← Volver a la ficha del textoEARLY BUDDHIST
THEORY OF KNOWLEDGE
BY
K. N. JAYATILLEKE
B.A.(Ceylon), M.A.(Cantab), Ph.D. (London)
Sometime Research Student in Moral Science, University of Cambridge
Nuffield Fellow in the Humanities, 1960-1
London
GEORGE ALLEN & UNWIN LTD
RUSKIN HOUSE MUSEUM STREET
FIRST PUBLISHED IN 1963
This hook is copyright under the Berne Convention .
Apart from any fair dealing for the purposes of
private study , research , criticism, or review, as per-
mitted under the Copyright Act, 19 56 , no portion
may he reproduced hy any process without written
permission. Inquiries should he addressed to the
publisher
© George Allen & Unwin Ltd, 1963
PRINTED IN GREAT BRITAIN
in 11 on 12 point Fournier type
BY UNWIN BROTHERS LIMITED
WOKING AND LONDON
DEDICATION
Ukkadharo Manussanam Niccam Apacito Maya
I shall always revere the Torchbearer of Mankind’
— Sutta Nipata, 336
FOREWORD
The present work forms an important contribution to the solution of a
number of problems more in particular pertaining to the earliest develop-
ments of Indian philosophy. In 1925 P. Tuxen observed that in any future
exposition of the history of this philosophy two factors should predominate :
1. the relation of early Buddhism to Indian thought; 2. the correlation of
the latter to the Indian science of grammar. 1 In 1927 the famous Russian
Buddhologist Stcherbatsky made the significant statement that even after
a century of scientific study of Buddhism in Europe, we were still in the
dark about the fundamental teachings of this religion and its philosophy.
At the current state of inquiry — thanks to the assiduous and penetrating
efforts of many scholars in West and East — a good deal of this ‘darkness’
has been dispelled. Yet, there are still various gaps in our knowledge to be
filled. For one thing, even though we are at present fairly well acquainted
with the later developments of systematic Indian philosophy, there is still
much uncertainty about the actual origin and incipient formative stages,
i.e. the ‘pre-history’ of its logical and epistemological and, to a less extent,
of its linguistic aspects. For another, even to-day too many misconceptions
about the exclusively mystic and recondite nature of this philosophy continue
to prevail, especially in non-professional circles. For the sphere of thought
indicated by the collective name of ‘Indian Philosophy’ is extremely complex.
Indeed, in terms of the history of ideas, its chief attraction must be sought,
not only in its spiritual and cultural unity or in the perennial truths of its
monistic-idealistic metaphysics, but rather in its rich diversity. For this is
indicative of its long development including an ever deepening confrontation
with fundamental philosophical problems. This complexity has led to highly
divergent value judgments on the part of Western philosophers as well as
professional scholars, mostly of an earlier generation. They included those
who regarded the very term ‘Indian philosophy’ as a ‘contradictio in adjecto’
and its teachings as vaguely indefinite displays of dreamy thoughts, lacking
in clear-cut concepts and proper definitions. However, other scholars were
convinced that it had reached a very high standard of development.
Stcherbatsky (e.g.) stated that, in addition to its systems of empirical idealism
and spiritual monism, it had produced an intricate logic and a remarkable
1 Cf. P. Tuxen, Zur Darstellung der indischen Philosophic, A.O., vol. IV, p. 118 f;
Th. Stcherbatsky, Conception of Buddhist Nirvana, Leningrad 1927, p. 1 ; id. Madhyanta-
Vibhanga, Leningrad, 1936, p. IV; H. Zimmer, Philosophies of India, London 1951,
p. 27 f; K. Kunjunni Raja, Indian Theories of Meaning, Madras-Adyar, 1963. B.
Faddegon, The Vaigesika-System, Amsterdam 1918, p. 12.
6
Foreword
epistemology and that the principal lines of its development showed parallels
With those of Western philosophy, including rationalism and empiricism.
Even though valid objections may be adduced to the theory of ‘parallel
development’, there are at present few doubts about the ‘high standard’.
Among other things, it is a fact that the consistent investigation of logical
fallacies and contradictions, on the basis of exact canons of reason, form an
essential part of nearly all the systems, orthodox and heterodox. And, in the
Words of Faddegon, already in early Vaisesika we find a purely theoretical
attitude of mind and not ‘that craze for liberation’ which dominates nearly
all forms of Indian thought . . . Rather, it is the theoretical desire for a
correct classification and system of definition. The variety of opinion,
mentioned above, is to a large extent induced by the problems of Indian,
i.e. Sanskrit, Pali and Prakrit, philosophical language which — as shown in
a number of recent publications — is itself correlated to the terminology and
categories of the highly developed Indian science of grammar. Especially,
the correct interpretation of the intricate technical terminology presents
many difficulties. In many cases, the same terms have different connotations,
or altogether different meanings, within different contexts and, historically,
at the successive periods of their application. Indeed, already in ancient
India, both the grammarians and the philosophers were concerned with the
problems of meaning and important works were written on this subject.
Long before this happened in the West, ‘semantics’ became a fundamental
part of the Indian philosophical discipline. Thus, in addition to a careful
historical consideration of the semantic theories, only a meticulous textual
analysis, on an extensive comparative basis, can produce valid interpretations
of Indian philosophical ideas in European languages which are both compre-
hensible and ‘intrinsic’. Moreover, to give adequate meaningful renderings
of the difficult texts, even a thorough grounding in modem philosophical
analysis is nowadays an indispensable prerequisite.
A further problem which has engaged the attention of scholars is the exact
position which early Buddhism occupied in the development of Indian
thought, the more so as it was regarded by some of them as a ‘foreign body’
in Indian philosophy. Moreover, they were of the opinion that the purely
philosophical quality of the Pali canon was surprisingly deficient. Again,
Stcherbatsky stated that the Pali-school of Buddhologists entirely overlooked
the system of philosophy which is present on every page of the Pali canon.
In his opinion, Buddhist authors played a leading part in the development
of Indian epistemology. This is certainly established for the later school of
Dignaga and Dharmakirti and their followers. Stcherbatsky’s views are
largely confirmed by the present work which is primarily concerned with
the earlier period. Dr Jayatilleke, who had the privilege of being admitted
to Wittgenstein’s classes, is that rare combination of accomplished philolo-
gist, historian and methodic philosopher. His book goes far beyond the
Foreword
7
indication of its title. On the basis of a profound analysis of the relevant
earlier and later texts as well as a critical re-examination of the works of his
predecessors in the field, he traces with great ingenuity and scholarly
thoroughness the epistemological foundations of Pali canonical thought,
from the Vedic period onwards. His fully connected account sheds new
light, not only on the problems of the earlier period which have engaged
the attention of scholars during the past forty years, but also on those of
the later developments. Moreover, with regard to the present day conflict
of metaphysics versus logical and linguistic analysis, the book contains
valuable material which elucidates from the Indian point of view some of
the basic problems of this conflict.
D. Friedman
PREFACE
The origins of the Indian empiricist tradition and its development in
Early Buddhism are largely unknown to Western scholarship, despite
the fact that T. W. Rhys Davids at a very early date compared Buddhism
with Comtism 1 and Radhakrishnan went so far as to say that ‘Early Buddhism
was positivist in its outlook and confined its attention to what we perceive*. 2
However, modem Western thinkers, who have dipped into the literature
of Buddhism, have sometimes been struck by its analytical and positivist
turns of thought. H. H. Price, who was the Wykeham Professor of Logic
at the University of Oxford, remarked that ‘there are indeed some passages
in the early part of the Questions of King Milinda which have a very modern
ring, and might almost have been written in Cambridge in the 1920V. 3 Aldous
Huxley was of the opinion that Early Buddhism for the most part respected
the principle of verification and confined its statements to verifiable proposi-
tions. In his own words: ‘Among the early Buddhists, the metaphysical
theory (i.e. of Brahman of the Upanishads) was neither affirmed nor denied,
but simply ignored as being meaningless and unnecessary. Their concern
was with immediate experience, which, because of its consequences for life,
came to be known as “liberation” or “enlightenment”. The Buddha and his
disciples of the southern school seemed to have applied to the problems of
religion that “operational philosophy” which contemporary scientific thinkers
have begun to apply in the natural sciences . . . Buddha was not a consistent
operationalist; for he seems to have taken for granted, to have accepted as
something given and self-evident, a variant of the locally current theory of
metempsychosis. Where mysticism was concerned, however, his operation-
alism was complete. He would not make assertions about the nature of
ultimate reality because it did not seem to him that the corresponding set
of mystical operations would admit of a theological interpretation*. 4
Huxley’s qualification that ‘the Buddha was not a consistent operation-
alist’ may not have been made had he been aware of the epistemological
basis and the nature of the Buddha’s positivism and had he not been misled
by scholars to think that the Buddha had dogmatically accepted the doctrine
of rebirth from the prevalent tradition (v. Ch. VIII).
1 Origin and Growth of Religion , London, 1881, p. 31.
2 S. Radhakrishnan, Indian Philosophy , Vol. II, London, 1931, P- 47 2 *
3 ‘The Present Relations between Eastern and Western Philosophy’ in
The Hibbert Journal , Vol. LIII, April 1955, p* 22 9 *
4 Grey Eminence , London, 194 2 , PP* 47-8.
10
Preface
Our findings about the Early Buddhist theory of knowledge are based
primarily on the source material afforded by the Pali Canon, studied histori-
cally and philosophically in the light of the contemporary, earlier and late*
literary evidence bearing on the subject. The literary, linguistic, ideological^
sociological and historical evidence still points to the high antiquity and
authenticity of the Pali Canon 1 , although what we learn from it abouf
Early Buddhism may have to be supplemented and, perhaps, even modified
at times in the light of what we can glean from the other literary tradition^
of Buddhism 2 . We may refer here to the recent opinion of a student of
religion, Dr Robert H. Thouless, who says that ‘it seems more likely that
Hlnayana was Buddhism as originally taught and the Mahayana was a,
product of development and conventionalisation’ 3 .
The present work seeks to evaluate the thought of the Pali Canon from a
new point of view and in the light of new material. In it an attempt is made
to uncover the epistemological foundations of Pali Canonical thought. One
of the main problems of epistemology is that of the means whereby our
knowledge is derived. In this work the questions pertaining to the means of
knowledge known to, criticized in and accepted by the Buddhism of th^
Pali Canon are fully discussed. A comprehensive survey of the historical
background (Chs. I, II and III) was indispensable for this purpose partly
because this throws considerable light on the Buddhist theory of knowledge
and also because part of the material for the study of this background is to
be found in the Canon itself. i
Apart from the inquiry into the means of knowledge, a number of
questions relating to the problem of knowledge have been dealt with. Thu^
we have endeavoured to show the kind of logic adopted by the Buddhists
in contradistinction to that of the Jains (Ch. VII). While Wittgenstein’s 4
imaginary tribes played hypothetical language games showing the various
possibilities in the use of language, we find here actual instances in which
different systems of logic were employed in order to cope with certain
conceptual situations. We have also investigated the role of analysis, the
theories of meaning and truth and the problem of the limits of knowledge^
as they appear in the Canon.
1 v. Dialogues of the Buddha , Part i, Tr. T. W. Rhys Davids, SBB., Vol. II,
London, 1956, pp. ix-xx; cp. M. Winternitz, A History of Indian Literature ,
Tr. S. Ketkar and H. Kohn, University of Calcutta, 1933, p. 18. j
2 E. Lamotte grants a primitive core of remarkably uniform material common
to the Pali Nikayas and the Agamas, v. Histoire du houddhisme indien , Vol. I,
Louvain, 1958, p. 1 71. For a sceptical view, v. J. Brough, The Gandhari Dharma-
pada , London, 1962, pp. 31 ff.
3 “Christianity and Buddhism” in Milla wa-Milla , No. 2, November 1962, p. 3.
4 The author had the privilege of being admitted to Wittgenstein’s classes
held in his rooms at Whewell’s Court, Trinity College, Cambridge, in the years
1945-47*
II
Preface
The student of Indian philosophy should find here material pertaining to
the ‘prehistory’ of systematic Indian logic and epistemology and the origins
of the Indian empiricist tradition. A student of Greek thought may be able
to see in these pages some parallel developments to his own field, as well as
the differences. Of particular interest to the student of Western philosophy
would be Chapters VI and VII dealing with ‘Analysis and Meaning’ and
‘Logic and Truth’ respectively, the anticipation of two theorems of the
propositional calculus (Ch. VIII, sections 702-710), the theory of causation
(Ch. IX, sections 758-782), the empiricism of the Materialists (Ch. II) and
the Buddhists (Ch. IX).
I would express my gratitude to Dr D. L. Friedman for patiently reading
through this thesis and offering many valuable comments, criticisms and
suggestions. I am also grateful to him for introducing me to literature
pertaining to this subject which I had failed to consult at the time of writing
my first draft. My thanks are also due to Professor A. L. Basham, who
evinced an interest in this work and very kindly read through the whole of
Chapter III. I must also place on record my indebtedness to Professor
O. H. de A. Wijesekera of the University of Ceylon, from whom I learnt the
first lessons in research, and who encouraged me to work on this subject.
I am grateful to Mr D. J. Kalupahana, my pupil and colleague who was
kind enough to undertake the task of preparing the index and to my wife
and other colleagues and friends for assisting me with the proof-reading
and advice. I must also thank the University of Ceylon, which with the
generous assistance of the Asia Foundation defrayed a small portion of the
cost of this publication.
University Park, Peradeniya, Ceylon.
19 May 1963
K. N. JAYATILLEKE
CONTENTS
FOREWORD 5
PREFACE 9
ABBREVIATIONS 1 9
I. THE HISTORICAL BACKGROUND I VEDIC 21-68
Subsections ( Numbers here denote paragraphs and not
pages). Introductory observations (1-5). The germs
of scepticism regarding the possibility of knowledge
(6-10). The quest for simplicity (11) and order (12).
Scriptural authority as a source of knowledge (13).
The role of analogy in reasoning (14). The place of
doubt in the Brahmanas (15). The importance of
knowledge (16-17) and the growth of speculation
(i 8 - 2 o). Reason in the construction of metaphysical
theories (21-44). The origin and development of the
institution of the debate; the study of rhetoric and the
elements of reasoning (45-66). The means of know-
ledge recognized in the Early Upanisads — scriptural
authority, perception and reason (67-72). Mystical
intuition as a new way of knowing in the Middle and
Late Upanisads (73-76). The interest in one of the
main problems of epistemology at the close of the
Vedic period (77). The chronological and historical
relationship between the Upanisads and Buddhism
(78-81).
II. THE HISTORICAL BACKGROUND II
NON-VEDICI MATERIALISM 69-108
Subsections. The rational origins of Materialist
thought (82-87). The three main types of Materialist
schools according to their epistemological doctrines
(88-89). The early existence of the school or schools
in which perception only (90-91) or perception and
verifiable inference were the means of knowledge
recognized (92-94). The nihilist school which denied
the validity of all means of knowledge may have had
its origin as an early (pre-Buddhistic) rival branch of
1 4
Contents
the realist school or schools (95-118). No evidence
from the Nikayas for the existence of Arthasastra Lo-
kayata (Machiavellian Materialism) (119). The
Materialist criticism of the argument from authority
(1 20-1 21). The kind of reasoning employed by the
Materialist (1 22-1 29). The empiricist criticism of
metaphysical concepts (130-133). The use of the
modus tollendo tollens (1 34-1 35). Experiments devised
to test an hypothesis (136-139). The Materialist
critique of the claim to intuitive knowledge on the
part of the Mystics (140- 142).
III. THE HISTORICAL BACKGROUND III
NON-VEDIC II — SCEPTICS, AjTviKAS
AND JAINS IO9-168
Subsections. Relation to earlier traces of scepticism
(143- 1 44). The conflict of theories the main motive
for the birth of scepticism (145- 15 2). The rational
( I 53“ I 54) an d the moral grounds (155-156) of
scepticism. Was the logic of four alternatives (the
four-fold schema) an innovation of the Sceptics
(157)? The first three schools of Sceptics mentioned
in the Pali Nikayas (158-170). The similarity with
Pyrrhonean scepticism (1 71-174). The fourth school
of Sceptics (175-181). Two theories as to the origin
of the four-fold schema (182-190). The relation of
scepticism to the relativism (syadvada) of the Jains
(1 91-194). The rational origins of the main concepts
of the Ajlvikas (195-200). Their interest in causal
arguments (201-208). Their use of reason (209-212).
Their claims to mystical intuition (2 13-2 15) and
belief in tradition (216). The three-fold schema and
the three standpoints of the Trairasika Ajlvikas
(217-230). The main elements of the Jain theory of
knowledge (231-243).
IV. THE ATTITUDE TO AUTHORITY 169-204
Subsections. The pre-Buddhistic thinkers classifiable
mainly as traditionalists, rationalists and ‘experien-
tialists* on epistemological grounds (244-247). The
Buddha identifies himself with the last group (248-
250). The various forms of argument from authority
(251-260). The criticism of anussava (revelation,
Contents
tradition, report) as a valid means of knowledge
(261-283). The Buddhist and the Materialist criticisms
of the Vedic tradition compared and contrasted
(284-288). The Buddhist criticisms in the light of the
later claims made in regard to the validity of the Vedic
tradition (289-293). The criticism of parampara
(claims based on the continuity of a tradition) (294-
296), of hearsay (itikira) and non-revealed tradition
(297-303). The criticism of the principle of ‘con-
formity with scripture’ (pitakasampada) as a criterion
of knowledge (304-305). The criticism of testimony
(306-307). The reason for the rejection of authority
(308). The general condemnation of authority (309-
311). Authority within Buddhism (312).
V. THE ATTITUDE TO REASON 205-27 6
Subsections. Who were the takki (sophists, casuists,
debaters, reasoners?) and what was takka? (3 13-3 19).
The counterpart of the Greek sceptic-sophist not found
in India (320-326). Was Dighanakha a logical sceptic
or a nihilist Materialist? (327-336). The evidence for
the existence of vitandavadins or casuists (337-348).
The earliest dilemmas and the study of rhetoric on the
part of the intelligentsia (349-359). The debate in
Early Buddhist times and the reasoning used in it
(360-377). The debated theories, the schools to which
they belonged and the types of arguments used in the
construction, defence or criticism of them (378-415).
Reason in the construction of metaphysical theories;
four types of ‘reasoners’ and the meaning of takkika
(quibblers or reasoners?) (416-418). Takki as
rational metaphysicians; their theories (419-435).
The critique of reason; its inadequacy for knowledge
(436-442).
VI. ANALYSIS AND MEANING 277-332
Subsections. The r^le of analysis in the Pali Canon
(443). The critical and analytical outlook (444-451).
The analysis of questions into four types (452-458).
The categorical question (459-461). The question
to be answered after analysis (462-465) or with a
counter-question (466-468). The question which was
1 6
Contents
to be set aside; the meaningless question and its
modern parallels (469-481). Other senses of
analysis (482). The nature of definition (483-496) and
classification (497-500). The problem of universals
(5 01 — 502). The delimitation of terms and the so-
called ‘applied logic of conversion’ (503-515). The
four kinds of analysis (catupatisambhida) (516-519).
The origins of Indian linguistic philosophy (520-522).
The nature of verbal disputes (523-526). The im-
portance of linguistic convention and the linguistic
origin of metaphysics (527-535). The concept of the
‘meaningless statement’ (536-559).
VII. LOGIC AND TRUTH 333 “ 3 ^
Subsections. The questions discussed (560). The
theory that the four-fold logic violates the law of
Contradiction or represents laws of thought (561-
563). The four alternatives in the light of Aristotelian
logic (564). Recent theories suggested to explain this
logic unsatisfactory or inadequate (565-570). The four
alternatives and what they mean (571-583). The
necessity to distinguish four-cornered negation from
four-cornered rejection (584-585). The difference
between Jain syadvada (the seven forms of predica-
tion) and the Buddhist catuskoti (the logic of four
alternatives) (586-591). Robinson’s theory not
acceptable (592). The analysis of statements according
to truth-value, utility, pleasantness and unpleasant-
ness (593-595). Truth as correspondence with fact
(596). Coherence as the criterion of truth and the
senses of consistency (597-598). Is there a conception
of ‘partial truth’? (599-601). Truth and utility; the
pragmatic criterion (602-605). Truth and verification
(606). The conception of the ‘middle’ view as true
(607-609). The two kinds of discourse and the theory
of double truth (610-621).
VIII. AUTHORITY AND REASON WITHIN
BUDDHISM 369-41
Subsections. The nature of the questions discussed
(622-624). That the Buddha uncritically accepts certain
dogmas from the prevalent tradition is the view of many
scholars (625-626). That such dogmas are not part of
Contents
Buddhism is another view (627). Both views are
unsupported by the historical and textual evidence
(629-639). The authority of omniscience not claimed
for the statements of the Buddha until the very latest
stratum of the Pali Canon (640-649). Buddhism not a
revelation (650). The nature and role of belief or faith
(saddha) and its relation to knowledge; provisional
belief for the purposes of verification not incompatible
with rejection of authority (651-673). Another con-
ception of faith (674-678). The nature and role of
reason in the Pali Canon (679). Buddhism not a ration-
alism in the sense of being an a priori system of
metaphysics (680-685). Reason has definite but limited
function (686-687). Reason in the criticism of other
doctrines and the use of the modus tollendo tollens
(688-701). The anticipation of two theorems of the
propositional calculus — the rules of Implication and
Contraposition (702-710). Refutation (niggaha) in
the Kathavatthu (71 1).
IX. THE MEANS AND LIMITS OF KNOWLEDGE
Subsections. The emphasis on direct personal know-
ledge (7 1 2-721). The similarity and difference be-
tween the way of knowing of the Middle and Late
Upanisads and of Buddhism (722-725). The causes
and conditions for the emergence of extra-sensory
perception or of knowledge in general (726-732).
Extra-sensory perception only a means to an end
(733-734)- The possibility of error (735). Personal „
knowledge not subjective (736). The Materialist and
Buddhist emphasis on perception (737). The objec-
tivity of knowledge and the necessity for eliminating
subjective bias (738-740). The recognition of percep-
tion, normal and paranormal (741-743). The theory
of perception (normal) (744-749). The nature of para-
normal perception (750-756). Inductive inference
(757-759)* The theory of causation (760-782). Induc-
tive inferences based on empirical evidence and causal
theory (783-784). Verification (785). The notion of
extra-sensory perception does not belong to the
mythical or miraculous element (786). Inductive
inferences, valid and erroneous, based on the data of
paranormal perception (787-791). The Empiricism of
17
416-476
i8
Contents
Buddhism (792-796). The importance of verification
(797). The conception of ‘intellectual release* without
complete verification (798-799). The limits of
empiricism and knowledge (800-817).
APPENDIX I
477-481
CHRONOLOGICAL TABLE OF SCHOOLS
482
BIBLIOGRAPHY
483-494
INDEX
495-519
ABBREVIATIONS
A.
Anguttara Nikaya
AA-
Anguttara Nikaya
Atthakatha, i.e.
Manorathapuranl
A.A.S.
Ananda Asrama Series
abori.
Annals of the Bhan-
darkar Research In-
stitute
Ait. Ar.
Aitareya Aranyaka
Ait. Br.
Aitareya Brahmana
AO.
Acta Orientalia
Ar.
Aranyaka
Ard. Mag. Ardha Magadhi
AV.
Atharvaveda
BEFEO.
Bulletin de fficole
Frangaise d’Extreme-
Orient
BHS.
Buddhist Hybrid
Sanskrit
Br.
Brahmana
Brh.
Brhadaranyaka
Upanisad
Ch.
Chandogya Upanisad
Corny.
Commentary
CPD.
Critical Pali Diction-
ary
C.S.
Caraka Samhita
D.
Digha Nikaya
DA.
Digha Nikaya Attha-
katha, i.e. Sumangala-
vilasini
Dh.
Dhammapada
DhS.
r\nn\ T
Dhammasarigani
DPPN. Dictionary of Pali
Proper Names
Encyclopaedia of
Religion and Ethics
The Book of the
Gradual Sayings, v.
Anguttara Nikaya
HIP.
History of Indian
Philosophy
HOS.
Harvard Oriental
Series
IHQ.
Indian Historical
Quarterly
IP.
Indian Philosophy
It.
Itivuttaka
J.
Jataka
JA.
Journal Asiatique
JAOS.
Journal of the
American Oriental
Society
JPTS.
Journal of the Pali
Text Society
JRAS.
Journal of the Royal
Asiatic Society
Katha.
Katha Upanisad
Kaus.
Kausitakl Upanisad
Kaus. Br.
Kausltaki Brahmana
Khp.
Khuddakapatha
K.S.
The Book of the
Kindred Sayings, v.
Samyutta Nikaya
Kvu.
Kathavatthu
KvuA.
Kathavatthuppaka-
ranatthakatha
M.
Majjhima Nikaya
MA.
Majjhima Nikaya
Atthakatha, i.e.
Papancasudanl
Manu.
Manusmrti
Mait.
Maitri Upanisad
M.L.S.
Middle Length
Sayings, v. Majjhima
Nikaya
M.S.
Mimamsa Sutra
Mund.
Mundaka Upanisad
Nd.I.
Mahaniddesa
20
Abbreviations
Nd.IL
Cullaniddesa
N.B.
Nyaya Bhasya
N.S.
Nyaya Sutra (SBH.
Edition)
NS.
v. N.S.
O.U.P.
Oxford University
Press
p.
Pali
PAS.
Proceedings of the
Aristotelian Society
PIPC.
Proceedings of the
Indian Philosophical
Congress
Prasna.
Prasna Upanisad
PTS.
Pali Text Society
PU.
The Principal Upani-
sads, Ed. S.
Radhakrishnan
Pug. Pan.
Puggalapannatti
PvA.
Petavatthu Atthaka-
tha, i.e. Paramattha-
dipani
Ram.
Ramayana
RV.
Rgveda
S.
Sarny utta Nikaya
SA.
Sarny utta Nikaya
Atthakatha, i.e.
Saratthappakasinr
Sam.
Samavayanga Sutra
$B.
Satapatha Brah-
mana
SBB.
Sacred Books of the
Buddhists
SBE.
Sacred Books of the
East
SBH.
Sacred Books of th^
Hindus
S.Br.
v. SB.
SDS.
Sarvadarsanasam-
graha
S.K.
Sankhya Karika
Skr.
Sanskrit
S.P.S.
Sankhya Pravacana
Sutra
Sn.
Suttanipata
SnA.
Suttanipata Atthaka-
tha, i.e. Paramattha-
jotika II
Su.
Sutrakrtanga
Svet.
Svetasvatara Upani-
sad
Tait.
Taittirrya Upanisad
Tait. Br.
T aittiriy a B rahmana
Th. I
Theragatha
Th. II
Therrgatha
Toev.
Toevoegselen op’t
Woordenboek van
Childers, H. Kern
U.C.R.
University of Ceylon
Review
Ud.
Udana
UdA.
Udana Atthakatha,
i.e. ParamatthadrpanI
Vbh.
Vibhanga
VbhA.
Vibhanga Atthakatha,
i.e. Sammohavinodani
Vin.
Vinaya
V.S.
Vaisesika Sutras
Y.B.
Yogabhasya
Y.S.
Yogasutras
CHAPTER I
the HISTORICAL BACKGROUND I— VEDIC
(1) When we consider the history of thought in Greece, we find that
metaphysics first develops out of mythology 1 and it is only when
metaphysical speculation attains a certain maturity and results in the
formulation of a variety of theories that an interest is shown in the
problem of knowledge and epistemological questions are first mooted . 2
If we turn to the Indian context we can trace an analogous though by
no means an identical development.
(2) The intense speculative interest, which is so evident in the tenth
book (mandala) of the Rgveda persists as an undercurrent in the
period of the Brahmanas and issues forth in the theories and intuitions
of the Upanisads, whether we consider them a linear development in
Vedic thought or as being due to the impact of an external element,
Aryan or non-Aryan. Contemporaneous with the Middle 3 or Late
Upanisads or perhaps even later, we find the existence of schools of
thought which either broke away from the Vedic tradition or grew up
in isolation from and in opposition to it. The thought of this period
displays a wide variety of views. It was probably during this period,
which is coeval with or immediately prior to the rise of Jainism and
Buddhism that there arose the first questionings about the nature,
scope and validity of knowledge, resulting in the emergence of the
Sceptics (Ard. Mag. annania = Skr. ajnanikah; P. amaravikkhepika,
1 v. J. Burnet, Greek Philosophy — Thales to Plato , London, 1943, pp. 3 ff.
2 Note that Book I in Burnet’s work {op. cit .) dealing with the pre-Socratics is
entitled ‘the World’ (p. 15) and Book II from the Sophists onwards ‘Knowledge
and Conduct’ (p. 103).
3 We shall be using the term ‘Early Upanisads’ to denote the ‘Ancient Prose
Upanisads’, ‘Middle Upanisads’ for the ‘Metrical Upanisads’ and ‘Late Upanisads*
for the ‘Later Prose Upanisads’ in Deussen’s classification; v. The Philosophy of
the Upanishads, tr. Rev. A. S. Gedden, pp. 23-5.
22
Early Buddhist Theory of Knowledge
v. infra ^ 147, 158) and the epistemological and logical theories peculiar
to Materialism, Ajivikism, Jainism and Buddhism.
(3) When a metaphysical theory is fairly well developed, there is a
tendency to inquire into the grounds of its truth. Similarly, where there
are a number of conflicting theories about a particular problem, it
would be natural to ask which of them was true. Both these queries
lead to an investigation of the nature of truth and knowledge, which
may give rise to logical and epistemological doctrines. This seems to
have been the general pattern according to which interest was first
stimulated and advances made in the solution of the problem of
knowledge both in India as well as in Greece.
(4) In this survey of the Vedic period we shall be concerned with what
the Vedic (Brahmanic and Upanisadic) thinkers assumed or thought
were the means of knowledge and in the origin and nature of reasoning
as we find it in this literature. Both these questions shall be considered
in the light of their bearing on the thought of the Pali Canon.
(5) The Rgveda does not betray any awareness of the nature of
problems of knowledge. If we accept the naturalistic explanation, the
Rgvedic gods were probably fashioned on the analogy of ourselves
by positing wills behind the dynamic forces of nature but there is no
indication whatsoever that the thinkers were consciously employing
an argument from analogy. The mechanical and organic views of
creation 1 seem to have been similarly arrived at, although here the
analogies with some observable facts of nature are more evident at
least to the reader. The tendency on the part of the mind to look for
simpler explanations in place of the more complex is perhaps respon-
sible for the emergence of monotheistic and monistic tendencies 2 in
the last phase of Rgvedic thought.
(6) Interest is almost invariably focused on the outer world and it is
rarely that we meet with a thinker in an introspective mood though
we find an instance of a person who asks himself in a sceptical tone:
‘I do not clearly know what I am like here; bewildered and bound with
a mind, I wander’ (na vi janami yadivedam asmi ninyah 3 samnaddho
1 Radhakrishnan, Indian Philosophy , I, pp. 99 ff. 2 Op. cit pp. 89 ff.
3 We have followed the commentators in translating this term; Madhava takes
it to mean ‘concealed’ (antarhitah, v. fn. 2) and Sayana following him says, ‘the
term “ninyah” denotes what is concealed (and means here that) he is concealed,
i.e. has a bewildered mind (antarhitanamaitat antarhito mudhacittah).’
The Historical Background 23
manasa carami, RV. 1.164.37). Madhava, the pre-Sayana commentator
of the Vedas, 1 interprets this statement to mean ‘ “I do not clearly
know”, i.e. I do not understand whether I am this (world), being
beyond nature (j>rakfti-)\ thoroughly bound by a mind attracted by
objects and being concealed, I wander’. 2 While Madhava speaks in
terms of Sankhya philosophy (v. prakrti -), Sayana tries to give an
explanation, consistent with the philosophy of Advaita Vedanta,
when commenting on this statement he explains it as: ‘I do not know
that I am that which is the existent, the intelligent and the blissful
(saccidanando)’. 3 Both commentators take ‘idam’ to mean ‘the
universe’ and Sayana makes this quite explicit (yadivedam yadapidam
visvam asmi, loc. cit. ). These interpretations, based as they are on
later philosophies, are inadmissible for the Rgveda and we have tried
to render the sentence literally taking ‘idam’ in its adverbial sense to
mean ‘here’. 4 With the exception of Wilson who as usual follows
Sayana closely, 5 the translations of scholars bring out the sceptical
nature of the utterance. Griffith has: ‘What thing I truly am, I know
not clearly: mysterious, fettered in my mind I wander’ 6 and Geldner
renders it as; ‘Ich verstehe nicht was dem vergleichbar ist, was ich
bin. Ich wandele, heimlich mit dem Denken ausgeriistet.’ 7 Prasad
denies that this verse betrays any scepticism, 8 but translates it as:
‘I do not know whether I am like this, ignorant, prepared I go about’.
Here the translation of ‘samnaddho manasa as ‘prepared’ is in contra-
diction with ‘ignorant’, but even this translation which differs from
that of Sayana reflects a little of the sceptical mood of the original,
though Prasad prefers to call this ignorance rather than scepticism
(y. op. cit., pp. 24, 28).
1 v. Rgarthadlpika on Rgvedasamhita by Madhava, ed. L. Sarup, Vol. I
Lahore, 1939, Preface, p. 15. The pre-Madhava commentary (v. op. cit., p. 16)
of Skandhasvamin pertaining to this section is not available in print.
2 Na vi janami na vijanami yadapyaham idam asmi prakrteh parah visayaparena
manasa samyakbaddhah antarhitah carami, op. cit., Vol. II, Lahore, 1940,
pp. 326, 327.
3 Yo’yam saccidanando’ sti so aham asm! ti na vijanami, loc. cit.
4 v. Macdonell, A Vedic Grammar for Students, p. 210, Section 178, 2(a).
5 ‘I distinguish not if I am this all: for I go perplexed and bound in mind*,
H. H. Wilson, Rigveda Sanhita, Vol. II, Poona , 1925, p. 77.
6 R. T. H. Griffith, The Hymns of the Rgveda, Vol. I, Benares, 1889, p. 291.
7 K. F. Geldner, Der Rigveda, Uberset^t und Erlautert, Gottingen, 1923,
Vol. I, p. 21 1.
8 History of Indian Epistemology, 2nd ed. Delhi, 1958, pp. 20 ff.
24
Early Buddhist Theory of Knowledge
(7) The very few stanzas which strike a sceptical note deserve gg
special mention. 1 Here for the first time there is an expression of
doubt about the possibility of knowing certain things and a dim
awareness that some sort of evidence was necessary before we can
afford to make factual assertions. What evidence is there for the
existence of Indra unless someone has seen him? One stanza in a hymn
says, ‘One and another say, “there is no Indra”. Who hath beheld him?
Whom then shall we honour?’ 2 Who again can be sure about the fact
or nature of creation when no one has beheld the spectacle, ‘Who has
seen that the Boneless One bears the Bony, when he is first born?
Where is the breath, the blood and the soul of the earth? Who would
approach the wise man to ask this?’ (Ko dadarsa prathamam jayama-
nam, asthanvantam yad anastha vibharti, bhumya asur asrgatma kva
svit, ko vidvamsam upagat prastum etat. RV. 1.164.4). It will be noticed
that the author of this statement is the same as the person who felt 1
uncertain about himself (y. supra , 6). Now Prasad has questioned the
propriety of concluding that these questions suggest an attitude of j
scepticism and says that ‘either they are simply meant to introduce a
discussion, or at the most they indicate a confession of ignorance on the
part of the individual, who puts them’ (op. cit. y p. 24). Prasad is quite
right in pointing out that this hymn contains the subject matter of a
brahmodya (y. infra , 46) at which questions of this type were asked,
but if we examine the nature of this question itself, it will be seen that
it cannot be explained away as a confession of ignorance on the part of
the author. The question expresses the puzzlement of one who cannot
understand (in a philosophical sense) how a Boneless Being can pro-
duce a Bony offspring — an apparent contradiction. Quite apart from
the contradictory nature of this statement, what evidence was there to
believe in it. People doubted the existence of Indra because they could
not see him and the Nasadlya hymn poses the problem, ‘the gods are
posterior to this creation: if so, who knows whence it evolved?’
(arvag deva’sya visarjanenatha ko veda yata ababhuva, RV. 10.129.6).
Surely it is in this same sceptical spirit that it is asked, ko dadarsa . . . ?
(who has seen . . . ?), meaning thereby ‘who could have seen this
spectacle for us to know that it did really happen?’. The fact that the
author of this hymn doubts his own nature and confesses in all humility
1 For a collection of sceptical stanzas in the Rgveda, y. Radhakrishnan and
Moore, A Source Book of Indian Philosophy, pp. 34-36.
2 N’endra asti ti nema u tva aha, ka Im dadarsa kamabhi stavama, RV. 8.100.3
(=8.89.3, Griffith’s Translation).
The Historical Background 25
that he is ‘asking these questions out of immaturity and ignorance’
(pakah prcchami manasavijanan, 1.164.5) does not mark him off as an
ignoramus, any more than Socrates could be deemed to be ignorant
because he confessed that he knew nothing.
(8) The Scepticism of the Nasadlya hymn (RV. 10.129), which has
been unanimously accepted by scholars, 1 is denied by Prasad 2 follow-
ing Sayana. The hymn ends on a sceptical note according to the usually
accepted interpretation 3 and the question as to whether this is scepti-
cism or not depends on the interpretation given to this last stanza,
which reads:
Iyam visrstir yata ababhuva
Yadi va dadhe yadi va na
Yo’ syadhyaksah parame vyomant
So anga veda yadi va na veda, 10. 129.7.
Let us consider Sayana’ s explanation, especially since Prasad claims
that it agrees with his. 4 Commenting on the first two lines Sayana
says: ‘The Highest Self which is the material cause from which this
creation (i.e. this diverse creation variegated by way of its mountains,
rivers, oceans, etc.) has evolved, i.e. has arisen, is indeed the One who
either bears, i.e. sustains or does not sustain this; and thus, who else
indeed would be capable of sustaining it: if (anyone) sustains it, it
must be the Lord Himself, who would sustain it and no other’. 5
1 In addition to the translators we may mention Keith, Religion and Philosophy
of the Vedas , HOS., Vol. 32, p. 435; Ranade, A Constructive Survey of Upani -
shadic Philosophy y p. 3; Barua, History of Pre-Buddhistic Indian Philosophy ,
p. 16; Winternitz, Geschichte der indischen Litter ature y Vol. I, pp. 87, 88.
2 Op. cit. y pp. 25 ff.
3 Griffith translates,
‘He the first origin of this creation, whether he formed it all or did not form it,
Whose eye controls this world in highest heaven, he verily knows it, or perhaps
he knows not*.
Op. cit. y Vol. IV, p. 368.
4 Cp. Prasad, ‘. . . compare Sayana’s interpretation of the verse which agrees
with that of mine* (op. cit. y p. 27, fn. 1).
5 Yata upadanabhutatparamatmana iyam visrstir vividha girinadlsamudradi-
rupena vicitra srstir ababhuvajata so’pi kila yadi va dadhe dharayati yadi va na
dharayati evan ca ko namanyo dhartum saknuyat yadi dharayedlsvara eva dhara-
yennanya iti, op. cit. y Vol. 6, p. 410.
2 6
Early Buddhist Theory of Knowledge
In the light of this comment Say ana’s translation of the first two lines
would be as follows:
‘That (Highest Self) from whom this creation arose,
Either sustains it or does not sustain it/
The comment on the next two lines is as follows: ‘The Highest Lord
who is such a person does indeed (“indeed” in the sense of “as is well
known”) know, i.e. understands: if he does not know, i.e. does not
understand, who else indeed would know: the sense is that the omni-
scient Lord alone would know about this creation and no other’. 1
This implies the following translation of the last two lines:
‘He who is the Lord in the highest heaven;
He verily knows, if (anyone else) does not know’.
Sayana’s translation of the first two lines is unobjectionable from the
point of view of grammar and syntax though his contention is that
the second line means, ‘the (Highest Self) alone sustains it and no one
else’, which is not apparent from even his literal rendering of the
sentence. But his translation of the fourth line is clearly at variance
with grammar, for he alters the subject of the verb ‘veda’ of the second
sentence from ‘sah’ to ‘ka anyah’ (understood) without any support
from the original. If we have misunderstood Sayana in attributing to
him such an unwarranted periphrasis, he is at least translating this line
as ‘he verily knows or does not know’ and interpreting it to mean ‘it
is only he who knows and no one else’, although it is evident that this
sentence cannot mean this either in a literal or a figurative sense. Now
Prasad, speaking of the second and fourth lines of this verse, observes,
‘These two clauses do not express doubt or ignorance, but mean and
that quite in accordance with idiom that it is only He who bore it, and
no body else and it is He who knows it and no body else respectively’
{op. cit ., p. 27, fn. 1), but he does not translate the verse or explain
how the only possible literal translation can idiomatically mean what
he and Sayana try to make it mean. It is evident that Sayana is really
trying to explain the verse away rather than to interpret what it
strictly meant since he could not countenance the claim that the sacred
scriptures contained statements sceptical about the knowledge or
power of the deity but we cannot be led by these considerations.
1 Idr§o yah paramesvarah so anga angeti prasiddhau so’pi nama veda janati,
yadi va na veda na janati ko nama anyo janlyat sarvajna is vara eva tarn sjrstim
janiyat nanya ityarthah, op. cit., Vol. 6, p. 471.
27
The Historical Background
(9) The scepticism of this hymn is interesting not merely because it
ends on a sceptical note but because it does so, after taking account of
almost every possibility with regard to the problem of the origin of the
world. If we consider the problem in the abstract at the purely philo-
sophical level, we can say that we can either know the answer to this
question or we cannot. 1 If we say we can know the answer, we can
suggest either that the world was created or it was not. If we say that
the world was created, we can say that it was created out of Being or
Non-Being. If we say it is out of Being we may say that it is created
either out of matter or out of spirit. An analysis of the hymn reveals
that all these suggestions are implicit in it, although it offers its own
theory tentatively by trying to synthesize the concepts of Being and
Non-Being, of matter and of spirit. We may diagrammatically repre-
sent the alternatives considered in the hymn in the light of its state-
ments as follows:
(yadi va) veda, 7 yadi va
either he knows na veda, 7
| or he does
| | not know
yadi va dadhe, 7 yadi va na, 7
either he formed or he did not
(created) it form it
sat, 1
1
asat, 1
being
non-being
na asat na u sat aslt, 1
there was neither non-being nor being
ambhah kim aslt, anidavatam, 2
gahanam gabhiram, 1 he breathed without breath
Was it water, unfathomable (spirit?)
and deep ? (i.e. was it matter ?)
1 We are leaving out the possibility that the question is meaningless in the
Positivist’s sense of the term.
28
Early Buddhist Theory of Knowledge
(10) Despite the (to us) dogmatic presentation of his own theory;
the sceptical conclusion, after taking account of almost all the possibly
answers to this question, nearly approaches scepticism with regard to,
the possibility of knowledge in respect of the problem of the origin,
of the world. This scepticism which is based on the consideration that
‘since the gods came after the creation (srsti-, lit . emission, emanation),
no one knows how the world began’ (arvag devasya visarjanenatha
ko veda yata ababhuva, 6) because no one was there to behold the
spectacle (cp. ko. dadarsa . . . ?, supra , 7), is soon forgotten in the
orthodox tradition. However, it leaves its mark in (or is rediscovered
by?) Buddhism, where Brahma, reputed to be ‘the creator’, (sajita,
D. I.i 8 < srj-i-ta(s) = Skr. srasta: cp. katta, nimmata, loc. cit.) is
said to be ignorant of his own origin (loc. cit ., v. infra , 645). More-
over, it is said that ‘it is not possible to conceive of the beginning of
the world: a first cause (lit. prior end) cannot be known’. 1
(11) The desire for simple and single principles of explanation, which
seems to have led to the emergence of the monotheistic and monistic
concepts in the final stratum of Rgvedic thought seems to have worked
its way into the undercurrent of speculation found in the Atharvaveda
and the Brahmanas, where the few philosophical hymns try to com-
prehend the entirety of the universe under some single concept such
as Time (Kala), 2 Eros (Kama), 3 Creative Power (Brahman), 4 Life
Principle (Prana) 5 or an Ontological Framework (Skambha). 6
(12) The same tendency is found in the Brahmanas. For although here
thought is subservient to the practical ends of the sacrifice, the uni-
verse, conceived on the analogy of the sacrifice, is regarded as a unity.
The unity is, however, not evident on the surface and is made up of
hidden bonds and relations lying concealed beneath the plural uni-
verse. 7 ‘What is evident ( pratyaksam ) to men is concealed (paroksam)
to the gods, and what is concealed to men is evident to the gods’ (yad
vai manusyanam pratyaksam tad devanam paroksam atha yan manus-
1 Anamataggo’yam . . . samsaro pubbakoti na pannayati, S. II. 178; the
translation of ‘anamataggo* is not without its problems (y.v. PTS. Dictionary)
but it is the etymology that is doubtful and not the sense, which is clear from the
contexts of its use in the Samyutta Nikaya; cp. CPD. Dictionary, s.v. anamatagga.
2 AV. 19.54, 53. 3 AV. 9.2. 4 AV. 19.42. 5 AV. 1 1.4. 6 AV. 10.7.
7 This is how one thing becomes the mystic name (nama), the mystic symbol
(rupa), the mystic body (tanu) and the mystic bond (bandhu) of another. For
examples and references to Brahmanic literature see Ranade and Belvalkar,
History of Indian Philosophy , II, pp. 62-5.
