rTME

NYAYA THEORY OF KNOWLEDGE

A CRITICAL STUDY OF SOME PROBLEMS OF
LOGIC AND METAPHYSICS

feY

S. C. CHATTERJEE, M.A.. Ph.D.

Premchand Roychand Scholar (Cal.), Lecturer in PhilosophYi
Calcutta University

PUBLISHED BY THE

UNIVERSITY OF CALCUTTA
1939
 PBINT^ ro InDiA

numD AMD PDBIiISHED BT BSnFBNDBAIiAL BAMEBIBI

AT IBB OALOmTA tmiTBBSm PBMB, BBNATB HODSB, OAIiCDTTA.

Reg. No 1117B—January, 1939—1!,

 In Memory
OF

MY MOTHER
AND

MY FATHER
 CONTENTS
Page
Preface xv

Chapter I

Introduction 1

BOOK I

THE METHOD OF VALID KNOWLEDGE

(PEAMiNA)

Chapter II

The Nature and Forms of Knowledge 9

1. Definition of Knowledge (buddhi) 9

2. Classification of Knowledge 21
3. Memory and Dream 23
4. Doubt (sam^aya) 30
5 Error (viparyyaya) 34
6. Theories of Illusion in Indian Philosophy 36
7. Hypothetical Argument (tarka) 47

Chapter III

Valid Knowledge and its Method (Prama and

Pramana) 53

1. Definition of prama or valid knowledge 53

2. Definition of pramana or the method of
knowledge 56
3. Nyaya criticism of the Bauddha views of

pramana 60
4. Nyaya criticism of the Mimariasa and
Samkhya views 66
 CONTENTS

Page
Chapter rv

The Factors of Valid Knowledge (Prami) 75

1. The subject, object and method of valid

knowledge 76
2. Distinction the method from the subject
of
and object of valid knowledge 78

Chapter V
The Test of Truth and Error 83

1. The problem^ and alternative solutions 83

2. The Nyaya theory of extrinsic validity and
invalidity 84
3. Objections to the theory answered by the
Nyaya 89
4. Criticism of the Samkhya view of intrinsic
validity and invalidity 98
5. Criticism of the Bauddha theory of intrinsic

invalidity and extrinsic validity 99

6. Criticism of Mimamsa theory of intrinsic
validity and extrinsic invalidity 101
7. Indian and Western theories of Truth 110

BOOK II

PEECEPTION AS A METHOD OF KNOWLEDGE

(PRATYAK9APRAMANA)
Chapter VI

The Definition of Perception 125

1. Primacy of perception over other methods of

knowledge 126
 CONTENTS IX

Page
2. The Buddhist definition of perception 129
d. The Jama, Prabhakara and Vedanta defini-

tions of perception 132

4 The Nyaya definitions of perception 136

Chapter VII

The Psychology of Perception 143

1. The Senses (indriya) 143

2. Function of the senses 151
3. The nature and function of the mind (manas) 157
4.
3. The self and its function in perception 161

Chapter VIII

Ordinary Perception and its Objects 166

1. Different kinds of perception and the cate-

gories of reality 166
2 Perception of substances or things (dravya) 168
Perception of attributes (guna) and actions
(karma) 174
4 The universal (sarnanya), particularity (vi^esa)
and the relation of inherence (samavaya) 180
5 Perception of non-existence (abhava) 191
6 Internal perception and its objects 199

Chapter IX

Three Modes of Ordinary Perception 207

1. Nirvikalpaka and savikalpaka perceptions 207

2. Eecognition (pratyabhijna) as a mode of
perception 224
ii— (1117B)
X CONTENTS

Page
Chapter X
Extraordinary Perception (Alaukika PRATyAK?A) 228

1. Samanyalak^ana or the perception of classes 228

2. Jnanalaksana or acquired perception 238
3. Yogaja or intuitive perception 248

BOOK III

THE THEORY OF INFERENCE (ANUMANA-

PRAMANA)

Chapter XI

The Nature of Inference 253

1. Definition of Anumana or Inference 253

2. Distinction between perception and inference 254
3. The constituents of inference 256

Chapter XII

The Grounds of Inference 261

1. The logical ground of vyapti or universal

relation 261
2. The question of peiitto principii in inference 274
3. Tlie psychological ground of inference (paksata) 277
4. Lingaparamar^a as the immediate cause of
inference 284

Chai^er XIII

Classification and Logical Forms of Inference 289

1. Bvartha and Parartha inferences 289

 CONTENTS xi

Page
2. Purvavat, Sesavat and Samanyatodrsta
inferences 291
3. Kevalanvayi, Kevala-vyatireki and Anvaya-
vyatireki inferences 292
4. The logical form of inference 297

Chapter XIV

The Fallacies of Inference 306

1. Distinction between a valid and an invalid

reason 306
2. The fallacy of savyabhicara or the irregular
middle 309
3. The fallacy of viruddha or the contradictory
middle 311
4. The fallacy of prakaranasama or the counter-
acted middle 312
5. The fallacy of asiddha or the unproved middle 314
6. The fallacies of kalatita and badhita or the
mistimed and contradicted middles 317
7. The fallacies of ohala, jati and nigrahasthana 319

BOOK IV

upamana OE COMPAEISON

Chapter XV
The Nature and Forms of Upamsna 326

1. The Nyaya definition of upamana 326

2. The Jaina, Mimamsa and Vedanta views of
upamana 327
3. The classification of upamSna 332
xii CONTENTS

Page
Chapter XVI

UPAMiNA AS AN INDEPENDENT SOURCE OF KNOWLEDGE

(Pramana) 335

1. Can upamana give ns any valid knowledge ? 335

2. Can uparnana be reduced to any other
pram ana ? 337
3. Conclusion 340

BOOK V
SABDA OR TESTIMONY
Chapter XVII

The Nature and Classification of Sabda 345

1. The Nyaya definition of sabda and its different

kinds 345
2. Other systems on the nature and forms of

sabda 347

Chapter XVIII

Of Words (Pada) 351

1. Sounds and words 351

2. Words and their meanings 353
3. The import of words 358
4. The unity of words and the hypothesis of
sphota 362

Chapter XIX

Of Sentences (Vikya) 366

1. The construction of a sentence 366

 CONTENTS Xlll

Page

2. The meaning of a sentence 371

3. The import of sentences 375
4. Sabda as an independent source of knowledge 381

Chapter XX

Other Sources of Knowledge 390

1. Different views about the ultimate sources of

knowledge 390
2. Arthapatti or postulation as a source of know-

ledge 393
3. Abhava and anupalabdhi as sources of know-
ledge 400
4. Smrti or memory as a distinct source of know-
ledge 405
6. Summary and general estimate of Nyaya
Epistemology 411

Index 419
 ABBREVIATIONS

BP. ... Bliasapariccheda (Nirnaya Sagar edn.).

NB. ... Nyaya-Bhasya.

NBT ... Nyayabmdutika (Bibliotheca Indica edn.).

NK. . Nyayakandali (Vizianagram edn.).

NL. ... Nyayalilavati (Nirnaya Sagar edn.).

NM. ... Nyayamanjari (Vizianagram edn.).

NS. ... Nyaya-Sutra.

NSV. ... Nyayasutravrtti.

NV. ... Nyayayarttika (Chow kbamba edn.).

NVT. ... Nyayavarfctikatatparyatika ( Do. ).

NVTP. ... Nyayavarttikatatparyapari^uddhi (Bib.Ind. edn.).

PS. ... Padarthadharmasamgraha (Chowkhamba edn ).

SC. ... Siddhantacandrica (Nirnaya Sagar edn.).

SD. ... Sastradipika ( Do. ).

SM. ... Siddliantamuktavali (

Do. ).

SS. ... Sarasariigraha (Benares edn ).

TB. ... Tarkabhasa (Poona edn.).

TC. ... Tattvacintamani (Bibliotheca Indica edn.).

TD. ... Tattvadipika (Calcutta edn.).

TK. ... Tarkakaumudi (Nirnaya Sagar edn.).

TKD. ... Tattvakaumudi (Calcutta edn.).

TM. ... Tarkamrta ( Do. ).

TR. ... Tarkikaraksa (Benares edn.).

TS. ... Tarkasamgraha (Calcutta edn.).

TTS. ... Tattvarthadhigama-Sutra (Bib. Jam. edn.).

VB. ... Vyasa-Bhasya.

VP. ... Vedantapanbhasa (Calcutta edn.).

VS. ... Vai^esika-Sutra (Chowkhamba edn.).

 PKEFACE
The history of [ndian philosophy is a record of many
different forms and types of philosophical thought. There
is hardly any system in the history of Western philosophy
which has not its parallel in one or other of the systems of
Indian philosophy. But of the Indian systems, the Vedanta
has received the greatest attention and it has sometimes
passed .ns the only Indian system worth the name. This is

but natural TIf Vedanta with its sublime idealism has

an irresistible appeal to the moral and religious nature of
man It has been, and wiii ever remain, a stronghold of
“
spiritualism in life and philosophy. It is like one of the
great living wells, which keep the freshness of the eternal,

and at which man must rest, get his breath, refresh

himself.” “The paragon of all monistic systems,” says

William James, “ is the Vedanta philosophy of Hindostan.”

Although we have not such a sublime monism in the Nyaya,
yet its contribution to philosophy is not really inferior in any
way. In fact, the other systems —the Vedanta not except-

ed — have been greatly influenced by its logical and dialectical

technicalities. In their later developments all the systems

consider the Naiyayika as the most powerful opponent and

try to satisfy his objections. The understanding of their

arguments and theories presupposes, therefore, the know-

ledge of the Nyaya.

As a system of realism, the Nyaya deserves special

study to show that Idealism was not the only philosophical

creed of ancient India. Then, as a system which contains

a thorough refutation of the other schools, it should be

studied before one accepts the validity of other views, if only

to ascertain how far those views can satisfy the acid test of
xvi NYiYA THEORY OF KNOWLEDGE

the Nyaya criticisms and deserve to be accepted. But above

all, as a thorough-going realistic view of the universe, it

supplies an important Eastern parallel to the triumphant

modern Realism of the West, and contains the anticipations
as well as possible alternatives of many contemporary real-

istic theories. The importance of the Nyaya is, therefore,

as great for the correct understanding of ancient Indian

philosophy, as for the evaluation of modern Western
philosophy.
The theory of knowledge is the most important part — in

fact, the very foundation of the Nyaya system. This book

IS an attempt to give a complete account of the Nyaya
theory of knowledge. It is a study of the Nyaya theory of

knowledge in comparison with the rival theories of other

systems, Indian and Western, and a critical estimation of

its worth. Though theories of knowledge of the Vedanta

and other schools have been partially studied in this way
by some, there has as yet been no such systematic critical

and comparative treatment of the Nyaya epistemology.

The importance of such a study of Indian realistic theories

of knowledge can scarcely be overrated in this modern age

of Realism
The scope of the book is limited to the history of the
Nyaya philosophy beginning with the Nyaya-Sfitra of
Gautama and ending with the syncretic works of Annam
Bhatta, Vi^vaiiatha and others. It does not, however, con-
cern itself directly with the historical development of the
Nyaya. There are ample evidences to show that Nyaya as an

art of reasoning is much older than the Nyaya-Sutra. We

find references to such an art under the names of nyaya and
vdkovakya in some of the early Upanisads like the Chandogya
(vii. 1.2) and the Suhdla (li). It is counted among the
updhgas or subsidiary parts of the Veda {vide Oaranavyuha,
li ;
Nyaya-Sutra-Vrtti, 1.1.1). It is mentioned under
the names of dnvlksikl and tarkaUstra in some of the
 PREFACE xvii

oldest chapters of the Mahabhdrata {vide sahhd, anuidsana

and ^dnti parvas). We need not multiply such references.
Those here given show that the Nyaya as an art or science of

reasoning existed in India long before the time of Gautama,

the author of the Nydya-Sutra. As a matter of fact,

it has been admitted by Vatsyayana, Uddyotakara, Jayanta

Bhatta and others that Gautama was not so much the
founder of the Kyaya as its chief exponent who first gave
an elaborate and systematic account of an already existing
branch of knowledge, called nydya, in the form of ifitras or
aphorisms. It is in these sutras that the Nyaya was deve-
loped into a realistic philosophy on a logical basis. What
was so long mere logic or an art of debate became a theory
of the knowledge ol leality. It is for this reason that the
present work is based on the Nydya-Suira and its main
commentaries.
So far as the account of the ancient Nyaya is concerned,
my sources of information are mainly the Nydyv-St'dra,
Nydya-Bhdsya, NydyaidiiUlca, Nydyaidrii-ikatatpaiytillid,
Tatparyapan^uddln, Nydyamafipn and NydyasutravitH.
In my account of the modern and sjneretist schools of the
Nyaya, I have mainly made use of Gahge^a’s T attvacxnidmani
with the commentary of Matburanatha, Jagadi^a’s TarMmr-
ta, AnnamBhatta’s Tarkasa'fhgraha and Dtptkd, Yaradaraja’s
Tdrktkaraksd, Kesava Misra’s Taikabhdsd and Visvanatha’s
Kdnkdvali with Siddhdntarnvktdiall and Dinakarl. I have
also consulted several English expositions of Indian philoso-
phy, like Dr. Jha’s Nydya Philosophy of Gautama, Sir B. N.
Seal’s Positive Sciences of the Ancient Hindus, Sir S.
Eadhakrishnan’s Indian Philosophy, Dr. D. M. Datta’s
Six W ays of Knowing Professor Keith' b Indian Logic and
Atomism, Dr. S. N. Dasgupta’s History of Indian Philosophy,
and MM. Kuppusw'ami Sastri’s Primer of Indian Logic.
My indebtedness to these and other works has been indicated
by footnotes in the proper places,

iii— (1117B)
 xviii NYiYA THEORY OF KNOWLEDGE

The method of exposition adopted in the book is com-

parative and critical. I have alM'ays tried to explain and
develop the ideas and theories of Indian philosophy in terms
of the corresponding ideas and concepts of Western philo-
sophy. The great danger of this is the tendency to read,
consciously or unconsciously, Western ideas into Indian
philosophy. I have taken all possible care to guard
against the imposition of foreign ideas on the genuine
thoughts and concepts of Indian philosophy. As a general
rule, the different parts of the Nyaya theory of know-
ledge have been first explained and compared with those
of the other systems of Indian philosophy. For the
sake of completeness, the Indian theories have sometimes
been elaborated in such details as to give one the im-
pression of prolixity I have then undertaken a discussion
of the Indian views from the standpoint of Western philo-
sophy No attempt has been made to a£6hate the Indian
views with parallel views in Western philosophy. Such an
attempt cannot surely do justice to the originality and
individuality of Indian thought. While bringing out the
points of agreement between Indian and Western philosophy,
their difference and distinction have not been ignored and
passed over. I have not been able to support or justify the
Indian theories on all points. It has been found necessary
to modify them in some places and supplement them in the
light of Western philosophy. At the same time, I have
duly emphasised the special contributions of Indian philoso-
phy towards the solution of the problems of knowledge dis-

cussed in Western philosophy.

In conclusion, I tale this opportunity to express my
gratitude first to the late lamented Professor Henry
Stephen, of revered memory, who by his life and teaching
made the study of Western philosophy popular among
Indian students and infused into my jouthful mind the
spirit of an intensive philosophical study. I have nho to
 PREFACE XIX

acknowledge my indebtedness to Sir B. N. Seal, who was a

versatile genius and an eminent authority in Indian and
Western philosophy, and from whom I received great inspira-

tion and valuable guidance in the early days of my re-

searches in Indian philosophy. I have to express further

my deep sense of gratitude to Professor K. C. Bhattacharyya,
a profound thinker and astute metaphysician, who for some
time held the George V Chair of Philosophy in the Calcutta
University. It was my proud privilege to sit at his feet, and
discuss and clear up some of the abstruse problems of logic
and philosophy treated in this book. I have to acknow-
ledge with thanks 'the great help I have received from
MM. Pandit Slt<lram f^astri, of the Calcutta University,
while studying some original works of the Nyaya
philosophy.
I have to express further my most grateful thanks to

the great savant, Sir S Radhakrishnan, George V Professor

of Philosophy, Calcutta University, and Spalding Professor
of Eastern Religions and Ethics, Oxford Ilniversity, for the
constant encouragement, help and guidance I have received
from him in completing this work. My thanks are also
due to my esteemed friend and talented writer. Dr. D. M.
Datta of the Patna College, for reading considerable parts
of the manuscript and for making valuable suggestions. I

am obliged to the authorities of the Calcutta University,

especially to Dr. Syamaprasad Mookerjee, its ex-Vice-

Chancellor, and Mr. Jogeschandra Chakravorti, its Registrar,

for kindly undertaking the publication of the book at the

University Press. I must thank also Mr. Dinabandhu

Ganguh, Superintendent, Mr. Bhupendralal Banerjee,

Printer, and Mr. Jatindramohan Roy, Reader of the

Press, for their help and co-operation in the printing of

this work.

January. 1939. S. C. Cha'iteejee

 CHAPTER I

INTRODUCTION
The Nyaya philosophy is primarily concerned with the
conditions of valid thought and the means of acquiring a
true knowledge of objects. Nyaya as a science lays down
the rules and methods that are essentially iieces.'-ary for a
clear and pi ecise understanding of all the materials of our
knowledge as these are deiived from observation and
authoiity. With this end in view, the science of Nyaya
deals with all the processes and methods that are involved,
either directly or indirectly, m the right and consistent
knowledge of reality. That this is so appears clearly fiom
the common use of the word dnvlhsikl as a synonym for

the NyayaSastia The name dnvtkstkl means the science

ol the processes and methods of a reasoned and systematic
knowledge of objects, supervening on a vague understand-
ing of them on the basis of mere peiception and uncnti-
cised testimony In other woids, it is the science of an
analytic and reflective knowledge of objects in continuation

of and as an advance on the unreflective general knowledge

in which we are moie recepli\e than critical. It is the

mediated knowledge of the contents of faith, feeling and

intuition. Accordingly, Nyaya (literally meaning methodi-
cal study) may be described as the science of the methods
and conditions of valid thought and true knowledge of ob-
jects. In a narrow sense, however, nyaya is taken to mean
the syllogistic type ol inference, consisting ol five proposi-
’

tions called its members or factors.

It should, however, be remarked here that the epis-

temological problem as to the methods and conditions of

valid knowledge is not the sole or the ultimate end of the

I Pruli;fiadipailtakas.miuilayat\uin nyoyaham, Didfnii on TC., II, Chapter on

Nyaya an<l A\ayava

2 NYiYA THEORY OF KNOWLEDGE

Nyaya philosophy. Its ultimate end, like that of the other

8} stems of Indian philosophy, is liberation, which is the

summum bonum of our life. This highest good is con-

ceived by the Nyaya a-; a stale of pure existence which is

free equally from pleasure and pain. For the attainment

of the highest end of our life, a true knowledge of objects

is the sure and indispensable means Hence it is that the

problem of knowledge finds an important place in the Nyaya
philosophy.
But an enquiry into the conditions of valid thought and
the methods of valid knowledge presupposes an account of

the nature and forms of cognition or knowledge in general.

It requires us ak'o to consider the nature and method of
valid knowledge in general and the nature and test of truth

or validity in particulai. Hence the pielinnnary questions

that arise in the Nyaya theory of knowledge aie What
IS cognition or knowledge as such ^ What are its different
forms ? What is valid knowledge ? What is meant by a
method of valid knowledge in general ^ What do we mean
by truth or validity ? What is the test of truth, the
measure of tine knowledge, the standard of validity

What are the constituents or factors of valid knowledge'?

It 18 a matter of historical interest to note here that,
among other things, the problems of knowledge in

general and those of the methods of valid knowledge in

particular were brought home to the Naiyayikas by the

Buddhists and other sceptical thinkers of ancient India in

the course of their scathing criticism of the realistic philo-

sophy of Gautama. '
They set at naught almost the whole

^ This 18 clear from the opening verie of Uddyotakara’s Nyayaidrittha, in whub

it 18 meniioned that the object of the Vdriiika la to reiiio\e the misconceptions
of the critics of Gautama’s teaching even though it was well explained by ihe great
commentator Vatsyftyana In explaining this verse V^aspati mentions the name of
Dignaga, the great Buddltist logician, as one of the hard critics of Gautama’s philo-
sophy. In the NydyavdrtUkatdtparya^ikd he has given a dear account of some of the
difficulties raised by the Buddhists ip connectiop with the question of fremdna
 iNTRODtJCTlON 3

of the Nyaya philosophy as an edifice built on sand. The

Nyaya teaches that the highest good is attainable only
through the highest knowledge. But the theory of know-
ledge in it IS a vicious circle. It takes upon itself the
futile task of Kant’s first Critique where he examines reason
in order to prove the validity of thought and reason. “If
it IS the business of Kant in his Critique of Pure Reason
to show how mathematics is possible, whose business is it

to show how the Critique of Pure Reason itself is

”
possible ? ’
With regard to the Nyaj a theory of knowledge
a similar question is asked by the Bauddha critics. It is

pointed out by them that a criticism of knowledge must be

made by the instrument under criticism and thereby pre-
supposes the very thing in question. Thus the validity of

knowledge is made to rest on tlic validity of the methods of

knowledge To maintain that our knowledge is true we

must prove that it is really so, tliat it is derived from a

valid method of knowledge wdneh always gives us true

knowledge and never leads to a lalse result. But, then,
how are we to know the validity of that method of know'-
ledge ? From the nature ot the case, the task is an im-
possible intellectual feat.
With regard to the knowledge of validity there are two
possible alternatives. The validity of knowledge may be
cognised by itself, i.e. be self-cognised. Or, the validity
of one knowledge may be cognised by some other knowledge.
The first alternative that knowledge cognises its own valid-

ity is inadmissible. Knowledge, according to the Nyaya,

cognises objects that are distinct from and outside of itself.

It cannot turn back on itself and cognise its own existence,

far less its own validity. Hence no knowledge can be the
test of its own truth. The second alternative, that the valid-
ity of any knowledge is tested by some other knowledge, is

1 The New Realisnit p. 61.

4 NYAYA THEORY OF KNOWLEDGE

not leas objectionable. The second knowledge can at best

cognise the first as an object to itself, i.e. as a paiticular

existent. It cannot go beyond its object, namely, the first

knowledge, and see if it truly corresponds with its own

object. An act of knowledge having another for its object

cognises the mere existence of the other as a cognitive luct.

It cannot know the further fact of its truth or "falsity.

Moreover, of the two cases of knowledge, the second,
which knows the first, is as helpless as the first in the
matter of its own validity It cannot, ex hypothesi, be the
evidence of its own validity. Hence so long as the validity

of the second knowledge is not proved, it cannot be taken

to validate any otuer knowledge. It cannot be said that
the second has .s’lf-evident validity, so that we do not want
any proof of it. This meins that one knowledge, of which
the validity is self-evident, is the evidence for the validity ol
another. But if the truth of one knowledge can be self-

evident, why not that of another ^ Hence if the second

knowledge has self-evident validity, there is nothing to
prevent the first from having the same sort of self-evidence.

As a matter of fact, however, all knowledge has validity

only in so far as it is tested and proved by independent

grounds. Truth cannot, therefore, be self-evident in any
knowledge. If, by such arguments, the validity of know-
ledge itself is made incomprehensible, there can be no possi-
bility of assuring ourselves of the validity of the methods of
knowledge, such as perception, inference and the rest. The
value and accuracy of a method of knowledge are to be
known from the validity of the knowledge derived from it.

It follows from this that if the validity of knowledge is un-

knowable, that of its method is far more unknowable. ^

Hence we are involved in a vicious circle ; the validity of

knowledge depends on the validity of the method of acqinr-

(I NVT.pj. 4-5
 INTKODUCTION 6

lug such knowledge, while the validity of the methods is to

be tested by the knowledge derived from them. As Hob-

house puts the matter
•
“ Our methods create and test our
knowledge, while it is onfy attained knowledge that can test
them.”' It is the contention of the Bauddha critics that

the Nyaya theory of knowledge is involved in such circular

reasoning in the attempt to prove the validity of knowledge.
This contention, if admitted, renders the Nyaya philosophy
uttterly worthless. It becomes a hopeless attempt to realise

the highest good by means of the highest knowledge which

IS impossible.
It was with the object of meeting the difficulties raised
by its critics that the old Nyaya entered on a critical study
of the problem of knowledge in its relation to reality. After
Vatsyayana’s first elaborate exposition of Gautama’s Nydya-
Sutra, his worthy successors had to defend the Nyaya
against renewed attacks. They discussed both the logical
and metaphysical problems more fully and also many other
questions of general philosophical interest. The result is a
^
fully developed and complete system ol philosophy.

The modern school of the Nyaya, beginning with

Gangesa, attempts to give greater precision to the thoughts

of the old school It lays almost exclusive emphasis on its

theory of knowledge. The forms and concepts invented by

it give the Nyaya the appearance of a symbolic logic. The
old theory of knowledge is a criticism of thought as related

to the real world of things. It is more empirical and prac-

tical, and it tries to discover the relations between reals.

The modern theory becomes more formal or conceptual.

It tries to find out the relations of meanings and concepts.
It develops into a formal logic of relations between concepts
and their determinants. The old Nyaya gives us w'hat may
1 Hobhouse, Theory of Knowledge^ p 487
2 An account of the controversy between the Ndi3ayikas and the Bauddha logicians
IS given in Dr S C Yidy&bhmixm's History of Indian Logic^ Bk IT, Ch 11
6 NYiYA theory OE KNOWLEbOE

be called philosophical logic, while the modern Nyaya is

formal logic and dialectic.

The Syncretist school develops the Nyaya further by

incorporating the "Vai^esika theory withini it. The cate-

gories of the Vaise§ika become a part of the objects of know-

ledge iprameya) in the Nyaya. But this synthesis of the

Nyaya and the Vai^esika does not ignore their differences

with regard to the theory of knowledge. One is as severe

as the other in its criticism of the opposed logical theories.

The Nyaya theory of knowledge is the cumulative body

of the logical studies and their results in the different schools

of the Nyaya. It may be said to have three aspects •

the

psychological, the lexical and the philosophical. The first

is concernd with the descriptive analysis of the facts of

knowledge. The second is interested especially in the

criticism of the forms and methods of knowledge. The

third 18 an attempt to determine the final validity of know-
ledge as an understanding of reality. These aspects of the
Nyaya epistemology, however, are not to be found in

abstract separation from each other. In the next chapter

we shall have to discuss the mainly psychological questions

as to the nature and forms of knowledge.
 BOOK I

THE METHOD OF VALID KNOWLEDGE

(PRAM ANA)
 CHAPTER II

THE NATURE AND FORMS OP KNOWLEDGE

1. Definition of Knowledge (huddhi)

If we take knowledge in its widest sense to mean any

way of cognising objects, then valid knowledge \^ill be a
special form of cognition {huddh). All cognitions are not

valid knowledge. Hence in order to understand the nature

of the method of valid knowledge {piamana), vie have to
consider first the nature and diffuient forms of cognition or
knowledge {pidna) as such.
In the Nyaya-Vaisesika philosophy cognition {huddhi)
IS taken to mean the same thing as apprehension {upa-

lahdhi), knowledge (p'ldna) and cognisance (pratyaya).^

Hence we say that knowledge means awareness or apprehen-
sion ol objects. It includes all cognitions that have a more
or less determinate objective reference. The object of appre-

hension may be a thing or a quality, an act or an emotion,

the existent as well as the non-existent. But in every case

in which there is knowledge there must be something that

stands out as the object of knowledge. Knowledge consists

simply in the manifestation (prakah) of objects. ® All things

are made manifest or revealed to us when they become ob-

jects of knowledge. Further, knowledge is said to be the

property of illumination or manifestation that belongs to the

^ Bu(ltJliirupa)ab<)ltir]naQamityanarthantaram, NB ,
1.1.15

BuddliirupalabJhirjnauam pratjaya iti part ayah, NK., p 171

2 Arfchapraka^o biuldluh, TK., p 0,

2— (ni7B)
10 NYiYA THEORY OF KNOWLEDGE

self. Without this luminous light of knowledge we lose the

ground of all rational practice and intelligent activity. It is

on the basis of knowledge of some kind that call living beings

deal with other objects of the surrounding world. Hence

knowledge (buddhi) is regarded as the ground of what may
be called the behaviour or conduct of a living being. A
living creature behaves differently in relation to different

objects because it somehow knows them to be different.

Then we are told more definitely that knowledge is that kind

of awareness which is meant when, by intiospection, one

^says ‘
I am knowing.’ ’ This means that knowledge is in-

tellection as distinguished from affection and volition.

Something different is meant by the phrase ‘lam knowing,’
from what is meant by saying ‘
I am desiring or willing or

doing something, or sintply being pleased or displeased with

it.’ Although knowledge is distinguishable, it is not sepa-

rable, from feeling and volition. In knowledge the knower

does not passively allow himself to be impressed by external
objects and end by having mental copies of those objects.

According to the Nyaya, the self is not a mere aggregate or

series of conscious phenomena, which is only acted on and
determined by sense-impressions, but has no power to react
on and determine them. This materialistic and sensa-
tionalist theory of the self is rejected by the Nyaya. On the
other hand, it conceives the self as a conscious agent which
receives impressions of sense, knows external objects through

them and acts upon things according to its subjective

purposes. Knowledge is a cognitive fact by which we have

‘an apprehension or understanding of objects. But it is

bound up with certain affective elements, namely, the

feelings of pleasure and displeasure, according as the known
objects are pleasurable or painful. Through such feelings

.knowledge leads to certain conations, viz. desire, aversion

1 TS and TD , p 32
 UATURB AND FORMS Ot KNOWLEDGE ll

and volition in the form of an exertion (samiha) to obtair

^
pleasurable objects and avoid painful ones.
Hence knowledge may be said to be a cognitive pheno-

menon which is always connected with conation through the

mediation of feeling In any particular act of knowledge of
an object, there is a feeling of being pleased or displeased
with it and an active attitude of desire or aversion which
may lead to ceitain overt movements towards or away from
the object. The Nyaya, however, does not go so far as to say
that knowledge is at once a phase of cognition, feeling and
conation. In cognising, an object we may also cognise its

pleasurable or painful character and also become conscious of

certain tendencies m relation to it. But the actual feelings

of pleasure and pain or the conative processes of desire, etc.,

take us beyond cognition. Knowledge is not a phase of

feeling or the will, although it may be always connected
with them It has a distinctive and self-sufficient character

of its own and should not be reduced to feeling or volition.

With regard to the essential nature of knowledge we may
ask Is knowledge a substance or an attribute ? Is it a

mode or an activity ? According to the Nyaya, knowledge

is an_ attribute of the self It is not a substance, since it

cannot be the siuff or the constitutive cause of anything, nor

IS it the permanent substratum of certain recognised and
variant properties. The Samkhya and the Yoga systems
look upon cognition as a substantive mode or modification
{vrtti) of the material principle called huddhi, as it reflects

the light or consciousness of the self in it This, the

Naiyayika contends, is unintelligible. We cannot understand
how the self’s consciousness, which is immaterial and intangi-
ble, can be reflected on any material substratum. We should
not speak of any reflection, but rather say that knowledge or
consciousness belongs naturally to huddhi itself. But this

> UB , 1 1 1-2
12 NYAiA TUL.ORY OF KNOWLEDGE

will commit us to the absurd hypothesis of two selves or

subjects for any case of knowledge. In truth, however, there
is but one conscious subject for ail cognitions in one person.
It is generally believed that knowledge is neither a mode
nor a substance but a kind of activity or function ihnijd).

The Bauddha and the Mimamsa systems agree in describing

knowledge as an activity, a transitive process. ^
The Nyaya
however emphatically repudiates the conception of knowledge
as an Jayanta in his Nijnyamaiijari (p 20) traces
activity.

the act theory of knowledge to a grammatical prejudice, a

lonfusion between knowledge as manifestation and the verb ,

‘
to know ’
as denoting an action . When we hear the c\-
pressions ‘
I know, ‘
I cognise,’ etc., we are apt to be misled
nto the belief that knowledge or cognition is an activity or
process. But this only shows how in piiilosophy we may be
deceived by the vague expressions of ordinary language
Knowledge, although it is not an activity ot any kind, is

still a transient phenomenon as appears from the throe

tenses of the veib ‘
to know ’
It is a dated event which is

to be regal ded as a quality and so can be perceived like

physical qualities Just as physical qualities are perceived

by their special sense organs, so knowledge is perceived by
the internal sense called minis.
'

But knowledge canjiot be

the quality of any material substance, since, unlike that,
it does not admit of external perception Physical properties
are perceived by the external senses, but knowdedge is not so
perceived. Being thus fundamentally different from all phy-
sical qualities, knowledge is to be regarded as the property of

an immaterial substance called soul Still, knowledge is not

an essential attribute of the soul The soul has acquired
this property in its bodily setting, i.e. in relation to a body.

» NB NVT NSY
, ,
ard NM .n NS, 1 I 15 See also NS 32 1 IT

2 Juandknya hi sakirmika, fiasUadtpika, p 56 T/ aho Nijayahindvtila, Ch I

^ NM , p 4(K) , TB , p IR
 NATURi: AKD FORMS OF KNOWLEDGE 1^

To the Ailvaita Vedanta, knowledge or consciousness is just

the sell, the very stull of it. For the Nyaja, knowledge
appeals as the result of a lelation between the soul and the
body, which in themselves are not knowledge liut when it

does appear, it has to exist as an attribute inhering in the

soul substance.

Knowledge, as an attribute ol the self, is always directed

to objects, it always refeis beyond itself, t .c. to objects out-

side of and different fro in itself Knowledge is never sclf-niani-

lested. The capacity of self-rnanifestation in knowledge is,

accoiding to the Nyaya, a mere hypothesis of the Samkhya-

Vedanta and the Piabhakara Miniamsa Cognition cannot
cognise itself, ft can giasp, not itself, but an ‘other.’
Knowledge is not indeed, like the will, a way of acting on
othei objects, only it refers oi points lo something else. We
shall have to considei later the question as to how' know-
ledge can be known. The direction towards an object is

what has heed called ‘

intentional inexistence ’
by Brentano
and Meinong. They take it as a character common to all

psychical phenomena. The Nyaya, however, limits it to

cognition and denies to cognition the capacity of being

directed to itself, i.c. being self-cognised
Fiom what has been said it w'lli appear that knowledge
IS conceived by the Nyaya in a very wide sense. In Western
philosophy thought or consciousness, as a cognitive fact,

has sometimes been regarded as an essential attribute of the

mind and a pervasive character of all mental phenomena.*
The Nyaya, however, does not pass over the distinction
between thought (yndna), on the one hand, and feeling,

including pleasure and pain, desire and aversion, and will,

on the other, bnder knowledge it brings together all cogni-

tive facts, like sensation, perception, inference, memory.

^ Vi’fianamanatmasainvetiandin, N VT , p 4

* Vide Russell, Analyi>is of Mmd pp 14 f

* Cf Descaites a distinction of m extensa and re* togitans.

14 NYAYA THEORY OF KNOWLEDGE

doubt, dream, illusion and tbe like. In this sense the

onddinot the Nyaya corresponds to cognition which,
placed by the side of feeling and will, gives us the tripartite
division of mental phenomena in the traditional school of

Western psychology. It stands, as Alexander also has said,

*'
for all kinds of apprehension of objects, whether sensation,
*
or thought, or memory, or imagination, or any other.”
So far the Nyaya view of knowledge seems to be just
and comprehensive. But, then, a more fundamental problem
IS raised. It is the ontological problem of the status of

knowledge as a fact of reality. Is knowledge a qua lity^ or ^

relation, or an activity ?
First, we have the act theory that knowledge is an acti-

vity. It IS not difficult to see what induced some philoso-

phers to accept this view of knowledge. There can
be no knowledge unless the mind responds to

the influences of the surrounding world At any

moment of inattention or absent-mindedness we do not per-
ceive sounds or know things other than those in which we
are engrossed, although the sounds or things may be acting
on our senses. If there is to be knowledge, the mind must
react to the actions of other things on it. Knowledge is not
a reflection of objects on the mind which receives them
passively like a mirror or reflector It is a process in which
themind actively reaches out to o biects and illuminates
them. Hence knowledge must be a kind of activity, rather
it IS a mental activity.
The act theory of knowledge has been accepted by
various schools of philosophy. In Indian philosophy, the
Bauddha and the Mimamsa systems uphold it For the
former, to exist is to act and so to change. Knowledge as
an existent fact consists in the act of showing and leading to
an object. According to the Mimamsaka, the act of knowing

^ Cj Spate, Time and Deity Vol II, p 82

 NATURE AND FORMS OF KNOWLEDGE 15

{]mnaknya) refers to au object. For Kant also knpwledge

^
involves the synthetic activity of the understanding. Spencer
tells us that consciousness arises when the tendencies
towards action counteract one another and are therefore
thrown back on themselves so as to become conscious of
their existence, and knowledge appears as an incident in the
adaptation of the organism to the environment. For Bergson
also consciousness is a ceaseless creative activity. The
voluntarists identify knowing with willing when they hold
that cognition is the will when it is thwarted by difficulties

and so looks for {i.e thinks) means to overcome them.

With the pragmatists knowledge is a belief determined by
the will. For nuo-idealists like Cioce and Gentile ^ knowing
IS the form of theoretical activ'ty and in thinking we create
the thought we think about. Alexandei,^ who is a realist,
seems to treat knowledge as a mental act when he says that
‘
every expeiience may be analysed into two distinct ele-

ments and their lelation to one another, namely, the act of

mind or awareness and the object of which it is aware, and
that the one is an-ing and the other an-ed ’
The Behavi-
ourists ‘ go to the other extreme and identify knowing with
the activity of the body. They hold that consciousness is

implicit behaviour, thinking is sub-vocal speaking, and

knowledge is a particular kind of behaviour in animals, or
such response to the stimulus as has the characteristics

of appropriateness and accuracy.

We may dismiss the behaviouristic contention that
knowledge is a particular kind of bodily behaviour. That
there is any behaviour, explict or implicit, can be known
only if there is a knowing subject. Behaviour cannot ex-
plain knowledge, but presupposes knowledge in order to be

^ Principles of Psychology^ \ol I, Ft IV

2 Of OEM Joad, Introduction to Modern Philosophy ^Ch 3
3 Space, Time and Deity, Vol. I, pp 11-12, Vol II, p 86
^ Cf. Watson, Behaviour , Bussell, Analysis of Mind, pp 256 fif.
1C NYiYA THEORY OF KNOWLEDGE

understood. Further, from all we know about the conduct

of living beings it appears that behaviour arises out of

conduct and is not identical with it. Behaviour may be

the objective side of knowledge It has also a subjective
side which is reflected in behaviour or overt action. This is

recognised by Russell in his Outline of Philosophy, in which

he supplements the objective view of knowledge as a way of

reacting to the environment, by the subjective view of it

as an awareness. The Naiyiiyika rejects altogether the act

theory of knowledge as a giarnmatical prejudice, and ex-

cludes knowledge from the category of karma or action.

E\en if we suppose that knowledge is an activity, the

question will an se Whit is tlie nature of this activity ?

It cannot be any kind ot phjsical activity, force or motion.

Nor can it be a psychical ac livity The existence of any
activity in the mind or consciousness is a highly question-

able tact. According to James, ^

the will is a relation

between the mind and its ideas, and in willing there is no

inneivation or putting forth of energy by the mind into the
Ixidy. Titcheiicr and some modern psychologists also

endorse this view and exclude the will from among the
elementary mental processes The Biiagavadglta '
antici-

pates thise modern psychologists when it says that ‘

all

actions take place in the material world and it is only

egoism that deludes the self into the belief that he is an
agent.’ In so far as this is true, we cannot speak of know-
ledge as an activity except by way of metaphor In know-
ledge itself as .an awareness we find an object that is cognised
and a subject or self that cognises it, but not any activation
or energisation. An act is as much an object of knowledge

as any physical thing, quality, or action. It is manifested

by knowledge, and is not identical with it. On the other

* The Principles of Psychology Vol II, Cli XX Vf

^ A Text-Book of Psychology Sec. 10.
3 Ad, III, 8127.
 NATURE AND FORMS OF KNOWLEDGE

Jiand, knowledge appears as a standing and an accompi

fact which manifests everything that comes before it

is more like a static illumination than a sweeping flow

conscious Hence knowledge is different from actiorr.
stuff.

Among modern writers Moore and Broad refute the act

theory of knowledge so strongly advocated by Dawes
Hicks.'
The second theory with regard to the nature of know-
ledge is that it is a relation between certain entities.
According to Meinong, tlie Austrian realist, and the
Critical realists, kno\yledge is a relation between three
tcjrnas, viz. a mind, an obiect, and a content. When
I know the table, my mind comes into relation
with a physical object through the content of tableness.
In The Problems of Philosophy Russell seems to accept
the view that knowledge. 4g^jau..^h/ee:terna„jre^^
Some other realists hold that knowledge is a relation between
two terms, namely, a mind and any object. Moore reduces
cognition to the holding of a relation between a sense datura
and a character. Broad also agrees with Moore in this
respect and denies the existence of any mental act. ^ Russell
in his work Our Knowledge of the External World reduces
the knowledge-relation to a two-terra process, i.e a relation
between the mind and the external world. The Neo-Realists
go further and reduce knowledge to a relation between one
kind of terms. According to them, “knowledge is not a rela-
tion between a knowing subject and an object known. It is

merely a special sort of relation between objects.” ® In the

words of the new realists, “Things when consciousness is

had of them become themselves contents of consciousness ;

and the same things thus figure both in the so-called external

^
Cf Tj A Belli, Knowledge and Trviht pp 186 f,

2 Cf L A Beid, op cit,

3 Essays m Cnttcal Reahsm, p 89,

3— (1117B)
18 NYiYA THRORY OP KNOWT.EDGE

*
world and in the manifold which introspection reveals.”
Russell advocates this theory in The Analysis of Mind.
James in his Essays in Radical Empiricism reduces knowledge
directly to a relation between one type of entities. According
to him, knowing can be easily explained as a particular sort

of relation into which portions of ‘ pure experience ’

may
enter. The relation itself is a part of 'pure experience,’
one of its terms becomes the subject or bearei of the
knowledge, the knower, and the otlier becomes tbe object
known.
The relation theory of knowledge does not stand the test

of sound criticism Although the relation between the

subject and the object takes the form of knowledge,
knowledge itself is not a relation. we seem to be
All that
justified in saying is that knowledge appears when the
subject becomes related to the object, but it is a new pheno-

menon other than the subject-object relation. According to

the Naiyayikas, knowledge may be said to arise ultimately

out of the relation between the soul and the body Still, it

is not merely a relation between the two, but a new property

accruing to the soul therefrom. Whether knowledge can
be treated as a quality or not, we shall consider next The
point we are to stress here is that a relation as such is not
a cognition but a cognitum^ i.e. an object ot cognition .

Of course, when a thing is known, it enters into what we

call the knowledge-relation. But this assumes the subject’s
awareness of the thing as the basis of the relation. So the
relation cannot constitute knowledge. As Reid has said,
‘
knowledge is not itself a relation but the apprehension of
relations’^ Supposing that knowledge is a relation, we
ask : How do we know it ? It must be through some other
knowledge which, therefore, transcends the relation and is

1 The New ReaUsm, p 85

3 Jj A Ueid, op at p 189, ,
 Nature and forms oN KNOwi-EbOE 19

not identical with it. In fact, the subject-object relation

does not produce knowledge bat only serves to manifest it,

just as the contact between the eye and a physical thing

serves to manifest its colour but does not produce it.

The third view with regard to knowledge is that it is a

quality. According to Descartes and his followers, thought
or cognition is the essential attribute of the mind or the soul
substance, just as extension is the essential attribute of
matter. The Samkhya and the Yoga systems look upoii
knowledge or cognition as a modification of buddhi or the
intellect which is its. substratum. The Ramanuja school
of the Vedanta takes knowledge as an essential quality of
the self. The self is not, as the Advaitins say, itself

knowledge but is qualified by knowledge Knowledge is not

the essence of the self, but an attribute owned by the self.

The Naiyayikas and the Vaisesikas also advocate the quality'

theory of knowledge. For them, knowledge is an attribute
which inheres in the soul substance which, however, is

separable from it.

But the quality theory of knowledge also involves certain

difficulties. It cannot account for the reference to objects
that IS inherent in knowledge A quality is an intransitive
property of a thing. It hangs on the thing and does not
point to anything beyond. It is in activity that we find a

tiansition from one to an ‘

other.’ For one thing, to act

means aggressively to reach another. But at the same time

we must not overlook the distinction between the ideal ‘

reference ’
to object that we find in knowledge and any form
of physical processor transeunt causality. Knowledge refers'

to Its object and is in this sense a cognition of the object. It,

docs not however move towards it. In it there is no transi-

tion from point to point in space In so far as this is

the case, the Nyaya is right in opposing the attempt to

identify knowledge wuth activity. But the view ol know-
ledge as a quality misses the other fact of objective
 20 NYAVA THEOfeY OF KNOWLEDGE

refereoce that we find in knowledge. Knowledge seems

to be what Reid calls ‘
a self -transitive process ’
Its self-

transcension is, as Hoernle points out, directly experienced

by us. *
So it seems to occupy a position intermediate
between quality and activity. To describe its self-transcen-
sion or objective reference and, at the same time, demarcate
it from physical activity, we may say that knowledge is

an ‘
ideal activity. ’
But after all the characterisation of

knowledge as an activity, be it physical or ideal, is only

a symbolic description. While physical activity is real

and intelligible, an ‘
ideal activity ’
can hardly be made

j
intelligible to us. Knowledge is, therefore, neither a

t quality, nor an activity, nor a relation.

^ Knowledge is the most fundamental fact of reality. It

is the intrinsic character of all reality. Wilhout pausing

here to discuss the ontological problem as to the nature
and constitution of reality, we may say that reality is a
living intelligent system. The ultimate constituents of
things are not material but living particles which not onl\
exist and interact with one another, but somehow experience
their existence and activity. These many living particles

are the differentiations of one universal spiritual life. A

pluralistic constitution of things is not inconsistent with
the unity of their ultimate ground which is the Absolute
Reality. The real is, therefore, an objective system, of

which existence and knowledge are two inseparable aspects.

It is true that what we know as the consciousness or
knowledge of human beings is a specific phenomenon
determined by certain conditions as the subject-object
relation or the activity of the nervous system and the brain.

But while these conditions explain the specific form and

character of this or that knowledge, they cannot account
for the original sentience or experience which is embedded

^ L ^ Hcjd, op at ,
p 188
 NATURE AND FORMS OF KNOWLEDGE 21

in reality and conditions those conditions themselves.

Hence we conclude that knowledge is present in all reality
and 18 manifested in a specific form in man by the subject-
object relation It does not require to be attached as a
quality to any other reality, say, matter or mind or soul. It

is just the self-expression of reality. In the words of

Bosanquet ‘
we may say :
“ Knowledge is an essential form
of the self-revelation of the universe ; experience as a whole
IS the essential form.”

ClaSsiftcation of Knowledge

Taking knowledge in the most comprehensive sense as

the cognition of objects, the Naiyayikas proceed to distinguish
between its different forms, according to the differences in
the nature and accuracy of cognitions. In view of these,

knowledge is first divided into anuhhava or presentation

and smrh or memory. ^ In anuhhava there is a presenta-
tional knowledge of objects and so it is felt to be given to
1^. It is oiiginal in character and not the reproduction
of a previous knowledge of objects. Smrti or memory, on
the other hand, is not the presentation of objects, but a
reprodu ction of previous experien ce. Here our knowledge
appears to be due not so much to objects themselves as to
out past cognitions of those objects. Each of these has

been further divided into valid (ijathartha) and non-valid

(ayathdrtha) forms, according as it does or does not accord
with the real nature of its object-

Under anuhhava or prescntative knowledge we have

the two kinds of valid and non-valid presentations. Of
these, the former is called pramd and includes all cases

of true presentational knowledge of objects. According

1 Logic, Vol IT. p 322

2 8a dvjvjtiba sjuftiraDubhava^ca, T8 and TD.| p 32
22 tJYAYA tHEOilY OF KNOWLEDGE

to tbe Nyaya, there are four distinct kinds of pramd or

valid presentation, namely, perception (pratyahsa), inference
(anumdna), comparison (upamana), and testimony (hhda).
In each of these there is a presentation of some object as
it really is. Hence pramd, according to the Nyaya, is not
any cognition nor any true cognition as such. It is a

valid presentational knowledge of objects. *

As a matter
of usage, however, the word ‘
knowledge ’
may be used for
pramd, according to the context.

There are some cases of knowledge which arc presenta-

tional in character but not valid These constitute the
class of apramd or non-valid presentations {ayathd)thdnu-
hhava), which includes all cognitions that are either false

or not-true but not false Hence under aprama the N}aya

includes doubt (samsaya) \\ith its varieties of conjecture

(tt/ia) and indefinite cognition (anadhyavasdya) as well as

error {viparyyaya) and hypothetical reasoning (tarAo). ^
It

should here be noted that cognitions w'bich do not agree

with the real nature of their objects are not always false

or erroneous (bhrama). There may be cognitions which

fail to give us a correct presentation of objects and so are
not true (pramd). But at the same time they may not
make any claim to truth, nor lead to any definite assertion.
Such 18 the case with doubt, conjecture, indefinite cognition
and tarka. •These are not true indeed, but yet they are not
false (viparyyaya). It is in view of such facts that the

Nyaya divides non-valid presentation (ayathdrthdnnhhava)

further into doubt (samiaya), error (viparyyaya) and hypo-
thetical argument (tarka). Hence it is not correct to speak
of apramd or non-valid presentation, always as a case of
bhrama or error. It becomes so w’hen it definitely con-
tradicts its object (viparltanirnaya)

^ Tsttvanubhava prama, Saplapadarfht, 140

* A 3 atbal'thanuhha^astn^^’h <li s.irfo^aynvjpHryyayatnrkabliedat, TP and TD
p 82 Ubanadbyavosayayoslu 8anj4 13 a eva, 6^ap/apaddrl/f*, 39
 NAiURE AND FORMS OF KNOWLEDGE 23

Memory is not prama or valid knowledge, since it does

not refer to presented objects. It may be of two kinds,
namely, true and false (yatharthamayathartham). True
memory is in accord with the real nature of the objects

rememi)ered, whereas false memory does not tally with the

real character of the remembered objects. In waking
life we have both these kinds of memory. In dreams our
cognitions are false mernory-cognitions. Dream is a kind
of memory that is not in agreement with the real nature

of the cognised objects. All knowledge, however, including

dream, refers to some real object ; only dream is a false
memorial representation of the real.
We may repiesent the Nyaya classification of knowledge
by the following table

Knowledge (huddhi)

Presentation {amibliava) Memory {smrU)

Valid {prama) Non-valid (apiama) True False

I
{yathdrtha) {ayathartha)
I

Doubt Error Hypotbetical Argument

{sarnaaya) (viparyyaya) {iarka)

Perception Inference Comparison Testimony

(pratyakaa) (anurndna) (upamdna) {iabda)

8. Memory and Dream

Memoiy (smrti) is knowledge of one’s own past. It is

a representative cognition of past experiences due solely to

the impressions produced by them. It is thus different from
recognition {pratyahhi]M) which, according to the Nyaya,

1 TB., p. 30
2 Samskaratnatrajanyara ]fianam smrti h, TS p
,
32,
24 NYAYA THEORY OF KNOWLEDGE

is a form of qualified perception and has reference to the

direct presentation of some object, although it involves an
element of representation. *
In memory, however, there is

only a revival of our past experiences, in the form of ideas

and images, in the same form and ordei in which they were
actually experienced by us at a certain point of past time.

The ground or condition of this revival is of course the

latent impressions left by our past experiences and retained

by the soul. ^ When the mind comes in contact with such
psNchic dispositions {bhdvana) there is a remembrance of

the corresponding original experiences. Memoiy being

thus a cognition, by the same self, of what has been once
cognised, is an evidence for the soul’s permanence. As to

the general character of memory we may, therefore, say

that It IS knowledge arising solely out of the impressions of

pre\ious experiences and pertaining to a permanent soul.

While memory has for its general conditions some

original past presentation (purvdnuhhava) and its impres-
sion (samskdra), it has a number of specific causes that

serve either to retain the impressions or revive them m

consciousness, and thereby bung about the phenomenon of
memory. Among these are (1) attention (pranidhdna)
which fixes anything m the mind, (2) association (mhandha)
which connects different experiences and makes them
suggestive of one another, (3) repetition {abhydsa) which
secures persistence for the impressions, (4) sign (linga) that
leads the mind to the thing signified, (5) characteristic
mark {laksana) that recalls the class to which an object
belongs, (6) similarity {sddHya) that associates the ideas
of like things, (7) ownership iparigraha) which is suggestive
of the owner or the thing owned, (8) dependence (dsraydhita-
sarhbandha) of which one term suggests the other, (9)

TD,p 83
Apubhuvajanya amj-tihetuibhavana, &tala^iatravfttl^^, TS , p 86,
 NATURR AND FORMS OF KNOWLEDGE 26

contiguity {anantaryya) which binds together successive

phenomena, (10) separation (mi/opa) that frequently reminds
one of what he is separated from, (11) identity of function
(ekakdryya) that recalls similar agents, (12) enmity (tirodha)
that suggests the livals in any sphere, (13) superiority
(ahhya) that reminds us of what it is due to, (14) acquisi-
tion (prapti) (hat frequently recalls its source, (15) covering
{vyai adlidna) that suggests what is covered, (10) the feelings
of pleasure and pain (sukhaduhkha), (17) desire and aversion
{icchddvem) (18) fear (hhaya), (19) need (arthitia)
that reminds one of their causes and objects, (20) action
(kriyd) which is suggestive ot the agent, (21) the feeling
of ahcction (/tf(,a) that often reminds us of its object,
(22) meiit (dhanna) and (id) demerit {adharim) that
are suggestive of tlie belief iii pre-existence and help
or hinder the retention of experiences. These causes of
memory (annot be simultaneously opeiative. Hence re-
collections are not simultaneous but successive in their
appearance in consciousness.'
Memory is of two kinds, namely, true (yathdrtha) and
false (ayathdrtha) It is true when it has its basis in some
valid piesentation {prarnd janya) and is in agreement with
the real nature of the remembered objects. On the other
hand, memory is false when it arises out of such original
cognitions as were erroneous [apmmdfanya) and so does not
accord with the natuie of the objects iccalled in it. Thus
the tiutli and falsehood of inomoiy depend on those of the
corresponding original preseutative cognitions {pummn-
hhava) that constitute the ground ot all memory ^ In
waking life we have both these kinds of memory. The
voluntary or involuntary recollection of past objects, when
we are awake, becomes true or false according as it is con-

‘ NS and NB ,
3 U 44
2 Purvanubhavasya yatharthatvayathartbatvabbyaoa smara^amapi ubhayaruparn
bbavufc], TM
4— (1117 B)
2(> NiAYA THEORY OF KNOWLEDGE

nected with right or wrong cognitions in the past and so, is

or IS not in accord with the real nature of the objects

remembered.
Dreams illustrate what is intrinsically false memory.
According to the Nyaya, dream-cognitions are all memory-
cognitions and untrue in character.^ They are brought
about by the remembrance of objects experienced in the

past, by organic disorders and also by the imperceptible

^
influences of past desires and actions {ailrsta) Hence
dream-cognitions have sometimes a moral value in so lar as

they produce pleasurable or painful experiences in the self

according to the merit or demeiit accruing from the actions

of waking life Dream-knowledge, however, is intrinsically

false It IS no doubt related to certain objects of the real

world But these objects as cognised in dream are not pre-

sent to serine. They are either past or remote. Still in

dreim, objects are actually lepresented as present Hence

there is in dream a false cognition of the real when it repre-
sents the not-present as the present, the ‘that’ as the
‘
this
’
^
It may so happen that dreams sometimes turn out
to be true and tally with the subsequent experiences of wak-
ing life. But such correspondence between dream-cognitions
and waking experience is neither normal nor invariable.
Hence dream can never be called pramana, or the source
of such presentative knowledge as has a real and an
invariable correspondence with the object
The Nyaya account of dream ignores the fact that dream-
cognitions are as good presentations as our ordinary percep-
tions Dreams have not the regularity and orderliness of
wn king perceptions. But otherwise the two are indistin-
guishable. The presentative character of dreams has been
rightly noted by other systems. The Vai^esika considers

^ Rvapne tu sarvarn^va jfianam smarapamayathirtharij ca, TB , p 80

* Svapoastu anubhufapadar^hasmarapaih adrsfena dhfttiidose^a ca janyate, TAT,
3 Do^ava^cna tadjti ath^na idantityudayat, TB , p 30,
 HaI'URE ANl) FORMS OF KNOWLEDGE 27

dream to be a kind of internal perception due to the inner

sense (manas) as aided by impressions of past experiences. It

IS felt as if coming by way of the external senses.^ The

Mimariisa does indeed take dreams as reproductions of past
experiences. But it admits that they appear as presenta-
tions and are indirectly connected with the real objects of

past experiences. The Advaita Vedanta hnds in dream a

phenomenon of some philosophical significance In it there
18 the mental creation of a world under the influence of
avidya as aided by the impressions of waking experiences.
But the dream-world is. quite analogous to the world of

sense and the sciences That the world of our ordinary

expel lence may be a dream is a hypothesis that is admitted
even by Russell ' to be logically possible, though not as

simple and preferable as the common-sense belief in an ex-

ternal world. In all this we have a just recognition of the
presentative character of dream-knowledge, even though it

IS eventually condemned as false and erroneous. The Nyaya

view of dream is defective in so far as it reduces dream-
cognition to false memory. Dream-cognitions are more like

perceptions than memory-cognitions. When w'e recollect a

dream we feel ‘
such and such objects Avere seen (not merely
remembered) in dream last night.’ Bream is a kind of false
perception. It may sometimes be excited by a physical
cause aswhen a bell ringing causes us to dream of going to
school. But although sometimes started by a physical cause
dreams do not follow such causes up to their end. Breams
aie generally independent ol the impressions produced by
physical causes on our body. Even when excited by a physi-
cal cause, the series of experiences through which a dream
progresses cannot be traced to a corresponding series of physi-
cal causes Still, our dream experiences are more like per-

1 PS„ pp. 91 f

^ Pfoblemb oj Philosophy di>(i 191

28 NYAYA IHEORY OF KNOWLEDGE

ceptions than anything else. These have not indeed the

force or zwang with which the data of sense come to us.

But they seem to possess the vivacity and spontaneity of our

ordinary sense perceptions. At least, they are directly given
to us like our perceptions. Dreams are experiences which
we have, and do not arrive at by any process of reasoning.
Hence it is that they are c.illed perceptions. But they are
false perceptions becauhc they are contradicted by oui waking
experiences. To the dieamei, however, they appeal as true

perceptions, because he cannot relate them to his waking

experiences and see how they ,iie contradicted by the latter.’
According to the Nyaya, memory {,smrtt) is not valid

knowledge (piatiul) We can speak of true and lalse

memory. But even true memory, which gives us a true

cognition of some past object, cannot be called pramu or
valid knowleiige. On this point Indian systems of philo-

sophy are divided in their opinion Some of them consider

memory to be as valid as perception and inference, and
look upon it as the source of our knowledge of past facts

The Vai^esika accepts memory as valid knowdedge distin-

guished from all forms of wrong cognition.^ So too the

Jama philosophy counts memory among the forms of valid

mediate knowledge iparoksa puma) The Advaita Vedanta,

we shall see, is not definitel}'^ opposed to memory being
regarded as valid knowledge.
The other sjstems, especially the Nyaya and the
Mimaibsa, refuse to recognise memory as valid knowledge
(pranid). The Mimamsa objection against memory, as we
shall see moie fully hereafter, is that it gives no new know-
ledge (anadhigata), but is only a reproduction of some past

1 C/ R S WoofJworil), Psychology, %U ecln , pp IIS-IC

“ Or you are fully

asleep, and then the image th i it come are dre »m8 an 1 Reeiri ci'tirely real, since contact

with the objective sitmjtion h.is been lost.’

2 Cf. PS , p 9t ,
NK , pp 256-57
’ Cf. TT8.. 1, 9.J3
 NATURE AND FORMS OF KNOWLEDGE 21)

knowledge. The Nyaya, however, does not admit the

Miuiamsa contention that any knowlege becomes invalid
simply because it refers to a previously known object (grhita-
grdhUakrta). According to it, what makes memory invalid
(apranvl) is the absence of the character of presentation
(anuhhuti) in it. Memory may, in some eases, correspond
to real objects. Still it is not valid knowledge, since it

does not correspond to given objects and does not arise

out of the objects themselves {artha]anya) In memory
we have not a cognition of given objects but a re-cognition
of what were given, ui the same form and order in which
they once existed in the past and have now ceased to exist.
That form and order are now past and therefore no longer
real, so that between these and we their niemory-irnages
cannot speak of a correspondence to the given. Even when
an object is first perceived and then immediately remem-
bered, so that perception and immediate memory refer
to one and the same object and are spoken of as
equally true, we are to observe that the state of memory
boirows its validity from the antecedent peiception which
^
produces and fashions it {ydcitamandanapmya) As a
matter of fact, however, the object ceases to be given
and to be the operative cause of knowledge in memory.
The recollection of long past or remote objects is clearly

independent of the co-operation of these objects {anapeksi-

tdrtha). Memory, being thus based on no given datum
{anarthajanya)^ fails to give valid presentational knowledge
(pramd), and so, is not a source of knowledge (pramdna).^
An examination of the view that memory is not valid

knowledge is postponed at this stage. We shall come

to it after we have got all that the Nyaya has to say about

pramd and the piamdnaff.

I
'I'R and SS , pp 4.S-46

» NM , pp 20-28
30 NYAYA 'YHEORY OF KNOWLEDGE

4. Doubt {sam^aya)

Doubt (saj»6'aj/a) is the cognition of conflicting notions^

(vimarh) with regard to the same object ^
It is the
mental reference of two or more contradictory properties
to the same object. In it the mind oscillates betw'een differ-/
ent alternate characterisations of some given object.^
Doubt thus consists m an alternation between different*
conflicting notions with regard to the same object. The'
alternatives between which the mind passes in succession
in the state of doubt are called kotis. These may be two'
or more in different instances of doubt. Sometimes they
are contrary terms (e g. post and man) and sometimes
contradictory terms (e.g. post and not-post) Doubt arises
when with regard to some perceived object there is the
suggestion of such conflicting alternatives but no definite
cognition of any differentia to decide between them Hence
doubt has been spoken ot as incomplete or indecisive cog-
nition {anaiadharanatmaha) But doubt is not merely
the absence a ' ot a&sured cognition {micaya). It is

not mere negation ot knowledge. It is a positive state of

cognition of mutually exclusive chaiacteis in the same
thing and at the same time
The state of doubt may be analysed into the following
factors. There is first the presentation of some existent
object. Next by virtue of association the presented lact
calls forth two or more apperceptive systems, each of which
tries to appropriate it but is counteracted by the rest. In
the absence of any definite cognition of such differentiating
characters in the presentation as answer to any of the
apperceptive groups, mind the oscillates between them
and we have the phenomenon of doubt. Hence doubt

> NS., 1 1 23.

* fikaBoiiu dharmipi viruddhanirjaJiiarmavai^i^tyajfiftnam 8diih4aya(;, TS.,
p. 82
 NATURE AND FORMS OF KNOWLEDGE 31

supposes the recollection of the differentiating characters

of an object but no corresponding presentation of them
{vUesasmrtyapeksa) The actual process of mental oscilla-'

tion in doubt is generally expressed in the form of an inter-

rogation, e.g. ‘ Is the yonder erect figure a man or a'

post or a tree-trunk?’
Doubt is of five kinds. First, it may arise from the per-
ception of such properties as are common to many things, as
when we perceive a tall object at a distance and are not sure
if it be a man or a post or a tree-trunk, because tallness is
common to them all. 'Secondly, it arises from the cognition
of any peculiar and unique property, as when the cognition
of sound makes us doubt if it is eternal or non-eternal, since
it is not found in eternal objects like the soul and the atom,
nor in non-eternal things like water and earth. Thirdly,
it may be due to conflicting testimony, as when the different
philosophical theories of the soul leave us in doubt as to the
real nature of the soul. Fourthly, it is caused by the
irregularity of perception, as when we doubt if the perceived
water really exists or not, since there is a perception of
water both in a tank and a mirage. Lastly, doubt springs
from irregularity of non-perception, as when we are not
sure if the thing we cannot see now really exists or not,
since the existent also is not perceived under certain condi-
tions.^ According to the later Naiyayikas, such as Uddyo-
takara, Vacaspati and others, there are not five but only
three or two kinds of doubt.* Irregularity of perception or
non-perception is not by itself a cause of doubt. So also
conflict of testimony is not an independent cause, but only
an auxiliary condition of doubt. In all cases of doubt there
is either the perception of common properties or the cogni-
tion of something quite new and uncommon. Gafige^a

1 Nil ,
1 1 23
« NV , and NVT., 1 1. 23
32 nysya theory of knowledge

speaks of only two sources of doubt, namely, the suspicion

of upadhi orconditioD, and the perception of a property com-
mon to many things without any presentation of their
^
differentiating attributes
According to the Vai^esika, there is fundamentally only
one kind of doubt since it is always due to perception of
properties common to many familiar objects^ Indefinite
cognition {anudhynvasaya) is a form of knowledge which is

quite different and distinct from doubt. Indefinite cognition

’

IS incipient knowledge of an object as a mere ‘

something
without any detinitioii or determin ition of what that some-
thing IS It IS illustrated by our ordinary nascent ex-
periences, as when we say ‘something passed by without our
knowing what it was,’ or when in the presence of an unfami-
liar living being we say it must be some kind of animal.’
’’
‘

Such indefinite cognition differs fio n doubt both in origin

and essence. Doubt arises from recollection, without any
accompanying presentation, of the specific characters of two
or more objects, of which we perceive the common proper-
ties. The indefinite (anadhyavasaya), however, is a possible

form of cognition with regard to objects whose specific

character had never been presented to us. Again, the

indefinite does not, like doubt, rest on two or more conflict-

ing noti ms with regard to the same subject.^ It should

however be noted here that some of the s>ncretist writers
on tiie Nyaya-Vaisesika include conjecture (uha) and indefi-
nite cognition (anadhyavasaya) under doubt. According to

Sivaditya and Madhiva, conjecture is that form of doubt, in

which one of the conflicting suggested altern itives becomes
more probable than the other, as when seeing a tall object
in the rice field we say ‘it is probably a tall man.’ Indefi-

1 TC , 11, pp 210-11
2 PS., T>p. 8.1 f
’ NK , p. 18S
< NIj.,pp 45-46, NK
 natdre and forms op knowledge ^8

nite cognition is that kind of doubt, in which both the alter-

natives are implicitly present but neither explicitly thought

of/ Samkara Mi^ra in. his Upaskara suggests that indefi-

nite cognition arises from cognition of a peculiar property

which 18 not found in other things. Hence it corresponds
*
to the second kind of doubt mentioned by the Nyaya.
Doubt 18 not valid knowledge {prama). It may some-
times have the character of presentation (anubhava) of an

object. But it has neither the mark of being an assured

definite cognition (asamdigdha) nor that of a true corres-
pondence with the object {yathdrtha), and so, does not lead

to successful activity. In doubt the oscillation of thought

between different ideas has no objective counterpart in the
real. Nevertheless doubt is not error {viparyyaya) Doubt
as a form of cognition, is neither true nor false. It carries

with it no definite assertion of any character with regard to

Its object. It makes no claim to be a true judgment of the

object and so the question of its falsity or contradiction does
not arise. The value of doubt lies in its being a great impe
tus to study and investigation. It is the starting-point of u

critical knowledge of objects. In this sense it may be said

to be the beginning of philosophy. The critical philosophy

of Kant is doubtless indebted to the scepticism of Hume.

Tha Nyaya account of doubt, it will be seen, gives us

some important truths.As a mental state, doubt is shown

to be different from both belief and disbelief. It neither

affirms nor denies anything, but only raises a problem for

thought. As such, doubt should also be distinguished from
‘
the mere absence of belief.’ There is absence of belief

even when we do not think of anything at all. In doubt,

however, we think of two or more alternatives in regard to

the same thing. It always has, as the Nyaya-Vai^e^ika

^ tJhanadhyavasayayoFtu saihfiaya eva» Saptapddarihit 38, cf. also sec aod

Madbava*B Commentary, Mttahha^mi on it

* UpaiHra^ 3. 2 17,

5—(1117 B)
Si NiiYA THEORY OF KNOWLEDGE

would say, some Ifotis which become contradictory when

referred to the same subject This point has been rightly

noted by Bosanquet^ when he says “a definite doubt is un-

questionably a disjunctive judgment.” He observes also

that “ when a man first doubts and then decides, on such

a question as whether the river he sees before him is safe to

ford, there must be a positive basis of the two or more

alternatives as well as one suggested alternative.” What
this positive basis is or what different bases of doubt there
may be have been elaborated by the Nyaya. That the alter-

natives are exclusive and contradict each other has also been
admitted by Bosanquet. But the Nyaya seems to show
better insight when it sajs that doubt is never a definite

cognition (a ead/mrana), but an indecisive questioning attitude

towards an object.^ It is not a judgment at all It does not.
assert anything When we are in doubt about anything we
do not really know nor do we claim to know what it is.

We cannot even say that 'it must be either this or that.’

All we can say is :

‘
Is it this or that ?’ It is on account of
this that doubt is neither true nor false. For, as Bradley
says, “partial ignorance does not make any knowledge
fallacious, unless by a mistake I assert that knowledge as
unconditional and absolute.”^

5. Error {viparyyaya)

Error (bhrama) is the reverse of valid knowledge

{prarm). While valid knowledge is the presentation of an
object as what it really is (tattvanubhava), erroneous know-
ledge is the cognition of an object as what it really is not
(atattvajmna).* In error an object is cognised as having

1 Logic, Vol I, pp 35, 279, 866

* Kiiiisvidityanyatarannavadharayati, NB ,
1 1 28
3 PrmctpleB of Logic, Vol. I, p 31
* Saptapaddrthi, eec 140
 NATURE AND FORMS OP KNOWLEDGE 36

certain characteristics that really fall outside of its being.

Hence it has been described as the wrong apprehension
(mithyopalahdhi) in which an obiect is taken for wh at it is

not (atasminstaditi piatyayal^)} The cognitions of a shell

as silver, of a rope as a snake, of a post as a man are all cases
of error or wrong cognition. In each there is the cognition
of an object as other than what it really is. Hence it may
be said that error consists in attributing such characters to
an object as are not to be really found in it (tadahMvavati
tatprakdraM) In it one universal is referred not to its own
locus but to that of a different universal. In the cognition
of a shell as silver, silverness is referred to a wrong locus,

namely, the shell ^ Hence it is a false characterisation of

the object by the negation of its real characters.

Thus erior is to be distinguished from doubt. Unlike

doubt, it IS not only non- valid knowledge (aprama), but is

positively invalid or false knowledge (hhrama). An

erroneous cognition goes beyond the state of uncertainty in
doubt and carries with it a definite assertion {avadharam or
niicaya) about some presented object But, then, it is an
assertion that contradicts the real nature of its object
(vtparUanirnaya). It is a false judgment of the real
through the attribution of such characters as are excluded by
it i'Diparltadharmddhydropena) We become conscious of

error when there is a contradiction between our cognitive

and volitional experiences. Erroneous cognitions do not

lead to successful activity. The cognition of silver in a

piece of shell is found to be erroneous when it fails to lead

up to the expected results. It is contradicted and finally

sublated by the unexpected experience of failure of the acti-

vity concerned in approaching and picking it up. Actions

inspired by wrong cognitions fail to realise their ends and

thereby expose the invalidity of those cognitions,.

1 NV 1. 1
,
2
> C/TD ,p P3, TC ,
I
pp 401. 418.
 36 NYiYA THEORY OF KNOWLEDGE

6. Theories of Illusion tn Indian Philosophy

The explanation of errors of perception has been a per-

plexing question for all philosophy. The question is this :

How are we to explain the false perception of silver in a

shell ? Is it due to the object itself ? Or, is it due to our
subjective attitude towards the object ? According to the
Nyaya, while valid knowledge (pramd) is objective in the

sense of being grounded in the object itself {arthajanya). all

error is subjective in so far as it is due to the introduction of

a certain foreign character into the object by the knowing
subject (adhydropa). In the case of the mirage, for

example, there is nothing wrong in the object. “ The

object all the while remains what it actually is. In regard
to the flickering rays of the sun, when there arises the

cognition of water, there is no error in the object : it is not

that the rays are not rays, nor that the flickering is not
flickering; the error lies in the cognition : as it is the cogni-
tion which instead of appearing as the cognition of the
flickering rays, appears as the cognition of water, i.e. as

the cognition of a thing as something which it is not.”'

From this it follows that there is no error in the simple
apprehension (dlocana) of the object. The object as given in

indeterminate {mrvikalpaka) perception consists of a number

of actually present flickering rays of the sun. But on
account of certain defects in the sense organ and the influ-
ence of association and memory, the given datura is mis-
interpreted as water in the determinate {savikalpaka) per-

/ ception of it. Hence the error lies not in the indeterminate

I
perception of the given but in the determinate perception of

/ it as worked up and modified by some representative

elements.®

1 NV„ 1. 1. 4.

i NVT., 1 1. 4,
 NATURE AND FORMS OF KNOWLEDGE 37

The modern school of the Nyaya shows great ingenuity

to explain the perceptual character of illusory experience.
That in illusion there is the attribution {aropa) of a false
character to a perceived fact is no doubt true. But the
questions that arise are : How do we come to ascribe the
false character ? How again does this false character
appear as something actually perceived in illusion ? The
Nyaya rightly points out that an illusory experience is a
single perception . It is not, as Prabhakara thinks, a com-
plex of perception and recollection with theii distinction
blurred by obscuration of memory. Thus when we have
the illusion of silver in a shell, we no doubt attribute
silverness to the shell which is not its proper locus. But at

the same time it is equally doubtless that the silver is some-

how perceived and not merely remembered in illusion. This
has been very well pointed out by A. C. Ewing when he
observes: “ The difficulty in the case of perception is not
the mere fact of error, but the demand that we should hold
both that what we immediately perceive is numerically
identical with a physical object or a part of such an
object and yet that it is quite different.” ^
To explain illu-

sion, therefore, we have to explain its perceptual character,

inst. ad of trying to explain it away.

Taking the illusion of silver in a shell as an illustration,

the Nyaya account comes to this. There is first the contact

of sense with something present before it. Owing to some

defects, the sense apprehends such general features of the

thing as its brightness, etc., but fails to discern its peculiar

and distinctive features. But the general features being

associated with some other thing (here silver) recall the

memory-images of the peculiar properties of that other

thing. Through such recollection there is a sort of contact

(jHanalaksand sunnikarsa) between sense and that other

1 ilfind. April, 1930, p 149.

 38 NYAYA THEORY OP KNOWLEDGE

thing {i.e. silver). Hence there is an actual perception of

silver in the illusion.' The perceived silver is then referred

to the locus (idam) or the something which is present before

and perceived by sense. Hence in the illusion there is

perception of both the ‘

this ’
and the ‘
silver,’ although in
different ways.^ So far there seems to be nothing wrong.
The error comes in and the illusion arises when the silver
that IS perceived elsewhere is referred as a predicate to the
‘
this ’
as its subject It is this determinate knowledge of
‘
the this ’
as qualified by ‘silverness’ (uhstajnana) that
can account for a man’s efforts to gain possession of the
illusory object. In recognition (ptatyabhtjM), in which we
say ‘
this is that man I saw yesterday,’ we see how certain
presentative and representative elements combine to make
up one single perception. Any ordinary valid perception
also illustrates how a given sensum combines with associated
ideas to make up one percept. But while in these, the
combination has its objective counterpart, in illusion the

relation between the perceived ‘

this ’
and ‘
silver ’
is not
objectively real. It is contradicted and sublated either

by a subsequent experience that corrects the illusory

experience of silver and shows it to be false, or by

the experience of disappointment which ensues when we
take possession of it. In the first case the cognition
of silver is shorn of its objective (vmyapahdra), and in

the second case we are put in possession, not of the silver,

but of the shell (phaldpahara) Hence the error lies not in

the presentations concerned in the perception but in the

determination of one presentation by another given through
association and memory {jdtyammskdrdt). And since this

determination results in a judgment of the object as some-

' C/.Woodworth, Psychology, p 110: “Memory im)ge8, then, are recalled

senfitiions, or have more or less of the quality of sensitions

^ Cf MamatvindnyajaDyatvat ]atyasaihskaracca saksatkaritvamevobhayatra,
TC.i.p. 525.
 NATURE AND FORMS OF KNO-WLEDGE 39

thing other than what it is, the Nyaya theory of error is

called anyathakhvdti or vimntakhydti. According to it, an
erroneous cognition is presentational in character and has
some basis in facts But the facts being misplaced and
misrelated, err^ beconaes a false apprehension of the

The above view of anyathakhyati is common to the

Nyaya-Vai^esika. It has been accepted in the main by
Kumarila, Ramanuja ^
and the Jainas. But the
Bauddha, the Prabhakara Mimamsa and Advaita Vedanta
systems oppose the Nyaya view and propose different theo-
ries. Hence the Naiyayikas proceed to repudiate the other
theories of error. According to the Yogacaras, there is nc
extramental reality, and things are only thoughts or ideas.
Reality is a stream of cognitions bifurcated into a subjective
and an objective series. Error consists in an illegitimate
process of projection of subjective ideas as objective and
extra-mental facts. All cognition of objects thus objectifies

the subjective and is therefore erroneous. This view is

called dtmakhyat i or jfidndkarakhyati, since it insists on the

sole reality of ideas and looks upon all objects as cognitions
wrongly taken for external things.

This theory, however, the Naiyayikas object, fails to

account for the facts of the case. On the theory of subjec-

tive idealism of the Yogacaras, there is no difference
between knowledge, and the subject and object of know-

^ Ramanuja bas proposed an alternative theory of illusion which is distinjfuished

from the above as ^atkhvatt. According to it, all cognitions are relatively true and

none absolutely false. The cognition of silver in a shell is true with reference to the

element of silver that is present in the shell la every object of the world the ekpjeilJfeS

of all other objects are present in different proportions So in the structure of a shell

an element of silver is present, although the shell element preponderates in it. Hence

the cognition of silver bas an objective basis, and is so far true But owing to certain

defects of the sense organs, there is a distortion of the shell element, and we have the

what is really a silver-shell The perception is wrong, not

perception of silver in
because a partial view or
the cognition of no fact or of the unreal, but
it is
because it js

an imperfect knowledge of the real. (C/. Srlhha^ya^ 1.1 1

40 NtlYA THBORY OB* KNOVVLEDGR

ledge, everything being an idea only. Hence the cognition

of silver should appear, not in the form of ‘
this is silver ’

but ‘I am silver’, which however is not the case. Then, if

everything be an idea we do not know how to account for

the difference between an idea and its corresponding percept.

Finally, the Nyaya view, that error is the cognition of an
object as what it is not, really includes the Yogacara theory
that in error the subjective is taken for the objective and is

so cognised as what it is not.^

The Madhyamika school of Bauddha philosophy negates

all existence. It holds the asatkhmU view that error con-
sists in the* manifes tation of the non-existent as existen t.
The cognition of silver in the shell is erroneous because it

manifests the non-existent silver as existent, and we become

conscious of this when our first cognition of silver is contra-
dicted by the subsequent cognition of shell. Against this it

has been urged by the Naiyayika that the illusion of silver

13 not entirely baseless, it cannot arise out of nothing.

What is absolutely non-existent can not produce even the
wrong cognition of silver The illusion of silver is due to
something in the nature of the shell. It occurs generally in

connection with a shell and the like, but not indifferently

with everything. Even if error is a cognition of the non-

existent as existent, it is the cognition of it as what it is not.

Hence we have in it a case of anyathakhyati which thus

includes the asatkhyati of the Madhyamika. In truth, how-
ever, the utterly unreal and non-existent cannot be the object
of any knowledge whatsoever.^
The Advaita Vedanta puts forward the view of anirmca-
nivakhvdti . This does not differ so widely from the Nyaya
anyathakhyati as may appear While the two
at first sight.

views agree so far as the nature and mechanism of illusory

1 NVT.pp 86 f , NM.pp 176,646-16

* Ibti
 NATURE AND FORMS OF KNOWLEDGE 41

perception are concerned, there is difference m one essential

point. According to amrmcanlyakhyaU, there is in the
illusion of silver a contact of the defective sense organ with
the glittering shell and then a mental modification answer-
ing to the form of ‘
this ohject.’ Now through the operation
of nescience (avidyd), as aided by the past impressions of
silver, revived by this object’s similarity, tlieie is the pro-
duction of some inexplicable silver which lasts so long as the
illusion lasts. It is nei ther real nor anreal, nor both real

indjyijjxcRlfc^buiiiKlefijiable and indeterminable Hence in

the illusion of silver there is an actual cognition of silver.

The illusion is preseiitative in its character so tar as it is

connected with some silver actually piesent to consciousness.

But while, according lo the Kya>a, this presentation ot

silver IS due to association and mernoiv (jdlijasamskdrat) to

the Advaitin, it is due to the production of the ‘

cognised
silver ’
for the time being. To this the Naiyayikas object
that if the silver is actually produced, there would be no
illusion but a v .^hd perception. If it be said that the silver

IS supernatural (alaiihika) and is erroneously cognised as

natural (Jauliha), we have in it just a case of anyathdkh-

ydh. If, on the othei hand, the supernatural silver is cog-
nised as supernatural, there cannot be anything wrong in

the cognition, nor any piactical activity in the cogniser

to obtain such supernatural silver.^
The Prabhakara school of the Mimamsa differs from all

others and advocates the view of akhitdti or vivekdkhydti.

According to it, error consists simply in the want of

discrimination between percept and image, or between

direct apprehension and memory. It is a sort of con-

fused memory (smrtipramosci). In the case of the

illusion of silver in the shell what happens is that there

is first the direct perception of an object with the

1 NVT , pp 88-87 , NM , pp 187 f.

6— (1117 B)
 42 NYiYA TflBORY OF KNOWLEDGE

attribute of brightness, etc. Then through association

by similarity the perceived bright object revives the image
of silver. Hence the state of cognition has the dua l
character of percent and image , of something seen and
something remembered. On account of certain abnormal
conditions, the two things are not kept distinct and are
allowed to fuse or coalesce, and we have the resulting
cognition of silver referred to the piece of shell. When the
illusion is corrected, there is no sublation of the silver but

only an explicit recognition of the presentative and represen-

tative factors of the wiong cognition. The distinction

between the two being cognised, there remains no confusion

as to the fact of silver being only remembered and not
'
perceived

To this theory the Naiyayika objects that it fails to

account for the presentative character of the illusion. So

long as we are under the illusion we have a consciousness

of the silver as something present and perceived, and not as

what was perceived before and is now only remembered.
Further, there can be no activity to secure the silver unless
there is a positive and determinate cognition of it. A con-
fused knowledge cannot inspire the confidence necessary for
practical activity. Non-discnmination, as mere confusion
of knowledge, cannot be the ground of such actions as are
generally connected with an illusory experience. Hence]
illusion must be a single determinate cognition of an object.
All this comes out in our subsequent judgment of the illusion
as it stands corrected and negated. It is in the form “ what
I had seen is not silver,” and not “ what I had remembered
is not silver.” This clearly shows that the illusion of
silver is an error of perception and not of memory. Finally,
even on the akhyati theory it must be admitted that in error

' Vtd$ Jba, Prahhakara School of PurvamlmathsSt pp. 28-32 , Sadholal Lectures on
Ny&ya^ Ch, IIL
 NATURE AND FORMS OF KNOWLEDGE 43

there is a cognition of the object as what it is not. This

willmean that akhyati is but a form of anyathdkhyati.
Hence the Naiyayikas conclude that anyathdkhyati is the
most satisfactory theory of error. It gives us all that the
other theories require, but is not vitiated by their faults.^
Among the theories of perceptual error or illusion as

explained above that of the Nyaya seems to be more accept-

able than any other. For the Bauddha idealist error consists

in the objective appearance of subjective ideas. But this

cannot explain the distinction between true and false

perception. In both,- the object of knowledge is not realty

other than knowledge or an idea, although it may appear to
be so. Hence both must be equally wrong. Further,

there being nothing but ideas, one idea may be mistaken

for another, but not for that which is no idea at all, i.e.

for extra-mental object. Perceptual errors cannot, there-

fore, be explained on the theory of subjective idealism of

“For,” as Ewing says, “ even in
the Yogacara type.
error we are concerned not with our ideas but with external

reality, .. .error is not a mere dwelling on our ideas but an

^
unsuccessful cognising of objects.”
The Prabhakara Mimamsakas treat error as the subjec-

tive appearance of an object. In it an idea or image of the

mind is referred to a given object so as to become one with

it. This explanation of perceptual error was once generally
accepted in European philosophy. According to most of

the Western systems, in illusory perception a real object is

modified by subjective factors supplied by the mind through

association, memory, emotion, etc. Among modern
Eossky holds that *
falsity is the subjective
thinkers,
appearance of the object since foreign elements can be intro-
duced into the object only by the knowing
subject.’®

1 NVT., * NM., tbtd.

2 Mind, April, 193i>, pp. 138-39.

Knowledge, pp 227 f., 267 f.

