much earlier letter, dated 14 September 1936, he records
that “two Parisians (Jules Bloch and Mrs Foucher 22) came from
19
“I Er egenskap av vikarierande turkologisk vägledare.”
20
“Ellen har lyckats få en liten jungfru för sommaren, vilket betyder att vi nog
får det lättare att ta emot våra vänner den här sommaren.”
21
“Ellen och jag fick leva nästan hela juni månad i Paris (rättare i Sèvres som
snyltgäster hos Jules Bloch), mitt i ett myller av orientalism och orientaler.”
22
The wife of Alfred Foucher.
23
Copenhagen and stayed with us”.
Smith speaks little of his university teaching. On 21 November
(probably 1937) he tells Jarring that “during the Spring seminar we will
deal with ‘style and metre in the Upani!ads’ and accept with gratitude
all parallels and contributions to the world’s art of verse from the Turk 24
in Lund himself”. In a letter dated 18 April 1937 Smith comments with
enthusiasm on Jarring’s suggestion of holding a series of joint seminars
on the Sakas, “Your proposal of joint seminars on the Sakas is genial,
… it should, however, be called ‘Every third (fourth) week H.S. Nyberg
and Helmer Smith will hold sem(inar) ex(ercises) on Central Asiatic
texts together with doc(ent) G. Jarring.’ ”
As can be seen from many remarks already quoted, Smith was not
only a loyal friend but also had a good sense of humour as well as a
wise and subtle irony which could concern also himself. He begins his
letter of the 29 August 1935 with the words : “Thus my metrical follies
have found you in the heart of Asia ! — in Pakistan …”.25 Although he
was very hard working, he did not despise festive occasions, even when
they tended to be somewhat excessive as, for example, the celebration
of “Dines Andersen’s last day in Sweden” (letter dated 19 August 1935)
which caused “fatigue and a nosebleed”. On 25 July 1937 Smith
recounts, “We work joyfully : two cigars and five fillings of the pipe are
the time-measure of a working day for Dines — then I have seventeen
hours for airing our study. We shall see if it will end on the 10th of
August.”26 On 25 May 1938 he communicated humorously to Jarring :
“You have thus seen [in the newspapers] that I have been considered
23
“Det kom tv parisare (Jules Bloch och fru Foucher) från K(ö)b(en)h(a)vn och
stannade en vecka hos oss —.”
24
“I vår behandlar vi ‘stil och metrum i upani!aderna ‘ och alla paralleler och
bidrag till världsverskonstens teori mottas med tacksamhet … från självaste
lundturken” (that is to say, Gunnar Jarring).
25
“Så har mina metriska tokerier funnit dig i hjärtat av Asien ! — i Pakistan… ”
26
Så vi arbetar gladeligen : två cigarrer och fem pipstoppningar är tidsmåttet på
Dines’ arbetsdag — sen har jag 17 timmar att vädra arbetsrummet på. Vi får
se, om det slutar 10de augusti.”
decrepit enough to enter the Academy of Antiquities… .”27 He also
makes fun of Charpentier and his name, when in one of his earliest
letters written on 17 March 1936, he requests Jarring to carefully
proofread some newly edited, unspecified text :28 “Well, this is what it
looks like. Would you kindly take the trouble of seeing to it that the
corrections are not misunderstood, then it can be printed. But we must
have a third proof of sheet 16, likewise of the remaining sheets.
Otherwise the whole thing will become Charpentier (the blessed man
considered that proofreading belonged to the lower classes — including
some German case- and number-endings). — And I am still far from
being appointed a Charpentier.”29
Some letters refer to his outdoor work in Kummelnäs which he did
willingly. although it often prevented him from studying. “Again
Kummelnäs,” he writes on 6 May 1937, “and this means Sakish, as well
as it may go, but the day after tomorrow we shall plant potatoes the
whole day.”30 And after a few days, on 9 May 1937 : “Now the potatoes
are in the ground and I have started presenting the meaning of one of
our Sakish pages (§§ 18–29 of a tractate which speaks of bodhisattvas
such as K!itigarbha and Mañjußr¥), unfortunately, the correct
understanding must be established from a Chinese source. But we can
look at it, can’t we?” And twelve years later, on 23 August 1949, Smith
writes, “Shortly my farm servant will come, and we shall fell three firs
with dry tops to get firewood for the winter. Farming takes at least six
27
“Du har således sett att jag befunnits skröplig nog att komma i Antikvitet
sakademien… .”
28
Probably Sköld’s Materialien zu den iranischen Pamirsprachen.
29
The whole letter runs like this : “Ja, så här ser det ut. Vill Du ha besväret att
övervaka att rättelserna inte missförståss, så kan det sedan tryckas. Men av
ark 16 måste vi ha ett 3dje korr(ektur) och så av de övriga arken. Annars blir
det Charpentier av det hela. (Salig människan ans"g att korrekturläsning hörde
hemma i underklassen — inklusive en del tyska casus- och numerusändelser.
— Och jag är ju ännu inte utnämnd till Charpentier på långa tag.”
30
“Kummelnäs igen, och det betyder sakiska, så gott det går, men i övermorgon
skall vi lägga potatis hela dagen.”
hours every day, and also other pleasant things can get in the way [to
keep me] from reading modern Hindi and Pali.”31
But what about Påli and Middle Indian ? May we guess that Smith
did not consider Jarring to be a correspondent sufficiently versed in this
field ? The letter dated (Uppsala) 11 April 1938 would perhaps not
corroborate this hypothesis, though it deals, on the other hand, more or
less exclusively with Middle Indian forms for “Turk(ish)” :
I have had trouble with a sporadic sound-law in Middle Indian — that
is to say, a sound-rule which seems to be valid for the adaptation of
learned Sanskrit words (respectively loan words) to Påli–Prakrit habits of
pronunciation, the clearest examples of which are :
Sanskrit mËrkha “idiot”, which in the regular way develops into Pkt
mukkha (rkh assimilated, u shortened before the [consonant] group) but
which (Pischel, Grammatik der Prakritsprachen § 139) also appears as
murukkha, whilst kh is geminated (a well-known rule in Sanskrit school
pronunciation) and anaptyctical u is developed.
Sanskrit pËrva “first”, etc., normally puvva (like above, but some-
times puruvva parallel with above).
If, while this rule functioned, one had borrowed the word turk, it
would have necessarily become an -a stem, hence *turka, and from it
(parallel with mËrkha) turukka. The form exists and is noted by Pischel
(op. cit. § 302), but as an example of loss of aspiration, as one normally
said Turukkha which comes from Sanskrit Turu!ka.
From this I draw a conclusion which is sure : that turukkha never had
a kkh but came directly from < turk. Secondly, I believe in the possibility
that the same turk > turukka could become turukkha in conformity with
the similar — and excuse me — therewith associated murukkha “thick-
skull” and milakkhu (°kkh sporadic in Påli) “barbarian (as concerns
language)”.
Thirdly, if the Sanskrit form Turu!ka is not older than that, it could be
a learned transformation of Turukkha. (The form Tura!ka, which is said
to exist, is without phonological value.)
31
“Nu kommer min gårdskarl, och vi skall fälla tre tallar med torrtopp till ved
för vintern. ‘Lantbruket’ tar minst 6 timmar var dag ; och även andra
angenäma saker kommer i vägen för läsningen av nyindiska och pali.”
32
You would know better what the scholars say about Turu!ka… .
The very last document of the “Collection Gunnar Jarring” is not a letter
from Helmer Smith addressed to Gunnar Jarring but the photocopy of
an undated letter Smith addressed to Nils Simonsson.33 In a few added
lines dated 30 November 1989 Jarring comments upon it as follows :
“Dear Per,34 This photocopy of a letter from Smith to Simonsson shows
Lund in a glorified light. May I suggest that you put it at the end of
Helmer Smith’s ‘“dossier’” in my collection of letters. By the way,
32
“Jag har havt bestyr med en sporadisk ljudlag i medelindiskan — d.v.s. en
ljudregel som tycks gälla för anpassningen av lärda sanskritord (resp. lånord)
till pali–prakritiska uttalsvanor — vars klaraste exempel är :
sanskrit mËrkha, som i rätlinig utveckling ger pråkr. mukkha > rkh
assimilerat, Ë förkortat [framför] gruppen, men som också (Pischel
Grammatik der Prakritsprachen § 139) uppträder som murukkha i det kh
gemineras (en känd regel för det skolmässige sanskrituttalet) och anaptyktiskt
-u- utvecklas.
sanskrit pËrva “först”, etc., normalt puvva (som ovan, men ibland puruvva
parallelt med föregående.
Om man, då den regeln verkade, hade lånat in ordet turk, skulle det med
nödvändighet ha blivit en -a-stam, således *turka, och därav (parallelt med
murkha) turukka. Formen finns och antecknas av Pischel (§ 309), men såsom
ett exempel på aspirationsförlust, eftersom det normalt heter Turukkha och
detta kommer av sanskrit Turu!ka.
Härav drar jag en slutsats som är säker : att turukka aldrig havt något kkh utan
kommit direkt < turk.
För det andra tror jag på möjligheten av att samma turk > turukka, kunnat bli
turukkha efter de snarlika — och ursäkta ! — därmed associerade murukkha
“tjurskalle” och milakkhu (°kkha sporadiskt i pali) “barbar [till språket]”.
För det tredje är sanskritformen Turu!ka icke äldre än att den skulle kunna
vara en lärd ombildning av Turukkha. (Formen Turaska, som förekommer,
saknar “fonologiskt” värde).
Vad de lärde säger om Turu!ka etc., vet Du nog bättre…”
33
Nils Simonsson (1920–1994) was Professor in Indian Languages and
Literature at the University of Oslo from 1963–1975 and Professor in Sanskrit
and Comparative Indo-European Linguistics at the University of Uppsala
from 1975 to 1985.
34
Chief librarian at the University Library in Lund, now retired.
Helmer Smith’s letters to me need not be kept in a closed envelope any
more but should be preserved in the same manner as the other letters.
Yours faithfully, Gunnar.”35 Smith’s letter (or a part of the letter) runs
as follows :
The commentary is in general well informed … but I am becoming more
and more convinced that there exist two kinds of Påli : one which was
spoken by Buddha and was written by Buddhaghosa [and] the È¥kåkåras,
Aggavaµsa, Moggalåna and other theras and was understood and read
by Fausbøll and Trenckner and others. This is the first kind of Påli ; the
other kind of Påli is more flexible and more suited to express the
fundamental doctrines of Christianity and the philosophy of Epicurus.
This is the Påli of Comparative History of Religions, it is spoken inter
alia in Lund ; it is a fortunate language, because its vocabulary is small
36
and it is not troubled by any grammar.
The contents of this document are unfortunately disappointing and

