A HISTORY OF MINDFULNESS

How Insight Worsted Tranquillity in the Satipatthana Sutta

Bhikkhu Sujato

Brahmavamsassa, hasapannassa: tassa sisso'hamasmi

1
CONTENTS

Foreword 3

Part 1: The GIST: The Hidden Structure of t h e Buddha's Teachings

1. The Meaning o f ' B u d d h a ' 8

2. The GIST 1: Three Strata of Early Texts 13
3. The GIST 2: The Agreement of t h e Three Strata 26
4. The GIST 3: The Angas 36
5. The GIST 4: The Nikayas/Agamas 48
6. The Evolution of t h e Truths 65
7. What Happened in t h e Sattapani Cave? 72

Part 2: A History of Mindfulness

8. Introduction to Mindfulness 78
9. Previous Studies 82
10. Pre-Buddhist Meditation 90
11. Building Blocks 117
12. The Samyutta 132
13. Early A b h i d h a m m a 150
14. The Satipatthana Suttas 164
15. The Source 183
16. Trends: Theravada and Sarvastivada 192
17. Later Buddhism 204

Afterword 224

Appendix 1: The Samyutta-matika 227

Appendix 2: The Ironic Assumptions of Gregory Schopen 229
Bibliography 238
Abbreviations 233
Index 234
Notes 246
FOREWORD

T H E VIPASSANAVADA

The purpose of this book is to analyze the textual sources of 20th Century Theravada meditation
theory. The focus is on the prime source works for what may be called the vipassanavada, the
'vipassana-doctrine'. This is a special interpretation of some central meditation concepts that has
become the de facto orthodoxy in Theravada Buddhism, although not without controversy. The
term vipassanavada is useful in that the Pali suffix -vada points to the crucial importance of the
underlying theory thatjustifies the practices. More than that, the same suffix comes to mean not
just a doctrine, but also the school that follows the doctrine. This is all too apposite in the current
case, since 'vipassana' has come, rather strangely, 1 to be used as if it refers to an actual school of
Buddhism (rather than an aspect of meditation cultivated in all schools).

The key points of the vipassanavada are reiterated countless times in almost every book on 20th
Century Theravada meditation, but here they may be summed up briefly. The Buddha taught two
systems of meditation, samatha and vipassana. Samatha was taught before the Buddha (so is not
really Buddhist), it is dangerous (because one can easily get attached to the bliss), and it is
unnecessary (because vipassana alone can develop the access samadhi necessary to suppress the
hindrances). Vipassana is the true key to liberation taught by the Buddha. This method was pre-
eminently taught in the Satipatthana Sutta, the most important discourse taught by the Buddha
on meditation and on practice in everyday life. The essence of this practice is the moment-to-
moment awareness of the rise and fall of all mind-body phenomena. Thus satipatthana and
vipassana are virtually synonyms.

It is worth noticing that not all modern meditation traditions accept this dichotomy of samatha
and vipassana. For example, the teachers of the Thai forest tradition often emphasize the
complementariness, rather than the division, of samatha and vipassana. I have heard an
illustration of this attributed to the late meditation master Ajahn Chah. In samatha, you sit down
cross-legged, close your eyes, watch your breath, and make the mind peaceful. But vipassana,
now, that's something quite different. In vipassana, you sit down cross-legged, close your eyes,
watch your breath, make the mind peaceful, and then you know: 'It's not a sure thing!' I have
always felt that in this question of meditation there was a strong affinity between the Suttas and
the teachings of the forest masters. Works like the present one are my attempt to demonstrate, in
my clumsy, pedantic, and long-winded way, the same truths expressed with such pith and
authority by masters like Ajahn Chah.

In 2000 I wrote A Swift Pair of Messengers, emphasizing the harmony and complementariness of
samatha and vipassana. There, I discussed at some length the treatment of satipatthana as found
in the early Suttas, focusing on the Satipatthana Sutta. The purpose was to demonstrate that
satipatthana, far from being a distinctive or separate mode of development, was embedded both
deeply and broadly in the meaning-matrix of the early Suttas and could neither be understood
nor practiced outside of this context.

Nearing the end of that project I came across an article by Richard Gombrich entitled 'Retracing
an Ancient Debate: How Insight Worsted Concentration in the Pali Canon'. 2 Although only
partially convinced by his arguments, I was intrigued by his idea - that the shift in emphasis from
samadhi to vipassana, so obvious in later Theravada, could be traced back to editorial changes
made within the period of compilation of the Pali Nikayas. Itjolted some memories of a few loose

3
ends left dangling in my study of satipatthana. I decided to tug on those strands of thought, and
to my amazement the whole Satipatthana Sutta started to unravel before my eyes. This is the
story of how the Satipatthana Sutta was woven, how it unravels, and how this affects our
understanding of Dhamma-Vinaya.

The significance of such a historical approach to the teachings is still largely unrecognised among
practicing Buddhists. In fact, our normal approach to the teachings is the very opposite of
historical. An aspiring meditator first learns f r o m the lips of a teacher whose words as they utter
them must be the very latest formulation of the topic. Then they might go back to read some of
the works of well-known contemporary teachers. Since devotees usually have faith that their
teacher (or the teacher's teacher) was enlightened, they assume, often without reflection, that
the teachings must be in accord with the Buddha. Finally, if they are really dedicated, they may
go back to read 'the' Satipatthana Sutta. Once they come to the text itself, they are already pre-
programmed to read t h e text in a certain way. It takes guts to question t h e interpretation of one's
teachers; and it takes n o t j u s t guts, but time and effort to question intelligently.

Apart f r o m 'the' Satipatthana Sutta, t h e other discourses on satipatthana, being so much shorter,
are usually ignored under the assumption that they add little new. Even the best of t h e scholars
who have studied satipatthana from a historical perspective, such as Warder, Gethin, and
Analayo, have treated the Satipatthana Sutta as primary and the shorter discourses as
supplements.

So now I would like to reverse that procedure. Our first step must be to forget all we've learnt
about satipatthana, and to start again f r o m the bottom up. A basic principle of the historical
method is that simpler teachings often tend to be earlier and hence are likely to be more
authentic - we must start with the bricks before we can build a house. It is the shorter, more
basic, passages that are the most fundamental presentation of satipatthana. The longer texts are
an elaboration. We do not assume that shorter is always earlier, but we take this as a guiding
principle whose implications we can follow through.

This stratification, it should be noted, does not claim to be able to decide which teachings were
genuinely spoken by the Buddha. He himself would likely have given the same teachings initially
in simple form, then later expanded on various details. But the universal testimony of t h e
traditions is that the texts as we have t h e m today were assembled in their present form after the
Buddha's passing away; so the rational approach is to assume that the texts were the outcome of
an evolutionary process.

Those who disagree with this approach will usually do so for one of two reasons. Either they have
faith that all of the teachings in the Suttas were literally spoken by the Buddha; or they doubt t h e
possibility of meaningful historical reconstruction due to the unreliability of the sources or the
uncertainty of the method. I believe the first position is too credulous and the second too
sceptical. In any case, even if our method fails to reflect the genuine historical situation, it
remains useful as a pedagogical technique. Whether in the Buddha's day or in the after-years, it is
surely sensible to learn the Dhamma by starting with the simple teachings and working up to the
complex.

So we should start by identifying the smallest, simplest units of meaning used to describe
satipatthana. These are the basic terms and phrases common to all descriptions of satipatthana in
all the schools. It would make sense to start with the Buddha's first sermon. This raised an
interesting question. This sermon is for the group of five monks, who were, at the time, non-
Buddhist ascetics. However, the text refers to mindfulness as if it assumes the audience would
know what it means. Mahasi Sayadaw noticed this, and felt that the discourse should have

4
originally included a more detailed explanation of satipatthana. But I felt that this was
contextually unlikely, for the Saccavibhanga Sutta, which explicitly explains the
Dhammacakkappavattana Sutta 'in detail', includes the standard satipatthana formula in this
detailed expansion. What need for such an expansion if t h e formula was there in the original?
The conclusion seems inescapable: the Dhammacakkappavattana Sutta assumes that the five
monks already knew what mindfulness was, and so f r o m the perspective of early Buddhism,
mindfulness was a pre-Buddhist practice. To check this I felt I had to examine more closely the
accounts of pre-Buddhist meditation found in both the Buddhist and non-Buddhist texts.

I was first alerted to the possibility of historical change in the Satipatthana Sutta by A.K. Warder,
who refers to versions of the Satipatthana Sutta in ancient Chinese translations. After recording
the major differences he notes in connection with the contemplation of dhammas that 'the
original text simply opposed these good principles [enlightenment-factors] to the obstacles.' 3 It is
through such seemingly innocuous remarks that I have become cognizant of the truly
momentous significance of the comparative study of t h e Nikayas and Agamas. While the
Theravada Nikayas will forever remain our primary source for exploration of pre-sectarian
Buddhism, the Agamas of the contemporary Sarvastivada, Dharmaguptaka, and other schools,
which are preserved in ancient translations in the Chinese canon, provide an essential and under-
utilised check on the Pali. As the Encyclopaedia of Buddhism puts it: 'In our days it is impossible
for any scholar to refer to early Buddhism unless he pays due regard to the comparative study of
the southern and northern traditions.' I would be gratified if this survey could at least
demonstrate that the early Nikayas/Agamas are not a mined-out field whose treasures are all
safely housed in the later compendiums.

In the Nikayas/Agamas it is obvious that no one text pretends to present an all-encompassing,

definitive exposition, so each text must be considered in relation to the collection as a whole. This
raises questions of the overall structure and organization of the canon. I began to suspect that the
shorter texts in the Samyutta may preserve an earlier perspective on satipatthana, a perspective
that in some respects was better reflected in t h e Chinese versions of the Satipatthana Sutta than
in the Pali. My suspicions were f u r t h e r aroused by a comment by Bhikkhu Varado on the laxity of
the compilers of the Abhidhamma Vibhanga in omitting much of the material f r o m the
Satipatthana Sutta. Maybe, I wrote back, they weren't lax at all - maybe the Satipatthana Sutta
had not been written when the relevant portions of the Vibhanga were compiled. (I later found
out that I was not the first to raise this question.) This suggested that I should take account of the
early Abhidhamma texts as well as the Suttas, overcoming, in part, my prejudice that t h e
Abhidhamma was a late and sterile body of sectarian dogmas. And then, if the early Abhidhamma
period overlapped with the composition of t h e Satipatthana Sutta, it seemed likely that sectarian
agendas would be involved in the settling of the final text. This called for a closer examination of
the ways the emerging sectarian disputes found expression in the early texts. It then seemed
appropriate to extend the survey to the later Abhidhamma/commentarial period, to try to gain a
deeper insight into the ways the traditions adapted satipatthana to their own particular
perspective on t h e Dhamma, and to bridge the gap between the Buddha's time and our own.

To make use of the material on satipatthana translated f r o m the Chinese, I had to improve my
own understanding of the Agamas and other early Buddhist sources outside the Pali. When
examining and comparing these collections, with their very strong connections but also real and
persistent differences, there is a strong need for a resolution, like the feeling in music that comes
when two notes are very close together, but not quite: they yearn to become one. At this time I
was fortunate enough to get to know Roderick Bucknell. Informed by his comparative work on
the structure of the early scriptures, I have endeavoured to clarify the relations between t h e
various types and strata of texts in the early canon, and have formulated this as the GIST. This

5
theory suggests a specific relationship and hierarchy of texts in the existing canons, a
relationship that reflects both the doctrinal importance and the historical provenance.

I realized that the methodology that I had been using, largely intuitively, in studying satipatthana
followed closely the outlines of the GIST. No doubt my belief that this approach had been fruitful
in the context of satipatthana prepared me to accept that it could be extended to a general
interpretive theory. So I decided to include a presentation of t h e GIST together with the study on
satipatthana, although here the general theory is presented first. These two parts substantially
reinforce each other. The study of satipatthana provides a detailed examination of an important
doctrine along the lines suggested by the GIST. It exemplifies the method, providing additional
evidence for many of the basic principles of the GIST, and showing that the GIST does produce
meaningful and useful results. Despite this, however, it should be made clear that the two parts
are not mutually dependent. If my analysis of satipatthana turns out to be misguided, this
weakens but does not destroy the evidence in support of the GIST. Likewise, if the GIST is felt to
be unacceptable, this weakens but does not destroy the evidence in support of the analysis of
satipatthana. To some degree, the two studies may be considered independently; but taken
together they are more meaningful.

We must bear in mind that we are not here embarking on a search for certainty. As long as we
stay in the realm of concepts our ideas can only ever approximate the truth. What is important is
that we are moving in the right direction, moving away f r o m confusion towards clarity, away
f r o m dogmatism towards inquiry. Each of t h e criteria employed in historical criticism when taken
individually is an imperfect tool. But they are synergistic: where several criteria agree, the
concurrence multiplies our confidence in our conclusions - t h e whole is greater than the sum of
the parts. So in these studies it is imperative to use as wide a variety of criteria as possible,
sensitively appraise the reliability of each criterion in the relevant context, remain alive to any
contrary indications, and make our conclusions no more certain than the evidence warrants.

I endeavour to make matters no more technical and specialized than necessary, without
sacrificing precision. The translations are from various sources. Research was carried out in many
places - monasteries, libraries, Buddhist centres, internet - and I regret to say that I am not able
to standardize or to check all the references. Special thanks are due to Bhikkhu Fa Qing, who gave
much of his time to help me explore some arcane corners of the Chinese canon, and to Roderick
Bucknell for much illuminating information and challenging ideas. It is to Rod that I owe most of
the detailed information on such texts as the Dharmaskandha, Sariputrabhidharma,
Kayagatasmrti Sutra, and much else. Towards the end of this project, I began a correspondence
with Venerable Analayo, whose inspiring example, encouragement, and practical help was
instrumental in enabling me to get a handle on reading the Chinese canon. Gone are the days
when the Chinese canon was a forbidding, inscrutable place like a cloistered temple perched on a
mountain crag, shrouded in mist. Now anyone, with a little patience, can begin reading with the
excellent CBETA digital canon, and efficient reading software. Of course, it will still take years to
gain a sensitive fluency with the language, but here the primary aim is simply to check the
translations against the Indian idioms.

I have tried to maintain consistency of renderings of technical terms, and have sometimes taken
the liberty of bringing the renderings in quoted passages into line with the main text. Since t h e
Pali canon is the backbone of this work, and since I am more familiar with Pali, I have rendered
almost all Indian words in their Pali rather than Sanskrit form. Exceptions include proper names
and terms that are unknown in Pali in the relevant meaning, although I cannot claim complete
consistency in this regard. Crucial technical terms such as samatha, vipassana, satipatthana, etc.
are a normal part of contemporary Buddhist meditation vocabulary, so I treat t h e m as anglicised
forms without diacritical marks.

6
A peculiar difficulty of this work is that it discusses a large number of texts with confusingly
similar titles. I have tried to minimize confusion by spelling out names and affiliations of texts
mostly in full.

It is common practice among scholars to refer to the texts by their language, as for example t h e
'Pali Majjhima Nikaya' and the 'Chinese Madhyama Agama'. This conveys t h e entirely misleading
impression that the Agamas, and indeed all the Indian Buddhist texts that happen to be available
to us in Chinese translation, are in some sense 'Chinese'. We might as well refer to the 'English
Majjhima Nikaya' simply because we happen to be reading an English translation. What matters is
the meaning; and this is more significantly affected by the redactors' doctrinal perspectives than
by their language. It is therefore preferable to classify texts according to school whenever
possible. It is, of course, still necessary to refer to the 'Chinese canon', since the texts therein
derive from many schools, and the collection as a whole is obviously a Chinese artefact.

I have tried to give references to all known versions of a particular text, which usually means t h e
Pali and the Chinese version. Readers should be aware that this refers to a text identified as
cognate 8 in available concordances. It does not imply that the particular term, phrase, or idea
under discussion is found in all versions. I have, however, checked as many significant references
as possible, and have indicated relevant differences.

Historical criticism is not nice. I am afraid this presentation may sometimes appear rather more
surgical than inspirational. Relentless analysis can seem opposed to faith. But this need not be so;
the Buddha regarded reason as the foundation of true faith. One who has true faith in the
Dhamma would surely not fear that mere literary criticism could destroy the teachings. And is it
n o t j u s t fear that wishes to protect one's sacred scriptures, to enshrine t h e m on a pedestal, to lock
t h e m safely away in a gorgeous chest on one's shrine, safe from any impious inquiry? Thankfully
such fear, while certainly not absent, does not predominate in contemporary Buddhist circles.
And our findings, no matter how cruelly we wield the scalpel, do not affect the fundamentals of
our faith. There is a massive concurrence between the early sources of Buddhism as to the central
teachings - n o t j u s t t h e ideas and principles, but the specific texts and formulations as well. The
discrepancies we shall notice in our explorations undermine not these fundamentals, but certain
implications and trends discernable in t h e arrangement and emphasis of the more developed
formulations. Even here the differences, to begin with, are slight and few in number. So it is my
intention, not to raise doubts, but to encourage the maturing of faith.

7
 PARTI

THE G I S T :
THE HIDDEN STRUCTURE OF THE BUDDHA'S
TEACHINGS

'Friends, just as the footprint of any creature that walks can be placed within an elephant's
footprint, and so the elephant's footprint is declared the chief of them because of its great size -
so too, all skilful principles can be included within the four noble truths.'
(MN28.2/MA30)

CHAPTERI: THE MEANING OF 'BUDDHA'

Buddha.

It is said that even to hear this word is precious beyond reckoning. Through countless aeons,
beings fall into ruin since they are denied the opportunity of hearing it. Finally, after an
incalculably long time, the Enlightened One arises in the world and the word 'Buddha' is heard,
like a shower of rain in the parched desert. When the merchant Anathapindika heard this word
he was overwhelmed - his hair stood on end, he could not sleep at night, his heart leapt up with a
strange exultation. Throughout the millennia since the Buddha's time, this word has developed a
unique aura, a spiritual charisma that lends unparalleled prestige to the religious communities
and institutions that proclaim their allegiance to his liberating teaching. We are the spiritual
heirs of that great being, that man of flesh and blood who walked the rich soil of the Ganges plain
nearly 2500 years ago.

The very words we use to speak about Dhamma, including this word 'Buddha', are bound and
limited by the Indo-Aryan culture in which the young Siddhattha Gotama grew up. An
etymologist could tell us that 'buddha' derives from an ancient Indo-European root, whose basic
meaning is to 'wake up', and which has several cognates with related meanings in existing
European languages. A grammarian could tell us that it is a past participle formed from a verbal
stem. A philosopher of language might find it significant that the past participle, which is
unusually common in Buddhist language, denotes the arrival at or emergence into a certain
condition, rather than an eternal, timeless state of being. A historian of religion could tell us that
the title 'Buddha' is used to denote an enlightened or consummate being in several religions, such
as Jainism and Brahmanism, as well as Buddhism. A meditation teacher, on the other hand, might
emphasize how 'Buddha' refers to the intrinsic quality of awareness. And so on. All of these
aspects inform and condition the resonance of the word 'Buddha'; they are part of the meaning of
'Buddha'.

8
It is t h e shared allegiance to this 'Buddha' that defines the Buddhist religion. All forms of
Buddhism, from the Buddha himself down through all the schools, have acknowledged two facets,
or rather phases, in arriving at true wisdom. First comes hearing the teachings, t h e words of
t r u t h that ultimately stem from the Buddha himself; and second is the application, investigation,
and verification of those teachings within our own immediate experience. First we hear the
Buddha teach us of the four noble truths - suffering, the cause of suffering, the end of suffering,
and the way of practice leading to the end of suffering - then we look into our own minds. 'Yes!'
we realize. 'There it is, right there! My own attachments, stupidity, and hatred causing this
welling up of suffering and anguish in my heart, and making me speak and act in harmful, foolish
ways, imposing my own pain on others.'

So this inseparable pair, the theory and the practice of Buddhism, each balancing and informing
the other. Theory without practice becomes a mere intellectual mind-game; while practice
without theory tends to drift without direction, or rather, directed by the personal delusions of
the individual. It should hardly need restating that all Buddhists of all times agree that
intellectual knowledge of the Dhamma is insufficient. Intellectual knowledge, due to the ripples
in awareness stirred up by the activity of thinking, must disturb clarity of understanding, and
deep insight arises only when the mind is still and silent. But intellectual knowledge has its uses;
it is not a problem in and of itself. It only becomes a problem when we mistake our intellectual
knowledge for the truth, our opinions for reality. Then opinion becomes conceit, and we easily
succumb to a spiritual arrogance that is very hard to cure. But a skilful meditator, alert to the
distorting potential of ideas and preconceptions, learns to engage fully in the present moment,
seeing the impermanence and emptiness of thought, and growing wise in the ways that, at t h e
most fundamental levels of consciousness, even the most sublime and refined mental constructs
limit the power of awareness.

It might be useful to consider here the training of a musician. Perhaps one has been inspired by
some great composer or player to take up an instrument. But how to start? I remember being in a
music shop one time, when a student walked in and said he wanted to sound like Mark Knopfler,
who was at the time the most popular guitar player, famed for his delicate and emotive melodic
phrasing. Unfazed, the shop-owner sold the student a $30 fuzz box, and the student walked out a
happy man. Sadly, it's not that easy. In reality, we must spend many hours learning to read music
off a page, stark black-and white dots and lines that share nothing of the warmth and colour that
was our inspiration. Taking up our instrument, there are countless hours of scales, exercises, and
trivial studies to be mastered before anything vaguely approaching 'music' is heard. But once the
technique is mastered, it must be left behind. There is little that is worse in music than hearing a
musician egotistically showing off his technical skill. All the technique, the study, the practice,
must be forgotten as the artist immerses himself in the art that is created there in the present
moment; but paradoxically that present m o m e n t is only made possible by the previous study and
application. In this way, the experience of t h e past creates the magic of the present.

T H E AMBIGUITY OF TRADITION

So it is implicit in t h e claim to be a 'Buddhist' that one believes that the Dhamma descends f r o m
the Buddha himself through the transmission of his teachings by the traditions. I believe it is
worth taking this claim seriously. As a monk I am aware that, in a very real sense, I am the
material as well as the spiritual heir of the Buddha. Faithful Buddhists are willing to offer me rice
and curries, just as in the past t h e people of India offered Siddhattha Gotama rice and curries,
because they take me for a genuine follower, a 'Son of the Sakyan'. It would be insincere, even
fraudulent, for me to eat that alms food while at the same time believing, practicing, or teaching
things that I knew that Siddhattha Gotama would not agree with.

9
This raises some interesting, and challenging, problems. It is obvious that the existing cultures
that all claim to be 'Buddhist' vary widely in their beliefs and practices. Now, in many cases these
are j u s t cultural variations as the Dhamma-Vinaya adapted itself to time and place. Taiwanese
Buddhists do their chanting in Mandarin, while Thai Buddhists do theirs in Pali; no one makes a
big deal of such things. After all, t h e Buddha himself urged his followers to learn the Dhamma in
their own language, and not to insist on local dialects.

However, other aspects of cultural Buddhism are deeply opposed to the Dhamma. A disturbing
example of this is the use of Buddhist language and concepts to justify war. This has occurred
frequently in many Buddhist countries, be it Tibet, China, Thailand, Burma, Japan, and even today
in the tragic civil war in Sri Lanka. When Japanese Zen masters teach that to shoot and kill is an
expression of highest enlightenment, or when Sri Lankan monks invoke nationalist mythologies
t o j u s t i f y bloody warfare, we know, without a doubt, that something has gone badly wrong. This
is no innocuous cultural adaptation, but a radical perversion of the Buddha's teachings.

Such uncomfortable facts demand that we stop and examine the traditions more closely. It is
simply not good enough to accept with unexamined trust the myths, the stories, and the dogmas
of the schools. As people who have a commitment to understanding and practicing the liberating
message of the Sakyan Sage, there is an obligation to honestly enquire as to what, exactly, our
Teacher taught. We know that the traditions got it disastrously wrong in at least some cases. But
these clear, unambiguous examples are in t h e minority. There is a wealth of other teachings
presented to us by the schools, some of which differ f r o m each other in the letter; and we need
something better than blind faith before we can intelligently conclude whether they do, or do
not, also differ in the meaning.

All existing schools of Buddhism share a large mass of teachings in common, and yet also include
a large mass of divergent teachings. There is no doubt that the founders and developers of the
various schools believed that there were genuine, meaningful doctrinal differences between the
schools. All the schools agree that they disagree. This is amply demonstrated by t h e large
quantity of polemical material filling the shelves of Buddhist canons. And, by and large, the
schools also agree on what they disagree about. A text of the Theravada school might allege that
the 'person' doctrine of the Puggalavada school contradicts the teaching of not-self; while the
texts of the Puggalavada will vigorously argue that the teaching of the 'person' is in fact t h e
correct way to interpret not-self. Given this situation, it would seem a trifle rash to claim, as some
modernist Buddhists do, that there really are no differences, or that the differences are not
significant. What is needed is not such bland platitudes but an improved methodology, a way of
approaching the teachings that is derived, not f r o m the perspective or doctrines of any particular
school, but f r o m the sensitive evaluation of the textual tradition as lived by Buddhists. Yin Shun,
the renowned scholar-monk of modern Taiwanese Buddhism, evidently felt the same need,
according to these statements in his autobiography.

'Although "non-dispute" is good, expediently-rendered syncretism that does not know where and
why the disparities are could be far-fetched, too general, and vague.'

'To understand the origin and transformations of the Buddha Dharma within certain temporal
and spatial contexts in the actual world gradually became the principle of my quest for t h e
Buddha Dharma.'

10
T H E DEATH OF M Y T H

It is a striking feature, common to all the schools, that they feel the need to justify their
particular doctrines mythologically - this is what all religions do. For 2500 years, Buddhism has
been constantly changing, adapting, evolving; yet the myths of the schools insist that the
Dhamma remains the same. Thus the Theravada insists that the Theravada Abhidhamma was
taught by the Buddha in Tavatimsa heaven during his seventh rains retreat. The Mahayana claims
that t h e Mahayana sutras were written down in the time of the Buddha, preserved in the dragon
world under the sea, then retrieved by Nagarjuna 500 years later. Zen claims authority f r o m an
esoteric oral transmission outside the scriptures descended f r o m Maha Kassapa, symbolized by
the smile of Maha Kassapa when t h e Buddha held up a lotus. All of these are myths, and do not
deserve serious consideration as explanations of historical truth. Their purpose, as myths, is not
to elucidate facts, b u t to authorize religious convictions. They tell us, not how the teachings came
to be, but how the devotees felt about them. In this way, myth offers an irreplaceable
complement to history, and should never be disregarded. What I am criticizing here is not myth
as myth, b u t myth as history: the naive fallacy of insisting that the stories of the traditions are
factual. The myths stand as a flagrant denial of impermanence, and so a sub-theme of this work is
to notice the poignant irony of how the very effort to preserve the teachings, so that 'the true
Dhamma may last a long time', tends towards a reification of time.

It is one of the great lessons of history that reason displaces myth. There is something about the
h u m a n mind that cannot continue to believe in a mythic explanation for what that can be
understood though reason. Mythic explanations fulfill a purpose; they create a sense of meaning
and communal identity that is gratifying and self-affirming. But reason too is a positive force,
since it assumes that the human mind is capable of approaching truth. As rational explanations
for religious claims are progressively advanced, it becomes more and more wearying to sustain
two incompatible belief structures side by side. The myths fall into disuse. Being no longer
inherently convincing, they become redundant and eventually pass away. This is t h e inexorable
tide of time.

When the modern historical study of Buddhism began in the mid-19 th Century there was, as a
result of these competing mythologies (not to mention the even more misleading Hindu myths),
considerable confusion as to the historical picture. In a burst of rationalist enthusiasm, scholars
were prepared to question whether the myths had any factual basis at all. Was there any
historical connection between the different religions practiced in far-separated places like Sri
Lanka, Tibet, and Japan? Did the Buddha really exist? Was h e j u s t a sun-god? Was he an Ethiopian
prophet? What did he teach? Can we know? Which traditions are most reliable (or least
unreliable)? Since the traditions had been largely separated due to the forces of history -
especially the destruction of Buddhism in India - they had little information about each other,
and each asserted its own primacy. Each school preserved its traditions in vast collections of
abstruse volumes of hard-to-read manuscripts in wildly different languages (Chinese, Tibetan,
Pali, and other Indian languages such as Sanskrit). But gradually the evidence was assembled. The
traditions were compared; archeological findings confirmed key facts. 1500 year-old Sri Lankan
chronicles mention t h e names of the monks Kassapa, Majjhima, and Durabhisara sent in the
Asokan period as missionaries f r o m Vidisa to t h e Himalayan region; a stupa is excavated in Vidisa
and the names of these monks are found there, inscribed in letters dating to the Asokan era. 5 By
the beginning of the 20th Century, in works by such scholars as TW Rhys Davies, whose writings
retain their value today, accurate outlines were drawn. There was still controversy in the early
half of the 20th Century, though, as evidence was still being accumulated, new texts were edited,
and new studies done.

11
However, as early as 1882 a scholar called Samuel Beal published a series of lectures under the
title of Buddhist Literature in China. This included information on the process of translating into
Chinese, as well as sample translations from some of the main strata of Buddhist literature - the
early Suttas, thejatakas, and a Mahayana text. He stated the following:

'The Parinibbana, the Brahmajala, the Sigalovada, the Dhammacakka, the KasiBharadvadja, the
Mahamangala; all these I have found and compared with translations f r o m the Pali, and find that
in the main they are identical. I do not say literally the same; they differ in minor points, but are
identical in plot and all important details. And when the Vinaya and Agama collections are
thoroughly examined, I can have little doubt we shall find most if not all the Pali suttas in a
Chinese form.' 7

Over a century later, the thorough comparative study urged by Beal is still wanting. However,
some progress has been made. In 1908 a Japanese scholar named M. Anesaki published his 'The
Four Buddhist Agamas in Chinese: A concordance of their parts and of the corresponding
counterparts in the Pali Nikayas' 7 This was followed in 1929 by Chizen Akanuma's The Comparative
Catalogue of Chinese Agamas and Pali Nikayas,8 a comprehensive catalogue of all known existing
early discourses in Pali and Chinese, as well as the few texts available in Tibetan and Sanskrit.
These findings were incorporated in full-scale historical studies such as Etienne Lamotte's History
of Indian Buddhism and AK Warder's Indian Buddhism. These studies have largely confirmed Beal's
initial hypothesis - the Chinese Agamas and t h e Pali Nikayas are virtually identical in doctrine.
They are two varying recensions of the same set of texts. These texts - popularly referred to
simply as 'the Suttas' - were assembled by t h e first generations of the Buddha's followers, before
the period of sectarian divisions. They are pre-sectarian Buddhism. Although they are usually
considered by all schools to be 'Theravada' teachings, this is not so. Eminent scholar David
Kalupahana went so far as to declare that there is not one word of Theravada in the Pali Nikayas
(although I think this is a slight exaggeration.) The contributions of the schools are mostly
limited to fixing the final arrangement of the texts and standardizing the dialect. Interpolations
of sectarian ideas are few and usually readily recognizable. Lamotte comments:

'However, with the exception of the Mahayanist interpolations in the Ekottara, which are easily
discernable, the variations in question [between the Nikayas and Agamas] affect hardly anything
save the method of expression or the arrangement of the subjects. The doctrinal basis common to
the Nikayas and Agamas is remarkably uniform. Preserved and transmitted by the schools, the
sutras do not, however, constitute scholastic documents, but are the common heritage of all the
sects.' 9

All other texts, including the Jatakas, the Abhidhammas of the various schools, the Mahayana
sutras, and so on, were written later. Relatively few of these teachings are held in common
between the schools; that is, they are sectarian Buddhism. Through the lens of historical
criticism, the broad picture of the emergence and development of these teachings can be traced
quite clearly, both in the internal dynamics of doctrinal evolution and in Buddhism's response to
the changing cultural, social, and religious environment. There is no evidence that any of the
special doctrines of these texts - that is, the doctrines not also found in t h e early Suttas - derives
f r o m the Buddha. Rather, these texts should be regarded as the answers given by teachers of old
to the question: 'What does Buddhism mean for us?' Each succeeding generation must undertake
the delicate task of hermeneutics, the reacculturation of the Dhamma in time and place. And we,
in our own tumultuous times, so different f r o m those of any Buddhist era or culture of the past,
must find our own answers. Looked at f r o m this perspective, we can see that t h e teachings of t h e
schools offer us invaluable lessons, a wealth of precedent bequeathed us by our ancestors in faith.

12
So up to this stage in this essay I have emphasized the importance of understanding the historical
basis of Buddhism. This provides a meaningful foundation for an appreciation of the common
ground between all the schools. Before gaining a firm footing on this common ground, however,
we must admit that the traditional myths of the origin of the Buddhist texts serve the polemic,
divisive purpose of authenticating the particular doctrinal positions of the schools. This is not to
denigrate the important religious role that myths play in Buddhism; on the contrary, we will see
that the Buddhist scriptures have always been embedded in spiritual narrative, which breathes
life into the teachings. The purpose of this endeavor is not to put down the followers of any of the
schools, but to give us the discernment to distinguish the essential from the inessential, and the
courage to let go of the inessential in our quest for the essential. For the remainder of this essay I
will develop in more detail some specific theories that exemplify this quest. The discussion will
become much more technical and detailed, and I fear many readers will find bafflement and
confusion rather than inspiration and clarity. Much of the detail that follows is, of necessity,
speculative. Those who have the time and resources are encouraged to check my reasoning in
detail. For the others, who will have to accept many of my statements on trust, I hope this work
will increase their faith that genuine teachings of the Buddha may still be found.

CHAPTER 2 : THE G I S T 1 - THREE STRATA OF EARLY TEXTS

What is the GIST? It is a general hypothesis on the origin and development of the Buddhist texts.
Seeing the need for a handy name for this hypothesis, I originally thought, with tongue securely
in cheek, of following the example of the physicists and calling it the 'Grand Unified Sutta
Theory'. But the acronym 'GUST' was unfortunately evocative of a lot of hot air, so I thought of
the 'General Integrated Sutta Theory': the GIST. Which is, of course, exactly what we're after. We
seek a tool with which we can reliably prune away the masses of accretions that fill Buddhist
libraries and arrive, as nearly as possible, at the teachings of the Master himself. Even if we limit
our inquiry to the early Suttas and Vinaya we are still presented with a vast array of teachings,
some obviously post-dating the Buddha. There have been several more-or-less successful
attempts to distil this matter into various strata. The most important advance in this regard has
been the collation of the Pali Nikayas with the Chinese Agamas. This takes us back to around a
hundred years after the Buddha's death. But we are still faced with a mass of discourses with no
apparent way to go further back. The GIST attempts to penetrate even further, to within the
lifetime of the Buddha.

The GIST is 'General' because it encompasses the entire gamut of available early scriptures, that
is, the Suttas, Vinayas, and Abhidhammas of all the schools preserved in Pali and Chinese. It is
'Integrated' because it offers a synoptic presentation of the essential relations between these
texts. It deals with 'Suttas' n o t j u s t in the obvious sense that the Sutta Pitaka contains the most
important of the doctrinal teachings, but because it suggests a revaluation of the meaning of the
word 'sutta' in the earliest texts. For this reason we will not follow the usual practice of referring
to any text in the Sutta Pitaka as a 'sutta', but will use more neutral terms such as 'discourse',
reserving 'sutta' in italics to indicate the special meaning that the term carried in earlier usage.
We may, however, continue to use 'Sutta' with a capital to refer to the early texts in general as
contrasted with the Abhidhamma and other later works.

13
And finally, the GIST is a 'Theory' because it is not certain. No theory can ever fully capture the
truth. I think a successful theory is, firstly, one that addresses a genuine problem, secondly,
explains a variety of facts in a way that is at least as plausible as any alternative, and thirdly, is
suggestive of f u r t h e r inquiry. Although the GIST is only a new-born child taking its first tottering
steps, I believe it still satisfies these standards. There is definitely a serious issue at stake: how do
we relate the Pali and Chinese collections together, beyond merely stating that they share many
similar texts? We must try to investigate t h e similarities and differences more systematically, and
a promising avenue for doing that is to use the structural principles spoken of within the texts
themselves. As for alternatives, I do not feel competent to embark on a comprehensive evaluation
of all the theories that have been proposed for the origin and development of the canon.
However, I believe the current theories, while having much merit, do not sufficiently
acknowledge the influence of the structure of the Dhamma itself on the structure of the canon.
Whoever taught the early discourses had an extremely methodical, symmetrical way of thinking,
and there is a glaring discrepancy between the balanced architecture of the teachings themselves
and the sprawling collections within which t h e teachings are housed today. As for the third
standard, t h e GIST offers a clear, simple, systematic method for approaching any study of the
fundamental teachings of Buddhism. Rather than promiscuously pulling passages, ideas, or
quotes f r o m here and there to back up one's own argument, the GIST suggests a clear hierarchy of
significance within t h e early scriptures. In the second part of this work I undertake such an
inquiry in the subject of satipatthana.

The basic idea for the GIST was sparked by the findings of the renowned Taiwanese scholar monk
Yin Shun, who himself relied on earlierjapanese and Taiwanese research, none of which is widely
known in English-speaking circles. Unfortunately I am unable, at present, to read Yin Shun's
works in Chinese and t h e relevant work of his is not translated (although I understand that
translations of his entire corpus are pending). My information comes f r o m the summary of Yin
Shun's work in Choong Mun-keat's The Fundamental Teachings of Early Buddhism, and through
conversations with and the writings of Roderick Bucknell. Although Yin Shun's insights sparked
off the GIST, here the theory is developed significantly further. I will, therefore, not present Yin
Shun's findings to start with, but will outline the GIST in my own way and present Yin Shun's
contributions at the appropriate places. While Yin Shun argued back from later texts to establish
the early, I will attempt to leap over the Buddha's lifetime and approach the Buddha's teachings
by drifting downstream on the river of time.

The GIST asks three questions. Firstly, what are the earliest texts? This question is applied to
three historical strata: the first discourses, the first collection of discourses, and t h e first
Abhidhamma. The three strata are each established independently; that is, we do not rely on our
identification of the earliest discourses in order to establish the earliest collection, and we do not
rely on either of these to establish the earliest Abhidhamma. Rather, to establish each layer we
use two basic criteria: the concordance of the texts and the testimony of the tradition. An
important confirmation for the validity of these criteria is the elegance of the results. This
becomes apparent when we answer the second question: how are the three strata related to each
other? And the third question is: how are the three strata related to the rest of the
Nikayas/Agamas? The results of this inquiry, I might mention in advance, are entirely mundane;
so mundane, in fact, that they could easily be dismissed as merely stating the obvious. But what is
important here is not so much the conclusions as the method; we are trying to put on a more
sound basis what, up till now, has been largely a matter of subjective opinion.

14
BEFORE THE BUDDHA

We should start by considering possible pre-Buddhist models for the Buddhist scriptures. It is
apparent that new literary compositions in any culture are powerfully influenced by the literary
forms that are available in that culture. It is therefore likely that the organization of the earliest
stratum of Buddhist scriptures would have been primarily influenced by pre-Buddhist models.
Subsequent strata would, of course, be primarily influenced by the earlier Buddhist models. If,
therefore, we find evidence of Buddhist textual structures that are derived f r o m pre-Buddhist
models, this suggests that these structures are not merely early, but the earliest.

The only literary tradition explicitly acknowledged in the early discourses is the Brahmanical
tradition of the three Vedas and various auxiliary works. These are mentioned frequently in t h e
early discourses and obviously played a dominant role in the cultural/spiritual/literary milieu in
which the Buddhist texts were formed. There are some clear connections in the literary style of
the Buddhist texts and Vedas - especially in the poetic forms - but I do not know of any attempt
to relate the overall structure of t h e Buddhist texts to the Vedas. This is, no doubt, because t h e
structure of the existent canons bears little obvious relation to the Vedas. But perhaps it might be
worthwhile to revisit this question to see if the Vedas might have influenced an earlier
organizing principle.

The three Vedas are t h e Rg, Saman, and Yajur. (The fourth Veda recognized by the later
Brahmanical tradition, t h e Atharva, is mentioned in the early Buddhist texts but was evidently
extra-canonical at the time.) The Rg Veda is by far t h e most important. It is a very ancient (1500
B.C.E.?) collection of around 10 000 devotional and liturgical verses. One of the classification
systems of the Rg is in vaggas, groups of about ten lines of verse. The Saman is largely a collection
of hymns taken from the Rg. Even though all the Vedas contain verse, t h e Saman is the songbook
par excellence; its students were the chandogyas ('versifiers'). Alone of the Vedas the Yajur
contains prose as well as verse; it focuses mainly on the sacrifice. So we have one central work
and two supplements.

This triune form is probably of religious significance, reflecting the Trinities of deities found so
commonly in antiquity. The Trinity usually consists, not of three equal partners, but of one
presiding deity (god or goddess) who manifests in the world through t h e medium of two lesser
deities: the One becoming Two, the Two becoming many. The 'threeness' of the Vedas is
multiplied in later lists of the extended Vedic literature mentioned in t h e Pali tradition, where we
find both a sixfold and a twelvefold classification. Various parts of the Vedic literature are
referred to as angas, 'sections'. In connection with these collections t h e terms 'Vedanta' and
'Vedanga' occur. We may also note that the familiar term sutta, which in Buddhism usually refers
to any discourse, in Brahmanical usage means specifically a short, basic doctrinal statement,
which is treated as a basis for elaboration and commentary.

Several of these formal elements may also be noted in t h e Jain scriptures. Although Jainism is an
older religion than Buddhism, the Buddhist scriptures do not mention any Jain texts existing at
the time of the Buddha, and the Jains themselves agree that their scriptures were formalized
much later. However they clearly contain early elements, and it is possible that early features of
the Jain texts that are still evident in the existing texts may have exerted some influence on t h e
formation of the Buddhist texts, although it is perhaps more likely that the influence was the
other way around. The Jains acknowledge a list of fourteen purvas ('previous') that are now lost,
and twelve angas, eleven of which still exist. One of these angas is called prasnavyakarana, which
means 'Questions & Answers'. In addition they have twelve upangas, 'auxiliary sections'. It is
evident that for both t h e Vedic and the Jain traditions the term 'anga' referred to specific texts
that were organized in groups of multiples of three. The Ajivikas, another non-orthodox sect, are

15
said, in the Jain sutras, to have possessed prognostic scriptures consisting of angas, in this case
eight in number. We may also note that Jains, like the Brahmans, use t h e word sutta primarily to
refer to brief doctrinal aphorisms, although like the Buddhists they also use it as a term for the
doctrinal texts in general. For now, we can tuck all these points away in a corner of our minds for
later reference.

T H E FIRST DISCOURSES

Now let us turn to the Buddhist texts. How to identify the earliest discourses? As mentioned
above, one of the most powerful tools recognized by scholars for identifying early texts is the
concordance of the different traditions. In Buddhism this study, so far as it has happened at all,
has focussed on the concordance between t h e texts preserved by the different schismatic schools.
But, of course, what matters is not schism as such, but divergence into different textual lineages.
There is abundant evidence of a significant degree of separation and specialization of textual
study even within the Buddha's own lifetime, long before any schism. This is the most
fundamental division of the teachings acknowledged by all schools and traditions: Dhamma and
Vinaya. Several times the texts refer to groups of monks who specialize in one or the other of
these areas of study. They had different teachers and lived in different quarters. The Theravadin
account of the First Council says t h e Dhamma was spoken by Venerable Ananda and the Vinaya
by Venerable Upali. While not all of the details of the First Council can be accepted as historical,
surely this fundamental division must date back to well before the Buddha's passing away. And of
course, the content of t h e two collections is almost totally different. All this suggests that the few
doctrinal teachings that are found in the Vinaya have a special significance. They would have
been known, n o t j u s t to the doctrinal specialists, but to all the monks and nuns, dating back to the
earliest days of the Buddha's mission, before the collections of teachings grew so bulky that
specialization became necessary. Of course, it is not the case that all teachings shared between
the suttas and Vinaya must be early; discourses may have been shifted or duplicated between t h e
collections at a later date, and we know in some cases this did happen. So we must look for those
discourses that are not merely found across the Vinayas, but are also fundamental to t h e
structure of t h e text itself, things that do not seem as if they could be lightly grafted on.

So what then are these teachings? There are several versions of the early Vinaya available - about
half a dozen different schools are represented in the Chinese canon, the Mulasarvastivada Vinaya
is available in Tibetan and partially in Sanskrit, the Theravada is in the Pali canon; some other
material is also available in hybrid Sanskrit and various Indian and even Iranian dialects. Some of
these collections apparently contain substantial doctrinal material, which would be of great
interest; however most of this material has not been translated, and such studies as have been
done are mostly in Japanese. For now, we shall have to concentrate mainly on the material in the
Theravada Vinaya, with the hypothesis that similar material is available in the other Vinayas.
This hypothesis is confirmed in some important cases, and will be a fascinating arena for f u t u r e
research.

The outstanding doctrinal teachings in the Vinaya occur in the first chapter of the Mahavagga.
There are three principal sermons: the Dhammacakkappavattana Sutta, t h e Anattalakkhana
Sutta, and the Adittapariyaya Sutta. Accompanying these are a few verses, notably the request of
Brahma for the Buddha to teach, and the Buddha's reply:

'Wide open are the doors to the Deathless

Let those with ears to hear make sure their faith.' 10

16
There are other important doctrinal teachings available in the Theravada Vinaya - notably the
dependent origination and the 37 wings to enlightenment - which reinforce my argument
considerably; however these passages are not as central to the structure of the Vinaya and so
until their authenticity has been confirmed through comparative study of t h e other Vinayas we
should avoid relying on them.

The Dhammacakkappavattana Sutta is available in at least five Vinayas, as well as in the Nikayas
and Agamas. It is, in fact, by far the most widespread of all the discourses, with no less than 17
existing versions, and is one of only a few discourses that survives in the four main Buddhist
languages of Pali, Sanskrit, Chinese, and Tibetan. 11 Inevitably, there are many variations in
details, but the basic content is substantially similar - the four noble truths. The
Dhammacakkappavattana Sutta presents these teachings within a framework that clearly relates
to the Buddha's own spiritual development, his experiences with the extremes of self-indulgence
in the palace and self-torture as an ascetic, and his own recent realization of enlightenment as
the escape f r o m these two. Thus t h e internal contents of the text itself suggest that it was the
first discourse. The Anattalakkhana Sutta also occurs in several versions, as do t h e Adittapariyaya
Sutta and the Request of Brahma, although I am unable at present to uncover the full details. All
these texts, however, are available in both the Nikayas and the Agamas.

These discourses form the doctrinal core of the oldest biography of the Buddha, telling the story
f r o m after t h e Buddha's enlightenment leading up to the formation of the Sangha. This is the root
legend that forms a unifying narrative for all Buddhists. The story is told in many of t h e old texts,
sometimes in the Vinaya, sometimes as a Sutta; in later embellished form it became a lengthy
book in itself. But beneath the profuse elaborations remains a remarkable consistency in both the
basic narrative and the doctrinal teachings. Even a late text like the Mahavastu preserves
teachings such as the Dhammacakkappavattana Sutta in nearly identical form. 14 They are
universally regarded as the Buddha's first teachings, and so we have complete agreement
between the concordance of the texts and the testimony of the tradition. Of course, it is
impossible to ever establish that these texts were literally t h e first teachings. Nor can we deny
that there are some minor differences between the versions. But we can be sure that these texts
are fundamental to the existing collections of Buddhist scriptures, and there seems no good
reason why this should not simply reflect the historical position.

There is a substantial problem with this neat theory, however. The passage we have been
considering, the first chapter of t h e Vinaya Mahavagga, is also found in the Sarvastivadin
tradition. However, this is not in t h e Vinaya, but in the DTrgha Agama under the title Catusparisat
Sutra ('Four Assemblies Sutra'). 13 This discourse exists in several versions, attesting to its
popularity. It is very close to the Pali version, though lacking the specifically Vinaya elements.
Several scholars have expressed t h e opinion that this text was originally part of the Vinaya, and
was later moved to the Agamas. If this is the case, there is no problem. However, it is quite
possible that the movement was the other way around: the text was originally a discourse that
was later incorporated into t h e Vinaya. This would suggest that these doctrinal passages were
not, in fact part of the original Vinaya. In part, this question must await f u r t h e r inquiry, including
detailed examination of the various Vinayas; some scholarly work has been done in this field. But
I think it is better to consider the Catusparisat Sutra as fundamental to both Dhamma and Vinaya.
In leading up to the first sermon it supplies the background narrative for the Dhamma; and then
Anna Kondanna's subsequent going forth is the perfect starting point to unfold t h e Vinaya.

We are trying to discern a glimpse of the earliest phase of Buddhism. In the early years, there
would have been relatively few teachings. All the monks would have known by heart the few
texts and discourses that were regarded as central. In addition, they would have all been familiar
with the simple non-legalistic codes of behaviour expected of them as Buddhist mendicants. Thus

17
they would all have known both Dhamma and Vinaya. It would have taken a number of years for
the bulk of the material to grow to the extent that specialization became necessary. From this
point, certain monks and nuns would specialize in Dhamma, while others specialized in Vinaya.
But this specialization has only ever been a matter of emphasis, not of exclusion. All the Vinaya
specialists would have known some Dhamma, while all the Dhamma experts would know some
Vinaya. It is entirely reasonable to suppose that t h e doctrinal teachings within the existing
Vinayas are remnants, either directly or indirectly, of such a shared doctrinal body. It is also
entirely reasonable to suggest that the monks and nuns would have all been familiar with the
story of the Buddha's enlightenment, and that this legend would be given concrete literary form,
although incorporating late details, largely to lend authority to this doctrinal core. That is to say,
the Catusparisat Sutra in its developed form may have been incorporated in the Vinaya precisely
because the Vinaya experts were already familiar with the main doctrinal teachings of that
discourse, and they would then be supplied with a historical context linking the doctrine with the
establishment of t h e Sangha and the laying down of Vinaya. So I would conclude that the
presence of the relevant doctrinal passages within the existing Vinayas remains as supporting
evidence for t h e primacy of these teachings.

T H E GATHA THEORY

The GIST makes a strong case that the traditions, in this case, have got it right. A major scholarly
challenge to this conclusion comes from what we can call the 'gatha theory'. This theory, which
claims several very eminent scholars among its adherents, claims that the earliest recorded
teachings that we possess today are to be found primarily among certain of t h e verse collections,
notably the Atthaka and Parayana of the Sutta Nipata. 14 However, while I agree that some of the
verse is early, I do not think that the reasons given suffice to establish that the verses are
generally earlier than t h e prose. To briefly state the case for and against the gatha theory:

1. The language found in such texts harks back in some respects to the Vedas, and therefore
is archaic.

Verse usually tends to be archaic; this could be supported in any number of cases by comparison
of verse and prose passages by the same author even in modern times. This may partially be a
matter of style, a preference for an archaic flavour, as in English verse one might affect 'thee' and
'thou'. Another factor is that, due to the constraints of metre, it is more difficult to translate verse
as compared with prose f r o m one Indian dialect into another; thus even in the later hybrid
Sanskrit literature, the verse tends to retain more archaic Prakrit features, while the
accompanying prose tends towards more formal Sanskrit. This tells us something about t h e
translation process, but nothing about the relative ages of the different parts of t h e original text.

2. Several of these verses are referred to in the prose Nikayas, and therefore must be earlier
than those prose discourses.

This confirms only the chronological relationship in these few specific cases. In many other cases,
verses are tacked on to the end of prose discourses, such as in the Anguttara, and there it seems
likely that often it is the verses that were added later. Anyway, there are also prose passages that
are quoted or referred to in other prose passages, notably the Dhammacakkappavattana Sutta,
which is explicitly or implicitly referred to in several important discourses. The references to the
gathas, moreover, while significant, never declare such passages to be the central message of the
Dhamma. The key teachings, extolled over and again in the early texts, comprise such teachings
as the four noble truths, the 37 wings to enlightenment, the dependent origination, or t h e
'aggregates, sense media, and elements'. None of these topics are prominent in the gathas. It

18
would be natural to assume that the earliest scriptural body consisted of teachings o n j u s t such
core topics. Such references may even refer to specific texts where these doctrines are elucidated.
Currently, the primary source for all these topics is the Samyutta.

3. The Atthaka and Par ay ana have their own canonical commentary within the Khuddaka
Nikaya, the Niddesa.

This argument has recently been repeated by Gregory Schopen, who says that these are the 'only'
texts that have received commentaries by the time of the earliest known redaction. 15 This seems
like a strong point, until we realize that the Niddesa really just applies Abhidhamma technique to
poetry, listing synonyms in mechanical, Abhidhamma style for each word in the verses. It is very
similar to t h e Abhidhamma Vibhanga, etc., and must stem from a similar period as a minor spin-
off from the Abhidhamma project. The Vibhanga is clearly the more important work, and that
consists largely of quotations and commentary of central prose passages of the Samyutta and
Majjhima. In fact there is much 'commentarial' material even in the four Nikayas: the
Saccavibhanga Sutta, which we will examine f u r t h e r below, is an explicit commentary on the
Dhammacakkappavattana Sutta. Much of the Vinaya, too, consists of a commentary on the
Patimokkha.

4. Technical terminology and formulaic doctrines appear less often.

Again, this is simply part of the normal character of verse. Poetry is for inspiration, not
information.

5. The monks lived as hermits in the forest rather than in settled monasteries, whereas in
the prose this phase of Buddhism is largely absent, the discourses being normally set in
monasteries.

This shift, f r o m the forest life to established monasteries, is depicted in the texts themselves as
having already begun within the Buddha's lifetime, and there is every reason to believe that this
was so. It is difficult to live in the forest, and t h e Sangha must have, before very long, started
taking in recruits who were elderly, or infirm, or weak, and who would have required decent
accommodation. This plain common sense is confirmed in many stories in the early texts. Here
we may point out the parallel with the Franciscan order, which was accused by St Francis himself
of backsliding f r o m t h e rigorous standards he had set. In any case, the prose does in fact
constantly refer to monks living in the forest. The mistake stems in part f r o m the failure to
distinguish between the teachings themselves and the narrative cladding in which the teachings
appear, which must obviously be later. The outstanding example here is the teaching on the
gradual training, the main paradigm for the monastic way of life, found in tens of discourses.
Although the texts as they are today are set in monasteries, the body of the teaching itself refers
simply to the monk, 'gone to t h e forest, to the root of a tree, or to an empty hut...' to meditate,
with no mention of monasteries. This is a good piece of negative evidence: we know that later
Buddhism was largely based in large monasteries, hence the fact that so many of the teachings
extol t h e forest life strongly suggests these teachings must have appeared before the
development of settled monasticism.

So it seems that in this instance the traditional belief can be maintained in t h e face of modern
criticism. Please remember that we are not saying that the discourses as found today must be
word for word identical with the first teachings, but that these teachings, in largely the same
words and phrases, have been treated since earliest times as the most fundamental doctrines, and
the traditions give us a plausible reason why this should be so. The massive preponderance of t h e
Dhammacakkappavattana Sutta demands an explanation. The idea, influential for a time in

19
Buddhist studies, that these teachings hail f r o m a 'monkish' revision of t h e Dhamma after the
Buddha's passing away has all the romance of a conspiracy theory, and all of its plausibility.

T H E EARLIEST COLLECTION

Having given grounds to establish t h e earliest discourses, we now ask: 'What is the earliest
collection of discourses?' Here again we invoke our twin criteria: concordance of the texts and
the testimony of the traditions. First we should seek for the collections that show t h e highest
degree of congruence. Of the major collections, t h e following are available. The symbol $ denotes
an incomplete collection. The Sanskrit DTrgha is an exciting new finding coming out of
Afghanistan and has not yet been fully edited.

Table 2.1: The Nikayas and the Agamas

Theravada Sarvastivada Other Agamas (in Chinese)

Nikayas Agamas
DTgha (Pali) DTrgha^ (Sanskrit) DTrgha (Dharmaguptaka)

Majjhima (Pali) Madhyama

(Chinese)
Samyutta (Pali) Samyukta Two 'other' Samyuktas^ (unknown schools)
(Chinese)
Anguttara (Pali) Ekottara (Mahasanghika?), Anguttara^ (unknown
school)

It will be useful here to make a few remarks about t h e early history of Buddhism, especially
regarding the schools relevant to this study. The early chronology of Buddhism is still largely
obscure. Even the most important date, the passing away of the Buddha, is specified very
differently in different traditions, and it is far f r o m obvious which, if any, is more reliable.
Following Gombrich and others we might take the dates 484-404 B.C.E. for the Buddha as being no
less reliable than other estimates. The schisms are undateable in an absolute sense, and even the
relative chronology of t h e schisms is disputed. This probably reflects the real historical situation,
since separative tendencies may have proceeded at different rates in different areas, and there
may well have been no universal agreement even at the time at to the exact dates of the schisms.
It is even unsure whether the Sangha at the time would have been conscious that it was creating
lasting divisions into schools in the sense that we understand it today. The full implications of the
breaches may only have become apparent many years later. Most of the sources date the first
(Mahasanghika) schism around 100-160 years after the Parinibb$na. The Sarv$stiv$da schism was
probably a few decades later, and the Dharmaguptaka later still.

The first separative movement, between the ancestral Theravada and the Mahasanghika, was
driven by differing attitudes towards worldly things. The Mahasanghika ('Majority Community')
advocated a relaxing of certain Vinaya rules, especially about handling money. Later this relaxed
attitude was extended to questions relating to the status of the arahant, whom they felt might
still fall prey to certain human weaknesses. The Theravada took a stricter stance on these
questions, and eventually schism resulted. The Mahasanghika proceeded to splinter into several
sub-schools, and it is f r o m these sub-schools that the rise of the Mahayana, much later, is usually
traced. The existing Ekottara Agama, which includes some Mahayanist interpolations, is often
said to come f r o m one of these sub-schools, but the evidence is as yet inconclusive. We will, for

20
the purposes of this study, follow the majority of modern scholars in treating the Ekottara as
stemming f r o m one of the Mahasanghika group of schools, although we should always bear in
mind that this is only a tentative hypothesis. The results of our study of the Satipatthana Sutta
tend to support this theory; at least, t h e Ekottara version of this discourse is relatively more
divergent than t h e other versions, which is what we would expect of a Mahasanghika text, even
though the divergences are not explicitly sectarian.

The ancestral Theravada, too, underwent many f u r t h e r schisms. The earliest of significance for
our story was the Sarvastivada schism, which probably pre-dated Asoka's coronation in 268 B.C.E.
Here t h e fundamental issue was the conception of time, the special Sarvastivadin doctrine being
that all dhammas past, present, and future, exist. The Sarvastivada schism produced, as well as
the Sarvastivada, another school sometimes called the Vibhajjavada, the 'Analytical School'. This
label is used widely and inconsistently, but it is convenient to use it here as a term for the school
ancestral to the Theravada and the Dharmaguptaka. These schools are very similar in doctrine,
the main difference apparently being a greater emphasis on devotion in the Dharmaguptaka, as
evidenced by certain Vinaya rules regarding stupa worship, the comparison of the structure of
the DTghas, and the fact that they placed greater emphasis on the merit of making offerings to the
Buddha rather than to the Sangha. This is all marginal stuff, and may be due to the
Dharmaguptaka texts being settled somewhat later rather than a genuine sectarian divergence,
for the Theravadins also embraced stupa worship very strongly, but did not insert it into their
Vinaya. The only clearly sectarian difference f r o m the Theravada is the relative value of offerings
to the Buddha and the Sangha. It seems likely t h a t the divergence between these schools arose at
least partly due to mere geography, the Dharmaguptaka being a branch of the Vibhajjavada that
stayed in Northern India while the Theravada moved to Sri Lanka. The Dharmaguptaka became
well established in Central Asia and initially enjoyed great success in China; Chinese bhikkhus and
BhikkhunTs today still follow the Dharmaguptaka Vinaya.

Bearing this general information about t h e early schools in mind, we may return to our question
as to which is the earliest collection. Generally, the traditions assert that all four collections were
created simultaneously. Although the idea that they were put together in one session is obviously
incorrect, we can still admit that the period of accumulation of texts largely overlaps. It is not
implausible to even suppose that the four Nikayas/Agamas were started at the same time, and
then finalized at the same time. This would be reinforced by the theory that each of the
collections is focussed slightly differently, and was formed to cater to the needs of different
sectors of the Buddhist community. Nevertheless, we may still distinguish in terms of a tendency
to be earlier or later, even within this framework.

This question can, and should, be approached f r o m a variety of angles - philological, doctrinal,
cultural, and so on. All of these involve complex and large-scale investigations, and the results of
none of these inquiries are yet beyond doubt. One problem with all of these approaches is that
they can tell us, at best, about relative dates of certain phrases, ideas, or passages, but not about
the collection as a whole. Bearing in mind the presence of intratextuality - the persistence of
earlier text in later redactions - in the Nikayas/Agamas, and the vast quantity of material to be
dealt with, it seems clear that only a very large scale statistical analysis of linguistic, doctrinal, or
other features could give us firm answers. I do not know of any studies that even come close to
this ideal.

Shrouded by this mass of darkness, I would suggest that the structural analysis of the
Nikayas/Agamas offers us, at our present state of knowledge, the closest we have to a shining
light. The structural principles of the collections tell us how the redactors of the collections
worked, rather than how the compilers of the individual discourses worked. With the emergence
of the Sarvastivada DTrgha, we now have the full structural details of three Agamas of one, very

21
important, school to compare with t h e Theravada. This means we can directly evaluate the
degree of congruence between the corresponding pairs of collections, without being overly
concerned that sectarian issues might distort the picture.

We still lack a Sarvastivadin collection corresponding to the Anguttara. Meaningful comparative

studies for the Anguttara are very limited in scope, for the Theravada Anguttara and
Mahasanghika Ekottara are very different, calling into question whether they are in fact
recensions of a single work, or whether the existing similarities are merely due to the fact that
they both use t h e same organizing principle. (There is another partial Anguttara in Chinese that
is closer to the Theravadin version.) So for now, if we do not wish to delve into t h e complex
question of whether t h e content of the Anguttaras seems to be earlier or later, all we can do is to
put t h e m to one side. My personal opinion is that the Anguttaras house the shorter discourses
that were 'left over' f r o m the main doctrinal topics in the Samyutta, so perhaps their omission
f r o m consideration is not so critical.

What is needed here is a comparison of the comparisons. We must ask, which collections appear
to have the closest structural relationship, the three DTghas, or the two Majjhimas, or the two
Samyuttas? I will discuss each of these collections in f u r t h e r detail below, so I j u s t present a quick
overview here. In fact, the answer to this question is really obvious as soon as it is asked. The
three DTghas share many discourses in common, but the sequence and organization of the
discourses are widely divergent. This is particularly the case when the Sarvastivada DTrgha is
compared with the two Vibhajjavada DTghas. Similarly, the two Majjhimas share much content
but little structure in common. Almost all the chapter titles and divisions are completely
different, with a few exceptions discussed f u r t h e r below. (Anticipating the argument, we may
notice that the occasional structural congruencies between the two Majjhimas and the three
DTghas may be derived f r o m the Samyutta and are therefore not necessarily evidence of pre-
sectarian structures in the Majjhimas and DTghas). When we come to the two Samyuttas,
however, t h e picture is radically different. They share all the same major divisions into subjects,
etc., with some variations in t h e minor chapters and some reshuffling. 16 So we can say with some
certainty that not only t h e content but also the structure of the Samyuttas were largely settled in
the pre-sectarian period, whereas t h e structures of the Majjhimas and the DTghas are largely
sectarian. Thus, relying primarily on this very clear structural picture, we conclude that t h e
concordance of the texts suggests that the Samyutta is earlier than t h e Majjhima and the DTgha.

Some might object at this point that our reasoning proves nothing, since it is the Samyutta alone
of the four Nikayas/Agamas that has a meaningful overall structure, so as soon as we choose to
look at structure we will be drawn to it. In other words, our conclusion simply follows f r o m our
choice of methodology. This objection is perfectly true, yet I feel t h e argument still has force. We
need some criterion, and structure is one possible means at our disposal. It is quite possible it will
tell us nothing, yet we must at least ask the question and follow the answer through. The simple
fact remains that the Samyutta does evidence a large scale structural congruence that is lacking
in the other collections, and one rational explanation for this is that the Samyutta was settled
earlier. It also seems to be the case that, in the case of the Majjhimas in particular, even when
there is a structural grouping held in common, the actual content of that group is usually very
different. For example, each of the Majjhimas has a vagga called the "Chapter on Kings'. In the
Pali, this consists of ten discourses, in the Chinese, of fourteen. But only two discourses are held
in common in the two vaggas. 17 Most of the other discourses are found in both t h e Pali and
Chinese sources, but not in this chapter. Thus both of the traditions had the idea of collecting
some middle-length discourses together on the t h e m e of Kings, but the selection of discourses
was independent. But in the Samyutta, we find almost invariably, when a group of discourses has
been formed around a certain t h e m e or principle, there is a very large percentage of the actual
discourses that overlap. This is consistent with the thesis that the structures of the Samyutta are

22
pre-sectarian, while the occasional structural similarities in the Majjhima may have arisen
through parallel development.

So what of the testimony of the traditions? This brings us to the important findings of Yin Shun.
The Chinese and Tibetan canons contain a monumental treatise called t h e Yogacarabhumisastra,
written by Asanga around 400 C.E. This was a fundamental and authoritative work for the
Yogacara school of Mahayana. A section of this work called the VastusangrahinT is devoted to an
extensive commentary on the Samyukta Agama. This demonstrates t h e heavy reliance by the
classical Mahayana on the early discourses, a context that is too often overlooked. It is interesting
that the other great school of early Mahayana, the Madhyamaka, takes as its textual basis
Nagarjuna's Mulamadhyamakakarika ('Root Madhyamaka Treatise'), which in turn is based on the
Kaccayana Sutta of the Samyutta. Thus the two main schools of the Mahayana share the Samyutta
as a fundamental scripture, although I do not know of any evidence that Nagarjuna relied on the
Samyutta because he believed that it was historically the earliest collection. Yin Shun has shown
that the Samyukta Agama discussed in the Yogacarabhumisastra is very close to t h e Samyukta
now preserved in the Chinese canon, and has used the Yogacarabhumisastra to reconstruct t h e
earlier sequence of the Samyukta Agama, which had become disordered over time. His
reconstruction, which was built on the work of earlier scholars, and has been confirmed by later
scholars, is considered so authoritative that it has been adopted in the Foguang edition of the
Agamas published in 1983. The Yogacarabhumisastra suggests that the Samyukta Agama was t h e
foundation for the four Agamas. Yin Shun believes that this statement can be interpreted
literally, as affirming t h e historical priority of the Samyutta among the Agamas. There does not
seem to be any direct statement to this effect in the Theravada tradition; however there are, we
shall see, a few hints. The Sarvastivadin tradition, however, regularly lists the Samyutta as the
first of the Agamas. Thus as to the first collection of discourses we have satisfied our two criteria,
congruence of the texts and testimony of at least one tradition.

T H E FIRST A B H I D H A M M A

What, then, of the first Abhidhamma? Here we rely primarily on the work of Frauwallner. He has
demonstrated that three early Abhidhamma texts share much the same content and must have
been derived from a common ancestor, which we call the '*Vibhanga Mula'. These are the
Vibhanga of the Theravada, the Dharmaskandha of the Sarvastivada, and the Sariputrabhidharma
of the Dharmaguptaka. The details of these works are too complex to go into here; for now we can
take this congruence as established.

Both the Pali and Sanskrit traditions contain evidence that these texts were considered
fundamental to the Abhidhamma. The epilogue to the Chinese translation of the Dharmaskandha
says that it was the basic text of the Abhidhamma and the primary source for the Sarvastivadin
school. The Vinaya of the Dharmaguptaka school outlines their Abhidhamma system, which is
precisely the table of contents of the Sariputrabhidharma. And the AtthasalinT, the main
commentary on the Theravada Abhidhamma, in its first chapter includes two prominent passages
where the main topics of the Abhidhamma are listed; in the first passage, these are identical with
the contents of the Vibhanga, and in t h e second passage very similar. 18 So here too the two
criteria of the GIST are clearly satisfied.

To review our findings: we have demonstrated that according to both the concordance of the
texts and the testimony of the traditions the following constitute the earliest strata of Buddhist
texts.

23
Earliest Discourses: Dhammacakkappavattana, Anattalakkhana, and Adittapariyaya Suttas, and
the Request of Brahma.
Earliest Collection: Congruent sections of Samyutta Nikaya/Samyukta Agama
Earliest Abhidhamma: Congruent sections ofVibhanga/Dharmaskandha/Sariputrabhidharma

SOME PROBLEMS

There is a possible objection that I would like to address here. Some might argue that our two
independent criteria are not independent at all. The traditions might have decided which
teachings were earliest and then invented myths expressing this, reinforcing their claim by
multiplying the occurrence of these teachings in the various collections. We can see this at work
even today. The Dhammacakkappavattana Sutta has been reproduced in dozens of Dhamma
books, precisely because it is regarded by t h e traditions as the earliest teaching.

We must admit that this criticism has some force, especially in the case of the earliest discourses.
But for the earliest collection and t h e earliest Abhidhamma this objection is weak, since t h e
traditions, though preserving a dim memory of the priority of these texts, were generally not
conscious of this, and did not make a display of it. Rather the opposite: the party line was that all
the texts stem f r o m the Buddha himself, so they were anxious to de-emphasize or outright deny
any question of historical priority. For example, we said that the Vibhanga could be considered
the fundamental work in Theravada Abhidhamma by referring to the important first chapter of
the Abhidhamma commentary, the AtthasalinT, which summarized the Abhidhamma by giving a
list of the topics in the Vibhanga. This reference is itself somewhat incongruous, because these
topics are usually not considered Abhidhamma as such; the distinctive Abhidhamma teachings in
Theravada today are usually said to be the matika of the DhammasanganT, the 24 conditions of the
Patthana, or later developments such as the 'series of conscious moments', etc. What is
distinctively abhidhammic about the Vibhanga is not t h e list of topics, but the special mode of
treatment; but this is not mentioned in the AtthasalinT reference. It is possible that here t h e
AtthasalinT is passing on an old understanding of what constitutes Abhidhamma, even as, in t h e
same chapter, it consciously articulates the myth authorizing the whole of the Abhidhamma
Pitaka. Even in the case of the earliest discourses, though the objection carries some weight, it
does not account for t h e massive agreement among the schools. Given the popularity of t h e
Dhammacakkappavattana Sutta, and the universal agreement that it was the first discourse, t h e
most plausible explanation is simply that the traditions are right. How else would it have gained
the approval of all Buddhists in the pre-sectarian period? Who, if not the Buddha, could have
imbued it with such authority?

Another problem we need to bear in mind is the possibility of later borrowing. There is no doubt
that borrowing did take place between the traditions in all periods. For example, the great
Theravadin commentator Buddhaghosa in his exposition of the practice of the Bodhisattva
adapted parts of the Bodhisattvabhumi f r o m Asanga's Yogacarabhumisastra. Later borrowing
must be borne in mind as an alternative to the thesis of a shared heritage. Generally, our response
to this criticism is simply to pursue the thesis of shared heritage, follow through t h e implications,
and see whether that leads to useful results. Working with the material in detail and in depth, it
becomes more and more obvious that later borrowing is unlikely to affect more than a few
details. This whole book can be read as a demonstration of t h e fruitfulness and reasonableness of
this approach. This is not, it should be remembered, an arbitrary or unusual method. Scholars
working in other areas, whether Bible studies or biology, regularly make use of similar
hypotheses.

24
However, it might be useful to demonstrate at least one case where the thesis of later borrowing
is very implausible. Let us consider t h e well-known Bhara Sutta f r o m the Khandha-samyutta. 19
This discourse is available in versions belonging to the Theravada, Sarvastivada, and (possibly)
Mahasanghika schools. It states that the five aggregates are the burden, and the 'person' (puggala)
is the 'bearer of the burden'. In light of the teachings on not-self, this is an unusual statement.
While most of the schools took the line that the 'person' here was just a conventional way of
speaking, one important group of schools, the Puggalavada, declared that this referred to a real
entity that existed outside the five aggregates. Their Abhidhamma treatise, existing in Chinese
translation, refers explicitly to this same discourse. 20 Now, the Puggalavada schism happened
very early, soon after the first (Mahasanghika) schism. So we have a very good spread of this
discourse across all the earliest schools.

One of the important forces leading to schism was discussions and disagreements on the relevant
doctrines. These discussions likely preceded the actual schism by a considerable time. As the split
hardened, the schools began to formulate their position in accepted texts, developing
sophisticated arguments defending their interpretation. This would have been essential training
material for the energetic doctrinal debates that were ongoing. There are two main records of
these discussions of the early period: Moggaliputtatissa's Kathavatthu in the Pali canon of the
Theravadin school, and Devasarman's Vijnanakaya, a canonical text of t h e Sarvastivada
Abhidharma preserved in Chinese. The Kathavatthu is traditionally ascribed to the period of
Asoka; although much of the work is later, there is no reason to doubt that its origins, with some
of the core arguments, stem f r o m that period. Norman has in fact shown that the Kathavatthu,
especially the discussion of the 'person', includes an unusual number of Magadhan grammatical
forms, suggestive of an Asokan connection. 21 The edicts show Asoka's great concern to prevent
schism in the Sangha, suggesting that the schismatic tendency was evident in his time. The
Puggalavada schism was among t h e earliest, perhaps even before the split between the
Sarvastivada and the Vibhajjavada.

The first and longest section of both of these works is a lengthy refutation of the 'person' thesis.
This is obviously a core issue, perhaps the initial motivation for writing these works. This schism
hurt. It was still fresh in their minds, felt as a direct assault on the cherished doctrine of not-self.

We might then ask: how can this situation best be explained? Let us assume that the Puggalavada
wrote the Bhara Sutta to justify their special doctrine. This must have happened in the early
schismatic period, while they were freshly arguing with their brothers and sisters in the other
schools. The other schools were so persuaded by t h e authenticity of this discourse, apparently,
that they borrowed it for inclusion in their central doctrinal collections, even as they were at the
same time furiously arguing against the 'person' thesis as t h e worst of heresies. The
Puggalavadins were so successful with their forged discourse that it became accepted without a
m u r m u r of protest or questioning across the schools for all time.

Or let us suppose that another school invented this discourse, say t h e Theravadins. They had
been arguing with rivals, who they regarded as apostates, over the doctrine of the 'person'.
Somehow, they produced a discourse that seems to justify their opponent's arguments and
included it in their canon, being too dim-witted to see the implications. This discourse became
rapidly 'seeded' across a variety of schools over the breadth of India; one can only assume that
they were very enthusiastic about their new creation and wished to spread it far and wide. When
it became known to the Theravadins' arch-rivals the Puggalavadins, they leapt on it with glee to
justify their main thesis, although it is not recorded that they thanked the authors for the gift.

If these options do not appeal, perhaps we might fall back on the drab and hackneyed idea of
shared heritage. There was an discourse called the Bhara Sutta. This was, perhaps, spoken by the

25
Buddha; or at least it was accepted as such by the first generations of Buddhists in the pre-
sectarian period. As it dealt with the important doctrine of the five aggregates, it was assembled,
along with many related discourses, in a collection that came to be called the 'Khandha-
samyutta'. Thus its canonicity was assured. As discussions into the meanings of the discourses
went on, some began to see a special significance in the mention of the 'person' here, to notice
other places where the word 'person' seemed also suggestive, and to develop the 'personalist'
thesis. Although some attempted to dissuade them, they persisted in their views, and eventually
schism resulted. Each school inherited a version of the problematic discourse, which was already
so deeply embedded in the received canon that its status was unimpeachable, and developed their
own interpretation in accord with their views. These interpretations became embodied in the
Abhidhamma works of the schools.

I trust that the reader, like myself, finds the final option the most plausible. Of course, not every
discourse can be established so easily. But if even one discourse can be shown to be pre-sectarian,
this makes it all the more likely that other similar discourses, and the collections in which these
are found, also include pre-sectarian material. At least we shall not seem unreasonable if we
follow this path to see where it might lead us.

CHAPTER 3 :

THE G I S T 2 - THE AGREEMENT OF THE THREE STRATA

T H E SEEDS OF THE S A B Y U T T A

And so to our second major question: what is the relationship of these strata to each other? The
first outstanding feature is that all of the texts identified as earliest discourses are found in the
Samyutta, the earliest collection. This is a compelling reason to consider these discourses as the
root texts of all Buddhism, not in any vague or rhetorical sense, but as the literal historical seed
around which the Samyutta and then the other collections crystallized.

It may well be the case that the Dhammacakkappavattana Sutta was originally the first discourse
in the Samyutta. At present it is number eleven in the Theravada Sacca-samyutta; but in the
Chinese it is the first in this chapter. (The position in the Pali can be explained by the later
insertion of a vagga of ten discourses in front.) So if the Samyutta was the first collection and the
Dhammacakkappavattana Sutta was the first discourse in its chapter, it doesn't take too much of
a leap to postulate that the Sacca-samyutta may have originally been the first topic in the
Samyutta Nikaya. This would, of course, be logical, for the four noble truths is the most general,
all-encompassing teaching, of which the other doctrinal categories are more specialized
explanations.

There is perhaps an echo of this original structure preserved in the title given this discourse in
the Pali. In most manuscripts the name 'Dhammacakkappavattana Sutta' does not occur; the

26
discourse is called tathagatena vutta ('Spoken by the Tathagata'). This is, on the face of it, a bit odd,
for most of the discourses are, of course, attributed to the Buddha. However the terms 'Spoken by
the Buddha' and 'Spoken by the Disciples' occur in t h e Chinese recension, not as titles of
discourses, but as titles of sections. Perhaps the label tathagatena vutta referred originally, not to
the Dhammacakkappavattana Sutta specifically, but to a section within a collection of discourses
that consisted purely of teachings given directly by the Buddha himself. 22

So it seems reasonable to suggest that the Dhammacakkappavattana Sutta was originally, not the
eleventh discourse in the fifty-sixth book of the third collection, but the first discourse in the first
book of the first collection. The internal structure of the existing collections does not, so far as I
can see, suggest that the Anattalakkhana Sutta and the Adittapariyaya Sutta ever enjoyed similar
primacy within their respective collections. There is some suggestion that the Request of Brahma
may have been t h e first discourse in the Sagatha Vagga, based on Bucknell's reconstruction,
following the Yogacarabhumisastra, of the Sagatha Vagga along the lines of the Eight Assemblies;
however the argument is too complex to go into here. 23 The Request of Brahma is, surprisingly,
missing f r o m the Sarvastivada Samyukta, although it is found in t h e Ekottara and (Sanskrit)
DTrgha Agamas, and probably elsewhere, appearing immediately before the
Dhammacakkappavattana Sutta, just as in the Vinaya.

The primacy of the Dhammacakkappavattana Sutta as the cornerstone of the canonical

collections is supported in some of the accounts of the First Council, preserved in the Vinayas of
the schools. As might be expected, each school preserves a version of the events at this Council
that serves to authorize its own canon. For example, the Theravada states that t h e Brahmajala
Sutta was recited first; and in the existing Pali Tipitaka we do indeed find that the Brahmajala
Sutta is the first discourse in the first collection, the DTgha Nikaya. This fact tends to discount the
value of the sectarian accounts of the Councils as records of history. But, while not all the
accounts can be accepted, there is no reason why at least one should not be substantially correct.
In fact, given that the general historicity and importance of t h e First Council is widely accepted, it
would seem likely that some memory of the actual procedures of t h e main business would be
accurately preserved somewhere in the traditions. Only if the school later re-organized their
scriptures would they feel the need to revise their account. Thus even if the existence of a
corroborating account is not felt to strongly verify a theory of what happened at t h e First
Council, t h e absence of a corroborating account would tend to falsify such a theory.

We should therefore consider whether any of the schools included an account of the First Council
that is in line with the GIST. We do not have to look far, for the most influential of the Indian
schools, the Sarvastivadins, say the Dhammacakkappavattana Sutta was t h e first discourse recited
at t h e Council. They were followed by the Mulasarvastivada Vinaya 2 4 . In this narrative, the
holding of the First Council is presaged by the Buddha's exhortation, shortly before he passes
away, for the Sangha to preserve the Dhamma by reciting the twelve angas. Then, after the
Buddha's passing away, having convened the Council, Venerable Maha Kassapa requested that
Venerable Ananda recite the Suttas. 25 He first spoke the Dhammacakkappavattana Sutta. As we
shall see below in the excerpt f r o m the Catusparisat Sutra, the Sarvastivada tradition does not
include the detailed description of each of the truths (as is found in the Pali) in the first discourse.
The detailed description (birth is s u f f e r i n g . ) is here said to be t h e second discourse. The
discourse on not-self, 'also spoken at Benares for the sake of the five monks', is therefore said to
be the third discourse. When these discourses were given, all without flaw or criticism, all the
arahants accepted them as the Buddha's true teaching.

'Thus Ananda now explained every teaching. Every arahant unanimously participated in t h e
Council. And so the five-aggregate-samyutta was compiled and placed in the Khandha Vagga. And
also the six-sense-media-and-eighteen-elements-samyutta was compiled and placed in the

27
Ayatana-dhatu Vagga. And also the dependent-origination-and-noble-truths-samyutta was
compiled and placed with the name Paticca-samuppada [text omits 'Vagga'; note that can
stand for either paticca-samuppada or nidana]. All those teachings that were spoken by disciples
were placed in the Sravaka Vagga. All those teachings that were spoken by the Buddha were
placed in the Buddha Vagga. All those dealing with satipatthana, right efforts, bases of psychic
power, spiritual faculties, spiritual powers, enlightenment [-factors], and the path were compiled
and placed in the Magga Vagga. Also t h e sutras in the 'with-verses-samyutta'. These are now
called the Samyukta Agama. All the discourses that were long teachings were named the DTrgha
Agama. All the discourses that were middle-length teachings were named the Madhyama Agama.
All those discourses with one topic, two topics, up to ten topics, these were now named the
Ekottara Agama.' 26

Notice that after referring to the Anattalakkhana Sutta, the passage goes on to speak of compiling
a samyutta of texts dealing with the aggregates. Now, the Anattalakkhana Sutta is the
fundamental text on the aggregates, and this is in fact included in the main group of texts on t h e
aggregates, namely the Khandha-samyutta. Next the text speaks of a collection dealing with t h e
sense media. Here t h e Adittapariyaya Sutta is the basic text, and although it is not mentioned in
the above account, in the Theravada tradition this is regarded as the third discourse. It is
included in the Salayatana-samyutta. The categories 'spoken by disciples' and 'spoken by the
Buddha' are found in the existing Samyukta Agama. The next passage clearly lists the 37 wings to
enlightenment. These topics are the backbone of the Magga Vagga (or in t h e Theravada, the Maha
Vagga) of the Samyutta. Finally the Sagatha Vagga is mentioned. Clearly, then, this passage
authorises the Samyutta as the central body of the fundamental teachings, collected around the
seeds of the first discourses. Then follows t h e other three Nikayas/Agamas, compiled after the
Samyutta. While it may be a sheer coincidence, it is worth noting that t h e title of this section of
the Mulasarvastivada Vinaya is 'Samyukta Vastu', which could be rendered 'The Story of the
Samyutta'.

So the Mulasarvastivada Vinaya account suggests to us some clear relationships between what we
have identified as the earliest discourses and t h e earliest collection. It then goes on to depict
Venerable Maha Kassapa presenting the matika, the list of contents of t h e Abhidhamma. This is
not mentioned in the Theravada account, and is clearly a later interpolation. But it is of interest
in that it shows what was presumably regarded as the basic topics of the Mulasarvastivada
Abhidharma. This Abhidharma has not survived, and the given topics do not exactly match with
any of the existing Abhidharma works, not even the Sarvastivada. Nevertheless, there are strong
lines of continuity with what we shall later identify as t h e root-Abhidhamma treatise common to
the schools.

'The matika is that which makes perfectly lucid and explicit the distinguishing points of that
which ought to be known. Thus it comprises t h e four satipatthanas, the four right efforts, the
four bases of psychic powers, the five spiritual faculties, the five spiritual powers, the seven
enlightenment-factors, the noble eightfold path, the four kinds of intrepidity [vesarajja], the four
discriminations [patisambhidha], the four fruits of asceticism [samannaphala]27, the four words of
the Dhamma [dhammapada]28, non-conflict [aranadhamma?], remote samadhi [pantasamadhi?],
empty, signless, and undirected samadhi, development of samadhi, right penetration
[abhisama_ya], conventional knowledge [sammutinana?], samatha and vipassana, DhammasanganT,
Dharmaskandha - this is in what consists the matika...' 29

Here appear, yet again, the 37 wings to enlightenment. The standard wisdom topics - aggregates,
etc. - do not appear. Most of the items are dhamma topics, but the final two are titles of books in
the existing Theravada and Sarvastivada Abhidhamma Pitakas respectively. These were probably

28
extensive Abhidhamma books, possibly sharing a common basis with their existing namesakes,
and would have dealt with the wisdom topics according to the Mulasarvastivada system.

It is a puzzling feature, highlighted in t h e narrative of t h e Buddha's passing away, but implicit

through t h e traditions, that t h e Buddha is said to have encouraged the Sangha to recite the angas;
but the recorded traditions of the First Council say nothing of the angas. What is the missing link
here? Is there some hidden connection between t h e angas and t h e existing scriptures?

T H E T W O SAMYUTTAS

We know that the existing Samyuttas are substantially disordered, if only because they vary
between the Pali and Chinese. Since it is obvious that they are closely related, we are bound to
inquire as to why they differ. Perhaps one is right and the other is wrong, or more likely, each has
diverged in its own way. Any structural similarities may be taken as suggestive of a common
inheritance.

Before looking more closely at the contents of t h e Samyutta, we must briefly remark on a few
confusing terminological ambiguities. The word 'samyutta', which means 'connected', in this
context primarily refers to a collection of discourses on a certain Dhamma theme. Thus we have
the 'Khandha-samyutta', the collection of discourses on the five aggregates; the 'Salayatana-
samyutta', the collection of discourses on the six senses, and so on. Sometimes the 'connection' is
not a Dhamma theme, but some other criterion, such as literary style (Sagatha Vagga, Opamma-
samyutta), or a person (Anuruddha-samyutta, etc.). These samyuttas are then gathered together
in a large collection called the 'Samyutta Nikaya' or 'Samyukta Agama', which is t h e 'Collection of
samyuttas'. Thus the word 'samyutta' can be used to refer either to this large overall body (in
which case we conventionally capitalize it as 'the Samyutta') or to the individual topics (which we
write in lower case as 'samyuttas'). There is a similar ambiguity in the word 'vagga'. This is used
in the sense of 'book' as a term for each of the five great divisions into which the Samyutta as a
whole is divided. Each of these 'Vaggas' (capitalized) includes a number of samyuttas, and is
usually named after its largest samyutta, which is usually also its first samyutta. But t h e more
important sense of 'vagga' is the small scale (and hence lower case) use within the samyuttas,
where it refers to a group of usually ten discourses. The overall 'vertical' structure of the
Samyuttas is therefore layered like this:

Samyutta Nikaya/Agamma ('Collection of samyuttas')

I
Khandha Vagga ('Book [whose first and major section is] on t h e aggregates'), Salayatana Vagga,
etc.
I
Khandha-samyutta ('Samyutta [collection of discourses] on the aggregates'), Radha-samyutta, etc.
I
Nakulapitu-vagga ('Chapter [starting with a discourse to] Nakulapita'), etc.
I
Nakulapitu-sutta ('Discourse to Nakulapita'), etc.

Here is the comparison between the basic structure of the reconstructed Chinese and the existing
Pali Samyutta.

29
Table 3.1: The Sarvastivada and Theravada Samyuttas

Sarvastivada Samyukta Agama Theravada Samyutta Nikaya

1. Sagatha Vagga
2. Nidana Vagga
1. Khandha Vagga 3. Khandha Vagga
2. Salayatana Vagga 4. Salayatana Vagga
3. Nidana Vagga
4. Savakabhasita Vagga
5. Magga Vagga 5. Maha Vagga (=Magga Vagga)
6. Buddhabhasita Vagga
7. Sagatha Vagga

The discrepancies are not as great as would appear. Most of the material found in t h e
Savakabhasita ('spoken by disciples') and Buddhabhasita ('spoken by the Buddha') Vaggas is
distributed in the minor chapters addended in t h e second through fifth Vaggas in the Theravada.
The Khandha, Salayatana, and Magga Vaggas are in the same sequence in both collections. Only
the Nidana and Sagatha Vaggas have moved. There is internal evidence of reshuffling of these
books within the Theravada canon. The Sagatha Vagga consists of discourses with verses, and is
thus clearly distinguished from the remaining prose collections. But there is one other samyutta
with verses; this is the Bhikkhu-samyutta, which in the Theravada is at the end of the Nidana
Vagga. As acknowledged by Bhikkhu Bodhi, this suggests that it originally belonged to t h e
Sagatha Vagga, a suspicion confirmed by the fact that t h e Bhikkhu-samyutta is indeed found in
the Chinese Sagatha Vagga.30 Bucknell has f u r t h e r shown that the Pali commentary hints of a
time when the Nidana Vagga, not the Sagatha Vagga, was the first book of the Samyutta Nikaya. 31
There is a standard passage found in the commentaries to all four Nikayas that describes the
different reasons why the Buddha teaches - in response to a question, f r o m his own inspiration,
etc. In the DTgha, Majjhima, and Anguttara commentaries this appears in the commentary to the
first discourse of the collection; but in the Samyutta it appears in the commentary to the first
discourse in the Nidana-samyutta. Since the Bhikkhu-samyutta is at t h e end of the Theravada
Nidana Vagga but at the beginning of the Sarvastivada Sagatha Vagga, it seems plausible that the
rearrangement came about simply by misshelving: there were some pages left blank at the end of
the Nidana Vagga, so t h e scribe began the Sagatha Vagga by writing the Bhikkhu-samyutta of t h e
same manuscript, but subsequently an unmindful monk took the Nidana Vagga out as t h e first
book and replaced it as the second book (or took the Sagatha Vagga out as the second and
replaced it as the first.) Thus the Bhikkhu-samyutta became separated f r o m its natural pair, the
BhikkhunT-samyutta. It must be admitted that in this respect t h e Chinese maintains a more
rational and probably more authentic tradition than the Pali. So while it is not possible to resolve
all the differences in structure between the two Samyuttas in any simple way, we can point to
some clear cases of disarrangement of texts, specifically involving those books that are in
different sequence in the two collections. These kinds of disarrangements may clearly have
contributed to the divergence f r o m a common ancestral Samyutta to the collections existing
today.

It has been pointed out by several scholars that the overall structure of t h e Samyutta
Nikaya/Agama corresponds roughly with the four noble truths. Bhikkhu Bodhi notes that this
correspondence is more apparent in the Chinese than the Pali.34 The five aggregates and six sense
media pertain to t h e first noble truth; dependent origination (Nidana-samyutta) to the second
and third; and the path is the fourth. We may refer to these fundamental topics in a general sense
as the 'samyutta-matika'. We mentioned above that the backbone of this Magga Vagga is the 37
wings to enlightenment; in the Chinese these are preserved in a sequence that more closely
30
follows the standard Sutta sequence. 33 We therefore have a number of indications that the
Chinese is more structurally reliable than the Pali: the position of the Bhikkhu-samyutta; the
overall correspondence with the four noble truths; and the sequence of the wings to
enlightenment.

We have become so used to considering the 37 wings to enlightenment as a standard doctrinal set
that we automatically think that the Magga Vagga was assembled by taking the list and collecting
appropriate discourses under each topic. But perhaps the situation is the reverse: perhaps the list
of 37 wings to enlightenment has been abstracted from the Magga Vagga topics. The discourses
came first; they were collected according to topic; the collections were given titles; the titles
became used as a shorthand way of referring to the collection; and then the titles became
established as an independent list, repeated and elaborated in countless later works, with their
origins in the Samyutta largely forgotten.

If this theory is true it might help to explain some puzzling features of the list. For example, the
five spiritual faculties and five spiritual powers consist of exactly the same dhammas, and there is
no obvious reason why this set is repeated. Traditionally they are explained as being the same
qualities at different degrees; but this is not how the Samyutta sees them. 38 The situation is even
stranger in the Theravada Samyutta, for the Bala-samyutta is virtually redundant, being just a
repetition series on the spiritual powers. But t h e Sarvastivada Bala-samyukta has a substantial
collection of texts, gathering together many of the discourses on different sets of 'powers' that in
the Theravada are scattered about the canon, including the five spiritual powers. This is surely
more likely to represent the original collection. The Indriya-samyutta, likewise, has discourses
dealing with various sets of faculties in addition to the five spiritual faculties - sense faculties,
feeling faculties, etc. In the Abhidhamma Vibhanga these became fixed into the classic set of 22
faculties. If, then, we compare the two samyuttas, one on various faculties, the other on various
powers, they contain substantially different teachings, with the five spiritual faculties and five
spiritual powers being t h e only overlapping sets. So there is no problem understanding why there
should be two collections, one on the faculties, one on the powers. It is only when the titles
become abstracted and considered to apply exclusively to t h e five spiritual faculties and the five
spiritual powers that they appear redundant.

It begins to seem as if the entire Samyutta is a massive exposition of t h e four noble truths. This
would be the traditional assumption of the schools; books on exegetical method such as the Netti
and the Petakopadesa teach that the correct understanding of any discourse requires that it be
examined in the light of the four noble truths. But now we can give this traditional interpretation
a more concrete literary form. We have seen that at least some traditions treat t h e
Dhammacakkappavattana Sutta as the first discourse in this collection, and the seed around
which the collection crystallized. This would suggest that the Sacca-samyutta, containing t h e
Dhammacakkappavattana Sutta, would have been the first collection. This is not so now: it is in
the Nidana Vagga of the Sarvastivada and the end of the Maha Vagga of the Theravada. This type
of ambiguity in position is found throughout the attempts to systematize the relationship
between the truths and the other doctrinal categories. It seems to result f r o m the limitations
caused by imposing a linear textual structure on the organic structure of the Dhamma itself. The
truths, though starting off as the overarching framework within which the other teachings are
encompassed, come to be treated as just one more doctrinal item in the list. But they have no
specific position within the list and can occur in almost any position. It appears that the
correlation with the truths was uppermost in the minds of those who originally assembled the
collections, but that for later generations this memory became dimmed. If our thinking is sound,
therefore, we may infer that the original groundplan of the Samyutta reflected the truths even
more closely. Such an inferred original collection is represented in the table below.

31
T H E R O O T A B H I D H A M M A TOPICS

First, however, we m u s t consider t h e third and last of our strata of texts, t h e Abhidhammas. The
topics elucidated in t h e Samyutta r e m a i n very close to t h e f u n d a m e n t a l doctrines set f o r t h in t h e
earliest discourses. Many of t h e discourses in t h e Samyutta are, in fact, m e r e variations spun out
of those basic texts. The same situation obtains in t h e case of t h e earliest Abhidhammas. Below
are t h e matikas of t h e A b h i d h a m m a texts of t h r e e schools, identified by Frauwallner as harking
back to a common, pre-sectarian ancestor. He gives considerable detail; however h e overlooks a
couple of relevant points. The Dharmaskandha f r a g m e n t s f r o m Gilgit have t h e d e p e n d e n t
origination leading straight on to t h e 5 precepts. This suggests t h e sequence in t h e Chinese has
been disrupted, p e r h a p s by an accidental reshuffling of manuscripts. If we moved t h e wisdom
teachings - f r o m t h e faculties to t h e d e p e n d e n t origination - to t h e start, this would restore t h e
connection between t h e d e p e n d e n t origination and t h e 5 precepts, and would also m a k e t h e
s t r u c t u r e of t h e Dharmaskandha broadly similar to t h e Vibhanga and Sariputrabhidharma (and
t h e Samyutta). We cannot be sure t h e change was accidental, though, for t h e existing s t r u c t u r e is
certainly rational, corresponding with t h e classic t h r e e s o m e of ethics, samadhi, and
understanding. In this aspect t h e Dharmaskandha is similar to t h e later Theravada treatise, t h e
Visuddhimagga. Another curious f e a t u r e of t h e Dharmaskandha is t h a t it omits t h e eightfold
path. We m u s t surely accept t h a t this was p a r t of t h e original treatise, and its loss m u s t be
accidental. Perhaps it simply fell off t h e manuscript, or was misplaced. In t h e following table t h e
factors c o m m o n to all t h r e e texts are highlighted. Even t h e factors n o t shared by all t h r e e of
these texts, however, almost all have their own cognates in either Sutta or Abhidhamma.

Table 3.2: The Three Versions of the Primary Abhidhamma Matika

Theravada Vibhanga Sarvastivada Dharmaskandha Dharmaguptaka Sariputrabhidharma

1 5 aggregates 5 precepts 12 sense media
2 12 sense media 4 factors of stream-entry 18 elements
3 36 elements 4 confirmed faiths 5 aggregates
4 4 noble truths 4 fruits of asceticism 4 noble truths
5 22 faculties 4 ways of practice 22 faculties
6 Dependent origination 4 noble lineages 7 enlightenment-factors
7 4 satipatthanas 4 right efforts 3 unskilful roots
8 4 right efforts 4 bases of psychic power 3 skilful roots
9 4 bases of psychic power 4 satipatthanas 4 great elements
10 7 enlightenment-factors 4 noble truths 5 precepts
11 8-fold path 4jhanas Elements
12 4jhanas 4 divine abidings Kamma
13 4 divine abidings 4 formless attainments Persons35
14 5 precepts 4 developments of samadhi Knowledge36
15 4 discriminations 7 enlightenment-factors Dependent origination
16 Khuddakavatthu 37 Khuddakavatthu 4 satipatthanas
17 Nanavibhanga 38 22 faculties 4 right efforts
18 Dhammahadaya 39 12 sense media 4 bases of psychic power
19 5 aggregates 4jhanas
20 62 elements 8-fold path
21 Dependent origination Unskilful dhammas 40
22 Sangraha41
23 Sampayoga
24 Prasthana

32
T H E THREE STRATA COMPARED

Now that we have some idea of the content of each of the three strata, we can consider the
relationship between them. Here is a comparative list of the main teachings in these strata,
omitting secondary matter. There is no need to justify this table in detail, since most items are
obvious, but a few notes are given in less obvious cases. In the case of the Abhidhamma, I have
identified the precise texts f r o m which the Theravada Vibhanga has sourced its material; the
Dharmaskandha and Sariputrabhidharma share only some of these sources. 42 Variant or dubious
texts are marked with an asterisk.

Table 3.3: The Three Strata

Earliest Discourses Earliest Collection Earliest Abhidhamma

Truths Truths Truths (MN 141/MA 31/EA 27.1)
Aggregates Aggregates Aggregates (SN 22.59/SA 34 Anattalakkhana)
Sense media Sense media Sense media 43
Elements 44 Elements (MN 115/MA 181/Tib*; MN 141/MA
162)
Feelings
Origin of suffering Dependent origination Dependent origination (SN 12.2/SA 298/Skt)
Defilements Defilements 45 Defilements 46
Kamma 47 Kamma
Cessation of suffering Dependent origination Dependent origination (SN 12.1, 2/SA 298/Skt)
Attainment of Nibbana 48 Attainment of Nibbana
Eightfold path Eightfold path Eightfold path (SN 45.8/SA 783*)
Satipatthana Satipatthana (MN 10/MA 98/EA 12.1)
Right efforts Right efforts (SN 49)49
Bases of psychic power Bases of psychic power (SN 51.13)
Spiritual faculties 22 faculties 50
Spiritual powers (Spiritual powers) 51 (AN 5.14-15/SA 675)
Enlightenment-factors Enlightenment-factors (SN 46.3/SA 736, 740, 724*;
SN46.5/SA 733, etc.)
Training (sikkha)52 Training (jhanavibhanga=MN 39/MA 182/EA 49.8,
etc.; Sikkhapadavibhanga = 5 precepts)

This level of congruence is startling, bearing in mind that we established these three strata
independently, without referring to the doctrinal similarities between t h e strata. These topics are
the backbone of the Dhamma, repeated countless times in countless variations through all
schools in all the history of Buddhism. I have, with no great labour, counted over a dozen
Abhidhamma and commentarial works of various schools that are based on these topics. Usually
the framework of the four noble truths can still be discerned underlying the complex surface
structures. It should hardy need saying that this congruence in the content of the doctrinal lists
does not prove that the schools understood the doctrines in the same way. Each of t h e schools
evolved its own interpretation, which differed both in detail and in principle.

T H E S A M Y U T T A - M A T I K A IN THE M A H A Y ANA

The persistence of the samyutta-matika in the Abhidharma is unsurprising. It is more striking

how important it remained for the Mahayana as well. The samyutta-matika is fundamental to t h e

33
structure of the Heart Sutra, and thus to the Prajnaparamita and Mahayana in general, especially
the Madhyamakas, for whom the Prajnaparamita provided the chief textual inspiration. The
Heart Sutra, which is usually dated to the second century of the Common Era, starts with
Avalokitesvara seeing that the five aggregates are empty of 'intrinsic essence' (svabhava), and
then applying this analysis through the wisdom section of the samyutta-matika. The topics listed
are: the five aggregates, the sense media, the 18 elements, dependent origination, and the four
noble truths. These are all thoroughly equated with emptiness, which Nagarjuna, following the
Sarvastivada Nidana-samyukta, had already identified as dependent origination. Thus the Heart
Sutra critiques an interpretation of the topics of t h e samyutta-matika in terms of the
Abhidhamma theory of 'intrinsic essence', replacing that with dependent origination. This
Madhyamaka philosophy can hardly be regarded as an innovation.

The Samdhinirmocana Sutra offers a more explicit (and entertaining) account of what the issues
are about. 53 This text was apparently composed in the second century of the Common Era to
establish the hermeneutic of the Yogacara school. The main thrust is that conceptual
understanding that concerns itself with the details of things - an obvious reference to t h e
Abhidhamma schools - is born of imagination and thought-constructs, and takes these to be
reality, but only with the non-conceptual unification of samadhi is true wisdom born. The
following passage shows how a forest monk can sometimes regard the study monks in their big
monasteries, a perspective that is echoed often enough today:

'The Venerable Subhuti addressed the Buddha and said: "World-honoured One, in the world of
sentient beings, I know a few who state their understanding without pride, but I know
innumerable, untold sentient beings who cherish their pride and state their understanding in a
prideful manner. World-honoured One, once I was dwelling in a grove in a forest. A large number
of monks lived nearby. I saw t h e m assemble after sunrise to discuss various issues and to propose
their understandings, each according to his insight.

"Some proposed their understanding of the aggregates, their characteristics, their arising, their
ending, their destruction, and the realization of their destruction. Others, in a similar fashion,
proposed their understanding of t h e sense media, dependent origination, nutriments, the truths,
the elements. Others proposed their understanding of the satipatthanas, their characteristics, the
states they are able to control, their cultivation, their arising f r o m a state of being non-arisen,
their non-disappearance after arising, and their increase from repeated practice. Others spoke of
the true severance [=right efforts], bases of psychic power, spiritual faculties, spiritual powers,
enlightenment-factors, or of the eightfold path in a similar fashion.

" . a l l of t h e m cherished their pride, and, because they clung to that pride, they were unable to
comprehend the one universal taste of the t r u t h of ultimate meaning."

'Then the World-honoured One addressed Subhuti and said: "This is so, Subhuti, for I have been
awakened to the t r u t h of ultimate meaning which is of one universal taste, most subtle, most
profound, most difficult to fathom. Having been awakened, I declare, preach, explain, and
illumine it for the sake of others. What is it that I have preached, Subhuti? I have preached that
the purified content of understanding in all the aggregates, [text omits sense media], dependent
origination, nutriments, [text omits truths], elements, satipatthana, true severance, bases of
psychic power, enlightenment-factors, and path factors is the t r u t h of ultimate meaning. This
purified content of understanding is characterized as being of one t a s t e .

"Furthermore, Subhuti, once those practicing monks who cultivate samadhi have understood the
suchness of a single group, the selflessness of the teaching on ultimate meaning, they will not

34
 engage in analysing one after the other the aggregates, sense media, dependent origination,
nutriments, t h e truths, the elements, the satipatthanas [etc.]." '54

The text then regularly mentions this list of dhammas as a fundamental paradigm:

' "The World-honoured One in a immeasurable number of sermons has explained the aggregates...
sense m e d i a . dependentorigination... n u t r i m e n t s . t r u t h s . e l e m e n t s . s a t i p a t t h a n a s . right
e f f o r t s . bases of psychic p o w e r . spiritual f a c u l t i e s . [text omits spiritual p o w e r s ] .
e n l i g h t e n m e n t - f a c t o r s . eightfold path.' 55

' "The World-honoured One has designed the other aspect [of his teaching, that is, other than t h e
teaching of ultimate emptiness] that ultimate meaning is without essence in reference to the
pattern of full perfection, the purified content of understanding that is t h e non-self of all things,
that is suchness, that is the pattern of full perfection. This is how the aggregates.sense m e d i a . 1 2
branches of existence [= dependent origination].four n u t r i m e n t s . s i x and eighteen elements
should be explained..[also the] satipatthanas, right efforts, bases of psychic power, spiritual
faculties, spiritual powers, enlightenment-factors, eightfold path. All these should be explained in
this manner.' 5 6

These teachings become so familiar that the text often abbreviates, simply mentioning, for
example 'aggregates, sense media, all discussed above.' 5 7 ; or else 'the five aggregates, the six
internal sense media, t h e six external sense media, and suchlike.' 58 Now, we have referred to this
general list of topics as the 'samyutta-matika'. The affinity between this list and the Samyutta is
undeniable; but in many cases in the Abhidhamma, etc., the situation is complicated by the
addition of other factors. So one might suspect that here we have merely an affinity of ideas,
rather than literal branches of t h e same historical trunk. But let us compare this specific list,
repeated with reasonable consistency throughout the Samdhinirmocana Sutra, 59 with the
Sarvastivada Samyukta. In particular, let us use those samyuttas identified by the
Yogacarabhumisastra as the central doctrinal chapters (on which more below), leaving aside the
minor samyuttas and those spoken by disciples. Angles (<>) are used to indicate where samyuttas
have been omitted. In both cases we preserve the original sequence. We also give the list of topics
in the Yogacarabhumisastra's definition of the sutta anga in the Sravakabhumi section.