The Historical Background 29
yanam paroksam tad devanam pratyaksam, Tandyamahabrahmana,
22.10.3). In the Rgveda there was a primitive conception of causality
underlying the idea of rta which seems to have denoted the ‘course of
things 5 or the observable physical order of the world before it acquired
a moral and theological connotation. But in the Brahmanas, which
value ‘what lies beyond the sphere of the senses 5 (paroksa-), 1 the
conception of a causal order gives way to that of a magical order. 2
(13) It is in the Brahmanas that we find developed what became for
orthodoxy the supreme source of knowledge — the revealed scriptural
text. As Ranade and Belvalkar say, ‘the Brahmanas came to invest the
mantras with the character of divine revelation. They are at times
spoken of as eternally self-subsistent and coeval with God-head — if
not actually prior to Him. At other times — and especially in the newer
Brahmana texts ( underlining mine) — they are described as creations of
Prajapati, the head of the whole pantheon 5 . 3 The hymns are said to be
seen, learned or found generally by some special insight on the part
of the seers and not made or composed by them. 4 5
(14) The reasoning in the Brahmanas is analogical and centres
round the symbolism of the sacrifice. The analogies are remote. A
fanciful etymology, a myth, legend or a vague similarity is sufficient
to establish a connection between two things. 5 An explanation to be
satisfactory has to be made in terms of a sacrificial analogy. Examples
of typically Brahmanic reasoning may be found at SB. 1 1.4.1. 12— 15,
which describes the debate between Uddalaka Aruni and Svaidayana
Gautama. The following are two arguments found there: (1) Atha
jWapuro’nuvakyaka prayaja bhavanti, tasmad imah praja’dantaka
jayante, i.e. and since the fore-offerings are without preliminary
formulae, therefore creatures are born here without teeth, 1 1.4.1. 12,
(2) atha jya^ajyahavisah prayaja bhavanti tasmat kumarasya retah
siktanna sambhavaty udakamivaiva bhavaty udakam iva hyajyam, i.e.
1 The expression paroksapriya hi devah, i.e. the gods love what is not evident,
is common in the Brahmanas; v. op. cit ., p. 63.
2 Oldenberg, Religion des Veda , pp. 315 fF., 321 fF.; Keith, HOS., Vol. 32,
pp. 379 fF.; Frauwallner, Geschichte der indischen Philosophic, Band I, pp. 41 fF.
References to magic are found in the Rgveda and the Yajurveda as well, v. H.
Liiders, Varuna, Band II, Gottingen 1959? PP- 5°9~45> Die magische Kraft des
Rta in den vedischen Mythen.
3 Op. cit., p. 56.
4 Keith, HOS., Vol. 32, p. 482.
5 v. Ranade and Belvalkar, op. cit., p. 63.
30 Early Buddhist Theory of Knowledge
and since the fore-offerings have ghee for their offering material, a
boy’s seed is not productive but is like water since ghee is like water.
This sounds utter balderdash, but just as much as a biological reason
would be given today as to why a ‘boy’s seed is not productive’,
nothing short of a ‘sacrificial’ reason would have satisfied a Brah-
manic thinker. Anything to be understood had to be explained on a
sacrificial analogy and discovering these analogies (bandhuta) was as
much an art as the reasoning itself. The reasoning in the above argu-
ment may be exhibited as follows since much is taken for granted in
the arguments :
1. Ghee fore-offerings are not productive (since ghee is like water — ■
v. udakam iva hyajyam — and water is not productive in a biological
sense)
2. Ghee fore-offerings are like the boy’s seed (since both are at the
beginning, ghee fore-offerings at the beginning of the sacrifice and
the boy at the beginning of life)
3. Therefore, the boy’s seed is not productive.
The form of this argument from analogy would be as follows:
1. A has the characteristic p
2. A is like B
3. Therefore, B has the characteristic p.
The remotest connection, natural or magical, between two things is
sufficient for the Brahmanas to draw the analogy that ( A is like B ’ on
the basis of which inferences are made.
(15) There is rarely any admission of the need for or possibility of
doubt and investigation (mlmamsa) 1 is always carried out with the
conviction that the correct interpretation of the revealed texts opens
the door to all knowledge but there is mention of vicikitsa, or the
doubt that premotes inquiry. 2 Vicikitsa or ‘doubt’ is in fact one of the
recognized states of mind. The Sathapatha Brahmana says, ‘wish,
conception, doubt, faith, lack of faith, determination, lack of deter-
mination, shame, thought, fear — all this is mind’ (kamah samkalpo
vicikitsa sraddhasraddha dhrtiradhrtirhrlrddhirbbhir ity etat sarvam
mana eva . . ., 14.4.3.9); thus, ‘Pratardana ... questions about his
doubt 9 (Pratardanah vicikitsatn papraccha, Kaus. Br. 26.5). Of specific
1 v. Keith, Religion and Philosophy of the Vedas, HOS., Vol. 32, p. 483.
2 Ibid,
The Historical Background 31
doubts there is mention of the doubt regarding man’s survival in the
next world. 1 Some of the doubts raised appear to be genuinely philo-
sophical as when it is asked how the raw and red cow can yield hot
white milk and how the boneless semen can produce creatures with
bones 2 but the answers given in terms of sacrificial analogies are,
needless to say, hardly satisfactory.
(16) In the Aranyakas knowledge comes to be greatly valued; where
the knowledge of the symbolism of the ritual was what really mattered,
the performance of the ritual itself may be dispensed with. The
knowledge is not prized for its own sake but is invariably considered
to be a means to an end. The usual formula would be that knowledge
of X gives Y y where X may stand for some item of empirical or meta-
physical knowledge and Y for anything from material gain to spiritual
reward. Thus we have the following statement in the Aitareya Aran-
yaka: ‘The Hotr mounts the swing, the Udgatr the seat made of
Udumbara wood. The swing is masculine and the seat feminine and
they form a union. Thus he makes a union at the beginning of the
uktha in order to get offspring. He who knows this gets offspring and
cattle . 93 The growing importance attached to knowledge, however, is
such that everything had to be subordinated and one’s entire life
geared to this end by the time of the Upanisads.
(17) In the Upanisads there is a continuation of the theme that know-
ledge gives some kind of reward. He who knows (veda, Brh. 1.3.7.),
for instance, the superiority of the breathing principle (prana-) over
the sensory and motor organs becomes his true self and the enemy
who hates him is crushed (loc. cit.). There is, however, no explanation
as to why this knowledge should give this specified result. One of the
rewards is immortality, conceived in the earliest Upanisads as the
escape from a second death (punar-mrtyu-): ‘He who knows that air
is the totality of all individuals conquers repeated death.’ 4
(18) This great importance attached to knowledge paves the way for
thinkers to speculate on the nature of reality and the problems of life
without being hampered by the limitations of the Vedic tradition. The
influence of the earlier mythology and theology is no doubt felt, but
1 v. Keith, Religion and Philosophy of the Vedas, HOS., Vol. 32, p. 483.
2 Ranade and Belvalkar, op. cit., p. 73.
3 Prenkham hotadhirohaty audumbarfm asandim udgata, vysa vai prenkho
yosa sandl, tan mithunam eva, tad ukthamukhe karoti prajatyai. Prajayate prajaya
pasubhirycz evatn veda, Aitareya Aranyaka, 1.2.4.10,11.
4 Vayuh samastih apa punar mrtyum jayati ya evam veda, Brh. 3.3.2.
32 Early Buddhist Theory of Knowledge
the thinkers bring a fresh mind to bear on the problems they seek to
solve.
(19) The tendency especially on the part of Indian scholars to regard
the Upanisads as presenting a single systematic and coherent philo-
sophy on the basis of the interpretations and expositions of either
Sankara, Ramanuja, Madhva or others has much obscured the in-
dependence and originality of the speculations of many thinkers of the
Upanisadic tradition. Such an attitude fails to take account of the fact
that although the Upanisadic thinkers owed allegiance to the Vedic
tradition, they were free to theorize on matters and topics that fell
outside the scope of that tradition. They not only belonged to separate
schools but were often separated and isolated geographically. Besides,
many generations would have lapsed between one outstanding teacher
and another. We find evidence of conflicting theories, of the criticism
and replacement of one theory by another and the influence of earlier
views on later thinkers, who build on them. All this would not have
been possible if there was a single uniform philosophy called the
vedanta , which is unfolded on every page of the Upanisadic texts.
(20) If we examine the Upanisadic texts, considering the theories
found in separate sections or ideological units separately, we would
find that the thinkers of the Upanisads can be classified into two dif-
ferent categories. Firstly, there are those who found and propound
their views by indulging in metaphysical speculation and rational
argument not without a basis in experience, despite the earlier myth-
ology weighing heavily on their minds. Secondly, there are those
who profess their theories as an expression and interpretation of what
they claim to have themselves experienced by the practice of yoga,
although in the form in which they are presented they are dressed in a
good deal of metaphysical clothing. The former set of thinkers are
usually met with in the Early Upanisads while the latter are generally
represented in the Middle and Late Upanisads, but no absolute division
is possible since the rational metaphysicians are found in some of the
Middle and Late Upanisads (e.g. Prasna) while references to yoga
philosophy and practice are not entirely absent in the Early Upanisads.
(21) The difference between these two types of thinkers, namely the
rational metaphysicians who found their theories on a priori and
empirical reasoning and the contemplative intuitionists who claim to
acquire special insights into the nature of reality by following certain
techniques of mind control and culture, would be clearer if we take
The Historical Background 33
samples of their theories and statements and examine the epistemo-
logical bases of their thought. Let us first consider some of the meta-
physical theories and see on what kind of thinking and reasoning they
are based.
(22) Let us take the philosophy of Uddalaka. His philosophy has
been treated separately 1 by both Barua 2 as well as Ruben. 3 Ruben
examines his ontology and calls it a ‘hylozoistische Monismus’, 4 and
refers to Uddalaka as ‘der alteste Materialist’ 5 and as a ‘Realist’. 6 Barua
starts with his theory of knowledge and is inclined to call him an
Empiricist. Since we are interested only in this aspect of his thought
we may examine Barua’s appraisal of it. He says; ‘ . . . Uddalaka
propounded an empirical theory of knowledge. Henceforth let no
one speak, he asserts, of anything but that which is heard, perceived
or cognized. He seems repeatedly to point out: — The only right
method of scientific investigation into the nature of reality is that of
inference by way of induction’ {op, cit ., p. 138). Later Barua seems to
qualify Uddalaka’s claims to be a pure empiricist: ‘According to his
own showing the senses furnish us with sufficient indications from
which the knowing mind can easily infer the nature and relations of
things in themselves’. 7
(23) Now the statement that Barua attributes to Uddalaka, namely
‘henceforth let no one speak of anything but that which is heard,
perceived or cognized’ does not seem to bear the meaning that Barua
gives to it, when we consider its literal translation in the context in
which it appears. Uddalaka propounds the elements of his philosophy
and then says, ‘Verily, it was just this that the great householders and
great students of sacred knowledge knew when they said of old, “no
one will now mention to us what we have not heard, what we have not
perceived, what we have not thought” ’ (etaddha sma vai tad vidvamsa
ahuh purve mahasala mahasrotriyah na no ’dya kascana asrutam,
amatam, avijnatam udaharisyatl ti, Ch. 6.4.5). This statement does not
1 1 .e. separately from the rest of Upanisadic thought.
2 A History of Pre-Buddhistic Indian Philosophy , pp. 124—42.
3 Die Philosophen der Upanishaden , Bern, 1947? pp. 156—77; cp. Geschichte
der Indischen Philosophic, Einfiihrung in die Indienkunde, Berlin, 1954, pp. 25—7,
81-94. 4 Die Philosophen der Upanishaden, p. 166.
5 Geschichte der Indischen Philosophic, p. 81.
6 Die Philosophen der Upanishaden, p. 156.
7 Op. cit., p. 140; v. his subtitle, ‘Uddalaka neither trusts nor yet distrusts the
evidence of the senses".
B
34 Early Buddhist Theory of Knowledge
seem to convey anything more than that anyone who has grasped the
theory set forth by Uddalaka knows all there is to be known and
therefore no one can teach him, i.e. make him hear, think or under-
stand anything new. It is a dogmatic assertion claiming finality for his
philosophy. It does not mean, ‘henceforth let no one speak . . .’ but
‘today (adya) no one (na kascana) will speak (udaharisyati) . . .’ and
no epistemological significance can be attached to it.
(24) Whether Uddalaka is an empiricist or not can only be deter-
mined by examining the epistemic origin of his theory and when we do
so, he appears to be basically a rationalist, who makes considerable
use of empirical premises to illustrate his theory and serve as a basis
for his metaphysical insights.
(25) Uddalaka for the first time in the history of Indian thought
expressly suggests a proof of the reality of Being (sat) instead of
merely assuming it, when he asserts, ‘some say that . . . from non-
Being Being was produced. But, verily, my dear, whence could this
be? ... how could Being be produced from Non-Being’? (taddhaika
ahuh . . . asatah saj jayata. Kutas tu khalu, saumya, evam syat . . .
katham asatah saj jayeta, Ch. 6.2.1, 2).
(26) Having proved the reality of being by pure reasoning, Uddalaka
had to explain how the world could have a plurality of things, if
Being (sat) alone were real. If Being was the only reality, plurality
is mere appearance. The different shapes and names that things have,
cannot be real, for Being is the one and only substance that exists. This
is illustrated by some empirical examples. When we see an object of
clay, we know that its shape and name can be changed but its sub-
stance cannot be changed for ‘the modification is merely a verbal
distinction, a name, the reality is just clay’ (vacarambhanam vikaro
namadheyam mrttikety eva satyam, 6. 1.4).
(^27) It is not only the present plurality that has to be accounted for
but the origin of this plurality. Here Uddalaka uncritically accepts the
earlier mythological notions and says that Being wishes to multiply and
procreate and produces heat (6.2.3). Heat (tejas) produces water
(apas) and water food (annam) (6.2.3, 4). Empirical evidence is adduced
in favour of this causal sequence, 1 where it is pointed out that when we
1 Note that heat, water and food are in the relationship of root and sprout;
‘. . . with food for a sprout look for water as the root. With water as a sprout
look for heat as the root’. . . annena sungenapo mulam anvicceha, abdhih . . .
sungena tejo mulam anviccha, 6.8.4).
The Historical Background 35
are hot, water (i.e. tears or perspiration) is produced (6.2.3) an< 3 that
water in the form of rain produces food (6.2.4).
(28) Original being conceived as an active and animating principle
now produces everything out of its three emergent products, heat,
water and food (6.3.2). So all things are made out of these three
constituents, which are called the three colours or forms (trim rupani,
6.4.1) because heat is supposed to be red in colour, water white and
food or earth dark (6.4.2), again presumably on empirical grounds.
Man himself is therefore a product of these three forms. But how can
the mind or voice be explained as a by-product of these three primary
products? Uddalaka here speaks of the coarse (sthavistah), medium
(madhyamah) and fine (anisthah) constituents of these products (6.5)
and argues that the finest essence of food moves upward on the analogy
of butter moving upward when milk is churned (6.6.1) and becomes
the mind. Food becoming mind is again proved empirically on the
grounds that if you refrain taking food while drinking only water
for some time you forget what is in your mind 1 (6.7. 1-3). Physio-
logical processes like hunger and thirst are likewise explained as being
due to the interaction of the primary products. You are hungry
(asana = as-a-nd) because water leads off (rcczyanti) the food eaten
(aiitam) (6.8.3). This argument is based on fanciful etymology and
is reminiscent of the type of reasoning found in the Brahmanas ( v .
supra , 14).
(29) We are therefore produced from Being though we do not know
it (6.9.2). We also reach Being at death for in the process of dying
there is a reversal of the process of production, the mind (the product
of food) goes into breath (prana, the product of water) and breath in
turn to heat and heat into the highest deity, at which point he knows
not (6.15.1-2), for he cannot recognize the people who gather round
him ( loc . cit,). What is empirically urged by observations made on the
dying man is also rationally arrived at where it is suggested that the
substance of our personality, constituting the mind, breath and voice
are so completely mixed up on reaching homogeneous Being that
there would be no separate mind to know that ‘I am this one’ (iyam
aham asmi) or ‘I am that one’ (iyam aham asmi) so that we know not
that we have reached Being (6.9.1, 6.10.1).
1 This is one of the earliest experiments performed with the idea of testing a
theory; cp. Ruben, ‘Er Hess seinen Sohn Svetaketu fiinfzehn Tage nicht fur
einen Ritus, sondern als materialistiches Experiment fasten. . . Geschichte der
Indischen Philosophic, p. 87.
3 6 Early Buddhist Theory of Knowledge
(30) Uddalaka ought to have said in accordance with his metaphysics
that all beings are produced from Being at birth without knowing i|
and subsequently reach Being at death without knowing it. But th«
necessity of assigning a special reward to knowledge as was th i
prevalent fashion of the times (y, supra , 16) probably makes him sayj
inconsistently that those who do not know his theory become tigers^
lions, etc., in the next life (6.9.3) while those who know his theory
(which is assumed to be the truth) reach Being and are merged in it
never to return, their belief in truth ensuring this just as much as he
who speaks the truth is saved in a trial by ordeal (6.16.1-3).
(31) At the end of the lecture, Uddalaka’s son understands (vijajnau,
6.16.3) the theory that was propounded. There is no suggestion or
implication whatsoever that the theory was to be comprehended by
the practice of special techniques such as yoga. It was merely this
rational understanding that was considered necessary for ensuring the
goal of reaching Being at death. The theory itself is clearly a product
of reason and speculation, as we have shown. The reasoning is partly
a priori and partly empirical, although the a priori reasoning is not
consistent and the empirical conclusions not warranted by the evi-
dence adduced. It is also necessary to note the impact of the earlier
mythology and the Brahmanic ‘reasoning’ on the thought of Uddalaka.
When the Buddha says that there was a class of brahmins who pro-
pounded theories on the basis of reason and speculation (y, infra,
420 ff.), was he thinking of thinkers of the type of Uddalaka, whose
name and the central theme of whose philosophy was known to the
Jatakas? 1
(32) Let us now consider another metaphysical theory which is a
product of rational and empirical reasoning and which is attributed to
Prajapati in the section 8.7-12 of the Chandogya Upanisad. Its interest
for Buddhism lies in the fact that it contains a kind of reasoning, which
is taken to its logical conclusion in Buddhism (y. infra , 39).
(33) The inquiry begins with the assumption that there is a soul
(atman) which has the characteristics, inter alia , of being free from
death (vimrtyuh), free from sorrow (visokah) and having real thoughts
(satyasamkalpah) (8.7.1). The problem is to locate this soul in one’s
personality.
(34) The first suggestion is that the soul may be the physical per-
sonality, which is seen reflected in a pan of water (8.8.1). But this
1 v. Barua, op, cit.> pp. 125-7.
The Historical Background 37
physical personality, it is argued on empirical grounds, is subject to
death; ‘it perishes in the wake of the perishing of this body’ (asyaiva
£arfrasya nasam anv esa na£yati, 8.9.1). So this will not do (naham atra
bhogyam pasyami loc . cit .). This conclusion embodies a rational
argument of the following sort: if X (the body, in this instance) has
the characteristic not -p (i.e. not free from death), then it cannot be
an instance of A (the atman), which necessarily has (by definition, i.e.
assumption) the characteristic p .
(35) The next suggestion is that the soul may be identified with the
self in the dream-state (8.10.1). This escapes the objection against the
previous suggestion for the dream-self ‘is not slain when (the body)
is slain’ (na vadhenasya hanyate, 8.10.2). The logic of the reasoning
is as follows: here is an instance of Y (the dream-self), which has the
characteristic p (free from death) and which therefore may be an
instance of A which must have the characteristic p . But this suggestion
too is turned down (r. naham atra bhogyam pasyami, 8.10.2) for not
only must Y have the characteristic p to be an instance of A , but it
must also have the characteristic q (free from sorrow, visokah). But
it is seen on empirical grounds that Y does not have the characteristic
q: ‘he comes to experience as it were what is unpleasant, he even weeps
as it were’ (apriyavetteva bhavati, api rodativa, 8.10.2). Therefore,
Y cannot be an instance of A.
(36) The next suggestion is that the soul may be identified with the
state of deep sleep (8.1 1.1). This, it may be observed, escapes the
objections against the two previous suggestions. The reasoning may be
represented as follows: here is an instance of Z (the state of deep sleep),
which has the characteristic p (free from death; v. etad amrtam, this
is immortal, 8.1 1.1) and also the characteristic q (free from sorrow
or grief; v. samastah samprasannah . . . etad abhayam, being composed
and serene . . . this is free from fear, 8.1 1.1) and which therefore may
be an instance of A , which must have the characteristics p and q .
But this suggestion too is turned down, (v. naham atra bhogyam
pasyami, 8.1 1.1) for not only must Z have the characteristics p and q
to be an instance of A , but it must also have the characteristic r (real
thoughts, satya-samkalpah). But it is seen on empirical grounds (i.e.
by introspection) that Z does not have the characteristic r; ‘in truth
he does not know himself with the thought “I am he” nor indeed the
things here — he becomes one who has gone to destruction’ (naha
khalv ayam evam sampraty atmanam janati, ayam aham asmi ti, no
Early Buddhist Theory of Knowledge
38
evemani bhutani, vinasam evapito bhavati, 8.1 1.1). The argument u i
to this point is that the atman cannot be identified with any aspect ojj
the personality, physical or psychological.
(37) Not satisfied with this purely negative conclusion empirically
arrived at, the metaphysical assumption is then made that the atman
(not identifiable with its states) must be an unobservable entity (3
pure ego 1 ) within the personality with all its aspects. ‘The body is
mortal but is the support of the immortal bodiless atman’ (martyarn
. . . idam sariram . . . tadasyamrtasyasanrasyatmano’dhisthanamj
8.12.1). Dogmatic utterances have now taken the place of rational
arguments and Prajapati now indulges in his own quota of Brahmanic(
‘reasoning’. He sees an analogy between ‘air, clouds, lightning and
thunder’ on the one hand and the atman on the other, since both ard
bodiless (8.12.2) and argues that since the air, etc., ‘reach the highest
light and appear each with its own form’ (param jyotir upasampadya
svena svena rupenabhinispadyante, 8.12.2), the atman too, similarly
(evam, 8.11.2), ‘rises up from this body and reaches the highest light
and appears in its own form’ (asmac charirat samutthaya param jyotir
upasampadya svena rupenabhinispadyate, 8.11.3).
(38) Incidentally, it is significant that elsewhere in the Upanisads the
soul (atman) is identified with the dream-state (Brh. 4.3.9, IO > Ch.
8.3.2) and with the state of deep sleep (Brh. 2.1.16-20) and therefore
this theory constitutes a criticism of these earlier theories ( v . supra , 19).
(39) It will be seen that the Buddha in advocating the theory of anatta
follows a pattern of argument very similar to that used by Prajapati
here in the earlier part of his theory. The Buddha like Prajapati takes
various aspects of the personality and shows that none of them can be
identified with the atman, since they do not have the characteristics of
the atman. The following is a sample of such an argument:
Buddha — What think you? Is the physical personality permanent or
impermanent? (Tam kim mannasi? . . . rupam niccam va aniccam
va ti, M.I. 232)
Saccaka — It is impermanent (aniccam . . ., loc . cit.)
Buddha — Is what is impermanent sorrowful or happy? (Yam pana-
niccam dukkham va tarn sukham va ti, loc . cit.)
1 On the use of this term, v. Broad, The Mind and its Place in Nature ,
pp. 214 ff., 278 ff., 558 ff., and 603 ff.
39
The Historical Background
Saccaka — Sorrowful (dukkham . . ., loc . cit .)
Buddha — Of what is impermanent, sorrowful and liable to change, is
it proper to regard it as ‘this is mine, this I am, this is my soul’?
(Yam pananiccam dukkham viparinamadhammam, kalian nu tarn
samanupassitum: etam mama, eso’ ham asmi , eso me atta ti, M. I.233)
Saccaka — It is not (No h’idam . . ., loc . cit.).
This same argument is now repeated for other aspects of the per-
sonality such as feeling (vedana), ideation (sanna), etc. (loc. cit.). One
may compare the expression used in the Pali passage, eso aham asmi
to indicate the identification, with the corresponding expression, ayam
aham asmi (Ch. 8.1 1.1) used for the same purpose, in the Upanisad.
The main difference in the attitude of Prajapati and the Buddha is that
the former assumes the existence of an atman and on failing to identify
it with any of the states of the personality, continues to assume that
it must exist within it and is not satisfied with the results of the purely
empirical investigation, while the latter as an Empiricist makes use of
the definition of the concept of the atman without assuming its exis-
tence (or non-existence) and is satisfied with the empirical investiga-
tion which shows that no such atman exists because there is no evidence
for its existence. Was it those who reasoned in this manner basing their
reasoning on definitions (laksana-), who were called lakkhana-vada
(Nd. I.294; v. infra , 367)?
(40) Whether these Upanisadic theories were known to Buddhism
and had an impact on the thought of Buddhism can only be determined
in the light of evidence. It is worth noting that Prajapati’s theory of
the state of the soul after death is utterly different from Uddalaka’s.
In the Brahmajala Sutta there is a reference to a theory held by a
school of brahmins (eke samana-^mA/ncz/za, D. I.30) to the effect that
the soul after death has form (rupi), is without defect (arogo, lit.
without disease) and is conscious (sanni) (D. I.31). Prajapati’s theory
assigned all these characteristics to the soul after death. The soul has
form since ‘it appears in its own form (svena r^enabhinispadyate,
8.12.3). It is without defect or disease since it is said that ‘when crossing
that bridge (to the next world) if one is blind he becomes no longer
blind, if he is sick he becomes no longer sick’ (etam setum tlrtvandhah
sann anandho bhavati, viddhah sann aviddho bhavati, Ch. 8.4.2).
And the soul is conscious because if it so desires it becomes conscious
of enjoyment with women, chariots or relations (8.12.3). According to
Uddalaka’s theory on the other hand the soul would be without form
40
Early Buddhist Theory of Knowledge
(arupi), without defect (arogo) and without consciousness 1 (asannijjj
being merged in Being, which is another theory about the after-lif|
ascribed to a school of brahmins in the Brahmajala Sutta ( y . D. I.32).
(41) Let us now briefly consider some aspects of the thought of Yaj-
navalkya. Yajnavalkya’s importance for us lies in the fact that he too i$;
a rationalist thinker, who popularized a double negative form of!
expression, used in the Buddhist texts. His theory of survival also seems
to be known to Buddhism. 2
(42) Yajnavalkya has been called a mystic (Mystiker) 2 by Ruben but
this is misleading since there is no reason to believe that Yajnavalkya’s*
theories are based on any kind of mystical experience as were the
views of most of the thinkers of the Middle and Late Upanisads.;
Besides, another fact that has not been considered in the treatment of
Yajnavalkya is that the teachings ascribed to him in different places in.
the Upanisads do not seem to be of a piece, consistent with each other.
For instance, on the one hand the neti neti doctrine or the transcendent
conception of Brahman, who is describable only in terms of negative
epithets, is attributed to him (Brh. 3.9.26, 4.5.13) and on the other
hand the pantheistic doctrine totally opposed to it to the effect that
Brahman is ‘made of this, made of that’ (idammayah adomaya iti,
Brh. 4.4.5). The probable explanation for this is that several incom-
patible doctrines were put in the mouth of an outstanding teacher.
(43) We shall confine ourselves to Yajnavalkya of the neti neti doctrine
and consider the passages ascribed to him in sections 2.4.1-14, 3.9.28
and 4.5.15 of the Brhadaranyaka Upanisad. Now Deussen quite
rightly traces ‘the primitive source of the entire conception of the
unknowableness of the atman’ 3 to the statements ascribed to Yaj-
navalkya in this Upanisad, but it is equally necessary to emphasize the
fact that the rational unknowability of the atman is rationally arrived
at and is not a product of mystic experience. In fact Deussen himself
points this out when examining (Brh. 2.4.12-14=4.5.13-15), the
locus classicus of this doctrine, he says, ‘On careful consideration two
thoughts will be found to be implied here: (1) the supreme atman is
unknowable because he is the all-comprehending unity, whereas all
1 Note his saying that on reaching Being after death, ‘they know not, “I am
this one”, “I am that one” * (Ch. 6.10.1).
2 Geschichte der Indischen Philosophie, p. 95.
3 The Philosophy of the Upanishads , Edinburgh, 1906, p. 79.
4i
The Historical Background
knowledge presupposes a duality of subject and object; but (2) the
individual atman also is unknowable because in all knowledge he is
the knowing subject, consequently can never be the object’. 1 We agree
with Deussen that two arguments are implicit in this passage, first
that since reality is one and knowledge is dual, we cannot have know-
ledge of reality and second that the subject of knowledge cannot be
known since it is never the object and ‘thou canst not know the knower
of knowing’ (na vij hater vijnataram vijanlyah, Brh. 3.4.2). This is
reasoning and not mystic experience.
(44) Yajnavalkya’s conception of the after-life is also a product of
reasoning. At Brh. 3.9.28, he compares man to a tree. Now a tree when
it is felled at the root grows up again from the root ‘but when a man
dies from what root can he grow up’ (martyah svin mrtyuna vrknah
kasman mulat prarohati, 3.9.28.4). One cannot say it is from semen
(retasa iti ma vocata, 3.9.28.5) for that is possible even while the person
is living. The answer given is cryptic for it is said that ‘when born he
is not born again for who would again beget him’ (jata eva na jayate,
konvenam janayet punah , 3.9.28.7). This is a plain denial of the
possibility of rebirth ( punar janman ) and his theory seems to be as he
himself states that ‘after death there is no consciousness’ (na pretya 2
samjnasti, 4.5.13). By this he means the absence of any sense-conscious-
ness, since this is possible only by the presence of the sense-organs,
the uniting place (ekayanam) of the sense-data (3.5.12). But he is at
the same time not denying that all consciousness is absent since the
subject of consciousness, conceived by him as ‘a mass of pure ex-
perience’ ( ayam atma , anantaro’bahyah, krtsnah prajhana-ghana eva ,
4.5. 1 3) persists. So the state of survival is one in which ‘there is no
samjna’ nor a lack of it, i.e. no asamjna. Nov/ in the Majjhima Nikaya
(II.231) there is a reference to a school of recluses and brahmins who
argued that ‘the state of being neither conscious nor unconscious’
(na-eva-sahha-na-asanna) was a peaceful (santa-) and an excellent
state (panitam) because on the one hand ‘normal consciousness is
defective, a disease, a thorn’ ^sanna rogo sanna gando sanna sallam,
loc. cit .) while ‘unconsciousness is utterly bewildering’ (asanna
sammoho ). Yajnavalkya’s conclusion is at least the same, though his
1 Op. cit., pp. 79, 80.
2 In all the sixteen contexts (s.v. ‘pre-’ i n Jacob, Concordance to the Principal
Upanisads ) in which this verbal form is used in the Brhadaranyaka Upanisad, the
reference is to departure at death.
42
Early Buddhist Theory of Knowledge
argument is somewhat different but we may observe the parallel
between the use of sammoha in the Pali passage with the use of moha-
to denote the bewilderment of Maitreyi (mohantam aplpipat) when
Yajnavalkya says that ‘after death there is no consciousness’ at which
Yajnavalkya replies, ‘I am not speaking of a state of bewilderment’
(na va are’ham moham bravlmi, 8.5.14) probably implying as in the
Pali passage to a state of utter blankness in which there is no con-
sciousness at all.
(45) We have illustrated the use of reason by the Early Upanisadic
thinkers by the examples of Uddalaka, Prajapati and Yajnavalkya.
While they give unmistakable evidence of the presence of reasoning
during this period, the Chandogya Upanisad, according to the inter-
pretation of Sankara, and the translations of Deussen, Hume and
Radhakrishnan speaks of Logic (vakovakya-) as being one of the
branches of study during this period (v. infra , 51). Before we can under-
stand what is meant here by vakovakya- it is necessary to study the
origins of the debate for there is reason to believe that it was in and
out of these debates that the first conceptions of valid and invalid
reasoning arose (v. infra , 348). Brough seems to believe otherwise when
writing the ERE. article on Logic he says: ‘The historical beginnings
of logical theory are to be found in the racial dispositions and social
conditions which gave occasion for the deliberate control of our trains
of thought. In India, it appears to have originated with rules in cere-
monial deliberation.’ 1 He contrasts the example of India with that of
Greece saying that ‘in Greece it originated with canons of public
debate and scientific instruction’. 2 Randle on the other hand com-
menting on the Kathavatthu says that ‘logic was preceded by attempts
to schematize discussion, attempts which were inevitable in view of
the habit of organized public discussion, which prevailed in early
India but which could not succeed until the nerve of argument had
been separated from the irrelevancies in which the early methodology
had obscured it and plainly exposed in the formulation of the syl-
logism’. 3 We cannot wholly agree with either of these verdicts. If by
‘logical theory’ Brough meant the problems of epistemology then
certainly some of these problems, such as whether testimony (sabda)
was a genuine means of knowledge, the meaning of words and
1 Vol. 8, p. 128. 2 Ibid.
3 H. N. Randle, Indian Logic in the Early Schools , Oxford University Press,
I 93°, p. 14; cp. Th. Stcherbatsky, Buddhist Logic , Vol. I, pp. 27 If.
The Historical Background 43
propositions, the problem of universal may perhaps be ultimately
traced to ‘ceremonial deliberation’ and a significant portion of this
was contributed by the Grammarians (Vaiyakaranas) but the first
awareness of the validity and invalidity of reasoning seems to have
arisen out of the debate as much in India as in Greece. On the other
hand, Randle seems to imagine that the only forms of valid reasoning
must be syllogistic and it was probably this which led him to ignore
the foreshadowings of some of the theorems of the propositional
calculus in the Kathavatthu ( y . infra, 703-710).
(46) The debate in the Indian context seems to have its historical
origins in the Vedic institution of the hrahmodya 1 (or brahmavadya ).
A brief glance at the history of the brahmodya seems profitable in so
far as it gives a picture of the origin and development of the debate.
The earliest brahmodyas are riddles or religious charades which are
to be found in the Rgveda (1.164, 8.29) or the Atharvaveda (9.9, 10).
They frequently occur in the Brahmanas. 2 Their general form is that
of question and answer though sometimes the answers are cryptic or
the questions presupposed. 3 When the sacrifice became the reigning
institution in Brahmanic society, the brahmodya was a minor diversion
within it. Bloomfield calls it in this context ‘a charade to enliven the
mechanical and technical progress of the sacrifice by impressive
intellectual pyrotechnics’. 4 Keith says, ‘it is a feature of the Vedic
sacrifice that at certain points are found Brahmodyas, discussions about
the Brahman, the holy power in the universe. Such theosophical
riddles are specially common at the horse sacrifice’. 5 The following is
an example of such a brahmodya as related in the SB. (13.5.2. 11 ff.):
11. They hold a Brahmodya in the Sadas . . .
12. Hotr — Who walketh singly? Adhvaryu — the sun.
1 v. Keith, Religion and Philosophy of the Vedas, HOS., Vol. 32, pp. 344, 345,
435; A. Ludwig, Der Rigveda oder die Heiligen Hymnen der Brahmana, Band III,
Prag, 1878, pp. 390 ff.; Bloomfield, Religion of the Veda , pp. 216 ff.; Bloomfield,
‘The Marriage of Saranyu, Tvastar’s Daughter’, JAOS., Vol. 15, pp. 172 ff.;
M. Haug ‘Vedische Rathselfragen und Rathselspruche’, Transactions of the
Munich Academy, 1875, pp. 7 ff. This last reference is quoted from Bloomfield,
JAOS., Vol. 15, fn. £, and I have not been able to find this article in the libraries
of London and Cambridge.
2 These references are given in the article of Bloomfield, JAOS., Vol. 15,
p. 172.
3 Bloomfield, JAOS., Vol. 15, p. 172.
4 Religion of the Veda, p. 215.
5 HOS., Vol. 32, p. 435.
44
Early Buddhist Theory of Knowledge
13. Adhvaryu — Whose light is equal to the sun?
Hotr — The Brahman.
15. Udgatr — Into what things has the spirit (purusah) entered?
Brahman — Into five things hath the spirit entered and they are
established in the spirit: this I reply unto thee: not superior in
wisdom art thou (to me). 1
(47) These brahmodyas were uttered in the form of a dialogue,
technically called vakovakya (lit. speech and reply? cp. vakovakye
brahmodyam vadanti, they utter the brahmodya in the form of a
dialogue, SB. 4.6.9.20). These dialogues are formal and stereotyped
and were probably learnt by heart. They seem to have been among the
earliest passages to be so learnt for they are mentioned along with the
study of just the rc, saman and yajus (madhu ha va rcah, ghrtam ha
samamrtam yajumsi, yaddha va ayam vakovakyam adhlte kslraudana-
mamsaudanau haiva tau, the Rc verses are honey, the Sama verses
ghee, and the Yajus formulae ambrosia, but when he utters the dialogue
it is both milk and meat, SB. 11. 5.7.5). Later the list of things to be
studied becomes longer and includes vidya, itihasapuranam, etc. (SB.
11.5.6.8). But what is important is that a time seems to have come
when the vakovakya was no longer a formal utterance but an ex
tempore performance and the study of vakovakya- would have then
become the study of the nature of discussion and debate, whereby one
could outwit one’s rival. It may be observed that this desire to outwit
one’s rival is already seen in the example we quoted above where it is
said by the Brahman ‘not superior in wisdom art thou (to me)’. Such
a ‘free’ vakovakya- to which Sayana has drawn our attention 2 is to be
found at SB. 1 1.4.1. 12-15, which is not even a debate in a sacrificial
session but an open contest for victory between Uddalaka Aruni and
Svaidayana Gautama. We have already studied samples of the reason-
ing found here (v. supra , 14).
(48) When we come to the Early Upanisads this analogical reasoning
tends to lose its magical character (not altogether) and becomes more
empirical, though here too the inferences are not strictly warranted by
the observations made. Thus the observation that we perspire when
it is hot was sufficient for Uddalaka to conclude that ‘Heat causes
Water’ 3 (v. supra , 27). At this stage we noticed a priori reasoning as
1 SBE., Vol. 44, pp. 388, 389.
2 v. SBE., Vol. 5, p. 98, fn. 3.
3 The Milesian philosophers seem to have been at the same stage of thought.
The Historical Background 45
well, such as ‘Being cannot come out of Non-Being’ (v. supra , 25).
Both these forms of reasoning were made use of to construct meta-
physical theories. Vakovakya- at this stage of its development may have
signified a general study of this kind of reasoning as well as of topics,
which would help to make one a good debater.
(49) The debate at this stage seems to have been carried over from the
sacrifice to the public assembly and become an institution important
in itself and not a minor feature of a sacrificial session. Svetaketu
Aruneya goes for the purpose of debating to the assembly of the
Pancalas, which is called pancalanam parisadam (Brh. 6.2.1) and pan-
calanam samitim (Ch. 5.3.1). Sometimes the brahmins would go to the
courts of kings to hold such debates. Yajnavalkya goes to king Janaka
of Videha ‘desirous of cattle’ (the prize of the debate) and subtle
questions (pasun icchan anvantan, Brh. 4.1.1), and holds controversy
with him. So does Balaki come to Ajatasatru (Kaus. 4.1) and debate
with him. It was these assemblies of the brahmins and the ksatriyas
which came to be known as the brahmana-parisa and the khattiya-
parisa respectively in the Pali Nikayas ( y . infra, 349).
(50) But the debate on the sacrificial ground also seems to have con-
tinued without a break, though it was no longer a formal brahmodya
but a heated contest. There is a description of such a debate at a
sacrifice at Brh. 3.1-9, which Janaka attends and where he offers a
prize to the victor (i.e. to the wisest brahmin, brahmistha). In the
Mahabharata, it is said describing the proceedings of a sacrifice that
‘as the sacrifice progressed eloquent reasoners (vagmino hetuvadinah)
put forward many theories based on reasoning (hetuvadan) with the
intention of defeating each other’. 1 It is probably these brahmins who
called the ‘brahmins addicted to the debate’ (brahmana vadaslla) at
Sn. 382 (v. infra , 375). It is also probably to them that the Mahaniddesa
refers by the term hetuvada (Nd. I.294) though the term need not be
restricted to the brahmins. 2
(51) We found that the term vakovakya- was used in the Brahmanas
to denote a branch of study and observed that at a certain stage in its
development, it probably meant ‘the study of the nature of discussion
1 Tasmin yajne pravrtte "pi vagmino hetuvadinah, hetuvadan bahun ahuh,
parasparajigisavah, Srlmanmahabharatam, Asvamedhaparva, 86.27, Ed. T. R.
Krishacharya and T. R. Vyasacharya, Vol. 14, Bombay 1909, p. 103.
2 The Materialists were called ‘haitukah’, and probably the vitandavadins
(casuists) as well; v. Das Gupta, History of Indian Philosophy, Vol. 3 , p. 5 X ^*
4 6 Early Buddhist Theory of Knowledge
and debate*. One of its latest occurrences is in the Chandogya Upanisad
(7.1.2: 7.2.1: 7.7.1), after which it does not occur at all in this sense in
Indian literature. Sankara commenting on this word explains it as
tarka-sastram} and vakovakya- is translated by Deussen 2 as ‘Dialektik’
and by Hume 3 and Radhakrishnan 4 as ‘Logic*. This is by no means
unreasonable, for the study of the debate may have led to or included
at this time the study of ‘the elements of reasoning* and so long as
vakovakya- as ‘logic* is not taken to mean what logic (nyayasastra-,
tarka-sastra-) later came to denote there is no insuperable objection to
this translation.
(52) That the brahmins were studying some kind of tarkasastra also
appears to be confirmed by the evidence of the Pali texts for here
lokayata - (D. I.11.88; A. I.163, 166; A. III.223;Vin. II.i39;Sn. p. 105)
is represented as one of the branches of study of the orthodox brahmins
and this is explained as vitanda-vada-sattham 5 or ‘the science of
casuistry’ (DA. I.24: SnA. 447) or vitanda-sattham . 6 As Prof. Rhys
Davids has shown, what is stated in the Canonical texts is confirmed
by a passage in the Mahabharata where ‘at the end of a list of the
accomplishments of learned Brahmans they are said to be masters of
the Lokayata’. 7 Thus both according to Sankara as well as the Pali
texts, the early brahmins were making a study of the elements of
reasoning or casuistry or debating topics and this is by no means
intrinsically improbable, when we find that these brahmins were
constructing the first rational metaphysical theories at this time.
(53) Faddegon has however questioned the translation of this whole
passage by Deussen and Hume on the grounds that ‘the commentator
has tried to find in the Upanisad-text all sciences known in his time’ 8
and dismisses the translation of vakovakya- as ‘logic’ as unsupported
by the use of this word at SB. 11. 5.7.5 (op. cit ., p. 47). He says ‘we
may conclude that vakovakya in the Chandogya-Upanisad cannot yet
1 Vakovakyam tarka-sastram, Chandogyopanisad, Ananda Asrama Series
No. 14, p. 393.
2 Sech^ig Upanisads des Veda , Leipzig, 1921, p. 174.
3 The Thirteen Principal Upanishads , pp. 250, 251, 254.
4 The Principal Upanisads , pp. 469, 470, 475.
5 Lokayatam vuccati vitandavadasattham.
6 Abhidhanappadlpika, 5.12, vitandasattharp vinneyyam yam tarn lokayatam,
v . Abhidhanappadlpika, ed. Muni Jina Vijaya, p. 18.
7 SBB, Vol. II, p. 169.
8 B. Faddegon, The Catalogue of Sciences in the Chandogya Upanisad, AO.,
Vol. 4, p. 44.
The Historical Background 47
have signified logic, since this science was developed many centuries
later as an outcome of the technical art of philosophical discussion’
( loc . cit.). He suggests for it ‘the general meaning of dialogue and
metaphorically ability and smartness in debating’ {loc. cit.) or ‘the
cleverness of arguing in dialogue’ (op. cit., p. 53). We would agree
with Faddegon that vakovakyam could not have meant ‘the elements
of reasoning’ at SB. 1 1. 5.7.5 but the case is different with its latest use at
Ch. 7.2.1, etc., after the institution of the brahmodya, as we have seen
(v. supra , 46-50) underwent many changes. As for Faddegon’s other
argument discrediting almost the entirety of Sankara’s interpretations
of this passage on the grounds that they are anachronisms, we would
like to point out that Sankara’s exposition on the whole is corroborated
by what the Pali Nikayas attribute to the brahmins as the arts and
sciences studied by them. In fact, this independently supports the
Chandogya list itself by showing that it does not contain later inter-
polations. It shows that the catalogue of sciences in the Chandogya
as well as Sankara’s comments on the whole are to be trusted as giving
a fair sample of Brahmanic learning at least at the time of the Pali
Nikayas. We may do this in the form of a table giving the Chandogya
catalogue, Sankara’s comment, the word in the Pali Nikayas which is
the equivalent either of the Chandogya catalogue or Sankara’s com-
ment. We have indicated in brackets the equivalents found only in a
Pali Corny.:
Chandogya list
r. atharvanam
caturtham
2. itihasapuranam
pancamam
3. vedanam vedam
4. pitryam
5. rasi
6. daivam
7. nidhim
8 . vakovakyam
Sankara s comment
bharatapanca-
manam
vyakaranam
sraddha-kalpau
ganitam
utpatajnanam
mahakaladinidhi-
£astram
tarkasastram
Pali equivalent
athabbanam, Sn. 927
itihasapancamanam
D. 1.88
veyyakarana-,
D. 1.88
saddhe, D. I.97:
ketubha-, Sn. 1020
ganana, D. I.n
utpatam, D. 1 . 8 , v. 1
?
lokayatam = (vitanda-
vada-sattham ,
DA. I.247)
48
Early Buddhist Theory of Kn
.owledge
Chandogya list
Sankara s comment
Pali equivalent
9 -
ekayanam
mtisastram
(nltisattham, DA.
