GOKULDAS SANSKRIT SERIES

NO. 36

TARKABHS
OF

KEAVA MlRA

Edited with Translation, Notes and an

Introduction in English
by

S. R. IYER
with a foreword
by

PANDIT GAURINATH SASTRY

( Ex-Vice-Chancellor, Varanaseya Sanskrit University and
President, Sanskrit Sahitya Parishad,
Calcutta )

HM! U!',111111 \ ORlENTALIi

House of Oriental and Antiquarian Books
VARANASI
DELHI

:
'
Publisher
- -;
CHAUKHAMBHA ORIENTALI A
P. O. Chaukhambha, Post Box No. 32
Gokml Bhawan, K. 37/109, Gopal Mandir Lane
VARANASI-221001 ( India )
Telephone : 52939
Telegram : Gokulotsav
BranchBungalow Road, 9 U. B. Jawahar Nagar
DELHI-110007

Chaukhambha Orientalia
First Edition 1979
Price : Rs. 25-00

PrintersVidya Vilas Press, Varanasi,

il ft: II

r.
S.

TT
'S.

cfto
,

%.
- ^ ^ ^oo

FOREWORD
Kesava Misra's Tarkabhs is a manual of the
sycretic school of JVyya-Vaisesifca. It is based on the
old Myya tradition and takes into consideration the
sixteen categories of the JVyayasutra of Gotama. This
proves beyond doubt that even in the age ofJVavyanyya the study of the old Myya texts were prevalent.
The study of Tarkabhs was very popular in all
parts of India as is proved by the large number of
commentaries written on it.

Kesava's style is lucid

and his arrangement of topics is scientific. Long ago

Bodas published an edition of TB with exhaustive
notes.

The present edition by Sri S. R. Iyer is highly

informative and analytical.

I am sure it will win the

admiration of the students and teachers of Myya and

will be considered as a landmark in the tradition of
the textual study of Nyya-Vaisesika.
Feb. 14, 1979

G a u r m a t h Sastri

WR- II

i 1/

CONTENTS
[ The letter (N) denotes that such subjects are described in
the Notes portion ]
Foreword
Abbreviations
Introduction

...
...
...

...
...

s:q^ra:

q^mf^i^

The defects of definition (N) 6

JWJT^^FTT^
STR^S^Oij

7
8

FW^^ 11 ^

35ROT3S^pJir

Explanation o f 3T?qsTTflf% ( N )
10
*ER0I5lfipfcfJ3[
12
crT?nRTeFRn3[
12
p \ /*^ fN ^^

1 'I

Explanation of cnrcU
TTT^. and ^l^ (N)
snfRTOW (N)

14
18

S<:cf%^|i {*fET

fyuiWJ

srwraif^rc&m

20

23

ffiT35HoTr
26
Brief description of "$*q, gar
and i of the Vaiesikas
(N) 26
s^l^ or negation (JNJ
%^fa^ TOTT^^^lf^^TT:
5TOTnf%vn*Ti

ou
33
36

5TC^T^^^OH3[

36

5T?T^3^0m

36

The divisions of SJR in

tabular form (N)
39

...

V
XI
13-44

Explanation of R R W and
ef33qq cognitions () 39
44
44
45
46
46
47
47
OT^f&infiroii^m: (N) 49
53
T^RTl1^5'9oTIWTT^[^ ( N )
^T^'a^^IMc^IignTi ( N )
^TT^>iR^^Qn!7^Hf%: ( N )

Di

54
55
56
58
58

Brief description of the

technical terms in 3ra*u
59
(N)
62
66
67
70
71
Explanation of the five
members of syllogism
if

( viii )

and comparison with

the Western or Aristotlean
system (N)
72
78
80
81
82
Comparison
with the
Aristotlean system (N) 83
Figurative representation of
the three kinds of nfcf(N)86
88
H
91
The definition of q^TT according to modern School(N)92
^*3T*TOli3n*W
93
Explanation of t*3T*n3Hr
(N)
93
srf^
96
Explanaion of sqfr (N) 100
103
104
Third variety of SR$lf?arsf>
(N)
105
107
108
Explanation of how ^ , sufrf
and fqgTOTO differ from
t ^ m r c r ' s (N)
109
Comparision with fallacies
in the Western system
(N)
111
Summary of s r g m ^ (N)
118
119
121
121

124
125
126
128
128
Mmamsa view

regardingy
129

Explanation

of 3Tf$ andi

130
Rhetorician's view regarding
(N)
137
The Naiyyika view of
3Ts^ter as compared to
that of the Grammarians
(N)
137
fqf%ft^q<W*
139
Explanation of the Mmamsaka view (N)
141
143
144
146
Explanation of the Mmamsaka view (N)
147
150

154

Dififereaces of the other

schools with the Naiyyikas(N)
. 155
Brief description
of
different views regarding
l

165

.
3

4. m:
0)
(ii)

196

166
169
170
172

(vi)

199

174

(vii)

200

174
175

(viti)

200

197
Explanation
198

(ix)firon:
: (N)

175
(x & xi)

(iii

202
204
204
205

176
(iv)

(v)
(vi)
(viii) ;
() HT:

177
179
180
181
183
185
187
187
189
189

Naiyyika view of
sleep (N)
191

(xiii)
(xiv)

(O

193
Explanation of

205
206

(XV)

(xvi to xxi)
209
210
210
212
213
214
215
215

(xxii & xxiii)

(xxiv) m
(3)
(4)
(6)
(7)
sr|%rt

192

(2)

201
202

216

5. |f: (

218

194
(ii) W
(iii)

195
196

7.
8.

220

9.
10.

232

221

12.
223

226
227
228
229
230
231

234
238
238
239
241
242
243
243
244
246

Index to authors and works

referred to in the Text
and Notes
247
231 Errata
249

ABBREVIATIONS
An. Bh,
Ans.
A. S, S.

= Annam Bhatta
= Answer
= Anandasraraam Sanskrit Series

R P )
ssBhsa Paricheda
B. S. P. S.
= Bombay Sanskrit and Prakrit Series
Com.
= Commentary
Ch. B. or Ch. Bh. =Chennu Bhatta
Ed. or Edn,
G. S or G. N, S.
H. , L.
H. N. N. M.
K. P. J. R. I.
Kar. or cjnf
K.M.
3RT. S".

s Edition
= Gautama Nyya Sutra
= History of Indian Logic
= History of Navya hyya in Mithila
=Ka Prasd Jyaswl Research Institute
=Krikava!
=Keava Miara
=Kvya-prakaa

Ms. or Mss,
N. S. or ?m. ^

s Manuscript or Manuscripts
- Nyya Stra ( The numbers following
denote the w^usr, ^lg^ and m^ )
=Nyaya-Vaiesika

N. V.

Q.

Question
=2=^4^! ( referring to

SHI. ^
S. B. E.
S. M.
T

=^TR%?fCR^rEr ( referring to 3 i f ! ^ )
= Sacred Books of the East
= Siddhnta-Muktvali

'^ 1

TwI

?'

Tarka-Sangraha
l

=Tarka-Sangraha-Dpika

V.

s Verse

V. S.

Vaiesika-SGtra

INTRODUCTION
The six orthodox systems of Indian Philosophy popularly
known as S T ^ R T R are H^T and *fta tafp and ?s

^tftaterT
The Six Systems
or
^^fHf^T

and Tattftartai ( or

Because of some affinity in thought and

doctrines these have traditionally been
recognised as three pairs of allied systems
( OTRcr^foT ), sometimes designated
as qfrT{r, ?m?T5 and rfteisrig^r, though each has flourished
independently and with a distinctive individuality," often
acquiescing in and attacking the others just like the friendly
quarrels among brothers in a large joint family. All these
systems were originally intended to help in the exposition
and interpretation of Vedic Texts which are believed to be
the revelations of God and therefore infallible. That is. why
they are all called aTfcf^^T^TTf^ in contrast to the heretic or
Tlfc^ systems such as those of the Crvkas, the Buddhists
and the Jainas, who do not believe in the authority and
omniscience of the Vedas. Amidst all their differences in>
doctrines and methods one thing is common to all the orthodox
systems, viz, that salvation is the sole end and aim of life
with knowledge as the only door leading to it. It is natural
to ask then which are the things worth knowing and which
are not. The propounders of each system have answered this
question in the light of their own thinking and speculation.
These teachings have been codified in cryptic terse aphorisms
called Stras modelled on the Stras of Pnini on grammar*
These Sutras form the basic foundation for each system, on
which later scholars have built up imposing superstructures,
by their . commentaries variously called as *u*q, snffe, ff%,
*mmi, Zt^l, etc. The Sutrakras, i. e. sages who first codified
the crystallised findings of |each system in the form of Sutras,
are according to tradition ( 1 ) ^sfass forOT*9*T;( 2 ) qSfs for
2 T.

14

INTRODUCTION

( 3 ) effort for ftfsE; ( 4 ) ^ for ?qm; ( 5 ) srfafq" for

and ( 6 ) qi^rraiT or S^TST for q^cT I Many a Western
scholar have questioned the authorship of these sages on the
ground that it is supported only by tradition and also because
the same sage is sometimes known under different names; e. g.,
e&n^ is known as 3&s and w * n , ftcra as ^nfrfr and sr^q^,
and ^T^TW as KTTH. But so far nothing has been brought
to light to discredit the belief in the tradition ascribing the
authorship as above. Except in the case of the \qq the Stras
of the other systems have come down to us more or less in
the original form. The original rfczr^t of 3>f$ is supposed
to have been lost, as the one extant now under that name
has been proved to be a very late production of the 14th
century or so. The original teachings of that system were
collected together in 72 small verses in what is called Hi^r33TR^T or SHSsrenEfcr by ^ar^&r, sometime before the second
century A. D. Since its appearance this work constitutes the
basic authority for that system just like the Sutras. Various
scholars have tried their best to arrive at the periods when
these Sutras were redacted or codified; but due to lack of
definite data all their findings have ended only in tentative
guesses which may conform to or confirm some subsequent
event or development of thought.
( 2 ) Gautama Buddha, the founder of Buddhism and
Vardhamna Mahavra, the founder of Jainism, are now
believed to have lived from 624 to 544 B. C.
The Origin and from 599 to 527 B. C. respectively accord
of sfijfsfcf
ing to their own traditions. Any date before
and ?i[pj
the rise of these two schisms is almost a blank
anybody's guess depending upon the particular
evidence or material on which emphasis is stressed. The
Buddhists and Jainas are believed to have collected together
the verbal teachings of their founders in the form of Tipitakas
and gamas a century or so after their Nirvanas. They might
have been prompted to do this by seeing the Hindus clinging
to their Vedic texts with unflinching faith and reverence.
However that be, the conflictpretty severe in many cases

INTRODUCTION

15

that arose among the adherents of the three faiths, and that
stirred the minds of all intellectuals in the decades following
the times of these two reformers," must have compelled all the
parties to systematise their doctrines and to give expression
to them in a methodical form and also to meet the arguments
put. forth by the opponents. The Buddhist Suttas definitely
refer to the ^ and iHtf$ systems and vaguely to the vftti
and z^m. The canonical scriptures of the Jainas compiled
about the same time contain a large number of logical terms
such as SK^, srgTTH, ^??T<n, rg, ifr, etc. which on the face
of it appear to be borrowings and not their own creations.
The mere reference to these systems positively proves the
existence of the Hindu Daranas in some form or other much
earlier. No definite evidence is yet available to set aside the
orthodox belief that all these systems had crystallised by this
time into some definite form and were being circulated in
learned circles from teacher to pupil. We are at present
concerned with the sritf^ and ?sn*T systems only, which
together constitute what is called Indian Logic. All the
internal evidence pointed out by various scholars tend to show
that the tstfsc Stras took shape sometime about 400 B. C.
and that the *mq Stras must have followed it not long afterwards, at any rate about 300 B. C.
( 3 ) Some scholars lay undue stress on the fact that
( circa 320 B. C. ) has not mentioned^ specifically the *%m or
as such among the important branches
learning which a king must learn. These
is ^"PT
branches of learning as mentioned by him in
Chapter 2, feiTOg^ section of his smstra, are
( i ) 3?l?gtf%t ( = logic and philosophy ); ( ii ) ^*ft ( a the
Vedic scriptures); ( iii ) sna ( = Economics, trade etc. ) and
( iv ) ^ftf '( =the science of polity). In the concluding
portion of the same chapter he explains the important subjects which come under these headings as follows :

16

INTRODUCTION

i srsTsr Verruf
I

consists of m%*t, n\u and l l ^ . If the first

two are taken to mean the philosophical systems of ^fq^s and
qcSfe, what does st^T^c connote ? It cannot be the Crvka
system as 3^fk^T refers to that system separately in the samesection. The word naturally means what has come down from
the world,- i. e, the course of reasoning which is in common
use as handed down from posterity and which the common
man accepts as reasonable on the. strength of custom and
practice.This is corroborated by the author when he explains
further egfwfsR^Tntmthat which investigates by reason.
If this is not logic what else could it be ? The verse at the end
stating that it is the beacon light for all branches of learning,
strengthens this view. The same verse is found quoted by
WE*UR in his n*nwT on 1-1-1 with the last line modified
as fei^it sr^facf, thereby indicating the source of the verse.
Perhaps the word s n ^ f a ^ t was more commonly used for
logic than ?qm during the time of ^fgq, There is every
reason to believe that the word mm was originally used to
denote the <?fii*ricn system. sTq*3**r and many ancient Smntis
use this word to denote the reasoning of efrffcr. Pnini uses this,
word in the same sense, vide his Sutra II-2-122. Many later
writers on''^h:ta3T use that word as part of the title of their
works, e.g., KivTOn^ of qwqwfifosr.sTiriiT^if^iWofm^srwi
PTftEgn of 3itfTTO, e t c. The word was perhaps appropriated
by the Naiyyikas later on to denote exclusively the fivemembered syllogism on which depends the whole course of
reasoning to convince the opponent in a debatecf. 3KcqTOf o n
1-1-1 under snras, 'citsqq*t wm ^V. As a result of the great
importance attached to this ?3T*? or five-membered syllogism*
the system itself came to be designated as ?^i^fgf?JT, 'WreiM,
etc., thus supplanting the original term sn^lfrtft. In this
connection the words of Mr. M. R. Bodas in his introduction

INTRODUCTION

17

to the edition of ^^T^ in the B. S. P. S. deserve close

consideration :"As a general rule we find that when a new
school arises it coins its own phraseology to distinguish itself
from its predecessors. In this case, however, the followers
of Gotama appropriated an old word and that word stuck to
them so fast as to become afterwards their exclusive property.
The explanation, it seems, lies in the fact that the science of
logic which afterwards developed into a separate system was
originally the child of ^Rtaferi" ( p. xxxi ).
( 4 ) The ^f^j;:* of ^T\\% as extant at present contains
ten ieqrre's or chapters, each divided into two
sf^fg^sraf
parts called n%E and in all consists of 370
Sutras. The first four read as follows :
( i ) amist *W s?rusn^w = Now we shall explain *ni
( ii ) ^cisvg^q'friarq'faf;! H ^"fl: = That by which one attains
prosperity and salvation is vro.
( iii ) egr^^iri^^ srmi^HssThe authority of the Vedas is
due to the exposition of this vro.
{ iv ) TO^qq;cr^55,gaT,spH, cnm^jfarttq, H ^ O T ^ T q^rcfai
Wf swnfar r^RTfe&TfnfiW = The attainment
of salvation comes about from the correct understanding
of the (six) categories by discriminating their similarities
and dissimilarities; ( a ) Substance, ( b ) Quality, ( c )
Action, ( d ) Generality, ( e ) Particularity, and ( f )
Inherence This correct understanding itself is caused
by a.special kind oft?;?.
The rest of the whole work is taken up by the definition,
exposition and elaboration of these six categories. That tangible
categories as such are only the first three was well known to
35cn3 as he definitely points out *4 r ^ q g t j ^ R g ( 8-2-3 ).
The next three are intellectual conceptions postulated to fulfil
certain logical needsH*nNr fit %fr ! a r ^ ' ( 1-2-5 ).
Though 35<ni3[ uses the word scmsr in its various senses at very
many places he did not mention it as a separate category. Its
inclusion as the seventh category was done by later scholars
though it is not certain who effected it. ^PU^T* in his

18

INTRODUCTION

classifies the categories as vrre ( = positive ) and

( = negative) while fref^fsr who followed 3<^jrf within
a short time divides q^sf's directly into seven classes, ther
six positive ones as above and s w ^ as the seventh. "While
the Aristotlean categories area logical classification of predicates only, the qitfep categories are a metaphysical classification of all knowable objects or of all reals. They are not,
as the Kantian categories are, mere moulds of understanding.
Nor are they, as the Hegelian categories are, dynamic stages
in the development of thought. Hegel's is a philosophy of
Absolute Idealism, a dynamic and concrete Identity-in-difference. The ^fq^> system is a pluralistic realism, a philosophy
of identity and difference, which emphasises that the heart of
reality consists in difference It is a mere catalogue of the
knowables, an enumeration of the diverse reals, without any
attempt to synthesise them. "( A Critical Survey of Indian
Philosophy by Prof. C. Sharraa, p. 176-Delhi edition )" Though
the ^ft^p system cannot stand the rigorous analysis of sffattHT
or ^T?f as a school of philosophy its emphasis on scientific
thinking and inductive reasoning is a marked improvement
over the materialistic speculations of the times. In spite of
its followers being dubbed as "s4hnfgT:"i its unrestricted
speculations on many a metaphysical topic exercised such a
profound influence on the development of other systems that
it earned the eulogium "zvnxi qfRt <a r&nfrrenreq t">
which sticks to it to this day.
( 5 ) The earliest commentary available at present on the
V. S. is the tf^TOTOHSlc by snascnn^. Though this is more an
independent treatise than a ws*T, it is consiWcfqrre and
dered as such as it supplements the original
commentators. Sutras in many respects. There is a reference to another commentary called Won the V. S., in the s^STOf^o, a gloss on ankara's
supposed to be written earlier than the 13th
century. But for this reference no further information is
available about this work. g^rccftH^ftcq' is all the more impor-

INTRODUCTION

19

tant in that it gave the V. S. a theistic interpretation, thus

redeeming the system from atheistic proclivities? Nothing
is however known about the date or place of the author. Mr.
Bodas in his introduction to the ft* & has tried to prove
that he is earlier to s n ^ r e T the author of ^ r e ^ W * . The
authority and popularity of sr. VIT can be gauged by the large
number of commentaries and sub-commentaries written on it
by eminent scholars. Three of them deserve special mention
and though they were printed some years back all of them
are out of print at present. They are (i) itorac
4
written before the 9th century; ( ii ) ^znqrq^ft by
who fortunately gives the date of its composition as 991 A. D.
He quotes and criticises the sito^aft at several placs; ( iii )
fi&Tflilqreft by s?3RT^T*T ( circa 1050 ). Though he does not
mention sftsnt by name he quotes from the f ^ R ^ t ( as also
from No. i ) at a number of places mostly for refutation. The
printed portion of the f%*aiT335t extends only up to the end
of jjcrr section and it is believed by many that the author did
not live long enough to complete it. Besides sub-commentaries
on these three, several reputed scholars like 3TgTfsr, TOWT!r34 etc. have also commented on the V. S.
has also written a very useful and independent gloss
on the V. S., called ^ q ^ R which is very popular among N. V.
students. It is really heartening to note that, even at the
present time when the study of Sanskrit) as a wholenot to
speak of these classics-1-is at a low ebb, there are scholars
trying to rekindle the interest in the V. S., vide the **Fsr*
Commentary on the V. S. by ^iHT^T^T^T printed from Madras
in 1941.
( 6 ) The N. S. of qfcr*f is spread over five srora's each
sub*divided into two s n i p ' s . The total number of Sutras
varies from 523 to 532 with different authorities.
^T^^f
The fact that some of these authorities have tried
of ^ftcFT t o fix the total number of Sutras in different ages
proves that these Sutras have been modified,
enlarged or otherwise tampered with from time to time. ( i )
has written a ?inqitfira?[ giving the text of

20

INTRODUCTION

the Sutras and stating the number of Stras under each topic
( VbVn ). This might have been written by him as a guide or
sequel to his classic commentary Hrra3lf$3KTO3fcT. At the
end of this 5^m^fsr?Er we read as follows ;
i srsqrqr: 51 snf^f^Tf^r 101

84 I g^Tfa" 530* I q^iPl 206 | 3TTTf*T 8411 |

11 ? n
i II ^ II
il ^ n

( Vide ?m-vtj published in the A-S-SL)

The word ^sp^omiat the end of the first verse is signicant in that during his time ( 10th Century ) the text of the
Sutras was in a state of confusion. He has also taken pains
to count even the total number of words and letters. The date
of composition is given as 898 but no era is mentioned. Up
till now scholars have taken it as the Vikrama Sainvat corresponding to 841 A.D. But the late Dr. D.. Bhattacharya in his
"History of Navy a Nyya in Mithila", on the strength of new
evidence brought to light such as the works of sn^sit and
S^ftfJ, has proved that it should be the aka era corresponding to 976 A. D.
( ii ) The second attempt at fixing the text of the Sutras is
seen in the irogjft^l* written by enTsqffsr T and printed at
the end of Part I of pqpTOSr by ^sprTOg in the Chovvkhamba
edition. It was probably a sequel to his commentary on N. S.
called cx3rest35 ( not published ). He enumerates 531 Stras
and the editor ( or someone else) has added short notes
* Actually the number of Satras is 528. But what is meant by
ty^f 206 is not clear as the words run to over two thousands. In the
Chowkbatnba edition of this ^FTiJ^fl' ( Vide Dr. Ganganath Jha's
edition with his own gloss ) the Sutras are mentioned as 528, but the
other two as "g^fa 196; 3T^Fnf*T 8385."

INTRODUCTION

"

21

under each page showing the authority for the inclusion .of
one. It ends with the verse :
11
The date given 1428, if taken as the Vikrama era, corresponds to 1371 A.D.; but if taken as the aka era to 1506 A.D.
The author's date has been fixed on good grounds between
1410 to 1490 ( or "a man of th whole century", vide H. N.
N. M. p. 158 ) and so the date given appears to be that of
copying the book and the verse also of the copyist. This
^ t ^ R contains 531 Sutras as also various readings and shows
how changes have taken place in the text of the Sutras within
about five centuries after sn^qfcrfirsr I( iii ) A third attempt is seen in the ?3TR^ffo of fe^mT<
^ , the author of vnwrfHfc and its commentary
this f fa written in 1634 A. D. was composed
by him not for fixing the text or total number of the Stras,
but only as a concise and useful guide for beginners without
entering into any discussions. This purpose it serves very well
even now. He has commented on only 523 Sutrasvide the
iqrjjlf fxr printed along with the ^zn^vn^ in the A. S. S.
( iv ) As late as 1924 Dr, Ganganath Jha in his edition of
*wr*SWT with his own gloss ^ t e and another called
wsr^sg*, has commented on 532 Sutras. He has however
mentioned in the introduction that he has followed the text
of the Sutras as fixed by cnxTfqfofiqsr 1 except in the 4 cases of
deviation on the authority of the *n*q or ^f^. The other
commentary rctKras^ of unknown date which ends abruptly at
111-2-17 has 3 more Sutras in the portion commented upon.
( 7 ) The N. S. begins directly by enumerating T6 q^yf's,
TO<, gifcr, etc. ( the first Sutra quoted and explained by
Keava Miara in the text ), a true knowledge
The Contents of which, according to Gautama, leads to
of N. S.
salvation. The first chapter gives the general
definitions of these 16 and the rest of the
whole work is devoted to the discussion on and exposition

22

INTRODUCTION

of these categories with a fair amount of digression on theological, physiological and non-logical topics. A study of the
N. S. as a whole will not fail to strike the reader that the
emphasis of the author is more on dialectics and the method
of argumentation for vanquishing an opponent, which was
evidently dictated by the rising influence of the Buddhists and
Jainas and to some extent of the other Hindu systems as well.
His emphasis on SflM before explaining the Nyya concept
of knowledge which is equated to f% or intelligence, his
devoting two whole chapters ( IV and V ) to two unimportant
categories, *nfei and fa^-n?!", and his inclusion of ^T?, SS3<?
and fsfcf^T among the categories^ strengthen the view that the
practice of argument even for a temporary triumph over the
opponent was the principal aim of the author. This may perhaps account for his scanty and cursory treatment of srite
which are only the seven categories of 35GTIT^. .His indebtedness
to 3FUQ[ is revealed by his incorporating in toto or in part a
few of the latter's Sutras in his text as clarified by suc^n^R.
The sophistic tendency of the- N. S. has led to a loose knitting
together of a fairly large mass of heterogenous matter, a good
portion of which can even be considered irrelevant. While
the N. S. as a whole is more radical, dialectical and realistic,
the V. S. is more symmetrical, metaphysical and uncompromising in outlook. In spite of all these drawbacks, later authors
have culled together the gems of truth and thought scattered
here and there and built-up the strong and imposing edifice
of 3TH3TP8 which has stood the test of time and adorns au
honoured place in the pantheon of philosophies.
( 8 ) Dr. S. C. Vidyabhushan is of the opinion that sngqi^ is
not another name for afhm but was a different person, that
both have contributed to the production of the N. S., that gitq,
^V$ and a portion of vmvn were first handled by tas under
the name 3n?eftf%#f that the other categories and a fuller and
systematic treatment of gi?<n came from sr^qi^ and that the
latter was the redactor of the 3uMf%# of ntaw into the
present N. S. just as ^ ^ was the redactor of srfi^ra^ or the

INTRODUCTION

2$

of an^". Since there is no proof to hold that the two

were distinct persons except that some commentators have
referred in their introductory verses that the N. S. came from
^, it is difficult to agree to any of these suggestions, m^5R?cr, 35*WWT, fta* and others refer to both the
naines without any distinction. The Buddhist and Jaina works
refer to tfcn^T^ though the name s^qT$ is also rarely
found in them. The learned Dr. further mentions that the N. S.
quotes from the 3slgT33H*33 of unknown date and from the
Wsrf^SW of T3T5jT (circa 250-320 A. D. according to him)r
and that the N. S. criticised by ^nnT^ refers to the *nm
portion as explained in the ^^H^cT. Actually it would appear
to be the other way round. More than 56 years, have elapsed
since Dr. Vidyabhushan first published his History of Ancient
Indian Logic and enough material has accumulated since then
which would require a reappraisal of the whole situation.
With the material available to him at that time, his pioneering
work no doubt still holds its position. But in such matters
where collateral and internal evidence is the sole mainstaysopinions have to change with the advent of additional material
or with a fresh interpretation of the old. Dr. D. R. Bhandarkar
in his Introduction to the edition of cfViTCT in the B. S. P6 S.
has assumed as a foregone conclusion that they are one and
the same person when he says that the N. S. is "ascribed to
iftew alias S^RT^." He has put forth certain other views such
as ( 1 ) the N and V systems belong to one and the sameschool, ( 2 ) the former sprang up from the ^fc^^fa and the
latter from the ^ ^ i q T ^ a r ^ ^ , ( 3 ) ^f^r does not include
r*?T*T or titfcp under su?sfti%3?t in his scheme of learning for
kings, as pointed out by Prof. Jacobi, even though the
33R ( TOWT*R ) equates ?qm with tfrep; ( 4 ) this
cannot be regarded as the author of the N. S. as it stands at
present due to many additions found in it from Buddhist
sources; ( 5 ) that sr^qT?, ^ T ? and one m& were the pupils
of a teacher cftTrer*^ who is supposed to be the 27th avaiara of
iva as mentioned in the 3Tg*fU<. As* already pointed out
earlier all these can be considered only as bold suggestions

24

INTRODUCTION

requiring corroboraron by more substantial proofs other than

Purnic references or benedictory verses.
( 9 ) The oldest available commentary on the N. S. is the
of 9TCTOT3R who is also called qfa^flwfr.. Nothing is
known about him except that he must
T
have lived before f ^ r m , the founder of
and Commentries Buddhist logic. As a pupil of ^g^rg fj3?*niT
is supposed to have lived between 450
and 500 A. D. According to Dr. S. C. Vidyabhushan W W R
was a native of Conjeevaram and lived about a century before
f^nm. The absence of any commentary on the N. S. during
this long gap of about six centuries is inexplicable. If there had
been any, none has come to light so far. snc^sp is a forcible
writer and a close reading of his *rr"T would show that he is
well conversant with qo *no. These two srr&sr's on the Stras of
the two systems are regarded as supreme authorities by later
authors to be commented and enlarged upon with due reverence and not to be corrected or refuted. This ?WTOT^ was
commented upon after a lapse of nearly two centuries by
rrcrsi-setet in his ntcsrifife. From the initial verse of this
work it is clear that the interpretations of the vnsq were
highly clouded by adverse criticisms from Buddhists and
Jainas and that he took upon himself the task of clearing all
these clouds. Nothing is known about the personal history of
except that he is being refuted and criticised by ^r, the greatest of all Buddhist logicians, who is supposed
to have lived about 650 A. D. A host of Buddhist and Jaina
scholars, all eminent in their own ways, flourished between
650 and 950 A. D. who criticised the orthodox Hindu, systems
especially the ?2T?T and ;$famT. schools. Hindu scholars also
were not lacking in their efforts to counter these onslaughts.
Recent endeavours of scholars have brought to light the names
and works of many such authors, e. g. srmcfer, 3TfS (logician \
OTRif, BrH^r, sftsfj, sftsfftjs, ^vr, etc. on the ^m side and
similarly on the sfrrifl side as well. But the ?*n*T works of
this period which were direct commentaries on the sjer,

INTRODUCTION

2$

or gfTf^ were all driven to oblivion by the outstanding:

compositions of gT^qffsr I ( circa 950-1000 ) and of
S ^ r e r m ( circa 1050-1100 ).
( 10 ) ^T^qfcrfasr I is a name to be conjured with in Hindu
Sastraic literature. There are very few scholars who excel
him in his depth of learning and in the inimitai
*d able style in which he presents the subject
matter. His pen has wielded all the six systems
of philosophy in a masterly manner, except,
proper. He wrote ^mcft and cr?3TiNi ( not yet discovered ) on ^?rr, ?*Trc^fa3> and C T ^ ^ OQ taraT, tf^~
3%?t on ^Nq", a^rarc?* on $m and the ?iTT^lf^T^4a^T
on ?qTO. We shall refer here only to this work. Its excellence
and authority were such that it soon eclipsed^ almost all the
works written after grtc^W up to his own times. It earned
for him the titles atef^ and cn?qqfeT*f. He refuted all theBuddhist objections up to his time, especially those of qpfeftfHe and his teacher f^^^H who is the author of a work called
ssTPTOSfift ( not yet discovered ) were the main targets of
later Buddhists, especially gRsft and TO$tf. Within less
than a century the great ^q^n^iaf, commented on the cTcqq^T in his iinaqRsf:. g^re's adoration of qr^qf verges
on veneration as is evidenced by the eulogiums he gives to
the latter such as :
i
5

{ ii

( At the end ofTOTOTsection of qR^f^ > .

The N. S. together with these four famous commentaries,
and qRgf%; is often collectively spoken
as the nrq^sft or ^3?l?sit according as the Sutras are included or excluded, S^ZR has written besides %WTgdt,
already mentioned under 0dfs9 and this qfrgfe two independent works ariuRcref^ and ?*u*THT^fe. The former

26

INTRODUCTION

was written with the sole purpose of refuting the Buddhists

and the Jainas and from the incisive nature of his remarks
in it, this may be his earliest work. The *vm<gsfTSfs was
composed with the main idea of establishing the existence of
God through inference and in the course of his elaboration
he touches upon almost all the important topics of the *mq
system. In fact, the 5gpmff along with the qftgf% may be
regarded to constitute the nucleus for the evolution of the
Modern School of Indian Logic, ^sq-pqi^ which was established on a sound footing by nf^qiswT, The influence and
authority of 3R among later authors were so profound that
he is exclusively referred to as sTT^mf: in the plural ( and sometimes also as frmrSB ) in subsequent *qm and other works.
( II ) The development of Indian Logic is usually divided
into three periods;- ( i ) the ancient period extending from the
Vedic times to the end of the fourth century
Division into A. D. or till the appearance of the r*nsmi^
three periods of WE^l^R. This has been briefly described
above. ( ii ) The medieval period spreading
from 400 to 1200 A. D. or till the end of 3^DR; and ( iii ) The
modern period from 1200 A. D. onwards. During the medieval period the Buddhist and Jaina Schools of Logic prevailed
predominantly over Hindu Logic which had to struggle
strenuously for its self-preservation. On all three sides veteran
scholars of extraordinary ability, genius and scholarship
sprang up at different periods just like dazzling comets from
unknown space dimming the brightness of existing stars for
a while and again diving into unknown space, many not to
be heard of again. The Jainas were not so vehement; or bitter
in their attacks on their opponents and their missionary and
propagandist zeal was not so pronounced as that of the
Buddhists. It was a three-cornered fight in which the Hindus in
the end emerged victorious, aided among other things by the
destruction of Buddhist monasteries and universities like Vikramasila by the Muslims. By 1200 A. D. Buddhism was practically
driven out of India. Many Buddhist scholars fled to Tibet and

INTRODUCTION

27

China carrying with them the originals ofmany Buddhist works

which they were able to salvage out of the vandalism. Many
of these were translated into Tibetan and Chinese languages
with the help of local interpretors. Some of the originals have
been irretrievably lost and some have been unearthed recently
by the endeavours of a number of scholars; e. g. the works of
etc. These works throw a flood of light on the development of
philosophical and logical thought in India during this period;
bring to light at least the names of many unknown and forgotten authors and their works; and help in fixing the
chronology of many writers not only in N. V. but also in
other systems.
( 12 ) There are frequent references to logical themes in
the Buddhist scriptures, the Tipitakas and Suttas. The ST^TT^fWT in III618 divides knowledge into
Buddhist logic six heads, those that arise from the six
senses including the mind. In the BTgSTTg ^ under ifwfa^TO of g^faqftf inference as prevailing at
that time is described in some detail. The qft^R of fffqfqa^
describes logical procedures for settlement of cases among
the monks. Examples of debate ( s&m ) are given in the mn\g^^q^or of srfwi?nfqa3s. The ffs^q^ written in the first
century A. D. which is considered equal in importance to
their scriptures, mentions that fif^ ( the Bactrian King
Menander) "was well-versed in holy tradition and secular law;
the mis?, ftT, y*isr and tfep systems of philosophy; arithmetic, music, medicine; the four Vedas, the Purnas and the
Itihsas; astronomy, magic, causation and spells; the arts of
war; poetry; conveyancingin a word the whole nineteen."*
About 78 A. D. the Kushana King Kaniska established a new
system of Buddhism called 3^pn*T, the great path or vehicle.
The followers of this system began to designate the old one
as promulgated in the Tipitakas as ta^Rthe lower or
small path. The former spread rapidly in Tibet, Mongolia,
* Translation by Rhys David in S. B E. Series Vol, 35.

28

INTRODUCTION

China, Japan etc. and the latter continued to flourish m

Ceylon, Burma, Siam etc., while both systems grew side by side
in India. Kaniska held a council of Buddhist monks, introduced the use of Sanskrit for Buddhist works in place of
Pali which was in vogue and encouraged the writing of
independent works on- various aspects of the teachings of the
Tipitakas. Thus arose a large number of Suttas which in
general constitute the Scripture of the Mahynists and
contain their philosophical, metaphysical, theistic and logical
speculations. Based on these Suttas scholars like sreqta,
^m^T, wris 3sp3 f^TOT, ^hlfi, etc. composed all their
philosophical treatises on Buddhism. But prior to f^RTT no
one appears to have written any systematic work on logic
proper, though his predecessors have used logical me hods
and principles in their works. fi^RTO wrote the sraiTOg^:.
?H5jteT, commentaries on these and some other smaller
tracts. The originals of these are not available at present, but
some of them have been restored in part into Sanskrit from
their Tibetan translations. Buddhist tradition says that logic
was handed down by f^RTT through a succession of ten
pupils from 3T$?3tfcl to ite*fg ( circa 635 A. D. ) who was
the teacher of Houen Tsang at Nalanda University. Before
fj^WT, n*TI5R had refuted the N. S., especially Gautama's
division of things into 16 categories and also adopted a threemembered syllogism. 3i?cqTO?f answered these objections in
his ?qravncq and f^u^T in turn criticised erftTOTZR- By about
600 A. D. setera^ in nis ?qra3if3 defended the Hindu position. But within half a century twsfir raised his cudgels
against Hindu logic as a whole. He wrote about a dozen
works the most important of which are the wn^frig and the
TOToi^ffibp. Except the first one none of his works were available in original for a long time. Recently, due to the efforts
of Mahpandita Rhul Snkirtyyana most of the originals were
recovered from Tibet and have now been published. tro^fS
is considered the greatest Buddhist logician. His criticism of
Hindu logic was so ruthless that for nearly three centuries
the latter had to struggle hard for its self-preservation.

INTRODUCTION

29

Scholars like wsHl, g^Rf^r 3^*3^5, f^freffi etc. had tried

their best through the ages to defend the Hindu position*
But, as pointed out earlier, it was snwcriisr and 33?^ who
successfully resisted the onslaughts of the Buddhists, with
the result that the efforts of their predecessors were eclipsed
by their monumental commentaries not only in logic but in
other systems as well ^T^qfcifwr and his immediate predecessors mentioned above were within a short period attacked
by slTsft of the Vikramasila University, the last great star
of appreciable lustre in the vast galaxy of Buddhist logicians.
^TRsft's works were recently discovered from Tibet and have
now been published from Patna. He claims to have broken
down the pillars of Hindu logic as he himself proudly
proclaims :

( Vide ^msftfa^ft^Hnrafe frm Patna, p, 159 )

The great a^sn^r who followed ^T^sft within a short
period answered all his objections ( as well as of others ) in
his aiici^fiiN and other works mentioned earlier. Some
of his rejoinders to jH?it with a slight twist in the latter's
verse e. g. sts^qf 3^^5333^1 etc. have become classical
literature. With s ^ r e came to an end the long drawn-out
fray between the Hindusand Buddhists, mainly due to the
extinction of Buddhist activities in India. This healthy conflict between the two systems gave a great impetus to the
growth of Indian logic and epistemology.
( 13 ) The Jainas, like the Buddhists, seem to have followed the Hindu s^n^T for their logical needs in the early stages,.
with suitable modifications to suit their theistic
Jaina logic beliefs. In their canonical literature like the
WFrwq5,-WWftqpl, etc. 5R0T is classified as
of four kinds as in Gautama's N. S. The word | g besides,
denoting reason, is sometimes used as a synonym for
3T.

30

INTRODUCTION

Kt is mentioned in the sense of debate; and points of defeatfi?f*2U^fare also described. ^ T f ( circa 433-357 B. C. )
appears to be the earliest writer to discuss certain logical
principles in his commentary ^^^ife^fgfe on his own
work. His main object in this commentary was to establish
certain principles of Jaina theism and for this he .adopted a
syllogism of ten members. These ten members are referred
to and refuted by sneFTTCR in his rqrawTctr. vy^sng mentioned
for the first time the WTI^ or HEm#fa^ ( or only ?w in
short ), the seven-fold method of asserting the possible existence of a thing. These seven steps are :
( 1 ) ^Tl^fcTs relatively, a thing is ( real ).
( 2 ) ^rnfe = relatively, a thing is not ( real ).
( 3 ) **T3f 3 *nfci = relatively, a thing is and is not (real).
( 4 ) WT^^rawi55relatively, a thing is indescribable.
( 5 ) * ^ f e r ^T^STf^s relatively, a thing is ( real ) and
indescribable.
( 6 ) ttraifla ^rcTO^ = relatively, a thing is not ( real )
and also indescribable.
( 7 ) ^^fo ^ sfe ^TS3TK*qs$ = relatively, a thing is and is
not ( real ) and also indescribable.
All these assertions are judgements from different standpoints. The word WTSI here does not mean 'may be' or 'probably', but means 'relatively'. What the Jainas want to convey
is that reality has infinite aspects, that all judgements are
conditional or relative, that absolute affirmation or absolute
negation is erroneous, and that the infinitely complex nature
of things admits of all kinds of opposite or contradictory
predications from different stand-points. A host of best'scholars
among the Jainas have spent their energies in the elaboration
and defence of this s?qf, with the result that their whole logic
and epistemology may be supposed to centre round this
theme. The next important and original writer we hear of is

INTRODUCTION

$1

( circa 1-85 A. D. ). In his ttt3mffmffiI5T and in

its vr^qr he divides SWOT as shown below. The word sm\u\
is used by him to denote both valid knowledge and its
instrument.

( Knowledge
( Includes
gained through 3Tg??, STFH,
the senses. )
wrf?, s?$i?qf%,
, and

( Pertaining ( Pertaining
to ordinary to yogins. )
men. )

It will be noticed that the words qiH and sre^ are used
by STHScnfr in senses quite opposite to what they bear in
Hindu logic as also in the works of later Jainas. He also
describes ?PT, the various methods of comprehending things
from particular standpoints. Still logic was -subservient ' to
theism and religion. It was with fo^ctaf^f? ( circa 480-530
A. D. ) that systematic logic is believed to have evolved
among the Jainas. He wrote the ^TfT^R, a short treatise
in 32 small verses, which itself forms a part of his greater
work crf^-gTfflraT. This nqTOcU? explains nmw and w
and divides the former as \m^ and qftq, direct and indirect in
the ordinary sense, sjcsr^ is again divided into ssngfnf^ or
that arising through the senses, and qRTfN or that arising
through perfect enlightenment of the Soul, qffa is divided
into two, 3THPT and 3T3^. The former is sqnra and mm which
has only five members as in the N. S. Fallacies are described
as those of H^, lg and ^si?cr and include both Iccnsn and
fn^sTR, though the method of their treatment differs con*
siderably from Gautama's. ^r^% is treated under 3RT which
is of seven kinds. Later authors have elaborated on these

32

INTRODUCTION

topics in their commentaries, sub-commentaries and indepeo, dent treatises, c w a * ^ , s^^f, flifawf*?, TOW5, t^R??,.
to name only a few, are some of the most outstanding authors
who have enriched Jaina logic through the ages. This line
of brilliant scholars extends up to srsftfin^nfr, a native of
Dabhoi in Gujarat who lived between 1608-1688 A . D . H e
studied the astras under Brahmin Pandits in Baaras and
wrote over hundred works, some of which were important
treatises in Jaina logic like ?qrast^}q, ^ra^zr fspvrreTj ?*n*T^a^iTSJ, etc. He was so proud of his talent and scholarship
that he does not flinch from criticising and satirising his
opponents violently, vide his fling on SspmfRmfT, the
greatest exponent of c^qra in Nadia3*u*q sfte^si^r *ra mmi
fttfrfK ( in wrTCra*^TO ). With him ended this long line of
scholars and there is no new development or fresh treatment
in any branch of Jaina philosophy.
( 14 ) Before referring to the modern period of Indian
lgica brief mention of tbe sj^ofs or introductory manuals
is quite useful. The literature on the V. S. and N.
sr^FTirfffT S. had become so vast by about the 10th century
that it was practically impossible for an ordinary
student to have enough knowledge of logic 'within a short
time for the pursuit of his studies in other branches. Teachers
in various seminaries must have realised this situation and
to overcome it have summarised the principles of the two
systems in simpler language avoiding discussions about rival
schools. These summaries came to be known as SHOT'S. These
may be classed broadly under three heads as follows.
( i ) etfesT^aiTfa or those that'.treat exclusively of the
six ( or seven ) categories of 35tf7[ and the two WTO'S wm^
and rgere as in that system. Important examples are ( a )
and ( b ) wnqtsfteracft. The former was written by
sometime about 1150 A. D. 'He divides things
into seven categories including ' aram, and then defines the
things enumerated without entering into any polemics. He.

INTRODUCTION

has six t^WTS's, sw^focT being the sixth. He has probably

followed the R. *r?. or the ?zrragT? of *TTCTsi% in this respect.
fref^ divides *TTW*T ( generality ) into two varieties mtfh
and sqifr, perhaps for the first time. This work must have
been popular for sometime as there are some half a dozen
commentaries on it. The pqrassteraeft1 also belongs to the
same period as the first and is written by <effsrrsu4. This
cannot be considered strictly a R3>f or suitable for beginners;
but is more an independent summary of the R, vn., written
in a forcible, argumentative and intricate style. The author
enters into polemics and refutes the Buddhists and exhibits
his dialectical skill under all important topics. The work has
attracted eminent scholars like 3^RP, 3l?;*3T, w*ra^,
^g^fRfT and others to write commentaries and sub9
commentaries on it.
*
( ii ) rosiw&nf1 or those which explain mainly the 16
categories of itaT, mentioning briefly about the ^tf^r^ categories under nfcr. Under this category can be classed ( a )
5*n*rcn* of ncracr, ( b ) cnfi^^r of 3?3?r5r and ( c ) 33&n*rT
of %^T^rfr^r. The uqrorrc treats of only three raw's omitting
3rq*rra, mentio.ns anro^focr as the sixth | ^ i ^ m and makes no
mention of the Htfa^ categories as such. The author was a
Kashmirian and lived about the early half of the 10th century.
He has himself furnished a voluminous commentary on it
called ?qra?TO which was discovered only recently and
published from Baaras in 1969. It is a highly polemical work,
is quoted profusely by later writers, both Hindus and Buddhists,
and earned for the author the name ^roisFTT by which he is
more popularly known. It is this *J5aT3ST who is referred to by
^m*xt by the word ?qTOT$iftui in the verse quoted on page 17
and whom he considers as one of the pillars of Hindu logic.
( b ) 3rc37l5T was a native of Andhra parts and is supposed
to have lived about the beginning of the 12th century. His
afw^FT is in three chapters written in small verses followed
by the author's own commentary. The first chapter deals with
fourteen out of the 16 categories of .fom and the seven N

34

INTRODUCTION

categories, are treated under yam but separately from the

twelve of the N. S. The second and third chapters are devoted
to the exposition of the 15th ( *i\fn ) and 16th ( fascwra )
categories respectively, (c) The gshs?TOT of ^f^ is described
below under paras 17 and 18.
( iii ) 3^qR<Jnf or such manuals which begin with the
seven ^ f $ categories as the basis but treat of the fuir SHIP'S
more prominently than other topics. These were written by
authors who belong to the T{^?^m school and so have merged
the two systems in a more rational manner than in the
other two kinds of s t a r ' s . Only three of the most important
ones are mentioned here as specimens. ( a ) cr^f^cr of HT^NT,
one of the ablest commentators on, the ^iffc of fnrmf, who
lived, between 1540-1600. The first part of the work deals
with the seven categories of ^OTT^ and the second part
beginning with 'sraTOT:4^i3' treats of the four TOWS eschewing discussions and the ^rcrwnq phraseology. This useful book
did not gain the popularity it deserves probably due to the
two snST&Ts described below taking the field within a short
time of its appearance. ( b ) The ^ m f of $r^; is rightly
the most widely read primer throughout India and has
supplanted all the previous manuals by its excellence which
is evidenced by the large number of commentariesover
fifty have been brought to light so farwhich it has evoked.
It begins with the enumeration of the seven iffaap categories,
elaborates the four SHT^'S under gfg, a quality, and is written
in a very simple flowing language avoiding all dialectics.
The author belonged to Andhra parts, studied at Baaras and
was a great polymath having written valuable commentaries
in ssnqyOTj' STO, sftofe and if^ci. He has furnished a valuable
commentary on his own text called ^tfq^ and also commented
on the 3TT3&E of *m%%. He lived in the early part of the 17th
century and there is a common saying in the south to the
credit of his scholarship. ''rftiwwisrit Tretograa f[^ t"
( c ) vnqiqf^ ( or crf&rasft ) of fewm ( circa 1634 ) whose
was referred to earlier, consists of 168 small verses.