would not be accepted by any of today’s scholars of Påli. The statement
that Påli is “spoken in Lund” and thus “shows Lund in a glorified light”
is absurd and tinged by strongly local patriotism. It is rather strange that
Jarring wished this message to be incorporated into his letter
35
“Käre Per, denna fotokopia av ett brev från Smith till Simonsson kastar ett
förklarat ljus över Lund. Får jag föreslå att Du lägger den i slutet på Helmer
Smiths “dossier” i min brevsamling. Helmer Smiths brev till mig behéver f.ö.
inte längre ligga i slutet kuvert utan férvaras på samma sätt som andra brev.
Din tillgivne Gunnar.”
36
“Kommentaren vet i allmänhet väl besked … men jag blir mer och mer
övertygad om att det finns tv “Käre Per, denna fotokopia av ett brev från
Smith till Simonsson kastar ett förklarat ljus över Lund. Får jag föreslå att Du
lägger den i slutet på Helmer Smiths “dossier” i min brevsamling. Helmer
Smiths brev till mig behéver f.ö. inte längre ligga i slutet kuvert utan férvaras
på samma sätt som andra brev. Din tillgivne Gunnar.” sorters pali : en sort
som talades av Buddha och skrevs av Buddhaghosa [och] †¥kåkåraerna,
Aggavaµsa, Moggalåna och andra theraer och förstods och lästes av Fausbøll
och Trenckner mfl. detta är första sortens pali ; den andra sortens pali är
smidigare och mera ägnat att uttrycka kristendomens grundläror och Epikuros
filosofi — det är den komparativa religionsforskningens pali, det talas bl.a. i
Lund ; det är ett lyckligt språk, för dess ordförråd är ringa och det besväras
icke av någon grammatik.”
37
collection.
On the whole, the “Collection Gunnar Jarring” gives us a rather
good picture, perhaps not so much of the work on Påli, but of the
personality and the various activities of Helmer Smith. It remains,
however, surprising that Smith, in the period between 1934 and 1938,
sent somewhat more than 150 letters to Gunnar Jarring, who was not an
Indologist, though he had numerous interests in common with Smith. A
field of intense interest shared by both of them was above all
lexicography. Jarring was undoubtedly an ambitious and extremely
talented young man, to whom Helmer Smith could easily take a liking,
and Smith was probably sincere, when he in his letter dated 19 August
1935 praised Jarring by jokingly quoting two lines from a students’
theatrical parody, a “spex” :38
For you know languages which no tongue speaks,
and you can interpret what nobody thought.39
Siegfried Lienhard
37
A contemporary of Helmer Smith who mastered the Påli language with
unparalleled ease was, however, Wilhelm Geiger (1856–1943) from the
University of Munich. In his book Wilhelm Geiger : His Life and Works
(Colombo : Tübingen 1977 (2nd ed.), p. 135), H. Bechert mentions “a famous
Sinhalese scholar who had corresponded with Geiger since 1928. … This was
Aggamahåpaˆ#ita Polvatt„ Buddhadatta Mahånåyaka Thera (1887–1962).
Buddhadatta Thera gives a detailed account of his acquaintance with Geiger
in his autobiography, ”Sr¥ Buddhadattacaritaya”, in which he also gave the
text of several of Geiger’s letters. The first letter from Geiger to Buddhadatta
was written in Påli, and was later included as a reading-exercise in
Buddhadatta’s text-book of Påli (see ibid., note 52 : A.P. Buddhadatta, Aids to
Pali Conversation and Translation (Ambalangoda 1951), pp. 130f. and the
plate between 80 and 81 : “First page of a Letter written by Wilhelm Geiger to
Ven. Sri Subhuti Thera”).
38
From Latin spectaculum. “Spexes” had become popular since about 1850,
especially at the University of Lund.
39
“För du kan språk som ingen tunga talar, och du kan tyda det som ingen
tänkt.”
An Index to JPTS Volumes IX–XXIX
An Index to The Journals of the Påli Text Society (1882–1927 = Volumes
I–VIII), compiled by P.D. Ratnatunga (Mudaliyar) and revised with an
Appendix and arranged by S.S. Davidson, was published by the Society in
1973. This index lists, by author, the articles published in the Journals since it
was revived in 1981. The years of publication are: IX (1981), X (1985), XI
(1987), XII (1988), XIII (1989), XIV (1990), XV (1990), XVI (1992), XVII
(1992), XVIII (1993), XIX (1993), XX (1994), XXI (1995), XXII (1996),
XXIII (1997), XXIV (1998), XXV (1999), XXVI (2000), XXVII (2002),