3 The IntmUve Basts of
44 NYiYA THEORY OF KNOWLEDGE

Bosanquet only puts the matter in objective terms when he

says that ‘
in falsehood something actual pretends to be
something else, or, like a false com, has not the significance
which it claims.’ ' But while this may be accepted as a
general definition of error, it does not explain how in

perceptual error certain subjective factors, ideas or images,

are actually felt as perceived, out there in space. It cannot
be said that we do not really perceive the illusory object
but only imagine that we do perceive it. Why should we
doubt the verdict of experience here if we do not doubt it
elsewhere? And experience clearly tells us that we do
perceive the illusory object.
To explain the perceptual character of illusory experience,
the Advaita Vedanta supposes the temporary production of
certain positive entities ol an indeterminate order
(amrvacanlyd pidhbhdsihi satta). These are neither real nor
unreal, but actual facts. The illusory silver is not real,

because it is contradicted by a closer experience. It is not

unreal, because it is perceived as an actual fact so long as
the illusion lasts. Hence it is an appearance which is

undeterminable as real or unreal, and is ultimately due to

avidyd or ignorance. But how can produce a

positive entity and then make us perceive it as an existent

fact IS, left unexplained. Farther, as Alexander has pointed

out, ‘error does not give us a new and more shadowy being
than the spatio-temporal reality, but is the world of
’
determinate being misread ^
Moieover, if in illusion

certain positive entities are really produced and perceived,

^ -

we should not be having an illusion, but a true perception.

According to the Nya}a, illusion is a misplaced fact.

All the factors of an illusory pciception arc real and perceived

facts, but they are brought into a wrong relation. In the

1 Logjc, Vol I, pp 67, 70

2 Alexander, Space, Time and Deity Vol I, p 202
 NATURE AND FORMS OF KNOWLEDGE 45

illuaory perception of silver, for example, we have an actual

perception of a certain locus and some ^ilver, both as real

facts. The error arises because the silver perceived at a

different time and place, is related to the time and place

occupied by the perceived locus, namely, the shell. Some
modern realists give a similar explanation of perceptual

error. According to Alexander, ‘

error consists in wrongly
combining the elements of reality Everything which is

illusory in the illusion does actually exist in correspondence

with the mental activity through which it is revealed ;
but
the personal character of the activity dislocates the real

object from its place in things, and refers it to a context to

winch it does nut belong. So when 1 fancy a horse’s body,

and complete it with a man’s head, the head exists in
reality, but not upon a horse’s body.’ ’
To quote Alexander s

own words “Some of our objects are illusory ;

they are

real so far as they are perspectives of space-time, but they

contain an element introduced by our personality, and do

”
not belong where they seem to belong ^ This however
does not explain how illusory objects can be perceived at the

time and place, to which they do not belong. How can we

perceive here and now something which exists elsewhere ?
W P. Montague tries to explain this by some distortion
of the real object in producing its effect on the brain. He
thinks that the so-called sensory illusions result from

certain physical or peripherally physiological distortions of

the real object underlying them.* This means that illusory

perceptions depend on certain objective and real conditions

in the same waywhich true perceptions are so conditioned.

in

E. B. Holt goes further than this and establishes the

^He thinks that all eirois are cases of
objectivity’ of error.

contradiction or contrariety The perception of silver is

1 Essays m Critical Realism, pp

^ Space, Time and Deify, Vol II, p 249
3 The New Realism, ipp 288-92
46 KYiYA THEORY OP KNOWLEDGE

illusory because it is contradicted by the experience of the

same object as shell and not silver. But neither the
experience of shell nor that of silver is subjective. Both of
the contradictory experiences are objective, since the real
object itself has contradictory characters. Holt says that
“ the case of hallucinations is paralleled by such cases as
that of mirrored space, wherein sundry mirrored objects occupy
the same spatial positions as are occupied by other ‘
real

objects situated behind the mirror.” Hence we are to say

that error consists in entertaining mutually contradictory
propositions, of which one may be preferable, but none
subjective, because the world is full of such contradictory

propositions.' On this view, however, the distinction

between truth and error becomes insignificant. The same
thing may, with equal truth, be called a shell or silver
The Nyaya does not go so far as to say that contradictory
characters belong to the same thing or that contradictory
propositions are equally objective. It is not the case that
the same real has the contrary characters of shellness and
silverness. It has really one character, namely, shellness.

But the silver is also a perceived fact. Hence the crucial

question is : How can the silver, which exists elsewhere,

be presented here and now

The N>aya explains this by
?

jMnalaksandpratyasatti which means a kind of sense-object

contact brought about by the impressional revival of the

past experience of an object. Hence there is a pidnalaksana

perception of the silver. As we shall see more fully here-

after, the perception of the silver is a case of what is called

“ complication ” by some Western psychologists. In it the
sensation of a particular bright colour calls up, by its

previous association, the impression of silver and w'e have

the perception of silver in the shell. The silver does not

appear as an idea or image of the mind, but is a content

* Ibid , pp 369-70.
 NATURE AND FORMS OF KNOWLEDGE 47

presented by the sensation of bright colour. Still, the

perception of the silver is illusory because the character of
silverness does not really co-exist with the given sensation
of bright colour. It is the presentation ot silver in a wrong
relation and so an illusion.

7. Hypothetical Argument (tarha)

Tarka is a type of implicative argument by which we

may test the validity of the conclusion of any reasoning
(or of any judgment) Heie wc ask whether any contradic-
tions would follow if the given conclusion be accepted as
true or rejected as false. If there is any contradiction in
accepting it as true we have no doubt that it is invalid.
But if in rejecting it as false we arc involved m a contradic-
tion, there can be no doubt that the conclusion must have
been valid.
The process of reasoning m tarka consists in the deduc-
tion of an untenable proposition from a certain position
(amstaprasanga) This has the logical effect of exposing
the invalidity of that position and thereby lending support to
the counter-position. Thus with regard to the inference of

fire from the perception of smoke, there are two alternative

positions, namely, that the smoky object is fiery, and that
it IS not fiery From the latter position we deduce the

proposition that the object is not smoky, which is contra-

dicted by our direct experience. This is expressed in the

form of a hypothetical proposition, viz. ‘
if the object be

fireless, it must be smokeless.’ Hence tarka validates the

inference of fire through the deduction of an inadmissible

proposition from the contrary hypothesis. The proposition

IS a deduction from the hypothesis in the sense that it follows

from it according to a general rule. It is a general rule

that whatever has a mark (t he vnama) . has that which it.

18 a mark of (the vyapaka). Now the absence of fire is a

48 NYiYA THEORY OF KNOWLEDGE

mark of the absence of smoke. Hence if it be said that

there is absence of fire in the object, we cannot resist the
conclusion that there is absence of smoke in it, i.e. it is

smokeless. Such a conclusion, however, is contradicted by

direct observation. Hence it is that tarka has been defined
by the modern Naiyayikus as the process of deducing from’

a mark that of which it is a mark, but is false (vyapydngJ-

^
kdrena amstapydpahapMsaniamiiipah)
When the proposition established by any method of

knowledge (piamana) is doubted or disputed, we should

have recourse to taika to la} the doubt or end the dispute.
In ta)ka we take the contradictory of the proposition in the

form of a hypothesis and leadily see how that hypothesis

leads to a contradiction Hence tai/ra selves as the limit

to doubt {^ai)kdi n Ihi) Since, however, the invalidity of a

position IS not a ground of the validity of its opposite,

tarka is an aid or auxiliary (sahakdrl) to pranidna, but not

pramdna by Thus when on seeing
itself. a table we say :

‘
there is no book on the table,’ we have a judgment of

perception expressed in a proposition If anyone doubts the

truth of this proposition we may effectively dispel it by an
argument like this ‘
If there were any book on the table,

it would have been perceived like the table ; but it is not

so perceived ; therefore it does not exist.’ But to argue in

this way is not to know the non-existence of a book on the

table. The knowledge of the book’s non-existence is a

matter of perception according to the Naiyayikas. Similarly,
to argue that ‘
if the object be fireless it must be smokeless,’
is not to know The knowledge that the
that it is fiery.

smoky object is fiery is acquired by means of inference

from smoke as a mark of fire. The hypothetical argumenl
only confirms this inference. Hence tarka does not ori-
ginate true knowledge, i.e. is not a pramdna, although if

1 TB.,p.&9.
 NATUllE and FOEMS OF KNOWLEDGE 49

contirms a pramana which brings about the knowledge in

question, i.e. is auxiliary to the pramana (pramdndnugra-
haka)}
There are five kinds of tarka . These are called
dtmdSraya, anyonydhaya, cakraka, anavasthd and tadanya-
hddhitarthaprasanga} In all of them the logical form and
character of the argument is the same, and they serve the
same end of testing the validity of some reasoning or
judgment.
Atmdhaya is an argument that brings out the incon-
sistency involved in - a reasoning which seeks to prove that
anything is dependent on itself in respect of its origin or
duration or cognition The argument may be stated in
this form ‘
If A must be different
is the cause of A, it

from itself, because the cause is different from the effect.’

Anyonydsraya is an argument whuih brings out the
contradiction involved in the judgment that two things
are reciprocally dependent on each other. The argument
may be stated thus :
‘
If A depends on B, and B depends
on A, A cannot depend on B.’ To say that ‘
B depends on
A ’
IS virtually to deny that ‘
A depends on B.’ The idea
of reciprocal dependence, which is so much favoured by
some Western thinkers, is rejected by the ancient Indian

thinkers as self-contradictory and absurd.

The third type of tarka is called cakraka. It consists

in exposing the fallacy of a reasoning in which a thing is

made to explain the pre-supposition of its own pre-supposi-

tion {tadapeksyapeksyapeksxtva). If A is pre-supposed in

B and pre-supposed in C, then to explain A by C is
B is

to reason in a circle, because C by its inherent limitations

leads us back to A. Starting from A we are referred to C
as the ground of its explanation, but to explain C we are

» TB p 32.
,

2 NSV 1. 1.,
40.

7— (1117 B)
So NYiYA THEORY OF KNOWLEtoGB

brought back to A as its ultimate ground or basis. Here the

curve of explanation makes a complete circle in so far as

our thought returns to its own starting-point through two

or more intermediaries. Thus if we admit that perception
is pre-supposed in inference and the latter is pre-supposed
in testimony, then to prove perception by testimony is to
commit the fallacy of circular reasoning. This may well
be exposed by a tarka like the following :
‘
If preception
depends on testimony, it must be independent of sense-
’
object contact
The fourth type of tarka is called anavastha. It is

an argument which brings out the absurdity of an indis-

criminate extension of the fallacy of undue assumption.
Here we expose the fallacy involved in the indefinite regress

of thought from point to point without any final resting

ground {aiyaiasthitapaiarnfamroya) There is an infinite

regress of thought (anavastha) when in an explanation we

make use of an indefinite number of principles, each of

which pre-supposes its Here our thought moves not

next.
in a circle, but up a staircase, as it were. Thus if we
explain A by B, B by C, C by D, and so on ad infinitum,
we do not really explain anything. Or, if we try to

deduce the ground of inference from inference we arc

logically committed to the fallacy of infinite regress. The
fallacy may be exposed by a tarka like this ‘
If inference

depends on inference for its ground, no inference is pos-

sible.’

The last type of tarka is called tadanyahdd}iitdrtha~

prasanga or pramdnahddhitarthaprasanga. It is an argu-
ment which indirectly proves the validity of a reasoning by
showing that the contradictory of its conclusion is absurd.

This may be done by opposing the contradictory of the

conclusion to some fact or some universal law. If, there-
fore, its contradictory be false, the original conclusion must
be true and based on a valid reasoning. Take, for example,
 NATURE AND FORMS OP KNOWLEDGE 51

the inference :
‘
Whatever is smoky is fiery ; this object
IS smoky ; therefore this object is fiery,’ If this con-
clusion be false, then its contradictory, ‘
this object is
not fiery ’
should be true. But the latter proposition
is found to be absurd by the following tarka.
‘
If
in the case of this object smoke is not related to
fire, then it cannot be an effect of fire. But it

must be due either to fire or to not-fire There is no third

alternative here. We do not find it to arise out of not-fire

Hence if it is not due to fire, it must be either an uncaused

effect or a non-existent phenomenon The first alternative

contradicts the law of universal causation and is therefore

untenable. The second alternative becomes self-contradic-
tory, since it commits us to the proposition that the smoky
object IS smokeless For, if A (smoke) be a mark of B
(not-fire), and B (not-fire) w'ere a mark of C (not-smoke),
then A (smoke) would be a mark of 0 (not-sraoke) In
view of such absurdities involved in the contradictory of the
original conclusion we must reject it as false and accept the
original conclusion as true and as based on a valid in-

ference.
It IS to be observed, however, that the Nyaya division

of tarka into five different kinds is logically unsound This

division has reference to the different kinds of reasoning

which may be tested by an argument like tarka. But

the classification of tailia should not be based on the kinds

of reasoning that may be tested by it, because these are

unlimited and quite external to the nature of tarka as a

type of argument. A classification of tarka must be based

on the logical character of the arguments employed in
different cases. Now having regard to its logical character,

we find that tarka is fundamentally of one kind. In every

case in which it is employed it has the form of an incon-

argument (amstaprasanga) developed out of the

sistent
reasoning or its contradictory. If
conclusion of a given
52 NYiYA THEORY OF KNOWLEDGE

this inconsistent argument arises out of the conclusion of

the given reasoning, we are convinced that the given rea-
soning is invalid. If it arises out of the contradictory of
the conclusion of a given reasoning, we know for certain
that the original conclusion and the given reasoning are
valid. As to its logical character, therefore, tarka seems
to correspond to the antilogisni in Western logic. Accord-
ing to some Western logicians,^ the antilogism is an incon-
sistent triad of propositions by which the validity of any
syllogism may be determined. A syllogism is proved to be
valid if by combining the contradictory of its conclusion
with the original premises we get an inconsistent triad.
If, however, the resulting triad is consistent, the original
syllogism IS invalid. It is also admitted by these Western
logicians that the inconsistent triad, like the Nai}ayika’s
tarka, is not itself an argument. It should however be
remarked here that the logical form of the argument in
tarka does not exactly correspond to that of the antilogism.
Tarka is put into the form of an implicative argument,
while the antilogism into that of a categorical syllogism.
Again, tarka may be employed to test the validity of any
reasoning, inferential or otherwise, and it may be developed
out of a given conclusion or its contradictory with or
without the original premises.

1 Cf Chapman and Henle, The Fmdamentals of Logic, 90 and 102.

pp
 CHAPTEE III

VALID KNOWLEDGE AND ITS METHOD

(prama and PRAMANA)

1 Definition of prama or valid knoioledge

In Chapter 11 we have considered the different forms of

non-valid knowledge (apramd) Here ue are to consider
the nature of valid knowledge {prama) and the general
character of the method of valid knowledge (pramana)
It may appear to some that the distinction between valid
and non-valid or invalid knowledge is not only unnecessary
but incorrect. Knowledge, in its strict sense, means a
true belief that carries with it an assurance of its truth.
Hence knowledge is always true. It is a tautology to speak
of ‘
valid knowledge ’
and a contradiction to speak of ‘
non-
valid or invalid knowledge.’ The latter is no knowledge
at all, since it does not stand for any belief which is true
and which gives us an assurance of
its truth. When we
speak of prama as valid knowledge, we do not forget the
strict sense of the word ‘
knowledge.’ But the word know- ‘

ledge ’
has been used in a narrow as well as a wide sense.
Hence in view of the facts that the Nyaya-Vai^esikas use
jMnam make a distinction
in a very wide sense, that they

between true and false pldnam, and that prama implies

something more than knowledge in its strict sense, we
propose to use the phrase ‘
valid knowledge ’
for prama.
’

As however we have already said, the word ‘

knowledge
may be taken to mean prama according to the context.

1
Cf, Russell, Problems of Philosophy, p. 217.
64 NYIYA THEORY OP KNOWLEDGE

Pramd has been defined by the Nyaya as true presen-

tational knowledge (yatharthanuhhava). It is a definite

and an assured {asamdtgdha) cognition of an object, which

is also true and presentational in character. Hence it is

that pramd excludes all kinds of non-valid knowledge, such

as memory, doubt, error, hypothetical argument (tarka),

etc Memory is excluded because it is not presentational

{anuhhava). Doubt and the rest are excluded either because
they are not true or because they are not definite and assured
cognitions It appears from this that pramd has three
main characteristics, namely, assuredness , truth an d presen-

tat i veness .

As to the first, we may explain it by saying that pramd

or valid knowledge is a definite categorical assertion as
distinguished from all indefinite, problematic and hypothe-
tical knowledge. In pramd there is a feeling of assurance
or conviction in what is known. That is, valid knowledge
IS always connected with a firm belief. All assurances or firm

beliefs, however, are not pramd In illusion (bhrama) we

firmly believe in what is false. Pramd implies something
more than a subjective certainty.
Hence the second characteristic of pramd is that it is

true or unerring (yathdrtha) knowledge. But what makes

knowledge true {yathdrtha)? In answer to this we are told

that knowledge is true when it is not contradicted by its

object (arthavyabhicdrl) This means that knowledge is

true when it reveals its object with that nature and

attribute which abide in it despite all changes of time, place
and other conditions ^
What is once true of an object is

always true of it, no matter what its position in space and

time may be. More definitely speaking, to know a thing
truly is to know it as characterised by what is a character-
istic of it (tadvati tatprakdraka). We have the truth about

> NVT.pp. 6,21.

 Valid knowledge and ifs method 55

a thing when we judge it to be such-and-such, and it is

such-and-such, t.e. as we determine it-
by qualities which
the thing does in fact possess.^ Hence, according to the
Nyaya, the truth of knowledge consists in its correspon-
dence to facts.

So far pranid may be said to mean the same thing as

knowledge in its narrow sense. Like the latter, it is a
true belief which is connected with an assurance or convic-
tion of its truth But the Nyaya goes further and adds a
third qualification to pmmd. According to it, pramd is not
only a true and an as.sured cognition, but also a presenta-
tional cognition ((inuhharu) Otherwise, memory will have
to be regarded as pramd. Memory-knowledge is both true
and definitely believed to be true. Still it is not pramd,
since it is not presentative but representative cognition.
What then is anvhhava ? To say, as some Naiyayikas
have said, that anubhava is knowledge other than
memory is just to beg the question. But the matter has
not been left there. We are told by others that anubhava
IS knowledge of given facts as distinguished from those
that are imagined or supplied by the mind.* Or, it may be
said that anubhava is knowledge which is grounded m and
due to the object itself (arthajanya) Or again, it may be

said that anubhava is a cognition that follows uniformly

and immediately on the presence of its special cause. This
means that a cognition is presentational if it is not separated

from the existence of its unique cause by any interval of

time As such, memory cannot be called anubhaba,

because its object is not a given fact, or because it is not

1 Cj ImBUj, The IniuiUve Basts of Knowledge, p. 227 "We have acquired

truth only when the differentiated appearance is con posed entirely of elements present

in the ob.eot itself and nothing has been introduced into it fiom without.”
2 Tattvamanaropitam rupain, tasya jnanamanubhavah, SaptapadartM,

sec 64,
66 NJTiYA THEORY OF KNOWLEDGE

due to any influence of the object, or because the impressions

{saihskdra), out of which it arises, are not immediately
antecedent to it in every case. On the other hand,
perception, inference, comparison (upamdna) and testimony
are all cases of anubhava or presentational knowledge.
That sense-perception is so, will be generally admitted.
But inference and the rest also are, according to the Nyaya,
presentational cognitions. Even prdtibha or intuitive
knowledge of future events is regarded as a perception due
to the object itself. The Nyaya holds that each of these
is a cognition of some objective facts and is conditioned by
those facts. Memory being only a reproduction of past
experience cannot be said to be due to its object and
is, therefore, other than presentational knowledge
(anubhava).^
Hence the Nyaya definition of pramd or valid knowledge
comes to this. Pramd is a presentational cognition
(anubhava), in which there are a characterisation, in thought,
of the object as it is in reality (yathdrtha), and a
definite assurance of its being objectively valid
(asamdtgdha )

2. Definition of pramdna or the method of knowledge

Pramdna derivatively means the instrument of valid

knowledge (pramdydh karanam). Hence, generally
speaking, we may say that pramdna is the means or
source of right knowledge. It is that which gives
us valid knowledge, and only valid knowledge of
objects. So it has been said ;
“ There cannot be
any right understanding of things except by means
of pramdm. A subject arrives at the valid know-
ledge of objects by means of pramdna, for the existence
and nature of objects are to be ascertained only by such

1 NM.,p 28; TR & 88., pp. 9-11

 VALID KNOWLEDGE AND ITS METHOD 67 .

cognitions as are based on pramana-” Again, we are told :

“ Pramana is the cause of valid cognition of objects, inas-

much as it gives us a knowledge of objects as they really
are and exist in themselves “ Pramana has a real
correspondence with objects, in the sense that the nature
and attributes of objects, as revealed by pramana, are
uncontradictorily true of them, despite all variations in
*
time, place and other conditions,”
So far we are given to understand, not what a pramana
exactly is, but what the general character of pramana must
be. We do not go beyond such general description of

pramana when we are told by others that “ pramana is that

which IS invariably related to prama,” or, ” to be pramana
is never to be disconnected from a knower possessing right
knowledge ” All this means only that pramana is the
karana or means of prama or valid knowledge. What then
is a karam and how is it constituted ? In order to answer
the first part of this question we should follow the distinction

between karana and karana, means and cause.

A cause has been defined as the invariable and uncondi-'
tional antecedent of an effect {ananyathasiddhamyatapurva-

hhavl). Conversely, an effect is the invariable and uncondi-

tional consequent.^ Or, an effect is what begins to be and

thereby negates its antecedent non-existence. There are

the constituent (samavdyi),

three kinds of causes, namely,
{nimitta).
the non-constituent (asamavdyi) and the efficient
The constituent causes is the substratum in which the effect

inheres, e.g. the threads of the cloth. The non-constituent

It determines the
cause is the mediate cause of an effect.

» NB ,
l.l 1 ,
4 2 29
* NVT .tbtd
Kv^uman, ah, i-5
3 S<»rf.ad<»r^ana«amgrohtt,ChapteronNy5yaPhilo8opIiy .

cause and effect is a one-one relation,

* This implies that the relation between
there being only one cause for one effect
and one effect for one ca^ It

common sense and ordinary text-books

by
the Idea of a plurality of causes as endorsed
of logic

8
58 NyiYA THEORY OF KNOWLEDGE

effect only in so far as it stands as an inherent attribute

of the constituent cause. Its causal efficiency therefore is

mediated through its intimate relation to the material or

constituent cause. In relation to the effect ‘
cloth,’ the con-

tact of the threads is the non-constituent cause. So also the

colour of the threads is the mediate cause of the colour of
cloth. The efficient cause is different from both the consti-
tuent and non-constituent causes It is not merely the
lassive substratum in which the effect inheres, nor any
nherent attribute of the substratum that indirectly

letermines the effect Rather, it is the agency that acts on

)oth the constituent and non-constituent causes and makes
hem produce the effect. In relation to the cloth, the loom
ind such other agents constitute the efficient cause. It is

ihe efficient cause that is to be regarded as larana or means,

lecause it is principally concerned in bringing about the
jffect. While the first two are general causes or rather
jonditions of the effect, the last is the actually operative
lause of it It is the special cause, or simply, the cause ot
^be jeffect.^

Now reverting to the definition of pramana, we may say

that it IS the specific cause of valid knowledge as distin-

guished from its general causes or universal conditions.

Pramana is the unique operative cause (karana) of right

knowledge (pramd). It does not, however, follow from this

that pramana is a simple concept denoting a single thing,

On we are told that it denotes a complex ol
the other hand,
many conditions which are partly physical and partly psy-
chical or mental in nature In fact, any instance of know-
ledge involves a long and complicated process which if

either physical and physiological or mental or both. Th(

perception of a jar, for example, is conditioned by physica
contact between the eyes and the object as well as by inter-

1 C/ TB., pp 2 f and T8., pp. 36 f.

 Valid knowledge and its method 69

nal operations of the visual or'^an, its contact with manas

or the mind, and that of the latter with the soul. In inferen*
tial and verbal knowledge there are such specific psychic
conditions as the knowledge of a universal relation and
understanding of the meaning of a proposition and so on.
Hence pramdna is taken to mean the entire complex oi

collocation of all the specific physical and psychical

conditions {bodhdbodhasva hhdod sdmag n) that are actually

operative m bringing about valid and assured cognition of

objects (pramd). This, however, does not include such uni-
versal conditions of all knowledge as subject and object,

time and space, etc., within the compass of pramdna or the

method of knowledge Hence the final definition of pramdna
is that it IS the complex of specific conditions, other than
the subject and the object, which does not normally fail
^
to produce valid knowledge
The Vaiisesika system defines pramana as the unique
operative cause {karana) of both true presentational know-
ledge and memory ^ It would take memory as a distinct
pramdna or method of knowledge like perception and infer
ence. The Nyaya restriction of pramdna to the ground o
presentational knowledge has been set aside and memory
has been rightly shown to be an independent method o.

knowledge by the Vaisesikas.’

The Jainas also take pramdna in a general sense so as
to make it applicable to both immediate presentational
knowledge (prat^afesa) and mediate knowledge (paroksa) so

far as they are true. Under mediate knowledge they include

sense-perception, inference, memory and recognition. In

this general sense, pramdna is knowledge that reveals both

itself and its object in a way that is not liable to contradic-

tion.

1 C/ NM , p
16
2 Smftyanubhavaaftdhara^am pmnakaranam praraa^am, TK , p* 6*

3 Vide tnfra, Bk V, Ch. XX, Sec. 4.

60 NYIYA theory OF KNOWLEDGE

The Advaita Vedanta defines pramdna as the operative

cause (karana) of pramd or true knowledge. It defines

pramd in two ways. First, pramd means knowledge that

has both the characteristics of novelty and uncontradicted-
ness {anadhigatahddhita) This means that true knowledge
is uncontradicted and original, i.e. gives us new inform-
ation. Secondly, pramd is taken to mean simply uncontra-
dicted knowledge of objects The result is that pramd is
made to exclude or include memory according as we accept
the one or the other way of defining pramd or true
knowledge,’

3. Nydya criticism of the Bauddha mens of pramdna

It has been generally admitted by all the schools of

Indian philosophy that pramdna is what gives pramd and
that pramd is true knowledge. But there is much difference

of opinion among them as to the nature of the truth, which

each of them claims for its pramdna.
The Buddhists generally take the truth of knowledge to
consist in its capacity to produce successful activity Pramd
or true knowledge (samyagjndna) is harmonious in the
sense that there is no conflict between the cognition of an
object and the practical activity to obtain it. In fact all
knowledge is meant for some action. We seek knowledge
because we want to act effectively in relation to other things.
Hence pramdna or the method of knowledge fulfils its
function when it shows an object in such a way as to enable
us to act successfully in relation to it. In short, pramd is
practically useful knowledge, and pramdna is the source of
such knowledge.®
To this the Nyaya objects that practical utility (artha-
siddhi) does not constitute the truth of any knowledge.

1 VP Ch,
I
8 NBT.,Ch. I.
 VALID knowledge AND ITS METHOD 61

If it did, the distinction between true and false knowledge

would be hard to maintain. The Buddhists suppose that a
knowledge is true when it is such presentation (pradariaka)
of an object as leads to the actual attainment (prdpaka) of
it. On this view, all inferences become invalid in so far
as the object of inference is not actually presented to senses.
Perception may be said to present an object, but it cannot
lead to the attainment of the presented object If the object
be a sensum, it must be fleeting on the Bauddha theory of
momentariness (fesam/mmda) . It cannot be made to wait
and persist m existence, so that our subsequent activity
may put us in possession of it. If however the object be
an ideatum ora cogitable entity (adhyavaseya), then for the

Buddhist, it is no reality but a fiction {kalpand), and we

cannot speak of any attainment of it. b’urther, if with the
Buddhist we accept the pragmatist conception of truth and
say that whatever is practically useful is true, we must
admit that all knowledge about the past and the future is

untrue, since it is not connected with any present practical

activity. So also with regard to the knowledge that is con-
nected with no activity but the absence of all activity, t.e.

the attitude of indifference towards the object of knowledge.

Some knowledge may involve a tendency tovrards what is

pleasurable, and some a tendency away from what is painful.

But there is also a knowledge of what is neutral {upeksanlya),
with regard to which we remain indifferent. Such know-
ledge may be as valid as any other, of which the Buddhists
can speak, although it is not connected with any useful

practical activity. Finally, with the pragmatist definition

of true knowledge ipramd), the Buddhists cannot but admit

the validity of memory and samkalpaka perception, since
both of them have great practical utility and Loth lead up
to the attainment of their respective objects.^

» NM., pp, 23 f. , TB , pp 14 f.
62 NYAIA THEORY OF KNOWLEDGE

The realistic schools of Bauddha Philosophy, namely, the

Sautrantika and the Vaibhasika, define prama as consisting
in the identity of content between a cognition and the
cognitum.^ A cognition becomes valid when it has the
same content as the object cognised by it. The proper
function of pmwflwa is to give a true knowledge of objects.
But that gives a true congnition of the object which
determines the cognition in question to have the same form
and structure, in which the object exists, so as to give it

the character of objectness (nsayaUl). The sense organs,

however, which arc generally supposed to be the organs of
knowledge (pulnakaiana) do not determine our cognitions to
have the same content with the objects cognised by them.
The cognition of blue colour is not certainly due to the
action of the eyes, for the same eyes are operative even in

the cognition of colours other than blue. It is the blue

content of the object that determines our cognition to be a

cognition of blue colour. The content of the object being
impressed on our cognition gives the same content to it,

and thereby reveals the object itself as having that content.

Therefore, the objective datum (arthdMra) is the pramdna
or the source of our knowledge of the object, inasmuch as

it IS the given datum that determines the object as well as

our knowledge of it one way or the other. The content of

the object is thus both the ground and the product of know-
ledge, the means and the end of the process of knowledge.
It should not be supposed that there is a contradiction in
the same thing being the content of both the object and its

knowledge. For here the object is only the object of

knowledge and the knowledge is a determination of the
object itself. “ When a free is known as hmiapd, the
nature of the hmUapd is the content of both the tree and
our knowledge of it. It is the object of our knowledgeas

* Visayasafupyam sakarasya vijfianasya, NVT.,p. 20.

 VALID KNOWLEDGE AND ITS METHOD 63

well as the ground of a discriminative cognition ot the

object in question. Hence the content of an object
{arthdhara) is pramdna m so far as it establishes an identity
between the object and our knowledge of it.”
^

The Nyaya rejects also the above view of the Bauddha

realist. To it, the view that the content of the object is the
ground of its knowledge because it ensures correspondence
between the two, is not intelligible. It cannot mean
that the content of the object reproduces itself as the
content of cognition, for the same thing can not act
as a cause in relatioh to itself Nor can mean that the
it

object’s content is revealed by its cognition The content

being identical with the cognition need not be revealed or
manifested by another act ot cognition. Nor again can it be
taken to mean that the object’s content is what discriminates
a cognition and thereby produces a discriminative knowledge
of itself. The content and the cognition being identical

the one cannot discriminate the other. The law of dis-

crimination requires that the discriminator must be some-

how different from the discriminated. When I discriminate

a blue colour, I am obviously different from and stand over

against the colour winch is an object of my thought. All

discrimination must take place in this way. The same

thing cannot therefore be both the object of knowledge and
^
the content of knowledge
The Nyaya criticism of the Bauddha view of correspon-

dence between knowledge and its object contains an element

of important truth. It has the effect of showing that the

correspondence betw'een knowledge and its object has no

meaning when, as on the Bauddha view, the two become

fused together as one stuff It is meaningless to speak of

correspondence between knowledge and its object, if we take

them as identicals or absolute similars. Correspondence

1 Vtde NVTP., pp 169-64 (Bib. Ind. Bdn.)

* Vide NVTP., pp 177 f.

 64 NliYA THEORY OF KNOWLEDGE

between cognition and its object is intelligible when each of

them has certain conditions and characters that are wanting
in the other, i e. when there are distinctive differences

between them. Hence we cannot accept the view of the

new realists *
that ‘
the content of knowledge, that which
lies in or before the mind when knowledge takes place, is

numerically identical with the thing known, and is not in a

class by itself. This means that things, when consciousness
is had of them, become themselves contents of conscious-
ness and the same things figure both in the so-called

external world and in the manifold which introspection

reveals Thus objects literally and actually enter into the
mind, and not subjective facts like cognitions or ideas.’ On
this view, the distinction between knowledge and its object

or between truth and error becomes meaningless.

The Bauddha idealists, namely, the Yogacaras give another
definition of pramdna. According to them, consciousness
{vijMna) as the principle of self-manifestation is the source
^
of all knowledge {pramdna) Having no determinations in

itself, consciousness comes to have certain determinate con-

tents in order to manifest itself and thereby gives us know-
ledge of a world of objects, A pramdna is that which
manifests objects, but manifestation as a conscious process
can belong only to that which is intelligent and conscious.
The sense organs being unintelligent and unconscious cannot
have the power of conscious manifestation. Hence the intellect
itself is to be recognised as pramdm by virtue of its intelli-

gent nature and capacity of manifestation.® It has neither

any permanent subject as its locus nor any objects thart are
external to and independent of it. It is the intellect that
accounts for both the subjective and objective aspects of

experience. With its beginningless tendencies conscious-

^ Cf, The New Reahsm pp 34-36

s Yij&anasyaiv&D&karasjatmanatmapraka^aQasamarthyain, NVT , p 20.

» NVTP.,p.l55.
 VALID KNOWLEDGE AND ITS METHOD 66

ness is manifested m two series, namely, the objective,

consisting of percepts or object-ideas, and the subjective,
consisting of perceptions or subject-ideas. We need not
posit the real existence of objects outside of consciousness.
Consciousness may be regarded as manifesting both itself
and the so-called objects from within itself. The series of
external objects, though not ultimately real, is yet set up by
consciousness lor the sake of practical activity through the
influence ol beginningloss desires and impressions (vasand)
that are inherent in every finite mind. The diversified
contents of experience arise out of the continuous operation
of desire, and their bifurcation into the subject-object series

IS the result of the will to live and act. “ As there are

ultimately no objects or pcrceptiblcs other than the intellect,
the intellect itself is to be recognised as manifesting itself

and IS its own perceptible, luminous with its own light,

’

like light The intellect or consciousness, therefore,

’
^
is

both pramd and pmmdna, the ground of knowledge and the

attained knowledge, since it is the cause of manifestation
and the object manifested m knowledge
As against the Bauddha idealists, the Naiyayikas point
out that the definition of pramdna as the power of self-

manifestation in the nature of consciousness is untenable.

A capacity or power is always a tendency to do something
not yet accomplished. But self-manifestation, being a fait

accompli inherent in the very nature of consciousness,

cannot be said to be the effect or product of any power or

capacity. Again, consciousness, as pramdna, being the

ground of the cognition of objects, can not at the same time

become the objects of cognition. One conscious state may
become object for another state of consciousness, but the same
consciousness can not be both the cognition of object and the
object of cognition. The Yogacaras however take the same

‘ Sarvadarianasorrigraha, Chapter on Bauddha philosophy.

9
66 NYiYA THEORY OF KNOWLEDGE

consciousness as cause and effect, means and result of the

process of knowledge. But it is absurd to speak of the same
thing as the subject and the object, the knower and the
known.
The force of the Naiyayika’s objection against the
Yogacara view of pramam lies in its insistence on a funda-
mental difference between knowledge or consciousness and
its object. Knowledge as manifestation presupposes some
object that is manifested by it. As against all idealism the
modern realists of the West point out that experience or
percipi presupposes existence or esse. For the Naiyayikas
experience or knowledge presupposes some object which may
be mental or physical, existent or non-existent. Hence it

IS meaningless to speak of knowledge as self-manifestation,

i.e. a manifestation of itself as object and by itself as

subject.

4. Nydya criticism of the Mimdmsd and Samkhya mews

In the Bhatta Mimamsa, pramd or true knowledge is

defined as primary and original knowledge (anadhigata).

Hence pramdna is that which gives us new knowledge, i.e.

a true cognition of objects of which we have had no know-

ledge in the past. Every case of knowledge, if it is to be
of any value, should be original in character. It implies a
new step, by which we advance from the known to what is

not yet known. Eeal knowledge is a synthetic process

adding new contents to the old stock of knowledge.
Pramdna is the means of acquiring knowledge, and so must
lead to the acquisition of such knowledge as is not yet
attained but is still to be acquired. If the objects are
already known, there can be no necessity of acquiring a
knowledge, of them. The method of knowledge, therefore,
must be concerned in knowing what has not been previously

» NVT.,p.21:NM.,p. 16.
 VALID KNOWLEDGE AND ITS METHOD 67

known. It follows from this that memory {smrh) cannot

be pramd or true knowledge, in so far as it is not a new
experience but the resuscitation of some old experience.
It has no new contents but refers only to the already
acquired contents of knowledge.^
Here the Naiyayikas point out that the dehnition of
prama, as knowledge which has the characteristics of truth
and novelty (ydthdrthya and anadhi-gatatva), is too narrow.
It excludes many cases of knowledge which are undoubtedly
valid but do not refer to absolutely new objects. Eternal
objects, such as space, time, soul, God, etc., cannot be
said to be wholly unknown to us. We consider them to be
eternal because their non-existence at any time cannot be
proved. These have a necessary existence both for our
thought and the things of the world. Our present know-
ledge of such objects comes to us as necessary knowledge.
Once we have such knowledge we cannot say either that the

objects had no existence before or that we had no know-

ledge of them prior to this. Rather we think that we had
an implicit knowledge of the objects, whatever may be the

degree of its clearness or distinctness. They are a priori

like Kant’s categories of the understanding. As Pringle
Pattison says :
“ Mathematical truths, as soon as we realise

them, are seen to be necessary, and we seem to have known

them always.”^ Plato supposes that our knowledge of
them is a recollection. Without going so far it may be

said that we have an a priori knowledge of eternal entities

in the same way in which Russell ® shows we have an a priori

knowledge of general principles. Nevertheless, they may

be better known or cognised by perception, inference and
testimony. But, on the Bhatta view, no knowledge about

1 YatharthamasrhTtagrahi:Mtiaih prama^amiti, 8D , p 45.

2 The Idea of Imnwriahtyt PP. 46-47.
8 Problems of Philosophy Cb. Vn.
68 Nl'iYA THEORY OF KNOWLEDGE

these eternal principles can be valid, since it cannot be

knowledge of what was not at all known before. Further,
our knowledge of ordinary objects more often than not, a is,

knowledge of what was previously known. But that does

not make them less valid than the most valid knowledge w’e
can have.

Again, the validity of pratyabh'i]na or recognition, as a

form of knowledge, becomes inexplicable. To recognise a
thing IS to know it as what was once known before. In it

the object that is now perceived is directly felt as the same

thing that was perceived before. We have, for example,
the judgment ‘this is that man whom I saw yesterday’
The ‘
this ’
of the present perception is identified with the
‘
that ’
of past perception. Recognition cannot, therefore,

be a knowledge of what was not known Still all men

including the Bhattas, admit that recognition is form of
prama or true knowledge But consistency requires that we
must either give up the idea of novelty {anadhigatatm) as a
characteristic of pramd or say that recognition {pi atyahhipid)
is not true knowledge, i.e. is apramd. In fact, however,
no knowledge is made true or false by reason simply of its

originality or unonginality The truth of knowledge does

not depend on the newness of its object.’

In the case of what is called dhdrdvdhikajndna or persis-

tent knowledge, the Bhatta definition pramd obviously
of

fails. When the same thing is known by a man for some

time there is a continuous senes of cognitions with regard
to it. Here all the cognitions, which succeed the first and
constitute the continuous series together with it, refer to the

same thing that has been previously known by the first

cognition. We cannot say that each member of the series
refers to a new object. Hence persistent knowledge is, as
the Bhattas themselves admit, valid knowledge, although

1 NM.pp. 21-22
 VALID KNOWLEDGE AND ITS METHOD 69

it IS not a knowledge of the new, but of the already known

{adhigata)}
Of course, theBhattas contend that the continuous cogni-
tion refers to new objects in all its parts The series of cogni-
tions occurs at different instants of time. The thing as thus
connected with different times, though apparently the same,
becomes really different objects for our persistent know-
ledge of it. The successive cognitions are valid in so far as
each apprehends the object as qualified by a different time
and theiefore as something new.^ The Advaita Vedanta
suggests another way* out of the difficulty According to it,

persistent knowledge is valid either because its different

parts perceive ctift'erent instants of time or because it is one

single cognition as long as it persists and no new mental
modification is produced. A continuous cognition is thus
one present knowledge manifesting one thing which was
previously unmamfested. So the question does not arise
as to whether the series of cognitions apprehends new ob-
jects or not.®

To this the Naiyayikas object that the instants of time

(ksana) cannot be perceived by us. The different instants,

entering into the persistent cognition, being unperceived,

cannot be said to constitute different objects for the series

of cognitions ^ Were these temporal differences apprehended

and wedged into the body of the continuous cognition, its

continuity would be broken up and our sense of continuity

be lost. As that is not the case, we are to say that in per-

sistent knowledge the series of cognitions refer to one and

the same object. Nor can it be urged that persistent know-

ledge is a single state ot cognition enduring for some time.

Although from a subjective standpoint continuous cognition

1 NVT.p n
2 SD.,p 45
3 VP Chap
, I.

1 NVT ,
tbtd.
 70 NYiYA THEORY OF KNOWLEDGE

may be considered to be one present state of conscious illu-

mination, yet objectively it is a series of cognitions. The

present is not a point or one instant of time, which is im-
perceptible, but a mass or block comprising several instants,
while a cognition cannot endure for more than one moment
or instant of time. Hence a present continuous cognition
is really a series of cognitions, of which those that succeed
the first are admitted by all to be as valid as the first. ^ It
cannot be seriously maintained that they open up new
aspects of the object. ‘
The palm of the hand seen a thou-
sandth time adds no new content to our previous knowledge
ofit.’^ In fact the validity of knowledge does not lie in
any character of novelty No knowledge is made true or
false by reason simply of its originality or unonginality.
The truth of knowledge does not depend on the newness of
its object. The validity of any knowledge comprises three
facts, namely, first a correct presentation ol the object ;

second, the practical activity of the knowing subject in

response to the presentation, and finally, fulfilment of the
activity in relation to the object. These three facts are

inseparably bound up with one another, so that with the

true presentation of the object, there follows the reaction
of the subject and the fulfilment of the reaction by way
of its producing the expected results. Now all the repeated
exjseriences of a thing are equally connected with these
three facts Hence there is no reason to think, as the
Bhatta view will lead us to think, that the first cognition

of a thing is valid knowledge, while all other subsequent

cognitions of it are invalid.®

' TC , 1, pp 379 f Compare the Nyaya account of *

the present ’
with James’s
‘
specious present
’
{Principles of Psychology) and Titchener’s *
tine-6eld ’
(Text-Book

of Psychology)
» NM..pp.21f.
3 NVT.. 21 NM., thtd.
 VALID KNOWLEDGE AND ITS METHOD 71

The Prabhakara Mimamsa defines pramd or valid know-

ledge as immediate experience {anubhuti). It is different

from memory which is due solely to the impressions of past

experiences. All immediate experiences have intrinsic

validity. There cannot be any question as to the validity
of immediate experience, because that is self-evident.
Memory however is mediate knowledge, being condi-
tioned by past experience. Hence the truth of knowledge
ipramdnya) is guaranteed by its having the character of
immediacy.^

The Naiyayikas bring forward the charge of inconsist-

ency against the Prabhakara definition of pramd. If all

cognitions are valid by themselves, there is no justification

for treating memory-cognition as invalid on the ground

that it IS not immediate experience. It is also curious that

the Prabhakaras take memory as valid so far as the mani-

festation of knowledge and the knower is concerned, but
invalid with regard to the manifestation of the object.

According to them, every cognition is a triune manifestation

(tnpuHsamint) It manifests the subject, the object and itself

at one and the same time. Memory as a cognition is valid

so far as it manifests the knower and itself (dtmasvdtma),

but invalid so far as it manifests the object (vedya). But
there is no sense in this invidious distinction Either memory
18 wholly valid or it is not valid at all. Further, it is difficult

to see what anubhuti or immediate experience really means.

It can not mean such knowledge as is not conditioned in its
originby some other knowledge. If it did, savikalpaka or
determinate perception and inference would become invalid,
since these depend on previous experience. Other possible
meanings of anubhuti also do not stand scrutiny. So the

1 Pr a-m^nam nniibhutiti sa smjteranya, etc ,

PrakaTanapaflcM, p. 43.
 72 NYiYA THEORY OF KNOWLEDGE

Prabhakara definition of prama and pramana is rejected as

unsound by the Nyaya.^
In the Samkhya system, it is the function of the
intellect (buddhivrtli) that is regarded as pramana or the
specific cause of true knowledge. The self knows an object
through a mental modification that corresponds to the im-
pression produced in the sense organ by the object in ques-
tion The object having impressed its form on the sense
organ, the mind presents it to the self through a correspond-
ing modification of itself Hence the mental function is

ptamana or the source of our knowledge of the object.

The Naiyayika rejects tins view also as untenable

According to him, it is unintelligible how a material and un-
conscious principle like Imddht can be the locus or the

substratum of knowledge. It is the self that has the cogni-

tion of objects and not any blind modification of unconscious
matter.^
The Nyaya criticism of the Bhattaviewof prama as know-
ledge of what was not previously known raises an important
problem of knowledge The problem is this Is knowledge .

a cognition of the known or of the unknown ? If it be a

cognition of the known, there is no need of it ;

if it be a
cognition of the unknown, there is no possibility of attain-

ing it. We do notknow a thing which is already

want to

known, and cannot seek to know anything which is absolute-

ly unknown. The Bhattas would say that since the known
need not be known again, all knowledge must be a cognition
of what was not known. On the other hand, the Naiya-
yikas point out that knowledge need not necessarily be a
cognition of what was not previously known. If it were so,

we could not speak of the development of knowledge or of a

knowledge of the old and the familiar as we have it in recog-

* TE and 8S., pp. 19 f

® NM.p.26.
 VALID KNOWLEDGE AND ITS METHOD 73

nition. It seems to us that the Naiyayikas are substantially

right in their contention that the knowledge of what is al-
already known is possible. All knowledge, except acquaint-
ance, admits of degrees of determinateness.' Our knowledge
of objects may pass from an indeterminate cognition of
their bare existence to a definite recognition of their nature,
character and past history. The more we know of the
characteristics of an object, the more determinate is our
knowledge of that What
known to have certain
object. is

characteristics may be further known to have other

important characteristics. It is in this way that our know-
ledge of an object develops and becomes more precise
and comprehensive. It is tiue that the other characteristics

were not previously known and so impart to the later

knowledge a character of novelty. This however does

not show that the object itself becomes new whenever we
discern new characteristics in it. Eather we are to say
that we know the same object which, in a way, we
already know. In fact, our response to an absolutely
new object is more like a shock of surprise than know-
ledge in the proper sense.
The Prabhakara view of prama as immediate experience
(annbhuti) is not really refuted by the Nyaya. Its criticism

of this view generally sounds like the ignoratio elenchi.

What it does is not to attack anubhuti as a character of

true knowledge, but to show its inconsistency with the

Prabhakara account of memory. In fact, the Prabhakara’s

anubhuti and the Naiyayika’s anubhava are cognate con-

cepts. Their use of these concepts to exclude memory

from prama or valid knowledge appears, as we shall see, to

be equally unsound. Anubhuti or anubhava, as a character

of prama does not necessarily imply that memory is not

prama or valid knowledge.

to Logic, p. 24.
1 C/ L 8 Stebbing, Modern Introductton

10
 NYiYA THEORY OF KNOWLEDGE

The Nyaya should not have found fault with the Samkhya
views of the intellect or the mind as an organ of knowledge

'{pramana). On its own showing, prawdp as the unique cause

of prama is a complex of physical and psychical conditions

(bodhabodhasrabhm smagrl). Hence there seems to be

nothing wrong merely in taking a material principle like

buddhi as the means or organ of knowledge (pmmdna).

74
 CHAPTER IV

THE FACTORS OF VALID KNOWLEDGE (PRAMi)

1. The subject, object and method of valid knowledge

Prama or valid knowledge has been defined by the Nyaya

as true presentational cognition (yathdrthdnubhava). If

we analyse this concep'tion of prama we shall get three

essential factors involved in all valid knowledge. Knowledge

as such implies a subject-object relation. In all knowledge,

be it true or false or neither, we see that a subject or

knower stands related to an object, in so far as the former

has a cognition of the latter. When however we have not

any knowledge or cognition in view but only true or valid

knowledge (prama), there must be another factor, namely, a
method knowledge (pramdna). Of course, any knowledge,
of

true or false, may be said to be produced by certain condi-

tions or causes. But the cause or ground of wrong know-
ledge is not pramdna or a method of knowledge. The cause
non-valid knowledge (apramd) is not,
or the ground of

therefore, a factor of valid knowledge (prama). Hence we

prama or valid knowledge implies
see that the conception of
subject, the object and
three necessary factors, namely, the
(pramdtd, prameya and pramdna)}
the method of knowledge
has been said
As to the first, namely, the pramdtd, it
a subject or knower, in which
that every knowledge involves
The subject is the
knowledge inheres as an attribute.
It is that which likes
substantive ground of all cognitions.

sat, pramavyaptatf pramanan,. SarvadarianattO^groha.

1 Sadhanafrayavyatinktatve

CbttpUr on Nyayr» plalosopby

76 NYiYA THEORY OF KNOWLEDGE

and dislikes things and acts accordingly. Hence the

pramdta is the self conceived as an intelligent agent. It is

also independent in the sense that it exists for itself and is

an end to itself. Its independence comes out in the facts

that it enjoys and suffers, it is the centre of all activities,

and that it directs the course of other things without being

determined in its course by them, i e. it is self-determined.
The pramdtd thus described by the Nyaya resembles the self

as that is conceived in modern self-psychology.* Like it,

the pramdtd is that which knows and strives, enjoys and

suffers, remembers and expects; it is an agent, a striver, a
desirer, a refuser.

Secondly, pramd or valid knowledge implies some

prameya or object, to which the process of knowledge refers

or to which it is directed The object of knowledge may be

either existent or non-existent. Both positive and negative
facts may become the objects of true knowledge, but the
knowledge takes different forms in the two cases. In the
case of existent objects our knowledge is positive and does not
depend on any objects other than its own The knowledge
of non-existent objects is negative and conditional on the
direct apprehension of similar existent objects. “ The light

of a lamp, which reveals the existence of certain perceived

objects in a dark room, manifests also the non-existence of
those that are not perceived, for if the latter had existed
there, they would have been perceived like the similar per-
ceived objects.” * Hence just as there can be no knowledge
without a conscious subject that knows, so there is no
knowledge without an object —a thing or an attribute, a
state or a process, a positive or a negative fact that is

known. Subject and object (pramdtd and prameya) are

strictly correlative factors involved in all knowledge. They

• Cf. McDoagall, Ovthne of Psychology, pp. 21, 426.

‘ NB., 1. 1 1.
 Factors of valid kkowledge 77

are distinguishable no doubt as the knower and the known,

but not separable in any act of knowledge.
Thirdly, all true knowledge must be connected with some
method of knowledge. In Western philosophy it is customary
to analyse the knowledge-relation into the three factors of
subject, object and process of knowledge These correspond
respectively to the pramatd, prameya and pramd in Indian
philosophy. In addition to these three, the Nyaya recognises
the special cause of knowledge (pramdna) as an important
factor. This is what in Western logic is usually called a
method of knowledge and not a factor of it. While it does
not overlook their distinction, the Nyaya rightly considers
the subject, object, method and resulting state of knowledge
{pramatd, prameya, pramdna and pramd) as mutually impli-
cated aspects of the whole truth. Each of. these is as essen-
tial to knowledge as the rest, and each of them involves the
rest by way of logical implication {prasaldt). So it has
been said that in these four principles, when taken together
in one whole, but never as disjoined, there is the realisation

of truth {tattbaparisamdpti) Truth is realised when the

subject having known the real by pramdna, as good or bad
or neutral, proceeds to obtain it o*" avoid it or remains in-

different to it, as the case may be, and ends iii actual attain-

ment or avoidance or mere apathy.

The real has thus a value for the knowing subject. Any
account of reality as absolutely foreign to our subjective in-

terests and personal values would be a fundamental misconcep-

tion of it. The different systems of Indian philosophy agree

in holding that the world of experience is a system of moral

dispensation, in which man has to work out his destiny in

obedience to the universal law of moral causation (karma).

Eight knowledge of reality is of supreme importance for

True knowledge is not

man to reach the destination of life.

a passive and lifeless reflection of reality . On the other

hand, it is that philosophic view of reality, which has its

 78 NYSYA THEORY OF KNOWLEDGE

basis in the vital needs of our spiritual nature and is essen-

tially conducive to the attainment of our supreme life-pur-
pose {nihireyasa) Herein lies one of the striking points of
contrast between Eastern and Western philosophy. In the
West, philosophy is mainly a matter of intellectual under-
standing. It IS indeed a reflective knowledge of the nature
and relations of things and beings. But such knowledge has
not a direct bearing on life and conduct. It is seldom carried
on with the express intention and determined will to solve

the problems and work out the final good of life. To the
ancient Indian thinkers, however, philosophy was not a
mere rational knowledge of things and theories, but a means
to the realisation of our life-end, the path to the final goal

of life.

2. Distinction of the method from the subject and

object of valid Imowledge

It will appear from the preceding section that the sub-

ject (pramdta), the object (prameya) and the method
(pramdna) are all necessary conditions of valid knowledge
ipramd). No valid knowledge is possible without any one
of them. Eor valid knowledge, the subject and object are
as much necessary as the method of it. Hence the question
naturally arises : How are we to distinguish the method
from the subject and object and say that the first is the spe-

cial cause (asadharana karana) of valid knowledge ? Why

is it that among the equally necessary factors of knowledge
some one should be marked off from the rest and called the

unique operative cause (karana) of it ? How do we know

that pramdna is the most eflScient ground tsddhakatama) of
knowledge, while the other factors are only the general
conditions or the logical implications of knowledge ?

In the Nyaya system the answer to this question is to be

found in two very important considerations. That pramdna

 FACTORS OF VALID KNOWLEDGE 79

is the supreme cause and the most important ground of valid

knowledge will, in the first place, appear from the following
facts.

First of all we see that there is a uniform relation of

agreement in presence and in absence between pramdm and
pramd, as between cause and eftect. A piamdna is always
accompanied by valid knowledge which, in its turn, can
never arise without the former. Of course, there can be no
valid knowledge w'lthout a subject and an object. But every
case of the existence of subject and object is not necessarily
a case of the appearance of pramd or valid knowledge. On
the contrary, in every case in which a pramdna is operative,
pramd or valid knowledge must appear as a natural sequence.
Thus a man has no perception of objects in relation to which
no sense organ is operative, although he, as subject, and
those objects exist side by side If, however, the objects are
in contact with his sense and his mind responds to the sense

impressions, he cannot but have perception of those objects.

Secondly, we observe that the pramdtd or subj'ect arrives
at a true knowledge of objects only when it is aided by a
piamdna or an operative cause of knowledge. Contrariwise,
we find that the subject and object cannot produce any know-
ledge in the absense of the pramdna or the source of know-
ledge. That is, the subject knows objects only when it
makes use of a certain method, but not singly by itself.

Thirdly, w'e see that pramdna is the last link in the chain

of antecedent conditions that lead to the knowledge of objects.

It is the immediate antecedent to the origin of knowledge.
The aggregate of psycho-physical conditions, on which
knowledge depends, is completed by pramdna, and knowledge
appears immediately as an effect. Pramdna is the cause of

knowledge inasmuch as it is the immediate antecedent, on

which knowledge follows first and immediate.^

> NV., pp 6-7 ; NVT , pp. 22-26

 80 NYIYA THEORY OF KNOWLEDGE

Lastly, it has been pointed out by the Naiyayikas that a

distinction between the different kinds of knowledge is made
by reference to the methods of acquiring knowledge.
Perception, inference, testimony, etc., are regarded as
different kinds of knowledge because they are due to different

pramdnas or methods o( knowledge. This cannot be due

to the subject or the object of knowledge, because these
may be the s.ime in what are generally admitted to be
different kinds of knowledge. The same subject may know
the same object first by inference and then by perception,
as when a in in confirms the inference of fire in a distant
place by approaching it. lienee the subject and object
cannot explain why one kind of knowledge is called percep-

tion and another inference Similarly, we find that the

mind’s contact with the soul is the common mediate cause
of all forms of knowledge. But the mode of this contact

is different ini different kinds of knowledge. We cannot

account for such different modes by the subject and object
of knowledge, for they may be the same in two kinds of
knowledge. It is the pramina that determines the mind’s
contact with the soul in different ways in the different
kinds of knowledge.’

There are two ways of classifying knowledge, i.e. by

reference to the nature of the objects known, and by
reference to the grounds of knowledge. According to the
first, we have as many kinds of knowledge as there are
kinds of knowables or possible objects of knowledge. This
way of distinguishing between the different kinds of
knowledge has been followed by the Jamas in their theory
of knowledge which divides knowledge broadly into the
two kinds of pratijaksa or immediate and paroksa or mediate.
Hobhouse als) follows the same principle in classifying the
methods of knowledge in his Theory of Knowledge. The

> Ibtd.
 FACTORS OF VALID KNOWLEDGE 81

second way, however, is generally accepted in Western

philosophy. According to this, there are as many kinds of
knowledge as there are ways of knowing or specific grounds
of knowledge.The Nyaya follows this way along with the
Vedanta and some other Indian systems. It shows also
that a distinction of knowledge into different kinds cannot
be based on the subject or the object of knowledge. The
conclusion drawn from this and other facts is that pramana
or the method is the operative cause of knowledge (pramd-
karanam).

The second consideration, on which the superiority of

pramana to the other farctors of valid knowledge is based, is
this. The primary function of knowledge is to give us
truth in the sense of real correspondence between idea and
object (arthavattvam). Now for the fulfilment of this
function knowledge is primarily and directly dependent on
pramana or the operative cause of knowledge. The other
factors of knowledge cannot lead to truth except through
'
the aid of pramana The objective validity of knowledge
is directly dependent on the efficacy of the method or
pramana employed to acquire it. The subject or pramdtd
cannot directly produce the validity of knowledge, because
as an agent it requires means to bring about a result and
cannot itself directly produce the result. Nor can the
object or prameya be said to produce the state of valid

knowledge, for in inference the object is absent and cannot,

therefore, be operative in producing a knowledge of itself in

the knowing subject. It may, of course, be said that once

we have the truth, we find it as belonging to the subject,

the object and the knowledge-relation between the two.
Still the subject, the object and the state of knowledge do
not produce the truth, but owe it to the functioning of

^ Arthavati ca pram&^e pramatS prameyaih pramitintyarthavanti bhavanti, NB.,

1 . 1 .1 .

11
82 NYAYA THEORY OF KNOWLEDGE

pramdm or the ground of knowledge. Hence pramdna or

the method of knowledge is the means or the operative cause
(karana) of knowledge, as distinguished from the pramata
or subject and prameya or object which are indeed logically
implied in all knowledge but are not directly concerned in
producing objectively valid knowledge {prama)}
Thus according to the Naiyayikas, the objective validity
of knowledge is due to pramdna or the method on which it
is based. The conscious subject and the cognised object
cannot account for the correspondence of knowledge with
real facts. The subject and the object participate in truth

in so far as they are made to do so by some efficient organ

of knowledge, the sense or the reason with which we are
endowed. The universal condition of all knowledge is indeed
consciousness. But from mere consciousness we cannot
deduce the specific modes of knowledge, such as perception
of the table, inference of coming rain, verbal cognitions

and so forth. Hence while consciousness seems to be the

first and the general cause of all knowledge, we require
certain specific second causes to explain the particular modes
of knowledge and their correspondence to particular objects
or facts of the world. Such specific causes of knowledge
are called pramdnas in Indian philosophy.

1 NVT , pp. 22, 29 f.

 CHAPTER V

THE TEST OF TRUTH AND ERROR

1. The problems and alternative solutions

In the preceding chapters we have considered the

different conceptions of bhrama or error and prama or true
knowledge. We have also seen that the specific modes of
knowledge arise from certain operative causes or specific
conditions {]ndnakarana) These are called pramana when
the knowledge is true and apramana when it is false. Here
we have to consider the following problems as to the truth

and falsity of knowledge. Admitting that kn(<wledge

depends on certain specific conditions for its origin, how
are we to explain its truth or falsehood ? How again are
we to know its truth or falsehood as the case may be ? In
other words, the questions are : How is the validity or

invalidity of knowledge constituted ? And, how is its

validity or invalidity known by us ? The first question

refers to the conditions of origin (utpatii), while the second,

to the conditions of ascertainment (jhapti) of truth and

falsity.

Generally speaking, two possible answers may be given

to the above two questions. First, it may be said that
knowledge is both made and ascertained to be valid or

invalid by the same conditions which bring about that

knowledge (jhdnasdmagri) Secondly, it may be said that the

truth or falsity of knowledge is both constituted and known

by external conditions. On the first alternative, both truth

and error would be self-evident (svatah). On the second

alternative, neither truth nor falsehood could be

self-evident,
 84 NYiYA THEORY OF KNOWLEDGE

but both must be evidenced by something else (paratah).

Or, a distinction may be made between the two cases of

truth and falsehood. It may thus be said that w'hile the
truth of knowledge is constituted and ascertained by intrinsic
conditions (svatah)^ its falsity is made so by extrinsic
conditions {paratah ) . Or, we may just reverse the order
and say that while falsity is self-evident (svatah), truth

requires evidence or proof by external conditions (paratah).

Thus we come upon four alternative solutions of the

problems set forth above. Each of these has been adopted
and supported by one or other of the systems of Indian
philosophy. Samkhya accepts the first alternative, namely,
that both the validity and invalidity of knowledge are self-

evident. The Nyaya is generally in favour of the second,

viz. that neither validity nor invalidity is self-evident, but
that both are constituted and known by external conditions.
The Bauddhas support the view that falsity is self-evident

in knowledge and that external conditions are necessary for

truth, if there be any. TheMimamsa and the Vedanta advo-
cate the theory that all knowledge has self-evident validity,

while falsehood is due to certain extrinsic conditions.^

2. The Nyaya theory of extrinsic validity and invalidity

According to the Nyaya, knowledge is just the

manifestation (prakaia) of objects. As such, it is neutral
to truth and falsehood. No knowledge is true or false on
its own account, i. e. simply because it is produced by cer-
tain specific causes (jMnasdmagri). The truth or falsity

of knowledge depends respectively on its conformity or non-

conformity to objects or facts. A knowledge is true when
it corresponds to the real nature and relations of its object
if not, it becomes false, provided it claims to be true. Thus

^ SarvaddT^cfUua^grahat Chapter on Jaimioi system.

 lEST OF TRtiTH AND ERROR 85

truth and falsity are characters that appear to be added to

knowledge which is indifferent to both, but may have either,
according to special circumstances/ Hence the conditions
of the validity or invalidity of knowledge must be different
from and other than the conditions of the knowledge itself.

If knowledge and its validity were conditioned by the same

conditions, then error too would become valid knowledge.
Even error is a form of knowledge and is conditioned by
the conditions of the knowledge. Hence it should be
as good as valid knowledge. That is, the validity of
knowledge being conditioned by the conditions of knowled-
ge itself, there cannot- be any false knowledge. On the
other hand, if knowledge and its invalidity were due to
the same conditions, there can be no valid knowledge.
Hence we must admit that the truth and falsity of know-
ledge are due to different special conditions other than the
conditions or specific causes of knowledge itself.

What then are the special conditions of the validity

and invalidity of knowledge ? According to the Nyaya,
the validity of knowledge is due to the efficiency of the
conditions of knowledge (Mranaguna), while its invalidity

is due to some deficiency in those conditions (karanadosa)

The efficiency or deficiency of the conditions is constituted

by certain positive factors. We cannot say that the effi-

ciency of the conditions of knowledge is simply the absence

of defects in them {dosahhavd), or that deficiency means only
the absence of efficient conditions {gunabhava). Both effici-

ency and deficiency stand for certain special positive conditions

which modify the general conditions of knowledge and make
it true and false respectively. In fact, the specific character

of an effect {kaiyavi^csa) is to be explained by some specific

character of the cause (kdranaviSesa) Truth as a specific

character of some knowledge and falsity as another speci-

1 YatharthetarasadbSraDO dharmo bodharupatvam, NM., p. 160.

 86 NYiYA THEORY OF KNOWLEDGE

fic character of some other knowledge must therefore be

due to different specific characters of the general conditions
of knowledge.^ Thus the contact of an object with a sense
organ is the general condition of perception. But its validity

depends on such special conditions as the health of the

sense organ, nearness of the object, sufficient light and
sense-object contact On the contrary, perception is invalid
when its general conditions are modified or vitiated by
such other special conditions as disease, distance, darkness
and slender sense-object contact Similar is the case with
inference and other kinds of knowledge. Hence both the
validity and invalidity of knowledge are conditioned by
extrinsic conditions other than the conditions of knowled-
ge itself iparatah utpattih).

Next we come to the question of the ascertainment

of validity and invalidity (pramanya-pramdnyamscaya).
How is the validity or invalidity of knowledge known ?
Are they known by the same conditions that constitute
knowledge or by any other external conditions ? Generally
speaking, the Naiyayikas maintain that the truth and'
falsity of knowledge are known by certain external condi-
tions other than the conditions of knowledge itself'

iparatah prdmdnydprdnidnya]hapti) If the validity of

knowledge be known intrinsically by the conditions of

knowledge itself, there can be no doubt with regard to

the validity of any knowledge. On the other hand, if the

knowledge of invalidity be due to the conditions of know-
ledge itself, there can be no wrong action. That is, if truth
be self-evident, there need be no doubt and dispute about
knowledge, and if falsehood be self-evident, there should be
no illusion and disappointment. In fact, neither truth nor
falsehood is known to belong to knowledge just at the time

1 Doso 'pra’naya janakab pram&yastu gupo hbavet, etc StddhdntakmvIavaJi^ 131 ,

prama jfianahctvatiriktahetvadhioft karyyatve ‘ati tadvii^e^atvat aprarnavat, Kusum^

dfljah, pp. 2 j7*09.
 TEST OF TRUTH AND ERROR 87

we have that knowledge. As a general rule, the validity or

invalidity of knowledge is known some time after the know-
ledge itself has appeared.*

What then are the external conditions of the knowledge

of validity or invalidity of knowledge The NaiySyikas
?
answer that both are known by inference. While know-
ledge may be internal l y perceived, its validit y or Jn validity
18 to he inlejr^ from such extrinsic conditions as its

capacity or incapacity to produce successful activity

(pra ( r/ hsaviarthy^ Knowledge is known to be valid
when it leads to successful activity in relation to its object.
We know objects by means of perception, inference, etc.
This knowledge sets up certain psycho-physical reactions
if ravrth) in the knowing subject. The success of these
reactions, i.c. their being duly connected with the expected
object, is the evidence for the truth of the knowledge. Men
sometimes act on wrong knowledge under the belief that it

is true and arises out of valid conditions. The reason for this
confusion between true and false knowledge is some point of
similarity between the two Even wrong knowledge is as
good a cognition of objects with their general characters
{samanyapancchedaka) as right knowledge is such. So one
is apt to be mistaken for the other. But in such cases what
distinguishes valid knowledge from the invalid is the test

of successful activity A valid knowledge not only gives us

a cognition of some object, but also leads to successful

actions on the part of the knowing subject. An invalid

knowledge, on the other hand, gives us a cognition of

objects indeed, but it fails to lead to successful activity.

If our knowledge does not correspond to its object, it can

not be practically efficient in relation to it. In the cases

of illusion, hallucination, dream, etc., our knowledge can-

1 TC , I, pp 184, 233 , NM., pp 160, 169 f

^ hi 8araa-rthaprav|'ttiji,nakattvftdaiiumeyftni, NVTP , 1.1.1

88 NYiYA THEORY OF KNOWLEDGE

not be the basis of effective actions. The Naiyayikas,

therefore, conclude that the truth of knowledge is not self-

evident in it, but is evidenced or known by inference from

successful activity. By successful activity, they mean the
volitional experiences (arthalmyajmna) of the expected
object (phalapidna). The perception of water in a certain
place is known to be true when by acting on that perception
we meet with the expected water. Contrariwise, a
knowledge is known to be invalid, when it is contradicted by
subsequent volitional experiences {pravrttii'isamvdda). That
IS, the invalidity of knowledge is inferred from the failure

of the practical activities based on it The perception of

silver in a shell is known to be illusory because the act

of picking up does not give the expected silver. Hence

pravrttisdmarthya means that the object as cognised is found
present when acted upon by us, i.e. it is given to

volitional experience just as it was given to the corresponding

cognitive experience.'

In the case of the knowledge of familiar objects

(ahhydsadaMjiidna), we do rot require the test of successful

activity or conative satisfaction (pravrttisdmaithya). In

this case it may seem at first sight that the validity or
invalidity of knowledge is self-evident (svatah). A habitual
experience is known to be valid or invalid even before we
proceed to act upon it and see if it leads to the expected
object or not. It would therefore seem that the validity or
invalidity of habitual experience need not be known by any
inference and, as such, is self-evident. But here the
Naiya^ikas point out that it is a contradiction to say
that the truth of the familiar is self-evident. The
knowledge of the validity of familiar knowledge is

conditioned by the conditions of its familiarity. The

familiarity of knowledge means its similarity to previous

» NM , pp 171-72.
 TEST OF TRUTH AND ERROR 89

knowledge. Hence if we have ascertained the validity of

the previous knowledge, we may very well know the validity
of the present familiar knowledge by an inference based on its

similarity to the former {tajjatiyatva) What happens here

is that the previous verification of knowledge by conative
satisfaction becomes a determinant of similar subsequent
knowledge. This shows that the validity or invalidity of
such knowledge as is not new is known by inference based
on essential similarity or generic identity This inference
is, in most cases, implicit and unconscious. But it is

never absent. We may put it explicitly in the form of a

syllogism like this ‘
All knowledge of a known character is

valid ; this knowledge is of that character; therefore this is

valid.’ So too, mutatis mutandis, for the inference by

which we know the invalidity of the knowledge of familiar
objects.' Hence the Naiyayikas conclude that knowledge
is both made true or false, and known to be true or false by
certain external conditions other than those conditioning
the knowledge itself.

3. Objections to the theory answered by the Nydya

According to the Nyaya, knowledge is not ascertained as

true or false at its very inception. To have knowledge is
not, at the same time, to know it as true or false. The
validity or invalidity of knowledge is first known by us
when we act upon that knowledge and see if the action is

successful or not. But with regard to the test of conative

satisfaction {pravrttisamarthya) as a condition of the
knowledge of the validity of knowledge, it may be asked :

how do we know that the feeling of satisfaction is true and

not false? The perception of water, for example, is to be

known as valid when it leads to the volitional experience of the

I NM , p. 174.

12
90 NYIYA THEORY OF KNOWLEDGE

expected results (arthakriydkhyaphalajmna). But how do

we know that the volitional experience is valid ? Does it

not require to be validated by other conditions? If it does,

there will be no end of the process of validation and the

^
first perception of water cannot be completely verified

To this question the Naiyayikas give two answers. First,

it has been said that the experience of expected objects
(phaJajhdna) does not ordinarily require any test of its

validity, because there is no doubt about it or because there

is the fulfilment of our purpose in it. As for instance, the
first perception of water in a mirage requires to be tested
because we have doubts about its validity, but that of a
man going into water need not be further tested, since it is

not infected by any doubt and it fulfils the man’s

expectations

Secondly, the volitional experience of expected objects

may, if necessary, be verified by certain special characteris-
tics of it. Thus the visual perception of water may be
validated by the expected tactual sensations of it, and the
latter may be further confirmed by the experiences of
bathing, washing, drinking, etc., which are usually associated
with water. It may be urged here that a man has the
whole series of experiences even in a dream. Hence it is at

least theoretically possible that the first volitional experiences

of water as well as those of its usual associates are as invalid

as dream experiences According to the Naiyayikas, this
hypothesis is untenable There is an obvious distinction
between dream consciousness and waking experience. While
the latter is clear and distinct, the former is confused and
indistinct. Dreams have not the order and uniformity of
our waking experiences. Dream experience is contradicted
by waking perceptions There cannot be any retrospection of
dream cognitions (anuvyavasaya) What is cognised in

» NM., p 172.
 TEST OF TRtTTH AND ERROR 91

dream cannot be the object of a later dream cognition as

something that was cognised. In dream all things may be
seen but none remembered as what has been previously seen.
Hence the waking volitional experiences cannot be reduced
to dream.
The second objection against the Nyaya theory of
extrinsic validity is that it involves the fallacy of reasoning
in a circle (parasparasraya) The knowledge of the vali-
dity of knowledge is said to be conditioned by successful
activity, which, in its turn, depends on the knowledge
of validity. Successful activity depends on two conditions.
First, it depends on a true knowledge of objects. Any
knowledge of objects cannot make our actions successful.
If it were so, even a wrong cognition of silver should lead
to the actual attainment of it. Hence successful activity
must always be due to a true knowledge of objects.

Secondly, successful activity requires a right understanding

of those objects as means to some end or good. We strive

for certain objects only when we know them as the neces-

sary conditions of realising some good Such knowledge
may, of course, be derived from inference If the present
objects are similar to other things which proved to be
effective means in the past, we infer that these too will

serve as means to the present end. This then implies that

successful activity requires a valid knowledge of objects

as means to some good. But we cannot know that we

have a valid knowledge of objects unless we already know
what the validity of knowledge means Hence it seems
that successful activity depends on the knowledge of vali-

dity, while the knowledge of validity depends on successful

activity. The two being thus necessarily interdependent,

neither can be made the ground of the other, and so the

validity of knowledge can never be known.

1 NM , Ibtd
s NM , pp 163 f.
92 NXlYA theory of KNOWLEDGE

The Naiyayikas meet the above objection with a just

recognition of the difficulties raised in it. According to

them, the validity of knowledge is not self-evident, but
must be ascertained from certain external conditions. In
the case of new objects of experience, such conditions are
to be found in the success of the practical activities based

on any knowledge. The validity of knowledge is to be

known from its capacity to produce successful activity.
Hence prior to any conative verification, tbe validity of
knowledge remains doubtful. It is also true that a valid
knowledge of objects is the basis of our successful actions
in relation to them. An action cannot lead to the expect-
ed results unless it is grounded on a true knowledge of
some objects as means to some end.
So far the Naiyayikas admit the contention of the
Critics. But they point out that this does not lead to the
conclusion drawn by them. It does not follow that there
can be no successful activity without prior knowledge of
the validity of knowledge. A true knowledge of objects is

by no means the necessary condition of our action {prairtti)

in relation to them. Any knowledge of objects, right or

wrong, IS the sufficient ground for producing certain modes
of action on the part of the knowing subject What
happens generally is that we act even in the midst of un-
certainty and that while acting we may have doubts as to
the success of our actions. Even if it be true that to act
for ends we must adopt means, it is not always necessary
that we must have a true knowledge of the means of actions.
A mere belief in the means as means will suffice for many
voluntary actions.' Again, successful activity may be
dependent on a valid knowledge of objects. But this does
not mean that we must have a knowledge of the validity

1 Cf Btebbicgi Logic tn Practice, p 99 .

“ Many of our most important actions
have to be performed in accordance with beliefs of snch a kind (i. e beliefs more likely

to be true)/*
 TEST OF TRUTH AND ERROR 93

of the knowledge, by which it is conditioned. To act

successfullywe must have true knowledge as the basis
of our activity.But we may have true knowledge even
when we do not know that it is true. To have true
knowledge is not necessarily to know the truth of that
knowledge.^

Hence it follows that we can act when we have some

knowledge of an object. The knowledge, by which our action
IS conditioned, may be true or false without our knowing it

as true or false. If then we find that the action is successful,

we come to know that the knowledge, on which it was

based, is true. If, however, the action becomes unsuccess-
ful, we are convinced that the knowledge, on which it was
grounded, must have been false. All that is necessary for
our actions is that we must believe in what we know, and
not that we must verify it as true before we proceed to act.

When we have ascertained the validity of knowledge

in some cases, some perceptions and inferences, we may in
other similar cases act with an assurance that we have a
true knowledge of some objects and that the means of our
actions are efficient. Here a present knowledge, a new
perception or inference, is known to be valid by reason of
its essential similarity with some past valid knowledge.
So it may be said that prior to successful activity we know
the validity of our knowledge. But even here it should be
noted that the knowledge of validity is not the determinant
or the cause (prayojaka) of practical activity. Such
knowledge may precede activity but it is not a neces-

sary condition of practical activity. If in the face

of this, the critic insists that no successful activity

is possible without previous knowledge of validity, the

reply is that the critical activity itself must stop. The

critic cannot be sure of the validity of his contention

1 NM., p. 178
94 NYSYA THEO, Y OF KNOWLEDGE

without the attainment of success. Hence if the attain-

ment of success presupposes knowledge of validity, there
cannot be any contention at all The contending will can
have no logical justification There is, therefore, no
necessity of the knowledge of validity either for activity

as such or for successful activity. The latter does not pre-

suppose the former Hence there is no fallacy of circular

reasoning involved in the view that successful activity is

the test of the truth of knowledge

The third objection against the ISyaj a view of extrin-

sic validity IS that it involves the fallacy of aigumentum ad
infinitum {anarastha). If the validity of a knowledge is to

be known from an extei nal source, i e by means of some

other knowledge, then we shall have to prove the validating
knowledge on other external grounds, and so on ad infini-

tum. Thus we are to say that the validity of perception

is known by inference, that of inference by comparison
{upamdna), that of the last by testimony, and that of tes-

timony by still other methods of knowledge. Hence the

methods of knowledge must be innumerable. It may be
said that to prove the validity of knowledge we need not go
beyond the four methods, but prove one individual perception
or inference by another perception or inference. Even then
we cannot avoid the difficulty of infinite regress What
will happen is that within the circle of the four metl ods of
the Nyaya, the process of validation of one knowledge by
another will go on for ever. Thus the perception of water
may be known to be valid by inference from successful
activity or essential similiarity. But how are we to know
the validity of the validating inference ? It must be by
some other perception or inference, and so on ad infinitum.
To avoid this difficulty the Naiyayikas cannot say that
while the validity of the primary knowledge is established
by the secondary, that of the latter is self-evident, and so
requires no verification. If the truth of the secondary
 TEST OF TEUTH AND ERROR 96

knowledge be self-evident, there is nothing to prevent the

primary knowledge from having self-evident validity.
Further, it will involve a surrender of the Naiyayika posi-
tion that the validity of all knowledge is constituted and
ascertained by external conditions. Hence it seems that
on the Nyaya theory of validity, the process of the verifica-
tion of knowledge will go on as an infinite chain of
arguments, in winch every link will hang on tlie next, but
the last link is never to be found (amldiparamj)ara)^
To this the Naiyayikas reply that the validit of a
knowledge must be known by extrinsic conditions wherever
it 18 necessary to know it at all But it i^ not always
ncces<^ary to ascertain the validity of a knowledge. It be-
comes necessary when any doubt as to its validity actually
arises. Thus when we have the visual perception of water
and have any doubt about its validity, we do of course
ascertain it by inference fiom some successful activity, t. e.

by touching or drinking the water. But the validity of

the verilying experience requires no further examination
or proof. There being no doubt about its validity we do
not feel any necessity to prove or ascertain it. Hence
the tactual perception of water validates the visual per-

ception of it even when there is no ascertain-

ment of its own validity. When however we have any

doubt about the validity of the tactual perception, we must esta-
blish it by other external conditions, such as the corrobora-
ting testimony of different {lersons. Thus it follows that

to know the validity of a knowledge by external grounds,

it IS not necessary to know the validity of those grounds
so long as they stand undoubted and uncontradicted
{sam^ayohhaca) Tf any one still doubts that the validating
ground may itself bo valid or not, then we have an unmean-
ing motiveless doubt which has no place in logic."

1 TC .
I, pp 276-77 , NM , pp. 1621.
2 NM , p 173 , TC ,
T, pp 277.79, 282 f
96 NYi¥A THEORY OF KNOWLEDGE

Lastly, the Nyaya discusses the sceptical contention that

there cannot be any valid knowledge. By valid knowledge
is meant such knowledge of objects as is due to some method
of knowledge (pramdna). But how is knowledge related
to its object in the order of time ? Does it precede or
succeed or synchronize with the existence of its object
(prameya) ^ Knowledge cannot be said to precede its object,

since no knowledge appears except as the knowledge of some

object. Nor can we say that knowledge succeeds or follows
its object. A thing becomes an object to us in so far as it

is known. There can be no object which is not the object

of some knowledge. Without knowledge there is no object.

If a thing can be an object independently of knowledge,

there is no need of a method of knowledge for it. Nor
again can we say that knowledge and the objects of know-
ledge co-exist in time. If that were so, all objects of the
world will be known at the same time, and there can be
no desire to increase our knowledge of things. Further,
this will contradict the Nyaya view of the serial order of

cognitions, from which the existence of manas or the

internal sense is inferred. Hence it follows that there can
be neither knowledge nor a method of knowledge {pra-
^
mdna).
This IS the sceptical objection against the possibility

of knowledge as such. It denies the possibility of know-

ledge on the ground that the reference of knowledge to its

object IS inexplicable in the order of time. To this we

may of course say with Green that, even if knowledge be taken
as an event in time, its reference to the object is timeless,
so that the question of the temporal relation between know-
ledge and its object does not arise. The Naiyayikas,
however, admit that knowledge refers to its object in the
order of time. But they point out that the temporal order

> NB., 2.1. ail.

 TEST OF TRUTH AND ERROR 97

between knowledge and its object is indeterminate. It is

not the case that knowledge must have a fixed temporal

order of priority or posteriority or simultaneity with its
object Knowledge arises out of certain causes and refers
to some objects. In some cases knowledge precedes the
existence of its object, as when we know something
that is to happen in the future. In other cases the
object as a physical tiling or event may be said to precede
our knowledge of it, as when we know that something was
or had happened in the past. In still other cases knowledge
and its object may be said to co-exist or to appear simultane-
ously in time. This is illustrated by the perception of pre-
sent facts and, still better, by introspective knowledge of
mental contents. Hence the Naiyayikas maintain that tl e
time-relation between knowledge and its object cannot be
objectively determined as something unalterably fixed like

that between cause and effect. It is a variable relation

which IS to be determined as of this or that kind by actual
observation of the instance of knowledge in question. In
fact, the same thing can, in different cases, be called the
knowledge of object (prama), the object of knowledge
ipramcya) and the operative cause of knowledge {pramana)
just as the same word may be subject and predicate in differ-

ent relations and positions. Hence the question of the

time-relation between knowledge and the object of knowledge
does not necessarily lead to tbe denial of all knowledge.
Further, scepticism, if it is to be consistent, cannot really
deny the possibility of knowledge. To deny knowledge is to

disbelieve it. But to disbelieve is to know that something is

not true. Hence the denial of knowledge must have a posi-

tive basis in some kind of knowledge. But it is a contra-
diction to deny knowledge by means of knowledge. This is
the Naiyayika’s reductio ad ahsurdum of scepticism.^

1 NB ,2112-16

13— (1117B)
98 NliYA THEORY OF KNOWLEDGE

4. Criticism of the Sdrhkhya view of intrinsic validity

and invalidity

According to the Samkhya, truth and falsity are inherent

in knowledge. A knowledge is both made true or false and
known to be true or false by the conditions of the knowledge
itself. Validity and invalidity cannot be produced in any
knowledge ah extra, but must belong to it ah imho. The
one IS as much intrinsic or internally conditioned as the
other. Hence knowledge must have validity or invalidity

on its own account and, as such, these must be self-evident.