INTRODUCTION

35

The first part deals with the seven categories as in the'-g^fj

omitting the 24 qualities which is relegated . to the second
part. The four OT^'S are treated in detail under *s{\m\ which
is defined as the seat of knowledge etc. The author himself
has furnished a very learned commentary on the verses called
%5?crg^cgft which is a popular introduction to the ^s?T?qTqr
techniques. This is read by beginners as the second primer
after the crssjcfof, before entering into the polemical works of
and
( 15 ) Ever since the times of 3r^R who has himself
written two T^O'S one each on srftfg and ?qpT systems, several authors have tried to merge the two systems
aad together in some form or other. But it was
left to the genius of TKfcTtqpKTra of Mithila to
fuse the two systems together boldly into a
grand comprehensive whole 33>5ncr. There is no parallel
in the whole range of philosophical literature where a single
book brought about such a revolutionary change in the
realm of logic and epistemology of the Hindus. The ^m
f^raft of his often quoted as Jfi wrought a change of outlook and treatment in discursive and dialectical works in all
systems of Hindu philosophy written after the 14th century.
The well-knit marshalling of all relevant dessertations into
a coherent treatise took the world of learning by surprise.
Matter gave place to methodology and attention became
focussed on what the argument is for and against a topic
rather than who argued it. The older school dealt mainly with
the nature of things while the modern school concerned itself
to the correct and meticulous definition of things. The
'categoristic' method of vtm was replaced by the 'epistemological'method of *r#$?,n which.^naturally involved "a huge
over-growth of inflated and hair-splitting logic-chopping."
The rcqfrscmfi is divided into four chapters, each dealing
with one SW\I beginning with sicsrw. Each chapter is divided
again into sections and subsections and in the course of an
elaborate exposition of the nature and content of each

INTRODUCTION

thoroughly discusses all the important and relevant topics

of both the N. and V. systems, comparing and contrasting
them with kindred topics in other systems in a masterly and
authoritative manner. As is the case with a majority of great
scholars of ancient India very little is known about the
personal history of TT^T. He was a Mithila Brahmin hailing
from a village named %\^ ( not identified ) as described in
two cwemfo's or family geneologies written about 1726 A.D.,
which do not mention even his father's name. Dr. S. C.
Vidyabhushan however states in his H. I. L. that "he was
born in a village named Karion on the banks of the river
Kamala 12 miles south east of Darbhanga." His' period of
literary activity has/been fixed by Dr. D. C. Bhattacharya
from 1325-1350 A.D. from an investigation of all the evidence available so far, vide H. N. N. M. pp. 100-104.
( 16 ) Within a century of its appearance the Tfa established a school by itself throughout India. Ti|Nfs son and disciple EPTOT^WTPT was also a great logiCommentaries : cian. He commented on all the important
Mithila School
works of sr^TH and also on the 5th
chapter of the N. S., the 3sfmcft of sr^m^ 4 and the ^o^TOa^TSi of sftf *?. He does not appear to have
commented on the wfoT of his father as he does not refer to
sucha commentary in any of his available works, but only
quotes from the Tifa whenever necessary. Dr. Bhattacharya
has unearthed the names of a number of scholors like sf^?,
^cfnsrfw, *Tf f^, ^ritqT^l^r, ^ q f etc. Almost all of them
have commented on the Sfa, mainly in an antogonistic
spirit as is inferred from their being quoted in later works.
They are all supposed to have lived perhaps within the lifetime or within a few decades';pf TTIN and/or of his son. q ^ R
might have naturally felt diffident to wield his pen against
so many scholars of his own time and place. All their names
have now been forgotten due to the authoritative and outstanding commentary of sraffsi popularly known as qw?:
{circa 1420-1500 ). His commentary named an^f covers

INTRODUCTION

37

only three chapters of the *rf&T, omitting the 3WF portion.

He took lessons from his uncle ^frfim and perhaps also from
^qffe whom he controverts in many places in the suter.
He is also different from the authors of the same name, of
the poem Tftrfir?3 and the drama srcrarcisrc. He is believed
to have earned the epithet q ^ T by his skill as a dialectician
according to a traditional saying q^rcqfrqit *r w t fsf^. It is also said that w\fa had encountered 83H3TCTO,
the great Mdhva Pontif of Udipi Mutt, who was religious
preceptor to Krishna Deva Raya of Vijayanagar and author
of ^rF^c and other polemical works, wtfa had a number
of distinguished pupils in his seminary like srft ( the son
of his professor sygqfa ), *nsra ( his own son ), gi$3 ( his
nephew ), ^faa, jata*, wftw Wf etc. All of them as also
other reputed scholars of the time like S^fosr and sra^qfofif^r II both of whom were about a generation older to re^,
have commented.on the fffoT ( a few on snsfce also ) some
for and others against Tfi?3T like Sfgfrsr. Most of them are
now known only from references to them in later works.
Only the commentary of ^ f ^ gained some popularity in the
south where a few scholars have written sub-commentaries
on it. Thus, while the 14th century was dominated by TTip^f
and his son, the next century saw a healthy conflict, between
the approvers and opposers of tTNTs commentators till at
last the ^ e s ^ of sra^f gained the upper hand. This period
is considered to be the most glorious one for ^s^r^T^ studies
in Mithila till this glory was shifted to Nadia in Bengal in
the next century.
( 17 ) Early in the 16th century *T$3i*nwb? of Bengal
( circa 1480-1539 ) established a school of logic in Nadia.
His pupilage under zm& of Mithila, as also
Nadia School that of his disciple fitaff&T, are now considered to be mere myths. His father vi&xifstR? was perhaps the first Bengali scholar to comment on
therfor;and this commentary is referred to by the son in his
STglfc *s reputed as the teacher of four emi-

38

INTRODUCTION

nent men :^^^rf^Tt^Rfo the logician; ( ii ) C^OTH?^ theTantrist; ( iii ) ?tgT?c[r the Jurist and ( iv ) # 3 ^ , the founder
of the Vaishnava movement; all distinguished in their line
of special study. Though 3f$^ is famous as the founder
of the Nadia School of rfsq?^*?, it was his pupil f^ftfoT who
raised it to university status. He initiated a special chair of
logic at Nadia and was its first occupant. Similarly two of
his co-pupils are reported to have established chairs for
Tantra and law. To occupy the chair of logic one had to
write an original commentary on the jfr ( or later on, on
the '^tfrfr of fififa ) and also to prove in an assembly of
i earned Pandits the exceptional merits of sucha work by
his dialectical skill, feisfr himself wrote learned commetaries called ^ffr on ( 1 ) portions of the sftfTthat on the
srgH portion being the most widely read. ( 2 ) sn^crccjh
fir of 33fq-Tr; ( 3 ) np^TWtJ^T5 and ( 4 ) ssteracftsrarer both
of ^fiTW, besides a few independent tracts such as q^m^rtf, anwraqriq, etc. In course of time as the number of
students increased, eminent professors seem to have
opened separate seminaries on their own and thus Nadia
became an important centre of learning, especially for
.logic. These seminaries produced some of the best scholars
of Indian logic like |<J[3T33TO*ita, ^R^stero, W^fl^ih?,
*W?^, florero, ?pFF?, Si^ter and rt^re?, to name only a
few outstanding figures. In Baaras also the study of the
f^wf and its commentaries flourished side by side and
we hear of such able scholars like iHfHfi, ^^TOsnraiqfw,
qSHRWT. **j$2f SWT*N^TSTS, etc, from that place, who have
all commented on the i%3cmfT; sat^ and/or ^fRi. The
efforts of all these brilliant scholars could not, however,,
supplant the sTT^t^ of stq^g which still holds its position
as the first important commentary of the f./ The fact that
almost all the Nadia scholars who have commented on the
^tffo of fsrif including sm^Nr and TT^m?, have also
invariably commented on the sn^fcs as well, would go a long
way to prove the merit of the latter. But for its interminable involved phraseology shrouded in a maze of intricate

INTRODUCTION

391

technical terms which to the uninitiated appears only as a

"vast mass of perverted ingenuity", this voluminous lite*
rature of the Neo^logicians would have attracted a wider
circle of readers. This age of expansion, however, ended
with zr^ra? ( 1603-1709 ), the prince of Indian dialecticians".
From the 18th century the Nadia professors rarely attempted
any commentary on the ifn or the sftffr, but confined themselves to writing short tracts on select individual topics..
These are called ccfeq^'s, most of which have not seen the
light of day. At the present time H^i^ and ^a^ST are more
popular in the North and TT^R in the South. Even so only
five or six select sections from their commentaries on the
chapter, such as stntgq^?, sqf^mOTqf^Wn^j.
^ R E , etc. are all that are taught and learnt in the
seminaries or Sanskrit Colleges. For other' portions of the
subject the teacher and the taught both have recourse to
the fo^crjtETsife of firaiTW ( mentioned under tT^sr^oT ) and
irs advanced commentaries.
( 18 ) The ramrai of&srafmis, as already mentioned, a
which explains the subject matter in an easy language.
The author does not enter into any detailed
discussions on points of difference with other
systems; but at the same time he discusses briefly
the view points of the Buddhists and fhiH^'s on important
topics just to show the logicians' dialectical methods. These
digressions serve well to relieve the otherwise monotonous
nature of the narrative. At the very outset he quotes the first
Stra of qfasf enumerating the categories. The exposition
of the four OTTO'S is spread over half the book and that of
the second category sta covejrs the major portion of the
second half. The other fourteen categories are just noticed
very briefly in the last few pages. The author's scheme of

40

INTRODUCTION

incorporating the tfsp categories will be clear from the

following chart.

reiser ^
1

23

4 5

10

11

12

7 8 9
^^5 snfc
14 15

13

16

I
1

J
I

gar
1 2

4 | 5

6" 7

I .

q p ^ HfT^
3
4

J
md^i H?3T*T
5
6

10 11 12

II
7

While describing the seven categories ^szr, gar, etc. of the

if system the author emphasises on important doctrines
of that system such as the atomic theory, the origin of the
notion of duality, fj^r, qra theory, fawwsffvrm, etc. In all
these matters he closely follows
(19 ) The author styles himself only as sra in the body
of the book, vide the concluding verse at the end of the vm\w
section, though in the colophon at the end
Kesava Misra
of the work it is mentioned as ^f ^^^fT^rand
quftm grqbmr H^BT. The word feT suggests
Commentators that he might be a native of Mithila,
Internal evidence is also meagre as he does
not quote any author except ^^^^T^T besides the N. V. S.
and their VTT^'S. He shows marked predilection to 33SR by

INTRODUCTION

41

the way he tries to defend the latter's definition of srfrf^r( vide para 136 ). Dr. Peterson in his catalogue of Ulwar
MSS. ( 1892 ) mentions that a C^VRTST^IT by WffTO together
with a sub-commentary on it by ^%^T is preserved in thelibrary of the Maharaja of Alwar ( vide H. N. N. M. p. Ill )..
If this is to be relied upon*no one seems to have seen and
described these since Peterson's notice of them-sn^T^T would
be the first author to comment on r^^T. This fact may
also point to sj^fsr's "place of origin as Mithila. He is.
supposed to nave lived in the later half of the 13th Century..
No other work of his has come to light and he has perhaps
not written any. After ^rffffl, the next early commentator is.
*J*TS who was a protege of king Harihara of Vijayanagar
( 1377-1409 A.D. ) whom he mentions at the end of his.
commentary. He is unaware of the great upheaval in logical!
studies in Mithila under TifpST. The large number of commentaries on the cfqroraTsome 28 are enumerated in the B. S. P.
S. editionand the great variants in the textnearly half"
the space in the text portion of the above edition covering
the 1st 96 pages is demoted to the variantsis evidence of
the popularity of the book in the different parts of India. All
the same, it is surprising to find that this book was not so
popular in Mithila itself as in the other parts of India. In.
the south it was being used as a beginners' text book till the
appearance of the 33m of ararog in the 17th century.
Among the commentators as mentioned in the Bombay
edition arefound names ofsuch scholars as nitavrg, f^leR^fg,
ftif'c, and iftqhrra. None of them appears to be from
Mithila except perhaps itqtam. Dr. Bhattacharya says that
he belonged to Mithila ( vide p. 64 of his H. N. N. M. ) while
Dr. Vidyabhushan states that fiIhe was a Maharatta and lived
in Baaras during the time of Raja Jai Sing" ( vide p. 485 of
his HiL ). The commentaries of %TOg and iTtqteTST were more
. popular in the South than in other parts as more copies of
* It is regretted that no reply could be elicited from the present,
custodians of this collection even after repeated reminders.

42

INTRODUCTION

these two are found there than in the North. The former is
to the point, free from ^sq^ra terminology, and generally
easy for the beginner to follow without outside help. Tftqtoro's
is a weighty and learned one, and though written under
the sr^rm influence like his 3#TTi?, he has tried to be as
clear as possible. This and the commentary o f ^ m (if
available now ) deserve publication. Two other commentaries
were published long ago, that of 37icn?r from Poona in 1894
and that of f^^OT^.from Baaras in 1901; but bgth are long
out of print and 1 have not been able to see either.
( 20 ) The tQxt of the rT^^HT followed in the present
edition is mainly based on a transcript which 1 had made
for ray use as early as 1918 from a palm-leaf
The Present MS, in Grantha script. This MS. belonged to
Edition
a learned Pandit, Shri Ramasubba Dikshitar,
a native of Tinnevelly District and a friend
of my Guru's father. Shri Dikshitar, I was told, came to our
village* in connection with the renovation of a temple and
after that func:ion was over, he was prevailed upon to stay
there for teaching Sanskrit to the youngsters of the family
of the local Zamindar. He stayed in our village for about
eight months and passed away there itself in September
1838, leaving behind his 3 or 4 Mss. One of these, a fairly
bulky one 14" x2J", contained the cf^mT in 18 leaves, the
commentary on it by *Wtap in the next 53 leaves and then
a portion of the commentary of^g^nsup to the beginning
of s^sRef^ ( ending with ^ sjsmg) in 21 leaves, and
then there were 25 blank leaves. It was written on both
sides of the leaf, 8 ( and sometimes 9 ) lines on a side and
not in a verv legible hand. The MS. was pretty oldat least
about 150 years old even in 1918the leaves had become
brittle and smoky at the edges; and it showed * signs of
constant handling as there were corrections in between
lines and in the margins on almost every page in different
* This village, Chend aman gal am,, Is situated at the mouth of the
Periyar river some 20 miles to the north of Cochin.

INTRODUCTION

43

hands. My transcript was carefully corrected by my Guru

whose photograph adorns the frontispiece here. It was
later used by a friend of mine, a Namboodiri from a village
near Ottapalam. He had compared it with two other MSS.
in Malayalam script got from two different sources. There
was no appreciable change in the readings except for the
addition or omission of such words as i?f, gt ^9 erg*, %f,
etc. On comparing this transcript with the text in the BSPS
edition of the cTcfcrRT it is found that it agrees in general
with the MS marked C in that edition. The text presented
here can therefore, I believe, be taken as the Kerala recension
of the 33TOreT. When it was decided that these notes, which
were originally given to certain students of the Delhi
University who approached me for private tuition when the
"book was first introduced as a text book for M. A. Classes,
should be printed in book form, I made some efforts to get
the above mentioned MS. from my village in 1969. But I
was told that all the MSS. there had been transferred elsewhere. So the original idea of including the commentary
- of atqtaTO was dropped, Besides the Bombay edition, I have
seen three other editions of this primer with detailed glosses
in Hindi.
1. With Hindi commentary1 by ^re*fMte?printed at
Baaras in 1953.
2 With Hindi commentary by srvtaraQgfprinted at
Delhi in 1968.
3. With Hindi commentary by ^^H^T%^printed at
Meerut in 1972.*
All these editions as also the English translation of
the book by Dr. Ganganath Jha ( published from Poona
* These three together are called A editions as there is pratically no
difference in their texts and the Bombay edition is called B. Another
edition with Sanskrit and Hindi commentaries published in the Haridas
Sanskrit Series from VrnasT ( 19 77 ) came to my notice after this Ma.
was sent to the press.

44

INTRODUCTION

in 1924 ), though it came to my hands at a very late stage,,

were useful to me in various ways. Noticeable differences
in readings of the text in these editions have been referred,
to in the notes here and there.
( 21 ) I shall be failing in my duty if I do not record the
help I derived from two similar publications, viz. ( 1 ) the
edition of cranc ( Text only ) by Dr.
Acknowledgement S. Kuppuswamy Sastri printed under
the title

A Primer oj Indian Logic*'

from Madras; and ( 2 ) the edition of the same book with

two Sanskrit Commentaries by Mr. M. R. Bodas in the B.
S. P. S. 1 am highly indebted to these two for the choice of
many an appropriate phrase in English and for deciding
how muck explanation of a particular topic should be in-.
eluded for a beginner in the subject, keeping in view the
back-ground equipment of a modern college student.
Suggestions for correction and improvement by readers of
the book will be gladly welcomed. My sincere thanks are
due to the famous publishers Chaukhambha Orientalia,
Varanasi, for readily agreeing to publish this book in spite
of the fact that books with translation and notes in Hindi
are more popular among students these days. I am also
extremely grateful to Pandit Shri Gaurinath Sastry, Ex. ViceChancellor, Baaras Sanskrit University and the present
President, Sanskrit Sahitya Parishad, Calcutta for kindly
consenting amidst his various other duties, to write a foreword to this attempt after going through the proof-forms
and for suggesting improvements and here words would
not be adequate to express my gratitude to Dr. C. R. Swaminathan,the present Principal, Sri LalbahadurKendriya Sanskrit
Vidyapeeth, Delhi, but for whose unstinted help an
guidance this humble attempt would not have reached its
present stage.

S. R. Iyer.

( 1)
( INTRODUCTION )
[1] sTMsfq ifr r
TranslationThis ( work ) Tarkabhsan exposition of
reasoningis brought forth by me for the sake of such
beginners who do not want to take pains and to exert, but
desire to enter into ( i.e. learn ) the Nyya system with a little
hearing ( i.e. learning from the teacher ). ( With this in view )
the subject matter has been condensed ( as much as possible.)
retaining all the important arguments ( for proper understanding ).
NotesThe word n\& here and in similar contexts means
a beginner who has the necessary background, cf. 3T3JT?rt
fra aski: ( T. S. ).
logic or reasoning;
by which the desired meaning is brought out clearly.
tth rqprcn&^TOT: w * ? a sRnT^a orfwfirfer a^tarar. The
arguments or reasonings on which the science of logic is based
are expounded here. i%q or condensation can be effected in
two ways, ( 1 ) by omitting verbal elaboration and ( 2 ) by
leaving out some ( items ) from the subject matter. By the two
words ^fe^andg^qf^c, the author implies that he haso mitted
only verbal elaboration and not anything important from the
subject matter. The latter would mean that only particular
portions of the subject have been explained, thus detracting its
usefulness as an elementary text for beginners.
1. A read f ^ though Gh. B. reads only
2. G reads f l R ^ g thouh Gh. B. has as printed here.

TARKABHAS (

It is customary with Sanskrit writers to indicate the scope

and purpose of their work before beginning the subject proper.
This usually consists of four parts collectively called arg^wT. ( 1 ) fsrw = the subject matter or scope of the book; ( 2 )
s the purpose; ( 3 ) aifo^TR^=the proper person for
whom the book is intended; and ( 4 ) H ^ ? ^ = relation. Here by
the word ^spwT^T the: subject matter, by nsRO girsT the purpose, and by ^T^5 the proper person for its study, have been
indicated. The relation, which in almost all such cases is
srfcTOsiSifcTCRpwrarthat of the treated and the treatiseis
implied in the word sraiw1. This last one is necessary so
that the author should not stray away from the subject proper
by digressing upon side issues, thus detracting attention from
the main point.
After introducing the reader to the scope and aim of his
work, the author directly enters into the subject by quoting the
first sutra of Gautama in his rm^?, which is the foundation
stone for the Nyya philosophy.

[2]

Translation"TOTUT
W
is the first among the Nyya
Sutras of Gautama. Its meaning is that by a correct and
true understanding of the sixteen categories of substances
beginning with srcnar one gets liberated from this world of birth
and death.
1. 5ITTT^ = Means or instruments of valid knowledge.
2. sito = Obj ects of valid knowledge or cognition.
3. H5T*T = Doubt.
4. q2t3R = Purpose or motive.
5. s 2 \ ^ = Example.
6. fcn5"Rf = Established conclusion.
7. srg?r^ = Members of syllogism.
8. TC = Rednctio ad absurdum.
9. foF = Decisive knowledge.
-.
10. 3^ sa Discussion ( to arrive at the truth ).

 11. 35^f = Argument ( both constructive and destructive) or

disputation.
12.fc^ST= Mere destructive argument or wrangling.
13. lqri^rf = Fallacies in reasoning or fallacious reasons.
14. [3? = Quibbling or specious argument.
15. *nf = Unavailing or futile objection.
16. fracture = Vulnerable point ( clincher ).
NotesGautama ( also called Aksapda ) is traditionally
believed to be the founder of the Nyya system of thought. It
would perhaps be more correct to say that he was the first to
codify the principles ofthat system in what are called Nyya
Sutras. Sutra here means a short pithy saying with the
minimum words for expressing a proposition or statement,
modelled on the Kalpa Stras and Panini's Stras on Grammar.
A popular definition of stra widely accepted is as follows :
f<g: It
A sutra should have the minimum words and should not
give room for doubts. It should be meaningful and capable
of covering all aspects of the point mentioned by it. It should
not contradict any previous or succeeding stra and should
be blameless in all ways.
The author, following Gautama, begins to explain these
sixteen categories with the minimum discussion and arguments^ The usual way of arriving at a conclusion is by putting
a question in the form of a doubt or objection which is called
^T^r-the opponent's viewpointand then stating the accepted
conclusion in the form of an answer, called
[3]

?r =ET 3TOT^

5T|ftT:,

1. A has
2. A & B have V ^ ^ W instead of W .

TARKABHASA (
Translationg^^ffi or correct understanding of these categories beginning with V[V(\T\ cannot be had unless their ^G^5
or enumeration, <5^OT or definition, and qO^T or analysis are
done. This has been said by the Bhsyakra, commentator on
the Nyya Sutras, ( Vtsyyana ):"The procedure in this
astra is threefold viz. enumeration, definition and analysis."
Enumeration is the mere statement of the categories by name
and this has been done here in the sutra quoted above.
Definition means the pointing out of their distinctive
qualities or properties; e.g. that of a bull is that it is an animal
with Qnm( = dewlap ) etc., as its distinctive characteristic.
Analysis consists in the examination as to whether the definition fits in with the thing defined. The first having been already
done, the other two, definition and analysis, have to be done
now for a correct understanding of mxvn and other categories.
Notes Enumeration also includes subdivision ( fevun )
which is only a statement of the classes into which a thing
could be divided. This is usually done after a thing has been
defined, as it is necessary that all the subdivided classes should
have the ^ o r or characteristic of the main thing; e.g. S^IOT is
first defined as the instrument of valid knowledge. Then only
arises the desire to know how many such instruments are
there and hence follows the division or classification of vmm
into four kinds; then the definition of each kind and their
subdivisions, ^ o r has been translated as definition for want
of a better word. The definition of cs^or itself is STcn^TRffl\ro; i.e. a specific characteristic which enables the thing defined
to be differentiated from other things as in the example given
above. This happens when the ^&r is free from the following
three defects usually associated with it. ( 1 ) ^%5^ig= overapplicability; i.e. the proposed ^ c n is applicable to things other
than the one defined (^s^cRcRR^ )3 e.g. if the ^^m of bulls
is stated as J%[ ( = the quality of having horns ), it is applicable to horned animals other than bulls and so goes beyond
the ^ 5 ^ or thing to be defined. ( 2 ) ai5^ifg = partial applicability ( 35^iR^t 3^T*3[ ); i.e. the &^m covers only a portion
of the 35^; e.g. if qjfq^ec (the quality of being brown coloured)
is stated as the ^oTSfor bulls it is applicable only to brown

coloured bulls and not to bulls of other colours. ( 3 )

total inapplicability ( ss^qi ar^cR^ ); i.e. the %m is not
applicable at all to the &*m; e.g. if q$?ra3C3 ( the ^feature of
having only one hoof ) is taken as the ^TOT of bulls, it is not
applicable to the bull species as they are all double-hoofed.
? Hence the 35^11 of a bull which is devoid of all the three defects
is 3TCHft[*ft3 (having dewlap etc ).
The purpose of having such a strict 3>^& is mainly differentia
ation and in some cases proper designation. ( in^f%3?fein^ 3T
35^oH5?T snitsR); e.g. 'The quality of having smell' Trarc?3 is
stated as the ^ o r of earth ( <j3 ) and it serves as a valid
reason for differentiating earthy things from non-earthy things;
for smell is the specific quality of the element earth and does
not abide in anything other than earth. What helps in differentiation also helps in specific designation.
The examination of a ^ m to find out that it is devoid of
all these defects and that it thus serves the purpose for which it
is proposed, constitutes the third and most important process in
Nyya. Hence its impact on the other systems loi thought
like Mimnis, Vednta, etc. All other systems adopt this
method of analysis.in the elaboration of their topics.

( 2)
[4]
1

vvii:
TranslationSo the 35^aT of signar, the category mentioned
first in the sutra, is being stated, g tiro is the instrument or
means of valid knowledge or right cognition ( SRT^OT ). Here
( the word ) SI^RIOT is the 3S^sr (object to be defined) and sjTn^&i
is the w^vi. ( Here the opponent puts in a query ). If SW^T is
the ^OT of valid knowledge, its result has to be stated, for as
a rule a ^^UT is always accompanied by a result. ( Answer )
True. The result is valid knowledge itself, i.e. the thing accom*

TARKABHASA (

pushed by the instrument, just as the result of the axe, the

instrument for cutting, is the cutting or breaking up itself.
Next the two v/ords in the ^rsT, sm and ^or, are being
explained.

[5]
sn=rr i
TranslationWhat then is SIH whose instrument is
means valid experience, i.e. an apprehension which accords
with the true character of the object or thing apprehended. By
the use of the word srq4 ( valid or real ), cognitions such as
doubt ( H^^T ), misapprehension ( fjqsfc ), and hypothetical
apprehensions leading to rednctio ad absurdum ( ^r^ ) are
avoided as all these are erroneous or faulty experiences. By
using the word srg^re ( in the definition of sm\ ) remembrance
( ^feff ) is excluded. [ After this some printed editions have
another sentence ^T3if%q4 gr ^(CT:Remembrance is cognition based upon what was experienced previously], s?g*ra
here means all cognitions other than remembrance.
NotesAfter defining *xm as ?reTO<f%rc the author
examines it to show that it is devoid of the defects mentioned
above. If we do not qualify eigvf^ by theadjective *??p-i, i.e.
if we say s*gsw $;RT, the OS^U will be applicable to doubt etc.
which are not 35^*r. Therefore the flaw of overapplicability
^Rrs^fH will crep in. To avoid this the word q^m is used by
which doubt etc. arc excluded from ss^sr as they are not valid
( s[snr ). Similarly the word ^tm
is used to exclude jgfc or
remembrance from the scope of the definition. Lastly to make
the definition more explicit the word sfg^ra itself is explained
so as to include all knowledge other than ^ % .
Next, the second word ^im in the definition of STRT is being
explained :-

[6] fp J T :
I

TranslationWhat does sp^or mean ( in the definition of

) ? It means the most efficient cause. That which is most

instrumental ( in bringing about the final effect ) is

and hence ^or means the most efficient cause ( = instrument ).
1
NotesThe definition of ^ T in the text is only a repetition
of Panini's Sutra, 1-4-42. Another definition usually adopted is
c
*SP3R0T 3JiTi ^OT^'that which is the most effective in producing the effect out of the various causes is ^ar.
The
author alludes to another definition of 3^* later on when he
defines STcsr^R^oT, vide notes under para 24.

[7] ;rg m%

Translation ( In the above explanation ) ^ ^ and spRf are

only synonyms and it is not known what a Claris. This is being
explained. .That which invariably precedes an effect and
is unconditionally necessary for it, is the cause of that
effect; e.g. the threads, loom etc. are the causes of the cloth.
When a cloth is being made, a donkey ( the weaver's own )
may be present there accidentally and so it exists before the
effect ( the cloth ). [ Therefore the donkey also may be considered a cause for the cloth ]. But its presence is not invariable
( fnra ). [ So it is not a cause ]. Again, the colour of the
thread is invariably precedent to the cloth. [ So let that also
be a cause of the cloth ]. But it is not an unconditional
necessity ( sr^zrerrf^S ). It is conditioned by the fact that the
colour of the threads is a cause for the colour of the cloth, and
if it is said to be a cause for the cloth as well, this would lead
to unnecessary assumptions. Thus causality (=ERtJKcO is defined
as invariable and unconditional antecedence to the effect; and
effectness ( ^W3 ) as invariable and unconditional consequence of the cause.
NotesFrom the above definition of a cause it is clear that
a cause should satisfy three conditions. (1) It must precede the

10

TARKABHASA ( a

effect ( qfcnf ). (2) Its precedence should be invariable

( fw ) and not accidental as the weaver's donkey etc. (3) It
should be unconditioned ( SR^sf^ ). This RtTOTf^,.
when referring to a cause, means that which could not be
c
made out to be otherwise than indispensable'. This reduces
to the fact that the invariable sequence between cause and
effect should be made out through the invariable concomitance
( strife ) between these two directly as also between their negationsin other words through a knowledge of sr^sr ( affirmative reasoning ) and srff^ ( negative reasoning ); e.g. A may
invariably precede an effect B; still, on analysis, it may be
found that A is not indispensable for producing B. If A were
to be a cause it should be amenable to the affirmative concomitance (arssptsq'Tfg), "If B comes to effect A should be present";,
as also to the negative concomitance (sqf^qpwnfO, "Since A is
not present B does not come to effect". In the case of the donkey,
whether it is there or not, the cloth is produced ( when all the
other causes are present ). It is thus wqsnfe;? and therefore
not a cause.
Later Naiyyikas have classified all possible dispensable
antecedents ( s^trEnfg^r. ) under five heads.

^t, 19-22 )
Dispensability can be judged in five ways :1. zm
rra:-Thread ( cf?g ) is the material cause of cloth.
This thread is known only with its delimiting adjunct 'threadness' ( cT=g^ )* Therefore the latter is known as an invariable
antecedent to cloth only through its counterpart the thread.
Such a thing ( gr?gcsi ) cannot be considered indispensable for
the effecthere the cloth.

V t

2. ^uui'"^- Any thing whose anffecedence can be known

only in relation to some other known cause, cannot be indispensable for that effect; e.g. the colour of the thread is an
antecedent only by virtue of its substratum and not on its
own. Therefore it does not come within the causal apparatus
and so becomes dispensable.
3. sjyq
SETrwWhen the antecedence of a thing is made
out only after it is known to be the antecedent of some other
effect, that cannot be taken as indispensable for the effect in
question. Ether ( BTT^T^T ) is known as the inherent cause of
sound; and if it were to be a cause of cloth its antecedence can
be known only after the former antecedence to sound is
understood. Therefore it becomes dispensable to cloth.
4. 3Rcp"**jna:When the antecedence of a thing can be
made out only through the person behind the effect, that becomes dispensable; e.g in making a cloth a weaver is necessery
but the weaver's father, though antecedent to the cloth in
question, comes into the picture only through the weaver and
not in his own right and is therefore discarded as dispensable.
5. arfcT"'*^^All things extraneous to the bare minimum
necessary for producing the effect must be considered dispensable; e.g. the donkey which has accidently strayed into the
place of weaving (already explained in the text).
This classification need not be taken too literally; but
only as an explanation of the different methods of determining
dispensability. A little thought will make it clear that the
difference between the first and second categories, as also between the third andjfourth, is too slender to deserve separate
classification. Moreover the fifth class is so comprehensive
that it includes all the other four. That is why the author of
the above verses concludes by saying that the last variety is
the most important and necessary one.

[8] qx^

Translation -Some one has defined cause as 'that which has

affirmative and negative concomitance with the effect.' This

12

TARKABHASA (

is not proper because in that case the eternal ( fkm ) and allpervading (firg) substances, viz. time, space and ether, will turn
out to be not causes-in-general, as their negation both in
respect of time and space is not possible.
Notes3T?3*r and sqfnt^ have been explained earlier. -Time
( *&]& ), space ( % ) and ether ( T^^T ), which are eternal and
all-pervading, are considered to be causes-in-general ( vide
next para ) for all effects. Since they are eternal their negation
in any form is not possible and therefore they cannot have
negative concomitance ( ssrfa^Rs^ifg ) with the effect,'and so
this definition will not cover such cases. Therefore this
definition is open to the defect of partial applicability (arsqifg).
The objection is only to the definition of cause in this form
and not to the fact that there should be invariable concomitance between cause and effect. This definition of cause is
accepted by the Mmmsakas and some Baddhists.

[9] T^ EpRof ffsr, W T T ^ S W f a ^ ^ S T e r I p

TranslationCause is of three kinds; inherent cause

), non-inherent cause (3Rcnfr^nCtfT) and causes-in-general
(fafiTR&T). Of these, inherent cause is that in which the
effect inheres when produced. Cloth inheres in the threads
and not in the shuttle etc. Therefore the former is the inherent
cause of cloth ( and not shuttle etc. ).
NotesThe inherent cause is usually the material out of
which the effect is produced and corresponds to the s^^ffi^Ro
of the Vedntins. It is also translated as consequent cause
and the other tv/o as non-consequent cause and efficient cause.
Here a new relation inherence ( HR3TO ) has been introduced and so the author begins to explain the same before
proceeding to explain the other.causes.
[10] ?T

TranslationQ,. fHere cloth has relation with the shuttle

etc. as it has with threads. Then how can you say that cloth
is produced as inherent in threads only and not in shuttle etc, ?
Ans, It is true ( to some extent ). Relation is of two kinds,
mere conjunction ( H^^T ) and inherence ( girare ). The relation between two inseparables is inherence, and that between
any two others ( which are not so intimately connected ) is
conjunction or contact.
Here the word 'inseparable' ( srgcfVf^ ) is used in a technical sense which requires explanation.

[11]

5F1"

i W3T d

TranslationQ,. What are 'inseparables' ( sg^H^ ) ?

Ans. When two things are so intimately connected such
that one subsists on the other so long as the latter is not destroyed, they are called inseparables. This has been said by
others. "Those two things arc to be known as inseparables
if one of them subsists on the other so long as the latter is not
destroyed." Such inseparables are (1) the parts and the whole;
(2) quality and substance; (3) motion and that which moves;
(4) generality and the individual things ( having it ) and (5)
speciality and the eternal substances. The whole etc. subsists
in the parts etc. only till they are not destroyed. When the
parts begin to perish, the whole etc. does not subsist on the
other, e.g. when the threads perish the cloth is destroyed.
Similarly when the substratum ( i.e. the substance ) is destroyed its quality also is lost. Th state of being destroyed
) means existence of all causes for destruction.

14

TARKABHAS (

NotesThe relation inherence ( mmm ) is accepted only in

the case of the five 'inseparables' mentioned above. The whole
( QT^fcft ) rests on its parts and is regarded as lost only when
the parts disintegrate as in the case of the cloth and its parts,
the threads. In the case of the next two, quality and motion,
they get lost when their substratum perishes. But when the
quality or motion is destroyed ( e.g. when a white cloth is
coloured blue or red, or when the motion of a ball is reversed)
the substratum does not perish, but only inheres in the changed
quality or motion. Generality ( 5f or TOf^ ) is a generic
feature which inheres in all the individuals ( strffR ) of a class;
e.g. T[\m ( cowness ) is the generality inhering in all cows and it
does not disappear till the cow perishes. In the fifth case of
inseparables, speciality ( ffN ) is accepted only in eternal
substances and therefore the question of their destruction does
not arise. The use of the word sfsUcSH in the text above is not
happy as it cannot be interpreted in its ordinary sense fiin the
same order'; but has to be construed 'as applicable in each case'
{ as has been explained above .
Here, inherence ( ro^r ), generality ( HRT?^ ) and speciality ( tfrftq ) are categories adopted by the Nyya school from
the Vaisesikas. Kanada, the founder of the Vaisesika system,
has classified all substances into six categories. (1) ^sq = substance, (2) goi = quality, (3) ^4=.motion (or action), (4)
generality, (5) f^N=speciality ( or differentia ) and (6)
inherence. Later authors have added a seventh one to this list,
sT*7tc = non-existence or negation. The first six are therefore
spoken of as sn^rq^m: = existent or positive entities. The Naiyyikas have included all these under the broad heading
=knowables), vide Kesava Misra's own treatment under cs
in part II; (STKVTFI). That the relation between the table and its
colour is not the same as that between the table and the book
on it, can be easily understood. The former is called inherence
( TWq" ) and the latter is conjunction ( Nft*T ). It would be
easy to remember that in the above classification categories 2 to
5 ( FT, JT, nsfpq1 and f^tq ) inhere in the first category ^sq.
Also when one substance is produced out of another ( e.g. a
cloth from cotton, a jar from clay, a table from wood etc. ) the
relation between the component parts ( cotton, clay etc. called

15

) to the finished products ( the whole or awrf?r ) is

not the same as the mere contact between the hand and a stick
or the table and a book; and so here also the relation is inherence. These five cases, where inherence is accepted, are summarised in the above verse ( of the text.).
Generality or iTO^ is also called 3fr and is an important
category for the logicians. When we see different men or different
beasts, our experience, however diverse it may be in other respects, shows a certain degree of uniformity which takes the form
'this is a man' or 'this is a beast' (am H^W, wr *im). Thisunifor;
mity in our experience can be explained only by assuming a generic feature common to all men or to all beasts. This generic
feature is wg^c^ (=manness) and ^r^ ( = beastness) which is
common to all men and all beasts respectively, however different
they may be in other respects. Thus wg^pg and Qm% are uniform generic characteristics of men and beasts, i.e. Qfn?ntt Sre:
or T W ^ . This generality is raised to the status of a category
( q^i4 ) as it will be easier to explain many factors, according
to the principle of economy in thought and words ( ^ER^TO ).
Briefly this is how the Nyya and Vaiesika systems recognise
generality as a distinct category. This is classed as high ( m )
and lower ( <tfq? ). ^ T is the highest generality as it inheres in
all the first three categories, while ^sq-g*, JPT^, 3W3 etc. are
lower when compared to rer. Similarly, e^fc^, 3 ^ 3 etc. are
lower generalities when compared to 55^3 and e^TT, but greater
( m ) when compared to qz&i, vzm, *t$m9 *gq^, etc. as these
latter inhere only in a smaller number of things when compared to the former.
This TRTT^ serves a double purpose; firstly to unify through
universals on the basis of observed uniformities, as in the case
of ritof ( cowness ) which gives an impression of all the individual heads under the term cow; and secondly to differentiate
wherever experience requires it; e.g. *\\m in particular cases
can also denote differentiation from all things other than
cows ( TiftR^ ). The first process of generalisation has led
to several generalities being recognised, while the second of
differentiation has resulted in postulating a self discriminatory feature called f^tq ( speciality or particularity ) being
postulated as a differentia for everlasting substances (

16

TARKABHASA (

Composite substances ( ST^^ ) like a jar or a

cloth can be distinguished from one another by means of their
component parts. Earth, water, fire and air are producible
substances-(TP^sqnfoi) and their component parts ultimately reduce to their lowest components, the atoms ( sr*3 or
TT*ng ) of each of these four. The ultimate atoms of these
four and the other five substances, STI3$ ( ether )3 ^t<? ( time ),
f[( space ), 3n**7^( soul ), and SRH (mind) constitute the
everlasting substances and a unique fta is ascribed to each
of these. Of all the Indian schools of philosophy it is only
the Nyya-Vaisesika that accepts this f?q as a special
feature. In fact.the term t^tfqsp for the school of q^oiT^ is itself
based on this hypothesis of fsrete- Since the atoms of the four
producible substances are alike in jpr, ^ and ^nfcf, there is no
criterion for distinguishing an earth atom from a water atom etc.
In order to account for their differentiation a fta or . speciality is attributed to each atom. Since atoms are countless and
eternal, fsf's also are innumerable and eternal, and their
relation to the everlasting substances ( as stated above ) is
inherence. They are accepted as self-differentiated as well as
self-differentiating as otherwise there would be endless regression ( 3RcR.SU ).
It may be shown that these three categories, OTRFST, ffrs
and CiTOT^ have been postulated to explain in their own way
the unity and diversity in the first three categories ( ^sq-, JJCTT
and EOT ) with reference to the principle of economy ( 351^ ).
Even among the logicians there is difference of opinion as to
whether all the eternal substances like ether, space etc. should
have each this firer. Each jiva with its mind has its distinct
experiences and other characteristics which could form its
differentia from other jivas. But liberated jivas ( gxET: ) cannot
have such differentia unless a ferita is ascribed to each, as all
such jivas are considered to be alike in all respects. Regarding
(ether) some logicians of the older school contend that a
should be ascribed to it as the delimiting determinant of
its causality to sound ( ^s^H^Tfo^TT'JRrrei^ ), while others
say it is not neccessary. Similarly ^ ^ and f%% are to be
differentiated by f^q if they are to be distinct categories. But
if they are classed under Iswara ( qsffif of the logicians ), as

17:

Raghuntha Siromani and his followers of ^ssrwnsr -do, ' this

acceptance of fsrere is not neccessary as a separate category,
as Iswara can be differentiated from all the rest; by His
eternal omniscience etc., without the aid of any man-made
It would be useful to have a brief idea of the views of
certain important Indian Schools of thought on this concept of
snfcT as compared to that of the Nyya School. The grammarians, besides using the term snf to denote a class attribute
as in the Nyya sense, use it for denoting caste, lineage and
adherence to particular Vedic branches. This latter usage
follows from their rules for derivation of terms denoting a
particular class ( cf. siqunf^TC under nf;^ ). 3iff according to
them, is not only a generic attribute, but is so essential to any
substance, that they call it SUTO^ ( = life-giver ). This idea is
well summed up by Bhartrhari when he says *?T% *ftt 3^<ta *fti
nnft:, *OTfacrwsn5*ft:-i.e., 'a cow neither becomes a cow
nor a non-cow merely because of its physical nature; it is a cow
because of its relation to cowness.' The grammarians do not
accept this relation as CUTOTO nor call it a separate category.
The Bhtta School of Mmmsakas accept snff asan eternal
and perceptible category, and though it is present everywhere
it is manifested only through a ssffo. But the relation binding
snfr to the 5irf%3 is not TTOPT. They call this relation 3T3*ur
which according to them means identity-cum-difference; $5Hf^T^ or identity compatible with difference. In a statement like 'this is a cow' ( ar ft: ) 'this' denotes the particular
siff and 'cow' denotes the snf cowness. Thus in this statement a snfr is equated to a s^rfs; but the question cannot be
absolute, as in that case 'this' .and 'cow' would become synonyms which they are not. Therefore they argue that on the
strength of the preceptual experience the relation 9T$3KT should
be accepted in such cases though, ordinarily difference is
opposed to identity; for after all compatibility or otherwise of
two things is determined by experience. Also, unlike the
Naiyyikas who ascribe craTO or m\fh to the first three categories only, the Bhftas ascribe the highest rfcf viz HTCI to
also besides to the other three,
2

18

TARKABHAS (

The Advaitins, who generally follow the Bhttas in such

matters, accept the Nyya theory of snff to their profit and
show that their highest snf, w$h is the one grand ultimate
reality called Brahman and that all the lower ones like TCCT,
3T>3c3 etc., are only appearances superimposed on this grand Httt*
The Prabhkara school of Mimmsakas accept Sffiai only
in perceptible things and not otherwise; i.e. while irrtet, ^3c^,
etc. are accepted as life's, greater generalities like
oT^ etc. are not accepted as such. They however
accept R^T^f as the relation between two inseparable things,
but consider it eternal only when the related objects are both
eternal and not otherwise. They dismiss the relation
as an impossible jugglery of words.
The Buddhist and Jain philosophers do not accept
e*7cT and fiN as separate categories and even in their refutation of the views of other systems there is nothing by way of
constructive criticism.
It is necessary in this connection to know the principles
which determine which of the attributes should be treated as
sufr's and which otherwise. This has been summarised by
Udayanchrya, one of the greatest logicians of India who
lived in the early part of the 11th century, as follows :-

( 1 ) rki^: = unity of the object, e.g. the sky isa single allpervading category and therefore ar^^r^ cannot be a 3f.
Similarly ^^c^C and f^c^. ( 2 ) s ^ F / ^ ^ = Identity of the
objects though the words may be different; e.g. ^(im and
g^c^ cannot both be snfc's a s ^ a n d wg^, the safe's denoted
by these attributes, are only synonyms for the same object.
( 3 ) 5T5P*. When two attributes are such that while they
exclude each other in some places they exist together at
another place, the defect is called gsR or nfeft; e.g. s^ccf exists
in the five qji's ^T", 3f, ^, m% and n^T^, and qjfer exists
in the first four "^'s, and also in n^ravThus these two attributes
co-exist in the first, four ^fi's but exclude iach other H
and SRH. This defect therefore prevents both sjje^g1 and

19
being accepted as suf's. ( 4 ) ^ r a ^ H = Lack of finality, or
endless regression, e.g. a fRr cannot have another 3?fcTf for in
that case, there will be no finality for, each time we add one
more m\ht it will require another one over it. ( 5 )
Violation of the distinctive feature of the object, e.g.
cannot have a snf,fr$T<?3,since by hypothesis they are selfdifferentiating and cannot have another generic attribute to
distinguish them. ( 6 ) Want of proper relation is sigssriqr. If
sr^fgfrcf is accepted as a sffr we shall have to say that it rests
on its substratum sr^m through the relation CfHgi3T which is
impossible. Similarly OT^fsrtf also cannot be a
From the above it is clear why SR can exist only in the first
three categories ^sqr, gar and za. This raises another question,
What is the position then of such common features as SaTCcr,
ftT^3 etc. ? Such of those attributes which cannot be classed
as snf's due to any of the preventives (n^raO mentioned above
are called snifTs. Even among tarqifr's some logicians make
a classification as psfa and 3??r ^qff's. CT3*tafr's are those
where the attributes can be explained by other means without
reference to their substrata, e.g., Jjftccj is explained as
= being the substratum of actions; arraTO^ as 3TS^^
the determinant factor in the inherent
cause of sound. Moreover the relation subsisting between them
and their substrata is not a direct one as HHWT orr^T. They
are said to be W H i H * ^ or indirectly connected, sr^atqfVs
are those attributes which do not need another definitive
expression for their understanding and which are connected
to their substrata by a direct relation. Only those attributes
which are connected with their substrata by means of R^FT
anj which do not come under the preventive conditions mentioned above, become aria's; e.g. ifft^, ^ ^ etc. All other
attributes where the relation, though direct, is not HflcfnT ( e.g.
W ) are _ simply called a^^qfVs, e.g. fqq^r = object;
jectness ) is its characteristic; and T%qsrcn?qr ( = being
objectness ) is an snc^irqifer and its link to fma is self-link
or ^r^qw^sr"i.e. the related thing itself constitutes its
relation. Similarly gfosrtfraT^ ( = being correltiveness ) is
another $^\qifw.'.lt would be advantageous to understand the

20

TARKABHASA (

subtle differences among nf's and sqif's at an early stage to

understand the laboured expressions of later Naiyayil<as.
Though almost all Naiyyikas use the word ?m^ as a
synonym for snf, that term really includes ^qif's also. The
division of HTW^ into m and srq^ relates to only suia's and
not to 3"qfi's. In fact, the e^^^ of Sivdiiya divides
into snfcT and sqifa and the Pirn's into m\, wmj and
. Ordinarily it is only the *nfh and o^aatqif that occur
in our perceptual experience and hence the greater importance
paid to them.

