PREFACE

THIS book is a revised version of my thesis, "Contemporary
Sinhalese Buddhism in its relation to the Pali Canon', approved
for the D.Phil. degree at Oxford University. The thesis was sub-
stantially written in the university vacations of 1966-7. On sab-
batical leave in Ceylon in the latter half of 1969 I took the oppor-
tunity to make some changes, but although I hope to have elimi-
nated some mistakes, I could not entirely bring it up to date, or
profit by all the publications which have appeared since I started
writing. Rather than delay publication still further I am letting it
go forward now, for revision is a process which can never be
completed. In particular, however, I must mention that Chapter 8
was written before I had read <1Homo Hierarchicus>1 by Louis Dumont
(Gallimard, Paris, 1966); I would formulate most of the first half
of the chapter very differently today, but I have let it stand because
I do not think that my conclusions have been invalidated. I regret
also that the following reached me too late to be used: Heinz
Bechert, "Einige Fragen der Religionssoziologie und Struktur des
su%dasiatischen Buddhismus', <1International Yearbook for the Socio->1
<1logy of Religion>1, vol. 4, 1968, pp. 25 1-95 ; S. J. Tambiah, <1Buddhism>1
and <1the Spirit Cults in North-East Thailand>1 (C.U.P., Cambridge,
1970).
  The greatest pleasure in finishing the book comes from the op-
portunity it gives me formally to acknowledge the help and kindness
I have received during its preparation. The field work on which it
is based was undertaken in Ceylon from August 1964 to August
1965, a visit wholly financed by a Treasury Studentship from the
British Government. My second visit, in 1969, was largely paid
for by a Hayter travel grant awarded to me by Oxford University.
The success of both visits was, however, entirely due to the material
and moral support given me by friends, and to my wife, who both
times accompanied me and kept house unaided in the village for
a total of nearly two years. In England before I left, Professor
Gananath Obeyesekere gave me invaluable advice about criteria
on which to choose my village, as well as a great deal of academic
advice and practical assistance when he had returned to Ceylon.
The Rev. Walpola Rahula received me in Paris, answered a lot of
viii                          <1Preface>1

questions, and has continued ever since to treat me with the great-
est kindness. I owe my introduction to both these scholars to that
renowned patron of Ceylon studies Professor E. F. C. Ludowyk.
When we reached Ceylon we would never have found a suitable
village had not Professor P. W. Vithanage devoted several whole
days to driving us round and giving us the benefit of his geo-
graphical knowledge and geological wisdom. Our first steps in
Sinhalese were guided by Mr. B. L. Fernando of St. Antony's
College, Kat*ugastot*a, who for two months took endless trouble in
teaching us, and has never since lost interest in his pupils. These
are more strictly academic debts in Ceylon. I am not thanking a
field assistant or interpreter because I had none.
   On the more personal side, I wish to thank Major and Mrs.
A. A. de Alwis, Professor and Mrs. E. F. Bartholomeusz, Dr. and
Mrs. W. M. J. Bartholomeusz, Mr. and Mrs. F. Lobo, Professor
and Mrs. G. Obeyesekere, Professor and Mrs. P. W. Vithanage,
Proctor and Mrs. H. A. C. Wickremeratne, Mr. and Mrs. L. Y.
Wickremeratne, and Mr. Jayatissa Yapa for hospitality such as
non-Ceylonese can hardly conceive.
  While the thesis was being composed Dr. Brenda Beck, Dr. K.
Malalgod*a, Professor Sir Karl Popper, and of course my super-
visor, Professor R. C. Zaehner, read all or part of the work and
made valuable suggestions. The typists involved were too many to
name individually, but I would especially like to thank the first, Mrs.
Sriani Fernando, and the last Mrs. Wendy O'Flaherty, for their
sympathetic co-operation. Mrs. Heather ron and my wife sacrificed
much time to proof-reading when they both had better things to do.
  The completed thesis was read by Dr. Bryan Wilson, Pro-
fessor Obeyesekere, and my father, all of whom made just and
stimulating criticisms from which I hope I have benefited. My
intellectual debt to these three people and to Professor Popper
cannot be adequately conveyed by references in the text. Mr.
C. H. B. Reynolds generously gave me Sinhalese lessons for a
whole year without prospect of reward. My mother and my wife
helped immensely with the mechanics of revision and production,
and spotted many slips. Finally I thank the staff of the Clarendon
Press for their encouragement and efficiency.
  I dedicate this imperfect draft to my informants, with gratitude
and respect.
                             CONTENTS

   ABBREVIATIONS                                             xi

   PRONUNCIATION                                           xiii

   Introduction: Aims and Scope of this Book                     1

   I. The Problem: Religious Change 1
   II. My Ceylon?                                               17
   III. The History, Geography and Economy of Ceylon            20
   IV. Field-work and Presentation                              37

1. Sinhalese Buddhism--Orthodox or Syncretistic?                40

2. The Basic Vocabulary of Buddhism                             57

3. The Buddha                                                   80

   I. The Buddha's Biography                                    80
   II. The Buddha as worshipped: Man or God?                   103

4. A Sketch of the Universe as seen from Mi$gala               144

   I. Causation                                                144
   II. The Contents of the Universe 153
   III. Attempts to influence non-human Beings                 191
   IV. Summary                                                 210

5. Total Responsibility in Theory and Practice                 214

6. The Ethic of Intention                                      244

7. The monastic Ideal and the Decline of Buddhism              269

8. Caste in the Monastery                                      294

9. Conclusion                                                  318

   APPENDIX                                                    328

   GLOSSARY                                                    339

   BIBLIOGRAPHY                                                349

   INDEX                                                       357
               ABBREVIATIONS


<1A.N.       Anguttara Nika$ya>1
C.U.P.     Cambridge University Press
<1Dh.A.      Dhammapada At*t*hakatha$>1
<1D.N.       Di$gha Nika$ya>1
<1J.         Ja$taka>1
<1J.A.S.     Journal of Asian Studies>1
<1J.R.A.S.Journal of the Royal Asiatic Society>1
<1Mhv.       Maba$vamsa>1
<1M.N.       Sajjhima Nika$ya>1
O.U.P.     Oxford University Press
P.T.S.     Pali Text Society
R.A.S.     Royal Asiatic Society
<1S.N.       Samyutta Nika$ya>1
<1Vin.       Vinaya Pitaka>1
<1Vv.        Vima$na-vatthu>1
<1Vv.A       Uima$na-vatthu Atthakatha$>1
PRONUNCIATION AND TRANSLITERA-

                         TION OF SINHALESE

                (with remarks on Sanskrit and Pali)


THE full Sinhalese alphabet contains all the letters used in Sanskrit
(ad therefore in Pali) with a few additions, but many of these
letters occur only in words loaned from Sanskrit. The Sinhalese
pronounce Sanskrit and Pali much as do European scholars. A
given letter has generally the same phonetic value (pronunciation)
in Sinhalese, Sanskrit, and Pali.
   Sinhalese vowels follow a phonetic pattern close to English
vowels minus the diphthongs. Sinhalese <1a, a%, e, a$, a%$, e$>1, approximate
respectively to the vowels in English <1bun, ban, Ben, bard, bad>1,
and the first element in <1bane>1; however, in an unstressed syllable
a is a central vowel, just as happens in English (compare the
vowels in <1man>1 and <1postman)>1, and <1a$>1 is pronounced short. Stress is
very light and is as far back as English pronunciation would
tolerate, tending to fall on a long syllable (i.e. on a long vowel or
a vowel followed by more than one consonant). The Sanskrit vowel
<1r>1 ~s pronounced as in American <1cur>1, long or short. In Sanskrit and
Pali the vowels <1e>1 and <so>s are invariably long, and the convention
is not to mark them.
  Of the consonants, <sg>s is always hard, <1c>1 is like English <1ch, j>1 as in
English, <1n+>1 (palatal n) like English <1ny>1 (as in <1onion>1). <1T*>1 and <1d>1 are
pronounced further back than English <1t>1 and <1d>1, but not as the true
retroflexes found in other Indian languages; <1t, d>1 are true dentals,
like those letters in French and Italian. The distinctions between
Sinhalese <1n*>1 and <1n, l*>1 and <1l>1 are historical only. <1Y>1 is as in English
<1yes. S/>1 and <1s>1 occur only in Sanskrit loanwords; in Sanskrit they
are pronounced much like English <1sh>1, but in Sinhalese this is
ignored (except by pedants) and both are pronounced as <1s. H>1
after another consonant denotes aspiration, not a change in
quality (English <1hothead>1 not <1bother)>1; this too occurs only in
Sanskrit loanwords, and tends to be ignored in Sinhalese pro-
nunciation. <1N# n#, m# are pronounced very lightly indeed. M*>1 is a
pure nasal like standard English <1ng>1 in <1singing>1; in colloquial
Sinhalese any final nasal is so pronounced.
xiv

   The letter which is here tranaliterated sometimes as <1v>1, some-
times as <sw>s, may strike the English ear as either, but falls in
between English <1v>1 (a labiodental fricative) and English <1w>1 (bilabial
semi-vowel): it is a bilabial fricative. The exception to this is that
in final position it is pronounced as a semi-vowel, i.e. <1pav>1 is pro-
nounced like the first syllable in English <1power>1. By scholarly con-
vention the transliteration is <1v>1, but the English administration
represented the letter as <1w>1, which is the custom in Ceylon to this
day.
  The scholarly system of transliteration which is standard for all
three languages has been consistently used so far as is feasible
without undue pedantry. Apparent inconsistencies may be due to
switching between languages, especially in the reproduction of
orally gathered material, since speakers tend to make free use also
of parallel Sinhalese forms (e.g. Pali: <1Maha$sena>1; Sinh: <1Mahase$na>1
or <1Mahasen>1).

                    INTRODUCTION: AIMS AND SCOPE
                          OF THIS BOOK
                  I. THE PROBLEM: RELIGIOUS CHANGE

    . . . in case it be said that we have only described the facts in
    relation to a theory of them and as exemplifications of it and
    have subordinated description to analysis, we reply that this
    was our intention.<s1>s
                                            EVANS-PRITCHARD



THIS is intended as a contribution to the empirical study of
religion, and in particular to the study of religious change.
  Religion is studied in many different ways, characteristic of
different academic disciplines. Theologians customarily study
their own religion, which they accordingly meet on its own terms,
and discuss problems posed by and within this historically given
framework of ideas. When they carry the study of other religions
beyond polemics the result is usually called comparative religion.
This vague term can of course always be applied to studies of any
type which compare two or more religions, but the study which
is now called comparative religion is practised and organized
principally by theologians. Philologists study the texts which form
the basis of at least the most famous religions, and speculate about
their origins and what the authors meant. Religious historians
study past religious doctrines and institutions where these are
accessible through documentary or perhaps archaeological
evidence. Sociologists and anthropologists of religion describe and
discuss, usually with some stress on their behavioural aspects,
existing religious practices and institutions; if any difference is
made between the two disciplines it is usually held to be that
sociologists study literate, social anthropologists illiterate societies,
or that sociologists study larger or more differentiated commun-
ities than anthropologists, or that sociologists are more likely to

2

concentrate on formal organization, anthropologists on more in-
formal behaviour. Social psychologists who work on religion are
like sociologists and social anthropologists in that they too study the
religions of contemporary social groups; but they probably
concentrate more on the ideas and attitudes of their subjects.
Other psychologists of religion deal with the religious attitudes of
individuals in various ways ; at one extreme are behaviourists, who
will study only religious behaviour, with a marked disinclination
to include in this category verbal utterance. At the other extreme
are psychoanalysts and other more philosophically oriented
psychologists, who are very interested in the words and notions of
their subjects, but share with the behaviourists a reluctance to use
their subjects' terminology or indeed conceptual framework in the
discussion of religion--unless of course the subject should happen
to hold the same views as they do themselves.
   It may be objected that this classification by subject matter is
naive, because I should have classified these studies by method,
certain methods being typical of the humanities (e.g. the philolo-
gists) and others of the social sciences. I have not done so because
I believe that scholarly research advances by only one method,
shared by the humanities, the social and the natural sciences: the
method of hypothesis and falsification, of theory and test, of con-
jecture and refutation; the method which has been formulated ad
explained by K. R. Popper.<s2>s Metaphysics, with which theologians
are concerned, are not usually amenable to this method; at the
other end of my list, the theories of psychoanalysts and psychia-
trists tend to wander over the border into philosophical specula-
tion or degenerate into technological rules of thumb; but the
method of conjecture and refutation is characteristic of all the
disciplines listed in between; and in so far as studies in them are
duly characterized by this method they constitute what in the
first line I called the empirical, or might call scientific, study of
religion. The method, then, is only one; the means--documents,
interviews, experiments--are infinite, whatever the researcher can

                                                                 3

lay his hands on as he does his best to test his old hypotheses and
to formulate new and better ones which may lead him towards the
truth.
   Again, it may be objected that my classification is jejune and
uninteresting. With this objection I agree; indeed, to elicit it was
one of my reasons for drawing up the list. Academic boundaries are
artificial: the realities are the problems. Problems have a way of
crossing these boundaries; while chasing one the hapless researcher
may wander, alone and unarmed, into the territory of a foreign
and possibly hostile discipline. If he keeps quiet he may escape
unobserved; if he is fool enough to raise his voice he will be
apprehended, and must throw himself on his captor's mercy. I
am in this unfortunate position. By education a philologist, far less
of a historian, and devoid of anthropological training, I have been
chasing the problem of religious change, and found myself far
from home, deep inside the territory of the social sciences, far
from Oxford libraries in a village in central Ceylon. The bulk of
this thesis consists in the presentation of a year's field work, the
very stuff of anthropology; and my problems and conjectural
explanations are of the type which hitherto have interested mainly
those trained in the social sciences.
   In extenuation of my presumption let me plead that some
eminent anthropologists of religion say that they are dissatisfied
with the limitations of much work in their field, limitations
imposed by the tradition that anthropologists study only what
they can observe at first hand. This especially hinders the study
of social change: an anthropologist who goes out to observe a
society naturally produces what is called a synchronic study in
which the society, including its religion, has its picture taken at
one moment and is thus presented as if static. Very rarely can an
anthropologist revisit a society to see for himself how it changes;
moreover anthropology itself is still so young that few communities
have been systematically observed twice even by different people.
Further, as Professor Evans-Pritchard has pointed out in a series
of lectures at Oxford,<s3>s by an unfortunate chance most of the
people who hive had interesting ideas about the anthropology of
religion were not themselves field-workers and lacked detailed and
accurate knowledge of any society on which they might have
tested their hypotheses: Tylor, Durkheim and Le/vy-Bruhl are

4

celebrated examples. Even Radcliffe-Brown and Malinowski, who
did field-work themselves, did not possess comparable data col-
lected in one society at different times, and so lacked material for
diachronic studies; this lack may have decisively influenced their
own lines of interest, which virtually ignored the problem of
social change. In the absence, then, of data collected in the field,
the student of social change must rely on glimpses of the past to be
found in myths and folklore, and on the more conventional written
sources. Professor Evans-Pritchard has invited athropologists to
become historians; and this gives heart to the philologist turned
anthropologist.
   Left me now return to my list of approaches to religion. Another
reason I had for drawing it up was indeed heuristic. There are
several distinct senses in which a religion may be described; some
of these correspond to the academic distinctions made in my list,
and some do not. "Jesus Christ is the son of God' is a theological
or metaphysical statement. "According to the Bible Jesus Christ
is the son of God' is a philological statement; a still more precise
one would be to quote the Bible in Greek. "People have believed
Jesus Christ to be the son of God' is a historical statement, and
in a sense "85 per cent of the present inhabitants of Cremona
believe that Jesus Christ is the son of God' could be called a
historical statement too. Where does this leave the other social
sciences? The academic distinctions between sociology, anthro-
pology, and psychology do not accurately correspond to different
ways of talking about religion; yet within the subject-matter of
these disciplines logical distinctions can and should be made.
   The distinctions I propose to make, and use throughout this
book, are between what people say they believe and say they do,
and what they really believe and really do. These are simple
categories which I have chosen because I find them heuristically
useful; in themselves they, and the words I have chosen to express
them, of course explain nothing. What people really do I shall call
simply religious behaviour. Mat they say about their beliefs and
practices I shall call "cognitive'. Mat they say may of course
vary according to circumstances, but though this point is of great
practical importance in field-work, it is not theoretically impor-
tant. For our present purposes I will assume that people who say
what they believe are speaking the truth; yet there may still be
something more to find out about their beliefs. What people really

                                                           5

believe I am aware to be ultimately unknowable; but this does not
mean that it is nonsense to talk about it. Only a pure behaviourist
refrains from making inferences from what people really do to
what they are thinking or feeling. If a lapsed Roman Catholic
states that he has lost all belief in God, and yet is seen to cross
himself at a moment of crisis, we infer from his actions that he
is operating on an ideological system which differs from his
explicit or conscious beliefs: he behaves <1as if>1 he still believed in
God. In old-fashioned terms it is the religion of the heart, not the
head. So I propose to call it "affective'<s4>s religion. However, what-
ever my terminology, behaviourists and other adherents of
scientistic creeds may still balk at my entertaining such a category
at all. In answer I would remind them that most religions include
ethics, which involve a value system, which I wish to study: and
in ethics the tensions between the cognitive ad the affective belief
system and value system are notoriously acute. Someone may tell
us quite seriously that he esteems a certain virtue, and yet act in a
contrary manner without apparent sense of guilt or incongruity.
In such cases we speak of unconscious hypocrisy, or say he is
deceiving himself (all concepts unacceptable to a behaviourist, who
is plainly, though unconsciously, not at home in politics). Nothing
depends on terms, and I shall be glad to use such everyday words
as "hypocrisy', but by using also the expression "affective ethics'
I shall try to keep their logical standing clear.
  Much confusion has arisen from statements which ignore these
distinctions. Let us take an example. A Catholic<s5>s theologian may
make the statement, "The souls of the dead do not return to this
world, but are in Heaven, Hell or Purgatory'; and other theolo-
gians may concur: i.e. the statement is theologically correct. A
corresponding empirical statement which looks correct is, "Accord-
ing to Catholic doctrine the souls . . . Purgatory'; the philologist
could produce chapter and verse. All too often, however, "Accord-
ing to Catholic doctrine . . .' is taken as equivalent to "Catholics
believe that . . .'. But if this substitution is made we have to ask,
"Which Catholics?' Our interlocutor may be a historian and

6

correctly specify a certain group. But if he is merely a non-specialist,
or even a theologian, he may rashly answer, "All Catholics.' Then
we have to jog his memory and ask, "What about ghosts?' Un-
doubtedly many Catholics admit to a belief in ghosts. This is their
cognitive position. Others may say they do not believe in ghosts,
but tell stories of experiences they declare to be inexplicable
in any other terms; the cognitive position of these people is incon-
sistent. Other professed unbelievers may refuse to sleep in what
they are told is a haunted bedroom--in that case their behaviour
is inconsistent with their cognitive position, and we may say that
affectively they believe that ghosts exist or at least might exist.
Having reminded ourselves of the importance of ghosts to many
Catholics, we may look back at the doctrinal position of Catholic
theology, and see that things are not quite so simple. There are
no ghosts in the Bible in the sense of spirits of dead people,<s6>s so
those who consider the Bible to be the foundation of their religion,
and yet believe in ghosts, are unorthodox or inconsistent by this
criterion. Yet in the Book of Common Prayer there is a form of
ceremony for exorcism, which is still occasionally used, and I have
established by questioning that Anglicans generally consider the
entities exorcised to be not devils, as happens in the New Testa-
ment, but ghosts, euphemistically known as "spirits of the
departed'. I have just mentioned the inconsistent position of a
Catholic who in effect states that he does and does not believe in
ghosts: is there a parallel inconsistency in doctrine, in Anglican
theology? Being no theologian I put this question to the Arch-
bishop of Canterbury at a public meeting in Oxford in December
1963. Referring to a recent exorcism performed by the Bishop of
Exeter which had been reported in <1The Times>1, I asked His Grace
whether Anglicans believed in ghosts. He replied that belief in
the spirits of the departed and their return to this earth was
neither accepted nor rejected by Christian doctrine. So even the
theological statement at the beginning of this paragraph was not
so impeccably correct as it seemed.
  The Archbishop's reply illustrates another point: while indivi-
duals are frequently inconsistent in their statements of beliefs,

                                                                 7

the theologies of the highly organized religions are rarely if ever
inconsistent on matters of importance. There are two ways, apart
from critically arguing the matter through to a (possibly negative)
conclusion, in which they avoid contradictions at difficult points.
One way is to declare the matter a mystery inaccessible to human
reason: this is a solution beloved of Protestants, but in contrast
not available to Therava$da Buddhists, who claim, as the Buddha
claimed, that their doctrines are founded entirely on reason. The
other way is to ignore the problem altogether, either by never
taking cognizance of it, or by explicitly stating that it is not their
concern to come to a decision. Inevitably there must be problems
of which any system of thought takes no cognizance; but these are
more likely to concern the philosopher than the empirical inquirer
whose subjects have never heard of them. Refusal by a religion to
come to a decision on certain points is a more important phenome-
non in our present context. The Archbishop's comment on ghosts
is one example. <1Mutatis mutandis>1 the same reply might have been
made by a Buddhist cleric: whether or not you believe in ghosts
is irrelevant to whether you are a "good' or orthodox Buddhist,
and the same incidentally is true of whether you believe in gods
(see Chapter 1). Perhaps the most celebrated exponent of the
undecided question was the Buddha himself. On one occasion<s7>s
he refused to answer a monk who asked him whether the world was
eternal, whether it was finite, whether the soul and the body are
identical, and whether an enlightened man in any sense existed
after death. The Buddha replied that he had never promised to
answer these questions; anyone who refused to lead the religious
life till they were answered would be like a man wounded by an
arrow who refused to have it removed till he knew the name and
caste of the man who shot it; his teaching was merely the practical
way to release from misery. At another times he condemned as
idle speculation a far longer list of metaphysical questions.<s9>s This

8

refusal to pontificate on what he regarded as inessentials was
what we might expect of a man who asked his listeners not to take
any doctrine on trust or authority but only to accept what they
themselves found reasonable; it certainly has been very influential
in the history of Buddhism, and goes far to explain why Buddhists
vary so much in some areas of belief and take little interest in
others: they are not covered by the Buddha's doctrine. Religion
provides a framework for interpreting reality. On certain points a
religion is specific; these we usually call dogmas; other points
are undecided. Religions are dogmatic on different matters, and
Buddhism has rather few dogmas altogether.


  I have so far discussed religion in general terms without
attempting to define it. This does not really matter, because every-
one knows what I have been talking about, and problems of
definition are essentialist problems, essentially trivial.<s10>s But I
cannot entirely fail to mention a matter so controversial. For a
long time it was stated or assumed that religion was essentially
characterized by belief in God or gods, but it was then objected
(e.g. by Durkheim)<s11>s that Buddhism, or at least Therava$da Bud-
dhism, lacked this belief, so on empirical grounds another defini-
tion seemed to be needed.  Moreover, on theoretical grounds
sociologists wished to broaden the definition to include institu-
tions. With this I have some sympathy. It seems to me an unneces-
sary duplication of terminology to focus on the belief or ideological
aspect of religion and so to define it that it coincides with theism,
the first meaning of which given in the <1Shorter O.E.D>1. is "Belief
in deity or deities'. So far as concerns Buddhism the whole con-
troversy has illustrated the confusion caused by not observing

                                                               9

the logical distinction between the cognitive and the affective level.
It is probably (arguably) not true that Therava$da Buddhism as a
philologist sees it, as a set of doctrines found in texts, involves
belief in gods. Although the Buddha himself believed, so far as
we can tell,<s12>s that supernatural beings exist, it can certainly be
argued, and is indeed believed by Sinhalese Buddhists, that this
is not a part of his teaching, any more than the proposition that,
say, animals exist. If we now inquire what Buddhists believe, we
find that most of them do indeed believe in supernatural beings.
But my landlord, a Sinhalese Buddhist villager, emphatically
stated that he believed in neither gods nor demons, nor did he ever
behave in such a way as to make me doubt his assertion. But he
called himself, and was admitted to be, a Buddhist. He believed
in the Buddha. What this involves will be discussed at greater
length in Chapter 3, where I will suggest that for Sinhalese the
Buddha is cognitively human but affectively divine. Unfortunately
even this does not settle the matter finally, because in this context
the categories of human and divine do not correspond to those of
natural and supernatural; it is not I think possible to translate
"supernatural' precisely into Sinhalese, for the Buddha is believed
to have had superhuman attributes which we, but not the Sinha-
lese, would consider contrary to the laws of nature. Whether a
definition which equates religion with theism holds on the cogni-
tive level is therefore controversial; the only level on which I think
it certainly holds is the affective: Sinhalese Buddhists behave, at
least with regard to the Buddha, as if they believe in a super-
natural being.
  Faced with these complications I for my part am content to
leave the definition of religion to the practitioners themselves.
Sinhalese Buddhists do call Buddhism (<1Buddha$gama>1) a religion
(<1a$gama>1), so I will not gainsay them;<s13>s similarly I would allow that
there was a religion wherever a group claimed <1bona fide>1 that it had
one. (This seems to be the logical result of adopting conven-
tionalism as distinct from essentialism.) I would merely note
that a religion combines certain systems, or structured sets of
associated phenomena, of which some systems are more important

10

in one religion and some in another. Dr. Percy Cohen's list
seems exhaustive : (1) cognitive system<s14>s (2) moral system (3) ritual
system (4) symbolic system (5) affective system (6) social system.
None of these systems are peculiar to religion (as Ruth Benedict
pointed out in her slightly different analysis), but their concur-
rence is probably shared only by a few political creeds, e.g.
Communism, which, apparently for this very reason, some
desire to call "religions'.
   What religion, however, does not include is magic, and I had
better explain here how I propose to demarcate magic from
religion on the one hand and science on the other. Very simply.
Science admits of falsification, which religion and magic do not.
The difference between magic and religion is that between a spell
and a prayer: magic is a technique or technology, which is to say
that if correctly performed it must effect the desired result;
religion requires aid which must be asked for and cannot be
compelled. Let me reiterate that this is no definition or exhaustive
description of religion, but merely an <1ad hoc>1 distinction--a state-
ment of where I consider magic to stop and religion to begin: the
Sinhalese Buddhist monk in his own language does not consider
that it is "religion' to ask for anything.
   Since writing the above I have read with admiration Professor
Melford Spiro's sensible and witty paper, "Religion: Problems of
Definition and Explanation'.<s15>s Spiro is led by his acquaintance
with Therava$da Buddhism in Ceylon and SE. Asia to remark that
Buddhists are not in fact usually atheists. Moreover he points out
that even if they--or other large groups of people--were atheists,
this need not debar us from defining religion in theistic terms, as
there is no <1a priori>1 reason for making the definition so broad as to
cover everyone. In fact a narrower definition may be more inter-
esting, and prompt research. Spiro thinks it useful to have a work-
ing definition of religion--and for his purposes it is, as he wants
to discuss religion in general. (He shows himself to be, like Popper
but unlike most famous anthropologists, both a deductivist and a

                                                            11

conventionalist.) He finally (p. 96) defines religion as "an institu-
tion consisting of culturally patterned interaction with culturally
postulated superhuman beings', a more precise, if less elegant,
reformulation of the old-fashioned view. If a definition of religion
is called for I am glad to accept this one, with the small modifica-
tion that I would except magic by the criterion explained in the
previous paragraph.


  Having explained my own concepts I had better say why I am
interested in religious change (and, as a corollary, in religious
conservation), and how I propose to deal with it. It is difficult to
establish causality within the field in which the social sciences are
interested, because chicken-egg situations are so often cropping up ;
the association of two phenomena is noted, but in the absence of
controlled experiments it is hard to say which has causal primacy.
Perhaps this is one reason why theoreticians of the social study of
religion have been much preoccupied with the problem of the
origin<s16>s of religion itself.  Unfortunately the origins of most
religions, and therefore of religion itself, are lost in the past, so
that answers to this historical question must be speculative and
untestable. Thus the problem seems to be insoluble ; the apparently
divergent answers have really been answers to different questions.
Intellectualist answers like Tylor's have in fact been attempts to
explain the origins of belief systems; sociological answers, like
those of Durkheim and Malinowski, have been attempts to explain
the origins of moral and social systems; etc. True, each theoreti-
cian laid claim to explailling religion as a whole by asserting the
causal primacy of the system they had explained; Ruth Benedict,
for example, inherited from Marx and Durkheim the theory that
beliefs were mere rationalizations of, i.e.  caused by, religious
behaviour. All these claims were demonstrably exaggerated, and
there seemed to be no reason for giving one system any primacy
over the other; rather they came to be viewed as potentially com-
pletely independent, as in Cohen's formulation above. Recently,
however, there has been a resurgence of intellectualism in a tradi-
tion which owes most to Tylor among anthropologists and Weber
among sociologists. This tradition focuses on the idea that people

12

often really do things--perform ceremonies, found and support
institutions, undertake moral acts--because of what they believe
to be the case. At the same time researchers have realized that
while we cannot get back to the origins of religion itself, we can
look at the origin of bits of religion, i.e. <1religious changes.>1 When we
have historical evidence for beliefs in former times, we may be
able to document changes in a belief system; and when we know
about a change in a belief system we can speculate on its causes.
This is why it seems important to study religious change and
why such change can best be studied in societies long literate, and
best of all in societies following religions based on books.
  It may seem paradoxical that the social study of religion should
concentrate on the rational, but Popper, the begetter of what he
calls the "rationality principle', has shown that there is no other
way. The rationality principle is the zero principle, the basic
assumption of the social sciences; it is the principle that in so far
as behaviour is amenable to social study, it is the behaviour of
people acting in what they conceive to be their best interests. Its
application to the sociology of religion has been expounded by
I. Jarvie in his interesting and provocative book <1The Revolution in
Anthropology>1.<s17>s If we see a man in an agricultural comunity
oppressed by a drought enter a special building (called, maybe, a
church), get on his knees, fold his hands and start muttering, we
may assume that he is insane, in which case we can offer no
further testable explanation of his conduct, or that he thinks his
actions may bring rain. And if that is what he thinks, we must
assume that he thinks some entity is capable of giving him rain,
and that that entity, which begins to look rather human in its
emotions, will be placated by his suppliant posture and humble
gesture, and listen to his words. The rationality principle refers to
rationality in behaviour in a situation (which includes a set of given
beliefs). It does not apply directly to the situation, or to the beliefs
included in the situation, although the question of the rationality
of those beliefs may be raised: a belief is rational if it is a good
solution, in a given intellectual situation, of some problem it tries

                                                                13

to solve. Conduct is rational if it consists in using what seem
appropriate means to attain given ends: the man praying for rain
is acting rationally. What are given to him are the beliefs that
make those means seem appropriate and the ends he wishes to
attain; in this case that rain should fall. The sum of all the ends a
man considers desirable is his value system; the sum of his beliefs
is his belief system. Ultimately values cannot be derived from facts
(nor facts from values) ; but frequently a man's beliefs about facts
ad his values have a common historical origin. Some of his
beliefs about facts he gained from his own experience, and many
of these he holds critically. The rest of his beliefs, both about
facts (what is the case) and about values (what is desirable) he
derived from other people, usually his parents and teachers; this
of course applies also to his beliefs about mundane and trivial
things, but important for us are only his beliefs about metaphysics
and ethics, which together constitute his religious belief system,
and which are all untestable.
  Social changes come about as the unintended or the intended
consequences of the actions and utterances of individuals. Religious
change is not often an intended consequence: when it is, the
individual who brings it about is usually said by sociologists to
have charisma, a term invented by Weber which stands in lieu of
an explanation. However, even these "charismatic' individuals
bring about many changes which they do not intend, and the
greater their influence the more unintended consequences of their
actions there are: Jesus Christ is surely not accountable for the
whole of Christianity.  Most social changes therefore are the
unintended consequences of other social changes. Moreover, the
exceptions just mentioned need not worry us unduly, for people
only become charismatic <1a posteriori,>1 because their ideas have
affected other men.<s18>s (This shows that the sociological concept of
charisma is quite distinct from the Christian theological concept
of a special favour vouchsafed by God,<s19>s which is the original
meaning of the word. The theological concept is unaffected by

14

my argument, as the grace of God does not depend for its existence
on human recognition.) Even a religious innovator must offer
people something they want; a man who preaches suicide will
only be listened to in times of exceptional misery. This shifts the
question back to what people want, their ends, aims, or values,
the sum of which we call a value system.
  The relation to each other of values in a value system may be
hierarchic or preferential. In a hierarchy of values only certain
things are desirable for their own sakes; and other things are good
or desirable as means to these higher-level values.<s20>s Preferences,
on the other hand, have no rigid structure (there is an infinite
number of possible preferential choices), and, especially at the
lower levels, tend to be <1ad hoc>1. Just now I prefer strawberries to
raspberries, an <1ad hoc>1 preference which may change by the next
meal; but unless I am unusually gluttonous or misanthropic I
always prefer friendship to good food. Even among hedonists only
the very sophisticated will claim that they value friendship only
as a means to happiness, and in practice it is highly doubtful
whether anyone goes through life making these calculations (i.e.
"If I help that friend will I increase my happiness?'), so the value
attached to friendship in our society does not seem to be hierarchi-
cally subservient to any higher-level aim or value. Therefore it is
safe to say that most contemporary western intellectuals have a
plurality of values. However, the value systems of most religions
are doctrinally, and perhaps cognitively, hierarchically structured,
at least at the higher levels. A Christian may express the view that
the attainment of Heaven is the <1summum bonum,>1 the one highest-
level aim to which all other aims are subservient. Below Heaven
in this hierarchy he might put the three cardinal virtues, faith,
hope, and charity, and all other values would be subsumed under
these. (A modern Protestant might say that these virtues too are
valued "for their own sakes'; but as the practice of these virtues
must by Christian doctrine invariably lead towards Heaven this
does not destroy the hierarchy.) This forms an interesting contrast
to Buddhism. A cornerstone of Therava$da Buddhist doctrine is
that <1nirva$n*a>1 is the <1summum bonum.>1<s21>s As <1nirva$n*a>1 is a kind of

                                                                 15

extinction it precludes the practice of virtues (it is beyond good
and evil) such as charity, which is admitted as a means, a sub-
sidiary goal, but not as part of the end itself. Some early Buddhists
obviously found this unsatisfactory: they developed the doctrine
of the Bodhisattva ideal.  In contrast to the Pratyekabuddha
("Enlightened singly'), who seeks <1nirva$n*a>1 for himself alone, the
Bodhisattva ("He whose essence is enlightenment') takes a vow
to attain <1nirva$n*a,>1 but once it is in his grasp pauses on the brink
and will not enter until all other creatures are similarly advanced.
Out of his infinite compassion he denies himself the supreme bliss,
but works in the world to bring others towards it. Love has
replaced <1nirva$n*a>1 as the ideal. This at least is the accepted view of
the distinction between Therava$da and Maha$ya$na ethics ; whether
it is correct will be considered in the last chapter.
  The high-level aims shared by members of a community we call
the ethos of that culture. (Like the aims of which it is constituted,
the ethos may be discussed on either the cognitive or the affective
level.) Unfortunately ethos is so general a thing that it is difficult
to observe it in the field and perhaps impossible to describe it in
testable and refutable statements. The best we can usually do is
to characterize it by negative example: an English gentleman
would never kick a dog or a lady. At this point anthropology has
become an art rather than a science; but that need not deter us.
  We can now see <1what>1 changes when a religion changes. A
religious action is based on a belief about facts and directed to
certain aims. Either the belief or the aims may change. What does
not change is the relation between these two: that the action
based on the beliefs and directed to the aims follows the rationality
principle. If it does not it is not amenable to systematic study.
We can also see <1why>1 the religion changes: as the consequence,
usually unintended, of the changed situation which has been
brought about by individuals. The types of explanation I am
therefore looking for will become plain if I quote, as briefly as
possible, three examples from my field-work.
  The first is an example of changing beliefs. There is a widely

16

held hypothesis, to which I also subscribe, that many of men's
beliefs about the invisible world mirror their experience of this
world. This applies especially to the power structure. Central
Ceylon a couple of hundred years ago had, in the precise sense, a
feudal social structure, and numerous political divisions and sub-
divisions. The gods were conceived on the same pattern: each god
was the main god, i.e. overlord of a certain territory, and had a
retinue of lesser gods who were in charge of lesser territories
within his; these in turn had their retinue, and so on down to
village level. When the British conquered central Ceylon in 1815
they at first left the structure of authority unchanged at all but the
highest levels, merely substituting government officials for local
noblemen, and imposing a far more comprehensive central
control. The correspondence between the divine and the human
political structure is explicitly recognized by some Sinhalese
themselves--in the modern terminology: an old monk with no
western-style education told me in so many words that a certain
god was like the Government Agent (G.A.), a lesser one like the
District Revenue Officer (D.R.O.), etc. However, an unintended
consequence of the British conquest, with its abolition of feudalism
and its centralization of authority, as well as of improved com-
munications within the island, is that local gods are losing their
authority, i.e. their interest for the villagers, to the extent that
some of their shrines are falling into disuse; and even of the four
gods who were said to have jurisdiction over the island three are
greatly declining, one greatly increasing in authority, so that
political centralization may eventually bring about a local mono-
theism. This trend has probably been hastened by the fact that
since Independence politicians have deprived the administrative
civil service of much of its power and autonomy; even the G.A.
(the senior administrator of a province) counts for less than he did
under the British. A more up-to-date villager, telling me of the
gods' arrangements, said that they have a sort of parliament.
  My second example is one of changing aims. It was the aim of
the Buddha to attain <1nirva$n*a,>1 a mystical release from normal states
of consciousness, it is attained in life, and someone who has
attained it is not reborn: he escapes from the wheel of rebirth,
to which all creatures are tied. Rebirth the Buddha considered
misery; the peace of <1nirva$n*a>1 was the only good worth having.
But most Sinhalese villagers do not want <1nirva$n*a>1--yet. They are
            17

like St. Augustine who prayed "Make me chaste and continent' O
Lord--but not yet.'<s22>s They say they want to be born in heaven;
some of them would even like to be reborn in a favourable station
on earth. They fear, as their Teacher did not, the extinction of
sensation, and want to go on feeling. Moreover, though they will
probably utter the impeccably orthodox sentiment that life is
sufferlng, most of them plainly do believe--and say so when they
are not speaking in a religious context--that people with wealth
and power are happy. This shifting of an aim, from <1nirva$n*a>1 to
heaven or even to earth, is thus explicable in terms of higher-level
changed aim--wish to go on living, and a changed belief--that
rich people, or at least gods, are happy. The next step is to inquire
why the higher-level aim and the belief have changed; but this
would take me too far afield just now. An unintended consequence
of the Buddha's teaching might be that it made life seem worth
living!
  My third example is an illustration of what might be a changing
ethos; it is very similar to one given by Jarvie (p. 114). Many
Sinhalese with the aim of temporal advancement have changed
their religion in the last 500 years. They became Catholics to
avoid being murdered by the Portuguese, Protestants to placate
the Dutch, Protestants again to please the British. Recently
many Christian Sinhalese have been changing back to Buddhism
for similar reasons. However, once a man has changed his religion,
he and his children are exposed to the whole ideology of that
religion, so that gradually the entire ethos of the family may
change. In this case a man's ethos, i.e. highest-level aims, have
changed as the unintended consequence of his, or his father's,
rational pursuance of lower-level aims.


                          II. WHY CEYLON?

  I have said why I think religious changes are worth studying,
what they are, and in what terms I propose to explain them; but
I have not yet said why I chose to study Sinhalese Buddhism. The
methodological reasons are these. I mentioned a while back that to
study religious change we have to be quite sure what the present
religion, which we can now see, has changed from; and as the
detailed and systematic observations of social scientists are a

18

recent event in history our best hope lies in studying religions
based on book; i.e. scriptures. I am not confusing history with
philology: I do not claim that these books tell us what the mass of
adherents to the religion at that time believed. But they do tell
us what their authors believed. In some celebrated cases, e.g. the
Bible, the authors of the scripture were unfortunately not the
founders of the religion, but for our problem this does not matter
too much, as it is Jesus Christ's doctrine as reported in the Bible
which has been influential. All the religions usually reckoned as
"world religions' have been founded on books. Further, a great
number of splinter groups within religions, known to sociologists
as sects, have been founded on books, especially in western society.
Some of these, notably Protestant sects, have been admirably
studied by sociologists since Weber: Bryan Wilson's book <1Sects
and Society>1<s23>s is an illustrious example. However, the study of
sects has its own problems, which are not mine: I am interested
rather in the broader religions, which sociologists call churches,
which are found in societies at large. The problems of studying
Christianity in any given society are especially complex, as no
religion has moved through so many cultures and undergone so
many upheavals. I do not know Hebrew, Arabic, or Chinese, so I
cannot adequately study the doctrines of Judaism, Islam, Taoism,
or Confucianism. The problems of Hinduism are again peculiar:
although there are scriptures the authority of which all must
acknowledge, namely the Vedas, they are so ancient, their scope
is so limited--for they are a collection of hymns and sacrificial
formulae--and their interpretation so obscure that their relevance
to what people believe is negligible; moreover, only a minority of
Hindus have been allowed access to them. There are innumerable
other texts, but some, like the Upanishads,<s24>s are esoteric, while
only a very few possess an authority not confined to certain sects
or strata of the population;<s25>s to make matters worse, the chron-
ology of Hindu scriptures is totally obscure.
   Buddhism has the advantage of a well-established chronology.

                                                               19

We know that the Buddha lived in NE. India and died there at the
age of 80 after preaching for forty-five years. It is fairly certain that
he died in 486 or 483 B.C., which would mean that his Enlighten-
ment, which is when Buddhism can be said to have started, took
place in 531 or 528 B.C. The Buddhist era, however, is dated from
the Buddha's death. But by the Buddhist era, used in Ceylon
and thence adopted by other Buddhist countries, the Buddha
died in 544/3 B.C., some sixty years earlier. The discrepancy is
believed to have arisen in the eleventh century, the dark ages of
Ceylon, when Tamil invasions smashed the ancient civilization
ad many cultural traditions were lost.  Of the two kinds of
Buddhism, Therava$da and Maha$ya$na, the latter is very hetero-
geneous : moreover the only Maha$ya$na country in which field-
work is now practicable is Japan, the furthest from the India
sources and the last to be converted. Therava$da Buddhism has the
further advantage of a unitary body of acknowledged scripture,
the Pali Canon; Pali is an ancient Indian language closely related
to Sanskrit. Three countries, Ceylon, Burma and Thailand, are
the principal bearers of the Therava$da tradition. Of these Burma
and Thailand were once Maha$ya$na countries; they adopted
Therava$da through contact with Ceylon, notably in the eleventh
ad twelfth centuries.<s26>s Ceylon has further advantages from my
point of view. It is an island, so that the influences which have
entered it can be comparatively well isolated and identified. It has
a written chronicle, of varying but sometimes considerable
excellence, which enables us to follow the fortunes of the Sinhalese
people from the very start, or at least from their conversion to
Buddhism in the third century B.C., and which, being the work of
monks, contains much information on Buddhism. Most of the
commentaries on the Pali Canon were composed in Ceylon, which
in ancient times was a centre of Buddhist learning. (The chronicle
and commentaries will be discussed in Chapter 1.) The Sinhalese,
who form the great majority of the inhabitants, have always been
preponderantly Buddhists, so that in an important sense there is
a uninterrupted tradition of over 2,000 years.<s27>s Finally, Sinhalese

20

is a language descended from Sanskrit and Pali, so that any con-
ceptual shifts of terms used in the texts in those languages should
be fairly easy for me as a Sanskritist to spot, being undisguised
by translation; moreover, I expected I could learn Sinhalese more
easily than Burmese or Thai, which are structurally unrelated to
Sanskrit. So Ceylon it was; and this work is about Ceylonese
Buddhism.<s28>s To explain why I chose to study it on the ground
in the way I did, I must first give a brief sketch of Ceylonese
history, geography and economy, which will further serve as a
background to my findings.


III. THE HISTORY, GEOGRAPHY AND ECONOMY OF CEYLON

   The history of Ceylon can conveniently be divided into two
main periods, ancient and modern; with the dividing line drawn
at 1505, when the Portuguese arrived in the island as the first
European colonial power. The ancient period can then be sub-
divided into two unequal parts, the Anura$dhapura and the
Polonnaruva periods, which are named after the capital cities of
the time. The Anura$dhapura period is from the fourth century
B.C. to the late ninth century A.D. ; the Pol*onnaruva period properly
ends in the thirteenth century, after which there is another period
of political instability, the capital shifting from place to place.
Although the Anura$dhapura period is always recalled as the golden
age of the Sinhalese nation and Sinhalese Buddhism, many of the
distinctive institutions of modern Sinhalese Buddhism came into
being during the Pol*onnaruva and immediately subsequent
periods.<s29>s For most of the modern period the island was politically
divided in two, and the political division acquired cultural impor-
tance. The European powers ruled only in the coastal areas, and
were based on Colombo on the west coast, with important ports

                                                                  21

at Jaffna and Galle, at the northern and southern ends of that
coast. The Portuguese were ousted by the Dutch in 1658, the
Dutch by the British in 1796. Meanwhile in the mountainous
interior the Kandyan kingdom maintained its independence till it
surrendered to British treachery in 1815. The British restored
independence to a unified Ceylon in 1947.
   Independent Ceylon has preserved the British administrative
division of the island into nine provinces, each divided into two
or (usually) three districts. Most of these divisions, even down to
district level, are good approximations to the boundaries of sub-
cultures in this amazingly diverse island, but only a few need
concern us here.
   The four principal cities are still Colombo, Kandy, Jaffna, and
Galle. The Central Province, in the centre of which is Kandy,
was the core of the Kandyan kingdom, which comprised roughly
this province and the four provinces adjoining it. Kandy is called
in Sinhalese simply "The City' (<1Nuwara>1). The English name is a
corruption of the Sinhalese phrase "On the mountain' (<1kande$>1 or
<1kande$ ud*a>1). It lies at 1,600 feet above sea level in the midst of very
hilly country. To the south of Kandy the mountains are far
higher; the plateau at Nuwara Eliya, only 25 miles as the crow
flies and 48 miles by road, is at 6,000 feet. The passes down to the
coastal plain from this eminence are precipitous and spectacular;
those to the west and east of Kandy are hardly easier--hence the
city's impregnability. To the north, however, the hills decline
rather more gradually till they give way to the plains of the North
Central Province, in which lie both Anura$dhapura and Pol*on-
naruva. This area was the principal seat of the ancient civilization,
but after its devastation by the Tamils it relapsed into sparsely
populated and malarial jungle, in which condition it remained till
this century, when archaeological interest, nationalist sentiment,
and above all the need for cultivable land led to systematic jungle
clearance and restoration of the ancient irrigation system.
Although the Europeans made Colombo the focal point of all
communication systems, so that the east coast is now the most
backward and inaccessible, in ancient times this was not so: it
was on the east coast that Mahinda landed when he brought
Buddhism to Ceylon in the third century B.C. Perhaps the last
famous visitor to arrive from that side was Robert Knox (see
below) in 1660, when the west coast was already developed, but

22

in European hands. A lesser centre of the ancient civilization was
in the now totally undeveloped south-east: it was to here that
the Sinhalese had to retire when beaten by Tamil invasions, and
from here that the Sinhalese national hero, Dut*uga%mun*u, marched
up the east coast, some way inland, to defeat the Tamils--a march
he would have difficulty in repeating today.
   Village agriculture is concerned mainly with the growing of
rice, locally known as paddy, which is the staple food. In the five
months which it takes to grow, rice requires to stand in non-
stagnant water. There are two monsoons, the SW. from May to
September and the NE. from October to January. Little rain from
the SW. monsoon clears the central mountain massif, so that the
north, the eastern coast and the south-east depend on the rain
which falls in three months. The Sinhalese who lived in these
areas in ancient times built huge reservoirs (<1va%va>1), locally known
as tanks, and irrigation canals (<1a%la>1), which are justly considered a
wonder of the ancient world: modern technology cannot improve
on them. The type of paddy cultivation which depends on such a
tank has been exhaustively described by E. R. Leach in his study
of a village in the North Central Province.<s30>s In the west, south-
west and centre of the island, where there is a likelihood of rain
for nine or ten months in the year, no such system of water
conservation is necessary, although there used to be some tanks
of unknown antiquity, even in the hills round Kandy; one in the
village where I worked had its bund (dyke) permanently breached
in the eighteenth century. Much of the land in this wetter half of
the island is now used as plantations of tea, rubber and coco-nuts,
the cash crops on which the country's exports depend. Rubber
and coco-nuts are commoner in the low country, while the higher
the hills the better the tea. Most of the higher land in the centre
of Ceylon was appropriated by the British in the nineteenth
century to plant fist coffee and then tea; as a labour force they
imported Tamils from South India, whose descendants and
relations still work on them today. The cultural isolation of the
plantations on the slopes from the villages in the valleys is amazing,
though it has been exaggerated by politicians: I found few
Sinhalese villagers in the area who could not speak at least a little
Tamil, and no doubt the reverse is also true.
   Most of the hills in the Kandy District of the Central Province,

             23

where I worked, are not so high nor the valleys so steep as in the
most famous tea country, but there is hardly an inch of flat
ground. This and the lush vegetation amid which the houses stand
make for rather diffuse settlements, in which often one house can
barely glimpse the next through the trees; sometimes a small
group of cottages forms a little hamlet' maybe even with its own
shop, as a distinct unit within the larger village. Since the British
eradicated malaria in the Second World War population pressures
have been very severe--31/2 per cent per annum growth for the
country as a whole. Thus the rice-growing villages of farmers, some
landless and all under-employed, sprawl out so that the boundaries
between one and the next are often hard to discern, and political
or administrative units may not coincide with traditional parish
limits. A village (<1gama>1) is therefore capable of various definitions,
but by traditional criteria it consists of an area of land, an estate,
all of which originally, in fact or fantasy, belonged to the families
of the founders of the village. It still consists in a fairly compact
body of land holdings (with bits taken out by estates, roads, or
government colonies), but many of these are nowadays owned by
absentee landlords. These holdings are part rice fields (<1kum#buru>1)<s31>s
and part dry land (<1god*a>1), which is usually known in English under
the rather misleading title of "village gardens'. In fact this <1god*a>1 is
not cultivated, but on it grow all the types of fruit a villager uses;
on it the livestock graze, and houses are built. The commonest
plants provide coco-nuts (<1pol>1), which are the basis of all local
curries, king coco-nuts (<1ta%m#bili>1), many kinds of bananas (<1kesel>1),
jak fruit (<1kos>1), bread-fruit (<1del>1), betel leaf (<1bulat>1) and areca nut<s32>s
(<1puvak>1) which are chewed together, limes (<1dehi>1), pineapple
(<1anna$si>1), various condiments and spices which are in daily use,
and finally the kitul palm. From the sap of its flower is made a
delicious brown sugar called jaggery (<1hakuru>1); no wonder the
villagers despise the sugar-cane (<1uk gas>1) which also grows wild.
(Nowadays it is cheaper to buy ordinary imported sugar.) This
same sap, like that of the ordinary coco-nut palm, can be drunk

24

at various stages of fermentation, as toddy (<1ra$>1) or arack (<1arakku>1).
Very rarely a house may have a cultivated vegetable garden, but
most vegetables are grown in a <1he$na>1, a casual cultivation of other-
wise unused land, which is usually cleared by burning. The <1he$na>1
used to be cultivated--illegally--on Crown lands, but the scope
for this is nowadays much restricted, and there is very little unused
land to be had even for rent, so most vegetables are bought at the
market, by those who can afford them. The most important
vegetables, from the villagers' point of view, are chillies (<1miris>1),
onions (<1lu$n*u>1), and lentils (<1parippu>1). The livestock in a Kandyan
village is mostly chickens (which supply eggs and meat) and goats
(which are eaten but not milked) ; water-buffaloes are used to draw
ploughs and are never killed, but their milk is sometimes used for
making curd, and is served with jaggery to monks and other
honoured guests. Other cattle are rare (on beef-eating see Chapter
6, pp. 261-2), and for milk, which is always used in tea, most depend
on tins of milk powder, or government supplies. Cheese is not
eaten, butter an exotic luxury. Pigs are almost unknown in the
hill country, and most villagers have never seen a horse or a
sheep. Nearly every house has a dog, and cats are common; those
who can afford to feed their domestic animals tend to give them
rice and other human foods. A rich man in the village where I
stayed owned an elephant, which was tethered to a tree outside
his house when not working for hire.
  Most houses are solidly built of cemented bricks covered with
plaster and whitewash. They have cement floors. Windows are
always barred, usually have shutters and occasionally glass. Roofs
are of red tiles or corrugated iron sheeting; ceilings are rare.
There are internal walls, partitioning the building into three or
four rooms; older or grander houses may have an open or half-
open verandah in front. Almost all houses are bungalows. Kitchens
are usually in a lean-to against the back of the house; cooking is
on wood fires, and there are no chimneys. Every house has a
separate earth privy in a building of its own. The very poorest
houses are of wattle and daub, with roofs of cadjan (plaited palm
fronds). (This is much commoner in the dry zone.) There are
cemented wells which the local council builds for drinking water,
and more informal unwalled wells for bathing, though river
bathing is preferred as more fun.
  The rice fields are terraces formed from the hillside; each step

                                                                 25

is surrounded by a low mud dyke (<1niyara>1) just wide enough to
walk along. Small gaps (<1vakkad*a>1) in these dykes keep the water
moving down the hillside. The rice is a cheerful light green. The
valleys with streams sparkling through them; above these the
dark and varied vegetation dotted with red tile roofs; above these
again the hills terraced in shimmering yellowy green, lined with
mud paths along which move a few figures in white and bright
colours; further up still the higher hills, their redder earth almost
hidden by dark green tea bushes; and at the back of all the
mountains blue in the distance: this vista, combining natural
grandeur and comprehensible humanity, is the most beautiful I
know.
    A small but important feature of such a landscape is likely to
be a cluster of buildings round a huge peepal tree and an object
shaped like a dome with a spire, between 10 and 40 feet high,
washed white or pale blue. The Bo tree, under which the Buddha
attained Enlightenment, and the stu$pa, which contains a relic,
indicate the presence of a Buddhist monastery, where dwell one
or more monks and novices, dressed in bright robes of yellow,
orange, or even red. These people and institutions are repre-
sentatives of an ancient religious tradition; let me now attempt
to summarize its history.


  The main primary source for the history of Ceylon in general
ad the Sinhalese in particular is the <1Maha$vam*sa,>1 a Pali verse
chronicle composed in four instalments, the first in the sixth
century (Anura$dhapura period), the second in the thirteenth, the
third in the fourteenth and the last in the eighteenth century.<s33>s In
the European edition (ed. Wilhelm Geiger)<s34>s only the first part is
called the <1Maha$vam*sa:>1 the three latter parts form the <1Cu$lavam*sa;>1
but the consecutive numeration of the chapters is preserved. The
whole work is written by Therava$da monks, so naturally it is
especially concerned with religious history; indeed, its identifica-
tion of the fortunes of the Sinhalese with those of Buddhism is
the main intellectual source of modern Sinhalese Buddhist
nationalism. The quality varies, especially with the distance of

26

events described from the times of writing, but archaeological
corroboration fills us with admiration for the oldest part; although
it contains many miraculous incidents these are mostly to be
regarded as embellishments of the true events. For the ancient
period there are many other literary sources in Pali and Sinhalese,
notably histories of relics; also much epigraphic and other archaeo-
logical evidence. For the modern period there are also literary
sources by Europeans; justly the most celebrated is the first book
ever written on Ceylon in English, maybe still the best and
certainly the most entertaining: Robert Knox's <1An Historical
Relation of the Island of Ceylon>1, first published in 1681.<s35>s Knox
(1641-1720) was a merchant seaman whose ship put into harbour
on the east coast of Ceylon after a storm late in 1659; in 1660 he
was captured with his father the captain, and several companions
by King Ra$jasimha II (1635-87), who collected European
prisoners, whom he maintained in comfort and honour in various
villages--probably over a thousand of them in the course of his
reign. Knox made good but never settled down, and escaped in
1679. Religion is the facet of Sinhalese culture he seems to have
understood least well--he always kept his Bible with him--but
even so his information is invaluable to us.
  The outstanding secondary source for the history of Ceylonese
Buddhism is the Rev. Walpola Rahula's <1History of Buddhism in
Ceylon>1,<s36>s but unfortunately, as the sub-title tells us, this covers
only the Anura$dhapura period. Space does not allow me to discuss
this invaluable work, which should be consulted by all serious
students of Buddhism. The Rev. Rahula reports the conclusions
of two articles on heterodox traditions by the highest authority on
Ceylonese antiquities, Professor S. Paranavitana, "Pre-Buddhist
Religious Beliefs in Ceylon'<s37>s and "Maha$ya$nism in Ceylon'.<s38>s He
also makes full use of Professor G. P. Malalasekera's <1The Pali
Literature of Ceylon>1,<s39>s which deals with this aspect of Buddhism
up to modern times. Unfortunately there are no other secondary

                                                                  27

sources of comparable usefulness available for the post-Anura$dha-
pura period, i.e. the last thousand years; Wilhelm Geiger's <1The
Culture of Ceylon in Mediaeval Times>1<s40>s (which appeared post-
humously, edited by Heinz Bechert), is a fine harvest of facts
gleaned from the chronicles, but its synchronic approach obscures
the course of development, and on Buddhism it is altogether
disappointing.
  The Sinhalese nation was allegedly founded by Vijaya, the
grandson of a lion (Skt.: <1sim*ha>1), a ne'er-do-well from NW. India
who landed in Ceylon on the day the Buddha died. The island
was at that time inhabited only by spirits and demons. More
prosaically, it was probably inhabited by Va%ddas, a people
athropometrically and culturally distinct from the Sinhalese; a
few Va%ddas can still be distinguished among the inhabitants of
the jungles of eastern Ceylon. Anura$dhapura was founded by
Panduka$bhaya<s41>s in the fourth century B.C.  His grandson,
Deva$nam*piya Tissa, was a contemporary of the emperor of
India, Asoka (ruled <1c>1. 269--<1c>1. 230), who was a convert to Bud-
dhism. Asoka sent religious ambassadors to countries as distant
as Egypt and Macedonia, and after the third Buddhist council
(according to Therava$din reckoning) monks were sent to all the
outlying regions and border countries round India to preach
Buddhist doctrine. To Ceylon was sent one of Asoka's own
sons, Mahinda. According to the chronicle he was accompanied
by four other monks, a novice, and a lay disciple. A single monk
pe1forms the lower ordination ceremony, by which a layman enters
the Order as a novice, but five monks are the quorum needed to
perform the higher ordination ceremony (<1upasampada$>1) by which
a novice becomes a monk; if such a quorum cannot be mustered
no ordination can take place, the line of succession is broken, and
after a time there are no more monks. Mahinda's group was
therefore the smallest possible paradigm community of male
Buddhists. The Buddhism which Mahinda brought was of the
Therava$da school, which has its scriptures in Pali; these scrip-
tures, the Pali Canon, were at that time not yet written down, but
Mahinda and his followers presumably brought them in their
heads. Tradition has it that they thus brought the commentaries

28

too, and that Mahinda translated these into Sinhalese. Mahinda
received as a gift from the king, who was the first Ceylonese
convert, a park in Anura$dhapura in which to found the Maha$vi-
ha$ra ("Great Monastery'); he then ploughed a furrow round the
city as a parish boundary (<1si$ma$>1), and by this token established
Buddhism in Ceylon.<s42>s The Maha$viha$ra was to become the bastion
of Therava$din orthodoxy for more than a thousand years, and it
is in this tradition that Ceylonese monks, especially those of the
Siyam Nika$ya (see below), stand today. The commentaries and
chronicles which survive from the Anura$dhapura period were
probably all written there (see Chapter I). Mahinda had his sister
Sanghamitta$, who was a Buddhist nun, bring from India a branch
of the sacred Bo tree under which the Buddha attained Enlighten-
ment, and it was planted with great ceremony, and saplings

                                                                29

distributed over the island. Sanghamitta$ also established in Ceylon
the Order of nuns, for only a nun can admit a woman to the
Order. Asoka sent as relics the Buddha's begging bowl and various
of his bones, of which the right collar-bone was enshrined in
Ceylon's first stupa, the Thu$pa$ra$ma. Both the tree and the stupa
stand in Anura$dhapura to this day.
In the second century B.C. Tamil invaders from South India
ruled in Anura$dhapura over the Sinhalese, who were apparently
already Buddhists. There was a Sinhalese kingdom in western
Ceylon, called Dakkhin*adesa,<s43>s and a third in the south-east,
called Rohan*a. From Rohan*a a national revival was led by
Dut*t*haga$man*i$ (101-77 B.C.), who proclaimed that he was fighting
not for a kingdom, but for Buddhism and put a relic in his spear;
monks were in the army, though they left the Order for the
occasion, and the war ended when Dut*uga%mun*u, as he is known
in Sinhalese, slew the Tamil king in single combat at the gates of
Anura$dhapura.  These events are the recorded beginning of
Sinhalese Buddhist nationalism. From now on Sinhalese kings
were regarded as defenders of the faith, and Buddhism acquired
official trappings. Dut*uga%mun*u erected in honour of Buddhism
enduring monuments of his reign. But within forty years of his
death Ceylon suffered civil war, invasion, and a famine so severe
that people resorted to cannibalism. The Maha$viha$ra was aban-
doned and many monks fled to India. Those who were left feared
that the oral traditions would be lost, and so about 30 B.C., for
the first time in history, they wrote down the Pali Canon, with
commentaries (presumably in Sinhalese). King Vat*t*aga$man*i$-
Abhaya (27-17 B.C.), who reconquered Anura$dhapura from the
Tamils, gave a new monastery called Abhayagiri to a monk from
another part of the country who had helped him; but this dis-
tressed the monks of the Maha$viha$ra, who expelled the monk.
However, he and his followers refused to accept the expulsion,
and thus arose the first schism in the Ceylonese Sangha. Within a
few years this schism acquired a doctrinal rationale: the Abhaya-
giri monks accepted the teachings of Dhammaruci, a monk of the
Vajjiputra sect in India, who taught for instance that there is a

30

personal entity which survives transmigration (<1puggalava$da>1), as
against the contrary theory (<1anattava$da>1) of the Therava$dins.
Abhayagiri monks then studied both Therava$da and Maha$ya$na,
while the Maha$viha$ra monks were conservative and regarded
them as heretical. Kings favoured now one, now the other, often
both and occasionally neither.
Late in the third century Maha$ya$na monks came to Ceylon from
India. Their teachings<s44>s were suppressed, but found favour at
Abhayagiri in the reign of King Maha$sena (334-61), who was
himself converted to them. At one point he even had the Maha$vi-
ha$ra demolished and the premises sown with beans, but he later
repented. In his reign is the first mention of a Bodhisattva image,
which the king had made by his son, who was skilful at carving
ivory; in his reign too the famous tooth relic was sent from
India, though it arrived only after the king's death. This relic
became the palladium of Sinhalese royalty till the British conquest
of Kandy in 1815, and it was worshipped by Dudley Senanayake
on his election as Prime Minister in 1965. Under Maha$sena also
was crystallized a further division in the Sangha. A group of
monks including one Sa$gala had in the reign of Maha$sena's father
left the Abhayagiri Viha$ra, disturbed by the introduction of
Maha$ya$na. After a revolt against the Maha$ya$nists Maha$sena built
for a follower of Sa$gala the Jetavana Viha$ra. For the rest of the
Anura$dhapura period the sangha was divided into these three
Nika$yas: Maha$viha$ra, Abhayagiri, and Jetavana. A Nika$ya does
not correspond to a sect or a denomination: the institution is
clerical only, and its characteristic is that monks of more than one
Nika$ya will not perform ceremonies (ordination, <1uposatha,>1 etc.)
together; customs and teachings (Sinh. <1sirit virit>1) may also differ
but need not do so. In ancient times, however, the doctrines of the
Nika$yas usually did differ: after Maha$sena the Maha$viha$ra was
Therava$da, the Abhayagiri Maha$ya$na, and the Jetavana alter-
nated. Ceylon was naturally influenced by Buddhist developments
in India. Indian chronology is uncertain, but Maha$sena's reign
must have been roughly contemporaneous with the rise of the

                                                            31

Yoga$ca$ra school of Buddhism in India, which brought irra-
tionalist doctrines and magical practices. In the sixth century,
when the great Buddhist logicians lived and disputed in India, a
Therava$din monk came from India and defeated the Maha$ya$nists
in public debate. In the seventh century, when the Chinese
traveller Hiuen Tsiang recorded great veneration in India for
Buddhist learning, there was so much interest in Ceylon in the
<1Abhidhamma>1 (Buddhist systematic philosophy) that one king
before setting out for war "asked his general to request his queen
to study the <1Abhidhamma,>1 preach it and transfer its merit to
him'.<s45>s In the middle of the ninth century Indian Buddhist
Tantrism (Va$jiriyava$da) came to Ceylon, and though its esoteric
doctrines diverged very widely from Therava$da it seems to have
been influential even at the height of the Pol*onnaruva period, over
300 years later.<s46>s There are other telling signs of Maha$ya$na
influence on the official religion of the later Anura$dhapura period,
culminating in a tenth-century inscription<s47>s which declares that
the Buddha promised that none but Bodhisattvas could become
kings of Ceylon.
  In the second century B.C. and fifth century A.D. there were
periods when Tamil invaders ruled in Anura$dhapura, but even
they professed Buddhism, or at least followed Buddhist customs,
and they do not seem to have occasioned any cultural discon-
tinuity. However, with the rise of Pa$n*d*yan power, invasions from
South India became more frequent and more severe, and the
Sinhalese and Tamils were repeatedly at war. Pol*onnaruva, further
from India, was a royal seat for the first time in the late seventh
century, and the capital gradually shifted to there; the last king
to rule in Anura$dhapura died in 896. Already the Tamil Hindus
must have been exerting cultural influence, for in the late eighth
century King Mahinda II feasted brahmins and gave them
sweetened milk to drink, a Hindu practice. Similar activities are
recorded in the ninth century. This is not surprising, as by no
means all relations with Tslnils were hostile: already in the
seventh century there were Tamil ministers, who are recorded to
have built Buddhist monasteries.  There were peaceful Tamil

32

colonists who intermarried with Sinhalese families, probably at
all social levels; some Sinhalese kings had Tamil wives. More-
over the Sinhalese kings themselves used Tamil mercenaries called
<1Velakka$ras.>1 Finally in A.D. 1001 the Col*a dynasty from South
India, at the height of their power, having already held Anura$dha-
pura for some years, captured and ruled in Pol*onnaruva. The
Sinhalese were confined to the mountains and the south, squab-
bling among themselves, while the Tamils sacked all the monas-
teries, and pillaged and destroyed the premises of luxury and
religion. In this period much of ancient Ceylon was lost. When
Yijaya-Ba$hu I reconquered the throne in 1065 it was no longer
possible to hold a higher ordination ceremony, as there were not
five monks left to form a quorum, so he had to send to Burma for
some monks to establish a new line of succession. The Order of
nuns had died out completely, and this could not be reinstated, as
there were no Therava$din nuns elsewhere, so that Order became
extinct in Ceylon. Even after Vijaya-Ba$hu's restoration the <1Velak-
ka$ras>1 were so powerful that not only did they mutiny and sack
Pol*onnaruva, but under one of his successors they were in control
of the Temple of the Tooth there, probably operating a "protection
racket'.<s48>s
  Parakkama-Ba$hu I (1153-86), known as the Great, decisively
defeated the Tamils and even invaded the Indian mainland. The
cultural activity of his reign has led Malalasekera justly to dub it an
Augustan age. The buildings both religious and secular erected
by him and his successor still astound the visitor to Pol*onnaruva.
There was great production of scholarly literature in Sanskrit,
Pali, and Sinhalese. The twelfth and thirteenth centuries were
probably the only period when Sanskrit was used in Ceylon. Its
rise, due perhaps to Maha$ya$na influence within Buddhism and
Tamil influence on society, had a lasting effect on the Sinhalese
language, which since this period has been the prevailing literary
medium. Modern Sinhalese is full of Sanskrit loanwords (i.e.
<1tatsamas>1) with Sinhalese grammatical endings, but has, so far as
I know, no Pali loanwords at all. This applies equally, and most
remarkably, to the religious terminology, which is either "pure
Sinhalese'<s49>s (i.e. the mediaeval, un-Sanskritized language, also

                                                                 33

called <1El*u>1) or Sanskrit, also known as "mixed Sinhalese'<s50>s. Lin-
guistically almost the only thing which is Pali about Sinhalese
"Pali Buddhism' is the sacred texts, which are still normally used
in the original.<s51>s
  Parakkama-Ba$hu held a council to reform the Sangha and estab-
lish the true doctrine. The <1Maha$vam*sa>1<s52>s records that at this time
"In the villages belonging to the Sangha the good morals of monks
consisted only in their supporting their wives and children.'
Senior monks of the three Nika$yas were convened, and after much
discord the king himself, with the aid of three learned monks,
decided the points at issue in favour of the Maha$viha$ra; he
accordingly suppressed the other two Nika$yas in 1165. The
Sangha was thus organizationally unified for the first time in
more than a thousand years. Pol*onnamva fell in 1215. Under the
pressure of Tamil attack and internal discord the Sinhalese kings
of the next three centuries moved their capital from place to
place, generally down the west side of the island and into the
mountains. Parakkama-Ba$hu II (1236-71) imported monks from
the Col*a country in South India "and established harmony
between the two teachings (<1ubhayasa$sanam*>1)',<s53>s so we infer that
though the Abhayagiri and Jetavana Nika$yas had formally been
extinguished some monks still held to Maha$ya$na doctrine. This,
however, is the last occasion on which we hear of them. It is
interesting that Therava$da Buddhism flourished in South India
at this period. Parakkama-Ba$hu IV (c. 1302-46) appointed as his
teacher a monk from the Col*a country, who taught him all the
<1Ja$takas>1 (stories of the Buddha's previous births), which were then
translated from Pali into what is still the standard Sinhalese ver-
sion; the translation is ascribed to the king himself.<s54>s In 1411 the
Chinese invaded the west coast and carried off the king. This was
probably the last Sinhalese contact with Maha$ya$na Buddhists till
recent times. The Chinese installed Parakkama-Ba$hu VI (1412-
-67), who ruled at Kot*t*e$, near Colombo, and was the last king
(till George III) to rule the whole of Ceylon.

34

  When the Portuguese arrived in 1505 the island was politically
divided in much the way that was to persist till reunification came
under the British: one king ruled at Kot*t*e$, another at Kandy.
King Dharmapa$la of Kot*t*e$<s55>s was converted to Catholicism and
became Don Juan Dharmapa$la; he lost all power, and at his death
in 1597 the line was finally extinguished, and with it all pretence
of political independence for the low country. The Kandyan
kingdom, after the brief reign of Ra$jasim*ha I (ruled 1580-91),
who though a Sinhalese made S/aivism the court religion and
persecuted Buddhism,<s56>s remained independent till 1815, and
preserved a Sinhalese Buddhist cultural identity, but till the reign
of Ki$rti S/ri$ Ra$jasim*ha (1747-81) Buddhism does not seem to have
prospered. In the intervening period no noteworthy works of
Sinhalese or Pali literature were written. Moreover, within just
over 15o years Kandyan kings sent abroad three times for monks
to come and re-establish the higher ordination in Ceylon. The
first two missions, received by Vimala Dharma Su$rya I (1592-
1604) in 1596<s57>s and by Vimala Dharma Su$rya II (1687-1707)
in 1697,<s58>s came from Burma and had curiously short-lived effects.
According<s59>s to an eighteenth-century source<s60>s these two missions
were invited not in order to initiate a genuine religious revival,
but to legitimize the claims of incumbents to their temple lands.
These incumbents, who had taken the lower ordination, claimed
thereby the status of novices, but some of them kept their lay
names and many were not even celibate. During the captivity of
Robert Knox (1660-79) those who thus made no pretence of
observing the monastic rules were known as <1ganinna$nse$s,>1 and wore
not yellow but white. This class, half-way between monk and
layman, was a phenomenon unique in Ceylonese history.
  In 1697 only one monk could be found to converse in Pali with

                                                              35

the visiting monks from Burma. It was a pupil of both of this
monk's pupils who instigated the revival of Buddhism in the
mid-eighteenth century.  Va%livit*a Saranamkara<s61>s founded a
renaissance of Pali and Sinhalese literature; soon the most
learned man of his time, he rapidly became the most influential,
and deserves to be called the father of modern Sinhalese
Buddhism. His patrons, oddly enough,<s62>s were members of a Tamil
royal line from Madurai who ruled Kandy after 1739. Applications
for monks were this time made to Siam; two embassies failed,
but the monks who arrived in response to the third held a great
ordination ceremony in Kandy in 1753 at Malvatta, Saranam-
kara's monastery. They made a <1si$ma$>1 there and a new <1si$ma$>1 at
Asgiri, Kandy's other main monastely, and then toured the
country on similar errands. Saranam*kara received the title of
<1Sanghara$ja>1 ("King of the Order') and founded the current lines of
pupillary succession at Malvatta and Asgiri. Shortly afterwards
the king put these two monasteries formally in charge of all other
monasteries in the country, and ordered that only they could
hold higher ordination ceremonies. They are still the headquarters
of the Siyam Nika$ya throughout the hill country, which is to say
of the largest body of monks. Siyam Nika$ya monks in the low
country have headquarters elsewhere, but this is purely a matter
of convenience. The founding in the nineteenth century of
the other two Nika$yas of today, the Ra$man+n+a and the Amarapura,
will be discussed in Chapter 8.
  When the British acquired Kandy in 1815 Buddhism ceased at
last to be the State religion, but by the Kandyan Convention they
guaranteed that "Buddhism and the Agama [religion] of the Devas
[gods] were inviolable' and monasteries were allowed to retain
their property and exemption from tax. In 1848 a Buddhist monk
was executed by the British for allegedly taking part in an abortive
revolt, but this was an isolated incident. In 1880 Col. H. S. Olcott
of the Theosophical Society, intrigued by newspaper reports of a
public disputation in 1874 in which a Buddhist monk had defeated
a Methodist missionary, visited Ceylon and founded the Buddhist

36

Theosophical Society, which set up its own schools in competition
with Christian schools and started a new stream of Buddhist
development--but this is another story, best left to Chapter 1.
  There are at present some 17,000 Buddhist monks in Ceylon,
in a population of 101/2 million of whom 71/2 million are Sinhalese
and 7 million Buddhists. Almost all the Buddhists are Sinhalese.
The other Sinhalese are Roman Catholics whose ancestors were
converted under the Portuguese; almost all of these live along the
west coast. Of the Sinhalese, 21/2 million are officially classed as
Kandyans. The other elements of the population are as follows.
There are over a million Tamils known as "Ceylon Tamils' or
"Jaffna Tamils' whose ancestors came over in ancient and mediaeval
times; they live mainly in the north of Ceylon and along the east
coast, and the great majority are S/aivite Hindus. There are about
a million Tamils known as "Indian Tamils' or "Indians' whose
families were imported by the British to work on tea plantations;
they mostly live in the hill country and are predominantly S/aivite
Hindu, though a sizeable minority has been converted to Chris-
tianity, mostly Roman Catholicism. There are half a million
Muslims, also known as "Ceylon Moors', who are scattered
throughout the island--mostly engaged in trade; their origin is
not quite clear, but their ancestors were probably Arab merchants
in the Middle Ages, and some of them at least must have arrived
via South India, as they speak a dialect of Tamil; but they have
intermarried a great deal with the local population, and in Sinha-
lese areas are usually bilingual. Of the smaller minorities, the
Burghers, of mixed European and Ceylonese descent, are Chris-
tians, and live in towns (mainly Colombo) where they have white-
collar jobs, or on estates as planters; there is a sprinkling of Malay
Muslims, descendants of mercenaries employed by kings of
Kandy; other Indian immigrant groups pursue trade in Colombo;
and the few remaining Va%ddas, speaking their archaic Sinhalese
dialect, live in the jungles on the eastern side of the mountain
massif, where no doubt their relations have imperceptibly merged
with the rest of the population.<s63>s

                                                           37


                   IV. FIELD-WORK AND PRESENTATION

     I wished to be a spectator of their sports and told them so; but
     finding they did nothing but look at me, I retired to a neigh-
     bouring house to take some refreshment . . .<s64>s
                                                JOHN DAVY

  The raw material on which this book is based consists princi-
pally of some three dozen interviews with monks in the area,
between ten and twenty miles from Kandy, where I lived for most
of the year August 1964-August 1965. These interviews were
conducted without interpreters, though during three of them
bilingual friends were present. I was able to check dubious points
and make a few revisions on a second visit to Ceylon in the latter
half of 1969; in particular I used this opportunity to broaden my
knowledge of lay opinion on religious questions. Let me briefly
explain the reasoning behind my procedure.
  In asking people their opinions I was always very much aware
of two dangers. The more obvious one is that people will tell you
what they think you like to hear. The Sinhalese are deeply
courteous, and do not consider it a virtue to tell unpleasant truths.
The answer to any leading question is therefore likely to be "Ov'
("Yes'). This word, ov, which we translate "yes', in effect has rather
the meaning, "I have heard'; it must not be taken as an answer to a
question without further probing and corroboration. The less
obvious danger is to force people to give an answer to something
they would otherwise never think about. It is not that the question
may be inappropriate, or phrased in the wrong terms; what I have
in mind is rather that a question may unnaturally enlarge a man's
cognitive position by forcing upon him a new thought. Such a
question would be, "What god or sort of god would you like to be
reborn as?'; for although most villagers say they want to be
reborn as gods, they have never thought about it more specifically.
If the man X replies, "As Kataragama', this is certainly of interest,
but it would be misleading to record "X wants to be reborn as
Kataragama', unless the question--and indeed the whole assess-
ment of the situation--were recorded too. This difficulty is to
some extent unavoidable, as no field-worker can sit around for ever

38

waiting for people to bring up subjects spontaneously and display
the contents of their minds before him unasked; but one must
exercise caution. Sitting around would be ideal; but it is too slow
in practice. I had to conduct interviews.
I did, however, feel that I should sit around for a time till I
got some idea of what would be the right questions to ask about
religion. After living for some seven months in the village which
I shall call Mi$gala I drew up a list of questions I thought I could
profitably ask of monks, and these became the backbone of my set
interviews. I added some questions as I went along, as I hit on
interesting points of controversy; other questions I came to spend
very little time on--though for the sake of uniformity I did not
entirely drop them--as I perceived that the answers were stan-
dardized. Why did I choose to interview monks? Firstly, because
this was considered a much more natural activity than to interwiew
laymen, who would have wondered what I was up to ; monks were
not surprised that someone should want to talk to them about
religion, and indeed some of them were quite flattered that an
Englishman should have come so far to do so. Connected with this
is the more important and rather obvious second reason: monks
have many more opinions on religious topics than do most laymen.
The list of religious questions I could have asked a layman which
he might himself have asked, leaving aside those to which answers
would be completely standardized, would be very short. Thirdly,
monks are a clearly definable class, so that if I interviewed all or
nearly all the monks in a certain area my study would be exhaustive
in at least one respect, and even if I failed in my wider aims I
would have done one thing thoroughly. Finally, interviewing the
monks gave me a pretext not merely for travelling round the area
in general, but in particular for visiting all the temples and thus
making without too much extra effort a survey of local religious
art and iconography. It was plainly convenient to interview monks
in an area roughly centred on Mi$gala; I made it my goal to inter-
view all the monks, or rather all the incumbents, in the same
parliamentary constituency, as this seemed to be about the number
I would have time for. In the event I nearly succeeded in this aim.
Altogether I visited thirty-nine local temples (including temporary
establishments) and conducted reasonably full interviews (some-
times in more than one session) with thirty-four local monks,
conversing at least briefly with all but two of the incumbents.

                                                               39

   At first it seemed to me that my book should consist of this
interview material, and the descriptive notes which go with it.
However, they would be rather meaningless to readers of a cultural
and intellectual background so far from that of a Kandyan monk
unless I explained the assumptions on which the questions and
answers rest. To do this would require a large part of the book
as it now stands (especially Chapters 2 and 4). Moreover, the work
would become unduly descriptive, with too little connection to the
problem situation of western academics.  Accordingly I have
framed my book with regard chiefly to the intellectual issues
involved, and have drawn on the interviews wherever relevant,
without exhausting the material they contain. In particular I
have neglected certain areas which they cover, such as politics,
which are separable from my main concerns and the inclusion of
which would make this book unduly long; I hope to publish some
of it elsewhere in the near future. I must also stress that although
my book is based on fieldwork in Sinhalese villages it is intended
as a study of traditional Buddhism, <1not>1 as a monograph on "religion
in a Sinhalese village'.
    A work so ambitious cannot but have serious shortcomings, but
to conclude this introduction I must mention what I consider to
be its major deficiency. Close acquaintance with other ancient
schools of Buddhism would certainly elucidate the origin of many
Ceylonese Therava$din beliefs and practices, especially those
branded as "popular' : veneration of the Buddha, attitudes towards
the dead, and "prayers' (<1pra$rthana$>1) are obvious examples. I hope
that future research by myself and others will throw light on these
problems.

                                I
                   SINHALESE BUDDHISM--ORTHODOX
                         OR SYNCRETISTIC?