XXVIII (2006), XXIX (2007).
Allon, Mark. A Gåndhår¥ Version of the Simile XXIX.229–62
of the Turtle and the Hole in the Yoke
Balbir, Nalini. The I.B. Horner Lecture 1997 : Jain– XXVI.1–42
Buddhist Dialogue : Material from the Påli Scriptures
———. Three Påli Works Revisited XXIX.331–65
Bangchang, Supaphanna. A Påli letter Sent by the
Aggamahåsenåpati of Siam to the Royal Court at XII.185–212
Kandy in 1756
Bareau, André. The Theravådins and East India IX.1–9
According to the Canonical Texts
Baums, Stefan, see Salomon, Richard
Bechert, Heinz. The Bauddhayåna of Indonesia : IX.10–21
A Syncretistic form of Theravåda
Blackburn, Anne M. Notes on Sri Lankan Temple XXVII.1–60
Manuscript Collections
Bodhi, Bhikkhu. The Sus¥ma-sutta and the Wisdom- XXIX.51–75
Liberated Arahant
Braun, Heinz. The Colophons of Burmese Manuscripts XXVII.147–53
Buddhadatta, Aggamahåpaˆ!ita Polvatte. XI.155–226
Paramatthavinicchaya by Anuruddha
Collins, Steven. Kalyåâmitta and Kalyåâmittatå XI.51–72
———. On the Very Idea of the Påli Canon XV.89–126
———. The Story of the Elder Måleyyadeva XVIII.65–96
———. See also Denis, Eugène
———. Remarks on the Third Precept: Adultery and XXIX.263–84
Prostitution in Påli Texts
Cone, Margaret. Patna Dharmapada. Part I : Text XIII.101–217
421
422 An Index to JPTS Volumes IX–XXIX
———. The I.B. Horner Lecture 1995 : Lexicography, XXII.1–34
Påli, and Påli Lexicography
———. caveat lector XXIX.95–106
Cousins, Lance S. The Pa††håna and the Development IX.22–46
of the Theravådin Abhidhamma
Crosby, Kate. Såriputta’s Three Works on the XVIII 49–59
Samantapåsådikå
———. Sa!khepasårasa!gaha : Abbreviation in Påli XXIX.169–74
Denis, Eugène and Steven Collins. Bra˙ XVIII.1–64
Måleyyadevatthera-vatthu
Dundas, Paul. A Note on the Heterodox Calendar and XXIX.76–92
a Disputed Reading in the Kålakåcåryakathå
Exell, R.H.B. RËpårËpavibhåga by Buddhadatta XVI.1–12
Filliozat, Jacqueline. Documents Useful for the XVI.13–54
Identification of Påli Manuscripts of Cambodia,
Laos and Thailand
———. A Survey of Burmese and Siamese Påli XIX.1–41
Manuscript Collections in the Wellcome Institute
———. The Commentaries to the Anågatavaµsa in the XIX.43–63
Påli Manuscripts of the Paris Collections
———. Catalogue of the Påli Manuscript Collections in XXI.135–191
Burmese and Siamese Characters Kept in the Library of
Vijayasundaråråmaya, Asgiriya
———. Survey of the Påli Manuscript Collection in the XXIV. 1–80
Bodleian Library, Oxford
———. Nine Påli Manuscripts in the Vatican Library XXVI.139–60
Gethin, R.M.L. Mythology as Meditation : From the XXVIII. 63–112
Mahåsudassana Sutta to the Sukhåvat¥vyËha SËtra
———. What’s in a Repetition? On Counting XXIX.367–89
the Suttas of the Saµyutta-nikåya
Gombrich, Richard F. A new Theravådin Liturgy IX.47–73
———. Old Bodies like Carts XI.1–3
———. Three Souls, One or None : The Vagaries of a XI.73–78
Påli Pericope
———. Two Notes on Visuddhimagga IX :1. The XII.169–71
Etymology of Puggala ; 2. An Imperfect Form in Påli
———. A Note on Ambapål¥’s Wit XV.139–40
———. Making Mountains Without Molehills : The XV.141–43
Case of the Missing StËpa
———. Why Is a Khattiya Called a Khattiya ? The XVII.213–14
Aggañña Sutta Revisited
———. The Monk in the Påli Vinaya : Priest or XXI.193–213
Wedding Guest ?
———. Report of the Påli Text Society for 1994 XXI.215–17
Grey, Leslie. Supplement to the Concordance of XXIV.103–47
Buddhist Birth Sories
Hallisey, Charles. Tuˆ!ilovåda : An Allegedly XV.155–95
Non-Canonical Sutta
———. A Propos the Påli Vinaya as a Historical XV.197–208
Document : A Reply to Gregory Schopen
———. Nibbånasutta : An Allegedly Non-Canoncial XVIII.97–130
Sutta on Nibbåna as a Great City
Hara, Minoru. A Note on vinaya XXIX.285–311
Hazlewood, Ann Appleby. A Translation of XI.133–59
Pañcagatid¥pan¥
———. Saddhammopåyana : The Gift Offering of the XII.65–68
True Dhamma
Hinüber, Oskar von. The Ghost Word Dv¥hitika and the IX.74–86
Description of Famines in Early Buddhist Literature
———. Two Jåtaka Manuscripts from the National X.1–22
Library in Bangkok
———. The Oldest Dated Manuscript of the XI.111–19
Milindapañha
———. An Additional Note on the Oldest Dated XII.173–74
Manuscript of the Milindapañha
———. Remarks on a List of Books Sent to Ceylon XII.175–83