[12] ^q-T^^q-^T^f^ft 1 cT

*&: i

i g gq-f i v[z?3 &mwt:

Translation( In the present example ) the threads and the

cloth are the component parts ( ar^sre ) and the whole thing
(sR*f5R) respectively. So their relation is inherence, they being
inseparables. The relation between the cloth and the shuttle
is not inherence, because these are not inseparables. The
shuttle does not exist in the substratum of cloth, nor the cloth
in the shuttle. Hence their relation is conjunction ( fftn )
only. Thus it is in the threads that the cloth inheres
and therefore threads are the inherent cause of the cloth and
not the shuttle etc. The cloth is the inherent cause of its colour
and other qualities ( as they inhere in the cloth ). Similarly
the lump of clay is the inherent cause for the pot, and the pot
itself for its colour etc.
Here the opponent ( <^qf%^) questions the validity of the
last statement that the pot is the inherent cause of its colour etc.

[13]

2T

TTORT sraft V

r faf ^ fet

Translation( Objection ). The colours etc. of the pot are

produced at the same time as the pot. Therefore there can be
no effect-cause relation between the qualities and the qualified
as they are produced simultaneously, just like the right and
left horns ( of a bull ) and as there is no sequence between the
two ( i.e. the pot and its colour ). As such the pot etc. cannot
become the inherent cause of the colour etc. in them; for the inherent cause is a specific cause.
This is answered as follows:-In fact the qualities and the
qualified are not born at the same time. But all things are
produced at first without any quality. Only afterwards qualities
are born as inhering in them. If both were to be born at the
same time they would have the same causal conditions. In that
case there would be no difference between the two ( the pot
and its colour ) . When two things are different it is necessary
that they must have different causal conditions. Therefore in
the first moment the pot is produced without any qualities and
is therefore antecedent to the qualities. As such the pot is the
( inherent ) cause for its qualities and there is difference in the
causal conditions as well. The jar cannot be its own cause as
no antecedence and consequence are possible with regard to one
and the same thing. A thing cannot become its own antecedent and at the same time its own consequent. Therefore,
being antecedent to its qualities, the pot becomes the (inherent)
cause to them.
NotesThe objector at the outset supposes that the pot and
its colour etc. are produced at the same time, in which case the
necessary condition that the cause should be antecedent to the
effect will not be fulfilled. The answer is that the supposition
of simultaneousness is wrong; that all things at the first moment

22

TARKABHASA (

of existence are born without any colour or other qualities and

therefore the condition of causal antecedence is fulfilled.
This raises another question which is put as the next objection by the opponent.

[14]

Translation( Objection ), In that case the pot in its first

moment of existence cannot be seen as it is a thing without
colour, just as wind; for ( the rule is ) that only that substance
which has the qualities ^m ( greatness ) and s^c^q ( manifested colour ) becomes visible. Moreover the pot will not be a
substance at all as it is not the abode of any quality. The
definition of a substance itself is that it should be the substratum
of some qualities.
AnswerWhat you say is true. But what hann is there if
the pot is not visible in the first moment of its existence ?
Even if it is conceded that the pot is born along with the
qualities ( colour etc. ), it does not become visible at the
moment of winking of one's eyes. It is therefore established
that at the first moment of its origination a pot is born without
any qualities and that it becomes visible from the second
moment onwards. Nor can it be objected that the pot is a nonsubstance at its first moment of origination for want of being
the abode of qualities. A substance can be defined as that
1. B. has the word qiJ^ before ^Z'. and Ch. B. approves ofthat
reading.
2. B. reads this sentance as JpnsR^ 3 ig
which is evidently an interpolation from Gh. B's. com.

23

which is an inherent cause. The other definition (being the

abode of qualities ) can also be justified in the case of the pot
at the first moment of origination, as it is capable of being such
an abode. Such capability here is only the absence of absolute
negation of qualities.
NoiesWhen a substance is defined as the substrate of
qualities what is meant is that it should be capable of being such
a substrate and it does not matter if at certain moments of its
existence it has no qualities for it takes up qualities at the second
instant and therefore absolute negation of qualities cannot be
predicated.
Here the sentence ( ^f%
Tilia ) within brackets is not
commented by Chennubhatta and perhaps does not form part of
the original text though the printed texts and some MSS. have
it. Moreover it has been alluded to previously and answered
( vide the sentence CTHT^rafcT^ etc. in the previous paragraph. )
After explaining the inherent cause in some detail, the
second type of cause, viz. the non-inherent one, is being
explained.

[15]

T
!
!; I

TranslationA non-inherent cause is that which adheres

closely to the inherent cause and at the same time keeps up its
causal ability; e.g. the conjunction of the threads is the noninherent cause of the cloth. This conjunction, a quality, inheres in the threads, the qualified, which are the inherent cause

24

TARKABHASA (

of the cloth. It also fulfils the causal condition of being an

indispensable antecedent. Similarly the colour of the threads
is the. non-inherent cause for the colour of the cloth. ( An
objection is being raised to this ), Cloth is the inherent cause
of its colour, and it would be more appropriate if some attribute of the cloth itself is made the non-inherent cause of the
colour of the cloth as that attribute will be more closely associated with the inherent cause; while the colour of the threads,
if made the non-inherent cause, doeS'not adhere closely to the
inherent cause ( viz. the cloth ). ( Answer ). Not so. The
adherence to the inherent cause need be only indirect such
as adhering in the inherent cause of its inherent cause.
NotesThe phrase 'non-inherent cause5 used to denote STCTT3Tfrwuir should not be taken to mean that the cause in question
does not inhere in any substratum. It only means that it can
never become the inherent cause to the effect in question ( ^
*TO31^*G ^CW^TRT^I?OT ); and it does inhere in the inherent
cause of the efifect directly as in the case of the conjunction of
the threads ( first example ), or indirectly as in the case of the
colour of the threads. The definition of non-inherent cause
given in the text does not make this clear though it has been
explained in the examples. A better and simpler definition is
that given in the Tarka Sangraba, "sffrra SFTT^ %J g%3fowr
mtcT e ^ WTOOTRTfrorcTO^. This means that the noninherent cause should, besides serving as a cause, co-inhere in a
common base (which is invariably the inherent cause) along with
the effect or its inherent cause. Taking the first example, cloth
is the effect, its inherent cause is the thread, the conjunction of
the threads and the cloth both inhere directly in the threads. In
the second example the colour of the cloth is the effect. Its
inherent cause is cloth which is the common substratum in
this case. The colour of the threads (the caused does not
inhere in the cloth and so its relation to the cloth has to be
sought through the threads ( its inherent cause ). This relation
is called ^e^^lf^H^^W, i.e. inherence through another substance in which it inheres. It is necessary that cause and
effect must have a common substratum ( WH^nf^T ) The
relation which joins the effect to this substratum is called
^ and that connecting the cause to it is

[ the relation that delimits (i) the effectoess and

(ii) the causality ]. The Naiyyikas would like to define the
two kinds of non-inherent causality including these relationships as follows :
The relation linking the mmm and the SERUTCU is known as
f^rf^q^ra,
i.e. that by which one co-forms or fixes the
other inclusively. In the first example the effectness in the
cloth ( epram ) is fixed by the causality ( epRor^T ) in the ?g^tT through the common relation w^^. This is then
expressed as follows :{T3rTri^i^T9rf^^Tl3[Tg[
f^^?gcfitTTcm^%'5f^RTcnr^' engaita: i
: q gHTCTrcrTftfrcaTq I Similarly the second example can
be put in the form H^reH3?m3%^qz^q^3

The inherent cause also can be expressed in the same

terminology as
^
; I Here the common substratum ( the base on
which the pot is produced) is the two halves of the pot and
the relation linking the two is inherence, being part and whole
( ar^qgri^irf^^ ). The halves exist in the substratum by themselves; Le. the relation is ^T^T5^T = self-link, ( i.e. they do not
require any other relation to delimit their existence ).
[ Such expanded expressions are quite common in almost
all Nyya works, especially in commentaries, which may appear
cumbresome and bewildering in the beginning. But the student
who puts in some effort at the outset to understand the Naiyayika terminology will find himself amply rewarded by the
precise and compact nature of such thought-provoking expressions. Though it is not intended to fill these notes with such
pedantic expressions, we shall advert to them occasionally to
make the student familiar with the Naiyayika phraseology. ]
Next the third kind of cause ( as per division mentioned in
para 9 ) is being defined and after that the discussion on causality concluded.

26

TARKABHS (

[16] (a) ff?r^Rq- .

m ^ ^T w r l
( b ) cr^TT

q; I "
(c)
Translation(a) That whichis^neither an inherent: nor a noninherent cause, but is still a cause is called a cause-in-general
or an occasioning cause, just as the loom, shuttle etc. ( other
similar accessories ) are for the cloth.
(b) These three kinds of causes apply only to
positive categories. For non-existence, only the occasioning
cause is possible as it does not inhere anywhere; and inherence
is the link between two positive substances only.
(c) Out of the three kinds of causes ( described
above ) that one which is most efficient ( cufosTq* ) in whatever
form in the production of the effect, is called its ^ROT or instrument. Therefore the definition of mwn as the instrument of
valid knowledge is well established.
Notes(a) While there is only one inherent cause and one
non-inherent cause for any effect, there are various other
accessories and factors neccessary for the production of the
final effect, and all these are classed as occasioning causes. All
the three kinds of causes are necessary for an effect; for a
bundle o threads does not become a cloth though the inherent
and non-inherent causes ( ^g and ^sgN ) are theoretically
existent in the bundle.
(b) The positive categories are six in number as mentioned
by Kanada in his V. S. ( 1-1-2 ).
1. "53!l = substance, 2. g<n = quality, 3. ^ 4 = action ( or activity ), 4. HWFsr = generality, 5. firto = particularity, 6. HH^I^ =
inherence.
Among those six, it is only the first three that are producible
). The other three are not producible, but are only

271
abstract concepts postulated to answer certain logical needs
(vide notes under para 11). Out of the three producible
things only a substance ( ^n ) can be the inherent cause of an
effect and any of the other two ( jpT or ^ ) can be the noninherent cause. In the case of a stone falling, the stone is the
inherent cause and for the first moment of its fall, the weight
of the stone ( gs?3", a quality ) becomes the non-inherent cause;
but from the second moment onwards of its fall the acceleration it gains during the just preceding moment (i.e. !*T, a
quality ) constitutes the non-inherent cause till it comes to rest.
Similarly in the case of water flowing, during the first moment
of flow the liquidity (or ^3^0 of the water is the non-inherent
cause and from the second moment onwards, the acceleration
or momentum of flow gained during the immediately preceding
moment becomes the non-inherent cause.
. " %
.
A
Since.the Naiyyikas accept the Vaesika categorisation of
all knowables ( r^nfr ), it will be useful and necessary to give
a very brief description of their categories.
...- .
( i ) "%zn'& ( substances ) are nine in number as follows :
1.
2.
3.
4.
5.

ifni = Earth
6. ?$\ = Time
snq: = Waters
7. f^= Space _,
sHj= Light ( including fire ) 8. STTC"?^= Soul
^Tg = Wind or air
9.
5^iifr = Ether

( ii ) gen's (qualities) are classified as twenty-four in number.

1. ^ q = Colour
2. ?H = Taste
3. TTO = Smell
4. ^q*jT =5 Touch
5. N^T = Number
6. quriTGT =Size or dimension
7. 2*re^=a Separateness
8. RfaT= Conjunction
9. fsmRT = Disjunction
10. q^qf = Remoteness
11. srq^q" = Proximity
12. f% = Intellect (or cognition)

13. ^ = Pleasure
14. : ^ = Pain
15. 5^T = Desire
16. g^r = Aversion or dislike
17. smm~Effort or volition
18. 3 ^ 3 = Weight or heaviness
19. 3 ^ 3 = Fluidity
20. *$ff =Viscidity (or oiliness)
21. H ^ R = Tendency
22. ^ = Merit
23. 8^*5 = Demerit
2i. SISG[ = Sound.

28.

TARKABHAS (

Only the first seventeen Gunas were enumeratad by Kanada

in his Sutra (1-1-6); the remaining seven were added by the commentator Prasastapda. One of the best deOnitions of Guna is

that given in B. P. (v. 85) 'arar ssqfra i h n faapn fafwT

g W - i . e . Gunas reside in substances and are themselves devoid
of attributes and actions,
Themodern school of Naiyyikas does not find sufficient justification to include f%vrm, ^ R * , W * and arq^c^r as gin's. Disjunction is only loss of contact or nt3T?rre. The second one is not
different from ^ and therefore comes under STW^. Remoteness and proximity are only notions in time-relations or spacerelations and are hardly separate from time and space. The
same argument used for excluding fg*TTT might hold good in
the case of g:^ and sew as well. But, on the whole, the Naiyyikas seem to hold that these apparently negative qualities are
not contradictory to the one previous to it; but that they represent positive notions.
It would also be useful to remember the classification of
Gunas into ftrita ( special ) and 3TRTO ( general ). A special
quality is one which resides in one substance only at one time
and not in two or more substances. The rest are naturally
flW or general qualities. The B. P. enumerates them as
follows : ^ : u 90 (b)

awft l i t f t ^ gun: I
^

U 91
." i 92 (a)

The six qualities beginning with intellect, the four ending

with touch, viscidity, natural fluidity, merit and demerit (together
termed as sreg ), impressions or memory ( 3TCRT which is a
variety of fqpR ) and sound are classed as fiHrpfs.
The seven qualities beginning with number and ending with
proximity, artificial fluidity, weight and velocity ( ^sr ) are
SttHlwrgoi's.' It should be noted thatwST?: ( tendency ) includes
the three varities, VHERT (memory), ^rr (velocity) and fer%**7TW
( elasticity ) and no reason is given as to why they are clubbed
together under one word ?SR^T^. For other classifications of
qualities the student may refer to the B, P. and its commentary

the S. M. They are also described in the second part oc this

book also, when defining substances.
( iii ) 3 5 P \ ( motion or activity ) is divided into five kinds:( 1 ) 3Tc%q<n = upward motion; ( 2 ) srq%qt = downward
motion; ( 3 ) 3TTgSR = contraction; ( 4 ) TOW = expansion and
( 5 ) aTW=: going or any movement that does not come under
the above four. Though this classification is evidently unsatisfactory, it is followed by almost all Naiyyikas in deference to
Kanda's original classification in his Sutra ( 1-1-7 ). The words
W^,fiR*n and c% are all used as synonyms in common
parlance; butin Sstraic usage, especially among theNaiyyikas,
there are certain differences. <f is the volitional impulse or
effort that invariably precedes any activity; 3 w H s the general
term for any motion and fhw denotes the particular activity
denoted by the context. This difference between the latter two
is no doubt too subtle and is not adhered to rigourously. But
it is not difficult to discern the difference in meaning in expressions like "sratf f*TT Tcfr fwi" etc.
For the definition of ^ ^ we shall follow Kanada who says

"cEgwregof

sttaf^r^^^H&iffii

gttfe^ai^" ( 1-1-17 ).

This means that action ( or motion ) inheres only in one substance at a time, is different from Guna or quality and is the
immediate ( i.e. non-inherent ) cause of conjunction and disjunction. The last portion ( a^nft^ifc^^r^ETCor ) is the
most important factor and is based on the essential fact that
every activity brings about first a disjunction (fsRHT), then loss
of conjunction with previous spot ( <^5TffcfaraTCT ) and thirdly
conjunction with another spot ( rcrc^TOratar); each of these
three factors occupies one moment ( gar ). Thus with the
moment of its origin an activity ( fsqr ) is said to last four
moments and it comes to an end in the fifth moment. It is
interesting to note that the Naiyyikas use any one of these five
factors as a delimiting condition ( sraifr ) of a ^or which is the
lowest unit of time conceived by them. [ Kesava Miara also
refers to this function of fflB^TT (or activity) in the next paragraph
in the text ].
The other three abstract categories TWT. f$N and sra^re
have been explained earlier when dealing with srgafrac. These

30

TARKABHASA ( <s

1hree together with the ultimate atoms of the first four substances
and the other five substances from ST^ST or ether,, are considered as eternal ( fac^r ) and hence non-producible ( sra^T )
as already explained under para II.
Branra as a separate category was not mentioned by Kanada
along with the other six categories. He however mentions it
in his Stras quite a number of times as though it is a foregone
and undoubted conclusion (e.g. ^R^Tm^T^rahrm : V.S. 1.1.4).
Later commentators ( some say it is the great Udayanchrya
but there is no definite proof for it ) rightly raised it to the
status of a separate q^%! or category and since then the number
of categories is accepted as seven by both Vaisesikas and Naiyyikas. The word has been translated into English as negation as well as non-existence, though both the words do not
represent the full implication of the Sanskrit word arara.
srvnsr is of four kinds; snTTWi = antecedent non-existence;
= consequent non-existence; 3K3T?cTraR = absolute non-existence; and 5T?it?^iTIf or ^=mutual non-existence or difference.
An expression such as "a pot will be produced here shortly'9
predicates the non-existence of the pot before it is produced.
This is therefore called snwra. This is a necessary cause for
any effect. It is beginningless, but gets destroyed as soon as
the effect is produced. When a pot is broken or lost,
then also we speak of its non-existence, which is then called
^TO. It has thus a beginning from the moment the effect is
destroyed, but continues endless afterwards. This is what is
meant by saying that a WiW has no wrara but has ssfe, while
has a STR^ra but no v^r. It will'be noticed that SHTFvriW,
3T and *cfr represent the three stages of ime; past, present
and future. The past has no beginning, the future has no end,
and the present has both beginning and end.
The third variety of smra occurs when we say 'the book
is not on the table' or 'a cow is not white'. In such cass
what is denied is the relation of the book with the table and
that of white colour with the cow. This is called ^c^?cn*tt3f
or absolute non-existence and it is supposed to be eternal.
This word has been explained as 'follows :-cg3T?

31
The fourth variety is called sf?ifc*nvnsr or simply il^ which
means difference or non-identity. In expressions like '^Z'- yi\ ^',
what is denied is the identity ( tfT^T^ ) of the cloth with the
pot and this makes it different from the previous three varieties.
In other words when we say 'sf qt ^* we deny the relation
of the cloth with the pot; but when we say 6vz: qt ^' we deny
the identity of the cloth with pot. The former is SK^cTO^ and
the latter SP^PTT^; and the distinction is brought out clearly
by the use of the locative case in the former, which at once
gives the notion of 3*fwqRtfi( = substratum). This variety is also
considered by the Naiyyikas as eternal just like ar^cttvn^.
The various factors involved in a negative statement are
five in number and should be carefully discriminated. They
are : (1) ^f^Rr^, counter-correlative i.e., the object the exis?
tence of which is denied; (2) ^ f ^ T O T ^ ^ E W , the delimiting
adjunct of the nf^tfti^ ; (3) 3*fitf^, the correlate substratum;
(4) sgiifTnHix^cs^l, its delimiting adjunct; and (5) Sifeitfr^ra^#^chP*> the relation which delimits the countercorrelativeness of the object with the substratum. In a cognition like * 5f ran qV TTfr the existence of a jar on a particular spot is denied, vz is. here sf^fr^ and its delimiting
adjunct is ^zm; for, what is denied is not any particular jar,
but the whole class of jar which is delimited by harness'.
Whenever the existence of an object is considered, one has to
conceive it only through a substratum to which it is bound by
some relation; and this holds good when its non-existence is
also being considered. Here rjra is the substratum, srgsrtfit^,
and its delimiting adjunct is evidently ^jjres^. As the jar is
conceived as existing on the ground only through its contact with it, its non-existence also is delimited by the same relation, contact or Hto, which is therefore the
F*?'<7 in the present case. When we say *<a
this relation is SRP?, and the statement-can be true even when
there is a jar on the ground. But, if by the word srsi the. component part of the jar ( spqre ) is what is meant, then the statementcannot be true, for the jar inheres in the cfiqres and this
cannot be denied. In gthe latter case ( when mm& is. the
^3*Tf^) we can say .'rer TCtito 3t ^T^T to' make it true. The,
Naiyayika-would thus expand the smre -in.1 the ..statement"i *$$

32

TARKABHS (
as follows :

or simply
:-the relation and the two delimiting
adjuncts being left out as self evident.
In the case of gprorc and ScH the <ffg^fn^is invariably the
inherent cause (OT3TfqsTCtflof the sfcritftnO and the delimiting relation is therefore ^rn^ms for it is in the component parts
( such as H^qrr^s etc. ) that the object ( such as sf etc. ) is said to
be produced or destroyed. In the case of a^stemn^, e.g. *sra:
qt q \ it is the 'sameness' of the fqtfT^ ( TO ) that: is denied
in the srgft{T^ ( 32; ) and the delimiting relation is always this
'sameness' or n^srcr as it is called in Nyya phraseology3^r
arTcHT serqr $ 3 sTTttT TOI ? ^; R^T VTT^: CT^TC?}. T h i s vwm

is

therefore called cn^T^TO^^refe^ST^ra or simply ? = difference; and the other three varieties are classed together as
where sigiT ( = relation ) means 'relation other than
It will be seen that, in the ultimate analysis, it is this
'relation' that forms the basis for denial in all negative statements. This will also illustrate the great importance which
Indian logicians ( especially of the Navya Nyya school )
attach to jnr ( relation ) in general.
To avoid endless regression arvn^wre of an object is, in
general, considered equivalent to the original *n* itself; eg. sfTviT^rencT becomes ^z itself. But sreTOT3^3[ may not always be sf.
This discussion is too abstruse for a beginner to follow and
need not be pursued here.
The concept of srqra, highly complex as it is, is of considerable moment to the pluralistic view of the universe o the
Naiyyikas. They define g f e ( or liberation from all worldly
existence ), which is the end and aim of all systems of Hindu
Philosophy, as srr^f?c^:^^r, i.e. annihilation of all unhappiness or evils. On the face of it, it leads to the supposition that
the Naiyyikas are pessimists. This is not so. The word ^ ^
in this context means all voluntary activities leading to birth
and death. Complete cessation of all such activities comprises
in itself all conceivable forms of happiness ( f^*$r ). By
defining gfe through the positive side in some such manner as
%
etc, the emphasis is more on

33

what one expects and not on what one should do to attain that
stage. This latter emphasis is reflected through the negative
form of definition ; for, after all, gfrG is the highest goal which
everyone should try to achieve.
( c ) Ail the above discussion on causes was necessitated
by the definition 'ffflT^cr flfHio}5 in para 5. Though wn has
been explained as the most efficient cause ( vide p.ira 6 ), we
have yet to know how to determine which is the most efficient*'
cause out of the three classes. In trying to explain this the
author uses the phrase qrqrqfqr, which still leaves it indefinite, as
each one can interprtait in his own way. The commentator
Chennu Bhatta offers the following remarks :6f^
:i

l|fcr \il This, in effect, is only a paraphrase of what the author

himself says in a subsequent paragraph ( vide para 18 ).
Naiyyikas of the older school define q^oT as
^iqruo WPi' ; i.e. a specific cause which has an activity
accompanying it in producing the effect. According to this a
potter's stick becomes HPOT as he uses it in producing the rotation of the wheel. By adding the adjective 'sqqriT^', they
appear to make the definition more cumbersome rather than
more specific. For the potter's wheel itself or the inherent
cause itself (OTT^K^T) can become the p^or. The NavyaNyya school define q^<u as '<fi3rataraf^9 ^T*OT\ i.e. 'that
cause the absence of which will not produce the effect even
though all other causes are present.' This will ultimately
reduce itself to saying that the 3T333Tfc$rcoT is itself the c^ur.
For after the two parts of the jar have joined together, the jar
is produced and no other cause operates in the formation of a
jar. To avoid further confusion, it would appear best to
adhere to the definition of PniniflTO$a w i (sreT.1-4-42) which
the author himself has adopted earlier (vide p. 6) and to adopt
that cause as %w\ which satisfies bast in any individual case.

[17] 3TTT srfr^ w<iMr raw ^ \

34

TARKABHS (

i Translation .'-Someone defines 3TOT<n as that ( instrument )

by which a thing not known previously is made known. This
is not correct ; because, based on the same jar ( perceived
already) a succession of cognitions as 'this is a jar' 'this is a jar",
each based on the immediately preceding cognition, arises and
all these will have to be treated as invalid. Nor will it be correct
to say that all these cognitions being momentary, each one has
for its object the jar of that particular moment ( which does
;not exist at the succeeding moment ) and thus the above definition of knowing what is not known earlier, can be satisfied.
Because in sense perception such -difference in the minute
moments of time is not cognized. If this difference in time be
cognizable, then the feeling of simultaneity of the four things
beginning with motion and ending with contact ( which is
experienced by all ) will become inexplicable.
Notes- This definition of snwr is that of the Buddhists who
believe that everything in this world lasts only for a single
moment ( ^Rn^fasTR^ ). In the case of these successive or
serial cognitions, the first cognition does not exist when the
second one arises and so the jar presented as object in each
cognition does not exist at the moment of second cognition,
and thus, they argue, their definition is flawless. The Naiy
yika does not accept this concept of momentariness. He
argues that such minute difference in time is not noticed, but
what is experienced is a simultaneity in such cases. For example
when a lotus (lower is pierced by a needle, we feel ( and
actually say ) that all the petals have been pierced simultaneously. Actually there are four moments of time involved in the
piercing of even one petal( 1 ) the motion of piercing the
first petal with the needle point; ( 2 ) the separation of the point
from that petal; ( 3 ) the destruction of the contact of the needle
point with, that; and ( 4 ) its contact with the second petal.
These four factors are expressed as feu, fqrt f?TRT:, f*TmT^
^TO^Rra?, ^^TO^ifrqfftfThis
sentence is seen
printed along with the text after ?rwra in almost all the edi-

tions; but it appears to have crept into the actual text from the
commentary of Chennu Bhatta. Otherwise the author would
not have used the abbreviated form 'fsRf[*faT*nlV which
the commentators explain as above.
Several moments of time elapse before all the lotus petals are
pierced; but still the experience is that all have been pierced in
a moment. Therefore subtle differences of time in such cases
are not cognizable and this holds good in the case of successive
cognitions referred to here. So the argument that the jar in the
previous cognition occurs as a fresh (or unknown) object in the
succeeding cognition does not hold good, as the idea of moraenfariness itself is contrary to experience.
Now the author concludes his arguments on the definition

[18] 3 swror: ^ i w f a srgfr srfcr,

TI

[ jnTRJT:

arfcRrfff

rar 1 ]

TranslationThere are varions causea for valid cognition

such as the cognizer, the object cognized etc. Can all these be
( called ) 3r*ffT's or not ? This is being answered. Even when
the cognizer and the cognized object are there, valid cognition
does not arise; but when there is contact of the sense-organ
with the object cognition arises without delay. Therefore senseobject-contact alone becomes SRGT, [Because of this peculiarity
it ( the sense-object-contact ) attains a more specific nature over
all other causes such as the cognizer etc; and that ( cause )
which is more specific than others is s?^ai ( as already explained
above ). Therefore sense-object-contact alone is e^ur and not
the cognizer etc. ]
NotesThe author emphasises the specific nature of the
sense-object-contact to show that that alone is HJROT in valid perception. This would imply that he rejects the definition
tf' (explained in the previous paragraph)

36

TARKABHSA ( mh

and adopts the modern view. But that he is not definite abonit
it is brought out immediately below when he says that "afwsjw^f?4" *5Vn' in the case of ftf$4TO13 or indeterminate
knowledge.
The portion enclosed in brackets [ srarar.
'...^S^i]
is omitted from the text in many MSS. Chennu Bhatta does not
comment on this portion as it is only a repetition,
[19] fft SRTTfa1 ^ f t I cT^TT ^ ?qR^f I "STcWTJTTTTt^R^T^T: STTPifV' sfcM ( ^ T ^ - M - 3 )
TranslationThese smm's are four in number as stated ( by
Gautama ) in the Nyya Sutra. ( 1 ) $f<ir$~the instrument of
perception; ( 2 ) srgwr = inference; ( 3 ) 3rqw?f=: comparison;
and $IS3[ = word or verbal testimony.

(3)
[20] f% 5?r: vRqti
i

crTT:

nvT

fft

TranslationWhat then is perception ? That instrument

which gives rise to direct valid cognition is called perception.
That alone is called direct valid perception which arises
through the sense organs. It is of two kinds( 1 ) jfa^qsps
determinate and ( 2 ) fq^c(fi a indeterminate. The instrument of perception is of three kinds. In certain cases it is the
sense organ itself ; in some casesjt is the sense-object-contact;
and in certain others it is knowledge itself.
[21] sfTciT 3pfVf?Si ^ " W T ? IRT f*rfpCFI>^<TT SPTT

5TR
TranslationWhen does the sense-organ become pfoi '?
When an indeterminate knowledge is the result ( of perception )

". . f

37

then the sense-organ is the instrument1. Tha various processes

involved in perception are as follows :first tlje soul prompts
the mind; the mind gets into touch with the ( particular ) sense
organ, which in turn goes forward to contact the object, because the senses have the function of contacting the objects.
Then through the sense-organ which has contacted the object,
indeterminate knowledge arises ( at first ) which visualises the
object alone as 'this is something', without any idea of its name
or generality ( or any quality making the object definite ). The
instrument of such a knowledge ( ffifcj^q^ ) is the sense-organ
just as the axe is that of cutting. The sense-object-contact is
the mediate activity, just as the axe-wood-contact is that of the
instrument of cutting, the axe. Indeterminate knowledge is the
result just as the actual severance is by the axe.

[22] qr^r j ^ f t f e q r ^ f o w : VRTH* ? q-?T

TranslationWhen does sense-object-contact become 3J*T ?

After ( the above-mentioned ) indeterminate knowledge has
arisen, it is followed by a determinate cognition, which visualises the object observed as something with a definite name as
Dittha, as something with a generality or belonging to a particular genus or class as a Brahmin and as something with specific attributes or qualities as black in colour ; i.e. a definite
knowledge wherein the attributes or qualities and the qualified
( and how they are related ) are all clearly brought out as he is
Dittha, a Brahmin with dark complexion. In this case ( i.e. in
determinate cognition ) sense-object-contact becomes SRTC, the
indeterminate cognition is the mediate activity and the determinate knowledge itself is the result.

[23]

TARKABHS

TranslationVf hen does knowledge itself become

After the above said determinate knowledge has been gained,
there arises the knowledge of discarding or accepting or indifference ( of the object perceived ). In that case indeterminate knowledge is $*UT. Determinate knowledge ( which
follows ) is the mediate activity ( **nqi* ) and the knowledge of
discarding or otherwise is the result.

[24]

TranslationSome one says ( as follows ) :Even for deter

minate knowledge and the rest, it is the sense-organ itself that
is the instrumental cause. Whatever contacts etc. ( of the senses
and objects ) occur in between, they all form the mediate
activity.
Notes( on paras 19-24 ). The word sr^T^r, it should be
noted, is used both for perception ( SKSf^sr^T ) and its instrument
( qPW ). [ Cf. wq^ffisROT csrgr l T. S. ] By stating the four
instruments the author has left it to the student to infer the
four kinds of cognitions ( Jim's) of which these are instrumentS0
They are :( 1 ) 3JTO^ = perception, ( 2 ) Sfgffr a cognition by
inference ; ( 3 ) sqfff a cognition by comparison; and ( 4 )
53^IH = cognition by verbal testimony. The divisions and
sub-divisions of ^rffi or <|f^ followed in other similar elementary
Nyya texts like ^ H f l f , vrrawfifc, etc. are more lucid and
this is given below in the form of a table.
After tffJ some printed texts have another sentence

:-The inclusion of this is not warranted by the context. Its

proper place, if at all it ia part of the text, should be under the definition
of ef^OT ( under para 6 or under para IB ).

or

in
the form
or

or

argfaf"

sqff

or =
or
first divided
into
and
and then again
divided into 6
varieties corrseponding to
the 6 f f

|
( No subfirst divided
divisions )
into 3 classes
and
^ ^
and again divided
as 4 and

means without qualification and rR^q with qualifications. In the cognition of any object, say of a jar, three
factors are involved, namely ( 1 ) the jar entity called fer^q1
or substantive, ( 2 ) its determinant jarness called f^fa<n or
, and ( 3 ) the relation linking the previous two, called
or ftf^W. The determinants of these three are respectively
cr ( or fiiqirRn ) and ramu. These play a very
significant) part in the phraseology of ^s^ysn^ ( and also in the
other systems of Indian philosophy which follow the Nyya
method of analysis ). In the cognition 'this is a jar', jar is the
object or f^ and the cognition itself which contains the fw
in termed fqf^. Their determinants fqsreu and fir^cU also
play important roles. The former three factors ( f%^?r etc. )
are included under fqq, as they are all presented in the cognition and the determinant Prqqn will include all the three

44

TARKABHAS (

TranslationThe relation of the senses with objects, which

is the cause of valid direct perceptual cognition is of six kinds
only as follows :( 1 ) contact; ( 2 ) inherence with the contacted object; ( 3 ) inherence with the inherent in the contacted object; ( 4 ) inherence; ( 5 ) inherence with the inherent,
and ( 6 ) the relation of being substantive and adjunct.
These six relations are being explained in order in the next
6 paras.

[26] q-^r ^ p r r srefpsRf ?TR

5TFT
TranslationWhen the cognition of a jar comes through the
help of the eyes, then eye is the sense-organ, jar is the
object, and the relation between these two is mere contact, as
they ( the eye and the jar ) are not inseparables. Similarly
when knowledge of the soul ( i.e. one's own self) arises, in the
form ' I am', through the internal sense, mind, then this latter
( i.e. mind ) is the sense organ, soul is the object, and their
relation is also mere contact.
NotesPerceptual
experience of all substances
comes through this relation ( H ^ T ). Though soul-mind-contact is a general cause ( f^fir^^ROT ) for all knowledge, mind,
which is considered the sixth and internal sense, can also
directly bring about perceptual knowledge like the other
senses ( sight, touch etc. ). The author wants to illustrate this
point by mentioning that the soul is the object of perception of
the mind. Another object of mentioning this specifically is to
emphasise that the Naiyyikas hold quite a different view point
from the Prbhkaras and the Advaitins with regard to the
perception of the s o u l The Prbhkaras hold that soul

45
itself the abode of the knowledge of 'self-effulgence*
and so is self-illuminating Le. does not need
the help of the mind-sense for its perception. The Advaitins,
on the other hand, assert that soul is self-effulgent knowledge
itself ( TOSWTO fr^ ) and as such is always self-illuminating
and non-disguised ( ma* ) i.e. illumines itself without the
intervention of the mind just as a lamp does not require another
one to illuminate it. The unflinching devotion of theNaiyayikas
to the pluralistic view of fkm and Pmk\ will not allow them
to compromise with the ^a^TCf theory.

cR

I
I
WT,

Translation-W'hen the colour etc. of a jar is perceived by

the eye etc, e.g. 'the jar is black in colour', then eye is the
sense-organ; the colour of the jar is the object; and the relation
between the two is 'inherence in the object contacted'; for
colour ( etc. ) inheres in the jar with which the eye is in direct
contact. Similarly the same relation holds good when pleasure,
pain etc. which are inherent in the soul is perceived by the
mind ( internal sense ). When perceiving the dimension etc.
of a jar ( by the eye etc ) then a set of four relations is also
postulated as an additional cause; for even when the relation
'inherence in the object contacted5 exists, the size ( or dimension) etc. of the object at a distance is not grasped in their
absence, This set of four relations is as follows :-the
relation of the constituent parts of the sense-organ ( 1 ) with
the constituent parts, and ( 2 ) with the whole, of the object
and the relation of the whole of the sense-organ, ( 3 ) with the
constituent parts and ( 4 ) with the whole, of the object,

42

TARKABHAS (

also as in &fta4tp&'but that the connections between them

are missed and that is why the observer is not) able to put it in
words. The Viistadvaitins also adhere to the above view of
the Prbhkaras in their explanation of
ff$5<r$.--.
:
The Naiyayikas refute all these views, but end up by establishing it as a necessary antecedant to their theory of complex
knowledge (fftggfffi). The Advaitins, however, carry it
forward to show that real ffiww is the highest form of
realising the truth ( Sfir ), the only absolute and permanent
reality ( *r^n ), untainted by any of the attributes ( ffw ).
This is one of the many instances where the followers of
Sankara's Advaita use the device of the Naiyayikas to explode
the pluralistic view of the latter.
The question also arises whether f^rfippqisg is a valid
perception (TOT) or only an illusion ( $W ). As our author
just mentions these at the end of the book (under <|f^-in part II)
a little explanation about ym would not be out of place here.
The simplest definitions of these two are ^ffjjT cR
vm and a3*n33f 33i$Tt W^ WBy the word ^ the
or attribute ( such as ^zm ) is meant and ajfr means
or sil; and cffiTOrcfcr means tre*3rara<lf i.e. ^f^F^sq^ etc. If
in the cognition of a jar, its own attribute 93*3 is presented as
the 5J35TT then the cognition is valid; otherwise it is invalid or
illusory, as,in the case of mistaking a piece of nacre as silver.
In f^rfi^q^ the inter-relation of the fraqrcn's are only missed
but not mistakenly perceived ( as silver is perceived in nacre )
So it is neither vm\ nor srar, as Viawantha points out explicitly.
"

5T TOT Tf SW ^ f i f i ^ q r ^ : I

TOWaTfroftf % ^^R^Tff cT^ll" ( ^T. 1. 135-136 )

The author began his'discussion on q^or (vide para 6) where
he used the words ^fT^^^JR^, srfiirerPftr, and s^g and now ends
up his discussions by laying stress on these same words in para
18. He has scrupulously avoided the definition "s^TTR^COTWT W qwA" and has made it clear that he does not
accept that view when he says "^f^rftnrcft *T^%tT sn^q1:
^W ?f^r*Nfauf^3 ^JT" ( vide para 18 ). Here in describing
the three star's of perception, he lays particular stress on

43
in all the 3 cases. (In some editions the definition of fNR is also included in the text which is considered
spurious in the text followed here, vide footnote on page 38 ).
When the author says there are only two kinds of perceptions
and immediately gives 3 SRG'S for the same, the question arises
whether there is a third variety and if so why he has not
included it in his statement of division. If, as the author has
explained, ^Ffft^Hif ^ a s ff^^sTH as its SRSF, does it
find a place under perception at all ? According to the accepted f^F^r all perception is produced by sense-object-contact
and its %Wl is either the sense-organ itself when you accept a
TNR in between or the sense-object-cotact when eqiTR is not
accepted, The author subscribes to the latter view as explained above.
In the case of all cognitions other than perception, knowledge
of something ( e.g. o sqrf, m&m etc. ) becomes ^<J and it is
only in perception that t&V*l is other than knowledge, vide the
table of sirffi given on p. 39. Actually knowledge whether the
object observed is to be rejected, accepted or left alone as
neutral comes from the knowledge of utility, which is gained
either by inference ( etgnra ) or from the advice of elders
). So l^q^Rf fe does not come under
in fact ViswanCa in is S. M. ( under e&TfpT
51 ) has defined perception as ^W^oep ^R. Moreover,
prfsN^^lffi cannot be considered even as an ordinary cause
as it is open to the flaw a*^*iftfe ( vide notes under para 11),
for its antecedence to rtq<?lg(^ can be established only
after it is known as antecedent to 3ff3<T9. These considerations lead one to douot the correctness of the text here or that
there is some inconsistency in the treatment of the subject.
Neither Chennu hatta nor Gopntha makes any mention of
it in their commentaries, it has not been possible to procure
any other commentary to verify the position. This conclusion
gains additional srtength by what the author says in para 24
as the view of someone, that the sense-organ is . m^m in all
cases of perception, without either supporting or refuting it,
though it is evident that he does not agree with it.

40

TARKABHS

determinants ( ffNraT etc. ) or any one of them according to

the context, The cognition of a jar will be found expressed as
and such expressions are not easily translatable in another
language.
By the word te^q ( in f^i^TO etc. ) all the 3 frroai's are
meant. The student can easily see that the absense of any one
of them will make the other two non-existent, and also that the
third one T^tT has an edge over the other two as it is the
binding link and so draws to itself the other two. Indeterminate knowledge is often defined asftqqftT3lr5 R or simply
*r8*lfa3W% J8U; and the determinate one as fWffiSfe TO
a cognition in which all the fqsra's are presented. This latter
is always of a complex nature as explained above and is called
fafmgre ( qualified knowledge ).
When the sense-organ comes in contact with an object, say
ajar, determinate knowledge of the jar does not arise the next
moment. But the observer perceives the object as a vague
something, without the fresRU's having any link with one
another. This is what is called frfih^w It is necessary to
have such an intervening knowledge because, Sf^*?5! being a
qualified knowledge, is dependent upon the knowledge of the
qualificationfffgffi^ f^^TWfa^^ . This rule comes
up by actual experience. Suppose A has a stick in his hand,
B, on observing him, only says A has something in his hand,
so long as he does not know what a stick is. So, just as the
knowledge ^i^t g ^ : ( the man has a stick ) is dependent on
the knowledge of the stick, the knowledge of a jar (we TO=TO3*
ffS! ) is dependent on the knowledge of the qualification ^2?^.
This knowledge of qznf is supplied by the judgement of the
observer through his latent impressions ( fcsflR ) and the
ffiwi* helps him as a catalytic agent. So the Naiyyikas
postulate ff^qq^ as a necessary antecedent to the sifsr^qsF;
the only difference between the two is that the former is not
capable of being expressed in words Cf.

P* 92, Benares Ed. )

4
Among the different Hindu Philosophers it is only the
grammarians who do not accept this fcONw. They hold
the view that no knowledge of an object impossible without the
help of the word signifying it. (cf. wff ^sffft $?^^^% ^513^*1
sriflmq> ). All other systems accept in some form or other
the difference between the two kinds of knowledge, though they
may differ in the explanation of the details. The Buddhist
logicians assert thatff$3<T$is the real valid perception while
flf%WT$ is all illusory ($Tfi). They take the word f^qf to mean
doubt or fabrication; e.g. the man in front may be called by one
as sNi and dark coloured, by another as N and brown coloured and so on. It is difficult to get at the truth after sifting
through all these fq^q's. fM3$9<i is free from all these f$??'s
and in that form the object perceived is presented as a*thing-initself which they call 3 3 ^ ^ , absolute and unrelated in its
momentary existence. ( cf. ^W f<W C^fon^ ( cnf-?n-fir )
Kesava Miara also alludes to it at the end of his treatment of
' ( vide para 33 below ).
?, the author of a Bhasya on the Vaiesika Sutras,
( circa 3rd century AD ), describes ffi=WP as an awareness
presented only in imagination ( 3TT&l^Tffn33[ )- The Bhttas,
while following him, use the same expression in explaining it,
and compare it to the vague unverbalised perception of a child.
This vagueness, they assert, is due to the fact that the generality such as 92*3 presented in the ftrfifaq<t$ is not understood
as pervading all the several individuals belonging to that class
( sf ), thus missing the idea of sameness or similarity ( srgf %
f )
when the same observer perceives another jar. , Moreover the
object in front is conceived as a mere awareness of an entity
( qrffK ) without any relation with any kind of attribute, with
the result that s*nf f% or difference from other classes cannot
be had. These two factors sfgffrf and 3*nff% constitute the
special feature of crf$vq$ and these are missing in
While the Bh^as hold along with the Naiyyikas that
OrfWW i s only to be inferred, the Prbhkaras say that it is
presented in itself at the time of sense-object-contact along
with the knower and the thing known, as in the case of other
cognitions, that the three frrofti's are presented in the

46

TkKABHS ( &

Noies-The postulation of the set of four relationsthat

of the whole and parts ( $%nfk\dLnd m*m ) with the whole
and parts of the object observed--is according to the older
school of Naiyyikas; and is not accepted by all as a general
rule. The later school hold that proper distance ( at which
the eyecao see clearly ) is a cause for such perception, or by
saying that undue distance is a hindrance for clear perception
and absence of hindrances is a general cause for all effects in
general^^fara sfr s r r v ^ * T m ^ | g w ^ - This will avoid
the postulation of four complex relations which will ultimately
lead to a doubtful conclusion.
By this relation things inhrent in the object observed, such
as its qualities ( jjur ) and actions ( ^ ) are perceived.