This view follows from the Samkhya theory of immanent
causality (satkaryavada). According to this, causation is

only manifestation of the effect that potentially pre-exists in

the cause. A cause can produce only that effect which is

inherent m the causal complex. Otherwise, any cause will

produce any effect, even the unreal and the fictitious. Hence
the validity or invalidity of cognitions as causally determin-
ed effects must be regarded as somehow inherent in the cog-
nitions. This means that validity and invalidity are inherent
in knowledge. Thus the validity and invalidity of know-
ledge are self-evident.^

The Samkhya view has been criticised by the Nyaya and

the Mlm5ihsa. The latter points out that the theory of
causality, on which the Samkhya view of the validity and
invalidity of knowledge rests, is itself untenable. Causation
or effectuation has no meaning if what is caused is pre-exist-
ent and so need not really be caused or produced. Causation
must be a process of real effectuation, i.e. it must be the
production of the new or the previously non-existent effect.
Further, it is a contradiction to say that both validity and
invalidity belong to the same thing, namely, knowledge.
How can such contradictory characters belong to the same

« BD. ft SC., p. 20 . NM., p. 160,

 tESt OP Tbutfi ANb ERROR 99

thing ? It cannot be said that while validity is intrinsic to

valid knowledge, invalidity is intrinsic to invalid knowledge.
Apart from any external conditions, a knowledge cannot
determine itself either as valid or invalid. If validity and
invalidity are equally intrinsic to knowledge, it must have
both at the same time.' The Naiyayikas reject the Saihkbya
view on the ground that it fails to account for the failure of
practical activities (pravrttitnsarhvdda). If the validity of
knowledge be self-evident, there cannot be unsuccessful
activity. If its invalidity be self-evident, there cannot be
any activity at all. The cognition of silver in a shell must
be either valid or invalid. If it is valid and known to be
valid by itself, then the act of picking it up should not lead
to disappointment. On the other hand, if it is invalid and
known to be invalid by itself, no one should strive to pick it

up. But illusions and disappointments are ordinary and

frequent experiences of life. Hence neither the validity nor
the invalidity of knowledge is intrinsic and self-evident.^

5. Criticism of the Bauddha theory of intrinsic invalidity

and extrinsic validity

According to the Buddhists, all knowledge is invalid by

its very nature The validity of knowledge consists in its

capacity to produce successful action. Hence prior to any

successful activity every knowledge is to be treated as in-
valid. We cannot say that validity belongs to knowledge
simply because it has come to be, or has appeared. In that
case, error willhave to be regarded as valid knowledge,
because error too appears as a form of knowledge. That
knowledge has been produced does not necessarily mean that
there is m it a true cognition of the object, since the

*SD. & 80 , pp 20-21 MSnomeyodaya,

: p. 76.

*NM.,p.l60;TC., I.p 184.

100 NYiYA THEORY OF KNOWLEDGE

knowledge is liable to contradiction. Hence we are to say

that invalidity belongs to knowledge at its inception and its

validity is due to the negation of invalidity by exter-

nal conditions. In fact, the question of truth and falsity

does not arise so long as we are concerned with mere belief.

We ma.y know things and believe in them without knowing
whether the belief is The question of the
true or false.
validity of knowledge arises first when a certain belief

is contradicted and we apprehend its invalidity. Hence m

logic we must start with the invalidity of knowledge.

From a sceptical or rather critical standpoint, the Bud-

dhists take all knowledge as intrinsically invalid and treat
validity as an extrinsic character which knowledge comes to

have by way of conative verification (pravrthsamarthya)

According to them, the truth of knowledge is constituted by
successful activity. Hence it follows that prior to success-
ful activity, knowledge is not-true. When any knowledge
leads to successful activity we know that it is not not-true,
i.e. it is true. So the Buddhists give a negative definition
of truth as what is not false {avisamvadakam) and conclude
that falsity is intrinsic and truth extrinsic to knowledge .

The Naiyayikas reject the Bauddha view of intrinsic

invalidity on the ground that it cannot account for unsuc-
cessful practical activity (pmcrttivisamvada) If the invali-
dity of knowledge be self-evident, why should a man run
after the false, knowing that it is false. Hence there can-
not be any practical reaction in connection with illusion.

Again, if the invalidity of knowledge be due to defects in

the conditions of knowledge and be known through contra-
diction, it cannot be held that it is intrinsically conditioned

and self-evident That invalidity is due to certain extra

conditions (karanadosa) must needs be admitted. Invalidity
is not merely the absence of validity, but a positive character

’ NET , pp 3 f
 TEST OF TROTH AND ERROR 101

of such forms of knowledge as doubt and error. Hence it

must be due to some positive conditions other than the

causes of knowledge. As such invalidity is not intrinsic to
knowledge.^

6. Gnhcism of the Mlmamsd theory of intrinsic validity

and extrinsic invalidity

According to the Mimamsa and the Vedanta, the valid-

ity of knowledge is due to conditions that are intrinsic to
knowledge, and the knowledge of validity is due to the
condition of knowledge itself. This means that knowledge
IS both made valid and known to be valid by its own intrin-
sic conditions.

It is the very nature of knowledge to reveal its object.

There is no knowledge which does not manifest the nature

of some object Hence it follows that knowledge requires
no other conditions than itself in order to reveal its object.

It cannot fail to give us truth if it is to be knowledge at all.

For how can it reveal its object without being true^ So w<
see that knowledge must be valid by its very nature, i.e.

the very conditions which condition knowledge must als(

condition its validity. It cannot be said that knowledge is

neutral cognition and that validity and invalidity are its

adventitious characters. There is no such thing as a neutral

or characterless cognition. Every cognition must be either
valid or invalid. There is no third alternative here. A cog-

nition which IS neither valid nor invalid is not a fact but a

fiction. As for doubt (samsuya), we are to say that it is not

a neutral cognition, but a form of invalid knowledge. Fur-

ther, if knowledge is not valid on its own account, it can

never be made valid on account of any external

condition. For the validating condition must itself be valid,

I NM , pp 160, 169 , 8D., p 21

102 NYIYA THEORY OF KNOWLMHGB

ated by other conditions, and these again by still othei

conditions and so on ad infinitum. This means that th«

validity of knowledge cannot be finally established. Hence

we must either say that knowledge is valid by its verj
nature or deny that there is any valid knowledge at all. The
validity of knowledge must thus be conditioned by the con-
ditions of knowledge itself, i.e. it must be intrinsic to

knowledge. The only external condition for validity is the

negative one of the absence of vitiating factors (dosdhhdva).
But this does not mean that validity is externally condi-
tioned, since the absence of a thing is not a positive factor
that contributes anything towards the validity of know-
ledge. The absence of vitiating conditions accounts for the
absence of invalidity and not for the positive fact of validity.
There is no evidence for any positive external condition
like special efficacy of the conditions of knowledge {karakati-
nktatadgataguna) So we are to say that the validity of
knowledge is due to the conditions of knowledge itself {sva-
’
rupasihitahetuja) ,
i.e. it is intrinsic to knowledge
Similarly, the validity of knowledge must be known from
the conditions of knowledge itself. A true knowledge is by
known to be true. It does not require anything else
itself

to show its truth. In fact, the truth of knowledge canno t

be known from any external condition. We cannot know
it from any special efficacy in the conditions of knowledge
(kdranaguna) There is no such thing as a special potency
of the causes of knowledge Even if there were such a
thing; it cannot be known, since it must pertain to the
sense-organs and, as such, must be supersensible. Nor can
we know validity from the experience of non-contradiction
(Jbddhakdbhavajmna) Non-contradiction cannot be a test

of the truth of knowledge. When we speak of non-contra-

diction we must mean either of two things, namely, that

1 VP., Oh VI ; 8D., pp. ao, 21 , 48.

 TEST OF TRUTH AND ERROR 103

there is no contradiction for the present or that there can-

not be a contradiction at any time. know that there But to

is non-contradiction for the present is not to know that the

knowledge must be valid. For what is not now contradict-

ed may be and often is contradicted in the future. That
there cannot be a contradiction at any time is what we can
never know unless we become omniscient Absolute non-
contradiction is therefore an impracticable test of truth. Norl
again can the validity of knowledge be known from itsi

coherence with some other knowledge {samvada). For thiaJ

will lead to infinite regress. How can we know that there is

real coherence between two cases of knowledge? How again
do we know that the second knowledge, with which the first

coheres and by which it is tested, is itself valid? To prove

the validity of the second we must show its coherence with
a third and so on ad infinitum Or, if we stop anywhere
and say that the last knowledge or the system of knowledge
as a whole has self-evident validity, there is no reason why
the first knowledge cannot be said to have the same self-

evident truth It cannot also be said that the coherence of

knowledge consists in the volitional experience of expected

results. The mere experience of desired objects is no evi-

dence ol the validity of knowledge. In dream we have vivid

experiences of many desired things. This however does not

make dream a valid knowledge at all. Thus we see that

there cannot be any extraneous test of the validity of

knowledge, like correspondence, coherence, non-contradiction
or pragmatic utility. Hence knowledge must test or

certify its own truths i.e. the validity of knowledge is

self-evident.^

As to invalidity or falsity, the Mimamsakas and the

Vedantists hold that it is extrinsic to knowledge.

While truth is organic to knowledge, falsity is accidental

» VP. A so., idKl ,

NM , pp. 162.66.
 104 NYiYA THEORY OF KNOWLEDGE

and externally conditioned. Thus the invalidity of knowledge

is due to some defects in the conditions of knowledge
(karanadosa). It is only when certain defects vitiate its

natural conditions that knowledge fails in its purpose,

namely, the attainment of truth. So invalidity is condi-
tioned by conditions other than those of the knowledge
itself. Similarly, the knowledge of invalidity depends on
such external conditions as the experience of contradiction

and the knowledge of vitiating conditions. This means

that knowledge per se is valid. If in any case it is rejected

as false, that is only because it fails to lead up to certain

expected results {viparitavyavaharaja), and not because it

tails to reveal its object. Every knowledge is intrinsically

valid. When any knowledge is, as a matter of practical

usage (vyavahara) called false, that is either because it is

contradicted by some other knowledge or because it is

perceived to arise from certain defects in its natural

conditions. The invalidity of knowledge is thus both
constituted and known by external conditions other than

the conditions of knowledge itself

In some cases it may so happen that when one knowledge

is contradicted by another, we may have some doubt with
regard to the second invalidating knowledge and require a
third knowledge to resolve the doubt. But even here we
are not involved in an infinite regress and bound to give up
the idea of the intrinsic validity of knowledge. If the
third knowledge is consistent with the first, then the
intrinsic validity of the latter remains established as before.
What the third knowledge does is not to validate the first

by reason of its coherence with it, but to dispel the false

doubt raised by the second contradicting knowledge. It,

on the other hand, the third is consistent with the second

it confirms the sense of contradiction and the first is known
• Ibtd.
 TEST OF TRUTH AND ERROR 106

to be invalid. Here the invalidity of the first is known

through something other than itself. That the third is
consistent with the second does not mean that it validates
the second. It means only that the doubt attaching to the
second is dispelled by its coherence with the third. Hence
it follows that every knowledge validates itself through
itself and is invalidated by some other knowledge. That
IS, validity is intrinsic to all knowledge, while invalidity is

extrinsic and accidental.'

The Naiyayikas reject the first part of the Mimamsa
theory, namely, that knowledge is both made valid and
known to be valid by -
its own intrinsic conditions. The
validity of knowledge cannot be due to the conditions of
knowledge as such. If that were so, there could not be any
invalid knowledge, since even invalid knowledge arises from
the conditions of knowledge. In fact a valid cognition is

more than a cognition as such. Hence it must be due to

some special character of the general conditions of knowledge
just as an invalid cognition is due to some positive factors

that vitiate the general conditions of knowledge. The mere

absence of vitiating factors cannot account for the
positive character of validity. Thus the validity of percep-
tion 18 due not merely to the absence of vitiating factors

like the diseased condition of the sense-organ, but to such

positive factors as the healthy condition of the sense-organ,

etc. Similarly, in all other cases the validity of knowledge

IS due to some special auxiliary conditions in the specific

causes of knowledge (kdranaguna). Such special conditions

may not be always perceived, but they may be known from

other sources, like inference and testimony. The special
efficacy of the sense-organs may be known from the medical
sciences. Further, if the validity of knowledge be due
simply to the absence of vitiating conditions, its invalidity

» NM.. W. 166-67.

14— (1117 B)
106 NYAYA THEORY OF KNOWLEDGE

may also be said to be due to the absence of efficacious

conditions. As such, we need not say that invalidity is due
to external conditions. Again, on the Mimamsa view, all

knowledge being intrinsically valid, the distinction between

truth and falsehood becomes insignificant. We should not
speak of any knowledge as invalid. It cannot be said that
when any knowledge turns out to be false, it altogether

ceases to be knowledge or cognition A wrong knowledge

is as good a cognition as a true one So, if cognition per se
be true, there cannot be any wrong cognition. But that

there are wrong cognitions, illusions, and hallucinations is

an undeniable fact. So it must be admitted that botli

validity and invalidity are externally conditioned.

Similary, no knowledge is by itself known to be valid,

i.e. the validity of knowledge is not self-evident Thus

the cognition of blue does not cognise its truth or validity
at the same time that it cognises the blue colour. It does
not even cognise itself immediately as a cognition of

blue, far less as a valid cognition of blue. On the Bhatta

view, a cognition is not immediately cognised, but is known
mediately by inference If so, the validity of knowledge can
not be immediately known by itself. Nor can we say that
with every cognition there follows immediately another
cognition which cognises the validity of the first. With the
perception of blue, for example, we do not find another
cognition immediately following it and cognising its validity

or invalidity. There is no introspective evidence for a

secondary cognition of validity appearing immediately after
the primary cognition of an object. Even if there were such,
the validity of knowledge will not be self-evident but
evidenced by another knowledge. Further, if the invalidity
of knowledge be known from its contradiction, we are to say
that its validity is known from the absence of contradiction.

1 NM.,pp 170-71; BM., 181-86,

 tEST OF TRUTH AND ERROR 107

On the other hand, if we accept the Prabhakara view

that knowledge is self-manifest and guarantees its own
truth without reference to anything else, we do not
see how there can be doubt and suspicion, or how
there can be any failure of practical activity. Since
validity is inherent and self-evident in knowledge, every
knowledge must carry m it an assurance of its truth and
we should have no doubt. Similarly, every knowledge
being true and known to be true by itself, there cannot
be any disappointment in practical life. But doubts and
disappointments are very common experiences of life. If it

be said that doubt ari-ses out of contradiction between two

cognitions and is resolved by a third cognition, we are
forced to give up the idea of self-evident validity. The
third cognition may not constitute the validity of the first

by reason of its coherence with it, but it at least condi-

tions our knowledge of its validity So tlie validity of one

knowledge is known by another knowledge. In fact

knowledge only reveals its object. To know that it is

valid, i.c. it truly reveals the object, we must have some
extraneous test like coherence with volitional experience
or some accredited past knowledge. Hence the validity

of knowledge must be known from external conditions.^

It will appear from the above discussion that the
Nai}ayikas are not prepared to accept the theory of intrinsic
validity. That truth is intrinsic to and self-evident in all

knowledge is not admitted by them. But that the truth

of some cases of knowledge is self-evident is admitted by
some Naiyayikas. There is on this point a difference of

opinion between the ancient and the modern exponents of

the Nyaya. The older Naiyayikas insist that a proof of
the validity of any knowledge requires the exclusion of other
suggested possibilities contrary to it. Hence we find that

1 Ibtd.
108 NliYA THEORY OF KNOWLEOOE

tDey try to establish the validity of every knowledge by

external grounds, even at the risk of an infinite regress.

To prove the knowledge we need not go beyond

validity of

perception, inference, comparison and testimony. But

one perception or inference must be proved by another
individual perception or inference. Hence the same
knowledge may sometimes be proved by another and some-
times taken to prove another (pradlpapraka^avat). This
seems to be a more empirical and common-sense view of the
validity of human knowledge.'
Later Naiyayikas however do not insist that every
knowledge must be tested and proved before we can accept
its validity. According to them, the validity of knowledge
need not be proved if there is not the slightest doubt about
it (sam^aydhhdva). Its validity is practically self-evident

so long as it is not contradicted. A motiveless doubt of a

possible contradiction is of no account. To say that a
knowledge is evidently valid it is not necessary to prove its

infallibility or to exclude all other possibilities contrary

to it. So if there be such knowledge as cannot reasonably
be doubted we are to say that it has self-evident truth.
Of course, when any doubt or dispute arises as to its truth,
we must prove it by some extraneous test. Among such
cases of knowledge the Naiyayikas include logical inference
and comparison, and cognition of the resemblance between
two cognitions (jMnagatasddr^yajmna), cognition of cog-
nition or self-consciousness {anuvyavasdya), and cognition
ofanything as something or as a mere subject (dharmi-
jmna). Inference and comparison (upamdna) as logical
methods are based on some necessary relation between two
terms, namely, a mark and the marked (hetu and sddhya),
a name and its denotation {samjna and (samjnl) respectively.
Hence knowledge by inference and comparison is known as

• NB , 2, 1, 17-19.
 TEST OE TROTH AND ERROR 109

necessary knowledge. These become the content of self-

conscious knowledge and, as such, their validity is self-

evident. Similarly, the cognition of a cognition or aware-
ness of awareness, the cognition of the similarity between
cognitions and the cognition of anything as a mere subject
are all cases of self conscious knowledge. In these we not
onlyknow something but also know that we know it.
That IS, we know that something is known. Hence these
cases of knowledge also are necessary knowledge having
self-evident validity.^
In the case of sense-perception and testimony, how-
ever, there cannot be any self-evident validity. These are
not based on any necessary relation between two terms.
There is no necessary relation between sense-perception and
its object or between words and their meanings. Hence
we cannot say that to know anything by external percep-
tion or testimony is also to know that we know it. These
cannot be the content of self-conscious knowledge and
their validity is not self-evident.*
Ii should however be noted here that with the
Naiyayikas the self-evident validity of some cases of know-
ledge does not exclude their liability to error. For the
Mimamsaka and the Vedantist, the self-evident character
of a truth means its infallibility which excludes the possi-

bility of any falsification, so that error pertains not to truth

but to its applications {vyavahara). For the Naiyayikas,
however, even necessary truths are empirical and so require
confirmation by fresh applications {i.e. •pravrttisdmarthya)

whenever necessary. But they are different in status from

ordinary observation and generalisation. They possess the

1 Anumanasya nirastasamastavyabhicara^ai^kasya svata eva prama^ya-

maDUUQeyavyabhicarilifiga'jamutthatvat, etc., NVT., pp 12-13; anamanopamfiD&-

nuvyavaa&yadhaniiijfianftiianiapi . svata 6va pramitpyagraha^ etc., NVTP., pp,

119-20; vide al-o TC..i,pp 277-79,282-84

* Pratyak§a4abdavijfiftnayonia Bvato Vyabhicaragraha iti, etc. NVT ,
110 KYiYA TflEORY OF KNOWLEOQF

highest degree of certainty which is humanly attainable.

Still they may change if, as Hobhouse has said, the constitu-

tion of the world changeth. Hence the Nyaya conception

of the self-evident validity of some knowledge is different

from the Mlmamsa and Vedanta view of the self-evident

validity of all knowledge.

7 . Indian and Western theories of truth

Here we propose to examine the Indian theories uf

truth, as explained above, in the light of parallel Western
theories. With regard to truth there are two mam ques-
tions, namely, how truth is constituted, and how truth is

known. The first question relates to the nature of truth

and the answers to it give us the definitions of truth The
second question refers to the ascertainment of truth and
the answers to it give us the tests or criteria of truth.
With regard to these two questions there seem to

be two possible answers Thus it may be said that truth

is a self-evident chnracter of all knowledge Every know-
ledge 18 true and known to be true by its very nature
Knowledge does not depend on any external conditions
either to be made true or to be known as true This is

the theory of the intrinsic validity (svatah pramdnya) of

knowledge as advocated by the Samkhya, Mimarnsa and
Advaita Vedanta systems of Indian Philosophy. According
to the last two schools, the truth of knowledge consists
just in its being uncontradicted {ahadhita) The absence
of contradiction, however, is not a positive but a negative
condition of truth. Knowledge is both made true and
known to be true by its own internal conditions It is

only falsehood that is externally conditioned. Bo truth

is self-evident, while falsity requires to be evidenced by
external grounds. The Samkhya goes further than this.
 TEST OF TRUTH AND ERROR 111

It maintains that both truth and falsehood are internally

conditioned and immediately known, i.e. are self-evident.
There is no exact parallel to the above theory of truth
m Western philosophy. It is true that in modern
European philosophy knowledge, in the strict sense, is

always taken to mean true belief. But truth or validity

IS not regarded as intrinsic to all knowledge, independently
of all external conditions. It is in the writings of Professor

L. A. Eeid, a modern realist who owns no allegiance to

the current schools of realism, that we some approach
find
to the view that truth is organic to knowledge. But even
Reid makes it conditional on knowledge efficiently fulfilling

its function, namely, the apprehension of reality as it is

He thinks that truth is nothing else but knowledge doing

its job. Thus he says “Truth is, indeed, simply, . .

the quality oi knowledge perfectly fulfilling its functions.’’

Again he observes “If knowledge were not transitive,

if we were not in direct contact, joined with reality, then

all our tests, coherence, correspondence, and the rest,
’’
would be worthless *
Here truth is admitted to be a
natural function of knowledge, but not as inherent and
self-evident in all knowledge. In the theory of intuition-
ism, we find a close approach to the view ol self-evident validi-

ty To the question How do we know that a belief

‘
is true

or valid ? ’
mtuitionism has a simple answer to give,

namelj, that we know it immediately to be such. As

Hobhouse puts the matter “ Intuitionism has a royal

way of cutting this, and indeed most other knots . for

it has but to appeal to a perceived necessity, to a clear

idea, to the inconceivability of the opposite, all of which

may be known by simply attending to our own judgment,
and its task is done.’’ ^ Among intuitiomsts, Lossky

has made an elaborate attempt to show that truth and

186, 199, 204.

1 L A. Reid, Knowledge and Truth, pp
* Hobhouse, Theorff of Knowledge, p. 488
112 NYiYA THEORY OF KNOT^PLEDGB

falsity are known through an immediate consciousness of

their objectivity and subjectivity respectively. For him,
truth is the objective and falsity the subjective appearance
of the object. But how do we know that the one is objec-
tive and the other is subjective ? The answer given by
Lossky as also by Lipps is that we have “ an immediate
consciousness of subjectivity ” and “an immediate conscious-
ness of objectivity.’’ To quote Lossky’s own words :

“It is in this consciousness of objectivity and subjectivity,

and not ... in the laws of identity, contradiction,
and excluded middle, that our thought has a real and
immediate guide in its search for truth.’’
It should be remarked here that the above theories
of self-evident truth or intrinsic validity give us a rather
jejune and untenable solution of the logical problem of
truth. They leave no room for the facts of doubt and
falsehood in the sphere of knowledge. But any theory
of truth which fails to explain its correlate, namely, false-

hood, becomes so far inadequate. Further, it makes a

confusion between psychological belief and logical certainty.
Psychologically a wrong belief may be as firm as a right
one. But this does not mean that there is no distinction
between the two. Subjective certitude, as such, cannot
be accepted as a test of truth. It is true that the theory of
intrinsic validity does not appeal to any test of truth
other than the truth itself. It assumes that the truth
of knowledge is self-evident, and that we cannot think
of the opposite. In fact, however, there is no such self-

evident truth. It is only in the case of the self that we

can speak of self-evidence in this sense. The self is a
self-manifesting reality. It is manifest even m any doubt
or denial of its reality. Hence self-evidence belongs
really to the self only. It is on the analogy of the self

Losskjt The Intuitive Bam of Knowledge, pp. 237*29,

 TEST OF TRUTH AND ERROR 113

that w© apeak of the self-evideoKje of any other truth. A

troth is self^rideut in so far as it has the evidence ol
the self or is evident like the self. But as we have
said, there is no such self-evident truth other than the
self itself. In the case of any other truth, we can always
think of the opposite in a sensible way. That ‘
two and
two make five’ is not as nonsensical as ‘abracadabra.’
Even if the opposite of a certain belief be inconceivable,
it does not follow that the belief is infallible. What was
once inconceivable is now not only conceivable but perfectly
true. Hence we c.innot say that self-evident validity is

intrinsic to all knowledge.

The second answer to the question How ‘
is truth
constituted and known T leads us to the theory of extrinsic
validity (paratah pramdnya) According to this, the
truth of any knowlenge is both constituted and known by
certain external conditions. As a general rule, the validity

of knowledge is due to something that is not inherent

in it. So also the knowledge of validity depends on certain
extraneous tests. Validity is thus assigned to one knowledge
on the ground of some other knowledge. This is the theory
of extrinsic validity as advocated by the Nyaya and the
Bauddha systems. In Western philosophy, the correspon-

dence, the coherence and the pragmatist theories of truth

all come under the doctrine of extrinsic validity. In each

of them the truth of knowledge is made to depend on
certain external conditions other than the knowledge
itself. According to almost all realists, old and new,
it is correspondence to facts that constitutes both the

nature and the test of truth.' Of course, some realists

differ from this general position and hold a different

view of the matter. Thus Alexander* makes coherence the

» Vide Tht New Realism and Essays m Critical Bealum,

* ipmUt Tim# and Detty, Vol. H, pp. 881 f.

16— (1117B)
 114 NYiYA THEORY OF KNOWLEDGE

ground of truth. But in speaking of coherence as determin-

ed by reality, he accepts indirectly the theory of corre-

spondence. Reid/ on the other hand, treats correspondence
to the given only as a test of truth RusselP defines truth in
terms of correspondence and accepts coherence as a test of
some truths, while others are said to be self-evident. In the
philosophy of objective idealism,® coherence in the sense of the
systematic unity of all experiences IS made both the ground
and the test of truth. The truth consists in the coherence of
all experiences as one self-maintaining and all-inclusive sys-
tem. It is in this sense that Bosanquet* says that ‘
the truth
is the whole and it is its own criterion. Truth can only be
tested by more of itself.’ Hence any particular knowledge is

true in so far as it is consistent with the whole body of ex-

perience. On this view, the truth of human knowledge be-
comes relative, since coherence as the ideal of completed ex-
perience IS humanly unattainable. For pragmatism,® truth is

both constituted and known by practical utility. The truth

of knowledge consists in its capacity to produce practically
useful consequences. So also the method of ascertaining
truth is just to follow the practical consequences of a belief

and see if they have any practical value. With this brief

statement of the realistic, the idealistic and the pragmatist

theories of truth, we proceed to examine the Buddhist and
the Nyaya theories of extrinsic validity.
From what we have said before it is clear that the
Buddhists adopt the pragmatist theory of truth and reality.

For them, practical efficiency is the test of both truth and

reality. The real is what possesses practical efficiency

* Knowledge and Truth, Chap VIII

* Problems of Philosophy, Chapa XII, XTII , Our Knowledge ""of the External
World, p. 68 , The Analysts of Mind, p 166.
* Vide Joachim, The Nature of Truth, Chap HI
< Logui,Vo\ n, pp 266-67.
* James, Pragmatism, L^t, VI; Perrjr, Present Philosophical Tendencies,
Ptf. IV, V,
 tEST OF TilUtfl AND ERROR 11§

{drthaknya) and the true is the useful and so practically

efficient (arthakriyasamarthya). But the pragmatic concep-
tion of truth 18 embarrassed by serious difficulties The
Nyaya criticism of the Bauddha conception of pramana has
brought out some of these difficulties. Here we may note
that to reduce the true to the useful is to make it almost
meaningless. It is by no means the case that truth is only
a matter of practicil utility The atomic and the electron
theories oT matter make very little difference in our practical
life. Similarly, the different theories of truth involve no
great difference in their practical consequences. But in the
absence of any other test than that of practical utility we
cannot say which one is true and which is false. Further,
there are certain beliefs which are admittedly wrong but
which are otherwise useful for certain purposes of life. But
no one would claim any truth for a wrong belief on account
of its practical utility. Hence the Buddhist and the
pragmatist theories of truth cannot be aceepted as sound and
satisfactory.

The Nyaya theory of truth, it will be seen, combines

the correspondence, the coherence and the pragmatist
theories with certain modifications. According to it, the

truth of knowledge consists in its correspondence with

objective facts, while coherence and practical utility are the

tests of truth in such cases m which we require a test. It

defines the truth of all knowledge as a correspondence of

relations {tadcati tatprakdraka) To know a thing is to

judge it as having such-and-such a character. This know-

ledge of the thing will be true if the thing has really such-

and-such a character ;
if not, it will be false. The Nyaya
view of correspondence is thus different from the new
realistic idea of structural correspondence or identity of

contents.^ That knowledge corresponds to some object does

1 Cf. Chapter III, Sec 3, shove.

 J«fi¥A THEORY OF KNOlVLRl)<jiF

toot, for the Naiy&yika, mean that the contents of the object
bodily enter into consciousness and become its contents.

When, for example, I know a table, the table as a physical

existent does not figure in my consciousness. This means
only that I judge something as having the attribute of
‘
taWeness ’
which really belongs to it. There is a subjec-
tive cognition of a physical object. The one corresponds
to the other, because it determines the object as it is, and
does not itself become what it is. If it so became the
object itself, there would be nothing left on the subjective
side that might correspond to the physical object. Nor again
does the Nyaya follow the critical realist’s idea of correspon-
dence between character-complexes, referred to the object by
the knowing mind, and those actually belonging to the object.
When we know anything we do not first apprehend a certain
logical essence or a character-complex and then refer it to

the thing known Our knowledge is in direct contact with

the object. In knowing the object we judge it as having a
relation to certain characters or attributes. Our knowledge
will be true if there is correspondence between the relation
asserted in knowledge, and that existing among facts. Thus
my knowledge of a conch-shell as white is true because there

18 a real relation between the two corresponding to the

•redation affirmed by me On the other hand, the perception
of silver in a shell is false because it asserts a relation
between the two, which does not correspond to a real rela-

tion between them.^

Wnile truth consists in correspondence, the criterion
of truth is, for the Nyaya, coherence in a broad sense
{sainvdda). But coherence does not here mean anything

1 Cf
**
8fiiith*s jodgmeot that it » the light uf a ahip is true jtjat because *
it/
light, it ID fact BO related to a real ship Jones' judgmeDt (that it it tht
light of a star), ofi the other haod, is false, because this thought if oot ao
apprehemton of the existing preseut complex fact. light>belongiDg-to-Bbip."-*Beid,
Knowledge and Truth, pp 889-10
 tEST OE TRUTH AND BRBOfe IjL^

of the kind that objective idealism means by it. The Nyaya

coherence is a practical test and means the harmony be-
tween cognitive and conative experiences (pravrttisdmar-
thya) or between different kinds of knowledge (tajjdtiyatva).
That there is truth in the sense of correspondence cannot,
as a general rule, be known directly by intuition. We
know it by inference from the fact that the knowledge in
question coheres with other experiences of the same object
as also with the general system of our knowledge. Thae
the perception of water is known to be valid when different
ways of reaction or experiment give us the same experience
of water. It is this kind of coherence that Alexander
accepts as a test of truth when he says: “If truth is

tested by reference to other propositions, the test is not one

of correspondence to reality but of whetlier the proposition
tested IS consistent or not with other propositions.”'
Hobhouse* also means the same thing by ‘consilience’ as
'’a measure of validity. According to him, validity belongs
to judgments as forming a consilient system. Of course,
he admits that such validity is relative and not absolute,
since the ideal of a complete system of consilient judgments
is unattainable. The Nyaya idea of samvada or coherence
may be better explained as a combination of Eeid’s methods
of correspondence and coherence. If we take the judgment
‘
that is the light of a ship,’ we can test its truth by what
Eeid calls the correspondence method “of approaching the
light and seeing a ship.” This is exactly what the Nyaya
means by pravrttisdmarthya or successful activity. Or, we
can employ, so says Reid, the cheaper coherence method

“of comparing this knowledge with other kinds of knowledge

and see if it is consistent with them.” * In this we have
the Nyaya method of testing one knowledge by reference to

1 Space, Time and Deity^ Vol. H, p 252

• The Theory of Knowledge, pp. 499-500
® Knowledge and Truth, pp. 208-4, 211-12
118 NYiYA THEORY OF KNOWLEDGE

some other valid knowledge {ta]]dtlyatva). But the Nyaya

goes further than this and accepts practical utility also as a
test of truth. Thus the validity of the perception of water
may known from correspondence and coherence in
be the
above sense. But it may be further known from the
satisfaction of our practical needs or the fulfilment of our
practical purposes in relation to water, such as drinking bath-

ing, washing, etc. But the Nyaya never admits the pragmat-
ist contention that the truth of any knowledge is constituted
by its utility or serviceableness. Knowledge is made true

by its correspondence to some reality or objective fact. It

IS true not because it is useful, but it is useful because it

IS already true Hence truth consists in correspondence and

is tested bv coherence and practical efficiency
But from the standpoint of the modern Nyaya, all truths
do not require to be tested. Some truths are known as
such without any test or confirmation. These are mani-
festly necessary and so self-evident truths. Here the Nyaya
view has some affinity with Itussell’s theory of truth. ^
In
both, truth is defined by correspondence to fact, but in
different ways. Although truth is thus externally condi-
tioned, some truths are admitted by both to be self-evident.

For the Nyaya, however, such truths are only necessary

truths or what Russell calls a pnon principles Of the
different kinds of knowledge by acquaintance — sensation,
memory, introspection, etc. — which are admitted by Russell
to have self-evident truth, it is only introspection or self-

consciousness (anuvyavasdya) that is admitted by the Nyaya

as having self-evident validity. The validity ot self-

consciousness is self-evident because there is a necessary

relation between consciousness and its contents. When I
become conscious of a desire for food, I find that my

consciousness is necessarily related to the desire, it is the

1 Problems of Philosophy^ Chaps. XI, XII, XIII.

 TEST OF TRUTH AND ERROR 119

desire itself as it becomes explicit.^ Here I not only know

something, but know that I am knowing it, i.e. the truth
of my knowledge is self-evident.
The different theories of truth discussed above may be
shown to supplement one anotlier and be reconciled as
complementary aspects of a comprehensive theory. The
first requisite of such a theory is the independent existence
of a world of objects If there were no such world, there
would be no ground for the distinction between truth and
falsehood Some of our beliefs are true or false according
as they are or are not borne out by independent objects or
facts. It IS because there aie certain independent objects,
'

to which our beliefs may or may not conform, that we

distinguish between truth and error. Hence we say that
truth consists in the correspondence of our knowledge with
independent objects or facts. The difficulty on this view,
it is generally remarked, is that if the objects are independent
of knowledge, we cannot know whether our knowledge
corresponds with them or not How can we know what is
outside and beyond knowledge, and see that true knowledge
agrees with it? The reply to this is that in the case of

external objects, physic il things and other minds, we can-

not straightway know the correspondence between our
knowledge and its we cannot deny the
objects. Still,

reality of these external objects. But for the independent

existence of other things and minds we cannot explain the

order and uniformity of our experiences and the similarity

of the experiences that different in lividuals may have under

similar circumstances. That some of our experiences

represent the real qualities of things may then known

be
from the fact that they are given in the same way to

1 C/ C Hartshorne’s article m The Momet (Vol XLTV, No 2, p. 171) “ Must

this (feeling) not be admitted to present an obvious dual aspect of being at once
subjective and yet a content or object of consciousness, at once a mode and a datijin
**
pf av^arepess ?
120 NYiYA THEORY OF KNOWLEDGE

different persons, or to the same persons through different

senses. As Professor Price has shown, “ sense-data

cohere together in families, and families are coincident with
physical occupants." ’
On the other hand, some of our
experiences are not taken to represent the qualities of things,
because they do not cohere with other experiences of the
same individual or of different individuals. The first kind
of experiences is considered to be true and objective, while
the second is judged to be false and subjective. Similarly,
our knowledge of other minds is true when it correctly

represents the contents of those minds It will be false,

if what we impute to them forms no part of their actual

contents. This shows that it is correspondence to facts
that constitutes the nature of truth, although we cannot
directlyknow such correspondence m the case of physical
things and other minds. To know this we have to consider
if one knowdedge coheres with others or the whole body of
human knowledge, and also consider if we can successfully
act on our knowledge. What is true works, although
whatever works is not true. Thus we know the correspond-
ence of knowledge with facts from its coherence and
pragmatic value. But to know that a certain knowledge
corresponds with ficts is to know its truth It does not

constitute its truth. The knowledge becomes true if, and

only if, it corresponds with facts. We know or test its

truth when we find that it is coherent with other parts of

our knowledge and our practical activities. So truth is

constituted by correspondence with facts and is tested by

coherence and practical activity.
The Vedanta view of truth as uncontradicted experience
logically implies the coherence theory of truth. That some
experience is uncontradicted means that it is different from
the contradicted. But to be different from the contradicted

^ Cf. Perception, p 802,

 TEST OF TRUTH AND ERROR 121''

means to belong to the body of coherent knowledge. We

do not and cannot rightly judge an experience to be un-
contradicted unless we relate it to other experiences and
find that congruous with them. A dream experience
it is

is wrongly judged by the dreamer to be uncontradicted and

true, because he cannot relate it to his waking experiences.

It cannot be said that a dream experience is true for the
time being and becomes false afterwards. What is once
true is always true. A dream experience may sometimes
be judged to be true, but it is really false for all time.
And its falsity appears from its incoherence with waking
experience. Hence we -are to say that an experience is

really uncontradicted when it is rel.ited to other experiences

and is found to be coherent with them
It may be urged against the above view that truth
consists in correspondence and is tested by coherence, that
it either assumes the truth of the testing knowledge, or
must go on testing knowledge ad infinitum. If knowledge is

true when it corresponds with facts, and if the correspond-

ence cannot be directly known, then the truth of every
knowledge must be tested by its coherence with others.

This, however, means that there can be no end of the

process of proving knowledge and, therefore, no final proof

of any knowledge. To solve this difficulty we must admit
that there is at least one case in which knowledge is, by
itself, knowm to be true. We have such a case in self-

consciousness. While the truth of all other knowledge is

to be tested by coherence, the truth of self-consciousness is

self-evident and requires no extraneous test. The self is a

self-manifesting reality. Hence the contents of our mind
or the self are manifested by themselves. They are at

once existent facts and contents of consciousness. To

become conscious of the contents of one’s mind is just to

make them explicit. What we are here conscious of are

not outside or beyond consciousness. Mental contents not

16--(1117B)
 NYiYA THEORY OF KNOWLEDGE

only are, but are conscious of themselves The state of

knowledge and the^object of knowledge being identical, we

cannot strictly speak of a correspondence of the one with
the other. Or, if we speak of a correspondence between
them we are to say that it is, directly known and so need
not be known or tested in any other way. When we feel

pain, or know something, or resolve to do anything, we may

be conscious of feeling it, or knowing it, or resolving to do
it. What we are here conscious of as objects are the objects
themselves as they become explicit or conscious of them-

selves. Such knowledge is, therefore, not only true, but

also known to be true by itself. Hence we admit that the
truth of self-consciousness is self-evident, while all other
truths are evidenced by external tests like coherence and
pragmatic utility.

122
 BOOK 11

PERCEPTION AS A METHOD OF KNOWLEDGE

(PRATYAKSAPRAMINA)
 CHAPTER VI

THE DEFINITION OF PERCEPTION

1. Primacy of perception over other methods of knowledge

Perception (pratyaksa), like inference and the rest, is

taken to mean both a form of \?alid knowledge (pramiti) and

the means or method of acquiring such knowledge (pramdna).
We are here concerned with them as methods of knowledge.

According to the Nyaya, there are four distinct and inde-

pendent methods or sources of knowledge, namely, percep-

tion, inference, comparison and testimony. Of these, per-

ception comes first and is the most fundamental. Of

course, the primacy of perception over the other methods of

knowledge is not due to anything in the nature of the

objects of knowledge. So far as the objects of knowledge

are concerned, the methods or ways of knowing stand on the
same footing in respect of their value and importance.

While there are certain objects which may be known by

any of the four methods, there are other objects which must
be known by a particular method and cannot be known by
any other. The existence of fire at a distant place may be

known from the testimony of a reliable person. It may

also be known by inference from the observed smoke as a

mark of fire. Or, if we take the trouble to go up to the

place from which smoke issues forth, we have a perception

of the fire on the spot. Hence with regard to such objects

as the fire, one method of knowledge is as good and valid as

any other. Contrariwise, there are certain cases in which

a truth is to be known by some special method. Scriptural

126 NYiYA THEORY OF KNOWLEDGE

testimony is the only source of our knowledge about supra-

mundane realities. That there is a future life, a heaven or
a hell, is to be believed on the authority of the scriptures and
cannot be proved by perception or inference. Similarly, our
knowledge of the unperceived cause of a perceived effect is

derived neither from testimony nor perception, but from

inference. Likewise, perception gives us the knowledge of
what is directly present to sense and we do not require any
inference or testimony for a knowledge of it. In relation to
the objects of knowledge, therefore, the methods are some-
times interchangeable and sometimes exclusive With
regard to any method of knowledge it may be said that some
contents of knowledge or some truths can only be given by
it, while others may come from this as well as from the
other sources of knowledge. Hence, so far as the contents

of knowledge are concerned, there is nothing to distinguish

one pramana or method of knowledge as more fundamental
'
than any other
Nevertheless there are certain important considerations in
favour of the Naiyayika view that perception is the most
primary and fundamental of all the sources of knowledge
recognised in any system of philosophy In the first place,

perception is the ultimate ground of all knowledge It is

true that all knowledge does not arise from perception The
empiricists including the Carvakas are wrong when they
suppose that sense-perception is adequate to explain the

entire body of human knowledge. According to the Nyaya,

perception is not the only '-oiirce of our knowledge, but it is

the basis of the other .sources or methods ol knowledge

Hence it has been said that all the other methods of know-
ledge presuppose perception and must be based on know-
ledge derived from perception.^ J S. Mill recognised this

1 NB , 1. 1. 3
2 Sdirvapramaijaaam pratyak^apurvakatvat, NVT ,113,
 THE DEFINITION OF PERCEPTION 127

truth when he said that “the truths known by intuition are

the original premises from which all others are inferred.’’ ^

For the Nyaya, however, perception is the basis on which

w'e have a knowledge ol other truths by inference as well as
by comparison and testimony. Inference as a method of
knowledge depends on perception The first step in infer-

ence IS the observation of a mark or the middle term (hnga-

dai Sana) ,
and the observation of the relation between the
middle and the major term. Hence, infeience is defined as
that knowledge which must be preceded by perception
^
{tatpurvakam) Similarly, upamana or compaiison as a
method of naming depends on perception of the points of
similarity between two objects So also sabda or testimony
IS dependent on perception inasmuch as the first step in it is

the visual or auditory perception of written or spoken words,

and such woids must come from a person who has a direct
or intuitive knowledge of the truths communicated by him.
So we see that perceptual knowledge is the ultimate ground
of all other knowledge by inference, comparison and
testimony.®
Secondly, the Naiyayikas observe that perception is the
final test of all knowledge We may question the truth of
the knowledge derived from infeience, testimony, etc., but
the truth of peiception is in a way beyond question. We
may know the same thing by means of testimony, inference
and perception But while the knowledge from testimony
and inference requires confirmation by perception, the per-
ceptual knowledge of the thing is in need of no further
confirmation A man may acquire the knowledge of a thing
from some authoritative statement, i e. the testimony of a

person. But this generally serves as the starting point for

further reflection on the matter and produces a desire to

> System of Logic, p 3

* NS & NB ,
1 1 5

3 NB &NV,1.1 6-7
12S NYiYA THEORY OF KNOWLEDGE

ascertain the truth in a more convincing way, say, by a

process of inferential reasoning. But the inference which
confirms the testimony may, in its turn, require further veri-
fication by way of direct perception of the object. Percep-
tual verification is thus the final test of all other knowledge
and, as such, perception is the most important of all the
methods or sources of human knowledge.’
In European philosophy the validity of perception
as a source of knowledge is rather overstressed by
the empiricists and some modern realists. According
to them, the truth of perception is unquestionable and
Thus S Mill remarks • “ Whatever
self-evident. J. is

known to us by consciousness (intuition), is known beyond

possibility of question. What one sees or feels, whether
bodily or mentally, one cannot but be sure that one
sees or feels. So also W. T. Marvin thinks that “ per-
ception is the ultimate crucial test, and as such, it does not
presuppose its own possibility. It simply is ;
and the
man who questions it assumes it in order to do the ques-
tioning.”” Similarly, Russell tells us repeatedly that the
truths of perception are self-evident truths, for which we
require no test at all.’ The Naiyayikas, however, do not
admit that the validity of perception as such is self-evident

and unquestionable. That perception is the final test

of all other knowledge does not mean that the truth of

perception is self-evident or that it cannot but be true.

From the standpoint of common-sense ^to ulism they

grant that, under normal conditions, whAv-'is directly

perceived is not doubted and so need not be further proved

or tested. When however any doubt arises with regard

1 Sa ceyaiii pramiti^ pratyaksapara, NB., 1.13,

2 System of Logic, p. 4.
3 The New Reahsm, pp 66-67.

* Problems of Philosophy, Chapter XI » Our Knowledge of the Eicfernal World,

 THE DEFINITION OF PERCEPTION 129

to the validity of perception, we must examine and verify

it as much as any other knowledge.

2. The Buddhist definition of pefception

The Buddhists define perception as the unerring cog-

nition of a given sensuin in complete isolation from all

ideata.* In it the object of cognition is a unique indivi-

dual {svalaksana) and the process of cognition is a*" mere

sensing without any element of ideation (kalpana) in it.

Vasubandhu, a Bauddha logician of the Yogacara school

(circa 480 A.. D ), characterises perception as a cognition
that IS directly produced by the object, of which it is the
cognition.^ The cognition of fire, for example, is a per-
ception, if and in so far as it is produced and wholly
conditioned by fire as an existent fact On the other hand,
the ‘
cognition of silver ’
in the presence of an oyster-shelly
cannot be called perception, since it is not caused by the
object, of which it claims to be the cognition. The silver is

non-existent at the time and place at which it seems to

be perceived and so cannot causally determine the cogni-

tion in question. So also the inferential cognition of fire

is distinguished from perception by the fact that it is

not produced directly and exclusively by fire as an object-

ive fact. The inferential knowledge depends on such
other conscious and unconscious conditions as the cognition
of smoke, the association between smoke and fire, memory
of the relation between the two and so on. For the same
reason, the Buddhists deny the perceptual character of

the so-called perceptions ot individual objects like the

jar, tree, etc. {samortijMna Y What we directly perceive

1 Prdtyak^arii kalpana po(}li imabhranfcam Nydyabmdu, Chapter I

2 Taio'rthadvtjMnaih pratyakfam^ NV, & NVTK., 114,

3 Ibtd.

17— (1117B)
130 NYiYA THEORY OF KNOWLEDGE

IS not the jar or the tree as a unity of the universal and

the particular, but some quality or part of it. What

is thus directly sensed is next combined with certain
images and ideas of other associated qualities or parts and
thereby produces the complex cognition of a jar or a tree.

In fact, such complex cognitions (satnvrtijndna) are not

perceptions, since these are not directly produced by the ob-
ject alone. Rather, they are wrong cognitions based on the
hypothesis of universal essences (]dtt) underlying the aggre-
gates of parts and qualities constituting individual objects
Dignaga, the greatest Bauddha logician (circa 500
A. D ), brings out the implications of Vasuhandhu’s defi-

nition of perception. If perceptual cognition is solely

determined by its object, it must be wholly given and not

anywise constructed by the mind Hence Dignaga defines

perception as a cognition which is not at all subjectively

determined and is not modified by ideas or concepts

(kalpana).' The concepts of name, class, quality, action
and relation do not enter into the perception of an object.
What is perceived by us is a unique individual that does
not admit of any description by concepts and words It

13 just what it IS immediately sensed to be Words and

concepts express such aspects of things as are general or
common to many things. But a thing is an individual
in so far as it excludes all other things from within itself.

Hence what is individual is to be directly felt or intuited,

but not expressed by words or concepts From this it

follows that perception is just the cognition of an immedia-

tely given datum and is completely free from all subjective
or conceptual determination It is a pure sensation
which cannot be properly described or embodied in verbal
judgments.^

^ Vide PramaQasamuocaya, Chapter I

* NBT., pp 7-12.
 THE DEFINITION OF PERCEPTION 131

The Buddhist
definition of perception has been criti-
cisedand rejected by the Naiyayikas It has been pointed
out by them that Vasubandhu’s definition of perception
is too wide. If by perception we are to mean a cognition
which IS objectively determined (tato’ rthadvijnanam) all

true knowledge will have to be regarded as perception.

As Bosanquet has rightly pointed out, “ reality is operative
in truth.” Thus a true inference has an objective basis
^

in so far as the conclusion expresses a real relation between

two things. So we may say that what is validly inferred
is an objective fact which is causally efficient towards the
inferential cognit ion •
Similar is the case with the other
kinds of valid knowledge Even the wrong cognition of
silver in a shell is not without some objective basis. The
wrong judgment, ‘that is silver,’ is based on the ‘
that’ as
an objective fact Further, on the Bauddha view of uni-
versal momentariness (hsanihavada) we do not see how
perception can have an objective basis. The object being
the cause of perception must be antecedent to it. So when
the perception is or appears, its momentary cause, namely,
the object, must cease to exist The object cannot therefore
be the cause of perception. But if perception be not
directly produced by the object, we cannot call it perception
at all.^

Later Bauddha logicians like Dignaga, Dharmakirti

and others reduce perception to a mere sensation free from
all conceptual determination. This, the Naiyayikas think,
IS logically indefensible and arbitrary. None of our ordinary

perceptions is a pure sensing of the given datum.

On the other hand, perception is the interpretation of

sensations by associated images and ideas. It is now a

commonplace of philosophy that “ perception contains not
merely sensuous and revived images, but a large element

1 Logic, Vol II. P 289

» NV & NVT., 1. 1 4.
132 NYiYA THEORY OF KNOWLEDGE

of meaning as well.”* Perception is not, as the Bud-

dhists think, an unmeaning sensation of an indeterminate
real called svalaksano. It has a definite meaning and
refers to a determinate object as that is revealed through
sensations. It is only because the Buddhists arbitrarily

deny the meaning element in perception that they are forced

to exclude the complex cognitions of a jar, tree, etc.

(samvrhjndna), from the range of perception. As a matter

of fact, these are as good perceptions as any other. If,

however, we allow with the Buddhists that perception

is a matter of pure sensation, we do not understand how
it can at all be conceived or logically defined A pure sen-
sation IS an unreal abstraction and not a psychological fact.

We cannot point to any of our actual experiences as a case

of pure sensation without any element of ideation in it.

Such an experience, even if it were real, can hardly be

described, far less defined The Buddhist definition of
perception is self-contradictory {vydhata) in so far as it

tries to define and determine what IS undefinable and inde-

terminate. Just as what is perfectly unknowable cannot
even be known as unknowable, so we cannot consistently
^
determine a perfectly indeterminate experience as perception

3. The Jaina, Prdhhdkara and Veddnta definitions oj

perception

It is customary to define perception in terms of sense-

functioning. The ordinary idea is that perception as a

1 Essays jn Critical Realism, p 91

2 NV & NVT ,1 1 4; NM pp 92-93, 97-100 SO., pp 38-39
, ,

Tbe notion of an eneffable sensam, like the Buddhist’s svalakfiana, has also been
repudiated of late by some eminent Western thinkers like Whitehead, Heidegger,
Hickert, Bosauguet, Dewey Whitehead speaks of it as the pensationalist fallacy and
”
Heidegger as the illusory notion of mere givenness, nntinged with the “ concern
which he holds to be constitutive of experience throughout See Charles Hartshorne’s
article on The Intelligibility of Sensations **
in The Monist^ July, 1934, pp. 161-85,
 THE definition OF PERCEFTION 133

form of knowledge is essentially dependent for its origin and

distinctive character on the stimulation of the sense-organs.

There is a departure from this common usage in the defi-

nition of pratyaksa or perception given by the Advaita

Vedanta, the Prabhakara Mimfimsa and the Jama system.
According to the Prabhakaras, perception is the direct
cognition of an object. It is the intuitive or immediate
knowledge that we may have of the subject and object of

knowledge or of knowledge itself. '

For the Jamas too,
pratyaksa is the direct and immediate knowledge of objects.
It is of two kinds . mukhya or the primary and samvya-
vahanka or the practical J'lie first is quite independent
of the mind and the senses. While the origin ot the second
IS conditioned by the mind and the senses, its essence lies

in the direct cognition of some object. Hence perception

IS in its essential nature a direct knowledge of objects.*

In the Advaita Vedanta, perception as a pramana is

the unique cause (karana) of perception as a form of valid

knowledge (pranid). In this sense, the sense-organs
constitute the karana or the unique cause of perceptual
cognition. The latter {t e. pratyaksa pramd), however,
IS defined as immediate and timeless knowledge {caitanya).

Such immediate knowledge is the self itself, because it is

only in the self that there is pure immediacy of knowledge.

The senses are the karana or the unique cause of percep-
tion as immediate knowledge in so far as the mental modi-
fication {antahkaranavrtti) which manifests it, is due to the
function of the sense-organs. What takes place in percep-

tion is this. The antahkarana or mind goes out through

the sense-organ which is m contact with a present percep-

tible object and becomes so modified as to assume the form

1 Saksatp'^atltih pr.ityak3ara meyamaTpraruasu sa, PraA,aranapa/lc?^a pp 51-62.

2 Vi^ada fianasvabhavam pratyaksim, avyavadbanena pramanaDtaranirapek^a*

taya pratibbaaanam vastuno’nubhavo vaiiadyam vijftanasyeti etc , Prameyakamala

m&rta^4a, pp. 57-67.
134 Ni'iYA THEORY OP KNOWLEDGE

of the object itself. The mind being a material principle,

it IS quite possible for it to move and attain the dimension

of the object of perception. Perception is the immediate
knowledge in which the mental modification is non-differ-

ent (ahhinna) from the object and is lit up by the self’s

light. The immediacy of perception, however, is not due

to its being produced by sense-stiinulatrm If that were
so, then inference would have been as immediate as percep-
tion, since, according to the Nai>ayikas, the mind as an
internal sense is operative in inference On the other
hand, there cannot be any immediate knowledge by intui-
tion, because it is not due to the senses The connection
of perception with sense-stimulation is more accidental
than essential.^
That there may be immediate knowledge without
any stimulation of sense is admitted by many leading
philosophers of the West. Any knowledge by acquaintance,
Russell ^ thinks, gives us a direct knowledge of things.
“ ” would be quite possible
Direct cognition,” says Ewing,®
without direct perception ” With regard to perception,
however, it is generally held in European philosophy that
it IS the cognition of an object through sensations Here
the process of perception begins with the action of an extei-

nal object. The object produces certain modifications in

the sense organ and the nervous system and, through these,

gives rise to a mental image corresponding to itself. In

the Advaita Vedanta the order of the process is reversed.
The mind goes out through sense and reaches the object,
and there becomes literally changed into the form of the
object. On this view, the perplexing question of the
correspondence of a mental image to the object, of which
it 18 the image, does not at all arise. The direct apprehen-

1 VP ,
Chapter I
2 The Problems of Philosophy, Chaptei V
3 Mmd, Apnl, 1930, p 140
 THE DEFINITION OF PERCEPTION 135

Sion of objects m perception is thus better explained by

the Vedanta. It goes further than this and maintains
that the essence of perception lies, not in its being produced
by sense-object contact, but in the immediacy (saksattvam)
of the knowledge given by it.

4 The Nyaya definitions of perception

The old school of the Nyaya defines perception in

terms of sense-object contact (indnyarihasannikarsa)

According to it, perception is the valid knowledge produced
by the contact of 'an object with a sense organ. ^ This
means that perception as a form of valid knowledge is

conditioned in its origin by the stimulation of the senses.

This definition of perception follows from the etymological

meaning of the word prafyaksa or perception. Pratyaksa
derivatively means the functioning or operation of the sense

organs, each in relation to its particular object (aksasydksasya

^
pratnnsayam vrttih pratyaksam) In relation to a particular
object, the sense may function in two ways. It may function
to bring about a contact of itself with the object. In this

case, the result is a cognition of the object. Secondly, a

sense organ may be operative to produce a cognition of some
object In this case the sense-function consists in sense-
cognition, and the result is a judgment of the cognised

object as something desirable or undesirable or neither.®

[But even sense-cognition as a form of sense-function is

Conditioned by sense object contact. It follows that

the fact of sense-object contact is involved in the very

meaning of the word pratyaksa, and is common to all

perceptions This common and essential character is

made the basis of a definition of perception. For all

1 NS 1. 1, 4.
,

2 NB, 1.1.3
» Ibtd.
 136 NYiYA THEORY OP KNOWLEDGE

the older Na^yikas, perception is the valid cognition

of an object, as distinguished from feeling and volition,

and as conditioned by the contact of that object with a

particular sense organ.
The above definition of perception as knowledge due to

sense-object contact is accepted by common sense and many

philosophic systems. In the Vaisesika philosophy ^ percep-,

tion IS described as knowledge which is conditioned by the

senses The Samkhya-Yoga system also defines perception

in terms ot sense-stimulation. According to the Samkhya,®

perception is the direct cognition of an object by a sense
when the two come in contact with each other The Yoga
system holds the same view while emphasising that percep-
tion IS especially a cognition of the particularity of an object.

For it perception is ‘
a mental modification produced by
sense impressions and mainly related to the specific characters
entering into the nature of the individual object of percep-

tion.’® In the MImamsa-sutra and the Bhatta school of

Mimamsa, perception is defined as ‘
the cognition which is
produced by the efficient contact of the senses with theirl/

objects.’ ^
Varsaganya, a follower of the Mimamsa school,

reduces perception just to the functioning of the sense

organs.®
Although the old Naiyayikas are disposed to find fault

with the definition of perception given in the other s} stems

of philosophy, they do not dispute the fact of sense-object
contact as constituting the essential nature of perception.

Thus the Bhatta definition is attacked on the ground that

it 13 not limited to valid perceptions but may apply also to

^ Akfamik^iin pratlryotpadyafe pratyaksam, PS , p 94.

* Vide Saihkhya sutra and Pravacanahhd^ya, 1 89 ; Sdmkhyakdnha^d,
3 Vy&sa-B}ia§ya on Yoga-siitrat 1 7

^ Satsamprayoge puru^asyendnyanftra baddhijanraa tatpratyak^ain, Mimdihsd sUtra,

1.14 Vide also ^dstradlpikd, p 35 ; Slokavdrttika, 4 84

5 ^rotrddivftimtt, Varf agapya quoted in NV. & NVT ,114.
 THE DEFINITION OF PERCEPTION 137

doubt and error as forms of perception, in which there is

sense-contact with some object.^ ^This means that a defini-
tion of perceptio ) as valid knowledge (prama) should
explicitly mention that validity is an essential character of
it/ But the context makes it unnecessary for the
Mimarhsaka. Similarly, the Samkhya view is considered
by the Naiyayika to be inadequate, because it does not
expressly state the fact of sense-object contact.^ Such
criticism however is vitiated by a sophistical spirit. It is

true that Isvarakrsna does not use just the phrase ‘

sense-
object contact ’
in his definition (viz pratiinsayadhyavasayo
drsiam). But this follows clearly when we collate the
sutra and the commentaiies on this point It is therefore

unnecessary to dwell at length on the Nyaya criticism of

the above definitions of perception. It will suffice for our

present purpose to say that the definition of perception as

knowledge produced by sense-stimulation or sense-object
contact IS common to the Nyaya-Vai^esika, the Samkhya-
Yogaand the Bhatta Mimamsa system of Indian philosophy.
As we have already said, the same definition of perception
IS generally accepted in Western psychology and philo-
®
sophy
The modern school of the Nyaya takes exception to the
old definition of perception in terms of sense-object contact.

VCLa^e^a, the father of modem Nyaya) opposes it on several

grounds. First, it is objected by him that the definition is

too wide, since it applies to inference and memory as forms

of knowledge in which there is sense-object contact. The
mind as an internal sense is operative and related to the

object known through memory or inference. Secondly, the

» Ibid.
* NM.,p 109
3 Cf, Sir J. H. Parsons, An Introduction to the Theory of Perception p 3 Sensory
:

presentations, as we experience them, invariably evolve perception, however naive it

”
may be, at the start but there is no perception without sensory presentation

18— (1117B)
138 NYiYA THEORY OF KNOWLEDGE

definition is too narrow in another sense. It precludes the

possibility of God’s omniscience which is a direct perception

of all truth and reality. If there can be no perception

without sense-object contact, ue can hardly speak of the

divine perception, since it is not a sensuous cognition in any

sense Further, we do not find that sense-contact, in one
or other of its forms is common to all perceptions. Again,
to define perception in terms of sense-stimulation is to beg
the question What a sense-organ or its stimulation means
IS to be known from perception. As a matter of fact, we
understand what perception is, long before ue know what
a sense is and which of the senses is concerned in producing
it Hence to explain perception by sense-stimulation is to
^
explain the ohscurum per nhseurius In view of such
defects in the old definition, the modern Naiyayikas propose
to define perception as immediate knowdedge It is the
character ol immediacy (sahsatl^dritvani) that is common
to all perceptions The visual, auditory and othei kinds
of perception are alike connected with the feeling that
something is immediately known by the subject or the
knower.’“ Another definition of perception, given by the
modern Nyaya,Ss that it is knowledge which is not brought
about by the instrumentality of any antecedent knowledge^*'
This definition applies to all cases of perception, human or
divine. At the same time it excludes all other kinds of
knowdedge, such as inference, comparison and testimony
Inference is due to the previous knowledge of a fixed relation

between two things (vydpti^ndna) In vpamdna oi com-

parison the operative cause of knowledge (pidnaJcarana) is

the knowledge of similarity between two things (sadr^ya-

jndna). Knowledge by testimony is brought about by the
understanding of the meanings of words (padajndna). So

1 TC , I, pp 639-13.
2 Pratyaksasya saksatkaritvarn lak^a^iaiu, TC ,
ihtd
3 JfianakaraijLakain jfifiDamiti tu vayam, TC , I, p 662
 THE DEFINITION OF PERCEPTION 139

also memory depends for its origin on some direct experience

in the past (purvanuhhava) It is only in the case ol

perception that our knowledge is not caused by previoui

experience. It cannot be said that perception is entirelj
undetermined by previous experience. A determinatt
{savthalpaka) perception ol an object, as having certain
attributes and belonging to a class, is conditioned by th(
previous knowledge of those attributes as standing for a

certain class of things But even here our pr^eption of the

object IS only c^ditioned b ut not caus ed by the previous
knowledge (] ndna any a and not] hdnakaranak a)}
j
Thepiesent
peiceplion of the book before me does not arise out of m>
previous knowledge .ibout books, altliough the character ot

that perception is determined by such antecedent knowledge.

At least the perception is nut the result of a conscious
application of antecedent knowledge to a present case. Tha
Buddhist definition of perception as knowledge whicii is

directly produced by the object alone, excludes all the con-

scious and unconscious conditions ol knowledge from per-

ception except the object The modem Nyaya, however, ex-

cludes all the conscious conditions of knowledge from tha

definition of perception above given. For it, perception la

not entirely undetermined by previous experience. But even il

it be determined by previous experience, it is not consciously

brought about by that experience. Hence perception baa
been defii ed as the knowledge which is not brought about by
the instrumentality of any antecedent knowledge.
In the syncretist school of the Nyaya, attempts have been
made to reconcile the older definition of perception with the
modern. Among the ancient Naiyayikas, Udayana, Vardha-
mana and others have suggested that the definition of per-

ception, as given in the Nyaya Sutra in terms of sense-object

contact, applies only to ordinary human perception {lauktka-

1
Op cit , pp. 552»63
140 NYiYA THEORY OF KNOWLEDGE

matravisayatvat) J The science of logic is concerned with

the conditions or grounds of valid knowledge. It is not

interested in the eternal and unconditioned knowledge of the

divine being. Hence the siitia gives a definition of such
perception as is due to certain specific causes (pram ana).
In view of this it has been suggested by some syncretist
logicians that for normal human peiception the definition is

that it is knowledge produced by sense-object contact But

when we take perception in a wider sense so as to cover
ordinary and extraordinary human perceptions as well as the
eternal perception of the divine being, we should define it as

immediate [oparohsa) knowledge, oi as knowledge which is

not brought about by antecedent knowledge (ihandkara-

naka).^ Kes'avamisra in his Tarkahhdsd goes further and
takes the old ancf modern definitions of perception to mean
the same thing (According to him, perception is the source
of valid immediate knowledge and immediate knowledge is

that which is brought about by sense-object contact/ Hence

for knowledge there is no distinction between what is imme-
diate and what is sensed Rathei immediate knowledge is

just the product of sense-stimulation.

Of the two definitions of perception given above, that

proposed by the modern Nyaya seems to be more acceptable.
That perception is generally conditioned by sense-stimula-
tion or sense-object contact is true as a matter of fact. Still

this fact does not constitute the essential or the universal

character of all perceptions There are cases of knowledge
which are undoubtedly perceptual in character and yet not
brought about by sense-object contact. Instances to the
point are to be found not only in the intuition of seers and

* Vide Udayana’s Kusumdfljalt and Vardhamana’s PrakdU, 4-6

^ Aparokaapramavyapfcdm pratyak^am anyatha laukikapratyaksamatravnakiayam-
iDdnyajanyapramasadhakatamam pratyak^am, TK and 8S pp 67-69 ,

3 Sak^atkaripramakarapam pratyaksam, Baksatkani^I ca pratna saivocyate yendri-

ya}ft, TB., p 6.
 THE DEFINITION OF PERCEPTION 141

God’s omniscience, but in such abnormal perceptions as

illusion, hallucination, dream, delirium and the like In these
cases we have perceptions of objects that are not in actual
contact with our senses. No doubt they are all cases of
wrong knowledge, but that they are actual perceptions we
cannot dispute. When we have such experiences or when
afterwards we recollect them, we feel that we did perceive
certain things which we should not have so perceived.
These then are significant experiences which show that
perceptual knowledge is not essentially a matter of sense-
object contact. Similaily, any ordinary valid perception is

found by analysis to contain elements which are not sensa-

tions in any sense When we peiceive an orange from a
distance, we have a sensation of its colour but not of its

smell, taste and touch Still we perceive all these qualities

as constituents of the total percept ^
Hence it may legiti-

mately be said that sense-stimulation is not the essence of

perception. Wherein that essence lies we may also find in
the light of the above instances They serve to show that
we feel sure to perceive anything that is directly piesented
to us. It does not matter much whethei the direct presen-
tation, which we call perception, is light or wrong, is by
way of sense or not. \Hence we may say that to perceive

a thing is to know it immediately^ The modern Nyaya

hits upon a truth when it defines perception as immediate
knowledge, although it recognises the lact that perception
is generally conditioned by sense-object contact. The
Prabhakara Mimamsa and the Advaita Vedanta are at one
with the modern Nyaya in recognising this truth about the
essential nature o^ perception This is also admitted by
Hobhouse when he says that the common and essential

^
(7/ L S StebbiDg, Logic m Practice, p 13 “ It is true that such judgments
(‘ That IS a cow,’ ‘
Here is a pen') go be 3 ond what la directly given to sense, but it does

not follow that they are reached by inference Certainly such judgments may be
erroneous. None of these is completely trustworthy, but they are all we have,**
 142 NYilA THEORY OF KNOWLEDGE

character of simple ideas of sensation and reflection lies,

not in their dependence on any sense organ, or any special

kind of physiological stimulus, but in then immediate

presence to consciousness. Hence while admitting that

apprehension, in the sense of sensation or perception, is

conditioned by both the sense organ and its stimulation,

^e defines it as the knowledge of what is immediately

present to consciousness/^

^ Hobhonse, The Theory of Knowledge^ Pt I, Ch I Dr Stebbing sepms to

endorse the definition of percepti n as immediate knowledge when she observes that in

perceptual judgirients we merely record what we take tu be dnecily given Vide op ext .
 CHAPTER VII

THE PSYCHOLOGY OP PERCEPTION

1 The Senses (indnya)

From the Nyaya standpoint pereeption is an immediate

valid cognition of reality, due to some kind of sense-object
contact. As such, perception involves four operative condi-
tions or causes, namely, objects, senses, mind and self.

What objects can be perceived and how they are perceived

by us, we shall see in the next chapter. Here we propose
to consider, from the Naiyayika standpoint, the psychologi-
cal questions as to the nature and function of sense, mmd
and self in perception.

It IS generally admitted that perception is primarily

conditioned by the activity of the senses in relation to

some objects. Hence perception is usually defined in terms

of sense-stimulation Even those who refuse to do so

have to admit sense-activity as a factor conditioning all

perception But there is some difference of opinion as to

the exact nature of the senses and their functions in percep-
tion, According to the Buddhists, the senses are the
external organs (golaka) occupying different parts of the
surface of the body ^
The visual sense, for example,
is the pupil of the eye, since objects can be seen only when
the pupil is in order but not otherwise Thus the end-
organs are entrusted by the Buddhists with the function

of perception and therefore spoken of as the senses.

For the Jamas a sense is the physical organ with a

1 Golakasyeodnyatvamiti Bauddha^^, NSV.> 3. 1. 30,

 144 NYiYA THEORY OF KNOWLEDGE

specific energy {^akti}.^ According to the Samkhyas,

the senses are not physical bodies like the pupil of

the eye, but modifications of the subtle material principle

called ahamkara. They think that a physical (bhautika)
organ cannot account for the perception of distant objects.
In perception the senses must function in direct contact
with the objects ol perception But a physical organ like

the eye-pupil cannot have direct contact with an object lying

at a distance or behind a glass. This is possible only if the
sense oigan be all-pervading in character and not a limited
physical substance So the Samkhyas think that the senses
are modifications of a subtle all-pervading matter (aham-
kanka) and are themselves all-pervading in character.^

The Nyaya rejects both the Bauddha and the Sarnkhva

view about the nature of the senses It agrees with the
Mimamsa and the Vedanta in holding that the senses are
neither the end-organs nor modifications of any all-pervad-
ing subtle matter According to these systems the external
senses are material sub.stances constituted by the physical
elements {bhautika) and localised in the defferent end-
organs *
In the Nyaya system, a sense is defined as a
supersensible organ of knowledge having its locus m the
animal organism “ A sense cannot itself be sensed oi per-

ceived. The existence of the senses is not a matter of direct

perception for us It is by means of inference or reasoning

that we know their existence. The eye cannot perceive it-

self. But that there is a visual sense, follows from the*

general law that every function is conditioned by some organ.
So it has been said that a sense is what cannot be sensed

1 Vide Prameyakamalamdrtandaf p 61
2 AhainkaropaJaa.ika[niQdnyam, Taitvakaumudl, 26 Cf NSV ,
ibid.
s NS,1 1 12, VP,Ch VII.SD,p 36
4 Saruasamjuktam jflanakarHujamatindnyamindriyam, TB , p 19 C/ H H Price,
Perception, p. 25 “ By ‘
genses ’
I do not mean ‘
gense-organg ’
but ‘
gense*f acuities,*

1 e the power of being acquainted wah this or that kind of sense-datum,”

 THE PSYCHOLOGY OF PERCEPTION 145

but must be inferred as an organ necessary for the function

of perception. Another definition of sense, given by some
Naiyayikas (e. g. GaAge^a and Vi^vanatha), is that it is

the medium of a contact between the mind and an object

to produce such knowledge as is different from memory.*
This definition, however, is not applicable to mind as a

sense, since it cannot be said to be the medium of contact

between itself and objects like pleasure and pain The
3astradlpikd^ defines sense as what produces a clear and
distinct knowledge of the object it is brought in contact
with. This definition holds good with regard to all the
senses including the mind.
According to the Nyaya and the Mimarasa, there are sik
sense organs. Of these some are c tiled external and some
internal (bahyamabhyantaram) There is only one internal
sense called manas or mind. This will be separately
dealt with in the another section. There are five external
senses, namely, the olfactory, the gustatory, the visual, the
cutaneous and the auditory. These senses are physical in

character, because they are constituted by the physical ele-

ments.® But for their physical character we cannot explain

the limitation of perception to a particular time and space.
If the senses were, as the Samkhyas say, non-physical and
all-pervading principles, we should have simultaneous per-
ceptions of all objects in the world. Each sense is capable
of revealing the existence of one particular class of objects.
A sense organ IS constituted by the physical element whose
qualities are sensed by it. It possesses the specific attribute

of its constituent physical element and is therefore capable

of perceiving the qualities belonging to it.*

1 SM, 68, TO ,1, pp 650 61

« Vide p. 36
3 NS andSD.ffcid
* NSV.,3132.

19— (1117B)
146 NYiYA THEOBY OF KNOWLEDGE

The olfactory sense {ghrana) is the organ of apprehend-

ing smell. It must have the quality of smell in it in order

to apprehend smell in other things As such, it is consti-

tuted by the earth, to which smell originally belongs as an

attribute. The sense of smell is said to have its seat

in the forepart of the nasal cavity (ndsagravartt).^ This,

however, is contradicted by modem psychology which holds

that ‘
the organ of smell is a mucous membrane lining the
roof and part of the walls of the extreme upper portion of
^
the nasal cavities.’
The gustatory sense Oasana) is the condition of taste-

sensations. It IS constituted by the physical element called

water, because it is possessed of the specific attribute of
that element, vtz. ta^te It is located in the forepart of the
tongue ijihiagraiarh) But the localisation of the sense of
taste in the tip of the tongue does not agree with the finding
of scientific psychology From it we learn that the end-
organs of taste, t.e. the taste-bulbs occur largely in the
posterior part of the tongue The tip of the tongue is

especially sensitive to sweet taste. As such, it may be said

''
to be the organ of certain tastes
The visual sense {cahsn) is the giound ol coloui-sensa-
tions and is itself coloured It has its locus in the pupil of

the eye. It is constituted by a luminous substance called

tejas or light In the case of the visual sense the consti-
tuent element of light has no manifest form and touch
(anudbhutarupasparh). The sun as a luminous orb has
both form and touch manifest in it Hence it is that the
one cannot be,*while the other is, perceived by us The
account of the visual sense given here is in substantial
agreement with the modern theory of vision, although it

lacks most of its finer details. It will have the support of

> TB.ib
® Vide Titchener, Text-Book of Psychology
3 Jbtd
 THE PSYCHOLOGY OF PERCEPTION 147

modern optics which recognises the dependence of visual

sensation on the eneigy of light and the presence of a
colouring matter behind the lens, and treats the retina as
the
organ of vision/
The cutaneous sense {tvak) is the source of touch-sensa-
tions and temperature-sensations. It is constituted by the
physical element an, because, like the air, ii manifests the
quality of touch. The locus or the end-organ of the sense
of touch is the whole skin of the body, from head to foot,
outside and inside its surface ^
From a common-sense
standpoint no distinction is here made between the different
sensations of touch or pressure and those of waimth and
cold. In truth, the cutaneous sense is highly complex.
Many psychologists distinguish between four cutaneous
senses, namely, those of pressure, warmth, cold, and pain."
The Naiyajikas do not go so far in their account of the
cutaneous sense, but describe both pressure and temperature
as touch sensations. Nor do they subsciibe to the view that
the sense of touch is the primitive sense, from which the
other external senses develop by increasing differentiation.

“Touch,” Aristotle observed, “ is the mother of the senses.”

Modern psychologists also think that “ starting from this

mode ot sensibility as a basis the other senses develop by

processes of increasing complexity and refinement.”'' The
Naiyayikas oppose this hypothesis on the ground that
the sense of touch cannot, in any degree, perform the
function of the other senses in those who are deprived of
them."
The auditory sense (srotra) is the source of sensations of
sound. It has its seat in the drum of the ear. It is possess-

1 Ibid
2 TB , pp 20, 24
3 Titcbener, op cit,

4 Hollmgworlb, Psychology p 466,

6 NB., 3.1 51-63
148 NYAYA THEORY OF KNOWLEDGE

ed of the quality of sound. As such, it is identical with a

portion of the physical element dkasa, as that is present in
and limited by the ear-hole. It is not a separate substance,
but is a limited portion of dka.^a itself, since the quality
of sound belongs originally to dkdh} According to the
Vedanta, however, sound is not exclusively a property of
dka^a, since it is perceived in the air and other elements as
well. But it admits that the sense of hearing is constituted
by the physical element of dkdh.^ The Mimamsa differs

here from both the Nyaya and the Vedanta in holding that
the auditory sense is a portion of space (digbhaga) enclosed

within the aural cavity.® Thus all the three systems

agree in holding that the five external senses are physical
entities constituted by the physical elements.
To the above list of the six senses, recognised by the
Nyaya and the Mimamsa, the Hamkhya system adds five

other senses. These aie the five senses of action {karmen-

driya). They are called speech, hands, feet, rectum and the
sex-organ, and perform respective!) the functions of speak-
ing, prehension, locomotion, evacuation and reproduction.^
Thus the Samkbya gives us a list of eleven sense-organs.
The Vedanta accepts this with one exception. It excludes
the mind or antahkarana from the list of the senses.’ The
Nyaya objects to this long list on the ground that the
assumption of a sense for every function will lead to count-
less senses. If by sense we mean, as we should, a bodily
organ of some special ._kind of perception {sdksdtpratlti-
sddhana), then the organs of action cannot be called sense
organs. They do not give us any perceptions of a new
thing or quality other than those connected with the six

1 TB p 20.
,

* VP ch vir.
,

3 SD , p. 36
* SB,AikhyakiTtk&, 26 & 28.
‘ VP., Ch, VII
 THE PSYCHOLOGY OF PERCEPTION 149

senses of smell, taste^ sight, touch, hearing and the

mind.^
The Naiyajikas’ enumeration of the senses, it will be
seen, is different from that usually given in Western psy-
chology. Of the six senses enumerated by them, mind as
an internal sense finds no place in modern psychology,
while a number of internal senses is added to the five exter-
nal senses recognised by them. These are the kinaesthetic
senses. Then we are told that there are certain organic sen-
sations which are derived from the internal organs of the
body. These too will have to be regarded as senses if we
seriously accept the view of organic sensations. Further,
the cutaneous sense is subdivided into the pressure sense,
the temperature senses and the pam sense. Of these, the
first two, namely, the pressure and temperature senses are
included by the Naiyayikas in the sense of touch, since,

according to them, warmth and cold are only different kinds

of touch sensations. Hence in addition to the five senses
of the Naiyayikas, we have to admit a pain sense, several
kinaesthetic senses and the vital or organic senses. These
are supposed to be necessary to explain the sensation of pain,
the perceptions of movement and position, resistance and
weight, and other organic sensations connected with the
abdominal organs, the digestive and urinary systems, the
circulatory and respiratory systems and the genital systems.
It seems to me, however, that this long list of the senses,

to which Western psychology commits us, has its basis in

an unsound epistemology . A careful epistemological analy-

sis of what are called sensations will show that they are
cognitive mental states which acquaint us with the qualities

of things. A sensation has, therefore, a cognitive value and

has reference to some given datum The cognitive value of

a sensation lies in its being the basis of our perception of

j NVT„ p. 681 j NM., pp. 482 f.

150 NYiYA THEORY OF KNOWLEDGE

the thing which produces ihe sensation Another charac-

teristic of sensations is that we can have images correspond-
ing to them at the time when they are past. Judged by such
criteria, it is very doubtful if we can legitimately speak of a
sensation of pain or of kmaesthetic and organic sensations.
While touch, warmth and cold are qualities ol things, pain
cannot be referred to anything as its quality. We feel pain
indeed, but do not perceive anything as paintul, just as we
perceive a rose as red when it produces the sensation of red
in us. We can hardly form an image of a pain previously
felt in the same way in which we can image a previously
experienced colour or sound Hence pain is bettei charac-
terised as a feeling, rather than as a sensation Similarly
the so-called organic sensations may be shown to be feelings
aroused by certain states of the vital organs, or by the ordi-
nary sensations of pressure, warmth, etc. Thus according
to Titchener,^ ‘
the sense of satiety, of a full stomach would
come from an upward pressure against the diaphragm. Thirst
appears as a diffuse pressure or as a blend of pressure and
warmth. There are times when the separate heart-beats are
’

clearly sensed as dull throbbing pressuie Titchener says

further that ‘
the special sensations of the genital system
appear first as an excitement, then as gratification and
thirdly as relief.’ But all these aie plainly feelings, and not

sensations in the proper sense Lastly, what are tailed

kinaesthetic sensations are analysable into certain feelings

and ordinary sensations produced by different kinds ol bodily

activity. ‘
A muscular sensation is ordinarily a dull and
diffuse pressure upon the skin. With increasing intensity

it takes on a dragging character and sometimes passes into

dull pain.’
‘
The articular sensation is said to be a massive

complex of sensations in the w'rist-jomt whose quality is not

distinguishable from that of cutaneous pressure.’ The per-

1 Vide Text-Book of Psycholoqy,^p 160-92

 THE PSYCHOLOGY OF PERCEPTION 151

ceptions of the movement and position of our limbs, with

closed eyes, and those of the weight and resistance of other
bodies are said to be based upon the articular sensations.
But we have no specific sensations of movement and posi-
tion, resistance and weight In fact, none of them can be
treated as a sense-datum like coloui and sound. Hence w(*

lequire no separate senses tor their perceptions If, then,

the so-c tiled miiscuiat and articular sensations can be analys-
ed into sensations of pressure and feelings of strain and
pain, we see no reason why we must admit separate senses
foi them We admit different external senses for the differ-

ent kinds ot sensations, like coloui, sound, etc ,

because we
cannot analyse any ot them into any other The sen'^ations

ot one sense are quite distinct from those of any other sense.
But even those who speak of the kinsesthetic senses would
admit that ‘
the kinaesthetic sensations are, in general, very
like the cutaneou-, and, in one case, indistinguishable from
cutaneous pressure.’ Supposing that there are kinaesthetic
souses, we ask : W’hat -ensa or sense-qualities do they
acquaint us with ^ Certainly, movement and position,

resistance and weight are not sensa like colours and sounds.
Theie being no other distinct sensa for them, we ate to say
that the kiiuesthetic sensations, like the alleged sensation
of inneivation, are really feelings produced by bodily

inovements. Accordingly, the so-called kinaesthetic senses

will have to be regarded as organs of action and not of
sensation

2 Function of the senses

The function of the senses is to produce perception of

objects. For a sense organ, to function is to give us im-

mediate knowledge about certain objects. How, then, do
the senses function to give us perception of objects ? Is the

activity of the senses conditioned by their contact with the

152 NYiYA THEORY OP KNOWLEDGE

objects of perception ? Or, do they give us perception of

objects without any contact with them ?

According to the Buddhists, the senses function without
direct contact with the objects of perception. They are all

“ distance receptors ” (aprdpyakari) and do not require im-

mediate contact with their objects This is especially seen
in the case of the senses of sight and hearing. We see far-off

objects that cannot have any direct or appriiximate contact

with the eyes We hear sound produced at a long distance
from our ears. Similarly, the eye perceives objects much
larger than itself and so incapable of being covered by it.