from Siam in the Eighteenth Century
———. Khandhakavatta : Loss of Text in the Påli XV.127–38
Vinayapi†aka ?
———. The Arising of an Offence : Ópattisamu††håna XVI.55–69
———. The Nigamanas of the Suma"galavilåsin¥ and XXI.129–33
the Ka"khåvitaraˆ¥
———. Chips from Buddhist Workshops : Scribes and XXII.35–57
Manuscripts from Northern Thailand
———. The Paramatthajotikåd¥pan¥, a Fragment of the XXIII.27–41
Sub-commentary to the Paramatthajotikå II on
the Suttanipåta
———. Tuva††ati/tuva††eti Again XXVI.71–75
———. Lån2 Nå as a Centre of Påli Literature During XXVI.119–37
the Late 15th Century
———. See also Mettanando Bhikkhu.
Horner, Isaline Blew. Keci “Some” in the Påli X.87–95
Commentaries
Hundius, Harald. The Colophons of Thirty Påli XIV.1–173
Manuscripts from Northern Thailand
Hüsken, Ute. The Legend of the Establishment of XXVI.43–69
the Order of Nuns in the Theravåda Vinaya-pi†aka
Jackson, P. A Note on Dhammapåla(s) XV.209–11
———. The Canonicity of the Netti and Other Works XXVIII. 61–62
Jaini, Padmanabh S. T¥rthaµkara-prak®ti and the IX.96–104
Bodhisattva Path
———. A Note on micchådi††hi in Mahåvaµsa XXIX.153–68
Jong, Jan Willem de. Fa-Hsien and Buddhist Texts IX.105–15
in Ceylon
Jurewicz, Joanna. Playing with Fire : The Prat¥tya- XXVI.77–103
samutpåda from the Perspective of Vedic Thought
Kahrs, Eivind G. Exploring the Saddan¥ti XVII.1–212
———. Commentaries, Translations, and Lexica: Some XXIX.137–51
Kalupahana, D.J. The Philosophy of History in IX.117–26
Early Buddhism
Khantipålo, Bhikkhu. Where’s That sutta ? A Guide X.37–153
to the Discourses in the Numerical Collection
(A"guttara-nikåya)
———. See also Laurence C.R. Mills
Kieffer-Pülz, Petra. Stretching the Vinaya Rules and XXIX.1–49
Getting Away with It
Lamotte, Étienne. The Gåravasutta of the Saµyutta- IX.127–44
nikåya and its Mahåyånist Developments
Lienhard, Siegfried. On the Corresponance of XXIX.409–22
Helmer Smith and Gunnar Jarring
Liyanaratne, Jinadasa. Påli Manuscripts of Sri Lanka XVIII.131–47
in the Cambridge University Library
———. South Asian Flora as Reflected in the XX.43–161
Twelfth-Century Påli Lexicon Abhidhånappad¥pikå
———. A Påli Canonical Passage of Importance for XXII.59–72
the History of Indian Medicine
———. Sri Lankan Manuscriptology XXVIII.39–48
Lottermoser, Friedgard. Minor Påli Grammar Texts : XI.79–109
the Saddabindu and its “New” Subcommentary
Manné, Joy. Categories of Sutta in the Påli Nikåyas XV.29–87
and Their Implications for Our Appreciation of
the Buddhist Teaching and Literature
———. Case Histories from the Påli Canon I : XXI.1–34
The Såmaññaphala Sutta Hypothetical Case History
or How to Be Sure to Win a Debate
———. Case Histories from the Påli canon II : XXI.35–28
Sotåpanna, Sakadågåmin, Anågåmin, Arahat –
the Four Stages Case History or Spiritual Materialism
and the Need for Tangible Results
Matsumura, Junko. Remarks on the Rasavåhin¥ XXV.153–70
and the Related Literature
Mellick Cutler, Sally. The Påli Apadåna Collection XX.1–42
Mettanando Bhikkhu & O. von Hinüber. The Cause XXVI.105–17
of the Buddha’s Death
Mills, Laurence C.R. The Case of the Murdered Monks XVI.71–75
———. See also Bhikkhu Khantipålo
Mori, Sodo. Uttaravihåra††hakathå and Sårasamåsa XII.1–47
———. Recent Japanese Studies in the Påli XXIX.175–90
Commentarial Literature
Nihom, Max. Kåmaloka : A Rare Påli Loan Word in XX.163–70
Old Javanese
Nolot, Édith. Studies in Vinaya Technical Terms I–III XXII.73–150
(1. saµgha-kamma ; 2. adhikaraâ ; 3. månatta,
parivåsa, abbhåna)
———. Studies in Vinaya Technical Terms IV–X XXV.1–111
(4. The disciplinary procedures of tajjan‰ya-°,
nissaya-°, pabbåjan‰ya-°, pa†isåraˆ‰ya-°, and
threefold ukkhepan‰ya-kamma (n.) ;
5. Nissåraˆå (f.)/nissåraˆ‰ya (n.), osåraˆå (f.)
/osåraˆ‰ya (n.) ; 6. Nåsan¡ (n.f.), “expulsion” ;
7. Daˆ"a-kamma (n.), “punishment” ;
8. Pakåsan‰ya-kamma (n.), “procedure of
proclamation” ; 9. Patta-nikkujjan¡/°-ukkujjan¡
(n. f.), “turning down/up the alms-bowls”)
Norman, Kenneth Roy. Devas and Adhidevas in IX.145–55
Buddhism
———. Påli Lexicographical Studies III : Ten Påli X.23–36
Etymologies
———. Påli Lexicographical Studies IV : Eleven Påli XI.33–49
Etymologies
———. Påli Lexicographical Studies V : Twelve Påli XII.49–63
Etymologies