[28]

Translation~-When the generality colourness' (^^cf) which

is inherent in the colour of the jar is perceived by the eye,
then afeo eye is the sense-organ, the generality ^q?^ etc. is the
object and the relation between these two is 'inherence in that
which is inherent in the contacted object*. Because ^:q?qf is
inherent in the colour which is inherent in the jar with which
the eye is in contact.
)
NoiesSimilarly, ^T?W^ is cognised by the sense of smel,
by the sense of taste, sqrf^ by the sense of touch and
etc. by mind ( internal sense ), by this relation.

[29]

HFG[T

^ w q - : ^mt

?T?T

i
Translation-When does inherence become the relation ?
When the sense of hearing receives sound then ear is the
sense, sound is the object, and their relation is inherency

47

Because the auditory sense is ether delimited within the space

of tke ear, ami sound is the ( special ) quality of ether, and
the relation between quality and the qualified is inherence
( vide para 11 )
;

[30]
^

Translation-When soundness ( is^tqr ), the generality inherent in sound, is grasped by the auditory sense, then the
ear is the sense, the generality soundness is the object and
their relation becomes 'inherence in the inherent', for soundness inheres in sound which itself is inherent in ether represented by the auricular orifice ( the ear ),
Notes-The$e two relations ( CRSfjW and pqacrcrTOTir ) are
formulated for the cognition of sopad ( by if^m ) and its
generality ( w ? ^ ) by the second relation. Sound is the
special quality ( %t ) of ether ( 8Ti3SftT ). Ether is present
everywhere but all sounds are not audible to the ear. Only
such sounds.that reach the ear are audible and so the ear ( or
more correctly the ether delimited by the auricular orifice)
is also considered as ether.

[31]

TranslationWhtn does the relation of being substantive

and adjunct become sense-object-contact ? When the eye, on

TARK ABHS ( *4
coming into contact with the ground* perceives the nonexistence of jar etc, in the form 'there is no jar in this place',
thefe'non-existence of jar' is the adjunct ( or qualification) of
the ( substratum ) ground which is in contact with the eye,
the ground is the substantive* and the relation between these
two is (what is called ) faatwfr^qrara. Similarly when
mind, in contact with the soul, feels the absence of pleasure
etc. in the form CI am devoid of pleasure etc.% then the 'nonexistence of pleasure etc' forms the adjunct to the (substantive)
soul which is in contact with the mind. Likewise, when the
absence of the generality m^ is grasped in the letter-sound n,
which inheres in the ear-sense, then this 'absence of qvaf becomes the adjunct to the letter-sound T which inheres in the
ear ( and which is the substantive here ). [ In both thee
cases also fta<nft$trai9 *s ^Q relation ]. This is how, in
brief, 'non-existence' is perceived by the senses through th
sense-object-contact called ft atyft &*?*?!(, which in turn is
based (and determined ) by the five kinds of relations ( $fqt*T
etc. explained previously ).
Similarly, ( the relation ) ^TOW-inherence-is also perceived
by the senses through the relation frctaoiftjNraTcf. Inherence
( of the threads ) in the cloth is grasped as substantive to the
( adjunct ) threads with which the eye is in contact, in the form
'Inherence with cloth exists in these threads',
Notes on paras 25-31..In the above descriptions of the six
affix's ( sense-object'coniacts ), the word gft$c| is used in a
special technical sense. The words 'relation* and 'contact* are
generally used for translating it, but actually it connotes a
special type of relation which represents the whole range of
perceptiv? reach of the sense-organs. It can easily be seen
that the contact ( = sWftT ) of the jar with the ground is not of
the same type as the contact of the eye with the jar. It is this
latter type of contact that is meant by the word g f c r ( We
shall also use this word hereafter with the following abbreviations :i) = *Nft*Tj (2) 4. = agoKrarcra, (3) -ercr.
(4) W = TTOW, (5) OT-H^^&cr'WWT and

49
Before explaining the Nyaya-Vaiesika conception of ^
or sense-organ the reader may refer back to the notes under
para 16 where the subdivisions of fs?T and g< are mentioned.
Out of the nine substances mentioned there the first five are
collectively termed as 'elements' or ^ f ( earth, water, light,
air and ether) and the last five as rBfa (ether, time,
space, soul and mind ), The first four Bhtas are each
classed under 3 heads viz. $rcfc = body, 5f?3*T = sense-organ, and
fai^q a object. The earth-body includes that of human beings
as well as of all the lower animals. Though the animal body
is made up of the five elements ( qrepftftR; ), it is called earthbody ( qffasRfar ) as earth is the main constituent ( i. e.
the inherent cause ) and the other four are only supporting factors, STTOW^'s. Similarly water, light, and air
bodies are also admitted; but these are supposed to exist only
in the worlds of nw, uf[WC and mg respectively. This view
is based on Purnic Cosmology and we may leave it
out of consideration. The sense-organs pertaining to these
four elements are respectively ( 1 ) the sense of smell, ai<n,
(of earth); ( 2 ) that of taste, ( mm of water ); ( 3 ) that of sight
**%:, ( of light ) and ( 4 ) that of touch, m%, ( of air ). The fifth
sensethat of hearingcomes under ether (cf. notes under para
29 ). When we say nose, tongue etc. they should be taken as
meaning sense-organs and not as limbs or parts of the body.
The latter are the seats or abodes of the senses. Anything
other than suft and ^f??^ is called a fkw, but these two also
come under 'firw when they are objects of knowledge. But
sense organs are not objects of perception ( srwr^r ) but only of
inferential or Verbal cognition.
The most common definition of %$%% is
TjtfRlsr HcT g^IWWWtasnT: ; i.e. it is the seat (or*abode)
ofthat contact of mind, which produces knowledge, but which
is not the seat of any manifested special quality except sound.
The reader should recall here the Naiyyika theory of evolution of perception stated by the author in para 21. "NOT *RHf
is this 7RRf^3$^ta that is meant by ' *RTOftiT ia the above
definition. The division of the 24 qualities into special
4

50

TARKABHS (

and general 03TW) has already been mentioned under para 16.
Sound is one of the special qualities and if it is not excluded
from the adjectival portion, the definition would not cover the
sense of hearing. The 'term ^ 3T has to be explained. All the
five senses fall under the elements earth, water etc. as explained
above. So they must have the special qualities of these
elements; i.e. nose should have smell' as it belongs to earth
which has smell as its essential special quality ; the eye has
colour as it belongs to light, etc. But these senses are not perceptible because the smell, colour etc. inhering in them are not
of such a character as to aid perception as the smell %f sandal,
the colour of the jar etc. These qualities are said to exist in
them in a latent unmanifested manner; i.e. seg^r, and hence
imperceptible. Only when they are capable of being perceived
by the particular sense-organ concerned they are said to be
Sfr or manifest. Whenever it is said that smell is perceived
by nose, colour by eye, etc. it means manifested smell, manifested
colour etc. If this?word is not used in the definition, it would
not cover the senses. So to avoid this flaw the word 3T^cC is
inserted as an adjective to goT. In this Way the definition is
made to cover all the senses including mind, which, though a
substance as per the Vaiesika classification, is considered as
an internal sense ( ST?cfff?^r ) in property.
The older exponents of the Nyya-Vaiaesika system hold
that all the external senses reach out to their respective objects
to produce their cognitions3?g sn*r . WT^qpjTwfTn?i; I
Their contention is that sense-object-contact is secured whether
the object reaches the sense or vice-versa. The modern school
holds t at it is only in the case of the eye, the sense ( in the
form of rays ) reaches the object to visualise it and so only the
visual sense can be called T^^TR:^ and all the other four
should be considered W F ^ f t ^ . Yet others include the
sense of touch a^ so under
It will be noticed that the first three ^rtrore's are intended
to cover respectively the perception of ( 1 ) the substances;
(2) the qualities and actions inherent in those substances
perceived under ( 1 ); and ( 3 ) the attributes' that are
inherent in the inherent qualities etc. perceived under ( 2 )*

51
The fourth Sffir^ is inteaded to cover the grasp of sound by
the ear and the fifth that of the inherent attributes in sound.
We have already referred to the fact that the qualities to be
perceived should be in manifest form ( g:3cr ). In the case of
perception of substances ( which happens only through the
senses of sight and touch ), these should also have ( 3?gcrq
( manifested colour ) and ^ ( largeness or proper size ) as
co-existing determinants in the sense-object-contact causality
( ^[Wcn^ij^E ). It is not enough if these are connected with
the substances perceived in some indirect form, as otherwise
EfMfcr in aja.r'and ^t^m in an earth atom would become
perceptible as these are connected in some distant form with
the 3??^ and 3f;q of the jar. This means that in the psrception of substances these two factors should be caused by
the relation HiT3TT3T, in the perception of qualities and action, by
the relation inherence in its substratum ( wsr^efl^T** ) a n d
in the case of perception by the third ^ % ^ , by the relation of
inherence in that which inheres in its substratum ( ^iSRCW^WWr). The older school hold that 3Rr?r is a cause for
all perception by external senses while the modern school
assert that it need be a cause for visual perception only.
Similarly an$tarata ( contact with light ) should also form
a determinant in the case of visual perception as objects in
darkness are not visible.
The sixth aft^qf ( ft. f. w. )also calledfirWcTTin shortfollows from the Naiyyika dictum :-"ffti *n rRB^ila1
nimm snfcrec^vTwsr ^f^ita ^"; i.e. when an individual
entity ( substance, quality or action ) is perceived by a particular sense-organ, the generic attribute of that entity and its
non-existence are also perceived by the same sense-organ.
While generic attributes such as STOTI and ra*3, ft&H! and
*r$W, s r s ^ and 373*3 etc. are perceived through a f i r e s 2, 3
and 5, non-existence ( as also inherence ) is perceived through
the last rfr^. This relation is a complex one as it is based
on one or other of the previous five. When one perceives the
non-existence of jar at a particular place, the eye is actually
in contact with its substratum, the floor. This relation is therefore H^^W^T-adjunctne^s tQ that which is contacted by

52

TARKABHS (

the sense-organ. The non-existence of colour in numbers (and

so forth) is perceived by f^E HfliffsrtaGT3Tadjunctness to that
inhering in the thing contacted by the sense, the non-existence
of colour etc. in OT1W etc, by the relation ^TfrOTifcweto
fiiifa*nm; because fsTi^ in the adjunct inhering in fen which
5 inherent in the thing perceived and therefore the relation inherence occurs twice; and similarly, other cases of perception
of non-existence in other complex cases also are explained by
the common and uniform relation of adjunctness (frtwcn) qualified in such a manner as to connect it to the substratum which
is contacted by the sense-organ. All these extensions apply
to the perception of inherence ( ragrTO ) as well.
In the case of perception of non-existence,
( = effectual non-apprehension ) is an important causal factor.
When a jar is present in front, the eye does not perceive the
non-existence of the jar, because the apprehension of the jar
is there. So the non-apprehension ( agq^fsqr ) of the countercorrelative ( lfaftPi^) is a necessary cause for perceiving its
non-existence. This non-apprehension should have also
effectualness ( or capacity ). This effectualness ( sftisffia ) is
determined by some such ratiocination as "If there be a jar
here, it ought to have been perceived by the eye, just as the
floor.'9 When contact with light, largeness and all other
necessary conditions exist, we can always assume that if there
had been a jar it 0-ughtJ.to be visible. Here it is the nonexistence of the jar that is perceived. But in darkness the
above assumption ( or ratiocination ) cannot be made. ( In
other words sftwi is not there.) Hence in, darkness there
can be no visual perception of the non-existence of the jar.
But we can have tactual perception ( of the non-existence );
for in darkness we can have the above q>qm as contact with
light is not a necessary condition for tactual perception.
Qualities like g^wr ( weight ) are not srwrg ( perceptible )
because they are not sftraj and so their non-existence is also
not perceptible. The non-existence of manifest colour
( 3 ^ ^ ) in air, of smell in stone, of bitter taste in gur, of
coldness in fire, of sound in ear and of pleasure in the soul
( and. similar nira's ) are all perceived through the respective

53
senses, as sfrrcccn of the counter-correlative in such cases can
be had through the above mentioned form of ratiocination.
In general the perception of eeT^TR ( the first three kinds of
non-existence ) depends on the effectiveness of the countercorrelative ( 5frRtn3fa*T3T ) and that of rwftn*W ( or difference ) on the effectiveness of the substratum ( &rarc'itarm ).
The perception of difference from ghosts in a pillar (^$f fwi^~
$<?! ) is an example in the latter case.
The above six efil^'s relate only to what is called normal
perception ( ^fsp^qpf ). The Naiyyikas recognise three
other types ~of super-normal perceptions ( ^r^f^^Rc^r^ ) which
have three kinds of super-normal relation or srsftfire ^r$<i
They are also called by the word sr^THff which means the
same thing as tffspq. They are ( 1 )
( 2 ) gm^oiTSKqrofo, and ( 3 )
fe
( .1 ) wrpTOVTOOTrf--^This relation is based on the
knowledge of cri?TT??T which is used here in the sense of
common attributeHflWTt tirf:. It may mean any common
attribute such as ^JR^ which is an eternal generality or it may
be any other thing such as a jar that is perceived as a
common attribute in all places observed. When one sees
smoke in the hearth he gets an idea of ^m also and through
that an idea of all smoke as possessed of. vgRm. Similarly
when one has observed ajar at a particular place or as inhering in its component parts ( ^ r ^ ) it enables him to visualise
all such places or 3Ri35's where the jar may be related by
contact or inherence. In all such cases the relation with
which the common attribute is observed in the first case should
be the same in the substratum in the second case also. Thus
at a place from where the jar lias been removed, one would
not have the knowledge of all places when the jar was related
as the wn^TOH is not there at that moment. It is therefore
necessary for super-normal perception to have all the conditions
necessary for normal perception by the same organ.
( 2 ) gffi^fflTSMHfr-This is a relation based, on knowledge of something which is always connected with the object
of cognition; e.g; as soon as one sees a rose flower he gets the
knowledge 'here is a fragrant rose'. Here rose is the object

54

TRKABHS ( a ^

of visual contact; but fragrance is not a quality capable of

being grasped by the eye. All the same fragrance is an adjunct
to rose in the knowledge. To account for such perceptions
the Naiyyikas assert that the knowledge of fragrance of rose
which the observer had experienced previously and which lies
latent in his mind (as a ft$l7 ) steps in as the relation to
connect it immediately with the obseived rose. Similarly when
one mistakes nacre as a piece of silver the knowledge of silverness ( <T5r?*3 ) gained elsewhere serves .as the-: relation connecting it with the nacre actually contacted by the eye. At the
moment of observation one does not know that the object
contacted by the sense-organ is not silver, and that is why
this perception is not valid, and is proved to be so after he
has taken the object,in hand and examined it. If this relation
is also based on a form of knowledge, in what way does it
differ from the previous one which ( as explained above ) is
based an "the'knowledge of ^RT?sr ?,. Through ^m?wfi5qT<8% one gets the knowledge of the substratum with which the
common attribute is connected. Through gH^ttiWUW one
comes to know the connection of that thing with the thing
observed. If this second relation is not admitted the knowledge of fragrancehood ( g*fn^ ) as an adjunct to rose and
of silverness ( ^S^CST ) as an adjunct to the object observed cannot be explained pibperly.
( 3 ) ^m^w^^tni^qi^f%By practising yoga exercises
one gains certain occult virtues or powers by which one is
enabled to have knowledge of things past, present and future
with or without any external sense-object-contact. This knowledge is considered as a form of internal perception ( ?RH~
sw^r) and to account for it the relation based on some powers
gained by yoga exercises is admitted by the. Naiyyikas. Since
yogins are of two types, g^s or those who have attained such
powers, and jp^H or those who are on the road for such
attamitientj this relation is ; also of two kinds. Yogins of the
first type can have knowledge of all things at all times, while
those of the'second type need the help of meditations before he
gets such knowledge.
' '>'
The first two kinds of super-normal perceptions, it will be
observed, are on the borderland of perception' as they really

55

form a combination of two cognitions, perception and recollection,

though without the knowledge of such a combination. The
Nyya school, relying on the after-thought ( sgsq^i^ ) that
occurs in the form of CI have had experience of a fragrant
flower9, explain such cognitions as perceptions and bring in
the recollection part within the causal frame work through
these super-normal relations. The Mmmsakas and Advaitins
do not admit of such super-normal perceptions in their schemes
and explain such cognitions in different ways.
The Nyya insistence on ft. ft. ^T. as a special nfwwi for
the perception of non-existence is also intended to refute the
views of certain opponents, especially the Bhtta Mmmsakas
and Advaitins. These two schools postulate another %mw
for the cognition of srvtt^1 alone and call it argqgsfssr (nonapprehension ). The Nyya realists hold that non-apprehension is a cause for mrrcqcJTO and argue, that it need not be
raised to the status of a ^or thus formulating a new SRI^T. The
opponents reply that it is easier to explain the cognition of srsu^
through a new srwn instead of clubbing it under ^ ^ ^ and
assuming various complex ^rf^^'s for its perception. The
Maiyyika, comes'out with the reply that ft.ft;.*n.-.-is.not;a new
relation but only a self linking relation-?;^*
aBdftit^cTare only of the form offirsftqaT, adjunct, and
substantive, which; are -frothing new The arguments on both
sides? are well balanced and no finality could be reached.
t^TJt-^f^1 ( as explainedearlier ) is a necessary condition
for smTcfSR^ for all schools who hold svf- as a separate
category. The Prbhkara Mimmsakas do not hold TVTTW as
a separate category, stating that it is only a particular form of
its substratum ( vide also paras 66-71 ).

[32] r

Translationr~r:ln this iWay the sense-object-contact has been

described in six ways. All these are summarised as foliows.

56

TARKABHS (

The cognition arising out of the senses is of two kindsdeterminate and indeterminate. Its instrument is of three kinds and
the sense-object-contact is of six types. The various things
cognised through these six contacts are respectively. ( 1 ) The
jar, ( 2 ) the blue colour of the jar, ( 3 ) the blueness ( ftRr)
in (2)? ( 4 ) sound, ( 5 ) the generality of sound ( IS^HT') and
( 6 ) non-existence and inherence.

[33J 3

Translation[ The Buddhist idealist objects ]. Indeterminate

cognition which has for its object the real ^ss^T-^the thing-initself untainted by any of the f^W^n's^^may be considered as
perception. But how can determinate cognition be perception
as it has generality as its object and conveys the comprehensive generic idea just like words and inferential probans
( f^ ) ? Only that cognition directly arising from a real
object can be perception and only an object which really exists
( qWPKU can,.produce such perception. Only ^^^or^^as
admitted by us ) is the real thing that exists ( in a positive
form ); and not generality which is insigtficant ( being an
3f*n? ) in as much as its positive character has been refuted by
all proofs.
ABS,T1IS is not so; for generality also is a real existing
entity ( and not insignificant as you have surmised ).

NotesA reference has been made earlier to the Buddhist

view of ffsi^q and the notes under para 24 may be recalled
here. The Buddhists do not admit.fimir^ or snf ( generality )
as a separate category. They explain the functions of
through a newly formulated w*f ( characteristic ) 'called
This is explained as SRTcj^sntfi or fWffirRrar-differbnce from

57
what it is not. Their arguments for not recognising it as a
positive entity proceed on the following lines. "Before the jar
was produced, jarhood ( wzm ) was not present in any of its
causes. After the jar is produced where does it come from ?
It cannot all of a sudden rise out of nothing. When the jar
is destroyed ^m is not destroyed as it is eternal. So where
does it go as it is not visible ? Because such things cannot be
explained properly r w r cannot be admitted as a real positive
entity. The similarity of cognition in a number of jars as
'this is a jar, this is a jar' ( fgf f^srcsre ) is explained by you
by postulating "sja^ as a common generic feature. We say that
a jar as soon as it is produced acquires the characteristic of
differentiating itself from all non-jar things. Because this
characteristic persists in all jars we get the similarity of cognition; and so on."
The Hindu Naiyyikas ( as explained under para 24 ) consider 9fR><T$ as a complex cognition embodying all the
fsRicT's. The Buddhists think that sense organs can contact
only positive real entities and not unreal things like HTTTT^
( which according to them reduces to a &wm ). They therefore
do not treat ^ f ^ ^ q ^ as a sense perception. The author here
dismisses the objection by the single sentence that aWT^T is a
real ppsitive entitymeaning thereby that it is not a 3T wr, that
the sense-organ can have contact with it through the substratum and so determinate cognition comes under sense perception. This is only a statement of one's own view and not a
refutation of the opponent's. It can be seen from the brief
reference to the Buddhist theory ( above ) that its refutation as
such means striking at the roots of fundamental concepts on
both sidesan attempt neither possible nor necessary in a
primer like the present one.
Thus perception has been explained.

58

TARKABHS (
(4)
INFERENCE

[34]

5
[35] f%

%* qrqfr i
. . Translatian 34Subsumptive reflection ( mm4 ) of the
reason ( few ) is ( called ) inference. The derivative meaning
of ^rgUR is that through which something is inferred. It is
( only ) through subsumptive reflection that something is
inferred. Therefore such a reflection becomes inference; and
that is the knowledge of ( the presence of) smoke etc. in the
subject; because that is the instrument for inferential cognition.
( ^l^f^ ) Such a cognition is the knowledge of ( the presence of) fire etc. in the subject and its instrument is the
knowledge of the presence of smoke etc.
35What is fi3 and its TOn are being explained..
Reason- ( fgs^F ) is that which causes a thing to be known by
the strength of invariable concomitance ( sqrr ); e.g, smoke is
the reason for ( inferring ) fire. It is in this manner. The
invariable co-existence in the form cwhere there is smoke there
is fire' is ( what is called ) ^T%. When that is grasped smoke
makes known (the presence of) fire. Because it causes the
inference of fire through the strength of %m8, smoke becomes
fe^ for ( inferring ) fire. The third knowledge of this f&f is
what is called ffq*T*fa. For instance, in the kitchen etc.
one sees repeatedly smoke in association with fire and through

59
these repeated perceptions he makes out a natural relation
( between the two ) in the form 'where there is smoke there
exists fire'.
Notes The word srgR means srg = after and H~ proof
or knowledge; i.e. it is a mental judgement which arises after
some thing has been heard or observed through certain steps
of reasoning based on the observed ( or heard ) things. The
author in his anxiety to establish the scarcer or instrumentality
gives the grammatical derivation of the word ( srgftq sr^f )
at the very outset. Vtsyyana's remark ( while commenting on N. S. 1. 1. 3 ) is more to the point "faete fitTOW vm**TR" the subsequent judgement or apprehension of a thing
from the reason previously known. That is also one of the
main reasons why Nyya writers prefer to treat 3rg*nr just
after 5R*re as the former is greatly dependent ( 3q*rtaro ) on the
latter which thus becomes sq^ftsq; and the propriety of treating
inference after perception is explained by the relation
Before going further, it would be desirable to have a
general notion of the technical terms involved herein. The
author no doubt gives their technical definitions as he proceeds
further and we shall also refer to them in detail at the appropriate places, argff ( Wr ) is the inferential cognition of
the form The mountain has fire', which arises after going
through certain reasoning processes. The mountain in which
fire is inferred is called ^^ or subject. Fire which is the thing
inferred, is called tt*% or probandum. Smoke which was
observed as coming from the mountain, causes the inference to
be drawn through a knowledge of its invariable concomitance
with fire and so becomes fff or | g or HP^ or probans
( reason ). The invariable concomitance between the gitsq- and
is called sqiftf, and with reference to ssu%, ^ q - becomes
(pervasive or of greater extent) and m%m becomes
( pervaded or of less extentf). Expressions like '3%oqn*qt
*|i;' and ^Ts^iq^ gf|: are quite common. By keeping to the
general form of ssnflc, ' ^ vcqRft^ifo:' it should not be difficult
to keep the fT^qr and sqiq^ distinctively. These three terms
q^, W&K and | g correspond to the minor, major and middle

60

TARKABHASA (

terms of the Aristotlean system. While the latter lays greater

emphasis on the middle term, the Nyya ?cgftt depends on the
relation between g and m^ technically called f
},. The words fs, g g and ^'T*i are all used as synonyms.
But the first one ff which actually means 'mark' or 'siga' is
used in a technical sense in sugiTT section of Nyya works.
Smoke by itself as seen trailing in the sky without knowing
the source of origin or its substratum, will not be a cause,for
inferring fire. Even if the source is known, if its co-existence
with fire is not ..recollected, then also there is no inference.
So the g, smoke, must have two distinct qualifications, ..to
enable it. ..to cause the inference of fire (1) it must be qualified by nfe ( i. e. stnfffirg ) and ( 2 ) it must be observed to
exist in a subject ( q^r ) i.e. it must be an adjunct of the q^
( W<W )'in other words it must have q^^Tlcn-the quality
of being a qiWH- When the | g is known to be snfiiffS then
it is technically called ff and this is what the author means
by saying $mfe^dpi*T3TO ff. Some commentators explain the
word as ^ftww1 sTfTiatfr f^#fef is that which makes us
ascertain the latent thing which may be the s?n% itself or through
it ( i.e. by the force of n%Bsrrfsrf ) the fixe or crTtrar.
Now the knowledge of the two qualifications of.tg may
remain as separate notions without any relation between
them e.g. \fjft tqnfrffre :' and %HI qcTffh Even, then it does
not cause inferenceunless these two merge together to form' a
third complex knowledge as ^ifBf^f^IS^: qiaffxft3 inference
does not take place. This is expressed in the text as ' ^ tjeft^r^rf'fTOren'. This complex knowledge is called variously
as" q^m^T, ff^W^T or ^cfmfefq^ii^. This is the instrument'
( ^im ) of inference according to .the author as he does not
admit of a mediate activity ( swiH ) in between the effect and
the ultimate cause. This does not appear to be consistent
with what the author has said under perception ( vide notes
on'p. 43 ). Aooambhatta in his ^r. ^ t says the same thing
( cfWlfefqwsifsg*n?m); but in his own commentary {$. cf. ^.)
he seems to have recollected the inconsistency* and adds as
* An Bb. says under perception

ji

61

an after-thought "sgnTWsPWri 3*aiffr *r TOTOgR n%5fr ^tinj". Viswantha in his 3>f?$T33ft has explicitly stated his
position as ics^rore*g wtrcr, ^ w *9Tf>#^^i argwii' ## '' of
( V. 66 ). ( In inferential cognition, q^OTST is the sqrq1!?, knowledge of s^ife is the HETfJ and the cognition itself is the *$&result ). His definition of q^T^r is also more straight-forward
'sin^f q^fi^^t: qR*W ^ q ' . The knowledge that the S
or concomitant exists in the subject ( y^ ) constitutes
, which we have translated, above as 'invariable concomitance', is at once the most important and most disconcerting term in the domain of aigFHR. Kesava Miara begins with
an apparently simple definition '{nfr3*rfo*r surfe5'
i
means co-existence which is the same thing as
This co-existence should be invariable ( faira ) and not adventitious; e.g. wherever the ^g is, the ctfT should exist in the
same place. This is fulfilled in the stock example of smoke
and fire, for at all places where we have observed smoke such
as kitchen, hearth, bonfire etc., we have seen fire also without
exception, i.e. the co-existence or concomitance is universal
or fqfr. We also know that smoke is an effect of fire. So
where the effect (smoke) is seen, the cause ( fire ) must be
invariably present. But the inverse is not invariably true,
for 'where there is fire there is smoke9 may be true in certain
cases ( as in the kitchen, hearth etc. ); but it fails in a redhot
iron piece. This shows that fire, besides existing in .all places
where there is smoke, exists also in places where there is no
smoke; i.e. it is of greater extent than smoke. sq% therefore
carries with it, besides co-existence,, the idea of greater extent
and less extent; i.e. ^ is s^nqsj? and H^R is
The word extent ( or extensive ) is ambiguous and is likely
to be mis-understood when bulk or numbers are involved. An
area of 10 acres is certainly of greater extent than one of 5
acres, but still a 10 acre-field is not smq^ to the Naiyyika.
It is just vice versa. For here the sajfe is 'wherever there is 10
acre field there is 5 acre field as well' and not the otherway
round. Similarly a greater sum ( say 1000 Rupees ) is not sqiqc^
to a smaller sum, ( say 200 Rupees ), for we do not have the
'wherever the sum 200 is, there exists the sum 1000%

62

TARKABHAS ( $<

This is why we have mentioned on prepage that the distinction

between zmy^ and swq c a n be had from the form of
The definition of z^fy as n;<er*fft*ra or f^r^ar^rf ^4 is
simple and intelligent enough for all practical purposes. It
follows the fa^cfssnff of ^ i W m f a . When slightly expanded
it takes the form c%gHm^^^i^?ai^^af^iti^i^frRTRqp*<ri" ( vide tr. T. $. ) which is explained by its commentator ft&$r as follows i^effe
|
i99. The expression
is only a paraphrase of the word fmx ( = invariability ).
Kesava Miara after explaining s*n% begins to explain how
qpflSHft becomes the third step in the process of inference and
incidentally mentions another definition of sqi% as ^mfp^ra*W?, a natural relation. 33!fTf$ means not brought about
by adventitious circumstances--srftqrrfre?and this requires
an explanation as to what s;qTf is. This leads him to a
digression in the next para. We shall also follow the digression before explaining something more about fc

[36]
I

: I %

63
Translation( f as mentioned before repeated observation
leads one to know the relation nf% ), the relation 'Whoever
is Maitri's child, must be dark' will also be valid and similar
to that between smoke and fire, as observing over and over
again is common in both cases, ( But it is not ); for the relation
between Maitri's child and dark complexion is not a natural
one, but one brought about by an adventitious circumstance,
such a circumstance being the effect of Maitri's feeding upon
certain kinds of vegetables etc. ( In other words) the darkcomplexion is brought about not by being Maitri's children,
but is due to some effect of Maitri feeding on vegetables etc,
This causal circumstance ( which is purely adventitious or
accidental ) is what is called sqifr. There is no such ^qifr in
the relation between smoke and fire. If there is one, is it
effective or ineffective ( in vitiating the relationship ) ? If it
is ineffective it need not be suspected to exist; and an effective
one, if present, should be noticeable..' Where there is an sq%
it is ( always ) noticed; for instance ( 1 ) in the relation of fire
with smoke ( i.e. in the eqifcl, *5ifTOrei ^ : ) contact with wet
fuel; ( 2 ) in the relation of slaughter with causing sin, prohibition; and ( 3 ) in the relation of being Maitri's child with dark
complexion, the effect of vegetable feeding, are ( observed as )
^qf's. Whereas in the co-existence of smoke with fire ( q^i
tjjflKf nf*r : ) no ^rqfw in noticed. Had there been one it would
have been observed. It is not observed; so there is none/ By
actual sense-perception aided by the above ratiocination and-by
the fact of its non-apprehension, the non-existence of ^qif^r is
determined. Here snfc between smoke and fire is determined
by this sense perception alone which makes us know the coexistence between the two and which is aided by the impression
of the knowledge that there is no 3qifr. Therefore the relation
between smoke and fire is a natural one and not produced by
any ^qTf^f. Such a natural relation ( untainted by ^qifw ) is
what is called egnf^r ( invariable concomitance ).
NotesIn the previous paragraph the author has mentioned
that by constant or successive observation ( sjsft ^fsr ) in
kitchen, in the hearth, in bonfires etc. one grasps the. coexistence of fire and smoke. But mere co-existence is not
sufficient for ascertaining the complex relation, S^T%, as it

64

TARKABHS (

involves the element of invariability as well. This invariability

is grasped by a sort of mental ratiocination ( 3P ) after coexistence has been ascertained. This amounts to the search
for the existence of influences brought about by adventitious
circumstances, vitiating the apprehension of co-existence. By
observing fire and smoke existing together at several places, if
someone were to grasp the strife 'where there is fire there is
smoke' it does not become invariable. On analysis of the fact that
fire is one of the causes of smoke, and by observing only the
effect one is naturally led to infer the existence of the cause; but
by observing the cause the inference of the effect is not always
warranted, as there may be other causes absent which have
prevented the effect coming into existence, In the case of a
red hot piece of metal there is actually fire but no smoke due
to absence of wet fuel which is also an invariable cause for
smoke. So wet fuel is an adventitious circumstance which
prevents the invariability of the relation 'where there is fire
there is smoke'. This is technically called ^qfr. In the case
of the nifca 'where there is smoke there is fire' no such
is observed and therefore the s^nf is invariable.
This is not the appropriate place for defining and explaining sqifr and so the author just gives a few examples to show
how sqfr prevents the invariability of smf- ( The author
gives the definition of sqifr under 'fallacies^ and we shall
advert to it under that paragraph vide p. )
The presence of 3Rif in any ssnf leads to the detection of
the fallacy srfRR ( t abiding in a place where the wy does
not exist ), which in its turn prevents the ascertainment of
So 5*n% is al s o defined as a relation not caused by
( SRtqTW 3 W * : ). This is based on the idea of
= non existence of the f g without or in the absence
of the mw. When expanded this definition takes the form
CTtnrc^'tnifcKr- This is called in Nyya works as i^q^rnfr
and lays more emphasis on the invariability part than on coexistence as is evident from the double negative involved in it
( sn^T and if ft ). Such a 3qy% will be found not to cover
cases where the non-existence of jT^or irg cannot be conceived
in any form as in the case of grftq*?, siifaw etc. More-

65

over, it will not apply to an inference like 'jjo

WQt? ( = a quality has existence because it has a generality).
Here &m is ^r^rr and it abides in spar, jjo? and cfff. So CTttr^qrs
are HWS and the other categories and only where the fg
J.S known to abide in any of these through the relation delimiting the fg*n here ( i.e. smarm ) its negation ( or a*fo?cf ) can
be postulated; for in any negation the knowledge of the
counter-correlate ( srfaratfir^) isa necessary cause. To avoid
these difficulties the definition of nfg based on 9T^rqf?ra( as explained under para 35 ) was put forward by Gagea
Updhyya. This latter is termed as isrvT?c3*nf;
The term 3j?t^h is rather ambiguous. It may mean
( 1 ) repetition of the same observation over and over again
( sjWT 35HTsri OTTfRi ) or ( 2 ) observation of several instances
o the TSS and | g ( m*m wmm?mm\ ^^ ) or ( 3 ) observation at several places of ww and f existing together ( r^rc^c*
filW^3 flWrrenftisN' )- It is evident that it is in the last
sense that the term is used here. Even then such observation
by itself does not enable one to conclusively determine the
unerring nature of nfa. For, even after observing in 1000
places the co-existence of smoke and fire if in a single instance
smoke is suspected to exist without fire, nfg cannot be made
out. To avoid this it is mentioned that co-existence should be
qualified by the.knowledge of the absence of Q^fSrarcv brought
about by ascertaining the non-existence of sqffnr ( sq^sroraf ta ) and this latter partnon-existence of
is made out by an inductive sort of ratiocination ( ^ )
which'takes, the form 4if there be no fire then there can be no
smoke a "well/ This is emphasised by the author in the
expression '^OTn^rartflr ^ H f ^tftw'.
Even then how does the observer arrive at the universality
of 5*n%?' Firstly, it is impossible for any one to observe all
instances where smoke and fire co-exist In many cases ufr
is exemplified as "*ra *ra 1311*39 3T3fT:" and the repetition
STSCSI is intended to emphasise the universality of the relation.
The notion of 'everywhere' is quite distinct from that of the
totality of particular perceptions however valid they may be.
The process of induction as understood in the English system
5

TARKABHS (

of logic ' is not accepted by the Naiyyikas. J. S. Mill

in his system f logic ( Peoples Edn. p. 188 ) defines induction
as "that operation of the mind by which we infer that
what we know to be true in a particular case or cases will be
true in all cases which resemble the former in certain assignable repects." The Nyya system explains 'that operation of
themind' by their 3j^f5f SRznj through cnsrT?^WT3T?UH%
by which one is1 enabled to conceive all tcr and all ^ff through
the delimiting attributes ^ J T ^ and ^fls which one cognises
thiougli the observed particulars ( see notes on p. 53 under
para 31). This provides the notion of 'everywhere' and
cnabS\on.e to grasp ^ T % in its universal form as 'wherever there
is. smote there; is fire.' In the English system induction is a
process of inference from known particulars t the unknown
general; while ^TW^^T!TS3?TO% is the process of cognising
tfye class notion from the perception of the individual through
the flaw ; of association. Because this cognition follows the
visual perception of the individual and because there is no
% Q RWtT in between, it is included under T^^. AS
there is no direct cfoproferftre*!, ' it is classed as r^f^
supernotmal-as already explained. To emphasise this our
aujbor has rightly pointed out 5r?*T%ora at both the places in
the last portion of the above parafor cognising the absence
as well as for ascertaining

[37]

I
l

--' - Trnslaion^The cognition of smoke in the kitchen,

while ascertaining the invariable concomitance of smoke with
fire in the,manner described,above,,is the firs,t Step. The pognition of smoke in the subject, mountain etc, is the second step.

67

After that (seeing smoke in the mountain) one recollects

the invariable concomitance and then subsumes as follows :
"This mountain has smoke which is invariably concomitant
with fire", and this cognitiom is the third step. This third
step should neccessarily be accepted. Otherwise our knowledge
will end at the cognition 'where there is smoke there is fire'.
How can this aons prove the presence of fire ( in the
mountain ) ? Therefore here also it is necessary to accept a
cognition of Smoke (which is invariably concomitant; with
fire ). This cognition is what is called fef q*m. T ^ i s being
the direct instrument of sfgff is gipfln* or inference. From this,
inferential cognition of the form "This mountain has fire"
;
arises.
.
NotesAll the three cognitions mentioned by the author
centre round the | g , smoke in the present case. The first isactually Tififra itself by observing. smoke and fire together
in kitchen etc. or in other words that smoke is an invariable
concomitant of fire ( wj sif|fnw ). The second is the sight
of smoke in the mountain. (wjq: q^cf # ). This second step
immediately makes one recoilect the s^^ 1 by the law of association. By this recollection, the two previous cognitions
( ^?t ^ I ^ T ^ and ipp q w f <$ ) merge together to form a
complex cognition as 3%ttnc*T{ff: TCftfr.or ^ s q r i t ^ ^ ^
t r ^ : . This is what is called *?<rm^T. Immediately after this
one gets the s?gf% 'The mountain has fire' and so it becomes
the ^T of the latter, which is the same as ^J^T^T. If the
third step^ tRwtf is not accepted, we would end with the
recollection of the ssnfr and that by itself will hot be sufficient)
to produce the cognition that the miq*6 or ??eqr exists in the
subject, as is accepted by the Mmmsakas and Vedntins. To
emphasise this point the author argues further in the next two
paragraphs, that e*iif$rqFr or qspmm^R by itself is not sufficient to bring about inferential cognition.

( a ) A has

68

TARKABHSA (

Translation Q. Why does not the first cognition of smoke

in the kitchen become the cause for inferring fire ? True.
But at that time the invariable concomitance between smoke
and fire has not been determined; and only after that has been
ascertained inferential cognition can arise. In that case,
when once the invariable concomitance has been ascertained
one can infer fire in the kitchen itself. Not so; because there
is no doubt, as fire is actually seen there; and only that thiog
the existence of which is doubted becomes an object of
inference. This has been declared by the author of the wssy
( on wn-^-1-1-1 ):-"Reasoning ( i.e. inference ) operates not:
on the unknown or the definitely known object, but only om
the doubted object"
NotesThe question itself is due to the loose wording of
the author in saying that, in all the three steps leading to
q*W& it id t i i e cognition of smoke that serves as cause for
inference; and as such this question and answer are not of
much import For 'all schools of throught etn^fra is the
most important and neccessary cause for inference ( vide S
M. quoted in p. 61 ) and therefore to question why there is no
inference when the causey is not there would be wide of the
mark. The second part of the objection ( why not infer fire
in the kitchen itself after ascertaining wn% ) has some meaning. But the answer given is according to the older school
of logicians who hold that q<f ( = subject ) is that in which the
existence of the - e w is in doubt ( rf^rcrapsircr^: ), to which
we shall advert in para 57 ). The modern school of logicians
would answer the objection as follows :"Yes; one can infer
fire in the kitchen also if he so wishes; but after enftTR, he
must have tRfflPlf- ( 3|^TOt|tfTr*i ) and in the next moment
he will have the inferential cognition ( 3%*T^*RT )." In
this connection a remark of Gagea under q^m is interesting.

f t f ^ 4 i

H f u " ( Lovers of logic

would like to establish an object by inference even if they have

ascertained it by actual perception. )

69

[39] srq- q^nraTTprq" <jft

frar11
TranslationQ. When a man," on reaching the mountain,
just sees smoke ( issuing from it ), why does not this knowledge of smoke lead him at once to infer the existence of fire ?
At that time doubt about the existence of fire is certainly there;
for, in the absence of any evidence for or against the presence
of fire, it is legitimate to doubt its existence. A. True; but,
just as one who has not ascertained s*n% does not have inferential cognition, similarly one who has ascertained it but has
forgotten it, also does not have such cognition; because the
recollection of3fT% (at the right time) is also a cause for
inferential cognition. On seeing smoke ( in the mountain ),
one usually recollects 537% ( by the law of association ) in
some such form as *all those (places) which are smoky are also
fiery.* After observing smoke ( in the subject ) and after recollecting ssrfi?, a third cognition ( connecting the two ) such as
'the invariable concomitant smoke exists in the mountain'
arises; only this, and not any other cognition, causes the inferential cognition ( of tire ). This cognition is argHR ( inference, the 5RC for srgfef ) and is identical with fe^TCTOI. Therefore the definition that argtiTO 5 ffTOTCT is
well established.
Notes The author's intention in the above two paras is to
emphasise that TOH3T as a third step is essential for bringing
about inferential cognition, as opposed to the view of the
Mmmsaks and other schools who hold that srgffr arises
directly from s^f^FT. The Bhttas hold that ssrife^ra is a
synthetic cognition with reference to certain observed particulars and is quite sufficient to lead one to infer the ^ ^ in a
subject where the i| is observed to exist, provided there is no

70

TARKABH^S (

suspicion of. ^frercpr. While the Bhttas do not deny

the existence or occurrence of qrorcr, they hold that it does
not Become a cause, much less a ^or, for sgffif^; and where
ever such a Complex cognition as tftFflr occurs^ that itself is a
case of inference. The Prbhkaras differ from the Bhttas in
their general outlook towards strife which ( io them ) is free
from all limitations of space and time and thus assumes the
fofmof universal generalisation sis in the Naiyyika system.,
But unlike the Niyyikas, they too do not accept WTO in
between-fenff and orf f%. Tue Vedntins follow the httas
in the tjreatrnent of the SOTTOS in general including inference,
It is nqj) - ^considered necessary here to encumber the beginner
with all the $ubt;l arguments for and against the . acceptance
of-ironf- Qf all ,the Indian schools of philosophy it is only
the Naiyyiks who insist on its acceptance, and we
shall advert to it again under members of syllogism ( vide
para?41 )- i

-, *:

. ..

...

. .. ...

;.

, . , = .:

* X ^rfrTF| i

is of two kind$; ^ m for -one's own

sake of-another.? The former is that
( process ) by which one, g:ets convinced in his own mind, Por
example, after ascertaining s^nf^c between smoke and fire in
the kitchen etc, by the special perception ( as mentioned
aboye;),; if one happens to go near a mountain and sees.an
unbroken .line ,of smoke, issuing from the mountain aod
jeacWog to the skj.es, he doubts the existeqce of fire there anf
(immediately } recollects the einf^T *where there is smoke there
is fire', through the impression that gets aroused by the sight
of smpke. Then he makes sure that such a (concomitant)
smoke exists in the subject ( w ) before him and, thereby

7t
concludes ( in his mind ) that there is fire her also ;( i.e.-. in
the mountain ). This is called ^rofgflK.
NoiesThe division of inference into ^ m and yw%
though not found in the Sutras of qterR and wnj%, was first
enunciated by TO^ftqi^ in his Bhsya on the Vaisika Sutras
ofspifT^. By ^ m ^ one clears his own doubts, and after
getting himself convinced about the existence of the probandum
in the subject, as described in the above para, he puts forward
certain propositions in a logical order, so that the hearer also
may get the same sort of conviction. The inference drawn by
the hearer,from the propositions so used by the former is called
, whLh is being explained in the next para.

[41] p^ ^fkqr FP ^TRf^j^nr yt 3ffafir5

^flf
:
I
Translation--When someone, after inferring for himself
fire fro in smoke ( as explained above), puts forth a fivemenbejre syllogism to carry conviction to another, that infer*
ence ( i.e. the one that occurs in the hearer's mind ) is calle, d
inference for another ). The five members are :
( 1 ) The mountain has fire; r
( 2 ) Because it emits smoke;
( 3 ) All that emits smoke, contains fire, as the kitchen;
( 4 ) This mountain is like that ( i.e. emits smoke concomitant with fire )'.
( 5 ) Therefore it contains fire.
By .means of this syllogism which contains the five members
ec
t > *ke ^W or probans happens to be represented as
possessing all the five requisite qualifications ( q^sftqqsr ); and
through this ( i.e. the fef ) the hearer also gets convinced of
the presence of fire ( in the mountain ). Therefore this is called
or inference for the sake of another.

TARKABHAS (
>tesThefivemembers of syllogism are :
1 ) srfsF =The statement of the proposition to be
inferred or proved, as q<&ft 3 % ^ .
2 )f3

The statement of the probans by means of

which the ww is inferred, as ^STg; or
t|*n^. The difference between these two
forms is that when we say t{jR3rc3T3., the
fact that the smoke exists in the subject
is brought out more forcibly than by
saying ftTr. In other words,
means 'because of smoke' while
means 'because of the existence of
smoke in it'. Naiyyikas prefer to use
this latter form when they want to state
a reason.