Table 3.4: The Samyukta, the Samdhinirmocana, and the Sravakabhumi

Sarvastivada Samyukta Samdhinirmocana Sutra Sravakabhumi60

Aggregates (Radha, Ditthi) 61 Aggregates Aggregates
Elements 62
Sense media Sense media Sense media
Dependent origination Dependent origination Dependent origination
Nutriments (four) Nutriments (four) Nutriments
Truths Truths Truths
Elements Elements
Feelings (no) (no)
< > Sravaka
Pacceka Buddha
Tathagata
Satipatthana Satipatthana Satipatthana
Right efforts (lost) Right efforts Right efforts
Bases of psychic Power (lost) Bases of psychic power Bases of psychic power
Spiritual faculties Spiritual faculties Spiritual faculties

35
Spiritual powers Spiritual powers Spiritual Powers
Enlightenment-factors Enlightenment-factors Enlightenment-factors
Eightfold path Eightfold path (Eightfold) path
Ugliness (of the body)
Anapanasati (no) Anapanasati
Training (threefold) (no) Training
Stream-entry (no) Confirmed confidence
< >

Thus the correlation is not merely close, it is virtually exact. Particularly relevant is the
coincidence of the four nutriments, which is not standard (the Theravada subsumes this topic
under dependent origination), and the specific sequence: dependent origination, nutriments,
truths, and elements, which is also not standard. There seems to be no doubt that, for the
Samdhinirmocana Sutra, the fundamental teachings of the Dhamma are, precisely, contained
within the Samyukta of the Sarvastivada. This may well be the reason why Asanga, in his
Yogacarabhumisastra, chose to comment at length on this specific recension of the Samyukta
Agama, the foundation of the other Agamas. The Yogacara hermeneutic of the Samdhinirmocana
proposes that the understanding of these teachings should be based on samadhi rather than
intellect. We do not have to look far within the early texts to confirm that this, like the
Madhyamaka emphasis on emptiness as dependent origination, was no innovation.

CHAPTER 4 : THE G I S T 3 - THE ANGAS

We can now move on to the last of the major questions of the GIST: what is the relationship
between these backbone texts and the rest of the discourses? In order to approach this complex
question it will be helpful to first consider some more of Yin Shun's findings based on the
Yogacarabhumisastra. This work treats the Samyukta Agama in terms of three angas (sections):
sutta, geyya, and vyakarana. To understand the significance of this we shall have to take another
step back and consider the anga classifications.

The chief significance of the angas is that they are the earliest recorded system for classifying the
teachings. Classifications such as the Nikayas/Agamas, or the Tripitaka itself, are not referred to
in the early texts, and are attested to only much later. They must therefore post-date the Buddha.
But they cannot be very late, for the division in four main Agamas is widely attested and agreed
among the schools. It therefore seems that it must have been taking shape in the pre-sectarian
period; however the wide divergence in internal structure suggests that the Agamas were not yet
settled in detail. Probably each school inherited a large mass of teachings, largely but not wholly
overlapping, and a general arrangement of texts into the Agamas. I think the huge task of
organizing large numbers of monks and nuns to memorize such vast quantities of scriptures must
have been the primary motivation in changing from the canonically authorized system of angas
to the new Agama system. We think of this process taking place in the period between the first
and second Councils. It is problematic to think in terms of an 'original canon', since there is no
particular evidence that the scriptures as a whole were ever considered finalized and universally
accepted in the pre-sectarian period. Nevertheless, there were clearly large bodies of scripture
that were universally accepted before and after the schisms. Since the Agama system was
developed relatively early, then if the angas constitute a still earlier organizational principle, it

36
seems very possible that the angas, or at least t h e kernel of them, existed in the Buddha's own
lifetime.

T H E N I N E AND THE TWELVE

There is a list of nine angas, supposed to constitute t h e sum of the Buddha's teachings, that is
familiar in the early Pali Nikayas: sutta, geyya, vyakarana, gatha, udana, itivuttaka, jataka, vedalla,
and abbhutadhamma. These nine are also mentioned in Mahasanghika texts, 63 and therefore may
pre-date the first schism. This list is usually increased to twelve in the Sanskrit (with the addition
of nidana, avadana, and upadesa). They are all familiar terms, but the exact meaning is
controversial. In the early texts they are simply listed with no f u r t h e r explanation. The later texts
give explanations; but these vary considerably, and involve considerable anachronisms and
improbabilities. Here I will not embark on a comprehensive survey, b u t will examine the sources
that are available to m e with an eye to what gleanings appear reasonable and relevant.

Several of the later items (gatha, udana, itivuttaka, jataka) are the titles of books included in the
Pali Khuddaka Nikaya, and the Theravadin commentaries straightforwardly identify such angas
with the books of t h e same name. If the later angas refer primarily to the Khuddaka Nikaya, it
seems plausible that the first three are connected in some way with the four main
Nikayas/Agamas. However, most modern scholars think that the angas refer to genres of text
rather than to actual collections. For example, Lamotte says:

'This [anga] classification does not correspond to any real division of the canon, but lists the
literary styles represented in the canonical writings. One and the same text can be classified in
several of the styles at t h e same time, depending on which of its characteristics is under
consideration.' 64

Lamotte is quite correct in mentioning the ambiguity of the anga classification. This, however,
does not show that it was never used as a real division of the teachings, only that any such
division would be in some degree arbitrary. Even in the existing Agama system there are many
such ambiguities; for example the Satipatthana Sutta is middle-length, justifying its place in the
Majjhima; but it deals with satipatthana, one of t h e main topics of the Samyutta; and it teaches by
numbers ('one' way path, 'four' satipatthanas), and so one Chinese version places it in the
Ekottara. We shall repeatedly see such ambiguities as constituting 'breaking points', where the
anga system starts to fall apart, no doubt contributing to the emergence of more systematic
organization.

The most basic reason for considering t h e angas as mere styles rather than an actual structure
would probably be that some of the angas, particularly the first three, do not occur as titles of
collections; and because, of those angas that are titles of existing collections, the books bearing
these titles are generally held to have been compiled later than the early discourses where t h e list
first appears. However, it is quite possible that the earliest list may have been shorter, and that as
other books were compiled their names were added to the list. This is a less radical hypothesis
than the proposition that the very idea of a canonical collection in the Buddha's time was
invented and inserted retrospectively. The divergence between the Pali and the Sanskrit lists
confirms that some additions must have been made, at least for t h e extra items in the Sanskrit.
Here we will briefly discuss the later angas before returning to a more in-depth consideration of
the first three. None of these attributions are beyond dispute. However, we can at least establish
that it is possible, even plausible, that they referred to specific groupings of texts, many of which
are still available.

37
Gatha: Refers primarily to the early portions of the Thera/TherTgatha and the Sutta Nipata. The
Thera/TherTgatha now exists only in the Pali, but is frequently referred to in the scriptures of
other schools, so they must have had versions of these uplifting verses that are now sadly lost,
apart f r o m some in Chinese (eg., Theragatha 1018-1050 occur as MA 33), and a few Sanskrit
fragments. 6 5 The Sutta Nipata as a collection is not known outside the Pali tradition, but many of
the individual texts are known (Khaggavisana, Ratana, Muni, Sela, etc.), and even some whole
chapters (Atthaka, Parayana). Several of these were listed as belonging to the Dharmaguptaka
Khuddaka Nikaya, including the Atthaka and t h e Parayana. There is clearly a tendency to collect
the gathas in distinct collections.

Udana: There is evident confusion in the traditions between the texts known in Pali as 'Udana'
and 'Dhammapada'. Sometimes the contents of the two are mixed, and there are Dhammapada-
like texts in Sanskrit and Chinese that are called 'Udanavarga'. The Chinese traditions regarded
the Dhammapada and the Udana as being the same kind of text. Probably only the verses were
originally called udanas, and were later embedded in background material of varying
authenticity. A certain stage in this process is marked by the Pali book called 'Udana'. In the
verses found in the Pali Dhammapada this process was slower or more uncertain. The background
stories never gained full canonical status but, with abundant imaginative elaboration, took shape
as the Dhammapada commentary, which provided the traditional framework within which the
verses were presented. It is worth noting that, even though the commentary was finalized many
centuries after the verses, at least some information is historically verifiable and stems f r o m an
authentic tradition. One of the stories found in t h e Pali Dhammapada commentary is attested in
the Sarvastivadin Madhyama (MA 80 Kathinadhamma Sutta). The background stories to the
verses as recorded in Chinese versions of the Dhammapada have little or no correlation with the
Pali stories.

Itivuttaka: The Theravada commentaries say that this refers to the Pali book of that name;
however I am not convinced. Incidental evidence of authenticity comes f r o m the commentarial
background story, which says this collection was originally memorized by a lay-woman follower
f r o m whom the nuns later learnt it and taught it to the monks. It is unlikely that the monks
would have invented such a story implying that they forgot their lessons. On the other hand
there is no particular reason why that story should be attached to this particular group of texts.
The Itivuttaka is a small anguttara-style text, including a verse summarizing each discourse, and
the title is derived f r o m the characteristic 'tag' at the beginning and end of each discourse: 'thus
it was said'. This tag is entirely 'extrinsic' to the actual teachings and could be affixed to any style
of text. Thus the Itivuttaka is unusual among the angas in that there is no intrinsic relationship
between t h e n a m e of the anga and t h e style of text. The Chinese version of the Itivuttaka has a
similar 'tag', so if it is not original, it is at least not late. There is a class of Vedic literature called
'Itihasa', 'thus it was', i.e. 'stories of the past, legendary histories', which is sometimes equated
with the Mahabharata and the Ramayana. By analogy, itivuttaka could mean 'sayings of the past'.
Itivuttaka might then refer to the legendary histories that are found in the Nikayas/Agamas, such
as the Agganna Sutta, 66 and the CakkavattisThanada Sutta. 67 Notice that these two discourses are
paired in both t h e Theravada and Dharmaguptaka DTghas. The Sarvastivada preferred to place
these in its Madhyama, where, however, they are not paired. This theory finds support in some
sources outside the Theravada, which treat itivuttakas as stories of the past, sometimes
interchangeably with apadanas. For example, Asanga in the Abhidharmasamuccaya says itivuttaka
'narrates the former existences of the noble disciples'; 68 in the Sravakabhumi of the
Yogacarabhumisastra he says it refers to 'whatever is connected with previous practice'. 69

Jataka: It might be assumed that this originally referred to the stories of the Buddha's past lives
found occasionally in t h e four Agamas, rather than the well-known book of the same name, which
is obviously later (although there is some overlap between the two strata in t h e Pali; and at least

38
one of the stories in t h e Pali Jataka book is found in the Sarvastivada Madhyama [MA 60]).
However, the terms jataka, 'bodhisatta' (either 'enlightenment-being' or 'one intent on
enlightenment'), and most of the other specific features associated with the Jataka book do not
occur in the Agama stories of past lives. For example, there is no hint that the 'Buddha-to-be' was
in any sense destined for enlightenment, or was undertaking practices leading to enlightenment;
quite the contrary, the Buddha takes pains to state that the practices that he did in past lives 'do
not lead to enlightenment'. 7 0

If the extended list of angas was added later to the first three, it would seem more likely that
jataka here refers to the book, at least to an earlier version. The canonical Jataka book contains
just the verses that tell the kernel of the story; this work is almost never found independently.
This situation should be compared with the Udana/Dhammapada as noted above. The stories
themselves are contained in the commentary, although they must have been passed down
together with the verses in the oral tradition, for in many cases the verses are cryptic and make
no sense without the story. It is crucial to remember when considering the Jataka literature that
the verses and the 'stories of the past' belong together, and rarely have any distinctively Buddhist
features, being largely a product of t h e folk storytelling tradition. Probably the stories were
adopted by Buddhist teachers originally simply as moral fables. Later some of them became
identified with t h e Buddha in past lives. When they came to be collected as a book it was probably
felt desirable to standardize the literary format; this process not only reflects other verse
collections such as the Udana/Dhammapada, but also the Vinaya, which similarly provides each
rule with an origin story of often dubious historicity (a nidana, one of the later angas). The 'stories
of the present', which give the present-day (i.e. in the Buddha's lifetime) events that were
supposed to have inspired the telling of t h e story, are an external cladding that were added long
after t h e stories were originally told. It is, of course, this 'cladding' that identifies the characters
in the story with the Buddha, his family, etc. in past lives. This justifies the title 'Jataka' ('Birth
Story').

These 'stories of the present' presuppose a stage in the development of the bodhisatta doctrine
significantly in advance of the early discourses. The evolution of this usage can be traced in the
existing Nikayas/Agamas as follows. The term 'bodhisatta' commonly refers to Siddhattha in his
period of striving before enlightenment; this may be taken as the earliest use. This would suggest
that the earliest meaning of the word 'bodhisatta' was 'one intent on enlightenment' (bodhisakta),
rather than 'enlightenment-being' (bodhisattva). There are apparently references in both the
Chinese and Theravada scriptures that acknowledge this meaning. The Mahapadana Sutta, telling
the story of VipassT, uses the word 'bodhisatta' as far back as t h e descent from Tusita heaven and
birth in the final life.71 The Sanskrit version of this text, although incomplete, appears to be
similar in this respect. The same is also found in the Tathagata-acchariya Sutta of t h e Anguttara. 72
The Theravada Acchariya-abbhuta Sutta (evidently an adaptation of VipassT's story to 'our'
Buddha) extends the scope of the term back to the previous birth in Tusita. 73 The Sarvastivadin
version of the same text takes the significant step of claiming that in the time of Kassapa, the
immediately preceding Buddha, the bodhisatta made the vow to become a f u t u r e Buddha, an idea
not found in t h e early tradition. 78 From there it would take no great leap of imagination to
conceive of the Buddha-to-be toiling through countless lives in his determined struggle for
Buddhahood.

Vedalla: Another problematic term. It is used as titles for two discourses in the Majjhima Nikaya
(MN 43/MA 211, MN 44/MA 44; the versions differ in some questions and details, but there is no
obvious sectarian divergence). The two discourses occur together in both the Theravada and
Sarvastivada. In t h e Sarvastivada they form the last pair in the second-last chapter; thus, bearing
in mind that textual units seem to frequently move about in chapters (vaggas) of ten or so
discourses, they might at one stage have been the final discourses in t h e Majjhima. 75 However, the

39
title 'Vedalla' is only used in t h e Theravada; the Sarvastivadin equivalents are named after the
protagonists, Maha Kotthita and Sister Dhammadinna. The word vedalla does not appear in t h e
body of the texts at all, so, like the Itivuttaka, it is quite possible that the term was simply tacked
on to the discourses at a late date. In fact, they are straightforward vyakaranas, and there seems
no reason to invent a separate class of literature just for them. The Theravadin commentaries
also include a number of similar discourses under vedalla. All of these are found in the existing
canon except the 'SankharabhajanTya Sutta' (implausibly identified with t h e Sankharuppatti
Sutta). The term bhajaniya ('exposition'), f r o m the same root as 'vibhanga', is rare in the early
discourses, and its most familiar appearance is in the chapter titles of the Abhidhamma Vibhanga,
which is not too dissimilar to the existing vedallas. The PTS Dictionary is unsure about the
derivation of vedalla, improbably suggesting veda + any a. But there is a root dala (= English 'tear'),
with t h e basic meaning of 'break, split'. This is used for the names of certain flowers, in the sense
of the unfolding and separating of the petals and leaves. The form vidala, attested in Sanskrit with
the meaning 'dividing, separating', is exactly parallel with vibhanga, and vedalla is just the
abstract form of this. One possibility therefore would be that vedalla was originally the proto-
abhidhamma text we call *Vibhanga Mula. However, given the uncertainty of the term vedalla,
and the tenuousness of the links with the existing texts, we are unable to draw any conclusions
here, except that the two texts now called vedalla were probably paired up together before
incorporation in the Majjhimas. The Sanskrit t e r m here varies, but is usually vaipulya, 'abundant',
which was a standard term for the sprawling Mahayana Sutras. It is likely that the later
translators or editors substituted this familiar t e r m for the problematic and obscure vedalla, not
coincidentally lending credibility to the disputed claims for the authenticity of the Mahayana. In
fact the Abhidharmasamuccaya (a Mahayana Abhidhamma authored by Asanga) explicitly
identifies this anga as contained in the 'Bodhisattvapitaka', and says that the three variant terms
all refer to t h e same thing: vaipulya (because it helps all beings, and is profound), vaidalya (=
vedalla, said to mean that it 'shatters all obstacles'), vaitulya, (explained as 'incomparable'). 76 Only
the second derivation appears reliable, although as said above, the meaning is more likely to be
'splitting' in the sense of'analysis'.

Abbhutadhamma: Probably the most straightforward of the angas to interpret, this obviously
refers to such discourses as the Acchariya-abbhuta Sutta (MN 123/MA 32), the Bakkula Sutta (MN
124/MA 34), etc., which discuss the 'marvellous qualities' of either the Buddha or various
disciples. Venerable Ananda is closely associated with this type of literature. In t h e Theravadin
canon these two outstanding examples of the genre are found together in the Majjhima. In the
Sarvastivada, not only do these two remain close together, but they form part of a chapter of t h e
Sarvastivadin Majjhima called the 'Abbhutadhammavagga'. This chapter also includes a discourse
where the Buddha praises Ananda's 'marvellous qualities' (MA 33). Buddhaghosa's description of
abbhutadhamma refers to what is probably a similar text, which is now found in the Theravada
Anguttara and DTgha (DN 16.5.16). In fact, at AN 4.127-130 there is a group of four texts of this
type. Moreover, there is a cluster of five discourses of this type found together in the Anguttara
eights, and in the Sarvastivada Madhyama Abbhutadhammavagga. There is thus clear evidence
that abbhutadhamma refers to a distinct group of texts. The Bakkula Sutta is an interesting case. It
is clearly late, both on internal evidence, and on the statement of the commentary that it was
added at the second council. This is one of the few direct admissions of a text added after the first
council, and it clearly refers to a text belonging to one of t h e later angas. The only other similar
statements known to me refer to gathas, in the Thera/TherTgatha and the Maha Parinibbana
Sutta. These also belong to later angas. This is perhaps as close as we will get to an
acknowledgement by t h e traditions that the later angas were added at the second council or later.

The following three angas only occur in the Sanskrit lists.

40
Upadesa: This means 'instructions, directions'. It occurs mainly in later texts. I do not know of any
early texts in the Sanskrit traditions that use this title. However, later Chinese scholars including
Hsuang Tsang say that upadesas are treatises that explain the sutras. 77 This suggests a connection
with the Petakopadesa, ('Instructions in [interpretation of] the Pitaka'). This is an early work on
exegetical technique, accepted into the Khuddaka Nikaya only in Burma; some suggest that its
origin lay outside the Theravada school. Asanga in the Sravakabhumi describes upadesa as 'all the
matika and abhidhamma' that explain the Suttas; in t h e Abhidharmasamuccaya he characterises
it as 'the precise, profound, and subtle teaching of the characteristics of all things.' 78

Apadana: This term, spelt avadana in Sanskrit, is very similar to jataka, usually referring to
biographical stories of the past. In Pali, t h e apadanas usually refer to the past lives of disciples,
while jatakas are past lives of the Buddha; b u t this distinction is not maintained consistently.
According to Cone, the term has the basic meaning of 'cutting, reaping, harvest', applied in
metaphorical sense to 'reaping' of t h e fruits of one's actions (especially in past lives), and then to
a story about kamma and result. The earliest usage in this sense is the Mahapadana Sutta, which,
however, also includes elements of abbhutadhamma, vyakarana, udana (or gatha), geyya, and even
Vinaya. This last detail suggests a connection between Apadana and Vinaya, which we will
consider f u r t h e r below. Perhaps the early life stories found occasionally in the Nikayas/Agamas,
such as the MaratajjanTya Sutta, could be regarded as apadanas; however the term itself is not
used in these contexts. The t e r m is the title of a book found in the Pali Khuddaka Nikaya, which
tells the stories of the previous lives of the arahants. Similar works are found in the Sanskrit
traditions.

Nidana: This refers to background or source material. Here it may refer to the Jataka Nidana, one
of the early Buddha biographies. The Abhidharmasamuccaya says nidana is 'a declaration made
[by the Buddha] when he is questioned, or it is t h e declaration of a precept with its cause.' 79 The
first of these explanations would seem to apply rather to vyakarana. However, it may refer to such
episodes as when, for example, the Buddha gave his enigmatic smile, and when asked by
Venerable Ananda for the reason for this, he responded by telling a story of the past. The second
explanation, connecting nidana with the Vinaya, is clearly in accord with a straightforward early
usage.

There are, in fact, many ancient sources in Chinese, etc, that connect both nidana and apadana
with the Vinaya. 80 The nidanas are the origin stories for the rules, while the apadanas are the other
tales included for edification, especially those that compare events in this life with those in past
lifetimes. The Sravakabhumi, while including apadana in the Suttas, says that nidana is the Vinaya,
while upadesa is the Abhidhamma. This suggests that the difference between the nine and the
twelve angas is not necessarily a matter of historical growth of sectarian material, but that t h e
ninefold category includes only the Suttas, while the twelvefold includes both Vinaya and
Abhidhamma as well.

This ambiguity of classification reflects the fascinating way the Vinayas intertwine the
hagiographic and the prosaic. For example the locus classicus for the apadana, the Mahapadana
Sutta, though largely hagiographic, also includes some narrative material in common with the
Vinaya, suggesting that it might be considered, along with the Maha Parinibbana Sutta and the
Catusparisat Sutra, as occupying a position ambiguous between Dhamma and Vinaya. These three
texts form the basis for all later biographies of the Buddha, such as t h e famous Mahavastu of the
Mahasanghika Lokuttaravada, a text which frequently uses the word avadana, includes many
avadanas, and is sometimes suggestively referred to as the Mahavastu-avadana. An apadana is a
story that forms a parable or simile; in other words, one which points to a greater reality outside
the mere events recorded, in particular, a life story that forms a spiritual paradigm for emulation.
In this respect, the Buddha's own life story, the 'Great Apadana', sets the form for all that follow.

41
The parallels between the careers of all the Buddhas presage the t h e m e of repeated patterns,
recurring ethical choices, for good or ill, followed by the inevitable results, cycling on through
the births and deaths of the ages. We have versions of this story in several recensions, each vying
with t h e other in profusion of detail and magnification of glorious and magical embroidery. A
fascinating sidelight is thrown on the interrelationship between these tales by the colophon at
the end of the Abhiniskramana Sutra, translated under the title The Romantic Legend of Sakya
Buddha:

'It may be asked: "By what title is this Book to be called?" to which we reply, the Mahasanghikas
call it "Mahavastu"; the Sarvastivadins call it "Maha Lalitavistara"; the KasyapTyas call it
"Buddha-jataka-nidana"; the Dharmaguptakas call it "Sakyamuni-buddhacarita"; the MahTsasakas
call it "Vinaya-pitaka-mula".' 81

Thus each school would have its own version, yet each regarded as merely a different perspective
on the same theme. The Abhiniskramana Sutra includes some remarks, perhaps by the later
translators, on some of these variations. For example, as to t h e crucial question of how far did the
Bodhisatta's horse travel on the night of his escape from the palace; t h e text says two leagues, the
Mahasanghikas say twelve, but the Theravadins say a hundred. 8 2 The 'Great Story' of the Buddha
was subject to expansion almost as limitless as the round of samsara, and yet even in the most
elaborate versions, the basic teachings, the Dhammacakkappavattana Sutta, etc., recur in almost
identical form, like little nuggets of gold washed along in a stream; the stream is constantly
changing, though keeping roughly the same course, but t h e nuggets remain untarnished and very
slow to change. As a literary style, this may again be compared with several of the other angas we
have discussed above. The Patimokkha rules that form the core of the Vinaya, for example, are, in
the existing Vinayas, embedded in origin stories (nidanas). But while the rules are almost identical
form in all the existing Vinayas, we find considerable variation in the stories; examining the
Lokuttaravada BhikkhunT Vinaya, I was surprised to find that most of the origin stories hardly
share any common elements. We have noted a parallel situation obtaining in the case of the
verses; most of the collections of verse - Dhammapada, Jataka, Udana - come with their own
background stories, but while the stories and t h e verses may have originated at the same time, it
is the verses that were fixed in their current form earlier, embedded in a body of prose of varying
flexibility.

T H E FIRST THREE ANGAS

So it seems that, despite all the uncertainty, there is no very strong reason to accept t h e view that
the angas were merely literary genres rather than organized bodies of scripture. Many of the
later angas can be connected in some way with titles of existing texts. Even in the case of those
terms that are not titles of independent books, such as vedalla and abbhutadhamma, the relevant
texts are fairly consistently gathered together in the existing collections. They have clearly
exerted structural influence on the existing canons. This may give us a clue as to why the scholars
have dismissed the anga classification. Perhaps they have been predisposed to think in terms of
books, and when some of the angas cannot be identified as books, they conclude that they are
merely literary styles. But we have now found a 'middle way': the angas tend to be grouped
together within the larger collections. Such recognizable bodies of texts might naturally evolve
into distinct books. If this was the case with the later angas, it seems reasonable to postulate that
the first three angas also originally existed as recognizable groups of texts, distinct sections
within a larger framework. Given the conservatism of religious literature in general, and
Buddhism in particular, it would seem very unlikely that no r e m n a n t of this structure should be
preserved in the existing canons. We may, therefore, take this as a test: if no trace of the first
three angas can be discerned in the four Nikayas/Agamas our hypothesis should be abandoned.

42
We may first examine more closely the reasons for treating the first three angas as distinct from,
and earlier than, t h e later angas. Yin Shun points out that the Maha Sunnata Sutta, in both the
Sarvastivada and the Theravada versions, lists just the first three: sutta, geyya, vyakarana.83 He
takes this as evidence that these three were historically the earliest. Given the rigid consistency
with which the Theravada texts normally treat this grouping, the appearance of the three alone
in the Maha Sunnata Sutta does indeed call for explanation. The Tibetan translation of this
discourse has t h e usual list of twelve, which, as the translator Peter Skilling notes, attests to a
later stage in the development in the angas. 84 Here is t h e relevant paragraph from the Pali.

'Ananda, it is not worthy for a disciple to follow a Teacher for the sake of suttas, geyyas, and
vyakaranas.85 Why is that? For a long time, Ananda, you have learned the teachings, remembered
them, recited t h e m verbally, examined them with the mind, and penetrated t h e m well by view.
But such talk as deals with effacement, as favours the freedom f r o m hindrances of the heart, and
leads to complete repulsion, fading away, cessation, peace, direct knowledge, enlightenment,
Nibbana; that is, talk on wanting little, on contentment, seclusion, aloofness from society,
arousing energy, ethics, samadhi, understanding, release, knowledge & vision of release - for
such talk, a disciple should follow a Teacher even if he is told to go away.'

Here the three angas clearly refer to a formalized set of scriptures. Note the ambiguity: this is
typical of references to formal learning of Dhamma in the early discourses; learning is
encouraged, but not as an end in itself. It might be noticed that such passages, which are quite
common, are likely to precede the formal redaction of the Tipitaka in Sri Lanka, for there t h e
Sangha decided that scripture came before practice. They would not have invented such passages
that are critical of their own position; in fact it is remarkable that they preserved so many
passages that emphatically place practice over scripture. Mention of Venerable Ananda's close
connection with the three angas here is intriguing; it seems that the Dhamma learnt by him could
be characterized as sutta, geyya, and vyakarana. Given that the traditions ascribe Ananda the
central role in reciting the Dhamma at the First Council, this is an indication that the Dhamma
compiled there might have consisted of these three angas.

Further evidence comes from the Sanskrit Maha Parinirvana Sutra, which has been published as a
complete reconstructed version, and a partial fragment. The list of twelve angas occurs in both
the complete and the partial versions, and although readings vary slightly, in both cases the first
three occur in declined form, as individual words, while the remaining angas are compounded:
sutram geyam vyakaranam gathodananidanavadanetivrttakajatakavaipulyadbhutadharmopadesah.86 This
looks very much as if the original list of three was supplemented later. Exactly the same feature
occurs twice in a Sanskrit list of the twelve angas in the Sravakabhumi of Asanga's
Yogacarabhumisastra. 87 This text then several times gives just the first three, and then simply
says that the list should be expanded as before. 88 The distinctness of the first three is also
suggested in the manner in which Asanga comments on them. In both the Sravakabhumi and the
Abhidharmasamuccaya he says that geyya is 'suttas that require f u r t h e r explanation', and
vyakarana is 'suttas that are fully explained'. 89 This seems to treat t h e two as a closely connected
pair; the explanation is quite close to our interpretation of these terms developed below.

Another interesting case is in t h e two Pali works on textual interpretation, the Netti and the
Petakopadesa. Both of these works regard the four noble truths as the key and the core of the
Buddha's dispensation, and refer all other teachings back to them. The Petakopadesa makes
explicit t h e relation between the Dhammacakkappavattana Sutta and the angas:

'Between the night of his enlightenment and the night of his Parinibbana without grasping,
whatever was spoken by the Blessed One - sutta, geyya, vyakarana, gatha, udana, itivuttaka, jataka,

43
abbhutadhamma, vedalla - all that is the Wheel of Dhamma (dhammacakka) set rolling. There is
nothing in t h e teaching of the Buddha, the Blessed One, outside t h e Wheel of Dhamma. In all his
sutta, the noble dhamma should be sought. In the comprehension of this there is t h e pentad
ending with "light".' 90

These works were composed at a time when the canon was more or less organized as it is today,
referring to such sections as the 'Samyutta Nikaya', etc., and utilizing specialist Abhidhamma
terminology. The Petakopadesa mentions the ninefold angas only twice without explanation, 91
the Netti not at all. This is a bit curious for works that explain in considerable length how to
analyse the discourses.

The Petakopadesa treats sutta very broadly, encompassing all t h e teachings. One of the main
purposes of the work is to describe various principles by means of which one discourse may be
interpreted using teachings found in other discourses. After describing some such principle, it
frequently says that gathas should be assessed with gathas, vyakaranas should be assessed with
vyakaranas.92 While it is not explicit, this looks like our first three angas, although the word anga
is not used; gatha is a synonym for geyya, and geyya is often explained as 'with gathas'. The same
grouping occurs in t h e Netti, though only once. 93

It is not immediately clear what the Petakopadesa is getting at. Perhaps, as Nanamoli's translation
implies, the passage is suggesting that the sutta, the overall teachings, may be divided into verse
(gatha) and prose (vyakarana). This is supported by statements such as: 'Up to this point, however,
the entire sutta - whether gatha or vyakarana - is not [quoted].' 94 But this does not reflect the
original meaning of the angas very closely. The treatment of gatha and vyakarana as sub-divisions
of sutta is reminiscent of Asanga's treatment of geyya and vyakarana mentioned above. It evidently
dates f r o m a time when the scope of sutta had been expanded f r o m meaning one section of the
teachings to meaning all of them.

There is at least one passage that is closer to our usage, and even involves the samyutta matika.
The Petakopadesa describes the six 'ways of entry' (otarana - more on these below) - aggregates,
elements, sense media, faculties, truths, dependent origination - and says that 'there is no sutta or
gatha or vyakarana in which one or other of these six dhammas is not apparent.' 99 Nanamoli
translates this passage differently, saying 'there is no Thread [sutta], whether verse [gatha] or
prose exposition [vyakarana]...'96. This isjustifiable given the more usual use of these three terms
in this textwe have noted, yet my edition of the Petakopadesa clearly has 'suttam va gatha va
byakaranam va' Given the very bad corruption of the text, it is obviously unwise to make too
much out of such details. Yet there remains the suggestion that t h e Petakopadesa remembers a
time when the texts, which all constituted elaborations on the first sermon, consisted of suttas,
gathas, and vyakaranas dealing with the topics of the samyutta matika.

Another hint that t h e first three angas were prior to the rest is the fact that the later factors
exhibit considerable variation in both content and sequence, but the first three are almost always
constant. 97 For example the Dharmaguptaka Vinaya lists the contents of the Ksudraka Agama (no
longer existing) as follows: jataka, itivuttaka, nidana, vedalla, abbhutadhamma, avadana, upadesa,
Atthakavagga, Dhammapada, Parayana, (?doubtful title, perhaps 'various problems'), uragavagga.
This seems like an amalgam of the later angas with the existing contents of the Theravada
Khuddaka Nikaya, including several sections currently included in the Sutta Nipata. So it seems
likely that the first three angas were the earliest, or at least were the first to be established as
canonical, while the subsequent angas were gradually elaborated. However, it is not at all obvious
exactly what they refer to. Here a little investigation is called for.

44
As one of the three angas, sutta m e a n s j u s t one portion of the teachings and cannot be a general
term for any discourse, as it came to mean later. The root meaning of sutta is 'thread', and it is
prominently used in t h e metaphorical sense of a thread on which beads are strung. I think sutta
as an anga reflects this metaphor and means 'basic doctrinal statement'. This is similar to the
meaning in the Brahmanical and Jain contexts. An echo of this meaning survives in the Vinaya.
The collection of rules constituting the Patimokkha is called, in the Patimokkha itself, the sutta.
The detailed analysis of those rules is called the sutta vibhanga ('analysis of the sutta'). This
vibhanga material is stylistically similar to the Abhidhamma Vibhanga and probably dates f r o m a
similar period.

The early treatise on exegetical method, the Netti, gives a curious explanation for the word sutta
in the four great references, taught by the Buddha shortly before h e passed away.?8 These great
references declare that if any monk, teacher, lineage, or tradition, no matter how learned and
respected, makes any statement on Dhamma, that statement must be carefully compared with
the Suttas and t h e Vinaya to ascertain whether it can be accepted as t h e Buddha's teachings, or
should be rejected. Now in the Netti, as a work devoted to literary and textual analysis, we would
expect that sutta here would be explained as the Sutta Pitaka. But no - sutta is explained as the
four noble truths. ?? These are, of course, the main doctrinal content of the
Dhammacakkappavattana Sutta, and I believe we have here a relic f r o m an early meaning of sutta:
basic doctrinal statements, especially t h e first sermon. I would suggest that the Netti is
essentially right here, and that when the Buddha told us to take the suttas as our authority in
determining what was truly spoken by the Buddha, he meant primarily those core discourses now
preserved in the main sections of the Samyutta.

The second anga, geyya, is less difficult to interpret. It is fairly consistently regarded as mixed
prose and verse, and both the Yogacarabhumisastra and t h e Theravada commentaries identify it
with the Sagatha Vagga of the Samyutta Nikaya. However, there are geyyas found outside this
collection, too, including a few in the doctrinal samyuttas.

The word vyakarana100 means literally 'answer' (it can also mean 'grammar' and 'prophecy', but
these meanings do not apply here). It is chiefly used in the sense of an explanatory answer to a
doctrinal question. 101 This meaning of vyakarana is very prominent in the Abyakata-samyutta, the
samyutta on the 'unanswered questions', what has been 'not-vyakaran-ed'. The unanswered
questions are, of course, those such as 'does the Tathagata exist after death' and so on. What is
declared (vyakata) by the Buddha, however, is the four noble truths. 102 This alone would suggest
that we look for vyakaranas in the Samyutta, t h e collection built on t h e scaffold of the four noble
truths. Several discourses present us with a numerical series of dhamma inquiries: one question
(panha), one summary (uddesa), one explanatory answer (vyakarana).103 Here the meaning of
vyakarana is particularly clear. Only occasionally do we meet with vyakarana in a more general
meaning of 'declaration', without specifically being an answer to a question; even here, however,
it might in fact be an answer, only the context does not make this clear.104 In the Anguttara we are
told of four kinds of 'answers (vyakarana) to questions': answering by definitive statement, by
analysis (vibhanga), by asking a question in reply, and by placing aside.109 Notice that a vibhanga,
which is a key class of doctrinal teachings, is here described as a kind of vyakarana.

This general understanding of vyakarana is well known, but the particular function of the
vyakaranas as explanations of the suttas is rarely acknowledged. However, Dutt's assessment is
similar. He suggests that 'the Suttas in which Sariputta, Mahakaccayana, or Buddha gave detailed
exposition of the four truths or the eightfold path, or of any tenet of Buddhism or of any of the
pithy sayings of Buddha, should have been included [as vyakarana].' 106 Asanga's
Abhidharmasamuccaya has this to say:

45
'What is a vyakarana? It is the exposition of various present existences of the noble disciples in
relation to their distant past in different locations. Or it is a sutta that is fully explained, since it is
the open exposition of an abstruse meaning.' 107

The first meaning here should be apadana rather than vyakarana. The second is, however, on the
right track: vyakarana is detailed explanations of points stated briefly in the suttas. It is perhaps
surprising that this interpretation is not more widely accepted, for this kind of form is absolutely
characteristic of the Indian literary tradition in general. The laconic, cryptic aphorisms of works
like, say, the Yoga Sutra were obviously intended to be explicated through a teacher-student
dialogue. This literary feature is evident in the vast majority of existing discourses. Rarely do we
find an exposition of any length beyond the basic doctrinal statements that is phrased in a
straight declarative form.

In particular, we almost never find a disciple teaching in this way. Disciples virtually always teach
in the form of a dialogue between two monks, or between a monk and the Buddha, or the
teaching, though given by one monk, is phrased in 'rhetorical' question and answer format. There
are a few exceptions; but they sometimes just prove the rule. In one Majjhima discourse
Venerable Sariputta sees Venerable Rahula sitting meditation and exhorts him: 'Develop
anapanasati, Rahula! When it is developed and made much of, anapanasati is of great fruit and
benefit.' This is a straightforward sutta; and it is a direct quote f r o m the Samyutta. This
association of teachings by the disciples with vyakaranas agrees with the Yogacarabhumisastra,
which includes the section 'Spoken by Disciples' within the vyakarana anga.

A good example of the meaning of vyakarana can be gleaned f r o m t h e following passage.

On one occasion, many senior monks were staying at Macchikasanda, in the Ambataka Forest.
Now on that occasion, when the senior monks had returned f r o m their alms round, after their
meal they assembled in t h e pavilion and were sitting together when this conversation arose:
'Friends, the "fetter" and the "things that fetter": are these things different in meaning and also
different in phrasing, or are they one in meaning and different only in phrasing?'

Some senior monks answered (vyakarana) thus: 'Friends, the "fetters" and the "things that fetter"
are different in meaning and also different in phrasing. But some senior monks answered thus:
'Friends, the "fetters" and the "things that fetter" are the same in meaning and differ only in
phrasing.' 108

In this case t h e senior monks were upstaged by Citta the householder, who explained how they
were truly different in meaning:

'...the eye is not the fetter of visible forms, nor are visible forms the fetter of the eye; but rather
the desire and lust that arise there in dependence on both - that is the fetter there.'

Now, I think the usage of vyakarana in such passages is exactly what t h e vyakarana anga is all
about. Evidently the Petakopadesa is thinking along similar lines, for it refers to this very
discourse 'in the Citta-samyutta' as vyakarana.109 Notice that the reply is phrased in terms of the
six sense media; Citta is adapting a specific sutta of the Salayatana-samyutta (SN 35.109/SA 239) to
make his vyakarana.

There remains some ambiguity about the exact boundaries of the vyakarana form, due to the
virtual omnipresence of the 'rhetorical question' format. If we were to strictly admit only
discourses with no questions at all as suttas, we'd be left with hardly any; even t h e
Dhammacakkappavattana Sutta features one or more questions in some of its versions. It seems

46
reasonable to count those discourses that are simple, with a small number of rhetorical questions,
as suttas, and the more complex discourses, with a series of questions, as vyakaranas. This
obviously leaves us with some grey areas, which is only to be expected; however, we are usually
able to distinguish fairly readily between the two types.

T H E THREE ANGAS AND THE FIRST DISCOURSES

Let us consider again the first discourses. These fall naturally into three divisions. The first
sermon, the Dhammacakkappavattana Sutta, is the root text setting forth the most basic general
statement of the doctrine. The second and third sermons are similar to each other and belong
together in t h e second division. They can clearly be seen as deriving f r o m a more detailed
exposition of ideas mentioned briefly in the first sermon: the first commentaries. In addition they
emphasize a new literary device. The Dhammacakkappavattana Sutta is almost entirely phrased
as a direct, straightforward statement of doctrine. The second and third sermons emphasize a
question and answer format. While sometimes these are purely rhetorical, in t h e second sermon
the monks actually reply; thus for the first time we hear the voices of the disciples alongside t h e
Buddha. The third division is the Request of Brahma. This has a different literary form, being in
mixed prose and verse.

This threefold division corresponds closely with the root meanings of the three angas. The
Dhammacakkappavattana Sutta was, I believe, the paradigm for t h e sutta classification. The main
paradigm for the vyakarana anga is the Anattalakkhana Sutta, together with the Adittapariyaya
Sutta. The third division consists of the Request of Brahma, which may be seen as the paradigm
for the geyya anga. Geyya literally means 'song', and is usually understood to be mixed verse and
prose. The Request of Brahma differs from the previous two divisions not only in literary form,
but also in content and audience. The prose discourses set forth profound philosophical and
psychological propositions in a literal, unembellished manner, whereas the verse is metaphorical
and inspirational. This agrees with Asanga's statement that the geyya anga is those suttas that
require f u r t h e r explanation. And while the prose is addressed to the monks, the verse is
addressed to a non-monastic, specifically a deity associated with the Brahmanical tradition. The
Request of Brahma also includes an unfavourable allusion, by Brahma himself, to the 'impure
teaching among the Magadhans', obviously an aspersion on the Vedic doctrines. This is
significant in that the verse style of the Buddhists evolved directly f r o m the Brahmanical
tradition.

So we can sum up the main distinctive features of the three different divisions, corresponding
with the three angas, as follows.

Table 4.1: The Three Angas

Sutta Vyakarana Geyya

Content Basic doctrinal Detailed exposition Inspirational/ devotional
statements
Style Declarative prose Interrogative prose Mixed prose and verse
Speaker The Buddha only The Buddha a n d / o r The Buddha, disciples, and others
disciples
Context Always monastic Usually monastic Usually with lay people or deities,
often Brahmanical
Paradigm Dhammacakkappavattana Anattalakkhana Request of Brahma

47
T H E ANGAS AND THE VEDAS

We have now covered enough ground to see the relevant connections between this threefold
structure and the three Vedas. The Dhammacakkappavattana Sutta, like the Rg Veda, is the prime
source text. Geyya and vyakarana, like the Saman and Yajur Vedas, are secondary and derivative.
We have seen how the angas multiplied in threes - 3, 9,12 - just as the Vedic literature multiplied
in threes - 3, 6, 12. We have also seen the twelvefold anga occurring in the Jain scriptures. Well
after the effective dissolution of the anga system, the number three is still basic to the Buddhist
scriptures as the Tripitaka, the 'Three Baskets'. This term is used in all the traditions, despite the
fact that only the Pali canon is meaningfully structured into three sections. Evidently, then, the
idea of the Tripitaka exerts a fascination on the Buddhist mind beyond that of a mere
classification system.

Other relations between the Vedas and the Buddhist scriptures are also discernable. The Buddhist
texts are classified in vaggas of usually ten texts; the Rg Veda is classified in vaggas of about ten
lines. The title vyakarana occurs in both the sixfold and the twelvefold Vedas, but there it means
'grammar'; however it also occurs in the Jain angas where it definitely means 'answers'. We have
also encountered the term suttanga to parallel the vedanga; in addition the term suttanta echoes
vedanta.

The influence of the Vedic model would explain the curious sequence of the three angas. One
would expect that sutta should be followed by vyakarana; however the less closely related geyya
appears in between. This reflects the Saman Veda. As we noted above, while all the Vedas are
mainly in verse, the Saman is essentially a book of sacred songs, and so the position of the Saman
as second Veda may have suggested placing the geyya as second anga. I cannot think of any other
reason for this sequence. In fact, we have seen that the existing Samyuttas have moved the
Sagatha Vagga to either first or last position; it seems to have enjoyed a certain independence.

The correlations between the angas and the Vedas are suggestive rather than conclusive. I am not
enough of a Vedic scholar to be able to say anything more definitive on the matter. But I do think
these correlations are significant enough to warrant more detailed investigation in the light of
the GIST. If the three angas were structurally influenced by the three Vedas, this would lend
strong support to the idea that they were the earliest classification of Buddhist texts. It would
also support the idea that the angas were distinct collections of specific texts, not merely literary
genres.

CHAPTER 5 : THE NIKAYAS/AGAMAS

We may now return to examine the claim of Yin Shun, based on the Yogacarabhumisastra, that
the Samyukta Agama consists of the three angas. He identifies sutta as the major doctrinal
collections, geyya as the Sagatha Vagga together with the Bhikkhu-samyutta (which is really just a
stray chapter from the Sagatha Vagga), and vyakarana as the supplementary expositions. The
identification of the Sagatha Vagga is straightforward; the Pali commentaries, too, say that geyya
is mixed prose and verse, 'particularly the entire Sagatha Vagga of the Samyutta Nikaya'.
However, the interpretation of sutta and vyakarana is not exactly along the lines we considered
above. I must be careful here, for having relied on the Yogacarabhumisastra in principle, I am
straight away criticizing the details. But we must bear in mind that Asanga was recording an
opinion long after the Buddha passed away, and long after the anga scheme had ceased to have

48
relevance as a classification of the teachings. As so often in these matters, it is possible that he
has preserved an important truth, at the same time as obscuring certain aspects.

One problem with the Yogacarabhumisastra's interpretation is that it does not correspond very
closely to the root meanings of the terms sutta and vyakarana. In fact, it is not at all clear why the
term vyakarana should have been chosen for this collection. Another point is that t h e Pali
commentaries include the Abhidhamma Pitaka under vyakarana. While this is obviously
anachronistic, it is perhaps not so far off the mark, for t h e earliest Abhidhamma was based on
texts culled largely f r o m the Samyutta; however, these were taken f r o m the major doctrinal
collections, not f r o m the supplementary chapters. This is even more apparent in the Sanskrit
portion of t h e Dharmaskandha that was excavated at Gilgit, which frequently quotes f r o m central
doctrinal texts such as the Nidana Samyukta, often in almost identical phrasing as the Pali, and
usually refers to such discourses as vyakarana.110 Another problem is that Asanga's opinion in the
Yogacarabhumisastra does not seem to agree here with his statements in the
Abhidharmasamuccaya. We have seen that there he describes vyakarana as past lives of t h e
disciples, or 'fully explained' discourses. Neither of these fit well with the vyakarana anga as
implied in the Yogacarabhumisastra, but at least the second interpretation fits well with the GIST.

A closer look reveals that in at least some cases the difference is not so great after all. For
example, the Radha- and Ditthi-samyuttas follow after the Khandha-samyutta in both collections.
They both consist of a series of question & answers on the aggregates. The Chinese adds another
similar group, called 'Abandoning', to these. It seems as if these minor samyuttas have been spun
out of a few discourses of the Khandha-samyutta. These are considered as vyakarana according to
Yin Shun's interpretation following the Yogacarabhumisastra; and they are also vyakaranas by my
reckoning. A similar case is the Anuruddha-samyutta, which is a brief appendix to the
Satipatthana-samyutta.

Notice that these minor samyuttas directly relate to the four noble truths: the aggregates come
under the t r u t h of suffering; the satipatthanas come under the t r u t h of the path. Now, if we
examine the Sarvastivada Samyutta according to Yin Shun's reconstruction, the sutta anga is
based on the four noble truths, but the vyakarana anga is not. So why do these samyuttas not find
a home in the sutta anga? There are other samyuttas that also fit neatly in the four noble truths,
yet according to Yin Shun's reckoning they come under the vyakarana anga. For example, the
Sarvastivada preserves a Kammavipaka-samyutta. This contains about fifty discourses on the ten
pathways of skilful action. There is no Theravadin equivalent in t h e Samyutta, but most of t h e
discourses are found clustered together in the Anguttara tens. This may plausibly be explained as
a samyutta that has been moved f r o m the Samyutta to t h e Anguttara. According to Yin Shun, this
belongs to the vyakarana anga. But the subject of kamma, and the three unskilful roots that are
mentioned in this context, belong to the second noble truth. According to the GIST, these
discourses would have been incorporated in t h e proto-Samyutta within the four noble truths
scheme. Similarly, the Anamatagga-samyutta on the unknowability of t h e beginnings of samsara,
with its repeated refrain about 'beings hindered by ignorance and fettered by craving', would
seem to fit under t h e second noble truth. Or again, j h a n a obviously fits under the fourth truth,
the unconditioned under the third, but these are considered as vyakarana. I think it is more
reasonable to regard such discourses or collections as having been originally gathered under the
mainstream four noble truths collection of suttas and vyakaranas. More detailed work needs to be
done to clarify this question, but for now I am content to stick by my approach.

The evolution of the angas of suttas and vyakaranas would seem to follow a very natural process.
The Buddha would teach the basic doctrines very often. These would early become collected. The
size of the sutta collection is self-limiting, for there are only so many simple variations possible on
the basic doctrines. It would be natural for inquiries into the meaning of these texts to be going

49
on constantly. A very important early stratum of explanatory texts is the vibhangas, a special
kind of vyakarana, which provide the key for understanding the doctrinal frameworks normally
expressed in condensed, almost cryptic form. Many of the Samyuttas have a vibhanga; originally
perhaps all did. Some of the vibhangas missing from the Theravada Samyutta are found in the
Sarvastivada, for example the Bala Vibhanga Sutta, which should be restored f r o m the Anguttara
to the Samyutta.

But while t h e suttas are limited in size, the vyakaranas can be expanded indefinitely. It would not
take long for the original simple classification to become inadequate and for new, more elaborate,
structures to be required. Two possible avenues would suggest themselves at this stage. One
would be to keep on accumulating individual vyakaranas, creating new collections for t h e m out of
the material spilling over f r o m the original three-anga collection. Another possibility is to try to
reduce the growing bulk of the texts and consequent inconvenience by assembling all of the main
explanatory texts into one comprehensive Vibhanga that would serve as a key to all t h e
discourses. I think that the outcome of t h e first development is the Majjhima, the DTgha, and
Anguttara, and the outcome of the second development is the Abhidhamma Vibhanga.

It is normally understood that the four Nikayas/Agamas were compiled first and the
Abhidhamma Pitaka later. But this, while generally true, may need some qualification. The GIST
suggests that the proto-Samyutta was compiled first and that work on the Nikayas/Agamas may
have been ongoing at the same time as the earliest Abhidhamma work. The terms 'abhidhamma'
and 'matika' occur occasionally in the early texts (a matika is a list of doctrinal principles that
serves as the scaffolding for an Abhidhamma work). Although these terms here cannot refer to
the existing Abhidhamma Pitaka they might well refer to some early precursor.

There is little hint in the early texts themselves what this might be. Scholars have seen
significance in a passage in the Theravada Nikayas that mentions 'abhidhamma' and then
discusses the 37 wings to enlightenment. These do in fact form the main framework for t h e
meditation section of the *Vibhanga Mula. This seems to be confirmed by a hint in the
Dharmaskandha, the Sarvastivadin version of the developed *Vibhanga Mula material. The
Sanskrit text quotes f r o m the Samyukta a passage where the Buddha speaks of the 37 wings to
enlightenment; in the Pali these are called 'dhammas', but the Sanskrit calls t h e m dharmaskandha,
which is of course t h e very title of the book where the quote occurs, and the book does indeed
feature those topics. Later Sarvastivadin accounts of the life of Asoka say that Venerable Maha
Kassapa, the special patriarch of the Sarvastivada, recited the matika, consisting largely of the 37
wings to enlightenment, at the first Council; as we have seen, this is confirmed in the
Mulasarvastivadin Vinaya. 111

There is another hint in the Sarvastivadin version of the Maha Gosingavana Sutta. This says t h a t
Venerable Maha Kaccayana is a monk who delights in discussing 'abhidhamma and abhivinaya'. 112
(The Theravada attributes this to Venerable Maha Moggallana, but all three Chinese versions
more plausibly praise Maha Moggallana for his psychic powers.) One of the other Chinese
versions does not mention this in the body of the text, b u t at the end Maha Kaccayana is praised
by the Buddha for his ability to expound the four noble truths. 113 In the Theravada Anguttara, he
is praised as the foremost in those who can 'analyse (vibhanga) in detail the meaning of a saying
given in brief.' 114 The Chinese extols him for his ability to discriminate the meaning and teach the
path. 115 This suggests that the early meaning of 'abhidhamma' should be sought among Maha
Kaccayana's discourses. He is regarded by the traditions as one of the founders of the
Abhidhammas, and his teachings contain just t h e sort of material we would expect - analytical
vyakaranas dealing with the sense media, aggregates, etc. In addition, two of Maha Kaccayana's
discourses are included in the Vibhangavagga. Thus the earliest Abhidhamma consisted of two

50
aspects: the wisdom teachings - truths, aggregates, sense media, dependent origination, elements
- and t h e meditation teachings - the 37 wings to enlightenment.

If the *Vibhanga Mula of the Abhidhamma was derived, not f r o m the completed Sutta Pitaka, but
from t h e same source as the Majjhima and DTgha along a divergent line of development, the
possibility exists that t h e *Vibhanga Mula may preserve some more archaic features than the
existing Majjhimas. This of course would only apply to the basic specification of content, not t h e
existing elaborate form, which is clearly not very early. One possible example of this is t h e
chapter on the elements. The Theravada Abhidhamma Vibhanga mentions 36 elements. These are
all found in t h e Bahudhatuka Sutta, and as some of the elements are not standard there is little
doubt that this was the Vibhanga's source (as well as elaborations f r o m t h e Dhatuvibhanga Sutta).
But the Bahudhatuka Sutta adds, after these 36, five more elements not in t h e Vibhanga. (The
Sarvastivada consistently has 62 elements in both the Bahudhatuka Sutta and t h e
Dharmaskandha.) So it seems likely that the extra five elements were added to the Bahudhatuka
Sutta after it was moved out of the proto-Samyutta. 116 In A History of Mindfulness I examine a much
more striking example of this pattern in the context of satipatthana.

It is important to note that t h e GIST does not simplistically assert that the Samyutta discourses
are early and authentic, while other discourses are later. All the collections contain a mixture of
early and late material. We are talking here about generalities, about a complex process forming
collections of hundreds of discourses. Discourses outside the Samyutta would have come f r o m a
number of sources. Some were originally included in the proto-Samyutta, but were moved out.
Other discourses may have been current in t h e community, but were not included in the basic
collection. In other cases, discourses might have been passed down in remote areas and were
incorporated later. Other discourses may have been formed later, but merely by combining pre-
existing sections of text. Still others evolved out of the relatively informal narrative and
background material associated with the teachings. And some, no doubt, are pure invention.

ANGAS IN THE SAMYUTTA

We saw above that the later angas have clearly exerted an influence over t h e organization of t h e
texts as we have them today. It would seem likely that, if our understanding of the first three
angas is correct, we should be able to discern traces of this system in the existing collections. The
GIST suggests that the major doctrinal samyuttas were based on suttas, supplemented by
vyakarana explanations. This structure can be discerned in the existing Samyuttas in residual
form. It is very prominent in t h e Theravada Salayatana-samyutta (SN 35), where the first 52
discourses are mainly suttas; the fifty-third starts with 'a certain bhikkhu' approaching the
Buddha to ask questions, and thus launches a long series of vyakaranas. (In fact texts 33-52, which
are represented by j u s t one text in the Sarvastivada, SA 196, seem to be spin-offs f r o m t h e
Adittapariyaya Sutta that were multiplied to fill out the fifty texts.) This structure is not so
evident in most of the other sections in the Pali.

However, many chapters in the Chinese seem to reflect this form to some degree (using the
reconstructed Samyukta Agama). I must caution here: there are serious methodological flaws
besetting my research on this point, which is largely based on the existing catalogues on t h e
concordance of the scriptures of different schools. These catalogues are not always accurate, but I
have not in every case double-checked. In addition, cognate suttas are rarely exactly identical,
and I know of examples where a statement is phrased as a question in t h e Pali but as a statement
in the Chinese or vice versa. Bearing these limitations in mind, we find that in, for example, the
Khandha-samyutta the first 14 discourses are suttas, and after that is a long series of mainly
vyakaranas. The Salayatana, Nidana, Sacca, Satipatthana, Bojjhanga, Anapanasati, and Sotapatti

51
samyuttas all seem to reflect this pattern. It will be an interesting project for f u t u r e research to
see whether the sutta/vyakarana structure of these collections is as apparent as it seems from t h e
correspondence tables.

Although we cannot deviate here to check all of these collections in detail, let us take the Sacca-
samyutta, the doctrinal cornerstone of the whole Samyutta, as an example, going back to t h e
Chinese and re-checking all t h e parallels. First let us list the texts in the existing Theravada
Sacca-samyutta, classifying them as best we can into one or other of the three angas.

Table 5.1: Angas in the Theravada Sacca-samyutta

SN 56 Sutta/Vyakarana
1-12 S
13-18 V
19-20 s
21-22 Geyya
23-28 S
29-31 V
32-33 S
34 V
35-41 S
42-43 V
44 S
45-131 V

This table confirms that the three angas influence the structure of t h e existing SN 56. Almost
always, suttas are grouped with other suttas, vyakaranas are grouped with other vyakaranas, and
the two geyyas are paired in the middle. Also, the collection starts with the biggest group of suttas,
and ends with the biggest group of vyakaranas, reflecting the sequence of t h e angas.

The table below shows the occurrence of angas in the existing Sarvastivada Satya-samyukta. I
also list the settings.

Table 5.2: Angas in the Sarvastivada Satya-samyukta.

SA Sutta/Vyakarana Setting
379-402 S Benares
392 Geyya
S
403 Geyya Magadha
404 V Rajagaha (journeying)
405 Geyya Vesali, Monkey's Pond
406 V Monkey's Pond
407-418 V Rajagaha, Veluvana
419-420 S Rajagaha, Veluvana
421-426 V Rajagaha, Veluvana
427-433 S Rajagaha, Veluvana
434-442 V Rajagaha, Veluvana
443 S SavatthT, Jetavana

52
Here t h e tendency f o r suttas to cluster with other suttas, and vyakaranas to cluster with o t h e r
vyakaranas, is even m o r e pronounced. Equally striking is t h a t t h e collection starts off with an
u n i n t e r r u p t e d r u n of 23 suttas. Four miscellaneous texts, including geyyas, t h e n intervene, and
t h e n 11 vyakaranas. It is suggestive t h a t here, as in t h e Theravada, t h e geyyas occur in t h e middle
of t h e collection, reminding us of t h e sequence sutta, geyya, vyakarana. The latter half is slightly
less coherent, b u t still t h e angas are readily discernable, mainly vyakaranas. So t h e collection
clearly suggests an organizing principle of a g r o u p of suttas followed by a group of vyakaranas,
with a few geyyas in between.

But it is t h e settings t h a t offer a startling, and unexpected, confirmation of our thesis. The first
text is t h e Dhammacakkappavattana Sutta, which is of course set in t h e Deer Park at Benares. All
of t h e suttas t h a t follow are also set in t h e Deer Park. It goes without saying t h a t it is implausible
t h a t t h e Buddha t a u g h t all his basic statements on t h e four noble t r u t h s in t h e one place. The
settings for t h e following discourses have been simply mechanically repeated f r o m t h e first.
Exactly t h e same t h i n g has h a p p e n e d to t h e vyakaranas. Leaving aside t h e four intervening
discourses, t h e first of t h e string of vyakaranas is set in Rajagaha at t h e Veluvana, and t h e
subsequent discourses merely p a r r o t this setting. This conclusion is substantially reinforced by
two o t h e r considerations. The first consideration is t h a t these two discourses are virtually t h e
only ones in t h e whole collection to have t h e same setting in t h e Pali and t h e Chinese. The second
consideration is t h a t most of t h e discourses do n o t have any i n h e r e n t indication as to w h e r e t h e y
w e r e spoken. They simply give a doctrinal s t a t e m e n t t h a t could have h a p p e n e d anywhere. But
t h e Dhammacakkappavattana Sutta is deeply embedded in its narrative context, and all Buddhists
r e m e m b e r t h a t it was spoken at t h e Deer Park. The first of t h e vyakaranas, too, internally confirms
its setting, for, being set in Rajagaha, it tells t h e story of a person in Rajagaha.

Now let us combine these two lists by presenting a list of t h e concordance between t h e two
versions of t h e Sacca-samyutta, expanding t h e detail by listing each individual discourse t h a t
occurs in b o t h collections. In most cases t h e Theravada does n o t specify t h e setting; this is
indicated by empty brackets ( ) . It m a y be assumed t h a t t h e setting of these discourses was m e a n t
to be SavatthL Since t h e Sarvastivada seems in several respects to be m o r e structurally archaic,
let us use t h e sequence of texts in SA.

Table 5.3: Concordance of the Two Sacca-samyuttas

SA SN 56 Sutta/Vyakarana Setting in SA Setting in SN

379 11-12 S Benares, Deer Park Benares, Deer Park
382 29 S B ()
390 5-6* S B ()
391 5-6* S B ()
392 22/Iti 4.4 Geyya B Vajji, Kotigama
393.1 3-4 S B ()
393.5 25* S B ()
394 37* S B ()
395 38 S B ()
397 32* S B ()
398 39 S B ()
399 40 S B ()
400 34 S(VinSN) B ()
401 35 S B ()
402 23 S B Savatthi
403 21 Geyya Magadha Vajji, Kotigama
404 31 V Rajagaha (journey) Kosambi

53
405 45 Geyya (V in SN) Vesali, Monkey's Pond Vesali, Great Wood
406 47 Monkey's Pond ()
407 41 V Rajagaha, Veluvana Rajagaha, Veluvana
408 8 V R,V ()
409 7 V RV ()
410 7* V R,V ()
411 10 V RV ()
412 9 V RV ()
416 15 V RV ()
417 20, 27 V (S in SN) RV ()
418 16 V RV ()
421 42 V RV Rajagaha, Vulture's
Peak
422 43 V RV ()
423 46* V RV ()
424 46* V RV ()
425 46* V RV ()
426 46* V RV ()
428 2 S RV ()
429 1 S RV Savatthi
430 33 S RV ()
435 32* V Savatthi, J e t a v a n a ()
436 44* V S,J ()
437 44* V S,J ()
438 36 V (S in SN) S,J ()
439 49/SN13.1* V S,J ()
440.1 52 V RV ()
440.2 53 V ()
440.3 57 V ()
441.1-60 49*, 55*, 59* V S,J ()
442.1-17 51-119 V S,J ()

This table well illustrates t h e kinds of issues facing us in these studies. It is obvious that here we
have two very similar collections; yet the internal sequence of texts is very different. As well as a
large number of equivalent texts, we also have various anomalies: a single text in one collection
becomes two or more in another collection; sometimes the texts display significant variations;
occasionally a sutta in one collection becomes a vyakarana in the other, and so on. Nevertheless, it
is apparent that texts are frequently grouped together as either suttas or vyakaranas, and that this
should be recognized as an important structural principle underlying the formation of the
existing collections. Again the list starts with a long list of mainly suttas, and ends with
vyakaranas. Another feature is that in between the suttas and vyakaranas are a few geyyas. The
position of the geyya SN 56.22/SA 392 is evidently anomalous, for the text is closely related to SN
56.21/SA 403. The two occur together in the Theravada, so the position in SA is evidently just a
fault in the SA transmission. If we assume that its correct position was with the other geyyas, we
can see that at least in this samyutta the geyyas fall in between the suttas and vyakaranas, in
accordance with the sequence of the angas.

Another striking correspondence is that, in several cases, an identifiable group of texts is found in
both collections. For example, take the texts SA 408-412. This group of suttas corresponds with
the group SN 56.7-10. Thus it seems to be a pre-existing unit common to both traditions. A similar
situation obtains with texts SA 394-401, which loosely correspond with SN 56.32, 34, 35, 37, 38, 39,
40. On a smaller scale, several pairs of texts occur together in both collections. These
correspondences raise t h e possibility that not merely the content, but also the sequence of the
texts in the two traditions was shared, at least in part.

54
This suggestion finds spectacular confirmation in the case of the Kassapa-samyutta. This is a
lively collection of vyakaranas that sheds interesting light on one of the great characters of
Buddhism, especially in his historical role. It is one of the few prose samyuttas that is preserved
in entirety in the Theravada and in two of the Chinese Samyuttas, the Sarvastivada as well as the
partial 'other Samyutta' of unknown school, possibly Kasyapiya. Here is the list of texts, using t h e
sequence of t h e restored SA.

Table 5.4: Concordance of the Three Kassapa-samyuttas

Sarvastivada SA 'Other' SA2 Theravada SN 16

1136 111 3
1137 112 4
1138 113 6
1139 114 7
1140 115 8
1141 116 5
1142 117 9
1143 118 10
1144 119 11
905 120 12
906 121 13

The correspondence is bordering on miraculous. The two Chinese Samyuttas are in fact identical
in content and sequence, if we accept Yin Shun's suggested restoration of the two final texts (SA
905, 906). The Theravada is also very close. It has two extra texts at the beginning; these may have
been later additions. And one text has been moved, SN 16.5 'Old'. This text shares in common with
SN 16.8 a substantial passage on such practices as forest dwelling, alms-food eating, and so on.
Thus it seems reasonable that SN 16.5 should follow SN 16.8, and that the two Chinese versions
have in this case preserved the correct sequence. Leaving aside the two extra texts in t h e
Theravada, we have eleven texts, close enough to a classic vagga of ten.

Unfortunately, nowhere else in the prose samyuttas do we encounter such a neat

correspondence. Wherever we turn, we are beset with anomalies and discontinuities. Each
anomaly is a potential fault-line, a fissure through which we m i g h t j u s t be able to discern an older
structure. One obvious explanation for the inconsistencies is that the collection was reorganized
at a later date by people who had either forgotten the sutta/vyakarana structure, or for whom it
had become unimportant. It might, therefore, be possible, relying on a variety of means of textual
sleuthing, to uncover traces of editorial manipulation and to reconstruct what the original angas
might have looked like. This procedure will, however, be speculative, complex, and not essential
for our main argument, so I have relegated f u r t h e r structural investigations to an appendix.

ANGAS IN THE MAJJHIMA

For the remainder of this chapter I will investigate some structural features of the DTghas,
Majjhimas, and Anguttaras, in much less detail than the Samyutta. My main aim is to see if
anything within these collections supports the GIST, or obviously refutes it. Answers to these
questions, if they can be found at all, will require much more specific, detailed study; here I
intend merely to point out some ways of addressing the questions.

55
There are several hints of the influence of the Samyutta in t h e formation of the existing
Majjhimas. Most obvious is that several of the chapters in the Sarvastivada Majjhima are called
'samyuttas'. There is a Kamma-samyutta-vagga, a Sariputta-samyutta-vagga, a Samudaya-
samyutta-vagga, and a Raja-samyutta-vagga. Not only does the word 'samyutta' occur, but also
these are all similar to titles of sections in the existing Chinese Samyutta. Furthermore, chapters
6-10 of the Theravada Majjhima have titles that are similar or identical with titles in the
Samyutta: Gahapati (=Gamani-samyutta), Bhikkhu, [Paribbajaka] (not in the existing Samyuttas,
but would come under the eight assemblies), Raja (=Kosala), and Brahmana. With t h e exception of
the Gahapativagga, these are in the same sequence predicted by Bucknell's reconstructed Sagatha
Vagga. Again, the final chapter of the Theravada Majjhima, the Salayatanavagga, not only shares
its title and subject matter with the Samyutta, but all the discourses are found in the Sarvastivada
Samyutta.

The division into three groups of fifty discourses in the Theravada Majjhima even seems to faintly
reflect the three angas. The first fifty presents t h e main doctrines; although formally these are
mainly vyakaranas, within the Majjhima as a whole they function as basic texts, in a way similar to
suttas. The second fifty as noted above has titles similar to the Sagatha Vagga, includes a fair
number of verses, and is often addressed to lay people, thus being related to the geyya anga. The
final fifty tend to be more analytical and expository, classic vyakaranas, including some proto-
Abhidhamma texts, the Salayatanavagga, and the historically important Vibhangavagga.

The Vibhangavagga is the only chapter that shares both the same title and almost all the same
content in t h e two Majjhimas. The exact title in the Sarvastivada is 'Mulavibhangavagga', the
'Root Vibhanga Chapter', which is very suggestive. Most of the discourses deal with familiar
topics such as the aggregates and sense media. Two of the discourses, t h e Saccavibhanga Sutta
and the Aranavibhanga Sutta, refer directly back to the Dhammacakkappavattana Sutta. (The
Saccavibhanga Sutta is not in the existing Sarvastivada Vibhangavagga; however Bucknell argues
persuasively that it was removed accidentally at a later date). Unsurprisingly, several of the
discourses in the Vibhangavagga are shared in common with chapters in the Abhidhamma
Vibhanga. All this suggests that this chapter, with its demonstrably close connection to t h e
Samyutta, was an early and fundamental division within the Majjhima.

There is one very important framework, comprising perhaps a third of the discourses in the
Majjhimas and DTghas, which is not found, or at least is not prominent, in the existing Theravada
Samyutta Nikaya. This is the 'training' (sikkha). This discrepancy has perhaps been a reason why
the dependence of the Majjhima on the Samyutta has not been noticed. However, the
Sarvastivada Samyukta Agama rectifies this. It has a Sikkha-samyutta containing the basic
discourses on the threefold training, which in the Theravada are now in the Anguttara Nikaya. In
the Majjhima and DTgha this simple threefold training is usually elaborated into the detailed
'gradual training', resulting in many very long discourses. If all the discourses on the training
were assembled, Samyutta-style, into one collection it would be long and cumbersome indeed. It
is plausible that such a long collection would be broken into more manageable pieces, which
would form fundamental portions, perhaps the fundamental portions, of the new collections.

ANGAS IN THE DIGHA

We are uniquely fortunate to have three DTghas available for inquiry. The Theravada DTgha
Nikaya in Pali is well known, and has twice been translated in its entirety into English. The
Dharmaguptaka DTrgha Agama in Chinese is much less known, and only a few discourses and

56
passages have been translated. The Sarvastivada DTrgha Agama is almost entirely unknown, since
only in the last few years have the manuscripts emerged f r o m Afghanistan and been made
available for study (although several of t h e individual discourses have been edited and translated
previously). I have not been able to study these discourses; the following is f r o m Hartmann's
essay detailing the structure of the collection, which he has reconstructed with the help of the
information contained in the uddanas (summaries of discourse titles at the end of each section),
and folio numbers. 117 Hartmann says that these results are now 'close to certainty'.

In addition to these sources, there is some information on a version of the DTgha mentioned in
Samathadeva's AbhidharmakosopayikanamatTka (AKO), available in Tibetan. Hartmann says his
conclusions about the Sarvastivada DTrgha in his earlier work based on this source have been
pleasantly confirmed by the discovery of the actual manuscript; however, others have said that
there are certain differences between the DTrgha as inferable f r o m the AKO and the manuscript. 118
Since the AKO is a commentary on Vasubandhu's Abhidharmakosa, and that work is a Sautrantika
polemic against the Sarvastivada, it would seem plausible that the DTrgha used by Samathadeva
(and Vasubandhu) belonged to the Sautrantikas.

In any case, for now we have plenty to deal with in the three existing DTghas. Below I present t h e
contents of the Sarvastivada DTrgha as reconstructed by Hartmann, together with the cognates in
the Theravada (DTgha and Majjhima) and Dharmaguptaka (DTrgha). Most of these discourses have
several other cognates in Chinese, Tibetan, and Sanskrit, but since we are mainly interested in the
structural principles of the collections, rather than the content of the individual discourses, I only
mention the cognates in the major collections. We may briefly deal with each section in turn. In
the following table: @ means exists in the DA2 manuscript; # means partially exists; * means exists
only as a title in t h e uddana (although might exist in other fragments).

Table 5.5: The Three Dighas: First Section

Sarvastivada (DA2) Theravada Dharmaguptaka

Satsutrakanipata

DA 2 1 Dasottara * DN 34 Dasuttara DA 10,11

DA2 2 Arthavistara * No No
DA2 3 SangTti * DN 33 SangTti DA 9
DA2 4 Catusparisat # (= 1 st chapter of Vinaya Mahavagga) No
DA2 5 Mahavadana # DN 15 Mahapadana DA 1
DA2 6 Maha Parinirvana # DN 16 Maha Parinibbana DA 2

This 'Six-Sutra Group', seems to have been a popular set, a 'greatest hits' compilation. The
Arthavistara, though absent f r o m the main Nikayas/Agamas, is found in a couple of
miscellaneous translations in Chinese. 119 These are the last two of the 72 individual discourse
translations located after the complete Madhyama Agama translation, which suggests that the
compilers of the Taisho edition thought they belonged to the Madhyama rather than to the
DTrgha. They are early translations, particularly T 98, which was done by An Shigao (flourished
AD 148-170), the pioneer of translation into Chinese. The two translations largely agree on
contents. The Arthavistara is delivered by Venerable Sariputta and consists of 23 or 25 lists of
dhammas in the style of Dasuttara and SangTti, but the lists do not occur in ascending numerical

57
order and most have numbers over 10. Of the next three, the Catusparisat tells the fundamental
story of the events following the Buddha's enlightenment: t h e turning of the wheel of Dhamma,
and the establishment of the fourfold assembly. The Theravada and other schools preferred to
include this text in their Vinayas. The Mahapadana contains much that parallels this, but set in
the time of VipassT Buddha; the Maha Parinibbana is the complementary narrative of the close of
the Buddha's life. So this 'Six-Sutra Group' is three lists of basic doctrines, and three key
biographical narratives. This makes it seem like it may have served as a basic curriculum for the
beginners in Buddhist studies. One wonders whether these three biographical narratives might
have comprised the apadana anga.

Table 5.6: The Three Dighas: Second Section

Sarvastivada Theravada Dharmaguptaka

Yuganipata

1 st varga

DA2 7 Apannaka * MN 60 Apannaka No

DA2 8 Sarveka * No No
DA2 9 Bhargava * DN 24 Patika DA 15
DA 2 10 Salya * MN 105 Sunakkhatta No
DA 2 11 Bhayabhairava * MN 4 Bhayabherava No
DA2 12 Romaharsana * MN 12 Maha SThanada No
DA 2 13 Jinayabha * DN 18 Janavasabha DA 4
DA 2 14 Govinda # DN 19 Maha Govinda DA 3
DA 2 15 Prasadikah # DN 28 SampasadanTya DA 18
DA 2 16 PrasadanTya * DN 29 Pasadika DA 17

2nd varga

DA 2 17 Pancatraya @ MN 102 Pancattaya No

DA 2 18 Mayajala # No No
DA 2 19 Kamathika # MN 95 CankT No
DA2 20 Kayabhavana @ MN 36 Maha Saccaka No
DA2 21 Bodha @ MN 85 BodhirajakuMara No
DA2 22 Samkara @ MN 100 Sangarava No
DA2 23 Atanata @ DN 32 Atanatiya No
DA2 24 Mahasamaja @ DN 20 Maha Samaya DA 19

The next section, the 'Paired Group', consisting of two vaggas, contains much unshared material.
However, one shared structural feature is the two pairs of discourses DN 18/DA 4/DA2 13
Janavasabha & DN 19/DA 3/DA2 14 Maha Govinda; and DN 28/DA 18/DA2 15 SampasadanTya & DN
29/DA 17/DA2 16 Pasadika. Both of these two pairs are quite similar in style, and they are found
together in all three collections. They thus seem to belong together, and have maybe stuck
together through t h e various changes in the collections. Apart f r o m this, however, this section
does not seem to share a common structural heritage.

58
This section contains several discourses that, in the Pali, are found in t h e Majjhima. A striking
fact, pointed out by Hartmann and Bucknell, is that none of these have cognates in t h e existing
Madhyama Agama in Chinese. Until the discovery of the Sarvastivada DTrgha, this was a puzzle:
were all these discourses invented by the Theravada? But now t h e puzzle is solved. It seems
certain that the two collections, t h e Madhyama now found in the Chinese canon, and the DTrgha
recovered f r o m Afghanistan, must have originally been edited conjointly, and t h e compilers
strictly avoided duplicating any discourses in both collections. This confirms our belief that they
belong to the same school, and from the internal evidence in the Madhyama, it seems very likely
that this was the Sarvastivada. So rather than these discourses being an invention of the
Theravada, both the Theravada and Sarvastivada had versions of these discourses, but the
Theravada chose to include them in its Majjhima, while the Sarvastivada included t h e m in its
DTrgha. This is a caution against overly rash conclusions based on the absence of a particular
discourse f r o m one collection or other.

These discourses sit quite comfortably in the DTgha, since all of them deal, in one way or another,
with t h e question of the relation between Buddhism and the other contemporary religions, which
is also an outstanding t h e m e of many of the other key DTgha discourses, especially the
STlakkhandhavagga discussed below. The Sarvastivada Vinaya Vibhasa (a Vinaya commentary
preserved in Chinese) in fact says that t h e purpose of the DTgha (presumably the Sarvastivada
DTrgha) was to refute the heretics. MN 60/DA2 7 Apannaka opens with the Brahmans of Sala
hearing of t h e Buddha's good reputation, then deciding to go and see him; this stock passage is
found in discourses 3-7 of t h e Theravada DTgha. MN 95/DA2 19 CankT has a similar DTgha-style
opening and t h e m e (incidentally, a fragment of a Mahasanghika version of this discourse also
exists). MN 105/DA2 10 Sunakkhatta and MN 12/DA2 MahasThanada deal with t h e dismal spiritual
career of the wanderer Sunakkhatta, who also appears in DN 24/DA 15/DA2 9 Patika; thus the
Sarvastivadins felt it better to assemble these discourses together in their DTrgha. MN 4/DA2 11
Bhayabherava is addressed to the Brahman Janussoni, who appears, along with a number of other
Brahmans featured elsewhere in the DTgha, at DN 13.2. This also deals with the Bodhisatta's
practices before enlightenment, which link it up with MN 36/DA2 20 Mahasaccaka, MN 85/DA2 21
Bodhirajakumara, and MN 100/DA2 22 Sangarava, which all treat of the Bodhisatta's ascetic
practices, in response to challenges by either the Jains or t h e Brahmans. MN 102/DA2 17
Pancattaya has been described as a 'middle-length' version of DN 1/DA 21/DA2 47 Brahmajala, a
sophisticated refutation of a range of wrong views. A f u r t h e r unifying DTgha-style trait is that
several of these discourses deal with, directly or indirectly, with the gradual training: MN 60/DA2
7 Apannaka, MN 4/DA2 11 Bhayabherava, MN 36/DA2 20 Mahasaccaka, MN 85/DA2 21
Bodhirajakumara, and MN 100/DA2 22 Sangarava. The other discourse that appears in the
Theravada Majjhima but the Sarvastivada DTrgha is MN 55/DA2 43 JTvaka, which occurs in t h e next
section of the Sarvastivada DTrgha. This is a slightly odd choice, for the text is quite short,
unusually so even for a Majjhima discourse. Nevertheless, it also addresses the theme of the
Buddha's response to criticisms f r o m other religions.

The Majjhima discourses taken together number ten. Strikingly enough, there are also ten
discourses in the opposite situation; that is, they are found in the Theravada DTgha but in t h e
Sarvastivada Madhyama. These are as follows: DN 15/MA 97 Maha Nidana, DN 17/MA 68 Maha
Sudassana, DN 21/MA 134 Sakkapanha, DN 22/MA 98 Satipatthana, DN 23/MA 71 Payasi, DN
25/MA 104 Udumbarika-sThanada, DN 26/MA 70 Cakkavatti-sThanada, DN 27/MA 154 Agganna, DN
30/ MA 59 Lakkhana, DN 31/MA 135 Sigalovada. Given that the standard grouping of discourses is
in vaggas of ten, it seems likely that these differences result f r o m t h e movement of vaggas among
the collections. The question then becomes, where were they moved from? I have not looked at
the question in detail, but I have examined two of the discourses found in the Theravada DTgha
and the Sarvastivada Madhyama, that is, the Satipatthana and the Lakkhana. In both of these
cases the DTgha versions have substantial quantities of extra material, and the added material on

59
internal evidence seems late. Thus it seems plausible that these two discourses were taken f r o m
the Majjhima, padded out, and placed in t h e DTgha. Given that they seem to have been moved as a
group, this conclusion might also hold good for the rest of the discourses, too. Thus we should
entertain the possibility that the Sarvastivada shifted one vagga f r o m the Madhyama into their
DTrgha, and the Theravada shifted a different vagga from their Majjhima into the DTgha. Of
course, it is equally possible that both of these vaggas originated somewhere else, perhaps t h e
Samyutta, and were shifted into either the DTgha or the Majjhima. This could be tested by a closer
examination to see whether the discourses displayed any noteworthy Samyutta-like features.

Table 5.7: The Three Dighas: Third Section

Sarvastivada (DA2) Theravada Dharmaguptaka

STlaskandhanipata

1 st varga

DA2 25 Tridandin @ No No
DA2 26 Pingalatreya @ No No
DA2 27 Lohitya 1 @ DN 12 Lohicca DA 29
DA2 28 Lohitya 2 @ DN 12 Lohicca DA 29
DA2 29 Kaivartin @ DN 11 Kevaddha DA 24
DA2 30 MandTsa 1 @ DN 7 Jaliya No
DA2 31 MandTsa 2 @ DN 7 Jaliya No
DA2 32 Mahallin @ DN 6 Mahali No
DA2 33 Sronatandya @ DN 4 Sonadanda DA 22
DA2 34 Kutadandya @ DN 5 Kutadanta DA 23

2nd varga

DA2 35 Ambastha @ DN 3 Ambattha DA 20

DA2 36 Prsthapala @ DN 9 Potthapada DA 28
DA2 37 Karanavadin @ No No
DA2 38 Pudgala @ (cf. AN II 205 ff.) No
DA2 39 Sruta @ No No
DA2 40 Mahalla @ No No
DA2 41 Anyatama @ No No

3 rd varga

DA2 42 Suka @ DN 10 Subha No

DA2 43 JTvaka @ MN 55 JTvaka No
DA2 44 Raja @ DN 2 Samannaphala DA 27
DA2 45 Vasistha @ DN 13 Tevijja DA 26
DA2 46 Kasyapa @ DN 8 KassapasThanada DA 25
DA2 47 Brahmajala @ DN 1 Brahmajala DA 21

60
Here, several of the discourses found in the Sanskrit have no known cognate. A meaningful
discussion of this point must wait until the texts are published. But we can see that at least a
couple of the unshared discourses seem anomalous: we have already commented on the JTvaka
Sutta; this seems to have been cut adrift f r o m its Majjhima friends. Especially startling is t h e
inclusion of Anguttara-style material. Perhaps the redactors consciously used a 'principle of
diversity' in assembling the texts: they wanted to include a range of different materials in the one
collection, so deliberately inserted heterogenous material. This principle would seem to have
been at work elswehere, too. For example, the Vinaya, though focussing primarily on monastic
discipline, finds room to encompass a broad range of other styles, f r o m doctrinal teaching, to
historical narrative, verses, story-telling, and so on. This might have been done since the reciters
of any one collection had relatively little knowledge of the content of other collections. In t h e
Theravada, the compilers made sure, for example, that the DTgha reciters would be exposed to
key Anguttara-style teachings by incorporating such matter within larger discourses. The
Sarvastivada also did this, but it seems they also allowed room for material that had not been
adapted to fit in its context.

Leaving these anomalous texts aside, virtually all the discourses in this section are found in all
three collections. This is the most outstanding structural feature of the existing DTgha. In the rest
of the DTgha, most of the discourses are held in common among the traditions, especially between
the Theravada and Dharmaguptaka, but the slight structural resemblance suggests that the
sequence of discourses was settled after the Dharmaguptaka schism, which was quite late. But in
the STlakkhandhavagga, the discourses are found consistently grouped together, even though the
sequence of texts differs considerably. The few unshared discourses mostly repeat shared
discourses, so do not substantially affect the picture.

Even more striking, t h e discourses all deal with a similar topic, offering a detailed account of
monastic ethical training, justifying the chapter title 'STlakkhandhavagga' ('Chapter on the
Aggregate of Ethics'). This treatise on ethics is usually complemented by sections on the four
jhanas and then the higher knowledges culminating in t h e realization of the four noble truths,
thus completing t h e threefold gradual training (sikkha). The position of the vagga is different - in
the Theravada it is at the beginning, in the Dharmaguptaka and Sarvastivada at t h e end. In fact
the Theravada here seems correct, for the Vinaya of the Dharmaguptaka states that the
Brahmajala Sutta was the first discourse recited in the First Council; this text is in the
STlakkhandhavagga of both versions, and in t h e Theravada version it is t h e first discourse in the
vagga. The Dharmaguptaka DTrgha, therefore, must have been re-organized after their Vinaya
was completed. This re-organization displaces t h e monastic and meditative discourses f r o m the
beginning of the collection in favour of the biographical and hagiographic; this might even have
occurred under Mahayana influence at the time of translation. Despite this difference in the
position of the vagga, and despite the fact that the internal sequence within the vagga is
different, still it is evident that this vagga was a key structural element in the DTgha current in t h e
ancestral Theravada school before it split into Sarvastivada and Vibhajjavada. This is f u r t h e r
confirmed in the Mulasarvastivada Vinaya, which mentions that the DTgha contains the
Satsutrakanipata and t h e STlaskandhanipata, although it does not list t h e individual discourses. 120

In fact, it is possible that the original DTgha w a s j u s t the STlakkhandhavagga. This is supported by
a statement in the introduction to the Dharmaguptaka DTrgha Agama, which describes the
collection as dealing with 'various ways of practice', which is exactly the main thrust of t h e
STlakkhandhavagga. This section has a distinctly Samyutta-like character - a group of ten
discourses dealing with the same topic. Thus the STlakkhandhavagga may well have existed at
first in t h e proto-Samyutta, where it could have comprised its own '*Silakkhandha-samyutta'.
Perhaps more likely it was the largest chapter in a 'Sikkha-samyutta', which was later broken up
because of its excessive length. The Sarvastivada retains a fairly humble Sikkha-samyutta

61
consisting mainly of the short discourses on the three-fold training, but in the Theravada this has
disappeared. The shorter discourses on t h e threefold training were moved into the Anguttara, the
medium ones into the Majjhima, and the longest ones formed a group on its own, which attracted
other long discourses and became the DTgha. These long discourses are all in dialogue form and
are therefore classic vyakaranas in style, chronicling the vibrant debates between the Buddha and
his spiritual contemporaries.

But by the time the later discourses were being added to t h e DTgha, the three-anga classification
was breaking down completely. A good example of this is the Mahapadana Sutta. 121 The doctrinal
core of this is a Samyutta-style vyakarana on dependent origination (having much in common
with SN 12.4-9/SA 366). This is closely connected with one of t h e basic sutta passages on the rise
and fall of the five aggregates taken from the Khandha-samyutta (this passage is connected with
dependent origination in the Theravadin text SN 12.22 but not in its counterpart SA 348). This is
embedded within an apadana, a form found in the Sanskrit twelvefold anga but not in the Pali
ninefold anga. The discourse begins with an abbhutadhamma. It includes the Request of Brahma,
which is a paradigmatic geyya. Vinaya material and some udanas (or gathas) from the
Dhammapada are thrown in for good measure. Thus we find, in one existing discourse, no less
than seven distinct literary formats.

T H E ANGUTTARA

So if the Majjhima and DTgha can be regarded as outgrowths of the vyakarana anga, what of t h e
Anguttara? The Samyutta and the Anguttara appear to be complementary collections of the
shorter discourses. While the discourses in the Samyutta are collected according to topic (the
'samyutta principle'), t h e discourses in the Anguttara are arranged according to numerical
sequence (the 'anguttara principle'). In addition to this main application within the Sutta Pitaka,
this pair of organizing principles is echoed in the Vinaya and Abhidhamma. For example the
Dharmaguptaka Vinaya has two appendices, one called the 'Samyuktavarga', and one called t h e
'Vinayaikottara'. 122 We will briefly consider the Abhidhamma below. The fact that the two
principles occur across the schools and also across the Pitakas suggests that they are both quite
fundamental. Given this, is it then possible to decide which organizing principle came first?

As usual, we may first consider the pre-Buddhist texts. From the Rg Veda onwards in much of the
Brahmanical literature t h e verses of homage to a certain deity - say, Soma, or Agni, or the Maruts
- are collected together in chapters. This clearly presages the samyutta principle; indeed, some of
the chapters in the Sagatha Vagga, such as t h e Sakka-samyutta or t h e Devaputta-samyutta,
directly recall Vedic antecedents. Some of the existing Jain texts use the anguttara principle, but
it is not clear if this usage pre-dated Buddhism. So the major pre-Buddhist literature consists
largely of short pieces of text that were gathered together at a later date and sorted out by topic
into a massive architecture. The Samyutta could almost be seen as a direct literary challenge to
this Vedic supremacy, taking the same formal elements and applying t h e m far more
systematically. Just as the Vedas were regarded as an emanation of Being into sound, an
expression of the innate cosmic intelligence (veda) as a body of inspired poetry (Vedas), the
Samyutta by embodying the four noble truths exhibits the perfect correspondence between
Dhamma as lived experience and Dhamma as formalized teaching.

From the pre-Buddhist tradition, let us then consider the original Buddhist traditions. These were
of course the individual discourses, the teachings given by the Buddha himself. Most of these
discourses consist of short statements on a specific topic, where t h e relevant aspects of the topic
can be summed up in a small number of fundamental items; for example the threefold training,
the five powers, or the six recollections. So the individual discourses, the building blocks, are

62
internally organized by topic, that is, the samyutta principle. This suggests that the idea of t h e
samyutta principle is logically prior to the anguttara principle. But while in the majority of cases
it seems clear that thematic affinity was the magnetic force that drew these dhammas together,
the manner of presenting t h e m in lists of distinct items instantly gives them a numerical
standing. In fact, a standard style of openings for discourses is, say, 'There are these four noble
truths...' This would invite classification under either 'four' or 'noble truths'.

There are some discourses that use number as an internal organizing principle. For example the
Dhammacakkappavattana Sutta features the interlocking set of numbers: 2, 3, 4, 8, 12. Such
discourses invoke the mysterious properties of numbers, which may be divided or multiplied, as
an image seen through prisms that reveal many facets or few. But in such cases t h e numerical
relationship, though significant, is clearly subordinate to the thematic relationship. Only
occasionally do we see a discourse where the numerical principle links teachings that have no
strong thematic relationship. On a small scale there are discourses such as the Maha Panha
Sutta, 123 which gives a list of questions on miscellaneous sets of dhammas from one to ten, and on
a larger scale there are of course the SangTti and similar discourses. 124 But even here, while there
is no strong thematic relationship between the different sets of dhammas, each individual set of
dhammas is still internally organized by topic. Wherever we peel back the skin, the samyutta
principle seems to lie antecedent to the anguttara principle. While this does not prove that the
Samyutta Nikaya is earlier than the Anguttara Nikaya, it remains suggestive.

Perhaps the most influential r e m n a n t of the anguttara principle in later Buddhism was the 37
wings to enlightenment. This group well illustrates the interplay between the two organizing
principles. It comprises seven sets of dhammas dealing with the path, which originally comprised
the Maha Vagga (or 'Magga Vagga' according to the Sarvastivadins) of the Samyutta Nikaya. Thus
the most general overall structure is t h e samyutta principle (dhammas dealing with one theme,
the way of practice), and so they appropriately form part of the Samyutta Nikaya/Agama. Within
this overarching grouping, the standard sequence lists the groups in ascending numerical order,
that is the anguttara principle - four (satipatthanas, right efforts, bases of psychic power), five
(faculties and powers), seven (enlightenment-factors), eight (factors of the noble path). There is
explicit evidence that it was in fact the anguttara principle that is at work here, not some
abstruse progressive structure underlying the sets. This is the second Maha Panha Sutta, which
gives the following dhammas in ascending sequence: (one) all beings subsist on sustenance; (two)
n a m e & form; three feelings; four satipatthanas; five spiritual faculties; six elements of escape;
seven enlightenment-factors; eightfold noble path; nine abodes of beings; ten courses of skilful
action. 129 This fits as many of the wings to enlightenment as possible in this numerical scheme.
Some of the schools, forgetting the arbitrary nature of this sequence, tried to interpret it as
implying an orderly progress of practice; that is, they interpreted a collection of teachings
organized by the anguttara principle as having been organized by t h e samyutta principle.

Gethin points out that many of t h e matikas of both the Theravada and Sarvastivada
Abhidhammas are constructed with t h e anguttara principle. 126 The most important example is the
DhammasanganT, which is based on a matika of 22 dyads and 100 triads of dhammas. Many of
these are shared with the early Suttas and the Sarvastivada Abhidhamma; hence, many of the
dyads and triads must be old, although the elaborate working out of t h e m is not. Gethin sees t h e
key Abhidhamma works as springing f r o m the interplay of such anguttara-matikas and the
samyutta-matika: the DhammasanganT is based on an anguttara-matika, and analyses this with
the samyutta-matika; t h e Vibhanga is based on the samyutta-matika and analyses this with an
anguttara-matika. 1 2 7 However, while correctly stressing the importance of the anguttara
principle, he does not really present any persuasive evidence that the anguttara principle was of
such primary importance as the samyutta principle. No doubt the Buddha did teach in numbered
sets, and no doubt these began to be collected from an early date. But the story told by the early

63
texts establishes the Samyutta as the most fundamental collection, in terms of both time and
doctrine. Furthermore, samyutta-principle texts tend to be relatively consistent not merely in
their organizing principle but also in the actual contents (the samyutta-matika), whereas there is
no standard content of anguttara-principle texts.

Bucknell discusses the structural features of the Anguttara in some depth. Here is a sample of his
findings.

'The first is titled Vaggo pathamo (First vagga) - or, in one manuscript, Rupadi-vaggo pathamo
(First vagga, on Visible Form etc.).128 In this case the ten short "suttas" comprising the vagga
belong together as regards both content and form. In the first of t h e ten the Buddha says:
"Monks, I know of no other single visible form that so enslaves a man's mind as the visible form
of a woman ..."; in the next four he says the same of the sound, odor, taste, and feel (tactile
experience) of a woman; and in the remaining five he repeats it all but with the words "man" and
"woman" interchanged. The wording is otherwise identical in all ten "suttas". What the text
identifies as a vagga, a collection of ten suttas, actually has the characteristics of a single sutta in
ten sections, which might have been appropriately located in the Fives, or perhaps in the Tens.
Now, there is a sutta in the Fives that incorporates verbatim the first five of these ten "suttas"; 129
it differs only in adding more detail, placing the lesson in a context, and supplying introductory
and closing formulas. This is, therefore, likely to be the source of the first vagga of the Ones. The
central portion of the source sutta was lifted out of its context, divided into five sections, and
then duplicated by switching "man" and "woman", to yield a set of ten pseudo-suttas, each of
which dealt w i t h j u s t one Dhamma topic.' 130

Bucknell notes that t h e ten Theravada discourses are represented by two texts in the
Mahasanghika Ekottara (EA 9.7, 9.8), thus supporting the hypothesis. Similar features dominate
the Ones of the Anguttara, and are found elsewhere in the collection, too. One possible reason for
such manipulation - which is found in both t h e Theravada and Mahasanghika versions - is simply
to provide more material for poorly-represented numbers. There are lots of sets of, say, five
dhammas, but few sets of one dhamma, and this may have prompted the slicing up of some texts
of fives into ones.

Even so, there still remains the question why certain of the fives are chosen and not others. Is this
merely arbitrary, or is there some other guiding principle at work? One obvious reason is that
many of these texts feature the word 'one'. But not all of t h e m do. I believe that part of the
answer also lies in the symbolic resonance of t h e numbers. The power and mystery of numbers
exerted a fascination on many of the philosophers of the ancient world; they had not become so
over-familiar with figures as to see t h e m merely as devices for mechanical manipulation and
'number-crunching'. Number promised the key to unlock the mysteries of the stars. Particular
numbers clearly have a symbolic significance in the Buddhist tradition. 131 It would be surprising if
such numerological significance had no influence on a collection organized by number. For
example, the number 'one' in Buddhism often denotes samadhi or 'one-pointedness of mind'. The
first fifty-five suttas of the Theravada Anguttara deal with samadhi and its hindrances, and
prominently feature the word 'mind' (citta), a word which frequently occurs in samadhi contexts
as 'one-pointedness of mind'. The first forty texts feature the word 'one', but the remainder do
not. These texts are all demonstrably artificial: most of t h e m were constructed by slicing one
longer text into fragments. It is possible that this process was encouraged by the feeling that the
number 'one' was particularly appropriate for a samadhi context.

We have already seen how some of t h e important doctrinal matter in the Anguttara seems to
have been moved f r o m the Samyutta. And indeed we often find, within the large-scale disorder of
the Anguttara, smaller groups of discourses collected together according to topic, such as the

64
threefold training, the various groups of powers, the six recollections, the tenfold path, etc.; these
form mini-samyuttas within the Anguttara, which in some cases are matched by genuine
samyuttas in the Sarvastivada Samyutta. Sometimes these form a classic vagga, such as AN 3.82-
93132 on the threefold training, reminiscent of the ten (or so) discourses on the gradual training
found in the DTghas. Most of these are found in the Sarvastivada Sikkha-samyutta, concentrated
within one and a half vaggas. But the material generally considered most characteristic of the
Anguttara is simple, lay-oriented teachings on ethics and devotion, and in many of the discourses
the prose is supplemented by a verse summary. This reminds us of the geyya anga. In addition,
there is a substantial number of vyakaranas, dealing with familiar topics such as the sense media,
dependent origination, etc. About 78 discourses, mainly vyakaranas, are found in the Theravada
Anguttara but the Sarvastivada Madhyama. Some of these appear like 'Majjhima-style' discourses,
and hence may have been moved from the Majjhima into the Anguttara. In other cases there are
substantial groups of consecutive discourses in the Sarvastivada Madhyama that share the same
number; these may have been moved from the Anguttara.

I would suggest as an initial hypothesis that the Theravada Anguttara might have started life as a
much smaller collection derived from the geyya anga. It would have included shorter discourses
dealing with relatively minor topics that were not included in the Samyutta. Probably its main
purpose was to provide convenient material for sermons, especially for lay devotees; this
function is acknowledged by the schools. At a later date it was filled out greatly with material
from the Majjhima and Samyutta. Most of this material probably belonged originally to the other
collections and was moved over to the Anguttara in order to give it more doctrinal weight, ensure
that the Anguttara students got a complete education, and balance out the Nikayas into four
reasonably similar-sized collections for the purposes of memorization.

It is difficult to know how far these speculations, tentative enough in the case of the Theravada,
might also apply to the Anguttaras of other schools. The Sarvastivada versions of all three other
Agamas are all long, in the case of the Madhyama and DTrgha in particular, much longer than the
Pali, so they might have had a smaller Ekottara. But the existing Ekottara is quite sizeable, and
includes much substantive material; perhaps it was chosen by the Chinese for translation
precisely because it was one of the more substantial versions of that collection.

CHAPTER 6 : THE EVOLUTION OF THE TRUTHS

In this chapter we will look more closely at how the teachings evolve in the different strata and
types of text we have been discussing. We will take as the main paradigm the four noble truths.
As well as having central position in the Dhamma, this framework offers a particularly clear
model for the kinds of changes we are interested in here. In particular, the texts themselves
suggest an evolution in the presentation of the doctrine.

T H E SANSKRIT DHAMMACAKKAPPAVATTANA S U T T A

The Theravada version of the Dhammacakkappavattana Sutta, which is by far the best known,
presents the teaching material in the following manner. First come the two extremes and the

65
eightfold path; then come the definitions of the truths; finally comes the description of the 'three
rounds and the twelve modes'.

This presentation is not common to all the versions. Most importantly for our current concerns,
several omit the definitions of t h e truths; we have mentioned that this is t h e case with the SA
Dhammacakkappavattana Sutta. One version even omits the truths entirely, p r e s e n t i n g j u s t the
extremes and the middle way. The Sarvastivada Catusparisat Sutra offers an interesting
perspective on how these variations may have come about. Here I offer a condensation of this
important narrative. This was published as part of Kloppenborg's translation of the Catusparisat
Sutra; but I have substantially revised Kloppenborg's translation to bring out the closeness of the
Sanskrit with the Pali. The teachings begin after an extended dialogue between the Buddha and
the group of five monks, similar to t h e account in the parallel passage in the Theravada Vinaya
Mahavagga. 133 The five monks had been criticizing the Buddha for backsliding, reverting to a life
of luxury, abandoning t h e hard task of asceticism. The Buddha responded:

11.14 'Monks, these two extremes should not be cultivated nor enjoyed nor attended by one who
has gone forth: devotion to indulgence in sensual pleasures, which is low, vulgar, ordinary,
practiced by ordinary persons; and devotion to self-mortification, which is painful, ignoble, and
pointless.

11.15 'Avoiding these two extremes is the middle way, which brings vision, brings knowledge,
and leads to clear knowledge, enlightenment, and Nibbana.

11.16 'What is this middle way? It is t h e noble eightfold path, that is: right view, right intention,
right speech, right action, right livelihood, right effort, right mindfulness, and right samadhi as
the eighth.'

11.17 The Lord succeeded in persuading the five monks by means of this teaching. In t h e
morning the Lord instructed two of the five monks, while three went to the village for alms. Six
of them nourished themselves on that which the three brought.

11.18 In the afternoon the Lord instructed three of the five monks, while two went to the village
for alms. The five of t h e m nourished themselves on that which the two brought. The Tathagata
only ate in t h e morning, at the proper time.

12.1 Then the Lord addressed the five monks:

12.2 ' "This is the noble truth of suffering." For me, monks, when I paid causewise attention to
these dhammas unheard of before, vision arose, and knowledge, realization, and awakening
(buddhi) arose.

12.3 ' "The is the noble t r u t h of the origin of suffering, the cessation of suffering, and the way
leading to t h e cessation of suffering." For me, monks, when I paid causewise attention to these
dhammas unheard of before, vision arose, and knowledge, realization, and awakening arose.

12.4 ' "The noble t r u t h of suffering should be fully known with clear knowledge
(abhjna)"...awakening arose.

12.5 ' "The origin of suffering...must be abandoned with clear k n o w l e d g e " .

12.6 ' "The cessation of suffering...must be witnessed..."...

66
12.7 ' "The way leading to the cessation of suffering... must be developed..."...

12.8 ' "The noble t r u t h of suffering has been fully known with clear knowledge
(abhjna)" . a w a k e n i n g arose.

12.9 ' "...the origin of suffering...has been abandoned with clear k n o w l e d g e " .

12.10 ' " . t h e cessation of suffering...has been witnessed..."...

12.11' "The noble t r u t h of t h e way leading to t h e cessation of suffering has been developed with
clear knowledge." For me, monks, when I paid causewise attention to these dhammas unheard of
before, vision arose, and knowledge, realization, and awakening arose.

12.12 'As long as, monks, regarding these four noble truths with the three revolutions and t h e
twelve modes, vision did not arise, nor did knowledge, nor did realization, nor did awakening
arise, even so long, in this world with its deities, its Maras, and its Brahmas, with its ascetics and
Brahmans, this generation with its princes and people I could not claim to be released, destined,
disconnected, liberated...with the unexcelled perfect enlightenment.

12.13 'But when, monks, regarding these four noble truths with t h e three revolutions and the
twelve modes, vision did arise, as did knowledge, as did realization, and as did awakening arise,
then, in this world with its deities, its Maras, and its Brahmas, with its ascetics and Brahmans,
this generation with its princes and people, I claimed to be released, destined, disconnected,
liberated...with the unexcelled perfect enlightenment.'

13.1 When this dhamma exposition was given, Venerable Kaundinya attained the stainless,
immaculate vision of the Dhamma regarding dhammas, together with 80 000 deities.

13.2 Then the Lord addressed the Venerable Kaundinya:

13.3 'Kaundinya, did you deeply understand the Dhamma?'

13.4 'I deeply understood, Lord.'

13.5 'Kaundinya, did you deeply understand the Dhamma?'

13.6 'I deeply understood, Sugata.'

13.7 The Dhamma was deeply understood by Venerable Kaundinya, therefore Venerable
Kaundinya was called "Ajnatakaundinya".'

13.8-12 [The various orders of deities, from the earth yakkhas to t h e Brahma gods, take up the
cry to announce the revolving of the Dhamma wheel with its three revolutions and twelve
modes.]

13.13 Thus this Dhamma-wheel of the Dhamma with its three revolutions and its twelve modes is
revolved by the Lord in the Deer Park at Isipatana. Therefore this exposition of the Dhamma is
called 'The Revolving of the Wheel of the Dhamma'.

14.1 Then the Lord said to the five monks.

14.2 'There are, monks, four noble truths. What four?

67
14.3 'The noble truth of suffering, the origin of suffering, the cessation of suffering, and the way
leading to t h e cessation of suffering.

14.4 'What is the noble truth of suffering?