———. Påli Lexicographical Studies VI : Six Påli XIII.219–27
Etymologies
———. Påli Lexicographical Studies VII : Five Påli XIV.219–25
Etymologies
———. Index to JPTS Volumes IX–XIV XIV.227f.
———. Påli Lexicographical Studies VIII : Seven Påli XV.145–54
Etymologies
———. Index to JPTS Volumes IX–XV XV.213f.
———. Påli Lexicographical Studies IX : Four Påli XVI.77–85
Etymologies
———. Påli Lexicographical Studies X : Two Påli XVII.215–18
Etymologies
———. Påli Lexicographical Studies XI : Six Påli XVIII.149–64
Etymologies
———. Index to JPTS Volumes IX–XVIII XVIII.177–80
———. External Sandhi in Påli (with Special Reference XIX.203–13
to the Suttanipåta)
———. Påli Lexicographical Studies XII : Ten Påli XX.211–30
Etymologies
———. Book Review (Catalogue of the Burmese-Påli XXVI.161–64
and Burmese Manuscripts in the Library of the
Wellcome Institute for the History of Medicine)
———. Index of Grammatical Points Discussed in the XXVI.165–68
Notes to Elders’ Verses I
———. The Anågatavaµsa Revisited XXVIII. 1–37
Oberlies, Thomas. Påli, Påîni and “Popular” Sanskrit XXIII.1–26
———. A Study of the Campeyya Jåtaka, Including XXVII 115–46
Remarks on the Text of the Sa"khapåla Jåtaka
Pecenko, Primoz. Såriputta and his Works XXIII159–79
———. L¥natthapakåsin¥ and SåratthamañjËså : The XXVII.61–114
Puråâ†¥kås and the È¥kås on the Four Nikåyas
Penth, Hans. Buddhist Literature of Lån Nå on XXIII.43–81
the History of Lån Nå’s Buddhism
Pind, Ole Holten. Studies in Påli Grammarians I : XIII.33–81
Buddhaghosa’s References to Grammar and
Grammarians
———. Studies in Påli Grammarians II.1 XIV.175–218
Pruitt, William. Reference to Påli in Seventeenth-Century XI.119–31
French Books
———. Burmese Manuscripts in the Library of XIII.1–31
Congress, Washington, D.C.
———. Additions to the Burmese Manuscripts in the XXIV.171–83
Library of Congress, Washington, D.C.
———. The Career of Women Disciple Bodhisattas XXIX.391–407
Rahula, Walpola. Humour in Pali Literature IX.156–74
Roock, A. Index of K.R. Norman’s Collected XXVI.169–231
Papers I–VII
Ruegg, David Seyfort. A Further Note on Påli GotrabhË IX.175–77
Saddhatissa, Hammalava. Påli Literature in Cambodia IX.178–97
———. NåmarËpasamåso : The Summary of Mind XI.5–31
and Matter
———. Nåmacårad¥pikå XV.1–28
Salomon, Richard, and Stefan Baums. Sanskrit Ik#våku, XXIX.201–27
Påli Okkåka, and Gåndhår¥ Ismaho
Schmithausen, Lambert. On MahåyånasËtrålaµkåra VII.1 XXIX.199–200
Schopen, Gregory. The StËpa Cult and the Extant XIII.83–100
Påli Vinaya
———. The Ritual Obligations and the Donor Roles of XVI.87–107
Monks
———. The Buddhist Bhik#u’s Obligation to Support His XXIX.107–36
Parents in Two Vinaya Traditions
Skilling, Peter. The Rak#å Literature of the Íråvakayåna XVI.109–82
———. A Citation from the *Buddhavaµsa of the XVIII.165–75
Abhayagiri School
———. Theravådin Literature in Tibetan Translation XIX.69–201
———. Vimuttimagga and Abhayagiri : The Form- XX.171–210
Aggregate According to the Saµsk®tåsaµsk®tavinißcaya
———. The Sambuddhe verses and Later Theravåda XXII.150–83
Buddhology
———. On the School-Affiliation of the “Patna XXIII.83–122
Dhammapada”
———. New Påli Inscriptions from South-East Asia XXIII.123–57
———. A Note on King Milinda in the XXIV.81–101
Abhidharmakoßabhå#ya
———. A Note on Dhammapada 60 and the Length of XXIV.149–70
the Yojana
———. Praises of the Buddha Beyond Praise XXIV.195–200
———. The Sixty-Four Destructions According to the XXV.112–18
Saµsk®tåsaµsk®tavinißcaya
———. Contents of the New Edition of the
Syåmara††hassa Tepi†aka††hakathå XXVII.153–56
———. Some Citation Inscriptions from South-East XXVII.157–72
Asia
———. Jåtaka and Paññåsa-jåtaka in Southeast Asia XXVIII. 113–73
———. Zombies and Half-Zombies : MahåsËtras XXIX.313–30
and Other Protective Measures
Somaratne, G.A. Intermediate Existence and the XXV.119–52
Higher Fetters in the Påli Nikåyas
Stargardt, Janice. The Oldest Known Påli Texts, XXI.199–213
Fifth–Sixth century. Results of the Cambridge
Symposium on the Pyu Golden Påli Text from
Ír¥ K#etra, 18–19 April 1995
Thiradhammo Bhikkhu. Corrections to The Book XIX.65–68
of the Discipline
Warder, A.K. Some Problems of the Later Påli IX.198–207
Literature
CONTRIBUTORS TO THIS VOLUME
Professor Dr Oskar von Hinüber Professor Gregory Schopen