Again, the eye and the ear can perceive the distance and
direction of sights and sounds respectively. This they
could not, if, like the senses of smell, taste, and touch, they
were in immediate contact with their objects. Lastly,
many of us can, at the same time, see the same object or
hear the same sound from different places. Conversely, one
man can, almost at the same time, see two things or hear
two sounds, fairly apart from each other. This shows that
the senses of sight and hearing may function without actual
contact with their respective objects.^
According to the Nyaya, Sarakhya, Mlmaihsa and
Vedanta systems, the senses can perceive only such objects
as are in direct or indirect contact with them (prdpyakari)
This IS obvious in the case of the so-called lower senses,
namely, touch, taste and smell. Sensations of touch and
taste arise only when the sense-organs are in immediate
contact with their respective objects. To taste a thing is

to place it in direct contact with the tongue. To touch a

thing IS to bring it in contact with the skin. Sense-object
contact as a condition of olfactory sensation is no less real,

although it may sometimes be less obvious. If the smelling

object be m our immediate surrounding, there is obvious

1 NV and NVT., 1.1.4.

 THE PSYCHOLOGY OF PliRCLPriON 163

contact of it with the olfactory organ. If it be at a distance

from us, then the odorous particles given off by the object
are brought in actual contact with the sense organ by cur-
rents of air.* In the case of the lower senses, all systems
of philosophy admit a direct sense-object contact. It is also

admitted that the senses of smell, taste and touch remain

passive in their end-organs where they are met by their
respective stimuli. The remaining two senses of sight and
hearing also act in contact with their objects, although not
quite as directly as the rest. According to the Vedanta, the
sense of hearing travels to the sounding objects and gives
us sensations of sound. ^ The Nyaya, however, agrees with
modern science in holding that sound-waves sent by the
object are received into the ear-passage and there perceived
as sound ’
According to both the hiyaya and the Vedanta,
the visual sense reaches out to its object and gives us colour-
sensations This IS why the eye and the ear can perceive
the distance and direction of their respective objects. While
in modern science visual sensation is believed to be due to the
transmission of light-waves from the object to the eye, in

Indian philosophy it is explained by the emanation of light-

rays from the eye to the object. Ordinarily the colour-rays
are not visible, but aremedium of sense-
inferred as the
object contact in visual perception. Still we may perceive

them under certain special circumstances, as when they

emanate from the eyes of cats and other animals in a dark
night. The uninformed and the uncultured may not admit
sense-object contact in the perception of distant objects,
because the medium of such contact is imperceptible. But
that there cannot be any perception without sense-object
contact IS implied in all cases of obstructed sense-activity

(avaranopapatU) We cannot see things hidden behind an

1 NM., p 479.
* VP Ch I
.

3 BP 166.66
,

20— (1117B)
164 NYIYA THKORY OF KNOWLEDGE

opaque body. Again, what seems to be a simultaneous per-

ception of objects occupying different places, is really a case

of different perceptions occurring in quick succession. We

perceive a near object earlier than a remote one, because
our senses take a longer time to reach the latter. Hence
sense-object contact is a condition of all perceptions.^
According to all the Naiyayikas, sense-object contact
(indnyarthasannikarsa) which is the cause of perception,
IS of SIX kinds First, we have a case of direct contact
which consists in the conjunction (samyoga) between sense
and its object. In the visual perception of a substance like

the jar, there is an immediate contact of the eyes with the

object. The visual sense finds its way to the jar and be-
comes conjoined with it Secondly, there may be an in-
direct contact of sense with its object through the mediation
of a third term that is related to both When we see the
colour of the jar, our eyes come in contact with the colour
through the medium of tbe jar The jar is conjoined with
the eyes on the one hand, and contains the colour as an in-
herent quality of it, on the other. Here the contact between
sense and object is due to the object’s (here colour) insepara-
ble relation to what is conjoined with sense. Hence this
sense-object contact is called samyuHa-samavaya or a rela-

tion of inherence in what is conjoined to sense. Thirdly,

sense-obtect contact may be more indirect than what we find
in the preceding case. It may be due to the mediation of two
terms which, by their relation, connect sense with its ob-
ject. When by means of the eyes we perceive a universal
like ‘
colourness ’
(rupatvddisdmdnya) inhering in the jar’s
colour, there is contact of the eyes with the object ‘
colour-
ness ’
through the medium of the two terms ‘ jar ’
and
‘colour.’ This sense-object contact is called samyukta-
samaveta-samavaya, since it is due to the object’s

1 NV. and NVT., 1.1 4 ; NM., pp. 479 f.

 THE Psychology of perception 155

inherence (samavaya) in something (here colour) which

IS inseparably related to (samaveta) what (here jar)
IS conjoined {samyukta) to sense. The fourth case ol

indirect sense-object contact is called samavaya or in-

herence. Here the sense is in contact with its object
in s) far as the object inheres as a quality in the sense itself.

This 18 illustrated in the auditory perception of sound. The

ear’s contact with sound is due simply to the latter’s inher-
ence as a quality in its own substance, akaia, which consti-
tutes the auditory sense. In the fifth case sense-object
conitact IS called samaveia-samavaya or the relation of inher-

ence in that which inheres in sense Here the sense is in

contact with its object through the medium ot a third term
that IS inseparably related to both Thus in the auditory
perception of the universal ‘soundness’ {•iahdatva), the ear

IS in contact with the object ‘

soundness,’ because it inheres
in sound which, in its turn, inheres as a quality in the ear.
The la^t type of sense-object contact is called visesanatd or
visesyavisesanahhava. In it the sense is in contact with its

object in so far as the object is a qualidcation {vi^esana) of

another term connected with sense It is by means of such

sense-object contact that the Naiyayikas explain the percep-
tions of non-existence (ahhdva) and the relation of inherence
{samavaya) It takes different forms according to the different
ways in which the mediating term is related to sense. Thus
it IS called samyukLa-vi^e^anatd when the object is adjectival

{oisesana) to that which is conjoined (samyukta) to the sense

organ. This is illustrated by the visual perception of the

non-existence of a jar in a certain place. Here the eye is in

contact with non-existence as a qualification (vUesana) of

the place which is in conjunction (samyukta) with the eye.
Similarly, it is called samyukta-samavetaviiesanatd when the
sense is related to an object that is adjectival (visesana) to

' BP. ASM, 69-62

156 NYiYA THEORY OF KNOWLEDGF

what inheres (samaveta) in that which is conjoined

(samyukta) to sense. Thus in the perception of the non-
existence of sound in the odour of the earth, our sense is in

contact with the non-existence as a qualification of odour

which inheres in the earth as conjoined to the sense Or,
It may be called sarnanetavisesanata when the object of pei-
ception IS a qualification {vi^esana) ot that which inheres
(samaveta) in sense Thus when we perceive that there is

no odour in sound, our sense is in contact with the non-

existence of odour as a qualification of sound which inheres
in the auditory organ. So too with regard to the other
forms of this kind of sense-object contact. These have been
collectively caWeA i lie sanatd or the contact of sense with that
'
which 18 adjectival to another term connected with sense
The Vedantins also recognise these six kinds of sense-object
contact But they do not admit the perception of non-
existence (abhdva) and inherence (samavdya) ^
We shall

discuss this point afterwards.

With regard to perception, it has been asked What is

the karana or the unique cause of peiception ? Is it sense

or sense-object contact or something else According to the
older Naiyayikas, '
sense-object contact (sannikarsa) is the
karana of perception, since it is the immediate antecedent to
the appearance of the phenomenon of perception. The
contact of sense with its appropriate object does not normally
fail to produce a perception of it. Hence it is that sense-
function or sense-contact should be recognised as the karana
or specific cause ot perception. Modern Naiyayikas, how-
ever, hold that the sense organ is the karana of perception.''
A karana is the unique operative cause of an effect (vydpdra-
vadasddhdrana-karana) It is something that produces the

> 8M , 61,
' VP , Chapa I * VI
» NB NV & NVT
, , 1.1. 3-4.
* BP, & SM., 68.
 THE PSYCHOLOGY OP PERCEPTION 167

effect by its operatiom and not the operation itself. The

senses are such causes as produce perception by their activity
or contact with the objects. As such, it is the sense
(indnya) that should be called the Hrana of perception.
Ke^ava Misra reconciles these views by reference to Ihe
different kinds of perception. He thinks that sense and
sense-object contact are the karanas of mrvikalpaka and
savikalpaka perceptions respectively. Even mrvikalpaka
cognition may be taken as karana in relation to cognitions of

the values of objects for us {hanopaddnopeksabuddhayah)

Others, however, think that sense is the karana of all kinds
of perception.’

3 The nature and function of th( mind (manas)

It IS with some hesitation that we use the woid m.ind for

manas in connection especially with the Nyaya philosophy
In Western philosophy mind is geneially taken to mean both
the subject ot consciousness or the self and the totality of

conscious states and processes in winch the self is manifested

In this sense mind corresponds roughly, not to the manas,
but to the atnian or soul in the Nyaya system Of course,
among European thinkers tlieie arc some, the materialists

and some behaviourists, who reduce mind or self to a func-

tion of Thus understood, mind stands for just

the body
what the Carvakas mean by atman or the self In deference

to the common usage of language \\e propose to use the word

mind for manas, pointing out the differences in their

meaning.
According to the Nyaya- Vaisesikas, manas or mind is a
sense like the senses of taste, smell, etc. It is an internal

sense having its locus m the heart (hrdaydnianarti):^ In

the Samkhya and Mimamsa systems also mind is treated as

1 TB., p. 6.

* TB., p ao
158 NriYA theory of knowledge

an internal (dntara) pense.^ The Sarhkhya considers it to be

^
an unconscious product of subtle matter (ahamkdra) For
the Nyaya-Vaise§ikas, mind is an eternal substance
which is different from the physical substances. Unlike
the external senses, mind is non-physical (ahhautika),
t.e. it 18 not constituted by any of the physical
elements of earth, water, etc It is not, like the exter-
nal senses, possessed of any - specific attiibute of the
physical elements, nor is it limited to the perception o" any
particular class of objects. As an internal sense it is

concerned in all knowledge in different ways It is atomic

and exists in contact with the soul {anvatmasamyogi)
The mind as a sense cannot be perceived, but is known by
inference (na prafyaksamapt ti'annmdnagamyam)
Just as external perception depends on the external
senses, so internal perception depends on an internal sense,

called manas Every perception requires the contact of an

object with its special sense organ We have perceptions
of such subjective facts as pleasure, pain, desire, aversion and
the like. These perceptions cannot be due to the senses
of sight, hearing, etc ,
since they arise even without these
senses. Hence there must be an internal sense to

produce internal perceptions The mind is also a

condition of external perception. The external senses
can perceive objects only when they are in contact
with the mind. To perceive an object the mind must
attend to it through the senses. We do not perceive things
in a state of absentmindedness, even though our senses be
in physical contact with them ® So also the mind is a con-
dition of such subjective states and processes as doubt and

> SD., p. 86
2 Samkhyakartka, 27.
3 NB.,1, l,4;TB.,pp. 23, 30.
* BP. 67,86.
s TB., Ibid.
 THE PSYCHOLOGY OF PERCEPTION 169

dream, memory and inference, etc.^ Some Naiyayikas

hold that although the mind is a condition of all knowledge,
yet it does not act as a sense m the case of memory,
inference, etc., because that will render them indistinguish-
able from perception.^ But it may be said that in memory
and inference tlie mind gives us a knowledge of objects, not
by its contact with them, but through some other know-
ledge, as a past experience and the understanding of a uni-
versal relation (tyapti). Hence memory and inference are
not cases of perception, although they are dependent on the
function of the mind as a sense.®
That manas or the mind is and is atomic follows also
from the order of succession among our cognitions. At
any moment of our waking life various objects are acting
upon our body. All the external senses may thus be in
contact with their objects at the same time. But we cannot
have more than one cognition at one moment. Hence it

follows that the senses of smell, taste, etc., must come in

contact with some internal organ in order to produce cog-

nitions. This internal organ is manas or the mind. It has
no extension or magnitude (avydpi), because it cannot come
in contact with more than one thing at one time. If the
mind were an extended organ, it would have had simul-
taneous contact with more senses than one and we could
have many perceptions at one and the same time. This
being not the case, we are to say that the mind is atomic
(anupanmdna) .*

The Vedanta view of the mind is different from those of

the other systems. According to it, manas is that function
{vrtti) of the antahkarana which is concerned in the state of
doubt. The same antaftkarana is called buddhi, ahamkdra

1 NB., 1.1 16
2 TC.,I.p. 660,SM.,61
» SD., p.86.
< N B., 1. 1. 16, 3. a. 60-68, BP., 86.
160 NYiYA THEORY OF KNOWLEDGE

and citta, according as it functions respectively in the states

of decision (ni^caya), conceit (garva) and recollection

(smarana). It is the anfahkarana which performs these

and other mental functions, such as cognition, feeling,

desire, etc. Hence by the mind we are to understand,

not manas, but anfahkarana as conceived by the Vedantist
According to him, the mind is not an atomic substance,
but an inert principle of limited dimension {pancchinna)
Although inert Qada) in itself, it manifests pure intelligence
{caitanya) and is therefore regarded as intelligent in a
secondary sense (jhanatvopacara). The mind is not a sense
(indnya) whose existence is proved by inference from the
perception of pleasure, pain, etc Direct knowledge or per-
ception IS not due to sense-object contact We have a
direct perception of the mind when we perceive the qualities
of pleasure, pain, etc,, in it And a perception ot these
mental states does not require any internal sense, called
’
manas, in the other systems
It IS to be observed here that the view of mind as sense
is not acceptable Those who take the mind as internal

sense deny that it is a physical {bhavhha) thing of any

kind. So the mind as sense cannot be a physiological
apparatus like the brain or any part of it that is directly

correlated to conscious processes. The mind as a non-

physical sense is analogous to the ‘
inner sense ’
conceived
as a special faculty of inner experience in traditional

Western psychology.^ But the one is quite as unnecessary

as the other to explain the facts of consciousness It is the
internal perception of pleasure, pain, etc., that is held to

require an internal sense. But if by sense we mean, as the

Naiyayikas do mean, a medium of contact between mind
and '
hen the mind itself cannot be a medium of
3

1 VP Ch
. 1
* Vide Klemm, History of Psychology
 THE PSYCHOLOGY OF PERCEPTION 161

contact between itself and objects like pleasure, pain, etc.

The medium must be something else which should be
called sense and not the mind. In truth, however, no inter-
nal or ‘
inner sense ’
is necessary for the perception of
pleasure, pain and other psychical processes These are
held by the Naijayikas to be attributes of the self. As
such, the> are parts of the conscious life of the self and are, by
their very nature, conscious or perceived facts The Vedanta
IS right in holding that the mind (antahkarana) perceives
itself and its functions without the help of any internal
sense. It agrees with modern psychology in holding that
mind is just the totality of conscious states and processes.
It IS involved in some difficulty by making conscious pheno-
mena qualities of a material substratum. How can the
antahkarana, which is inert and material in itself, become a
conscious and an intelligent mind '
By the self’s relation
to or reflection in it,’ says the Advaita Vedantist The self

(atman), which mind nor matter, is the ground of

is neither
both mental and material phenomena The Advaita Vedan-
tist would thus agree with the new realists who hold that
mind and matter are not two opposed substances but different

arrangements of the same neutral stuff. Or, as Kussell has

said “ Matter is not so material and mind not so mental
If so, mind and matter need not
’
as IS generally supposed ^

be two contradictory terms or irreconcilable opposites, but

may become related to each other. Hence mental functions

may belong to an apparently material substratum like the

antahkarana.

4. The self and its function in perception

By the self {dtma) we are to understand the individual

soul ifivatma) in connection with perception. The self, in

1 Analysts of Mtnd, p 86.

21— (lil7B)
 162 NYiYA THEORY OF KNOWLEDGE

this sen'?e, has been conceived in different ways by different

schools of philosophy. We find four main views of the self

in Indian philosophy. According to the Carvakas, the

self is either the body with the attribute of intelligence or

the aggregate of the senses of sight, hearing, etc. This

is the materialistic conception of the self. The
Buddhists reduce the self to the mind as a stream
of thought or a series of cognitions. Like the

empiricists and the sensationalists, they admit only the

empirical self or the ‘
me.’ Among the Vedantists, some,
the Advaitavadins, take the self as an unchanging selt-

shining intelligence {smprakaia caitanya) which is neithei

subject nor object, neither the ‘

I ’
nor the ‘
me ’
Other
Vedantists, the Vi^istadvaitavadins, however, hold that the
self IS not pure intelligence as such, but an intelligent
subject called the ego or the ‘
I ’ Qfiatahamartha
evdtma)}
The Nyaya-Vai^esikas adopt the realistic view ol the
self. According to them, the self is a unique substance, to

which all cognitions, feelings and conations belong as its

qualities or attributes.^ Desire, aversion and volition, plea-

sure, pain and cognition are all qualities of the self. These
qualities cannot belong to the phy.sical substances, since
they are mental. Hence we must admit that they are the
peculiar properties of some substance other than the physical
substances. The self is different in different bodies, because
their experiences do not overlap but are kept^ distinct The
self is indestructible and eternal (nitya). It is ubiquitous or
infinite (vibhu), since it is not limited in its activities by
time and space.® The body or the external senses cannot
be called self because intelligence or consciousness cannot be
their attribute. The body, by itself, is unconscious and

1 Ftie 1 1.1,
* NS, 1.1.10, FS.pp. 80 f.

» BP, 61
 THE PSYCHOLOGY OF PERCEPTION 163

unintelligent. The senses cannot explain all mental func-

tions, Imagination, memory, ideation, and the like are

independent of the senses of sight, hearing, etc. The mind

too cannot take the place of the self If the mind be, as the
Nyaya-Vai^esikas hold, an atomic substance, then the
qualities of pleasure, pain, etc ,
in it must be as impercep-

tible as the mind itself If, on the other hand, the mind
be a series of cognitions, each manifesting itself, then
memory becomes inexplicable No member ot a mere series

ot cognitions can know what has preceded it or what will

‘
succeed it (vasandyah samkramdsamhhavdt) . A succession
of ideas is not an idea of succession.’ The Advaita
Vedantiii’s idea of the self as eternal self-shimng intelligence
IS no rnoie acceptable to the Naiyayikas than that of the
Buddhists. There is no such thing as pure intelligence
unrelated to some subject and object Intelligence cannot
subsist without a certain locus. Hence the self is not
intelligence as such, but a substantial principle owning
intelligence as its attribute. The self is not mere knowledge,
but a knower, an ego or the ‘I’ {aiiamkdrdhaya).^
Still knowledge or intelligence is not an essential and in-

separable attribute of the soul. The soul is, in itself,

neither material nor mental, but a neutral substance which

comes to have the attribute of intelligence or consciousness
“
in Its relation to the body
According to the Nyaya, the self is the fundamental
giouiid of all mental functions. It is involved in all

cognitions, affections and volitions. All the experiences of

an individual, whether cognitive or otherwise, must inhere

in the self and cannot be separated from it. What are

known as innate faculties of the mind, the reflexes, instincts

and inborn feelings of fear, hatred, etc., are all conditioned

1 BP and SM 48-50, ,

I!
NV ,1122, NM., p 432.
 164 NYiYA THEORY OF KNOWLEDGE

by the past experiences of a self in this or a previous

life.’ Even the body and the senses cannot function unless
they are connected with the self. Hence the self is called
the guiding principle of the body and the senses (indriya-
dyadhisthata) In the case of external perception the
self comes m contact with the object through the medium of
the mind and the external senses. ‘
When we have the
perception of an external object, the self is in contact
with the mind, the mind with the external sense concerned,
and the external sense with the object of perception.® In
the case of internal perception the mediation of external
sense is unnecessary. Here the object is in contact only
with the internal sense, called manas. We shall consider
internal perception later on.
Like the older school of Western realism, the Nyaya-
Vai^esika system accepts the substantialist theory of the
self. But their view of the self differs from that of the
realists in two important respects. For the realists
consciousness or intelligence is an essential and inseparable
attribute of the soul. The soul cannot exist without the
attribute of consciousness. For tlie Nyaya-Vai^esikas,
however, the soul is in itself an unconscious substance.
Consciousness is an accidental property of the soul, due
only to Its temporary connection with the body. Then,
while the realists conceive the individual soul at least to
be a limited substance, the Nyaya-Vai^esikas take all souls
as unlimited and all-pervading substances {vihhu). But
the Nyaya view of the self is untenable. If the soul be a

substance we do not understand how it can exist without

its distinctive attribute of consciousness and still be called
the self. Without consciousness the soul is indistinguishable
from matter. Again, to say that the soul is in itself pure

» NS. 3 1.19 ff

* BP., 47.
> TB., p. 6.
 Tj^B PSYCHOLOGY OF PERCEPTION

substance is to say that it is a substance without attri-

butes, which, however, is a contradiction. . Further, if the
soul be an independent entity of the realistic type we
cannot explain relation to consciousness or mind or body.
its

The soul-substance is not in its own essence related to

anything else. Hence it can only be externally related to
other things through the medium of a third thing. In
perception it is said to be related to the object through the
medium of the senses. But how are we to explain its rela-

tion to the mind or other senses? That must be by some other

medium. Again, that medium must require another and so
on indefinitely. So we are to give up the idea of the self as

a neutral substance externally related to consciousness.

In truth, the self is a self-conscious reality. We may calf

it a substance in the sense of an existing ‘

continuant, ’
but
that continuant is psychical and not physical. It is not a
metaphysical surface on which consciousness is accidentally
reflected. Rather consciousness or intelligence is the
intrinsic character of its existence, the core of its being.
It is not indeed the passing thought, which James pro-
poses to call the self. The Naiyayika is right in insisting

that fleeting ideas or cognitions cannot take the place of the

permanent self. So also an unconscious substance cannot be
the conscious self. "We should say that the self is the
intelligent reality or being which is the ground of all thought
and experience. It is the eternal self-manifesting real which
witnesses, but is not involved in, the flow of events.
 CHAPTER VIII

ORDINARY PERCEPTION AND ITS OBJECTS

1. Different hinds of perception and the categories of

reality {padartha)

Taking perception as a general name for all true cogni-

tions produced by sense-object contact the Naiyayikas
distinguish between different kinds of it. First we have
the distinctio n between laukika or ordinary and alaukika or
extraordinary perceptions. This distinction depends on
the way in which the senses come m contact with their
objects. ' We have laukika or ordinary perception when
there is the usual sense-contact w'lth objects present to sense.
In alaukika perception, however, the object is such as is

not ordinarily present to sense but is conveyed to sense

through an unusual medium. Ordinary perception, again,
is^jif two kind^, namely, external (hahyaj and internaj
(mdnasa). The former is due to the external sense/? of

sight, hearing, touch, taste and smell. The latter is

brought about by the mind’s contact with certain objects.

Thus we have six kinds of ordinary perception, namely, the
visual, auditory, tactual, gustatory, olfactory and the

internal.^ In this chapter we propose to consider the objects

of external and internal perceptions. The special cases of

perception, called alaukika, will be discussed in a later

chapter.
According to the Nyaya-Vai^esikas, there are two main
types of reality, namely, being and non-being {bhdvo'hhd-
mSca). Being as a category {padartha) stands for all that

1 BM., 62. 63.

 ORDINARY PERORPTION AND ITS OBJECTS 167

ts, or for all positive realities. Similarly, non-being stands

for whatever is not, i.e for all negative realities. That a
thing is not, or does not exist is as much a real fact as that
it IS, or does exist. Hence ahhdva or non-existence is as
good a category of reality as bhdva or being. There are
SIX kinds of being or positive reality These are substance
{dravya), quality iguna), action (karma), the universal
(samdnya), particularity (vUesa), and the relation of
inherence (samavdya). Of these, the first three are exist-
ents. These participate in existence (sattd). On the other
hand, the universal, particularity and the relation of in-
herence are positive realities but not existent facts. These
do not participate in existence (sattd). T hey pos sess being
b ut not existence. Hence there are two types of being or
positi\e reality, mz the existent and the subsistent. Of
the six kinds ot positive reality recognised by the Nyaya-
Vai^esikas, the first three are existents and the last three
subsistents. Thus we have altogether seven categories of
reality, namely, substance, quality, action, the universal,
particularity, the relation of inherence and non-existence.^
All objects of the world or all realities have been brought
under these seven categories by the Nvaya-Vai^esikas.
Their scheme of the classification of realities may be com-
pared to that of the modern realists. Among modern
realists Russell,^ Alexander ® and others recognise the reality

of negation or non-existence. According to them, negative

or non-existent facts are as real and objective as positive
facts. It IS also held by them that among positive facts

some have existence in a particular time and space, while

others have subsistence only is so far as these are free from

limitation to one particular space and time. According to
them, univers-ils and relatio is ate such subsistent realities.

1 BP and SM , 2, 14

2 The Analym of Mind, pp 275 76

Space, Tme and Deity, pp 200 22,

168 NYiYA THI ORY OF KNOWLEDGE

But for the Naiyayikas, particularity also is a subsistent

fact and the relation of conjunction {samyoga) has existence

as an attribute of the substances related by it.

All realities, we have said, are brought under the seven

categories. But all of them cannot ordinarily be objects

of perception. Hence with regard to the different objects

included under each of the categories we have to distinguish

between those that are perceived and those that are imper-
ceptible.

2. Perception of substances or things (dravya)

A substance is defined as the constitutive cause of things

or as the substratum of qualities.' There are nine kinds

of substances. These are : earth, water, light, air, dkaia,
time, space, soul and mind.'* Of these, the first four stand

for both the atoms of earth, water, light and air, and the
compounds formed by these atoms. The atoms of earth,
etc., cannot be perceived. A compound of two atoms, called
dvyanuka or the dyad, is also imperceptible, because, like an
atom, It has no dimension and manifest quality (mafeaftea
and udbhutatva) Mind (manas) as another atomic sub-
stance IS not an object of ordinary sense perception So
also dkd^a, time and space are, according to the Nyaya-
Vai^esikas, imperceptible substances.® Ikdia is an all-

pervading medium which is inferred from the phenomena

of sound. Space and time are conceived as two receptacles
of unlimited dimension holding all things and events within
them. These two are the most fundamental conditions for
the existence of finite objects and are therefore called the
origin of all that IS originated (jani/anamjan aka A). Each

1 SiimavayikaraQaui ilravyn'ii, guija^rayo \5, TB , p 20.

» BP ,3
3 TB. and TM Ch. 1.
 ORDINARY PERCEPTION AND ITS OBJECTS 169

of these is said to be one, eternal and infinite.* As infinite

wholes these cannot be perceived, since the conditions of
perception, viz. limited dimension and manifest quality,
are absent in them.
It follows from the above that the perceptible substances
are earth, water, light, air and the soul. Of these, the
soul is the obje ct of internal perception which we shall
consider afterwards With regard to air as a substance,
there is some difference of opinion among the Naiyayikas.
According to the older Naiyayikas, there are two conditions
of the external perception of a substance, namely, that it

must have a limited dimension and manifest colour (waliattre

^
sail udhhutarupavattvam) On this view, air becomes
imperceptible, since it has no manifest colour in it. Its

existence is therefore to be known by inference from the

quality of touch which subsists in the air. According to
the modern Naiyayikas, however, colour is not a condition
of all external perception of substances. It is only in the

visual perception of substance that manifest colour is an

essential condition. The sense of touch also perceives
substances in which the quality of touch is manifest
(udbhutasparh) Hence we may have a perception of air as
®
a substance possessing the quality of manifest touch
Admitting that we have a perception of the substances
of earth, water, light and air, it should be noted that what
we perceive is neither an atom {paramanu) of earth, etc ,

nor a compound of any two atoms only {dvyanuka). To be

perceived, a substance must have a limited d imension. It

must be neither infinite like space, etc., nor infinitesimal

like an atom or the compound of two atoms. Hence the
perceptible substances are finite things from the triad

’ BP und SM ,
44-46
* IM.Ch 1.

S SM 56.
,

22— (1117B)
170 NYiYA THEORY OF KNOWriBDGE

(trasarenu) upwards. A triad (tryanuka) is a compound of

three dyads or three compounds of two atoms each. It is

the minimum perceptible substance in Indian philosophy.

As a matter of fact, therefore, the substances that are per-

ceived by the exteimal senses are complex finite things like

the jar, table, tree, etc.

In the external perception of substances or finite things

the senses come in direct contact with the things. This

sense-object contact is called samyoga or conjunction Of
the five exter nal senses, it is the sen->e of sight toucji

that c an give us a perception of things,. The e\e perceives

things that have manifest colour, t e are visible The sense

of touch perceives things that possess manifest touch or are

tangible. In both ca-^es the perceived thing is conjoined
to the sense. They are not always or inseparably related
but are two substances that come in actual contact with
each other at the moment of perception (ayutasiddhija'

bhavat). The other senses of taste, smell and bearing

cannot give us perceptions of things. These can perceive
the qualities of taste, etc., but not the substances or things,
in which the qualities inhere Hence we have only visual

and tactual perceptions of physical things To perceive a

thing is to perceive it as having a limited dimension in
space. The organs of sight and touch, being extended,
can perceive things as having a limited extension The
other senses cannot perceive extension and are therefore
incapable of perceiving things as extended m space ’

The things that are perceived by the external senses

possess a limited dimension (mahattva). This means that

J Gandha^rayagraba^e ta ghra^asyasamarthyam, etc SM,53, ghraijarasana.

6rotraj>i dravyftgrahakS^i, cak^ustvanmanamsi dravyagrahakani, TK p 9 H H Price
also thinks that our beliefs concerning material things are based upon visual and
tactnal experiences^ and that other modes of Beose-experience, e g hearing and
smelling cannot by themselves give us any knowledge of the material world. See
his Pereeftum, p, 2
 ordinary PUftCEPTION AND ITS OBJECTS lYl

they are made ap of parts (avayava). The magnitude of a

thing depends on the aggregation of a number of parts

composing it. Hence it seems that to perceive a thing we

must peiceive all its component parts at one and the same
time. But a simultaneou-! perception of all tlie parts of a

thing IS not possible. In the visual perception of a tree,

for example, the eye comes in contact with only a part
of its front side. There is no contact of the eye with the
other sides of the tree or other parts that fall outside the

visual field. How then can we have a perception of the

tree when only a part of it is actually perceived ?

This question has troubled psychologists for a long

time The answer given by the associationists is generally

accepted by other schools of psychology, such as struc-

turalism, functionalism and self-psycliology. According to
the associationist psychology of Hume, Mill and others,

a thing IS an aggregate of its parts We perceive the

different parts one after the other and, it may be, on different
occasions. It is because the different parts are always

found to go together that their corresponding ideas become

associated in our minds Hence the perception of one part

recalls the ideas of the other parts, and all of them

associated together give us the perception of the tree.

Accoiding to other psychologists, the peiceptioii of the

tree is no doubt due to the combination of the presented

part with ideas or representations of the other parts But

this combination is effected, not by the association of ideas,

but by the synthetic activity of the mind or the self.

Among Indian thinkers the Buddhists adopt the associa-

tionist explanation of the perception of things as wholes

made up of parts.
Here the Naiyayikas point out that the associationist

explanation account for the perception of a thing.

fails to

According to it, the perception of a thing consists in having

or images of other parts when one
part of it is
ideas
172 NYiYA THEORY OF KNOWLEDGE

actually perceived. But these other parts are as far from

being the thing as the perceived part itself. Further, to
think of the other parts as connected with the perceived
part is not to perceive them, but to infer the unperceived
from the perceived. Hence what we call the perception of a
thing 18 really an inference or remembrance of it. It
cannot be said that to perceive a thing the mind is to
synthesise the presentation of a part with the representa-
tion of other parts. In that case we have a mental con-
struction and no perception of the thing. So the Naiyayi-
kas maintain that we have a direct perception of the thing
as a whole along with the perception of any part of it

According to them, a thing is not a mere aggregate of

parts (avayavasamuha) but a whole which is distinct from
any_ or all of the parts constituting it (dravyantara) It

subsists in the parts not by fractions, but wholly and

indivisibly. To perceive any part of a thing as part is also

to perceive the whole to which it belongs. When we

perceive a book we apprehend it directly as a whole of
parts. We have not to construct it from successive per-
ceptions of different parts or from perception of some and
ideal representation of others. we have not a direct
If

perception of the thing as a whole, we cannot perceive it

at all. If we are to construct the thing from sense-
impressions of its parts, perception would become inferential
knowledge.'
Some modern psychologists confirm the Nyaya view
of the direc t perceptio n of a thing as a whole. H. H.
Price ® rejects the associationist and the rationalist explana-
tion on this point and holds that in perception ‘
what we
accept 18 not simply a surface (though this is the most that
can be present to our senses) but a complete material thing

1 NB ,
2 1 28-34
2 PercepUon, pp 163*54
 ORDINARY Perception and its objects i73

as a whole. When we sense the sense-datum the house

just presents itself to us as a whole, without any reasoning
or passage of the mind.’ Similarly, the Gestalt ps>cho-
gists ^
show that the perception of a thing, say an
orange, is not a colour experience somehow combined
with the experiences of a certain shape, taste, touch and
smell. On the other hand, it is a whole of experience which
gives us knowledge of the thing as a whole, i e. as a round
fruit, soft to touch and with sweet acid taste. We try to
account for this whole of experience by saying that it is a
compound of certain simpler ideas like those of colour, taste,

smell, etc. But here'we tail to notice that the experience-

whole is what we have to start with and the simpler ideas
are discovered by subsequent analysis. Hence we are to say
that there is first the perception of a thing as a whole and
that its parts are next perceived by focu''SiDg attention on
this or that aspect of it.

But while the NaiyayiKas are right in bolding that we

have the d irect perception of a thing as a whole, _
they seem
to limit arbitrarily the range of such perception to the tac-
tual and the visual field. They deny the capacity of per-

ceiving things to the senses of taste, smell and hearing. But

it IS dogmatic to say that tasting or smelling or hearing a
thing IS not perceiving it. It is true that taste, smell and
sound are the qualities of substances or things. But so also

are colour and touch. Hence if the senses of sight and

touch can perceive things when they sense their colour and
touch, there is no reason why the other senses should fail to
perceive things when they perceive their other qualities.

This IS all the more necessary for the Naiyayikas who hold

that to perceive a quality the sense must first come in con-

tact with the substance, of which it is the quality. For the

perception of a quality, the sense must be related, through

1 Psychologies of 1925, Kohler, Gestalt Psychology

174 NYiiA THEORY OF KNOWLEDGE

the substance, with the quality {sathyukta-samavaya). But

if there is contact of sense with the substance there must be
a perception of it. Even if it be said that perception is not
determined by sense-object contact but by the character of
immediacy, we have to admit that the senses of taste, smell

and hearing give us a perception of things. An appeal to

direct experience shows that the gustatory, olfactory and
auditory cognitions of things are as immediate as their
visual and tactual perceptions. .

3, Perception of attributes (gum) and actions (karma)

An attribute (guna) is defined as that which exists in a

substance and has no quality or activity in it ’

A substance
exists by itself and is the constituent (samavayi) cause of
things. An attribute depends for its existence on some
substance and is never constitutive of things. It is a non-
constituent (asamamyi) cause of things in so far as it deter-
mines their nature and character All attributes must be
owned by substances. So there cannot be an attribute of
attributes An attribute is itself attributeles. (nirguna). An
attribute is a static property of things It hangs on the
thing as something passive and inactive (nisknya). So it is

different from both substance and action. There are altogether

twenty-four kinds of attributes. These are: colour (rupa),
taste (rasa), smell (gandha), touch (sparh), sound (^ahda),

number (samkhyd), m ignitude (pariindna), differentia

(prthaktva), conjunction (sathyoga), disjunction (vibhdga),

remoteness (paratoa), nearness (aparatva), fluidity (drava-
tva), viscidity (^neha), knowledge (buddhi), pleasure (sukha),
pain (duhkha), desire (icchd), aversion (dvesa), effort

(prayatna), heaviness (gurulva), merit (dharma), demerit

(adharma) and faculty (samskdra)

^ Dravya6rit& jfieya nirguiaa nisknya gupah, BP ,

86
2 BP., 8-5
 ORDINARY PERCEPTION AND ITS OBJECTS 176

All attributes do not admit of sense perception., Hence

we are to exclude the imperceptible attributes from the ob-
jects of perception FacnltLy (samskara) is of three different
kinds- velocity (i;egfa), which keeps a thing in motion;
mental disposition (bhavana), which enables us lo remember
and recognise things; and elasticity (sthiUsthapaka) by
which a thing tends towards equilibrium when disturbed.
Of these, mental disposition and elasticitj' are regarded as
imperceptible attributes. So also merit (dharma) an<^

d^ejjjfimi ladtiarmaj are considered to be supersensible attri-

butes of the soul ’
With regard to effort (prayatna) some
Naiyayikas make a 'distinction between three different
kinds of it, namely, vitil effort (jlvanayoni), which main-
tains the flow of life fprawasawcarafietu), and positive and
negative volitions {praiirtti and nivrtti) According to them,
vital effort {fivamyom prayatna) is an attribute which is not
perceived, but inferred as the cause of vital functions.
Modern Naiyayikas, however, do not recognise the attribute
of vital effort “
\lmost all the Indian systems agree in
holding that heaviness or weight (gnrutva) is a quality
which cannot be perceived, but must be inferied ^
The
reason why these qualities are treated as imperceptible is

that they cannot be perceived by any of the senses lecog-

nised in Indian philosophy. They are supersensible entities

and hence there cannot be any kind of contact between these

and our senses. In the -Jaina system, however, light and
*
heavy are treated as touch-sensations
The remaining attributes are regarded as capable of

being pi-rceived by the senses But those belonging to the

soul, s uch as knowledge, pleasure, pain, desire, aversion and
volition, ar e the_ obje ct s of i ntern al perception w in h we

1 BP ,
93-9#
* SM & Dmakari, 149 52
^ BP., 153.
* Cf, PaflcasUMyasara^ Ch. 11.
176 NYiYA THEORY OF KNOWLEDGE

shall afterwards consider. These apart, the attributes that

are left are the objects of external perception. Some of

them are perceived only by one sense, and some by two

senses. Thus colour (rupa) is perceived only by tbe eye.
Taste (rasa) is a quality of substances that is perceived only
by the gustatory sense {rasam). Smell (gandha) can be
perceived only by the olfactory sense (ghrana), and touch
only by the cutaneous sense (tvak). The perception of
these four attributes has two conditions, namely, that they
must be manifest (udhhuta) and that they must belong to

substances possessing a limited dimension (mahadvrttitva) ’

In the perception of these attributes the second form of

sense-object contact {viz samyukta-samavdya) obtains.

The attributes come in contact with their respective senses
through their inherence (samavdya) in the things that are
conjoined (samynkta) to the senses. Conjunction or direct
contact is possible only between two substances. Hence
there cannot be any direct contact of the attributes with
their special senses. But the attributes of colour, etc ,

inhere in certain things or substances, such as a jar, an

orange, a table, etc. When these things come in actual

contact {i.e. are conjoined) with the eye, etc., there is a

perception of the qualities of colour, taste, smell and touch

as inhering {samaveta) in them.
Sound is an attribute of dkdia, and is perceived only by
the auditory sense (hotra). In the perception of sound the
fourth kind of sense-object contact ( viz. samavdya) bolds
good. Sound is perceived when it comes in contact with
the auditory sense. 'I’his contact cannot be a relation of
conjunction {samyoga), since one of the terms is an attribute
The sense of hearing is a portion of dkdh, in which sound
inheres as an attribute. Hence the contact of sound with

the auditory organ means its inherence {samavdya) in the

1 BP , .53-56, 92-^3.
 ORDINARY PBRCBPTION AND ITS OBJECTS 177

latter.' Sound is first produced by the conjunction or dis-

junction of things. But this sound cannot be perceived
unless it reaches the passage of the ear. So the first sound
produces other sounds which either undulate towards the ear
or move in straight lines mall directions {vlcitarahganyayena
kadamhamukulanyayena vd) In this way the senes of sounds
meets the ear. The last number of the series which strikes
upon the ear-drum is perceived, while the first and the inter-
mediate ones are not pi'rceived. So it is not correct to say
that we perceive sound at a distance.'' The Vedantist,
however, thinks that in the perception of sound it is the ear
that meets sound and not vice versa. We are not wrong
when we say that we hear the sound of the distant drum.
For there is nothing to contradict the obvious experience of
distant sounds.®
The attributes that admit of perception by both the'

senses of sight and touch are number, magnitude, differentia,

conjunction, disjunction, remoteness, nearness, fluidity,

viscidity and velocity.* These are perceived by the eye when

connected with light and manifest colour, and by the
tactual sense when connected with manifest touch.® Of
visible and tangible things there may be respectively a visual
and tactual perception of their number, etc. In the percep-
tion of these qualities we have the second kind of sense-

object contact, viz samyukla-samavdya. The objects {i.e.

number and the rest) come in contact with the senses

through their inherence (samavdya) in certain things that

are conjoined {sain]ukta) to the senses.

In the perception of magnitude jpa nmana) we are to
admit a further fourfold cQntactJbetween sense an d the,

> BP ,
63 60
• BP , & SM ,
165 66
3 VP .
Chap 1,

4 BP & SM ,
93
6 IbiJ, 64-56

23— (1117B)
178 NYiYA THEORY OF KNOWLEDGE

things co ncerned. By magnitude is here meant a limited

dimension that belongs to ordinary things like tables and
jars, and neither the unlimited dimension of the all-pervad-

ing substances, nor the minute dimension of atoms and

dyads, for these are imperceptible. The magnitude
of ordinary sensible things is due to the number and magni-
tude of their component parts. Hence to perceive the
magnitude of a thing we are to perceive the co-existence
and relative position of the constituent parts. This is
rendered possible by four kinds of contact {catustayasanni-
karsa) that between the different parts of an extended
sense-organ and the different parts of the thing, that
between the different parts
sense-organ andof the
the whole of the thing, that between parts of the thing
and the whole of the sense-organ and that between
the whole of the sense-organ and the whole of the
thing.* by virtue of such contacts between sense and
It IS

things that we can perceive their magnitude from a distance

The Naiyayika has to take the help of so many kinds of
sense-object contact because he believes in the direct visual
perception of the magnitude of distant things. The mus-
cular sensation of movement is not admitted by him as a
factor in the perception of magnitude or limited extension.
Differentia (prthaktva) is a positive character of things.
That one thing from another, e. g. a cow and a
is different

horse, does not simply mean that the one is not the other.
Difference does not consist in the mutual negation (anyonyd-
hhdva) of two things. One thing is different from other
things, not simply because it excludes or negates them, but
because it has a distinctive character of its own whereby it is
differentiated from them. This distinctive character consti-
tutes its differentia from other things. Differentia is thus an
objective character or attribute of things and is perceived in

> TB., p. 6.
 ORDINARY PERCEPTION AND ITS OBJECTS 1*79

things that are perceptible.^ The Vedanta, however, takes

difference to be a case of mutual non-existence and holds
tliat It is known not by perception, but by non-perception
{anwpalahdht) Modern Naiyayikas also do not treat
difference as a separate quality, but reduce it to mutual non-
existence.®
Conjunction (samyoga) is a relation {samhandha) that is

perceived as an attribute of the things related by it. Dis-

junction or separation (vibhaga) is not a relation {sam-
bandha). Eather, it is the negation of the relation of con-
junction between two things It is also perceived as an
attribute of the things which are disjoined. Space and
time as infinite wholes are imperceptible substances. But
the remoteness or nearness of things in time and space is a
perceptible quality of the things. Things are far or near in

space according as they are separated from our body by a

larger or smaller number of contacts with space-points.
Similarly,' things are near or remote in time according as
they have a smaller or larger number of contacts with time-
mstants. Such position m time and space becomes an at-

tribute of things and is perceived by the senses of sight and

touch.^ Fluidity, viscidity and velocity are the qualities of

certain things and are perceived by the senses of sight and

touch like other perceptible qualities. Here, again, the
modern Naiyayikas do not lecognise remoteness and nearness
as separate qualities, since these are due to varied contacts
of an object with points of time and space.®
Action (karma) is physical movement. Like an
attribute, it inheres only in substance.® It is different

1 SM 114
,

* VP Ob VI.
,

5 Dinakan, 114.
4 BP., 64-66, 121-24
6 Dtnofcarl, 124.
® CalaDatmakaii) karmai ivadravyamatraTftti, TB » p, 28
180 NYiYA THEORY OT' KNOWLEUGB

from both substance and attribute. Substance is the

support of both action and attribute. An attribute is

a static character oi things, but actions are dynamic

While an attribute is a passive property that does not take

us beyond the thing it belongs to, an action is a transitive

process by which one thing reaches another. So it is

regarded as the independent cause of the conjunction and

disjunction of things An action has no attribute because
the latter belongs only to substance. All actions or move-
ments must subsist in limited corporeal substances {murta-
dravyavrtti) Hence there can be no action or motion in
the all-pervading substances. There are five kinds of action
such as throwing upward, throwing downward, contraction,
expansion and locomotion {utksepanapahsepanakuncana-
prasdranagamana) These actions belong to such substances
as earth, water, air, light or fire and the mind. Those
inhering in the mind are imperceptible, since the mind is so.

The action or motion of the perceptible substances can be

perceived by the senses of sight and touch In the percep-
tion ot movement the second kind of sense-object contact,
viz. samyukta-samavaya, is operative. The senses come in

contact with movement through their conjunction with the

*
things in which it inheres

4. The umvenal {sdmdnya) particularity {viksa) and

the relation of inherence (samavdya)

There are three views of the universal or the class-

essence in Indian philosophy. In the Bauddha philosophy
we have the nominalistic view. According to it, the indi-
vidual alone is real and there is no class-essence or universal
other than the particular objects of experience. The idea of
sameness that we may have with regard to a number of indi-

1 BP., 64-66.
 ORDlNAKY PERCEPTION ANt) ITS OBJECTS 181

viduals of a certain class is due to their being called by the

same name. It is only the name that is general, and the
name does not stand for any positive essence that is present
in all the individuals. It means only that the individuals
called by one name are different from those to which a
different name is given. Thus certain animals are called
cow, not because they possess any common essence, but
because they are different from all animals that are not cows.
So there is no universal but the name with a negative
connotation.'
The Jamas and the Advaita Vedantins “
adopt the con-
ceptualistic view of the universal. According to them, the
universal does not stand for any independent entity over and
above the individuals On the other hand, it is constituted
by the essential common attributes of the individuals.
Hence the universal is not separate from the individuals,
but IS identical with them in point of existence The two
are related by way of identity. The universal has existence,
not in our minds only, but in the particular objects of
experience It does not however come to them from out-
side, but IS just their common nature. On this view,
“ individuals have,” as Mill sajs, “ no essences.” **

The Nyaya-Vaisesikas accept the realistic view of the

universal. According to them, universals are eternal entities

which are distinct from, but inhere in many individuals

*
(mtyatve satyanekasamavetatvam) The universal is the
ba«i8 of the notion of sameness that we have with regard to
all the individuals of a certain class. It is because there is

one single essence present in different individuals that they

are brought under a class and thought of as essentially the

1 Vtde TB , p 28, Stx Buddhist Nydya Tracis^ Ft 5, on Samanya-dn^a^a-dik-

prasarita.
* VP Ch ,
I , Outlines of Jainism^ p 116
3 J 8. Mill, System of Logic, p 78
* SM.,8.
 182 NYiYA THEORY OE KNOWLEDGE

same.* Like Plato’s “ ideas,” or “ essences ” of the

mediaeval realists, sdmdnya or the universal is a real entity
winch corresponds to a general term or class-concept in our

mind. Some of the modern realists also hold that a

‘
universal is an eternal timeless entity which may be shared
by many particulars.’ The Naiyayikas agree further with
the modern realists in holding that universals do not come
under existence (sattd). These do not exist in time and
space, but have being and subsist in substance, attribute
and action There is no universal subsisting in another
universal {sdmdnydnadhiharanatvam), nor is there any
universal for particularity (miesa), inherence (samaodya) and
non-existence (abhdva).^ Modern realists, however, do not
admit with the Naiyayikas that all universals pertaining to
sensible ot)jects ate capable of being perceived by the senses.

According to the former, we can perceive only such univer-

sals as may be called ” sensible qualities,” as for example,
colour, whiteness, hardness, etc.®

According to the Naiyayikas, the universals that subsist

in supersensible objects are imperceptible (atindnyavrltlni
atlndnydni) The universals that inhere in perceptible
objects (pratyaksavrtflni) are perceived by the senses which
perceive their locus. The universals of substances or things

(dravya) are perceived by the senses of sight and touch

'*

provided they are visible and tangible things respectively.

Thus jarness {ghataiva), treeness {vrksatva) and manhood
are universals that inhere respectively in all individual jars,
trees and men. When perceiving any of these individuals, we
directly cognise also the universal inhering in it Here the
second form of sense-contact, viz. samyukta-samavdya,
functions. The universal “ jarness ” comes in contact with

'
TB.P 28,P8,p 164
» BP. ASM., 14, 16
5 Vide, EotaeW, Problem* of Phtloiophy, Oils IX, X
*
TK., p 9.
 ORDINARY PERCEPTION AND 1TB OBJECTS 183

the visual or the tactual sense through its inherence (sama-

vaya) in the jar that is seen or touched and is thus conjoined
(samyukta) to the sense concerned.
The perception of universals pertaining to attributes
(gum) and actions {karma) generally takes place through
the third kind of sense-contact, viz. samyukta- samaveta-
saniacaya} When perceiving such qualities as a red
colour, a fragrant smell, a sweet taste and a bard touch, we
also know directly, %.e. perceive the universals of colour,
smell, taste and touch as such. when we per-
Similarly,
ceive a particular kind of movement we know directly what
motion in general is To perceive anything as thrown
upward or downward, is also to perceive the universal
of upw'ard or downward impulsion {utksepanatvadi).
Smellness (gandhatoa) as a universal pertaining to different
kinds of smell is perceived by the olfactory sense. Taste
(rasatva) as the genus of different kinds of taste is perceived
by the gustatory sense. Similarly, the universals of colour
(rupalva) and touch (sparhlva) are perceived by the senses
of sight and touch respectively. The universals pertaining

to other sensible attributes and actions or movements are

perceived by both the senses of sight and touch. ^ These
universals are perceived when they come in contact with
the senses through their inherence in attributes or actions
which inhere in certain things that are conjoined with the
senses. When, for example, I see a white paper, I perceive
the universal ‘
whiteness ’
as intimately related to the

particular kind of white colour in the paper which is in

conjunction with rny eyes Soundness (Sabdatva) as the

genus of different kinds of sound is perceived by the auditory

sense through the fifth kind of sense-contact, viz. samaveta-

samav&ya. The universal ‘

soundness ’
is in contact with

1 TB., p. 6! BP., 64-66, 69-60.

BP., 68-66.
184 NYAYA THEORY OF KNOWLEDGE

the ear through its inherence (samavdya) in sound which

inheres as a quality (samaveta) in the auditory sense.’

The universals of actions or motions belonging to perceptible

things are perceived by the senses of sight and touch
through the third kind of sense-contact, viz. samyukta-
samaveta-samavdya. The universal ‘
motion ’
is in contact

with the visual or cutaneous sense in so far as it subsists

in a particular kind of movement inhering (samaveta) in

something that is seen or touched (i e. is conjoined with

the visual or tactual sense).’*

According to the Vedanta, the universal, as constituted

by the common attributes of the individuals, is perceived
along with the perception of the individuals. The percep-
tion of the different kinds of universals is mediated by
different kinds of sense-contact The universal of sub-
stances IS perceived by samyuktataddtmya, that of attributes
or actions by samyuktdbhinnatddatmya, and that of sound
by tdddtmyavadahhinna forms of sense-object contact.
These three forms correspond respectively to the second,

third and fifth forms of sense-contact admitted by the

Naiyayikas. But where the latter speak of the relation of
inherence (samavdya), the Vedantist puts in the relation
of identity, since inherence is not admitted by him as a
distinct category and the relation between substance and
attribute, or universal and particular is said to be one of

identity (tdddtmya), so that they require no tertium quid

like inherence to relate the one to the other.®

Particularity (viksa) is the extreme opposite of the

universal (sdmdnyaj. It is the ultimate ground of the
differences of things from one another. Things are ordi-

narily distinguished from one another by means of their

component parts or aspects. But the differences of parts

> BP , 68. 61
» BP., 54.66.
> VP., Ch. I
 ORDINARY PERCEPTION AND ITS OBJECTS ^.85

or aspects rest on those of other smaller parts or finer

aspects. In this way we are led to the ultimate differences

of the simple substances, beyond which we cannot go.

The ultimate differf'nces are due to certain unique characters

which distinguish one thing from all other things of the
world. Particularity is such unique character of the simple
and eternal substances.^ It is completely different from
universdls or the things coming under any universal.
Hence by particularity we are to understand the unique
individuality of space, time, dkasa, mind.s, souls, and the
.itoras of earth,, water, light and air. Particularity is thus
eternal and subsists lu the eternal substances (nitya-
^
nityadravyaorttih) Tliere are innumerable particularities,
since the individuals in which they subsist are innumerable.
Other things are distinguished by their particularities, but
the latter are distinguished by themselves (svata eva
vydvrttdh). Hence particulaiities are so many ultimates
(antydh) in the analysis and explanation of things. There
cannot be any perception of them, since they are super-
sensible entities (atlndnyah) Some modern Naiyayikas,
however, do not admit that particularity is a distinct

category. If the particularities can he distinguished by

themselves, without having any distinguishing character in
them, the eternal substances also may be distinguished by
themselves without requiring any character like particular-
ity (viie§a).*
There are two main relations recognised m the Nyaya-
Vai^^ika philosophy These are the relations of conjunc-
tion {samyoga), and inherence {samaodya).^ A relation is

here conceived as a positive connection between two facts.

1 BP & SM 10 ,

« TB p ‘28, PS
, , p 168
3 TS.pp 11,88
* Dtftakari, 10
s TB , p 2,

24— (1117B)
186 NMYA THEORY OP KNOWLEDGE

Hence it is that the oomnoonly recognised negative relations

of disjunction, spatial and temporal separation, etc., are
treated as qualities and not relations. Conjunction
(samyoga) is a transient relation between two things which
may and normally do exist in separation from each other.
Two balls moving from opposite directions meet at a certain
point of space The relation which holds between them
when they meet is one of conjunction. It is a temporary
contact between two things which may again be separated
(.yutasiddha) So long as it is, it exists as a quality of the
terms related, but it does not affect the existence of those
terras. It makes no difference to the existence of the balls
whether they are conjoined to each other or not. Thus
conjunction is an external relation which exists as .in

accidental quality of the terms related by it.

Svnaodyi is an etern il and natural (ayutasiddha) relation

between two facts, of which one inheres in the other. ^ It

18 a necessary relation in so
far as the related terms or at
least one of them cannot exist without being related to tht
other. Like conjunction, it is distinct from the termi
related by it.^ But while conjunction exists as an adventi-
tious quality of the related terms, samavaya does not exis
as a qualit} but a lways su bsists between the things relatgj]
The relation of samavaya holds between such entities as
whole and part, attribute or action and substance, the uni
versal and the individual, particularity and the simplt
eternal subs‘anc 's. Of these pairs, the first cannot exisi
without being related to the other. The whole is alway'
related to its parts, attribute or action is inseparabh
related to some substance, the universal must always stibsH'
111 the individual, and so particularity (vthsa) in th^^ simpk

1 Ayutanddhapoh taihbandhafy ecmaviyalf etc., TB . p 2, samavSyatoath mfy

taihbandhatram etc , 8M ,11.
$ Svatathbandhibhmno nttyali aadibandhafy iamavdya^, TM Ch. I
,
 ORDINARY PliRCBPTION AND ITS OBJECTS 181

substances.* But the part- can exist without being actuall>

re'ated to the whole. A substance m ly have being withoui
relation to its attributes or actions.® So we see that
saniavdya is a necessary relation for one of the relata aac
not for both. It cannot be called an internal relation
since the related entities are not affected by it. It standi

as a natural link between two facts, each of which

has a distinct existence of its own. A substance and its

attribute are not made to be such by their relation to each

other. Hence, like conjunction, samavaya is an external
relation. But, unlike conjunction, it is not produced, nor
does it exist as a quality in time and space. It always
subsists between two facts, which are naturally related to
each other Some modern realists treat relations as
universals that do not exist but subsist.® For the
Naiyayika, however, the relation of conjunction is an
attribute and has existence in time and space. The relation
of inherence is a subsistent fact but not a universal. It is a

category distinct from the universal and the rest. Both

conjunction and inherence are objects of perception. The
relation of inherence is perceived by the senses of sight and
touch. In perceiving this relation there is contact between
sense and object by way of vHesanatd. We directly

perceive that the cloth inheres in the threads. But the

contact of the relation of inherence with our sense is

indirect. It comes in contact with the sense of sight or

touch through being adjectival to the threads which are

conjoined with either of them.^ The Vai^esikas, however,'

hold that samavaya cannot be perceived, but must be known

by inference. They agree with the Naiyayikas in maintain-

1 BP and SM , H
* Dravyani nirgii^auteva prathainamutpadyate parent tataamavetft ntpadyante,

etc , TB . p 8.

s Rusatll, Problems of Philosophy, Cli TX.

‘ BP., 64-66, 62
188 NYAYA THEORY OP KNOWLEDGE

ing that samavaya is one and eternal. We cannot dis-

tinguish between different kinds of samavaya. because it is

the same in ail Some modern Naiyayikas, however,

cases ^

contend that samavaya is not of one, bat many different

kinds, since one thing is found to inhere in other things in
^
different ways
Svarupasamhandha * is a third relation admitted by the
Naiyayikas. It has been introduced to explain such cases
of connection between things as are neither conjunction
(samyoga) nor inherence Unlike the relations of conjunc-
tion and inherence, svarupasamhandha is not distinct from
the terms related by it Rather, the relation is itself

constituted by one of the relata What is the relation

between an object and our knowledge of it? It cannot be

inherence, since knowledge, i.*- a quality, inheres not in the
object, but in the soul Nor can it be conjunction, because
that is possible only between two substances, while know-
ledge is an attribute and not a substance Nor can we say
that we do not require any relation. The two being
distinct entities require somehow to be related. Hence it is

said that a new relation, c tiled svarupasamhandha, is to

relate the two The object is related to knowledge in so far

as it is known The relation of knowledge to the object is

thus the object itself as known {arthasvarupa) The object
as known is what we mean by its relation to knowledge, so
that the relation is constituted by the object, or is due to

the nature of the object itself

The Nyaya doctrine of samavaya has been severely
criticised and rejected by the Advaita Vedantm. Samkara
in his commentary on the Brahma-Sfitra shows that the
necessity of the category of samavaya arises from the
Naiyayika’s conception of a thing as a collection of distinct

1 PS pp ,
172-176
* Dtnakartfll,
^
BM.f 11; Ny^yakoh, p. 1067,
 ORDINARY PERCEPTION AND ITS OBJECTS 189

and different entities like substance, attribute, action, the

universal an 1 particularity. If these are so many distinct

entities, we have to explain how they are united in the

thing so as to make it one whole This cannot be explain-

ed by the relation of conjunction, since it is a case of the
dependence of attribute, action etc., on substance, while
conjunction is a relation between independent substances
which can exist in isolation So we have to introduce a
new relation to explain the intimate connection between a
substance and its attributes, actions, etc., and this is

samavaya or the relation of inherence. But if attributes,

actions, etc., are distinct and different from substance,

there is no more reason to suppose that they depend on
substance than there is to think that such independent
objects as man, horse and cow, depend on one another. If,

on the other hand, attribute and the rest depend on substance

and cannot exist without it, we should say that they are

not different entities hut onlj different aspects of the sub-

stance itself The same thing is called by the different

names of attribute, action, etc , according to its different

organizations (somsthana), in the same way in which the

same man is a fathei, a son, a brother, etc ,
under different
conditions A substance and its attribute or action, the

universal and the individual, the wdiole and the part are
such that we cannot have any experience of the one
without the other They aie inseparable both in our
experience and in point of existence It follows from this

that they are not different entities, but aspects of the same
thing To say that tliev aie distinct and diffennt entities

which require to be related by samavaya, which also is a

distinct entity, is to couit the fallacy ol infinite regress

{anavastha) How- is samataya itself lelated to each of the

terms 1 elated by it To explain this w'e have to bring in

new relations which being distinct entities will also require

other relations and so on ad infinitum. Hence we are to

190 NYSYA THEORY OF KNOV\ LEDGE

say that attributes actions, universals, etc., are not

independent entities, but aspects of the same substance, and
that we do not require a new relation like samatdya to relate
them to substance
The Vedantin’s criticism of the Njaya view of samavdya
seems to mjss the essential point. From the standpoint of
a common-sense realism the Naiyayika maintains the
distinction between a substance and its attributes or actions,

the universal and the individual, the whole and the part.
It may be conceded to the Vedantin that we have no
experience of attributes and actions without a substance, of

the universal without the individual, of the whole without

its parts. It may also be admitted by us that attributes,

actions, universals, etc ,

do not exist except in a substance.
But from this we cannot conclude that a substance is

identical with its attributes and actions, or that it is an

aggregate of the aspects of attribute, action, universal, etc.
Two entities may be inseparable and yet not identical, only

if they are different and distinct in our experience. We

cannot, indeed, have any experience of attributes without a
substance But the same experience tells us that a sub-
stance IS distinct from its attributes, and the attributes are
distinct from the substance. A substance is not an attribute,
nor a group of attributes. There can be no attributes
without some substance. Hence the existence of attributes

presupposes the distinct reality of a substance. That we

ascribe different sense qualities to the same substance also

shows that the substance is not identical with any of them,

but is distinct from them all. What we mean by a
substance is, therefore, different from what we mean by an
attribute. The substance stands for the ‘
continuant ’
or

the reality underlying the changing characters of a thing,

and the attributes stand for its properties or powers of

manifesting certain characters under certain conditions.

Although an attribute is not a substance, yet it is inseparably

 ORDINARY PERCEPTION AND ITS OBJECTS 191

related to or rooted in it. This inseparable relation or the

fact of subsistence of the attribute in a substance is

samavaya. The same reasoning applies mutatis mutandis to

the relation between substance and action, the universal and
the individual, the whole and the part, particularity and the
particulars In each case we have an inseparable relation
between two distinct and different entities. It is a relation

like samyoga or conjunction. But, unlike sainyoga which

isaqualit}, samavaya is a special relation that cannot be
reduced to any other category. It is not a substance because
it has no qualities. It is not a quality because it is not
limited to substances, nor does it qualify substances. For
the same reason it cannot be regarded as an action It is

not a universal nor paiticiilanty, because it is neither the

common essence of many things nor the peculiarity of
anything It is the objective fact of an inseparable connec-
tion between two other facts which are distinct and different

Hence we have to admit a separate category called samavaya

or the relation of inherence.

5 Perception of non-existence (ahhava)

The above categories of substance and the rest stand

for positive realities, whether existent or subsistent. The
category of abhai a stands for all negative or non-existent
facts. The>e facts are as real and objective as positive facts,

only they have no being. While positive facts are, i. e.

possess being, negative facts are not, i. e. possess non-

being.’ i4b/iaua or non-existence IS defined as that which
IS nut inherence and cannot be in the relation of inherence
to anything else (asainavayatre satyasamacdyah) It is

not inherence because it does not subsist as a positive fact.

It cannot have the relation of inherence to anything, be-

1 Abh»^vat\arr) dravyadi atkany<‘Dy5bbavavrtttvarn, , 12,

2 Sarvadarhnasafngraha, Chapter on Vaide^ika Ryatem.

192 NYiYA THEORY OF KNOWLEDGE

cause non-existence cannot be the constitutive element ol

anything and also because inherence is possible only be-

tween two positive entities

As to the nature of non-existence there are different

conceptions According to the Samkhya and the Prabha-
kara Mltnamsa, the non-existence of one thing in anothei
means the mere existence of the latter. For example,
the non-existence ol a jai on the table means the existence
of the table per se Hence non-existence is the mere exist-
ence ol a locus { adh ikara akaivaliiam ) The Nyaya objects
to this view and treats non-existence as an entity distinct’

from Its locus {atinktapadarLha). If non-existence were

the simple existence of the locus, it would become indistin-
guishable from existence The table exists as much when
the jar is non-existent as when it is existent on the table.

So we may speak of a jar’s non-existence on the table even

when it actually exists there Non-existence is not the
^
locus as such, but an objective character (vi^esana) of it

It IS not, like the colour of a table, a guna or attibute of the

locus. It is only adjectival to or a determination of the
locus which exists as a positive entity Non-cxistence or
negation is thus a real and distinct entity which is adj’ec-

tival to some positive fact. According to the Vedanta and

the Bhatta Mimamsa also, non-existence is an entity
distinct from the locus All objects h ive two characters, a
positive and a negative A thing exists positivel> in itself

and 13 characterised negatively by the absence of other

things in it. This absence oi non-existence of other things
is not the mere existence of the thing in itself If the

non-existence of a jar on the ground were the mere exist-

ence of the ground, we cannot speak of its non-existence

when there is a cloth on the ground and so no cognition ol

the mere ground. On the other hand, we may speak of the

1 TKD .p 60, SD.pp 83-84

t TO.. I, pp. 898f , 8M , 2 & 12.
 ORDINARY PBROBPTION AND ITS OBJECTS 193

jar’s non-existence even when it exists on the ground, for

the ground as such exists while there is a jar on it. Hence
non-existence is something distinct from its locus.'

Abhava or non'Cxistence is of four kinds, namely, prQga-

bhUca, pradhvams&hhava, atyantabhava, and anyonyubhdva.*
Some NaiySyikas, however, bring the first three under the
head of sathsurgabhdoa and recognise only two kinds of
non-existence.® SaiHsargabhava or the negation of a con-
nection is th it which is dififerent from anj/onyabftara or the
negation of identity. In the one the connection between
two things IS denied, in the other we deny the identity or
sameness of two different things or concepts.
PrQ,gabhS,va or antecedent non-existence is the non-exist-
ence of a thing prior to its production, e. g. the non-exist-
ence of an effect in the cause. So long as the effect is

not produced, it is non-existent in the cause. This kind of

non-existence is said to be without a beginning but not
without an end {anddih sdntah). It is subject to cessation
(vinasya).* The effect never existed before its production,
so that its non-existence has no beginning. Its production
at any time means the end or cessation of its previous non-
existence. Pragabhdva thus refers to the past non-exist-

ence of a thing and implies the possibility of its future

existence. Like the past it has no beginning but has an

end, since it ends just when the thing begins to exist. The
present existence of a thing ends its past non-existence,

just as the past has its end in the present.

Pradhvamsdbhdva or emergent non-existance is the non-

existence of a thing to its destruction, e.g, the ?

non-existence of an effect when it is destroyed. This kind
of non-existence has a beginning but no end (sadiranantah).

> SD.pp. 83-84.

• TS., p. 89 C/ VP., Ch. IV.
• NVT., BP ASM, 12
2. 2. 12 5

< 8M,12 jTB.p. 99:T8.,p. 89.

25— (IIITB)
194 NYIYA theory of KNOWLEDGE

It is subject to origin in time (janya), but not to cessation.^

When a jar is broken its existence ends, and its non-exist-
tence begins but can never be ended, because the same jar
cannot be brought back to existence. Pradhvamsdhhdva
may thus be said to refer to the future non-existence of a
thing in relation to its present existence. Like the future it

has a beginning but no end. It begins just when a thing’s

present existence ceases, and continues for all time to come
just as the future begins with the cessation of the present
and extends indefinitely forwards Thus while prdgahhdva
IS beginningless, pradhvamsdhhdia is endless. The one
is an infinite senes backwards, the other is an infinite

series forwards. The one has an end, while the other has
a beginning. That is, an infinite series may have an end
or a beginning. Kant in bis first ‘
antinomy ’
failed

to see this when he argued that what is infinite can

neither begin nor end. This antinomy may be solved in
the light of the Naiyayikas’ finding that some infinite

series have ends.

Atyantdbhdva or absolute non-existence is the negation
of a connection between two things for all time (mtyasamsar-
gdbhdva). It is subject neither to origin nor to cessation
{ajanydvindB).^ Thus it is both beginningless and endless.
This IS illustrated by such cases of non-existence as a hare’s

horn, the colour of air, etc. The non-existence of a

connection between horns and the bare, or colour and the air
is true at all times, past, present and future. Thus
atyantdbhdva or absolute non-existence is not a cipher
which IS the property of a general term that applies to no
object.
Anyonydbhdva or reciprocal non-existence is the negation

of identity or the difference between two things, e.g. a jar is

* Ibid.
^ Nityasaiiisargibbftvatvaa atyaDtabh&vatTam, SM ,
ibtd ;
traik&Uko'bbftro-
*tyanlibh&va]^f TB.» ibid.
 ORDINARY PERCEPTION AND ITS OBJECTS .19^

not a cloth.* A jar and a cloth mutually exclude each other,

and so each is non-existent as the other. Anyonyabhava

stands for this mutual negation of the relation of identity

between two things. Like absolute non-existence (atyanta-

hhdva), reciprocal or mutual non-existence {anyonyabhava)
is eternal, i.e. both beginningless and endless (ajanyavina^i).
But the distinction between the two is this. Absolute non-
existence has a material aspect. In it there is the affirma-

tion of something actual (.eg he hare or the air), and the

negation of a relation with regard to it (e.g. the relation

between hare and horn, or colour and air). On the other

hand, reciprocal non-existence is only a logical or formal
negation with reference to the relation of identity between
two things which need not be actual. That ‘
X is not Y,’
‘
a red star is not a blue star ’
is true even if no such things
actually exist. In absolute non-existence we deny the
connection between two things, while in reciprocal non-
existence the identity or sameness of two things or concepts
18 denied. The opposite of absolute non-existence is a con-

nection between two things, while that of reciprocal non-

existence is the identity between them ^ The proposition

‘
there is no colour in the air ’
implies the absolute non-exist-

ence of colour in the air. The opposite of this will be a

proposition which connects colour with the air, e g. there

IS colour in the air.’ The proposition ‘

a jar is not a cloth

implies the difference of the one from the other. The

proposition which identifies the
opposite of this will be a
‘

two, e.g, a jar is a cloth.’

As to how non-existence or negation is known by us,

there is a sharp difference of opinion among the different

Indian philosophy. According to the Bbat^a

systems of
Mlmamsa and the Vedanta, non-existence is known by non-

AnycnyflbhSvastii tidStoyapratiyogiko’ bbavut.

*bid
>

* TS and TD . pp 89 f
 niiVa theoki of knowleixjb

perception of what should have been perceived (yogyinu-

palahdhi). It can not be known by perception. Perception
requires sense-object contact. But there cannot be any
contact of sense with non-existence or negation. It is

not, in some cases at least, known by inference. When

we know the non-existence of a jar on the ground before
us, we have a direct knowledge which is not mediated by
any inferential reasoning. In such cases our knowledge of
non-existence comes from non-perception as a distinct

source of knowledge. ‘ As we have already said, non-existence

IS, according to the Bhat^a Mlmam^a and the Vedanta, both
an objective character of things, and a character of the
presentation of things. But it is not perceived like the
whiteness of snow or the redness of a rose. On the other
hand, we have an immediate feeling of it as a character of
the presentation just when we have that presentation. This
subjective feeling of the presented character as distinct or
the discriminative feeling of it is what we mean by our
knowledge of it. Anupalabdhi is this subjective feeling
and IS an independent source of the knowledge of non-
existence.
The Vai^esika and the Prabhakara school bold that

non-existence is known by inference. According to thi

former the non-existence of the cause is inferred from th(

non-existence of the effect, just as its existence is mferreu

from that of the effect.* In the Prabhakara Mimaihsa also

non-perceptiou is not regarded as a distinct source of the

knowledge of non-existence. Bather, the non-perception of
a thing is the condition from which we infer its non-
existence.* The Samkhya * and the Nyaya system agree in
holding that non-existence is known by perception. Accord-

> VP .Ch. VI; SD.pp 88-87

• P9 ,p 111 ; Upoikaratp, 928
» SO., pp hS f.

* TKD , pp. (K)-51,

 ORDINARY PERCEPTION AND ITS OBJECTS td7

ing to both, non-existence is an objective deterrainatioii

(parindma or vi^esana) of some positive entity (JbhStapadSr*
tha) and can be perceived by the senses. Tbe Nyaya
explains the perception of non-existence as due to an indirect
sense-object contact, called vi4esanatd. The non-existenee
of a jar on the gronnd is an adjective or determination
(vtiesma) of the ground which is its substantive (vUesyd).
This non-existence comes m contact with our sense through
being adjectival {vUesana) to the ground which is in con-
junction {samyul'ta) with the sense Thu® non-existenee
is perceived as an adjective or objective character of some
positiive thing which is in contact with our sense. ‘ That
for the knowledge of non-existence there must be a contact
of sense with its locus is also admitted by the VedSnta.
And, that the knowledge of non-existence is aided by non-
perception is admitted by the Nyaya. We know that a jar
does not exist on the ground when we feel that it must have
been perceived if it were on the ground.* Thus the Nyaya
and the Vedanta agree in holding that to know non-existence
there must be a perception of the locus and non-perception
of what does not exist in it. But while the Nyaya takes
non-perception {anupalahdhi) as an auxiliary condition
(Jcarana) of the perception of non-existence, the Vedanta takes
it as the unique cause (karam) of the knowledge of non-
existence. For the Nyaya, the senses which perceive the
locus do also perceive non-existence as a determinant
§ana) of it. For the Vedanta, the senses are solely concent'

ed in the perception of the locus and do not go further than

that And while the senses perceive a certain locus, the
ground, they do not perceive a jar on it This non-percep-
tion, therefore, of what might have been perceived if it

> TB . p 6 . TC., I. pp «7t.76,

* Yadi syftdupalabliyetetyavaiM yatia pra^ajyate, BP., 6^2 , yogy&cupaiabdhyl
abbftva^ pralyalcfa^i TIC.
 198 NYiYA TUEORY OF KNOWLEDGE

existed, is the source of our knowledge of the jar’s non-

existence.^

Ansong Western thinkers it is now generally recognised

•
that negative facts are as real as positive facts. But there
.
is much difference of opinion as to how negation or non-
existence is known. According to Alexander, ““
‘
negation is

not merely a subjective attitude of the rnind. That is only

an instance of negation, in the region of mental acts. Nega-
tion or negativity is a real character of things, which means
exclusion or rejection. Not-white is the character which
excludes or is different from white.’ This then would sup-
port the Nyaya view that non-existence is perceived as a
determination of some positive entity. This seems to be
implied also in the '
lew of negation held by Bradley and
Bosanquet. Bradley* says : “The affirmative judgment
qualifies a subject by the attribution of a quality, and the
negative judgment qualifies a subject by the explicit rejection
’’
of that same quality According to him, ‘
the truth of the
negative lies in the affirmation of a positive quality In “ A
is not B’’ the real fact is a character x belonging
to A, which is incompatible with B. The basis of

negation is really the assertion of a quality that excludes

(x). It IS not the mere assertion ol the quality of exclusion
(not-B).’ So too Bosanquet^ holds that affirmation is prior

to negation as supplying the reality within which alone

negation has a meaning. In this sense the non-existence
of a jar on the ground will be equivalent to the existence of

the ground as such. It is to be observed, however, that

while the ground is perceived, its exclusion of a jar is not
so perceived. The fact of there being no jar on the ground
may be a given fact, but it is not given by way of sense

I Vide TC., I, Ch on Anupalabdhi , VP ,

Ch. VI.
* SpacBf Tme and Detiy, p. 200
^ LfigWf Vol. I, pp. 1 16-17
* Logtc^ Vol I, p 261
 ORDINABY PERCEPTION AND ITS OBJECTS 199

perception. As Russell ’
h;is said, negative facts are real
but not sensible facts. That
‘
A is not B,’ or that ‘
a jar
does not exist on the ground ’
is raore primarily a subject
ive feeling of privation with regard to A, or the ground,
than a perception of any sensible fact. Hence the Vedanta
seems to be nearer the truth when it takes non-perception
(anupalabdhi) as an independent source of the knowledge
of non-existence or negation.

6. Internal perception and its objects

Internal perception is due to the internal sense or

manas. Hence it is called mdnasa or dntara pratyahsa.
It IS the knowledge of mental facts brought about by their
contact (sannikarsa) with the inner sense or manas. Thus
mdnasa or internal perception is, like introspection, the
source of our direct knowledge about mental or subjective
facts. But while modern introspectionists take introspection
to consist in the mind’s knowledge of its own contents, the
Naiyayikas treat internal perception as knowledge of certain
subjective facts other than, but due to, the mind as a sense.

Generally speaking, the self and its contents are the objects

of internal perception These are perceived when they

come in contact with manas or the mind. In introspection
the mind or self turns back on itself and perceives what is
going on there withjut requiring any sense. The
Naijayikas, however, like the older introspectionists, believe
that the self requires an “ inner sense ” to perceive psychi-

cal facts, just as it requires the external senses to perceive

external objects.
Among the objects of mdnasa or internal perception the

Bhasapariccheda mentions the feelings of pleasure and pain,

desire and aversion, cognition or knowledge and all kinds of

1
Analysts of Mtnd, p 276^
2D0 NYAYA THBOBY OF KNOWLEDGE

mental effort or volition.^ To these we may add the univer-

sal of each of these attributes, their non-existence and
inherence in the self, and the self itself “ All of these are

perceived when there is contact (sannikarsa) in some form

or other, between them and the internal sense of manas
I#et us now consider the process involved in the perception
of these objects.
According to the Nyaya, pleasure and pain, desire and
aversion, cognition and volition are attributes of the self.

Their relation to the self is one of inherence (samavaya).

They are perceived when the mind as a sense comes in

contact with them. This sense-object contact is not one of

direct eon)unction (samyoga). It is an indirect contact
called sa&iyukta-samavaya. Pleasure, pain and the rest as
particular facts, come in contact with the mind through their

inherence (samavaya) in the self which is conjoined

(samyukta) with the mind. Similarly, the universals of

pleasure, pain, etc., are perceived through that kind of

indirect sense-contact which is called samyukta-samaveta-
samaaaya. The universals of pleasure and pain (sukhatva-
duJ^khatva) subsist in particular pleasures and pains by
way of inherence (samavaya). The p.irticular pleasures
and pains exist in the soul as its inherent attributes
(^amaveiagundh) Hence the mind comes in contact with
the universals of pleasure and pain through their inherence
in what inhere in the soul which is conjoined to the mind.
In perceiving any particular pleasure or pain we do perceive
its pleasurableness or painfulness quite as directly, although
the process of perception is more mediated and complicated.*
So also, we perceive that pleasure, pain, etc , inhere in the
self so long as they exist or are present. And just as we
perceive their existence so also we perceive their non-exist-

^ MaDogrfthyain snkbadi da^^khamiochft dveyo BP., 57.

» TKm p. 9.
> Vidt 8M , 57 1 TB., p. 6,
 ORDINARY PBRCRPTION AND ITS OBJECTS 201

ence or absence. That ‘

I am unhappy,’ or - I have
ceased to be angry ’
is as much a matter of direct percep-
tion as that ‘
I am happy,’ or ‘
I am angry.’ The procest
of perception is, however, somewhat different. The percep-
tion of the inherence (samaoaya) of pleasure and pain, a!
also of their non-existence (ahhdca) in the self is mediated
by the indirect sense-object contact called vi§esanatd. Both
the inherence of a present pleasure and the non-existence
of a past one are determinations {viSesana) of the self.

They are perceived when the mind as sense comes in

contact with them through its conjunction with the sef
which has those determinations.^ It is only in the percep
tion of the self that there is a direct sense-object contact
The self as a substance comes in actual contact (samyoya
with manas or the mind as another substance, and thereby
becomes an object of internal perception.* It cannot h
perceived by the external senses, since it possesses neithei
a limited dimension (mahattva) nor any manifest (udhhuta,
colour or touch.* According to some Naiyayikas, th{
pure self cannot be an object of perception. The self is
perceived only as related to some perceptible attribute like

cognition, pleasure, etc. We do not perceive the self as

such but as feeling or knowing or doing something.

Hence the self is perceived through the perception
of this or that state of consciousness. While one’s own
self can be perceived, other selves can only be inferred
from their bodily actions or behaviours.^ According to the
Vedanta, pleasure, pain, desire, aversion and volition are
perceived, but their perception requires no sense organ
like manas or the mind. They are the different parts oi

1 TB.. p 6.

* Manabantareijiendriye^a yadatmavj§ayakam jfiaDaib janyate* hsmiti tadfi iDao

indriyamatmarthalj, Ibtd,
S BP & SM ,
49-50

Ibtd

26— (1117 B)
202 NYiYA THKORY OF KNOWLEDGE

aspects of the antahkarana. As such, there is a natural

identitication between these and the antahkarana or the
mind. This identification means a perception of all

that IS identified with the antahkarana. In short, mental

states are perceived facts because they are mental, and so
do not require any sense to perceive them.’’
As to the question how cognition or knowledge is known,
there is a sharp difference of opinion among philosophers.
Some thinkers who fieny the possibility of introspection
would s.iy that knowledge can never be known. This is the
position taken up by Comte, Dunlap and others. Comte
thought that knowing cannot be known, since it involves a

division of the mind into two parts, which is impossible So

too, Dunlap, in his article “ The case against Introspec-
tion,”^ urges that there IS a dualism of subject and object,
that the subject can never become object, and therefore there
can be no awareness of an awareness. He says •
“Knowing
there certainly is ; known, the knowing certainly is not.”
Again he says :
“ I am never aw.ire of an awareness ” But
if this IS so, how do we know that there is any knowledge
or awareness at all ? Dunlap says that it is
‘
by being aware
of something ? ’
This means that when I am aware of some-
thing, I am aware To know something
of being aware of it

16 thus to know that something is known. Hence it cannot

be denied that knowledge is somehow known, be it by intro-
spection or not. As Russell ® has pointed out, ‘the statement
“ I am aware of a colour ” is assumed by Dunlap to be known
to be true, but he does not explain how it comes to be
known.’
Hence the next question is How is it that knowledge
is known ^ x^ccording to the baihkh) a, the Prabhakara
Mimamsa and the Advaita Vedanta, knowledge is known by

» VP Ch , 1.

* Psychological Review 8«pt., 1912

s Analysts of Mind^ p 1 1^,
 ORDINARY PERCEPTION AND ITS OBJECTS 203

itself: Cognition or knowledge is a conscious fact and it fe

the very nature of consciousness to be aware of itself. The

point has been elaborated by the Prabhakaras m their theory
of triputlsamvit or triune perception.' According to it, every
knowledge manifests itself at the same time that it mani-
fests an object and the knowing subject. It i s at once a
manifestation of three things, namely, knowledge, the
object and the knower. The Jamas also take a similar view
with regard to the nature of knowledge. The AJvaita
Vedanta takes knowledge or intelligence to be the essence
of the self, the very stuff of it. As such, knowledge is self-

manifest and self-shining {simprakd§a) It does not require

anything else to manifest or know it On this view, every
cognition is self-cognised, and consciousness is full and
complete awareness of something by a self. But that ‘
every
knowledge is self-conscious knowledge,’ or ‘
to be aware of
something is also to be aware of that awareness ’
is a propo-
sition which IS not borne out by psychological facts. Sub-
conscious or unconscious experiences of the mind cannot be
said to be full and explicit awareness of themselves. Fur-
ther, as Bussell * has remarked, it is highly probable that
children and the higher animals are aware of objects, but
not of their own awareness.
According to the Bha^ta Mimamsa,^ knowledge cannot
be directly known. We can never know any knowledge
immediately by itself or by any introspection called internal

perception. That we have an awareness or a knowledge of

some object is no doubt a matter of knowledge for us. But
this latter knowledge is not at ail immediate and perceptual
knowledge ; it is only mediate and inferential knowledge.
When we are aware of something, it comes to have the

1 Vtdt Prakara^pafietki, p 69.

J Vide VP Ch,
1.

3 Analyn) of Mtnd, pp 116-16.

* Vide SB , pp 66-57
204 NYiYA THEORY OF KNOWLEDGE

character of ‘
being an object of our knowledge ’
(jMtatd).
But how can a thing have this character of ‘
being known,’
unless there was previously some knowledge ot it ? Hence
’

from the character of ‘

being known ’
or ‘
being cognised
in the known object we infer the antecedent existence of

knowledge or cognition. Thus knowledge is neither self-

manifested nor directly perceived, but inferred from the

character of ‘knownuess ’
or ‘
cognisedness ’
{jndtatd) in the

object that has been known or cognised, The Naiyayikas

reject this view on the ground that ‘
knownness ’
cannot be
a character of objects, for objects acquire no character from
their relation to knowledge.
The Naiyayikas, as we have already seen, hold that
knowledge is known by introspection oi internal perception
(mdnasa pratyahsa). According to them, cognition or
knowledge manifests its objects, but not itself. It points
bejond itself and can never be directed to itself. Hence
cognition or knowledge cannot be self-manitested. It does
not, however, follow that knowledge cannot be at all known
or manifested. Just as an object is manifested by a cogni-
tion of it, so one cognition is manifested by another that
follows it and makes it an object to itsell. First there is the
cognition of an object (lyavasaya) and then another
cognition coming after it cognises the first, t.e. there is an
after-cognition (anuvyavasdya) of the first cognition.' It

follows that every cognition is not necessarily cognised,

that awareness of an object is not always an awareness of
itself. It is only when the self or mind attends to, and
casts an introspective glance at it, that one cognition or
knowledge is known or perceived. This view of the Naiya-
yikas has the support of many modern introspectionists like
Stout, Laird and others. Thus Stout observes “ Psychical :

states as such become objects only when we attend to them

» TR , p 53 , TD , P 32,
 ORDINARY PBRCRI’TION AND ITS OBJECTS 206

ill an introspective way. Otherwise they art not themselves

objects, but only constituents of the process bj which objects
are cognised.” ^
So too, Laird says :
” Certainly, our cogn^
tive processes are, in their usual exercise, processes with
which (not at which) we look ; and none of them, perhaps,
can look at itself. It does not follow, however, that
another (introspective) look cannot be directed towards this
process of looking, . . This means that one cognition is

known by another by way of introspection. But for the

Naiyayikas, introspection involves a peculiar difficulty. It

supposes the simultaneous presence of two cognitions, which

IS not admitted by the Naiyayikas. Hence we are to say
that the cognition, which is cognised by another cognition,
IS past in relation to the second cognition which is present.
This implies that introspection is really memory or retro-
spection of what is past. But there cannot be any
memory without a previous perception corresponding to it.
Hence we are committed to the view that every cognition
somehow cognises itself. It may not have an explicit
awareness of itself, but only an implicit or vague feeling of
its presence. As Stout has elsewhere said :
“ The stream
of consciousness feels its own current.” Hence the way
in which cognition or knowlpdge (or for the matter of that,
the mind) knows itself is quite different from that m which it
know's an object external to itself. This has been very well
recognised by Alexander in his distinction between an
enjoying and a contemplating consciousness. He says that
‘
in any experience the mind enjoys itself and contemplates
its object, that the mind is not a contemplated object to
itself, and that introspection is not contemplation.'^ Hence

' Manual of Psychology f p 184

* Contemporary BnUsh Philosophy, Fir»t Series, p. 227,
* Analytic Psychology, Vol I, p. 160
* Space, Time and Deity, Vol 1, pp 12-17
 NiiyA THEORY OF KNOWLSbO^ •

we conclude tliat knowledge is known directly by itself.