3 ) gr^^a = The statement of the nfr between the

c m in the first member and the | g in
the second, followed by an example
( other than the subject ) where the coexistence has been observed as sft ^t
vjjWT^tf ^fssflT^*n nrarg:. nfar is
stated in two forms of which the first
is that used in the text, in which the ^g
and *TI*ST are stated as qualifications of
the common substratum. The other
from is *ra re 1ITC93 tfssf: in which
the subject is referred to in the locative
gender and the | g and nsq are predicated as existing in it. The duplication of the relative and the correlative
in both forms is only to emphasise the
universality of the 3sni%. The former
method of stating the nf^a1 conforms
more with the form of the other members of the syllogism where the emphasis is more on the subject or w ; while
the latter form would present the ssnf^r
more naturally and so is preferred to

73.

express s^ifi^r alone but not as a member

of the syllogism. If the latter form is used
\ the example following it will also have to
be put in the locative ( *w\ ^Trfe ). Even
then it would not join up naturally with
the next statement tfn ^ R ^ .
4 ) ^tRq* Statement of the TOTO i.e. combination
of the sqriffSn^ and qTOR$cTT9R (as explained earlier ) as nn\ tgrraq. The word
*r!T stands for *nrei<^%g*!T^ and SRT for
ther subject in question.
5 ) f*m*R = The conclusion as aturan; the word
standing for ns^F;. The subject are is
supplied from the previous sentence.
The first member prepares the hearer as to what exactly he
is expected to infer. Only when his curiosity is aroused, it is
natural for him to ask "How or on what grounds do you say
that ?" So the reason is stated in the second sentence. Not
satisfied by hearing the naming of the reason, he further enquires
how this particular reason proves the existence of the *nr.
So the third member states the concomitant relation
between the | g and mw as also a particular instance where the
hearer might have actually observed the concomitance. As
soon as ssnff is realised the hearer is asked to combine the
two cognitions gained by the second and third members into a
complex one such as CTcqcqcqlgm^ m :- When this tfcro^
arises and when there is no impediment, the inferential cognition arises and to show that the ^g proves the propositi'on conclusively without being vitiated by any counteracting influence or by any stultifying evidence, the proposition is again
stated as proved ( cf. Q. E. D. under Euclid's theorems the
demonstration of which has some resemblance to this fivemembered syllogism. ) These are called sr^q^rs as they form
parts or limbs of the body of q^mgTR. The syllogism is also
termed ^rra or sqmsRt and q*ii?<|5m as ^resrh?* or
The Mmmsakas and Vedntins hold that all these five
members are not necessary, but that either No. 1, 2 and 3 or

74

TARKABHS (

Nos. 3, 4 and 5 would be sufficient for <rcraf|7TH. The Buddhists

go a step further saying that only two members, Nos. 3 and 49
would be sufficient for the hearer to form an inference. The
Naiyyik stand is that ra these ases of truncated or incomplete
syllogisms th hearer's mind would not pass through a methodical system of reasoning, that he would be put to a lot of
menial searching foi the missing links, that it is because of
the quickness of the mental activity that this searching is not
taken into accriM, and that therefore it is better to state all
the steps in the form they advocate to avoid all doubts or
stultifying influences. There is no ftnd to the arguments oo
both sides and the difference, Which is mainly based on the
fact that these-sahools do not accept qwref as a cause of srgwt,
will ultimately, reduce itself to one of form rather than of
substance,. In this connection the following quotation from
is significant
ft

re he, Nyyja^syllogism;Jv^ith that of

Aristotle. ,rjit) mustrbe reniembered tha.t the N^ya system does
not recognise anything cprresponding to the syllogistic figures
and moods accepted by Aristotle. A typical Nyya syllogism
can be compared in some manner to the Barbara type of the
firstfiguJT^rOf,Aristptle ; j^hicibt consists of three members as
f o l l o w s ;vj

.-..

, i

., . . . . . . . .

L All men are mortal( called the major premise )

2. Socrates is a man (

minor

)
3. ,\ Socrates is mortal-(

conclusion
)
The first step expresses the ranfe and therefore corresponds
to the ?3Tf *ui ( without the specific example ) of our author.
The second step states the existence of the reason in the subject-i.e. W^HVT; and corresponds to the 2nd member %g. This
may also,be : taken as corresponding to the 4th member
as i is stated immediately after sqT%, thus representing
ftfifSTOW^T or q^m^. Aristotle, however, does not postulate
anywhere such a complex cognition as q^rw^f in the process of
inference. The last step evidently corresponds to No. 5 of
the Nyya system ( %w*T ) ( or to the 1st member sjfr^ ).

75

To satisfy the Naiyyika fully the above syllogism should

be modified a little and rearranged as follows :
= Chaitra is mortal.
= Because h$ is a man.
3 ) ^rt ^ft w g ^ : ff w4*l
= Whoever is a man is mor
.nnff:
( ^^gfO )
tal; just as Maitra.
4 ) Hri ^ T ^ ( i.e. waftgsin^5 ) cTOOTcWT ( i.e. WW> )~
)

= Chaitra is like that.

= Therefore
mortal.

Chaitra

is

Similarly the Nyya syllogism can also be reduced to the

Aristotelian form by changing the order of ( 2 ) and ( 3 ) and
by omitting ( 1 ) and ( 4 );.e.g s
Nyya
1. ^ra* J&WtarM

Aristotle
LAU who are virtuous are
laudable.
2. p^ffij
2. Chaitra is virtuous.
3. sft wt gcrj^T^ H $T ^T^rftiv 3. Therefore vChaitra is laudar I
able: *

It will be noticed that this form of Aristotle compares

well with that accepted by the Vedntins ( cf. quotation
aboye from ^. *?> on pre-page )., f
-.;.. ;,.,.,_.
As regards the other three figures and the other three moods
of the first figure of Aristotle, they have all to be reduced to
the Barbara mood ( wherever possible ) and then reduced to
the Naiyyika form. It may however be observed that the
moods Darii aud Ferioque of the first figure are not admissible
under the Nyya system as it does not recognise a particular
conclusion. M ore ver, our Nyya system does not attach any
importance to the artificial distinctions f A, E, I and O propositions which play such an important role under the Aristotelian system.

76

TARKABHASA (

Some may still doubt about the necessity or utility of all

the five members of the Nyya system especially when the same
conclusion is arrived at by using only three of the members (or
even two as the Buddhist contends ). One of the best answers
to the above question will be found in the following quolation:"If syllogistic expression, like any other expression directly or
indirectly presupposes a hearer to whom it is addressed, if
Nyyaprayoga or syllogistic expression finds a place only in
inference for others ( <ftTO<[9TT?f ) and if the process of reasoning in inference for oneself ( oiraf^TTR ) is not syllogising, a
strictly logical debate, as recognised by Gautama and his followers, must involve a self-contained syllogistic expression as its
main part. The aim of such a self-contained syllogism is to
enable the hearer, first to specifically think of what has to be
demonstrated; secondly to learn what the reason is, thirdly to
understand how the universal and invariable relation which
forms the basis of inference is arrived at through observation,
fourthy how the reason actually relied upon is identical with
what is known to be invariably concomitant with the pro*
bandum and fifthly to realize that the probandum is conclusively
proved by a probans which is not vitiated by a counterprobans
or by a stultifying proof. As already indicated these five
requirements can be fully met by the five members of a syllogism, viz. sff^, 3> 3?lSW, s q ^ and fc*R. It will be
seen from this that the five-membered syllogism of Gautama,
far from comprising any superfluous member, is the only
complete form of Syllogistic expression which could enable a
hearer's mind to pass in a methodical way through each of the
five stages of demonstrative reasoning, as indicated above."
(A Primer of Indian Logic by Prof. S. Kuppuswami Sastri
p. 226. ).
The five requisite qualifications of a f^ referred to by the
author at the end of para 41 of the text ( ' qRre<frTTOI%ff ^ ) will
be further explained by the author under para 48.
Gautama, as already stated above, has not expressly
divided inference into *3Rand q^af. His division of inference
into three types as given in Sutra 1-1-9 is as follows:( 1 ) ^ ^ r ^
i.e. reasoning from cause to effect as in inferring an ensuing

77

shower from the appearance of dense clouds; ( 2 ) i f a ^ , i.e.

reasoning from effect to cause as in inferring a past shower
by seeing an overflow of the river, and ( 3 ) TOP^rat ss which
comprises all other inferences that follow from reasons other
than cause-effect relations. The inference of soul (S?S^) as
a substratum for gf|[ ( knowledge ), ^X^T ( desire ) etc., may be
cited as an example of the third type. Commentators have
given various other interpretations for these words ijsr^ etc;
but these have only added to the already existing vagueness of
the Sutra, with the result that it is not possible to get at the
criterion which Gautama had in mind for this division. The
TS^pqw school beginning from Gangea takes no notice
of this division and has stuck to the division of inferences into
SWfST and q^m as given by E^RCN^ in his commentary on the
V. S. It was not borrowed from the Buddhist Naiyyika
Dinnga as Prof. A. B. Keith would like us to believe ( vide his
'Indian Logic and Automism' p. 106 ).

This division of srgWT into wrf and q^m is one of the

most vital topics in Indian logic, it clearly enabled the Nyya
system from falling into the restricted scope of formalists logic
( as in the case of Aristotles's system ). It shows how the
deductive and inductive processes of reasoning merge together
to form the complete process of philosophical thinking called
srgsiR. It never viewed that formal validity is something
separated from and independent of material validity. Those
critics who regard the five-membered syllogism as "in part
vestigial, rather than determined by the requirements of logical
analysis1," would do well to ponder over the subjoined remarks
about Indian Logic from the pen of a great western philosopher.
"This error, which appeared very early in our western world
has spread during the centuries and yet dominates many minds;
so true is this that 'logic' is usually understood to mean
*illogic' or 'formalistic logic'. We say our western world,
because if Greece created and passed on the doctrine of logical
forms, which was a mixture of thoughts materalised in words
and if words become rigid in thoughts, another logic is known
which, as it seems, developed outside the influence of Greek
1. Vide Randle's Indian Logic in its Early Stages p. 167.

78

TARKABHS (

thought and remained immune from the form'alistic error. This

is Indian logic, which is notably anti-verbalist
Indian
logic studies the naturalistic syllogism in itself as internal
thought, distinguishing it from the syllogism for others, that
is to say, from the more or less usual, but always extrinsic and
accidental forms of communication and dispute. It has not
even a suspicion of the extravagant idea ( which still vitiates
our treatises ) of a truth which is merely syllogistic and formalistic and xvhih may be false in fact. It takes no account of
the judgement, or rather it considers what is called judgement,
and what is really the proposition, as a verbal clothing of
knowledge; it does not make the verbal distinction of subject,
copula and predicate; it does not admit classes of categorical
and hypothetical, of affirmative and negative judgements^
All these are extraneous.to.logic whose subject is the constant,
^'knowledge considered in itself.'' .[Logic as the Science of
Pure Concept by Benedetto Croce, translated into English by
Douglas Ainslie ]

[42] sr^r ^sfafflPT irrst, mt

f()

|g: \ ^

"

[43]

11
121-122-123
a
b

79
Translation42Here ( in the above mentioned syllogism )
tlie presence of fire in th hill is the probandum ( sni )=that
which is to'be proved or inferred, and the presence of smoke
is the reason or probans ( $g or snw^r ). This reason ( smoke )
is of the affirmative-negative kind, in as much as it has invariable concomitance (with fire) both in the affirmative and.
negative ninrier. For example, 'where there is smoke there is
fire, as in the kitchen' is the affirmative type of wif^af, for in
the kitchen the presence of smoke and fire are observed directly.
Similarly, *where there is no fire there is no smoke as well, as
in a. lake' is the negative type of ssnfr, for in the lake the
absence of both fire and" smoke are observed.
43The( nature of negative ssnfr (i.e. the peculiar chaacteristic which makes it distinct from the other type ) is this;
viz.. it is the negative of the prevaded ( s*npr i.e. the reason )
iti the affirmative concomitance, that becomes the pervader
{ 3sn<735 i.e. the probandum) and the negative of the pervader
(in the affirmative variety) that becomes the pervaded
(in the negative s^ifr ). This has been stated ( by Kumrila
in his loka-Vartka ) as follows :
1The character of being 'pervaded5 and of being the
'pervader' existing between two positive things gets
subverted when the negations of the same things are
:
considered.
2In the affirmative *m$tt, the probans is the 'pervaded'
and the probandum the pervader. In the negative nfr,
the negation of the probans becomes the 'pervader' and
the negative of the probandum becomes the 'pervaded'.
3While stating aqnf, the 'pervaded5 is stated first and
then the'pervader'. When nf^f is stated expressly
i this form then it becomes clear in its true form ( i.e.
the character of being 'pervaded' and 'pervader's
s^TqWT^becomes easily discernible. )

[44];

80

TARKABHS ( TO

I ],

TranslationThus in the case of the reason, presence of

smoke, s^nlfor exists both in the affirmative and negative
manner. But in the example portion of the syllogism ( 3TT?*r
ST3% ) only the affirmative concomitance has been presented.
This is because the statement of one form of ssnfr alone
will serve the purpose. Even there, the affirmative one has
beeo chosen because it is more direct ( or striaght-forward of
the two ). When a thing can be accomplished by the direct
method, it is not proper to resort to the indirect ( or complicated ) method. Therefore ( it is established that ) the reason,
TTOSR^, is of the affirmative-negative variety To the same class
(i.e. that which has both kinds of concomitance) belong reasons
such as *created-ness' ( ^cff ) etc* which are used to establish
probandums like aimpermanence' ( aife^qf ) etc. [ For
example, sound is not permanent ( or is non-eternal, probandum, ) because it is created ( probans ); where, there is
createdness there is impermanence as in the case of a pot
(affirmative concomitance), where there is the absence of
impermanence there is the absence of createdness as well, as
in the case of ether (OT=STCI ) 1

[45] ^fq^g;: ^^qf-^fft i

( b ) This sentence cf*(S|T...*TJT^" is not found in the other Mss.

It is possible that it might have got mixed up with the text from some
commentary ( cf Ghennu Bhatta ). The B. S P. S. edition has rightly
omitted this sentence from the text.

81

a f f a ^ ^ sre qrar ^ i

Translation 45Some reasons are such as have only negative

concomitance; e.g; the reason 'presence of breath etc*, when
put forward to prove the probandum 'presence of soul' The
syllogism in this case is as follows:( 1 ) The living body
has a soul; ( 2 ) because it breathes; ( 3 ) that which has no
soul does not breathe, as a pot; ( 4 ) This living body is not so
( i.e. it is not non-breathing );( 5 ) therefore it is not so ( i.e.:.
not without a soul, i.e. it has soul ). Here, in the living body
( subject )9 the 'presence of soul' is nq ( probandum ) and
'presence of breath etc' is the reason. This (reason) has
only negative concomitance, as there exists no affirmative
concomitance. For ( if there had been one ) it should be of
the form 'that which has breath etc. has a soul.' But there is
no example'as in the case of this', as all living bodies have
been included in the subject ( *m ).

[46]

f )
i

Translation 46All definitions are also instances of reasons

having purely negative concomitance, as when we define
'earth' (substance ) as that which has smell ( as its characteristic ). The syllogism would be :The thing in question ( fqW5[*%^ ) is to be classified as earth; because it has smell; that which
cannot be classified as searth' does not have smell, e.g., water..
6

82

TARKABHS ( a

Similar is the definition of sniTor as the instrument of valid

cognition. The syllogism is as follows :Perception etc. are to
be classified as SWS; because they are instruments of valid
cognition; that which cannot be classified as a STWT cannot be
an instrument of valid cognition, e.g, an invalid sense perception ( 5rsn?T*T*T ) etc. If in this case, there be affirmative etnfr
in the form 'that which is an instrument of valid cognition is
always classified as vmvn, no corroborative example could be
pointed out ( as invalid sense-perception shown above ),
because all sren^s are included in the subject (TI^).
In this
inference the probandum is the fact of 'being classified as qswn9
and not that of 'being EOTn'; because SUO being synonymous
with 'instrument of valid cognition' the probandum would . be
identical with the probans which will then lead to the fallacy
'non-difference from the probandunC ( 'n^t?' ) Thus have
been explained reasons which have purely negative concomitance.

'[47]

Translation AlSome other reasons are such as have only

affirmative concomitance; e,g.; ( i n the syllogism)

Speciality ( fk$m ) is namable.

Because it is cognisable.
What is cognisable is also namable, for instance
a pot;
This ( frtq ) is also such ( cognisable ),
Therefore it is such ( namable ).
( a ) In this syllogism ail the printed texfs except B read ^53[ as ^
instead of ik^m wMch is the reading accepted by the commentator
Ghennu

83

Here in 'specialty5 ( W )> namability is the probandum and

cognisability is the reason which has only affirmative concomitance. The negative concomitance ( if there be one )
would be of the form 'that which is not namable will also be
not cognisable5 but no corroborative example 'like this' could
be pointed out ( as cpot' in the previous case ), for in all such
cases only a well-known thing can be cited as example and all
such things are cognisable and namable.
Notes ( 42-47 ) After explaining the Nyya syllogism the
author proceeds to explain the nature of the f^ ( reason ).
It will be noticed thaHn the S^T^OI^HCT of the syllogism .the
s?n8 is stated directly ( or affirmatively ) as *m tfrca3Tf&:, *mi
s?gra. This 5*jif% can be stated in an indirect manner also, i.e.
negatively, as 3*31 3$3v?i533 ^OT**: WT $. Those acquainted
with Aristotle's system can observe that this negative 3in% is
obtained from the affirmative one by obverted conversion
as follows :
AffirmativeAll smoky places are fiery.
{*ra
By obversionNo smoky places are non-fiery.
( m *Rt f
By conversionNo non-fiery places are smoky.
( ers qf <m
By Contrapositionall non-fiery places are non-smoky.
It will be observed that by introducing two negatives in
the aveera r affirmative 3*H%, the s^i^ (probans ) and sqm
(probandum) in it automatically change into the sqq^ and
sqi^ respectively in the s$Rr^ or negative sqif^cT. There
should not be any difficulty in discerning this if the
is remembered in its simplest form ^ *m&i t^ s^iw ( =
and y
All fgs however are not capable of taking both the forms
of ssnfe as in the case of the stock example with tjyi as STI^T
and C% as s^iqs^. Some are such as can have only either of
the two. When ssnfg can be had only negatively that f&w iscalled ^fftqfcrf^ and when only affirmatively that {&$ is

TARKABHS (

84

those which can have both forms o

are called

The full syllogism in all the three cases are repeated below
to make the distinctions clearer :
(a)
1)
2)
3)
4 ) *r

[ cT*.TT ( =

5)
)
(b)
1)
2)
3
4)
5)

(c)
Affirmatively

1 ) qsrat *fam\
2 ) ^R^c^i^L

3 ) ^^^5^^i W
4 ) m *zt ( =
5 ) r^^^^T ( s

negative )
( The human body has a soul ).
( because it breathes etc ).
( What is soul-fes does not breathe
etc. as a pot.
(This is tfsoi.e. not non-breathing etc. )
( .'. this is not so i.e. not soul-few
=this has soul ).
affirmative )
( The category fiH is namable ).
( Because it is cognisable ).
( What is cognisable is namable
as a pot ).
( This is so (i.e. cognisable),
( .'. This is such ( i.e. namable ).
affirmative-negative )
Negatively
same
same

(
RIT ( = H

In cases of negative ssnfB the conclusion is in the negative

form and only after combining the >wo negatives we arrive at
the positive conclusion.
If in the %35Krf<l^f example, we try to put the wnfr in
the affirmative form as cwhat breathes etc. has a soul', we have;

85

to seek an example to corroborate this ( similar to the 'pot'

exemplified in the negative sn% ). Such an example is not
forthcoming, because call that breathe etc' are included in the
q^r ( = subject ) and an example should be one other than the
qf itself. In the case of the ^T^fr, the negative ssnfr would
be of the form "whatever is not namable is not cognisable' and
here also no corroborative example could be had, as there is
nothing which is not namable and at the same time not cognisable. In the former case an example based on identity
( *rw ) with the q^ is wanting as the q^ is co-extensive
with the abode of the mw. So the % ^ ^ S ! T ^ % is also distinguished as nq^SHT?c^fr. In the case of the ^9&T?3fc,
an example based on difference ( fsn4 ) with the q^ is lacking
as both the m^ and fef are co-extensive everywhere and their
negation cannot be had anywhere. So this is termed as frq^rand fq^ are being explained below in para 51.]
The importance of stating a corroborating instance in the
would be e'vident from what has been stated above.
Gautama's definition of 3r?TfjTi<n in the Nyya Sutra is as
follows :H!*raw*T?qL ^cwreft 1&FQ 33[nK*tt ( 1-1-36 ) and
'afjE^talT fqOq' ( 1-1-37 ). It is on the basis of these two
Sutras thatJ later commentators and authors have elaborated
sqrfcrinto affirmative aad negative forms and from that the fef
which takes both or either of the forms. Freely rendered the
substance of the above two sutras would be ^ioT is the statement of an example which corroborates the relation between
the probandum and the probans ( ttirorarasrarc: = nf ) through
some characteristic ( which it possesses) based on identity with
or difference from the probandum. The Bhsyakra taking the
example c$is^isf%<q: ?*q%K3fq? comments further r'
^O^W

\
l

That a mere enunciation of a corroborative instance as

or rTF TTsf^9 does not constitute 3?i*< is evident
from the fairly long and pretty vague qualification attached to
the above sOtra. Moreover, there would then be no

TARKABHASA ( <r

86

difference between gFc-~ which is a category by itself according to Gautamaand ^ T ^ O T which is only a member of
the syllogism. The JWHfK tries to bring out this point by
showing the derivation of the word 33T5W in the last
sentence quoted above. 'That which illustrates the relation
( 32TTpcT ) between two things which function as %ww and ^re^r
is 33.fCOT.' If the corroborative instance does not illustrate the
csrTfe whafc-else is it supposed to illustrate ? It is thus fairly
evident that both the Strakra and the Bhsyakra have in
their minds the enunciation of the *m$B also in the 35157^
"3TW, which is followed by the instance where the hearer is
supposed to have observed or understood the *mf%. Why
then did the Strakra or the Bhsyakra not state it explicitly
is a query wide of the mark5*fton rgsftTRfV^ \
The Mmmsakas and Vedntins do not admit isrararfsF as a means of inference, but include it under a separate
. j called wrqf or 'presumption' which the Naiyyikas do
not accept as such. This will be adverted to further on
under para 65 where our author refutes snjiqfr as a sepa*
rate

STOW.

The threefold divison of smfr and similar points connected

with n^THnicwra can be illustrated figuratively as well,*
A
B

&

1
trr

F
VTrt

If
D

^- %
G

1ST

(a) This figurative illustration has been borrowed from the edition of
T a r k a Sagrah in the B S P S with grateful aknowledgement but with
some modifications.

87

In the figure on pre-page which shows three squares one

inside the other, the largest square A C represents any
class higher than the Ht^, gf%, say 53^; the middle square
H K the sphere of ?fl> and the smallest square EG the
sphere of t?. The larger gnomon AJC will then represent the
sphere of 3qw3; the largest gnomon AFC the sphere of
SJTmm and the gnomon HFK, the sphere of 3% without tjjl
( or with tgtrrara ). In 9f?eros*nfrwhere there is smoke
there is fireonly the square E G comes into picture as that
is the area where smoke and fire co-exist. Also smoke occupies
less space than fire and so the former is *m*it and the latter
s^Tfq^. If we state the above snfor in the wrong way 'where
there is fire there ' is smoke', the sq. HK comes into picture
where there is both smoke ( sq. EG ) as well as 'no-smoke'
( gnomon HFK ) and so the sqrfr is not valid as the
( fire ), occupies a larger space than the s*nqsp smoke.
In strRrc^arra'where there is no fire there is no smoke
the relative clause takes us to the gnomon AJC where there
is 'no smoke' also. So the ssrrRcT is valid and the *TT<ST 'no
fire' occupies less space than the S^TT^'no-smoke'. If the
same stnfr be stated in the wrong way, 'where there is no
smoke there is no fire', the gnomon AGC comes into picture
_where there is 'no-smoke'; but that includes the gnomon HGK
where there is fire. Here the e*nqr$ 'no-fire' occupies less
space than the S*TO 'no-smoke' and so the nfar is invalid.
In both cases the gnomon HFK represents the area of invalid
ssT^r or the area of reasons vitiated by sqfr such as wet fuel
( vide p. 63 ). The discussion of ^Tifr is taken up by the
author under 'fallacies', vide para 53 below.
In sferasfg^fsp^ such as 'the living body has a soul because it breathes', the q^, HTT and | g are all co-extensive
and all three are represented by the middle square HK. Since
a corroborative example has to be one other than the trg we
have to seek it from outside the square HK, i.e., the gnomon
AKC which is the area representing 'no-soul' and 'non-breathing'. This is the same as saying that no cTOBircT can be
had and so no affirmative nf^9 can be possible in such cases.

TARKABHASA ( * &
In sfe^T^f^ such as c fta *s namable because it is cognisable', the tfpar and |? are co-extensive and both are
represented by the largest square AC and the qg is only a
particular thing from it. So another particular thing such as
ET2, sre? etc. can be pointed out from the same area; but no
example can be had from outside that area as there is
nothing which is 'not namable and not cognisable' at the same
time. This is what is meant by saying that a sN^^flg has
nofq^TOT?aand so has no negative ssnfr.
In the case of eroqeirfcN^ ^snfor the square EG, the
gnomon AJC|and the gnomon HFK represent respectively
the areas of rq^r, fqw and gqfr; and the q^ happens to be
any particular from the square EG when the nnfr is affirmative, or any particular from the gnomon AJC when the
ssnf is negative. In the case of fcwqfTf there is no Hq^
area other than that of the q^ ( square HK ) and the gnomon
AKC represents the firq^r area. For a similar reason there is
no flrq^ area in the case of sfc^pcffi'g as the whole square
AC represents the srq^1 area.
Having explained the f^ in detail the author proceeds
tp explain the characteristics which it should possess to lead
to a valid inference.

[48]

89

Translation 48Out of these three reasons, viz. 'affirmativenegative', 'exclusively affirmative' and 'exclusively negative',
the'affirmative-negative' is capable of establishing itsprobandum
only when it has all the five characteristics and not when it
lacks any of them. These five are :
( 1 ) q ^ i ^ r s Subsistence in the subject.
( 2 ) srq% m^ = Existence in/positive instances' ( = ro^,i.e.
all places where the probandum is definitely
known to exist ),
( 3) fq^i^Tff: = Non-existence in 'negative instances'
( = fq^, i.e., all places where the probandum is definitely known not to exist ).
( .4 ) aTsff^fq^crr s Not subject to any stultifying inference.
( 5 ) 3TOiRr<ret3q[ = Not liable to be counteracted by any
reason to the contrary.
All these characteristics exist in all ( valid ) 'affirmativenegative' reasons such as 'presence of smoke' etc. It is like
this. ( 1 ) Smoke is an adjunct to the subject hill as it is observed in the hill. ( 2 ) It is seen existing in a positive instance,
kitchen, ( along with fire the probandum ). ( 3 ) It does not
exist in a negative instance such as a pond ( where fire also
does not exist ). ( 4 ) Similarly the 'presence of smoke' is not
subject to any stultifying influence for the object of the reason

90

TARKABHS (

( tjwsr ) is the probandum ( srffW^ ) itself and this is not

stultified or contradicted by any means of cognition ( vtmm ),
Similarly tiraicT does not face any opposing or counteracting
reason proving the contrary of the probandum. A contrary
reason is one which will prove just the opposite of the probandum in question. Such a reason is not observed to exist in
the case of the reason 'presence of smoke'. Thus the reason
sp^at possesses all the five characteristics and is therefore
capable of establishing the 'presence of fire'.
49That fire is an adjunct of the subject ( hill) is established by the strength of the reason being an adjunct of the subject in this manner. Inference involves two distinctive factors;
( 1 ) the invariable concomitance and ( 2 ) the adjunctness of
the reason to the subject. By invariable concomitance is
established the presence of the probandum in general (without reference to any particular subject ), and by the adjunctness of the reason to the subject is established the presence
of the probandum in the subject ( in question ) in
particular. Fire is to be inferred as existing only in the hill
( the subject in question ) by the presence of smoke in it.
Otherwise there would be no necessity for inference at all as
the presence of the probandum in general has already been
established by the knowledge of the invariable concomitance
( the first factor ).
50AH other reasons which have both affirmative and
negative kinds of s*nfr ( as smoke ) become valid reasons only
if and when they possess all the five distinctive characteristics
( explained above ). Otherwise ( if they lack in any one of
them ) they become fallacious reasons. Reasons which partake only the affirmative s^nfccf are endowed with only four of
the above characteristics as they do not satisfy the condition
'non-existence in negative instances' as there is no 'negative
instance' in their case. Similarly all reasons with only Negative' onf are endowed with only four of the characteristics
as they do not exist in 'positive instances' for want of such
an instance.
NotesThe five characteristics mentioned in para 48 are
nothing new, but have been collected together from the dis-

91
eussions about ssnf, wsmm and f>f in the earlier paragraphs, to enable the reader to judge whether a particular
reason is valid ( i e. capable of establishing the probandum ) or
fallacious. The first one q ^ m m is self-evident as without the
reason actually existing in the <?W no inference is possible.
This wards off the fallacy called arfef= that of unfounded
reason. The second eq% w*k warns against the fallacy forta =
that of contradictory reason. The third frq^Tfsffi keeps the
reason from slipping into the fallacy swfrH = that of inconstant reason. The fourth characteristic sr^lfcrf^rac?
saves the reason from being stultified by any other reason,
which is the fallacy called ^TST. The fifth one srtffq^R^
protects the reason from being dragged off by any other
reason proving the opposite of the probandum, which fallacy
is named Hcsrfq^m. These characteristics help one in an
argument to steer clear of the fallacies which the opponent
may try to haul up if possible. What these fallacies are and
how they prevent valid inferences being drawn are being
explained further on from para 52.
Para 49 briefly summarises the Naiyyika attitude towards
inference in general. In this connection the notes under paras
39 to 42 may be referred to.
Similarly para 50 is a recapitulation of the three kinds of
f<?5^ explained in the earlier paras.
In the above discussions the meanings of the terms q^r, *T<rer
and fq^ have been takn for granted without being defined
properly. The author now proceeds to define them in the
next para.

[51] %JT: ^-im-f^FTH ? 3^Rt I

ra Wf qren I mi wTTj^r" <T#CT: l f ^ r ^ n ^ ^ anff

TranslationWhat is meant by qw, R^T and fq^r is being

explained ( below )> q^* ( = subject ) is that in which the
presence of the probandum is in doubt; e.g; in the inference of
fire ( in the hill ) by seeing smoke issuing from it, the hill is the

92

TARKABHS (

subject, rq^1 ( positive instance ) is that in which the probandum is definitely known to exist; e.g., the kitchen in the same
inference as above, fq^ ( = negative instance) is that in
which the probandum is definitely known not to exist; e.g., the
pond in the same inference.
NotesIt is doubtful whether any real purpose is served
by the words w and qrf in the above definitions, Chennu
Bhatta's comments here are interesting,
i" This is not a satisfactory
explanation as any statement of a definition is always backed
by authority and the use of the word t=nif does not make it
more authoritative. He indicates this clearly in the next
sentence : "tf^rcisi^ W, fafT3*r mW-,
This definition of q^ as 'that in which the existence of the
probandum is in doubt' is according to the older school and is
open to the fault of sisinfr. A man inside the house hears
thunder and immediately infers that the sky is cloudy and he
had no doubt previously whether the cloud was there or not.
This definition would also preclude a man, who had already
ascertained fire in the hill by other means, from inferring it
again on seeing smoke issuing from there, as he has no doubt
and hence no w . The modern school has therefore adopted
another definition, fqrafaqTf*$f^EfH^viTsr: q^cT' and q^
as the substratum of this u^m. fpw$fp* means 'the desire to
prove the existence of the probandum'' and ff^'is the definite
knowledge of its existing in the subject, and this has been
qualified by the adjective fqTOfwfftf. So this definition
means that q^r^i is the absence of certain knowledge about
the existence of the probandum, which (knowledge) should
at the same time be bereft of the desire to infer the same.
This expression is not so confusing as it might appear to be
at first sight It is common experience that one can have
inference if he so desires irrespective of whether he has f?f%
or not. So this gives two ways of having q^ar. When there
is no desire and also no fe% then also there is q ^ n as both
the negatives are present directly. Only when there is

93
and at the same time no desire to infer, there is no weT; for
in this case we have fqrafrSTf^frfEfof and not its swm.
It is clear therefore that either the desire to infer or the
absence of ff is sufficient to provide the needed q^Rn. Here
f% is an obstructing agent of inference ( argfflraf^ra; )
and fnusrfqT acts as a neutralising agent ( 3%3$ ) against
the obstruction frfe Since ^vrcrar9ftfogsrfq?*reTn3 is a
cause for any effect in general, the above definition of q^rar is
only a particular instance of this principle as applied to
inference.

[52]
qr SWT: I

TranslationThus the three kinds of reasons viz., ( 1 ) those

that have both affirmative and negative *m%; ( 2 ) those that
have only affirmative 3*n%; and ( 3 ) those that have only the
negative one, have been explained. All others are fallacious
or erroneous reasons. They are of five kinds only as follows .:( 1 ) rf^ = The unfounded or unknown reason.
( 2 ) f^3[ * The contradictory reason.
( 3 ) anrarfeftee = The inconstant or straying reason.
( 4 ) SROTR = The counterbalanced or opposing reason.
( 5 ) ^5!^5T^^Tqf^S = The stultified or belated reason.
NoiesFrom what has been stated so far it is clear that the
W or | g is the most important factor in inference in as much
as it leads through s^ift and q*ra3T to a valid conclusion. It
is therefore equally necessary to know why particular reasons
do not give rise to valid conclusions in order to steer clear of
them even in ordinary parlance. Such reasons are technically
called fhcTOrar. and their treatment forms an essentially
important, though difficult, topic under
as such is the thirteenth category of Gautama, vide
his stra quoted on page 4 under para 2; and our author treats
of them again in detail from paras 136 to 148 under Part II.
There also he distinctly states :<C

94

TARKABHSA (

I" It means that those reasons which

fall short of any one if these five essential characteristics, q^*W^T etc., become fallacious reasons. This may be taken as
the author's definition of g^Tmn in general.
The word t^T*TO c a n be interpreted in two ways according as how the compound is dissolved. If it is dissolved as
tcftTfVTO: it means %&$*% or fallacy of the reason. If it is
derived as f ^ ^ W H ^ ?% I^T^T; it means | g | g or fallacious
reason. Here the author has given the classification of fallacious reasons according to the second meaning. The fallacies
would then be named ( in the same order ) as ( 1 ) srff^ ( 2 )
r, ( 3 ) 3T?NST^ ( or ssrfo^R ), ( 4 ) smom^qf ( or
), and ( 5 ) ^ ^ i ^ ^ i q ? ^ ( or era ).
As a matter of fact the above-mentioned five-fold division
is based on Agtes and not on ^slgs and therefore the distinction between the two meanings becomes significant. A
reason may have two or more of the above fallacies; e.g. in
the inference qi^t t p r a i ^ ^ the | g has the fallacies ( 1 ) and
( 3 ) of the above list; in %$t ^%^\^^^? the t g has ( 1 ), ( 4 )
and ( 5 ); in l^t vgtn\\ 3f :, the | g has four of them ( all
except 2); and in an inference like m^?^^T^ ?&gt the | g
possesses all the five. In such cases the classification of the
^g|gs under any particular head would be misleading and so
the Naiyayikas have classified all errors that are possible i a a
t g under five broad heads. A ra;g has to fulfil only five conditions and a reason can become <|i by not satisfying any one
or more or all the five of these, conditions. A | g can have
more than one fallacy at a time, but the fallacies themselves
are distinct and separate just like a man who is rich and pious
can at the same time be stingy and crueh This is expressed in
Sanskrit as ^qiWTOf^gqi^OTf?', i-e. the vitiated things may
merge in one substratum, but the vitiating circumstances themselves do not merge into one ( but remain separate ).
Another definition of t^&ro adopted by later Naiyayikas
is
as follows : igfiT%?icwoiT^grwRrsrs:wspiin4s3T'ift^?Hr
'twrmaw*. it means that a I^W^r is that which is the subject
of a valid cognition which obstructs the inference directly

95
or indirectly by preventing the ascertainment of its instrument
qRST or sinfe. From the examples given below under individual fallacies it will be seen that fallacies ( 1 ), ( 2 ) and ( 3 )
prevent nPcfSfR or mHR$T and thus prevent inference indirectly,
while nos. ( 4 ) and ( 5 ) obstruct inference directly. Hence
the necessity of the phrase ^gfafa-cresRoiT^Rrc in the above
definition. At what stage in the process of inference these
fallacies become obstructions to valid conclusions is illustrated
under each fallacy in the subsequent paragraphs.
There is some difference in the number and names of these
fallacies between the Vaisesikas and the Naiyyikas. Kanda's
Stra enumerating the fallacies is pretty obscure 3?qfH3[ts^qtsfrsH^ Hfc?'srt I (V. S. Ill-1-15 ). Later writers generally
follow the interpretation of the commentator STCFTCcrqr who
sums up the view of the Strakra in the following half-verse,
'f^f^f^wfsf
sGTWSBrsftet r Ksyapa ( i.e. Kanada )
declared ^ ^ , srf^; and nf^'W as the three non-reasons
( srf^ ) or fallacious reasons. The^ first two correspond to
nos. ( 2 ) and ( 1 ) of the list given by Kesava Miara, while
the third c^sf is equated to ershncra or eqfroTT. si^srqT?
adds a fourth one sw^focT to the above three; but later
commentators include it under the sigTOf variety of
The Buddhist logicians (vide t*. ql. n.fa. Ch. III. ) in
general follow Kanada both in the number and naming of the
fallacious reasons, as they attribute only three of the essential
characteristics ( omitting the last two ) to a valid reason.
Gautama in his y^u-^-1-2-4 enumerates the five "fallacious
reasons as follows :-^^5^fT^H-i^^-^og^-ei|H-^T^^^i
^WClT-Of these the first one is equated to art%fcre by
Gautama himself in his next Stra as aiTreif^f: ^Wr^R: ( ^ l ^-1-2-5 )0 3?ij$m corresponds to arfe^ and sR^citef to ^T@?Tqf^S Gangesa, the founder of Navya-Nyya, following
Gautama, enumerates them in the same order, but with slight
change in the names as c
Next the author begins to describe the individual fallacious
reasons in the order he has followed.

96

TARKABHS (

[53] (a)
i
!

(b)

(c)
1 srq*^; ^^Tf^rr^T^rq; 1

1
t
1

ffroqimra<r?r 1
Translation (a)Of these ^rfe^is of three kinds^i 1 )
j that reason whose subject ( snsrcr or *i^ ) is unknown.
( 2 ) 33qrf3[ = that reason whose very form (^^q)
is
known net to exist in the subject. ( 3 ) ^^^fe^ = that
reason which has its invariable concomitance with the probandum unfounded. An example of the first ( snsnnf^ ) is
the syllogism, *a sky-lotus is fragrant, because it is a lotus, like
a lotus in the pond.' Here sky-lotus in the subject and that is
totally non-existent.
( b ) An example of $3CTI%3[ is 'sound is transitory because it is visible like a pot.9 Here visibility is the reason, but
that is quite unknown in sound as it is only audible.

97
( c ) The third s^^^fj^ is of two kinds:one caused
by the absence of any evidence to grasp the sn%5 and the
second caused by the presence of ^*Tfer or adventitous circumstance ( vide notes under para 36 ). An example of the first
kind is this: tf Whatever is an entity is momentary, like a
mass of clouds. Sound etc. are entities ( and hence
.momentary )". Here there is no evidence which enables one
to ascertain the ssn% ^^H^cT^fftaFq'. If this fallacious
reason ( w^ ) is also to be included under the other kind of
sgn^snfcfig caused by 3fqif, it would amount to accepting the
view that the probandum ( ^fcrwaf ) exists in the subject.
( d ) An example of the second variety of
( caused by ^qifer ) is this :'Killing in a sacrifice causesdemerit because it is an act of killing, just like killing outside sacrifices'. In this case 'killing' as such is not a cause
( suitSI35 ) for demerit; but 'the fact of being prohibited by
scriptures ( fafRp^ )' *s t^ e cause for demerit. So 'prohibited
ness' acts as 3wf in this ca$e. it is like this. The definition
of grqrfRr is that it is a thing which is co-extensive with the
probandum in question and which at the same time is not
co-extensive with the probans. ff^-3. is co-extensive with
the probandum wrtfsnwsT^, because there exists the ssnff
4
where there is 3TWif<3TO^ there exists fcfcp?'. At the same
time it does not pervade the probans f&i?3, for there is no
such n% 'where there is |^T^ there exists ff^^', as the
latter does not exist in sacrificial killing. Because ff^gr acts
as-^qf in this case, the reason fn?3, whichjts dependent on the
invariable concomitance of some other thing, becomes
( as defined above ).
NotesThe fallacy afrf^ actually means 'the fact of not
being known or ascertained.' To understand the technical
sense in which it is used here the reader may recollect the
general form of q^T^T explained under para 35^l^^sqi^^gW^ qsf : or asofefgf^tgw^ q^:This involves <^ as the
substantive, | g as its direct attribute and m\B as a direct
attribute to | g and hence indirectly of w . The absence of
any one of the three will prevent the complex TOira being
grasped. The unknown or unfounded thing in the word
7

98

TARKABHAS (

refers to these three parts of TOTO and thus we get

three varieties of srfef^. If the w is unfounded the fallacy
is called susmfa^ ( anarcr = q^r or subject ). If the | , is
unproved the fallacy becomes ^^qfjf^ ( ^ ^ q s t h e essential
form of the $g ). If the s?n% is the unproven thing it gives
rise to the third variety iTOftnfef% ( n ^ ^ s w n f g ) .
Though there are thus three distinctive fallacies they are
classed under one broad head rfei% with three sub-divisions
because all three in effect prevent w$?3T, the instrument for
inference.
In the example of 3U$r?nfcl3E given in the text the q^r is
and its delimiting adjunct,
q ^ m s f ^ ^ W , is
( = growing in the sky ). This when applied to qualify
lotus reduces the latter to an imaginary thing and thus the
q^ becomes fR. Another instance of 3n$Kttfl3[ occurs
when the syllogism is expressed as HOTITO: q^tft ^flp^ treibt
Here the delimiting adjunct 3&i^W=l may not reduce the^hill
to an imaginary thing; but it certainly is a superfluity acting as
an unnecessary bar on the ^ when one is trying to infer fire in
it. From these examples nsnnfefic can also be defined as
%
or simply
The second variety ^^qfrf ( as explained above ) can be
defined as q"% i|*Wcr: and the example given makes it clear.
Other examples such as g^t ^TIW^ ^ T ^ 5 'm^roPi' ^cT^^f^
etc. can be easily gathered. When it is known that the |<$ is
absent in the qre, the question of q^psnfcn does not arise and
thus the first requisite of a g$g is non-existent. The author
himself explains in some detail the various ways in which a
%g becomes ^ q i f ^ under para 138 at the end of the book
when he treats of t^TW^'s again as a separate category.
9 the third variety, has already been explained
.as q%^rc33T3*Mfc$TrT3:> This can happen in two ways:
,( 1 ) When there is no evidence for ascertaining ssnfg; or ( 2 )
When the apparent ssnfg is proved to be false by an adventitious
circumstance or factor technically called ^qf. These" two
are sometimes called -n-ttnsBfS^ftcft l : and #qfrit Ig
te^pectivly. The example for the first kind is borrowed by

99
the author from the Buddhist logicians who hold that all
ntities ('rer:) are momentary (^ffoi^).- The syllogism is
therefore stated with only two members as the Buddhists do :
y%TOLTO^fr#, *m &&3W ( TOTO ); and ^ar sis^f?: ( %qm ).
Though the Buddhist may try to justify the syllogism in his
own way the Naiyyika does not see any means of ascertaioiog
the 5^1% here between w% and forour*.. Even if someone
were to doubt that there might be some such ssnfg, a doubtful
one is as good as no 3*nfe and the | g still remains
To avoid the controversy later Naiyyikas give the example
forCTznsr'9fcri3as qsrt 3 % m ^ ^ O T f f ^ n . The addition
of the attribute SREI^PTO to the %g makes it different from
ordinary smoke of the kitchen etc. with which only its ssnfn
with fire has been ascertained by the observer. In as much as
there is no wnfe of ^i^%i7^T^r with fire such a sqrolg does not
exist in the q^ and so the eg becomes fallacious. In
the %g as such does not exist in the q^1; while in
it is the s q p ^ g that is absent for want of the adjunct
We shall have occasion to refer to this point again when we
come to R^if^r^, where also the preventive factor (
) is due to lack of
For the second type of ttTTO&nf5, i.e. etqf^5p|g5 the
author has already given three examples of sqff when he
describes sinft as R W R T ^ ^ I ^ ^ : ( vide para 36 ). Here he
defines 3-qffo in technical terms and shows how the definition
applies to one of the examples alluded to earlier, gr^fw is
derived as 37 - nfrrcffii sT5f=^T^m' ^Tr^%, ^qT^r:;
meaning thereby that 3"rTfr is such a thing that, for the time
being, imparts some of its properties to another thing which
All the printed editions except B read the text here as tf%

fa#,

vw*imyz*z \ ra ^ W3p{?: i"

But eh.

Bh., one of the oldest commentators, does not approve of this reading as
he clearly says " ^ ^ c T ^ c i R ^ ^ ^ ^ q f ^ ^ ^ q ^ ^ g ^ : " I This
syllogism occurs in ifRS^'s ^"U^^T^l^1 as the first line of the
initial verse 'q**ra^e?f<f ^l^To^C^F^Eg *TOT %& ( vide
published by KPJR, Patna ). *

100

TARKABHASA (

comes in contact with it When a piece of crystal is placed

over or near a red flower the latter imparts its colour to the
crystal which then appears to be red. The flower acts as
OTfw here for the appearance of redness in the crystal In the
stock example ^4f!ft ^U^T^l*, contact with wet fuel, anifs^TRtT# plays the role of ^qiR*. It is well known that smoke
is caused only when fire comes in contact with wet fuel. Wet
fuel is as intimate a cause for smoke as fire is and so we have
the valid wnf% *q& ^^3Tf^TOfa:\ In other words enifaf*
sftor becomes *mq*G to %$$the m^ in the above example.
But there is no such connection of fire with wet fuel for either
can exist without the other as in a red-hot piece of metal or
in a fuel stack exposed to rain. In other words sd^wwitel is
not asqprq^ of fire, the reason here; and so satisfies the con
dition for ^qif^. The same result can be deduced by a different]
reasoning. Both fire and wet fuel are causes for smoke and so
ares^Tq^'s in relation to It. By seeing the effect ( $TO.) we
can infer the cause ( qsi^or ) and not vice versa In 6^m^'^%if
the attempt is to infer the effect ( STJT ) from one of the causes
( fs ) and so the other absent cause ( wet fuel ) points out a
place where the co-existence of the f^ with *n5T is. absent.
( eug. red-hot metal ). This is what the author means when

he says '^ita7<nfcft<;pr" in para 36 ( p. 62 ).