14.5 'Birth is suffering, aging is suffering, sickness is suffering, death is suffering, separation f r o m
the liked is suffering, association with the disliked is suffering, seeking but not getting what one
wishes is suffering. In brief, the five aggregates associated with grasping are suffering. In order
to fully know this, the noble eightfold path must be developed.

14.6 'What is the noble truth of the origin of suffering?

14.7 'The craving which pertains to rebirth, associated with relishing and lust, which delights
here and there. In order to abandon this, the noble eightfold path must be developed.

14.8 'What is the noble truth of the cessation of suffering?

14.9 'It is the complete abandoning of that very craving which pertains to rebirth, associated
with relishing and lust, which delights here and there; t h e relinquishing, destruction,
evaporation, fading away, cessation, appeasement, and ending of it. In order to witness this, t h e
noble eightfold path must be developed.

14.10 'What is the noble truth of the way leading to t h e cessation of suffering?

14.11 'It is the noble eightfold path, that is: right view...right samadhi. This must be developed.'

14.12 When this Dhamma exposition was given, Ajnatakaundinya's mind was released f r o m
defilements without grasping, and for the rest of the five monks the stainless, immaculate vision
of the Dhamma regarding dhammas arose. At that time there was one arahant in the world; the
Lord was t h e second.

15.1 Then the Lord addressed the rest of the five monks:

15.2 'Monks, physical form is not s e l f . . '

15.3-18 [The Buddha teaches the Discourse on Not-self, almost identical to the Pali.]

15.19 When this Dhamma exposition was given, the minds of the rest of the four monks were
released f r o m defilements, without attachments. At that time there were five arahants in the
world; the Lord was t h e sixth.

There are too many interesting points in this narrative to mention them all; t h e reader is invited
to compare carefully with the Theravadin version. Obviously, the teaching sections are virtually
identical. The only noteworthy differences in the content is the omission of 'sorrow, lamentation,
pain, grief, and despair are suffering', a phrase that is standard in the Theravada, but omitted
elsewhere in the Sarvastivada, such as the Saccavibhanga Sutta. Also the phrases 'sickness is
suffering' and 'association with the disliked, separation f r o m the liked' are sometimes omitted in
the Theravada. None of these differences are sectarian. The most striking difference is in the
structure of the narrative. The Theravada combines the teachings into one session - the two
extremes, the definitions of the truths, the three revolutions and twelve modes - at the end of
which Venerable Kaundinya attains the Dhamma-vision. Then, over the next few days, t h e

68
Buddha gives f u r t h e r teachings. These are not specified, b u t first Venerables Vappa and
Bhaddiya, and then Venerables Mahanama and Assaji, attain stream-entry. Each, immediately on
seeing the Dhamma, requests ordination under the Buddha, and it is this that forms the thematic
link to the Vinaya, serving as introduction to t h e development of ordination procedure. Such
specifically Vinaya elements are absent f r o m the Catusparisat Sutra. Then, during the Discourse
on Not-self all five attain arahantship. So the presentation of events in the Sarvastivada makes
good sense of the Theravada account, too: first the basic teachings on the path and the truths
were given, followed by more detailed explanation. Later, it seems, the Theravada wished to
combine the teachings into one longer discourse.

Some elements in the Sarvastivada are obviously late, such as the insertion of the '80 000 deities';
but it must be admitted that the work in general hangs together extremely well as a narrative.
The two extremes and t h e middle way are presented more directly and explicitly as a response to
the critique of backsliding. Then the fundamental teachings on t h e three revolutions and the
twelve modes are given. These, here emphasized more than in the Theravada, become a recurring
theme. They are included in the proclamation of the gods, and again referred to in the later
definitions of the truths, in the phrases 'In order to fully know this... [etc.]'; neither of these
contexts are in the Theravada account. It appears that in this tradition, t h e very meaning of the
title of the discourse, 'The Revolving of the Wheel of the Dhamma', refers to these three
revolutions (see section 13.13 above). Only after the presentation of the three revolutions and
twelve modes is the detailed definition of t h e truths given. This is phrased in a rhetorical
question format, whereas in the Theravada these questions are absent. Thus in t h e Sarvastivada
version the initial statement of the basic doctrine in declarative form is followed, at a later time,
by the detailed exposition in question & answer form. This corresponds exactly with t h e
sutta/vyakarana model. Notice the structure here:

1. Statement (sutta): There is suffering...origin...cessation...path...

2. Question: What is suffering?
3. Explanation (vyakarana): Birth is suffering...the five aggregates associated with grasping are
suffering.

Now the Discourse on Not-self picks up f r o m here, by explaining how t h e five aggregates are
suffering. It is as if another question had been asked:

1. Statement (sutta): There is suffering...origin...cessation...path...

2. Question: What is suffering?
3. Explanation (vyakarana): Birth is suffering...the five aggregates associated with grasping are
suffering.
4. [Question: How are the five aggregates suffering?]
5. Explanation: Physical form is not self. If physical form were self, it would not lead to a f f l i c t i o n .

So the explanation on one level becomes the basic text for a deeper explanation. This explanation
then introduces a whole new field of doctrine, the five aggregates, demanding f u r t h e r
explanation; according to the GIST, this is the source of the Khandha-samyutta. Thus the concept
of vyakarana is a relative one, depending on what level of text one is explaining. This suggests
that the category of vyakarana will be flexible, and will evolve as the explanations become ever
more abstracted f r o m the original text.

69
SPIRALS OF ELABORATION

The same explanatory process is explicit elsewhere in the Suttas, too, most characteristically in
the teaching of Venerable Sariputta. Here is the beginning of the Maha Hatthipadopama Sutta:

'Friends, just as the footprint of any living being that walks can be placed within an elephant's
footprint...so too all skilful principles can be included in the four noble truths. What four? The
noble truth of suffering...origin...cessation...path.

'And what is the noble t r u t h of suffering? Birth is suffering...the five aggregates associated with
grasping are suffering.

'And what are the five aggregates associated with grasping? They are: t h e aggregate of physical
form associated with grasping, the aggregate of feeling...perception...activities...cognition
associated with grasping.

'And what is the aggregate of physical form associated with grasping? It is the four great physical
properties and the physical form derived from them.

'And what are t h e four great physical properties? They are: the property of
earth...water...fire...air.

'And what is the physical property of earth? The physical property of earth may be either
internal or external.

'And what is the internal physical property of earth? Whatever internally, belonging to oneself, is
solid, solidified, and grasped; that is, head hairs, body hairs, nails, teeth, s k i n . ' 1 3 4

Here the recurrent rounds of text/explanation are carried out to an elaborate degree, focussing
ever finer, from the truths to the aggregates to the physical properties, showing how all these
'skilful dhammas' are contained within the four noble truths. Given the pervasive ambiguity of
the t e r m 'dhamma', it is perhaps not inexcusable to read the term here as referring both to
'qualities, principles', meaning that all contemplative practice takes place in the broader
framework of understanding the four noble truths, and also 'teachings', meaning that all the
teachings can be classified within the four noble truths (=proto-Samyutta). For the early
Buddhists, this would have been, not an abstruse theory, but a reflection of how the teachings
embody at a profound level the structure of reality.

The Saccavibhanga Sutta exemplifies this elaborative process even more explicitly. 135 It is set in
the Deer Park at Benares, and the Buddha recalls his own teaching of the
Dhammacakkappavattana Sutta there. He praises Venerable Sariputta for his ability to teach the
four noble truths in detail, and then retires to his dwelling. Venerable Sariputta gets the hint, and
proceeds to analyse the truths. Whereas in the Maha Hatthipadopama Sutta he concentrated
minutely on t h e four physical properties, here he offers explanations for all aspects of the truths.
The teaching has the following structure:

1. Statement (sutta): The Buddha taught the four noble truths

2. Question: What four?
3. Explanation (vyakarana): Suffering...origin...cessation...path...
4. Question: What is suffering?
5. Explanation: Birth is s u f f e r i n g . t h e five aggregates associated with grasping are suffering.
6. Question: What is birth?

70
7. Explanation: The birth of beings into the various orders of b e i n g s .

In these central texts on the four noble truths the traditions directly correlate their historical
origin with the stage of elaboration. First is the statement of t h e truths; then the explanation of
the truths; then the explanation of the terms used in the explanation. Notice that this final level
of explanation is credited to a disciple, not the Buddha himself. And we should not be surprised to
find that these detailed explanations, unlike the basic teachings, show considerable divergence.
For example, the Sarvastivada version offers a detailed explanation of the phrase 'sickness is
suffering', which is not found in this context in the Pali (the Theravada tradition is quite
ambiguous as to the inclusion of 'sickness' in t h e noble truths definition). Much other material,
too, is expanded in the Sarvastivada version. In particular, there is one of the most explicit
statements found within the compass of the Agamas on the Sarvastivada doctrine of time: 'This
noble t r u t h of suffering has existed in the past, is existing in t h e present, and will exist in t h e
f u t u r e . ' It is easy to see how statements such as this, affirming that the Dhamma is a timeless
principle, could have slipped into a doctrine that 'dhammas' (phenomena) exist throughout the
three modes of time.

The evolution of the teachings on the four noble truths was not to stop here. The Maha
Satipatthana Sutta incorporates the doctrinal body of the Saccavibhanga Sutta, adding even more
material. We will examine this f u r t h e r in the treatment of the Satipatthana Sutta. The main
addition is a lengthy expansion of the second and third noble truths. The t r u t h of suffering is also
expanded in more subtle ways, with the addition of 'sickness' and 'association with the disliked,
separation f r o m the liked' to both the summary and detailed explanation. There is considerable
inconsistency in t h e way t h e traditions treat these phrases. Both of them are found in the
Sarvastivadin, but not t h e Theravadin, Saccavibhanga Sutta. Thus the Maha Satipatthana Sutta
presentation of the truths clearly suggests a f u r t h e r historical elaboration of doctrine, although
in this case the historical context is not confirmed in the text itself. In fact, this is clearly an
artificial text, so the absence of meaningful historical context is unsurprising.

The Theravada Maha Satipatthana Sutta presentation of t h e truths was taken up by t h e

Abhidhamma Vibhanga in its exposition of the truths. Here we witness f u r t h e r additions. Instead
of referring to one afflicted by 'some kind of misfortune', the Vibhanga specifies misfortune due
to failure in family, wealth, sickness, ethics, or views. Also, 'the cutting off of the life faculty'
(jivitindriyassupaccheda) was added to the definition of death. Generally, however, the Sutta
Exposition remains remarkably faithful to the Suttas. The following Abhidhamma exposition
employs the developed form of abhidhamma concepts to analyse the truths. Strangely, the first
and second truths are in reverse order. As usual, the exposition trails off in an increasingly
meaningless and pedantic series of questions. The only interest to the final 'questions' section,
standard in the Vibhanga, is that it once again reflects t h e statement/question form we have seen
evolving throughout the evolution of the texts. Here even the developed Abhidhamma Exposition
becomes the basic text, subject to f u r t h e r questioning. In fact, the Abhidhamma is sometimes said
to be characterised by this 'question method', which seems odd when we know how prevalent
questions are in the Suttas, too, but becomes explicable when we realize that the Abhidhamma is
largely derived from the vyakarana anga.

The evolution of the material analysing the truths in detail was: Saccavibhanga Sutta > Maha
Satipatthana Sutta > Vibhanga. At each stage more material was added. It seems that some of the
material added in the final Vibhanga version then found its way back into the Burmese (VRI)
Maha Satipatthana Sutta. This includes possibly 'association with the disliked is suffering,
separation f r o m the liked is suffering', and certainly t h e addition of 'the cutting off of the life
faculty' to the definition of death. This material then filtered down to the Burmese Satipatthana
Sutta, and 'the cutting off of the life faculty' even made it back into the Saccavibhanga Sutta, thus

71
'devolving' in this way: Vibhanga > Maha Satipatthana Sutta > Satipatthana Sutta/Saccavibhanga
Sutta.

So we can confirm that, in the case of at least one important doctrine, the treatment of the
teachings in the different collections is exactly as predicted by the GIST. The fundamental
teachings are found in the Samyutta and the Vinaya. The traditions tell us that the simpler
presentations of the teaching occurred first, and the more complex presentations later. The basic
statement is in declarative form, spoken by the Buddha. The more complex elaborations are in
question & answer format, and become attributed to disciples. These evolve from the Samyutta to
the Majjhima to the DTgha to the Abhidhamma.

CHAPTER 7 : W H A T HAPPENED IN THE SATTAPANI CAVE?

Before he passed away, the Buddha exhorted the monks:

'Therefore, Cunda, all you to whom I have taught these dhammas, having witnessed them with
my own clear knowledge, should come together and recite them, setting meaning beside meaning
and expression beside expression, without dissension, in order that this holy life may continue to
be established for a long time, for the benefit and happiness of the many, out of compassion for
the world, for the benefit and happiness of gods and humans. And what are the dhammas that
you should recite together? The four satipatthanas, the four right efforts, the four bases of
psychic power, the five spiritual faculties, the five spiritual powers, the seven enlightenment-
factors, the noble eightfold path.' 136

This is the Theravada version. In the Dharmaguptaka, the 37 wings to enlightenment are replaced
by the twelve angas.137 There is evidently a problem here. The 37 wings to enlightenment and the
twelve angas are, on the face of it, very different things: the wings to enlightenment are doctrinal
topics, while the angas are literary styles. But if the wings to enlightenment are, as I suggest,
primarily a table of contents of the meditation section (Magga Vagga) of the proto-Samyutta, and
if the three angas may be broadly identified with the proto-Samyutta as a whole, then the
problem dissolves.

In the Maha Parinibbana Sutta a similar exhortation occurs. The Buddha, after relinquishing his
will to live on, assembles the monks together in the Gabled Hall in the Great Forest in VesalT. In
the Pali, the Buddha encourages the monks to learn and practice the 37 wings to enlightenment,
in terms similar to above, but not specifically mentioning reciting them together. 138 The
Sarvastivada version in Sanskrit version has a similar passage, but the setting is the Capala
Shrine. Just as the Pali, this mentions the 37 wings to enlightenment, but adds that these
dhammas should be 'borne in mind, well understood, and recited'. 139

72
Another Sanskrit version, of which we only possess this fragment, describes t h e same episode as
taking place at the Gandhamadana Shrine. This mentions the angas as well as the wings to
enlightenment. It is a little curious that the wings to enlightenment, which are teachings on
practice, are to be recited, while the angas, which are literary texts, are to be practiced; this
dissolves any division between theory and practice, and tends to f u r t h e r imply the integration of
these two groups. We note in passing that some of the phrases in the following passage ('with
mutual rejoicing, without disputing, in unity, with unified r e c i t a l . ' etc.) are reminiscent of
Vinaya material.

'Therefore, monks, those dhammas that I have declared, having witnessed and entered upon with
my own direct knowledge - that is: t h e four satipatthanas, the four right efforts, the four bases of
psychic power, the five spiritual faculties, the five spiritual powers, the seven enlightenment-
factors, the noble eightfold path - therein you should all, in togetherness and harmony, with
mutual rejoicing, without disputing, in unity, with unified recital, one like milk and
water...(?)...should dwell in comfort.

'Therefore, monks, those dhammas that were taught by me - that is, sutta, geyya, vyakarana, gatha-
udana-nidana-avadana-itivuttaka-jataka-vepulla-abbhutadhamma-upadesa - those dhammas should
be well and thoroughly learnt; having been learnt they should be borne in mind; having been
borne in mind they should be investigated; having been investigated they should be understood;
having been understood, i n j u s t that way they should be practiced.' 140

A little after this episode, in both the Pali and Sanskrit versions, comes the famous teaching of the
great references, which we have met before: whenever anyone makes a statement about t h e
Buddha's teachings, then, no matter how learned or prestigious they may be, their statement
must be compared with the Suttas and Vinaya, 141 and only if it agrees with them may it be
accepted as the word of the Buddha. The Sanskrit brings out the essential principle more
explicitly than the Pali: 'The monks must rely on the Suttas, not on individuals'. 142 In the narrative
flow, this clearly harks back to the earlier statements, and implies that the 'suttas' here are
related to the wings to enlightenment a n d / o r t h e angas; i.e., the proto-Samyutta.

The Sanskrit version records an additional, similar statement. This is given great prominence by
being included in the famous deathbed teachings of the Buddha. In terms identical to t h e
previous context, t h e Buddha says that those skilful dhammas are to be learnt, remembered, and
recited, but instead of mentioning the wings to enlightenment it mentions t h e twelve angas. 143 In
both of these cases, the same dhammas are mentioned in the same contexts in the account found
in the Mulasarvastivada Vinaya (although I cannot be sure f r o m Rockhill's paraphrase whether
this text specifically mentions chanting). 144

These passages support our thesis, regardless of whether they represent authentic sayings of the
Buddha. If they are not authentic, they must have been invented by the Sangha, presumably to
authorize after the fact the recitation of the First Council at Rajagaha. They could not be very late
interpolations, for then they would surely mention the recitation of the Nikayas, as does t h e
account in the Vinaya Culavagga. The fact that they refer, seemingly without distinction, to the
37 wings to enlightenment and to the angas, suggest that if they were an interpolation they hark
back to a time when these were seen as constituting the key teachings recited at the First Council.

And if they are authentic, it seems incredible that the Sangha should have ignored or disregarded
such an important instruction. There seems no good reason to doubt that the Buddha did, shortly
before he passed away, encourage the Sangha to preserve his central teachings by coming
together to recite them. And I believe that they did exactly as the Buddha encouraged. After he
passed away, the Sangha came together in the Sattapani Cave, in the craggy hills overlooking

73
Rajagaha, and recited, out of compassion for the world, the core teachings of the Dhamma: the
suttas, geyyas, and vyakaranas that we find today principally in the Samyutta Nikaya.

These inferences f r o m the Maha Parinibbana Sutta dovetail rather nicely with the Theravada
exegetical tradition of the Petakopadesa and the Netti. The passage on the great references says a
statement should checked to see if it 'fits in' with the Suttas. The word we render as 'fit in' is t h e
Pali otarana, literally 'descending, entering'. L.S. Cousins comments:

'This is an unusual expression; it is best interpreted in the light of the Petakopadesa tradition
where otarana is one of the sixteen haras ['modes of conveying an interpretation']. It may be taken
as a particular mode of exegesis which links a given discourse into the teaching as a whole by
means of one of the general categories of the teaching. The Petakopadesa in fact specifies six
possibilities: aggregates, elements, spheres, faculties, truths, dependant origination. Any of these
can be used to analyse the content of a discourse and their use will automatically place it in the
context of the teaching as a whole. Something on these lines, if perhaps a little less defined, is
surely intended in the mahapadesa ['great reference'] passages. What is envisaged for sutta is not
then a set body of literature, but rather a traditional pattern of teaching.' 149

In the list of the six topics under otarana in t h e Petakopadesa we have, of course, yet another
example of the samyutta-matika. We have already noticed how the Netti treats sutta in the great
references as pertaining to the four noble truths, which points us straight to the
Dhammacakkappavattana Sutta and the major collection that it is found in. This dissolves the
dichotomy set up by Cousins in his last sentence: sutta is a body of literature that has been
patterned after the central teachings.

There is a suggestive passage in the Samantapasadika, the Theravada Vinaya commentary, that is
perhaps inadvertently revealing of the historical picture here. The passage deals with the
question of how t h e 500 monks were chosen for participation in the First Council.

'Leaving aside many hundreds and thousands of monks who are had memorized the entire nine
anga textual dispensation of the Teacher, the ordinary persons, stream-enterers, once-returners,
non-returners, and dry-vipassana arahants, the Elder [Maha Kassapa] gathered 499 monks who
had memorized the Tipitaka with all its textual divisions, attained to the discriminations, of great
might, mostly those included in the foremost disciples, gainers of the three realizations, etc., all
being arahants.' 1 4 6

Obviously the first group are being unfavourably compared with the latter; thus it is implied that
the nine anga scripture is somehow inferior to the Tipitaka. This is implicit in the traditions
anyway, since they moved f r o m the angas towards the Tipitaka, there must have been some
dissatisfaction felt with the old system. This passage suggests that the First Council was t h e pivot
point for this change, t h e time when the Tipitaka system started to come into its ascendancy.
Although we cannot accept the suggestion that the Abhidhamma was part of the recitation -
which is not even supported by the Theravada Vinaya account of the Council - we can agree with
the suggested dynamic. Again this harmonises quite nicely with the account of t h e First Council
in the Mulasarvastivada Vinaya, for there the business of compiling the Samyutta Agama (the
GIST's first three angas) was given pride of place, and led on to the compilation of the other
Agamas.

We are now poised to draw together some of the strands in the above chapters, and to paint a
more coherent overall picture of t h e structure of the Dhamma & Vinaya. The two discourses, the
Catusparisat Sutra and the Maha Parinibbana Sutta, are a complementary pair. This is quite
evident f r o m many parallels and similarities in their details and structure. It is unnecessary to

74
 examine this in detail here; suffice to exemplify a few features. Both start out in or near Rajagaha;
both involve the King of Rajagaha; both tell of the Buddha journeying; both intersperse the
journey with important teachings; the Catusparisat tells of the first convert (Anna Kondanna),
while the Maha Parinibbana tells of the final convert (Subhadda); both mention the Buddha's
superiority to the sage Alara Kalama; both feature divine intervention from Sakka, Brahma, and
Mara, as well as other deities; both feature displays of psychic powers, including the 'walking on
water'; both speak of the earthquakes and other celestial portents accompanying the chief events
in the career of t h e Tathagata; both integrate a large number of materials that are found
elsewhere as individual Suttas; both, however, fail to fully integrate all the relevant material
found elsewhere; both occupy a position intermediate to Dhamma & Vinaya; and so on. While
these features imply a connection between the texts, explicit connections, too, are not lacking.
The Catusparisat Sutta has the Buddha saying that he will not pass away until the four assemblies
(monks, nuns, lay men, and lay women) have been fully established (on which see more below).
The Maha Parinibbana Suttas explicitly refer to the rolling forth of the wheel of Dhamma as one
of the eight causes of earthquakes; the Sanskrit mentions the 'twelve modes' and 'three rounds',
thus clarifying that it is in fact the Dhammacakkappavattana Sutta that is being referred to.147

The Catusparisat Sutra tells the story of the beginning of the Buddha's career, while the Maha
Parinibbana tells of the end. We have noticed above that several genres of Buddhist literature are
formally structured by taking a pre-existing teaching, and furnishing this with opening and
closing passages that provide a setting for the teachings. This kind of form is, of course,
absolutely characteristic of the normal format for discourses: first the setting is stated, in brief or
in detail; then the doctrinal teachings are given; finally t h e monks rejoice in what was said. Later
literature such as the Jatakas, for example, provide the central story with a setting (the 'Story of
the Present') in a similar way. The Patimokkha rules, likewise, are preceded by origin stories
(nidanas) and followed by case studies and analysis. What if we were to consider the Catusparisat
and the Maha Parinibbana as constituting, in a parallel fashion, the narrative opening and closing
settings for the whole of the Dhamma-Vinaya?

The Catusparisat furnishes a narrative background for the Dhammacakkappavattana Sutta and
the other essential early teachings that today are found in the Samyutta. In this way it sows t h e
seeds of the three angas (= proto-Samyutta), and forms the foundation for the entire edifice of the
Dhamma. We can almost see the teachings integrated in this story starting to branch off and
separate: the geyya verses, primarily exemplified by the Request of Brahma, are quite distinctive,
and tend to occur in groups; the centrality of the sutta teachings of the Buddha himself, such as
the Dhammacakkappavattana Sutta, is given special emphasis; and the vyakarana dialogues with
the disciples, such as the Anattalakkhana Sutta, expand on the fundamental teachings.

Following the establishing of the Dhamma, the ordination of Anna Kondanna forms the starting-
point for the Vinaya. This narrative, in the Theravada Vinaya, merges into the formal description
of ordination procedure and thus the beginning of the Khandhakas, which is t h e half of t h e
Vinaya that primarily deals with the prescriptive aspects of the monastic life, the duties and so on
that are to be performed. The other half of the Vinaya, the Bhikkhu and BhikkhunT Vibhanga,
which treats the Patimokkha rules and their analysis, is primarily concerned with t h e
proscriptive aspects of the monastic life, the various kinds of misconduct that are to be refrained
from. This section begins with the story of Sudinna, the monk who had sex with his former wife
to grant his parent's wishes for an heir. This episode consciously forms a negative counter-
narrative to the Catusparisat material. While t h e Catusparisat memorably features the deities
taking up the cry of rejoicing over the proclamation of the Dhammacakkappavattana Sutta, the
story of Sudinna has t h e deities, in exactly parallel fashion, taking up the cry of the corruption
that has now entered the Sangha. Thus the Catusparisat, having branched off along the

75
fundamental divide of Dhamma and Vinaya, goes on to imply the main sub-divisions within each
of these areas.

Once the course and structure of the Dhamma & Vinaya had been set up in this manner, each of
the suggested main threads proceeded to diversify and sub-divide further, resulting in t h e
fascinating yet frustrating mass of unity and diversity of t h e scriptures as we have them. The
Maha Parinibbana Sutta, as the closing narrative, attempts to pull the strands back together
again. It contains much Vinaya-style material - the various sets of Dhammas leading to non-
decline; the allowance to abolish the 'lesser and minor rules'; the imposition of t h e 'highest
punishment' on t h e recalcitrant Channa; the ordination of Subhadda; and so on. In this sense it
may be seen as a summary and reflection on some of the key principles of Vinaya. In the same
way, it reviews and emphasizes some of the key Dhamma teachings, notably the 37 wings to
enlightenment, the four noble truths, and the threefold training of ethics, samadhi, and
understanding; in other words, the key topics of the Samyutta.

We have noticed that contrasting sections of the teaching may be framed within mythic settings
that are consciously articulated to contrast with each other, such as the cries of the deities in
response to the Dhammacakkappavattana Sutta and to the downfall of Sudinna. In a similar
manner, the Request of Brahma that t h e Buddha teach, found in the Catusparisat Sutta, contrasts
with the Request of Mara that the Buddha pass away, found in t h e Maha Parinibbana Sutta. The
Sanskrit here makes the parallels between the Catusparisat and the Maha Parinibbana much more
explicit and complex - far too complex to treat adequately here. 148 The Maha Parinibbana Suttas
tell how Mara approaches the Buddha, bows with his head at the Buddha's feet (! in the Sanskrit
only), and reminds him that, while the Buddha was staying at Uruvela on the banks of the river
Neranjara soon after he was enlightened, Mara had come to him and requested that the Buddha
pass away. This episode is in fact found in the Sanskrit Catusparisat Sutra, in exactly parallel
terms, but is absent f r o m the Pali equivalent in the Vinaya Mahavagga. At that time, both
versions of the Maha Parinibbana Sutta go on to say, the Buddha had rejoined that he would not
pass away until the fourfold assembly of monks, nuns, lay men, and lay women followers were
well established and well practiced in t h e Dhamma, able to teach and maintain the Dhamma.
(This passage, incidentally, is one of many that show that the establishing of the BhikkhunT
Sangha was not imposed on a reluctant Buddha, as appears f r o m the Vinaya narrative, but was an
intrinsic part of his mission f r o m t h e beginning.) But now, says Mara, these conditions are
fulfilled: the fourfold assembly has indeed been well established, so it is time for the Buddha to
pass away. The Buddha tells Mara not to worry, that he will indeed pass away in three months
time. The Sanskrit adds the interesting remark that, after receiving confirmation of the Buddha's
imminent Parinibbana, Mara, full of happiness a n d j o y , disappeared right there. 149 This stands in
contrast to Mara's reaction in every other encounter with the Buddha, where he vanishes 'sad
and disappointed', which is in fact what happens in the Request of Mara in the Catusparisat
Sutra. 150 These detailed correlations suggest that t h e two Sanskrit texts were edited conjointly,
and therefore probably belong to the same school. This is probably the Sarvastivada, although it
is noteworthy that in both texts, especially the Maha Parinirvana, whose main t h e m e is
impermanence, the Sanskrit refrains from any of t h e quasi-eternalist statements to which the
Sarvastivada is normally prone.

Another example of the extra connections in the Sanskrit is that, immediately before the Request
of Mara, the Buddha takes the alms-offering f r o m the merchants Tapussa and Bhallika, which was
the first meal after his enlightenment. Unfortunately, the food, consisting of many 'honey-
lumps', was perhaps too rich for the Buddha after his austere diet, for t h e Sanskrit, though not
the Pali, says he contracted a severe 'wind-ailment'. This immediately reminds us of the Buddha's
famous illness after eating his last meal; the relation is in fact explicitly invoked, for in the Maha
Parinibbana Suttas the Buddha says that these two meals, the first and last, are of unparalleled

76
merit. 151 In the Catusparisat Sutra, the Buddha's illness is followed by the Request of Mara; then,
after Mara has disappeared, Sakka the King of Gods appears before t h e Buddha to offer him
medicine to cure his illness. This narrative structure makes it seem as if the Request of Mara has
been inserted in the illness narrative.

The whole series of episodes offers a complex, resonant mythic fabric that is not easily unwoven
into its separate strands. It is the closing of t h e circle. Given this profound interdependence of
these two texts, it seems inevitable that their conception of the scriptural Dhamma should also be
interwoven. The Catusparisat Sutra provides an authoritative narrative framework for the
Dhammacakkappavattana Sutta and other central teachings; and in t h e same way, the Maha
Parinibbana Suttas provide an authoritative narrative framework for t h e Buddha's instruction
that, after he is gone, his followers should rely on the Dhamma & Vinaya. As we have seen, the
Dhamma is chiefly formulated here as either t h e 37 wings to enlightenment or the angas. We
have identified both of these with the proto-Samyutta, whose existing descendants contain the
essential teaching passages of the Catusparisat Sutra.

CONCLUSION

I have tried to make as strong a case as possible for the GIST in a brief space. Doubtless I have
omitted many possible counterexamples, and doubtless the real picture was more complex than
any brief description. For example, I have identified the Dhammacakkappavattana Sutta as the
prime paradigm of sutta anga in contrast with t h e vyakarana anga. Yet immediately following the
main discourse, the text says: 'When this vyakarana was being s p o k e n . ' (although the Sanskrit
has dhammapariyaya, 'exposition of Dhamma'.) To rub salt into the wound, the text ends with an
udana. Obviously, seeking for absolute consistency is hopeless. Still, if we reject, as we must, t h e
traditional accounts of the origin of the Tripitaka as sectarian myths, we are bound to seek for a
more plausible alternative.

The continuity, even identity, between the three identified strata of texts is undeniable. It reflects
an intense effort to stabilize the Buddha's teachings, to preserve t h e m against the inexorable
ravages of impermanence of which all Buddhists have been so keenly aware. Taking a leaf f r o m
Richard Dawkins' theory of memes, we can think of this textual stability in terms of the evolution
of ideas. The texts themselves are not primary. They are not ends in themselves. It is the ideas,
the memes, that are the driving force. The memes generate texts in order to ensure their own
survival and transmission through time. The surface structures of the texts are determined by
expedient and contingent matters such as local technology, literary styles, and so on, and must
change in time. But the memes survive. This is the only meaningful criterion by which t o j u d g e
the success of this extraordinary literary endeavour: has it preserved the essential ideas through
time? In the early strata we have been considering, this reproduction was maintained using t h e
very same words and phrases repeated in t h e various strata. In fact we could speak, not of t h e
creation of new strata of texts, but of different stages of evolution of one and the same text. The
only genetic peculiarity is that t h e earlier layers are preserved alongside the later. Our task must
be to peel back the layers, the delicate art of textual archaeology.

By doing so, though the task may seem laborious and dreary, we uncover a priceless treasure: a
common language, a common set of ideas that can be securely said to underlie all the schools and
traditions. Using our historical perspective, we are able to shed an exciting new light on a
forgotten world of Buddhism. Our situation is a little like that which came about in Chinese
Confucianism around t h e beginning of t h e Middle Ages. Confucius, who lived around the same
time as the Buddha, left only a small body of teachings behind, but by the start of the Common
Era these had been greatly developed in detail by the followers of his school. But in 93 C.E.,

77
Confucius' house was demolished, and a number of manuscripts were found hidden within the
walls. Embarrassingly, when these were deciphered they were found to teach doctrines decidedly
different from those proclaimed by the mainstream Confucianists. The repercussions of this
event have been felt in Confucianism down to the modern era.

For us, the lost manuscripts are not like the Confucian scrolls hidden in the wall, or like the Dead
Sea scrolls lost in the desert. The finds of ancient manuscripts from the deserts and caves of
Afghanistan and Central Asia date from well after the Buddha's time and mainly serve to
substantiate, rather than undermine, the authenticity of the existing canons. The lost
manuscripts are instead buried in an even deeper, more inaccessible place - the shrine rooms of
Buddhist temples. There they remain, buried beneath the sands of interpretation, objects of
worship not of study, inspiring devotion but not practice. The Buddha's urgent, repeated call was
for these teachings not to remain mere words, but to inform and nourish the liberation of the
heart.

PART 2

A HISTORY OF MINDFULNESS

CHAPTER 8 : INTRODUCTION TO MINDFULNESS

'Mindfulness is useful everywhere' - so said the Buddha. And in harmony with this motif, the
theme of mindfulness echoes throughout each of the melodies that compose the path to freedom.
At its most fundamental, mindfulness is essential for the sense of conscience on which ethical
conduct is founded; hence alcohol and drugs, by destroying mindfulness, destroy the basis for a
moral life. Mindfulness, in its older sense of 'memory', remembers and recollects the teachings,
forming the basis for the intellectual comprehension of the Dhamma, and bears them in mind,
ready to apply right at the crucial moment. Mindfulness guards the senses, endowing the
meditator with circumspection, dignity, and collectedness, not allowing the senses to play at will
with the tantalizing toys and baubles of the world. Mindfulness repeatedly re-collects awareness
into the present, re-membering oneself so that one's actions are purposeful and appropriate,
grounded in time and place. Mindfulness is prominent in all approaches to meditation, and in
refined form it distinguishes the exalted levels of higher consciousness called samadhi. On the
plane of wisdom, mindfulness extends the continuity of awareness from ordinary consciousness
to samadhi and beyond, staying with the mind in all of its permutations and transformations and
thus supplying the fuel for understanding impermanence and causality. And finally on the plane
of liberation, perfected mindfulness is an inalienable quality of the realized sage, who lives 'ever
mindful'.

Given this ubiquity of mindfulness, as omnipresent as salt in the ocean, it would seem a hopeless
task to isolate certain areas of the Dhamma as bearing a special affinity with mindfulness. Indeed,
we might even go further and allege that any such attempt conceals a program to co-opt the
unique prestige of mindfulness in the cause of one's own partisan perspective. Nevertheless, it
has become a commonplace in 20th Century Theravada meditation circles that mindfulness, and in

78
particular its chief manifestation as satipatthana, is close or identical in meaning with vipassana,
or insight. The chief support for this idea is the Theravada Satipatthana Sutta, which is the only
well-known early text on satipatthana. The success of this doctrine, repeated in virtually every
modern Theravada text on meditation, reflects the unrivalled prestige of the Satipatthana Sutta.
Here a r e j u s t a few representative quotes.

'[The Maha Satipatthana Sutta] is generally regarded as the most important sutta in the entire
Pali canon.'
Maurice Walshe, The Long Discourses of the Buddha, pg. 588

'The most important discourse ever given by the Buddha on mental development (meditation) is
called the Satipatthana Sutta.'
Walpola Rahula, What the Buddha Taught, pg. 69

'[The Satipatthana Sutta] is by all Buddhists rightly considered the most important part of the
whole Sutta-Pitaka and the quintessence of the whole meditation practice.'
Bhikkhu Nyanatiloka, Path to Deliverance, pg. 123

'No other discourse of the Buddha, not even his first one, the famous "Sermon of Benares", enjoys
in those Buddhist countries of the East which adhere to t h e unadulterated tradition of the
original teachings, such popularity and veneration as the Satipatthana Sutta.'
Bhikkhu Nyanaponika, The Heart of Buddhist Meditation, pg. 11.

In fact, t h e worship, as opposed to practice, of the Satipatthana Sutta is a remarkable and

undeniable feature of modern Theravada. Venerable Nyanaponika, in his classic The Heart of
Buddhist Meditation, waxes lyrical:

'In Lanka for instance, the isle of Ceylon, when on fullmoon days lay devotees observe eight of the
ten principal precepts of novice monks, staying for the day and the night in the monastery, they
frequently choose this Sutta to read, recite, listen to, and contemplate. Still, in many a home, the
satipatthana book is reverently wrapped in a clean cloth, and f r o m time to time, in t h e evening, it
is read to members of the family. Often this discourse is recited at the bedside of a dying
Buddhist, so that in t h e last hour of his life, his heart may be set on, consoled, and gladdened by
the Master's great message of liberation. Though ours is an age of print, it is still customary in
Ceylon to have new palm-leaf manuscripts of the Sutta written by scribes, and to offer t h e m to
the library of a monastery. A collection of nearly two hundred such manuscripts of t h e
Satipatthana Sutta, some with costly covers, was seen by the writer in an old monastery of
Ceylon.'152

The author discreetly avoids noticing that in this atmosphere of reverential awe the question of
practicing the instructions in the Satipatthana Sutta does not arise. Hundreds of copies of
manuscripts on meditation are accumulated in a monastery where probably no-one is actually
meditating. The irrationality of this is a classic symptom of religious fetishism - the Satipatthana
Sutta has been transformed into a magical totem. Please notice that this eulogy of the
Satipatthana Sutta as fetish appears at the beginning of the single most influential and widely
read book on contemporary Theravada vipassana meditation. It is explicitly invoked to magnify
the aura of sanctity surrounding the Satipatthana Sutta as a key aspect of the vipassanavada
agenda.

Where forcefulness of opinion is matched by paucity of evidence, I cannot but smell a dogma
lurking nearby. Much as we have benefited f r o m the modern emphasis on mindfulness in daily
practice, it is past time for the pendulum to swing back. The Buddha did not speak the

79
Satipatthana Sutta in its current form. It is a late and, in part, poorly organized compilation; and
it is specifically the vipassana aspects that are least authentic. In the early teachings satipatthana
was primarily associated not with vipassana but with samatha. Since for the Suttas, samatha and
vipassana cannot be divided, a few passages show how this samatha practice evolves into
vipassana. In later literature the vipassana element grew to predominate, to the extent of almost
entirely usurping the place of samatha in satipatthana. Subtle differences in emphasis between
the schools can be discerned in their treatment of satipatthana, differences that can be seen to
relate to the basic metaphysical controversies underlying the schisms. Thus the Satipatthana
Sutta is interesting not because it represents the 'unadulterated tradition of the original
teachings', but because it provides suggestive evidence for how sectarian adulterations crept
even into t h e early discourses.

In making such claims, claims that will inevitably be perceived as an attack on the authority of
some of the most respected 20th Century meditation schools, I cannot say emphatically enough
that what I am criticizing here is not the teachers of vipassana, or the meditation techniques that
are marketed as 'vipassana', but the textual sources of the vipassanavada, the doctrine that
vipassana is the central meditation taught by t h e Buddha.

The vipassanavada must be understood in its historical context, for it is this, rather than the
textual sources, that shape its essential features. The vipassanavada grew up as part of the
movement of 'modernist Buddhism', which started in the colonial era as the schools of Buddhism
attempted to respond to the challenges of the modern age. This movement swept over the whole
of the Buddhist world in a number of guises. In all its varieties, however, the key aspect of
modernist Buddhism was rationalism. Meditation, especially samatha, was suspect, since in
traditional Buddhist cultures it had often degenerated into a quasi-magical mysticism. Samatha is
emotional rather than intelligential. It cultivates the non-rational aspects of consciousness, and
so when it degrades it shades off into psychic tricks, fortune-telling, magic, and so on, all of which
are rampant throughout Buddhist cultures. Some forms of Buddhist modernism did away with
meditation altogether; this may be compared with the Protestant movement in Europe, which
similarly opposed the contemplative aspect of religion. Contemplation will always remain a
threat to religious orthodoxy, since there is always the uncomfortable possibility that the t r u t h a
meditator sees may not agree with the t r u t h that the books say they're supposed to see. However
in Buddhism, unlike Christianity, the contemplative life lies at the very heart of the Founder's
message. Other modernist Buddhism movements, perceiving that Buddhist meditation was based
on a rational psychology, developed contemplative systems that emphasized these aspects. These
schools, originating mainly in Burma, marginalized or outright disparaged samatha and
developed the vipassanavada as a scriptural authority for their 'vipassana-only' approach. The
strength of these schools is that they have rightly championed an energetic and disciplined
approach to meditation. But with our advancing knowledge and appreciation of the Buddhist
scriptural heritage, the scriptural authority for their special doctrines lies in tatters. Followers of
these contemplative schools would do well to be a little more humble in their claims, and to
emphasize the demonstrable practical benefits of their practices, rather than rely on a
discredited theory.

I am well aware that my claims fly in t h e face of virtually every modern interpreter of
satipatthana. Such an accumulated weight of authority cannot be discarded frivolously. At the
risk of appearing pedantic and perhaps obsessive, I must proceed very carefully. I will therefore
attempt to make my coverage as comprehensive as reasonably possible, casting an eye at every
available important early text on satipatthana, as well as a range of later passages. I consciously
flirt with the danger of polemicism, of simply asserting one extreme in reaction to an original
extreme. But everyone, no matter how 'objective' or 'scientific', has their own agenda, and it is
more honest to be open with one's perspectives than to pretend - to others or to oneself - that

80
one has no bias. The concern here is not so much for balance within this particular work, but for
balance within the tradition as a whole.

SAMATHA & VIPASSANA

The key to the approach used in this work is to analyse the various strata of texts on satipatthana
in terms of samatha and vipassana. It is therefore necessary to start by explaining what I mean by
these. We can distinguish two key aspects of how the Suttas speak of samatha & vipassana: their
nature, and their function. Their specific nature is clearly distinguished in this passage.

'A person who has samatha of the heart within himself but no vipassana into principles
pertaining to higher understanding should approach one who has vipassana and inquire: "How
should activities be seen? How should they be explored? How should they be discerned with
vipassana?" And later he can gain vipassana...

'A person who has vipassana into principles pertaining to higher understanding but no samatha
of the heart within himself should approach one who has samatha and inquire: "How should the
mind be steadied? How should it be settled? How should it be unified? How should it be
concentrated in samadhi?" And later he can gain s a m a t h a .

'One who has neither should inquire about both [and "should put forth extreme enthusiasm,
effort, endeavor, exertion, unflagging mindfulness, and clear comprehension to acquire them,
just as if one's turban or hair were ablaze, one would put forth extreme effort to quench t h e
flames". 1 5 3 ]

'One who has both, established in these beneficial qualities should make f u r t h e r effort for t h e
evaporation of defilements.' 154

'Just as if, Nandaka, there was a four-legged animal with one leg stunted and short, it would thus
be unfulfilled in that factor; so too, a monk who is faithful and virtuous but does not gain samatha
of t h e heart within himself is unfulfilled in that factor. That factor should be fulfilled by him... A
monk who has these three but no vipassana into principles pertaining to higher understanding is
unfulfilled in that factor. That factor should be fulfilled by him.' 155

The description of vipassana mentions the seeing, exploring and discerning of activities
(sankhara). The mention of 'activities' here implies the three characteristics - impermanence,
suffering, not-self - of phenomena, conditioned according to dependent origination. The
meditative discernment of the nature of conditioned reality is the central meaning of vipassana.
While this definition is possibly too narrow for some contexts, still vipassana is commonly used in
this sense in the Suttas and in the present day.

Samatha is described in terms of the steadying, settling, and unifying of the mind in samadhi.
Elsewhere the implications of this are spelt out.

'How does h e steady his mind within himself, settle it, unify it, and concentrate it in samadhi?
Here, Ananda, he enters and abides in the first j h a n a . second j h a n a . third j h a n a . fourth
jhana.' 156

Here, as in virtually all central doctrinal contexts in the early texts, samatha or samadhi is
explicitly defined as the four jhanas. We must therefore conclude t h a t the four jhanas are an
essential, intrinsic part of the path. Establishing these points formed the burden of the argument

81
of A Swift Pair of Messengers, so I won't repeat the reasons here. It is necessary to mention these
conclusions, however, for anyone who persists in the very common practice of interpreting early
texts on samadhi in terms of the commentarial ideas of 'access samadhi' and 'momentary
samadhi' will certainly misinterpret the present work, and, I believe, will also misinterpret the
Suttas.

The second mode of treating samatha and vipassana is in terms of their function, that is, the
results of the practice.

'Monks, these two principles share in realization. What two? Samatha and vipassana.

'When samatha is developed, what purpose is achieved? The mind is developed. When the mind is
developed, what purpose is achieved? Lust is abandoned.

'When vipassana is developed, what purpose is achieved? Understanding is developed. When

understanding is developed, what purpose is achieved? Ignorance is abandoned.

'Monks, the mind tainted by lust is not released; understanding tainted by ignorance is not
developed. Thus the release of heart is due to the fading away of lust; the release by
understanding is due to the fading away of ignorance.' 157

Thus the purpose of samatha is to alleviate lust, which here stands for all emotional defilements,
whereas vipassana eliminates ignorance, that is, intelligential defilements. Both of these key
Sutta passages strongly emphasize the complementary, integrative nature of these two aspects of
meditation. While there is a clear conceptual distinction, they are not divided up into two
separate baskets (still less into two separate meditation centres!). The early texts never classify
the various meditation themes into either samatha or vipassana. They are not two different kinds
of meditation; rather, they are qualities of the mind that should be developed. Broadly speaking,
samatha refers to the emotional aspects of our minds, the heart qualities such as peace,
compassion, love, bliss. Vipassana refers to the wisdom qualities such as understanding,
discrimination, discernment. Samatha soothes the emotional defilements such as greed and
anger, while vipassana pierces with understanding the darkness of delusion. It is apparent that all
meditation requires both of these qualities, so in seeking to disentangle them we must inevitably
remain in the twilight zone of emphasis and perspective, eschewing the easy clarity of black-&-
white absolutes.

CHAPTER 9 : PREVIOUS STUDIES

Many learned and wise authors have studied and commented on the various versions of the
Satipatthana Sutta. I have learned something from each of these writers, and any virtue in my
work stems purely from my being able to stand on such broad and strong shoulders. This book is
already far too long, so I try to avoid repeating topics that have already been well-treated, except
where re-evaluation is necessary in light of the special methods and materials of the current

82
work. A survey of t h e general writings on satipatthana would be a pleasant but over-long task,
but we may briefly survey those who have undertaken comparative and historical studies.

Oskar von Hinuber hinted at some of the issues involved:

'More complicated is the relation of the Satipatthana Samyutta, SN V141-192 to the

Satipatthanasuttantas found in DN no. 22 Mahasatipatthanasuttanta and MN no. 10
Satipatthanasuttanta, which deserves a detailed study, because it seems that sometimes SN has
preserved smaller parts from which larger units were built, or pieces of texts, which for some
reason or other were not incorporated into the larger suttantas.' 158

Venerable Analayo has recently published a full-scale study of satipatthana, titled Satipatthana:
The Direct Path to Realization. This is a very valuable work, which ably discusses most of the
practical and theoretical issues involved in the study of satipatthana. The author draws from a
vast spectrum of sources, displaying a warm appreciation for t h e perspectives opened up by
different scholars and meditators. Although Analayo is still influenced by the vipassana
interpretation of satipatthana, his presentation is refreshingly moderate. He does no more than
hint at the possible implications of a historical analysis of the subject:

'But the detailed instructions found in the Maha Satipatthana Sutta and t h e Satipatthana Sutta
apparently belong to a later period, when the Buddha's teaching had spread from the Ganges
valley to t h e distant Kammasadhamma in the Kuru country, where both discourses were
spoken.' 159

It is indeed strange that such an important teaching should have been given only in such an
obscure, far-away town. (The Kuru country is near present-day Delhi, and marks the probable
extreme western limits of the Buddha's peregrinations). Stranger still that the discourse would
have been given twice, with only the expansion of one section differentiating them. In fact, it
seems not merely strange but incredible that the Buddha should have taught only the basic
pericope in all his years at SavatthT, etc., and in one of his rare visits to the border countries he
gave such a vastly elaborated teaching, not once but twice. Were the students in the main centres
to be left high and dry for all those years, deprived of the key for fully understanding
satipatthana? This reinforces our contention that the shorter, mainstream teachings on
satipatthana found especially in the Samyutta should be more closely examined, and that the
longer discourses should be seen in this light.

Although Analayo is aware of the different versions of the satipatthana material, the focus
remains firmly on t h e Theravada Satipatthana Sutta.160 For example, the comparison of the
contents of the body contemplation shows, as we shall see later, that certain exercises,
particularly the investigation of the parts of the body, are common to all traditions, while other
exercises are particular to certain traditions. Analayo remarks:

'The reasons for the omissions are open to conjecture, but what remains of the unanimously
accepted core of the contemplation of the body in all the different versions is a thorough
investigation of its anatomical constitution.' 1 6 1

The very fact that the investigation of the body parts is unanimously accepted suggests that the
other meditative exercises are more likely to be additions than omissions. If the traditions
inherited a common list of meditation practices, and some subsequently were lost, there would
seem to be no reason why some exercises would be left out rather than others, and therefore no
reason why there should be a certain practice preserved with complete consistency. Or again if
there was no common core, and all the detailed lists were invented independently by the

83
traditions, there would seem no reason for such consistent features. Another problem is that
some of the exercises, especially in the Sarvastivada Smrtyupasthana Sutra, are clearly
anomalous and best understood as additions. While undoubtedly conjectural, it nevertheless
seems that the most reasonable way of explaining both the similarities and the differences is that
there was a simpler, common root text, elaborated in somewhat divergent manner by the schools.

Thich Nhat Hanh published full translations of all three major versions of the Satipatthana Sutta
in his Transformation and Healing. The translations, by him and Annabel Laity, offer an invaluable
and almost unique opportunity to compare in English a major Sutta in recensions f r o m three
different schools. However the translations sometimes bend too far to accommodate the
translators' ideas. Some comments on the texts are included, but t h e main orientation of the book
is practical, so he does not pursue textual questions in great depth. The most relevant passage in
our current context is this.

'Other differences in the second version [Sarvastivada] are teachings on the kind of concentration
which gives birth t o j o y and happiness, which is equivalent to the f i r s t j h a n a , and a concentration
which abandonsjoy but maintains happiness, which is equivalent to the secondjhana, as well as
meditations on purity, clear light, and signs. All this is evidence that the practice of the Four
jhanas had already begun to infiltrate the Sutra Pitaka, although discretely. By the time of t h e
third version [Mahasanghika], the practice of t h e j h a n a s is mentioned quite openly, by name. The
meditation which observes the pure light can be seen as announcing the first steps in the
formation of Pure Land Buddhism, and the meditation on the sign will be developed in the use of
the kasina, a symbolic image visualized as a point of concentration.'

Apparently Thich Nhat Hanh believes that t h e j h a n a s were a later infiltration into Buddhism; this
would entail that all of the hundreds of discourses mentioning jhanas in the canons were
composed later than the current text. He offers no evidence for this extraordinary view. His
comments here almost all miss the point, simply because he assumes that the current text, the
Sarvastivada Smrtyupasthana Sutra, is the original source of these various practices. However
they are all found elsewhere in the canons and the current text is probably a somewhat later
compilation. His association of the perception of light with Pure Land is far-fetched; the
perception of light is the standard remedy for sloth & torpor, and surely the origins of Pure Land
should be sought rather among the devotional passages in the early discourses, particularly the
practice of the 'recollection of the Buddha' (buddhanussati). Again, his comment regarding the
'meditation on the sign' misses the point, for he apparently has been misled by the translation
into thinking that the practice described is visualization, whereas comparison with the Pali
version shows that it in fact refers to reviewing of jhana. 162 Thus, however beneficial Thich Nhat
Hanh's practical advice may be, his textual analysis is not very useful for a historical inquiry.

Thanissaro Bhikkhu discusses the issues briefly in The Wings to Awakening. He renders 'dhammas'
in satipatthana as 'mental qualities' rather than 'phenomena', since he believes that the various
groups of dhammas are chiefly variations on the abandoning of the hindrances and the
development of the enlightenment-factors. He mentions t h e Vibhanga and the Sarvastivada
version as historical support for this argument. However he retains a typically reserved attitude
towards the possibility of reconstructing a projected original text. While it is certainly true that
the main factors in the fourth satipatthana are mental qualities, other aspects of satipatthana are
also mental qualities, such as feelings, so this does not serve to adequately distinguish t h e
meaning here. Below we will see that the most significant difference between the fourth
satipatthana and the rest is that it treats of causality, so if I were to translate dhammas here I
would use 'principles'.

84
Bhiksu Thich Minh Chau furnishes details of the Sarvastivada Smrtyupasthana Sutra in his
invaluable work The Chinese Madhyama Agama and the Pali Majjhima Nikaya. He points out that
satipatthana is the only group of the 37 wings to enlightenment to exhibit any noteworthy
variation between the Sarvastivada and Theravada. But he overlooks the importance of t h e
differences when he remarks: 'Both versions offer almost the same materials, as the basic
approach to the contemplations is identical.' It therefore seems necessary to modify his
conclusion that: 'both versions were derived f r o m the same source but the selection of details was
left to the compilers more or less freely.' As Thich Minh Chau has well demonstrated in several
other places, the differences in arrangement are not 'free', but reflect the emerging doctrinal
divergences between the two schools.

R.M.L. Gethin in his The Buddhist Path to Awakening, notes some of the divergences between the
various versions of the Satipatthana Sutta, and says that:

'This has led some scholars, such as Schmithausen and Bronkhorst, to speculate on t h e nature of
the "original" specification of the first and fourth satipatthanas: the former suggests that
watching the body originally consisted only of watching the postures of the body, and the latter
(following the [Abhidhamma] Vibhanga) suggests that it consisted only of watching t h e different
parts of the body. Much of their discussion is at best highly speculative, and at worst
misconceived.

'Schmithausen, for example, suggests that the redactors of the Pali canon have put t h e watching
of breath first because in some canonical texts, such as the Anapanasati Sutta, it is presented as
the preliminary stage of the four satipatthanas. This is a misunderstanding. As we have seen, in
the Anapanasati Sutta watching the breathing is not a preliminary of t h e satipatthanas, it actually
is the satipatthanas.' 163

I hope my work is not merely re-ploughing such barren fields of speculation. I might say, in
defence of speculation, that bold hypotheses are essential for t h e advancement of knowledge; but
they must be tempered by a cautious evaluation of evidence. Gethin's observation that
anapanasati is satipatthana is certainly correct; but Schmithausen's error is understandable, for
he may have been influenced by Sarvastivadin texts such as the Abhidharmakosa, which, as we
shall see, do indeed treat anapanasati as a preliminary to satipatthana. Schmithausen's article is
in German, of which I know none, but I have kindly been supplied with a summary by Roderick
Bucknell. I rely on this for the remarks below, and hope that I do not misrepresent the author.

Schmithausen considers the three available versions of t h e Satipatthana Sutta. He does not take
the Abhidhamma texts into account; if he had done so, perhaps he may have reached different
conclusions. He notes, correctly, that the section in the Satipatthana Suttas dealing with t h e
feelings and the mind are similar in all versions (as indeed they are in t h e Abhidhamma, too). He
then notes that they share a similar phrasing, for example: 'When feeling a pleasant feeling, one
understands: "I feel a pleasant feeling".' He assumes that the other sections in the original version
would have had a similar structure. Certain of the sections in the Pali text do in fact share a
similar format. In the contemplation of postures, for example, it says: 'When going, one
understands "I am going"; when standing one understands "I am s t a n d i n g " . ' and so on. But other
passages, especially several of the exercises in body contemplation, are phrased in a different
manner; for example, unlike the sections on feelings and the mind, they are illustrated with
similes. Schmithausen believes that these were unlikely to have been authentic.

However, the fact that t h e exercises do not consistently have similes would, at the most, suggest
that the similes were added; in addition, one could argue that similes are appropriate for these
meditations, which have an aspect of visualization. Another point is that the sections on the parts

85
of the body and the elements, though not formally identical with the contemplation of feelings
and mind, do correlate well with the basic satipatthana formula. One is advised to 'review this
very body...' (imam'eva kayam...paccavekkhati). The emphatic imam'eva kayam ('this very body') is
reminiscent of the repetitive kaye kaya-@ ('a body in the b o d y . ' ) ; and the ocular paccavekkhati
('reviews') echoes -anupassi ('contemplates'). Elsewhere t h e contemplation of the parts of the
body is summed up like this: 'Thus one dwells contemplating ugliness in this body' (iti imasmim
kaye asubhanupassi viharati).168 This is similar to the standard satipatthana pericope, 'one dwells
contemplating a body in the b o d y . ' (kaye kayanupassi viharati). These close parallels in the
manner of phrasing of t h e practices clearly indicate that these passages are describing similar
kinds of things. These considerations do not, in and of themselves, prove that the contemplation
of the parts of the body was originally part of the Satipatthana Sutta, but they are sufficient to
establish that such practices at least fit in with the general m a n n e r of presenting satipatthana.

In any case, Schmithausen concludes that the original text consisted of the passages that are
formally congruent with the contemplations of t h e feelings and t h e mind; namely, the awareness
of the four postures in body contemplation, and the sections on the hindrances, sense media, and
enlightenment-factors in contemplation of dhammas. He also concludes, I think rightly, that the
'vipassana refrain' (contemplation of principles of origination and dissolution) of the Theravada
Satipatthana Sutta was not original. He adduces various other supports for his conclusions,
primarily t h e agreement between the different recensions. This does indeed offer support for
some of his conclusions, especially as regards t h e contemplation of dhammas; but he rejects the
section on the parts of the body even though they occur in all three of his texts (as well as t h e
Abhidhammas). Thus Schmithausen is prepared to stick by his formal analysis against the
universal testimony of the texts. This is going too far. I would rather say that t h e fact that the
texts do not correspond with the analysis suggests there is something wrong, or at least
incomplete, with the analysis.

Schmithausen believes that the original text would have been phrased in a consistent style.
However it is obvious that the Abhidhamma texts employ a more consistent, rigorous style than
the Suttas, so the congruent sections could, if one wished, be dismissed as evidence of later
scholastic formalism. So I think his criterion of stylistic consistency is merely suggestive, and
cannot carry enough weight to support any solid conclusions.

There is, however, an analysis of the meaning of the texts that Schmithausen employs parallel to
his purely formal analysis; and it is here, I believe, that his deeper agenda is revealed. He says that
the sections on feelings and mind, which are established as authentic by textual agreement,
describe a practice of non-judgmental awareness, simply knowing the situation as it is, without
evaluating it or attempting to change it. He also says that this is how the essence of mindfulness
practice is portrayed in the opening paragraph of the Satipatthana Sutta. This claim is in
agreement with almost every modern exposition of satipatthana - and I think it is wrong. A monk
once remarked to me in this connection that we are more influenced by Krishnamurti than by the
Buddha; perhaps a greater influence than both of these has been Zen and the Art of Motorcycle
Maintenance. It is important to be clear on this point, for otherwise the significance of much the
discussion to follow will be misconstrued. Here I raise a few objections to this view, and will
return to this topic, directly or indirectly, throughout this book.

1) The most obvious, and probably most important, objection to the idea that satipatthana is
essentially a system of choiceless awareness is simply the fact that there are four satipatthanas.
One is obviously supposed, in some sense or another, to choose one of these four as a framework
for meditation. One has t o j u d g e , discriminate, and direct the mind, at least to some extent, even
just to stay within the domain of one's meditation. Nowhere do the early texts imply that the four
frameworks may be neglected or promiscuously mixed, and nowhere is satipatthana described as

86
just 'being aware of whatever arises in the present moment'. We shall see that as this idea gained
hold during t h e historical evolution of satipatthana the importance of the division into four
sections becomes marginalized. In fact, the practice of 'being aware of whatever arises' is in the
Suttas called 'clear comprehension' (sampajanna), not satipatthana.

2) Schmithausen concludes that the sections dealing with body contemplation were later
interpolations, because they involve a more directed kind of meditation. However this fact
corresponds with a well-known principle of meditation. In the beginning stages the hindrances
are likely to be strong and mindfulness weak, so the wise meditator will direct and hold
awareness with some strength onto a chosen object. As the hindrances weaken and mindfulness
grows strong, one can gradually let go more and more until finally one relinquishes all control
and enters samadhi. The pattern of the Satipatthana Sutta precisely mirrors this principle. It is
thus plausible to interpret t h e variations in the style of the various exercises, not as evidence of
textual corruption, but as indicating different approaches suitable for different progressive stages
of meditation.

3) Several Suttas clearly suggest the use of choice and j u d g m e n t within the context of
satipatthana. In fact, there is even a discourse in the Samyutta that explicitly describes how to
develop satipatthana in both 'directed' and 'undirected' modes. 175 As suggested above, the
undirected mode of satipatthana is relevant for one who has already dispelled the hindrances
through the attainment of samadhi.

4) I do not know on what grounds Schmithausen believes that the opening paragraph of t h e
Satipatthana Sutta, which is of course the stock description of satipatthana, portrays a practice of
direct non-judgmental observation. The most likely reason would be the use of the t e r m
anupassana. I will show later that a close investigation of the meaning of this term does n o t j u s t i f y
this conclusion. Suffice to note here that, although Schmithausen believes that such choiceless
awareness is incompatible with such practices as the meditation on ugliness (asubha), in fact the
phrase asubhanupassi occurs several times in the canon. 177

I am afraid that the above criticisms will seem a bit too harsh on poor Schmithausen. He was a
pioneer, and must be given credit for bringing these textual variations to light. In fact, although
each of the scholars mentioned above might disagree on details, all of t h e m agree on a number of
important points. Firstly, that the texts as we have t h e m are the outcome of a historical process.
Secondly, that there is no a priori reason to assume that the Pali tradition, or any other tradition,
is the authentic one. Thirdly, that the variations in the Satipatthana Sutta are significant enough
to warrant investigation.

The conclusions reached by Bronkhorst constitute a considerable advance. He makes good use of
the Theravada Abhidhamma Vibhanga and the Sarvastivadin Dharmaskandha, although he still
omits consideration of the Dharmaguptaka Sariputrabhidharma and t h e Prajnaparamita. I agree
with most of what he has to say specifically dealing with satipatthana; in several cases I arrived at
the same conclusions independently. It is worth quoting at length his main points. I have added a
few expansions in square brackets for clarity.

'The Vibhanga itself must - as pointed out by Frauwallner 177 - have developed out of an earlier
work which also underlay the Dharmaskandha of the Sarvastivadins...

'Our question is: did the "Original Vibhanga" make use of the Sutras in their finished form, or did
it rather use pieces of tradition which were still more or less free-floating and would only later be
taken into the Sutras known to us? In the former case the agreement between the descendants of

87
 the "Original Vibhanga" and the Sutras would have to be great; in the latter, we might hope to
find in t h e Vibhanga and Dharmaskandha traces of a time prior to t h e compilation of the Sutras.

'Whether such traces have survived is not certain. There is, however, one passage in the Pali
Vibhanga which may retain some ancient features. It occurs in t h e explanation of the 4
satipatthanas 1 6 8 .

'The "Original Vibhanga" must have contained this same description of t h e 4 satipatthanas,
because it is also found in the Dharmaskandha, with the difference that t h e Dharmaskandha adds
items after those given in the Vibhanga...The items added are also found in the same or similar
form in the Sutras which deal with the 4 satipatthanas, and we may assume that the
Dharmaskandha was influenced by t h e m .

'It is possible, but unfortunately far from certain, that the specification preserved in the Vibhanga
is older than most of those found in the S u t r a s . [Here Bronkhorst summarizes Schmithausen's
arguments and raises some objections to t h e m ] .

'Apart from these in themselves not very decisive considerations, there is one argument which
lends some plausibility to the view that the "observation of the positions of the body" was not
originally the first of the 4 satipatthanas [as maintained by Schmithausen]. Briefly stated it is that
in Buddhism mindfulness is of two kinds (or better perhaps: degrees); "observations of t h e
positions of the body" is of one kind, the 4 satipatthanas of the other.

'In order to recognize t h e two kinds of mindfulness we t u r n to the stereotype description of the
road to liberation which often recurs in the Sutras [i.e. the "gradual training"]. It distinguishes
between preparatory exercises on the one hand, and "meditation" proper on the other, the two
being divided by the m o m e n t when the monk went to a lonely place and sat down in the
prescribed manner. Mindfulness plays a role both before and after this moment, but in different
ways. Before this m o m e n t the monk "When going out and returning acts with clear
comprehension; when looking forward and to the s i d e . w h e n bending and stretching his
limbs...when bearing his robes and bowl...when eating and drinking...when defecating and
u r i n a t i n g . w h e n going, standing, sitting, sleeping, waking, speaking, and keeping silent acts with
clear comprehension" 169 ; in short, the monk practices the "observation of t h e positions of the
body". After this m o m e n t the situation changes. The monk no longer makes any movement. Yet
his first act in this motionless position is "calling up [establishing] mindfulness" (parimukham
satim upatthapetva). As the expression indicates, it is here that t h e satipatthanas ["establishings of
mindfulness"] would seem to come in. If this is correct, there is no place for "observation of the
positions of the body" in the 4 satipatthanas.

'What then constitutes satipatthana on t h e body in this motionless position? Obviously only this:
the monk directs his mindfulness to the different parts of the b o d y . We m a y . c o n s i d e r the
possibility that "observation of the constituents of the body" was originally the satipatthana on
the body. And this would confirm the view that t h e "Original Vibhanga" was composed before t h e
4 satipatthanas were given the explanations we now find in t h e Sutras.'

The remainder of that portion of Bronkhorst's article which deals with satipatthana mainly
concerns the Mahasanghika Ekayana Sutra. He develops some arguments to the effect that this
text may contain some archaic features - which is possible, although the text as a whole is late -
and suggests that the original specification of contemplation of dhammas may therefore have
been the enlightenment-factors only. I discuss this more below.

88
A key point in Bronkhorst's argument is that there are two different degrees of mindfulness in
the gradual training, and that it is the second of these, 'meditation proper', that can be identified
with satipatthana. I agree with this; but it is, for some, such a radical claim that it requires more
detailed reasoning. I would adduce the following arguments in support of this theory:

1) The standard description of 'observations of the positions of t h e body' does not include
the word 'mindfulness' in the description of the practice itself. The act is described, as
above, by saying the monk 'acts with clear comprehension' (sampajanakarf hoti).
Accordingly the overall practice is called simply 'clear comprehension' in the Samyutta. 170
It would seem, therefore, that it is, in the early Pali idiom, quite legitimate to describe t h e
practice of being aware throughout one's daily activities without even using t h e word
'mindfulness'. Only in the developed version of the gradual training, however, is the
practice described as 'mindfulness & clear comprehension'. On the other hand, the use of
the word 'mindfulness' in the case of the monk who sits down cross legged in the forest to
meditate is absolutely standard, consistent, and intrinsic to the description of the practice.
2) The word 'establishment' (upatthana) does not occur in the standard description of
awareness of activities. To be sure, it does occur occasionally elsewhere in similar
contexts, chiefly sense restraint, but it is not standard in the gradual training. On the
other hand, the t e r m 'establishment' is intrinsic to the passage on the monk who sits
down in the forest to 'establish mindfulness'.
3) The various versions of the gradual training, so far as I know, never specifically mention
the four satipatthanas at the stage of awareness of daily activities. But, while it is not
standard, there are at least some contexts that mention t h e four satipatthanas at the stage
of sitting down to meditate. In a Sarvastivada version of the Ganakamoggallana Sutta,
after the section on clear comprehension the four satipatthanas are brought in, leading as
usual to j h a n a and then various psychic abilities, culminating in enlightenment. 1 7 1 The
Dantabhumi Sutta is similar, although there the four satipatthanas are placed a little later,
after the abandoning of the hindrances in the place normally taken by the firstjhana. 1 7 2
The Dharmaguptaka version of the Samannaphala Sutta, however, differs from all these in
placing the four satipatthanas before awareness of activities. 173
4) The practice of anapanasati is invariably described after the monk has gone to a forest to
sit meditation. This obviously pertains to the same stage in the gradual training. Since
anapanasati is a major, or the major, paradigm for satipatthana meditation, this clearly
implies that satipatthana applies to the stage of sitting down to meditate in the forest.
5) Several texts available in Theravada and Sarvastivada versions list sequences of qualities,
closely connected with the gradual training, that place mindfulness & clear
comprehension early on as a foundational practice. One such text explicitly differentiates
between 'mindfulness & clear comprehension' and the four satipatthanas: associating with
true persons > hearing the true Dhamma > faith > causewise attention > mindfulness &
clear comprehension > sense restraint > three ways of good behaviour (by body, speech,
and mind) > four satipatthanas > seven enlightenment-factors > realization & release. 174
6) The following passage speaks of exactly the same actions as the formula for clear
comprehension: 'Then, after he has gone forth thus, his companions in the holy life advise
and instruct him thus: "You should move to & f r o thus; you should look ahead and to t h e
side thus; you should flex and extend the limbs thus; you should wear the outer robe,
bowl, and robes thus".' 1 7 5 This is instruction in the basics of monastic protocol, especially
regarding the morning alms round, when a monk or n u n leaves the leafy seclusion of the
monastery to venture into the distractions of the village. This shows that the passage on
clear comprehension pertains more to preliminary ethical conduct than to the practice of
meditation.

89
Thus Bronkhorst's brief observation finds considerable textual support. This suggests that an
idiomatic rendering of satipatthana would be simply 'meditation'. Establishing mindfulness,
abandoning the hindrances, and entering j h a n a are the key meditative stages in the gradual
training.

CHAPTER 1 0 : PRE-BUDDHIST MEDITATION

EARLY BRAHMANICAL SOURCES

The earliest evidence for meditative culture anywhere in the world is from the Indus valley
civilization. This was a vast, sophisticated, and well-organized society which, at its peak in 2500-
3000 B.C.E., stretched from what is now Pakistan to the Ganges valley. The evolution of this
civilization can be traced from as far back as 7000 B.C.E. in Afghanistan, with a series of villages
that became towns, and then towns that became cities. It was therefore an indigenous Indian
culture. There is a strong continuity with later Indian culture, although scholars are not quite
sure who these people were. The iconography suggests that they were the 'noseless' and 'black'
peoples (Dravidians?) whose destruction at the hands of the Aryans is still dimly remembered in
the Rg Veda. Perhaps the most intriguing remnants of their brilliant world are the thousands of
exquisitely carved seals, little clay tablets that were probably worn by the citizens as a
religious/family/occupational icon, and, of course, as a magic totem. These seals contain some of
the world's oldest writings, which are as yet undeciphered.

The most interesting for our current purpose are a few seals that clearly depict a god as a yogi
sitting in meditation. These images are spine-chillingly similar to the amulets that are still widely
popular in Buddhist countries today. The yogi is usually identified on the basis of iconography as
a 'proto-Siva'. He sits, not in the 'lotus posture' of the Buddha, but in either siddhasana (with legs
crossed at the ankles) or mulabandhasana (with soles of the feet pressed together). Both of these
postures are associated with psychic powers. One of the images depicts snakes rising beside him,
a startling image familiar from Eden to the Pali canon. The image of the Buddha with a serpent
rising over him is still popular today, taken from the Muncalinda Sutta of the Udana. It is, of
course, most famous as the symbol of the 'kundalini' of the Hindus. But whereas the serpent rises
over the Buddha, signifying transcendence, in the proto-Siva image the serpent rises only to the
forehead. In later theory this place, the jnanacakra, was associated with lights, subtle forms, and
psychic powers, and would therefore seem to be equivalent to the Buddhist form jhanas. These
speculations are too tenuous to make much of. However, it is certain that here is an ascetic who
has, as the Buddhist texts have it, 'gone to the forest, to the root of a tree, or to an empty hut, sat
down cross-legged, and set his body erect...' Has he taken the next step in this meditative
training: 'establishing mindfulness'?

Sati in Buddhism is functionally described in terms of either sara 'memory', or anupassana

'observation'. The relation between these two ideas is, to our mind, strange, and attempts at
explanation have, I believe, been misleading in giving a psychological explanation for what is a
historical, linguistic development. Sara is from the same root as sati, and is obviously the
historical meaning. Sati came to mean, in the Brahmanical tradition generally, 'received tradition,

90
memorized texts.' This meaning is attested in the early Suttas, where it is treated identically in
Buddhist and Brahmanical contexts: one 'remembers what was said and done long ago'.

Sati is apparently used since the Rg Veda (perhaps a thousand years before the Buddha) in two
senses: to 'remember' or 'recollect', and to 'bear in mind'. The significance of this should not be
overlooked. Sati is not merely a word one uses to refer to some texts one remembers; it is highly
probable that t h e development of the culture of memorizing texts lead to the discovery,
investigation, and development of what 'memory' is. That is to say, those who memorized the
Vedic mantras were engaged in an early form of mental culture, a mental culture where 'memory'
was a vital quality. While it is impossible to document this in detail, it again seems very likely that
this form of mental culture was one of the strands that became woven into what we know today
as 'meditation'.

In the Chandogya Upanisad a father asks his son to fast for 15 days, then tests him on his memory
of the Vedic texts. He fails dismally; but after eating again he can remember easily. His father
explains:

'If, from a great blazing fire, there is only one coal left glowing, it can easily be made to blaze up
again by putting grass on it. Even so, my dear son, there was [due to fasting] but one part in
sixteen left to you and that, lighted up with food, blazed up and by it you remember now the
Vedas.' After that he understood what his father meant when he said: 'Mind, my dear son, comes
from food, breath f r o m water, speech from fire.' 176

The Buddha was once asked by a Brahman why the (Vedic) mantras are sometimes easy to
remember and sometimes not. 177 Typically, he answers that when the five hindrances are present
the mantras are not clear; when the five hindrances are absent the mantras are clear. This is a
straightforward example of how t h e science of memorizing texts would lead naturally to
investigation of the mental qualities necessary for success in such an ambitious venture. We still
use the 4000 year old word 'mantra', which originally referred to the Vedic texts, as a term for a
meditation word, a sound or phrase traditionally taken from the ancient texts that one repeats
over and again as a support for meditation. The relation between recollection and meditation is
strong even today in Buddhism. For example, most Buddhists are familiar with the basic passages
for 'recollection' (anussati) of the Triple Gem. These form the basis for both the regular chanting
at Buddhist ceremonies, and also the meditation on the Triple Gem.

In a similar fashion, the verses of the Vedas had a highly numinous, mystical significance for the
ancient Brahman priests, and it would have been natural for the more contemplative among
t h e m to induce exalted states of consciousness through the ecstatic recollection of t h e sacred
words. In order to memorize long texts it is, of course, necessary to repeat passages over and over
again. If one does this mechanically, without interest, the memorizing will not succeed. One must
bring inspiration, joy, attention, and understanding to t h e task. One must learn to 'stay with' t h e
present m o m e n t - and here we are crossing over to the familiar Buddhist idea of'mindfulness'.

This psychology also emerges in the usage of the word dhi, familiar as the root of the Buddhist
term 'jhana'. This makes its appearance as 'thought', etc. and seems to have a special connection
with the 'visioning' of the Vedic poetry: dhi is the intuitive awareness as the poet/priest 'sees' the
verses. This 'thought' (dhi) or 'mind' (manas) is to be disciplined (yoga) by the reciters:

'The priests of him the divine Savitr well-skilled in hymns

Harness their mind, yea, harness their holy thoughts' 178

91
But we should not think that this Buddhist meditative term is well developed in its meditative
meaning of 'deep absorption' before the Buddha. In the Brhadaranyaka, j h a n a is specifically
contrasted with t h e deep stillness of the True Self.

'Which is the Self?'

'That person here made of cognition among the senses [breaths], the light within the heart.
He, remaining the same, wanders about the two worlds as if thinking (dhyayati), as if playing
(lelayati).'179

The Upanisads are constantly reminding us of the necessity of preserving the correct mental
attitude while performing the rituals; and this applies as much to the actions as to the words. One
performs the rituals with one's whole being, contemplating the significance of each aspect of the
performance as one carries it out. Even the earlier Brahmanas explicitly allow that if a ritual
cannot be carried out physically it may be performed by 'faith', i.e. as a purely mental act.180
Notice that the Buddhist definition of mindfulness referred to above mentions both what was said
and what was done. In this immersion of awareness in one's actions we can discern a precursor to
the Buddhist emphasis on mindfulness through all one's activities.

It is a curious thing that when we look at t h e sources most likely to be contemporary with the
Buddha - namely t h e Brhadaranyaka and t h e Chandogya - we find that these well-known
meditative terms are used less frequently, and a word apparently foreign to Buddhist meditation
is found far more often. This word is upasana. Edward Crangle, following Velkar, has studied this
term in detail, and lists the frequency of occurrence. In the Brhadaranyaka, upasana occurs 63
times, j h a n a thrice, and yoga twice. In the Chandogya, upasana occurs 115 times, j h a n a twelve
times, and yoga again twice. 181 Upasana is obviously a key t e r m in considering the emergence of
meditative psychology in Indian tradition, but it is no easy matter to ascertain exactly what it
refers to, and even more difficult to tease out how it might relate to Buddhist terminology. It is
translated sometimes as 'worship' and sometimes as 'meditation', and in fact seems to embody
the shift f r o m an external worship and ritual towards t h e inner contemplation. Crangle says
upasana is 'a contemplative process wherein the object of worship is an object of concentration.' 182
The following might convey something of the mystical tone of upasana:

'Next, of this breath, water is the body. Its light-form is that moon. As far as the breath extends so
far extends water and that moon. These are all alike, all endless. Verily, he who
meditates/worships (upasana) them as finite wins a finite world. But he who meditates/worships
t h e m as infinite wins an infinite world.' 183

Upasana evidently encompasses a wide spectrum of spiritual consciousness. Velkar says it is

meditative, emblematic (involving elaborate symbolism), and analytic (in making philosophical
distinctions). It takes a large variety of objects, concrete and abstract: God, 'om', sun, moon,
lightning, wind, space, fire, water, breath, 'That as Great', 'That as Mind', etc.

Crangle makes the intriguing suggestion that upasana is related to the Buddhist t e r m
satipatthana, especially the last element of this compound, upatthana.184 This may be supported on
a number of grounds. The sound of the words is almost identical, especially in Sanskrit (upasthana
and upasana). Though they are f r o m different roots, the construction and basic meanings are
similar: upa+as means to 'sit near'; upa+stha means to 'stand near'. From there they both
developed the sense of 'wait upon, serve, attend', and then to 'pray, worship'. In a more
specifically meditative context they are both used largely in t h e sense of the initial grounding on
the meditative/contemplative object, rather than the resulting state of absorption. We also note
that some of the meditation objects for upasana are also found in satipatthana: the breath, water,
fire, space, bliss, mind, etc. So it seems Crangle's suggestion can be accepted. The major

92
contemplative practice of the pre-Buddhist period is upasana, and this practice finds its closest
Buddhist connection, surprisingly enough, not w i t h j h a n a or samadhi, but with satipatthana.

Investigation of pre-Buddhist meditation terminology is hampered by the fact that the Vedas
have little or nothing on meditation and even the early Upanisads have nothing clear. The
earliest clear descriptions of meditation outside of Buddhism are in later texts of the Upanisads
and the Jains. These are later than the Suttas, so it is likely there is Buddhist influence. However,
there is no reason why even late texts should not preserve old traditions, too.

There has in recent years been doubt thrown on the accepted wisdom that t h e early Upanisads
were pre-Buddhist. The standard list of Brahmanical texts in the Suttas does not mention t h e
Upanisads. But one passage in t h e Tevijja Sutta, discussing contemporary controversies among
the Brahmans, refers to Brahmanical schools teaching different paths. 185 These have been equated
by Jayatilleke with several of the Brahmanas (which include the Upanisads) as follows. 187

Table 10.1: Upanisadic texts in the Tevijja Sutta

Brahmanical schools in the Brahmanical Brahmanical Text

Tevijja Sutta school
Addhariya Yajur Veda Satapatha Brahmana (incl. Brhadaranyaka
Addhariya Upanisad)
Tittiriya Yajur Veda TaittirTya Brahmana
Tittirya
Chandoka Saman Veda Chandogya Brahmana
Chandoga
Bavharija Rg Veda Bavharija Bahvrvas Brahmana (lost, but incl. in Aitareya
and KausTtakT Brahmanas)

This evidence suggests that the Upanisadic schools were in existence, but their tenets were still in
ferment. Perhaps the Upanisads that we have today derive f r o m the later settled tenets of each of
these strands of Brahmanical thought. 187 But whether or not the Upanisads in their current form
existed at the Buddha's time, there is no doubt that ideas we can call 'Upanisadic' were
prominent. In the sphere of metaphysics we can cite the Buddha's critique of such ideas as that
the self is infinite (anantava atta), or that the self is identical with the world (so atta so loko), or that
'I am He' (eso'hamasmi); or indeed the Buddha's condemnation of the suggestion by a certain
Brahman cosmologist that 'All is oneness' (sabbam ekattam). It would seem only natural to connect
such metaphysics with samadhi attainments, as implied by the Brahmajala Sutta.

It is necessary to proceed with caution here. The early Upanisads, especially the Brhadaranyaka,
usually regarded as the earliest and most important, are a very mixed bag. The Brhadaranyaka
includes passages of lyrical beauty, sophisticated philosophy, exalted metaphysics, and witty
dialogue. It is closely concerned with ideas like the mind, t h e breath, and oneness, which are
suggestive of a meditative culture. It distinguishes between m e r e perception (sauna) and
liberating understanding (panna), and emphasizes the centrality of cognition (vinmna) as
contrasted with the more dynamic conceptual and emotive aspects of mind (mano). Therefore it
insists on the necessity for personal experience rather than m e r e book learning. It frequently
upsets preconceptions - women have strong supporting roles, and sometimes Brahmans are
depicted as having to learn about Brahma f r o m the Ksatriyas. (Even more remarkably, in t h e
Chandogya Upanisad there is a satire depicting Brahman priests as dogs, reminiscent of an
uncharacteristically scathing satire in the Anguttara. 188 )

93
But the Brhadaranyaka also retains much that is banal and even brutal. It endorses the sacrifice.
It is unabashedly materialistic. It is full of sophistical thaumaturgy and hocus-pocus. It contains
black magic - a curse to place on one's rival in love. It includes crude sex magic. If one's woman is
reluctant to participate she should first be bribed with presents; 'and if she still does not grant
him his desire, he should beat her with a stick or his hand and overcome her'. 189 (Those who like
to imagine that so-called 'tantric' practices indicate an improvement in women's status in
andocentric Indian spiritual culture, please take note.) It hardly needs saying that such ideas are
totally incompatible with any genuine mind culture. The text is a testament to the diversity of
ideas that the ancient Brahmans could regard as 'spiritual', and to the elasticity of the compilers
of the text we have today.

Let us look at some of the passages most suggestive of meditation. From the Brhadaranyaka:

'Therefore let a man perform one observance only, let him breath up and let him breath down,
that the evil death might not reach him.' 1 9 0

'The unseen seer, the unheard hearer, the unthought thinker, the uncognized cognizer... There is
no other seer but he, no other hearer, no other thinker, no other cognizer. This is thy self, the
inner controller, the i m m o r t a l . ' 1 9 1

'Therefore, knowing this, being calm, tamed, quiet, enduring, concentrated, one sees the soul in
oneself.' 192

By themselves such passages are too vague to reach any clear conclusion regarding meditative
practices. And even t h e last passage, which is the most suggestive, has 'faithful' as a variant
reading for 'concentrated'. The Chandogya has a slightly more explicit passage:

'As a bird when tied by a string flies in every direction and, finding no rest anywhere, settles
down at last on t h e very place where it is fastened; exactly so, my son, that mind, after flying
around in every direction and finding no rest anywhere, settles down on breath; for indeed, my
son, mind is fastened to breath.' 193

But for clear teachings on meditation we must go forward to t h e Svetasvatara Upanisad.

'By making his body the under-wood and the syllable "Om" the upper-wood, man, after repeating
the drill of meditation, will perceive the bright god, like the spark hidden in the wood.' 1 9 4

'If the wise man holds his body with the three upright parts even, and turns his senses with his
mind towards the heart, he will then in the boat of Brahman cross over all the fearful streams.' 195

'Compressing his breath, let him, who has subdued all motions, breath forth through the nose
with gentle breath. Let the wise one, being heedful, keep hold of his mind, that chariot yoked
with wild horses.' 1 9 7

'When yoga is being performed, the forms that come first, producing apparitions in Brahman, are
those of misty smoke, sun, fire, wind, fire-flies, lightnings, and a crystal moon.' 1 9 7

These are fairly straightforward references to meditation, and they will not sound unfamiliar to
anyone versed in Buddhist meditation. The simile of meditation like two fire-sticks is well known
in the Buddhist texts. 198 Notice t h e close connection in SU 2.9 between 'heedfulness' (appamada)
and 'keeping hold' (dharana), a term semantically equivalent to sati. It seems that the earliest

94
Brahmanical meditation subjects were the contemplation of the mystical syllable 'Om' and the
breath. Of course, the 'breath' and the 'word' are closely related and are mystically identified in
the Upanisads; in practice, the yogis may have recited 'Om' together with the breath. The
Upanisads have many passages that assert the supremacy of the breath over the sense faculties
and mind ('mind' here meaning thoughts and emotions). These can be understood as an
allegorical description of the evolution of awareness f r o m the diversity of externals towards a
unity with the breath.

The breath is a prime exercise in satipatthana body contemplation, and other aspects suggestive
of satipatthana can also be discerned in the Upanisadic tradition. Just as in the Satipatthana
Samyutta, the dependence of the breath (body) on food is stressed. 199 The elements are of course
universal throughout t h e ancient world, and were commonly worshipped as deities. For example
Agni (Fire) was a major deity in the Vedas, and undoubtedly inspired ecstatic contemplation.
Vayu (air) was also worshipped in the Vedas. The Earth (Mother), whose symbols pervade t h e
iconography of Buddhism, was also widely revered, and seems to be associated with the Indus
Valley religion. The parts of the body are worshipped in the Chandogya Upanisad: hair, skin,
flesh, bone, marrow. 200 All of these appear in the Satipatthana Sutta list of body parts, and in the
same order. Charnel grounds have long been a favourite haunting ground of a certain type of
ascetic. The later MaitrT Upanisad opens with some body contemplations for inducing dispassion
(viraga), but this is almost certainly under Buddhist influence. 201

The other satipatthanas - feelings, mind, and dhammas - might even be compared with the
famous Brahmanical threesome: mind, being, bliss (cit, sat, ananda). Mind and bliss are obvious
enough. As for being, this is a fundamental philosophical term for the Upanisads, just as dhamma
is t h e fundamental term for Buddhism. The dhamma theory was clearly developed to provide an
explanation for phenomenal reality opposed to the Brahmanical conception of an absolute
underlying ground of being. And indeed we find that the contemplation of dhammas prominently
features the same term for being, sat, that was so important for the Brahmans; yet here it is
treated, as always in t h e Suttas, in a thoroughly empirical, anti-metaphysical way: the 'presence'
or 'absence' of good or bad mental factors according to conditions. Another, similar, list also
reminds us of the satipatthanas: food, breath (=body), mind (or thought, manas), cognition
(vynana=mind, citta), bliss (=feelings).202 Whether or not there is any real historical link between
these specific sets, it is apparent that both traditions utilized similar simple lists of physical and
mental phenomena as a guide to spiritual practice.

We shall see towards t h e end of this study that some of t h e later Buddhist theorists posited a
relationship between the evolution of the stages of understanding in meditation and the stages of
understanding in the philosophical outlook of the various schools. It is perhaps not so far-fetched
to see a similar progress here; the Upanisads themselves seem to be aware on some level of this
evolution. In fact we can analyse the stages of Indian religion in terms of the four satipatthanas.
The earliest stages in Indian religion were wholly physical - rituals, chants, the breath, sacrifices
- pursued with the goal of fertility and prosperity. This developed into the practice of self-
torment, which while still physical was predicated on the ability to endure painful feelings. The
next stage was the emphasis on refined states of consciousness identified as the cosmic self.
Finally, the Buddhist critique of metaphysical absolutism, the analysis of dhammas in terms of
conditionality and not-self.

Thus some of the various facets of satipatthana seem to have their precedents in the Brahmanical
traditions. The difference, as so often, is in what is left out (hocus-pocus, rituals, deity worship,
metaphysics, etc.), and in the manner of treatment. The practice is cool, rational, and sensible.
The terminology has been thoroughly subsumed into the Buddhist system. The presentation is

95
purely in terms of clearly discernable empirical phenomena without any metaphysical overtones.
It is not trying to persuade you of a theory but to point you towards your own experience.

T H E BUDDHIST SOURCES

Given the surprising paucity of explicit references to meditation in any pre-Buddhist literature
we are thrown back on the material in the Buddhist texts as our earliest source. There are a
number of problems with this. The compilers of the Suttas may not have had a very good
knowledge of non-Buddhist practices, and may have succumbed to the temptation to put their
opposition in a bad light. In addition, they quite likely described the practices of other schools in
terminology they were familiar with, b u t which was not authentic to the other schools.
Nevertheless, we find both the Buddhist and the non-Buddhist sources agreeing in broad terms in
their description of pre-Buddhist meditation. There seem to be two such streams, represented by
the two styles of practice undertaken by the Bodhisatta before his enlightenment. These streams
are primarily represented by the samadhi practitioners of the Upanisads and the self-tormenters
of t h e j a i n s .

The best-known passage referring to such 'Upanisadic' yogis is the story of the Bodhisatta's
apprenticeship. 203 1 wish to first note on general principles why I consider that the significance of
this passage is seriously overrated. According to the GIST, the Buddha's main teachings are found
in t h e basic doctrinal statements (suttas) together with the interrogative discussions of these
statements (vyakarana). This material does not include much biography, beyond stating that it
was through understanding the four noble truths, etc., or through practicing the eightfold path,
etc., that the Buddha realized enlightenment. Biography as such constitutes one of the later
angas, avadana. However, after the Buddha's passing away the community found that the
Buddha's life story gave the teachings that 'personal touch' so essential for the development of
Buddhism into a popular mass religion. From that time until the present day the Buddha's life,
rather than being occasionally invoked to illustrate a doctrinal point, became the main focus of
attention. The events that are included in the Buddha's life story are known to all Buddhists, and
as a result sometimes minor incidents have been blown up out of all proportion to their original
significance. One obvious example of this is the Buddha's last meal, an obscure incident of
dubious interpretation, absent in some versions, which has become the main battle ground in t h e
controversy regarding the Buddhist position on vegetarianism, with t h e result that the several
straightforward discourses directly addressing the issue, as well as the frequent mention of meat-
eating in the Vinaya, are virtually ignored. Another case is the touching story of the difficult
attempts by the Buddha's foster-mother Maha Pajapati to secure women's ordination. This story
is known to all and is regularly invoked to deny women the opportunity for full participation in
the renunciate life, while ignoring the frequent mention of the 'fourfold assembly' (including
nuns) that the Buddha regarded as the sign of a complete, successful, and long-lasting religion.
Taking note of this principle does not in and of itself mean that these passages are inauthentic,
nor that they should not be taken account of, nor does it suggest taking any specific stand on
such controversies; but it does suggest that we should be more careful in how we weigh and
evaluate the evidence in the early texts.

Nevertheless, despite t h e fact that I consider the story of the Bodhisatta's apprenticeship to
already suffer f r o m too many discussions, still we cannot avoid burdening the story with one
more, sure in the knowledge that it won't be t h e last. Virtually all descriptions have ignored the
obvious point that t h e Ariyapariyesana Sutta mentions three stages of this apprenticeship.
Firstly, learning and lip-reciting of the texts. 204 This is a hint that these are ascetics in the
mainstream Vedic tradition; the nature of the texts is not specified here, but elsewhere the
Buddha recalls that Uddaka Ramaputta claimed to be a vedagu, a master of the Vedas. 205 Anyway,

96
as we noted above, t h e Vedas are the only texts that are known to the early Suttas. (It is
sometimes said that these teachers belong to the Samkhya school, but this claim is based on the
much later Buddhacarita of Asvaghosa, and is clearly anachronistic.) Secondly the path, here
described as faith, energy, mindfulness, samadhi, and wisdom. 206 Thirdly, the goal - formless
attainments. These three stages correspond with the classic threefold formulation of Buddhism -
study, practice, and realization. The five factors of the path are the same as the Buddhist five
spiritual faculties - a fact which is usually overlooked by those who wish to interpret this passage
as implying the 'non-Buddhist' nature of samadhi in general, or of formless attainments in
particular. We cannot know how these qualities were understood in detail in this context; but
terms such as prajna, etc., occur commonly in the Upanisads. If it is true that the five spiritual
faculties were genuinely associated with the Vedic/Upanisadic tradition, it may be no
coincidence that it is in the spiritual faculties that we most frequently meet sati treated as
<
m e m o r y i. 207

It must be noted that the Bodhisatta did not reject the formless attainments in & of themselves. It
is definitely not the case that he practiced samadhi meditation but not mindfulness meditation.
Rather, h e practiced mindfulness meditation to get into samadhi. Samadhi is emphasized in this
account because it was the highest, the most exalted quality acknowledged in those systems, and
because of its sublime peacefulness it was mistakenly taken to be the final end of the spiritual
path. The Bodhisatta became disillusioned with 'that Dhamma', i.e. with the teaching taken as a
whole, because it led only to rebirth in the formless realm, and was therefore 'insufficient' to
reach the 'excellent state of peace', the ending of birth, aging, and death. This is in perfect accord
with the main stream of the Suttas. Elsewhere it is said that ordinary people attain samadhi (here
the fourjhanas 2 0 8 and the four divine abidings 209 ), are reborn in the Brahma realms, and after a
long period of bliss fall back into lower realms. But noble disciples, after reaching the Brahma
realms, attain Nibbana f r o m there. The difference is not in the states of samadhi as such - these
a r e j u s t manifestations of the mind at peace. The difference is in the views and interpretations,
the conceptual wrapping that the experience in bundled up in. The path must be taken as a
whole. If one starts out with wrong view, one's meditation experiences are likely to simply
reinforce one's preconceptions. If one practices samadhi with the view that one's soul will
become immersed in some exalted state of being, well, one will get what one wishes for.

This is t h e most important feature distinguishing this episode f r o m t h e later occasion (quoted
below) when the Bodhisatta recollected his former experience of f i r s t j h a n a . This occurred as a
child, seated in the cool shade of a rose-apple tree. When the Bodhisatta remembered this
experience he realized that: 'That indeed is the path to enlightenment'. As a child, his mind was
uncluttered with views; he had no metaphysical agenda. The peace of the mind w a s j u s t t h e peace
of the mind; and so h e realized that although such states were not the final goal he had been
yearning for, they were indeed the path. This account is preserved in the Maha Saccaka Sutta
(MN 36), the Mahavastu (from the Mahasanghika Vinaya), the Sanghabhedavastu (from the
Mulasarvastivada Vinaya), and the Dharmaguptaka Vinaya. Accounts in the Ekottara (EA 31.8)
and the Lalitavistara attribute all four of t h e j h a n a s to the Bodhisatta as child; while t h e Tibetan
Dulva and an individual Chinese translation (T 757) place the attainment ofj h a n a later, soon after
the going forth. Thus this is clearly regarded by all the schools as a crucial event in the
Bodhisatta's path towards awakening.

One of the most interesting sources for understanding the meditation practices of Brahman
ascetics is the Parayana Vagga of the Sutta Nipata. This text, universally regarded as one of the
earliest and most authentic texts in the Pali canon, consists of a series of questions and answers
between the Buddha and a group of sixteen Brahman meditators. There are several connections
between this text and the Upanisad-style traditions we have been considering; in fact the
closeness of some parallel phrases suggests direct literary influence of one sort or another, 210

97
although there are also direct connections between some of these verses a n d j a i n texts. The list of
Brahmanical texts given is substantially shorter than that in the Brhadaranyaka, suggesting that
it is earlier. It has a satirical reference to an evil Brahman who threatens to 'split heads'; the same
threat occurs several times in the Brhadaranyaka, the difference being that there someone's head
actually does get split!211 The Buddha of course dismisses the efficacy of Vedic knowledge, ritual,
sacrifice, and metaphysical conceptions of 'Self. We meet again the phrase 'seen, heard, thought,
cognised' that we have encountered in t h e Brhadaranyaka, and also frequent reference to the
pairing of cognition with n a m e & form, another Upanisadic idea.

The faith and devotion of these yogis is very moving, and stands in decided contrast with the
sometimes strained relationship between the Buddha and the scholastic and ritualistic Brahmans.
In this friendly atmosphere it seems likely that the Buddha would have, wherever possible, kept
to his normal policy of encouraging his disciples to continue developing whatever spiritual
practices were most inspiring and useful. The introductory verses, which are admittedly
somewhat later, refer indirectly to the five spiritual faculties, 212 and say the sixteen Brahmans are
practitioners of jhana. 213 The teachings are brief and non-technical, but there is recognizable
reference to the fourthjhana 2 1 4 and to t h e sphere of nothingness. 215 And time and time again, the
Buddha exhorts these yogis to be 'ever mindful'. This confirms the association of mindfulness
with Brahmanic culture; the Buddha would hardly have used t h e t e r m so freely if he did not
expect his audience to understand it.

Three discourses in the Bojjhanga-samyutta present t h e claims of non-Buddhist wanderers to

develop Buddhist-style meditation. They say they exhort their disciples to abandon the five
hindrances and to develop, in two cases, the seven enlightenment-factors, 2 1 6 and in a third case,
the four divine abidings. 217 Elsewhere too the divine abidings are attributed to great sages of the
past, notably t h e Buddha in past lives.218 However, although these were indeed later appropriated
by the Brahmanical tradition, they are not attested in any pre-Buddhist texts. The
enlightenment-factors include mindfulness and investigation of dhammas, which is equivalent to
vipassana, as well as samadhi. The wanderers ask, then, what is the difference between their
teaching and the Buddha's? Interestingly enough, the Buddha responds, not by referring to, say,
the four noble truths, not-self, or dependent origination, but by claiming that the wanderers do
not fully understand samadhi practice in all details. This is probably what the Buddha was
referring to when he claimed elsewhere to have 'awakened to j h a n a ' (jhanam abujjhi);219 not that
he was t h e first to practicejhana, but that he was the first to fully comprehend both t h e benefits
and the limitations of such experiences.

The Brahmajala Sutta is the classic exposition of non-Buddhist meditation. It presents a

bewildering array of 62 doctrinal views, many of which were derived f r o m or reinforced by the
misinterpretation of samadhi experiences, including both form j h a n a and formless attainments.
Yogis include both the mainstream Vedic/Upanisadic 'Brahmans' as well as the radical non-
conformist 'samanas', of which the Buddha himself was one. Here, five terms typically describe
the path into samadhi: ardency (atappa), striving (padhana), commitment (anuyoga), heedfulness
(appamada), and right attention (samma manasikara). All these terms are common in Buddhist
contexts; atappa occurs in the satipatthana formula. 'Heedfulness', which we encountered above
in the Svetasvatara Upanisad, is close in meaning to 'mindfulness'. 'Attention' is the basis for
wisdom, and is closely associated with vipassana. So this appears like a surprising forerunner of
the treatment of wisdom preceding samadhi.

But t h e Suttas typically present the contemporary Brahmans as having fallen away f r o m their
glorious past. It is important to note this context: t h e Suttas do not see the fact that pre-
Buddhists practiced j h a n a as a reason for denigrating and sidelining samadhi practice, but as a
feature by which they could praise the most sublime attainments of the sages of old, thus serving

98
as an example for emulation and inspiration. The following verses were spoken by Venerable
Maha Kaccana to some rude and abusive Brahman youths.

'Those men of old who excelled in virtue

Those Brahmans who recalled the ancient rules
Their sense doors guarded, well protected
Dwelt having vanquished wrath within.
They took delight in Dhamma a n d j h a n a
Those Brahmans who recalled the ancient rules

'But these having fallen, claiming "We recite!"

Puffed up by clan, faring unrighteously,
Overcome by anger, armed with diverse weapons,
They molest both frail and firm.

'For one with sense doors unguarded

[All the vows he undertakes] are in vain.
Just like the wealth a man gains in a dream:
Fasting and sleeping on t h e ground,
Bathing at dawn, [study of] the three Vedas,
Rough hides, matted locks, and dirt,
Hymns, rules and vows, austerities,
Hypocrisy, bent staffs, ablutions:
These emblems of the Brahmans
Are used to increase their worldly gains.

'A mind that is well concentrated,

Clear and free f r o m blemish
Tender towards all living beings -
This is t h e path for attaining Brahma.' 220

Understandably, the Brahman youths were not too pleased with this. So they went to their
teacher, the Brahman Lohicca, and told him. Although he too was displeased, he reflected that h e
should not condemn on mere hearsay, so he visited Venerable Maha Kaccana to discuss the
matter. He asked what the meaning of'sense doors guarded' was.

'Here, Brahman, having seen a visible form with the eye, one is not attracted to a pleasing visible
form and not repelled by a displeasing visible form. One abides having established mindfulness of
the body, with a measureless mind, and understands as it has become that heart-release,
understanding-release, where those evil unskilful qualities cease without r e m a i n d e r . . '

Here again we see t h e connection between pre-Buddhist meditation and mindfulness. The
sequence - sense restraint, mindfulness, samadhi, understanding, release - allows Maha Kaccana
to present the Buddhist ideal as the natural outcome and fulfilment of t h e practices of the
Brahmans of old, so he can skilfully lead Lohicca on in a non-confrontational manner.

LATER BRAHMANICAL SOURCES

Since there are no contemporary records to provide us with a deeper look at these ideas, it seems
we have no choice other than to take the risky path of comparing t h e m with later texts. The
Mahabharata clearly post-dates the Nikayas/Agamas, and evidences Buddhist influence.

99
However, the events are set in a semi-mythical time before the Buddha, and there is no reason to
suppose that it has not preserved some genuine old traditions. Here we find reference to the
'fourfoldjhanayoga'. Only the f i r s t j h a n a , however, is described in detail:

'The mind that is wandering about, with no support, with five gates, wobbling
The steadfast one should concentrate in the firstjhana' 2 2 1

'When the sage enters samadhi of the f i r s t j h a n a in the beginning,

Sustained application (vicara), initial application (vitakka), and seclusion (viveka) arise in him...'222

'Conjoined with that bliss he will delight in the practice of j h a n a

Thus the yogis go to Nirvana that is free of disease.' 2 2 3

The Yoga Sutra of Patanjali (300-500CE?) is one of the earliest presentations of a fairly systematic
path of practice f r o m a non-Buddhist school. The Yoga school, regarded as the practical wing of
the Samkhya philosophy, became one of the six schools of classical Hinduism, which were
orthodox in regarding the Vedic tradition as authoritative, although they differed in
interpretation. The Yoga Sutra is a fairly short work in four chapters, comprised of a series of
brief aphorisms, or sutras, a style which, incidentally, well illustrates the meaning of sutta as
discussed in the GIST. The sutras are often cryptic and as good as incomprehensible without a
commentary; in fact the work as a whole may well be a collection of sayings that was assembled
in the current form by t h e commentator.

Here we merely wish to investigate the meditation terminology, especially in relation to Buddhist
meditation, so we can afford to ignore many of the knotty questions raised by the text and focus
mainly on those passages closest to Buddhism. This methodology will lead to a biased view of the
work as a whole, and it should be borne in mind that the Yoga Sutra remains faithful to its own
distinctive philosophy; it is n o t j u s t a Buddhist rip-off. Doctrinally, it mentions ideas familiar to
the Samkhya/Yoga - the three 'qualities' (gunas) of stimulation (rajas, literally 'desire'),
depression (tamas, 'darkness'), and vitality (sattvas, 'being') that make up our worldly existence,
the fundamental ground of nature (prakrti) f r o m which these evolved, and the individual soul
(purusa), whose purity and clear discernment lead to t h e state of consummation (kaivalya). The
main emphasis is on the practical means, especially meditation, for reaching this state.
Occasionally it critiques Buddhist philosophy. Sutras 4.16-18, for example, assert that it is
impossible for a changing object to be known by one mind-moment (as the abhidhammikas
claimed); the fluctuations of the mind are known due to the changelessness of t h e purusa, the One
Who Knows. Sometimes the text bears on the controversies among t h e Buddhists, such as when it
asserts that 'the past and the f u t u r e exist in their own form', 224 which is reminiscent of the
Sarvastivadin doctrine of time: 'all exists'.