1 - 93 )
10.
devavidya
niruktam
sivavijja?, D. I.9
11.
brahmavidya
rgyaj uhsamakhya-
syavidya
tevijja, Sn. 594
12.
bhutavidya
bhutatantram
bhutavijja, D. 1.8
13 -
ksatravidya
dhanurvedam
khattavijja, D. I.9
14.
naksatravidya
jyotisam
nakkhattam, Sn. 927
J 5 -
sarpa (vidya)
garudam
ahivijja, D. 1.8
1 6.
devajanavidya
gandhayukti-nrtya-
gita-vadya-silpa-
naccam gitam
vaditam, D. 1.6
(54) We do not propose to scrutinize this list item by item, as it
would divert us from our present problem. But if we examine this list
as a whole, it would be noticed that five of the Pali items (i, 2, 12, 13,
14) are identical in word and meaning with the Chandogya list, while
one of them (15) is identical in meaning and one (10) doubtful. Of the
rest, no less than six Pali items (3, 4, 5, 6, 11, 16) are more or less
identical in language and sense with the comments of Sankara. This
lends authenticity to Sankara’s comments in that it shows that these
sciences were cultivated by the brahmins at least during the time the
Pali Nikayas were composed, if not earlier and Sankara could therefore
not have been making arbitrary comments particularly with regard to
items 4, 5, 6, 11, 16, which have been questioned by Faddegon. All
this implies that when Sankara was commenting on vakovakya- as
tarkasastra- there is no reason to think that he was trying to find a place
for tarkasastra in this list, but that he was probably recording a genuine
tradition, particularly when we observe that the Buddhists have credited
the brahmins with making a study of what they in their poor opinion
of them have called the vitanda-sattha or the ‘art of casuistry’.
The fact that when Pali commentaries came to be written Lokayata-
exclusively meant Materialism is perhaps an added reason why the
comment vitanda-sattha, quite independently of the corroboration from
Brahmanic sources is to be considered as preserving a genuine tradition.
(55) The etymology of the word lokayata- however, does not even
remotely suggest any connection with logic or casuistry. On the other
hand, all the explanations of the etymology of the term by scholars 1
1 v. Chattopadhyaya, Lokayata, A Study in Ancient Indian Materialism , New
Delhi, 1959, pp. 1-4. Das Gupta, A History of Indian Philosophy , Vol. Ill,
pp. 51 2 ff-
The Historical Background 49
on the presumption that the word directly or indirectly means ‘the
philosophy of Materialism* are utterly mistaken since the earliest use
of the word, as we have seen, does not at all betray such a connotation.
As Prof. Rhys Davids was the first to point out in his study of the
meaning of the term, 1 lokayata as used in the Nikayas is one of the
branches of learning of the orthodox Vedic brahmins. Speaking of the
context in which the word appears he says: ‘The whole paragraph is
complimentary. And though the exact connotation of one or two of
the other terms is doubtful, they are all descriptive of just those things
which a Brahman would have been rightly proud to be judged a master
of. It is evident, therefore, that the Dictionary interpretations of the word
are quite out of place in this connection .* 2 It is necessary to point this
out since this statement seems to have fallen on deaf ears in the field of
scholarship 3 and no attempt has been made to explain the meaning of
this earliest use of the term.
(56) Prof. Rhys Davids himself suggested that the word ‘probably
meant Nature-lore — wise sayings, riddles, rhymes . . .* (op, cit ., p. 171).
He even gave a list of passages in the Brhadaranyaka and Chandogya
Upanisads and the Aitereya Aranyaka (loc, cit .), which he believed
contained the subject-matter of lokayata- and suggested rather half-
heartedly that with the growth of this branch of learning it came to be
associated with ‘sophists and casuists* (loc, cit.). All this was pure
surmise, based on his belief that loka- meant ‘nature* and that lokayata-
meant the ‘study of nature* and that there was evidence for this in the
Upanisadic and Aranyaka literature. Against this, Tucci has pointed
out that loka- by itself does not mean ‘nature* in the Pali literature and
that for this purpose the word bhajana-loka is used. 4
(57) It is surprising that Prof. Rhys Davids and after him all the
scholars who discuss the meaning of lokayata- missed both passages
in the Nikayas which could have given some information about the
subject-matter of lokayata, one occurring in the Samyutta Nikaya
(II. 77) and the other in the Aiiguttara Nikaya (IV.428). The former is
1 SBB., Vol. II, pp. 166-72. 2 Op. cit., p. 166.
3 E.g. Chattopadhyaya uncritically quotes Prof. Rhys Davids to show that
the early brahmins studied lokayata- in the sense of Materialism (v. op. cit.,
p. 32) and says, ‘Evidences like these perhaps indicate that we are in need of
revising our notion of the Brahmana, particularly of the Brahmana of Buddhist
India’ {op. cit., p. 33). ‘A Sketch of Indian Materialism’, PIPC., 1925, p. 40.
4 ‘A Sketch of Indian Materialism’, PIPC., 1925, p. 40.
50
Early Buddhist Theory of Knowledge
quoted by Dr Malalasekera in a paragraph he has written about the
‘lokayatika brahmana’ ( s.v . DPPN.) ignoring the problem of the
meaning of lokayata altogether, while the latter passage which also
mentions lokayatika brahmana ( loc . cit .) is not mentioned by him as
well. 1 In both these contexts, ‘two lokayata brahmins’ (dve lokayatika
brahmana) approach the Buddha. This expression is translated by
Mrs Rhys Davids as ‘two Brahmins wise in worldly lore’ (K.S. II.53)
and by Hare as ‘two brahmins skilled in metaphysics’ (G.S. IV.287).
The Corny, to the Samyutta Nikaya explains lokayatika as ‘one
versed in Lokayata or the science of casuistry’ (lokayatiko ti vitanda-
satthe lokayate kata-paricayo, SA. II.76) and in the Anguttara Corny,
the word is explained as ‘students of Lokayata’ (lokayata-pathaka,
AA. IV.200). The term lokayatika- seems to describe the brahmin who
made a special study of that branch of Brahmanic learning known at
the time as Lokayata.
(58) These doctrines are specified in the Samyutta context as follows:
(1) Sabbam atthl ti, i.e. ‘that everything exists’, which is called the
oldest (jittham, Skr. jyestham) lokayata-doctrine.
(2) Sabbam natthl ti, i.e. ‘that nothing exists’, called the second
(dutiyam) lokayata-doctrine.
(3) Sabbam ekattan ti, i.e. ‘that everything is a unity’, called the third
(third) lokayata-doctrine.
(4) Sabbam puthuttan ti, i.e. ‘that everything is a plurality’, called the
fourth (catuttham) lokayata-doctrine.
(59) It may be observed that all these theories are about sabbam or
sarvam, which is found in the Upanisads to denote the ‘cosmos’ or
the universe as a whole (r. infra, 65). It will also be seen that these
four doctrines are presented in two pairs as thesis and anti-thesis:
the second and the fourth are the anti-theses of the first and the third
respectively. The Corny, explains that the first and the third are
Eternalist views 2 (sassata-ditthiyo) while the second and the fourth are
Materialist views 3 (uccheda-ditthiyo, lit Annihilationist views). This
dialectical opposition in these pairs of views reminds us of the Vedic
institution of the brahmodya , which found expression in the form of a
1 Lokayatika- is not even mentioned in the Volume of Indexes (A. VI) of the
Anguttara Nikaya.
2 Evam ettha sabbam atthi, sabbam ekattan ti ima dve pi sassata-ditthiyo, SA.,
11 , 76 .
3 Sabbam natthi, sabbam puthuttan ti ima dve uccheda-ditthiyo, ibid .
5i
The Historical Background
vakovakya- (y. supra. , 47), which was originally a dialogue and later a
debate (dialectics) in which one tried to outstrip the other by argu-
ments, designed to disprove one’s opponent’s thesis and prove his
own. The other main deduction that we can make from the above
passage is based on the commentarial identification of the second and
the fourth views as those of the Materialists. This suggestion is con-
firmed by the fact that there is evidence of the existence at this time
of these two schools of Materialists referred to, the pluralist school
(or schools, v. infra , 115) and the nihilist pragmatic school, which we
have argued was the school to which Dighanakha belonged (y. infra ,
334) and which adumbrates the later philosophy of Jayarasi (y. infra ,
11 6). The fact that Lokayata is the term which later comes into cur-
rency as a general term for these Materialist schools of thought also
supports this identification. If this is so, then the later use of the term
Lokayata to denote exclusively the Materialist doctrines is a one-sided
application and development of a term, which had a wider coverage
earlier, denoting as we see not only the Materialist doctrines but their
anti-theses, the Eternalist doctrines as well. In fact, it may be noted
that according to this passage the oldest lokayata is not the Materialist
doctrine but the eternalist doctrine.
(60) In the Anguttara context too, ‘two lokayata brahmins’ (dve
lokayatika brahmana), loc . cit., meet the Buddha to discuss the problem
of the extent of the cosmos (loka-). They say that Purana Kassapa and
Nigantha Nataputta are ‘directly opposed to each other’ (annamahnam
vipaccanlkavadanam, M. I.429) in regard to the views that they hold
about the extent of the universe, one holding that ‘the universe is
finite’ (antavantam lokam, loc . cit.) and the other that ‘the universe is
infinite’ (anantam lokam, loc. cit.). It is possible that these two theses
constituted a pair of lokayata-doctrines, in which case loka- is here
used in the sense of the ‘cosmos’ (y. infra , 65), and lokayata would
mean ‘what relates to the cosmos’ or the problems of the nature and
extent of the cosmos, studied as debating topics and based on reasoning.
(61) The Lankavatara Sutra 1 also records an encounter between the
Buddha and a lokayatika brahmin. This gives a long list of lokayata-
doctrines and although it is less reliable than the Nikaya passages with
regard to what it tells us about Brahmanical doctrines, we can never-
theless glean some information. This passage too has been ignored by
scholars in discussing the meaning of lokayata- and Suzuki has
1 Ed. B. Nanjio, Kyoto, 1923, Bibliotheca Otaniensis, Vol. I, pp. 176—9.
52 Early Buddhist Theory of Knowledge
consistently mistranslated the term as ‘materialism’ 1 though it is
obvious from the context that it could not mean this. Here thirty-one
lokayata-doctrines are mentioned as follows:
(1) Sarvam krtakam, i.e. everything is created, called the first
lokayata- theory (prathamam lokayatam).
(2) Sarvam akrtakam, i.e. nothing is created, called the second
lokayata- theory (dvitiyam lokayatam).
(3) Sarvam anityam, i.e. everything is impermanent.
(4) Sarvam nityam, i.e. everything is permanent.
(5) Sarvam utpadyam, i.e. everything is resultant.
(6) Sarvam anutpadyam, i.e. everything is not resultant, called the
sixth lokayata- theory (sastham lokayatam).
(7) Sarvam ekatvam, i.e. everything is a unity.
(8) Sarvam anyatvam, i.e. everything is different (the world is a
plurality).
(9) Sarvam ubhayatvam, i.e. the world is a duality.
(10) Sarvam anubhayatvam, i.e. the world is a non-duality.
*(n) Sarvam karanadhinam, i.e. everything is subject to causation
since they are seen to proceed from a diversity of causes
(vicitra-hetu-prapatti-darsanat).
(12) Sarvam avyakrtam, i.e. everything is inexplicable.
(13) Sarvam vyakrtam, i.e. everything is explicable.
(14) Asty atma, i.e. there is a soul.
(15) Nasty atma, i.e. there is no soul.
(16) Asty ayam loko , i.e. this world exists.
(17) Nasty ayam loko , i.e. this world does not exist.
(18) Asti paro loko , i.e. the next world exists.
(19) Nasti paro loko , there is no next world.
*(20) Nasty asti ca paro loko , i.e. there is and is no next world.
(21) Asti moksah, i.e. there is salvation.
(22) Nasti moksah, i.e. there is no salvation.
(23) Sarvam ksanikam, i.e. everything is momentary.
(24) Sarvam aksanikam, i.e. nothing is momentary.
(25) Akasam pratisankhyanirodho nirvanam krtakam, i.e. space,
non-wilful destruction and nirvana are conditioned.
(26) Akasam pratisankhyanirodho nirvanam akrtakam, i.e. . . . are
not conditioned.
(27) Asty antarbhavah, i.e. there is an intermediate existence.
1 v. D. T. Suzuki, The Lankavatara Sutra , London, 1932, pp. 152-5.
The Historical Background 53
(28) Nasty antarbhavah, i.e. there is no intermediate existence.
(29) Ajnana-trsna-karma-hetukam . . . tribhavam, i.e. the three-fold
world is caused by ignorance, desire and karma.
(30) (tribhavam) ahetukam, i.e. the threefold- wo rid is not caused:
and it is said that ‘this pair too constitutes a lokayata (dvaya-
tnapy etat . . . lokayatam).
*(31) Sva-samanya-laksana-patita sarva-bhavah, i.e. all things are
classifiable under their specific and general characteristics.
(62) The section ends by saying that ‘there is lokayata as long as the
mental activity of the dogmatic construction of the external world
persists’. 1 This is an attempt to explain the origin of lokayata-theories
on the basis of the assumptions made in the Lankavatara Sutra itself
and it is therefore not very enlightening. Even the list cannot be
considered to give us an account of Brahmanic doctrines. For instance,
theories (25) to (28) are topics on which, les sectes du petit vehicule,
to use Bareau’ s expression 2 were divided. Thus the two theses, namely
that ‘there is an intermediate existence’ and its opposite ( v . 27, 28)
are considered a pair of lokayata-theories. This is a doctrine on which,
as the Kathavatthu (VIII.2) shows the Buddhist sects were divided
and it was a subject of debate between the contending parties. 3 This
use of lokayata- to refer to the debating topics, mentioning thesis as
well as anti-thesis, on which the Buddhist order was divided seems
indirectly to throw some light on the earlier use of lokayata to refer
to the debating topics of the brahmins, on which opposing views were
found within the orthodox circle of brahmins. It will be seen that all
the above topics excepting (11), (20) and (31) — marked with an
asterisk — are stated in the form of thesis and anti-thesis and the fact
that they were considered in pairs appears to be confirmed by the
statement made about (29) and (30) namely that ‘ this pair too ( dvaya -
mapy etat) constitutes a lokayata ( v . supra, , 61). This would have been
1 Yavad . . . manovispanditam bahyarthabhinivesavikalpasya taval lokayatam,
op. cit., p. 178.
2 v. Andre Bareau, Les Sectes Bouddhique du Petit Vehicule, Publications de
1 ’ecole Francaise d’extreme-orient, Saigon, 1955.
3 As Bareau has shown {op. cit.) the Purvasailas (p. 101), Vatsiputriyas (p. 119),
the Sammatiyas (p. 124), the Sarvastivadin Vaibhasikas (p. 142) and the Late
Mahlsasakas (p. 188) were of the view that an antara-bhava exists, while the
Mahasanghikas (p. 68), the Vibhajyavadins (p. 172), the Mahlsasakas (p. 184),
the followers of the Sariputrabhidharmasastra (p. 194) and the Theravadins
(p. 223) were of the opposite view.
54 Early Buddhist Theory of Knowledge
unnecessary unless they were debating topics, representing the thesis
upheld by one party and the anti-thesis defended by the opposing
party. We may therefore surmise on this ground that the lokayata-
theories mentioned in the Samyutta Nikaya were probably stated in
the form of thesis and anti-thesis ( [v . supra , 59) because they were the
subjects of debate among those brahmins, whom the Suttanipata has
described as being ‘addicted to the debate 5 (v. infra , 375).
(63) The passage in the Samyutta Nikaya said that the oldest lokayata
thesis was ‘that everything exists 5 while according to the Lankavatara
Sutra, the first thesis was ‘that everything was created 5 . Now creation
theories have undoubtedly to be reckoned among the first cosmological
theories or the first attempts to comprehend the origin of the cosmos.
That Being or sat was the primary cause or the ultimate reality of the
universe appears to have been one of the earliest cosmological theories,
which was probably followed not very much later by the theory that
‘nothing 5 (asat, v. RV. 10.72, 2) 1 really exists and the Nasadlya hymn
probably attempted a synthesis of these two theories (v. supra , 9),
Even if we treat these theories as pre-philosophical, we notice that
the subject is treated at a philosophical level at Ch. 6.2.1. Here the two
theories are clearly contrasted: one is that ‘Being is the only reality 5
(sad eva . . . aslt) and the other, which is quite clearly held by ‘certain
people 5 (taddhaika ahuh) in opposition to this theory is that ‘nothing
is real 5 (asad eva . . . aslt). This may be deemed to be a reference to
the nihilist school of lokayata, which according to the Samyutta
Nikaya held the tenet that ‘nothing exists 5 (sabbam natthi, v. supra
58) and which is described in the Lankavatara Sutra as the school
which held that ‘this world does not exist 5 (v. 17). The other pair of
lokayata-theses mentioned in the Samyutta Nikaya (i.e. sabbam
ekattam ; sabbam puthuttam) also appear to have had their origin in
the Upanisadic period. The Bhagavadglta speaks of ‘some who worship
with the offerings of knowledge with (theories) of unity as well as of
plurality 5 (jnana-ya]nena. ca’pyanye . . . upasate ekatvena prthaktvena ,
9.15). We may note here that ekatva- and prthaktva- in Sanskrit give
rise to ekatta- and puthutta- in Pali according to the usual phono-
logical rules. Now the Isa speaks of the absence of delusion on the
part of those who see the universe as a unity ( ekatvam anupasyatah, 7)
and the Katha holding that ‘there is no diversity in the universe 5
(neha nanasti kincana, 2.1.11) criticizes ‘those who see diversity in it 5
asatah sad ajayata, RV. 10.72.2, 3.
The Historical Background 55
(mrtyos sa mrtyum gacchati ya iha naneva pasyati , loc, cit .); this
criticism must have been directed against a pluralistic theory of the
universe and it is likely that this is the Materialist theory mentioned at
Katha 1.2.6 ( v . infra , 116) where it is said that ‘he who thinks “this
world exists, there is no other” repeatedly comes under my (i.e. of
Death) control ’ (ayam loko nasti para iti mam, punah punar vasam
apadyate me, loc . cit) for the same fate (cp. mrtyos sa mrtyum gacchati)
was held out against those who were convinced of a pluralistic theory
of the universe.
(64) The main concepts of the lokayata theses also appear in the
Mulapariyaya Sutta (M. I.i ff.), which gives a list of categories or
concepts having a cosmological significance. Mulapariyaya has been
translated by Miss Horner as ‘synopsis of fundamentals’ (M.L.S.
I.3) but this translation does not make sense. We believe that mula-
here means the ‘root cause’ or the primary cause of the world. It is in
this sense that the word is used at Aitareya Aranyaka 2.1. 8.1, where
the cosmological theory that water is the first or primary cause of the
world is mooted and it is said that ‘this (water) was the root (cause)
and that (i.e. the world) was the shoot ’ (i.e. the effect) (etad vai miilam
adas tularn). In this Sutta we observe that this theory, namely that
water or apa- is a mula- or a root cause is mentioned along with a
number of such cosmological theories. Pariyaya here probably means
‘the nature of’ as at Sn. 581 - 1 Mulapariyaya Sutta, therefore probably
means ‘the discourse on the nature of primary causes or concepts’.
Among such causes or categories explaining the origin or the nature
of the universe, we find the concepts of ekatta -, nanatta- (= puthutta-
in sense), and sabba- (M. I.3).
(65) All this points to loka- in lokayata meaning not ‘nature’ as Prof.
Rhys Davids imagined but the ‘cosmos’. It may be seen that in the
Lankavatara list the lokayata-theses were about sarva- (i.e. the cosmos)
or loka -. The references in the Nikayas confirmed this (v. supra , 60).
Now the word loka- is used in a collective sense, to denote the entire
universe and this sense is in fact clearly defined at Brh. 1.5.17, where it
is said, ye vai ke ca lokah, tesam sarvesam loka ity ekata, i.e. whatever
worlds there are, they are all comprehended under the word ‘world’.
We also notice that in this same context loka- is used synonymously
with brahman: tvam brahma , tvam yajna, tvam loka iti. Brahman is
1 Tasma dhlra na socanti viditva loka -pariyayam, i.e. therefore the wise do not
grieve, knowing the nature of the world.
56 Early Buddhist Theory of Knowledge
also sometimes used synonymously with sarva-: etad brahma etad
sarvam , Brh. 5.3.1. In the light of all this evidence we may conclude
that these lokayata-theses were promulgated at the brahmodyas which,
as we have shown, developed from a simple formal dialogue into a
lively debate and it was probably in preparation for these debates that
the lokayata-theses would have been studied. They would have
constituted the possible answers to problems about the cosmos, along
with the reasons on which they were based. The study of the reasoning
would have been at first not strictly divorced from the theories them-
selves and it is to these studies of the brahmins that we have to trace
the beginnings of metaphysics as well as of logic and epistemology. A
verse in the Mahabharata describing the sage Narada shows that
logical studies (nyaya-) went hand in hand with the study of meta-
physical concepts such as ‘monism’ (aikya-, cp. ekatta) and ‘pluralism*
(nanatva-, cp. puthutta) and that this was part of Vedic studies as a
whole:
vedopanisadam vetta rsih suraganarcitah
itihasa-purana- j nah, purakalpa-visesakrt
nyaya Wd/dharmatattvajnah sadangavid anuttamah
a^ya-samyoga-zzcz/za^vcz-samavaya-visaradah, 1
Sabhaparva, 5.2-3.
This verse may very well reflect a time when the Nyaya or logical
studies were accepted by orthodoxy and admitted into the rank of
Vedic studies but in the light of the above evidence from the Buddhist
Nikayas, confirmed and corroborated by the Brahmanic literature, we
have to presuppose that there was a period when the study of lokayata
or the ‘elements of metaphysics and reasoning’ formed a part of Vedic
studies. However, a time seems to have come when some of the loka-
yata-theses propounded in the process, were seen to oppose or under-
mine the fundamental doctrines of the Vedic tradition and it no longer
seemed desirable for orthodoxy to allow brahmins the free exercise of
reason and speculation. Thus the rule was laid down that ‘the brahmin
who despises the roots (of Vedic tradition) because of his dependence
on the science of reasoning ( hetu-sastra -) should be cast out by the
good (brahmins) as a nihilist, who scorns the Vedas’ (yo’vamanyeta
1 ‘The supreme sage who was revered by the gods, knew the Vedas and
Upanisads, the histories and Puranas, was a specialist in ancient rituals, was
versed in logic, the truths of justice and the six branches (of learning) and had an
expert knowledge of the (concepts of) monism , conjunction, pluralism and
inherence/
The Historical Background 57
te mule hetusastrasrayad dvijah sa sadhubhir bahiskaryo nastiko
vedanindakah: Manusmrti II. n). At this time Lokayata- as a branch
of study would have been taboo to the brahmin orthodoxy and the
word lokayata survived to denote those very doctrines, which were
opposed to Vedic teachings but which were once nurtured within the
orthodox fold itself.
(66) This sense of lokayata- appears to be preserved in the Arthasastra,
where it is said to form part of anvlksikl or ‘philosophy’, comprising
both metaphysics and logic: samkhyam yogo lokayatam anvlksikl . . .
hetubhir anviksamana anvlksikl lokasyopakaroti, 1 i.e. Samkhya, Yoga
and Lokayata (constitute) philosophy ... by investigating with
reasons it serves the world; the Corny, (modern) explains lokayata
here as ‘the science of reasoning as taught by Brahma and Gargya’
(nyaya-sastram Brahmagargyoktam, Vol. I, p. 27). Anvlksikl was
rendered by Jacobi as ‘philosophic’. 2 But Hacker in an article entitled
‘Anvlksikl’ 3 has questioned this translation on the ground that since
anvlksikl according to Kautilya’s own comment means ‘examining by
reasons’ and this is practised in all the sciences the term does not
exclusively mean ‘philosophy’. Yet he too admits that ‘anvlksikl or
reasoning’ is ‘habitually applied to systems of philosophy because these
cultivate argument and logical thinking’ (op. cit ., p. 82) and his main
objection is that these terms are ‘never synonymous with philosophy’
(loc. cit.). It does not therefore disprove our contention that lokayata-
in its earliest use meant the study of metaphysical topics along with
the reasoning involved, with the idea of gaining success in debate.
(67) The ways of knowing recognized at this time are, as Keith has
shown, 4 stated in the Taittiriya Aranyaka as pratyaksa (perception),
anumana (inference), smrti (scripture) and aitihya (tradition). Keith
1 The Arthasastra of Kautalya , Ed. T. G. Sastri, Vol. I, p. 27.
2 v. ‘Zur Friihgeschichte der indischen Philosophic’ in Sitzungberichte der
Preussische Akademie der Wissensch often, Berlin, 19 1 1, pp. 733 ff.; cp. A. Foucher,
Le Compendium des Topiques ( Tarkasamgraha ) D’ Annambhatta, Paris, 1949, who
translates ‘anvlksikl’ as ‘investigation rationelle’ (Introduction, p. xi), which he
says is ‘completement independante des textes sacres et uniquement fondee sur
l’experience courante; et celle en embraissait, nous dit-on, le Sankhya, le Yoga et
le Lokayata, toutes doctrines originairement agnostique et realistes’ {loc. cit.).
He classifies Nyaya and Vaisesika under ‘Lokayata’ {loc. cit.).
3 Paul Hacker, ‘Anvlksikl’ in, Wiener Zeitschrift fur die Kunde Sud- und
Ostasiens und Archiv fur indische philo sophie. Band II, 1958, pp. 55 ff.
4 Religion and Philosophy of the Vedas , HOS., Vol. 32, p. 482.
58 Early Buddhist Theory of Knowledge
thinks that this represents ‘a late popular view’, 1 but the use of the
word smrti- to denote ‘scripture’ rather than the late word ‘sruti’,
the absence of the use of the word pramana and the general context of
the passage does not favour Keith’s view.
(68) When we analyse the language of the Early Upanisads we find
the use of three or four verbal forms juxtaposed to signify the ways of
knowing accepted at the time. Thus at Ch. 7.24.1, ‘pasyati . . . srnoti
. . . vijanati’, seems to sum up the different ways of knowing things.
The root drs- is used in the Early Upanisads predominantly to denote
the ‘seeing’ of visual objects with the eye (caksusa rupani pasyati , Brh.
3.2.5). In this sense it is found very frequently either singly 2 or in
lists where visual sensing or perception is contrasted with other forms
of sensing or sensory perception. 3 These lists do not however mention
all the five senses. Brh. 2.4.14 mentions smelling (jighrati), seeing and
hearing, while Brh. 4.5.15 refers to seeing, smelling, tasting (rasayati)
and hearing and Brh. 4.3.31 and 4.4.2 add touching (sprsati) to the
list, making five in all. When the verbal forms of y'drs- are used
without mention of the other forms of sensing or sensory perception,
it seems to denote not just visual sensing or perception but perception
in general. Thus, seeing (drstih) is used to denote perception in general
where it is defined that seeing in this instance consists in perceiving
the warmth of the body by touch (Ch. 3.13.8). The fact that the forms
of y'drs- were used predominantly to denote visual perception is
undoubtedly due to the simple fact that perhaps the largest number of
our perceptions are visual perceptions so that the word for visual
perception is gradually extended to denote perception in general.
(69) Yet auditory perception was precluded from being denoted by
V drs- since the verbal forms of y / sru had to be used side by side in
contexts, where ways of knowing were referred to, because of the
tremendous importance traditionally attached to hearing at this time.
This importance is due undoubtedly to the respect and reverence in
which the sacred scriptures were held and these scriptures could not
be seen 4 but had to be learnt by hearing them from one’s teacher. The
veneration in which hearing and learning from teachers was held is
clearly seen from one of the earliest references in the Upanisads. It is
1 Religion and Philosophy of the Vedas, HOS., Vol. 32, p. 482.
2 Brh. 1.3.4; 1.4.1; I.5.3.; 5.4.3; Ch. 1.2.4; 2.4.7; 2.24-4; 3 - 6 - 1 ; S - 1 - T , 5 - 12 - 1 ;
6.12.1; 7.11.1, etc.
3 Brh. 2.4.14; 4.5.15; 4.3.31; 4.4.2.
4 See, however, supra, 13.
The Historical Background 59
said that the eye is one’s human wealth for one finds it with the eye,
but that the ear is his ‘divine wealth’ (srotram daivam) for he hears it
with his ear (Brh. 1.4. 17). Here the divine wealth referred to is un-
doubtedly the sacred scriptures and the use of the epithet ‘divine’ to
describe what is heard as opposed to what is seen is indicative of the
authority attached to the former. It is important, however, to notice
that even in the Early Upanisads when it came to a matter of deciding
between the evidence of seeing and the testimony of report or hearing
about matters of fact in the everyday world, the decision was made in
favour of sight against hearing as being the more reliable. It is said:
‘Truth is sight. Therefore if two persons come disputing, one saying
“I saw” and the other “I heard” they should trust the one who says
“I saw” (caksur vai satyam, . . . tasmad yad idanim dvau vivadamanau
eyatam aham adarsam, aham asrausam iti. Ya evam bruyat, aham
adarsam iti tasma eva sraddadhyama, Brh. 5.14.4).’ We find this idea
persisting later in the Maitrl Upanisad, where it is said that ‘here the
evidence is what is observed (by the senses)’ (atra drstam nama prat-
yayam, 6.10). This is possibly the reason why the Taittlriya Aranyaka
distinguishes between two kinds of knowledge obtained from hearing,
namely what is authoritatively heard and remembered (smrti), that is
the sacred scriptures and what is learnt from report or tradition
(aitihya) with regard to other matters, a distinction which led to the
necessity to separate divinely revealed scripture (sruti) from fallible
human tradition (smrti).
(70) In addition to perception or hearing (or learning) there is mention
of thinking (y^man; vi + Vjna; ni -f- \/dhya) as a means of know-
ledge at this time. The verbal forms used cover the rational reflective
sources of knowledge, which the Taittirlya Aranyaka appears to indi-
cate by the word anumana, i.e. reasoning, or inference. The thinking
process is sometimes described by the single word vijanati but at
other times a distinction appears to be drawn between the two cognitive
processes of mental conceiving and rational understanding, a dis-
tinction which is not very clear. We may list the references to ways of
knowledge as reflected in the language of this period as follows:
Text seeing or hearing or mentally rationally
perceiving learning conceiving understanding
Brh. 2.4-5) drastavyah srotavyah mantavyah nididhyasitavyah
4 - 5*6
60 Early Buddhist Theory of Knowledge
Text seeing or
hearing or
mentally
rationally
perceiving
learning
conceiving
understanding
Brh. 3.4.2 pasyeh
srnuyah
manvlthah
vijaniyah
Ch. 3.13.8 drstih
srutih
Ch. 6.1.3
srutam
matam
vijnatam
Ch. 7.25.2 pasyan
manvanah
vijanan
Ch. 7.26.1 pasyatah
manvanasya
vijanatah
(71) Let us consider an example. It is said that ‘the atman should be
perceived (drastavyah), learnt of (srotavyah), conceived of (mantavyah)
and rationally understood (nididhyasitavyah) (Brh. 2.4.5, 4*5 «6). This
is put in the mouth of the Yajnavalkya, who is soon going to prove by
rational arguments that the atman cannot be apprehended by any of
these standard ways of knowing (Brh. 2.4.14, 4.5.15), but if we con-
sider this passage in the light of other passages bearing on it in this
stratum of thought, we see that there were thinkers at this time who
believed that the atman could be known by all these usual ways of
knowing. The atman could be seen or empirically perceived if it was
a matter of seeing your figure in a pan of water (Ch. 8.8.1) or of
perceiving the warmth of the body (Ch. 3.13.8). It could be heard or
heard of, if it was a case of hearing the sound as of a fire blazing on
closing one’s ears (Ch. 3.13.8) or of hearing about it from a teacher
when ‘what was not heard of’ (asrutam) presumably in the sacred
scriptures becomes heard (srutam bhavati) (Ch. 6.1.3). It could like-
wise be metaphysically conceived of and rationally understood by
thinking (e.g. vijajnau, Ch. 6.16.3). I* significant that even Sankara’s
comment on ‘mantavyo nididhyasitavyah’ (Brh. 4.5.6) is that it can
be known through ‘argument and reasoning’ (tarkenopapattya). This
was the atman of the Early Upanisads, that could be known by the
then accepted ways of knowing, that is by perceiving empirical in-
stances, by instruction, or by metaphysical reasoning or rationally
demonstrated to be unknowable in these ways.
(72) These ways of knowing are recognized in the Buddhist texts
which employ the same terminology to denote them. These terms
occur mostly in contexts which criticize these very Upanisadic doc-
trines of the atman. For example it is said that one should not regard
as the atman ‘what is seen, heard, thought of, understood or attained . . .’
( dittham sutam mutam vihhatam pattam . . ., M. 1 . 135). If we leave out
the last (pattam, which is a way of knowing recognized by the Middle
and Late Upanisadic thinkers, v . infra , 73, 74) we notice that the others
The Historical Background 6 1
are the same as the concepts occurring in the Upanisadic list. Likewise,
in the Suttanipata we find that the forms, ‘ dittha -, suta- y muta -’ often
used to denote the corresponding ways of knowing in the Upanisads
O'- Sn. 793, 79 8 > 8o2 > 8 i 3> 9 01 )-
(73) While perception, scripture, and reasoning were regarded as the
usual ways of knowing in this period, we find that the verbal forms
from \/drs- acquire a new meaning (other than that of sense-percep-
tion) in the Middle and Late Upanisads. The atman now has to be
directly seen but this cannot be done by means of perception 1 (v.
praptum sakyo na caksusa, i.e. one cannot attain it with the eye,
Katha, 2.3.12 cp. Katha, 2.3.9, na caksusa pasyati kascanainam, i.e.
no one sees it with the eye). Nor can it be had from the sacred
scriptures (nayam atma pravacanena labhyo, i.e. this soul is not to be
attained by means of scriptural instruction, Katha, 1.2.23). ‘Manifold
instruction’ is of no avail (na bahuna srutena , Katha 1.2.23 == Mund.
3.2.3). The mention of manifold instruction (srutena) as distinguished
from scripture (pravacanena) is probably a reference to the diverse
metaphysical theories about the atman in the Early Upanisads. ‘Nor
is this apprehension attainable by reasoning (naisa tarkena matir
apaneya, Katha, 1.2.9). The atman is ‘not to be reasoned about*
( atarkyah , Mait. 6.17, cp. Katha, 1.2.8, anlyan hy atarkyam anupra-
manat, i.e. for it is inconceivable being subtler than the subtle): it
‘cannot be had by the intellect ’ (labhyo na medhaya, , Katha, 1.2.23 =
Mund. 3.2.3). The traditional ways of knowing hitherto accepted are
discarded as far as the knowledge of the atman goes and ‘seeing*
acquires the new connotation of extrasensory perception. Thus the
atman which is hidden within all things and does not shine forth is
seen ( drsyate ) by the subtle seers by their subtle awakened intuition
(drsyate tvagryaya buddhya suksmaya suksmadarsibhih, Katha,
1.3. 1 2). One sees (jpasyate) while in meditative rapture (dhyayamanah)
by the purification of knowledge ( ' jnana-prasadena ) and not by any
of the sense-organs (Mund. 3.1.8). As the Svetasvatara puts it, one
would see (pasyet) God hidden as it were by practising the drill of
meditation (dhyana) (1.4). Here was a new way of knowing, un-
recognized in the earlier tradition, acquired by means of meditation
(dhyana = P. jhana) though the vision or revelation itself was said
1 Cp. ‘He is not grasped by the eye . . . nor by the other sense-organs ’ (na
caksusa grhyate . . . nanyair devaih, Katha, 3.1.8).
62
Early Buddhist Theory of Knowledge
to be due to the grace of atman or God. 1 Thus by the time of the Middle
or Late Upanisads the following ways of knowing appear to have been
recognized, viz. (i) normal perception, (2) extrasensory perception,
(3) scriptural or traditional authority and (4) reason.
(74) The word that is most frequently or almost invariably used to
denote the knowledge derived from this means of extrasensory
perception is jhana . Thus it is said that the atman is obtained ‘by right
knowledge’ (samyag-jnanena, Mund. 3.15) or by the ‘peace of know-
ledge’ (jnana-prasadena, Mund. 3.1.8) and those who obtain it are
‘satisfied with their knowledge’ (jnana-trptah, Mund. 3.2.5). This
atman or God as ‘knower’ is jnah (Svet. 6.2, 16, 17) and knowing God
or having the right knowledge is denoted by verbal forms of
(jnatva, Katha 1.2.16; 17; 2.3.8; Svet. 1.11,2.16, 3.10; jnatum, Katha
1. 2. 21). But the word jnana is not entirely confined to this usage for
at Katha 2.3.10 it is used in the plural to mean the ‘knowledge of the
five senses’ (yada pancavatisthante jhanani manasa saha, when the
five sense knowledges together with the mind cease). Likewise other
cognitive verbs are at times employed to denote the above sense but
their occurrence is sporadic and very rare: e.g. matva (Katha 1.2.22),
matih (Katha 1.2.9), viditva (Katha 2.1.2), viduh (Katha 2.3.9),
vidyam (Katha 2.3.18) and veda (Svet. 3.8).
(75) Although the contemplatives claimed a direct experience of
reality totally different in character from any kind of metaphysical
insight, it must be said that their description of these experiences is
not without interpretation and is bound up with a good deal of meta-
physics and theology. A knowledge of the Vedas was in theory no
more necessary than it was in the earlier metaphysical phase. 2 But
tradition could not be entirely done away with and particularly at a
1 There is a doubt whether dhatuh prasadat ought to be translated as ‘through
the grace of the Creator’, since Sankara interprets the phrase as ‘dhatusamprasa-
dat’, i.e. through the tranquillity of the senses, an interpretation which is sup-
ported by usage in this stratum of thought as Hume has shown (op. cit., p. 350,
fn. 1). But this does not alter the fact that it is to be conceived as a revelation as
well, since it is expressly stated that the atman reveals himself (Katha 1.2.23;
Mund 3.2.3.).
2 In this phase it said, for instance, that Janaka has no metaphysical knowledge
of what happens to him after death although he has ‘mastered the Vedas’ (adhlta-
vedah, Brh. 4.2.1); Svetaketu Aruneya returns proud and conceited ‘after studying
all the Vedas’ (sarvan vedan adhltya (Ch. 6.1.2) but without knowing the nature
of reality (loc. cit.).
The Historical Background 63
time when the growing influence of heterodox schools of thought was
felt, 1 it would have seemed desirable for the teachers of the Vedic
tradition to close their ranks and give a definite place to traditional
Vedic learning. This was done by the expedient of saying that there
were two kinds of knowledge, the higher and the lower and while
this new way of knowledge was regarded as the higher knowledge, the
Vedas are given a definite place in the scheme of things by calling it
lower knowledge. Thus the Mundaka says, ‘two kinds of knowledge
are to be known . . . the higher and the lower’ (dve vidye veditavye . . .
para oAvapara ca, Mund. 1.1.4), the lower being the study of the
Vedas and the ancillary sciences (op. cit. y 1.1.5) <an ^ the higher that
by which the imperishable is apprehended’ (atha para yaya tad
aksaram adhigamyate, loc. cit .). The use of the word vedanta (veda +
anta, the end or consummation of the Vedas) to denote this higher
knowledge (Mund. 3.2.6, Svet. 6.22) also reveals the same attitude of
maintaining the continuity with the Vedic tradition while regarding
this knowledge as final.
(76) Thus by the time of the Late Upanisads there were three main
schools of thought in the Vedic tradition. Firstly, there were the ortho-
dox brahmins who believed in the supernatural revelation of the Vedas
and held the Vedas to be the supreme source of all knowledge.
Secondly, there were the metaphysicians of the Early Upanisadic
period, who held that the highest knowledge was to be had by rational
argument and speculation based on their faith in or acceptance of
premises, which they believed in, either because they were tradition-
ally unquestioned or because there were rational or empirical grounds
for believing in them. Thirdly, there were the contemplatives, who
believed that the highest knowledge was personal and intuitional and
was to be had by an extrasensory perception, acquired partly by the
practice of a technique, though dependent ultimately on the will of
the atman or Isvara. Each of these forms of knowledge was believed
to result in salvation, so that salvation was conceived to be possible,
inter alia , (1) by one’s metaphysical beliefs, (2) reliance on scripture,
and (3) intuitional knowledge. When, therefore, the Buddha ‘says that
there is no salvation through metaphysical beliefs, revelation or
intuitional knowledge’ (na ditthiya, na sutiya, na nanena . . . visuddhim
aha, Sn. 839), in speaking to a brahmin, it is probable that he was
referring to the theories of the above three classes of thinkers ^cp.
1 Deussen, op. cit., p. 70.
64 Early Buddhist Theory of Knowledge
‘the sage does not say that they are “experts” by virtue of their (meta-
physical) beliefs, scriptural learning and intuitional knowledge’ (na
ditthiya , na sutiya, , na hanena , muni . . . kusala vadanti, Sn. 1078).
(77) During the close of this period we find in the Maitri Upanisad
the use of the word pramana (a valid means of knowledge) in a tech-
nical sense and a growing realization that our claims to knowledge
must be backed up by their being made on valid grounds. We talk
about time but how do we know that such a thing called time exists.
This Upanisad suggests that we measure or know time from observing
the movements of the sun across the constellations. It is said: ‘Because
of its subtlety this (course of the sun) is the proof for only in this way
is time proved (to exist)’ (sauksmyatvad etat pramanam anenaiva
pramlyate hi kalah, 6.14). This is followed by the significant statement
that ‘ without a valid means of knowledge there is no apprehension of
objects {lit. of what is to be proved)’ {na vina pramanena prameya-
syopalahdhih , loc . cit.). The importance attached to the study of the
pramanas or the valid means of knowledge (the central problem of
epistemology) in Indian thought may be gauged by the fact that every
school of thought, orthodox or heterodox had its theory of pramanas.
When the Greeks (Strabo) referred to Indian philosophers as the
‘pramanika’, 1 it is not clear whether this was a reference to all the
Indian philosophers at the time (of whom they were aware of), who
claimed to base their theories on valid means of knowledge or a class
of ‘epistemologists’, who made a study of the valid means of know-
ledge; in any case it shows the importance of pramanas for Indian
thinkers at this time, as confirmed by the reference in the Maitri
Upanisad. There is also a reference to pamanika in the Anguttara
Nikaya and since this is not far removed in time from the Maitri
Upanisad and the Greek reference, we may translate the term pamanika
as ‘epistemologists’ since it fits the context: ‘In this matter the epis-
temologists 2 (?) argue thus; this person and the other have identically
1 They are described as a class of brahmins ‘contentious and fond of argument’
called the Pramnai\ v. J. W. M’Crindle, Ancient India , p. 76. Cp. The Cambridge
History of India, Vol. I, p. 421, where E. R. Bevan, the author of the article says,
‘The people intended were undoubtedly the pramanikas , the followers of the
various philosophical systems, each of which has its own view as to what con-
stitutes pramana a “means of right knowledge”.’
2 The Corny, has ‘those who form judgments with regard to individuals,
judge, i.e. ought to weigh and consider’ (puggalesu pamana-gaha paminanti,
pametum tuletum arahanti, AA. V.53). The PTS. translation reads, ‘those who
measure thus measure . . .’ (G.S. V. 98).
The Historical Background 65
the same traits, why then is one of them (considered) inferior and the
other superior’ (Tatra . . . pamanika paminanti : ‘imassapi te’va
dhamma aparassapi te’va dhamma, kasma nesam eko hlno eko panlto
ti’ (A. V.140). The context indicates that pamanika here are a class of
people who judge the truth- value of a statement in the light of evi-
dence and is therefore strongly suggestive of the sense we have given
to it,
(78) In the above discussion we have assumed that Upanisadic
thought was known to Buddhism and has had an impact on it. The
problem of the relation between the Upanisads and Buddhism deserves
to be reviewed in respect of three questions, (1) the question as to
whether there was any contact between Buddhism and the Upanisads,
(2) if there was contact at what point (chronological) did it occur,
and (3) the question whether Buddhist thought can be considered as
a continuation of or a reaction against the main trends of Upanisadic
thought. We shall, of course, not attempt to answer any of these
questions here, but it is necessary to state that with regard to the first
question, we assume contra Thomas 1 that there was contact and the
knowledge that Buddhism shows of Upanisadic thought would we
believe justify our assumption. With regard to the second question
we find that while many scholars are inclined to place the rise of
Buddhism close to the period of the Katha Upanisad 2 others prefer a
date long after even the Late Upanisads (e.g. Maitri) had been com-
pleted. 3 We would prefer to date the rise of Buddhism somewhat
before the Maitri Upanisad, which we believe refers to a rising Buddhist
movement.
(79) Hume, while stating that ‘the usual date that is thus assigned to
the Upanishads is around 600 bc just prior to the rise of Buddhism’, 4
1 The History of Buddhist Thought , p. 90; most scholars, however, have
admitted influence of interaction — v. Hume, The Thirteen Principal Upanishads ,
pp. 6 f.; J. Przyluski and E. Lamotte, ‘Buddhisme et Upanisad’, BEFEO., Vol. 32,
1932, pp. 141-69; Helmuth von Glasenapp, ‘Vedanta und Buddhismus’, Abhand-
lungen der Geistes- und Sofialwissenschaftlichen Klasse Jahrgang 1950, pp. 1013 ff.
2 v. Oldenberg, Buddha, Calcutta, 1927, pp. 53-8; J. Charpentier, The
Kathaka Upanisad’, Indian Antiquary > Nov. 1928, p. 207; J. N. Rawson, The
Katha Upanisad , Oxford University Press, 1934, pp. 42-8.
3 E.g. Ranade and Belvalkar, who speak of the necessity of postulating a period
of thought-ferment between ‘the end of the Upanishadic movement and the
commencement of the Jain-Buddhistic movements’ ( History of Indian Philosophy
Vol. 2, Poona, 1927, p. 443).