LET me at the outset state a general conclusion. I found the
Buddhism which I observed in Kandyan villages surprisingly
orthodox. Religious doctrines and practices seem to have changed
very little over the last 1,500 years. But Buddhism is 2,5oo years
old, so why do I say 1,500 years, and what do I mean by orthodoxy?
I mentioned in the introduction that Buddhism came to Ceylon
in the days of the Indian emperor Asoka, about 243 B.C.<s1>s This was
Therava$da Buddhism. The contents of Therava$da Buddhist oral
traditions go back uninterrupted to the time of the Buddha himself,
but we only have histories written nearly a thousand years after
the Buddha's death,<s2>s so the accuracy of these traditions for the
early period, especially before Buddhism came to Ceylon, is
controversial. To enter this controversy lies outside my present
scope, and so does the adjudication of Therava$din claims that they
alone represent true Buddhist orthodoxy, and that other sects are
heretics. But whatever we make of the Therava$din accounts which
tell how three Buddhist councils within a couple of centuries of
the Buddha's death established the scriptural canon and expelled
heretics, it is certain that in Asoka's day schisms had occurred,
that there were different sects and different (though perhaps only
very slightly different) scriptures, and that from some of these
sects was to grow Maha$ya$na Buddhism, which became so in-
fluential in India, Tibet, and the Far East.
Therava$din orthodoxy rests on the texts of the Pali Canon; it is
so defined by Therava$da Buddhists themselves, in so far as they
are sophisticated enough to pose the question. More precisely,
we may add, it rests on the Pali Canon as interpreted by the
ancient commentaries, an interpretation which is accepted

                                                                 41

without question in Ceylon and with only very minor disagree-
ment by most western scholars. At first glance scriptural authority
thus seems to be two-tier; but this picture is rather over-simplified,
because the Canon itself contains ancient commentaries, and its
chronology is largely problematic. The Canon is called in Pali
the <1Tipit*aka,>1 because it consists of three <1Pit*akas>1 ("baskets'): the
<1Vinaya Pit*aka>1 (monastic rules), <1Sutta Pit*aka>1 (sermons consisting
in turn of five <1Nika$yas,>1<s3>s of which the first four contain the sermons
and the last miscellaneous texts, many of them in verse), and
<1Abhidhamma Pit*aka>1 (systematic philosophy). Buddhists believe a
passage at the end of the <1Vinaya Pit*aka (Cullavagga>1 11 and 12)
which says that two councils of monks were held, one just after
the Buddha's death and one a hundred years later, and that at the
first council were composed the <1Vinaya>1 and <1Sutta Pit*akas,>1 in their
final form.<s4>s The passage, being a part of the <1Vinaya Pit*aka,>1 con-
tradicts its own literal accuracy by mentioning the second council.
Most other texts in these two <1Pit*akas>1 claim to be, and therefore
according to Buddhist tradition are, the very words of the Buddha;
but even in the four <1Nika$yas,>1 the collection of sermons setting
out the Buddha's teachings, there are some sermons attributed
not to Buddha but to disciples. Some of these are said to have
been delivered in the Buddha's lifetime; but a few explicitly state
that the Buddha is dead, or refer to events after his death.<s5>s Part

42

of the <1Abhidhamma Pit*aka>1 is said by Therava$din tradition to have
been compiled at the third council, held <1c>1. 247 B.C. under the
auspices of Asoka, and whether or not such a council took place
scholars agree with this relative chronology, and go further by
assigning to the whole <1Abhidhamma>1 a date when Buddhism was
already organized and scholastic. Without discussing the dates
of the miscellaneous texts (<1Khuddaka Nika$ya>1) of the <1Sutta Pit*aka,>1
of which scholars consider some as likely to be the Buddha's words
as anything, some to be certainly much later, and some to be
based perhaps even on pre-Buddhist material, we have found a
measure of agreement between religious and scholarly opinion:
the Pali Canon as we have it was not all compiled at one time.
Therefore, we may add, it cannot reflect the state of the Buddhist
religion at one given moment; nor indeed is it likely that so large
and heterogeneous a body of material should be entirely free from
contradictions.
  Whatever its origins,<s6>s the Pali Canon must have existed more or
less as we have it now in the time of Asoka, when Mahinda brought
it over to Ceylon. An Asokan inscription lists seven texts, five of
which we can identify in the Canon we know, though mostly
under different titles, which suggests that the other two also may
be in our Canon. Among the sculptures at Bharhut and Sanchi
(second century B.C.) are pictures of stories in the Canon. These
are by far the oldest records of the Canon physically to survive:
the Pali Canon was not even written down till the first century
B.C., and of course no manuscript from that era exists.
  That Mahinda and his colleagues brought the Pali Canon (in
their heads) to Ceylon may be a slight over-simplification, but
cannot be far from the truth. We owe the story to the great com-
mentator Buddhaghosa, who wrote in the early fifth century A.D.
Buddhaghosa adds that they brought commentaries in Pali, which
Mahinda translated or recomposed in Sinhalese. If Pali originals
of the Sinhalese commentaries ever existed, they had been lost

                                                                43

before Buddhagosa's time. Adikaram<s7>s has shown that the Sinha-
lese commentaries on which Buddhagosa's work is based were
principally compiled in the first century A.D. The <1Maha$vam*sa>1
(XXXVII) says that Buddhaghosa came from North India to the
Maha$viha$ra in Anura$dhapura because in India the commentaries
(<1at*t*hakatha$>1) had been lost; his teacher told him that the Sinhalese
commentaries were the genuine work of Mahinda, and he should
translate them into Pali. His first work in Ceylon was the <1Visud-
dhimagga>1<s8>s (<1The Path of Purity),>1 an exhaustive summary in Pali of
Buddhist doctrine, still the best work of its kind.<s9>s He then wrote
commentaries in Pali on most works in the Canon.<s10>s The Sinhalese
books on which these commentaries are based have perished,
because he superseded them. Though he may at first have intended
only to translate them he in fact edited and systematized them.
He quotes nearly every earlier work of Pali literature known to
have existed. His interpretations are mutually consistent. To this
day Buddhaghosa's Buddhism is in effect the unitary standard of
doctrinal orthodoxy for all Therava$da Buddhists, whether or not
they are educated enough to be aware of the problem.
  The authors of the Pali commentaries on the parts of the Canon
which Buddhaghosa did not treat do not conflict with him, so the
commentatorial stand-point is unified and homogeneous. We
know very little about these other authors, but they probably were
contemporaries of Buddhaghosa or lived slightly later. The two
more important ones are Buddhadatta<s11>s and Dhammapa$la.<s12>s

44

Tradition has it that Buddhadatta was born in the Col*a kingdom
but became a monk at the Maha$viha$ra; there is an unlikely story
that he was on a ship going back to India when he met Buddha-
ghosa on his way out. Dhammapa$la wrote in South India and was
probably a Tamil too; "he states . . . that he follows the traditional
interpretation of texts as handed down in the Maha-Vihara'; and
"it is quite likely that he had the advantage of studying the Tamil
commentaries (of which we know that at least two existed) as
well'.<s13>s A further anonymous author wrote the Pali version of the
<1Ja$taka>1 commentary which has come down to us; the <1Ja$takas>1 are
stories of the 550 previous births of the Buddha; their Pali verses
are canonical, but the prose parts, which carry the story forward
so that without them the verses are mostly unintelligible, rank as
commentary and for a while existed only in Sinhalese, like the
other commentaries.<s14>s The author says that he follows the tradi-
tions of the Maha$viha$ra at Anura$dhapura. The commentary on
the <1Dhammapada,>1 which is not so much a true commentary as a
collection of stories much in the style of the <1Ja$taka>1 commentary,
is like the latter traditionally but probably wrongly attributed to
Buddhaghosa. The other two authors<s15>s of commentaries whose
names are known also lived at Anura$dhapura and wrote in the
same tradition.<s16>s
  Though the Pali Canon is strictly the only sacred work and the
supreme authority, there is another book which though neither
Canon nor commentary has in Ceylon received the kind of fame
and respect usually reserved for sacred books.  This is the
<1Maha$vam*sa,>1 the Pali verse chronicle of Ceylon, mentioned in the
introduction. The first part of it was written in the late fifth
century by a monk called Maha$na$ma who lived in a monastery at
Anura$dhapura belonging to the Maha$viha$ra. Like the commen-
taries, it was based on much Sinhalese written material, which
has long since disappeared. It was based also on another similar

                                                             45

Pali chronicle, the <1Di$pavam*sa,>1 which was written also in Anura$dha-
pura, perhaps by nuns,<s17>s about a century earlier, which makes it
the oldest work written in Ceylon to have survived. But the
<1Maha$vam*sa>1 follows it so closely, quoting long passages verbatim,
and at the same time is so superior, that the <1Di$pavam*sa>1 is compara-
tively neglected. Both chronicles are--luckily for us--predomin-
antly histories of the Buddha's teaching (<1sa$sana),>1 and focus
throughout on episodes of religious interest, beginning with
mythical accounts of the Buddha's visits to Ceylon. The earliest
date for which the <1Maha$vam*a>1 is authoritative may thus be fixed
by the most sceptical at the fourth century A.D., by the sanguine
at the third century B.C. (Mahinda's mission to Ceylon), and by
the pious during the lifetime of the Buddha himself. The chronicles
are not primarily expositions of doctrine (albeit each chapter ends
with a pious verse reflecting on the vanity of worldly pomp or some
such orthodox sentiment); and though they afford fascinating
material for the history of Buddhist ideas, they plainly do not
form a touchstone of doctrinal orthodoxy as I have defined it; but
they give accounts of religious practices and especially festivals
which are rather outside the scope of the canonical literature. So
though we may not be able to call a practice canonically orthodox,
because the Canon does not mention it, we may often say that it
stands in an orthodox tradition if it occurs in the <1Maha$vam*sa.>1 And
it cannot fail to strike a modern observer that many festivals des-
cribed in the ancient <1Maha$vam*sa,>1 as well as in its continuations,
differ little, except in scale, from festivals in Kandyan villages today.
  This then is what I meant when I said at the beginning of the
chapter that I found the Buddhism which I observed orthodox:
that the doctrines of the villagers would have been approved by
Buddhaghosa and that most of their religious practices would have
been familiar to him and his contemporaries. But I also said that
I found this orthodoxy surprising. Why so? I suppose it is unusual
for the religion of a society to change so little over l,5oo years;
certainly it is unusual for societies of which we have the records.
But my surprise was caused rather by the frequency with which
I had been told, by books and by people, that Sinhalese village
Buddhism was corrupt. "Corrupt' was the word generally used by

46

laymen; in academic circles this word sounds too pejorative to be
respectable, and is replaced by "syncretistic', or some periphrasis.
One example, from a noted authority: Sir Charles Eliot wrote, "At
present there may be said to be three religions in Ceylon; local
animism, Hinduism, and Buddhism are all inextricably mixed
together.'<s18>s Let me examine this belief that village Buddhism is
corrupt, and explain why I think it is mistaken.
   Sinhalese villagers have been judged corrupt Buddhists because
they say and do things which the judges think are incompatible
with what is said in the Pali Canon. In particular, most Sinhalese
believe gods and demons to exist and make offerings to them under
various circumstances. (The relevant beliefs are described in
Chapter 4.) The judgement that this is corrupt Buddhism is based
on a misunderstanding which has arisen because the original
people to make it were westerners, raised in a Christian culture,
whose background made them think of religion as god-centred.
The problem of God is central to the Semitic, the monotheistic
religions: it is essential that there is only one God, and all other
gods are false gods, which means that they do not exist. It is true,
though almost ludicrously inadequate, to say that Christians, Jews
and Muslims consider God the most important thing in the
universe. Buddhists with whom I talked, however, will admit the
existence of the Christian God with perfect indifference,<s19>s just as
they would no doubt admit the existence of any other god proposed
to them; gods are not a problem in Buddhism. Some Buddhists
recognize this explicitly: one old monk said to me, "Gods are
nothing to do with religion.'<s20>s Gods are powerful beings to be

                                                                   47

supplicated for worldly goods; Buddhism, as we shall see in the
next chapter, is concerned with ethics, future lives, and release
from worldly existence. An ordinary Buddhist layman is no more
concerned to accept or deny the existence of some new super-
natural being than a western layman is concerned to accept or
deny the existence of some newly discovered type of nuclear particle
or natural force. Both are just facts of life.
    A Buddhist who acknowledges gods can thus be cleared of the
charge of doctrinal incorrectitude. But of course some gods are
much more important to him than others. These gods figure in
what any casual western observer would call his religion; they
come under any western academic definition of religion, including
Spiro's ". . . interaction with culturally postulated superhuman
beings', which I discussed in the Introduction. I explained there
that I prefer the conventionalist use of terminology, to call some-
thing a religion if it is so called by its own practitioners. But I
would also like to deny that the Simalese Buddhist religion, in
ay sense of the term, is syncretistic. I take it that we talk of
cultural syncretism when elements from different cultures com-
bine or fuse to produce a composite which is not normal or not
acceptable in either culture. There is of course no dividing line
between what is syncretism and what is not, because syncretism
has no "essence' ; but it seems to make sense to say that the culture
of an Englishman who goes to dinner in an Indian restaurant is
not thereby syncretistic, whereas the culture of an Englishman
who preserves some English habits, but refuses cutlery and eats
all his food with his right hand is: he has broken the unwritten
rules of his original culture, without entirely adopting a new one
(a feat which is rarely possible). Syncretism is a concept easier to
apply with clarity to a system of ideas like Buddhist doctrine, both
because there the rules are mostly written and explicit, and because
strict logical incompatibility can occur between ideas but not
between behaviour patterns.
  It is claimed that Sinhalese religion is syncretistic when
Buddhists acknowledge "Hindu gods', by which is meant gods of
whom they have learnt from Hindus. (This theory is implicit in
the title of an article by E. R. Leach: "Pulleyar and the Lord
Buddha: an aspect of religious syncretism in Ceylon'.)<s21>s Let us
leave aside the question of whether these gods are part of Sinhalese

48

Buddhist religion, for we can at least agree that they are part of
Sinhalese Buddhist culture. I still deny that the presence of
Hindu gods in Sinhalese Buddhist culture is a sign of cultural
syncretism. Buddhism has always acknowledged Hindu gods. This
is not surprising because the Buddha, as Indians like to point out,
was a Hindu, just as Christ was a Jew. (This does not mean that it
makes much sense to think of Buddhism as a kind of Hinduism;
the analogy with Christianity holds.) However much speculation
may be devoted to dividing the Pali Canon into layers of authen-
ticity, no rationalist western scholar has ever reached a bedrock
stratum of "primitive Buddhism' in which no gods and demons play
a part. This statement will be amplified in Chapter 4; suffice it here
to say that supernatural beings were as much a part of the Buddha's
universe as they are of a Buddhist villager's universe today.<s22>s
As Sinhalese Buddhism (like Buddhism in ancient India) has
throughout its existence been geographically adjacent or even
intertwined with Hinduism it is natural that theological (in the
strict sense) developments accepted by the Hindus, and in particu-
lar by Tamil Hindus, have been accepted by Buddhists also. Over
the ages the personnel of the Sinhalese pantheon has somewhat
changed; contact with Tamils has brought the relative new-
comers Kataragama and Pattini$ into prominence, but some old-
timers like Vishnu, whom the Buddha is believed to have made the
guardian of Buddhism in Ceylon,<s23>s are still powerful. What
applies to these "Hindu' gods applies equally to the lesser beings,
the demons, etc., of Eliot's "animism'. In contrast the structure
of the pantheon, which is brilliantly described in Obeyesekere's
articles, has changed little ; the kinds of change which have occurred
are suggested in the Introduction. But the status of the whole
pantheon in its relation to the Buddhist religion has changed not
at all. The traditional Sinhalese Buddhist framework for inter-
preting reality has remained unaltered.
It will be seen that my argument against the alleged syncretism
of Sinhalese Buddhism has two levels. On the lower level I am
claiming that the presence of "Hindu' or "animist' supernaturals

                                                                49

in the Buddhist's universe is not a novel or syncretistic feature,
but has always been the case. This is a historical and textual argu-
ment. Argument on this level should be sufficient to refute the
charge ; but the argument on the higher level is theoretically more
interesting. I mentioned that Buddhist villagers will acknowledge
the existence of the Christian God. I have never known a Buddhist
actually to pray to this God for health, wealth, or other benefits in
their life; but if one did so I should still not consider his religion
syncretistic; because such a prayer would still be consonant with
Buddhist doctrine. This higher-level argument, the point I
especially wish to make, is doctrinal and logical. So long as
Buddhists continue to treat gods as a kind of supermen, able to
grant favours to suppliants, but still ultimately of limited life and
powers and subject to moral law, their beliefs are not syncretistic.
Belief in gods like this is not logically (or otherwise) incom-
patible with Buddhist doctrine. Who the gods are is irrelevant to
this point. That Sinhalese Buddhists have adopted gods from
Hindus rather than from Christians is an accident of historical
contiguity. Despite invasions and wars, relations with Tamil
Hindus have been more intimate than with the European Chris-
tians. The period of friendliest contact with Europeans was prob-
ably the late nineteenth century, and under these circumstances
Christianity began to influence Buddhism to the point of genuine
religious syncretism.  Copleston reports that during Buddhist
festivals in Colombo walls would exhibit the slogan, "God bless
our Lord Buddha'.<s24>s This is syncretism, because it shatters the
Buddhist framework. But I saw no such slogans; they are now
only part of history.
  The Buddha's teaching was limited to what he thought con-
ducive to enlightenment. Most people have broader interests, so
Buddhists are almost bound to have other beliefs besides those of
Buddhist doctrine. To sum this up in a word, Buddhism in real
life is <1accretive.>1 This conclusion harmonizes with that of Professor
Dumont when he writes of the currents in early India which pro-
duced Buddhism (and other religious movements) : "An individual
religion based upon choice is <1added on>1 to the religion of the
group.'<s25>s (My italics.) I am not at all averse to describing the

50

original part of the Buddha's message as an "individual' as against
a "group' religion; we must only remember that this "individual'
religion has never existed <1in vacuo.>1
Evidently I could not be further from agreeing with an observer
like Paul Wirz, who has written of Ceylon,<s26>s "In reality, it is the
same here as in other Buddhist countries; only very few compre-
hend the true Buddhist dogma in its real profoundness; the rest
are Buddhists in name only, among them also a great part of those
who wear the yellow gown. . . . They are too deeply rooted in their
old conceptions and customs. . . .' In another recent book which
deals with the subject of contemporary Sinhalese religion, the
American sociologist Professor Bryce Ryan says that, "Pure
Buddhism is a philosopher's abstraction.'<s27>s While I have tried to
show in my Introduction that the concept of pure Buddhism can
be meaningful, and in this chapter what I mean by it, it is doubtless
to passages such as the above that Professor Ryan is referring-<s28>s
The reader may be wondering what "the true Buddhist dogma' is
which Dr. Wirz comprehends "in its real profoundness' while it
eludes so many Buddhist monks. Clearly I cannot give a definitive
answer to this, but I think I can indicate what Dr. Wirz has in
mind, and the matter is worth a digression because Dr. Wirz's

                                                           51

remarks are typical of a view of traditional Buddhism by no means
confined to scholars and Europeans, but common also to English-
educated Sinhalese.
  Buddhism was discovered for the West mainly by British mis-
sionaries and civil servants in Ceylon in the nineteenth century.
George Turnour, a civil servant, pioneered the study of Ceylon
history when he discovered, edited and translated the <1Maha$vam*sa,>1
the first important Pali text to be published.<s29>s But it was the
missionaries who took more interest in Buddhism, in order to
know their enemy better. In particular the works of a Methodist
minister, the Rev. Robert Spence Hardy,<s30>s were important
sources for western knowledge of Buddhism, and were excerpted
by encyclopaedias of religion.<s31>s The missionaries decided, with
some justice given their own frame of reference, that Buddhism
was an "atheist creed',<s32>s since they were told that the Buddha was
not a god. This information was seized on by western rationalists,
atheists and agnostics, who were delighted to hear of a religion
which was atheistic, and even claimed to be rational; Buddhism
as they understood it seemed an ideal ally in the fight against
Christianity. Indeed, Bishop Copleston, Bishop of Colombo,
writing in 1892, saw in Buddhism's "alliance with Western scepti-
cism' one of its few hopes for survival!<s33>s
  In the meanwhile two Ceylon civil servants moved into the
forefront of Buddhist and Pali studies. R. C. Childers began to
publish Pali texts in 1869, and from 1872-5 compiled the first
western Pali dictionary. T. W. Rhys Davids, probably the most
influential Pali scholar of modern times, founded the Pali Text
Society for the publication of Pali books in 1881; his book
<1Buddhism,>1 first published in 1877<s34>s and many times reprinted, may

52

be said to mark the close of the pioneer period of Buddhist studies.
Rhys Davids was an excellent scholar, but he naturally stressed
the rationalist elements in Buddhism, because they formed the
most striking contrast both to Christian, and, in Buddhist eyes, to
other Indian traditions. I suspect that he also found them the
most sympathetic. Buddhism became familiar to the educated
public when in 1879 Sir Edwin Arnold published <1The Light of
Asia>1;<s35>s this long poem on the Buddha, while presenting Buddhism
  m the way usual at the time, suggested an analogy between the
Buddha and Christ which doubtless contributed to its popularity.
  In their campaign against Christianity English secularists and
free-thinkers soon found some curious allies, less sympathetic but
more militant than the Buddhists of Ceylon. The president of the
National Secular Society, Charles Bradlaugh, sponsored and gave
lectures on comparative religion to disprove Christianity's claims
to uniqueness. In 1874 Mrs. Annie Besant heard him lecture on
similarities between the stories about Krishna and about Christ.
She joined the N.S.S., and soon became Bradlaugh's chief assis-
tant and companion, and co-editor of the <1National Reformer,>1 the
N.S.S. newspaper. In 1882 Mrs. Besant devoted considerable
space in the <1National Reformer>1 to first criticizing and later attack-
ing the Theosophical Society, which had been founded in New
York in 1875 by Madame Blavatsky, a Russian, and Col. H. S.
Olcott, an American, and in 1879 had moved to India. One of
the earliest theosophist classics was entitled <1Esoteric Buddhism,>1<s36>s
and the movement was sometimes referred to as Buddhist Theo-
sophy, though it had in it no more of Buddhism than of divine
wisdom. Early in 1889 a leading member of the N.S.S. contributed
a series of articles also entitled "Esoteric Buddhism' to the <1National
Reformer,>1 and Mrs. Besant, who had studied the book of that
name, lectured on "Buddha: His Life and Teachings'.<s37>s In the
same year, to the infinite distress of Bradlaugh and her other
rational friends, she was converted to theosophy; when Madame
Blavatsky died in 1891 she became the leader of the movement,<s38>s
and held this position till her death in 1933. In 1891 theosophy

                                                           53

was described by her friend William T. Stead as "the natural child
of the marriage of Christianity and Buddhism'.<s39>s Mrs. Besant
visited Ceylon for a week in November 1893 (when she was on
her way to India for the first time), and received spectacular wel-
comes in Colombo, Kandy, and Galle. The arrangements had been
made by Col. Olcott.
  Col. Olcott himself first landed in Ceylon on 17 February 1880,
a day which is still locally celebrated as "Olcott Day'. (There are
so far as I am aware no longer any special festivities, but articles
on Olcott appear in the newspapers.) On this occasion he stayed
for five months. His influence on Buddhism was in the first
instance organizational. He founded the Buddhist Theosophical
Societies, which ran schools which emulated and competed with
the schools of Christian missionaries. (These schools remained
under the B.T.S. until the government's takeover of schools in
1961.) Olcott drew up a "Buddhist catechism', and encouraged
Buddhists to celebrate Wesak (the anniversary of the Buddha's
birth, enlightenment and death) with songs modelled on Christ-
mas carols. It was Olcott who gave the impetus which before long
resulted in the founding of further Buddhist institutions to parallel
Christian ones: the Young Men's and Young Women's Buddhist
Associations, and the Buddhist Sunday Schools (<1daham pa$sal>1) at
which monks teach the village children religion from books pro-
vided by the Y.M.B.A. The latest example of Buddhist imitation
of Christian institutions occurred in 1966 with the introduction
of the <1poya>1 weekend: the weekly holiday of Sunday with Saturday
afternoon has been replaced by the <1poya>1 day (full moon, half moon
or no moon) with the preceding half-day. This alone will suggest
how much modern political Buddhism owes to Col. Olcott. Per-
haps it also owes something more directly to the Christian
missionaries.<s40>s In 1929 Sir Hugh Clifford, Governor of Ceylon
1925-7, wrote in the <1Encyclopaedia Britannica,>1<s41>s

54

Women's Buddhist Associations, and Islam the doffing of shoes on
entering Buddhist Temples. Simultaneously the broad toleration of
other faiths, which from the earliest times has been so marked and so
notable a feature of pure Buddhism, has shown a tendency to be
replaced by vulgar abuse of tenets of other creeds of a kind with which
the more ignorant types of Christian Missionaries "to the heathen' have
unhappily made the world familiar.

Is it entirely a coincidence that of Ceylon's Prime Ministers since
independence the least demonstratively Buddhist, Sir John Kotel-
awela, was the only one not educated at a Christian mission school?
  From his first visit to Ceylon Olcott took back to Madame
Blavatsky in India a young disciple, Don David Heva$vita$rana,
later known as Anaga$rika<s42>s Dharmapa$la. Though he spent most
of his adult life in India, running the Maha$ Bodhi Society, which
he founded, Dharmapa$la exerted crucial influence on the forma-
tion of modern Buddhist ideology in Ceylon. In 1893, under the
auspices of Mrs. Besant, he represented Buddhism at the World
Parliament on Religions in Chicago. At that time Professor Max
Mu%ller wrote to him in a letter,

You should endeavour to do for Buddhism what the more enlightened
students of Christianity have long been doing in different countries of
Europe: you should free your religion from its later excrescences, and
bring it back to its earliest, simplest, and purest form as taught by
Buddha and his immediate disciples. If that is done, you will be sur-
prised to see how little difference there is in essentials between the
great religions of the world. And this must be done with perfect honesty.
Nothing not quite sincere or truthful should be tolerated. Nothing has
injured Buddhism so much in the eyes of scholars and philosophers in
Europe as what goes by the name of Esoteric Buddhism. Madame
Blavatsky may have been a dear friend to you, but Truth is a dearer
friend.

  Here then meet the two streams of western thought which have
deeply influenced English-educated Buddhists in Ceylon: theo-
sophy, and "the scholars and philosophers of Europe' seeking
Buddhism's "purest form as taught by Buddha and his disciples'.
Modern Buddhist fundamentalist movements and institutions,
such as the Vajira$ra$ma in Colombo, plainly owe much to the pre-

                                                                55

conceptions, as well as the researches, of scholars like Max
Mu%ller and Rhys Davids. The editions and translations of the
Pali Text Society are more used than any indigenous materials by
Sinhalese scholars. Several Europeans, especially Englishmen and
Germans, have gone to Ceylon to become monks, and passed their
lives there with great reputations for holiness. They probably owe
their first acquaintance with Buddhist texts to the late nineteenth-
century translations of Rhys Davids or Karl Eugen Neumann, and
have contributed the results of their own western education to the
current of Buddhist ideas. One community of these monks, headed
till his death by a learned German, the Rev. N+a$natiloka, is very
near to the place where Dr. Wirz stayed. Most recently a German
disciple of the Rev. N+a$n*atiloka, the Rev. N+a$n*aponika, has from
his Forest Hermitage in Kandy helped to found the Buddhist
Publication Society, which distributes Buddhist pamphlets
throughout the world; since 1958 they have published more than
a hundred <1Wheel>1 booklets and <1Bodhi Leaves.>1 At first these were all
in English, but some are now appearing in Sinhalese. The scholar-
ship of the Rev. N+an*aponika is unimpeachable and his piety
renowned; my purpose is merely to indicate the direction of his
influence. At least half of these missionary publications are the
works of Englishmen or Germans, many of them reprints of old
works, going as far Jack as <1The Light of Asia,>1 which Mrs. Besant
read in 1879.
  The aim of this excursus has been to suggest that there are in
Ceylon today broadly two types of Buddhism, the traditional and
the modern. It is not surprising if those Sinhalese Buddhists who
have been educated in English are themselves uncertain under
these circumstances about the validity of indigenous traditions,
and tend to cull their own Buddhism at least as much from
western sources as from local clergy and customs; and this may
be said with equal force of the Buddhist nationalist movements and
organizations which have arisen within the last ten years, always
under educated urban inspiration and leadership. I disagree with
those Europeans, like Copleston and Clifford, who consider the
last hundred years to have produced a Buddhist revival: it was the
previous hundred years, the period in which the three modern
Nika$yas were founded (see Introduction, p. 35, and Chapter 8,
pp. 307-9), in which Sinhalese Buddhism revived; this other
Buddhism is something new. When I say that Sinhalese Buddhism

56

is orthodox it is not to these recent movements that I refer; indeed,
I wonder whether they are not heading towards the first genuine
syncretism in Ceylonese Buddhist history.<s43>s
  It may seem disappointing that a work professedly on religious
change should devote its first chapter to saying that there has been
little change. Little, but not none : we shall see, especially in
Chapters 7 and 8, that there has been enough change to merit an
examination. But this relative changelessness has its compensa-
tions. For if the Buddhism we can observe today is like the
Buddhism of the Pali Canon, it follows, obviously, that the
Buddhism of the Pali Canon was like what we can observe today.
The Buddhism of the Pali Canon is at least 1,500 years, probably
1,900 years, old; how much older it may be we can only surmise.
The Sinhalese Buddhist villager today may have a religion quite
like the Buddhist villager nearly 2,500 years ago. With this
exciting thought let us examine that religion.

                                2
                     THE BASIC VOCABULARY OF
                             BUDDHISM

IN this chapter I shall try to assemble the elementary facts about
Sinhalese Buddhism and to introduce those technical terms which
are indispensable. To give some understanding of how a Sinhalese
Buddhist sees his own religion I shall preface my exposition with
remarks on the Sinhalese words used to denote the main religious
attitudes, thus revealing certain contrasts with, for example, the
vocabulary of English Christianity.
  The beliefs, behaviour, institutions, and language which I shall
describe in the rest of this work, and which for convenience I
refer to as Buddhist, are those of the Kandyan Sinhalese inhabi-
tants of Mi$gala and surrounding villages as they explained them
to me; in particular the beliefs are those of the local monks (see
Introduction, section 4). Detailed and systematic introductions to
Sinhalese Buddhist doctrine have been published in both English
and Sinhalese, notably in books which are used as textbooks in
Ceylonese schools. These books are authoritative, in the sense
that they are written by learned and distingnished Ceylonese
monks; they will doubtless mould the thinking of this and future
generations. They are essentially summaries of the doctrine of
the Pali Canon as understood by Buddhaghosa. Among such works
I cannot too highly recommend the Rev. Narada's <1The Buddha
and his Teachings>1.<s1>s My interest, however, was to find out what
a group of people actually believe, so I shall not refer to the Ven.
Narada's book, but use the information given me by the monks and
villagers themselves. Though I am mainly interested in Buddhist
beliefs and ethics, these cannot be described or understood with-
out some reference to Buddhist institutions, physical and social,
so I shall insert such descriptions at places where they may not too
much disrupt the argument.

  I am concerned with religious beliefs. But does our very con-
cept of religious belief correspond to a concept of the Sinhalese?

58

  The Sinhalese for "religion' is <1a$gama. A$gama>1 is a Sanskrit word
which in Sanskrit usually means "tradition', "sacred tradition', or
"sacred text'. It is not surprising that a people whose religion
depends on writings should use for religion a word which used to
connote texts. In Sinhalese, however, the word is fully naturalized
and no longer has this connotation except by etymology; it trans-
lates "religion' very well. Buddhism is <1Buddha$gama,>1 Hinduism
<1Hindu a$gama,>1 Islam <1Muslim a$gama,>1 etc.<s2>s
  The adjective derived from <1a$gama>1 is <1a$gamika,>1 meaning "reli-
gious', but it is not very idiomatic, nor is there an idiomatic word
for "secular'. These meanings are expressed by a periphrasis such
as <1a$gama vas/ayen,>1 "by way of religion', "to do with religion', and
<1melova vas/ayen,>1 "to do with this world'. Ames has written an
interesting paper<s3>s in which he discusses Sinhalese religion in
terms of Durkheim's categories of the profane and the sacred. He
claims that these categories exist in the Sinhalese terms <1laukika>1
and <1lokottara.>1 It is perfectly true that these two words would be
fair translations of "profane' and "sacred' respectively: <1laukika>1
literally means "of the world', <1lokottara>1 "above the world'. But
these terms are pure Sanskrit, and purely learned; I have never
heard them used in conversation, and to most villagers they are not
even intelligible. The sacred-profane dichotomy is not concep-
tualized by ordinary Silmalese--which is far from saying that he
does not know what is religion and what is not.
  "Religion' is easy to translate into Sinhalese: "belief' is a little
harder. The verb "to believe' or "to believe in' is most often
translatable by <1adahanava$.>1 But <1adahanava$>1 has a wider meaning.
Geiger's <1Etymological Glossary of the Sinhalese Language>1<s4>s gives
as translations "believe in', "trust in', "worship'. This covers what
seems like a whole range of religious attitudes. In the right
context <1adahanava$>1 does adequately translate "to believe': <1Budda-
game$  adahanava$>1--"(to) believe (in) the Buddhist religion';
<1Buddha$gamaya a%ttamayi kiyala$ adahanava$>1--" (to) believe that the
Buddhist religion is the very truth'. But unfortunately there is no

                                                                      59

noun which corresponds to <1adahanava$>1 as "belief' corresponds to
"to believe'. <1A%dahi$ma>1 will sometimes do, but it is basically a verbal
noun, "believing'; it is awkward and unidiomatic in the plural.
<1A%dahilla>1 is a more literary word, and is specialized to denote not
beliefs but acts of worship and devotion. Thus if one needs a
noun to express "beliefs' one will be forced to fall back on a more
general word. <1Adahasa>1 (which is etymologically unrelated to
<1adahanava$>1) most usually means "intention', but it can lose the
volitional element and mean "idea' or "opinion'; <1mataya,>1 slightly
less colloquial, means "opinion'. So in asking a Sinhalese his
"religious beliefs' a noun phrase would be <1a$gama ga%na adahas,>1 or
<1a$gama ga%na mata,>1 "opinions about religion'. But we can improve
on this by turning to verbs. <1Adahanava$>1 comes from Sanskrit
<1s/raddha$>1 ; the Pali noun <1saddha$>1 derived from this appears frequently
in the Pali Canon as a commendable quality; Sinhalese
<1s/raddha$va,>1 locally pronounced <1sardha$va,>1 is used in the same
context. <1Sraddha$/saddha$>1 has usually been translated "faith'. But
in her regrettably unpublished thesis for the University of
Ceylon, "The Role of the Miracle in Early Pali Literature', Dr. E.
Ludowyk Gyo%mro%i shows that this translation is a bad one.
Faith is a Christian concept; faith should lead to the acceptance
of doctrines which have not been accepted by reason. Christianity
admits that it contains doctrines which man will have to take on
trust. But Buddhism makes no such admission: not only are all the
truths it teaches intelligible to human beings, but doctrines should
even be rejected if reason finds them wanting. <1Saddha$>1 written
in religious context might therefore be better translated "confi-
dence' ; it is the confidence that Buddhist doctrine is correct, the
confidence indeed that permits the issuing of such a challenge.
It is trust, but not irrational or uncritical trust. Of course this
is on the cognitive level; that the trust is in fact sometimes
irrational or uncritical is not relevant here.
  The point that emerges from this etymologizing is that
Buddhism and Sinhalese do not make a distinction between faith,
i.e. taking something on trust, on the one hand, and any other
kind of belief on the other. The word which corresponds semanti-
cally in Sinhalese to Pali <1saddha$>1 is <1Vis/va$sa,>1 which means "trust'
or "confidence' in any context. It can be transformed into a verb
by adding <1karanava$: ma$ vis/va$sa karanava$>1 "(to) trust me'. <1Vis/va$sa
karanava$>1 can always function as a synonym of <1adahanava$>1 in the

60

sense of "believe': <1Buddha$gamaya vis/va$sa karanava$>1, "(to) believe
in Buddhism'.<s5>s But unlike <1adahanava$>1 it is not confined to religion:
<1Engelante$ hima va%t*enava$ kiya$ mama vis/va$sa karanava$>1--"I believe
it snows in England.' Faith or "fiduciary' belief is not distinguished
from any other belief, however obviously rational. A better way,
then, of asking a Sinhalese Buddhist about his religious beliefs
will be, "A$gama ga%na monava$da adahanne?' or "A$gama ga%na
mokadda vis/va$sa karanne?'--"What do you believe about religion?'
   Even these sentences, however, though linguistically unexcep-
tionable, still sound a trifle unusual. This is because culturally the
only religious beliefs which matter are the true ones : Buddhist
doctrine--<1Dharma>1. (This is the Sanskritic form; Pali <1dhamma>1 and
old SiMalese <1daham>1 or <1dam>1 are confined to special contexts.)<s6>s
<1Dharma>1 can be and has been translated in a thousand ways:
"righteousness', "truth', "the Way', etc. It is best not translated at
all. The scope of <1dharma>1 will be explained in the next section; it
does not cover all a Buddhist's beliefs, but if asked about his
religious beliefs he will take the question to refer to <1dharma>1.
   Let me interpose here that the above remarks are written from
the point of view of one seeking to translate English concepts into
Sinhalese; if we look at Sinhalese itself we must add an important
item to our vocabulary. <1Dharma>1 covers Buddhism's doctrinal side,
its belief system. The normal Sinhalese word for Buddhism's
institutional aspect, Buddhism as a spatio-temporal phenomenon
is <1s/a$sanaya. Buddha$gamaya, dharmaya>1, and <1s/a$sanaya>1 may thus
all be translated into English as "Buddhism'; and of these three
words <1Buddha$gamaya>1 is certainly the least frequent.
   For an investigation of the belief system of a religion I have
selected the term "belief' as crucial in English, and the words
<1adahanava$, vis/va$saya>1, and <1dharmaya>1 (which is more limited in
scope) as covering this semantic area in Sinhalese. Is it possible to
decide on a similarly central concept for a religion's action system?
I think English "worship' and Sinhalese <1van+dinava$>1 will serve our
purpose well: one is regularly used to translate the other, but their
differences are again instructive.

                                                                61

   Christianity is theocentric, Buddhism is not.  Christianity is
based on a concept of personality, that of Jesus Christ; for
Buddhists the Buddha and his personality are extremely impor-
tant, but their religion is based on a doctrine. There is a corres-
ponding contrast between the attitude of Buddhists and Christians
to the founders of their respective religions. Of course I am talking
here of their cognitive attitudes, the attitudes they express in words ;
whether their affective attitudes, their true feelings, are similarly
contrasted is a question which cannot be approached through
language, but which I shall return to at the end of the next chapter.
If a Christian were to be asked some such curiously vague ques-
tion as, "How do you stand to Jesus?', typical answers would be,
"I worship him', "I adore him', "I love him', "I trust him'. The
answers all speak as of someone still existent, which according to
Christian doctrine Jesus is, in that he is God, who is eternal. That
Jesus is God also makes it possible to say "I believe in him'. No
one will say "I believe in Julius Caesar': he lived, is dead, and that
is that. The Buddha is in the same case as Julius Caesar. We shall
see that affectively he may be something more; but cognitively he
is a dead man. The most we normally do for a dead man is to
revere his memory, and that is just how Buddhists, if asked, claim
that they view the Buddha. Perhaps they do love him in the way
that Christians love Jesus Christ; but they never say so. The
nearest they get to it is taking joy (<1pri$ti>1) in contemplation of the
Buddha or in making offerings before his relics; but these are still
the joys of pleasant thoughts, or even, as I was told, the pleasure of
looking at the beautiful flowers which are being offered.
  The typical verb for a Buddhist's attitude to the Buddha is
<1van+dinava$>1. In Ceylon this is taken as an exact translation of
"worship', which can be most misleading.<s7>s The verb has two

62

meanings, a physical gesture and a mental attitude. In the first
place it denotes <1any>1 respectful gesture of salutation. The least such
gesture is merely to place the palms together in the Indian gesture
called in Sanskrit <1an+jali>1, which is like the Christian gesture of
prayer. There are even gradations in the position of the hands--
the higher the hands the greater the respect. Before a superior this
is accompanied by bowing the head or the body, squatting or
kneeling. When kneeling one may unclasp the hands for a moment
and bend one's forehead to touch the ground, or the foot of the
person saluted. Complete prostration I never saw practised by a
Sinhalese, but it is used by Tamils before their gods, and gets the
same name in Sinhalese. <1Van#dinava$>1 is the name of such a gesture
in any context. Applied to the mental attitude which is obviously
supposed to correspond to this physical gesture, namely reverence,
the word is strictly religious: to use it of an attitude towards a
layman is extreme, and clearly metaphorical. Accordingly the
least misleading interpretation of the Buddhist's claim that he
<1van#dinava$>1 is to say that he "reveres' the Lord Buddha, and that
when he is in a temple or in front of an image in his own home he
behaves accordingly.
  I wrote towards the end of the previous chapter of the influence
of the West on Buddhism. Those remarks supplied the historical,
the above lines the conceptual background to the curious judge-
ments on Buddhism which were made to me by some of those
people, even in the village, whose education had brought them in
touch with western thought. I mention these judgements here to
contrast them with the traditional attitudes. The western mis-
understanding of Buddhism as atheistic explains the statement (in
English) of a Sinhalese estate clerk that Buddhism is "not a religion
but a practice'. Later in my stay the news that Buddhism is not a
religion reached almost every villager, many for the first time. The
Ministry of Education published a new primer for teaching
Buddhism in schools,<s8>s intended for study by six-year-olds, no
doubt the first book in a projected series. In the first paragraph of
the introduction occurs the sentence, "The Buddhist religion is
not a religion' ("Buddha$gamaya a$gamayek no ve$'). This is immedi-
ately amplified: "In the Buddhist religion there is no religion

                                                                   63

limited to offerings and devotions. The Buddhist religion<s9>s is a
way of life.' ("Puda pu$ja$, a%dahilivalat*a si$ma$ vu$ a$gamayek Buddha$-
gamaye$ na%ta. Buddha$gamaya ji$vana kramayeki.') Didactically this
may be admirable, but the amplification was not enough to avert
attention and wrath from the simple statement that Buddhism is
not a religion, a statement which through the very exigencies of
the Simhalese language stands in a self-contradictory form ! There
was a political storm: the Marxists who were in the coalition
government had obviously infiltrated the Ministry of Education;
this was part of their campaign to abolish Buddhism. Some of
those who started these charges may have been disingenuous, but
the Sinhalese villagers who heard them may be forgiven if not
only the six-year-olds were unable to grasp the sophisticated
claim that their religion was no religion.
  The youngest monk in Mi$gala temple was in much the same
current of opinion as the writers of the textbook. (The fact that
the monk considered himself a Marxist, however little he knew
about Marxism, suggests that the political charges against the
textbook may not have been quite groundless.) This monk had
spent the last ten years of his life being educated in the Low
Country, and had just got his B.A. from one of the two monastic
universities near Colombo. During these ten years he had never
returned to the village for more than a few days. Intellectually he
no longer had anything in common with the villagers or with his
two elder colleagues; when I appeared he seized me like a vulture
swooping on a lump of carrion. It was not so much that I was a
link with the outside world: his intercourse with me was his
chance to demonstrate his superiority, while to me he could air
his contempt for his surroundings. One day I appeared with a
small book, <1Bauddha A%dahilla>1 ("Buddhist Devotion'), a religious
pocket book so popular that in its various editions it must have
reached the majority of the literate Buddhist public, at least
among the middle classes. It contains all the Pali verses which a
layman is supposed to know or need, with Sinhalese translations
and explanations.<s10>s I asked him if he had read it. He told me with
great emphasis that he had never even looked at it. That was all

64

just <1vis/va$saya>1 ; true Buddhism was the <1dars/ane$>1--philosophy. The
young monk later told me that all the worshipping, etc.--what
might be summed up as "popular Buddhism'--was a Maha$ya$na
accretion. After six months he had made himself so unpopular--
not through his views on religion, which he prudently kept to
himself, but by his arrogant and insolent behaviour--that he was
more or less forced to leave the village and return to Colombo.
   Warned by his fate, let us approach the subject through the
categories of the villagers themselves. To give a succinct account
of an interlocking system is difficult. I shall try to summarize very
briefly what classes of Buddhists there are and what they believe.
I shall conclude the chapter with a general description of the
physical context of village Buddhism, namely the temple. In
subsequent chapters I shall expand on details of this summary.


   The most frequent religious act of Buddhists is the recitation of
a few lines in Pali. Many Buddhists recite them daily in private;
their recitation begins, and often punctuates, every public religious
occasion. Here they are, with approximate translation.

<1Namo tassa Bhagavato arahato samma$ sambuddhassa>1.
Worship to the Blessed <1arhat>1<s11>s truly fully enlightened. (Three times.)

<1Buddham* saran*am gaccha$mi>1.
I go to the Buddha for refuge.

<1Dhamnam* saran*am* gaccha$mi>1.
I go to the Doctrine for refuge.

<1Sangham* saran*am* gaccha$mi>1.
I go to the Order for refuge.

<1Dutiyam* pi Buddham* saran*am* gaccha$mi>1.
And a second time I go to the Buddha for refuge.

<1Dutiyam* pi Dhammam* saran*am* gaccha$mi>1.
And a second time I go to the Doctrine for refuge.

<1Dutiyam* pi Sangham* saran*am* gaccha$mi>1.
And a second time I go to the Order for refuge.

<1Tatiyam* pi Buddham* saran*am* gaccha$mi>1.
And a third time I go to the Buddha for refuge.

                                                      65

<1Tatiyam* pi Dhammam* saran*am* gacchami>1.
And a third time I go to the Doctrine for refuge.

<1Taliyam* pi Sangham* saran*am* gaccha$mi>1.
And a third time I go to the Order for refuge.

<1Pa$na$tipa$ta$ veraman*i$ sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from taking life.

<1Adinna$da$na$ veraman*i sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from taking what is not given.

<1Ka$mesu miccha$ca$ra$ veraman*i$ sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from wrong conduct in sexual desires.

<1Musa$va$da$ veraman*i$ sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from telling lies.

<1Sura$meraya-majja-ppama$da-t*t*ha$na$ veranan*i$ sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from intoxicating liquors which
  occasion heedlessness.

The first line is a salutation to the Buddha. The next nine lines are
the Three Refuges (<1tisaran*a>1) thrice repeated. The last five lines
are the Five Precepts (<1pan sil>1).
    A person who takes the Three Refuges and Five Precepts is
thereby a Buddhist layman. There is no ceremony for conversion
to Buddhism beyond the recital of these lines, so anyone who says
these words and means them can rightly call himself a Buddhist.
To go so far and then to keep the precepts is considered sufficient
for great religious progress. A Buddhist layman is called an
<1upa$saka>1 (feminine: <1upa$sika$>1)--at least in religious discourse ; in lay
idiom the term is generally reserved for people of notable piety.
The Five Precepts should be recited daily (at least), and kept always.
  The status of <1upa$aka>1 denotes a religious attitude, a certain
relationship to the Dhamma. An <1upa$saka>1 is also supposed to
support those of higher religious status than himself (principally
monks), and by virtue of this relationship to the Sangha he is
called a <1da$yaka>1 ("giver'). This sounds like a function. It is possible
for a <1da$yaka>1 not to be an <1upa$saka>1 (i.e. for a non-Buddhist to
make gifts to monks, etc.), though of course this is not normally
envisaged. However, <1da$yaka>1 is also a status. When an English
parson would refer to his "parishioners', a monk calls the laymen in
his village not <1upa$sakayo$>1 but <1da$yakayo->1.

66

  Buddhists have the custom (no doubt it is pre-Buddhistic) that
certain days are holier than the rest. These days, called <1poya>1 days,
are the quarter days of the lunar month: full moon, no moon,
half moon. They are holy in that order; all the crucial events in
the Buddha's life are believed to have happened on full moon days.
Theoretically on any day, but by custom only on <1poya>1 days, a
layman may take the Eight Precepts (<1at*a sil>1). These are the Five
Precepts with this addition:

<1Vika$la-bhojana$ veraman*i$ sikkha$padam* sama$diya$mi->1.
I undertake the precept to abstain from eating at the wrong time.

<1Nacca-gi$ta-va$dita-visu$ka-dassana-ma$la$-gandha-vilepana-dha$rana-
  man*d*na-vibhusana-t*t*ha$na$ veraman*i$ sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from seeing dancing, music vocal and
  instrumental, and shows; from wearing garlands, perfumes and
  unguents, from finery and adornment.

<1Ucca$ayana-maha$sayana veraman*i$ sikkha$padam sama$diya$mi>1.
I undertake the precept to abstain from high beds and big beds.

  The sixth precept means no solid food after midday. The
seventh means dressing in plain white, with no more ornament
than maybe a rosary of 1o8 brown beads as a necklace. The eighth
means sitting or lying only on mats on the ground. Traditionally
it has been customary only for elderly people to take the Eight
Precepts, except perhaps once a year at Wesak, the full moon in
May, the anniversary of the Buddha's birth, Enlightenment, and
death. Hence those laymen taking the Eight Precepts are com-
monly known as "laywoman mummy' and "layman daddy'--
<1upa$sikamma$>1 and <1upa$sakappucci>1 were the local terms.
  There is a further stage possible for laymen: the Ten Precepts
(<1dasa sil>1). These ten are formed by splitting the seventh into two,
and adding

<1Ja$taru$pa-rajata-pat*iggahana$ veraman*i$ sikkha$padam* sama$diya$mi>1.
I undertake the precept to abstain from accepting gold or silver.

In fact, although to do so is theoretically possible, no one takes
the Ten Precepts for a limited period of time. They involve the
same abstentions as those practised by the clergy, so men of such
piety are likely to enter the Order, unless they are too old to make
the break. Laymen (male) who take the Ten Precepts (<1dasa sil
upa$sakayo$>1) wear white. They may be "home-dwelling' (<1gedara>1

                                                                     67

<1inna>1 or <1gr*hastha>1) or "homeless' (<1anaga$rika)>1. Anaga$rika Dharma-
pa$la (see p. 42) was a "homeless' layman, but I have never
heard of a villager's becoming one. Nor did I meet even a <1gr*hastha
dasa sil upa$saka>1, but I heard of two in my area, probably elderly
widowers. On the other hand, the Order of nuns (<1bhiksun*i$ sangha>1)
being extinct, to take the Ten Precepts is the highest religious
status a woman can attain. Such women shave their heads and
wear yellow, sometimes (informally) white. There are a few urban
communities of them, but in villages they are infrequent and
usually solitary. On the edge of Mi$gala were two, who lived in
adjacent caves. Though one of them was not old, they were known
as <1upa$saka ma%$n*iyo$, ma%$n*iyo$>1 being an extremely honorific word for
"mother'. Such ladies, like monks and Dharmapa$la, renounce their
lay names and take Pali names, to which they add the title <1silma%$n*iyo$>1.
  On entering the Order (<1mahan*a vi$ma>1) one becomes a novice.
Novices are usually boys, and known as "little monks' (<1podi
ha$muduruvo$>1 or <1pun+ci ha$mudaruvo$>1).<s12>s A novice has to take and
keep the same Ten Precepts. Even a monk has to do no more:
he has to submit to 227 rules, but these rules are an elaboration,
not an extension, of the precepts, and their detailed application
to the life of a monk. Monastic life has its duties and even its rites,
but most of these can be avoided by becoming a hermit; in the
way of morality the Ten Precepts sum up all that is asked of a
Buddhist. Monks and novices wear robes of some yellow or
orange (no significance attaches to the shade). There are various
words for monk,  most of which have the honorific ending <1vahanse$>1
--literally "sandal-shadow', but the meaning has been forgotten.
Related to the Pali <1bhikkhu>1 is <1bhiksunvahanse$>1. A rather surprising
word for "monk'  is <1sanghaya$>1 ( +- <1vahanse$>1), which in form is merely
Sinhalese for <1sangha>1, the Order, which indeed is its usual meaning
in the literary langnage. But the common words in the village, for
both reference and address,<s13>s are <1sva$minvahanse$>1 and <1ha$muduruvo$>1,
both meaning "lord', but no longer used to denote temporal power.

  The objects of the Three Refuges are the Three Jewels (<1triratna,
teruvan):>1 most people asked to explain Buddhism (<1Buddha$gama>1)

68

would begin here. The Buddha is the founder of the religion, the
Dharma is the truth he discovered, the Sangha is the vehicle for
preserving and propagating that truth.
   For the moment the Buddha and the Sangha require no further
explanation than to say that doctrine views the Buddha as a
human being who is now dead, but whose relics may be venerated,
and the Sangha as other human beings who are supposed to
emulate him and are venerated for this reason. But the Dharma
must be further discussed. I have said that if you ask a Buddhist
his religious beliefs he will assume you are talking of Dharma.
But these beliefs operate in a context of other beliefs, of more
basic assumptions. This is true both logically and historically: the
Buddha grew up in a Hindu society and accepted many Hindu
assumptions. A few words about these assumptions are necessary.
  The universe is full of living beings, in hierarchically ordered
strata. Men are somewhere in the middle (and have their own
hierarchically ordered strata--castes). Above them are various
classes of gods and spirits, below them are animals, ghosts, and
demons. Above this world are heavens, below this world are hells.
By and large, power, well-being and length of life increase as one
goes up the scale. So do the power and inclination to do good.
But at all levels there is death, the ineluctable reminder of the
unsatisfactoriness of life. Death supplies the mobility between the
different levels. Everywhere, constantly, are death and rebirth.
One's station at birth is determined by <1karma. Karma>1 is a Sanskrit
word simply meaning "action', but it has acquired this technical
sense. It is the principle of "Be done by as you did'; a good action
will improve your station, a bad action depress it. One's <1karma>1
may affect one's fortunes in other ways besides one's station at
birth, but this is its most important effect. All this is accepted by
all kinds of Hindus and by Jains--by all the major Indian religious
systems. However, Buddhism was the first system completely to
ethicize the concept.  For Buddhists <1karma>1 consists solely of
actions morally good or bad, not of other actions such as ritual.
  <1Karma>1 is the most powerful, large-scale system of causation.
(Buddhists do not believe that any creature, however exalted, has
the power to override <1karma>1.) Operating within this system are
others, which affect the details of specifically human affairs. Such
systems are astrology and the (free-will) activities of supernaturals.
These or similar systems are again accepted, with or without

                                                                69

inconsistency, by the other major Indian systems. In Sinhalese
Buddhist ideology these systems fit into each other without
inconsistency, as will be explained in Chapter 4.
  The belief system described so far is not <1Dharma;>1 it is just the
outline of a cognitive map of the universe, a map which the
Buddha and his followers have not called in question. Taking all
this for granted, the Buddha saw the truth--the Dharma.
  In one formulation there are Four Noble Truths. They are

(1) Unhappiness.
(2) The arising of unhappiness.
(3) The destruction of unhappiness.
(4) The path leading to the destmction of unhappiness.

  (1) The word I have translated as unhappiness is Pali <1dukkha>1,
Sanskrit <1duh*kha>1, Sinhalese <1duka. Dukkha>1 notoriously lacks an
English equivalent, though the concept is not obscure. It is simply
the opposite of <1sukha>1 (in Sinhalese of <1sa%pa>1), which means "well-
being'.  Unfortunately "ill-being' is not an English locution.
Translators have used "suffering', "unsatisfactoriness', etc. "Frus-
tration' would be philosophically accurate, but sounds too petty.
"Unhappiness' is also inadequate because it is only mental, not
physical. Armed with this apology, I shall use <1dukkha>1.
  In the Buddha story <1dukkha>1 is symbolized by an aged man, a
sick man, and a corpse. But it is far wider than the positive ills of
disease and death. It is anything that crosses us, it is not getting
what we want. It even exists for the normally content, because
he wants to go on living, but must face the prospect of death.
Death shows that all goods are impermanent (<1anicca)>1. In this sense
<1dukkha>1 exists even in the highest heaven and for the greatest god.
The Sinhalese often say "ji$vite$ dukayi'--"life is suffering' or just
"life is sad'.
  (2) <1Dukkha>1 exists because of desire: it is desire's corollary. If
we have no wants we shall suffer no disappointments. There are
many words in Pali for this desire: a common one is <1tan*ha$>1--thirst.
The usual one in Sinhalese is <1a$sa$va>1. This derives from Sanskrit
<1a$s/a$>1, which is often translated "hope'. It is desire, notably the
desire for life, which gives the impetus for rebirth.
  (3) The destruction of unhappiness is consequently the des-
truction of desire. Those who cease to desire life will not be
reborn, and experience no more <1dukkha>1. They, and they alone,

70

experience liberation (<1moks*a>1) from the wheel of rebirth, i.e.
phenomenal existence (<1sam*sa$ra>1), and attain <1nirva$n*a>1. Etymologl-
cally <1nirva$n*a>1 means "blowing out'. It is indescribable, being
beyond the realm of words, which describe the world we know;
it is defined negatively. The Sinhalese usually talk of "seeing' or
"gaining' <1nirva$n*a (nivan dakinava$>1 or <1labanava$>1); it is an intuition,
a sudden awareness, a mystical state. Enlightenment comes during
the course of a lifetime ; one who has it is an <1arhat>1 (Sinhalese <1rahat>1,
respectfully <1rahatanvahanse$)>1. An <1arhat>1 lives out the fated span of
his physical life; its end is called <1parinirva$n*a>1 (Sinhalese
<1pirivan)>1.<s14>s
  (4) The path leading to the destruction of unhappiness is the
Noble Eight-fold Path. Rough translations of its eight constituents
are 1 . Right views. 2. Right resolve. 3. Right speech. 4. Right
action. 5. Right livelihood. 6. Right effort. 7. Right mindfulness.
8. Right meditation. These eight are often summarized as <1si$la,
sama$dhi, pan+n+a$. Si$la>1 (morality) covers nos. 3-5; <1sama$dhi>1 (medita-
tion) nos. 6-8 ; <1pan+n+a$>1 (wisdom) nos. 1 and 2. is Each of these
three is to some extent a prerequisite for the next. Nos. 1 and 2
come both at the beginning and at the end : intelligence and
motivation are needed to make a start; at the end they ripen to the
true wisdom of Enlightenment.
  The Four Noble Truths and the Noble Eight-fold Path are
described in the Buddha's first sermon.<s16>s This short text makes
one other point, with which indeed it begins. Because to see the
Truth one must steer between indulgence and moltification of the
senses, Buddhism is the Middle Path.
  The path described by Buddhist doctrine is ethical, intellectual,
and mystical. The innovations of the Dharma are very simple, and
consist mainly of a recipe for action. This is easy to grasp. None
of the doctrines I have written of so far would be unfamiliar to a
villager. He could probably not list the Four Noble Truths or the
constituents of the Noble Eight-fold Path; he might not even know
the titles of the doctrines. But the substance of the argument he
would know; indeed he could probably express most of it himself
if occasion arose. If asked to describe his religion he will mention

                                                                   71

most of what I have mentioned, from the Three Jewels on, prob-
ably with some more detail on the Buddha and on good and bad
deeds. The Buddha I shall deal with in the next chapter. Before
I  say more on good and bad deeds there is one metaphysical
question to be disposed of. It is too difficult for many villagers to
hold views on it, but it is a necessary logical link for a basic under-
standing of Therava$da doctrine, so I had better dispose of it
here.
   What is it that transmigrates?  Buddhist philosophy (<1abhi-
dhamma>1) revolves round <1anicca, dukkha>1, and <1anatta>1, which are said
to be three facets of the one truth. <1Dukkha>1 and <1anicca>1 I mentioned
above under the first Noble Truth. <1Anatta>1 is Pali for "non-self'.
In philosophy it came to be applied to everything, as a denial of
essence, but originally it was merely a denial of the existence of a
soul, an immortal part in living beings. Therava$da Buddhism is
<1ana$tmava$da>1--the doctrine of no soul; this much is known to
innumerable people who could not explain it. Most monks, how-
ever, can explain that <1a$tman>1, self, is merely the name of an aggre-
gate of mind and body which dissolves at death; it has no indepen-
dent existence. What survives is <1karma>1, the effect of actions. If I
may supply my own illustration, every act for good or ill is like a
boomerang, which comes back at you. Whether you throw the
boomerang is a matter of free will, but once it is thrown conse-
quences must follow. Buddhist doctrine holds that if I die before
the boomerang gets back someone else will have been born in
my place and reap those consequences ; moreover the nature of that
someone is conditioned by what I was.
   Plainly this is a rather subtle and difficult doctrine. The Buddha
himself is supposed to have realized that his teaching was "against
the current'.<s17>s Even so, this doctrine understood only by the few
has permeated the language. Sinhalese has no word for soul, in
the sense of an immortal part or adjunct of a human being. The
Sanskritic word <1a$tma>1 is still in use, but it has acquired the meaning
of "life-span'. "[I shall not do it] in this life'--"me$ a$tme$ di$'. "[We
met] in a previous life'--"issara a$tme di$'.
   I asked the incumbent of Mi$gala temple about <1a$tma>1 and rebirth.
We have no <1a$tma>1, he said, because <1a$tma>1 implies something change-
less, and we change all the time. Even if we become gods we shall

72

still be constituted of the five components (<1skandhas>1); we shall
have lost only the name of man (<1ma$nus*yana$ma)>1. We use the word
<1a$tme$>1, and talk as if it existed, for convenience (<1pahasu pin+isa)>1. This
is the realm of conventional truth (<1sammuti satya>1); ultimate truth
is different ("parama$rtha satya venayi'). The being who is born is
neither the same as, nor different from, the one who died: he
quoted the Pali "Na ca so, na ca anyo' ("Both not he and not
another').  The relation is like that between caterpillar and
butterfly.
   The doctrine of no-soul, <1ana$tmava$da>1, is canonical according to
the interpretation of Buddhaghosa.<s18>s
   But does not death nevertheless mark a decisive break? The
doctrinal answer is no: our thought and even our bodies are
changing at every moment. The element of continuity is pre-
cisely <1karma>1, the force of past actions; <1karma>1 not merely binds
one life to the next, but one collection of the <1skandhas>1 to the col-
lection existing in the next moment. Just as a man can remember
what he did yesterday, he can, with training, remember what he
did in a former life; the difference is merely one of degree.
   Although there were probably not more than half a dozen monks
in my area who could have expounded the argument stated in the
previous paragraph, I am sure all would have assented to it. It is
the doctrinal position which everyone claims to accept, however
little they are able to articulate it. On the other hand, people
certainly talk as if they conceived a series of lives in terms of
survival of the personality ; terms such as "My next life', sentiments
such as "We must have met in a former life', are freely expressed.
In other words, a Buddhist seems to think of rebirth in very much
the way that the concept first appears to a westerner brought up

                                                                  73

with the idea of a soul which survives death. Is the Buddhist then
not being logically inconsistent? I think not. If it is merely con-
ventional truth (<1sammuti satya>1) to talk of "I' when in reality
(<1parama$rtha vas/ayen>1) I am nothing but a series of groups (<1skan-
dhas>1) strung together on the string of <1karma>1, why should it not be
equally permissible to use convention to talk of "my next life'?
Logical inconsistency can only arise between cognitive positions;
on the cognitive level the Buddhist is being consistent.
  Where the clash may, and I think does, come is between the
cognitive and the affective level. Despite the doctrine that we are
but a series of groups connected by actions, people do in fact
think of themselves as having a more or less stable and concrete
existence. From this it is a short step to conceiving this existence
as extending beyond death. That people affectively believe that
<1they>1 will survive death and be reborn cannot, by the nature of what
I called affective beliefs, be demonstrated directly; but I think it
can be inferred from other statements and actions, especially
those discussed in Chapter 5. Moreover, I think that this affective
belief in personal survival, clashing with the cognitive belief in
merely karmic survival, is the basis for a whole system of affective
religion which diverges from official doctrine. I shall return to
this in my last chapter.
  It would have been misleading to leave the impression, even in
a summary, that villagers grasp all the subtleties of orthodox
doctrine on <1karma>1 and rebirth. It would be equally misleading to
leave the impression that <1nirva$n*a>1 is their immediate religious goal.
I doubt whether this would be true of anyone I met, cleric or lay.
Whether interested in <1nirva$n*a>1 as an ultimate goal or not, people
set their sights on whatever they consider the most desirable
rebirth. What makes for a good rebirth is merit (<1pin>1), and what
makes for a bad rebirth is demerit (<1pav>1). <1Pav>1 is usually translated
"sin', but "demerit' brings out its neat opposition to <1pin. Pin>1 and
<1pav>1 are conceptualized as quantifiable and (in a sense--see
Chapter 5) transferable. If you ask a villager how he acquires <1pin>1
or <1pav>1 the commonest answer will be, by keeping or breaking the
Five Precepts. The Five Precepts are, however, negative : there is
a list of Ten Good Deeds, which is not canonical, but widely
influential. In an unmethodical way this list subsumes the precepts
and includes all possible ways of earning <1pin>1. The Ten Good
Deeds (<1dasa kusala karma>1) are listed in a Pali stanza:

74

                <1Da$nam* si$lan+ ca bha$vana$>1
                <1Patti patta$numodana$>1
                <1Veyya$vacca apaca$yan+ ca>1
                <1Desana$ suti dit*t*hiju>1.<s19>s

These were paraphrased for me by a monk as:
  Giving (material), keeping morality (i.e. the precepts), meditat-
ing, giving (transfering) merit, rejoicing in (another's) merit,
giving service, showing respect, preaching, listening to preaching,
right beliefs (<1samyak dr*s*t*iya--v/is/va$saya)>1. In an admirable article<s20>s
J. F. Dickson gives a succinct explanation (pp. 2o3-7) of each of
these terms; I could not improve on his account, and refer to it
the reader who would like a fuller preliminary description before
I deal in Chapters 5 and 6 with specific problems arising from the
list. The term <1pinkama>1 literally means "act of merit', and so could
apply to any item in this list; however, as Dickson remarks
(p. 207), idiomatically the term is only applied to the outward
manifestation of merit-earning--the public, not the private event.
Canonical Buddhism is a religion for the individual: excepting
only a few simple ceremonies for monks it neither prescribes nor
contains any ritual. From the beginning there were inevitably
religious occasions of a public nature, i.e. involving more than one
person: the two typical occasions are feeding monks and preach-
ing, which very often go together. No ritual was laid down for
these occasions, but already in the Canon, which after all consists
mainly of sermons and is full of invitations to meals, they are
described in uniform language and the procedure is plainly
standardized. In modern Ceylon any public Buddhist occasion is

                                                                75

called a <1pinkama>1. The typical <1pinkam>1 are still feeding monks
(<1da$ne$>1--the word literally means "giving' but has become a tech-
nical term) and preaching (<1ban*a>1), plus one related to <1ban*a>1, a
particular kind of recitation of sacred texts called <1pirit>1. The
morning meal is called <1hi$l da$ne$>1, the (main) midday meal <1daval
da$ne$>1, the evening snack <1gilampasa da$ne$. Hi$l da$ne$>1 should be given
at about 7 a.m., <1daval da$ne$>1 at about 11.30 (to be over by noon),
and <1gilampasa>1 normally between 5 and 6, but <1gilampasa>1 may also
be served just before the beginning of a <1pinkama>1 in the early part
of the night. For a <1da$ne$>1 or <1ban*a>1 the ritual is usually scarcely more
elaborate than in canonical times.  There are, however, other
<1pinkam>1, of various degrees of canonical authenticity, and indeed
villagers apply the term to any public activity in which monks take
part: when just before the national election sixty monks gathered
at Mi$gala temple and walked in procession to the town council
grounds to hold a political meeting, this was a <1pinkama>1 too.


  Most big <1pinkam>1 take place at the temple, and in order to give
a general picture of what a <1pinkama>1 involves I shall conclude this
outline with a brief description of the physical setting of Sinhalese
Buddhism. Most villages--and the proportion is increasing--
contain what, following Ceylonese practice, I call a temple. The
language of Christian culture is again misleading. The essential of
a Buddhist temple, the fact by virtue of which it is in use, is that
one or more monks live there. The building in which the monks
live, which is normally an ordinary bungalow like a layman's, is
called a <1pansala>1. Typically a temple consists of a group of features,
but any or all of these features may be absent except the <1pansala>1.
When there is nothing but a <1pansala>1, which in practice means that
the temple is new or inhabited only part-time, the temple as a
whole is called an <1a$va$saya;>1 otherwise, i.e. normally, it is called a
<1viha$raya>1. The properly equipped temple of a certain village, e.g.
Mi$gala, is in conversation always referred to as, e.g. <1Mi$gala
viha$raya>1. Formally, however, the <1viha$raya>1 usually has some given
Sanskrit name, and this almost invariably ends in <1a$ra$maya:>1 e.g.
"S/ri$  Saddharma$ra$maya'--"The Blessed Temple of the True
Doctrine'. The average number of monks in a temple in the Mi$gala
area is two, though single monks are very frequent. There are
usually anything up to four novices, but they spend most of their
time away at school. (An educational establishment for monks is

76

called a <1piriven*a>1.) A temple servant (<1a%bittaya$>1) also often lives in the
<1pansala>1.
  The principal common features of a <1viha$raya>1, besides the
<1pansala>1, are : a building containing religious art, including at least
one sculpture of the Buddha; a peepal or sacred fig tree, locally
known as a Bo tree; a relic mound, in English usually called a
stupa; a preaching hall; and a building for monks to perform their
communal rites (<1vinayakam)>1. Of these only the first two were
virtually constant features in the area where I worked, though the
third, the stupa, an almost invariable feature of Low Country
temples, was also very common. The first contains, the second and
third are themselves objects of veneration.
The building containing religious art (for which I can find no
better English than the awkward locution "image-house', a literal
translation of <1pat*ima$ghara>1, its Pali name) is, confusingly, also
called a <1viha$raya>1, or <1viha$ragedara>1 or <1viha$rage$ (gedara>1 and <1ge$>1 are
usually interchangeable synonyms meaning "house').  It is of
varying size, ornamentation and complexity. A Buddha image,
often over lifesize and invariably seated or lying, is opposite the
entrance door of the innermost shrine, and before it is always some
kind of ledge or table on which offerings, principally flowers, can
be laid. In theory, and usually in practice, the image contains a
relic. Anyone who comes to the temple for religious purposes
comes to pay his respects before this image. There are other
statues, usually including Buddhas and gods, and the walls are
painted with similar figures and with decorations. Almost invari-
ably the entrance to the <1viha$ragedara>1 or to the shrine room (some-
times to both), and often the main Buddha image too, is framed
by a <1makara toran*a>1 ("sea-monster arch'), a traditional arrangement
of highly stylized motifs surmounted by a monster's head, <1en face>1,
with mouth agape<s21>s
The Bo tree (<1Bo$ gaha, Bo$dhinvahanse$>1), is the kind of tree under
which the Buddha attained Enlightenment (<1bo$dhi>1, old Sinh. <1bo$>1);
moreover, many trees are claimed to be grown from cuttings of the
Bo tree at Anura$dhapura, which is itself grown, according to the

                                                                      77

<1Maha$vam*sa>1, from a cutting of the original tree. It is usually sur-
rounded by a parapet (<1ba%mma>1), which includes a ledge or altar
(<1mal a$sana>1) for offerings of flowers. Many of these trees are old and
magnificent, and I suspect that often they have determined the
situation of the temple.
  The stupa has many names. <1Stu$pa>1 is itself Sanskrit. Old Sinha-
lese is <1da$gaba>1 or <1da$ga%ba>1, which indicates precisely what it is, a
"relic-container'. Another old Sinhalese word, now confined to
proper names, is <1sa%$ya>1 (derived from <1caitya)>1. The words I found
mostly used were <1caitya>1<s22>s and <1vehera. (Vehera>1 is merely the old
Sinhalese form of <1viha$raya>1, another specialization of this versatile
word!) The stupa probably existed in India in pre-Buddhist times
as a burial mound. When the Buddha died his relics were distri-
buted and buried in such mounds, and it soon became the custom
to build them for other holy monks (<1arhats>1) too.<s23>s In Ceylon they
are invariably built of solid masonry (with a small sealed chamber
in the centre for the relics), and their shape is standardized. On a
base, usually circular or octagonal, rises a dome, topped by a cube
(<1haraskot*uva>1) from which rises a thin spire which is usually of
bronze. Modern stupas do not approach the scale of the ruins of
Anura$dhapura and Pol*onnaruva; in villages they range from
perhaps fifteen to forty feet in total height. By a stupa again it is
usual to build at least one flower-altar.
  The preaching hall (<1ban*a mad*uva>1 or <1ban*a gedara>1) is a simple
building, usually square with walls only to half height so that
people outside can see and hear what is going on inside. The
central part of the floor is slightly raised; here monks sit (though
laymen may do so too), and the monk who is preaching has a
special chair, or one covered with the richest cloth available.
   The building for monastic ceremonies, called a <1poya gedara>1, may
be of any sort, and often one of the other buildings is used for this
purpose. Normally, a <1poya gedara>1 is enclosed, as acts of the clergy
are private.

78

  Most buildings are washed white or blue, and red-tiled roofs
are common. In the hill country most temple premises are on a
series of terraces, surrounded by foliage, and during a <1pinkama>1
people stream up and down the steps. Ideally at the temple they
wear white, but many wear their ordinary clothes of any colour,
the women usually bright. The monks wear their robes, which
range from yellow through orange to red. The flower offerings and
five-coloured Buddhist flags add to the gaiety. <1Pinkam>1 are noisy
specialists provide a lot of drumming (on drums called <1he$visi>1 and
<1bera>1) and playing on a piercing wind-instrument not unlike an
oboe (called <1horan*a%$va)>1. Proceedings are punctuated by cries from
the crowd of "Sa$dhu', which means "good' or "holy', and is used
rather like the Christian "Amen' to express participation and
assent, but it tends to come in a triple burst of "Sa$dhu sa$dhu
sa$a$a$', which feels more cheerful, like "Hip-hip-hooray'.
  Both monks and the <1viha$ra>1 fixtures are treated with decorum,
which is not the same as solemnity. The decorum extends to the
treatment of certain other Buddhist objects. The preaching chair
(<1dharma$sana>1) has just been mentioned. When preaching a monk
has a special kind of fan (<1vat*a$pata>1) with a decorated handle, which
he sometimes holds to conceal his face. When I was given such a
fan and had it in my house the villagers insisted that it be placed
high up, and did not like my wife to handle it. But an even more
sacred object is a manuscript of a sacred text (<1ban*a pota)>1. (By mass
production printed books--and pictures--have been largely
desacralized, though a village neighbour did put a tattered book
of religious verse (<1kavi>1) which he lent us on the flower ledge just
below the picture of the Buddha.) The times are past when,
probably under Maha$ya$nist influence,<s24>s sacred texts inscribed on
metal would be buried in stupas; but the palm leaf manuscripts
are kept wrapped in rich cloths, and when they are taken out for
<1ban*a>1 or <1pirit>1 they are borne, like relics, on a man's head.
   The rest of this book will have little to say about communal
worship, which is neither doctrinally nor behaviourally nearly as
important in Buddhism as for instance in Christianity. Pilgrimages
apart (see pp. 108-12), there is very little concerted action involved
in Buddhist worship. Monks may organize committees to promote
temple welfare, but even fund-raising tends to be <1ad hoc>1, and to
most temples meals are supplied according to a roster of house-

                                                                  79

holds only in the four months of the "rainy season' of the liturgical
calendar. The biggest effort of co-operation probably goes into
supplying the food for a major <1pinkama>1 at the temple ; on such
occasions huge pots of rice, innumerable curries, cakes, sweet-
meats, and fruits are carried up to the temple, and there is vast
waste, from which beggars profit. Then the temple is thronged
with people, many of them eager to help serve the food, and
several hundred people may take the Five Precepts, invite the
gods to share their merit, and assent to the monk's wish that they
should all be reborn in the time of the next Buddha, Maitri$, and
under him realize <1nirva$n*a>1. On arrival most of these people went
up to the <1viha$ra>1 with a few flowers and worshipped the main
Buddha image, and maybe the Bo tree and <1caitya>1 too, perhaps
muttering a few Pali verses, or merely remaining squatting with
hands folded to forehead and head bowed; after the <1da$ne$>1 many of
them evaporate before the sermon, and the congregation becomes
even more predominantly feminine, though to be sure a few elderly
men in white are always conspicuously zealous. Even on a major
full-moon day I have not known more than a hundred people, or
more than twenty adult males, to take the Eight Precepts and
spend most of the day at the temple, repeating Pali verses after a
monk early in the morning, hearing a sermon at midday and
maybe again in the evening, and in between chatting or meditating
by droning Pali verses over a rosary. In towns of course the pious
bourgeoisie are more assiduous.

                                3
                            THE BUDDHA
                     I. THE BUDDHA'S BIOGRAPHY

GOTAMA BUDDHA, under whose teaching (<1s/a$sane$>1) we live, was
the twenty-fifth in the line of Buddhas.<s1>s The twenty-sixth will be
Maitri$ Buddha and we do not know who or how many will come
after him. The first Buddha was called Di$pam*kara, and it was on
seeing him that our Buddha, who in that life was a young brahlnin
ascetic called Sumedha, made the vow that he too would become
a Buddha. A being who makes such a vow is called a Bodhisattva
(Sinh.: <1Bo$sat)>1. The 550 lives which our Buddha lived as a <1Bo$sat>1
before he was born as Gotama are the subject of the book of
<1Ja$taka>1 stories. The introduction to this book is called the <1Nida$na-
katha$>1. It begins with the story of Sumedha, and then gives the life
of Gotama (his family name), both before enlightenment, when
he was a prince called Siddha$rtha (Sinh.: Siduhat), and the first
two years after it, when he can properly be called the Buddha.
The relationship of the <1Nida$nakatha$>1 to other biographies of the
Buddha, which are in Sanskrit or based on Sanskrit originals, was
long in doubt. Recently Frauwallner has shown that all the bio-
graphies probably derive from a lost one composed little more
than a century after the Buddha's death. Therefore, although
nearly all current Sinhalese beliefs about the Buddha's biography
are based on the <1Nida$nakatha$>1 and closely related texts (see below,
p. 100), the main points of the story are of extremely ancient
origin.
   Sinhalese Buddhists today say the same as the Canon. The
Buddha was a man; his status was human, not divine. He was of
course the best possible man in every respect, a man of superla-
tives: born in the best caste, living the most comfortable life,

                                                                81

capable by destiny and talent of ruling the world as a temporal
monarch. He had a wife and son. By leaving all this he therefore
made the greatest possible renunciation. In his search for the
truth he then underwent the most extreme austerities. Even after
his Enlightenment he was just a man. It is true that he performed
some miracles, or rather wonders (for miracles imply the humanly
impossible), and had marvellous powers ; but these wonders are
effected by the power of truth, the truth which it is open to any
man to realize, and the powers likewise are those of any <1arhat>1--
indeed some of the more trivial ones, such as levitation, may be
attained by meditation even before the final goal of <1nirva$na>1 has
been reached.
   Siddha$rtha was 29 when he renounced the world. He spent six
years as an ascetic, seeking the truth, and found it at the age of
35. After seeing <1nirva$n*a>1 he spent forty-five years helping others to
the same realization, and entered <1parinirva$n*a>1 when he was 80.
Since then he is dead. Being dead he is of course powerless for
good or ill. He cannot hear prayers or aid supplicants. Moreover,
even if he could it would not help very much. The Buddha was
concerned only with spiritual goods, salvation; worldly goods he
renounced. Each man has to work out his own salvation by his
own effort. Man is a free agent, and makes his own <1karma>1, which
determines where he will be reborn. The important things are
not in the gift of any power in the world--except the free will of
the individual himself. In the Canon's account of the Buddha's
last days<s2>s we read the Buddha saying that one should rely only
on oneself, only on the Dhamma;<s3>s his last words were, "Com-
pounded things by their nature decay; be attentive in your
efforts.'<s4>s
  This is the cognitive position, which is not affected by a slight
over-simplification which I have made.  Buddha means "en-
lightened' in Sanskrit and Pali, so strictly speaking anyone who
has seen <1nirva$n*a>1 is a Buddha. A monk explained to me that strictly
there are three kinds of Buddhas. Those usually referred to as
Buddhas, the twenty-five plus, are <1samyak sambuddha>1, "truly fully
enlightened'. (Recall the formula: <1Namo tassa Bhagavato arahato
samma$ sambuddhassa>1.) They are <1lokottara>1, supramundane. Of
course strictly speaking all Buddhas are <1ipso facto lokottara>1, having

82

risen above the world of transmigration ("sam*sa$ra lo$kin uttara
vela$'), but by usage (<1vyavaha$ra>1) only the <1samyak sambuddha>1 get
this title. Such Buddhas found a <1s/a$sane$>1, they teach. Then there is
the <1pratyekabuddha>1 (Sinh. : <1pase$budu>1), "enlightened by himself'. He
realizes <1nirva$n*a>1 for and of himself alone ("tamange$ nivan avabo$dha
paman*ayi'). Like the <1samyak sambuddha>1 they reach <1nirva$n*a>1 with-
out the help of a doctrine ready at hand, for they are born in
periods when there is no <1s/a$sane$>1 ("Budu s/a$sane$ lo$ke$ pavatina na%ti
ka$laka a%ti venne$'); but they do not found one themselves, so they
help no one. They mostly occur in the Himalayas and such remote
regions. By definition none exist now. Finally, there are those
who become enlightened (<1buddha>1) as disciples (<1s/ra$vakaya$ vas/ayen>1)
under the <1samyak sambuddha>1--the ordinary <1arhats>1, if we may so
speak. All these three types of Buddhas are still normal human
beings, and do not further affect the argument; we will revert to
reserving the title Buddha for the <1samyak sambuddha>1 like Gotama.
The <1pratyekabuddha>1 will hardly occur again, and the <1s/ra$vaka
buddha>1 will be simply called an <1arhat>1.
   The cognitive position is logically coherent and canonically
correct. Very little occurs which cannot be rationalized in terms
of these beliefs. However, the Buddha has long since become the
victim of a personality cult; and throughout recorded history
behaviour towards him has been such that we may well question
whether affectively he is viewed as no more than a historical
personage long dead. Even the cognitive position expressed in
words is not quite consistent.
   The question whether Buddhists really view the Buddha as the
mortal founder of their religion or as a being with at least some of
the attributes of the Christian God (such as eternal life, omni-
potence, and responsiveness to prayer) is the one that has most
interested western observers, and I shall return to it at length in
the second half of this chapter. The subject is a large one, but to
arrive at a tentative answer it will be sufficient to focus attention
on behaviour towards Buddhist images. This behaviour will turn
out at times to be discordant with the cognitive position just stated.
However, before we go on to examine these divergencies, and
then to establish the Buddha's position in affective religion, more
can and should be said about the cognitive position. For it would
do violence to my chosen method of taking full account of the
villagers' own views were I to restrict my statement of their beliefs

                                                                   83

about the Buddha to the one fact, most elementary from their
point of view, that the Buddha was mortal. So the rest of the first
half of this chapter will be devoted to what villagers know (or
believe they know) about the Buddha. (I use the word "know'
because they do; I do not wish to imply by it any agreement or
disagreement on my part.)
  To indicate the state of knowledge about the Buddha I offer
two kinds of evidence. Every image-house (<1viha$ragedara>1) is
adorned with paintings and sculpture which depict incidents from
the Buddha's life, and in some cases his former lives, and the lives
of previous Buddhas. Most of the image-houses I saw were the
works of local craftsmen; some, especially the more recent, were
the work of craftsmen from the Low Country (usually around
Ma$tara in the extreme south). Villagers do not customarily pay
a great deal of attention to the art in their temples--certainly
detiils of artistic finesse or even of iconography have no meaning
for them--but they naturally have some idea of the incidents
represented.
  My other evidence is the pair of thin paper-back books which
together form the primer for the bottom class, children about 5
years old, of a Buddhist Sunday School (<1daham pa$sala)>1. These
equivalents to Sunday Schools were started about the turn of the
century by the Young Men's Buddhist Association (Tarun*a
Bauddha Man*d*alaya), which publishes the books used. (It still
owns the rights to most of them, but those for the first five classes
are owned by the Government's Department of Culture.) Books
are supplied free to all registered schools. These can be run by
anyone, but in practice they are normally run by a monk at the
village temple with or without the assistance of other monks and
local schoolmasters. They are free and the books are loaned to the
children. The children sit for annual exams set by the Y.M.B.A.,
and receive appropriate certificates.  There were <1daham pa$sal>1
attached to about half the temples I visited ; they meet on Sundays,
but since I left the Government has substituted for the western
weekend the <1poya>1 day and the half-day preceding it, so they now
meet on <1poya>1 days. In Mi$gala the <1daham pa$sala>1 was the concern
of the chief monk, who took great pride in it. Children came from
four villages, and most of the younger children in Mi$gala attended.
The chief monk claimed that when I enrolled (in the bottom class)
I was the 518th pupil, but I think attendance rarely exceeded 300.