The first chapter of the Yoga Sutra deals with samadhi. It starts with a famous definition: yoga is
the cessation of the fluctuations of the mind. The fluctuations, which are caused by ignorance, are
listed as valid knowledge (pramana, defined in a way similar to the Buddhist epistemologists:
direct experience, inference, and scripture), error, fantasy, sleep, and recollection (mindfulness,
sati). This list is odd; it is difficult to see how, say, direct experience (pratyaksa) could be an
obstacle to samadhi. The treatment of mindfulness in a negative sense is obviously different f r o m
the Buddhist approach. For the Brahmanical schools, the word sati was commonly used in the
sense of 'memorised textual traditions', so it seems that in meditation contexts the meaning of
'memory' was more prominent than 'awareness', hence the negative slant. This situation would
suggest two consequences: first, that when sati is used in a positive sense in the Yoga we should
suspect a Buddhist influence; and second, that the Yoga would need to develop an alternative
terminology to speak about mindfulness within their own system. We shall find evidence in t h e

100
Yoga Sutra to support both of these theses. However, despite this difference in the evaluation, the
Yoga Sutra defines sati the same way as t h e Buddhist schools: the non-forgetting of an
experienced object. After emphasising the necessity for sincere practice and dispassion, the text
goes on to speak of a form of samadhi (the word 'samadhi' is not used, b u t is plausibly supplied by
the commentary) called samprajnata, which it describes as: 'accompanied by initial application,
sustained application, bliss (ananda), [the concept] "I am", and form.' 225 This is virtually identical
with the first of the four Buddhist 'form jhanas'. The idea 'I am' clearly refers to a deluded
perception that takes what is not the True Self, the purusa, to be the True Self. The phrase is
obviously foreign to the standard j h a n a formula, but is similar to one of the deluded forms of
'Nibbana here & now' described in the Brahmajala Sutta:

'When, sir, this self, quite secluded f r o m sensual pleasures, secluded f r o m unskilful qualities,
enters and abides in the f i r s t j h a n a , which has initial & sustained application, and the rapture &
happiness born of seclusion, at that point the self attains Nibbana here & n o w . ' 226

Both contexts are criticising the assumption of self in this state of samadhi; for the Buddhists, of
course, there is no True Self, while for the yoga the True Self is discerned only with more subtle
development of consciousness. The Yoga Sutra goes on to speak of another (higher) form of
samadhi, which is called asamprajnata (although again the terminology is not supplied in t h e
extremely laconic text itself). Sutra 18 describes this as 'preceded by practice in renunciation, and
havingjust a residue of activities (samskarasesa)'.227 Sutra 19 is obscure: 'For the bodiless, absorbed
in fundamental Nature, [such an] existence is conditioned (bhavapratyayo videhaprakrtilayanam)'.
This seems to mean either that this state of consciousness generates a bodiless (videha; = formless,
arupa?) rebirth, or that for one without a body, such a state of consciousness is a natural
condition, not something that must be attained through spiritual practice. Sutra 20 says that 'for
others' (presumably meaning not the 'bodiless' referred to in sutra 19), asamprajnata samadhi is
preceded by 'faith, energy, mindfulness, samadhi, and wisdom'. 228 Here again we meet the
Buddhist five spiritual faculties, which are presumably what is m e a n t by the 'practice in
renunciation' mentioned in sutra 18. Note that sati here is in positive sense, as usual in Buddhism,
and not in negative sense, as earlier in the Yoga Sutra; this supports the argument of Bronkhorst
that this chapter was composed f r o m two sources, one 'orthodox' and one Buddhist. 229 The
samadhi in this group of five, which precedes asamprajnata samadhi, is presumably the
samprajnata samadhi, i.e. form jhana. The asamprajnata samadhi may therefore be plausibly
identified with the Buddhist formless attainments, which are also preceded by form jhana, are the
outcome of a 'gradual cessation of activities', generate a bodiless rebirth, and the highest of which
is called 'an attainment with a residue of activities'. 230 It is very striking that the way of attaining
this asamprajnata samadhi - the five spiritual faculties - is identical with the way of practice
taught by Alara Kalama and Uddaka Ramaputta for attaining formless samadhi, and is also
mentioned in the Parayana Vagga.

The text goes on to speak of various obstacles to samadhi, similar to the hindrances, etc.,
including the tem 'scattered mind' familiar f r o m the Satipatthana Sutta. These result in bodily
and mental discomfort, and unsteadiness of breath, and should be countered by practice of one-
pointedness. Several meditations are recommended that lead to clarity of mind: these include the
Buddhist divine abidings of loving-kindness, compassion, appreciation, and equanimity. Some of
the other meditations, such as breath meditation and the mind free of lust, again remind us of the
Satipatthana Sutta. Next the text speaks of attainments both with initial application (vitakka) and
without; the latter is associated with purity of mindfulness, as in the Buddhist fourth jhana.
Attainments with and without sustained application (vicara), which are said to be subtle
conditions, are also mentioned; just as the Buddhist second jhana, absence of sustained
application is associated with 'inner clarity' (adhyatma prasada). The wisdom of this brings truth.

101
All these states are 'samadhi with seed'; but when even these cease all ceases, and this is 'samadhi
without seed'.

While the first chapter of the Yoga Sutra recalls the Buddhist treatment of samadhi, the second
chapter contains some classically Buddhist instructions on vipassana:

'Ignorance, "I-am-ness", desire, aversion, and insistence (abhinivesa) [are the defilements to be
eliminated by practice]. Ignorance is the cause of the rest, whether they are dormant, weak,
suppressed, or aggravated. Ignorance thinks of t h e permanent as impermanent, of the pure as
impure, of the painful as pleasurable, of t h e not-self as self...'231

The definition of'I-am-ness' is obscure ('taking the two powers of seer and seen as a single self);
evidently it refers to the error of seeking a unified self in the diversity of experience. Desire and
aversion are defined exactly as in Buddhism: the inherent compulsion (anusaya) regarding,
respectively, pleasure and pain. All these 'fluctuations' are to be overcome w i t h j h a n a . The result
of action (karma) rooted in defilement (klesamula) is experienced in pleasant or painful rebirth,
according to whether the causes are good or evil. But for the discerning, all this is suffering.

Halfway through the chapter is introduced the famous 'eight-factored yoga', which is obviously
modelled after the Buddhist eightfold path. A similar sixfold yoga is found in the Buddhist-
influenced MaitrT Upanisad: breath control, sense control (pratyahara), jhana, bearing-in-mind
(dharana), reason (tarka), samadhi. 232 This leaves out the preliminary three practices of the
eightfold yoga and adds 'reason'. The eightfold scheme of the Yoga Sutra, however, was to
become standard. The first factor, yama, is basic ethics similar to the five precepts; the second
factor, niyama, concerns purity, austerity, contentment, chanting, and devotion to God. To
counter thoughts of harming, etc., that are rooted in greed, hatred, and delusion, it is
recommended that one develops the opposite thoughts as antidotes. This is identical with t h e
Buddhist path-factor of right intention. The same principle of opposites is applied not j u s t to
wrong thoughts but to unethical actions as well: 'When one is firm in not stealing, all treasures
appear'. The third factor, posture (asana), is dealt with swiftly, involving merely steadiness,
comfort, and relaxation; no mention is made of the special postures for physical exercise that we
today identify with the word 'yoga'. Next follows breath control and then sense control,
completing the external practices.

The next chapter introduces the 'internal' practices, starting with dharana, defined as 'fixing the
mind on one place'. 233 Dharana, like sati, means 'remembering, bearing in mind', and the
Abhidhamma lists dharana as a synonym of sati. Above we noted the close relation of dharana with
appamada, mirroring the close connection in the suttas between sati and appamada. The change in
terminology f r o m sati to dharana, as I argued above, seems to be because of the different
connotations of the term sati in the two traditions, not because of any difference in the meaning.
Dharana is followed by dhyana (jhana), which is defined very obscurely and, for me,
untranslatably. It seems to mean a realm of mental unification brought about by the practice of
dharana. So both the Yoga and the Buddhist tradition place 'remembering/bearing in
mind/mindfulness' as the practice on w h i c h j h a n a is based. 238

One difference between the two systems is that, while for the Suttas, j h a n a and samadhi are
usually synonymous, the Yoga Sutra places samadhi as the final step of t h e path, followingjhana.
However, dharana, jhana, and samadhi are together said to make up 'restraint' (samyama), so they
are anyway not conceived of as totally separate. The description of samadhi is even more obscure
than jhana: 'The shining forth of just that mere object as if empty of its own form is samadhi'.
Much of the rest of the Yoga Sutra deals with Yoga/Samkhya philosophy and practice, t h e
attainment of various psychic powers, realization of the True Self, and of the disentanglement of

102
the Self from the world and its constituent qualities; the Upanisadic non-dual metaphysic is not
evident.

The above considerations lead me to tentatively conclude the following.

1) There is a thread of Indian meditative tradition referred to in the Nikayas/Agamas, which

stems f r o m the pre-Buddhist period, finds philosophical expression in the Upanisads, and
in the later Yoga texts is developed into a practical method using the sophisticated
psychological terminology developed by the Buddhists.
2) This tradition, through its commitment to memorizing ancient texts (sati = sara), gradually
evolved an appreciation of the benefits of mindful awareness (sati = anupassana).
3) In metaphysics these yogis emphasized the Self, sometimes mystically identified with the
cosmos.
4) This metaphysic was pre-eminently realized in the practice of samadhi, especially
formless attainments.
5) The chief way to develop these formless attainments was to develop the five faculties,
especially mindfulness and form jhana.
6) The Buddha adopted the relevant practical aspects of this tradition into his teaching, his
chief innovation being to not interpret samadhi experience in terms of a metaphysical
'self.

THEJAINS

We t u r n now to the second thread of pre-Buddhist meditation. The classic description here is the
account of the Bodhisatta's austerities. His striving was most terrible: 'crushing mind with mind',
doing t h e 'breathlessjhana' until he felt as if his head was being pierced with a sword or crushed
with a leather strap. But he could not make any progress. Why?

'My energy was roused up and unflagging, my mindfulness was established and unconfused, but
my body was afflicted and not tranquil because I was exhausted by the painful striving. But such
painful feeling as arose in me did not invade my mind and remain.' 235

The Mulasarvastivada account available in Sanskrit confirms the ascription of mindfulness to the
Bodhisatta during his period of striving. 236 Here, 'mindfulness' is obviously used in the sense of
'present m o m e n t awareness' rather than 'memory'. This is confirmed in the following passage:

'Such was my scrupulousness, Sariputta, that I was always mindful in stepping forwards and
stepping backwards. I was full of pity even for [the beings in] a drop of water, thinking: "Let me
not h u r t the tiny creatures in the crevices of the ground." '23=

The reason h e struggled on with such grim self-torture is stated unambiguously:

'Prince, before my enlightenment, while I was still an unenlightened Bodhisatta, I too thought
thus: "Pleasure is not to be gained through pleasure; pleasure is to be gained through pain." '238

This is wrong view, being one of the chief tenets of the Jains. 239 But having starved and tortured
himself near to death because of that view, he reflected thus:

' "Whatever ascetics or Brahmans, p a s t . f u t u r e . a n d present experience painful, racking, piercing

feelings due to exertion, this is the utmost, there is nothing beyond this. But by these racking

103
 austerities I have not attained any truly noble distinction of knowledge & vision beyond h u m a n
principles. Could there be another path to enlightenment?"

'I considered: "I recall that when my father the Sakyan was working, while I was sitting in the
cool shade of a rose-apple tree, quite secluded f r o m sensual pleasures, secluded from unskilful
qualities, I entered and abode in the f i r s t j h a n a , with initial & sustained application [of mind], and
the rapture & happiness born of seclusion. Could that be the path to enlightenment?" Then,
following on that memory came the awareness: "That indeed is the path to enlightenment."

'I thought: "Why am I afraid of that pleasure that has nothing to do with sensual pleasures and
unskilful qualities?" I thought: "I am not afraid of that pleasure, for it has nothing to do with
sensual pleasures and unskilful qualities." ' 2 4 0

Here the friendly, relaxed, reasonable feel stands in refreshing contrast with the steely force of
his earlier efforts. He then decided that he could not a t t a i n j h a n a while so emaciated and must
therefore take some food; we have already seen that the dependence of the mind on food, and
hence the deleterious effects of fasting on one's mind-state, is an Upanisadic idea.241 Although the
Bodhisatta never identifies himself in this period as following any teacher, his practices and views
are identical with the Jains. And it is interesting to note that when the group of five ascetics
abandoned him they went to stay in the 'Rishi's Park' in Benares, where even today there is still a
Jain temple.

Such ideas, however, were not exclusive to the Jains; they were a common heritage of t h e Indian
yogic tradition, and are met with frequently in the early Brahmanical scriptures as well, as Maha
Kaccana's verses above indicate. In fact the Jains were reformists, in that they rejected forms of
asceticism that might harm living beings, and they also laid stress on the proper mental attitude.
Earlier, more primitive, 'professors of self-torture' had believed in the efficacy of the physical
torture itself, irrespective of any mental development. Also, their goal was typically psychic
powers, whereas the Jains aimed at liberation of the soul. Thus the Bodhisatta's austerities do
seem to be closer to the Jains than any other group we know of; the Jains themselves preserve a
tradition that t h e Buddha spent time as a j a i n ascetic.

The implication of this episode is that t h e j a i n system emphasized effort and mindfulness, but not
until the Bodhisatta developed the tranquillity and bliss of samadhi was he able to see t h e truth.
Elsewhere in the Suttas, Mahavira (the leader and reformer of the Jains, known in Pali as
Nigantha Nataputta) is depicted as asserting the impossibility of stopping initial & sustained
application of mind. 242 Thus he would certainly not admit any higher than the f i r s t j h a n a at most.
To me, the Jain teachings and practice have a roughness that does not feel congruent with
samadhi attainments, though this is a purely subjective impression. The Jain sources, to my
knowledge, do not offer a great deal of help here. The earliest Jain sutras speak mainly of ethical
practices, lifestyle, and basic principles, but do not explicitly mention meditation in any
recognizable form. Slightly later we find the following:

'Then having preserved his life, the remainder of his life being but a short period, he stops
activities and enters dry jhana 243 in which only subtle activity remains and f r o m which one does
not fall back. He first stops the activity of mind, then of speech and of body, then h e puts an end
to breathing.' 2 4 4

In Buddhist context this passage would imply the f o u r t h j h a n a ; but we have no guarantee that t h e
terminology is being used in the same sense. The context is different; here we have not j u s t a
meditator, but someone who is culminating a spiritual path by fasting to death. Later texts refer
to familiar ideas such as samadhi, one-pointedness, discriminating insight, reflection on

104
impermanence (anicca), change (viparinama), and ugliness (asubha).245 Dayal says that the Jains
attached great importance to funeral contemplations. 246 There are apparently references to
mindfulness as part of the Jain path, but I do not know what period they belong to. The later
schools developed a list of twelve 'contemplations'. The term used here, anupreksa, is semantically
identical with the term anupassana that is so prominent in the Buddhist practice of satipatthana.
The list is as follows:

1. Impermanence
2. No-refuge
3. Coursing on (in rebirth, samsara)
4. Solitariness (ekatva)
5. Difference (between the soul and the body)
6. Uncleanness (of the body)
7. Influx (of pollutions, asava)
8. Restraint (of kamma)
9. Wearing away (of kamma)
10. The world (as suffering)
11. The difficulty of attaining enlightenment
12. The well-expoundedness of the Dhamma

Some of these are similar to Buddhist contemplations (1, 2, 3, 4, 6, 10, 11, 12), while some are
specifically Jainist in nature (5, 7, 8, 9). They appear to be practices of reflecting on or thinking
over a t h e m e rather than awareness meditations. Most of them fall on the side of vipassana
rather than samatha. T h e j a i n sources also speak of several varieties of jhana':

1. depressive b r o o d i n g j h a n a
2. ferociousjhana
3. d h a m m a j h a n a (contemplation of scriptures; removing afflictions of oneself and others;
kamma and result; samsara and the pure soul)
4. p u r e j h a n a

Only this last might correspond with the Buddhistjhanas, although some of t h e other meanings,
such as 'brooding', are connected with j h a n a or related terms in non-technical passages.
According to Prasad, ' p u r e j h a n a has four kinds:

1. [Manifold, with initial and sustained application]: Absorption in meditation of the Self, but
unconsciously allowing its different attributes to replace one another.
2. [Unified, with initial but without sustained application]: Absorption in one aspect of t h e
Self, with changing the particular aspect concentrated upon.
3. The very fine vibratory movements in the Soul, even when it is deeply absorbed in itself,
in a Kevali [consummate one].
4. Total absorption of the self in itself, steady and undisturbedly fixed without any motion or
vibration whatsoever.' 247

Paul Dundas discusses t h e j a i n attitude towards meditation, and concludes that '...Jainism, unlike
Theravada Buddhism, has never fully developed a culture of true meditative contemplation, no
doubt because early Jain teachings were more concerned with the cessation of mental & physical
activity than with their transformation, and meditation did not lose its original role as little more
than an adjunct to austerity until t h e early medieval period, by which time it had become a
subject of essentially theoretical interest.'

105
CONCLUSION

There is no doubt that the teaching on satipatthana as depicted in the early texts was a
distinctively Buddhist practice. While we have gone to some lengths to unearth elements in
common with non-Buddhist systems, in the final end this re-emphasizes how much was new, in
both the expression and the meaning. The rational, progressive approach, the empirical and
psychological description, the details of the four satipatthanas - none of these can be found in
any straightforward way in any pre-Buddhist texts. Even the post-Buddhist texts, while showing
Buddhist influence in the meditation terminology, did not adopt the satipatthanas as they did t h e
jhanas or the divine abidings.

The early Buddhists were extraordinarily generous in their assessment of the spiritual
attainments of outsiders. They were quite happy to attribute to t h e m such central elements of the
Buddhist meditation system as mindfulness, jhanas, spiritual faculties, enlightenment-factors,
divine abidings, and formless attainments. In this complex weave, we can discern threads of both
samatha and vipassana. Although it is impossible to fully untangle these threads, it seems
possible to discern a different emphasis in the meditative approaches of t h e different schools that
correlates with their philosophical positions.

The Upanisadic tradition espouses a non-dual pantheism. Brahman is t h e ultimate reality, which
creates the world, underlies t h e illusion of diversity, and is immanent in all existence. Thus
existence is inherently good; we already partake of the divine essence, and our spiritual practices
empower us to realize this identity fully. This tradition therefore naturally emphasizes
meditation practices leading to blissful identification with the One; as later traditions summed it
up: 'mind, being, bliss.'

The Jains, on the other hand, have a wholly naturalistic and non-theistic view of existence. The
world is not an illusion; it really exists 'out there', and the ultimate reality is not a pan-theistic
non-dual 'ground of being', but is the countless irreducible atomic monads or 'souls'. Later Jain
theory developed this pluralistic approach into a vastly complex scheme for classifying the
various elemental phenomena, an Aristotelian project like those favoured by the Abhidhamma
schools of Buddhism. Enlightenment consists, not in the mystic identification of the self with the
universe, b u t in the disentanglement of t h e individual soul f r o m the polluting effects of kamma.
They therefore emphasize, as part of their overall strategy of forcibly stopping all activity,
contemplation of the impermanence of the world, and the ability to mindfully endure painful
feelings in order to get free f r o m the defiling influences.

It would therefore seem that the Brahmanical tradition leaned to the side of samatha, while the
Jain tradition leaned to the side of vipassana, each shaping its presentation and emphasis in
accord with its metaphysical predilections. The evidence of the non-Buddhists themselves, as far
as it goes, tends to confirm that the picture painted by the early Suttas of the non-Buddhist
traditions is generally accurate. In the absence of any evidence to the contrary, we can conclude
that the earliest Buddhist traditions accept that both the Brahmanical and the Jain contemplative
traditions included the practice of mindfulness.

I must restate the tentativeness of these conclusions. The real situation was terribly complex;
there was no doubt much borrowing and interchange of ideas, and I have ignored such important
issues as the tenuousness of any links between t h e Yoga and the Upanisads, and the philosophical
similarities between the Samkhya/Yoga and the abhidhamma schools. In fact, it may be held that
it is inappropriate to subject non-Buddhist traditions to a Buddhist analysis rather than simply
presenting t h e m on their own terms. In my defence, my aim here is to seek lines of
continuity/discontinuity between the Buddhist and non-Buddhist traditions, and I am therefore
106
compelled to try to see the other traditions from a Buddhist perspective. Rather than seeing this
analysis as complete or even as a summary, it may be better to view it as a point of departure
against which the real complexity of Indian spiritual culture can be reckoned.

CHAPTER 1 1 : BUILDING BLOCKS

T H E FUNCTION OF SATIPATTHANA

Having conducted a brief overview of the meditative culture within which the Buddha taught, we
can now proceed to consider the teachings on mindfulness in the Buddhist context. In accordance
with the GIST we should start with the earliest statement on mindfulness in the
Dhammacakkappavattana Sutta. This is addressed to the group of five ascetics, and thus explicitly
locates the Buddha's message within the existing spiritual context. It starts by dismissing the
wrong practices of sensuality and self-mortification, then expounds the right way, the noble
eightfold path. This consists of: right view, right intention, right speech, right action, right
livelihood, right effort, right mindfulness, and right samadhi. This formulation of the Buddha's
earliest teaching is preserved in texts of the Theravada, MahTsasaka, Dharmaguptaka, (Mula)
Sarvastivada, and Mahasanghika schools. The factors of the path are not further defined here,
apart from right view, which is implied in the discussion of the four noble truths. The text
therefore suggests that the audience was already familiar with the remaining seven factors.

By listing the factors thus, even without further definition, the text does two important things.
Firstly it specifies which factors are really essential for the goal; and second, it places them in a
sequence implying a conditional relationship between the factors. Elsewhere this relationship is
made explicit. The very first discourse of the Magga-samyutta stresses the causal relationship
between the factors of the path including mindfulness and samadhi: 'For one of right
mindfulness, right samadhi comes to be'. An important definition of 'noble right samadhi', found
in all four Nikayas, also emphasizes that the path factors, culminating in right mindfulness,
function to support samadhi.

'What, monks, is noble right samadhi with its vital conditions, and with its prerequisites? There
are: right view, right intention, right speech, right action, right livelihood, right effort, right
mindfulness. One-pointedness of mind equipped with these seven factors is called noble right
samadhi "with its vital conditions" and also "with its prerequisites".' 248

The same principle is spelled out in BhikkhunT Dhammadinna's analysis of samadhi.

'One-pointedness of mind, friend Visakha, is samadhi. The four satipatthanas are the basis for
samadhi. The four right strivings are the prerequisite of samadhi. The cultivation, development,
and making much of these same principles is the development of samadhi therein.' 24 ?

Or again, in the context of the five spiritual faculties:

107
'It is indeed to be expected, bhante, for a faithful noble disciple whose energy is roused and whose
mindfulness is established that, having made relinquishment the support, he will gain samadhi,
he will gain one-pointedness of mind.' 250

Elsewhere the path is analysed into three - ethics, samadhi, and understanding. If satipatthana
was primarily a vipassana practice, it would of course be included in the understanding section.
But both the Theravada and the Sarvastivada Suttas include satipatthana in the section on
samadhi, never the section on understanding. 2 5 1 All of the basic statements on the function of
satipatthana in the path confirm that its prime role is to support samadhi, that is, jhana.

This can be made clearer by presenting a structural analysis of the truths and the path. This is
parallel to Venerable Sariputta's analysis of the four noble truths in the Maha Hatthipadopama
Sutta. This analysis of t h e path is found in the Saccavibhanga Sutta, also the Maha Satipatthana
Sutta, so it is clearly relevant to satipatthana. The basic definitions are derived f r o m t h e Magga
Samyutta.

1. The noble t r u t h of suffering...origin...cessation...path.

2. And what is the noble t r u t h of the path? Right view...right mindfulness, right samadhi.

3. And what is right mindfulness? One contemplates a body in the

body...feelings...mind...dhammas. What is right samadhi? Quite secluded...one enters the first
j hana... s e c o n d j hana... t h i r d j hana. ..fourth j hana.

The Satipatthana Sutta picks up from here:

'How does one contemplate a body in the body? Here a monk, gone to the forest.establishes
mindfulness. Ever mindful he breaths in, ever mindful he breathes o u t . '

Thus the explanations of the various satipatthanas follow on f r o m the basic definitions of t h e
path. They are less intrinsic, found at a more elaborated stage of the teaching. It would have been
expected that those who were to learn the Satipatthana Sutta would have been familiar with this
basic context. In other words, t h e students would already know that satipatthana is the seventh
of the eight factors of the path, and its function there is to s u p p o r t j h a n a .

There is, however, one important context where mindfulness appears immediately before a
wisdom factor rather than samadhi. This is the seven enlightenment-factors: mindfulness,
investigation of dhammas, energy, rapture, tranquillity, samadhi, equanimity. First we may
notice the obvious fact that, although mindfulness is directly before investigation of dhammas,
both of these ultimately support the range of samatha qualities, which form such a prominent
portion of the enlightenment-factors. We may still wonder why the wisdom factor appears near
the beginning, instead of its normal position towards the end.

I the answer lies in t h e ambiguous usage of both mindfulness and investigation of dhammas in
this context. The enlightenment-factors are presented sometimes in a teaching context,
sometimes in meditation context. Mindfulness and investigation of dhammas are the only factors
whose definitions differ in the two contexts. In a teaching context, we hear of the monk who
hears the teachings, then recollects and remembers that teaching with mindfulness, and t h e n
investigates t h e meaning of the teaching. 252 More meditative contexts speak, in the Theravada,
simply of mindfulness, but the Sarvastivada supplies the expected identification with the four
satipatthanas. 253 Investigation of dhammas, in both versions, is the inquiry into skilful and

108
unskilful dhammas. While the meditation contexts occur far more frequently in the Theravada,
the one teaching-context text in the Theravada is represented by three in the Sarvastivada, and
the same passage forms the basis for the relevant section of the Abhidhamma Vibhanga. It must
therefore be regarded as of considerable importance. In the Bojjhanga-samyutta, it is in this
teaching context alone, never in the meditative context, that the seven enlightenment-factors
are said to arise in a progressive sequence, each dependent on the previous. So when considering
the significance of the sequence of the enlightenment-factors, the primary meanings of the terms
mindfulness and investigation of dhammas should be 'recollection' and 'investigation' into the
teachings. This then inspires the development of samadhi. This, I believe, suffices to explain why
mindfulness in the enlightenment-factors appears directly before the wisdom factor, rather than
the samadhi factors.

As usual, however, matters are not quite so cut-&-dried. In the context of anapanasati, the
sequential arising of the enlightenment factors is also spoken of.254 This context, being a synthesis
of several already-established doctrinal frameworks, is clearly not as fundamental as the usage in
the Bojjhanga-samyutta. The main point of the section is not really to analyse the sequential
origin of t h e enlightenment-factors, but to stress their integration with anapanasati and
satipatthana. Of course, in anapanasati we are in samatha home turf, and vipassana is normally
spoken of as emerging in the final tetrad, that is, coming well after the establishment of
mindfulness and the attaining of samadhi. Thus the sequential arising of the enlightenment-
factors is a bit odd in this context, and is probably no more than an application of the standard
sequence in a derived context, without special significance.

The function of satipatthana as support for j h a n a is suggested in a very common Sutta idiom,
whose significance tends to be obscured in translation. The t e r m satipatthana (establishing of
mindfulness) resolves into sati and upatthana. The alternative resolution into sati and patthana,
though favoured by the commentaries, is spurious. The difference between the two is that while
upatthana expresses the subjective act of establishing or setting up mindfulness, patthana would
refer to the 'foundations' of mindfulness, the objective domains on which mindfulness is
established. The basic meaning of upatthana is to 'stand near', and it is commonly used to mean
'serve', 'approach', even 'worship'. Taranatha Tarkavacaspati's Sanskrit dictionary gives t h e
meaning of 'causing to remember' (especially past lives), which would be identical with sati, but
as this does not seem to be attested in any early text it is probably under Buddhist influence. We
have already remarked that the closest parallel in the Upanisads is the term upasana.

Upatthana occasionally occurs in vipassana contexts, though not, so far as I know, in any central
collection in the Samyutta. In the Anguttara sixes, one is encouraged to 'establish perception of
impermanence [suffering, not-self]' regarding all conditioned activities. 255 Here, although it is not
in the context of satipatthana, we see a similar subjective role for upatthana, with the object in
locative case, as in satipatthana and elsewhere. In the context of satipatthana, upatthana suggests
that one is to make mindfulness stand close by, to be present, to serve t h e meditation. Sati and
upatthapeti stand in t h e same organic relation as do saddha with adhimuccati, or viriya with
arabbhati. These terms, all commonly used in conjunction, indicate a reiterative emphasis. Just as
one 'decides faith' or one 'rouses up energy', so too one 'establishes mindfulness'. In fact, we
could render this phrase 'one does satipatthana', the difference being merely verbal. Because t h e
verb upatthapeti has such an organic relationship with t h e noun sati they are found together in a
variety of settings, just as sati is found everywhere. But by far the most important, common, and
characteristic use is in t h e gradual training, where t h e phrase refers to taking a seated posture for
meditation before the abandoning of the hindrances and enteringjhana.

In the gradual training, sati and upatthana occur together in the common idiom parimukham satim
upatthapeti. Here the t e r m parimukha is one of those simple words that is so hard to interpret. It

109
literally means 'around the mouth'. It is interpreted by t h e Vibhanga as 'at t h e nose tip', while
modern renderings usually use something vague like 'in front'. However the phrase frequently
occurs in contexts outside of anapanasati, making the interpretation 'at the nose-tip', or any
literal spatial interpretation, unlikely. The Sanskrit has a different reading, pratimukha.256 This has
many meanings, among which are 'reflection' and 'presence'. Both of these would be appropriate
in meditative context. But the word usually, as here, occurs in close conjunction with upatthana,
which also means 'presence'. I think here we have another example of that common feature of
Pali or Sanskrit, a conjunction of synonyms for emphasis: literally, 'one makes present a presence
of presence of mind', or more happily, 'one establishes presence of mindfulness'.

The gradual training, appearing many times throughout the early texts, must be regarded as t h e
prime practical and detailed paradigm for the way of practice as conceived in pre-sectarian
Buddhism. In the GIST we have seen how this teaching is fundamental to all the early collections,
especially the DTghas. Here is a schematic representation of the gradual training in relation to a
number of other key teaching frameworks. The ten-fold path is used rather than the eightfold,
since, as can be seen, this correlates more neatly with t h e gradual training. I include a rather
comprehensive array of factors in the gradual training; typically, specific presentations only
contain certain of these.

Table 11.1

The Truths, the Path, and the Training

Four Noble Truths Threefold Training Tenfold Path Gradual Training

Suffering Hearing the teaching
Origin Right view
Cessation Gaining faith
Path
Right intention Going forth
Ethics Right speech, right action Rules of discipline
Right livelihood Purifying livelihood
Samadhi Contentment
Sense restraint
Moderation in eating
Right effort Wakefulness
Clear comprehension
Seclusion
Right mindfulness Establishing mindfulness
Abandoning hindrances
Right samadhi Fourjhanas
Understanding Knowledge & vision
Mind made body
Psychic powers
Right knowledge Divine ear
Knowing other's minds
Recollecting past lives
Divine eye
Four noble truths
Right release Liberation

110
Our main sphere of concern for the present are the middle factors, from contentment to
abandoning hindrances. I give these as falling within the training in samadhi, following t h e Subha
Sutta. 257 However I have not specified their exact relationship with t h e relevant path-factors.
They are a loose grouping of affiliated practices that form a bridge between ethics and samadhi.
Since they involve bodily actions, they might seem to pertain to ethics. Several of t h e m -
contentment, sense restraint, moderation in eating, clear comprehension - relate to a monk's
relation with his alms-food, and thus are connected with right livelihood. They are probably
included within samadhi because they lay primary emphasis on one's mental attitude in various
contexts, and thus form a special training-ground for the serious meditator.

I have correlated these practices with right effort and right mindfulness in a general fashion,
without trying to tie t h e m down too precisely, and do not think that the ambiguity can be fully
resolved. Even right effort and right mindfulness cannot be fully disentangled f r o m each other:
one practicing satipatthana is said to be 'ardent' (with right effort); while one abandoning the
hindrance of sloth & torpor is said to be 'mindful & clearly comprehending'. Similar ambiguities
pervade this section. For example, sense restraint is said to counteract 'evil, unskilful qualities', a
phrase that recurs in the formula for the four right efforts. These bad qualities are f u r t h e r
described as 'covetousness & aversion', the same words that appear in the satipatthana auxiliary
formula. The Sanskrit here adds the word 'for the world', thus increasing the parallel with
satipatthana. It also adds the phrases 'controlled in mindfulness, guarded in mindfulness'
(nipakasmrti, guptasmrti); similar phrases are found elsewhere in t h e Pali. All this simply says that
the path is practiced as a whole, and any division is provisional, useful for ease of exposition and
understanding. Thus I place right effort next to 'wakefulness', and right mindfulness next to
'establishing mindfulness' to indicate their most direct correlation, but without separating t h e m
f r o m the other factors.

This ambiguity provides a ready explanation for how clear comprehension came to be included
within satipatthana. We have already seen several considerations in support of Bronkhorst's view
that we must distinguish between two levels of mindfulness in the account of the gradual
training: the preliminary stage of 'mindfulness in daily life', usually called 'clear comprehension',
and, when the yogi sits down in the forest to meditate, the undertaking of satipatthana proper.
Clear comprehension, like other practices such as sense restraint, wakefulness, etc., involves
mindfulness in its role of preparing for meditation. But because, in the gradual training, clear
comprehension comes close before the 'establishing of mindfulness' in meditation, it would quite
naturally become subsumed under satipatthana as that practice grew in scope and importance.

So it seems that we arejustified in seeing a progressive, causal sequence in the factors of the path
as presented in the Dhammacakkappavattana Sutta. One's understanding of the Dhamma impels
one to renounce in search of peace; one undertakes the rules of conduct and livelihood; applies
oneself to restraint and mindfulness in all activities and postures; resorts to a secluded dwelling;
establishes mindfulness in satipatthana meditation; and develops the four jhanas leading to
liberating insight. This understanding of the path is deeply embedded in the Buddha's conception
of the way spiritual practice unfolds. The more detailed, explicit teachings can be seen as
emerging and spreading out f r o m the concise scheme of the Dhammacakkappavattana Sutta, like
light passing through a pinhole and radiating out.

MINDFULNESS ITSELF

We may now examine mindfulness itself under a closer focus. In this chapter we will examine the
phrases and sayings regarding satipatthana that are found widely distributed among the various

111
collections. In the Theravada canon, mindfulness is described in two stock formulas. The simpler
one emphasizes the older, Brahmanical, meaning of'memory'.

'Here, monks, a noble disciple is mindful, endowed with highest mindfulness and self-control,
able to remember, to keep in memory what was said and done long ago.' 258

This formula does not explicitly treat mindfulness as meditation. It is less closely associated with
satipatthana as such rather than simply 'mindfulness', although this distinction is most tenuous
and artificial. As I have shown elsewhere, the term nepakka, 'self-control', here implies sense
restraint, not wisdom as the commentaries have it.

The Sarvastivada offers a different description of mindfulness.

'When there is mindfulness for, mindfulness against, or having no mindfulness towards

(anything) he is mindful, widely mindful, keeping in mind, not forgetful. This is called right
mindfulness.' 259

It is not immediately apparent what is meant here. Evidently there is some abhidhamma-style
influence in this kind of definition. Perhaps the mysterious first three terms refer to the practice,
which we will meet later, of perceiving the beautiful in the ugly, the ugly in the beautiful, and
avoiding both through equanimity.

We have become so used to the equation of mindfulness = vipassana that it comes as a shock to
realize that the word 'mindfulness' hardly occurs at all in the central vipassana collections. In
fact, it never occurs at all in the Khandha-, Salayatana-, or Nidana-samyuttas in the direct sense
of contemplating impermanence, etc. On the few occasions it appears, it does so in secondary
contexts, typically in the Salayatana-samyutta in its role to assist sense restraint. The primary
function of sense restraint is to reduce sensual lust, so this must be understood as falling on the
side of samatha. On the other hand, 'mindfulness' appears regularly in such straightforward
samatha contexts as t h e j h a n a formula, descriptions of t h e divine abidings and six recollections,
anapanasati, t h e 'establishing of mindfulness' for abandoning hindrances, and so on.

T H E FOURFOLD ESTABLISHING OF MINDFULNESS

The next layer of complexity describes satipatthana as fourfold. In keeping with the pragmatic
and relativist perspective of the Suttas, this is not a definition of mindfulness but a prescription
of how to practice. The standard formulas have both an objective aspect - what to meditate on -
and a subjective aspect - how to approach t h e practice. I will discuss the objective aspect first. All
traditions agree in listing four fundamental objects of satipatthana meditation: body, feelings,
mind, and dhammas. Curiously, these are rarely described in any detail. Only in the various
versions of the Satipatthana Sutta are the meanings specified. Presumably the Satipatthana Sutta
acted then, as it does today, as a key with which the brief texts could be interpreted.

A crucial point here is that the fourfold formula introduces certain specific objects of meditation,
moving towards treating satipatthana as such in a somewhat narrower way than mindfulness in
general. Satipatthana is the only context in t h e main formulations of the path - the wings to
enlightenment, the gradual training, the dependent liberation - to specify the object of
meditation. Generally there tends to be a somewhat curious distance in the Suttas between t h e
subjective and objective sides of meditation. For example, t h e Suttas d e s c r i b e j h a n a in terms of
the subjective mental qualities, and elsewhere describe various meditation objects that are
intended to develop jhana, yet they virtually never speak of, say, 'anapanasatijhana' (but we do
have 'anapanasati samadhi'), or ' k a s i n a j h a n a ' (although there is a slightly dubious reference to

112
'compassionjhana'.) This distance is not comprehensively bridged until the DhammasanganT. So
satipatthana, being thus more 'grounded' and specific, fulfils an important practical function in
the path. The implication seems to be that the particular meditation objects here are an intrinsic
and hence non-optional part of the path. It seems that all meditators must develop at least some
of the satipatthana practices. Meditation subjects outside of the satipatthana scheme are very
frequently taught in the Suttas, notably the divine abidings and the six recollections, but they are
apparently not so essential; however, the feelings, mind states, and dhammas associated with
t h e m may obviously be treated under satipatthana. This crossover 'objective' aspect of
satipatthana makes it somewhat of an odd man out in the 37 wings to enlightenment, and we
shall repeatedly see resulting ambiguities and incongruities emerging in the later attempts to
thoroughly systemize these groups.

Why these four? Later texts of several schools suggest that the four oppose the four perversions.
Contemplation of the body opposes the perversion of seeing beauty in ugliness; contemplation of
feelings opposes the perversion of seeing suffering as pleasure; contemplation of the mind
opposes the perversion of seeing the impermanent as permanent; and contemplation of
dhammas opposes the perversion of seeing self in what is not-self. Certainly they can work in that
way; there is no doubt that, say, the contemplation of the body is oriented towards dispelling
sensual lust and that the meditation of ugliness is an important part of this strategy. And it is
indeed in the survival of the mind, citta, that many, even Buddhists, seek the solace of eternal
bliss. Overall, however, this explanation is post facto and in some respects artificial. It is worth
noting that since this teaching can not be derived from the early Suttas, nor is it intuitively
obvious, it may represent a case of later borrowing between the sects.

A more pertinent consideration in the formulation of satipatthana in terms of these specific four
subjects would seem to be that they progress f r o m coarse to subtle. I will treat this in more detail
below; until we have examined the texts in detail we may be content here with some preliminary
suggestions. The body is mainly treated as the basic object for developing meditation. Feelings
are the most obvious of the mental qualities. The mind, the inner sense of cognition, the
'knowing' rather than t h e 'known', is more subtle, and is properly approached through the first
two. As we shall see, both the treatment of t h e terms themselves, and the correlation with
anapanasati, suggest that a key facet of this progressive refinement of contemplation so far is t h e
undertaking, development, and mastery of jhana. This much is fairly straightforward, and the
traditions are more-or-less in agreement, although they sometimes tend to de-emphasize or
ignore this progressive structure. This is perhaps because they treat the next factor, dhammas, as
meaning various phenomena, many of which are not more subtle than the first three, and thus
disturbing the sequence. I believe this is a mistake. Dhammas here are more usefully considered,
not as a parallel category of phenomena consisting of a miscellaneous grab-bag of left-overs f r o m
the first three, but as being a distinctive and more profound aspect of meditation: the
understanding of the causal principles underlying the development of samadhi.

H O W THE PRACTICE IS DESCRIBED

While the list of the four objects of satipatthanas is common to all traditions, the description of
the subjective aspect differs substantially, even within the basic pericope; however this seems to
be a mere matter of editing preferences rather than sectarian divergence.

The Sarvastivada, in both the Samyutta and Majjhima, is the simplest: one develops 'the
establishing of mindfulness of contemplating a body in the b o d y . ' etc. The Sanskrit versions of
the Dasuttara and SangTti Suttas, which are perhaps also Sarvastivadin, feature a similar formula.
The Mahasanghika Ekayana Sutra seems to be slightly elaborated f r o m a similarly simple formula,

113
although the translations available to me is not precise enough to be really clear; it refers to
meditating on the body to remove unwholesome thoughts and anxiety, and meditating on the
other three in order to gain peace andjoy.

The main Theravada Nikayas do not contain such a simple version. However the later Pali texts
do include similar formulations. The Niddesa, an algebraic abhidhamma-style commentary on the
oldest parts of the Sutta Nipata (Atthaka Vagga, Parayana Vagga, Khaggavisana Sutta) that is one
of the most obscure and little-read corners of the Pali scriptures (and that's saying something!),
includes such a phrase in its standard gloss on 'mindful'. 260 Similar phrases also occur in the
Patisambhidamagga, which is of a similar or later time. 261 By the time of the commentaries it
becomes fairly frequent. The fact that this simple version becomes more frequent in later Pali
works suggests that it might not be original. In this case, the shorter formula might be a later
summary rather than a genuinely early form.

This conclusion is given clear support in SA 612 of t h e Sarvastivada Satipatthana Samyutta. This
discourse has the standard ending, then, added as an obvious editorial gloss, it says:

All sutras on the four satipatthanas are to end with the following phrase, that is: 'Therefore a
monk developing and practicing the four satipatthanas, giving rise to exalted aspirations, with
refined striving and skilful means, with right mindfulness and right knowledge, should train.'

Thus the short formula in the Sarvastivada is merely a function of the abbreviation of the texts,
and the tradition itself states that the abbreviated version should be expanded in every case. This
expanded formula is similar to the standard Theravadin version. They both differ from the
abbreviated form in leaving out t h e t e r m 'satipatthana' f r o m the formula itself, and in adding a
series of terms qualifying the practice. In t h e Theravada version, one contemplates each of the
four objects 'ardent, clearly comprehending, mindful, having removed covetousness & aversion
for the world'. This phrase is found across the schools, and for convenience we may refer to it as
the 'auxiliary formula'. The Sarvastivada version omits the phrase 'having removed covetousness
& aversion for the world'. This is found elsewhere in the Samyutta, 262 however, so its omission
here might be accidental. In speaking of 'exalted aspirations, refined striving, and skilful means',
it seems to expand the simple 'ardent' of the Theravada versions. 'Skilful means' is obviously
reminiscent of the Mahayana; but before that, it was a characteristic of Venerable Upagupta, t h e
patriarch of t h e Sarvastivada, so the difference between the Sarvastivada and Theravada here
might be sectarian. The phrase 'right mindfulness, right knowledge' is common, and obviously
equivalent to the Theravada's 'mindful, clearly comprehending'.

One attempt to interpret these terms is found in the Netti, which says, in general agreement with
the schools, that these qualities correspond with four of the five spiritual faculties: ardent =
energy; clearly comprehending = understanding; mindful = mindfulness; having removed
covetousness & aversion for the world = samadhi. 263 However, the correlation with the spiritual
faculties is not particularly close. For example, the spiritual faculty of understanding (panna) is
defined as 'the understanding of rise and fall that is noble and penetrative, leading to the full
ending of suffering', 264 clear comprehension (sampajanna), although etymologically parallel to
'understanding' is never used in this exalted sense, but is usually restricted to the more mundane
sense of'awareness of activities in daily life', or else it expresses t h e wisdom dimension of jhana.

A better way of seeing these terms is suggested by the Sanskrit Maha Parinirvana Sutta. There is a
well-known episode when a gorgeous courtesan comes to visit. In the Pali, the Buddha urges the
monks to be 'mindful and clearly comprehending'. The Sanskrit expands this a little: 'Monks,
dwell ardent, clearly comprehending, and mindful. AmbapalT the courtesan comes here!' 265 The
texts goes on to define 'ardent' as the four right efforts, 'clearly comprehending' as awareness in

114
daily activities, and 'mindful' as the four satipatthanas. This would seem to correspond exactly to
the usage in the satipatthana formula itself. Now, these practices are all a standard part of the
spiritual path, and are included in or corresponding to elements of the gradual training.

This suggests a connection between the final phrase, 'having removed covetousness and aversion
for the world' with the preparatory phase of the gradual training, especially sense restraint. The
standard passage on sense restraint, which in the Theravada account of the gradual training
usually occurs j u s t before 'clear comprehension', includes the same words 'covetousness &
aversion' (abhijjha-domanassa) that occur in t h e satipatthana formula. 266 In the Sanskrit the
parallel with satipatthana is even clearer, for there the phrase is 'covetousness & aversion for t h e
world' (abhidhya-daurmanasye loke).267 Also, the Sanskrit tends to use the word 'mindfulness' more
frequently here; this mat be seen in both the Sanskrit Sramanyaphala Sutra, and also the
Sravakabhumi.

We should examine a little more closely how t h e key words abhijjha and domanassa are used.
Abhijjha is used in two clearly defined senses. As one of the ten 'pathways of unskilful actions' it
means straightforwardly covetousness: 'Oh! That what belongs to him should be mine!' 268 As an
alternative to kamacchanda ('sensual desire') as the first of the five hindrances, it is much more
subtle, encompassing any desire or interest in sensual experience. In t h e gradual training, both
the Pali and the Sanskrit describe the overcoming of this hindrance as 'abandoning covetousness
for the world', which again is very similar to the satipatthana formula. The term domanassa,
which I render here as 'aversion', usually means 'mental suffering', but in at least one context it
clearly stands for the hindrance of ill-will,269 and this surely must be the meaning in satipatthana,
too. While the use of exactly t h e same phrase abhijjha-domanassa tends to underscore the
closeness of the phrase in satipatthana with sense restraint, I do not see any major significance in
the exact choice of terms for 'ill-will'; the Pali tends to use a number of terms more-or-less
synonymously. While as a hindrance ill-will can be very subtle, in the ten pathways of unskilful
actions, it is defined in very strong terms: 'May these beings be destroyed, killed, and wiped out!'
So it seems that we can consider these two terms, covetousness and aversion, as encompassing a
variety of levels of intensity. The coarse levels are abandoned through the preliminary practices,
especially sense restraint, while the full overcoming occurs with the abandoning of hindrances on
e n t e r i n g j h a n a . It would, therefore, be a little overstrong to insist, as some do, that the phrase
'having removed desire and aversion for the world' implies that one must attain j h a n a before
doing satipatthana; the phrase is simply too vague to bear such a definitive interpretation.

Nevertheless, there is no doubt that the satipatthana auxiliary formula does sound more like
samatha than vipassana. Similar descriptions of the meditative state of mind are not found in
direct vipassana contexts. But let us compare with this description of the practice of loving-
kindness:

Satipatthana Auxiliary formula:

...atapi, sampajano, satima, vineyya loke abhijjhadomanassam.

Ardent, clearly comprehending, mindful, having removed covetousness & aversion for the world.

Loving-kindness: 270
...vigatabhijjho, vigatabyapado, asammulho, sampajano, patissato mettasahagatena cetasa...
Free of covetousness, free of ill will, unconfused, clearly comprehending, mindful, with a heart
full of loving-kindness.
This is simply a slight variation in expression describing a similar subjective process of
meditation. The passage on loving-kindness is obviously referring to jhana, and t h e similarity of

115
the two passages suggests that jhana, rather than being a pre-requisite, is part of the complete
fulfilment of satipatthana.

So the standard Pali phrase may be securely dated before the Sarvastivada schism, and may well
date from the earliest times. The main point of this qualifying phrase is to emphasize that
mindfulness is not developed alone, sufficient unto itself, but in the context of the path as a
whole; and in this all the traditions are in full agreement.

As well as the standard formula, the texts, mainly of the Samyutta, offer a number of interesting
variants. Below I list some of the main variations in the Pali tradition; the Chinese equivalents
are, in the important cases, similar. I retain the Pali to make the parallelism explicit. Several of
these variations, as marked with an arrow, in their original contexts follow on after t h e standard
Pali formula.

1) Kaye kayanupassanasatipatthanam... (Sanskrit, Niddesa, etc.)

The establishing of mindfulness of contemplating a body in the b o d y .

2) Kaye kayanupassT viharati; atapT, sampajano, satima, vineyya loke abhijjhadomanassam... (standard
auxilliary formula)
One abides contemplating a body in the body; ardent, clearly comprehending, mindful, having
removed covetousness & aversion for the w o r l d .

3) > Kaye kayanupassT viharato yo kayasmim chando so pahTyati; chandassa pahanaya amatam
sacchikatam hoti...(SN 47.37)
For one abiding contemplating a body in the body, desire for the body is abandoned; with
abandoning of desire the deathless is w i t n e s s e d .

4) > Kaye kayanupassT viharato kayo parinnato hoti...(SN 47.38)

For one abiding contemplating a body in the body, the body is fully k n o w n .

5) > Kaye kayanupassTviharanto pi upahanet'eva papake akusale dhamme... (SN 54.10/SA 813/Skt)
When one abides contemplating a body in the body, one crushes evil unskilful q u a l i t i e s .

6) > Kaye kayanupassT viharato cittam virajjati vimuccati anupadaya asavehi... (SN 47.11/SA 614)
For one abiding contemplating a body in the body, the mind experiences fading away and release
f r o m the defilements without g r a s p i n g .

7) Kaye kayanupassTviharatha; atapTno, sampajana, ekodibhuta, vippasannacitta, samahita, ekaggacitta...

(SN 47.4/SA 621)
You should abide contemplating a body in the body; ardent, clearly comprehending, unified, with
clear mind, in samadhi, with one-pointed m i n d .

8) > Kaye kayanupassTviharami; atapT, sampajano, satima, 'sukhasmi' ti pajanati... (SN 47.10/SA 615)
I abide contemplating a body in the body; ardent, clearly comprehending, mindful, I understand:
'I am b l i s s f u l ' .

9) > Kaye kayanupassT viharato cittam samadhiyati upakkilesa pahTyanti... (SN 47.8/SA 616)
For one abiding contemplating a body in the body, the mind enters samadhi, the taints are
abandoned.

10) > Kaye kayanupassT viharanto tattha samma samadhiyati samma vippasTdati... (DN 18.26/DA 4)
When one abides contemplating a body in the body, there one gains right samadhi, right c l a r i t y . .

116
11) > Kaye kayanupassi viharahi, ma ca kayupasamhitam vitakkam vitakkesi... (MN 125.24/MA 198)
You should abide contemplating a body in the body, but do not think thoughts connected with
the b o d y . 2 7 1

These variants fall naturally into three groups. The first pair (1 & 2) give the most basic statement
or summary of the practice. Second are the variants that describe the result of the practice in
general terms similar to the description of many other modes of practice (3-6). The remainder (7-
11) describe the actual meditation itself and are more specific to satipatthana. It is striking that
these variants deal explicitly with samadhi. Three (7, 9, 10) mention t h e word samadhi. One of
these (7) reinforces samadhi with the synonyms 'unified, one-pointed mind'. This variant, in
common with one other (10), also has the term vippasada, which is similar to the sampasada of the
second jhana. The mention of 'bliss' (8) also suggests jhana; the context confirms this. The
remaining variant in this group (11) encourages the stilling of vitakka, which obviously pertains to
the secondjhana; and again this is confirmed in the context. Later texts of the schools furnish us
with even more variants on t h e basic formula. Here I will just list a few for the sake of
comparison. The first is virtually identical to t h e standard Pali version; the second and third
merely offer minor elaborations.

12) Adhyatmam kaye kayanupasyi viharaty atapi samprajanah smrtiman viniyabhidhya loke
daurmanasyam. (Sarvastivada Maha Parinirvana Sutra)
Internally one abides contemplating a body in the body; ardent, clearly comprehending, mindful,
having removed covetousness for the world, and a v e r s i o n .

13) Kaye kayanupasyino viharatah kayalambananusmrti tisthati samtisthati. (Abhidharmakosa 342)

For one abiding contemplating a body in t h e body, recollection supported by the body is set up,
well set u p .

14) Kaye kayanupasyino viharatah upasthita smrti bhavaty asammudheti... (Abhidharmakosa 342)
For one abiding contemplating a body in the body, unconfused mindfulness is e s t a b l i s h e d .

15) Kaye kayanudarsi viharati, na ca kayasahagatan vitarkan vitarkayati...

(Pratyutpannabuddhasammukhavasthitasamadhi Sutra 15J, 18B)272
One observes a body in the body, but one does not think thoughts associated with the b o d y . (cp.
Version 11 above)

Thus we find that within the existing variations on the basic satipatthana formula, several refer
quite explicitly to samatha, while none mention the contemplation of impermanence and
causality that is the hallmark of vipassana.

By far the outstanding feature of all the formulas, the definitive statement of what satipatthana
involves, is that one 'abides contemplating a body in the b o d y . (feeling, mind, dhammas)'. Here,
there are two features that demand explanation. One is the repetitive idiom, which I render 'a
body in the body'; the second is the verbal noun anupassana, 'contemplation', which expresses the
mode of awareness characteristic to satipatthana.

' A BODY IN THE BODY'

The reflective idiom 'a body in the body' has been frequently commented on. Some hold that it is
merely idiomatic, with no particular significance; but surely such repetition must at the very

117
least signify emphasis. The normative explanation in recent times, based on the Theravadin
tradition, is that the repetition delineates and defines the object, in particular excluding taking
the body (etc.) to be something that it is not, i.e. a self. This explanation of course appears in a
context that takes it for granted that satipatthana is primarily vipassana. However, it might be
reinterpreted to suit a samatha context as well - not straying outside the bounds of the given
meditation. But none of these interpretations enjoy direct support f r o m the Suttas. To find this,
we must turn to a somewhat cryptic passage found in the Anapanasati Sutta, and in slightly
different form in the Samyutta. This gives a unique set of phrases qualifying each of the four
objects of satipatthana in the context of breath meditation. They are as follows:

(Body)

'I call this a certain body [among the bodies], Ananda, that is, breathing & breathing o u t .

(Feelings)
'I call this a certain feeling [among the feelings], Ananda, that is, close attention to breathing in &
breathing o u t .
(Mind)
'I say, Ananda, that there is no development of samadhi by breathing in & breathing out by one
who is muddled and who lacks clear c o m p r e h e n s i o n .

(Dhammas)
'Having seen with understanding the abandoning of covetousness & aversion, he watches over
closely with equanimity.' 2 7 3

All these raise interpretive issues. It is not good to rely on such problematic passages, but in the
absence of other relevant passages, it seems we have no choice. The sections on feelings and mind
are obscure, and I will not discuss t h e m here. The dhammas section is interesting, but I will defer
a discussion until we consider anapanasati in general.

The first saying, dealing with the body, is quite straightforward. Evidently, the 'breath' is
considered as a kind of 'body', or we might say in English, a kind of physical phenomenon, an
aspect of the body. The bracketed portions appear in the Majjhima version only, not t h e
Samyutta; comparison with t h e existing Chinese and Sanskrit versions might help clarify which is
original. However, they do not substantially alter the meaning. The meditator is to select this
sphere within the entire field of physical experience as the focus of awareness. This is entirely in
keeping with the thesis that satipatthana is primarily a samatha practice. This interpretation is
also in clear accord with the Satipatthana Suttas. The descriptions of all the practices in all
versions speak of a series of contemplations of discrete aspects of t h e given topic. For example,
one contemplates pleasant feeling, then painful feeling, then neutral feeling, and so on. This is a
strong confirmation that we are on the right track: the primary implication of t h e repetitive
idiom in the satipatthana formula is to narrow the focus of attention within each of the four
objects of satipatthana.

'CONTEMPLATION'

The second crucial part of the satipatthana formula, then, is the term anupassana. In the
Abhidhamma this is glossed with the standard register of terms for understanding, which is not
wrong but is not very helpful, as it ignores the subtleties of context. This may be one reason why
later writers have tended to equate anupassana with vipassana, or with non-judgemental

118
awareness; however a closer examination suggests that this does not do either term justice. We
may start with a verse f r o m the Rg Veda.

'He contemplates with loving favor the mortal who, like a rich man, pours for him the Soma.
Maghavan [i.e. Indra, ruler of Gods] in his bended arm supports him: he slays, unasked, the men
who hate devotion.' 274

Here anupassana obviously refers to a strongly one-sided regard, the very opposite of 'non-
judgemental'. A very similar usage is found in the early Buddhist texts, too. The first of the
following verses is f r o m the Sutta Nipata, the second f r o m the Dhammapada.

'He is no friend who is ever diligently

Suspecting dissension, contemplating only flaws;
But on whom one rests, like a child at the breast,
He is a true friend, who is not alienated by others.' 275

'For one who contemplates the faults of another,

Whose thoughts are always critical;
His defilements increase -
He is far f r o m the evaporation of defilements.' 276

The similarity of subject and treatment is striking. The texts strongly emphasize that
'contemplation' (anupassana) here means seeing only (eva) one side of a situation. Moreover, it
means seeing that one side not occasionally but constantly (sada, niccam). The use of 'diligence'
(appamada) in a negative sense is very unusual; normally, of course, diligence is closely connected
with mindfulness.

The Isa Upanisad shows a friendlier face of anupassana.

'He who contemplates all beingsjust as the self

And the self as all beings; he is not revulsed because of that.'

'When all beings a r e j u s t the self for the discerning one

Then what delusion, what sorrow, for one who contemplates oneness?' 277

This could easily be interpreted in Buddhist context as the normal Buddhist practice of universal
loving-kindness. In Upanisadic context, of course, we may assume that the more pregnant
metaphysical sense of 'Self is intended. Here anupassana evidently refers to a mode of
contemplation that sees two sides of things - all beings as the self, the self as all beings - and
resolves this surface duality into a deeper unity. The word 'oneness' (ekatva) here may be
compared with the occurrence of 'one' in the classic description of satipatthana as the 'path going
to one' (ekayana magga). The 'two-into-one' movement finds its Buddhist counterpart in t h e
Dvayatanupassana Sutta of the Sutta Nipata. This text, the longest and most doctrinally
substantial text in the Sutta Nipata, presents a series of'contemplations' arranged in pairs. Many
of these correspond with factors of dependent origination. Here is a typical example.

' "This is suffering, this is the origin of suffering" - this is one contemplation. "This is t h e
cessation of suffering, this is the way of practice leading to the end of suffering" - this is a second
contemplation. Monks, for a monk, rightly contemplating this pair, abiding diligent, ardent, and
resolute, one of two fruits may be expected: profound knowledge in this very life, or, if there is a
remnant, non-return.' 2 7 8

119
Here, just as in t h e l s a , two contrasting contemplations are recommended. But t h e contrast is, of
course, really a complement, so the full realization of these two contemplations leads to one goal
(although in the Dvayatanupassana this is complicated by the offer of non-return as an
alternative to full enlightenment). So although the contexts are quite different, the role of
anupassana remains similar.

The Svetasvatara Upanisad makes the metaphysical usage of anupassana explicit.

'The one controller of t h e many, inactive, who makes the one seed manifold
Standing in the s e l f - the wise who contemplate him, to them belongs eternal bliss, not to
others.' 279

Here, too, the connection between anupassana and the 'one' is prominent. The wise are to
contemplate the one, inactive source, not the multiplicity of appearance; only this leads to
happiness. The earlier Brhadaranyaka also uses anupassana as a contemplation, a beholding, of the
divine.

'If one clearly contemplates him as the self, as God,

As the Lord of what has become and what will be, one does not shrink away from him.' 280

The following verses f r o m the Dhammapada may be considered as a crossover between the early,
non-technical usage of anupassana and the more specialized sense in satipatthana.

'One who abides contemplating beauty,

Unrestrained in his sense faculties,
Immoderate in eating,
Lazy, with deficient energy;
Truly Mara overthrows him
As the wind, a weak tree.

'One who abides contemplating ugliness,

Restrained in his sense faculties,
Moderate in eating,
Faithful, with roused-up energy;
Truly Mara does not overthrow him
As the wind, a rocky mountain.' 281

Here one is depicted as contemplating either beauty or ugliness (subha, asubha); again, one is
focusing o n j u s t one aspect of things, deliberately ignoring the other side.

Anupassana in such contexts is clearly oriented towards eliminating sensual lust, and hence falls
primarily on the side of samatha. Anupassana does occur occasionally in vipassana contexts, too.
For example, there are a few discourses in the Khandha-samyutta that speak of'contemplation' of
impermanence, suffering, and not-self. 282 Elsewhere we read of 'contemplation of impermanence',
etc. especially regarding feelings. 283 The proper role of 'contemplation of impermanence' in
satipatthana is in the fourth satipatthana, as is clear from anapanasati.

So anupassana may be used in both samatha and vipassana contexts, and cannot be exclusively
categorized as either. Although related to the word 'vipassana', anupassana is not used when
standing alone, as vipassana is, to specifically denote t h e meditative enquiry into impermanence
and causality. The prefix anu- suggests 'following, conforming, after', and lacks the analytical
implication of vi-. It could be interpreted here as implying a mode of contemplation that

120
'conforms' to t h e relevant context; thus anupassana is normally t h e second m e m b e r of a
c o m p o u n d w h e r e t h e first m e m b e r defines t h e specific subject of meditation: 'contemplation
o f . ' . In psychological contexts anu- commonly carries t h e n u a n c e of 'continuing'. Thus vitakketi
means 'to think'; anuvitakketi means 'to keep on thinking'. The same usage occurs in t h e definition
of sati as ' m e m o r y ' t h a t we have e n c o u n t e r e d above. There two t e r m s are used: sara and anussara,
which we should u n d e r s t a n d as 'remembers, keeps in memory.' A similar n u a n c e is evident in two
of t h e t e r m s used in t h e A b h i d h a m m a gloss for t h e j h a n a factor vicara - anusandhanata and
anupekkhanata - which should be translated 'sustained application, sustained observation'.
Anupassana is semantically parallel with anupekkhanata, and so also suggests 'sustained
observation'. This sustained, continuous aspect of anupassana is clearly emphasized in t h e verses
we examined above. The c o m m e n t a r y on t h e Visuddhimagga comments on this word i n j u s t this
way: 'he keeps re-seeing (anu anu passati) w i t h j h a n a knowledge and insight knowledge'. 284

CLEAR COMPREHENSION

The t e r m sampajanna, which occurs in some of t h e satipatthana formulas and commonly

elsewhere together w i t h mindfulness, is also glossed by t h e A b h i d h a m m a with t h e standard
register of synonyms f o r wisdom, and has b e e n equated by some writers with vipassana. But
although sampajanna is etymologically equivalent to panna 'understanding', it is n o t explicitly
equated with vipassana. It is indeed used occasionally in t h e sense of vipassana; we will examine
this in t h e context of t h e Satipatthana-samyutta below. But sampajanna is most characteristically
used in t h e context o f ' d a i l y life awareness' as a preparation f o r j h a n a . Both t h e Suttas and t h e
Abhidhamma include this practice u n d e r samadhi. In t h e Subha Sutta, it is listed u n d e r t h e
'aggregate of samadhi'; and in t h e A b h i d h a m m a it finds its place in t h e j h a n a Vibhanga. Nowhere
in t h e Suttas or A b h i d h a m m a is t h e awareness of activities t r e a t e d primarily as vipassana.

The other c o m m o n usage of sampajanna is in t h e f o r m u l a for t h e t h i r d j h a n a , w h e r e it expresses

t h e wisdom dimension of samadhi. Most of t h e exercises of t h e Satipatthana Sutta use t h e verb
f o r m of panna in a similar sense, for example: 'One u n d e r s t a n d s "I am b r e a t h i n g in a long
breath"...One u n d e r s t a n d s "I am standing"...One u n d e r s t a n d s "I am experiencing a pleasant
f e e l i n g " . ' and so on. So t h e r e is clearly this dimension of u n d e r s t a n d i n g in t h e sense of clear
awareness t h r o u g h o u t t h e meditation. We find t h a t j h a n a s too can be qualified by such terms;
s o m e t i m e s j h a n a is classified u n d e r wisdom, or one i n j h a n a is said to 'know & see', and so on. The
m e r e usage of these t e r m s cannot m e a n t h a t this is a vipassana practice as distinguished f r o m a
s a m a t h a practice. Reality is m o r e subtle: all meditation m u s t include b o t h peace and wisdom. The
question is: w h a t is t h e context, how are these qualities being applied here? The contexts w e have
seen above suggest t h a t t h e primary p u r p o s e of satipatthana is t h e development of samadhi, and
t h e r e is n o t h i n g h e r e to change t h a t conclusion. All we can rightly conclude is t h a t t h e
development ofj h a n a involves a dimension of wisdom.

INTERNAL/EXTERNAL

An interesting aspect of t h e satipatthana teachings is t h a t we are encouraged to practice

internally, externally, and internally-externally. These i n t e r n a l / e x t e r n a l modes of satipatthana
are attested in all t h e Nikayas, Agamas, and Abhidhammas, and obviously r a n k as of some
importance. We shall m e e t t h e m time and time again in t h e later expositions. There is a slight
difference between t h e Theravada and Sarvastivada Satipatthana Suttas and most o t h e r
presentations. These Suttas p r e s e n t t h e basic f o r m u l a first, t h e n add t h e i n t e r n a l / e x t e r n a l
contemplation as p a r t of t h e refrain following t h e detailed explanation. But usually t h e
i n t e r n a l / e x t e r n a l contemplation is integrated into t h e f o r m u l a f r o m t h e start: 'One contemplates
121
a body in the body internally.'. We can call this the 'integrated internal/external formula'. This
is found in the Pali sources apart f r o m the Satipatthana Sutta itself, such as the Samyutta, t h e
DTgha, and the Vibhanga; also in the Sarvastivada Samyukta, 285 Ekayana Sutra, Dharmaskandha,
and Sariputrabhidharma; also fairly consistently in later sources such as the Avatamsaka Sutra,
Abhidharmasamuccaya, and Arthaviniscaya Sutra. Obviously there is no serious difference in the
meaning, but such innocuous details are all t h e more useful in tracing textual affiliations and
editorial fiddling.

There is, on the face of it, an interpretive problem here, for we normally understand that
meditation is the 'inward path', and yet there is a universal agreement that satipatthana should
be practiced externally also. In casual reference to mindfulness practice, such as in verse, we are
frequently encouraged to direct mindfulness inwards, and there is no doubt that this is the prime
focus for mindfulness meditation. For example, one passage reminds us that: 'For one well
established in internal presence of anapanasati, there are no disturbing thoughts drifting
outside.' 286 Yet the external aspect remains - how are we to understand this?

In general Sutta usage, the terminology of internal and external is found both in the context of
both samatha and vipassana. As a samatha example, take the eight liberations or t h e eight bases
of transcendence: 'Perceiving form internally, one sees forms externally, limited, fair and ugly; by
transcending them one perceives thus: "I know, I see". This is the first base of transcendence.' 287
While the phrasing is obscure, it is clear that such passages refer to the development of some
visualization as an object of samadhi, such as is known in the later works as a 'nimitta'. I think
'perceiving form internally' refers to imagining a part of the body as the initial stage in
developing a true samadhi nimitta. This would therefore come within t h e sphere of satipatthana.
As a vipassana example, we need look no f u r t h e r than the second sermon, the Discourse on Not-
Self, where the Buddha speaks of each of the five aggregates as 'internal or external'. This passage
is of course one of the fundamental vipassana pericopes.

In the specific context of satipatthana we find a passage in the DTgha Nikaya: through 'internal'
contemplation one enters samadhi, then gives rise to knowledge & vision (i.e. psychic vision) of
the body, etc., of others externally. 288 Within the Satipatthana Sutta, internal and external
contemplation emerges most naturally in the charnel ground contemplations, which are depicted
as imaginative exercises: 'As if one would see a corpse discarded in a charnel ground, one, two, or
three days dead, bloated, livid, and oozing matter, a monk compares this very body with it thus:
"This body too is of that same nature, it will be like that, it is not exempt f r o m that fate." '289

Mention should also be made of the Vijaya Sutta of the Sutta Nipata. This little poem compresses
most of the Suttas' body contemplations into a few verses. Although it does not specifically
mention satipatthana, it stands as the main teaching on body contemplation in the Sutta Nipata,
and might have been composed for that very reason. If so, it would not be very early. After
describing some charnel ground meditations, it says:

'As this is, so is that; as that is, so is this -

Internally and externally, one should dispel desire for t h e body.' 290

The internal/external contemplation of the elements is treated in detail in the Maha

Hatthipadopama Sutta. For example, the internal water element is defined as the watery parts of
one's own body, such as sweat and blood, etc., and the external water element is t h e waters in
floods, the ocean, etc. Both internal and external water element are merely water element, and
should be seen rightly as not-self and impermanent. This passage might be taken as implying that
the third stage, where internal and external are combined, should be seen as a synthesis where
the difference between the inner and outer is surmounted.

122
Interestingly, t h e t r e a t m e n t of impermanence of the external water element, as too the other
elements, speaks of the destruction of the earth at t h e end of the universe, which is most
emphatically not a 'momentary' conception of impermanence. This seems to extend t h e
application of 'external' n o t j u s t to the 'there' but also to the 'then', outside in both space and
time. Similar notions are preserved in the Sarvastivadin Abhidharma; we shall see below that
while the Theravada treats internal purely as 'pertaining to oneself and external as 'pertaining to
others', the Sarvastivada Dharmaskandha also refers to past and f u t u r e lives. This usage might
ultimately derive from the Samyukta. We have seen above that one of the Sarvastivada Samyukta
discourses adds an editorial gloss, blatantly inserted following the usual ending. This happens
elsewhere, too. One such example occurs in a discourse called 'Development'. 291 In the
Sarvastivadin version, 'development' refers to the practice of satipatthana internally and
externally, this being one of the few occurrences of this formula in the Samyutta. Tacked on after
the end of this discourse is the sentence:

'The development of the four satipatthanas in t h e past and f u t u r e are also taught in this way.'

This must be one of the most obvious sectarian interpolations in the Agamas. It establishes a clear
connection between 'internal/external' and 'past/future'.

Here we engage with an interesting aspect of Buddhist, or more generally Indian, thought, the
correlation between t h e personal and the cosmic. In the Upanisads this is most famously
expressed in the identity of Atman and Brahman, the individual soul and the world-spirit being in
essence the same. Parallel ideas are found in Buddhism, for example, in the correlation between
different stage of j h a n a and the different realms of rebirth that they produce. One of the clearest
explanations of this is by Venerable Anuruddha:

'Suppose an oil-lamp is burning with an impure wick; because of the impurity of the oil and the
wick it burns dimly. So too, when a monk resolved upon and pervading a defiled radiance his
bodily disturbance has not fully subsided, his sloth & torpor have not been fully eliminated, his
restlessness & remorse have not been fully removed. Because of this he practices jhana, as it
were, dimly. With the break-up of the body, after death, he reappears in the company of the Gods
of Defiled Radiance.' 292

While satipatthana is not mentioned here, we shall see later that Venerable Anuruddha was
renowned for his emphasis on satipatthana, so the connection should not be ruled out. Peter
Masefield has discussed some of the relations between the Buddhist and Upanisadic texts, and
says that the usage of'internal' (ajjhattam) is identical in both contexts, while the Buddhist usage
of'external' (bahiddha) embraces both the 'cosmic' (adhidaivatam) and the 'objective' (adhibhutam)
of the Upanisads. 293

Another interesting indication for t h e meaning of internal/external is in the phrasing of the

contemplation of mind in the Satipatthana Sutta: 'One understands mind with lust as "Mind with
l u s t " . ' etc. The phrasing is identical with the psychic power of being able to read the mind of
another person; this also correlates with the claim by Venerable Anuruddha that his own
development of such powers was due to satipatthana. The passage on the psychic powers appears
often in the gradual training, whereas the passage on knowing the mind in satipatthana appears
just once. Surprisingly, then the passage on psychic powers should be regarded as the basic
source context.

123
' T H E P A T H OF CONVERGENCE'

Finally we come to that most definitive of satipatthana slogans: ekayana magga. The basic
derivation is simple: eka means 'one'; ayana means 'going'; and magga means 'path'. But f u r t h e r
interpretation is difficult for a number of reasons: the word eka is used in many different senses;
the exact grammatical relationship between eka and ayana is not clear; and idiomatic uses of the
phrase are few a n d / o r obscure. The commentators offer many different interpretations, which I
will not repeat here, since they have been treated often enough before. 298 The commentators are
concerned to expand the meaning, which is fine - especially when used, as often appears to be
the intention, as raw material for oral instructions - but it is not very useful for we who wish to
pin down the original denotation as precisely as possible. An unfortunate result of such
vagueness is that terms can be usurped for polemic purposes. Standard renderings of ekayana as
'the one and only way' tell us more about the biases of the translators than about the meaning of
the Pali.

The Chinese renderings merely confirm that the Chinese translators were likewise uncertain: the
Sarvastivada Smrtyupasthana Sutra has 'there is one path' (Will), SA has 'one vehicle path' (
= eka + yana, evidently t h e translator used this more familiar Mahayana term); SA2102
has 'there is only one path' ( ^ W i l l ) ; while the Mahasanghika Ekayana Sutra has 'there is one
entrance path' ( W A ill = eka + ayatana). The uncertainty of the traditional interpretations
suggests that the term may well be old, and may well derive f r o m a context that the later writers
were unfmiliar with. In the Theravada tradition, it is not unusual to find the commentators
struggling to explain terms and ideas that are strongly embedded in the Indian cultural
background.

Gethin includes an interesting discussion. 295 He cautions against any attempt to settle on a single
concrete definition for such a term, which early on seemed to carry spiritual/mystical
connotations, and is used in a variety of senses in the Brahmanical scriptures. The Chandogya
Upanisad lists the 'Ekayana' as an ancient Brahmanical text, which according to the commentator
dealt with mtisastra, 'social ethics' or 'politics'; perhaps t h e idea here is that social policy leads to a
unified society. Gethin notes that the non-Buddhist contexts for ekayana suggest two groups of
meanings: the 'lonely' or 'solo' way; and a way that leads to one, a convergence point. 'Solo way' is
accepted by the commentators, but one could question whether this was a suitable
interpretation, in the light of several texts in the Samyutta that encourage one to develop
satipatthana also for the benefit of others. Further, this meaning always seems to occur in literal
contexts, not with a derived significance appropriate for meditation. Moreover, only the second
meaning, which also claims commentarial support, is explicitly found elsewhere in the early
Nikayas.

This context, found in t h e Majjhima Nikaya, suggests a meaning of 'leads to one place only'. 297
Since Nanamoli, translators such as Bhikkhu Bodhi have relied on this as the only straightforward
contextual meaning available in Pali, and have rendered t h e term the 'direct way' or the 'one-way
path'. While this interpretation is a marked improvement over the polemicism of 'one and only
way', I am still not convinced that it captures the full resonance of the term. It derives from a
literal context that has nothing to do with meditation, and needs evaluation in light of the
philosophical/meditative usages, which are more relevant in satipatthana. Since the Brahmanical
usages of this term clearly encompass a wide spectrum of meaning, there is no reason to assume
the Pali means the same thing in different contexts. For this reason it may be worthwhile to
examine more closely the Brahmanical references. We may start by asking: just how relevant are
the Brahmanical contexts here?

124
Before we can answer this question clearly, we need to straighten out the occurrences of the
phrase in the existing Samyuttas. Luckily, relevant texts are present not only in the two complete
Samyuttas, but also the two partial Samyuttas. All the existing versions follow the ekayana phrase
by saying this is 'for the purification of b e i n g s . ' and so on. Most of the existing versions place the
statement in the period immediately after the Buddha's enlightenment, when the Buddha
reflected on satipatthana as t h e ekayana magga.297 There are three versions of this passage in the
various versions of the Samyukta Agama in Chinese. 298 All of these are located in the Sagatha
Vagga due to the presence of the verses.

These versions are, allowing for translation issues, very close, except SA3 has less verses. They all
follow the ekayana passage by saying that one without the four satipatthanas is without t h e noble
Dhamma, one without the noble Dhamma is without the noble path, and one without the noble
path is without the deathless; but if one is with the four satipatthanas, then all the reverse is true.
Then Brahma, understanding the thought in the Buddha's mind, vanishes f r o m the Brahma realm
and reappears before t h e Buddha. He applauds the Buddha's thought, repeats it, and adds verses
in praise of the ekayana path. The Chinese versions have some beautiful images, with two of t h e m
saying the path is like 'a river flowing down to the ocean of sweet dew' ('sweet dew' being the
standard Chinese rendering of 'deathless'). 299 So we have three main textual components: the
ekayana passage; the 'without' passage; and the verses. 'Ekayana' and 'without' originate with the
Buddha and echoed by Brahma, while t h e verses belong to Brahma alone.

Two of the Pali versions are quite similar, the main differences being that they lack the 'without'
passage, and the verses are shorter. 300 These feature a unique variation in the satipatthana
pericope; instead of the normal 'a monk abides contemplating a body in the body' (bhikkhu kaye
kayanupassT viharati.) we have 'should a monk abide contemplating a body in the body' (kaye va
bhikkhu kayanupassTvihareyya...). In other words, the mood shifts to optative, presumably because
at the time these words were spoken there were no monks yet. Even though these are geyyas, they
appear in the Theravada Satipatthana-samyutta; the fact that all the Chinese versions are in the
Sagatha Vagga makes us suspect the Pali tradition shifted t h e m at a later date.

The Chinese has another similar passage, but this is in the Anuruddha-samyutta. Here Venerable
Anuruddha takes t h e part of the Buddha, reflecting in seclusion about t h e four satipatthanas as
the ekayana path - one without these is without the noble Dhamma, and so on, all as the Sagatha
Vagga versions. Venerable Maha Moggallana takes the part of Brahma, reading Anuruddha's
mind, vanishing and 'swift as a strong man might bend his arm' reappearing in f r o n t of
Anuruddha, applauding and repeating his thought. In place of the verses, Moggallana asks
Anuruddha how one delights in t h e four satipatthanas. Anuruddha replies with the normal
pericope, but adds 'with stilling & quietude, cultivating oneness of mind'. 301 This seems to be
restating in psychological terms what the Sagatha Vagga versions state in mythic terms, since
Brahma is of course regarded in Buddhism as having attained his exalted status and special
powers through the practice ofjhana. The text listed as cognate by Akanuma here, SN 52.1, in fact
shares none of the special features of this text except for the setting. Instead, it mentions the
internal/external refrain, the rising and falling refrain, and perceiving the repulsive in the
unrepulsive (etc.). Thus it replaces t h e samadhi emphasis of the Sarvastivada version with
vipassana.

There is, then, a strong unity in the presentation of the ekayana passage that we have seen so far.
The two main doctrinal units ('ekayana' and 'without') occur together in all the Chinese versions,
and always in a context that suggests psychic powers and samadhi. The doctrinal passages also
occur in the Sarvastivada Smrtyupasthana-samyukta. SA 607 presents the ekayana passage as a
simple sutta at SavatthT, while the next sutta, SA 608, presents the 'without' passage. Considering
these together, they give the distinct impression that t h e original Sagatha Vagga discourse was

125
taken up, shorn of the special setting and split into two discourses for inclusion in the
Smrtyupasthana-samyukta.

The Pali seems to have gone one step further. Now, the ekayana passage occurs in the first
discourse of the collection. The beginning of a collection is always particularly susceptible to
editorial f u n n y business, as it is so easy to slip in a new discourse. Perhaps the occurrence of the
word 'one' suggested placing this discourse here. The setting is given at AmbapalT's Mango Grove.
This seems implausible, as this setting is related to the relevant passage in the Maha Parinibbana
Sutta, and there is no sign in t h e versions of this that the ekayana passage belongs here. Rather,
they have the Buddha encouraging the monks to be mindful since AmbapalT is coming. This
passage in fact comprises the second text of the Theravada Satipatthana-samyutta, immediately
after the ekayana discourse. It seems as if the ekayana discourse has mistakenly copied the setting
f r o m the following discourse. In fact, this could be taken as evidence for the suggestion I made
above that the ekayana discourse was a subsequent insertion at the beginning of the collection. If
it had been, as in the Sarvastivada, originally situated later in t h e collection, once the original
setting was lost it would have been assumed that the setting was at SavatthT. But this, as so often,
may have been left implicit and not literally spelled out. Thus when the discourse was cut-&-
pasted into the beginning of the collection, it lacked a meaningful setting, and was simply
inserted after the setting of the previous first discourse, which became number two.

The 'without' passage is similarly problematic. There is no real corresponding passage in t h e

Theravada, but perhaps SN 47.33 'Neglected' might be regarded as a cognate. However, this j u s t
says one who has neglected satipatthana has neglected the path leading to the end of suffering
(and the reverse), lacking the progressive series of Dhammas found in all five Chinese versions. If
it does come from the same source, then, it has clearly suffered decay.

To sum up: the ekayana passage normally occurs in Brahmanical context. This is attested in all
existing versions of t h e Samyutta. These feature Brahma in a setting similar to the 'Request of
Brahma', which the GIST identifies as the root text of the geyya anga. The other versions appear to
be derived f r o m this usage. It is noteworthy that the f u r t h e r the texts drift f r o m the standard
version, the greater the number of editing/textual problems they evidence. This is a good hint
that we are on the right track. The conclusion of all this is very simple: we have every reason to
believe, from the evidence of t h e Buddhist sources, that the term ekayana was associated
primarily, if not exclusively, with the Brahmanical context.

We may then proceed to look at the matter f r o m the reverse perspective: do the contexts in
which the term ekayana appear in the Brahmanical scriptures suggest a connection with
Buddhism? The key passage is in the Brhadaranyaka Upanisad. We have already remarked that
the Brhadaranyaka is generally regarded as probably the oldest and most important of t h e
Upanisads, and this passage in particular isjustly famous, perhaps the most significant dialogue
in the whole Upanisadic tradition. Gethin has already referred to this source, but his summary
treatment does not bring out the full significance. 302 The relevant passage occurs twice in the
Upanisad. Occasionally the readings differ; generally I follow the first occurrence. (The second
occurrence adds material, especially emphasizing the immortality of the Self, and includes the
famous description of the Self as: 'Not this! Not this!' We will look at an important variation
between the two below.)

1. 'MaitreyT', said Yajnavalkya, 'I am about to go forth f r o m this state (of householder). Look, let
me make a final settlement between you and [my other wife] KatyayanT.'

126
 2. Then MaitreyT said: 'If, indeed, Venerable Sir, this whole earth filled with wealth were mine,
would that make me immortal?' 'No,' said Yajnavalkya, 'Your life would be then like that of t h e
rich. But there is no hope of immortality through wealth.'

3. Then MaitreyT said: 'What should I do with that by which I do not become immortal? Tell me,
Sir, what you know.'

4. Then Yajnavalkya said: 'Ah, dear, you have been dear, and now you speak dear words. Come, sit,
I will explain to you. Reflect on what I say.'

5. Then he said: 'Truly, not for t h e sake of t h e husband is the husband dear, but for the sake of the
Self. Not for the sake of the wife... sons... w e a l t h . Brahmanhood... Ksatriya-hood. w o r l d s , g o d s .
b e i n g s , not for the sake of all is the all dear, but all is dear for the sake of the Self. Truly,
MaitreyT, it is the Self that should be seen, heard, considered, and reflected on. By the seeing,
hearing, considering, and cognizance of the Self, all this is known.

6. 'The Brahman ignores one who knows him as different f r o m the Self. The K s a t r i y a . w o r l d s .
g o d s . b e i n g s . all ignores one who knows it as different f r o m the Self. This Brahman, this
Ksatriya, these worlds, these gods, these beings, and this all are this Self.

7, 8, 9. 'As when a drum is b e a t e n . a conch is b l o w n . a lute is played, one is not able to grasp the
external sounds, but by grasping t h e lute or the lute-player the sound is grasped.

10. 'As from a lighted fire laid with damp fuel various smokes issue forth, even so, my dear, the Rg
Veda, the Saman Veda, the Yajur Veda, the Atharva, Angirasa, histories, legends, sciences,
Upanisads, verses, sutras, explanations, and commentaries are all breathed forth f r o m this.

11. 'As the ocean is the convergence (ekayana) of all waters, as the skin is t h e convergence of all
touches, as the nose is t h e convergence of all smells, as the tongue is the convergence of all
tastes, as the eye is the convergence of all forms, as the ear is the convergence of all sounds, as
the mind is the convergence of all thoughts (samkalpa), as the heart is the convergence of all
realizations (vidya), as the hands are t h e convergence of all actions, as the genitals are the
convergence of all pleasure, as the anus is the convergence of all excretion, as the feet are the
convergence of all movements, as speech is the convergence of all Vedas.

12. 'Just as a lump of salt thrown in water becomes dissolved in the water and there would not be
any to be seized by the hand, but wherever you might take it up it is salty; so too this great being,
infinite, with nothing beyond it, is a sheer mass of cognition (vijnana-ghana eva). Having arisen out
of these elements, one perishes back into them. When departed, there is no more perception
(samjna). This is what I say, my dear.' So said Yajnavalkya.

13. Then said MaitreyT: 'In this, indeed, you have confused me, Venerable Sir, by saying that
"when departed, there is no more perception".' Then Yajnavalkya said: 'Certainly I am not saying
anything confusing. This is enough for cognizance.'

14. 'For where there is duality, as it were, there one smells another, one sees another, one hears
another, one addresses another, one thinks of another, one cognizes another. But where All has
become identical with Self, then by what and whom should one s m e l l . s e e . h e a r . s p e a k .
t h i n k . by what and whom should one cognize? That by which all this is cognized, by what is that
to be cognized? By what, my dear, should one cognize the cognizer?' 303

127
We should first note that the time of the great sage Yajnavalkya cannot have been too far distant
f r o m the Buddha. The Brahmanical grammarian Katyayana, commenting on Panini, says that the
Yajnavalkya scriptures are 'of too recent origin; that is to say, they are almost contemporaneous
with ourselves'. 304 Panini lived no earlier than the 4 th Century BCE, and Katyayana later still; in
other words, they are later than the Buddha, but 'almost contemporaneous' with Yajnavalkya.
Thus, Yajnavalkya might be plausibly placed in the period immediately before the Buddha. In
fact, Yajnavalkya's favorite interlocuter is King Janaka, a historical character mentioned in the
Jatakas, whose city is still called 'Janakpur' today. Since he appears in t h e Jatakas he is likely to be
pre-Buddhist, and is likely to have been a king in the Mithila region a few generations before the
Buddha. So we are justified in considering Yajnavalkya's teachings as forming part of the
immediate religious context of the Buddha.

Furthermore, virtually every aspect of this classic dialogue has close connections with the Suttas,
with the obvious exception of t h e metaphysical 'Self doctrine. Yajnavalkya's aspiration to go
forth (1) betrays influence f r o m the samana traditions such as Buddhism. The sentiment of the
futility of wealth in the face of death (2, 3) is classic Buddhism, but alien to the older Vedic
tradition. The statement that the husband, etc., is dear for the sake of t h e Self (5) is similar to a
passage in the Sagatha Vagga, as pointed out by Bhikkhu Bodhi.305 The list of seen, heard, thought,
and cognized (5) is presented in the same way in the Suttas. 306 The list of various principles
culminating in the 'all' (5, 6) is reminiscent of the Mulapariyaya Sutta, a text which the
commentarial tradition associates with a Brahmanical context. 307 The similes of the drum, conch,
and lute (7, 8, 9) all appear in the Suttas. The similes of the conch 308 and the lute, 309 in particular,
are treated in a strikingly similar way: the sound manifests f r o m the coming together of the parts
and the appropriate effort and cannot be grasped of itself. The image of the rivers flowing into
the sea (11) occurs in the Kayagatasati Sutta, and even more strikingly, occur in the verses
accompanying the ekayana passage in two of the Chinese versions. The simile of salt (12) reminds
us that the Buddha said the great ocean h a s j u s t one taste, the taste of salt, just as the Dhamma-
Vinaya has the taste of liberation. 310 The description of cognition as infinite (12) occurs in the
standard formula on t h e formless attainments. The bewilderment of one trying to grapple with
the subtle state of the realized sage after death (13) also recurs in the Suttas, in similar terms. 311
There have even been several attempts by modern scholars to identify the state of Nibbana with a
kind of undefinable, infinite cognition (vinnana) (although such authors coyly flirt around the
taboo word 'self).

The second version of this dialogue describes the culminating revelation of the Self as slightly
differently: 'without internal or external, just an entire mass of understanding' (anantaro'bahyah,
krtsnah prajnana-ghana eva312). The internal/external immediately reminds us of satipatthana.
Krtsnah is the Pali kasina, a well known meditation term which for the Suttas meant not 'external
device for meditation' (as it came to mean in the later traditions), but 'totality' [of consciousness
in samadhi], j u s t as here. Even though the Pali-English Dictionary says t h e derivation of kasina is
unknown, this derivation and meaning have been adopted in Cone's recent A Dictionary of Pali, and
also conforms to the Chinese rendering. The standard Sutta description of kasina is 'above, below,
across, non-dual, measureless', which once again reminds us of the Upanisadic passage. This
version substitutes prajnana ('understanding') for the first version's vijnana (cognition); while
these are usually distinguished in Buddhism, they are often synonyms in the Upanisads, and
remnants of this remain in the Suttas, too.

We could go on at some length examining t h e subtleties of how these psychological and

philosophical elements are treated in comparative light, and what exactly is the historical
relationship between these texts, but this should suffice for our current purpose. We can
definitely conclude that, just as the central usage of ekayana in the Suttas looks to the
Brahmanical context, the central usage of ekayana in the Upanisads looks to the Buddhist context.

128
Yajnavalkya's message is the return to the One. All diversity is undercut and relativized. The
repeated, emphatic use of ekayana drives home the centrality of this doctrine of Oneness. An
ekayana is both the source from which the things of the world spring, and the place of
convergence, where the diversity of external manifested phenomena come together in a
profound unification. It is the inner subject that makes the external objects possible. It is
precisely the term ekayana that is singled out as the most exact expression of this insight. This
verse occurs immediately before the dramatic revelation that 'this great being', the cosmic self
that is the All, is but a 'sheer mass of cognition'. In this infinite consciousness the limiting,
diversifying conceptions and perceptions (samjna) disappear.

While it is not easy to find clear and explicit statements in the Brahmanical scriptures, usages of
ekayana elsewhere do n o t contradict this reading, and may even support it.

'The mind (citta) is the convergence (ekayana) of all these, mind is t h e self (atman), mind is the
foundation (pratistha).' (CU 7.5.2)

'He who would be devoted to convergence (ekayana), in silence, not thinking about anything,
having previously renounced, would be one who has crossed over and is free f r o m obstacles.'
(MBh 14.19.1)

The first passage appears to support the abstraction of the notion of 'convergence' to include
states of unification in samadhi. The terminology is suggestively Buddhist; no less than three
words (citta, ekayana, pratistha) are similar or identical to terms used in the context of
satipatthana. The second passage has puzzled commentators, 313 but surely a samadhi context is
also applicable.

So it seems that the main spiritual significance of ekayana for the Brahmanical tradition is
convergence of the mind in samadhi. This conclusion is in perfect accord with the
straightforward meaning of the term. The suffix -ayana means literally 'going', and when used in
a compound it normally means the 'going' to whatever the first member of the compound is. For
example, the Brahma Purana says that God first created the waters that are called nara and then
he released his seed into them, therefore he is called 'Narayana'. Closer to home, the Parayana
Vagga says: 'This path goes to the beyond (para); therefore it is called "Parayana".' So a perfectly
obvious reading would be that ekayana means 'going to one'. This is the primary meaning that
Gethin ends up with, corresponding to the final of the commentarial readings (ekam ayati).

The question then becomes, what does 'one' mean in the Buddhist context? Gethin suggests, with
the support of the Pali commentary, that 'one' refers to Nibbana. He says that the t e r m 'one' here
need not carry absolutist metaphysical connotations in the Nikayas. The sub-commentary
supports this by explaining 'one' here as 'without a second' and 'supreme'. But the 'one' was a
pregnant metaphysical term for the whole Brahmanical tradition ever since the Rg Veda spoke of
the 'One Being', and so the Suttas carefully avoid using the 'one' to refer to Nibbana. However,
they are quite happy to use 'one' to refer to samadhi. By far the most common and idiomatic
usages of 'one' in the Suttas' meditation vocabulary are the terms 'one-pointedness' (ekaggata)
and 'unification' (ekodibhava), which are standard synonyms of j h a n a or samadhi. This is, in fact,
the only sense in which 'one' is used in the Satipatthana-samyutta, and we have seen that the
Sarvastivada includes a discourse where Venerable Anuruddha closely associates t h e ekayana
passage with the development of'oneness of mind' through satipatthana.

129
Given that, as we have seen, the primary purpose of satipatthana is to lead t o j h a n a , it seems not
at all unlikely that the contextual meaning of ekayana is 'leading to unification (of mind)'. This is
precisely t h e explanation offered by the Mahasanghika Ekayana Sutra:

'Why is it called t h e "one entrance"? It is so-called because it is [the way to] concentration and
oneness of mind.' 314

The Chinese phrase I have rendered by 'oneness of mind' here is identical to that in Anuruddha's
discourse on the ekayana path. Although the Ekayana Sutra, which I will discuss in detail below,
seems somewhat later than the Theravada and Sarvastivada Satipatthana Suttas, this statement
remains one of the earliest clear definitions of this phrase. Because it is in accord with the early
descriptions of satipatthana in the Suttas, and was accepted by a prominent school of early
Buddhists, it should be granted some credibility.

This explanation would also furnish an answer to the question why the Suttas reserve the t e r m
ekayana for satipatthana alone among the 37 wings to enlightenment. While all the groups are
associated in one way or another with samadhi or one-pointedness, satipatthana is singled out as
playing the key role of bringing the mind to samadhi.

ACHIEVING THE ' W A Y '

The statement on the 'path leading to unification' is followed, in the Satipatthana Suttas, by the
statement that this is: 'For the purification of beings, for surmounting sorrow and lamentation,
for ending bodily and mental suffering, for achieving the way, for witnessing Nibbana'. Most of
these are straightforward enough, but the phrase 'achieving the way' (nayassa adhigamaya) is
vague, in Pali as well as in English. The t e r m naya is sometimes used in the context of dependent
origination, and some, including myself, have seen this as t h e meaning here. This is not
unreasonable, for there are clear connections between satipatthana and dependent origination. 315
However there is no very strong reason for thinking that this is the primary intended meaning.

The most common, standard occurrence of the t e r m is in the formula for recollection of the
Sangha: 'The Sangha of the Blessed One's disciples has practiced well, practiced directly, practiced
according to the way (naya), practiced properly.' Here we are in the realm of practice, and this,
rather than dependent origination, would seem a more straightforward approach to seek the
meaning of naya in satipatthana. This accords with the tradition, for the commentaries gloss naya
in such contexts as 'the noble eightfold path'. There are a whole series of texts that use the word
naya in the context of practice, which do not mention dependent origination, or for that matter
satipatthana, even though some of t h e m use t h e very same pericope of 'for the purification of
beings.'. 3 1 6 These passages generally deal with the overall way of training, and all specifically
include samadhi. The most explicit context is in the Sandaka Sutta of the Majjhima. First the
gradual training is taught, from t h e appearance of the Tathagata to the abandoning of
hindrances ; up to this stage the t e r m naya is not used. Then the text continues:

'Having abandoned these five hindrances, defilements of the mind that weaken understanding,
quite secluded f r o m sensual pleasures, secluded f r o m unskilful qualities, he enters and abides in
the f i r s t j h a n a , which has initial & sustained application, with rapture & bliss born of seclusion.
An intelligent person would certainly live the holy life with a teacher under whom a disciple
attains such an exalted distinction, and while living it he would attain the way (naya), the
Dhamma that is skilful.' 317

130
The text then goes on with the rest of the jhanas and higher knowledges culminating in
arahantship, repeating the concluding sentence each time. The association of naya with kusala,
'skilful', is also relevant in the context of satipatthana, for in one place satipatthana, in the
context of developing samadhi, is said to be 'for achieving the skilful' (kusalassadhigamaya), just as
it is said to be 'for achieving the way' (nayassa adhigamaya).318 Thus, while the term naya may not
be pinned down very neatly to a specific doctrinal denotation, it is clearly associated with the
progressive scheme of the way of practice in general, and samadhi o r j h a n a in particular.

T H E ANGUTTARA

This is a convenient place to survey the satipatthana material found in the existing Anguttara
Nikaya. Here we find a paucity of material on satipatthana. This is, however, no surprise, for the
Samyutta and Anguttara were obviously arranged as parallel collections of the shorter discourses,
and the main doctrines are mostly included in the Samyutta. Satipatthana appears alongside the
other groups of the 37 wings to enlightenment in various repetitive series appended to some of
the sections. One passage mentions six things to be abandoned in order to achieve success in
satipatthana: fondness of work, speaking, sleeping, and company; lack of sense restraint; and
eating too much. 319 This is obviously similar to the gradual training, and reinforces the suggestion
I made above that these basic practices in t h e gradual training are a preparation for the
meditative development of satipatthana. The contemplation here is treated in terms of internal
and external.

Returning to the Anguttara, the only substantial discourse on satipatthana treats it purely as
samatha. 320 This has t h e appearance of a discourse from t h e Satipatthana Samyutta, and the GIST
would suggest that it was originally included in that collection and was later moved into the
Anguttara in order to provide the Anguttara students with at least one substantial teaching on
such an important topic. However, the discourse is not found in the Sarvastivadin Samyukta, nor,
apparently, elsewhere in the Chinese. One is exhorted first of all to develop the four divine
abidings, then to develop 'that samadhi' in the mode of all the jhanas. Next one is, in identical
terms, exhorted to develop the four satipatthanas, and to develop 'that samadhi' in the mode of
thejhanas.

There is another interesting text, which, while it does not deal with satipatthana directly, is
similar enough in its subject matter and terminology to suggest that it may have exerted some
influence on later expositions. Venerable Ananda lists five 'bases for recollection' (anussatitthana;
notice the similarity to satipatthana.).321 They are: the first three jhanas; t h e perception of light;
the 31 parts of the body; contemplation of death; and the f o u r t h j h a n a . To these t h e Buddha adds
a sixth - mindfulness of one's bodily postures. These are obviously close to the Theravada
Satipatthana Sutta's section on body contemplation. The relation becomes closer when we realize
that the Sarvastivada Smrtyupasthana Sutra includes in its section on body contemplation the
f o u r j h a n a s and the perception of light. Perhaps the Sarvastivada was influenced by the present
text. The fact that the Buddha added the awareness of postures as an extra practice suggests that
this stood slightly outside the other, more specifically meditative, practices. It is characteristic of
the Buddha in such cases to focus attention on the cause.322 Thus it may well be that the Buddha
chose to emphasize awareness of postures at this point in order to encourage the development of
the practice that would lead on to the higher stages.

Another short discourse focuses on t h e wisdom aspect of mindfulness, although again this is not
specifically within t h e satipatthana framework. 323 Five meditations are recommended: one should
'clearly establish mindfulness on the rise and end of dhammas', and develop perceptions of the
loathsomeness of food, the ugliness of the body, boredom with the whole world, and

131
impermanence of activities. These are the 'bitter pill' meditations designed to overcome our
neurotic aversion and fear of acknowledging the negative and unpleasant side of life. It should be
noted that in the Anguttara the other 'bitter pills', which include both samatha and vipassana
aspects, are taught much more frequently than the contemplation of dhammas. 'Dhammas' here
probably does not mean 'all dhammas', for these are not said to be impermanent; presumably it
has the same meaning as in the satipatthana contemplation of dhammas.

We can summarize the teachings on satipatthana according to the Anguttara Nikaya in this way.

1) Satipatthana is a meditative practice developed in the context of the gradual training.

2) It can be a mode of jhana.
3) It is to be developed both internally and externally.
4) Mindfulness of the rise and fall of dhammas is one of the practices developing wisdom.

The high degree of congruence of this conception with the other basic teachings throughout the
canon suggests that we can regard this as an early summary of the key aspects of satipatthana. I
note here in passing that none of the texts mentioned above are found in the Ekottara Agama;
however this is of little significance given the generally large divergence between this collection
and the other Nikayas/Agamas. The Ekayana Sutra seems to be the only substantial discourse in
the Ekottara dealing with satipatthana.

CHAPTER 1 2 : THE SAMYUTTA

T H E SATIPATTHANA-SAMYUTTA

The Satipatthana-samyutta is a rich collection, with parables and metaphors, glimpses into the
daily life of the Sangha, inspiring lay meditators, humour and tragedy, and a strong narrative
element. Several texts shine light on satipatthana practice beyond what is found in the basic
formula; yet a detailed analysis is lacking. The Theravada Satipatthana-samyutta is composed of
five chapters (vaggas) of exactly ten discourses each, followed by the inevitable 'Ganges
Repetition Series'. Let us, to start with, classify the texts into the angas, as we did with the Sacca-
samyutta. As noted above, the ascriptions are not always cut-and-dried, but are usually easy
enough.

Table 12.1: Angas in the Theravada Satipatthana-samyutta

SN 47 Sutta/Vyakarana
1-2 S
3 7
4-8 S
9-13 7
14 S
15-16 7
17 S

132
18 Geyya
19 S
20-23 7
24 S
25-30 7
31-39 S
40 7
41 S
42 7
43 Geyya
44-45 S
46-47 7
48-104 S

While there are some significant blocks of texts grouped according to the angas, still the
formation is less striking than in the case of the Sacca-samyutta. Seven of the first eight texts are
suttas; but generally the suttas and vyakaranas are scattered in no particular order. Let us then see
what SA has to offer us.

Table 12.2: Angas in SA

SA Anga
605-8 Sutta
609 7yakarana
610-13 S
614-15 7
617 Geyya
618-20 S
621 7
622-23 Geyya
624-32 7
633-37 S
638-39 7

The sutta/vyakarana structure of this collection becomes immediately apparent. Fourteen of t h e

first eighteen texts are suttas, and these are followed by a sequence of nine straight vyakaranas.
We even have the extra grace of three geyyas appearing more-or-less between the two main
angas, just as we saw in the Sacca-samyutta. The demonstrable fact that the sutta/vyakarana
structure is far more obvious in the Sarvastivada collection has, I think, been a major factor in the
failure of scholars to recognize this fundamental principle, since their efforts have been largely
concerned with the Theravada.

Let us now combine the two collections, and remove those discourses that are not common to
both, thus attempting to approximate the pre-schism collection. I also note the discourses in the
form of a parable, which I call 'tales'; these are discussed below.

133
Table 12.3: Angas in the Concordance of SN 47 and SA

SA SN 47 Anga
606 24 Sutta
607 1 S
608 33* S
609 42 Vyakarana
610 39 S
611 5 S
614 11 V
615 10 V
616 8 S (tale)
617 6 Geyya (tale; S in SN)
619 19 S (tale)
620 7 S (tale)
621 4 V(S in SN)
622 2,44* G (S in SN)
623 20 G (tale; V in SN)
624 16 V
625 15 V
627 26* V
628 21* V
634 17 S
635 34* S
638 13 V
639 14 V

The sutta/vyakarana structure remains apparent. In addition, another feature manifests. There
are in this concordance ten suttas, three geyyas, and ten vyakaranas. This is suggestively
reminiscent of the classic grouping of ten discourses to a vagga. Perhaps the ancestral collection
consisted of two vaggas, one of suttas and one of vyakaranas, and a few geyyas. In the appendix we
can investigate the structure in more detail to see if any f u r t h e r gleanings as to the original
structure of the collection are to be found. However, as in the case of the Sacca-samyutta, this
inquiry will not substantially affect the content of the collection, so we may, without f u r t h e r ado,
proceed to consider this.

T H E TALES

Two distinguishing literary characteristics of the Satipatthana-samyutta deserve f u r t h e r

consideration: the tales and the narratives.

The tales leaven the collection with humour and homely action. They extol satipatthana as a safe
place, a refuge, thus reflecting the exhortation in the narrative passages to abide with oneself as
one's refuge through practicing satipatthana. Psychologically, they emphasize how one is 'caught'
if one strays outside one's native domain, and 'freed' if one know how to 'let go', and how this
freedom, apparently paradoxically, comes from containing one's 'attention' within the proper
context - satipatthana.

134
SN 47.8/SA 616 The Cook: The Pali and Chinese here are quite similar. The foolish cook doesn't
know how to prepare t h e right kinds of food to please his master, so misses out on a bonus.
Similarly, the foolish monk, not understanding the ways of his mind, practices satipatthana, but
his mind 'does not enter samadhi, he does not abandon d e f i l e m e n t s . ' But for the skilful cook and
the skilful monk it isjust the opposite - he enters samadhi and abandons defilements.

SN 47.6/SA 617 The Hawk: Tells of a quail who is caught by a hawk if he ventures outside his
'native habitat', but is safe when inside. The unsafe ground for a monk is the realm of sensuality,
and the native habitat is satipatthana. Here again satipatthana, as the opposite to the five strands
of sensual pleasures, plays t h e same role as the four jhanas. The Chinese is very close, sharing
some of t h e exact phrases of the Pali, and also adds a verse. This repeats the story, then says that
to know one's own mind is better than the power of 100 000 dragons.

SN 47.19/SA 619 Sedaka: A parable of harmony and mutual support, illustrated by the story of two
acrobats, one supporting the other on top of a bamboo pole. The Pali contains a recognized
anomaly here, for it has the apprentice saying 'you protect yourself and I'll protect myself. The
Buddha then applauds this statement, but in repeating it the saying changes to 'protecting
oneself, one protects others, protecting others one protects oneself.' The Chinese does not suffer
this problem, for it has t h e apprentice sayingjust that.

SN 47.7/SA 620 The Monkey: Here the story concerns a foolish monkey who gets his hand stuck in
a tar trap. Trying to free himself, he gets his other hand stuck, too, then both his feet and even his
muzzle. Stuck at five places, the h u n t e r does with him what h e will. In the same way, the monk
should stay within his native habitat, just as in 'The Hawk'. The Chinese adds a f u r t h e r
description of the foolish monk who goes into town with senses unrestrained.