This knowledge of knowledge however is neither an explicit

manifestation nor a definite perception of it, but a feeling

or an enjoying consciousness of itself.

 CHAPTEB IX

rHEEE MODES OF ORDINARY PERCEPTION

1. NirviMpaka and savikalpaka perceptions

As we have noticed in a previous chapter, perception has

been divided by the Naiyayikas into two broad classes,

namely, laukika or the ordinary and alaukika or the extra.-

ordinary. This division depends on the nature of the

sense-object contact that is involved in all perceptions.
Ordinary perception again, has been divided into the six
kinds of olfactory, gustatory, visual, tactual, auditory and
mental perceptions. Such classification of ordinary percep-

tions has reference to the senses concerned in perception.

According to another classification, ordinary perception is of

two kinds, namely, nirvikalpaka or the indeterminate and

savikalpaka or the determinate. Here the principle of
classification is the character of the perceptual knowledge
which arises from sense-object contact. To these two kinds

of perception we may add pratyahhijM or recognition as a

special form of determinate perception. Thus keeping in
view the nature of perception, the NaiySyikas distinguish
between three modes of ordinary perception, namely, the
nirvikalpaka, the savikalpaka and pratyahhijM. Extra-
ordinary perception being explicit and definite knowledge,
’
has but one mode, namely, the savikalpaka or determinate.
While the distinction between nirvikalpaka and savu
kalpaka perceptions is generally recognised in Indian
philosophy, there is much difference of opinion, among the

I tu , ,
sayikalpakamevn, NyHyakoSut p.
208 NYiYA THEORY OF KNOWLEDGE

different systems, as to their exact nature and validity.

The grammarian philosophers (dahdikas) along with others
take the extreme view that all perceptions are saviltalpaha
or determinate, since every perception must be expressed
in a verbal proposition ‘and ns consequently predicative in
its clucactec. This is met-by- another, extreme view,. held
by the Huddhists and some V’edantists, that mfvtkaJpaka or
indeterminate perception alone is valid, while savikalpaka

or determinite perception is false knowledge Between

-these ^ two extremes we may place the other systems of
philosophy which accept both ntrvikalpaka and savikalpaka
perceptions as true knowledge^ Thus among the different
theories of perception in Indian philosophy there seems to
be a sort of gradation from tlie most abstract to the most
jeOncrete \iew of percepUon.

In the A.dvaita Vedanta system we seem to have the most

abstract view of nirvxkalpaka perception. According to it,

nirjvikalpaka is the knowledge of pure bei.ng (sanmdtram}.

It 18 a cognition of the ‘
this’ or the existent as such, but
not as determined by anything. To aletermine a thing is

to characterise it by this or that quality and is thus to

diStmguisli it from other things having different qualities.
All tins, however, is the work of discriminative thought

(vikalpa). P/ior to discrimination tliere cannot be any

cognition of an object as such-and-such, i.c as a deter-
minate reality. Hence ntrvikalpaka perception must be the
cognition of pure indeterminate being. It is m savikalpaka
perception that an object is determined by certain qualities
and is distinguished from different objects. But all objects

being ultimately one undifferentiated uniti of intelligence

{caitanya), their distinction or difference is only an

appearance. The view of the world as a plurality of
independent reals is not only opposed to scriptural testimony,
but also logically untenable and self-contradictory. Hence
we are to say that savikalpaka gives us a knowledge of
 THREE MODES OF ORDINARY PERCEPTION 209

appearances, while nirvikalpaka gives us the ultimate

^
truth.
The epistemology of the Neo-Advaita Vedanta is m
agreement with the general metaphysical position of the
earlier school. From the vydvaharika or practical standpoint
it makes a distinction between saoikalpaka and nirvikalpaka
perceptions. According to it, the saoikalpaka is the know-
ledge of a tiling as qualified by an attribute (vaiHHydvagdhi).
In it the thing is related as subject to the attribute as its

predicate Hence the essence of saoikalpaka perception is

the determinatiop of its object by way of predication. The

perception of an object as a jar is saoikalpaka, since the
quality of ‘ jarness ’
is predicated of the object. On the
other hand, nirvikalpaka is the knowledge of a thing as not
related to anything, not even to itself {samsargdnavagdhi).
It is the knowledge of a thing as the identical excluding all

relations. Hence in nirvikalpaka perception there cannot be

any predication of the object in terms of the subject-predicate
or substantive-adjective relation {vi§esya-viiesana-samhandha).
As instances of such nirvikalpaka or non-relational know-
ledge, the Vedantist mentions the propositions ‘
this is that

man,’ that art thou.’ When in the presence of a man

we have it said ‘
this is that Devadatta,’ we have the
perception of Devadatta as the same identical man. This
knowledge of the man is nirvikalpaka, since its object is not
qualified by or related to anything. What we apprehend
here is the simple identity of the man in the past and the
present, but not his relation to any time, space, or anything
else. It IS true that the man may have certain attributes
and IS related to this or that time and space. But in the
perception in question we are not at all concerned with his
relation to any attribute, but only with his identity as such.
It may be urged here that since the knowledge is grounded

1 Vide RSm&niijA’s commentarv on the Veddnfasufra, 1 1.1 , BD . p dO,

27— (1117B)
210 NYiYA THEORY OF KNOWLEDGE

on a proposition (vdkyajanya) it cannot but be relational

in character. A proposition is the assertion of a relation
between two things, of which one is the subject and the
’

other predicate. Hence the knowledge ‘

this is that man
must be savikalpaka or predicative. To this the Advaita

Vedantist replies that a proposition which asserts the

identity of an object does not come under the law of predi-
cation. When we say ‘
this is that man we do not ’

predicate ‘
thatman of ’ ‘
this ’
It is not the case that we
relate ‘
this man with ’ ‘
that man,’ but simply assert the
identity b?tween the two. An identity proposition (akhan-
dartha vdk'ja) thus gives us non-predicative or non-relational
knowledge {mrvikalpaka jMna).^ Thus we see that,

according to the Advaita Vedanta, nirvihaJpaka perception

is a judgment of identity expressed in a proposition The
identity that is perceived in niroikalpaka does not pertain
to any of the specific attributes or parts of the perceived
object. It refers only to the unity of the object as an
unrelated essence, i.e. as pure being. But while the
‘
identical ’
perceived in mrvikalpaka is an abstract unity of
being, our knowledge of it is a propositional judgment of the
non-predicative order {akhandarlha vQkya). The Veiantist
further holds that nirvikalpaka perception is self-manifest
or self-conscious knowledge. It is perceived by itself

ipratyaksa), and does not require any other knowledge to

manifest or perceive it. It follows also that we have
first the savikalpaka perception of an object as related to
certain qualities and then a nirvikalpaka perception of it as
a unity that remains identical with itself under different
conditions.

According to the Buddhists, nirvikalpaka is the only type

of valid perception. It is such cognition of an object as
contains no element of thought or ideation in it (kalpand

» vp.,ch.i.
 THREE MODES OF ORDINARY PERCEPTION 211

•podham)} Ordinarily, knowledge involves two elements,

namely, the given or the sensed and the meant or the
ideated. The Buddhists hold that what is given is a unique
individual (svalaksam) that belongs to no class and is not
related to anything.* We may call it by a name, bring it

under a class and think of it as having certain qualities,

actions and relations. But its name, class, quality, action

and relation are not any part of what is directly given ; these
are the contributions of our mind (kalpand) to the given
experience. Hence nirDikalpaka perception is a cognition
of the given datum as such, i. e. as not modified by any

idea or concept like those of its name, class, etc. (nama-

]dtyadyasamyutam). It is a pure sensation of the simples of
experience and does not lend itself to any verbal expression
(abhildpasamsargayogyapratibhdsam). As contrasted to

this, savikalpaka perception is a verbalised experience, in

which the object is determined by the concepts of name,
class, relation, etc. Here we think of the object as a complex
of parts and attributes, bearing a certain name and having
certain relations. Such knowledge, however, is false, since

it IS not due to the given object, but to our conceptual con-

struction of it.* Tlius the Buddhists reduce nirvikalpaka to

pure sensation which is valid but blind, and savika’paka to

conceptual knowledge which is definite but false. Be it

noted, however, that the Buddhist’s nirvikalpaka as a cogni-

tion of the simples or unique individuals of experience is

less abstract than the Advaitin’s nirvikalpaka as a cognition

of pure being.
In the Mimamsa, the Samkhya and the Nyaya-Vai^e^ika
system we have what may be called a concrete view of

perception. According to these realistic schools, what is

1 Pramanasamvccayat Ch I ; NBT pp 9 f.
» Cf “ Apare tu svalak^apamatragocaraifa nirvikalpakamicchanti,” SD.,p, 41.
3 NVT 1. 1 4, MadhavacSryya, Sarvadarhnasathgtaha Chapter on Baaddhil
,

philosophy Vide also NM , pp 92 f.

212 NYiYA THEOHY OF KNOWLEDGE

given m perception is not, as the Advaitins think, a pure

unity of being or the abstract identity of the ‘
tins ’
and
‘
that ’
of experience. Nor is it, as the Buddhists suppose,

the abstract individual, the bare particular or the mere

‘
this ’
of experience. On the other hand, it is held in

these systems that any perception, mivikalpaha or savihal-

paka, IS a direct cognition, of the real individual which is a

unity of the universal and the particular The distinction

between mrvikalpaka and savikaJpaka is not due to any

difference in the contents of perception, but to the way in

which the same contents of expeiience may be ordered and

arranged. As w'e shall see, however, there is some difference

of opinion as to the nature and structure of mrvikalpaka

perception between the Bhatta Miaiamsa and Samkh}a
systems, on the one hand, and the Piahhakara and Nyri>a-

Vai^esika s} stems, on the other

According to the Sarnkhy.i and the Bhatta Miraamsa,
both niTixkalpaka and savikalpaha aie equally valid and
necessary modes of perceptual knowledge By mrvikalpaka
they mean that cognition which spontaneously niises at the
first moment of contact between sense and object '
It is a

knowledge of the object as one individual whole of generic

and specific attributes. There is no differentiation betw'een
the universal and the paiticular that are combined in the
body of the individual Hence there is only an apprehension
of the individual as an indefinite object (sammugdhaiastu-
matra), but no definite understanding of it as this or that

kind of object (vasluvi^esa) ^ Nirvikalpaka perception thus

resembles the perception of the children and dumb persons.
Like the latter perception, it is a simple apprehension of
an object as something, but not as this or that kind of

1 Ak^a'^anDtpatSn tntarft£Davivjktasainanyavi^e?avibbaga£D sammugdhavaBtumftbra-

goraramalocanajfianarjj, SD , p, 40,

2 Sdinmugdham vastumatraDtu praggrhpantyavikalpitarri, tatsanianyavii^eeabhyftm

kalpayanti mantsipal;, TKD ,

27.
 THREE MODES OF ORDINARY FliRCEPTlON 213

thing. So, likewise, it cannot be expressed in words, i. e.

embodied in propositions. Thus nirvikalpaka perception is

a simple apprehension, in which the ‘

this ’
of experience

is brought under the general idea of ‘

something ’
(vastusd-
mdnya). It is a judgment without words, i e. a non-propo-
sitioual judgment.^ That nirvikalpaka is areal mode of percep-
tion appears from the lact that it is the basis of what we call
reflex actions and savikalpaka perceptions. When a man
suddenly withdraws his Anger from a pricking pin, we cannot
say that there is in him a clear understanding of the pricking
object as a pin. Yet without some knowledge of the object
his action remains unintelligible. What prompts the man’s
action in this case is properly described as nirvikalpaka per-
ception of ‘
something pricking him.’ Again, savikalpaka
perception is understood by all as the predicative know-
ledge of an object {vaihstydvagdhi jndna). In it the homo-
geneous indefinite object of nirvikalpaka perception is ana-
lysed into the universal and the particular (sdmdnya-vi^esa),
and the two are then related by way of predication. Thus
when I perceive an animal as a cow, my perception is savi-

kalpaka, since I definitely predicate ‘

cowness’ of the animal
before me. Now such predication is possible only through
comparison of the animal in question with other objects of
our past experience and recollection of their respective class-

names. But the impetus to the acts of comparison and

verbal memory must be given by a previous cognition of the
animal as some kind of thing. This first nirvikalpaka cog-
nition of the animal as something {vastusdmdnya) is neces-

sary for the subsequent savikalpaka understanding of it as

this particular kind of thing {vastuvUesa) , i.e. as a

cow. The indefinite individual of the first experience is

analysed into ‘
this thing ’
as a particular, and ‘
cowness ’
as a

1 Ihti.
 214 NYiYA THEORY OF KNOWLEDGE

universal in relation to it and qualifying it as a predicate.

This IS the saoikaJpaka knowledge of the thing in terras of
the substantive-adjective relation, and it is expressed as a
verbal judgment of the form :
‘
this is a ow.’ ^ Thus we
see that, according to the Simkhya and Bha$ta MimamsS
scheols, niroikalpaka and saoikalpaka are but two stages in
the perception of the sime fact. The former is a stage of
unverbalised judgment of an object as an individual whole,
while the latter is a verbal judgment of it by way of predi-
cation.

In the Nyaya-Vai^asika and Prabhakara Mimaihsa sys-

tems we find a theory of perception which is in substantial
agreement with that just explained above. According to
the Nyaya-Vaides'kas, ordinary perception is of two kinds,
namely, niroikalpaka and saoikalpaka, both of which are
equally valid and grounded in reality.^ They hold that
niroikalpaka is not merely a cognition of the bare particular
(soalaksana), since it manifests the universal {samanya) as

well. If the universal is not cognised at the niroikalpaka

stage, our knowledge of it at a subsequent saoikalpaka stage
becomes inexplicable. But for a direct knowledge of the
universal as a unity of the particulars we could not recognise
them as similar and group them together into one class. The
universal cannot be constructed by our mind at any stage
of our knowledge, unless we start with it as a directly given
fact.® Nor again, is niroikalpaka a knowledge of the abstract

1 SD. & sc., p 40 , TED , 27

^ Some commentators on ilie Ny^ya system think that the distinction between
mrvtkalpdka and aomka/paka is not recagnis3d in the Sutr i, Bh&^ya and V&ntika,
and that It was introduced into the a philosophy by later logicians V5''a8pati

Mi^ra however, in bis T dlparya\ika {p 125) traces the distinction to NySya-Sutra,

1.14 Following his t acher, Frilocana, he takes the words avyapade^yam and vya-
tasUyatmdkam contained in this sutra to mean respectively ntrvtkalpaka and savt-

kalpaka perceptions Later Naiyayikas follow Vftcaspati in this interpretation of the

sutra and hold that it distinguishes between ntrvtkalpaka and savtkalpakn as two
kinds of ordinary perceptions.

3NM.P 98
 THREE MODES OF ORDIUARY PERCEPTION ‘216

universal (sdnianyamUtra), or of pure being {sattd). Our first

cognition of a thing is not merely an apprehension of its

unity, but also of its differences. When we perceive anything,

its mamfoldness is as much manifest to consciousness as its
unitary character. It is cognised as a unity of many parts,
qualities and a-’pects. Likewise, if nothing but mere
being (saltd) be the content of immediate apprehension, we
do not know how to account for our knowledge of the parti-
culars {vxiesa) of experience. Further, pure being which
is nothing in particular cannot be the object of our know-
ledge ’
Hence -we are to admit that in mrvtkalpaka per-
ception there is a cognition of both the universal and the
particular, the generic and specific properties of an object
as such ^ It is a knowledge of the perceived object with all

the wealth of its concrete characters in themselves. It cog-

nises the universal or the class-essence present in the object
of perception as well as its colour, form, structure and other

specific characteristics. Thus the nirvikalpaka or indeter-

minate perception of an orange is the cognition which is

produced immediately after the contact of the senses with

the object, and which manifests its generic and specific
properties in their isolation It gives us a knowledge of
the orange, not as orange, but as the grouping of a certain
colour, taste, smell, etc., with a certain universal called
orangeness.® But while both the universal and the particu-
lars constituting an object are cognised in nirvikalpaka
perception, they are not brought under the substantive-
adjective relation (vUesyaviiesanasambandhdnavagdhi).* In
nirvikalpaka perception these are cognised as unrelated units

1 Ibid

* Sftmfinyuvi^e^efu svarupalooanamStrauo prnlyaksaai, PS., p. 187 , nirvikalpakaiu

. . . a&na&ujam cobh tyaaiapi gfhQ&ti, NK., p 189,

5 SM.,58.
^ Avyapatlelyam jfttyAdisvarupftvagahi oa tu jfttyadinatn imtbo vi^e^a^avjficftya-

bbavavagfthTti, NVT ,
p 185,
216 NYiYA THEORY OF KNOWLEDGE

of reality.. Here then the object of perception is not

known as a particular related to a certain universal. It

is not judged as an individual belonging to .i certain class

and bearing that class-name Hence niroikalpaka is the
knowledge ot an object as not characterised in any way
{nisp^ akaraha) It is an apprehension ot ilie object as
‘
something,’ but not as related to a class and called by a
xiiuue{nama]dtyddtyo}andhtnam . . . kincididamiti) Hence it

IS not a judgment ol the object in terms of the subject-

predicate relation (vaihstydnavagdhi). On the other hand,
it IS a simple apprehension of the existence and attributes
of an object without any corresponding judgment of it in

words, or by way of predication

Accoiding to some linguistic thinkers there cannot be
any nirvikalpaka perception in the sense of an unveibalised
experience a-, e-tplained above They hold that we cannot
think things except through words. All objects are in-

separably connected with the words by which they are

denoted. To cognise a thing is to know it as such-and-such

and so to relate it to a denotative word {odcakahbda). Like-

wise, we can act in relation to a thing only when we know
it precisely as ot this or that kind, i. e. determine it by
means of a class-name. In fact, all our cognitions are
embodied in veibal propositions, such as ‘
I know a colour,’
‘
I have a taste,’ ‘
it is a smell,’ and so on. All cognitions

being thus inseparable from verbal expressions, there can

be no mn i/ca/pa/ca or unverbalised cognition* The Nyaya-
Vai^esikas repudiate the linguistic contention on the follow-
ing grounds. In the ca-e ot children and dumt) persons
there is undoubtedly a knowledge of many objects, but no
verbal expression of that knowledge. Even in the case

1 Nlmaj&tyadiyo'aDarahitain vai^i^^y&Davagahi nisprakarakam nirvikalpakaui,

TC . I. p 809.
^ Nirvikalpaka ID vn^ei^faprakar&dirabitaiii yaatvavarupamitrajfiAnam, TM.| Ob II.
3 NB„ U 1.4, NK., p 1B9.
 THREE MODES OF ORDINARY PERCEPTION 217

of grown up persons, who are in the habit of expressing

their thoughts in words, there is such a thing as sensation
which is a bare apprehension of something, but no verbal
judgment of it. The first stage of perception is a sensory
cognition arising just with the contact between sense and
object. Like the perception of the child or the dumb person,
it does not require and has not the time to develop into
verbal judgment The verbal expression of sensory cogni-
tions is a later stage of perception, which serves the
purpose of sociil intercourse and com nunication. The
linguistic contention that objects are inseparable from their
corresponding words leads to absurd consequences. If it

were true that all objects are inseparably connected with all

the words denoting them, even children and idiots should

know all those words when they perceive the objects. It

follows also that a man who perceives the words ‘

colour,’
‘
sound,’ etc ,
should have a knowledge of the objects denoted
by them, even though he may be deprived of their special
sense organs. Hence the Naiyayikas conclude that all cog-
nitions need not necessarily be verbally expressed knowledge.
x\s a matter of fact, w'hat enables us to recall the words
with which an object is associated is a previous nirvikdpaka
cognition of it as an existent fact. Our first experience
of an object is a simple apprehension of its existence apart
from any verbal association. Such simple unverbalised
experience is the ground of our subsequent judgment of it

in words or predicative propositions. Hence nirvikdpaka

is a real stage of perception.*
According to the Naiyavikas, nirvikalpaka is a real

but not a perceived fact {atlndriya)} It is a conscious,

but not a self-conscious state. The Naiyajikas hold that to

be self-conscious means, for a conscious state, to be per-

I NB and NK ,
tbtd.
^ Jannirvikalpakbyath tadattodnyamifya-tei BP, and 8M , 58,

28— (1117B)
IS OTiYA THEORY OF KNOWLEDOE

‘ived by another state of consciousness. In self-conscious-

ess (anuvyavasiiya) one cognition is cognised b> another
hich follows the first and appreliends it as an object to

self. But there cannot be a cog.iition of nircihaJpaka

jgnition. To cognise a cognition is to know it explicitly

} a cognition of this or that object To become conscious

[ a mental state is to refer it explicitl} to the self, on the
oe hand, and some definite object, on the other. Thus
ly awareness of a perception (anuoywasaya) appears in the
irm ‘
I know this jar,’ or ‘
I know tins table.’ Nirvikal-
il>a perception is a knowledge of the uncharacterised
bject It IS an un lifferentiatcd feeling of the iiulefinite.

.8 such. It cannot be known a-> a fierce ition of tlii' or that

bject. Hence the Naiyayikis differ from the Auvaitins
1 holding that mrat/i:a/paA:a pet cepii >11 cinnot be perceived
r directly known. ^
Alllimghwe cannot perceive it, we
in logically pr ive it. The exist nee of niroikalpaka
erception is proved by inference. In saoikalpaka perci'p-

on an object known as related to certain qu ill ties

is But
re cannot know the relation between a thing and us
ualities unless we previously know these in themselves.

Hrvikdlpaka is this prior knowledge of the thing and its

ualities as unrelated entities. It is the ground of our

ivikalpaka or relational knowledge about the thing. Hence
re must admit nirvikalpaka as the fir^t stage of all giades
f perception, since a simple cognition of existents as such
» the precondition of all complex cognitions of their different

elations. Thus the reality of nirvikalpaka is inferred from

avikalpaka perception.*
With regard to savikalpaka perception, the Nya^a-
/^aitesikas agree with others in holding that it is the cog-

^ Nirvikalpakaiii na pratjak^am . Ta]$j?(7§Davag&bi;fiaoa3ya pratyak^a h na

savati, gha^amabam pratyayat, etc ,
SM , 58.
^ Vii$i§tavai4t9(yii]flft(iaiii prati bi uavaccbedaka-prakardka'ii jQ&nad)
irapam, etc., TC., I, p, 819,
 THREE MODES OP ORDINARY PERCEPTION '219
a

nition of an object as qualified by certain attributes (vi6i§ta-

jMna). In it the object of perception is known as charac-
terised by some qualities. So it is a judgment in which
certain attributes are related to the object by way of predica-
tion, e.g. ‘ this is a cow.’ ^
Hera the attribute of cowness
is predicated of the presented object. Hence saoikalpaka
perception is always expressed in a proposiMon, of which
the subject is the thing perceived and the predicate is the
attribute present in the thing From this it follows that
the contents of sainkalpaka perception are the same as those
of the mrvikalpaka. It is the same object that is cog-
nised in both. But while in the former the object and its

attributes stand in the subject-predicate relation, in the

latter they are n<,t so related. Hence the tv\o differ, not

in their object or content, bu' only in point of predicition.

Satihaipala is a judgment <•{ the object in words or jirop isi-

tions lut nirvika'pika i^ a jiidgtnein of it not in words or

propositions The one i> a predicative, while the other is a
®
non-ptedic.itive judgment of the same object or fact.

While ninikaJpaka is the fiist, saoikaJpaka is the second

stage of an ordinary peiception The first stage develops into

the second in the case of all normal individuals wh) know

the use of any language. The process of dcvel ipmeiit from

the one to the other is explained by association and memory.

In the case of the perceptional jiidgme it :
‘
this is a cow.’

the first step is the contact of sense with the object, which
immediately leads to a nirvikalpaka perception or simple
apprehension of the cow as something indefinite. The
indefinite object of nirvikalpaka perception being associated

with a certain class-name in our past experience revives the

^ ‘^avikalpakarboa fianam yatl.ft gaurayamiti ibtd., p 839 8avikalp»kam

nftma'&ty&(]iyo ao& makaiii* TB p C

* Taflii>at>aeva\astvMma savikalpakaaya gocaral? aa eva nirvikalpakaaya iandol*

lakhavi varjual^ • Iha iiliJ&nusandhSniiDatratDabhyadhikani paraoif \i 9 ayfi na tu

bhedo'tti aavikalp&vikalpayoll^, NM., p. 99
220 NYiYA THEORY OF KNOWLEDGE

word-image answering to that name. With this we remem-

ber the class-name of the perceived object and call it by that
name. It is here that we have a savikalfala perception of
^
the object expressed in the proposition ‘
this is a cow.’
The Buddhists deny tbc validity of the sabikalpaha mode of
perception. They contend that what is given in perception
is a bare particular which belongs to no class and bears no
name, i.e. has no relations. Its class, name, etc., are only
thought-relations which do not exist in the object, but are
introduced into it b} the thinking subject to meet the needs
of our practical life.^ From the standpoint of the Njaya
realism, however, there is no error in the savihalpalia

perception of an object as qualified by certain attributes

and called by a name. According to it, what is given

in nirvxkalpaka perception is neither a characterkss nor
an uncharacterisable object, although it be not so far

characterised in any way. In reality the object is a con-

crete individual in which certain particulars or specific attri-

butes are united with a certain class-essence or universal.

W’hile in mrvikalpaka the object is apprehended as an
undifferentiated w'hole of the universal and the particulars,
in savikalpaka these are analysed, unfolded and recombined
into the substantive-adjective relation. Hence it cannot be
said that savikalpaka is concerned only with thought-rela-
tions which have no objective basis. Bather, it unfolds all

that IS implicitly involved in the mrvikalpaka stage and

expresses it in the form of a proposition. It does not add
anything that is not contained in the object itself. It repre-

sents no change or development in the object of perception.

On the other hand, it marks a change in the perceptive
consciousness of the object, a development of it from a dumb
feeling of ‘
something there ’
to an articulate expression of

j Vide NVT., p. 128, NK., p, 192.

* Vtdt NVT., pp. 188 f„ and TR pp. ,
60-61.
 THREE MODES OF ORDINARY PERCEPTION 221

the feeling as a cognition of this or that definite thing.

Hence sacikaJpaka is as valid as, but more expressive than,

mrmkalpaka perception.^
Yet another theory of perception, which we have to
consider no>v, presents wlut miy be called the most concrete
view ol perception. It takes up the extreme position that
all perceptions are sacikalpaka or determinate and that there
IS no such thing as a perlectly indeterminate {nirvikalpaka)
perception. This view of perception is shared by the Carva-
kas, the Jamas, the ancient Sabdikas and the Visistadvaita
Vedanta ot Ramanuja According to Ramanuja, to know a
thing is to know it as possessed of certain attributes. A
thing’s existence cannot be separated irom its nature and
attributes. To know the ‘
that ’
or existence ot a thing is

also to know the ‘

what ’
or the nature of its existence. All
knowledge is, therefore, a definite cognition ot some object
as related to a certain class and qualifie.. by certain attributes.
It IS always a deterniinate (saoikalpaka) c 'gnition of the
object as this or that kind of thing. There cannot be any
knowledge of the perfectly indeteiminate That which is no
thing in particular cannot be the object of our knowledge.
Hence there is no such thing as indeterminate {nirvikalpaka)
knowledge in the sense of a cognition of what is not determined
or characterised in any way {mroisesa). There being thus no
absolutely indeterminate knowledge, the distincticn of nirvi-
kalpaka and savikalpaka perception is a relative distinction.

It IS a distinction between two perceptions, both of which

cognise an object as somehow qualified and determined
{savi^esavisaya). But while in mivikalpaka the object of
perception is partially determined, in savikalpaka it is deter-
mined more fully and clearly. Thus the first perception of
a cow IS nirvikalpaka in the sense of being a cognition of it

as an animal of a certain make-up, or of some kind, but not

1 Vtd0 NVT., PP 187-U, NM., pp. 64-89.

 222 NYiYA THEORY OF KNOWLEDGE

of this or that particular kind. On the other hand, the per-

ception of the same object, in the case of an adult who
knows cows as a class, will be savikalpaka in so far as the
object is here further determined and definitely known as
belonging to the class of cows. Savikalpaka and nirvikalpaka
perceptions are thus cognitions, not of the characterised and
uncharacterised, but of the more or less definitely charac-
terised object.*

The Carvakas, the Sabdikas and the Jamas go further

than Ramanuja and hold that nirvikalpaka perception is not
real in any sense. According to the Jamas, all true know-
ledge must be a definite and an assured cognition of objects
(vyavasayatniakam pianam).'^ What distinguishes true
knowledge from doubt, error and the rest is the fact that it

is a firm belief which is also true. It is a definite judgment

of ail oliject as this and not as that.® In it there is a definite
affirmation or denial that an object is or is not such-and-
such. In the so-called nirvikalpaka perception, however,
there is no such definite assertion of an> thing about any
object Hence it cannot be recognised as a form of valid

knowledge. Further, all knowledge being implicit m and

manifested by the self, perception is only conditioned and
not produced by the function of the senses Every percep-
tion, just when it occurs, will be a complete manifestation of

the object. In perception there need not be a transition

from an initial stage of vague and unorganised sen^^e-impres-
sions to that of distinct and determinate knowledge. All

true perceptions are, therefore, determinate {savikalpaka)

cognitions of objects as they really are in themselves.^
The same conclusion has been reached by the Sabdikas
or grammarian philosophers on the ground of the intimate

^ Pratyak9a9ja mrvikalpakasavikalpakabhedabbinntisja Da nirvi^e^avastUDi

pramft^abb&vat^f etc., ^ribhd^ya, 1.1

S Vide Prameyakamalamdrtanda, 1.

^ Tannikajfttxnakaxh Bam&ropaviraddbatvftdanDm&oavat, ibid, 8.

i lbid.,p.8.
 THREE M0BB8 OF ORBINAEY PBRCBFTION 22S

relation between thought and language. According to them,

all objects are inseparably connected with the words or
terms denoting them. All our thoughts and cognitions of
things are expressed in words and propositions. We cannot
think of things except through their corresponding denota-
tive terms.* Bhartrhari, a grammarian philosopher, lays

it down as a general rule that there can be no thought

without language and that all knowledge must be verbalised
experience ® It follows, therefore, that all our perceptions
must be cognitions of objects as denoted by certain names
or words. They must be expressed in propositions, in
which the perceived thing and its qualities are related as

subject and predicate. Ht nee there can be no nirvikalpaka

perception in ihe sense of a cognition which is independent
of verbal expression and free from association with words
or general terms.® All perception is thus savikalpaka or
determinate knowledge of objects as qualified by the
atiriimtes piedicated of them. The Carvakas recognise only
saviKaJpaka or determinate perception, in which we cognise
objects as possessed of a number of perceptible qualities.

For them, ninikaJpaka perception is a hypothesis w'hich

cannot be verified by actual experience. It is something
which cannot be perceived and is therefore unreal.
With this we pass from the extreme view of perception
as blind sensation to what appears to us to be another

extreme view ol it as a fully developed judgment expressed

in a predicative proposition. This is met by an inter-

mediate position that distinguishes between two modes of

perception, namely, the nirvikalpaha and the savikalpaka,

of which the former is a simple apprehension or judgment

of an object without words, and the latter a predicative

1 Parve*rth&b sarvathS sarvadft aarvatra nSinddheyanvitSl^, etc., NVT , p 126.

* Na ao*8ti praty lyo loke dibd&ougainadfte, anaviddhaiDiva ifianam sarvaiii

ilabdena gamyate. (Bhaitfbari, K&nka quoted in Siddh&ntacanink&t pp. 89-40.)

9 NVT., pp. 125.26.

 224 NYAYA THEORY OF KNOWLEDGE

judgment of it in a word-proposition. This intermediate

position seems to be a more reasonable view of perception.
The pure sensation, to which the Buddhists reduce percep-
tion, is a psychological myth. Then the linguistic I'lew of
perception as alw.iys a fully developed propositional judg-
ment IS contradicted bv such perceptions as ante-date
language, e.g. the perceptions of children and higher
animals. Even in the case of adults who know a language,
an indefinite cognition of an object as given in sensation
may very well precede a definite perception, in which it is

recognised as a particular kind of thing. According to

most of the Indian s} stems, the former is the mrvihalpa'ka

and the latter the saukdpaka mode of perceptual knowledge.
It IS also generally held by them that there is no room for

error in mrukalpaka perception, since it is a bare apprehen-

sion of the given object without any judgment of it ns this

or that. Hence it is alw.iys true. It is the savikalpaka

perception of an object as a particulai kind of thing that is

liable to error, since our judgment here may not conlorm to

the real nature of the object.^

2. Recognition (pratyahhijM) as a mode of perception

Kecognition may be understood in two senses In a

wide sense, recognition means understanding the nature or
character of a thing. In this sense, to recognise a thing
is to know it as such-and-such, as when I know that the
animal before me is a cow. It is generally admitted that
reco.,nition in this sense is an ordinary mode of perception,
which 18 called saitkaJpaka and which relates a thing and
its qualities by way of predication. In a narrow sense,

^ Jayanta Bhatta however, maintains that hability fo error is common to both

nirtJifca7pfl/fa and I erceptjoDB TheillusKDs cf sense, like the perception

of two moons, are instanced by him as cases of nimkaJpaka perception which ar^

•rroneons. (Vide NyHyafnaHjartf p. 97.)

 THREE MODES OF ORDINARY PERCEPTION 225

however, means knowing a thing as that which

recognition
was known before. To recognise thus means to cognise
once again that which we are aware of having cognised
before. Pratyabhijm is recognition in this sense. It

consists in knowing not only that a thing is such-and-such

but that it IS the same thing that we saw before. According
to the Naiyayikris, pratyahhijM is the conscious reference of
past and a present cognition to the same object. I see a
a jar, recognise it as something tliat was perceived before,
^
and say this is the same jar that I saw.’
‘

With regard to the nature of pratyahhijM or recognition

the question is : Is it a simple or a complex cognition? Is

it a case of pure perception or memory^ According to the

Buddhists, recognition is a mechanical compound of percep-
tion and memory. It cannot be called perception, because
it rebites to a past object with which there cannot be any
sense-contact It cannot be called pure memory which
refeis only to the past, while recognition refers to a present
object as the ‘ this.’ Nor can we say that recognition
IS a synthesis or unitary product of perception and memory.
Perception and memory arise respectively out of sensation

and imagination, and are incapable of fusing into a single

effect Admitting that recognition is a unitary product,

what is the nature of its object‘d If the object be past,

then recognition is not different from memory. It it be in

the future, recognition becomes a form of imagination

(samMpa). The object cannot be a merely present fact,

since in recognition the object is identified with something

of our past experience. To say that the object of recogni-
tion exists m the past, present and future is a contradiction

in terms. Hence the Buddhists conclude that pratyahhijM

IS a dua cognition including both perception and memory

lidl^Qrvaparsyorvijllan&yofekaTi^&yo pratisaDdhiifiaDaiii pratyabbi]fiftuaixi, £to., NE.,

3 1. .
8 2. 2
. .

29— (1117B)
226 NYiYA THEORY OF KNOWLEDGE

which refer respectively to the two aspects of an object

as ‘ tv>^" ’
and ‘
that,’ or as present and past/
TheJainas take pratyahhijm to mean recognition in

the sense of both understanding the nature of an object

and knowing that it w^as perceived before. To recognise
a thing is to know that it has this or that property, or

that it is the same as what was seen before. It is not true

to say that pratyahhipid is a dual cognition consisting of

perception and memory. Although conditioned by percep-

tion and memory, it is a new kind of knowledge winch
cannot be resolved into them The testimony of introspec-
tion clearly tells us that pratyahhijm is a unitary cognition

and a distinct type of knowledge What the lSai}ayikas call

upamdna or comparison is, according to the Jamas, a lorm
of pratyahhijm as understood by them.^
According to the Naiyayikas, piatyahhijhd consists in

knowing that a thing now perceived is the same as what

was perceived before.® That pratyahhijm or recognition, in

the second sense, is a single psychosis appears clearly from

the fact that it refers to one and the same object The
cognitions of a jar and a cloth are two different psychoses,
because they are evidently related to two different objectives.
Recognition refers to only one thing and is therefore a
simple and unitary cognition. The unique cause {karana) of
the phenomenon of recognition is constituted by the senses
and the effects of past experience. Recognition is brought
about by sense-impressions as modified by the effects of

previous experience of an object.'' It gives us the knowledge

of an object as existing in the present and as qualified by
its relation to the past." A thing’s relation to past time

* NM , pp 448-49
* Vide Prameyakamalamdrtanda, pp 97-100
3 So’yam Dtfvadatta ityatltavartamanakalavi^iBtavi^ayakara ]fianam pretyabhijfiS,

Mttabhaitnifp 25
* Samsk&raBahitamiDdriyamasy&tjL prattteb kara^ara etc ,
NM » p 459
5 AtItakaIavi4ifl|o vartamftnakalavacchinna^cdrtha etasyamavabbftsate, tbid.
 TfiRBE MODES OP ORDINARY PERCEPTION 22?

or a past experience is a character which qualifies its present

existence. To know this is just to know that we have per-

ceived it before, i.e. to recognise it. Hence recognition is
a special mode of samkalpaka perception. In all but the
nirvtkalpaka mode of perception there is the influence of
past experience on our present knowledge. All savtkalpaka
or determinate perceptions of objects consist of certatm
given or presentative elements and certain representative
factors like ideas and imiges of similar objects experienced
in the past. In an ordiniry savikalpaka perception the re-

presentative factpr^i do not remain distinct but are assimilated

with the presentative elements to make up one percept In
pratyahhijM or recognition, the representative factor has the
form of a definite recollection of some past experience of an
object and modifies the present perception of it Still it is

perception, since it is brought about by sense-object contact,^

Hence the Naiyayika concludes that recognition is a kind
of qualified perception, in which the present object is quali-
fied by a distinct recollection of its past experience. The
Mimamsakas and the Advaita Vedantins also hold that re-

cognition IS a kind of perception. The Mimamsaka«, how-

ever, do not distinguish it from an ordinary savikalpaka
perception. According to them, recognition is that kind of
perception in which the object is determined by the name
by which it is called, e.g. ‘ this is Devadatta.’ *
For the
Advaitin, pratyahhijm is a perception of the mrvikalpaka
kind, since there is in it no predication of anything about
the perceived object, but an assertion of its identity amidst
changing conditions.®

^ PorvabhijlIaaavi^istagrahyaiii&Qamisjatfiiii pratyabhijfiaoam, NM,, p 461,

2 8D., p. 42
s yp., Ch. 1.
 CHAPTER X

EXTRAORDINARY PERCEPTION (ALAUKIKA

PRATYAKSA)

1 Samanyahksana or the perception of classes

In the ancient school of the Njaya vve do not meet with

the distinction between laidika or ordinary and alaukika
or extraordinary perception This distinction appears in
the modern Nyaya beginning with ftangesa In laukika or
ordinary perception there is a normal sense-contact with
objects present to sense In alaukika perception, however,
the objects are not actually present to sense, but are con-
veyed to it through an extraordinary medium In it there
is a special kind of sense-object contact {alaukila-sanmkarsa).
Extraordinary perception is of three kinds, namely, samanya-
laksana, imnalaksana and yoga]a.

Sammy alak Sana is the perception of a whole class of

objects through the generic property {samdnya) perceived in
any individual member of that class Thus when we per-

ceive something as a pot we judge it as belonging to the

class of pots. But to know that the thing belongs to the
class of pots IS also to know all other pots belonging to the
same class. To say that ‘
this is a pot ’
is to know, by
implication, what all other pots are. Hence in perceiving
one thing as a pot we perceive all other pots. But the other
pots are not present to sense in the same way in which one
is present. How then can there be any perception of the
other pots? If there is to be any perception of the other
pots^ they must be in some sort of contact {sannikar§a) with
 liXTRAORDINARY PERCEPTION 229

our sense. According to the Naiyayikas, when we per-

ceive one pot we perceive the universal ‘ potness ’
as its
defining property. It is this perception of the universal
‘ potness ’
in the present pot that serves the purpose of
contact (asatti) between sense and all other pots. The'
knowledge of the universal (samanyajmna) is here the
medium of sense-object contact, by which we have a percep-
tion of all pots when one is perceived,^ Ti may be objected^
here that make each of
sdmanyalaksana perception would
us omniscient If all the objects of a class are known with

the perception -of any of them, we should know all knowable

objects when we perceive anything of the world as belonging
to the class of knowables. The Naiyayikas reply that in

sdmdnijalaksana we perceive only one member of a class as

an individual with its specific and generic properties, while
the other members are known as possessing the generic
property or the universal alone. Hence we cannot expect to

have that full and detailed knowleilge of all things, which is

implied by omniscience
That sdmdnyalaksam is a type of real perception is

supported by the Naiyayikas on the followung grounds.

Without it we cannot explain the knowledge of universal
propositions (vydpti), which is presupposed in inference.
How know that all smokes are related to fire ?
do we We
cannot know this if our perception be limited to parti-
cular smokes, for any number of particulars will not make
up the universal. Hence we must admit that while perceiv-
ing one smoke as related to fire, we perceive all smokes,
through the universal ‘
smokeness,’ as so related. It cannot
be said that it is unnecessary to assume that we perceive all

smokes as related to fire, because we cannot even doubt if

all smokes are related to fire or not unless all smokes are

somehow presented to us when we do perceive one as related

1 isattirS^rayaijara to a5many»jftanami?yate, etc., BP. and SM ,

64*65.
230 NYiYA THEORY OF KNOWLEDGE

to fire.^ Again without sdnidnyalak^ana we cannot explain

negative judgments of perception like ‘
this cloth is not a
pot,’ ‘
this IS not a cow,’ etc. To say that ‘
this is

not a cow ’
is to know the class of cows, i.e all cows.
This can be known only if when perceiving one cow we per-

ceive all other cows through the universal ‘

cow ness ’
as
perceived in the present and perceived cow. Further, we
cannot explain the voluntary effort to attain a pleasure with-
out the help of sdrndnyalaksana. To strive consciously for a
pleasure is somehow to know' it as something to be, but not

yet, experienced. But how can we know a pleasure which w'e

have not jet had? This is possible if, when experiencing

one pleasure, we know' all pleasures through the universal
‘ pleasurableness ’
as belonging to the perceived pleasure.^

The Nyaja view of sdrndnyalaksana has been severeh

criticised and hnally rejected by the Vedanta. According to
it, sdnidnya or the universal is a group of essential and
common attributes belonging to a number of individuals.
While the universal, as such, may be perceived along with
the perception of an individual, it does not give us a percep-
tion of all the individuals possessing the same universal.

Nor IS there any valid ground to believe that in perceiving one

individual we must perceive all other individuals of the same
class. Thus in inference, say of fire from smoke, it is suffi-
cient if W'e know ‘
smokeness ’
as related to ‘
fireness.’ It is

not at all necessary for us to know that all smokes are related

to fire. If all were known, then there would be no need for

any inference in a particular case. Similarly, the negative

judgment ‘
this is not a cow ’
is quite possible if we only
know' what \cowness ’
is and not what all cows are. So,
too, there may be a conscious pursuit of some future pleasure
if the pursuer knows it to be similar to his previously ex-

* TC.,lI,pp. 290f,, 8M.,65.

^ Ibtd , pp. 283-90.
 EXTRAOEDINARY PERCEPTION 231

perienced pleasures. For this, there need not be a sdmanya-

laksana perception of all pleasures when one is actually
perceived.^
Now we are to observe that the reality of samdnyalaksam
as a type of perception depends on the presentative know-
ledge, if any, of a class of things If there is any such
knowledge with regard to a whole class of things, we have to

admit samdnyalaksam as a type of genuine perception. It

IS pointed out by the Naiyayikas that any knowledge of the

genus or the class does not justify us in believing in
sdmdnyalaksana. To know the general character of an
atom IS to have only a conceptual knowledge (manasabodha)
of all atoms. Similarly, we may think of the class of pots
m a-daik room But that does not mean that we have a
presentation of all pots through any sense for which we
require the help of a sdmdnyalaksana perception. It is only
when in perceiving some individual object there is the
presentation of the class of objects to which ’t belongs that
we have to admit sd many alah Sana to exjilain the presentative

knowledge of that class. Hence the crucial question is this :

Is there any presentative or direct knowledge of a class of

things^ The Naiyaj-ikas contend that there is such a know-

ledge and seek to prove it on such grounds as we have
explained above. Before we come to these we may state the

following facts as more or less non-controversial.

When I know an individual as belonging to a certain
class 1 know the univeisal or class-essence undei lying it. To
know the universal, however, is to know a character or

group of characteis which belongs to all the members of the

class. So in knowing the universal I know all the indivi-

duals of that class — past, present and future — as participat-

ing in that universal. To know a horse as horse is to

know horseness, and to know horseness is to know that it

I Ytde Advattastddhif pp. 187 39.

232 NYiYA THEORY OF KNOWLEDGE

belongs to all horses, or to know all horses as possessing it.

Again, a knowledge of the class seems to be implied m any

generalisation. When from such particular cases as ‘
A is

mortal,’ '
B is mortal,’ ‘
C is mortal, ’
and so on, we con-
clude that ‘
aW men are mortal,’ we somehow know that
mortality is true of the class of men. Now the question is

How do we know anything about the whole of a class of

things from the observation of some of its members ? In
other words, how do we get general propositions from the
observation of particular facts?
In Western logic this is explained by inductive infer-
ence.’ The Naiyajikas, however, hold that the knowledge
of the class or the general proposition is given by perception
of an extraordinary kind. The Vedantms, on the other
hand, argue that a geneialisation is the statement of an
invariable relation between universals and that it is known
through the observation of their concomitance in one or
more instances. Thus the geneial proposition ‘
all men are
mortal,’ or ‘
whatever is smoky is fiery
’
is the expression of
an invariable relation between manhoid and mortality, or
smokeness and fireness We have a knowledge of such
general propositions when in any particular instance we find

manhood to be related to mortality, or smokeness to fireness.

The Vedanta view of generalisation thus corresponds to what

“ intuitive induction ” in Western logic and
is known as

is explained as a “ process by means of which we apprehend

a particular instance as exemplifying an abstract generaliza-

tion,” as when ‘
from the apprehension of this red patch as
being darker than that pink patch we may know immediately
that every such red patch {i.e. redness) is darker than every
such pink patch (t.e. pinkiie-s).’ It seems to me that

while the abstiact principles of mathematics may be taken as

^ Vfde Mill, System of Loqic ; Stebbing, Logic in Practice pp. 19-20

^ Vide Stebbing, Modern Introduction to Logic, p. 243,
 EXTRAORDINARY PERCEPTION 233

statements ot necessary relations between certain universal

concepts, all our empirical generalisations, including those
here given, are truths about classes of things. In fact, such
principles arc what may be better called the necessary laws
of thought lather than truths about any universal that has a
denotative reference to a class of things. But, if we take
the above cases as abstract generalisations exemplified in
particular instances, we do not understand how they can
be called, as some Western logicians have called them,
“
intuitive induction ” It by such induction we are to mean
the immediate apprehension of an axiom by means of its

exemplification ‘in a particular instance,” we must admit

that there is no room for any induction or inference or
reasoning in it. It we have an immediate knowledge of
anything, there is no need tor any inference or reasoning
with regard to it. A description of such knowledge as
“ intuitive induction ” seems to be no less objection- me
to

able than a desciiption of it as “ perceptual inference.”

If, however, by ” intuitive induction ” we mean the intuitive

knowledge of a general principle through a particular

instance, we should make it plain that such instance is not
the basis of an inference with regard to the general principle.
Rather, the general principle is apprehended along with
the particular instance, although an explicit statement of it

may be separated from the latter by an interval of time.

To guard against a possible misunderstanding, therefore,
it is better to characterise our knowledge of the general

principles of logic and mathematics as intuition than to

an “
^
call it intuitive induction.”
Our ordinary generalisations, however, are different

1 This J8 really adm’ttf^d by Dr Stebbing although there seems to be some

wavering at certain places, Cf “ Thus we may be said to see the general principle in
** **
apprehending the particular case The mtwttion is of the form but it relates to the

material exemplified in the form fitalics (Modern Introduction to Logic,

pp 96 97,244)

30— (1117B)
 234 NYIYA THEORY OF KNOWLEDGE

from the abstract principles of logic and mathematics.

While the latter are truths about certain universal concepts,
the former are truths about classes of things. When we
lay down the proposition ‘
all men are mortal,’ or '
all smoky
objects are fiery,’ what we really want to convey is, not
that there is a necessary relation between manhood and
mortality, or between smokeness and fireness, but that
mortality is true of the class of men, or that all smokes are
connected with fire. Such propositions are empirical
generalisations in the sense that these are assertions about
whole classes of things, which are true, and that these are
arrived at from observation of particular instances. But
what IS the nature of the process of knowledge that is

involved when we generalise from some to all ‘ ’ ‘ ’

in a logical-

ly valid way ? Is it any kind of induction ? If so, it must

be either perfect induction
‘
or “ Induction by Simple’

Enumeration.” A perfect induction‘ is one in which ’

‘from the consideration of each of the members of a limited

class we pass to a generalisation concerning all the members

of that class.’ This is eKeraplified when on examining

every boy of a class one says “all the boys of this class are
intelligent.” With regard to this Stebbmg *
say-s: ‘Mr.
Johnson has suggested the convenient name summary
induction for thismode of inference, and that it is certainly
a more appropriate name than “ perfect induction.” It ’

seems to me that both names are equally inappropriate, and

that for the same reason. A summary of a number of
observed facts is not an induction at all. To call it an
induction, be it perfect or summary, is to misjudge its
epistemic character as inferential If on examining every
patient in a sick-room a physician says “ all the patients
in this room have got fever,” then his judgment is not to
be described as an induction or inference in any sense. It

1 Modem Introduction to Logic, p. 244,

 jiXtraordinary perception 235

IS but a memory-synthesis like the one we have when we

understand the meaning of a sentence by a synthesis of its

constituent words and their meanings as that is effected by

memory. Hence our knowledge of a general proposition
like ‘
all men are mortal ’
cannot be called a perfect induc-
tion,” even if it v/ere possible for us to examine all men.
But that is not possible as a matter of fact, since in man
we have, not a limited, but an unlimited class that has
“ an infinite number of members.”
Let us next consider whether an empirical generalisation
like ‘
all men are mortal ’
can be explained by “ Induction
by Simple Enumeration.” ” Generalisation from a number
of examined instances which are not assumed to constitute
all the instances of the given class is now usually known
by the name Induction by Simple Enumeration.” ^
Such
induction may, therefore, be put in this form :
‘
A, B, C
are mortal, therefore all men are mortal.’ But this is not
a form of valid inference It obviously violates the general
rule of inference that we must not go beyond the evidence,
since in this inference the conclusion makes a state-
ment about all men on the ground of what is observed in
some men. Further, if it were a form of valid inference, it
would validate any argument that might be put in this
form, just as the forms of deduction guarantee the validity
of a reasoning that may be put in any of them. While,
however, we accept the argument all men are mortal, be- ‘

cause A, B, C are mortal,’ we do not acquiesce in the

proposition all men are white, because A, B, C are white.’
‘

But why ? If we examine the two arguments we shall

them is this. In the first,
see that the distinction between
mortality which is found in some men is predicated of all
men. In the second, whiteness which is observed in some
men is predicated of all men. But then, we find A, B, C

> Op Ca ., p
. 246.
236 NYSYA THEORY OF KNOWLEDGE

to be mortal, not because they are A, B, C, but because

they are men. On the contrary, we know that A, B, 0 are
white because they are A, B, C, and not simply because
they are men as such. This means that while mortality
is related to tlie essential nature o^ A, B, C, whiteness is

not so related to them That individual men like A, B, C

possess a certain essential common nature which is to be
found m men is borne out by the fact that we put
all

together all men into the class man and exclude all other
‘ ’

animals from that class If, then, we find that mortality

18 related to the essential nature of some men, we know

that all men must be mortal. That is, we know all men
to be mortal when we know that mortality belongs to the

essential nature of some men like A, B, C. But the first

knowledge does not follow fiom the second To know
mortality to be related to the essential nature of some men
IS just to know that it is related to all men or the class of

men What is related to the essential nature of some

‘

men must be related to all men is a truth which is known

’

directly or immediately, and for which we require no infer-

ence or reasoning. Hence our knowledge about the whole

class IS here an intuitive knowledge due to the knowledge
of the class-essence or the univers.il It cannot be said
that the knowledge of the class-essence oi the universal is

got by inductive inference. The latter presupposes the

former and so cannot be the ground of it. It seems to me
that the universal, underlying a class of things is either
directly known or never known. Observation of and experi-
ment on things help us to find or discover the universal
that is in them, but not to make or construct it out of
them. And when by observation and experiment we find

that the universal or class-essence is related to something,

we know at once that all the members of the class are
related to that thing. According to the Naiyayikas, it is

the perception of the whole of a class as related to an

 EXTRAORDINARY PERCEPTION 23 '?

attribute, which is simultaneous with the perception of the

class-essence as so related. To distinguish it from ordinary
sense perception they call it alauktka or extraordinary per-
ception. If this be, as it very likely is, so, what is known
as inductive inference may be reduced to the Naiyayika’s
sammyalaksana perception in the sense ot intuition of a
general proposition through the knowledge of the class-,

essence or the universal.^

The Naiyayikas further point out that without the per-

ception of the whole of a class of things the problem of
generalisation cannot even arise. If, when perceiving a
particular smoke as related to fire, we ask are ‘
all

smokes related to fire? ’

it is because the class of smokes

1 Some Western logicians hilly realise the futility of all iniaction as a form of
valid inference. They do not, howevei, so clearly realise the imphcation of this
futility. If there is no form of inductive inference, there can be no such thing as
inductive inference in logic If this be admitted, as it should be, then the problem of
generalisation takes a different form A generalisation is no longer a matter of
inference, but is to be explained by way of intuition That our knowledge of general,
principles is intuitive will be admitted by many But very few, if any, will admit
that an empirical generahsxtion also is a matter of intuition based on the knowledge
of class-essences or universals Some Western logicians, however, seem to tend
towards this view when they try to establish a general proposition on the ground of the
knowledge of “important resemblances” or “common properties” or “class characters”
of things Consider, for example, the following statements from Dr Stebbing *
Such
classes as swans and men differ from suJi classes as scarlet things and sour things in
the fact that every member of the class swan, for instance, has several properties m
common with all the other members, whereas the members of the class scarlet things

have few properties in common which are l ot also possessed by things that are not
”
Mill, “ natural kinds
*
scarlet Such classes as swans are callei by Simple enumera-
tion IS not, then, to ba regar led a process simply of counting ; it is a counting of
instances recognised as having certain properties in common The inference is depen-
dent upon recognition ot resemblances * (Modern Introduction
pp 248-49 ) to Logic,

‘
It 18 the fact that certain properiies are found together that makes class names so
useful If we know that there is a set of properties such that no member of the set
18 ever found without other members of the set, then we have a basis for inference

It is because this appears to be the case with natural kinds that generalisation about

natural kinds, such as crows, acids, men, seems to be plausible.’ {Op. cit , p. 261
“ As Mr Keynes points out *
Scientihc method, indeed, is mainly devoted to dis-

covering means of so heightening the known analogy that we may dispense as far as

possible with the methods of pure induction

*
” (Op cit., p. 266 )
238 NYiYA THEORY OF KNOWLEDGE

is somehow presented to us. It cannot be said that

we merely think of all smokes as a concept or
general idea, and that there is no presentation or direct
experience of them. We can think of such particular
smokes as were previously experienced by us. But the idea
of particular smokes always falls short of the class of smokes
or all smokes. And there cannot be any idea of the class
of smokes without a corresponding direct experience of it.

It is only when, in perceiving one smoke as related to fire,

all smokes are presented to us through the perception of

the class-essence ‘smokeness,’ that we can legitimately

ask the question Are all smokes or the class of smokes
related to fire? Hence we conclude that sdmanyalaksana
is a type of genuine perception It should, however, be
borne in mind that we have not such a perception whenever
we perceive any individual possessing a class-essence. It

18 only when the perception of the class-essence of an

individual has a direct reference to the class of things to

which it belongs that we have to admit a sdmdnyalaksana

perception of that class of things through the perception of
the class-essence. In what cases other than those mention-
ed above there is such a reference is a matter of phenomeno-
logical observation which we need not discuss here

2 . Jfidnalaksana or acquired perception

The second type of extraordinary perception is called

jMnalak§ana. It is the perception of an object which is in
contact with sense through a previous knowledge of itself.^

When on seeing something one says ‘

I see a piece of
fragrant sandalwood,’ he has an immediate knowledge or
perception of its fragrance. This cannot be explained
without the help of jmnalaksana. How can he perceive

^ Vjfayi yasya tasyaiva vyapftro jfiSnalak^a^ab, BP., 66.

 EXTRAORDINARY PERCEPTION 239

the fragrant sandalwood, seen at a distance? Its fragrance

is not then smelt by him and so does not come in contact
with the sense of smell. Nor can there be any ordinary
contact between smell and the sense of sight. Still he
perceives it as fragrant sandalwood lying at a distance.
Hence we are to say that there is some extraordinary contact
between fragrance and the sense of Here our past
sight.
experience of fragrance in the sandalwood does the work of
contact between sense and object. Our past knowledge ol

fragrance {saurahhajmna) brings about the present percep

tion of it, although it is not actually smelt by us. It cannoi

be said that the present perception of fragrance as a parti

cular IS brought about by sdmanyalaksana cognition of Ihi

class of fragrants. The latter cognition supposes ar

ordinary perception of the genus of fragiance through sense

contact which is not to be found in the present case.* Thus
we see that in both sdmdnyalaksana and jndmlaksana per-
ceptions sense-object contact is mediated by some kind of
knowledge. In the former, the knowledge of a universal
and, in the latter, some past experience is the medium of
contact between sense and the perceived objects. But the
distinction between them is this While in sdmdnyalaksana
the knowledge of the universal leads to the perception of the
individuals in which it inheres (dSraya), in jndnalaksana,
a past knowledge leads to the present perception of its own
object {yadvtsayakam jmnarh tasyaiva pratydsathh)

The Naiyayikas explain illusions by the help of the

theory of jmnalaksana perception. The illusory silver is

perceived because it is presented through our previous

knowledge of silver as seen at some other time and place.

But the Vedanta objects to the Nyaya theory of pddnalaksana.

It argues that to recognise jndnalaksana as a type of genuine

1 SM., 66.

* Ibid.
240 NYiYA THEORY OF KNOWLEDGE

perception is to reduce ordinary inference to perception.

If the fragrance of sandalwood be perceived because we
have a previous knowledge of it as connected with sandal-
wood, then in the inference of fire from smoke, the fire

may be said to be perceived since we have also a

previous knowledge of it as related to smoke. There

is nothing to distinguish between the two cases. In
the one w'e have a perception of sandalwood and a
previous knowledge of its relation to fragrance. In the
other, we have a perception of smoke and a similar know-
ledge of its ndation to fire Hence if the fragrance be in
extraordinary contact with sense and so perceived, there
may be such an extraordinary perception of fire and, for
the matter of that, of all objects of inference So the
Vedanta holds that our knowledge of the fragrance of
sandalwood, seen at a distance, is due to inference and not
any extraordinary perception like the Naiyayika’s pidna-
laksana}
Now let us consider whether the knowledge of the
fragrant sandal that is involved in the judgment “ 1 see

a fragrant sandalwood ” is really a case of perception as

held by the Naiyayikas, or a case of inference as urged by
the Vedantins. Students of Western philosophy will
readily recognise that such knowledge is of the same kind as
what IS called “ complication ” by some psychologists. The
judgment “ I see a fragrant sandalwood ” is in fact equiva-
lent to the judgment “ The sandalwood looks fragiani.”
And this is really anothei instance of what Stout, Ward and
Wundt call “ complication ” and illustrate by such
judgments as “ Ice looks cold,” “The armour looks hard,
smooth and cold.” ^
We may go further and say that the
Naiyayika’s ]Mnalaksana is similar to the visual perception

1 VP , Ch 1

2 Vide Stout, Manual of Psychology, p 102, Wundt, Human and Animal

Psychology, pp 285-86,
 EXTRAORDINARY PERCEPTION 241

of distance, since the eyes have no independent means of

apprehending those relations of surfaces and lines which
presuppose the third dimension.” If there can be a visual
perception of distance, coldness, hardness, etc., there can
also be a visual perception of fragrance. Hence the funda-
mental question to be discussed here is this Can there be,
:

and is there in fact, a visual perception (i) of distance,

(ii) of coldness, hardness, etc., and (Hi) of fragrance.
With regard to the first case (i), J. S. Mill was
strongly of opinion that the perception of distance by the
‘

eye is, in reality, an inference grounded on experience ;

though in familiar cases it takes place so rapidly as to appear

exactly on a par with those perceptions of sight which are
really intuitive, our perceptions of colour.’ Modern psycho-
logy, however, has outgrown this view and finds no difficulty

in admitting that there is a visual perception of distance.

With regard to the other cases, however, there is much
difference of opinion among philosophers, both Indian and
Western. While the Advaitins would bring all such cases
under inference, the Naiyayikas are in favour of treating
them as genuine perceptions. Among modern thinkers some
psychologists like Stout, Ward and Wundt take at least the
second group of cases (m) as a form of perception, although,
to distinguish it from ordinary perception, they give it the
name of “ complication.” Many other psychologists,
however, would reduce them to some kind of rapid or
implicit inference. As for the third case (Hi), the Naiyayikas

are perhaps the only realists who would say that we have a
visual perception of fragrance.

How are we to deal with the second and the third case ?
We have to raise two questions, viz. (i) how are we to

distinguish between perception and inference, and (ii) where

are we to draw the line between perception and inference ?

1 System of Logic, p 4.

31— (1117B)
242 NYiYA THEORY OF KNOWLEDGE

Perception may be defined as an immediate knowledge

of objects or a knowledge of objects which is not brought
about by any other knowledge. On the other hand,
inference is the indirect knowledge of an object through
the mediation of some sign which is known to be always
related to it. In both perception and inference there is

an interpretation of some datum. In both we may dis-

tinguish between a given and a suggested content. In

perception there is a synthesis of some sensuous elements
and some non-sensuous presentations But in perception
these two elements are not kept distinct They fuse into
one whole or form parts of one complex object. There is

no transition of consciousness from the one to the other.

We do not think of the one apart from and independently
of the other. The one immediately suggests the other and
gives rise to the cognition of an object as one whole, of
which they are the parts. Thus in an adult’s visual

perception of a rose, the sensory element, namely, its colour,

immediately presents its other qualities, namely, its touch

and smell, and the sensed and the presented elements blend
into the perception of the rose.
Inference is distinguished from perception by the fact

that it gives us a knowledge of some fact through the

mediation of some other fact on the basis of a uniform
relation between them. Of these two facts the second
suggests the first through a knowledge of their uniform
connection with each other. But the suggestive fact and
the suggested fact remain distinct and we are conscious of
a transition of thought from the one to the other. These
do not fuse into one whole as they do in perception.
Rather, they stand out as two wholes or two distinct facts,

either of which is thought of independently of the other.

Thus in the inference of fire from smoke my mind passes
from one thing to another which is distinct from it and is

thought of independently of it. Hence the distinction

 EXTRAORDINARY PERCEPTION m
between perception and inference is briefly this : Perception
is the integral immediate consciousness of an object.
Inference is a multiple mediated consciousness of an ob'ject

which may be expressed as a this-therefore-that conscious-

ness.
In view of the above distinction between perception and
inference we cannot admit the Advaitin’s contention that to

recognise jHanalaksana as a form of perception is to

obliterate the distinction between perception and inference.

When we pass such judgments as ‘
the rose looks soft,’
‘
the stone looks hard,’ we do not pass from the colour of
the rose or the stone to its tactual quality. Likewise, when
we say ‘
ice looks cold,’ or ‘
the sandal looks fragrant,’
there is no transition of thought from ice to coldness, or
“
from the sandal to its fragrance. We do not say because
ice, therefore cold,” or ‘
because sandal, therefore fragrant.’
On the other hand, coldness or fragrance is a part of the

presentation of the ice or the sandal. What we say is :

“1
see the cold ice or the fragrant sandal,” just as we say
“ we see the distant hill.” In the inference of fire from
smoke, however, there is a tran.^ition of our thought
from smoke to fire as two distinct objects. We never say
” I see the fiery smoke.” In fact, the fire is here only

thought of by us and not presented to us. To recognise

^iidnalaksana, therefore, as a form of perception is not to

ignore the fundamental distinction between perception and

inference.
It may indeed be contended here that while in the visual
perception of a rose or a stone or a block of ice, there is a

presentation of some tactual quality, there is no such pre-

sentation of fragrance in the visual perception of sandal-
wood. To this we are to say that if the eye can present a

tactual quality like softness or hardness or coldness which

it is not fitted, by nature, to perceive,
there is no inherent

impossibility in the eye being made competent to perceive

 244 NYiYA THEORY OF KNOWLEDGE

smell. We are, therefore, to consider under what condition

or coaditions perceptioa takes place. When we have discover-

ed these conditions, we shall see that, although logically per-

ception rtnd inference are two fundamentally distinct ways of

knowing, yet from the psychological standpoint, it is not

possible to draw the line between them.

As we have already said, perception is constituted by
the union of certain sensuous elements with certain non-
sensuous presentations into one whole. The sensuous ele-

ments are the given and the non-sensuous elements are

presented by the given. As a general rule, perception
occurs when there is either a natural or a habitual associa-
tion between the sensuous and the non-sensuous elements of

perception, so that the one immediately calls up the other

and the two are fused into one whole of presentation of some
object. By ‘
natural association ’
I mean an association, in
our minds, of one sense quality with such other sense quali-
ties as are, by nature, connected with it It follows natur-

ally from this that such associations will be most frequently

experienced by us and therefore ingrained in our minds.
The question as to how many senses and sense qualities
there are is not relevant for our present purpose We adopt
here the universally accepted view that there are five senses
and five corresponding sense qualities, namely, smell, taste,

colour, touch and sound. Taking these sense qualities in the

order in which they have been mentioned here, we may say that
there is a natural connection of that which precedes to those
which succeed it, but not vice versa Thus a smelling object
has generally some taste, colour, touch and sound. But a
sounding object like space or air has no colour, taste and
smell. So also, to smell a thing is generally to see and
touch it. But to see a thing is not to taste or smell it so
generally. Hence it is that a smell or taste sensation
naturally calls up the colour and touch of an object.
On the other hand, a ‘
habitual association ’
is the asso-
 EXTRAORDINARY PERCEPTION 245

cidtion, in our minds, of one sense quality with those which

are not, by nature, connected with it. Hence it is less

frequently experienced by us and is, in many cases, looser

than a natural association. ‘ Still when a habitual associa-

tion between them is once formed in our minds, one sense
quality immediately calls up others and all of them are
combined into the perception of an object. Thus sound is

not naturally connected with touch as there is sound in

space but no touch. So also, touch has no natural relation
to colour, as there is touch in the air but no colour. Simi-
larly, there is no natural connection of colour with taste and
smell Thus there is colour in light but no taste and smell.
Hence it is that an association of colour with smell is looser
than that of smell with colour, with which it is naturally
connected While the mere smell of kerosene immediately
suggests its colour, the latter by itself hardly suggests the
former. But by repeated experiences of their uniform con-
nection with one another a habitual association between
them may be established in our minds, and one sense quality
may present another which is not naturally connected with
it. It is in this way that the sound of a bell presents its

tactual and visual qualities and we have the auditory percep-

tion of a bell, just as the sight of the bell presents its tactual
and auditory qualities and we have a visual perception of it.

Thus we say “ I hear the bell,” just as we say “ I see the

bell.”
The two alternative conditions of perception, viz. a
natural and a habitual association between its two elements,
are realised in the life of an individual through repeated
experiences of the objects of the world. When we speak of

^ This IS also admitted by some Western thinkers Cf Stout, Manual of Psychology

pp 102-03 Ifi the qualification of actual touch e>perience by revived visual experience
‘

we find the union of the constituents of the complex much looser/ Cf also Bosanqueti
The Essentials of Logic, p 31 The judgment of sight perception, “ That (which I
• *

see) is a cab,” though its terms are more mextrtcahly interwoven, has just the same
elements m it as the judgment of sound perception,
**
That (which I hear) is a cab.*’
*
 NYIYA THEORY OF KNOWLRDOB
24C

a natural association between them we do not mean that it

is congenital, so that the association is formed in the mind

of every individual from his or her birth. If it were so,
every baby should have as good perceptions as any grown-up
person. All that we mean by a natural association is that it

has a basis lu the constitution of things and that it is most

frequently met with in the experiences of an individual. If,

therefore, both natural and habitual associations are produced

by repeated experiences, two important consequences would
follow : (1) It follows that the knowledge of the same object

may, under the same objeetpe conditions, he a matter of per-

ception or inference for different individuals, according as
there is or is not a natural or a habitual association as the
basis of their knowledge (2) It follows that, under the same
oh]ective conditions, the knowledge of tlie same object may at

first be a matter of inference and subsequently of perception,

and vice versa, for one and the same individual. The know-
ledge of a rose from its sight is a perception for the person

in whom the repeated experiences of roses have established

a natural or a habitual association between its colour and
touch and smell. The knowledge of the same rose, under
the same external conditions, will be an inference for

another person when, for want of repeated experiences, its

colour does not immediately call up its touch and smell, but
suggests them as distinct ideas or images on the ground of
its similarity to the colour of roses. For illustration we
may refer to the difference between our knowledge of a new
variety of the rose and that of the gardener who presents it to

us. An armour may look hard, smooth and cold to an adult

who is familiar with it. A child may just imagine it to

have these or very different tactual qualities. The fragrance

of sandalwood may be directly known from its sight by those
who are closely acquainted with it, but for others its visual
appearance may only be a sign from which to infer its fra-

grance. When I go to a foreign country I can barely infer

the distance and size of an object from its Tisual appearance,
but a native of the country who is familiar with the environ-
ment has a perception of these from the same position,

which is essentially visual. If, however, I stay there for

some months and become familiar with the environment, I

may have a visual perception of the same facts under the

same objective conditions. For the illustration of the con-
verse case, i e. the relapse of perception into inference, we
may refer to (i) any case of senility in which a person has
but a doubtful inferential knowledge of things and persons
under the same objective conditions under which he or she
once used to perceive them, and (ii) the common though
curious instance of forgetfulness in which we fail to recog-
nise an old acquaintance whom we tiave not seen for many
years and try to infer his identity by putting certain
questions to him and judging their answers.

The foregoing discussion leads us to the conclusion that there

cannot b e an absolute line of demarcation between perceptjnn

and inference^ We cannot say that under the same objective
conditions the knowledge of an object must always be a per-
ception for every individual and that it can never be other-
wise. Nor can we aver that such knowledge must always
be an inference for all individuals and that it can never be a
perception for any. On the contrary, we are to admit that it
may be either, for different individuals, or for the same in-

dividual at different stages of his or her life, according to

his or her or their mental equipment.

In view of the answers which we have given to the two

questions stated above we are justified in saying that, under
certain conditions, it is quite possible that there may be a
presentation of fragrance in relation to the activity of the

visual sense. When these conditions are fulfilled, there is

nothing to prevent the eye from giving us an immediate

knowledge of fragrance. And that is why of all people the

Yedantins must admit that there may be a visual perception

 BOOK III

THE THEORY OF INFERENCE

(ANUMANAPRAM5.NA)
 CHAPTER XI
THE NATURE OF INFERENCE
1 . Definition of Anumdna or Inference

Anumdna literally means such knowledge as follows

some other knowledge. It is the knowledge of an

object due to a previous knowledge of some sign
or mark (linga).^ The previous knowledge is the know-
ledge of the linga or mark as having a universa
relation with the sddhya or major term and as bein^

present m the paksa or minor term.^ Hence anumdna has

been defined in the Nyaya system as the knowledge of ar

object, not by direct observation, but by means of thf

knowledge of a linga or sign and that of its universal relatior

(vyapti) with the inferred object.®

The object of inference is some fact which follows froir

some other fact because of a universal relation between the

two. With regard to something of our experience we wani

to know by means of anumdna that which may not be per-
ceived but 18 indicated by what is perceived in it. Anumdna
as a pramdna is therefore the source of our knowing through
the medium of a sign or mark that a thing has a certain
character. It leads to the knowledge of a thing as possess-
ing a character, say fire, because of its having another
character, smoke, which we apprehend and which we know
to be always connected with it. Thus anumdna we arrive
in

at the knowledge of an object through the medium of two

acts of knowledge or propositions.

^ Miteaa liAgena IiJ^gino'rtbasya pa^c&nmUnamaQum&nam, N£ ,

1 1 8
2 VyaptivijSifta-pak^adharmatajfiana^anyam, etc., TC n, p
, 2.

8 NM.,p. 109.
254 NyiYA THEORY OF KNOWLEDGE

All systems of Indian philosophy agree in holding that

anumana is a process of arriving at truth not by direct
observation but by means of the knowledge of vyapti or a
universal relation between two things. The Nyaya view is

stated already. According to the Vai^esikas, anumana is

the knowledge derived from the perception of a hnga or

sign which is uniformly connected with something else,

such as cause, effect, co-effects and correlative terms. ^ The

Buddhists take anumana to consist in the perception of
that which is known to be inseparably connected with
another thing Such inseparable connection between two
things is due either to the law of causality or the principle of
^
essential identity {tadutpatti and fadatmya) So also the
Jainas hold that anumana is the method of knowing an un-
perceived object through the perception of a sign and the
recollection of its invariable concomitance with that object.®

The Sarnkhya and the Yoga, the Miraainsa and the Vedanta
system too define anumana as the knowledge of one term
of a relation, which is not perceived, throiigli the knowledge
of the other term which is perceived and is explicitly under-

stood as invariably related to the fiist term.^ In anmnana

what IS perceived leads us on to the knowledge of what is

inferred through the knowledge of a universal relation

(vyapti) between the two

2. Distinction between perception and infeience

Perception and inference are equally valid methods of

human knowledge (pramana). But while perception is in-

dependent of any previous knowledge, inference depends on

previous perception I Inference is sometimes defined as

1 PS pp
,
99f
2 NyayahmdUt Chapter II
3 Pfameyakamalamdrtanda, p, 101
* TKD , 5 , VB., 1 7 ; SD & SC , p 00, VP ,
Ch. 11.
 NATURE OF INFERENCE 265

knowledge which is preceded by perception.) It depends

on perception for the knowledge of the linga or the middle
term as subsisting in the paksa or the minor term. It

depends on perception also for the knowledge of vyapU or

the universal relation between the middle and major terms
of inference.^ It is only when we have observed two things
to be always related that from the perception of the one
we infer the existence of the other. Thus inference is
knowledge derived from some other knowledge, while per-
ception is not derived from any other knowledge. That is,,

inference is mediate and perception immediate knowledge of

an object.
All perception is essentially of one kind, namely, that it

IS a knowledge of what is given But there are different

kinds of inferences based on different kinds of vyapti or

universal relation. Perception is generally due to some con-
tact of our sense-organs with the objects perceived by us. It
gives us knowledge of only those objects which he within
the range of the senses. Hence it is limited to the here and
the now, i.e. to present objects Inference, on the other
hand^ is due to the knowledge of vyapti or universal relations

among objects. It is by means of such universal principles

that inference gives us a knowledge of objects beyond the reach
of our senses. It extends our knowledge from the present
to the past^ distant and future. Ordinarily we perceive
objects that are in actual contact with our senses, but we
infer those that are not open to sense perception.® Percep-
tion usually excludes inference but not vice versa. What is

perceived or directly known does not ordinarily require to

be known indirectly by means of inference. Inference
functions with regard to neither what is absolutely unknown
nor what is definitely known. It relates to objects that are

1 NS ,
1 1 6
» NB., 1. 1 6
5 NV,2181.
 256 NYiYA THEORY OF KNOWLEDGE

doubtful, i.e. objects which we have reasons to believe in,

but which are not yet established facts. Hence inferences

generally require confirmation by means of perception.^

3. The Gonshtuents oj Inference

From the definition of inference {anumana) it will appear

that there must not be less than three propositions and
more than three terms m any inference In inference we
arrive at the knowledge of some unperceived character of
a thing through the knowledge of some hnga or sign in it

and that of vyapti or a universal relation between the sign

and the inferred char.icter. There is first the knowledge
of what IS called the hnga or mark in relation to the paksa
or the subject of inference This is generally a perceptual
judgment relating the hnga or middle term with the paksa
or minor term of inference (hngadarhna), as when I see
that the hill is smoky, and infer that it is fiery. ^ It is a
proposition in which the hnga is predicated of the paksa
and thus corresponds to the minor premise of a syllogism.
Secondly, inference requires the knowledge of vyapU or a
universal relation between the hnga and the sadhya, or the
middle and major terms This knowledge of the lihga or
middle term as always related to the sadhya or major term
IS the result of our previous experience of their relation to

each other. Hence it is a memory-judgment in which we

think of the hnga as invariably connected with the sadhya
‘

(vydphsrnarana), e.g. all smoky objects are fiery.’ Thirdly,

we have the inferential knowledge (anumiti) as resulting
from the previous knowledge of the Hnga and that of its

^ Agnerd]r§t&tvena samdeba8y&nuda)at. Saihdigdbaic&rtho 'namlyate, TB., p. 8

2 It should bcj observed here that the Brst step of inference may be either the

perception of the middle term or a proposition stating that the middle term is related

to something.
 NATURE OF INFERENCE 257

universal relation (vyapti) with the sadhya. It is a propo-

sition which relates the paksa or minor term with the
‘
sadhya or major term, e.g. tlie hill is fiery.’ The inferen-
tial cognition (anumiti) is a proposition which follows from
the first two propositions and so corresponds to the conclu-
^
sion of the syllogism.

Corresponding to the minor, major and middle termi

of the syllogism, inference in Indian logic contains thre(
terms, namely, paksa, sadhya and hetu. The paksa is the
subject under consideration in the course of the inferential
reasoning Every inference proceeds with regard to some
individual or class ot individuals about which we want to
prove Hence the paksa is that individual or
something ‘

class about which we want to establish something or pre-

dicate an attribute which is suspected but not definitely

known to be present in it. ^ That which possesses the
inferable character is called paksa or minor term of in-
‘
ference, e g. the hill ’
when we want to prove that it is

fiery. In relation to the paksa or minor term in any

inference, a sapaksa or homogeneous instance is that which
IS decisively proved to be related to the inferable character,
‘
e.g, the hearth ’
in relation to ‘
the hill.’ ® Con-
trariwise, a vipaksa or heterogeneous instance is that which
IS definitely known to be characterised by the absence of
the mfeiable character, e.g. ‘
water ’
as marked by the
absence of ‘
fire.’^

While the paksa is the subject, the sadhya is the object

of inference. It is that which we want to know or prove

by means of any inference. The sadhya is that character

of the paksa or minor term which is not perceived by us,

’ SM ,
66 67 , NM , p. 109.
2 SaihdigdhaBadhyavan paksa TS., 44 Of Nanupalabdhe na nirpTte’rthe

nyajalj pravartate, kintu sam^ayite, NB., 1.1

3 Ni^citasadhyavan 8apaksa^^, TS., 44.

* Ni^citasadhyabhavavan vipaksal?, Ibtd,

33— (1117 B)
 258 NYiYA THEORY OF KNOWLEDGE

but indicated by some sign present in it. In short, it is

the inferable character of the minor term and thus corre-

sponds roughly to the major term of the syllogism. It is

that character which is predicated of the minor term m the

resulting inferential knowledge or the conclusion of the
syllogism.

With regard to the exact nature of the sadhya there is

some difference of opinion among the different systems of

Indian philosophy. According to the Advaita Vedanta, what

is inferred is the unperceived character of the subject or minor
term of inference. In the inferential knowledge that ‘
the
hill is fiery,’ it is ‘
the fire ’
that is inferred and not ‘
the
hill ’
which is but perceived. The Buddhists contend that
‘
the fire ’
cannot be the object of inference from smoke.
We know it just when we know the smoke as related

to fire. So there remains nothing more to be inferred.

Nor do we inter the relation between ‘

the fire ’
and ‘
the
hill.’ We cannot speak of a relation unless there are two
things to be related. But in inference we have only one
thing, namely, the hill, since the fire is not perceived.
The hill being perceived cannot be said to be the object of
inference. What is therefore inferred is ‘
the hill as

possessed of fire.’ ^
The Mimamsakas also hold that what
we infer is the subject or minor term as related to the
predicate or The Naiyayikas however
the major term. ^

maintain that the object of inference may be different in

different cases. What is inferred may be either the subject
or minor term as related to the major term, or the major
term as related to the minor, or the middle term taken as a
particular individual and related to the major term. ® When
we perceive smoke in a hill, what we know by inference is

1 Pramdnasamuccayat Chapter 11 Ffde also NVT,, pp 179-80.

* Tasmftt dharmavi^i^tasya dharallJ^a^;l syit prameyata, SlokavdrtUkat Chapter or
Inference, Agnivi^iftastu parvato anameyalj, 8D , p 63
> and NV , 1,1.36,
 NAtURE OF INFERENCE 269

either ‘
the hill as related to fire,’ or ‘
fire as related to the
hill,’ Bat when the site of the smoke cannot be perceived,
what we infer is that the perceived individual smoke is
^
related to fire.

The third term of inference is called the linga or sign

because it serves to indicate that which we do not perceive.
It is also called the hetu or sadhana in so far as it is the
ground of our knowledge of the sadhya or what is inferred.
Like the middle term of a syllogism, it must occur at least

twice in the course of an inference. It is found once in

rel ition to the pahsa or minor term and then in relation
to the sadhya or the mijor term. It is through a universal
relation between the hetu and the sadhya, or the middle
and major terms that the paksa or minor term, which is

related to the middle, becomes connected with the sadhya

or major term ^ That is, the paksa is related to the sadhya
through their common relation to the hetu or middle terra.

There are five characteristics of the middle term ^

The first

IS paksadharmata, or its being a character of the paksa.

The middle term must be related to the minor terra, e.g.
the hill IS smoky (S is M). The second is sapaksasattva
or its presence in all homogeneous instances in which the
major The middle must b e distributively related to
exists.

the major, eg. all smoky objects are fiery (M is P). The
third IS vipaksasattiia or im absence in all heterogeneous
instances in which the major is absence. gf. whatever is

not fiery is not smoky (No not-P is M) The fourth is

abddhitamsayatm, or the uncontradictedness of its object.

The middle term must not aim at establishing such absurd

and contradictory objects as the coolness of fire or the
squareness of a circle. The fifth character of the middle
IS asatpratipaksatua or the absence of counteracting reasons

‘ Vide NVT , p 182

2 V}aptibalenarthap:amakajti lifigain, TB , p 7.
3 Vtde NM , p 110, N8V ,124
 NYiYA TfiEORY OF KNOWLEDGE

leading to a contradictory conclusion. These five character-

istics, or at least four of them, must be found in the middle

term of a valid inference If not, there will be fallacies.

We shall have to consider these points more fully later

on.