84

The senior class was mostly taken by the monk, all the other
classes by senior children. The lower classes, which were larger,
were subdivided by sex. The classes spread across a grassy space
at the foot of the stone steps leading up to the temple which served
as a kind of village meeting-ground for the rare organized game or
visit of the Government film unit; they occupied a permanent
school-type building also used for meetings of the village council,
etc., and a shed which will no doubt be replaced in due course
by something permanent; but the roof space was inadequate for
the masses of children. Proceedings were supposed to last from
9 to 11, timed by an alarm clock the monk had carried down from
the monastery. He would administer the Three Refuges and Five
Precepts; the children then divided into their classes and took
turns in reading out the lesson. In our class we read through the
lesson twice and then answered in writing the questions on it given
in the book; the teacher corrected our answers. There then was
recitation of Pali or old Sinhalese verses, also from the books, which
had to be learned by heart, and the class usually managed to get
away before the two hours were up.
   The pattern of the textbooks for the lower classes was that for
each class there were two volumes which we used on alternate
weeks. The arrangement of my books was typical: the first volume
contained the chapters on the life of the Buddha, in fact on his
life till he became Buddha (i.e. saw <1nirva$n*a>1), and some Pali
stanzas which everyone has to know, with their explanations in
Sinhalese; the second book had five <1Ja$taka>1 stories and a selection
of moral stanzas in old Sinhalese by a fourteenth-century poet.
In the Appendix I give a translation of the first two chapters of the
Buddha's early life, to give some idea of their flavour, the full table
of contents of the primers of the two bottom classes, and a list of
the other <1daham pa$sala>1 books in use at our temple. The wide
circulation of these books, which pass from hand to hand, can be
judged from the fact that of the first grade primer which we were
using the first volume was the third edition, of 27,500 copies,
printed in 196o, and the second volume was the fifth edition, of
20,000 copies, printed in 1961.
  It will make for clarity if I begin by summarizing without com-
ment the best known incidents concerning Gotama Buddha. If
any reader finds that I have omitted a detail, or even a whole story,
which he considers crucial, or introduced something unimportant,
  It will make for clarity if I begin by summarizing without com-

                                                                 85

I can only apologize. That my selection should to some extent be
arbitrary is inevitable, and I have tried to justify my choice by the
documentation which immediately follows the summary. Finally
I give some brief indication of the ancient sources on which this
documentation rests. Well over half my summary is concerned with
the life of Gotama before he became Buddha; this disparity reflects
the ancient sources which the Sinhalese artists and authors have
used.
   (1) <1Sumedha>1. He became a Bodhisattva in the time of the (first?)
Buddha, Di$pam*kam, when he was a young brahmin ascetic called
Sumedha. Di$pam*kara was visiting a city with his monks, and the
inhabitants were preparing the way. Sumedha offered to prepare
a muddy stretch. He had not finished making it dry when Di$pam-
kara Buddha appeared, so he threw himself prone in the mud so
that Di$pam*kara Buddha and his monks could walk over him and
thus avoid the puddles. On seeing Di$pam*kara Buddha he made a
resolve that he too would become a Buddha. Di$pam*kara Buddha
stopped before him and prophesied that after countless ages he
would become a Buddha called Gotama.
   (2) <1The other Buddhas>1. He performed great services for twenty-
three more Buddhas after Di$pam*kara Buddha, and all prophesied
that he too would become a Buddha. These Buddhas before our
Buddha are collectively known as the twenty-four (<1su$visi>1) Buddhas.
The next Buddha will be Maitri$ (Pali: Metteyya). All the Bud-
dhas have similar biographies ; e.g. they have two chief male
disciples and attain Enlightenment under a particular tree which
becomes sacred to them.
   (3) <1Ja$takas>1. The future Gotama Buddha lived through many
lives as a Bodhisattva, accumulating the Perfections (<1dasa pa$rami$>1
or <1pa$ramita$>1) which are essential to a Buddha. These lives are the
material of the <1Ja$taka>1 stories. Many <1Ja$takas>1 are well known, but
none better than the story of his last human birth before Buddha-
hood, as King Vessantara. King Vessantara practised the perfec-
tion of giving (<1da$na pa$ramita$>1) by bestowing on suppliants not
only his wealth but even his wife and children. However, all was
restored to him within his lifetime.
   (4) <1The gods' invitation>1. Like every Buddha, Gotama was born
in one of the lower heavens called Tusita, and when in due course
the gods there requested him to be born on earth to become a
Buddha he had to consider the right time, place, continent,

86

family, and mother. Our Buddha decided to be born then, while
the term of human life was 100 years, at Kapilavastu (Silm.
Kim#bulvat) in India in the family of Suddhodana, king of the
S/a$kyas, from his chief queen Maha$ma$ya$.
  (5) <1Conception>1. At the A%sal*a full moon <1poya>1 (in July), ten lunar
months before Wesak, Maha$ma$ya$ was observing <1at*a sil>1 and sleep-
ing alone. She dreamt that a white elephant carrying a white lotus
in his trunk entered her right side.
  (6) <1Birth>1. Exactly ten lunar months later Queen Maha$ma$ya$ was
on her way to the city of Devadaha where her parent lived,
planning to give birth there. On the way she turned aside at a
grove of <1sal>1 trees called the Lumbini$ grove. There the birth-pangs
came upon her, and she gave birth standing, holding on to a
branch. The future Buddha descended spotless from the womb,
and was caught by gods. He took seven steps and exclaimed, "I am
supreme in the world. This is my last birth.'
   (7) <1Asita>1. A wise old man called Asita or Ka$l*a Devala (both
mean "black') heard of the prince's birth and asked to see him.
When he saw him he smiled and then wept. Asked the reason, he
said he smiled because this child would become a Buddha, but
wept because he realized that he would not be alive to see it (he
was going to be reborn in a formless heaven). Having left the
palace Asita told his nephew to renounce the world in expectation
of the time when he could become a monk under the Buddha.
   (8) <1Kon*d*an+n+a's prophesy and the name-giving'>1. For the name-
giving ceremony King Suddhodana invited 108 brahmins. The
eight wisest brahmins examined the prince's body, and seven
prophesied that he would become a universal monarch or a
Buddha, but the youngest, Kon*d*an+n+a, said that he would surely
become a Buddha after seeing an old man, a sick man, a corpse,
and an ascetic. Kon*d*an+n+a later became one of his first disciples.
The name given was Siddha$rtha, meaning "He whose purpose is
fulfilled'. (Siddha$rtha is the correct Sanskrit form, Siddhattha the
Pali, Siduhat the Sinhalese.)
  (9) <1Ploughing festival>1. Once at a ploughing festival, while his
father was ploughing, Prince Siddha$rtha sat under the shade of a
rose-apple tree and meditated. He attained the first trance (a
technical term of Buddhist meditation) and the tree's shadow did
not move away, whereupon his father worshipped him.
  (10) <1Education>1. At his studies Prince Siddha$rtha was out-

                                                             87

standing, but he took no lessons in the use of arms, and the
rumour spread among his warrior kinsmen that he would be useless
in war. Accordingly he arranged a demonstration at which he
strung and shot the most difficult bow.
  (11) <1Marriage>1. Prince Siddha$rtha at the age of sixteen was
married to Princess Yasodhara$, the daughter of King Suppa-
buddha of Devadaha, his mother's brother. He was then officially
installed as heir apparent (<1yuvara$ja$)>1.
  (12) <1The three palaces>1. King Suddhodana did not want his son
to renounce the world, so he kept him in the royal park, into
which he allowed no old or sick men, no corpses and no ascetics.
He built his son three palaces, one for each season, and here the
prince lived surrounded by every luxury.
  (13) <1The four omens>1. When he was 29 years old the future
Buddha asked to go out to a park with his charioteer Channa. On
the way he met an old man. Channa explained the sight to him,
and told him that he too was subject to ageing. At this Prince
Siddha$rtha lost his inclination for amusements and returned home.
In the same way he saw a sick man and then a corpse. On their
fourth excursion they saw a man who had renounced the world
and looked serene. This sight filled Prince Siddha$rtha with joy
and he resolved to renounce the world himself. He proceeded to
the park, where he received a message that Yasodhara$ had borne
him a son. At this he remarked, "Ra$hula is born' (Ra$hula is sup-
posed to mean a bond, though this popular etymology is in-
correct), so his son was called Ra$hula.
  (14) <1The great renunciation>1. That evening Prince Siddha$rtha sat
thinking in his palace while slave girls danced and sang before
him. He fell asleep, whereupon they too lay down just as they
were and slept. During the night, the prince awoke and saw them
lying around ; he was disgusted, and resolved to renounce the world
instantly. He summoned Channa and had him prepare the royal
steed Kanthaka. He went to Yasodhara$'s bedchamber and from
the threshold gave her a last look as she lay sleeping, holding
Ra$hula to her. Then he rode out of the city on Kanthaka, while
Channa clung to the horse's tail. Ma$ra, a wicked god, tempted him
to stay and become a universal monarch, but he paid no attention.
  (15) <1Hair-cutting>1. By morning the Bodhisattva had travelled
thirty leagues and reached the broad River Anoma$. He spurred
Kanthaka across in one bound. On the sands of the opposite shore

88

the Bodhisattva dismounted and cut off his hair with his sword.
The great god S/akra took the hair to heaven where it is kept in a
stupa called the Sil*umin*i sa%$ya. He also took off his royal garments
and replaced them with mendicant's robes, supplied by the great
god Brahma$. He sent back Channa and Kanthaka, and went on
alone. Channa later became a monk, but Kanthaka died of a
broken heart and was reborn in heaven.
  (16) <1Self-mortification>1. The ascetic Siddha$rtha began to beg for
his food. He visited the city of Ra$jagr*ha, capital of Magadha,
where King Bimbisa$ra offered him the kingdom; he refused, but
promised to come back to that kingdom first when he had found
Enlightenment. He meditated under the direction of two teachers,
but finally left them dissatisfied. With five other ascetics, headed
by Kon*d*an+n+a, he undertook the most rigorous asceticism, starving
himself until he nearly died. But he decided that this too was use-
less, and started taking proper food again. At this his five disciples
left him in disgust.
  (17) <1Suja$ta$'s milk-rice>1. After taking a proper meal he went and
sat at the foot of a banyan tree called Ajapa$la ("goatherd'). A mer-
chant's daughter called Suja$ta$ had prepared an offering of milk-
rice for the guardian deity of that tree. When her maid saw
Sidda$rtha she thought he must be the tree-deity making himself
visible, so she fetched Suja$ta$ who came and offered the rice in a
golden bowl. The Bodhisattva went and bathed in the River
Neran+jara$. Then he divided the milk-rice into forty-nine parts,
and having eaten it all he launched the bowl in the water, wishing
that if he was that night to attain Buddhahood the bowl should be
carried against the current. It swam upstream.
  (18) <1Battle with Ma$ra>1. That night, the full moon night of
Wesak, his thirty-fifth birthday, the Bodhisattva sat comfortably
cross-legged at the foot of a peepal tree, vowing not to get up till
he had attained Enlightenment. Now Ma$ra brought up all his
hosts of demons to scare the Bodhisattva, but though all the gods
fled in terror to the corners of the earth the Bodhisattva meditated
unmoved on his ten perfections, so nothing was able to touch
him. As he was alone Ma$ra then sought to confound him by
boasting of his own past charity. All the demons gave witness that
Ma$ra had given alms, and he challenged the Bodhisattva to produce
a witness himself. The Bodhisattva said he had given alms in his
birth as Vessantara and called the earth to witness. The earth

                                                           89

spoke in witness, and Ma$ra fled with his hosts, leaving the
Bodhisattva victorious.
  (19) <1Enlightenment>1. The Bodhisattva then meditated through
the three watches of the night. In the first watch he acquired
omniscience about the past, in the second omniscience about the
present, and in the third he understood the chain of dependent
origination which reveals the origin of suffering. At dawn he
attained <1nirva$n*a>1, at which the earth quaked and many marvels
were seen.
  (20) <1The seven weeks>1. The Buddha remained for seven weeks
deep in meditation. Each week was signalized by a different event.
In the fifth week the three daughters of Ma$ra tried to seduce him
by appearing to him as beautiful women, but he did not look at
them. At the end of the seventh week two merchants, Tapassu
and Bhalluka, passed and gave him alms. They took refuge in the
Buddha and the Dharma, thus becoming the first <1upa$sakas>1.
  (21) <1Invitation to teach>1. The Buddha doubted his ability to
teach what he had realized, but the great god Brahma$ begged him
to teach. On consenting he wondered to whom he should preach
first. He thought of his five disciples, who were now in the Deer
Park at Benares. On his way there he met an ascetic called Upaka,
whom he told of his enlightenment. Upaka said "It may be so',
and went on. The Buddha reached Benares on the full moon day
of A%sal*a.
  (22) <1First sermon>1. When his five former disciples saw him coming
they did not intend to rise for him, but he concentrated his
benevolence on them till they were affected, so they paid him the
usual respects. He preached to them the first sermon, "The
Turning of the Wheel of the Doctrine' (Pali: <1Dhammacakkappa-
vattana Sutta>1; Sinh. : <1Damsakpa%vatum Su$traya>1).<s5>s On hearing it
Kon*d*an+n+a entered the path to <1nirva$n*a>1, as the other four did in
the next three days. On the fifth day he preached the sermon "On
the Non-Existence of the Soul' (<1Anattalakkhan*a Sutta>1),<s6>s and all
five attained <1nirva$n*a>1.
  (23) <1Rainy seasons>1. The Buddha preached for forty-five years.
Most of the time he was travelling with some of his monks, but
each year the rainy season, which lasts three lunar months starting
on A%sal*a full-moon day, was spent at one place. There is a list of

90

where they were spent. Two were in heavens, one of them preaching
to his mother who had died ten days after his birth and been
reborn in the Tusita heaven, the other, the seventh, expounding
the <1Abhidharm*a>1 to the gods of the Heaven of the Thirty-Three
(Pali: Ta$vatim*sa ; Sinh. : Tavatisa$ Devlova).
  (24) <1Visit to S/a$kyas>1. During the first year of his ministry the
Buddha paid a visit to the S/a$kyas at Kapilavastu. As many of them
doubted his abilities he caused rain to fall on only those who
wished to get wet, and made marvellous phenomena appear in
pairs. He ordained his half-brother Nanda by a trick, and when
Yasodhara$ told Ra$hula to go and ask his father for his inheritance
he ordained him too. King Suddhodana was distressed, so the
Buddha promised that in future no one should be ordained with-
out the parents' permission.
  (25) <1Devadatta>1. The Buddha was subject to various hostile
attacks, but he always quelled them as he had quelled Ma$ra. His
great enemy was Devadatta, a son of King Suppabuddha, who
became a monk but created schisms, and even plotted against the
Buddha's life; he sent a rutting elephant against him, but the
elephant fell worshipping at the Buddha's feet. The Buddha
explained that Devadatta had been his enemy in many previous
births.
  (26) <1Disciples>1. The Buddha's two principal disciples were Sa$riputta
and Moggalla$na (Sinh.: Sa%riyut and Mugalan). His personal
attendant was Ananda. Other famous monks included Upa$li, who
had been a barber and became the greatest authority on the rules
of monastic discipline, and Maha$ Kassapa ("Kassapa the Great';
Sinh. (= Skt.): Maha$ Kas/yapa). A monk called Si$vali sometimes
attended the Buddha on journeys because he was always able to
find food. There is a list of eighty famous monks. The greatest
<1da$yaka>1 was Ana$thapin*d*ika (Sinh.: Ane$pid*u) who bought for the
Order the Jetavana$ra$ma at S/ra$vasti$ by covering the ground with
gold coins; in this grove the Buddha then spent many rainy
seasons. The greatest <1da$yika$>1 (female <1da$yaka>1) was Visa$kha$.
  (27) <1Visits to Ceylon>1. During his life Buddha paid three visits
to Ceylon (see p. 108).
  (28) <1Death>1. The Buddha entered <1parinirva$n*a>1 at Wesak at the
age of 80. He became ill after eating pork at a <1da$ne$>1 offered by a
smith called Cunda. In a grove at Kusina$ra$ he lay on his right side
between two <1sal>1 trees and died while meditating, surrounded by

                                                                    91

hosts of men and gods. After his cremation his relics were distri-
buted by a wise brahmin called Dron*a (Pali: Don*a).

   The traditional themes of Kandyan temple art are rather
restricted.  Sculpture is generally confined to images of the
Buddha (often with <1makara toran*a>1), Maitri$, and a few gods.
Paintings, leaving aside donors' portraits, <1arhats>1 (including
Sa%riyut and Mugalan), gods, and decorative motifs, mainly show
the twenty-four (<1su$visi>1) previous Buddhas and the seven weeks
(<1sat satiya>1) passed by the Buddha immediately after his Enlighten-
ment. Other traditional subjects, slightly less frequent, are the
sixteen (or eight) sacred places in Ceylon (see pp. 109-10), the
battle against Ma$ra, S/akra's heaven (both the last two tend to be
painted on ceilings), and the <1Vessantara Ja$taka>1. Before about
1900 other subjects are exceptional, and these subjects are still the
main stock-in-trade of local Kandyan artists. Round 1900, how-
ever, temple artists from the Low Country, especially from the
area round Ma$tara in the extreme south of Ceylon, began to invade
the hill country. Today at a guess more than half the temple art
extant even in the Kandy District itself is Low Country work,
though in my area the proportion is a bit lower. Influenced by
European painting, the Low Country craftsmen attempted pers-
pective, used a far wider range of colours (which they tended to
buy commercially rather than concoct them according to the
traditional recipes), and--most relevantly for our purposes--
introduced a far wider range of subjects both in painting and in
sculpture. The most important of these subjects were scenes (often
shown in a cycle) from the <1Buddhacarita>1, the life (especially the
early life) of the Buddha, a subject not unknown but rare in earlier
Kandyan painting. The Kandyan painters have gradually assimi-
lated the subject matter and some of the methods (notably the use
of mass-produced glossy paints) of their more prestigious Low
Country colleagues, so that today there is not so much difference
between the two styles. Regional differences in subject matter, in
so far as they exist, can probably be ascribed to the taste of the
patrons, that is the monks, who at least in theory decide on the
themes with which their <1viha$ras>1 are to be decorated. Typical
subjects gaining in popularity through Low Country influence are
sets such as ten <1Ja$takas>1 illustrating the Ten Perfections, or the
eight "victories' hynmed in successive verses of a mediaeval

92

Pali poem, the <1Jayamangala Ga$tha$>1. Some differences in manner
are still discernible: Kandyan artists still sometimes eschew per-
spective, and the brightly coloured sculpted tableaux depicting
scenes from the <1Buddhacarita>1, the <1Maha$vam*sa>1, or <1Ja$taka>1 stories
are not yet, to my knowledge, in their repertory--they keep to
painting for narrative and confine sculpture to isolated figures.
  I visited thirty-two temples in my electorate with finished
examples of religious painting and sculpture. (In seven temples the
image-house was not yet begun or was under reconstruction.)
Usually, but not necessarily, the art is confined to one image-
house; my remarks apply to whole temple complexes. Of these
thirty-two art complexes, nineteen were, I think, by Kandyan
artists only, eight, I think, only by Low Country artists, three by
both and two not ascertained. I ascribe the comparative paucity of
Low Country art to the relative poverty, remoteness, and conserva-
tism of much of the area.
  The subjects of every paragraph except 27 in my summary
above are depicted in local temples. I shall confine myself to
mentioning those shown more than five times. The paragraph
numbers below refer back to the summary.
  (1, 2) There are fourteen representations (two in one temple !) of
the twenty-four previous Buddhas; of these all but three are
paintings, and all but one of the sculpted sets are by Kandyans.
(The exception is doubtless an imitation of the sculpted <1su$visi>1 in
an older temple in the same village belonging to the rival Nika$ya.)
When painted they are shown seated; when sculpted, standing.
The twenty-four figures are identical; they are differentiated by
the worshippers kneeling, very small, to their left (our right--but
the side sometimes changes). These worshippers represent the
future Gotama Buddha renewing his vows as a Bodhisattva before
each Buddha. In most of these lives the Bodhisattva was human,
but for instance in the time of Paduma, the eighth Buddha, he
was a lion. Sumedha is not depicted separately but appears at the
beginning of the <1su$jisi>1 lying prone at Di$pamkara's feet.
  There are ten statues, all standing, of Maitri$ Buddha; he is
identified by his tall headgear in which sits a small meditating
Buddha. This iconographic detail shows the representation of
Maitri$ to be descended from representations of Avalokites/vara, a
Bodhisattva of Maha$ya$nist provenance, statues of whom, similarly
identifiable by a small Buddha seated in the headgear, have

                                                           93

survived from the late Anura$dhapura period. When represented by
a statue Maitri$ Buddha is painted white or pink and holds a lotus.
Most of the statues of Maitri$ are Low Country work. The tradi-
tional Kandyan way of representing Maitri$ is to paint him seated
cross-legged in the Tusita heaven awaiting the gods' invitation to
be reborn on earth; in this style he has neither lotus in his hand
nor Buddha in his hat. In three temples he is so painted over
entrance doors, and in seven on the ceiling, though about this last
figure there is some doubt because some informants said that the
person shown in heaven was S/akra. However, in eighteenth-
century temples (e.g. Dam#bulla) we know that similar scenes are
intended to portray Maitri$. In one temple is a cycle of eighteen
paintings depicting Maitri$'s life. The incidents are all repetitions,
with new proper names, of the listed incidents in Gotama's
life.
  (3) I have counted altogether forty-two depictions of twenty-one
different <1Ja$taka>1 stories, disregarding the fact that some are shown
in cycles and some in single scenes. There is a general tendency
for those <1Ja$takas>1 which are canonically associated (see p. 101)
with the Bodhisattva's acquisition of a particular perfection
(<1pa$ramita$>1)<s7>s to be more widely known. Other popular stories are
the <1Telapatta J>1. (96),<s8>s <1Culladhammapa$la J>1. (358), <1Sa$ma>1
and <1Daham*son*#d*a J>1., a post-canonical Sinhalese story. Most shown
locally are the <1Vessantara J>1. (547) (eight times) and the <1Daham*-
sonda J>1. (six times).
  (4) The scene in which the gods request the Bodhisattva, seated
at ease in their midst, to be born on earth is depicted six times.
  (6) There are ten representations of the birth. The sky is full of
gods. The scene is depicted at the moment when the Bodhisattva
has taken the seven steps; to mark each step there is a lotus, and
he is standing on the last. One cycle precedes the birth with the
queen and her party setting out for Devadaha. Two show the
return to Kapilavastu.
  (14) The Great Renunciation lends itself to portrayal in three
separate scenes:

94

   (i) Siddha$rtha awakes to find the dancing girls asleep and repul-
       sive. Shown once.
  (ii) Siddha$rtha gazes on his wife and child. Shown five
       times. Usually the Bodhisattva is shown standing at the door
       with a calm expression, but in one picture he is walking away
       with what can only be described as a sneer. Maybe this is
       due to a maladroit executing.
 (iii) The Bodhisattva leaves the city on his horse. This is shown
       six times. In the sky in front of him invariably appears
       Vasavatti Ma$ra, tempting him to return. In this episode Ma$ra
       is coloured green.
  (15) This episode is especially popular with artists, and divides
into two scenes:
  (i)  The leap across the Anoma$ River, shown five times.
 (ii)  The Bodhisattva cutting short his hair, shown nine times.

Most representations include S/akra catching the hair and turban
in mid-air, and some have a picture of the Sil*umin*i sa%$ya in heaven.
Some add the god Brahma$ on the other side of the picture, bring-
ing robes. According to one version the Sil*umin*i sa%$ya is likewise
balanced by the Sal*umin*i sa%$ya, a heavenly stupa in which are kept
the clothes the Bodhisattva discarded on this occasion. There are
also two separate depictions of the Sil*umin*i sa%$ya and the Sal*umin*i
saya, which might be considered the heavenly prototypes of our
man-made stupas, and one of the Sil*umin*i sa%$ya alone.
   (17) Suja$ta$ offering the milk-rice is depicted six times. The
Bodhisattva is shown twice launching the bowl upstream on the
Neran+jara$ River.
   (18) The battle against Ma$ra is shown in six versions; the scene
gives the artist ample scope for the exercise of the imagination in
creating hideous monsters. In two of the representations the
Bodhisattva is shown calling the earth to witness; a small female
figure, shown from the waist up, rises up from the ground in front
of him; this is Mahi$ Ka$nta$va, the earth personified.
   (19) The actual moment when Gotama became the Buddha is
not represented as an episode in his life, though it could be said to
constitute the first part of the <1sat satiya>1 (see below). On the other
hand this moment furnishes the paradigm for the <1sama$dhipil*imaya>1
("meditation image'), in which the Buddha is shown in a symmetri-

                                                             95

cal pose, legs crossed, hands folded in lap, eyes closed or half-
closed. Successful <1sama$dhi pil*ima>1 convey a great tranquillity. This
seated <1sama$dhi>1 posture is the commonest one for a Buddha statue,
the one opposite the entrance to the shrine, which contains or is
supposed to contain the relic. The Buddha was perhaps first shown
seated, and the famous stone seated Buddha on the Outer Circular
Road at Anura$dhapura, which may be the most ancient Buddha
image to survive, is in the <1sama$dhi>1 position. A <1sama$dhi pil*imaya>1 is
not normally thought of as representing the Buddha at a particular
moment in his life, but it seems to me possible that it originated
as a representation of the Enlightenment.
   In an image-house a <1sama$dhi pil*imaya>1 usually has its back
against a wall, in which case the wall is painted with a large halo
(<1budu ra%s>1) in white, yellow, orange, red, and blue. Such a halo may
also be attached to a Buddha image in some other position.
  The enlightened Buddha is usually represented by a statue
rather than a painting. This reminds us both that it is his En-
lightenment which made the Buddha worshipful (in contrast to
Jesus, who is sacred from birth), and that the Buddha image, at
least as an object of devotion, originated as a relic container (pp.
112-14 below), a function which a two-dimensional painting cannot
fulfil.
  (20) There are nine representations of the <1sat satiya>1. The events
of the first six weeks are listed in a Pali stanza:

       <1Pat*hamam* bodhipallam*kam* Dutiyam* ca animisam
       Tatiyam* cam*kamanam* set*t*ham* Catuttham* ratana$garam
       Pan+camam* ajapa$lanam*, Mucalindona chatthamam>1.<s9>s

  First the seat of Enlightenment; second unwinking; third the
Most excellent ambulatory; fourth the house of jewels; fifth
Ajapa$la(na); sixth with Mucalinda.

  (i) The Buddha remains in <1sama$dhi>1 without moving for a whole
      week.
 (ii) Without blinking he contemplates the <1bodhi>1 tree.
(iii) He makes ajewelled ambulatory in the sky up and down which
      he walks in meditation.

96

 (iv) Similarly, he makes a jewelled house.
  (v) On return to earth he goes back to sit, for the next three
      weeks, under the banyan tree where Suja$ta$ saw him. The
      three daughters of Ma$ra try to tempt him. They are depicted
      as seductive women playing musical instruments.
 (vi) During a storm the Buddha is sheltered by the supernatural
      serpent king (<1na$gara$ja$>1) Mucalinda, who puts his great hood
      over him. The Buddha is shown sitting on the snake's coils,
      half enveloped in the five-headed hood.
(vii) The visit of Tapassu and Bhalluka, which occurred at the
      end of the seventh week, is not always included in it. In such
      cases the Buddha may again just be shown in meditation, or
      the cycle may be limited to six weeks, the main <1sama$dhi
      pil*imaya>1 in the shrine perhaps standing for the seventh.


  (22) There are eleven representations of the first sermon. It is a
favourite subject for a tableau, in which the five disciples sit in
front of the Buddha, facing him, and gods protrude from the
walls all round, arriving on clouds to listen.
  (23) There are several preaching scenes depicted, often as com-
panion pieces to the first sermon. The favourite (six times) is the
Buddha expounding the <1abhidharma>1 in the Heaven of the Thirty-
three.
  (26) (i) The Buddha's two chief disciples, Sa$riputta and Moggal-
la$na, are shown in ahnost every temple, invariably as a pair (I
counted twenty-six pairs, one per temple). They are known as the
Buddha's right-hand disciple (<1dakun*at sav>1) and left-hand disciple
(<1vamat sav>1) respectively, and flank the Buddha's image accordingly.
Sa$riputta always has a pink, Moggalla$na a blue face ; otherwise they
are represented like any other monks, and always mirror each
other in every other respect. Generally they flank the main Buddha
image in the shrine, especially so if it is a seated image, and they
usually are shown facing or inclined towards this image in a
reverent posture.
  (ii) The other two disciples who fairly often have statues to
themselves are A$nanda and Maha$ Kassapa. Neither of them seem
to have any iconographic peculiarities, so it is hard to tell them
apart. A monk told me that Maha$ Kassapa could be distinguished
by his size, but I found this unreliable. To make matters worse,
both are especially likely to be portrayed standing near the feet of

                                                                 97

a Buddha who is entering <1parinirva$n*a>1 (see para. 28 below), and
the monk in this scene was identified for me now as one, now as
the other. In the canonical account of the Buddha's death A$anda
figures prominently; Maha$ Kassapa arrives after the Buddha has
died, and worships his feet as the body lies on the pyre. Either the
iconographic tradition has intentionally conflated the two events,
which is unlikely, or one of the two disciples was originally meant
and it has been forgotten which. It seems more likely that the
person represented is Ananda, as there is no sign of a pyre; in that
case the position near the Buddha's feet would be a mere coinci-
dence. If I count under A$nanda those statues accompanying the
<1parinirva$n*a>1 which were identified for me as Ananda, and the
others as dubious, leaving to Kassapa only his isolated representa-
tions, the totals are six statues of A$nanda, three of Maha$ Kassapa,
and three dubious.
   (iii) The only other disciple to rank a statue is Si$vali (one). In
an <1a$va$saya>1 with no image-house but a shrine at one end of the
room' I noticed a framed picture of Si$vali as the only wall-
decoration. Because he is connected with a constant food supply
Si$vali is usually considered a patron more appropriate for laymen
than for monks.
   (iv) There are two pairs of portraits of Ana$thapin*d*ika and
Visakha$.
   (28) When the principal Buddha image in a shrine was not seated,
it was invariably lying in a Pose called the "lion-lying' (<1sim*ha
seyya$va>1); the Buddha lies on his right side, his head on a pillow
on which his right hand also lies, his left arm straight down by his
side and his feet one on top of the other. According to my infor-
mants this pose can represent the Buddha meditating in a lying
position; usually, however, it is clear that the <1parinirva$n*a>1 is
intended. (I have heard that if the toes are all in line he is alive,
but if the big toe of the left foot is raised he is dead; however, it is
rare for the toes not to be aligned, so I am not sure that this inter-
pretation is to be credited.) The lying Buddha is called a <1sa%tapena
pil*imaya>1. I counted thirteen such images.


   The evidence from <1daham pa$sala>1 textbooks can be more rapidly
summarized, as it is presented in the Appendix. For each year
there are two textbooks, of which the first is principally concerned
with the life of Gotama, the incidents being arranged in ten

98

chapters. Naturally each book is rather more detailed than its
predecessor, and written rather less simply, but there is a good deal
of repetition from year to year; for instance, the Bodhisattva's
birth is described in each of the first three years. In comparison
to visual representations there is far less emphasis on the spectacu-
lar; for instance the accounts of the birth omit the seven steps and
subsequent pronouncement, and the accounts of the Enlighten-
ment omit the battle with Ma$ra, and describe the event rather in
terms of my paragraph 19 than of my paragraph 18. This probably
reflects western influence. On the other hand, the early books
maintain the concentration on Gotama's life before Enlighten-
ment: the first year ends with the barest mention of that event;
the second year has six chapters on his life as Siddha$rtha, the
seventh chapter takes us to the first sermon, then there are two
chapters on the years of teaching, and the last is on the <1parinirva$n*a>1.
   The events in the first book, which I would assume to be the
best known, are covered at not much greater length than I have
given them, in the very simple and rather repetitious style suitable
for small children. There is a fair amount of descriptive detail,
especially about the three palaces (my para. 12), which I have
omitted, but not much which I would judge substantive. On the
other hand some important details are missing. Each chapter is
two small pages long in the original. This said, a fair notion of the
primer's coverage and emphasis can be gathered from the follow-
ing tabulation of its contents.

<1Chapter of priner Paragraphs in my account>1

        1            1
        2            4, 5, 6 (no details on birth)
        3            7
        4            8
        5           10
        6           11
        7           12
        8           13
        9           14 (no Ma$ra), 15 (no gods, no dismissal of
                       Channa and Kanthaka)
       10           16 (no Bimbisa$ra or teachers), 17 (no bowl),
                       19 (no details)

   The primers also include a number of <1Ja$taka>1 stories. These
hardly coincide with those represented in temples, nor am I able

                                                               99

to discern a pattern, such as illustration of the Ten Perfections,
underlying their selection, so I shall merely refer the interested
reader to their listing in the Appendix.
  The two sources cited, temple art and the <1daham pa$sala>1 primer,
account for all paragraphs in my account except paragraph 27,
the Buddha's visits to Ceylon. On what are these contemporary
sources based?  Almost entirely on ancient commentatorial
literature. Almost everything mentioned under my paragraphs 1
to 22 and 24 is based on the <1Nida$nakatha$>1. Paragraphs 23, 25 and
26 are gleaned from various canonical sources; more relevantly,
perhaps, those events and persons are frequently mentioned in the
introductions (<1paccuppanna-vatthu>1) to individual <1Ja$taka>1 stories
which were presumably composed at the same time as the
<1Nida$nakatha$>1, the introduction to the whole book. Paragraph 27
is based on the first chapter of the <1Maha$vam*sa>1. Paragraph 28 is
based on the canonical <1Maha$parinibba$na Sutta>1.
  Someone who reads the <1Nida$nakatha$>1 will find in it almost every-
thing now known about the Buddha by a Sinhalese villager. It is
not my concern here to trace the origins of the contents of the
<1Nida$nakatha$>1 in any detail--that has already been well done by
Dr. E. J. Thomas in his learned work, <1The Life of the Buddha as
Legend and History>1,<s10>s which covers the main stories found in all
ancient Sanskrit and Pali sources. However, a few more words
about the <1Nida$nakatha$>1 in its relation to other literature will be
 appropriate.
  The <1Nida$nakatha$>1 forms the introduction to the Pali prose
commentary on the <1Ja$takas>1. As explained in Chapter 1, the Ja$taka
stories are themselves technically a commentary, i.e. the prose
stories are a commentary on the verses which form their nucleus;
but the work of which the <1Nida$nakatha$>1 is a part is a commentary
on the <1Ja$taka>1 stories taken as a whole. It is not the first such
commentary, because it mentions an earlier one with which it
disagrees ; this is probably the Sinhalese version on which it is
based. Its introductory verses indicate<s11>s that it was composed in
the Maha$viha$ra in Anura$dhapura, but it is anonymous. It can-
not be securely dated, but as will appear from the analysis
of its contents, it is likely to be roughly contemporary with
Buddhaghosa, and as Buddhaghosa was the first to translate

100

Sinhalese commentaries into Pali this work cannot precede him.<s12>s
  After brief (but involuted) introductory verses the <1Nida$nakatha$>1
is in three parts: I <1Du$re-nida$na>1 ("Distant Origin') II <1Avidu$re-
nida$na>1 ("Not so Distant Origin') and III <1Santike-nida$na>1 ("Proxi-
mate Origin'). I is half the length of the whole, II and III a quarter
each. I contains the story of Sumedha, including his decision to
attain each of the Ten Perfections; matter on Di$pam*kara and the
rest of the twenty-four Buddhas; and a bare mention of various
births in which he attained the Ten Perfections. II contains the
life of Gotama from his decision to be born to his Enlightenment-
III tells of the first year of his preaching, ending with Ana$tha-
pin*d*ika's gift of the Jetavana$ra$ma.
  Three other texts between them cover almost everything that
is in the <1Nida$nakatha$>1. Two are canonical, though generally con-
sidered late: the <1Buddaavam*sa>1 and the <1Cariya$pit*aka>1. The third is
the commentary on the <1Buddhavam*sa>1, the <1Madhuratthavila$sini>1
by Buddhadatta. The <1Nida$nakatha$>1 uses the first two, but seems
itself to be used by the third.
  The <1Buddhavam*sa>1, in verse, has twenty-eight chapters, all but
one of them short. The first sets the scene: the Buddha tells how
he came to attain Buddhahood. The second chapter is by far the
longest, and tells the story of Di$pamkara and Sumedha; all of it
but the last verse is quoted, with acknowledgement, by the
<1Nida$nakatha$>1 (as vv. 12-222). Chapters 3 to 25 inclusive tell of
the other twenty-three Buddhas before Gotama. Just the first
verse of each chapter is quoted (with some variant readings)
the <1Nida$nakatha$>1 (as vv. 223 and 225-46). The 26th chapter gives
a biography of Gotama, as brief as the preceding chapters, and is
not itself quoted. Chapter 27 is a list of Buddhas, ending with
Metteyya, who is not mentioned elsewhere in the work; chapter 28
gives a list of relics and their distribution. Neither of these last
two chapters has a commentary, so they are probably late additions.
So much for the <1Buddhavam*sa>1.

                                                                  101

  So far (up to page 44) the <1Nida$nakatha$>1 consists entirely of these
verses from the <1Buddhavam*sa>1 and of prose passages which enlarge
on them or (in the case of Buddhas after Di$pam*kara) paraphrase
the verses not quoted. The <1Nida$nakatha$>1 then lists the Buddhas
before Gotama, a list which is also found in the <1Buddhavamsa>1
commentary (pp. 131-2). Next there are a few lines summarizing
the story so far and listing the qualifications (such as male sex)
for becoming a Buddha. The last three pages of the <1Du$re-nida$na>1
then list births (all of them <1Ja$taka>1 stories) in which the Bodhisattva
attained one of the Ten Perfections, and for each perfection a
verse is quoted. At the end the author says that the stories men-
tioned can be found in full in the <1Cariya$pit*aka>1. Interestingly
enough this is not true of the <1Cariya$pit*aka>1 as we know it. The
<1Cariya$pit*aka>1 is a canonical book which sim
stories to illustrate the attainment of the Ten Perfections, and
retells them briefly in verse. But the selections made by the
<1Nida$nakatha$>1 (by title only) and the <1Cariya$pit*aka>1 differ in about
half the cases, and of the eleven verses quoted, despite the con-
trary impression given by the author and the contrary statement
by Rhys Davids,<s13>s only six<s14>s are in fact taken from our <1Cariya$-
pit*aka>1. This strongly suggests that the version of the <1Cariya$pi.taka>1
then known was not the one which has reached us.<s15>s So much for
the <1Cariya$pit*aka>1.
  This disposes of the <1Du$re-nida$na>1 of the <1Nida$nakatha$>1: it
depends entirely on the <1Buddhavam*sa>1 and <1Cariya$pit*aka>1; all of it
not duplicated elsewhere is of a purely commentatorial nature.
The rest of the <1Nida$nakatha$>1 is largely duplicated by the com-
mentary on the <1Buddhavam*sa>1. This was noted by Rhys Davids,<s16>s
and Miss Horner cites the parallel passages in the footnotes to her
edition of the commentary. The commentary on the story of
Di$pam*ara and Sumedha is largely the same as the first part of
the <1Nida$nakatha$>1. The first Part of the commentary on the 26th

102

chapter of the <1Buddhavam*sa>1, the one about Gotama Buddha, is an
abbreviated version of most of the rest of the <1Nida$nakatha$>1, adding
only a few short bridge passages and a few verses. This section of
the commentary (p. 272, line 6--p. 291, line 31) takes the story as
far as the first sermon, covering <1Nida$nakatha$>1, p. 47 (beginning of
the <1Avidu$re-nida$na>1) to p. 82, line 12. The remaining twelve pages
of the <1Nida$nakatha$>1 are not to my knowledge duplicated or closely
paralleled in another Pali text, except that parts of it recur in the
introductions to individual <1Ja$taka>1 stories, which are usually pre-
sumed to be by the same hand.
   The relevance of this analysis to our subject is that the tradi-
tions I encountered do not derive from the <1Nida$nakatha$>1 alone. For
example, the Sunday School primer gives the names of Sid-
dha$rtha's three palaces as Ramma, Suramma, and Subha. In the
<1Nida$nakatha$>1 they are not named. But these names occur in the
<1Buddhavam*sa>1 (27, 14)<s17>s and hence in its commentary (p. 278).
The commentary goes on to give the story about the feat of arms,
more or less as given in the Sunday School primer. In the <1Nida$na-
katha$>1 this incident is mentioned only briefly, but the author refers
one for it to the <1Sa$rabham'a Ha$taka>1 (522), in which a similar, but
even more elaborate story is told of the Bodhisattva in a <1former>1
life. The version before us therefore seems to come from the
<1Madhuratthavila$sini$>1.
   Sometimes one has to look even further. When a monk was
telling me about the birth of the Buddha (see Chapter 7, p. 288)
he recited a pair of Pali verses. The first is not in the <1Nida$nakatha>1-,
but is in the <1Buddhavam*sa>1 and in its commentary. Its original
appearance in the <1Buddhavam*sa>1 is in a very short reference to the
Buddha's birth, but it is quoted again in the commentary in the
context of the full story. The second verse, however, is in none
of these texts, but in an even shorter version of the whole early
biography, perhaps potted from the <1Madhuratthavila$sini>1. This very
short version is in the commentary to the <1Dhammapada>1 (vol.
1, pp. 83 et seq.). This text, like my informant, closely follows
the first verse with another;<s18>s so this is presumably his source.

                                                                    103

   Of course the traditions under discussion do not necessarily or
even probably derive <1directly>1 from the Pali texts. In most cases
the Pali texts have been mediated through the great Sinhalese
classics, prose accounts of the Buddha's life such as the <1Butsa-
ran*a>1, the <1Pu$ja$valiya>1, the <1Saddharma$lam*ka$raya>1 and the <1Saddhar-
maratna$valiya>1, and through the folk literature deriving from them.
But these Sinhalese works closely follow the Pali texts discussed
above.


          II.  THE BUDDHA AS WORSHIPPED: MAN OR GOD?

   Now that I have reviewed the Buddha's life and works (<1caritaya>1),
the primary interest of the Buddhists themselves, it is time to
return to his ontological status, if I may so call it, which has been
the primary interest of academic observers, and to discuss the cult
of the Buddha and his image. The bridge to this cult of the Buddha
is the relic (<1dha$tu, dahatunvahanse$>1). We already hear of relics in
the Canon. Shortly after the Buddha's last words (quoted on p.
81 of this chapter), in the same text, is recounted how after the
Buddha's body had been cremated all the peoples of that area sent
asking for his remains. They were divided into eight parts by
Don*a; a tribe who applied late got the embers of the pyre, and
Don*a kept the vessel in which the remains had been collected.
The original text ends by telling us that there were then ten por-
tions, and for each a stupa was built and a festival celebrated. To
this is appended a verse, which Buddhaghosa says was added in
Ceylon, which adds to the earlier eight portions of bodily remains
four teeth, one of which was allegedly in heaven, and another in
Kalinga in South India. This latter became of great importance
to the Sinhalese, for it was brought to Ceylon in the fourth
century, and from that time on accompanied the king of Ceylon,
and had a temple in its honour in his capital. Its history is
chronicled in Pali verse in the <1Da$t*ha$vam*sal>1<s9>s ("The Chronicle of
the Tooth'), composed by Dhammakitti in the twelfth century.
The Portuguese claimed to have taken it to Goa and destroyed it
in the sixteenth century, but the Sinhalese claim that by miracu-
lous intervention they falled to do so. The Sinhalese kings built a
temple for it in Kandy which has the title, unique for a religious
edifice, of "palace'--the Dal*ada$ Ma$liga$va (literally "Palace of the

104

Tooth Relic'). The temple is controlled by a lay administrator,
the Diyavadana Nilame$, and is likewise unique in not belonging
to a monastery, though monks of the two great monasteries of
Kandy, Malvatta and Asgiriya, alternately officiate there, preaching
on <1poya>1 days, etc. It is nowadays--as has perhaps always been the
case--exhibited only to people of the greatest importance, such as
heads of state and prime ministers. But once a year it is taken in
procession (<1peraha%ra>1) inside its casket (<1karan*#d*uva>1) on the back of
the largest available elephant; though it is said that these days an
empty substitute casket is taken, in case of accidents. This is the
A%sal*a Peraha%ra, Ceylon's most famous tourist attraction;<s30>s it takes
place on the ten nights leading us to the full moon of the Sinhalese
month of A%sal*a (in August) and on the day itself. The tooth is
preceded by the <1peraha%ras>1 of the four gods whose temples are near
by; the five processions coalesce into one great spectacle of
elephants and men, officials in pompous Kandyan costumes of
white, purple and gold, drummers bare-bodied leaping and turning,
a din of their drumming and chanting, the whole illuminated by
the flames of resin torches. Probably no other occasion attracts so
many Sinhalese visitors from all over the island and specially the
Kandyan provinces; the procession is certainly a <1pinkama>1 and
attendance is meritorious; but the Tooth Temple is not one of the
sixteen great places of pilgrimage, because the Buddha never
visited Kandy. We shall return to this question.
   To trace the progress of other individual relics is beyond the
scope of this work. Relics have always followed close upon the
spread of the doctrine; the arrival in Ceylon of the Buddha's
begging bowl and right collar-bone and a branch of the sacred
Bo tree were mentioned in the Introduction. Most recently the
Ceylonese Buddhist Mission to England has been supplied with
relics from Ceylon; the first relic was judged perhaps too obscure
when Mrs. Bandaranaike, then Prime Minister, arrived in London
with a second, taken from Mihintale$ (the site of Mahinda's arrival
and conversion of King Deva$nampiya Tissa), and astonished the
inhabitants of Chiswick with a <1peraha%ra>1 including an elephant
from a circus.<s21>s
   Relics may be of the Buddha or of any <1arhat>1, male or female.

                                                                105

By the excavation of a stupa in Kashmir in which was found an
inscribed reliquary containing the bones of Majjhantika, striking
confirmation was given to the historical accuracy of the <1Maha$-
vamsa>1, which records this <1arhat's>1 mission to Kashmir in chapter
XII. However, in practice only relics of the Buddha are impor-
tant.<s22>s They have been divided by doctrine<s23>s into three classes :
<1sa$ri$rika>1 (pieces of the body), <1pa$ribhogika>1 (things he used) and
<1uddesika>1 (reminders, i.e. representations). It is these three classes
of relics, or their most typical examples, which are listed in this
Pali verse, known by heart by most villagers and recited by the
pious laymen who take the Eight Precepts at the temple on <1poya>1
days :

   <1Vanda$mi cetiyam* sabbam* sabbat*t*ha$nesu pat*it*t*hitam*
   Sa$ri$rika dha$tu maha$ bodhim Buddharu$pam sakalam sada$>1.

   I worship always every shrine, standing in every place, the bodily
   relics, the great Bodhi tree, and evely image of the Buddha.

   The worship of relics, as Obeyesekere put it in his lectures,
resolves the clash of the cognitive fact of the Buddha's absence
with the psychological (I would say affective) fact of his presence.
In a recently published article<s24>s he suggests the linguistic and
doctrinal origins of the belief that an enlightened person is
immanent in his relics. Linguistically, <1dha$tu>1 in Sanskrit and Pali,
and hence in learned Sinhalese, means an "element'; for instance,
it can refer to the essences of man and woman embodied respec-
tively in semen and a posited equivalent female fluid. This does
not mean (<1pace>1 Yalman)<s25>s that an ordinary Sinhalese villager
knows that the word can mean "semen' and hence that relics have
connotations of pollution or sexuality.<s26>s It does, however, mean

106

that unlike the word "relic', which suggests something merely left
behind, the word <1dha$tu>1 suggests a basic constituent, perhaps even a
sort of essence. This of course applies in the first instance to the
ancient Sanskrit and Pali, and is not a Sinhalese innovation. Far
more interesting, to my mind, is Obeyesekere's hypothesis that
this conception of relics derives from one of the "undecided ques-
tions' mentioned in the Introduction (p. 7). The Buddha would
not say, when asked, whether an <1arhat>1 after death exists or does
not exist or both or neither. Though that can hardly have been his
intention, his silence leaves room to conjecture some sort of after-
life for the enlightened. Moreover, the doctrine that Enlighten-
ment, the <1summum bonum>1, is a "blowing out' of the personality is
counter-intuitive. "Thus the dha$tu is the visible representation of
the immortal nirva$na state.' From the orthodox doctrinal point of
view this is a misunderstanding, but it could date from the earliest
times. As Obeyesekere shows, it fits in well with the popular
ascription to relics of the same abnormal powers of locomotion as
<1arhats>1 are supposed to possess, typically the ability to levitate and
fly.
   Such holy powers should, by the logic of this argument, inhere
only in bodily relics. Indeed, certain functions are reserved for the
bodily relics. With a few celebrated exceptions, such as the
Buddha's begging bowl (<1pa$tra>1), they are the only ones to be
enshrined. Enshrinement can be in stupas, in relic caskets (which
usually have the form of miniature stupas), or in images--which
will be explained below. Relics are of course usually to be found
in temples, though exceptionally one is owned by a layman. In
theory it is not a relic which makes a temple holy--it is monks who
sanctify it by their presence--but in my researches I came across
only one temple which did not claim to possess a relic. Though
they are of course handled with the greatest veneration, in a wider
sense these relics are casually dealt with: I invariably asked after
the origin of a relic, but never got any reply more interesting than
that it was inherited from the monk's teacher (<1parampara$vin>1). This
is not the uncritical attitude of fanatical regard, which would
retail some elaborate myth, but an indifference entirely proper to
monks. However, I think the indifference is shared by laymen: I

                                                                  107

was sometimes told quite casually that some relic had been lost
("na%ti vuna$'). These village relics are indeed not very impressive
objects: as a special favour I was shown those in Mi$gala, precious
casket removed to reveal precious casket, until the last tiny stupa
contained a couple of minute white balls of what I presume was
bone.
   When monks formally visit a layman's house, typically to
receive a meal or to recite <1pirit>1, they probably take with them the
small temple <1karan*#duva>1 containing relics, and there is a <1peraha%ra>1.
(A <1perahara>1 need not be religious at all, but usually it is. The sort
of <1peraha%ra>1 I am discussing involves monks or relics, probably
both. There are also <1peraha%ras>1 for gods, in which their images are
carried.) On the way to or from a <1pirit>1 the sacred books (<1pirit pot>1),
palm-leaf manuscripts, are carried and treated in the same way as
the relic. The relic casket, covered with a rich cloth, is carried on
a cushion on a man's head, and a canopy is held over it. It is
preceded by drumming and, on grander occasions, dancing;
small boys run ahead throwing fire crackers, other children carry
white pennants tied to sticks, and maybe Buddhist flags, and all the
laymen frequently shout "Sa$dhu sa$dhu sa$'. The monks walk
behind the relics, carrying their umbrellas. When the procession
reaches its destination it halts for a final climactic salvo of drum-
ming and then dissolves. Before monks enter the building they
take off their sandals, if any, and as they reach the threshold their
feet are washed by one layman and dried by another. (The reader
who recalls biblical parallels should be reminded that in Ceylon
such personal services to monks are always performed by men.)
In a well-conducted ceremony the monks from this point walk
on white cloths, the provision and spreading of which is a special
function of the washer caste. At a really big <1peraha%ra>1 like the
Kandy A%sal*a Peraha%ra these cloths are used throughout the pro-
cession, but this is too extravagant for vnlagers. Inside the house
the relic casket (with manuscripts, if brought) is put on a table;
the essential is that it should be higher than the seats of those
present. When monks are fed, a portion may be set aside in front
of the relics. I saw this done several times by middle-class people
in cities. It is a substitute for the normal <1Buddha pu$ja$>1 (offering to
the Buddha) which is made at every temple before every meal.
Such offerings are discussed below. In the village, however, the
<1Buddha pu$ja$>1 is usually sent straight up to the temple to be offered

108

in the normal way, so no food need be reserved at the scene of the
<1da$ne$>1. When the monks have been satisfied the remainder of the
food is served to the guests, but the <1Buddha pu$ja$>1 may be given
only to animals or beggars; in cities it is often just thrown away.
The return procession is similar to the arrival but much less
ceremonious.
  The commonest <1pa$ribhogika dha$tu>1, object used by the Buddha,
is a Bo tree. Under its shade he attained Enlightenment. There is
one in almost every temple, and some grow elsewhere besides.
Wherever a Bo tree grows it is sacred: it may not be cut down, and
it may receive offerings. The nature and rationale of offerings
(<1pu$ja$>1) made before relics will be extensively discussed below.
Here I shall only remark that one type of offering seems to have
entered Buddhism from earlier tree worship via the Bo tree: a
piece of cloth, nowadays formalized as a flag (<1kod*iya>1), may be tied
on to a branch.
  It is the <1pa$ribhogika dha$tu>1 which provide occasion for pil-
grimage. In the canonical text already quoted which describes the
Buddha's death is a passage in which the Buddha, on the last day
of his life, recommends pilgrimage to four places: where he was
born, attained enlightenment, preached the first sermon, and died.
These places should be seen and admired, and if anyone dies with
a contented mind while on a pilgrimage (<1cetiya-ca$rika$>1) they will
be reborn in heaven.<s27>s These four points, or the places believed
to be them, are the supreme goals of pilgrims; every year tours
are arranged from Ceylon, and several of my monastic informants
had been on them. However, more places of pilgrimage have been
found nearer home, on the above model: anywhere visited by the
Buddha provides occasion for religious awe, so he visited Ceylon.
  According to the first chapter of the <1Maha$vam*sa>1 the Buddha
paid three visits to Ceylon, arriving by air. His first visit was to
Mahiyangana, at the eastern foot of the central massif; his second
to Na$gadi$pa (Tamil: Nainativu), an island near Jaffna in the
extreme north. On the third visit he went to Ka%lan*iya, on the west
coast near Colombo, and on his way back planted his footprint on
Siripa$da, known in English as Adam's Peak, in the south-west
part of the hill country, the second highest mountain in Ceylon
and by far the most spectacular; he further tarried and meditated
in a cave at the foot of Siripa$da, at Di$ghava$pi$ on the east coast, and

                                                                   109

four or five spots in Anura$dhapura. According to another ancient
version (<1Samanta-pa$sa$dika$>1, I, 89) the third visit included Muti-
yangana, in the south-eastern highlands. What these places have
in common is that, like the Bo tree, they were used by the Buddha.
At Mahiyangana the Buddha also left some hairs of his head and
that temple later miraculously acquired a bone of his neck;<s28>s at
Na$gadi$pa he left the seat from which he had preached and the
tree which had served as his parasol, themselves <1pa$ribhogika
dha$tu>1; but at the other places he originally left nothing, and
Siripa$da, the place of pilgrimage <1par excellence>1, has nothing
matenal to offer the worshipper except the Buddha's footprint,
recently covered in concrete.
  There are altogether "sixteen great places' (<1sol*os maha$stha$na>1)
which are said to have been visited by the Buddha and are therefore
sites of pilgrimage. Such a group of sixteen is alluded to in the last
chapter of the <1Cu$lavam*sa>1 (C. 128 and 253) but for reasons given
below I suspect that the group is far older. The list is not stable,
and I think I have discovered the basic reason for its instability:
in the passage in the <1Maha$vam*sa>1 describing the Buddha's third
visit to Anura$dhapura, and in the passage in the <1Di$pavam*sa>1 on
which it is at least partially based, there is an ambiguity,<s29>s so that
they can be interpreted to mean that the Buddha stopped at the
future site of the Bo tree <1at>1 the future site of the Maha$viha$ra, or
to make these into two separate stops. The other points visited in
Anura$dhapura, as listed in the <1Ma$havam*sa>1, are the Ruvanva%li
sa%$ya, the Thu$pa$ra$ma and the Lanka$ra$ma. There are thus eleven<s30>s
<1or>1 twelve<s31>s sites which according to fifth-century sources (the
<1Maha$vamsa>1 and the <1Samanta-pa$sa$dika$>1) were visited by the
Buddha, and the number sixteen has to be completed by the
addition of four or five names. Candidates for the other places are
Tissamaha$ra$ma near the south-east coast (always included);
Kataragama, near Tissa; Mihintale$, very near Anura$dhapura;
and at Anura$dhapura the Mirisava%t*i sa%$ya, Abhayagiri, and

110

Jetavana. When the Buddha is supposed to have visited these
places is obscure to me, but their other sacred associations are all
recorded in the first part of the <1Maha$vamsa>1.
   What can we deduce from the size and shape of the list about its
history? Though I know of no ancient evidence for a list of sixteen
places of pilgrimage, the number sixteen has ancient connections
with stupa worship in Ceylon:

"On the lowest terrace of a stupa there were sixteen marks of footsteps
known as <1pa$dapi$thika$>1 fixed at regular points round the cetiya. They
indicated the places where the pilgrim should stop and kneel down and
worship in the course of his circumambulation, after offering flowers at
the upper terraces. . . .'<s32 ">sGurul*ugomi . . . (12th c.) says that there were
sixteen <1pa$dapi$thika$s>1 fixed at the Ruvanvalisa%$ya indicating the points
at which sixteen golden Buddha-images were enshrined inside the
cetiya. . . .'<s33>s

  As for the list's composition, it is notable that none of the
superb sites of Pol*onnaruva are included, but that all the places
named have sacred associations going back at least to the fifth
century A.D. Moreover, it can hardly be a coincidence that half the
places listed are so well distributed round the country, while the
other half are in the ancient capital. It certainly looks as if the list
antedates the Pol*onnamva period, and symbolically brings the
whole of Ceylon under the religious suzerainty of Anura$dhapura.
  These then are the sixteen places of pilgrimage, fifteen of which
the devout Buddhist hopes to visit at least once in a lifetime. (Fif-
teen, because the site of DiWaguha$ ("Divine Cave'), in which the
Buddha spent a day at the foot of Siripa$da, has not been deter-
mined.) A pilgrimage is simply called a <1vandana$va>1 ("worshipping')
or a <1pin gamana>1--a "merit journey'. It may well be the longest
journey in a villager's life. It is normally undertaken in large
groups, without differentiation by age or sex. A monk may go too,
in which case he is of course provided for by the others, but I do
not think monks travel on more pilgrimages than do devout
laymen. Pilgrimages may take place at any time, but there are
recognized seasons for the most famous spots, and nowadays the
Government makes appropriate arrangements. The season starts<s34>s

                                                                111

with Siripa$da, which is usually climbed in the four months ending
at Wesak (in May), because that is the part of the year when it is
least likely to rain there and the sunrise may be seen in its full
glory. Siripa$da is the only pilgrimage necessarily to involve physi-
cal exertion, but the 3,000-foot climb up steep stone steps, up by
night and down in the morning, deters neither the aged nor the
pregnant woman with another baby on the hip; I went along
with such a cross-section of Mi$gala, and all of us made the ascent in
three hours and the descent in one and a half. Poson, in June, is
the time to visit Anura$dhapura, because it is the anniversary of
Mahinda's arrival in Ceylon. The fortnight before the full moon
in August (A%sal*a) is the time of Ceylon's two most famous
religious festivals, though neither of them is on our list of pil-
grimages: the procession for the tooth relic in Kandy and the
festival of the god Kataragama at his eponymous centre in the
south-eastern jungle. A Buddhist's visit to Kataragama, whether
it is considered one of the sixteen places or not, is inevitably com-
bined with a pilgrimage to nearby Tissama$ha$ra$ma; witnessing the
procession of the Tooth is anyway meritorlous. Finally, September
is the month of the <1peraha%ra>1 at Mahiyangana.
  In the old days, of course, people walked on most pilgrimages,
but nowadays they hire a bus from one of the companies which
cater principally for this demand.  Some enterprising people
decide to organize a trip, calculate that they can fit fifty or sixty
bodies into a bus, and divide the cost equally among the lay
travellers. Some buses take pilgrims to all the sixteen places, and
several besides, in a tour lasting a week or more. These buses
have materially altered the character of pilgrimages. The senti-
mental outsider may deplore the rapid disappearance of the
religious picnic party adventuring through the forest towards some
scene of ruined grandeur. The safer bus journey to a repaired and
whitewashed antiquity, electrically illuminated by night, may seem
less romantic, and may exclude the very poor, but few Kandyan
villagers in the old days can ever have reached Na$gadi$pa, and the
wider travel may even have marginal political advantages in in-
creasing national sentiment. The ritual character of the pilgrimage
has been equally altered: all improvements in communications
work to the detriment of purely local deities. Less mobile villagers
used to worship the local god before leaving his territory; then the
god's overlord, the bigger god who controlled a wider area; and

112

so on till at their destination they reached the Buddha, lord of all
and god above the gods (<1deva$tideva>1). The remains of these cus-
toms, if any, are hard to recognize. On leaving Mi$gala for Siripa$da
no one ever dreamt of bothering with our local god, Pit*iya Deyyo$;
but on our way we stopped in Kandy to pay our respects to the
Buddha at the Temple of the Tooth, and as we climbed the
mountain we chanted.

           <1Ape$ Budun api vann*#dinna
           Saman deyyo$ pihit*a vent*a>1.

           We praise our Buddha; god Saman, help us.

That Saman is the god of Siripa$da everyone knows, but few know
how far his territory extends. Much less obvious is the vestigial
nature of our stop in Kandy: to go from the Buddha's tooth to the
Buddha's footprint is a geographical, but not a religious ascent;
however, we were worshipping the local and familiar relic before
venturing into the unknown.
  The two classes of relics already discussed have probably been
venerated since the Buddha's death, worshipped by pilgrimage
and prostration, like an old and respected teacher. The third class,
the <1uddesika dha$tu>1 which "indicate' the Buddha to remind one of
him, are an innovation, though an ancient one. Though the very
shape of a stupa, through its association with physical relics, has
itself acquired the status of an <1uddesika dha$tu>1, the most important
example is the image of the Buddha, his representation in painting
or sculpture. The Sanskrit word for any image, <1pratima$>1, has
become specialized: in Ceylon it and its derivatives, such as
Sinhalese <1pil*imaya>1, refer only to sacred sculpture, for the sculp-
tured image is culturally the important one.<s35>s
   In the earliest Buddhist art, which is of course Indian, the
Buddha was represented only symbolically by an empty seat
under a Bo tree, a footprint, an umbrella, or the <1dhammacakka>1, the
wheel of the law which he discovered. Or probably it would be
more accurate to say that the Buddha was not shown at all, to
symbolize the fact that he was <1nibbuta>1 ("extinguished'), and certain
emblems served to indicate the scene portrayed. Where and when

                                                              113

his body was first depicted is still disputed: art historians say that it
was either in Gandha$ra on India's north-west border, where art
was influenced by contact with the Roman Empire, or well within
North India in the Mathura$ school of art, and that it probably
happened some time in the last two centuries before Christ. They
thus reject or ignore the evidence of the <1Maha$vam*sa>1, which says<s36>s
that a stone image of the Buddha was set up by Deva$nampiya
Tissa, i.e. in the third century B.C. This image was the most
famous one in ancient Ceylon; it is frequently referred to in the
chronicles, and was probably the one which so impressed the
Chinese pilgrim Fa Hsien in the early fifth century A.D. (He says
the image he saw was made of green jade, but that is most improb-
able.) It is possible that it is the same as the famous <1sama$dhi
pil*imaya>1 (seated Buddha image) still visible (though disfigured by
restoration) on the Outer Ring Road at Anura$dhapura; though
this statue is far less tall than the two <1chang>1 (about 22 feet) vaguely
claimed by Fa Hsien. This <1Maha$vam*sa>1 testimony occurs in an
account of events in the third century A.D., and was written still
later; the author may have been mistaken about the statue's
origin.<s37>s Error is, however, even less likely in the long account
(chapter XX:) of Dutuga%mun*u's decoration of the relic chamber
in the early first century B.C. ; he had a golden Buddha image made,
and many other sculptures and paintings which are enumerated.
The similarity of the subjects then chosen to what I saw in the
village temples I visited is quite remarkable.
  Buddha images are sacred objects. They are made by special
people : their manufacture is--or was before the days of mass
production--the prerogative of <1sittaru>1 (literally :  "painters').
<1Sittaru>1 do all the art-work, both painting and sculpture, in a
temple. They form a sub-caste of the <1navandanno$>1, the caste which
includes all types of craftsmen, from carpenters to goldsmiths.
Another person, an artist, may in these modern times make a
Buddha image, but if he has any trace of traditional sentiment
left he will refrain from giving it the finishing touch, which con-
sists in putting in the eyes, and leave this to a <1sittara$>1, who will
perform the appropriate ceremony, a <1ne$tra pinkama>1 ("eye festival').

114

Only then is the image sacred. As Knox put it: "Before the Eyes
are made, it is not accounted a God, but a lump of ordinary
Metal, and thrown about the Shop with no more regard than any-
thing else. . . . The Eyes being formed, it is thenceforward a God.'<s38>s
Though in this passage Knox calls the image a god, he elsewhere
shows that in the seventeenth century the cognitive position was
just as I have explained it: "As for these Images they say they do
not own them to be Gods themselves but only Figures, repre-
senting their Gods to their memories; and as such they give to
them honour and worship.'<s93>s The <1pil*imaya>1 is an <1uddesika dha$tu>1
But it is also usually the repository of a <1sa$ri$rika dha$tu>1. For
Buddhaghosa, it seems, an image was not itself a relic (<1dha$tu>1), and
was of importance only if it contained one: he only mentions
regard being paid to an image (<1pratima$>1) which has a relic (<1sadha$-
tuka>1). I have mentioned above that to this day portable relics are
always in a container shaped like a stupa, or else in a small Buddha-
statue; this must have been how Buddha images acquired their
sacred character; and the doctrine of the <1uddesika dha$tu>1 is still
very close to the viewpoint, which seems to have been Buddha-
ghosa's, that an image is worshipped only for the relics it contains.
    A relic is not worshipped merely by contemplation or physical
gesture. Offerings are made, and such an offering is called a
<1pu$ja$>1: the word does not differentiate between the act and the
thing offered. Typical offerings which may be made before a
<1Budupil*imaya>1, a Bo tree or a stupa are flowers, incense and lights,
flowers being much the commonest. Villagers visit the temple on
<1poya>1 days and other religious occasions to make these offerings
but more often they are made at home. Some homes have a picture
or small image of the Buddha,<s40>s before which these offerings are
laid on a small altar. The word which here and throughout I
translate "altar', <1mal a$sana>1, literally means "flower seat'. In addition
the main <1Budupil*imaya>1 in a temple is offered the same meals as the
monks.
   Most of the offerings I am about to describe are accompanied
by Pali verses. With one unimportant exception (<1Pu$yemi Bud-
dham*>1 . . . , the second verse cited below) all these verses, as well

                                                                   115

as several more, are reproduced by Hocart in his account of the
ritual at the Temple of the Tooth in Kandy.<s41>s It seems a fair
guess that the verses are the same age as the rituals they accom-
pany, but unfortunately their origin is unknown. The best that we
can do at the moment is to surmise that these and the other
popular devotional verses printed in the <1Bauddha A%dahilla>1 and
elsewhere, including the <1Jayamangala Ga$tha$>1, were composed in
mediaeval Ceylon between say 1200 and 1500. In my opinion a
later date is most unlikely, as the sixteenth century saw a great
decline in Pali learning. This does at least suggest that Christian
influence can have played no part in their formulation. Rules for
the worship of the Tooth, quoted in Hocart (pp. 34-7) from the
<1Dal*ada$ Sirita>1 give no details of the rituals but show that they
were already elaborate. They date from soon after 1300.
  The reader interested in ritual procedures should study Hocart.
(I am told by Mr. H. L. Seneviratne that what Hocart describes
still takes place.) I describe only the rituals I witnessed in the
village, which are simpler forms of the same; moreover, my main
interest in the rituals is in their ideological implications, and I
arrange them accordingly. The rituals in the Temple of the Tooth
form an ordered sequence not perceptible in village temples.
  I shall begin with those offerings which are both the commonest
and doctrinally the least problematic. In towns flowers can be
bought outside every temple; in villages they are readily available.
A flower should be offered without leaf or stalk, and each bloom
is laid down to face the image (or Bo tree, etc.). This is merely the
kind of good manners one shows to any superior, to whom one
does not present one's feet or one's back. According to a sermon
on the subject delivered in Mi$gala at Wesak, the custom of offering
flowers was started (and is still practised) by the gods. When
offering a flower it is usual, but not essential, to recite these two
Pali stanzas.

<1Van*n*a-gandha-gun*opetam etam* kusuma-santatim*
Pu$jaya$mi munindassa siripa$da-saroruhe>1.

This heap of flowers, which has colour and scent, I offer at the blessed
  lotus feet of the lord of sages.

116

<1Pu$jemi Buddham* kusumen' anena pun+n+ena-m-etena ca hotu mokkham*.
Puppham* mila$yati yatha$ idam* me ka$yo tatha$ ya$ti vina$sabha$vam>1.

I make offering to the Buddha with this flower, and by this merit may
  there be release (<1moks*a>1). Just as this flower fades, so my body goes
  towards destruction.

   Not every kind of incense is appropriate to the Buddha; those
with grosser fragrance are reserved for lower beings. The Buddha
receives joss sticks, which are lit and left to smoke before him.
They are generally confined to the worship of an image, maybe
because the sweet smell would be lost in the open air. These are
the verses which accompany respectively an offering of incense
(<1suvan#da>1) and of "fragrant smoke' (<1suvan#da dum>1), i.e. burning
incense.

<1Sugandhi-ka$ya-vadanam* ananta-gun*a-gandhinam
Sugandhina@ham* gandhema pu$jaya$mi Tatka$gatam*>1.

To him of fragrant body and face, fragrant with infinite virtues, to the
   Tatha$gata <i.e. the Buddha> I make offering with fragrant perfume.

<1Gandha-sambha$ra-yuttena dhu$pena@ham* sugandhina$
Pu$jaye pu$janeyyam* tam* pu$ja$-bha$janam uttamam*>1.

With this fragrant smoke full of perfume I make offering to the supreme
   recipient of offerings, worthy to receive them.

   The lights offered nowadays by individuals are sometimes
candles, the influence of western technology and the Roman
Catholic church, aided maybe by Col. Olcott. In villages the usual
light offered is still an oil lamp, a little clay dish containing a strip
of cloth lying in coco-nut oil. On special occasions after dark the
entire premises of a temple may be outlined with such small
lamps; in Mi$gala this was done to commemorate the anniversary
of the death of D. S. Senanayake, independent Ceylon's first
prime minister. A few temples in the area have provision for a
perpetual lamp to burn, a <1dol*os maha pahana>1 (literally "twelve-
month-lamp'). This is still a coco-nut oil lamp as described, but a
big one on a separate structure, e.g. a short pillar, standing in the
courtyard, with some protection from the wind and rain. Here is
the Pali verse for offering lights (<1pahan>1).

<1Ghana-sa$ra-ppadittena di$pena tama-dham*sina$
Tiloka-di$pam* sambuddham* pu$jaya$mi tamo-nudam*>1.

                                                                  117

With this lamp which blazes with firm strength, destroying darkness, I
   make offering to the truly Enlightened lamp of the three worlds, the
   dispeller of darkness.

   The order in which I have presented these verses corresponds
to how well they are known: the flower verses are in the first book
of the Sunday School primer, and known to every normal person,
whereas I think only the very devout layman would know the last
two.
   Flowers, incense and lights are the only offerings commonly
made within the home. It is the remaining types of <1pu$ja$>1, those
more peculiar to public places, which raise the more complicated
problems. Before considering these, however, let me deal with the
question raised by any offerings before relics and images : can such
worship be reconciled with the cognitive position that the Buddha
is dead and powerless to help? Yes certainly--it can and it is.
Such offerings raise not only the question of the ontological status
of the Buddha, with which we are here concerned, but also con-
nected problems concerning "prayer' (which will be discussed in
Chapter 5) and concerning ethics (see Chapter 6) of which I can
here do little more than take cognizance. However, it is essential
for an understanding of the Therava$din position that I here give a
brief explanation of the rationale for all such offerings.
   Ask any monk, and the cognitive position is quite clear: no
offering, no flowers, no recital of verses has any intrinsic merit; it
is the thought that counts. The Buddhist ethic is an ethic of inten-
tion; and doctrine is consistent on this point. We may accept this,
but question whether the intention in offering flowers or food is
not to receive some favour in return. The answer is no. What then
is the intention? There are two ways of answering this. The first
is to say that there is no further intention: the thought itself, the
emotion on the mind of the worshipper, if it is pure (<1s/uddha>1),
makes for what we might translate as spiritual development (<1hita
diyun*uva>1)<s42>s which is furthered by the earnest aspiration to
achieve <1nirva$n*a>1 which should accompany the offering. This
answer is doctrinally orthodox and clear. The other answer might
be to say that there is an intention to acquire merit (<1pin>1). This
perhaps begs some further questions, as it seems to be coming
close to granting to an act the intrinsic merit which has just been
denied; but the answers of an adept informant will take us back

118

to the same position as that disclosed by the first answer. The
relation between these two answers, which cognitively can be
harmonized, but affectively seem to differ, will be examined in
Chapter 6. For my present purpose it suffices to note that all
answers come down to talk of pure thoughts, and emphatically
deny that the Buddha is seen or considered as a god, still alive or
powerful.
   The nature of the emotion which the worshipper is supposed
to feel might be conjectured from some of the verses. The stanza
on lights contains a play on words which gives it a philosophical
twist: the darkness which the Buddha dispelled was the darkness
of ignorance. The second stanza in the flower offering contemplates
the transience of the body. Are such philosophic moods the ones
that are experienced or supposed to be experienced? Very many
people, though they do not know Pali, have had the meanings of
the stanzas explained to them, and monks certainly understand
them precisely. However, when I asked the head monk at Mi$gala
about <1mal pu$ja$>1 he said that they are not for the Buddha, who
is dead, but for us to derive joy (<1pri$ti$>1) from looking at them.
Buddhaghosa says<s43>s that one gets <1Buddha$lambanapi$ti>1, joy derived
from contemplation of the Buddha, by looking at a Bo tree or
<1caitya>1. The head monk's reply was therefore in this tradition in
describing the emotion of a worshipper as <1pri$ti>1, joy, though he
gave it an interesting and perhaps idiosyncratic twist by making the
flowers themselves the immediate reason for joy. What occasions a
pure emotion is, however, irrelevant, so the discrepancy is unim-
portant. It is more interesting that despite the clear contrary
implication of some of the verses recited for particular offerings,
the general emotion felt to be appropriate to <1pu$ja$>1 is joy.
   We now come to consider offerings and ceremonies which are
commonly said to be borrowed from Hindu practices, and thus
incur the suspicion of Hindu ideological backing. Most Ceylon
Tamils (the Hindus in question) worship at temples containing
images which are believed to incorporate a god and thus to be in
some sense alive and divine. It is, however, possible, <1prima facie>1,
that the Buddhists should have adopted ceremonial forms and
changed their meaning; this becomes all the more plausible if
the forms turn out on closer inspection to have been modified.

                                                                  119

   Here is the first example. Before every meal in the monastery
the principal Buddha image has a portion of the food, etc., laid
before it. The shrine is emptied for a couple of minutes; then the
food is taken out again, and thrown away or handed to dogs or
beggars. No self-respecting Buddhist would touch it, or anything
which has been offered to the Sangha, which is in the same cate-
gory. In Mi$gala I think it was usually eaten by dogs, as a normal
<1Buddha pu$ja$>1 is too small to be worth the trip for a beggar. But
when lots of monks were fed and so plenty of food was left over
Tamil beggars assembled behind the temple and had a feast. In
their eyes such food is not even degrading, for after a Hindu <1pu$ja$>1
the worshippers share the food which has been offered; but to the
villagers this was a vivid illustration of the beggars' degradation,
and among themselves they made some quite nasty comments to
this effect. The food is offered before the Buddha image by who-
ever is giving the meal to the monks, or by a monk if it comes from
their provisions. (Food is supplied by laymen in some monasteries
always, in others almost never, in the rest somewhere in between.)
Monks have two meals, breakfast and the midday meal, and in
the evenings they drink tea, maybe suck sweets, and are offered
also the materials for chewing betel and other things they may
like to chew. A portion of all these is first offered to the Buddha
image; e.g. the midday meal, which the monks eat at 11.30, is
offered to the statue at 11 or soon after. The Pali stanza which
may accompany this runs:

<1Adhiva$setu no bhante bhojanam* parikappitam*
Anukampam upa$da$ya pat*igan*ha$tu-m-uttamam*>1.

May the monk accept the food we have prepared; taking compassion
   (on us) may he receive the best.

In my translation the verse is addressed to a monk; the Pali says
<1bhante>1, which is a normal way of addressing a monk in the Canon.
The Buddha too was a monk, so the word makes sense here. It
crops up here because <1adhiva$setu no bhante>1 is a common phrase
which (like <1anukampam upa$da$ya>1) may be used when asking a monk
to administer the precepts (<1Pan sil>1 or <1at*a sil>1). According to the
<1Nida$nakatha$>1, the two merchants who give the Buddha his first
food after Enlightenment say:

<1Pat*igan*ha$tu no bhante Bhagava$ imam a$ha$ram* anukampam upa$da$ya>1.

120

May the monk, the Blessed One, accept this food from us, taking
   compassion (on us).