Kartäuserstr. 138 Dept of Asian Languages and
D–79102 Freiburg Cultures
Germany 290 Royce Hall
University of California, Los Angeles
Dr Petra Kieffer-Pülz Box 951540
Wilhelm-Külz-Str. 2 Los Angeles, CA 90095
99423 Weimar U.S.A.
Germany
Dr Eivind Kahrs
Venerable Bhikkhu Bodhi Faculty of Asian and Middle Eastern
Chuang Yen Monastery Studies
2020 Route 301 University of Cambridge
Carmel NY 10512 Sidgwick Avenue
U.S.A. Cambridge
CB3 9DA
Dr Paul Dundas U.K.
Department of Sanskrit
University of Edinburgh Professor Padmanabh S. Jaini
7 Buccleuch Place Department of South & Southeast
Edinburgh Asian Studies
EH8 9LW 7233 Dwinelle Hall # 2520
U.K. University of California,
Berkeley, CA 94720–2520
Dr Margaret Cone U.S.A.
Faculty of Asian and Middle Eastern
Studies Dr Kate Crosby
University of Cambridge Dept of the Study of Religions,
Sidgwick Avenue SOAS
Cambridge Thornhaugh Street
CB3 9DA Russell Square
U.K. London
WC1H OXG
U.K.
429
430 Contributors
Professor Sodo Mori Professor Dr Nalini Balbir
No.3-13-50 Saiwai-cho, Shiki-shi 32 rue des Bruyères
Saitama 353-0005 F–92310 Sèvres
Japan France
Professor Dr Lamber Schmithausen Dr R.M.L. Gethin