4 The Thirteen Principal Upanishads , p. 6.
C
66 Early Buddhist Theory of Knowledge
finds ‘evidences of Buddhist influences’ 1 in the Brhadaranyaka, Mundaka
and Prasna Upanisads. 2 * The examples he gives are of a very doubtful
character and can hardly be considered evidence of influence. Thus
it is arbitrary to say that the karma doctrine at Brh. 3.2.13 is due to
Buddhist influence and the few philological affinities that he has shown
between the language of the early Buddhist texts and the Upanisads
ipso facto prove little. Deussen on the other hand has rather indirectly
suggested Buddhist influence on the Maitri Upanisad. He speaks of
‘the polemic against the heretics which is found in Maitr. 7.8-10’,*
and says that ‘Brahmanism in view of the consequences which the
attitude of the earlier Upanishads had entailed in Buddhism and similar
manifestations, returns to its original position’. 4 Later he states
more expressly that ‘in the Maitr. Upan. is revived the ancient Vedic
standpoint in regard to tapas in the presence of Buddhist and other
errors’. 5
(80) In the section in the Maitri Upanisad, where this polemic occurs,
there is a reference to a sect wearing a ‘ruddy robe’ (kasaya-, 7.8).
It is said that they convert their opponents by ‘rational arguments
and examples’ (tarkadrstanta-, loc . cit.), deny the doctrine of the soul
(nairatmyavada-, loc . cit.), call attention to ‘a dharma which is destruc-
tive of the Vedas and orthodox scriptures’ (vedadisastra-himsaka-
dharmabhidhyanam astv iti vadanti, 7.9) and ‘whose goal is the mere
attainment of pleasure’ (ratimatram phalam asya, loc . cit.). We may
take it that the reference could be either to the Materialists, the
Ajivakas, the Jains or the Buddhists. There would have been many
sects other than the Buddhists wearing the red robe, although the
Dhammapada seems to regard it almost as a distinctive feature of the
Buddhist monk where it is said that ‘he who dons the red robe, not
free from stain and lacking in restraint and truth is not worthy of the
red robe’ (anikkasavo kasavam yo vattham paridahessati, apeto dama-
saccena na so kasavam arahati, 9). Now the Materialists did not value
dharma : 6 but not only is dharma one of the central concepts of
Buddhism, the doctrine being known as ‘the dharma’ (= P. dhamma-,
M. I.37), the Buddha is known to the brahmins, according to the
evidence of the Pali Nikayas as a ‘teacher of the dharma’ (dhamma-
vadi-); we find the brahmin Assalayana (Skr. Asvalayana) saying:
1 The Thirteen Principal Upanishads , p. 6. 2 Op. cit., pp. 6, 7.
3 Deussen, op. cit., p. 65. 4 Ibid. 5 Op. cit., p. 70.
6 v. D. Chattopadhyaya, Lokayata, p. 228.
The Historical Background 67
‘the recluse Gotama is a dharma-vadin; dharma-vadins are difficult to
argue with . . (samano . . . Gotamo dhammavadi , dhammavadino
ca pana duppatimantiya bhavanti, M. II. 147). The Ajivakas and the
Jains may have also had their dharma, but they retained the doctrine
of the soul (even if their concept of a soul differed from that of the
Upanisads: the Jains ‘upheld the soul’, ayaval = Skr. atma-vadi,
Acarariga Sutra, 1.1.5) 1 and neither the Jain nor any of the Ajivaka
doctrines could be called a nairatmya-vada (a doctrine denying the
soul), by which Buddhism was known. The reference, ratimatram
phalam asya (mere pleasure is the fruit thereof) clearly rules out the
Jains and most of the Ajivakas, who were given to extreme forms of
ascetism and shunned pleasure; it seems to suggest the Materialists
but it should be noted that the Buddhists had a strong reputation of
being hedonists at this time. The Buddhist texts say that it was the
opinion of the ‘other religious teachers and wandering ascetics’ that
‘the recluses who are sons of the Sakyan live in luxury (lit. are addicted
to pleasure )’ (. . . anna-titthiya paribbajaka evam vadeyyum —
’Sukhallikanuyogam anuyutta samana Sakya-puttiya viharanti ti,
D. III. 1 30). This is confirmed by the Sutrakrtanga, where it is said
that ‘some men, Sramanas and Brahmanas, who ignore and deny these
true words, adhere (to their own tenets) and are given to pleasures
(SBE., Vol. 45, p. 236) and which according to the Corny, is a refer-
ence, among others, to the Buddhists (loc. cit ., fn. 2). Considering this
evidence the inference is unmistakable that the reference is to a rising
Buddhist movement and while the Maitri forbids the brahmins to
study what is not of the Veda (navaidikam adhiyitayam, 7.10), it is
noteworthy that there is much material in it which has a Buddhist
flavour. The instances are too many for us to discuss here, but mention
may be made of the contemplation of the organic substances of the
body. 2 Similarly, the concept of ‘the sheath of Brahma’ (brahma-
kosa -, 6.38), which of all the classical Upanisads appears only in the
Maitri, is known to Buddhism (e.g. kosani viceyya kevalani, dibbam
manusakan ca brahma-kosam , Sn. 525), but it will be noticed that while
1 The Ayaramga Sutta, ed. H. Jacobi, Part I — Text, London, 1882, p. 1.
2 The Upanisad mentions fifteen organic substances (1.3) while the Buddhist
texts mention thirty-one (M. I.57) and sometimes thirty-two (Khp. 2). The belief
that the earliest list contains thirty-two items is a common mistake; v. Warder,
Early Buddhism and Other Contemporary Systems, BSOAS., 1956, p. 51, fn. 1,
‘the stock list of thirty-two organic substances of the body — to which the brain
is appended ... as 33rd’.
68
Early Buddhist Theory of Knowledge
the Buddhist text speaks of three kosa-s, the Maitri has the concept of
a four-fold kosa- (caturjalam brahmakosam, loc . cit.).
(81) We may conclude from the above that the rise of Buddhism is
not far removed in time from, though it is prior to, the Maitri Upani*
sad.
CHAPTER II
THE HISTORICAL BACKGROUND II—
NON-VEDIC I— MATERIALISM
(82) The impact of Materialist thinking on the thought of the Canon
is strong (y. infra, 6yj) and it therefore seems desirable to study the
epistemological doctrines of the Materialists in so far as they seem to
have a bearing on the thought of the Canon.
(83) It is customary to regard Materialism as a heterodox school, which
grew up in violent opposition to Vedic thought, but our study of the
concept of Lokayata has shown that Materialist philosophy emerged
within the Brahmanical fold as part of its logical and metaphysical
(lokayata-) speculations. Even scholars, we notice, trace the origins of
Materialist thinking to the thought of the Early Upanisadic period.
(84) Das Gupta, followed in this respect by Jadunath Sinha 1 and Chat-
topadhyaya, 2 finds ‘references to the lokayata doctrine (by which he
means Materialism) ... in the Chandogya Upanisad VIII, 7, 8, where
Virocana . . . went away satisfied with the view that the self was identical
with the body’; 3 but this is not full-fledged Materialism since this
particular brand of dehatmavada , as Das Gupta calls it (Joe. cit.) y
entertains a belief in the after-life which is quite explicit (Ch. 8.8.5).
Ruben, on the other hand, as we have seen (v. supra, 22) traces the
origins of Materialism to the thought of Uddalaka even calling him a
Materialist. There are undoubtedly certain materialistic trends in the
thought of Uddalaka 4 but we must remember that his ultimate reality
Being (sat) has the quality of sentience (tad aiksata bahu syam, Ch.
6.2.3) an d creatures in some sense survive bodily death (Ch. 6. 9.2.3).
It nevertheless could have furnished the germs of a Materialist philo-
sophy if these inconsistencies were got rid of.
1 History of Indian Philosophy , Vol. I, p. 230. 2 Lokayata , p. 45.
3 A History of Indian Philosophy , Vol. Ill, p. 528.
A Note his materialist conception of the mind (manas), which is formed of the
finest essence of food (anna-) (Ch. 6.6.1).
70
Early Buddhist Theory of Knowledge
(85) It is perhaps not without significance that the second school of
Materialists (i.e. according to the interpretation of Silanka 1 ) mentioned
in the Sutrakrtanga (2.i.io=SBB. 2.1.22, Vol. 45, p. 343) makes use
of Uddalaka’s a priori premiss that ‘Being cannot come out of Non-
Being’ (katham asatah saj jayeta, Ch. 6.2.2) when it says, sato natthi
vinaso asato natthi sambhavo, which Jacobi translates, ‘what is, does
not perish, from nothing nothing comes’ and which literally means,
‘there is no destruction of Being, there is no origination of Non-Being’.
If so Uddalaka would be the father of Indian Materialism, in the sense
in which all Materialism, according to Burnet, 2 is said to depend on the
theory of Parmenides, who was the first philosopher of Being in
Greek thought to make use of this premiss.
(86) Finally, it will be noticed that the Materialists themselves seem
to trace their doctrines to the Early Upanisads when they quote a
statement attributed to Yajnavalkya in the Upanisads in support of
their doctrines. 3 It is significant that this same Yajnavalkya asks a
question which has materialistic implications when, comparing man to
a tree, he says, ‘a mortal when cut down by death by what root does he
grow up? For if with its roots they should pull up a tree, it would
not come into being again.’ 4 We find this same analogy used in the
Mahabharata where, as the context shows, the materialist conclusion is
clearly drawn: ‘If the root of a tree that is cut down does not grow up
again, though its seeds germinate, where is the person who having died
comes back again.’ 5 It is therefore highly significant when the com-
mentary to the Digha Nikaya speaking of the epistemic origins of the
materialistic theories says that some ‘accept Materialism on the basis
1 On Su . 2. 1. 10, Vol. II, fol. 17, sa ca Samkhyamatavalambi . . . Lokayatama-
tavalambl va nastiko, i.e. he depends on the Sankhya theory ... or is a nastika
depending on the Materialist theory.
2 Early Greek Philosophy , London and Edinburgh, 1892, pp. 194-5, Hegel,
Erdmann, Schwegler et aL traced the origins of idealistic thought to Parmenides
but Burnet says, ‘Parmenides is not as some have said the father of idealism. On
the contrary, all materialism depends upon his view’.
3 ‘tad ahuh, “ vijndnaghana evaitebhyo bhutebhyah samutthaya tanyevanuvinasyati
na pretya samjfidsti ” ti,’ Sarvadarsanasamgraha, by Sayana-Madhava, Ed. V. S.
Abhyankar, Second Edition, Poona, 1951, p. 5; the quotation is from Brh. 2.4.12.
4 yat samulam avrheyuh vrksam na punar abhavet martyah svin mrtyuna
vrknah kasman mulat prarohati, Brh. 3.9.28.6.
5 Chinnasya yadi vrksasya, na mulam pratirohati, bijanyasya prarohanti
mrtah kva punar esyati, Srfmanmahabharatam, Santiparva, 184.14, Ed. T. R.
Krishnacharya and T. R. Vyasacharya, Vol. 12, Bombay, 1907, p. 294.
The Historical Background 71
of such arguments as ‘beings are like tree-leaves (or trees and leaves),
which when they fall, do not grow up again’. 1
(87) If Materialism grew up as a product of the incipient rational
temper of this period, it is not surprising that references to this doc-
trine should be found by the time of the Katha Upanisad, which
mentions a class of people who hold ‘this is the world, there is no
other’ 2 and deny survival. 3 Later, in the Svetasvatara Upanisad, it had
to be reckoned among the theories current at that time, for here
reference is made to several speculative theories about the nature of
reality and one of them is the ‘doctrine of elements’ (bhutani, 1.2).
This may be identified either with the Materialist theory in the Buddhist
texts, which upheld the reality of the four elements, viz. earth, water,
fire and wind (catummahabhuta-), 4 or the Materialist doctrine men-
tioned in the Jain texts, which held that the five elements (pancama-
habbhuya, Su. 2.1. 10), viz. earth, water, fire, wind and air were alone
real and that all things were composed of them.
(88) Since we are concerned only with the epistemological theories
propounded and the nature of the arguments adduced in support of
their doctrines by the Materialists contemporary with the rise of
Buddhism, we shall confine ourselves to these aspects of their doctrines.
It is, however, difficult to determine with any degree of exactness what
portion or proportion of these doctrines could have been contemporary
with this period, since most of the informative accounts that we have
of the Materialists are of a later date. We would therefore adopt the
method of stating those doctrines, which we suspect have a bearing,
direct or indirect, on the thought of this period, even when the form
in which they are stated is comparatively late and then endeavour in
the light of the material available from the Early Jain and Buddhist
sources to sift what may be early from the late.
(89) When we consider the epistemological theories of the various
schools of Materialists, we find that with regard to the opinions and
theories held on the problem of the means of knowledge, it is possible
to classify them into three groups, viz. (1) those who upheld the
1 Yatha rukkhapannani patitani na puna viruhanti, evam satta ti adina
takkena va ucchedam ganhanti, DA. I.120.
2 ayam loko nasti para iti mani, 1.2.6.
3 yeyam prete vicikitsa manusye ’stityeke nayamastlti caike, 1.1.20.
A Cp. atta rupl catummahabhutiko, D. I.34; catummahabhutiko* yam puriso,
D - 1*55*
72 Early Buddhist Theory of Knowledge
validity of perception alone and denied inference and other forms of
knowledge, (2) those who upheld the validity and priority of percep-
tion, but admitted inference in a limited sense, denying other forms of
knowledge, (3) those who denied all means of knowledge including
perception.
(90) Group (1) is the best known. In Madhava’s (14th c.) Sarvadar-
sanasamgraha, which is the locus classicus for a clear and concise state-
ment of what appears to be the widest known school of materialism,
it is said that ‘this school holds that perception is the only source of
knowledge’ (pratyaksaikapramanavaditaya, p. 3). Earlier in the
Sarvasiddhantasamgraha of Sankara (8th c.) it is stated that according
to the Materialists ‘only the perceived exists, the unperceived does not
exist by reason of its never having been perceived’. 1 Perhaps, still
earlier, in a reference to materialistic doctrines, where the account is
not technical or elaborate, we find it said that the Materialists held
perception to be the only source of knowledge: ‘Understand, intelligent
One, that no one exists hereafter; regard not that which is beyond
the reach of your senses, but only that which is an object of percep-
tion’. 2 It is not possible to determine how early this reference in the
Ramayana could be but we find that the theory set forth here is
associated with the Materialists and is criticized in the Pali Nikayas
D. II.328, 330). Kassapa is here arguing with the ksatriya Payasi
(y. infra, 136-9), who has performed a series of experiments, all based
on the assumption that it is possible to verify the existence of a soul by
sense-experience. This shows indirectly that the Materialists repre-
sented, could only be satisfied by the evidence of the senses. The
argument of the Materialist is stated by his opponent as follows: ‘I do
not know this, I do not see this; therefore (tasma) it does not exist’
(Aham etam na janami, aham etam na passami, tasma tarn natthi ti,
loc . cit.). This is countered by the argument that it is wrong to infer
from ‘I do not see X’ that ‘X does not exist’. 3 An example is given of a
man born blind (jaccandho puriso) who says he cannot see black or
1 pratyaksagamyamevasti nastyadrstamadrstatah, 2.2.3. Pd. M. Rangacarya,
Madras, 1909, p. 5.
2 sa nasti param ity etat kuru buddhim mahamate, pratyaksam yat tad atistha
paroksam prsthatah kuru, Ayodhyakanda, 108.17, v. Ed. Srinivasa Sastri Katti,
Vol. 2, p. 992.
3 Cp. Cowell, SDS., p. 14, ‘when you deny the existence of an object on the
ground of its not being perceived, you yourself admit an inference of which
non-perception is the middle term’.
73
The Historical Background
white forms (kanhasukkani rupani), forms of various colours, the stars,
the sun or the moon and argues that since he does not see them, such
things do not exist. Payasi is made to admit that such things do exist
and that therefore the argument that what one does not see, does not
exist, is false. We saw in the above quotation from the Ramayana that
it was implied that there was no hereafter, since the hereafter is beyond
the reach of our senses (i.e. of perception) and therefore the hereafter
does not exist. It is the logic of this same argument that is assailed here.
The Materialist could, however, still maintain his case for perception
by arguing that even though one may be blind, visible objects exist
because they are perceived by others, whereas the other world is in
principle unobservable by anybody and therefore cannot be presumed
to exist. This objection is implied in the question that Payasi proceeds
to ask, viz. ‘who tells you, Kassapa, that the gods exist* (ko pan’etam
bhoto Kassapassa aroceti atthi deva . . ., loc. cit.). What is meant is that
in the case of physical objects we can go on the information of others
who have perceptual evidence of them, even if we are blind but in the
case of the hereafter we cannot expect such information since no one
can be presumed to have any perceptual evidence of its existence. This
is met by the rejoinder that the ‘other world cannot be observed
in the way he thinks by the human eye* (na kho . . . evam paraloko
datthabbo yatha tvam mannasi imina mamsacakkhuna, loc . cit.) but
that it is still observable by some by means of ‘clear, paranormal,
clairvoyant vision’ (dibbena cakkhuna visuddhena atikkantamanusa-
kena, loc. cit.). It is claimed that there are recluses and brahmins who
devote their lives to meditating in the forest and developing their
faculty of clairvoyant vision, whereby they can observe this world and
the next (imam eva lokam passanti param eva, loc. cit.). The Materialist
is not impressed by this argument since he repeats that he is still of the
former opinion, 1 presumably because he does not believe in the possi-
bility of extrasensory perception and further discussion on these lines
is dropped.
(91) It is, however, clear from the above that the Materialists at this
time attached great importance to perception as a means of knowledge,
1 Kinca’pi bhavam Kassapo evam aha, atha kho evam me ettha hoti; itipin’atthi
paraloko, n’atthi satta opapatika, N’atthi sukatadukkatanam kammanam, phalam
vipako ti, i.e. although the reverend Kassapa says so, I am still of the opinion
that there is no other world, no surviving beings, and no result or effect of good
or evil deeds. Later (D. I., p. 352) it is said that he was convinced by these
arguments.
C*
74
Early Buddhist Theory of Knowledge
even if it is not clear whether they did hold that it was the only means
of knowledge. It is not possible to ascertain whether the metaphysics
of materialism preceded its epistemology or vice versa but there is
undoubtedly an intimate connection between them. Silanka, the ninth-*
century 1 commentator on the Sutrakrtanga, seems to think that the
denial of the unverifiables on the part of the Materialist is based on
their epistemology and results from the acceptance of perception alone
as the only means of knowledge. Speaking of the Materialists he says,
‘they argue (pramanayanti) as follows: there is no soul apart from the
(material elements such as) earth, etc., because there is no means of
knowledge to apprehend it and the only means of knowledge is percep-
tion and not inference or the rest since with the latter there is no direct
sense-contact with the object and error results; as a result of error and
owing to the presence of obstacles they (i.e. these means of knowledge)
would be of a defective nature and one cannot have confidence in any
of them; it has been said that “one who runs on uneven ground groping
his way about (lit. trusting on his hands, feet, etc.) depending largely
on inference is bound to fair’ — this is the characteristic of inference,
scripture and the rest (of the means of knowledge) for (with them) one
has to move as it were by groping one’s way because there is no direct
contact with objects; therefore, perception is the only means of knowledge
(pratyaksamevaikam pramanam) and by means of it one cannot apprehend
a soul different from the elements and as for the consciousness (caitanyam)
found in their midst, it manifests itself only when the elements
come together in the form of a body like the intoxicating power when
the ingredients are mixed ’. 2 This passage tells us why the Materialists
relied only on perception and how their philosophical beliefs depend
on this. Perception is the only valid means of knowledge since the
others are liable to error, as there is no direct sense-contact with the
object in their case; therefore, there cannot be a self-identical soul since
1 Glasenapp, Der Jainismus, p. 107.
2 On Su. 1. 1. 8, Vol. I, fob 15, tatha (te) hi evam pramanayanti — na prthivyadi-
vyatirikta atma’sti, tadgrahakapramanabhavat, pramanam catra pratyaksameva,
nanumanadikam tatrendriyena saksadarthasya sambandhabhavadvyabhicarasam-
bhavah, sati ca vyabhicarasambhave sadrse cabadhasambhave dusitam syad iti
sarvatranasvasah, tathacoktam — ‘hastasparsadivandhena, visame pathi dhavata
anumanapradhanena, vinipato na durlabhah’ anumanam catropalaksanamagam-
adlnam api, saksadartha sambandhabhavaddhastasparsaneneva pravrttiriti,
tasmat pratyaksamevaikam pramanam, tena ca bhutavyatiriktasyatmano na
grahanam, yattu caitanyas temupalabhyate, tad bhutesv eva kayakaraparinatesv
abhivyajyate madyangesu samuditesu madasaktivad iti.
75
The Historical Background
one cannot perceive it. This shows that their metaphysical beliefs had
an epistemological basis according to this account of Silanka. It may be
seen that from the earliest times the more sceptical minded were
inclined to doubt or deny the existence of what they did not see. Much
of Rgvedic scepticism was based on this principle (v. supra, 7). It
would therefore not be implausible to suggest that the birth of the
Materialist philosophy in India may have taken place when the
principle that what one does not see does not exist, was more or less
systematically worked out. 1
(92) Whatever the views held by the Materialists contemporary with
or prior to Early Buddhism, there is every reason to believe that group
(1) taken as a whole denied the validity of inference altogether. This is
evident from the accounts given of the Materialist criticisms of
inference in Santaraksita’s Tattvasamgraha (1457-9), Kamalasila’s
Panjika (ibid.), Jayanta’sNyayamanjari 2 and the Sarvadarsanasamgraha
(Ch. I). In this respect group (2) is in agreement with group (3), which
also criticizes anumana (inference). A fairly adequate account of these
group (1) criticisms of inference have been given by Das Gupta 3 and
Jadunath Sinha 4 and we do not propose to repeat this here. A brief
account of the group (1) criticisms of inference as taught in the Nyaya
school is given by Radhakrishnan and Moore 5 although the criticisms
of Jayarasi Bhatta are specifically directed against the conceptions of
inference found both in the schools of the Nyaya as well as the Budd-
hists (v. infra, 105, 106). Of these accounts, Jayarasi’ s criticisms are the
most specific and elaborate while the simplest and the most general
account appears to be that given in the Sarvadarsanasamgraha. The
gist of the argument here is that inference cannot be shown to be a
valid mode of knowledge unless it can be proved that there are good
grounds for knowing the truth of universal propositions (vyapti) as
well as their necessity. Now universal propositions or universals cannot
be known by perception, for perception whether external or internal
(i.e. introspection) is of particulars, with which we are acquainted
through sense-experience or introspection. They cannot be claimed to
be known through inference for this would lead to infinite regress. It
cannot be testimony for this is either a form of inference or implies
1 Cp. those who did not believe in gods or sacrifices in the Rgveda (RV. 8,
70.7, 71.8; 10.38.3); these contexts mention the ‘godless man' (adevah).
2 Ed. Pandit S. S. Narayana Sukla, Benares, 1936, pp. 108, 109; v. anumanap-
ramanyaksepah.
3 HIP., Vol. Ill, 533 ff- * HIP., Vo1 * T > 2 35 ff - 5 °P- ciu > PP- *36-46.
7 6 Early Buddhist Theory of Knowledge
inference and if we accept testimony we would have to believe anything
whatsoever on authority. Nor can it be analogy (upamana) for this
merely relates a name to a thing named. Likewise, the necessary con-
nection between causes and effects asserted by universal propositions
cannot be established and the connection may very well be a coinci-
dence.
(93) It is obvious that even this simplest account is far too sophisticated
to have its roots in the period of Early Buddhism. Does this mean that
the criticism of inference is a later development and that in the earliest
period inference of some kind was admitted along with perception?
If the Materialists were among the first thinkers in India to argue and
thus develop the tarkasastra — it seems prima facie unlikely that they
would have discarded anumana so early, especially after realizing that it
was the mainstay of the hetusastra. They had to argue very sharply
against their opponents and they would have cut the ground beneath
their feet if they denied the logical basis of their reasoning altogether
and admitted its total invalidity. When we examine the reasoning
behind some of Payasi’s experiments ( y . infra, 136-9) we notice that
he makes use of inference, though it is inference based on sense-percep-
tion, despite his fundamental assumptions, namely that the soul is
visible or has weight being mistaken. Besides, the argument of the
Materialists is put by the Buddhist in the form i etam na janami etam na
passami, tasma tarn natthi’ and we have some reason to believe that the
phrase ‘janami passami’ is used in the Buddhist texts (v. infra , 783) to
denote ‘perception and inductive inference based on perception’
though Buddhism uses ‘perception’ in a wider sense to include extra-
sensory perception. If this is so, then in the context of the Materialist
this phrase should mean ‘sense-perception and inductive inference’.
In other words perception has priority as the basic means of knowing
though inference also plays a limited part when what is inferred is in
principle verifiable by sense-perception.
(94) Another reason for surmising that inference in this sense is
possibly a part of the early doctrine of Materialism of at least one of the
schools is that it appears to have been held by group (2), represented
by the views ascribed to Purandara in Kamalasila’s Tattvasamgraha-
panjika. 1 The statement ascribed to Purandara 2 is as follows:
1 1482-3, p. 41; v. Das Gupta, op. cit Vol. Ill, p. 536, fn. 2.
2 Tucci has shown that Purandara was a ‘Carvaka-mate granthakara" (an
author of a book on Materialism); v. ‘A Sketch of Indian Materialism", PIPC.,
1925, p. 36.
The Historical Background 77
Purandaras tv aha, lokaprasiddham anumanam Carvakairaplsyata eva
yattu kaiscillaukikam margamatikramya anumanamucyate tannisid-
dhyate, i.e. Purandara says that it is well known that even the
Materialists accept inference although they object to people (kaiscit)
employing inference beyond the limits of sense-perception {lit. beyond
the path of this world). This view attributed to Purandara is confirmed
by the references made to him by the Jain commentator Vadideva Suri
who, as Das Gupta has pointed out, 1 quotes in his commentary
Syadvadaratnakara (II.131) on his Pramananayatattvalokalankara a
sutra of Purandara, viz. pramanasya gaunatvad anumanadarthanis-
cayadurlabhat, i.e. from the very nature of this means of knowledge, it
is difficult to determine (the existence of transcendent) objects by means
of inference(P). The sense of this sutra is however made clear by
Vadideva’ s comment, laukikahetunam anumeyavagame nimittam sa
nasti tantrasiddhesv iti na tebhyah paroksarthavagamo nyayyo ‘ta
idamuktam anumanad arthaniscayo durlabhah. Das Gupta’s translation
of this passage appears to be more of the nature of a commentary than
a translation. 2 We may translate it more or less literally as follows:
‘since in transcendent proofs (tantrasiddhesu, lit . what is proved in
religious texts) the basis for inference is absent unlike in the case of
perceptual inferences, a knowledge of transcendent objects cannot be
had (nyayyo, lit. inferred) by them; therefore has it been said that “it is
difficult to determine (the existence of transcendent) objects by means
of inference”.’ This shows the existence of a school of Materialists who
admitted perception and empirical inference but discarded metaphysical
inference on the grounds that what was in principle unperceivable was
unknowable. For a valid inference to be possible, it is necessary to
establish the truth of a universal proposition (vyapti), which reveals a
concomitance between a hetu and an anumeya (=sadhya-, cp. laukik-
ahetunam anumeyavagame) and this is not possible unless both are in
principle observables. It is difficult to say whether this school asserted
that there was a necessary connection between cause and effect or
merely held that the concomitance or sequence was only probable and
therefore the inference was only probable. It is worth noting that one
of the objections against inference brought out by group (1) was that
1 Op. cit., Vol. Ill, p. 536, fn. 2.
2 ‘Thus since in the supposed supra-sensuous transcendent world no case
of hetu agreeing with the presence of its sadhya can be observed, no induc-
tive generalization or law of concomitance can be made relating to that
sphere/
78
Early Buddhist Theory of Knowledge
there was no necessity in the concomitance, despite our repeated
observation of several instances. As Jayanta says in elucidating the
theories of the Materialists, ‘universality (vyapti) is not established
even by the observation of several instances (bhuyodar§anagamya’pi)
since there is the possibility of error even after observing a thousand
instances: though we come to the conclusion that smoke and fire are
concomitant (sahacari, lit . go together) by observing several instances
we cannot know that there is no smoke in the absence of fire despite
this repeated observation’. 1 But whether they made this latter qualifica-
tion (which, incidentally, is the same as the objection that Hume raised
against causation and inference 2 ) or not, it is clear that in limiting the
inferable to the sphere of the verifiable, they were tacitly assuming the
truth of the Verification Principle 3 and it is therefore this group rather
than group (3) (y. supra , 89) which deserves to be called a positivist 4
school of thought. Purandara’s statement that it is well known
(lokaprasiddham, lit, known the world over) that the materialists
accepted inference does not make sense unless they or the majority of
them had in fact accepted the validity of both perception and inference
in the above sense up to that time. That the reference to this school is
not confined to Purandara’s statements and their exposition is evident
from the reference made to it by the Jain commentator Gunaratna who
in his Tarkarahasyadlpika commenting on the phrase, manam tvaksaja-
meva hi, in verse 83 of the Saddarsanasamuccaya says that ‘the particle
“hi” in this phrase is added to denote a distinction, the distinction
being that at times (kvacana) the Carvakas welcome inferences such as
“smoke” (implies fire) which are limited to stating what is within the
reach of the world but not transcendent inferences (alaukikamanu-
manam) which (claim to) establish (the existence of) heaven, what is
1 bhuyodarSanagamya’pi na vyaptir avakalpate sahasraso’pi taddrste vyabhica-
ravadharanat bhuyo drstva ca dhumo’gnisahacarltigamyatam anagnau tu sa
nastlti na bhuyodarsanadgatih, Nyayamanjarl, p. 109.
2 A Treatise of Human Nature, Vol. I, Part III, Section XIV. Cp. p. 169. ‘If
we define a cause to be an object precedent and contiguous to one another and
where all the objects resembling the former are placed in a like relation of priority
and contiguity to those objects that resemble the latter; we may easily conceive
that there is no absolute nor metaphysical necessity, that every beginning of
existence should be attended with such an object/
3 Warnock, English Philosophy since 1900, pp. 44 ff.
4 Basham speaks of ‘the positivism of Ajita’ (op. cit p. 271) but he does not
clarify his usage. Warder (v. infra , 97) uses the term of group (3) but as we have
shown this is quite unjustifiable.
The Historical Background 79
unseen (adrsta), etc.’ 1 The fact that Vadideva Suri and Gunaratna are
Jains and the other reference was in a Buddhist work shows that at
least the Buddhist and Jain tradition was well aware of the existence of
this school. When we consider this in the light of what we know of the
Materialists from the Early Buddhist and Jain sources it seems probable
that these early Materialists or at least one school among them believed
in the validity of both perception and inference while giving priority
to perception and restricting inference within the limits of the verifiable.
(95) The third group of Materialists, as classified according to their
epistemological theories, is represented by Jayarasi Bhatta’s Tattvo-
paplavasimha, which is the only extant authentic text of the Materialists
(lokayata). Although this work was published in 1940, very few
scholars seem to have taken note of it. Ruben ^ 1954), 2 Jadunath Sinha
(1:956) 3 and Sharma (i960) 4 make no reference to it in discussing the
philosophy of the Materialists and Chattopadhyaya (1959), who
professes to make a specialized study of lokayata-, 5 begins his book by
lamenting the lack of any treatise of this school.
(96) The Tattvopaplavasimha refers to another work of the same
school, the Laksanasara (p. 20) or the ‘Essence of Definition (?)’,
which may be his own work since after criticizing two of the charac-
teristics of perception (avyabhicari, vyavasayatmakam) according to
the Nyaya definition (N.S. 1.1.4) he refers the reader to the above
work for the criticism of the other characteristic (avyapadesyam). As
the editors have pointed out (pp. iii, iv), the reference in Sri Harsa’s
Khandanakhandakhadya to a school of the Lokayatas, which like the
Madhyamika school of Buddhism and the school of Sankara is said
to have denied the validity of all means of knowledge (pramanas) is
most probably a reference to this school.
1 P. 306, hi sabdo’tra visesanartho vartate, visesah punas Carvakairlokayatra-
nirvahanapravanam dhumadyanumanam isyate kvacana, na punah svargadrstadi-
prasadhakam alaukikam anumanam iti.
2 Op. cit.; however, he has more recently written an article on this subject
entitled, ‘Uber den Tattvopaplavasimha des Jayarasi Bhatta eine Agnostizistische
Erkenntniskritik , appearing in, ‘Wiener Zeitschrift fur die Kunde Siid-und
Ostasiens und Archiv fur indische Philosophic, Band IP, 1958, pp. 140-53.
3 Op. cit.
A C. D. Sharma, A Critical Survey of Indian Philosophy, London, i960.
5 Op. cit., p. 6, ‘. . . in the ocean of uncertainty concerning the lost Lokayata
the only piece of definite information is that we are left with no original work
on it\
80 Early Buddhist Theory of Knowledge
(97) The interest of this school for us lies in the fact that it seems to
throw some light on a school of thought mentioned in the Nikayas,
which represented a standpoint of absolute nihilism or logical scepti-
cism in rejecting all views but which at the same time has been called
an ‘annihilationist’ (ucchedavada-) or Materialist school ( v . infra y 335).
It also appears to explain a certain usage ( v . infra y 116) of a phrase
attributed to the early Materialists which would otherwise be inexplic-
able. Warder has seen in this branch of Lokayata philosophers ‘not
materialists but positivists according to modern ideas’ (op. cit. y p. 52)
and says that ‘we may perhaps connect Jayarasi Bhatta’s theory with
Sanjaya Belatthiputta in the Samannaphala Sutta, which, however, is
stated as merely agnostic or sceptic without positivist content’ (ibid.,
p. 53). He calls Jayarasi Bhatta ‘a positivist’ (ibid.) and this branch of
Lokayata as ‘the positivist branch’ (ibid.) which rejected perception
whereas the ‘ordinary Lokayata as described in the Tattvasamgraha
and elsewhere allows perception as the only means of cognition’ (ibid.).
He adds that this ‘positivist trend may have been a later development
in the Lokayata- Carvaka school rather than an early rival branch of
Barhaspatya’ (ibid.). The editors of this text have also expressed the
view that this work ‘carries to its logical end the sceptical tendency of
the Carvaka school’ (p. i) and have raised the question as to whether
the author of this work is a mere sophist who has no views of his own,
although they themselves do not think so (p. xiii).
(98) We may state at the outset that we do not agree with Warder’s
assessment of this philosophy as positivism and our objection is not
that he is, as he says at the end of his paper, applying ‘modern philoso-
phical terms’ to ‘ancient doctrines’ (v. op. cit. y p. 62). Nor can we see
much of a connection between this philosophy and the views expressed
by the sceptic Sanjaya. And since this early school of absolute nihilists
or logical sceptics, who have also been called materialists, seems to
contain the basic concepts of this philosophy we are more inclined to
entertain the possibility that the germinal ideas or the roots of this
school go back to the period of the Pali Nikayas and that this school
was possibly an early rival branch of the other school which at that time
accepted at least the validity of perception if not of inference as well.
(99) Since this work has been untranslated 1 and largely ignored since
its publication it seems desirable to give a brief account of its nature
1 Except for a brief extract of the criticism of anumana in the Nyaya school,
given in Radhakrishnan and Moore (op. cit. y pp. 236-46).
The Historical Background 81
and contents before we form any conclusions about it. The work
claims to be, as the editors have shown (pp. xi, xii) a text of the
Lokayata school. The author quotes Brhaspati who is sometimes
mentioned by name with great respect (pp. 45, 88) and is once called
the Sutrakara- (p. 79). In the second paragraph of his work he cites
the proposition ‘earth, water, fire, air are the real elements (tattvani);
by their combination (arise) the body, the senses, objects and con-
sciousness’ 1 which Gunaratna in his Tarkarahasyadipika quotes 2 as the
statement of Vacaspati (—Brhaspati). He also quotes with approval
the sayings ‘the worldly path should be followed . . . fools and the
wise are alike in the eyes of the world’ 3 which he attributes to the
wisest of men (paramarthavidbhih).
(100) And now begins the problem. He speaks of the tattvas (four
elements) of the Lokayata, but shows that we have no grounds for
affirming that they are real. He uses an epistemological argument:
‘We can talk of a means of knowledge (mana) only if it is valid
(sallaksananibandhanam manavyavasthanam, lit. the determination of
a means of knowledge depends on its having the characteristic of
existence) and the proof of the (existence of the) objects of knowledge
(meyasthitih) depends on the means of knowledge but if there is no
means of knowledge (tadabhave) how can we speak of the real existence
of both (objects as well as means of knowledge)’. 4 This is not claimed
to be a disproof of Brhaspati’s proposition (quoted above) for it is
said that in asserting that earth, etc., were tattvas he was indirectly
referring to (pratibimbanartham, lit. reflecting) the fact that if even
what is widely accepted as real does not bear critical examination
(vicaryamanani na vyatisthante), then what of other things (kim
punar anyani). But this is surely a departure from the materialist thesis,
for how can a person who does not believe in the objective existence
1 prthivyapastejovayuriti tattvani tatsamudaye sarlrendriyavisayasamjna, p. 1.
2 Gunaratna’s quotation adds caitanya as a by-product of the rest, yaduvaca
Vacaspatih, prthivyapastejovayuriti tattvani tatsamudaye Sariravisayendriyasa-
mjna, tebhyascaitanyam, op. cit p. 307. It may also be noticed that visaya- is
placed before indriya- in this. The addition of caitanya- strongly suggests
that this was the view of the school which admitted an emergent atman con-
sidered a by-product, which Silanka distinguishes from the other school (v.
infra, 115).
3 laukiko margo’nusarttavyah . . . lokavyavaharamprati sadrSau balapanditau,
P* r -
* sallaksananibandhanam manavyavasthanam, mananibandhana ca meyasthitih,
tadabhave tayoh sadvyavaharavisayatvam katham . . . p. 1.
82
Early Buddhist Theory of Knowledge
of the material elements be called a materialist? JayaraSi, therefore,
cannot be reckoned a materialist, as far as his theory of the external
world is concerned. He seems to deny the real existence of both this
world as well as the next, in denying the reality of all tattvas and his
work as its name implies is intended to ‘upset all principles’ (tattva-
upaplava-) epistemological as well as ontological, and he claims to
have done so at the end of his work (tadevam upaplutesu tattvesu,
P- I2 5 >-
(101) His epistemological argument (assuming that his disproof of
the pramanas is valid) only goes to prove that we do not or cannot
know that there are real objects of knowledge and not that such objects
do not exist. In other words, his argument should have led him to
scepticism and not to nihilism. But it is important to observe that
nowhere in his work does he claim to be a sceptic or grant the possi-
bility of the existence of things even if he cannot know them. On the
contrary he even uses metaphysical arguments (v. infra , 104) to dis-
prove the existence of the soul. He is therefore not a sceptic but an
absolute nihilist in his metaphysics though he may be called a logical
sceptic in so far as he is sceptical of (i.e. doubts or denies) the possi-
bility of knowledge.
(102) Though he is not a materialist, we may perhaps concede that he
shows a certain partiality for materialism in that he seems to imply that
the material tattvas have a greater claim to reality by the common
consent of the world (loke prasiddhani, p. 1). On pragmatic grounds
(vyavaharah kriyate) he says that we should believe in the existence of
the body, of physical objects (ghatadau) and of pleasure (sukha-)
(p. 1) and recommends as a wise saying that the ‘way of the world
(laukiko margah) should be followed’. As he thus seems to recommend
the materialist way of life, we may call him a pragmatic but not a
metaphysical materialist.
(103) Jayarasi’s work is almost exclusively devoted to epistemology,
if not for a brief section in which he criticizes the atman-theories of
Nyaya (pp. 74-8), Jainism (pp. 78-9), Sankhya (pp. 79-81), Vedanta
(pp. 81-2) and Mimamsa (pp. 82-3). We shall translate a section in
which he criticizes a Vedanta theory (not that of Sankara) of the atman
since this would throw some light on the nature of his reasoning and
the question as to whether he is a positivist. He is criticizing the theory
that the soul (or pure ego) is of a blissful nature (anandarupam) and is
The Historical Background S3
absolute (kaivalyam). He proceeds as follows: 1 ‘Those who posit the
blissful nature and absoluteness of the soul do not speak with reason.
Why? If the soul has a blissful nature and it is introspective (svasam-
vedyam), then this experience will be present (prasaktam) in its sam-
saric state, in which case the effort to attain salvation is futile. If on
the other hand this is not experienced in the samsaric state, the soul
would have the nature of being enveloped with primeval defilements.
Just as a jar when concealed under a cloth is not recognized as a jar,
the soul when smeared with defilements is not known as a soul but this
(argument) is false, since there is no congruity (vaisamyat) between
the instance and the example. In the case of our not recognizing the jar
when it is hidden beneath the cloth there is no contact of the jar with
the organ of sense, owing to the cloth concealing it and in its absence
the sense-cognition of the jar does not take place. But here in the case
of what is covered with defilements, what is concealed? The conceal-
ment of the experiencer and the object of experience cannot take place
for the experiencer (vedaka-) and the experienced (vedyam) are of the
nature of the soul. As in the case of the consciousness (vijnanam) of the
Buddhists (Bauddhanam),it is experienced in the presence or the absence
of objects. Since introspectibility is of the nature of the soul, it is ex-
perienced in the presence or the absence of defilements, as owing to the
inactivity (akincitkaratvat) of the defilements the soul persists as a dif-
ferent object. But if the defilements are identical with the soul, then in
saying “the defilements are removed”, are you not saying that the soul
is removed, in which case it can be objected that there is no salvation!’ 2
1 Ye’pi anandarupam atmanah, kaivalyam abhidadhati te’pi yuktivadino na
bhavanti. Katham? Yady atmanah anandarupam svasamvedyan ca, tada sam-
saravasthayam api tat vedyam prasaktam; tatas ca moksarthaprayaso nisphalah.
Atha samsaravasthayam na vedyate anadimalavagunthitam atmanah svarupam,
yatha patantarite ghate ghatabuddhir na bhavati, evam malalipte atmani atma-
buddhir na bhavati; tad etad ayuktam, drstantadarstantikayoh vaisamyat —
patantarite ghate patabuddhir na bhavati patantardhane sati indriyena sakam
sambandho nasti tadbhavad ghate nendriyajam vijnanam sampadyate. Iha tu
punah malavagunthanena kasya vyavadhanam kriyate? na vedyavedakayor
vyavadhanam kriyate. Vedyam vedakan ca atmasvarupam eva — yatha Baudd-
hanam svasam vedyam vijnanam tac ca visayasadbhave 'pi vedyate tadabhave'pi
vedyate, yathatmanah, svasamvedyam svarupam malasadbhave’pi vedyate
tadasadbhave’pi vedyate, malasyakincitkaratvad atmano ‘rthantaratvenavas-
thanat. Atha tadatmyena sthitani malani; tada'malany apaniyante' kimuktam
bhavati? Atma’panlyate, tatas ca moksabhavaprasangah, pp. 81, 82.
2 For a similar argument, v. Sthiramati, Madhyantavibhagatlka, Edition par
Susumu Yamaguchi, Nagoya, 1934* PP* 60 ff.
84 Early Buddhist Theory of Knowledge
(104) In all his criticisms of the atman- theories, Jayarasi is employing
dialectical arguments to disprove his opponent’s thesis. In the above
instance, it would be seen that he takes up the proposition p (atma
anandarupam), put forward by his opponent. He then says that p im-
plies the truth either of q (atma anandarupam svasamvedyam) or not-q
(atma anandarupam avedyam). Both lead to contradictions showing
that p is false, q implies r (moksarthaprayaso nisphalah), which con-
tradicts one of the propositions or assumptions of his opponent’s
system. Not-q is self-contradictory, since q is self-evident (cp. vedyam
vedyakan ca atmasvarupam eva). Therefore, his opponent’s thesis p,
is false. They are not the arguments of a positivist, 1 who wishes to
show that no meaning can be attached to the concept of an atman
(soul) and hence it should be dispensed with, but the kind of argument
that any metaphysician would employ against (to use a phrase of F. H.
Bradley) 2 his ‘brother metaphysician’.
(105) The rest of Jayarasi’s work is devoted entirely to the discussion
of epistemological topics. He criticizes theories of the validity of
knowledge put forward by the Mlmamsa school (pp. 22-7) and the
Buddhists (pp. 22-32). Almost half the work is concerned with the
criticism of the validity of perception as upheld in the Nyaya (pp. 2-22),
by the Buddhists (pp. 32-58), in the Mlmamsa (pp. 58-61) and the
Sankhya (pp. 61-4). It is followed by the criticism of the theory of
inference (anumana) in the Nyaya (pp. 64-74) and of the Buddhists
(pp. 83-109). The concluding section is a criticism of knowledge based
on authority (sabda-, aptokti-), where the apauruseya- theory (pp.
116-20) and the views of the grammarians (pp. 120-5) are discussed.
A page or two is devoted to the criticism of the arthapatti (presump-
tion) — theory of Mlmamsa, as well as upamana (comparison) and
abhava (negation) as means of knowledge (pp. 109-13). Sambhava-
(inclusion) and aitihya- (report) are dismissed in two sentences (p. 1 13),
the former being subsumed under inference (anumana) and the latter
under scriptural tradition (agama). The space devoted to each possibly
reflects to some extent the importance attached to these theories at the
period in which he wrote, but we cannot fail to observe that he begins
his work with the criticism of perception and then only goes on to
discuss the problems of the validity of knowledge in general. Consider-
ing also the space allotted to the criticisms of perception one gets the
1 Cp. Wittgenstein, Tractatus Logico-Philosophicus , p. 151, Sections 5.631-3.
2 Appearance and Reality, Second Edition, London, 1906, p.i, ‘He is a brother
metaphysician with a rival theory of first principles/
The Historical Background 85
impression that the author thought that if perception, which was
universally accepted and that even by some of the Materialist schools,
was demolished, it goes without saying that no means of knowledge
could be valid.
(106) His technique of argument is throughout the same and his
criticisms are almost invariably directed against some specific theory.