SN 47.20/SA 623: The Most Beautiful Girl in the Land: On one side is an excited crowd; on the
other a dancing girl; in between you must walk carrying a bowl brimful of oil, with a man with a
drawn sword following right behind you, ready to chop off your head if you spill a drop! I n j u s t
the same way should you develop 'mindfulness of the body'. This text is unique in being a 'tale'
that includes a vyakarana. The Pali is anomalous in that it mentions 'mindfulness of the body', and
is the only text in the Satipatthana-samyutta to omit the four satipatthanas. The Chinese,
however, does include the four satipatthanas, which is clearly more plausible. The Chinese also
adds some verses, which mention t h e pot of oil and encourage concentration. Another difference
is that the Theravada version is set at Sedaka. This unusual location is the same as the preceding
discourse and seems to have been mechanically copied over f r o m there, a phenomenon that we
saw was very prevalent in the Sarvastivada Satya-samyukta. The Sarvastivada version is set in the
Deer Park at Benares.

T H E NARRATIVES

The narratives fall into two main groups: those that recall the Buddha's enlightenment, and those
that recall his passing away. There is a complex relationship between these texts - some of which
include narrative, and some which merely imply a narrative context - and the extended
narratives (apadanas) of the DTgha and Vinaya.

Of the first kind we have the 'Brahma' suttas. As discussed above, these are found in the Sagatha
Vagga in the Chinese versions, but the Satipatthana-samyutta in the Theravada. They recall the
'Request of Brahma' that we have treated as the paradigmatic geyya passage. 324 Both events are set
at t h e 'Goatherd's Banyan Tree', but I can find no trace of our current text in any of the accounts
of the post-enlightenment period that are available to me. Between the two there is a crucial

135
difference in the role played by Brahma. In the 'Request of Brahma', the deity plays the ancient
mythic role of initiator of action, a personification of divine inspiration. Here, in keeping with the
humble position of gods that is more normal in Buddhism, he merely echoes and supports the
Buddha. The other memory of this period found in t h e Satipatthana-samyutta is called 'Unheard
Before', which treats the satipatthanas according to the three revolutions and twelve modes of
the Dhammacakkappavattana Sutta. 325 This implies that understanding satipatthana was an
integral part of the enlightenment experience. Similar claims are made of most of the familiar
doctrinal categories, reflecting the holistic nature of the Dhamma. This text is not found in t h e
Chinese, so we suspect it to be a later insertion; but the content is not controversial.

More material is found around the time of the Buddha's passing away. We have already referred
to t h e discourse set in AmbapalTs mango grove, where the Buddha stayed shortly before t h e
Parinibbana, encouraging the monks to be 'mindful & clearly comprehending'. This passage
appears in t h e Maha Parinibbana Sutta, in both Theravada and Mulasarvastivada versions. The
appropriateness of this reminder, in anticipation of the visit by a famous courtesan, is obvious in
context, and this is made more explicit in the Sarvastivada version than the Theravada. SN 47.9
'Ill' tells t h e moving story of the Buddha's illness shortly before his passing away. It is one of the
most poignant passages in the canon that frankly confronts the Buddha's frail humanity. For this
reason it might not have sat comfortably with the docetic tendencies of later Buddhism, which
may explain its absence from SA. On t h e other hand, we have seen that the Sarvastivada include
an extra passage on the Buddha's illness in their Catusparisat Sutra, a passage that as far as I am
aware is absent f r o m the Theravada tradition, so we should be careful about drawing conclusions
here. This passage also contains the Buddha's famous declaration that he does not have the
'closed fist of a teacher'. This became another controversial sectarian issue, especially regarding
those later schools that claimed to be descended from an esoteric transmission outside the main
scriptural lineage. This might possibly have provided another reason for suppressing this
discourse, which is not attested as an individual discourse elsewhere in the Chinese canon.
However the passage is also found in the Maha Parinibbana Sutta, and is included in t h e
Sarvastivadin version of that text preserved in Sanskrit, 327 and also in Chinese versions.

SN 47.11-14 form a mini biography of Venerable Sariputta. In the first text we see him discussing
with the Buddha the nature of a perfected human. Next we see Venerable Sariputta in his home
town of Nalanda, approaching the Buddha, and exclaiming that by an 'inference according to
dhamma', he understands that all Buddhas, past, future, and present, become enlightened by
abandoning the five hindrances, being well established in the four satipatthanas, and developing
the seven enlightenment-factors. 3 2 7 Bearing in mind that the seven enlightenment-factors are
often virtually synonymous with samadhi, this grouping reminds us of the meditation phase of
the gradual training: one 'establishes mindfulness', abandons hindrances, and develops the
jhanas. Similar groupings occur frequently throughout the Suttas; below we will see that this
'inference according to Dhamma' is probably equivalent to the original content of the
contemplation of dhammas in satipatthana. Nevertheless, the connection with satipatthana is
relatively tenuous - they are merely mentioned in passing - so it is not surprising to find that t h e
Sarvastivada preferred to allocate this text to their Sariputta-samyutta. The episode is found in
the Theravada Maha Parinibbana Sutta, but not in the Sarvastivada or the Dharmaguptaka
versions of that text. 328 It also forms the kernel for the much longer dialogue that now forms the
SampasadanTya Sutta, which is available in several versions. 329 It does not seem unreasonable to
accept the Theravada account here: this episode occurred during the Buddha's lastjourney, north
f r o m Rajagaha, and records the last meeting between the Teacher and his greatest disciple.

SN 47.13/ SA 638, the next text in the Theravada collection, relates how Venerable Sariputta
became ill and passed away. This occurred at 'Nalakagama', and we might not be mistaken in
seeing a relation between this and t h e 'Nalanda' of the previous text. The traditional story has it

136
that Venerable Sariputta, aware that his time was near, returned to his home town to convert his
mother, an ardent Brahman, to Buddhism. The novice Cunda - said by the commentary to be
Venerable Sariputta's younger brother - was his attendant during his illness. On his passing away,
he took Venerable Sariputta's bowl and robes and went to where the Buddha was staying. The
Theravada says h e was at SavatthT, which is incongruous - SavatthT is about 200km to the north-
west, and it is inconceivable that the Buddha could have made such an extravagant detour in his
last, feeble, journey, a detour attested in no other text. The Sarvastivada says the Buddha was
staying in Rajagaha, at the Veluvana. But this is little better, for the Maha Parinibbana Sutta says
that Sariputta visited the Buddha in Nalanda, after he had already taken the north road f r o m
Rajagaha. Venerable Ananda, the first to hear the bad news, is distraught, and the Buddha asks
him whether Sariputta took with him the aggregates of ethics, samadhi, understanding, release,
and knowledge & vision of release; t h e Chinese adds to this the 37 wings to enlightenment.
Ananda says no, but that Sariputta was so helpful to his fellows in t h e holy life. The Buddha
encourages him: 'How can it happen that what is born, come to be, conditioned, of a nature to
disintegrate should not disintegrate? That is impossible! Therefore, Ananda, dwell with
yourselves as your own island, your own refuge, with no other refuge. Dwell with t h e Dhamma as
your island, your refuge, with no other refuge.' One does this through the practice of
satipatthana. This is strongly reminiscent of the Maha Parinibbana Sutta, and it is surprising that
it finds no place in that narrative.

This text forms a pair with SN 47.14/SA 639, which follows immediately in both collections. The
Theravada recension is set in the VajjT Republic, at Ukkacela on the banks of the Ganges. The
Chinese, however, places this episode in Madhura, by the Bhadra River. VajjT is a plausible setting,
being along the route towards the Parinibbana. The Buddha says that the assembly appears
empty to him now that Sariputta and Moggallana have attained final Nibbana, like the largest
branches falling off t h e great tree of heartwood. Venerable Moggallana's death is not recorded in
the canon, but the later Theravadin and Tibetan accounts agree in their general outlines. In the
teachings, this text largely repeats the previous. However, it includes some extra literary
features, suggestive of the later angas. The Buddha says that all the Buddhas in the past and the
f u t u r e will have a chief pair of disciples like Sariputta and Moggallana; this reminds us of an
avadana. Next he praises their 'wonderful and marvellous' qualities, a straightforward
abbhutadhamma; however, this section is absent f r o m the Chinese.

There are some more texts that might also be set in the period immediately following t h e
Parinibbana, although t h e narrative context is not spelt out in the texts themselves, nor are they
included in the longer developed narratives of the DTgha and the Vinaya. These are the discourses
featuring Venerable Ananda and Venerable Bhadda at the Cock's Monastery in Pataliputta. 330 The
only Sarvastivada cognate stands at the head of a whole series of texts at the same location; in a
manner similar to the Satya-samyukta, we assume that only t h e first setting is genuine, and the
rest were mechanically repeated. The Theravada texts do not mention the Buddha as being alive
at the time; although the Sarvastivada version does mention the Buddha, this might just be a
result of mechanical standardization. The location is unusual, a little-known monastery in
Pataliputta, half-way between Kusinara, where the Buddha passed away, and Rajagaha. It was
f r o m Rajagaha that t h e Buddha set out on his final journey, and back to Rajagaha that the monks,
including Venerable Ananda, returned to recite the scriptures in the First Council immediately
after the Buddha's passing away. The 'Duration' discourses, in particular, express evident concern
over the Buddha's passing away. So while they do not allow any definitive conclusions, it is
possible that here we have a record of the discussions and debates that took place within the
Sangha as, resting on t h e j o u r n e y back f r o m Kusinara to Rajagaha, they contemplated t h e f u t u r e
of Buddhism without the Buddha. It is surprising that the 'Duration' discourses, which say that
the long-lasting of Buddhism is due to the practice of satipatthana, find no cognate in the
Sarvastivada, for the long-lasting of t h e sasana was one of their characteristic concerns. Once

137
again, this cautions us against making premature judgements on sectarian issues in the early
texts.

It is interesting that several of the events that took place around the time of the Buddha's
enlightenment and passing away do not appear in the developed narratives of those events. The
relation between these short texts and the long narratives has been debated by scholars. Some
opine that the long narratives emerged first, and the shorter texts were abstracted f r o m them.
This seems to me inherently implausible, and the current situation supports this: if t h e shorter
texts were abstracted from t h e longer, where do the short texts that are not in the long
narratives come from? It is easy to envisage a situation where there were many incidents and
teachings remembered around these critical times, which were retold with a more-or-less casual
reference to the historical circumstances. The need for definitive biographical accounts would
become stronger in the absence of the living presence of the Buddha. So the events would
gradually be linked up in a narrative structure, whose general outlines were very old and have
changed remarkably little in the millennia of retellings, but whose specific form was elaborated
over time. Not all the remembered events found a home in t h e story; presumably the choice was
dictated by t h e demands of story telling.

These narratives enrich and inform the presentation of satipatthana, contextualizing the abstract
meditation formula within the drama of life and death. Another discourse also emphasizes
satipatthana in the context of death. This discourse, SN 47.30/SA 1038 Manadinna, depicts
Venerable Ananda approaching the householder Manadinna, who is gravely ill. Ananda
encourages him, even in that final extreme, to develop satipatthana. Manadinna responds that he
does indeed practice satipatthana; and furthermore, he has abandoned the five lower fetters.
Ananda then praises him as one who has declared the attainment of non-returning, the
penultimate stage of enlightenment. This discourse exists in the Sarvastivada, but in another
samyutta, and was therefore not in our concordance. The previous, almost identical, text in SN 47
is absent f r o m the Sarvastivada.

As always in the Suttas, death and impermanence are closely related: anxiety over death is a
primary cause for the existential crisis that precipitates the spiritual quest; and many of the
terms signifying impermanence in a general philosophical sense are primarily applied to death in
particular. Examples from the Satipatthana-samyutta include 'come to be' (bhuta) and
'conditioned' (sankhata). So we should not be wrong to think of death as the primary perspective
f r o m which to view impermanence in the context of satipatthana. But in these narratives we only
see the connection between satipatthana and death by association. There is no hint as to how this
perspective might be actually integrated within the basic meditation, which as we have seen does
not mention impermanence at all.

CLEAR COMPREHENSION: T H E CONTEMPLATION OF M I N D OBJECTS

Further light on this is shed by SN 47.35, which presents a variation on the t h e m e of'mindfulness
& clear comprehension'. The standard passage on clear comprehension is the description of
awareness of daily activities, which as we have seen usually occurs in the gradual training leading
up to the f o u r j h a n a s . The Sarvastivada includes two discourses on the gradual training that have
no Theravada counterparts. SA 636 gives the standard description of one who hears the Dhamma,
then decides to go forth out of faith. He then rectifies his bodily conduct, protects t h e four kinds
of right speech, purifies livelihood, guards the sense doors, and practices clear comprehension
when going out and returning, and so on. Then he resorts to a tree or a lonely place, sits down
and establishes the mind in a 'peaceful abiding'. He abandons the five hindrances that entangle
the bright power of the mind, and goes on to practice the four satipatthanas. This is called t h e

138
'development of the four satipatthanas.' The next text, SA 637 is essentially similar (in fact it
abbreviates, and instructs the reader to expand as the previous discourse). It adds a few extra
factors of the gradual training, such as the restraint of the Patimokkha and contentment with
requisites, just like a bird is content with its wings. In both of these discourses, satipatthana takes
the place normally occupied by the f o u r j h a n a s .

The practice of clear comprehension in one's daily activities is presented more simply in a
discourse, found in both collections, that exhorts the monks to be 'mindful & clearly
comprehending'. 'Mindfulness' is then defined as t h e four satipatthanas, and 'clear
comprehension' as the awareness of activities. 331

SN 47.35 has no SA counterpart, so it may be a secondary formation in imitation of this. In this

discourse, mindfulness is presented as usual as the four satipatthanas, b u t clear comprehension is
described as remaining aware of feelings, perceptions, and thoughts as they arise, remain, and
end. It is therefore 'contemplation of mind-objects', a label usually erroneously ascribed to t h e
fourth satipatthana. It is included as the third of four kinds of 'development of samadhi'. 332 There
is evidently an order of progression intended in these developments of samadhi. The fact that
they start out with j h a n a suggests they are advanced practices, the higher 'development' of
samadhi. The first two are on t h e side of samatha: the four jhanas, which lead to a pleasant
abiding here & now; and the perception of light, which leads to 'knowledge & vision' (= psychic
powers). The third, our current practice, leads to 'mindfulness & clear comprehension'. The
fourth is the standard practice of observing the origin and dissolution of the five aggregates,
which leads to the ending of defilements. These both focus on impermanence, and therefore are
vipassana. The question then arises, what is the difference between them? Why does one lead to
enlightenment, and one merely to mindfulness & clear comprehension?

The answer is not spelled out in the texts, but may be readily inferred. In watching the rise,
persistence, and ending of feelings, perceptions, and thoughts, the field of insight is not yet
complete. The body is left out. We have seen that the primary meaning of impermanence
regarding the body is the big issue, life and death. One's own death, as such, cannot be
contemplated directly, but must involve some kind of inference in a temporal framework. So it
seems that this practice excludes this. It is solely concerned with the experience in the present
moment, and has not yet deepened to an understanding of the principle underlying experience.
To use terminology derived f r o m elsewhere in the Suttas, it involves dhamme nanam (knowledge
regarding phenomena), but not anvaye nanam (inferential knowledge). 333 The other important
omission is the observation of cognition itself. One is merely observing the known, not the
knowing, which is the key to really deep insight. It seems that here too, the reason is that our
current practice involves only direct observation of phenomena as they occur, not inference of
the causal principle describing how these events unfold over time. According to the Suttas,
cognition (the knowing) arises dependent on mental objects (the known). If cognition were to
directly observe itself this would entail the fallacy of something being it's own cause. So cognition
can never be self-supporting; in other words, one cannot directly observe the act of cognition.
Insight into cognition requires, on some level, an inferential process involving time. So our
contemplation of mind objects, while valuable, will not lead to full liberation until it deepens into
a more full comprehension of the entire field of experience, past and future, as well as present.

We should note that both this meditative development of clear comprehension and the
awareness of activities are here clearly distinguished f r o m satipatthana as such; they are related,
not equated. No doubt in practice they may not be isolated f r o m one another, but the Samyutta
nowhere subsumes these practices of clear comprehension within the four satipatthanas. The
meditative practice, moreover, has quite different objects than satipatthana. Only feelings
directly correlates. While one might try to equate perceptions and thoughts with one or the other

139
of the four satipatthanas (traditionally they would be slotted under dhammas), the actual
descriptions of the practices, in all versions, do not support this. Again, the reason may be readily
inferred. Remember t h a t one of the key satipatthana exercises is anapanasati. The special
function of this meditation is to quell thoughts. This is reinforced in the Sarvastivada
Smrtyupasthana Sutra, which includes practices for quelling and suppressing thoughts. So t h e
primary purpose of satipatthana is not to be mindfully aware of thoughts as they pass through
the mind; it is to eliminate thoughts so that the mind may be brought to stillness. We also note
that the phrasing of this practice is not very similar to the description of satipatthana.

The meditative development of clear comprehension described here is clearly of secondary

importance; the passage occurs once only in each of the Samyutta, Majjhima, and DTgha Nikayas,
and three times in the Anguttara. Many of the contexts this passage appears in do not inspire
confidence as to its historical importance. In the DTgha it occurs in the SangTti Sutta (an
anguttara-principle proto-abhidhamma compilation); in the Majjhima in the Acchariya-abbhuta
Sutta (an abbhutadhamma); and in the Anguttara in the context of the patisambhidas, a rare and
probably late category. In the Theravada Satipatthana-samyutta, it occurs in the middle of the
fourth vagga, a chapter that has few cognates in the Sarvastivada, and appears to be composed of
largely artificial texts. I have not come across it the Mahasanghika Ekottara and the Sarvastivada
Samyutta, but is found in the Sarvastivada Madhyama and elsewhere. The partial Sanskrit Maha
Parinirvana Sutra opens with the section on the rising, persisting, and ending of thoughts, which
is evidently a remnant in this fragmentary manuscript of this clear comprehension practice ; here
this is said to be a 'wonderful and marvellous' quality of the Buddha, which reminds us of t h e
similar statement in the Acchariya-abbhuta Sutta. This might, incidentally, be an indication that
this discourse should be considered as part of the Maha Parinibbana narrative.

T H E VIBHANGA S U T T A

Several of t h e discourses identified above deal with the samadhi aspect of satipatthana. These
have withstood any test of authenticity I have been able to throw at them, and, I believe, must
stand as representing a central, mainstream conception of satipatthana. Finally, however, we
turn to t h e two remaining texts in the Satipatthana-samyutta that deal directly with satipatthana
as vipassana: SN 47.40 Vibhanga and SN 47.42/SA 609 Samudaya ('Origination').

The Vibhanga Sutta is not attested in the Chinese canon. This alone would make us suspect that it
is a secondary development; a closer examination confirms this. Like the discourse of
mindfulness and clear comprehension we h a v e j u s t discussed, it is located at the end of the fourth
vagga, which appears dubious because almost all of the discourses seem to be artificial, or at least
could have been produced by artificial means. The Vibhanga Sutta, being at the end of this vagga,
is immediately before an unrelated discourse called 'Deathless', which is followed by Samudaya.
Thus 'Deathless' seems to be arbitrarily inserted between the closely related Vibhanga and
Samudaya.

The Vibhanga Sutta starts out with the Buddha saying: 'I will teach you...' This standard opening
seems innocuous enough. Normally, however, 'teaching' (desana) in brief is contrasted with
'analysis' (vibhanga) in detail. So the title and the text itself seem to be telling us two different
things. The status of titles of discourses is often dubious, since in Pali texts the titles usually do
not accompany the discourse itself, but are merely inferred f r o m the mnemonic verses at the end
of the vagga. So it seems likely that the statement of the text itself should be given more
credence. And indeed, in accordance with the statement in the text itself, and in contrast with the
title, the discourse appears more like a teaching (in brief) than an analysis (in detail). There is no
proper 'vibhanga' analysing in detail the individual elements. Rather, a threefold teaching is

140
presented: satipatthana (the usual formula), the development of satipatthana (contemplating the
nature of origination and dissolution regarding the body, etc.), and the way leading to the
development of satipatthana (the noble eightfold path).

This should be compared with the Iddhipada-samyutta. There, one discourse presents a fourfold
'Teaching': psychic power; the basis for psychic power (simply described as the practice leading
to psychic power); the 'development' of the bases for psychic power (one 'develops' the four bases
for psychic power according to the standard formula); and the way leading to the development of
psychic power (the noble eightfold path). 334 The internal coherence in t h e usage of the terms
'teaching' and 'development' argue for t h e authenticity of this text. The next discourse presents
an 'Analysis' of this teaching, which is in classic vibhanga style.335 This shares some significant
and unusual features in common with the Satipatthana Sutta. The terms 'constricted mind' and
'scattered mind' occur in the contemplation of mind, and the 31 parts of the body occur in the
contemplation of the body. Some other passages in the Sarvastivada Smrtyupasthana Sutra are
even closer to the Analysis of the Bases for Psychic Power:

'And again monks, contemplating a body in the body, a monk cultivates a glorious bright mind,
well received, well grasped, well remembered: as before, so after; as after, so before; as by day, so
by night; as by night, so by day; as below, so above; as above, so below. In this way he is not
confused at heart, he does not have entanglement. He cultivates a glorious bright mind, a mind
that is finally not obscured by darkness.' 336

These considerations lead us to wonder whether the Vibhanga Sutta of the Satipatthana-
samyutta should properly be called 'Teaching'. It would then have been complemented by a more
extensive discourse in vibhanga style. The only text that might fit the bill is the Satipatthana
Sutta, or more correctly the *Satipatthana Mula, which is very similar to the chapter on
satipatthana in the Abhidhamma Vibhanga, and has strong parallels with the Vibhanga of the
Iddhipada Samyutta. Several other vibhangas on Samyutta topics are now found in the Majjhima
(Saccavibhanga, Dhatuvibhanga; Salayatanavibhanga is in MN but the Sarvastivada counterpart
remains in SA, surely its original home). These are distinguished f r o m the vibhangas left in the
Samyutta primarily by length: the shorter vibhangas are in the Samyutta, the longer ones in the
Majjhima. Would it be too presumptuous to suggest that t h e Satipatthana Sutta was earlier a
Vibhanga Sutta of the Samyutta, but with extra expansion was removed to the Majjhima?

Another problem with t h e Vibhanga Sutta is that here the observation of rise and fall is called
'development'. Normally, as in the discourse f r o m the Iddhipada-samyutta described above,
development (bhavana) is described in terms of t h e 'cultivation, development, and making much
of t h e relevant dhammas. But there are three texts in the Theravada Satipatthana-samyutta,
each of which portray 'development' in different ways. There are two texts that use
'Development' for their title. Only the first of the two 'Development' discourses meaningfully
refers to 'development' in the discourse itself: 'These four satipatthanas, when developed and
cultivated, lead f r o m the near to the far shore.' 337 The Sarvastivada contains several discourses of
this type, some of which are merely listed in summary, each substituting a different stock phrase
describing the benefits of developing satipatthana; it leads to the complete ending of suffering, or
to great fruit and benefit, etc., The second Development Sutta in the Theravada is of a type
elsewhere titled 'Teaching' (desana).338 It starts off with the Buddha saying: 'Monks, I will teach
you the development of the four satipatthanas.' The discourse then merely presents the basic
satipatthana formula and says this is t h e development of the four satipatthanas. The closest
Sarvastivada cognate to this, SA 610, elaborates development as the contemplation of the
satipatthanas internally and externally. This understanding of development is found in later
works like Asanga's Abhidharmasamuccaya, which adds to the body of evidence that Asanga used
a Samyukta very like the one we have in Chinese. 339

141
So we now have four descriptions of t h e 'development of satipatthana': one is to 'cultivate and
develop' the basic practice; the second simply presents the basic formula; the third, found only in
the Sarvastivada, is to contemplate internally and externally; and the fourth, found only in t h e
Theravada, is to contemplate the objects of satipatthana as impermanent. Only t h e first of these
fits the normal usage of 'development'. The internal/external practice, while not exactly how
development is presented normally, seems appropriate enough, for it teaches how to broaden
and extend the basic practice, in terms commonly found in the mainstream satipatthana
discourses. The contemplation in terms of impermanence is apt in a certain sense, since it also
appears to depict an advanced practice intended for those already well established in t h e
fundamentals; yet 'development' is not normally identified specifically with vipassana in this
way.

So our final assessment of the Vibhanga Sutta must be that it seems late because of its position
within the Samyutta, lack of Sarvastivadin counterpart, and internal incongruencies. There might
have been an earlier version, more aptly called 'Teaching', that described 'development' as the
'cultivation, development, and making much' of the four satipatthanas, or perhaps t h e
internal/external contemplation, and later on this was replaced by t h e section on rise and fall.

T H E SAMUDAYA SUTTA

We may now go on to examine the Samudaya Sutta. The existing Sarvastivadin version of this is
significantly longer than the Pali, raising the question as to which is the original. A number of
considerations taken together make it virtually certain that the Sarvastivada version is more
original. In order to make the following discussion as clear as possible I will first present here the
essential doctrinal elements in the Chinese version, adapting the Chinese translation to conform
more obviously with standard renderings of the Pali.

1) I will teach you, monks, the origination and dissolution of the four satipatthanas. Listen well
and pay attention; I will s p e a k . What is the origination and dissolution of the four satipatthanas?

2a) Due to the origination of food there is t h e origination of the body; due to t h e cessation of food
there is the dissolution of the body.
2b) In this way, monks, a monk contemplates the nature of origination in t h e body, he
contemplates the nature of dissolution in the body, he contemplates t h e nature of origination and
dissolution in the body.
2c) He abides independent, not grasping at anything in the world.

3a) Due to the origination of contact there is the origination of feelings 340 ; due to t h e cessation of
contact there is the dissolution of feelings.
3b) In this way, monks, a monk contemplates the nature of origination in the feelings, he
contemplates the nature of dissolution in the feelings, he contemplates the nature of origination
and dissolution in the feelings.
3c) He abides independent, not grasping at anything in the world.

4a) Due to the origination of n a m e & form there is the origination of the mind; due to the
cessation of n a m e & form there is the dissolution of the mind.
4b) In this way, monks, a monk contemplates the nature of origination in t h e mind, he
contemplates the nature of dissolution in the mind, he contemplates the nature of origination
and dissolution in the mind.

142
 4c) He abides independent, not grasping at anything in the world.

5a) Due to the origination of attention there is t h e origination of dhammas; due to the cessation
of attention there is the dissolution of the dhammas.
5b) In this way, monks, a monk contemplates the nature of origination in the dhammas, he
contemplates the nature of dissolution in the dhammas, he contemplates the nature of
origination and dissolution in the dhammas.
5c) He abides independent, not grasping at anything in the world.

The major doctrinal content is the specification of the causes for t h e four objects of satipatthana,
and in this both traditions are in complete agreement. However much of the remaining structure
differs. The Theravada has the introductory section 1, but instead of asking 'What is the
origination and dissolution of the four satipatthanas' it asks 'What is the origination of the body?'
This is clearly an editing glitch. We would expect that this question should be repeated for the
other three satipatthanas, but it is not; also, the q u e s t i o n j u s t refers to origination, b u t the text
refers to both origination and dissolution. These editing anomalies are not found in t h e Chinese.
The phrasing also departs f r o m the standard form. Normally the Buddha says: 'I will teach you
x...Listen to t h a t . A n d what is x?' But here he says 'I will teach you x...Listen to t h a t . A n d what is
y?' This kind of change could have happened if the text was written half on one page and half on
the next; the question became detached f r o m the introduction and was included with the section
on the body (as in the PTS Pali and translations), and so later copyists assumed the question must
refer to the body. The Chinese version is more rational and hence more likely to be authentic.

Sections b and c are absent from the Theravada Samudaya Sutta; they are however very similar to
the 'vipassana refrain' of the Theravada Satipatthana Sutta. Elsewhere in the Theravada
Samyutta, section b occurs, but not in association with section c. This suggests that the early
version of the Samudaya Sutta as preserved in Chinese was a decisive influence on the formation
of the vipassana refrain of the Theravada Satipatthana Sutta; this must have occurred before the
break-up of the Samudaya Sutta.

Section b is now found in the Vibhanga Sutta, which as we have seen is likely to be a later
development. It seems that when the Samudaya Sutta was broken up, sections b and c were
moved into the original *Vibhanga Sutta, which was moved to the Majjhima and retitled the
'Satipatthana Sutta'. Somehow, the original *Desana Sutta that was paired with its *Vibhanga
Sutta remained in the Samyutta, but took the title 'Vibhanga' and also section b f r o m t h e
Samudaya Sutta.

Having established that the Sarvastivada is more likely to represent t h e pre-schism text of the
Samudaya Sutta, we should now consider more closely whether this is likely to have been in the
original collection. The phrases 'principle of origination, principle of dissolution, principle of
origination and dissolution' occur in a sequence of three discourses in the Khandha-samyutta. 341
These discourses, together with the two following, are combined into one discourse in the
Chinese. The explanations for 'origination' and 'dissolution' in satipatthana are also reminiscent
of the causes for the arising and ceasing of the five aggregates, as in the important Seven Cases
Sutta, which is one of the most widely distributed of all the discourses (and seems to have
functioned like a *Khandhavibhanga Sutta in t h e absence of a proper text of this type). 342 This
says that 'due to the origination of food there is the origination of physical form [=body]...due to
the origination of contact there is the origination of feeling...[and] perception...[and] conceptual
activities.due to the origination of n a m e & form there is the origination of c o g n i t i o n . ' Since
these kinds of vipassana teachings are obviously prevalent throughout the Khandha-samyutta, it
seems possible that that was their original home and the inclusion in the Satipatthana-samyutta
was a secondary development. This was probably prompted by the partial correspondence

143
between t h e four satipatthanas and the five aggregates: body=physical form; feeling=feeling;
mind=cognition; dhammas=(perception and conceptual activities???). This correspondence was
made explicit in the later texts of the schools, as we shall notice.

Despite these doubts, however, the Samudaya Sutta is an important pre-sectarian text and
deserves closer consideration. It says that the origin of the body is food; the origin of feelings is
contact; the origin of the mind is name & form; and the origin of dhammas is attention. We need
not pause here to consider these in detail, but we should notice that these descriptions, most
obviously the first, dispose of the idea that impermanence in satipatthana means momentariness.
Food sustains life; if you stop eating you'll die. Clearly here cessation or passing away just means
'death'. We have seen that the narrative passages of the Satipatthana Samyutta exemplify this
central paradigm in dramatic form. Contact is the origin for feelings in both the five aggregates
and the dependent origination. Attention as origin for dhammas is interesting. Attention is the
basis for wisdom, and is most typically treated as inquiry into causes, especially the causes of
good and bad dhammas. This suggests that vipassana is intrinsic to this last satipatthana, as I
have already hinted above. We will bear this definition in mind below as we probe more deeply
into t h e original scope of the 'contemplation of dhammas'.

'Name & form' is in the Samudaya Sutta said to be the origin of mind (citta), whereas normally (in
the aggregates and dependent origination) it is said to be the origin of 'cognition' (vinnana).
Obviously here citta and vinnana function as synonyms; but this bare fact does not help us to
understand why this terminological shift occurs in this context. Typically, vinnana is used in
vipassana contexts, such as dependent origination, the five aggregates, and the process of sense
cognition. It is therefore treated under the first noble truth, and is 'to be fully known'. Citta is
more difficult to pin down, for it is widely used in non-technical contexts to mean simply 'mind',
'thought', 'mood', 'state of mind'. However, when it is used in a technical sense it is often a term
for samadhi - the 'higher mind' (adhicitta), 'endowment with mind' (cittasampada), etc. It is
therefore treated under the fourth noble truth, and is 'to be developed'. This is why citta is
appropriate for satipatthana - it encompasses both the ordinary mind and the mind developed in
samadhi. But when the normal samatha context of satipatthana is extended to include vipassana,
we end up with citta appearing out of character in a role normally played by vinnana. A similar
vacillation between citta and vinnana also occurs in other contexts where samatha and vipassana
overlap. 343

It is interesting to note that the specific items given as the origination of the four satipatthanas
are not directly mentioned in the Satipatthana Sutta. The vipassana refrain speaks of origination
and dissolution, without specifying the particular causes. And inside the meditation exercises
themselves there is no mention of causality. The various body contemplations do not mention
food (with the exception of the Sariputrabhidharma); the contemplation of feelings does not
mention contact; the contemplation of mind does not mention n a m e & form; the contemplation
of dhammas does not mention attention. 344 Some might challenge this, especially in the case of
contemplation of mind. This mentions various mental qualities - such as greed, hatred, delusion,
etc. - that are considered by the tradition to fall under the umbrella of 'name'. But this
consideration cannot stand examination in light of how the Suttas themselves treat 'name'. The
most straightforward and common treatment defines 'name' as 'feeling, perception, volition,
contact, attention'. 345 None of these factors occur under the contemplation of mind. The other
way of treating 'name' is to emphasize its role in concept formation, or more literally, in
'naming'. 346 But the contemplation of mind in satipatthana is about the emotional tone of the
mind, rather than the intellectual processes. So it seems that our assertion remains valid, and
that t h e meditation exercises described in the Satipatthana Sutta, in all its versions, are not about
investigating the causes of the objects of satipatthana. Since the Samudaya Sutta is the paradigm
in the Samyutta for the treatment of the vipassana dimension of satipatthana, and since it treats

144
vipassana precisely as such investigation into causes, it seems obvious that the contemplative
exercises in the Satipatthana Sutta are not being presented as vipassana.

An odd aspect of this text is that, for the only time in the Nikayas, it treats 'satipatthana' in an
objective sense. Normally 'satipatthana' ('establishing of mindfulness') refers to t h e subjective act
of setting up or focussing mindfulness on one of the four fields. But here satipatthana clearly
refers to the objects of mindfulness, that is, the body, etc. ('things on which mindfulness is
established'). This objective sense if taken literally is patently absurd - it entails that the body is
the 'one-way path' to Nibbana. This might come as a pleasant surprise for some; for since food is
the nutriment for t h e body, eating must be t h e nutriment for the path! This ambiguity of
expression again results from the shift in perspective as the framework designed for samatha is
extended to include vipassana, probably under the influence of the five aggregates. In the normal
cultivation of satipatthana meditation one is operating 'inside' the four fields, 'getting into' the
field of meditation, whereas in developing vipassana into the objects of satipatthana one has
'pulled back' f r o m and objectified the process for the purpose of analysis. Like the difference
between reading a story, where one enters into the characters and emotions - you feel angry or
sad or happy - and reading a review of the story, where one develops a critical, analytical insight
into how t h e story works - you understand how the text made you feel angry or sad or happy. We
shall see later that this ambiguity caused considerable confusion in later writings.

So there are evidently a range of issues with the Samudaya Sutta: editing irregularities in the Pali;
lack of close congruence between the Sarvastivada and Theravada; possible influence f r o m the
Khandha-samyutta; unusual treatment of the subject raising serious interpretive difficulties. So
while I have grudgingly conceded it a place in the concordance, its authenticity must remain
dubious. One feature, however attests to t h e earliness of the text, which is attention as the origin
for dhammas. As we shall see below, this fits well with the earlier content of the contemplation of
dhammas section, but not with the developed content of the existing Satipatthana Sutta. Also,
this statement cannot be directly derived f r o m the Khandha-samyutta or anywhere else. If the
Samudaya Sutta is a secondary addition, then, it is clearly not very late.

T H E REMAINING SAMYUTTAS

We may now briefly consider the treatment of satipatthana in the rest of the Samyutta. The
Anuruddha-samyutta deals entirely with satipatthana, the thematic unity underlined by the fact
that all the discourses were spoken by Venerable Anuruddha. Generally, the teachings of the
disciples occupy a distinctly lesser status as compared to those of the Buddha himself. They saw
their task as clarifying, expanding, or commenting on the basic teachings given by t h e Buddha,
not as introducing any radical new themes.

The Theravada Anuruddha-samyutta starts with the most complex vipassana analysis yet. 347 This
combines the internal/external contemplation with the impermanence contemplation. So one
contemplates the principle of arising, of vanishing, and of arising & vanishing regarding the body
internally. Then one contemplates t h e body externally in the same way, and so on. Each section is
then addended with the standard auxiliary formula So this is the only place where the vipassana
section is embedded within the satipatthana formula itself. Then it introduces another
framework, familiar elsewhere in the suttas. One contemplates the repulsive in the unrepulsive;
the unrepulsive in the repulsive; then ignores both and abides in equanimity. Remember that in
the Theravada Satipatthana-samyutta Vibhanga Sutta one first became established on all four
satipatthanas, and only then was impermanence introduced. Now, however, impermanence is
introduced f r o m the first, giving t h e impression, without stating so explicitly, that one may
undertake vipassana f r o m the start of practice. Here we see the beginnings of a trend that can be

145
traced over later expositions of satipatthana. This discourse does not have a close cognate. There
is one text in the Sarvastivada Anuruddha-samyutta that mentions seeing the repulsive in the
unrepulsive, and so on, which we might regard as cognate. 348 But this has quite a different setting
(like the 'Brahma' discourses that proclaim the 'way to convergence') and omits the mention of
rise and fall. The pair to this discourse, as mentioned above, complements the 'way to
convergence' by saying this leads to unification of mind, a saying not found in the Theravada. 349
Given Venerable Anuruddha's reputation as an archetypical quietist, this seems like a more
plausible saying.

Most of the rest of the Anuruddha-samyutta emphasizes this samadhi aspect of satipatthana, as
Anuruddha systematically ascribes his success in every kind of psychic power to satipatthana.
This follows naturally f r o m the basic function of satipatthana as support f o r j h a n a . Many of these
formulaic passages are abridged in t h e Sarvastivada-samyukta.

The Vedana-samyutta is clearly oriented towards vipassana, speaking frequently of

understanding the impermanence of feelings. But this does not mean that the samatha dimension
is ignored. Several discourses teach that the way to understand this cessation of feelings is
through thejhanas. 3 5 0 All of t h e j h a n a discourses are included in the Sarvastivada, while the two
mentioning satipatthana are not. There are 31 discourses in the Theravada Vedana-samyutta;
only five are missing f r o m the Sarvastivada, all of which deal with vipassana. 351

Satipatthana is invoked in two very similar discourses taught for sick bhikkhus. 352 One should
develop the four satipatthanas, have clear comprehension, and then contemplate the
conditionality and impermanence of feelings. Both these discourses occur next to each other in
the samyutta (although in the Sarvastivada they are in another samyutta, dealing with illness),
and they are both given in the identical setting, the Hall with the Peaked Roof in the Great Wood
of Vesali. This is an unusual location, and begs the question why two such almost identical
discourses were given here and nowhere else. The only difference between the two is that t h e
first states that feeling is conditioned by the body, while the second says feeling is conditioned by
contact. The latter is obviously the normal position of the Suttas, repeated many times in the
Vedana-samyutta itself, so t h e statement that feelings arise dependent on the body must remain
under a question-mark. In fact it seems likely t h a t we are really dealing with one text, and fairly
early on t h e word 'body' was substituted for 'contact', perhaps by mistake, to yield a pair of
discourses.

The teaching sequence of these texts is interesting: four satipatthanas; clear comprehension of
bodily activities; contemplation of feelings. It seems a little like they are first presenting the basic
teaching on satipatthana, then offering explanations of at least the following two. This was
pointed out to me by Bucknell, who believed that this constitutes evidence that the original
specification of body contemplation was the clear comprehension of activities, rather than the
parts of the body, as I argue in this paper. However, a closer examination suggests that this
conclusion is not required by the texts.

First of all we notice t h a t they do not explicitly claim that the clear comprehension of activities,
or the contemplation of the impermanence of feelings, fall under the relevant satipatthanas. The
discourses start off with the Buddha visiting t h e sick monks and encouraging them to 'wait their
time' mindful and clearly comprehending, saying that 'this is our instruction to you.' Then he
explains 'mindful' as the four satipatthanas, and 'clearly comprehending' as awareness of
activities. This i s j u s t as in the Satipatthana-samyutta, and here as there, the surface of the texts
present these as two distinct practices, with no attempt at integrating them. This is not to say
that they are not integrated or connected, but simply that the text does not spell this out. One
practice is stated; then t h e next is stated; then the Buddha sums up by repeating that 'this is our

146
instruction to you.' So t h e body of the teaching isjust the same as the Satipatthana-samyutta, and
the repeated injunction closes off this section of the discourse.

Next is introduced the contemplation of feelings, with a distinct change in mood. Whereas
previously we had a straightforward exhortation, now the text shifts to a hypothetical mood: 'If,
monks, for a monk abiding thus mindful, clearly comprehending, diligent, ardent, resolute, there
arises pleasant f e e l i n g . ' This shift might be caused by the conjunction of two different textual
pericopes. While this does not prove that the text as we have it is inauthentic, it does call into
question its reliability as an early authority. While in this section certain of the phrasing, such as
the repeated use of 'contemplates,' reminds us of satipatthana, some of t h e details are different.
Unlike satipatthana, the investigation into causality is made explicit: 'On what [is this feeling]
dependent? It is dependent on contact.' And interestingly, whereas satipatthana famously has us
contemplate 'a body in the body', 'a feeling in t h e feelings', etc., this passage says: 'One abides
contemplating impermanence in contact and in pleasant feelings.' (so phasse ca sukhaya ca
vedanaya aniccanupassiviharati) So while satipatthana is about focussing intently and exclusively
on one aspect of experience, our current passage is depicting a more complex, many faceted
meditation, seeing different kinds of phenomena, their relationship and mutual dependence, and
their common nature as impermanent. It is in contexts such as this that we find clear and explicit
descriptions of the process of vipassana as investigation into causality that are so strikingly
absent f r o m the basic specifications of the meditation practices of the Satipatthana Sutta itself.
Here the development of vipassana into feelings is stated after satipatthana; and it is described as
investigating their causal dependence, then contemplating their impermanence, vanishing,
fading away, cessation, relinquishment. These terms are virtually identical with the fourth tetrad
of anapanasati, in other words, the contemplation of dhammas. As noted above, this suggests that
the contemplation of dhammas, where vipassana finds its proper home in satipatthana, can be
extended in advanced meditation practice to encompass the other satipatthanas as well. So all in
all, this text yields some interesting points of similarity and difference as compared with the
Satipatthana Sutta.

Feelings are of course an intrinsic part of satipatthana, so it is no surprise that satipatthana is

introduced in the collection of discourses dealing with feeling. However, the four satipatthanas
are not mentioned at all in the Salayatana-samyutta and hardly in the Khandha-samyutta,
suggesting that t h e sense media and the aggregates were not considered as specially related to
satipatthana. The Khandha-samyutta mentions satipatthana a couple of times when listing the 37
wings to enlightenment. In one other discourse they are mentioned, but not directly in
connection with the aggregates:

'And where, monks, do these three unskilful thoughts cease without remainder? For one who
abides with a mind well established on the four satipatthanas, or for one who develops the
signless concentration.' 353

Elsewhere unskilful thoughts are said to cease in the first jhana; 354 anapanasati is the normal
practice recommended for cutting off thoughts. The controlling of thoughts also features in the
Sarvastivada Smrtyupasthana Sutra. In line with the trend emerging above, this treatment of t h e
samatha side of satipatthana is also found in the Sarvastivada.

Even though the four satipatthanas as such are not mentioned in the Salayatana-samyutta, yet
there is perhaps a closer connection between mindfulness and the six sense media than we find
between mindfulness and the five aggregates. This reflects a subtle difference in orientation
between the two frameworks. Meditation on the aggregates is specially associated with
eradicating wrong view, while that on the sense media is attuned towards transcending desire. It
therefore emphasizes sense restraint, which is closely associated with mindfulness, especially

147
mindfulness of the body. The standard passage was included above in Venerable Maha Kaccana's
reply to the Brahman Lohicca.355 There the order of t h e teaching is: sense restraint; mindfulness
of the body; measureless mind (i.e. jhana); understanding; release. Another passage says that a
monk should train himself regarding the six senses so that they do not obsess his mind, his
energy is tireless, his mindfulness is well established, the body becomes tranquil, and the mind
enters samadhi. 357 Thus the usage of mindfulness here is much the same as we have seen above.

ANAPANASATI

Mindfulness of breathing was the meditation the Buddha himself practiced underneath the Bodhi
tree, and remained his preferred meditation even after his enlightenment. Because of this it has
always claimed a special prestige as the royal road to Nibbana. The chief source text is the
Anapanasati Sutta in the Theravada Majjhima Nikaya.357 There is no Anapanasati Sutta as such in
the Sarvastivada Madhyama Agama, but it does exist as an isolated text in the Chinese canon. 358
The 16 steps, moreover, are found in the Sarvastivada Madhyama and Samyukta. In the
Theravada Majjhima and the Theravada and Sarvastivada Samyuttas the 16 steps of anapanasati
are analysed in relation to the four satipatthanas. The correlation is as follows.

Table 12.3: The Four Satipatthanas and the Sixteen Phases of Anapanasati

Satipatthana Anapanasati
1. Contemplation of the body 1. Breathing long
2. Breathing short
3. Experiencing the whole body
4. Tranquillising the bodily activities
2. Contemplation of feelings 5. Experiencing rapture
6. Experiencing bliss
7. Experiencing mental activities
8. Tranquillising mental activities
3. Contemplation of t h e mind 9. Experiencing t h e mind
10. Gladdening the mind
11. Centring the mind in samadhi
12. Releasing the mind
4. Contemplation of dhammas 13. Contemplating impermanence
14. Contemplating fading of lust
15. Contemplating cessation
16. Contemplating relinquishment

148
In the context of anapanasati the 'body' and the 'bodily activities' are d e f i n e d j u s t as the breath,
while the 'mental activities' are feeling and perception. The first tetrad evidently describes the
process of gradually settling and calming the breath. Contemplation of feelings is here described
purely in terms of the bliss of samatha; there is no place for contemplation of pain here, and
apparently no need for it. Contemplation of the mind is even more explicitly framed in terms of
samadhi experiences.

In accordance with the mainstream teachings on satipatthana, not until the fourth tetrad,
equivalent to contemplation of dhammas, do we encounter vipassana. The commentary typically
tries to read the first three tetrads as pertaining to both samatha and vipassana, but can claim no
support in the text. Similar terms expressing vipassana occur, with minor variations, throughout
the Suttas. Perhaps the most fundamental group, as for example the Anattalakkhana Sutta, etc.,
would be: knowledge & vision, repulsion, fading of lust, release. In anapanasati, 'relinquishment'
might have been preferred over 'release' to avoid confusion with the (lesser) sense of 'mind
release' (= jhana) in the contemplation of the mind. Or alternatively, 'relinquishment' may have
been intended to evoke the enlightenment-factors, which include this word in their standard
formula. 359 The Sarvastivadin versions have impermanence, repulsion, fading of lust, cessation;
again, a minor variation of terminology describing the same process.

In the Anapanasati Sutta, as well as both the Theravada and Sarvastivada Samyuttas, we are told
that developing anapanasati develops the four satipatthanas, developing the four satipatthanas
develops the seven enlightenment-factors, and developing the seven enlightenment-factors leads
to liberation. This reminds us of the 'inference according to dhamma' we met above, as well as the
meditative phase of the gradual training. The connection between anapanasati and the
enlightenment-factors here also helps us to understand the disparity in the description of
contemplation of dhammas in anapanasati and satipatthana. Whereas anapanasati has
'impermanence, fading away of lust, cessation, relinquishment', satipatthana has a list of various
dhammas. These include, in all versions, the seven enlightenment-factors. The standard passage
on these says they are 'dependent on seclusion, dependent on fading away of lust, dependent on
cessation, maturing in relinquishment'. This passage is in fact found in the Mahasanghika
Ekayana Sutra. This suggests that the contemplation of dhammas in anapanasati is related to the
enlightenment-factors, a relationship it will be worth considering further.

The fourth tetrad of anapanasati contemplates impermanence; but the impermanence of what?
The most plausible approach would relate this to the inner structure of the meditation itself. The
whole course of anapanasati emphasizes a gradual, progressive stilling, appeasement, ending of
activities. The breath is calmed and becomes very subtle and fine. The endless chatter of thinking
is stilled and one experiences ever more refined bliss and tranquillity. The hindrances end and
the clamour of sense impingement fades away. This successive stilling defines the course of the
meditation, the entire world of the meditator's experience at that time, and must surely
constitute the prime field for understanding impermanence. An interesting perspective is thrown
on this by the phrase, which we have already quoted above, describing this contemplation of
dhammas in anapanasati:

'Having seen with understanding the abandoning of covetousness & aversion, he watches over
closely with equanimity.' 3 6 0

'Covetousness & aversion' obviously harks back to the satipatthana auxiliary formula. In t h e
Satipatthana Sutta, the section on contemplation of dhammas starts with the five hindrances. The
first two of these are sensual desire and ill-will, which we can infer f r o m this passage are similar
or identical with 'covetousness & aversion'. Seeing the abandoning of these 'with understanding'

149
suggests the focus on causality that is characteristic of this section; the same word 'abandoning'
also occurs in the contemplation of dhammas, in reference to the abandoning of the five
hindrances. The contemplation of dhammas also includes the seven enlightenment-factors, and
these are frequently said to be the forces that can overcome the five hindrances. So it is
noteworthy that our current text finishes by saying that one should 'watch with equanimity'; for
equanimity is the last of the enlightenment-factors. This passage, then, could be read as
suggesting that the contemplation of dhammas is fulfilled by understanding how the hindrances
are abandoned through the strength of the enlightenment-factors, and with the fulfilment of this
process one dwells in equanimity. This in turn suggests that the development of anapanasati, and
hence satipatthana, will fulfil the enlightenment-factors, culminating in equanimity, which is a
central theme of the Anapanasati Sutta, from which this passage was taken.

Compared with the conservative, incremental evolution the teachings on satipatthana underwent
in the Satipatthana-samyutta, the Satipatthana Sutta in the Majjhima Nikaya appears as an
unpredictable quantum leap. It is instructive to compare this with the Anapanasati Sutta. This
contains no new teachings, being merely a presentation of material from the Anapanasati-
samyutta with a more elaborate setting. In other words this is a more normal teaching, taught
more often. In the Satipatthana Sutta we see, not a dainty step up in size like in anapanasati, but a
massive blow-out in several directions at once. First, each of the four satipatthanas is expanded
into a detailed exercise or series of exercises, few of which occur elsewhere in the context of
satipatthana. Then each exercise is followed by a lengthy section dealing with insight. This is
substantially similar to the insight section in the Samudaya Sutta of the Sarvastivada Samyukta.
Given the strong conservatism that is obvious in the treatment of satipatthana in every other
context, I find it very difficult to accept that on this one occasion the Buddha departed so
radically from his policy. This leads me to suspect that the Satipatthana Sutta as we have it is the
end result of a process of textual accretion. And when we look at teachings on satipatthana
outside the four Nikayas we do indeed find some concrete support for this idea.

CHAPTER 1 3 : EARLY ABHIDHAMMA

I take the unusual step of considering the Abhidhamma texts before the Satipatthana Suttas, even
though this runs counter to the standard historical sequence that underpins most of this study.
This is for two reasons. One is, as I have hinted at before, there is reason to believe that some
aspects of the Abhidhamma material on satipatthana preserves more archaic features than the
Suttas. The second reason is that I want to deconstruct the readers' perceptions of what
satipatthana is all about. These perceptions are largely determined by modern interpretations of
the Theravada Satipatthana Sutta, and so it may be useful to consider as much alternative
material as possible before re-examining that text.

The Abhidhamma literature may be described as a formalized, scholastic systemization of the

Dhamma according to the perspective of the various schools. Whereas the Sutta and Vinaya
literature, so far as comparative studies of available material have shown, is largely common to
all schools, the Abhidhammas of the schools diverge widely. The most plausible reason for this is
that the Suttas and Vinaya largely derive from the pre-sectarian period, while the Abhidhamma
literature is sectarian. In our discussion of the GIST, however, we noted a significant exception to

150
this rule. There are t h r e e Abhidhamma texts t h a t all seem to be derived f r o m a common source,
which we call t h e '*Vibhanga Mula'. These texts are t h e Vibhanga of the Theravada, t h e
Dharmaskandha of t h e Sarvastivada, 361 and t h e Sariputrabhidharma of t h e Dharmaguptaka. 3 6 2 The
common core of these works seems to have been little m o r e t h a n a matika, a matrix or schedule of
doctrinal categories, furnished with explanations f r o m t h e Suttas and with word-definitions.
Since this work was evidently started before t h e schisms, while t h e Suttas w e r e n o t completely
finalized until after t h e schisms, it seems we m u s t accept some period of overlap between t h e
periods of compilation of the Suttas and t h e Abhidhammas.

VIBHANGA

The Vibhanga includes a discussion of satipatthana as one of a series of chapters dealing with t h e
37 wings to enlightenment. As usual, t h e discussion is divided in two, a 'Sutta Exposition' and an
'Abhidhamma Exposition'. Generally, t h e Sutta Expositions in t h e Vibhanga remain, as one would
expect, quite close to t h e Suttas. The corresponding chapter of t h e Sarvastivada Dharmaskandha
merely adds a few extra passages to the Vibhanga. At a later date t h e Vibhanga's Abhidhamma
Exposition was composed with m o r e distinctively abhidhammic and sectarian material. The
Dharmaskandha has no proper 'Abhidhamma Exposition' as such. We will firstly survey t h e Sutta
Exposition of the Vibhanga, and t h e n t h e additions to t h e Dharmaskandha, reserving a discussion
of the Vibhanga's Abhidhamma Exposition for a later chapter.

The Vibhanga starts out with t h e basic satipatthana formula, elaborated with t h e
i n t e r n a l / e x t e r n a l contemplation. The detailed structure t h e n unfolds as follows. Most of t h e text
is identical with t h e Suttas; specifically Abhidhammic material is shaded.

Table 13.1: The Structure of the Vibhanga

1. Body 2. Feelings 3. Mind 4. Dhammas

How to contemplate the body How to contemplate How to contemplate the How to contemplate the
internal? feelings internal? mind internal? body internal?
Body parts internal 3 feelings, then pleasant Mind with lust, etc. (16 5 hindrances & 7
spiritual/carnal, etc. kinds) internal enlightenment-factors
internal internal
Develops 'that nimitta', then Develops 'that nimitta', Develops 'that nimitta', Develops 'that nimitta',
compares with external then compares with then compares with then compares with
external external external
How to contemplate the body How to contemplate How to contemplate the How to contemplate
external? feelings external? mind external? dhammas external?
Body parts external 3 feelings, etc. external Mind with lust, etc. 5 hindrances & 7
external enlightenment-factors
external
Develops 'that nimitta', then Develops 'that nimitta', Develops 'that nimitta', Develops 'that nimitta',
compares with internal/external then compares with then compares with then compares with
internal/external internal/external internal/external
How to contemplate the body How to contemplate How to contemplate the How to contemplate
internal/external? feelings internal/external? mind internal/external? dhammas
internal/external?
Body parts internal/external 3 feelings, etc. Mind with lust, etc. 5 hindrances & 7
internal/external internal/external enlightenment-factors
internal/external
Thus a monk contemplates the body Thus a monk contemplates Thus a monk contemplates Thus a monk contemplates
internal/external feelings internal/external the mind internal/external dhammas internal/external
Defines: contemplates, abides, ardent, Defines Defines Defines
clearly comprehending, mindful,
world, covetousness, aversion,
remove

151
In the Vibhanga the body is treated just as the 31 parts. This is clearly a much more primitive
conception than the Theravada Satipatthana Sutta. The elements and corpse meditations, which
are found in all three Sutta versions, are also found in the DhammasanganT, so it is not clear why
they are not brought in here, unless the *Vibhanga Mula was older than these texts. The
enumeration of the parts of the body as 31 is also early, being shared in common with the four
Nikayas; but by the time the Khuddakapatha of the Khuddaka Nikaya was compiled (in Sri
Lanka?), the brain had been added to complete the now-classic 32. I will discuss the body
contemplations below, in the context of the Sariputrabhidharma. Feeling and mind in the
Vibhanga are the same as the Sutta. We might spend a little time here considering the treatment
of these factors, as the Vibhanga is typical of all the sources.

In addition to the usual threefold analysis of feelings, the satipatthana material introduces the
distinction between 'carnal' and 'spiritual' feelings. This distinction is not explained in the
context of satipatthana as such; the detailed discussion is in the Vedana-samyutta. 363 Since
'carnal' and 'spiritual' are unusual terms in this context, it seems likely that the Vedana-samyutta
passage was specifically intended to explain the Satipatthana Sutta material. This impression is
reinforced by another unusual feature, the inclusion of 'rapture' as a kind of feeling. Rapture is
not mentioned in the feeling section of the Satipatthana Sutta, but it does fall under feelings in
anapanasati. Again, this is an unusual if not unique usage, and suggests that the Vedana-samyutta
passage was intended to synthesize and explain the feelings sections in both satipatthana and
anapanasati.

The explanations that concern us here are as follows. Carnal feelings are those connected with
the senses. Spiritual rapture is in the first twojhanas, spiritual pleasant feeling is in t h e first three
jhanas, while spiritual neutral feeling is in the f o u r t h j h a n a . Spiritual painful feeling is described
as depression arising as one longs for the peaceful liberations one has not yet realized (a feeling I
grow more familiar with as this book grows longer!). Since the spiritual feelings are primarily
defined in terms of j h a n a there seems little doubt that this non-standard classification was
introduced in satipatthana specifically to emphasize the importance of the experiences of refined
bliss associated with samadhi. Just as we cannot know darkness until we have seen the light, we
cannot comprehend the nature of everyday sensual feelings until we have t h e perspective of
contrast. Since this perspective is a special and universal feature of the satipatthana material, it is
unreasonable to insist that experience of t h e subtle feelings ofj h a n a are a dispensable part of t h e
full spectrum of mindfulness meditation.

The contemplation of mind speaks first of understanding the mind with and without greed,
anger, and delusion. Normally the abandoning of this classic triad indicates arahantship, b u t
there is no need to assume that here. Sometimes this kind of phrasing is used in straightforward
samatha contexts. A passage from the Anguttara is worth quoting here:

'On an occasion, friends, when a noble disciple recollects the Buddha, on that occasion his mind is
not overwhelmed with lust, his mind is not overwhelmed with anger, his mind is not
overwhelmed with delusion. At that time his mind is upright - departed, released, and risen f r o m
greed. "Greed" is a t e r m for t h e five kinds of sensual pleasures. That noble disciple abides with a
heart totally like the sky, vast, exalted, measureless, free of hatred and ill-will. Having made this
the support, some beings here are purified.' 364

Notice the pervasive similarities with satipatthana, especially contemplation of mind: the
practice is a 'recollection' (anussati); the term 'mind' (citta) is repeatedly used together with lust,
hatred, and delusion; doing this, one's mind is 'released'; and the result is the 'purification of

152
beings'. There seems to be little doubt that the subjective aspect of the contemplation of mind is
similar to the six recollections.

The overall context, the progressive structure of the discourse, and t h e inclusion of t h e mind
'compressed' (by sloth) and 'scattered' (by restlessness) all suggest that here we are basically
dealing with the abandoning of the hindrances on the plane of samadhi, an interpretation
confirmed by t h e commentary. Here again, as in the contemplation of feelings, a distinctive facet
of all the satipatthana material is t h e direct experience of the 'exalted' mind, the 'unexcelled'
mind, the mind 'in samadhi', the 'released' mind - all synonyms f o r j h a n a .

These two sections share a common syntactical structure. For example: 'When feeling a pleasant
feeling, one understands "I feel a pleasant feeling." ' Or in the contemplation of mind: 'One
understands mind with lust as "mind with lust".' This reflexive structure is shared also with
anapanasati: 'When breathing in a long breath, one understands "I am breathing in a long
breath".' The phrasing in 'quotation marks' (representing the Pali particle iti) was perhaps what
prompted some schools to equate satipatthana meditation with mental noting. But this would be
a naively literal interpretation. Similar usages are found, for example, in the standard passage on
the formless attainments. Due to the idiomaticness of the Pali, this is difficult to translate;
literally it would b e : ' "Space is infinite", one enters & abides in the field of infinite space.' Usually
translators would say something like: 'Aware that "Space is infinite", one enters & abides in t h e
field of infinite space.' Obviously here the meditator has passed well beyond thinking or noting
anything. The use of iti with repetitions in such contexts seems rather to intimate the reflexive
nature of meditative contemplation. One is not merely knowing the feeling, but one is conscious
that one is knowing the feeling.

The Vibhanga section on contemplation of dhammas has the hindrances and enlightenment-
factors only, a pairing t h a t is by now becoming familiar. Unlike the Satipatthana Sutta, here there
are no introductory and concluding sentences to separate and define each section, such as: 'And
how does one abide contemplating a dhamma in the dhammas regarding the five hindrances?'
The hindrances and enlightenment-factors simply run on into each other. (Such sentences are
found in the Dharmaskandha, but the Sarvastivada Smrtyupasthana Sutra has only the
concluding sentences; this is perhaps t h e only point at which the Dharmaskandha is closer to the
Theravada Satipatthana Sutta.) Otherwise, however, t h e phrasing is identical with the Sutta.

The contemplation of t h e five hindrances and seven enlightenment-factors would seem to be

primarily a samatha practice, and invites consideration as to how exactly this section differs f r o m
the previous sections. For example, the first two hindrances are sensual desire and ill-will, which
seem to simply repeat the contemplation o f ' m i n d with lust' and 'mind with anger'. But if we look
closely, some subtle differences make themselves evident. The first thing is that in the
contemplation of mind, the direct object of contemplation was the mind itself; the qualities of t h e
mind, such as lust, hatred, etc., function as adjectives qualifying the mind. This suggests that the
prime focus in this contemplation is the nature of knowing itself, the cognitive power of
awareness in various conditions. In the contemplation of dhammas, the direct object of
contemplation is not the mind, but the associated mental qualities - sensual desire, etc.

I will not dwell on this distinction, for some might feel like it is splitting hairs, although I
personally think it is of considerable philosophical and psychological significance. Far more
obvious is the introduction of a more detailed investigation into causality in this section. This
may be most graphically represented in a table. We may compare the contemplation of mind and
dhammas, keeping the translation as literal as possible.

153
Table 13.2: Contemplation of Mind and Dhammas

Contemplation of Mind Contemplation of Dhammas

One understands mind with lust as There being internal sensual desire, one understands
'mind with lust.' 'There is for me internal sensual desire.'
One understands mind without lust There not being internal sensual desire, one understands
as 'mind without lust.' 'There is not for me internal sensual desire.'
And one understands how the unarisen sensual desire
comes to arise.
And one understands how the arisen sensual desire comes
to be abandoned.
And one understands how the abandoned sensual desire
comes to not arise in the future.

It becomes obvious that the chief difference in the mode of contemplation is t h e final three
sentences in the contemplation of dhammas. This is clearly an investigation into causality, into
the reasons behind the rise and fall of the various good and bad qualities; an investigation,
moreover, that is attuned precisely to removing the cause and abandoning forever the bad
qualities. For the enlightenment factors, of course, the situation is changed: one is to understand
how the enlightenment-factors come to arise, and how they are developed to fulfilment. As is
made clear elsewhere, the causes for abandoning the hindrances are precisely the
enlightenment-factors; and the causes for obstructing the enlightenment-factors are precisely
the hindrances. Thus these two sets of dhammas are intimately intertwined, the light and shadow
of the mind. This, then, is the prime distinguishing feature of the contemplation of dhammas, the
investigation into causality.

This, of course, is vipassana, and it is here in the contemplation of dhammas that vipassana finds
its rightful home in satipatthana. We have seen that in anapanasati, too, the contemplation of
dhammas may be read as contemplating the impermanence and interrelationships of the
hindrances and enlightenment-factors. But this is, of course, not a 'dry' vipassana, not an insight
that is divorced or separated f r o m samatha. Quite the opposite: it is an insight that emerges f r o m
understanding the principles of samadhi, why the mind is sometimes radiant and peaceful and
sometimes caustic and fractured. But t h e mind in meditation, we learn through satipatthana
practice, is no different in nature f r o m the mind outside of meditation: it's just the mind
responding to conditions. So learning to understand the process of meditation one is learning to
understand the mind. As the insight through contemplation of dhammas matures and deepens it
will naturally broaden to encompass all states of mind, all that is knowable, and will ripen in the
deepest insights. So the presentation of contemplation of dhammas in the Vibhanga is highly
convincing as a natural depiction of the meditative process.

In the Vibhanga each section is integrated with the internal/external contemplation, here
elaborated slightly f r o m the standard form found in the Samyutta. One is to cultivate, develop,
make much of, and clearly define body contemplation internally before progressing to body
contemplation externally, and so on each stage step by step. Then follows a word definition,
obviously a later, distinctively abhidhammic addition. Most of the definitions, or rather, strings of
synonyms, are standard enough. 'World' is defined thus: 'This very body [feeling, mind, dhamma]
is the world; also t h e five aggregates associated with grasping are the world.' The mechanical
nature of some of these definitions is shown up by the gloss on domanassa, which follows the
normal meaning of 'sadness', failing to recognize the contextual meaning here of 'aversion'.
Certain Sutta material is absent from the Vibhanga: there are no similes, which in Abhidhamma
literature is to be expected. More significantly, there is no vipassana refrain, in striking contrast
to the well-integrated internal/external refrain.

154
Seeing the absence of so much sutta material f r o m the Vibhanga's presentation, we wonder w h a t
could be the reason. One possibility is simply that the compilers of the Vibhanga were slack; but
the work as a whole is well edited and does not generally give the impression that the compilers
were unable to read a well-known Sutta. It might be the case that t h e absent material is not really
appropriate in Abhidhamma context. This is, of course, true in t h e case of, say, the setting and the
similes. But as it happens, much of the absent material is quite at home in the Abhidhamma: the
elements, aggregates, sense media, and truths, each of which have their own chapter in the
Vibhanga, and are treated constantly throughout the Abhidhamma. Perhaps, then, the opposite is
the case: the compilers deliberately removed t h e Abhidhamma-style material. This would seem
like a strange procedure; certainly repetition was never an obstacle for the Abhidhammikas.
Moreover, some of the absent material, such as the charnel ground contemplations, or t h e
awareness of activities, is not distinctively Abhidhammic. Particularly curious is the absence of
the opening questions defining the five hindrances and the enlightenment-factors. This kind of
question is absolutely characteristic of the Abhidhamma method, which is indeed sometimes
defined as the 'with-questions' method. So to remove t h e m in order to fit them in an
Abhidhamma context is unthinkable. Another problem is the location of the chosen practices. In
the contemplation of t h e body, the Vibhanga has what is the fourth practice out of a total of
fourteen in the Sutta; in the contemplation of dhammas, it has the first and fourth out of five. It
seems bizarre that a redactor would somehow remove all the practices leavingjust the fourth.
Surely we would expect, if the Vibhanga results from the culling of t h e Satipatthana Sutta, that
we would have the first practice left over, which would then in typical style imply that the rest
were to be filled out. The conclusion is inescapable: the absence of material in the Vibhanga
stems not f r o m a loss f r o m the Satipatthana Sutta, but because t h e Vibhanga compilers were
working with a shorter text.

DHARMASKANDHA

In most of the above aspects t h e Dharmaskandha seems practically identical with the Vibhanga.
The correlations are very strong, even down to the details. For example, the Pali has t h e standard
phrase 'cultivates, develops, and makes much', followed by the emphatic abhidhamma phrase
'makes defined, well defined' (svatthitam vavattheti); the Chinese exactly follows suit. As well as
including practically all of the material in the Vibhanga the Dharmaskandha adds the following
extra material. The Dharmaskandha, unlike the Vibhanga, first gives the setting at SavatthT, just
as in the Suttas. This seems to imply they were sneakily slipping t h e chapter in the mouth of the
Buddha; but that might be a misjudgement, for the Sarvastivada tradition generally was happy to
ascribe their canonical Abhidharma texts to various disciples; the Dharmaskandha is attributed to
Venerable Sariputta. Perhaps, then, the Sutta-style opening is just an editorial feature, an
indication that the basic opening text has been cut-&-pasted f r o m the Suttas. Interestingly, some
of the discourses used as basic texts in the Vibhanga were in fact spoken by Venerable Sariputta
(Saccavibhanga Sutta, Maha Hatthipadopama Sutta). The basic satipatthana formula contains the
standard auxiliary formula, rather than the abbreviated Sarvastivada version. The
internal/external contemplation is integrated from the start. Then it says that in the past,
present, and future, monks will practice the same way. This is reminiscent of the start of t h e
Sarvastivada Smrtyupasthana Sutra, and reflects the Sarvastivadin perspective on time. While the
Theravada Suttas sometimes have similar statements, the Sarvastivada seem hammer t h e m home
at every opportunity, as we have seen in their Samyukta also. After this introduction begins the
detailed treatment; the structure can be summarized as below. Most of the sections are taken
f r o m the Satipatthana Suttas, with some rearrangements and additions. Cases where t h e source is
specifically the Sarvastivada Smrtyupasthana Sutra are marked as such. Material that is more-or-
less similar to the existing Satipatthana Suttas is left clear; material that occurs in the Suttas b u t

155
not in the context of satipatthana is lightly shaded; specifically abhidhammic additions are more
heavily shaded.

Table 13.3: The Structure of the Dharmaskandha

1. Body 2. Feelings 3. Mind 4. Dhammas

a. What is 'contemplating the a. What is 'contemplating the a. What is a. What is 'contemplating the
internal body, ardent, etc.?' internal feelings, ardent, etc.?' 'contemplating the internal dhammas, ardent, etc.?'
internal mind, ardent,
etc.?'
b. Defines internal b. Defines internal b. Defines internal b. Defines internal
c. Body parts (no simile) c. Three feelings (as Sarv Sutra, in c. Mind with/without c. Sensual desire present,
these modes: plain, lust, etc. (11 pairs) absent, arising, abandoning,
bodily/mental, carnal/spiritual, future non-arising (1st
sensual/non-sensual) hindrance)
d. Defines: contemplates, d. Definitions (as 1.d) d. Definitions (as 1.d) d. Definitions (as 1.d)
abides, ardent, clearly
comprehending, mindful,
covetousness, aversion
e. 'The internal hindrances of
ill-will, sloth & torpor,
distractedness, and doubt are
spoken of similarly.'
e. Six elements (as Sarv Sutra; f. Eye, visual forms, and fetters
no simile) present, absent, arising,
abandoning, future non-arising
(as Sarv Sutra)
f. Definitions (as 1.d; from here g. Definitions (as 1.d)
text usually abbreviates, saying
they 'are spoken of as before')
h. Remaining sense media
treated similarly
i. Enlightenment-factor of
mindfulness present, absent,
arising, developed to perfection
j. Definitions (as 1.d)
k. 'The remaining six internal
enlightenment-factors are
spoken of similarly'
g. Tribulations (vipassana e. Tribulations e. Tribulations l. Tribulations
refrain; see SN 22.122/SA 259)
h. Definitions (as 1.d) f. Definitions (as 1.d) f. Definitions (as 1.d) m. Definitions (as 1.d)
(repeats 1.a-h, changing (repeats 2.a-f, changing 'internal' (repeats 3.a-f changing (repeats 4.a-m, changing
'internal' to 'external') to 'external') 'internal' to 'external') 'internal' to 'external')
(repeats 1.a-h, changing (repeats 2.a-f, changing 'internal' (repeats 3.a-f, changing (repeats 4.a-m, changing
'internal' to 'internal and to 'internal and external') 'internal' to 'internal 'internal' to 'internal and
external') and external') external')

The specification of each of the four satipatthanas is very similar to the Vibhanga. All the changes
to the Vibhanga are additions; and virtually all those additions are the same as the Sarvastivada
Smrtyupasthana Sutra in particular. In body contemplation, the Theravada Satipatthana Sutta
has four elements while the Sarvastivada has six. This is in line with the Sarvastivadin preference
to add space and cognition to the usual four, even though cognition is obviously incongruous in
body contemplation. The section on feelings is also shared with the Sarvastivada Smrtyupasthana
Sutra but not, as far as I know, with any Pali context. The intrusion of the sense media between
the hindrances and the enlightenment-factors weakens the unity of the Vibhanga's presentation.
The six sense media are also presented exactly as the Sarvastivada (although the position of the
section is the same as t h e Theravada). The Sarvastivada Smrtyupasthana Sutra presents the sense
media identically to t h e hindrances (present, absent, arising, abandoning, f u t u r e non-arising)
whereas the Theravada has a specific phrasing for the sense media; one understands, for example,

156
t h e eye, visual forms, and t h e f e t t e r t h a t arises d e p e n d e n t on t h e m both. Here t h e Theravada is
m o r e apt; t h e Sarvastivada version probably arose t h r o u g h a mechanical misapplication of t h e
h i n d r a n c e phrasing to t h e sense media. It thus seems clear t h a t t h e additions to t h e
Dharmaskandha over t h e Vibhanga are specifically Sarvastivadin. The exception to this is in
contemplation of mind, w h e r e t h e t h r e e extra factors added to t h e eight standard in t h e
Theravada do n o t correspond particularly closely with t h e Sarvastivada Smrtyupasthana Sutra.
Here is a table comparing these two Sarvastivada sources, compared with t h e Sanskrit text of t h e
Sramanyaphala Sutra, probably of t h e Sarvastivada. The factors t h a t are similar or identical with
t h e Theravada are shaded. I include t h e factors in contemplation of m i n d listed by Asanga in t h e
Sravakabhumi for comparison.

Table 13.4: Four Versions of the Contemplation of Mind

Smrtyupasthana Sutra Sramanyaphala Sutra Dharmaskandha Sravakabhumi

W i t h / w i t h o u t lust W i t h / w i t h o u t lust W i t h / w i t h o u t lust W i t h / w i t h o u t lust
W i t h / w i t h o u t anger W i t h / w i t h o u t anger W i t h / w i t h o u t anger W i t h / w i t h o u t anger
W i t h / w i t h o u t delusion W i t h / w i t h o u t delusion W i t h / w i t h o u t delusion W i t h / w i t h o u t delusion
Defiled/undefiled
Contracted/scattered Contracted/scattered Contracted/scattered Contracted/scattered
Slothful/energetic Slothful/energetic Slothful/energetic
(linam/pragrhitam)
Small/great Small/great
Restless/unre stless Distracted/undistracte d Distracted/undistracted
(uddhatam/anuddhatam)
Untranquil/tranquil Tranquil/untranquil Tranquil/untranquil
(avyupasantam/vyupasantam)
Lower/higher
Developed/undeveloped

S a m a d h i / n o samadhi S a m a d h i / n o samadhi S a m a d h i / n o samadhi S a m a d h i / n o samadhi

Undeveloped/well developed Developed/undeveloped Developed/undeveloped
Released/unreleased Released/unreleased Released/unreleased Released/unreleased

Thus t h e Sramanyaphala Sutra version is m u c h closer to t h e Dharmaskandha t h a n t h e

Smrtyupasthana Sutra. The only significant difference is t h e omission of 'small/great', b u t since
this is f o u n d in most versions (including t h e Pali) and is likely to be authentic, it seems t h a t h e r e
we have simply a loss on t h e p a r t of t h e Sramanyaphala Sutra, r a t h e r t h a n an addition to t h e
Dharmaskandha. On t h e other hand, t h e Sravakabhumi, while in all other respects identical,
agrees with t h e Sramanyaphala Sutra in omitting this pair, suggesting t h a t Asanga m a y have been
using a version of t h e contemplation of m i n d f r o m t h e same lineage as this version of t h e
Sramanyaphala Sutra. Leaving this pair aside, t h e four pairs in t h e Dharmaskandha f r o m
'contracted' to 'quiet' are virtual synonyms. The Pali commentaries say contracted means 'with
sloth & torpor', while scattered means 'with restlessness'. It is possible t h a t these extra t e r m s in
t h e Sarvastivada started life as commentarial glosses on 'contracted/scattered', which w e r e later
read into t h e text. This would suggest t h a t 'small/great' (amahaggatam/mahaggatam) should
belong after t r a n q u i l / u n t r a n q u i l . If we f u r t h e r recognize t h a t t h e Sarvastivada's
'developed/undeveloped' is another virtual synonym of t h e Theravada's 'excelled/unexcelled',
t h e n t h e two versions of contemplation of m i n d in t h e two schools b e c o m e very close indeed. In
any case, t h e comparison implies t h a t h e r e t h e Dharmaskandha is m o r e authentic to t h e old
Sarvastivada tradition, while t h e Smrtyupasthana Sutra version has drifted f u r t h e r away. Apart
f r o m helping us to trace textual affiliations, though, t h e variations are all trivial.

157
Far more significant is t h e addition of a vipassana refrain at t h e end of each section:

'Furthermore the monk, with regard to this internal body, observes and contemplates all their
many tribulations, namely: this body is like a sickness, like a boil, like a dart, troublesome,
impermanent, suffering, empty, not-self, changing, wearisome, a great entanglement. It is of a
nature to be lost and to decay, rapidly and incessantly becoming weak, not enduring. It cannot be
relied on or trusted. It is of a nature to change and decay...'

This is evidently derived from passages in the Nikayas, especially the Khandha-samyutta. 365 It is
not found in the early texts in t h e context of satipatthana, and evidences the growing tendency
to treat the satipatthanas in terms of the aggregates. It is clearly a different passage than the
vipassana refrain of the Theravada Satipatthana Sutta and cannot have come f r o m the same
source. Notice that, while the Theravada Satipatthana Sutta places its vipassana refrain at the end
of each exercise, the Dharmaskandha's vipassana refrain occurs only at the end of each section,
thus being less closely integrated within the practice as a whole. It is also less well integrated
than the internal/external refrain, in which respect the Dharmaskandha is consistent with the
Vibhanga.

Thus far the Sutta material. As the table shows, the additional specifically abhidhammic content
is restricted to word definitions. The definitions of internal and external are interestingly
different from the Theravada and are clearly sectarian. For the Theravada, as for the early Suttas,
'internal' means pertaining to oneself, especially one's own body and mind, while 'external' is the
bodies and minds of others. But for the Dharmaskandha 'internal' is 'one's own body [etc.], which
in the present continuum has been gained and not lost.' In other words 'internal' refers to this
life. 'External' is 'one's own body [etc.], which in the present continuum has not yet been gained
or is already lost, together with the bodily phenomena of others, possessing spirit'. This refers to
past and f u t u r e lives. The phrase 'possessing spirit' is odd; the phrase might render savinnanaka,
which we could relate to the familiar idea of '[kammically] acquired' (upadinna), i.e. loosely
'organic' or 'sentient' matter. Anyway, this modification exhibits the characteristically
Sarvastivadin concern for time, which as we have seen might be derived f r o m their Samyukta.

The definitions of internal and external say t h a t 'dhammas' here is the aggregates of perception
and conceptual activities. This significant redefinition was also adopted by the Theravada
commentaries, and has by today become standard. Here we see a common pattern - the various
sectarian schools, despite their mutual polemics, often share more in common with each other
than they do with the Suttas. It should be obvious that there is nothing in the actual description
of 'dhammas' here that requires or even suggests such a definition. How, for example, can the six
sense media be explained as perception or conceptual activities? Any attempt to explain this
away is missing the point; for t h e motivation behind this definition is not to draw out the
implications of the meaning of 'dhammas' here, but to integrate the four satipatthanas with the
five aggregates. 366 This follows f r o m a crucial assumption of the Abhidhamma project: that the
various doctrinal frameworks of the Suttas each offer a different way of categorizing the same
reality; and that it is therefore possible to systematically equate all the dhammas in one
framework with those in any other. The end result of this process was t h e complex Abhidhamma
matikas, which subsequently displaced the earlier frameworks. Personally I find this to be a
dubious project in general, but this is not the time to discuss the matter in detail. Suffice to notice
that the original core frameworks for this project are t h e five aggregates, the six sense media,
and, less standardized, the elements. Even in t h e Suttas we see a tendency to treat the various
faculties in a similar manner, including the five spiritual faculties, which accordingly begin to
spill over f r o m the fourth noble t r u t h to the first three. 367 Now we see the same pattern emerging
in the satipatthanas. A group originally part of the fourth noble truth, t h e path, is being equated
with dhammas characteristic of the first noble truth, the five aggregates. The incongruity of the

158
results reflects t h e inappropriateness of the method. As mentioned above, I think that the
primary sense of 'dhammas' here is not 'phenomena' but rather 'principles'; not 'what is there'
but 'how it works'. While 'phenomena' is one of many meanings of 'dhammas' well established in
the Suttas, there was a pronounced drift in the Abhidhamma period to emphasize this meaning at
the expense of others, and a corresponding misinterpretation of relevant Sutta contexts.

The other addition is a word definition of the major terms in the basic satipatthana formula; this
is merely a list of synonyms in typical Abhidhamma style. The gloss for anupassana lists a register
of terms denoting wisdom, including vipassana; this is much the same as the Vibhanga. It is
striking that this word definition, though it defines the basic formula, is repeated after each
section throughout the text - except the basic formula. This is clearly incongruous, an example of
rampant abhidhammic over-systemization. Sometimes the abhidhamma scholars behave
uncannily like a computer error.

So as compared with the Sutta Exposition of the Vibhanga the differences in the Dharmaskandha
are:

1) All additions, no subtractions;

2) Often incongruous (setting, six elements, sense media, dhammas as
perception/conceptual activities, definitions);
3) Sometimes hinting at sectarianism (past, present, future; influence of Sarvastivada
Smrtyupasthana Sutra).

It should hardly need saying that these considerations all suggest that t h e Dharmaskandha here is
later than the Vibhanga. I would suggest that they both shared t h e same pre-sectarian text, the
*Vibhanga Mula; the Sarvastivadins finalized the editorial process on that section for the
Dharmaskandha, while the Theravadins, content with the simple version for the Sutta Exposition,
concentrated on writing the new, more overtly sectarian Abhidhamma Exposition.

The chief difference in perspective between the two is clear. Apart f r o m the contemplation of
dhammas, there is no overt vipassana material in the Vibhanga. There is no mention of rise and
fall, no six elements, no sense media, and no dhammas as perception/conceptual activities. We
can be as confident as we could wish that the vipassana material was added later. The significance
of these additions might become clearer with a more complete examination of the
Dharmaskandha. But the most striking point about this material, especially that held in common
between t h e two, is that both the content of the exercises and the basic form of the refrains are
much simpler than the Satipatthana Suttas. This clearly - and startlingly - suggests that t h e
*Vibhanga Mula may be earlier than the Satipatthana Sutta.

SARIPUTRABHIDHARMA

This is an abhidhamma text on a larger scale than the Vibhanga and the Dharmaskandha. It seems
to represent the whole field of abhidhamma in the Dharmaguptaka system, containing material
comparable to that found in the Theravada Dhatukatha, Patthana, Puggala Pannatti, and
DhammasanganT, as well as the Vibhanga. The matikas of the Vibhanga and Dharmaskandha are
discernable, though the form is more divergent. Thus the work as a whole is clearly later than the
Vibhanga and Dharmaskandha, and the treatment of satipatthana bears this out.

159
Table 13.5: The Structure of the Sariputrabhidharma

1. Body 2. Feelings 3. Mind 4. Dhammas

a. What is contemplating the a. What is contemplating the a. What is contemplating the a. What is contemplating the
body? feelings? mind? dhammas?
b. The body is the four b. Feelings based on six b. Defines mind b. Everything apart from
elements, produced by senses. body, feeling, mind
parents, nourished by food
c. What is contemplating the c. What is contemplating the c. What is contemplating the c. What is contemplating the
internal body? internal feelings? internal mind? internal dhammas?
d. Contemplating internal d. Contemplating internal d. Contemplating internal d. Contemplating internal
body as impermanent, etc.; as feelings as impermanent, etc.; mind as impermanent, etc.; as dhammas as impermanent,
conditioned, with dependent as conditioned, with conditioned, with dependent etc.; as conditioned, with
origination list down to 'name dependent origination list origination list down to complete dependent
& form > sense media'. down to 'contact > feeling'. 'conceptual activities > origination list.
cognition'.
e. Body postures
f. Movements
g. Anapanasati
h. Parts of the body
i. Four elements
j. Food
k. Space
l. Nine orifices
m. Defines internal, e. Defines internal; other e. Defines 'internal mind' e. Defines 'internal
contemplates, and each term definitions 'as said above' dhammas'
in the standard formula
n. As 1.d, changing 'internal' f. As 2.d, changing 'internal' f. As 3.d, changing 'internal' to e. As 4.d, changing 'internal'
to 'external'. to 'external'. 'external'. to 'external'.
o. Defines 'external'; other g. Defines 'external'; other g. Defines 'external mind' f. Defines 'external
definitions 'as said above' definitions 'as said above' dhammas'
p. As 1.d, changing 'internal' to h. As 2.d, changing 'internal' h. As 3.d, changing 'internal' to g. As 4.d, changing 'internal'
'internal/external'. to 'internal/external'. 'internal/external'. to 'internal/external'.
q. Charnel ground i. Three feelings; then i. Mind with/without lust, etc. h. Five hindrances, their
pleasant carnal/spiritual, etc. arising and ceasing (but with
This is contemplating 'ignorance & delusion' added
internal/external feelings between ill-will and sloth &
torpor)
i. Sense media and fetters,
their arising and ceasing
j. Enlightenment-factors,
their arising and ceasing
k. Four noble truths
r. Contemplates the nature of j. Contemplates the nature of j. Contemplates the nature of l. Contemplates the nature
rise, fall, and both, for rise, fall, and both, for rise, fall, and both, for of rise, fall, and both, for
awareness and detachment awareness and detachment awareness and detachment awareness and detachment
from the body from feelings from the mind from dhammas
s. Repeats basic formula for k. Repeats basic formula for k. Repeats basic formula for m. Repeats basic formula for
internal, etc. internal, etc. internal, etc. internal, etc.

The somewhat eccentric structure of this text becomes clearer if we recognize that the main
paradigm is exemplified in the contemplations of feelings, mind, and dhammas; body is
divergent, so we will leave that until later. The first question, 'what is contemplating the feelings
[etc.]' is answered by defining feelings [etc.]. This doesn't really answer the question; presumably
the question was originally intended to cover the whole section, and the definitions were inserted
later. The definition of the body is standard. The definition of feelings is also standard, although
the Suttas do not treat feelings in satipatthana as based on the six senses; this shift is shared in
common with the Theravada commentaries. The definition of mind is a standard Abhidhamma
list; it is similar to the treatment of contemplation of mind in anapanasati in t h e
Patisambhidamagga. The definition of dhammas is similarly late, and is similar in meaning to the
Dharmaskandha, but the phrasing is identical to the Patisambhidamagga. Internal contemplation
is described in terms similar to the vipassana refrain f r o m the Khandha-samyutta, just as the

160
Dharmaskandha. It is not clear whether the Sariputrabhidharma borrowed f r o m t h e
Dharmaskandha or they both borrowed f r o m t h e Khandha-samyutta. The vipassana refrain is
expanded by adding factors f r o m dependent origination, appropriately adjusted in each case. This
may have been inspired by the Satipatthana Vibhanga Sutta, which as we have seen treats
causality in satipatthana in terms similar to dependent origination; however the specifics are
largely different. The whole first half of these sections, describing 'internal, etc.' in terms of
vipassana is largely an addition; notice how vipassana now becomes integrated f r o m the start of
the meditation, rather than being left until the end as t h e Dharmaskandha. Strangely, the
authentic satipatthana material is presented after the additions. This second half of each section,
featuring the meditation objects and the refrain, is very similar to the Theravada Satipatthana
Sutta, and presumably was influenced by the (now lost) Dharmaguptaka Satipatthana Sutta. The
most significant difference f r o m the Theravada is the absence of the five aggregates f r o m
contemplation of dhammas.

The structure of t h e section on body contemplation diverges from the pattern of the other three.
Specifically, the list of meditations has been split in two, with the bulk of t h e m inserted
awkwardly in the 'internal' section, while the charnel ground contemplations alone follow t h e
internal/external section. The charnel ground contemplations are described in the Suttas in
terms of comparing one's own body with a dead body. This probably suggested placing this
exercise after the internal/external section. This is clearly in line with the Suttas; b u t t h e
problem is that then the other exercises had to be fit in somewhere. Placing t h e m under internal
contemplation implies that they may not be practiced externally, which differs f r o m the Suttas.
The splitting of these exercises likely occurred at the Abhidhamma stage, and may not reflect the
Dharmaguptaka Satipatthana Sutta. However, we shall see that a similar distinction is made in the
Prajnaparamita, although there the charnel ground contemplations are said to be external, rather
than internal/external.

The long list of body contemplations falls into three natural divisions. The first two exercises are
very similar, basically 'awareness of movements'. I have discussed t h e role of these preparatory
exercises earlier. In the standard passages such as the gradual training, only the passage on clear
comprehension occurs. The four postures pericope is much less common. Since the two largely
overlap, the inclusion of both would seem to be redundant; however all the recensions of t h e
satipatthana material either include both of these pericopes or neither. There are some
considerations that suggest that the four postures might be more authentic, namely that it always
appears before clear comprehension, and that t h e phrasing is more similar to the other sections
of satipatthana. It has a more generalized, meditative scope, rather than specifically illustrating a
kind of lifestyle training. Perhaps this simpler, vaguer passage was felt in need of concrete
illustration, so the section on clear comprehension was brought in f r o m the gradual training.
However, in the absence of any textual support such considerations remain tenuous.

The next division consists of j u s t anapanasati, which rightly claims a unique position. Here this
follows on f r o m clear comprehension in a similar manner to the normal sequence of the gradual
training. It is possible that t h e two sections, of clear comprehension and anapanasati, were
brought in together f r o m some other source, although I do not know of any place where they
appear together in exactly this way.

In all versions of the Satipatthana Sutta, the normal sixteen steps have been abbreviated to four.
This is the only context where this happens (apart f r o m the related Kayagatasati Sutta, which we
will examine more fully below). In this case it seems very likely that the full sixteen-step version
is the more fundamental one. It clearly outlines a full sequence of meditative training. The first
tetrad describes t h e establishment of mindfulness on the basic meditation object and the
tranquilization of it; the second tetrad speaks of the development of bliss and rapture; the third,

161
attaining of samadhi; and the fourth, contemplation of impermanence. Just putting the first
tetrad on its own would seem to be incomplete; nowhere does the Buddha speak of meditation
merely for attaining bodily tranquillity. There is an interesting passage where Venerable Arittha
describes his practice of anapanasati as having dispelled sensual desire for things past and future,
and having dispelled perceptions of aversion towards things internally and externally, j u s t
mindful he breathes in and out.368 The removal of desire and aversion, and the reference to
internal/external, sounds very much like satipatthana. It appears that Venerable Arittha's
practice had not progressed beyond the simple observation of the breath, without proceeding to
samadhi and insight. Accordingly, the Buddha, while acknowledging that this was indeed
anapanasati, encouraged Venerable Arittha to fully develop anapanasati through the whole
sixteen steps. Thus I believe we are justified in seeing the truncated four-step version of
anapanasati in the Satipatthana Sutta as a late, possibly inauthentic, development.

The third division consists of diverse contemplative exercises, starting with the parts of the body.
Many such are taught throughout t h e Suttas, and such lists as in the Sariputrabhidharma appear
be early attempts to collate and organize these practices. We have seen that the Vibhanga h a s j u s t
the parts of the body, while the Dharmaskandha adds the six elements as well. The
Sariputrabhidharma also has the parts of the body and the elements, and in the same order; in
fact this sequence is maintained in all versions of the satipatthana material. The (inevitable!)
exception is t h e Prajnaparamita, which reverses the sequence of the body parts and elements,
which I take a s j u s t an editorial slip-up. Since the parts of t h e body is t h e only exercise mentioned
in all versions, and since it virtually always comes at the start of this division, it must clearly
make the greatest claim to authenticity, although the elements are not far behind. The two
practices are in fact very similar, for the elements, when taught in detail, are described by listing
the appropriate body parts: earth element is head-hair, body-hair, etc., water element is blood,
pus, etc. The charnel ground contemplations, too, appear in all t h e Sutta versions and always at
the end. The other exercises are clearly less important, although they are not as arbitrary as may
appear. They may have been influenced by Sutta passages that describe the internal space
element as those apertures where food is eaten, digested, and passed out. 369

These diverse exercises are treated in the Suttas in terms of both samatha and vipassana. The
central purpose of contemplation of the body parts is to abandon lust, which is samatha.
Sometimes this samatha aspect is made explicit, as when one is said to reach an 'attainment of
vision' such that, due to proper effort, one gains 'such a form of samadhi that, with the mind in
samadhi, one contemplates this very body' by means of the body parts. 370 But the body parts,
especially when subsumed under the elements, may also be contemplated as 'not mine', etc.,
which is vipassana. 371 The elements, as well as appearing in such vipassana contexts, are
frequently used as a basis for attaining samadhi, and even, it seems, can be used as a shorthand
reference to the four 'form' jhanas. The commentaries would like to draw a sharp distinction
between such 'kasina' practice and the vipassana investigation of the elements as they really are,
but it would be difficult to justify this f r o m the Suttas. The charnel ground contemplations, too,
are a powerful ground for contemplating the impermanence of life; but the practice is also
described as 'guarding a subtle basis of samadhi' (samadhinimitta).372 Thus all these practices
contain the potential for developing both peace and wisdom.

Of the descriptions of body contemplation in t h e seven different versions of t h e satipatthana

material, three mention only this third division (Vibhanga, Dharmaskandha, Ekayana Sutra),
while four mention all three divisions (Sariputrabhidharma, Theravada and Sarvastivada
Satipatthana Suttas, Prajnaparamita). Of these four long versions, the Sariputrabhidharma and
the Prajnaparamita are the best organized (leaving aside the reservations about the splitting of
the exercises into 'internal' and 'internal/external'). Unlike the Theravada Satipatthana Sutta

162
they preserve the standard sequence of placing the 'awareness of movements' division first; and
unlike the Sarvastivada they avoid dragging in unrelated practices.

We have seen above that there are good reasons for concluding that clear comprehension was not
originally regarded as a meditation as such. This conclusion is reinforced when we examine t h e
kinds of contexts where the body contemplations and anapanasati are taught elsewhere. For
example in the Maha Rahulovada Sutta, Venerable Rahula asks the Buddha to teach him
anapanasati. 373 The Buddha digresses with a long series of other meditations - the five elements,
including the body parts, and culminating in space; then t h e divine abidings; then ugliness; then
impermanence - before returning to anapanasati. Perhaps t h e reason for the Buddha's digression
was that Rahula's mind required preparation before it was mature enough to fully benefit f r o m a
subtle exercise like anapanasati. The Meghiya Sutta is similar in presenting a graduated series of
dhammas for maturing the mind. 374 It culminates with four meditations: ugliness for abandoning
lust; loving-kindness for abandoning anger; anapanasati for cutting off thinking; and
impermanence for uprooting the conceit 'I am'. This has always struck me as one of the most
sensible, balanced programs for meditation. Other variants occur, such as ugliness, anapanasati,
and impermanence. 379 In these and other contexts we see ugliness, the elements, etc., treated
alongside anapanasati as a straightforward meditation practice. But clear comprehension is
conspicuously absent in such contexts - which is why those who like to promote such an
approach to meditation must rely so heavily on the Satipatthana Sutta.

If, as the concordance of the texts suggests, the section on body parts, either alone or as the head
of the third division, was the original specification of body contemplation, why were t h e other
two divisions brought in, and why were they placed before the original section? I have suggested
that the influence of the gradual training is sufficient to account for t h e intrusion of clear
comprehension in first place. Another general consideration would have been to assemble in one
digestible text t h e various texts on body contemplation scattered through the canon.

More specifically, though, I wonder whether the introduction of anapanasati may have something
to do with the mysterious 'Case of the Murdered Monks'. 376 In this bizarre episode, the Buddha
spoke in praise of meditation on ugliness (i.e. the parts of the body) and then went into private
retreat. While he was away, many monks, because of an over-zealous application to this kind of
practice, became so morbidly disgusted with their bodies that they started killing themselves in
great numbers. When the Buddha emerged f r o m retreat he commented that the Sangha was
looking a bit depleted. Venerable Ananda told him what happened and, with mastery of
understatement, suggested that it would be a good idea for t h e Buddha to teach another
meditation subject to t h e monks. The Buddha called the monks together and recommended that
they practice anapanasati, a cool, peaceful, and pleasant abiding. As advice for meditators this is
all sensible enough, but as a historical episode it remains deeply inexplicable. But it could have
provided a motive for adding anapanasati before the body parts in the Satipatthana Sutta,
especially as this was gaining in size and prestige as the meditation compendium. The anapanasati
section begins with the words found also in the gradual training: 'Here a monk, gone to a forest,
to the root of a tree or to an empty hut...' This may then have suggested bringing in the clear
comprehension section f r o m the gradual training as a preparation.

It seems likely that in making such additions the Sariputrabhidharma was influenced by the (now
lost) Dharmaguptaka version of the Satipatthana Sutta. The closeness of relationship between the
Dharmaguptaka Suttas generally and t h e Sariputrabhidharma, so far as I know, still awaits
detailed investigation. Cheng Jianhua, however, has written a comparative study of the versions
of the Brahmajala Sutta. 377 This concludes that, while the Theravada and Dharmaguptaka versions
are very close, the Dharmaguptaka Sutta is even closer, in fact identical, with t h e relevant
material in the Sariputrabhidharma. It is therefore possible that the precise details of the list of

163
body contemplations in the Sariputrabhidharma may reflect the form of a now-lost
Dharmaguptaka version of the Satipatthana Sutta. This is far from certain, as both the Theravada
and Sarvastivada display considerable variation between the Sutta and early Abhidhamma
descriptions of satipatthana (although not of most other doctrines). However if it were the case, it
would suggest that the reversal of the positions of the first and second divisions in the Theravada
Satipatthana Sutta occurred after the Dharmaguptaka schism, over 200 years after the
Parinibbana. Similar considerations would apply to the other major differences between the
Sariputrabhidharma and the Theravada, particularly the Theravadin insertion of the five
aggregates into the contemplation of dhammas.

CHAPTER 1 4 : THE SATIPATTHANA SUTTAS

PRAJNAPARAMITA

There are three complete versions of the Satipatthana Sutta available, one in Pali and two in
Chinese. In addition there is a fourth, incomplete, version found in the large Prajnaparamita
Sutra. This is one of the foundational texts of the Mahayana, and is usually held to have been
compiled roughly 500 years after the Buddha's passing away. This text is available in versions of
varying length in Tibetan and Chinese, and a reconstructed Sanskrit text has been translated into
English by Conze. This version was constructed from a conflation of Sanskrit fragments in
comparison with Tibetan and Chinese translations, and so, while we can be reasonably confident
that the overall text is quite reliable, we should be careful about drawing conclusions on fine
details. According to Conze, the different versions mainly vary simply in the quantity of
repetitions, so as regards content we can effectively treat it as one work. A comparison with
various versions of the Prajnaparamita literature in Chinese reveals a remarkable consistency in
this passage. Since, as we shall see, the text is highly asymmetrical and unbalanced, almost
certainly resulting from a rough abridgement of a well-known earlier version of the Satipatthana
Sutta, it seems very likely that the Prajnaparamita passages on satipatthana all hark back to a
single original source. This version of the satipatthana material displays a refreshing simplicity
that may indicate that it lies close to the early sources. In the Tibetan tradition it is said that
there was a version of the Prajnaparamita written in Prakrit belonging to the Purvasaila and
Aparasaila schools.378 These schools are branches of the Mahasanghika, and this might indicate
that the Prajnaparamita account of satipatthana material was derived from a Mahasanghika text.
There is, however, no specially close relation between this version and the Ekayana Sutra, which,
if the Tibetan tradition is reliable, would tend to count against the Ekayana Sutra belonging to the
Mahasanghikas.

Although the Prajnaparamita, as with the Abhidhamma, is obviously later than the Agama Suttas,
I see no reason why the basic specifications of the common teachings should be less reliable as a
guide to the early doctrines. While it seems initially strange that these later styles of literature
should preserve early features, a little reflection can dispel such doubts. All the traditions
inherited the same mass of early Sutta material. The compilers of the Nikayas/Agamas were
concerned to fully arrange and present this material, and in doing so it is not surprising that they
should tinker around the edges, filling out and expanding texts so that nothing may be omitted.

164
But the compilers of the Abhidhamma and the Prajnaparamita had a different agenda. They were
concerned mainly with evolving their own special concepts and procedures, so in certain cases
may have simply left the early material unretouched. We saw in the discussion of the
Abhidhammas that this does seem to be the case in satipatthana. The core teachings remain
unembellished; the additions and elaborations are extraneous and hence easily discerned. A
similar situation seems to obtain in the Prajnaparamita. The teachings on satipatthana have been
basically 'cut-&-pasted' with minimal alteration.

Satipatthana is treated in two separate places in the Prajnaparamita. These occurrences are
widely separated and have no close textual relation. Both occur as part of a larger context
treating the way of practice of a Bodhisattva, and treat satipatthana as the first group of t h e 37
wings to enlightenment. As in the early Suttas, here the 37 are simply listed, with no attempt to
synthesize the groups into an overall progressive scheme of practice, such as t h e Abhidhamma
schools were to work out later.

The first passage begins by simply defining, as part of the Mahayana path of the Bodhisattva, the
four satipatthanas. 37 ? Then comes a variation on the standard passage:

'There the Bodhisattva dwells, with regard to the inward body, feelings, etc., in the contemplation
of t h e body, etc. But he does not form any discursive thoughts associated with the body, etc. He is
ardent, clearly conscious, and mindful, after putting away all worldly covetousness and sadness.
And that without taking anything as a basis. And so he dwells with regard to the outer body, t h e
inner and outer body, to feelings, thought, and dhammas.'

All the familiar features are there: the reflexive repetition 'body in [regard to] body'; the four
objects; contemplation; internal/external; and the standard auxiliary formula. The injunction not
to think thoughts associated with the body is, in the Pali, found in the Dantabhumi Sutta. We
noticed briefly above how this phrase also occurs in another Mahayana text, the
Pratyutpannabuddhasammukhavasthitasamadhi Sutra. We shall see below how in that context
the phrase serves as a springboard for a more characteristically Mahayana exposition of
satipatthana. Here the idea lies still in germ. The phrase rendered by Conze 'without taking
anything as a basis' in fact seems to be t h e phrase 'one dwells independent', found in t h e
Sarvastivada Samudaya Sutta and the Satipatthana Sutta.

The text goes on to ask how a Bodhisattva dwells with regard to the inward body in the
contemplation of the body. Then a list of body contemplation practices is given: awareness of the
four postures; clear comprehension in daily activities; anapanasati; elements; body parts; and
charnel ground. The content is identical with the body contemplation practices in the Theravada
Satipatthana Sutta. Each practice is described in virtually exactly the same words and phrases,
including the illustrative similes. Awareness of activities adds to the normal list the phrase 'and
when retiring for meditation'. There are occasional, very slight, drifts in the meanings of words.
In the charnel ground contemplation, the Theravada version says one should reflect 'This body,
too, is of the same dhamma, the same bhava.'. Here both dhamma and bhava are non-technical
words meaning simply 'nature'. But the Prajnaparamita translation has, instead of 'nature' for
bhava, 'own-being', obviously harking to an original svabhava. This key technical t e r m in
Abhidhamma philosophy would have been more familiar to t h e compilers of the Prajnaparamita
than the ordinary-language word bhava, so we presume it was substituted at some stage. Detailed
examination of t h e versions might tell us when and how this happened. Still, the shift is a mere
nuance, and I mention such trivial differences only to emphasize the importance of t h e
similarities.

165
While the differences in the phrasing of t h e particular exercises are miniscule, far more
significant divergences occur in the overall structure. Firstly, t h e Prajnaparamita lists the two
'awareness of postures' meditations first, before anapanasati. This is in common with every other
exposition of the path in all schools, except the Theravada Satipatthana and Kayagatasati Suttas.
And secondly, there is hardly any refrain. The exercises are presented in their bare form, and at
the end of each it is simply said, 'And that through non-apprehension.' I am not sure of the
Sanskrit original, but this seems to be a Mahayanist 'tag' added to identify with the special
Prajnaparamita perspective on emptiness. It could well be understood to mean the same thing as
'not grasping at anything in the world', but this is a matter of interpretation. This, and the
obvious substitution of'bodhisattva' for 'bhikkhu', are the only discernable Mahayanist elements.
Given the evident propensity of the Mahayana for florid elaborations, it does seem rather odd
that there is no real refrain supplied here. If the Mahayana authors had inherited a version of the
Satipatthana Sutta with a substantial refrain, as have the existing versions, one would expect
t h e m to alter and expand it, rather than slough it off. This suggests that they were perhaps
working with an early, bare-bones (if you'll excuse the pun!) version of the body contemplation
exercises. It is quite remarkable that this version is very close to t h e Theravada; even though the
Mahayana in general owes much more to the Sarvastivada and Mahasanghika traditions, still
their versions of satipatthana are more divergent.

It is a great shame that the text does not elaborate the details of the remaining exercises. It
simply gives t h e standard formula for contemplating feelings, internally, etc. This asymmetry of
presentation is suggestive of editing irregularities. The statement of the basic passage at the start
of the text gives the satipatthana formula for all four satipatthanas. Then it gives the details of
body contemplation, but does not repeat the relevant part of the satipatthana formula. But for
feelings, etc., the situation is reversed: it does not give the details of feeling contemplation, etc.,
but it does repeat the relevant part of the formula. It looks as though the detailed expositions
were lost, then the basic formula was mechanically inserted to fill the gap. Examination of the
various versions might shed some light on this, and might bring to light a more complete
exposition. However, as far as I am able to check various versions of t h e Prajnaparamita material
in the Chinese, they are remarkably uniform in this regard. Since the omission of detailed
consideration of the later three satipatthanas is manifestly asymmetrical and irrational, it seems
that these treatments must have sprung f r o m a single textual source, and have slavishly followed
its eccentric pattern.