The above verse is merely a versification of this prose formula.
There are minor variants for this verse, most of which are to
modify its application: instead of <1bhojanam*>1: "food' one may offer
<1vyan+janam*>1: "curry', <1taralam*>1: "rice gruel', <1khajjakam*>1:  "sweets',
<1pa$ni$yam>1: "water', <1pa$nakam*>1: "drink', <1bhesajjam*>1: "medicine', or
<1gila$napaccaya>1, which also means "medicine' (literally "requisites
for disease'), but in its Sinhalese form <1gilampasa>1 is the generic
term for the evening <1da$ne$>1.
   At this point it may be objected that the production of senti-
ments of joy or philosophic contemplation may be an adequate
explanation for the comparatively insubstantial offerings of
flowers, incense, and lights, but that when food is offered surely
something else must be involved. The <1ide/e rec|ue>1 runs something
like this: flowers, etc., are the true Buddhist offerings; food is
offered to their gods by Hindus (who make no careful distinction
between the image and the presence of a divinity); food is laid
before the Buddha image just as it is laid before statues of Vishnu
or Kataragama (which may even be found in the same shrine);
therefore the offering of food to the Buddha is a practice borrowed
from the Hindus. The fact that the food thus offered is rendered
sacred, in that no Buddhist will afterwards eat it, is sometimes
adduced in support of this theory. To take the supporting argu-
ment first: the sacredness of the offering is of a very different
character, for in a Hindu <1pu$ja$>1 food which has been offered to the
god is then distributed among the worshippers in a kind of com-
munion. The practice implies that one derives benefit from the
close contact of one's food with a divinity. The Buddhist attitude
we saw to be just the opposite: food offered before an image is
treated as if it were <1sa$nghika>1, i.e. like something given to the
Order, and for this reason no longer allowed for lay use.<s44>s This
distinction is perfectly clear, though maybe not so theoretically

                                                                121

expressed, to the Buddhist inhabitants of Mi$gala. Some of them
occasionally visit a Hindu temple to Kataragama about 5 miles
away which is run by and mainly patronized by Tamils. Like the
Hindus they take food, offer it to the god, and then eat their share
of what has been offered. They accept the Hindu theory that
Kataragama has enjoyed the food, and have no theory about why
they then eat it themselves. But all this has no effect on their
behaviour in a Buddhist temple. An interesting question is what
happens to food which is offered to, say Kataragama, in a Buddhist
temple. On this I cannot speak from experience. In Mi$gala temple
there were no statues of gods at all, which is rather unusual. In
most temples in the area there are statues of gods, but most of
them are purely decorative. Even those which are worshipped are
never, to my knowledge, given offerings of food. But in some
Buddhist temples in other districts a god is offered e.g. the <1daval
da$ne$>1 after another portion has been offered to the Buddha. Some
of this can be returned to the worshipper according to the Hindu
custom. Every god who is worshipped and receives offerings has
his own priest (see next chapter), who in fact of course receives
the offerings for him; this would apply also to offerings of food,
which are not <1sa$nghika>1 and are therefore fit for human con-
sumption.
   But the offering of food to the Buddha need not be explained
by a diffusion of Hindu practices in Ceylon. Buddhaghosa says<s45>s
that wise men before a meal offer food and drink to an image or
casket (<1cetiya>1) containing a relic which they place before them.
This is exactly the practice we saw to be followed today at a <1da$ne$>1
when the <1Budu pil*imaya>1 in the temple is too far away for con-
venience. The offering before the portable relic is the older cus-
tom: to make it before the large, stationary image which contains
a relic is but a logical extension. This is not to deny that Hindu
influence may have been helpful in formulating modern proce-
dure : the <1Buddha pu$ja$>1 at mealtimes is no doubt sociologically over-
determined. This conclusion is reinforced by the tenor of other
offerings and ceremonies.
   The crucial respect in which modern practice seems to have
changed from that recommended by Buddhaghosa is the recitation
of the verse quoted above, in which the Buddha, being asked to

122

accept the food, is addressed as if he were alive. This looks like a
break-through of what I shall show to be the affective attitude
to the cognitive level: feelings that the Buddha is a living presence
here seem to find expression in words. The case is not clear-cut:
the man reciting the Pali formula <1Adhiva$setu no bhante>1, to which
he is also accustomed in other contexts, may not fully understand
its meaning, let alone its implications, nor is he likely to under-
stand the exact meaning of the rest of the verse; after all, it is one
thing to compose such a verse oneself, quite another to mouth a
half-understood conventional phrase in a dead language. Still, when
all is said and done, I think the words do amount to a cognitive
inconsistency, which is not removed by saying outside the
immediate context of the <1Buddha pu$ja$>1 that of course the Buddha is
not there to hear the words or accept the food.
   One very particular <1da$ne$>1 had best be mentioned here, the
<1kiripid*u pa$traya>1 ("bowl of milk-rice'). This is--or was--an annual
harvest festival, an offering to the Buddha of first fruits in the form
of a huge bowl of milk-rice. The festival is obsolete in my area,
and my information is unfortunately scanty. I was told that the
rice was prepared in the presence of only male villagers, some of
whom danced as it was cooking. In these respects the festival
sounds similar to the <1ad*ukku>1, offerings of first fruits to the local
village gods, which I mention in the next chapter. On the other
hand I was told that originally the rice was buried after being
offered, though in more recent times it had been given to beggars
which is to say that it got the normal treatment of a <1Buddha pu$ja$>1,
whereas <1ad*ukku>1, like other offerings to gods made off <1viha$ra>1
premises, are eaten by the participants. Moreover, the dish of
milk-rice was offered in commemoration of the dish of milk-rice
which Suja$ta$ offered to Gotama the day before his Enlightenment,
cooking it with honey (though the villagers just boil the rice in
cows' milk). The rice was cooked on temple premises (in Mi$gala
on the porch of the <1viha$ra gedara>1), and if the ceremony was cor-
rectly performed the relic in the stupa emitted rays (<1Budu ra$s>1) in
the five colours of the Buddha's halo. This happened in the days
of my informants' grandfathers. (No doubt the event was always
ascribed to the past.) The ability to emit such rays is a property of
<1sa$ri$rika dha$tu>1 and does not appear as anomalous or miraculous to
those who do not share western ideas concerning the laws of
nature. A schoolmaster was, however, worried by the idea and

                                                                   123

suggested that the rays were an illusion arising from the "fumes' of
the cooking rice.
   We have now enumerated all the types of offerings which are
accompanied by Pali verses recorded in <1Bauddha A%dahilla>1, which
may suggest that those still to be described are doctrinally less
orthodox. Though most of the standard objects which can be
offered have been listed, any offering for the embellishment of a
sacred object or improvement of the shrine can count as <1pu$ja$va>1,
and such miscellanea as a vase for flowers or a table to serve as a
<1mal a$sana>1 need not detain us. However, the offerings of cloth
(<1redi>1) deserve special mention. Before the main <1Budupil*imaya>1
there is always a curtain rail on which hangs a curtain, which is
often of fine material, and may have the name of the donor em-
broidered on it. The custom is an imitation of the curtains which
hag before Hindu images, and hence also before many Buddhist
statues of gods. However, the contrast with Hindu usage is again
instructive.
   In a Hindu temple (<1kovil>1) the image, which is conceived of as
incorporating a living presence, is kept veiled behind its curtain or
curtains and can never be seen by a casual observer. The statues of
gods in Buddhist buildings devoted to them (<1de$va$le$>1) are also kept
behind curtains, but the curtain will always be drawn by the
attendant priest (<1kapura$l*a>1), or indeed by other people, on request.
Those statues of gods which are inside a <1viha$ra gedara>1 may not
even have curtains, and be treated with no respect at all, their
importance diminishing with their proximity to a Buddha image.<s46>s
All this is as much as for the Buddhists to say, "We know that as
Hindu gods you are entitled to such prerogatives as curtains, but
here you are quite secondary to the Buddha.' The living Hindu
images, then, are entitled to privacy, or even secrecy; but there
should be no restrictions on access to a Buddha image, in the first
place because it is not alive, and in the second place because, even
if it were, Buddhism is exoteric and anyone should have access to
the Buddha. I shall have occasion to return to this point.
   The influence of Hindu practice is also suggested by the various

124

cloths which sometimes cover the Buddha's hands or feet and
legs. A standing Buddha has the right hand raised in blessing, and
in some temples this hand is covered by a cloth. A seated Buddha
may have a cloth covering the entire lap. Apart from a cloth, a
particular kind of ornament, which I might call a cloth rosette
but is doubtless intended as a lotus, is often placed on the folded
hands, and maybe elsewhere. This "cloth lotus' is a round cloth
disc, made in patchwork and multi-coloured, with many tiny
"petals' radiating from the centre. The idea of covering the hands
and feet (both of which are painted with auspicious marks) may be
an analogue with Hindu practice; the form of the cloth lotus I can
only surmise to show the influence of flower offerings.
  A final specific offering of cloth is the offering of flags (<1kod*i>1), or
rather pennants. These pennants may be strung up in rows in an
image-house; but the usual practice is to tie them to the Bo tree.
Here the influence is not specifically Hindu; ancient evidence
shows the custom to be a survival of tree worship which has
occasionally gravitated into the shrine. To place <1kod*i>1 in the
<1viha$rage$>1 was not customary in my area, and in Mi$gala and its
immediate surroundings they were not even tied on to trees, so I
am no authority on the practice. It seems clear, however, that
originally any strips of coloured cloth were used, as they still are
in some parts of Ceylon, and the pennant shape is a mere formaliza-
tion, perhaps under western influence. Such pennants, made of
paper, are carried on sticks in village <1peraha%ras>1 by children, and
their shape is that of some church flags, of which they could be a
miniature version.
  Claims made by the worshippers that they are merely com-
memorating a dead man, who lived as a simple monk, become in-
creasingly difficult to justify in the context of the ceremonies still
to be discussed. For clarity of analysis I have now stated the
cognitive claims as two: that the Buddha was a man, and that he
was a monk on the same social level as any monk who attains
<1nirva$n*a>1. The doctrinal position is unequivocal on both points; but
in our context the former seems the more crucial: it is less jarring
to treat an image of a dead monk as if it were a dead king than as
if it were alive. It is over the question of the Buddha's earthly
status that the first discrepancies seem to arise; and I shall show
that they date from ancient times.
   The only common offering which I have not so far discussed is

                                                                    125

the "offering of sound' (<1s/abda pu$ja$va>1), which consists essentially
of drumming. This drumming is the prerogative and duty of a
certain caste, the <1berava$yo$>1. In Mi$gala these drummers are attached
to the temple as feudal villeins; they have a hereditary right to
tenure of lands owned by the temple in return for which they must
perform hereditary services. If the temple agrees the service may
be commuted for a cash rent, but the tenure is always revocable
by the temple. This system of social organization will be men-
tioned at greater length in Chapter 8; for present purposes only
a passing remark on it need be made. Originally there was a wide
variety of such service tenures; indeed it is arguable that the
great majority of Sinhalese castes originated in this way. Norm-
ally, of course, the services are due to people, and in the case of
<1viha$ragam>1, villages owned by monasteries, they are due to monks.
   In the 150 years since the end of the Kandyan kingdom the
feudal organization has gradually declined, and though most
people still work on land owned by others, service tenures in the
strict sense have become a rarity. Drummers, however, perform
a service which is principally required for the Buddha, and is
therefore in continued demand. Any temple of some age--say
fifty years--is almost certain to own land, and some of this land
is probably leased to drummers in return for their professional
services. Although drumming is not the only caste-bound pro-
fession--laundering is another--it is the only one, at least in the
area where I worked, which is regularly associated with service
tenures, and this to such an extent that the term <1ra%$jaka$riyo$>1
literally "workers for the king', which used to designate all feudal
villeins, is now widely taken to refer only to drummers living off
temple lands.
   There are various types of Sinhalese drums, and their use shows
some regional variation. In the Mi$gala area the basic and essential
drum on religious occasions is the <1bera>1, the long drum, tapering at
both ends, which gives the caste its name. The drum is worn
slung round the neck, and struck with the hands at both ends.
In Mi$gala it was invariably accompanied by the <1horan*a%$va>1, a shrill
wind instrument looking like a large and simplified oboe, with a
limited range of notes. The players of the <1horan*a%$va>1 are hereditary
specialists, also of the drummer caste; the playing does not seem
to require much skill, as the notes produced follow no discernible
pattern. In Mi$gala drumming usually consists of one or two <1bera>1

126

and a <1horan*a%$va>1, sometimes accompanied by a <1he$visi>1, a squat
cylindrical drum. The other instruments, a couple of which are
mentioned in my article on the <1ne$tra pinkama>1, are not used in
Mi$gala.
  The paradigm case of a <1s/ubda pu$ja$va>1 is that which occurs at the
temple itself on <1poya>1 days and any other special occasions, such as
a big <1pirit>1. The drums are beaten for a few minutes at three times
on those days, to coincide with the three offerings of food. Ideally
the drumming begins when the food is taken in to the shrine, con-
tinues for a few minutes while it is left there, and stops when the
food is taken out again. Drumming may also occur at other times
of the day, but this is said to be "for no particular reason' (<1nikam>1)
not "as worship' (<1pu$ja$vat*a>1). However, a formal <1s/abda pu$ja$va>1 can
be offered at other times, as an item by itself. I observed this
after a <1tun ma$sa da$ne$>1, at which a householder feeds monks three
months after the death of a relation: drummers stood outside a
shrine and played for a few minutes without any other ritual. The
giver of the <1da$ne$>1 acquired merit by this (merit which he was to
pass on to the dead man), in that he paid the drummers for their
services. In Mi$gala and most other old temples, drummers are
attached to the temple by feudal ties, and so do not get paid.
Temples which do not command the services of drummers almost
invariably hire them on <1poyas>1, though a poor temple may miss out
the quarter (<1atavaka) poya>1 and even the <1ma$sa poya>1 (when there is
no moon). The going rate for such drumming is two rupees a day.
In Mi$gala there was a drummer living near the temple who usually
performed, but for full moon days and other big occasions drum-
mers came from the temple's <1viha$ragama>1 a few miles away,
bedding down in a corner of the <1pansala>1. Even in such cases, when
the drummers perform unpaid, as a duty, the mere making of the
offering is meritorious, and all who participate in thought earn
merit. To the end that more might hear the drumming and hence
have the opportunity to participate in the offering, the head monk
had built while I was there a <1he$visi nan*d*apa>1, a small roofed
enclosure on a rocky prominence near the temple, from which it
was estimated that the sound could carry to the furthest pads of
the village.
   Drumming also accompanies any <1peraha%ra>1 or <1pinkama>1 worth the
name, attending the relics whenever they leave the temple; but
this is not strictly a <1pu$ja$va>1, for such processions with drumming--

                                                                    127

and sometimes with dancing too--are by no means necessarily
religious, but may take place for any secular festivity, from
occasions of state to the homecoming of some newly-wed rich
villager. This identity of religious and secular ceremonial is the
point to which I wish to draw attention.
   The <1s/ubda pu$ja$va>1 is last on my list of regular offerings, and the
only common one for which no Pali verses exist. It is not hard to
imagine why this should be so. In the first place, anyone who has
experienced such drumming inside a <1viha$raya>1 will realize that
during it one cannot hear oneself think, let alone recite. More
important, perhaps, is the non-religious nature of the offering, on
which I shall expatiate after describing the <1su$visi pinkama>1. Finally,
there is a subtle difference in the form of the <1pu$ja$>1, in that other
offerings are presented by the worshipper himself, whereas in this
case he is merely paying others--professionals--to make the
offering.
   The three remaining ceremonies to be described, which deviate
increasingly from doctrinal expectations, are all rather major
affairs reserved for special occasions. The fist can be viewed as an
elaborate extension of the drumming offerings described above. It
is called a <1su$visi pinkama. Su$visi>1 is the old Sinhalese for "twenty-
four', and the reference is to the twenty-four previous Buddhas.
The ceremony, if that is the correct word for an occasion which
does not seem to be highly formalized, consists in Kandyan
dancing by a troupe of dancers and drummers, who meanwhile
chant religious poems, in particular about the lives of the previous
Buddhas. The show goes on for three (or six) nights, theoretically
with eight (or four) dances a night; in my experience the dances
invariably start so late that there is not time for eight before dawn
at 6 a.m., when the meeting breaks up ; but nevertheless the second
and third nights always start with the ninth and seventeenth
Buddhas respectively. Sometimes the three-night <1su$visi>1 is supple-
mented by a fourth night on which the <1sat satiya>1 is danced, one
dance for each of the seven weeks Gotama Buddha spent immedi-
ately after Enlightenment.
   The <1su$visi>1 and similar dances are called <1pante$ru na%t*um>1, "tam-
bourine dances', because the dancers carry tambourines. The
<1pante$ru na%t*um>1 were devised by Mr. H. M. Kiri Banda$ of Hala$-
giriya in Ka%$galla District (between Kandy and Colombo); they
were inspired by certain rituals at the Temple of the Tooth, and

128

first performed in 1912. Initially confined to that part of the
Ka%$galla District known as the Four Ko$rale$s, they have in the last
twenty years become very popular in the Up Country. Despite
their modern date their inspiration is wholly traditional.
  A troupe is led by a "teacher' (<1gurunna$nse$>1), who in the per-
formance faces the dancers, keeping time with a tiny pair of
cymbals called <1ta$lampata>1 and leading the recitation. The poetry
recited may be his own composition, or his teacher's, or derive
from a published source like the <1Arahata Vandana$va>1, a long and
rather miscellaneous religious poem, dating from the Kandyan
period, which Mr. Kiri Ban*d*a$ used before writing his own poems.
In the <1su$visi>1 the dance for each Buddha begins and ends with
verses about him, but the major part in the middle is danced to
religious verses on any subject. The actual drummers always
belong to the drummer caste (<1berava$yo$>1), but the <1gurunna$nse$>1 and
his dancers may belong to any caste. Formerly only males could
dance, but recently--"by popular demand'--girls have been
included: this appearance of dancing girls in a Buddhist temple is
flagrantly untraditional.
  I attended three different <1su$visipinkam>1 for a total of five nights,
and though detailed accounts would here be out of place, it will
be convenient to describe them in outline. They were not arranged
for a date of any religious significance. They were major features
in fund-raising drives for new temple buildings, and from the
point of view of their organizers, the temple incumbents, were
purely commercial undertakings. This is certainly the standard,
perhaps the invariable pattern. The performing troupe was hired
for so many rupees a night (R.150 in one case), but the organizers
could be sure of collecting enough money from lay spectators to
make a considerable profit. Although this commercial motive was
obvious to all, for the spectators the occasion offered both enter-
tainment and religious merit. In this it resembled all <1ban*a>1 (preach-
ing); indeed, a <1su$visi pinkama>1 has much in common with <1ban*a>1,
and will have to be mentioned again under that head; here I am
concerned only with its relevance to worship of the Buddha. The
entertainment element is more obvious than usual with <1ban*a>1, as
spectators could watch singing and dancing; religious merit could
be acquired both by listening to the words about previous Buddhas
and by contributing financially--and of course, as always, by
rejoicing in the acquisition of merit by others.

                                                                129

   The night's events began with two pure fund-raising activities.
First came an auction (<1salpil*a>1) of sundry objects donated by lay-
men. There followed a <1malvattiya vende$siya>1, the sale of a tray of
flowers to the last bidder, a universal feature of temple fund-
raising occasions. An arrangement of opulent flower blossoms
on a tray is shown to the assembled laity, and young men who
act as ushers then move among them collecting donations. In my
experience this takes the form of a competition between villages,
but this competitive element seems quite artificial. Each donation
is announced by an usher standing in a commanding position and
using a microphone at maximum volume; he announces the
amount, and the village of the donor. Each large amount is
greeted by "Sa$dhu sa$dhu sa$'. The event continues for two or three
hours, and the ushers make frenetic efforts to maintain enthusiasm.
The money is not added up, but the flower-tray is "won' for his
village by the last person to make a donation--in other words,
when no more money can be elicited. What is really won is the
right to offer the flowers to the Buddha--not that this confers any
religious distinction, as the flowers are first carried round for
everyone to touch with folded hands and thus share in the merit
of the offering.
  It is only at this point--which is usually after midnight--that
monks appear on the scene. They take no particular notice of the
images of the twenty-four Buddhas; they participate throughout
the proceedings <1de haut en bas, not>1 as worshippers. They are first
offered requisites (<1pirikara>1).<s47>s At one ofthe <1su$visipinkam>1 I attended,
twenty-four sets of monks' robes (<1sivuru>1) had been bought
beforehand by donors who subscribed at least ten rupees (their
names were read out), and the idea was to offer one of these before
each dance. Before a robe was offered it was carried round among
the audience for people to put coins (<1pan*#duru>1) on it as an additional
contribution. Each set of robes was offered before the pictures of
the twenty-four Buddhas; moreover, there were twenty-four
Buddhas but not twenty-four monks present; so in some sense the
robes were being offered to the Buddhas though they were to

130

become the property of the monks. The offering of robes will be
discussed below in connection with a ceremony which affords a
more clear-cut case. Towels (also reckoned as <1pirikara>1) were also
there, to be offered by any person or group who put up two
rupees, a feat only achieved after considerable exhortation from
the incumbent monk. These were offered to the monks direct.
  Before each dance a monk gives a "sermon', which consists of a
reading from the <1Nida$nakatha$>1 or one of its Sinhalese versions in
the <1Pu$ja$valiya>1 or the <1Saddharma$lam*ka$raya>1. He reads out the
story of each Buddha whose story is about to be danced. After the
sermon the dancers appear and begin their performance. They
dance before a screen on which have been put up pictures of the
twenty-four previous Buddhas. On two occasions this took place
in the <1ban*a gedara>1 (preaching hall), on the other under a pavilion
(<1man*d*apa>1) erected <1ad hoc>1. On no occasion therefore were they
dancing in the presence of a consecrated Buddha image, nor do
I imagine that they ever do so, as there would be no room for the
dancers, let alone the spectators, in a shrine room. Nevertheless
I feel entitled to suggest that they dance primarily for the Buddhas,
only secondarily for the spectators. No Sinhalese would put it like
this; they consider the object of the <1pinkama>1 to be to gain merit
and to raise funds, and the question "for whom' the dancing is
intended does not arise explicitly. However, the dancers behave
as if performing a religious ritual: they begin by taking the Three
Refuges and Five Precepts. Moreover, they then dance as if the
pictures were the principal audience, facing them most of the
time; most of the human audience view the performance from
<1behind>1, the monks and other privileged people usually getting a
side view.
  These notes on the <1su$visi pinkama>1 may be completed by an
impresaion of the first troupe I saw (not the family troupe). There
were six dancers, accompanied by three drummers on <1bera>1. The
costume showed red, white, and gold against the dark bodies.
Each dancer had a white headdress, a kind of turban with two
long ends falling waist-length behind him. His only other garment
was a white sarong, bunched in a frill at the waist in front. Over
this he wore a red belt and a red kerchief tied round the waist,
and he had coloured side-pockets on the sarong. Feet, like torso
and arms, were bare, but he wore golden armbands, heavy golden
anklets, and long golden earrings. He held a tambourine, which he

                                                            131

frequently passed between his hands or tossed in the air. The
dancing varied between alow and fast tempi, and during the slow
passages verses were chanted in an unmelodious monotone. For
an outsider the fast passages were much the more exciting; egged
on by the drummers, they went faster and faster, till at the climax
the pattern of group movement would be broken by individual
displays of twirls, somersaults, etc., which reached great acrobatic
virtuosity.
  To explain the form of this ceremony, of which there is no
published account, has involved some digression from my theme,
the compatibihty of such rituals with explicit doctrine. The
seventh precept listed in Chapter 2, which is taken by all monks
and all laymen undertaking any special religious observance, is
"to abstain from seeing dancing, music vocal and instrumental, and
shows'. A <1su$visi pinkama>1 is sponsored, organized, and presided
over by monks, and watched by laity, however pious, as an act of
religious merit. No villagers (or village monks) seem to regard this
as an infringement of the seventh precept, because the <1su$visi>1 and
allied <1pinkam>1 are classed as <1ban*a>1. One might add that many monks
do not scruple to attend performances of Kandyan dancing even
in a completely secular context, classifying it as "culture' (<1sam-
skr*tiya>1) rather than mere entertainment (<1vino$de$>1). Kandyan dancing
is widely supported by the Government and officially considered
educational.
  However, my principal concern here is not the keeping of the
precepts, but the behaviour towards the Buddha (and his prede-
cessors). In conaidering the first set of offerings we saw that the
Buddha is usually addressed as a monk in the accompanying
verses, and behaviour is compatible with this conception. To
offer musical performance in the <1s/ubda pu$ja$va>1, let alone a night-
long combination of music and dancing, is plainly not to treat him
as a monk should in theory be treated. Rather it is to treat him like
a king. This interpretation does not originate with me but with
some of my informants. It is a doctrinal inconsistency of which they
are aware, and which they tend to excuse by saying that these
things were done for the kings of Kandy, who ordered that they
should be done for the Buddha likewise.
   In fact, however, both the dancing and singing before relics and
the explicit attribution of kingship considerably antedate Kandyan
times, for they occur in the first part of the <1Maha$vam*sa>1. It is

132

recorded (chapter XXXIV) of King Bha$tika$bhaya (A.D. 38-67) that
he not only offered the Great Stupa (the modern Ruvanva%lisa%$ya)
incredible masses of flowers, perfumed unguents, lamps, water (as
an offering to stupas outside my experience, but perhaps a transfer-
ence from watering the Bo tree), and jewellery (including "lotuses
of gold, the size of wheels' (line 47), but also arranged in honour
of the Great Stupa "various dances and plays with various instru-
mental music' (line 60). His brother King Maha$da$t*hikamaha$na$ga,
who succeeded him, seems to have been equally munificent; when
he had built the Ambatthala stupa on Mihintale$ he covered it with
a red woollen blanket festooned with pearls and gold balls (line
74), put up flags, triumphal arches, etc., and had dancing, singing,
and instrumental music (line 78). The entire description of this
ceremony is very reminiscent of modern practices, though if the
account is credible modern times have not witnessed anything
comparably lavish. The fact that the focus of a <1su$visi pinkama>1 is
likely to be a <1Budupil*ima>1 loses most of its significance as possible
image-worship if such festivals have traditionally been associated
with all types of Buddha relics, and the claim of the worshippers
that the image is quite secondary gains plausibility.
   The ascription of temporal sovereignty to relics dates virtually
from the introduction of Buddhism to Ceylon, and subsequently
was to become almost a commonplace. King Deva$nampiya Tissa
worshipped (Pali: <1pu$jesi>1) some young Bo trees by offering them a
white umbrella, emblem of royalty, and performing the royal
consecration (<1abhiseka) (Mhv>1. XIX, 59). Dut*uga%mun*u did exactly
the same (same Pali roots) for the relics he was enshrining in the
great stupa (XXXI, 90), and gave them sovereignty over Ceylon
for seven days (XXXI, 111). The purely symbolic nature of such
gestures is surely indisputable; they do not ascribe life or the
capacity to rule to the saplings or the bones. Though, as I must
repeat, the ceremonies I have described are no doubt sociologically
over-determined, it seems to me more relevant for their interpreta-
tion to recall Buddhist practice of 2,000 years' standing than to
equate them with the usage of Hindus who treat their images as
gods and their gods as kings, and provide them with music at meal-
times much as the drums are beaten during the <1Buddha pu$ja$>1.
   Before passing to the next ceremony, let me finally note the
linguistic evidence for what an outsider might call "confusion of
spiritual and temporal power' in the Sinhalese concept of the

                                                              133

Buddha. I have already mentioned that the commonest Sinhalese
word for monk is <1ha$muduruvo$>1, meaning "lord'. Like the English
equivalent, the word can refer to spiritual or temporal authority,
so when the Buddha is referred to as <1Buduha$muduruvo$>1 this
ambiguity is preserved. The only other common title of the
Buddha is <1Budu raja$manvahanse$>1, which may be translated "His
Majesty King Buddha'. In writing of the Buddha, or speaking of
him with formal correctitude, certain honorifics are used; for
instance the ending -<1se$ka>1 after a verb is reserved for him alone.
I believe that all these honorifics were formerly used for kings of
Kandy. Finally, the word <1te$va$va>1, which used to denote personal
services (e.g. fanning) involved in waiting on the king, was used
as a synonym of <1pu$ja$va>1 in the context of the next ceremony to be
described, in which several such operations are performed.
Linguistically the Buddha is the last remaining, the permanent
Sinhalese king.
   The inconsistency with orthodox doctrine in treating a Buddha
relic or image like a king may be said to be interesting but not
glaring. After all, it is natural that respectful behaviour should be
modelled on situations in human society; moreover, it can even
be objected that the Buddha <1could>1 have become a universal
monarch, but rejected temporal power for a career in which even
kings bowed down to him. But, most important of all, the con-
fusion perhaps began with the Buddha himself. In the canonical
text already quoted for evidence of relics and pilgrimages, the
Buddha tells his attendant A$nanda that he should be cremated
like a great king, and a stupa (Pali: <1thu$pa>1) should be erected for
him at a crossroads. "And whoever shall put there garland or
perfume or paint, or worship there (<1abhiva$dessanti>1) or calm their
minds, that will long conduce to their welfare and happiness.' He
goes on to say that he, the Samma$sambuddha, a Paccekasam-
buddha, any other <1arhat>1 (a <1Tatha$gata-sa$vaka>1),<s48>s or a righteous
emperor of the world is worthy of a stupa, because on recollecting
their excellence people will become tranquil and go to heaven
after death.<s49>s This reason for relic worship is exactly that which
I was given by villagers and have already recorded above. Whether
these passages are interpolations I cannot here discuss ; even if they

134

are, they are extremely ancient (before Buddhism came to Ceylon);
and for the Sinhalese the question does not arise.
  So far it is only the Buddha's monastic status which has been
compromised, and for this he himself sowed the seed. That he was
mortal is still not in doubt: when I asked about this at the <1su$visi
pinkama>1 a monk told me firmly that the whole performance was
merely out of respect (<1gauravayen>1) for the Buddha's memory.
   There are, however, two further ceremonies involving a Buddha
image which cannot be explained in the same way--at least, when
I questioned the monks present at the ceremonies no one attempted
a rationalization, but explained them as traditions of no religious
significance. These two ceremonies also appear to be less ancient.
We have seen that chapter XXXIV of the <1Maha$vam*sa>1 mentions
every <1pu$ja$>1 with which I have so far dealt<s50>s (with the exception of
the food offering mentioned by Buddhaghosa slightly later), and a
few more besides. The next ceremony concerns the anointing of an
image, and I know of no references to it in early literature.
However, the <1Cu$lavam*sa>1 (chapter XXXIX) tells how a general
under King Kassapa I (<1478-496>1) was refused permission to
perform an <1abhiseka>1 festival for a Buddha image (1, 7), but held
it under his successor. <1Abhiseka>1 means anointing: though it is the
same word as that used for royal inauguration, that meaning can
hardly fit here. Rahula<s51>s and Geiger<s52>s say they do not know what
ceremony is being referred to, and we shall never know for
certain. But it does seem possible that it was related to the
<1na$numura mangalya>1 which I shall now describe, as that word
means "Festival of Anointing'. The <1na$numura mangalya>1 is men-
tioned in the regulations of Parakkama Ba$hu IV (early fourteenth
century).<s53>s
   The <1na$numura mangalya>1 is performed only once a year,<s54>s at the

                                                                    135

Sinhalese New Year. The Sinhalese New Year centres on 14 April,
but various observances and celebrations stretch over about a
week. Astrologers declare a day, shortly before or after 14 April, on
which people are supposed to bathe for the New Year; and the
ritual has been extended to the Buddha image.<s55>s There is no
special auspicious moment (<1na%ka%t ve$la$va>1) at which the ceremony
has to be performed, but it must be over in time for the monks
to go and bathe in the river themselves, in the stretch reserved
for them, and return in time for the <1da$ne$>1 at 11.
  The ceremony I witnessed was at Mi$gala at about 8.30 a.m. on
10 April 1965, and was performed by the two adult monks in
residence at the time. At Mi$gala there are two <1viha$ragedara>1, one
dating from the eighteenth century. The incumbent explained the
ceremony as a relic of the days of the Sinhalese kings, now pre-
served in only a few old temples; the things done for the Buddha
image, he said, used to be done for the kings, who then ordered
that they be done for the Buddha likewise. The crucial difference
in performance was that the Buddha was only anointed on his
mirror image. Normally no laymen witnessed this ceremony, but
I was allowed to do so as a favour. To this and the use of the
mirror I shall return below.
  In the <1pansala>1 the unguents (<1na$nu>1) are prepared on a tray:
saffron bark (<1kokum potta>1), margosa leaves (<1kohom#ba kol*a>1), tur-
meric (<1amu kaha>1), sesamum (<1tala>1), green gram (<1mu$n a%t*a>1), punac
(<1mimura>1), coco-nut (<1pol>1), red and white sandalwood (<1rat>1 and <1sudu
han#dun>1), and "all kinds of sour things' (<1okkoma a%m#bul ja$ti>1): limes
(<1dehi>1), sour oranges (<1a$m#bul dod*am>1) and tangerines (<1mas na$ram>1).
The pieces of red and white sandal, from which a little has been
shaved, are said to date from the Sinhalese kings. Each of the
ingredients is mixed with oil and made into a paste; then a little of
each paste is taken and mixed into one paste in the centre of the
tray.
  There is a small <1peraha$ra>1 up to the temple, with the usual
drummers (<1bera>1 and <1horan*a%$va>1). Someone carries on his head under
a canopy not only the unguents but also a mirror, a kettle of water,
a fly whisk (<1ca$mara>1), a small bell, and a set of the eight requisites

136

in a package. At the temple an oil lamp is lit on either side of the
door, and the junior monk sweeps out the temple. Someone blows
a conch shell many times, and the drums and fife play. The shrine
room is shut, with only the two monks and me inside it, and for
light a single candle is placed by the head of the main Buddha
image, which is a recumbent one. All is now ready, and the chief
monk shouts to the drummer through the door "Strike up'
("Gaha$pan'). The two monks then kneel before the Buddha image
and worship, bending their heads to the floor three times. They
are very serious. The chief monk then wedges the mirror on the
pedestal by the Buddha's head, so that it reflects the head at an
angle, and uses a flower head left there in offering to smear paste
onto the mirror. The junior monk pours water over the mirror
from the kettle, and the chief monk removes the paste with the
same flower. The same is repeated for a smaller Buddha statue
in the same shrine, a seated image. The chief monk then offers the
robes to the lying Buddha, and in turn waves the fan and the yak's
tail before him and rings the little bell. After each action he bows
deeply with folded hands. Finally both monks kneel and mutter a
few words beneath their breath. Exactly the same sequence of
events, minus the initial worshipping, was repeated at the other
<1viha$ragedara>1. The chief monk, who had a skin complaint, preferred
to bathe in hot water and did not proceed to the river, but the
junior monk went with two novices and many male hangers-on to
the river, where before bathing they all anointed their heads with
a mixture of sour stuffs from a brass urn.
  This simple ceremony combines elements which can be said to
correspond to at least three different conceptions of the Buddha.
The analogy with a secular king was sufficiently explained by the
chief monk and I need only add a reminder that a yak's tail is a
traditional Indian accoutrement of royalty, and the waving of a
fan and ringing of a small musical bell (in addition to the usual
drumming) likewise were the actions of royal attendants per-
forming <1te$va$va>1.
  A different conception is reflected in the action of offering the
eight requisites. It may at first seem implausible to argue that
food laid before a statue argues a different attitude from clothing
similarly offered, and certainly the two can be seen as a con-
tinuum; nevertheless there are certain contrasts which I wish to
stress. First, I must repeat that the former (offering food) was

                                                              137

justified by the monks themselves in canonical terms, whereas the
latter (offering robes) was not, so that there is a clear distinction for
the participants themselves. Secondly, the offering of food has
been shown to stem from an ancient tradition, whereas I know of
no ancient authority for offering robes to a relic of any kind.
Finally, food may be offered to anyone, but the requisites can
only be offered to monks and novices (<1pa%vidi>1). At first sight this
looks orthodox: did we not say that the Buddha was a monk? But
there is something odd about a monk's offering the requisites,
because while occasions can and do arise (as when a junior monk
waits upon the others at meals) when a monk offers food (and the
monk who offers food before the Buddha can be said to act on this
analogy), on no occasion does a monk offer requisites--they are
always offered by laymen. The requisites, in short, are always
offered to someone of distinctly higher religious status. Therefore
when the chief monk offered the requisites before the Buddha
image he was behaving not merely as if the Buddha were a monk,
but as if he were a person on another religious plane. While the
respect shown in making the offerings first discussed (from flowers
to food) was shown to be modelled on that shown by laymen to
monks, here it is monks who are treating the Buddha as a super-
monk.
  The Buddha has been treated as if he were a king and/or a
super-monk. Whether he has been treated as alive in the statue<s56>s
is on the evidence so far still ambiguous; but it is this question to
which we must now turn. One detail which seems incongruous with
the claim that the Buddha was a mere mortal is the case of the
mirror. Kings have yaks' tails waved before them, and monks are
offered robes, but no human, dead or alive, has his face anointed
in its reflection only. A possible interpretation of this use might be
that it arose merely from a veneration so great that the monk
considered himself unworthy to touch even a representation of the
Buddha's face. This, however, becomes implausible when we find

138

that a monk has no general reluctance to touch representations of
the Buddha. A far stronger interpretation emerges when we take
into account another ceremony in which a mirror is used, my final
example.
  However, before leaving the <1na$numura mangalya>1 I must mention
a less tangible point: the atmosphere of secrecy surrounding the
main part of the ceremony. Esotericism and secrecy are wholly
untypical of Therava$da Buddhism. Access to the sacred is free for
all. A Hindu deity is tucked away in his shrine, usually visible
only to the attendant priest, and in many Hindu temples permitted
proximity is graded by the caste and ritual purity of the wor-
shipper. The contrast between Buddhist and Hindu practices in
Ceylon is neatly exemplified by the use of the curtain before the
image, to which I have already referred. The shut doors of the
shrine--which we shall also find again in the next ceremony--
suggest not merely Hindu forms but also Hindu ideology: the
image has become a god--only affectively, of course.
  My final example of a ceremony involving a Buddha image is
the <1ne$tra pinkama>1, or "eye festival'; the final touch, which com-
pletes an image, is to paint in its eyes. I have described this
ceremony at length elsewhere,<s57>s so I shall restrict myself to the
barest recapitulation of my conclusions. Certain elements of the
ceremony made me agree with Knox's interpretation (quoted
above on p. 114) that, "The Eyes being formed, it is thenceforward
a God.' That adding the eyes should be the action which conse-
crates the image is itself aignificant; the features of a mere image
could presumably be finished in any order. The ceremony is
undertaken by the specialist craftsmen who made the image as the
culmination of long and elaborate ritual. It seems very dangerous,
for it can only be undertaken at an extremely auspicious moment.
The craftsman and his assistant shut themselves into the shrine
and everyone is warned to keep clear even of the outer door. A
mirror is again used; the craftsman does not paint the mirror
which would be ineffective, but he looks only at the mirror while
painting the statue's eyes sideways or over his shoulder. After-
wards he is led out blindfold, and the covering is removed when
his eyes will first fall on something, such as water, which he can
then destroy symbolically with a sword-stroke.

                                                              139

What the craftsman fears is obviously the gaze of the image,
which is conceived as being so powerful that it cannot be borne
direct; everyone else must put closed doors and as much space
as possible between themselves and the gaze, while he who has
seen it come into being, even reflected in a mirror, has dangerous
potential in his own gaze till the evil has been transferred and
killed. The concept of the dangerous look or gaze (<1ba%lma>1) is com-
mon currency among the Sinhalese, and of course in many other
cultures : for instance illnesses (notably psychological troubles) for
which medical diagnosis is not accepted are usually ascribed to the
gaze of some evil spirit (<1yaka$>1). Indeed, the <1ne$tra pinkama>1 contains
elements common to most Sinhalese exorcism.
That the circumstances of a <1ne$tra pinkama>1 imply the attribution
of life to the image cannot be and was not denied; the monks pooh-
poohed the whole thing as a meaningless (but harmless) tradition.
In the terms of my analysis, the ceremony treats the Buddha as a
god. But what is especially striking is that his gaze is treated as
dangerous, potentially maleficent. It is not an excess of holiness
which the craftsman fears from the new-born gaze. None of the
other ceremonies contained this element of fear and danger. With
knowledge of the <1ne$tra pinkama>1 it is possible to attribute the use
of the mirror in the <1na$numuru mangalya>1 to a similar fear, though
the emotion may no longer be felt. (I mentioned that in fact the
whole ceremony is almost obsolete.) Certainly the monk did not
conspicuously avoid the gaze of the image's eyes, and he was only
grave where the craftsman seemed positively scared. But the
totally undoctrinal danger of the <1ne$tra pinkama>1 is clear and un-
ambiguous. Moreover, the introduction of the <1ba$lma>1 makes it
impossible to claim that the image is only being venerated for the
relic it contains.
After stating the doctrinal position that the Buddha was a monk
who is now dead, I have shown that certain ceremonies performed
before his images imply othervise. Though nearly all the cere-
monies can be explained, as they are, as gestures of respect for his
memory, certain others can not, and in turn imply minimally that
he was a king, that he was (or is) a monk above other monks, that
he is alive in the image, and that his living presence is potentially
maleficent. The hierarchic order of these assumptions could be
slightly varied, for instance by claiming that it is implied that he is
a living rather than a dead king and that this is further from

140

doctrine than the implication that he is a live super-monk. This is
not important if all the implications listed are granted; the most
interesting result, which seems to me incontestably proven by the
last two examples quoted, is that certain ceremonies only make
sense on the assumption that the statue is alive, which I take to
amount to calling the statue a god.
   It is, of course, open to the outside interpreter of these rituals to
work in the opposite direction, and to claim from the evidence of
the <1ne$tra pinkama>1, etc., that the Buddha always appears to his
worshippers as a god. This is in effect what has usually been done,
though interpreters usually stop short at extrapolating from the
<1ne$tra pinkama>1 the danger of the Buddha's gaze (of which they
may not be aware) when they declare the Buddha's image to be
considered a living idol but admittedly benevolent. I have sug-
gested that this is too simple a view. I prefer to see in each cere-
mony the minimal supernaturalism, and hence the minimal
deviation from the canonical doctrine, that the facts will bear,
because this approach pays due attention to the views of the
participants themselves. Moreover, it allows me to preserve the
distinction between what people say they believe and what they
behave as if they believe--what I called cognitive and affective
beliefs. In this context I must stress that it does not follow at all
that because an image is treated as alive in one ritual it is generally
so viewed: many villagers are not even aware of the proceedings
at a <1na$numura mangalya>1 or <1ne$tra pinkama>1, and those who are may,
like the monks, explicitly repudiate their implications and declare
their ceremonies meaningless.
  The importance of context clearly emerges when we investigate
a final piece of evidence. In all the complex evidence adduced from
<1Buddha puja$>1 there is only one statement in words to imply that
the Buddha is still a living presence. There is, however, a widely
known Pali couplet which runs as follows:

           <1Ka$yena va$ca$ cittema pama$dena maya$ katam*
           Accayam* khama me bhante bhu$ripan+n+a Tatha$gata>1.

           Forgive me my transgression committed through
           carelessness by body, word or thought, O Tatha$-
           gata of great wisdom.

   This verse is quite unlike that used in the <1Buddha pu$ja$>1 in that
it is no mere formula; in particular, the Buddha is here addressed

                                                                     141

not only as <1bhante>1, which can apply to any monk, but as <1Tatha$gata>1,
which applies only to him. Apparently he is being asked to forgive
sins, just like the Christian God.  This would contradict two
cardinal points of doctrine: that the Buddha is dead, and that sins
can never be cancelled by forgiveness.
   At first sight one suspects Christian influence. But it turns out
that this is of a piece with the other Pali verses cited earlier; like
them, it originated in the context of the daily ritual at the Temple
of the Tooth, of which indeed it is the final verse. The monk
performing the ritual addresses the Buddha image, and asks to
be forgiven if he has made any mistake in the ritual. The mistake
need not be mechanical : impure thoughts too require forgiveness.
   There can be no doubt that in its original context this verse
involves addressing the Buddha image as if it contained or
embodied the living presence of the Buddha. The question is
rather how it is generally used and understood. Inquiries among
the village laity suggest that there the verse is indeed confined to
the ritual context: it is recited at the end of <1Buddha pu$ja$>1, either
in the temple or at home, and taken to refer to the period of the
preceding <1pu$ja$>1 only. But in the middle classes--and possibly else-
where--the verse seems to be of more general application: it acts
as a request for pardon for wrong-doing, as an attempt to elude
the laws of moral causality. Middle-class people are no doubt
aware, if the matter is brought into discussion, that this is con-
trary to doctrine, and hence attempt the sort of rationalization I
was given by the monks I questioned: that of course the Buddha
is dead and can <1not>1 forgive sins, but this is just a way of saying one
is sorry, a device to purify one's own thoughts. This is the same
as the explanation, quoted earlier, which monks gave me for the
verses that accompany offerings; in this case I have ventured to
call it a rationalization because to me it carries no conviction.
  Village laity, however, are generally simply inconsistent.  If
pressed they reveal that although the Buddha is dead there is a
certain Buddha force (<1Budubalaya>1), which will exist till at the
end of this dispensation, in another 2,500 years, the Buddha's
relics (including images) will collect at the Ruvanva%lisa%$ya at
Anura$dhapura and after a last sermon the Buddha will finally dis-
appear. (For the details and origin of this belief see the last part of
Chapter 7.) But the inconsistency is interestingly limited, because
this <1Budubalaya>1 does not seem to operate outside the ritual

142

context as part of the world at large, but only in the presence of
images; it is virtually a force inherent in the image. To say that
this force is put in by the craftsman who completes an image by
painting in the eyes would be a logical final step, but we cannot
take it because the Sinhalese do not take it. Things are just not so
tidy.
  Our general concluaion must therefore be that cognitively the
Buddha is dead and without further power, but that even this
cognitive position is not consistently maintained in the ritual
context; we may, however, suspect, especially after taking into
account some of the material in the next two chapters, that
affectively the Buddha is felt still to be potent, even when an
image is not present.
  After all, the very phrase <1Buddham* saran*am* gaccha$mi>1 --"I go to
the Buddha for refuge'--has a theistic tinge, however it may be
interpreted by pundits, and any inspection of Sinhalese religious
literature or behaviour suggests that few Sinhalese Buddhists
would object--though they would probably not originate such
words--to the sentiment of the great seventh-century Indian
Maha$ya$nist poet, S/a$ntideva:

          <1Ati$tya yusmadvacanam* sa$mpratam* bhayadars/ana$t
          s/uran*am* ya$mi vo bhi$to bhayam* na$s/ayata drutam>1.<s58>s

          I transgressed your command; now, on seeing my
          danger, in fright I go to you for refuge: quickly
          destroy my fear.

  I hope in the above discussion to have illustrated some of the
principles stated in the first part of my Introduction. I have been
concerned with several logically distinct but contextually related
problems raised by worship of the Buddha. The only ones which
I have attempted to answer in full, are what the Sinhalese say
about the Buddha, how they behave towards him (on which I
shall have more to say), and how their behaviour compares with
their professions. The answers to the first two are descriptive
(although all description of course involves interpretation), to the
last interpretative. I have also touched on the historical problem
of the evolution of the ceremonies among Buddhists and the
question, also primarily historical but doomed to remain largely
matter for conjecture, of what has been borrowed from Hindu

                                                             143

neighbours. The subject holds other problems, of course, includ-
ing ones which I have not even noticed. But if the Buddha is
generally said to be dead, this raises in the western mind many
questions about the Sinhalese universe and its laws, especially
about places filled and functions performed by the Christian
God. To answer some of these general questions, and to provide
a background for others more strictly doctrinal (to a Buddhist
mind), will be the purpose of the next chapter.

                               4

                 A SKETCH OF THE UNIVERSE AS SEEN
                          FROM MIGALA

                         I. CAUSATION

THE Christian God is the First Cause and the Final Judge. The
Buddha showed little interest in eschatological questions, which
he pronounced irrelevant to man's problems, and the Sinhalese
have inherited his attitude. The universe, according to general
Indian belief, goes through cycles of evolution and devolution;
though our physical environment is periodically destroyed it
comes into being once more. The Buddhists see no beginning or
end to this process, so have no need of a first cause, nor of any
cause to bring the world to a halt. Nor can there be any question
of an omnipotent judge; for while the universe lasts, so long
results, pleasant and painful, follow from the inexorable law of
<1karma>1.
The law of <1karma>1 would fill a central position in any account of
Buddhism (or any other religion) which attempted to treat their
problems as counterparts of those of the great monotheistic
religions, for it replaces the omnipotence and omni benevolence
attributed to the Christian God, and resolves that conflict between
them technically known as the problem of theodicy. Problems and
modifications in the operation of <1karma>1, which fall more properly
in the sphere of Buddhist doctrine and my subject matter, will be
discussed in the next chapter ; but here it concerns us more gener-
ally, as the foundation of the Sinhalese view of causation and the
way in which the universe functions.
The traditional Sinhalese view of the universe, what we might
call their cognitive map of the world, does not have a concept
corresponding to our "laws of nature'.<s1>s Things operate according

                                                                145

to the way they are, their "own essence' (<1svabha$vaya>1); of course
this usually leads to nothing unexpected; if a man walks into a wall
he will stop with a bump. But if he practises meditation success-
fully he may be able to go through the wall. The regularities of the
physical world are not "laws', and may on occasion be evaded.
   There is, however, one universal law of which everyone is
aware, and that is the law of <1karma>1. This law applies to all living
beings, always. <1karma>1 means "deed' or "act'. Strictly speaking your
<1karma>1 is what you do, not what is done to you--that is the matura-
tion (<1vipa$ka>1) of your <1karma>1, the result of one of your previous acts
in this or one of your former lives. <1Karma>1 is a doctrine of free will.
Indeed, will is paramount: the doctrine is that what counts is the
intention, not the effect. "It is intention (<1cetana$>1) that I call <1karma>1,'
said the Buddha.<s2>s
  For all the inconsistencies which will be probed in Chapters 5
and 6, everyone knows and broadly subscribes to this doctrine,
even if perhaps they cannot formulate it. The doctrine is logically
very far from fatalism; indeed it might be said to go to the other
extreme. On one level the imprecision of linguistic usage is mis-
leading. When a villager meets with misfortune he is liable to
say, "It's <1karma'>1 ("Karume$'), by which he means--and will, I
think, say so if questioned--that the misfortune is the <1result>1 of
his former bad actions. To this extent "it can't be helped', but
that does not mean that it is not his fault, or that he is not respons-
ible. Determinism is a heresy (<1mithya$ dr*s*tiya>1).<s3>s Moreover, <1karma>1
is not necessarily responsible even for everything that happens to
you: it operates, if I may so put it, in a gross way. One monk said
that <1karma>1 determines the station in which you are born, and your
luck (<1va$sana$va>1); after that, it is up to you--the pre
"Luck is one's former habits' ("Va$sana$va kiyanne pera purudda'),
said another monk. The idiom is as in English: "luck' usually
means "good luck'. If he has good fortune, a villager will probably
say, "Va$sana$va'; if bad, "Karume$', but he knows that his <1va$sana$va>1
is due to his <1karma>1.
  But the response to misfortune "Karume$' does in fact have a
fatalistic ring, which suggests that on the affective level the

146

villager may not accept the total responsibility which he admits
cognitively. In practice the distinction between these two logically
so different positions may be a fine one: a thief is not allowed by
doctrine to claim that he is stealing because of his <1karma>1, but he
may--and probably does--claim that it is his <1karma>1 which has
caused his rebirth as a man who has to steal to keep alive. In fact
villagers do often say "We are sinners' (<1api pavka$rayo$>1), not in any
mood of contrition or of defiance but simply meaning, "Look, we
are so poor that we have to sin to keep alive', and probably
further, "and we are born poor because we sinned before'. More-
over ~~als too--to say nothing of lowly spirits--are as they are
because of former misdeeds; now they have little chance to do
good, but must live by killing others, etc.; this bad <1karma>1 can set
up a vicious circle, only to be broken by the greatest willpower.
In this light the world of the Poor villager seems a pathetic
spectacle.
  Theodicy (popularly known as "the problem of evil') is in the
first place a logical problem: if God is Love, and can do anything,
how is it logically possible that he allows suffering? In Buddhism
the <1karma>1 theory disposes of this problem on the cognitive level:
suffering is a punishment for sin committed either in this or in a
former life. There exists, however, no general consolation for
undeserved suffering, so that psychologically <1karma>1 does not solve
the problem; in fact, by making a man responsible for even his
"luck' it may be said to aggravate it. In other words, <1karma>1
resolves the problem of theodicy on the cognitive, but not on the
affective level.
   One of the affective legacies of the <1karma>1 theory is an uncertainty
very similar to that felt by Calvinists anxious to know whether they
were predestined to salvation or damnation. The Calvinists inter-
preted worldly success as a sign of divine favour and so worked
hard to achieve it; the Buddhists--rather more logically, if I may
say so--tend to consult astrologers. You cannot avoid the results
of bad <1karma>1, but you can at least be prepared for it. A man's
luck, which is really his current stock of <1karma>1, is largely reified
in his horoscope (<1ke$ndraya>1): if his <1karma>1 is good he will be born at
a moment when the stars are favourable. According to a Mi$gala
astrologer, everyone has three <1ke$ndra>1, based on his sign of the
zodiac (<1lagna>1), the sun (<1ravi>1), and the moon (<1candra>1), but the
first of these is the most important. The planets (<1grahayo$>1) are

                                                                      147

pan-Indian:  <1ravi>1 (sun), <1candra>1 (moon), <1kuja>1<s4>s (Mars), <1budha
guru>1 (Jupiter), <1s/ukra>1 (Venus), <1s/ani>1 (Saturn), <1ra$hu>1 and
<1ke$tu>1. (The last two are the head and tail of a mythical demon who
is supposed to cause eclipses.) They may be in any of the twelve
lunar mansions (<1na%ka%t>1--the word is also used to denote the stars
in astrology generally), and the horoscope is drawn up like

this.          Lunar  mansion  number one  is top centre and

they are numbered consecutively anti-clockwise. I believe that
mansions 1, 4, 7, and 10 are the most important, and if they are all
empty the outlook is very bad. The horoscope indicates the out-
for material prosperity, and if it is very poor a solution for
Tales is to make them monks, in which condition they are sup-
posed to have no stake in good fortune. So at least I was told by
laymen, though monks usually denied it. Of course I was not so
impolite as to ask a monk outright whether he had a bad horoscope ;
their denial came by implication because they would say that a
monk needed a "suitable horoscope' (<1sudusu ke$ndrayak>1); many said
they did not know in what this consisted, but one specified that it
should have Saturn in the ascendant (<1senasuruyo$ge>1), while another
added to this that in a monk's horoscope (<1pa%vidiyo$ga>1) five planets
should converge in mansions 1, 4, 7, and 10. Altogether nineteen
monks out of twenty-one asked stated that a suitable horoscope
was needed to become a monk, although a couple added that
exceptions could be made; and the general implication was that
the suitable horoscope was a favourable one. However, a contrary
case gives some confirmation to lay gossip: one monk told me,
with disarming frankness, that when he was 22 years old his horo-
scope was inspected as he was about to marry, and it was found
to be such that instead of marrying he hastily entered the Order.
  Horoscopes are always matched in arranging weddings to make
sure that the couple are suitable; but otherwise villagers rarely
consult them, as the above story illustrates, and often they even
lose them. Another branch of astrological practice, besides casting
horoscopes, is determining auspicious (<1subha>1) days and moments;
almanacs listing these are widely diffused, but auspiciousness may

148

be general (for everybody) or particular (for oneself) and the latter
can only be determined by consulting an astrologer. Although Mrs.
Bandaranaike was plausibly rumoured in December 1964 to be
awaiting an auspicious moment for the dissolution of Parliament,
villagers have little interest in these mornents and observe them,
if at all, with very little precision. Altogether villagers tend to
pay less attention to astrology than do members of the middle
and upper classes. Partly no doubt this is because astrology--
like magic--caters for the emotional insecurity of those who have
left behind the ascribed status of traditional village life for the
competitive strains of a more "open' society. Dr. Obeyesekere,
however, suggests a more specific explanation: that here again we
encounter the anxiety caused by the doctrine of <1karma>1. The more
conscious a man is of the doctrine, the more anxious he becomes
to know what the future holds in store for him; the more
Buddhism, the more astrology. I think this an interesting and
plausible hypothesis.<s5>s Traditionally astrology would predict only
the events of this life; yet I came across the case of an elderly
professional man who, while he was turning to religion, had
astrologers predict for him his next rebirth.  His doctrinally
induced anxiety must have been unusually great. Dr. Obeyesekere
further substantiates his hypothesis by a comparison of villages
he knows; one in remote parts, with little knowledge of Buddhist
doctrine, has correspondingly little interest in astrology, and many
of the villagers never even had horoscopes. I made the same obser-
vation: middle-class people in and around Mi$gala were punctilious
in observing auspicious practices at the New Year to which
villagers paid little or no attention.<s6>s But on the personal level
among the villagers I may have detected a contrary tendency:
those who put most trust in Buddhism were least inclined to astro-
logy. This I would explain by canonical tradition: the Buddha
himself condemned astrology, palmistry, and all similar practices,

                                                               149

though his condemnation was specifically directed against their
practice by monks: he did not deny their possible validity, but
declared them a distraction from the road to salvation. This attitude
was exemplified by the monk, devout and well versed in the
scriptures, who told me that astrology is not lies (<1na$ka%ts/a$stre$ boru
na%$>1), but to become a monk no horoscope is necessary, only faith
(<1ke$ndara o$na%$ na%$, sraddha$va>1). But another monk who struck me as
being unusually devout took the opposite view : one needs a suitable
horoscope to become a monk, and a novice who is found to have
an unsuitable one must leave, because his religious conduct would
turn out poorly; moreover one should be ordained at an auspicious
time. This I would judge a flagrantly uncanonical position in its
implied assimilation of spiritual to temporal success. The be-
haviour of the monk who took a correspondence course in astrology
and has a government certificate to prove that he passed the
final examination also strikes me as uncanonical, even though
he told me that the study of astrology is traditional in his line of
pupillary succession. A few other monks said they knew some
astrology but only one said he practised; several said it was wrong
for a monk to do so. However, astrology was till recently taught
at the primarily monastic university (<1piriven*a>1) of Vidyodaya.
  A sceptical western reader will not be surprised to learn that
even the predictions of astrology are usually imprecise; a villager
who wants a specific piece of information will probably consult
another specialist who employs other means (see below). What
astrology generally predicts is an <1apale$>1,<s7>s an unlucky period, which
may be either general or individual, just as a day may be lucky
either for everyone or for an individual in accordance with his
horoscope. There was an <1apale$>1 all over the Indian cultural area in
February 1962 when there was a conjunction of planets so dire
that Indian predictions of the end of the world reached western
newspapers. In Ceylon there was widespread ritual chanting of
Buddhist texts (<1pirit>1, discussed below), and in Mi$gala temple
thirty monks chanted (in rotation) for a week, ending with a
colourful ritual (the <1dorakad*asne$>1). One of the monks described it
to me with great enthuaiasm and said the final ritual was very fine
(<1hari lassanayi>1); there was a photo of the participants hanging in
the <1pansala>1, and everyone seems to have had a lot of fun. Equally
cheerful was a young villager who told me his horoscope showed

150

he had an <1apale$>1 of fifteen years, of which he had about eight
years to go. In the end, he said, things would turn out OK. If he
did <1pin>1 he would be reborn in a better state. On the other hand a
middle-class village for whom an <1apale$>1 was diagnosed immediately
had a specialist come and perform white magic, and stayed home
from his job when the astrologer so advised.
  Bad <1karma>1 means that one is due for misfortune. An <1apale$>1 is not
a specific misfortune, but a period of being prone to misfortune.
Specific misfortunes are caused by other beings--gods, men, or
devils--who operate as freewill agents, or they may result from
natural causes such as eating the wrong food. These causes inter-
lock and cannot be rigidly schematized. A man who falls ill will
probably first try western medicine at the local hospital, and if
that fails try Ayurvedic ("Sinhalese') medicine administered by a
village specialist. If that fails his next resort will be determined by
his sub-culture and individual temperament. He may ascribe it to
human agency (black magic), and employ suitable counter-
measures (white magic). He may ascribe it to demons of disease
or to malign planetary influences (i.e. simply the influence of the
<1apale$>1) personified as planetary divinities, and banish or appease
them by more or less elaborate exorcistic ceremonies. He may
ascribe it to the actions of a god, or rather the failure of that god's
protection, and make the god a vow, promising him some present
or service if he recovers. If the remedy does not work this may
be due to a wrong diagnosis, or, much more likely, it may be
because the man's <1karma>1 is too bad, and he is due to suffer longer.
  The theories and remedies listed are not mutually exclusive : for
instance, a man may make a vow to a god at the time when he
enters hospital, or think that the failure of a god to offer protection
has let a malign demon do his work. What they have in common is
that none of them have anything to do with religion as the Sinha-
lese understands the term. The western observer who makes cross-
cultural classifications could find in the above list first science,
then magics and finally religion; but to the Sinhalese this has no
relevance: all are means of combating worldly misfortune and
therefore are worldly and <1ipso facto>1 not religion. Nor are the beliefs
on which these practices are founded considered religious beliefs :

                                                                    151

they form no part of the Buddha's message, but are just facts of
life. From this point of view there is no difference between gods
and demons, between the higher and the lower supernaturals,
although Dr. Obeyesekere has made the interesting point that it
is easier for a Buddhist to deny the existence of gods than of
demons, as the activities of demons--principally inflicting disease
--are the more obvious. In fact the only person I found positively
to deny the existence of gods was a lay villager, though some
monks were sceptical about their powers. However, the Sinhalese
are aware that other religions set store by gods, so that when I asked
monks about gods a couple of them made a point of stressing that
their gods have nothing to do with Buddhism. Behef in gods is
not contrary to Buddhism, but has no connection with it; belief
in gods is not a matter of religion (<1a$gama vas/ayen adahanne na%$>1);
Buddhists as such have no devotion (<1bhaktiya>1) to gods and make
no demands (<1illi$ma>1) on them; to worship just gods (<1deviyanma
a%dahi$ma>1) is heretical (<1mithya$ dr*s*t*iya>1). It is crucial to the Buddhist
theory of monasticism (<1mahan*akama>1) that a monk should not be
concerned with worldly goods and ills; but it is only worldly goods
and ills that gods and devils (whom for convenience I shall call
supernaturals, though the concept is un-Sinhalese) have it in their
power to bestow or remedy. Accordingly, while theory has
nothing to say against the existence of these supernaturals, it does
not become a monk worthy of the name to make any use of them
or their institutions; to seek worldly gain is improper, and so it is
improper to require the favour of a divine or a human potentate.
Both are sometimes solicited, from an ignorance which does not
know of this impropriety or a greed which chooses to ignore it;
but supernaturals are far less often solicited than men, whether
because such intercession with a god or demon, requiring as it
does a lower-class human intermediary, is a more conspicuous
lapse, or because efforts at propitiating human favour are regarded
as the more fruitfully expended, we can only conjecture. I do not
know how many of the monks I interviewed had at some time or
other been involved in civil litigation over temple property, but
the number was at least double the two who said that monks do
occasionally make vows (<1ba$ra venava$>1) to gods, of whom only one
admitted doing so himself; and a monk in a law-court is a far
commoner sight than a monk in a god's shrine (<1de$va$le$>1). To this
ideological reason why a monk may not placate gods or demons

152

the sociologist may conjecturally add another: that it is the very
function of the Buddhist clergy to preserve Buddhism, not merely
in the sense that they preserve the texts and are themselves coter-
minous in time with Buddhism, as they themselves admit, but also
in their keeping the doctrine distinct from the theocentric religious
tendencies of the laity. To preserve the Therava$da tradition the
Order need not deny the existence of gods, but it must keep
sufficiently aloof from theocentric practices to allow laymen a
vision of a life lived independent of supernatural favour.
  The distinction between propitiation of super-humans and
religious actions is clear-cut, for, however ambiguous the onto-
logical status of the Buddha, he cannot be dealt with like the gods
on a <1quidpro quo>1 basis, nor of course can he be compelled by spells
like a demon. In Buddhist theory, however, the distinction is only
incidentally one of means, namely that gods and demons (being
alive) can be reached and affected while the Buddha (being dead)
can not; essentially it is one of ends: gods and demons can be
asked for any worldly benefit, from the destruction of an enemy
to recovery from diseases; the Buddha is in fact asked to forgive
sins. But it is also possible to use religious instruments to further
certain limited worldly ends. (This raises no more problems of
definition than getting a priest to post a letter.) Buddhist texts are
used (and have perhaps always been used) for what at least to an
outsider looks like white magic, and it might even be argued,
although I would hesitate to agree, that in certain contexts the
Buddha himself is invoked for similar worldly ends, namely pro-
tection from sickness and danger.
  The use of Buddhist texts to gain protection and relief from
misfortune is generally formalized in a ceremony called <1pirit>1; the
text used in informal contexts which could be taken as an invoca-
tion to the Buddha is the "<1Itipi so' ga$tha$>1. Both will be discussed in
detail at the end of this chapter, where they have been placed at
the end of the section on white magic. I have included them in this
chapter rather than in the next one, which is on some problems of
doctrine, and deals with some other Buddhist practices which
might be described as magical, both because they seem behaviour-
ally more on a par with other practices for counteracting mis-
fortune and because they are doctrinally unproblematic, which
the <1pra$rthana$>1 discussed in Chapter 5, being concerned with future
lives and religious goals, are not. On the other hand I juxtapose

                                                                 153

<1pirit>1 and <1pra$rthana$>1 to draw attention to their similarity, exempli-
fied in the words of the <1Maha$ Jayamangala Ga$tha$>1. To elaborate
would be premature; I mention this rather complex argument
here merely to point out, in accordance with my principle of
paying attention to the sentiments of the Sinhalese themselves,
that their explanation of <1pirit>1, etc., does not exactly parallel their
explanation of the cases listed above, in which gods and devils
were propitiated directly and remedy was clearly matched with
cause. <1Prima facie>1 it might seem that the obvious explanation for
the recitation of sacred texts at a aick-bed would be to say that as
the situation is due to bad <1karma>1 the direct remedy is to acquire
better <1karma>1 by the performance of meritorious actions. But this
explanation is rarely if ever used, because it does not really work:
the doctrine of <1karma>1 requires that bad acts bear fruit, so that the
<1pin>1 of the sacred text will be rewarded in the future, but can
have no effect on the current punishment of <1pav>1. On the other
hand <1pirit>1, etc., can by more than one argument be made to
harmonize with the rationalist position of canonical orthodoxy.
I shall return to these interpretative problems at the end of the
chapter.

                II. THE CONTENTS OF THE UNIVERSE

The bulk of the remainder of this chapter must be devoted to a
descriptive account of the non-human inhabitants of the Sinhalese
Buddhist universe and the institutions directly connected with
them. Although this subject matter mainly<s9>s falls outside the local
definition of Buddhism, much of it is necessary for a wider under-
standing of the way in which the Buddhist sees the world, and in
particular for an appreciation of the ambiguous position of the
use of Buddhist texts just alluded to, to which I shall return at the
end of the chapter. For the most part I confine myself to the
minimum information necessary for this purpose, but at times I
shall supply some further data. I hope that these descriptive
digressions may not be felt to impede the flow of the argument,
but I think it right to attempt to complement the ethnographic

154

data collected by my predecessors in the field. There is plenty of
detailed literature on the Sinhalese pantheon, ritual and magic,<s10>s
but most of it depends<s11>s on material collected in the Low Country,
and in these matters there is much variety between the Low
Country and Kandyan sub-cultures, and even between smaller
areas.
  In treating this kind of material much use--and misuse--has
been made of Redfield's distinction between great and little
traditions. The great tradition in a community consists of the
common inheritance of the whole civilization to which it belongs,
and is transmitted principally by the learned (usually in books);
the little tradition consists of the other, local elements of a culture,
varying between communities, for the most part transmitted
orally and ignored in the works of litterati. Dr. Obeyesekere in his
article "The Great and Little Traditions in the Perspective of
Sinhalese Buddhism' has brilliantly demonstrated the utility and
the limitations of these concepts in the interpretation of field-work
material, using as his example the subject matter of this chapter.
He himself has drawn<s12>s a useful distinction between the ideal and
the actual pantheon, between supernaturals whom people merely
believe to exist and those with whom they interact. I think it is
possible usefully to relate this latter distinction to that of Redfield
and thus clearly to schematize the whole situation. The graph
paper on which this schema will be drawn is the Sinhalese map of
the physical universe.

                                                            155

   According to canonical Buddhist doctrine the world consists of
three spheres or areas (<1dha$tu>1), each with many named subdivisions.
The lowest sphere is that of desire (<1ka$madha$tu>1), the next up is the
sphere of form (<1ru$padha$tu>1), and the highest the sphere of non-
form (<1aru$padha$tu>1); they are inhabited respectively by creatures
with sensual desires, with no desires but formed bodies ("made of
mind'), and with no bodies. This categorization is scholastic, so
we need not further pursue such questions as whether the spheres
are literally on top of each other. The sphere of non-form is so far
out that no one knows anything about its formless inhabitants
except that they are very near attaining <1nirva$n*a>1. At the top of the
sphere of form are five heavens called "pure abodes' (<1suddha$va$sa>1);
as their inhabitants are immaterial they cannot be illustrated or
usefully described beyond saying that they are minds sunk in
meditation. The lower heavens sometimes are depicted on temple
ceilings; there are at the lower end of the sphere of form seven
Brahma$ heavens (<1brahmaloka>1 or <1bam#balova>1), and below them are
six divine heavens (<1divyaloka>1 or <1devlova>1) which constitute the top
segment of the sphere of desire. The inhabitants of these <1divyaloka>1
enjoy divine bliss (<1dev sa$pa vin#dimava$>1). These heavens have a
drawback: they are so pleasant that no one bothers to meditate
and make spiritual progress. Most of my informants agreed that
an inhabitant of one of these heavens, i.e. a god, cannot attain
<1nirva$na>1; that can only be done by a human being or an inhabitant
of one of the higher spheres. Nevertheless, these <1divyaloka>1 are the
heaven (<1svargaya>1) in which most people aspire to be reborn.
  The heavens are a place up in the sky; <1nirva$n*a>1 of course is not
a place but a state. A schoolgirl, 12 years old, said that <1nivan>1 is in
the sky above the heavens, and the stars are lamps which they light
in heaven, but adult villagers are more sophisticated, if less poetic.
But her statement that in <1nirva$n*a>1 one has no body, only thoughts,
is interesting in that it answers the canonical account of the higher
spheres and illustrates how these have resulted from attempts to
imagine rarefied worlds to correspond to the most sublime mystical
experiences. In fact one canonical formulation has it that there are
four stages or stories (<1bhu$mika$>1): sphere of desire, sphere of form,
sphere of non-form, and supra-mundane (<1ka$madha$tu, ru$padha$tu,
aru$padha$tu>1 and <1lokottara>1). "Supra-mundane' is of course <1nirva$na>1.
It is this highly specialized word, <1lokottara>1, that M. M. Ames
takes as a translation for "sacred' as opposed to "profane', which

156

illustrates the pitfalls in matching Buddhist with Durkheimian
concepts.
  The bottom stratum of the sphere of desire, and thus of the
universe, consists of various hells under the ground, hells hot
and cold in which demons torment those who must pay for
past sins. Black monsters are shown in temples inflicting on their
naked victims the most sadistic tortures. Often the punishment
fits the crime: in a special hell depicted in one temple those who
have killed hair-lice are crushed between huge boulders which
appear to them as human finger-tips must appear to a hapless
louse.
  In between the heavens and the hells, roughly in the middle of
the sphere of desire, is the human world (<1manus*yaloka>1) we know,
which includes the sun and moon and other planets in the sky, as
already mentioned, and a great variety of other non-human
beings, including gods [<1sic>1], animals, demons, and goblins.<s13>s
  This map of the universe taken from the Pali Canon has not,
and probably never had, any clear effect on religious behaviour, but
it illustrates a pattern which I see repeating itself. In the scheme
of the three spheres man's world is rather near the bottom of the
pile. However, when we focus on the world in which we as men
have any immediate or practical interest we ignore the top two
spheres and concern ourselves only with the sphere of desire, a
world of which man is the centre, with gods in heavens above him,
demons in hells below him, and other men, animals, goblins and
ghosts around him; a world moreover in which man is unique in
his ability to attain <1nirva$n*a>1. Even for the Buddhist in ancient times
there was a difference between what Obeyesekere calls the ideal

                                                               157

and the actual universe: some parts he just knew about from
teaching, while some were psychologically real because he felt he
Tight easily visit them himself. I suggest that this narrowing of focus
has repeated itself: the heavens and hells of the canonical sphere
of form are now psychologically more distant, and the super-
naturals with whom man has any contact, who constitute the ideal pat-
tern, are only those of the human world (<1manus*yaloka>1) in its centre.
   As with areas of the world, so with their inhabitants: no one
has ever met one of the disembodied denizens of the Pure Abodes,
but some gods and many ghosts are regular village visitors. This
was as true in ancient times as it is today. The non-human inhabi-
tants of the canonical universe have been catalogued by Father
Joseph Masson; a glance at his survey of what he calls (counter
to the tenor of this thesis) "popular religion' in the Pali Canon<s14>s
reveals even within the Canon a difference between an ideal and
an actual pantheon, between supernaturals who appear in formal
catalogues and those who figure in stories. For anyone who wishes
to "get the feel' of a religion this distinction is surely most impor-
tant. It is a distinction which cuts right across the question of
origins with which Redfield's dichotomy between great and little
traditions is properly concerned. When the relevant part of the
Canon was compiled<s15>s the great tradition was that of Vedic
Brahmanism, and fragments of the Vedic pantheon survive in the
Buddhist texts, though the only parts of it to occur in the actual
Buddhist pantheon have been transformed and "Buddhicized'. The
Vedas were, however, composed at least 500 years before the
Buddhist scriptures, much further west, by high-caste Aryans
(i.e.  speakers of Vedic Sanskrit).  Buddhism therefore incor-
porated material from the little tradition of a different time, a
different place, and people with a different ethnic, social and
linguistic background. Some of this too was obsolete or obsoles-
cent. To all this were added the innovations of the Buddhist
<1litterati>1, a new great tradition which made a rather successful
attempt to systematize the jumbled heritage.
  This situation finds its parallel today; if anything, the historical
constitution of the modern Sinhalese pantheon is the less

158

complicated. It is less complicated because there has been no struc-
tural innovation; the custodians of the Buddhist great tradition are
still the monks, and the systematized schema they offer is the very
one invented by their predecessors more than 2,000 years ago. It is
the schema which I have just introduced. It is ideal, not actual, but
so it always has been. The most important gods of the actual
canonical pantheon, Sakka and Maha$ Brahma$, were borrowed
from the Hindu great tradition, but changed in character and
station so as to appear suitably Buddhist; the most important gods
in the modern Sinhalese pantheon, notably Vishnu and Katara-
gama, have likewise come in through contact with Hindus, and
Vishnu at least has been Buddhicized by being made a Bodhis-
attva. (Kataragama is rather a special case, as we shall see below).
Sakka and Maha$ Brahma$ were assigned their heavens in which
to live; the universe now being partitioned there are no new
worlds for the new gods to inhabit, so they live in the human world
with us. (For gods living in the human world there are canonical
precedents). As for the demons and goblins, both then and now
they are mostly local individuals fitted into traditional categories.
  Although the modern Sinhalese have preserved the memory of
several other categories of canonical supernaturals (most of which
were probably ideal, not actual, even when the Canon was com-
posed), there are only three categories of supernaturals in the
actual pantheon: god (<1deyyo$>1 or <1deviyo$>1<s16>s--this plural form is used

                                                                159

also to denote the singular), devil (singular <1yaks*aya$>1 or <1yaka$>1,
plural <1yaks*ayo#>1 or <1yakku>1), and ghost (singular <1pre$taya$>1 or <1pere$taya$>1,
plural <1pretayo$>1 or <1pere$tayo$>1). Of these three classes the first two are
much the more important.  "Devil' and "ghost' are extremely
inaccurate translations, as I shall soon explain, so I shall refer
to these two categories by the Sanskrit terms <1yaks*a>1 and <1preta>1.
Added to men and animals this makes five classes of beings, and
so corresponds to the standard canonical list of five <1gati>1 (states in
which one can be reborn); in ascending order these are hell,
animals, <1preta>1, men, and gods.<s17>s
  The category of god, as conceived both now and in the Canon,
is well-defined and fairly straightforward: a meritorious person is
liable to be reborn in heaven, and the inhabitants of heaven are
all gods. The only problem arises in connection with the gods of
this world. In the canonical account of the Buddha's first sermon
it says that the gods of the earth (<1bhummadevata$>1) were delighted
with it and their applause attracted the attention of the gods in the
various heavens;<s18>s in the account of the Buddha's death the
Buddha tells A$nanda that there are gods of the earth with percep-
tions of the earth (<1santi devata$ pathaviya$ pathavi-san+n+iniyo>1), who
weep and mourn at his passing.<s19>s These gods of the earth may be
the minor spirits who proliferate in the Canon rather than gods of
the power and importance of Vishnu in Ceylon, but they do
provide a precedent for locating gods in our world rather than in
heaven: one monk told me that Vishnu, Kataragama, etc., live
mainly in this world; whether or not they go up to the heavens
(<1divyaloka>1) we do not know. The general opinion concerning
Kataragama, the most important god in modern Ceylon, was that
he was powerful but a sinner, and definitely an inhabitant of our
world. The opinions of one monk deserve to be reproduced in
detail. Concerning Vishnu he held the unusual theory (which he
told me was his special private theory) that he is a personification
of nature ("Visn*u kiyanne svabha$vaya'); this does not mean that it
is futile to pray to him--on the contrary, he is very powerful. The
other gods he divided into "big gods' (<1loku deyyo$>1) and "little gods'

160

(<1pod*i deyyo$>1).  (Compare the canonical distinction quoted by
another monk between gods of little and of great power (<1alpe$sakya
mahe#sakya>1).) The "big gods' he mentioned were the canonical gods
who survive in the Sinhalese ideal pantheon, namely S/akra and
the Four Guardian Deities, who live in their heavens; the little
gods, he explained, are merely humans who had developed minds
(<1hita diyun*uva>1) in life and have been deified. He applied this
to all the non-canonical gods except Vishnu. We can see the
process with Mr. S. W. R. D. Bandaranaike, who will soon be
<1Ridave$ D*ayas Ban*d*a$ra>1.<s20>s Similar cults await other men who
have little passion, hatred, and delusion (<1ra$ga dve$sa mo$ha adu
minissu>1). It would be invidious to reproduce the rest of our dis-
cussion.
  The above exposition suggests a rather mixed intellectual back-
ground. According to Buddhist doctrine all gods were men in a
former life, so it should not be necessary to specify this for any
particular gods. On the other hand the same Buddhist tradition
makes euhemerism a favourite explanation of gods among sceptical
Buddhists. The difference between the Buddhist and the euhemer-
istic position is that the former allows the gods to exist in their own
right, the latter claims that they exist only in the popular imagina-
tion. My informant was probably implying that the "big gods' of
the Canon really exist but the "little gods' do not, a position not
illogical but incongruous.
  The categories of <1yaks*a>1 and <1preta>1 are far less clear, and the
muddle goes back to the Canon itself. Both then and now the
category <1yaks*a>1 overlaps with that of god on the one hand and that
of <1preta>1 on the other. The former overlapping is a mere semantic
unclarity, the preservation (or repetition?) of which is a historical
curiosity; the latter reflects the more fundamental difficulty, met
with also among Christians, of reconciling doctrine on the fate of
men--especially bad men--after death with a half articulated
belief in their return as ghosts.
  The range of meanings of <1yakkha>1, the Pali equivalent of <1yaks*a>1
in the Canon, can best be studied in the Pali Text Society dictionary
entry for that word. The term covers a great variety of supernatural
beings, usually of a religious and moral status somewhere near
that of man. Two kinds of beings seem principally to be covered

                                                                   161

by the term: morally neutral or ambivalent nature deities who
live in trees or on mountains; and extremely malevolent and
hideous demons who frequent cemeteries and lure lonely travellers
to sudden death. The female of the former class is frequently por-
trayed, according to the general interpretation, in the lush female
figures at Bharhut and Sanchi; the female of the latter class is apt
to beguile the wayfarer or shipwrecked sailor by appearing as a
lovely woman, only to turn into a cannibal ogress. This latter,
demoniacal, class widely supplants in Buddhism the similar
<1ra$ks*asa>1 (Sinh.: <1rakusa>1) of Sanskrit literature. However, the word
<1yaks*a>1 has a wider meaning of "supernatural being', for it is even
applied to th e Buddha. I think the importance of this passage can
be over-stressed,<s21>s for it is in a long verse eulogy during which
the Buddha is also called <1Sakka purindada>1 ("S/akra sacker of
citadels'); but the great god S/akra is himself called a <1yakkha>1.<s22>s
At the other extreme one kind of <1preta>1, those with great power
(<1mahiddhika>1), are all <1yaks*as>1. For the middle ground let me adduce
the passage<s23>s in which <1yaks*as>1 are cited as a distinct group parallel
with three other kinds of demigods: <1gandhabba, kumbhand*a>1 and
<1na$ga>1.<s24>s Their king is the guardian of the north, Vais/ravan*a (Pali:
Vessavan*a <1alias>1 Vessamuni <1alias>1 Kubera, Sinh.: Vesamuni), who
is elsewhere said to be the god of wealth and live in the Himalayas.
He has survived till modern times. He lives in the lowest of the
<1divyaloka>1, but to judge from what I was told can properly be
called a <1yaksa>1 himself.
  The semantic spread of the term <1yaks*a>1 in modern Ceylon is
almost the same, except that it is never used of the Buddha or a
"big' god; it implies at least potential sinfulness and generally

162

residence in the human world.<s25>s There are native divinities who
live in big trees, especially <1Bo$>1 (sacred fig) or <1na$>1 (ironwood)
trees;<s26>s these are vaguely conceptualized and poorly integrated
with the rest of the pantheon, but the day before cutting down a
big tree the woodmen will inform the resident divinity by men-
tioning the fact as they pass by, and maybe decorate another
tree with lights and strips of cloth to indicate it is a suitable
alternative residence. (<1Bo$ and na$>1 trees are not cut down at all.)
The more important and better-known sort of <1yaks*a>1, however, is
distinctly malevolent, and in the Low Country is exorcized by
devil-dancing. These <1yaks*as>1 generally bring disease, and certain
diseases are even associated with particular <1yaks*as;>1 but in my area
these individually named <1yaks*as>1 have little currency, and mis-
fortune may be ascribed to any <1yaks*a>1 fairly indiscriminately. There
is a clear tendency for the named <1yaks*as>1, those with individual
personalities, to be assimilated to the important class of minor
deities called "the Twelve Gods' and known to scholarly litera-
ture as "the <1Ban*d*a$ra>1 cult'. The Twelve Gods will be discussed in
detail below; here suffice it to say that they are known collectively
as gods (<1deyyo$>1) and most of them include "god' or "deity' (<1de$vata$>1)
as part of their name, and yet most of them are freely described
as <1yakku>1. The expression may be pejorative, but this depends on
the speaker's expression and tone of voice; the term is not itself
loaded, but retains its canonical ambiguity. At the other extreme
are the devils who torment sinners in hell: for these too there is

                                                                       163

no word but <1yaka$>1. But whereas all the other <1yakku>1 are part of the
actual pantheon the devils in hell are only part of the ideal
universe.
   Already in the Canon the <1pretas>1 (Pali: <1peta>1) are an adaptation
of an older idea to the Buddhist conceptions of rebirth and
punishment, and seem to be poised on the borderline between the
actual and the ideal; they are still in exactly the same position. In
the Vedas there is no cycle of rebirth; the dead go down to some
kind of Homeric underworld, where they exist as ancestor spirits
or "fathers' (<1pitaras>1). These <1pitaras>1 were to receive offerings (<1bali>1)
from their male descendants, and this obligation persisted,
illogically, despite the new doctrine of rebirth first met in Upani-
shads. Buddhism resolved the illogicality but retained some incon-
gruity. A class of <1pretas>1 was invented, hungry ghosts who roamed
the dark corners of the earth just slightly better off than if they
were in hell, and liable to reform just like all other classes of
beings. <1Preta>1 and <1peta>1 mean "departed', the term being used to
mean "dead' as in English; but the term is linguistically confused
with the <1pitaras>1'<s27>s thus revealing the ambiguity of the concept.
In fact, as B. C. Law writes,<s28>s "with the Buddhists the <1Petas>1 are
dead kinsmen'. Another compromise with the older ideas was
achieved when in non-canonical Pali literature the <1pretas>1 were
divided into four<s29>s (or three)<s30>s categories of which one was
designated as "living on what others give' (<1paradattu@paji$vino>1); but
in this case what they receive is not a physical offering, but merit
given them by human beings, notably by relatives at the funeral
feast. These feasts and their underlying ideas will be discussed
in the context of merit transference at the end of Chapter 5. Here
it must suffice to say that the semantic overlap between <1pitaras>1
and <1pretas>1 is still a living reality, in this sense : although cognitively
--and logically--<1pretas>1 can be anyone's relations, the only <1pretas>1
of whom people usually think and with whom they interact are
their own dead relations.
  I mentioned above that the <1pretas>1 "of great power' are said in
the Canon to be <1yaksas>1. In modern Ceylon the categories <1yaksa>1
and <1preta>1 are never formally stated to overlap; but they seem

164

somewhat confused in matters concerning their manifestations and
treatment--in a word, when they are ghosts. Sinhalese has a word
appropriately translatable as "ghost'--<1holman>1. It is important to
note that a <1holman>1 is not a separate class of being; it is a mere
apparition, a manifestation (<1avata$raya>1) of a being.<s31>s (The word
<1avata$raya>1, which has a distinguished history in Hinduism, has
in Mi$gala so far come down in the world as to be used as a
<1symonym>1 for <1holman>1.) These <1holman>1 are much associated with
scenes of death, which leads the westerner to expect that they
might be manifestations of <1pere$tayo$>1, but in fact they are generally
said to be <1yakku. Pere$tayo$>1 are said to have huge bellies and very
small mouths (their uncouth appearance and insatiable appetite
perhaps justify their being called "goblins');<s32>s <1holman>1 on the other
hand never have this form, but appear typically as a naked white
figure or as a dog. They are, however, not necessarily seen; they
may be heard (as pursuing footsteps), or even felt (as a hand on
the back). They appear at night, and the hours of 6 p.m., 6 a.m-
(twilight) and midnight are said to be especially dangerous. The
most dangerous place is a cemetery--and here we are reminded
not only of ghosts but, perhaps more pertinently, of the necro-
phagous <1yaks*a>1 of the Canon. The worst such ghoul is Mahaso$na$,<s33>s
who may lay his great hand on the back of a lonely wanderer near
a cemetery at midnight. People dead of shock have been found to

                                                                     165

bear his livid imprint on their backs,<s34>s a phenomenon com-
parable to the bleeding stigmata of Padre Pio and other Christians.
Though I heard of no such extreme case in Mi$gala, the villagers
do not allow their womenfolk near the cemetery at night, and I
heard of one lady who had come across Mahaso$na$ and been
frightened into a fever. However, <1holman>1 do not necessarily do
you any harm; they can be kept at bay by reciting a Buddhist text
(the "<1Iti pi so' ga$tha$>1--see end of chapter) or a consciousness of
rectitude, and some seem to view them with more curiosity than
alarm, though such people, as one might expect, have usually not
met one--and some do not even believe that they are anything but
figments of the imagination.
  I am not an authority on <1pere$tayo$>1, because so far as I could dis-
cover the inhabitants of Mi$gala do not interact with them; outside
the context of the funeral feast I think they are part of the ideal
rather than the actual pantheon. It is in the Low Country that
they are actual: Obeyesekere writes,<s35>s "<1Pretas>1 ... are given inferior
foods like <1pulula>1 or marijuana and sometimes fecal matter'; but
I came across no such offering, and an informant wise in local
lore even refused to draw the Low Country distinction between
offerings to <1yaks*as (pide$ni>1) and offerings to <1pretas (dola>1), but said
that as both were <1amanus*ya>1 (a Sanskrit word literally meaning
"non-human') both received <1dola>1. Obeyesekere tells me that in the
<1Low>1 Country <1pretas>1 are quite frequently perceived as troublesome
ghosts, poltergeists, etc., and pacified or exorcized in a spirit of
loathing and contempt; psychologically this is very interesting,
for <1pretas>1 are one's own relations--typically parents. Thus one
may offer one's own father blood and pus to eat, or confine one's
own mother in a small box and dispose of it in a latrine.
  Another term for creatures in this murky area is <1bhu$la>1 (Sinh.:
<1bhu$taya$>1, pl. <1bhu$tayo$>1), which literally means "has-been'. In the
Canon the word is common only in the compound <1bhu$tavija$>1,
literally "science of <1bhu$tas>1', which is the normal word for exorcism.
The word is not commonly used in Mi$gala--though again I believe
it to be common in the Low Country--but I have had to explain
it in order to render intelligible the opinions on this subject of the
euhemeristic monk quoted above; his opinion on the lower spirits
will illustrate how the line between <1yaks*a>1 and <1preta>1 is blurred. At