-.-:

Three examples of ^qfw are mentioned in para , 36. Of

these the first two have been examined above-^npsrTOtT
and f%f^?w. These two become sn^s^nq^ absolutely, i.e.
without the intervention of any other attribute to their
In other words the stnfTs in the two cases, ' ^ ^ ^ ^
cfcftiTs' and 'sraSW^roreRcc ^ fcfr[<3q' a re valid as they are.
Such instances are classed as *$fCn^53T^qTfe. There
are however instances where sqfw becomes n^wriq^R only
when the H ^ is. qualified by some other attribute. Such
attributes can be taken from the ^TR ( = fcg)# or the y^ of the
syllogism in question or from any other source outside these
two. These are then termed, ( 2 ) ^ W ^ ^ ^ ^ ^ T m ^ , ( 3 )
q^^iri^f^spifOTaiiqiqc and (4 ) ^#^h^i^i^^^TT^%. An
instance of the second kind occurs in the inference of black
complexion in Maitri's child, the third example for

loi
mentioned by the author in para 36. Here the full syllogism
WOUld be
fefe?
W wftRTO^ 3TOH
where ^TcEqT^xgr is shown to be an ^rifr. This is not
to the pwbandum imm% as it is, for the asufi?, ^ CTPN1 a
is not true, e.g. in a crow or a black vessel. But if the
is qualified by the | \ j itself and the safr is stated as
*& Hftft^^re^p^^RSf -^ ^TO<TH33 it is found to be true
in the case of the other children of ^if^ft" who are observed to
be dark.
An example of the third variety of srqf^ can be seen in a
syllogism like mqi STHTO:''gKT^^f&^r^ where
( manifest colour ) becomes an ^qf. Here also if the
is taken witl} the stated m*v[ as sr^f W9K rte^ 1 ?, this is
found to fail in wro???^. If however the m*?* is qualified by
some attribute taken from the q^r ( m^ here ) and the *wf$$
is stated as ^ ^ ^ f | ^ ^ m f ^ ^ R ^ ^ ^ ^ ^ q ^ it is found
to be valid.
The fourth variety of ^ir occurs in syllogisms like
^ ^ T ^ , where ?tm<^ (the fact of being a positive
entity ) acts as ^qifr. To make it eT^^Tq^, the mw has to
be qualified by some such attribute as 5RT?3; for the ssof^T,
ra-fRTRrw cr^ *tmi fails in ^TTVTI^ which is destructible but
not a vm. If the smf^ef be stated as n% W^mmfe^
^ vtmt it is found to be trae, pitfer being neither a
nor a wwst, y$m% becomes an instance of
In the last two varieties of 3fw, ^rT^RT^n^^ should
also be taken with the *TTSR qualified by the same attribute
used to qualify the w*W, e.g. n^
^i^m^^^wn^mvtmu
fl ^f^c ( vtm W^ ). Similarly in the case of sn^sr
Otherwise these will not become sqf's and thus there
will arise the defect fsqtfe. Again these two, ^nssTssnq^ccf
and H T ^ l ^ N ^ f , should be governed by the same relation
( ^ r ^ r ) to avoid the defect ^Tf^i^. Otherwise in a valid
reason like 3%m^i|in^, one can always find an ^qfw which
co-exists with the ^ns^r through one relation and does not coexist with the m*? through another relation. For instance

102

TARKABHAS (

and 99
Thus by resorting'to different
relationships m$t%> (light or brightness) cao be made to act as
even in a safo. This is expressed in Sanskrit as e ^ The vicious nature of 3rqfi? can be represented graphically
as we did in the case of ST%g9 vide p 86. In the figure on
that page it is shown that the gnomon HFK represents the
area where m*n alone exists without the | g , i.e. the area of
3}R*ite^ and similar ^qrfo's. In the case of a #qTW35|g such
as qffircr^:,^ being OTTO occupies the larger square H K
and 9S, the 91*9, occupies the smallest square E G. siuf WSR^q\T9 the ^qrfs being etnTO to the W^9 W ought to occupy
the whole square DB. If so, in the gnomon HFK there should
exist wet fuel and smoke but no fire. This is an impossibility
as smoke is the outcome of contact of fire with wet fuel. ' Thisimpossibility is what anOTirindicates. If 3% were to be a
valid | g for inferring smoke, nif^nNta would become ssnqsc
of rff, which it is not as is seen in a red-hot piece of irpn.
Because ^% strays away ( rfTif1 ) from the ^^ifiy, it wilt
stray away from the sin^T of the ^mfk according to the general
rule s^t ^rniTTO^fr4Rt e cr^^i^ii. 9
This also explains why grqrifw does not prevent ioference or
its instrument directly, but only points out a way to find out
s^rfirgH or inconstancy in the proposed | g . This is expressed
as STTfe^ff^Hte^i^. For the same reason swfw as such is
not considered as a |<cn*UJ. Some later Naiyyikas do
not include *ftqTfc$|g under irr^wffl^; e.g. firwpi mentions
only iihs^T'as an example of ^n^nf^9[ ( Kar. 87 ) and treats
of sqf at the end of the book saying *s*fawwT3*n*g<lr*S.
H' (Kar. 140). Such authors would like to include
under s r ^ f ? ^ . Keava Miara also hints at this
view indirectly when he uses the phrase '
1
in the last sentence of para 53 (d).
One point in the last sentence of para 53 ( c )
,requires some

103
explanation which was postponed till after 3"qif was explained. The author has stated that there is no evidence to grasp
the 32n% '?frat cc^fr^'. Naturally the question ariseswhy
not then find out an sror and classify e*3f as a #qfw3%5 ?
Such an sqTfw exists in sreqfrrwsr or refer or TOsrfctatfrra; for
we have ^ ^o^ cT^qfrW^" qw ^s|*nssT6snq$cgrand at
the same time we do not have '^rw ^# ffsftqfw^*, as such a
n% fails in eterna! things like ether etc.HnOTtnTOKr This
would mean that ^tqffTCKcr is a cause for ^rf^ and that even
though it could not be established by the reason
could be established by sqfw?^ just as sm by
and srcnjgm?* by ff^WT- This implies that ^f^^ is being
accepted; but this position is against the accepted doctrine of
the Naiyayikas who contend that any produced thing can be
destroyed only in the third moment from origin and not in the
second moment as the Buddhists argue. This;in brief, is what
the author means in the above cryptic sentence.
It will be of interest to note that the example of OTTRT cited
in the textfft^^is borrowed by the author from the Jain
authors, just as the previous syllogism is borrowed from the
Buddhists. The Jains hold that all killing without exception
leads to sin. This syllogism, according to them, is quite valid
and is put forward by them to refute the stand of the Hindus.
Readers who find it difficult to understand the characteristic
of ^qfw through its technical definition given in the text may
remember the following verse of fRftr, which in effect means
the same thing but clothed in simpler words.
H ( Kar. V^ 139 )
i.e. All g^ifs are co-existent with the proposed SWfcr; and
in some substratum of the proposed | g , both that ^qiRf and
W^t will be absent. 3T*ft*ites& and qNforf$n are examples of
such substratum.

[54] TT^f^wV^T^ tgf%^: I iT WT

(a)

(a)

104

TARKABHASA (

TranslationThe contradictory reason is that which has

invariable concomitance with the negation of the prohandum
( in question ); e.g. sound is eternal because it is a product,
like ether. The fact of being a product is actually concomitant with non-eternalitys the opposite of the probandum
here. What is produced is always non-eternal and not
eternal. Therefore the reason 'being a product9 becomes a
contradictory one.
NotesFrom the definition it is clear that a i ^ ^ ^ g can
exist only in fijw's and so lacks the characteristic fcws?*.
In si^orT it is a second | g that proves amarrare whereas in
f?l^ it is the proposed %g itself which proves it. The use of
a f%^lg in an argument exposes not only the person's inability
but his ignorance as well. r

[55J ^^f^Rs^f^^: I

TranslationThe straying reason is that which is inconstant

( Sfs^rftrgT* ). It is of two kinds, the common strayer and the
peculiar strayer. That which exists in all the three* w , wn
and f^q^r, is the common strayer; e.g. ( in the syllogism )
sound is eternal because it is cognisable, the reason cognisability exists in all eternal and non-eternal things. That ('reason )
which is absent from bothOT^rand fk*m but exists only in the
q^f, becomes the peculiar strayer, as ( in the syllogism ) 'earth
is eternal because it has smell' the reason smell does not exist
in any *cq^ or fw, but exists only in earth ( W ).
NotesThe word sfl^snf^sK means 'that which does not
stick to one thing, but strays in different ways'.
* These three have been explained in para 5 1 .

105
effect, means the same thing. The two words are
synonymous and one is not the definition of the other. A
general definition of this fallacy is ntTO&nsRraTTclthe fact of
being anyone out of its different varieties W etc. The
author recognises only two sub-divisions of this |c3TOT9
$WT*ar and gfraron. If a | g does not reside in the w the
fallacy becomes S^q&f^ and so the word q^ris not necessary
in the definition of crarca?. Because it resides both in ^^
and fijq^i it gives rise to two contradictory cognitions as | g -TOiHfttn^: and tg^esTORSsn^: and hence no valid s*nfe can
be had. It differs from fk^% in that the latter resides only in
fww and gives rise to the notion ^^HT^T^TT^^T^:. In other
words in k^ the s^fr is contrary while in $TOH;oi33rfTOrc it
is only doubtful or imperfect. Other examples of this variety
are common due to hasty unfounded generalisations, e.g.
The second variety aSTW occurs when the | g does not
exist in both ^q^r and fq^y. Its absence in f%qf is of course
what is needed, but its inconstancy comes from its not occupying its proper place. In other words ^n^Roi is so wide that
it trespasses into forbidden area fq^j1, while the other, its
opposite, 33TTW:, is so narrow that it does not occupy even
its allotted space n^?. So both are equally faulty. Another
example of this kind is ^s^t faw 5F5^3T^ where the |
does not exist in ftq or wkm things, but only in the q
There is another variety of anr^TNsp called argqsTf^ or
^ mentioned by almost all later writers. From the way
r speaks of this t^mtr, it is evident that this threefold division is n accepted fact before his timescf. q^T

eraRm$eFTO^TS3q^r?: srqfte %f s $ ^ ( vide s*nsrqfts%

under N. S. 1-2-6 ). The word Q g q ^ T ^ means one without
a conclusion and so the t^wraTft can be translated as "the
inconclusive strayer.' This has been explained as follows in

the aspsag^t :"^g^rmq^itsgq^f xft \

fc

I." The idea behind this argument

106

TARKABHS ( w

is that when 'all things'fare included under q% there is no

place where the co-existence of the | g with nss can be pointed
out positively or negatively and so no s^rifncan be ascertained*
This is the view of the older logicians. The modern school
does not agree to this totally. They contend that in a %^*srfrtg like tf^rrqsiitaHfTq,, a reasonably valid zmB can be
had by taking particular cases where the coexistence has been
observed, as ^ siiN&r snVrta^ 3n ^X *fc etc. Till a
contradictory instance is pointed out the conclusion drawn
can also be reasonably valid. They therefore define Tgqcrtnfi^
s ^q?mW3rfHtfrcn?5Tf; and hold that this prevents the
grasp of e*rfc9n% ( Vide S. M. under Kar. 71 ).
The example given in the text for H^T^ is
that for 3TWfrfR^ quoted above is ^4*
The remark of our author under rqrTW *s
Ig
T^fRl' would apply equally to egtfR^ as well. In
what respect then does the latter differ from the former ?
The advocates of ^TOinf^ argue that a certainty or uncertainty ascertained from particular individuals comprising:
the whole may not be true when applied to the whole as such,
there being an element of doubt still lurking behind. Those
who do not subscribe to this view like %^r&$T would include
one or other of the above two varieties of
according as emphasis is laid on fifq^H^ or
If we look more closely at the principle behind this sub
division of STTRnfci^ there appears to be some justification for
accepting srgWff^as a third variety. It has been mentioned
earlier that the absence of one or more of the essential conditionshere HWS^ and fwq^f woull xnake the $g
fallacious. This can happen only in three ways :( 1 ) by
the absence of faR^TSK! when the %| is named UWSG; ( 2 ) by
the absence of ^q^f^g when, it is named euraran; and ( 3 ) by
the absence of both when it is called
It should be noticed that the preventive factor in
lies in the imperfect r doubtful ssnf^r between the proposed

107
Ig and JSS and not in the absence of ?q^ and fpjnj instances^
Which only provides a convenient means of detecting the
fallacy. This can be understood clearly by a close look
the following examples where examples of i^OT^f and %
are given side by side of the three kinds of

II

III

iv

m*
The distinction between szrfi'gfTT and
important since1 the preventive factor in both is due to the
absence of s?nfg. TRT^R is what prevents sq\f directly while
n^l'wffr^ only indicates the absence of ssrrf. In the
syllogism q ^ ^ % m ^ 5 T ^ ^ , the %g is too wide to contain
within its scope the nr and so the S?IT% sf^ q ^ ^ c?^ ^ % ^
is immediately seen to be invalid. In ^ e r ^ ^ i 9 , till a place
like ST^fat^ is found, it is impossible to say whether fire has
^TB with smoke or not. In the syllogism ^Pw. ^src^ also,
the sqi% can neither be asserted nor denied and therefore
remains in doubt. In ssyf-^ there is the certainty that **}f&
is invalid while in ^I^^lff" there is the uncertainty that
^1%is not valid. So the former is stronger and more palpable of the two. In this connection the notes under s^rrccrcgfife^
may be read again, where it has been made out that ^T^ is
only an indicator of s^i

[56]

108

TARKABHS (

'TranslationThs incongruous reason is that which is opposed by another reason which proves the existence of the opposite
of the proposed probandum. For example sound is non-eternal
as it is devoid of all qualities of eternal things. This is opposed
by 'Sound is eternal because it is devoid of all qualities of
eternal things? This is also called Rufrq^sone having an
adversary.
Notessaturo;! is explained by Gautama as W ^ sRapnft*en *r f%afcTs?*Hf^g: THROOT (N-S-1-2-48 -). srew here means
the argument in question and its consideration ( fpg ) raises
an expectation l FPigr ) which requires to be fulfilled by
valid conclusion. But another |; is put forward which instead of fulfilling its purpose leaves the gsgso where it was at
the beginning. This is a roundabout way of explaining the
fallacy. Later authors have therefore adopted a simpler word
for it, ^2jf%q^r = one having an adversary, i.e., adversely.
The former word emphasises on the inconclusiveness arising
from the fallacy, while the latter term aims at the mutually
preventive nature of the two opposing reasons. It must be
noted that the two opposing reasons should be of equal weight
as otherwise the weightier one would prevail over the weaker
and then there will be no &tfTOcTT3<*3*fcraf3 siwfcPre?3
^Tsg^rtftf fnw i In this fallacy the two reasons give rise to
two opposing qwpfl's of equal weight, which do not lead to a
valid conclusion either way,

[57] 1

TranslationA stultified reason occurs when the opposite of

the proposed probandum is known to exist in the subject by
any other (more trust-worthy) meaos of cognition. It is
called ^nfffii as well as ^^Tc^^Tqf^S. Take for example the
syllogism,'fire is not hot, because, it is a product like water*.

Here the probandum proposed to be proved by the reason *being

a product3 is 'absence of heat'. But its opposite 'presence of
heat' in the subject is already ascertained through perception,,
as heat is experienced through tactile perception.
Thus has been explained inference.
NotesThe term sf^WtfTO^H means a | g put forward
after the lapse of proper time; i.e. after the negation of the
probandum has been ascertained by a more trust-worthy proof
as shown in the example given. Because the ^g proposed
stultifies the knowledge that the hearer has in the form q^r:
*nttn*n39Fl, it is named also sfera. In ^srfw?^, because of
the two opposing syllogism, one is always left in doubt as to
whether the %tm or the ^^WT^ exists in the w . In ^fRT one
is sure of the existence of the *nWTTW through other means.
Even then it can be argued that there is little justification for
being classified separately. In an example like 3 %?<!*&[:
Iff' is found to exist both in ^<% aod fijq^ and it
can come under H-PTwqfT^Tft^. Suppose the | g in the
above syllogism is changed to ' srcKsrrat^ Still the fallacy m%
exists. But the | does not exist in the q^r and it becomes a
Such over-lapping is common in a majority of
|
and the classification is based on the preventive factors
as has been explained earlier ( vide notes under para 52 ) , mw
is the direct result of not fulfilling the fourth condition of a.
$
( WTJaf^ftwi ) mentioned in para 48 and prevents
directly irrespective of the | g proposed. Secondly in
a syllogism like 3sqfe$rasto]E|t *T*W^ sfMtanTCL > tle WTH is
quite valid and there is no question of the | g having any of the
other fallacies. It has, however, been established for other
reasons that at the moment of origin ( ^qf^fst^at ) all
products emerge without any rpr and so there is no smell in
the pot at that moment. In such cases no fallacy other than
^m could be noticed.
Before taking leave of | ^ [ 9 ' S it is necessary to understand their difference from the other kinds of defects which
Gautama has classied under ^ , snfi and ffmT (vide
N.S.L-4-l-quoted on p. 4 ) Keava Msra also speaks about
these at the end of this book; but we shall refer to them here

 10

TARKABHS ( &

itself to bring out their contrast with ^qnWTft. ^ is quibbling

or equivocation. If, on seeing a person or thing, someone Were
to explain orft W , and if the. hearer responds ?rf|, ?rf|,
anwc ^ti's it is a case of ^35, as it is evidently a perverse
misinterpretation of the speaker's intention. Such instances
occur with double entendre words or ambiguous usages where
it is possible for the hearer to pervert the speaker's meaning,,
An example of the latter occurs when the hearer replies r:
*8r q WSfr for the remark it rofr. If the speaker were to
express himself more clearly as ^ts 3 ^ tsrf this reply would
not be possible. Such instances arise out of imperfect
or improper usages and have nothing in common with
Futile or irrelevant rejoinder is called sufr or snsgTR. One
proposes a statement like sissfrsfw fcfsR^l^ srag^, and if the
opponent, instead of trying to examine it for its validity or
otherwise, puts up a counter-proposition ^ 4 s r e ^ srso[t ^r
rwtsfq m\n,, it becomes 5ntga?. This arises by perversely
shifting the significance of the example pot. Gautama has
enumerated 2\ varieties of snfr based on similarity or dissimilarity with the example.
fwWR means vulnerable point and arises out of misunderstanding or non-understanding the intention of the speaker.
In a proposition like ^s^tsfqw ^33R3T3 the opponent points
out the fallacy w r . If, instead of trying to defend his position, the speaker were to say q%$ SF3ts*g qf : it amounts to the
vulnerable point srfQITfr. Such shifting of one's position
amounts to inability to defend one's stand in an argument.
The stock example for 'shift of accent' seen in Western books
on Logic is an instance of the fst^R called smfec^. If, in
the Biblical injunction "Thou shalt not bear false witness
*againsf thy neighbour5', the emphasis is stressed on 'agains?
it may mean "Thou shalt therefore bear false witness for thy
neighbour'. But if the emphasis is shifted to 'thy neighbour'
it may mean *Thou shalt bear false witness for others."8 Both
^35 and -frotrora. indicate the inability of the speaker to
defend his position in an argument. | ^ m H can become a
f
as auch as it points out the defects in the

opponent's arguments and thus causes his defeat. But these

three, ^ etc, can have no place in the search for the validity
or otherwise of a proposition. That is why Gautama mentions
il^w? as a separate category before ^ in his first Sutra
and again mentions it while enumerating 22 varieties of
<N.S. V-2-1).
A brief look at fallacies as expounded in Western logic
will not be out of place here, in as much as we have referred
to their method of syllogistic reasoning under wrw... Firstly,
the difference in the structure of a syllogism in the two systems
should be clearly borne in mind. Secondly, the word'fallacy'
in their system denotes defects and errors in all the three
terms, major, minor and middle, as also those arising out of
the arrangement of the premises, reduction from one mood to
another, etc. Thirdly, the multiplicity of moods and figures,
the rules for reduction and conversion, the distribution of
terms both in quality and quantity, the various theories about
predication, and similar things cratela good amount of coniusion with the result that detection of fallacies is pretty
difficult. Fourthly, there is no uniformity or unanimity in the
classification and elucidation of fallacies, though they follow,
directly or indirectly, the lead given by Aristotle. He divided
them as fallacies in dictione and fallacies extra dictionem; i.e,;
fallacies of speech and those outside speech. He included
six varieties under the first and seven under the second head.
This classification was chiefly aimed at the Sophists of his
times, a set of philosophic thinkers who were supposed to be
the pioneers of advanced learning and education in Ancient
Greece. In fact these fallacies were often named Sophisms by
Aristotle himself. Notable writers on Western logic, while
expressing their dissatisfaction about Aristotle's classification,
treat the subject each in his own way. Mill gives a five-fold
classification such as fallacies of ( 1 ) simple inspection, ( 2 )
observation, ( 3 ) generalisation, ( 4 ) ratiocination and ( 5 )
confusion. Whately divides them firstly into logical and nonlogical fallacies and the former again.into purely logical and
semi-logical .ones,: thus giving in effect three varieties. Later

112

TARKABHAS (

authors criticise all these classifications, but have nothing

better to offer. Weitem thinks "a rigidly ^scientific classification of fallacies" as just not possible. De Morgan goes to
the extreme when he says : "There is no such thing as a classification of the ways in which man may arrive at an error'5.*
Gautama, who might even be a contemporary of Aristotle,
has narrowed down the circle of fallacies in the Nyya system
to a single head t^ftTO by making the whole structure of
srgfflR hinge round the | g . The result is that a fallacy, in
whatever part of the syllogism it may occur, can be traced to
an imperfect or erroneous use of the i|g in the 'nfe. You
have only to state the ssnfij in the standard formsna f i p ^
?cnr: or TCO3^3*3knar93k~-and then examine the material
truth of the statement It is here that the %^i^ helps and
hence its importance. The un-limited resources of the Sanskrit
language have enabled the Naiyyikas to avoid all negative
moods and.figures. You have only to add the word wctm or
its equivalent .affix at the proper place to make the meaning
negative, and still keep the form positive. This has helped
among other things ( 1 ) to avoid the difficult process of obverr
sion etc.; ( 2 ) to keep all their syllogisms to one single mood
Barbara; ( 3 ) to ward off fallacies due to distribution or nondistribution of terms and such latent errors as denying the
antecedent, affirming the consequent etc.; ( 4 ) to narrow
down formal fallacies to the single instance of 'ambiguous
middle3; and ( 5 ) to reduce all material fallacies to a *%?$NR
or 3ifrf[. "That this common featureviz. vitiating the probative value of a probansis the fundamental basis of
Gautama's classification of defective reasons is implicitly
conveyed by the significant name of f^OTS used by Gautama.
* The remarks in this, and the next para are meant only to
bring out the contrast between the two systems and not with any
idea of showing the superiority of one over the other. Both the
systems have their merits and defects and still serve well the purpose which the founders of the two systems had in view.

It may-be noted here that the same philosophic instinct, that

helped the Nyya theories of inference and syllogism over the
fmalistic barriers which Western Logic still finds it difficult:
to surmount, has also made it possible for the Nyya system
to equip itself with a really helpful scheme of defective probans,
hinging on the concept of | g which forms the main ground
of syllogistic reasoning." ( A Primer of Indian Logic by Prof.
S. Kuppuswam astr, pp. 240-41 ).
Let us now examine a few examples of the more important
;
fallacies in th:? Western, system.
,
I. First and foremost is the 'undistributed middle' which
arises from a breach of one of the most important of their
general rules; viz. the middle term should be distributed at
least once in either of the premises. The middle term, being,
the connecting link between the other two, will not join up
the major term with the minor if it is not distributed at least
once.* In other words the | g will not become ^^r^l^
when it will become ssrfsref^and in many cases it will
become T^mW^when it turns as
Ex, (I) All virtuous men are happy; (Major premise, A )
All rich men are happy;
(Minor ..;..'V. A )
.\ All rich men are virtuous. ( Conclusion, A )
This is Barbara mood of first figure. Both the premises being
A propositions do not distribute the predicate 'are happy\
which is the middle term here. So the conclusion is wrong.
Such mistakes are quite common even in ordinary talk due to a
tendency to regard r the premises as mathematical equations.
In statements like 'the rich are happy5 or 'the poor are not
happy', the predicate does not equal its subject, but is only
affirmed or denied of the subjet. 'Sappy men' include various
types of men, the virtuous, the rich* the poor, the care-free*
the contented etc. The two premises state two such types of
* A term is said to be distributed when the whole of it is
covered by the predication; and if only a part of it is covered
by the predication it is said to be undistributed. The reader is
advised to keep any elementary text-book on Western logic ( deductive ) for reference when going through these portions.

114

TARKABHAS (

men without any connection between :tfaem. A and B may!

both: be happy but that fact does not warrant-the conclusion
t h a t : A is B, as it is well known that many rich men are not
virtuous and many virtuous men are not rich.
-.-:
Translated and re-arranged Q the Nyya
syllogism becomes,
^

form

this

( srfcT^T = Conclusion )
( | = Minor premise )
( sr^f *<? = Major premise )*

In the iri% we at once see that the ftsi and | g have

interchanged places with the result that the | g is stated to be
^PfcfssnqsF. So it becomes Bl^TT^ or : c&qrflro s there are causes
other than happiness for virtousnss.
[ ;. Ex. ( 2 ) All that glitters is not gold; ( O ) ...
. ;:) H;
Glass glitters;
(A )
il. j , /.Glass is not Gold, v - : (O )
:
Here the | g , 'that Which glitters' is undistributed in both the
so;the <?oncl^^onisn;Qt;! correct." >
"Translatedftboines,

The l g is easily seen to-be ^TWT^r as aiflsr^ exists-in

)i
( not-gold-things ) as well as in fq^r ( gold )i
* 'Let,us take an exarpile ^ch of Htqjl^ and TROT from
the Nyya system and,jurn it intp the Western system.

Ex.{ 3

* In this and in the examples below a corroborative instance such as

W* and the other two members ^ q ^ and flR can easily be
supplied by the reader.
;; : ,
'?.-

J15

,. In English formKilling causes sin;

Sacrifice involves killing;
/. Sacrifice causes sin.
If in the above, the subject means call kinds of killing' as is
the intention^ it leads to the defect of "undue assumption5, as
an exception is made in the case of killing in sacrifice. But
if it is taken to mean 'some kind of killing' as is not warranted,
this becomes a case of 'undistributed middle'; and then the
conclusion should be particular.
Ex. ( 4 ) oraronqsfcft 3%trr, S W ^ W L ; *ra sr&Rqr 3ra
3%t??3q, becomes in English
" All cognisable things are fiery:
The hill is cognisable;
.*. The hill is fiery.
Here the major premise is wrong as it is a case of 'undue
assumption'.
II. Next in order of importance is the 'illicit process' which
may be of the major or of the minor term. This happens when
the second general rulethat no term undistributed in its
premise may be distributed in the conclusionis violated.
Ex. ( 5 ) Illicit majorAll cows are quadrupeds;
Dogs are not cows;
.% Dogs are not quadrupeds.
Here the major term 'quadrupeds' is undistributed in the
major premise, but is distributed in the conclusion, which therefore becomes wrong. In Sanskrit it would read
Here the s?n% ought to have
been stated as qg iftfrarai srsr ^g*<nfewsr which is a etffftf^.
The spcR form of this should read as ^^i ^pjir gf^ lh and
mot as stated in the syllogism. Either way the | g is easily
seen to be
Ex. ( 6 ) Illicit minorNo man is perfect;
All men are animals;
.". No animal is perfect.
Here the minor term 'animal' is undistributed in the minor
premise, but distributed in the conclusion. So the conclusion,

116

TARKABHAS

though a factual statement, does not follow from the premises.

In the Nyya form it would read mfwrr raVjais'r lfa^^ \
*ra snfr cT^i ^ c i g ^ ^ ^ ^ . The ^g is one which exists only
in the q^ and as such becomes
Ex. ( 7 ) Taking an example
and expressing it in English it would read : All things that have soundness are eternal;
Sound is a thing which has soundness;
.\ Sound is eternal.
Here the syllogism is technically correct in form. But the
fault lies in the meaning. The minor premise does not predicate anything and amounts to saying 'Sound is sound*. Thus
the major premise and the conclusion become identical and
this is no proper syllogism. Also sound becomes distributed
in the conclusion without being distributed in its premise.
III. We shall illustrate in brief a few examples of the third
important fallacy, 'the ambiguous middle', under which a
variety of defects will be found included in the text-books.
Ex. ( 8 ) Light is pleasing;
His loss is light;
.' His loss is pleasing.

Sanskrit form

Here'light'is used in two senses in the two premises and

it is a 'fallacy of equivocation'. In Nyya it will be a case of
^ and not ^^TOTO. In the Sanskrit form the ambiguity has*
been removed, but still the | g is :
Ex. ( 9 ) Begging is punishable;
Owners of charitable
institutions beg;
.". They are punishable.
Here 'begging' denotes two different types of asking in the
two premises, and this is an instance of the 'fallacy
of accident' In Nyya it is, at best 5n?gTi?. But in the
Sanskrit form of the syllogism the %g is
being the !

117

Ex. ( 10 ) The angles of a triangle are two right angles;

A is an angle of a triangle;
\ A is two right angles.
Sanskrit formot

Here the middle term, 'angles of a triangle', is taken collectively in the major premise and distributively in the minor
premise. The word 'together' after triangle in the major premise would no doubt make the statement correct; but then the
minor premise and the conclusion do not arise. This is called
a 'fallacy of division'. In the Sanskrit form, the %g is sFTftf
as we already know that one angle of a triangle is always less
than two right angles.
Ex. ( 11 ) Five and six are odd
and even;
Five and six are
eleven;
.*. Eleven is odd and
even.
In English it is called the 'fallacy of composition' and
appears to be a riddle. Here also the middle term is used
distributively and collectively in the two premises. In both
systems it could be an instance of
Ex. ( 12 ) God created man;
Man created sin;
.\ God created sin.
Here there are actually four terms; God, Creator of man,
man, and creator of sin and a syllogism can have only three
terms. This is called the 'fallacy of four terms' or the 'logical
quadruped'. In the Sanskrit version the portion cGod created
man' is missing, as here also there can be only three terms. As
it is all the five t?3T*TO's can be traced in it.
The above examples will show that ( 1 ) there is no correspondence between the fallacies in the two systems; ( 2 ) the
fallacies in the Western system vary according to the mode

118

TARKABHSA (

of reduction; ( 3 ) it is easier to convert the Nyya syllogism to

English than the other way as the former is more strict and
circumscribed; ( 4 ) a majority of the so-called 'material9
fallacies in the Western system centre round ssrf^f^, the most
common fallacy in the Nyya system; and ( 5 ) it is easier to
point out the fallacy in the Nyya system by expressing the
in the standard form,
.
The whole of STfSR has been summarised by fviro in his
^ ( V. 66-78 ) in 12 small easy verses so well that it
deserves quoting in full and remembering by every interested
student.

n
\
*w
i

n
\

( End of Anumna )

list

COMPARISQN

TransJatiQn~KnovfldgQ through comparison is that which,

is gained by the similarity of one thing to a known thing like
a cow, when this similarity is aided by the recollection of an
assertion made byy some knowingg person
to that j ejtfct. ^b
p

example an urban man who does not know what thfefe Word
*T%% denotes, hears from a forester that <;*TR (-GayalBs
Gaveas ) is ' an animal similar to a co\v'. Afterwards he
happens to go to a forest where he see some animal similar
to the cow. ' jif at that time he recollects the words of the
forester, he gains the knowledge that the animal in ffont of
him is what is called im%9 which is similar to 'the cow. This
knowledge is ^qi and it is the instrument of the -cognition
sqffr. After knowing that the animal in front is similar
to the cow, the cognition that this animal is to be called 3T^T?
is OTf; which is the grasp of the relation between the word
%1%W and the thing denoted by it (or in other words the
denotative significance of the word ^qfcNn^f^H^^^fT! ),
This is the ^suit ( of 3qw?r ) and as it is not ascertained by
the other sraw's, perception and inference, ^q*rm is a
separate simm.
(Thus comparisin has been explained.)
Noteszmm*l is the knowledge that the thing ( or animal )
in front has similarity to the cow. This similarity was made
known by a third person as the forester. On seeing the
similarity ( 9re?39H )> the words of the third person come

120

TARKABHAS ( a &

to mind ( T ^ T 5 | T ^ W ^ O T ). t. The former is the ^m and the

latter the mediate activity CpmR ). This is the view of the
moderners. The ancients would say that srf^cnwm^H
is the mm and s n s ^ g m is the e ^ T R or ssnq^. Kesava
Miara prefers the former view, but he does not go all the way
With the moderners as he distinctly says that 3qff is of the
form 3T*ft (fq<*g!) T ^ q ^ ^ i - T h e moderners say that ^qfafi
is of the form ^ TOraqq^rra: . The difference between the
two is that in the former case only the particular animal in
front is denoted by the word T T ^ ; while with' the moderners
the whole class of animals, of which the one in front is an
individual, is denoted by the word TT^^T. Either way sjft^l*
WPTOW'T and WSCTS^nf are both necessary for sqffr, which
is ascertaining the denotative significance of a word^^feSJ*3PErg"R or ^fific-( for which vide the next section sis^ ).
Just as ^qffr arises out of 'rcg^SFTO it can arise through a
knowledge of dissimilarity ( = 1^?!^^ ), as weir as through a
knowledge of some uncommon, or peculiar characteristics
( = w e i ^ ^ ^ l ^ ). An example of the former kind is ^ | t
W R ^ . e^m^i^^^li ; !T # .--a camel does not have a level back
and a short neck like a horse; arid of the latter ^ i ^ t fijfe^Sfl^^W : a rhinoceros has a single horn protruding from
its nose; Viewed in this way ^q^R can be of three kinds
( vide f^F^r^t^Er a commentary on T. S. of An. Bh ).
The Viesikas, Smkhyas and Buddhists do not accept gqfnsr
as a separate vmm, but include it ander ^cgHlT. The Naiyyika's reply is typical and is su;mmarised by Nilakantha as.
followsl^fq^i^g q^r^^c^^n^sceT

\ The inference by which

they include sqsR under inference would be something like
Even if this were valid, experience shows that the denotative
import of a word is grasped without the aid of such a s^i%.
WSSTOI in his comments on N. S. 11-1-44-48 points put the
practical utility of knowing what certain technical names
of .medicinal herbs in Ayurveda
denote
from the
directions of an authoritative person or book (i.e. through

121
), On the other hand the fact that the knowledge
arising out of similarity can only be tentative and not positive, may still stand against its acceptance as a separate RHIT.
g^zRraro discusses all this in his ?^M>f5iTTife and states his
conclusion thus :

Our author has also stated the same idea in the last
sentence of the text above.

WORD OR VERBAL TESTIMONY

fi
I

>r :

TranslationThe utterance of a trust worthy person (

constitutes verbal testimony.. ^1B is one who speaks about a
thing as it really is and ^TSRT ( sentence ) is a collection of
words which have the three characteristics ( at the same time ):
( 1 ) aurafat = verbal expectancy; ( 2 ) sfosrm = congruity and
( 3 ) ^[f^% = proximity. For this reason a collection of words
lik^bw;, horse, man, elephant' is not a sentence as it is
wanting in verbal expectancy among them. Similarly 'spray
with fire' is not a ( meaningful ) sentence as the two words

122

TARKABHS ( *

are lacking in congruity/ Here 'fire' and 'spraying5 are not

capable of being construed together, because the instrumental
case in srfosTT denotes that fire-is the instrument for the act
of spraying and fire is'not "capable of becoming such an instrument in as much as 'fire1 and the act of spraying' do not
join together through caiise-effe.ct relation. Therefore the
words ^fewi f^do not constitute a , ^meaningful) sentence.
Similarly if the words Tn^.and w ^ r are uttered one by one
with an interval of an hour between them and pot together,
the two will not become a sentence. Even t&ough there exist
sn^i^T and '%smT between them the two words'lack proximity.
Therefore a sn^R or sentence is only such a collection of wo^ds
which have all the three chiaractristics ( t the same time),
desirous of .attairiipg heaven.. sha^ld perform, T the sacrifice
ssitfeetes. ^#^f| q'^lw g?cfthere .are five fruits on the
bank of'the river;are_( ( correct ) sentences.1 Sinlrly Ltlie
words TnT^r become a real sentence if they are uttered
without ( undue ) delay-between*h two.
MotesIn ^ns^sjrj = cognition from verbal testimony, knowledge of the words .( *R5Fi) is the ^OT, recolectio of the
meaning of the words ( q ^ m U ) is the tmqw, knowledge of
the denotative force of words ( ^rffK^ ) is an accessory and
verbal comprehension ) is the;result, cf.

( E^T?. V-81 )

Keava Miara begins with the definition of the cj^or as

and then goes on with the explanation of the words
and m$q. These definitions involving ^yi^^ ( statement of the reality or truth ) are significant as they indicate
that to the Naiyayika the test of validity of ^is^^t^ lies not in
the authority of the speaker but in the fact of his words being
based on truth.
A sentence consists of meaningful words, but these words
become meaningful only when they fulfil all the three prerequisites required of them, OT^H, rtrm and sfrf qr

123
Though these have not been defined as such the author has
given sufficient explanation and examples to make clear what
these words stand for. When a word or a set of words is not
capable of producing a cogent meaning for want of another
word or words there occurs 3H3ngfTT3TO r^asramg^T?^Ts^5^T^3sn35|fr. When someone utters ^ ^ r and then
keeps quiet, the hearer expects to know what further is predicated about itwhether sandal is to be brought ( m*m ),
whether it is to be removed ( w<om?f ) , whether it is fragrant
( ?f*7 ), or cooling ( sftar ) etc. This expectancy is what is
meant here by WSiffi. By simply saying nfc,9W;,'5qr:, ^ f t etc.
the words do not convey a complete or cogent meaning and the
hearer expects to know 'what about these?' If some'such
words as 'air WifNf or gU^TOifr are added this expectancy
is calmed and the verbal comprehension ( 3TTS^^ ) is complete and valid.
The second requisite ^ R H is variously translated as congruity, compatibility, coherence. It occurs when the meanings
of two words are not contradictory or incongruous. If one
were to say 'sprinkle with fire' the hearer's knowledge that
fire is incompatible with the act of sprinkling prevents him
from forming sus^sftqr as he would do if he were told'sprinkle
with water'. Other instances are ^^ 3ir, fi ffafr, a
^^fa1, RfjiafafcT etc. Such usages may be common in sf^rcW ^ where the meanings would be quite different, and so they
are out of consideration here.
The third requisite ciff also called qrrafc, is defined as
q5RTOflr5?^tafa*4'the fact that the words are uttered
without undue interval of time between any two. Such delayed
utterances prevent the hearer from connecting together the
meanings of the words and thus prevent him from getting a
verbal cognition. It is common experience that the meaning
of verses arises only after the syntactic order of the words
is re-arranged so as to have the proper
Some authors add a fourth requisite, c U ^ ^ r a or knowledge
of the intention of the speaker. This is necessary especially
with double entendre words or where the implied sense has

124

TARKABHSA ( w

more importance. An instance of the first type is when someone orders ^WHST, where the word e^sr is capable of the
two meanings, salt and horse. Which meaning is intended
by the speaker is to be determin by the context in which it
is uttered. If the speaker is taking food when he says it, the
word means salt, or if he is preparing to go out it would mean
horse. Examples of the second type occur when someone
exclaims 3JJ sraiT^ ! or g^Tfi ! meaning thereby 'you are a
coward !' or 'what you said is quite absurd !' according to
the context. Those who do not mention cTTcqq^R as a separate requisite would include it under sT^PT^ or sftKTcH. Either
way the necessity of accepting it as an accessory for susspt^
cannot be outlooked.
It should be remembered here that it is the knowledge of
the existence of these requisites that causes 311539)3- and not
their mere presence as will be evident from the following
example. When somone says simply 51*, the hearer nearby
immediately closes the door if there is a gust of wind, or opens
it if there is a knock, according to thenecessity of the occasion.
Here the word cfq&ff[9 or 'atqif <g% though not actually uttered,
is understood from the context; and it is the knowledge of the
meaning of these words that causes the hearer to have the
proper ?nT*3rt^ and through that the appropriate action.

Translation( An objection ). Even here ( i.e. in sentences

like 3R(iteft$ <ra ^Tfa sfsr) it is not the words that have ST^T^T
but it is their meanings; because the fruits as the supported
things, call for a substratum which is furnished by the bank.

125
In fact, if we think a little more deeply, even the meanings
cannot be said to have au^TlgT, as that being a form of desire
can exist only in a conscious being. ( Answer ). True. When
a man hears certain words and understands their meanings
these latter produce in his mind this srrasiffT, and so being
productive of the snsFTjjfT the meanings themselves are said to
have 3?$l||*. Through them the words that generate those
meanings are said to possess 9iT^T^I by implication ( ^<rarc or
35^r&!T ). Or, the words themselves having denoted their meanings may be said to create in the mind of the hearer the
expectancy of one meaning being construed with another by
implication. Thus the meanings endowed with 3U35T|ff become
capable of being construed together. Through these meanings the words themselves ( by implication ) are said to be
capable of being construed properly.

[61]

TranslationProximity of words occurs when they are

uttered by one and the same person without undue delay in
the intervals between two words. This happens directly in
the words themselves and not through their meanings. Thus
this meaning ( i.e. the detailed definition of a sentence ) is
arrived at. A sentence ( m*w ) is a collection of words which
are such that ( 1 ) they produce in the mind of the hearer,
through the denotation of their meanings, an expectancy
regarding the words themselves being interdependent or
regarding their meanings being interdependent; ( 2 ) they
convey meanings that are capable of being construed without
incompatibility or contradiction; and ( 3 ) they are close
enough ( to produce the desired meaning without undue
delay ).
NotesIn para 60 the author discusses how w$T|fr can be
a characteristic of words which are uttered by the speaker.

126

TARKABHASA ( < ^

is a kind of expectancy or desire arising in the mind of

the hearer. So the objection is how it is stated to exist in the
words of another man or in their meanings. The reply is that
these words create the expectancy in the mind of the hearer;
or better still, the words denote their meanings which in their
turn create this expectancy or desire to know what is being
said next such that its meaning may join properly with that
of the previous word or words. Thus the words or their
meanings as the cause of this expectancy is stated by implicationsrq^uto possess the expectancy, just as we say
zft 3?1?3*. Here sweetness is a quality in the hearer's mind or
ear, but is transferred to the cause of the feeling, the song. The
second requisite sfas^n, it is evident, should exist in the meanings of the words as explained in the previous para. The
third srfrf should exist in the words themselves as mentioned
in para 61. In the final definition of sn^sr the three requisites are
shown in their expanded form by the three adjectives qualifying the word ^ W ? . This shows that Keava Miara does
not want to include ^nwrensf as a separate, requisite
[62] qt ^ ^ E W | [ : '
^ ^

TranslationA word is a collection of letters; and collection here is such that the letters in it constitute a single conception. When a word is uttered the letters in it appear one
by one in a certain order and each letter fades away by the
time the next letter is uttered with the result that there is no
possibility of any of the previous letters remaining ( in the
ear ) at any one moment ( or in other words only one letter
remains in the ear at any single moment. But their impressions
( sfc$TC ) a r e teft behind in the mind of the hearer. As soon
as the last letter is heard, the auditory organ combines it with

 " "' " "

127

the impressions.of all the previous letters and conveys to the

mind the idea of the thing denoted by that collection of letters
even .though all the letters except the last one are not there at
that moment. The T .auditory organ is endowed with this
special faculty of picmg together the impressions of the previous letters with that of the last one to create a single conception from the whole. rTis. is because of the efficiency: of
the accessory causes jast as in a case of recognition.(*K*Tf:9f ).
When recognition occurs, the previous experience though long
past, is conveyed at the moment as though it were part of the
present time.( through the strength of the impression embedded in the mind), The idea of a sentence consisting of a
collection of words is also conveyed by the auditory organ in
a similar manner after combining the denotation of the last
word with the impressions of the several previous words heard
in succession., 1; :
. , ,
n

Notes The discussion following the definition of 'word'

centres round how the several letters in a word are capable of
demoting a single entity. Each letter as it is uttered caves
behind an impression^ itself in the mind of the hearer and
fades away. After the last letter, the mind combines the impressions of the previous letters and this \combination makes him
recollect the knowledge ofthe entity denoted by it, which is
already latent iii:he;mmd of the hearer, .When one utters
words like ft, srsr, wre* ^TCOTfl, etc., only he who has
previously understood the denotative significance of such words
makes out what; the speaker pme^ns. This fact of bringing up
the latent idea of a s?rf^ by all the letters comprising the word
is what is meant here by the phrase q^Rfre?itara. Other
authors express the same idea by defining 'word' as ^pgq^
a word is that which has a denotative force or saf. What
this mfe is will be explained in the.notes further on
The author's main intention in the above discussion appears
to be to refute the theory of tie" grammarians who postulate
Hste as anew thing for generating'the idea of a single entity
on hearing an orderly set of letters. *<st? is a kind of eternal
sound ( ^ff ) which is made explicit by the letters<st35
^ ^ ^
^ 2 : ~ T h i s in its turn creates the

128

TARKABH?&S& (

notion of the entity denoted by that collection of letter s ~ : ^

c^swf^fr site; wsfqlrtt^:J-^3rhe grammarians attach so
much importance to this *O\z that they raise it t the- status of
^^5fr^j which again is equated to q ^ F ^ of the Vedntitis.
This ttz theory f the grammarians however is not accepted
by most of the other schools of thought Our author quotes
the instance of g?*rfvrt? or recollection in support of his argument that the mind processes the impressions of the successive
letters and words to a judgement leading to Sns^rtey. In
srcrfT^rT one recognises another pefs ( or thing ) whom he
has met some time ago, as 'this is that Devadttalats^lf^:.
Here *W represents the past experience recollected through the
impression left behind on seeing him earlier and ar' is the
subject of the present visual perceptions In ordinary talking or
silent reading these mental processes are not-noticed because
of the quickness with which the mind is able to process the
latent. impressions. Tins quickness also accounts for s.ome.
students understanding the teacher more quickly than others.