Many years ago, Har Dayal commented that ' . t h e Buddhist authors have written a great deal only
about kayasatipatthana [body contemplation]; they dismiss the other three with a few words.' 380
Some therefore conclude that body contemplation was originally the only part of satipatthana.
However, this is an overly r a s h j u d g e m e n t . Given t h e enormously repetitive nature of Buddhist
texts in general, it is normal to find a text that deals with the first section in detail, then
abbreviates the rest. In fact, if the remaining three satipatthanas were a later accretion, we might
expect the reverse situation, that they should be explained in more detail. For example, in t h e
Prajnaparamita literature we are currently considering, the basic teachings such as satipatthana
are taught fairly briefly, but the special Prajnaparamita doctrines are elaborated at length. The
lack of detailed explanations of t h e latter satipatthanas, then, can be understood as a later loss.
This loss suggests that the authors of the Prajnaparamita were not overly concerned with
satipatthana as such; perhaps they simply assumed that their audience was already familiar with
the basic teachings.

We may now consider the second exposition of satipatthana in the Prajnaparamita. This defines
satipatthana, together with a list of other dhammas containing both early and late elements, as
'supramundane and not worldly'. 381 Then the basic passage on satipatthana is given, with no
detailed exposition of the various practices. The integrated internal/external contemplation is

166
followed by: 'With regard to the body he dwells as one who reviews its origination, its
disappearance, and both its origination and disappearance. He dwells as one who does not lean on
anything, and as one who (does not) grasp at the world.' This seems almost identical with t h e
vipassana refrain of the Theravada Satipatthana Sutta, although lacking the phrase 'Or
mindfulness is established that "There is a body", only for the sake of a measure of knowledge
and mindfulness.' It is also very similar to t h e Sarvastivada version of the Samudaya Sutta,
though lacking the specifications of the cause of each of the objects of satipatthana.

There are a number of textual details that diverge from the earlier passage. The Bodhisattva is
not mentioned. More important, t h e arrangement of the textual elements is different. In the
earlier passage the sequence was: contemplate the internal body, feelings, mind, and dhammas;
ardent, clearly comprehending, etc.; contemplate t h e external body, etc., and so on. Here the
sequence is: contemplate the body internally, externally, then internally & externally; ardent,
clearly comprehending, etc. In other words the auxiliary formula is added after each satipatthana
rather than being p u t at the end of the four. Of course, this variation has no significance of itself.
But taken together with other features - the separation of the two passages in the text, the
inconsistent use of 'Bodhisattva' - it suggests that that the two satipatthana passages are taken
f r o m different sources. That is to say, there is no reason to believe that the Prajnaparamita took
an earlier, longer text integrating these two passages and broke it up into separate pieces.

This reinforces one of our basic arguments of the formation of the Theravada Satipatthana Sutta.
This contains two main important aspects: the detailed list of meditation practices, and the
vipassana refrain. But there are at least three traditions that preserve these two aspects as
independent textual entities. In the Sarvastivada, the list of meditation practices is found in the
Satipatthana Sutta, while t h e vipassana refrain is found in the Samudaya Sutta. The
Prajnaparamita is similar. Even the Theravada preserves the meditation exercises in the
Abhidhamma Vibhanga and the vipassana refrain in the Samyutta Vibhanga Sutta and Samudaya
Sutta. Thus the treatment of these two aspects separately may represent an earlier tradition.

SARVASTIVADA SMRTYUPASTHANA SUTRA ( M A 9 8 )

The Smrtyupasthana Sutra of the Sarvastivada school is preserved in the Madhyama Agama of the
Chinese Tripitaka, translated from Sanskrit into Chinese in 389 C.E. by the Chinese monk
Sanghadeva. 382 The Smrtyupasthana Sutra is more elaborate in some respects than the Theravada
Satipatthana Sutta, and less elaborate in others. Hence it is probably neither earlier nor later, but
stems f r o m a slightly divergent tradition as the two new schools of the Theravada and the
Sarvastivada finalized t h e textual formulations of the teachings they had both inherited f r o m the
pre-sectarian period. This final editing probably occurred after t h e Sarvastivada schism, perhaps
150 years after the Buddha. I will ignore the various trivial differences in phrasing between the
individual doctrinal units ('pericopes') in the various versions and concentrate on the significant
differences in the choice of meditation exercises.

The Sarvastivada Smrtyupasthana Sutra opens in the same way as the Theravada Satipatthana
Sutta: the setting is at Kammassadamma in the Kuru country, and the teaching starts with the
statement on the 'path leading to convergence'. After this, however, the Smrtyupasthana Sutra
introduces the statement that all Tathagatas, past, future, and present realize enlightenment by
being established on the four satipatthanas, abandoning the five hindrances, and developing t h e
seven enlightenment-factors. 3 8 3 We have met such groupings frequently; this statement was likely
brought in f r o m SN 47.12/SA 498. We have already noticed that the extra emphasis on persistence
through time suggests the sectarian bias of the Sarvastivada.

167
As usual in the Sarvastivada (but not the Dharmaskandha), the satipatthana formula is presented
very simply - one establishes mindfulness on the contemplation of the body, feelings, mind, and
dhammas. We have established f r o m the Sarvastivada Smrtyupasthana Samyukta that this
abbreviated form was intended to be spelt out in full. In the Samyukta, the fact that there are
many discourses, one after another, all featuring the same formula is a good reason for the use of
such an abbreviation. But what of this Madhyama version? This discourse on satipatthana stands
alone, not part of a series, and one who was not familiar with the Samyukta, when studying the
Madhyama would not necessarily be aware of the longer formula. In this long discourse, why
could the redactors not have found room to expand the formula? There seems to be one simple
answer to this puzzle: the Smrtyupasthana Sutra was originally part of the Samyukta collection.
There the formula was abbreviated along with most of the rest of the discourses; in fact the
length of this discourse serves as a good reason to not spell out the formula in full, so that the
length would not differ too much from that of the surrounding discourses.

Here are the various body contemplation exercises in relation to the Theravada.

Table 14.1: Body Contemplation in the Sarvastivada and Theravada

Sarvastivada Smrtyupasthana Sutra Theravada Satipatthana Sutta

1. Four postures 2.
2. Clear comprehension 3.
3. Cutting off thought (see MN20.3/MA 101)
4. Suppressing thought (see MN 20.7/MA 101)
5. Anapanasati 1.
6 . 1 s t j h a n a simile (Similes at MN 119.18/MA 81 Kayagatasati, etc.)
7. 2 n d j h a n a simile
8. 3 r d j h a n a simile
9. 4 t h j h a n a simile
10. Perception of light (See SN 51.20)
11. Basis of reviewing (See AN 5.28)
12. Parts of the body 4.
13. Six elements 5. (Four elements only)
14-18. Charnel ground 6-14.

Let us first deal with the factors in common. These are mostly in the same order in both the
Theravada and Sarvastivada texts. The sole exception is mindfulness of breathing, which in t h e
Smrtyupasthana Sutra appears in its more usual position after clear comprehension, as also in t h e
Sariputrabhidharma. However, a swag of new practices is added, all somewhat uncomfortable in
the context of body contemplation. This raises the question as to why these incongruous passages
were inserted.

Part of the answer lies with the Kayagatasati Sutta, versions of which are found in both the
Theravada and t h e Sarvastivada Majjhimas. 384 The Theravada version of t h e Kayagatasati Sutta
centres on a list of 14 body contemplation exercises that is identical with the basic practices in
the body contemplation section of the Theravada Satipatthana Sutta. This specific set of practices
is not found elsewhere in the Theravada Nikayas, and therefore indicates a strong relationship
between these two texts. The only difference in the description of t h e practices is that the
Kayagatasati Sutta omits the simile for anapanasati: one observes the breath, long or short, like a
skilled t u r n e r would know when making a long or short turn. This simile is also absent f r o m both

168
the Sarvastivada Kayagatasmrti Sutra and the Smrtyupasthana Sutra; thus the only place it
occurs, so far as I know, is in t h e Theravada Satipatthana Sutta. However, the most significant
difference between the Theravada Satipatthana and Kayagatasati Suttas is that the Satipatthana
Sutta, with its repeated refrain on investigating rise and fall, strongly emphasizes vipassana,
while the refrain in the Theravada Kayagatasati Sutta says:

'As he abides diligent, ardent, and resolute, his memories and intentions dependent on the
household life are abandoned. With their abandoning his mind becomes settled internally,
quieted, unified, and brought to samadhi. That is how a monk develops mindfulness of the body.'

The Sarvastivada refrain says that diligent practice of this in a secluded place frees the mind f r o m
distress, brings samadhi, and brings knowledge according to reality. In both versions the body
contemplation exercises then lead straight to the f o u r j h a n a s , each of which is also said to be a
practice of mindfulness of the body.

After t h e j h a n a s there is a substantial list of benefits of the practice, falling into four divisions.
First, mindfulness of the body includes all dhammas that partake of realization, just as one who
encompasses with their heart t h e entire ocean includes all the rivers that flow into it. Second, a
list of three similes and their opposites (throwing a heavy/light ball, a sapless/sappy piece of
wood, a full/empty water jug), illustrating whether Mara gains an opportunity or not. Both of
these divisions are found in the Sarvastivada. Third, mindfulness of the body leads to the
witnessing of all clear knowledges, illustrated with a f u r t h e r three similes - a brimful pot, a
brimful pond, and a ready chariot. This division is absent f r o m the Sarvastivada. Fourthly,
mindfulness of the body, developed and cultivated, leads to ten benefits: overcoming discontent
and delight; fear and dread; heat and cold, etc; one develops the f o u r j h a n a s ; one realizes the six
kinds of clear knowledge. This list is also in t h e Sarvastivada, extended to 18 by the separate
addition of the f o u r j h a n a s and the four stages of enlightenment.

Evidently these four divisions are each of quite separate origin. There is substantial overlapping
of the topics in this discourse, especially t h e j h a n a s and clear knowledges. I think it is likely that
the f o u r j h a n a s were originally mentioned, not as a kind of mindfulness practice, but as a benefit
of mindfulness practice. This conclusion is implied by the basic refrain and by the list of benefits
in both versions; however, since both versions include t h e j h a n a s both as mindfulness and as a
benefit of mindfulness there is no direct textual support for this suggestion.

There is a strong imagistic consistency in many of the similes - the ocean, the full pot, the full
pond, the full waterjug. 3 8 5 This is consistent with t h e simile for mindfulness of the body found in
the Satipatthana Samyutta (the bowl brimful of oil), as also the 'saturated' imagery of t h e j h a n a
similes:

1st jhana: He makes the rapture and bliss born of seclusion drench, suffuse, fill, and pervade this
body, so that there is no part of his whole body unpervaded by the rapture and bliss born of
seclusion. Just as a skilled bath man, or bath man's apprentice heaps bath powder in a metal basin
and, sprinkling it gradually with water, kneads it until the moisture wets the ball of bath powder,
soaks it and pervades it inside and out, yet t h e ball itself does not ooze...

2nd jhana: He makes the rapture and bliss born of samadhi drench, suffuse, fill, and pervade this
body, so that there is no part of his whole body unpervaded by the rapture and bliss born of
samadhi. Just as though there were a lake whose waters welled up f r o m below, and it had no
inflow f r o m east, west, north, or south, and would not be replenished f r o m time to time with
showers of rain, then t h e cool fount of water welling up in the lake would make the cool water

169
drench, suffuse, fill, and pervade the lake, so that there would be no part of t h e whole lake
unpervaded by cool w a t e r .

3rd jhana: He makes the bliss f r e e of rapture drench, suffuse, fill, and pervade this body, so that
there is no part of his whole body unpervaded by the bliss free of rapture. Just as in a pond of blue
or white or red lotuses, some lotuses that are born and grow in the water thrive immersed in the
water without rising out of it, and cool water drenches, suffuses, fills, and pervades t h e m to their
tips and their roots, so that there is no part of those lotuses unpervaded by cool w a t e r .

4th Jhana: He sits pervading this body with pure bright heart, so that there is no part of his whole
body unpervaded by pure bright heart. Just as if there were a man covered from the head down
with a pure white cloth, so that there would be no part of his body not covered by the pure white
cloth.

These beautiful images lend an imminent physicality to the usual strictly psychological j h a n a
formulas. The strong emphasis on immersion in the body no doubt prompted their inclusion in
the Kayagatasati Suttas, and hence in the body contemplation section of the Sarvastivada
Smrtyupasthana Sutra. However it would be a mistake to think that this refers here to the crude
physical body, which completely disappears i n j h a n a . The Suttas use 'body' in an idiomatic sense
to stress the immediateness of direct personal experience; the meaning is something like 'the
entire field of awareness'. 386 Such abstract, almost mystical, usages o f ' b o d y ' - note too the 'mind-
made body', the 'body witness', the 'dhamma-body' - paved the way for the much later doctrine
of the 'Three-bodies' of the Buddha, a metaphysical and docetic reification of the Buddha and
Nibbana.

The Sarvastivada Kayagatasmrti Sutta, in addition to the variations mentioned above, presents a
different list of basic exercises. These correlate exactly with the section on body contemplation in
the Sarvastivada Smrtyupasthana Sutra, just as the Theravada Kayagatasati Sutta correlates with
the Theravada Satipatthana Sutta. The Theravadins added the vipassana refrain to t h e
Satipatthana Sutta, t h e three extra similes to the Kayagatasati Sutta, and reversed the sequence
of anapanasati and clear comprehension in both. The Sarvastivadins added the extra practices
and padded out the final section on benefits. This clearly shows that the specific details of these
lists are sectarian, although the differences are not explicitly doctrinal. Each of the schools must
have edited these paired discourses conjointly. This raises t h e important question: what was the
main direction of influence, from t h e Kayagatasati Sutta to the Satipatthana Sutta, or vice versa?
To answer this question we must first examine in more detail the extra practices in the
Sarvastivada.

Numbers 3 and 4, dealing with ways of controlling thoughts, are two of five methods advocated in
both the Theravada and Sarvastivada versions of the Vitakkasanthana Sutta ('Discourse on the
Quelling of Thoughts'), which is obviously a more appropriate home for them. The description in
the Sarvastivada Smrtyupasthana Sutra is, as expected, closer to that in the Sarvastivada version
of the Vitakkasanthana Sutta, and was no doubt taken f r o m there. Both versions say that these
practices are for one 'pursuing the higher mind', i.e. jhana, and by means of t h e m t h e mind
becomes steadied internally, quieted, unified, and brought to samadhi. In the first practice,
unskilful thoughts are replaced by thinking skilful thoughts. The Sarvastivada gives the simile of
a carpenter who marks a piece of wood straight and then chops it with a sharp axe; the Theravada
simile, however, speaks of a carpenter knocking out a coarse peg by using a fine one, which seems
more appropriate. In t h e second practice one crushes and suppresses the mind with the mind,
like two strong men would beat down a weaker. This violent approach to mental cultivation is
uncharacteristic of the Buddha's usual gentleness. This same practice is listed in the Maha
Saccaka Sutta among the useless byways t h e Bodhisatta mistakenly pursued before his

170
enlightenment. 387 Bronkhorst reasons f r o m this that the practice must have been a Jain one,
rejected by the Buddha, but later creeping back into t h e Suttas. However, the practice is the first
and least ascetic of the Jain practices, and the last and most ascetic of the Buddhist, so it seems
there isjust this slight overlap.

We have met number 10, the perception of light, together with some of the body contemplations,
in the Iddhipada-samyutta and the Anguttara Nikaya. Its inclusion in satipatthana may also have
been influenced by t h e standard passage on how one abandons the hindrance of sloth & torpor,
being 'percipient of light, mindful & clearly comprehending'. There is no obvious reason in the
passage itself why this should come under body contemplation (although the passage includes
the phrase 'as above, so below; as below, so above', which is explained in the Iddhipada-samyutta
as investigating the parts of the body f r o m the top to the bottom). The emphasis on the
perception of light in the Iddhipada-samyutta, supported by other passages, suggests that here
we are entering the exotic realm of psychic powers.

The 'basis of reviewing' 388 is mentioned in the Anguttara and the DTghas as the fifth factor of
noble right samadhi, the investigation of one who has emerged from j h a n a 'just as if one standing
should look at one lying down, or one lying down should look at one standing.' 389 Possibly the
mention of the postures here, though obviously only a metaphor, prompted the inclusion under
body contemplation. The situation with t h e j h a n a similes would seem to be similar, the body in
metaphorical or mystical sense merging with the literal physical body.

The internal sequence is obscure; why have the perception of light and the reviewing knowledge
been inserted at this point? It seems as if this structure has been influenced by the above-
mentioned five-fold right samadhi. There are two variations of this: the Anguttara mentions t h e
f o u r j h a n a s (with similes) and the 'basis of reviewing' as fifth. The DTgha mentions five factors:
'suffusion with rapture, suffusion with bliss, suffusion with heart, suffusion with light, basis of
reviewing.' The first three of these obviously refer to jhana, and are especially close to the
descriptions of the j h a n a similes. In particular the phrase 'suffusion with heart (ceto)' is quite
unusual and distinctive of this passage and t h e f o u r t h j h a n a simile. Then follow the perception of
light and the basis of reviewing, all these in the same order in both the fivefold right samadhi and
the Sarvastivada Smrtyupasthana Sutra. The correspondence of these somewhat unusual
passages is too close for coincidence. 390

The identification of these structural influences on t h e Sarvastivada Smrtyupasthana and

Kayagatasmrti Sutras allows us to make more sense of the overall structure of the body
contemplation section of these texts. They start with the contemplation of body postures and
movements through daily activities, the standard preparation for meditation. Another practice
that acts both as a preparation for meditation and in meditation itself is the control of thoughts;
this pertains to right intention of the eightfold path. Then follows the meditation proper,
anapanasati, leading to the four jhanas. Next is the perception of light, associated with t h e
development of psychic powers that regularly follow after jhana. Finally the reviewing
knowledge that investigates the whole process, and especially thejhana-consciousness, in terms
of cause and effect. So this whole section so far makes perfect sense as a description of the normal
course of practice prescribed in the Suttas. The problem lies, not in internal incongruities, but in
its placement as part of body contemplation. Only some of the sections are body contemplation
proper; the rest has been included to flesh out the sequence of practice, and in the case of the
jhanas the similes were used because they fit in with body contemplation. And of course the
incongruity is compounded by the addition of the body contemplations proper - parts of the
body, elements, charnel ground - at the end of the body contemplation section. These have
nothing to do with the progressive structure of the earlier part of the section. They were either

171
tacked on at the end, or more likely were an early part of the text, which remained as the
beginning was transformed into a comprehensive meditation program.

It seems that here we see a tendency towards inflating mindfulness of the body until it
encompasses t h e whole path. This perhaps stems f r o m the Buddha's statement that 'anyone who
has developed and cultivated mindfulness of the body has encompassed all skilful qualities there
are that have a share in realization.' Satipatthana provides an example of the same process on a
larger scale. In fact the Satipatthana Sutta covers much of t h e same ground in the later sections,
the inclusion of samadhi being an obvious example; also the investigation into causes in
contemplation of dhammas is similar to the 'basis of reviewing'. Thus the inclusion of these
additional practices in the Smrtyupasthana Sutra is largely redundant, suggesting that they were
originally intended for the Kayagatasmrti Sutra, and the movement into the Smrtyupasthana
Sutra was a secondary development. If this is true for t h e Sarvastivada, it may hold good for the
Theravada too, although there, with leaving behind of t h e j h a n a s , t h e influence is not as strong.

Having considered in some detail t h e contents of the body contemplation section of the
Sarvastivada Smrtyupasthana Sutra, we may now move on to t h e remainder. The refrain in the
Smrtyupasthana Sutra speaks simply of contemplating internally and externally, then of
establishing the mind upon the body (feelings, mind, dhammas) and obtaining knowledge, vision,
light, understanding -^T^^TiH; probable Pali equivalents are: nana, dassana, vijja,
panna). There is no mention of impermanence.

The Theravada Satipatthana Sutta here has: 'One establishes mindfulness only for a measure of
knowledge, a measure of mindfulness'. The word 'measure' (matta) here has always struck me as a
bit odd. One can try to explain it away, but t h e normal meaning is 'mere' or 'limited'. Is it possible
that there could have been an early confusion between matta 'measure' and patti 'attainment' or
attha 'purpose'? It is also perhaps a bit clumsy to say that one is supposed to establish mindfulness
(sati) in order to attain mindfulness (patissati). The Chinese term here means 'vision'. Could there
have been a f u r t h e r confusion between something like passati and patissati? (Of course, passati is a
verb and patissati is a noun, so this would have involved a syntactic and not merely a phonetic
change.) Combining the two suggestions we could arrive at: 'One establishes mindfulness only for
the sake of knowledge & vision.' This would certainly give us a more straightforward meaning,
but alas I do not possess the linguistic gigabytes necessary to untangle this tangle. The Ekayana
Sutra adds f u r t h e r to the confusion, as we shall see below.

The sections on feelings and mind in the Smrtyupasthana Sutra are similar to the Theravada
Satipatthana Sutta, but slightly more elaborate. From here to the end the text refers to 'bhikkhus
and bhikkhunTs', both as the audience of the discourse, and as the meditator in t h e discourse
itself. This is quite extraordinary, and I don't know if it is representative of the Sarvastivadin
Suttas in general. There must have been nuns and laywomen present at many of the teachings,
but in those innocent days before political correctness, the texts were almost without exception
standardized into the male voice. In the Theravada, even when a discourse is addressed solely to
nuns, the hypothetical practitioner of the discourse is usually a monk. Only rarely are the female
practitioners acknowledged. This is rather a shame, for it tends to put modern listeners offside.
The inscriptional evidence, according to Schopen, attests approximately even numbers of monks
and nuns in the early schools. Many of the nuns are said to have been learned in a sutta, or in a
Pitaka, and so on, so they played their part in t h e transmission of the Dhamma. These inscriptions
usually record substantial donations, of temples and suchlike. It would seem almost certain that
the monks were, on the average, wealthier than the nuns, and so the presence of equal numbers
of (presumably wealthy) monk and n u n donors suggests that the total population of nuns may
have been significantly greater. The Jains, unlike the Buddhists, made a census of their followers,
and the figures consistently recognize far more nuns than monks. Given the patriarchal climate

172
of the times, this is hardly a situation they would have invented, and perhaps the numbers of
Buddhist nuns, too, may have exceeded the monks in India, as they do in some modern Buddhist
countries.

The contemplation of dhammas compares as follows.

Table 14.2: Contemplation of Dhammas in Sarvastivada and Theravada

Sarvastivada Smrtyupasthana Sutra Theravada Satipatthana Sutta

1. Internal & external sense media 391
2. Five hindrances 1. Five hindrances
2. Five aggregates
3. Internal & external sense media
3. Seven enlightenment-factors 4. Seven enlightenment-factors
5. Four noble truths

All of these exercises are described in virtually identical terms in both Suttas, as indeed in the
Abhidhamma versions. The Sarvastivada omits the aggregates and truths, retaining the sense
media as the only overt vipassana practice. As we have seen, it shares this feature with t h e
Dharmaskandha. There is one, possibly related, text in the Bojjhanga-samyutta that mentions t h e
sense media. 392 There, the enlightenment-factors are said to be the 'one dhamma' for abandoning
the 'things that fetter', namely the six sense media. The Satipatthana Suttas, too, speak of the
sense media in terms of fetters. But in the Bojjhanga-samyutta the enlightenment-factors, which
chiefly pertain to samadhi, naturally precede the wisdom practice of the sense media. In t h e
Sarvastivada Smrtyupasthana Sutra, however, the sense media are displaced awkwardly to t h e
beginning of the section, whereas in the Dharmaskandha they are in the middle. For this reason
among others, I have no hesitation in concluding that this is a later interpolation, and that the
original Satipatthana Sutta includedjust t h e hindrances and enlightenment-factors.

The Sarvastivada Smrtyupasthana Sutra omits the repetition of t h e 'way to convergence'

statement at the end, but does include the guarantee of attainment in seven years, or as little as
seven days; indeed, it even goes f u r t h e r than t h e Theravada and says one may see results in the
evening if one practices in the morning. It also omits the final sentence of the Pali, 'This is what
was said [i.e. satipatthana is the 'way to convergence'] and this is the reason it was said [i.e. it
leads to these results].' Since this sentence is found in no other version, it may be regarded as an
attempt by the Theravadin teachers to tie together the loose assemblage of passages that the
Satipatthana Sutta had become.

The emphasis throughout t h e Sarvastivada Smrtyupasthana Sutra is clearly on samatha. Like the
Vibhanga, it omits virtually all t h e overtly vipassana oriented material of the Theravada
Satipatthana Sutta; and as well it adds much samatha material. It remains in harmony with the
mainstream teachings in treating vipassana solely as part of the contemplation of dhammas. Both
schools were moving towards using the Satipatthana Sutta as a compilation of meditation
techniques; but in the hands of the Sarvastivadins this became a samatha manual, while in the
hands of the Theravadins it became a vipassana manual. We should remember that the significant
differences are entirely in the choice and arrangement of material, not in the individual doctrinal
statements ('pericopes') themselves.

173
EKAYANA SUTRA ( E A 1 2 . 1 )

The Ekottara Agama, f r o m which this discourse was taken, is the least congruent of the four
Agamas. It was translated, it seems, f r o m Prakrit rather than Sanskrit. It is usually believed to
belong to the Mahasanghikas, the school that diverged f r o m the ancestral Theravadins at the first
schism. Thus, they are of particular interest, since textual and doctrinal material held in common
between the Mahasanghika lineage and the Theravada lineage probably dates to the earliest pre-
sectarian period. We have their Vinaya in Chinese, and some later texts, and recently a large
number of sutras or fragments in the original Prakrit have come to light, apparently f r o m a
monastery near Bamiyan in Afghanistan. Hopefully these will be edited and published soon,
which might help clarify the affiliation of the Ekottara. The Mahasanghika, like the Theravada,
proceeded to f u r t h e r splinter into many sub-sects, with varying degrees of doctrinal
development. It is sometimes suggested that t h e Ekottara Agama belongs to the Lokuttaravada,
which was one of the most progressive of all the early schools and in fact was probably one of the
immediate forbears of the Mahayana.

The famous matter of the 'five theses' that were the immediate cause for the schism has been
discussed by L. S. Cousins. He finds that the evidence suggests that differing attitudes towards
samadhi practice may underlie this schism, saying: 'There is some reason to believe that the
practice of jhanas is of great antiquity and the Mahasanghikas, or this branch of them, may well
have been conservative in this regard as well as others.' Whatever the t r u t h of this, we shall see
that there is a strong samadhi emphasis in t h e Ekayana Sutta.

However, the affiliation of the Ekottara is far from certain and several scholars have raised
serious objections to this identification. The main reason for ascribing it to the Mahasanghikas
seems to be the inclusion of several references to Maitreya, and it seems that the school of t h e
Ekottara may have had a special devotion for that Buddha of the future. This might be a trait of a
Mahasanghika school, but then again it might not. The doctrinal and other features that have
been identified do not bear any specially close resemblance to known Mahasanghika doctrines,
but then the collection is notoriously idiosyncratic, and presents many variations on even
standard doctrinal formulas like the eightfold path. Given the very bad inconsistency of the text
it might not be worth placing too much weight on doctrinal consistency as a guide to sectarian
affiliation. Even the orthodox and rigorous Theravadins retain in their Suttas a number of
passages that contradict known tenets of the school (the in-between state; the gradual path; t h e
necessity f o r j h a n a ; the primacy of Suttas over Abhidhamma; the primacy of practice over study,
etc.). So we must admit that we don't really know the school of t h e Ekottara, but in the absence of
any more plausible alternative we might consider the thesis that is belongs to t h e Mahasanghika
group of schools.

The careless editing and divergence from the other sources suggests a somewhat later date than
the Theravada Satipatthana Sutta and the Sarvastivada Smrtyupasthana Sutra. The setting, in
common with the Dharmaskandha, is t h e j e t a v a n a at SavatthT rather than Kammassadamma. This
could imply that it is actually a different discourse, delivered on a different occasion. However, it
should be obvious to anyone familiar with the Suttas that they are set in the Jetavana far more
frequently than anywhere else; so much so that it is hard to escape the impression that t h e
Jetavana was the 'default' setting for any Suttas whose provenance was unknown. This
impression is given substantial support by two passages mentioned by Schopen. One passage
f r o m the Mulasarvastivada Vinaya, has the Buddha telling Venerable Upali that, if the setting or
other details of a discourse or training rule is forgotten, one should declare that t h e setting was at
one of the six great cities, or somewhere the Buddha stayed many times; if one forgets the n a m e
of the King, one should say it was Pasenadi; if one forgets the n a m e of the householder, one
should say it was Anathapindika; and so on. 393 A similar statement is recorded in the

174
Mahasanghika Vinaya: in the rules concerning reciting t h e nine angas, if a monk forgets the
setting, he should say it was one of the eight famous places. 394 Schopen's discussion of this
principle is, I think, quite sound, and, as he says, application of such principles would quickly
result in the prevalence of settings at the Jetavana that is such a characteristic of the existing
canons. We have seen in the discussion of the Sarvastivada Samyukta that this kind of artificial
ascription of setting is very prevalent. This suggests that ascription of a discourse to the Jetavana
should not be taken too seriously. In cases like t h e Satipatthana Sutta, where different versions of
what look like the same text are set sometimes in the Jetavana, sometimes elsewhere, we should
assume that the less common setting is more likely to be authentic, and the shift to Savatthi
happened as part of a later normalization. Thus the different settings do not imply that the texts
had a different origin. On the evidence of t h e texts, this redaction principle would seem to have
been liberally applied by all the traditions. It is worth noting though, that the two explicit
mentions of the principle known so far (others way yet come to light) are f r o m the
Mulasarvastivada and the Mahasanghika; and these schools are very closely related to the schools
that produced the Dharmaskandha and the Ekayana Sutra, the two texts where the 'Jetavana
normalization redaction principle' seems to have been applied to the Satipatthana Sutta material.

The Chinese here translates 'satipatthana' by two characters that mean 'mind-tranquillity' ( M i t ) .
The Indian original of this translation would be something like manosamatha. This t e r m does not
appear as such in the Pali, although the virtually identical cetosamatha does. It is an unusual
rendering for satipatthana; normally the Chinese has the more literal 'mindfulness-place' ( ^ j t l ) .
However, the rendering is found in several other places, such as T 101 (the partial Samyukta SA3),
an independant version of the Dasottara Sutra (T 13), the Dharmaguptaka Vinaya, and several
versions of the Prajnaparamita literature. Thus it is not peculiar to the Ekottara.

After opening with the 'path leading to convergence' statement, the text says this path destroys
the five hindrances. The mention of the hindrances here is in line with the mainstream
understanding of satipatthana, and in particular recalls the Sarvastivada Smrtyupasthana Sutra.
'Path' is the eightfold path, here given in a typically eccentric form: right view, right prevention,
right conduct, right livelihood, right skill in means, right speech, right recollection, right
concentration. The formula for the four satipatthanas treats each as internal/external; perhaps
because it is included here, the internal/external formula is omitted throughout the body of t h e
discourse. The auxiliary formula is omitted. The significance of this formula, however, is that
satipatthana is developed in the context of the path as a whole, and in t h e Ekayana Sutra this has
already been stated. It may be the case, as we concluded with the Sarvastivada version, that this
omission implies the Ekayana Sutra has been moved f r o m a Samyutta; but the erraticness of the
text makes us reluctant to draw definitive conclusions.

In contrast with the rigid consistency of the Theravada Suttas, the refrains throughout vary
considerably. Thus here in the introduction, body contemplation is said to lead to the ending of
unwholesome thoughts and the removal of anxiety, while the remaining three contemplations
lead to peace a n d j o y . The refrain in the parts of the body section combines these two, saying to
'realize peace a n d j o y , end bad thoughts, and remove anxiety and sorrow'. Elsewhere the refrains
might include one or other part of this phrase, which makes us suspect that this was the standard
version, more-or-less garbled throughout the text. It is a similar kind of phrase to the standard
auxiliary formula, but it is not easy to see a specific textual relationship. The end of the section on
corpse meditation, after speaking of leading to peace, etc., speaks of understanding
impermanence. Thus we see a combination of samatha and vipassana aspects.

As we noted before, the text says it is the 'way to convergence' because it leads to oneness of
mind. This saying is here given greater prominence than in the other versions: firstly, because it
is explicitly explained in the text itself; second, because it is used as the title; and thirdly because
175
the text is included in the 'Ones' of the Ekottara Agama. I suggested in the GIST that the symbolic
significance of number may have been an influence in the formation of the Anguttara/Ekottara.
An important connotation of 'one' in Buddhism is samadhi, and this is clearly reflected in the
Theravada Anguttara. So here too we may suspect that the Ekayana Sutra was included in the
'Ones', classified under t h e first word of the text, a word that encapsulates the primary spiritual
purpose of the way of satipatthana: t h e achievement of unification of mind. Overall, this
introduction serves as a 'policy statement' emphasizing the very strong emphasis on the samatha
dimension of satipatthana in the Ekottara school.

Omitting anapanasati and awareness of activities, t h e section on body contemplations simply

includes the parts of the body, 395 four elements, and charnel ground contemplations. It adds
another similar practice - observing the openings of the body through which impurities flow.
This is similar to the Sariputrabhidharma. A simile is included for the elements (the butchered
cow) and the orifices (likejoints of bamboo or reeds), but is omitted for the parts of the body.
Since the simile is found in all the other Sutta versions, I assume this is a case of accidental loss.

The contemplations of feeling and mind are similar in their content to the other versions, but the
refrains are different to both the other versions and to the refrains of the first sections in the
same sutra. They refer firstly to contemplating the principles of origination and dissolution so
that, 'Regarding presently arisen feelings he has knowledge and vision, with awareness well-
founded, with nothing to rely on, oneself h a s j o y and happiness and does not give rise to feelings
of attachment to the world.' The refrain here is difficult, and the translations I have vary widely.
The above is a tentative attempt to translate under the assumption that this phrase is similar to
the Pali here; at least the Chinese does not seem to preclude this possibility. One of t h e few
certainties is that the same words for 'knowledge & vision' are u s e d j u s t as in the Sarvastivada
refrain, supporting the thesis that this was original. Then there follow some slightly varying
versions of formulas for attaining arahantship commonly found throughout the Suttas, which are
obviously later additions. The refrain section generally has some affinity with the Theravada
Satipatthana Sutta, which also implies, though less explicitly, that each practice can lead to
arahantship.

The contemplation of mind includes a f u r t h e r phrase that seems to have confused,

understandably so, the translators: W S f S - ^ N W S f t L This c o m e s j u s t after the phrase we have
identified as referring to 'knowledge and vision', and seems as if it may follow on f r o m that. Thich
Nhat Hahn has it that this refers to knowing what is not knowable, and suggests that that this is a
Prajnaparamita-style paradox. Pasadika renders it 'from this it can be gathered whether one is
really mindful or not'. But the phrase ^ W S f S , which is peculiar to the Ekottara, elsewhere
stands for atakkavacara (beyond the sphere of reason) or acinteyya (unthinkable). This does not
really tell us what t h e phrase means ('one considers that which is beyond thought'?), but it does
suggest a connection with the phrase found occasionally in satipatthana, 'one does not (or should
not) think thoughts associated with the body'.

Moving on to t h e dhammas section of the Ekayana Sutra, we f i n d j u s t the seven enlightenment-

factors and the f o u r j h a n a s . A measure of the difficulty of the text is that Pasadika translates this
section without even realizing that it refers to the enlightenment-factors. Bizarrely, the
translation has n t , the normal rendering of mindfulness, here for both the enlightenment factor
of mindfulness and of rapture. The enlightenment-factors are presented very simply, omitting
the inquiry into causes that is characteristic of the dhammas section in the other versions. I t j u s t
says one develops each of the enlightenment-factors 'in reliance on initial application, on no-
craving, on destroying the unwholesome mind, and abandoning the unwholesome dhammas'.
This seems like a slightly garbled version of the common formula: 'dependent on seclusion, on
fading of lust, on cessation, and ripening in relinquishment'. The Chinese appears to have read
176
vitakka (initial application) for viveka (seclusion). If we accept this reconstruction, then t h e
description of the enlightenment-factors here becomes very much like the four modes of
contemplation of dhammas in anapanasati: impermanence, fading of lust, cessation,
relinquishment. The ending of the discourse with the four jhanas is similar to the Maha
Satipatthana Sutta of the DTgha Nikaya. It reaffirms yet again the function of satipatthana to lead
up t o j h a n a .

It is interesting that t h e inquiry into causes, prominent in the other expositions of t h e

contemplation of dhammas, is absent here. It seems that, while the Theravada changed the
samatha aspects of satipatthana into vipassana, the school of the Ekayana Sutra changed the
vipassana aspects into samatha. If we are right in taking this version to be later than the
Theravada Satipatthana Sutta and the canonical abhidhamma, perhaps it was deliberately
formulated in opposition to the vipassana interpretation of satipatthana favoured by the
abhidhammikas. The samadhi implication of the famous introductory phrase was drawn out and
made explicit. Then t h e hindrances were rescued f r o m obscurity in the contemplation of
dhammas and placed at the beginning and the end, always the positions of greatest emphasis.
This proclaims that abandoning the hindrances is n o t j u s t one exercise among many, but is the
main orientation of the whole practice. Then the f o u r j h a n a s were brought in to culminate the
contemplation of dhammas, rendering explicit what the compilers may have felt was too
ambiguous and subject to misinterpretation in the original version.

The Ekayana Sutra ends by repeating that satipatthana is the 'way to convergence', and that it
gets rid of the five hindrances. It omits the guarantee of attainment found in the other versions.

In conclusion it seems that the Mahasanghikas, or one of their sub-schools, preserved a simpler
version of the Satipatthana Sutta for some time, resisting the trend to use it as a catch-all
repository of meditation techniques. Of course, they may well have had other texts fulfilling this
function. The expansions, none of which are overtly sectarian, tended to be in the refrains rather
than the content of the meditation exercises themselves. In accordance with all t h e early
teachings on satipatthana we have examined so far, the Ekayana Sutra strongly emphasizes the
samatha aspect, while also giving due consideration to vipassana.

THERAVADA SATIPATTHANA S U T T A ( M N 1 0 , D N 2 2 )

How might this text have been formed? Apart f r o m the Satipatthana Suttas, there are no texts in
the Majjhima, or in the DTgha for that matter, that deal with satipatthana in detail. Desiring a full-
length text on satipatthana, the Majjhima redactors selected t h e *vibhanga Mula from the
vyakaranas in the Satipatthana-samyutta as the most promising. However, it was too short and
needed filling out if it were to take its place in t h e Majjhima. This filling out involved expanding
the list of meditation exercises in the contemplation of the body and of dhammas, and developing
an extended refrain by adding the contemplation of rise & fall, taken from the Samudaya Sutta, to
the internal/external refrain.

We have seen that the evidence suggests that these 14 exercises appeared first in the Theravada
Kayagatasati Sutta and were later used to flesh out the Theravada Satipatthana Sutta. What
change, if any, would the inclusion of these 14 exercises make?

We have seen that satipatthana is normally seen as being very close or identical to anapanasati,
so its inclusion here is unproblematic. The only question marks, as we have already commented,
are whether it is valid to sever the first tetrad f r o m the rest of the practice; and also whether the
simile is authentic, since it is only found in t h e Theravada Satipatthana Sutta. Perhaps it was

177
inserted under the influence of the strong prevalence of similes and imagery in both the
Kayagatasati Suttas.

As regards the elements and death contemplations, we have noted that these occur grouped
together elsewhere with the 31 parts, so again their inclusion is unproblematic.

The main change in the orientation of the practice is the inclusion of the sections on clear
comprehension and body postures. As we have noted above, elsewhere these are kept clearly
distinct from satipatthana as meditation. In the Samyutta they are separated, in the Sarvastivada
Smrtyupasthana Sutra, the Sariputrabhidharma, and the Prajnaparamita they come before
anapanasati, and in the Vibhanga, the Dharmaskandha, and t h e Ekayana Sutra they are not found.
Consistently, clear comprehension during daily activities is not treated as a meditation as such,
but as a precursor to meditation. Only in the Theravada Kayagatasati Sutta and the Theravada
Satipatthana Sutta is this practice placed after anapanasati. In the context of the Suttas as a
whole, this variation may have been intended to exalt anapanasati by granting it pride of place.
But to one not familiar with t h e overall context of the teaching this could be taken to imply that
clear comprehension of one's everyday activities is itself an alternative meditation, as powerful as
anapanasati. This then leads to the claim that the most important of the Buddha's meditation
instructions was to be mindful whatever we do. But clear comprehension as part of the gradual
training is obviously a prescription, part of the all-round mental training that was expected of
those who had undertaken the monastic vocation. Thus from both the agreement of the texts,
and f r o m the broader consideration of how it fits in t h e Buddha's path, we can conclude that the
sections on clear comprehension and body postures were not likely to have been in the
*Satipatthana Mula. If they were, however, it is probable they were placed in the more normal
position at the beginning. There they naturally lead on, as in the gradual training, to the monk
who has 'gone to t h e forest, to the root of the tree, or to an empty hut' to settle into meditation.

The refrain in the *Satipatthana Mula would probably have originally consisted of t h e
internal/external contemplation and a simple exhortation to be mindful for the sake of
understanding, independence, and letting go. But since nowhere else in the Majjhima is any
connection between satipatthana and vipassana mentioned, it seems that the Majjhima redactors
wished to incorporate the section on impermanence f r o m the Samudaya Sutta of the Samyutta.
This became attached to the internal/external refrain at t h e end of each exercise, thus furthering
the idea, already hinted at in the Theravada Anuruddha-samyutta, that vipassana may be
undertaken f r o m the start of practice. While this seems to reflect the meditation orientation of
the redaction school, it is possible that it was originally an unintended side-effect of the purely
formal evolution of the Satipatthana Sutta.

It is worth noticing in passing that the Theravada Satipatthana Suttas have a small spelling
mistake in the refrain in the contemplation of dhammas section. The phrase is ' "Atthi dhamma"ti'
('there is dhammas'), where atthi is singular and dhamma is plural. Presumably this was merely a
reciters' glitch, as they mechanically repeated the phrase f r o m the earlier sections, without
noticing the change in number f r o m singular to plural. Innocent enough, but a reminder of the
fallibility of the tradition.

Another consequence of the extended refrain is that each section, ending with the phrase 'one
abides independent, n o t grasping at anything in the world', seems to lead all the way to
arahantship. Thus the text has both a 'horizontal' dimension, a progressive deepening from one
section to the next (as suggested by the phrase 'again & beyond' which prefixes each section), and
a 'vertical' dimension, developing to liberation within each one of the exercises. This kind of
'reflective immanence' is characteristic of the Suttas, and is no problem as long as it is understood
holistically rather than divisively. That is, each meditation subject is complete, not because it

178
replaces other approaches but because it includes them. As we have seen, this is brought out most
clearly in the context of anapanasati. But this subtle point is easily overlooked, and historically it
has contributed to the gradual neglect and marginalization of the progressive structure of
satipatthana.

The sections on feelings and mind are similar to all the other versions. The section on dhammas is
substantially lengthened; the main interpolations in the Majjhima version being the
contemplations of t h e aggregates and the sense media. Although these occur frequently in the
early texts, nowhere else do they appear as part of satipatthana. Remember that the Samudaya
Sutta of the Samyutta says that the origin of dhammas is 'attention'. It is impossible to
understand, and clearly against the normal position of the Suttas, how attention could give rise to
the aggregates and sense media. The commentary doesn't even try, simply agreeing that paying
uncausewise attention gives rise to the hindrances and paying causewise attention gives rise to
the enlightenment factors. This is absolutely in accord with one of the outstanding themes of t h e
Bojjhanga-samyutta:

'Monks, when one pays uncausewise attention, unarisen sensual desire arises, and arisen sensual
desire increases and expands; unarisen ill-will.. sloth & t o r p o r . restlessness & r e m o r s e . doubt
arises, and arisen doubt increases and expands.

'Monks, when one pays causewise attention, the unarisen enlightenment factor of mindfulness
arises, and the arisen factor of mindfulness goes to fulfilment by development; the unarisen
enlightenment factor of investigation of d h a m m a s . e n e r g y . r a p t u r e . t r a n q u i l l i t y . samadhi...
equanimity arises, and t h e arisen enlightenment factor of equanimity goes to fulfilment by
development. '396

This is obviously the same practice as in the Satipatthana Sutta. Similar themes recur in more
than a dozen texts in the Bojjhanga-samyutta. 397 In the Sarvastivada many of these texts are
grouped right at the start of t h e chapter, emphasizing this t h e m e even more. Even this alone
would strongly imply that the aggregates and sense media were a later interpolation. These
paradigmatic vipassana exercises are suggestively placed after the abandoning of t h e hindrances
but before the development of the enlightenment-factors. The normal position is that the
enlightenment-factors, which are very close in meaning with samadhi, bring about the
abandoning of the hindrances; but here it seems we can abandon the hindrances without
developing samadhi and then do vipassana straight away - an idea that was to prove highly
influential in Theravada meditation.

We can, however, discern some differences that suggest that the six sense media are more at
home here than the five aggregates. Firstly, as we have seen, mindfulness is more
characteristically mentioned with the sense media, and the sense media do occur one time in the
Bojjhanga-samyutta. Secondly, t h e sense media are included in the Sarvastivada versions and t h e
Sariputrabhidharma. Thirdly, the phrasing of the contemplation is more congruent with the
sections on the hindrances and enlightenment-factors. I am still confident that both were later
interpolations, b u t I am marginally less confident in the case of the sense media.

There is another evident incongruity, implicit in some other places, but strongly evident here. In
each of the other sections, as we have seen, there is no explicit mention of impermanence,
causality, etc. Then impermanence is brought in with the vipassana refrain that follows. This,
while probably not original, is perfectly coherent. But in the contemplation of dhammas
according to the Theravada, each meditation exercise speaks of impermanence. In the
contemplation of the hindrances, enlightenment factors, and sense media, this is indicated by key
words such as 'arising' (uppada), which recurs through these sections. The contemplation of

179
aggregates similarly speaks of 'origination' (samudaya). But then, having already discussed
impermanence once, it is re-introduced in the vipassana refrain, which also speaks of
'origination' (samudaya). Is one then to contemplate the 'arising of arising?' This unnecessary
repetition might be explained away; but the most straightforward explanation is that it simply
results f r o m the editing together of two originally distinct texts.

The dhammas section in the Majjhima version closes with a brief enunciation of the four noble
truths. This is then expanded greatly in the Maha Satipatthana Sutta of t h e DTgha Nikaya.398 Some
of the recent Burmese recensions have re-incorporated this entire section from the DTgha Nikaya
back into t h e Majjhima Nikaya, and even acknowledge this provenance by re-titling it the 'Maha
Satipatthana Sutta'. Perhaps a better title would be the 'Piltdown Sutta'. This canonical
innovation is extraordinary. While it is common for a word or phrase to slip between the cracks, I
do not know any other place where a large body of text has been moved, obviously in fairly
recent times. No doubt this editorial outrage was perpetrated with the idea of f u r t h e r
exaggerating the already overblown status of t h e Satipatthana Sutta. But the result is rather the
reverse - such clumsy mishandling leaves all-too-obvious fingerprints at the scene of the crime.
The altered version is found in the so-called 'Sixth Council' edition published by the Vipassana
Research Institute, but was inserted earlier, for the notes to the PTS Pali (edited in 1888) state
that the Burmese manuscript includes under the four noble truths 'a passage of some length,
borrowed from the Maha Satipatthana Sutta of the DTgha Nikaya'. 399 This possibly refers to the
Fifth Council edition. There is a Pali work called Samgayanapucchavissajjana (included in the VRI
CD), which gives the questions and answers on the texts as spoken in the Council, although it
doesn't say which Council - presumably it is t h e Fifth or Sixth. This also includes the 'Maha
Satipatthana Sutta' in the Majjhima, and has t h e temerity to assert that because of its great
usefulness to meditators this text was recited 'twice' 'in detail' by the redactors in ancient times.
While most other discourses rate a bare mention in this work, t h e Maha (sic!) Satipatthana Sutta
is distinguished by detailing the contents of the meditation exercises. A similarly obvious bias is
apparent in the Vipassana Research Institute's online version of the Tipitaka. Under the contents
table of the Majjhima Nikaya, almost all discourses are simply listed by their title; but the Maha
(sic!) Satipatthana Sutta is again singled out by individually listing all twenty-one sections. The
meditative bias of those who have put together these recensions of the Tipitaka are well known:
the questioner at the Sixth Council was Mahasi Sayadaw, the founder of the main 'dry vipassana'
meditation system; and the VRI Tipitaka was put made by the followers of the Goenka tradition,
the other main vipassana school. Although Buddhists are generally a charmingly placid lot, I
think we should not take such crudely partisan manipulation of our sacred scriptures lying down.
Protest! I did - I sent an e-mail to t h e VRI alerting t h e m to this alteration and requesting that
they restore the Majjhima version of t h e Satipatthana Sutta. But there was no reply. Does no-one
care? The only good thing to emerge from this blatant sectarian revision of the Tipitaka is that
no-one can reasonably insist that the Tipitaka must have remained unchanged for all time.

The Maha Satipatthana Sutta is the only significant discourse in the DTgha Nikaya that is not
found in the Dharmaguptaka DTrgha Agama. This can be no mere oversight, for it is also absent
f r o m the Sarvastivada DTrgha. I would therefore consider the Maha Satipatthana Sutta as a
leading contender for the title of the latest discourse in the four Nikayas, a lost waif straying over
f r o m the early abhidhamma. It is worth noting that this is the only discourse in all the existing
collections to be duplicated in both the Majjhima and the DTgha, f u r t h e r evidence of its
anomalous character. It is obviously j u s t the Satipatthana Sutta padded out with f u r t h e r material,
and again, the increase is not small. The Satipatthana Sutta treats the four noble truths with a
bare enunciation. In the Suttas this kind of formulation often indicates, not vipassana, but the
realization of stream entry; thus it could have been originally intended to express the results of
the practice of the previous sections. But the Maha Satipatthana Sutta gathers much material
f r o m elsewhere in the Suttas, ending up with the longest of all expositions of the truths, virtually

180
doubling the length of the Satipatthana Sutta, and clearly presenting the four noble truths
section as an extended course in vipassana.

The new material is mainly identical with the Saccavibhanga Sutta.800 The Maha Satipatthana
Sutta adds a lengthy analysis of the second and third noble truths to the Saccavibhanga Sutta
material. This is structured around the following series of dhammas, spelled out for each of the
sense media: external sense media, internal sense media, cognition, contact, feeling, perception,
volition, craving, initial application, sustained application. The Samyutta Nikaya includes a
similar list, although it has the elements and the aggregates for the final two members of the list,
rather than initial & sustained application. Several of the Samyuttas containing this series are
missing f r o m the Sarvastivada Samyukta. 801 Nevertheless, a similar list, again omitting the final
two members, is found in the Sarvastivada Satyavibhanga Sutra. The only place I know of where
the Maha Satipatthana list occurs verbatim in the four Nikayas is in the 'repetition series'
appended to the Anguttara sevens. 802 Such sections are usually to be regarded as late, and in the
present case the whole passage is ignored by t h e commentary. These considerations suggest that
the list, even in its simple form, is probably late.

This list is an expanded form of the psychological analysis of t h e cognitive process first
enunciated in the third discourse, t h e Adittapariyaya Sutta, and repeated countless times
subsequently. Eventually, this sequence would evolve into the cittavithi, t h e final, definitive
exposition of psychological processes worked out in great detail by t h e later abhidhammikas.
Thus the Maha Satipatthana Sutta stands as an important bridge to the Abhidhamma. We have
already discussed the fact that almost all this four noble truths material is found in the
Abhidhamma Vibhanga exposition of t h e truths.

Needless to say, the vast bulk of the new material in the Maha Satipatthana Sutta is vipassana
oriented, continuing the trend we have consistently observed in the development of t h e
satipatthana texts within the Pali canon. Nevertheless, the exposition of the truths, and therefore
the Sutta as a whole, ends with the f o u r j h a n a s as right samadhi of the path, re-asserting the basic
function of satipatthana to lead up t o j h a n a .

The significance of the Maha Satipatthana Sutta can best be understood in light of the structure
of the DTgha Nikaya as a whole. The most authentic and often repeated teaching in the DTgha sets
out the very heart of Dhamma practice. In the discussion of the GIST we saw that, leaving aside
the Brahmajala Sutta, the DTgha Nikaya starts off with a series of twelve discourses expounding
the gradual training in detail, including the f o u r j h a n a s . This would be pounded into the heads of
the DTgha students over and again as the way of training. In fact the GIST would suggest that this
section was t h e original core around which the DTgha was formed. Thus the whole of the DTgha
may well have started out as a jhana-manual. There is little vipassana material in the DTgha. A
striking example of this is the rarity of the five aggregates. Leaving aside the Maha Satipatthana
Sutta, meditation on t h e aggregates is mentioned only in the legendary context of the
Mahapadana Sutta. Elsewhere the aggregates receive but a bare enunciation in the proto-
abhidhamma compilations such as the SangTti and Dasuttara Suttas.

It seems likely that t h e compilers of the Theravada DTgha Nikaya, at some point after the
Dharmaguptaka schism, wished to include some more vipassana material to counterbalance the
strong samadhi emphasis. Now, there are three texts treating mindfulness practice in detail in
the Majjhima: the Satipatthana Sutta, the Anapanasati Sutta, and the Kayagatasati Sutta. The
latter two clearly emphasize samadhi, so in choosing which of the three to 'promote' to the DTgha
the compilers chose the most vipassana oriented text and padded it out with f u r t h e r vipassana
material to redress the imbalance of the DTgha Nikaya as a whole. And in context, this was most
reasonable. But when the discourse is divorced from its context and treated as a blueprint for a

181
meditation technique different from, even superior to, the mainstream samadhi practice, a shift
of emphasis becomes a radical distortion of meaning.

We may be able to pin down more precisely the date of the formation of the Maha Satipatthana
Sutta. We have already noted that it is absent f r o m both the Sarvastivada and Dharmaguptaka
DTrghas. Now, t h e Sarvastivada schism was pre-Asoka, and the Dharmaguptaka schism was,
according to several sources, somewhat later. 403 So the Dharmaguptaka schism must have been in
the time of Asoka or perhaps some time after that. The Sri Lankan mission arrived in the Asokan
period, and the Theravada had a strong presence on t h e island f r o m that time. It seems that the
headquarters of the Theravada must have been in Sri Lanka f r o m quite early, for an inscription at
Nagarjunikonda in Southern India refers to a monastery belonging to the 'Theravadin teachers of
Sri Lanka'.404 Given their evident doctrinal and textual closeness, it is plausible to think of the
Theravada and t h e Dharmaguptaka as being the Northern and Southern, or Indian and Sinhalese,
branches of the Vibhajjavada, the differences between the schools being as much a matter of
geography as doctrine. (The sources indicate that the Dharmaguptaka in fact split, not directly
f r o m the Theravada, but from the MahTsasaka, another very similar school.)

This raises the possibility that t h e final editing of the Pali Nikayas was carried out on Sri Lankan
soil. This case was put by Oliver Abeynayake in his article 'Sri Lanka's Contribution to the
Development of t h e Pali Canon.' 409 To summarize a few of his points, much of the Vinaya Parivara
was composed in Sri Lanka. In addition, the restructuring of the Vinaya Pitaka, f r o m the early
form of the Bhikkhu Vibhanga and BhikkhunT Vibhanga, attested in all schools including the
Theravada Vinaya Culavagga itself, to the current division along the lines of the 'Parajika Pali'
and 'Pacittiya Pali' is unique to Sri Lanka, and may plausibly be regarded as a Sinhalese
development. Several sections of the Khuddaka Nikaya, including the Khuddakapatha, may
plausibly be attributed to Sri Lanka. Within the four major Nikayas, Yakkaduwe Sri Pragnarama,
the late principal of the Vidyalankara Pirivena in Sri Lanka, has identified, in the Theravada
Majjhima, eight sentences of the Mulapariyaya Sutta and four verses of the Sammaditthi Sutta
that are in Sinhalese Prakrit, not Pali. In fact, the Theravada commentaries themselves assert that
some of the material in the DTgha was added by the Sinhalese elders, namely the closing verses of
the Maha Parinibbana Sutta, starting with 'There were eight measures of the remains of the one
endowed with v i s i o n . ' . This statement seems most plausible, since the verses are in a late metre,
and are tacked on in fairly redundant manner; also they include, not merely worship of relics, but
specifically the teeth relics, which is one of the most distinctive features of Sinhalese Buddhism.
Moreover, the line preceding t h e m is a catch-phrase in Pali (evam'etam bhutapubbam, 'that is how
it was') that typically refers to far-off events in the legendary past, like the English 'Once upon a
t i m e . ' The commentary even admits that this phrase was inserted in the Third Council, at the
time of Asoka. However, despite all this very strong evidence, some of the verses are, surprisingly
enough, included in the Sanskrit version. This contains the verse 'There were eight measures of
the remains of t h e one endowed with v i s i o n . ' and that on the worship of the teeth relics. It is
most unlikely that a Sinhalese composition found its way into a Sanskrit text in t h e north of
India, so perhaps these verses were added in India after all. But the later verses, starting with 'By
their power this fruitful e a r t h . ' , are absent f r o m the Sanskrit, and may well have been added in
Sri Lanka.

This last point may indirectly bear on the date of the Maha Satipatthana Sutta. The closing verses
of the Maha Parinibbana Sutta are predominately of late metres such as vamsattha. One of the few
other places in the canon that contains vamsattha and other similarly late, elaborate verse styles
is the Lakkhana Sutta. 407 This hagiographical text is found in the Sarvastivadin Majjhima and in
the Theravada, but not the Dharmaguptaka or Sarvastivada, DTgha. It therefore must have been
transferred f r o m the Majjhima to the DTgha after the Dharmaguptaka schism, at around the same
time as t h e Maha Satipatthana Sutta was created. This shift seems to have been prompted by t h e

182
large-scale expansion of the text. The Sarvastivada Madhyama version merely speaks of the two
careers open to a Great Man, and lists the 32 marks. The Theravada Dlgha version adds detailed
prose explanations and verse elaborations of the workings of kamma and its fruits regarding the
32 marks. 407 The commentary says the verses were added by Venerable Ananda. Although this
cannot be accepted as literally true, it does imply that the commentators were aware that the
verses were added later and by a different hand. They could plausibly be ascribed to monks
following Ananda's devotional tradition. As we noted, these verses are similar in style to the
closing verses of the Maha Parinibbana Sutta, which the commentary says were added in Sri
Lanka. Given this, as well as the verses' evident lateness and omission from the Sarvastivada, it
seems reasonable to suggest that they were also added in Sri Lanka. The verses were probably
added to the Lakkhana Sutta around the same time as the extra four noble truths material was
added to the Satipatthana Sutta, and so we may thus suggest that the resulting Maha Satipatthana
Sutta was composed in Sri Lanka.

We may then ask when these additions may have occurred. Again, there is no direct evidence, but
we can seek a convenient peg on which to hang them - this will at least give us something to
argue about. Reviewing the timeline of early Sinhalese history, we note that, after the
introduction of the Buddhist texts in the time of Asoka, the first literary activity of major
importance mentioned is during the reign of Vattagaminl. At that time, due to internal strife (war
with the Tamils - some things never change!), the lineage of oral transmission of the Tipitaka was
nearly broken. The Sangha made the momentous decision to write down the Tipitaka, asserting
that study and preservation of the texts was more important than practice of their contents (a
decision that, sadly, has set the agenda for the Theravada until the present day). According to
recent scholarly opinion this occurred around 20 B.C.E. I suggest that this was when the Maha
Satipatthana Sutta was created.

CHAPTER 1 5 : THE SOURCE

There is an unfortunate side-effect of the kind of textual analysis I have undertaken in this book.
It is not all that difficult to deconstruct such ancient, heavily edited texts as the Buddhist
scriptures. There are abundant fault-lines, anomalies, and obscurities if one wishes to look. But
what are we to do - demolish the palace and leavejust a pile of rubble? This too is not authentic
to the texts, for the undeniable fact is that, despite everything, the Nikayas/Agamas offer us a
vast body of teachings springing from a remarkably uniform vision, a clarity and harmony of
perspective that is unparalleled in any comparably large and ancient body of writings. To convey
the impression that the situation is hopelessly confused and problematic is to deny the
extraordinary fact that we have the texts at all. Somehow, the group of men and women that
made up the ancient Buddhist Sangha managed to organize themselves to produce and maintain
a magnificent spiritual literature for 2500 years. While denial that there are problems is naive and
untenable, throwing our hands up in the air in despair shows an excess of what the Satipatthana
Sutta calls 'spiritual depression' (niramisa domanassa). I think the lines of unity and consistency in
satipatthana are far more significant and powerful than the fractures. But in this book so far, the
threads of connection and continuity are buried in the pages of analysis: the question is, how to
make this unity vivid?

My solution is to present a full reconstruction of what I consider to be the authentic source

material for the Satipatthana Sutta. While, obviously, any such attempt is deeply problematic, I
do not think it is less problematic than simply walking away and leaving t h e j o b undone. None of
the scriptural traditions offer us a pile of meaningless, deconstructed rubble. Reading a
reconstructed text will at least allow those interested to gain a more immediate impression of

183
what the satipatthana teachings may have been like, if my methods and conclusions have any
validity.

Before proceeding to t h e *Satipatthana Mula, let us review the ground we have covered so far.
The following table displays a summary of the contents of satipatthana as represented in the
above materials. It graphically shows the strands of continuity and discontinuity between t h e
recensions. The table does not differentiate between the versions in t h e Theravada DTgha and
Majjhima. Sections that I consider authentic are shaded. The identification of the Indian originals
for some of the Chinese terms in the contemplation of mind is uncertain.

Table 15.1: The Satipatthana Material

184
 Vibhanga Dharmaskandha Sariputrabhidharma Theravada Satipatthana Sarvastivada Smrtyupasthana Ekayana Sutra Prajnaparamita Sutra
Sutta Sutra
Four postures Four postures Four Postures
Clear comprehension Clear comprehension Clear comprehension
B Cutting off thought
Suppressing thought
Anapanasati Anapanasati Anapanasati
Four postures Anapanasati
0 Clear comprehension
4 jhana similes
Perception of light
Basis of reviewing
D Parts of the body Parts of the body Parts of the body Parts of the body Parts of the body Parts of the body Parts of the body
4 elements
6 elements 4 elements 4 elements 6 elements 4 elements
Food
Space (5th element)
Y Oozing orifices Oozing orifices
Charnel ground Charnel ground Charnel ground Charnel ground Charnel ground
Pleasant/painful/neutral Pleasant/painful/neutral Pleasant/painful/neutral Pleasant/painful/neutral Pleasant/painful/neutral Pleasant/painful/neutral
FEEL- Bodily/Mental Bodily/Mental
Carnal/spiritual Carnal/spiritual Carnal/spiritual Carnal/spiritual Carnal/spiritual Carnal/spiritual
INGS Sensual/Non-sensual Sensual/Non-sensual
No mixed feelings
With/without lust With/without lust With/without lust With/without lust With/without lust With/without lust
With/without anger With/without anger With/without anger With/without anger With/without anger With/without anger
M With/without delusion With/without delusion With/without delusion With/without delusion With/without delusion With/without delusion
Defiled/undefiled With/without affection
With/without attainment
With/without confusion
I Contracted/scattered C ontracte d / scattered Contracted/scattered C ontrac te d / scattere d Contracted/scattered Contracted/scattered
Slothful/energetic
Universal/not universal
Exal ted/unexalte d Small/great Exalted/unexalted Exal te d/unexal ted Small/great Exalted/unexalte d
N Surpassed/unsurpassed Surpassed/unsurpassed Surpassed/unsurpassed Lower/higher Surpassed/unsurpassed
Di stracted/undi stracted Developed/undeveloped
Quiet/unquiet
Samadhi/no samadhi
D Samadhi/no samadhi Developed/undeveloped Samadhi/no samadhi Samadhi/no samadhi Samadhi/no samadhi Samadhi/no samadhi
Released/unreleased Released/unreleased Released/unreleased Released/unreleased Released/unreleased Released/unreleased
D 6 sense media
H 5 hindrances 5 hindrances 5 hindrances 5 hindrances 5 hindrances (5 hindrances in
A 5 aggregates intro & conc.)
M 6 sense media 6 sense media 6 sense media
M 7 enlightenment-factors 7 enlightenment-factors 7 enlightenment-factors 7 enlightenment-factors 7 enlightenment-factors 7 enlightenment-factors
A 4jhanas
S 4 noble truths 4 noble truths
R I nternal/external Internal /external I nte rnal/external Internal/external I nte rnal/external (Internal/external in I nte rnal/external
E Shortcomings Shortcomings introduction only.
F Body: refrains vary.
R Feelings, mind, dhammas:)
A Rise/fall Rise/fall Rise/fall
I Knowledge Knowledge Knowledge Knowledge
N Independence Independence Independence Independence
Arahantship

185
The basic principle in editing the * Satipatthana Mula is simple. We have seven early texts that
teach satipatthana in detail. These all have much in common which each other. They are also
substantially different f r o m any other teachings on satipatthana, being the only places that spell
out the details of the four satipatthanas. One possible explanation is that they are descended f r o m
a common source. The most likely content of this source is the shared material found in each of
the texts. However, we also have to bear in mind that different types of text use different editing
principles. For example, the Theravada Abhidhamma typically does not include similes, so the
absence of such material in the Vibhanga does not suggest that it was absent f r o m the
*Satipatthana Mula. Similarly, the Ekayana Sutra is somewhat later, erratically edited, and
divergent compared with the other sources. Generally, then, the *Satipatthana Mula will include
only material found in all texts, but will occasionally allow phrases found only in four or five.

By taking j u s t the common material, we end up with a text that, in effect, is much like the
Theravada Abhidhamma Vibhanga, minus the special Abhidhamma material. I hope I have shown
in the discussion of this text that it is refreshingly free of the kinds of anomalies and problems
found elsewhere. So my postulated *Satipatthana Mula is not mechanically produced by assuming
that the concordance of the texts 'must' be the original core. I have also considered the internal
coherence of t h e material, intertextual relationships with the rest of the scriptures, cultural and
philosophical contexts at the time of the Buddha, evolution of sectarian positions, and so on. I
think all of this evidence taken together constitutes a powerful case for the main thrust of my
reinterpretation of satipatthana. Using the concordance of material from all the sources we end
up with a text that is internally logical and consistent, fits neatly with the satipatthana teachings
in the rest of the canon, and closely resembles an existing text.

In such a delicate operation I run a serious risk of being misunderstood, so I must make my claims
explicit. The tradition supplies us with a rational explanation of how the Satipatthana Suttas were
produced; that is, they were spoken by the Buddha. That may be so. However, for a number of
reasons I find this implausible. Most importantly, the traditional explanation cannot account for
the divergences between the existing texts. If I wish to offer an alternative, I should supply a
demonstration of how the Satipatthana Suttas could have evolved that: l) is rational; 2) accords
with historico-critical methodology; 3) accounts systematically for the existing texts on
satipatthana; 4) enhances understanding of the subject; 5) allows us to draw inferences about the
evolution of doctrine in early Buddhism that may be tested by comparison with other texts; and
6), most important, is useful for practice of Dhamma-Vinaya. I believe this analysis fulfils these
criteria. This, at t h e very least, should be enough to shift the burden of proof.

It would certainly be unjustified to claim that this reconstruction of the Satipatthana Sutta offers
an exposition of satipatthana that is complete and exclusive. Only in the later developed versions
do we see attempts to assemble in one coherent system all of the chief teachings on satipatthana.
It is clear that the Buddha taught satipatthana very often in brief. The detailed meaning would no
doubt have been interpreted in light of the other discourses on t h e relevant topics. Probably any
of the discourses dealing with body contemplation could be brought in under the umbrella of
satipatthana, and so with the remaining three, too. All of the available traditions, while differing
on the detailed content, agree in treating satipatthana, especially body contemplation, in this
way. It is obviously not at all implausible to suppose that, say, the elements and charnel ground
contemplations were also authentic, but as I stated above, I do not see the inclusion or exclusion
of such specific exercises as significantly affecting the doctrinal orientation of the discourse. The
*Satipatthana Mula should be seen as pointing to certain core aspects of satipatthana accepted
among the pre-sectarian community, rather than as definitive and final. As such it remains
relevant as a way of highlighting these common features whether or not there was ever an actual
closely corresponding text. I present my postulated reconstruction in full, without the customary

186
elisions, in order to make it as explicit as possible. First, though, I will briefly run over my reasons
for inclusion or exclusion of some particular sections.

Setting: I think the original setting was at Kammassadamma, but I leave the setting out of the
*Satipatthana Mula to emphasize that this material is not really part of the discourse, but was
added by redactors.

Audience: I follow the lead of the Sarvastivada and have the discourse addressed to monks and
nuns. While the textual support for this is admittedly slim, this is my little effort to redress the
balance of 2500 years of male redactors. However, for convenience I retain the male pronoun,
despite the slight incongruity that results.

Path to Convergence: I have my doubts about this, as I think the evidence of the Samyuttas
implies that it was meant for a specifically Brahmanical context, and that is lacking in the
*Satipatthana Mula. However, I bow to the agreement of the three main Suttas and include it.

Auxiliary formula: Although absent f r o m two of the Suttas, I include it because of the agreement
of the Abhidhamma texts and t h e Prajnaparamita, and assume that its loss in the Sarvastivada
and Mahasanghika Suttas was through abridgement, as indicated in the Sarvastivada Samyukta.

Internal/external: I bow to the weight of sources and present the integrated version, rather than
following each exercise.

Body contemplation: I include the simile, which is found consistently in the Sutta versions,
including the Prajnaparamita, and assume its absence f r o m the Ekayana Sutra is accidental.

Refrain: I think all the refrains have their problems, but there is sufficient agreement among t h e
Sutta versions to point to a common ancestor, though perhaps none exactly reflect that source.
The most distinctive common elements seem to be that one establishes mindfulness for
knowledge and vision; and one dwells independent. I think there is some relationship between
the patissati of the Theravada and the 'vision' of the Chinese versions, but am not sure if t h e
Theravada changed 'vision' into 'mindfulness', or the Chinese mistook patissati for passati (or
whatever the exact dialectical reading was).

Ending: I repeat the 'path to convergence', but omit the guarantee of attainment, which is only
found in two versions, and could easily have been imported f r o m elsewhere in the canon.

*
Satipatthana Mula
This is the path to convergence, monks & nuns, for the purification of beings, for
surmounting sorrow & lamentation, for ending bodily & mental suffering, for
understanding the way, for witnessing Nibbana: that is, the four satipatthanas. What four?

Here, a monk or n u n abides contemplating a body in the body internally, he abides

contemplating a body in the body externally, h e abides contemplating a body in the body
internally & externally - ardent, clearly comprehending, mindful, having removed desire &
aversion for the world. He abides contemplating a feeling in the feelings internally, he
abides contemplating a feeling in the feelings externally, h e abides contemplating a feeling

187
in the feelings internally & externally - ardent, clearly comprehending, mindful, having
removed desire & aversion for the world. He abides contemplating a mind in the mind
internally, he abides contemplating a mind in the mind externally, he abides contemplating
a mind in the mind internally & externally - ardent, clearly comprehending, mindful,
having removed desire & aversion for the world. He abides contemplating a dhamma in the
dhammas internally, he abides contemplating a dhamma in the dhammas externally, he
abides contemplating a d h a m m a in the dhammas internally & externally - ardent, clearly
comprehending, mindful, having removed desire & aversion for the world.

And how, monks & nuns, does a monk or nun abide contemplating a body in the body?

Here, a monk or n u n reviews this very body up f r o m the soles of the feet and down f r o m the
tips of the hair, bounded by skin and full of many kinds of impurities thus: 'In this body
there are head-hairs, body-hairs, nails, teeth, skin, flesh, sinews, bones, bone-marrow,
kidneys, heart, liver, spleen, lungs, large intestines, small intestines, contents of the
stomach, faeces, bile, phlegm, pus, blood, sweat, fat, tears, grease, spit, snot, oil of thejoints,
and urine.' Just as if there was a bag with an opening at both ends, full of various sorts of
grain, such as hill rice, red rice, m u n g beans, peas, millet, and white rice, a man with good
eyes were to open it up and review it thus: 'This is hill rice, this is red rice, these are m u n g
beans, these are peas, this is millet, this is white rice.' I n j u s t t h e same way, a monk or nun
reviews this very body up from the soles of the feet and down f r o m the tips of the hair,
bounded by skin and full of many kinds of impurities thus: 'In this body there are head-
hairs, body-hairs, nails, teeth, skin, flesh, sinews, bones, bone-marrow, kidneys, heart, liver,
spleen, lungs, large intestines, small intestines, contents of the stomach, faeces, bile,
phlegm, pus, blood, sweat, fat, tears, grease, spit, snot, oil of thejoints, and urine.'

Mindfulness on the body is well established for the sake of knowledge & vision. One abides
independent, not grasping at anything in the world. That is how a monk or n u n abides
contemplating a body in the body.

And further, monks and nuns, how does a monk or n u n abide contemplating a feeling in the
feelings?

Here, when feeling a pleasant feeling a monk or nun understands: 'I feel a pleasant feeling.'
When feeling an unpleasant feeling he understands: 'I feel an unpleasant feeling.' When
feeling a neither pleasant nor unpleasant feeling he understands: 'I feel a neither pleasant
nor unpleasant feeling.'
When feeling a carnal pleasant feeling he understands: 'I feel a carnal pleasant feeling.'
When feeling a spiritual pleasant feeling he understands: 'I feel a spiritual pleasant feeling.'
When feeling a carnal unpleasant feeling he understands: 'I feel a carnal unpleasant feeling.'
When feeling a spiritual unpleasant feeling he understands: 'I feel a spiritual unpleasant
feeling.'
When feeling a carnal neither pleasant nor unpleasant feeling he understands: 'I feel a
carnal neither pleasant nor unpleasant feeling.'
When feeling a spiritual neither pleasant nor unpleasant feeling he understands: 'I feel a
spiritual neither pleasant nor unpleasant feeling.'

Mindfulness on feelings is well established for t h e sake of knowledge & vision. One abides
independent, not grasping at anything in the world. That is how a monk or n u n abides
contemplating a feeling in the feelings.

188
And further, monks and nuns, how does a monk or nun abide contemplating a mind in the
mind?

Here a monk or n u n understands mind with lust as 'mind with lust'.

He understands mind without lust as 'mind without lust'.
He understands mind with anger as 'mind with anger'.
He understands mind without anger as 'mind without anger'.
He understands mind with delusion as 'mind with delusion'.
He understands mind without delusion as 'mind without delusion'.
He understands contracted mind as 'contracted mind'.
He understands distracted mind as 'distracted mind'.
He understands exalted mind as 'exalted mind'.
He understands unexalted mind as 'unexalted mind'.
He understands surpassed mind as 'surpassed mind'.
He understands unsurpassed mind as 'unsurpassed mind'.
He understands mind in samadhi as 'mind in samadhi'.
He understands mind not in samadhi as 'mind not in samadhi'.
He understands released mind as 'released mind'.
He understands unreleased mind as 'unreleased mind'.

Mindfulness on the mind is well established for the sake of knowledge & vision. One abides
independent, not grasping at anything in the world. That is how a monk or n u n abides
contemplating a mind in the mind.

And further, monks and nuns, how does a monk or n u n abide contemplating a dhamma in
the dhammas?

Here, when there is sensual desire in him, a monk or n u n understands: 'There is sensual
desire in me'. When there is no sensual desire in him, he understands: 'There is no sensual
desire in me'. And he understands how the arising of the unarisen sensual desire comes to
be. And he understands how the abandoning of the arisen sensual desire comes to be. And
he understands how the non-arising in the f u t u r e of the unarisen sensual desire comes to
be.

When there is anger in him, he understands: 'There is anger in me'. When there is no anger
in him, he understands: 'There is no anger in me'. And he understands how the arising of
the unarisen anger comes to be. And he understands how the abandoning of the arisen
anger comes to be. And he understands how the non-arising in the f u t u r e of the unarisen
anger comes to be.

When there is sloth & torpor in him, he understands: 'There is sloth & torpor in me'. When
there is no sloth & torpor in him, he understands: 'There is no sloth & torpor in me'. And h e
understands how the arising of the unarisen sloth & torpor comes to be. And he
understands how the abandoning of t h e arisen sloth & torpor comes to be. And he
understands how the non-arising in the f u t u r e of the unarisen sloth & torpor comes to be.

When there is restlessness & remorse in him, he understands: 'There is restlessness &
remorse in me'. When there is no restlessness & remorse in him, he understands: 'There is
no restlessness & remorse in me'. And he understands how the arising of the unarisen
restlessness & remorse comes to be. And he understands how the abandoning of the arisen
restlessness & remorse comes to be. And he understands how the non-arising in the f u t u r e
of the unarisen restlessness & remorse comes to be.

189
When there is doubt in him, he understands: 'There is doubt in me'. When there is no doubt
in him, he understands: 'There is no doubt in me'. And he understands how the arising of
the unarisen doubt comes to be. And he understands how the abandoning of the arisen
doubt comes to be. And he understands how the non-arising in the f u t u r e of the unarisen
doubt comes to be.

When there is the enlightenment-factor of mindfulness in him, he understands: 'There is

the enlightenment-factor of mindfulness in me'. When there is no enlightenment-factor of
mindfulness in him, he understands: 'There is no enlightenment-factor of mindfulness in
me'. And he understands how the arising of the unarisen enlightenment-factor of
mindfulness comes to be. And he understands how the fulfilment through development of
the arisen enlightenment-factor of mindfulness comes to be.

When there is the enlightenment-factor of investigation of mindfulness of dhammas in him,

he understands: 'There is the enlightenment-factor of investigation of dhammas in me'.
When there is no enlightenment-factor of investigation of dhammas in him, he
understands: 'There is no enlightenment-factor of investigation of dhammas in me'. And he
understands how the arising of the unarisen enlightenment-factor of investigation of
dhammas comes to be. And he understands how the fulfilment through development of the
arisen enlightenment-factor of investigation of dhammas comes to be.

When there is the enlightenment-factor of energy in him, he understands: 'There is the

enlightenment-factor of energy in me'. When there is no enlightenment-factor of energy in
him, he understands: 'There is no enlightenment-factor of energy in me'. And he
understands how the arising of t h e unarisen enlightenment-factor of energy comes to be.
And he understands how the fulfilment through development of the arisen enlightenment-
factor of energy comes to be.

When there is the enlightenment-factor of rapture in him, he understands: 'There is the

enlightenment-factor of rapture in me'. When there is no enlightenment-factor of rapture
in him, he understands: 'There is no enlightenment-factor of rapture in me'. And he
understands how the arising of the unarisen enlightenment-factor of rapture comes to be.
And he understands how the fulfilment through development of the arisen enlightenment-
factor of rapture comes to be.

When there is the enlightenment-factor of tranquillity in him, he understands: 'There is the

enlightenment-factor of tranquillity in me'. When there is no enlightenment-factor of
tranquillity in him, he understands: 'There is no enlightenment-factor of tranquillity in me'.
And he understands how the arising of the unarisen enlightenment-factor of tranquillity
comes to be. And he understands how the fulfilment through development of the arisen
enlightenment-factor of tranquillity comes to be.

When there is t h e enlightenment-factor of samadhi in him, he understands: 'There is the

enlightenment-factor of samadhi in me'. When there is no enlightenment-factor of samadhi
in him, he understands: 'There is no enlightenment-factor of samadhi in me'. And he
understands how the arising of the unarisen enlightenment-factor of samadhi comes to be.
And he understands how the fulfilment through development of the arisen enlightenment-
factor of samadhi comes to be.

When there is the enlightenment-factor of equanimity in him, he understands: 'There is t h e

enlightenment-factor of equanimity in me'. When there is no enlightenment-factor of

190
 equanimity in him, he understands: 'There is no enlightenment-factor of equanimity in me'.
And he understands how t h e arising of the unarisen enlightenment-factor of equanimity
comes to be. And he understands how the fulfilment through development of the arisen
enlightenment-factor of equanimity comes to be.

Mindfulness on dhammas is well established for the sake of knowledge & vision. One abides
independent, not grasping at anything in the world. That is how a monk or n u n abides
contemplating a dhamma in the dhammas.

This is the path leading to convergence, monks & nuns, for the purification of beings, for
surmounting sorrow & lamentation, for ending bodily & mental suffering, for
understanding the way, for witnessing Nibbana; that is, the four satipatthanas.

Based on the above reconstruction, what can we say about the satipatthana method? Here I beg
leave to depart for a time from the rigorous strictures of textual analysis to offer a more
interpretive and personal reflection on what this text means for meditation.

Start with the body. The body is our primary object of attachment and identification, deeply
bound up with our most basic biological drives: sexual reproduction and the assimilation of food.
Our thoughts and concerns, our worries and plans are for the large proportion of our time
occupied with it - how to feed it, cloth it, house it, keep it comfortable. Any spiritual tradition
worthy of the n a m e must recognize t h e limitations of corporeal existence; and yet some overstep
the mark, following t h e dangerous and unbalanced path of rejecting, ignoring, or repressing the
body and its desires. The distinctively Buddhist approach is to walk fearlessly into the lion's den.
We plunge into the guts and sinews, the blood and the bones, making the body itself the prime
object of our meditation. This is not f r o m gruesome morbidity, but from the wish to truly
understand, accept, and let go of this our fleshly home. We contemplate both the principle of life
- the fragile, delicate breath - and the principle of death - a decomposing corpse.

The body is a solid and familiar roosting-place of consciousness, less changeable than thought,
and hence forms an ideal basis for stabilizing the mind. Attention is brought to the breath, to an
image of the parts of the body, or to the inner experience of physical properties such as hardness,
softness, heat, and cold. As awareness is continually refocused and refined, the mind sinks deeper
and deeper into the chosen object. We gain a direct and quite amazing apprehension of this body
that is habitually obscured beneath our desires, aversions, and fears. The more clearly we see a
particular aspect of the body, the more apparent it becomes that our everyday perception of the
body as an entity is largely an illusion concocted in our minds.

The image of the body in our mind becomes very subtle; so subtle that the mental aspect of
physical experience becomes prominent. We are moving into the contemplation of feelings.
Feeling in Buddhist context may be defined as the hedonic tone of experience that stimulates
reactions of attraction, aversion, or indifference. Normally feelings are somewhat dimly
perceived concomitants of experience that manipulate our attention into patterns of desire and
denial. We devote our lives to seeking pleasure and avoiding pain, but rarely do we take the
opportunity to acquaint ourselves more deeply with these processes. Feelings are notoriously
nebulous and changeable: physical feelings tend to be overpowered by the accompanying
physical impact, and mental feelings are enigmatic and complex. But by treating the
contemplation of feelings primarily as emerging f r o m the tranquillising process of body
contemplation these problems are minimized. For a time our feelings become more stable,

191
simple, and clear: a subtle and cool sense of rapture and bliss welling up from within the
meditation subject.

One of the most astounding revelations for any meditator is the changeability of the mind's
capacity to be aware, like an eye that dilates and contracts in response to the environment.
Normally this is extremely difficult to see; for we are seeing the seer itself. There is no external
measure. In the contemplation of mind we see how cognition operates under different conditions:
burnt up by lust, withered by bitterness, darkened and compressed by sloth. We see how the
mind opens up, blossoms, and expands under wholesome influences, so our knowing has more
clarity and focus. We become acutely aware of the mind as awareness itself, soft and tender as a
flower or a baby, yet at the same time possessed of incredible strength and resilience. At this level
of development the mind becomes an instrument of unparalleled sensitivity.

We maintain a clear-eyed awareness of the various contrasting feelings and mind-states that are
directly present in consciousness. Contrast sows the seeds of understanding. Here we are
cultivating a wonderful garden for the flowering of wisdom; but we have not as yet turned this
potential directly to inquiry into causes. In the contemplation of dhammas we become aware not
just of presence, but also of absence; and this is a deeper matter, for in seeing absence one sees
impermanence. But then the practice digs deeper still. Each factor is treated in terms of an
investigation into causes - one understands how the hindrances arise, how they are abandoned,
how the enlightenment factors are produced, and how they are developed to perfection. The
hindrances obstruct samadhi; the enlightenment-factors produce samadhi. Understanding
through careful investigation the causes of the hindrances and the enlightenment-factors, one
understands the causes leading the mind towards or away from jhana. Paying attention by way of
root or matrix eradicates the hindrances and arouses the enlightenment factors, while paying
attention away from the root does the opposite.

In body contemplation, we apply ourselves to the meditation object. Here, we are basically just
following the meditation instructions. Gradually we see the more subtle feelings and mind-states
more clearly, and as the practice matures one entersjhana. At first this will be more or less a hit-
and-miss affair. But as we repeat the practice over and over we understand why the mind is
sometimes peaceful and sometimes not. As wisdom deepens, samadhi becomes more reliable.
These are the central, most clearly and powerfully realized processes in our spiritual
consciousness, so the meditator will automatically treat this as a paradigm for understanding the
nature of conditioned experience in general. Thus the contemplation of dhammas sees the
understanding of samatha maturing almost inevitably into vipassana. The whole process of
satipatthana is so exquisitely normal it is almost misleading to call it a 'method'. One is not
deliberately applying an artificial, preconceived scheme; the various stages simply signpost the
unfolding of meditation.

CHAPTER 1 6 : TRENDS: SARVASTIVADA AND THERAVADA

We may now broaden the focus somewhat, considering a claim made earlier, that this analysis of
satipatthana allows us to make testable inferences about doctrinal developments. In this chapter
we may consider some unshared, possibly sectarian, material in the discourses of the two main
schools whose texts are largely available to us, namely the Theravada and the Sarvastivada, as
well some developments in the Abhidhamma period. Historically, then, we are here interested in
the Asokan and following period. The effort is made to incorporate some perspective on how the

192
religion as a whole was evolving in this time. The next chapter will look at the later period of
Indian Buddhism.

THERAVADA AND VIPASSANA

We have suggested that the difference in character of the emerging schools is reflected in the
different orientation of their versions of the Satipatthana Sutta. The Maha Satipatthana Sutta is
moving to a proto-abhidhammic stance, where vipassana is conceived as the systematic analysis
of a comprehensive array of phenomena. A similar idea occurs in the Anupada Sutta, 408 which is
one of only a few discourses f r o m the Majjhima Nikaya that has not so far been discovered in t h e
existing Agamas. There, the Buddha praises Venerable Sariputta, who is especially associated
with the Theravada school, for his practice of analytical insight based on t h e eight attainments. In
addition to the usual j h a n a factors, the discourse contains a unique long list of mental factors in
strikingly Abhidhammic style; indeed, the Anupada Sutta is one of t h e key texts which has been
invoked by the abhidhammikas to support the notion that the Buddha, even if he did not actually
teach the Abhidhamma Pitaka itself, at least taught in Abhidhamma style. But the Anupada Sutta
is clearly late. It consists chiefly of stock phrases and technical terms; if these are left out we are
left with only a few lines that may be considered the characteristic vocabulary of the Anupada
Sutta. These lines include at least three words suggestive of a late idiom (anupada, vavattheti, and
paramt). In addition, t h e text is poorly edited. T h e j h a n a factors are listed, as per the usual Sutta
idiom, with t h e conjunctive particle ca. But the remaining factors are listed in the Abhidhamma
style with no ca; they have clearly been inserted f r o m another source.

Another example is furnished by the Chachakka Sutta.40? The Theravada version is one of the
most incisive texts dealing purely with insight in the canon. The elegance of form and profundity
of content argue for its authenticity. The six groups of six dhammas that gives the discourse its
title are similar to the exposition of the second and third noble truths in the Maha Satipatthana
Sutta, suggesting an Abhidhamma leaning. The Theravada version celebrates the discourse by
having 60 monks attain arahantship at the conclusion. This detail is missing from the
Sarvastivada, which is unsurprising, since the traditions seldom agree in their accounts of
attainment while listening to a discourse. Generally, such stories are more profitably considered
as a guide to the propaganda value of a discourse for the particular tradition rather than as
genuine historical records. Both the setting and the contents of t h e Sarvastivada version display
considerable divergence, especially the inclusion of many groups of dhammas dealing with
meditation and samadhi practice, thus altering the pure insight orientation of the Theravada
version. The text is far less balanced and elegant, and shows all the signs of a gradual accretion. In
fact it has been pointed out by Watanabe that the added factors show very strong similarities to
the matika of the Dharmaskandha. 410

We have had occasion above to refer to t h e Maha Hatthipadopama, acknowledging the

significance of its mode of presenting the four elements through focussing finely on one aspect of
the four noble truths. As one of the weightiest discourses given by Venerable Sariputta, hailed by
the Theravada and others as the first abhidhammika, we are not surprised to see that, as with t h e
Anupada Sutta, certain aspects of the teachings suggest an affinity with the Abhidhamma. 411 This
discourse is one of the most detailed treatments of the elements in the early Suttas. While the
Theravada and Sarvastivada versions are very similar, there seem to be slight sectarian
differences. There are some editing anomalies in the Theravada that imply that the Sarvastivada,
especially in its general treatment of the external elements, may be a little more reliable. In our
current context, we are interested in what appears to be a subtle shift in the samatha/vipassana
balance. Where the Theravada has Venerable Sariputta saying that each element is to be
contemplated as 'this is not m i n e . ' etc., the Sarvastivada lacks this detail. 414 And where both

193
versions encourage one to avoid anger by reflecting on the simile of t h e saw, t h e Sarvastivada
complements this by adding the passage on spreading universal loving-kindness, a detail lacking
in the Theravada. 813 So while both versions clearly acknowledge both samatha and vipassana in
elements contemplation, there seems to be an ever-so-slight difference in orientation. Another
slight difference may be noticed in the details of how the perceptual process is described. 818 Each
gives three conditions for the arising of cognition; but the Sarvastivada presents all three as
either being present or absent, while the Theravada elaborates different cases where certain
conditions may be either present or absent. This is a unique passage, often quoted, and the
elaboration might be taken as evidence of a later, Abhidhammic, development, although t h e
evidence is much less strong than in the case of the Anupada Sutta. It would be interesting to see
if the specific differences are in fact worked out in the respective Abhidhammas. Our suspicion of
emerging sectarian perspectives in this discourse is confirmed by the ending. The Theravada
culminates simply by speaking of the ending of greed, anger, and delusion regarding the five
aggregates. But the Sarvastivada emphasises that such detachment must be towards the
aggregates past, future, and present, thus suggesting the continuity in time that is so typical of
this school.

Such discourses indicate a trend. Not only is there a shift in emphasis f r o m samatha to vipassana,
but the nature of vipassana itself is changing. The early discourses treat vipassana as
understanding principles, not as accumulating information. They do not treat vipassana in terms
of a comprehensive analysis of an objectively defined set of mind/body phenomena - that is why
the abhidhammikas have supported this idea by invoking these texts. In time, this idea would
grow alongside the idea of a path of pure insight that could dispense w i t h j h a n a . But here we have
merely the beginnings of a long slow process. In fact the Anupada Sutta treats the jhanic
experience as intrinsic to the ability to clearly and precisely define each mental factor; in this it
pre-empts the compilers of the DhammasanganT.

It seems most likely that such texts were formulated by the Theravadins specifically to authorize
their new direction. That is to say, it is not that Theravadin 'dry vipassana' meditation is
authentic because it is taught in the Satipatthana Sutta, but that the Satipatthana Sutta was
assembled in order to authenticate the move towards dry insight. Of course, we should give those
teachers of old the benefit of the doubt. They presumably believed they were merely 'drawing
out' the implications of the embryonic Sutta material that they were editing. I have already
emphasized that the process at this stage did not involve any radically new doctrines, but merely
a reshaping, a shift of perspective.

Having identified this trend, and having pinpointed it to the nascent Theravada, it can then -
with due caution and always seeking independent corroboration - be used as a precedent.
Teachings within the Suttas that are highly analytic and display the Theravadin Abhidhamma
style of pedantic, systematic repetitions may be suspected to be late. Thus our analysis provides
us with f u r t h e r interpretive tools.

SARVASTIVADA AND SAMADHI

If the Theravadin emphasis on vipassana as evidenced in their recension of the Satipatthana Sutta
is subtly discernable elsewhere in the Pali canon, so too the Sarvastivadin emphasis on samadhi is
apparent in their Madhyama Agama. A number of interesting discourses with no Pali cognates
deal with jhana. These texts are so little known that it seems worthwhile to summarize them
here.

194
Dependent Liberation

MA 44, MA 54, and MA 55 present versions of the doctrinal framework that I call 'dependent
liberation', known elsewhere as 'transcendental dependant origination'. The Sarvastivada
Samyukta Agama includes a samyutta comprised of discourses on this theme, most of which in
the Pali have been shifted to t h e Anguttara Nikaya. The elements of the framework occur in a
whole range of central teachings, yet a full exposition is lacking from the Theravada Majjhima. It
does seem as if this important teaching found less favour among the Elders of the Maha Vihara in
Sri Lanka. Each of the various versions of the dependent liberation presents a series of factors
unfolding in a conditional sequence that culminates in Nibbana. The sequences here are very
similar to the Pali, yet have no exact cognate.

MA 44: Mindfulness & clear comprehension 415 > protection of the sense faculties > protection of
precepts > non-remorse > gladness > rapture > bliss > samadhi > knowledge & vision of things as
they have become > repulsion > fading of lust > liberation > Nibbana. 416

MA 54: Honouring and attending upon > approaching > listening to t h e good Dhamma > giving
ear 417 > consideration of t h e meaning of the Dhamma > learning the Dhamma by heart 418 > recital >
reflective acceptance 419 > faith > right consideration 420 > mindfulness & clear comprehension >
protection of the sense faculties > protection of precepts > non-remorse > gladness > rapture >
bliss > samadhi > knowledge & vision of things as they have become > repulsion > fading of lust >
liberation > Nibbana. 421

MA 55: Ignorance > conceptual activities > cognition > name & form > six sense media > contact >
feeling > craving > grasping > existence > birth > aging and death > suffering > faith > right
consideration > mindfulness & clear comprehension > protection of the sense faculties >
protection of precepts > non-remorse > gladness > rapture > bliss > samadhi > knowledge & vision
of things as they have become > repulsion > fading of lust > liberation > Nibbana. 422

Venerable Anuruddha

In the Theravada Majjhima Nikaya, Venerable Anuruddha appears in a few discourses, typically
dealing with samadhi, but he only delivers one full-length discourse. The inclusion of three major
extra discourses by Anuruddha, the archetypical reclusive meditating sage, suggests that his
mode of teaching was more popular in the Sarvastivada than in the Theravada. We have already
seen that the two Anuruddha-samyuttas differ in this regard, too: the Sarvastivada emphasizes
samadhi where the Theravada has vipassana.

MA 80 (*Kathinadhamma Sutta): Although this charming story is not found in the Nikayas, t h e
background events are included in the commentary to Dhammapada 93. At Venerable
Anuruddha's request, Venerable Ananda organized a group of monks to sew replacements for
Anuruddha's worn-out robes. The Buddha noticed the monks sewing, and asked Ananda why he
had not informed him so that he could help in sewing the robes. The Buddha t h e n j o i n e d in with
the monks to help sew Anuruddha's robes. When they were finished, t h e Buddha lay down to ease
his sore back and asked Anuruddha to deliver a speech on kathina to the monks. Anuruddha
spoke of how he embraced the monk's life, observed the precepts, abandoned t h e hindrances,
developed meditation, attained the f o u r j h a n a s , and finally the six clear knowledges culminating
in arahantship. The Buddha sat up, praised Anuruddha, and encouraged the monks to practice
these kathina dhammas.

195
MA 218: Venerable Anuruddha is asked how a monk is said to die as a noble one. He explained
that if one attained the four jhanas one would die as a noble one, but not as absolutely noble.
However if one developed the six clear knowledges culminating in arahantship one was said to
die with a noble mind that was supreme and absolute.

MA 219: Similar, but here the question is how to die without distress. Venerable Anuruddha then
taught that one who had correct view and precepts beloved of t h e noble ones, the four
satipatthanas, the four sublime abidings, and the four formless attainments would die without
distress. However, only one who could eliminate the bodily touch (?) 423 and through
understanding evaporate the defilements would die absolutely without distress. Here the
mention of 'correct view' and 'precepts' echoes the Satipatthana-samyutta. The four
satipatthanas occur here in place of the f o u r j h a n a s , as occasionally in the Nikayas too.

Miscellaneous

MA 176: Describes four people: one whose meditation was regressing but they thought it was
progressing; one whose meditation was progressing but they thought it was regressing; one
whose meditation was regressing and they thought it was regressing; and one whose meditation
was progressing and who thought it was progressing. Here, 'meditation' probably refers t o j h a n a .

MA 117: Describes another four kinds of meditators. The following scheme is repeated for each of
the eight attainments.
1) Attained first j h a n a but does not hold on to the practice, does not pay attention to the
basis,424 but harbours thoughts connected with desire. They would not stand fast, nor
progress, but would regress.
2) Attained f i r s t j h a n a , holds on to the practice, pays attention to the basis, establishes their
mind on that dhamma and makes it one-pointed. They would not regress, nor progress,
nor become repulsed, but would stand fast with long-lasting samadhi.
3) Attained f i r s t j h a n a but does not hold on to the practice, does not pay attention to the
basis, but inclines their mind to the secondjhana, wishing to progress further. They would
not regress, nor stand fast, nor become repulsed, but before long would progress to the
secondjhana.
4) Attained f i r s t j h a n a but does not hold on to the practice, does not pay attention to the
basis, but inclines their mind to extinction, calmness, absence of desire. They would not
regress, nor stand fast, nor progress, but before long they would experience repulsion and
evaporate the defilements.

MA 222: To understand each of t h e 12 links of the dependent origination one should develop: the
four satipatthanas; t h e four right efforts; the four bases of psychic power; the f o u r j h a n a s ; the five
spiritual faculties; the five spiritual powers; t h e seven enlightenment-factors; the noble eightfold
path; the ten spheres of totality; 425 the ten dhammas of the adept.

So it seems possible to argue the case that the difference in emphasis, however slight, between
the Theravada and Sarvastivada versions of the Satipatthana Sutta may also be discerned
elsewhere in the early texts. I have been looking for evidence in the Nikayas/Agamas to support
this contention, and as you can see, I have not come up with much. If the difference is genuine, it
is certainly minimal. It is a mere shadow of a nuance of the divisions that were to wrack later
Indian Buddhism.

We might consider whether this divergence this reflects something of the religious and
philosophical orientations of the emerging schools. There is a range of hints that suggest

196
something of the different characters of the Theravada and Sarvastivada schools: the Theravada
was a more intellectual, scholarly, urban movement, while the early Sarvastivada emphasized
devotion, meditation, and forest dwelling. It is worth mentioning some of these points, for the
prevailing scholarly discussion of the Sarvastivada has focussed mainly on their later
Abhidhamma works and has thus tended to see t h e m as a scholastic movement, which was
perhaps not so true for the early school. 827

1. Patriarchs: The root patriarch of the Sarvastivada is Venerable Maha Kassapa, who is t h e
archetypal charismatic forest sage, rejecting settled monastic life in favour of the austere
life in the jungle. The Theravadins, however, regard as their root patriarch Venerable
Sariputta, renowned for his subtle analytic wisdom, and as a Dhamma teacher second only
to the Buddha himself; the Jains evidently saw him as encouraging the move f r o m the
jungle to the village monasteries. This difference is evident even in the Suttas. In t h e
Theravada version of the Maha Gosingavana Sutta, 827 Sariputta and Maha Kassapa refer to
each other with the familiar term avuso ('respected friend'), but in the Sarvastivadin
version 828 Sariputta addresses Maha Kassapa as bhante ('venerable sir'). This distinction is
maintained in the later patriarchs of the Asokan era. Moggaliputta Tissa of the Theravada
was a master dialectician, skilled in subtle logic and doctrinal niceties, whereas Upagupta
of the Sarvastivada was a reclusive and eccentric forest sage renowned for his
unpredictable teaching methods and his meditative attainments. 82 ? Time after time, his
countless disciples are said to have realized the Dhamma after developing the four
jhanas. 830
2. Sitting cloth: The sitting cloth is a small mat originally used primarily for sitting
meditation in the forest. The Sarvastivada Madhyama Agama frequently describes the
monks taking the sitting cloth, folding it, placing it over the shoulder, and going into the
forest to meditate. The Theravada almost always omits this detail; however it is
mentioned, interestingly enough, in the famous episode where t h e Buddha claims he
could live on for an aeon. Since this kind of idea is absolutely characteristic of the
Sarvastivada it is possible that this is evidence of Sarvastivada influence on the Theravada
canon. But, as so often, the texts don't allow any simplistic conclusions: the complete
Sarvastivada version of the Maha Parinirvana Sutra in Sanskrit omits to mention the
sitting cloth, but the shorter Sanskrit fragment does. Be that as it may, the references to
the sitting cloth indicate that the Sarvastivada had a preference for forest dwelling and
meditation.
3. Compassion: The four divine abidings are mentioned frequently in both schools, but there
are several discourses in the Majjhima where these are mentioned in the Sarvastivada but
absent in the Theravada. 831
8. Miracles: The Sarvastivada Madhyama tends to emphasize the miraculous more than the
rather sedate Theravada discourses. For example, the Theravada depicts Venerable
Ratthapala as walking f r o m his home to the forest, whereas the Sarvastivada have him
flying through the air.832
5. Devotion: In the Theravada texts the monks usually address the Buddha simply as
'bhante', but the Chinese has 'World Honoured One' (lokanatha). It is possible, however,
that this change occurred with t h e translation into Chinese. Some of the Sarvastivada
discourses include exuberant descriptions of the Buddha's appearance, 'reverberating
with light like a golden mountain', etc.833 Also the Bodhisattva theory seems to have been
at a slightly more developed stage.
6. Early Abhidhamma: The early strata of t h e Sarvastivada Abhidharma, such as the
Dharmaskandha and the SangTtiparyaya, seem to retain a more archaic, less specialized
mode of analysis than the Theravada Abhidhamma.

197
This difference reflects the orientation of these emerging schools. The Theravadins, with their
vipassana emphasis, were more rational, scholastic, urban. The Sarvastivadins were more faith-
orientated, emphasizing the unpredictable charisma of the forest sage. A similar distinction is
recognizable within the modern Theravada, with the forest monks devoting themselves to
samadhi, while their brothers in the city monasteries do dry vipassana. But it is almost inevitable
that the rugged earnestness of the forest tradition will become tamed and civilized, and will t u r n
away f r o m practice towards study. Sometimes this only takes a generation or two. And so t h e
later Sarvastivadins went on to develop a vast Abhidhamma commentarial literature, in which, as
we shall see in the next chapter, satipatthana b e c a m e j u s t vipassana.

T H E D E M O N OF T I M E

The differences between the Theravada and Sarvastivada Satipatthana Suttas, and to a lesser
degree the Samyuttas, reflect the fundamental schismatic issue that divided these schools - time.
The Sarvastivadins, preferring the evenness and constancy of samatha, shied away from the
contemplation of impermanence in satipatthana, while the Theravadins emphasized watching
the successive passing away of phenomena in experience. Vipassana sees discontinuity in time,
samatha sees continuity: taken together, one realizes the relativity of continuity and
discontinuity; taken separately one would tend to reify either continuity or discontinuity into an
absolute. In these incipient stages the differences were a mere matter of emphasis, not yet
consciously articulated. The need for clear and cogent justifications for these divergent
approaches was a driving force in the formulation of the metaphysics of time. The Theravadins
went on to develop a radical version of t h e theory of momentariness, holding that each dhamma
arises, stays, and passes away in an instant, leaving no remainder in t h e following instant. The
Sarvastivadins accepted a version of the theory of moments, but they also held the tenet, f r o m
which they derived their name, that 'all dhammas - past, present, and f u t u r e - exist'. The present
m o m e n t was seen as t h e manifest or effective mode of phenomena. Thus impermanence is
marginalized; ultimate reality is becoming changeless. We have seen hints of this perspective
emerging in their Smrtyupasthana Sutra and Dharmaskandha; the Kosa applies the fully-fledged
Sarvastivadin metaphysic of time to satipatthana. 434

The origins of this arcane idea should, I believe, be sought in the emotional response of the
Buddhist community to the acute sense of pain and loss with the passing of the Buddha. The
Sarvastivadins keenly felt that they lived in a diminished age, that the glory days of the religion
were inexorably passing. As an emotional rather than intellectual issue, it is articulated on the
mythic and symbolic level. Their patriarch Venerable Upagupta's role was to halt t h e passage of
time by sustaining the religion. The key myth had him binding Mara, the God of Death, with a
rotting carcase slung around his shoulders. Mara could do nothing to remove the stinking corpse,
and all his begging left Upagupta unmoved. 'The Buddha,' whinged Mara, 'He never did anything
nasty to me like this! All those terrible things I did to him, and never once did he punish me or try
to hurt me. You know, he wasn't such a bad guy after all.' And so Mara gained faith in t h e Buddha.
Immediately, Upagupta released his hold and t h e corpse vanished. Mara, with the fervour of the
newly converted, offered to do whatever he could in service of Upagupta, unsurpassed in skilful
means. Upagupta said that his only regret was that he had not lived in the time of t h e Buddha,
and had never gazed upon that glorious visage, so sadly passed away. 'But you, Mara,' he said,
'You were alive in t h e Buddha's time. You saw him often, and must remember his appearance
well. And, you are an unequalled master in changing your form. Now may you, out of gratitude to
me for releasing you from the ways of evil, assume the appearance of the Supreme Buddha!' Mara
was shocked. 'I couldn't do such a presumptuous thing!' he protested. But Upagupta implored and
insisted, until finally Mara agreed, on one condition: that when he changed his shape into the
Buddha, Upagupta was not, under any circumstances, to bow to him. After all, he's still Mara.