166

first he said that <1yakku>1 are wicked men, such as the black cannibals
in Africa; the <1yakku>1 who were in Ceylon when the Buddha<s36>s
visited it were of this sort. This statement is ambiguous: it could
mean that wicked men are reborn as <1yakku>1, which would be
traditionally orthodox, or that they tend to be mistaken for <1yakku>1,
a rationalization on the euhemeristic lines of his remarks about
gods. Then, however, he added that some <1yakku>1 are also <1bhu$ta;
bhu$ta, pere$ta>1 and <1holman>1 are the same, though <1pere$ta>1 is really the
general name for a dead man. He then spoke of <1yakku>1 possessing
people: when the possessed person is exorcized the <1yaka$>1 shouts
"Hu$' and leaves by the mouth. <1Yakku>1 are very fond of drumming :
he told a story of a woman who while possessed could drum
marvellously, but lost the art when exorcized. There are many
<1yakku>1 in Colombo, but some everywhere--even in Buckingham
Palace there are ghosts.
   The above is the only note I have of someone specifically con-
necting <1holman>1 with <1pere$ta>1 (rather than with <1yaks*a>1); but equally
significant is the remark of a schoolgirl that there are <1holman>1 only
when bad men, not when good men, die. Moreover, to add to the
confusion there is a category of <1yaks*a>1 called <1mal*a yaka$>1 ("dead
<1yaks*a>1'). I only came across the term once in my field-work (des-
cription of Kad*avara--see below, p. 187), but Gooneratne gives
this account of them (p. 19): "If a man, who lives at enmity with
another, remember, on his deathbed, just before he dies, and at
the very moment of his expiring, any thing relating to that enmity
and if, instead of a feeling of forgiveness, resentment and hatred
take possession of his mind, he is supposed to become, after
death, a demon of this kind. These demons are not so powerful,
as [big demons like Mahaso$na$], nevertheless they too cause
sickness.'
  The indeterminate position of the wicked dead seems estab-
lished. It is aggravated by the ideas about hell. In the doctrinal
picture hells have two classes of inhabitants, the wicked and the
devils who torment them, but this is incongruous with the picture
of heavens in which the good are reborn <1as>1 the gods. Moreover,
people are supposed to identify with the tormented, not with the
tormentors; but it is dubious which should be considered the
worse fate, for the tormented are expiating their sins and so will
be reborn elsewhere, whereas the tormentors by their cruelty are

                                                               167

sinning more. My objections are a purely academic quibble, for
the hells are part of the ideal rather than the actual universe; they
are certainly believed in, but as men cannot interact with their
inhabitants the only effect they can have on behaviour is to deter
from sin, an effect not measurable by the social scientist. How-
ever, these slight inconsistencies even on the cognitive level do
suggest that affectively the doctrine of unpleasant rebirth for bad
<1karxa>1 matching pleasant rebirth for good <1karma>1, all rebirths
being transient states, does not carry full conviction.  The
ambiguous nature of the <1preta>1    and their imperfect demarcation
from <1yaks*a>1 I have shown to be canonical; the dilemma about hell
is canonical also. In the canonical list of five rebirths (<1gati>1) the
others are all as types of creature; but the worst is just "hell',
whether as tormentor or tormented being unspecified. There is in
fact no rebirth as devil mentioned in this list: it was no doubt to
remedy this deficiency that the otherwise obsolete Vedic <1asura>1
were introduced to extend the list of <1gati>1 to six. The Sinhalese
<1holman>1 are a local product of the very widespread human hopes
and fears that the dead may revisit their former dwellings and
acquaintances; but in so far as such an event would be inconsis-
tent with the doctrinal Buddhist scheme of things, the incon-
sistency is nothing new.
  Working up from the bottom of the universe in the same way as
the list of <1gati>1, we have now dealt with hell, <1preta>1, and indeed
<1asura>1. Animals require only a few words. They too are vaguely
considered to form some kind of hierarchy, with the elephant,
hero of Sinhalese folklore, at the top. As in western culture,
animals traditionally symbolize certain vices: the pigeon stands
for lust, the pig, dog, and crow for rapacious greed. A man
subject to a particular vice is likely to be reborn as the appropriate
animal. The cobra (<1maya$>1--Skt. <1na$ga>1) has a peculiar position:
though it is always a sin to kill, to kill a cobra is especially sinful
and extremely unlucky--two different ways of saying that it is so
bad that bad results are sure to ensue before long. If a cobra is
known to be living somewhere he is not disturbed, but often offered
saucers of milk. Cobras sometimes are said to guard sacred places.
In the Canon <1na$gas>1 are a separate class of beings, semi-divine five-
headed cobras, and the cobra's special status is a survival of this
belief.
  The next on the list of <1gati>1 is man. The human world is itself

168

hierarchically structured in the caste system (see Chapter 8). To
be reborn in a good human station is more desired by some than
to be reborn in heaven (see Chapter 6), nor is this wish necessarily
irreligious, as it must seem to a westerner, for men are held to
have a better chance than the lower gods of attaining <1nirva$n*a>1.
Any human being, irrespective of status, may attain <1nirva$n*a>1, but
it is widely believed to be much harder for women. The relations
of all human beings to the physical world are the same, except
that enlightened people (<1arhats>1) have supernormal powers (<1r*ddhi>1)
which enable them to pass through solids, fly through the air, etc.
(In the Canon the acquisition of these powers comes several stages
before Enlightenment, but in the popular imagination this gap has
disappeared.) <1Arhats>1 are finished when they die, and are not
normally worshipped, though in ancient times (when there <1were>1
<1arhats>1) their relics were venerated.  However, an <1arhat>1 called
Si$vali occupies a somewhat anomalous position. He is little men-
tioned in the Pali texts, and then mainly in connection with his
birth; but the Buddha called him "pre-eminent among recipients
of gifts',<s37>s and the merit he had gained by giving honey to a
former Buddha was so great that in this life he would never go
hungry, so the Buddha took him along on journeys through
difficult terrain.<s38>s He is portrayed as an ordinary monk, seated,
holding an alms bowl, and his picture is in many homes (including
some <1pansal>1), and even on buses over the driver's seat (when the
driver is a Buddhist). He does not receive prayers or offerings; his
picture merely seems to serve as a general "lucky chanu'.
  We are now back with the gods and the problems they raise. I
have already shown that all the main Sinhalese beliefs about the
workings of the universe and the pantheon are part of the great
tradition. Even Vishnu and Kataragama (described in more detail
below), who are most typically cited as innovations in Therava$da
Buddhism, are certainly not ascribable to a "little tradition'; they
are pan-Sinhalese, not local gods; worshipped by members of the
educated classes (probably more than by poor villagers); his-
torically derived principally from the great tradition of Hinduism;
long absorbed into Sinhalese literate culture as their presence in
the <1Maha$vamsa>1 will testify; and fitting comfortably into the
Buddhist doctrinal scheme of things. Certain individual local

                                                                169

godlings and demons have no such respectable histories to offer,
and can certainly be ascribed to little traditions, but this is
trivial--the point is that they all fit into the structure which the
great tradition has provided. To my mind the principal interest of
little traditions (sub-cultures) in Ceylon lies in their <1selectivity>1
from the great tradition. In Mi$gala there is a proliferation of minor
gods, but little stress on <1yakku>1, and <1pere$tayo$>1 are so little considered
that they have by and large retreated from the actual to the ideal;
a glance at the work of Ames or Wirz will show by contrast the
enormous exuberance in the Low Country of belief and practice
concerning all the lower and nastier spirits; and I believe that
this is a general contrast between the Low Country and the
Kandyan provinces. On the other hand I think that Obeyesekere's
distinction between the ideal and the actual pantheon can be
explained in terms of the system itself: it is the inhabitants of our
world, the human world--"little' gods, men, animals, <1yaks*a>1, and
ghosts--who are actual, and <1preta>1 are on the borderline because of
their unclear status and their confusion with <1holman>1.
  The two gods who were the most important in the Canon--so
much so that they alone deserve to be called "actual' canonical
gods--are the two who have survived to head the ideal Sinhalese
pantheon, and of these two S/akra (Sinh. : S/akraya$) is--as he was--
more important than Maha$ Brahma$. S/akra has remarkable staying
power because under the name of Indra he was the most impor-
tant god in the R*g Veda. However, after a vigorous youth as a
rumbustious warrior god and a respectable middle age as a highly
moral supporter of the Buddha, he has declined into a feeble and
somewhat disgraceful senility; for he is now little considered, and
his only appearance in ritual of which I am aware is as a lecherous
old man!<s39>s In the Canon S/akra rules a heaven (called in Pali the
Ta$vatim*sa heaven) which is one away from the lowest; the lowest
is ruled by the Four Great Kings or "world guardians' (<1lokapa$la>1),
one for each point of the compass: Dhrtara-s*t*ra, Viru$d*ha, Viru$-
pa$ks*a and Vais/ravan*a rule E., S., W., and N. respectively. Except
for Vais/ravan*a, king of the <1yaks*as>1, these four are totally colourless.
These six gods, S/akra, Maha$  Brahma$, and the four world
guardians, are frequently portrayed together at the entrance to
image-houses, usually forming a standardized framework for the

170

<1makara toran*a>1 on the outside entrance wall of the central shrine.
S/akra and Maha$ Brahma$ may also appear in representations of
scenes from the Buddha's life, even attending on him like divine
correlates of the two chief disciples; but the four world guardians
have no more than a decorative function. Vais/ravan*a may be said
to lead a double life: as king of the <1yaks*as>1 he is imagined as a kind
of super-<1yaks*a>1 himself, but in temple iconography he is a good-
looking young king not distinguishable from his colleagues.
   Below the four world guardians we enter the human world and
the actual pantheon, gods with whom the Sinhalese interact. At
the top of the hierarchy are the four guardian deities (<1hatara
varam deyiyo$>1) of Ceylon, two of whom are always Vishnu and
Kataragama. Traditionally these gods were associated with a
particular area of Ceylon in which also stood their principal
shrine: for instance, Kataragama is at home in Ruhun*u, the south-
eastern part of the country, which contains his eponymous shrine.
However, the country was in ancient times divided into three
parts (Pihiti, Ma$ya$, and Ruhun*u) not four, and the list of guardian
deities fluctuates. That their number was fixed at four only
because of the influence of the four world guardians of the Canon
is virtually proven by the evidence of those Sinhalese books retail-
ing canonical material in which the four world guardians are
confusedly called the <1hatara varam deviyo$>1.
  Now that we are getting down to the human world it is time to
remark on the structure of the pantheon. Here I can do little
better than recapitulate the conclusion of Dr.  Obeyesekere's
brilliant article. He has shown that the structure of the pantheon
follows the structure of authority in the feudal system which was
once so exactly represented in the Kandyan kingdom: all power
derives ultimately from the Buddha, who has given the gods
<1varam>1 ("warrants'), and similarly the greater gods give warrants to
the lesser gods, who form their retinue (<1pirisa>1 or <1parivara>1).  I
entirely concur with this interpretation: indeed I was given it
spontaneously by an elderly monk who can never have heard of
Dr.  Obeyesekere's work, a startling tribute to Obeyesekere's
acumen. There are various symbolic ways in which the gods are
made to defer to the Buddha's superior authority, and a few of
them will be pointed out in the remainder of this chapter; but none
of them require that the Buddha be conceived as still living.
Obeyesekere presents his feudal model with the Buddha as the

                                                                     171

king at its head, alive and active at least to the extent that he is still
able to issue new warrants. He is fully justified in so doing, as he
cites a Low Country drama as evidence for the active Buddha.
However, it is not logically necessary that the Buddha be alive
for the warrant mechanism to function if he has already delegated
his power to some high god or gods; new warrants can be issued
by further delegation.<s40>s In fact I never heard anyone suggest
even in this context that the Buddha is still alive or active. On the
contrary, my informant clearly stated that S/akra was king of the
gods--a position which I may say has been attributed to him in
very different times and circumstances. At the top of the pyramid
of current divine authority I would accordingly substitute for
Obeyesekere's Buddha my informant's S/akra. S/akra is the king;
if the Buddha is anything in this model we will have to transfer to
the feudalism of mediaeval Europe and call him God who has
legitimated the occupant of the throne.
  Theoretically the four guardian deities of Ceylon derive their
power from S/akra, or his council of ministers or parliament or
whatever device he governs by, and the local gods, each with his
rigidly defined territory (<1si$na$va>1--"boundary'--the same term as
is used for parishes) are retainers of the guardian deity in charge
of their area; but in fact there are several gods of indeterminate,
or rather varying, status, and the whole thing does not work out so
neatly. Probably it never did, but I suspect that some of the con-
fusion has arisen through the gradual dissolution and merging of
local cultures which has followed political centralization, improved
communications, and the population explosion. Especially in the
last twenty years, since the eradication of malaria and general
advance in public health, villages have spilled over till the old
boundaries have only legal or historical existence, and the people
are no longer quite sure where the jurisdiction of the local god
ends. Improved communications mean the spread of ideas: out-
siders give the villagers new ideas about the gods, as about every-
thing else; and a god who hits the headlines, as it were, is likely to
displace his old colleague or rival. Political centralization has made
the old model of regional autonomy obsolescent, and the replace-
ment of feudalism by modern parliamentary and bureaucratic

172

government has blurred the clarity of the traditional hierarchy.
Some now compare the god's arrangements to Parliament. The
main beneficiary of all these changes has been Kataragama, who
seems to be on his way to ousting all the other guardian deities.
Obeyesekere even suggests<s41>s that within a few years this stern
and punitive father-figure, originally a war-god, will have made
Ceylon monotheistic (my term) for the first time in history. I saw
his power rising locally: while I was there the first Sinhalese shrine
(<1de$va$le$>1) to Kataragama in the area was opened. Moreover, he was
said to be the god from whom the local village gods derived their
warrants; but as the god of Ruhun*u I would think that he can
originally have had no business in the hill country, and I would
guess that he has filched the retinue of Saman or one of his other
declining colleagues. This last is, however, rather light-hearted
conjecture; against it may be adduced Kataragama's long estab-
lished position as one of the Four Gods in nearby Kandy, where
Saman has no shrine of any importance.
  Kataragama has long been the most important deity of the
Tamil Hindus of Ceylon, who are all S/aivites and whose religiosity
contrasts markedly with that of the Kandyan Sinhalese--Ruth
Benedict, using Nietzschean concepts, would have called the
former Dionysian, the latter Apollonian.<s42>s Coloured red, he has
six heads and twelve arms, which hold weapons; his mount
(<1va$hana>1) is a peacock. He is S/iva's second son, known to Sanskrit
mythology as Skanda,<s43>s Kuma$ra, or Ka$rttikeya, and to his Tamil
worshippers by these and several other names;<s44>s the name
Kataragama, far the most commonly used by the Sinhalese, is no
doubt originally the name of his shrine in the extreme south-east
corner of Ceylon.<s45>s Here he has two wives, Te$va$ni$ and Valli$
Amma$. Both are in fact Tamil names for his wife, who in Sanskrit
is called Devasena$. Sinhalese belief has it that Te$va$ni$ is the Tamil
wife he brought with him from India, with whom he lives most of

                                                                173

the time; Valli$ Amma$ is the Sinhalese or Va%dda wife he met in
Kataragama, whom he visits during the annual festival there in
July/August. Such annual festivals are celebrated in all his Hindu
temples, and involve processions (<1peraha%ra>1) and fire-walking; the
festival at Kataragama itself is well on the way to becoming, if it has
not already become, the most important event in the religious cal-
endar, attracting even more pilgrims than the A%sal*a Peraha%ra for
the Buddha's Tooth in Kandy. Though the Kataragama festival is
primarily for Hindus, drawing Tamils from all parts of Ceylon and
even from India,<s46>s it has since ancient times been a place of Bud-
dhist pilgrimage also, and Sinhalese in increasing numbers are going
there ad making vows to the god. (There is now also a Muslim
shrine there so that Muslims may have a religious excuse for join-
ing the pilgrimage. Christians have no such excuse--but go.) The
character of some of these vows inclines towards the gruesome: in
their fulfilment men--and women--not only shave their heads (or
half their heads) and walk on fire, but put skewers through their
cheeks, or even hang in mid-air from strings passed through the
tendons of their backs. The more arduous of these devotions are
practised mainly by Tamils, but fire-walking is not entirely alien to
the Sinhalese tradition, and more and more non-Hindus are under-
taking such vows, and coming under the influence of Kataragama's
dark and violent spirit.
  Most of the monks I asked (twenty out of twenty-seven), had
visited Kataragama, several more than once, though they usually
hastened to add that they went just to look, not for worship
(<1bhaktiya>1) or to make a vow (<1ba$ra venta>1).  Several, however,
mentioned that (like other gods)   he should be given merit, and
some said he gave protection (<1a$raksa$va>1) to Buddhism (<1sa$sane>1) or
Buddhists.  Opinions and knowledgeability about him varied
greatly, and seemed to constitute one of the best indices of tradi-
tionalism, the more traditional monks knowing and caring little
about him and holding him in low esteem. A very old monk, who
had never been to Kataragama, seemed to put him on a level with
<1yakku>1, saying that he rules over cemeteries (<1balapa$nne$ sohonata>1).
Several mentioned that he inspired fear, but one said there is no
reason to be afraid; he came to Ceylon in flight from India after
losing a war, and people just believe that he has supernormal
power (<1te$jasa>1).  His harshness to Tamils was stressed, and is

174

explained by a myth: he came to Ceylon from India, and assumed
the form of a boy. When he landed it grew dark and he needed
shelter for the night. He knocked at the hut of a Tamil, who
refused him. He then went to the hut of a Sinhalese, who not
merely asked him in but went out himself, to give the whole hut
to the visitor. For this Kataragama is now making the Tamils pay
(<1vipa$ka denava$>1) and the main priest (<1kapura$l*a>1) of his shrine has
to be a Sinhalese. The hut in which he stayed is reconstructed in
front of his <1de$va$le$>1. The monk who told me this was just back
from a trip to Kataragama, and far more positive about him than
most: Kataragama, he said, was very good and truthful (<1satyava$di$>1).
It is widely known that Kataragama is supposed to have killed
<1asuras>1 in a great battle; the above monk who favoured Kataragama
justified this conduct, explaining that the <1asuras>1 were very wicked,
committing the ten sins (<1dasa akusal>1), and therefore deserwed to
be punished. Another monk, less pleasantly, explained that the
<1asuras>1 were in fact Tamils. The sufferings which Tamils under-
take in his worship are felt to require explanation, and this
explanation coincides with Sinhalese nationalist prejudice. The
monk who made it connected Kataragama with Dut*uga%mun*u's
victory over the Tamils--not unreasonably as Dut*uga%mun*u grew
up in and began his march from the immediate vicinity of the
shrine, which indeed another monk said he built in fulfilment of a
vow to the god. I asked some monks whether killing did not make
Kataragama a sinner (<1pavka$rayek>1), and they usually agreed.
Moreover, said one, he has servants who are <1yaks*as>1, and he also
sins through them. Only one monk held him to be a Bodhisattva,
and he was a man of deviant opinions who nevertheless cannot
have held him in high regard, as he had never been to Kataragama.
As for his history, he was generally connected with the human
world and with India. One monk said that most of the gods were
just men (<1minissu nayi>1) who were born in India, and because of
their merit were reborn as powerful gods. The statement of
another monk that Kataragama is not a man but a god by birth
(<1utpattiyen deyyo$>1) does not contradict this; it is rather aimed
against the euhemeristic type of explanation which I have already
cited, and of which I came across several more examples in this
context. To say clearly that he was a man <1in a previous life>1 fits
the traditional Buddhist scheme of things; but it is hard to know
how to interpret the remarks of the monk, an intelligent and

                                                                 175

modern-minded person, who said that all these gods were
originally men, and there is a popular feeling that they have been
reborn as gods; this implied no scepticism, for he also said that
Kataragama had the power to help people, and he knew of cases
in the village where he had cured barrenness.
   Kataragama's arrival in Buddhism as a Hindu alien was re-
marked on by several, not only in metaphorical terms. One monk
by implication put the Hindu influence back to the time of
Dutuga%mun*u (c. 100 B.C.) or earlier. A second told me that <1viha$ras>1
and <1de$va$le$s>1 have been built together since the Pol*onnaruva period.
A third associated a more recent dating with an unusually out-
spoken condemnation:  popular beliefs about Kataragama are
heretical (<1vaituli adahas>1). The last kings of Kandy, he said, were
of S/aivite origin and brought S/aivite queens from South India.
These queens believed in Vishnu and Kataragama (who was once a
man in India), and so they became popular. When the kings went
to worship the Buddha, the queens wanted to go and worship
Kataragama, so they put the shrines together for convenience;
this is how shrines to gods came into temple compounds. Then
according to S/aivism (<1S/iva$game$ ha%tiyata>1) you must make a vow,
and give a hand for a hand, etc. [a reference to <1ex voto>1 offerings];
this is contrary to Buddhism (<1Buddha$gamat*a viruddhayi>1). But
Kataragama has no real power; his influence (<1balapa%$ma>1) depends
on belief (<1visva$sa>1), and, like a magician, he works on credulity.
  According to Professor Hans-Dieter Evers there are many
myths according to which a <1viha$ra>1 was built by the king, an
accompanying <1de$va$le$>1 by the queen. Though this was the only case
I came across, I can well believe that the assertion is frequent,
for it is a natural way of expresaing that the two systems repre-
sented by these buildings are conceptually distinct but com-
plementary.
  The above monk was the only one to say that making vows was
"contrary to Buddhism', though the maverick monk who said
that Kataragama was a Bodhisattva came rather near it when he
said that to worship gods is heretical (<1mithya$ dr*s*tiya>1), and it is
wrong (<1varada>1) to believe in gods alone (<1deviyanma a$dahi$ma>1),
and it is wrong to ask for individual gain, as people do at Katara-
gama, rather than for liberation (<1moks*a>1). I am, however, sure that
these two monks did not mean to express any disagreement with
the clearly-expressed opinion of several others, that the worship

176

of Kataragama is simply irrelevant to Buddhism.<s47>s Belief in gods
is not contrary to Buddhism, but has no connection with it, said
one.  Buddhists <1as such>1 have no devotional worship (<1bhaktiya>1)
and make no requests (<1illi$ma>1) of gods, said another. Belief in gods
is not a matter of religion (<1a$gama vas/ayen adahanne na%$>1) said a third,
who certainly believed in some of them. The same attitude was
expressed by numerous monks who said that one should give the
gods merit but nothing else. What the monk who associated
Kataragama with S/aivite queens had in mind was that worship of
Kataragama is specifically S/aivite and not found in the texts; he
also was probably thinking of what the maverick monk went on to
say: that vows to Kataragama are often of a dubious moral
nature--victory in law suits, or worse--and people who ask the
god for favours against others are reborn as <1yaks*ayo$>1.
  So much space has been given to Kataragama because it was
concerning him, as by far the most important god in modern
Ceylon, that I mainly questioned the monks, and their attitudes
to him exemplify their attitudes to "popular beliefs' in general.
The other guardian deities of Ceylon can be disposed of more
briefly, especially as they are already the subject of much scholarly
literature. There are in Kandy four <1de$va$le$s>1 which have for many
centuries combined to hold the A%sal*a Peraha%ra<s48>s (see Chapter 3),
and they belong (in order of ceremonial precedence) to Na$tha,
Vishnu, Kataragama, and Pattini$. This tetrad is not the same as
that which appears as the protectors of Ceylon in inscriptions and
literature of the fourteenth and fifteenth centuries: Upulvan (i.e.
Vishnu--see below), Saman, Vibhi$s*an*a, and Skanda-Kuma$ra
(= Kataragama).<s49>s Only Kataragama and Vishnu are in both lists,
so it is not surprising that when I asked people to name the Four

                                                             177

Gods (<1hatara deyyo$>1) they were the only two always named.
Were the analogy of divine with human government consistently
applied we would expect the Four Gods to have Ceylon, or perhaps
Sinhalese Ceylon, divided between them into four non-overlapping
territories (<1si$ma$>1), but the reality is not so tidy. Of Vishnu I was
often told, in accordance with the <1Maha$vam*sa>1 story, that his
territory is the whole of Ceylon.<s50>s His chief shrine in Ceylon is at
Devundara, on the south coast. Ancient and mediaeval references
which are commonly assumed to be to Vishnu, including the
<1Maha$vam*sa>1 passage and references to the Devundara shrine,
in fact say Upulvan/Uppalavann*a. The name means "having the
colour of a blue lotus', and Vishnu has. Professor Paranavitana
questions whether Upulvan really was Vishnu,<s51>s but as the two
have been popularly equated the question is purely academic.
Vishnu's traditional character is benevolent and colourless;
his character as <1fidei defensor>1 is his only salient characteristic.
As in India he is shown as blue, with four anns, and his mount is a
kind of eagle (<1garud*a>1).
  Na$tha (the name means "lord' in Sanskrit and Pali) has almost
been forgotten. His main shrine is at Tot*agamuwa on the south-
west coast, but to my knowledge he is never considered the
guardian deity of that or any other particular region. His Kandy
shrine is probably the oldest of the four <1de$va$le$s>1 there, being built
in the second half of the fourteenth century, and was used for
ceremonies at the inauguration of kings of Kandy.<s52>s The dis-
appearance of the court dealt him a severe blow. Na$tha was once a
mere epithet of Avalokites/vara, an important Bodhisattva of the
Ma$ha$ya$na Buddhism which came to Ceylon in late classical times,
but this has been forgotten--Avalokites/vara is a name not known
to villagers. However, he was twice named to me when I asked which
gods were Bodhisattvas, so a memory lingers. A middle-class lady
in Kandy told me that people do not make vows to him as to the
other gods, and explained this by saying that he is to be the next

178

Buddha Maitri$, so presumably has lost interest in vows. Monks
too said that Na$tha is Maitri$. It is also notable that he does not
share his shrine with other gods, but lives apart. Maitri$ today
shares an iconographic feature, the small seated Buddha figure in
the headdress, with the classical Avalokites/vara.  Maitri$ and
Avalokites/vara Na$tha are several times mentioned together in
inscriptions and literature from the fourteenth to the eighteenth
century.<s53>s There has clearly been some conflation. However, I
doubt the Kandyan lady's explanation of why people make no
vows to Na$tha, although it would support my thesis; I suspect
that once they did, but that he has simply drifted out of the actual
pantheon for reasons unconnected with his Bodhisattva nature,
and so lost power and prestige.
  Pattini$ too is losing ground. She still has shrines all over the
country, including two in my area of research, but few people
name her as one of the guardian deities. She is a goddess from
South India, with a highly distinctive cult which has been ex-
haustively studied by Obeyesekere. She is notable as the only
lady in the pantheon, actual or ideal. Her rich background of
ritual and mythology has been largely forgotten in my area: I was
merely told that she was powerful (<1satta>1) and could cure pox;
one monk identified her with Na$gamma$, a mother goddess who
has a temple in Na$gadi$pa near Jaffna. Her headquarters are at
Navagamuva near Colombo.
  More commonly listed as one of the Four Gods (<1hatara deyyo$>1)
than either Na$tha or Pattini$ is Saman. His historical origins are
the subject of another learned monograph by Dr. Paranavitana,<s54>s
but again its contents are not in any sense known to the villagers.
Saman is <1par excellence>1 the god of Adam's Peak, which in Pali is
perhaps named after him (Samantaku$t*a)--or he after it--and so he
is the god of the Kandyan hill country. In fact he has another
major centre, even closer to Kandy, at Mahiyangana. It is interest-
ing that in both places he is intimately associated with the Buddha,
for these are the sites of the Buddha's footprint and his first visit
to Ceylon respectively. There is thus something very Buddhist
about Saman; his figure appears frequently in <1viha$ra>1 in my area
and one monk went so far as to say that he is the very highest god,
next to the Buddha in rank, with Vishnu, Kataragama, and

                                                                179

Alutnuvara below him. He is represented as white, and his mount
is a white elephant. I was told in Mi$gala that the season for
pilgrimages to Adam's Peak ends at Wesak; on the two nights of
the Wesak full moon festival no pilgrims, not even monks, make
the climb, but the way is left clear for Saman's white elephant,
who lives secluded in the jungles on the mountain alopes, to
ascend and pay solitary homage to the Buddha.
  Vibhi$s*an*a, the last god with a historical claim to be one of the
Four, rarely cropped up in my informants' lists. In Sanskrit
texts this is a name of S/iva, but about Vibhi$s*an*a nothing seems
to be known these days; he is totally colourless and iconographically
bears no resemblance to S/iva, but is not individuated. His main
shrine is at Ka%lan*iya near Colombo. Mr. Malalgod*a tells me that
it was razed by the Portuguese; this must be a cardinal reason for
his decline.
  The gods mentioned so far are all, I believe, of nationally
recognized status. Of more strictly local reputation is Da%d*imunda
alias Alutnuvara Deyyo$ <1alias>1 De$vata$ Ban*d*a$ra Deyyo$. (Three such
different names may suggest the conflation of three gods; if this
is so the conflation is complete, as several informants told me that
these names denoted the same person, nor were there any state-
ments to the contrary.) Da%dimunda's status and function are very
differently assessed. Some say he is in the retinue of Kataragama,
some (more circumstantially and so more plausibly) of Vishnu;
but some even counted him (under the name of Alutnuvara)
as one of the Four Deities, and so he must be mentioned here.
His main shrine (under the names of Da%dimunda and Alutnuvara)
is some twenty miles west of Kandy, near Ma$vana%lla in the
province of Sabaragamuva.<s55>s I was told at the shrine, which is
large and extremely prosperous, that it is about 800 years old,
and was originally to Vishnu, but since Vishnu's shrine was
moved to Kandy in 1748 the chief god is Vishnu's minister

180

(<1ama$tyaya$>1), Da%d*imunda De$vata$ Ban*d*a$ra. According to a Mi$gala
villager, Da%d*imunda got this shrine built by <1yaks*a>1 labour recruited
from Bengal and South India. The image at the shrine is a gold
figure of a prince, only about a foot high; I saw no mount there,
nor did he have one in Mi$gala--a most distinctive feature. I was
given more news of him by various informants in Mi$gala. He is
the son of the <1yaks*a>1 Pu$rn*aka and a <1na$gakanya$>1 ("cobra maiden'):
Pu$rnaka figures in the <1Vidhurapand*ita Ja$taka>1,<s56>s so this gives
Da%d*imunda canonical antecedents. The main story about him is
this. When Gotama was seated under the Bo tree striving for
Enlightenment, Da%d*imunda was one of the godlings in attendance
on him. At the approach of Ma$ra all the godlings fled, but Da%d*i-
munda, being rather slow, could not make his getaway, and hid in
the Buddha's robe. When Ma$ra and his host had been routed the
Buddha looked round, and at that moment Da%d*imunda appeared
before him. "You alone of all the gods did not desert me,' said
the Buddha, and in gratitude put him in charge of Ceylon. So
I was told that Da%d*imunda is a guardian of Buddhism (<1sa$sanarak-
saka>1) for all Ceylon, and this fits him to Vishnu, who received a
similar trust. I shall have more to say of Da%d*imunda below when
he crops up in less exalted company.<s57>s
   Before going on to describe those deities who are definitely
lower in status, I should explain how the <1Bodhisattva>1 (Sinh.:
<1Bo$sat>1) fits in. Bodhisattva is a religious, not (like god) an onto-
logical status;<s58>s a Bodhisattva is simply a being who has taken a
vow to become a Buddha, or, in the phrase used by a monk of
Vishnu, a person fulfilling the requirements for Enlightenment
(<1bo$dhi sambha$ra puran*a aya>1). Whether someone is a Bodhisattva
may be known only to himself. (This does not imply that <1one>1
would keep it a secret.) In the <1Ja$taka>1 stories Gotama Buddha's
lives as a Bodhisattva are recounted, and he is always born as a
god (usually quite a minor deity), a man, or one of the higher
animals. A Bodhisattva would be virtuous, so that a lower rebirth
would be most unlikely. We have seen that in tenth century Ceylon

                                                                  181

all kings were believed to be Bodhisattvas. Nowadays, however,
only a few of the ancient kings (Dut*uga%mun*u, Deva$nampiya
Tissa, Maha$sena) and some of the higher gods are named as
Bodhisattvas, and on asking for names of Bodhisattvas I found
disagreement. Most claim that Vishnu is one, but one monk denied
it. The claim was varioualy made for Saman, Alutnuvara, Na$tha,
Vais/ravan*a, Kataragama, and perhaps Ta$ra$;<s59>s most monks denied,
either directly or by implication, that Kataragama was a Bosat.
Saman seemed to have the best religious reputation; he was given
me as an instance of a good Buddhist, and another monk told me
that he is the only god free from sin (<1keles>1) and has entered on the
path to enlightenment (<1sovan>1). One monk held deviant views which
must reflect some reading: Ta$ra$ is supposed to be above the gods;
Maha$ya$nists believe in her and think she is a Bosat; she is power-
ful in India. This was the monk who rejected Vishnu, but accepted
Kataragama as a Bosat.
   The rest of the pantheon in my area consists of gods who are
permanent or floating members of a group called the Twelve Gods
(<1dol*os deyyo$>1), and the group is usually considered to form the
retinue of Kataragama. Most of these twelve gods are village gods
(<1gamba$ra deyyo$>1)--that is, they are individually believed to be in
charge of a village, or rather a group of villages. There is said to
be an analogy between the human and the divine power structure;
using the title of the administrative officers who superseded the
feudal nobility, I was told that the Four Gods were like Govern-
ment Agents (who were in charge of a district) and the Twelve
Gods like District Revenue Officers (their immediate subordin-
ates). The Twelve Gods thus hold power on warrants from one of
the Four Gods--nowadays from Kataragama. An implied modifi-
cation of this analogy was, however, supplied by the old Mi$gala
<1kapura$l*a;>1 he said that though our local god, Pit*iya Deyyo$, is a
retainer of Kataragama, he is in absolute charge of his own territory,
so that if Kataragama wishes to visit it he has to get Pitiya Deyyo$'s
warrant. This seems to take us back to the days before the bureau-
cratization of the feudal structure, when nobles had full local
autonomy.
   Local gods (like the Twelve) are of rather different character
from regional gods (i.e. the Four). Local gods in some sense

182

validate the social structure: their rituals emphasize the traditional
status differences of caste and sex, and they are supposed, through
the <1kapura$l*a>1, to ensure good harvests and freedom from disease.
Traditionally they are the only gods of whom the members of the
village community make requests. But once a year the villagers are
supposed to go to the shrine of the regional god, just as the share-
cropper pays his annual viait of homage to the landlord, who may
live in the city. Villagers used to be somewhat frightened of
leaving the territory of their local god--as they walked through the
forest on their way to the regional centre they would repeatedly
say "Karun*a$vayi' ("Mercy') and hang strips of cloth on trees to
placate the spirits through whose territory they were passing. The
ritual at the regional centre paid no attention to status distinctions
--in the <1de$va$le$>1 of a major god all are equal; nor was this the place
to ask for anything specific:  one just worshipped and received
edification. This pattern is rapidly disappearing. As it becomes
easier for people to get to the bigger gods, by bus, they visit them
more often and informally, and ask them favours; the visits for
purely religious edification are those paid to the Buddha's relics.
The regional centre for Mi$gala should be Mahiyangana; a large
group of villagers went there, but it was at Poson, a day associated
with the Buddha, and it was to see the <1viha$ra>1 and worship the
relics that they went. It is not only because villagers now go to the
big gods that the little gods are in decline; the big gods are also
coming to the village. What this amounts to is the supersession of
specifically Kandyan tradition by the religioaity of the Low
Country and the middle classes--which is the same thing. Low
Countrymen outnumber Kandyans by about five to one; they have
also enjoyed the dubious advantage of about 300 years more of
European rule. Their traditional society was modified long ago
under European laws along European-built roads. With greater
mobility, and more education, they became economically domin-
ant; the trade in Kandyan town is almost entirely in the hands of
Tamils, Muslims, and Low Country Sinhalese. Geographical
(horizonal) mobility long ago disposed of local gods in the Low
Country and introduced Low Countrymen with their emphasis on
big gods and little devils into Kandyan towns and villages.
Social (vertical) mobility produced among Low Countryman a
middle class, something which hardly yet exists among Kandyans,
whose feudal aristocracy still own most of the land; the emancip-

                                                        183

tion of the middle classes disinclines them towards village gods
whose rituals remind them of their caste status and narrower social
horizons, while the notorious strains of social mobility incline
them to belief in astrology and even in black magic. Meanwhile
they enter the professions and the bureaucracy, and become the
models for ambitious villagers to emulate. A lady in the village
took to visiting regularly the Hindu temple to Kataragama a few
miles down the road, though she said she was afraid of him, and
spoke ill of our local village <1kapuva$>1 as a fake (though she still had
her daughters stand well clear when he became possessed); the
explanation lies I think not in her very pleasant personality,
but in the fact that as a low-caste woman she would be at a
double status disadvantage in the local cult, and visits to Katara-
gama made her feel a more progressive and socially superior
per son.
   It is in the context of the worship of the Twelve Gods (though it
would also apply to the worship of the major gods) that I came
across the concept of ritual impurity (<1killa>1) in Mi$gala. <1Killa>1 is
conferred by menstruation, by parturition, and by contact with
death. Caste impurity is <1not>1 covered by the term, and thus is not
conceived of as having anything particularly to do with ritual; there
is thus a disjunction between temporary and permanent impurity,
which contrasts with the Hindu situation. This is just what Pro-
fessor Dumont's theories would lead us to expect of a society
subscribing to a "world-renouncers" religion. Moreover--and
still in conformity with Dumont's theories--<1killa>1 affects only the
cult of the gods; there is no <1killa>1 for monks, and it has no relevance
to the worship of the Buddha or to anything else that the villagers
themselves conceptualize as "Buddhist'. For this reason I need say
no more about it.
   The worship of the Twelve Gods is sometimes known as the
"<1Ban*da$ra>1 cult'. <1Ban*d*a$ra>1 was a Kandyan title, and the word
figures in the names of many of these gods. The term "<1Banda$ra>1
cult' would be a good one, were it taken merely to imply that these
gods function in ways analogous to noblemen under the old order.
However Rahula (<1History>1, p. 37) calls it "the adoration of deceased
chiefs and prominent ancestors', an explanation to which the term
lends itself but which I believe to be only partially correct; so I
refer to them simply as the Twelve Gods.
  Ethnography on the Twelve Gods is rather scanty. Seligmann


184
wrote a short article<s60>s on them which does not entirely agree with
any other data I have come across. However, Barnett (op. cit.)
retails lots of stories about them which he extracted from the
manuscripts of folk poetry collected by Hugh Nevill in the late
nineteenth century and deposited in the British Museum. Wirz
(op. cit.) has data on the more maleficent spirits (see below). There
is also much relevant information in Parker's <1Ancient Ceylon>1,<s61>s
especially in Chapter IV. This is interesting, because Parker's
book is ostensibly about the Va%ddas; it shows how hard it is to
disentangle Va%dda and Kandyan Sinhalese culture. The stories--
and even some of the names--of the gods show too how strong is
the Tamil element in this cultural area.
   There were in or near Mi$gala three shrines to Pit*iya Deyyo-,
the local patron, each with its own priest called <1kapuva$, kapura$l*a>1,
or <1kapumahattaya$>1. The title is used for a marriage broker, so it
might be translated "go-between'; it indicates that the priest in
this cult acts as a medium, or shaman. The priests of these shrines
were related, an uncle (whom I shall call K 1) and two nephews
(K 2 and K3). K 1 and K 2 (who were not on speaking terms) were
my main informants on the Twelve Gods, and I asked them to list
the gods. K 1 said that there are altogether sixty-seven gods, of
whom he made offerings to twenty-four; of these, twelve are the
main ones and twelve are their attendants. Asked to name the
twelve he then listed fifteen, and when I pointed this out eliminated
three, whom I have put at the end of his list. The second list is that
of K 2, who initially listed twelve, but later in conversation
referred also to Amusiri Deyyo$. The third list was given to
me by the ex-headman (<1a$racci>1) of a village about eight miles
from Mi$gala.  It is worth reproducing because he seemed a
competent and spontaneous informant on this subject and his
list seems the least repetitious. It does not include Pitiya
Deyyo$, whom he confirmed to be in charge of the Mi$gala area,
but said his own area was under Kande$ Deyyo$, who protected
the village and cured diseases. I reproduce the titles exactly as
given.

                                                                      185

      <1K1                     k2                       A$racci>1
Pit*iya Deyyo$               Pit*iya D.
Gan#ge$ Ban*d*a$ra Deyyo$                              Gan#ge$ B.D. Gan#ge$ B.D.
De$vata$ Ban*d*a$ra Deyyo$                             Alutnuvara D. De$vata$ B.D.
   = Da%dimunda
Ki$rti Ban*d*a$ra Deyyo$                               Ki$rti B.D. Ki$rti B.D.
Kal*u Ban*d*a$ra Deyyo$                                Kal*u B.D.
Vanniya Ban*d*a$ra Deyyo$                              Vanniya B.D. Vanniya B.D.
Kal*ukuma$raya$              Kal*ukuma$ra D.           Kal*ukuma$raya$
Kohom#ba Deyyo$              Kohom#ba D.               Kohom#ba D.
Kadavara                     Kad*avara Deyyo$          Kad*avara Deyyo$
Kandakuma$ra                 Kande$ Deyyo$             Kande$ D.
Amusiri Yaka$                Amusiri Deyyo$            Amusiri Deyyo$
Mangara Deyyo$
Hu$niyam                     Hu$niyam De$vata$ Ban*d*a$ra
Valko$n Ban*d*a$ra Deyyo$
Abayako$n Ban*d*a$ra Deyyo$
                             Siddamula$ Hu$niyam
                             De$vata$ Ban*da$ra
                             Palleba%dda Deyyo$        Palleba%dda Deyyo$
                                                       Kiriamma$ Deyyo$

Translations :
   Pit*iya:       a tract of flat ground
   Gan#ge$:       in/on/at/of the river
   Ki$rti:        fame, glory
   Kalu:          black
   Vanniya:       of the Vanni, a desolate area in northern Ceylon
   Kalukuma$ra:   black prince
   Kohom#ba:      Margosa (a kind of tree)
   Kandakuma$ra:  mountain prince
   Kande$:        on/of the mountain
   Amusi$ri:      raw blood
   Hu$niyam:      void
   Kiriamma$:     milk-mother (term for wet-nurse, also for grand-
                    mother)
   Palleba%dda:   a place name

Allowing for some regional variation we seem to have a well
standardized list. I agree with K 1's second thoughts, for his last
three names do not really belong: Hu$niyam is a pan-Ceylonese
deity who will be discussed below; the other two were not named
by anyone else. Similarly, I would eliminate Siddamula$ Hu$niyam
from the second list--he looks like a makeshift variant on Hu$niyam

186

to complete the   number. Kiriamma$ Deyyo$ is the only female.<s62>s
   K 2 told me that some gods <1areji$vama$na>1. This is a Sanskrit word
meaning "alive'. I deduce, however, that he used it to mean
"perceptible' or "perceived'; its opposite he expressed by the
sentence "when they are born they live without being visible'
("A$tmayak laba$gena innava$ pe$nen na%tuva'). He said Alutnuvara
Ban*d*a$ra Deyyo$ and Hu$niyam De$vata$ Ban*d*a$ra were <1ji$vama$na>1:
Alutnuvara can go on flowers, over rocks, or in the water; Hu$niyam
can take any form, e.g. a black dog or a tall white figure. K 1 said
that none of the Twelve Gods was married; ten were born of
woman (<1ma%n*ike$>1) like us, but Gan#ge$ Ban*d*a$ra Deyyo$ was born of a
blue gem in a blue sea, and Kandakuma$ra was born from the
mountain at Kataragama, so these two were born without pollution
(<1killa>1) and no one polluted may approach them. Another kind of
status difference was indicated by K 2: the only gods who can be
served by a <1kapuva$>1 who is not <1goyigama>1 ("high-caste'--see Chapter
8) are Kad*avara Deyyo$ and Amusiri Deyyo$. This seems to match
the fact that neither of these gods is ever called <1banda$ra>1 or
<1kuma$ra>1. The word which K 2 used for "service' was <1ra$jaka$riya>1,
which historically means the service obligations attached to feudal
land tenure--yet another analogy (if that is not too weak a word)
between the religious and the secular.
   As all the above deities except Valko$n and Siddamula$ Hu$niyam,
on neither of whom do I have any more information, have their
stories told by Barnett, I shall only report from my own notes
enough detail to suggest their general character. Certain stories--
of a prince or noble banished for a misdemeanour; of the miracul-
ous removal of a rock impeding irrigation--are variously told of
most of the Twelve; as is common with such groups, their indi-
vidual characteristics are largely interchangeable.
   I was given most information on Pit*iya Deyyo$, our local patron.
For 50 cents K 3 showed me his picture, or photograph (<1cha$yaru$->1
<1paya>1) as he called it, which he claimed to be ancient and could well
be fifty years old. It is a picture of a man in the costume of a
Kandyan nobleman (<1radala>1) with a horse (Pit*iya Deyyo$'s invari-
able mount) and four attendants, painted on a piece of wood
shaped rather like a wine glass, with ornamentation round the

                                                               187

borders. It is kept under a cloth and hence is well preserved. The
picture of Pit*iya Deyyo$ produced by K 1 was a less venerable
object, a cardboard cut-out of a bearded man with a white horse.
K 2 at his shrine also showed me a Kandyan noble with a white
horse: he also had pictures of Palleba%dda Deyyo$, over whose
"photograph' cockroaches were crawling, Vanni$ya Band*a$ra Deyyo$
and Alutnuvara Ban*d*a$ra Deyyo$. All look like Kandyan noblemen.
Vanniya B.D.'s mount is an elephant; Alutnuvara has no mount.
   Pit*iya Deyyo$, says K 2, escaped from India after committing
some misdemeanour there and came here in the days of Ki$rti
S/ri$ Ra$jasim*ha [i.e. about 200 years ago]. He viaited sixty-seven
places and from each acquired a follower; all of them are now
gods and form his retinue (<1pirisa>1); Palleba%dda Deyyo$ was his first
and chief follower. According to K 1, he came from India five or
six hundred years ago; his story went on to connect him with
various local places: where he first stopped, ate, bathed, etc. He
is in charge of everything locally, and (like all gods) knows the
future, on which he can be consulted at his shrine via the <1kapuva$>1.
Not all, however, agreed that Pit*iya Deyyo$ came from India;
a monk identified him as the son of the righteous Tamil king of
Ceylon, El*a$ra. While driving in his chariot that unfortunate prince
accidentally ran over a calf, for which, on complaint from the calf's
mother, he was executed by his father. (The story comes from the
<1Maha$vam*sa>1, XXI, 15-18.) Another informant told me the same
story but located it in India. All agree that Pit*iya Deyyo$ is (the
reincarnation of) a prince, that he is connected with a (perhaps
unintentional) misdemeanour, and that he was Tamil.
  Kal*u Ban*d*a$ra Deyyo$ has a <1de$va$le$>1 on a large mountain nearby,
but no one knows where it is. Once an old farmer lost his two
buffaloes and went searching for them with his son. They found
them in a pond. They were so tired that they too bathed in the
pond, and when they came out the father was young again, and so
were the buffaloes. When they returned home the farmer's wife
at first refused to have anything to do with him, and was convinced
only with difficulty that it was he. The villagers then went out
looking for the pond, but it had disappeared.
  Kad*avara Deyyo$ is a <1yaks*a>1 who lives in trees in the jungle. He
used to be a minister in India of a certain King of Malaya (i.e.
South India), but having died as a man he became a <1yaks*a (manu-
s*yek ma$rila$ yaks*ek vuna$>1); he is a <1mal*ayaka$>1 [literally "dead <1yaks*a'>1,

188

i.e. the reincarnation of a man]. He cut a rock to make a water-
course at the Temple of the Tooth in Kandy. It is tempting to
associate this with the fact that according to the dictionary
<1kad*avara>1 means a ditch or channel, but the meaning is given as
lexical only, so the connection may be fortuitous or the dictionary
mistaken. Parker (p.  157) shows that the name is Tamil and
means "the Celestial who escaped', which fits another myth.
    Kal*ukuma$ra, the black prince, is unambiguously maleficent.
As his name indicates, he too is or was a prince, and he is said
by K 2 to have 6,000 followers. According to K 1 he was the son
of a Kandyan king. He and his father each reared a <1kon*d*eya>1 bird.
One day he let out the two birds and they fought, and his bird
killed his father's. At this the king sentenced his son to death, but
on the insistence of his ministers the sentence was commuted to
banishment, and he came to a river called the Maha Oya and
settled down there in the Maha Oya De$va$le$. [The Maha Oya
is in south-western Ceylon.] But the main point about Kal*uku-
ma$ra is that he afflicts women, appearing to them in dreams
or otherwise casting his gaze (<1ba%lma>1) upon them to disturbing
effect--I gather he is a kind of incubus and may then have to be
exorcised. Though he usually confines his attention to women, an
elderly man in Mi$gala told me that Kal*ukuma$ra was always in
him and would possess him during ceremonies so that he was
carried away (<1va%hemava$>1); it might be aignificant that the man was a
bachelor.
    Hu$niyam or Su$niyam has a most peculiar history, and it is
doubtful whether he should properly appear here, up among the
deities, or down among the <1yakku>1 and <1holman>1. He is especially
liable to appear at night as a black dog or naked white figure;
according to K 2 he can bring good or evil, but another villager
said he was definitely benevolent. This opinion contrasts violently
with his character in the Low Country, where he is predominantly
associated with black magic, which may even generically be called
<1hu$niyam>1. He certainly is fond of assuming different shapes: I
was told that he has eighteen different forms or disguises (<1ve$sa>1)
and wears clothes of five colours with a garland of flowers, usually
white. He is often associated with Da%d*imunda. At Da%d*imunda's
big temple near Ma$vana%lla he is the only other god represented,
and is said to be Da%d*imunda's minister. One monk even identified
Hu$niyam with Alutnuvara. I found him sharing one <1de$va$le$>1 in

                                                              189

the grounds of a <1viha$ra>1 with Vishnu, Mahase$na Kataragama,<s63>s
and Da%d*imunda. On this occasion I was told that Vishnu is in
charge of the whole world, Su$niyam of Ceylon, and Da%d*imunda
of the Kandyan provinces (<1ud*a rat*a>1). Of Su$niyam there was a small
statue dressed in white as an ascetic (<1ta$pasa>1) taking the precepts
(<1sil>1), and a picture of him as a village deity (<1gamba$ra deyyo$>1), which
l was told to be his own usual form (<1niyama svaru$paya>1), with four
arms. Whether his appearance as a virtuous Buddhist is more or
less surprising than his nocturnal behaviour as a black dog depends
on one's view of the etymology of his name. According to De
Lanerolle in Geiger's <1Etymological Glossary>1, it has evolved (via
Tamil) from Sanskrit <1s/u$nya>1 "the void', the dominant principle
According to those Maha$ya$na philosophers who declared the world
to be "empty' in that things have no essence. (In this they were
developing the concepts, which they shared with Therava$da, of
impermanence (<1anityata$>1) and non-self (<1ana$tmata$>1); though Su$niyam
be a degenerate, I would not wish to call him a heretic !) But accord-
ing to Geiger himself, in an article<s64>s which De Lanerolle refers to
in his entry but apparently disagrees with, <1hu$-niyam>1 is El*u for
"thread-method', i.e. magic, so that Hu$niyam is a personification
of magical practice. Geiger's theory seems to me the more probable.
The above information on the Twelve Gods, although sketchy,
sufficiently demonstrates that to describe their worship as an
ancestor cult is a gross over-simplification. Even the title of
<1ban*d*a$ra>1 is no guarantee of a god's human origin (in the euhemer-
istic sense), for it is accorded to Hu$niyam too; it is rather a
Kandyan naturalization certificate. The most conspicuous char-
acteristic of the shrines of the Twelve Gods was that though
themselves unimpressive they were always associated with a
striking natural feature: a huge rock, a cave, above all a spot with
a wonderful view. All the three local shrines to Pit*iya Deyyo$
were on or by huge rocks, one of them in a cave in the jungle just
above a great sloping rock alab, another on a rock overlooking the
river and covered with temple trees, a third on a great rock which,

190

though not very high, commanded a magnificent view of the
village territory. The shrine to Kande$ Deyyo$ near the village of
the headman quoted above was on a mountain, and so inaccessible
that it was now visited only once a year. Gan#ge$ Deyyo$'s shrine
was on an island in the middle of a big river. Moreover the names
of at least four of the gods (Pit*iya, Gan#ge$, Kande$, and Kohom#ba)
associate them with natural phenomena. In this they recall the
canonical divinities of trees and parks (<1vanadevata$, a$ra$madevata$>1),
and the <1yaks*as>1 associated with trees and mountains: a group of
morally neutral nature divinities.
A second current associates them with the more distinctly
malevolent type of canonical <1yaks*a>1. The confusion of this group
with the nature divinities is, we recall, a canonical feature. Here
probably fit those gods who are most often called <1yakku>1 : Kad*avara,
Amusiri, Kal*ukuma$ra, and Hu$niyam. It is notable that these are
the very members of our list who figure in Wirz's account of the
demonology of the Southern Province.
Thirdly there is indeed the ancestor element, represented in the
titles <1ban*d*a$ra>1 and <1kuma$ra>1, suggested also by such names as Ki$rti
("Glory') and by Alutnuvara and Palleba%dda, which are place
names. Local people themselves incline to this type of explanation
of the cult; we have already discussed the ambiguity involved in
this interpretation among Buddhists, who are anyway committed
to a belief in the rebirth of all beings and therefore to a type of
euhemerism as an explanation of any god or demon. What is note-
worthy is that the stories so often involve the banishment of a
prince or minister for some misdemeanour. This has the ring of
historical authenticity, as the mountains have long served as
strongholds for disgraced potentates and rebel chieftains. The fact
that they are usually said to be Indians, or at least Tamils, likewise
sounds authentic. This true <1ban*d*a$ra>1 element therefore provides a
third strand in the constitution of the Twelve Gods. What must,
however, be noted is that two or three of the elements--nature,
maleficence, and ancestry--combine to form the individual
divinity. Pit*iya Deyyo$ has a name associating him with nature, is
supposed to have been a prince, and left India under a cloud;
Kal*ukuma$ra was also a prince who left under a cloud, brings a
particular kind of disease, and has a shrine in a river.
Our catalogue of the inhabitants of the universe seen from
Mi$gala is now concluded.

                                                                191


            III. ATTEMPTS TO INFLUENCE NON-HUMAN BEINGD

      Buddhism being considered to be the <1sacred>1 religion, while
      Demonism is only a religion relating to one's temporal
      interests, it is natural that the influence of the former should
      to a certain extent be felt on the latter.
                                             GOONERATNE, p. 15

   The institutions and ceremonies appropriate to each class of
being have been described at length by such authors as Wirz and
expertly classified by Ames. Tambiah has contributed an account
of the hierarchy of languages used in ritual.<s65>s In the following
paragraph I wish to do little more than give a picture of the types
of worship distinctive to Mi$gala, and this will again involve a
concentration on the Twelve Gods.
   The higher and more powerful the divinity, the more imposing
the edifices and personnel devoted to him; the lower and weaker
the devil, the more the attitude of human beings, both laymen
and specialists, changes from worship to control.

The <1Institutions>1

   (1) Conceptually the Buddha stands outside this scheme, for
in his case control is out of the question; but from the sociological
(i.e.  behavioural) point of view, as Ames has demonstrated,
monks, <1viha$ras>1, and <1Buddha pu$ja$>1 fit in well. The <1viha$ra>1 is described
in Chapter 2, <1Buddha pu$ja$>1 in Chapter 3. The main ritual language
is Pali.
   (2) After <1viha$ras>1 the largest (and richest) religious institutions in
the area are the Tamil temples (<1kovil>1) to Kataragama. In each
there is a brahmin in permanent residence; the one I visited was
specially brought down from Jaffna. The ritual is entirely S/aivite
Hindu, and the great majority of the cliente\le are Tamils from the
plantations. The main ritual language is Sanskrit.
   (3) The Buddhist god has a <1de$a$le$>1 served by a <1kapuva$>1, though
the major gods are also usually represented in <1viha$ra>1, in which
case they generally have no <1kapuva$>1 (and no organized worship).
In accordance with the reformist principles of the Ra$man+n+a
Nika$ya (see Chapter 8) their <1viha$ras>1 have no gods or <1de$va$le$s>1
on the premises; there were only three Ra$mannn+n+a <1viha$ras>1 in my

192

area. <1De$va$le$s>1 to the Twelve Gods exploit natural sites and often
lack buildings; there may be no more construction than a shelf to
serve as altar. The <1kapuvo$>1 are always of <1goyigama>1 caste, except, I
was told, for the gods Kad*avara and Amusiri, whom we have seen
to be more properly called <1yakku>1; the <1kapuva$>1 of one Kad*avara
<1de$va$le$>1 was said to be <1padu>1, a low caste. By tradition the office
of <1kapuva$>1 runs in families, but where new shrines are opened to
regional gods new men are sometimes recruited. The main ritual
language is Sinhalese.
   (4) Minor spirits have no buildings, and the specialists devoted
to them could appropriately be called exorcists. Material on this
subject is more prolific and classification more precise in the
Low Country. An exorcist of <1yakku>1 is called a <1yaka$dura$>1, of evil
planetary influences a <1balia%dura$>1, but usually these are the same
person, according to context. A supplier of amulets (<1yantra>1)
and charms (<1mantra>1) (who may in a specific context function as
<1yaka%dura$>1 or <1balia%dura$>1) is a <1kat*t*ad*iya$>1. These exorcists are usually
(but not necessarily) of low caste. The drummer caste (<1berava$yo$>1)
are supposed to be skilled in these matters, and are especially
associated with appeasing the planetary deities. A general term for
a magician is <1mantraka$raya$>1, "charm-maker'. His spells are in a
hotch-potch of languages, normally unintelligible.
   There is a general tendency, exemplified by the last paragraph,
for those who are specialists in one type of causation to pick up
knowledge and practice connected with other types. We have
already seen how monks, who may be said to have specialist know-
ledge of <1karma>1, and are involved in rituals to avert and cure mis-
fortune through <1pirit>1, tend also to become experts in astrology
and Ayurvedic medicine. The same goes for the other specialists:
astrology and Ayurvedic medicine are very often found together
also in laymen. As both are associated with learning they tend not
to be practised by the low-caste drummers, who are often socially
depressed; but these drummers, who participate in rituals for the
Buddha, are knowledgeable in ritual at all levels and are generally
reputed to be good at magic. <1Kapura$l*as>1 also tend to be suppliers
of <1yantra mantra;>1 and the one man in Mi$gala who knew astrology
(which he had learned from a monk) practised Sinhalese medicine
(part-time and for little or no money), and was the son and brother
of <1kapura$l*as>1.
   Before I describe the local rituals, it will give a fair idea of their

                                                                   193

relative importance if I list the religious institutions in my area.
There were about forty <1viha$ra>1, which do not concern us in this
chapter, and two Hindu <1kovil>1 to Kataragama, which will not be
mentioned again. There are two old shrines to Pattini$ (whose
priests are called <1pattini$ra$l*as>1). Otherwise the only <1de$va$le$>1 of any
age to a god of national standing was a <1de$va$le$>1 to Kataragama,
which was in a village entirely surrounded by tea estates, and had
a Tamil as <1kapuva$>1; though it was within the grounds of a <1viha$ra>1
it therefore hardly qualifies as a Sinhalese institution. With this
dubious exception twenty years ago there was not a single <1de$a$le$>1 to
Vishnu or Kataragama in the entire area. However, this is changing
fast. A Vishnu <1de$va$le$>1 recently opened next to one of the <1kovils>1,
ad during my stay a Kataragama <1de$va$le$>1 started in a cave in the
woods, very much in the style of the homes of local gods. The
habit of having a <1de$a$le$>1 in the <1viha$ra>1 grounds, extremely common
in the Low Country and a feature of some of the principal old
Kandyan temples, was gaining ground even faster: the first had
been built in 1949--and was run by a <1kapura$l*a>1 imported from
Ma$tara in the south--and four more were being built or projected.
I am excepting an old <1de$va$le$>1 to Pit*iya Deyyo$ which stood in one
small <1viha$ra>1 (commanding, as ever, a superb view), and a tiny
shrine to Vishnu in an uninhabited monastery. But I must not give
the impression that Vishnu and Kataragama are strangers to the
area. Most image-houses contain their statues, or at least that of
Vishnu, standing behind a curtain in the Hindu fashion, and the
worshipper can always worship them, drop a coin in the box
provided, or even make a vow if he should be so inclined.
   Accurate information on <1de$a$le$s>1 of the Twelve Gods is hard to
get, as many of them are falling into disuse. I asked about them
when interviewing monks and my list may contain duplications.
I listed fifteen <1de$a$le$s>1, all said to be to Pit*iya Deyyo$, Kande$
Deyyo$, Kad*avara Deyyo$ or Ki$rti Ban*d*a$ra Deyyo$. All of these
except the one already mentioned had no physical connection with
a <1viha$ra>1. Though I could not investigate them, I think the two
shrines to Pattini$ properly belong in this category, not merely
because of their age but also because they must have performed the
functions of shrines to village gods.
   The meaner spirits are only noticed when something goes
wrong, and as far as most people except the <1kapura$l*a>1 are con-
cerned this is true also of the Twelve Gods, excepting perhaps the

194

annual harvest festival. Statues of gods in <1viha$ra>1 are of course
accesaible to viaitors at any time, as the <1viha$ra>1 is always open:<s66>s
but any <1de$a$le$>1 which can be locked up is only opened by the
<1kapura$l*a>1, who attends on certain days called <1kemvara>1.<s67>s The
<1kemvara>1 in my area are Sundays, Tuesdays, and Fridays. In
other parts of the country they are Wednesdays and Saturdays.<s68>s
But this does not seem to be purely a question of locality. A
village lady told me that it varies from god to god; that Alut-
nuvara's days are Wednesday and Saturday. The days at the big
Alutnuvara <1de$va$le$>1 at Ma$vana%lla are in fact Wednesday, Saturday,
and half-day Monday (<1varu kemvara>1--afternoon only), but
whether this is determined by the identity of the god or by the
shrine's location (in Sabaragamuva Province) I have not enough
data to judge. The same lady considered that of the three <1kemvara>1
the most auspicious day on which to visit Kataragama's <1kovil>1
(possibly because it is generally an auspicious day) is Friday.
On the other hand a little used <1de$va$le$>1 to Pit*iya Deyyo$ opened
only, if at all, on Sundays. When a <1de$va$le$>1 to one of the Twelve
Gods is in operation the <1kapura$l*a>1 becomes possessed (<1ma$yam
venava$>1), but interestingly enough he will not do this on a <1poya>1
day. On the one hand the god gives precedence to the Buddha;
on the other there is something not quite proper about becoming
possessed, even if it is your role. I have never heard this something
made explicit, but I judge that it violates the Buddhist ideal of
self-control. To become possessed on a <1poya>1 day would therefore
be like getting drunk on a Sunday morning.
   Traditionally, as I have mentioned, people do not make vows
(<1ba$ra venava$>1) to the higher gods, but only worship (<1van#dinava$>1)
them for their blessing, which is conceptualized as "peace' (<1s/a$ntiya>1
or <1set s/a$ntiya; set>1 is merely the old Sinhalese form of <1s/a$nti>1).
There is some disagreement whether the gods do in fact do good,

                                                                      195

or merely give protection (<1a$raksa$va>1) from hann, especially the
hann brought about by <1yakku>1; in any event it is sufficient to trans-
fer Terit to them and perhaps occasionally to give a coin towards
their expenses.
   To the village gods, however, people hardly ever go unless
they have some specific request; they go to make contact with the
god, the <1kapura$l*a>1 acting as intermediary. Even if no one turns up
he is supposed to perform some rudimentary ritual (<1pu$ja$va>1)
for the god; what in fact happens no one can know, but I am sure
the ritual is most perfunctorily performed, for I never met anyone
who seemed to hold the local gods in awe. I saw <1pu$ja$>1 to the Twelve
Gods at all three local <1de$va$le$s>1 to Pit*iya Deyyo$. The <1kapura$l*a>1 is
naked to the waist and wears a white sarong with a red cloth
lucked in at the waist; he always wears his hair in the traditional
style, long and done up in a bun (<1kon*de$>1) at the back of the
head. The basic <1pu$ja$>1 consists merely of his blowing a conch shell
and then reciting an invocation (<1ya$tika$>1) while shaking a timbrel,
ringing a bell, and burning a little incense (<1dummala>1) before the
god's image. He then smears ash or saffron paste on his throat and
chest. Anyone who wishes to be associated with the offerings
must give a coin (<1pan*#d*uru>1--the plural form is used proleptically)
purified with lime juice and wrapped in a betel leaf.  Usual
additional offerings are candles, coco-nut oil (for lamps) or even
flowers--much the same things as are commonly offered before
the Buddha. An offering entitles the worshipper to be mentioned
in the invocation--the <1kapura$l*a>1 makes an entreaty (<1kannalavva>1)
to the god for his welfare. All invocations to these gods begin with
the words "Ai bo$' several times repeated; this is a clipped version
of "Ayu bo van', the normal greeting between human beings,
meaning "May you live long'. During the invocation the wor-
shippers are supposed to squat with hands respectfully folded.
At the end the <1kapura$l*a>1 makes a mark (<1tilaka>1) with saffron paste
on the forehead of each. Further services cost extra. At some
<1de$a$le$s>1 I have seen tariffs posted, but the only tariff I saw in my
area was at a <1kovil>1, and I doubt whether the institutions of the
Twelve Gods will not disappear altogether before they become so
formalized. What most people want is to be told either their future
or the cause of a malady, and this usually costs a rupee for a
villager, though maybe much more for a well-to-do visitor. To tell
the future (or the unknown past) is generally called <1s/a$stra kiyanava$>1;

196

I was told that there are four main methods: <1pe$na kiyanava$>1
(which is what the <1kapuva$>1 does), palmistry, looking at a man's
horoscope, and asking at what moment he left his house to come
and ask his fortune (we might subsume the last two under astro-
logy). To tell the future a <1kapuva$>1 becomes possessed (<1ma$yam
venava$>1), which he achieves by dancing on the spot, shaking the
timbrels, and breathing very rapidly; he tosses his head till his
long hair falls about his face. He then gasps out the answers. I saw
another method employed. When a village lady wanted to know
the cause of a disease the <1kapuva$>1 had a small palm-leaf manuscript
book with a long thread tied to one end like a bookmark; the
lady plunged this thread into the book at random while he held
it loosely shut, and he then opened it at that place and gave a

                                                                    197

villages some time in April, when the spring harvest (<1maha>1) has
been gathered in. Most informants held that this vow and
offering were made to Vanniya Ban*d*a$ra Deyyo$, though he was not
the village deity; but the headman whose list I cited above said
it was to all Twelve Gods. Some middle-class friends gave theirs
to Kataragama. The <1ad*ukku>1 themselves are a portion of rice which
is set aside; if offered in the morning, I was told, the rice is then
sent up to the temple for <1Buddha pu$ja$>1 at midday, but if offered in
the afternoon people just eat it themselves. People can give <1ad*ukku>1
individually for the success of their own crop, but the tradition is
to hold a collective ceremony at which the <1kapura$l*a>1 becomes
possessed and dances to each of the gods in turn to the accompani-
ment of drumming and chanting.<s70>s After dancing he tells the
future (<1pe$na kiyanava$>1). This type of ceremony, which locally is
to my knowledge held only for the Twelve Gods, is called "to
dance <1kemvara' (kenvara matanava$>1), and <1kemvara>1 in this sense is
used as a synonym of <1gam mad*uva>1<s71>s (verbally: <1gam mad*uva
nat*anava$>1). (This differs from Low Country usage, where a <1gam
mad*uva>1 is usually in honour of the goddess Pattini$ and is a much
more elaborate affair lasting several nights with a variety of dramas
and other rituals.) It takes place on a Tuesday or Friday night
I was told that only men are present and that some of the spectators
become possessed and dance too (my informant had done so
himself), which is ascribed to the gaze of the god falling on them
<1via>1 the <1kapuva$>1. This is the ideal picture of the ceremony, but it is
locally in decline and held only irregularly. Though occasionally it
still approximates to the ideal, the only one I saw was on a small
scale, half-way between a communal and a domestic rite: it was
held by an individual fanner in front of his house, and though
two <1kapura$l*as>1 took turns to get possessed, dance, and tell the
future, only about a score of male spectators came, while the
organizer's female relatives and neighbours watched from the
verandah. None of the spectators was inspired to join in.
  The object of an exorcism is to free the patient from the evil
influence (<1dos>1 or <1vas>1) which has fallen on him from the gaze

198

(<1ba%lma>1 or <1dr*s*t*iya>1, generally pronounced <1ba$luma>1 and <1dis*t*iya>1) of
some evil spirit or person. (The person need not be malevolent:
one can have the evil eye without knowing it.) All large-scale
exorcisms are somewhat akin to the <1kemvara>1 just described: a
specialist dances all night to the drumming and chanting, becom-
ing possessed by (or impersonating--the distinction is often
unclear) the spirits to whom this is principally addressed. Explana-
tions vary: the <1yaks*a>1 is forced or persuaded by the exorcist to
leave the patient, or banished by a more powerful <1yaks*a>1 or <1god>1
who is his master. Sometimes the patient becomes possessed too
and joins in the dancing.  The famous devil-dancing of the
Southern and Western Provinces, in which men of the drummer
caste exorcize demons of disease, are of this type. Another spec-
tacular type of ceremony in the Low Country is <1bali>1, which is
directed to appeasing the planetary deities.  However, devil
dancing is unknown in my part of the country, and <1bali>1 (usually
called <1tovil>1 or <1bali tovil>1) are rare: I do not think there had been one
in Mi$gala for three years. I saw a <1bali>1 ceremony in a nearby town,
but it was far simpler than I understood the Low Country cere-
monies to be. <1Bali>1 ceremonies feature an image of the planetary
deity, which may be huge, and is itself called a <1bali>1. A monk told
me that these are of three types: <1malbali>1 (made of flowers),
<1an#damabali>1 (drawn or painted), and <1am#banabali>1 (moulded, i.e.
three-dimensional figures).  A schoolmaster (not locally born)
said that the dances for <1yakku>1 and for <1bali (yakna%t*uma>1 and
<1balina$tuma>1) were the same, being differently named according to
context.  Often I found my informants on this topic unwilling
to make the distinctions which Ames has documented for the
Low Country; they obviously had little experience of these
matters.
   In view of this it is perhaps not untypical that the only lime-
cutting ceremony I attended in Mi$gala was in the house of an
immigrant from the Low Country, and performed by a <1yaka%dura$>1
from another village. The man, whom I shall call Perera, had been
in poor health and felt that he was resented as an outsider; he
suspected that malicious neighbours had practised black magic
and brought a <1yaka$>1 into his house. He got a <1yaka%dura$>1 to come from
some way away so that people--I suppose he meant his enemies--
should not know about it. The lime-cutting (<1dehi ka$pi$ma>1) was
to nullify the spell. Detailed descriptions of similar ceremonies

                                                                 199

have been published, so I shall only summarize. The exorcist
prepares indoors the offering (<1pide$niya>1) for the <1yaka$>1; the items in
it, which are minutely specified, are called <1pulut*u>1 (literally "burnt':
the word properly refers to the five kinds of parched grains which
are included) and include meat. He then takes up in a leaf a little
of each ingredient and leaves it, in great secrecy, at a place where
three ways meet. The <1yaka$>1 is presumed to have gone out for the
food, and then not to find his way back. Indoors a bowl containing
five oils is then put over a fire; Perera covers his shoulders with a
white cloth and sits on a mat; he is now "the patient' (<1a$turaya$>1).
The chair on which lies the rest of the <1pide$niya>1 is put next to him
with limes and an egg. The exorcist burns joss sticks before
pictures of the Buddha and gods, which are hanging on the walls,
and begins. Each lime in turn he holds against the patient's body
at some crucial anatomical point, or moves it backwards and
forwards between such points, and mutters very fast and low;
at the end of each spell he cuts the lime with cutters and tosses
the two halves into the sizzling oil. He uses twenty-one limes and
covers the whole body, then puts the cutters too into the oil.
Finally he takes the egg, passes it from Perera's forehead down
to his feet, cracks it on the ground and throws it into the pot.
The <1dosa>1 has all been drained out and burnt.
  The piece of meat included in the <1pide$niya>1 was the only element
in the rite which jarred with Buddhist principles, and after all,
men as well as <1yakku>1 eat meat. In the Low Country, exorcisms
sometimes involve the sacrifice<s72>s of a living victim (<1billa>1)--in
fact a chicken--to the <1yaks*a>1, and I did hear of one such case in
Mi$gala. My informant, a very poor and uneducated local man,
unfortunately would not tell me whether the patient too was a
local, as the matter was supposed to be somewhat secret. The
exorcist was again procured from outside the village, and the rite
took place in the cemetery. The patient had to give as <1bili>1 a cock,
a bottle of toddy and a bottle of arack. My informant was present
ad said it was all fake (<1boruva>1), for who but a man takes toddy
and arack?
  The only kind of magic which my informant believed in, he
told me, was the tying on of thread, because this was accompanied
by reciting the qualities of the Buddha. This thread is a simple

200

amulet, and the measure is a prophylaxis rather than a cure.<s73>s
It was often performed by a <1kapuva$>1, but I found out that my
informant, who was also high-caste, had tied on a thread himself
for someone who had been frightened by a <1holman>1. Besides the
thread some people supply amulets (<1yantra>1),<s74>s pieces of copper
foil on which are engraved linear diagrams with letters of the
alphabet and which are then rolled up and kept in a small cylinder
round the neck. Children usually wear one. The <1yantra>1, as well
as parts of the spells, are in Ma$laya$la$m, and are the same as those
used by the Tamils in the area. The spells (<1mantra>1) are used
when making and applying these amulets; normally a <1mantra>1
should be recited over a <1yantra>1 108 times, but further repetition
will make it stronger. I was told that besides Sinhalese and
Ma$laya$la$m they use Tamil and Telugu.<s75>s The word <1mantra>1 is
generally used only to denote spells of white magic; they begin
with at least a brief invocation of the Buddha (e.g. "O$m* Buddha$ya'),
tantamount to an acknowledgement of his supremacy over all
forces.
  White magic, however, usually implies the existence of black
magic to be counteracted; but to believe that black magic is
being used against you argues a certain degree of paranoia. The
villagers of Mi$gala were not much given to paranoia, and generally
conceptualized the evil influences to be counteracted as <1yakku>1.
I did not come across cases, common enough in the Low Country
where people were believed to have an evil eye or evil tongue, nor
did anyone ever tell me that he was the victim of black magic. I
would not deny the possibility of these things going on in Mi$gala,
but they are certainly not obtrusive. The only story of black magic
which came to my ear sounds a cheerful note. A group of women
disliked a lady teacher in the village. They approached a <1kapura$l*a>1
to put a spell on her, so that she would have some misfortune and
wish to leave. He asked 150 rupees (nearly a month's salary for a
junior teacher), 50 of it in advance. He then went to the husband of
the unpopular lady, who had on occasion helped him out finan-

                                                                 201

cially, and told all, saying that as they were good people no harm
would come to them, but he would like the lady to feign illness and
stay away from school. Accordingly while teaching at the school
next day she complained of a violent headache, and immediately
took three days' aick leave. The women were much delighted
with their quick results, and handed over the outstanding 100
rupees. In their glee they went--as would be customary--to call
on the sick lady, pretending to condole. But to their chagrin, after
three days she reappeared at the school, hale and hearty. They went
to complain to the <1kapura$l*a>1; at first he claimed to be nonplussed,
and questioned them, only to be told they had visited the victim.
No wonder she had got well! They had spoilt it all by going to see
her. Now some of it might even have rubbed off on to them. Would
they not like some protective amulets? ...
  Earlier in this chapter (p. 150) I listed the ways in which the
Sinhalese feels he can cope with misfortune, and control his en-
vironment; I have now placed these ways in their setting of belief
and ritual. Most of them have nothing to do with Buddhism
(<1Buddha$gamaya>1). But Buddhism does have some contact with
secular welfare, not through any attempt to propitiate the Buddha
himself, but through the use of sacred texts, the Buddha's word
(<1Buddhavacane$>1). Behaviourally viewed, certain canonical texts
are used as spells. However, for the Buddhists there is no in-
congruity in most of these uses; they are still quite explicable in
their own terms; moreover they are of great antiquity. The princi-
pal such use of texts is institutionalized in a ceremony called <1pirit>1.
  <1Pirit>1 is derived from the Pali word <1paritta>1 meaning "amulet'.
The word now denotes a prescribed text or set of texts (which may,
however, be abridged) recited in a particular ritual. This ritual
has certain features, shared with other Sinhalese rituals, to which no
reference is made in the texts recited and for which no rationale is
given. A reel of thread or string (<1pirit nu$l>1) is held by all the monks
reciting and passed round to enclose the spectators, any of whom
may also clasp it. At least one bottle of water (<1pirit pa$n>1) is kept on
or under the table on which the manuscripts of the texts (<1pirit pot>1)
lie during the recitation, and at the end of the ceremony a monk
sprinkles this water over everyone present. These features, not
rationally explicable in doctrinal terms, allow us to mention
<1pirit>1 here in the context of magic as well as in Chapter 5 in the
context of soteriology.

202

   There is to my knowledge only one text which has a specialized
application: the <1Angulima$la Paritta>1, for use in childbirth (on
which see Chapter 5, p. 224). All the other applications of <1pirit>1,
which are listed below, make use of a collection of Pali texts,
mostly <1suttas>1 from the Canon, made in Ceylon by the tenth cen-
tury<s76>s called the <1Catubha$n*ava$ra>1 ("Four sections of sermons').
The texts which make up this collection are listed, not in strict
order, by Rahula;<s77>s the list dictated to me from memory by a local
monk coincides with Rahula's in all but one particular which is
plainly a alip. The <1Catubha$n*ava$ra>1 begins with three texts, the
<1Mangala (alias Maha$mangala) Sutta>1, the <1Ratana Sutta>1, and the
<1Metta (alias Karan*i$ya-metta) Sutta>1, all of which are from the
canonical book called the <1Sutta-nipa$ta>1 and also from part of
the canonical chrestomathy, the <1Khuddaka-pa$tha>1. Two <1sil>1
<1na%$n*iyo$>1 who lived together near my village told me that whenever
possible they recite these three texts twice a day, at 5 a.m. ad
7 p.m. These texts, followed by verses called the <1Maha$ Jayamangala>1
<1Ga$tha$>1, constitute the <1Maha Pirita>1 ("Great <1Pirit>1') which begins
and ends long <1pirit>1 ceremonies, and is always recited by all monks
present. The other text which must be singled out for mention
is the <1A$t*a$na$t*iya Sutta>1 (from the <1Di$gha Nika$ya>1); this is always
recited twice and comes last, immediately before the final recitation
of the <1Maha Pirita>1, which means that its recitation falls at about 4
in the morning.
   <1Pirit>1 may be "said' by anyone (even a layman, if no monk is
available), and there is no limit to the number of participants.
During long ceremonies however the full complement usually is
present only for the <1maha pirita>1; most of the time the recitation
is carried on by two monks (<1yuga pirita>1) or four monks, who work
in two-hour shifts. Continuity must be preserved, not necessarily
by an uninterrupted flow of sound, but by the continuous presence
of at least one monk in a preaching seat holding the thread.

                                                                      203

  A <1pirit>1 ceremony using the full range of texts (i.e. the <1Catu-
bka$n*ava$ra>1) may last all night (<1sarvara$trika pirit>1), which in practice
means roughly from 9 p.m. to 6 a.m.; for thirty-six hours (<1tun
tis pa$ya pirit>1), roughly from 9 p.m. to 6 a.m. a day later; or for a
week (<1sati pirit>1),<s78>s as happened when the whole world seemed in
danger from the fearsome conjunction of the planets. (At the end
of a week's <1pirit>1 there is a ceremony called the <1dorakad*asne>1.)<s79>s A
standard form of shorter <1pirit>1 is <1varu pirit>1, which is said for an
hour in the evening before the monks' <1gilampasa pu$ja$>1, an hour
before breakfast (<1hi$l da$ne$>1) the next morning, and an hour the
next evening, i.e. roughly at 5 p.m., 5 a.m. and 5 p.m. However,
<1pirit>1 can come in much shorter chunks: Radio Ceylon began every
day with about fifteen minutes of it, immediately after the Three
Refuges and Five Precepts, broadcast at 6.30 a.m. However, <1pirit>1
by radio obviously becomes a totally different affair, as the thread
and water are missing. It is, probably consciously, a reversion to
canonical practice, in that the texts are used, theoretically at least,
for the meaning they convey; for <1pirit>1 originated as a kind of
preaching--to a supernatural audience.
  The use of water in this connection is, however, very old. In
the reign of Upatissa I (end of the fourth century A.D.) Ceylon

... was afflicted by famine and disease. The king inquired from the
Sangha if anything was done by the Buddha in such a situation to
alleviate the suffering of the people, and the monks described to him
how the <1Ratana-sutta>1 was recited by the Buddha when Vesali was
visited by such a calamity. Thereupon the king had a golden image of
the Buddha made, and placing in its hands the Buddha's stone alms-
bowl filled with water, mounted it on a chariot. Then he organized a
great almsgiving and ordered the citizens to observe the moral precepts
(<1si$la>1), himself observing them.
  The city was beautifully decorated, and a large crowd of monks
following the chariot with the golden Buddha image walked the whole
night round the streets reciting the <1Ratana-sutta>1 and sprinkling water.
The king himself took part in the ceremony, walking with the monks.
Rains came and famine and pestilence disappeared. Upatissa decreed
that this ceremony should be performed whenever there was a similar
calamity in the Island.<s80>s

204

The Ratana Sutta is still one of the texts comprising the <1maha>1
<1pirit>1. In a footnote to this passage Rahula gives the story of the
Buddha's alleged use of <1pirit>1.