Dobbertinweg 24 Dept of Theology and Religious
21266 Jesteburg Studies
Germany University of Bristol
3 Woodland Road
Professor Richard Salomon Bristol
Department of Asian Languages and BS8 1TB
Literature U.K.
University of Washington
225 Gowen Dr William Pruitt
Box 353521 36 Lake View
Seattle, WA 98195–3521 Calne, Wiltshire
U.S.A. SN11 8JA
Dr. Peter Skilling Professor Siegfried Lienhard

129/123 Muu 3 Perfect Place August Wahlströms väg 1, 8 tr
Soi Sai Ma, Ratanathibet Road S–18231 Danderyd
Bang Rak Noi, Muang District Sweden
Nonthaburi 11000
Thailand
PTS Research Grants
In keeping with the aim of the Pali Text Society to foster and
promote the study of Pali texts, the Council of the PTS invites
applications for PTS Research Grants from suitably qualified
persons working in the field of Påli studies. Applicants’ research
will normally be expected to lead towards material suitable for
publication by the PTS, and conditions of any grant awarded will
be that the PTS has first option on the publication rights of the
research and holds the copyright of any material it publishes.
Applications should be submitted by 28 February and 31
August of each year for consideration by the Council of the PTS
in March and September respectively. Applicants are asked to
submit (1) a statement briefly outlining their research project
(two A4 pages with an appendix if necessary) clearly stating the
purpose of their research, indicating a timetable for its
completion, and stating whether they will be working on the
project full-time or part-time, (2) a CV, (3) the names of two
academic referees, (4) a statement of their financial circum-
stances. Grants will not normally be awarded to those in full-time
academic employment.
Grants may be awarded to cover a period of research (full-
time or part-time) up to a maximum of one year, but with a
possibility of renewal. Applications should be sent either by
email ([email protected]) or by post to the President of the Pali
Text Society, 36 Lake View, Calne, Wilts SN11 8JA.
431