260
 CHAPTER XII

THE GROUNDS OF INFERENCE

1. The logical ground of vyapti or universal relation

In inference our knowledge of the sadhya or major

term as related to the pahsa or minor term depends on the
knowledge of ryaph between the middle and major terms.*
It is on the ground of vyapti or a universal relation that
the middle term leads to the knowledge of the inferred
object {vydptihalendrthagamakam lingam). Every inference
is thus logically dependent on the knowledge of vyapti.
Hence the questions that we have to consider here are
(i) What IS vyapti^ and {ii) how is it known?
With regard to the first question we have to say that
vyapti literally means the state of pervasion or permeation.
It thus implies a correlation between tw’o facts, of which
one is pervaded (vyapya) and the other pervades {vydpaka),

A fact IS said to pervade another when it always accom-

panies the other. Contrariwise, a fact is said to be
pervaded by another when it is always accompanied by the
other. It follows from this that the vydpaka or the
pervader is present in all the places in which the vyapya or

the pervaded is present. In this sense smoke is pervaded

by fire, since all smoky objects are also fiery. But while
all smoky objects are fiery, all fiery objects are not smoky,
e.g. the red-hot iron ball. Similarly, all men are mortal,

but all mortals are not men, e.g. birds and beasts. A
vyapti between terms of. nequal extension, such as smoke
and fire, men and mortals, is ca lled asamavydpti or visama-
vvavti. It is a relation of non-equipollent concomitance
262 NYiYA THEORY OF RNOWLKOGE

between two terms, from one of which we may infer the

other, but not vice versa. Thus we may infer fire from
smoke, but not smoke from fire. As distinguished from this,

a vyapti between two terms of equal extension is called

samavyapti or equipollent concomitance. Here the vyapti

holds between two terras which are co-extensive, so that we
may infer either of them from the other. Thus there is a
samavyapti between cause and effect, substance and attribute.
We may infer the cause from the effect, the substance
from the attribute, or vice versa Thus whatever is produced
IS non-eternal, and whatever is non-eternal is produced.
It will appear from the above that visamavydpti is a
universal proposition, of which only the subject is distri-

buted, i.e. taken in its entire extension A samavyapti, on the

other hand, is a universal proposition which distributes both
the subject and the predicate. They would thus correspond
respectively to the universal aflSrmative and universal nega-
tive propositions in Western logic. It is to be noted how-
ever that there are some universal affirmative propositions
which distribute both their subject and predicate Thus
‘whatever is produced is non-eternal,’ ‘men are rational
animals ’
are cases of samavyapti or universal affirmative
propositions in which both the subject and the predicate are
distributed.

For any inference the minimum condition is some kind

of vyapti between the middle and major terms. It does not

matter whether the vyapti is sama or visama, i.e. equipollent

or non-equipollent. This satisfies the fundamental law of

syllogistic inference that one of the premises must be
universal. Now the vyapti between the middle and
major terms means generally a relation of coexistence
{sahacarya) between the two, e.g. wherever there is smoke
there ia fire. ‘
Every case of coexistence, however, is not a

^ Yatra dhumastatragninti sahacarya niyamo vyaplili, TS p ,

45
 GROUNDS OF INFERENCE 263-

oase of oyapti. Thus all the children of a certain father

may be dark. But this does not mean that there is vydpti
or a universal relation between a particular parentage and
dark complexion. In many instances fire may coexist with
smoke. Still there is no vydpti or universal relation between
fire and smoke, since there may be fire without smoke.
The reason is that in such cases the relation of coexistence
IS dependent on certain conditions (upadhi) other than the
terms related. Thus tlie darkness of complexion is deter-
mined by certain physiological conditions, and the presence
of smoke in fire is conditioned by moisture in the fuel.

Hence we are to say that injdpti is that relation of co -

middle and^ major terms which is

ind^pendent^of all conditions {vpddhi)^ It is an invariable

and unconditional lelation {myata anmpddhika savihandha)
of concomitance between the middle and major terms. This
means that there is no exception to the relation of con-

comitance between the two, no instance in which the middle

IS present without the major. Hence vydpti as the logical

condition of inference may be defined either positively or

negatively. Positively speaking, vydpti is the uniform
existence of the middle term in the same locus with the
major term such that the major term is not absent in any
locus in which the middle term exists. In the terminology
of the Navya Nyaya, vydpti is such a relation of coexistence
between the middle and major terms that the major is not a
counter-entity to any negation abiding in the middle, i.e.

it IS none of those things which are absent in the middle

term. ^ Vydpti has been negatively defined as the non-
existence of the middle term in all the places in which the
major term does not exist. ® That there is vydpti between

1 Vide TB pp 7-8
,

^ Athav& hetumanni^thavirahapratiyogina sadhyena hetoraik&dh’kara^yam vyipti-

rucyate. BP , 69 Vide also T C IT, p 100. Cf ,
VP Ch .
II.

^ Vyaptih sadhyavadaDyasmmuaaambandha, etc., BP., 68,

 264 NYAYA THEORY OF KNOWLEDGE

the middle and major terms means that the middle (M)
never is, if the major ( P) is not. These two definitions of
vyapti give us two universal propositions, one positive and
‘
the other negative, e g. all cases of smoke are cases of

fire,’ and ‘
no case of not-fire is a case of smoke ’
(All M is

P, and No not-P is M). This means that the vyapti or

universal proposition which is the ground of inference may
be either affirmative (anvayi) or negative {vyatirehi) Hence
vyapti is said to be of two kinds, namely, anvaya or affirma-
tive and vyatireka or negative While in anvaya-vyapti or
the universal affirmative proposition the middle term is

vyapya or subject and the major is vyapaha or predicate, in

vyatireka-vyapti or the universal negative proposition the
contradictory of the vyapaha or predicate becomes vyapya or
^subject and the contradictory ot the vyapya or subject
becomes the vyapaka or predicate '
Thus ‘
whatever is

smoky is fiery,’ or ‘
All M is P ’
is an anvaya-vyapti, of
which the corresponding vyatireka-vyapti will be ‘
whatever
is not-firey is not-smoky,’ or ‘All not-P is not-M.’ The
logical ground of inference then is ai/dph in the sense of a
universal proportion which may be either affirmative or
negative.
^
So much for the definitions of vyapti oi the universal
relation between the middle and major terms of inference.
The next question is ; How is vyapti known? How do we
pass from parti lular cases of the relation between smoke
and fire to the universal proposition ‘
all cases of smoke are
cases of fire ’? This is the problem of induction, which is

not separately treated in Indian logic, but is made a part

1 VideTB.p 9

2 Gange^a in his Tattvaemtamant discusses at length the dehnitions of vyapti,

lam'.ering twenty in all Of thes^ he rejects all but one as either too wide or tjo

larrow or otherwise objectionable The last definition of vyapti^ accepted by him as

ralid, has been noted above, along with another to cover both the cases of anvaya and
jyaUreka vyapti or the universal affirmative and the universal negalrve proposition.
 GROUNDS OF INFERENCE 265

of the general theory of inference. Indian systems of

philosophy take inference as a process of reasoning which is

not only formally valid but also materially true. Hence

in an inference of any kind the question arises : How do
we get the universal proposition (vyapti), on which inference
depends ?
The Carvakas, who are radical empiricists, contend that
all knowledge is limited to particulars. We cannot pass
from the knowledge of particular cases of the relation
between two objects to that of all possible cases. There is

no successful method of generalisation from particulars.

Perception, which is the only source of human knowlege,
does not help us to establish a universal proposition.
It is limited to present facts and cannot tell us anything
about the past, distant and future. Hence from perception
we know what is true of a thing or a limited number of
things. That is, perception gives us particular, but no
universal propositions. If perception cannot give us a
knowledge ot vyapti or universal proposition, inference and
the other alleged sources of knowledge certainly cannot do so.

All sources of knowledge except perception depend on vyapti

or .1 universal relation between two things and cannot,
therefore, be made the ground of our knowledge of it.
*
To take them as such is to reason in a vicious circle.

The Buddhists meet the Carvaka contention in two ways.

First, they point out that the Carvakas’ refutation of
inference is itself a process of reasoning which, on their
own admission, depends on some kind of vyapti. As such,
it practically amounts to a refutation of their own position,

namely, that no process of reasoning including inference

IS valid. As a matter of fact, the Carvakas employ the
method of inference more than once in their philosophy.
For example, it is by means of inference that they can know

^ Vide Sarvadar^anasamgraha, Chapter 1,

34— (1117 B)
266 NYIYA THRORY OF KNOWLEDGE

that other people differ from them with regard to the

question of inference, or that other sources of knowledge
are as fallacious as inference, or that God, soul, etc., do not
exist because they aie not perceived.
Next the Buddhists proceed to show how vyapti or a

universal proposition may be based on the principles of

^
causality and essential identity and tadwtpatti).
When two things are related as cause and effect, they are
always and everywhere lelated to each ether There can
lie no exception to their relation, since the cause cannot he
separated from its effect, nor the effect from its cause. To
say that there is no necessary relation between the cause
and the effect, or that there may be an effect without its

cause is not admissible, because such a hypothesis involves

self-contradiction and makes life impossible. Hence we are
to take the law of causality as a universal law. 'J'o

determine whether the relation between two objects is causal

or not we are to apply the test of pancakaranJ According
to it, there are five steps in the deteimination ol a causal
relation First, the effect is not perceived before it is

produced. This means that the effect is an event which

appears after another phenomenon that is its cause The
causalphenomenon is thus antecedent to the effect-
phenomenon. Secondly, the cause is perceived, le. there
IS a change in the existing order of things. Thirdly, the
effect-phenomenon appears in immediate succession.
Fourthly, the cause is made to disappear. Fifthly, the
effect disappears in immediate succession. The Buddhist
method of determining the causal relation corresponds to
Mill’s method of difference in its double application. If, all

other conditions remaining the same, the appearance of one

phenomenon is immediately followed by that of another,
^ Op. at , Ch ir
* Op at ,
Ch II; Nydyahttidu^ Cb II
' K&ryasyotpatteli prftgantipalarhbbab etc , SarvadarSanasaihgraha, Ch II.
 GltOUNDS OF INFEtlENCte 267

and its disappearance is immediately followed by the dis-

appearance of the other, then the two are related as cause

and effect. When once we know them to be related as cause

and effect, we may very well take them as universally
related. Similarly, the principle of essential indentity
{tadatmya) is another ground on which we may base a
universal proposition. A thing is always related to what is

identical with it. Identity does not mean a mere repetition

of the same thing, e.g. ‘
A is A ’
Nor can there be any
identity between things that are absolutely different, e.g.
a horse aud a cow By identity we mean the relation
between two different things that coexist in the same
locus {sdmdnadhtkaranyam) Thus there is identity between
the genus and the species coming under it, or the class
and the individuals included in it A hm§apa is identical
with a tree, in so far as the two refer to the same object.
From this we know that all hmhpds are trees, .since

hmsapds will cease to be hm^apds, if they are not

trees Thus oydpti
'
or a universal proposition is to

be based on the necessary principles of causality and

identity Experience, oi observation and non-observation
cannot be the sure ground of generalisation. Empirical
knowledge, however well lounded, can never be necessary
and universal. There is an clement of doubt and uncertain-
ty in ail empirical generalisations Causality and identity
being the presuppositions of all experience are necessary and
universal truths, to which all sense experience must conform.
Hence any generalisation based on either of these two
principles is universally valid and not open to any doubt.
The Naiyayikas criticise aud reject the Bauddha method
of ascertaining vydph on the following grounds. According
to the Buddhists, oydpti or a universal relation between the
middle and major terms is to be deduced from the relation of

^ Sarvadarsanasamgraha and NydyahtndVithtd

268 NYiYA THEORY OF KNOWLEDGE

causality or identity between the two. This, however, is not

true. There are many cases of vyapti or universal relation
which is independent of the notions of causality and identity.
Thus there is a universal relation of succession between
day and night, or between the different seasons, or between
sunset and the appearance of stars Similarly, we find a
universal relation of coexistence between a certain substance
and its attributes, or between a certain colour and a certain
taste. Here we have vyapti or a universal relation between
terms which are neither cause and effect nor identical with
one another^ but fiom one of which we can validly infer the
other. Further, the relation of identity between two things
can hardly be treated as a ground of inference from the one
to the other. If the two things be identical, then both must
be equally perceived or inferred. The tree being identical
with the Hm^apd must be perceived just when we perceive
the latter, and so need not be inferred. For the same rea-

son, we infer the tree from

if the hmiapd and say “all
Mmiapds are trees,’’ we should be able to infer the Mm§apd
‘
are hmhpds
’

from the tree and say ‘

all trees The two’

things being identical, we should be able to infer either of

them from the other. Nor again does the abstract principle
of causality help us to draw inferences in particular cases
Granting that there is a universal and an unconditional
relation between the cause and effect, it is extremely difficult

for us to determine whether the relation between two parti-

cular things IS causal or not. The test of paflcakdranl

recommended by the Buddhists is not an absolute guarantee

for there being a causal relation between two things. That
test applies when all the conditions of a certain relation

remain the same. But it is only with regard to the known or

the perceptible conditions that we may be sure whether they

remain the same or not. With regard to the imperceptible
conditions we cannot be absolutely certain that no change in
these corresponds to a change in the relation between two
 GbOUNDS OF INFERENCE 269

things. Thus in the relation of fire to smoke it is just

possible that an invisible agent {pUaca) always intervenes
between the two and produces the smoke. Moreover, there
is such a thing as a “ plurality of causes,” which makes it

hazardous to infer any particular cause for any single effect.

Thus we may admit that fire is the cause of smoke ma

particular case, and yet say that it is not a cause in other
cases, or that there are other causes producing smoke in
other instances. Hence it is not always safe to infer a
^
particular cause from an effect as such.

According to the Vedanta, * vyapti or a universal propo-

sition IS the result of an induction by simple enumeration.
It rests on the uncontradicted expeiieiice of agreement in

presence between two things. When we find that two

things go together and that there is no exception to their

relation, we may take them as universally related. The

Nyaya agrees with the Vedanta in holding that vyapU is
established by means of uncontradicted experience of the

relation between two things It is based, not on any a

pi ion principle like causality oi identity, but on the uniform
experience of concomitance between two objects The
Nyaya, however, goes further than the Vedanta and supple-
ments the uncontradicted observation of agreement in pre-
sence by that of agreement in absence and tarka or indirect
proof.* The Nyaya method of induction or generalisation

may be analysed into the following steps. Fir^st we

observe that there is a uniform agreement in presence
(anvaya) between two things, or that in all the cases in
which one is present the other also is present.

Secondly, we see that there is uniform agreement in

absence (vyatireka) between them, i.e. in every case

' NVT, pp 168-64, NM , pp. 113.17.

2 VP ( h II.
,

^ Vyabb’oar^syagrBho’tba eabacaragrabastatbft, betnrvyaptigrabe tarkah kvaci-

( t bnnkanivartaknh BP , 137.
2t0 NYiYA THEORY OF KNOWLEDGE

in which the one is absent the other also is absent.

So far we see that the two things go together both in
their presence and absence, or that there is positive and
negative coincidence between them (sahacdra). Tbirdl^^ we
do not observe any contrary instance in which one of them
is present without the other (vyahhicdrdgraha'^ From this

we conclude that there must be a natural relation of invaria-

ble concomitance between the two things.^
Still, we are not sure it their relation is dependent on
any condition (upddhi) or not. Vydpti or a universal
relation between two things is that relation of concomitance
between them which is independent of all upddhis or con-
ditions An upddhi or condition is a term which is co-

extensive with the major but not with the middle term of
an inference ^
Thus when one infers the existence of

smoke from fire, he relies on a conditional relation between

fire and smoke, since fire is attended with smoke on condi-
tion that it 18 fire from ‘
wet fuel ’
It will be seen here that

the condition of ‘
wet fuel ’
is always related to the major
term ‘
smoke,’ but not so related to middle term ‘
fire,’

as there are cases of fire without '

wet fuel ’
Hence to

make sure that a certain relation of uniform concomitance

between two things is a vydpti or a universal relation, we
must eliminate all conditions. This can be done by repeated
observation {bhuyodar§ana) of their agreement in presence
and absence under varying circumstances. Here if we see

that there is no material circumstance which is present oi

absent just when the major term is present or absent, we

are to understand that its concomitance with the middle
term is unconditional. In this way we can exclude all the
suspected conditions of a relation of concomitance between
the middle and major terms and say that it is a relation of

> 8M..137
* A-vyaptasadhano sadhyasamavyaptirucyate sa up&dbinti, Sarvadaritma*
samgraha, Chapter I. Vide also BP., 138
 GROUNDS OF INFERENCE 271

vyapti or unconditional concomitance.^ If even after

repeated we have any doubt as to there being
observation
vyapti or a universal relation between the middle and
major
terms, we are to have recour se to tarJca or in direct proof
to end such dou bt. Thus the universal proposition, *
all
cases of smoke are cases of fire,’ may be proved indirectly
by disproving its contradictory If this universal proposi-
tion be false, then its contradictory, ‘
some cases of smoke
are not cases of fire,’ must be true. This means that there
may be smoke without fire. But the supposition of smoke
without fire is contradicted by the known relation of
causality between'fire and smoke. To say that there may
be smoke without fire is to say that there may be an effect
without its cause, which is absurd. If any one has the
obstinacy to say that sometimes there may be effects without
causes, he must be silenced by the practical contradictions
(vydghdta) involved in the supposition If there ean be an
effect without a cause, why should he constantly seek for
fire to produce smoke or for food to alleviate his hunger ?
Thus its contradictory being proved to be false, the universal
proposition ‘
all cases of smoke are cases of fire ’
comes
out as true, i.e. there is vyapti or a universal relation
between smoke and fire.

So far the Naiyayikas try to establish vydpti or a

universal proposition by the method of simple enumeration
supported by tarka or a hypothetical reasoning which
indirectly proves its validity By examining a number of

positive and negative instances ot agreement in presence

and absence between two things, they conclude that there

IS a universal relation between them. This conclusion is

then indirectly confirmed by showing that a denial of the

universal relation between these two things leads to con-

tradictions. But as we have already seen in connection

1 Vide, TB , pp. 7-8

• Vide SM., 187 : TC ,
II, pp 210-12.
272 NYAYA THEORY OF KNOWLEDGE

With their theory of samdnyalakmia perception, a genenil

proposition like ‘
all smoky objects are fiery ’
cannot
”
be logically proved by ‘^Induction by Simple Enumeration
In simple enumeration we pass from some observed cases
of the relation between two things to a statement about
their relation in all cases. Thus from some observed cases
of the relation between smoke and fire we infer that all

smokes are related to fire. But this inference is not valid,

since it violates the general rule of inference that we must
not go beyond the evidence. The method of simple enu-
meration cannot, therefore, conclusively establish vydpti or
a universal proposition. Hence the question is : How
from the observation of some smokes as related to fire do

we know that all smokes are related to fire The

Naiyayikas explain this by the help of sdmdnyalaksana
perception. The universal proposition ‘
all smokes are

related to fire ’
cannot be explained by the perception of

particular instances of smokes as related to fire, for any

number of particulars cannot make up the universal. For
this we require a perception of the whole class of smokes as
related to fire. We have such a perception through the per-
ception of the universal ‘
smokeness ’
as related to ‘fireness.’

In perceiving particular smokes we perceive the universal

’

‘smokeness’ inhering in them. But to perceive ‘

smokeness
18 to perceive, in a non-sensuous way, all smokes so far as

they possess the universal ‘

smokeness.’ Hence the uni-
versal proposition ‘
all smoky objects are fiery ’
is given by
a non-sensuous perception of all smokes as related to fire
‘
through the perception of smokeness as related to fireness
The Nyaya method of establishing vyapti brings out
the importance of class-essences or universals for induction.

1 SamaiiyaUkBa^^m vina dbumatvena Bakaladhutn&Bam vaimitvena sakala-

vahnlD&di ca bhanam katbam bbavet tadartham samaDjalaksai^a sviknyate, etc.»

SM., 65 Vyaptigraha^ca BamaDyalai{9anapratjaBatty& sakaladhumadiviilfayakal^, etc,,

TC , II, pp 153-54,
 GROUNDS OF INFERENCE 273

It shows how the validity of a generalisation from the

particulars of experience depends ultimately on the discovery
of certain common essences or universal characters of
particular things. From the observation of a limited
number of instances ol the relation between two things
we cannot know anything for certain about all possible
instances of them unless we find that the things possess a
certain essential nature whicli is the basis of their relation
in some cases. The particular objects of experience lend
themselves to a generalisation when they are recognised as
instances of a class and possessed of some essential common
nature. A number of things are arranged in one class in
view of such common essence or universal which is present
in all the members of that class, but absent id those of
a different class Hence if in some cases we see that some-
thing IS related to the essential nature or the universal
underlying a class, we know that it is related to all the
members of that class. The observation of particular
instances is important because it helps us to find the uni-

versals underl;ying different classes of things and their

relations with i/iie another. Hence the problem of induc-
tion IS the problem of the discovery of class-essences or
universals in particular things
exemplified As we have
already remarked, some Western logicians are slowly recog-
nising the truth of the Nyaya view that an inductive
generalisation must be based on the knowledge of class-
essences or universals embodied in particular things. But
they do not go so far as to say with the Naiyayikas that an
empirical generalisation from particular instances is a

matter of uon-sensuoiis intuition based on the perception

of universals. They would generally treat it as an infer-

ence from known resemblance or as a perfect analogy. Mr.

Eaton, however, goes further and maintains that the first

step in induction is a direct perception of the universal in

the particular. He says :

“ Induction proceeds from the

35-(1117B)
 274 NYiYA THEORY OF KNOWLEDGE

particulai to the general, but not from the sheer particular.

The particular must be seen embody some characters or
to
relations, to cxomplily somefoim Given a pailicular, let
us say a blinding streak of light, and another particular, a
loud crash following immediately after, we must be able to
characterise these occurrences and frame a geneialization
‘
lightning IS followed l)y thunder,’ in ordei that induc-
tion may have a beginning This most primitive of all

inductive steps can he described as the direct peiception

of the universal in the particulai A generalization relevant
to particulars must be framed if it is to he tested, and this
primary relevance of a generalization to particulars cannot
be manufactured from pirticulars as mere ihiscs and thats
There is no process by which this relevance can be inferred
It can only be directly apprehended ” *
To this first stage of
the inductive procedure, Mr Eaton adds a second, in which
isolated generalisations are made more probable by the
elimination of irrelevance and by fresh evidence, and a
third, in which generalisations reinforce one another by
entering into logically organised systems -
It is to be
observed, however, that a geneialisation is framed at the
very first stage, and that the second and third stages only
help us to test and confirm it. Hence so far as the know-
ledge of the general proposition is concerned, we are to say
that it is given to us by way of a direct perception of the
universal in the particular.

2. The question of pelitio principii m mferenee

As we have already seen,' every inference involves the

knowledge of vydpti or a universal relation between the
major and the middle term Without a universal relation

^ R M Eaton, General Loqtc, p 496

2 Op cif , p 501
 GROUNDS OF INFERENCE 275

between the two, no valid conclusion can be drawn from the

premises. It is when we know that smoke is univer-
only
sally related to fire that we can conclusively prove the exist-

ence of fire in a lull in which we see smoke. Otherwise,

the inference will be inconclusive and invalid. On the
other hand, it would seem we know smoke to be
that if

universally related to fire, we already know the smoke in

the hill to be related to fire. The truth of the universal
proposition ‘
all cases of smoke are cases of fire ’
involves,
nay, depends on the truth of the proposition ‘
this case of
smoke is a case ot fire.’ Thus it would seem that the
major premise 'ot an inference, which is a universal
proposition, assumes what we want to prove in the con-
clusion, i c. an infetenee involves the fallacy of petitio

pnncipii or begging the question.

The above dilemma of inference has been anticipated
and solved in Indian philosophy. The solution is generally
based on the distinction between the knowledge of the uni-
versal and that ot the particulars coming undei it When
we know that smoke is always related to fire, we know them
in then general character as two universals This does not
imply that we know the relation between all particular
smokes and fires Thus the Mimamsakas^ argue that the
knowledge of vydpti or a universal relation between smoke
and fire does not necessarily involve any knowledge of a
particular instance of fire, e.g the fire in a hill. When
we know the universal proposition ‘
all cases of smoke
are cases of fire,’ we do not know anything about the hill,

far less, about its relation to fire. If that were not so, or,

if we knew anything about the fiie in the hill, there could

be no necessity for the perception of smoke in the hill, in

ordei to know the existence of the fire in it. Hence it

follows that the conclusion of the inference, namely, ‘that

1 Vide SU , pp. 62-03.

 276 NYiYA THEORY OF KNOWLEDGE

hill is fiery,’ is new knowledge which is not involved

a
in the knowledge of its premises. The Naiyayika view
of vyapti as covering all the individual cases of a
relation seems to commit inference to the fallacy of
vetitio vrincAmi . Thus it has 1 een held by the Naiyayika
that when vse know the vyapti or the universal relation

between smoke and fire, we know all the individual cases of

smoke to be related to fire. we cannot account Otherwise,
for the inference of fire from the smoke in a hill. If we do

not know that the hill-smoke is related to fire, we could not

possibly pass from the one to the other. But then the diffi-
culty IS that if we already know the hill-smoke to be related
to fire, there is no room for an inference to arrive at a new
truth The conclusion of such an inference will only repeat
what is already stated in the premises. This difficulty in

the Nyaya may however be explained

view of inference
According to the Naiyayika, to know that smoke is univer-
sally related to fire is indeed to know that all cases of smoke ‘

are cases of fire.’ But the knowledge we have of all fires

and smokes is mediated by the knowledge of the universals

‘
fireness ’
and ‘
smokeness ’
(samanyalaksanapratydsatti)
This means that we know all fires and smokes in so far as

they participate in ‘fireness’ and ‘smokeness,’ t.e. ic

their general character without any reference to their speci-

fic characters. So while the vyapti gives us a knowledge of

the relation between smoke and fire in general, an infer-

ence based on it gives us the knowledge of the relation of

fire to a particular object, namely, the smoky hill. The.
major premise of the inference ‘
all cases of smoke are cases
of fire ’
does not by itself lead to the conclusion that there is

fire m the hill. It is only when the major premise is com-

bined with the minor, ‘
there is smoke in the hill,’ that we
draw the conclusion ‘
there is fire in the hill.’ This shows
that the truth of the conclusion is not epistemically involved
in that of the majoi premise or the universal proposition.
 GROUNDS OF INFERENCE 277

Hence we are to conclude that inference is neither inconclu-

sive nor a petitio principii, since it gives us a new know-
^
ledge.

3. The psychological ground of inference (paTcsata)

Just as inference depends on the knowledge of lyapti or

a universal relation between the noiddle and major terms, so
it depends on the relation of the middle term with the
minor term. ^ In inference the minor term becomes related
to the major through its relation to the middle term Every
inference proceeds with regard to some object about which
we want to establish something on the ground of a vyapti or

a universal proposition Hence the minoi term is as much

necessary for infeience as the middle teim The minor term
being called paksa in Indian logic, paksatd is treated as a

necessary condition of inference If there is to be any in-

ference, there must be a paksa or a minor term Hence the
question is Under what conditions do we get the minor
term of an inference^ Or, under what conditions do we
draw inference with regard to anything? While the vali -

dity of inference depends on its_pQssibility depend s

onjiaksjM^ Inference takes place wdien there is a paksa or
subject of inference, it becomes valid when based on vyapti
or a universal relation between the middle and the major
term. Hence while vyapti is the logical ground of infer-

ence, paksato is its psychological ground or condition.

From the fact that the minor term is an object about
which we want to infer something, it will appear that the
two obvious conditions of a mino r term are the absence of
certainty about something {s-iddhyahhava'^ and the will to

1 TB p 11 TO , II, pp 290-91
, ,

2 Anumfinnsya dvf ai^ge vyaptih paksadharmata ca, etc TB , p 11 Vyapyasy

parvatadiv|ttiivam pak$adbartnaia> T8 , p 46
 278 NYAYA THEORY OF KNOWLEDGE

infer it (stsMhapsa) The old Naiyayikas' and the

Vedantists ^ accept both of these conditions when they say
that paksata consists in the presence of doubt about the
sddhya or the major term {sadhyasamdeha) We have a
paksa or a minor terra when we are in doubt whether a
certain subject is related to the sddfij/a or the major term.
Now doubt implies not only the absence ol certain know-
ledge about sometliing but also a positive desire or will to

know It Hence doubt as a condition of mteiencc involves

both the absence of certainty about something and the desire
to have certain knowledge about that thing
The modern Naiyayikas take exception to the above view
oipaksitd. According to them, neither the absence of cer-
tainty nor the will to infer is a necessary condition of infer-
ence There may be inference even in the presence of cer-
tainty. A logician may, it he so will, infer the existence of
an elephant from its trumpeting voice even when he has
perceived it and so acquired certain knowledge about it Or,
a man may infer the exi-itence of the self even when he has
acquired certain knowledge about it from the scriptures.
Again, there may be inference even when there is no will to

infer, as when one involuntarily infers the existence of

clouds from the roar of thunder This case shows also that
the presence of doubt is not an essential condition of infer-

ence, since there is in it no previous doubt as to the exist-

ence of clouds in the sky Thus we see that inference takes

place under the following coiuhAions (a) when there are

absence of certainty and presence of tfie will to infer ,
(b)

when there is absence of both certainty and the will to infer;

(c) when there is presence of both certainty and the will to

infer. But no inference takes place when there are pre-

^ Na mri^ibe’rthe nyaya^? pravartate kintu sam^ayite, NB ,

1 IJ Samdigdha*
sadhyadbarmft dhatml pak^ah, TB p 11
,

^ Pak^atvarn tii aadhyasaradehavattvam afidhyagocarafladhakamanabhavavattvara

vl, Advatfasuldht, p 29
 GROUNDS OF INFERENCE 279

seDce of certainty and absence of the will to infer. Hence

to combine the first three cases and exclude only the last, we

are to say that inference takes place in all cases excepting

that in which there are presence of certainty and absence
of the will to infer This is expressed by the modern
Naiyayikas by saying that paksata consists in the absence
of that condition in which theie are the presence of
'
certainty and absence of the will to infer
The conditions of valid inference have of late been dis-
cussed by some Western logicians. All of them, however,

do not sufficiently realise the importance of the psycho-

ingical condition of inference, which Indian logicians discuss
so thoroughly under the theory of paksata Russell seems
to think that all that is necessaiy for inference is the logical

condition of a relation of implication between piopositions

According to him, the psychological element, namely, our
knoioledge of the propositions and their relation, is not a
necessary condition of inference. Thus he says ‘
It is

plain that wdiere we validly infer one proposition

from another, we do so in virtue of a relation
which holds between the two propositions whether
we perceive it oi not . the mind, in fact, is as purely receptive
in inference as common sense supposes it to be in perception

of sensible objects ^ Some ’

other Western logicians like

Ml Johnson and Dr. Stebbing ^ have recognised the im-

portance oi both the logical and psychological conditions of

inference. According to them, there are two kinds ol condi-

tions for any valid inference. The first kind of conditions

refers to the propositions and the relations that hold between
them These conditions are said to be independent of the

1 Sisadhayisaviraha-visistaBiddhyabhavalj paksata Yatra siddhirnasti tatra

Bi?adliayisayam satyamasatyainapi paksata Yatra sisadhayisasti tatra siddhau

flatyamaaatyamapi paksata Yatra siddhjrasci sisadhayisa ca nasti tatra na paksata, etc ,

m , pp 809 10. Vtde also TM .

Ch II , TC . II, pp 407-32
^ Russell, Principles of MaihemaitcSt p 33
3 Stebbmg, Modern Introduction to Logic pp 216-16,
 280 NYAYA THEORY OF KNOWLEDGE

thinker and are called by Mr. Johnson the constitutive

’
conditions .
’
In order that the proposition q may be lor-

mally inferred from p, it is necessary that p should logically

imply q and also that p should be true. The other kind of
conditions refers to the relation of the propositions to what
the thinker may happen to know. Since in inference a
thinker passes from something known to something inferred,
it follows that the pro})ositions and their relations must be
known by us. It follows also that what is inferred must
not be already known as true or false In order that q may
be validly inferred from p, it is necessary that p must be
known to be true, and also that p must be known to imply q
without its being known that q is true. These conditions are
dependent upon the relation of the thinker to the proposi-
tions involved in inference, and are called “ the epistemic
condition s ” of inference.
It would appear from the above that there is a consensus
of opinion among logicians, both Indian and Western, that
a valid inference must satisfy at least two conditions,
namely, that there must be a true proposition and that it

must imply another proposition. There is, however, some

difference of opinion among them as to how these conditions
condition inference. While a realist like Kussell seems to

think that they condition inference even w'hen they are not
known, Indian logicians maintain that they can condition
inference only when they are known by us. According to
them, while perception may be said to be conditioned by
the existence ol the sense organs, inference is conditioned,
not by the mere fact, but by the knowledge of something as
a sign and that of its invariable relation to something tdse,

although the reality of these things and their relation is inde-

pendent of our mind. ’
These two views seem to be reconciled

1 8a (v} apti^) ca aattaya cak^uradivaDnafipabhavam bhajatc kmtu j0a*^atay&,

80rvadarhna$afhgrahat Ch. I Cf also BP , 66 ,
VP , Ch. II,
 GROUNDS OF INFERENCE 281

by Mr. Johnson who holds that for inference there must not
only be a true proposition and a relation of implication
between propositions, but that these must be known by the
thinker who is inferring.
With regard to what we have called the psychological
conditions ol inference, there is a sharp difference of opinion
among logicians.The question here is Under what condi- :

tions does inference take place ? The answer given to this

question by the old Naiyayikas and the Vedantins is that
inference takes place when there is a doubt about what is

to be inferred. This is perhaps the most plausible view

that would be readily acceptedby common sense. No man
takes the trouble to infer or prove anything unless he is
in doubt about it. This view, however, is contradicted by
the inference of clouds from the sudden roar of thunder,

since it is not preceded by any doubt in the mind of the

thinker who infers.

But then it may be said that want of certainty, if not a

positive state of doubt, is the essential condition of infer-

ence In the Advaitasiddhi this view is accepted as an

alternative to the first given above, ivhen it says that

paksotd consists in the absence of proof relating to what is

to be inferred.^ Among Western logicians. Dr. Stebbing
also supports this view when she says :
“ Since inference

is a process in which a thinker passes from something

known to something injctfcd, it is clear that we would not say

we had infcTTcd q if we had already asserted q It is,

obvious that must not be kTiowti to be true,

therefore, q
^
and equally obvious that q must not be known to be false.”
There is a strong presumption in favour of this view. Infer-
us certain
ence as a source of knowledge aims at giving
knowledge about things. So it is obvious that if we want to

1 8aclhyagocBragildhakan)Sn5bh6v»vsMv»m va, Advattattddht, p. 29,

* Modem Introduction to Logic, p. 316.

36— (1117B)
282 NYiYA THEORY OF KNOWLEDGE

know anything by inference, it is because we lack certain

knowledge about it.

Now let us consider if the second view can explain all

the cases of inference mentioned by the modern Naiyayikas.

There seems to be no difficulty so far as the first two cases
are concerned. In the first case (a), we have inference when
there is the absence of certainty together with the will to
infer, e.g. the inference of future ram from the appearance
of dark clouds in the sky In the second case (h), we have
inference when there is the absence of both certainty and
the will to infer, e g the inference of clouds from the
roar of thunder While there is the absence of certainty
in both these cases, the will to infer is absent in the second.
This seems to suggest that the absence of certainty is the
essential condition, and the will to infer oni} an accidental
condition of inference. But when we come to the third

case, we are confronted by an exception to the rule that

every inference is conditioned by the absence of certainty.
Thus in case (c), we have inference when there is certainty
together with the will to infer If this be so, we have to

reject the view that the absence ot certainty is an essential

condition of inference and recognise the importance ot the
will to infer as a condition of inference. But the question is

Is there really any case in which inference takes place in

spite of certainty and in virtue of the will to infer ? The
examples cited by the Naiyayikas are rather doubtful cases.
Thus it m;iy be said that it a logician infers the existence
of an elephant perceived by him, it must be because he has
some doubt, however slight, about the truth of liis percep-
tion. Similarly, we may say that when a person infers the
existence of the self known by him through the scriptures,
it must be because he is not absolutely sure of the truth of
his scriptural knowledge. But there are certain cases of
inference which may be taken as crucial instances. The
path described by a falling body may be deduced by a physi-
 GROUNDS OF INFERENCE 283

cist from certain laws of motion, even when he sees it and

has no doubt about the reality of what he sees. ‘
We might
prove, to a person who doubted the correctness of our memory,
that it rained yesterday, by pointing to other facts with
which rain is necessarily connected.’ A lawyer may produce
evidences to prove a case of which he has a personal know-
ledge. Some theorems of Geometry prove what is otherwise
obvious or clearly perceived. At least, the geometrician
who proves them has no doubt about their truth. It is true

that 111 some of these cases there is some doubt m the mind
of the person or persons for w'liom these inferences are made.
But we must frankly admit that there is no doubt in the
mind of the person who makes the inference. It cannot be
said that the presence of doubt in one mind conditions the
process of inference occurring in a different mind. Hence
we are to admit that there may be inference in the face of
certainty, only if we have the will to inter It may, of
course, be asked here •
What does the will to infer aim at

in such a case ^ To this we reply that it aims at demon-

strating a known fact by showing its necessary connection
with other facts It cannot be said that the demonstrative
knowdedge of the fact being absent befoie, the inference is

really conditioned by the absence of ceitaiiity. So fai as

the know'Iedge of the fact is concerned, its demonstration

adds nothing to the certainty with which it was otherwise
known before Nor can we say that what the demonstrative
inference proves is not that there is such-and-such a fact,

but that such-and-such a fact follou s from certain other

facts. That a fact follows from other facts is no part of
the conclusion of an inference, but a part of its grounds or
premises. Hence we are to say that the conclusion of the
demonstrative inference states the same fact that was previ-
ously known by perception or memory, only it arrives at the

fact by way of inference. And, as Prof. Creighton says : “It

is not necessary for inference that the conclusion reached
284 NYiyA THEORY OF KNOWLEDGE

should be a fact which was not hitherto known.” ^

So we
conclude that the modern Naiyayikas are justified when they
emphasise the function of will in inference, and define
paksatd as the absence of the condition in which there is

certainty, but no will to infer.

4. Lingapardmarh as the immediate cause of inference

As we have already seen, eveiy inference must involve

at least three steps. There is first the knowledge of the
middle term as related to the minor term (lingajftdna).

Secondly, there is the knowledge of a universal relation

between the middle and the major terms (vydptijndna)

Lastly, there is the conclusion, in which the major term is

predicated of the minor term. The conclusion is the result

of inference as a process of reasoning Now the question
is : What is the special cause (karana) that brings about the
conclusion? Is it the knowledge of the linga or the middle
term ? Or, is it the knowledge of vydpti or the universal
relation between the middle and major terras ?
According to the older logicians, including the Buddhists,
the Jainas and some Naiyayikas, it is the knowledge of the
linga or the middle terra that leads to the conclusion.^

Hence the middle term, known as such, is to be taken as the

karana or operative cause of inference For the Mimamsakas
and the Vedantists, the knowledge of vydpti or the universal
relation between the middle and major terms is the karana
or special cause of The knowledge of vydpti is
inference.
revived in our mind when we see the linga or the middle
term as related to the paksa or the minor term, and this
leads to the conclusion. On this view an inference will

’ Introductory Logic, p 432

* F;rfeSM,p286
 dROtJNDS OP INFERENCE m
include the following steps •
(1) the knowledge of vyapti

or the universal relation between the middle and major

terras, e.g. all cases of smoke are cases of fire ; (*2) the
perception of the minor term as qualified by the middle
term, e.g. the hill is a case of smoke; (3) a revival of the
impression of vydph previously acquired, viithout any
necessary recollection of it; and (4) the conclusion that the
minor term is related to the major term, e.g. the bill is a
case of fire.

According to the modern Naiyayikas, the linga or the

middle term cannot be the karana or operative cause of
inference. The middle term may be a thing of the present
or the past or the future. But it cannot function in an
inference when past or future The middle term cannot
lead to the conclusion except through the knowledge of
or the universal i elation between it and the major
term. Hence the knowledge of vyaph or the universal
relation between the middle and major terms should be
taken as the special cause {karana) of inference. But the

knowledge of vydpU does not immediately lead to the

conclusion. It has for its function (vydpdra) a synthetic

view of the middle term as related to the major, on the one
^
hand, and the minor, on the other (bngapardmar^a). I his

is called trtiyabngapardmarh, i.e. a consideration of the

middle term for the third time. The middle term, e.g.

smoke, is known first when we acquire the knowledge of its

invariable relation with the major term ‘

fire ’
in the kitchen,
etc. It is known for the second time in relation to the
minor term, e.g. the hill. It is considered for the third

^ Vide Siddh6nfamuktdvaH with DtnakarJ, pp 288 89 Cf Veddnta^Panhhdfdt

Chapter, II '*
evaih ca *
eyam dhflmavan’ iti paksadharmatftjfi&ne ‘
dhflmo vahnivylpya*
ityaDubhaTftbitaBaihgk&rodbodfae ca sati *
vahnim&n *
ityanuiQitirbhayati, na ta madbye
vyaptiamarai^aih iajjanyam *
vahmvjftpyadhumavknayam *
ity&di viie^ai^aviibi^ta*
**
Iftftnam
* Vyaparastu paramari5a|? karapaiii vyaptidhirbhavet, etc., BP. and SM., 66-67
 286 NYiYA THEORY OF KNOWLEDGE

time when we know it as that character of the minor term

which is universally related to the major, as when we say
‘
the hill is posses-^ed of such smoke as is always related tr

fire.’ ^
It is through such lingapardmarh or knowledge o
the middle term as universally related to the major and a
characterising the minor, that the knowledge of vydpti leads
to the conclusion. Hence while the knowledge of vydpti is
the special cause (karana) of inference, hngaparamarh is the
immediate cause {caramakdrana) of the conclusion ^ Some
modern Naiyayikas go further and say that lingapardmaHa
itself is the operative cause (karana) of inference. ”
On this

analysis inference involves the following steps ; (1) the

knowledge of t'l/dpft or the universal relation between the
middle and major terras, e.g all cases of smoke are cases
of fire ; (2) the perception of the minor term as related to

the middle, e g. the hill is a case of smoke; (3) a recollec-

tion of the vydpti between the middle and major terms; (4) a
Contemplation of the middle as correlating the major and
minor terms, eg. the hill is a case of smoke peivaded by
fire ; (5) the conclusion relating the minor term with the
''

major, e.g the hill is a case of fire

As it has been pointed out by Dr D M Datta, the

Nyaya view of hngapardmar&a as the immediate antecedent
of the conclusion agrees with Biadley’s analysis of inference.
According to Bradley, an inference is always an ideal con-

^ Mahtna^^adau drs|aote . dhumajfianam prathamam, parvatadau dbumam

vyapyatvena tatsmarapam dvitiyam, tataatatraiva vyapyatvena dhurajBya
)aramar6o vabnivyapyadhumavanayamityevamrupo jayata iti tftTyatvam, Tarkakau-
mudi, p 10
» Ftde TC., n, pp 521-61
3 Vtde Tarkasamgraba, p 60
< Yena purusepa rnabanasadaa dburae vabaivyaptirgrblta patot sa eva purusah

kvacitparvatadavaviocbiDiiamu’am dhumari kham pa^yati, tadanaotaram dhumo vahm*

vyapya ityevamrupam vyaptismarapam tasja bhavati pa^cacca vabnivyapyadhumavanaya-
miti i&aLam bbavati sa eva paramar^a ityucyate iadanantaratb parvato vahnijxi&Ditya-
numitirjayate, Siddb5ntamuktSvali, pp 284-86.
• Vtde The SiX Ways of Knowing, p. 207.
 GROUNDS OF INFERENCE 287

struction resulting in the perception of anew connection.

The premises of inference are the data, and the process of

inference consists in joining them into a whole by an ideal

construction. “ We must fasten them together, so that

they cease to be several and are one construction, one
individual whole. Thus instead of A-B, B-C we must have
A-B-C ” Take for example the inference ;
‘
Man is mortal
and Caesar is a man and therefore Caesar is mortal.’ In this
inference there is first a construction as Caesar-man-mortal,
and then by inspection we get Caesar-mortal.” ^
So also the

Naiyayikas hold Ihat without lingapardmarsa as a synthetic

correlation of the minor, middh; and major terms we can-
not explain the transition from the premises to the con-
clusion. If no such synthesis was required, then the
premises could, even in their isolation, lead to the conclu-
sion. ^ The Vedantists, however, contend that the two
premises, taken together, lead to the conclusion. When 1

see smoke in a hill and remember that smoke is always'

related to fire, I am at once led to the knowledge that there
IS fire in the hill. It is not necessary for me to stop and
say further, ‘
the smoke in the hill is a smoke pervaded by
fire.’ ® As we shall see it more fully later on, in the case

of inference for oneself we do not require anything more

than the major and minor premises to arrive at the
conclusion. There is a natural transition of thought from
these premises to the conclusion When, however, we
are to demonstrate the truth of the conclusion to other

persons, we must state the identity of the middle terra

occurring in the two premises and exhibit it in a third

premise which relates the same middle to the minor and

major terms. Here we assert that the same M which

^ Bradley, Pnncvples of Logtc, Vol. I, p. 269.

* SM.,68
5 VP„ Ch TI
 NYiYA THEORY OF KNOWLEDGE

is always related to P is present in S. Hence we conclude^

that lingapardmark is not an essential condition of all

inferences, although it serves to make an inference most

cogent and convincing.

288
 CHAPTER XIII

CLASRIPTCATION AND LOGICAL FORMS OP

INFERENCE

1. Svartha and Parartha inferences

In Indian logic an inference is a combined deductive-

inductive reasoning consisting of at least three categorical
propositions. All inferences are thus pure syllogisms oJ

the categorical type which are at once formally and mate-

rially valid. Hence we have not a classification of mterences
into deductive and inductive, immediate and mediate,
syllogistic and non-syllogistic, pure and mixed. The
Naiyayikas give us three different classifications of inference.
According to the first, inference is of two kinds, namely,
svartha and parartha. This is a psychological classifica-

tion which has in view the use or puipose which an in-

ference serves According to another classification,

inference is said to be of three kinds, namely, purvavat,

iesavat and sdmdnyatodrsta. This classification has
reference to the nature of the vydph or the universal

relation between the middle and major terms of inference.

Purvavat and §esavat inferences are based on causal uni-
formity, while sdmdnyatodrsta is based on non-causal
uniformity. According to a third classification, inference

IS distinguished into kevaldnvayi, kevala-vyaiireki and

anvaya-vyalireki. This classification is more logical inas-

much as it depends on the nature of the induction by

which we get the knowledge of vydpti or the universal

37— (1117B)
290 NliYA THEORY OF KNOWLEDGE

proposition involved in inference. These different kinds

of inference we shall have to consider one after
another.
All inferences must have one of two ends in view.

They are meant either for the acquisition of some new

knowledge on our part or for the demonstration of a known
truth to others. Accordingly, all inferences are classed
under the two heads ot svartha or inference for oneself and
pardrtha or inference for others. '
An inference is called

svartha when it aims at the knowledge of an unperceived

object on the part of a man who employs that infeience.

In this kind of inference a man seeks only to reach the

conclusion for himself by relating it to the major and minor
premises. This is illustrated in the case ol a man who
infers the existence of fire in a hill because he first per-
ceives a mass of smoke in it and then remembers that there
IS a universal relation between smoke and fire. On the
other hand, an inference is parartha when it aims at

demonstrating the truth of the conclusion to other people.

In this inference there is a justification of the conclusion

through a justification of the middle term that leads to it.

It IS here specifically pointed out that the same middle term

which IS universally related to the majoi is also pre&ent in
the minor term. The conclusion is thus found to follow
necessarily from a synthesis of the major and minor
premises. This synthesis is embodied in a third premise
which relates the minor, middle and major terms of the
inference. A pardrtha anumdna is illustrated when a man
having inferred the existence of fire m a hill lays it down
as a thesis and proves it as a conclusion following from the
major and minor premises and their combination into a
^
third premise.

' TaccanuBi&Dam aviviShcim, s\irtham paiartliam cetj, TB p , 9,

* TP , pp 46-49.
 CLASSIFICATION OF INFERENCE 291

2. Purvavat, 3esavat and Samanyatodrsta inferences

In the Nyaya-Sutra^ inference is distinguished into

three kinds, namely, purvavat, ^esavat and samanyatodrsta.
There are different views with regard to the nature of these
inferences. According to one view, a purvavat inference is

that in which we infer the unperceived effect from a per-

ceived cause. Here the hnga or the middle term is related
to the sadhya or the major terra as its cause and is, there-
fore, antecedent to it. In this inference we pass from the
knowledge of the .antecedent cause to that of the consequent
effect. This is illustrated when from the presence of dark
heavy clouds in the sky we infer that there will be rainfall.
A ^esavat inference is that in which we infer the unperceived

cause from a perceived effect. Here the middle term

^
is

related as an effect to the major term and is, therefore,

consequent to it In this inference we pass from the
knowledge of the effect-phenomenon to that of the antecedent
causal phenomenon. This is illustrated in the inference of

previous rain from the rise of the water in the river and
its swift muddy current It will be observed here that in
both purvavat and ^esavat inferences the vyaph or the
universal relation between the major and middle terms is a
uniform relation of causality between them. These in-

ferences thus depend on scientific inductions. In samdnyato-

drsta inference, however, the vyaph or the universal
relation between the major and middle terms does not depend

on a causal uniformity. The middle term of the inference

IS related to the major term neither as a cause nor as an
effect. We infer the one from the other, not because they
are causally connected, but because they are uniformly
related to each other in our experience. This is illustrated,

when one infers that the sun moves because, like other

> Vide NS , 1. 1. 6.
 NYiYA THEORY OF KNOWLEDGE

moving objects^ its position changes, or, when we argue

that a thing must have some attributes because it is like a
substance. Here the inference depends not on a causal
connection, but on certain observed points of similarity
between different objects of experience. So it is more akin
to an analogical argument than to syllogistic inference.
According to a second interpretation, a purvavat inference
is that which is based on previous experience. If two
things have always been found to be related in the past,
then from the perception of the one we infer the existence
of the other, as when we infer fire from smoke. Similarly,
a desavat inference is taken to mean inference by elimina-
tion, which the inferred character is the residuum of a
in
process of elimination which excludes other characters.
This IS illustrated when one argues that sound must be a
quality because it cannot be a substance or an activity or a
relation and so on. So also idmanyatodrsta inference is
explained as that in which we do not perceive the relation
between the major and middle terms, but find the middle
to be similar to objects which are related to the major
term. This is illustrated when one argues that the soul-
substance exists because the quality of consciousness must,
like other qualities, inhere in a substance ’ According to a
third view * these three kinds of inferences may be taken to
mean kevalanvayt, kevala-vyatireki and anvaya-vyahreki
inferences which we are to consider next.

3. Kevaldnvayi, Kevala-vyatirekt and Anvaya-vyatireki

inferences

In view of the different methods of establishing vydpii

or a universal relation between the major and middle terms,
inferences have been classified into the kevalanvayi, the

1 Vtde NB., 1 1 5
* Vide NV. & NSV ,116.
 CLASSIFICATION OF INFERENCE 293

kevala-vyatireht and the anvaya-vyaUreki. An inference is

called kevalanmyi when it is based on a middle terra which

IS always positively related to the major term. Here the
knowledge of vydph between the middle and major terms is

arrived at only through the method of agreement in presence

(anvaya), since there is no negative instance of their agree-
ment in absence. ^
This is illustrated in the following
inference

All knowable objects are nameable;

The pot is a knowable object;
Therefore the pot is nameable.

In this inference the major premise is a universal

’

affirmative proposition in which the predicate ‘

nameable
is affirmed of all knowable objects. This universal proposi-
tion IS arrived at by simple enumeration of the positive
instances of agreement in presence between the knowable
and the nameable Corresponding to this universal affirma-
tive proposition we cannot have a real universal negative
proposition like ‘
No unnameable object is knowable,’ for
we cannot point to or name anything that is unnameable. *
The minor premise and the conclusion of this inference
are also universal affirmative propositions and cannot be
otherw'ise. Hence with regard to its logical form the
kcvaldnvayi inference is a syllogism of the first mood of
the first figure, technically called Barbara.
A kevala-vyatireki inference is that in which the middle
term is negatively related to the major term. It depends
on a vydpti or a universal relation between the absence of
the major term and that of the middle term. Accordingly,
the knowledge of vydpU is here arrived at only through the
method of agreement in absence (vyattreka), since there is

* Yatra sailhyavyatireko na kutrSpynati sa kevalSnvayi, TM ,

Ch. IT.
* TB., p 10.
294 NYiYA THEORY OF KNOWLEDGE

no positive instance of agreement in presence between the

middle and major terms excepting the minor term. *
This
may be illustrated by the following inferences :

(1) No non-soul is animate

All living beings are animate ;

Therefore all living beings have souls.

(2) What is not different from the other elements has

no smell ;

The earth has smell ;

Therefore the earth is different from the other

®
elements.

Symbolically put the inferences stand thus

No not-P IS M ;

S is M ;

Therefore S is P.

In the second inference above, it will be seen, the middle

term ‘
smell ’
is the differentia of the minor term ‘
earth.’

An inference which is thus based on the differentia (laksana)

as the middle term is also called kevala-vyatireki In it the
minor term is co-extensive with the middle. Hence we
have no positive instance ol the coexistence of the middle
^
with any term but the minor. So there can be vydpti or
a universal relation only between the absence of the middle
and the absence of the major term. We cannot point to
any positive instance of agreement in presence between the
major and middle terms, except those covered by the
minor term. Hence the major premise is a universal
negative proposition arrived at by simple enumeration of

^ Kevalavjatireki tvasatsapak^K), etc , TC , II, pp 682 f. Vyatirekavyaptau lu

sSdhySbhSvo vySpyah hetvabhavo vyapaks^, TM , Ch II.

2 r»deTB.,plO
3 VideTS p 62. ,

< Lak^a^amapi kevalavyalireki heta^i, TB., p. 10

^ Yatra sadbyapiasiddbib pak^atirikte D&sti, sa kevalavyalireki, TM., Ch. II.
 CLASSIFICATION OF INFEBENCF 295

negative instances of agreement in absence between the

major and middle terms. ^
The minor premise is a universal
affirmative proposition. But although one of the premises
is negative, the conclusion is affirmative, which is against
the general syllogistic rules of Formal Logic. Hence we
see that kevala-vyatireki inference is not any of the valid
moods of syllogism recognised by Formal Logic. The
validity of such inferences, however, has been admitted by
^
Bradley as a special case of negative reasoning
An inference is called anvaya-vyatirekt when its middle
term is both positively and negatively related to the major
term. ^ In it there is vydpti or a universal relation between
the presence of the middle and the presence of the major
term as well as between the absence of the major and the
absence of the middle term. The knowledge of the Dydpti

or the universal proposition, on which the inference

depends, is arrived at through the joint method of agreement
111 presence and in absence {anvaya and vyatireka). The
iiydpii or the universal proposition is affirmative (anvayi)
when it is the result of an enumeration of positive instances
of agreement in presence between the middle and major
terms. It is negative (oyatireki) when it is based on the
simple enumeration of negative instances of agreement in

absence between the middle and major terras.^ The

difference between the universal affirmative and universal
negative propositions (anvaya-vyapH and vyatireka-vydpti)
is that the subject of the affirmative proposition becomes
the predicate, and the contradictory of the predicate of the
affirmative proposition becomes the subject in the corre-

sponding negative proposition ®

Hence an anvaya-vyatireki

' Vide TB , p 10
2 Vide Bradley, Principles of Logw^ Vo’ I, pp 274-82
3 Vatra sadhyam sadhyabhava^ca aoyatra prasiddhab so ’nvayavyatireki, etc.»

TM., Ch. II
* Sa oftnvayavyatireki, anvayena vyatirekena ca vyiptinaattvSt, etc , TB.| p 9,

5 Ihide
296 NYiYA THEORY OF KNOWLEDGE

inference may be based on either a universal affirmative or

a universal negative proposition as its major premise. It

is illustrated in the following pair of inferences :

(1) All cases of smoke are cases of fire ;

The hill is a case of smoke ;

Therefore the hill is a case of fire.

(2) No case of not-fire is a case of smoke ;

The hill IS a case of smoko ;

Therefore the hill is a case of fire.

The Vedantists do not recognise the above classification

of inference into kevalmvayi, kevala-vyatireki and anvaya-
vyatireki. According to them, inference is logically of one
kind, namely, anvayt. An inference must be based on
vyapti or the universal relation between the middle and
major terms. The knowledge of vyapti is arrived at
through the observation of agreement in presence between
the middle and major terms with the non -observation of an\
contrary instance. Hence for the Vedantists, both the
premises as also the conclusion of an inference must be
universal affirmative propositions. That is, all inferences
must be in the technical form of Barbara. But there
cannot be any kevalanvayi m the sense of an inference in
which the major term is a character that is not anywhere
non existent. In kevalanvayi the middle term is only
positively related to the major terra, since there is no case
of their absence. This, however, is not true. The whole
system of finite categories being transcended and negated
in Brahman or the Absolute, we cannot have any term
which IS never non-existent As for the Nyaya view of
vyatireki, the Vedantists contend that as a reasoning based
on a universal negative proposition it is not to be regarded
as an inference, but as arthdpatti or postulation. An
inference is a knowledge of the major term through that of
 CLASSIFICATION OF INFERENCE 297

the middle term. This knowledge is based on the vyapti

or the universal relation between the presence ot the middle

and the presence of the major terra. one infers When

fire from smoke he depends on the knowledge of vyapti,

not between the absence of fire and the absence of smoke,

but between the presence of smoke and the presence of fire.
There being no such thing as vyatireki inference, we cannot
admit the possibility of anvaya-vyatirehi inference which is but
*
a synthesis of the anvayi and vyatireki forms of inference.

Accoiding to the Naiyayikas, however, aithapatti is not a

separate method, oi knowledge, but a form of inference. We
shall have to consider this question more fully hereafter.

4. The logical form of inference

All thesystems of Indian philosophy agree in holding

that the syllogism represents the typical form of an inferen-
tial reasoning In inference we arrive at a truth through
the medium some other truths
of Like the conclusion of a
syllogism, inferential knowledge is a deduction from certain
propositions. There is however some controversy among
the different systems as to the number of the constituent

parts or propositions entering into an inference (avayava).

According to some old Naiyayikas, there are ten members
or constituent parts of an inference. These are (1) jijMsd
or the desire to know the truth, (2) sam4aya or doubt about
the real nature of a thing, (3) Sakyaprapti or the capacity of
the pramanas to lead to true knowledge, (4) prayojana or the
purpose of making an inference, (5) samhya-vyuddsa or the
removal of all doubts about the truth of an inference, (6)

pratijhd or the first proposition, (7) hetu or the reason, (8)

udaharana or the example, (9) npanaya or the application of

®
the example, and (10) nigamana or the final conclusion.

1 Fide VP Ch
,
H
> NB ,
1. 1 32.

38— (1117B)
 298 NYiYA THEORY OF KNOWLEDGE

The above view of the syllogism as consisting of ten

parts or members idaSdoayava) has been criticised and
rejected by the later Naiyayikas, from Vatsyayana down-
wards. According to them, the first five factors, mentioned
above, are unnecessary for proving anything by means of an
inference. They represent not so much the logical steps in
drawing a conclusion as the psychological or epistemological
conditions involved in inference. Thus the desire to know
(]t]Msd) may be taken as a condition of all knowledge, by
which we want to realise some end. But such desire does
not prove anything to any person and cannot, therefore, be
regarded as a factor of inferential reasoning, Similarly,
doubt is the impetus to a desire to know the truth and is,

in this sense, a condition of knowledge. But to doubt is

not to prove anything The validity of all knowledge

depends on the validity of the methods of knowledge {hhya-
prapti). But the validity of the methods cannot be put
forward as a part of the argument to prove a conclusion.
So also the purpose or the end, which an inference serves, is
no part of the inference itself. The removal of doubt
{sathsaya-i yudaso) consists in repudiating all views which
contradict the conclusion of an inference. This serves to
lend indirect support to the conclusion, but does not really
prove it. Hence it has been held by the Naiyayikas that the
syllogism consists of the last five members mentioned above,
since they are all necessary for proving or demonstrating a
truth.' The Sarhkhya* and Vai4e§ika® systems also
accept this view of the syllogism as consisting of five
members or propositions. The five members of the
syllogism have been explained by the Naiyayikas as
follows.

* NS. * NB., 1. 1 32.

* Vide Sdrhkhya’Sutra^ 5 27
5 In Pra6aBtapSda*B Paddrthadharmasafhgraha (p 114) the five members of the
syllogism are called praitjM, apadeSa, nxdarima^ anusaihdhana
and pratydmndya
 CLASSIFICATION OF INFBRENCB 299

(1) The first member of the syllogism is called the

pratijM or the propositum. It is just a statement of one’s
position and consists in the assertion of some unperceived
quality or character in relation to some object of experience.
Tlie assertion mav be affirma t ive or negative. Hence in the
pratijM a certain predicate is either affirmed or denied of a
certain subject, e.g. ‘ the hill is fiery,’ or ‘sound is not
eternal.’ The pratijna includes a subject {paksa) and a
predicate (sddhya), but no copula or verb to relate the two,
e.g. ‘ parvato vahnimdn.' It thus corresponds to a proposition

without any copula. It is to be proved and established by

other propositions in the course of the inference. The
pratijM simply tells us what the subject of the inference is
*
and what we want to infer or prove with regard to it.

(2) The second member of the syllogism is called the

hetu or the reason. It consists in the statement of the

mark or the sign {linga) which being present in the subject
or the minor term suggests that the latter possesses a certain

pioperty predicated of it. It is the assertion of the middle

term by which we know that the paksa or the minor

term is or is not related to the sadhya or the major
term. It may thus be called the middle premise or the
middle proposition of the syllogism. But while the pratijm
IS a proposition of two terms, the hetu is a one-term
p roposition. Thus for the propositum ‘
the hill is fiery,’

the hetu or the middle proposition is ‘

dhumat,' i.e.
‘ ®
because of smoke.’
The
(3) third member of the syllogism is called

udahmana or the example It consists in the assertion of a

universal relation {vydpti) between the major and middle

terms with reference to some apposite instances. The hetu
or the middle term proves the presence or the absence of the

» NS & NB., 1. 1. 83.

* JbM., 1.1.84-86
300 NYiYA THEORY OK KNOWLEDGE

major in the minor only as it is connected with the minor,

on the one hand, and uniYersally related to the major, on
the other. Hence the universal relation between the major
and middle terms must be duly asserted as an essential
member of the syllogism. This assertion is a universal
proposition which may be either affirmative or negative.
It IS a universal affirmative proposition when it indicates
the agreement in presence between the major and middle
terms as supported by a positive instance, e.g.
‘
all cases of
smoke are cases of fire, to wit, the kitchen ’
It takes the
form of a universal negative proposition when it shows the
agreement m absence between the two, as supported by a
negative instance, e.g ‘no case of not-fire is a case of
^moke, to wit, the lake.’ *
The third member of the Nyaya
inference thus corresponds to the major premise of the
syllogisms in the hist figure. As a universal pioposition
supported by certain instances, the third member of the
syllogism is found to be an inductive generalisation based
on actual facts of observation. It thus shows that an
inference is both deductive and inductive, formally valid
and materially true. As Dr. Seal rightly observes :
‘
It
harmonises Mill’s view of the major premise as a brief
memorandum of like instances already observed, with the
Anstoteliani view of it as the universal proposition which
^
IS the formal ground of the inference.’
(4) The fourth member of the syllogism is called upanaya
or the application. It consists in the application of the
universal proposition with its example to the subject or the
minor term of the inference. While the third member of
the syllogism states the universal relation between the majoi
and middle terms, or between their absence, the fourth
member shows the presence or the absence of the middle

1 Ibid. ,1 1 36*37. Of V}apt3pradar6anavi^ayo NVT., p 304.

* The Positive Sciences of the Ancient Hmdus, p. 262.
 CLASSIFICATION OF INFERKNCB 301

in the minor terim^ It may thus be called the minor

premise of the syllogism, and may be a universal affirmative
or negative proposition. It is affirmative when it is the
application of an affirmative major premise with a positive
‘
instance, e.g. so, like the kitchen, the hill is smoky.’ It

IS a negative proposition when it is the application of a

negative major premise with a negative instance, e.g.
‘
unlike the lake, the hill is not not-smoky, i e. it is smoky.’
The fourth member of the syllogism is not merely a
repetition of the second or the middle proposition. It brings

out the identity between the middle term mentioned in the

second member and that which is stated to be universally
related to the major m the third member of the syllogism.
As such, it IS a synthesis of the second and third members
of the syllogism. It shows that the same middle which is

universally lelated to the majoi term is also present in the

minor term, and is, therefore, very useful for the purpose
oi pi oof.

(5) The fifth and the last member of the syllogism is

called mgamana oi the conclusion. Here the preceding

four steps are brought to a point so as to demonstrate the
truth of the first proposition, with which the inference
starts. It consists in the re-statement of the pratijM or the
propositum as proved by the major and minor premises,
‘
e.g. therefore the hill is fiery.’ It should not be supposed
that the conclusion is an unmeaning repetition of the first

proposition. What is at first put forward as a hypothesis

or bare assertion is asserted in the conclusion as a firmly
established truth. What appears in the first proposition as

a judgment to be proved, does indeed re-appear m the

conclusion, but as something proved and demonstrated by
other propositions.^

> NS. and NB, 11.88 ;TR,pp 1811,

» Jbtd. ; NS. and NB„ 1.1.39.
302 NYiYA THEORY OF KNOWLEDGE!

Tbe logical form of an inference or the syllogism,

according to the Naiyayikas, may be illustrated in tbe
following ways

(1) S is P;
s is M
M 18 P
S IS M
.-.s IS P.

Or,
The hill is fiery

Because it smokes
Whatever smokes is fiery, e.g. tbe kitchen ;

So the bill smokes;

Therefore it is fiery.

( 2) S is P;
V S 18 M ;

No not-P is M;
S is not not-M;
SisP.

Or, The hill is fiery ;

Because it smokes
Whatever is not-fiery does not smoke,
e.g. the lake :

Not so the hill (i.e. the hill does smoke)

Therefore tbe hill is fiery.

Tbe Mimamsakas and the Vedantists join issue with the

Naiyayikas on the question of the parts or members of a
syllogism. ‘
They agree with the Naiyayikas in holding that
the syllogism is necessary only for pararthmumSina or
demonstrative inference and that svdrthanumana or inference

I Vide SD.,p. 64; VP., Cbi^*. H.

 CLASSIFICATION OF INFERENCE 303

for oneself requires no verbal statement in the form of the

above syllogism. But they decline to accept the NySya
view that the syllogism consists of five members or proposi-
tions. According to them, a syllogism does not require
more than three members to carry conviction to anybody.
The two essential conditions of a valid inference are vydpti
or a universal relation between the major and middle
terms and paksadharmata or the presence of the middle term
in the minor. Hence the full force of a syllogism comes
out in the body of three affirmative propositions, two ol

which stand for the grounds of inference and one for th(

conclusion. These three propositions are either the pratijM.

hetu and udaharana, or the udaharam, upanaya anc
nigamana. * Hence we will have two forms of the syllogism

which may be put thus :

(1) SisP;
S IS M ;

All M IS P.

Or, The hill is fiery ;

Because it smokes
Whatever smokes is fiery, e.g. the kitchen.

(2) All M IS P ;

S isM ;

.*. SisP.

Or, Whatever smokes is fiery, e.g. the kitchen ;

The hill smokes ;

Therefore it is fiery.

^ The Buddhists go further than the Mimamsakas and the Vedftntists m reducing

the syllogism to two propositions only, nam6ly» the uddharcn^a and the upanwya, but
no mgamana or conclusion As Mr. Joseph and Dr Stebbuig also point out, we may
sometimes put an argument in the form of a smgle proposition as when we aay *
if

wishes were horses, beggars could ride.* (C/. Joseph’s Introduction to Logic, p. 852;
Stebbiug's Modem Introduction to Logic, p^ )
 304 NYiYA THEORY OF KNOWLEDGE

It will be observed here that in the first form of the

syllogism given above, the inference starts from the conclu-
sion, and then the premises are stated to justify it. In the
second form, the premises are given first and then the
conclusion is drawn from them That inferences may take
both forms has been recognised by some modern Western
logicians like H. W. B. Joseph,’ L. S. Stcbbing, ^ F M.
Chapman and Paul Henle.'* But it will be admitted by all

that while the second form (in which the premises come
first and the conclusion last) has a rigidly formal character,
it IS the first (in which the conclusion comes first and the
premises last) that is ordinarily when we
used by us
actually infer anything “In ordinary speech we more
often state the conclusion first and then state the premises.
This gives emphasis to the conclusion and also aids in

showing the direction of our aiguiuent,’’ ’

If this be so, then
we must say that flic Nja>a form of inference, in which the
conclusion-to-be-pioved comes first and the premises last, is

the natural or actual form of reasoning But it should be

remarked that the first proposition cannot be strictly called
the conclusion, since a conclusion is what follows from
certain grounds or premises. So it seems better to speak

of it, like the Naiyayika, as just a pratijM or probandum,

i.e. something to be proved This proposition stands out
as the conclusion when it is seen to follow logically from
certain other premises That there must be two such
premises, viz. the major and the minor, all logicians would
readily admit. And that there must also be a third premise
to synthesise these two seems to be admitted onlj by a
'''

logician like Bradley It will, however, be admitted by

1 Introduction to Logic ^ pp 255-56

2 Modern Introduction to Logic, pp 82, 84
3 Chapman and Henie, The Fimdamentah of Logic, p. 98
« Ibid
® The Principles of Logic, Vol. I, p. 259,
 CLASSIFICATION OF INFERENCE 306

others that there is no logical necessity for any thinker to

infer the existence of fire in a hill unless it is shown that

the smoke in it is just that real natural smoke which is

pervaded by fire. If the smoke in the hill be an illusion,

like the mirage, then we cannot conclude that there is fire

in the hill, although we may thinh that there is. So if

there is to be no gap in the chain of reasoning that is to

establish the conclusion, we are to have a third premise
to bring out the identity of the middle term in the preceding
two premises. These three premises together with the
conclusion and the probandum give us the five-membered
form of the Nyaya syllogism which, therefore, seems to be
both psychologically correct and more conclusive for demon-
stration.

39— (1117B)
 CHAPTER XIV
THE FALLACIES OP INFERENCE
1. Distinction between a valid and an invalid reason

In Indian logic the fallacies of inference are all materia

fallacies. So far as the logical forms of inference are con
cerned, there can be no fallacy, since they are the same fo
all valid inferences. An inference, therefore, becomes falla
1
cions by reason of its material conditions The Nyay;
account of the fallacies of inference is accordingly limitei
to those of its members or constituent propositions, an(

these have been finally reduced to those of the heiu or tin

reason.^ For the purpose of proof an inference is made t(

consist of five members, namely, pratijfia, hetu, udaharana

upanaya and nigamana As such, the validity of ai

inference depends on the validity of the pratipm and othei

constituent parts of it. If there is anything wrong witl
any of its members, the syllogism as a whole becomes
fallacious. Hence there will be as many fallacies o
inference as there are fallacies of its component parts
from the first proposition down to the conclusion. So w(
may speak of the fallacies of the pratijm, etc., as coming
under the fallacy of inference (nyayabhasa).^ But it mus
be admitted that the validity of an inference dependi

Cf The AristoteliaQ cla8si6catioD of fallacies into those m cftcfiotie and thoi

nnem This agrees with the NySya classification m excluding the formi
^distributed middle, illicit process, and so on, Civ m ibe list of fallacies.

2 j^nydyabhdsa has been used by Vatsyayana in a technical sense to mea

3 cijJ. as are opposed to preception and scriptural testimony (Vtde Nydyt

4 however, it is used in its literal sense to mean the fallacies <

B The
 FALLACIES OF INFERENCE 30^

ultimately on the validity of the hetu or the reason employed

in it. So also the members of a syllogism turn out to be
right or wrong according as they elaborate a right or wrong
reason. The fallacies of inference ultimately arise out of
the fallacious reason. So the Naiyayikas bring the fallacies
of inference under the fallacies of the reason (hetvdhhSsa)
and consider a separate treatment of the inferential fallacies

due to the propositum, example, etc. (pratijnabhdsa, drstantd-

hhdsa) as unnecessary and superfluous.^

Now the question is What is a fallacious middle
(hetu)? How are we to distinguish between a valid and
an invalid middle ? Literally speaking, hetvahhdsa or the
fallacious middle is one that appears as, but really is not,

a valid reason or middle term of an inference. It appears

as a valid ground of inference because it satisfies some of

the conditions of a valid middle term But on closer view
it IS found to be f.illacious because it does not fulfil all the
conditions of a valid ground of inference. ^
As we have seen
before, there are five conditions of the hetu or the middle
term of an inference. First, the middle term must be a
characteristic of the minor term ipaksadharmata) Secondly,
it must be distributively related to the major term, i.e.

the major must be present in all the instances in which the

middle is present (sapaksasattva) Thirdly, and as a
corollary of the second condition, the middle term must be
absent in all cases in which the major is absent {vtpaksd-
sattva). Fourthly, the middle term must not relate to
obviously contradictory and absurd objects like the coolness
of fire, etc. {abadhitavisayatva). Fifthly, it must not itself

be validly contradicted by some other ground or middle term

(asaipratipaksatta) Of these five conditions, the third

does not apply to the middle term of a kevalanvayi inference,

j Vide TR.. p. 1236 i NM., p. 672,

» Vide NV., 1.2.4.

 308 NYiYA THEORY OR KNOWLEDUlS

because it is such that no case of its absence or non-exist-

ence can be found. Hence, with regard to it we cannot
say that the middle term must be absent in all cases in
which the major is absent. Contrariwise, the second
condition does not apply to the middle term of a kevala-
vyattreki inference, since here the middle term is always
negatively related to the major term There is a universal
relation between the absence of the middle and that of the
major term. Of such a middle term we cannot s ly that
wherever it is present the major must be present. It is only

in the case of anvayavyatiicki inferences that the middle

term must satisfy all the five conditions Hence it has been
said that a valid middle teim is one that satisfies the five
or at least the four conditions as explained above '
As
contrasted with this an invalid middle term (hcivahhusa) is

that which violates one or other of the conditions of a valid

ground of inference (hetn). It may be employed as the
hetu or the middle term of an inference, but it fails to prove
the conclusion it is intended to prove. There are different
forms of the fallacious middle according to the different
circumstances under which it may arise. All fallacious
middle terms have been classified under the heads of the
savyahhicara, viruddha, prakaranasama or i>atprahpaksa,
sadhyasama or asiddha, kalatUa and hadhiia. ^
Kesava Misia
observes that the fallacies of definition such as aiivyupti or

1 VideNSV,1.2 4, NM.p no
* Id both the old and the modern schools of the Nyaya, the inferential fallacies
have been classified under five heads. The first four kinds of fallacies bear the same
naines or at least the same significance in both the schools The last kind of fallacy,
however, is not only callef by different names, but bears substantially different
meanings m the two schools It is in view of this fact that I have taken the two
names to stand for two kinds of fallacies of the middle term Hence we get six kinds
of fallacies in plac*e of the five enumerated in the Nyaya treatises {Vide Nydya-Siitra,
124; Tattvacmiamam, II, p 778 ) In the Prabhakara Mimamsa the fallacies of
the middle term are called the asddhdrana, hadhiia, sadharana and asiddha (vide Jha,
Prahhdkara Mimamsa, p 46) In Paddrthadharwasamfjraha the fallacies are called the
Oitddha, virvddhat sarhdtgdha and anadhyavasita
 THE FALLACIES OE INFERENCE 309

‘
the loo wide,’ avyapti or ‘
the too narrow ’
and asambhava
or ‘
the false ’
also come under the fallacies of the middle
^
term.
t

2. The fallacy of savyabhicara or the irregular middle

The fiist kind of inferential fallacy is called the

savyabhicara. In it the hetu or the middle term is found
to lead to no one single conclusion, but to different opposite
conclusions. This fallacy arises when the middle term
violates its second condition, namely, that it must be
distnbutively related to the major term. This condition
requires that the middle term must be pervaded by the
major term, or that the major must be present in all the
cases in which the middle is present The savyabhicara
hetu, however, is not uniformly concomitant with the major
term It is i elated to both the existence and the non-
existence oi tlie major term. It is therefore called anaikan-

tilm or an irregular concom itant of the sadhya or the major

tertm Hence from such a middle term we can infer both
the existence and the non-existence of the major term .
^ 01
such savyabhicara or irregular middle there are three kinds,
namely, the sadharana, asadharana and anupasamhan,
The sadharana or the ordinary fallacy of the irregulai

middle occurs when the middle term is in some cases related

to the major and in other cases related to the absence

of the major. “ This is illustrated in the following
syllogism .

All knowable objects are fiery ;

The hill is knowable

Therefore the hill is fiery.

1 Vide TB p 37.,

2 Anaikantikah savyablac&ra^i, NS , 1.2.6.

3 Sadharajjal? s5dhyavat tadaDyavfttilj, etc , NSV., 1.2,6.

310 NYiYA THEORY OF KNOWLEDGE

Here the middle term ‘

knowable ’
is indifferently related to
both fiery objects like the kitchen, and fireless objects like

the lake. Ail knowables being thus not fiery we cannot

conclude that a hill is fiery because it is knowable. Rather,
it is as much true to say that, for the same reason, the hill

is fireless.

The second form of the savyabhicara is called asadharana

or the extraordinary. It is a peculiar form of the fallacy

of the irregular middle. In it the middle terra is related

neither to things in which the major exists nor to those in

which it does not exist. ^ Hence from such a middle term

we can infer neither the existence nor the non-existence of
the major term. Or, such a middle term may be employed
to prove both the existence and the non-existence of the
major term. This is illustrated when one argues that sound

IS eternal because there is hhdatva or ‘

soundness ’
in it.

Here the middle term ‘

soundness ’
is related only to the

minor terra ‘sound.’ It is found neither in eternal objects

like the soul nor in other non-eternal things like the pot.
Hence we do not know if soundness is universally related to
the eternal or the non-eternal The middle term being
undistributed one way or the other cannot lead to any valid
conclusion.
The third form of the savyabhicara is the anupasamharl
or the indefinite. ^ Here the middle term is related to a

minor term that stands not for any definite individual or

class of individuals, but indefinitely for all objects. Hence
the distribution of the middle term cannot be proved either
positively or negatively. To prove that the middle term is
distributively related to the major we are to point out either
the positive instances of their agreement in presence or the
negative instances of their agreement in absence. Since,
however, the minor term stands for all possible objects, we
^ As&dbara^a^ sapak^iavipak^ayyftvj^tab etc., thtd,

I AnvajavyatirekadirftftQtarabito *napasazhharl, etc., TS., p. 66.

 THE FAIiLACIES OF INFERISNCB 811

cannot go beyond them and get any case in which the

middle coexists with the major, or the absence of the
major is concomitant with that of the middle term. This
is illustrated in the inference that ‘
all objects are eternal,
because they are knowable.’ The validity of this inference
depends on the validity of the major premise, namely, ‘
all

knowables are eternal.’ But the validity of the major

premise cannot be proved, since beyond all objects we have
no instances of the concomitance between the knowable and
the eternal.

3. The fallacy of viruddha or the contradictory mtddle

There are two different explanations of the fallacy of

viruddha According to the Nyaya-Sutra '
and Bhasya, the
fallacy ol the viruddha consists in the opposition of one
doctrine to a previously accepted doctrine, both belonging
to the same system of thought. It is n contradiction
between the different parts or doctrines of a system of
philosophy. As an example of this Vatsyayana cites two
contradictory statements from the Yoga-Bhasya, namely,
(i) that the world ceases from manifestation because it is

not eternal, and (m) that even then it exists because it

cannot be destroyed.^
In the above sense the viruddha as a fallacy means the
contradictions and inconsistencies involved in any school
of philosophy. As such, how'ever, it is not an inferential
fallacy, but the fallacy ot self-contradiction in which any
theory or philosophy may be involved. Hence the first

explanation of the iiruddha as given above does not appear

to me to be acceptable.

According to the later Naiyayikas, from Uddyotakara

downwards, the hetu or the reason is called viruddha when
I Siddli&Dtaoaabliyupeija tadvirodhl viruddhaj^, NB , 1.2.0
» Vide NB., 1.2.6,
 312 NYiYA THEORY OF KNOWLEDGE

it disproves the very proposition which it is meant to prove.

This happens when a middle term exists, not in the
objects m Avhich the major exists, but in those in which the
major does not exist That is, the vtrtiddha or the contra-
dictory middle is that which is pervaded by the absence of
the major term ^ The result is that such a middle term in-
stead of proving the existence of the major in the minor term,
which IS intended by it, proves its non-existence therein.
It contradicts and sublates the prnhjfid or the proposition
which it IS employed to prove and establish Thus if one
argues ‘
sound is eternal, because it is caused,’ we have
a fallacy of the viriiddha or the contradictory middle. The
middle term ‘
caused ’
does not prove theeternality of sound,
but its nou-eternality, because all that is caused is non-
eternal. Hence the distinction between the fallacies of the
savyabhicara and the inruddha is that while in the former the
middle term is universally related neither to the existence of
the major nor to its non-existence, in the latter the middle
term is universally related to the non-existence of the major

term. As a consequence of this, the sovyabhicaia or the

irregular middle only fails to prove the conclusion, whereas
the viruddha or the contradictory middle disproves "it or
proves the contradictory proposition.

4. The fallacy of prahaiamsama or the counteracted middle

The third inferential fallacy is called the prakaranasama.

Literally, it means a reason which is similar to the point at
issue iprakarana). We have a point at issue when there
are two opposite views with regard to the same subject, both
of which are equally possible, so that they only give rise to
a state of mental vacillation as to the truth of the matter.

^ PratijMhetvorva virodhal^ etc , NV., 1. 2 6

^ SadhjSbhavavyapto heturviruddhal? etc., TS., p. 67.
 THE FALLACIES OF INFBREKCE 313

Now when a middle term does not go further than producing

a state of mental oscillation between two opposite views we
have a case of the prakaranasama middle. This happens
when one reason seeks to prove the existence of the major
in the minor, but there appears some other reason to prove
the non-existence of the major, and both of them are found
to be equally strong Here the opposed reasons counteract
each other, but neither can suhlate the other They may
indeed be employed as the middle terms of an inference, but
(‘ach being neutralised or counterbalanced by the other
(satproltpnksito) fails to establish a sure conclusion and is

therefoie fallacious Hence the prakaranasama is also called

saipiatipaksn or that which is opposed by an equally strong
hefti or middle term. ^
This is illustrated in the following
arguments :
‘
sound is eternal, because the properties of the
non-eternal are not found in it ’
; and ^
sound is non-eternal,
because the properties of the eternal are not found in it.’

Here both the inferences are fallacious, because there is

nothing to distinguish between the two middle terms leading

to opposite conclusions.^ The two middle term^ being
counteracted by each other cannot lead to any definite con-
clusion and we are left with the same question with which
wc started, namely, whether sound is eternal or non-eternal.
The lallacy of the praharana}>ama is distinguished from that
of the saoyabhiedra by the fact that while in the latter one
and the same character of the minor is taken as a middle
term that may lead to opposite conclusions, in the former
two different characters of the minor are taken as the middle
teims leading to opposite conclusions. It is also distin-

guished from the inruddha or contradictory middle which by

1 Yaamat prakaraijacinta sa nirnayarthanoapadistab prakaranasumal.!, N8 1. 2. ,

27 Yasya pratipak?abhutam hetvantaram vidyate aa prakarapasamal}, sa eva sat*

pratipaksal^ etc.i TB , p. 30.
2 Vtde NB , 1. 2. 7.