We may illustrate this by taking a few passages. In criticizing the
Buddhist theory of perception, he takes up the definition of perception
in the Nyayabindu (1.4), viz. ‘perception is free of construction
(kalpana’ podham) and is incorrigible (abhrantam)’. 1 He argues as
follows: ‘One should not say this since the sense of “free” (apoha)
in the phrase “free of construction” is not to be found. Then is kalpana
itself to be excluded (apohya)? What is kalpana? Is the consciousness
that arises with qualifications of quality, motion, species, etc., kalpana,
or is kalpana the consciousness that produces memory or is it of the
nature of memory, or does it arise from memory, or is it a reflection
of the contact with speech, or is it the apprehension of speech, or is it
of an unclear nature, or has it the nature of apprehending unreal objects,
or is it itself unreal or is it the seeing of objects accompanied by
inference (trirupallinga- 2 gato’rthadrs), or is it a reflection of objects
past or future?’ 3
1 Pratyaksam kalpanapodham abhrantam, p. 32.
2 Lit. ‘the middle term which has the three characteristics (of a valid syllogism)’
viz. ‘the existence (of the middle term) in the probandum, in what is like the
probandum and its absence in what is not (like the probandumf (anumeye’tha
tattulye sadbhavo nastitasati), v. Randle, Indian Logic in the Early Schools , pp.
181 ff.; cp. H. N. Randle, Fragments from Dihnaga , London, 1926, pp. 22-5;
also, Bochenski, Formale Logik , p. 503, 53.10 and 53.11. He gives the formal
rules as
(1) M is present in S (the fire on the mountain)
(2) M is present in XP (there is smoke in the kitchen which has fire)
(3) M is not present in X-Not-P (there is no smoke in the lake which has no fire).
3 Iti na vaktavyam, kalpanapodhapadasya apohyarthasambhavat. Nanu
kalpanaiva apohya? Ke’yam kalpana? Kim gunacalanajatyadivisesanotpaditam
vijnanam kalpana, aho smrtyutpadakam vijnanam kalpana, smrtirupam va,
smrtyutpadyam va, abhilapasamsarganirbhaso va, abhilapavatl pratltir va
kalpana, aspastakara, va, atattvikarthagrhitirupa, va, svayam va’tattviki,
trirupallirigagato’ rthadrgva, atltanagatarthanirbhasa va?
Tad yadi gunacalanajatyadi visesanotpaditam vijnanam kalpana; tat kim
avidyamanagunacalanajatyadivisesanotpadyatvena kalpana, uta vidyamanot-
padyatvena? Tad yadi avidyamanagunacalanajatyadivisesanotpadyatvena
kalpanatvam tad ayuktam; avidyamanasya janakatvabhavad eva akalpanatvam.
Atha vidyamanagunacalanajatyadivisesanotpadyatvena kaplana, tat kim savisayarp
86
Early Buddhist Theory of Knowledge
‘Now if kalpana is the consciousness that arises with the qualifica-
tions of quality, motion, species, etc., does it arise from unmanifest
qualifications of quality motion, species, etc., or from manifest (qualifi-
cations) ? If it is from unmanifest qualifications, it will not do, as there
would be no kalpana since it cannot be produced by what is unmanifest.
But if kalpana arises from manifest qualifications of quality, motion,
species, etc., does the knowledge of kalpana (kalpanajnanam) have an
object or have no object? If the knowledge of kalpana has an object,
it will not do, for even when it produces quality, motion, species, etc.,
it will not be kalpana, owing to the incongruity of the presence of an
appropriate object. If kalpana has no object, then the absence of an
object being itself the cause of kalpana, there would not arise the
qualifications of quality, motion, species, etc. If it is without an object,
then there would be no knowledge of kalpana, nor knowledge of
no kalpana (akalpanajnanam) but pure knowledge. If kalpana has
the nature of knowledge, all knowledge would be knowledge of
kalpana.’
‘Now if kalpana is the knowledge that produces memory, it will not
do, for memory arises even from the seeing of forms, etc., and that is
not kalpana.’
(107) It will be noticed from the above that Jayarasi’s method of attack
consists in taking the concept of kalpana, suggesting various alternative
definitions, showing that some of these definitions (e.g. smrtyutpada-
kam jnanam kalpana) do not apply, while others (e.g. gunacalana-
jatyadivisesanotpaditam vijnanam kalpana) lead to contradictions. The
concept of kalpana is therefore presumed to be self-contradictory and
a definition which contains this concept is untenable. Since the best
definitions of perception are all untenable, it is assumed that no true
account of perception is possible and therefore perception as a means
of knowledge is invalid.
(108) The criticism of the Nyaya account of perception proceeds on
similar lines. The author takes up the definition of perception as given
kalpanajnanam, nirvisayam va? Tad yadi savisayam sat kalpanajnanam, tad ayuk-
tam, gunacalanajatyadijanyatve’pi na kalpanatvam arthasamarthyasamudbhavat-
vasyanativrtteh. Atha nirvisayam sat kalpana tada nirvisayatvam eva kalpanatve
karanam na gunacalanajatyadivisesanajanyatvam; yadi ca tan nirvisayam, tada
na kalpanajnanam, napyakalpanajnanam, jnanamatrata syat, jnanatmataya ca
kalpanatve sarvam jnanam, kalpanajnanam syat.
Atha smrtyutpadakam jnanam kalpana, tad ayuktam, rupadidarsanad api smftir
utpadyate, na ca kalpanatvam. Op. cit., pp. 32-3.
The Historical Background 87
in the Nyayasutra (1.1.4), viz. ‘Perception arises from contact between
sense-organ and object, is determinate (avyapadesyam), non-erroneous
(avyabhicari) and non-erratic (vyavasayatmakamy. 1 Jayarasi argues
as follows: 2 ‘It is non-erroneous (avyabhicari) . . . (the text is here
defective and words have been omitted) . . . Does its non-erroneous-
ness consist in its arising from an abundance of non-defective causes or
in the absence of obstacles or in the efficiency of the process or in any
other way? If its non-erroneousness arises from an abundance of non-
defective causes, in what way is the non-defective nature of the causes
known? It is not from perception, since the proficiency of the eye, etc.,
is beyond the grasp of the senses. Nor is it from inference, since one
does not apprehend a basis for inference (lingantara-). Is not (then) this
very knowledge the basis, which gives rise to the knowledge of its
excellence? If so, the mutual dependence results in a difficulty. And
what is it? A suspicion of defect in a cognition which arises in depen-
dence on the virtues and defects of the senses, is not dispelled as in the
case of a consciousness of sound produced by the effort of a person’.
(109) Jayarasi seems to have picked out the characteristic of avyabhi-
cari despite the fact that avyapadesya- occurs earlier in the definition
in order to spotlight the fact that since perception cannot be shown to
be non-erroneous it must be erroneous. This he demonstrates by
suggesting different senses of avyabhicari and arguing that the truth
of none of them can be established.
(no) We may now draw our conclusions. The term ‘positivism’ has
been applied to characterize the philosophy of Comte and his successors
because of their anxiety to rid philosophy of speculative elements and
have its basis in the data and methods of the natural sciences. 3 Empiri-
cists like Hume and Mach have been called positivists because of their
forthright rejection of metaphysics and attempt to confine philosophy
1 indriyarthasannikarsotpannam jnanamavyapadesyamavyabhicari vyavasayat-
makam pratyaksam, p. 2.
7 Tac cavyabhicari . . . kim adustakarakasandohotpadyatvena, ahosvid badhara-
hitatvena, pravrttisamarthyena, anyatha va? Tad yady adustakarakasandohot-
padyatvena avyabhicaritvam, saiva karananam adustata kenavagamyate? Na
pratyaksena, nayanakusalader atlndriyatvat; napyanumanena lirigantaranavagateh.
Nanu idam eva jnanam lingam taduttham tasya visistatam gamayati; yady evam
itaretarasrayatvam duruttaram apanipadyate. Kinca indriyanam gunadosasrayatve
tadutthe vijnane dosasanka nativartate pumvyaparotpaditasabdavijnana iva, op.
cit., p. 2.
3 Baldwin, Dictionary of Philosophy and Psychology, s.y. Positivism.
88
Early Buddhist Theory of Knowledge
to the results of observation. More recently, the term has been used of
the philosophy of the Logical Positivists 1 (the Vienna Circle, Witt-
genstein of the Tractatus Logico-Philosophicus , Ayer of Language ,
Truth and Logic) who while rejecting metaphysics have broken away
from the narrow empiricism and psychological atomism of positivists
like Mach and Hume and have endeavoured to base their positivism on
logical foundations. 2 The term is also sometimes loosely employed to
refer to the modern Analytical Philosophers, who are really the succes-
sors of Russell and Moore. None of these positivisits have attempted
to disprove the validity of perception and inference by metaphysical
arguments as Jayarasi does. On the contrary, they have been anxious
to preserve the validity of perception and inference as recognized
methods of knowing in the natural sciences, although they have tried
to rid these concepts of speculative assumptions and linguistic con-
fusions. The only point of comparison that we can see is that like
Jayarasi the modern positivist will also say that there are no ultimate
‘tattvas’ in a metaphysical sense, but the latter would not try to deny
or disprove their existence and would merely hold that assertions
about such super-sensuous realities are strictly meaningless. We cannot
therefore agree with Warder’s description of Jayarasi’s school as
‘positivists according to modern ideas’.
(hi) The anxiety on the part of the positivist to save science and
eliminate metaphysics led him to formulate the Verification Principle,
the acceptance of which almost became some time ago the hallmark of
a positivist. When we observe that the second group of Materialists
(group (2)) did almost the same for similar reasons in trying to dis-
tinguish between empirical or verifiable inference and unverifiable or
metaphysical inference (supra, 94), it is this school which best deserves
to be called the positivist school in Indian thought.
(112) Nor can we see the connection that Warder sees between
Jayarasi’s theory and the thought of Sanjaya. The most we can say is
that if Jayarasi’s denial of knowledge led him to scepticism rather than
to nihilism, as it ought to have, then we may have argued that it was
possibly similar to the grounds on which Sanjaya accepted scepticism,
though we have no evidence whatsoever as regards the basis of the
latter’s Scepticism. All that we do know was that Sanjaya was a Sceptic,
1 v. Warnock, English Philosophy since 1900, Ch. IV. Warnock uses the term
‘positivist’ of the Logical Positivist (v. pp. 56, 58, 60).
2 v. Ayer, Language , Truth and Logic , pp. 136 ff.
The Historical Background 89
who granted the possibility of transcendent truths (e.g. hoti Tathagato
parammarana, etc., v. infra, 176) without denying them outright and
this is a sufficient ground for us to distinguish between the philosophies
of the two.
(113) The suggestion made by the editors of this text and by Warder
that this school of thought represents a later trend which carried to its
logical end the sceptical tendency of the Lokayata school cannot
entirely be put aside. Once the validity of inference was denied, as it
was, at some time, in the main branch of this school, it is evident that
perception could not stand for long on its own feet. Besides, it is clear
that Jayarasi is criticizing the views prevalent during his time and from
these criticisms alone we cannot deduce that there was a primitive core
of beliefs in this school, which go back to earlier times. But when we
find a reference in the Pali Nikayas to the existence of a school of
Lokayatikas, who were absolute nihilists and who probably denied
the truth of all views, it raises a strong presumption as to whether we
should not trace the origins of the school of Jayarasi to an early rival
branch of the other realist school or schools of Materialism.
(114) As we have already seen ( v . supra, 57, 58), in the Samyutta
Nikaya there is a mention of two brahmins, called lokayatika, who
interview the Buddha. One of the views that they hold is that ‘nothing
exists’, which according to the Corny, was a Materialist view 1 ( v .
supra, 59).
(114A) The view that ‘nothing exists’ is in fact occasionally mentioned
elsewhere in the Nikayas in contrast to its opposite, namely that
‘everything exists’ (sabbam atthi), both of which are said to be two
extreme views, which the Buddha following the middle way avoids. 2
In a similar manner is juxtaposed the ‘view of personal immortality’
(bhavaditthi) and the ‘annihilationist view’ (vibhavaditthi). 3 It there-
fore seems reasonable to suppose that the view that ‘nothing exists’ is
also a v/Mavaditthi. Now this latter term seems to denote the Materialist
philosophies mentioned at D. 1 . 34, 5, all of which are said to ‘posit the
cutting off (ucchedam), the destruction (vinasam) and the annihilation
(1 vibhavam ) of the person’ 4 . This means that it is very probable that the
1 Sabbam n’atthi sabbam puthuttan ti, ima dve uccheda ditthiyo ti veditabba,
SA. II.76. 2 sabbam atthiti eko anto sabbam natthlti dutiyo anto, S. 11.76*
3 Dve’ma ditthiyo bhavaditthi ca vibhavaditthi ca. Ye . . . bhavaditthim
alllna . . . vibhavaditthiya te pativiruddha. Ye vibhavaditthim alllna . . . bhavadit-
thiya te pativiruddha, M. I. p. 65.
* sattassa ucchedam vinasam vibhavam pannapenti, D. 1 . 34.
90 Early Buddhist Theory of Knowledge
theory that ‘nothing exists’ was either one of or was closely associate*}
with the Materialist theories at the time of the Pali Nikayas. ,
(115) If so, what could these two lokayata-materialist 1 views, one
holding that ‘nothing exists’ and the other that ‘everything is a
plurality’, be? We may identify the second with that school of
Materialists who upheld the reality of the elements, which is repre-
sented by Ajita Kesakambali in the Nikayas who speaks of the existence
of at least the four elements, earth, water, fire and air (D. I.55) and
which appears to be similar to if not identical with the first school of
materialists 2 propounding a theory called ‘the-soul-is-the-same-as-the-
body’ theory (tajjivataccharlra-) in the Sutrakrtanga (2.1.9=2.1.19,
SBE., Vol. 45, p. 342). According to Silanka’s interpretation (v.
supra , 85) there is another school of Materialists mentioned in the
Sutrakrtanga at 2.1. 10 (=2.1.21, 22, SBE., ibid.) which speaks of five
elements, including ether (akasa). If Silanka’s identification is correct,
this latter theory clearly brings out the plurality and the reality of
elements, which are described as uncreated (animmiya, akada), eternal
(sasata) and independent substances (animmavita, no kittima, avanjha).
If the identity of the pluralist school with one of these schools is
correct, then the other lokayata theory, which denied the reality of all
things looks very much similar to the absolute nihilism of Jayarasi.
(116) Haribhadra in his Saddarsanasamuccaya speaks of the lokayatas 5
(lokayatah) being of the opinion that ‘this world extends only as far as
what is amenable to sense-perception ’. 4 From this one may argue that
‘lokayata-’ means ‘what pertains to this world’ or the ‘philosophy of
this- worldliness or materialism’ as Chattopadhyaya has done . 5 We
cannot agree that this was the original meaning of the word ( V . supra 9
65, 66) but there is no reason to doubt that at least one of the schools
of the Materialists believed in the reality of this world and it is signifi-
cant that the Materialist theory referred to in the Katha Upanisad
(1.2.6) speaks of the existence of this world and the denial of the next,
ayam loko, nasti para iti, which is translated by Hume as ‘This is the
world! There is no other!’ (op. cit.> p. 346) and by Radhakrishnan as
1 lokayata- is used in other senses and lokayatika- for non-materialist views as
well in the Nikayas, v. supra , 59.
2 I.e. on the basis of the language used to describe them, v. infra.
3 Op. cit., verse 80, lokayata vadantyevam . . ., p. 301.
4 etavaneva loko' yam yavanindriyagocarah, op. cit ., verse 81, p. 301.
5 Op. cit., p. 3, ‘Thus Lokayata meant not only the philosophy of the people
but also the philosophy of this worldliness or materialism/
The Historical Background 91
‘this world exists, there is no other’ (PU., p. 610). Now Medhatithi
defining haitukah at Manu. IV.30 (op. cit ., Vol. I, p. 342) asserts that
the nastikas upheld the doctrines of ‘nasti paraloko, nasti dattam,
nasti hutam iti’, i.e. ‘there is no next world, no (value in) giving, no
(value in) sacrifice’. But the theory of the Materialists as defined in the
Pali Nikayas is somewhat different. Whilst mentioning ‘natthi dinnam
natthi hutam’ (— Skr. nasti dattam, nasti hutam), it also has the phrase,
‘ natthi ayam loko y natthi paro loko’ (D. I.55, M. III.71). Prof. Rhys
Davids has translated this phrase as ‘there is no such thing as this
world or the next’ but the phrase as it stands literally means ‘this
world does not exist, the next world does not exist’. This has always
presented a problem for while it is well known that the lokayata-
materialists denied the existence of the next world, it appears to be
strange that they should be spoken of as denying the existence of this
world as well, particularly when they were elsewhere supposed to
affirm positively the existence of this world. It is the discovery of the
philosophy of Jayarasi which makes it possible for the first time to see
that there was a lokayata-materialist school which denied the existence
of this world as well.
(117) We have, however, to face the problem as to how this theory,
which denies the existence of this world as well as the next, comes to be
associated with Ajita, who is represented as believing in the reality of
the four elements. Was Ajita also a pragmatic Materialist like Jayarasi?
The more probable explanation seems to be that the Buddhists identi-
fied all the known materialist views with Ajita, who symbolizes the
philosophy of Materialism, inconsistently putting together the tenets
of mutually opposed schools since they both (or all) happened to be
in some sense (metaphysical or pragmatic) materialists. This is also
possibly the reason why Ajita, while propounding the theory of the
four elements (catummahabhutiko’yam puriso) like the first school of
Materialists, mentioned in the Brahmajala Sutta (D. I.34, ayam atta
rupl catummahabhutiko . . .) also inconsistently speaks of the existence
of akasa (akasam indriyani samkamanti).
(1 1 8) The above evidence seems to point to the existence of at least
two schools of lokayata-materialists, the pluralist school of meta-
physical materialists, who believed in the reality of the elements and
denied only the existence of a next world and the nihilist school of
pragmatic materialists, who denied the reality of this world as well as
the next. Since the materialist philosophies (in India) as a whole and
92 Early Buddhist Theory of Knowledge
Jayarasi’s lokayata in particular seem to be based on epistemological
foundations, it seems not unlikely that this early nihilist school of
lokayata was a product of an epistemological nihilism, which denied
even the validity of perception and paved the way for the birth of
philosophical Scepticism, which almost immediately succeeds it. There
is good reason to believe that the early lokayata speculations were
closely associated with the study of reasoning or the cultivation of the
tarka-sastra and lokayata-materialism seems to have been an offshoot
of lokayata speculations in general, which were a branch of brahmanical
studies at one time (v. supra , 65). It is therefore very probable that it
was this same school of nihilist lokayata, which is represented as a
school of logical sceptics in the Dlghanakha Sutta (M. I.497-501),
which denied the truth of all views, since a representative of this school
is called a materialist (ucchedavado, v. infra, 121) in the commentary
(MA. III. 204) and as we have shown there is textual evidence to
confirm this view (v. infra, 334). In the light of the evidence we have
cited, it is difficult to resist the conclusion that this early lokayata school
of absolute nihilists, logical sceptics and pragmatic materialists were
the precursors of the philosophy of Jayarasi and were in time at least
contemporaneous with the existence of the Pali Nikayas.
(119) Dr Warder says that ‘another materialist school seems to have
appeared among the kings themselves and especially their ministers,
including perhaps the celebrated Vassakara of Magadha, who in the
Anguttara Nikaya, Vol. II, p. 172, expresses a realist view in conformity
with Arthasastra Lokayata’ (op. cit., p. 55). But the context hardly
warrants such a grandiose conclusion. Here Vassakara says that he
holds the following view: ‘If he who speaks of what he has seen as
“thus I have seen”, there is nothing wrong in it ... of what he has
heard as “thus I have heard” ... of what he has sensed as “thus I have
sensed” ... of what he has understood as “thus I have understood”,
there is nothing wrong in it’. The Buddha does not wholly agree with
this point of view and says that one should not assert even what one
has seen, heard, sensed or known, if it is likely to be morally undesir-
able. The Buddha makes the same point elsewhere (M. I.395) where
he says that one’s speech should not only be true but also morally
useful (atthasamhitam) and not morally harmful (anatthasamhitam).
Vassakara on the other hand seems to be satisfied if someone states and
confines himself to the bare truth, as he has experienced it, irrespective
of its moral consequences. This is not the doctrine of the Arthasastra,
The Historical Background 93
which recommends the utterance even of untruths for the sake of
political expediency but appears to be his own personal view. The
context is ethical and one can hardly draw epistemological or philoso-
phical conclusions from it, especially since Vassakara’s statement is
compatible with any philosophical standpoint, idealist, phenomenalist
or realist. The fact that Vassakara as an important Magadhan official
may have studied the Arthasastra and the Arthasastra gives a naive
realistic account of the world has, in our opinion, little to do with the
view expressed here.
(120) Whatever differences existed among the Materialists on epis-
temological matters they seem to have all agreed in criticizing the
authority of the Vedas and the argument from authority. This probably
goes back to the earliest times. In fact, the original stimulus in the
genesis of the Materialist philosophy may have been provided by the
dissatisfaction with the Vedic tradition at a time when those who
would not still break with tradition found they could no longer agree
with the old traditional knowledge and sought to replace acceptance
of tradition and revelation with metaphysical inquiry. The statement
attributed to the Materialists in the Sarvadarsanasamgraha that ‘the
impostors, who call themselves Vedic pundits are mutually destructive,
as the authority of the jnanakanda (section on knowledge) is over-
thrown by those who maintain the authority of the karmakanda
(section on ritual), while those, who maintain the authority of the
jnanakanda reject that of the karmakanda’, 1 may have a history that
goes back to the earliest phase of Materialism, though this particular
criticism itself would not have been possible at least before the termina-
tion of the Early Upanisadic period for it was probably at this time
that the original Vedas as well the traditional lore including the
Upanisads (v. Brh. 2.4.10, 4.5.1 1) are said to have been breathed forth
by the Supreme Being.
(121) According to the Sarvadarsanasamgraha, the Materialists criti-
cized the sruti or the revelational tradition as a valid means of know-
ledge on the grounds that the Veda is ‘invalidated by the defects of
falsity, contradiction and repetition’ (anrtavyaghatapunaruktadosair-
dusitataya, p. 4). When therefore the Nyaya Sutra very much earlier
says that ‘(according to some the Veda) is unreliable since it has the
defects of falsity, contradiction and repetition’ (tadapramanyamanrtavy-
aghatapunaruktadosebhyah, 2.1.58) using identical language it is
1 The Sarvadarsanasamgrahah , Trans. E. B. Cowell, London, 1882, p. 4*
94 Early Buddhist Theory of Knowledge
probably trying to meet the criticisms of the Materialists in particular,
although this view was shared by the other heterodox schools as well.
At the time of the Pali Nikayas we find the statement attributed to
Ajita, the Materialist, who says that ‘those who uphold the atthikavada
are making a false and baseless lament’. 1 Here the criticism is limited
to the defect of falsity, probably because the term atthikavada, is, in
this context, used in a wide sense to denote not only the traditional
philosophy of the Vedas but the philosophies of those heterodox sects
as well, which believed in the concepts of soul, survival, moral respon-
sibility or salvation. The common factor of these heterodox schools
barring the Materialists was the belief in survival; 2 so the absence of a
belief in survival is taken to be the defining characteristic of a Material-
ist, who is as a consequence called one who subscribes to the natthi-
kavada in the Pali Nikayas. This is clear from the use of the term
natthikavada in the Appannaka Sutta (M. I.403), where it is employed
to denote the theory that ‘there is no next world’ (natthi paro loko,
M. I.403) and we observe the following polarities of usage:
micchaditthi natthikavado, M. I.403
„ akiriyavado, M. I.406
„ ahetuvado, M. I.408
sammaditthi atthikavado, M. I.404
„ kiriyavado, M. I.407
,, hetuvado, M. I.409
When therefore these terms are employed together, e.g. ahetuvada,
akiriyavada, natthikavada (M. III.78, A. II.31), they are not to be
treated as synonyms but as variants of micchaditthi. 3 Atthikavada-,
therefore, as used by Ajita has a wide connotation and we cannot
presume that his criticism is limited to the Vedas though it certainly
would have included it.
(122) Let us now examine the kind of argument that the Materialist
during the time of the Pali Nikayas used in defending or proving his
own beliefs and in criticizing the theories of others. We have for this
1 Tesam tuccham musa vilapo, ye keci atthikavadam vadanti, M. I.515.
2 Even the Sceptics seem to have believed in survival in a pragmatic sense
(y. infra, 163).
3 The definition of natthika- as a ‘sceptic, nihilist’, of natthikaditthi- as
‘scepticism, nihilistic view, heresy’ and natthikavada- as ‘one who professes a
nihilistic doctrine’ in the PTS. Dictionary (j.v.) is inacurate and misleading, in
the context of the Pali Canon.
The Historical Background 95
purpose to rely mainly on the account given of the Materialist schools
in the Sutrakrtanga, since the Pali Nikayas (and their commentaries)
which briefly state the doctrines of the Materialist schools but not the
reasoning behind them, are not very informative on this subject.
(123) As for the nihilist school of lokayata-materialists, we have no
more information than what we have stated elsewhere (v. supra , 112;
infra, 333, 334). As we have said they appear to have been logical
sceptics, who denied the truth of all views, probably on epistemological
grounds since there was no means of knowing anything, as even the
validity of perception could not be relied on. The school on which we
have the most information seems to be the positivist school (v. group
( 2 ), supra , hi), which upheld the priority of perception without
denying empirical or verifiable inference. But before we deal with
them it is necessary to dispose of another school of Materialists, which,
if Silanka’s interpretation is correct, relied on metaphysical or a priori
arguments to construct their thesis of Materialism.
(124) The second book of the Sutrakrtanga speaks of four kinds of
people representing four types of philosophies. Of these ‘the first kind
of man is the person, who asserts that the-soul-is-the-same-as-the-
body’ (padhame purisajae tajjivatacchariraetti, Su. 2.1.9), 1 i.e. the
Materialists who identified the soul with the body. This seems to be the
same as the first of the seven schools of Materialists mentioned in the
Brahmajala Sutta, which asserts that the ‘soul is of the form of the body
and is composed of the four great elements’ (atta rupl catummaha-
bhutiko, D. I.34). It was also probably the philosophy attributed to
Ajita, who speaks of the ‘four elements composing the person’
(catummahabhutiko ayam puriso, D. I.55) and much of whose
language is in common with the account of the Sutrakrtanga, 2 though
as we have suggested, doctrines attributed to him seem to be of a
composite character ( y . supra, 117). It also appears to be the minimum
doctrine accepted on pragmatic grounds by the nihilist school of
materialists as well. 3
1 The reference in Jacobi’s translation is different (SBE., Vol. 45). It will be
mentioned where relevant.
2 Cp. Ard. Mag. kavotavannani atthini bhavanti with P. kapotakani atthini
bhavanti and Ard. Mag. asandlpancama purisa gamam paccagacchanti with P.
asandipancama purisa matam adaya gacchanti.
3 v. Jayarasi’s quotation, prthivyapastejoyvayuriti tattvani (op. cit ., p. 1) and
the Buddha’s statement to Dighanakha the materialist and logical sceptic, ayam
. . . kayo rupi catummahabhutiko , M. I.500.
96 Early Buddhist Theory of Knowledge
(125) Now, it is said that ‘the second kind of person is one who asserts
the existence of the five great elements’ (docce purisajae pancama-
habbhutietti, Su. 2.1. 10). This would appear to be a second school of
Materialists, asserting the reality of the five elements including akasa
(akase pancame mahabbhute, loc. cit.) if not for a qualification made
towards the middle of this passage and the fact that we were led from
the context to expect a different kind of philosophy. The Ardhama-
gadhi text reads as follows: pudhavi ege mahabbhute, aii ducce maha-
bbhute . . . iccete pancamahabbhuya animmiya . . . satanta sasata
ayacchattha, puna ege evam ahu- sato natthi vinaso asato natthi
sambhavo (loc. cit.). The presence of the word ayacchattha- seems to
mean that the person who held the reality of the five elements also
believed in the reality of the atman as a sixth element, in which case
this is not a Materialist philosophy at all and the passage may be
translated as follows: ‘Earth is the first element, water the second
element . . . thus these five elements are uncreated . . . independent and
eternal with atman as the sixth (element); further, some say that,
“there is no destruction of that which is and no origination of that
which is not”. Jacobi translates differently following Silanka 1 taking
“puna ege evam ahu” with the previous sentence as follows: ‘Earth
is the first element, water the second element . . . These five elements
are not created ... are independent of directing cause or everything
else, they are eternal. Some say, however, that there is a self besides the
five elements. What is, does not perish; from nothing, nothing comes’
(SBE., Vol. 45, p. 343). This translation is permissible though it devi-
ates from the form in which the text is printed, but it does not solve the
problem for it means that this passage is introducing not one but two
theories, one a Materialist theory and the other a Realist theory which
asserts the substantial existence of the soul as well. Silanka, as we pointed
out (v. supra , 85), interprets the two theories as the Lokayata (lokayata-
mata-) and the Sankhya respectively. He distinguishes this lokayata
from the former which he calls Tajjivatacchariravada 2 following the
Sutrakrtanga though however he still considers this a species of
lokayata. 3
1 Tadevambhutani pancamahabhutanyatmasastani punareke evamahuh, op. cit.,
Vol. II, fol. 18 on Su. 2.1 .10.
2 Ayanca prathamo purusastajjlvataccharlravadT, op. cit., Vol. II, fol. 17 on
Su. 2.1.9.
3 v. Te caivamvidhastajjivataccharlravadino lokayatikah, op. cit., Vol. II, fob
16 on Su. 2.1.9.
The Historical Background 97
(126) If we accept this dualist interpretation and that one of the
theories spoken of is Materialism, it is necessary to reconcile ourselves
to the fact that it seems to be different from that of the nihilist and the
empiricist schools in that it is a product of pure reasoning. The belief
in the plurality of the elements is probably grounded on some such
premiss as ‘what is distinguishable is separable in reality 5 . 1 Since we
can distinguish between the qualities of earth, fire, etc., they have a
separate reality. Now each of the real elements being real must have
the characteristics of Being. That which is real cannot be destroyed
since ‘there is no destruction of Being 5 (sato natthi vinaso, loc . cit .); so
each of the elements is indestructible and hence eternal (sasata) and
without end (anihana=Skr. anidhanah). Likewise since ‘nothing can
come from Non-Being 5 (asato natthi sambhavo), they must have had
the quality of Being for all time; so that these elements could not have
been created directly or indirectly (animmiya animmavita akada no
kittima no kadaga, loc, cit,) and hence have no beginning (anaiya,
apurohita). Again, each of these elements cannot affect the other
elements for in such a case there would be loss of their Being and
hence they are independent (satanta= Skr. svatantrah) substances
(avanjha, i.e. not void, being plenums and not vacuums like the atoms
of Democritus). This rational Atomistic Materialist school seems
therefore to have made considerable use of Uddalaka’s a priori premiss
(v, supra , 25) that ‘Being cannot come out of Non-being 5 much in the
same way in which Empedocles and the Greek Atomists, Leucippus
and Democritus, made use of Parmenides 5 a priori reasoning about
Being in the history of Greek thought. 2 The only reference to this
school outside Silanka 3 that we have been able to find is by
Gunaratna, who after describing the nastikas who ‘spoke of
the world being composed of the four elements 5 (caturbhu-
tatmakam jagadacaksate, op, cit., p. 300) says: ‘But some who are
somewhat like the Carvakas (Carvakaikadeslyah) think that akasa
is the fifth element and speak of the world as being composed
1 Hume makes good use of this premiss or principle (as he calls it) in a different
connection, v. op. cit., p. 35. What consists of parts is distinguishable into them
and what is distinguishable is separable. Cp. p. 32.
2 v. J. Burnet, Greek Philosophy, London, 1943, pp. 69, 95, 197.
3 Silanka mentions a school of Materialists who believed in akasa as the fifth
element even when he is commenting on the first school as follows: kesancillo-
kayatikanamakasasyapi bhutattvenabhyupagamadbhutapancako’panyaso na dosa-
yeti, since some Materialists consider ether as an element the reference to five
elements is not wrong, op. cit., Vol. II, fol. 16 on Su. 2.1.9.
D
98 Early Buddhist Theory of Knowledge
of five elements’. 1 But whether they grew out of the deductive
rational basis of their doctrines and fell in line with the epistem-
ological outlook of the other schools, we cannot say in the
absence of any evidence and we have therefore not included this
school in our classification of Materialist theories according to their
epistemological doctrines, particularly since Silanka’s interpretation of
the passage in the Sutrakrtanga itself is doubtful.
(127) The other Positivist school of Materialists appears to have been
the more vigorous and the better known, since it seems to have
made a strong impact on the epistemological theories of Early Budd-
hism. Most of the later accounts of this school take it for granted that
its Materialist beliefs are a product of its epistemology. We have
already quoted the views of Sllanka, who was of the opinion that since
the Materialists held perception to be the only source of knowledge
they disbelieved in the existence of a soul ( v . supra , 91). Gunaratna
says the same: ‘therefore, the soul, good and evil and their fruits,
heaven and hell, etc., which others speak of, do not exist since they are
not perceived (apratyaksatvat)’. 2
(128) The beliefs attributed to Ajita Kesakambali are precisely these
and we may presume that they were arrived at by this principle of
empirical reasoning, which as stated in the Nikayas was of the form,
aham etam na janami, aham etam na passami, tasma tarn natthi, ‘I do
not know and see this, therefore it does not exist’. Ajita’s beliefs are as
follows:
(i) There is no soul . ‘A person is composed of the four elements*
(catummahabhutiko ayam puriso).
(ii) There is no value in morals or religious practices . ‘There is (no
value) in sacrifice or prayer (natthi yittham, natthi hutam)’, ‘there is
(no value) in giving (natthi dinnam)’; ‘there are no good and evil
actions, which bear fruit’ (natthi sukatadukkatanam kammanam
phalam vipako); ‘there are no (obligations to) one’s parents’ (natthi
mata, natthi pita).
(129) In holding sense-perception to be the ultimate basis of know-
ledge they seem to have criticized not only the claims to the authority
1 Kecittu Carvakaikadeslya akasam pancamahabhutam abhimanyamanah,
pancab hutatmakam jagaditi nigadanti, op. cit., p. 300.
2 Tato yatpare jlvam punyapape tatphalam svarganarakadikam ca prahuh,
tannasti, apratyaksatvat, op. cit. } p. 302.
99
The Historical Background
of the Vedic scriptures ( v . supra, 121) but the claims to extrasensory
perception or higher intuition (abhinna) on the part of some of the
religious teachers of their times. This seems to be the significance of
Ajita’s remark that ‘there are no well behaved recluses and brahmins
of good conduct, who can claim to know the existence of this world as
well as the next by realizing this themselves with their higher intuition*
(natthi loke samanabrahmana sammaggata sammapatipanna ye
iman ca lokam paran ca lokam sayam abhinna sacchikatva pave-
denti, loc . cit.).
(130) That empiricism was the keynote of their arguments is evident,
when we examine the few arguments of the first school of the Material-
ists recorded in the Sutrakrtanga. One of the arguments is that you
cannot observe a soul separate from the body and therefore there is no
soul apart from the body. The inference is directly drawn from obser-
vation and is inductive: ‘As a man draws a sword from the scabbard
(kosid asim abhinivvattittd ) and shows it saying, “this is the sword and
that is the scabbard” (ayam . . . asi ayam kosi), so nobody can draw
(the soul from the body) and show (it saying), “friend, this is the soul
and that is the body” (ayam . . . aya iyam sariram). As a man draws a
fibre from the stalk of munja grass (muni ad isiyam) and shows it saying,
“this is the stalk and that is the fibre” (ayam . . . munje iyam isiyam ) . . .V
We have underlined these examples given to illustrate the fact that the
argument may have been suggested by what their opponents who held
that ‘the soul was different from the body’ 1 2 were claiming. For, in the
Katha Upanisad (2.3.17) it is stated that ‘one should draw up from
one’s own body the inner-atman (antaratman) like a fibre from a stalk
of munja grass’ 3 (antaratma . . . tarn svaccharirat pravrhen muhjad
ivesikam ). This was possibly the subjective experience of a Yogin. The
Buddhists while not committing themselves on this question as to
whether the body was identical with the soul or was different from it
since it is one of ‘the things on which no definite view was expressed’
1 Se jahanamae kei purise kosio asim abhinivvattittanam uvadamsejja ayamaiiso
asi ayam kosi, evam eva natthi kei purise abhinivvattittanam uvadamsettaro,
ayamaiiso aya iyam sariram. Se jahanamae kei purise munjao isiyam abhinivvattitta
nam uvadamsejja, ayamaiiso munje iyam isiyam . . . Su. 2.1.9, Vol. 2, fol. 11.
2 anno jivo annam sariram, ibid. Cp. annam jivam annam sariram, Ud. 67,
where it is a theory put forward and debated by some recluses and brahmins.
3 Radhakrishnan has mistranslated the phrase munjadivesikam as ‘the wind
from the reed’ (PU., p. 647).
IOO
Early Buddhist Theory of Knowledge
(maya anekamsika dhamma desita, D. I.191), themselves claimed that
in certain jhanic states one could ‘create psychic selves out of this body*
(imamha kaya annam kayam abhinimminati . . . manomayam, D. I.77;
cp. the attapatilabhas 1 or ‘the attainment of selves’, D. 1 . 195), where
this ‘self’ (kaya-, attapatilabha-) appears to be different from the body
in the same way in which ‘the stalk of munja grass is separate from the
fibre, the stalk being the one thing and the fibre another, although the
fibre is pulled out of the stalk’ ( ayam muhjo ayam isika, anno muhjo
anna isika, munjamha tveva isika pavalha) (D. I.77) or in the same way
in which ‘a sword is different from the scabbard, the sword being one
thing and the scabbard another, although the sword is drawn from
the scabbard’ ( ayam asi ayam kosi , anno asi anno kosi, kosiya tv’eva asi
pavalho , loc . cit.). When, therefore, we consider the context of this
argument it would appear that the Materialists were questioning and
contesting the objective validity of these claims on the ground that
one could not demonstrate for all to see that such a soul or ‘self’ was
different from the body, since such claims could not be verified from
sense-experience.
(13 1) The importance that the Materialists attached to verification in
the light of sense-experience is brought out in these arguments. The
point of the above argument seems to be that no meaning can be
attached to the concept of ‘different from’ unless it was possible to
observe a soul as separate from the body in this verifiable sense of
‘difference’. In the other argument the importance of verifiability is
more explicitly brought out. One cannot speak of the existence of the
soul unless the soul is verifiable by sense-experience and since no such
soul is perceived, it is those who say that it does not exist (asante) or
it is not evident (asamvijjamane) who would be making the ‘right
statement’ (suyakkhayam==Skr. svakhyatam) about it. This argument
seems to have had its repercussions in Buddhism, where the Buddha
appears to be making a similar criticism of the concept of Brahma,
(v, infra , 550, 552) and we may state it fully following Jacobi’s transla-
tion: ‘Those who maintain that the soul is something different from
the body cannot tell whether the soul (as separated from the body) is
long or small, whether globular or circular or triangular or square or
hexagonal or octagonal, whether black or blue or red or yellow or white,
whether of sweet smell or of bad smell, whether bitter or pungent or
astringent or sour or sweet, whether hard or soft or heavy or light or
1 V- infra , 528 - 535 .
The Historical Background ioi
cold or hot or smooth or rough. Those therefore who believe that
there is and exists no soul speak the truth’. 1 The argument is that the
soul cannot be seen since it has no visible form (long, globular)
or colour (blue); likewise, it cannot be smelt (sweet smell), tasted
(bitter) or known by touch (heavy, cold). Hence it cannot be per-
ceived and one cannot speak of that which is not perceivable as
existing. The Materialists seem to have adopted Berkeley’s empiricist
principle of, esse est percipi, and argued that, non percipi est non
esse.
(132) It is, however, necessary to observe that even this argument is
not an abstract one, entirely evolved by the Materialists, but seems to
have been suggested by and specifically directed against their oppo-
nent’s theories about the atman or jlva. It would be seen that the
atman has shape and size according to some Upanisadic conceptions.
At Katha 2.3.17, the atman is ‘of the size of a thumb’ (angustamatrah)
and at Cbandogya 3.14.3, it is said to be ‘smaller than a grain of rice’
(anlyan vriheh). Likewise the Jains held that the soul (jlva) took the
shape of each body. Some of the AjTvakas seem to have believed that
the ‘soul was octagonal or globular and five hundred yojanas in
extent’ 2 (jlvo atthamso gulaparimandalo, yojanani sata panca, Pv. 57,
verse 29). As Basham has shown, according to late Ajivika sources the
soul was blue in colour. 3 The abhijati doctrine 4 may possibly have
been based on beliefs about the colour of the soul and it may be
noticed that the colours mentioned here are also the colours of the
abhijatis and are stated in the same order though the distinction between
the white (sukka) and the pure white (paramasukka) is not made.
These conceptions may have been suggested by experiences in trance-
1 Anno bhavati jive annam sarlram, tamha te evam no vipadivedenti, ayamaiiso,
aya diheti va hasseti va parimandaleti va vatteti va tamseti va caiiramseti va
ayateti va chalamsieti va attamseti va kinheti va nlleti va lohiyahalidde sukkileti
va subbhigandheti va dubbhigandheti va titteti va katueti va kasaeti va ambileti
va mahureti va kakkhadeti va maiieti va gurueti va lahueti va sieti va usineti va
niddheti va lukkheti va evam asante asamvijjamane jesim tam suyakkhayani
bhavati, Su. 2.1.9. Vol. 2, fol. 11.
2 The commentary (Paramatthadlpani, III. 253) says that ‘the soul is some-
times octagonal and sometimes globular’ (jlvo kadaci atthamso kadaci gulapari-
mandalo).
3 Op. cit., p. 270; ‘Jlva . . . was the colour of a palai fruit’, which is blue.
4 Other explanations are, however, possible; v. G. P. Malalasekera and K. N.
Jayatilleke, Buddhism and the Race Question , pp- 38-9 and p. 39, fn. 1.
102
Early Buddhist Theory of Knowledge
states as Buddhaghosa has suggested. 1 Even the Buddhists while dis-
pensing with the concept of a substantial soul, speaks of the experience
in jhanic states of a ‘consciousness (vinnanam) as being attached to
(ettha sitam, ettha patibaddham) though separate from the body of
the four elements . . . like a blue, orange, red, white, or yellow string
running through a diamond, bright, of the purest water, octagonal in
shape (atthamsa-), well-cut, clear, translucent, flawless and perfect in
every way’. 2 The Materialist criticism was, therefore, probably directed
against the objectivity of these claims in view of the fact that they could
not be demonstrated as verifiable in the light of sense-experience.
(133) That the positivism of these early Materialists was perhaps not
entirely based on this psychological empiricism is suggested by an
argument against the concept of the atman based on an elementary
linguistic analysis. This argument occurs as late as the verses quoted
in the Sarvadarsanasamgraha but there is some reason to suppose that
these verses preserve some of the primitive views of the Materialists.
Besides, the argument has its counterpart in the early Buddhist texts,
where the Buddha says that one should not be misled by language in
talking about an atman (v. infra, 533). The question as to whether the
Buddhists borrowed the argument from the Materialists (or vice versa)
or whether they both used it more or less contemporaneously for a
common purpose depends on the methodological criteria that we
adopt 3 but there is no gainsaying the fact that both the Materialists as
well as the early Buddhists appear to have used this argument against
the atman-theorists, whether they were influenced by each other or
not. The argument is, however, more explicitly stated by the Material-
ists and seems to be a criticism of one of the earliest conceptions of the
1 He says (DA. I.119) that those who consider that the soul has material or
visible form (rupl attati) do so on the grounds that the colour of their meditational
device (kasinarupam) is the soul, taking the consciousness that prevails in relation
to it as his own; he, however, distinguishes the Ajlvikas and others who arrive at
similar conclusions on purely logical grounds. Rupl atta’ti adisu kasinarupam
attati tattha pavattasannam c’assa sanna ti gahetva va Ajlvaka’dayo viya takkamat-
ten’eva va.
2 Ayam kho me kayo . . . catumahabhutiko . . . idanca pana me vinnanam ettha
sitam ettha patibaddhan’ti. Seyyatha pi . . . mani veluriyo subho jatima atthamso
suparikammakato accho vippasanno anavilo sabbakarasampanno, tatra suttam
avutam nllam va pltam va lohitam va odatam va pandusuttam va, D. I.7 6.
3 I.e. if we go strictly by the principle that whatever occurs in a later text is in
fact later in origin, we would have to say that the Buddhists were the first to use
this argument but this need not necessarily be true.
The Historical Background 103
meaning of words, namely that the meaning of word is an object. The
word for ‘meaning’ and ‘object’ in Indian thought is the same word
‘artha’ and the orthodox conception as noted by Kalidasa is that ‘the
word and the object are closely allied’. 1 According to the Purva
Mimamsa, the relation between the word and its meaning is natural,
necessary and eternal. 2 This means that the word ‘I’ must have an
object which must be the substantial ego. Arguing from logic to reality
one may hold that ‘I’ — statements must have as their subject a sub-
stantial pure ego, which is the ontological subject of the predicates.
The Materialists contested this belief or argument urging that the
subject of statements such as ‘I am fat’, etc., is the body which alone has
the observable attribute of fatness, 3 while phrases such as ‘my body’
have only a metaphorical significance 4 and would mean ‘the body that
is I’ just as when we speak of ‘the head of Rahu’ we mean ‘the head
that is Rahu’. 5 The Materialist thus seems to have pointed out on the
basis of an elementary linguistic analysis that it is false to conclude that
every proper or common name or grammatical subject entails the
existence of a specific ontological entity, to which it refers.