JPTS 2007 Xxix

Uploaded by

Document Informationclick to expand document informationJPTS_2007_XXIX

Document Informationclick to expand document information

Copyright:

Available Formats

JPTS 2007 Xxix

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

JPTS 2007 Xxix

Uploaded by

Copyright:

Available Formats

Journal of the Pali Text Society

Pali Text Society

PALI TEXT SOCIETY

Festschrift in honour of the 80th birthday of

All rights reserved. Subject to statutory exceptions, no part of this

© Pali Text Society 2007

isbn 0 86013 438 5

First published in 2007

Printed in Great Britain by

(Photo courtesy of Geoffrey Wells)

Two particularly auspicious events in the history of the Pali Text

anniversary of the foundation of the Pali Text Society in 2006, and in

Jacobi (1850–1937), “The Âyâraµga Sutta of the Çvetåmbara Jains”

3A translation of the Jåtakamålå by Jacob Samuel Speyer (1849–1913) “The

member of the board in J P T S 1 9 0 7 /8 ) from 1 9 2 3 until 1942.

from 1 9 5 0 to 1958 ; William Henry Denham Rouse (1863–1950),

President from 1942–1950 and immediate successor to the Rhys Davids

6After his death, Fausbøll was succeeded by Dines Andersen (1861–1940).

translation, both by Jinadasa Liyanaratne. Very recently in 2005, the

14IIJ35 (1992), parts 2/ 3 , pp. 81–272 with a bibliography only of books

* This article is an outcome of my work on “Die in der Vajirabuddhi-†¥kå

See Kieffer-Pülz 2001.

be handled according to the guidelines (mahåpadesa) handed down in the

with the 227 rules constituting the Påtimokkha, to be recited every

6The mention of an atikhuddakå s¥må presupposes a definition of the smallest

own decision as to the authority he attributes to a certain text. This

9Mhv 33,100–101!: pi†akattayapåliµ ca tasså a††hakathaµ pi ca mukhapå†hena

two from Burma written in the seventeenth and nineteenth centuries

12For instance the Anugaˆ†hipada or the Vajirabuddhi-†¥kå on the opinion of

katå. dhammasa!gåhakehi ca på†hasa!g¥tiµ ni††håpentehi sanni††hånaµ

2. Examples of stretching the rule

16Sp V 995 ,15ff. ad Vin I 71,32–73 ,20 (BD IV 89ff.).

might have prevented his ordination as a monk. The list of possible

have his parents’ permission ; he has to be at least twenty years old ; he

(ku††ha), boils (gaˆ"a), eczema (kilåsa), consumption (sosa) or epilepsy

108f.)!; one who gained access to the community by theft (theyyasaµvåsaka),

sheds a Tathågata’s blood (lohituppådaka), and a hermaphrodite (ubhato-

upasampadå, Vin I 93,24–32 (BD IV 120)!; Sp does not comment on it.

(BD IV 93)!; Sp V 997,10ff. !; on a thief broken out of jail (kårabhedaka cora),

hata), whose movements are destroyed (chinna-iriyåpatha), who are weak of

“You are not a slave!?”, and an additional question is må vikr¥takaß (Tib.

såmaµ vå dåsavisayaµ upeti, akåmako vå dåsavisayaµ upeti.

††håpenti, dåsiµ upa††håpenti ... na dåso upa††håpetabbo, na dås¥ upa††håpe-

The eighth Prak¥rˆaka of the Mahåsåµghika-Lokottaravådins quoted by

donation, and this presumably could not be an acceptable behaviour for a

§!187 , Gunawardana 1979, pp. 97ff.

in the Brahmajåla-sutta and elsewhere, D I 5 ,14f.!: dåsidåsapa†iggahaˆå pa†i-

dåsidåsapa†iggahaˆå pa†ivirato hot¥ ti (D I 5 ,14f.) vacanato. “Even for one

may be accepted by the sa"gha if they are part of the donation of a

c’ eva dåsapa†iggahaˆadukka†añ ca hoti dåsidåsapa†iggahaˆå pa†ivirato

gomahisaµ demå ti vadanti, på†ekkaµ gahaˆakiccaµ natthi. påsåde

42Vin-vn, v. 665 !; Sp III 683 ,17– 18, see below, n. 43.

44A II 78,31 (åråmikasamaˆuddesesu)!; III 109 ,31, 32 (a prophecy that in future

II 232 ,8 (BD V 322, uddåna to the preceding). It is hinted at in the word-by-

twice in the word-by-word commentaries!: Niss 18 Mk (prohibition of the

mediator between king and monks ; 56 as one who guards the

are similar to those of a novice. 59 According to the Samantapåsådikå,