When Vesa$li was afflicted by famine and pestilence, the Buddha
visited the city on the invitation of Licchavis, and recited the <1Ratana-
sutta>1. (It is included in the <1Khuddakapa$t*ha>1 as well as in the <1Sutta-
nipa$ta>1.) The Buddha first taught this sutta to Ananda and requested
him to go round the city accompanied by Licchavi princes, reciting the
sutta and sprinkling water from the Buddha's almsbowl. The city was
saved from the calamity. A great festival was held in honour of the
Buddha's visit. Two boats on the river were joined together and a
pavilion was built thereon. After this successful mission to Vesa$li,
the Buddha returned to Ra$jagaha along the Ganges. This journey is
called Ganga$rohana, and the name was given to the festival itself.
(SnA. pp. 204-5; CBhA. pp 97 ff)<s81>s

This story comes from Buddhaghosa's commentaries; the use of
water has no canonical authority. Rahula also quotes (p. 279) a
description from the commentary on the <1Di$gha Nika$ya>1 of how
<1pirit>1 should be recited for a sick man. He then says that "...
the <1pirit>1 ceremony as we know it today can be seen only after the
Pol*onnaruva period'.<s82>s I am not sure what he is referring to.
The description in the <1Di$gha Commentary>1 does not mention water
or thread, but adds other minor non-canonical features, now
obsolete, such as that the monk reciting the <1A$ta$na$tiya Sutta>1
should not have eaten meat or flour. <1Pirit>1 water, however, we have
just seen to be ancient; for <1pirit>1 thread I know of no reference
unambiguously connected with a ceremony, but in a famous
<1Ja$taka>1 story Prince Telapatta "had a <1pirit>1 done by <1paccekabuddhas>1,
and took the <1pirit>1 sand and <1pirit>1 thread' (<1accekabuddhehi parittam*
ka$ra$petva$ parittava$likam* c'eva parittasuttakam* ca a$da$ya>1),<s83>s and
later when in danger sprinkles the sand on his head and ties on
the thread.<s84>s For me tghis is evidence enough that no important
feature of a modern <1pirit>1 ceremony post-dates commentatorial
times.
  There is frequent reference in the Canon to monks being dis-
turbed in their meditations in forest solitude by troublesome
<1yaks*as>1.  The beginning of the <1A$t*a$na$t*iya Sutta>1 (D.N. xxxii) refers

                                                                      205

to this circumstance; Vessavan*a, king of the <1yaks*as>1, comes to the
Buddha saying that <1yaks*as>1 trouble monks in the forest and recites
to him a long poem consisting largely of the names of various
spirits, to keep them at bay; the Buddha then repeats it to his
monks. There are several other verse texts which, though they lack
a similar introduction, can be deduced to have been intended from
the first to serve a similar purpose. The commentary says as much
of the <1Karan*i$ya-metta Sutta>1<s85>s which preaches universal love and
good-will; and it is probably significant that the <1Ratana Sutta>1,
which has a similar purport, begins by addressing "Whatever beings
of the earth or sky are here assembled'. In Buddhist terms, the
<1yaks*as>1 are being preached to, converted to Buddhist benevolence
so that they will no longer want to hann anyone. (The Rev.
Rahula contrasts the Christian expulsion of devils with the
Buddhist way of winning them over by the exercise of loving-
kindness.)
  I was present at several <1pirit>1 ceremonies, but as descriptions of
<1pirit>1 have been published<s86>s I shall refrain from reproducing my
own. It will suffice to remark on the extremely unspecific nature of
the ceremony, which we can see from Rahula's work to have been
the case in ancient times also. <1Pirit>1 is used at a sick bed, to com-
memorate a death, to consecrate a new building, to avert a public
misfortune, to celebrate the opening of Parliament, or simply to
acquire merit.
  The canonical explanation of <1pirit>1 is that it converts, or at least
mollifies, the <1yakku>1. As a variant, when the Buddha said <1pirit>1
on behalf of the boy Di$gha$yu there was such a press of gods that

206

the <1yaks*a>1 who was due to eat the boy could not get in.<s87>s Another
explanation, also with some canonical authority, is the general one
that it is meritorious to recite canonical texts or have them recited.
The recitation of texts is always considered equivalent to preach-
ing (<1ban*a>1), and is listed (as <1desana$>1) among the Ten Good Deeds
(<1dasa kusala karma>1) (see Chapter 2, p.  74).  Moreover,
listening to preaching (<1suti>1) is similarly listed, being considered
the equivalent of "causing to preach'. This rather formalistic
attitude to the sacred texts clashes with the Buddhist doctrine of
intention (<1cetana$va>1) (see Chapter 6). What is here more to the
point is that this explanation does not fit the case of <1pirit>1 recited
(as so often) to allay an already present misfortune. To earn merit
in any way, including this, will improve one's <1karma>1, but <1karma>1
is a long-range affair, and there is no reason why the merit just
gained should take immediate effect, so as to make a sick man well.
To say that <1pin>1 can cancel out <1pav>1 is in fact a heresy, discussed at
the beginning of Chapter 5. Indeed, if one is thinking in terms of
<1karma>1 the presumption must be rather the other way: if a man is
ill because of a past sin he will go on being ill till the sin is expiated,
despite any <1ad hoc>1 remedial action. In the case of misfortune, to
explain <1pirit>1 as "merit in a hurry' will therefore not wash.
This line of explanation <1is>1, however, applicable when <1pirit>1
is used at a celebration--indeed it sometimes fits the case far
better than the explanation involving <1yakku>1. On the occasion of his
fiftieth birthday the chief monk of Mi$gala temple had <1pirit>1
chanted for thirty-six hours and gave alms to fifty monks, includ-
ing the Maha$na$yaka of Malvatta. He was of course not trying to
dispel any <1yakku>1, but simply making himself a birthday present of
a vast quantity of merit.<s88>s To be sure, there is no reason why the
texts used should have been <1pirit>1--any <1ban*a>1 would have served
the same purpose--but he used the institution which was at hand.
(The celebration of his own birthday--with great advance pub-
licity--strikes me as flagrantly untraditional conduct in a monk,
but was well thought of by the villagers, who gained both the
sight of a festivity and the chance to acquire merit by sympathy.)
What <1pirit>1 is in public, the "<1Itipi so' ga$tha$>1 is in private. The

                                                                      207

"<1Iti pi so' ga$tha$>1 is a short prose<s89>s formula, beginning with the
words "Iti pi so', which eulogizes the Buddha. There are similar
formulae eulogizing the Dhamma and the Sangha, which are
often recited in succession; the three together form a statement of
faith somewhat analogous to the Christian Creed. These Pali
sentences occur frequently in the Canon.<s90>s The "<1Iti pi so' ga$tha$>1
is also known colloquially as <1Budugun*a>1 (or <1Buddhagun*e$>1), "the
qualities of the Buddha'. The <1Budugun*a>1 are the most frequently
employed spell in white magic. Someone frightened by a ghost will
recite the <1Budugun*a>1 and thus keep the ghost at bay, even if he has
to repeat them all the way home. The <1Budugun*a>1 are highly
recommended in Mi$gala. The ex-headman said that in an <1apale$>1
(astrologically dangerous period) one should just recite the
<1Budugun*a>1,  and not bother about <1bali-lovil>1 and such ceremonies.
The old <1kapura$l*a>1 (K 1), who did quite a bit of business with
amulets, etc., himself said that the <1Budugun*a>1 were the best guard
against <1yakku>1, while the most effective cure for any ill is to recite
the <1Maha$mangala>1,<s91>s <1Ratana>1, and <1Karan*i$ya-metta Suttas>1 morning
and evening.  Moreover, when I asked him for some <1yantra
mantra>1 (apotropaic spells), the first one he recited to me was the
Pali

            <1Sabbapa$passa akaranam* kusalassa upasampada$
            Sacitta-pariyodapadam* etam* Buddha$nusa$sanam*>1

to be repeated three times, with Dhamma and Sangha substituted
for Buddha at the second and third repetitions. The verse means,
"To abstain from all evil, to do good, to purify one's mind, this is
the teaching of the Buddha/Dhamma/Sangha.' It is a famous
verse from the Canon,<s92>s alightly misquoted: in the original it ends
<1Buddha$na sa$sanam*>1, "the teaching of the Buddhas'. The reason for
this misquotation is probably that in its original form the verse
would require more change to accommodate the Dhamma and the
Sangha, a variation which patterns the verse on such familiar

208

models as the Three Refuges (pp. 64-5). Examples of similar verses
in spells could no doubt be multiplied by research.
The question again arises whether the <1Budugun*a>1 and similar
Buddhist formulae are mere spells, or whether their use can be
rationalized: affectively they are surely a spell, but does their
use involve a <1cognitive>1 inconsistency? Perhaps not, because to
recite the Buddha's qualities is to remind the evil spirits of his
power and goodness, so that they will lose the will to attack.
According to one monk, the Buddha himself recommended the
practice, saying that even S/akra (who is addressed in the <1Dhajagga
Sutta>1, one of the best known <1pirit>1 texts) has passion, hatred and
deluaion (<1ra$ga dosa moha>1), "so if you want anything just think of
me (<1mama sihi karaganta>1)'. If the qualities of the Buddha fail,
recite the qualities of the Dhamma; if that still fails, those of the
Sangha. Recital of the Three Jewels will get rid of all fear and
danger. This explanation touches on the belief in the power of
truth, which will be discussed in the next chapter.
However, Obeyesekere has told me of a twist in the use of
<1Budugun*a>1 which defies even this type of explanation. Among
villagers out on a hunt he heard the <1Budugun*a>1 muttered by a man
to prevent his rival from catching anything. The qualities of the
Buddha prevent violence--even when used with ulterior motives!
  Having found Buddhist texts used as spells, it will be less
surprising to find a relic used as a talisman. I did not come
across a case in practice; but I was told that someone who has
a Buddha relic on his person is invincible at cards! But lest this
be taken for a modern degeneration, let me refer to <1Maha$vam*sa>1,
XXV, 7, which tells us that Dut*uga%mun*u had a relic put in his
spear.<s93>s Nothing reported by any ethnographer in modern times
approaches this flagrant violation of both the letter and the spirit
of the doctrine.
The use of spells and talismans characterizes what we are calling
"magic'. There is a gap here between our categorization and that of
the Sinhalese, because where the spells are Buddhist texts or the
talismans, Buddhist objects, the practices with which they are

                                                               209

associated are generally considered to be part of Buddhist religion
(<1Buddha$gamaya>1); in fact it is largely these practices which
Buddhist modernists call Buddhist religion (<1Budda$gamaya>1), as
opposed to Buddhist philosophy (<1Bauddha dars/anaya>1) which
they conaider the true, original Buddhism. While in this chapter
I am mainly concerned with statements by Buddhists which inter-
pret <1pirit>1, for example, as something other than magic, its re-
semblance to magic will strike the western reader; I will return to
it in Chapter 5. At this point it is sufficient to point out that magic
<1per se>1 is not contrary to Buddhist doctrine, provided it serves only
worldly ends. In this respect it is on a par with gods and other
spirits; a picture of the world which includes these forces, and
considers a spell of similar efficacy to a medicine or a poison, is not
logically incompatible with a belief in <1karma>1.
   Black magic is contrary to Buddhist ethics, not to Buddhist
metaphysics: a sin, but not a heresy. White magic to secure
worldly ends is contrary to neither. A monk who practises black
magic is doctrinally on a par with one who drinks; a bad Buddhist,
if you like, but bad in the sense of wicked, not of inconsistent.
A monk who says <1pirit>1 to cure sickness, whatever may be his theory
to explain its efficacy, is a good Buddhist in every sense.
   Although we have come across certain beliefs and actions which
cannot be reconciled with the doctrine of the Canon, I would
suggest that these are somewhat trivial and peripheral. They are
also ancient; changes since ancient times even seem, oddly
enough, to have been more towards than away from orthodoxy.
In the ninth century King Sena II "had the Ratana Sutta written
on a gold plate and made offerings to it';<s94>s there is nothing like
that today. Finally, are the "deviant' practices so thoroughly
unjustifiable? Only on one level. All can be explained by the
Buddhist doctrinal belief in the power of human thought, to
which we shall have frequent cause to refer in the next chapters.
As a simple traditional monk remarked to me, diseases are usually
curable by medicine or by <1bali>1, in which they use the <1Budugun*a>1;
the reason one gets better is because of one's own ideas (<1tamange$
adahasa>1). Another monk said the same of all worship of gods, and
specifically of <1pirit>1: it will only do you good if you have faith
(<1visva$sa>1); <1you are protected by your own belief>1.

210


                           IV. SUMMARY

    There are few or none zealous in their worship, or have any
    great matter of esteem for their Gods. And they seldom busie
    themselves in the matters of their Religion, until they come
    to be sick or very aged. They debar none that will come to see
    the Ceremonies of their worship; and if a stranger should
    dislike their way, reprove or mock at them for their Ignorance
    and Folly, they would acknowledge the same, and laugh at the
    superstitions of their own Devotion, but withall tell you that
    they are constrained to do what they do, to keep themselves
    safe from the malice and mischiefs that the evil spirits would
    otherwise do them, with which they say, their Country
    swarm... .
      It is a usual saying, and very frequent among them (if their
    Geraha [<1graha>1], which is their fortune, be bad) What can God
    do against it: Nay, I have often heard them say, Give him no
    Sacrifice, but shit in his Mouth, what a God is He? So slight
    an estimation have they of their Idol-Gods; ...
                                                KNOX, p. 132

I have attempted to present the subject matter of this chapter
analytically. To give a clear picture of local beliefs in a brief
synthesis let me reproduce the opinions of the three men in
Mi$gala whom I knew best. Our neighbour, a very simple and
uneducated man, had seen the <1bali>1 with cockerel and arack, and
said it was all false. Similarly, he said, all the stuff done by the
<1kapura$l*a>1 (K 1) was false (<1boru>1). These <1kapuvo$>1 ask for money, but
why should a god need money? I asked if he believed in Pit*iya
Deyyo$. He did not answer me directly--plainly he took Pit*iya
Deyyo$'s existence for granted--but replied with a vehement
assertion that the gods need nothing but merit. I asked about
tying on thread. No, that is not false, because that is <1Buddhagune$.
Yakku>1 only frequent the cemetery; if you go past there at mid-
night or twilight you may get a <1ba%luma>1 and get scared, and tying
on the thread cures you of this fright. I found out that he himself
had tied thread for someone so frightened. "It is all in the mind
(<1okkoma hite$>1)'; he meant not that <1yakku>1 do not exist but that they
only trouble those in a receptive state of mind. He too has seen
<1holman>1 near the cemetery, and heard their noise, but nowadays he
is never troubled. I ask if <1holman>1 are people who sinned in previous
births. No, <1holman>1 are rightly called <1yakku>1.

                                                              211

   Another villager whom we got to know very well, a share-
cropper of very little education and no better than he should be,
struck me as the perfect "man in the street' of Mi$gala. He did not
go to the temple often, and had very few opinions on religious
topics, but he did tell me that he doesn't go in for this <1yaks*a>1
stuff. When he was seriously ill he did not get a <1yaka%dura$>1 (exor-
cist), but had monks say <1varu pirit>1. On the other hand when his
son was ill he made a vow to Pit*iya Deyyo$, and gave the god a
small weapon when his son recovered. Some people are frightened
of the dark and hear footsteps behind them, especially near the
cemetery,  but he has never seen anything [and he lived fairly
near it]. Some people in Mi$gala have met Mahaso$na$ there and
fallen ill of fright; a female couain of his even died of fright one
night; a year after her marriage she and her husband were asleep
in the house when at midnight a bull lowed out at the back; when
they went out they heard footsteps and her husband ran to look,
but found no one. One may meet Hu$niyam, but he is benevolent.
A figure of a white naked man is a <1holman>1; his wife is frightened of
these. Once she went out to the latrine at night and screamed to
him that she heard footsteps, so he went out; there was nothing
there, but he pretended to be shooing away a dog and so her
fear was appeased.
  Our landlord, a small cultivator in his fifties who had been a
bus-driver, was in most respects a typical villager, but held some
unusual opinions. It is possible that he had been influenced by
spending some time at a mission school when a boy, but I would
be inclined to discount this, as the same experience has not had
the same effect in the other cases with which I am acquainted.
He considered himself a devout Buddhist, but said he believed in
no gods at all--it was all lies (<1deyyo$ okkona boru>1). One of the
village shrines was robbed of all its cash, and this caused him end-
less amusement, for, he said, if the god was so powerful why did
he not stop the thief? He also refused to observe any of the tabus
connected with the Sinhalese New Year celebrations and his wife
served us a cooked meal at the time when the fire is supposed to be
extinguished. He had been to Kataragama "for fun' (<1vino$de$t*a>1), but
Kataragama too was "lies'; it was all for the <1kapu>1 to make money:
they rake in the coins (<1pan*#d*uru>1) in the morning and drink toddy
in the afternoon. Why should gods need money? One doesn't give
rupees to the Buddha.

212

   Such extreme scepticism, is, however, unusual. To conclude
this chapter I reproduce my notes, taken almost on dictation,
from an elderly monk who gave me in re/sume/ a classic exposition
of the Sinhalese view of the world--an anthropologist's dream.
   He has never been to Kataragama--people go there mainly for
vows (<1ba$ravalat*a>1), not to worship (<1van#dint*a>1). Kataragama was not
a man but a god by birth (<1utpattiyen deyyo$>1) and Ceylon is his
sphere of responsibility (<1ba$raman*d*ale$>1).
   The gods have a system of government like the human world.
S/akra is the chief of the gods, and he has a retinue of ministers
(<1ama$tyapirisa>1), the four gods (Dhrtara$s*t*ra, etc.) who live in the
Ca$turmaha$ra$jika heaven. Below them are gods in charge of
various countries. Ceylon is in the charge of four gods: Vis*n*u,
Kataragama, Saman, and Vibhi$s*an*a. Each rules a quarter of the
country; this part is under Saman. Pattini$ is all over Ceylon in
charge of one department (<1eka am*se$t*a pradha$ne$>1). Da%d*imunda is
fierce (<1can*de>1)--a <1yaks*a>1. Below these come the gods in charge of
villages (<1gamba$ra deyyo$>1) [here the monk wrinkled his nose], under
the former as D.R.O.s are under G.A.s.<s95>s'Those fellows are the
bosses of tiny little parts (<1E$ aya pod*i pod*i kot*asala lokko$>1). 'The
local D. R.O. equivalent is Pit*iya Deyyo$.
   There are three classes of beings:  <1deyyo$, yaks*ayo$, pre$tayo$>1.
The latter two groups are ruled by Vessavan*a, who is the boss of
the <1yaks*as (yakunne$ loka$>1). He is very tough (<1sa%rayi>1), and if someone
breaks his laws he looks at them in a certain way and his gaze
burns them to powder (<1a%sdrs*t*iya pit*i pit*i Piccenava$>1).<s96>s Gods are
pleased by the Five Precepts (<1pan sil>1) and <1pinkam>1, and like good
people; <1yaks*as>1, on the other hand, break all the five precepts and
like wicked people.  But they cannot harm the good people,
because they are scared of Vais/ravan*a.  <1Yaksayo$>1 and <1Pre$tayo$>1
live on smelly foods like meat; to get people to make offerings to
them they show them <1holman>1, which are just their own forms
<1svaru$pa>1) apparitions (<1avata$ra>1); these <1holman>1 frighten people
into making offerings. However, they may be exorcized by the
recital of the Three Refuges (<1un saran*a>1), at which <1yakku>1 and
<1Pre$tayo$>1 will flee for miles.
   Astrology (<1na%ka%t s/a$stre$>1) can tell only trends and dispositions,
and give auspicious and inauspicious times. To say that it can

                                                                 213

predict the future in detail is contrary both to truth and to
Buddhism.
  The planetary deities <1(grahayo$)>1 are not a class of <1yaks*a (yakku>1
<1vargayak);>1 they are ways of effecting the good and bad <1karma>1
which a man is born with. In other religions gods make creatures
--but who makes the gods? <1Karma>1 alone determines a man's
birth, but <1karma>1 is long-term <1(diga);>1 it does not determine the
details of one's life, but only those things which can't be helped--
principally the condition of one's birth. To escape particular
misfortunes may be possible by e.g. <1bali>1 ceremonies, in which the
planetary deities are propitiated <1(s/a$nti karanava$). Karma>1 is purely
the result of one's own efforts and thoughts.

   "All that we are is in the result of what we have thought', says
the first verse of the <1Dhammapada>1.<s97>s To test whether this doctrine
is still consistently upheld will be the purpose of the next chapter.

                                5
                  TOTAL RESPONSIBILITY IN THEORY

                           AND PRACTICE

AT the beginning of the previous chapter was stated the theory
of <1karma,>1 according to which a man's destiny depends on his own
efforts. However, in the course of the chapter were mentioned
certain beliefs which seem <1prima facie>1 to contradict this. In this
chapter we shall examine these beliefs and scrutinize their con-
gruity with stated doctrine. Our principal topics are the doctrine
of superseded <1karma (ahosi kamma),>1 the religious wish <1(pra$rthana$)>1
and the transference of merit <1(patti>1 and <1patta$numodana$).>1
  <1Ahosi kamma>1 is a Pali technical term, literally meaning "has-
been <1karma'.>1 It refers to <1karma>1 which is superseded by the course
of events. Though logically it could apply to deeds good or bad,
the term is in fact used only in the context of <1bad karma>1 which
does not come to fruition. That even good <1karma>1 may not bear
fruit is logically implied by the very common expression 'Pin
siddha ve$va$' ("May the good deed be successful') which is said
by the recipient of any good action, the monk who is reverently
greeted or the beggar who gets a coin. (The saying could also be
given as an example of the type of religious wish which I categorize
below (p. 218) as a "mere expression of benevolence'.) However, I
here discuss only <1ahosi kamma>1 in the sense of superseded bad <1karma.>1
   If someone attains <1nirva$n*a>1 he will not be reborn, so much of his
<1karma>1 will never bear fruit and becomes obsolete <1(ahosi venava$).>1
Even an <1arhat,>1 however, can never escape the consequences of
(i.e. <1ahosi karant*a)>1 one of the five great sins: killing mother, father,
or an <1arhat,>1 causing schism in the Sangha, or shedding the blood
of a Buddha. The venerable Mugalan, one of the Buddha's two
chief disciples, suffered terribly at the hands of robbers long after
he had attained Enlightenment.  They surrounded his house,
threatening to murder him, but he got out by his supernormal
powers <1(iddhi).>1 The same happened a second time. When he was in
the same situation for a third time<s1>s he examined his past lives

                                                                   215

with the eye of wisdom <1(n+a$n+acakkhu)>1 which all <1arhats>1 acquire,
and realized that this was the result of his having killed his
aged parents in a former life; so he submitted to his fate and was
beaten up by the robbers who left him at the point of death.
This story from the commentary to the <1Dhammapada>1<s2>s was told
me by a monk. The circumstances under which he had killed his
parents were these. They were very old and blind, and he was
living with them and his wife. The wife said his parents were
such a nuisance that he must choose between them and her. To
deceive him she would throw the scum of the boiling rice on to
the floor and then tell him his parents were spitting everywhere;
similarly she simulated fallen white hair. Finally the son pretended
to his parents that he had to take them on a journey through a
dangerous forest. On the way he suddenly exclaimed that robbers
were attacking. His parents told him to flee and save himself; he
ran off and returned pretending to be a robber, and killed them.
For this he came to be brutally murdered, <1arhat>1 or no.
   I questioned a monk in Mi$gala about <1ahosi kamma.>1 By doing
some great act of merit, he said, one may anticipate the maturation
of one's <1karma (karmavipa$ka)>1 and reach <1nivan>1 first. There are
four kinds of <1karma>1 :
   (1) <1dr*s*t*adharmavedani#ya>1: will bring results within this life
<1(me$ a$tme$ma vipa$ke$ denava$).>1
   (2) <1upapadya:>1 will bring results in the next life <1(ilanga a$tme$ di$
vipa$ke$ dendva$).>1
   (3) <1apara$pariyavedya:>1 there is no limit--will bring results
when opportunity arises <1(sima$vak na%$--avastha$vak la%buno$t vipa$ke$
denava$).>1
   (4) <1ahosi:>1 disappears <1(na%tivenava$).>1
   For example, Dut*uga%mun*u killed many Tamils in war, which
is <1pav,>1 but he did it to save Buddhism, and then did so much for
Buddhism (founding monasteries at Anura$dhapura, etc.)  that
his <1pin>1 so far outweighed his <1pav>1 that he will stay in heaven
<1(divyalo$ke$)>1 till the time of Maitri$, the next Buddha, when he will
be reborn as his right-hand disciple <1(dakun*at sav,>1 i.e. the equivalent

216

of Sa%riyut for Gotama Buddha) and attain <1nirva$n*a.>1 His <1pav>1 will
therefore never mature, there being no results <1(vipa$ke$)>1 of bad
<1karma>1 in heaven.
That my informant gave Dut*uga%mun*u as his example is no
accident, for apart from <1arhats>1<s3>shis is the only specific case of <1ahosi
karma>1 ever heard of. The prediction concerning him is contained
in <1Maha$vam*sa,>1 XXXII, so my informant was fully justified in
saying that Dut*uga%mun*u would never pay for the sin of killing.
Even so the monks I interviewed were divided on this question:
some concurred in saying that Dut*uga%mun*u would get off while
others asserted that at some point he must surely pay for his sins.
Although all the stories deal with the extreme case, it is not
merely the attainment of <1nirva$n*a>1 which can prevent the matura-
tion of sin. Common sense suggests that it may not be possible to
reap the recompense for every action. As an old <1upa$saka>1 put it to
me, though a man commit fifty murders he can only be hanged
once. This was just an analogy fitted to my western understanding ;
if he is constantly reborn he can of course be hanged fifty times;
yet if he persists in so murderous a course it becomes impossible
for him to pay for all his crimes. This is the point of the above
classification given by the Mi$gala monk quoted above, as inter-
preted by his colleague. There are some sins [these would be the
lesser ones] which are superseded <1(ahosi venava$)>1 if you do not pay
for them in this life or the next--"there is a time limit' ("Ka$lasi$-
ma$vak tiyenava$'). Some on the other hand may crop up any time
there is a chance. Another analogy: if he (the speaker) wants to do
me harm while I am in Ceylon but gets no chance before I leave
for England, the matter lapses <1(ahosi venava$).>1 [But killing is very
bad, so] we cannot say about Dut*uga%mun*u; finally he will be
reborn as Maitri$'s right-hand disciple, but we do not know what
lives he will pass through before then.
I have presented <1ahosi kamma>1 in a rather scholastic light for
this was the only light under which it ever appeared to me. I
never came across it as part of behavioural or "actual' religion,
and most laymen have no idea what it is. Obeyesekere, however,
has come across a popular variant which by a small change runs
clean counter to canonical doctrine. If sufficient merit can ensure

                                                                  217

that one will never pay for sins, it is not a big step to saying that
merit can cancel out sin. Obeyesekere has met this argument in a
remote area where villagers subsist partly by hunting, thus
flagrantly violating the first precept not to kill; they hope that by
becoming good <1upa$sakas>1 in old age they can nullify the effects of
their past sins. This is a clear case of religious change on the
cognitive level; its intellectual origin and psychological motivation
are clear, but it is not, I think, widespread, and so is of limited
interest. On the other hand it seems close--just a step over the
line into doctrinal unorthodoxy--to the belief in the power for
good of the death-bed wish, discussed in the next section.

  Far more important and problematic, because a common
feature of the life of every Buddhist, is the <1pra$rthana$va>1 or religious
wish. The same word is used by Sinhalese Christians to translate
"prayer'; I have avoided this translation in the Buddhist context,
as it would prejudice a balanced exposition. The term "prayer'
implies a person who could answer it, and suggests that he could
intervene in the course of events, thus interfering with the
suppliant's <1karma.>1 Neither implication is justified on the cognitive
level, though affectively the <1pra$rthana$va>1 does have some affinity
to prayer of a very particular type. If a tranalation is required it
would be best to use some such phrase as "earnest wish', which is :
indeed the primary meaning of the Sanskrit word and of its Pali
equivalent <1patthama$>1.
   <1Pra$rthana$>1 are made at the end of every <1pinkama.>1 Most com-
monly they occur at the end of a sermon (however short): the
monk expresses the hope that all beings may attain <1nirva$n*a>1 (e.g.
<1siyaludena$t*ama>1 or <1siyalu sattvayant*a nivan sa%pa la%be$va$)>1 at which
remark everyone present raises their folded hands to or above the
head, if they are not there already (for this is the posture in which
the pious listen to sermons), and exclaims, "Sa$dhu sa$dhu sa-',
exactly like a Christian congregation saying "Amen' at the end of a
prayer. Notice that in this context the <1pra$rthama$va>1 is always
exactly the same, namely to attain <1nirva$n*a.>1 The same <1pra$rthana$va>1
is used at funerals, where <1nivan sa$pa la%be$va$>1 concludes the eulogies
of the dead person which are printed and handed round, or at
poor funerals merely recited, and the same words are printed on a
banner at the entrance to the cemetery and perhaps elsewhere on
the funeral route. In this context the <1pra$rthana$va>1 is closely parallel

218

to the Christian <1Requiescat in pace,>1 with the difference that peace
is requested for the Christian immediately in the grave, long
before the Day of Judgement, whereas <1nirva$n*a>1 is the Buddhist's
long-term goal, certainly to be attained after further rebirths.
Under Buddhist influence Ceylon Catholics have turned to consider
their own long-term goal, and instead of R.I.P. their funeral ban-
ners read "<1svarge$ sa%pa la%be$va>1'--"May he attain the bliss of Heaven'.
Another common occasion for <1pra$rthana$>1 is in letter-writing
and here I wonder whether there has not been western influence.
It is customary to conclude a personal letter with a formulaic
religious wish, usually to the effect that the recipient should
find refuge in the Buddha or in the Three Jewels (e.g. "teruvan
saran*a la%be$va$ kiya$ patami'). There is nothing particularly un-
canonical about hoping that someone will find refuge with the
Three Jewels, in which he probably takes refuge daily, but these
formulae are not used outside letter-writing, and the practice
looks as if it is modelled on the Christian writing "May God be
your aid' or some such benediction.
The <1pra$rthana$>1 already discussed are made for others, which
makes it easy to explain them (as is done) as mere expressions
of benevolence. But the remaining ones are made for oneself,
which puts them in a different category, constituting a more
serious doctrinal problem. If one posits, as I think one must
that <1pra$rthama$>1 arose from an emotional need, one would expect the
<1pra$rthana$>1 for oneself to occur earlier than the <1pra$rthana$>1 for
others, and we shall see that such is indeed the case, only the
former occurring in the ancient literature.
Our account above of the <1pra$rthana$>1 for <1nirva$n*a>1 was not quite
complete. If the concept of <1pra$rthana$>1 is mentioned to a villager
he immediately thinks of the actual form which <1pra$rthana$>1
usually take on public occasions. Only modernist monks confine
themselves to the bald wish that all may attain <1nirva$na.>1
Traditionally (see for example the <1pin amumo$dana>1 in the Appendix)
the form of such a wish is for the monk to hope that the donors,
or all present, or those for whose sake the ceremony is taking place,
may in their future lives enjoy human and divine bliss (i.e. be
reborn as well-off men or as gods) till they are reborn in the time
of Maitri$, and <1then>1 attain <1nirva$n*a.>1 All then say "Sa$dhu', and by their
assent make this <1pra$rthana$va>1 for themselves.  It is only a few
fundamentalists who eschew this formula and condemn this

                                                                        219

postponement, urging the necessity of striving for <1nirva$n*a>1 here
and now. Their personal opinions deserve respect, but we shall
see that even Buddhaghosa would have earned their censure.
    The <1pra$rthana$>1 for oneself may have originated with the dying
wish, which--as we shall see below--is referred to in the Pali
Canon. There is a belief that if a person on his death-bed aspires
to be reborn in a particular state, his wish may well take effect.
At first sight this may seem to run clean counter to the doctrine of
<1karma;>1 but on the cognitive level this is not so. A man's dying
thoughts, so the argument runs, are necessarily conditioned by the
life he has led, so that to say that they determine his fate after
death is only to say that this fate is determined by his conduct in
this life. Unless his life has been pure he will not be capable of
making a <1pra$rthana$>1 for <1nivan,>1 but will aspire to some lowly form
of existence.  Conversely the good man on his death-bed will
remember some former good deed of his which will help him to
aspire to a good rebirth. Though this is doctrinally shipshape it
sails very close to the breeze: a village girl told me that the man on
his death-bed "remembers according to his luck' <1(tamange$ va$sana$va
anuva matak vemne$),>1 and she had introduced this topic in the
context of <1akosi karma;>1 though "luck is <1karma'>1 (p. 145) the
connotation is different, and the clear implication of the girl's
exposition was that a "lucky' man on his death-bed will recall an
act of merit and thus get a good rebirth, causing a host of minor
sins, which but for his dying thought would have carried more
weight, to become <1ahosi>1--inoperative.
   Affectively here is an attempt to mitigate the rigour of <1karma>1
by the possibility of a death-bed--not conversion exactly, but
shall we say sudden improvement. From this situation (if I am
right--I doubt whether the matter can be proved or disproved)
there developed long ago the more general practice of making
<1pra$rthama$>1 for oneself, ideally to attain <1nirva$n*a>1 but in fact probably
for all kinds of satisfaction. To what extent <1pra$rthana$>1 are made,
and to what end, I cannot say, as they are for the most part a
private affair, nor need they be made out loud. The usual occasion
for <1pra$rthana$>1 is some act of merit, typically at a <1da$ne$.>1 On receiving
a <1da$ne$>1 a monk recites a Pali verse in thanks:<s4>s

220

         <1Icchitam patthitam* tuyham* khippam eva samijjhatu
         Sabbe pu$rentu attasam*kappa$ cando pannarasi$ yatha$.>1
         May what you wish and desire<s5>s very soon be success-
         ful; may all your wishes for yourself be fulfilled like
         the full moon.

The content of this verse indicates that anything may be
desired, not just <1nirva$n*a.>1 (Indeed a reformist monk [not in my
area] pointed this out to me as his reason for never using the verse.)
The donor makes a wish and the monk hopes it will be fulfilled.
This is not strictly in contradiction to <1karma>1 theory, because it can
be said that the wish will only be fulfilled if the <1karma>1 is good
enough, and the merit gained just before a <1pra$rthana$>1 is made
should ensure this. However, there is little doubt that affectively
the donor feels he is achieving a certain result by a certain action
in an automatic, magical way. Moreover, the explanation just
given only avoids contradicting canonical doctrine in cases where
the <1pra$rthana$>1 is for something other than <1nirva$n*a.>1 This is not
because <1nirva$n*a>1 cannot be achieved by mere wishing (for it can
be argued that the <1pra$rthana$>1 is only for an advance towards
<1nirva$n*a>1 through mental purification, not a direct means to
<1nirva$n*a>1 itself), but because the Buddha specifically condemned
desire for <1nirva$n*a,>1 as he condemned all desire ; one should practice
meditation for its own sake, in the spirit in which a hen sits
on her eggs, and the results will follow. The later canonical
literature shows us that this advice was soon disregarded, as might
be expected, at least by the common run of Buddhists.
But before examining these ancient antecedents of the modern
practice I must give the explanations of <1pra$rthana$>1 offered by two
monks. <1Pra$rthana$>1 for <1nirva$n*a,>1 they said, are merely expressions
of one's desire for <1nirva$n*a,>1 a reminder of one's ultimate goal.
<1Nirva$n*a>1 can only be attained by effort <1(utsa$ha)>1 over many lives,
so one has to keep oneself up to scratch. Success in the human
world and in heaven <1(ma$nus*ya sampat, diviya sampat)>1 come
automatically when one has done merit, said one of them; but to
attain <1nirva$n*a (nivan sampat)>1 one must make a <1pra$rthana$va.>1
This, he said, is because to attain <1nivan>1 requires constant mind-
fulness and constant striving. If I may put this in a different,
unSinhalese way: <1pra$rthana$>1 for <1nirva$n*a>1 are mere epiphenomena

                                                                 221

of the pure thoughts of a man who, by reason of those thoughts,
is advancing spiritually. This explanation seems to me canonically
orthodox in avoiding all suggestion that <1pra$rthana$>1 are prayers
and thus avoid <1karma;>1 it fails only to satisfy the Buddha's in-
junction that one should not desire <1nirva$n*a>1--a text which is not,
I think, widely known. However, it also comes close to saying that
<1nivan>1 is achieved by something other than <1pin>1 (merit). <1Nivan>1
is in fact, as explained in Chapter 2, the result of morality, medita-
tion, and wisdom, all of which are <1pin;>1 to say that anything else is
necessary would be heretical.
   This very heresy was put forward by a monk whom I shall call
monk 6. I have so far not taken account of his opinions, because
his is a very odd case: he is not a fully ordained monk, nor recog-
nized by the other monks, and probably alightly crazy. He is a
poor villager, most unlikely ever to have been exposed to external
cultural influences; yet he has turned Buddhism into a devotional
religion in the Maha$ya$nist style of such sects as Jodo-Shin--a
genuine heretic by local standards. The monks just quoted said
that no external help is possible in your quest for <1nirva$n*a,>1 and
every right-minded Therava$din would agree; but monk 6 ex-
pressed the directly opposite opinion: all <1pin>1 is fruitless <1(ahosi>1
<1venava$),>1 and no one can attain <1nivan>1 without the help of Maitri$.
"Only that man can be called a Buddhist <1(Buddha$gama-ka$rayek)>1
who prays to reach <1nirva$n*a>1 in the time of Maitri$.' Monk 6's
opinions are more fully reproduced at the end of Chapter 7. Here
we must return to the history of <1pra$rthana$>1.
  <1Patthana$>1 for oneself are found in the ancient Pali literature
which I have taken as my basis for comparison with modern
times; they are, however, restricted to commentatorial literature.
In the Four <1Nika$yas,>1 the main body of sermons, the <1patthana$>1
practices described above do not occur, but the word is found
in the same general senses. In one text<s6>s it seems to be a synonym
of <1a$ya$cana$,>1 which means "request' or "prayer'. The meaning
"prayer' is unambiguous in the <1Nida$ma-katha$>1<s7>s (which strictly
speaking is part of a commentary), when Suja$ta$ makes a <1patthana$>1
to a <1yaksa>1 that if she gives birth to a son she will make offerings
to the banyan tree in which the <1yaks*a>1 lives.
   In other canonical passagess <1patthana$>1 is a synonym for <1pan*idhi>1,

                                                                     222

a word which can be translated as "earnest wish' or "aspiration'.
It seems to me that sometimes in the Canon <1pan*idhi,>1 with its verb
<1pan*idheti,>1 is used to give the same sense as <1pra$rthana$>1 today.
<1Panidhi>1 is the technical term for the vow to become a Buddha, i.e.
someone who has made a <1pan*idhi>1 is <1ipso facto>1 a Bodhisattva. One
monk said that a <1pra$rthana$>1 is necessary to become a Bosat or just
to see <1nirva$n*a;>1 here he was using the word <1"pra$rthana$'>1 as the
translation of <1pan*idhi>1 (which has no other Sinhalese equivalent);
the use has, however, been alightly extended, so that it is made not
only by those exceptional creatures who will become Buddhas,
but by all seeking enlightenment--hoping to become <1buddha>1 with
a small <1b.>1 (As mentioned in Chapter 3, anyone enlightened is
<1buddha,>1 nor is it peculiar to Maha$ya$na (as is sometimes thought)
to grade <1buddhas,>1 the Bodhisattva being superior to the mere
<1arhat.)>1
   The <1pan*idhi>1 is also the canonical equivalent of the modern
<1pra$rthana$>1 in another context, that of the death-bed wish. In one
story<s9>s a dying householder is advised by tree-deities to aspire to
become a universal monarch ("<1Pan*idhehi>1 . . . <1ra$ja$ assam* cak->1
<1kavatti$ti').>1 They think that this aspiration will succeed as he is
virtuous. But the householder says that kingship is impermanent,
and instead recites the qualities of the Three Jewels, giving his
relations faith <1(pasa$deti)>1 in them. It is perhaps notable that the
householder does not make a <1pan*idhi>1 (or <1patthana$)>1 for <1nirva$n*a,>1
thus preserving the pristine purity of the teaching; plainly,
however, this is only a short step away. Before leaving the <1pan*idhi>1
we may conclude that though the distinction between the spirit
behind a <1pan*idhi>1 to become a Buddha and a <1patthana$>1 to attain
<1nirva$n*a>1 may be clear in theory, it is a nice one which is liable to
become obliterated in practice.
  The word <1patthana$>1 occurs precisely in its modern sense in the
commentary to the <1Dhammapada.>1 Here are three examples,
corresponding to the three main kinds of <1pra$rthana$>1 for oneself
described above as current. First the dying wish : a woman makes a
<1patthana$>1 to be reborn as a female <1yaks*a (yakkhini),>1 and her rival
similarly achieves rebirth as a leopard.<s10>s Notice that these are
wishes by sinful people for very low rebirths. Second, <1patthana$>1
for some worldly end: a wrestling champion put two coloured
banners on the shrine of a former Buddha and wished that all

                                                                      223

women who saw him, except only his relations, should fall in love
with him; so in his future lives other men's wives on seeing him
could not control their desire.<s11>s For this magical use of an act of
merit to gain an immoral end I have no evidence today; I only
suspect that it <1may>1 occur--probably even the wrestler told no one
of what he was doing. Thirdly, <1patthana$>1 for <1nirva$n*a>1 (and for
lower grades of spiritual attainment) occur<s12>s in a story which
could be describing procedure today. After feeding a <1pratyeka-
buddha>1 with sugar-cane a man makes a <1patthana$>1 for "success
among gods and men, followed by enlightenment', whereat the
<1pratyekabuddha>1 says, "So be it' <1(evam hotu),>1 and recites <1"Icchitam*
patthitam*,>1 etc.' What is more, he then "forms the resolution that
he [the donor] should one day comprehend the law' <1(yatha$ so
passati evam* adhit*t*hahitva$).>1 This is wildly undoctrinal! Later, on
hearing of the incident, the man's elder brother makes a <1patthana$>1
for <1nirva$na.>1
   We see therefore that though <1patthana$>1 are foreign to the
earliest Buddhism, they appear in their "modern' form in ancient
times. They do undoubtedly contravene the principle that one
should not desire <1nirva$n*a,>1 but they can be, and are, explained in
such a way that they do not contravene the doctrines of <1karma>1
ad intentionality <1(ce$tana$va).>1 At the same time it is clear that
this cognitive congruence is achieved by a rather subtle argument
which can bear little relation to the feelings of most participants.
The man who gave the sugar-cane doubtless considered that the
<1pratyekabuddha>1 was granting his wish, and so did the story-teller.
The notion that anyone, even a <1pratyekabuddha,>1 can resolve on
someone else's spiritual progress is a more flagrant deviation from
doctrine than anything I came across in the field, and suggests the
Maha$ya$na. The elder brother probably felt that his own <1patthana$>1
would produce its result automatically. The elder brother's
affective beliefs suggest magic, the younger brother's theism. I
shall return to this point in a moment.
   The <1patthana$>1 just quoted occur in a text which is mainly a collec-
tion of stories. But the most typical and "popular' <1pra$thana$va,>1 the
wish to be comfortably reborn till one finally attains <1nirva$na>1
under Maitri$, occurs in the most serious and respectable ancient
source outside the Canon itself. Here are the closing lines of
Buddhaghosa's <1Visuddhimagga.>1

224

       <1Yam* siddham* imina$ pun+n+am*, yan*# c'an+n+am* pasutam* maya$
       etena pun+n+akammena dutiye attasambhave
       Ta$vatimse pamodanto si$la$ca$ragun*e rato
       alaggo pan+ca ka$mesu patva$na pathamam* phalam*
       Antime attabha$vamhi Metteyyam* munipungavam*
       lokaggapuggalam* na$tham* sabbasattahite ratam*
       Disva$na tasa dhi$rassa sutva$ saddhammadesanam
       adhigantva$ phalam* aggam* sobheyyam* Yinasa$sanam.>1

By the merit which has accrued through this <book> and by the rest
which I have generated: may I in my next life rejoice in the
Ta$vatimsa heaven, delighting in the quality of virtuous conduct, and
unattached to the five desires may I attain the first fruit; in my last
life may I see that bull among seers Maitn$, the world's highest
person, the Lord who delights in the welfare of all beings; may I
hear his preaching of the true doctrine, attain the highest fruit, and
beautify the religion of the Conqueror.

  These lines are missing in Burmese editions, so they could be an
interpolation. But no matter:  the same idea is unmistakably
alluded to in the body of the text (I, 135) and there ascribed to a
monk called Maha$sangharakkhita, who may or may not be the
same as the monk of that name whom Adikaram dates to the first
century B.C.,<s13>s but must in any case antedate Buddhaghosa himself.
  Before leaving the <1pra$rthama$>1 I must mention a canonical
doctrine which logically and affectively comes very close to it:
the act of truth (Pali: <1saccakiriya$).>1 It is believed that a solemn
asseveration of one's righteousness in some particular respect
can, by its truth, produce some desired result.  Most of the
canonical examples occur in the <1Ja$taka>1 stories,<s14>s but the <1locus
classicus>1 is in the <1Angulima$la Sutta>1 in the <1Sajjhima Nika$ya.>1
The former brigand Angulima$la, now a monk, sees a woman in a
painful and difficult labour; on advice from the Buddha he says
to her, "Since being born in the noble birth [i.e. entering the
Order] I have never intentionally taken life; as this is true may
you and your foetus be well.' And it comes to pass. Angulima$la's
words, known as the <1Angulima$la pirit,>1 are still used in Ceylon to
ease the pains of a woman in labour.
  The "act of truth' is today, not surprisingly, a matter of theory
rather than of practice; perhaps no one alive is holy enough to

                                                                 225

bring it off. The last recorded "act of truth' of which I am aware
was by Dut*uga%mun*u; when his soldiers in the confusion of battle
mistook each other for the enemy he asseverated, "Not for the
pleasure of sovereignty is my striving, but always to establish the
doctrine of the Fully Enlightened one. As this is true, let the armour
on my soldiers' bodies acquire a blazing hue.'<s15>s And so it hap-
pened.
   To harmonize the "act of truth' with the doctrine of <1karma>1
would not be logically impossible, as it could be claimed that
truth is a natural force, which is here being harnessed to a righteous
end; but this would be far-fetched. Affectively the "act of truth',
like the <1pra$rthana$>1, seems to be a short cut.
   To the outsider it would seem that the <1saccakiriya$>1 and the
<1pra$rthana$>1 might be grouped with <1pirit>1 and other practices akin
to white magic mentioned at the end of Chapter 4, in that all
seem to be based on an implicit belief in the power of religious
utterances. I was in fact told by a schoolmaster that the sound of
the <1A$t*a$na$t*iya Sutta>1, which on a night of <1pirit>1 is chanted in the
small hours of the morning, produces vibrations which keep the
<1yaks*as>1 at bay. This explanation is reminiscent of Hindu theory
and is unlikely to be traditional among Buddhists; but though
untypical it is worth quoting, as it explicitly describes <1pirit>1 as a
spell of automatic efficacy. Moreover this is almost done by <1apirit>1
text itself. The <1Maha$ Jayamangala Ga$tha$>1, verses which conclude
the <1maha$ pirit>1, begin

            <1Maha$ka$run*iko na$tho hita$ya sabbapa$n*inam
            Pu$retva$ pa$rami$ sabba$ patto sambodhim uttamam
            Etena saccavajjena hotu te<s16>s jayamangalam>1.

            The lord of great compassion for the welfare of
            all alive fulfilled all the perfections and reached
            the supreme Enlightenment. By this statement
            of truth may the blessing of victory be yours.<s16>s

   The last line recurs several times in the poem. The <1saccavajja>1
(which literally means "truthfulness') is plainly in the context a
<1saccakiriya$>1, "act of truth'; the text, the latest used in the <1pirit>1,
thus gives this explanation of itself. Though this step has never,
to my knowledge, been made by Buddhists, the most economical
explanation of the <1pra$rthana$>1 as behaviour would be to put it in

                                                                     226

the same category of white magic which works through the
efficacy of religious utterance.  It would of course be easy to
describe the automatic use of prayer by a Christian in difficulties
in similar terms; but we must remember that despite the prayer-
like appearance and history of the <1pra$rthana$>1 it is, at least nowadays,
more akin to magic than to theistic religion, for even though the
benediction of a monk or <1pratyekabuddha>1 may somehow help it
along, it is not considered to depend for its effect on the whim of
any external agent.

  The third contemporary belief which seems at first sight to run
counter to <1karma>1 theory is the transference of merit, usually
called in Sinhalese "giving merit' (<1pin di$ma>1), a term which can be
misleading. Giving normally implies that the giver parts with the
thing given, but in giving merit one merely offers to others the
chance to earn merit by rejoicing at one's own. It was several
times explained to me that giving merit is like giving light when
one lamp is lit from another. This disposes of the problem as far
as the giver is concerned; what about the recipient? Can one
acquire merit from an act one has not done oneself? Here I must
refer the reader back to the list of good deeds on page 74. The
fifth is <1patta$numodana$>1, rejoicing in another's merit (and the
fourth is <1patti>1--giving someone a chance to rejoice in your merit,
i.e. "giving' it). If someone does a good deed you may earn as much
merit as he does--or even more--by being really pleased about it.
If this seems to the action-oriented westerner to be an easy way
out, let me remind him that this doctrine is fully consonant with
Buddhism's ethic of intention: it is the thought that counts, and
merit bears fruit for the doer because of the pure thought that
accompanies it; therefore if the thought of a spectator is purified
too, so much the better for him. The high priest of Mi$gala,
having spent 5,000 rupees on a <1pinkama>1, told me that a villager
who took sympathetic joy in the merit he was earning might
thereby earn more merit from it than he did himself, without
spending any money. This seems to me a humane and rational
equivalent of the New Testament story of the widow's mite.
Theoretically indeed this could be an endless process, pleasure at
the merit (and consequent spiritual progress) of others enhancing
your own good <1karma>1, and their pleasure at your progress in turn
enhancing theirs. All this is <1patti>1 and <1patta$numodana$>1.

                                                               227

    This ingenious and pleasant doctrine has a curious history,
which well illustrates the interrelation between the cognitive
and the affective which is a theme of this work. In this generalized
form it is not a part of the original doctrine, and the terms <1patti>1
and <1patta$numodama$>1 in this sense are not found in the Four
<1Nika$yas>1.  The verb <1anumodati>1 is used in these early texts with
two closely related meanings: "to agree with'; or "to receive with
gratitude', i.e. "to thank'. In the first meaning, Sa$riputta says of a
doctrinal debate:<s17>s <1na me koci bhikkhu anumodati:>1 "no monk agrees
with me'.  Parallel to the second meaning is the noun <1anumodana>1<s18>s
"gratitude' or "thanks'. This word from the beginning<1<s19>s>1 is mainly
used as a technical term for the thanks uttered by a monk on being
given alms. This usage has been preserved unchanged till today in
the Sinhalese <1anumo$dan>1. The passage in the <1Vinaya Pit*aka>1<s20>s in
which the Buddha prescribes the utterance of the <1anumodana>1
and says that it should be said by the eldest monk present does not
specify the content of what is said: though the Buddha doubtless
composed his thanks variously to suit the occasion, one may
assume the early use of some such benedictory formula as "<1Icchitam*
patthitam*>1...' quoted above (p. 220).
   Merit is typically transferred after a <1da$ne$>1. The donor (<1da$yaka>1)
presents food (i.e. a <1da$ne$>1) to the monk, who then recites the Pali
verse <1Icchitam* patthitam*>1, followed by a long formula in highly
Sanskritized Sinhalese prose (see Appendix), telling the donor to
transfer the merit to the gods. This the donor does, either by
reciting the following Pali verse or by giving his assent ("Sa$dhu
sa$dhu') when it is recited by a monk.

           <1A$ka$sattha$ ca bhummat*t*ha$ deva$ na$ga$ mahiddhika$
           Pun+n+am tam* anumoditva$ ciram rakkhantu sa$sanam*>1.

           May air-dwelling gods and ground-dwelling gods,
           na$gas of great power, having rejoiced at the merit,
           long protect the Teaching.

The verse may be repeated, with "me' substituted for "the Teach-
ing'. It is also used at the end of any <1pinkama>1 or after private
worship.

                                                                       228

   This transfer of merit to the gods is canonical. In the <1Maha$pari-
nibba$na Sutta>1 the Buddha receives a meal from tvo ministers.
"The Blessed One thanked (<1anumodi>1) them with these verses:
Wherever a wise man dwells, he should feed the virtuous and
restrained ascetics there, and dedicate the gift (<1dakkhin*am* a$dise>1)
to whatever deities are there; when worshipped they worship,
when honoured they honour him.'<s21>s
   The text does not use the term <1Patti>1, though the meaning
expressed is the same; <1patti>1, however, is the commentatorial
gloss:<s22>s <1Ta$sam* dakkhin*am* a$dise ti sanghassa dinne catta$re paccaye
ta$sam gharadevata$nam* apadiseyya pattim* dadeyya>1. '"He should
dedicate the gift" means that when the Sangha have been given
the four requisites one should dedicate, give the merit to those
house-deities.' And the commentator significantly continues:
' "When worshipped they worship": they think, "These people
are not even our relatives, and even so they give us the merit." '
   Though the commentator is much later than the text, I think
he is right about the implied origin of merit transference. We come
here to a complex of ideas centring on those funeral feasts for dead
relatives which are common to so many cultures. <1Prima facie>1
they are perhaps an unexpected phenomenon in a Buddhist
context, but Buddhism inherited them from its Indian Hindu
background.  Professor von Fu%rer-Haimendorf describes how
Chetris, a high Hindu caste in Nepal, can gain merit by having
brahmins recite sacred texts for a week (<1sapta$ha):>1 "If performed
as a memorial rite the <1sapta$ha>1 involved the gift of an entire set of
household goods to the senior brahmin priest, and it is popular
belief that as a result of this donation corresponding objects of
personal use will be available to the departed for his life in the
next world. The fact that such an idea is inconsistent with the
belief in the immediate reincarnation of every human being in a
shape conditioned by his earlier deeds does not seem to disturb
the Chetris, who like other Hindus see nothing incongruous in the
holding of apparently inconsistent views.'<s23>s Buddhists are more
concerned with doctrinal consistency, so that this case affords a
fascinating study of the interaction between behaviour and doc-

                                                                 229

trine. But before reverting to ancient times let me describe what
goes on today.
   It is customary to feed monks, if possible in one's home, at
certain fixed intervals of time after the death of a relative. The
number and size of such <1da$ne$s>1 depends mainly on the wealth
and social status of the family; however, a <1da$me$>1 on the seventh
day after the death is obligatory, and one three months after
the death is almost equally common, while annual <1da$ne$s>1 on the
anniversary of the death are very common, especially for dead
parents. (This practice is not confined to laymen: the chief monk
In Mi$gala gives two annual <1danes>1 to commemorate the deaths of
his parents, and a third to commemorate the death of the previous
incumbent, his father in religion. <1Sil ma%$n*iyo$>1, who have more fully
renounced the world, in some cases give up the rite (both for their
parents and for their teacher), but do not seem to regard it as
improper for them: it is not branded as worldly.) All these <1a$ne$sd>1
given for the dead are called <1mataka da$ne$/da$na>1 ("dead' <1da$ne$>1) to
distinguish them from <1ji$va da$ma, da$ne$s>1 given by and for the
living;<s24>s but the <1da$ne$>1 after seven days is the <1mataka da$ne$ par
excellence. A mataka da$ne$>1 should properly be <1sa$nghika>1, which
means given to the Sangha as a body, not to individual monks. The
donor of a <1sa$nghika da$ne$>1 must approach a monk (in fact the chief
monk at his local temple) and issue an invitation by saying that he
wishes to give a <1sa$nghika da$ne$>1 for so and so many monks (at least
five), without naming any particular monk he wishes to come. The
monk who receives this invitation simply sends along the requisite
number.  Five is the minimum required for any official act of the
Sangha (<1vinayakama>1). However, poor people who can afford
only the seventh day <1da$ne$>1 may even then feed fewer than five
monks.
  The <1da$ne$s>1 for the dead vary in character. The seventh day
<1da$ne$>1 is a private affair for the family and maybe close friends. It is
preceded by a sermon (<1ban*a>1) late on the previous evening.<s25>s
The three month <1da$ne$ (tun ma$sa da$ne$>1) on the other hand is an.

                                                                       230

occasion for conspicuous consumption, and is usually preceded
by <1pirit>1, to which all acquaintances are invited--though in the
village explicit invitation is generally unnecessary. The annual
<1da$ne$>1 lies somewhere between these two. However, at <1mataka>1
<1da$ne$s>1 the ritual for transferring merit to the dead is the same, and
takes place after the midday meal (<1daval da$ne$>1) which closes the
whole <1pinkama>1, even if monks are also fed on the previous evening
and at breakfast-time.
  The meal is offered to the Sangha with the usual Pali sentence,
repeated three times:

              <1Imam* bhikkham* bhikkhusamghassa demi>1.

              I give this alms-food to the community of
              monks.

Sometimes a little food is thrown outside the house; it is variously
said to be for the <1pretas>1 or for the crows. When the monks have
eaten (and, usually, been offered "requisites' (<1pirikara>1), objects
they are allowed to own such as pillow-slips) a small bowl or dish
and a pitcher of water are set before them and the immediate
relatives of the dead man come to the front. Everyone assumes a
reverent posture and the head of the household slowly pours the
water into the dish till it overflows, while the monks intone in
unison :

           <1Yatha$ va$rivaha$ pu$ra$ paripu$renti sa$garam
           Evam eva ito dinnam peta$nam* upakappati>1<s26>s
           As the full water-bearing [rivers] fill the ocean,
           so indeed does what is given here benefit the
           dead (<1preta>1).

           <1Un*n*ane udakam* vat*t*am yatha$ ninnam pavattati
           Evam eva ito dinnam* peta$nam upakappati>1.<s26>s

           As water rained on a height reaches the low
           land, so indeed does what is given here benefit
           the dead (<1preta>1).

  When the water has overflowed the monks stop chanting and
everyone says "Sa$'. The proceedings conclude with a short sermon
on transitoriness (<1anityata$>1) to console the mourners, at the end of

                                                                 231

which, instead of the general formula transferring merit to the
gods given above, the householder says (perhaps on prompting
from a monk),

           <1Idam ne>1<s27>s <1n+a$ti$nam hotu. Sukhita$ hontu n+a$tayo>1.

           May this be for my relatives. May my relatives
           be happy.

  The doctrinal explanation of all this is rather complicated. It
was explained thus by the Mi$gala incumbent. The death is primar-
ily an occasion for doing merit (<1pina>1) oneself; secondarily for
offering it in case the dead man is expecting it. He can, however,
only rejoice and benefit from the merit if reborn as a <1pere$ta>1,
because if he is higher than that he does not need the merit, if he
is lower, in hell (<1apa$ya>1), he cannot get it. [There is in fact a
further refinement: only the top class of <1pre$ta>1, "those who live on
merit given by others' (<1paradattopaji$vin>1) are able to sympathize
with the merit of others.]<s28>s But this does not mean, he said, that
the relatives giving the <1da$ne$>1 assume that the dead man is now a
<1pre$ta>1, for if we pay a call we take food along as a gift, but if the
person is out we eat it ourselves; similarly, the <1pinkama>1 earns
merit for the living whatever the fate of the dead. The origin of the
custom of offering it to the dead is this, he said. Once King
Bimbisa$ra gave the Buddha a <1da$ne$>1, and his ancestors, who were
<1pre$tayo$>1, came to see it. He offered them no <1pin;>1 so they were sorely
disappointed. In the night there was a great noise round his palace,
so next morning he went to the Buddha and asked if this boded
ill for himself or his kingdom. The Buddha said not so, but
explained what had happened. So to remedy his oversight Bim-
bisa$ra gave him a <1da$ne$>1 that day too, and offered the <1pin>1 to the
<1pere$tayo$>1. My informant added that to offer <1pin>1 one has really to
<1think>1 of offering it while doing it--a mere form of words, or
afterthought, will not do.
  When introducing the <1pre$tas>1 in Chapter 4 I dwelt on their
ambiguous status throughout the recorded history of Buddhism,
and the above statement illustrates this very well. It contains
small but telling points of doctrinal inconsistency: it is not clear

232

why <1pre$tas>1 who have already come to the <1da$ne$>1 (as in the story of
Bimbisa$ra) should be unable to rejoice at the merit unless it is
specifically offered; nor why gods, who in other contexts are
always offered merit, are in this context said not to need it.
However, the latter discrepancy can be removed by saying that
while the gods are in no hurry for the merit the <1pretas>1 need it
urgently, because they are suffering in a state of woe, from which
only <1patta$numodana>1 can relieve them, because they have little
or no opportunity for performing meritorious actions independ-
ently. Remember, moreover, that length of life is one of the things
that decreases as one goes down the cosmic scale; the Rev. Rahula
has told me of a belief that <1pretas>1 live only seven days, which
makes it essential to catch them at the point before they sink
further. He very plausibly connects this with a belief found in other
schools of Buddhism (Maha$ya$na and Sarva$stiva$da) in an "in-
between state' (<1antara%bhava>1) lasting seven days; during this
period the person is suspended between death and rebirth, and
any improvement in <1karma>1 will of course make the next birth a
better one. I must stress that this <1antara$bhava>1 is not a Therava$da
belief; it goes towards explaining the seventh day <1da$ne$>1 on the
historical, not on the doctrinal level. From this historical angle it
is also interesting to notice that in the Pali words just quoted the
donor transfers merit to all his relatives, not just the recently dead
man, thus reminding us that Hindu offerings are to ancestors
(<1pitaras>1).
  The <1mataka da$ne$>1, though not described in every ritual detail,
is canonical in Therava$da Buddhism. In one <1sutta>1<s29>s a brahmin
says to the Buddha that brahmins give funeral feasts (<1s/ra$ddha$>1--
Pali <1saddha$>1), praying that the gifts [given to brahmins on their
behalf] may be enjoyed by their dead relatives, and he asks
whether this really works. The Buddha at first replies that it does
not work if the relative is reborn in hell, as an animal, as a human,
or as a god, but works if he is reborn as <1apreta>1, in which case he
lives on what his friends and kinsmen supply. In reply to further
questions the Buddha says that if the particular relative the donor
had in mind is not a <1preta>1, other relatives who are <1pretas>1 will
enjoy it, and it cannot happen that <1no>1 relatives are re-born <1aspretas>1;
but anyway no donor is without reward (<1da$yako>1...<1anipphalo>1). In
this text no reference is made to the merit of the act; the gift is

                                                                         233

said to benefit (<1upakappati>1) the relatives and they to enjoy
(<1paribhun+jati>1) it, so presumably the object passes to them direct.
That all this is addressed to a brahmin points up the fact that the
Buddhists were consciously adapting the Hindu custom.
   There are many references to the custom in a canonical book,
the <1Petavatthu>1, which consists entirely of poems about <1pretas>1.
The three verses cited above come from a poem<s30>s entitled the
<1N+a$tidhamma Sutta>1 ("The Sutta of behaviour due to relatives').<s31>s
(It also occurs under the title, taken from the first words, of <1Tiro-
kudda Sutta>1 as the seventh item in the nine-item canonical
chrestomathy, the <1Khuddakapa$tha>1). Here is Bhikkhu N~a$namoli's
translation of the whole poem.

          1.  Without the walls they stand and wait,
              And at the junctions and road-forks;
              Returning to their erstwhile homes,
              They wait beside the jambs of gates.
          2.  But when a rich feast is set out
              With food and drink of every kind,
              The fact that no man does recall
              These creatures stems from their past acts.
          3.  So they who are compassionate
              At heart do give for relatives
              Such drink and food as may be pure
              And good and fitting at these times:
          4.  "Then let this be for relatives;
              "May relatives have happiness.'
              These ghosts of the departed kin
              Foregathered and assembled there
          5.  Will eagerly their blessing give
              For (plentiful) rich food and drink:
              "So may our relatives live long,
              "Owing to whom we have this gain;
          6.  "For honour to us has been done,
              "No giver ever lacked the fruit.'
              Now there is never ploughing there,
              Nor any cattle-herding found,
          7.  Nor merchandizing just the same,
              Nor bartering for coin of gold:

234

              The ghosts of the departed kin
              Live there on giving given here;
          8.  As water showered on the hill
              Flows down to reach the hollow vale,
              So giving given here can serve
              The ghosts of the departed kin.
          9.  As river-beds when full can bear
              The water down to fill the sea,
              So giving given here can serve
              The ghosts of the departed kin.
         10.  "He gave to me, he worked for me,
              "He was my kin, friend, intimate'.
              Give gifts, then, for departed ones,
              Recalling what they used to do.
         11.  No weeping, nor yet sorrowing,
              Nor any kind of mourning, aids
              Departed Ones, whose kin remain
              (Unhelpful to them acting) thus.
         12.  But when this offering is given
              Well placed in the Community
              For them, then it can serve them long
              In future and at once as well.
         13.  The True Idea for relatives has thus been shown,
              And how high honour to departed ones is done,
              And how the bhikkhus can be given strength as well
              And how great merit can be stored away by you.<s32>s

The verses cited above were, in order, the ninth, eighth and
first half of the fourth. As Stede has remarked,<s33>s verses 11 and
12 (his 10 and 11) look like an addition. Moreover, till verse
12 there is no trace of Buddhism; dead relatives are to get food
and drink, and benefit their donors in return. Not till the end is
there mention of the Sangha or of merit, and they are not well
integrated. In the context of the poem the "this' which is given to
the relative in verse 4 line 1 is food and drink; only in the ritual as
now performed and explained is it merit, or rather the chance to
rejoice at merit. It is this rather complicated explanation which has
circumvented a doctrinal incongruity which originally must have
been glaring. A vestige of the originally Hindu practice of actually
offering food has, moreover, been preserved in the optional custom,

                                                                   235

noted above, of throwing a little food outside the house. Those who
say this is for the crows are rationalizing, though in accordance
with Buddhist ethics (kindness to animals); that it is for the <1pretas>1
must be the ancient explanation.  Note that the custom has
persisted, with no logical congruity, throughout the doctrinal
discussions about to be presented; but as their effect has been to
shift the emphasis of the ritual entirely away from the food it is
not surprising that the handful of food for the <1pretas>1 has declined
to a barely noticed, even an optional, detail.
  The story of the origin of the <1mataka da$ne$>1 told me by the
Mi$gala incumbent is taken from the commentary on this poem,
Buddhaghosa's <1Paramatthajotika$>1. In the commentary the thing
offered is still explained as the food, etc.' not the merit; but this is
perhaps not an important point, as the general interpretation is
clearly the modern one: the king gives a <1da$ne$>1 to the Buddha and
dedicates it (<1uddisati>1) to the <1pretas>1. More important is the story of
this event given in the commentary to the <1Dhammapada>1.<s34>s The
Buddha explained to Bimbisa$ra that after he had given the first
feast the <1pretas>1 had made a row because "when you gave the food
they did not get the merit' (<1da$ne dinne pattim* alabhama$na$>1). When
Bimbisa$a therefore fed the Buddha the next day, "he gave the
merit, saying "Sir, may the divine food and drink from here accrue
to those <1pretas." (Bhante, ito tesam peta$mam* dibbannapa$nam sam-
pajjatu$ ti pattim* ada$si>1). They get the food, then show themselves
to the king naked. The king therefore gives robes (<1ci$vara$ni>1)
the next day to the Buddha and his disciples, and the <1pretas>1
accordingly are clothed in heavenly garments. At this they leave
the condition of <1preta (petattabha$va>1) and become gods (<1dibbatta-
bha$va>1). "The teacher, giving thanks, used the words of thanks,
"Outside the walls they stand", etc.' (<1Sattha$ anumodamam* karonto:
tirokud*d*esu tit*t*hanti$ti tirokud*d*a$numodanam* aka$si>1).
  In this account the spirits are getting not mere merit, but more
tangible benefits--food and clothes. However, they get the food
ad clothes as a result of getting merit--plainly an ambiguous
situation. The question whether the <1pretas>1 could actually eat the
food was controversial in ancient times even though that they
could is the natural interpretation of the <1sutta>1 I have quoted.
In Nyanatiloka's summary of the <1Katha$vatthu>1, a late canonical
book (probably third century B.C.), question 69 is, "Can alms which

236

are given here be enjoyed by beings elsewhere (e.g. by the ..
Petas)?' and we are told that two sects think that they can, but that
Therava$dins hold that "the mind of the Petas might be favourably
influenced, but the material food cannot be enjoyed by them.'<s35>s
Though this seems to contradict the Therava$din commentaries
just quoted, which in their present form are many centuries
younger than the <1Katha$vatthu>1, it is very likely that the comment-
aries on this ancient custom are quoting an old story. But whatever
the date of the final victory of orthodoxy, it is clear that sensible
Therava$din monks decided that food being visibly consumed by a
monk could not possibly be eaten by someone else, so that if
people persisted in their habit of feeding dead relatives the custom
required reinterpretation. What the relatives were really getting
was something else--merit.
  Though we saw above that the offering of merit to the gods
occurs in the Four <1Nika$yas>1, the doctrine of <1patta$numodana$>1,
the acquisition of merit by anyone through empathizing in
another's merit, does not. According to modern doctrine such
empathy can take place whether one is "offered' the merit or not.
But the whole point of the story about Bimbisa$ra's ancestors was
that they had to be offered the merit before they could get it, and
the implication of offering merit to the gods is that they too cannot
get it just by being present--which they are in any case (note that
the reference in the <1Maha$parinibba$na Sutta>1 is specifically to the
local gods).  Moreover even now in all standardized situations
a verse or formula is recited offering the merit to specific bene-
factors. Why should the Pali verse be recited at the <1da$ne$>1 offering
the merit to the gods? The doctrinal answer (same monk) is that
one is simply drawing their attention to the merit. The same
reason can be given for the less institutionalized practice of carrying
round among the laymen present an offering which one is about
to make to the Buddha or the Sangha: the laymen fold their
hands, touch or make to touch the offering with their fingertips,
and then raise their hands to their foreheads in the gesture
of worship: the person who is physically making the offering
is drawing the attention of the others to his act of merit, and
their gestures symbolize their participation. In this case the
doctrinal rationale fits. It does not, however, adequately explain
what goes on at the ritual for the dead.

                                                                      237

  The reinterpretation of the <1mataka da$ne$>1 is the nub of this
problem; it can be followed by tracing the evolution of the meaning
of <1anumodati>1 from "thank' to "empathize', "rejoice in another's
merit'. In the poem quoted above the verb <1anumodati>1 is used of the
<1pretas>1 (poorly translated by <1N+a$n*amoli>1 as "their blessing give').
Then in the commentary the Buddha too does an <1anumodana>1.
Now observe what happens. The Buddhist givers of funeral feasts
have been told by the monks that their relatives are not getting the
food, but are getting something else--merit. The whole point of
the rite is to <1give>1 your dead relative something, so the donors are
satisfied, provided they still have the feeling they are <1giving>1.
They give their merit away to the dead, like goods or cash, and the
dead--presumably--say thank you (<1anumodanti>1). But no, say the
monks; the doctrine of <1karma>1 will not allow this; you cannot really
give your merit away; you are just allowing the dead to improve
their minds by expressing sympathetic joy at your good action in
feeding us. So though the text used in the rite says that the <1pretas
anumodanti>1, this does not mean that they thank you for a gift,
for nothing passes between you; they are just rejoicing.
   This is where the meaning of <1anumodati>1 undergoes its crucial
change : as part of the doctrinal response to undoctrinal behaviour.
Once <1anumodati>1 comes to refer just to a pure mental state, to
empathy in doing good, it is of course open to anyone to do it at
any time, without having to wait for an offer, and this is the
modern position: the villagers could rejoice at the monk's fiftieth
birthday celebrations without specific invitation.
   But since <1anumodati>1 was originally used both of the monk who
gets the food and of the god or <1preta>1 who gets the merit, the way is
now open to linguistic confusion. That is indeed what we find in
modern Sinhalese, for while what the monk recites at a <1da$ne->1
is still called the <1anumo$dana>1, according to modern doctrine it is
not he who <1anumodati>1, rejoices, but the gods--the third parties.
We thus get the following table, which I owe to a conversation with
the Rev. Rahula. In the table the three participants (or groups of
participants) are on the left: in the centre are the actions they are
performing in Pali, Sinhalese, and a literal English translation;
on the right is the type of good deed (according to the Pali list of
the good deeds) which the action represents. The lines from top to
bottom are in chronological sequence.

238

Donor   deti            denava$         gives      da$na
Monk    anumoda$peti   anumo$dan        causes to   desana$
                           karavanava$  cause to
                                          rejoice
Donor   anumodeti      anumo$dan        causes to  patti
                           karanava$      rejoice
Gods     anumodanti    anumo$dan        rejoice     patta$numodana$
                             venava$

   The Pali alone suffices to show that something has got twisted
up here. In tact the double causative form <1anumoda$peti>1 is not in
the dictionaries, and the plain causative <1anumodeti>1 is cited only
once,<s30>s in the quite different meaning "get the approval of'.
  The popular understanding of what goes on--which I might
also call the common-sense view--is rather different. It also
corresponds to the historical view. The monk is understood to be
saying "Thank you', as indeed he was by original doctrine. The
donor is then understood to be giving the gods his merit as a
<1quid pro quo>1, as if he were buying their protection for cash.
  This sort of transaction is only possible after the complete
separation of the merit earned by a gift from the gift itself, a
separation which we saw take place at the <1mataka da$ne$>1. In that
context the "transference of merit' was made in the <1Katha$vatthu>1
to conform to <1karma>1 doctrine by talking of the good intentions of
all concerned. But this explanation becomes more strained in our
next example, in which the merit of a gift is given retrospectively.
In a <1Ja$taka>1 story<s37>s the Bodhisattva, born as a brahman merchant,
has gained merit by feeding a <1pratyekabuddha;>1 he is shipwrecked,
and while swimming in the sea with an attendant is picked up by a
deity whose duty it is to protect virtuous men in misfortune; she
did not notice the attendant, so "the brahmin gave him the merit
of his good deed, and he received it gratefully' (<1bra$hman*o attana$
katakalya$n*ato tassa pattim* ada$si, so amumodi>1) and was picked
up.<s38>s Here we are very close to the idea of a fund of merit' like a
bank account, to be drawn on at will. I stress that even this passage
can be rationalized by reference to the doctrine of intention;
but the more obvious interpretation of the passage would be to
regard merit as a kind of spiritual money. And a characteristic

                                                                    239

of money is that when you have used it you no longer have it.
  The equation of merit with money becomes virtually explicit
in the story from the commentary to the <1Dhammapada>1 about the
two brothers and the sugar-cane (see p. 223). When the younger
brother feeds the <1pratyekabuddha>1 sugar-cane from his brother's
field, he thinks, "If my elder brother demands the price [of the
cane] I shall give him the price, if the merit I shall give the merit'
(<1face me jet*t*habha$tiko mu$lam a$hara$pessati mu$lam* dassa$mi sace
partim a$hara$pessatipattim* dassa$mi>1).<s39>s Of course this again is not a
doctrinal treatise; but it is clear that the merit and the money are
on an equal footing, and that the giving of the merit will have
nothing to do with benevolence or purity of thoughts. Here the
incongruity of the transaction with doctrine is not confined to the
affective, and reaches the cognitive level.
  In texts quoted so far the verb <1anumodati>1 has only been used of
receiving something offered, so that it has involved thanking for
something, not just rejoicing at it. There is nothing especially
meritorious about thanking someone for a gift; rejoicing at their
merit, however, is a good psychological state, and thus meritorious.
In our next example it is not yet clear whether the merit was offered
or not, though I think the latter. Here<s40>s a monk talking to a <1preta>1
says, "You never gave anyone a gift with your own hands, but
rejoicing at the gifts of others (<1parassa da$nam* anumodama$no>1)
stretched out your hand and spoke.' However, by this means the
<1preta>1 had acquired enough merit to be reborn in comfort.
  In the next stage a man rejoices (<1anumodati>1) at merit, not merely
without being offered it, but without being present. He merely
hears that there is going to be a <1da$ne$>1 and expresses his joy. This is
the modern doctrinal position.
  Finally let me quote from a late commentary a passage in which
both <1patti>1 and <1anumodana>1 occur. The words, though purporting
to describe Visa$kha$ and her friends after Visa$kha$ has built a <1viha$ra>1
for the Sangha, could, translated into Sinhalese, be an exact
transcription of modern procedure. Visa$kha$ says to her com-
panions,<s41>s " "The merit I have earned, rejoice at it, I give you the
gift of this merit.<s42>s With gladly trusting minds they all rejoiced.
"Oh, it is good, oh, it is good." One good lady who was there

240

concentrated especially on that gift of merit. Soon after she died
and was reborn in the Heaven of the Thirty-Three.' ( <1Yam* maya$
pun+n+am* pasutam*, tam* anumodatha, pattida$nam* vo dammi$ ti.
Aho sa$dhu aho sa$dhu$ ti pasannacitta$ sabba$ pi anumodimsu. Tattha
an+n+atara$ upa$sika$ pi visesato tam* pattida$nam* manasa$ aka$si>1.) This
<1anumodana$>1 of hers she describes<s43>s as a "pure rejoicing' <1(suddh'
anumodana$)>1 and hence truly meritorious. (This is in the canonical
text, not merely the commentary.) Though in this case the merit
was offered, that "rejoicing' and not "thanking' is the appropriate
translation is clear both from this passage and from the two
previous passages cited.
   We have traced here a correspondence between affective
religion and an early behavioural deviation, appearing in Buddhist
stories, but never explicitly accepted by doctrine; doctrine has
then made a come-back and harmonized practice with canonical
theory, though not without becoming exceedingly tortuous (and
philologically barbarous). The stages through which behaviour
evolved have been traced above; the behaviour represented in
passages dealing with <1patti>1 displays a gradual deviation frorn
doctrinal orthodoxy, while the changed meanings of <1anumodama>1
and its verbs represent the rationalizations of doctrine to accom-
modate the behavioural deviations. On the behavioural level the
passages first quoted for <1patti>1 represent sommeone's giving their
merit to another person, and that person's saying thank-you. As
the idea that one can give away merit contradicts a fundamental
doctrine, this clear implication has to be explained away, which is
done, most ingeniously, by changing the meaning of <1anumodati>1.
Although I know of no passage which is quite explicit on the point,
the identity of the last passage quoted with modern practice
strongly suggests that the modern doctrine too had been evolved
by the time it was written. For this it is significant that the good
lady attributes her rebirth in heaven to the purity of her rejoicing
(<1anumodana$>1). This comes not in the commentary, but in a canon-
ical text, albeit a late one, the <1Vima$navatthu>1. The entire evolution
of doctrine and behaviour with which we are concerned therefore
took place, in all probability, within what for the purposes of this
work I am calling the ancient period; and what I have called the
"modern' doctrine, though clearly different from the original
doctrine, may be as much as 2,000 years old.

                                                                     241

  In the <1mataka da$ne$>1 we encounter an especially striking case of
incongruence between the cognitive and the affective--a very
wide gap between the explanation of the ritual and what the chief
participants feel about it. It is true no doubt that they are not
calculating whether the dead person has been reborn as a <1para-
<1dattopaji$vin preta>1 so that he can benefit from their actions, but
the reason is not one of subtle doctrine. They feel emotionally
that they owe this ceremony to the dead man and that he will
benefit from it; they would act in a very similar way without the
doctrines to which they subscribe. Even the overflowing water
may have been reinterpreted. The pouring of water in Sinhalese
ritual frequently signalizes a solemn act (e.g. marriage); but here
it seems most relevant to recall the libations poured to the Manes in
other, older cultures.
  The water-pouring ritual just described takes place at the
funeral (<1avamangalya>1) itself. When the coffin has been placed in
the pyre or over the grave, according to whether the body is to be
cremated or buried, a white cloth is laid on it. One of the monks
present--there may be any number, down to one--gives the Three
Refuges and the Five Precepts as usual; then everyone present
repeats three times after the monk,

            <1Imam matakavattham* bhikkhusanghassa demi>1.

            I give this corpse-clothing to the Sangha of
            monks.

At this the monks spread the cloth out across the coffin, and
chant a Pali stanza:

            <1Anicca$ vata sam*kha$ra$ uppa$da-vaya-dhammino,
            <1Uppajjitva$ nirujjhanti, tesam* vu$pasamo sukho>1.

            Impermanent indeed are compounded things,
            whose nature is to arise and pass away; having
            arisen they are destroyed; their being stilled
            is pleasant.

This couplet is supposed to have been recited by S/akra at the
death of the Buddha;<s44>s its first words, <1Anicca$ vata sam*kha$ra$>1,
dead the printed announcements of the death which are posted
all over the village on the day of the funeral, and their general
purport is probably understood by everyone. The monks then

242

pick up the cloth, symbolically appropriating it, and someone
takes it away. The water is then poured by a near relation as at a
<1da$ne$>1, the monks reciting the same verses; as it is an outdoor
ceremony the water is sometimes that of a young coco-nut, which
is split at the appropriate moment. A monk then preaches. At a
village funeral which I witnessed the monks then left, before the
coflin was lowered into the grave, perhaps to avoid being present
during the lamentations which naturally accompany the body's
final disappearance.
  The cloth given to the monks is known as <1pamsuku$la>1 (literally
"dust-cloth'), a name which has also been extended to the cere-
mony of giving it. <1Pamsuku$la>1 is the name of the first of the classical
list of thirteen ascetic practices (<1dhutanga>1), optional for monks
who wished to display more than the usual ascetic rigour.<s45>s
A <1pam*suku$la>1 monk dresses only in rags picked up at cemeteries.
The modern <1pam*suku$la>1 ceremony is a curious fusion of this with
a <1da$ne$>1. By picking up the cloth from the coffin the monk is
symbolically taking the winding sheet, or some other item of
the corpse's clothing, and thus conforming to the letter of
<1thepamsuku$la>1 practice. On the other hand the dead man's next of
kin are giving the cloth, which therefore is the best new white
cloth, to enhance the value of the gift; they have made the funeral
an occasion for transferring the merit earned by a gift to the
Sangha, thus destroying the spirit of the <1pam*suku$la>1 idea so that
the original meaning of the term has been completely lost.
  The doctrines discussed in this chapter avoid serious incon-
sistency with orthodoxy.  Though some of them have been
described as "modern' they are so only in relation to the Four
<1Nika$yas>1; in simple chronological terms they are ancient.  (In
our discussion of merit transference we have demonstrated
development <1within>1 the corpus of early Buddhist writings, an
exercise not attempted elsewhere in the book.) The latter and
more important doctrines discussed are rationalizations, or if you
will Buddhicizations, of practices (the dying wish, the <1mataka
da$ne$>1) which afford some psychological relief from the oppressive
doctrine of man's total reponsibility for his own fate. Beliefs
which these practices <1imply>1 have never entered the mainstream of
Therava$da Buddhism: the cognitive position has been main-

243

tained with remarkable tenacity. The canonical theory of <1karma>1
survives intact--cognitively; affectively its rigour is sometimes
avoided. Similarly, though the doctrine of <1anatta>1 can be salvaged
(see p. 72) by the claim that the personality continuing through a
series of births has as much reality as the personality within one
life, <1pra$thana$>1 for happy rebirths and the transfer of merit to
dead relatives show that the <1anatta>1 doctrine has no more affective
immediacy with regard to the next life than with regard to this,
and that belief in personal survival after death is a fundamental
feature of Sinhalese Buddhism in practice.