(134) The other arguments recorded of the Materialists are all of the
nature of destructive hypothetical syllogisms of the form modus tollen-
do tollens, 6 where the implicate is a proposition which is observably
false or absurd entailing the falsity of the implicans. This seems to have
been a favourite type of argument employed by disputants against
their opponents during the time of the Pali Nikayas and the Buddhists
also use arguments of this same kind against their opponent’s theories
(y. infra , 693-710). It consists in taking an assumption or proposition
of your opponent’s system (say, p) and showing that it implies a
proposition q, which is observably false (or absurd), 7 thus implying
that the original assumption or proposition is false and untenable, viz.
if p, then q
not q
Therefore, not p
1 Vagarthavivasamprktau, i.e. united like the word and its object, Raghuvamsa,
1. 1. 2 v. Hiriyanna, Outlines of Indian Philosophy , pp. 309 ff.
3 aham sthulah krso’smi ti samanadhikaranyatah, dehah sthaulyadiyogac ca sa
evatma na caparah, Sarvadarsanasamgraha, by Sayana-Madhava, Ed. V. S.
Abhyankara, Second Edition, Poona, 1951, p. 7.
4 Mamadeho’yam ityuktih sambhaved aupacarikl, op. cit., p. 7.
5 Mama sariram iti vyavaharo rahoh sira ityadivad aupacarikah, op. cit., p. 6.
6 Stebbing, op. cit., p. 105.
7 This is popularly known as reductio ad absurdum .
104 Early Buddhist Theory of Knowledge
In Indian logic this would fall under tarka (v. Nyaya Sutra, 1.1.40) i.e.
indirect proof or confutation. Here again the evidence is from the
Sarvadarsanasamgraha but the subject-matter appears to be early. The
arguments are sometimes stated in the form of rhetorical questions but
they can be easily converted into propositional form. We may illustrate
this by stating the arguments in propositional form and comparing
them with the actual form in which they are stated. Most of the argu-
ments are against the validity of the sacrifice:
(i) If ‘beings in heaven are gratified by our offering the sraddha
here’ (p), then ‘food given below should gratify those standing on the
housetop’ 1 (q), but q is observably false and absurd, implying the
falsity of p. The implicate (i.e. q) is however stated in the form of the
rhetorical question, ‘then why not give the food below . . .’.
(ii) If ‘the sraddha produces gratification to those who are dead’ 1
(p), then ‘(offerings in their home should) produce gratification to
travellers’ (q). But q is observably false and absurd. Here the implicate
is stated in the form of the proposition ‘here, too, in the case of
travellers when they start, it is needless to give provisions for the
journey’. 1 This is really an implicate of the implicate but the logic of
the argument remains the same.
(135) There is a similar argument implying the falsity of the belief in
survival:
(iii) If ‘he who departs from the body goes to another world’ (p)
then ‘he would come back for love of his kindred’ ^q). But p is observ-
ably false implying the falsity of p.
(136) This last (i.e. iii) is among the propositions which Payasi puts
to the test by devising experiments to test its validity instead of being
merely satisfied with anecdotal or common-sense observations. Payasi,
who also appears to belong to the Positivist branch of the Materialists,
deserves to be mentioned separately since he adopts the Materialist
philosophy of life on the basis of empirical arguments and experimental
evidence. 2 The dialogue between Payasi and Kassapa, which is re-
corded in the Payasi Sutta (D. II.316 ff.) is said to have taken place
some time after the death of the Buddha. 3 It shows that at least by this
1 Coweirs Translation, SDS., p. 10. We have not quoted the Sanskrit text
here since it does not affect the form of the argument.
2 v . Ruben, op. cit p. 109; Payasi machte noch ein anderes konigliches
Experiment.
3 v. Prof. Rhys Davids, Dialogues of the Buddha } SBB., Vol. Ill, p. 347.
The Historical Background 105
time, if not earlier, 1 some people had thought of consciously devising
experiments to test the validity of a theory, however ill-conceived and
ill-devised the experiments may have been, and either accepted or
rejected the theory on the basis of the results obtained.
(137) Payasi recounts a series of such experiments that he has per-
formed with negative results in order to test the validity of the belief
in survival. He approaches those who have led an immoral life when
they are grievously ill and about to die and enjoins them to return to
him if they survive in an unhappy state and inform him about their
condition (D. II.320). He likewise approaches those who have led a
moral life and instructs them accordingly (D.II. 323, 326). These
experiments, he says, had negative results since none of the subjects
came back after surviving death to tell him about their plight.
(138) The next set of experiments he mentions are designed to test
whether a soul escapes from the body at death. However crude his
experiments are, he seems to have taken great care in arranging them.
He puts a man (a thief) alive into a jar, closes its mouth securely,
covers it over with wet leather, puts over that a thick cement of moist
clay, places the jar on a furnace and kindles a fire. When he believes
that the man is dead, he takes down the jar, unbinds and opens its
mouth and quickly observes it with the idea of seeing whether his soul
issues out (D. III.332, 333). His failure to observe such a soul is taken
as evidence that there is no soul. Another experiment that he performs
is that of weighing (tulaya tulayitva) a man’s body before and after
death. It is presumably assumed that if the weight is less after death,
then something has left the body, namely his soul, but Payasi finds to
his consternation that after death the body was heavier (garutara-) so
that it was evident to him that no soul had left the body (D. III.334).
In yet another of his gruesome experiments he kills a man by stripping
off cuticle, skin, flesh, sinews, bones and marrow, turning him around
when he is almost dead to see whether any soul escapes from his body
(D. II.336). Again, he flays a man alive cutting off his integument,
flesh, nerves, bones and marrow to see whether at any stage he could
observe a soul. This is probably based on the conception of the souls
at Taittiriya Upanisad 2.1-5, which speaks of five souls 2 , the one
encased in the other. All these experiments assume that the soul is
either an observable or material substance, possessed of weight, located
in the body and passing out of it at death.
1 v. Uddalaka’s experiment, supra , 28. 2 The pancakosa theory.
D*
106 Early Buddhist Theory of Knowledge
(139) Whether the problem of man’s survival can be studied experi*
mentally, as some modern psychical researchers believe, or not, it is
clear that Payasi’s experiments were misguided and ill-conceived and
no results could be expected of them. But the fact that he devised and
carried them out in order to test a theory shows that he had a funda-
mentally unbiased and scientific outlook to the study of a problem.
(140) We have suggested that some of the arguments of the Material-
ists implied a criticism of the objectivity of the claims of the mystics
(y. supra , 130) in that what was objectively verifiable was limited to
what was based on sense-experience. If the account given of the
Materialist schools in the Brahmajala Sutta is to be trusted, there seems
to have been a class of Materialists who, while valuing the attainment
of yogic states from a purely pragmatic point of view, denied the
epistemic claims made on their behalf.
(141) Of the seven schools we identified the first with the first school
of the Sutrakrtariga ( v . supra , 115) which asserted that the soul was
not different from the body. The second school which speaks of a
‘higher 1 soul’ (atta dibbo) still assuming the shape of the body (rupi)
is probably the same as the school referred to in Gunaratna’s quotations
from Vacaspati (v. supra , 99) which spoke of a Materialist school
holding that caitanya or consciousness was a distinguishable by-product
of the material entities. The description of the third to the seventh
schools are similar to the accounts given of jhanic states. Take the third
school. It is said to posit the existence of a ‘higher soul’ (atta dibbo)
which is described in the following phrase, rupi manomayo sabbahga -
paccangi ahinindriyo ( loc . «*.). The description is identical with the self
which is said to be created by the mind in jhana, viz., so imimha kaya
annam kayam abhinimminati rupim manomayam sabbahgapaccahgim
ahinindriyam (D. I.77); it is the same as the ‘mental self’ (manomayo
attapatilabho) attained in jhana, viz. rupi manomayo sabbangapaccangt
ahinindriyo , ayammanomayo attapatilabho (D. 1 . 1 95). The souls
posited in the remaining four schools are identical in description with
the states of the four arupajhanas. As Materialists, they are said to hold
that all these emergent souls are destroyed with the destruction of the
body. But the identity of the description of the souls with the jhanic
states makes the very existence of these Materialist schools suspect.
The possibility that they are hypothetical schools concocted by the
1 For this sense of ‘dibba-’ v. O. H. de A. Wijesekera, ‘Upanishadic Terms for
Sense Functions’, UCR., Vol. II, pp. 23, 24.
The Historical Background 107
author of the Sutta who was anxious to present sixty-two theories in
this Sutta cannot be ruled out especially since there seems to have been
a belief at this time that there were 'sixty-two ways of life’ (dvatthi
patipada, 1 D. 1 . 54) which means that there would have to be sixty-two
theories on which these were based. On the other hand, since the
majority of the views stated here are, in our opinion, traceable to non-
Buddhist sources we need not be too sceptical even of this list. 2 Even
if five schools, each according to the state of jhana mentioned, did not
exist, we need not doubt the existence of at least one school of Material-
ists who claimed to attain jhanic or yogic states, while denying the
ontological or epistemological claims made about them, especially
since we seem to find some hints about the existence of such a school
from other sources. Gunaratna says that there were some yogis (yo-
ginah) who were nastikas, where the context shows beyond doubt that
he is using the term nastika- to refer to the Materialist schools. His
statement reads as follows: kapalika bhasmoddhulanapara yogino
brahmanadyantyajatasca kecana nastika bhavanti, te . . . caturbhutat-
makam jagadacaksate (op. cit p. 300); here whether we take yoginah
as qualifying kapalikah or as a class by themselves it is clear that some
yogis were Materialists. In the Taittiriya Upanisad, we find the cryptic
statement, asadeva sa bhavati asadbrahme’ti veda cet (2.6.1.), which is
translated by Radhakrishnan as 'non-existent, verily, does one become,
if he knows Brahman as non-being’. If the statement that ‘Brahman is
non-being’ was made by a person who had attained the yogic state
described as the 'attainment of Brahman’ (brahmaprapta-, Katha,
2.3.18), he would be a Materialist as defined above.
(142) Now if there was a class of Materialists who had attained one of
the arupajhanas, we can make some interesting deductions about their
beliefs. For it is stated that when the fourth jhana is attained immedi-
ately prior to entering the arupajhanas (formless mystical states) the
mind is ‘clear and cleansed’ (parisuddha-, pariyodata-, D. I.75-6) and
1 This is one of the Ajivika doctrines propounded by Makkhali Gosala (v.
Basham, op. cit., p. 242). Basham takes it to mean ‘religious systems of conduct,
of which the majjhimd patipada of Buddhism was one' but has apparently not
noticed the correspondence of the number sixty- two with the sixty- two theories
frequently mentioned in the Buddhist texts.
2 N. Dutt following Thomas does not think that the list of views in the
Brahmajala represents actual views current at the time; v. Early Monastic
Buddhism, 2nd Edn., Calcutta, i960, p. 36; cp. E. J. Thomas, The Life of the
Buddha, New York, 1927, p. 199*
108 Early Buddhist Theory of Knowledge
that when the mind is clear and cleansed, it acquires certain extrasensory
faculties whereby it is possible to have a vision of one’s past births
(pubbenivasanussatinana-, D. I.82) as well as the ‘decease and survival
of beings’ (satte cavamane upapajjamane, D. I.82-3). If these Material-
ists acquired these ‘extrasensory faculties’ which ostensibly gave
alleged evidence of survival, why is it that they believed in the annihila-
tion of the soul at death? Did they like some moderns hold that these
mystic states and the visions had in them, though real as experiences
were nevertheless hallucinatory, delusive and non- veridical. The com-
mentary seems to offer an explanation though it does not appear to be
satisfactory. It says that ‘there were two types of Materialists (lit.
annihilationists), those who have attained jhana (labhl) and those who
have not (alabhl). Those who have attained it observe the decease (of
beings) but not their survival (cutim disva upapattim apassanto-) with
the clairvoyant vision of the worthy ones; he who is thus successful in
observing only the decease but not the survival of beings accepts the
annihilationist theory’. 1 The explanation is logically sound but it does
not appear very plausible. It would be more likely that this school of
Materialists asserted the possibility of attaining these mystical states
but denied any claims regarding the validity of extrasensory perception
in that they were private experiences which gave us no objective
information.
1 Tattha dve jana ucchedaditthim ganhanti labhl ca alabhl ca. Labhl arahato
dibbena cakkhuna cutim disva upapattim apassanto, yo va cutimattam eva datthum
sakkoti na upapattim so ucchedaditthim ganhati, DA. I.120.
CHAPTER III
THE HISTORICAL BACKGROUND III—
NON-VEDIC II— SCEPTICS, AJlVIKAS AND
JAINS
(143) In this chpater we propose to make a detailed study of the
doctrines of the Sceptics, which are mentioned in the Pali Nikayas.
They have influenced Early Buddhism (v. infra, 739, 813) and directly
concern us. We shall also briefly examine the epistemological and
logical doctrines of the Ajivikas and Jains, which seem to have a bearing
on the thought of the Canon.
(144) Traces of scepticism and agnosticism we find from the time of
the Rgveda onwards (v. supra , 7). These instances are sporadic and
there is no evidence of any widespread scepticism. Radhakrishnan says
that the hymn to faith (sraddha, R.V. 10. 1 51) ‘is not possible in a time
of unshaken faith 5 1 but there is nothing in the hymn itself to indicate
the presence of scepticism at the time. This scepticism, as we said,
found its latest expression in the Nasadiya hymn ( y . supra , 8-10),
where it was extended to the very possibility of arriving at a final
solution to a specific problem. This Rgvedic scepticism did not develop
any further but we found certain undercurrents of doubt (vicikitsa) in
the Brahmanas ( v . supra, 15). The doubt with regard to survival was
first mooted in the Brahmanas and appears in the Early Upanisads,
where it was asked whether man can survive death, when nothing is
left over to germinate in a next life ( V . supra, 15). On the other hand,
we found in the Upanisads a rational agnosticism approaching Kantian
agnosticism, where Yajnavalkya rationally demonstrated the impossi-
bility of knowing the ultimate reality or the atman (v. supra, 43).
Nevertheless, it was not an agnosticism proper in that it differed from
Kantian agnosticism in one significant respect. For, although it was not
1 Radhakrishnan and Moore, op. cit., p. 34*
no
Early Buddhist Theory of Knowledge
possible to apprehend ultimate reality conceptually it was still con-
sidered possible to have some sort of direct acquaintance with it in
deep sleep, in the next life or in a mystical state.
(145) These sceptical hints of the earlier Vedic thought and the
agnostic trends of the Upanisads could have paved the way for the
growth of sceptical schools of thought, but the impetus and the
occasion for their arising seem to have been provided by the presence
of diverse, conflicting and irreconcilable theories, pertaining to moral,
metaphysical, and religious beliefs. When there is a welter of contend-
ing views, people naturally become curious as to which view is true and
in the absence of a safe criterion of truth become suspicious as to
whether any view at all could be true.
(146) When a school of thought strongly urged the belief in survival
and another vehemently denied it and both were able to adduce on the
face of it equally strong arguments for their respective points of view,
one becomes doubtful as to which view if at all could be true. When
the Katha said, ‘this doubt (yicikitsa) there is with regard to a man
deceased; “he exists” say some, “he exists not” say others’ (1.1.20), it
is probably echoing at least the uncertainties with regard to the
problem of survival entertained by the intellectuals at that time in the
presence of a school of Materialists, who strongly denied survival.
(147) That the intellectual confusion resulting from the presence of a
diversity of views seems to have been the main motive for the birth of
scepticism is apparent from the sayings and opinions ascribed to the
Sceptics (ajnanikah, ajnaninah) by Sllanka commenting on the
Sutrakrtanga. One has, however, to be cautious in picking out the
sayings ascribed to the Sceptics from those attributed to the ajnanikah
or ajnaninah in general, since Sllanka uses these terms in at least three
senses. Sometimes he employs the word to denote ‘the ignorant’
religious teachers 1 following the usage of the Sutrakrtanga which uses
the term in this sense at times (y. annaniya, 1.1.2.16). He also uses the
term of the Buddhists who, he says, are ‘more or less ajnanikas since
they consider that karma done out of ignorance (he probably means
“unintentionally”) does not result in bondage’. 2 Most often, however,
1 samyagjnanavirahita sramana brahmanah, Vol. I, fob 35 on Su. 1.1.2.16.
2 Sakya api prayaso’ jnanika avijnopacitam karma bandham na yatltyevam
yatas te ’bhyupagamayanti, Vol. I, fob 217 on Su. 1.12.2.
The Historical Background ill
ajnanikah, or ajnaninah is used as a technical term to denote the Sceptics
following the usage of the Sutrakrtanga (e.g. 1.1.2.27, 1.12.2) and in
this sense the word is defined either as ‘those who claim that scepticism
is best’ (ajnanam eva sreya ityevamvadinam, Vol. I, fol. 35 on Su.
1. 1. 2. 1 7) or as ‘those in whom no knowledge, i.e. scepticism, is evident’
(na jnanamajnanam tadvidyate yesam te’jnaninah, loc . cit.; cp. ajnanam
vidyate yesam, Vol. I, fol. 215 on Su. 1.12.2). The term is also sarcastic-
ally defined as ‘those who move in ignorance or those who show
themselves off to the extent of being extraordinarily wise’ (ajnanikah)
(ajnanena va carantltyajnanikah; ajnanika va tavat pradarsayante,
Vol. I, fol. 215 on Su 1. 12. 2). When, however, Silanka makes the
statement that, Ajlvikadayo Gosalamatanusarino’jhanavadapravrttah,
i.e. ‘the Ajlvikas and the others who are the followers of Gosala’s
doctrines are a product of ajnanavada’, 1 ( op . cit., Vol. I, fol. 36), one
is at a loss whether to translate ajnanavada here as ‘ignorance’ or as
‘scepticism’ in the general or special senses. Since Silanka elsewhere
identifies the ‘followers of Gosala’s doctrines’ as the Vainayikas, 2
(moralists) which Professor Basham finds a ‘puzzling reference’, 3
it is unlikely that he thought of them as an offshoot of the Sceptics
(ajnanavada-) since he distinguishes the Vainayikas from the Ajna-
nikas.
(148) Despite these variant uses of the terms ajnanikah and ajnaninah
on the part of Silanka, it is not difficult on the whole to pick out the
references to the genuine Sceptics from the context. In one of these
contexts he ascribes a statement which, if true, leaves us in no doubt
that the conflict of theories and the consequent difficulty of discovering
the truth was the raison d’etre of scepticism. Barua has translated a part
of this passage, leaving out the latter part (which is somewhat obscure)
and has concluded from it that the Sceptics were stressing the moral
dangers of subscribing to conflicting views as the reason for their
scepticism. 4 He has mistranslated the phrase, bahutaradosasambhavat,
after reducing it to ‘bahu dosah’ on his own and rendering it as ‘many
1 Professor Basham has not mentioned this statement where he has made a
study of similar statements, v. op. cit., pp. 174-7.
2 Op. cit., Vol. I, Fol. 15 1 on Su. 1.6.27.
3 Op. cit., p. 177. Basham’s attempt to explain the Vainayikas as a later schis-
matic sect of the Ajlvikas(?) is unsatisfactory since the Vainayikas are known as
early as the Pali Nikayas (cp. venayiko, M. 1 . 140; cp. D. 1 . 174, santi eke saman-
abrahmana sllavada).
4 Barua, op. cit ., p. 330.
1 12
Early Buddhist Theory of Knowledge
moral injuries’, whereas the context would have made it clear, had he
translated the whole passage that bahutaradosa- here means ‘the
multitude of (intellectual) confusions’ or the ‘magnitude of mistakes
(arising from claims to knowledge)’. We may render this passage as
follows: ‘For they (i.e. the Sceptics) say that those who claim know-
ledge (jnaninah) cannot be stating actual facts since their statements are
mutually contradictory, for even with regard to the category of the
soul, some assert that the soul is omnipresent ( sarvagatam ) and others
that it is not omnipresent ( asarvagatam ), some (say) it is of the size of
a digit (< ahgustaparvamatram ) others that it is of the size of a kernel of a
grain of millet (syamakatandulamatram) some say it both has form and
is formless ( murtamamurtam ), some that it resides in the heart
(hrdayamadhyavartinam) and (others) that it is located in the forehead
(, lalatavyavasthitam ), etc. — in respect of every category there is no
uniformity in their assertions; there is no one with an outstanding
intellect whose statements may be regarded as authoritative; even if
such a person existed, he cannot be discovered by one with a limited
vision according to the maxim that “one who is not omniscient does
not know everything” for it is said “how can one desiring to know
that a certain person is omniscient at a certain time do so if he is devoid
of that person’s intellect, his knowledge and his consciousness”;
owing to the absence of the knowledge of the means, it cannot properly
be accomplished; it cannot be accomplished because of the mutual
dependence (of the two); for it is said “without a super-knowledge
(visistaparijnana-) the knowledge of the means is not attained
and as a result there is no attainment of the super-knowledge
of the object”; knowledge cannot completely comprehend the nature
of the object of knowledge, for it is said, “whatever is apprehended
should have the parts, near, middle and outer but here only the near
part is apprehended and not the others since it is determined by it
(i.e. the nature of the object)”; as for exhausting the atom (paramanu-
paryavasanata?) with the (knowledge of) the near portion, considering
the unrepresented parts out of the three parts, it is not possible to
apprehend the atom by those with a limited vision owing to the excel-
lence of its nature; therefore, since there is no omniscient person and
since one who is not omniscient cannot comprehend the nature of an
object as it is constituted, since all the theorists (sarvavadinam) have
conceived of the nature of the categories in a mutually contradictory
manner and those who have claimed super-knowledge (uttarapari-
jnaninam) are at fault (pramadavatam) Scepticism is best owing to the
The Historical Background 113
magnitude of the mistakes that arise (from claims to knowledge)
(bahutaradosasambhavat)’. 1
(149) Even if Sllanka, writing in the ninth century, has rightly repre-
sented the views of the Sceptics, we have no right to assume that in the
day of the Pali Nikayas they also held the same view. The sophisticated
argument based on certain conceptions about the nature of knowledge
in order to disprove the possibility of omniscience certainly appears
prima facie to be late but the general thesis of the Sceptics, that the
possibility of knowledge is doubtful since the claims to knowledge
were mutually contradictory, may well go back to the period of the
earliest Sceptics. Sllanka often quotes this idea sometimes as a maxim
of the Sceptics as, for instance, when it is said that ‘they posit the theory
that since those who claim knowledge make mutually contradictory
assertions, they cannot be stating the truth’ 2 and sometimes without
specific reference to the Sceptic as for instance when he says that
‘since the various theories claiming knowledge (jnanam) have arisen
in contradiction to one another, they are not true; therefore, Scepticism
is best of all.’ 3 Sllanka speaks of the kind of investigation (mimamsa-)
and reflection (vimarsa-) which leads to Scepticism as follows: ‘Is this
theory claiming knowledge (kimetad-jnanam) true or false? Scepticism
1 Tathahi te ucuh -ya ete jnaninas te parasparaviruddhavaditaya na yathartha-
vadino bhavanti, tathahi-eke sarvagatamatmanam vadanti, tatha’nye asarvagatam,
apare arigustaparvamatram, kecana syamakatandulamatram, anye murtamamur-
tam hrdayamadhyavartinam lalatavyavasthitamityadyatmapadartha eva sarvapa-
darthapurahsare tesam naikavakyata, na catisayajnanl kascidasti yadvakyam
pramamkriyeta, na casau vidyamano’ py upalaksyate’rvagdarsina, f na sarvajnah
sarvam janati’ti vacanat, tatha c’oktam-’ sarvajno’saviti hyetattatkale’pi bubhut-
subhih tajjnanajneyavijnanasunyairvijnayate katham? , ; na ca tasya samyak
tadupayaparijnanabhavat sambhavah, sambhavabhavascetaretarasrayatvat, tathahi
— visistaparijnanamrte tadavaptyupayaparijnanam upayamantarena ca nopeyasya
visistaparijnanasyavaptir iti, na ca jnanam jneyasyasvarupam paricchetum alam,
tathahi -yatkimupalabhyate tasyarvagmadhyaparabhagairnetarayoh, tenaiva
vyavahitatvat, arvagbhagasya’pi bhagatrayakalpanat tatsarvaratiyabhagaparikal-
panaya paramanuparyavasanata, paramanosca svabhavaviprakrstatvadarvag-
darsaninairi nopalabdhir iti, tadevam sarvajnasyabhavadasarvajhasya ca yatha-
vasthitavastusvarupaparicchedat sarvavadinam ca parasparavirodhena padar-
thasvarupabhyupagamat yathottaraparijnaninam pramadavatam bahutarados-
asambhavad ajnanameva sreyah, op. cit. 9 Vol. I, fol. 215, 6 on Su. 1.12.2.
2 Tairabhihitam jnanavadinah parasparaviruddharthavaditaya na yathartha-
vadinah, op. cit Vol. I, fol. 216 on Su. 1.12.2.
3 Na ca tani jnanani parasparavirodhena pravrttatvat satyani tasmadajnanameva
sreyah, op. cit. 9 Vol. I, fol. 34 on Su. 1.1.2.14.
H4 Early Buddhist Theory of Knowledge
is best since with an excess of knowledge, there is an increase of
mistakes (dosatireka-)’. 1
(150) Even the reference to the conflicting theories about the atman
may be attributed to the early Sceptic, since each one of the theories
stated were current by the time of the Pali Nikayas and all of them
could be traced to the period of the Early Upanisads. Thus the
pantheistic atman, which is ‘made of everything’ (sarvamayah,
idammayah adomayah, Brh. 4.4.5) would be omnipresent (sarvagatam)
while the transcendent atman defined negatively (neti neti, Brh. 3.9.26)
would not be so. Again at Katha 2.3.17, the atman is of ‘the size of a
digit’ (< angustamatrah ), while at Chandogya 3.14.3, the atman is ‘smaller
than a kernel of a grain of millet’ 2 (atma aniyan syamakatandulat).
Again at Brhadaranyaka 2.3.1, Brahman which is identical with the
atman is said ‘both to have form and also be formless’ (murtam
caivamurtan ca). Likewise at Katha 2.3.17 the atman ‘resides in the
heart’ (hrdaye sannivistah) while at Aitareya Aranyaka 2. 1.4.6 it (i.e.
brahman = atman) is located in the head (siro’srayata). It is not at all
surprising that the Sceptics would have been quick to see these con-
tradictions in the Upanisads in an age when the Vedantic interpretation
(or for that matter, the interpretations of Deussen or Radhakrishnan)
which tries to synthesize all these contradictions, was not known.
(151) The sophisticated argument against the concept of omniscience
appears to be too involved or complicated to belong to the early
Sceptics but here again we need not doubt that they would have
questioned the possibility of omniscience in an age when there was
more than one claimant to omniscience. The leader of the Jains claimed
omniscience according to the evidence of both the Buddhist as well as
the Jain texts (v. infra , 311) and so did Purana Kassapa (v. infra , 19 6).
Omniscience is claimed for Makkhali Gosala in the later Tamil texts
Civananacittiyar and Nllakeci as Prof. Basham has shown (op. cit . 9
p. 276), though there is no evidence that he himself claimed omniscience.
It is not unlikely that since the Buddha argued .against the claims to
1 Kimetadjnanam satyamutasatyamiti? Yatha ajnanameva sreyo yatha
yatha ca jnanatisayastatha tathaca dosatireka iti, op. cit Vol. I, fol. 35 on
Su. 1. 1. 2. 17.
2 Here the Upanisad is itself possibly trying to explain the contradictions in
previous theories by turning them into paradoxes since it also says that the atman
‘is greater than the earth’, etc.
u5
The Historical Background
omniscience on the part of religious teachers ( v . infra, 31 1), the Sceptics
would likewise have done so. The sayings which state that with a
limited knowledge no one can know that any person is omniscient,
e.g. ‘nasarvajnah sarvajnam janati’; ‘sarvajno’ saviti hyetattatkale
pi bubhutsubhih tajjnanajneyavijnanarahitair gamyate katham’ are
also quoted elsewhere 1 and may possibly have been old sayings of the
Sceptics. Another saying bearing on this topic specifically attributed
to the Sceptics and criticized, reads as follows: ‘All teachings are like
the utterances of barbarians since they have no (epistemic) basis’
(chinnamulatvat mlecchanubhasanavat sarvam upadesadikam, op. cit .,
Vol. I, fol. 35 on Su. 1. 1. 2. 17). This was possibly directed mainly
against those who claimed to speak with authority on the presumption
of their omniscience. On the basis of these sayings we may perhaps
surmise that they argued that since the human intellect was limited no
one could claim to know everything with such a limited intellect. They
may have even extended this argument to arrive at their Scepticism.
None of the metaphysical theories claiming to be true, which are the
products of such a limited intellect, can be known to be true, since they
are mutually contradictory. Now, no new theory can also be true since
it is bound to contradict one or more of the existing theories. Therefore
nothing can be known to be true. Thus the contradictions of meta-
physics and the impossibility of omniscience may have led them to
accept Scepticism. One thing we need not doubt and that is that these
Sceptics more than any other thinkers of their age appear to have been
struck by the fact that the conflicting theories not of one tradition but
of all schools seemed to cancel each other out. And in this respect the
Sceptics were really the children of the age in which they lived.
(151A) That the period immediately preceding the rise of Buddhism
was one in which there was an interminable variety of views on matters
pertaining to metaphysics, morality and religion is clear from the
references to them in the Buddhist and Jain texts. The Brahmajala
Sutta (D. I.12-38) refers to fifty-eight schools of thought other than
the four schools or types of Sceptics referred to. It is not improbable
that some of these are only possible schools not current at the time
( v . supra, 1 41) but there are good grounds to think that many of them
were actually existing schools in view of the independent literary
sources which refer to them. Similarly, the Sutrakrtanga mentions three
hundred and sixty-three schools. This list is artificially made up mainly
1 v. op. cit., Vol. I, fol. 35 on Su. 1.1.2.16.
ii 6
Early Buddhist Theory of Knowledge
but not solely on the basis of the categories of Jainism itself 1 but there
would be little reason to deny the existence of the four main schools
of Kriyavada, Akriyavada, Ajnanikavada and Vainayikavada and
probably several subgroups among them. A brief account of the doc-
trines of some of these schools is given in several contexts of the
Sutrakrtanga (i. 1.1.8-18, 1.2.1-28, 1.6.27, 1.12.2-11, 2.1. 14-30, 2.2.79)
and these accounts do not appear to be in the least artificial.
(152) When we consider this historical background, it is only to be ex-
pected that the Sceptics should appear at this time. In the Sutrakrtanga,
they are called the ‘annaniya’ (Skr. ajnanikah), i.e. the ‘ignorant
ones’ or ‘sceptics’ or ‘those who deny knowledge’ (v. supra , 147),
translated as ‘agnostics’ by Jacobi (SBE., Vol. 45, pp. 241, 315). They
are mentioned in a few places (Su. 1.1.2. 17, 1.6.27, 1. 21. 1-2, 2.2.79)
and are considered one of the four important schools of thought. But
the information given about them in the texts themselves is meagre. It
is said that ‘the speculations (vlmamsa) of the Sceptics do not land
them in ignorance (as they ought to); when they cannot instruct
themselves in the truth (param), how can they instruct others’
(annaniyanam vimamsa annane na viniyacchal, appano ya param
nalam, kuto annanusasiiim), Su. 1.1.2. 17). Jacobi translates annane na
viniyacchal, as ‘cannot lead to knowledge’ (op. cit ., p. 241) but this is
not supported by the text or the commentary. Even if we translate, na
vi-niyacchai (Skr. na vi-niyacchati) as ‘cannot lead to’, annane (Skr.
1 Sllanka makes up the list of 363 ‘schools’ as follows (v. op. cit., Vol. I, fol.
21 2 > 3 ):
(i) Kriyavadins 180 (kriyavadinamasltyadhikm satama bhavati)
(ii) Akriyavadins 84 (akriyavadinam . . . caturaslti)
(iii) Ajnanikas 67 (ajnanikanam . . . saptasastih)
(iv) Vainayikas 32 (vainayikanam . . . dvatrimsat)
363 Total
(i) The 180 Kriyavadins are as follows: the variables are — the nine categories
of Jainism such as jiva -, etc., the two principles of svatah and paratah , the two
attributes of nitya- and anitya -, the five concepts of kala savabhava niyati-,
isvara- and atman-. This gives 9X2X2X5=180.
(ii) The variables are — the 7 categories of jiva-, etc., taken negatively, the
two principles of svatah and paratah , the six concepts (note the difference of these
concepts from those enumerated in (i)), viz. kala-, yadrccha-, niyati -, svabhava-,
isvara-, atma-. This gives 7X2X6=84.
(iii) For Ajnanikas, v. infra, 157.
(iv) The variables are the four duties (of manas-, vdk-, kaya -, and dana-')
towards seven types of people; 7X4=28.
The Historical Background 117
ajnane) is not ‘knowledge’ but the opposite of it; the commentary
explains the phrase as follows: ‘Ajnane’ ajnanavisaye ‘na niyacchati’
na niscayena yacchati, navatarati (op. cit ., Vol. I, fol. 35), which means,
the ‘(speculations) do not definitely take them or place them in the
realm of ignorance’. What is meant is that their scepticism should
lead them logically to the conclusion that they know nothing whatso-
ever, but in fact their ‘reflections have the features of knowledge’
(paryalocanasya jnanarupatvat, loc. cit.) and ‘one cannot understand’
(na budhyate, loc. cit.) how they claim to know such propositions as
‘ignorance is best’ (ajnanameva sreyah, loc . cit.), etc. So when they
claim that they are Sceptics they are (according to this Jain criticism)
in fact claiming to have some knowledge as revealed by their dicta and
thereby they are contradicting themselves. The other context in which
something informative is asserted about the Sceptics states that ‘these
Sceptics being “experts” are uncommitted’ (asamthuya—asambaddhah)
(commentary, op. cit., Vol I, fob 2x5); Jacobi translates as “reason
incoherently” (op. cit., p. 316) but they have not overcome doubt;
unskilled they teach the unskilled and utter falsehood without dis-
crimination’ (annaniya ta kusalavi santa, asanthuya no vitigicchatinna,
akoviya ahu akoviyehim, ananuvittu musam vayanti, Su. 1.12.2).
( I 53) Though the Sutrakrtanga itself tells us little, Sllanka’s commen-
tary, as we have already seen, is more informative. The main difference
that we notice between Sllanka’s account and that in the Pali Nikayas
is that the former stresses the intellectual grounds for their scepticism,
while the emphasis in the latter is on the practical value or the prag-
matic reasons for Scepticism. While the general argument for scepti-
cism appears to have been the one we outlined above (v. supra, 148),
an often quoted saying of the Sceptic throws a little more light on the
rational basis of their scepticism. It is said that the Sceptics hold that
‘scepticism is the best since it is difficult to gauge the thought processes
of another’ (paracetovrttlnam duranvayatvadajnanameva sreyah, op.
cit., Vol. I, fob 35 on Su. 1. 1.2. 17; cp. paracetovrttmam duranvayatvat,
op. cit., Vol. I, fob 216 on Su. 1.12.1). The difficulty of knowing
another’s mind seems to be one of the reasons why the Sceptics held
to their other dictum that ‘all teachings are like the utterances of
barbarians since they have no (epistemic) basic’ (v. supra, 151).
Silanka himself following the Sutrakrtanga makes use of this idea in
another connection and observes that ‘owing to the difficulty of know-
ing another’s mind, they do not grasp what is intended by the words
1 1 8 Early Buddhist Theory of Knowledge
of their teacher and thus repeat the other’s words like a barbarian with-
out understanding the real meaning’. 1 This idea seems to be borrowed
from the Sceptics. The fact that Silarika himself does not as a Jain believe
that one cannot know another’s mind is clear from his criticism of this
sceptical view. He says it is false (asat), because the Sceptics themselves
cannot believe this. For they put forward views such as ‘scepticism is
best’ (ajnanameva sreyah, v. supra > 147) intended to instruct others.
He quotes in his favour a non-sceptical view which says that ‘the inner
mind of another can be apprehended by his external features, gestures,
movements, gait, speech and the changes in his eyes and face’. 2
(154) Here again, we do not know on what grounds the Sceptic held
the view that one cannot know another’s mind but it is evident that
this theory itself could have led him to scepticism. If one cannot know
another’s mind, communication is impossible and knowledge no longer
becomes objective. We may profitably compare this view with that of
the Greek sophist, who believed in the incommunicability of what we
claim to know. In Gorgias’ book 3 on ‘Nature or the Non-existent’, he
sets forth three propositions, viz. (1) that nothing exists, (2) that if
anything exists it cannot be known, and (3) that if it can be known, the
knowledge cannot be communicated. The Ajnanikas seem to have
agreed with propositions (2) and (3) but not (1) since quite con-
sistently with their scepticism they could not categorically hold that
‘nothing exists’ but only that ‘nothing could be known to exist’. This
is the same as proposition (2), thus granting the possibility of existence.
Now Gorgias proves proposition (2) by showing that knowledge is
identical with sense-perception and that since sense-impressions differ
with different people, no two people can have the same sense-impres-
sions with regard to an object. Therefore knowledge, which must
necessarily be objective, is not possible because of this subjectivity.
For the same reason this knowledge being identical with sensation,
cannot be communicated. The Indian Sceptics may possibly have
reasoned on similar lines, though one cannot be quite certain about
this, due to the lack of any definite evidence. The argument against the
possibility of complete knowledge ( v . supra , 148) seems to give a faint
1 Evam paracetovrttlnam duranvayatvadupadesturapi yathavasthitavivaksaya
grahanasambhavanniscayarthamajanana mlecchavadaparoktam anubhasanta eva,
op . cit, y Vol. I, fol. 35 on Su. 1.1.2.16.
2 akarairingitairgatya cestaya bhasitena ca netravaktravikaraisca grhyate
’ntargatam manah, op. cit., Vol. I, fol. 35 on Su. 1.1.2. 17.
3 W. T, Stace, A Critical History of Greek Philosophy > pp. 116-7.
The Historical Background 119
indication of this. Here it was said that whatever we apprehend has the
three parts, near (arvak-), middle (madhya-), and outer (para-) and
that we in fact apprehend only the near part (tasyarvagbhagasyevopala-
bdhih, loc. cit.). Now this is certainly true of visual perception and in a
sense of sense-perception in general. We see only the near side (the
side facing us) of objects, so that what each person sees of the object
would be different according to the individual perspective. So if we are
arvag-darsinah (a term which is frequently used in the sayings of
Sceptics) or ‘near-side-seers’, our knowledge at least of physical
objects, being dependent on our individual perspectives, would be
subjective since these perspectives would be different with different
individuals. In the absence of objectivity, there is no knowledge at all
and the private experiences or impressions of the different individuals
would be incommunicable. Whether the early Sceptics would have
employed such reasoning or not it is difficult to say but they certainly
seem to have held that one could not know another’s mind and this
seems to have been one of the grounds of their Scepticism.
(155) As we have seen, Silanka’s account stresses the intellectual basis
of their scepticism rather than the pragmatic or moral reasons for it,
but the fact that they were also present is evident from some of his
observations about the Sceptics. According to Silanka ‘the Sceptics . . .
conceive that even if there was knowledge it is useless (nisphalam)
since it has many disadvantages (bahudosavat)’ (Ajnanikanam . . .
jnanam tu sadapi nisphalam bahudosavaccetyevamabhyupagamavatam,
op. cit. y Vol. I, fol. 215). This shows that they not only considered
knowledge to be impossible but that it was useless. In enumerating the
sixty-seven ‘types’ of Sceptics, Silanka puts the question of the
Sceptic in two forms, viz. ‘Who knows that the soul exists? Of what
use is this knowledge? Who knows that the soul does not exist? Of
what use is this knowledge? etc.’ (san jivah ko vetti? kim va tena
jnanena? asan jivah ko vetti? kim va tena jnanena? op. cit.y Vol. I,
fol. 36 on Su. 1. 1. 2. 20; also Vol. I, fol. 212). The second of these forms
is clearly meant to imply that they adopted Scepticism on pragmatic
considerations as well.
(156) Silanka does not shed any more light on what the Sceptics
considered as the defects or disadvantages (dosa-) of knowledge but
as we shall see the accounts in the Pali Nikayas pay a good deal of
attention to this aspect of their scepticism. But since we shall be dealing
with each school of Sceptics mentioned in the Brahmajala Sutta
120 Early Buddhist Theory of Knowledge
separately, we may briefly state here what the Sceptics seem to have
considered to be the defects or disadvantages of knowledge. The sense
in which the term dosa- is used by Silanka in elucidating the views of
the Sceptics is not very clear. In the passage in which knowledge was
condemned ((v. supra, 148) as giving rise to a multitude of dosa-
(bahutaradosasambhavat), it was apparent from the context that the
word meant ‘intellectual confusions’ and not ‘moral injuries’ as sugges-
ted by Barua (y. supra, 148), who was probably influenced by the picture
of the Sceptic as drawn in the Buddhist texts. The other uses (e.g. ‘the
greater the knowledge the greater the dosa’, yatha yatha ca jnanatisayas
tatha tatha ca dosatireka-, op . ciu, Vol. I, fol. 35 on Su. 1.1.2.17)
were less clear and dosa- could here have meant (ambiguously) ‘moral
disadvantages’. In the Brahmajala Sutta, however, we find that the
first two schools of Sceptics held that there were undesirable psycho-
logical and moral consequences of claiming knowledge under condi-
tions, when it was impossible to know the facts for certain. According
to the first school (v. infra, 159 ), we have a liking or bias for (chando,
rago) a proposition and a dislike or a bias against its contradictory
(doso, patigho), when we come to accept it as true without valid
grounds. Since this is grounded on one’s prejudices for and against,
the proposition itself is said to be false and its acceptance wrong or
mistaken (musa). Now, uttering a falsehood or doing a wrong thing
is a source of remorse (vighato) and is a moral danger (antarayo).
According to the second school (v. infra, 166), the bias for or against
is an entanglement (upadanam) which is again a moral danger
(antarayo). The third school (v. infra, 167) seems to have been
impressed by the psychologically and morally disastrous consequences
of debating their theories, on the part of those who claimed to know
and believe in them. We must not forget that during this period not
only were there a variety of theories but a good many of them were
being hotly debated (v. infra, Ch. V), resulting in one party having to
undergo the miseries of defeat. Sometimes these debates seem to have
given rise to bickering and quarrels among the contending parties.
This third school of Sceptics, if not the first and second as well, seem
to have concluded that all this self-imposed unhappiness was due to
baseless claims to knowledge and that Scepticism was superior to
making such claims. It is probable that these were among the defects
or disadvantages (dosa-) of knowledge spoken of in some of Sllanka’s
quotations from the Sceptics. If so it would be seen that there were
both pragmatic as well as intellectual grounds for their scepticism.
The Historical Background 121
(157) As we have mentioned (v. supra , 15 1) Silanka speaks of sixty-
seven ‘types’ of Sceptics in order to make up the figure three hundred
and sixty-three, the number of schools of thought, mentioned in the
Sutrakrtanga. But they are neither schools nor types and the list is
artificially made up mainly but not solely out of the concepts of Jainism
itself. He takes the nine categories (navapadartha-) of Jainism, each
according to the seven forms of predication (saptabhangakah). This
gives sixty three (i.e. 9X7) forms of sceptical questions, which are
considered to represent sixty three ‘types’ of Sceptics asking these
questions. The last four ‘types’ are more interesting and possibly
represent a kind of question, which the Sceptics themselves asked.
They are as follows:
(i) Sati bhavotpattih ko vetti? Who knows whether there is an
arising of psychological states?
(ii) Asati bhavotpattih ko vetti? Who knows whether there is no
arising of psychological states?
(iii) Sadasati bhavotpattih ko vetti? Who knows whether there is
and is no arising of psychological states?
(iv) Avaktavyo bhavotpattih ko vetti? Who knows whether the
arising of psychological states is impredicable? Silanka comments
that the ‘other three possibilities of predication do not apply in the
case of the arising of psychological states’ 1 . The question of ‘the
arising of consciousness’ (sannuppada-) is one on which there
seems to have been a good deal of speculation during the period
of the Pali Nikayas and four different theories on this subject
are mentioned in the Potthapada Sutta (D. I.180). This could have
easily provoked these sceptical questions, but what is interesting is the
fourfold mode of predication adopted. It is possible that Silanka did
this merely to complete the figure of sixty-seven and his explanation
that this subject does not admit of the other forms of predication is too
puerile to be taken seriously. But it is also not unlikely that the Sceptics
in fact adopted a fourfold scheme of predication as we have suggested
(v. infra , 184) in discussing the evidence from the Pali texts.
(158) The Pali term used to refer to the Sceptics, namely, Amaravikk-
hepika y seems to be a nickname and has probably been correctly
translated as ‘eel-wrigglers’ (Prof. Rhys Davids, SBB., Vol. II,
pp. 37 ff.). It is however a word whose meaning is obscure and the
1 Uttaram bharigatrayam . . . bhavotpattau na sambhavatlti, op . cit., Vol. I,
fol. 213.
122
Early Buddhist Theory of Knowledge
commentary offers two explanations of it. It is said that it may mean
‘those who are confused by their endless beliefs and words 5 . 1 The
alternative explanation is that amara stands for a species of fish, who
are in the habit of running about in the water, constantly emerging and
diving down so that it is difficult to get hold of them and that similarly
this theory (Scepticism) runs hither and thither without arriving at a
definite conclusion. 2 The fact that the commentator gives two alterna-
tive explanations of the word shows that he himself was uncertain
about its meaning. The latter is probably to be preferred since amara
as meaning endless (pariyantarahita-) is far-fetched. Vacavikkhepa- is
used as a synonym of amaravikkhepa-, 3 and probably means ‘verbal
jugglery 5 in view of the fact that these thinkers would have appeared
in the eyes of their opponents to evade committing themselves with
regard to the truth or falsity of a proposition. When Ajatasattu refers
to the theory as just vikkhepam he probably means the same, i.e.
‘jugglery 5 or ‘confusion 5 .