40— (1117B)
314 NTiYA THEORY OF KNOWLEDGE

itself proves the opposite of what it is intended to prove,

while here the opposite conclusion is proved by a different
middle term {hetvantara)

5. The fallacy of asiddha or the unproved middle

The fourth kind of fallacy is called the sadhyasama or the

asiddha. The word sadhyasama means a middle term which
IS similar to the sadhya or the major term. The sadhya is a
character which we want to prove in relation to the paksa or
the minor term. Hence the sadhyasama stands for a middle
term which requires to be proved as much as the major term.
This means that the sadhyasama middle is not a proved or
an established fact, but an asiddha or unproved assump-
tion.^ The fallacy of the asidd/ifl occurs when the middle
term is wrongly assumed in any of the premises and so
cannot be taken to prove the conclusion. It follows that
the premises which contain the false middle become them-
selves false. Thus the fallacy of the asiddha virtually stands
for the fallacy ol false premises, which is a form of the
material fallacies in Western logic.
There are three main forms ot the fallacy ol asiddha,
namely, (i) the airayasiddha (ii) svarupasiddha and {in)

vyapyatoasiddha/ Of these, the airaydsiddha is a middle

term which has no locus standi. One condition of a valid
middle term is that it must be present in the minor term.
The minor term is thus the locus of the middle. Hence if

the minor term is unreal and fictitious, the middle cannot

be related to it. The result is that the minor premise, in
which the middle is related to an unreal minor, becomes
false. This is illustrated in the inference ‘
the sky-lotus is

> Vtde N8V., 1 2. 6

* Bftdhyatvftt s&dbyaBamab, NS 1 2 S
» TB.,p.68! TB.,p. 81.
 I^B FALLACIES OF INFERENCE 3l6

fragrant, because it belongs to the class of lotus.’ Here

the minor term ‘
sky-lotus ’
is unreal, so that the middle
‘
class of lotus ’
cannot subsist in it. The middle term
having no locus standi, we have a fallacy of the a&rayisiddha
^
or the baseless middle.
The svarupdsiddha is a middle term which cannot be
proved to be real in relation to the minor term. It is a

middle term which is not found in the minor term. * The

existence of the middle in the minor being unreal, the
minor premise which relates it to the minor term becomes
false. Thus if one argues sound is eternal, because it is
* •
‘

visible,’ he commits this fallacy. Here the middle teim

‘
visible ’
is wrongly assumed in the minor term ‘
sound ’
and
IS not justified by facts. If the minor term stands for a

number of things and the middle is found in some but not

all of them, we have the fallacy of hhagasiddha or efradeia-

siddha. To illustrate
•
‘
the four kinds of atoms of earth, etc.,
are eternal, because they are fragrant.’ Here the middle
‘
fragrant ’
is related only to a part of the minor term,
namely, the atoms of earth, but not to the other
kinds of atoms. Hence the middle term is partly false

and so equivalent to the svaiupasiddha middle. The

fallacies of hhagasiddha or ehade^asiddha are therefore
included witliiu the fallacy of soarupasiddha. It includes also
such other fallacies as (i) oiksandsiddha, where the middle
term has a false adjunct, as when one argues ‘
sound is

eternal, because being a substance it is intangible,’ while

sound IS not a substance but a quality ;
(u) viksyasiddha,

where the middle is an unreal substantive of a real adjective,

e.g. ‘ sound is eternal, because it is an intangible substance ’

(Hi) asamarthavtksanasiddha, vheie the middle has an

^ There are two kinds of this fallacy, vtz, aaatpak^a and stddhas&dhati^, The
^

first is explained above. The aecond means a middle whuh seeks to prove a proved or

undoubted fact, c jy.

*
a body has h nbs, letai te it is so perceived.* Cf, TM., Ch. IT.

* T8., & TB., f6id.

316 NYaY\ TH'SORY of knowledge

untneaniQg adjunct, e g. ‘sound is eternal, because being a

quality it has no cause,’ in whicli the adjunct ‘

being a
quality ’
has no force oi sense in the argunient ;
(ic)

asamarthaoisesyasiddha, where the middle is an unmeaning

‘
substantive of a significant adjective, c g sound is eteinal,

because It IS an uncaused quality,' in which the adjective

’
‘
uncaused’ renders the word ‘ quality ’
(juite supeifluoiis

The oyapyatbasiddha is a middle term w hose concomitanee

yepdph) with the major cannot be pioved • A valid middle

term must be univeisally lelated to the major term. II

a middle term is not known to be universally concomitant

with the major, it becomes invalid The result is that the
major premise which should express a i ijdph or a univeisal
relation between the middle and major teims becomes mate-
rially false. The fallacy of the lyapyatiasiddha may arise in

two ways It may be due to the non-eoneomitanee of the

middle term with the majoi, as m the iiileieiiee ‘

all reals

are momentaiy, sound is a real, thereloie sound is moment-

ary.’ Here the major premise is false, because theie is no
’

universal relation between the ‘

real ’
and the ‘
momentaiy
Or, it may be due to the presence of an upadhi oi condition,

on which the lelation between the middle and majoi terms

depends. Here the middle teini is not, as it should be, un-
conditionally related to the major and is, theiefore, lalse.

It IS illustrated m the infeience ‘

the hill is a case ol

smoke, because it is a case of fire ’

This inleience is in-

valid, because the relation of the middle teim ‘fire’ to the

major ‘
smoke ’
is conditional on its being ‘
hre fiom wet
luel.’ This fallacy of the conditional middle is technically
called anyathasiddha. ‘

1 Vide TB , pp 34-86.
* Vyapyatvagiddhastu gu eva yatia lielorvyapiirEdvagauiyale, llid.
3 TB , p 35.
*
Vid«NVT,p 846.
 tflli FALLACIHS OF INFERENCE 317

(). The fallacies of Mlatiia and badliita or the mistimed

and contradicted middles

The hdlatUa literally means a middle term which is

vitiated by the lapse of time *

In this fallacy the middle

teim consists of two or moie events which succeed one

anothei in time. But on the analogy of the given example,
these events must be simultaneous if the middle term, consti-
tuted by them, is to prove the conclusion. Since, however,
they are successive, the middle term becomes inappropriate
111 the order of time and is therefore called kdldtUa or the
mistimed middle. It is illustiated m the inference ‘
sound
IS dm able, because it is manifested by conjunction, like
colour.’ The colour of a thing is manifested when the
thing comes m contact with light, although the coloui
exists betoie and alter the contact. So also, it is argued,

muiid winch is manifested by the contact beliAeen two

tliingb (samyogavyangya) must be durable, i.e exist before

and aftei die contact. But the aigument is fallacious be-

cause its middle term is vitiated by a limitation m time. In

the case of colour the manifestation takes place simul-
taneously with the contact between light and the coloured
object. The manifestation of sound, however, is separated
by an interval of time fiom the contact between two things.
In fact, w'e hear the sound when the contact betw'een the
two has ceased. Hence it cannot be due to the contact,
because when the cause has ceased, the effect also must
cease. The middle term being incongruous with the given
example fails to prove the conclusion and is therefore falla-
cious.^ In this sense the kdldtita means a middle term
which 18 subject to different conditions in the two premises

* KalatyayapadJ^ial^ kalabifcal.i, NS ,
1 0.

2 Kalatyayena yukto yasyftrthasyaikade^o ’ padi^yam&naBya aa kalatita

udahara^asadhariDyaByabb&vadasadbaDaiuayam, NB., 1.2 9.
3] 8 NYiYA THEORY OF KNOWLEDGE

of the syllogism As such, it becomes a kind of fallacy that

corresponds to the fallacy of accident in Western logic.

According to a second interpretation, the hdldtUa is the
fallacy of a wrong order of the different members of the

syllogism, ft IS illustrated when there is an inversion of

the natural order of the prc.nises and the conclusion, as
when we put the premises alter the conclusion. On this

view, the kdldtlta corresponds to the fallacy of hysteron

proteron. But this view of the matter is not accepted by
the Naiyayikas. A change in the order of the members of a
syllogism does not really affect its validity nor render it

fallacious. Further, such a change does not involve a fallacy

of the middle term or an inferential fallacy It constitutes

a defect in the method or procedure and is, therefore, de-

scribed as the clincher of the inopportune (apraptaMla

^
nigrahasthdna) .

Although the fallacy of the hndhita has been treated by

some writers as another name for that of the kdldtlta^ yet it

seems to me better to distinguish between the two in view

of the sharp contrast in their meanings While the kdldtlta
stands for a middle term vitiated by a limitation in time,

the hddhita means a middle terra which is contradicted by

some other source of knowledge (pramdndntarena) A middle
term is contradicted when it leads to a conclusion, the
opposite of which is proved to be true by some other
pramdna. This is illustrated by the argument ‘
fire is cool,

because it is a substance ’
Here the middle term
‘
substance,’ which seeks to prove that fire is cool, is

contradicted because we know from tactual perception that

^ Ibtd It should be remarked here that although it be usual m a syllogiBm to put

the premises before the conclusion, yet that is neither logically necessary nor psycho-
logically correct It is now generally recognised by logicians that a syllogism may
take another form in which the cxinclusion comes first and the premises follow it

Hence we see that a change in the usual order of the propositions in a syllogism
involves neither the fallacy of hystgron proteren nor the clincher of the inopportune.
 THE FALLACIES OF INFERENCE 319

fire is not cold but hot. The fallacy of satpratipaksa, as

explained before, is different from this fallacy of bddhita

because in the former one inference is contradicted by

another inference, while in the latter an inference is contra-
‘
dicted by a non-mferential source of knowledge.

7. The fallacies of chala, jdti and nigrahasthdna

Apart from the fallacies of inference, the Naiyayikas

deal with certain other fallacies which occur in connection

with the art of- debate. These are called chala, jdti and
mgrahasthdna. The fallacy of chala consists in using the

same word to mean different objects in the course of a

debate. It thus corresponds to the fallacy of ambiguity in
Western logic. It is of three kinds, namely, vahchala,
sdmdnyachala and upacdrachala. In vdkchala or the fallacy
of equivocation the same word is used in different senses.
This is illustrated when one man says the boy ‘
is nava-
kamhala ’
(possessed of a new blanket), and another objects
‘
he IS not naca-kambala (possessed of nine blankets). In
sdmdnyachala the same word is taken to mean an individual
and the class to which it belongs, e.g. one man says ‘
this

Brahmin is a learned man,’ and another objects ‘

all Brah-
mins are not learned men ’
In upacdrachala or the fallacy
of figure of speech, a confusion is made between the
figurative and literal senses of an expression, e.g. when one
says ‘
the scaffolds cry out,’ and another objects ‘
scaffolds
^
cannot cry.’
Jdti is the fallacy of irrelevance. In it a futile argument
is based on some irrelevant consideration which does not

really prove the conclusion. There are twenty-four kinds of

jdti or futile arguments. The first is called sddharmyasama,

1 Vide TB., p. 87 , TS., pp. 68-60.

* Vide NS. & NB., 1 2.10.14,
320 NYiYA THEORY OF KNOWLEDGE

where an argument is based on some kind of similarity

‘
between two things, e.g. sound is eternal because it is

incorporeal ! The second is laidharmyasama,

ike the sky.’

where an ^argument is based on some kind of dissimilarity

between two things. The ulhotsasama, apalarsasama,
vamyascima, ararnyasavia, i ihalpasama and sddhyasama a»e
futile arguments in whicli the character of the minor teim
or the example is altered or they are unduly assumed without
sufficient reason The prapbsania and apraptisama are

futile objections based on the wrong implications of the

coexistence between the middle and major terms or their
absence The piaf^angasama and piatidrstaniasavia are futile

objections based on the giound that the given example has

not been proved by a senes of arguments, or that tlieie is a

countei -example The amilpalttsa7na is an objection based

on the gionnd that the middle teim ol the given uigiiment
cannot exist in the minor term before it comes into existence.
The samhijasama is an objection based on the doubt
aiising ftom a middle teim with opposite example^. The
prnharanaf'ama is an objection based on the giound ol a

middle teim which is i elated to both the sides ol a contio-

veisy. The ahetmania is an objection which is based on

the ground that the middle term is unintelligible in the thiee

orders of time The arthapatti^ania is an argument based

on mere presumption The anae^iasama is an argument to
prove the identity of all things on the ground ol their having

existence in common. The vpapathsama is an objection

based on the ground that there is a counter-argument to

the gnen aigument The vpalahdhisaina is the objection

to a given argument based on the giound that we can

perceive the truth of the conclusion even without the
argument. The anupalabdhisama is an argument to invali-
date a given argument from non- perception, on the ground
that non-perception cannot be perceived. The nityasama
is an argument to prove the eternality of all non-eternal
 THE FALLACIES OF INFERENCE 321

things on the ground that they are eternally non-eternal

The anityasama is an argument to prove the non-eternality
of all things on the ground of their resembling a non-eternal
thing in some respect or an other. The Mryasama is

argument opposed to a given argument from the nature of

an effect, on the ground that an effect may have very different
natures, and so cannot be taken to lead to a single conclu-
^
sion.

The mgmhasthani which literally means a ground of

defeat, is a fallacy which is due either to a misunderstanding
or to the want of understanding. It is said to be of twenty-
two kinds. These are : pratijmhdm or weakening one’s
proposition by adducing such examples as run counter to it

prati^nantara or shifting the proposition ;

pratijnavtrodha or
contradicting the proposition ;
pratijMsannyasa or renounc-
ing the proposition ;
heivantara or shifting the ground ;

arthantara or shifting the topic ;

nirarthaka or the
meaningless statement like abracadabra; avi]nalartha or the
unintelligible statement ;
aparthaka or the incoherent state-

ment apraptakdla
; or the wrong order of the parts of an
argument nyuna ;
or the suppression of any part of an
argument ;
adhika or the duplication of the middle term
or the example ;
punariikta or the meaningless repetition
of any part of an argument ;
ananuhhdsana or the refusal
to answer a question ; apidna or ignorance of the proposi-
tion ;
apraUbhd or the inability to give a reply to the

argument ;
viksepa or evasion of the argument ;
matdnupid
or admission of the defect in one’s argument ;
paryya-
nnyojyopek§ana or overlooking a defect in the argument ;

niranuyojydnuyoga or finding fault with the faultless

apasiddhdnta or the deviation from an accepted position ;

and hetvdbhdsa or the fallacy of the middle term.^

1 r»de NS, 12.18, 61 Iff

» Vid« NS 1.2 19 621 ff.
, ;

41— (1117B)
 NYIYA THEORY OF KNOWLEDGE

It will appear from the above that some of the fallacies

of chala, jdti and nigrahasthdna come under the inferential

fallacies, while others are either semi-logical or non-

logical fallacies. These relate either to the meaning
of words and propositions, or to the conduct of the parties

concerned in any discussion. Hence any elaborate account

of these three kinds of fallacies with their many subdivisions

is not necessary in connection with the Nyaya theory of

inference.
 BOOK IV
UPAMiNA OR COMPARISON
 CHAPTER XV
THE NATURE AND FORMS OF UPAMANA
1. The Nyaya definition of upamdna

The word upamdna is derived from the words upa

meaning sadr^ya of similarity, and mdna meaning cogni-
tion. Hence upamdna derivatively means the know edge l of

^e similarity between two things,. This derivative mean-

ing, however, requires certain qualifications in order to
give a complete definition of upamdna. As a pmmdna, upa-
mdna is the source of our knowledge about the relation
between a vvoid and its denotatioa {samjnd-samjnisamhan-
dha) ’ We have such knowledge when first we are told by
some authoritative person that the word denotes a class of
objects of a certain description and, secondly, finding some
objects of that description we recognise them as denoted by
that word. The description of the unknown objects denoted by
the word is generally given in terms of their similarity to some
familiar object of experience. Hence upamdna is generally

defined as the ground of our knowledge of a thing from its

similarity to anothei thing previously well-known. ^

Thus
a man, who does not know what a gaiaya or wild ox is,

may be told by some forester that it is an animal like the

cow. When next he meets with such an animal in the

forest, he knows that it is the gavaya. But the description

of the unknown objects denoted by a word may also be given

in terms of their dissimilarity to certain known objects or

1 SamjilasaibjilisaiubsDdbajfi&DamupamitiJ; etc., TS., p. 62.

* Vtdt NS. & NB., 1.1.0
326 NYiYA THEORY OF KNOWLEDOE

their peculiar properties. * Hence upamana or knowledge

by comparison is not always due to the knowledge of

similarity or dissimilarity between things. The knowledge

of similarity or dissimilarity is an accidental character of
this or that kind of upamana. What, however, is common
to all cases of upamana is the knowledge of the denotative
relation between a word and a certain class of objects. Hence
the Naiyayikas finally define upamana as the process of
reasoning by which we know that a word denotes a certain
class of objects on the basis of some authoritative state-

ments
Analysing the process of reasoning in upamana we get

the following steps. First, we have an authoritative state-

ment (attdesavakya) that a word denotes objects of a certain

description, e.g ‘ the gavaya is like the cow.’ Secondly,
when one observes any such object, be has the knowledge that
it answers to the given description (sadrsyadln). Thirdly,
there is a recollection of the descriptive statement received
from authority {vakyarthasmrti). Lastly, there is the
resulting knowledge that this kind of objects is denoted by
^
the word in question {upamxti) Thus a man, who does
not know what objects are denoted liy the word gamya, may
have it from some authority that the word denotes animals
resembling the cow. When next he happens to find such
animals, he perceives their striking similarity to the cow.
Then he remembers the authoritative statement that animals
resembling the cow are gavayas. With this he comes to the
'

conclusion that the word ‘ gaoaya denotes this class of

animals.
It may here be asked : W^hich of the four factors men-
tioned above is the karana or operative cause of the know-
ledge derived from upamana or comparison ? It cannot

> Vide NV. k NSV., 1.1.6.

* Vide BP., 79-80.
 NATURE AND FORMS OF UPAMiNA 327

obviously be the last, since that is the resulting cognition,

of which we want to know the principal cause. According
to the older Naiyayikas, the first factor, namely, the descrip-
tive statement of some authority is here the karana or special
cause of the knowledge of denotation of words. The modern
Naiyayikas, however, hold that the perception of similarity,
etc., IS the special cause whose function (vyapara) is to revive
in memory the authoritative statement and thereby lead to the
knowledge in question. A man recognises a gamya as such
just when he perceives its similarity to the cow and remem-
bers the statement ‘
the gavaya is an animal resembling the
^
cow.’

2. The Jaim, Mimdthsd and Vedanta views of upamdna

According to the Mimamsa and the Vedanta, upamdna is

the source of our knowledge of the likeness of things. In

some cases we may get the knowledge of likeness from per-
ception, as when we perceive a gavaya and know it to be like
the cow. From the perceived likeness of the gavaya to the
cow we next know that the cow is like the gavaya, although
the cow IS not now perceived by us. This latter knowledge of
the likeness of the unperceived cow to the perceived gavaya is

due to upamdna or comparison. It cannot be due to percep-

tion, since its locus, namely, the cow is not now perceived,
but only remembered. It is no doubt conditioned by the
perception of likeness in the gavaya. But when from
this perceptual knowledge of likeness we pass to the know-
ledge that the cow, not now perceived, is like the gavaya, it

is no longer perception. Nor is this second knowledge of

likeness due to inference. In inference the paksa or the

minor term is an object of perception and the hnga or the
middle term is present in the pak§a. In the alleged inference

> Vide NM., pp. 141-42.

328 NYiYA THEORY OF KNOWLEDGE

of the cow’s likeness to the gaoaya, the pak^a, i.e. the cow
cannot be an object of perception, and the linga, i.e. the
likeness of the gavaya would be present not in the paksa
‘cow,’ but in the gavaya. Further, when from the one like-
ness we know the other, we are not conscious of any
inferring, but of comparing. Introspection tells us that the
actual process of reasoning involved in the second knowledge
of likeness is not inferential ‘
Similarly, when we perceive
a horse and know it to be unlike the cow, our knowledge of
the unlikeness is perceptual. But when from this we
conclude that the cow, not now perceived, is unlike the
horse,we depend, not on perception and inference, but on
upamdna or comparison Thus the Mimamsaka and the
Vedantist admit that there is a perceptual element in

upamdna. But they go further and prove that the reason-

ing about likeness and unlikeness, based on some percep-
tion, cannot be fully explained by perception or inference.
It constitutes an independent source of valid knowledge
ipramdna), to which they give the name of upamdna
The Kaiyayikas criticise and reject the above new on
the following grounds First, they point out that it violates

the ordinary rule of upamdna oi compaiison In all cases

of upamdna we compare the unfamiliar object with some-
thing well-known in order to understand it better. In the
above view the well-knowm cow is compared with the
strange gaiaya But this cannot give us any new knowledge
about the cow which is already too well-known to us.

Secondly, the knowledge of the cow’s likeness may be

explained by memory and so does not require a separate
source of knowledge like upamdna. When we perceive the
gavaya we are reminded of the cow and not of other things.

The reason is that there are certain points of resemblance

between the two and that these were previously perceived

• Vide 8D , pp 74-70 ; VP., Ch III.

 NATURE AND FORMS OF UPAMiNA 329

with the perception of the cow. Hence we liave a memory

of the cow which was previously perceived to have
as that
some resemblance with the gavaya which is now perceived.
Hence there is no need for an independent pramana called
upamana to explain the knowledge of likeness and unlike-
^
ness
It IS to be observed here that Nyaya criticism has so far
very little force It is true that we ordinarily understand
an unfamili.ir object by comparing it with what is familiar.
But this does not prevent us from comparing the familiar
with what lb new and unfamiliar Nor does it invalidate
our subsequent knowledge of the former as like or unlike the
latter Further, we cannot say that all knowledge of
likeness is memory A man who has seen a gaoaya may,
when perceiving a cow, know its similarity to the gavaya.
It IS possible only for him to remember that a cow was

found to be like a gavaya at the time when he perceives the

gavaya. But a man who has never seen a gavaya cannot
know that a cow is similar to it. When the perception of
a gavaya suggests to him tliat the cow is like it because the
gavaya is like the cow, we cannot say that he only remem-
bers the cow’s likeness, since there was no previous percep-
tion of it

The Naiyayika is perhaps conscious of the weakness of

his first two arguments and so brings forward a third one to

supplement them. He thinks that even if upamana be

different from memory, we may very well exp'aiu it as a

form of inference. From the perception of the gfami/o we

know that it has some points in common with the
remembered cow. This leads to the inference that the

remembered cow is like the gavaya, because it has some

points in common with the gavaya. The \edantist’s
upamana is thus reduced to a mediate syllogistic inference :

1 Fide NM., pp 116-47

42— (1117B)
330 NYAYA THEORY OF KNOWLEDGE

“ Whatever has certain points in common with another

thing IS like that thing ;
the remembered cow has some
points in common -with the perceived gavaya ;
therefore it is
’
like the gavaya.”
The Naiyayika seems to be on strong ground when he
reduces the reasoning about likeness and unlikeness to infer-
ence. The Vedantist’s wpamanfif, when analytically consider-
ed, deals with our knowledge of the relations among correlative
terms. Ordinary syllogistic inference is concerned with the
relations of subject and predicate among different terms.
But there are other relations which turnish grounds of
inference. These are the relations among correlative terms
The doctrine of correlation (prafiyogitva) and the relations of
correlative terms have been ranch elaborated in the modern
Nyaya. There are two kinds of correlation, namely,
nhhavripinfiy:gitia or the correlation existing between
a te I and its contradictory, and satpratiyogitva or the
correlati ui cm-, mg between relatiie terms. The relations
among comalictory terms like A and not-A, red and
not-red aie the grounds of immediate inferences by con-
version, obversion, contraposition, etc. The correlations
among relative terras are the grounds of both mediate and
immediate inferences There are different tjpes of such
c ('relation. Bradley^ enumerates four types of these rela-

tions. These are (1) The synthesis of identity, e g

A IS the father of B, B is the son of A; or, A is the brother

of B, and B of C, then A is the brother of C (2) The
synthesis of degree, e g. X is greater than B, B is less

than A ; or, A is hotter than B, and B than C, therefore A

than C. (3) The 8\nthesis of time, e g. X is earlier than
B, B is later than A ; or, A is a day before B, B contempo-
rary with C, therefore C a day after A. (4) The synthesis

1 Vide NM., p. 148.

2 Bradley, LogtCt Vol I, pp. 264*66
 NATURE AND FORMS OF UPAMANA 331

of space, e.g. A is north of B, B is south of A ; or, A is

noith of B and B west of C, therefore C south-east of A.

The Vedantist’s upamdna deals with the correlations of
likeness and unlikeness which, following Bradley, we may
call the synthesis of comparison It consists in reasoning
from the proposition ‘
A is like B ’
to the proposition ‘
B is

like A’ ; 01 from ‘
A is unlike B ’
to ‘
B is unlike A ’
It
IS really a kind of immediate inference. But there being
no such thing as immediate intorenee in Indian Logic, the
Naiyfiyikas pul it in the lorm of a syllogism which has
the additional advantage of testing the validity of such
reasoning.
The Jamas do not lecognise uyamana as an independent
source of knowledge, but include it under p}atyahht]i~id or
recognition They understand pratyahhijM in a very wide
sense so as to cover all such cases of knou ledge as this is '

that object,’ ‘
this is like that,’ ‘
this is unlike that,’ ‘
this

is at ii distance from us,’ ‘

this is a tiee,’ etc. It is clear

from this that pratijahhijm heie stands for any knowledge

which IS conditioned by perception and memory The
leasoning from the pioposition ‘
the gavaya is like the cow,’
to the proposition ‘
the cow is like the gavaya ’
is not a
case of upamana, but of piatyahhipid, since it can very
well be explained by the perception of likeness m the
gavaya and memoiy of the cow. So also the Naiyayika’s
upamana as a source of the knowdedge of denotation of

words IS, according to Jamas, a case of piatyabhijM or

recognition. The knowledge ol the likeness oi unlikeness
through which we lecognise an object is given by percep-
tion. The knowledge ol the object as a cow oi a gavaya
IS due to the recollection of their description as supplied to
'
us by some competent person
It is to be observed heie that the Jama view of upamana

1
Piiimeyahamalamniiandat pp 9740^
332 NYiYA THEORY OF KNOWLEDGE

as a form of pratyahhijM or recognition rests on a wrong

assumption. They seem to think that a knowledge is

explained when we explain the constituent parts of it. But

to explain the component parts of knowledge is not to
explain knowledge itself. To say that it is so is the funda-
mental error of all associationist theory ol knowledge. If it

were really so, the Jama view of praiyabhipid itself as a

distinct type of knowledge will have to be discarded, since
it IS constituted by perception and memory. On this

a'.sumptiou we may reduce all kinds of knowledge to

perception, since the constituents of all knowledge ulti-

mately come from perception. That we recognise other

kinds of knowledge than perception is due to the fact that
the combination of elements derived from perception
involves new principles which take us beyond perception.
We shall consider hereafter if the Naiyayika’s upamana
involves any new principle of combination so as to justify
us in treating it as a new kind of knowledge.

3. The classification of upamana

Upamana was at first regarded as only of one kind,

namely, as the knowledge of a thing as denoted by a word
through its similarity to a well-known object of experience.
Later Xaiyajikas, however, distinguished between different
kinds of upamana. according as they are based on the
knowledge of dissirnilaiity between things, or on that of
then peculiai propertie.s Thus the Naiyayikas generally
recognise three kinds of upamana or knowledge by compari-
son, namely, sddharmyopamdna laidharmyopamdna and ,

dharmamdfiopamdna. '

In sddharniya-upamdna we start from the description of

an unknown object given in terms of its similarity to a well-

' V„h Tit H S8 , PI.

S1..S8
 NATURU AND FORMS OF DPAMiNA 333

known object by some authoritative person. If then we

find any object or objects that answer to the given descrip-
tion, i.e. are similar to the things they are compared with,
we know that they belong to this or that class. Here then
we apply a class-concept to certain facts on the basis of
some obseived similarity between them and other known
facts. The concept is given to us and the facts to which
it applies are ^elected by us This kind of uparndna is

illustrated in the citizen’s application of the name gavaya

to the wild oxen because they aie found to satisfy the
description of the gavaya as an animal similar to the cow.
Ill vaidharmya^upantdna the objects denoted by a word
aie described in terms of their contiast or dissimilarity to
some well-known objects of experience. This negative
description enables a man to recognise certain objects as
denoted by a word or as belonging to a certain class in so
fai as he finds that they fit in with the given description.

This It illustrated when a man recognises certain animals

as belonging to the class of horses because, unlike the cow,
they have no cloven hoofs
In dha} mamdha-upamdna the objects denoted by a name
are described in terms of their peculiar attributes or any
combination of attiibutes which is peculiar to them. This
description enables us to discriminate the things denoted by
thename from all other things and consequently apply the
name to just that class of things. This is illustrated when
from the description of ‘
man ’
as a cooking animal or of
the hatahha as a loiig-necked animal with projecting lips

and feeding on thorns, we recognise the animals denoted by

these words It may be observed here that these three kinds
of uparndna are illustrated also by medical students when
they colK'ct herbs and plants according to the descriptions
given of them in the mateiia medica.
Visvanatha in his Nydija-Snfra-Vrtti mentions another
Kind of upomuna which is slightly different from the above
 NYAYA THEORY OF KNOWLEDGE

three. Here upamana consists, not in the knowledge of

the denotation of a word, but in that of some unknown
property of an object through its similarity to a known
thing. Thu& if on hearing that a certain herb resembling
the mudgapainl is an antidote and then finding such a
herb we conclude that it is an antidote, out knowledge is
^
due to upamana or comparison.
It is sometimes held that the Naijayika’s upamana is

an analogical aigument There are two facts that lend

colour to this view First, the Nyaya-Sutia defines

upamana as the knowledge of an object (mdhtja) from

some recognised similaiity between tw'o things (prastddhn-
sadharmyat) Secondly, the last kind of upa)Hdna men-
tioned by \’i3vanathn \er) clo^ely appioximates an analogi-
cal reasoning wdien tioin the ob^elved U'semblances betw'een
two things we aigue to the I'lcsence of some unobserved
property in one of them But fiom the >3\aya ficcoimt

of upamana a' given above it will appeal that it docs not

leally coirespond to an analogical aigument In analogy
we infer one resemblance fiom othei lesemblances ,
e g
when we say A lesembles B in having the propeities
X and y, therefoie it lesembles B in having the property z
Butin upamana wt aigue as much from lesemblance as'

from contrast and p*culiariiy Fuithei, upamana leads not

to the knowledge of resemblance bctw'cen things, but to

that of the denotation of a w'ord, or to the application of a

name to a class of objects Even the special kind of upa-

mana mentioned by Vn^vanatha is not a knowledge of
resemblance, but is the identification of an object from a

given description

^ Vide Nydya’Sutra-Vrtti and Nyaya-Sutra^-Vivarana 116

334
 CHAPTEE XVI

upamana as an independent souece

OF knowledge (PRAMANA)

1 . Can npamana give us any valid hnowledge?

With regard to- the Nyaya view ot upamana it has been

urged by the Carvakas that it cannot give us any true know-
ledge about the denotation of words as maintained by the
Naiyayikas. In it we aie lo know th(' objects denoted by a
word from their similarity or dissimilarity to certain well-
known things or from their peculiarities. But mere resem-
blance 01 difference without any universal relation cannot be
the ground of a certain conclusion In the stock example of
sMharmya-upamana w’e are to know that a certain animal
must be a gavayn because it is similar to the cow. If the

similarity betw'een the two be perfect, then they become

identical with each othei Hence on the ground of such
perfect similarity it is as much true to say that the animal
IS a cow as to ^ay that it is a gaiaya II, on the other hand,
the similarity be semi-peilect oi considerable, then the word

gavaya may be taken to denote buffaloes in so far as they are

consideiably similar to the cow' If, again, the similarity

be imperfect oi shuht, there is nothing to prevent the appli-

cation of the name gavaya to cats and dogs in so far at

least as they are animals like the cow Similarly, any de-

scription of a class of things in terms of then dissimilarity to

certain well-known things or in those of their peculiarities

does not always help us to recognise them as such-and-such,

or know them as denoted by this or that word. This shows
336 NYiYA THEORY OF KNOWLEDGE

that upamana or mere comparison between things is not a

’
valid source of knowledge.

Now the Naiyayikas meet the above sceptical aigument

against the validity of upamana in two ways. First, they

point out that the argument rests on a misunderstanding as

to the real nature of upamana or comparison as a method of

kniwledgc It is not the case that when upamana is based

on similarity it IS committed to one of the tliiee degrees of
perfect, serai-perfect oi imperlec, similarity Far from this
being so, it has been expressly laid down that the similarity
must be one that has an accredited bearing on the subject in
question (piasiddha). Tlie siunlanty must be essential

and requisite, and serve as sufficient ground foi the

recognition of a class of things ,is denoted by a word
Upamana or comparison as a source of knowledge operates
through su di observed Minilaritv or dissimilarity as is rooted
in things and limits the denotation of a word to them As
a matter of fact, there is no such rule that the similarity

must be of a particular degree as perfect or semi-perfect

or imperfect. What particular sort of resemblance is meant
by the similarity in question depends on the special circum-
stances of the case and the context in which an argument
through comparison occurs As such, the given description
m terms of similarity, etc ,
makes a selection of its own
objects and brings them under a class-concept or a name m
the light of our previous experience. In the stock example,
’

the judgment ‘
this is a gauaya is brought about, not by the
degree of the similarity between the cow and the wild ox,

but by the sugg'^stiveness it has acquired in relation to our

past and present experiences. It is this suggestive character
of the similarity that restricts the denotation of the word
gavaya to the wild ox and excludes the buffalo and the like.

Secondly, the Naiyayikas do not deny that upamana some-

1 Vtde NS. & NB., 2.1.42.

 UPAMiNA AS INDEPENDENT PRAMiNA 337

times leads to wrong judgments, e.g. the judgment of a

buffalo as a gamya. But then this difficulty is not peculiar

to upamana. All of our perceptions and inferences are not

ipso facto valid. Still we recognise perception and inference
as valid sources of knowledge. If so, there is no reason why
we should deny that upamana is a valid source of the know-
ledge of some objects. The wrong judgments of comparison
may, like those of perception, be explained as due to the
wrong application of a right principle and not to the logioal
*
principle itself.

2. Can upamana he reduced to any other pramdna ?

Admitting that upamana is a valid source of knowledge

iprarndna), it may be pertinently asked : Is it an indepen-
dent source of knowledge, irreducible to any other ? This
question has been answered in the negative by some systems
of Indian philosophy These systems reject the Naiyayika
view of upamana as an independent method of knowledge
We have already considered the attempt made by the Jainas
to reduce it to pralyabhijhd or recognition. In some other
systems the attempt has been made to reduce it to perception
or inference or testimony. Hence the Naiyayikas discuss the
question of reducing upamana or comparison to some other
pramdna.
According to the Buddhist logicians upamana is a valid

but not an independent source of knowledge. It can be

explained as a combination of perception and verbal testi-

mony. There are two factors in upamdna, namely, the

knowledge of the similarity or dissimilarity between two
classes of things and the knowledge of the fact that things
of a certain class are denoted by a certain word. As to the
first factor we see that it is obviously given by perception.

1 Vide NB & NVT ,

2 1 43.

43-(1117B)
 338 fTYIYA THEORY OF KNOWLEDGE

When we see two things together we perceive that they are

similar or dissimilar to each other. As to the second factor,

namely, the knowledge of the denotation of words, it is

derived from the statements of authoritative persons, i.e.

testimony. Hence upamana need not be given the status of

an independent source of knowledge. Now the Naiyayikas

point out that the Buddhist contention rests on a complete
misunderstanding of the real nature of an argument by
upamana or comparison. The vital point in upamana is

neither the perception of similarity nor the verbal knowledge

I' of the denotation of a word, but the recognition of certain
objects, not known before, as belonging to a class

sand denoted by a class-concept. The similarity of those

objects to other things may be perceived and the class-

concept may be given by testimony. But the application of

the concept to a particular class of things cannot be

due to perception or testimony. Hence upamana cannot be
^
reduced to perception and testimony.
In the Samkhya ^ and the Vaisesika ® system upamana
IS explained as a form of inference. It is here admitted that
the Naiyayika’s upamana is neither the perceptual nor the
verbal cognition of the similarity between two objects, e.g.

the cow and the gavaya. On the other hand, upamana

really aims at the knowledge of the denotation of a word or
class-concept. But this can very well be explained as due
to inference. Analysing the Naiyayika’s upamana we
get three component factors. First, there is the com-
munication of some knowledge about the denotation of

a word by authoritative statements. This is obviously

a case of knowledge from testimony and is in the
form of the sentence ‘
the word gavaya denotes animals
resembling the cow.’ Secondly, there is the observation

1 Vide mv ,1.1 e
* Vide Tattvakaumudi and Avaranavirini, 6,
s Vide PS , p. 109.
 Tjt>AMiNA AS INDEPENDENT PRAMiNA 839

of a certain animal resembling the cow. This gives

us a knowledge of the animal’s similarity to the cow,
which is undoubtedly perceptual in character, since it

is due to sense-object contact. Thirdly, there is the

knowledge that the word gavaya denotes animals of the
same class as this particular animal now observed. This
last cognition is wrongly supposed by the Naiyayikas to be
due to upamana But it is really an inferential cognition
based on the knowledge of vyapti or a universal relation
between the word gavaya and animals resembling the cow.
The inference may be put m the form of the following
syllogism :

All animals resembling the cow are gavayas ;

This is an animal resembling the cow ;

Therefore this is a gavaya.

As against the above attempt to reduce upamana to

inference, it has been pointed out by the Naiyayikas that

the knowledge of the denotation of a word, which upamana
aims at, is possible without the knowledge of vyapti or a
universal relation between two terms. An argument by
upamana or comparison does not consist in an inductive
generalisation and its application to a new case. It consists

in the application of a class-concept to some objects because

they fit in with a given description. Upamana being thus
possible without the knowledge of vyapti cannot be reduced
to inference which is never possible without a knowledge

of vyapti or universal relation between two things. Further,

there is an unmistakable difference between the forms of
the cognitions in inference and upamana. In upamana the
resulting cognition is always expressed in terms of likeness,
etc., while an inferential cognition is expressed in terms of
the relation of ground and consequence. In inference the
introspective consciousness is a feeling of the *
therefore-
 340 NYiYA THEORY OF KNOWLEDGE

relation,’ while in upamana it is a feeling of similarity, etc.

In upamana we are not conscious of inferring but of

comparing. Inference is distinguished from perception

because our cognitions are distinctly different in the two

cases. Just for the same reason upamana must be distin-
*
guished from perception, inference and testimony.

3. Conclusion

The question discussed in the Nyaya theory of upamana

is this . How do we know the denotation of a word or a
class-name ? There are various ways in which we may
know it. In the first place, the objects denoted by the
word gavaya may be pointed out to us by any person who
knows its denotation and we may be told that these objects

are denoted by the word. In this case we know the

denotation of the word from direct testimony, because here
in the presence of the denoted objects we are told by some
authority ‘
These are the objects denoted by the word
gavaya.’ But it is not always possible for us to know the
denotation of words from direct testimony, for we cannot
always be brought to the presence of the denoted objects and
told that they are denoted by such and such words. There
are, however, other ways open to us to know the denotation
of words. We may know the denotation of a word from
its accepted definition or from a description of the objects
’
denoted by it. Thus from the definition of the word ‘
man
as a rational animal we understand what animals are denoted
by it. Similarly, from the description of the gavaya as an
animal resembling the cow, we can recognise the class
of animals called gavaya. Now the question is : What is

the nature of the process of knowledge involved m our

^ Vide TaTkasathgraha’Dtptka-Prakd^af p 63 , Dtnakcirl and R&inarudri on
Siddhantamukidvali^ pp 354-55.
 tfAMiNA AS INDEPENDENT tKAMUlA 341

understanding the denotation of words in this latter way ?

Is it perception or inference or testimony or any combination

of these ? According to the Nyaya, it is a distinct method of

knowledge called upamana or comparison. It is no doubt

true that the process involves an element of perception and
testimony. The definition or the description comes to us
as a spoken or written statement of some authority and, as

such, is but a form of testimony. Similarly, we know by

perception that certain objects possess the attributes or
characteristics included in the definition or the description.

But from this we cannot conclude that the process involved

in the knowledge of those objects as denoted by a word
IS a combination of perception and testimony. To explain
the elements of a knowledge is not to explain the knowledge
itself, if it has a distinctive character of its own. Percep-
tion does not become ideation because it involves certain
ideas and images. So too inference cannot be reduced to

perception and testimony even though it includes certain

elements derived from them. For the same reason the
process of knowing the denotation of a word should not be
reduced to perception and testimony, since they explain
certain elements of the process but not the process itself. The
next question is •
Can we not explain the knowledge of deno-

tation by an inference ? This can be done if we show that the

knowledge of denotation follows as a conclusion from a univer-
sal proposition as major premise through the mediation of

the minor premise. Now the knowledge of the denotation

of a word can be deduced, at least theoretically, from a
universal proposition like ‘
all animals resembling the cow
are gavayas.’ But to show that the denotation of words

can be known by inference is not to show that it is actually

so known. When we know the denotation of a word we

do not argue syllogistically from premises to the conclusion,
but simply compare certain facts with a given description.
To know that these facts fit in with the description require
 NYiYA THEORY OF KNOWLEDGE

a selective activity of the mind which is distinct from

perception, inference and testimony. Hence we conclude
with the Naiyayikas that upamana or comparison is an
independent source of our knowledge of the denotation of
words.

342
 BOOK V
SABDA OR TESTIMONY
 CHAPTER XVII

THE NATURE AND CLASSIFICATION OF SAB DA

1. The ^lyaya definition of M)da and its different kinds

3ahda literally means verbal knowledge. It is the

knowledge of objects derived from words or sentences. All
verbal knowledge, however, is not valid Hence hhda^ as a
pramano, is defined in the N\aya as valid verbal testimony
It consists in the assertion of a trustworthy person. ' A verbal
statement is valid when it comes from a person who
knows the truth and speaks the truth about any-
thing for the guidance ot other persons. ^ But it is a
matter of common observation that a sentence or statement
is not by itself sufficient to give us any knowledge of things.
Nor again does the mere perception of the words of a sen-
tence lead to any knowledge about objects. It is only when
one perceives the words and understands their meanings that
he acquires any knowledge from a verbal statement. Hence
while the validity of verbal knowledge depends on its being
based on the statement of a trustworthy person, its possibi-

lity is conditional on the understanding of the meaning of

that statement Hence hbda or testimony as a source of

valid knowledge consists in understanding the meaning of >

*
the statement of a trustworthy person.
It will appear from the above definition that the first step

in hbda or testimony is the perception of the words of a

sentence or proposition set forth by some trustworthy person.

^ Ipfopade^H^ ^abdal^, NS ,
1.1 7

2 Fide TR , pp. 94-95.

3 Vide BP. & SM ,
81 Cf Vakyarthajfianara ^abdajfianam, TS p ,
73

44-(ni7B)
 346 NYiYA THEORY OF KNOWLEDGE

In the case of a spoken sentence we have an auditory percep-

tion, and in that of a written sentence we have a visual per-
ception of the constituent words. Secondly, there must be
an understanding of the meaning of the words perceived
by us. It is through this understanding of the meaning of
words that we come to the final step, namelj ,
the verbal
knowledge of objects or the truth about certain objects.

Thus the katana or the special cause of ^ahda or valid ver-

bal knowledge is the knowledge of words (padapiana)^ which
leads to the knowledge of objects through its function
(vydpara) of recalling to our minds the meanings connected
with words or sentences. ’
Thus ^ahda is distinguished froni
the preceding praindnas by the fact that it is due to the
knowledge of words oi sentences, while jierception is due to

sense-object contact, infeience to the knowledge of vyapti or

universal relation, and upamdna or comparison to the percep-
tion of similarity or dissimilarity.
There are two ways in which all verbal knowledge has
been classified in the Nyaya system According to Vatsya-

yana, verbal knowledge is of two kinds, namely, drstartha or

that relating to perceptible objects and adrstartha or that
^
relating to imperceptible objects. 'I he first is limited to
the ordinary sensible objects of this world, while the second
relates to supersensible objects which cannot be known by
means of perception Under the first head we are to include

the trustworthy as-ertions of ordinary persons, the saints

and the scriptures in so far as they bear on the perceptible
objects of the wwld. Thus the evidence given by witnesses
in law courts, the knowledge about plants that we get from
a reliable farmer, the scriptural injunctions about certain
rites and ceremonies for rainfall, birth-control and the like

are illustrations of drstartha iahda. The second will include

^ Pada]5ftoarii to kars^aih dvaram fcatra padftrthadhUi, etc , BP & SM , 8J,

* Ba (Ivividho dfs^&dr^^ftrthafvat, NS dt NB ,118,
 iCATURfi AND CLASSIFICATION OF ^ABDA 347

all the trustworthy assertions of ordinary persons, saints,

prophets and the scriptures in so far as they bear on super-

sensible realities. Thus the scientist’s assertions about
atoms, ether, electrons, vitamins, etc., the prophet’s in-
structions about virtue and vice, the scriptural texts on God,
heaven, future life and the like are illustrations of adrsiflrtha
iabda.
According to the later Naiyayikas, there are two kinds
of iahda or verbal testimony, namely, vaidika or the scrip-
tural and laukika or the secular. In the first we have the
words of God. The Vedas are created by God and are there-
fore valid on .all points. Vaidika or scriptural testimony is

thus perfect and infallible by its very nature. As distin-

guished from this, laukika or secular testimony is not all

valid. It is the testimony of human beings and may there-

fore be true or false. Of laukika testimony, only that
which proceeds from trustworthy persons is valid, but not
the rest. *
It will be observed here that while the first classi-

fication of sahda depends on the nature of the objects of

knowledge, this second classification has reference to the
nature of the source of knowledge in testimony. All
Naiyayikas, however, agree in holding that testimony
must always be personal, i e based on the words of some
trustworthy person, human or divine In respect of truth or
validity there is no difference between the trustworthy asser-
tions of an ordinary person, a saint, a prophet and the
scriptures as revealed by God.

2. Other systems on the nature and forms of iabda

Of the other systems of Indian philosophy, the Carvaka,

Bauddha and Vai^esika do not recognise iahda or testimony
as a distinct pramdna or source of knowledge. According to

^ Vakyam dvividbaai, vaidikam laukikaficai etc., TS.^p. 73,

348 NYiYA THEORY OF KNOWLEDGE

the Carvakas, there is no logical ground or justification for

our believing in anything simply on the statement of another

person. If it were so, we shall have to believe in many
absurd and fictitious objects about which an\ fool may tell
us. If, however, hbda or testimony be constituted by the
statement of a trustworthy person, it is only a case of
inference from the character of a man to the truth of his

assertion. But inference cannot be accepted as a valid

source of human knowledge. Hence hhda or testimony
should not be recognised as a pramana or valid method of

knowledge. ^
According to the Buddhist logicians, ^ahda is

not an independent source of knowledge, but a form of per-

ception or inference If by hhda we mean to prove that
the person who makes a certain statement is trustworthy,
we reduce it to inference. If, however, we u<-e it to prove

that there are actual facts corresponding to a statement, we

reduce it to perception ^ According to the Vai4esikas, §abda
as a form of knowledge is to be included in inference, since

the ground of our knowledge is the same in both.

Just aj „ LP .
inference we know an unperceived objectJrora
the perception of something which is related to it, sojn
Mbda from the perception of words we know the objects
^ •» I* •* ••

which are unperceived but related to the words perceived

°
by us.

In the Jaina system sabda is recognised as a separate

pramana or source of knowledge It consists in the know-
ledge derived from words which, when properly understood,
express real objects and are not inconsistent with the
evidence of perception. It is called laukxka or secular
testimony when the words come from an ordinary reliable
person of the world. It is called idstraja or scriptural
testimony when it proceeds from a liberated self of extra-

*
Vide SarvadarSanasaihgrahat Chapter I
* Vide S C VidySbhu^aija’s History of Indian Logic pp. 287-88
3 Vide PS.,pp 108-8, VS .
923.
 NATURE AND CLASSIFICAIION OF ^ABDA 349

ordinary powers and knowledge, and relates to supersensible

realities. ^
Thus while in the N3aya system scriptural
testimony depends on divine revelation, in the Jama it

comes from the perfected and omniscient finite self. In

the Samkhya-Yoga system too we find a recognition of hbda
or testimony as a valid method of knowledge. ^ But while
in the Saiiikhya, scriptural testimony is regarded as im-
personal and therefore possessing self-evident validity, the
Nyaja takes it as neither impersonal nor self-evidently
valid. It holds that the scriptures have been created by
God and require to be proved by reason as much as any
other form of knowledge. According to the Mimamsa
system, ^abda as a pramdna consists in the* true knowledge
of objects, derived from the understanding of the meaning
of a sentence. It is called pauruseya or personal when con-
stituted by the words of trustworthy persons, and apauruseya
or impersonal when constituted by the w’ords of the
Vedas. The Prabhakara school of the Mimamsa, how'ever,
takes ^abda to mean only vaidika or scriptural testimony
about the existence of supeisensuous realities According
to the Vedantists, ^abda or agama as a source of valid
knowledge consists in sentences or propositions which assert

a certain relation between things, that is not contradicted

in any way. ® It is a verbal knowledge of objects, which is

not validly contradicted by any other knowledge. While

this IS implied in the Nyaya definition of Sabda, there is

some difference between the Nyaya and the Vedanta with

regard to the nature of vaidika or scriptural testimony.
According to the Nyaya, scriptural testimony is personal,

since the Vedas have been created by the supreme person

' Vtde Pramcyakamala y pp 112 13

2 Vtde TattvaJmmndi and Avaranavdrtm, 5 , Yoga-Bhdsya, 1.7
3 Vtde SD . p 72
4 Vtde Jlia’s Prahh&kara School of Purva-Mmaihsd, p. 62
6 Vide VP ,
Chapter IV
 NYiYA THIilORY OF KNOWLEDGES

or God. For the Vedanta, it is impersonal inasmuch as

God does not create but only reveals the contents of the
Vedas, which are eternal truths independent of God. So
also the Mimamsakas look upon the Vedas as a system of

necessary truths or eternal verities which are independent

of all persons and therefore purely impersonal in character.

For the Naiyayikas, the Vedas as a system of truths embody

the will of God. They express the eternal reason of the
divine being in the order of time.

.150
 CHAPTEE XVIII

OF WOEDS (PADA)

1. Sounds and words

In the last chapter we have seen that §ahda as a pramdna

consists in sentences or propositions put forth by some
trustworthy person. Now a sentence is a group of words
(pada) arranged in a certain way. To understand a sentence

(vdkya) we have to understand its constituent words. Hence

we propose to consider here the nature and meaning of
words, as well as other questions in connection with the
understanding of words.
Sabda literally means sound. In linguistics it means
also words or sentences. A word is a particular kind of
sound. So also a sentence is a group of sounds arranged

in a certain order. How then is a word related to ordinary

sounds ?
According to the Nyaya, sound is a physical phenomenon.
It is the attribute of an intangible and all-pervading sub-
stance called dkdia or the ether. Air is not the substratum
of the quality of sound, but the medium of its transition

from one place to another. Sound is a product of the

conjunction of two bodies or of the disjunction of the parts
of one composite body. It is therefore non-eternal or

subject to origin and cessation in time. ^

The Mimamsakas
here controvert the Nyaya position and bold that sound is

eternal, since it is not produced, but only manifested by the

contact of two bodies. It is unnecessary for our present

» TB , pp. 26-27.
352 Nm’A THEORY OF KNOWLEDGE

purpose to enter upon the long controversy between the

Nyaya and the Mimarhsa on this point.
’

Sound is of two kinds, namely, dhiani and varna. ® A

dhvani is an inarticulate sound, e.g. the sound of a bell or
a drum. It is a confused mass of sound-sensations having
no order or arrangement of its parts. It has no fixed nature

of its own, nor any fixed relation to other like sounds.

Dhvani is thus incapable of forming parts of any language.
On the other hand, a varna is a sound produced by the

action of the vocalorganof human beings, e.g. the alphabet.

A varna is a letter which has a fixed character and a definite

place in the alphabet of any language All varnas or letters

are constituents of human speech. They may be either

spoken or written. Spoken letters are auditory sensations
of significant sounds, while written letters are visual sensa-
tions of coloured figures. From the standpoint of

linguistics, the cries of birds and beasts, and even of new-

born babies are dumb and inarticulate They are as variable

and disorderly as sounds produced by physical things.

These do not lend themselves to any use as parts of any
language. Hence they are included within dhvani and not
made a separate class
A word IS a group of varnas or letters arranged in a
certain fixed order. The oider of the letters in a word
cannot be changed or reversed in any way without altering
its meaning. Thus the word ‘
cow ’
is a grouping of the
letters c-o-w in the given order If we change this order
’

we destroy the word itself. Similarly, the words ‘

won
and ‘
own,’ which contain the same letters, become different

because the fixed order of the letters is different in the two

cases. While a letter is a significant sound, a word is a

symbolic sound of a higher order. A letter signifies only

' Vtde NM., pp 206-82

^ ^abdo dhvant^ca varnaha, efcc , BP ,
164 65
 OF WORDS 353

a part of the alphabet, but a word stands for some thing

or some idea. Like letters, words may be either spoken or
written. A spoken word is the object of auditory perception
and a written word that of visual perception. Thus words
are symbolic sounds constituted by letters arranged in a
definite order. A word is not a mere collection of letters,
but a definite whole of letters or syllables which are its
parts and have a fixed order in the whole. It is a
unity of the parts in so far as it is the object of a single
cognition. ' The question as to how the constituents of a
word are synthesised so as to form one whole, will be
considered later on.

2. Words and their meanings

A word IS defined as a group of letters arranged in a

fixed order. This definition, however, has in view the
existence or the constitution of a word The essential nature
of a word lies in its meaning.^ Logically a word is a sound
that bears a certain meaning. The meaning of a word con-
sists in its relation to the object which it signifies. A word
may have different meanings according to the different ways
in which it may be related to an object. According to the
Naiyayikas, there are three kinds of meaning of a wofd,
namely, ahhidhd, panhhdsa and Idksana. ® Let us here con-
sider these different kinds of meaning of a word.

By abhidha is meant the primary meaning of a word.

It IS also called iakydrtha, idcydrtha and mukhydrtha. The
relation between a word and its meaning may be either

samketa or laksana. Samketajs th£ direct relati on between

a word and its meaning, such that the knowl edge of (he

^ Padam ca varijasarauba^^i, etc ,

TB , p 14
^ Cf * padaih ca vanjasamubah ’ {TarkahhUsd^ p. 14),
*
iaktaib padam *
{Tarkct

safhgraha, p. 64).
3 Vide SahdaSakU-prakaStka, Introduction (C. U. Edn.)

46— (1117B)
364 NYAYA THEORY OF KNOWLEDGE

word leads immediately to the knowledge of its relation tc

that meaning. Now saniketa or the direct relation betweer

a word and its meaning may be either eternal or non
eternal. When eternal and unchanging, it is called oi

the inherent potency ot a word Thus the relation betweer

the word jar and the object called jar is a direct and eterna
relation called &akti This hkh or potency of a word is dui

to the will of God that such and such a word should mear
such and such an object. According to the Mimamsakas
the 6akti of a word is its natural relation to the object wind
it signifies Just as fire possesses the power of burning, s(

words possess a natural (lotency to mean certain thingi

independent!) of The Nai 3 ayiUas

the will of any person.
however, contend that the relation between a word and iti
meaning is not a natural but conventional relation. Whei
the relation is established by God it is called hkti, anc
when it is due to the usage of mankind it is called pan
phasoL Now the meaning called up by the i^akti or inheren
potenc) of a word is its ahhtdha or .hkydrtha, i e. primar
meaning. The word wdnch jiossesses such a meaning ii

^
called a sakta or vacaka word
When samketa or the direct relation between a w'ord ant
its meaning is non-eternal or changeable, it is called pan
bhasd. This is due to the wall of the authorities in anj
science that such and such a word should mean such anc
such an object. The meaning called up by the conventioi
established by authorities is the panhhdsita or technica
meaning of a word. Words which bear such meanings an
called parihhdsika or technical words, e g. the wordi
‘article’ in giammar, ‘premise’ in logic, ‘court’ inlaw
‘
catc;:ory ’
m philosophy.
^

By laksand is meant the secondary meaning of a word

* Vide T8 and TD , p. 64, SahdcSakti^prakdhkd, pp 55 f,

^ Vide S(ihda4akti‘prakd4tkd^ pp, 54-65,

 OF WORDS 356

It is the indi rect or implied meaning in which we should

under stand a w ord w hen its direct or primary meaning does
not consist with other words or the context. A word in-
directly means an object when it is related to it because of
its direct relation with something else with which the object
IS somehow associated. When we are told ‘
the house is

on the Ganges,’ we take ‘the Ganges’ not in its primary

meaning of ‘
the current of water,’ but in the secondary
meaning of ‘
the bank of the Ganges.’ Here the secondary
meaning is suggested through its association with the
primary meaning There are three kinds of laksand or
secondary meaning, namely, jahallaksand, ajahallaksand and
jahadajahallaksand. In jahqllaksam^ no part of the prim^
‘
meaning is retained ,^e g. the scaffolds cry out.’ In ajahal-
laksand, the primary meaning of a word is also retained jn

the implied meaning, e.g ‘ a blue jar ’

meaning ajar with the
attribute of bluenes^ In jahada]ahallaksand, a part of the
primary meaning is retained, while another part is discard-
ed, e.g. ‘this IS that man,’ meaning the identity of the
man leaving out the attributes of ‘
this '
and ‘
that
’

' The
Naiyayikas do not admit with the Vedantists that not
only words but sentences also may have secondary meanings
(laksand) ^

The dlamkdnkas or ihetoiicians recognise another kind

of meaning of words, namely, cyaniand. This stands for

such meanings of w'ords as are neither directly nor indirect-
ly related to them, but only suggested by them. Thus the
sentence, ‘
the house is on the Ganges,’ may be taken to
mean that the house is cool and sacred. This meaning iq

called vyanjand or the suggested meaning The Naiyayi

kas do not recognise lyafijand as a different type of thv.

meaning of words, but include it within iakti and laksand

' Vide SahdaMi-tnaH,<ikd,pf -W f Vidf also TaUi arfipilca, p. 67.

2 Vtde VP ,
Ch IV, Sahdaiaktf’prahdhka, pp. 01 f
566 NYiYA THEORY OF KNOWLEDGE

or the primary and secondary meanings. The vyahgyartha

or suggested meaning of a word is really inferred from its

primary and stcondaiy meanings and is not separate from

^
them.
How do we learn the meanings of words ? There are
different ways in which we may learn them First, we learn
the meanings of the radicals, verbal roots, suffixes, etc., from
grammar. Secondly, we know the meanings of certain
general names by means of upamma or comparison, as
when we know the gavaya from its similarity to the cow.

Thirdly, we learn the meanings of words from dictionaries

Then we may know the meaning of certain words from

authority, as when a connoisseur tells us that such and such
objects are denoted by a certain word. Or, we may know
it by induction from the different uses of words by authori-
tative persons, as when we know the meaning of the w'ord

cow from the different uses made of it by our elders in

relation to a particular kind of animal. we may know Or,
’
the meaning of a word from its context, as when the chair ‘

means the chairman in a meeting

‘
Or, we may know
’

it from a given explanation, as when we understand a word

from any of its synonyms Finally, we may know the
meaning of a word from its application in connection with
a familiar word, as when we understand the meaning of the

word pika from the sentence ‘

the pika is crying cuckoo
’ ^
on this tree

That there are so many different ways of knowing

the meanings of words proves that the relation between
words and their meanings is not a natural but a
conventional relation. If there were a fixed natural
relation between a word and its meaning as between fire

and burning, then the word should have always coexisted

1 Vide Tattvadiptk&,)? ^ SahdaSakti-prak&hkat pp 64 f

* ‘
Saktigraham vySkarapopamana,* etc ,
8M , pp 359-72 C/ SahdaSahtt-
prakdhkdf 20
 OF WOllDS 357

with the object signified by we should have known

it and
their relation simply by perception. But a word does not
coexist with the object denoted by it. The word fire does
not coexist with the object ‘
fire ’
and produce any burning
sensation in us when we utter the word. Nor do we per-
ceive the relation between a word and its meaning in the
same way that we perceive the relation between fire and
burning. Further, the conventional character of the
relation between words and their meanings is evidenced
by the different meanings in which the same word is used
by different people. The variation in the meanings of

words cannot be explained on the hypothesis of a natural

relation between words and their meanings. It appears

also in the use of different words to mean the same thing,
e.g. aqua, water, ]ala. The convention {samaya) that such
and such words should n ean such and such objects is esta-
blished by God where the relation between words and their
meanings is a fixed and eternal relation called MU or
denotation. It is established by human beings living in a
society where the relation between them is a changeable
relation called panhhasa or lahsam.^
Words are divided into four kinds according to the
different ways in which their meaning is determined. A
word IS called yaugtka or etymological when its meaning
IS solely determined by those of its component parts, e.g.

the word data or giver meaning one who gives. It is called

rudha or conventional jwhen its meaning is determined by

the whole independently of the part meanings, e.g. the word
go meaning, not one who goes, but the cow. Soine
words are called yoga-rtldha^ or etymologo-conventionai
when the meaning determined by the whole agrees with
that determined by the the parts, e.g. the word
pankaja meaning a water-lily which grows m the mud.

’ 2 1 ^abdaialctt-firaJcditka, tbtd
358 NYAYA THEORY OF KNOWLEDGE

Lastly, certain words are called yangtka-ru dh^ or etymolo-

gical-conventional when then meanings are determined
either by the potency of the whole or by those of the parts,
e.g. the word udbhid meaning a germ or the sprouting of a
^
seed or a sacrifice.

riig tT/nyoii oj wor^

What IS the piimary m “aning of a word ? Does a word

mean an individual (vyalh), oi a pirticular form (dkrti)

or a umvers.il {jdti) Theie are different views about the

import of words These have been explained and exa-
mined by Vatsjayana m the Nydya-Bhdsya

According to some thinkeis, including the Samkhyas, a

-
woid denotes an individual object {cyakh). By an indivi-
dual IS meant a composite material b)dy possessing specific

properties. It is a sub-iance which has a limited dimension

and may have such qualities as smdl, taste, colour, touch,
etc It IS manifested and open to sense p?rc‘ption. It

follows from this that the principle of individuation is

matma signata or quantitatively determined matter and the

individual must have a manifest body {murti). * That
such individual objects are denoted by words is evident
from the established usage of mankind When we use
such expressions as ‘that cow stands,’ ‘ a herd of cows,’
‘
he gives a cow to the Brahmin,’ etc ,
we evidently mean
the individuals called cows. Such expressions cannot refer

to the genus or the universal ‘

cowiiess,’ since the universal
IS one and eternal, and so cannot be specified as this or

that, or spoken of as a collection of miny objects. Further,

if words do not by themselves mean individuals, we cannot

’ V«l«8M,pp 381.85
^ NS & NB 2 2
Fide , 57, Vnarai^aprameyasamgraha 181
^ Vyaktirgu;^ivi»C8l8rayo rourtil?, NS ,
2 2 64 Vtde also NB ,
tbul
 OF WORDS 359

explain tbeir reference to individual objects by any process

of transference of meaning.
The Naiyayikas reject the above view of the import of

words. If a word mean an individual as such, then any

word could mean any and every individual. A word,
however, does not mean any individuals, but the individuals
o f a certain clas^ In such expressions as ‘that cow stands,’
etc., what is meant by the word ooio is not the mere
individual by itself but the individual as distinguished by
the generality of cowness. Hence it is not true to say that
words denote individuals only. Although words do not, by
themselves, mean individuals, yet they may refer to indivi-

dual objects by reason of the individuals’ association or

^
connection with the primary meaning.
The second view about the import of words, which is

accepted by the Jainas and others, is that a word denotes

the particular form or configuration of individuals. ^ The
form (akrti) of a thing consists in the particular arrangement
of its component parts and the constituent particles of those

parts “The form of a thing is that which indicates the

generality and its characteristics.’’ Things are distinguished
from one another by their peculiar forms. The cow is

differentiated from all other animals by its form which con-

sists in the collocation of the dewlap, etc. Words denote
objects only as they express their forms or configurations in
spice, by which their nature is determined. Hence a word
must primarily mean the form or the structure which deter-

mines the individuality of an object. The iNai.iayikas reject

this view also on the ground that the form by itself is not
sufficient to constitute the nature of a thing. The clay

model of a cow is not what we mean by a cow, although it

possesses the form of a cow. Hence a word should not be

1 Vide NS. & NB., 2.2. 68.69.

2 Vide Vivofanc^prameyasarhgraha, p. 181,

360 NYiYA THUIORY OF KNOWLEDGE

taken to mean only the form or the physical shape of an

^
individual apart from its generality or class-essence
In view of the above difficulties in the ‘
individualistic’
and the ‘
configuration ’
theories, the Mlinamsakas and
Ve lantists propose a third theory, according to which a
word means the genus or the class-character of individuals.
The genus is th*^ basis of similar cognitions with regard to
different in livi luals. ft gives us a comprehensive know-
ledge of miny things as similar in essential points.

Wiir ls primarily mean such universals or genera as distin-

guish the pirticulars of experience. ^ If a word were to

mean the individual, then it must have as many meanings as

there are individuals meant by it. This, however, goes
against the law of parsimony which requires that a word
should have one primary meaning Although words prima-
rily mean universals, there is nothing to prevent them from
referring to the individuals. We know the individual at
the same time that we know the universal, because these are
inseparable in respect of both knowledge and existence. Or,
it may be said that while the universal is the primary mean-
ing of a word, the individual is its secondary meaning
(laksand). Thus the word blue primarily means ‘blue-
ness’ as a universal, but in the phrase ‘
the blue pot ’
it

means, by implication, the individual with the attribute of

blue colour. In the same way, although the woid cow
means ‘
cowness,’ yet by implication it means the individual
'

possessed of the generic attribute of cowness.

According to the Naiyayikas, words do not mean
universals only, since these cannot be understood apart from
the individuals and their particular forms A genus can be
recognised only through the individuals that constitute it and
their peculiar configurations. Hence the true view is that a

< Vide NS. k NB. 2.2.60-61 , 2.2.65

* Vide NS * NB, 2.2 61 ; 2 2.66
« Vide VP., Ch'IV; 8D., Ch. J.
 OF WORDS 3ei

word means all th e tbree^ namely, the individual, the con|;,-

guration, and the generality. *

It cannot be said that a word
directly means the universal and indirectly the individual,
for it has only gne primary meaning. The universal, the
individual and the form enter into the full meaning of a word
which does not exclusively mean any of them. All the three

factors are present in the meaning of a word in the same

way, though with different degrees of prominence. Hence
it in actual usage we do find only one factor to be evident,
that IS not because the other two are absent but because we
are not interested in them for the nonce. When we are
interested in the'difference or distinction of one thing from
others, what we do is to emphasise its individuality in the
meaning of the wor<l used for it, eg. when we say ‘
that
cow is standing.’ But when we want to stress the unity or
similarity of things, we give prominence to the generality as
a factor in the meaning ot the word Used, e g. when we say
‘
the cow IS eternal.’ Thus the old Naiyayikas conclude that
ev .'ry word means the univeisal, the individual and some
paiticular form, and that one of these is predominant, while
^
the re^t are subservient factors in the meaning of a word.

Among the mode, n Naiyayikas, however, some hold that a

w ild means an individual as charaolerised by the universal
{jdhvihstavyakti),^ while others maintain th it it means an
individual as qualified by both the universal and the configu-
ration {jdtydkrtimhstavyakti).'^ It follows from this that

there are three aspects in the meaning of a word, namely, a

pictorial, a denotative and a connotative. A word calls up
the form, denotes the individual, and connotes the genus or

* Vide NS & NB ,
S 2 (>2-63
> Vide NB ,
2 2 C3
3 Vtde Dmakari on Suldhaniamukidvalit 81
< Vide RVl ,
81, Sabda'^aktt-prahUikd, 19 Cf L. S Siebbi ig, Modern
Introduction to Logic^ p 500 “ Tbe demonstrative symbol means its denotation, i e ,

it sta’ids for the object denoted ,

whereas ths descriptive phrase means the properties and
"
not the objects (if any) denoted

4 r>-( 1117 B)
362 NYiYA THEORY OF KNOWLEDGE

the universal. Every word will therefore be coniiotative in

so far as it means the generic properties of the individuals

denoted by it. Indian logic thus leaves no room for the so-
called non-connotntive terms of Formal Logic in the West.

4. Che unity of loords and the hypothesis of sphota

A word is a group of letters having a certain meaning.

The letters composing a word have a definite otdei of suc-
cession among them We perceive the constituent letters
one after the other But the letters or syllables composing
a word cannot mean anything by themselves. Letters and
syllables bear a meaning only when they are combined into

one whole called a word To put the same thing in a differ-

ent w'ay, a w'ord means an object when it is j.erccived as the

unity of a number of letteis or syllables Thus the w'ord

cow means a paiticular kind ol animal when the letters

c-o-u ,
are percened and formed into the unity of one w'ord
Hence the question arisen How aie the letter^ in a word
combined into one whole ^ Is the unity of the w'ord due to
a synthesis of perception or memory or the intellect or some-
thing el'C ^
According to the Naiyayikas, the letteis composing a

word cannot be simultaneously perceived We can perceive

only one thing at one instant Hence the letter> of a wmrd
must be successively pe'ceived by as But in the order of
succession when one is present, the others are mther past or

future How then can there be .i synthesu of them all

into one word ^ The Naiyayikas hold that it is by means

of memory. It is true that we perceive the different
letters one after the other But when we come to the last

letter, the impressions of the preceding letters are retained

in our mind. Hence the paiception of the last letter as

aided by the impressions of the preceding letters presents

the word as a whole of many letters, and its meaning is
 OS’ WORDS 363

understood according to convention. Thus the word ‘cow’

IS perceived by the auditory sense all at once when it

becomes related to the last letter and is aided by the im-

pressions left by the preceding letters, although they are
past. ^
The unity of a word is thus explained by the
Naiyayikas as due to memory and association between the
letters composing it. According to the Vedantist, it is due
to Ithe synthetic activity of the intellect. The separate
experiences of the constituent letters come to us successively,

but they are synthesised into the perception of one word

by the intellect that holds together these experiences
^
(sainastapraLyaDamarsinl huddht).
The above explanations of the unity ot a word as due
to the synthesis of rncoiory or the intellect involve certain
difficulties for which the grammarians propose the theory
of the sphota When we perceive the last letter of a word,
we have no perception of the preceding letters. AH that we
can have at that inoinent is a memory of this or that pre-

ceding letter, but not of all. Strictly speaking, the

Naiyayikas cannot allow more than one cognition, a per-

ception or an image, to be in the mind at one moment.

Even if it weie possible for us to have the i tnpressions of

all the piecediug letters, they will serve to give us a know-

ledge of those letters by way of memory, but not of the

thing signified by a word. Then the Vedantist simply

assumes that the intellect holds together the experiences

of all the letters, but does not show how these fleeting and
successive experiences can be snmiltaneously present before
the same intellect. In truth, a woid is not a unity, but a

sei ics of successive sounds called letteis. These letters

cannot be unified into the experience of one word which,

therefore, cannot signify an object. All that the series of

1 SU ,82, TB., p 11.

* Vide Samkara-Bhafya an.i Bhamaii, 1 3 28.
 3^4 NYiYA THEORY OF KNOWLEDGE

letter-sounds does is to manifest one inarticulate sound-

essence called sphota which is the real unity of a word and
brings about the cognition of the object said to be meant
by the word. Like the genus, the sphota is an eternal
essence which is common to all the utterances of a word.
Corresponding to every word there is such a sphota or
sound-essence which is gradually unfolded by the letters

of a word. When a particular word is uttered, its sphota

or unitary principle is manifested and that directly presents

the meaning of the word. Hence the sphota is the real

word that means an object and there is no such thing as a

word of letters meaning things.'

The theory of sphota has been justly repudiated by many

right-thinking philosophers The sphota is not only, as
Thibaut remarks, a grammatical fiction, but is also useless as

an explanation of the unity of words. It has been severely

criticised and rejected by Sarnkara, Kumarila, Vacaspati
and others ^ It cannot be denied that words mean objects
and that they consist of letters or syllables arranged in a

definite order. When a thing is expressed by a word, all

that we perceive are letters and no sphota Even if there be

such a thing as the sphota, we do not understand how it can
mean an object when it is gradually unfolded by the letters
of a word. If a senes of successive sounds called letters
cannot form a single word, how can the successive stages
of the manifestation of the sphota or sound-essence be syn-

thesised into a unitary whole ? The theory of the sphota

does not bring us nearer the solution of the problem as to

how there can be a simultaneous perception of successive

facts as we find it in the perception of a word. Neither the
Naiyayikas nor the Vedantists give a satisfactory answer to
this question. They forget that a synthesis of the letters

> Vide NVT ,3 2 .57 ;

Samkara-BhStya, 1 8 28
* Vide Sarhkara-Bhdsya, 1.^ 2S NVT, 2 2 57 ; Sdstradiptkdf pp 95-97,; Sloka-
vdrttikat pp 610-44
 OF WORDS

in a word by memory or the intellect is not the direct know-

ledge that we mean by the perception of a word. The
Naiyayikas were forced to draw this conclusion by their
view of the mind as atomic and therefore incapable of having
more than one cognition at one instant Had they fully
realised the implication of their view of ‘
the present ’
as a
block of time comprising several instants, they could have
easily solved this problem. As many modern psychologists
like James, Titchener, Royce and others have sho\^n, our
present consciousness is not like an indivisible mathemati-
cal point, but IS extended like the saddle-back. It has a
span or. duration of its own. It extends both backward into
the past and forward into the future. Hence in the present

consciousness we may have a numbei of successive facts,

although that is very limited in our case Thus there is a

simultaneous perception of all the letters of a word, although

these are successively read or heard by us Hence we con-

clude that the unity of a word is due to the synthesis of
^
perception and not of memory or anything else

^ Vide 3 Pnnnples of Psychology Vol T, pp 608-10, Titchener, Text-Book

of Psycholog i/y p, 341 , Royce, The World and the Indtvtdvaly II, in.

366
 CHAPTER XIX

OF SENTENCES (VIKYA)

J The comiuichon of a sentence

A sentence (( (7/.'i/a) is a cimhination ot words liaving a

certain meaning An> combination of woids, however,

does not make a sigaincant ^enimce. The (on&truetion of

an intelligible scnteiite must conloim to foui conditions.

These are ahendbh, yoijyaid, '^annidhi and tdtpaiyya.^

B} dhdnksd or ex[)ecianc} w meant tliat quality of the
words of a sentence In which they expect oi imply one
another. A woid cannot b^ itsell coiive\ a complete iiican-

ing. It must be biought into lelation wiUi other words in

order to express a full judgment ^

When one heats the w'ord
‘bring’ uttered before him, lie at once a^ks ‘
what ^

The verb ‘
bring has a need foi some olhei words denoting
some object or object', c (/. ‘the jai
’
In the absence of

such words, it has no meaning and falls short of a complete

judgment. Similaily, a w'ord in the nominative case re-
quire.s a verb to convey a complete meaning. Generally
speaking, the dhanksa or exjiectancy of words is the relation
between kriydfva and kdrakatia, the verb and the case-
endings implied by it.'* When I say ‘dog,’ ‘horse,’
‘
cow,’ ‘
man,’ etc., I simply utter a string of names which
do not imply one another and cannot theref'oie constitute a

sentence. The reason is that there is no kriydtca and

' Vide BP ,
&‘2

* Cadaiya padaDtaravyatirekttprayuktananvay4naoubh4vakatvamak6fiksi, TS , p 72
3 Vide 8M , pp 423 f , TC ,
IV, pp 218 f.
 OF SENTENCES 367

hdrakatva between these words in the strictly grammatical

sense. ’
In some cases, however, we may have a sentence
without the relation of knydtva and kdrakaLva between its

constituent words, e.g. “ so ’yarn Devadatta.” In the case

of the relation of identity the ordinal y conditions of knydtva
and kdiakatva are not necessary Still, we cannot deny the
expectancy of the words in an identity proposition. The
words imply each other in so lar as one means the same
thing as is meant by the other. When we say this is
‘

that Devadatta,’ the ‘

tins’ and the ‘
that ’
mutually imply
each other. According to the Vedantist, there is dkanksd
or expectancy bet w'cen words, not onlywhen one actually
implies the other, but may possibly imply it. Thus when
I say bung the cow,’ one may ask what kind of cow
‘ ‘
’

Hence the word coiv may imply adjectives like black, old,
etc. There cannot be a significani sentence unless its terms
^
are thus capable of implying one another.
The second condition of the combination of woids in a
sentence is then yogyatd or mutual fitness. It consists in

the absence of contradiction in the relation of the objects

denoted by a sentence. When the meaning of a sentence is

not contiadicted, there is yogyatd or fitness between its

’

constituent words. The sentence ‘

moisten with fire

{agnind siiicet) is wanting in fitness because there is a

contradiction between fiie and moistening Hence there

must not be any incompatibility between the meanings of

the different words so as to render the whole sentence itself

meaningless. Some modern Naiyayikas do not consider the

knowledge of fitness to be a necessary condition of verbal

knowdedge. According to them, what prevents the under-

standing of a sentence is the knowdedge of the incompati-
bility between its words. As such, we may very well have

1 Vide TB , p. 13.
> VP Ch , IV.
368 NYiYA THEORY OF KNOWLEDGE

a verbal cognition only if we are not aware of any inconsis-

tency between the words of a sentence. We do not require

a further knowledge of their consistency or fitness with one
^
another.
Sannidhi or dsatti is the third condition of verbal
knowledge. It consists in the propinquity or proximity
between the different words of a sentence. If there is to be
an intelligible sentence, then its constituent words must be
continuous with one another in time or space. Spoken words
cannot make a sentence when separated by long intervals
of time Similarly, written words cannot construct one
sentencewhen the> are separated by long intervals of space
Thus the words bring a cow will not make a sentence
‘ ’

when uttered on three days or written on three pages, even

though they possess the fir^t two marks of expectancy and
^
fitness.

Tdtparyya as a condition of verbal knowledge stands for

the meaning intended to be conveyed by a sentence. A
word may mean different things in different cases. Whether
it means this or that thing in a particular case depends on
the intention of the person win uses the word. To under-
stand the meaning of a sentence we must consider the
intention of the writer or the speaker who uses it. Thus
when a man is asked to bring saindhava, he is at a
loss to understand whether he is told to bring salt or a
horse, for the word means both This can be ascertained
only if we know the intention of the speaker. Hence
the understanding of a sentence depends on the understand-
ing of its tatparyya or intended meaning In the case
of ordinary sentences used by human beings, we may ascer-
tain their tdtparyya from the context (prakarana) in which
they are used For the understanding of the Vedic texts

* Arthabadho yogyatA, TS p ,
72
9 Pad&cftmavilambenocc&rapaih sapnidbilji, ihtd.
 OF SENTENCES 369

we are to resort to the logical rules of interpretation syste-

matised by the Mlmamsa.^
With regard to the importance of tdtparyya or intention
as a condition of verbal knowledge there is much difference
of opinion among Indian thinkers. Some hold that a
definite knowledge of the tdtparyya or the intended meaning
is an essential condition of verbal knowledge. Others
think that an understanding of the tdtparyya is necessary
only in the case of equivocal terms and ambiguous expres-
sions having two or more possible meanings. Others again
maintain that while tdtparyya is a condition of verbal know-
ledge, it is not to be admitted as a separate condition, but
should be included within the first condition of dkdnksd
or syntactic expectancy. By dkdnksd we mean the need
that one word has for another in order to convey the intend-
ed meaning of the speaker As such, tdtparyya or the
intended meaning is a part of the dhdnksd or expectancy of
words. ^ The Vedantists, however, contend that tdtparyya
in the sense of the intended meaning is not a condition
of verbal cognition When the parrot imitates such human
expressions as ‘
who comes,’ ‘
who goes, ’
etc., we cannot
say that there is any intention behind its imitative cries.

Yet we have no difficulty in understanding the meaning of

these expressions. Or, when one utters the Vedic texts
without understanding their meaning, he cannot be said to
intend the meaning which his hearers interpret out of them.
The Vedantists, therefore, urge that tdtparyya as a condi-

tion of verbal knowledge is not constituted by the meaning

intended to be conveyed by the speaker, but by the fitness
of the words of a sentence to give a particular meaning
(tatpratitijananayogyatvam). Thus the sentence ‘
the jar is

in the room ’
is fit to denote the relation of the room to the

^ Tatprfttltlcchaya uccantatvam tatparyyajfjanam ca vaky&rthajflane hetul^ etc.,

Tattvadiptkd, p 68 Vtde also BP. & SM 84 ,

* Vtde SM., 84. Vide also Kuppusw^mi Sistri, Primer of Indian Logic, p. 836,

47— (1117B)
370 NYiYA THEORY OF KNOWLEDGE

jar, but not to the cloth. In the case of equivocal words,

like saindhava, etc ,
which may have more than one meaning,
we are to say that the tatparyya lies in their fitness to

yield a "particular meaning in the absence of some other

intended The word saindhava is fit to mean
meaning.
salt in the absence of any intention to mean the horse If,

however, it be used to mean both salt and horse, we are

to say that it has the fitness to mean both in the absence of

any intended meaning other than the two. Thus while the
\ edantists admit that tatparyya is a necessary condition in
the understanding of words or sentences, they reduce it to
the fitness of words themselves to give a particular meaning
apart from the intention, if any, of the speaker ^

It IS to be observed here that the difference between the

Nyaya and the Vedanta conception of tatparyya is ultima-
tely due to their different notions about the meaning of
words. For the Vedantist and the Mimamsaka, the
primary meaning (sakyartha) is a power inherent in words,
while for the Naiyayika it is imported into the words by the
intention of the person who uses them Hence the Vedan-
tist’s idea of tatparyya is vitiated by the initial assump-
tion that the fitness of a word to mean something is an
independent thing by itself, that it is a sakH or power
inherent in the word, but distinct from both the word and
the object denoted by it He is thus led to think that
tatparyya as a condition of verbal knowledge is constituted
by the inherent fitness of words to convey a particular
meaning independently of the will or intention of the speaker.
A word, however, is a significant sign or symbol. It acquires
a meaning or significance in so far as it is ‘consciously
designed to stand for A newly coined word
something.’
ISsuch a sign used by some one to signify something. We
understapd a word when we know what it is that a person

T F.d«VP,Ch IV.
 OF SFNTBNCES 371

using it means to signify, otherwise we misunderstand it.

As Dr. Stebbing’ observes: “A hearer understands a

word used by a speaker when he is referred to that which
the speaker intended to indicate to him.” But for the

speaker’s intention a word cannot have different meanings

in different contexts. Hence we cannot ignore the aspect
of intention in the meaning of a word. In fact the
Vedantists have to recognise it in the case of equivocal

words which may have two meanings if it be so intended

by the speaker or the writer It is also indirectly admitted
by them when they say that the tatparyya of a word
depends on the context {prakarana) in which it is used.

2 . The meaning of a sentence

A vakya or a sentence is a combination of padas or

words, which conforms to certain conditions. Just as words

mean objects, so sentences mean the relations of objects.

A sentence bears a certain meaning like the constituent

words. Hence the question here arises ’

How are the

meanings of the separate words constituting a sentence

related to that of the sentence as a whole ? Is the

meaning of a sentence merely the sum of the meanings

of its words ? Or, is it something new, but determined

by the meanings of the component words ? Or again,

does a sentence convey a meaning of its own independently

of the words constituting it ?

One theory of the relation between the meaning of a

sentence and those of its constituent words is known as

ahhihitdnvaya~vQda. According to it, the meaning of a
sentence is merely the synthesis (anvaya) of the meanings
of the separate words composing it When we read or

hear a sentence we have first an understanding of the

1 Logic m PracUcet p 66
3^2 NYiYA THEOttV OF KNOWLEt)QE

separate meanings of the words one after the other. Then

by putting together the meanings of all the words, according
to their expectancy, proximity, fitness and intention
{dkanksa, sannidhi, yogyatd, tatparyya), we arrive at the
construed meaning of the whole sentence. On this view,
then, the expression of the meanings of words precedes
the construction of a sentence, i e. there is a construction
of the meanings as expressed in the words {abhthitdnvaya)
As to how the different meanings, which are successively
expressed by the words, are put together, we are told that
it IS by means of memory. We understand the meanings
of the words successively ; when we come to the
but
last word of a sentence we remember the meanings of
all the preceding words. The meaning of the last word
being combined with those of the preceditlg words by
means memory, we have an understanding of the meaning
of
of the sentence as a whole. The theory of ahhihttdnvaya
is advocated in the Nyaya, the Bhatta Mimain^a and the
Vedanta system. It is generally supported by the follow-
ing reasons. If the words of a sentence have no
separate meanings of their own, then the classification of

words into nouns, adjectives, verbs, etc ,

becomes meaning-
less. Further, in every case in which we are to understand
the meaning of a sentence, we must first understand the
meaning of its component words. Without a previous
understanding of the words no one can understand the
meaning of a sentence. Moreover, if the meaning of a.

sentence were quite independent of the meaning of its

constituent words, then any sentence could convey any

meaning. Lastly, when we understand the meaning of

a new verse, we do so obviously on the basis of our

knowledge of the words and their separate meanings. This
cannot be explained by any understanding of the sentences,
since they are new and unintelligible to us. So it is con-
cluded that the meaning of a sentence is just the synthesis
 OF SBNFENCFa 373

of the separate meanings of its words/ Russell subscribes

to this view when he observes that ‘
a sentence may con-
sist of a single word, or of a wink ; but generally it consists
of several words. In that case it has a meaning which
is a function of the meanings of the separate words and
^
their order.’

Another theory of the relation between the meaning of

a sentence and those of its constituent words is known as
anvitahhidhana-mda. According to it, the meaning of a
sentence is not merely the aggregate of the separate
meanings of its constituent words The sentence has a
unitary meaning ot its own which cannot be resolved into
the complex meaning of its words Every sentence means
an action (kriyartha) It either commands or forbids us
to do something Hence the kriya or the verb is the
central unit of a sentence. All the other words of a sentence

develop or particularise the action which is the central mean-

ing of it The constituent words possess meaning only as

they are related to the action meant by the sentence. Thus

in the sentence ‘
bring the cow,’ the word cow means,
not the cow as such, but as the object of the verb bring.