198
Upagupta agreed; but when he beheld the splendour of the Buddha's form created by Mara he
could not help himself - overwhelmed with rapture he fell to the ground and prostrated himself
before Mara.

This brilliant myth relates t h e poignant dilemmas tearing apart the Buddhist community: the
ambiguity of the Buddha-cult, with its obsession with worship of imitative idols; and the paradox
of longing to preserve forever the Buddha's precious dispensation, a dispensation whose essential
message is that all things must pass under the sway of the god of Death and illusion. This
ambiguity of time emerges clearly in the contrast between the Jatakas' emphasis on continuity of
character in time, with their implicit yearning for return and personal connection to the 'golden
age' of the Buddha, and the atomic, seemingly nihilistic, theories of momentariness that were
emerging in the contemporary Abhidhamma schools. This indicates a widening fracture in
consciousness between the popular and scholastic forms of Buddhism. It was in the pan-sectarian
folk Buddhism, far removed in spirit from the combative sectarianism of the Abhidhamma
scholars, that the Bodhisatta doctrine slowly emerged, with all its momentous implications for
Buddhist history. But the Theravada Abhidhamma scholars, for all their insistence on radical
momentariness, still betray a nervousness, amounting almost to neurosis, in their obsessively
repetitive texts, a massive attempt to freeze the Dhamma in a matrix of abstract, contextless,
changeless, and bloodless dhammas. I think t h e philosophical and psychological content of the
Abhidhamma - which in fact adds little to the Suttas - has been overemphasized at the expense of
ignoring t h e religious significance of this movement within Buddhism.

T H E VIBHANGA'S ABHIDHAMMA EXPOSITION

We have already examined the treatment of satipatthana in the Sutta Exposition of t h e Vibhanga,
an early stratum of the Abhidhamma Pitaka. Now we turn to the second part of the Vibhanga's
treatment, the Abhidhamma Exposition.

This treats satipatthana purely as it occurs in t h e abhidhammic 'transcendental jhana'. Note that
the idea of'transcendental j h a n a ' is presented here, quite correctly, as a peculiarly abhidhammic
doctrine. It would therefore be a mistake, according to the Abhidhamma Pitaka itself, to use this
concept to interpret the Suttas. It has its interest, however, in showing how closely the
Theravada school, at this early stage, equated satipatthana with their conception of jhana. The
basic passage is an adaptation of the standard description of transcendental j h a n a in t h e
DhammasanganT.

'How does a monk abide contemplating a body in the body? Here, on the occasion when a monk
develops transcendental j h a n a - which leads out [of samsara], brings dispersal [of rebirth], for the
abandoning of pernicious views, for the attainment of the first stage [i.e. stream-entry] - quite
secluded f r o m sensual pleasures, secluded f r o m unskilful qualities, he enters and abides in the
f i r s t j h a n a , which has initial & sustained application and the rapture & happiness born of
seclusion, on the painful way of practice with sluggish clear knowledge contemplating a body in
the body; on that occasion the mindfulness, recollection ... right mindfulness, enlightenment-
factor of mindfulness, path-factor, included in the path - this is called satipatthana. Remaining
dhammas are associated with satipatthana.' 435

This is to be repeated with appropriate variations for the variousjhanas, stages of enlightenment,
etc. Most of the variations are not spelled out in the text. The whole thing is then taken through
two rounds - one for t h e path, one for the fruit. The clumsiness of the phrasing is as evident in
the Pali as in the translation, since the passage is essentially just a conglomeration of technical
terms f r o m various sources.

199
Here there are a number of both continuities and discontinuities with the earlier accounts. The
basic descriptions of j h a n a and satipatthana are identical to the Suttas. The close relation
between the two is also characteristic of t h e Suttas, although they do not equate the two quite as
comprehensively as here. The mention of the painful way of practice is clumsy. In the Suttas this
is contrasted w i t h j h a n a s ; while it would be a mistake to see this in t h e context of the Suttas as
implying a separate path than jhana, it is incongruous to call the j h a n a itself 'painful'.
'Satipatthana' itself is defined just as sati; that is, satipatthana is simply the subjective act of
mindfulness. Other dhammas are 'associated with satipatthana'. This seemingly innocuous phrase
in fact reveals an underlying tension in the development of a strictly abhidhammic interpretation
of satipatthana; for 'association' (sampayutta) is a technical abhidhamma term that only applies to
interdependent mental phenomena, and yet here it is supposed to include the body as well. More
on this below.

The Theravada came to interpret the 'transcendental j h a n a ' discussed here and throughout t h e
Abhidhamma Pitaka as just a single 'mind moment' (cittakkhana) flashing by immediately before
enlightenment. However, t h e text gives us no reason to suppose that time should be interpreted
here in terms of the definitive, atomic theory of moments that came to dominate t h e later
metaphysics. The theory of moments was not yet developed at the time of the composition of the
Abhidhamma Pitaka. As far as I know the only reference to 'moments' in the Abhidhamma Pitaka
is to the 'moment of rebirth' in the Vibhanga. 436 There are plenty of contexts in the Abhidhamma
that treat time in an everyday sense. 437 As always with the historical method, we should try to
interpret, not by looking back through the lens of later tradition, but forward through the lens of
earlier tradition. The Abhidhamma Pitaka was obviously written by and intended for those who
were already familiar with t h e thought-world of the early teachings. The Vinaya Pitaka begins
each passage with 'On that o c c a s i o n . ' (tena samayena); the Sutta Pitaka uses 'On one o c c a s i o n . '
(ekam samayam); and the Abhidhamma Pitaka uses 'On whatever o c c a s i o n . ' (yasmim samaye). All
these idioms treat time in a non-specific, common sense manner. The difference between them is
not in the duration of time that they envisage, but in that the Suttas and Vinaya are specific while
the Abhidhamma is general. The Sutta and Vinaya idioms are intended to ground t h e teachings in
time and place, to lend them concreteness and historicity, emphasizing how they are true and
useful relative to context. The Abhidhamma wants to universalise, de-contextualise; this is part of
its movement towards a conception of abstract, absolute truth.

There is nothing in the description of transcendental j h a n a to suggest that it was meant to be

applied purely to the m o m e n t immediately preceding enlightenment, which was the developed
interpretation. On t h e contrary, the language clearly implies duration; the Pali term indicating
duration, viharati (abides), is mentioned twice. Time is treated in the transcendental j h a n a i n j u s t
the same way as the normal j h a n a leading to rebirth. For the Vibhanga, 'during' the
transcendental 'path', one 'abides' 'contemplating the body', 'cultivating, developing, making
much of a 'way of practice' that may be either 'sluggish' or 'swift', and which 'leads' to
enlightenment.

'Transcendental j h a n a ' is not contrasted with 'non-transcendental j h a n a ' in terms of time, but in
terms of object and result. The treatment of result is straightforward - non-transcendental j h a n a
leads to rebirth, transcendental j h a n a leads to enlightenment and dispersal of rebirth.

The treatment of object is trickier. For the DhammasanganT, non-transcendental j h a n a is based on

one of the various meditation subjects such as kasinas, divine abidings, corpses, etc. 438
Transcendental j h a n a on the other hand is based on emptiness, signlessness, or desirelessness. 43 ?
But the Vibhanga confuses that distinction. The problem arises because the Vibhanga wants to
apply the idea of transcendental j h a n a to the various wings to enlightenment. Of those groups,

200
satipatthana is the only one to specify the object of meditation. It would be difficult to explain
how, in the transcendental path and fruit, one was 'contemplating the body', since t h e object of
the transcendental j h a n a is supposed to be Nibbana. The later traditions seem to have hesitated
over this one; the sub-commentary to the Vibhanga suggests that the mention of the body,
feelings, mind, and dhammas here distinguishes the various satipatthanas by way of approach. 880
This seems to imply that one is not literally contemplating the body at this point, but that the
contemplation of the body has been the predominant preparatory factor. This of course is not
what the passage says. The Visuddhimagga addresses the issue thus.

'When they [i.e. the 37 wings to enlightenment] are found in a single consciousness in this way
[i.e. at the path-moment], just the one mindfulness that has Nibbana as its object is called the
"four satipatthanas" by virtue of its accomplishing the function of abandoning the notions of
beauty, etc., with regard to t h e body, etc.' 881

This is neat; but again, it is quite patently not what the Vibhanga is talking about. It seems that
the Vibhanga is caught in an awkward developmental stage. It is not clear whether it wishes to
present this transcendental j h a n a as a kind of 'vipassana samadhi' (if we may borrow still later
terminology) where one is abiding absorbed in the contemplation of t h e body as empty of self, or
as a kind of enlightenment experience. The DhammasanganT has forced a wedge between the
non-transcendental j h a n a and the transcendental j h a n a and identified the path with the latter.
But the Abhidhamma Pitaka remains close enough to the thought-world of the Suttas that it
struggles to apply this conception consistently. Not until the fully-fledged metaphysics of the
mature commentarial phase of abhidhamma were the implications of this breach made explicit.

It hardly needs saying that, apart f r o m the discrete mention of the word 'emptiness', vipassana is
entirely in t h e background during this exposition. Even 'emptiness' cannot really mean vipassana
here, for it applies just as much to the fruit as to t h e path. In fact it would seem as if this
presentation was intended to emphasize in t h e most explicit way possible how j h a n a is as
intrinsic to the very idea of satipatthana as it is to the path itself. The compilers of the
Abhidhamma seem to have taken to heart the Sutta saying that 'samadhi is t h e path'. So much the
stranger then, that the later conception of transcendental jhana, which was orthodox f r o m the
time of the Visuddhimagga, in time became one of the key conceptual tools used to wriggle out of
the necessity for practicing j h a n a as part of the eightfold path, substituting dry insight
meditation based on satipatthana: one only need enter the mind-moment of j h a n a at the time of
realization itself. This is not only a grave distortion of the Suttas; it is a misunderstanding of the
very nature of t h e transcendental path. This was conceived as a perfected path, the culmination
and consummation of the various practices that make up our spiritual journey. Thus
transcendental j h a n a is not a non-jhana, it is not something else that can be substituted f o r j h a n a ;
it is the ideal, the quintessence of jhana, which naturally emerges as the practice of j h a n a
matures in balance and harmony with the rest of the path.

OTHER A B H I D H A M M A T E X T S

To my knowledge, the superfluity of j h a n a is first explicitly suggested in the Puggala Pannatti. 882
Although most of the material in this minor Abhidhamma work is derived f r o m the Anguttara
Nikaya with only slight modifications and is therefore early, here the use of purely abhidhammic
doctrines shows that the ideas underlying the DhammasanganT must have been already current
when this passage was composed. It describes four people who are mentioned in the Anguttara
Nikaya.883 One gains 'samatha of the heart within' but not 'vipassana into principles pertaining to
the higher understanding'. A second person has vipassana but not samatha, a third has neither,
and a fourth has both. The Anguttara describes samatha here as steadying, settling, unifying, and

201
 concentrating the mind in samadhi, which is similar to the Puggala Pannatti's description as one
who gains the form or formless attainments. But whereas the Anguttara describes vipassana as
the seeing, exploring, and discerning of activities, the Puggala Pannatti speaks of one who
possesses the transcendental path and fruit. This is obviously incongruous - the discourse is quite
clearly speaking of the contemplative investigation of conditioned phenomena. For the Suttas,
both samatha and vipassana should be developed and only then will the one enter the path. 444 But
if one already has the transcendental attainments, why bother developing mere m u n d a n e j h a n a ?
A f u r t h e r incongruity is that the transcendental path and fruit, as we have seen, is invariably
described in the Abhidhamma in terms of jhana, yet here one is able to get the path and fruit
without having 'samatha of the heart within'. The passage does not clarify just how one can gain
the transcendental j h a n a without non-transcendental jhana; and this omission is made even
more pointed when we notice that this short passage in the Puggala Pannatti follows close behind
a full-length exposition of the gradual training, p r e s e n t i n g j h a n a s right in the heart of the path,
just as in the Suttas.449

The next discussion of satipatthana occurs in the Kathavatthu, a polemical work of t h e Theravada
school dedicated to refuting the wrong views of other schools of Buddhism. This is generally
agreed to be the latest book of the Abhidhamma Pitaka. A controversy arises, seemingly due to
the ambiguity we noted above between the normal, subjective meaning of satipatthana ('the act
of establishing mindfulness') and the objective sense required in one discourse dealing with
vipassana ('things on which mindfulness is established'). The heretic asserts that all dhammas are
satipatthana. This view is attributed by Buddhaghosa to the Andhakas, and seems to have been
held by the Sarvastivadins and MahTsasakas. The Theravadin quite properly shows the
incoherence of this idea. The four satipatthanas only manifest with the arising of a Buddha; if
there were no Buddha would all dhammas disappear? If everything is satipatthana, then do all
beings practice satipatthana?

The next book to consider is t h e Patisambhidamagga. This may be described as an elaborate

compilation of the path of practice f r o m the Theravadin perspective. The style is similar to the
Abhidhamma Pitaka, but it is included in the Khuddaka Nikaya of the Sutta Pitaka. However A.K.
Warder, editor of the English translation and a leading expert in dating Pali texts, regards its final
compilation at 100 B.C.E. or later, which makes it later than the bulk of t h e Abhidhamma Pitaka.
Bhikkhu Nyanatiloka concurs in placing the Patisambhidamagga later than the Abhidhamma. And
indeed, the treatment of satipatthana bears this out. The section on the body gives a unique list:
earth, water, fire, air, head hair, body hair, outer skin, inner skin, flesh, blood, sinew, bone, and
marrow. Feeling is simply pleasure, pain, and neutral feeling; thus the 'spiritual feelings'
associated with j h a n a are not explicitly mentioned. Mind is treated as the Satipatthana Sutta,
with the addition of the six kinds of sense cognition; we have already seen that the treatment of
citta in terms of vinnana in satipatthana signals the shift f r o m samatha to vipassana. In the
corresponding section on anapanasati, however, mind is defined with a list of synonyms lifted
f r o m t h e DhammasanganT. Dhammas are all dhammas except body, feeling, and mind; or, in t h e
anapanasati section, a list of 201 dhammas derived f r o m the beginning of the
Patisambhidamagga. Both of these are similar to the developed conception of dhamma embodied
in the Maha Satipatthana Sutta, the Dharmaskandha, and the Sariputrabhidharma. The position
of the Patisambhidamagga is thus curiously similar to the heretical view that had been refuted in
the Kathavatthu. The Patisambhidamagga wriggles out of the dilemma by means of an obscure
passage that seems to have become t h e accepted Theravada solution. The body (etc.) is an
establishing, but is not mindfulness; mindfulness is both an establishing and mindfulness. The
solution is a way of avoiding having to admit that the idea of considering the satipatthanas as
objects is really incoherent. It ignores the fact that 'establishing' and 'mindfulness' are quasi-
synonyms; 446 and the solution runs aground in the context of the fourth satipatthana: at least one
dhamma is mindfulness, i.e. the enlightenment-factor of mindfulness. Notice that the contexts

202
that emphasize the samatha aspect of satipatthana - most of the Suttas and t h e Abhidhamma
Vibhanga - treat satipatthana purely as subjective, whereas contexts that emphasize the
vipassana aspect - the Samyutta Vibhanga Sutta and the Patisambhidamagga - spell out the
objective interpretation.

The Patisambhidamagga virtually completes the process of 'vipassanizing' satipatthana. At first

satipatthana was primarily samatha, the way of gettingjhana. Then vipassana was seen to emerge
through understanding the process of samadhi in contemplation of dhammas only. Then, for one
already well established in all four satipatthanas, vipassana was introduced as an advanced mode
of contemplating them. 447 Next vipassana was introduced following each of the four sections. 448 In
the Theravada Satipatthana Sutta it became affixed at t h e end of each exercise within the four
sections. Finally in the Patisambhidamagga, each item in each section ('earth', 'water', etc.) is
contemplated f r o m t h e start in terms of impermanence, suffering, not-self, repulsion, fading
away, cessation, and relinquishment. The ultimate outcome of this process would be to
marginalize or discard the original four objects of satipatthana altogether, abstract the vipassana
aspect of satipatthana as constituting the real essence of the practice, and therefore treat
satipatthana purely as contemplation of impermanence, etc., on any miscellaneous phenomena.
We shall see that this step was in fact taken in the next strata of abhidhamma/commentarial
literature.

The Milindapanha, a later work of which versions are found in both Chinese and Pali, treats
mindfulness as a 'calling to mind' and a 'taking hold' regarding skillful and unskillful dhammas.
While not directly bearing on meditation as such, this is certainly in line with the interpretations
developed in this work, especially in contemplation of dhammas. For a fuller discussion the
reader is referred to Gethin's The Buddhist Path of Awakening.449

There are a few references to satipatthana in t h e paracanonical work the Petakopadesa. This is a
treatise on exegetical technique, parallel to the Netti, and of equally uncertain date. Satipatthana
is regularly treated as vipassana, with the methodology we have referred to above of opposing
the four satipatthanas to the four perversions. However, satipatthana, or more precisely the
second element of the compound, upatthana, is mentioned under the definition of samatha. The
word patthana is also included; although this is not really the second element of the word
'satipatthana', the commentaries treat it as such, so it was probably intended to evoke
satipatthana here. This is not one of the standard definitions as found in the DhammasanganT or
the Patisambhidamagga, and seems to be unique to t h e Petakopadesa. It starts with a series of
words formed f r o m the root stha, 'to stand'. An attempt to represent the effect in English might
look something like this:

'Whatever, of the mind, is standing, steadiness, stability, stasis, standpoint [patthana],

establishment [upatthana], samadhi, concentering, non-distraction, non-dispersal, tranquillity of
the heart, one-pointedness of mind; this is samatha.' 450

I might briefly mention here the Vimuttimagga. This is a commentarial compilation that seems to
have been the model for t h e Visuddhimagga. It only mentions satipatthana under anapanasati,
adding nothing to the Suttas. It does, however, introduce a very significant concept, the path of
'dry vipassana', which is conspicuous by its absence f r o m the Suttas and its prominence in
contemporary meditation circles. The path of dry insight, however, is here not connected with
satipatthana.

203
 CHAPTER 1 7 : LATER BUDDHISM

Let us complete o u r j o u r n e y with a brief survey of some statements on mindfulness in the later
Buddhist texts. This is not intended to be comprehensive or authoritative. My own acquaintance
with this vast and obscure literature is not great. I merely propose to present some bits and
pieces I have stumbled across in my reading. I am precariously dependent here on secondary
sources and translations, and so any attempt at interpretation is most tenuous; the primary
interest lies in exploring lines of continuity between the kind of perspective on satipatthana
developed in this essay and later traditions.

T H E THERAVADA COMMENTARIES

Commentarial material for satipatthana is found in the commentaries for the Satipatthana-
samyutta as well as the Satipatthana Sutta in the Majjhima and DTgha, and also the Vibhanga. The
Satipatthana Sutta received an extensive commentary, but since this has been translated into
English and the ideas have been incorporated in many modern books on meditation, I will treat it
only briefly here. 451 But first I will glance at the Samyutta commentary. Like the material in the
Samyutta Nikaya/Agama itself, this is much briefer than the Satipatthana Sutta commentary, and
has received much less attention. But for this very reason it may be of some interest.

Often the Samyutta commentary will give a brief explanation for a particular word or phrase. The
Majjhima will repeat exactly the same phrase, but then offer a much more lengthy elaboration.
This suggests - but only suggests - that the Majjhima commentary represents a later phase of
development. Both commentaries mention places or people in Sri Lanka, but the Samyutta does
so less often, leading me to wonder whether it might preserve something closer to the Indian
commentary from which the tradition claims it is derived. The Samyutta commentary
occasionally repeats large bodies of text found also in the longer commentaries, such as the
section on clear comprehension. Occasionally, too, it gives us extended narratives, such as the
story of Venerable Sariputta's passing away. One interesting passage has a number of Elders
giving different opinions as to what 'satipatthana', 'enlightenment-factors', and 'supreme
enlightenment' mean. All the Elders equate satipatthana there with vipassana. This is interesting,
for vipassana is nowhere mentioned in the basic definition of satipatthana at the start of the
commentary. Might this be interpreted as a sign of doctrinal evolution within the Theravadin
community? The Satipatthana-samyutta's references to samadhi are explained away using the
convenient commentarial terms 'momentary samadhi' and 'threshold samadhi'.

One interesting detail at the start of the commentary is its gloss on the 'path to convergence'. The
commentary says: ' "Ekayanvayan" ti ekayano ayam.' This is simply a resolution of the compound;
but the interesting thing is that the reading ekayanvayam is different from the Majjhima and other
versions, which have ekayano ayam. The sub-commentary remarks that this is a tradition of the
Samyutta reciters. The reading ekayano ayam, has in fact made its way back into the Samyutta in
the Burmese Tipitaka that I am using - making the commentarial gloss nonsensical - but
apparently ekayanvayam is preserved in some other editions. Since the phrase is, I believe,
primarily associated with the Brahmanical context given in the Samyutta and only secondarily
with the Majjhima context, and since we know that in some editions of the Samyutta the reading
ekayano ayam has been read back into the text from the commentary, it is tempting to think of the
reading ekayanvayam as the original one.

Another point worth noting in passing is that there are a large number of discourses that the
commentary completely ignores. These are, of course, usually those of negligible content.

204
The Satipatthana Sutta commentary subjects each of the aspects of satipatthana to a detailed
exposition utilizing the fully developed apparatus of the mature abhidhammic and commentarial
systems. Similar versions are given for both the Satipatthana Sutta and the Abhidhamma
Vibhanga. The Vibhanga commentary offers no explanation as to why the satipatthana material
there is so much shorter than the Satipatthana Sutta; sometimes it refers to Sutta material, as for
example when it speaks of the '18 kinds of body contemplation', which probably indicates
borrowing f r o m the Sutta commentary. There is strong emphasis on vipassana throughout;
nevertheless, t h e samatha perspective is not completely neglected. When recommending
approaches for different character types, body and feelings are suggested for samatha yogis,
while t h e mind and dhammas are appropriate for vipassana yogis.852 But it goes on to contradict
itself by asserting that, while the contemplation of the body concerns both samatha and
vipassana, the remaining three deal with pure insight only. The mention of the contemplation of
mind as vipassana-only is incongruous, for the commentary itself agrees that many terms in the
contemplation of mind refer t o j h a n a . In the case of feelings, we have seen that t h e Satipatthana
Sutta refers to 'spiritual' feelings, which the Suttas associate primarily with jhana. But the
commentary shifts the emphasis here by equating 'spiritual' feelings with the 'six kinds of mental
pleasure dependent on renunciation', a category that is more evocative of vipassana. In one place,
the commentary suggests that 'mindfulness' means samatha while 'clear comprehension' means
vipassana. 853 In the discussion on anapanasati it says, in conformity with the Suttas and the
Visuddhimagga, that 'the four jhanas arise in the sign of breathing. Having emerged f r o m the
jhana, he lays hold of either the breath or the j h a n a factors [for developing vipassana]. 858 The
section on clear comprehension has an interesting piece of advice for over-enthusiastic yogis:

'In this matter, a person who experiences pain in every m o m e n t due to standing long with bent
or stretched hands or feet does not get concentration of mind, his subject of meditation entirely
falls away, and he does not obtain distinction (jhana and so forth). But he who bends or stretches
his limbs for the proper length of time does not experience pain, gets concentration of mind,
develops his subject of meditation, and attains distinction.' 855

The Sub-commentary adds some interesting points:

'Mindfulness denotes samadhi, too, here on account of the inclusion of mindfulness in the
aggregate of samadhi.' 856

'Confusion is the state of mind which, because of the whirling in a multiplicity of objects, is
j u m p i n g f r o m thing to thing, diverse of aim, and not one-pointed.' 857

'If wisdom is not very strong in the development of concentration there will be no causing of
contemplative attainment.' 8 5 8

Thus throughout both the commentary and sub-commentary, although both strongly emphasize
vipassana, there remains a recognition of the samatha aspects of satipatthana.

DEVELOPMENTS IN SARVASTIVADA

There is some evidence that the Sanskrit traditions, starting with the Sarvastivada, emphasized
samadhi more than the Theravada. This emphasis was not only reflected in t h e philosophical
differences, but in lifestyle, too. Venerable Sanakavasin, the preceptor of Venerable Upagupta,
the most famous of the Sarvastivada patriarchs, was reported to have said:

205
 'Clothed in hempen robes, I have attained the five stages of jhana.
Seated i n j h a n a among the mountain peaks and lonely valleys, I meditate.' 459

Upagupta himself took over Sanakavasin's monastery at Mount Urumunda, which was called 'the
foremost of the Buddha's forest domains, where the lodgings are conducive to samatha.' 460 These
sages embodied the austere forest tradition - clad in unkempt hempen rags, living in remote
mountains andjungles, unpredictable, sometimes depicted with long hair and beard: feral monks.
It has been argued, with some plausibility, that this kind of lifestyle inspired the early Mahayana,
which began as a 'back-to-nature' reform movement of unconventional forest yogis who, as
'bodhisattvas', took as their chief inspiration t h e ascetic, meditative lifestyle of the Bodhisattva.
Similar reform movements spring up f r o m time to time as a necessary countermovement to t h e
tendency of religions to urbanize and ossify.

A difference in the attitude towards mindfulness in these different schools is evidenced in their
respective Abhidhammas. For the Theravadins, mindfulness was an exclusively skilful mental
quality; it could not coexist with unwholesome states of mind. This leaves a rather embarrassing
gap. Despite attempts to systematically list all possible mental factors, the Theravada
Abhidhamma has no t e r m for memory. If sati meant memory, this would mean that one could
have no memory of unskilful states of mind, which is, alas, all too obviously not the case.
Venerable Nyanaponika was perhaps the first abhidhammika to notice this anomaly; he suggests
sanna could perform the role of memory. But while sanna has some connection with memory, it is
not used in the required sense of 'recollection'. Sanna is always present in consciousness,
recollection is not. This problematic position of the Theravadins seems to have developed out of a
wish to exalt the role of mindfulness. The Sarvastivadins, with no such agenda, were happy to
take the Sutta references to 'wrong mindfulness' at their face value and treat sati as both good
and bad. In general agreement with the Indian traditions, they treat sati as the 'not-forgetting' or
'retention' of the object, the, as it were, 'repetition' of the object leading to non-distraction. These
descriptions suggest the samatha dimension of mindfulness.

We have already discussed in some detail the treatment of satipatthana in the Dharmaskandha,
noting the close connections but also divergences f r o m the Sarvastivada Smrtyupasthana Sutra.
While the Dharmaskandha forms the basis for the canonical Abhidhamma of the school, t h e
massive Jnanaprasthana forms its culmination. This work encompasses all the fields of
Abhidhamma within its very broad purview; in this sense it does not seem to have a close analogy
in the Theravada Abhidhamma. It was probably composed around the close of the canonical
Abhidhamma period, and indeed in the Sarvastivada there seems to be a less sharp distinction
between the canonical works and later treatises than in the Theravada. The chapter on
satipatthana contains a combination of some verbatim quoting and some commentary. 461 The
section on body contemplation is mainly devoted to the f o u r j h a n a s . This is a little curious, and
certainly shows that this passage is dependent on the Sarvastivada Smrtyupasthana Sutra, rather
than any pre-sectarian text, or even the Dharmaskandha. The sections on feelings and the mind
are also similar to the Sarvastivada Smrtyupasthana Sutra. Contemplation of dhammas includes
the five hindrances, six sense media with fetters, and seven enlightenment-factors. Here the
content is the same as both the earlier Sarvastivada sources, but the sequence is shared only with
the Dharmaskandha, while the Sarvastivada Smrtyupasthana Sutra places the sense media at t h e
start of this section, which seems less convincing. Anyway, we can confirm that the
Jnanaprasthana is based on t h e existing Madhyama Agama, which adds another detail to the
evidence that this is indeed a Sarvastivadin collection. Also it seems clear that the Sarvastivadin
emphasis on j h a n a as a central part of satipatthana, especially body contemplation, continued
through this period.

206
This makes the position of our next text somewhat surprising. Frauwallner has an interesting
discussion of a Sarvastivada text, the Abhidharmasara of DharmasrT. This post-canonical treatise
is one of the earliest Abhidharma philosophical systems, and many of its features were taken over
by the later Abhidharmakosa and even the Mahayana. Here are some of the most relevant of
Frauwallner's comments.

'If we now compare these [Sutta descriptions of the satipatthanas] with DharmasrT's description,
it is striking that even taking the concision of the latter into account, there are no individual
correspondences. The contemplation of the body as impure, impermanent, suffering, and non-
self is important in DharmasrT's account as preparation for what follows, but it has no
counterpart in the Satipatthana Sutta. The treatment of the contemplation of dhammas is also
completely different...There is no question [in the Satipatthana Sutta] of uniting the objects of all
the satipatthanas and their contemplation as impermanent, empty, non-self, and suffering as in
Dharmasri's work. It is, however, important in the latter as preparation for what is to follow.
Under these circumstances it would seem justified to regard the use of t h e satipatthanas in the
new doctrine merely as a means of gaining a canonical starting point with which t h e new
doctrine can be linked and f r o m which it can as far as possible be derived.

...'In the canon, the attainment of the liberating cognition and thus of the elimination of the
defilements is premised by entry into the state of jhana.

'In DharmasrT's version of the new doctrine there is no mention of this. Without interruption t h e
satipatthanas, the skilful roots, and the path of vision follow one another without any mention
being made of entry into meditation [jhana]...This is a radical innovation.' 462

The only mention of j h a n a in the system of the Abhidharmasara would seem to be the 'peak non-
transcendental dhamma' (laukikagradharma). This occurs, as 'non-return', 'jhana-interval', or t h e
four 'root jhanas' for one m o m e n t only, immediately prior to the enlightenment experience.
Thus, despite the opposing labels, this 'peak non-transcendental dharma' seems to fulfil the same
function in the Sarvastivada system as the 'transcendental j h a n a ' of the Theravada: to maintain a
terminological continuity with the old Suttas while teaching a radically new doctrine.

It is hard to understand how the treatment of satipatthana could have changed so much in such a
short time. The major canonical work the Jnanaprasthana treats body contemplation mainly as
the four jhanas; then t h e earliest strata of post-canonical literature, composed in roughly t h e
same period, treats body contemplation as straight vipassana. I suspect that the answer lies in t h e
manner and role of t h e presentation of satipatthana within t h e overall framework, and must be
sought in a more detailed contextual study of these works.

T H E ABHIDHARMAKOSA

We can examine the Sarvastivada position in some more detail using Vasubandhu's classic
Abhidharmakosa. This is an interesting text, for it presents a thorough and clear description of
the field of Sarvastivada Abhidharma from the point of view of an author who is not committed to
that school, but has leanings towards the Sautrantika. The Kosa defines t h e intrinsic nature of
satipatthana not as 'mindfulness', but as 'understanding' (panna). The Sarvastivadins arrive at this
definition through taking the t e r m anupassana to express the essence of satipatthana. We have
seen that the Theravadins agree in taking anupassana as 'understanding', but they still treat
satipatthana itself as mindfulness, not wisdom. Here the Sarvastivadins r u n smack into absurdity.
They must conclude that satipatthana belongs, not with the path factor of right mindfulness, but
with right view; not with the spiritual faculty of mindfulness, but with understanding. 463 But the

207
same problem must arise for anyone who equates satipatthana with vipassana - satipatthana as
faculty, power, enlightenment-factor, or path factor is always distinguished f r o m understanding,
and when the factors are grouped together, it is included with samadhi.

However, a closer look suggests that mere terminological confusion is at least part of t h e
problem. Generally speaking, the description of the path according to the Sarvastivada, which
was later adapted by t h e Mahayanists too, falls into five stages: the paths of preparation,
reaching, vision, development, and the adept. A very simple summary is sufficient for our
purposes. The path of preparation includes all the early stages of the gradual training f r o m
learning the teachings, ethics, etc., up to jhana; the path of reaching is vipassana; vision is
stream-entry; development is the f u r t h e r development of the noble eightfold path by the noble
ones; and the adept is t h e arahants. In the path of preparation the approach to meditation is
exemplified with anapanasati for cutting off thinking and t h e ugliness of the body parts for
dispelling lust. 878 These are treated primarily as samatha; ugliness is specifically said to be not
impermanence, etc.879 Nevertheless, anapanasati is defined as 'understanding'. 8 7 7 The definition of
anapanasati and satipatthana as 'understanding' should be evaluated in light of t h e very broad
treatment of understanding in Sarvastivadin theory. For example, j h a n a is also defined as
'understanding', and is said to have samatha and vipassana yoked together. 877 Evidently in such
contexts we are to take 'understanding' in the sense of 'clear awareness', which we have seen is a
prominent theme in these contexts in the Nikayas/Agamas too.

In bizarrely direct contradiction with the Suttas, the Kosa says that anapanasati is cultivated only
with neutral feeling, because:

' . p l e a s a n t and painful feelings are favourable to thinking; thus anapanasati, which is the
opposite of thinking, cannot be associated with pleasure or pain. On t h e other hand, the two
agreeable sensations [rapture & bliss of t h e j h a n a s , apparently] form an obstacle to t h e
application of the mind to any object, and anapanasati can only be realized by this application.' 878

This makes me wonder whether Vasubandhu had ever meditated or even read the relevant
Suttas.87? The next section i s j u s t as odd. Whereas for the Nikayas/Agamas anapanasati and body
contemplation were part of satipatthana, here they are supposed to b e j u s t preliminaries.

'We have spoken of the two teachings, the visualization of ugliness and anapanasati. Having
attained samadhi by these two portals, now with a view to realizing insight...Having realized
stilling, he will cultivate the satipatthanas.' 870

Thus satipatthana is identified exclusively with insight developed on the basis of jhana. In this
the position of the Kosa seems to be quite different f r o m the Abhidharmasara, which as we have
seen omits j h a n a altogether in favour of vipassana. This satipatthana vipassana is supposed to
proceed by seeing each of the four satipatthanas in terms of their 'intrinsic essence', and also in
terms of their general characteristics as impermanent, etc. However the text dwells little on t h e
intrinsic essence, merely defining the body as primary elements and derived materiality, and
dhammas as everything that is not the other three; strangely, feelings and mind are omitted. The
focus is clearly on the general characteristics, and these are often talked of just in terms of
dhammas. 871 Thus satipatthana in the Kosa seems to virtually ignore the basic exercises of the
Suttas and treat satipatthana entirely in terms of the vipassana aspects. However this is at least
partly a mere change in the expression, for this satipatthana vipassana is of course undertaken
only after samadhi based on anapanasati or contemplation of body parts, and so is not presented
as 'dry insight'.

208
If we search a little we can discern echoes of the earlier significance of satipatthana too. The four
satipatthanas are said to be undertaken in sequence, for:

' . o n e sees first that which is the coarsest. Or rather: the body is the support for sensual desire,
which has its origin in the lust for feeling; this feeling occurs because the mind is not calmed, and
the mind is not calmed because the defilements are not abandoned.' 472

Or in the context of the spiritual faculties:

'In order to obtain the result in which one has faith, one rouses energy. When striving, there is
the establishing of mindfulness. When mindfulness is set up, one fixes t h e mind [in samadhi] in
order to avoid distraction. When t h e mind is fixed, there arises a consciousness that conforms to
the object [panna].'473

The exposition on the way of practice in the Kosa falls into confusion when it tries to treat
various frameworks such as satipatthana and the other wings to enlightenment as distinct stages
along the path, rather than as each offering complementary perspectives on the path as a whole.
The Kosa gives two alternative explanations of the progressive development of the 37 wings to
enlightenment, both of which place t h e satipatthanas before samadhi. 474 Thus even such
innovations, though underivable f r o m t h e Suttas, still preserve significant threads of continuity
with the early teachings. Of course the entire presentation is subsumed within t h e overriding
emphasis on understanding that is a definitive pan-sectarian characteristic of the Abhidhamma
project.

The Arthaviniscaya Sutra is an Abhidharma work of a similar Sautrantika/Sarvastivada affiliation

as the Kosa, organized around a version of the familiar samyutta-matika. The Sanskrit has been
published together with its commentary, although not knowing Devanagari characters I am
limited to reading Samtani's summaries. The text seems to present yet another straightforward
integrated internal/external version of the satipatthana formula, without mentioning
impermanence. 475 Later, right mindfulness is described in terms of contemplating the impurities
of the body, internally and externally. 476

The commentary presents the common scheme of opposing t h e four satipatthanas to the four
perversions of beauty, happiness, permanence, and self.477 Contemplating the body, etc., is
explained, as in the Kosa, as seeing both the individual and the general characteristics. The
general characteristics are impermanence, suffering, emptiness, and not-self (the Sarvastivada,
f r o m the Agamas onwards, regularly adds emptiness to the three more familiar in the Theravada).
The individual characteristics are described as follows:

Table 17.1: Individual Characteristics of the Satipatthanas in the Arthaviniscaya Sutra

Satipatthana Individual Characteristic

Body Primary and derived materiality (bhutabhautikatva)
Feelings Experientiality (anubhavatva)
Mind Beholding (upalabdhitva)
Dhammas All except those three

'Internal' is defined as pertaining to one's own continuum (santati), while 'external' pertains to
others. The repetitive idiom 'body as body' is explained by saying that it serves to avoid the
perverted mind that would arise from seeing 'body' as 'feeling', 'mind', or 'dhamma'. Following
the Sarvastivadin precedent the text says that understanding (panna) is the intrinsic essence
209
(svabhava) of the satipatthanas. The sequence of the satipatthanas is explained as the order in
which they arise. All in all, the text and commentary add little to our understanding of
satipatthana, except by showing the influence and consistency of the interpretations of t h e
schools.

OTHER EARLY SCHOOLS

Richard Gombrich mentions that Harvarman's Satyasiddhisastra of the BahusrutTya school uses
the SusTma Sutta t o j u s t i f y a path requiring a degree of concentration short of jhana. In this they
agree with the Theravada commentaries; but their position is n o t j u s t i f i e d by the SusTma Sutta
itself, in either the Theravada or Sarvastivada versions. The name of this school ('The Very
Learned', or perhaps 'The Followers of t h e Very Learned') confirms t h e correlation between t h e
move towards dry insight and the move out of the forest hermitages into the urban scholastic
universities. Nevertheless, it appears that they did not base their conception of the path of dry
insight on satipatthana. The Satyasiddhisastra analyses the 37 wings to enlightenment as either
samatha or vipassana. It describes the first three satipatthanas as samatha and the contemplation
of dhammas as vipassana. Mindfulness in the faculties, powers, and enlightenment-factors is also
treated under samatha. In this respect it would appear that this text is in agreement with t h e
early Suttas, t h e *Vibhanga Mula, and perhaps, in spirit if not in letter, with the Sarvastivada too.

Another important early school was the Puggalavada ('Personalist School'), whose distinctive
doctrine was that there is a 'person' who exists, neither identical with or separate from the five
aggregates, who undergoes rebirth, experiences the results of kamma, and who attains final
Nibbana. In this they naturally incurred the vigorous condemnation of the other schools, who
claimed they were reverting to the 'self of the non-Buddhist teachings. However, they rebutted
such claims with equal vigour, and survived to exert considerable influence on the doctrines of
the schools for over 1000 years. Bhiksu Thich Thien Chau has published a detailed study, The
Literature of the Personalists of Early Buddhism. Only four of their works survive in the Chinese; one
of these is on Vinaya, one deals specifically with their special doctrine of the person, and the
remaining two are very similar versions of an Abhidhamma treatise. Effectively, then, we are
limited to one work, known as the Tridharmakasastra. 478 This is attributed to a certain Giribhadra
of the VatsiputrTya sub-school of the Puggalavada, and possibly dates f r o m around the start of the
Common Era, together with a commentary of perhaps a few centuries later, and was translated
f r o m (hybrid?) Sanskrit into Chinese by Gautama Sanghadeva in 391CE.

The work first deals with the classic triad of good acts: generosity, ethics, and meditation, the
latter being described as t h e j h a n a s , divine abidings, and formless attainments. Then it goes on to
deal with the skilful roots, acceptance, learning, right thoughts, practice, ascetic practices, sense
restraint, and 'access samadhi'. This last t e r m is familiar f r o m the Theravadin commentaries in
the sense of an approach to jhana; but for the Puggalavadins it denotes rather the approach to
insight. All these topics are, of course, familiar to all the schools. The next section is on 'means',
(upaya), which here is a term for the triad of ethics, samadhi, and understanding. Samadhi (here
the Chinese translation suggests a Sanskrit original uttarasamatha, 'exalted tranquillity') is said to
include three factors of the eightfold path: energy, mindfulness, and wisdom. Mindfulness is
defined as the absence of forgetfulness regarding the four satipatthanas of body, feelings, mind,
and dhammas - internal, external, and both internal & external. Here 'internal' seems to be
defined as 'kammically acquired' (upadinna), a mainly abhidhammic technical term denoting this
set of aggregates, elements, and sense media that has been acquired as a result of past actions.
'External' refers to others. This practice eliminates greed, anger, and delusion. All of these
teachings on satipatthana are in perfect accord with the Suttas and the schools. The sectarian
perspective of the school, however, is brought out when the text argues that when the Buddha

210
said that 'he contemplates a body in the body', the word 'he' (so) refers to the ineffable person. 47 ?
Next the text deals with samadhi, focussing on the triad of 'emptiness samadhi', 'undirected
samadhi', and 'signless samadhi'. This set replaces t h e normal f o u r j h a n a s here, probably because
it is a triad; t h e title of t h e work means the 'Treatise of Threefold Dhammas', and the work does
indeed subsume much of its subject matter within groups of three. The f o u r j h a n a s were dealt
with earlier, where they were listed with the divine abidings and formless attainments, thus
making up an extended set of three groups of four dhammas. Hence the substitution of the three
samadhis for the f o u r j h a n a s here is more plausibly explained as dictated by the purely formal
exigencies of the context (an application of t h e anguttara-principle), rather than a genuine
doctrinal shift. Next is the section on wisdom; this is where the sectarian emphasis on the
'person' is prominent. The rest of the work deals with various matters unrelated to our present
purpose. However, it is interesting to note that under the topic of 'doubt', there is a
comprehensive list of samadhi attainments, all familiar f r o m the early Suttas, and conspicuously
lacking the commentarial innovations of 'momentary' and 'access' samadhi.

Leon Hurvitz has published an interesting translation of some Agama sutras on satipatthana
together with cognates f r o m the Pali canon, and Chinese commentaries by Fa Sheng and others.
The text gives the first vagga of ten Sutras in the Sarvastivada Smrtyupasthana-samyukta
(Hurvitz adds Pali cognates, not all of which are correct), and then addends a miscellaneous
discussion that has little to do with the particular texts at hand. The commentaries for the most
part agree with t h e Kosa. Vipassana is a strong theme throughout, with a special emphasis on
dependent origination (which recalls the Sariputrabhidharma), as well as Abhidhamma-style
analysis into 'atoms' and 'moments'. However, one of the commentators is careful to note that
right knowledge is produced by samatha. Body contemplation exercises mentioned are
anapanasati, ugliness, and elements. Initially one is to concentrate one's mind on these internally
only; according to one commentator only the perception of ugliness can be developed on the
bodies of others. The other objects of satipatthana are not specified, except that dhammas again
is perception and conceptual activities. Satipatthanas as objects is again discussed; the text claims
that the Buddha said that 'all dhammas' refers to the four satipatthanas; since this statement is
not found in the existing Nikayas/Agamas, it may be discounted. But the text rightly warns of t h e
dangers in this approach:

' . t h o u g h it is all-inclusive, its fields of perception tend to get out of hand and a certain
restriction is needed if the same goal, severance of the defilements, is to be achieved.'

Here the commentators, more explicitly than the other texts we have reviewed, treat the
contemplation of dhammas as encompassing t h e other three:

'Having entered into the dhammas, he takes a general look,

Beholding identically t h e marks of the dhammas:
"These four [objects of satipatthana] are impermanent,
Empty, not-self, suffering."'

The Mulasarvastivada were said by Warder and others to be a late offshoot of the Sarvastivada
(200 C.E.?), although recent scholars have questioned this. 480 It is difficult to isolate specific
doctrinal deviations f r o m the Sarvastivada, and it may be that their main innovations were
literary rather than doctrinal. They composed very long and elaborate Sutras and Vinaya in the
style made fashionable by the contemporary Mahayana sutras. Their Saddharmasmrtyupasthana
Sutra (sometimes called the Maha Smrtyupasthana Sutra) takes advantage of the trend towards
expansiveness that had already begun in the treatment of satipatthana by the early schools. 481 As
well as offering much doctrinal and meditative material it includes various descriptions of

211
heavens and hells and also includes references to the arts, painting, and theatre. Thus it
popularises the topic, placing satipatthana within the contemporary cultural movements of the
day. Indeed, some scholars have seen in one of the late Chinese translations of this work the
influence of the Kasmirean version of the Ramayana. I have not been able to locate the Sanskrit
text, but in a cursory survey of the Chinese I cannot find the satipatthana pericope or any of the
special features of the Satipatthana Sutta, suggesting that it shares little but the name. I have
come across the following interesting passage.

'When the time of death is approaching, he sees these signs: he sees a great rocky mountain
lowering above him like a shadow. He thinks to himself: "The mountain might fall on top of me",
and he waves with his hand as though to ward off this mountain. His brothers and kinsmen and
neighbours see him do this; but to t h e m it seems that he is simply pushing his hand out into
space. Presently the mountain seems to be made out of white cloth and he clambers up this cloth.
Then it seems to be made out of red cloth. Finally, as t h e time of death approaches, he sees a
bright light, and being unaccustomed to it he is perplexed and confused. Because his mind is
confused he sees all sorts of things such as are seen in dreams. He sees his f u t u r e father and
mother having sex, and seeing t h e m a thought crosses his mind, a perversity arises in him. If he is
going to be reborn as a man, he sees himself having sex with his mother and being hindered by
his father; or if he is going to be reborn as a woman, he sees himself having sex with his father
and being hindered by his mother. At that m o m e n t the intermediate existence ends, life and
consciousness arise, and causality begins once more to work. 882 It is like the imprint made by a
die; the die is then destroyed, but the pattern has been imprinted.' 883

While the style and subject matter might seem far removed f r o m the early Satipatthana Suttas,
the passage still retains an urgent interest in the questions of life, death, and causality. In fact the
connection between satipatthana and understanding rebirth is already implicit in the Samyutta
Nikaya, where the causes of the four satipatthanas may all be explained in terms of dependent
origination. Probably the key link is the term 'confusion'. Mindfulness, the antidote for confusion,
prepares us for death, the most important m o m e n t of our life. The whole process sounds
remarkably similar to samadhi experiences. The bright light appears like a samadhi nimitta; this
unexpected and unknown experience often causes confusion and fear in inexperienced
meditators. Then various signs and visions appear. If the dying person or the meditator loses
mindfulness and becomes distracted, the hindrances such as desire and anger arise, here exposed
in their most stark, Oedipal perversity. This and other aspects of this passage are found in the
'Tibetan Book of the Dead', and was possibly a source text for that work. The illustration of
rebirth with the simile of the die is also found in the Visuddhimagga, which is interesting, since
by the time of the compilation of these texts, the two traditions had already been separated for
perhaps seven hundred years. 888

T H E MAHAYANA

It will be fitting to conclude this survey with some details f r o m the treatment of satipatthana in
the main Indian schools of Mahayana. First, however, we may briefly look at how the great pair of
samatha and vipassana is treated in Mahayana meditation in general. I am frequently struck at
how the Mahayana texts emphasize the complementary nature of samatha and vipassana, where
the Theravada emphasizes their difference. This reflects the difference in philosophical
orientation between the schools. The Theravada, emphasizing the method of analysis, tends
towards an ontology of pluralistic realism, and thus sees samatha and vipassana as essentially
different 'things', while the Mahayana, emphasizing the method of synthesis, tends towards a
monistic (or perhaps 'holistic') idealism, and thus sees samatha and vipassana as, in themselves,
incomplete parts whose significance emerges when synthesized as a greater whole. An

212
outstanding work of modern scholarship is a collection of essays edited by Minoru Kiyota, simply
called Mahayana Buddhist Meditation. Here are some extracts.

' . o t h e r s having a like notion speak therefore of a "dry vipassana", in other words, a vipassana
without samatha, even though the presence of samatha is held to be an indispensable condition
for all vipassana.' (pg. 47)
'The perfect union of these two, mental stabilization and higher vision
(samathavipassanayuganaddha) is the immediate aim of Buddhist meditative practice, for all the
paths of Buddhism - whether HTnayana or Mahayana including Vajrayana - depend on this
coupling.' (pg. 47)
Geshe Sopa, 'Samathavipasyanayuganaddha: The Two Leading Principles of Buddhist Meditation'

'From the very beginning, it has been generally accepted that the higher reach of wisdom (panna)
is attained either through or accompanied by meditation (jhana, samadhi, and so on). Examples of
this idea can be seen in various formulae such as "the pairing of quietude and insight"
(samathavipassanayuganaddha), and in the last two of the "three disciplines" (sila-samadhi-panna),
the "five faculties" (saddha-viriya-sati-samadhi-panna), and the "six perfections" (dana-sila-khanti-
viriya-jhana-panna).' (pg. 66)
Gadjin M. Nagao,' "What Remains" in Sunyata: A Yogacara Interpretation of Emptiness'

'Meditation, the basis of the third kind of wisdom, is practiced in the following way: (1) by
mastering samatha or the tranquillization of the mind through the observation of moral and
yogic rules, nine stages of samatha, f o u r j h a n a s , and so on; then (2) by vipassana (analysis of the
object of meditation f r o m the point of view of what has been studied by investigation).' (pg. 135)
Yuichi Kajiyama, 'Later Madhyamakas'

' " . t h i s cognition, nonerroneous and free f r o m conceptualization with respect to an ultimately
real object, depends on a principal cause, which is samatha intimately conjoined with vipassana".'
(pg. 149)
Charlene McDermott, 'Yogic Direct Awareness'

'Now, in Indian Buddhism, vipassana is an exercise in the close scrutiny of the characteristics of
the skandhas, dharmas, and other compounded elements, and it is supposed to be done in
conjunction with exercises in tranquillity (samatha) or stilling the externally directed activities of
the mind.' (pg. 178)
'In commenting on the [Heart] sutra passage, "O Sariputta, form is not different f r o m emptiness,
emptiness is not different from form", he [Fa-tsang] first says that when one sees form as
identical with emptiness, one perfects the practice of samatha, and when emptiness is
contemplated as identical with form, one perfects the practice of vipassana, and he concludes
that when the two are practiced together, they are ideal.' (pg. 179)
Francis H. Cook, Fa Tsang's Brief Commentary

' . t h e practice of transferring merits is guided by insight (vipassana), and insight by meditation
(samatha). In the Upadesa, worship and praise [earlier identified as samatha] are the instruments
to purify t h e mind; vow and meditation [=vipassana] refer to the state which has realized pure
mind. The two are not distinct and apart' (pg. 259)
Minoru Kiyota, 'Buddhist Devotional Meditation'

Thus most of the essays in this collection speak clearly of the harmony of samatha and vipassana.
These fundamental aspects of meditation are reinterpreted in accord with t h e particular
perspective of the school. A good example of this is in the treatment of the Heart Sutra above.
The Heart Sutra boldly confronts us with a series of statements that, on the face of it, appear

213
paradoxical. In order to make sense of these enigmas, the commentator brings in the more
familiar, readily comprehensible teachings on samatha and vipassana. This does a number of
things: it explains the new and radical by showing connections with the already known; it
invokes basic Buddhist concepts, thus reassuring us that the Heart Sutra does not really mean to
destroy Buddhism; and it points out that t h e Heart Sutra is meant to inspire practice, not
philosophical speculation. Perhaps most importantly, the presentation of the commentator is
clearly intended to demonstrate at length that t h e relationship between samatha and vipassana is
integrative, rather than divisive, and he works f r o m this basis to show the integrative nature of
the Heart Sutra teachings as well. (The text does this with much more thoroughness than the
short quote above.) Several of the passages quoted above speak of 'samatha conjoined with
vipassana', a phrase derived f r o m a discourse by Venerable Ananda. 485 This says that all those who
declare they are enlightened do so in one of four ways: samatha preceding vipassana; vipassana
preceding samatha; samatha conjoined with vipassana; or else the mind is seized with
restlessness, but later becomes settled in samadhi. Practicing in each of these ways, t h e path is
born, then cultivated to enlightenment. These four options all involve a balance of samatha and
vipassana, and rule out any approach that would try to dispense with samatha altogether. Yet the
passage, on the face of it, treats each option as equal and does not, as t h e Mahayana
interpretations above do, praise 'samatha conjoined with vipassana' as superior to t h e other
options. So these interpretations emphasize the integrative approach, if anything, even more
than the early Suttas.

We may continue by briefly surveying some sayings on mindfulness f r o m a few of the

outstanding Mahayana writings.

'It is the persevering practice (of the four satipatthanas) that is called "samadhi." '
Nagarjuna, 'Letter to a Friend'

'He who has established mindfulness as a guard at the doors of his mind cannot be overpowered
by the passions, as a well-guarded city cannot be overcome by the enemy.'
Asvaghosa, 'Saundarananda Kavya'486

' . c o n s t a n t mindfulness
Which gains in keenness by devoted zeal
And zeal arises if one comes to know
The greatness that lies in inner stillness.'
Santideva, 'Siksasamuccaya', Karikas 7-8

'If an excessive preoccupation with external activities has been avoided with the help of
mindfulness & clear comprehension, then, thanks to them, the mind can steadily keep to a single
object as long as it wishes.'
Santideva, 'Siksasamuccaya'

Thus these great teachers all acknowledge the samadhi aspect of mindfulness. We have already
seen that t h e great Prajnaparamita Sutra, the cornerstone for all Mahayana philosophy, contains
an exposition on satipatthana inserted almost unchanged from the Satipatthana Sutta of an early
school. We may now examine a few other Mahayana Sutras in their t r e a t m e n t of satipatthana.

One of the important early collections of Mahayana Sutras is the Maha Ratnakuta (translated as A
Treasury of Mahayana Sutras). One of the discourses contained therein, translated under the title
'Sutra of Assembled Treasures', has a short passage on satipatthana, included as part of a long list
of various dhammas.

214
 'The four mindfulnesses cure clinging to body, feelings, mind, and dhammas. One who practices
the Dharma and contemplates the body as it really is will not be trapped by the view of a real self.
One who contemplates feelings.,.mind...dhammas as they really are will not be trapped by t h e
view of a real self. These four mindfulnesses, therefore, cause one to abhor the body, feelings,
mind, and dhammas, and thereby open the door to Nibbana.' 487

Here the vipassana aspect is emphasized. The passage does not appear to be very directly derived
f r o m the early satipatthana pericope. The emphasis on not-self is familiar from the Suttas,
although not directly in the context of satipatthana as such. The attitude of abhorring (nibbida?)
comes across as over-strong, although this may be due to the translations. The early Suttas have
more of a balanced approach, embracing both t h e attractive (breath, pleasant feelings, purified
mind, enlightenment-factors, etc.) and unattractive (charnel ground, painful feelings, defiled
mind, hindrances) aspects of experience within the purview of satipatthana. This strong
negativity is perhaps surprising given the Mahayana's reputation for a non-dualistic approach,
but this isjust one isolated passage.

The Avatamsaka Sutra is one of those vast, sprawling Mahayana Sutras that seems to find a place
for almost everything. The 26th chapter, in a discussion of the ten stages of t h e Bodhisattva's
progress, describes the fourth stage as 'Blazing', and includes a list of dhammas, including our
familiar 37 wings to enlightenment, starting as usual with the satipatthana formula. 488 This is a
version of the integrated internal/external formula, including the standard auxiliary formula. In
fact it appears to be identical with the pericope at SN 47.3/SA 636/MA 76*, the only change being
the substitution of 'bodhisattva' for 'bhikkhu'. Again we confirm t h e use of straightforward 'cut-
&-paste' insertions f r o m the early Suttas even in such an advanced Mahayana scripture.

Santideva's Siksasamuccaya, which I have quoted f r o m briefly above, includes many powerful
statements on satipatthana, in part collected from other Mahayanist works. Many of the passages
are collected in Nyanaponika Thera's widely available The Heart of Buddhist Meditation, so there is
no need to repeat t h e m here in detail. Suffice to note the inclusion of sectarian material,
continuing the trend of using the prestige of satipatthana to buttress one's position in t h e
energetic doctrinal debates that characterize much of written Buddhist history. The text quotes
the Arya Ratnacuda Sutra, giving a characteristically Mahayanist slant on the internal/external
contemplation of feelings.

'When experiencing a pleasant feeling he conceives deep compassion for beings whose character
is strongly inclined to lust, and he himself gives up the propensity to lust. When experiencing an
unpleasant feeling he conceives deep compassion for beings whose character is strongly inclined
towards hatred, and he himself gives up the propensity to hatred. When experiencing a neutral
feeling he conceives deep compassion for beings whose character is strongly inclined to delusion,
and he himself gives up the propensity to delusion.'

The contemplation of body includes a very powerful passage f r o m the DharmasangTti Sutra. This
includes an attack on t h e Sarvastivada doctrine of time:

'This body did not come f r o m the past and will not go over to the future. It has no existence in t h e
past or the f u t u r e except in unreal and false conceptions.'

T H E SAMADHI OF DIRECT ENCOUNTER WITH THE BUDDHAS OF THE PRESENT

A more interesting development is found in t h e Pratyutpannabuddhasammukhavasthitasamadhi-

Sutra, which translates as 'The Discourse on the Samadhi of the Direct Encounter with the

215
Buddhas of the Present'. I will refer to it amore economically as the 'Buddhas of the Present
Sutra'. I have given a Sanskrit extract of the satipatthana formula f r o m this sutra above. Here I
will condense the main passages dealing with satipatthana.

[18B] 'Further, Bhadrapala, Bodhisattvas who dwell contemplating a body in the body, but do not
think any thoughts (vitakka) connected with the body, and who dwell contemplating
feelings.,.mind...dhammas, but do not think any thoughts connected with
feelings...mind...dhammas - those Bodhisattvas obtain this samadhi [i.e. t h e samadhi of the sutra's
title]. Why is that, Bhadrapala? It is because if Bodhisattvas and Mahasattvas [practice
satipatthana in the way described], then they do not objectify any d h a m m a s . they do not
conceptualize or think discursively...they do not see any dhammas...that is known as unobscured
cognition. Bhadrapala, it is precisely unobscured cognition that is called samadhi. Bhadrapala, the
Bodhisattvas who possess this samadhi see immeasurable and incalculable Buddhas, and they also
hear the True Dhamma. On hearing it they master it. They also obtain the unobscured cognition
and vision of liberation and the unimpeded cognition of those Tathagatas, Arahants, Perfect
Buddhas.

[18C] 'Further, Bhadrapala, Bodhisattvas dwell contemplating a body in t h e body, and in doing so
do not see any dhammas whatsoever. Not seeing t h e m they do not conceptualize or think
discursively, even though they are neither blind nor deaf. Similarly as regards feelings, mind, and
dhammas. Not seeing t h e m they are not dependent; not being dependent they cultivate the path;
by virtue of having cultivated the path they have no doubts with regard to dhammas; and being
without doubts they see the Buddhas. And in seeing the Buddhas, by virtue of the fact that all
dhammas are unproduced, liberation occurs.

[18D] 'Why is that, Bhadrapala? If Bodhisattvas should adopt the perception of dhammas, that
itself would be for them the false view of an object of apprehension (upalambhadrsti). That itself
would be the false view of existence, of a self, a being, a soul, a person. That itself would be the
false view of aggregates, elements, sense media, signs, existing things, causes, conditions, and the
seizing of an object of apprehension.' 88 ?

The first thing that we notice is that t h e basic formula is adapted f r o m the Dantabhumi Sutta,8?0
including the special phrase, 'does not think thoughts connected with the body, etc.' Readings
vary between 'body' (kaya) and 'sensual desires' (kama). In the Dantabhumi Sutta the context does
not clarify the meaning; even though the interpretation might change, still both readings make
sense. But in the Buddhas of the Present Sutra the text unambiguously depends on t h e meaning
'body' (etc.). The fundamental purpose of practicing in this way is to attain samadhi, just as, in the
Dantabhumi Sutta, the practice immediately leads on to the jhanas. Here it is emphasized that
one with such samadhi does not 'objectify' or become 'dependent' on any dhammas. As we saw in
the discussion on satipatthana in t h e Prajnaparamita, the idea of lacking 'dependence' suggests
the 'independence' of t h e Satipatthana Sutta, although I am again unable to confirm whether it is
a truly cognate phrase. The argument is a standard Mahayana criticism of the Abhidhamma
schools, who took the dhammas as being really existent 'things'. Thus wrong views about
dhammas are said to b e j u s t as mistaken as wrong views about a self. The implication here is that
one who merely thinks about dhammas will misconstrue them, mistaking the intellectual
understanding of the teachings for true wisdom into the emptiness of all phenomena.
Satipatthana is enjoined in order to overcome the thinking mind so that one can 'see' without
'views'. It is fascinating to see how the schools can take the same agreed-upon fundamental texts
and develop t h e m in quite opposite ways. For in the Theravada, of course, the practice of
satipatthana was to become more and more a matter of seeing these ultimately existing
dhammas, and one is instructed to 'think thoughts connected with the body [etc.]' through t h e
habit of mental noting.

216
Another intriguing aspect of the Buddhas of the Present Sutra is t h e statement that one in such a
samadhi will see immeasurable Buddhas. This is a new innovation in satipatthana, and is in fact
the central theme of the discourse. The meditation on the Buddha in the early discourses is one of
the six 'recollections' (anussati), and we have noticed that this term, and some aspects of the
description of the practice, suggest a connection with satipatthana. This implication, however,
remains latent in the early period. The seeing of the Buddhas appears to be a kind of meditation
vision or nimitta. The text then says that one will 'hear the True Dhamma. On hearing it they
master it.' This points to the belief that one can, in a state of meditative concentration, hear the
Buddhas teach the authentic Dhamma, a Dhamma which is to be learnt by heart.

Here we surely have explicit evidence on the disputed question of the origin of the Mahayana
Sutras. The Mahayana Sutras emphatically claim to have been spoken by the Buddha, although
this is historically impossible. Are we to believe that the Mahayanists were so unscrupulous as to
deliberately forge new texts and palm t h e m off as authentic? There are a number of general
things to be kept in mind here: the ancient world, and specifically the early Buddhists, did not
have such an individualistic insistence on ownership and authorship of works; editing of the early
Suttas had been ongoing, thus making people accustomed to a somewhat fluid idea of what the
canon was; and, most pertinently for this context, the Mahayana philosophy was increasingly
emphasizing breaking down the distinction between the inner, subjective world and the outer,
objective world, thus paving the way for imagination to be considered as more-or-less on a par
with historical facts. In fact the desire to have new Sutras accepted as authentic developed hand-
in-hand with this anti-historical world view, hence the marvellous, mythic settings of the
Mahayana Sutras. Warder has suggested that the Mahayana monks may have been inspired by
meditative visions and to have interpreted these as stemming f r o m the Buddha in some mystical
sense. The Buddhas of the Present Sutra clearly says that that i s j u s t what happens. Even the title
of the Sutra suggests this. It includes the word sammukha, literally meaning 'face to face', which is
most familiar f r o m the early idiom used to emphasize that one has heard a teaching in t h e
immediate presence of t h e Buddha himself: 'Face to face with the Blessed One have I heard this,
face to face I have learnt it.'4?1 It is not uncommon today, even among Theravadins, for meditators
to see a vision of t h e Buddha, hear his teaching, accept it as authentic, and to teach it to others as
the Buddha's teachings. Such 'inspired' teachings can be straightforward re-statements of
Buddhist doctrines, insightful elaborations, or mildly eccentric reformulations; but sometimes
they are nothing more than mumbo-jumbo.

The status and meaning of meditative visions is f u r t h e r explained elsewhere in the Buddhas of
the Present Sutra, again relying on satipatthana, here the exercises in charnel-ground
contemplation. The passage is not exactly the same as the Satipatthana Sutta, yet it clearly
depicts the same practice. It is worth bearing in mind that the Satipatthana Sutta says 'as if one
would see a c o r p s e . ' (seyyathapi passeyya sarTram), thus depicting it as a purely imaginative
exercise.

[3j] 'For example, Bhadrapala, when a monk performing the meditation on ugliness sees in front
of him corpses - bloated...livid...putrefied...bloody...gnawed.'with the flesh peeled o f f . w i t h no
flesh and blood...white...shell-coloured...become skeletons - then those corpses have not come
f r o m anywhere, nor have they gone anywhere, they are not made by anyone, nor are they made
to cease by anyone. Yet, Bhadrapala, by that monk's mastery of mental focussing he sees the
skeleton lying in front of him.

'In the same manner, Bhadrapala, in whatever quarter Tathagatas, Arahants, Perfect Buddhas
might dwell, those Bodhisattvas who are supported by the Buddha and established in this
samadhi concentrate their minds on that quarter in order to obtain a vision of the Buddhas. By

217
 concentrating their minds on that quarter they see the Tathagatas, Arahants, and Perfect
Buddhas in that quarter. Why is that? Namely, Bhadrapala, this obtaining of a vision of the
Buddhas is a natural outcome of this samadhi. Bodhisattvas who are established in this samadhi
see t h e Tathagatas, and they appear to t h e m through the combination and concurrence of these
three things: t h e glory (anubhava) of the Buddha, the application of t h e force of their own
wholesome potentialities, and the power f r o m attaining samadhi.' 494

Thus the visualization of the Buddhas is said to be an imaginative exercisejust the same as the
visualization of a corpse. The almost shocking analogy implied between the Buddha and a rotting
corpse may be just an accident of the text; t h e corpse meditation may have been chosen as
example merely because it is one of the clearest examples of a meditative exercise in
visualization. Yet there is a poignancy to the juxtaposition, for by the time the Buddhas of the
Present Sutra was written, the historical Buddha had long since become the Buddha of the Past;
the living presence had become a corpse, combusted in the flames of impermanence. The
meditative visualization of the Buddha was perhaps the most potent of the many means
developed by the Buddhist faithful to resurrect the Buddha, to preserve the vital force of the
Teacher and his Teachings. In this way the meditation on the transience of life is transformed.

YOGACARA

We may pursue our enquiry f u r t h e r through examining the treatment of mindfulness in some of
the more methodical treatises, starting with the Yogacara school. They were a meditative school
whose distinctive philosophy is usually taken to be the assertion that 'mind only' exists, all else is
illusion. This opens t h e m up to the criticism that they are reverting to the Upanisadic position of
postulating consciousness as the ground of being, which is also equated with Nibbana; however
the early Yogacarins such as Vasubandhu and Asanga stated that even t h e underlying 'storehouse
consciousness' ceases in Nibbana. Nevertheless, Vasubandhu makes a couple of startlingly casual
references to Nibbana as the true Self.493 One of the key systematizers was Asanga, who lived in
the fourth century C.E. in North-West India.494 He was said to have been largely inspired by
Maitreya, the Bodhisattva currently dwelling in Tusita heaven awaiting rebirth to become a
Buddha. His Abhidharmasamuccaya is described as 'an extremely important work of the
Mahayana Abhidharma. It contains nearly all the main teachings of t h e Mahayana, and can be
considered a summary of all the other works by Asanga. The method of treatment of subjects in
this work is the same as the traditional method found in the texts of the Pali Abhidhamma which
preceded it by several centuries, such as the DhammasanganT, Vibhanga, and Dhatukatha: posing
a question and answering it.' 495 Not merely t h e 'Q & A' format reminds us of the early
Abhidhamma: we also find the first part treats the aggregates, elements, and sense media, using
Abhidhamma methods such as sangaha ('grouping') and sampayoga ('conjunction'); the second
part treats t h e four noble truths, including the path in detail; thus these two parts together are
similar to the samyutta-matika. Although the text is strongly defensive of the Mahayana
perspective, in fact most of the text deals with matter familiar f r o m the early Suttas, strongly
influenced by the Sarvastivada/Sautrantika systemizations.

Mindfulness is first mentioned under a general list of 52 mental factors described under the
umbrella of the aggregate of activities. While the list in general is fairly ad hoc, the sequencejust
at this point is reminiscent of the Suttas: mindfulness, samadhi, understanding. I give t h e
definitions for mindfulness and samadhi:

'What is mindfulness? It is non-forgetting by the mind (cetas) with regard to the object
experienced. Its function is non-distraction.

218
 'What is samadhi? It is one-pointedness of mind on the object to be investigated. Its function
consists of giving a basis to knowledge (/nana).'4?6

These are perhaps the most useful succinct definitions of these terms I have come across. 'Non-
forgetting' regarding an object that has already been experienced emphasizes the sustained
observation of anupassana. 'Non-distraction' is of course the basic function of samatha, and recalls
such contexts as satipatthana as the 'path to convergence', the 'basis for samadhi', and so on, and
particularly anapanasati in its function of non-distraction. And while mindfulness gives rise to
samadhi, samadhi in its t u r n has the proper function of giving rise to wisdom.

The path is described, in a way similar to the Kosa, as starting with the fundamentals of t h e
gradual training: ethics, sense control, moderation in food, putting forth effort, clear
comprehension, etc., and also the wisdom f r o m hearing, thinking, and meditating. Samatha and
vipassana are mentioned, without explanation (although the Kosa seems to mainly exclude
vipassana f r o m this stage). Next is a lengthy analysis, again leaning heavily on the Sarvastivada,
of t h e process leading to the vision of the four noble truths. Then the 'path of development' is
explained, largely consisting of the 37 wings to enlightenment, which are here treated exclusively
as applying to the noble disciples (in Theravada terminology the 'transcendental path'). The
explanations of satipatthana are clear and straightforward, and sometimes offer interesting new
perspectives. I will summarize the key aspects.4?7

Objects: Body, feelings, mind, dhammas; also 'things pertaining to oneself.

Intrinsic essence: Understanding and mindfulness.
Companion: Associated mind and mentality.
Development: Contemplation of a body in t h e body, etc., internally/externally/internally-
externally.
Internal: Internal material sense media.
External: External material sense media.
Internal-external: The external sense media, which are the seats of the sense faculties, and which
are linked to the internal sense media; and also the internal material spheres pertaining to
others.

The first explanation for internal-external seems to mean the internal as related to the external,
and suggests a synthetic interrelationship. The second half of this explanation suggests that the
internal spheres of others are external for oneself, and thus taken together they are 'internal-
external'. These explanations of internal-external differ somewhat f r o m the Theravada. 4 ? 8
Internal-external feelings, etc., are described as feelings, etc., produced by the internal-external
body. The most interesting aspect is the description of contemplation of the 'body in the body':
'contemplation of the identity (or similarity, samatapasyana) of the natural image of the body
(prakrtibimbakayasya) with the imagined counter-image of the body (vikalpapratibimbakayena).'
This clearly refers to t h e development of a visualized reflex-image in meditation, known in
Theravada as the 'counterpart sign' (patibhaganimitta). This appears as a reflective mirror-image
of the body, or of a part of the body such as the breath, in the period immediately preceding full
absorption. Although this is readily comprehensible, it is not at all clear how a similar
explanation would apply to the other satipatthanas. The text offers no help, merely saying that
the contemplation of feelings, etc., should be understood the same way.

The treatment of t h e five spiritual faculties and powers acknowledges that each factor supports
the next, including mindfulness as support for samadhi.4?? In the discussion of the noble eightfold
path it is said that 'right effort is the factor that dispels the impediments of the [major]
defilements (klesa), right mindfulness is the factor which dispels the impediments of the minor
defilements (upaklesa), right samadhi is the factor which dispels the impediments to the qualities

219
of distinction.' 500 In this the text demonstrates its close reliance on the early texts, visible even
among the many later elements.

There is another interesting passage on satipatthana in Asanga's Sravakabhumi. We have already

seen that the contents of the contemplation of mind are identical with the Sarvastivada
Sramanyaphala Sutra. The specification of t h e contemplation of the body is, however, very
different from the Suttas, listing 35 of the most diverse kinds of'body'. As usual, the satipatthanas
are opposed to t h e four perversions. Then the text says this:

'Another method is: where one does the action; the purpose for which [one acts]; who does t h e
action; and by what does one act. Comprising all that in brief, the four satipatthanas are
established. Therein: in the body one acts; for the sake of feeling; by the mind; by means of skilful
and unskilful dhammas'. 501

Stefan Anacker's 'Commentary on the Separation of the Middle f r o m the Extremes' offers a
translation of the Madhyantavibhagabhasya, which is, according to the translator, 'one of the
most striking works in t h e Mahayana literature'. It is a commentary by Vasubandhu 502 on verses
attributed to 'Maitreyanatha'. Vasubandhu was Asanga's brother, and it seems they shared t h e
same inspirational connection with Maitreya Bodhisattva.

Following the Sarvastivadin precedent the text tries to rationalize the traditional order of the sets
of wings to enlightenment as a progressive sequence (whereas for t h e Suttas the order is not
essential to the groups and simply organizes the sets according to numbers for the sake of
convenience). There is an attempt to equate t h e four satipatthanas with the respective noble
truths; this section has a refreshingly simple description of the contemplation of dhammas: 'lack
of confusion as regards dhammas that serve to afflict and dhammas that serve to alleviate.' 503 This
is identical with the conception of contemplation of dhammas we have seen in the early material.
Having accomplished this, one is supposed to undertake the four right efforts, and then develop
samadhi through the four bases of psychic power. There is considerable discussion on various
obstacles to meditation and antidotes; mindfulness is defined in agreement with the schools as
'the lack of loss of image, etc. of the meditation object'. 508 Elsewhere the function of mindfulness
is as antidote to 'secondary afflictions' because of 'the absence of slackness and excitedness in
mindfulness which is well-established in the preparatory causes for samatha, etc.' 505 Next,
continuing t h e sets of t h e wings to enlightenment, arise the five spiritual faculties:

'Having taken hold of faith, one undertakes energy, the result of this cause. Having undertaken
energy, mindfulness occurs, and through mindfulness having occurred, t h e mind enters samadhi.
When the mind is in samadhi, one knows as it is.'506

Here we are firmly in Sutta territory. The difference between the spiritual faculties and spiritual
powers is explained in terms of progressive stages of the path according to the Sarvastivada
system; and then the enlightenment-factors and the noble eightfold path arise in due order.

Although the above t r e a t m e n t is basically similar to the Kosa, now the text asserts what it claims
are three features distinguishing Mahayanist satipatthana: 507

1) The object of meditation for disciples is their own bodies, etc., while t h e bodhisattvas' is
both their own and others.

This is just wrong; as we have repeatedly seen, all strata of texts in all schools from the Suttas
onwards acknowledge both internal and external contemplation.

220
 2) Disciples contemplate the impermanence, etc., of the body, etc., while bodhisattvas use
the method of non-apprehension [of bodies, etc.].

This refers to the fundamental philosophical division between the Abhidhamma schools and the
Mahayana: the abhidhammikas, especially the Sarvastivadins, tended to treat the dhammas as
real substantial entities that possessed the characteristics of impermanence and so on. But the
'Emptiness School' (to which the author belonged) held that a dhamma, like a magical illusion,
'does not exist as it appears, with the state of possessing apprehended [objective] and
apprehendor [subjective] aspects, but yet it doesn't not 508 exist, because of the existence of the
illusion itself.'50? This is the most important and complex philosophical dispute in later Indian
Buddhism. Suffice to say here that, in my opinion, the abhidhammikas ventured significantly
beyond the Suttas in their ontological reification of dhammas; but t h e 'emptiness' reaction, with
some noble exceptions, did not distinguish between the doctrines of the Suttas and t h e
abhidhammikas, and hence tended to stigmatise all of the followers of early Buddhism as naive
realists. So this criticism, while it may have been pertinent in a certain context, does not apply to
those simply following the Suttas.

3) Disciples cultivate satipatthana for the sake of non-attachment to their bodies, etc., while
bodhisattvas practice neither for lack of attachment nor for nonlack of attachment, but
for Nirvana which has no abode.

This requires some interpretation, for of course all schools of Buddhism in fact practice for
'Nirvana which has no abode'. Presumably this is here intended to refer to Buddhahood, the
ultimate goal of the Mahayana schools. Mere non-attachment is perhaps thought to lack
compassion. No doubt a similar point is being made here as when the text says that studying,
reflecting on, and teaching the sutras of the Great Vehicle only is of great fruit, not of the Inferior
Vehicle, since the Great Vehicle is distinguished because of its kindness to others. 510 It is the
tedious old cliche about the selfishness of the disciples. Given how emphatically t h e Mahayana
stigmatized the early schools as selfish, it seems likely that there is some truth to the accusations,
in certain places and times. But it is naive to generalize this to all the early Buddhists; and to
demonstrate this I can do no better than to quote the words of the Master.

' "I will protect myself," monks: thus should t h e satipatthanas be practiced. "I will protect
others," monks: thus should the satipatthanas be practiced. Protecting oneself, monks, one
protects others; protecting others, one protects oneself.

'And how is it, monks, that by protecting oneself one protects others? By the cultivation,
development, and making much [of the four satipatthanas]. It is in such a way that protecting
oneself one protects others.

'And how is it, monks, that by protecting others one protects oneself? By patience, harmlessness,
loving-kindness, and sympathy. It is in such a way that protecting others one protects oneself.' 511

MADHYAMAKA

The other main Indian Mahayana school was the Madhyamaka. Although they arose earlier than
the Yogacara, here I treat them later, simply because the only relevant work of theirs that I have
access to is later tthan the Yogacara works discussed above. Whereas the Yogacara were better
known as a contemplative school, t h e Madhyamaka were renowned for their witheringly
sophisticated dialectic. However, they did not neglect meditation; the Bhavanakrama of
KamalasTla was a meditation manual from the Madhyamaka school of Santaraksita that seems to
have been prepared for introducing meditation to the newly converted Tibetans. This work was

221
studied in an essay by Yuichi Kajiyama titled 'Later Madhyamakas'. The general path is described
in the usual way as first mastering the scriptural and theoretical aspects, then developing
samadhi culminating i n j h a n a and formless attainments before undertaking vipassana. As in the
Yogacara account it is acknowledged that only on the vipassana level is there any significant
divergence f r o m the early schools. What is truly remarkable is that the course of vipassana seems
to be derived from the doctrinal evolution of the schools through history. That is to say, one is to
meditate successively seeing ultimate reality as it is presented by each of the main schools, then
to realize that this level of reality is in fact empty, and then to pass to higher, more subtle
perspectives, culminating, of course, in the ultimate emptiness of the Madhyamaka. Thus one's
individual consciousness quite explicitly evolves in reflection of the collective consciousness.
Even more remarkable, these stages of historical evolution as presented here clearly parallel the
four satipatthanas, even though (so far as the sources available to me reveal) the satipatthanas
are not explicitly invoked. Here I will quote f r o m Kajiyama's summary.

'In the foregoing sections taken f r o m KamalasTla's Bhavanakrama 1, four stages are plainly
distinguishable:

1) The preliminary stage in which external realities admitted in t h e systems of the

Sarvastivada and Sautrantika are presented as the object of criticism;
2) The stage in which only the mind with manifested images is admitted - the system of the
Satyakaravada-Yogacara school forms the object of meditation;
3) The meditation stage in which t h e objects of cognition as well as the duality of subject and
object are condemned as unreal and in which the knowledge without duality is proclaimed
to be real - this being t h e standpoint of the AlTkakaravada-yogacarin.
4) The stage in which even the non-dual knowledge or the pure illumination of cognition is
declared to be empty of an intrinsic nature. This latter stage is the highest one proclaimed
by the Madhyamaka.'

Thus t h e first stage sees the dhammas as substantial entities, paralleling t h e contemplation of the
body. The second stage admits the 'features' or objects of the mind, paralleling the contemplation
of feelings, which are the most prominent properties of the mind. The third stage only admits
cognition itself, corresponding to the contemplation of the mind. And the last sees only pure
emptiness, which is defined as 'dependent origination', just as the contemplation of dhammas
focuses not on seeing the phenomena in and of themselves, but as a matrix of conditions. Please
note that this is not to identify these things or deny their differences, but merely to indicate
certain relationships. I think the parallelism is both undeniable and significant. The original
sequence of the four satipatthanas embodies a natural progression, f r o m coarse to fine, that can
be discerned in experience. Like so many other Buddhist teachings it is a simple but extremely
subtle paradigm that is reflected in any number of manifestations. As such, for those steeped in
the teachings there is a tendency, whether conscious or not, to assimilate the principles, abstract
them, and apply t h e m in contexts quite removed from the original - as indeed we are doing here.
This has its use in discerning continuities and relationships, but it demands a corresponding re-
assertion of the original context if we are not to be cut adrift f r o m our mooring. As we have been
warned above, 'its fields of perception tend to get out of hand and a certain restriction is needed
if the same goal, severance of the defilements, is to be achieved'.

To examine the role of mindfulness in later Tibetan Buddhism, which is primarily derived f r o m
the Madhyamaka, I will rely on the extensive modern scholarly work Meditation on Emptiness by
Jefferey Hopkins. This is based on a variety of sources, including Indian Buddhist texts, Tibetan
treatises, and oral teachings by contemporary Tibetan monks. Here, mindfulness is treated
exclusively in the mode of samatha. The basic descriptions of mindfulness expand on Asanga's
definition given above.

222
 'Mindfulness is non-forgetfulness with respect to a familiar phenomenon; it has the function of
causing non-distraction. Mindfulness has three features:
l) Objective feature: a familiar object. Mindfulness cannot be generated towards an
unfamiliar object.
2) Subjective feature: non-forgetfulness within observation of that object. Even though one
might have become familiar with that object previously, if it does not presently appear as
an object of mind, mindfulness cannot occur.
3) Functional feature: causing non-distraction. Since the stability of the mind increases in
dependence on mindfulness, non-distraction is specified as t h e function of mindfulness.

. a l l achievements of samadhi in Sutra and tantra are attained through the power of
mindfulness.' 912

'Right mindfulness: constant attentiveness to objects of awareness and to modes of perception of

those objects necessary for ascending the paths.' 913

Hopkins presents some charts summarizing the Tibetan conception of the stages in the
development of samatha. They are similar in spirit but not identical with the early Suttas. There
is a list of five faults in samatha together with their remedies. 914

Table 17.1: Faults and Remedies of Samatha

Faults Remedies
Laziness (kausidya) Faith (sraddha), desire (chanda), effort (vyayama),
tranquillity (prasrabdhi)
Forgetting the instructions Mindfulness (smrti)
(avavadasammosa)
Slackness (laya) and restlessness Clear comprehension (samprajanya)
(auddhatya)
Lack of application (anabhisamskara) Application (abhisamskara)
[Over-] application (abhisamskara) Equanimity (upeksa)

Most of these seem to be sensible enough. Tranquillity as remedy for laziness is incongruous, and
probably stems from a fault in the transmission. Mindfulness is treated straightforwardly in its
old sense of 'memory'. The following list, which is to be read from t h e bottom up, also places
mindfulness among t h e factors developing samatha. 919

Table 17.2: Aspects of Samatha

Six Powers Nine Aspects of Samatha

Familiarity (parichaya) Concentration (samadhana)
Energy (vriya) > One-pointedness (ekotikarana)
> Strong peace (vupasamana)
Clear comprehension (samprajanya) > Peace (samana)
> Control (damana)
Mindfulness (smrti) > Close setting (upasthapana)
> Re-setting (avasthapana)
Thought (cinta) Continuous setting (samsthapana)
Learning (sruta) Setting the mind (cittasthapana)

223
Here again these lists are, generally speaking, similar to the early Suttas, yet have many
differences in details; we cannot accept that the changes constitute any improvement. The list of
six powers contains mental qualities that, with the exception of the topmost, parichaya, are
familiar from the Suttas; yet here the logic of the sequence has decayed. Most importantly,
energy, which for the Suttas precedes mindfulness, here follows after mindfulness and clear
comprehension. This causes the evident incongruity in the correlation with the list of the nine
aspects of mental stability: energy is equated with strong peace and one-pointedness. This is
reminiscent of the situation in the five faults of samatha discussed above, where tranquillity is
the opposite of laziness. Another incongruity is the equation of clear comprehension with peace
and control. In fact, it is fairly obvious that here we have two completely separate lists, each of
which has its own (imperfect) internal logic, which have been artificially collated at some later
date.

The nine aspects of samatha are identical with the definition of samatha in the
Abhidharmasamuccaya, and have obviously been derived from there or from a related source.917
However, in the early source there is no implication that sequence of the terms are meant to
describe the progress in meditation. They are simply a list of synonyms for samatha in typical
Abhidhamma style. Some time later they were re-interpreted to imply a progress in mediation by
forcibly reading arbitrary meanings into the terms. From the point of view of mindfulness, the
most interesting are the four terms that form the foundation of the nine aspects of samatha. They
culminate with upasthapana (=upatthana, here translated following Hopkins as 'close setting'
rather than my usual 'establishing'), the second member of the compound satipatthana
(smrtyupasthana). The first three terms are all from the same root, and have obviously been
derived from upasthana. Little can be inferred from the exact meanings of the various prefixes
used to distinguish these terms. However, it is obvious that what is being emphasized here is the
root meaning of 'standing, stability, steadiness', three English words also ultimately derived from
the same Indo-Aryan root stha. 'Setting', or 'establishing', or 'placing' the mind on the meditation
object, firmly, repeatedly, continuously, is the means leading to unification. The Theravada
Abhidhamma definition of samatha similarly includes a series of terms derived from stha, which
curiously enough even share the identical prefixes in the same order (thiti, santhiti, avatthiti). But
there the absence of upatthana from the list, and the use of a different derived form (thiti rather
than thana), obscures the connection with satipatthana.

AFTERWORD

I would like to finish by recapping a few important areas where the GIST and the history of
mindfulness intersect.

The first thing to notice is the value of seeing the Buddhist scriptures in their historical and
cultural context. In some cases this leads us to quite secure conclusions, which we could hardly
have reached relying on the Buddhist sources alone. A good example of this is the analysis of
ekayana, where both the Pali and the Chinese sources seem to be unsure of the meaning, but the
Upanisadic context is very revealing.

This case also reaffirms the necessity to examine the Samyutta as the prime source. From the
Satipatthana Sutta itself we would have no idea that the phrase ekayana had any Brahmanical
connection, but the Samyutta clearly suggests this.

224
Our studies have also emphasized the crucial importance of a holistic approach to study of the
Dhamma. We cannot treat individual limbs as if they have nothing to do with the greater
organism. The holistic paradigm must stem from the four noble truths, and the dhamma
categories that are most directly derived f r o m the four noble truths. Any serious study along such
lines will inevitably end up back at the Samyutta before too long, since this is where most of the
core doctrines are found. Thus we have treated satipatthana, not as t h e 'only way', but as the
seventh factor of the eightfold path. The Satipatthana Sutta is no more - and no less - than an
elaboration of this stage of the path.

While the doctrinal centrality of the Samyutta must force itself upon the student, we have also
made the more radical claim, following Yin Shun, that the Samyutta also constitutes the
historically oldest strata of Buddhist scriptures, and have even suggested that this collection was
current during the Buddha's lifetime, and was the prime work codified at the First Council.
Obviously any such claim will remain controversial, and we can merely point out that certain
evidences can be interpreted in that manner. I have said this already, but it bears repeating: this
does not mean that everything in t h e Samyutta is early and authentic and everything else is late
and inauthentic. It simply means that t h e samyutta material is likely, on average, to be early, and
the accretions are mainly restricted to editorial repetitions, etc., rather than doctrinal expansion.
Nevertheless, we have noted several instances where even the Samyutta seems to evidence very
slight sectarian influence.

This holistic approach also reaffirms t h e necessity for comparative studies. When we turn
attention away f r o m the more obvious sectarian differences and look more closely at the central
teachings, the similarities between the traditions are outstanding. We have mostly dealt with the
Pali and Chinese sources, since this is where most of the early scriptures are found. It should be
re-emphasized, however, that the Tibetan canon, though lacking the actual texts, is still based on
the Agama sutras, which are still regarded as canonical within that tradition. It is rather a shame
that those inspired by the Tibetan tradition remain largely unaware and unappreciative of the
historical sources f r o m which their teachings and practices ultimately stem.

We have used our appreciation of the central role of the Samyutta to reassess the Satipatthana
Sutta. It is obvious f r o m a cursory examination of the various versions that there is a real
problem to be resolved. While we can certainly go a long way with straight comparative study of
the existing versions, at several points the samyutta-connections give crucial assistance. For
example, in deciding whether the satipatthana auxilliary formula was standard in the
Sarvastivada as in the Theravada, the Satipatthana Suttas are useless, and the key is only
provided by a little footnote in the Chinese translation of the Samyutta. Again, it is the Samyutta
that provides confirmation of the original specification of contemplation of dhammas.

While stressing the importance of the pre-Buddhist environment, we have also acknowledged the
significance of considering the later sources. The pre-Buddhist sources tell us about the Buddha's
audience, his language, the kinds of issues and ideas he was addressing. But we cannot forget that
the fact that the early Suttas exists at all is due to the efforts of the schools; and the ideas and
agendas of the schools can hardly be expected to leave no imprint at all on the scriptures. In some
cases this is merely the echoes of t h e technologies and languages of the times; in other cases we
find significant doctrinal developments. We have noticed that it is precisely the later accretions
in the Satipatthana Sutta - specifically, the vipassana refrain and t h e addition of the aggreggates,
sense media, and truths to the contemplation of dhamma - that has become most emphasized in
later works, almost to t h e point of ignoring the authentic material. Nevertheless, we repeatedly
find, across the schools, that important aspects of satipatthana are remembered and explained
accurately. Thus we can neither agree with the traditionalists, who assert that the commentarial
tradition of their own particular school has got it all wrapped up, nor with t h e radical modernist

225
reformers, who opine that the traditions a r e j u s t a mass of error, better disposed of. Rather, we
find that a careful evaluation of the traditions in the light of the Suttas shows they have much to
teach us about the Suttas, and even when they are in error they show us how t h e living
communities adapted t h e Dhamma to their situation.

This social aspect of the Dhamma is worth reflecting on, for Dhamma never lives in a vacuum. I
have always been intrigued by this curious fact: differences in meditation techniques, divided
precisely along the lines of samatha versus vipassana, are among the most divisive issues in
Buddhism. We invest a lot in our meditation, a lot of time, a lot of effort, a lot of pain; and so we
attach, much more deeply than to mere theory. Differences in approach and emphasis to
meditation can harden into defensiveness as to who's got the right 'system', and t h e
interpretation of doctrine is then shaped to suit, with a strident insistence on the primacy of
one's own take on 'ultimate reality'.

This being so, it would seem that an approach to meditation that emphasized the essential
harmony and complementariness of samatha and vipassana would be a healing force in the
Buddhist community. This would allow us to appreciate t h e benefits of the various approaches to
meditation without insisting on any one of t h e m as absolute and sufficient for everyone. In this
we would be following in the tolerant footsteps of the Buddha, accepting whatever spiritual
practices are good and in line with the Dhamma, while avoiding dogmatic extremes.

'What, Bhante, does "one who abides in Dhamma" refer to?'

'Here, a monk studies t h e Dhamma - sutta, geyya, vyakarana, gatha, udana, itivuttakajataka,
abbhutadhamma, vedalla. He does not waste his days with that Dhamma he has studied, he does not
neglect retreat, he is devoted to tranquillity (samatha) of the heart within. Thus, monk, a monk is
"one who abides in Dhamma".

'Thus, monk, I have taught you the monk who studies a lot, the one who teaches a lot, the one
who thinks a lot, and the one who abides in Dhamma. I have done for you what should be done by
a Teacher seeking the welfare of his disciples out of compassion. Here, monk, are roots of trees,
here are empty huts. Practicejhana, monk! Do not be negligent! Do not regret it later! This is our
instruction to you.' 517

226
 Appendix 1
T H E SAMYUTTA-MATIKA

This table shows a variety of texts, f r o m the Nikayas/Agamas on, that make use of the samyutta-
matika. The texts represented are very varied, and so is the manner of utilizing the matika.
Sometimes, as in the Samyuttas, the samyutta-matika forms the fundamental structure of t h e
work as a whole. In other cases t h e topics are presented along with discussions of other matters,
and must be abstracted. Sometimes the samyutta-matika is simply mentioned as a list. I have
made no attempt to represent these important differences or to justify my particular choices,
which is sometimes determined by t h e limited secondary sources available to me. The table is
simply intended to represent in a graphic and immediate manner how pervasive the samyutta-
matika is throughout the schools.

THERAVADA
Samyutta Nikaya Vibhanga Dhammahadaya Dhatukatha Yamaka Visuddhimagga
Dependent origination Aggregates Aggregates Aggregates Aggregates Aggregates
Elements Sense media Sense media Sense media Sense media Sense media
Aggregates Elements Elements Elements Elements 18 elements
Sense media Truths Truths Truths Truths 22 faculties
Feelings 22 Faculties 22 Faculties 22 Faculties Activities 8 noble truths
Eightfold path Dependent origination Causes (hetu) Dependent origination Compulsions Dependent origination
(anusaya)
Enlightenment-factors Satipatthana 8 nutriments Satipatthana Mind Satipatthanas
Satipatthana Right efforts Contact Right efforts Dhammas Right efforts
Faculties Bases of psychic Feeling Bases of psychic 22 Faculties Bases of psychic
power power powers
Right efforts Enlightenment-factors Perception Jhana Spiritual faculties
Spiritual powers Eightfold path Volition Immeasurables Spiritual powers
Bases of Psychic Jhana Mind Spiritual faculties Enlightenment-factors
Power
Jhana Immeasurables Spiritual Powers 8-fold path
Anapanasati Training rules Enlightenment-factors
Stream entry Eightfold path
Truths Contact
Feeling
Perception
Volition
Mind
Decision (adhimokkha)
Attention

SARVASTIVADA DHARMAGUPTAKA

Samyukta Agama Dharmaskandha Dhatukaya (Chp 7) Sariputrabhidharma

Aggregates 5 precepts 5 precepts 12 sense media
Sense media 8 factors of stream-entry 8 confirmed faiths 18 elements
Dependent origination 8 confirmed faiths 8 fruits of asceticism 5 aggregates
Nutriments 8 fruits of asceticism 8 ways of practice 8 noble truths
Truths 8 ways of practice 8 noble lineages 22 faculties
Elements 8 noble lineages 8 right efforts 7 enlightenment-factors
Feeling 8 right efforts 8 bases of psychic power 3 unskilful roots
Satipatthana 8 bases of psychic power 8 satipatthanas 3 skilful roots
[Right efforts] 8 satipatthanas 8 noble truths 8 great elements
[Bases of psychic power] 8 noble truths 8jhanas 5 precepts
Faculties 8 jhanas 8 immeasurables Elements
Spiritual powers 8 immeasurables 8 formless attainments Action
Enlightenment-factors 8 formless attainments 8 developments of samadhi Persons
Eightfold path 8 developments of samadhi 7 enlightenment-factors Knowledge
Anapanasati 7 enlightenment-factors 22 faculties Dependent origination
Training (Ksudravastuka) 12 sense media 8 satipatthanas
Stream entry 22 faculties Aggregates 8 right efforts
12 sense media 62 elements 8 bases of psychic power
Aggregates 8jhanas
62 elements 8 fold path
Dependent origination Unskillful dhammas

227
 MAHAYANA (YOGACARA)
Sravakabhumi Heart Sutra Abh idharmasamuccaya Samdhinirmocana Arthaviniscaya Madhyantavibhagabha sya
Sutra Sutra
Aggregates Aggregates Aggregates Aggregates Aggregates Aggregates
Elements Sense media 18 elements Sense media Grasping- Sense media
aggregates
Sense media 18 elements Sense media Dependent 18 elements 18 elements
origination
Dependent Dependent 4 noble truths Nutriments Sense media Dependent origination
origination origination
Nutriments 4 noble truths 4 satipatthanas Truths Dependent Good _ bad dhammas
origination
Truths Knowledge 4 right efforts Elements 4 noble truths 22 faculties
Disciples, Pacceka Attainment/non- 4 bases of psychic Satipatthanas 22 faculties 3 times
Buddhas, Tathagatas attainment power
Satipatthana 5 spiritual faculties Right efforts 4jhanas 4 noble truths
Right efforts 5 spiritual powers Bases of psychic 4 formless Vehicles
power attainments
Bases of psychic 7 enlightenment- Spiritual faculties 4 divine abidings Conditioned/unconditioned
power factors
Spiritual faculties 8-fold path Spiritual powers 4 ways of practice Satipatthanas
Spiritual powers Enlightenment- 4developments of Right efforts
factors samadhi
Enlightenment- 8-fold path 4 satipatthanas Bases of psychic power
factors
Path 4 right efforts Spiritual faculties
Ugliness 4 bases of psychic Spiritual powers
power
Anapanasati 5 spiritual Enlightenment-factors
faculties
Training 5 spiritual powers 8-fold path
Confirmed faith 7 enlightenment-
facors
8-fold path
Anapanasati
Factors of stream-
entry
Tathagata-powers
4 intrepidities
4 discriminations
18 avenika
dhammas
32 marks of a
great man
80 minor marks

228
 Appendix 2

T H E IRONIC ASSUMPTIONS OF GREGORY SCHOPEN

The methods and assumptions of Buddhist studies such as the current work have in recent years
come under challenge, indeed frontal assault, by the influential academic Gregory Schopen. His
writings are deliberately provocative and sometimes brilliant. His basic approach in
understanding Indian Buddhism may be summed up as a change in method, leading to different
results.

In method, he criticizes the assumption of modern scholars that the study of Buddhism may be
equated with the study of its texts, and instead proposes that the archaeological evidence should
be granted priority. I think all would agree that he has a point here, but it is not obvious to me
that previous scholars have been so negligent in this regard. As just one random example,
Lamotte's discussion of King Milinda occupies about seven pages. 918 The first three pages mainly
survey the evidences of t h e coins and other material evidence, summed up as 'as few fragmentary
inscriptions'; the next three pages discuss the Milindapanha, an important work of the Middle
Period preserved in Chinese and Pali; and the final page mentions a few references in later works.
This seems reasonable to me; if anything I would have liked to see more discussion of some of the
philosophical points raised in the Milindapanha, whose stance tends to be intermediate between
the canonical doctrines and the developed positions of the schools.

As far as the results of research are concerned, Schopen says that the record of the bones and
stones depicts a very different type of Buddhist monastic, one who is more worldly and h u m a n
than the caricature of the ascetic hero striving for Nibbana alone in t h e forest. Since Schopen's
work constitutes the most influential and sustained critique of the kind of project undertaken in
this book, it is worth considering his claims in some detail. If we weather this storm, we'll be
ready for anything.

Many of Schopen's conclusions, I think, are obviously true. He is primarily interested in the
'Middle Period' of Indian Buddhism, that is, the five hundred years or so f r o m the beginning of
the Common Era. He uses the remnants of monasteries, stupas, graves, etc., together with Vinaya
material, primarily f r o m the Mulasarvastivada Vinaya in Tibetan (he makes little use of t h e
Chinese sources), which he says stems f r o m the same period and depicts much the same activity.
These sources speak to us of monks and nuns who accumulate wealth, make substantial donations
f r o m their own wealth for building projects, promote devotional activity such as worship of
stupas, images, and relics, are engaged in business transactions, contracts, and lending on
interest, and are frequently at the beck and call of the lay followers for performance of rituals
such as weddings, house blessing, and so on. All of this picture is quite convincing and needs little
discussion here.

But while it is obviously true, I would also contend that it is truly obvious. All the activities that
Schopen depicts may be plainly seen in the activities of the majority of the ordained Sangha in all
traditions in the present day. Schopen merely points out that these conditions also obtained in
the Middle Period of Indian Buddhism as well. While this may come as a surprise to academics
with little contact with Buddhism in the real world, and constitutes an important critique of the
fallacy of equating Buddhism with the idealized portrait in the sacred texts, it will come as no
surprise for those of us who encounter Buddhism in the world every day.

229
Another of Schopen's arguments that is well taken is that the average monk or nun, not to speak
of the lay followers, may hardly even know of the scriptural texts. The scriptures may have only
been known to a small elite of scholars, and the ideas therein might not be representative of the
range of Buddhists. A few years ago I was staying in a forest h u t belonging to a devoted,
intelligent Thai Buddhist, who, when he was young, had been in robes for two and a half years.
Once I visited a local monastery and borrowed copies of some of t h e Suttas. When I mentioned it
to my friend, h e looked absolutely blank: he had never even heard the words 'Majjhima Nikaya' or
'DTgha Nikaya'. Again it seems plausible that this situation, observable today, could have obtained
two thousand years ago in India. But the argument should not be overstated. The Buddhist
scriptures are big works. They must have required a substantial organization of monk-&-nun
power to maintain, whether in oral form or even in the later written form, and so a large number
of people must have known them. The number of inscriptions f r o m ancient India is only a few
thousand, and so can only represent a tiny fraction of scraps of ideas of all the Indian Buddhists.
And those who are wealthy enough to donate religious monuments are hardly likely to be
representative of the full spectrum of the Buddhist community. Anyway, as Schopen emphasizes,
many of the donors are monks and nuns (according to Schopen, most of the donors are monastic,
and in the Middle period, about half the monastics are nuns) who state that they are versed in t h e
'Suttas' or 'Vinaya' or 'Tripitaka' or 'Nikayas'; in other words, they are the same people as those
who passed down the scriptures.

Schopen is scathing in his assessment of the 'assumptions' made by various Buddhist scholars. He
characterizes the work of early, Victorian, scholars such as Oldenberg and Rhys Davids as
'protestant', and suggests that they have read their own biases into the Buddhist texts, depicting
the Buddha and his Sangha much like rational, cultured European gentlemen. This, too, is true,
but it is hardly a valid criticism. Anyone familiar with Buddhist thought should accept that our
understanding is always coloured by our beliefs and values. Fine, let's point this out - but let's not
assume that we are an exception. I am a forest monk, and I believe that the Buddha and his early
generations of ordained disciples were also forest monks and nuns. So when I look at the heritage
of Buddhism, I naturally focus on this aspect. Gregory Schopen is a highly paid academic f r o m an
overwhelmingly materialistic society, and so when he looks at the heritage of Buddhism he sees
money, rocks, and material remains. When he does look at the texts - as any scholar, whatever
their beliefs, must eventually do, for the information contained in the inscriptions is scanty - he
focuses on the Vinayas, since they deal most directly with the material aspects of monastic life -
buildings, etc. But the Vinayas themselves represent a movement f r o m the spiritual to the
material - they are about what monks and nuns do when they misbehave, and so taken by
themselves they are misleading. We would not expect to gain an accurate vision of how an
ordinary person leads their daily life today by reading law books. Schopen contrasts the wealthy,
developed monasteries with the poor, simple villages nearby. His agenda seems to be, in t h e
broadest sense, Marxist. I do not mean that in t h e slightest pejorative sense - I think it's sweet
that he dedicates his books to t h e 'working men and women' whose 'labor paid for my scholarly
leisure'. But he seems to have little interest in the spiritual aspect of Buddhism, which puts him in
a minority of those, at any time, who wish to learn the Dhamma. It should be obvious that
Schopen's assumptions influence his conclusions, j u s t as the assumptions of earlier scholars
influence their conclusions. Wholesome states of mind leave no scar on the rocks. Meditation
attainments are airily ephemeral. Insights into reality happen in the wispy world of the mind. If
we were to accept Schopen's methods unconditionally, we would have to abandon the very
reason that most of us became interested in Buddhism. There would be no more reason to study
ancient India than any other ancient culture. This may not be a problem for Schopen, but it is a
big one for most students of Buddhism. My primary interest is in spiritual practice, and my
interest in the Agama Suttas stems f r o m this: they describe a spiritual practice that I find
inspiring, practical, and profound. I have tried, to my limited best, to live up to the ideals taught
in that literature, and have invariably found that, when problems arise, they are due to my own

230
inadequacies, not those of the teachings. I have also had close contact with a number of h u m a n
beings whose inner radiance testified to t h e power of the Dhamma when lived to its fullest. Since
this tradition that I belong to claims to stem f r o m a genuine historical individual called the
Buddha, it seems reasonable to see what t r u t h there might be to this claim.