(159) The Buddhist texts refer to and briefly define the views of
different schools of Sceptics. They are spoken of collectively as ‘some
recluses and brahmins who wriggle like eels. For when a question is
put to them on this or that matter they resort to verbal jugglery and
eel-wriggling on four grounds 5 . 4 The first of these schools is described
as follows: ‘Herein a certain recluse or brahmin does not understand,
as it really is, that this is good or this is evil. And it occurs to him:
I do not understand what is good or evil as it really is. Not under-
standing what is good or evil, as it really is, if I were to assert that this
is good and this is evil, that will be due to my likes, desires, aversions
or resentments. If it were due to my likes, desires, aversions or resent-
ments, it would be wrong. And if I were wrong, it would cause me
worry (vighato) and worry would be a moral danger to me (antarayo).
Thus, through fear of being wrong (musavadabhaya) and the abhor-
rence of being wrong, he does not assert anything to be good or evil
and on questions being put to him on this or that matter he resorts to
verbal jugglery and eel-wriggling, saying: I do not say so, I do not
1 Amaraya ditthiya vacaya vikkhepo ti amaravikkhepo. DA. I.115.
2 Aparo nayo. Amara nama macchajati, sa ummujjana-nimujjanadivasena
udake sandhavamana gahetum na sakkoti. Evam eva ayam pi vado ito c’ito ca
sandhavati gaham na upagacchatl ti amaravikkhepo vuccati, loc. cit .
3 . . . vacavikkhepam apajjati amaravikkhepam, D. I.27.
4 Santi . . . eke samanabrahmana amaravikkhepika, tattha tattha panham puttha
samana vacavikkhepam apajjanti amaravikkhepam catuhi vatthuhi, D. 1 . 24.
I2 3
The Historical Background
say thus, I do not say otherwise, I do not say no, I deny the denials
{lit. I do not say, “no, no”)’. 1
(160) This description as it stands obviously admits of more than one
interpretation. On a very literal interpretation of it, it need not repre-
sent the point of view of a Sceptic at all, unless we mean by a Sceptic
a person who suspends his judgment on the truth or falsity of a
proposition, in the absence of adequate evidence or grounds for believ-
ing in its truth or falsity. If this class of thinkers merely refused to
assert that something was positively good or evil, unless they had
objective grounds for doing so, without being misled by subjective
bias due to their likes and dislikes, they would not be Sceptics but
critical thinkers recommending the outlook of science or intelligent
commonsense. The only difference from scientific scepticism, which
advocates the suspension of judgment in the absence of good evidence
or valid grounds for asserting the truth or falsity of a proposition,
would be that these thinkers did not merely consider that it was
intellectually unsatisfactory not to suspend judgment under such
circumstances, but that it was a moral danger (antarayo) as well not
to do so.
(161) Such an evaluation, however, is prima facie improbable. For if
they suspended judgment only until knowledge was possible without
ruling out the possibility of knowledge altogether, they would not
have been known to their opponents as having persistently refused to
commit themselves by asserting or denying all the logically possible
alternatives at least in respect of ethical propositions. It is, therefore,
very probable that they not merely denied knowledge of ethical pro-
positions but claimed that such propositions were, in principle,
unknowable, and that if we held that such propositions were either
true or false, as the case may be, we would be guided by our prejudices.
1 Idha . . . ekacco samano va brahmano va idam kusalan ti yathabhutam na
ppajanati, idam akusalan ti yathabhutam nappajanati. Tassa evam hoti: Aham
kho idam kusalan ti yathabhutam nappajanami, idam akusalanti yathabhutam
nappajanami. Ahan c’eva kho pana idam kusalan ti yathabhutam appajananto,
idam akusalan ti yathabhutam appajananto idam kusalan ti va vyakareyyam,
idam akusalan ti va vyakareyyam, tattha me assa chando va rago va doso va
patigho va tarn mam’assa musa. Yam mam’assa musa so mam’assa
vighato. Yo mam’assa vighato so mam’assaa ntarayo ti. Iti so musavadabhaya
musavadaparijeguccha n’ev’idam kusalan ti vyakaroti na pana idam akusalan ti
vyakaroti, tattha tattha panham puttho samano vacavikkhepam apajjati amaravik-
khepam: Evam pi me no. Tatha ti pi me no. Annatha ti pi me no. No ti pi me
no. No no ti pi me no ti. D. I.24-5.
124
Early Buddhist Theory of Knowledge
Hence it would be wrong on our part to make these assertions or these
assertions (or denials) would be false. Although their scepticism, with
regard to ethical propositions (what was good and evil) is specifically
referred to, it is not unlikely that their scepticism extended to all
propositions and that they denied the possibility of any knowledge
whatsoever in view of the fact that they are said to have given sceptical
answers ‘when questioned on this or that matter’ (tattha tattha panham
puttho, loc. cit .).
(162) This school of Sceptics is differentiated from the others on the
grounds that they adopted scepticism ‘through fear or aversion to
asserting what was false (musa-vadabhaya musavadaparijeguccha)’,
since what was asserted (or denied) would be false if the assertion was
due merely to one’s likes or dislikes. Strictly speaking, however, an
assertion made out of subjective bias need not necessarily be false,
although it would be wrong to make the assertion unless there were
good grounds for doing so. Therefore what is probably meant in this
context is that in the absence of objective criteria for judging what was
good or evil (or for asserting any proposition) we are led to hold some
view or another out of subjective bias and that this is wrong. Whatever
the explanation may be, it was the fear of believing in a proposition
out of prejudice in the absence of certain knowledge that made them
Sceptics. Their Scepticism is therefore due primarily to intellectual
reasons but from the account given of it a moral reason was also
present in that they hold that doing the wrong thing or uttering a false-
hood could cause worry or remorse (vighato) and be a moral danger
(antarayo) as well.
(163) This shows that despite their scepticism with regard to the
objectivity or the knowability of moral judgments, they held certain
subjective traits to be desirable. The commentary explains antaraya-
as ‘a hindrance to heaven or salvation’. 1 If this comment is relevant
then this class of thinkers were not purely intellectual Sceptics but seem
to have adopted scepticism on the grounds that knowledge was not
only impossible but was a danger to moral development and salvation,
a view which may have influenced Buddhism in regard to its attitude
to the ‘indeterminate questions’ (avyakatas) (y. infra , 813). In the
Sutrakrtanga, it is said that the Sceptics (Ajnanikavadah) along with
the other three main philosophical schools (Kriyavada, Akriyavada,
Vainayikavada) ‘teach final beatitude and final deliverance’ (2.2.79,
1 Saggassa c’eva maggassa ca antarayo, DA. 1 . 155.
The Historical Background 125
SBE., Vol. 45, p. 385). As Jacobi points out, in the commentary on the
Uttar adhyay ana Sutra (18.23), it is said that the Sceptics (Ajnanava-
dinah) ‘contend that knowledge is not necessary for salvation but
tapas x is, and Jacobi commenting on this says that ‘this seems identical
with the karmapatha’. 1
(164) The sacrificial brahmins of the Brahmanas proper, who continued
in the Upanisadic period recommending the path of action (karma-
marga-), were undoubtedly against the claims to metaphysical and
intuitive knowledge on the part of the Upanisadic thinkers. However,
there is no reason to think that they were sceptics or agnostics in the
matter of knowledge. They certainly claimed the veracity of certain
ethical propositions. The modified theory of the path of action (karma-
marga- or karmapatha-) to be found in the Isa Upanisad and later in the
Bhagavadglta is less averse to knowledge. The Isa, while condemning
those who delight in knowledge as being in greater darkness than
those who are ignorant , 2 nevertheless speaks of some kind of know-
ledge with which immortality is obtained (vidyayamrtam asnute, op .
cit., 11). Barua calls the thinkers of the Kena Upanisad, viz. the Keni-
yas, ‘sceptics’ (op. cit., pp. 261, 319). However, they were properly
agnostics, who denied the possibility of conceptual or sensory know-
ledge of reality 3 while not denying that reality ‘was known by an
awakening’ (pratibodhaviditam 4 ). Likewise, faith (sraddha) in the
Upanisads was never divorced from knowledge and there is no
evidence of a faith movement in the Upanisads, which decried know-
ledge. It is those who have both knowledge and faith 5 who attain
immortality. Moreover, faith (sraddha) is said to accompany both tapas
and knowledge . 6
(165) It is not intrinsically impossible that there were a set of thinkers
in the Vedic tradition who, because they believed in the efficacy of
1 SBE., Vol. 15, p. 83, fn. 2.
2 Tato bhuya iva te tamo ya u vidyayam ratah, 9.
3 Na tatra caksur gacchati na vag gacchati na manah na vidmo na vijanimo
yathaitad anusisyat, 3; cp. 7; 2.2.3.
4 2.4. Radhakrishnan translates the phrase as ‘when it is known through every
state of cognition" and quotes in support a cryptic comment of Sankara (bodham
bodham prati viditam) but this surely contradicts what is stated one verse earlier,
namely that ‘it is not understood by those who understand it" (avijnatam vijana-
tam, 2.3).
5 Brh. 6.2.15, te Y a evam etad viduh , ye cam! aranye sraddham satyam
upasate, . . .
6 Mund. 1. 2. 1 1, tapah sraddhe ye hy upavasanty aranye santa vidvamso . . .
126 Early Buddhist Theory of Knowledge
rituals and the value of faith, pronounced that knowledge, whether
empirical, metaphysical, or intuitive was a hindrance to moral progress
and salvation. But their existence is not testified in the literature itself
and it would be methodologically unsound to believe in their existence
since we would have to depend on an argumentum e silentio . Even if
they existed without leaving a trace of their doctrines in the Vedic
literature, they cannot be identified with the first school of Sceptics as
we know them from the passage under discussion. For in that case
they should not declare the impossibility of knowledge but merely
denounce it as morally dangerous. As we have shown ( v . supra> 148-
150) it is unlikely that this first school of Sceptics would have come
into being before the air was polluted (or rather enriched) with a
multitude of contending theories. Thus people came to feel on the one
hand that knowledge was uncertain and on the other that claims to
knowledge were morally dangerous in that one might believe in what
was in fact untrue and/or lead a factious and contentious life engaged
in dispute and debate in defending one’s beliefs. It is in such circum-
stances that we can expect to find an intellectual scepticism at the
theoretical level coupled with the practice of the good life as it was
traditionally known or best understood at the time.
(166) The description of the second school of Sceptics is almost
identical with that of the first except for the difference that according to
these thinkers, to be led to believe in a proposition by one’s likes,
desires, aversions and resentments would be an entanglement (upada-
nam, D. 1 . 25-6). Such entanglement would be a source of worry
(vighato) and as such a moral danger (antarayo). Upadana- literally
means ‘grasping’ or ‘clinging’ (PTS. Dictionary, s.v. sense 2) but since
these words express a pro-attitude 1 in that we grasp what we like or
desire but not what we hate or are averse to, it would be better to
translate the word as ‘entanglement’ or ‘act of involvement’. For it is
obviously intended to include the objects that we like as well as dislike.
Prof. Rhys Davids translates the word as ‘grasping condition of the
heart which causes rebirth’ (op. cit., p. 38) but this, as standing for a
concept of the Sceptics, need not, and indeed cannot, from the context
have the same technical significance as it has in Buddhism. In the
Buddhist context the word commonly means ‘the entanglement or
involvement that leads to becoming or survival in the next life’
1 On the use of this word see P. H. Nowell-Smith, Ethics , Penguin Books,
pp. 1 12-21.
The Historical Background 127
(upadanapaccayabhavo, D. 11-57) but there is no need to believe that
these Sceptics, would have had such a definite belief in survival or
rebirth, though they would not have dismissed the possibility. If as
was said in the Sutrakrtanga, the Sceptics too entertained beliefs about
heaven and salvation (v. supra , 613) it is possible that they held them
on pragmatic grounds without claiming actual knowledge. For, if
Silanka’s observations are correct, a favourite dictum of theirs was
‘Of what use is this knowledge’ (kim va tena jnanena? v. supra , 155)
as they did not believe that claims to knowledge had any pragmatic
value. However, based on what is implied from the context, the more
probable explanation is that this school of Sceptics merely considered
it undesirable to be involved in beliefs based on one’s likes or dislikes.
They held this view not on the grounds that such involvement would
lead to rebirth or survival but rather because such beliefs would be a
source of worry and mental disquietude (vighata-). In any case, it is
clear that this school of thinkers, unlike the first, adopted Scepticism
primarily out of moral considerations rather than for intellectual
reasons although the latter were not absent.
(167) The next school of sceptical thinkers is said to argue as follows:
T do not know, as it really is, what is good and what is evil and not
knowing, if I were to pronounce that this is good or this is evil, then
I would have to join issue, argue and debate with recluses and brahmins,
learned, subtle, hair-splitters, skilled in controversy, who go about
debunking with their intellect the theories of others. If I were to join
issue, argue and debate with them, I would not be able to explain to
them. If I were unable to explain to them, that would cause me worry
(vighata-) and be moral danger (antarayo).’ Thus because he fears and
detests interrogation (anuyoga-) 1 he does not ‘pronounce this to be
good nor that to be evil’. 2
1 This is a technical term associated with the debate defined in the Caraka
Samhita, v. infra, 322.
2 Aham kho idam kusalan ti yathabhutam nappajanami, idam akusalan ti yatha-
bhutam nappajanami. Ahan c’eva kho pana idam kusalan ti yathabhutam appa-
jananto, idam akusalan ti yathabhutam appajananto idam kusalan ti va vyakarey-
yam — santi hi kho pana samanabrahmana pandita nipuna kataparappavada
valavedhirupa vobhindanta marine caranti pannagatena ditthigatani -te mam
tattha samanuyunjeyyum samanugaheyyum samanubhaseyyum. Ye mam tattha
samanuyunjeyyum samanugaheyyum, tesaham na sampayeyyam. Yesaham na
sampayeyyam so mam’assa vighato so mam’assa antarayo ti. Iti so anuyogabhaya
anuyogaparijeguccha nYv’idam kusalan ti vyakaroti, na pan’idam akusalan ti
vyakaroti . . D. I.26.
128 Early Buddhist Theory of Knowledge
(168) This passage is again obscure. As it stands it merely states that
there were a set of thinkers, who, because of their ignorance of the
truth or falsity of moral propositions, did not desire to engage in
debate with skilled dialecticians. For they would not be able to con-
vince them of their scepticism and this would be a source of worry
and a moral danger to them.
(169) What is not clear from the above account is whether they
wished to avoid debate because they were Sceptics or whether they
adopted Scepticism because they wanted to avoid debate. The passage
seems to be suggesting or saying both at once. On the one hand it
seems to be saying that these thinkers ‘do not know’ that something is
definitely good or evil and that their scepticism leads them to avoid
debate, while on the other hand it seems to be saying that they do not
want to ‘pronounce that this is good or this is evil’ because they feared
debate. In the former case, as Sceptics, they would probably have found
that they partially agreed with any or every thesis that their opponents
put forward except of course the thesis that ‘there is knowledge’. They
had no particular thesis of their own that their opponents could dis-
prove unless it be their scepticism itself. Thus it would have been
difficult to convince their opponents of their scepticism inasmuch as
according to the current rules of debate it was required that one party
put forward a definite proposition to be proved ( v . infra, 344). In the
latter case, they would have adopted scepticism either because they saw
the futility of debate where skilled dialecticians could apparently prove
thesis as well as anti-thesis and/or because they saw the moral dangers
of debate since debates resulted in the defeat of one party or the other
and frayed tempers as well. They would have seen that there was no
point or purpose in debate since one was nowhere nearer the truth at
the end of it and at the same time feared debate because it could result
in loss of their mental equanimity which they valued. This seems to be
the more probable explanation judging from what we learned about
these sceptics from Silanka’s account.
(170) We have hitherto spoken of three schools of Sceptical thinkers,
namely those who adopted scepticism primarily through fear of false-
hood (musavadabhaya), through fear of involvement (upadanabhaya)
or fear of interrogation in debate (anuyogabhaya). All three schools
considered the consequences of falsehood, involvement and interroga-
tion psychologically undesirable in that they cause remorse or worry
(vighata-), which was a (moral) danger or hindrance (antaraya-). It
The Historical Background 1 29
seems to be clear from this that there was much in common between
these three schools or types of Sceptics and that they valued mental
stability if not the cultivation of some sort of ideal state of mind.
(171) Apart from the few hints that we get about these Sceptics from
Buddhist and Jain sources, we have not been able to trace any positive
reference to them elsewhere in Indian thought. But the account given
of the scepticism of Pyrrho, who is said to have been influenced by
Indian thought, 1 bears a remarkable similarity to the point of view
of these Sceptics. The quotation preserved by Aristocles from one of
Timon’ s prose works and which is supposed to represent the views of
Timon’s teacher, Pyrrho, reads as follows: ‘He himself (Pyrrho) has
left nothing in writing but his disciple Timon says that the man who is
to be happy must look to these three things: (1) what is the nature of
things, (2) what attitude should we take to them, and (3) what those
who take this attitude will gain by it. He says that he declared that
things were in an equal degree indifferent and unstable and incapable
of being tested. For this reason neither our senses nor our opinions are
true or false. So we must not put our trust in them but be free from
beliefs and inclinations and unshaken, saying of each thing in turn that
it no more is than it is not or that it both is and is not or that it neither
is nor is not. And those who take this attitude, Timon says will first
gain speechlessness ( abasia ) and then imperturbability (arapa^ta)’. 2
(172) One difference that we seem to observe on the surface is that
Pyrrho’s scepticism appears to be all-embracing while the scepticism
of the three schools outlined above seems to have been more or less
confined to moral propositions. But this appearance is deceptive. With
regard to all three of these schools, it is stated, that they gave sceptical
answers ‘when questioned on each and every matter’ (tattha tattha
panham puttho, D.I., pp. 24, 25, 26). The reason for high-lighting
the ethical examples was probably due to the Buddhists themselves
being mainly concerned with this aspect of their teachings just as much
as the account given by Silanka of the different ‘types’ of Sceptics
(v. supra , 157) gave one the impression that the Sceptics were mainly
interested in the concepts and categories of Jainism.
(173) If we compare the doctrines of the above three schools of
Sceptics with the account given above of the scepticism of Pyrrho, it
1 ERE., Vol. 11, p. 228, v. ‘Sceptics’; he is said to have ‘studied philosophy
under Indian Gymnosophists and Chaldean Magi’.
2 Loc. cit. } p. 229.
E
130 Early Buddhist Theory of Knowledge
would be interesting to observe that both hold that (1) there were no
beliefs or opinions which were true or false and therefore (2) we should
give no positive answer to any of the logical alternatives. It would also
be seen that (3) the four logical alternatives mentioned in Timon’s
account (i.e. is, is not, both is and is not, neither is nor is not) are
identical with that of Sanjaya, the Buddhists and perhaps also of the
three schools of Sceptics as we have shown below (v. infra, 184).
Lastly (4) the value of the sceptical attitude is said by Pyrrho to lie in
the fact that it promotes speechlessness (aphasia) and mental imper-
turbability (ataraxia), which seem to be the states of mind regarded as
ideal by the above schools of Sceptics since they held that anything
that caused mental instability was a hindrance. Because of Pyrrho’s
love of quietism, Burnet 1 thinks that Pyrrho is more of a quietist than
a sceptic and is inclined to regard him as being nearer the Buddhist
ideal: ‘We see that those who knew Pyrrho well described him as a
sort of Buddhist arhat and that is doubtless how we should regard him.
He is not so much a sceptic as an ascetic and a quietist’. But when we
see that Pyrrho’s scepticism as well as his quietism are shared by the
above schools of Sceptics, it would be more appropriate to regard him
as having a closer kinship with them rather than with the Buddhists,
who were opposed to their scepticism.
(174) Barua compares the school of Pyrrho with that of Sanjaya {op.
cit p. 32) but as indicated below (v. infra, 180) it is not said of the
school of Sanjaya, unlike in the case of the previous three schools, that
it held non-scepticism to be a source of vexation or a hindrance. We
therefore have no evidence that the school of Sanjaya valued mental
equanimity. In the circumstances we would have to hold that Pyrrho-
nean scepticism would be nearer the three schools mentioned above
than the school of Sanjaya, which in our opinion does not seem to have .
valued mental quietude at all.
(175) The fourth school of Sceptics is described in language identical
with that used to define the philosophy of Sanjaya so that we may
presume that Sanjaya was one of the foremost representatives if not
the leader of this school. Sanjaya is described along with the other five
teachers, who were contemporaries of the Buddha as being a well-
known (nato), celebrated (yasassi) teacher and a leader of a sect
(titthakaro) who was held in high esteem by the common folk (sad-
husammato bahujanassa); he is also said to have a following (sanghl
1 Loc. cit., p. 229.
The Historical Background 13 1
gani). This description occurs in a stereotyped sentence, 1 which is used
of all the six heretical teachers several times in the Nikayas (D. 1. 150,
M. I.4, S. 1.68, J. I.509). But since it is complimentary we need not
doubt that it contained a good deal of truth. He may have been
Sariputta’s teacher prior to the latter’s conversion to Buddhism. 2
Barua doubts this. He says that ‘one may reasonably object to the
identification of Sanjaya the Sceptic, who is designated in the Samanna-
phala Sutta as Sanjaya Belatthaputta (or Belatthiputta) with Sanjaya
described in the Vinaya Mahavagga and the Dhammapada commentary
as a Paribbajaka’ (op. cit p. 325). However, if the commentarial
tradition that Suppiya Paribbajaka was a disciple of Sanjaya 3 is correct,
we need not doubt that Sanjaya himself was a Paribbajaka and was
sometimes known as such.
(176) The sceptical philosophy of this school is defined as follows:
‘Herein a certain recluse or brahmin is dull, stupid. And by reason of
his dullness and stupidity, when questioned on this or that matter, he
resorts to verbal jugglery or eel- wriggling: “If you ask me whether
there is a next world, then if it were to occur to me (iti ce me assa) that
there is a next world, I would pronounce that there is a next world.
Yet, I do not say so, I do not say thus, I do not say otherwise, I do not
say no, I deny the denials. Similarly with regard to the propositions,
“there is no next world”, “there is and is not a next world”, “there
neither is nor is not a next world”, “there are beings who survive
(death)”, “there are no beings who survive”, “there are and are no
beings who survive”, “there neither are nor are there no beings who
survive”, “there is a result and a consequence of good and evil
actions”, “there is no result or consequence of good or evil actions”,
“there is and is no result or consequence of good or evil actions”, “the
Perfect One (Tathagato) exists after death”, “the Perfect One does not
exist after death”, “the Perfect One both exists and does not exist after
death”, “the Perfect One neither exists nor does not exist after death”’. 4
1 sarighino ganino ganacariya nata yasassino titthakara sadhusammata ca
bahujanassa, loc. cit.
2 v. Malalasekera, DPPN., s.v. Sanjaya Belatthiputta.
3 Paribbajako ti Sanjayassa antevasi, DA. I.35.
4 Idha . . . ekacco samano va brahmano va mando hoti momuho. So mandatta
momuhatta tattha tattha panham puttho samano vacavikkhepam apajjati amaravi-
kkhepam: ‘ Atthi paro loko?’ ti iti ce mam pucchasi, ‘Atthi paro loko* ti ce me
assa, “atthi paro loko” ‘ ti iti te nam vyakareyyam. Evam pi me no. Tatha pi me no.
Annatha pi me no. No ti pi me no. No no ti pi me no. “Natthi paro loko?” ti . . .
pe . . .’ Atthi ca natthi ca paro loko? N’ev’atthi na n’atthi paro loko? Atthi satta
132 Early Buddhist Theory of Knowledge
(177) Professor Basham, dissenting from Barua’s view that the passage
mentioned above represents a doctrine that was held in good faith by a
school of Pyrrhonists, thinks that the passage is ‘probably satirical, a
tilt at agnostic teachers who were unwilling to give a definite answer to
any metaphysical question put to them’ (op. cit ., p. 17). He adds that
‘its agnosticism was never a part of the Ajivika creed, and it may be
omitted from further consideration’ (loc. cit.).
(178) This seems to be unclear. For it may be asked whether (1) there
was at least one school of Sceptics, (2) Sanjaya was a Sceptic, and (3) if
Sanjaya was a Sceptic, he was an Ajivika. If (2) and (3) are denied there
is a contradiction inasmuch as elsewhere Professor Basham ascribes
the ‘sceptical philosophy’ outlined in the Sandaka Sutta to Sanjaya
(op. cit., p. 19). Moreover, he states (y. infra , 195) that ‘the Sandaka
Sutta seems to embrace all six of the heretical teachers ... in the
category of Ajlvikas’ (op. cit., p. 96). From this it may be inferred that
he considers ‘agnosticism’ or ‘scepticism’ (v. op. cit., p. 19 — both words
are used indiscriminately) as part of the Ajivika creed(s) in Barua’s
second sense of the term Ajivika, 1 with which Basham agrees. 2
Nevertheless, at the beginning of his work he states the very opposite.
Moreover, Professor Basham omits to discuss Sllanka’s statement to
the effect that ‘the Ajlvikas and others, who are followers of Gosala’s
doctrines are a product of ajnanavada’ (supra, 147), whatever ajnana-
vada may mean here.
(179) We are not anxious to prove that scepticism is part of the
Ajivika doctrines or not, though we would like to be clear about the
use of the term Ajivika, so as to avoid confusion (v. infra, 196). But if
Prof. Basham is saying that there was no school (or schools) of Sceptics,
but only ‘agnostic teachers’ who were sceptics only with regard to
metaphysical questions, it is necessary to urge that, as we have shown
above, the independent evidence and testimony of both the Buddhist
as well as the Jain texts seem to point in the opposite direction. But
opapatika? N’atthi satta opapatika? Atthi ca natthi ca satta opapatika? N’ev’atthi
na natthi satta opapatika. Atthi sukatadukkatanam kammanam phalam vipako?
Natthi sukatadukkatanam kammanam phalam vipako ? Atthi ca natthi ca sukata-
dukkatanam kammanam phalam vipako? N’ev’atthi na n’atthi sukatadukkatanam
kammanam phalam vipako? Hoti Tathagato parammarana? Na hoti Tathagato
parammarana? Hoti ca na hoti ca Tathagato parammarana? N’eva hoti na na hoti
Tathagato parammarana? ... D. 1 . 27.
1 ‘Ajivika — what it means’, ABORT, Vol. 8, 1927, p. 183.
2 Op. cit., pp. 96, 97. ‘We have seen that the second usage is very common in
early Buddhist literature’ (p. 98).
The Historical Background 133
Sanjaya’s scepticism may very well have been confined to metaphysical
questions, as we ourselves consider to be a possibility following the
suggestion of Jacobi (v. infra, 181).
(180) We are inclined to agree with Basham when he says that the
above passage is ‘satirical’ but only in the sense that it seems to give a
rather inexact version of the philosophy of Sanjaya, to whom the
Buddhists seem to have been somewhat antipathetic. In fact, the
account given gives the impression that Sanjaya was a naive Sceptic,
who adopted Scepticism out of sheer stupidity, either because he did
not know the answers to the questions put to him or the fact that one
of the logical alternatives must be true. Both the Brahmajala Sutta and
the Samannaphala Sutta (D. 1 . 58-9) versions emphasize the dullness
and stupidity of this thinker as a result of which scepticism is the out-
come. In the Brahmajala Sutta, out of the sixty two philosophical
schools, whose views are stated, this is the only one that is picked out
as being ‘a product of sheer stupidity’ (mandatta momuhatta, loc. cit .).
In the Samannaphala Sutta, it is stated as the impression of Ajatasattu
that Sanjaya ‘was the most foolish and stupid’ 1 of all the recluses and
brahmins. If Sariputta, who is lauded for his intelligence, could have
been at one time the disciple of Sanjaya, 2 Sanjaya could not have been
as stupid as he is made out to be and besides he would not have
attracted such a large following. What then could be the motive for
singling out this particular school of Scepticism as a product of folly?
One difference that we notice is that in the former three schools of
Sceptics there seems to be some conception of the good life, whether
they believed in salvation or not, and the sceptical attitude seems at
least to have been regarded as psychologically desirable in promoting
one’s peace of mind. Sanjaya on the other hand may have been a more
thorough-going sceptic, who made no pretence about the desirability
of scepticism as a way of life. He would thus have been much more
outspoken and critical of the views of his opponents. As a result the
Buddhists may have regarded him as being more deluded than the
other Sceptics who in spite of their theoretical scepticism had the good
sense to cultivate the tranquillity of mind, which was highly valued in
Buddhism as well.
(181) Sanjaya’s scepticism may have extended to the whole field of
knowledge for he too is supposed to have given sceptical answers to
1 Ayan ca imesam samanabrahmananam sabbabalo sabbamulho, D. I.59.
2 v. DPPN., s.v. Sariputta.
134 Early Buddhist Theory of Knowledge
all questions put to him. But if we examine the actual list of proposi-
tions mentioned (and here a whole list is given unlike in the case of the
previous schools), it would be seen that they pertain to metaphysics
(the next world, survival, transcendent existence) and morals (the
consequences of good and evil). It is therefore not improbable that his
scepticism was directed only to those questions, the answers to which
were unverifiable and therefore unknowable or as Jacobi says ‘trans-
cendent or beyond human experience 5 (< op . cit., p. xxvii). In this respect
the philosophy of Sanjaya may be compared with that of the positivist
branch of the Lokayata (i.e. group (2), v. supra , 89, 94), the difference
being that these positivists seem to have denied the truth of these
propositions or suggested that they do not make sense 1 while Sanjaya
seems to grant the possibility of their truth, though denying that we
have any means of knowing this. If so Sanjaya seems to have examined
the truth-value of propositions in the light of relevant evidence. His
philosophy, therefore, is as Ui sums it up ‘a scepticism on the one hand
and a primitive step of criticism of knowledge on the other hand, like
the sophists 2 in the Greek philosophy 5 . 3
(182) Another fact to be noted in the account given of Sanjaya’s
philosophy is that the propositions are arranged in a four-fold order
of expression and the logical alternatives are not confined to simple
assertions and denials. For instance, we find not only the expressions
‘there is (atthi) a next world 5 and ‘there is no (natthi) next world 5 but
also the forms ‘there is and is not (atthi ca natthi ca) a next world 5 and
‘there neither is nor is there no (n 5 ev 5 atthi na n’atthi) next world 5 . This
four-fold mode of expression, as we have shown later (r. infra, 581)
appears to have been adopted in the Pali Nikayas alongside the usual
two-fold mode. Keith gives the credit to Sanjaya for initiating this
new four-fold logic: ‘he seems as an agnostic to have been the first to
formulate the four possibilities of existence, non-existence, both and
neither . . . 54
(183) This is certainly a possibility that cannot be ruled out. His very
scepticism may have led him to include the modes of expression ‘both
is and is not 5 and ‘neither is nor is not 5 , both of which are expres-
sions sometimes used in common speech in addition to the ordinary
1 The closest that the early Materialists came to saying this is their statement
recorded in the Sutrakrtanga (v. supra, Ch. II, p. 46), viz. ‘it is those who say
that the (soul) does not exist or is not evident, who would be making the right
statement about it*. 2 v. infra, 326.
3 The Vaisesika Philosophy, p. 23.
Buddhist Philosophy, p. 303.
The Historical Background 135
assertions and denials, so as to make his scepticism and his scruples for
truth appear more comprehensive. On the other hand there are two
other alternatives worth considering, which appear to be more plaus-
ible than the one suggested.
(184) One of the alternatives is that the four-fold schema was not the
innovation of Sanjaya but was held in common by all the schools of
the Sceptics; in such a case Buddhism would have either borrowed this
classification from the Sceptics or shared it with them. The other
alternative is that the innovation was on the part of the Buddhists and
that the Sceptics themselves including Sanjaya were not concerned
with such problems.
(185) Let us consider the first alternative. We have already seen how
Silanka arranged the only example he seems to have taken from the
Sceptics themselves in a four-fold schema (y. supra , 157). He may have
done this merely to complete his figure of sixty seven ‘categories’ of
Sceptics: but the fact that the example he took was not based on Jain
concepts and his own admission that this particular example did not
admit of more than a four-fold order of predications are possibly
pointers to the fact that he was borrowing not only the example but
the four-fold formula itself from the Sceptics. The adoption of such
different schemas was perhaps characteristic of this period and was
probably necessitated by the variety of doctrines, which had to be
considered apart (y. infra , 573). Dr Basham has some evidence to show
that the Ajlvikas under Makkhali Gosala and the schismatic Jain sect
of the Trairasikas adopted a scheme of classifying propositions into
three logical ‘heaps’ (rasi) or categories (v. infra , 217-20). Now there
is no known school of Indian thinkers apart from the Buddhist (barring
Sanjaya who is known only through the references in Buddhist litera-
ture) who adopted a four-fold schema, but the five-fold formula of
denial, which according to the accounts given, is common to all the
schools of Sceptics, seems to be based on the acceptance of a four-fold
form of predication. Let us examine this formula:
1. Evampi me no — I do not say so.
2. Tatha pi me no — I do not say thus.
3. Annatha pi me no — I do not say otherwise.
4. No ti pi me no — I do not say no.
5. No no ti pi me no — lit. I do not say ‘no, no.’
The commentary offers two explanations 1 of the meaning of this
1 DA. I.115-6.
136 Early Buddhist Theory of Knowledge
formula. According to the first explanation, proposition (1) is an
indefinite rejection or denial (aniyamitavikkhepo. DA. I.115). Pro-
position (2) is the denial of a specific proposition, e.g. the denial of the
eternity view (sassatavada-) when asked whether the world and the
soul are eternal. Proposition (3) is the denial of a variant of (2), e.g. the
rejection of the semi-eternal theory (ekaccasassatam), which is said to
be somewhat different from (annatha) from the eternity theory.
Proposition (4) is the denial of the contrary of (2), e.g. the denial of
the annihilationist theory (ucchedam) when asked whether a being
(tathagato) does not exist after death. Proposition (5) is the rejection
of the dialectician’s view (takkivadam) 1 of a double denial, e.g. denying
the position if asked whether a being neither exists nor does not exist
after death.
(186) If we adopt the notations p. notp, 2 p.notp, not (p.notp) to
represent the usual four-fold propositional formula of predication in
Buddhism (i.e. corresponding to, is, is not, is and is not, neither is nor
is not), we may represent the above commentarial explanation in
symbolic form as follows, using the notation p= for an indefinite
proposition; and— -to express denial:
r - -(P=)
2. — (p)
3. -(p.notp)
4. —(notp)
5. — (not(p.notp))
It will be seen from the above that 2, 4, 3 and 5 (in this order) are the
denials of the usual four propositional types in the order in which we
stated them. The identifications of the commentator has some basis in
the wording. He has seen that there was a double ‘na’ (two ‘nots’) in 5
{no no ti pi . . .) and identified it as a denial of a proposition of the form
na eva . . . na (na) . . . Where he saw a single ‘na’ in 4 (no ti pi . . .)
he identified it as a denial of a proposition of the form ‘na . . .’. He then
identified 2 (tatha pi . . .) as the denial of a simple assertion. In 4
(annatha pi . . .) he saw a slight variation of 2 and identified it as a
denial of a partial assertion. But what is most unsatisfactory is the
1 This is called takki-vadam, i.e. the thesis of the sophist (vitandavadin) since
he rejects or argues against both thesis as well as anti-thesis; he neither asserts
p nor not-p.
2 We are not using the negation sign (~p) or the form ‘not p’ since ‘notp’ is
not the contradictory of p (v. infra, 575).
137
The Historical Background
identification of i. What does Buddhaghosa mean by an aniyamita-
vikkhepo? He can only mean the rejection of any one of the logical
alternatives (which he has exhausted in 2, 4, 3 and 5) without specifica-
tion, but the language (evampi me no) hardly suggests this.
(187) According to the second explanation, proposition (1) is the
denial of an assertion e.g. if asked whether this is good, he denies it.
Proposition (2) is the denial of a simple negation, e.g. if asked whether
this is not good, denies it. Proposition (3) is a denial that what you are
stating is different from both (1) and (2), e.g. if asked whether his
position is different from both (1) and (2) (ubhayato annatha), denies
it. Proposition (4) is a denial that you are stating a point of view
different from the above, e.g. if asked whether his thesis (laddhi) is
different from the three earlier points of view (tividhena’pi na hoti),
denies it. Proposition (5) is a denial of the denials, e.g. if asked whether
his thesis is to deny everything (no no te laddhi ti) he denies it. Thus
he does not take his stand (na titthati) on any of the logical alternatives
(ekasmim pi pakkhe). We may represent this explanation using sym-
bols as explained above, as follows:
i- — (p)
2. — (notp)
3 - -(— (b 2 ))
4* “ (— (1, 2, 3))
5- — 0, 2, 3, 4)
We have used the numerals as well, as equivalent to the formula that
follows for otherwise the notation would appear too complicated and
the point of adopting it would be lost. Thus, 3 means that ‘you deny
that your thesis is different from both your previous denial of the
assertion and the denial of the negation’. The purpose of the com-
mentator in both his explanations has been to show that the Sceptic
does not take his stand on any of the logical alternatives. But this
second explanation is less satisfactory than the first, since the explana-
tions of propositions 3 and 4 appear to be very arbitrary and hardly
related to the language used.
(188) We would like to suggest a third alternative explanation, which
has the merit of being the simplest and the one having the closest
affinities to the language used. Buddhaghosa’s second explanation
made the suggestion that the last proposition (no no ti pi me no) is a
denial of the rejection of all the possible logical alternatives. This
appears to be plausible since the statement literally means ‘I do not
138 Early Buddhist Theory of Knowledge
say “no, no” \ If so the others constitute the rejection of the usual
four-fold logical alternatives. We may take them in the usual order
and when we do so it would be noticed that on the whole they corre-
spond with the language used. We may state this explanation in
symbolic form, as follows:
1. -Cp)
2. — (notp)
3. — (p.notp)
4. — (not(p.notp))
5 - — (1, 2, 3, 4)
(189) If this five-fold formula of denial implies or is based on the four-
fold modes of predication of logical alternatives, then in the light of
the independent evidence from Sllanka as well (v. supra , 157), the
credit for adopting this schema should not go to Sanjaya alone, but
should be shared by all these Sceptical schools of thought.
(190) It also appears to be equally plausible that it were the Buddhists
who were the first to innovate and adopt this four-fold schema. We
noticed that when Sllanka tried to explain the existence of sixty-seven
‘categories’ of Sceptics, he did so by making them ask questions
according to the seven-fold mode of predication (saptabhangi) adopted
by the Jains. From this we cannot argue that the Sceptics were the first
to adopt the saptabhangi formula, as Keith has done in the case of
Sanjaya on precisely the same kind of evidence. In order to explain
their sceptical attitude it was necessary for their opponents to represent
them as not committing themselves on any one of the logical alterna-
tives and it is natural for them to do this by showing them as dismissing
the logical alternatives as they themselves understood them. It is
therefore not surprising that the Jains should represent them as dis-
missing a proposition in all the seven modes of predication known to
them, while the Buddhists picture them as discarding the four. Both
these alternatives are more plausible than the one that Keith has
offered in that they have some independent evidence to confirm them.
We cannot therefore agree with Keith, when he dogmatically gives the
credit to Sanjaya for being the ‘first to formulate the four possibilities’
(Joe, cit . ), when we know nothing about Sanjaya apart from the
accounts we get of him in the Buddhist texts.
(191) Jacobi thinks that ‘in opposition to the Agnosticism of Sanjaya,
Mahavlra has established the syadvada’ (op, cit,, p. xxvii). Superficially,
The Historical Background 139
there seems to be some truth in this observation. The Jain syadvada
appears to be the opposite reaction to that of the Sceptics when faced
with the same epistemological problem. The Sceptic doubts or denies
all the logical possibilities, whereas the Jain asserts that they are all
true in some sense or another. But this appearance of a radical contrast
is deceptive and in fact although the two have to be distinguished, it
would be quite wrong to consider them as being poles apart.
(192) The Buddhist in depicting the Sceptic as denying all the logical
possibilities and denying these denials as well, has not given an accurate
account of the point of view of the Sceptic in his anxiety to show that
the latter is making self-contradictory assertions. It would appear that
in denying the denials (no no ti pi me no, loc . cit.) the Sceptic was
contradicting himself, but in fact he does not seem to have denied the
possibilities outright. He would most probably have merely stated that
he does not agree that p is the case quite categorically (as his opponent
would have liked him to), since p may be true or p may be false and
one cannot know this. This is different from a categorical denial of the
possibilities. The position of the Sceptic would in fact be disclosed as
follows:
1. p may (or may not) be the case
2. Notp „
3. p.notp
4. Not(p.notp) „
(193) We may compare this with the standpoints of the Jains, which
we may state as follows, confining ourselves to the first four possi-
bilities only for the sake of the comparison:
1. p may be the case
2. Notp „
3. p.notp „
4. (p. is inexpressible) „
syadasti
syannasti
syadastinasti
syadavaktavyah
(194) It would now appear as if, far from being poles apart, it is
difficult to distinguish the two points of view. The difference is no
doubt there for, by say, syadasti, the Jains do not mean that ‘p may be
the case’ in the sceptical sense but that ‘p is in fact the case from a
certain point of view (naya)’ (v. infra , 236-8). Instead of one develop-
ing in opposition to the other as Jacobi has suggested the two seem to
have a common origin, though they part company at a certain point.
When both were faced with the problem of diverse theories (which
140 Early Buddhist Theory of Knowledge
could be stated as logical alternatives), both wondered whether any of
them could be true. But while the Sceptic seems to have concluded that
none of them could be known to be true, the Jain appears to have
formed the conclusion that each one may be true (v. infra , 236). In the
eyes of their opponents, both would have appeared to contradict them-
selves, the former by violating the Law of Excluded Middle (or rather
the Law of Exclusion, since there were more than two logical alterna-
tives, v . infra, 582) and the latter the Law of Contradiction (y. infra ,
582).
(195) Another influential class of religious teachers who made their
own contribution to the development of logical and epistemological
thought were the Ajivikas, who are distinguished from the Jains
(niganthas) in the Suttanipata (v. infra, 375). Although the word,
Ajivikah, was used primarily of the followers of Makkhali Gosala and
secondarily in a loose sense, as shown by Barua 1 and Basham, 2 there
is no evidence that the term was used of the Jains at least at the time
of the Pali Nikayas. Dr Basham’s deduction that ‘the Sandaka Sutta
seems to embrace all six of the heretical teachers, including the great
leader of the niganthas, Nigantha Nataputta or Mahavira, in the
general category of Ajlvikas’ (op. cit., p. 96) seems to be based on a
misinterpretation. Since we have inferred that some of the doctrines
criticized in the Sandaka Sutta are those of the Brahmanic tradition
(v. infra, 196) and Basham’s assumption that ‘the propagators of all the
objectionable teachings (i.e. in the Sandaka Sutta) are classed together
under the broad title of ajlvikas’ (op. cit., 20) would adversely affect
some of our own conclusions, it seems necessary to point out why and
where we differ from Professor Basham.
(196) The Sandaka Sutta criticizes four types of religions which are
false (abrahmacariyavasa) and four which are unsatisfactory (anassasi-
kam brahmacariyam) but not necessarily false. Nowhere in the Sutta
are these teachings associated with the names of individuals. It is
Basham who identifies these teachings with those of certain teachers
on the basis of the wording (op. cit., p. 19). He seems to have identified
‘the teacher claiming omniscience’ (loc. cit.) with Mahavira, for other-
wise he would not have come to the above conclusion. This identifica-
tion is arbitrary for the Sutta itself as we have said, mentions the
omniscient teacher as a type and the Pali Nikayas themselves refer to
1 B. M. Barua, ‘Ajivika-What it means’, ABORI., Vol. 8, 1927, p. 183.
2 Op. cit., p. 97 ff.
The Historical Background 141
both Purana Kassapa (v. infra, 383) as well as Nigantha Nataputta as
claiming omniscience. Secondly, Basham has refrained from identifying
the traditionalist (anussavika-) as well as the rationalist (takkl vimamsi).
Had he done so he would have found that the traditionalists were
mainly though not solely the Vedic brahmins as defined at M. II.21 1 —
santi, Bharadvaja, eke samana-brahmana anussavika . . . seyyatha’pi
brahmana Tevijja. The ‘rationalists’ as we have shown, were both
brahmins as well as Samanas (v. infra , 37 5). Now Basham says that
‘the conclusion of the Sutta is surprising’ (op. cit ., p. 19) but had he
made the above identification, to make his list complete, he would have
found that his own conclusion would have been still more surprising,
namely that even the Vedic brahmins would have to be called Ajlvikas
according to this Sutta. Thirdly, Dr Basham seems to have assumed
without justification that the ‘Ajlvika’ in the quotation on which his
entire conclusion is based, viz. Ime pan’ Ajlvika puttamataya putta,
attanan c’eva ukkamsenti pare ca vambhenti, tayo c’eva niyyataro
pannapenti, seyyath’Idam Nandam Vaccham, Kisam Sankiccam,
Makkhali Gosalan ti (loc. cit., fn. 7) are identical with the religious
teachers with whom he has associated the teachings mentioned in it.
But the context of this quotation seems to tell a different story. It
occurs in a digression at the end of the Sutta, when Sandaka Paribbajaka
asks Ananda a few questions, the last of which is, Klva bahuka . . .
imasmim dhammavinaye niyyataro ? How many saints 1 are there in
this religion? This question seems to have little to do with the earlier
sermon of Ananda on the different types of religions or religious
teachers. Ananda replies that there are over five hundred, to which
Sandaka Paribbajaka, who is probably an Ajlvika in the loose sense of
the term, says that as for the Ajlvikas ‘they (can) claim only three
saints’ (tayo c’eva niyyataro pannapenti). Lastly, Professor Basham
quotes Chalmers’s translation, which is inaccurate: ‘yet they have only
produced three shining lights’; even if we retain ‘shining lights’ as a
free rendering of niyyataro, which means ‘those who have attained
salvation’ (y. fn. below), pannapenti (=Skr. p raj napay anti, from pra-j-
Vjna+ causative suffix and not from pra--)- \/j an ) can on ly mean
1 Formed from nis +VV a T t T> it is the intransitive sense that is evident in the
usage; e.g. niyyanti dhira lokamha, the wise go out of the world, S.V.6 — in this
sense niyyataro would be ‘those who have gone out*, i.e. ‘the saved’, ‘the saints’;
cp. so niyyati . . . samma dukkhakkhayaya M. 1.68. If as the PTS. Dictionary
suggests (s.v.) the word is formed from nis + \/y am we should have niyyanta-,
niyyantaro on the analogy of ganta- from \/ gam. It cannot therefore mean
‘guide, leader’.