Hence in a s entence there is firs^a construction {anvay^

of__the words with one another and then an expression

{ahhidhana) of the construed meaning of the whole sentenM,
i.e. there is an expression of the construed meaning
(anvitahhidhana) The theory of anvitdhhidhdna is advo-
cated by the Prabliakara Mimamsakas and the grammarians.
There is, however, some difference of opinion between
them with regard to the function of the words in the
construed meaning of the sentence. According to the

grammarians, the constituent words have no separate

meanings of their own They convey only the integral

1 Fid* TB., p. 14 : NM, pp. 396-96 ,

Vwaranaprdtneyasarhgraha, pp. 267 f. Vide

aldO VP., Ch. IV ,

8D., p. 163.
» Vide Outline of Philosophy, p. 266.
374 NYAYA THEORY OF KNOWLEDGE

meaning of the sentence in different ways and degrees

Hence the words lose their individual meanings in the
unitary meaning of the sentence. The Prabhakaras, on
the other hand, maintain that the words convey both their
separate meanings and the construed meaning of the
sentence Just as in a machine the parts perform their
respective functions and at the same time contribute to

the function of the whole, so the words in a sentence present

their individual meanings till these are construed into the
unitary meaning of the sentence Hence the meaning
of a sentence is neither the aggregate meaning of the woids
nor is it quite independent of their separate meanings.
Rather, the sentence is a new combination of the individual

meanings of the words and, therefore, conveys a irma

meaning The Prabhakaras agree with others in holding

that the combination of the separate meanings of the words

IS effected by memory, since the woids appear in succession

and their meanings are only remembered by us when we

*
come to the end of the sentence
Of the different views about the meaning of a sentence,
that of the Prabhakaras seems to be the best If the mean-
ing of a sentence be, as the grammarians think, quite
independent of the w'ords, then we can have no other way
of knowing it than a personal explanation from the speaker
or writer of it. If, on the other hand, its meaning be
merely the aggregate of the word-meanings, we do not
see how any sentence can convey a new meaning to
meet a new situation. If the word-meanings are
not modified in the meaning of the sentence, according
to its context, no sentence can go further than the
old meanings of its words. On the other hand, without
something of their old meanings persisting in the words,
the new meaning of a sentence cannot be understood by us.

^ NM., pp 387-98, Vivaranaprameyasarhgrahaf^p 257-60.

 OF SENTENCES 375

Hence we are to say that in the meaning of a sentence the

separate meanings of the words are so modified as to fit in
with the context in which the sentence is used/ That the
meaning of the sentence dominates the meanings of its

words will appear also from the fact that in the life of the
individual the judgment precedes the separate concepts
related in it. The child makes assertions about objects
before he understands the separate meanings of words. It
IS the sentence and not the word that is the starting-point
of oui thought Hence the meaning of a word
and speech
should follow the meaning of the sentence in which it is
used. The meaning of the sentence is not a function of the
meanings of the separate words, rather, it functions in and
determines the meanings of its words. This appears from
the fact that we cannot ascertain the meaning of a word
unless we know the sentence in which it is used.

3 The import of sentences

For the Naiyayikas, a sentence is the verbal expression

of determinate knowledge (savikalpakajhana). It is only
determinate knowledge that can be conveyed by a sentence.
Indeterminate knowledge (ninikuIpakajMna) cannot be
expressed in w'oids or sentences. Now determinate know-
ledge IS the knowledge of a thing as qualified by an attri-

bute {visesanavi^esydvagaht) In it we know something to

be related to something else as substantive to adjective.

Hence a sentence as the verbal expression of determinate

knowledge must contain two terms and express a relation

betw'een them. Of these two terms one is called uddeSya

or the subject about which something is asserted. It is also

^
Cf Schiller, Logic for Use, p 66 ‘A successful transfer of meaning has to
satisfy iwo conditions. (1) It has to presuppose and respect old meanings and to
era pl(iy old truths, but it has also (2) so to arrange them in their contexts as to

develop ne^ meanings out of them, in order to express new truths,’

 376 NYiYA THEORY OF KNOWLEDGE

called viiesya or the substantive which is regarded as the

seat or locus of some quality. It may tlius be called the
determinandum or what is presented to be determined and
characterised by thought. In relation to it, the other term
is called vidhcya or that which is asserted about the
subject and is therefore a predicate It is known also as
the vUesana or the adjective which is referred to the subject.
It IS that w'hich determines the subject and may thus be
called the determinans in relation to it. On this analysis of

it, a sentence corresponds to a proposition in Western lomc.

But unlike the propositions of Formal Logic, the sentence
has no need for a copula. That there must be a copula or
a verb in a sentence is not admitted by the Naiyayikas
and many other Indian thinkers. The analysis of a sentence
into the subject, the predicate and the copula is repudiated
by the Naiyayikas as utterly groundless.^ We can very well
express a complete meaning without the copula, as when
we say “ parmto vahmman.” That the copula, as some
form of the verb ‘
to be,’ is not an essential part of the
proposition is also recognised by modern logicians like

Bradley,^ Bosanquet'* and Johnson.^ The Naiyayikas go

further than this and hold that no verb is necessary for a
sentence. It may be said that a verb is implied, if not
expressly mentioned, in a sentence. When we say ‘
a
fiery hill,’ or ‘
a red colour,’ we imply the verb ‘exists’
or ‘
IS ’. For the Naiyayikas, however, such verbs stand for

a subjective mode of our assertion, but not for any part of

the asserted fact or content The ‘hill as fiery,’ or the

^ Enyarahitain na vakyaCTiastltyadikaatu pracam pravado niryuktikatvada6radd-

heyat;t, SahdaSaHuprakd^ikd, p 28
2 Principles of Logic, Vol I, p 21
3 Logic, Vol I, p 81.

^ Logic, Pt I, pp 30-11
® 6’/ 8 H Mellone, Introductory Text-Book of Logic, p. 10 ‘
There is no separate
existence in thought corresponding to the separate existence of the copula in the typical

proposition, S is P.*
 OF SEKTENCEIs 377

‘
colour as red ’
is the content of our assertion. The verb
is or exists stands for no objective content. Hence a
sentence does not require a verb as an essential part of
its content. So .also we may have a proposition without
the copula or the verb ‘
to be.’ But we should observe
that although the sentence as a predicative judgment
(viSesvaiUesa navaadhi) corresponds to a proposition, yet it

is in itself wider than a There are sentences proposition.

which do not express any relation between subject and
predicate, or in which there may not be any subject or
‘
predicate, e.g. a dog runs,’ go there,’ etc. These are
‘

sentences, but not propositions expressing a relation betweer

two terms. The Naiyayikas, however, take the sentence
as equivalent to a proposition.
It will appear from the above that, according to the
Naiyayikas, the import of a sentence or proposition is the
predication of an attribute with regard to some thing or
things. It expresses the relation between a substantive and
an adjective (viSesyaviiesana) The substantive is some
thing or real, while the adjective is some other fact or real

found in relation to it. Hence we may say that both the

subject and the predicate are real facts forming one complex
whole. The proposition does not bring the one intc

relation with the other, but finds them as related. The

Naiyayikas, therefore, cannot agree with Bradley ^
and
Bosanquet “ who hold that a proposition is the reference ol

an ideal content to reality, or that a proposition characterise!

some part of reality, with which we are in immediate

contact, by referring an ideal content to it. For them,
the predicate is not an ideal content but a real fact. In the
proposition ‘
the ball is red,’ the redness is as much a

perceived fact as the ball, and so also their relation is no'

1 Prtnoplrs of Logic, Vol T, p 1,0.

^ Logic, \on,v S3.

48--(1117B)
378 NYIYA THEORY OF KNOWLEbGB

ideal but an actual fact. They do not accept the subjective

view that a proposition expresses a relation between two
ideas, or the idealistic view that it is the reference of an
ideal content to reality. As radical realists they are in
favour of the objective view that the proposition expresses a
real relation between two facts or reals. This naive view
of the Naiyayikas has been ably supported by Mr. Gotshalk’
who opposes the idealistic view and shows that ‘
the subject
of an ordinary judgment is not Reality itself but merely
and simply that limited situation within Reality engaging
attention,’ i.e. a finite and limited reality. So also what
IS predicated of the subject is some real fact, a thing or
quality, etc., and not a mere piece of meaning or an ideal
content referred by a judgment to an existent reality.

The above view of the Naiyayikas that all propositions

express the subject-predicate relation betw'een a substantive
and an adjective has been opposed by the Mimamsakas, the
Vedantms and other logicians. According to the gramma-
rians and the Piabhakaras, •
eveiy significant propo'^ition
means an action. If a proposition is to give us any new
knowledge, it must not relate to matters ot lact
(siddhapadartha) for these may be known by means of
perception and infeience. On the othei hand, the knyd or
the verb is the central unit of a sentence or proposition.
The subject and the predicate have meaning only as
they are related to the verb by the nominative and
objective cases. Hence the import of a proposition
lies, not in the subject-predicate relation between two
terms, but in the action denoted by its verb. Every propo-
sition expresses a command and is, therefore, an imperative
proposition. According to the Advaita Vedantins, " all pro-

^ Vide Mindf Jac ,

2 Vide VnaranaprafneyasaiHgraha
pp ‘257-68
3 Vide VP , Ch. I.
 OF SRNTENCRS 379

positions cannot be brought under the subject-predicate

form. Identity propositions, like ‘
this is that man,’ cannot
be construed according to the subject-predicate relation.
These propositions do not express any relation between two
things, but the simple identity of a thing with itself. We
cannot say here that ‘
that man ’
is the predicate or adjec-
tive of ‘
this man.’ These are non-relational and therefore
non-predicative propositions. Russell ^
also opposes the
view that all propositions are reducible to the subject-predi-
cate form. He thinks that the propositions which assign
the qualities of things come under this form, e.g. “ this
thing IS round, and led and so on ” On the other hand,
the propositions which express relations cannot be reduced
to the subject-predicate form Thus in the propositions
‘
A is like B,’ ‘
B IS the brother of C,’ ‘
C is greater than
D,’ we cannot say that the terminal term is predicated of
the initial term. They express respectively a symmetrical,
a non-symmetncal and an asymmetrical relation between
different terms, of which one cannot be regarded as the
quality of the other

When we consider the different views about the import

of propositions, we are led to think that a distinction should
be made betw'een predicative and non-predicative proposi-
tions. In a predicative proposition a subject is related to a

predicate as substantive to adjective. All propositions,

however, are not predicative in this sense There are many

propositions which cannot be brought under the subject-

predicate form. Thus Russell’s relational propositions, ‘

A
is like B,’ ‘
C is greater than D,’ do not conform to the

subject-predicate form. It may be said that these propo-

sitions are predicative because in the one ‘

likeness to B ’
is

predicated of A, and in the other ‘

being greater than D ’

» OtiT Knowledge of the External World, pp. 46-60,

3S0 NYiYA THEORY OF KNOWLEDGE

IS predicated of C. This will mean that A has the attribute

of being like B, and C has the attribute of being greater
than D. But ‘
being like B ’
is not a quality of A, nor
‘
being greater than D ’
of C, in the same way in which
the red colour is a quality of the ball. Similar!}, the
Vedantm’s identity proposition ‘this is that man’ can
hardly he reduced to the subject-predicate form For the
Naiya}ikas, this proposition is predicative in so far as it

means that ‘
this man is characterised by a past existence.’
In it a man’s existence at some other time and space is

predicated as a character of bis present existence. Although

the proposition may be interpreted in tins wa), yet it loses

its real force when so interpreted. The proposition

expresses a judgment of recognition {pr(ityabhi]na) In
recognition we are primarily interested in the identity

of aman from the past to the present. To recognise

a man as that Devadatta ‘ ’
is to know not only that
he was known before, but that be is identical in the past
and the present. Hence the proposition ‘
this is that
man ’
does not characterise a man by his past existence
and 18, therefore, non-predicative Finally, the sentences
which mean action cannot be called predicative proposi-
tions by any stretch of imagination. The sentence ‘
a
dog runs ’
is not a predicative proposition, because there
IS in it no subject-predicate lelation between two terms.
To make it predicative it may be converted into the logical
form ‘
a dog is a running animal.’ But this form of
the sentence does not bring out its real sense. It is an
altogether different proposition, and a false proposition
too, lor dogs do not always run. Similarly, sentences
expressing commands or imperatives are not predicative
propositions in any sense or form. ‘
Thou shalt not steal,’

‘ pray to God ’
are sentences which enjoin certain duties on
us, but do not assert any relation, predicative or otherwise,
between two ideas or things.
 OF SENTENCES 381

4. 3abda as an independent source of knowledge

According tg many schools of Indian philosophy, iahda

or verbal testimony is an independent pramdna like per-

ception and inference. As we have already seen, there is

some difference of opinion among them as to the nature

of Sahda or verbal testimony. There are two main views
with regard to it. On the one hand, the Jainas^ and the
Naiyayikas take ^ahda as the statement of a perfectly
reliable person. In this sense, hbda as pramdna means
a sentence which is spoken or written by a trustworthy
person, or the statement of some authority. On the
other hand, hbda as a pramdna is taken by the
Mimamsakas ® and the Vedantins '
to mean a sentence or
pioposition whose import is not contradicted in any way.
On this view, a sentence as the significant combination
of words, according to the four conditions of expectancy,

compatibility, proximity and fitness, is pramdna or a valid

source of knowledge.

Let us now consider whether sabda can in any sense be

regarded as an independent method of knowledge {pramdna).
This question resolves itself into two other questions. The
first question is : Can ^ahda give us a true knowledge of
objects ? If it can, it will have to be regarded as a pramdna
or source of knowledge. Then the second question will be
this : Is the way in which hhda gives us a knowledge ol

objects different and distinct from perception, inference and

the rest ? It does not matter if the same objects can be
known by perception or inference. So long as we cannot
reduce ^abda oi the verbal knowledge of objects to the condi-

1 Aptena praijitam Tacananiaplavacanam, Prameyakamalamarian^a, p 112.

2 Cf. Padarthabbidhanadvarepa >adv&ky&rtbavijflaDam taccb&bdaih n&n^a
pramftQiam, dD., p 72.
3 Vfde VP., Chap IV.
38’2 NYiYA THEORY OF KNOWLEDGE

tions of any other kind of knowledge, we must recognise it

as an independent method ot knowledge.

The first question deserves an affirmative answer. 3ahda
or testimony gives us true knowledge about many things.
The Buddhists, however, contend hbda winch consists
that
of words cannot give us any knowledge. Words are physi-
cal objects and cannot, therefore, take the place ot an organ
of knowledge like the senses or the reason. Further, there
IS no part or aspect of reality nhich cannot be known by
perception or infeience, and for which we require a different
method like Mbda or testimonj. This contention, however,
rests on a misunderstanding Just as the same objects may
beknown by perception and inference, so they may be
known by hbda or testimony- Or, we may say that ^abda
has to do with supraraundane realities which cannot be
known by perception oi inference Further, words as
physical sounds or coloured figures do not give us any know-
ledge of facts '
3abda as the understanding of sentences or
propositions gives us more knowledge about the world than
perception and inference A man’s knowdedge w'ould be
very meagre if be were to depend solely on his own
experience and leason. The bulk of our knowledge
comes from the testimony of our fellow beings, e.g books
and speeches. We accept on trust by far the greater part
of what we hold to be true Hence it is established that
hbda does give us true knowledge of facts and is, there-
fore, a pramdna or source of true knowledge.

Turning to the second question, we ask if ^abda oi verbal

testimony can be reduced any other method or form of

to

knowledge. While standing on the bank of an unknown

river I am told b) a local gentleman “ This river is :

fordable.” I know the depth of the river from this state-

ment. Can this knowledge be explained by perception, or

I
Ytde NV. A NVT., 2.1.41).
 OF SENTENCES 383

any other source of knowledge ? It cannot be a case of

perception, because I cannot directly see the river’s depth
nor measure it before going into its water. It cannot be
explained by memory, for there is no previous experience
corresponding to my present knowledge of the river’s depth.
I cannot now remember that the river is fordable because
I have not perceived it to be such in the past. It cannot
be said that m> present knowledge of the river as fordable
is the result of the synthesis of my ideas of a river and of
fordabihty acquired from the previous experiences of other
fordable rivers. Even il I have such ideas or memory-
images from previous experience, they will not explain my
knowledge of this river as fordable, because there is no
^
previous experience in relation to it.

Next we are to consider whether sahda or verbal testi-

mony can be reduced to inference It has been held by

many thinkers, both Indian and European, that knowledge
from testimony is really a foim of inference. The Buddhist
logicians hold the generally accepted view that testimony
IS a kind of inference, because in it we infer the truth or
falsity of a statement from the character of the person who

makes that statement But this view makes a confusion

between two different questions To determine whether
testimony is a separate source of knowledge or not, we are

only to see if it gives us a true knowledge of facts, and not

how its truth is known oi tested by us We can very well
know the meaning of a sentence even before we enquire
into its source, or when its source cannot be known. In
fact, testimony is the source of the greater part of our
knowledge of the world. Thus the Buddhist contention
falls to the ground. - The Vaisesikas try to reduce testimony

1 So’yam niyantiitarihaUaDDu pratyaksam na canuma na casaii smirtil)

saruanakara-samskaraprabbavatvat, SahdaMii prahd.silai pp 3-4

2 Vide NV., 117. Cf

*
Vakya^iavapantarnmeva liyaptanaptajfiananapek^aueva
padfirtbairvdkyartho ’vagaiuyate, SD , p. 73.
 884 NYAYA THEORY OF KNOWLEDGE

to inference on more plausible grounds. ^

According to them,
knowledge from testimony is governed by the fundamental
law of inferential reasoning. Just as in an inference we
know an unperceived fact from the perception of a sign that
is universally related to it ({vydptUdli-hnga)), so in testimony
we have the knowledge of some unperceived facts from the
perception of the words by which they are denoted. I hear
the sentence ‘
there are five trees on the river-bank.’
With this I have an auditory perception of a number of
words. I know that each of these words has a fixed and
universal relation with the object meant by it Hence to

know these words is, for me, to know the objects denoted by
them, just as to know smoke is, tor a man who knows the
universal relation between smoke and fire, to know the
existence of fire in relation to it. In testimony our know-
ledge about facts is brought about by the knowledge of

words as signs or middle terms {hngd] and that of their

invariable relation (lydpti) with those facts. Hence knnw'-
^
ledge from testimony is really inferential in character.
But this attempt to reduce testimony to a kind of inference
reminds us of the Procrustean method. It violently turns
and twists the nature of testimony to make it conformable
to the standard of inference, and yet it cannot succeed.
Inference depends on the knowledge of vynjdi which is a

natural relation of coexistence between two things, e g.

smoke and fire. No knowledge of the meanings of words

is necessary to infer the existence of fire from smoke. From
the mere perception of smoke in the hill we know that there

must be fire in it. If there were such a natural relation of

coexistence between words and their meanings, then an
illiterate man should have had as good knowledge from
words as any man of letters. Fuither, the elation hetwwn i

^ ^abdopaiDanayornaiva pithakpraniapyairiisynte, anuiiiftniigatfirthatvadiii vai<e-

^ikam matam, BP ,
140*41.
2 6abdo 'numaDam vyaptibalenarihapratipadakatv&ddhuiiiavat, NK , p 218.
 OF SENTENCES 385

the middle and the major term of an inference is such that

if the middle is present in a certain locus, the major also
must he present in it. If there were such a relation between
words and the objects denoted by them, we should expect
to find the objects in the body of the person or the book in
which the words occur In truth, the relation between
words and their objects is quite different from that which
holds between the middle and the major term of an infer-
ence. Words mean certain objects, but do not coexist with
them. So also, the conditions of verbal knowledge or testi-

mony are specifically different from those of an inference.

It IS true that both depend on a mental construction of
certain given data But in inference the construction is

limited to only three terms and proceeds according to their

relations of inclusion and exclusion. In verbal knowledge
there is a construction of the meanings of any number of
words constituting a sentence, according to their syntactical
*
expectancy, propinquity, mutual fitness and intention
Even if our understanding of the meaning of a word may be
said to be conditioned, like inference, by a fixed association

between the two, we cannot speak of any fixed relation

between a sentence and its meaning. A sentence conveys
different meanings according to its different constructions.

Its meaning depends on such specific conditions as the

expectancy, fitness, proximity and intended meaning of

words, which are not to be found in inference or any other

kind of knowledge. Finally, the evidence of introspection
(anuvyavasaya) confirms the result of the logical analysis

that testimony is distinct fiom inference. It clearly shows

that we cannot detect any inferential process in the know-
ledge from testimony. When from the sentence ‘
the cow
exists ’
I know that a certain cow does exist, I feel that I do
not infer the existence of a cow, but understand it from a

1 Yogyatarthagata ’kafik?a iabdanis^ba ’nubbavika, pratyekam vft militva v5 naite

Sabda^aktuprakahka kar 4
liAgainaBiddhitalji,

49—(1117B)
 386 NYiYA THEORY OF KNOWLEDGE

sentence. *
Hence we conclude that testimony is a distinct

method of knowledge which cannot be reduced to any other

method.
Some Western thinkeis now recognise testimony as a
separate source of knowlege. Thus Eussell ^ frankly admits
that ‘
testimony is essential to science, although it is open
to cnticism by the sceptic.’ So also Montague ''
thinks that
‘
testimony that is open to free and honest study remains as
legitimate a souice of knowledge as any other.’ He, how-
ever, holds that testimony is not a piiniary but a secondary
source of knowledge For him the weakness of testimony
consists first in the fact that authorities conflict But on his
own admission this difliculty is not peculiar to testimony ;

it IS present in each of the other methods. There are

conflicting perceptions and inferences, like conflicting testi-
monies. Hence this cannot be a ground of distinction
between one method as primary and another as secondary
The second and more serious source of weakness in testi-
mony IS, for him, its dependence on some other method for
establishing its validity When questioned as to why we
should accept a given authority, the answer must be that
the authority knows the truth direct through some other
method like experience or intuition. This shows that the
truth of testimony ultimately depends on the direct
experience, reason or intuition, of some person In testi-
mony the individual’s relation to truth is not direct but
indirect, for it depends on the direct knowledge of a second
individual, while m sense and reason the individual is in
direct relation with the truth of things Hence testimony
' Vaitoto asttnena gamanutninomltyaderanuvyavasaynsya latrftaaltvat, pratyuta
gaurastUi vakyadastUvena gaulj ilruto na tvanumita ityevanubliavacca, MdMti-
prakaitkd, p 7
2 Outline of Philosophy p 6
» The Ways of Knowing, Chap I It may be noted here that L 8 Stebbing algo
inchnea to the view that while tegtimony IS a source of bumao knowledge, it is not an
independent scarce, but can be reduced to inference {vide Logic in Practice,
p 102)
 OF SENTENCES 38?

cannot be a primary and an ultimate source of knowledge

like the other methods. According to the Naiyayikas also,
the validity of the knowledge derived from testimony
depends on the reliability of the person who is its source.
So also the truth of testimony is to be proved or tested by
successful activity or verification in direct experience. Still

the Naiyayikas recognise testimony as an independent source

of knowledge like perception and inference
The Mimamsakas and the Vedantins go further than the
Naiyayikas and hold that the truth of the knowledge from
testimony is both constituted and known by itself. By
testimony they mean a significant combination of ideas
expressed by words, according to their expectancy, com-
patibility, propinquity and fitness It is a sentence in
which the ideas expressed by the words are consistent with
one another and also with the facts denoted by them. The
meaning conveyed by a sentence is not only consistent in
itself but also with the facts of experience. As such, it

naturally leads to a knowledge of the truth. The truth of

the knowledge derived from testimony is thus constituted
by its own intrinsic conditions. And truth is, wherever
it IS, known by itself. It is a self-evident character of
knowledge and requires no other test than itself in order to
be known as true. For example, a true perception is by
itself known as true. Similarly, the knowledge derived

from a proposition or sentence is true and known as

is true,

if there be no ground to contradict or doubt it. The truth

of testimony is both constituted and established by itself.

Although we do not go so far as to say with the

Mimamsakas and the Vedantins that testimony has self-

evident validity, yet we find no reason to deny that it is an

independent or ultimate source of knowledge. Whether a
certain source of knowledge is independent or not depends
on two things : (i) whether it gives us a true knowledge of

facts, and (m) whether its conditions are distinct from those
388 NYAYA THEORY OF KNOWLEDGE

of any other source. It does not matter if the truth of the

knowledge is constituted or ascertained by some other source.

Thus the validity of an inference depends on the validity

of our perceptions of the middle term and its relation to

the major term. So also the tiiith of its conclusion is

tested by verification in direct experience. Still no one

denies that inference is a primary and an ultimate source of
knowledge. If it be so, why should we not recognise testi-
mony also as an ultimate source of knowledge ^ We have
already seen that it gives us a true knowledge of facts in a
way distinct fiom any other way of knowing them. If

testimony depends on perception to prove its validity, per-

ception also depends on inference to prove its own validity

when that is doubted or questioned by any one. The

dependence of one method on another for its pioof or
verification is a difficulty, not peculiar to testimony, but
common to all the methods of knowledge. Further, there
are certain crucial instances in which we cannot go beyond
testimony and prove its truth by some other method. If

to the question as to why a given authority should be

aceepted, the almost inevitable answer is, as Montague
thinks, that the authority possessed a direct knowledge of
the truth, what should be our answer to the next question
as to how we know that he had a direct knowledge of the
truth. Here we have to depend on the statement of the
authority himself. We do not require any other proof of
his direct knowledge of the truth. Similarly, we learn that
a name denotes a class of things from the testimony of our
elders and they from their elders and so on indefinitely
without there being any direct knowledge of the word’s
meaning on the part of any one. Even in the case of the
man who first used the name in relation to certain things
or one who first found it to be so used, we cannot speak of
a direct knowledge of its meaning, for the name could not
be perceived like a label attached to those things. Thus
 OF SENTENCES

ue see that, like perception and inference, testimony does

not always require to be proved by direct knowledge,

although it may be so proved when necessary We should

therefore recogni'se it as an independent and ultimate source

of knowledge like perception and inference.

389
 CHAPTER XX

OTHER SOURCES OF KNOWLEDGE

1. Different vieies about the ultimate sources of hnowledge

There is much difference ot opinion among Indian

thinkers as to what the ultimate sources of liuman know-
ledge are For the Carvakas, who are radical empiricists,
perception is the only valid source of our knowledge and all

true knowledge comes from peiception. The Buddhists

hold that perception and infeience are the two ultimate
sources ol true knowledge, which include other sources like
upamana and hhda According to the Samkhya and the
Yoga s}siem, sabda oi veibal testimony also should be
recognised as an independent source of knowledge like

perception and inference. The Samkhya includes upamana,

arthapalti and sambhaia uudei inference, and ahhma under
perception. ^
The Naiyayikas aie in favour ol the \iew that

there are lour independent sources of -laiowledge, namely,

perception, inference, testimony and upamana oi com-
parison, According to them, the other sources ol knowledge
may be included within these lour and so need not be
taken as ultimate or independent sources of knowledge
According to the Vaisesikas, there are four kinds of vidya
or true knowledge, namely, perception, inference, memory
and intuitive experience (drsajwdnfl). They include sabda,
upamana, arthapatti, abhnva, sambhava and aitihya within
inference. But there is some difference of opinion as to

whethei all the four kinds of ridyd or knowledge are

' TKD ,
kar 5, Yogu-iulr<i,l.'i
* NK.,pp 213-31.
 OTHER SOURCES OF KNOWLEDGE 391

independent sources of knowledge or not. It will follow

from the definition of vidya as definite knowledge which is

free from doubt and contradiction that all kinds of Didyd are
pramdm or independent sources ot knowledge. S5ridhara m
his Nydyahandall '
at first tells us that perception, inference

and smrti or memory are treated first because the> are

lauhiha pravidna or ordinary souices of knowledge, and
then ar.sa because it is an extraordinar} source of knowdedge.
But in the course of the discussion on smrti he observes
that it IS not a pramami because it depends on previous
experience to -give us knowledge of past objects The
-
Nydyallldvatl. compendium ol the V.usesika philosophy,
a

establishes the view that imrB or memory is an independent

source ol knowledge like perception, inference and intuitive
knowledge. In the later works ol the jS'yaja-Vaisesika
philosophy, however, it is generally maintained that the
Vaisesikas accept only perception and inference as two
independent sources of knowledge ’
According to the
Jamas, perception, both ordinary and extraoidinary,
inference, testimony, pratyabhijm or recognition and smrti
or memory are all independent sources of knowledge,
although they may be classified under the two heads of
pratyaksa and paroksa, immediate and mediate knowledge. ‘

The Prabhakaras hold that aithdpatti or postulation should

be accepted as a separate source of knowledge like peiception,

inference, testimony and comparison. The Bfiattas and
the Vedantins add anupalabdhi or non-perception to these

five and maintain that there are six distinct sources of

knowledge. The Pauramkas go further than this m holding

that sanibhava or piobability and aitihya or tradition also
are to be recognised as separate sources of knowledge like

1 Fidepp 186,267.
* Soiftirapi manintarameva, etc , NL , p 67 (Bombay Edn.).
s
Vtde TR., p 60 , TK , p 7
< Vide TTS , 1, 10-13.
 392 NYAYA THEORY OF KNOWLEDGE

the six pramanas admitted by the Bhattas and the

*
Vedantins.
Now the question is tins : How many independent
sources of knowledge are we to accept^ The Naiyayikas
accept only four sources of knowledge as distinct and
independent These are perception, inference, comparison
and testimony as explained and discussed before What
then are they to say with regard to such alleged sources
of knowledge as aitihya, sambhara, ahhai a or anupaJahdln,
arthdpatti, smrh, pratyahhipm and dnapldna? Accoiding
to them, drsap'idna or intuitive knowledge is a kind of
extraordinary {alanlika) perception, while pralyahhpnd is

only a kind of qualified perception.^ These two come under

perception as an ultimate source of knowledge and are not
themselves separate sources of knowledge. As regards
aitihya or tradition, the Naiyayikas hold that it is a kind
of testimony, of which the source is not definitely known.
’

Tradition means the continuous communication of a body

of ideas and beliefs from one generation to another It has
its origin in no living individual, but is enjoyed by all

individuals as the common property of the race. Now the

body of ideas and beliefs constituting a paiticulai tradition

IS accepted as true on the authority of some person or

persons, whoever they may be. We believe in tradition

because we are pretty confident that it must have originally

emanated from some reliable persons As such, tradition
is a form of vague testimony, in which w’e know certain

things on the authority of some unknown persons Similarly,

samhhava may be included within inference. It may be
taken to mean either probable knowledge or the knowledge
of numerical inclusion. In the first sense it is illustrated

when we expect rain from the appearance of clouds in the

1 VideTB-.v 66
2 Vide Chaps IX & X ante
3 Aiiirdi§tapravaktj*kain pravaduparamparyam aitihyam, NB ,
2 2 1
 OTHER SOURCES OF KNOWLEDGE 303

sky. Here we think that there will probably be rainfall

because we know that clouds are generally followed by
rain. But such probable knowledge is not pramd or valid
knowledge and so requires no pramdna or source of
knowledge to explain it.^ In the second sense, however,
samhham ^ means the knowledge of the part from that of
the whole within which it is included. Thus we know
that there is a hundred within a thousand, a seer within
a maund. Such knowledge is really inferential in character,
since it depends on the knowledge of vydpU or invariable
concomitance -between the part and the whole. Hence
samhhava need not be taken as a separate source of
knowledge other than anumdna or inference. The question
as to whether arthdpath, abhdva or anupalabdhi, and smrti
should be recognised as separate sources of knowledge or
not will be separately considered in the following sections.

2. Arthdpatti or postulation as a source of knowledge

Arthdpatti as a source of knowledge consists in the

supposition of some unperceived fact in order to explain
a given fact. When a given or perceived fact cannot be
explained without some other tact we have to presuppose

or postulate the existence of this other fact even though we

do not perceive it. A phenomenon is presented to our
experience and we find that there is a seeming contradiction
involved in it. We try to get over this contradiction by
supposing some other fact which explains away the

contradiction. The given fact which is to be explained is,

called the upapddya, and that which explains it is called |

the upapddaka. Hence here we proceed from the knowledge"

of something to be explained to the knowledge of that

^ PrftcQrasfthacaryassxDvedaD&t buddhirabadhiti saiftbha'va^^, 6tc.| NL • p. 57.

* Saihbhavo Dfima avinftbbftvmo rtbasya aattagraha^adanyasya sattftgrahaijaiu ..

tadapyanuminaineva, NB., 2 2 1*2

00— (1117B)
394 NYIYA THEORY OF KNOWLEDGE

which explains it, i.e. from the consequence to the ground.'

Thus when a man who is growing fat says that he fasts,

we find an apparent contradiction between his increasing
fatness and his fasting. We get out of this contradiction

by the supposition that the man eats at night, because a

man who fasts at day cannot grow fat unless he takes food

at night. Or, a man, who is living, is not found in his

house. To explain the absence of the man from his house

we suppose that he is somewhere outside his house, because

a living man cannot be absent from his house unless he lives

outside it.

According to the Advaita Vedanta and the Bhatta

Mimamsa, arthapatti is a separate source ot knowledge,
because it gives us a knowledge of facts which cannot be
otherwise explained, ft cannot be explained hy perception,
since the fact known through arthapatti is not perceived
by us. That the fat man eats at night is not a matiei of
perception for us. Nor can we explain this knowledge by
inference According to the Advaitins, arthapatti is not

an inference It cannot be reduced to anvayi inference,

because there is no anvaya or agreement in piesence between
fatness and eating at night as between smoke and lire

We cannot say that wherever there is fatness there is eating

at night, just as we can say that wherever there is smoke
there is fire. Nor can arthapatti be reduced to nyatirehi

inference, because there is no such thing as vyatirehi

inference Further, the direct report of oui consciousness
IS against the supposition that arthapatti is an inference. In
anuvyamsaya or introspection of the knowledge by aithapatti
we do not feel to have ‘
inferred ’
anything, but simply
to have supposed or presumed something in order to explain
^
something else

^ Upapadyajfl&oeoa upapadakajflanatD srthapattil;» ttc,VP,Ch V. Arthapatti-

rapi df^ab ^ruto va ’rtho’njatha nopapad^ata ityarthakalpana, SD p ,
7«
2 VP,Ch V
 OTHER SOURCES OF KNOWLEDGE '395

The Naiyayikas, Saihkhyas and others object to the

above view of arthapath as a separate source of knowledge.

According to the Naijayikas, arthapatti may be reduced to
an inference of the vyatireki type. It is not indeed an
anvayi inference in which the major premise expresses a
positive relation of agreement in presence between the
‘
middle and the major term, e.g. whenever there is

fatness, there is eating at night.’ On the other hand, it

is a vyatireki inference in which the major premise expresses

a universal relation between the absence of the major and
the absence of the middle. Thus the above example of
arthapatti may be reduced to the following syllogism :

A man who does not eat at night while fasting by day

is not fat
This man who fasts at day is fat ;

This man is not a man who does not eat at night, i.e.

he eats at night.

As arthapatti may thus be reduced (o vyatireki inference, the

Naijajikas lefuse to acknowledge it as a separate source of

knowledge.^ So also the Samkhya philosophers explain

arthapatti as a form of inference Taking the second
example of arthapatti given above, Vacaspaii points out that
•
it can be reduced to the following inference

If a living individual is absent somewhere, he is present

elsewhere ;

Devadatta who is living is absent from home ;

•
He IS somewhere outside his home.

Here a man’s existence outside his home is inferred from

‘
his absence from home ’
as the hnga or the middle term.
There is a relation of vyaptr or universal concomitance
between a man’s presence somewhere and his absence

^ Yastu Da r&trau bhui^kte nasau di\abhufl]anatve sati pIno, etc ,

TB , p. 15.
39b' NYiYA THEORY OF KNOWLEDGE

elsewhere. Every man finds this to be true m bis own

case. Hence when we know the one from the other we
simply infer it from its hnga or universal concomitant, ]ust

as we infer fire from smoke. ’

The Bhatta Mimamsakas expose the futility of the

attempt to reduce arthapath to inference They point out
certain fundamental differences between inference and
arthapath, which make it impossible for us to reduce either

of them to the other. It may seem at first view that

inference and arthapath involve the same process of

reasoning. In aithapath we pass from the knowledge of an

observed phenomenon to that of an unobserved phenomenon
without which It cannot be explained. In inference also

we pass from the observed smoke to the unobserved fire as

that which alone explains the smoke But a closer view'
of the matter reveals certain important and unmistakable
differences between the tw'o In inference we proceed from
the gajiiaha or the evidentiary fact to the garnya oi the
evidenced fact, while in arthapath we pass from the garnya
or the fact to be evidenced and explained to the yamaka or

that which evidences and explains it. Again, in arthapath

we are confronted with an apparent conflict between two
facts, e.g. a man’s fatness and fasting by day, or, a man
being alive and yet absent from home. In order to resolve
this conflict we have to presuppose or postulate another fact,

namely, that the man eats at night, or that the man has
gone out. So long as we do not make this supposition wc
are in doubt as to whether the man really fasts, or whether
he really exists or not. Such doubtful facts cannot be the
lihga or the mindle term of any valid inference. In
arthapath we get over this state of doubt and conflict by
supposing something which explains them nw-ay Henci

* Kvarnartbapattirapi ua prama^aDtaram yada khnlvavyapakab aannekatra nas

tadaoyatraaU, etc , Tattvakaumudi, p. 46
 OtHER sources of KNOWLEDGE 397

while in inlerence we pass from an undoubted fact {niicita

(jamaka) to its invariable concomitant, in arthdpatti we

proceed from a doubtful fact {samdigdha gamaka) to some-
^
thing which explains it and saves us from the doubt.
Finally, if arthapatti is to be reduced to inference, we must
show what the Unga or the middle term of that inference

is. When we argue that Devadatta exists outside his

house because he is living and yet absent from home, we
cannot take mere ‘
absence from home ’
as the middle
term, for the man may be dead and cease to exist at all.

Nor can we say that ‘

living ’
is the middle term, because
a living man may exist inside his house. Nor again can
It be said that Devadatta’s ‘
living together with his absence

in the house ’
is the middle term from which we infer his
existence outside the house. In an inference we first know
the hnga or the middle term and then, through it, the hhgi
or the major term The two are not known together, but
one after the other In the case of Devadatta, however, we
cannot connect his living with his absence from the house
except through the idea of Ins existence outside the house.
Hence to know the alleged middle term, namely, ‘
his living
together with his absence from the house ’
is just to know
his existence outside the house This being known along
with the alleged middle term, there remains nothing more
to be inferred from it So arthapatU is not the inference
of the major term from the middle term, but the pre-

supposition of one fact in order to explain another, in which

^
IS involved a seeming contradiction.

1 S^adevam, yadyaDupapannaui gamakam syat, iha tu yaonopapadyate tadev

gamyam yatha .
canam&ne m^citafii gamakara, e\auuartbapattau eamdigdbHi
gamakaroiii, etc., SD , pp 70-77
^ Na tavadKrh&bhavaioalraiii hOgain mile'pi saihbbavat, na jivauamatraii
gjho’pi sadbhavat, ato jivanasamaj'^to gjrhabhavo lifigauiiti vaktavyaiu, prathamam c

hi^gamavagamya paiicallibgyaQura&nena ibbavitavyam...atra ca na bahirbbavi

vagauaamantara^a g|*b&bbftvo ivanam oa 3aib8;*9tam pratyetum ^akyate virodlik

etc., BD., p. 78.

398 NYAYA THEORY OF KNOWLEDGE

The Bhatta Mlmanisakas next discuss the question of

reducing inference to atthapatti. It may be said that if we
accept arthapatti as a separate source of knowledge, there is

no more necessity of recognising inference as a different

source of knowledge. Inference may be sho\\n to be the

same as we can analyse an inference in the
arthapatti, for

following way. When I see smoke in the lull, I think that if

there were no fire, this smoke would be unaccounted for.
Therefore, either there is no smoke in the hill or the
universal proposition, '
wherever there is smoke, there is

fire,’ IS false. But neither of the alternatives can be

accepted. The universal proposition has been established
w'lth rigorous certainty and the smoke is an object of

perception. Hence the apparent contradiction is resolved

by the supposition that there is fire. Thus inference
becomes identical with arthapatti. To this the Bhattas
reply that inference may be said to be arthapatti onl} ifwe
admit that the universal proposition was not pieviously known
by inference. In certain instances we know that .<'inoke is

lelated to fire. Fiom this we infei that all smoke is related

to fire. It cannot be said that without the universal

proposition our knowledge of the relation between smoke
and fire in certain instances invohes a contiadiction which
18 resolved by the postulation ot it Hence the knowledge
of the universal proposition requires to be explained by
inference as a separate source of knowledge.
Now we are to observe that arthapatti as explained
above is an independent piamana like perception, inference
and the rest. It cannot be reduced to inference as the
Naiyayikas and the Samkbyas endeavour to do. The
reason for this, however, is not, as the Advaitins suppose,
that there is no .such thing as nyatmki inference, to which

* Syadevain yarti HarvBdbOrnavatftniaKiiiinattvaiDaDiiiiiaD&daDyeuftvagalaiii syal,

etc ,
SB , p 7!l
 OTHEB SOURCES OF KNOWLEDGE 399

arthapatti may possibly be reduced. The Advaita Vedantins

lose their case against those who prove that vyahreki is a
genuine type of inference, or reduce arthapatU to some
other kind of inference like the hypothetical-»categorical or
the disjunctive-categorical syllogism. The real reason is,

as the Bhattas point out, that arthayat ti cannot be reduced

tojany kind of inference. The fundamental condition of all

inference is the i elation of vyapti or invariable concomi-

tance between the major and the middle term. In every
inference the conclusion foUoxvs from a universal proposi-
tion which 18 the result of a previous induction. The
knowledge of the universal proposition is derived from the
uncontradicted experience of agreement in presence or in
absence between the middle and the major term. In any
inference we apply a universal proposition, which is already
known, to a particular case. To reduce arthapatU to
inference we must, therefore, show that here our knowledge
of the unobserved fact follow's from a universal proposition
which IS already known by induction. The Naiyayikas and
others would sa} that the knowledge giv en by arthapatU
does follow from certain universal propositions. That
Devadatta eats at night follows from the universal proposi-
tion, “ A man who does not cat at night while fasting by

day IS not lat.” Similarly, the fact that he is out follows

from the pioposition, “ A living man is eithei at home or
out of it
” But these propositions are not cases of real

oydpU or induction. They are not generalisations from the

particular facts of experience The universal proposition,

“ Wherevei there is smoke there is fire,” is derived from

the particulai instances of their coexistence. So also, the

proposition, “ Wherever there is no fire there is no smoke,”
is derived from the particular instances of their agreement
in absence. But we have no previous experiences of the
agreement in absence between eating night and
‘
at ’

fatness. We have previous experiences of the concomitance

400 NYiYA THEORY OF KNOWLEDGE

l)etween eating and fatness or between their absence.

Hence to explain the apparent contradiction between fat-
ness and absence ol eating by day we have to suppose that
there is eating at night Our knowledge of the fact that
Devadatta eats at night does not follow from any universal
proposition which is already known, because there is here
no universal proposition at all. It is the result of an
attempt to correlate his fatness with the absence of eating
by day — a process of reasoning which is different from that
involved in inference Similarly, the proposition, ‘
A
living man is either at home or out of it,’ is not a genera-
lisation from particular instances of the concomitance
between a man’s absence from home and presence outside.
We cannot say that the one coexists with the other, just
as smoke coexists with fire. Hence we cannot deduce
our knowledge of the fact that Devadatta is out from any
such universal proposition. Rather, it follows in the wake
of any attempt to reconcile the facts that Devadatta lives
and yet he does not live in the house. In fact, the so-
called universal proposition is itself a statement of the
conclusion in general terms and cannot really explain it.

Hence arthapatti is not a form of inference, but a separate

source of knowledge.

3 Ahhavft and unvpalnbdhi as sources of Inowledge

Abhava may be taken to mean either contrast or non-

cognition. In the first sense it means a relation of contiast
or antithesis between two things as between existence and
jion-existence. ’
When there is such a relation of contrast
or contradiction between two things, then from the exist-
ence of the one we may know the non-existence of the
other and vice versa. Thus from the non-existence of ram

' Abh&vo viroribl, abhutatfa bhutasytt, etc

,
NB., 2 2.1.
 OTHER SOURCES OF KNOWLEDGE 401

we know the existence of some contact of the clouds with

high winds which prevent rainfall. It is on account of the
obstruction offered by high winds that rain drops do not fall

to the ground, as they otherwise would by the force of

gravity. The Naiyayikas hold that this argument from non-
existence to existence is really a form of inference, because it

is bised on a uniform relation of concomitance between two

opposite or contradictory things. Two contradictory objects
are so related to one another that the existence of the one
implies the non-existence of the other and vice versa. Hence
abhdva or non-existence as a source of knowledge is to be
included within inference. ^ The Vai^esikas also reduce
abhdva to inference According to them, the non-existence
of the effect indicates the non-existence of the cause, just
as its existence indicates the existence of the cause. Hence
abhdva or non-existence gives us the knowledge of that
which is uniformly related to it, like the lihga or the
middle term of an inference. The argument based on
abhdva or non-existence is thus really a kind of anumdna
^
or inference.
Later Naiyayikas take abhdva to mean the absence of

cognition and not the relation of contrast or opposition

sbetvveen two things. ” In tins sanse abhdva coincides with
anupalabdht or non-cognition. According to the Bhatta
Mimamsa and the Advaita Vedanta, anupalabdhi is an
independent pramdna or source of knowledge. It is the

unique cause of such presentative knowledge of non-

existence as is not duo to inference or any other kind of
*
knowledge. Thus the non-existence of a jar on the table
which I see before me is known from the absence of its cogni-

J Ibid
« NK,p. 226; vs, 9.2.1.

» TB.,p 16, N.L,p. 67

< JfianakaroQa^ny&bhSiv&Dubhav&B&dhara^akarai^amaDupalabdhirQpaiii pramft^atD,
\P., Ch. VI.

61~(1117B)
402 NYAYA THEORY OF KNOWLEDGE

tion or ita non-percep tion {(inupalahdhi) I judge that the jar

does not exist on the table because it is not perceived, while I

know that it would have been perceived if it existed there.

This knowledge of non-existence cannot be explained by

inference, since it is not brought about by the knowledge
of vydpti or a universal relation between two terms. It

cannot be said that the non-existence of the jar is inferred

from its non-perception which is known to be universally

related to non-existence. The knowledge of a universal re-

lation between non-perception and non-existence requires a

previous knowledge of non-existence as such, which cannot
be given by any inference. *
Nor can we explain the know-
ledge of the jar’s non-existence by comparison (upamana)
or testimony, since it is not due to any knowledge of
similarity or of words and sentences Hence to explain

the direct knowledge of the jar’s non-existence we have to

recognise anupalabdhi or non-perception as a separate and

independent source of knowledge. All non-perception,
however, does not prove the non-existence of what is not
perceived We cannot perceive such supersensible entities
asdharma and adharma, ether and atom Yet we do not judge
them to be non-existent Non-perception gives us the
knowledge of the non-existence of such objects as should
have been perceived if they existed. If a thing should
be perceived under certain circumstances, then its non-
perception under those circumstances is a proof of its

non-existence. It is this appropriate non-perception

(yogyanupalahdh) that is the source of our knowledge of
^
non-existence.
The Naiyayikas, Samkhyas® and others controvert the
above view of anupalabdhi as an independent source of

' N&pyanumeyatr, ajCatena teoa kasjacilliAgasya aarubaDdhagtahapasatobbafit,

8D,p 87.

* VP *SD,»bjd
* Evamabb&vo ’pi pratyakjameva, TKD.,
p 60.
 OTHEK SOURCES OF KNOWLEDUB 409

the knowledge of non-existence. According to them, such

knowledge does not require anupalahdhi as a separate
source of knowledge, but is a special case of perception.
Just as we perceive the existence of objects, so also we
can perceive their non-existence under certain conditions.
When there is a jar on the table before me T perceive its
existence through a direct contact between my senses
and the object, jar Hence the existence of the jar is

directly perceived by me But when there is no jar on

the same table, I perceive its absence or non-existence as
a characteristic of the table The table is characterised
by the absence of the jar. Hence the absence of the jar
comes in contact with my senses through being adjectival
(viSesana) to the table which is in direct contact with the
senses So when I directly perceive the table, I indirectly

perceive the absence of the jar on it. This perception

of the absence or non-existence of a thijig, however, requires
two negative conditioma, namely, the non-perception
(anupalambha) of that thing and the hypothetical reasoning
(tarka) that if it existed it would have been perceived
like the table. Before we come to know the absence of
the jar on the table we must be sure of the fact that we
do not perceive it there Further, we must be sure that
all the conditions that are necessary for its perception are
present at the time when it is not perceived The absence
of the jar is perceived by me when 1 do not perceive it

on the know that it would have been perceived

table but
if it existed there. The non-existence of the jar is thus
known by means of perception when it is combined with
the non-perception of the jar and the hypothetical reasoning
about its existence. ^ That this knowledge of non-existence
IS a form of perception is directly felt by us. We are

immediately aware of the fact that the non-existence of

1 TBrkHsiiliakanpanupalathbhabanftthemt pratyak^eijaivabhftvagrahaDat, TB.. p 11,

404 NYiYA THEORY OF KNOWLEDGE

a jar on the table is directly known or perceived by us.

Anupalahdhi or nou-perception of the jar is a negative condi-

tion of the perception, and not the source of our knowledge of

its non-existence. If non-perception be taken as a source of

knowledge, then it must be either cognised by some other
non-perception or not cognised at all On the first alter-

native we are landed in the fallacy of argumentum ad

infimtnm. On the second, non-perception becomes identical

with perception, since, like perception, it is knowledge
which IS not pioduced by any other knowledge. Hence
the Naiyayikas conclude that non-perception is not a

separate source of knowledge, but a special case of per-

’
ception .

We have already seen how the knowledge of non-

existence, which IS not due to inference or any other kind
of reasoning, cannot he explained by perception. A negative
fact like the non-existence of a jar is not a sensible fact

like the existence of the table. We do not understand

how our senses can come in contact with the absence of

a thing. A man says at noon that nobody came to his

house in the morning. Here we cannot suppose any

contact between .sense and nobody. It cannot be said

that we perceive the non-existence of a thing as a charac-

teristic or quality of the locus in wliich it is non-existent.
To know a certain locus as characteiised by the absence
of an object is to know beforehand what absence or non-
existence IS. Hence our primary knowledge of non-exist-
ence cannot be a perception of it as the quality of any
locus, like the red colour of ;i rose Nor can it be said
that we are immediately aware of the fact that the non-
existence of a thing iH perceiced What we immediately
know IS that we do not peicene a thing in a certain place.

1
Abhavaprat)^ak9a8y.inul)havjkatv{ldarniptjlainl)ho pj na prjiiiapaDtarflni etc
, 8M ,
 OTHER SOURCES OK KNOWLEDGE 405

This absence of perception gives us the knowledge of its

non-existence. The Naiyayikas practically admit this

when they take non-perception as the antecedent condition
of the perception of non-existence. If to perceive the non
existence of a thing we are to make sure that we do not
perceive it under favourable circumstances, then we are to
say that it is the absence of perception that assures us of
the thing’s non-existence Hence we conclude that ami-
palabdhi or non-perception should be recognised as a separate
source of knowledge to explain our primary cognition of
the non-existence of objects.

4. Smrti or memory as a distinct source of knowledge

We have already given an account of the views of

the different schools of Indian philosophy with regard to
smrti or memory. It may be recalled here that with the

exception of the Jaina, the Vai^esika and the Advaita

Ved&nta system, all the schools are definitely opposed
to smrti being regarded as a form of valid knowledge.
All these schools agree in holding that smrti is know-
ledge which is solely due to the impressions of past

experiences In it there is a revival ol the impressions

of some old experience and consequently a repetition

of the experience itself in the form of images. Smrti

as a revival of past experiences has been excluded from
the forms of valid knowledge on two main grounds.
First, it has been urged by the MImamsakas that smrti does

not give us anynew knowledge but is only the revival of

some old knowledge. In it we do not know anything new,
but only remember that we knew something before. Then
the Naiyayikas, who do not accept this as a conclusive
reason, argue that smiti is not pramd or valid knowledge
because it is not anubhava or presentative knowledge. In

it we have the knowledge of what was once given in our

 406 NViYA I'HEORY OK KNOWLEDGE

experience, but has now ceased to be given and presented

to us. It is not the presentation of any objective fact but
the representation of what was once presented. The object
as remembered is different from the presented object, since
the object as presented before has now ceased to exist.

Hence we cannot speak of a true correspondence between

memory and its object (yatharihya)
The Jamas ^\ho accept smrti or memory as a source of

valid knowledge refute the above grounds urged against it.

According to them, snuti is not merely a revival of the

impressions of past experience. While the origin of
memory is conditioned by the revival of impressions of past
experiences, its essence lies in the knowledge of something
as ‘
that,’ i e. as past {ia htydhlta) ’ It is the knowledge
of a previously experienced object as past. To put it in the

words of Hobhouse, ‘mcmorv is an assertion of the past as

past.’ That memory refers to a previously experienced
object, or that it is an asseition of the past, is known from
memory itself Such knowledge of the past by means of
memory is valid, since, like perception, it leads to success-

ful activity. We cannot deny the validity of the knowledge

by memoiy simply on the ground that it refers to a previ-

ously known object If memory becomes invalid because

its object IS previously known, then the perception of the

fire wdiich is already known by inference would become

invalid. Lastly, if memory be invalid, then all inferences
which are based on the remembrance of vydpti between the
major and the middle term would become invalid ^ How
can any knowledge be valid when it has its basis in memory
which IS invalid? The validity of memory is presupposed
in the validity of inference as a source of knowledge.

^ Tadityak&TanubhfitarthaviBaya hi praUti^ Hmjtintjrucyale* J'rumeyJuuunJa-

martan4af p ^
^ Na c&atvapramai^aiu samv&dakatvat . . . ko hi BmftipurvakamaTiimiari&tJDnbhyu

pagamya punasl&rii nu&kurj&t, etc., thtd,, pp. U(V97

 OTHER SOURCES OF KNOWLEDGE 407

Some Vai^esika writers also aie m lavuur of accepting

smrti as a source of our knowledge of the past. Laugaksi
Bhaskara recognises both memory and presentative cognition
(smrtyanuhhava) as forms of valid knowledge and their
instrumental or special causes as sources of knowledge.
According to him, smrti or memory arises out of the
impressions of past experience and is the knowledge of an
individual object as ‘
that ’
or as something previously
experienced, e.g. ‘
that bathing ghat,’ ‘
that city of

Benares.’ *
Vallabhacarya proves on strong grounds that
smrti or memory also is an independent source of know-
ledge. Smrti is a separate pramdna because it gives us a
true knowledge of certain facts (arthani§cayahetutvat). The
fact that it depends on previous experience is no reason for

denying its independence, for that is something common to

all the pramanas or sources of knowledge. It cannot be
said that it is merely the repetition of some previous
experience. It is something more than the faint repetition
of a past experience If it were not so, we could not at all

know that the experience is past. In smrti or memory we

know an object as that which is past. The awareness of
its ‘ pastness ’
is no part of our previous experience of it.

It IS memory that gives a knowledge of this new element,

namely, the ‘
thatness ’
or the ‘ pastness ’
of an object, and
is, therefore, an independent source of knowledge. Hence
the Mimarhsaka contention that memory does not give ns
any new knowledge falls to the ground Then the Naiya-
yikas’ objection that smrti or memory does not correspond
to its object is also untenable It is true that in memory
an object IS thought of as being present at some time in the
past ipuri'anntnmdnalfdldiacchinna) and that its once present
condition has now been extinct {nivrttapuridiiastha). But

^ Saihskararaatrajanyam jfianam 8mrtil>, yatba ea manikarijiika . 8Qi|‘tyanu*

bbavagadhara^arfi pramakarapaih pramapam, Tarkakaumudi^ p 6.

408 NYiYA THEORY OF KNOWLEDGE

this IS Qo good ground for denying the coriehpondencc

between memory and its object. When we remember an
object, we are aware of representing it as no longer present
or with its old conditions as now extinct. The object is

therefore faithfully represented in memory. It follo\^8 from

this that memory is the true knowledge of an object. ’
We
may add also that memory is a presentative knowledge
{anuhliava), since it is based on an objective order of

things in the «orld. As we have already seen, anubhava

or presentative knowledge is the cognition of what is

objective (tattva) as distinguished from the false or the

subjective (nropita). In presentative knowledge the object
need not be directly given as in jxTception, for that
will exclude inference, comparison and testimony from
the field of (inubhata or presentative knowledge. All

these, howevei, are recognised by the Naiyayikas as

forms of presentative knowledge. What is common to all

these recognised forms of presentative knowledge is not

that they give us an immediate knowledge of some object,
but that they refei, either directly or indirectly, to

an objective fact oi an objective order of facts. In this

sense smrti or memory is as good a presentative know ledge
as any other recognised by the Nyaya or any other school
of Indian philosophy. The fact that an object is past is as
objective as the piesent existence of another.’ Hence
memory as the knowledge of the past as past is a true
presentative knowdedge (yathaithanubhava) To explain
such knowledge of the past we have to accept smrti or
memory as a separate source of knowledge (pramana).

1 Brnrfcirapi manaDtararoeva, arthaniAcajabetutvdt, aDabhavapArataDtrj&nnaivAm

iti cet, na, utpattiparatantryasja prama^antaraBamyftl Adhikaparicchede ca
praiD&^atv&t, anjatbd tadvyavasthdoopapatle)?, tatrftvacchmaih hi amflirarthatn-
dkalajati, ca jadi purvftnubbavasydpi gocarahi tatr&pi tadityullekba^ ayht . . •

nacet amftireva tatr&napek^eti mdoaio, NyayaVilavati^ pp 67-G8

* C/. H, H. Price, Perception, p 11 : “ The paat it as much a part of the
*'
ireal world as ibe present, and quite t>9 interesting
 OrHBR SOURCES OP KNOWLEDGE 409

Among Western thinkers Kussell, Hobhouse and others

recognise memory as the primary source of all our know-
ledge concerning the past. They agree m holding that we
may know the past in other ways too, for example, by
reading history or by inference. But these cannot give us
anv knowledge of the past unless we have already a direct
knowledge of it through memory. Thus Bussell says
“ It is obvious that we often remember what we have seen
or heard or had otherwise present to our senses, and that
111 such cases we are still immediately aware of what we
remember, iq spite of the fact that it appears as past and
not as present. This immediate knowledge by memory is

the source of all our knowledge concerning the past; without

it, there could be no knowledge of the past by inference,
since we should never know that there was anything past
to be inferred.” ^
In another place Russell observes that
memory resembles perception in point of immediacy and
*
differs from it mainly by being referred to the past.
Similarly, Hobhouse ® shows that memory is neither the
retention of past experience, nor a mere image of past
experience, but an assertion of the past as past on the basis
of such retention and images. Without a direct knowledge
of the past by memory we cannot understand retention as
an effect of past experience nor an image as an image of
the past. It cannot be said that we may know the past by
inference from the retention or impression of past experience
*
or from its revival as an image. For, as both Prabhacandra
and Hobhouse point out, inference in its turn involves
memory. Further, we cannot understand anything as a
sign or mark from which to infer the past unless we first

^ The Problems of Philosophy p. 76

• The Analysts of Afind, p. 178.
The Theory of Knowledge^ Pt 1, Ch. IV
< Vide Prameyakamalamlirtania, thtd,

62— (1117B)
410 NYiYA THEORY OP KNOWLBDOB

know that past as signified or marked by such and such

things Hence Hobhouse takes memory as a fresh postulate
of knowledge. “ It is,” he says, ” a direct or immediate
belief about the past, not a belief based on some other
truth.” ^
A. C. Ewing * also thinks that ‘
the direct view
of memory is we have any knowledge of the
clearly true if
past at all. If we know the past, it is the past we know
and not our present ideas.’ It is a mistake to suppose, as
the Naiyayikas do, ‘
that if we are directly aware of the
past, the past must be, so to speak, bodily present to our
mind or occupy, the same position as our present objects of

perception.’ Thus according to these Westein thinkers,

memor> gives us an immediate knowledge ot the past just

as perception gives us an immediate knowledge of the

present. Hence there can be no objection to memory being
regarded as true presentaiive knowledge (yathdrthdnuhhava).
This view of the matter removes the last vestige of the

difficulties in the way of taking smrti or memory as a

pramdM or source of valid knowledge. In fact, it stands
next to perception in the order of priority among the
sources of valid knowledge. All sources of knowledge
other than perception involve memory of some kind as one
of their conditions. Inference cannot take place without
the memory of a universal relation between two things
(vydpti). Upamdna or comparison depends on memory of
the knowledge communicated by a reliable person. In iahda
or testimony we depend on memory in order to understand
the meanings of words and to 8}nthesise the meanings of the
separate words into the construed meaning of the sentence.
Arthdpatti or postulation involves a conflict between the
order of our past experience as remembered and that of our

present experience. Anupalabdhi or non>perception also

* Op. crt.iP. 76.

< Mtiul, April, 1980, p. 142.
 OTHER SOURCES OF KNOWLEDGE 411

implies a contrast between the memory of something and

the absence of perception with regard to it. Hence, finally,

we have to aci^t jeveii s^arate sources of knowledge

which may be arranged m order of priority as follows :

perception, memory, non-perception, inference, comparison,

testimony and postulation.

5. Summary and general estimate of Nyaya Epistemology

The Nyaya theory of knowledge discusses all the import-

ant problems of logic and the relevant problems of meta-
physics. It formulates a realistic theory and tries to meet
the idealist’s objections against realism as a system of
philosophy. According to it, knowledge is a quality of the
soul, which manifests the objects of the world. All know-
ledge of objects, however, is not valid. For knowledge, to
be valid means to be given (anubhava) in some way or other,
and to have an assurance of truth in it. The truth of know-
ledge consists in its correspondence to real facts and the test
of truth lies in its pragmatic value and the coherence or
‘
consilience ’
of its different parts. It follows from this that
memory and dream, doubt, error and hypothetical reasoning
(tarka) cannot be regarded as valid knowledge, since they
are either not given or not true cognitions of objects.

These are, therefore, brought by the Naiyayikas under the

class of non-valid knowledge which includes all cognitions
which are either not given and true, or are false. The
falsity of knowledge is constituted by its non-correspondence

to facts and is known through failure of the practical acti-

vities inspired by it. It follows that truth and falsity are

not intrinsic to knowledge and that these are extrinsic

characters determined by external conditions like corre-
spondence and non-correspondence to reality respectively.

So also, no knowledge is by itself known to be true or false.

That is, truth or falsity is not self-evident in any know-

 414 NYiYA THEORY OF KNOWLEDGE

and non-syl logistic, pure and mixed. Having regard to

their purposes, or the nature of vyapti, or the nature of the

induction on which it is based, inferences are classified into

svartha and parartha, or into pUrvavat, Sesavat and
samanyatodr§ta or into kevalanvayt, kevala-vyatireki and
anvaya-iyatireh The fallacies of inference are all material
fallacies which affect the truth of the propositions involved
in inference. They ultimately arise out of a fallacious reason
or middle term. There are six kinds of fallacious middle
terms which violate one or other of the conditions of a valid
middle term. A logically valid inference must be free from
all kinds of fallacies.
Comparison is the source of our knowledge of the denota-
tion of a word on the basis of a given description of the
objects denoted by it. Thus a man may be told : “A
papaya IS an animal resembling the cow” If, on subse-
quently seeing a gavaya, he is able to give its name, we are
to say that he understands the denotation of the word
through comparison. Comparison is of different kinds,

according to the different terms in which the description may

be given. It is true that comparison involves an element of
perception and of testimony. The description comes to us

as the statement of some authoiily and, as such, is a kind of

verbal testimony. So also, we know by perception that

certain objects possess the chaiacters mentioned in the given

description. Btill comparison cannot be reduced to percep-
tion and testimony, because tliese will not explain the
application of the name to the relevant objects, which is the
essence of comparison. Nor can we explain it by inference,
for when we know the denotation of a word from a given
description, we do not reason syllogistically, but simply
compare certain objects with a given description. To
understand the denotation of a word in this way
requires a selective activity of the mind, which is

different from perception, inference and testimony.

 OTHER SOURCES OF KNOWLEDGE 416

Thersfore, compansoD is a distinct method or source of

knowledge.
Testimony is the statement of an authoritative person,
which serves to give us true knowledge about certain objects.
Itmay come to us in the form of either spoken or written
words and may relate to perceptible or imperceptible objects.
In any case, there must be a significant combination of the
words according to four conditions. They must imply one
another and express compatible ideas There must be ade-
quate proximity among them, and they must convey the
intention of the speaker or the writer who uses them.
Testimony is the source of the greater part of a man’s know-
ledge of the world. As the verbal knowledge of objects, it is

distinct from all other kinds of knowledge. Perception, in-

ference or comparison cannot take the place of testimony,

although there may be in it an eleinpnt of this or that other
knowledge. It is true that testimony ultimately depends on
perception or inference for its validity or for the proof of its

validity. Again, there may sometimes be a conflict of autho-

rities, Since, however, these difficulties are not peculiar to
it, but rather common to all the sources of human know-
ledge, there is no reason why testimony should not be recog-
nised as an independent method of knowledge like perception

and inference. If in spite of the conflict of perceptions or of

inferences, and the need of their mutual verification, we

accept them as independent methods, we must accept testi-

monv and comparison also as equally independent sources of

knowledge. All other sources of valid knowledge including
non-perception and postulation are brought by the Naiyayikas
under perception, inference, comparison and testimony.
Non-perception need not be admitted as a separate source of
knowledge to explain our knowledge of non-existence, for it

may be perceived by us as adjectival to the existent object

which is its locus. So also, postulation may be reduced to
oyitireki inference ani need not be made a separate method
412 NYiYA THEORY OF KNOWLEDGE

Udge, but must be evidenced by external conditions like the

success or failure of practical activity.
There are four kinds of valid knowledge and so four
distinct and independent methods of knowledge. These
are :
perception, inference, comparison and testimony.
While the old Naiyayikas define perception as an unerring
cognition produced by sense-object contact, the moderns
define it as immediate knowledge or as knowledge not

brought about by any antecedent knowledge. There are five

external senses and an internal sense called manas which is

necessary to explain the perception of the soul and its stales

and processes. The individual soul is an eternal and all-

pervading substance which is not essentially conscious, but

has the quality of consciousness when it comes into relation

with external objects through the senses. Corresponding to

the six senses, there are six kinds of ordinary perceptions
which give us direct knowledge of all perceptible objects
including substances, their qualities and actions, universals,

relations and the four kinds ot non-existence. There are

seven categories of reality, ot A\ljicb six stand foi positive,

and the last for negative facts. Of positive tacts, sub-

stance, attribute and action aie said to be existents, while
geneiality, particularity and inherence are called subsistents
Non-existence is a negative but real fact and, according to
the Naiyayika, there may be a direct perception of it along
with that of the positive fact which it qualifies Of ordi-
nary perception, there are thiee modes, namely, the mrvt-
kalpaha, the saiikalpala and pralyahhvjM or recognition.
These represent different stages in the development of our
perceptual consciousness, but they are equally valid and
refer to real contents of the objects of perception. In
addition to ordinary perception, the modern Naiyayikas
distinguish three kinds of extraordinary perception called
simAnyalaksana, jMnalaksam and yogap. The first two
are recognised by them as necessary to explain the peroep-
 OTHER SOURCES OF KNOWLEDGE 413

tion of objects by senses which are not ordinarily competent

to perceive them, and the last to explain the supernormal
cognition of objects, which cannot be brought about by any
sense.
Inference is a type of syllogistic reasoning in which we
pass from the apprehension of some mark or sign as related
to an object, to something else, by virtue of a relation of
invariable concomitance between the two. It is an argument
in which some thinker asserts that a certain proposition is

true because certain other propositions, which imply it, are

asserted to be true. Thus inference is a combined deductive-
inductive process which ensures both the validity of the
reasoning employed and the truth of the conclusion reached.
An inference must have as its constituents three terms and
at least three propositions. There are three conditions of
valid inference, namely, vydpU or a universal relation
between the middle and the major term, pahsata or the
assertion of the minor term, and ImgapardmaHa or a
synthetic view ot the middle term as related to the major,
on the one hand, and the minor, on the other. Vydpti is

the logical ground on which the validity of inference

depends. It is an inductive geneiahsatioii based ultimately
on the direct perception of the universal in the particular.

Pak^atd is the psychological ground which conditions the

possibility of inlerence and is defined by the modern Naiya-
yikas as the absence of the condition in which there is

certainty but no will to infer. Lingapardmar&a as the

correlation of the major, middle and minor terms is useful

for demonstrating the truth of the conclusion. These three

steps, together with the initial statement ot the object of

inference and the final conclusion, give us the five-membered

form of the syllogism. Since inference is a combined

deductive-inductive reasoning in the form of a categorical

syllogism, we have not a classification of inferences into

deductive and inductive, immediate and mediate, syllogistic

416 NYiYA THRORY OF KNOWLEDGE

of knowledge. For the Naiyayikas, then, there are four

distinct and independent sources of knowledge.
As a realistic theory of knowledge, based on the evidence
of direct experience, the Nyaya epistemology has a strong
appeal to our common sense. It has also a great value for
the orientation of philosophical problems from the common-
sense standpoint. But undue reliance on uncriticised ex-
periences and common sense has been the cause of certain
defects in the Nyaya theory. The Nyaya conception of

knowledge as an adventitious quality of the soul substance is

true neither to the nature of knowledge nor to that of the

soul. To say that knowledge is a quality is to leave un-
explained the fact of self-transcendence and ideal reference

to objects, which is inherent in knowledge. As we have

already observed, knowledge is the most fundamental fact of
reality. The distinctions of substance and quality, subject
and object, all fall within knowledge and are intelligible only
on the ground of knowledge. In this sense knowledge is

the essence of the ultimate reality which we call the soul or

the self. It does not require to be attached as a quality to
any other reality, say matter or mind or soul. It is just the
self-expression of reality itself. If this be true, then we
must give up the Nyaya theory of the individual self as a
substance which is not essentially conscious, but is accident-
ally qualified by consciousness when associated with a body.
Such a view of the self is contradicted by the evidence of our
introspective consciousness which reveals the self as a con-
scious subject and not as a thing with the quality of con-
sciousness. Further, on this theory, the disembodied soul will
have no consciousness and will, therefore, be indistinguish-
able from a material substance So also, we must give up
the idea of an ultimate dualism or opposition between subject
and object, mind and matter. To the ordinary understand-
ing, these appear to be two opposed substances w'hich c&n
hardly come into any relation with one another. The
 OTHER SOURCES OF KNOWLEDGE 417

Naiyayika does not go far beyond this common-sense view

when he treats the psychological distinction between know-
ledge and its object as the ground of an ultimate dualism
between soul and matter as two realities. In truth, however,
the distinction between subject and object, mind and matter
is a relative distinction made within knowledge. So it

presupposes the reality of a transcendent self which makes

the distinction and is the ground of both the objective and
subjective, the material and mental orders of existence.
On the purely logical side also the Nyaya theory
appears to be inadequate on some points. The view that
truth IS not self-evident in any knowledge, but requires in
all cases to be evidenced by independent grounds, logically
commits us to the fallacy of infinite regress. But, as we
have already seen, the self is a self-evident reality which does
not require and possibly cannot admit of any other proof,
for every proof presupposes the reality of the self as con-

cerned in the act of proving. Some Naiyayikas practically

admit this when they say that the truth of self-consciousness
(anuvyavasaya) is self-evident. So also, there seems to be|
no good ground for the Naiyayika’s refusal to admit memory,
non-perception and postulation as independent ways of

knowing. But for memory, we cannot have any knowledge*

of the past. Memory cannot be explained as a reproduction
of past experiences, due solely to the impressions left by
them. We could not know certain states as impressions or

reproductions of past we had not already

experiences, if

known the past directly through memory. We may have a

direct knowledge of the fact of non-existence, just as we have
that of existent facts. But from this we should not con-
clude that this direct experience is as much a matter of
sense perception in the one case as in the other. The truth
of the matter is that while the existent is perceived, what is

non-existent is wot perceived, and that directly in both cases.

Postulation is not the deduction of a conclusion from given

63— (1117B)
418 NYAYA THEORY OF KNOWLEDGE

premises, but the necessary supposition of a general principle

as the only explanation of some given facts. For Kant the

existence of God is a postulate of the moral life, not in the sense

that it is deducible from certain ethical propositions, but in

the sense that it is the only principle which can explain

ethical propositions concerning the moral life. So tve have
to admit memory, non-perception and postulation as three
distinct ways of knowing in addition to the four recognised

by the Kai}ayika.
Nothing that has been said above by way of criticism
should give one the impiession that the N\aya epistemology
has no value Such an impression would be entirely wrong
In fact, the contiilmtion of the Nyaja theory of knowledge
is not reall} infenoi to that of any othei theory, Indian oi

Western. The method ot logical analysis employed by the

Nyaya in the stud} of the problems of logic and metaphysics
is a valuable asset for any system ot philosophy The
charge is often heard against Indian philosophy that its

theories are not based on logical reasoning but on religious

authority and, theiefore, they are dogmatic, rather than

critical The Nya\a philosophy is a standing repudiation of

this charge The theoiy of knowledge, formulated by the

Nyaya, is made the basis not only of the Nyaya- Vai^esika,

but also of othei Indian systems, with slight modifications

The Nyaya applies the method of logical ciiticism to solve

the pioblems of life and reality. It is by means of a sound

logic that it tries to ascertain the truth and dehmd it against
hostile criticism Many of the contributions of this logic

are of great value even at the present d.iy The lealistic

logic, or moie geneially, epistemology of the Nyaya will not

suffer by companion with the modern leahstic theories of

the West.
 INDEX
Abhava (Kon exiatence), natuie of, Dl f
,
Gautama, 1, passim
kinds of, knowledge of, 195 f
193 f , , Gange^d, Tattvacmtamani 5, 31, 32, 137,
as a source of knowledge, 40(j f 264, 308, passim
AbbiliKanvaya-vada, 371 f
Action (Karma), nature of, 179, kinds of,
180, perception of, 180 Hobboust, L T . Theory of Knowledge, 5,
idvaitasiddln, 211,278,981 80, 111, ir, 142, 409 410 ,

Akhyati (oi Vi\ekakh}dt]) 41 f ,

H(>lIln^zv^orrh, H L , Psychology, 147
Alexander, Spa^e, Ttme, and Dciiy^M Hypotbetical Argument (Parka), nature
15, 41, 45, 113, 117, 1()7, 198, 2<r, of, 47*18 kinds of, 49 f Western ,

Anirvac anlyakbyati, 10 1 paiallel of, 52

Anupalabdhi (Non-cognition) , nature ind
validity of, 400 f ,
different views of,
401 f 117
,
Indefinite cognition, 32 r
Anviksikl, 1 Inference^, definition of, 253 f , distinguish-
Anvitabbidliana ^ada, 373 f ed fioiij perreption, 254 f , constituent^'
Anyuthakb}aM, 39 f of, 256 f vyapti ns logical ground of,
,

Artl.apitt’, nature and validitv f, 3^3 f <

,
261 f ,
question of peHUo principii in,
diflf rent views of, 391 f 117 IS ,
271 f paksatd as psychological ground
,

Asatkbyati, 40 f of,277 f lin.'aparamar^d as immediate

,

Atmakhyati, 39 f cause of, 284 f class’fication of, 289 f , ,

Attribute, nature and kinds of, 174, per Icgical form of, 297 f fallacies of, ,

ccpnon of, 175 306 f gemi-logical fallpcies of, 319

,

4iaTanavarmi, 338,349

lihagai adyiid, 10 Jaini, J Outlines of Jainism, 181

,

BhdniQtl, 363 James W

Principles of Psychology, 16,
,

Bhdsdpartccheda (ui Kanhatnlt), xvii 70 3d5 Radical Empiricism, 18,

pas<:im Pragmatism 114
Bosanquet, B , Logic, 21, 31, 44, 114, 131, Jba 't N Prahhdkara School of Purva
19s, 376, 377 Essentials of Logic, 215
,
Mhndntsd 12, 398, 349, Sadholal
Br.idley, F H ,
Principles of Loqu 34, Lcctuie^ on The Kydya Philosophy oj
198, 287, 295, 304, 330, 376, 377 Gautama, win 42
Joachim, H H
The Nature of Truth, 114 ,

Joad, C E M
Introdurtion to Modern
Cause, definition and different kinds of, Philosophy, I*)
57 58 Johnson. W
E Logic, 2j4, 37() ,

Cbaoinan and Hen’o, FundanieniaJs of To 3 epb,H \\ ?> Introduction to Logo,

Logic. 52,304 303, 304
Contemporary Briiush Philosophy, 205
Cre ghton J E Introductory Logic, 283
,

84 iveitf. B, Indian Loqu and Atomism,

C nticdl Realism, 17, 15, 113, 132 XVJl
Kant 3, 194
Klemrn 0 History of Psychology, 160
,

Dusg'ipta, S N History of Indian ,

Knowledge (Buddlii), definition of, 9f,
Philosophy, xvii ontological problem of 14 f , act theory
Dot ',
I D MSir Ways of Knotting, xvii,
,
of 14 f , 1 elation ti^eory of 17 f ,
quality
2R(. tbeoiv of, 19 f ' ( lassification of, 21 f ,

DTdbiti, 1 11(>17
Dwakan, 175, 179, 185, ISS, 285, 310, 361 Kohler W ,
Gestalt Psiphology, 173
Doubt, nature of, 30 f ,
kinds of, 31 f Kuppuswami Sfistii, Primer of Indian
Dieam, nature of, 2 > f Logic, XVII, 369
Kusumdiljah, 57,86 lUf

Eaton, R M
General Logic
,
273 71
Error, nature of, 34 35, theories of, 36 f Tjossky, N 0 , Intmtne Basis of Know
problems of, 83 84, tests of, 84 f ledge, 43, 56, 112
 420 INDEX

Manameyodaya, 99 f Badhakrishnan, S., Indian Philosophy,

Manas (Mind), nature and function of, XVII
157 f passim
, BEinanuja, iSrihhdiya, 89, 162, 209 222
McDougall, VV , Outline of Psychology, 76 Rdmarudri, 810
Meaning, nature and kinds of, 353 f Reid, L. A, Knowledge and Truth, 17,
Mellone S H
Introductory Text-Book of
, 18,20.111,114, 116, 117
Logic, 376 BLlation, nature and kinds of, 185 f per ,

Memory (Smfti), definition of, 23-21, ception of 187 f

conditions of, 24-25 kinds of, 25 f , as Boyce, J , The World and the Individual,
an independent source of knowledge, 365
405 f 4JU, 417
, Bussell, B., Analysis of Mmd, 13. 15, 114,
Mill, J S System of Logic, 127, 128, 161,167,199 20i i09, Our Knowledge
181, 241 of the External World, 17, 114, 128,
Mimdmsd-Sutra, 186, passim 379, Outhne of Plitosophy, 16, 371,
Mind, d7, 43, iM, 378 886, Principles of Mathemntics, 279,
Mitabhdsint 33, 226 Problems of Philosophy, 17, 27, 53, 67,
Monistf 119, 132 114, 118, 128, 134, 184, 187 , 4U9
Montague, W P , The Ways of Knowing,
386

Sabda (Testimony), definition of, 345-46,

New Realism, 3, 18, 45, 64, 118, 128 kinds of, 346 f different views of, 347 f
, ,

Nyaya-Bha^ya, xvii, 214, 806, passim independent validity of, 381 f

Nydyahmdu, 129, 254 266 f 3abda,^akti-prakdSikd, 353 f 361, 376, 383, ,

Nydyahmdutikd, 12, passim a95, 386

NyayakanddTt, 391, passim S.imavaya, nature of, 186 f perception of, ,

Nydyakoia, 188, 207 187 f criticnm of, 188 f

,

Nydyahldvatl, 391, 408 Samkara, Commentary on Brahma-Sutra,

Nydyamanjari, xvii, 12, passim 188, 36'1.64
Nydyavdrttika, xvii, 2, 214, passim Sdmkhyakdrtkd, 136, 148, 158
NydyavdrtUkatdtparyapariSuddhi, xvu, 109 Sdmkhyapravacanabh&sya, 136
NydyfitdrUikatalparyatikd, xvii, 2, 214, Sdmkhya-sutra, 136
passim Saptapaddrthl, 22, 83, 34, 55
Nydya-Sutra, xvi, xvu, 5, 214, 308, passim Sarvadarianasamgraha, 57, <
5, 75, 84,
Nyay i-Sulra-Vxtti, xvu, 334, passim
xvi, 191 211, 265 f 270, 280, 348
,

Nydya-Sutra Vttarana, 334 Sdstradtptkd, 12, 136, 364, //awfw

Schiller, F C S , Logic for Use, 875
Sell, B N The Positive Sciences of the
,

Paddrthadharmasamgraha, 298, 308 Ancient Hindus, xvu, 300

Paflcdsttkdyasdra, 175 Self, nature of, 161-63, function in per-
Parsons, J H
Introduction to the Theory • eption, 161 f realistic thejry of, 164,
,

of Perception, 137 416-17

Particularity iVi^esa), nature and kinds of, Sense, nature of, 143 f k.nds of, 145 f
, ,

184 f not perceived, 185

, function of, 161
Perception, primacy of, 125, definition Sentence, nature and construction of,
of, 129 f ps>cboioi^y of, 143 f
,
kinds , 366 f meaning of, 371 f
,
import of, ,

and objects of, 199 f Nirvikalpaka and , 875 f , proposition and, 376 f
Savikdlpaka 207 f Recognition as a , Siddhdntacandnkd, 223, passim
mode ol, 224 f extraordinary 228 f
, S id dhaniamukt avail, 86, 286 f , 840, 861,
Perry R B , Present Philosophical Ten- passim
dencies 114 Six Buddhist Nydya Tracts, 181
Persistent kn )w ledge, 68 f Slokavdrttika, 186, 258, 864
Prakaranapaflctkd, 71, 133, 202 Speneer, H Principles of Psychology, 16
,

Praiiia, characte istics of, 64 f Spho^a, theory of, 362-66

Pramiil^a, different views of, 59 f Stebbmg, L 8 ^4,
Modem Introduction
Pramdnasamuccaya. 13*>, 211. 258 to Logic, 78, 232 f 279, 281, 808, 804,
,

Prameyakamalamdrtan^a, 133, 144 222, , 361; Logic in Practice, 92, 141,142,

226. 254. 331 349, 381, 406, 409 232, 871, 386
Price, H H, Perception, 120, 144, 170, Stout, Q F.. Analytic Psychology, 205;
172, 408 Manual of Psychology, 206, 240, 246
PriDgle-Pattison, Idea of Immortality, 67 Substance, nature and kinds of, 168, per-
Proposition, sentence and, 376 f import , ception of, 169 f
of, 377 f , subject-predicate and impera-
tive, 378, identity and relational, 379 f
Psychologies of 1926, 173 Tarkahhdfd, xvii, 140, 858, pasiim
Psychologu.al Review, 202 Tarhakaumudl, 407, pasiim
 INDEX

Tarkamfia^ xvii, passtm Vcdanta-Patihha§d, 285, passm

Tnrkasarhgrahaf xvn, 286, 863, passm Vidyabhusaija, S C History of Indian
,

Tattoacintamam, xvji, passm Logic 5, 348

TaHvadiptk&t 366, passm Vtvarnnnprameyasamgraha, 368, 859, 878,
Tattvakaumudi, 144, 286 338, 319, 896 374, 378
Titcli^-ner, E B , Tesii-Book of Psychology, Vydsa-Bhd§ya, 136
16, 70,146,117, 150, 3t)5

Truth, definition of, 64 f ,

problems of, 88,
tests and tlieories of, 88 f Watson, J B Behaowur, 15
,

Woodworth, K 8 Psychology 28, 38

,

Words, sounds and, 3il f nature of, 352 , f

Universal, d fferent views of, 180 f

,
per- meanings of, 353 f kinds of, H67 , f ,

ception of, 1^2-184 import (»f 358 f ,

unity of, 362 f

Upatoana (Comparison) definition of, Wundt, W , Human and Animal Psycho

326 f ,
Jama, Mimamsa a* d Vedanta logy, 240
views of, 327 f , cLissification of, 832 f

independent validity of, 836 f Yoga-sutra, 136, 390, passtm

UpaskdTQi 33, 196 Yoga-Bhasya, 349, passtm

421