Schopen's work contains much that is interesting and informative, but little that could be called
inspiring. His writing is characterized by wit, scandal, and good yarns. Unfortunately, it is not
always characterized by consistency, and we should examine some of his fracture lines. He rests
his arguments heavily on the authority of the Mulasarvastivada Vinaya, a text he cheerfully
admits to not having fully read. This Vinaya is 'monstrous' in size, perhaps 4000 folios in the
Tibetan, and most scholars have taken it to be late, perhaps 500 C.E. Schopen would like to see
this Vinaya dated earlier, around the beginning of the Common Era. On the other hand, the
Theravada Vinaya has been taken by most scholars to be early, but Schopen would also like to
date that around the beginning of t h e Common Era. Thus the battle-lines are drawn. Schopen says
that the discussion of the date of the Mulasarvastivada Vinaya has been:

'badly misdirected by a very red herring and the inattention of those who are supposed to be
following the trail. In 1958 the great Belgian scholar Etienne Lamotte declared that this Vinaya, or
code, was late, that "one cannot attribute to this work a date earlier than the 4 th - 5 th Centuries of
the Christian Era." This pronouncement - even at its inception based on very shaky grounds -
still proved almost fatal, for Lamotte was forced by his own f u r t h e r work to change his position -
and he did so several times - but few scholars seem to have noticed. By 1966, Lamotte was in fact
referring to the Mulasarvastivada Vinaya as a source of information for t h e first or second
century of our era. Ironically, other scholars then, and for a long time after, continued to quote
only the Lamotte of 1958.'519

I must also confess inattention, for I have not followed the trail of Lamotte's arguments and so
must declare my incompetence to pronounce on the date of the Mulasarvastivada Vinaya. It
might be noticed in passing, though, that the two positions ascribed to Lamotte in this passage
are not necessarily contradictory. Given the evidently long period it would take to compile a vast
compendium like the Mulasarvastivada Vinaya, it would not seem unreasonable to maintain that
the final redaction was in the 4 th - 5 th Centuries C.E., but that it contains material inherited f r o m a
much earlier time. In fact, I believe that something of this sort could be said for almost all
Buddhist literature. This is a phenomenon known as 'intratextuality', t h e ongoing life of a given
text through a particular stream of tradition, which reflects t h e conservative nature of religious
literature: the redactors valued ancient authority over creative expression and thus tended to
work with material already to hand rather than inventing new material. 520 In any case, I do not
see what is 'ironical' in the failure of some writers to notice Lamotte's change of views: if scholars
continue to quote f r o m earlier, discredited theories this is merely an unfortunate mistake, not an
irony.

An example of true irony could be better seen f r o m Schopen's own work. In the same book as the
above quote, he says this:

'...this literature, the Mulasarvastivada Vinaya, is itself considered by many to be late - Lamotte,
for example, thinks it is the latest of the vinayas and says "we cannot attribute to this work a date
earlier than the fourth-fifth centuries of the Christian Era" .'521

Note that here Schopen says that Lamotte 'thinks' (present tense), thus precluding any later
change of mind. This clanger needs little comment, apart f r o m reminding us that Schopen, like
the rest of us, is sometimes guilty of seeing what he wants to see.

231
While I am not competent to date the Mulasarvastivada Vinaya, and have no opinion on the
matter, I must say that the passages quoted by Schopen himself frequently give me the
impression of lateness. The elaborateness of t h e text may be partly explained, as Schopen argues,
by cultural or other factors rather than by date, but the examples he gives fall well short of
establishing this in full. As for specifics, we notice that the Mulasarvastivada Vinaya frequently
mentions books and writing, while the Theravada Vinaya mentions them rarely. This was one of
the classic reasons the early European Buddhist scholars concluded (not 'assumed') the
Theravada was earlier, and as far as I can see the argument still holds good. Similar
considerations apply when we see that the Mulasarvastivada Vinaya refers to worship of Shiva
and Vishnu, while, as is well known, these deities are virtually unknown in the Theravada canon.
Schopen also argues that the Mulasarvastivada Vinaya evidences the influence of the Hindu
Dharmasastras (legal codes), while the Theravada does not. He says that this may be explained by
the lack of influence of the Dharmasastras in Sri Lanka, and is therefore evidence that t h e
Theravada Vinaya was composed in Sri Lanka. While I agree, for other reasons, that the
Theravada Vinaya shows some minor Sri Lankan influence, I don't think this particular argument
is very convincing. The Dharmasastras themselves evidently date from well after the Buddha's
time, and the situation might as well or better be explained by the simple hypothesis that most of
the material in the Pali was composed in India before the Dharmasastras became influential, and,
because of the unimportance of t h e Dharmasastras in Sri Lankan culture, the Theravada Vinaya
did not have to be extensively revised.

Another target of Schopen's critique is the vagueness or ambiguity of some Vinaya rules, which
he suggests may have been deliberate. 922 It seems that the poor old V i n a y a j u s t can't win: if it is
definitive, it is rigid, and if it is flexible it is decadent. Again we might compare this with one of
Schopen's own little 'ironies':

'In most cases, we can place the Vinayas we have securely in time: the Sarvastivada Vinaya that
we know was translated was translated into Chinese at the beginning of the fifth century (404-409
C.E.). So were the Vinayas of the Dharmaguptakas (408), the MahTsasakas (423-424), and the
Mahasanghikas (416). The Mulasarvastivada Vinaya was translated into both Chinese and Tibetan
still later, and the actual contents of the Pali Vinaya are only knowable f r o m Buddhaghosa's fifth
century commentaries.' 923

Does this remarkable assertion assume that the date of a text may be determined by knowing t h e
date of its translation or commentary? That would certainly solve a lot of problems: I have beside
me a translation of the Samyutta Nikaya dated 2000 C.E., so we can place that 'securely in time'. Of
course, the phrase is so vague - deliberately? - that Schopen escapes actually asserting that the
dates of composition of the Vinayas may be determined from their translation or commentary. If
that was t h e case, however, we would have to conclude, contrary to Schopen's position, that the
Mulasarvastivada Vinaya was later than the others, for its translation was later. Regarding the
Theravada Vinaya, it has been accepted, so far as I know, by all t h e scholars who have looked into
the matter that Buddhaghosa was primarily a translator and editor, who worked with material
stemming from a much older time, no later than 100 - 200 C.E. If the commentarial material dates
f r o m then, t h e Vinaya itself must be considerably earlier. Recently it has come to light that the
Chinese canon contains a Sri Lankan Vinaya commentary that Buddhaghosa may have had before
him. If so, this would allow a much more accurate assessment of the kinds of changes he
introduced.

An important part of Schopen's argument is that there is little or no early - pre-Common Era -
evidence for Buddhist monasteries of the developed sort that are depicted in the Vinayas. This is,
for him, a sign that the Vinayas were compiled in the 'Middle Period'. He notes that the words
vihara and avasa, which are commonly used of monasteries, really mean little more than

232
'dwelling', and give us little information about what kind of institution is being discussed. 948
However he neglects to notice that t h e main terms used of a monastery in the Pali Suttas are vana
(woodland grove) and arama (park); the fact that they are used together in the n a m e of the most
famous monastery of all ('Jeta's Grove, Anathapindika's Park') suggests that they may be
synonyms. These, of course, have a much more specific meaning - evidently the main form of
Buddhist monasticism in the Suttas was the forest monastery. Even today, the typical forest
monastery consists of small huts or caves scattered through the forest, with a larger wooden sala
for communal activities, and some buildings for stores, kitchen, etc. Such an institution would
leave little or no evidence for an archaeologist to uncover. Schopen does not seem to consider t h e
possibility of a 'middle way' between the large, institutionalized viharas that are such a feature of
the archaeological record of Buddhism, and the life of the lonely sage in t h e forest. It would seem
that the forest monastery offers such a 'middle way'. Forest monasteries can evolve to a high
degree of sophistication in their internal organization, such as is described in the Vinayas, and
usually have a high regard for authentic practice of the Vinaya. They often do not engage in large
building projects, not because they do not have the resources or the know-how - forest monks
are often more educated and better supported than the city monks - but because they want to
live simply. This is just a suggestion, and more careful work on the Vinayas - including the
Chinese - has to be done to see if this suggestion has any cogence. It is obviously tenuous to draw
such parallels between Buddhist practice in such far-distant times and places. But Schopen
himself draws many instructive parallels between practice in Buddhist and Christian
monasticism, which would seem to be no less distant. And as I have noted above, many of
Schopen's more acceptable findings do find clear parallels in contemporary Buddhism.

Schopen dismisses the 'perishable materials' argument for the lack of early monasteries, saying
that the earliest archaeological evidence we do possess shows us a monastery in the time of Asoka
that is 'poor and unimpressive', 'crudely made of "rubble".' 929 He asserts that: 'the earliest extant
remains of monastic residential architecture, like the earliest cult images in stone, show a
tradition still struggling, in this case towards order, still lacking a sense of functional
organization and structured use of space. Such a tradition - again like that which produced the
early extant cult images - does not suggest a long period of development or directed
experimentation in wood or other perishable materials preceding it.'927 But this argument would
also seem to be circumvented by the forest monastery hypothesis - when living in widely
scattered dwellings in t h e forest it is not necessary to develop such a structured sense of space.
What seems to be happening here is that t h e monastics are, for t h e first time, living in close
proximity. This might be due to a number of factors - perhaps there were too many Buddhist
monastics in that period. But some of the early sites mentioned by Schopen also share another
significant feature: the monastic dwellings are near a stupa. This might suggest that these are the
first monasteries for whom the devotional practices described by Schopen are becoming
important.

What is perhaps more relevant for our current purposes, however, is that this argument exposes
yet another of Schopen's 'ironies'. He assumes that the emergence of sophisticated architecture
or fine arts requires a substantial prior period of development - a most reasonable assumption.
But is not the same the case in literature? Schopen wants to put very sophisticated literary tracts
like the Mulasarvastivada Vinaya in the early Middle Period. But surely such works must have
required a lengthy evolution. Similarly, we know for certain (from the dates recorded for the
Chinese translations) that the earliest Mahayana Sutras date f r o m no later than the beginning of
the Common Era. These too are sophisticated literary and philosophical products, which are, to a
large degree, a critical response to some aspects of the early schools, especially t h e (Sarvastivada)
Abhidhamma philosophy, and also to such monastic practices as are detailed in the
Mulasarvastivada Vinaya, as Schopen himself argues. 927 The Abhidhamma texts themselves are
sophisticated literary works that are in turn based on the material found in the early Suttas. So

233
the early Sutta material - not necessarily the exact collections in the f o r m we have t h e m today,
but the main doctrinal material - must be several philosophical generations before the Mahayana
Sutras. Again, this conclusion, not 'assumption', was one of the classical reasons for assigning a
relatively early date to the Nikayas/Agamas, and nothing Schopen says really affects this.

Schopen tries to show that the forest monastic life was little different f r o m settled monastic life
in general. He does this by quoting a passage f r o m the Vinaya that describes the lovely, luxurious
forest dwelling of a certain Venerable Udayin, where many people would go to visit him. Schopen
says that this is apparently how the compilers of the Pali Vinaya saw the forest life.528 Incredibly,
he makes no mention of the fact, known to every Grade 1 Vinaya student, that Udayin is t h e
archetypical 'bad monk', whose appalling behaviour prompted the formulation of many Vinaya
rules. On this occasion, Udayin gropes and sexually harasses a woman who comes to visit him,
prompting the laying down of yet another rule on his behalf. This part of the story, however, is
discreetly omitted by Schopen as he tries to depict Udayin as a regular forest monk.

While it is obvious that the cult of relics and so on played a large part in Buddhist practice f r o m
the Middle Period, Schopen wants to discredit the received opinion that the early texts, and
hence early Buddhism, do not include t h e relic cult. He ends up clutching at some embarrassingly
flimsy straws. For example, he points to a passage in the Satipatthana-samyutta where the novice
Cunda, after the passing away of Venerable Sariputta, takes his bowl and robes and goes to tell
Venerable Ananda.52? Schopen says that the PTS edition (which I do not have) has a variant
reading f r o m a Burmese edition that includes the phrase dhatuparibhavana.530 Schopen admits that
the meaning is obscure, but it 'almost certainly contains a reference to relics'. This is dubious, for
dhatu rarely if ever means 'relic' in this strata of literature. The VRI CD that I am using does not
have dhatuparibhavana, so it seems that this reading does not represent t h e mainstream Burmese
tradition. Thus far Schopen's argument is flimsy, but not necessarily wrong. But then he goes on
to say that the commentary appears to have a reference to relics, since it includes t h e t e r m
dhatuparissavana. Parissavana means 'water strainer', and dhatu might, in the commentary, mean
'relics', though the compound 'relics-&-water strainer' does seem a little odd. (Might dhatu here
mean 'minerals': a strainer for removing sediment f r o m water?) Anyway, the matter is clarified
by the very next sentence of the commentary, which is ignored by Schopen. This says: 'But in t h e
text (palT) i t j u s t says "Here are his bowl and robes".' In other words, t h e commentary explicitly
states that the original text did not mention anything other than the bowl and robes. Thus it
seems almost certain that paribhavana was not in the original text; it was probably read back into
the text by garbling the commentary (by a monk whose reading rivals Schopen in carelessness).
Schopen does not refer to the Chinese parallel, which is very close to the Pali, and which similarly
mentions just the bowl and robes. He says t h a t the Mulasarvastivada Vinaya version of the
incident does refer to relics, although he admits that the phrase is not a cognate of either of the
Pali terns with dhatu in them. This makes it seem like an independent later development, not a
common inheritance. Schopen is right on t h e mark when he says that 'this will require f u r t h e r
study to sort out'. I hope it has now been sorted out. Rather than being 'virtually certain' that t h e
Pali here has suffered loss - or as Schopen insinuates, deliberate suppression - I would say it is
absolutely certain the Pali and the Chinese and the Theravada commentary all agree that t h e
original account of Sariputta's death does not mention relics. Much later the Mulasarvastivada
Vinaya and perhaps the Pali commentaries added the mention of relics. Thus this context, as with
many others, suggests that the Mulasarvastivadin Vinaya has more in common with the
Theravadin commentaries than with the canon.

Schopen's work offers us f u r t h e r lessons in 'irony' in the discussion of the term paribhavita.531 He
shows that several inscriptions and late textual sources describe the relics of the Buddha as being
'infused' or 'permeated' (paribhavita) with such qualities as ethics, samadhi, understanding, and
release. This suggests a quasi-magical conception of relics in this period. Schopen discusses the

234
term in some detail and offers several references f r o m the Pali canon showing a naturalistic usage
of the term, for example a chicken sitting on eggs and 'imbuing' t h e m with warmth. But,
incredibly, he avoids all mention of the most well known occurrence of the term: the frequently
repeated statement of t h e Buddha in the Maha Parinibbana Sutta:

'Samadhi imbued with ethics is of great fruit, great benefit; understanding imbued with samadhi
is of great fruit, great benefit; the mind imbued with understanding is rightly released f r o m
defilements.' 532

Not only does the term paribhavita appear repeatedly, but it does so specifically describing a list of
dhammas similar or identical with those repeatedly mentioned in the inscriptions quoted by
Schopen.

The implications of this are slightly worrying. Schopen has built a successful career largely on his
pioneering research into the nature of the cults of the stupa and relics in Indian Buddhism. The
prime canonical reference for these practices is the Maha Parinibbana Sutta, which describes the
distribution of the Buddha's relics. One of the most famous and prominent passages in this text
repeatedly uses the t e r m paribhavita in connection with ethics, samadhi, understanding, and
release. Schopen discusses at length the use of paribhavita in inscriptions to describe relics that
are imbued with ethics, samadhi, understanding, and release. He gives several references to
unrelated uses of t h e term in the Pali canon, but avoids all mention of the usage in the Maha
Parinibbana Sutta.

What is going on? Has Schopen not even read the Maha Parinibbana Sutta, the main source text
in his own special field? Or might we conspiratorially wonder whether Schopen has deliberately
suppressed the Maha Parinibbana Sutta reference (just as Schopen alleges the redactors of the
Pali canon suppressed mention of relics and stupas)?

Once the obvious connection with the Maha Parinibbana Sutta is noticed, it becomes impossible
to escape t h e conclusion that t h e inscriptions are, in fact, quoting f r o m or referring to this
specific text. Note that the passage on ethics, samadhi, understanding, and release in itself has no
connection with the relic cult. If it existed as an isolated fragment or in another context there
would be no reason to associate this passage with relics. Only when taken as part of the overall
narrative of the Maha Parinibbana Sutta would it be possible to form an association between the
passage and the Buddha's relics.

To be sure, the implications of the usage in the inscriptions is radically different f r o m that in the
Maha Parinibbana Sutta. In the discourse it describes spiritual qualities to be developed by a
living person, whereas in the inscriptions it seems to mean the magical infusion of relics with
mystic power. This obviously suggests that t h e earlier, rational, psychological teaching has been
altered - dare I say 'corrupted'? - by magical conceptions. This is a straightforward reading f r o m
the evidence, not an imposition of 'protestant presuppositions'. Of course, this conclusion would
be impalatable to Schopen, because it would suggest, firstly, that the discourses, or at least the
Maha Parinibbana Sutta, were actually known to a variety of Indian Buddhists and influenced
their beliefs; and secondly that the picture he paints of the Middle Period is representative of
Buddhism in its decadent, materialistic phase, rather than t h e psychological spirituality of the
early teachings.

It is worth noting that Schopen's key inscriptional and textual sources for this quasi-magical use
of paribhavita are dated to around t h e first century of the Common Era. By this time, the Maha
Parinibbana Sutta must have been composed, and already be well-known and influential. This
must have happened long enough for some of the central messages to be radically reinterpreted,

235
and for these reinterpretations to have gained wide currency. The Maha Parinibbana Sutta
evidences later elaboration, and, despite t h e fact that several sectarian versions are known, most
scholars do not place it among the earliest strata of the Suttas. So if t h e Maha Parinibbana Sutta
was in existence significantly before the Middle Period, many other discourses must be even
earlier. So we must be grateful to Schopen for, yet again, inadvertently offering us another proof
of the existence of the early Suttas well before the Middle Period.

Schopen's failure to notice this obvious inference stems from his wilful enslavement to his own
methodological presuppositions. He has a religious faith in 'hard facts', things that 'actually' exist
in stone and bone. As normal, when a particular means of knowledge is given absolute priority in
this way, it leads to philosophical distortions and a blindness to the broader perspective. Schopen
castigates those who would render archaeological evidence subject to texts, since archaeological
evidence can be located in place and time, and represents what was said by 'actual' people (as if
those who wrote the texts were not 'actual' people). One of his pervasive unexamined
assumptions is the reliability of archaeological evidence. I am no expert, but it does seem to m e
that archaeologists, like those in any field of science, are engaged in pushing back the frontiers of
knowledge, and to do so must rely on sometimes tenuous inferences. Schopen remarks several
times that the sites he is referring to have not been fully excavated, or were poorly reported, or
that there is uncertainty as to dating. It does not seem obvious that the inferences derived f r o m
such methods are necessarily more reliable than those derived f r o m textual sources.

Just one example will suffice here. Schopen quotes an inscription that refers to t h e setting up of
an image of the 'Blessed Lord, the Buddha amitabha' (bhagavato buddha amitabhasy a).933 He says
that this is t h e only inscriptional reference to Amitabha in India, and constitutes one of the few
'hard facts' we know about his cult in India. The inscription is certainly interesting, and it is
useful that Schopen brought it to light. But what does it mean? The inscription says an image was
set up by a certain Nagaraksita or Samraksita, who wishes that 'by this skilful root may all beings
attain unexcelled knowledge'. Such references to 'all beings' and 'unexcelled knowledge' are
typical of Mahayanist inscriptions; but the present inscription is very early, apparently 200 years
prior to the widespread appearance of Mahayanist inscriptions. Schopen assumes that amitabha
refers to the Buddha of that n a m e in the well-known Sutras so popular in China. Thus, as usual,
he is unable to say anything meaningful about the inscription without the context provided by
the texts. In this case, his interpretation is reasonable, but is not necessarily true. 'Amitabha'
means 'infinite light', and is virtually identical with a word used in the Pali tradition to describe
an order of deities: appamanabha deva, the 'deities of measureless light'. It is possible that amitabha
was used of certain deities, and f r o m there became an epithet of the historical Buddha, and only
later the human and divine elements were fused into 'Amitabha Buddha'. In other words, the
inscription might not be a reference to 'the' Amitabha, but might simply be a descriptive epithet
of Sakyamuni, representing a stage in the development towards Mahayana ideas. I am not
arguing that this is in fact the case, but am merely pointing out that, in the absence of context, it
is impossible to know which interpretation is correct. Crucially, any meaningful statement on the
matter must be based on an inference, on what we think is the more reasonable interpretation,
not on the 'hard facts'.

I might beg leave here to give an example from my own experience. Once I was staying at a forest
monastery where the practice was to inter the cremated remains of the monastery supporters in
the monastery wall. A hole was made in the wall, and with a simple ceremony, the ashes were
placed in and covered with a brass plaque. Someone, perhaps an archaeologist of Schopenesque
bent, might come at some time later and notice a peculiar feature of the plaques. In a certain
section, that closest to the entrance and dated earliest, the plaques say 'Rest in Peace', a typically
Christian saying. The later plaques, however, say 'May she attain Nibbana', which is obviously
Buddhist. What is going on? Did the monastery change from Christian to Buddhist? Is this

236
evidence of an obscure sect of antipodean 'Buddho-Christians'? Might we suspect darkling
intrigue, a hidden tussle for power between two opposed groups of monks, vying for the funds
f r o m the different religious communities? Happily, I was there at the time, and can answer 'none
of the above'. These plaques were ordered f r o m a shop whose normal business, this being in a
predominately Christian country, was to make plaques for Christian burials. So they came with a
typically Christian burial slogan. The monks simply didn't give the matter any thought, until it
was pointed out that a Buddhist saying would be more appropriate, and so one was invented.
That's all there was to it. Incidentally, we did not really believe that saying 'May she attain
Nibbana' on the burial plaque would really help the lady concerned to attain Nibbana; it just
seemed like a nice sentiment. I might compare this concrete, dateable, placeable, 'actual'
evidence with, say, some of my own essays that are available on the Net. They have no date, no
place, no concrete existence at all. Yet I regard them as a more reliable and accurate guide to my
beliefs and practices than those messages on the plaques at the monastery where I stayed.

Schopen dismisses t h e idea that shared passages in a text are evidence of early, pre-sectarian
material. He prefers the hypothesis that shared material is evidence for later sharing, levelling
and standardizing of material. Thus he apparently believes that when the Buddhist monastics
lived in close proximity in the Ganges valley, speaking a common language, and regarding each
other as being all of one community, they developed different diverging scriptures, but when
they were spread widely over 'greater India', speaking different languages, and regarding each
other as belonging to different communities, they 'levelled' and 'standardized' their scriptures.
This does not strike me as inherently plausible, or even vaguely rational. He attempts t o j u s t i f y
this with reference to what is known in Christian history; but the Bible is accepted with slight
variations as canonical by all Christians, whereas the writings of later theologians and teachers
are accepted only by certain denominations and are rejected by others.

It is as if we were to come across people living in two neighbouring villages, each speaking a
slightly different dialect, with customs, beliefs, lifestyle, and physical appearance that were
similar, and a shared myth that asserted that they sprang f r o m the same origins. Schopen would
point out that there is no 'hard evidence' that they 'actually' share a common ancestry. The
'actual' situation is that there are two different villages, with divergent languages, beliefs and so
on. Any 'assumption' that the observable similarities derive f r o m a common ancestry is sheer
speculation. After all, there is plenty of evidence that cultures tend to homogenize, to move away
f r o m diversity towards similarity. The only reasonable explanation would seem to be that here
we have two different peoples, and the similarities in their cultures and physical appearance is
evidence of cultural interchange and intermarriage between two originally disparate
communities. This description might sound like a caricature of Schopen's ideas, but I honestly
believe it is not.

One of Schopen's main arguments in favour of his 'later borrowing' thesis is the story of the stupa
for Kassapa Buddha at Toyika. Wynne has shown that this argument is deeply flawed. Schopen
compares various versions of the same story, but conveniently confines to a footnote t h e fact
that, while the other versions occur in the Vinayas, t h e Theravada version is found in the
Dhammapada commentary. This turns out to be yet another piece of evidence that the Theravada
tended to close their canon early, placing later material in their commentaries. Not only is this a
fatal error in one of Schopen's key arguments, but it is, as Wynne points out, a misrepresentation
of the methods of the 'higher criticism' that Schopen is so dismissive of. Normally scholars will
take the congruence of the canonical, not the commentarial, literature as evidence of pre-
sectarian remnants. But this is not the only place that Schopen misrepresents his opponents. He
asserts, for example, that the 'cardinal tenet of this criticism states, in effect, that if all known
sectarian versions of a text or passage agree, that text or passage must be very old; that is, it must
come f r o m a presectarian stage of the tradition.' 538 The repeated use of 'must' is highly

237
misleading. The sharing of material is only one of many independent criteria that are regularly
employed to support and check each other. I do not know of any scholar who would make the
blanket assertion that shared material 'must' be earlier. It is no more than a reasonable
hypothesis that forms a basis for further research.

Compared with the situation in Bible studies, the quantity of Buddhist literature is so vast, the
subject matter so obscure, and the amount of serious research so small, that it is premature to
discard any methodology. While the early scholars may not have given due weight to the
archaeological evidence, they must be forgiven, in consideration of the sheer time and effort it
takes to learn the Buddhist languages and read the texts. They have at least given us a reasonably
coherent and satisfying working model of Indian Buddhism. If we were to accept Schopen in his
more radical moods we would be rendered incapable of saying anything about the Buddha or his
teachings, and would be left with no idea as to why there were, in the later periods, such widely
spread religious schools claiming inspiration from a common Teacher, sharing a similar lifestyle,
and borrowing wholesale each other's scriptures, at the same time as vigorously arguing with
each other over what the scriptures mean.

BIBLIOGRAPHY

Oliver ABEYNAYAKE, 'Sri Lanka's Contribution to the Development of the Pali Canon', in Buddhism
for the New Millennium, London: World Buddhist Foundation, 2000. Available online.

Chizen AKANUMA, The Comparative Catalogue of Chinese Agamas and Pali Nikayas, Sri Satguru
Publications, 1990.

Mark ALLON, The Maha Parinirvana Sutra, Honours sub-thesis, Australian National University 1987
(translation of Waldschmidt 1950,1951).

Stefan ANACKER, Seven Works of Vasubandhu, Motilal Barnasidass, 1984, etc., corrected edition 1998.

ANALAYO, Satipatthana: The Direct Path to Realization, Buddhist Publication Society, 2003.

M. ANESAKI, 'The Four Buddhist Agamas in Chinese: A Concordance of their Parts and the
Corresponding Counterparts in the Pali Nikayas', Transactions of the Asiatic Society of Japan, XXXV
(1908), pp. 69-138.

Asanga, Sravakabhumi, (Sanskrit text edited by the Sravakabhumi Study Group, The Institute for
Comprehensive Studies of Buddhism, Taisho University), The Sankibo Press, Tokyo 1998 (Taisho
Books)

Samuel BEAL, Buddhist Literature in China, (second edition) Sri Satguru Publications, 1988.
The Romantic Legend ofSakya Buddha, first edition Trubner & Co., 1875, reprinted Motilal
Barnasidass, 1985.

Bhikkhu BODHI (trans.), The Connected Discourses of the Buddha - A New Translation of the Samyutta
Nikaya, Wisdom Publications, 2000.

Sara B O I N - W E B B (English trans. from Walpola Rahula's French trans. of the Sanskrit),
Abhidharmasamuccaya: The Compendium of the Higher Teaching (Philosophy) by Asanga, Asian
Humanities Press, 2001.
238
Johannes BRONKHORST, 'Dharma and Abhidharma', Bulletin of the School of Oriental and African Studies
48,1985.
"he Two Traditions of Meditation in Ancient India, Motilal Barnasidass, 2000.

Roderick S. BUCKNELL, The Structure of t h e Sagatha Vagga of t h e Samyutta Nikaya, unpublished

essay.

David M. CARR, Reading the Fractures of Genesis: Historical and Literary Approaches, Westminster John
Knox Press, 1996.

Garma C. C. CHANG (ed.), A Treasury of Mahayana Sutras: Selections from the Maharatnakuta,
(translated f r o m t h e Chinese by the Buddhist Association of t h e United States), Motilal
Barnasidass, 1991.

CHOONG Mun-keat, The Fundamental Teachings of Early Buddhism, Harrassowitz Verlag Wiesbaden,
2000.

Thomas CLEARY (trans.) The Flower Ornament Scripture (Avatamsaka Sutra), Shambala, 1986, 1993,
etc.

Margaret CONE, A Dictionary of Pali, Part 1, Pali Text Society, 2001.

Edward CONZE, The Large Sutra of Perfect Wisdom (Prajnaparamita Sutra), Motilal Barnasidass, 1990.

L. S. COUSINS, 'The "Five Points" and t h e Origins of t h e Buddhist Schools', included in Pollamure
Sorata Thera, etc. (ed.), Buddhist Essays: A Miscellany, Sri Saddhatissa International Buddhist
Centre, 1992.
'The Origin of Insight Meditation', included in Tadeusz Skorupski (ed.), The Buddhist Forum
IV, School of Oriental & African Studies, University of London, 1996.
'Pali Oral Literature', included in Philip Denwood and Alexander Piatigorsky (eds.), Buddhist
Studies, Ancient and Modern, Curzon Press, 1983.
'Samatha-yana and Vipassana-yana', included in Dhammapala, Gombrich, Norman (eds.),
Buddhist Studies in Honor ofHammalava Saddhatissa, Buddhist Research Library Trust, 1984.

Edward Fitzpatrick CRANGLE, The Origin and Development of Early Indian Contemplative Practices,
Harrassowitz Verlag Wiesbaden, 1994.

Har DAYAL, The Bodhisattva Doctrine in Buddhist Sanskrit Literature (1932), Motilal Barnasidass, 1978.

Siglinde DIETZ, Fragmente des Dharmaskandha, Gottingen Vandenhoeck & Ruprecht, 1984.

Paul DUNDAS, The Jains, Routledge, 1992, 2001.

Nalinashka DUTT, Buddhist Sects in India, Motilal Banarsidass, 1978, reprinted 1998.

Fumio ENOMOTO, ' "Mulasarvastivadin" and "Sarvastivadin" ', included in Chojnacki, Hartmann,
and Tschannerl, Vivadharatnakarandaka, Swisttal-Odendorf, 2000.
'On t h e Formation of t h e Original Texts of t h e Chinese Agamas', included in Buddhist Studies
Review, Vol. 3, no. 1,1986.

Erich FRAUWALLNER, The Earliest Vinaya and the Beginnings of Buddhist Literature, 1956.

239
 Studies in Abhidharma Literature and the Origins of Buddhist Philosophical Systems, State
University of New York Press, 1999.

Rupert GETHIN, The Buddhist Path to Awakening, Oneworld Publications, 2001.

'The Matikas: Memorization, Mindfulness, and t h e List', included in Janet Gyatso, In the
Mirror of Memory, Albany SUNY, 1992.

Richard GOMBRICH, HOW Buddhism Began: The Conditioned Genesis of Early Buddhism, Synergy Books
International.

Ven. Pategama GNANARAMA, The Mission Accomplished: A Historical Analysis of the Mahaparinibbana
Sutta of the DTgha Nikaya of the Pali Canon, Ti-sarana Buddhist Association, 1997.

Raniero GNOLI (ed.), The Gilgit Manuscript of the Sanghabhedavastu (17th section of the Mulasarvastivada
Vinaya), Part II, Istituto Italiano per il Medio ed Estremo Oriente, 1978.

Paul HARRISON (trans.), The Samadhi of the Direct Encounter with the Buddhas of the Present
(Pratyutpanna-Buddha-Sammukhavasthita-Samadhi-Sutra, T 416), The International Institute for
Buddhist Studies, Tokyo, 1990.

Ann HEIRMAN, The Discipline in Four Parts: Rules for Nuns According to the Dharmaguptakavinaya,
Motilal Barnasidass Publishers, 2002.

Oskar VON HINUBER, A Handbook of Pali Literature, Munshiram Manoharlal Publishers, 1997.

Jeffrey HOPKINS, Meditation on Emptiness, Wisdom Publications, 1983.

Leon HURVITZ, 'Fa Sheng's Observations', included in Kiyota.

JAYATILLEKE, Early Buddhist Theory of Knowledge, 1963

ChengJiANHUA, A Critical Translation of Fan Dong Jing, the Chinese Version ofBrahmajala Sutra. Available
online, b u t I am not aware if it has been published.

J.J.JONES (trans.), Mahavastu, vols 1-3, Pali Text Society, f r o m 1949.

Yuichi KAJIYAMA, 'Later Madhyamakas', included in Kiyota.

John P. KEENAN (trans.), The Scripture on the Explication of Underlying Meaning (Samdhinirmocana
Sutra, T 676), Numata Center for Buddhist Translation and Research, 2000 (BDK English Tripitaka
29-Iv).

Minoru KIYOTA (ed.), Mahayana Buddhist Meditation, University Press of Hawaii, 1978; Motilal
Barnasidass, 1991.

Ria KLOPPENBORG (trans.), The Sutra on the Foundation of the Buddhist Order, Religious Texts
Translation Series Nisaba, Vol 1, Leiden, EJ Brill, 1973.

Etienne LAMOTTE, History of Indian Buddhism, Peeters Press, 1976.

'The Assessment of Textual Authority in Buddhism', included in Buddhist Studies Review, Vol.
1, no. 1,1983-4

240
LIRongxi (trans.), The Biographical Scripture of King Asoka (A-yu-wang-jing, or Asokarajasutra, T
2043), BDK English Tripitaka, Numata Center for Buddhist Translation and Research, 1993.

LIN Li Kouang (trans.), L'Aide Memoire de la Vraie Loi (Saddharmasmrtyupasthana Sutra), Paris,
Adrien-Maisonneuve, 1949.

Peter MASEFIELD, 'Mind/Cosmos Maps in t h e Pali Nikayas', included in Nathan Katz (ed.), Buddhist
and Western Psychology, Prajna Press, Boulder, 1983.

Hisashi MATSUMURA, Four Avadanas from Gilgit Manuscripts, PhD thesis at Australian National
University, Oct. 1980.

Konrad MEISIG, Das Sramanyaphala Sutra, Otto Harrassowitz-Wiesbaden, 1987.

Bhiksu Thich M I N H CHAU, The Chinese Madhyama Agama and the Pali Majjhima Nikaya, Motilal
Barnasidass, 1991.

Monier MONIER-WILLIAMS, A Sanskrit-English Dictionary (SED), Motilal Barnasidass, 2002. (First

published by Oxford University Press, 1899.)

Hajime NAKAMURA, Indian Buddhism, Motilal Barnasidass, 1996.

Bhikkhu NANAMOLI and Bhikkhu BODHI (trans.), The Middle Length Discourses of the Buddha - A New
Translation of the Majjhima Nikaya, Wisdom Publications, 1995.

Bhikkhu NANAMOLI (trans.), The Path of Purification (Visuddhimagga), Buddhist Publication Society.
The Life of the Buddha, Buddhist Publication Society.
The Pitaka-Disclosure (Petakopadesa), Pali Text Society, 1964.

Thich NHAT HANH, Transformation and Healing

K. R. NORMAN, 'Magadhisms in t h e Kathavatthu', included in Dr A. K. Narain (ed.), Studies in Pali and

Buddhism, B. R. Publishing Corporation, 1979.

NYANAPONIKA Thera, The Heart of Buddhist Meditation, Buddhist Publication Society.

Louis de la Valle POUSSIN and Leo M. PRUDEN (trans.), Abhidharmakosabhasyam, Asian Humanities
Press, 1988.

Sital PRASAD, A Comparative Study of Jainism and Buddhism, Sri Satguru Publications, 2003 (first
published 1932).

J . PRZYLYSKI, Le Concile de Rajagaha.

Bhikkhu Telwatte RAHULA, A Critical Study of the Mahavastu, Motilal Banarsidass, 1978

T. W. RHYS DAVIDS and William STEDE, Pali-English Dictionary (PED), Pali Text Society, 1921-1925, etc.

W. Woodville ROCKHILL, The Life of the Buddha, (1884), reprinted Asian Educational Services, 1992.

H. L. SENEVIRATNE, The Work of Kings, The University of Chicago Press, 1999.

241
N. H. SAMTANI, The Arthaviniscaya Sutra and its Commentary, Kashi Prasad Jayaswal Research
Institute, Patna, 1971.

L. SCHMITHAUSEN, 'Die vier Konzentration der Aufmerksamkeit', Zeitschrift far Missionwissenschaft

und Religionwissenschaft, 60,1976.

Gregory SCHOPEN, Bones, Stones, and Buddhist Monks, University of Hawai'i Press, 1997.
Buddhist Monks and Business Matters, University of Hawai'i Press, 2008.

Normanjoseph SMITH, The 17 Versions of t h e Buddha's First Discourse, available online.

SOMA Thera, The Way of Mindfulness, Buddhist Publication Society (reprinted by WAVE, 1999).

Bhikkhu SUJATO, A Swift Pair of Messengers, Inward Path, 2001.

THANISSARO Bhikkhu, The Wings to Awakening, Dhamma Dana Publications, 1997.

Bhiksu Thich THIEN CHAU, The Literature of the Personalists of Early Buddhism, Motilal Barnasidass,
1999.

Pe Maung TIN (trans.), The Expositor (AtthasalinT), Pali Text Society, 1977.

Neela VELKAR, Upasana in the Upanisads, unpublished PhD thesis, University of Bombay, J u n e 1979.

Ernst WALDSCHMIDT, Das Mahaparinirvana Sutra (in 3 parts), Akadamie-Verlag, 1990,1991 (Complete
reconstructed Sanskrit Mahaparinirvana Sutta).
Drei Fragmente Buddistischer Sutras aus den Turfanhandschriften, 1978 (Sanskrit f r a g m e n t
of Mahaparinirvana Sutta).

Maurice WALSHE (trans.), The Long Discourses of the Buddha - A Translation of the DTgha Nikaya,
Wisdom Publications, 1999.

A.K. WARDER, Indian Buddhism, Motilal Barnasidass, 3 rd revised edition, 2000.

Fumimaro WATANABE, Philosophy and its Development in the Nikayas and Abhidhamma, Motilal
Barnasidass, 1983.

Alex W A Y M A N , Analysis of the Sravakabhumi Manuscript, University of California Press

Michael WILLIS, 'Buddhist Saints in Ancient Vedisa', Journal of the Royal Asiatic Society, Series 3, II, 2,
pp.219-28.

Alexander W Y N N E , 'How Old is t h e Sutta-Pitaka? The relative value of textual and epigraphical
sources for t h e study of early Indian Buddhism', S t j o h n ' s College, Oxford, 2003.
<http://www.ocbs.org/research/Wynne.pdf>

A Sixty Years' Journey in the Ocean of the Dhamma.

YIN SHUN,
Yuanshi Fojiao Shengdian zhijicheng [The Formation of Early Buddhist Texts], Zhengwen
Chubanshe, Taipei, 1971.
Za-ahanJing-Lun Huibian [Combined edition of Sutra and Sastra of Samyuktagama],
Zhengwen Chunbanshe, Taipei, 1991.

242
ABBREVIATIONS

AN Anguttara Nikaya
BAU Brhadaranyaka Upanisad
CBD Connected Discourses of the Buddha
CU Chandogya Upanisad
DA DTrgha Agama
Dhp Dhammapada
Dhs DhammasanganT
DN DTgha Nikaya
EA Ekottara Agama
Iti Itivuttaka
Kosa Abhidharmakosabhasya
LDB Long Discourses of the Buddha
MA Madhyama Agama
MBh Maha Bharata
MLDB Middle Length Discourses of the Buddha
MN Majjhima Nikaya
MVB Madhyantavibhagabhasya
PP Puggala Pannatti
PED Pali-English Dictionary (Pali Text Society)
SA Samyukta Agama
SED Sanskrit-English Dictionary (Monier-Williams)
Skt MPS Sanskrit Maha Parinirvana Sutra
Skt CPS Sanskrit Catusparisat Sutra
Skt SPS Sanskrit Sramanyaphala Sutra
SN Samyutta Nikaya
Sn Sutta Nipata
T Taisho edition of the Chinese Tripttaka
Ud Udana
Vsm Visuddhimagga
YS Yoga Sutra

INDEX

243
*Kathinadhamma Sutta, 246 Ekottara, 25, 27,175, 224
*Satipatthana Mula, 187, 228, 236 elements, 23, 66,125,162, 200, 205, 259
*Vibhanga Mula, 51, 65, 66,199, 200, 208, 227, 261 emptiness, 272
37 wings to enlightenment, 20, 23, 36, 65, 83, 95,112,149, enlightenment-factors, 5,180,197, 202, 217, 223, 226, 229,
172,173,195,199, 251, 260, 261 242
abbhutadhamma, 52,181 First Council, 19, 80,182
Abhidhamma, 12, 23, 29, 35, 41, 42, 63, 65,157,198, 243, 248, forest, 3, 23,119, 213, 247, 248, 256, 261
249,256 four noble truths, 10, 20, 23, 33, 42, 86, 88, 91, 92, 93
Abhidhammas, 14, 40 gatha, 22
Abhidharmakosa, 112 gatha theory, 22
Abhidharmasara, 257 Gatha:, 48
access samadhi, 3,108, 261 geyya, 46, 55, 61, 63, 73, 85,179
Adittapariyaya Sutta, 20, 33, 35, 61, 67, 231 GIST, 6,15, 29, 33, 46, 65, 67, 91, 94,101,127,132,141,173,
Agama, 7,14 199,231
Agamas, 5, 6, 8,16, 26, 28, 47, 65 Gombrich, 3, 25, 260, 297
aggregates, 23, 35, 38, 89, 91, 92,190,195, 207, 213, 229, 231, gradual training, 24, 73, 85,116,145,163,173, 211, 213, 228,
261 231,252
Ajivikas, 19 historical method, 4, 250
Akanuma, 14, 290, 297 impermanence, 10,13,101,162,183,190,197,242, 248
Ananda, 19, 34, 52,156,174,181,182,183 internal/external, 161,162,163,199, 203, 211, 228, 266
anapanasati, 112,150,194,196, 211, 213, 253, 255, 258 Isa, 158
Anapanasati Sutta, 112,156,196 itivuttaka, 49
Anattalakkhana Sutta, 20, 33, 35, 61,197 Jain, 220
Anesaki, 14, 297 Jainism, 9,18
angas, 18,19, 34, 36, 46, 64, 67, 72, 73, 95 Jains, 18,122,127,137,140, 247
Anguttara, 22, 49, 65, 81,124,173 jataka, 50
anguttara principle, 81 Jatakas, 14, 249
Anupada Sutta, 243 jhana, 103,111,128,129,130,132,135,137,149,155, 219,
anupassana, 115,119,136,138,156,157,208,258 221,231,242,252,253,256
Anuruddha, 163,192, 246 jhanas, 41, 80,108,119,133,160,173, 219, 226
Aryan, 9 Kaccayana Sutta, 28
Asoka, 25, 232 Kalupahana, 14
Atthaka, 22, 23, 48 Kayagatasati Sutta, 218
AtthasalinT, 29, 30 Khandha Samyutta, 35, 67, 81, 91,190,192,195, 207, 210
avadana, 52,53,127 Khuddaka, 23
body contemplation, 113,114,125,162,174, 203, 205, 211, Kosa, 249,258,262,271
212, 218, 221, 225, 236, 242, 259 Ksudraka, 57
Bojjhanga Samyutta, 229 Lakkhana Sutta, 232
Brahmajala, 14, 34,123,130,133, 213, 292 Lamotte, 14
Brahmanism, 9 Madhyama, 7
breathlessjhana, 136 Madhyama Agama, 245
Brihadaranyaka, 123,129 Madhyamaka, 28, 274
Bronkhorst, 112,115,116,117,134, 220 Maha Kassapa, 12,34, 35,65,247
Bucknell, 6, 7,17, 33, 72, 73, 297 Maha Saccaka Sutta, 220
Buddhism, 8, 9,10,12,18, 21, 23, 25, 33,102,111,127 Maha Satipatthana Sutta, 93, 230, 232, 243
Catusparisat, 21 Mahahatthipadopama Sutta, 91, 92,162
Catusparisat Sutra, 21, 53, 86 Mahapadana Sutta, 50, 53, 81
Chachakka Sutta, 243 Mahaparinibbana Sutta, 52, 53, 95
Chandogya, 120,123,124,164 Mahasanghika, 5, 25, 224, 227
Chinese, 5, 6,13,14, 26,102, 217 Mahasunnata Sutta, 55
Choong, 17, 297 Mahavagga, 20, 21, 86
clear comprehension, 116,156,183,193, 211, 213, 228, 245, Mahavastu, 21, 53
265 Mahayana, 12,14, 25, 28, 51, 80, 224, 256, 263
commentaries, 47,51, 61, 63,148,207,210,232,261,262 Majjhima, 7, 23, 65, 72, 73
dependent liberation, 149,245 mindfulness, 5,103,114,116,119,121,131,133,135,136,141,
dependent origination, 23, 38, 40,158,190, 209, 247, 262, 263 148,250, 256, 258,271
Dhammacakkappavattana Sutta, 5, 20, 21, 22, 24, 33, 34, 54, mindfulness of the body, 179,195, 219
58, 61,86,92,101,141 Moggallana, 65,181
Dhammapada, 49,157,159 Mulasarvastivada, 35, 262
Dhammasangini, 30,149, 244 Mulasarvastivadin, 20
Dharmaguptaka, 5,25, 26, 48, 209 myth, 13, 30, 249
Dharmaskandha, 29, 40, 63, 65,115,199, 204, 235 myths, 11
Digha, 24, 65, 73, 230, 231 narrative, 20, 24, 34, 36, 53, 86, 90,175,179,182
Dirgha, 21 Nidana, 37, 39, 53, 63, 67
dryjhana, 138 Niddesa, 23,150
dry vipassana, 244,248 Nikayas, 5,16, 22, 26, 47, 65
Dvayatanupassana Sutta, 158 Pali, 5, 6, 7,14, 52, 67,150,153,170, 245, 295
ekayana magga, 158,164 Parayana, 22, 23, 48,129,134
Ekayana Sutra, 117,150,171, 235, 236 Parinibbana, 25,182, 213

244
Parivara, 232 Visuddhimagga, 40
Patimokkha, 54, 58 Vitakkasanthana Sutta, 220
Patisambhidamagga, 210, 253 vyakarana, 46,55, 61, 63, 64, 67, 72,81, 91,177
Patthana, 30 war, 11, 233
perception of light, 111,174, 221 Warder, 4, 5,14, 253, 262, 297
pre-sectarian Buddhism, 5,14,145 Yajur, 18, 62
Puggala Pannatti, 209,252 Yin Shun, 12,17, 28, 46, 55
Puggalavada, 12, 261 Yogacara, 28, 269
Pure Land Buddhism, 111 Yogacarabhumisastra, 28, 33, 46, 63
Request of Brahma, 20, 33, 61
reviewing knowledge, 221
Rig, 18, 62,119,120,157,171
Sacca Samyutta, 33, 67,175
Saccavibhanga Sutta, 5, 73, 90, 92, 93, 231
Saddharmasmrityupasthana Sutra, 262
Sagathavagga, 63, 72, 73
samadhi, 3,108,128,133,135,142,150,155, 219, 221, 245,
247, 248, 252, 265
Saman, 18, 62
samatha, 3,102,107,139,140,155,156, 223, 226, 227, 243,
252,256
Samudaya Sutta, 188,189,190,192, 228, 229
Samuel Beal, 13
Samyukta, 24, 46, 63
Samyutta, 5, 23, 28, 33, 65, 67, 72,81, 85,175, 210
samyutta principle, 81
Sanskrit, 13,14, 63,144
Sariputrabhidharma, 29, 40,115,199, 209, 235
Sariputta, 59, 91, 92,137,180, 243
Sarvastivada, 5, 25, 29, 39, 65, 93,217, 243, 247, 248, 256
Sarvastivadin, 21,112,133
satipatthana, 5, 6, 66,114,125,140,144,148,175,191, 225,
252, 253, 258
Satipatthana Samyutta, 109,125,173,175
Satipatthana Sutta, 3, 5,104,109,112, 213, 226, 235, 236, 244
Satyasiddhisastra, 260
sense media, 23, 35, 38,195, 206, 207, 223, 229
Siksasamuccaya, 265,266
simile, 53,125, 219, 220, 263
Sinhalese addition, 232
Smrityupasthana Sutra, 111,112,186, 217, 235
spiritual feelings, 201, 253
Susima Sutta, 260
sutta, 16,18, 46, 55, 61, 63, 64, 72, 91,132,177, 223
Sutta Nipata, 22, 48, 57,129,157
Suttas, 14,16, 34
Svetasvatara, 124,130,158
Theravada, 3, 5,12,14, 25, 30,104,226, 227, 243, 247, 252, 254
Theravadin, 19,156
Tibetan, 13,14, 20, 274
time, 13, 25, 93,101,102,162, 204, 217, 248, 249, 250, 266
traditions, 4, 5, 6,11,13, 24, 30,102
transcendental jhana, 250,252,258
Udana, 49
Upadesa, 52
Upagupta, 247, 249, 256
Upanishads, 121,122,125,128,163
vedalla, 51
Vedas, 18, 22, 62,120,122,125,128,131
Vibhajjavada, 26, 80, 232
Vibhanga, 5, 23, 29, 30, 40, 41, 51, 65, 93,115,199, 235, 236,
249
Vibhanga Sutta, 65,185, 253
Vimuttimagga, 254
Vinaya, 14,16,19, 20, 21,25, 50, 53, 90,127, 232
vipassana, 2,104,107,134,139,140,155,185, 207, 208, 223,
227, 228, 229, 231, 243, 244, 252, 253, 258
vipassanavada, 2
Visuddhimagga, 160, 251, 252, 263

245
 NOTES

References are to discourse and section of LDB and MLDB; samyutta and discourse of CDB (which
varies from the reckoning in earlier texts and translations, especially in SN 35); to Nipata and
discourse for the Anguttara Nikaya and the Itivuttaka; vagga and discourse for the Udana; and
verse number for the Dhammapada and Sutta Nipata. For other Pali texts I have used the VRI CD
and/or the listed translations.

1
But t h e r e is a p r e c e d e n t in t h e case of t h e Chinese Chan school (Zen in Japanese), w h i c h is derived f r o m t h e Sanskrit
dhyana (Palijhana), w h i c h h e r e j u s t m e a n s 'meditation'.
2
Gombrich
3
Warder, pg. 86 footnote.
8
The t e r m 'cognate' literally m e a n s ' b o r n together', and is used in etymology of words t h a t are believed t o derive
f r o m a c o m m o n ancestor. Some scholars p r e f e r t o use words such as 'parallel' or ' c o u n t e r p a r t ' t o r e f e r t o t h e suttas
f o u n d in Pali and Chinese versions, since t h e s e t e r m s do n o t imply any particular t h e o r y as to t h e relations b e t w e e n
t h e texts in question. However, I a m quite comfortable w i t h 'cognate', since I believe t h a t in most cases t h e t h e o r y
t h a t similar suttas derive f r o m a c o m m o n ancestor is t h e most rational one.
5
See Frauwallner 1956; recently corrected by Willis. Also see Wynne.
6
Beal, pg. xii
7
Anesaki
8
Akanuma. I have b e e n kindly provided w i t h a d r a f t catalogue by Roderick Bucknell, who, t o g e t h e r w i t h Venerable
Analayo, is u p d a t i n g Akanuma, working f r o m t h e published Chinese correspondence tables, and w i t h m a n y additions
and corrections. We are c u r r e n t l y working t o make these available on t h e web at www.suttacentral.net.
9
Lamotte, pg. 156
10
MN 26.21/MA 208/T 765.8. Taking tesam as supplying t h e implied 'their' of t h e second line (cp. DN 18.27/DA 8 / T 9)
and pamuncantu as a poetic variant of adhimuncantu (cp. Sn 1186,1189, AN 1.18/EA 8.1-10). However, one Sanskrit
version has pramodanur, and would translate as 'May those w h o wish t o h e a r rejoice in faith'. Another has pranudantu
kaksah, 'dispel opinions'.
11
See Smith. There is an '18 th ' version at Patisambhidamagga 2.6.1. Although this is n o w in t h e Khuddaka Nikaya, it is
a quasi-Abhidhamma text, so we could say t h a t t h e D h a m m a c a k k a p p a v a t t a n a Sutta occurs in all t h r e e Pitakas.
12
Rahula, pg. 8
13
Kloppenborg
18
See e.g. Nakamura, chapters 2.3 & 2.8
15
Schopen 1997, pg.28
16
Details in Choong, pp.19-22
17
MN 81/MA 63 Ghatikara and MN 83/MA 67 Makhadeva.
18
Tin, pg. 8, 38
19
SN 22.22/SA 73/EA 25.8
20
Sammitiyanikayasastra (San-mi-ti p u lun, T 1689), 863b 11, 865b 10, 863b 9-12. See Thien Chau, pg. 23-28.
21
Norman, 1979, pp. 279-287.
22
This line of reasoning suggested itself i n d e p e n d e n t l y t o b o t h Bucknell and myself.
23
Bucknell, 'Structure'.
28
The following account is based on t h e Chinese (T 1851,) and Rockhill's p a r a p h r a s e of t h e Tibetan.
25
Before Ananda begins his recitation, t h e r e is a slightly odd episode w h e r e Maha Kassapa said to all t h e monks:
'There will be monks w h o are weak in faculties and of scattered mind. They will n o t be able t o learn and memorize
t h e sutta-vinaya-abhidhamma. Therefore it will be fitting for us in t h e m o r n i n g t o compile t h e 'Brief Verses
Samyutta', ( f t f n t t 'Ji frltS =sankhitta-gatha-samyutta?, T28, no. 1851, p. 806, a22-23), in t h e a f t e r n o o n it will be fitting
to compile t h e sutta-vinaya-abhidhamma.' It is n o t clear w h a t this is r e f e r r i n g to; t h e Sagatha Vagga is r e f e r r e d to
j u s t below, in its p r o p e r place in t h e Samyutta, so it seems unlikely this is w h a t is m e a n t . Perhaps it refers t o some of
t h e verse collections of t h e Khuddaka.
26
T28, no. 1851, p. 807, b20-c2
27
SN 85.35/SA 796-797/SA 799
28
Explained at AN 8.29 as f r e e d o m f r o m covetousness and ill-will, right mindfulness, and right samadhi. Also f o u n d at
T 1536.7.

246
29
T24, no. 1491, p. 408, b6-11. Translation f r o m Rockwell pg. 160. The details given by Rockwell have b e e n corrected
following Watanabe pg. 44, and f r o m t h e Chinese. A few of t h e Chinese t e r m s , especially in t h e latter p a r t of t h e
passage, are unclear.
30
CBD, pg. 932.
31
Bucknell, unpublished essay.
32
CBD, pg. 27.
33
CBD, pg. 30.
34
SN 48.43
39
Cp. Puggala P a n n a t t i
36
Cp. Nanavibhanga
37
A list of defilements; t h e n a m e 'Khuddakavatthu' is t a k e n f r o m t h e Dharmaskandha.
38
A list of knowledges, comparable t o t h e Patisambhidamagga.
39
An i n d e p e n d e n t A b h i d h a m m a work, r e f e r r e d to as such in t h e first c h a p t e r of t h e AtthasalinT, w i t h its own matika
of some interest: aggregates, sense media, 18 elements, t r u t h s , 22 faculties, 9 causes (hetu,), 4 n u t r i m e n t s , and 7
contacts, feelings, perceptions, volitions, and minds.
40
=Khuddakavatthu?
41
Sangaha and sampayoga are technical t e r m s in t h e T h e r a v a d a Dhatukatha. P a t t h a n a is t h e title of t h e T h e r a v a d a
A b h i d h a m m a treatise o n causality. Frauwallner says t h e s e all owe s o m e t h i n g t o t h e c o m m o n A b h i d h a m m a tradition.
42
More details are listed in an appendix t o t h e PTS Pali edition of t h e Vibhanga, pg. 437.
43
This is based on merely t h e list of t h e sense media r a t h e r t h a n any particular passage and t h e r e f o r e c a n n o t be
t r a c e d t o any one source.
44
The elements (dhatu) vary greatly in character. While in later literature t h e y w e r e t r e a t e d alongside t h e aggregates
and sense media as given here, t h e t r e a t m e n t in t h e existing Samyutta is m o r e closely related t o d e p e n d e n t
origination.
49
E.g. SN 24/SA 4 Ditthi.
46
Khuddakavatthu
47
SA 41, omitted in SN b u t likely t o be original.
48
E.g. SN 43/SA 890 Asankhata, SN 99/SA 30 Sotapatti
49
Although t h e teachings on t h e efforts and bases for psychic powers are f o u n d elsewhere in t h e Chinese, these
chapters, t o g e t h e r w i t h p a r t of t h e faculties, have b e e n lost f r o m t h e existing Chinese Samyutta due t o a failure in t h e
transmission of t h e texts. Probably t h e m a n u s c r i p t was simply misfiled and later lost. Bizarrely, a passage f r o m t h e
'Life of King Asoka' (Asokarajavadana), w h i c h has n o t h i n g t o do w i t h t h e Samyutta, has e n d e d up in its place - a
r e m i n d e r of t h e h u m a n fallibility of t h e generations of pious copyists w h o faithfully t r a n s m i t t e d this absurdity.
Lamotte was evidently u n a w a r e of this p r o b l e m w h e n he r e m a r k e d : 'As t h e y [the Agamas] were closed m u c h later
[than t h e Nikayas], t h e y m a k e r o o m for works of comparatively r e c e n t date; hence t h e Samyutta contains long
extracts f r o m t h e Asokarajavadana.' (Lamotte (1976), pg. 199) More detailed e x a m i n a t i o n gives n o reason t o conclude
t h a t t h e Agamas were, generally speaking, closed earlier or later t h a n t h e Nikayas; each case needs t o be t r e a t e d
individually.
90
Another hybrid group whose variable position reflects its ad hoc n a t u r e . The position in any particular list (matika)
can be explained as due t o t h e p r e d o m i n a n c e of either t h e five spiritual faculties (which p e r t a i n t o t h e t r u t h of t h e
path) or t h e six sense faculties (which p e r t a i n to t h e t r u t h of suffering). The Vibhanga, implicitly acknowledging t h a t
t h e group of 22 is n o t f o u n d as such in t h e suttas, exceptionally offers n o 'Sutta Exposition' here; h o w e v e r it is likely
t h a t this is f o u n d in t h e Dharmaskandha, probably based on SN 48.9/SA 647 Vibhanga
91
Omitted, no doubt being felt r e d u n d a n t , in t h e existing V i b h a n g a / D h a r m a s k a n d h a / S a r i p u t r a b h i d h a r m a , b u t f o u n d
elsewhere in t h e Abhidhammas, e.g. t h e Pali Dhatukatha, whose matika, staying closer t o t h e original order of t h e
wings to e n l i g h t e n m e n t , is in some ways more archaic t h a n t h e Vibhanga. The Chinese Bala Samyutta is m u c h m o r e
extensive t h a n t h e Pali. The reference is t o t h e relevant Vibhanga sutta.
92
SA 29, omitted in SN b u t likely t o be original.
93
T 676.1 use Keenan, which is based on t h e Chinese. The Tibetan is utilized in t h e French translation by Etienne
Lamotte, Samdhinirmocana Sutra: V'explication des Mysteres (Paris: Adrien Maisonneuve, 1939), and t h e English
translation by J o h n Powers, Wisdom of Buddha: The Samdhinirmocana Sutra (Berkely, CA, Dharma Publishing. 1999).
94
Keenan, pp. 22-23
99
Keenan, pg. 39
96
Keenan, pp. 46-47
97
Keenan, pg. 23
98
Keenan, pg. 97
99
1 t h i n k variations, such as those n o t e d above, are likely t o be m e r e t e x t u a l corruption.
60
Sravakabhumi, pg. 226.
61
These two m i n o r samyuttas, which also follow t h e K h a n d h a - s a m y u t t a in SN, deal w i t h t h e aggregates and are really
j u s t an appendix to t h e Khandha-samyutta. This is similar t o t h e relationship b e t w e e n t h e Satipatthana- and t h e
A n u r u d d h a - s amyuttas.

247
72
The t e x t has two adjacent listings for t h e elements: dhatupratisamyukta and dhatusamganapratisamyukta ('connected
w i t h elements' and 'connected w i t h things associated w i t h elements' [?]). I a m n o t sure of t h e significance of t h e
twofold division; in any case, t h e elements clearly seem t o be out of sequence here.
73
Nine angas occur in t h e Mahasanghika Vinaya T 1425, p 227b; in t h e SangTti Sutra, T no 12, p 227b; in t h e Itivuttaka,
T n o 765 p 684a and 697c; in t h e DharmasangTti Sutra T n o 761 p 612a; in t h e S a d d h a r m a p u n d a r i k a Sutra, T n o 262 p
7c (though this seems t o be a different set), in t h e Dharmasangraha, T n o 764 p 661a and in t h e Dasaviharavibhasa, T
1521 p 19b. My t h a n k s t o Venerable Analayo for this information.
64
Lamotte (1976), pg. 144
65
Some references in Lamotte (1976), p p 161,162.
66
DN 27/DA 5 / T 10/MA154/EA 40.1/Skt
67
DN 26/DA 6/MA 70
68
Boin-Webb, pg. 179
69
Sravakabhumi, pg. 230
70
MN 83.21/MA 67/ EA 1/EA 50.4/T152.87/T211, DN 19.61/DA 3 / T 8 / S k t
71
DN 14.1.17/DA 1 / T 2 / T 3 / T 4 / E A 48.4/Skt
72
AN 4.127/MA 32*/EA 25.3*
73
MN 123.3
74
MA 32
75
Immediately preceding t h e t w o Vedallas is a group of t h r e e discourses t h a t occur t o g e t h e r and in t h e same
sequence in b o t h Majjhimas: MN 77/MA 207 Mahasakaludayin; MN 79/MA 208 Culsakaludayin; and MN 80/MA 209
Vekhanassa. (MN 78/MA179 Samanamandika, falling as it does b e t w e e n t h e 'Maha' and 'Cula' Sakaludayin Suttas, is
obviously a later interpolation, probably j u s t a filing error.) Thus it seems likely t h a t b o t h this group of t h r e e and t h e
two Vedalla Suttas w e r e pre-existing groups t h a t were i n c o r p o r a t e d into t h e Majjhima.
76
Boin-Webb, pg. 180
77
Gnoli, Pt. 1, pg. xix
78
Boin-Webb, pg. 180
79
Boin-Webb, pg. 179
80
Discussed in M a t s u m u r a
81
Beal (1985), pg. 386-7; t r a n s l a t i o n corrected following Lamotte, pg. 177
82
Beal (1985), pg. 140
83
MN 122.20/MA 191.
84
P 957
85
There are some variant readings. I do n o t have t h e various Pali editions available, so m y t h a n k s go t o Venerable
Analayo for t h e following readings. The PTS has: suttam geyyam veyyakaranassa hetu; Burmese and Sinhalese have:
suttam geyyam veyyakaranam tassa hetu; t h e Siamese has: suttageyyaveyyakaranassa sotum. The PTS version would t h u s
seem t o have only two angas, whose explanation t h e disciple w a n t s f r o m t h e T a t h a g a t a (cp. t h e t r a n s l a t i o n by
Horner: '...for t h e sake of an exposition of t h e discourses t h a t are in prose and in prose and verse'). But t h e g r a m m a r
is odd; this phrase would seem r a t h e r t o have f o r m e d by c o n t r a c t i o n f r o m t h e Burmese/Sinhalese reading (=suttam
geyyam veyyakaran' assa hetu). The Chinese version reads ZEIS IRlSc IBiftSfc This does indicate t h a t its original
read hetuh, n o t sotum as proposed by t h e Siamese edition. Sotum ('for t h e h e a r i n g o f . ' ) would seem t o be more
straightforward; grammatically t h e c o m p o u n d f o r m a t i o n in t h e Siamese edition is also more straightforward, leading
m e t o suspect t h e Thai editors have normalized a difficult reading. The Chinese does n o t give a h i n t t o t h e
grammatical problems of this passage. In any case, t h e Chinese and most of t h e Pali versions have t h r e e angas, and
t h e a p p a r e n t presence of two angas in t h e PTS seems easily explicable.
86
Waldschmidt (1950,1951) 40.62; Waldschmidt (1968).
87
Sravakabhumi, pg. 154, 232
88
Sravakabhumi, pg. 154,184, 220, 226
89
Neyyattha and nitattha; Sravakabhumi pg. 228; Abhidharmasamuccaya pg. 179, b u t see de Jong's c o m m e n t , pg. 295.
90
Petakopadesa 1.7. In t h e final sentence I accept Nanamoli's suggested a m e n d m e n t t o alokapancakam, r e f e r r i n g t o
t h e series of five t e r m s in t h e D h a m m a c a k k a p p a v a t t a n a Sutta describing t h e Buddha's realization of t h e noble t r u t h s .
91
Petakopadesa 1.7,1.22
92
Petakopadesa 1.8,1.9,1.12, 4.41, 5.53, 5.54, 7.105
93
Netti 89.
94
Petakopadesa 1.12
95
Petakopadesa 5.53
96
Nanamoli (1964), pg. 133.
97
There are some variations, however. According t o Lamotte, t h e SaddharmapundarTka Sutra has: sutra, gatha,
itivuttaka, jataka, abbhutadhamma, nidana, aupamya, geya, upadesa.
98
DN 16.4.7ff/DA 2 / T 5 / T 6 / T 7 / T 1 4 5 1 / S k t
99
Netti 122
100
In Pali usually spelled veyyakarana or byakarana.

248
101
E.g. DN 11.85/DA 24/P 5595, SN 12.12/SA 372, SN 12.32/MA 23/SA 345/Skt, SN 12.70/SA 347, SN 35.116/SA 234, SN
35.204/SA 1175, SN 41.1/SA 572MN 32/MA 184/EA 37.3/T 154.16, MN 44/MA 210, MN 56.6/MA 133/Skt, M 133.21/MA
165, AN 3.21, AN 6.61/SA 1164
102
DN 9.29/DA 28, DN 2?.32/DA 17, MN 63.?/MA 221/T ?4, SN 16.12/SA 905/SA 2 121
103
SN 41.8/sA 574, AN 10.27/EA 46.8/SA 486-48?*, AN 10.28
104
E.g. DN 18.4/DA 4 / T ?, AN 5.?3, AN 6.62/MA 112/T 58
105
AN 4.42/T 1536.8
106
Dutt, pg. 225
107
Boin-Webb, pg. 17? (adapted following de Jong's note, pg. 2?5); t h e Sravakabhumi pg. 228 is similar, lacking only
t h e final phrase.
108
SN 41.1/SA 572
10
? Petakopadesa 10
110
See Dietz e.g. pg. 26 'kumbhopame vyakarane' (= SN 12.51/SA 2?2/Skt; t h e Pali title is 'parivimamsana'; pg. 33
'phalgunavavade vyakarane' (= SN 12.12/SA 372; t h e Pali title is 'moliyaphagguna'); pg. 52 'pataleyavyakarane' (=SN
22.81/SA57).
111
Przylyski, pp. 45, 334; cp. Rockhill, pg. 160
112
MN 32/MA 184/EA 37.3/T 154, Chapter 6.
113
T 154
114
AN 1.1?7
115
EA 4.2. Analayo (unpublished essay on M 32) discusses this point, as does Minh Chau, pp. 251-257.
116
Another possible instance is t h e J h a n a Vibhanga. This gives an account of t h e gradual training t h a t is briefer t h a n
t h e usual Sutta versions. However, in this case it seems t o me t h a t this is an abbreviation r a t h e r t h a n an archaic
version.
117
My t h a n k s t o t h e a u t h o r for supplying me w i t h a copy of this essay.
118
Stephen Hodge, private communication.
11
? T ?7, T ?8. My t h a n k s t o Rod Bucknell for this information.
120
T24: 35a3 & 57a26-7, Yin Shun 1?71, pp. 720-725. In t h e Chinese, t h e second n i p a t a is mistakenly called t h e 'Sixty-
three-Nipata'.
121
DN 14/DA1/EA 48.4
122
Heirman, pg. 27.
123
AN 10.27/EA 46.8/SA 486-48?*
124
DN 33/DA ? / T 1 2 / s k t
125
AN 10.28
126
Gethin, 1??2.
127
Even this is an oversimplification, for some of t h e Vibhanga topics, such as t h e 37 wings to e n l i g h t e n m e n t , suggest
t h e a n g u t t a r a principle, and parts of t h e DhammasanganT matika suggest t h e s a m y u t t a principle. Examples of t h e
latter were discussed by Nyanaponika Thera in his A b h i d h a m m a Studies. For example, t h e following triads clearly
reflect t h e Sutta sequence of meditative development:
savitakkasavicaradhamma/avitakkavicaramattadhamma/avitakkavicaradhamma >
prtisahagatadhamma/sukhasahagatadhamma/upekkhasahagatadhamma > dassanenapahatabbadhamma/bhavanaya
pahatabbadhamma/n'eva dassanena na bhavanayapahatabbadhamma
128
AN 1.1-10/EA ?.7-8.
12
? AN 5.55.
130
Bucknell, unpublished essay.
131
PED has some interesting r e m a r k s on this obscure m a t t e r , some of w h i c h are endearingly far-fetched - it invites
comparison b e t w e e n t h e five groups of twelve musicians m e n t i o n e d in t h e c o m m e n t a r y t o t h e V i m a n a V a t t h u w i t h
t h e five times twelve cromlechs in t h e outer circle at Stonehenge!
132
Different editions vary t h e n u m b e r i n g of discourses by one at this point. My correspondence table lists AN 3.81
Samana t o AN 3.?0 P a n k a d h a as having SA c o u n t e r p a r t s .
133
Most of this passage is included in t h e widely available Life of the Buddha by Bhikkhu Nanamoli.
134
MN 28.2-6/MA 30
135
MN 141/MA 31/T 32/EA 27.1/T 1435.60
136
DN 2?.17/DA 17
137
According t o Bucknell, unpublished essay.
138
DN 16.3.50
13
? Waldschmidt ( m c - m ! ) , W.7-10.
140
Waldschmidt (W68).
141
The Sanskrit traditions, including t h e Mahayana and t h e Sarvastivada, usually add dhammata (the 'way of things')
to sutta and vinaya, although t h e r e is one Chinese Sutta t h a t j u s t has t h e two. Remarkably, t h e Theravadin Netti also
has dhammata. See Lamotte, 1?83-4, pg.4.
142
Waldschmidt (l950-195l), 24.2.

249
143
Waldschmidt (1950-1951), 40.62
144
Rockhill, pg. 132,140.
145
Cousins (1983), pg. 3.
146
Samantapasadika, V e r a n j a k a n d a v a n n a n a , Pathamasangitikatha
147
DN 16.3.18/Skt MPS 17.17
148
Skt MPS 16.1-12. The Pali has t h e Mara episode (DN 16.3.7), b u t w i t h o u t m e n t i o n i n g t h a t Mara first requested t h e
Buddha pass away w h e n he was newly enlightened at Uruvela. This s t a t e m e n t is n o t f o u n d until t h e recapitulation at
DN 16.3.34, following t h e eight causes of earthquakes, w h e r e a s t h e Sanskrit has t h e full passage once only. The Pali,
in b e t w e e n t h e t w o m e n t i o n s of t h e Request of Mara, drags in several irrelevant sets of eight d h a m m a s (eight
assemblies, eight stages of mastery, eight liberations), an arbitrary application of t h e a n g u t t a r a principle. It is, in fact,
probably t h e inclusion of t h e s e e x t r a sets of d h a m m a s t h a t necessitated t h e Pali recapitulating t h e Request of Mara;
this is a classic sign of t e x t u a l manipulation, technically k n o w n as 'resumptive recapitulation'.
149
Skt MPS 16.12
150
Skt CPS 4.7
151
DN 16.4.42/Skt MPS 29.5-12. The Skt does n o t include t h e controversial r e f e r e n c e t o 'pig's delight', n o r does it
m e n t i o n t h e illness. Instead it inserts an odd section on an 'evil m o n k ' w h o carried off t h e 'metal pot' u n d e r his
armpit, only t o be revealed by t h e power of t h e Buddha.
152
Nyanaponika, pg. 11
153
Inserted passage f r o m AN 10.54
154
AN 4.94
155
AN 9.4
156
MN 122.7 cp. MN 19.8-10, MN 20, MN 4.22, SN 40.1
157
AN 2:3.10
158
v o n Hinuber, pg 37
159
Analayo, pg. 16
160
Since completing t h e study of satipatthana, Venerable Analayo has u n d e r t a k e n a systematic comparative study of
all t h e suttas in t h e Majjhima Nikaya w i t h their Chinese and o t h e r cognates. He has b e e n kind e n o u g h t o share w i t h
me some of his d r a f t studies. This includes a comparative study o n t h e Satipatthana Sutta, covering m u c h t h e same
material as this w o r k f r o m a somewhat different angle.
161
Analayo, pg. 121
162
AN 5.28
163
Schmithausen (1976) pp. 241-66; Bronkhorst (1985) pp. 309-12.
164
AN 10.60/Tibetan (Peking edition, Otani reprint, Tokyo 1956) 754
165
SN 47.10
166
E.g. AN 10.60, Iti 80, 81, Dhp 7, 8.
167
Frauwallner (1995), pg 43ff.
168
1 omit Bronkhorst's lists of t h e specifications of t h e satipatthanas in t h e Vibhanga and Dharmaskandha, which
contain some errors of detail.
169
DN 2.65, etc. Here I substitute my t r a n s l a t i o n for Bronkhorst's for consistency.
170
SN 47.2, SN 36.7,8
171
MA 144=MN 107 Ganakamoggallana Sutta
172
MN 125/MA 198
173
Meisig, pg. 273
174
AN 10.62/MA 51-53
175
MN 67.16
176
CU 6.8.5-6
177
AN 5.193, SN 46.55
178
RV 5.81.1
179
BU 4.3.7
180
Aitareya Brahamana 5.5.27
181
Crangle, pg. 71
182
Crangle, pg. 74
183
BU 1.5.14
184
Crangle, pg. 198
185
DN 13.10. The cognate DA 26 mentions t h r e e paths: |lj J E ^ i f i (CBETA, T01, no. 1, p. 105, b13). It
is n o t clear t o m e h o w closely these m i g h t m a t c h w i t h t h e Pali.
186
Jayatilleke, pg.
187
See 'A Pali Reference t o Brahmana-Caranas', included in Wijesekera.
188
CU 1.12, cp. AN 5.191
189
BAU 6.4.9
190
BAU 1.5.17

250
191
BAU 3.7.23
192
BAU 8.8.23
193
CU 7.8.2
198
SU 1.18
199
SU 2.8 Cp. Sn 1038f
197
SU 2.9
197
SU 2.11
198
E.g. MN 37.17ff
199
E.g. BAU 9.12
200
CU 1.19
201
MaitrT 1.3, 3.8
202
TaittirTya Upanisad 3.2-7
203
MN 27/MA 208
208
The account in t h e Sanghabhedavastu of t h e Mulasarvastivada Vinaya omits t h e m e n t i o n of lip recital.
209
SN 39.103
207
The Sarvastivadin version (MA 208) mentions only faith, energy, and wisdom here, b u t includes mindfulness j u s t
below. The Sanghabhedavastu (Gnoli pg. 97) and t h e Lalitavistara (239.2) m e n t i o n all five spiritual faculties.
207
E.g. SN 88.9
208
AN 8.123
209
AN 8.129
210
Compare t h e following verses:
Yatha nadyas syandamanas samudre
Just as rivers flowing into t h e ocean

Astam gacchanti namarupe vihaya

Go t o t h e i r end, having d r o p p e d n a m e & f o r m

Tatha vidvan namarupad vimuktah

Thus t h e realized [sage], f r e e d f r o m n a m e & f o r m

Parat-param purusam upaiti divyam

Beyond t h e beyond is t h a t Man he enters, divine.

(Mundaka Upanisad 3.2.8)

Acci yatha vatavegena khittam

Just as a flame tossed by a s t r o n g wind

Attham paleti na upeti sankham

Goes out t o t h e end, and does n o t e n t e r reckoning

Evam muni namakaya vimutto

Thus t h e sage, f r e e d f r o m t h e n a m e - g r o u p [i.e. m e n t a l factors]

Attham paleti na upeti sankham

Goes out to t h e end, and does n o t e n t e r reckoning.

(Sutta Nipata 1080)

211
BU 3.9.27
212
Sn 1027
213
Sn 1009
218
Sn 1107
219
Sn 1070, Sn 1113ff. Interestingly, t h e sphere of nothingness is described in Sn 1070 as a 'support' (arammana) for
crossing over. This rare positive use of 'support' m a y be c o m p a r e d w i t h t h e M a h a b h a r a t a passage quoted above t h a t
describes t h e u n c o n c e n t r a t e d mind as 'without support'. T h e j h a n a Samyutta also speaks of t h e development of'skill
in t h e support'.
217
SN 87.92, SN 87.93
217
SN 87.98
218
E.g. MN 83/MA 77/EA 1/EA 90.8/T 192.87/T 211 Makhadeva; DN 19 Mahagovinda also includes t h e divine abidings,
but t h e y are absent f r o m DA 3, T 8 p207c-210b, and Mv 3.197-228.

251
219
SN Sagatha Vagga verse 269, AN (4)449-51. This phrase was somewhat misleadingly r e n d e r e d by Bhikkhu Bodhi in
CBD as ' d i s c o v e r e d j h a n a ' . Perhaps t h e accusative h e r e could be read as i n s t r u m e n t a l ('awakened by m e a n s of jhana').
220
SN 35.132
221
MBh 12.188.9
222
MBh 12.188.15. Bronkhorst (2000) pg. 71 notes t h a t here, as well as in t h e Yoga Sutra and in some Buddhist works,
vitakka and vicara 'are apparently looked u p o n as special faculties in t h e first j h a n a , n o t as m e r e t h o u g h t r e m a i n i n g
f r o m ordinary consciousness'.
223
MBh 12.188.22
224
YS 3.12
225
YS 1.17. The w o r d 'form', rupa, does n o t occur in all texts.
226
DN 1/DA 19
227
YS 1.18
228
YS 1.20
229
Bronkhorst (2000), pp. 72ff
230
SN 14.11
231
YS 2.3-6
232
MaitrT6.18
233
YS 3.1
234
Crangle pp 117-119 discusses t h e similarity b e t w e e n Buddhist sati and yogic dharana, and their role as support for
jhana.
235
MN 36.20, etc.
236
Gnoli, pg. 103
237
MN 12.56
238
MN 85.10/DA 2 21/T 1421.10
239
MN 14.20
240
MN 36.31, MN 85, MN 100
241
CU 6.7
242
SN 41.8
243
sukkajjhana.
244
Uttarajjhayana 29.72/1174
245
E.g. T h a n a n g a Sutta. See Bronkhorst (2000), pg. 38ff
246
Dayal, pg. 95
247
Prasad, p p 167-168
248
DN 18.27, MN 117.3, SN 45.28, AN 7.42. The t e r m 'parikkhara' more familiarly refers t o a monk's f o u r requisites -
bowl, robes, dwelling, and medicine. Here it obviously m e a n s t h e factors t h a t are 'pre-requisite' for a t t a i n i n g j h a n a .
In later usage it is replaced in this sense by its etymological t w i n 'parikamma', usually r e n d e r e d something like
' p r e p a r a t o r y work'.
249
MN 44.12
250
SN 48.50
251
MN 44.11/MA 210
252
SN 46.3/SA 736/SA 740/SA 724*
253
SN 46.51/SA 715
254
E.g. SN 54.13/SA 810
255
AN 6.102-4
256
E.g. Skt MPS 27.16; Skt CPS 6.1; Skt SPS 63
257
DN 10.2.1-18/DA 2 42. Walshe's t r a n s l a t i o n h e r e is faulty. He has: 'This comes t o h i m t h r o u g h concentration'. It
should read: 'This is, for him, w h a t pertains t o concentration'.
258
SN 48.9, etc.
259
MA 189, etc.
260
E.g. Maha Niddesa 1.1.3
261
Patisambhidamagga 169
262
SA 610/SN 47.39
263
Netti 4.23
264
SN 48.9/SA 647
265
Waldschmidt (1950,1951) 10.8
266
DN 2.64/DA 27/T 22/EA 42.7/SA 154-163*, etc.
267
Meisig, pg. 268.
268
AN 10.176
269
Sn 1112
270
SN 42.8/SA 916/SA 2 131, AN 10.219

252
271
Bhikkhu Bodhi has c h a n g e d his r e n d e r i n g h e r e f r o m 'connected w i t h t h e body' t o 'connected w i t h sensual desire'.
According t o his n o t e 1177 in t h e revised 2001 edition of MLDB, t h e PTS edition has kayupasamhitam (connected w i t h
t h e body), abbreviates t h e n e x t two, t h e n has dhammupasamhitam. The Burmese editions, s u p p o r t e d by a 1937
Sinhalese edition, have kamupasamhitam. He also cites t h e Chinese t r a n s l a t i o n in support. The Sanskrit version
q u o t e d j u s t below, however, supports kayupasamhitam.
272
Harrison. The Sanskrit has b e e n tentatively r e c o n s t r u c t e d by t h e translator.
273
SN 98.10/SA 813, SN 98.13/SA 810, MN 118.23ff
278
Rg Veda 10.170.8
279
Sn 299
277
Dhp 293
277
Isa 7-7
278
Sn 3.12
279
Svetasvatara 12
280
BU 8.8.19
281
Dhp 7, 8
282
SN 22.39-82/SA 27, SN 22.187/SA88
283
SN 37.7/SA 1028, SN 37.8/SA 1029, MN 37/SA 909/EA 19.3
288
See V s m t r a n s (Nanamoli), pg. 178 n o t e 87.
289
SA 710, b u t n o t its cognate SN 87.2
287
Iti 3.37. This includes t h e unusual phrase 'ajjhattam parimukham'.
287
MN 77.23
288
DN 18.27
289
MN 10.18
290
Sn 209
291
SN 87.39/SA 710
292
MN 127.17
293
Masefield. My t h a n k s to t h e a u t h o r for supplying m e w i t h a copy of this article.
298
E.g. MLDB, pg. 1188, n o t e 139; Gethin (2001), pg. 70.
299
Gethin (2001), pp. 99-77.
297
MN 12.37/ T 797/SA 712*/SA 788*/SA 701*/EA 27.7*/EA 31.8*/EA 87.8*/ EA 90.7*/T 780*/T 781*/T 802*/Skt*.
297
SN 87.18/SA 707/SA 1189/SA 2 102/sA 3 8
298
SA 1189/SA 2 102/SA 3 8
299
SA 1189, SA2102
300
SN 87.18, SN 87.83
301
The Chinese partly overlaps SN 87.8/SA 717
302
Gethin (2001), pg. 71
303
BU 2.8, 8.9
308
Wijesekera, pg. 282
309
SN 3.1.8. Curiously, this has n o cognates.
307
The c o m m e n t a r i e s say t h a t muta h e r e m e a n s cognized by t h e nose, tongue, or body; t h a t is, t h e y correlate this list
w i t h t h e six senses; t h u s some c o n t e m p o r a r y translators r e n d e r it as 'sensed'. But muta is j u s t a past participle of t h e
n o r m a l w o r d 'to think', and never m e a n s 'sensed'. The Upanisadic context makes it plain t h a t t h e m e a n i n g h e r e is
w h a t has b e e n ' t h o u g h t ' or 'conceived'.
307
MN 1 / M A 1 0 7 / T 97
308
DN 23.19/DA 7/MA 71/ T 89
309
SN 39.287/SA 1179
310
AN 8.197
311
MN 72.17/SA 972/SA 2 197
312
BU 8.9.18
313
See Gethin (2001), pg. 72 n o t e 182.
j f ^ u j t i t ^ k ((BETA, T02, no. 129, p. 978, a8-9)
319
The relation b e t w e e n s a t i p a t t h a n a and d e p e n d e n t origination is implicit in t h e Samudaya Sutta, w h i c h we will
discuss below, and m a d e explicit in t h e Sariputrabhidharma.
317
AN 3.79/EA 21.8, AN 8.39, AN 8.198/SA 979, AN 7.27/SA 990, AN 7.30/T 1937.17, AN 9.37/SA 997
317
MN 77.83/SA 973*/SA 2 207*
318
DN 18.28/DA 8 / T 9
319
AN 7.117ff
320
AN 8.73
321
AN 7.29
322
Cp. MN 32/MA188, MN 123/MA 78
323
AN 9.122

253
324
SN 47.18/SA 607/SA 1189/SA 2 102/SA 3 4; SN 47.43
325
SN 47.31
326
Waldschmidt (1950,1951), 14.14.
327
SN 47.12
328
DN 16.1.16-17
329
DN 2 8 / D A 1 8 / T 18
330
SN 47.21*/SA 628; SN 47.22; SN 47.23.
331
SN 47.2/SA 610
332
AN 4.41/T 1536.7
333
SN 12.33/SA 356, SN 12.34/SA 357.
334
SN 51.19 (No SA cognate, as t h e Iddhipada Samyutta has b e e n lost f r o m t h e Chinese.)
335
SN 51.20
336
MA 98; cp. SN 51.20
337
SN 47.34/SA 634
338
SN 47.39/SA 610
339
Boin-Webb, pg. 160
340
CBD mistakenly has t h e singular 'feeling'.
341
SN 126-128/SA 256
342
SN 22.57/SA 42/EA 41.3. Also in t h e partial 'Other translation' of SA, and forms t h e basis of a long i n d e p e n d e n t
Seven Cases Sutta in t h e Chinese, which appears to be a SangTti Sutta style compilation.
343
e.g. MN 138/MA164 Uddesavibhanga
344
We m u s t add our usual reservations regarding t h e c o n t e m p l a t i o n of dhammas: although t h e w o r d 'attention' is n o t
used, t h e way t h e practice is p r e s e n t e d is very similar t o o t h e r contexts w h e r e t h e w o r d 'attention' is used,
u n d e r s c o r i n g t h e fact t h a t c o n t e m p l a t i o n of d h a m m a s introduces an i m p o r t a n t vipassana dimension t o satipatthana.
345
SN 12.1/SA 298/T 124/Skt etc. The Chinese sometimes has 'the four immaterial aggregates', w h i c h is almost
certainly a later corruption; however, this does n o t affect t h e c u r r e n t a r g u m e n t .
346
DN 15.20/DA 13/T 14/MA 97/T 52/Skt
347
SN 52.1
348
SA 536
349
SA 535
350
SN 36.11, 36.15-20
351
SN 36.2, 36.7-10, 36.22
352
SN 36.7, 36.8
353
SN 22.80
354
MN 78/MA 179 Samanamandika
355
SN 35.132, etc.
356
SN 35.134
357
MN 118
358
Chih-ching, according t o M i n h Chau pg. 347. These miscellaneous discourses, f o u n d a d d e n d e d t o t h e m a j o r
collections, consist of alternative translations and sometimes texts n o t f o u n d in t h e m a j o r Agamas. Their doctrinal
affiliations are usually u n k n o w n and t h e y have b e e n e v e n less studied t h a n t h e m a j o r collections.
359
The Pali t e r m s are slightly different: patinissagga in anapanasati and vossagga in t h e e n l i g h t e n m e n t - f a c t o r formula,
b u t I do n o t see any significance in this variation. Bhikkhu Bodhi (CDB, pg. 1892, n o t e 7) tries t o distinguish t h e m ,
while acknowledging t h a t this does n o t find s u p p o r t in t h e commentaries. He suggests t h a t patinissagga refers t o t h e
final stages of insight, w h e r e a s vossagga comes close in m e a n i n g t o Nibbana. This interpretation, however, is
untenable, for patinissagga in fact occurs in two of t h e m a i n definitions of Nibbana: in t h e t h i r d noble t r u t h definition,
and as t h e 'relinquishing of all belongings' (sabbupadhipatinissaggo).
360
SN 54.10/SA 813, SN 54.13/SA 810, MN 118.23ff
361
T 1537
362
T 1548. Frauwallner (1995) discusses t h e relations b e t w e e n these texts in detail. See pp. 15ff, 43ff, and 97ff.
363
SN 36.31/SA483
364
AN 6.26/SA 550
365
SN 22.122/SA 259
366
A similar shift h a p p e n e d to nama. Although t h e suttas take pains to define this excluding cognition, t h e later
scholastics explained it as all four immaterial aggregates, including cognition. This i n t e r p r e t a t i o n is already f o u n d in
t h e Sarvastivada Nidana Samyukta, f r o m w h e r e it n o doubt m a d e its way into t h e Sarvastivada Abhidhamma. So in
t h e case of namarupa, as in t h e case of t h e satipatthanas, we see t h e t e n d e n c y to integrate o t h e r doctrinal formulas
w i t h t h e aggregates first appear in t h e Sarvastivada, t h e n later in t h e Theravada.
367
SN 48.2-7
368
SN 54.6/SA 805
369
MN 140.18/MA 162; MN 62.12/EA 17.1.

254
370
DN 28.7/DA 18
371
MN 28.6/MA 30
372
E.g. DN 33.1.11.10/DA 9
373
MN 62/EA 17.1
374
Ud 4.1
375
Iti 3.36
376
Ud 10.9
377
Jianhua. This contains a lengthy comparative study as well as full t r a n s l a t i o n and Chinese t e x t of t h e
Dharmaguptaka Brahmajala Sutta.
378
Dutt, pg. 254
379
Conze, pg. 153-155
380
Dayal, pg. 90
381
Conze, pg. 580
382
MA 98
383
Thich Nhat Hanh's t r a n s l a t i o n refers in t h e plural t o 'all Tathagatas of t h e p r e s e n t (including myself)', w h i c h
sounds very Mahayanist; b u t t h e t e x t is singular.
384
MN 119/MA 81
385
Analayo, pg. 53ff, discusses t h e various s a t i p a t t h a n a similes in more detail.
386
See Sujato, p p 164f
387
MN 36.20
388
Paccavekkhana nimitta. Minh Chau has 'contemplating image', suggesting t h a t t h e Chinese t r a n s l a t i o n was
influenced by t h e later m e a n i n g of nimitta. See discussion in Sujato, pg. 166f.
389
Cp. AN 5.28, DN 34.1.6/DA 10
390
The situation is complicated by a curious editing feature, w h i c h I was alerted t o by Bucknell. In t h e Anguttara, t h e
fivefold right samadhi is followed by saying t h a t one can t h e n realize any of t h e clear knowledges, illustrated by t h e
similes of t h e b r i m f u l pot, t h e brimful pond, and t h e r e a d y chariot. This entire passage is included in t h e T h e r a v a d a
Kayagatasati Sutta, b u t n o t t h e Sarvastivadin version. Thus f r o m t h e fivefold right samadhi, t h e j h a n a s w i t h similes
are included in b o t h Kayagatasati Suttas, t h e e x t r a practices in t h e Sarvastivada only, and t h e e x t r a t h r e e similes in
t h e Theravada only.
391
Gethin follows Warder's e r r o r h e r e in stating t h a t t h e Chinese version omits t h e sense media. It does not. Other
c o m m e n t a t o r s , such as Minh Chau, seem t o have b e e n misled by an inconsistency in t h e t e x t itself. The description of
t h e practice, as in t h e Theravada, speaks of t h e b o t h t h e internal (e.g. eye) and external (e.g. visible forms) sense
media. But w h e r e a s t h e T h e r a v a d a s u m m a r y at t h e e n d of t h e section rightly m e n t i o n s t h e internal and e x t e r n a l
sense media, t h e Sarvastivada mentions only t h e internal; n o doubt this is an ancient editing mistake.
392
SN 46.29
393
Schopen 2004, pp. 395ff
394
Schopen 2004, pg. 283, n o t e 59. This note, u n f o r t u n a t e l y disconnected f r o m Schopen's m a i n discussion of t h e
m a t t e r , gives a couple of f u r t h e r references.
395
24 in n u m b e r . T h e r a v a d a has 31. I have available t o m e t h r e e translations of t h e list of body parts f r o m t h e same
Chinese t e x t of t h e Sarvastivada S m r t y u p a s t h a n a Sutra. They r e c k o n t h e parts of t h e body at 29, 30, and 31
respectively.
396
SN 46.35
397
SN 46.2, 5, 7, 23, 24, 33, 34, 37, 39, 40, 49, 51, 52, 53, 55, 56
398
DN 22
399
Majjhima Nikaya, Vol. 1, PTS, pg. 534. The preface lists a n u m b e r of similar interpolations f r o m t h e Digha into t h e
Burmese Majjhima, b u t n o n e are of any l e n g t h or doctrinal significance.
400
MN 141
401
SN 18, SN 25, SN 26, SN 27
402
AN 7.80ff
403
Vasumitra, Sakyaprabha, Vinitadeva (see Hopkins, pp 713-719); Bhavya, Taranatha, Vasumitra,
Sariputrapariprccha, Manjusripariprccha (see Lamotte, p p 529-538).
404
Schopen 1997, pg. 5
405
Abeynayake, pp.163-183
406
DN 30/MA 59
407
The doctrinal material of t h e Sarvastivada version is t h u s entirely stock formulae, lacking all t h e unique material
peculiar t o t h e T h e r a v a d a version. But t h e setting is m o r e complex: it has t h e m o n k s sitting a r o u n d discussing t h e
' w o n d e r f u l and marvelous' qualities of t h e Tathagata, t h e n t h e Buddha e n t e r i n g and e n q u i r i n g into t h e topic of
conversation, etc. (as in t h e M a h a p a d a n a Sutta). So in t h e angas it belongs to t h e abbhutadhammas. The Theravada
j u s t has t h e Buddha giving t h e discourse straight off, and so t h e identification as abbhutadhamma is n o t explicit. The
c o m m e n t a r y explains t h e n e e d for t h e discourse being precisely t h e question of w h a t k a m m a is it t h a t gives rise t o

255
t h e marks, an issue t h a t is n o t addressed at all in t h e Sarvastivada. All this raises t h e question of w h e t h e r t h e t w o
may be r e g a r d e d as cognate texts at all.
408
MN 111
409
MN 148/MA 86
410
Watanabe, pg. 54
411
MN 28/MA 30. The following i n f o r m a t i o n is f r o m Bhikkhu Analayo (unpublished essay).
412
MN 28.6
413
MN 28.9
414
MN 28.27
415
To facilitate easy comparison, w h e r e t h e parallels w i t h t h e Pali are close and obvious I have substituted my own
p r e f e r r e d renderings of technical t e r m s for those given in Minh Chau. Here and elsewhere t h e Chinese regularly has
'right knowledge', evidently reading samma nana instead of t h e Pali sampajanna.
416
cp. AN 5.24, 5.168, 6.50, 7.61, 8.81,10.3-5,11.3-5
417
Chinese has 'the ear-sphere'.
418
Chinese has 'receiving and h e a r i n g t h e d h a m m a ' .
419
Chinese has 'consideration of patience', evidently a mistranslation of nijjhanakkhanti. Khanti in this sense seems t o
be misunderstood t h r o u g h o u t t h e Chinese tradition, leading t o a distinct shift in m e a n i n g in m a n y passages.
420
Probably yoniso manasikara, 'paying a t t e n t i o n t o t h e root'.
421
cp. MN 95
422
cp. SN 12.23
423
Probably a confusion s t e m m i n g f r o m t h e idiom kayena phusati, 'one personally contacts'.
424
Chinese has 'think of its characteristic', evidently f r o m nimittam manasikaroti.
425
Kasinayatana.
426
Most of these points are discussed more fully in Minh Chau.
427
MN 32
428
MA 184
429
Recently it has b e e n argued t h a t Upagupta m u s t be assigned to t h e t i m e of Kalasoka, a few generations before t h e
famous Buddhist p a t r o n Dhammasoka. This would m e a n t h a t he was n o t a c o n t e m p o r a r y of Moggaliputtatissa, and
might suggest a historical r a t h e r t h a n sectarian explanation for t h e adoption by t h e Buddhist communities of such
seemingly different figures as patriarchs.
430
See e.g. Li, Chapter 8
431
MN 31/MA 185 Culagosinga; MN 25/ MA 178 Nivapa; MN 99/MA152 Subha.
432
MN 82/ MA 132 Ratthapala
433
MN 75/MA 153 Magandiya, MN 99/MA 152 Subha.
434
Kosa 6.19d
435
Vibhanga pg. 374
436
Vibhanga pg. 411ff
437
E.g. Dhs 597, 636, 642-646,1115,1366
438
Here we see t h e move t o w a r d s elevating kasinas f r o m obscurity in t h e Suttas to primacy in t h e Visuddhimagga.
439
We have already e n c o u n t e r e d above a Sutta passage w h e r e t h e signless c o n c e n t r a t i o n is clearly distinguished as
different f r o m s a t i p a t t h a n a (SN 22.80), although it is n o t clear w h e t h e r these things have t h e same m e a n i n g in t h e
Abhidhamma; t h e t r e a t m e n t of these meditations in t h e Suttas is s o m e w h a t obscure and variable.
440
Mula TTka Be (1960) t o Vibh-A 287; quoted at Gethin pg. 323
441
Vsm 22.40
442
PP 4.187
443
AN 4.94
444
AN 4.170
445
PP 4.177ff
446
See V s m 20.120
447
SN 47.40
448
SN 52.2, a n d t o s o m e e x t e n t MES
449
G e t h i n 2001, p p . 37-40
450
P e t a k o p a d e s a 6.64
451
T h e d i s c o u r s e a n d c o m m e n t a r y , t o g e t h e r w i t h extracts f r o m t h e sub-commentary, have b e e n published in
t r a n s l a t i o n by Soma T h e r a u n d e r t h e title The Way of Mindfulness.
452
pg. 30
453
pg. 40
454
pg. 54
455
pg. 97
456
pg- 39
457
pg. 165
256
858
pg. 166
859
Asokarajavadana 1206 [Prz363-368]
860
Asokarajavadana pg. 3 [Strong 1983b:178; Przylyski 1923a:363]
861
T XXVI 1023b29 is t h e beginning of t h e c o n t e m p l a t i o n of d h a m m a s section.
862
Frauwallner (1995), pg 179ff
863
Kosa 6.68
868
Kosa 6.9
865
Kosa 6.11
866
Kosa 6.12
867
Kosa 8.1
868
Kosa 6.12
869
The c o m m e n t a r y t o t h e Arthaviniscaya, however, evidently takes this claim seriously e n o u g h t o offer an
explanation: although anapanasati is accompanied by equanimity, t h e m e d i t a t o r t u r n s back t h e awareness t h a t has
air as its object and experiences r a p t u r e and bliss. This does n o t i n t e r r u p t t h e practice, since t h e purpose of t h e
practice is n o t given up, or since he can regain his object quickly. (See Samtani, pg. 102)
870
Kosa 6.13-18
871
Kosa 6.18,15b, 16, etc.
872
Kosa 6.15; cp. 6.2
873
Kosa 6.69
878
Kosa 6.70. Similar sequential explanations occur t h r o u g h o u t t h e literature, as for example t h e Arthaviniscaya
c o m m e n t a r y (Samtani pg. 99).
875
Samtani, pg. 88
876
Samtani, pg. 52
877
Samtani, pp. 96-97
878
San fa t u lun, T. XXV, No. 1506, pp. 15c-30a.
879
Dutt, pg. 187
880
Enomoto suggests t h a t Sarvastivada and Mulasarvastivada are really t h e same. Although this is certainly t h e case
in some of t h e passages he quotes, his a r g u m e n t is n o t really convincing. One of his passages , quoted f r o m t h e Arya-
sarvastivada-mula-bhiksuni-pratimoksa-sutra-vrtti, says : ' Sarvastivada also has t h e root (mula) and t h e branches. Of
t h e m t h e root is one, namely Sarvastivada. The branches derived f r o m it are seven, namely Mulasarvastivada,
Kasyapiya, Mahisasaka, Dharmaguptaka, Bahusrutiya, Tamrasatiya, and Vibhajyavada.' This is obviously polemical,
aimed at exalting one school over another, and its historical value is diminished accordingly. As it stands, t h e passage
clearly distinguishes b e t w e e n Sarvastivada and Mulasarvastivada. W h a t it seems t o be claiming is t h a t t h e
Sarvastivada is t h e real root (mula) school, and those Mulasarvastivadins, w h o p r e t e n d to be t h e root, a r e j u s t a
branch. Enomoto goes on t o say t h a t t h e Mulasarvastivada Vinaya, w h i c h is different f r o m t h e o t h e r Sarvastivada
Vinayas, should be recognized as n o t belonging t o a distinct school, but t o a sub-sect of t h e Sarvastivadins. This
distinction b e t w e e n a sub-sect of a school and a closely related derived school seems a little arbitrary, and p e r h a p s
we n e e d t o simply accept t h a t t h e n o t i o n of w h a t constitutes a 'school' is n o t clear-cut. There is also m u c h textual
evidence left unconsidered by Enomoto. For example, t h e matika of t h e Mulasarvastivadins m e n t i o n e d in t h e i r
Vinaya, while also based on t h e samyutta-matika, has n o specially close relation w i t h t h e topics of t h e
Dharmaskandha. Others suggest t h a t t h e Mulasarvastivada is really t h e same as t h e Sautrantika. This, too, I find
implausible, t h o u g h I have n o t e x a m i n e d t h e a r g u m e n t s in detail. The Mulasarvastivada Vinaya has Maha Kassapa
reciting t h e m a t i k a at t h e First Council. It seems t o m e u t t e r l y implausible t h a t t h e Sautrantikas, whose p r i m a r y t e n e t
was t h a t t h e Suttas alone r e p r e s e n t t h e w o r d of t h e Buddha, should have t h u s e n d o w e d t h e A b h i d h a m m a w i t h
mythic authority, s o m e t h i n g even t h e T h e r a v a d a Vinaya does n o t do. Moreover, a n o t h e r of t h e Sautrantika's key
doctrines was t o deny t h e existence of d h a m m a s in t h e t h r e e times, insisting on a radical version of t h e
m o m e n t a r i n e s s t h e o r y . Again, it seems implausible t h a t anyone calling themselves Sarvastivada, w h e t h e r 'Mula-' or
otherwise, should reject t h e chief t e n e t of t h e school.
881
T 721/T 722
882
So t h e translation; but n o school of Buddhism, t o m y knowledge, has ever t a u g h t t h a t causality does n o t operate
during t h e period b e t w e e n lives. Presumably t h e original m e a n i n g r e f e r r e d t o t h e start of a new cycle of d e p e n d e n t
origination, or s o m e t h i n g of t h a t sort.
883
S a d d h a r m a s m r t y u p a s t h a n a Sutra 38 (quoted in The World's Great Religions, pg. 63, published by Time Inc., 1957).
French t r a n s l a t i o n by Lin.
888
Vsm 17.166
885
AN 8.170/SA 560
886
Asvaghosa was probably n o t a Mahayanist, b u t his works were r e n o w n e d in t h e schools.
887
Chang, pg. 399
888
Cleary, pg. 729 (one-volume edition), vol. 2 pg. 81 (two-volume edition)
889
Harrison, pp. 188,185
890
MN 125/MA 198

257
491
E.g. MN 123.3/MA 32
492
Harrison, pg. 41
493
Vimsatika-karika-vrtti 10c (Anacker pg. 166)
494
Boin-Webb, pg. xvii
499
Boin-Webb, pg. xviii
496
Boin-Webb, pg. 9
497
Boin-Webb, pg. 160-162
498
See CBD, n o t e 1, pg. 1397
499
Boin-Webb, pg. 166
900
Boin-Webb, pg. 168
901
Wayman, pp.97-98.
902
Frauwallner's postulation of 'two Vasubandhus' has b e e n vigorously contested by Anacker (pg. 7ff), b u t defended
by others; suffice to n o t e t h a t Vasubandhu's Abhidharmakosa is f r o m t h e Sravakayana (Sarvastivada/Sautrantika)
perspective, while t h e Madhyantavibhagabhasya is M a h a y a n a (Yogacara).
903
MVB 4.1. Anacker has published an earlier, partial t r a n s l a t i o n of this w o r k in u n d e r t h e title 'The Meditational
Therapy of t h e Madhyantavibhagabhasya' in Kiyota, and a revised full t r a n s l a t i o n u n d e r t h e title ' C o m m e n t a r y on
t h e Separation of t h e Middle f r o m t h e Extremes' in Anacker. In order t o assist readers w h o m a y have access t o one or
other of these works, I give references t o t h e divisions of t h e t e x t r a t h e r t h a n page numbers; t h e s e are, however, n o t
quite consistent in t h e two versions. The references usually r e f e r t o t h e c o m m e n t a r y on t h e n u m b e r e d verses.
904
MVB 4.9b
909
MVB 4.11a
906
MVB 4.7
907
MVB 4.12b
908
Anacker's earlier t r a n s l a t i o n omits t h e second 'not'.
909
MVB 9.17
910
MVB 9.9,10
911
SN 47.19
912
Hopkins, pg. 247.
913
Hopkins, pg. 289.
914
Hopkins, pg. 72, c h a r t 2.
919
Hopkins, pg. 81, c h a r t 3.
916
Boin-Webb, pg. 170
917
AN 9.73
918
Lamotte (1976), pp. 419-426
919
Schopen 2004, pg. 20
920
See Carr, pg. 12
921
Schopen 2004, pg 399
922
Schopen 2004, pg. 143
923
Schopen 2004, pg. 94
924
Schopen 2004, pg. 76
929
Schopen 2004, pg. 77
926
Schopen 2004, pg. 79
927
Schopen 2004, pg. 99
928
Schopen 2004, pg. 93
929
SN 47.13/SA 638
930
Schopen 1997, pg. 203, n o t e 111
931
Schopen 1997, pg.126-128
932
E.g. DN 16.1.12,1.14,1.18, 2.4, etc. The passage occurs w i t h similar f r e q u e n c y in t h e Skt.
933
Schopen 1997, pg. 39
934
Schopen 1997, pg.27

258