142
Early Buddhist Theory of Knowledge
‘proclaim’ or ‘claim’ (lit. make known). Thus the interpretation of
‘the omniscient teacher’ as a personal reference to the Jain leader
irrespective of the other claimants, the failure to see the references to
the Vedic religion and the Brahmanic tradition, the failure to note the
strict context of the quotation and perhaps even the mistranslation
seem to have led Dr Basham to the above conclusion, with which we
cannot agree for the reasons stated.
(197) In the general sense of the term Ajlvika-, even the Sceptics, whom
Silanka seems to have associated with them were Ajivikas. However,
since it is necessary to consider the doctrines of the Sceptics separately
we shall confine our usage of Ajivikas to denote those Samanas, who
were neither Jains, Materialists or Sceptics.
(198) To turn to our main problem, the Ajivikas seem to have been
influenced both by the rational tradition of the Early Upanisads as well
as by the claims to intuitive knowledge on the part of the Middle and
Late Upanisadic thinkers. One of their main metaphysical interests
seems to have been the problems of time and change. Basham does not
believe that Ajlvikism ‘derived from Vedic or Brahmanical sources’
(op. cit., p. 98) but the hymns to Kala in the Atharvaveda (19. 53, 54)
seem to contain the germs of the determinist thesis, if determinism
(niyati-) was one of the main doctrines of the Ajivikas. In these hymns,
Time (Kala-) conceived as an hypostatized entity having everything
under its control and ‘beyond which there is no other greater force’, 1
is said to have ‘produced both the past and the future’ 2 while it is itself
eternally existent. 3
(199) Now the main argument for niyati seems to have been based on
the same a priori premiss of Uddalaka (v. supra , 25), which led to
Metaphysical Materialism on the one hand (v. supra, 85, 115) and to
the proto- Vaisesika Realism of another Ajlvika thinker, Pakudha
Kaccayana (v. infra, 428). As Dr Basham has shown, the Jain com-
mentators Silanka, Jnanavimala and Abhayadeva quote a verse ascribed
to the niyativadins, which has the significant statement ‘na “bhavyam
bhavati na bhavino” sti nasah’ (op. cit ., p. 221, fn.i) which means
‘that which is not to be will not be, nor does that which is to be perish’.
This is very similar to the a priori premiss, sato natthi vinaso asato
1 Tasmad vai na’nyatparamasti tejah, 53.4.
2 Kalo ha bhutam bhavyan ca . . . ajanayat, 19.54.3.
3 Kala- is said to be ‘eternal’ (ajarah, 19. 53.1) and ‘its axle is immortality’
(amjrtam nyaksah, 19.53.2).
The Historical Background 143
natthi sambhavo (v. supra 9 85), the only difference being that what is
conceived to be Being is not ‘what is’ but ‘what will be’ and what is
Non-heing is not ‘what is not’ but ‘that which is not to be’. Now Being
cannot be destroyed nor Non-Being come to be. Therefore ‘what will
be’ cannot be destroyed, i.e. cannot be otherwise and ‘that which is not
to be’ cannot come into being, i.e. will not be. The apparent self-evi-
dence of the proposition that ‘what will be cannot be otherwise’ is
based on the misconception that the future event which actually comes
to pass in the course of time, ‘exists’ or has Being. Once this is accepted,
the determinist conclusion follows; the same argument can be used to
show that the past could not have been different from what in fact it
was. So everything, past, present and future is unalterable and fixed.
It is probably this very argument of the niyativadin, which provoked
the Buddha to draw attention to the past, present and future usages of
the verb ‘to be’ (ahosi, atthi, bhavissati) and enjoin the desirability of
keeping ‘these three linguistic conventions’ (tayo’me niruttipatha)
apart without confusing them, so that one may see that one cannot
argue that ‘what will be’ (bhavissati) has existence (atthi) for the future
that has not come into being and manifested itself has to be reckoned
as ‘what will be’ (bhavissati) and cannot be reckoned as ‘what is’
(atthi) h It is significant that it is said at the end of this section that even
‘the ahetuvadins, the akiriyavadins and the natthikavadins cannot
afford to condemn or reject these three linguistic conventions for
otherwise they would be liable to censure’. 2 Here the akiriyavadin is a
reference to Purana Kassapa, whose doctrine is called akiriyam at
D. I.53, while a doctrine stated in identical language is called akiriya-
vada- at M. I.404-5 (y. supra , 121). He seems to have been an outright
niyativadin as his later reputation 3 shows. Ahetuvadin on the other
hand is probably a reference to Makkhali Gosala, but his doctrine is
called samsara-suddhi (salvation by transmigration) at D. 1 . 5 3 though
at M. II.408, the same stated in identical language minus the cosmology
is called ahetuvada-. His ahetuvada- is evident from his language in
1 S. III.71, Ta yo ‘me niruttipatha adhivacanapatha pannattipatha asamkinna
asamkinnapubba na samkiyanti na samkiyissanti appatikuttha samanehi brah-
manehi vinnuhi. Katame tayo. Yam rupam . . . vedana . . . sahha . . . sankhara . . .
vinnanam atltam ahoslti tassa sankha, na tassa sankha atthlti na tassa sankha
bhavissatlti. Yam rupam . . . pe . . . vinnanam, ajatam apatubhutam bhavissatlti
tassa sankha na tassa sankha atthi ti na tassa sankha ahosi ti . . . (v. infra, 527).
2 v. infra , 527.
3 According to Gunaratna, Purana holds the view that the world is a product
of niyati; Purano mj/ati-janitam, op. cit. y p. 20.
144 Early Buddhist Theory of Knowledge
that he denied a cause (hetu) for moral degeneration or salvation 1 but
at the same time it may be noted that it was Purana who was called a
ahetu-vadin in the Mahabodhi Jataka (J. V., pp. 237, 246; cp. Basham,
op. cit ., p. 18). Makkhali seems to have been a syncretist thinker whose
doctrine was highly eclectic in character. He appears to have believed
in niyati - as well as svabhava (= bhava :) and yadrccha (—sangati) and
possibly even in parinama. This is probably the reason why these
central concepts of different schools are welded together in his doc-
trines. According to him all beings (sabbe satta . . . bhuta . . . jiva)
undergo development {parinama). This culminates in the course of
time (samsarasuddhi) in final salvation to which all beings are destined
under the impact of the factors of niyati , bhava and sangati (niyati-
sangati-bhava-parinata). It is probably this eclecticism which helped
him to bring together the scattered forces of the Ajivikas differing
among themselves and earn their leadership. 2 It was probably this same
eclecticism which made it difficult for others to specify exactly what
his doctrine was. Hence he has been called (in addition to ahetuvadin)
a Vainayikavadin (y. Basham, op. cit., p. 176) and an Ajnanavadin
(v. supra , 147) by Silahka, while the Mahabodhi Jataka calls him a
Theist (Issarakaranavadin, v. Basham, op. cit., p. 18). His belief in
‘sixty two ways of life 53 to be lived out in samsara shows that he
believed that all doctrines had their part to play in man’s development,
though man himself had no contribution to make.
(200) This differs from Dr Basham’s assessment of Makkhali’s doc-
trine and depends on what interpretation is given to the phrase niyati-
sangati-bhava-parinata-. We cannot discuss this problem in detail
since it does not directly concern us. Professor Basham himself trans-
lates the above phrase following Buddhaghosa as ‘developed by
Destiny (niyati), change (sangati) and nature (bhava)’ (op. cit., p. 225)
and affirms that he prefers ‘to follow Buddhaghosa and to take the
three first elements of the compound as in dvandva relationship,
translating the phrase as above’ (loc. cit.). Two pages later, however, he
says ‘ sangati and bhava , the manifestations of niyati in individuals,
were only apparent and illusory modifications of the one principle,
and did not in fact introduce new causal factors into the universal
process’ (op. cit., p. 227). We differ from Dr Basham in following
Buddhaghosa’s interpretation consistently (since it is supported by
1 Natthi hetu natthi paccayo sattanam samkilesaya . . . visuddhiya, D. I.53.
2 v. Basham, op. cit., p. 34. 3 Dvatthi patipada, D. I.54.
145
The Historical Background
other evidence) without giving an exaggerated importance to the
concept of niyati-. According to this interpretation Makkhali does not
become a Strict Determinist since the opposite category of ‘chance’ or
Indeterminism plays a significant part in his system. He therefore
subscribed to niyativada- only in the sense that he thought that some
future events like salvation for all ( v . samsarasuddhi sabbesam, J. VI,
p. 229) were strictly determined. In holding thus that some events of
the future had Being he would also have shared in the above a priori
argument. But this does not mean that he thought that human effort
had anything to do with shaping the future since he denies this
altogether. 1
(201) We cannot also entirely agree with Professor Basham’s theory
that ‘for the niyativadin causation was illusory’ (op. cit ., p. 227). Since
the causal conceptions of the niyativadin may be, in our opinion,
important for understanding the Buddhist concept of causation we
may pursue this problem here. Basham is led to this opinion on the
basis that if Time was illusory, then motion and change are illusory
and causation which is intimately bound up with these concepts must
be illusory too (v. op. cit., p. 23 6). He thinks, however, that this was a
later development 2 influenced by ‘the new doctrine of avicalita-nity-
atvam or a completely static universe’ (op. cit., p. 236). This assumption
appears to be incorrect for, on the contrary, there is evidence to
show that this doctrine of avicalita-nityatvam or the concept of a
universe, motionless and permanent, was known in the time of the
Pali Nikayas (v. infra, 402-8). As Dr Basham himself has suggested
this doctrine was probably the result of the same kind of a priori
reasoning as found in Parmenides (loc. cit.). Whether a niyativadin like
Purana would also have shared this concept, it is difficult to say. It is
certainly one of the logical implications of the doctrine of the unreality
of Time, which seems to have been either a corollary of or the basis of
the determinist thesis. If it was held by the niyativadin, then the doc-
trine of a double standard of truth (v. Basham, op. cit., p. 230) could
have been utilized to resolve the contradiction, which is probably what
Parmenides himself does in speaking of a ‘Way of Truth’ and ‘Way of
Opinion’. This would mean that deterministic causation had a relative
reality and not that it was entirely illusory.
1 v. natthi attakare . . . purisakare, D. I.53.
2 v. ‘The universe seems to have been thought of as a continuous process,
which was recognized by some later Ajlvikas to be on the ultimate analysis
illusory*, op. cit., p. 227.
146 Early Buddhist Theory of Knowledge
(202) Whether for the niyativadin, causation in the ultimate analysis
was unreal or not, the fact of causation seems to have been accepted
by him at least at the level of conventional truth. That even the later
Ajivika rigidly believed in causation in a strictly determinist sense is
evident from the argument for niyati in Gunaratna’s (14th C.) work,
according to which it is necessary to posit the existence of the force of
niyati in order to account for causes and effects (karyakarana). Pro-
fessor Basham has given the gist of this argument but has failed to
quote the sentence, which seems to imply a belief in causation on the
part of the niyati-vadin. We may translate the argument as stated by
Gunaratna as follows: ‘Whatever happens at any time, anywhere, is to
be conceived as happening in the form of niyati only. Otherwise there
would be no definite sequence of causes and effects ( karya-karana -
vyavastha) or a fixed pattern of anything, owing to the absence of a
controlling agent’. 1
(203) When we consider the arguments of the niyativadin as stated by
Sllanka (9th C.) we find that arguments based on the two principles of
causal determination play a fundamental role. As Mill stated in his ‘A
System of Logic’, the methods of discovering a causal connection are
‘two in number’ {op. cit., 253) of which ‘one is, by comparing together
different instances in which the phenomenon occurs. The other is, by
comparing instances in which the phenomenon does occur, with
instances in other respects similar in which it does not. These two
methods may be respectively denominated the Method of Agreement
and the Method of Difference’ ( loc . cit.). These two principles known
in Indian logic as the anvaya-vyatireka-rlti first appear in a concrete
form in the causal formula of the Pali Nikayas. But this notion is
constantly made use of by the niyativadin in arguing against his
opponent. One of the arguments of the believer in karma, is that there
is a causal connection between good karma and pleasant consequences
and evil karma and unpleasant consequences. The niyativadin shows
by applying the two principles of causal determination to what is
observable in this world, that there is no such causal connection. Good
is not always followed by happiness nor evil by grief, nor is the
absence of good followed by the absence of happiness and the absence
of evil by the absence of grief. The argument is actually stated as
1 Yadyada yato bhavati tattada tata eva niyaten’aiva rupena bhavad upala-
bhyate. Anyatha karyakaranavyavastha pratiniyatarupavyavastha ca na bhavet,
niyamakabhavat , op. cit., p. 12. Dr Basham in his quotation {op. cit., p. 235,
fn. 2) has omitted the phrase underlined.
147
The Historical Background
follows: ‘In this world, grief does not arise for a man even though he
delights in evil courses, while for another person who does good, it
does’. 1 It is thus concluded that there is no causal connection between
evil and grief or good and happiness.
(204) The argument against the causal connection said to hold
between human exertion and its fruits is again shown to be false by
appeal to observation based on the two principles. If there was a causal
connection then we should have the following sequences:
Presence of human exertion Presence of its fruits
Absence of ,, „ Absence of „ „
But what we observe is as follows:
Presence of human exertion Absence of fruits
(Purusa-kare sati) (phala’praptih)
Absence of human exertion Presence of fruits
(V yapara’bhave) (visistaphala’ vaptih)
Thus it is argued that ‘nothing is achieved by human effort’. 2 If we
examine this argument we find that it is based on the assumption that
‘equal effort (samane purusakare sati) must be followed by equal
results’, 3 which is based on the principle that variations in the cause are
correlated with variations in the effect. This is explicitly stated in the
argument that ‘Time is not a causal factor, for since Time is uniform,
its effects in the world could not be multiform. There are variations in
the effect only when there are variations in the cause (Karana-bhede hi
karya-bhedo bhavati, n’abhede)’. 4 This is similar to Mill’s ‘Fifth
Canon’, 5 an extension of the two original principles.
(205) These subtleties probably developed later, but there is some
reason to believe that the basic argument based on the belief in the
1 Atr’aikasy asadanusthanaratasy ’api na dukkham utpadyate, parasya tu
sadanusthayino tad bhavati, Sllanka, op. cit., Vol. II, fol. 26, on Su. 2.1. 12;
quoted by Basham, op. cit., p. 234, fn. 3.
2 Yadi purusakarakrtam sukhadyanubhuyeta tatah sevakavanikkarsakadmam
samane purusakare sati phalapraptivaisadrsyam phala’praptisca na bhavet.
Kasya cittu sevadivyapara’bhave ’pi visistaphala’vaptir drsyata iti. Ato na puru-
sakarat kincid asadyate. Sllanka, op. cit., Vol. I, fol. 30 on Su. 1. 1.2.2.
3 The phrase, samane purusakare sati, implies samana-phalapraptih as the
expected consequence.
4 Na’pi kalah karta, tasy’aikarupatvaj jagati phalavaicitrya ’nupapatteh.
Karanabhede hi karyabhedo bhavati, na’bhede, Sllanka, loc. cit.
5 ‘Whatever phenomenon varies in any manner whenever another varies in
some particular manner, is either a cause or an effect of that phenomenon, or is
connected with it through some fact of causation’, op. cit., p. 263.
148 Early Buddhist Theory of Knowledge
two principles of causal determination went back to the earliest times,
since we seem to find a reference to it in the Sutrakrtanga itself.
(206) In introducing the ‘fourth type of person’ (catitthe purisajae),
who is a niyativadin (niyativaietti), the Sutrakrtanga (2.x. 12) places
the causal argument in the mouth of the Determinist as follows: iha
khalu duve purisa bhavanti — ege purise kiriyamaikkhai*, ege purise
nokiriyamaikkhai, je ya purise kiriyamaikkhai je ya purise nokiri-
yamaikkhai, dovi te purisa tulla egattha, karanam avanna, i.e. here are
two persons, one person maintains (the efficacy of) action while the
other person denies (the efficacy of) action, but both of them are
(ultimately) equal and alike on account of the cause (being niyati)’.
Here the crucial phrase karanamavanna seems to have presented a
difficulty to the translator. Jacobi renders karanamavanna as ‘they are
actuated by the same force’ (op. cit ., p. 317) but in a footnote (fn. 2) he
says that ‘this is the interpretation of the commentators. But to the
phrase karanamapanna they give here a meaning different from that
in the following paragraphs. I therefore propose the following trans-
lation of the end of the paragraph: “are equally (wrong), (err) alike as
regards the cause (of actions)” ’. Dr Basham follows Jacobi’s first
translation, viz. ‘Both equally and alike are affected by (a single) cause’
(op. cit., p. 233). We have closely followed Silanka who says, ata
ekarthavekakaranapannatvaditi niyati-vasen’aiva tau niyativadam
aniyativadam casritaviti bhavah, 1 i.e. thus (both are) alike since they
are affected by the same cause, the sense being that by the force of
niyati alone they have followed the niyativada and the aniyativada.
Jacobi’s second translation cannot be accepted since it is too much
of a periphrasis which introduces concepts like ‘wrong’ and ‘err’ which
are not found in or suggested by the context. But whether we translate
karanamavanna as ‘on account of the (same) cause’ or ‘as regards the.
cause’ the significance that karana- had for the person using this term
is clear from the example cited. He takes the case of two persons alike
in other respects except for the fact that one is a kiriyavadin and the
other an akiriyavadin and finds that latterly they are both still alike.
This is the application of Mill’s Method of Difference with negative
results and may be represented symbolically 2 thus:
ABC— b c
BC— be
where A= belief in niyativada. From observing the two sequences one
1 Op. cit., Vol. II, fol. 25 on Su. 2.2.12. 2 Stebbing, op. cit., p. 334.
The Historical Background 149
may draw the conclusion that A is not a causal factor; so is not-A not
a causal factor for the absence of A has made no difference to the result.
So both those who claim that kiriya (the act) is a cause or that akiriya
(the non-act) is a cause are equally wrong as regards the cause — this is
in fact supported by Jacobi’s second translation, though it is not
supported by the actual wording of the argument. Although this
negative conclusion — that belief in kiriya or akiriya cannot be the
cause — seems to be implied, the actual conclusion that is drawn is a
positive one, namely that niyati must be the causal factor. But this is
an assumption, since niyati is a metaphysical factor, which is unobserv-
able and cannot be discovered experimentally. Yet what is most
significant is that the niyativadin seems to have been convinced in some
sense of the fact of causation and made use of the causal argument,
based on a belief in the principles of causal determination to show that
his opponent was wrong. The problem is whether these concepts were
borrowed from another school or were intrinsic to his own system.
(207) According to the argument of the niyativadin as stated by
Gunaratna (v. supra, 202), the niyativadin believed in a ‘fixed pattern
of causes and effects’ (karya-karanavyavastha) but we saw at the same
time that Purana, the niyativadin, was called an upholder of the
‘doctrine of causelessness’ 1 (ahetuvada-). The reason for these con-
tradictory evaluations would be clear if we can comprehend the niya-
tivadin’s concept of causation. He denied whatever was held as the
causes of events natural or metaphysical, by his opponents (e.g.
purusakara-, karma-, kala-, Tsvara-). This would have made him appear
in the eyes of his opponents as one who denied all causes, internal or
external, of events. But all his criticisms imply a belief in causation,
which in the ultimate analysis turns out to be a belief in niyati conceived
as the first and the efficient cause of all phenomena. Nature to him was
a single rigidly deterministic system, in which no individual or separate
causal lines 2 or processes were discoverable or distinguishable. All
events and processes were caused but caused by the all-embracing
metaphysical principle of niyati. For such a rigid determinist individual
causal processes could not be conceived in isolation from the entire
system. This would have appeared to be the very denial of causation
as understood by some of their opponents and it is difficult to believe
' v. supra , 199. The term may also have been employed in the Nikaya period
to denote yadrccha-vada or ‘Indeterminism’ (— sangati) and Makkhali was
probably an ahetukavadin in both these senses.
2 On the use of the term ‘causal lines’, v. Russell, Human Knowledge, P* 333 ^
150 Early Buddhist Theory of Knowledge
that the conception of the principles of causal determination could have
arisen in such a system. On the other hand if niyati was only a meta-
physical controlling force which guaranteed the operation of the laws
of cause and effect, as Gunaratna’s argument of the niyativadin would
make us understand, we would have to say that he believed in causation
with the difference that for him the supreme Cause of causes and causal
processes was the factor of niyati. In such a case he would be an
ahetukavadin only in the sense of denying the metaphysical validity
of what his opponents conceived to be the hetu-s of things.
(208) Whatever the answer to this question, it seems probable that it
was people who made use of the causal arguments of the kind em-
ployed by the niyativadins, who were called the ‘causal argumenta-
tion^’ (karana-vada) 1 in the Mahaniddesa (v. infra, This work
comments on and defines the different types of ‘expert debaters’
(kusala) referred to in the Suttanipata, among whom the niyativadins
have undoubtedly to be reckoned.
(209) The other arguments of the niyativadin as given by Silanka
have been translated by Basham (op. cit ., pp. 231-4) and if we study
them it would be seen that almost all of them are dialectical arguments
having the following form: If p is true, then either q or not-q is true,
but q implies r and not-q implies s, each of which (i.e. s, r) contradicts
one of the assumptions or propositions posited by his opponent.
(210) We may illustrate this with an example. 2 Thus the theory that
happiness and grief come about through the agency of God (Isvara-)
is criticized by showing that the concept of God’s existence leads to
contradictions, as follows: ‘ “If God exists” (p), then either “God has
form” (q, Isvaro murtah) or “God has no form” (not-q, Isvaro’
murtah). If q, then “he, like an ordinary person, is not omnipotent”
(r, prakrtapurusasy’eva sarvakartitvabhavah); and if not-q, then “his
inactivity is greater than space” (s, akasasy’eva sutaram niskriyatvam).
Now both r and s contradict the definition of God, that he is omnipo-
tent and all-active.
(21 1) This is one of the standard dialectical metaphysical arguments
employed in later times and one may well doubt whether they were
those of the early niyativadin, but we need not doubt that the early
1 Karana-vada can also mean ‘those who debated about the first cause"; cp.
Kim karanam? Brahma?, Svet. 1.1.
2 v. Basham, op. cit., p. 231, where the full text and the translation is given.
The Historical Background 15 1
niyativadin did argue against his opponents and that his opponents
were these same protagonists as mentioned by Silanka. Silanka states
the fact that the niyativadin argued against those who claimed that the
effort of the person (purusakara-), Time (kala-), God (isvara-),
Intrinsic Nature (svabhava-) and Karma were respectively the causes
of ‘pleasurable and painful experience 5 (. sukhadukkha -). The Svetas-
varara Upanisad mentions that one of the questions debated by the
brahmavadins is as follows: adhisthitah kena sukhetaresu vartamahe,
where Sankara explains sukhetaresu as sukhadukkhesu 1 and the sentence
may be translated as ‘governed by whom (or what) do we live in
pleasure and pain 5 . Now, it is recorded in the Pali Nikayas not only
that this was one of the topics that was hotly debated at the time
(v. infra , 395) but that the parties to these debates were these same
theorists. The opponents of the niyativadin according to Silanka are
the following:
1. Yadi purusakarakrtam sukhadukkhadyanubhuyeta . . . v. Basham,
op. cit ., p. 230, fn. 1.
2. N 5 api kalah karta . . . ibid., p. 231, fn. 1.
3. IsvarakartxkPpi sukhadukkhe na bhavatah . . . ibid., fn. 2.
4. Svabkdvasya.pi sukhadukkhadikartrtva . . . ibid., p. 232, fn. 1.
5. Karmanzh, sukhadukkham prati kartrtvam . . ., ibid., fn. 2.
Now it would be noted that four of these theories are specifically
mentioned in connection with this very problem in the Devadaha
Sutta (M. II.222). We may state them under the numbers correspond-
ing to Silanka 5 s list:
1. Ditthadhamma-r//»a>t^ma 2 -hetu sukhadukkham patisamvedenti.
3. Issaranimmanahetu sukhadukkham patisamvedenti.
4. Sangati-M<mz 3 -hetu sukhadukkham patisamvedenti.
5. Pubbekata- A hetu sukhadukkham patisamvedenti.
(212) We may conclude from what we have said above that some of
the Ajivikas were rationalists who not only constructed their theories
by reasoning but also defended them against their opponents by
1 Adhisthita niyamitah kena sukhetaresu sukhadukkhesu vartamahe, Ananda
Asrama Series, No. 17, p. 18 on Svet. 1.1.
2 This is probably the same as purisa -parakkama-, A. IV. 190, which is a
synonym of purisa-thama- and purisa-viriya-.
3 bhava-=sabhava-, v. Basham, op. cit., p. 226.
4 Cp. the theory of puratana-karma-krtam, mentioned by Gunaratna, op. cit.,
p. 20.
152 Early Buddhist Theory of Knowledge
employing reason to demolish their theories. But there is also some
reason to think that these theories were not merely constructed a
priori but had some basis in claims to paranormal perception or super-
human insight as well.
(213) Basham says that both ‘Nigantha Nataputta and Makkhali
Gosala certainly seemed to have laid claim to full enlightenment’ ( op ,
cit, , p. 92). In the case of Nigantha Nataputta we have sufficient
evidence to believe that he claimed omniscience. But in the case of the
latter, although omniscience is claimed for him in later times {v, supra ,
1 51), there is no evidence that he himself claimed omniscience. Yet as
Basham has pointed out, he is said to have practised penance in order
‘to acquire magic power and superhuman insight’ {op, cit,, p. 50). Dr
Basham goes to the extent of saying that on the evidence of the Budd-
hist and Jain texts ‘it appears that he was capable, either honestly or by
fraud, of producing psychic phenomena’ {op, cit,, p. 51). The belief in
prophecy, it would appear, should be the natural outcome of their
determinist theory; if the future was wholly or partly determined, it
should be possible to know this in some way or another, because the
future exists in the same sense in which the present exists {v. supra, 199),
which was the niyativadin’s assumption. But it is also possible that
for the niyativadin part of the reasons for believing in his determinist
thesis were actuated by his belief in prophecy as the story of ‘Gosala
and the Sesamum Plant’ {op, cit,, pp. 47-9) seems to suggest. Basham
does not exclude ‘the possibility that the story has some basis in fact’
(p. 49). One of the central features of the story was that it was possible
to have precognitive experiences about at least some events in the
future by means of one’s intuitive knowledge. In fact it is said that it
was for the purpose of acquiring this kind of intuitive knowledge that
Gosala practised meditation and penance {op, cit,, p. 50). On the evi-
dence alleged, the possibility cannot be altogether excluded that these
ascetics may have had or seem to have believed that they had a few
precognitive experiences of the future, which either led them to or
reinforced their determinist thesis. Nothing is knowable unless it is a
fact; if the future is knowable it is a fact and this is not possible unless
the future exists in some sense in or like the present — which is the
determinist thesis.
(214) It is probable that some of the Ajivika beliefs about the size and
colour of the soul {v, supra, 132) are an externalization of experiences
had in trance-states. It is, however, curious that Buddhaghosa says
i53
The Historical Background
quite the opposite. In explaining the epistemic origin of the belief^in
the soul having size and form after death, he says that these are 5 the
result of meditative experiences. 1 As an alternative, he suggests that it
is the result ‘ purely of reasoning as in the case of the Ajivikas and
others’ (ajlvikadayo viya takkamatteri ev a va, DA. I.119). The
historical truth behind this assertion is probably that Buddhaghosa was
greatly impressed by the rational tradition of the Ajivikas. While he
was also aware that claims with regard to the size and colour of the
soul were made on the basis of trance experiences, he did not identify
these with the Ajivikas because he thought that they were mere
dialecticians (takkl-), which they probably were at the time at which
he wrote.
(215) While there is no evidence to show that Makkhali Gosala
claimed omniscience, there is good evidence that Purana Kassapa, the
pure Determinist, did so. Two brahmins meet the Buddha and tell him
about Purana’s claim to omniscience and what he claims to know in
the following words: ‘Purana Kassapa claims to be omniscient and all-
seeing and to be possessed of an infinite knowledge and insight such
that whenever he walks or stands, sleeps or keeps awake, his knowledge
and insight is constantly present continuously at all times. This is what
he says, “I abide knowing and seeing a finite world with my infinite
knowledge” ’. 2 As a determinist he probably claimed to know fully not
only the past and the present but the future as well. The theory that
‘the world was finite’ (antava loko) was one of those, which was
debated at this time (v. infra, 382, 383). It is likely that arguments were
evolved to ‘prove’ the validity of theories believed in on the basis of
mystic experiences.
(216) While the doctrines of the Ajivikas appear to have been held
mainly on the basis of reasoning and perhaps of personal claims to
supernormal insights as well, we cannot entirely discount the belief in
tradition on the part of even the early Ajivikas. According to a state-
ment of Sllanka, quoted by Basham (op. cit,, p. 175, fn. 3) they seem to
1 DA. 1 . 1 19. Rupi atta’ti adisu kasinarupam atta ti tattha pavattasannan c’assa
sanna ti gahetva . . . (They hold that) the soul has form (after death), etc., think-
ing that the soul has the form (colour) of the meditational device and taking its
after-image as their own consciousness.
2 A. IV.428. Purano . . . Kassapo sabbannu sabbadassavi aparisesananadas-
sanam patijanati carato ca me titthato ca suttassa ca jagarassa ca satatam samitam
nanadassanam paccupatthitan ti. So evam aha ’aham anantena nanena antavantam
lokam janam passam viharami ’ti.
154 Early Buddhist Theory of Knowledge
have had their own traditional scriptures: ‘the Trairasikas, who follow
the doctrine of Gosala, and who have twenty-one sutras, arranged
according to the order of the Trairasika siitras in the Piirvas\ Now, the
Suttanipata makes a reference to ‘the Vedas of the Samanas as well as
those of the brahmins 5 . 1 Since the Samanas were classified in the
Suttanipata as the Ajivikas and the Niganthas ( v . infra , 375), it is likely
that these collections of scriptures of the Ajivikas are among the
‘Vedas of the Samanas 5 referred to. It shows that at least some of the
Ajivikas had a sacred scripture as early as the period of the Pali
Nikayas, a fact which is confirmed by the quotations or adaptations
from them, found in the early Buddhist and Jain texts. 2
(217) The reference to the ‘followers of Gosala 5 (Gosalamata’nusa-
rino), elsewhere called the Ajivikas, as the Trairasikas in the above
statement of Silanka, is significant and points to the contribution made
by them to epistemology and logic. While the Sceptics and the
Buddhists evolved or adopted a four-fold logic, the Ajivikas who were
the followers of Makkhali Gosala appear to have classified propositions
into three mutually exclusive categories and had a theory of three-fold
standpoints (naya-). Haribhadra in his Vrtti on the Nandi-Sutra
identifies the Trairasikas with the Ajivikas: Trairasikasc Ajivika
ev’ocyante . 3 As Basham has shown 4 5 6 Abhayadeva states in the com-
mentary to the Samavayanga-Sutra that ‘these Ajivikas were called
Trairasikas 5 . 5 On the basis of Silanka’s statements Hoernle 6 had
identified these Trairasikas with the schismatic Jain sect of Trairasikas,
which came into existence over five centuries after Mahavira, but we
agree with Dr Basham’s contention 7 that the two have to be dis-
tinguished.
(218) Dr Basham says that ‘the distinctive characteristic of the
Ajivika system of epistemology . . . was the division of propositions
into three categories 5 (op. cit.> p. 274); but in fact there is another
distinctive feature, namely the adoption of three standpoints (nayas)
instead of the seven (or the two?) of Jainism. The passage quoted by
1 Vedani viceyya kevalani, Samananam yani p’atthi Brahmananam, Sn. 529.
2 Basham, op. cit., p. 214 ff., 216 ff.
3 Nandi Sutra, by Devavacaka (Devarddhi Gani). With the Curni (gloss) of
Jinadasa Gani and the Vrtti (commentary) of Haribhadra. Ed. Vijayadana Suri,
1931, Fol. 1 14. 4 Op. cit., p. 179, fn. 3.
5 Ta eva c’ Ajivikas Trairasika bhanitah, fol. 120 on Sam. 147.
6 ERE., Vol. I, p. 262. 7 Op. cit., p. 178.
The Historical Background 155
Basham from the Nandi commentary, fol. 1 13 (v. op. cit p. 274, fn. 5)
does not in fact appear in the edition of the commentary, that we have
used (i.e. Nandi Sutra, with the comms. of Jinadasa and Haribhadra,
ed. Vijayadana Suri, Indore, 1931). In this edition, there are two
passages more or less identical, which explain these aspects of the
epistemology and logic of the Ajlvikas: one is by Jinadasa Gani in his
Curni appearing on fob 1 10 and the other is in the Vrtti by Haribhadra
in another form of Prakrit appearing on fols. 113, 114. The difference
is merely dialectical, e.g. the Curni has ‘te c’eva Ajlvika terasiya
bhanita’, etc., while the Vrtti reads, ‘te c’eva Ajiviya terasiya bhaniya’.
These passages differ in a significant respect (v. infra) from the passage
cited by Basham.
(219) We may take the Sanskrit version in Abhayadeva’ s commentary
on the Samavayanga-Sutra, which reads as follows: ‘These Ajlvikas
are called Trairasikas. Why? The reason is that they entertain (icchanti)
everything to be of a triple nature, viz. soul, non-soul, soul and non-
soul; world, non- world, world and non-world; being, non-being,
being and non-being, etc. Even in (api) considering standpoints they
entertain a three-fold standpoint such as the substantial, the modal and
the dual’. 1 Thus according to Abhayadeva, they are called Trairasikas
for two reasons, in having a three-fold mode of predication and a
three-fold set of standpoints.
(220) The Prakrit versions are almost identically similar to this, the
only difference being that instead of sarvam tryatmakam icchanti they
have, savvam jagam tryatmakam icchanti, which makes no material
difference. But the passage quoted by Basham has significant variations.
It defines Trairasikah in an additional sentence as follows: Tatas tribhl
rasibhiscarantlti Trairasikah, i.e. thus, since they work with three
heaps they are (called) Trairasikas. Since this definition occurs
immediately after mentioning their three-fold standpoints (naya), the
‘heaps’ (rasi) seem to refer to the different types of predication as well
as the standpoints. The notable difference in this passage is that it speaks
of the three nayas as ‘dravy’astikam paryayastikam ubhayastikan ca’,
whereas Abhayadeva has (v. fn. supra) ‘dravyarthikah paryayarthikah
1 Ta eva c’ajlvikas Trairasika bhanitah. Kasmad? — ucyate, yasmatte sarvam
tryatmakam icchanti yatha jlvo’ajivo jlvajlvah, loko’loko lokalokah, sad asat
sadasat ityevam adi, nayacintayam api te trividham nayam icchanti tadyatha
dravyarthikah paryayarthikah ubhayarthikah, Samavayangasutram, with
Abhayadeva’s commentary, Ed. Naginadasa Nemachanda, 1938, fol. 120 on
Sam. 147.
156 Early Buddhist Theory of Knowledge
ubhayarthikah’ corresponding to the Prakrit ‘davvatthiko pajja-
vatthiko ubhayatthiko ya’ (Nandi Sutra) (fol. no). Are we to regard
the latter as preserving a more faithful tradition, since the two funda-
mental nayas of the Jains are called the dravyarthikarrdyz- and
parydyarthikanayz - 1 and the Prakrit versions of the Nandi commen-
taries agree with the Sanskrit version of the Samavayanga com-
mentary?
(221) According to Dr Basham, ‘the Ajivikas . . . seem to have
accepted the basic principle of Jaina epistemology, without going to
the over-refined extreme of sapta-bhangi , as in the orthodox Jaina
syadvada and Nayavada (op. cit ., p. 275). This implies that the
Ajivikas were aware of the seven-fold formulae of the Jains and sim-
plified them. But judged by the fact that the three-fold schema of
predication is simpler than the four-fold schema of the Sceptics and
Buddhists and the corresponding seven-fold schema of the Jains, it
would appear to be earlier than both the Buddhist and the Jain
schemas, with which the Ajivikas could not have been acquainted
when they evolved theirs.
(222) In fact, it can be shown that in the earliest Buddhist and Jain
texts the very doctrine of the Trairasikas, which seems to have
necessitated the three-fold schema, is mentioned, thus making it highly
probable that it was at least earlier than the Jain schema . For while the
earliest stratum of the Pali Nikayas knows of the four-fold schema,
one of the earliest books of the Jain Canon, the Sutrakrtanga, which
makes an independent reference to this Trairasika doctrine, does not
mention the seven-fold schema, although it is aware of the basic
principle of syadvada (y. infra, 233).
(223) The Brahmajala Sutta mentions a class of religious teachers,
who were semi-eternalists (samana-brahmana . . . ekacca-sassatika
ekacca-asassatika, D. 1 . 19), who hold that the world and the soul were
partly eternal and partly not (ekaccam sassatam ekaccam asassatam
attanan ca lokan ca pannapenti, loc. cit.). It is probably this same theory
that is elsewhere referred to as the view that holds that the soul and
the world are both eternal and not eternal (sassato ca asassato ca atta
ca loko ca, M. II.233, Ud. 69). Four varieties of these semi-eternalists
are mentioned in the Sutta, of which the second believes in the existence
of an ethereal group of Khidda-padosika gods (santi . . . Khidda-
padosika nama deva, loc. cit.). Now it is said that those who over-
1 v. Guerinot, La Religion Djaina, pp. 130-1.
The Historical Background 1 57
indulge in sporting in this Heaven 1 lose their memory, fall from this
state and are reborn on earth (tamha kaya cavitva itthattham agacchati,
loc. cit .). Such a person leaves the household life (anagariyam pabba-
jati), practises meditation and attains a jhanic state (atappam anvaya . . .
ceto-samadhim phusati, loc . cit.), whereby he sees this past life of his
and realizes that in that world there are beings who do not over-
indulge, and who are eternal (ye . . . na ativelam hassa-khidda-rati-
dhamma-samapanna viharanti . . . te . . . na cavanti, nicca dhuva
sassata aviparinamadhamma sassati-samam tath’eva thassanti, loc. cit.),
while the others are liable to fall. The account given of this school may
perhaps have undergone some distortion, but we can gather from what
is stated that according to this school, there are three types of beings:
1 . the eternal beings (sassata) who live for ever in that state (sassati-
samam tath’eva thassanti, loc. cit.).
2. the temporal beings who live in this world.
3. the partly eternal and partly temporal beings (ekaccam sassatam
ekaccam asassatam) who fall from the eternal state and perhaps
go back again after a life of restraint and meditation.
(224) We can see here more than the rudiments of the doctrine of
mandala-moksa or cyclic salvation (v. Basham, op. cit., 257-61). The
eternal beings would correspond to the cempotakars of the Civariana-
cittiyar, while those who fall would be the mantalars ( v . op. cit.,
p. 260), the main difference being that a different reason is given here
as to why the mantalars return from that state. Now, Dr Basham
says that this doctrine of cyclic salvation ‘appears to have emerged
some time after the death of Gosala’ (y. op. cit., p. 259). He does not
explain why it was necessary to await the death of Gosala for the doc-
trine to emerge, but he has seen that it is mentioned as early as the
Sutrakrtariga. However, it is not correct to say that ‘it is first men-
tioned in the Sutrakrtariga’ (loc. cit.). We may quote the original
version in the Sutrakrtariga since the identity in language with the Pali
version is significant:
Suddhe apavae aya iham egesim ahiyam
Puno kidda-padosenam so tattha avarajjhai
Iha samvude muni jae paccha hoi apavae
Viyadambu jaha bhujjo nirayam sarayam taha.
Su. 1. 1. 3. 1 1-2.
1 We have used capital H for ‘Heaven’ here to denote that in the opinion of
this school it was an eternal state.
1 58 Early Buddhist Theory of Knowledge
We may translate this as follows: ‘It is said by some that the soul is
pure and sinless, but again it sins (avarajjhai) in that state owing to
kidda-padosa - (pleasure and hatred? corruption through pleasure?);
born here, he later (paccha) becomes sinless as a restrained ascetic. As
pure water free from pollution becomes again polluted (so does he
again become sinful).’ This stanza appears in fact to summarize what
was stated in the Pali version. The Pali version explicitly stated
that some beings were eternal although all the beings were
called Khidda-padosika deva and Khidda- padosa - (Ard. Mag.
kidda-padosa -) was the cause of the fall according to both the
Buddhist and the Jain accounts.
(225) Now, as Dr Basham has pointed out, Sllanka identifies this
doctrine of the Sutrakrtanga with that of the Trairasika followers of
Gosala {loc. cit .). But interpreting this verse Sllanka gives a different
explanation of kidda-padosa- from that suggested in the Pali texts,
which tries to make out that the cause is excessive debauchery (ativelam
hassa-khidda-rati-dhamma-samapanna, loc . cit.). Sllanka gives a more
sublime reason for their fall, which was probably the reason that the
Trairasikas themselves would have given, namely that the eternal soul
has feelings about the true religion and ‘is elated when his religion is
revered (on earth) and other religions are looked down upon and is
angry when his own religion is despised’ 1 explaining kidda- and padosa-
as this joy and anger respectively. It is possible that when the Buddha
warned his disciples not to be elated when people praise his religion
and not to be angry when people condemn it (Brahmajala Sutta,
D. I.3, v. infra , 739) he was influenced by what he believed to be the
plight of the Ajivika as a result of his elation and anger about his own
religion. At the same time the Buddhist criticism that falling a prey to
temptation was the cause of their fall may have had some basis in the
Ajivika beliefs themselves, since one of the greatest fears of the
Ajivika ascetic was that he may succumb to the caresses of ‘the gods
Punnabhadda and Manibhadda’ who tempt him on the verge of death
(v. Basham, op. cit., p. 257 ff.).
(226) Basham does not mention this Pali parallel to the verse in the
Sutrakrtanga, which was the reason why he thought that this doctrine,
1 Op. cit., Vol. I, fol. 45 on Su. 1. 1. 3. 1 1.2. Svasasanapujam upalabhya’nyasas-
ana-parabhavam c’opalabhya . . . pramodah sanjayate, svasasananyakkaradars-
anac ca dvesah.
The Historical Background 159
was not so important for the early Ajlvika. 1 But unless it was one of
the cardinal doctrines both the Jains as well as the Buddhists would
not have stated it in summarizing their views.
(227) This Trairasika doctrine which found it necessary to posit three
kinds of souls and perhaps three kinds of worlds corresponding to
them as well as three kinds of being, appears therefore to be quite early.
The evidence points to its having its origin in a sect of Ajlvikas (in the
loose sense) independently of Gosala, though it may have accepted
Gosala’s leadership or merged with the followers of Gosala later on.
It is to this doctrine that Basham traces the necessity for the Trairasika
to posit a third possibility: ‘The Ajlvika postulate of a third possibility,
neither being nor non-being, must have formed a convenient logical
basis for the unusual doctrine that some souls were compelled to return
even from nirvana . These would be classified in the third category,
sadasat — emancipated from samsara and yet not emancipated’ (op. cit .,
p. 275). We agree with this conclusion though not in the form in which
Dr Basham states it, since the third possibility is not ‘neither being
nor non-being’ but ‘both being and non-being’ (sadasat), which has
to be distinguished from the former since the distinction was drawn in
the time of the Pali Nikayas. The thesis of this school is, as we said,
stated in the Pali Nikayas as ‘ sassato ca asassato ca atta ca loko ca
which would probably have been equivalent to ‘ sanasanjlvasca lokasca
in the terminology of the Trairasika. This, it may be observed, is not a
logical proposition which is contradictory as would appear from its
form (since it seems to violate the Law of Excluded Middle) but an
empirical proposition which is contingent (v. infra , 579). Thus, for this
school the three logical alternatives would be: (1) p, (2) notp, (3)
p.notp and not the usual two (i.p, 2.not-p) according to the Aristo-
telian schema.
(228) We are on less certain ground with regard to its doctrine of
nayas in respect of its antiquity and significance. Dr Basham assumes
that it is a simpler version of the seven-fold nayas of Jainism (op. cit.,
p. 275). But there is another possibility.
One has to compare these three nayas considering the terminology
with the two fundamental nayas of Jainism:
1 v. his remark, ‘This doctrine is not elsewhere mentioned in the Pali or Jaina
Prakrit texts, and seems not to have loomed large in the minds of the earlier
Ajlvikas’ (op. cit., p. 259).
i6o
Early Buddhist Theory of Knowledge
Ajivika
1. Dravyarthika-
(Substantial)
2. Paryayarthika-
(Modal)
3. Ubhayarthika-
(Dual)
Jainism
Dravyarthika- (3) naigama-,
samgraha-,
vyavahara-.
Paryayarthika- (4) rjusutra-,
sabda-,
samabhirudha-,
evambhuta-.
It will be seen that the Ajivika is more complex if we consider the fact
that Jainism has nothing corresponding to the ubhayarthikanaya-, but
on the other hand the Ajivika has not subdivided (as far as our
knowledge goes) the first two nayas. The fact that the first two nayas
are held in common, points to a common origin, though later the
Jains made further elaborations of these while the Ajivikas added the
third.
(229) One suspects a close connection between the three forms of
predication and the three nayas. Are we to say that each of the forms
of predication was possible only from one of the nayas, viz.
(1) Sat — according to the dravyarthikanaya-
(2) Asat — according to the paryayarthikanaya-
(3) Sadasat — acc
Early Buddhist Theory Of Knowledge
Anónimo