                                6
                      THE ETHIC OF INTENTION

IN the second chapter were listed the Five Precepts (<1pan sil>1),
the Ten Good Deeds (<1dosa kusala karma>1) and the Noble Eight-fold
Path (<1a$rya as*t*a$ngika ma$rgaya>1), three different ways of formulat-
ing the conduct expected of a Buddhist. How do these three formu-
lations relate to one another? The first one which a westerner
usually comes across is the Noble Eight-fold Path, for it is one of
the Four Noble Truths and occurs in the Buddha's first sermon;
but when a villager is asked about <1karmaya>1 he will rarely volunteer
information on the Eight-fold Path, and it is comparatively little
known. The reader will recall that its eight factors have been
grouped from the beginning into morality (<1si$la>1), concentration
(<1sama$dhi>1) and wisdom (<1pan+n+a$>1). Although the impediments to lay
meditation are practical, not doctrinal, they are, or are considered
to be, so great that very few laymen in traditional Ceylon meditate
before old age. Meditation and its peculiar prize, <1nirva$na>1, will
concern us in the next chapter; they are the sphere of what Weber
called virtuoso religion. The religious concerns of the normal
Buddhist layman are centred on morality and merit-making
(<1pinkam>1), which aim at a good rebirth, but rebirth none the less.
To see the merit-making and meditation as a clear dichotomy, as
has been done in sociological works, I do not find satisfactory in
terms of doctrine or of analysis, but as a principle of arrangement
it is convenient: this chapter will be concerned principally with
merit-making.
  The ways in which merit can be made are listed as the Ten Good
Deeds. We notice immediately that the third of them is <1bha$vana$>1,
meditation, which shows the difficulty of dichotomizing between
merit-making and meditation if we are to use the Buddhists'
own terms. The second of them is morality; so what is the relation
of morality to merit-making? By <1si$la>1 is meant the Five Precepts,
which should be called the five abstentions, and these are negative.
They are undertakings to abstain from killing, stealing, sexual
misconduct, lying, and drinking intoxicants. From the point of

                                                                  245

view of the individual, to keep to these undertakings is a sufficient
basis, as expressed in the Eight-fold Path, for meditation and for
the wisdom which attains <1nirva$n*a>1. Cognitively they are quite
sufficient--but affectively perhaps they suffice only for the
virtuoso, if even for him. A negative formulation sounds too cold-
blooded and requires some supplement. The supplement can
come from other canonical formulations, such as the four holy
states (<1brahmaviha$ra>1) of loving-kindness (<1metta$>1), compassion
(<1karun*a$>1), sympathetic joy (<1mudita$>1), and equanimity (<1upekkha$>1);
or simply from such an inclusive list as the Ten Good Deeds, which
present a Buddhist's ideal behaviour--I hesitate to use the word
"duty' in the context of Buddhism, which says "take it or leave it'
--in a positive form. Moreover, the Five Precepts seem to concern
a man in isolation; this is all very well for the virtuoso or the
recluse, but an ordinary layman misses in them any allusion to
such things as his interaction with the Sangha, which is his contact
with organized religion. A final shortcoming of the Five Precepts
from a practical point of view is that it is virtually impossible
to keep to them, so if there is no easier goal the outlook becomes
depressing. The Ten Good Deeds make the Five Precepts but one
item in a list; if you fail to abstain from all sin you can at least
compensate yourself by doing good in other directions. Weber
called the Five Precepts an "insufficiency ethic for the weak'.<s1>s
It is quite true that Buddhism recognizes different stages of
development, at which a man can legitimately make different
demands of himself. These different levels of demand were
summarized in Chapter 2. A monk has to do more than keep the
Five Precepts; put in general terms, meditation and wisdom are
necessary to attain the highest goal, but are expected of only a few
people. Yet to call the Five Precepts an "insufficiency' ethic seems to
indicate a judgement that they are rather paltry stuff. On the
contrary, being absolutes they are extraordinarily demanding. I
am not alluding merely to the fact that nearly everyone lies. Even
the first precept, not to take life, is necessarily broken by any
agriculturalist, and the Sinhalese villagers are vividly aware of the
fact that they are sinners (<1pavka$rayo$>1).
  If anything deserves the name of an insufficiency ethic--a
term I had rather avoid--it is the list of the Ten Good Deeds. To

246

perfect them all is tantamount to attaining <1nirva$n*a>1, but at least
one can make measurable progress by positive acts; unlike the
black and white of the Five Precepts, they allow of shades of grey.
  To sum up, the Five Precepts are negative--to do no evil--
while the Ten Good Deeds are positive--to do good. For the
Christian the latter is neceasary for salvation; for the Buddhist
only the former is absolutely required, but both the emotional
needs of the individual and the practical needs of society seem
better answered by the latter, broader formulation. I speak here
only of ethical actions in the sense of actions affecting others, ad
disregard other recommended behaviour, notably faith (for the
Christian) and meditation (for the Buddhist), to both of which I
shall return.

  Before discuasing the substance of Sinhalese Buddhist morality
I must draw attention to a critical difference between Buddhist
and western moral philosophy. According to generally accepted
western ideas an act may be good which is clearly not to one's own
advantage: one may sacrifice one's life for another. According to
traditional Christian ideas this is only a question of acting in one's
own long-term rather than short-term interests, because the
virtuous man will be suitably rewarded in Heaven. The prudent
man will therefore do good, because it pays; this is called the
<1prudential>1 attitude in ethics. Ever since Immanuel Kant many
people have considered this attitude somewhat base. Kant held
that any act, of whatever apparent goodness, done with thought of
reward, in this life or the next, was <1ipsofacto>1 not moral ; all rewards
were on a par with bribes. For Kant and his followers there is
thus a fundamental opposition between prudence and true moral-
ity. To the Buddhist, however--as perhaps to the traditional
Christian--this is a sheer nonsense. Buddhist doctrine agrees with
Kant that what counts is intention, not effect. (This will be further
discussed below.) <1Karma>1 is nothing more or less than intention
(<1ce$tana$va>1). But by the law of <1karma>1 every intention good or bad
will eventually be rewarded or punished, so prudence and true
morality must necessarily coincide.<s2>s
   It is therefore not surprising that prudential arguments in

                                                               247

favour of moral conduct have a long and respectable history in
Buddhism. When preaching to a lay audience the Buddha listed
the five advantages of morality:<s3>s wealth, good repute, self-
confidence in public, an untroubled death, and rebirth in heaven.
The disadvantages of immorality are the converse. A mediaeval
Sinhalese poet, Vi$da$gama Thera, wrote a century of religious
stanzas, the <1Lo$va$d*a San+gara$va>1,<s4>s which is read in the<1daham pa$sal>1;
nearly every one of these stanzas inculcates virtue under the threat
of tortures in hell. What is perhaps more surprising is that the
prudential attitude is often applied to this life with the mundane
air of "Honesty is the best policy'. I had a conversation with
two monks which brought this out clearly. They asked me why I
thought one should keep the Five Precepts. When I said that
otherwise one would get a bad rebirth they said yes, but what
about the results in <1this>1 life? I suggested remorse, but this was
impatiently brushed aside. If one breaks the precepts one will go
to prison; if one kills a man one is hanged. I protested that for
lying one does not go to prison, but one of them pointed out that
one does go to prison for perjury, so I failed to make my point.
The matter well illustrates the Buddhist belief in meeting the
audience on their own level. Not only did the Buddha, of course,
elsewhere produce far more elevated ideas on ethical motivation;
I also found later that the monk who spoke on hanging held the
pure canonical views on the supremacy of intention and the
supreme importance of purifying one's thoughts.
  In this chapter I shall concentrate on the first two of the Ten
Good Deeds, <1da$na>1 (giving) and <1si$la>1 (morality), which are far the
most important for the study of lay ethics. I shall focus on two
theoretical problems:  the problem of intentionality--whether
the doctrine of pure intention is adhered to or is compromised by a
utilitarian ethic of works; and the related problem of literalism--
whether people pay more regard to the letter or to the spirit
of moral precepts. A literalist interpretation of moral rules
(pharisaism) is incompatible with an ethic of intention.
  Giving comes first in the list of the Ten Good Deeds, and it is
easy to guess why. The existence of the Sangha, and hence of
Buddhism, depends, in theory at least, on the generosity of the
laity. Since the Pol*onnaruva period many temples in Ceylon have

248

owned lands which supply them with food and maybe additional
income, but monastic landlordism was not part of the original
scheme of things. There are, moreover, still many poor temples
where the monks depend wholly or in part on the laity for their
food, which is brought to the temple, and every now and again
reformist monasteries or individual monks revert to the original
practice of begging food from door to door. Beside these circum-
stances peculiar to Buddhism we may remark that all institutions
seem to need cash, and that fund-raising is a prominent feature of
all organized religion under the sun.
   Here, however, Buddhism faces a difficulty. If generous in-
tention is all that counts (and remember the monk who told me
that a poor villager even without giving a cent might earn more
merit than he from the <1pinkama>1 on which he had spent 5,000
rupees), why should people give to the Sangha rather than to
anyone else? This problem seems to have been acute from the
earliest time, for already in the Canon we find the highly ambivalent
doctrine of the <1suita%ble recipient>1. In the very formula describing
the qualities of the Sangha (mentioned above, p. 207) it is de-
scribed as "the best field of merit in the world' <1(anuttaram* pun+n+a>1-
<1kkhettam lokassa>1), which is to say that a good deed done towards
the Sangha (especially a gift given to it) will bear more fruit for
the doer than if bestowed elsewhere. In one sermons the Buddha
says that of the five timely gifts the most important is the gift
given to a virtuous person at the time of the first fruits. In another<s6>s
he says that even scourings thrown into a ceaspool gain merit by
feeding the creatures there, but gifts to the good are more fruitful
(i.e. gain more merit) than gifts to the wicked. This was another
question which I investigated in some of my interviews.
  The fullest answer given by a monk went thus. There are two
kinds of giving (<1(dan di$ma>1): that with thought of worship (<1pu$ja$-
buddhiya>1), which is motivated by respect (<1gaurava>1), and that
with thought of favour (<1anugrahabuddhiya>1), which is motivated
by pity (<1anukampa$va>1). The former is exemplified by a gift to the
Sangha, the latter by a gift to a beggar. For both, the accompany-
ing thought is all-important (<1ce$tana$va pradha$mayi>1), but the former
is superior, i.e. brings more merit. [When I spoke up for the
latter he showed no comprehension.] Merit varies with the virtue
of the recipient, but only in so far as that virtue is known to the

                                                                    249

giver: if the monk practises secret vice this does not diminish the
merit of his <1da$yakas>1. Someone practising the supreme quality of
giving (<1da$na pa$ramita$va>1) hike Vessantara--see below] gives to
everyone, regardless of their virtue; but for the best results one
should find out about the recipient beforehand. "The Buddha
commended giving with discrimination' (<1Viceyyada$nam* sugatappa-
sattham*>1--a Pali quotation).<s7>s
  Other answers on this topic were all in substantial agreement
with the above. One monk elucidated the point about the virtue of
the individual monk. The Sangha is full of good qualities <1gun*asam-
panna>1); when one gives to the Sangha, the virtues of the individual
monk who receives the <1da$ne$>1 do not affect the amount of merit,
because he has virtues, as it were, inherited (<1parampara$vin>1)
from the Sangha throughout history, and the gift is made not to
him, but to the Sangha as a whole. (Strictly speaking, as explained
on page 229, a gift (such as a meal) can only be given to the
Sangha if received by more than five monks, but in this context
this would be considered too subtle a consideration.)
   It already looks as if this doctrine of the suitable recipient is
compromising the supremacy of intention, and another monk's
answer makes this fairly explicit. The merit accruing from
liberality depends on the intention, but it does vary according to
the recipient: if the recipient has noble qualities (<1us/as gun*a>1)
it increases. There is an analogy with sowing seed on fertile or
barren ground. An elephant needs more than an ant, so it must be
more meritorious to feed him. [Here I detected another element
creeping in: to keep the argument on one track I pointed a choice
between feeding an unhungry monk and a hungry beggar.] If
one has to choose between an unhungry monk and a hungry
beggar one feeds the beggar because of the exigency of the
moment, but one gets less merit. Moreover, if with equally good
intention one feeds a monk and a beggar, although the thought is
the same (<1hita eka vuna$t*a>1) it must be more meritorious to feed
the monk.
  The anti-utilitarian streak is interesting: one <1would>1 feed the
hungry beggar, but it would be less meritorious. The question of
which is "really right' cannot be put; the nearest one might get
is to distinguish, as one monk did (see p. 257), between an action
from the societal <1(sama$ja vas/ayen)>1 and the religious (<1a$gama>1

250

<1vas/ayen>1) point of view.s By dwelling on this we must, however,
not forget that no Buddhist denies that one should <1ago>1 feed
hungry beggars. A monk said that the main purpose (<1paranua$rtha>1)
of the Sangha is religious progress, so it is best to give to them, but
this does not mean that one should not give to the laity: one must
help anyone in need; one has a duty to feed the hungry, and gains
merit by doing so. In a truly Buddhist society the rich would give
to the poor, and all would be well. Two other monks also spoke
against social and economic inequality. However, this they seem to
consider a question of politics rather than of religion. In the same
way, one of them denied that capital punishment was contrary to
the precept against taking life. "That's politics' (<1e$ka desapa$lane$>1),
he said. The failure to apply the ethical standards of religion to
politics seems to be universal.
  But let us return to intentionality. It has been compromised by
the variant of the recipient. The other variant is the size of the
gift. Here, too, the evidence is highly ambiguous. In theory
Buddhists believe in the widow's mite, but on closer inspection the
picture is not so clear. For instance, one monk told me that the
amount given is irrelevant, it is the effort that counts. But he went
right on to say that a wealthy person like Mrs. Bandaranaike is
lucky because she can give a lot, which is clearly inconsistent.
Another monk also said that only intention (<1ce$tana$va>1) counted.
To illustrate this he told me a story from the commentary to the
<1Dhammapada>1 about an ascetic who was very poor and so could
give the Buddha nothing of value, but did his best, as a reward for
which he was reborn in a very wealthy family and could thus give
a great deal! Note that this ambivalent tale dates from classical
times, so that this too is a dilemma of respectable antiquity.
  I suggest that the ethic of intention has taken some blows
because it is to some extent counter-intuitive: it seems obviously
better to do good, with whatever motive, than merely to mean
well. I would connect this with my remarks above (p. 245) on
the relation of the Five Precepts and Noble Eight-fold Path to the
Ten Good Deeds. I said that if anything deserves the name of a
insufliciency ethic it is the Ten Good Deeds, which offer a positive

                                                                251

formulation psychologically more reassuring than the negative
absolutes of the Five Precepts. I suggest that the list of Ten Good
Deeds, which is post-canonical,<s9>s is the doctrinal come-back after
experience had shown that the doctrine of pure intentionality was
not fully accepted in practice. Just as the doctrine of merit trans-
ference which I discussed in Chapter 5 was an ingenious legitimation
of the practices of those people (i.e. everybody, or nearly every-
body) who could not accept a particular consequence of the inten-
tionality doctrine, namely that they could do nothing for their dead
relatives, so the Ten Good Deeds, with their blanket coverage of
every potentially meritorious action, offer a general, though merely
implicit, legitimation for the feeling that an act must surely be
better if you have something to show for it. Historically and
emotionally, though not logically, this preference for positive <1pin>1
rather than mere abstention is linked to a general preference for a
happy personal future existence rather than a "blowing out' of the
personality. Most Hindus have believed since before the time of
the Buddha that good works, which they judge largely by effect,
result in rebirth in heaven. (This course, the life of the house-
holder, came to be contrasted to the ascetic's renunciation, which
leads to his freedom from rebirth.) <1Pinkam>1, which include absten-
tion from wrong-doing, meditation, and "right views', but also
easier and more specific goals, are thus the practical result of the
affective ethics of people who hope that through some combination
of good intentions and formally approved actions good works can
ensure for them a pleasant rebirth.
   But it would be a grave mistake to deduce from the above
negative evidence that the ethic of intention receives only
service, if that, and has little importance for the Buddhist ethos.
English has the proverb, "The road to hell is paved with good
intentions', but I cannot see how to put this into Buddhist Sin-
halese. Good intentions pave the road to <1nirva$n*a>1. Whether you
use the Sanskritic word <1ce$tana$va>1, as my informants usually did in

252

the religious context, or talk merely of thought (<1adahas>1 or <1hita>1),
it is the mind that counts. A canonical text says<s10>s that among the
seven<s11>s motives for giving alms those alms bring the best results
which are given "for the decoration and equipment of the mind'
(<1citta$lam*ka$ram* cittaparikkha$rattham*>1). Sinhalese speak of purifying
the thoughts (<1hita suddha kara gant*a>1) or mental progress (<1hita
diyun*uva>1)--terms which are applied equally to meditation and to
other forms of merit-making. I have already mentioned (Chapter
3, p. 118) how the Mi$gala incumbent told me that the point of
offering flowers to the Buddha was to feel joy (<1pri$tiya>1). Professor
Arnold Green has been given exactly the same explanation of the
<1da$ne$>1 (which in Chapter 7 I show to be ancient): the point of the
<1anumo$dan>1, the short speech which the monk makes on receiving
the food, is to fill the donors with joy (<1pri$tiya>1), so that they do not
think of how much it has cost them, but purify their thoughts and
thus make mental/spiritual progress. In the <1Dhammapada>1 com-
mentary story of the two brothers and the sugar-cane, several
times cited in Chapter 5, the elder brother, just before making his
successful wish (<1pra$rthana$>1) for <1nirva$n*a>1, is said to have "his body
suffused with joy' <1(pi$tiya$ phut*asari$ro>1).<s12>s This religion which is
founded on the premiss that the world is sorrow thus attaches
great spiritual value, under the right circumstances, to the feeling
of joy.
   A corollary of these beliefs is the further belief in the <1powe>1
of the mind, a power we have already seen in operation with the
religious wish (<1pra$rthana$>1) and religious resolution (<1pran*idhi>1).
Another theory, which does not logically follow from this but
seems to be connected, is the belief in the strength of mental
impulse. On this again Professor Arnold Green has interesting
information. He was told that someone who left valuables lying
all around was as culpable as the thief if they were stolen, for no
normal person can reasonably be expected to withstand such
temptation. This was never expressed to me in so many words, but
may help to explain the extraordinary insistence of the villagers
that we keep our house locked. Corroborative evidence is the

                                                                   253

insistence of the chief monk that it is <1natural>1 for the poor to envy
the rich. Green was also told that one should not hesitate over
difficult moral choices, but act on impulse. I would connect this
value attached to impulsivity with the Buddhist theory of the
mind as a rapid series of discrete impulses, not an uninterrupted
stream of consciousness as we have thought of it in the West. For
this, however, I do not have any evidence.

  We must now attempt a discussion of "morality' (<1si$la>1) in theory
and practice. <1Si$la>1 means the Five Precepts. These are the most
obvious principles to examine <1qua>1 Buddhist values
life, steal, commit sexual misconduct, lie, or get drunk. They do
not together add up to a whole ethos, because an ethos includes
values of greater generality (e.g. kindneas) and also contains values
not of religious origin (e.g. loyalty) which may not be moral values.
These values of secular origin, of which I judge the most important
to be honour, lie outside the scope of this study, but they are there
in the background, as it were; like other peasant societies, the
Sinhalese villagers experience shame as well as guilt, and perhaps
fear the former more than the latter. With this in mind we in-
vestigate the knowledge and practice of the Five Precepts.
   The Five Precepts, preceded by the Three Refuges, are taken so
frequently that they may be called the functional equivalent of the
Lord's Prayer. They are repeated after a monk at the beginning of
all <1pi#nkam>1 and again at odd moments throughout. This raises the
question in the minds of a spectator whether their recitation has
not lost its meaning. Plainly any religious community anywhere
in the world has not always got its mind on the texts it repeats;
but how far has this gone? One monk told me that one always had
the duty to keep the Five Precepts, whether one recited them or
not; but would this be clear to ordinary villagers? I accompanied a
group of some sixty villagers from Mi$gala on a pilgrimage to
climb Adam's Peak. At frequent intervals in our progress we
repeated the Five Precepts after a <1dasa sil upa$sika$>1, the person of
highest religious status present. On the way back, being tired and
impatient, I remarked to someone at a halt for such a recitation
that I was not going to join in this time, because my last recitation
of the Five Precepts had been so recent that I could not yet have
broken them even had I wanted to. My remarks were passed on
and discussed, and to my surprise it was soon decided--without

254

any counter-argument--that I was right and another recitation
would be pointless. The Five Precepts were not taken again for the
rest of the trip.
  The only systematic investigation of knowledge and interpreta-
tion of the Five Precepts of which I know was conducted by Obey-
esekere and his research assistants in a remote Kandyan village.
They asked twenty-two adults to recite the Five Precepts (in Pali,
of course); all but one were able to do so. The informants who
could recite the precepts were then asked for the meaning of each
precept. The number who were able to aasign meanings to the
undertakings they had given in Pali can be tabulated thus:

                    Number of        By how
                    precepts           many
                    understood        people
                         5                5
                         4                3
                         3                5
                         2                0
                         1                2
                         0                6
                                               --

                                         21

  The precepts recited were thus understood about half the time.
However, this question turned out to be of limited significance,
a test rather of linguistic than of ethical knowledge; for when
questioned in a different context everyone knew about the ideas
expressed by the Five Precepts, and moreover identified them as
Buddhist values.  My own unsystematic observations would
entirely support these findings; the Five Precepts are almost
universally known by rote, and their meanings also are generally
known and understood, though the specific meanings of the Pali
words are often not known--but this knowledge is on the increase
because of <1daham pa$sal>1 (Sunday schools) and religious instruction
in state schools.
  Before considering specific precepts a remark concerning their
form may be in place. Obeyesekere has been at pains to stress, in
accordance with Buddhist tradition, that the Five Precepts are
not commands like the Hebrew "Thou shalt not', but undertakings
in the first person. Cognitively this is indeed a crucial point. Yet
even here I doubt whether doctrine matches emotional reality.
                      255

When monks in sermons allude to the precepts, as they very
often do, they express them in the form of negative commands,
e.g. "Do not kill' (<1marant*a epa$>1). They can answer that strictly they
are speaking to their audience, not <1translating>1 the precepts; but I
do not think that is the impression that they leave the layman
with, and I strongly suspect, though he does not say so, that
Obeyesekere elicited from his informants the meanings of the
precepts expressed not as undertakings but as prohibitions.
  Obeyesekere's informants seem to have equated sin (<1pav>1)
with transgreasion against the letter of a precept. At first sight
this looks like subscription to an ethic of works, not of intention;
but it turns out that some things were said to be sins but neverthe-
less "all right'. To explain this we must take a closer look at his
data on the interpretation of the precepts.
  The precepts against killing and stealing are unambiguous and
their interpretation was not probed. Even a professional hunter
said killing was always a sin. The precept against sexual mis-
conduct (<1ka$mesu miccha$ca$ra>1) has been variously interpreted in
different Buddhist societies, which is quite in keeping with both
its letter and its spirit. Everyone asked agreed that adultery was a
sin (<1pav>1), though some said that it was much less sinful if com-
mitted with the knowledge of the spouse. Some said that pre-
marital intercourse was a sin, though most of these added that it
was not a sin if the parties subsequently married. A few said that
adultery in secret was a sin but all right, and the same was said of
premarital intercourse.  One man, who admitted he was an
adulterer, said adultery was a sin but all right. He went on to say
of another man that he claimed to be a good Buddhist because he
observed the Five Precepts, but he practised anal intercourse with
his wife and therefore was observing the letter, but not the spirit
--"a "good Buddhist" but a sodomist' (Obeyesekere's report of the
statement).
  Lying is sin, as it is against the fourth precept. Asked whether
they would lie in court to protect a kinsman accused of murder
some said that under those circumstances it would not be sin,
others that it would be a sin but they would do it. Of the fifth
precept against alcohol there was a more flexible interpretation:
seven said that to take alcohol was always a sin; seven that it
might be taken without sin for medicinal purposes ; six that it was
not a sin taken in moderation.
256

  As the Pali words of the fifth precept can be taken to mean
abstention from strong drinks <1which>1 are the occasion of intoxication
and carelessness or <1when>1 they are the occasion for intoxication and
carelessness I would not like to say which interpretation of it
answers even to the letter of the law, let alone the spirit, so this
evidence had best be disregarded. The case of the man who
judged anal intercourse permissible according to the letter of the
precept, but nevertheless wrong, is to some extent a red herring,
because I think he misjudged the case: anal intercourse would
come under "sexual misconduct' no leas surely than adultery if
(as in the Sinhalese case) society disapproved of the practice. The
quotation was worth including because of the insight it gives into
how a villager thinks that <1other people>1 interpret the precepts. These
two cases apart, the problems raised by the (regrettably sparse) data
are what is meant by "wrong but all right' (said of secret adultery),
and what attitude informs "a sin, but I'll do it' (said of lying to
save a kinsman). (In the context I take "wrong' and "sin' to be
synonyms, as they would be in Sinhalese (<1va$radi>1 and <1pav>1).)
The two statements seem to me to have quite different implica-
tions. Secret adultery may be all right because one will get away
with it, or because "what the eye doesn't see the heart doesn't
grieve over', so nobody is hurt by it. The former reason embodies
a prudential attitude taken by someone who does not really believe
in <1karma>1, the latter a utilitarian ethic which judges actions solely
in terms of their effects on others. It is rather the latter inter-
pretation which seems to fit the case when premarital intercourse
is said to be a sin but all right. The latter interpretation is, more-
over, clearly the only one which fits the case of the lie to save a
kinsman; here there is clearly added the notion that the morality
of an action varies with its context so that literal adherence to
moral precepts may not be the most moral course. (I prefer this
to talking of "the ends justifying the means', a phrase which may
cover much confusion.) This idea of a higher morality behind
the letter of the law is implicit in the whole doctrine of intention,
and does not jibe well with the utilitarian ethic which measures an
action by its effect on others.
   Though killing is always a sin, is it equally bad to kill a man and
a louse? After all, both are transmigrating beings who in a few
years might change places; on the other hand, a man's life seems
intuitively more valuable than that of a louse. As I mentioned in
                       257

Chapter 4, there is depicted in a local temple a special hell reserved
for people who kill lice; but I think you would have to go a long
way to find someone who took this seriously. A monk explained the
situation to me: the amount of <1sin (pava>1) varies with the intention
(<1ce$tana$va>1). It is equally bad to kill an elephant or an ant, a good
man or a bad, if the intention is the same; there is a different feeling
about it in society, because of the different values of the things
killed, but from the religious point of view both are the same
(<1a$gama vas/ayen dekama ekayi>1).
  One of the questions I frequently raised in my interviews was
the morality of Dut*uga%mun*u in making war on the Tamils for
the sake of Buddhism. In considering the answers one must bear
in mind that in the <1Maha$vam*sa>1, where the events are recounted,
Dut*uga%mun*u is presented as the national hero and his actions
are justified, so that to condemn him runs counter to tradition
and requires an independent mind. Though the answers varied
slightly, their general tenor is well illustrated by the reply of the
monk just quoted, who said that Dut*uga%mun*u's killing of Tamils
was sin, but not great, because his main purpose (<1parama$rtha>1)
was not to kill men but to save Buddhism; he did not have full
intention to kill. But to say that he will not pay for his sin (<1ahosi>1
<1veyi>1) is wrong. This last point I have discussed in the previous
chapter. The rest of the answer, which here concerns us, seems
clearly to bring out two points already evident in Obeyesekere's
village material. On the one hand killing is always, intrinsically,
<1pav>1. On the other hand, what really counts is the intention;
Dut*uga%mun*u's intention was good, so his sin was "less' (<1pav>1
<1ad*uyi>1). This latter opinion can also be expressed by saying that
his main purpose (<1parama$rtha>1) was good. One monk actually
quoted the <1Maha$vamsa>1 to the effect that Dut*uga%mun*u was told
by <1arhats>1 that it was right for him to fight the war; he <1thought>1
(the monk said) that because of his ultimate purpose he was in the
right, though the speaker disagreed with him. However, it is
the intention that counts:  what Dut*uga%mun*u had in mind was
not killing men, but the raising of religion (<1a$gama na$ga hita>1
<1vi$ma)>1.
  This monk was the only one to say that in his opinion Dutu-
ga%mun*u was wrong. Only two monks on the other hand went so
far as to say that he did <1not>1 sin. Most extreme in expression was
the monk who said that Dut*uga%mun*u did not sin in killing the
258

Tamils, as they had wrong views (<1mithya$ dr*s*t*i>1) [i.e. were not
Buddhists]. It is always wrong to kill good people (<1si$lavanta aya>1),
but not wrong to kill in order to save religion; killing in self-
defence is also problematic. Though this sounds rather ferocious
I should add that this monk seemed a kindly old person who went
on to say that one must do one's best to spread loving-kindness
(<1maitri$ paturuvant*a>1): he was, however, exceptionally unsophistic-
ated. The other monk to say that Dut*uga%mun*u did not sin
brought the subject up spontaneously; though killing is a sin, he
said, if Buddhism is in danger it is no sin to kill in its defence--as
did Dut*uga%munu.
  This last answer exemplifies the ambiguity of the situation.
Killing is <1pav>1--but sometimes it is not. Plainly the word is being
used in two different ways; and these two ways correspond to the
formalistic ethic of effect and the ethic of intention. According to
doctrine it is only the ultimate purpose which counts, so that the
intention must be the yardstick by which to measure right and
wrong; but the word <1pav>1 has come also--perhaps primarily--
to mean "sin' in the sense of transgression against the letter of the
Five Precepts. The answers of my informants who merely said that
Dut*uga%mun*u did "sin' do not therefore reveal whether the
speaker thinks Dut*ugaa%mun*u was right or wrong to act as he did.
At the time I did not fully realize the ambiguity of the term "<1pav>1',
so I did not probe more deeply, but to do so would have been very
difficult because the double meaning of <1pav>1 is not explicitly
realized by the monks. In fact I think that they would have split
about evenly for and against Dut*ugaa%mun*u ; but the results would
have been tendentious, in that without my asking them they
would never have considered the matter so deeply. (On the danger
of "pistol point' answers see Introduction, pp. 37-8.)
   Should any distinction between "sin' and doing wrong seem too
abstruse an extrapolation from the evidence, perhaps it will
appear justified when we turn from the negative to the positive
side of ethics. Here the accepted English translations of Buddhist
terminology already seem to acknowledge a distinction between
formalistic and intentional ethics, for <1pin>1, which is undeniably
the exact opposite of <1pav>1, is translated not "right' or "good', but
"merit', and <1pinkama>1, or its equivalents in Sanskrit and Pali, is
translated "merit-making'. I contend that "merit-making' has
formalistic connotations which are in many circumstances quite
                      259

correct; sometimes, however, the translation is unnecessarily
pejorative and "doing good' or "doing right' would be more apt.
Again, according to theory it is only "doing good'--the pure
intention--which counts; but in practice the performance of
certain prescribed acts seems often to be held--in Buddhism as in
 other religions--sufficient to ensure rewards.
   In the absence of statistical evidence for Ceylon and indeed
of comparative data it would be invidious to attempt an assess-
ment of the extent to which the Five Precepts are observed.
Robert Knox, who lived about 20 miles from Mi$gala in much the
same cultural area, gave a character of his fellow-villagers which
strikes me as quite up-to-date and applicable to my own ex-
periences. He is so impreased by their mendacity that he makes
three references to it within a couple of pages: "For they make no
account or conscience of lying, neither is it any shame or disgrace
to them, if they be catched in telling lyes: it is so customary';<s13>s
later qualified by ". . . approving lying in themselves, but mis-
liking it in others'.<s14>s On the other hand "Of all vices they are least
addicted to stealing, the which they do exceedingly hate and
abhor; so that there are but few robberies committed among
them. They do much extol and commend Chastity, Temperance,
and Truth in words and actions; and confess that it is out of
weakness and infirmity, that they cannot practice the same, ack-
nowledging that the contrary Vices are to be abhorred, being
abomination both in the sight of God and Men. They do love
and delight in those Men that are most Devout and Precise in their
Matters. As for bearing Witness for Confirmation in any matters of
doubt, a Christians word will be believed and credited far beyond
their own: because, they think, they make more Conscience of
their words.'<s15>s
  The difficulty is, of course, that neither Knox nor I have been
similarly exposed to other peasant societies for comparison.
I did not have the impression that killing, stealing, sex crimes,
or drunkenness were more prevalent than in any other com-
munity with which I am acquainted. Statistics do of course exist
for serious crimes such as homicide, but these can only constitute
a tiny part of the percentage of infringements of the precepts.
The homicide rate in Ceylon is notoriously high, but it varies so
260

much between different parts of the island (being quite low in the
Central Province) that even this is not a sound basis for generalized
judgements. Theft is very much feared by the villagers, as may be
the case in all peasant communities; we were constantly being
advised to lock up all our possessions, and thieves were spoken of
with much vehemence and expreas disgust. The proximity of
Buddhist institutions has no deterrent influence on thieves. One
must enter a Buddhist temple barefoot, leaving one's shoes at the
entrance; it is notoriously risky to leave shoes unguarded at the
entrance to the Temple of the Tooth in Kandy. My own sandals
--far too large, one would think, for Sinhalese feet--were ab-
stracted from the entrance to a village temple during a <1pinkama>1;
after a fruitless search the horror-stricken incumbent confessed
to me that he himself had had a pair of new sandals stolen the
previous week from his very bedroom in the <1pansala>1. I had the
impression that sex crimes were infrequent; on the other hand
Obeyesekere reported from his remote (and <1ipso facto>1 untypical)
village that nearly all the adult males there had committed adultery
when opportunity arose. (We already saw that adultery was not
very serioualy considered by some members of the community).
Drunkenness seems no commoner than in rural England, probably
less.
  But it will be thought that in one respect the conduct of Sin-
halese Buddhists must differ very sharply from that of westerners :
in the killing of animals. There is a prevalent misconception in the
West that Buddhists are vegetarians. Yet there is no Buddhist
country of which this is true on the societal level: vegetarianism
is a minority practice freely chosen by individuals, both cleric
and lay, much like abstention from alcohol inthe West. The Buddha
did not forbid meat-eating. In the <1Vinaya>1 rules monks are for-
bidden ten kinds of meat, ranging from human flesh to lizards,
but none of the creatures normally consumed are mentioned.
The only rule is that a monk may not eat an animal killed specially
for him; but if it is already dead, he may as well eat it. The Buddha
himself is reported as eating meat, and it seems that his last meal
was pork.
  The result of this tradition is that villagers, especially in isolated
communities, occasionally boycott butchers' shops, so that no
cattle or goats can be killed in the area; in this way they are
directly saving lives. I believe that such boycotts tend to be
                      261

organized by monks. Mere abstention from meat is, however,
considered useless. Nor must one exaggerate even the reluctance
to kill animals. To be a butcher is considered wrong for a Buddhist,
and most of the butchers in Ceylon are indeed Muslims. Villagers
in my area raised goats which in due course they sold to Mualims.
That the Muslims happened to be butchers who then killed the
animals they could regard as not their fault; on the other hand I
encountered reluctance to admit to the situation and was told that
the goats were raised as pets. Once again <1pav>1 seems to be the
literal transgression of the precept, but this interpretation does
not satisfy the conscience. In remoter areas villagers go hunting,
and I am sure they make no secret of it. As they themselves would
say, they are forced to it by poverty; the extra twist which they
give is that their poverty is due to sin in previous lives, so previous
sin produces conditions for further sin, and they are caught in a
vicious circle from which only the exceptional person is likely to
escape.
 Vegetarianism I found universally admired, but rarely practised.
There are, moreover, degrees of vegetarianism. It is worse to eat
beef than other meat (which in practice means chicken or goat-
meat), worse to eat meat or fowl (<1god*a mas>1) than fish (<1diya mas>1);
eggs seem to be worse than fish, but not as bad as meat. A true
Buddhist vegetarian eats no eggs; even educated people think that
all eggs are fertilized, and when I tried to tell people that a hen
could lay without prior assistance from a cockerel, I was never
believed, but probably considered crazy. To break an egg is to
take life, which means that in one way eggs are worse than any
meat, because the "killing' is committed on the spot by the cook.
Bareau reports that eggs are not used in monastery kitchens; for
similar reasons middle-class urban Buddhists tend to buy "Bud-
dhist eggs', which are already cracked. The hierarchy of value of
meats appears to me to relate to the size, value, and apparent close-
ness to humanity of the animal slain, eggs counting roughly as
chickens. This may also be part of the original reason for the special
disfavour of beef-eating. In terms of immediate history this atti-
tude is due to cultural contact with the Hindus; in 1815 the British
found that among the Kandyans beef-eating was tabu, as it still
is among orthodox Hindus. (This does not mean that either in
Ceylon or in India beef was never eaten by poor and remote
villagers.) Taking a longer historical view, the Hindu tabu on
262

beef-eating owes something to Buddhist influence, for the Buddha
drew attention to the plight of the cow when beef-eating seems to
have been general. Explanations given by Buddhists for the dislike
of beef-eating, where forthcoming, tend to be a mixture of the
economic (buffaloes are useful) and the moral (cows give us milk,
so we should not be so ungrateful as to kill them). (Goat milk is
rarely drunk.) I have even heard the psychological argument that
at they give us milk, are like our mothers. Villagers are
well aware, however, that Buddhists may eat beef but Hindus
may not; a lady told me that before visiting a Hindu temple to
Kataragama one had to abstain from beef for a week, and ingeni-
ously explained this by saying that Kataragama would be angry
because a bull (Nandin) is his father's (S/iva's) mount.
   In killing smaller animals Buddhist villagers seem to behave
much the same as other villagers the world over, and do not
display the compunction or squeamishness sometimes found in
the urban middle class. I came upon a monk who with the help of
the young temple servant (<1a%bittaya$>1) was clearing out a storeroom.
A cockroach ran out. "Kill it' (<1marant*a>1) he shouted. The boy,
however, who was piously minded, swept it over a small precipice
and thus saved its life. Monks will also slap at insects in a way
scandalous to pious Hindus. Not that they would ever deny this
to be a sin--they just don't care that much. The one animal which
villagers have great compunction about killing is a cobra; as
mentioned in Chapter 4, there is a belief that the cobra is an
animal of special worth and dignity, and to injure one is considered
unlucky.
   We are now in a position to hazard some remarks on ethos, or at
least on religiously inspired moral values. Ethos I take to be
something on the affective level--sincerely held high-level values,
whatever the <1fable convenue>1. Frequent failure to live up to an
expressed ideal may make us suspicious whether that ideal
represents an affective value, but is not conclusive: discrepancy
between theory and practice may result simply because the ideal
is so diflicult. Truth I would hardly describe as a <1major>1 value
in fact--affectively--however much lip service is paid to it. The
very frequency in villagers' conversation of the sentence "Boru
kiyant*a hon#da na%$' ("Lying is bad'), usually uttered with a light
intonation in the context either of mild accusation of a third party
                      263

or of protestation of one's own sincerity (compare the English,
"To be perfectly frank . . .', which regularly precedes a lie), may be
offered as evidence for the frequency of lying; while its tone
suggests that the offence is not really considered heinous. Lying
is bound to be frequent in a culture much concerned with the
preservation of status (<1tattvaya>1) and dignity (<1nambuva>1)--saving
face; the most trivial matter which might in any way appear dis-
creditable to the speaker is concealed almost as a matter of course.
Finally, although <1religious>1 truth, the Dharma, is of course a
matter of paramount importance, the idea of truth as an autono-
mous secular value, the truth which scholars seek or are supposed to
seek, has no firm base in village society. The <1Kalama Sutta>1 enjoins
each man to seek the truth for himself, but here too the reference is
to <1religious>1 truth conducive to salvation, not to truth as an absolute
value, and the passage is remarkable rather as a charter for toler-
ance and individualism--but in any case the <1Kalama Sutta>1 and
similar texts are not generally known to villagers. Truth is a
value, certainly, in so far as <1telling>1 the truth is enjoined by the
fourth precept, and <1seeking>1 the truth of a particular kind is an
important aspect of meditation; but where truth comes into con-
flict with more important values--honour, loyalty, kindness--
it can be lightly discarded.
  Piety receives less lip service than truth, but affectively I would
judge it to receive about equal weight: not a major value, but
definitely admired. It is conceived as a character trait rather than
as an attribute of a particular speech or action. The adjective
used is <1bhaktimanta>1, derived from <1bhakti>1, "religious devotion',
the key concept of Indian devotional religion; besides <1bhaktiya>1
the noun <1sardha$va>1 (see p. 59) is used in a closely allied sense.
Both nouns refer to the kind of faith, of loving trust, which is
recommended by theistic religions, for all that according to some
authorities this is supposed not to exist in Therava$da Buddhism.
The sentiment in Sinhalese Buddhists is certainly a mild one
compared to, for instance, the ecstasies of their S/aivite neighbours,
and its focus is rather difluse: the devotion and trust are chiefly
directed towards the Buddha, but may include other religious
characters such as the great disciples or good Buddhist gods like
Vishnu, venerable objects such as Bo trees or books of scripture, or
even particular monk. Though such veneration is accounted for in
rational terms, as discussed in earlier chapters, the religious feeling
264

shown by some pious laity--the sort of people who at funerals are
accorded the title of <1upa$saka>1 whether or not they took the Ten
Precepts--goes beyond what reason or doctrine would warrant.
Nevertheless, no one would <1admit>1 to going beyond reason, in
contrast to the Christian, whose faith beyond reason is specifically
commended.
     Christianity speaks of three cardinal virtues: faith, hope, and
charity. Buddhists have no similar list of virtues in common use,
but--in keeping with the tendency to negative expression--they
have a very commonly used list of three vices. These are <1lobha>1
(or <1ra$ga>1), <1dosa>1, and <1moha>1, commonly translated as greed (or
passion), hatred, and delusion. One monk translated them into
Sinhalese as <1a$sa$va, taraha>1, and <1mo$d*akama>1, which are naturally
tranalated as desire, anger, and stupidity. Of these three there is
no doubt that the most cited and the most abhorred is desire.
When a neighbour wanted to convey to me his disgust for another
man he said he was <1lo$bayi>1 (greedy); there is no other single word
which carries such a weight of moral disapprobation. When our
chief monk heard that some men had gone out ploughing on
Wesak--a day of all days on which one should avoid killing even
worms--he got really angry. I suggested that it was just stupidity.
No, it was desire (<1a$sa$va>1). Remember also the peculiar horror
(noted by Knox) of stealing. Though the accusation of <1lo$ba>1
on the lips of poor villagers often contains more than a modicum
of jealousy, the coincidence with the second Noble Truth--that
desire is the root of all evil--is striking. In practice Knox found
the Kandyans "near and covetous',<s16 which he seems also to have>s
been himself; and certainly it is hard for the poor not to care.
Nevertheless I make bold to say that freedom from attachment to
worldly goods is a true Buddhist value.
  The question of anger is more complex. A harsh, irritable, or
aggressive person is characterized as <1sa%rayi>1. Like <1boru>1, the word
<1sa%ra>1 is very frequently and lightly used. It may even carry a
connotation of approval, like the English word "tough'. This
approval even became explicit in one interview: a monk told me
that though one should generally be <1s/a$nta da$nta>1 (mild and re-
strained--see below) there are times when threatening (<1tarjana>1
<1karanaya>1) is necessary and one has to be tough (<1sa%rayi>1)--though
one should never carry this beyond words. This limitation is
                       265

important: violence is certainly condemned; even the censorious
Knox wrote, "They are not very malitious one towards another;
and their anger doth not last long; seldom or never any blood shed
among them in their quarrels. It is not customary to strike; and it
is very rare that they give a blow so much as to their Slaves; . . . <s17>s
Moreover, the most highly esteemed form of human activity,
meditation, is passive (however strenuous) rather than active.
Passivity in the West carries rather negative connotations which
it certainly does not possess in Sinhalese society, and children are
brought up to behave even more passively than long physical
dependence on the mother might warrant. Yet inborn human
aggression cannot vanish without trace. Compared to western
children, Sinhalese village children fight and hit each other far less;
but in playing with smaller children or western children (mine),
when it is doubly tabu to hit, they give them potential weapons
such as sticks and stones and court attack; when duly attacked
they then sometimes go and complain to authority. Such complaint
of course provides vicarious satisfaction of aggressive impulses,
also common in the West; but it is by no means always resorted
to, and the courting of attack by others is clearly masochistic.
This masochism seems to me to find expression in the gruesome
scenes depicted in many temples. Not only are paintings of tortures
in hell fairly common; a couple of famous temples have permanent
exhibitions in sculpted tableaux of old-time Sinhalese tortures,
and these are visited by tourists from all over the country. One such
torture shown in three-dimensional lurid colour is the impalement
of a man on a sharp pole which enters the anus and comes out
at the chest. The flimsy pretext for these displays, which are highly
profitable, is that they illustrate the perils of wrong-doing. What
chiefly distinguishes them, to my mind, from their western
equivalents such as Madame Tussaud's Chamber of Horrors, is that
in the West the audience is generally supposed to identify with the
executioners who are giving the victims "their just deserts', whereas
the Buddhist shows explicitly identify the spectator--you miserable
sinner--with the sufferer in hell or on the stake. The spectator's
psychology need not comply with this prescription, but my own
conclusion (which might be empirically tested by a researcher with
a strong stomach) is that these horrible spectacles are outlets for
suppressed aggression, but primarily in masochistic rather than
266

sadistic form.<s18>s Aggression, in particular physical violence, is thus
strongly condemned in Sinhalese society, and its absence is a
high-ranking Buddhist value.
  The word <1mo$d*a>1 means "foolish' or "stupid' and carries about the
weight of the latter. The Indian cultural tradition, very alien to
the western Christian, strongly associates intelligence with virtue.
The highest of the three stages of ideal Buddhist conduct is
<1pan+n+a$>1, wisdom, and this is given a fairly intellectualist interpreta-
tion in Sinhalese tradition: study of the <1abhidharma>1 is these days
generally considered the highest-level prerequisite for Enlighten-
ment (see <1vipassana$>1 meditation in the next chapter). Education
and learning, however formalistically conceived, are thus not
merely technical accomplishments but morally positive. In con-
demning a monk who was in fact disreputable for quite different
reasons a man dismissed him as <1ugat madi>1, "insufficiently learned'.
Learning, too, is a Sinhalese Buddhist value.
  I listened to people's judgements on their fellow men, and in
particular to funeral orations. The two main values expressed
seem to be kindness, especially as expressed in liberality, and
above all moderation. There is no doubt that the most frequent
phrase used to convey approbation of someone is <1sa$nta da$nta>1, nor
is there any doubt that this is a pure Buddhist value. <1S/a$nta da$nta>1
are two Sanskrit past participles for the roots <1s/am>1 "to appease'
and <1dam>1 "to tame'; together they express quiet self-control.
Appeasement of the passions, taming of the senses--these were
the Buddha's constant themes. The oldest and most famous books
of Buddhist religious poetry, the <1Dhammapada>1 and the <1Sutta-nipa$ta>1,
repeat these phrases and their synonyms so often that at times they
seem to consist of little else. <1S/a$nta da$nta>1 continually occurs at
every level of conversation. It is also a value in politics. Our village
neighbours, who were supporters of the U.N.P., whose party colour
is green, taught their children to sing,

             <1Ape$ pa$t*a kol*a, api na%$ kalabola>1.
             Our colour is green, we are not rowdies.

The standard criticism of the other side which we heard from
U.N.P. supporters was that they scolded too much (<1baninava$ va%d*i>1).
                       267

The <1s/a$nta da$nta>1 man (as they all conceived themselves) keeps a
civil tongue in his head. <1S/a$nta da$nta>1 is non-aggressive (<1sa%ra na%$>1),
but also something more. He is polite, in the fullest sense.
  The other word most frequently used in praise of someone is
<1karun*a$vanta>1, "kind', "compassionate'. This, too, is a truly tradi-
tional Buddhist value. I mentioned above the four holy states
(<1brahmaviha$ra>1) in the Canon, of which this is one.  Though
scholastically <1karun*a$>1 is differentiated from <1maitri$>1 (loving-kindness)
in that <1karun*a$>1 applies only to people in misfortune, in Sinhalese
<1karun*a$vanta>1 governs both meanings, like English "kind'; nor is it
an accident that the most popular meditation among monks seems
to be the meditation on kindness (<1maitri$ bha$vana$va>1), in which one
is supposed to suffuse the world with one's kind thoughts.
  The clearest mark of kindneas is generosity, and no one with
any experience of Sinhalese villages will demur at my saying that
to westerners their generosity borders on the incredible. There are
of course two sides to this, as I have already indicated: giving has
become institutionalized to a high degree, not only as a means of
making merit, but in more obvious forms of exchange and purely
social relationships. A westerner is also struck by the preference,
explained above, for giving to a wealthy temple rather than to a
poor layman. Nevertheless sheer generosity of both goods and
effort (known in a formal context as <1s/amada$na>1) is abundant, and
disregards the letter for the spirit. Westerners are not favourably
impressed by the famous <1Ja$taka>1 story of King Vessantara. This
was the last birth of the Bodhisattva before he became Gotama
Buddha, and he was to achieve in it the supreme quality of liber-
ality (<1da$na pa$ramita$va>1). To this end he gave away everything, even
his wife and children. This strikes us as excessive. It strikes the
Sinhalese in the same way. The two monks with whom I brought
up the subject both said that Vessantara was <1wrong>1. Generosity
is very well, but even there one must exercise moderation.
  Is there then any difference between the ethos of Sinhalese
Buddhists and that of other civilized communities? The modera-
tion and self-restraint of <1s/a$nta da$nta>1 are perhaps not very far from
the Greek <2kalo\s kagaqos>2, with his <2mhden agan>2 and <2sofrosunh>2,
or from the English gentleman ideal. But when we consider the
three cardinal virtues of Christianity we come to a startling
contrast. Faith we discussed above; the Christian ideal of faith
beyond reason is certainly not a Buddhist virtue, and might even
268

be considered a vice. What about hope? There is no Sinhalese
word for hope. There are words for expectation, but none for the
mixture of expectation with longing which we mean by "hope,<s19>s
The Sanskrit word usually translated "hope' is <1a$s/a$>1, the very word
that in Sinhalese is most commonly used for desire! In so far as
hope is desire, it is the supreme Buddhist vice! Only charity--
<1karun*a$va/maitri$>1--is a Buddhist value and that indeed is all-
important.


                              7
                    THE MONASTIC IDEAL AND THE
                      DECLINE OF BUDDHISM

ANY of the Ten Good Deeds may be performed by anyone, monk
or layman.  However, two of them, teaching (<1desana$>1) (which
includes preaching (<1ban*a>1)) and meditation (<1bha$vana$>1), are more
typical of monks; laymen may do either, but traditionally laymen
only teach or preach when monks are unavailable, and they gener-
ally preach merely by reading or reciting the scriptures. I came
across cases where monks through infirmity or laziness delegated
their preaching duties to lay <1upa$sakas>1, but this is not standard
practice. Similarly, meditation is recommended for all, but it is
recognized that in practice laymen will lack the necessary time and
tranquillity. On <1poya>1 days laymen who take the Eight Precepts are
expected to meditate, and the principal goal of retirement late in
life as a <1dasa sil upa$saka>1 or <1upa$sika$>1 is also meditation. As the Eight
Precepts are usually taken communally by a group of people up at
the temple there is probably not much opportunity for meditation
then; but as meditation is virtually the sole purpose of taking the
Ten Precepts I have little doubt that <1dasa sil upa$sakas>1 meditate,
probably more than most monks.
  It is sometimes assurned that meditation is the ideal for monks,
more active religiosity for laymen; but this is a misconception.
Given that everyone should meditate if possible, both laity and
clergy are in this respect divided into two groups of unequal size.
Just as laymen are in practice divided into the small minority of
meditating <1upa$sakas>1 and the large majority of people who do not
meditate, so monks are divided into two categories with the
primary duties of teaching and of meditation, and the former group
is many times the size of the latter. Those who are primarily
supposed to teach are called <1granthadhura>1 (Pali: <1ganthadhura>1)
(literally "having books as their burden'), and the others are called
<1vidars/anadhura>1 (Pali: <1vipassana$dhura>1) (literally "having medita-
tion as their burden'). These categories are first found in the
commentaries, and the distinction probably does not antedate
270

the Christian era.<s1>s In theory all are supposed to do some medita-
tion, just as all are supposed to preach if requested, but what is
theoretically a mere emphasis tends in practice to become an
absolute distinction.  There is another traditional distinction
which might be supposed to correspond to that between <1grantha->1
<1dhura>1 and <1vidars/anadhura>1, namely the distinction between village-
dwelling (<1gra$mava$sin>1) and forest-dwelling (<1a$ran+n+ava$sin>1 or <1vana->1
<1va$sin)>1 monks; however, the categories do not in fact correspond
at all. The former division is still taken to mean what it says, while
the latter dichotomy has for the most part lapsed into meaningless
scholasticism. There are still a few monks who live in forests,
while most monks live in villages (or towns), but village dwellers
are often called <1a$ran+n+ava$sin>1. The theory is that in the Siyam
Nika$ya Malvatta monks are <1gra$mava$sin>1 and Asgiri monks are
<1a$ran+n+ava$sin>1. This ascription dates from the re-establishment in
1753 of these two monasteries and the branches (<1pa$rs/va>1) they
head; it continues the tradition of a dichotomy formalized by
Para$kramaba$hu I when he unified the Sangha.<s2>s In fact, however,
all Asgiri monks (so far as I know) live in villages just like Mal-
vatta monks. At least within the Siyam Nika$ya the status of
<1a$ran+n+ava$sin>1 has thus become meaningless and is not really a
matter of choice. On the other hand whether a monk is <1grantha->1
<1dhura>1 or <1vidars/anadhura>1 is largely a matter of individual choice,
and I met a monk who was <1vidars/anadhura>1 (= <1bha$vana$ma$rgaya>1)
though living in a village as an ordinary parish monk. In theory
monks are supposed to go through both stages, passing from
book-learning to meditation, but this happens only in exceptional
cases. The meditating village-dwelling monk had entered the
Sangha as an adult adn had not received a monastic education;
these circumstances, though unsual, are catered for in both theory
and practice, and such a monk is not expected to undertake the
<1granthadhura>1.
  There is a conventional tripartite division of Buddhism into
learning (<1parya$pti>1), practice (<1pratipatti>1 and realization (<1prative>1-
<1dana$>1). Each part is higher than the preceding but also depends
upon them. <1Parya$pti>1 means the preservation of the scriptures;
          271

though this is a lowly thing, it is the basis for everything higher,
and thus indispensable. <1Granthadhura>1 monks are the vehicle for
<1parya$pti>1; their function is to hand on the doctrines by teaching
other monks and preaching to the laity. (That they are also
supposed to practise what they preach (<1pratipatti>1) and strive
towards Enlightenment (<1prativedana$>1) goes without saying.) Some
western authors have given the impression that Therava$da
presents the "selfish' ideal of attaining one's own Enlightenment
and that a monk has no duty but to strive for his Enlightenment;
according to them it would therefore be anomalous to speak of a
monk's "function'. Whatever happens in practice, they are right to
the extent that a monk is not traditionally conceived as having
pastoral duties towards the laity as individuals under his care; but
monks--or at least most of them--do nevertheleas have the
function, formalized as <1granthadhura>1, of preserving religious
tradition. It is with this same context in mind, incidentally, that
<1suti>1, listening to religious teaching, is listed as one of the Ten Good
Deeds.
   I was interested in finding out the extent to which monks actually
do teach and preach, at least in a formal context. The role of the
village monk as a teacher of the next generation of monks has
largely been superseded by monastic schools, the <1piriven*as>1. Out
of thirty-five monks I questioned on this point, only eight had
received their education entirely from their teachers, i.e. the
monks who gave them the lower ordination; the rest had attended
a <1piriven*a>1, all but one of them for more than a year, typically for
the several years between entry into the Order and the higher
ordination (<1upasampada$>1) at the age of twenty or twenty-one. (I
disregard any lay education received before entering the Order--
this never amounted to more than a few years at the village school.)
Most of the nine monks who had been educated entirely or nearly
so by their own teachers were elderly monks of the Siyam Nika$ya,
but one was in the Amarapura Nika$ya, nor does the practice
seem to be entirely dead: one young monk, not interviewed and
so not included in the above figures, whose higher ordination
ceremony I attended, had been educated in his own monastery
by a monk who had sent his previous three pupils to <1pirivenas>1.
The <1piriven*as>1 are new but not untraditional institutions, being
modelled on monastic schools and universities of ancient times.
The two most important (and two of the oldest) of the present


272

<1piriven*as>1 are Vidyodaya and Vidya$lam*kara, founded in 1873 and
1875 respectively in suburbs of Colombo. These now differ
somewhat from the run of <1piriven*as>1 in that they are devoted to
higher education and have become largely secularized; though
popularly known as "Buddhist universities' they have no oflicial
religious affiliation and have for some years admitted laymen, and
in 1966 Vidyodaya started admitting girls as well. A few of my
informants had gone on to these universities (none had been to the
purely secular University of Ceylon); the <1piriven*as>1 most of them
had attended could, however, more aptly be described as monastic
schools or colleges than as universities.  Such <1piriven*as>1 have
proliferated only in the last fifty years; before that education with
the teacher had been the norm for several centuries, but it is now
too late to catch more than the remnant of the last generation to be
generally so educated. Monastic education has now become more
standardized, and there can be little doubt that the general level
has risen. One of the monks educated by his teacher made the
unique admission (unique in frankness, not necessarily in ignor-
ance) that he knew no Pali, and a couple of the others said they
knew very little indeed; none of them had more than perhaps a
smattering of Sanskrit. On the other hand all pupils at <1piriven*as>1
are taught Sanskrit and Pali and consequently say that they know
them; of course I could not test the content of their knowledge,
but at least a certain expectation of learning has become normal.
Monks usually take a paternal interest (as the Buddha recom-
mended)<s3>s in their pupils, and finance their education at <1Piriven*as>1,
or get the parents to pay half. When their education is complete
most monks return to their villages; the rest for the most part
become teachers, either at a <1piriven*a>1 or at a lay school, where they
teach Buddhism and other subjects. The role of a monk as a
teacher has thus not disappeared, but has changed with the times
and become more specialized; monks who teach usually do so
professionally, on the same terms (including salary) as lay school
and university teachers.
  Religious education is, however, not confined to <1Piriven*as>1 and
ordinary lay schools: mention has already been made (see es-
pecially Chapter 3, pp. 83-4) of the <1daham pa$sal>1 or "Buddhist
Sunday schools'. These are usually headed and organized by local
monks, though the monks do not necessarily teach in them. Of the
                                                                273

thirty-six villages for which I have data eleven had such <1daham>1
<1Pa$sal>1 (one had two); in another nine the monks said that there had
been a <1daham pa$sala>1 which was now defunct, usually for lack of a
teacher. These figures show, however, that many monks, even
among the <1granthadhura>1, neither personally educate their own
pupils nor teach in an institution, and so probably do not teach at
  It is widely believed by westernized people in Ceylon that
monks preach every <1poya>1 day, just as the Christian clergy preach
every Sunday. There is to my knowledge no good historical or
doctrinal basis for this belief, and my investigations showed it to
be ill-founded. During the ten months in which I was at Mi$gala
the resident monks never preached on <1poya>1 days,<s4>s though on the
most important full-moon <1poya>1 days they arranged for other
monks to come and preach. The amount of preaching in other
villages I could not establish with accuracy, because once I asked
about it there was a feeling that there <1ought>1 to be preaching and so
its actual occurrence was probably exaggerated.  An indirect,
though equally uncertain, way of estimating such preaching is to
ask about <1at*a sil upa$sakas>1 on <1poya>1 days, as a monk would be very
unlikely to preach unless there were several such <1upa$sakas>1 present,
who would be his principal audience. Data given me on both
preaching and attendance of <1at*a sil upa$sakas>1 varied a great deal
from village to village, a variation which my own observation
confirmed. However, a survey of my data suggests that the
average monk preaches at his temple on full-moon <1poyas>1 and a few
other big occasions, such as some opening ceremony, in all maybe
fifteen to twenty times a year; the more conscientious monks
preach also on the no-moon <1(ma$aa) poyas>1; and a few preach twice
on the fortnightly <1poyas>1, or even preach every <1poya>1 (two cases
recorded), maybe twice (one case recorded). Those who preach
frequently tend also to be in demand at other temples; our temple
was visited at Wesak by three monks from the area, all of whom had
several engagements to preach that day. On these festivals a large
part of the village would attend to hear preaching, especially the
sermons after nightfall; but normally the audience was more or
less confined to the <1upa$sakas>1 taking the Eight Precepts at the
temple. These varied in number from about a hundred at Wesak
in temples with a large catchment area (such as mine), to a couple
274
or even none on half-moon <1poyas>1; on a full-moon <1poya>1 the average
temple in the area probably has forty or fifty people taking the
Eight Precepts, and about half that number on the <1ma$sa poya>1.
Most of these are old people, especially women, who outnumber
men taking <1at*a sil>1 by at least four to one; there are also a few
school children, which I believe to be a modern development.
Very few people indeed take the Eight Precepts between the ages of
twenty and fifty, except perhaps at Wesak.
The above figures on preaching do not include <1pirit>1, which
though technically a form of <1ban*a>1 is in a class by itself, and is so
irregular in its occurrence that any small-scale statistics would be
worthless. Nor does it include the essentially private preaching
which accompanies <1da$ne$s>1, especially the ceremonies connected
with death. The sermons of a Christian priest follow not only the
calendar but also the life cycle of the individual. Buddhism as
such, however, lacks sacraments; there is no equivalent of baptism,
no pubertal rite parallel to confirmation or the first communion,
and marriage is solemnized by a secular ceremony.  The only
life crisis with which monks are traditionally associated is death.
The grief and separation which death involves are a suitable
occasion for the exposition of the <1Dharma>1 of sorrow and imperm-
anence; so monks, who were originally refugees from lay life,
have become involved in it at this point. The ceremonies at the
funeral and on subsequent commemorative occasions, described in
Chapter 5 in connection with the transference of merit, are thus
over-determined, providing occasion also for sermons on the
fundamentals of Buddhist doctrine. Funeral sermons are usually
rather short (e.g. ten minutes), and contain a eulogy of the dead
person; the sermon which closes a <1da$ne$>1 (after the monks have
eaten and the merit has been transferred at the water-pouring
ceremony) are similarly short; both categories inevitably consist
mainly of reflections on the transitorineas of all worldly things.
Only the <1ban*a>1 which take place in the evening a week after the
death, or as a replacement for <1pirit>1 at one of the subsequent
commemorations, is a full-scale sermon, lasting perhaps an hour,
and may range over wider topics.
Like <1ban*a>1 on a public occasion, this sort of sermon commences,
very much in the Christian style, with a quotation from the Pali
Canon, which is then expounded and enlarged upon, largely with
the help of further quotations; there is more quotation and less
                                                               275

invention than is usual in Christian preaching, but in my ex-
perience a monk preaching never confined himself to a mere
recitation and paraphrase of a canonical text. While preaching a             
monk sits on a special chair <1(dharma$sama)>1 or at least on a chair
covered with a clean white cloth, and holds a preaching fan. He
may hold the fan in front of his face, which serves, I was told, to
depersonalize the sermon, but this is often neglected. A member
of the congregation, usually an elderly man, takes it upon himself
to act as respondent <1(pratyuttaradenna$)>1; he squats in the front
row, hands folded, and every so often says "Ehe$ ha$muduruvo$',<s5>s
which is as much as to say "Amen' in church; in theory he may
also interrupt and ask for clarification, but I never saw this
done.
  The types of <1ban*a>1 were enumerated for me by one of my in-
formants, a well-known preacher. All the commoner ones we have
already covered: <1mataka ban*a>1 is preaching in the house of a dead
man; <1pamsuku$le$ ban*a>1 is preaching at a funeral; <1sa$ma$nya bana>1
(literally "ordinary <1ban*a')>1 is preaching on <1poya>1 days. <1Dharma->1
<1sam*va$da>1, "doctrinal discussion', is usually carried on in an in-
formal context. <1A$sama deke$ ban*a>1 ("preaching from two seats')
seems once to have been common;<s6>s there was none in my area
while I was there, but one was scheduled for just after I left. In a
<1a$sana deke$ ban*a>1 one monk reads out or recites a text from the
Pali Canon, and the other reads out the <1sanne$>1, a close paraphrase
in Sinhalese; my informant said that the usual text preached like
this is the <1Damsak Pa%vatum suta>1--the first sermon. In ancient
times there were apparently three monks preaching in a full-scale
<1ban*a>1, one reciting a text, one paraphrasing in Sinhalese, and one
giving a sermon proper;<s7>s but I did not come across such an event.<s8>s
My informant did, however, list an <1a$sana hate$ ban*a>1, "preaching
from seven seats', which he said was used for the <1Brahmaja$la>1
<1Sutta>1, the first sermon of the <1Di$gha Nika$ya>1. I assume from the
nature of that text that this was a kind of part-reading perfor-      
mance, one monk taking the role of the Buddha and the others
reciting the arguments of his opponents. This kind of <1ban*a>1 he
276

said was obsolete.<1<s9>s>1 However, my informant was himself respon-
sible for an even more vivid form of <1ban*a>1, a full dramatization of a
canonical story in which the Buddha converts a man-eating
<1yaks*a>1 called Alavaka. The original version of this story is a short
poem with prose preface in the <1Sutta+nipa$ta>1, but the version my
informant had devised took about three hours to perform a nd was a
veritable folk drama. It was called <1Alavaka+yaks*a+damane$>1, "The
Taming of the Devil Alavaka'. He told me that he had himself
devised it with the help of five drummers <1(berava$yo$)>1 from his
village who performed the other parts while he sat on a preaching-
seat facing the audience, with the actors before him, and to
convert the <1yaks*a>1 gave a sermon exactly as he would have done
had there been no play. Whether similar dramatizations of <1ban*a>1
have taken place in the past I cannot say; this one has so far
remained unreported, though it might be said to represent a dis-
tinctive genre of folk-art, and it may have unreported predecessors.
The nearest thing to it in my experience is the <1su$visi pinkama>1,
which others called <1ban*a>1 though my informant did not mention
it in his list; but the <1su$visi>1 is highly stylized, as my account in
Chapter 3 shows. This particular drama is, however, the spon-
taneous creation of a local monk and his parishioners.
   The last kind of teaching <1(desana$)>1 which must be mentioned
is, like <1pirit>1, not idiomatically referred to as <1ban*a>1, because though
logically a form of preaching it is highly stylized and specialized.
This is the <1pin anumodana$l>1<s0>s already referred to in Chapter 5;
we saw then that the original meaning of <1anumodana$>1 is "thanks' or
"grateful acceptance'. In the commentary to the <1Dhammapadall>1
the term is already paraphrased <1anumodana$-dhamma-desana->1
(teaching the doctrine by way of thanks).  In the old sub-
commentary <1(pura$n*a$t*i$ka$)>1 on the <1Ja$takasl>1<s2>s the word <1anumodana->1
is glossed "teaching the doctrine to arouse joy in creatures' <1(satta$n->1
<1am piti-janaka-dhamma-desanam*)>1. Professor Arnold Green has
pointed out to me that there is a popular explanation of what the
monk is doing which similarly goes beyond the idea of mere
thanks: he is said to be causing the donors to feel joy <1pritiya>1;
                                                               277

he is actually doing them a favour, by causing them to forget
the trouble and expense involved in offering a <1da$ne$>1 and rather
to feel unalloyed pleasure at having given, thus purifying their
thoughts and advancing towards <1nirva$n*a>1. In this way his
<1anumodana$>1 is not mere thank, but a <1quid pro quo>1. This explana-
tion is of course compatible on all levels with the previously
mentioned functions of the <1anumodana$>1 as thanks and as an
exhortation to share the merit gained; it provides yet another
example of the over-determination of a ritual act.
  Though the recitation of the <1anumodana$>1 may look to a western
newcomer like the saying of grace before a meal, it lacks even
external resemblance in that it only happens when the meal is
given ; when the monks eat their own food, cooked in the monastery
kitchen, nothing is said. The original practice was for monks to
eat only what they had begged; but when monasteries acquired
possessions this became unnecessary, and when they were ex-
tremely large it also became impractical. In Anura$dhapura, we
learn from Fa Hsien and Hiuen Tsiang,<s13>s vast numbers of monks
and nuns were fed at a central kitchen supplied by the King, i.e.
at public expense. In the tenth century the King granted villages
to the monasteries of Jetavana and Abhayagiri to maintain their
refectories. <s14>s Throughout ancient times, however, very many
monks, perhaps all those outside the capital, customarily went on
the daily alms-round <1(pin*dapa$ta)>1. The <1pin*d*apa$ta>1 became excep-
tional presumably only when the landless monastery became
exceptional, i.e. in the Kandyan period.
  There are today no monasteries except the <1piriven*as>1 so large that
it would be impractical for all the monks to beg locally; but few are
so poor as to be quite dependent on <1da$yakas>1. Most monasteries
own a little land, and a few old ones own a great deal. I had the
impression, however, that the extent to which meals are supplied
daily to the monastery depends far less on the monks' need than on
more intangible factors such as the character of the monk and the
degree of organization of the laity. A few really organized monas-
teries had a roster so that each household could see when it would
be their tum to supply the <1da$ne>1. Our temple had such a roster for
four months of the year only, the three months of <1vas>1 (see below,
p. 279) and the following month, during which it is also customary
for monks to remain at home; in this period every Buddhist

278

household in the village had a turn, supplying either the midday
meal (<1daval da$ne$>1), or the evening requirements (<1gilampasa>1) and
breakfast (<1hil da$ne$>1) the next morning. Even this was economically
unnecessary to the temple, which was rich, but it was explained to
me that every family was thus given the chance to earn merit.
This argument was, however, not applied the rest of the year,
when the monks ate food provided from their own resources and
cooked in the monastery kitchen, except when they were invited
out to a <1mataka da$ne$>1 or when there was some big <1pinkama>1. On
these special occasions, e.g. Wesak and the head monk's fiftieth
birthday <1pinkama>1, the food was supplied by the whole village,
nearly every family supplying something; it was carried up to the
temple in a procession (<1peraha$ra>1), the rice arriving in a huge pot
suspended on a pole borne by a man before and a man behind it in
the traditional Sinhalese manner.
  The alms round (<1pin*d*apa$ta>1) is practised nowadays only by
modern fundamentalists, notably the famous Vajira$ra$ma in
Colombo; it is a difficult practice to maintain, because a monk
who goes begging for his food is usually regarded as remarkably
holy and everyone is so eager to give him food that a roster has to
be set up. There is a forest retreat called Salgala where since 1932
some monks of fundamentalist persuasion, including at least one
European, have retired to meditate in the style of the original
forest-dwelling hermits (<1a$ran+n+ava$sin>1); to give a meal to these
monks you have to put down your name more than a year in
advance. The fact that the place where they eat it is called a
<1pin*d*apa$ta s/a$la$va>1 does not make it different from the ordinary
<1da$na s/a$la$va>1 (monastery refectory). On the other hand a monk may
still beg for alms in case of need. Monk 6 lived entirely by begging
because he had no alternative. He was not fully ordained, but a
mere <1sa$man*era>1, who had entered the Order as an adult, and he was
somewhat odd ; actually I never heard anyone say either that he was
crazy or that he was heretical, but there were complaints that he
was not learned, and some held it against him that he had a family,
although originally it was of course quite normal to become a
monk when an adult with a family--as the Buddha did himself.
Be that as it may, he lived in the direst poverty with no perceptible
possessions, and every morning at about 6.30 would set off with
begging bowl and sun-glasses, hoping to collect enough for the
whole day so that he would not have to go out again later. In our
          279

village, to which he occasionally came as it lay less than two miles
from his temple, he was received by only a few households, so far
as I could observe by low-caste people, who probably felt sympathy
with him against the <1goyigama>1 (high-caste) monks of the Siyam
Nika$ya, who held all the temples in the immediate vicinity and
would have nothing to do with him, though he was Siyam Nika$ya
(and therefore high-caste) himself. Although his was a very un-
usual case, I must mention here that the lot of a poor monk may be
most undesirable. In one monastery which I came to I found an
aged monk, quite blind and nearly deaf, living all alone in filth
and neglect; he was kept alive by food brought by a local family
who took pity on him, but seemed to have no other company or
support. Both the children who brought the food and the monk at
the next temple, which was very nearby, said that the old man was
too bad-tempered (<1sa%ra va%d*i>1) for it to be possible to deal with
him; his pupil lived elsewhere and visited him about once a year;
whether in fact other people ever visited him I cannot say, but I
did not get that impreasion.
  <1Sil ma%$n*iyo$>1 likewise depend on charity. Often they live at or
near temples and get food brought much like monks. Two such
ladies who lived apart in a cave under a Bo tree near my village
were differently organized: on full-moon days the senior one went
round the nearby villages collecting goods and cash, on which they
then lived, doing their own shopping and cooking.
  Monks are generally most active as preachers during <1vas>1.
<1Vas>1, which originally means "rain', was the three-month monsoon
period in north-east India when the Buddha told monks to desist
from their wanderings and stay in one sheltered place. Though in
Ceylon this period is not climatically so distinctive, monks still
have to observe <1vas>1 and stay in one place for the three months of
<1vas>1 and the next month.<s15>s "Stay' means minimally that they should
not be away for more than six consecutive nights. In fact <1vas>1
is sometimes broken; and when a monk went on a pilgrimage to
India during <1vas>1 no one disapproved. Usually they stay in their
own monastery, but sometimes in another, or a layman provides a
special lodging. In the last case the monk is led in procession to his
280

retreat. A shopkeeper in the village next to Mi$gala who had built
a new house invited a monk who was from another village but
locally respected to spend <1vas>1 there, thus earning much merit and
giving his house an auspicious start. Under such circumstances a
monk is especially likely to preach often; at the least he constantly
has to preach <1anumo$dan>1, as described above. <1Vas>1 is also the only
period during which monks of the Siyam Nika$ya in my area meet
on full-moon and <1ma$sa poya>1 days for the <1uposatha>1 ceremony. It
is a basic <1vinaya>1 rule that monks who live in the same <1si$ma$>1 (see
Introduction, p. 28) should meet every fortnight and confess their
transgressions when the oldest monk present recites the <1pa$ti>1-
<1mokkha>1, the canonical list of 227 <1vinaya>1 regulations. In practice
the monks who so unite are the inhabitants of an informal group
of up to about ten monasteries which acknowledge the same head
monastery (e.g. Asgiriya);<s16>s and in my experience only monks of
the Amarapura and Ra$man+n+a Nika$yas (see Chapter 8) meet out-
side <1vas>1, this being the one tangible respect in which I found the
Amarapura more orthodox than the Siyam Nika$ya. The Ra$man+n+a
and Amarapura Nika$yas, so far as I could observe, only hold the
<1pa$timokkha>1 ceremony once a month, on full-moon <1poyas>1. More-
over in practice the Siyam Nika$ya monks (I cannot vouch for the
other two Nika$yas) do not rehearse the whole <1pa$timokkha>1 but
only the first section, which deals with <1pa$ra$jika>1 offences, for
which the penalty is defrocking. This confirms the ritual's purely
formal character: no confessions ever take place. Monks who are
notoriously guilty of a <1pa$ra$jika>1 offence (typically sexual inter-
course) may be refused recognition by other monks and by the
laity, but this never seems to happen as the result of a confession.
Sometimes such monks continue to wear yellow robes and even to
live in their monasteries, though it is legally possible for rival
claimants to displace them.
  The extent to which meditation (<1bha$vana$va>1) is actually prac-
tised by monks is extremely difficult to ascertain, not because
meditation practices are esoteric, as some western authors seem to
believe, but because the question is so involved with a monk's
prestige, poasibly even with his self-respect, that honesty is hardly
possible. The monks in my area who answered my questions on
          281

meditation fall into three groups of roughly equal size: those who
said, or as good as said, that they did not meditate; those who
claimed that they did some but certainly do very little; and those
who seem to meditate regularly. The statement of the monk who
said that there was only one monk in the whole area (another of my
informants) who meditated, was certainly a gross exaggeration.
His implication that <1granthadhura>1 monks could not be expected to
meditate was, however, accepted by many; though no one denied
that meditation is necessary for <1nirva$n*a>1, several said that the life
of a village monk was too busy, or hinted that they hoped for
better opportunities in the next life. Against this, one monk went
so far as to say that it is no good becoming a monk unless you
meditate--and doctrine is of course on his side; but another, who
was also in the Ra$man+n+a Nika$ya, explained to me as a matter of
course that as a <1granthadhura>1 monk his knowledge of meditation
was purely theoretical.
  The subjects--and no doubt the techniques--of meditation
closely follow the instructions laid down by Buddhaghosa in his
<1Visuddhimagga>1. This is not necessarily a modern feature; it may
go back to the eighteenth-century revival, when Siamese monks
came to teach meditation (<1vipassana$>1).<s17>s However, the tradition
must at least have been reinforced by monastic education.<s18>s
When I asked monks on what subjects they meditated their
answers never went outside the list of forty objects of meditation
282

(<1kammat*t*ha$na>1) and fundamental doctrinal principles listed by
Buddhaghosa. Popular subjects seemed to be the meditation on
kindness (<1maitri$ bha$vana$va>1) and the meditation on the impurities
of the body (<1pil*ikul/asubha bha$vana$va>1).<s19>s One monk said he
meditated every day on <1maitri$>1,<s20>s and quoted to me the <1Karan*i$ya>1-
<1metta Sutta>1. The <1pil*ikul bha$vana$va>1 consists in recalling the list of
the thirty-two constituents of the body<s21>s and then applying these
thoughts to oneself and others. This is traditionally the main
subject for meditation by the elderly <1upa$sakas>1 who take the Eight
Precepts on <1poya>1 days; a very old monk told me he meditated on it
"to get used to <1samsa$ra' (sasarat*a purudu vent*a>1). Similarly medita-
tion on the Buddha (<1Buddha$nusmr*ti>1) centres on the "<1Iti pi so>1'
<1ga$tha$>1 mentioned in Chapter 4. The only other text which was
mentioned to me (by two monks) as a subject of meditation was the
<1Satipat*t*ha$na Sutta>1,<s22>s which contains a wide-ranging analysis of
the individual; the popularity of this text is also traditional.<s23>s
   I was told in Mi$gala that the previous incumbent of the temple
and his chief pupil, the present incumbent, used to enter the
<1viha$rage$>1 on <1poya>1 days at some point in the afternoon, sit down
cross-legged on their leather mats near the main Buddha image,
and meditate--on what is not known. No doubt this presented an
edifying spectacle to the old people taking the Eight Precepts that
day. But most of the meditation in Mi$gala is undoubtedly done by
the <1dasa sil ma%$n*iyo$>1; they usually meditate for a couple of hours in
the afternoons, mainly the <1pil*ikul bha$vana$va>1. They meditate by
telling beads and muttering <1ga$tha$>1 (Pali verses and formulae), which
is not the manner of meditation prescribed or practised by monks,
so it is perhaps no wonder that the lady I talked to seemed rather
pessimistic about their meditation, declaring that even to be
reborn in heaven was an impossibly high goal and that they only
          283

aspired to be reborn as human beings so that they could go on
meditating. On the other hand she said that it is written in the
scriptures that if you do the <1Buddha$nusmr*ti>1 with perfect concentra-
tion for two or three hours you can see <1nirva$n*a>1; but such perfect
concentration was beyond their powers. Her selection of the most
devotional topic of meditation as the key to salvation is note-
worthy.
  According to Buddhaghosa and to some of my informants there
are two kinds of meditation, <1samatha>1 and <1vipassana$>1, tranquillity
and intuition.<s24>s (In the texts <1samatha>1 is often called or equated
with <1sama$dhi>1 and <1vipassana$>1 with <1pan+n+a$). Samatha>1 is similar to--
and largely borrowed from--non-Buddhist meditation techniques
of ancient India, while <1vipassana$>1 is a distinctively Buddhist
contribution. Buddhist interpretations of meditation have varied
greatly, but the Therava$din interpretation at least since Buddha-
ghosa unequivocally equates practisers of <1samatha>1 with what
Eliade calls "partisans of <1yogic experience>1' and practisers of
<1vipassana$>1 with his "partisans of <1knowledge>1',<s25>s and asserts the
superiority of the latter. The objects of <1samatha>1 are the forty
<1kammat*t*ha$na>1; the principal objects of <1vipassana$>1 are the three
cardinal Buddhist principals of <1anicca, dukkha, amatta>1 (imperman-
ence, sorrow, and non-self). It was interesting to find that these
topics were those named as the objects of their concentration by
the monk who had a reputation for meditation and by monk 6,
who was a religious enthusiast, albeit heterodox and eccentric.
As neither of these monks had ever attended a <1piriven*a>1 we can
certainly speak here of a living tradition, though whether it has
been resuscitated only recently it is impossible to judge.
   There are in Ceylon today forest-dwelling monks who devote
their lives to meditation; one group of these, that at Salgala just
referred to, numbered ten in 1969; a few others dwell alone or
with their pupils, in greater or lesser isolation, ranging from those
who live near big cities and thus receive frequent visits to those in
truly remote parts of the island such as the south-east coast.
Some of these monks are Europeans, and their entire behaviour
is intended as a revival (though to some extent it may be a
284

pseudo-revival) of ancient practices which have long lapsed. They
are part of the phenomenon which Heinz Bechert calls Buddhist
modernism, to which I have alluded in the latter part of Chapter 1.
  In some cases western influence is extremely indirect. A monk
meditating with his five pupils on the eastern coastal plains, inter-
viewed in 1969, told me he knew no Europeans and used no
European publications; he had studied for four and a half years in
Rangoon and then lived for three years and eight months in the
jungle at the foot of Adam's Peak, not even going begging but
living on roots and fruits. He is only accessible to visitors for two
hours a day (as against four hours at Salgala), and he and his
pupils certainly spend most of their lives practising meditation.
As at Salgala, they use a skeleton for meditation on the body; the
chief monk also has in his cell an earth <1kasina>1, a disc used for
concentration, described at length in <1Visuddhimagga>1, IV. Like the
Salgala monks he said he made much use of the <1Visuddhimagga>1,
but unlike them practised the meditation on breathing (<1a$n*a$pa$na>1-
<1sati>1) according to the Burmese interpretation of the <1Maha$>1
<1Satipat*t*ha$na Sutta>1, which method, he said, is not fully expounded
in any publication, but must be learnt from a teacher.<s26>s In
saying that meditation cannot be studied merely by following
written instructions this monk sided with ancient tradition against
the modernists including the Salgala monks.
   The comparative rarity of meditation is closely connected with
the widespread belief in the decline of Buddhism, a topic which I
usually included in my interviews. Gotama Buddha is said to have
predicted<s27>s that his <1s/a$sane$>1 would last for 5,000 years (5,500
according to two informants; this belief goes back to the com-
mentaries on the Canon.<s28>s (The figure 5,000 lends a certain
         285

poignancy to the celebration held in 1956 to mark the 2,500th
anniversary of Buddhism; we are half-way through, and a monk
remarked that the assassination of S. W. R. D. Bandaranaike in
1959 signalized the beginning of the decline.) Only during a
<1s/a$sana>1 are normal men able to gain Enlightenment; between
<1s/a$sana>1 the only enlightened beings are the <1pratyekabuddhas>1
(see Chapter 3, p. 82), who have a store of merit so immense
that they can realize <1nirva$n*a>1 under such adverse circumstances,
and are accordingly unlikely to be sojourning among us now. The
theory that a man attains Enlightement purely by his own effort
is preserved, but this is impossible without a knowledge of the
religious truths which a normal man is incapable of ideating for
himself, so that in practice one who has not heard a Buddha's
teaching has no chance. A village girl said that in a Buddha-less
period one must keep trying (<1utsa$ha karant*a>1), but only limited
progress is possible. A monk said that the Four Noble Truths
always exist, i.e. are true' but between <1s/a$sama>1 they exist undis-
covered, like electricity before man discovered its uses. It is
further believed by the majority of monks, at least of those
whose general attitudes can be characterized as traditional,
that this <1s/a$sane$>1 has already declined so far that it is no longer
possible for men to attain <1nirva$na>1. This opinion is very prevalent
among the laity and, interestingly, was even strongly put to me by
a <1sil ma%$n*iyo$>1. One monk even specified that till Maitri$ came it was
not even possible to become <1sovan>1 (Pali: <1sota$panna>1), a religious
state in which one will attain <1nirva$n*a>1 after seven more lives. The
last <1arhat>1 is commonly said to have been Maliyadeva, though most
people are very vague about when he lived. (In fact he is men-
tioned in <1Maha$vamsa>1, XXII, 30 and 49, as an <1arhat>1 in the time
286

of Dut*uga%mun*u; there are several stories about him in com-
mentaries<s29>s.)
  Others say that there <1may>1 still be human <1arhats>1, but it is
unlikely and/or undiscoverable. (One monk pointed out that an
enlightened person is not supposed to boast about it.) One monk
compared the <1s/a$sana>1 to a worn-out organism: very few can attain
<1nirva$n*a>1 now, just as a tree grows barren when its fruit is picked
too often, and the seventh child is weaker than the first. Some said
there might be enlightened beings still alive in one of the five "pure
abodes' (<1suddha$va$sa>1) at the top of the <1brahmaloka>1, the other part
of the universe besides mankind in which Enlightenment is
possible. The average view, perhaps, was that of the monk who
said that it is not impossible to attain <1nirva$n*a>1 now, but "religious
practice is weak; it is hard to believe that there is anyone alive
who has become an <1arhat' ("pratipatti durvalayi, arhat vecca aya>1
<1innava$ kiyala$ hitant*a ama$ruyi>1'),
   On the other hand a monk who knew the Canon unusually
well told me that it was a grave mistake to state that no <1arhats>1
now exist, for the Buddha's relics still exist, and so do <1parya$pti>1,
<1pratipatti>1, and <1prativedana$>1 [see pp. 270-1]; there must be some
<1arhats>1 left on earth, maybe in Burma.<s30>s The Buddha once told a
brahmin that his teaching (<1s/a$sane$>1) would last so long as monks,
nuns, laymen, and laywomen practised mindfulness (<1satipat*t*hana>1),
and of these four groups only the nuns have disappeared. Another
monk, also well-informed, put the matter precisely: books tell us
that it is not impossible for anyone to realize <1nirva$n*a>1 now, but
we know of no cases. The <1s/a$sana>1 declines because it is subject
to the universal law that there is nothing eternal "(sada$ka$lika
deyak na%$'). A third monk utterly rejected the suggestion that there
is a decline (<1pirihi$ma>1) in the <1s/a$sa$me>1, but he seemed altogether to
be on the defensive against the foreign intruder.
   Over what will constitute the disappearance of the <1s/a$sa$ne>1 there
was some disagreement. One monk said that the Sangha will die
          287

out, and that will be the end of Buddhism. A second, however,
included the relics and laity as physical indices of Buddhism's
existence. The formulation of a third was the widest: after 5,000
years will come the disappearance (<1antardha$ne$>1) of five things: relics
(<1dhatu$>1), realization (<1prativedana$>1) (which he glossed as "the fruit
of Enlightenment' (<1rahat phala>1)), monks and novices (<1linga>1)
(which he glossed as "monachism' (<1mahan*akama>1)), practice
(<1pratipattiya>1) (which he glossed as "morality' (<1sil>1)), and religious
learning (<1pariya$ptiya>1) (which he glossed as "<1Dharma>1'). This keeps
very close to the commentatorial text,<s31>s which is also the passage
on which is based the doctrine of the paramount importance of
learning. (See below, Chapter 9, p. 320.) The text says that prac-
tice and realization depend on the continued existence of learn-
ing.<s32>s The canonical texts will finally disappear in reverse order:
fist the <1Abhidhamma Pit*aka>1, beginning with the last book;
then the four <1Nika$yas>1 of the <1Sutta Pit*aka>1, beginning with the
<1Anguttara>1 and ending with the <1Di$gha Nika$ya>1; then the <1Ja$takas>1;
finally the <1Vinaya Pit*aka>1. Last to go is the <1Uposatha-kan*d*a>1.
When a king offers a purse of gold to anyone who can recite a four-
line verse of the Buddha's teaching, and there are no takers,
learning will have disappeared.
   For another version of what happens next, let a Mi$gala monk
take up the story.
  The <1sa$sane$>1 of a Buddha lasts for 5,000 years, which period is
called a <1Buddhotpa$daka$laya>1 (time producing a Buddha), and is
followed by an <1a-Buddhotpa$daka$laya>1 (the opposite). During all
this time men get wickeder and wickeder, and their life span gets
shorter, till it is only ten years. In these ten years people become
adults but remain like children.  Then comes the <1lokavina$sa>1
<1alias sattuvina$sa>1 "(destruction of the world' or "of living beings').
At that time a king comes down from the <1Ca$tummaha$ra$jika>1
<1divyaloka>1 ("Heaven of the Four Great Kings'), which is one of the
<1ka$ma$vacara>1 ("sensual sphere', i.e. lower) heavens. He beats a <1bera>1
288

(drum) and announces that after seven days will start a heavy rain,
the <1Murugasam*varsa$va>1 ("wild beast deluge'). "Don't get wet,'
he warns, but most men are so wicked that they don't believe
him; only those with <1pin>1 believe, and they prepare provisions and
shelter in caves, etc. Then a light rain (<1siriva%ssa>1) begins to fall.
All who get wet begin to look like game animals, so people start
shooting each other for meat. In this way all the wicked kill each
other till nothing is left but a great pool of blood. When the rain
finally stops those who had some merit left in them emerge from
their shelters, and on seeing the slaughter they feel compassion
(<1anukampa$va>1). This good thought (<1hon+da adahasa>1) produces a
revulsion from killing, and they thus relinquish the first of the
<1pas pav>1 ["five sins', the subjects of the Five Precepts]. At this the
maximum life-span (<1parama$yusa>1) goes up from ten to twenty.
Men then relinquish each of the <1pas pav>1 in turn, and the <1parama$>1-
<1yusa>1 rises to thirty, forty, fifty, sixty. They then get rid of the
<1dasa akusal>1 ["ten bad actions" another list of sins: these lists are
not mutually exclusive--the story is not strictly logical] and other
bad things till in due course the maximum life-span rises to
infinity [<1sic>1]. But this is not a suitable time for a Buddha to be
born, as men witness no death or decay and so see no reason to do
<1pin>1. It is when the <1parama$yusa>1 goes down again to 80,000 years
that the time is ripe. Then the Bodhisattva who is [just about] to
become a Buddha (<1buduvema bodhisattvayo$>1) is always born in the
<1Tusi$pura divyaloka>1 [another of the lower heavens, called in
Sanskrit the Tus*ita heaven]. Then the gods invite him to be born
on earth saying

          <1Ka$llo 'yam te Maha$vi$ra uppajja ma$tu kucchiyam*>1
          <1Sadevakam ta$rayanto bujjhassu amatam padam>1.<s33>s

           This is the time for you, great hero, to be rebom
           in your mother's womb, and saving the whole
           world<s34>s to realize the deathless state.