[63]

[64]
-~. Translation( 63 ) When such a sentence (as defined above)
is uttered by a trust-worthy person it becomes the !jwn called
?!*% ( or verbal testimony ). The result of this sr^nar is under*
standing the meaning of a sentence. This ^TS^GT, is similar
in human utterances as well as in Vedas. There is however
this difference. Among, men only some and not all are trustworthy and only such utterances that come from these trustworthy people are mtm. The Vedas are all uttered by the
greatest wm, lwara or God, and so every sentence in them is
as they all come under the definition

'-S

.-

"

128

(6.4 )< Thus have been explained the four instruments of

valid cognition. Any instrument other than these four is
not accepted as SWTOI ( in wn*rai*5 ). Such others (asare
proposed in other astras ) come under one or other of these
four.
NotesPara 63 gives a two-fld division of ^^ as
( Vedic ) and afte$ ( human ). Under the former are included
%$% and frf's are digests of rules and regulations and
are based on gRren^s and their authority is limited to that
extent. Some would like to include JjnoT and s[fn|TO also
under srf^; but these have comparatively little authority and
are cited only when no higher authority could be pvut forward.
That everything in the Vedas is TOW rests on one's faith. But
here is a tough bone of contention between the sTaiyayikas
and the Mmamsakas. The latter contend that the Vedas are
not produced by anyone ( entier ), but are revelations of
God and so eternal. The several sages mentioned along with
the Rks are only the original 'seers of the hymns'
This is based on their view that sound is eternal,
The Naiyyikas on the other hand hold that sound is non*
eternal as it is a produced thing and that the Vedas are qhifc;
but because of the trancendental wisdom and truth contained
in them they could originate only from an Omniscient Being
who is none other than God. The Advaitins try to reconcile
both these extreme views by saying that the Vedas as handed
down to us are scf as they are copies of what existed in
previous spar's ( cycles of ages )q^rom^TOCTOifc This in
effect means that gfr is eternal only in its meaning ( sro )9
not in its form ( sngpi^f ).
This controversy whether the Vedas are ql^sr or srifc^qr
is not of much consequence. The real difference among the
various schools of thoughtespecially among the Naiyyikas,
the Grammarians and the Mmamsakaslies in their explaining
how words uttered by one person denote the desired meaning
in the hearer. The Naiyyikas accept two significative functions for a ; wbtd' called ^is^f^'s. The first: and primary
9

or STpR^fT ( or srtcT* also sometimes), which

can %e translated as denotative force. The second is 35^O or
implication which is always dependent on the forir. ^ffe
is defined ais the relation between the word and its meaning,
It is an established convention that such
9-nd such a word denotes such and such a thing; as ?ft
^2'f ^f%: etc. The meanings of these words are handed down
to us from time immemorial and because their originators are
not known their ^f is attributed to God's will3^ q^T^Tsfsj fri^r j^ftia\^T q^rfei. Proper names such as ^% ira
etc. also have this $if as they denote particular persons ( or
sqfR's) even though their originators are known. Similar
is the case with the innumerable words that are being coined
in modern times such as ion, proton, ohm, watt, etc.
The ivimamsakas want this safer to fee treated s a separate
over and above the seven accepted by the ftfe's mainly
on the ground that it exists in ^ssr, goi and ^ 4 just like the
latter two abide in ^s^r. The Naiyyikas arge that this will
lead to accepting as many crft's as there are Words ( as also
other 55^'s etc.), as well as several.'trff^' for denoting a single
s^tfff. This cumbrous position ( *ffc^ ) could be avoided by
accepting ^ffl^r as a potent relation by which a word revives
the meaning imposed on it by divine or human will.
Though all schools accept the above two ^rs^ff^i's there
is great divergence among them regarding the locus of Sflcr.
What does a word like ft or ?iZ exactly denote ?Whether it is
( a ) the animal or thing as such ( i.e. s^fcT ); ( b ) the common
notion underlying all animals or things of that class (i.e. mf);
{ c ) both together; or ( d ) something else outside these two ?
These are the four important views on the subject and can
be stated in Sanskrit as follows :( 1 ) Stas&OTot ^fR^s; (2 )
tfcr.; ( 3 ) ^T%f^{^rss^qr^ r&a:* ( 4 ) srtf or
The first views is that of the modern Naiyyikas,
the second that of the Mmamsakas and Grammarians, the
* ^lpr metis convention-It is^ indiscriminately used as a synonym
fc ^f^^ifand* sometimes as its cause ( Vide 35T- 5f. Ch-. II. )

131
third of the older Naiyyikas and the fourth of the Buddhists.
Each of these schools look at the denotative force of a word
from a different angle and each would appear to be correct
in its own way.
When a sentence like SSW is being uttered the intention
is that the particular entity jar ( sres^f^ ) is to be brought
and not the snf ( sr2H$ ) as the capacity for being acted upon
( ^ffsn^ftwf ) exists only in the wrfa* On observing this one
who does not know what a QZ denotes previously assumes that
that word represents such and such a thing. This assumption
is confirmed when next the observer sees the jar being
removed. From this it must be admitted that the denotation
of the word srz is primarily ascertained in the ^fftffft as
otherwise the hearer would not bring or remove the particular
object. But then it may be argued that safgr is grasped only
on one particular s^fi^T and there are innumerable crcosffa's
in which ^ftrf is not grasped. This objection would appear
to be valid on the face of it; but is contrary to ordinary
experience. When one has ascertained sflf^r on a particular
he sizes up the general shape and form of a ^2 as a
and through it he conceives the general notion
underlying all such jars. In fact ^z&t is nothiag but:
shorn of all the elements that constitute the
shape of the object, such as a narrow neck, bulging belly,
rounded bottom, etc. sraK3T is the abstract formless notion
while ^gftef^f is the same notion in concrete form which
is called srfsffMfo It is this snfftfrH which enables one
to distinguish it from other things3rag^T!%. If one has
understood what is meant by the word srs, or in other words,
if one has grasped the denotative force of SB, it means that he
lias a definite conception of ( 1 ) the-3Z*9pMr;-( 2) the general
notion --^sur and ( 3 ) its peculiar shapesnfpfM^Was a
q^gsftcnf^frcg. This is what Gautama wants to convey in
his ssrra^ II-2-64, ''uunsfeiraras qf\wS-~-the meaning of
a word rests on all the three, mf, 3rr$f and
comment of this Sutra is explicit and to the point<g

notions

132

TARKABHASA (

and s^rfft are but different phases of one and the same>
thing serving distinct functions in logical explanation and as
such inseparable from one another.
Later on, as niceties of logical dialectics became more and
more polished, philosophers of the different Schools considered
it more expedient to restrict ^fira to a single locus on grounds
of simplicity ( ^ra? ). The Grammarians emphasize that
the denotative force acts primarily on the snfe as that is ther
life-giving element in any substance, mT^ (vide p. 17 j . Th
conception of the sqfiff is by a later mental process such as
sgns?, smqfr, or even by s^TgRif ( suggestion ).. As the visual
organ is in direct contact only with the strfi and as 'swjifraT-mfcm rests on it, it is only the s^f^par that can be brought when
some one says 3?H?ra. The Mmmsakas, Advaitins and the
Rhetoricians follow the Grammarians in this respect and they
all embellish the latter's arguments with their own, even
though they do not go hand in hand with the Grammarian
regarding his wte theory,
The modern Naiyyikas argue that it is easier on grounds
of &m% to postulate 5feR on the sqfa; mainly on the ground
that there would be no difference in meaning between TfTOH^F
and 3itecT3TT^r if 3f%> is accepted on the sn%. In this case
becomes the ^Rn4 and its delimiting factor,
will be t^^ which becomes definitely inexplicable and so
leads to fhftT* ?ftm itself is fairly imaginary or abstract, but
has to be accepted as a logical necessity and can be equated
to ^^^f^sT^Sf^Tfoi^. Both sides claim the advantage of
tsvsm and both have to resort to' 3TT%^presumption or
inferencefor the notion of scf%ar or sn% as the case may be.
There are however, even among the more modern Naiyyikas^
some who would prefer to go back to the
based on the N. S of Gautama quoted earlier, (cf.
rS.'M. under V. 81 ).
. The Buddhists who want to refute both the Mmmasaka
aod the Naiyyika at the same time strike a new path and
contend that the primary import of words is qt or srg|[tiff%^
i.e. differentia from all things other t}ian that. In othei; words.,

133

primarily denotes a thing different from non-jar

things and the siiff and ssffr are only aids for this differentia.
( I t should be remembe:red, that they do not accept a snfg1 as
explained by the Naiyyikas other than this sfqtf ). This is
in consonance with their negative approach to most of the
philosophical questions. This idea of differentia may sometimes be necessary to clarify the intention of the speaker; e.g.
in a statement like ^ten^, where the speaker wants to
emphasize that the animal in front is'not a buffalo', but an
elephant. Even here to say that gf^r is not on the canimal
elephant' is, evidently contrary to experience; and the notion
'not a buffalo' comes in as an additional factor through
implication dictated by the intention of the speaker.
The difference between denotation and connotation on
which great emphasis is laid by Western logicians, does not
find a place in any of the above views. If we may use these
two words in their most general meaning, ifc may be said in
general that the cftra ( r Slfa ) of the Naiyyikas comprises
both these notions, 3% corresponding to the connoted attribute and softer to the. denoted object as qualified by the 5%.
Viewed from this angle the weight of reason would sway to
siTfirftfeesrf^TO^vide f o r a ' s view quoted in an earlier
para.
, This sf; is of three kinds, *ftiT, ^fk and ^T^sf. sfar is
or etymology. A *ftfh$3l53[ is one which conveys
only that meaning which its component parts-usually a
root with particular affixesimpart to it. Examples are:-
ps$3, which is composed of &fk9 a preposition,+ (root =
t o g o ) + g ( a 3Scrrc?!3?*r ), means one who learns; sissnqsfs
+ foi^+^ffsrc^^r ) means one who teaches, ^:f is
i.e. where the denotation comes from convention
and not from its parts; e.g. sB = a jar and not ^z^ %fk nz: from
etymology; TSf? = an elephant and not sracftfi *izr- il?r^^
words are those which retain the etymological meaning bus
restricted by convention to a particular thing within that
meaning;; e.g. : ^5 = one that grows in or from mud and in
that sens it mayniean a lotus, lily, frog, weed or any such
r but it is restricted by convention to a lotus only.
SOT^RS

134

TARKABHS (

(=bee) %r^( = elephant) etc, are other examples. Ordinarily

one should not use such words in their etymological sense as
the general maxim is rfrn^j^fcfcrcft. Some authors have a
fourth classification also called tftfrr^S' words; i.e. those
which can be interpreted both by-ei^zmif; o r ^S^TO^f independently. Such are w m j ^ o n e w ^ich emerges by itself
as well as God Brahma; sg;faa[ = one which sprouts as well as
a tree or shrub. In their case convention has not grown to
such an extent as to restrict the sense to one particular thing.
All these different kinds of words require 'SfeRf on the
P or the ^ g ^ q or on both as the case may be. This
happens in various ways, which are thus enumerated!
in an ancient verse.
^

II

is ascertained in the following ways :

( 1 ) From
3*n^r = Grammar, which gives the meanings of the roots, the root endings
and terminations, the various cases,
prefixes and suffixes, compounds,,
and similar things; e.g. *T*3[ = to go;
^f^ = goes ( present tense ); n*3
= one who goes; fr = gait; ^m
with an = to come; ercftarrei = one
who has learnt the atras; etc.
( 2 ) From
3q*rH = the instrument of comparison, vide
para 58 of the text and the ntese
thereunder.
( 3 ) From
^)^T = dictionary like the mrosfcr which
gives synonyms as well as different
meanings for the same word.
( 4 ) From
scTraf3 = the statement of a trustworthy
person as ssstwq^^
( 5 ) From
a^^R = common usage of elders; e.g. when
one man says TURFTO and after the
cow has been brought ^prcnTO,
and after grass also has been
brought, another man who observes

135

( 6 ) From

( 7 ) From

these-< actions determinesth denotation of *t and ^ur.

rre sreta = context, as in the Vedic sentence
*snrcw#tf* where the denota*
tio of ^n is grasped from the
words that follow
fffci

= explanation

or commentary

as

( 8 ) From
contiguity with known word or
words, as in %$ *r?3TC^flW?>5f
the words *fCT7re and
whose meanings are known
determine that of fq^F a the
cuckoo.
To this list Rhetoricians add gestures and signs by hands
eyes, etc. ( vide SRT. *r. Ch. II ).
The second srs^crfe is ^m=implication.
It is defined as
^WW^^fs which means relation with the denotative sense of
the word. This is resorted to when the primary meaning
is unsuitable or incompatible. In a sentence like
the primary sense of the word TTr is the stream or
current of water and it is well known that a hut ( ^f ) cannot
stand in a stream. So to make the meaning compatible the
word TT is taken as signifying its banks contiguous with the
stream. The older logicians say that sr?gmgqqfa ( = discrepancy in the direct construing ) is the condition for resorting to 3S^OTT. But this is found to fail in expressions like 5jfl%fr
^f ^^913? where there is no discrepancy in w w . The
intention of the speaker of such a sentence is clearly to protect
curds from *all birds that eat it away' and not merely from
crows alone. So the later Naiyyikas stipulate ^
( = discrepancy in the intention of the speaker ) as the necessary
condition for resorting to ^ o n .
This ^ o f t is also of three kindsIf <*3iqf, mq^MPlf and
In gH1*it-ihe primary sense
4

TARKABHS (
gives place to another meaning ( 'connected with the
as in ifarai sto: or in wmv- gstsrf^. In the latter example the
5ftqT*i of *s _( = a cot) is given up as it is not capable of crying
and the word is taken to imply the children lying in those cots.
fe
is resorted to when a word while retaining its
implies something more than that, but connected to it
in someform; e.g. spsst ?frT^mr or %famt mfor. In the
latter case; those who hold umbrellas and those who do not
are both impliedby the word ^fo^as they all go in a crowd.
fe
happens in cases where some part of the
is retained and the other part is discarded ( 3ffq^s discarding ancl S T ^ ^ = retaining ). Examples of this type are
STte^Ibi:, <r?^ir etc. In the former the pronoun sn denotes
the man who was seen in the past time and STST denotes the man
in front at the present time. To maintain the identity of the
two the adjectival position 'seen in the past time' is discarded
and the substantive 'man' alone is retained. Similarly in the
second example g1^ denotes fiijpra^ and *# the ^pisite. To
identify between the two these words give up their attributesfaaTJTand ^;pi as they are opposed to each other. Then
We et the intended^T^t^, sfrsft WinfF^.
... These subdivisions are surpmed up in the following table
as an aid to memory, the examples are enclosed in brackets
under each.
-

etc.)

etc.)

etc.)

.-;/v^

337

< The Grammarians and following them the Rhetoricia&s

accept a third-51333% called, SG^^T or suggestion. This has
mainly two varieties; ( 1 ) ^ns$ = that arising directly, from the
word itself; and ( 2 ) sniffsthat arising from the meaning.
The former occurs in such sentences like J^sfe:, The purpose
of saying in an implied manner is to suggest the intelligence
etc. of the teacher in the pupil, which will not be so emphatically brought out if it were to be expressed in a direct manner
like g^g^roffe:. snsff ^^f^fT occurs when one exclaims
2&J *T3T^ ! on seeing him running away. Here the suggestion
is that he is a great cowardjust the opposite of the word ^jj.
Another instance is Wfqf 5F?? cft^T ^TT^^ ftt STWf^ \ Here the
suggested sense is 'after your departure I shall not live' and
shows the intensity of grief of the speaker. These schools
argue that such suggested meanings cannot come through
ftrer or 3$^<m and so a third function ( S^ST^T ) has to be
accepted to account for them. The Naiyyikas ( as also the
other schools who not accept S^^^T as a third ^T^f^r ) argue
fhat the suggested sense comes, through a mental process of
inference after knowing the frf^T^ or ^s'qirm as the case may
be.

or

'

5i^ite is the purpose or result ( <ESS ) of

and its essential features are briefly described below.
Any sentence or proposition, as all knows, consists mainly of
two parts, the subject or the substantive part called ^ ^ about
which something is predicated/ and the predicative part called
f^N". According to the Naiyyikas the meaning of the substantive in the nominative case gWT"3m, is the principal
concept, gwffiw, to which th meanings f all other words
join up directly or indirectly. There is however an additional
element conveyed by the sentence as a whole over and above
the concepts of the different words. This is the intended
relation of the various individual cDcepts, q^TOfcr, which
is brought forth not by the significative force of the words,
but by the syntactic juxtaposition of the various words,
technically termed as ^ T W ^ T . A sentence like 4fe
would give rise to a ^ns^^rta like "Cft^^-T^^
W^^-#f%^T^
form. This is
called ^rj^^prf^t^^^r; where the action denoted by the verb

138

TARKABBSS ( <n&n<iT )

qualified by all its adjuncts rests on the principal concept, the

subject,' and this is the distinctive feature of their
i
The Grammarians whose rules are the final authority for
the correctness of a word and its syntactic usage in combination with other words in a sentence, do not agree to the above
view of the Naiyyikas. They hold that in Sjs^ite the action
denoted by the verb is the principal concept and the meanings
of all other words includiug the subject are sub-ordinate to it.
This is called srrcqrergqqfgrit^srte. The above sentence would
according to them, give rise to a judgment like qT^CT^-W^^^xi^sp^-^crT^fe'lrfT^Rflf:.
The
Mirnmsakas
follow the Grammarians in general with this difference, vi.
that the arider or verb ending connotes an additional element
besides fifr which they name ^T^RT, "will-to-do". There is
wide difference between the Bhttas and the Prbhkaras, the
two chief rival schools among them in the exposition of
Ssste centering round this rfR..- These details need not
detain us for our present purpose. What is to be noted is the
fundamental difference between the Naiyyikas and the other
schools, whether the principal concept in a verbal judgement
is the subject or the predicate. "If from the stand-point of
logical analysis the subject is the central concept of a judgement the meaning of the root of the finite verb may be regarded as its central concept from the stand-point of linguistic
analysis; or the 'will-to-do* ( sn^TT )> denoted by the ending of
the finite verb, may be viewed as its central concept from the
stand-point of Mimms lgalisme* [ A Primer of Indian
Logic by Prof : S. K astry. p. 259 ].
The purpose of reiterating in para 64 that only four
are recognised by the Naiyyikas is to refute the two extra
TOTE'S advocated by the Mmmasakas and the Advaitins. In
this connection the number of siwn's accepted by the different
schools of Indian philosophy is noteworthy in as much as it is
an index, in general to the standard and level of epistemological skill attained by each school and may serve in many
cases as a clue to their mutual priority and indebtedness The
who may be considered the staunch Indian materialist

139
accepts ac'W as the only smvn anddoes not give credence to
any other. The iNtf, *ftq; and ?r have two, *j?*re and
srgrra. The gfair, *t*T and frfre^c schools accept three
adding srs^ to the above two. The ^*nf?E has four with wmm
( as described here ). The STO35Ttara$ adds yet another,
amfafr ( vide para 65 ) to the above four. The YNgtfbttere and
the 3*|fe^ have six gwi's adding smra or srgq^fsHr to the
above five. The <ftafr^ and cnfero sometimes reckon tivm,
^ H and %gT also. *h*sr occurs in sayings like cfifty is included in hundred9. \fam depends on belief in such myths as
'there is a ghost in this tree'. % occurs when some-one raises
two fingers to indicate the number two. These last three rest
on very shaky grounds and so our author does not consider it
necessary or fit to refute them. The real controversy thus
centres round STOTT and srgq^fssr. The *faTr% is as stubborn
and vehement in defending their separate existence as the
RfsF is in repudiating their independence and including them
in argwrand srre respectively.

[65]

I
i

Translation( objection ). There is another independent

instrument of cognition called areiqf ( = presumption). When
one finds that a certain fact is inconsistent on the face of it

140

TARKABHAS ( dfcrRT )

and when he presumes some other fact tooremove or explain

that inconsistency, such a presumption is what is called s?qf%.
For example, when one hears or, sees that Devadatta is fat
without eating by day he assumes that D must be eating by
night, because his fatness is not consistent with his not-eating
by day, (and it is consistent if it is presumed that he eats by
night). This presumption based on the inconsistency regarding fatness constitutes the instrument for cognising the fact of
D's eating by night. This is distinct from 5K*re and other
TOM'S, because the fact of eating by night cannot be ascertained
by the other Slum's.
( Answer). This is riot so as the fact of eating by night
is the object of ( i.e. is arrived at by ) inference. The syllogism
is as follows :
1. This Devadatta eats by night.
2. Because he is fat and yet does not eat by day.
3. He who does not eat by. night cannot be fat and at
the same time be not-eating by day, just like another
one who does not eat by day or night is not fat.
4. He is not such, i.e., D is not not-fat and yet does not
eaiD by day.
5. Therefore he is not such, i.e. D is not non-eating
by night, i.e. he is eating by night.
When by means of this negative inference one can ascertain
the fact of eating by night where is the necessity for stipulating
smqf% as a separate instrument of cognition V
NotesThe word smqf is used both for the SHW and the
5THT arising from it just like sjew, When it means SWTO it is
derived as wifcq1 aarrqf^: = Jrfqftt TSRq; S, when it is called
^qqrei^rriT == that which has to be explained or accounted for.
When it mans; SOT it is deprived as v&lwt snqfr: = ^^R, when
it becomes ^qqoj^R s that which explains or accounts for
the inconsistency in the former. Thus the knowledge of a fact
which is not accountable except by presuming another fact is
the instrument and the knowledge so presumed is the resultant
cognition. The fact that D is fat -without bating by day is
what has ;io be accounted for ( srqqrasgR ). r TJais can oMinari-

141

ly be explained only by presuming that he eats by night

( 37<n595fi?f ) Examples of this type are quite common. A
met B in Madras in the morning and then flew to Delhi. B did
not travel with him but A meets him again in Delhi the same
evening. A has then to presume that B must have travelled
by another flight the same day as otherwise his presence in
Delhi the same evening could not be accounted for.
In page 86 it has been mentioned that the Mmamsakas
and Advaitins do not accept wrf^gH but include all such
sgH under a separate vmw ( srohfr ). The Bhtta
Mmamsakas explain that W?qf% consists in removing a conflict between two valid cognitions, one of which is general,
^TORor, and the other specific, ergTOftof* which leads to the
conflict or inconsistency in the former. This fact necessitates
the presumption of a third knowledge which removes this
conflict In the stock example D is known to be fat,
5?R. He is also found to be not eating by day,
which raises the question, 'How then can he be fat ?\ This
conflict between the general and specific knowledges is not of
an irreconcilable nature as between two contradictories; but
only partial as between a general rule and exception; and this
is found to be easily accounted for by presuming the third fact
of his eating by night. Without this presumptive knowledge
the validity of one of the previous cognitions, general or
specific, becomes doubtful. So the Bhttas argue that the
inconsistency between the general and specific cognitions
directly leads to the presumptive cognition ( srifqrf% ). It is
here that the other school of Mmaniasakas, the Prbhkaras*
differ from the Bhttas. They contend that it is not the inconsistency between the general and specific cognitions that
leads to WJ?qf%; but it is the doubt of one of them becoming
invalid that leads to it. A closer look at the Bhtta view
would reveal that their exposition is nothing but an elaboration of the oqfa^TgWT of the Naiyyikas in substance and
as such it would be difcujt to extricate Wqf% from the
clutches of such an inference. While the Prbhkara view
may somehow try to redeem it from being engulfed by ssff^,; itsweak point is that it exalts doubt to ,the status of a

142

TARKABHS (

as mentioned above. The Prbhkaras would welcome

the objection as a compliment as they hold that all experience
is valid ( sfg^tn WT ). But both the Bhttas and the Naiyyikas do not go hand in hand with the Prbhkaras and so the
three-cornered controversy goes on for ever.
The real difference is between the Bhttas who want
to be treated as a distinct SRTT and the Naiyyikas who
include it under inference of the ^ssrf^f type; and each
side claims the advantage of ^ ^
over the other. The
Mimtaaskas do not admit sqf^^igTi^, and so for them it
is simpler and easier to explain cognitions like D's eating by
night by stipulating a distinct smqfr. Moreover they argue
that what the Naiyyikas call ssrfrc^TgFrFr is no inference at
all as in all such cases one has to pass through a state of 3f?^r5$% before arriving at the resultant cognition which is always
positive. They also hold that the after-thought ( ^ 3 0 ^ ^ ^ )
that occurs in one's mind after presumptive cognition is of the
type *I presume' and not 'I infer' ar^sngqq^n ^q^TTs,
H g sgfrstf. The Naiyyikas counter by saying that this
presumptive knowledge is only a foreshadowing of the imagination of what really results from a process of inference
through negative concomitance, and in as much as s^f^^^Tg^R
has to be accepted for other reasons 3qfi can and does come
under that heading and need not be raised to a distinct Sft(jm,
which would certainly involve *rW^. The real objection
against, accepting ffq&igjRH lies in the fact that a negative
ssnfn leads to a positive conclusion, which may be contrary to
experience. The weak point in the Bhtta view is that, if their
arguments are stretched a little further, the entire srptHspiraT
is in peril of being swallowed by $irqf%a position which
none of the schools acquiesces in. The Bhttas appear to be
aware of this fitfall and try to escape from it by saying that
the universal **nf of all fires and all smokes through a few
observed cases and many unobserved on&s cannot be accounted

for by any W*T other than srgiR. This is evidently an afterthought and does not convince any logician. Chennu Bhtta
in his commentary quotes a verse which gives a general
maxim aptly applicable to cases like this :3

143
II* This means
that when the objection and the rejoinder are similar on both
sides, either side should not be questioned 'why}' while
examining such a topic* It is perhaps for this reason that he
does not enter into any arguments, but has simply stated the
position of the two schools, unlike in the case of srora which
follows immediately below.
The Bhttas recognise two kinds ofwtfqft, ^cTTOfarT and
wrafafe according as the .jarqqraqH comes through verbal
testimony or by perception. To include both varieties under
gw the author uses the phrase %f s$ ^ ^T in the opening
part of the objection. In the answer portion he has given the
syllogism complete with all the five members so that there
may be no doubt about the merging of sroqffr with argSTm.
Next the author takes up the refutation of the other
or
f

( 8)
[66]

fa

?RPT

trf

Translation( objection). There is another distinct

named STTT^ = negation and it has to be accepted for the
cognition of 'non-existence' (of things). For instance, the
absence of jar etc. is cognised by srgqgsfsgf ( non-apprehension )
of jar etc. Non-apprehension is only the negation of apprehension and so the non-existence of jar etc. is cognized by the
instrument snra or
f
( Answer ). This is not so; because non-existence is
grasped by sense-perception itself which is aided by a ratiocination ( gn ) like 'had the jar existed here it would have been
Ch. : Bb. : quotes this verse on p 167 (Bombay Edition ) under
But it applies here also.

U4

TARKABMSA? ( n

percevtf just like' the ground', in combination with ifsonapprhension.

[67J ^fNfeqTfq -iTRamTpPT I CTqif^ I

!

Translation^ The Mmamsaka objects and the objection is

continued till the end of para 70 ). The sense-organs cognise
those objects only that are connected to them. This is proved
by a syllogism HJce this :The sense-organs cognise objects on
coming into contact with them, because they are instruments
( causes ) of cognition like light Or the syllogism can be
stated this way :The eye and the ear cognise their objects
after reaching them, because they are external sense-organs,
like the organ of touch. That the tactile organ apprehends
an object after its coming into contact with the object is
admitted by both sides. Like that there is no relation between
the sense:organ and omtt. Conjunction and inherence are
th only two ( tangible ) relations. Neither of these is possible
between ijf?^ and srarc, as conjunction can take place only
between two substances and srara is not a substance. Also
inherence is out of question as these two are not inseparables.

[68] f^hr^MN^T^T FSfrer.er mf i

ft i^
:I

TranslationThe relation existing between qualification

R i ) and the qualified (ffcsr) is no relation at all, as
it is not a single entity subsisting in two things distinct from.
itself. A relation should always be ( 1 ) distinct from the
relata, ( 2 ) should subsist in both the relatedjthings, and (3,)
should be a single entity; just like the conjunction between
the drum and the stick. This conjunction is different from:;
the drum and the stick, subsists in both of them and is only
one. The relation between qualification and the qualified
fMHGfrteW^is not such ( i.e. does not satisfy any of.
these conditions ). For example the relation between the
stick, the qualification, and the man, the qualified (in a
cognition like ?%&t g^q: ) is not something different from these,
two; for the character of being a qualification (frfa<JRn)
existing in the stick is not something distinct from the stick
itself; nor is the character of being the qualified ( f i ^ r T )
existing in the man anything distinct from the man himself.
Both these characters-frftfren and "'ft$c*cnare identical
with the forms of these two thingsfrtaw and frN?r. snrra
also has both these characters ( i.e. it becomes bothfrefaoTand
f^sr ) If it were different from these it should 'come under
any of the six positive categories ( ^sq-, ajar etc ). But none
of these categories subsists in smra. Therefore the f^rtaoi&r
subsisting in erarrar.is nothing other than Jits own form which
consists in its capacity to cause the knowledge of the thing;
related to it.
[69] q;^ ^T^^c^^TT^^T^^^irT* 1

iTcf ^TT'^I' 5 ^ I fT^TT%

TranslationThe same argument should be applied to
relationships such as *qTre?gr, 7WK3, etc; ( i.e. the relation
between the pervaded and the pervader ^^smqeEW^that
between cause and effectw^RWreetc. ). The character
of pervading subsisting in fire etc. ( ssnqs^ ) is identical with
that form of fire etc. which enables the cognition of something
concomitant with it ( i.e. its ssri^ ). Similarly the character
10

146

TARKABHS (

of being a cause subsisting in the thread etc. ( 35RGR3 ) is nothing other than its form which enables the grasp of the concomitance, both affirmatively and negatively, with its effect,
swrer also becomes suqsp and spH^r. ( If this character were
to be something different from swre itself, it should be a
generality like ^^m ). But such a generality does not subsist in

Translation-Thus the relation {ir^^{^5t^*T^ is not

distinct from lthe form of the qualification and from that of the
qualified* Nor does it subsist in both. In the qualification
there exists its character only and not that of the qualified
and in the qualified there exists its character alone and not
that of the qualification. It is also not a single entity ( as
or ^R3T* ); for the word *tra occuring at the end of a
compound ( fffa&i frM ^ fip^f^ OTte* ) should
be construed with each of its members, a t this gives rise to
two characters ( fifihaTWsr and f!rcfc*rarraO. But a relation
can be only one ( as already stated above ). Therefore fNuTfn*RT*n3r does not become a relation at all. Similar is the
case with s^T^^Tq^m, etc. The application of the word
^JFspqf to these is only by implication, due to their similarity
( with tffaT or SWWT ) in co-forming both the relata ( through
some sort of co-existence ). As a result of the above discussion
mm is jaot capable of being perceived by the sense organs as
it can have no relation with them. (So a new r5T*nBi should
be postulated for the cognition of mm ).

Ui

[71] mm i
i

Translation( Answer ). ( In the inference quoted earlier

that sense-organs cognise those objects which come in contact
with them ) the snfr is to be limited to positive objects
( *iraqrifc ). This means that sense-organs require connection with the objects only when they cognise positive things
and not in the case of WTIST. The cognition of WTre takes
place through the sense-organs by fcawfMtKnTro { as explained under perception ) and this is our established conclusion
( ff^T^^T ). The objection that sense-organs would in that
case cognise things not in contact with them has been answered by us by introducing the relation f^offiT ( vide p. 51 and
55). The objection is common to the opponent's view also
( where they meet it in their own way ).
NotesThe reader may again go through the notes on
pp. 30-32 where the N. V. views about snrra have been
explained briefly. The Prbhkaras do not accept smtsr as a
separate category but hold that it is identical with its substratum ( etfmraroq ). A cognition like 'there is no jar at
this spot' actually refers, according to them, only to the
empty spot. This is only a round-about way of accepting
3WT3 without admitting that fact. The Bhtta Mmmsakas,
who are the chief opponents here not only accept smre as a
distinct category but stipulate a separate mw for its cognition. Their arguments are briefly stated here as jcfq^ ( objection ) in paras 67 to 70 and the Naiyyika replies briefly in
para 71, as it has already been explained under para 31 ( as
also in the notes on p. 55 ).
* The verse quoted in the notes on p. 143 ( rtTsft": etc ) is found
printed after ^fOT^ as part of the text in all the published texts of
t^^TNT. It is not found in the MSS. relied here nor is it commented
upon by Ch. Bh. ( vide footnote on p. 14 3 ).

148

TARKABHAS (

The Bhttas in their anxiety to cut at the roots of the

Nyya view, begin by establishing that in s^w the sense
organs should have some sort of purposeful connection with
the objects they cognise, that they cannot have any connection with smR as it is a non-existent entity, that it cannot be
cognised by any of the other accepted STOP'S and that therefore a new TOM, which they call agq<siw or srora* should be
accepted for its cognition. They argue that fo. f. ?n. and
similar relations like ^ntiprara, ^TOawratt?, etrcrro&raro,
etc. cannot be considered as purposeful relations like ^zi or
WRrer. A relation, w^?^, according to them, should rest on
both the things they relate, should be something different
from the relata and should be only one in its form. All these
pseudo-relations do not satisfy any of these conditions and
cannot be considered as relations at all. They are just certain
temporary characteristics which help in their own form to
bring together the knowledge of the two things related.
Because of this similarity the word ^WHT is used to denote
them in an implied sense. In this way their reluctance to
accept fw. ik. *n. as a relation compels them to resort toas a distinct
The Naiyyika counters by saying that fit fr. w . is nothing new as it has been accepted by all in all cognitions. Since
it occurs in cognitions of ntffa like %% sret ?nfe it satisfies
all the conditions for sense-object contact and so there is no
need to postulate a new srepfl. Moreover, if sgq^fss is posed
as a new.swo, that itself being an arare should have another
sjgqosfssr for its apprehension, which in its turn will require
another, and so on. This, therefore, leads to endless regression a^Rc^Tand thus the Naiyyika claims the advantage
of ^ivm on his side.
This controversy really centres round the cenceptions of
fRssjf and 3*n3[ which differ considerably with both the
schools. To the Nyya realist relation presupposes diflerence
* The word ^f^^ itself is used to denote the instrument of its*
cognition, To avoid confusion we shall stick to

1$

149

to the total exclusion of identity and this makes Ms relation

as something wholly external and never internal. This in its
turn necessitates the acceptance of the view that the parts
are distinct from the whole and that they do not lose
their separate entity when they become the relata of some
relation. With this background he has no hesitation
to accept .^pTCPCF* like ft. fir. *n. etc. as a purposeful relation like spita or ?3*raTO. But then he has to
explain his position which reduces #?itT to a gif, but raises
3TO3W to a q^isf and makes *vTO3?qr to assume the forms of
the two relata which can be any two things out of the seven
categories. The Bhttas hold that difference is not incompatible with identity; that all relations involve difference and
identity in however small measure it may be; that no relation
can exist between two dissimilar things like Delhi and
Monday; that all relations are internal and their function is
to unify and not merely to bring two things together. This
argument of theirs would spell the extinction of hifar which
they support by making it closer to HSWT. Their contention
that m^icTara^cr is wholly internal while ^n^TOnwsj is no
relation at all is contrary t experience and is totally* unpalatable t the Naiyyika to whom difference ( ifcj ) is uncompromising even to the extent of repudiating rH^TcTO. These and
similar differences among the various schools of thought have
persisted ever since their origin atid their speculations have
contributed considerably to the advancement of philosophical
thought. So the position remains that the Bhtta ( and
others of his ilk mentioned earlier ) refuses to accept
J
and fsf. k. *3. as a *?ftf<! and instead postulates
and sroq&fai to account for the respective cognitions.
The author has therefore rightly ended the discussion by
saying that this is our fcngi?er and that loop-holes for objection
are common on both sides, which can always be answered in
consonance with their own dogmas.
After describing the four instruments of cognition as
accepted by the Nyya system the author enters into another
important topic connected with the validity or otherwise of
the cognitions arising through these. This discussion is
Commonly known as

150

TARKABHASA (
( 9 )

TranslationThis is being considered now. (In general) when

a man knows that there is water ( at some place ) he proceeds
to get it only after ascertaining the validity of his cognition.
In some cases he may act even on the basis of doubtful validity and if he finds water after his effort he determines that
his knowledge is valid. This is the actual state of affairs. On
this some people ( the Bhtta Mmamsakas are meant here)
object as follows :'Before his effort a man ascertains the
validity of his knowledge and then acts, because validity is
ascertained intrinsically.9 This means that when a man
gains a certain knowledge through whatever means, he gathers
it validity also through the same means. No cause other than
those that created that knowledge, is required for ascertaining validity which is a characteristic of the knowledge. Thus
intrinsicality ( ^3RW3 ) of validity constitutes non-dependence
on causes other than those that produced the knowledge.
Knowledge precedes effort ( f fr ) . Otherwise how can there
arise doubt about its validity or otherwise ? No doubt can
arise about a thing before it is known. Therefore it has to be
accepted that before he proceeds to act man gains knowledge
through smqf%5W based on the unaccountability of the

151
'known-rness9 ( SfjRRTT ), and he apprehends along with it the
validity of that cognition as well through that arsrqfr itself,
and theti he proceeds to apt. It is not correet to say that at
first knowledge alone is gained and thecu its validity is determined by seeing the fruitful result of his action. [ Up to this
beginning with ^This mean^ is the author's explanation of the
Bhtf objection ]>

[73]

1
?rrr

w^r^ 1

Translation[ The above objection is being answered as

follows by the Naiyyika ]; Firstly we do not admit what you
say about knowledge being grasped through snqf based on
the uiiaccountability of gRRn. So what to speak about its
validity through the same means ( sriqf% ) ? To explain,
this is what the opponent means. When one cognises a jar
( etc. ) he also apprehends the steel about the jar in the form
c
l have gained knowledge of the jar'. Through this it is
inferred that a characteristic called 5nft3T is also produced
when the knowledge (of the jar) occurs, ^mm does not
arise before this knowledge, but does arise after the knowledge; and because of these positive and negative con*
comitances it is determined that ^nm is produced by the
knowledge ( of the jar ). As no effect is produced without
cause, this fRRn produced out of the knowledge, cannot be
accounted for without its cause, the knowledge. This is how
through srajqfrraTO knowledge ( the cause ) is brought about
by oracTT ( the effect ). This view is not correct for there is
nothing like ^rram apart from the fact of being the object of
the knowledge.

ill

TARKABHp (

^TcRTT

%fe:

Translation( The Mmasaka again objects). The fact

thai jar |becomes the object of a cognition is nothing other
than that of its being the substratum of 9rcrit that is produced
by the cognition of the jar. It is like this. The object-ness
of the jar is not brought about by the relation m^T?*q[ because
no identity is admitted between the firesr, jar, and the fsr^f^;,
its cognition. If it is contended that the objectness of the jar
is due to its producing the cognition, then the sense-organs
also would become objects of cognition which is produced
by them. Therefore it is inferred that the jar .acquires a
characteristic ( through the cognition ) by which that alone
becomes the object of that cognition and not other things
( like the sense organs ). It is in this manner that ^fffiRH is
proved to exist by sfsrfqfrr caused by the un accountability of
$he object-ness, and not merely by sense-perception.

[75] Am i ^rrwRT^r i^r^ftwf^>q^t: )

i
on(Jh& Naiyyika answers); This is not correct;
for the characters of becoming object and subject ( firra^T
and firmen ) arise from the very nature of a cognition.
Between a thing and its cognition there is this natural peculiarity that the former becomes the object ( fam ) and the latter
the subject ( f*ff*nO in relation-to each other. If; this were
not so no object-ness could occur in past and future things
( which are presented in one's cognition ), because SFmm does
not exist in them in as much as no characteristic ( >*H like
^Tcrm ) cari be produced in a thing ( crfFRj that is not there.
Moreover SJTCRT itself is the object' of a cognition ( presumption
as explained by the opponent in para 73), wMh would require
another ^rer?T ( for the support of that cognition ) , a n d that
another and so on endlessly. Thus it Would lead to the
defect of endless regression 3TT3W. If (-t avoid this
defect ) you say that without-a second ^RTST the first one can
become the object on its own, then why accept the first sjiRT^T
( as existing in the jar when the jar itself can become the
object on its,own ). Granting ( for the sake of argument )
the existence of ^rirKTT, even then knowledge alone (and not
its validity ) can be produced by that ^fRmr. Validity can
come only through a particular sgrt"' which does not stray
away from a valid cognition and so how can it be said that
validity is grasped by the same means that produce the cognition ? If some such particular sTRm which does not stray
away from a valid cognition were there by virtue of which
knowledge and its validity are grasped simultaneously, the
same argument can be applied to invalidity also, viz. that
here is some particular gyrerar which does not stray away from
an invalid cognition and by virtue of which that knowledge
and its invalidity are grasped simultaneously. In this manner
invalidity also would be grasped intrinsically ( which position
is not acceptable t both sides ). If in spi^e of this, invalidity
is held to be grasped extrinsically, then validity should also
be held to be grasped extrinsically, i.e. both are grasped by
means other than those that produce cognition.

154
i

%fr i cT^r T

io

Il
translation-{The
Naiyyka explains his position).
Knowledge is gained by mental perceptiprr and its validity by
inference.^ It is like this. After cognising water ( a t a place )
a person effort in search of it eads either way, fruitful or
fruitless, ^he;n it is fruitful it is successful ( rw ). Through
such, a fruitful effort is inferred the true-to-fact nature of that
knowledge, which ris what is called 5Ti*tta^ ( validity ). The
syllogism is as follows. The knowledge of water which is
under dispute is valid, because it causes fruitful effort. The
mB ( here ) cthat (knowledge ) which.is not valid does not
cause fruitful effort, like an invalid knowledge', is a purely
negative one. Here 'knowledge of water which causes a
fruitful effort' is the qf, and snsn^ ( validity ) is^the probandutn. simi^ here means /the fact of being true to its
nature' and not the *fabt of being instrument of a valid cognition'. ( If the latter meaning is adopted ) there will occur
the faclt.oifiFW? 'in remembrance ( which is also a kind of
knowledge ). ^g or probans is the 'fact of causing successful
effort'. After having ascertained through negative inference

155

the validity of a knowledge which has led to effective action,,

the validity of another knowledge which has not led to such
effective action is also inferred, on the analogy of the previous
inference, even before the effort for search of water, through
positive and negative concomitance of tho.prpbans which has
(the added advantage of ) similarity to. the previous %g
( assn^HNrfeifar ). Therefore we conclude .tji^t the validity
of knowledge is grasped extrinsically and not by the same
means that produce the knowledge.
;
;.
There are only four spron's and Keava has explained ail
four according to the Nyya astra, for the benefit of beginners, mentioning only the bare minimum of arguments bearing
on each topic.
Thus has ended ( the explanation of )|he category
NotesHaving explained the instruments of valid cognition it is but natural to enquire into the fact whether the
knowledge gained through these instruments is true to fact or
not, or in other words valid or invalid. A thirsty man on
perceiving or hearing from someone that there is water yonder,
goes to the place indicated. If he finds water there he determines that his knowledge of the existence of water is valid,
and if otherwise it is not valid. On the analogy of such
experiences the Naiyyiks hold that validity or invalidity of
a knowledge is determined only after the object of the cognition has been effectively tested. In other words neither
validity nor invalidity RWT^ or armw*^ of knowledge is selfevident or intrinsic = *3cftUI. *^ft3*U^, or in short ^cr^r,
is defined as ^TTTOf ssnfft^R^si, or the fact that validity or
invalidity is grasped independently of causes other than those
that produced the knowledge; i.e. by the same causes of that
knowledge. Its opposite, extrinsicality, q^STiR^ or -qro;?
in short, is ^TT^Tf ^ffTtRT^c^, the fact that either is revealed
by causes other than those that produce knowledge. With
these definitions in mind we get four possible views regarding
validity and invalidity.
1. Validity and invalidity both intrinsically or ^rer:
view held by the SmkhyaSo
2. Both extrinsically or m$:~-view, held by the, Naiyyiks.