Thereupon the Bodhisattva looks out for a suitable occasion.
                                                               289

           <1Ka$lam desam* ca di$pam* ca kulam* ma$taram eva ca>1
           <1Ete pan+ca viloketva$ uppajjanti Tatha$gatha$>1.<s35>s

           Time, place, continent' caste<s36>s and mother--
           after examining these five are Tatha$gatas (i.e.
           Buddhas) born.

  Gotama Buddha was born when the maximum life-span was
120, and lived to the age of 80. The maximum life-span has now
declined to 100 years.
  The last part of this myth, which I recount exactly as I heard it,
is the episode which in Chapter 3 I called "The Gods' Invitation'.
There seems to be a slight discrepancy in the myth as the monk
recounted it, in that he says that the appropriate time for a Buddha
to be born is when a life span is 80,000 years, but Gotama was
born when the life-span was 120. This can be straightened out
by comparison with another account: it is Maitri$, the next Buddha,
who will appear when the maximum life-span is 84,000 years.
(This latter account gives the maximum life-span at present as
120.) Incidentally, that the Buddha did not live his full natural
term agrees with the canonical account in the <1Maha$parinibba$na>1
<1Sutta>1<s37>s of his voluntary renunciation of life.
  This story of the decline in world morals leading to general
destruction is found in the <1Cakkavattisi$hana$da Sutta>1 of the
<1Di$gha Nika$ya>1,<s38>s the text which also contains the prophecy of the
future Buddha Metteyya (Skt. and Sinh.: Maitri$). It is generally
believed that Maitri$ will be the Buddha in the next world cycle,
and people express the hope to be reborn (as human beings)
during the period of his teaching. Chances of this are compara-
tively good, as after attaining enlightenment he will live and
preach for 80,000 years. The monk said that his two chief dis-
ciples (corresponding to Sa%riyut and Mugalan for Gotama Buddha)
will be Dut*uga%mun*u and his brother Saddha$tissa, and his parents
will be Dut*uga%mun*u's parents, Ka$vantissa and Viha$ra Maha$
De$vi$. This prediction is found in the <1Maha$vamsa>1 (XXXII,
290

81-2). The belief that Dut*uga%mun*u will be Maitri$'s right-hand
disciple is general, but some say that his left-hand disciple will be
Deva$nampiya Tiasa, the king under whom Buddhism came to
Ceylon.
  The general opinion is that Maitri$ will be born after an ex-
tremely long time (3,000 years? (a village girl); 10,000 years?
(a monk); ten million years? (another monk)), but we had better
start now our preparations for that event, so that we shall then be
in a fit state to receive and profit by his teaching. Mainly we must
be sure to be born as humans or higher at that time. (Answers at
this point reflect the general disagreement over whether gods can
attain <1nirva$n*a>1.) Some say vaguely that "people who have done
merit' (<1pin karapu aya>1) will hear him; others lay more stress on
the difficulties. The village girl says that if one meditates now it
will come easier in the time of Maitri$, but one also needs faith
(<1sardha$va>1). One who in this cycle has committed one of the five
great sins<s39>s will not see Maitri$. To accumulate enough merit not
to be born in a hell (<1apa$ya>1) under Maitri$ one must keep the moral
precepts, do meditation, etc., and one must work out one's bad
<1karma (pav ahosi karant*a>1) by rebirth in unpleasant existences.
However, if one misses Maitri$ there will still be an infinite number
of Buddhas after him under whom one can obtain release, though
it is true (but perhaps not widely known) that after Maitri$ there
will be a particularly long gap. Monk 6 was the only person whom
I ever heard speak of Maitri$ with any fervour; generally I got the
impression that monks regarded him with no more than academic
interest, politely assumed for the occasion. Nor was he ever
spontaneously mentioned to me by a layman.
  To close this chapter I shall reproduce the eschatological opinions
of monk 6, to whose eccentric doctrines and unusual poverty I
have already alluded. The heterodox tenor of these opinions, with
their strongly theistic flavour, makes it difficult to subsume them in
the general scheme of this work, but even if the deviance that they
represent should be no more than a quirk of personal psychology,
the fact that I encountered them locally virtually obliges me to
record them, and their vivid fervour makes me not sorry to do so.
  The present state of religion is very low, he said. No one can
now attain <1nivan>1 until Maitri$ comes. No laymen now observe the
           291

Five Precepts, while monks multiply pictures and clocks in their
monasteries till they are smarter than hotels. The Buddha in
Jetavana$ra$ma lived in utter simplicity.
   Only that man can be called a Buddhist who prays to reach
<1nirva$n*a>1 [in the time of Maitri$] <1(nivan labant*a pra$rthana$ kalot>1
<1Buddha$gamka$rayek, na%tmam na%$>1). Religion is the use of truth
(<1a$gama a%tta pa$viccikiri$mayi>1). [What he meant by this is obscure.]
<1Nirva$n*a>1 is purity of thought (<1hita pirisudu kiri$ma>1). A man who has
attained <1nirva$n*a>1 is no longer liable to birth, decay, sickness, and
death; moreover his body has no weight, so that he can fly through
the air, and nothing will harm him: even if he treads on a snake,
the snake will not bite him. If he takes the Three Refuges and
observes the Five Precepts it is enough, for only sin weighs him
down and makes him subject to mortal changes. But he cannot
attain <1nivan>1 without the help of Maitri$; and <1pina>1 by itself is use-
less, for despite <1pinkam>1 one may be reborn in some low state like a
louse, and then what use is one's merit? So <1pima>1 is <1ahosi, nikam>1
(useless), unless accompanied by <1pra$rthana$>1. Even the merit
acquired by hoping that a crow will drink the water you spit out
after rinsing your mouth is sufficient basis for a <1pra$rthana$>1.
We attain <1nirva$n*a>1 by our <1pra$rthana$>1 and earnest wish to reach it;
as this wish grows more intense we go through the four stages on
the path, till in the final stage we reach <1nivan>1, the supramundane
good, which no termite or mouse, no accident can destroy (<1lokot->1
<1tara kusalat*a kisi ve$yyek, mi$yek kanne na%$, ana$turak na%$>1).
  Concerning Maitri$ I was given many more details than I can
reproduce; I believe they all derive ultimately from a post-
canonical Pali text, the <1Ana$gata-vam*sa>1. He will be eighty cubits
high, and live in a capital city, Ketumati$, a hundred times better
than a city in one of the heavens nowadays. Its inhabitants,
Maitri$'s congregation, will be deities (<1divyara$jayo$>1 and <1divya$nga->1
<1na$vo$>1) who did their meditation here on earth and so no longer
need to. All their physical wants will be automatically supplied:
divine trees will clothe them; when they are hungry or thirsty
food and drink will spontaneously appear; and when they want to
go outside a lavatory will be before them.  These remarks--
heretical in their depreciation of meditation--show that the
prayers to be reborn under Maitri$ are aimed at a goal more
succulent than <1nirva$n*a>1.
  Gotama Buddha has not yet completely seen <1nivan>1, in that he
292

has three <1nivan>1. The first two are his original Enlightenment and
his <1parinirva$n*a>1. The third will come at the end of the 5,000 years
for which the <1sa$sana>1 (doctrine) will last. For these 5,000 years the
Buddha is alive in the three worlds (<1ka$maloka, ru$paloka, aru$pa->1
<1loka>1). At the end of this period, all relics and images of the Buddha
will reassemble and join into one. (<1Arhats'>1 relics, etc., will not be
affected.) In Ceylon every relic, every image, every picture of the
Buddha will go to the Ruvanva%lisa%$ya at Anura$dhapura. From
Anura$dhapura all these will go to Rangoon, and at the Rangoon
<1caitya>1 the corporeal Buddha will come to life again. Simul-
taneously the branch of the Bo tree at Anura$dhapura will rejoin
the <1Maha$jaya bo$dhinvahanse$>1, the original Bo tree. Then the Buddha
will return to the foot of the Bo tree, and seated there will preach
his last sermon, for seven hours, and all the gods will come to
listen. At its conclusion the Buddha will finally enter <1nirva$n*a>1,
and all trace of his doctrine will have disappeared from the earth.
  The final paragraph of this account is based on the same Pali
commentaries as the account of the disappearance of the scriptures
quoted above. According to what is probably the later of these two
texts<s40>s there are three stages in the disappearances of the doctrine:
its existence (<1t*hiti>1) (during which it continually declines), its
withdrawal (<1osakkana>1), and finally its disappearance. When the
Sangha has died out the stage of "withdrawal' has been reached.
At this point the text, like monk 6, says that there are three
<1nirva$n*as>1 of the Buddha (<1parimibba$na$ni>1): his Enlightenment (<1kilesa>1-
<1parinibba$ra>1--"extinction<s41>s of the defilements'), his death (<1khan>1-
<1dhaparinibba$na>1--"extinction of the aggregates [constituents of the
person]') and <1dha$tuparinibba$na>1--"extinction of the relics'. All the
Buddha's relics in Ceylon (here conceived only as the physical
(<1sa$ri$rika>1) relics) will indeed assemble at the Ruvanva%lisa%$ya
(Maha$cetiya), and then stop at Na$gadi$pa, on their way over not to
Burma, but of course to India. At the foot of the original Bo Lee
(<1maha$-bodhipallanke>1) they will join up with the relics from the
world of the <1na$gas>1<s42>s and the various heavens, <1devaloka>1 and
<1brahmaloka>1. There, according to the other text<s43>s, they will re-
constitute the Buddha, complete with his thirty-two major and
                                                               293

eighty minor distinctive features, and he will perform the miracle
of the pairs. (There is no mention of preaching.) No human being
will go there, but deities will assemble from the 10,000 world
systems and lament the approaching darkness. The body will
emit a flash, and disappear. The later text does not say that the
relics will reconstitute the body, only that they will form a pile
like a heap of gold, which will emit rays of the six colours, which
will pervade 10,000 world systems. From all these world systems
the deities (<1devata$>1) will assemble for a last look; then the flash
from the relics goes up as far as the <1brahmaloka>1, they vanish,
and "the doctrine has truly disappeared (<1sa$sanam* antarahitam>1
<1na$ma hoti>1)'.<s44>s
                              8
                      CASTE IN THE MONASTERY
   The structure of Sinhalese society was based on the institu-
      tion of caste. Is caste in accordance with Buddhism?<s1>s
                                             DR. G. C. MENDIS

ALTHOUGH in the Introduction I stated that my main interest
was in religious change, the tenor of this work has rather been to
demonstrate that Sinhalese Buddhism has been surprisingly
conservative. Before summarizing those conclusions in a final
chapter, I wish to discuss the most striking exception, a change
certainly since the time of the Buddha and probably even since the
time of Buddhaghosa, which has so far been our basis for com-
parison on most points. <1Obiter dicta>1--e.g. allusions to sectarianism
and to monastic possessions--will already have suggested to the
reader that I have found more evidence for change in the practices
of the Sangha than in doctrinal belief--that change has been
organizational rather than ideological. Anyone who approaches
modern Buddhism in any Therava$da country after a study of
"primitive Buddhism' through canonical texts or such excellent
secondary sources on the early Sangha as Sukumar Dutt's <1Early>1
<1Buddhist Monachism>1<s2>s is immediately struck by these changes in
monastic life. The most striking ones in Ceylon are: the holding of
often considerable property by monasteries ; the custom of passing
pupillary succession to a relative, who thus effectively inherits the
monastic property; the invasion of the Order by caste, and its
division into parallel groups (Nika$yas) split mainly on caste lines ;
the virtual disappearance of the alms-round (<1pin*dapa$ta>1); the entry
of monks into politics; the employment of monks in galaried
positions. Of these six features the former three can properly be
called organizational, while the latter three more strictly concern
the conduct of monks as individuals. The disappearance of the
                       295

alms-round has already been discussed in Chapter 7; I shall here
discuss caste, which will involve some reference also to the first
two features, but for further data on these questions I must refer
the reader to works listed in my Bibliography. To the participa-
tion of my informants in politics I have devoted a separate article.
I cannot here go on to the question whether such participation is
justifiable, beyond remarking that whenever Ceylon has had
Buddhist rulers monks have held political power. On monks'
salaries I have hardly any data: no monks normally resident in
my area were employed, though I interviewed one who was.
Moreover, in all Ceylon only a small minority of monks hold
salaried positions, while monastic participation in politics, ten
years ago the pastime of a small minority, is still not universal.
Similarly, not all monasteries hold property, and only a very few
have any considerable wealth, while the passing on of such property
to relatives is very far from universal. The pervasion of monastic
organization by caste is, however, complete.<s3>s But this is not my
only reason for choosing to talk about caste rather than any other
feature of monastic change. Nor is my decision primarily deter-
mined by the fact that others of the features listed have recently
been written on.<s4>s My main reason for choosing the caste question
is that it has the greatest ideological significance. Unlike the
customs concerning property and succession it is admitted by
traditionalists to be doctrinally indefensible, and it is of course
heartily attacked by modernists. Some modernists go so far as
to say that the Buddha was against caste altogether: this is not the
case, but is one of the mistakes picked up from western authors.
As their belief is so prevalent I shall give more space than has
been my custom to examining the ideas on caste found in the Pali
Canon, before describing the entry of caste into the Sangha and
the opinions of my informants on the subject.
   The standard work on the Sinhalese caste system is <1Caste in>1
<1Modern Ceylon>1 by Bryce Ryan;<s5>s Chapter 9 (pp. 196-238) of his
book, <1Caste in the Kamdyan Highlands>1, is especially relevant to my
296

experience; indeed the section of this chapter called "Caste in the
Simple High Caste Village' (pp. 204-11) might almost have been
written about Mi$gala. The chapter on caste (pp. 58-95) in Nur
Yalman's book, <1Under the Bo Tree>1,<s6>s and his paper "The Flexi-
bility of Caste Principles in a Kandyan Community' in E. R. Leach
(ed.) <1Aspects of Caste in South India, Ceylon, and North West>1
<1Pakistan>1<s7>s are based on field-work in an area not far from my
own. These are more detailed and technical works; I shall present
here only such facts about Sinhalese caste as are relevant for a
discussion of the relation of caste to Buddhism; my picture,
moreover, is based on my own field experience, and thus differs
in some details from previously published materials.
  The Sinhalese caste system is historically and conceptually
related to the Indian; but there are fewer castes, and there is less
scope for ritual pollution through the violation of caste tabus than
in India. The social distances between most of the castes are very
small, and indeed are so unobtrusive in their operation that casual
visitors to Ceylon are hardly aware of caste, though the effect on
attitudes, which find expression for instance in political activity,
is more far-reaching. By comparison with India there is far less
inequality. This is partly because among the Sinhalese, as they
see it, both the top and the bottom strata of Hindu society are
missing. Of the four classical Hindu classes of brahmins, warriors,
agriculturalists with traders, and menial castes (<1bra$hman*a>1,
<1ksatriya, vais/ya, s/u$dra>1), the Sinhalese lack the first two: their
highest caste, <1goyigama>1 or <1goyikula>1, consists of <1goviyo$>1 or rice
cultivators. This top caste, in striking contrast to the Hindu situa-
tion, is by far the largest: in fact in the area where I worked, as
perhaps in Sinhalese society at large, it had an absolute majority
of the population. The <1goyigama>1 is the only caste found in the hill
                                                               297

country which contains sub-divisions, but in most contexts these
are not important, and all <1goyigama>1 are known among themselves
as "good men' (<1hon+da minissu>1), in distinction to the "lesser castes'
(<1ad*u kula>1). At the other end of the traditional social scale, the
Sinhalese hardly have anyone to correspond to the vast array of
Hindu out-castes (who among the Tamils in Ceylon may be as
numerous as those within caste lines).
   The <1goyigama>1 are sometimes equated with <1vais/yas>1. This has
considerable historical justification.  The Sinhalese words <1goyi>1
and <1govi>1 are derived from the Pali <1gahapati>1, which literally means
"householder'. Socially it means something like "bourgeois'. In
his study<s8>s of the society depicted in the <1Ja$takas>1 Fick showed that
the word <1vais/ya>1 was never used to describe actual people, but the
term <1gahapati>1 denoted those who stood in the social structure
below the nobility and above the artisans, i.e. where we would
have expected to find the <1vais/ya>1. All the Sinhalese castes below
<1goyigama>1 are considered, by those who bother their heads about
such matters, to correspond to Hindu <1s/u$dras>1 (the menial castes),
until we come to the <1rod*i>1 at the very bottom. I have not heard
the <1rod*i>1 explicitly equated with modern Hindu out-castes, but
they are regularly equated with the lowest class in ancient times,
the <1can*da$las>1 who occur in canonical texts and the <1Maha$vamsa>1
as out-castes (i.e. below <1s/u$dras>1), so I think the equation could
properly be taken a step further. The <1rod*i>1 are beyond the pale of
traditional Sinhalese society, the only caste to be "significantly
differentiated culturally';<s9>s they seem not even to have been
allowed into Buddhist temples--a point to which I shall return.
There are interesting myths about their origin, which all concur in
claiming that the <1rod*i>1 have been degraded from high (royal) status
for some heinous offence, e.g. eating human flesh. Traditionally
<1rod*i>1 are itinerant beggars, and they are rarely if ever refused,
because it is held that their curses are singularly potent. By an
extension of this useful belief, <1rod*i>1 are considered experts in black
magic. However, their existence does not much disturb the
comparative homogeneity of Sinhalese society, because they are
extremely few, between 1,500 and 3,000. Most of them have now
settled in <1rod*i>1 villages (<1kuppa$yama>1), and in fact I never saw one in
298

my area, where their place at the bottom of the social ladder was
taken by exceptionally depressed Tamils.
  Before I give a brief description of the operation of caste
distinctions in Mi$gala, I would like to make a brief excursus into
the history of the caste system in Ceylon. In this history different
types of social stratification seem to have existed in different
periods, all of them usually called "caste' by European scholars;
rather than use question-begging terms like "caste' and "class' we
might do better to employ some such neutral term as "status
group'; but for simplicity I shall here keep to conventional
terminology. The <1Maha$vamsa>1 and <1Cu$lavamsa>1 preserve intact the
old Pali terminology according to which society is divided into
the four classes <1bra$hman*a, ks*atriya, vais/ya, s/u$dra>1, and the out-
castes beneath them; but, as in post-vedic India, this division
seems to have been primarily theoretical, and there were in fact
many more than four status groups. There were never many
brahmins in Ceylon, but one village still has the name Bamun*u-
gama, which means "brahmin village', and even a thirteenth-
century text mentions that a certain monk was of brahmin
descent.<s10>s <1Ks*atriyas>1 were the nobles, the royal and potentially
royal families, and they seem to have been divided into clans. The
terms <1vais/ya>1 and <1s/u$dra>1 did not correspond to any clear-cut social
units, even in the ancient period, but various groups were sub-
sumed under each term, and possibly the <1vais/yas>1 were farmers
and the <1s/u$dras>1 artisans, as is theoretically prescribed, though I
know of no clear evidence for this. In mediaeval times (say A.D.
500-1500), though society was still said to consist of the four
classes, this classification seems to have become irrelevant, and
the <1Cu$lavamsa>1 usually divides society into <1kuli$na$>1 ("people of
family') and <1hi$na$>1 ("inferiors'), the former apparently being the
nobility, the latter the common people. The modern caste group-
ings are never mentioned in the chronicles, which give very
scanty information about the sixteenth and seventeenth centuries,
supplying only a bare outline of their political and religious
history; nor are there any other sources to supply us with data on
Kandyan society till the European accounts.
  The Kandyan system,ll which is known to us principally from
                       299

the account of Robert Knox (see p. 26) written in the late seven-
teenth century and, in more detail, from nineteenth-century official
British observations and reports, was a centralized feudalism in
which the king had a basic right to all land, and granted some
villages to nobles, temples and shrines, while other lands were held
directly under the crown by the village cultivators. But in all
cases in return for the land which they cultivated villagers gave
service (<1ra$jaka$riya>1), either to the court, directly or through an
administrative official, or to the nobleman or the religious institu-
tion which functioned as landlord.  In the Kandyan period
brahmins and <1ks*atriyas>1 have completely disappeared from Sinha-
lese society, except in so far as the kings, who latterly were
imported from South India, claimed <1ks*atriya>1 status. Society was
thus divided into two principal classes, <1goyigama>1 farmers and the
service castes, identified by the learned as <1vais/ya>1 and <1s/u$dra>1
respectively; but although <1goyigama>1 people still call themselves
"good' (<1hon#da>1) and the rest "inferior' (<1ad*u>1) this dichotomy does
not correspond directly to the mediaeval one between the <1kuli$na$>1
and the <1hi$na$>1, if only because, as far as we can tell, the <1kuli$na$>1 were
a clear minority of the people, whereas the <1goyigama>1 were and are
an absolute majority. On the other hand there is not the slightest
doubt that in the Kandyan period all castes, including the
<1goyigama>1, were differentiated by occupation: in other words they
were professional groups. I am not asserting that occupation was
in all cases the original factor of demarcation between groups.
Some of the castes were originally ethnic groups, bodies of
immigrants from South India; but such groups adopted particular
occupations, which came to characterize them.  (The same
phenomenon continues today; e.g. all Pathan immigrants seem to
be money-lenders.) How groups came by their professions is not
entirely mysterious, for we know of cases in which the king
ordered certain groups to undertake certain professions. Ralph
Pieris writes:<s12>s
... in Kandyan times it was considered the lawful function of the
king to ordain appropriate functions to various castes: he could also
degrade certain villages or families of high caste to a lower status, and
there are certain degraded <1gattara>1 villages [<1gattara>1 are the so-called
<1goyigama>1 out-castes, the lowest sub-caste of <1goyigama>1] in existence to
this day. The caste system thus acquired a certain flexibility, and
300

certain groups of people performing specialized functions became, in
effect, separate castes--the <1hala$gana>1 people for instance [who became
cinnamon pickers, and are now so known] were weavers imported from
India some seven hundred years ago.<s13>s
I would add that we do not know how many of the Kandyan
caste groups, which indeed survive today, existed as groups before
the Kandyan period. Notice also that the social groups so created
have long survived the disappearance of the Kandyan state. The
caste system of the Kandyan feudal period is gradually being super-
seded by the class system of a more mobile and industrialized
society, in which wealth, education, and power (especially associa-
tion with the government) confer prestige, but this development
has not yet had far-reaching effects on Kandyan village life.
   Though the central government and its modern apparatus
replaced the court in 1815, a few villages are still owned by
nobles, temples, or shrines.  The present government has
announced its intention of finally abolishing the noblemen's
feudal villages (<1nindagam>1), but vested interests on both sides of
the House are in no hurry to pass legialation to this effect and the
institution is likely to survive for some time.  Moreover,
the abolition of <1nindagam>1 need not involve villages owned by
<1viha$ras (viha$ragam>1) or by <1de$va$le$s (de$a$layagam>1). Mi$gala temple, for
instance, owns and will doubtless continue to own most of the
rice fields of a village higher in the hills. Such ownership means
that the temple gets half the produce from a large block of rice
fields; the other half goes to the cultivators, who in return also
perform services for the temple, which nowadays are virtually
confined to ritual drumming. Nowadays if you ask a monk whether
his temple has <1ra$jaka$riyo$>1 (feudal villeins) he will probably under-
stand the term to mean drummers who are obliged to give their
services (rather than hire them out as free agents), whereas for-
merly the term would have covered all service castes--washermen,
smiths, jaggery makers, etc. There are in Ceylon, and have been
at least since the seventeenth century, some two dozen castes, but
only about half of these are found in any number in the hill
country. Most of the inhabitants of Mi$gala (I estimate over 80 per
cent) were <1goyigama>1 (rice farmers), many of them of the <1patti>1
(herdsman) sub-caste and one family of the noble (<1radala>1) sub-
caste, which owned a lot of the village rice fields; the only other
                      301

castes indigenous to the village were grass cutters (<1panna>1),
washermen (<1rada$>1), and a family of drummers (<1berava$>1).<s14>s There
also lived in the village, in descending order peripheral to its
society, Tamils, Muslims, and Low Country Sinhalese; but my
remarks apply to local, that is Kandyan, families. Living as I did
in a village so predominantly <1goyigama>1, with washermen, who are
the most leniently treated of the low castes, the largest minority
group, I probably have a rather rosy picture of inter-caste rela-
tions among the laity. Ryan and Yalman record more discrimina-
tory behaviour. But there are many Kandyan villages of caste
structure similar to my own, so my experiences are unlikely to be
completely atypical.
   There is no visible difference between members of different
castes. In the Kandyan kingdom there were distinctions of dress,
but these were gradually outlawed by the British or eroded by
western contact; for instance, it used to be forbidden for low-
caste women to wear blouses, but this mark of social degradation
now has no legal sanction, and is confined to itinerant <1rod*i>1 beggars,
rarely seen in the area, and low-caste Tamil women from the
estates. There is free social intercourse between all the poorer
Sinhalese members of the village; people are quite likely to drop
in on members of another caste for a chat, although it is unusual to
pass beyond the verandah. (Members of different castes do not
pay on each other those formal calls which Ryan calls "inter-
familly visiting', but such visits altogether hardly occur except on
ritual occasions, i.e. maybe as little as once a year.) The only
houses definitely outside this social round are those of the two
richest families in the village, one of them the <1radala>1 mentioned
above, the equivalent to the lord of the manor, the other an un-
popular parvenu with a white-collar job in Kandy; both these
houses are distinctive, in their very different styles, by their size
and situation. Even among the rest of the villagers it might be
found that social familiarity went more by wealth and influence,
i.e. "class', than by caste, although as most or all members of one
caste in a given village are also relations, who naturally see more of
each other, this is hard to asseas. Within the framework of this
302

ease in casual intercourse certain distinctions are unvarying and
taken for granted. A lower-caste person must sit on a lower seat,
someone lower still must stand: the drummer family, being the
lowest caste in my village, stood in the presence of anyone else.
The drummer had become comparatively wealthy by growing
vegetables on temple lands leased to him and he rented out a
house to a schoolmaster, who was <1goyigama>1; but the school-
master, a man who had spent a year in the West, told me he
could not bring himself to offer his landlord a seat when he came
to see him and so he avoided having him in the house at all. Yet
he teaches pupils of all castes seated indiscriminately, and I never
found children who sit together at school unwilling to do so else-
where. A high-caste person will not accept food or drink from a
low-caste person, though he will accept a chew of betel, which
keeps things sociable: he may offer a lower-caste visitor refresh-
ment, but in a vessel which he does not use himself, e.g. a chipped
cup; most often tea, the standard drink, is offered in a coco-nut
shell which is then thrown away. To what extent these distinctions
are observed among the low-caste (non-goyigama) I cannot say;
I am sure that they used to be, but nowadays low-caste people are
becoming an articulate and often solid minority, increasingly
opposed to caste discrimination, and their opposition may gradu-
ally extend to discrimination from their inferiors as well as from
their superiors. Marriage outside the caste is still very rare; there
was one case in Mi$gala, a <1goyigama>1 woman married a low-caste
man, but he was from Colombo and worked far away, so that he
was rarely seen in the village. He was also a communist trade union
oflicial--a typical syndrome. Predictably this girl was generally-
regarded as a bad lot, but to us, perhaps because we were wester-
ners known to care little about caste, her badness was said to
consist in her communism, which was purely marital, and her
private life, which was subject to various unsavoury rumours, not
all of which, I suspect, were untrue. However, this woman was
not boycotted by everyone in the village, nor as far as I know was
she considered to have lost caste, a leniency in striking contrast to
custom under the kings of Kandy, where in theory at least mar-
riage to a man of lower caste was punished by death. Her children
would be ascribed to their father's caste, but will probably move
to the anonymity of Colombo, or at least to a government post far
away where their antecedents are not known.
                       303

   Despite the paucity of easily identifiable forms of discrimina-
tion, it lnust not be thought that villagers are for a moment un-
aware of caste distinctions. My evidence is linguistic. There are
in Sinhalese several personal pronouns of address and reference,
graded by status, as well as various forms of address: for instance,
an equal or superior is not normally addressed by name. A per-
son's attitude to caste is thus discernible in the pronouns he uses.
My experience here is alas rather limited, especially as I myself
always used polite fox for low-caste people, so that my inter-
locutors might politely adopt my style; however, I think that
<1goyigama>1 adults rarely address, and among themselves never refer
to, low-caste members of their village as equals. There are of
course degrees of inferiority, and I have heard the lowest pronoun,
"u$', which is primarily used for animals, used to refer to a low-
caste person, though it is very rude for anyone, however socially
exalted, to use this form to another person's face. The only time
I heard this done was by a young monk, avowedly a communist,
to an old service drummer with whom he was annoyed, and it had
a most unpleasant ring. This brings us back to Buddhism, and it
is time for me to consider the Buddha's own attitude towards
caste, as manifested in the Pali Canon.
  In the Buddha's environment (north-east India round 500 B.C.)
the caste system was by no means so rigid and elaborate as it has
become in India, or even in Ceylon, and so we must not expect to
find in the scriptures pronouncements dealing with precisely the
situation obtaining today. The theory of the four great classes of
<1bra$hman*a, ks*atriya, vais/ya>1, and <1s/u$dra>1, to which I referred earlier' as
well as the out-castes below these, already existed. There were also
a few domestic alaves (as was also the case in the Sinhalese kingdom)
who could in theory come from any caste though I suspect that
they usually came from low castes; what their caste position in ser-
vitude was is not quite clear, but from a reference in a slightly later
Sanskrit text, the <1Arthas/a$stra>1, which says that high-caste slaves
could not be compelled to do defiling duties, it seems that like their
later Sinhalese equivalents they were not necessarily degraded. The
institution of slavery, a peripheral phenomenon in Indian civiliza-
tion, should be treated with caste because the children of slaves
were also slaves (unless specifically manumitted), so that their
status was wholly ascribed, not achieved. There was neither
304

inter-marriage (though hypergamy seems to have been allowed
after the first wife) nor commensality between all members of one
class, and <1a fortiori>1 not between members of different classes, so
the modern system was beginning to appear, but we lack details.
Hindu law-books, of a slightly later date but pre- dating the modern
system, show that ideally people followed the profession of their
fathers, but there was provision for a different way of life in mis-
fortune (<1a%pad dharma>1), by which for instance a brahmin could
engage in trade without losing caste. In north-east India, the are a
where Buddhism arose, the major question of the day seems to
have been whether <1bra$hman*a>1 or <1ks*atriya>1, priest or warrior, spiritual
or secular power, was the higher caste. Further to the west the
brahmins had already established their supremacy. The Buddha,
however, came from a tribe in the Himalayan foothills, the S/a$kyas,
who claimed <1ks*atriya>1 status. There is a Buddhist tradition that
Buddhas are born into either the <1bra$hman*a>1 or the <1ks*atriya>1 cage,
whichever is higher at the time. Most of the Buddha's sermons
dealing with caste are concerned with the problem, "Who is the
true brahmin?  Some western interpreters have jumped too
quickly to the conclusion that the Buddha ethicized the concept
of caste and argued against all such social distinctions. He did
ethicize the concept, but he did not argue against the continuing
use of the social distinction within its own context; moreover, this
ethicization is only half the story. In one text<s15>s a caste-proud young
brahmin is insolent to the Buddha and disparages the S/a$kyas. The
first half of the Buddha's response shows the brahmin that
<1ks*atriyas>1 are superior to brahmins, for they are more rigorous and
exclusive; for instance, if we are to believe the text, a <1ks*atriya>1
outlawed by his own caste would be ritually acceptable to the
<1bra$hman*as>1, whereas the converse would not hold. The second half
then proceeds to expound the often-repeated doctrine that holi-
ness is not inherited but acquired, that the true brahmin is not a
born <1bra$hman*a>1, but someone who has achieved enlightenment.
Obviously the Buddha preferred to argue from strength; had he
merely told the brahmin that birth was irrelevant he would have
been open to the charge of "sour grapes'. Similarly, in an amusing
mythical account<s16>s of the origin of the social order, the Buddha
heartens two young brahmin converts, who are being reviled by
brahmins for associating with him and the other monks, by telling
                                                               305

them that the <1ks*atriya>1 order is top because it came first, from men
selected to govern the rest; at the same time he is at pains to point
out that they were chosen from among equals as <1primi inter pares>1.
But in talking of laymen the Buddha does not go much further
than saying:s17 "Not by birth is one a brahmin, not by birth an
out-caste; by deeds is one a brahmin, by deeds an out-caste.'
Inter-caste behaviour is not mentioned in discourses on lay ethics,
ad there are certainly no exhortations to social equality. Indeed,
caste is accepted as a fact of life on a par with other forms of
status; to be reborn in a low and unimportant family is the result
of demerit in this life, typically of failing to respect those of
superior status.<s19>s
  Questions of inter-caste ethics among laymen seem to have
been confined in primitive Buddhism to rivalry between <1ks*atriyas>1
and brahmins for the top position. In all Buddhist texts, contrary
to practice elsewhere, <1ks*atriyas>1 are mentioned first when the four
classes are listed. Presumably the Buddha's repudiation of ritual,
by denigrating the brahmins' religious importance, involved also
a denigration of their social importance. Buddhist ethics show a
general tendency to esteem wealth and power, rather than birth,
as criteria of worldly success, and I see the raising of the <1ks*atriya>1
above the brahmin as fitting into this pattern. However, those who
lead the holy life are supposed to pass beyond such rivalries and
distinctions. The Buddha draws the line: "In the supreme per-
fection of wisdom and righteousness there is no mention of notions
of birth, of lineage, or of the snobbery which says "You are as
good as I" or "You are not as good as I". It is when there is
marriage or giving in marriage that there is mention of such
notions. For whoever are bound by notions of birth, of lineage, of
snobbery, or of marriage ties are far from the supreme perfection
of wisdom and righteousness. Only having got rid of the bondage
of these notions may one realize this supreme perfection.'<s19>s
  Monks and nuns could come from any caste. This was appar-
ently not a Buddhist innovation, but traditional among those who
"went forth to the homeless life' and became hermits or wandering
ascetics. In the aetiological myth just referred to the Buddha says



306

that members of all four orders became the first ascetics. In the
<1Ja$takas>1 we hear of a potter<s20>s and a <1can*d*a$la>1 (out-caste)<s21>s who
became non-Buddhist ascetics.
  That all castes were admitted to the Sangha is streased both by
the Buddhists and by their brahmin opponents, who call them
menial, dark-skinned <1s/u$dras>1, or even address the Buddha himself
as an out-caste,<s22>s which from their ritual standpoint he was,
because of his contacts with out-castes. We know of several monks
and nuns that they were low-born, and Upa$li, the monk who was
the greatest authority on the monastic rules (<1vinaya>1) had been a
barber, one of the lowest occupations. As for their relations with
the laity, monks were told to accept food from anyone<s23>s (the
doctrine is ascribed to Kassapa, a former Buddha, but I see no
significance in this) so they did not recognize ritual impurity in
laymen any more than they did among themselves. The Buddha's
last meal was a dinner of pork given him by a smith.
  In no other respect has Buddhism in Ceylon come so far from
its original state as in the invasion of the Order by caste, yet
another testimony to the remarkable strength of that insidious
institution. By most monks the contradiction is realized and
admitted; it is not a subject on which they are happy to talk. This
shame shows that doctrinally they have not changed, though I
shall recount one remarkable piece of sophistry, a reinterpretation
of a scripture just quoted. Behaviour, however, is now completely
at variance with the original doctrine in all the points I have set
out concerning the internal conduct of the Order, as well as in the
conduct of many monks towards the laity.
  There are two pieces of evidence to suggest that caste dis-
crimination within the Sangha is ancient in Ceylon. The introduc-
tion to a <1Ja$taka>1 story told by the Buddha says that the occasion
for the story was a dispute when monks who had been <1ks*atriyas>1
and brahmins claimed preferential treatment in the monastery.
This introduction is commentatorial, and in its present form was
written in Ceylon about the fifth century A.D., but the commen-
tator surely did not invent it himself. At the same time I doubt
whether the tradition goes back more than the 700 years which
                    307

would take its origin back to India rather than Ceylon. The most
likely conjecture is that this episode reflects conditions very early
in the history of Ceylonese Buddhism. The other piece of evidence
has more radical implications but is leas certain. Two of the
monasteries founded during Mahinda's mission were called in
Pali Iasarasaman*aka and Veasagiri, which names mean respectively
"That which has ascetics who are lords' and "The hill of the
<1vais/yas>1'. The <1Maha$vamsa>1 (XX:, 14-15) says that the former
monastery was entered by men from noble families, the latter by
<1vais/yas>1. This would show caste differentiation on what we shall
see is the modern pattern. I say that the evidence is uncertain
because the <1Maha$vam*sa>1 was composed over 600 years after the
events described, albeit on the basis of earlier materials, and its
author may be conjecturing the origin of place names. If, however,
he is right, this would suggest to me not merely that monks in
Ceylon were always divided on caste lines, but that this was
probably so in India at that time, some 250 years after the Buddha's
death; for I cannot believe that Mahinda would have sanctioned
so radical an innovation.
   However, there is absolutely no clear evidence for subsequent
caste discrimination in the Sangha during the Anura$dhapura and
Pol*onnaruva periods. The origin of the restriction of entry to the
Siyam Nika$ya to the <1goyigama>1 caste is the subject of much learned
controversy.<s24>s Though the Siyam Nika$ya was founded in 1753,
the custom has far older roots. The first reference to the problem
occurs in the Dam#badeni Katika$vata,<s25>s dated 1266, which says
that at the <1upasampada$>1 (higher ordination) ceremony the ordinand
is to be asked his caste.<s26>s
  The Portuguese historian Queyroz, writing in the 1670s,<s27>s says
that only men of noble birth (<1appuha$mis>1) became monks.<s28>s Knox
in the same period reports likewise: the only people admitted into
the Sangha are "persons of the most noble birth, and that have
308

learning and be well bred'.<s29>s Both these passages suggest that
only <1goyigama>1 men became monks. However, at the great <1upasam>1-
<1pada$>1 ceremony in Kandy in 1753 ordination was not in fact
restricted to members of the <1goyigama>1 caste: among the ordinands
were a member of the <1dura$va>1 (toddy-tappers) caste, Sit*ina$maluve$
Dhammajoti, who became incumbent of the famous ancient
monastery of Mulkirigala in the Southern Province, and Ve$ha%lle$
Dhammadinna, who by caste was a drummer (<1berava$>1). These two
were both pupils of Va%livit*a Saran*am*kara, and most leading monks
in the Low Country Siyam Nika$ya today trace their pupillary
succession back to them. However, in about 1765 a royal decree,
claiming that unworthy monks had been ordained, restricted the
holding of <1upasampada$>1 ceremonies to the monasteries of Malvatta
and Asgiriya in Kandy. Such a restriction is contrary to the
<1vinaya>1 rules, but has been in force ever since. Though peculiar
to Ceylon, it has been copied by the other Nika$yas; they all
restrict ordination ceremonies to Nika$ya headquarters. Its pur-
pose must have been to tighten central control, in particular royal
control, over the Sangha; whether caste had anything to do with
it we cannot tell, but the fact that the king, Ki$rti S/ri$ Ra$jasim*ha,
issued a <1katika$vata>1 repeating the above-mentioned extract from
the Dam#badeni Katika$vata inclines one to suspect it. Already in
1769 monks replying to a questionnaire from the Dutch governor
of the coastal provinces stated that according to the <1vinaya>1 any
caste was allowed in the Sangha, but that in fact admission was
restricted to high-caste men. John Davy, who was in Kandy from
1816 to 1820, reports that only <1goyigama>1 men can become monks.<s30>s
There are cases between 1773 and 1798 of novices from the fisher
(<1kara$va>1) caste which ranks second to the <1goyigama>1, receiving the
<1upasampada$>1, but these seem to have been exceptions allowed on
orders from the head (Maha$na$yaka) of the chapter.<s31>s
  It was in 1798 that Am#bagahapit*iye$ N#a$n*avimalatissa, a member
of the cinnamon-picking (<1sala$gama>1) caste, went to Burma to be
ordained; on his return in 1802 he founded the Amarapura
Nika$ya. He was soon imitated by members of the fisher (<1kara$va>1)
and toddy-tapping (<1dura$va>1) caste. These three castes are virtually
                     309
confined to the Low Country, and are there reckoned below only
<1goyigama>1 in status. The Amarapura Nika$ya thus contains several
ordination traditions (<1parampara$>1). In 1864 the Ra$man+n+a Nika$ya
was similarly founded by Am#bagahavatte$ Inda$sabhavaran+a$na.
   If the Amarapura Nika$ya contains more than one ordination
tradition, it may well be asked what it is that constitutes a Nika$ya.
The answer to this question is complicated, but only because the
word <1nika$ya>1 has come to have two closely related but distinct
uses. In the Introduction was described how throughout most of
the Anura$dhapura period there were in Ceylon three <1nika$yas>1, go
that the term "the three <1mika$yas>1' was locally synonymous with
"the Sangha'. It may be this historical fact which has determined
the modern usage of referring once again to "the three Nika$yas', by
which is meant the Siyam, Amapura, ad Ra$man+n+a Nika$yas. (In
this sense I am spelling Nika$ya with a capital N, and not italiciz-
ing it.)<s32>s This usage is universal and has some organizational
correlates. From the point of view of ordination tradition, how-
ever, none of these Nika$yas is a unity. Nor is its origin the crucial
criterion for establishing what is a <1nika$ya>1 in the narrow, technical
sense of the word. A <1mika$ya>1 is a body of monks with a head
(<1maha$na$yaka>1) which holds independent <1upasampada$>1 ceremonies
for its novices, whether the line of pupillary succession has been
renewed from abroad or there has merely been a fissure caused by
a group of monks stating their autonomy. A <1nika$ya>1 in fact can be
said to exist when it has been recognized by the Registrar-
General.<s33>s
  A list of the <1nika$yas>1, based principally on the report of the
S/a$sana Commission,<s34>s appears on pp 263-5 of Bechert's book.<s35>s
As Bechert remarks in his note 718 (pp. 212-13), there are also
some monasteries in Ceylon which do not belong to any Nika$ya:
these are the creations of modernists, often Europeans, and need
not concern us.
  There are absolutely no differences in doctrine between the
310

three Nika$yas, merely a few differences in customs. These are listed
by Bechert on p. 215. The Siyam Nika$ya as he points out, carries
on the old Ceylonese traditions, whereas the other two Nika$yas
have imported some customs from Burma. The Ra$man+n+a Nika$ya,
the most recently formed, has not entirely lost its original reforma-
tory zeal, and its monks, like those of some small Amarapura
<1nika$yas>1 with similar origins, observe the <1vimaya>1 rules more strictly
in certain particulars, also mentioned by Bechert.
   The main principle of differentiation between <1nika$yas>1 is, how-
ever, caste. In general this is well known both to western scholars
and within Ceylon; but I found a state of affairs more extreme
than that reported by Bechert, whose work on the subject is the
most complete and up-to-date published.<s36>s Between the various
branches of the Siyam Nika$ya I know of no caste differences,
though it is rumoured that certain monasteries in Malvatta and
Asgiriya (which are really two groups of monasteries sharing
certain facilities) admit only <1radalas>1, the top <1goyigama>1 sub-caste.
That the Amarapura Nika$ya is divided mainly on caste lines was
already stated by Ryan in a passage quoted by Bechert (pp. 219-
20).<s37>s It does not normally recruit from the <1goyigama>1 or the very
low castes, though in this certain <1nika$yas>1 are exceptional (see
below). Bechert, following Ryan,<s38>s is wrong in stating that the
Ra$man+n+a Nika$ya is casteless, even though this was its original
intention. It is generally believed in Ceylon that the Ra$man+n+a
Nika$ya occupies an intermediate social position between the other
two Nika$yas, and this appears to be correct. I cannot confirm or
refute the popular opinion that in the Low Country it recruits
predominantly from the top two castes, <1goyigama>1 and <1kara$va>1;
I was, however, told by all the Ra$man+n+a monks whom I inter-
viewed (and am sure it is true) that the Kandyan <1(ud*a rat*a)>1 branch
admits only <1goyigama>1 men, just like the Siyam Nika$ya. Indeed,
caste has penetrated the Ra$man+n+a Nika$ya so deeply that one of
their young monks tried to convince me that they are socially
superior to the Siyam Nika$ya, claiming (untruly) that the Siyam
Nika$ya admit some low-caste men while the Ra$man+n+a does not.
                     311

   In reviewing the answers I got to my questions about caste I
must distinguish between the members of the three Nika$yas,
because the six Amarapura monks were themselves low-caste
while the rest were <1goyigama>1, and the Ra$man+n+a Nika$ya has an
official policy of cestelessness while the Siyam Nika$ya does not.
Common to all, however, was a readiness to describe the differ-
ences between the three Nika$yas in terms of caste; a couple of
monks delicately referred to varying "customs' (<1sirit virit>1). The
answers of Amarapura monks held no surprises: they were
against caste altogether, said they made no distinctions even
against <1rod*i>1, and in support of their views even cited the passage
given above from the <1Vasala Sutta:>1 "Not by birth is one a brahmin
. . .'. Even they, however, made no bones about stating that the
Nika$yas are mainly differentiated by caste. As might be expected,
they were perhaps readier than the Siyam Nika$ya monks to claim
social intercourse (<1a$sraya>1) with the other Nika$yas, if such a state-
ment can be made on the basls of so small a sample.
  The attitude of monks of the Siyam Nika$ya towards caste
ranged from the monk who thought all caste distinctions were
false and wrong ("<1boru va%d*ak>1') to the unquestioning acceptance
of many monks, especially the older ones, coupled perhaps with
some dismay at the disappearance of caste barriers. A young
westernized monk claimed to be against ceste distinction
(<1kulabhe$daya>1), but belied his statements by his conduct, addressing
even an old man of the drummer caste by a very low pronoun
(<1u$>1). Very few monks regard caste distinctions in lay society as
something undesirable or troublesome to the conscience; it is a
fact of life, and the scriptural position that low caste is the result
of sin in a former birth is widely upheld. The origin of caste was
referred to the canonical text on the subject, which tells of the
first king (Maha$sammata) and how the four classes arose by a
social contract. One monk slightly perverted this story by claiming
that cestes were created by kings to punish people for their bad
acts--a reference perhaps to the origin of certain Sinhalese
degraded castes. He then astonished me by interpreting the Vasala
Sutta in this way: people are high or low ceste "not by birth', but
by their acts--so it is the low castes' own fault! Even he, however,
said that essentially all men are equal (<1a$tme$ vas/ayem ekayi>1), by which
he meant that men of all castes have the capacity for Enlighten-
ment. This I think no Sinhalese Buddhist could be found to deny.
312

  Siyam Nika$ya monks are generally well aware that in the time
of the Buddha the Order was open to all. Only one monk stated
the opposite. One monk represented such freedom as exceptional,
due to special circumstances in their former lives. As an instance
he told the story of Upa$li: six cousins of the Buddha went to join
the Order and took along Upa$li, their barber. In the forest they
prepared themselves for ordination, and gave him their lay clothes.
He thought that if he returned alone with the clothes of the six
princes he would be assumed to have murdered them for their
valuables, so he took fright and decided to join the Order too. The
Buddha ordained him first so that he was senior to the other six
and they had to worship (<1van#dint*a>1) hirn; he did this to ensure that
they would not return to the lay life, because they would be so
ashamed at having worshipped a low-caste man. Upa$li had been
born in a low caste because in a previous life he was haughty
towards his inferiors; but he had good <1karma>1 to get him into the
Order because for innumerable ages (<1kalpa laks*ayak>1) he had wished
to be a monk under the Buddha. My informant has related the
story as it is told in the <1Vinaya Pit*aka>1,<s39>s with one twist: in the
original text it is the princes who suggest, as a token of humility,
that Upa$li be their senior; but in the new version it becomes the
Buddha's idea, to rub in the degradation of prostrating oneself
before a barber.
  The story which Siyam Nika$ya monks will almost invariably
tell when asked why they only admit <1goyigama>1 recruits is, in view
of the historical antecedents quoted above, presumably apocryphal.
The king (often specified as Ki$rti S/ri$ Ra$jasim*ha, or the last king
of Kandy, S/ri$ Vikrama Ra$jasim*ha), who is of course ritually
inferior to the meanest novice, was paying a call on a monastery,
when a low-ceste monk so far forgot himself as to rise from his
seat at His Majesty's entrance. The king was so embarrassed by
this breach of decorum that he decreed that henceforth only men
of good birth could enter the Order. Monks explain that they are
bound to obey the temporal power. I would suggest that the
temporal power of today does not demand ceste exclusiveness in
the Sangha, and most had no answer to this, but one replied that
the king's injunction is embodied in a <1katika$vata>1, and as it has
never been rescinded it still has the force of <1vinaya>1. I asked several
monks what would happen if they gave the lower ordination to a
                     313

low-caste pupil, and all agreed that he would be refused the higher
ordination at Malvatta or Asgiriya. At least half the monks ques-
tioned on this point seemed, however, somewhat ashamed of the
state of affairs, and candidly admit that entry into the Order should
be unrestricted. In strong contrast, one monk said that all respect-
able (<1sa%lakiyayutu>1) people joined the Siyam Nika$ya; some <1goyigama>1
joined the Ra$man+n+a Nika$ya, but this was not good [i.e. letting
the side down].
  As mentioned above, the three monks of the Ra$man+n+a Nika$ya
whom I interviewed were also <1goyigama>1, and I gathered from them
that locally, and probably in the whole of the Kandyan area, their
Nika$ya too was restricted to this caste. One of them, who went so
far as to tell me that his Nika$ya was purer caste-wise than the
Siyam Nika$ya, behaved very like a Siyam Nika$ya monk in telling
me that he personally disapproved, but his headquarters would
not allow him to ordain a low-caste pupil.
  Caste also plays a part in determining the behaviour of monks
to laity. The extreme cese is that of the <1rod*i>1, who were not admitted
to monasteries or temples. One monk said that the Sinhalese kings
prohibited it. John Davy wrote:<s40>s "Prohibited from approaching
temples, there is a solitary instance on record of a priest going and
preaching to them, for which, having incurred his sovereign's dis-
pleasure, he replied, "Religion should be common to all".' That
priest would nowadays find more approval but still few emulators.
I believe that the settlements of the <1rod*i>1 and the <1kinnara>1 (the next
lowest caste) lack monasteries, so that they do not normally hear
preaching; but a famous monk from Vajira$ra$ma, Na$rada Thera,
has gone and preached to them, and I doubt whether anyone now
would censure him. Not surprisingly, the only monks who indi-
cated that they would not allow <1rod*i>1 in were three elderly members
of the Siyam Nika$ya. Another monk, who was generally reactionary
on this subject, said they could go into the <1viha$ra>1 but not come
into the <1pansala>1. One of the Ra$man+n+a Nika$ya monks said that of
course <1rod*i>1 could come if they wished, but they would normally
not enter a <1pansala>1. Most monks, however, seemed to think that
they both could and would come, and welcomed the change.
   One Siyam Nika$ya monk, an amiable character, went so far as
to specify that <1rod*i>1 could even give a <1da$ne$>1. Were such a situation
ever to arise I wonder what would happen. The question whether
314

<1goyigama>1 monks (perhaps I should say Siyam Nika$ya monks) will
accept food and drink from a low-caste person is most curious.
There is no doubt that according to doctrine they should, and I
was always assured that they do; however, I witnessed a counter-
example. A large group of Siyam Nika$ya monks were invited to a
washennan's home for their midday meal; they came and ate the
food, but brought with them a lot of king coco-nuts, from which
they drank the liquid instead of accepting ordinary water from
the householder. Though no one would adrnit to there being any
particular reason for this I cannot think that they were actuated
merely by gormandize. They might take pork from a smith, but
they refuse water from a washerman.
   In considering the attitude of the laity towards monks of
different caste, we come perhaps to the nub of the problem.
Brahmins reviled the Buddha and his followers for their ritual
impurity, but they did not mind; their successors have proved less
stoical. Davy brings up this point:<s41>s "Only the Goewanse\ ... can
belong to the priesthood; not that religion excludes the rest, but
pride--people of the first caste not being able to condescend to pay
the respect due to a priest to a low-caste individual.' A <1goyigama>1
friend of mine confessed to me that he still feels the same way. The
Siyam Nika$ya story about the king and the low-caste monk is no
doubt an attempt to show this tension between high-caste layman
and low-caste monk in another light. In the area which I have
observed, the usual solution to this problem is avoidance: high-
caste laymen just do not normally go to Amarapura monasteries.
If they do for some reason have to go they fail, unless they are
very religious individuals, to pay their traditional respects by
kneeling and touching or making to touch the monk's feet. High-
caste monks condemn this failure, but in fact they themselves
seem to avoid formal contact with their low-caste brethren; a
Siyam Nika$ya informant had been invited by an Amarapura
Nika$ya informant to the <1ne$tra pinkama>1 at his temple, at the other
end of the same village, but had not gone and had never even been
to visit him.
  So long as this remains the prevalent attitude among <1goyigama>1
laymen it is hard not to sympathize with the following argument.
The monk who offered it was a <1goyigama>1 member of the Ra$man+n+a
Nika$ya, who amply demonstrated his own freedom from traditional
                                                          315

prejudice by planning an alms-giving to be attended by members
of all three Nika$yas, an event which would I believe be unique in
that area; he said he had friends in the Amarapura Nika$ya, and
had no time for ceste distinctions. However, he had taken only
<1goyigama>1 pupils. He lived, he explained, in a village which was
almost solid <1goyigama>1; if he took a low-ceste pupil they would find
fault, and some of them might not even bow down to him. Plainly
--if I may expand on his statement--a low-caste pupil would be a
dissster' as his parishioners would stop attending the monastery,
which would deprive him of his subsistence; some would go to a
monastery elsewhere, and some would stop going at all, which
would deprive them of merit. In all practical ways the cause of
Buddhism would suffer, merely to accommodate a young man
who could always find a niche in the Amarapura Nika$ya.
  The involvement of monks in the caste system is the result of
their active involvement in a caste-ridden society. The Buddha
told his monks<s42>s to travel around "for the public welfare and
happiness, out of compassion for the world', i.e. to keep in touch
with the laity. The major unintended consequence of this involve-
ment was the evolution of the Sangha in the first centuries after
the Buddha's death from a fraternity of wandering mendicants
to a loose federation (or group of federations) of settled monastic
communities with communal property. It was given some rationale
by the decision in ancient Ceylon (ascribed by Rahula to the first
century B.C.)<s43>s that learning (<1pariyatti>1) was of paramount impor-
tance, and therefore book-duty (<1gantha$dhura>1) was more essential
than meditation-duty (<1vipassana$-dhura>1). In this decision was
implicit the previous ethical decision that the preservation of the
doctrine (Pali: <1sa$sanat*t*hiti>1) was of paramount importance, a social
aim not necessarily compatible with the original doctrine of indivi-
dual salvation. Although the preservation of the doctrine was seen
largely in terms of the preservation of the Sangha, and book-duty
meant preservation of the texts within the Sangha, a Buddhist
laity is essential for recruitment to the Sangha and for its economic
support; book-duty therefore necessarily involves preaching and
other religious education, and therefore residence near the laity
and considerable contact with lay society. Compared to this general
abandonment of the wandering life, a development which began
before Buddhism left India, the secularization of monastic
316

organization in mediaeval Ceylon is of secondary importance.
Monasteries became landlords and their incumbents (<1viha$ra$dhi>1-
<1pati>1) feudal lords, so that the monks added temporal power to
their traditional prestige. An incumbency (<1ayitiva$sikama>1) became
hereditary, so that monks were tempted, not to put too fine a point
on it, to keep wealth in the family, and passed their position on to
relatives.<s44>s This tendency for the families of monks to acquire
wealth and power, and the general approximation of the role of
monk to that of feudal landlord, must both have worked towards
keeping monasteries in the hands of the top caste ; since the pattern
of family ties and caste exclusiveness had been established it was
difficult for other Nika$yas, although founded with reformatory
ambitions, not to go the same way.
   If I am right, the demands of modennists that monks should
involve themselves still further in society, after the model (usually
not expressed) of Christian priests, would tend to perpetuate
rather than diminish caste in the Sangha. This would be an
unintended consequence of their actions; their intentions are of
course quite in the opposite direction, for like early reformatory
movements (e.g. the Ra$man+n+a Nika$ya) they hope to restore the
primitive castelessness of the Sangha, or even to restore an
imagined primitive castelessness among Buddhist laity. My
hypothesis is unfortunately unlikely ever to be put to the test,
                    317

because the influence of the modernists is a puny force compared
to the general trend towards modernization which is breaking
down caste barriers and other traditional societal patterns. How-
ever, my observations may support this interpretation. The atti-
tude monks have to caste among the laity, as exemplified in their
attitude to <1rod*i>1, has become far more liberal, in confonnity with
general modern trends in society and also with the views of
Buddhist modernists; but I found no signs that caste within the
clergy is losing its grip, for here the ever greater involvement of
monks in society works in the same direction as traditional
pressures, and the economic forces which play so much part in
breaking down caste in society at large have no influence.
  It only remains to consider why ceste should be so much less
oppressive in Sinhalese than in Indian society. What has already
been said of Buddhism must render the answer obvious. (See also
Chapter 4, p. 183.) Pollution plays far less part in Sinhalese than in
Hindu society. The Buddha did not recognize the existence of
such a thing as far as he and his fellow monks were concerned. The
only ethic with which he was concerned was a universalist ethic:
there is no mention in the scriptures of caste <1dharma>1, no suggestion
that an action which is right for one man may be wrong for
another. According to Buddhist theory, caste originated as a matter
of convenience and has no more ethical implications than wealth or
beauty. It is part of Ceylon's Indian heritage; but in Buddhist
society it has lost its rationale. So although caste is not condemned
it is de-emphasized. The traditions of Hinduism tend to stress the
differences between men; the traditions of Buddhism lay the
accent on their similarity.
                    CONCLUSION



     What this Niwane$ is, is a religious mystery: priests are
     rather averse from answering questions on the subject;
     they say, it is forbidden to discuss its nature, and on the
     principle that if men understood it they would not like it
     but prefer worldly things, as flies do bad smells.<s1>s


AT the outset of this work I said I would distinguish what people
say from what they do. At its conclusion I wish to emphasize what
this means. It is not the same as the distinction made between
behavioural and textual Buddhism; for that is a distinction, as I
understand it, between what people do now and what they said
(or wrote) a long time ago. My investigation centred on what
people <1say now>1, as well as what they do now, and I found, as will
not surprise anyone at all familiar with Ceylon, that there are
discrepancies. What people say now has turned out, however, to
be remarkably similar to what was being said about 1,500 years
ago. But are we sure that in those days they really did what they
said, that they practised what they preached? I doubt it. The
evidence of the <1Maha$vam*sa>1 suggests to me that there were then
discrepancies between Buddhist principle and Buddhist practice
quite as gross as anything found today. Around 100 B.C. Dut*u-
ga%mun*u puts a relic in his spear (magic talisman!) to fight a holy
war (counter to doctrine!) for Buddhism, makes a statement of
truth (magic spell!) to protect his army, is told by monks that
killing non-Buddhists is not killing (intolerance! ignorance of first
precept!), puts up vast relic chambers with a statue of the Buddha
(devotional theism!), makes Buddhism the state religion (Bud-
dhism in politics!), etc. The monks in the fifth century who com-
piled the <1Maha$vamsa>1, and predicted Dut*uga%mun*u's rebirth as
the chief disciple of the next Buddha, obviously thought his holy
war fully justified. Not all my informants were so sure.
  The Reverend Rahula, who knows as much about the subject
                                                          319

matter of this book as anyone alive, likes to call all this "popular
Buddhism'. Dut*uga%mun*u's Buddhism was presumably "popular'.
I find this term dangerously misleading. It implies, I think, a
decline by most people, the <1plebs>1, from an ideal standard which is
maintained by a few spiritual aristocrats, a relationship analogous
to that between "popular' and "classical' music. By some it might
even be understood to imply the religion of the laity as distinct
from that of monks. To this latter assumption, which would be
especially erroneous, I shall return below. To the former I would
reiterate a request for demonstration of "classical' Buddhist
behaviour, outside the edifying stories in scriptures and textbooks
which are inserted for didactic purposes.  Monks fought in
Dutuga%mun*u's army and sanctioned what he did. And how many
Buddhists in Ceylon, even in this post-Dharmapa$la era, never take
part in <1pirit>1 or transfer merit, practices they brand as "popular'?
I think the only sense in which it is accurate to describe these
beliefs and practices as "popular' is the everyday one of "widely
liked' or "prevalent'. But if so used "popular' no longer dis-
tinguishes merit transference from, say, the Four Noble Truths or
alms-giving.
  Gananath Obeyesekere in his lectures has pointed out that
Weber too mixed up what the texts say and how people actually
behaved. In <1The Religion or India>1 he paints a picture of the
passionless <1arhat>1, the materials for which are drawn (mainly at
second hand) from Pali scriptures, deduces from this picture of
the ideal Buddhist personality that the first Buddhists were in
fact like that, and then from this draws conclusions about their
social and economic behaviour. "The mystic, acosmic love of
Buddhism (<1maitri$, metta$)>1) is psychologically conditioned through
the euphoria of apathetic ecstasy.... [The <1arhat's>1] temper...
remains cool and aloof...'.<s2>s Weber then has trouble harmonizing
this picture with the <1Theraga$tha$>1 and <1Theri$ga$tha$>1, the two canonical
collections of religious poems composed by the monks and nuns
themselves. "The tone to which the hymns of ancient Buddhism are
attuned is triumphant joy.'<s3>s
  This should have brought Weber face to face with the issue
which, however, he fudged, of whether the Buddhist ideal was
really negative, as he contended, or positive, which he claimed to
be a later development. The Five Precepts, like the ontology, are
320

negatively formulated, but have they ever constituted the whole
ideal? To put the matter more clearly, is the ideal self-restraint or
love? Not only are there two ideal values, which can be character-
ized as negative and positive; there are two possible ideal goals,
<1nirva$n*a>1 and heaven, which can be similarly contrasted; and two
possible courses of action which characteristically lead to these
goals, self-restraint and the outward passivity of meditation, or
active "merit-making', typically giving alms.
   It is in the light of these and related dichotomies that I would
like to survey the material collected in this work. In Chapter 6
we saw that these two values, self-restraint and love, usually
expressed by the adjectives <1s/a$nta da$nta>1 and <1karun*a$vanta>1, are
dominant among the Sinhalese today. The ideals are very different,
and not necessarily compatible. The contrast--if you like, the
tension--between them illuminates the history of Buddhism.
   We saw at the end of the previous chapter that early in the
history of Ceylonese Buddhism the question arose, "Which is the
basis of the doctrine, learning or practice?',<s4>s and that underlying
this controversy, which foreshadowed the division of monks into
book men and meditation men, was the essentially altruistic con-
cern for the preservation of Buddhism. The same dilemma,
though perhaps unrealized, already underlay the Buddha's injunc-
tion to monks to keep moving among the laity. This conflict
between self-restraint and love, between selfishness and altruism,
between negative and positive ideals--put it how you will--can, I
think, be traced back to the Buddha himself. In his biography,
which though preserved in late versions Frauwallner has shown
to date from only about a century after his death, he hesitates on
becoming Enlightened as to whether he should go forth and preach
the truth. Only with difficulty is he persuaded to do so. By
acceding to the gods' request that he preach he gives himself
forty-five years of unnecessary trouble; but he acts from his
universal compassion, and comes as near as anyone can in Bud-
dhism to saving mankind. Of course, being already Buddha he has
it both ways: he fulfils the ideals of both self-restraint and love.
When he decides to preach he is already an <1arhat>1; by founding a
<1s/a$sana>1 he becomes a <1samyak saddha>1. What many scholars,
especially the older ones on whom Weber relied, have failed to
                                                          321

notice is that for <1all>1 schools of Buddhism the <1samyak sambuddha>1,
who has added love to self-restraint, is infinitely superior to the
mere <1arhat>1, who does not benefit others after achieving Enlighten-
ment himself. Even those who regarded the <1arhat>1 ideal as the
<1summum bonum>1 did not consider it the <1solum bonum:>1 for all
Buddhists love is a great good.
  For the Buddha there was no incompatibility between the ideals
of self-restraint and of love; he could pursue both. Most people
are not so lucky. In practice it usually comes down to making a
choice. There is not much hurry, because we have many lives, but
in this life which shall we make our goal--our own Enlightenment
through self-restraint, or the service of mankind?
  This tension between the two ideals seems to me fundamental to
the distinction between Therava$da and Maha$ya$na. To equate
Therava$da with self-restraint and Maha$ya$a with love is mis-
taken: it is a question of emphasis. Maha$ya$a, which faced the
difficulty more explicitly, elevated the ideal of the Bodhisattva,
who postpones his own Enlightenment till the goal of his love, the
Enlightenment of all others, has been reached. Therava$da too
presents the Bodhisattva as a noble ideal--but he does not fit
in very well. What is emphasized is not his postponement of his
own salvation, but his determination eventually to become a
<1samyak sambuddha>1, and thus to realize both more or less simul-
taneously.
  In this respect Therava$da Buddhism has preserved what seems
to me, from the logic of later developments, must have been a
feature of Buddhism from the very beginning. If the <1summum>1
<1bonum>1 had always been love, and all other goods had been sub-
ordinate to this, there would have been no problem. and it would
be hard to account for the appearance of the opposite doctrine,
the <1arhat>1 ideal in the Pali scriptures. If however, the <1summum>1
<1bonum>1 was declared to be self-restraint and one's own salvation,
there must have been immediate obstacles to the translation of this
ideal doctrine into actual behaviour. Firstly, if the ideal is con-
sistently carried out, the religion would have little chance of
survival, depending purely on a host of individual conversions
unaided by any organization. If the Buddhists were utterly con-
sistent this would not matter. But the second obstacle must have
been human nature. All the evidence shows that many Buddhists
were neither so consistent nor so averse to the altruistic ideal as
322

Weber suggests and, perhaps, as they were intended to be. Those
who kept themselves to themselves as forest-dwelling <1arhats>1
have departed unsung (except by themselves), and few of their
names are recorded in history. But throughout the history of
Therava$da Buddhism, by which phrase I would like to subsume
nearly 2,500 years, there has been this problem: according to
doctrine self-restraint is the <1summum bonum>1, to which all other
values are subservient; but in practice there has been no such
clear hierarchy of values, but rather a preferential value system,
which at the highest level comprises a choice between self-restraint
and love.
  It is, I think, this dilemma which has underlined many of the
differences between doctrine and behaviour, between what I have
called cognitive and affective religion. (My choice of the word
"affective', with its overtones of "affection', turns out to be for-
tunate.) Again, I am certainly not suggesting that the cogritive
religion should be equated with the ideal of self-restraint, the
affective religion with the ideal of love--it is rather a question of
which way they lean. Nor, I must repeat, should "self-restraint'
and "love' be interpreted too narrowly: it is here a question of
whole syndromes, the one negative and world-denying, the other
positive and world-affirming. (My choice of words in this sentence
is an attempt to connect my concepts to those of Durkheim and
Weber.) Amplification will make this clear.
  The cognitive religion regards withdrawal for meditation as the
consummation of religious behaviour, and the utterly negative
(and rationally inconceivable) goal of <1nirva$n*a>1 as the ideal destiny.
In fact most people, even monks, say that meditation is impractic-
able, offering some such excuse as the pressure of affairs or the
decline of the doctrine, and they display very little liking for
<1nirva$n*a>1. Accordingly, <1nirva$n*a>1 has been postponed in the popular
mind to the coming of Maitri$, in some remote future era; it is so
difficult to attain, and we are not sure that we want it anyway.
Most people, monks included, devote themselves exclusively to
acts of merit (<1pinkam>1), the aim of which is a good rebirth in heaven
or on earth. The aim is made conformable to doctrine by saying
that it all counts towards <1nirva$n*a>1 in the long run; the list of
<1pinkam>1 conforms to doctrine by including, as two of the Ten Good
Deeds (<1dasa kusala karma>1), meditation and the negative "morality'
(<1si$la>1) of the Five Precepts; but if people really want to be sure of
                                                         323

merit they do not sit around practising self-restraint: they go out
and do something positive like giving alms, which is emotionally
more satisfying.  Cognitively <1pin>1 and <1nivan>1 are hierarchically
associated as means to an end; affectively, however, the two com-
plement each other, and the values are preferential.
  Finally, it must surely be significant that the name of the next
Buddha, Maitri$, <1means>1 "love'.  By postponing meditation and
<1nirva$na>1 till the time of Maitri$ people are thus rendering explicit,
on the symbolic level, that they attach at least as much value to
the positive as to the negative ideal.
  In various parts of this work, especially Chapter 5, I have shown
how doctrine has to some extent come to terms with these emo-
tional needs. The history of parts of Therava$da doctrine can
indeed be seen as a running battle between doctrine and senti-
ment. Formulations of ethical prescriptions have moved from the
negative to the positive, as just mentioned, without sacrificlng con-
sistency. An even better example is the history of the transfer of
merit, where the emotional need to do something for your dead
relatives, which according to original doctrine was impossible, has
been elaborately rationalized to make it compatible with the
fundamental doctrinal principle of self-help. Again, self-restraint
has accommodated to love.
  Though one of the tenns in my explanation is a component of
human nature, I have come nowhere near to saying that human
nature and human religion are everywhere much the same. The
feature of Therava$da Buddhism which I find most striking is
indeed the counter-intuitive persistence of the self-restraint ideal,
and the victories of doctrine. The latter are achievements of the
scholastic intellect; the former is exemplified in the prestige of
meditation and the extraordinary veneration in which any self-
styled hermit is still held in Ceylon. The phenomenon of Salgala
has already been cited in Chapter 7. Here monks have retired to
meditation in caves, but they are pursued by what Obeyesekere
dubs "the relentless piety of the masses', who cement the caves and
book a year in advance to offer food. In Obeyesekere's words, "The
<1arhat>1 ideal is still the index of religiosity.'
  The fate of the monks of Salgala conforms to a pattern which
recurs throughout Buddhist history: the few men who really wish
to follow the <1arhat>1 ideal break away and retire to jungle caves, but
as they are dependent on the laity for their food they do not break
324

off all contact with the outside world, and so they or their suc-
cessors run the risk of reverting to the state of an institution.
Traditionally these hermits--Pam*suku$likas, A$ra+n+ava$sins--have
been members of the Sangha, but this need not always be so:
recently the short-lived <1ta$pasa>1 movements consisted of hermits
who considered the Sangha corrupt and not worth joining. The
recurrence of these movements shows that there is often (or
always?) a small group of people for whom the unadulterated
<1arhat>1 ideal is attractive and not as psychologically implausible as
it is for most. We may conjecture that some of the Buddha's
original followers were of this type.
   This brings us to the whole question of the monk-layman
dichotomy. The cleft between them is obvious and needs no
denying. But writers on Ceylonese Buddhism tend to associate
with the two groups in a one-one correlation the two different
types of religiosity which I have typified by reference to the
ideals of self-restraint and love. The monks are all supposed to be
"salvation-strivers', and the laity interested mainly in heaven
attained by faith and good works. This finds some justification in
the texts, but none in social reality. Many monks reject this view
of themselves, and I think they are right to do so. They point out
that they need <1pin>1 as much as anyone else. Just like laymen they
observe astrological recommendations at the New Year or when
embarking on a journey, and they give <1mataka da$me$s>1 for their dead
parents; on analogy with laymen they treat their pupils as sons,
their fellow pupils as brothers, and give <1mataka da$ne$s>1 for their
dead teachers. Indeed, all the inhabitants of Mi$gala monastery
used to prostrate themselves every evening before the photograph
of the dead incumbent. Some <1pansal>1 display the picture of Si$vali,
the symbol of aspiration to material prosperity. I found monks, as
a group, very kind, hospitable, and worldly people. Ethically they
are mostly like laymen--but the point is whether they are like
<1good>1 laymen.
  It is perfectly true that this development is only to be expected
when monks are selected in childhood, and therefore constitute,
so far as concerns their personalities, a random sample of the
population rather than a group with a religious vocation. Bechert
suggests that this juvenile recruitment is bad for Buddhism, and
prefers the system in other Therava$da countries by which most
                                                         325

males enter the Order for at least a short period. Whichever is
more beneficial for society, I am sure the Buddha would have
found the latter system at least as surprising a development as the
former. However, the point that I wish here to make is that under
any system the majority of monks are not likely to be unambigu-
ously devoted to the <1arhat>1 ideal alone. Already in a canonical text
we find references<s6>s to <1puthujjana-kalya$n*aka-bhikkhu$>1, quaintly
rendered by Mrs.  Rhys Davids<s7>s as "hoi-polloi-good-fellow-
bhikkhus'. <1Puthujjana>1 is originally a technical term for an un-
enlightened person; it comes to connote, it seems, a person with
lay ethics. And, once again, this shift from the <1arhat>1 ideal has been
officially accommodated: these monks are designated village-
dwelling book men <1(gra$mava$sin>1 and <1granthadhura>1). It is doc-
trinally impossible entirely to abandon the self-restraint ideal, but
it has been pushed right into the background, and these monks
too are tacitly sanctioned in a course of making merit like the laity.
If they do not leave the monasteries, either reverting to lay status
or at least abandoning more <1vinaya>1 rules than they do, it is because
of the fear of disapproval, the material benefits of monastic life
and the continuing demand and admiration for examples of self-
restraint.
  If we wish to observe where the self-restraint ideal voluntarily
functions, apart from the sporadic hermits who furnish the
extreme case, we must look not to the ordinary village-dwelling
monk, but to the <1upa$saka>1. Though an <1upa$saka>1 originally meant
any Buddhist layman, the term is now used mainly for those who,
usually late in life, renounce worldly ties. Like the early members
of the Sangha, they choose their status rather than having it
thrust upon them, and free choice is a prerequisite for the whole-
hearted pursuit of any moral ideal. Some old village <1upa$sakas>1 have
a knowledge of scriptures and doctrines which would be remark-
able in a person of monastic education. Approaching death, how-
ever kindly they may be, they conceive <1maitri$>1 primarily as an
exercise in solitary meditation; their goal is unambiguously
selfish. Might we here find Weber's Therava$din recluse, "cool
and aloof'? Obeyesekere interviewed such a man. An inhabitant
of a remote village, he had been illiterate but could now read both
Sinhalese and Pali; he led a very simple life, and was respected
326

by the villagers, to whom he preached. One can also assume that
he meditated. Yet while he talked of the foulness of the female
body, he gave glowing descriptions of the damsels in heaven, which
he envisaged as a paradise of sensual delight. According to doc-
trines which he had studied and accepted, the next life conditioned
by his <1karma>1 would be experienced by a new constellation of
impermanent constituents (<1anatta>1 doctrine); all rebirth is un-
happiness, and <1nirva$n*a>1 the only goal worth seeking; and pious
works are in vain unless they are accompanied by freedom from
desire, which constitutes the only true progress. But these doc-
trines are counter-intuitive.  He would probably never have
admitted it, or been able to see the discrepancy so apparent to us,
but we can surmise that he was not interested in <1nirva$n*a>1 so much
as in the female inhabitants of a heaven with whom he felt that the
renunciation and other good works of his old age entitled him to
associate. He practised restraint now so that he could indulge later;
action rationally based on premisses inconsistent with those he
professed. I do not belittle him. I find his principles noble and
profound, his conduct virtuous. But the two are not related quite
as he imagines. Though he would probably never have admitted
that he was more interested in heaven than in <1nirva$n*a>1, that was
plainly where he hoped to be reborn.
  The <1upa$saka's>1 discordant aspirations exemplify the tension
between the two ideals, as well as the dichotomy between cognitive
and affective religion (rather than between belief and action, for
here the action suits both the cognitive and the affective belief
system, and the behaviourist would have nothing to go on) with
which this work has been largely concerned. They neatly illus-
trate also the unintended consequences of the adoption by a whole
population of a religion which teaches that total renunciation of
life is the only final answer to life's problems. Those who cannot
make that renunciation require something else, and another
system, logically subsumable to the former but emotionally
alternate, comes into being. Both my argument and my evidence
suggest that this happened very early in Buddhist history, at least
before Dut*uga%mun*u, 2,000 years ago. The adaptation was
excellent: the accommodated system has survived with little
change to this day.
  Early Therava$da was an uncompromising doctrine. It is con-
tinually difficult to be entirely responsible for your fate. "For
                                                         327

emotional mass religiosity there have been and are but two pos-
sible types of soteriology: magic or a saviour.'s Sinhalese "mass
religiosity' has accepted in deed, though not in word, a modicum
of both: the saviour we met in Chapters 3 and 4, the magic in
Chapters 4 and 5. These acceptances are as old as the commen-
taries, some of them far older. The words are hardly changed at
all. If this is popular Buddhism, could it be that

                   <1Vox populi vox Buddhae?>1