156

TARKABBSA (

. 3. Validity ^TCT and invalidity q^rr:-^-view held by the

1
Mmmsakas.
4. Validity <r*?r: and invalidity wnview held by a
section of the Buddhists
By explaining the arguments of the chief exponents of
ssfawr and TCf<K3f the topic can be clarified sufficiently. So
the author after stating the Nyya view in general in para 72,
summarises th view of the Bhtta Mmmsakas as <^q^ in
para 72, 73, and 74 and finally establishes the Nyya stand in
paras 75 and 76.
There is considerable difference in the theory of knowledge
among the various schools. Here we shall brieflly refer to
two of the contending schools as relevant to the topic under
discussion. According to the Naiyyikas a cognition like ^
Tsrer or arar fs is called wreaT^R. This cognition is processed by the mind into a judgement which is called STgssrastt^
( after-cognition ). Being a purely mental process it is considered to be JWgi$re*r^ = mental perception ( vide the author's
opening sentence in para 76 ). This usually takes the form
?5 SRTftr or srwf, embodies the s^srregTsr as its
object and produces the knowledge of the object ( silver, water
etc. ) of the cognition. It does not reveal the validity or
otherwise of the cognition or of itself. If this knowledge
were to do so it would not be possible to explain the
doubt that sometimes occurs whether the object cognised
is real silver or not, as no doubt could arise about a
thing the validity of which has already been ascertained. The
validity or invalidity comes in by a different cause by inferenceafter the person has picked up the object in front and
tested it. If his effort proves successful, i.e. if he finds it is
real silver, he infers as follows :53 *sr3^?f W , e^srfr
5R5Ptmil m sressFf f%5ff%c n sw, S W srttFaiq H If however he finds that it is not silver, but only a piece of tin or
nacre, he infers the invalidity also by a similar inference; viz,
t In both cases there can be ssrferfo sqr% as
well, viz. 3TT *r SOTT HT *r CTfi^Rf f^ srfa^r. STSTSSWT, etc. ( as
pointed out in para 76 ). Thu the Niyyik standf is that

knowledge is presented by srgs^qcnq as a mental perception,

then comes the effort ( Sff% ) prompted by the need or desire
to secure the object of the cognition, and lastly the validity or
invalidity of both the cognition and the knowledge from it
3W& get ^through a process of inference ( q^g: ) based on
the fruitfulness or fruitlessness of the effort.
There are three distinct schools of thought among the
Mimmsakas; ( 1 ) The Bhttas or followers of gssnft^ .*rg
( circa 8th century A. D. ), also called m f i ^ r c as he wrote the
3$ta3if andftTsrarf&gon ^3**n^ o v e r t^e- Mimms Sutras;
( 2 ) The Prbhkaras or followers of sim^t also named g^,*,
an younger contemporary and student of ?*nft3S syg, from
whom he differed considerably in the interpretation of the
Mimms Sutras as well as its *n<*T; and ( 3 ) The Miaras or
followers of ^RfrfTa ( not later than the 12th century ). All
the three schools hold the view of ^citsiTirc for validity and of
r*ctilsr for invalidity, but in their own method- To the
Prbhkaras all experience is sm\ ( sTJ^f: STO as already
mentioned ) and any invalid cognition ( or 3R as it is called )
is only a case of non-discrimination. So there is no question
of invalidity in their theory. For them knowledge is selfluminous ( ^s^^r ) and in every cognition the knower fTOT),
the object to be known (.$*)," knowledge itself ( sn*f ) and
its validity ( SWT?3[ or suwqr ) are all presented at the same
time. They make no distinction between sq^rw and
is supposed to be a Kerala Namboodiri who went to
Benares for further studies. There is an interesting story as to how he
came to be named (J^. Along with *F*grT, ^*%^5 and others he was a
student of i|Wfc5 *Tg. One day while explaining a topic, the teacher
was confronted with a sentence ^T^g ^TR, Sf^f *fa, ^^^T^T^t'tfi^ in
the middle of the text. The teacher was a little puzzled over it when
in his enthusiasm explained, " g ^ ^ f ^ I <Ts( rp 3 " ^ ^ I 8J3f
^fc^; i %W ^ ^ S ^ g 1 ^ ^ l" ^Tf^5 was struck by the
genius of his pupil and in his joy proclaimed, '^TTcf JJ^* Since then
th name stuck to him and his followers are quoted as J J ^ : -by later
scholars. Even to-day there are more followers of 5*3? in Kerala than
in other parts of the country.

158

TARKABHS (

^ R ' s . The Miaras, like the Naiyyikas, accept

and make it the cause for the knowledge from the cognition.
But, unlike the Naiyyikas, hold the validity of the sireraro is
revealed by the afgetragra simultaneously, thus, advocating
T for validity.
The Bhtta Mmiisakas who are the chief opponents here,
hold that all knowledge is transcendental, aTft{?^sr. This
means that a sense-organ f say the eye ) when it coins into
contact with an object (say a jar ) does not immediately
convey to the mind the knowledge, SR gs: as knowledge is
ST^ftfiF; but it causes the mind to apprehend a peculiar
characteristic ( w ) in the ^2 which is of the form c5Rt ^R^T sre:'
and hence called srreraTknown-ness. This ^rm^T can happen
only after something is known and so to account for it the
knowledge of the jar is presumed. It is thus that the knowledge
of the jar is presented by ^rram through srofqfi and is expressed in the text as ^ra^Wgqqfs^^TSsq^ ^"R *Jf. When
this knowledge is thus presented by ^mm> along with it is
presented its validity also by the same ^ m , and no other
cause is needed for it. Only after knowledge together with
its validity has been grasped in the above manner does a
person proceed to , act.
In short what the Naiyyikas
explain by sgssRSnsr the Bhttas try to explain through this
sTRfHi in consonance with their theory that knowledge is
transcendental.
The objections raised by the Naiyyika against the
hypothesis of ^cHSSl f the Bbttas are three-fold. Firstly he
does not subscribe to the postulation of an imaginary ^TRRIT
which, according to him, is nothing, other than the objectness
(fpWctt) of the thing presented in a cognition. Accepting as
he does ^i4^H&Trra as an effective relation ( vide p. 147 ) it is
easy for the Naiyyika to account for the f^m in the jar as
a natural one without any external support for its presence.
Since the Bhtta does not accept 3H3fTCuW3 as a relation at
all and since ( as pointed out earlier ). knowledge to him is
he has to explain fRSR through his ^TrigfT as ^rrn^'i^cr^. The objection of the MImmsaka to this
point is that if jar is presented in the cognition as object

159
because it produces the cognition* then the sense organs and
similar other causes should also be presented in it for the same
reason ( vide 5*q*iT 3 fw% etc. in para 74 ). Though this
also stands refuted by the total rejection of ^m%\, it is evident
that the causality in the object is quite different from that in
the seiise-organs etc., in as much as the object is the content
of the cognition.
,
The second objection is that sTKRTT being presented in the
form ra sra?, itself forms the object of its cognition. To
establish this object-ness even according to the opponent's
method would need another 5H33T to support it, which in tarn
would require another and so on. To avoid such endless
regression if it is contended that the first ^T^TT is such that rit
becomes naturally the. object pf its cognition without the
aid of another, then there is no necessity for postulating
the first sil^clT itseli The .-fg*tT can be directly attributed to the object itself without the intervention of an
imaginary
r
Thirdly, even if for the sake of argument, f ^ i is conceded
for causing the knowledge of an object, how doe it reveal at
the same time the validity of the knowledge ? Validity is
grasped by 3*fiVR..{ or bysqf according to the opponent )
and in either case the j|g, sTR^TT should be such that it does
not stray away from a valid knowledge ( Rf^RB^ifT^if^^).
This means that gRT from a valid cognition alone can prove
its validity, but at this stage we do notf know whether the
cognition is valid or not. If ^Krai of water in re^rra proves
its validity also, how does. a person after a fruitless search
conclude that the cognition was invalid as no new STRT^T has
come in between ? This would in effect amount to accepting
the Pfbhkara view that all experience is valid. Moreover
the same argument can be extended.to invalidity as well and
it would be more reasonable to accept that validity and invalidity are both revealed intrinsically on the analogy of
validity as the opponents-hold; or that both are revealed
extrinsically as the Naiyyikas explain. i;> Either way there is

TARKABHSA :(

no, reason to hold that validity alone is Scfatsnir ;vyhile in?validity is i^t the same time qfertofr. '
It should be remembered that the Mmmsakas who
maintain that sound is eternal ( 3T5^faw ) and that the Vedas
are HTTf? and srqk^r ( vide p. 129 ) have to stick; to their
^r^^ theory as otherwise the authority of the Vedas will bs
open to question under external evidence. The Naiyayikas
too accept'the authority of the Vedas, but as the creation of an
Omnisciejat (3rbd and so advocate q?33*3 view and niaintain
^jjWT^,through His Omniscience.

In this connection it is necessary to know a little more

about invalid cognitions or ^t^ as any theory of knowledge
will be of little use if it does not enable one to distinguish
between sr*n and srsrtruth and error. At the outset our
author begins by defining sr*n as ^wrfgvrc and by the word
*TJW he excludes all doubts and erroneous experiences, which
are STSTOTO( videp. 5); This in. effect amounts to saying
that ^rm is ,ihat which is not. 3rrncf.< In Part II of this book
while treating about <gfj (vide para 121c) he explains
sr^TOP as wlTf^r sm*nsrar.. This does not however convey
directly what:constitutes the truthfulness or erroneousness of
a cognitioni

The following definitions from the y^m? of srswp will be

found more instructive, in this respect; gfgff^ ^w^Rtsg^r^t
?Trm: and;! WjprTMrf; ^5%lTts^^tOTim:. In both sentences
the pronoun H ^ denotes the ym* = ( fk%m = attribute ). These
mean that ( 1 ) a valid cognition ( TO ) is that in which the
Qualified thing is presented as qualified by its own attribute^
and ( 2 ) an erroneous cognition ( $wm or ^m ) is that in
which it is presented as qualified by an attribute other than
its own. If on seeing" a piece f rope one grasps it as mrtag',.
the rope is presented in the dgnition as qualified by its own
attribute ?3g^ and therefore it is %m. If however he cognises
it as oui iq:!5 the thing in front is presented to him as possessing an attribute not existig i it ( e 4 ^ ). Therefore it becomes $rtr. The theory of xplainig such erroneous cognitions is called in general ^%m^ and forms an important
part of the theory of knowledge in all schools of thought

ri*; There are") five important views ? about *s*flfraS ' which
are enumerated in the following verses. The second verse
gives the chief advocates of each of these five in th 'order
enumerated in the
first.

fg: H
s. i t 1

( ) sic^ifr or the theory of self-apprehension is the one

advocated by the Yogchra school of Buddhists. They
postulate that all determinate cognitions are erroneous*
According to them, when one sees a white piece of metal in
front, his self or sicT^ which is identical with one's conciousess, f%^, transforms itself into silver due to its similarity to
the object in front; and therefore the latter appears as silver
in the cognition.
( 2 ) The Tnsqft^ or nihilistic school of Buddhists advocate
the 3[cij?rf^rsT^ or the theory of apprehension of a nonexisting thing. They hold that even non-existing things are
capable of being cognised without actual sense-objecfrcontacL
There is therefore every possibility of the thing in front being
cognised as silver due to defects in the person himself or from
external sources,
,( 3 ) To the Prbhkara&i all; cognitions are 5j*n and all
erroneous apprehensions are cases of non-discrimination (as
already mentipned pn p. 157). So their theory is called sc^jfg ==
non-apprehension. Their contention is that in mistaking a,
piece of tin as silver, two cognitions are involved, ( i ) the
perception ofthe thing in frot as 'this' ( f^ ) and ( 2 ) the
recollection of silver cognised by him previously elsewhere^ but
without his knowing it as recollection. These two occur so
quickly that the observer does not discriminate between the
two, with the result that the silver from the recollective part
merges into tfie perceptive part and gives rise to the cognition;
3
:
' \
( 4 ) Wn^T% is the theory of the Naiyyikas. All erroneous
cognition arise out of certain defects, ^i^'s, in the observer, im
11

62

T A R K A B H A S ((

ra

the object or in any of the attributive causes. On seeing an

object lying in front in a zig-zag position a man grasps it as
only 'this* due to bad light, distance or some defect in his
vision. The close similarity of the object in front to that of a
previously cognised object brings out the recollection which
is embedded in his mind as a cfepR according to the general
rule ^e^f^rgTTTTq^H^^R^. So far the Naiyyika
does not differ from the Prbhkara. Thereafter he does not
agree with the latter in saying that somehow the two apprehensions coalesce into one and the observer blunders into the
cognition sr wu To account for the recollected object
presenting itself in the perceptive portion without sense-objectcontact, the Naiyyika postulates a super-normal relation,
^T^&TW*n[fl which has been explained in pp. 53-54,
Through this relation he has no hesitation to say that a
real object which does not have a certain attribute is
presented in the cognition as having that attribute
( 5 ) It will be observed that while all the above four schools
of thought try io explain somehow the presence of a nonexisting thing such as silver or snake in an erroneous cognition
none of them accounts satisfactorily for apprehending the nonexisting thing in a perception. The Advaitin sees through
the pitfalls in all the above methods and tries to surmount
them through the help of w n or beginningless nescience
WfTSlfcen. His theory is called ^ f a ^ f a m T f a or the apprehension of th indefinable. When the sense of sight comes into
contact with the object in front the above-mentioned nescience
throws a veil over the nacre-limited-spirit ( S^*Wl^4^T )
and thus prevents the apprehension of the nacre-ness and so
the cognitive function, s ^ ^ a i f f a , apprehends the object as
only'this*^spn^Rlfr. The nescience, which continues to
Veil the sgfe^, aided by the observer's prepossessions and by
the similarity of the thing in front to silver, undergoes a
change with the result that silver comes into being and superimposed itself on the sub-stratum observed as 'this'-3^3%?^^ta^ifafi^TsfsiT ^^n^K^ai qfrfl. This silver has thus
paly a; relative reality called SifgnfrspTi. It iserfMksfor as
it:iscno.t capable of bing defined as; .-legd, r^ > pr as unr$a].

163
^, or as both, H^fRJ. Thus the Advaitin considers sra as
a complex cognition consisting of two distinctive factors;
TIle
( i ) the apCTqnorff^r a n d ^ i J * t h e ^ ^ f ^ object" of
$N being relatively real, there is no need to accede to the
extreme view of ---rcROTTfct o r t 0 t l i e acceptance .of a supernatural relation proposed by the Naiyyika.
The Viistadvaitins, followers of Rmnujcrya, follow the
qpRlf theory of the Prbhkaras in general with minor
modification to suit their postulates and give it a new name
^Wlf. The Bhttas accept sr?*mnsnfr without the supernormal relation. In its place they substitute their 3TT3^n,
already explained above. The Dvaitins or followers of
Mdhvcrya also accept the Naiyyika theory, but explain it
in another form without the supernatural relation or the SjTTcRTT
of the Bhttas. Because of the novel way of their exposition
they take pride in calling their theory as ^frsf3I?qmraznf. The
Smkhyas explain sm in their own way calling it r^*s*nfci
which on analysis amounts to a mixture of the Naiyyika and
Mdhyamika theories, without, of course, the super-normal
relation of the former. In fact there are good grounds to
believe that the *nrfasp propounded his argqpinf by modifying the Smkhya theory. Jaina philosophers of the older
school follow the Smkhyas and those of the modern school
appear to fall in line with the Naiyyikas. Though each
school of thought shows slight modifications here and there
in their exposition and terminology these minor differences
need not detain us here.
A comparative study of the various theories of jsm9 would
reveal the inter-relation of the different schools of thought and
also provide an index to the epistemological acumen attained
by each. It would appear that the author of the two verses
quoted at the beginning of this discussion had in mind the
ascending order of merit of each theory. Before concluding
one is tempted to quote the words of the learned Professor.
"If one could imagine that epistemological thought starts with
afRIJisrTfl as centre and, in its endeavour to escape from it,
swings forcibly between the two diametrical termini of

TARKABHSS
and armfo, it would not be difficult) to imagine
that such thought inevitably describes a comprehensive epistemolgica! circle in the form Of ^f^^^q?s^T%, which easily
accomodates itself to sr^if in respect of the non-discrimination of the two' fftr's constituting *TO and to roqtnf
by complete surrender in the case of HtqfrsRW". [ S, K. S.j A
Primer of Indian Logic, pp. 130-131 ]...
END OF PART I

PART II

( i )
[ 7 7 ] SPTTOTHJirTfr I 3TT

TranslationThe instruments of valid cognition have been

explained. Now the objects of such cognitions are being
explained. They are ( twelve in number as ) enumerated in
G. N. S. ( 14-9 ).
1. mm\ = The self or soul 2.
3. ^^qr = Sense-organ
4. w4^Objects
5. gf^ = Knowledge
6. SR:=Mind
7. siff = Action
8, ^N=? Defect
9. ifc^igr = Re-birth
10. *S35 = Result
IL :?rs Misery
12. arqe(*T = Salvation.
Noiesaiter is the second category out of the 16 mentioned
in N. S. 1-1-1 on page 4. The propriety of enumerating the
's in the above order is explained by Ch. Bh. as follows,
is mentioned first as that is the agent enjoying all
experiences of pleasure, pain, etc. srck is the second as it is the
vehicle through which SW^ enjoys such experiences. Next
comes the chief cause of all knowledge, f^qrcfoi, which form
parts of the body. What the senses grasp, srovare stated after
them. Next comes gf%; or knowledge which the preceding two
together convey to the soul. It is only through *RH that
knowledge is carried to the soul and so that follows %.
Mind is followed by ngffr as all action to procure the desirable and to avoid the undesirable arises after knowledge.
Next comes 3[ta which, through these actions, leads to merit
and demerit. Birth is mainly due to one's past actions and so

166

TARKABHAS (

follows as a consequence of Sff and ^ta. The results

) which are to be enjoyed in this birth due to one's
past actions are mentioned after birth. The combined effect
of all these beginning from $rifa[ is suffering and so <|:^ follows
W . The cessation of all sufferings in salvation which ends
the cycle of birth and death and so it is stated at the end
The attainment of salvation is the end and aim of life and all
the Sastras are means to that endvide tfrR's first Sutra
i . n w ^ [ 7 8 a ] cr^TT^pFn^^FTTrr i "r ^

[78b]

*rerfr 1

[78c] ^ XT ^ T ^T#raWTTfs^: I

167
Translation(a) Of these twelve soul is that which possesses
the generality soul-ness. It is distinct from the body, senseorgans^ etc; also different in each body, eternal and omnipresent. It is the object of perception by the mind ( vide para
26 on p. 44 ). If there is any doubt ( about the existence of a
separate soul ) it can be inferred by such reasons as the
presence of special qualities like knowledge etc. The inference
is like this. Knowledge etc. are verily qualities, because
besides being non-eternal they are capable of being cognised
by a single sense-organ only, like colour. Also qualities must
subsist in some substance or other which they qualify. Here
knowledge etc. are not the qualities of the five elements (earth,
water etc. ) as these qualities (subsisting in the soul ) are
perceived by the mind. Those qualities that subsist in the
five elements are not perceived by the mind like colour, smell
etc. They are also not qualities subsisting in space, time and
mind ( the other three substances ) due to being special
qualities. The qualities of space, time and mind like number
etc. are not special ones; those that subsist in these three like
number etc. are not ( considered ) special qualities as they are
common to all the substances. Knowledge etc. are special
qualities because they are such as are capable of being grasped
by a single sense-organ like colour, and so they do not subsist
in space, time and mind. Therefore some substance other
than these eight ( earth etc. ) should be postulated as the
substratum of knowledge etc. and that is the soul.
Translation(b) The syllogism ( for postulating soul as the
ninth substance ) which can be a negative or a positive one is
as follows :Knowledge etc. subsist in a substance other than
the eight ( earth etc. mentioned above ), because they are
qualities which do not belong to these eight. The negative
strife is this sA quality that is not one which does not belong
to these eight substances is not one which subsists in a sub-,
stance other than the eight; e.g. colour etc. The positive sufr
( for the same syllogism) can be stated thus :that quality
which does not subsist in one substance must subsist in any
other different from the former; for instance sound which

tm

TARKABHS (

does not belong to earth etc. is found to subsist in effier which

is different from earth etc. In this manner soul is established
:
as a substance distinct from earth etc.
Translation (c) It is omnipresent because its effect is seen
everywhere* By 'omnipresent' is meant that it has the greatest
magnitude ( i.e. is all-pervading ). Being omnipresent it is
eternal like ether. It is also different from body to body since
the experience of pleasure and pain varies with each soul.
NotesThere would appear to be some incongruity, at
least for the beginner, when a thing is defined as having its
own generality, as in the definition of sn^RR^ and in many
cases below. The generality BTTW?^ is dependent on 3n<^and
so can be ascertained after sn**^ is known. This definition
therefore means that anw^ possesses a characteristic which
differentiates it from all other similar and dis-similar things
and this characteristic is what is being established by a series of
inferences, as the fact of being the substratum of knowledge
etc.STHsra^r. How it becomes such a substratum is also
proved by the same inference and after such a differentia is
established it is found as a common one in many asitenw^and
so it is accepted as the generality 3W*R3* It should be remembered that snw^here means only ^ft^T?^ which the Naiyyikas
ill ways differentiate from ^JC^ or God. Some authors
expressly divide snW3[ into these two classes after defining it
as the seat of knowledge. Cf. "gwfHT^WimRT I
\" ( ^T. f ). Even when speaking of
again under ^^i in para 91 below the author does not mention
of 'rsnsR^ but only of SNTSR^ to whom 14 qualities are
ascribed. By knowledge etc. ( f ^T^ ) the author includes the
other five as stated in G. N. S. ?^T-^-5iq^-[^-|:^-^T?n^TOTt f&m ( 1-1-10 ), meaning thereby that any one of these six
qualities can serve as a f^ for the inference he has given. For
the classification of qualities as fc^ta and m m ^ ones vide
notes under para 16, Parti. The author wants to contradict
the Buddhists and Crvkas who do not believe in w*T^ by
saying that the separate existence of the soul can be established

169
both by mental perception and by inference with both positive
and negative 3*T%. The followers of WTTgsf contend that sfttf^
is srg i.e. with the least magnitude as opposed to the Naiyyika
stand that it is fag i.e. having the greatest magnitude. The
pluralistic realism of the Naiyyika does not allow him to
accede to the Advaitin's view that sficfT?*!^ and q^wwi^ are
identical. To emphasise all these disagreements with the
opposing schools the author repeats in para 78 (c) what he has
expressed in the earlier portion, happily without any arguments as in the case of srqfr and

2.
*T).: I. *r

TranslationThe receptacle situated within which the soul

enjoys its experiences is the body. Here *iha( = enjoyment )
means experience of pleasure or pain. The soul enjoys such
experiences only when it is limited by some receptacle and
that is the body. Or, body can be defined as the sub-stratum
of action. %2T ( action ) here is not mere motion ( ^q*^ ),
but means such activity that is conducive to gain the desirable
and to discard the un-desirable.
NotesSHUT^ being omnipresent like ether, has contact
everywhere; but pleasure or pain occuring anywhere is not felt
by everyone. Only such pleasure or pain that affects one's
body is felt by the soul inside that body, just as sound that
comes in contact with one's auditory organ is all that one
hears and not all sounds produced in ether, (cf. ;3>ofopTThe intention of giving a second definition seems to be to
conform to the N S. where ifom defines body as '^fesqnsrfsw
Oq' ( L l - l l ) . The word 'IST has a limited sense as
explained in the text.
The printed editions have the word sreqsnrfir between
and 5Rfa*3( in the above definition. Chennu Bha|ta

170

TARKABHS

does not appear to have it in his text but Gpinatha notices it

saying that it is used to avoid the definition over-lapping into
limbs of the body such as hands, feet etc.

3. 5 % ^ [80a] wftlgt ^ ^ ^ ^ f e ^ I
i

^TTtferr^fvr^sfr^^: i
I

3TrT

[80b]

[80c] ^
i
[80d] ^q^wfr^g;miTffrfef

[80e]
1

[80f] ^

[80g]

ara1

171

[80h] rf5

tfe

H
Translation( 80a ) That which forms part of the body^
which is the instrument of cognition and which is not perceptible by the sense-organs, constitutes ^fir or sense-organ.
If it is defined as that which is not perceptible by the senseorgans, the definition would be so wide as to include time
etc, and therefore the qualification 'which is the instrument of
cognition' is added. Even then the definition overlaps 'senseobject-contact' and so the qualification 'which forms part of
the body' is added. If it is defined as that which forms part
of the body and which is an instrument of cognition (i.e.
without the word STcftf^T ) the definition would apply to light
etc. and so the qualification sref^sr is necessary.
( 80b ) These sense-organs are six in number, viz nose,
tongue, eye, skin, ear and mind. [ As already explained on p.
49 these are the seats of the sense-organs ]. Of these that
which is the instrument* for the cognition of smell is the
olfactory sense-organ and it is situated at the tip of the nose.
It belongs to the substance earth as it has smell like a jar. It
has smell because it apprehends smell. ( The general rule is
that ) that sense-organ which apprehends any quality out of
the five, colour, taste, smell, touch and sound, is possessed of
that quality as is observed in the case of the visual organ
which apprehends colour and so possesses colour.
( 80c ) That sense-organ which apprehends taste is the
gustatory one and it abides at the tip of the tongue. It belongs
to the substance water as it has taste. That it has taste is
proved by the fact that out of the five qualities colour etc. il
apprehends only taste, like saliva.
( 80d ) That one which apprehends colour is the visual
organ and it is located in the black pupil of the eye. It
belongs to the substance light because it has colour. That it
has colour is proved by the fact that out of the five qualities
colour etc. it apprehends only colour like a lamp.

172

TARKABHSf A (

(80e) That sense-organ which apprehends touch is the tactile

one. It exists all over the body ( i.e. in the skin which covers
all parts of the body ). It belongs to the substance air as it
has touch. That it has touch is proved by the fact that out of
the five qualities colour etc. it apprehends only touch* like the
wind from the fan, which manifests the cool touch of the
water in contact with the body.
( 80f ) That which apprehends sound is the auditory organ.
It is nothing else but ether limited by the tympanum. That
it belongs to the substance ether is inferred from the fact of its
possessing sound. That it has sound is proved by the fact that
it apprehends only sound out of the five qualities colour etc.
( The general rule as already mentioned under smell is that )
that sense-organ which apprehends any one out of the five
qualities colour etc is possessed ofthat quality as seen in the
visual and other organs which apprehend colour etc. The
auditory sense apprehends sound and so it is possessed of
sound.
( 80g ) That sense-organ which causes the apprehension of
pleasure, pain etc. is the mind. It is atomic in size and is
situated inside the heart.
( 80h ) What is the proof for the existence of these senseorgans? The answer is inference only. It is like this. The
cognitions of colour etc. are brought about only through
certain instruments, because they are forms of action, like the
action of cutting ( which is brought about by the instrument
axe. These instruments constitute the sense-organs ).

4 . 3 w [81] sm: ^-qrpT: i cf

TranslationObjects are the six categories; substance,

quality, action, generality, speciality and inherence. Even
though S1HM and the rest ( as mentioned in G. N. S, 1*14 vide
p. 4 ) can be included in these six, they are all mentioned
separately with a definite purpose.

r ;_'

173

NotesThese six objects are the six categories of 3>TT^

mentioned in V. S. 1-1-4, which encompass all conceivable
things in the world. Knowledge or jgf[ is only one of the 24
qualities according to him and swn's as causes for knowledge
come under g % The objects of knowledge, srta's, constitute
these six qsjr's together with negation ( a**n^ ) which came
to be added as the seventh one as an after-thought ( vide pp,
26-32 ). Gautama's classification of things into 16 categories
is based n the notion that everything in the world depends
on knowledge and so gw rightly finds first place in his
scheme. Kanda's classification is based on physical considerations and so knowledge and its instruments are accorded
secondary treatment. This fact also keeps the two systems
separate even after the epoch-making attempt of Gangea to
amalgamate the two into one 33hHTC3* That both are complementary to each other is evident from the many borrowings
between them, s u ^ s R in his rqfHnffTO gives expression to
this idea m the following words at the very outset (vide his
comments under 1-1-1 ) :-

:i

Translatin~-{ Objection). "It is useless to mention

and the rest as separate categories. They, all come
or Si$m as the case imay be and so are not distinct from these
two. ( Answer ). This is true. The four branches of learning
[ 5P, ^Tri, ^a^nf and 3?^%^, vide quotation in the introdiictinfrom the'-awrfi, which the ^T^^R has in mind when
he writes this ] are distinct subjecls propounded for the benefit of mankind. H^T^T and the rest are distinct topics
coming under the fourth of these, -3TFftf%3S or ^T^f^. If
these are not-enumerated and explained separately from the
rest this -sn^f^^ft would reduce itself to the science of the
soul3^i5f^T coming under jftjust like the Upanisads.

174

TARKABHS ( <ra*TRT )

Therefore sfero and the rest are explained separately in this

atra." Keava Mira's cryptic remark
is evidently prompted by the above words of the
l ? 5[s*r [82] cf^r
TranslationOf these substance is that which is the inherent or material cause ( of all things ). It can also be
defined as the sub-stratum of qualities. It consists of these
nine only :( 1 )'sfaeft = earth; ( 2 ) snq: = waters; ( 3 ) ttaqx
light; ( 4 ) ^^: = air; ( 5 ) snwfl: = ether; ( 6 ) m\i = time; { 7 )
fi[ep= space; ( 8 ) ar^Ff^asoul and( 9 ) 3R^=mind.
NotesThe two definitions of ^"5^ have been discussed
earlier in paras 13 and 14 when explaining
1. sflFEft [ 8 3 ]

^ I

fcT

TranslationOut of these nine earth is that which possesses

the generality *earth-ness\ It consists of a peculiar aggregate
of parts, sometimes hard, soft etc. It exists in the forms of nose
( sense*organ-.),' body, and such things as clay, stone, trees etc.
it possesses the 14 qualities; colour, taste, smell, touch,
number, dimension, separateness, conjunction, disjunction,
remoteness, proximity, weight, fluidity and tendency. It is of
two kinds, eternal and non-eternal. The eternal variety consists, of atoms and the non-eternal is of the form of various
products* In both kinds the qualities colour, taste, smell and
touch are non-etenaal and caused by baking ( tjra ). qT^ is
due to.application of heat. Through this the former colour etc.
ill earth are destroyed and (in their places ) new ones are
created.

175

NotesFor the definition through the generality 'earth-ness'

see the remarks under para ( 78a ) Another definition more
commonly used is 'that which has smell', n ^ c f t sfMt. The
suffix ^ or 3 ^ in such cases shows that the quality inheres in
the substance. All important points about the Vaiesika
categories have already been explained in Part I vide pp
26-32.
By baking, yffi, change occurs only in the four qualities,
colour, taste, smell and touch and that too only in the substance earth. This is called wsftqfr and was originally
propounded by the Vaiesikas and later accepted by the Naiyyikas with some difference in the exposition of the process.
The Vaiesikas hold that when a raw pot is placed in the kiln
for baking the heat causes the whole pot to disintegrate to
its ultimate parts, the atoms, which then lose their original
colour, taste etc Again due to heat new colour etc. originate
in the atoms which in turn join up to form the new baked pot
with changed colour, taste etc. Because the change, according
to them* takes place not on the pot as a whole, but from the
ultimate atoms ( qt<g > the Vaiesika theory is called qteprasngj.
The Naiyyikas on the other hand say that the composite
whole is not disintegrated but only the original colour etc. are
replaced by new ones due to the special type of heating
fsnciteiftaTand there is no reason to suppose that the
whole disintegrates and that the atoms begin afresh to make
the baked pot. Moreover on taking out the pot from the kiln
we get the impression that this is the same one that was kept
in it for baking. Their theory is named fqs*qt^m$ (fire* meaning here the whole lot).
2. srqs [84] src^gTTfpq-p^T sjrqr: I

Translation-Water is that which has the generality 'waterness9. [ Another definition is tf ^TOfcfiur ST^that which has
cool touch ]. It consists of the sense of taste, water-body,
rivers ceas, snow, haily^etc. It also has the 14' Qualities

176

TARKABHS:(

mentioned for earth, but excluding smell and including viscidity ( ?5^ ). It is also of two kinds, eternal and non-eternal.
Colour etc, abiding in eternal water atoms are eternal and
those in other waters non-eternal.

3. sr: [ 8 5 ] N ^ ^ p n ^ P ^ r - 1

TranslationLight is that^which has the generality^ Mi'ghtB'CQsists of the^visual sese-brgn,light-body, the sun>
, fire, lightning and similar things. It has eleven qualities:
dimension^ separateoess, conjunction,
ipniiafflity, liquidity and tendency (the
Variety calledI^TT or velocity )i Its division into eternal and
non-eternal is similar to that of water. It is of four kinds :
( 1 ) light in which both colour and touch are manifested; ( 2 )
that in which both are oinmanifested; ( 3 ) that in which colour
is unmaiaifsted and touch is manifested; and ( 4 ) that in
which colour is manifested while touch is unmanifested. The
first variety ( both manifested ) is seen in such things as light
of the sun, masses of fire, etc. Gold has its colour ad ibuch
manifested but suppressed. Its colour and touch are not of
the unmaifstd type. If its colour is unmanifested it would
not be perceived by the eye. If its touch is unmaifested it
would not be felt by the skin. Suppression in this case is
brought about by the homogeneous colour and touch of efarth
particles mixed up with it. Light in which both colour and
toych are unmanifested occurs in the visual sense-orga.B. Light

in which the first is un-manifested and the second manifested,

occursfin the flight* (=heat ) in heated water. Light of the,
fourt^ variety is that seen in the halo around a brightly
burning lamp ( and similar things ).
NotesA "qualitative definition of light is y
q
The classification given above (i.e. four varieties ) is according to the rSTTsmFS. All the printed editions except that f
have another sentence in the text after f^TV
j , erhpr, S^^^'but there is no connection1
for it with either the preceding or succeeding sentences. Ch :
Bh : s comments here are clear,
ft

er ^

^ ^ s w i sjfftw i

r3*sfcr, *r rs^ Hf' After this e He!

introduces the four-fold division a i i the l text here

ir '
Xhe Naiyyikas bring out a negative inference to establish
the inclusion of gold under this category 'light', g^uf. afa^>
s:-, i.e. gold is a form of srr , because when there
is no counteracting influence ( or impediment ) it does not
lose its liquidity even under the application of intense heat;
that which has not this | g does not belong to fof ( iffEwf). To account for its yellow colour; cold ouch, and
weight it is postulated that these are due to the earth particles
adhering to it, which suppress its natural colour ad touch
( bright white and hot touch ) and also lend it weight. A;
modern student of science will hardly agree to this. The
Mmmsaks hold that gold and other metals quarried from
mines ( IT^5I ) constitute a separate substance other than
these nine. ; The more modern among the Naiyyikas ( i.e;
the followers of ^?rm%?tiif&T of Nadia ) include gold and
:
other metals under'earth'.

t 86 ! /
I if
12

178

TARKABHS (

TranslationAir is that substance which is connected with

^air-ness9 ( i.e. which possesses this generality ). It consists
of the tactile sense-organ, air-body, wind, breath of animals,
e t c It has the nine qualities, touch, number, dimension,
separateness, conjunction, disjunction, remoteness, proximity
and velocity ( a variety of STWT ). Air is to be inferred from
its touch. It is like this. When wind blows a touch which is
neither hot nor cold is felt. This touch being a quality cannot
have an existence separate from the thing it qualifies and so
causes the qualified thing to be inferred. This qualified thing
is the substance air, as earth and other substances do not
become this substratum. Moreover this touch which is
neither hot nor cold does not abide in substances other than
earth and air. Air is of two kinds, eternal and non-eternal;
the former consists of air-atoms and the latter the products
( wind, breath etc ).
Notes-Though there is a slight change in the wording of
tne definition of air, it is the same as saying gg^HWTO^
M\Q: as in the previous definitions. Another useful definition
is ^q*ff ^TO^rig; eng:. Because it is not perceptible by the
eye the attribute ^q<Tf?3 is used.
There is some difference of opinion between the ancient
and modern Naiyyikas as to whether air is perceptible or
not. The ancients say that air is not perceptible even by the
tactile sense, but can only be inferred. This inference is
what is explained in the text. The controversy centres round
the question whether s^f^q should be a necessary condition
for all src^r^r or only for ^T^qssrg (vide p. 50). The
inoderners hold the latter view saying that only s^cPEq^r is
necessary for tactile perception and not s ^ ^ q . Our author
apparently holds the ancient view and so gives ah inference
to establish it, similar to that in the case of 3TSR5^
These four substances have some sort of similarity in
their being eternal aod non-eteraal, the former Consisting of

atoms and the latter of the products. The non-eternal variety

of each is again classified as *rfta ff*3*T and f^T^r^r. Keava
Misra has not expressly stated this division but has hinted
it when he mentions what each substance consists of. It is
only these four that are capable of being produced Ost?** or w )
and so the author discusses in the next two paras the process
of their production and destruction-

[87]

TranslationNow the processes of production and destruction of the four producible substances beginning with earth
are being described. When two atoms combine as a result of
motion in them a dyad is formed. The two atoms constitute
its inherent cause; their conjunction the non-inherent cause
and cfate' etc. the general cause. Again due to motion in
them three dyads join together to form a triad, The dyads
constitute its inherent cause and the other two causes are the
same as in the case of the dyad. Similarly from the triads
a tertiary is produced; from the tertiaries larger things and
from these still larger ones are produced. In this manner
are produced the great 'earth', the great'waters', the great
""light' and the great 'air'. The colour and other qualities in
these products are produced from the colour etc. of their
inherent causes which constitute their substrata, according
to the general rule that *the qualities in the inherent cause
generate the qualities in the product/

TARKABHASA (

frf5i; [88]

: 1

^^i

TranslationWhen a product like a jar etc. is thus produced

with colour and other qualities, and when motion starts in
the component parts of the jar etc as a result of a jolt or a
blow, this motion causes the disjunction of the component
parts, which in its turn destroys the conjunction of the parts,
which is the nonrinherent cause of the composite whole. Thus
4? caused the destruction f t t e product, the composite jar
etc. From this it is seen how a substance is destroyed when
its non-inherent cause is destroyed. In some cases the
destruction of a substance is' caused by that of its inherent
cause. For instance when the Almighty desires the dissolution of the aforesaid earth etc., this desire brings about motion
in the atoms comprising the dyads, which in its turn causes
the disjunction of the atoms. Due to'this the conjunction of
the two atoms is lost and so the dyad is destroyed. With the
destruction of its substratum, the dyads, the triad is destroyed
and in this order the earth and other substances get destroyed.
Another instance is seen in the destruction of the cloth when
the threads get destroyed. The destruction of colour and
other qualities of tlie products is brought about by that of
their substrata. In some cases such destruction ( f qualities )
takes place even when the substratum is there, due to the
appearance of opposing qualities as in the1 case of jar etc*

during baking ( when the original qualities are replaced by

new ones ).
NotesIn para 87 the evolution of things, from atoms is
described. While a dyad is formed out of two atoms a triad
requires three dyads as its components and not two. Why it is
so will be explained in the subseqaent paragraphs after the
existence of atoms is established. In the text of para 87 all
the printed editions read ^ ^%% 3gfrag*tg$r; but the word
^gin seems to be out of place ( even wrong ) for the reason
to be stated by the author himself in para 90 below.
In para 88 the dissolution or destruction of produced things
is explained in two ways, by the destruction of either the noninherent cause or of the inherent one. The first view is that
of the modern Naiyyikas and the second that of the older
school. According to the latter, the ultimate atoms get disjointed firs and due to this the dyad gets destroyed. Only
in this case the destruction of the non-inherent cause ( 3T<gg*TNt*T vide para 87 ) acts as cause for that of the effect ( the
dyad here ). From the triad onwards destruction is brought
about by that of the inherent cause which is invariably the
component parts. The moderners for the sake of simplicity
( 3>sre ) want one uniform rule for all and say that the destruction of the non-inherent cause leads to that of the effect as
is seen in the case of a jarw^srsTta^nst ^WT^T. Either
way the view is based on the assumption that destruction
proceeds from cause to effect. While refuting this N. V, view
in his commentary on sffq^ II314 the great aankarcrya
gives an apt comparison by saying that it is absurd and
impossible to pulldown the foundation of a building without
removing the roof,*walls and other overstructures.
Next the author explains the conception of atoms or the
atomic theory of the ancients which was first propounded in
the V. S.

; [89] f% 5 ^ qrq-prngf

SPTFFC

gai

182

TARKABHS ( w

sfir STTR

TranslationWhat is the authority for (accepting) the

existence of atoms ? This is being explained. That very
minute particle of dust seen ( floating ) in a ray of sunlight that trickles through a slit or hole, is made up of
substances with smaller dimension, because it is a produced
thing like a jar. That substance ( with smaller dimension )
is also a product ( consisting of component parts ) in
accordance with the general rule that 'a thing which produces
another of greater magnitude should itself be a product'
This establishes the thing called dyad. This dyad is also
made up of things having smaller dimension which constitute
its inherent cause, because it is a product like a jar. That
which makes up this dyad is what is called an atom; and it i&
not made up of further component parts. ( Objection ). Having
accepted the general rule that a thing which produces another
must itself be a product, how is it that the atom ( which
produces the dyad ) is itself not composed of smaller parts ?
( Answer ). It is liket his. If we accept the objector's view
it would lead to an endless series of effects and causes. If
this latter position were to be agreed to it will lead to the
( absurd ) conclusion that the ( great ) mountain Meru and
the ( minute ) mustard seed have equal dimensions ( and so are
equal in size ), because both are made up of an endless
number of component parts. Therefore it has to be accepted
that the atom is not composed of still smaller parts.
NotesThe course of reasoning behind the N. V. atomic
hypothesis may be summarised as follows. Firstly from actual
observation it is seen that for visual perception substances
must have a minimum dimension ( qfno ) called *f[^
(largeness ). This sjissgr and its counterpart 3T<g<3( ( smallness )
vary between two extreme limits, the highest and the lowest

The highest limit is q?iwc*9 ascribed to all-pervading substances

such as ether, time, etc. The minimum w^qr necessary to
make a thing visible is that observed in the dust particle in a
sun-beam and so this particle is assumed to be the smallest
thing, visible to the naked eye. Similarly the lowest limit of
smallness ( srojgf ) is qTOigKi also named as qTWT^T. It is
also an established fact that all products are composed of
component parts ( ro*rc ). With these premises it is first
inferred that the sun-beam-dust-particle is composed of smaller
component parts. The syllogism stated in the text can be
expressed in another form as welln^^nrfc**} ^
^TS^raU * z ^ * This particle is named as grg or
and its component parts are called gops or dyad. This dyad
is not visible as it lacks H%^ and so has only srjg^. All the
same it is capable of .producing a substance with ivgm, the
s^tg^s mentioned above. On. this basis it is again inferred
that the dyad is also made up of component parts, just like
the two halves of a jar or the threads of a cloth. [In the
second inference the example given in the text is -9333. It
would have been more apt and direct if it had been mentioned
as 3qrT&33(J. The component parts of a dyad being established
in this manner, they are named <*$ng with the lowest limit of
3rprifwT^r. The same reasoning used in the case of the
dyad can be extended to this atom also, but then it will lead
to endless regression leading to no definite conclusion. The
extreme lower limit of dimension can never be reached with
the result that the dimensions of the products will become
incomparable. This will lead to equating a mountain with a
tiny mustard sedin size as both are made up of innumerable
component parts and there is no standard for measuring
dimension. So a limit is placed at the component parts of the
dyad, the atoms, as it meets with the logical needs of the
Naiyyikas. So the atoms are not divisible into smaller com-*
ponent parts and for that reason eternal. Even in modern
science matter is indestructible and can only be transformed.
This raises another question which is being discussed in the
next para.

TARKABHAS ( <s4

TranslationThe dyad is made up of two atoms only as a

single one cannot begin a product and there are no grounds
to postulate more than two. A triad however is composed of
three dyads as one is unproductive, ( as mentioned above ) and
two do not account for the greater dimension T?<3T seen in
it. This is because of the general rule that the greater dimension in the product is caused either ( 1 ) by such a dimension
in its component parts, the inherent cause, or ( 2 ) by the
:
greater number of such parts. In the case of the triad the
first cause is impossible (as the dyads have only <g^ and
not Tf-cof ) and therefore the second cause, ( greater number
of the parts), should be resorted to. Since three dyads
are considered sufficient to produce this greater dimension
(in the triad) there are no grounds to stipulate more than
three.*

' "

"

"

. *

'

' '

'

Notesn para 87 it has been mentioned that two atoms

join together to form a dyad and three dyads are required to
make a triad. The question naturally arises how this number
three has b^en arrived and why not two. In this connection
the general rule mentioned in that para^RorgoTT: ^mgoiTmay be recalled. According to this the dimension
in the triad is caused by the dimension in its inherent
cause, the dyad. But thp latter has only $T<g<rf*H<n and it
can transmit to the triad only that dimension which may be
of a higher degree as per the rule qftTnoTRt ^e^Tc^T^t^S^f^'HTCWB^f'W. The application of this rule would lead
to the result that the dimensions in dyads and triads should
bej of a higher degree of arp* ( i.e. they should be further
smaller than the atom ). This is contrary to experience
* There is an oft-quoted ar/cient verse in this connection :>

185
s we see the triad as having vt%m. To avoid this discrepancy
the dyad and triad are exempted from the above rule and in
their two cases the number of the component parts as the
cause for their dimensions. So t account for the difference
in the dimensions of a dyad and a triad there should be a
greater number of component parts in the triad than in the
dyad. Since the number of parts in a dyad is two, the next
higher number three is postulated as the minimum number of
parts to produce T(%m in the triad. As three satisfies the
logical requirement in this case there is no necessity to think
of a higher number. In the case of products greater than
the triad the above mentioned rule ( qftTT^^T^ etc.) can
apply as it is and there is no necessity to stipulate or restrict
the number of the component parts. ( Cf. ^g^T^^qf^ fsnwi^r i ^-TORfRwOT^W-Ch. Bh.,). With this reasoning it
will be found that the addition of the word ^rgfr: in the
sentence '^i^nj^.^a^froPfl' in para 87 is not warranted (as
already mentioned in the notes under that para ).

destara; i

TranslationEthQT is the substance which has sound as its

distinctive quality. The qualities sound, number, dimension,
separatenessj conjunction and disjunction abide in it. It is
one, all-pervading, eternal and inferrable from the reason of

186

TARKABHAS ( tv

having sound. It is inferred from this reason by a type of

inference alled qiffo, a kind of elimination. It is like this.
Sound is a special quality, because while having a generality
it is capable of being grasped by a single external sense-organ
existing i living beings like us, as in the case of colour etc.
Being a quality it must subsist in a qualified substance. This
substance cannot be anyone out of earth, water, light, air, and
soul as sound is grasped by the ear. The qualities abiding in
these five are not cognised by the ear, as is the case with
colour etc. and sound is grasped by the ear. Nor is it a
quality of time, space or mind as sound is a special quality.
For these reasons a qualified substance other than thse eight ,
should be accepted as the substratum of sound. This substance
is what is called ether.. It is only one as there is no proof to
suppose different varieties of it. Also a single ether satisfies
all ( logical ) purposes. Being one there is no generality,
srfcTsr?^, in it as any generality will have to be observed as
existing in many similar things. It is all-pervading, which
means that it has the greatest dimension, because its effect is
felt everywhere* On that account, i.e. being all-pervading,
it is eternal also.
Notesqfkstm*R\7t is_explained by Ch. Bh. as
simsr^Rt qffrssOT ew^W ^ftsfar:.* When the existence of
something is suspected in a number of things and when it is
proved that it could not exist in some of these, the knowledge
that it should exist in the remaining thing (or things), constitutes qfosfa. It is a process of elimination used earlier also,
in the case of STTS?^ and srig. 3 ^ as a gor must exist i one
or more of the nine ^s^r's. Having established this search for
its substratum begins; and for reasons stated in the text the
eight ^"sq's, earth, etc, are eliminated. Thus the remaining
one, ether, is determined as its substratum. That grsrcRar
cannot be a generality, iR^, has been explained under para
11, p. 18.

_LjJ

__

, , '

* A-editions'have this sentence as part of K. Ms7 text; but B rightly

omits it.

.'....

'

' .< "7- .

187

6. R'[92]

Translation-Time ( as a substance ) is to be inferred from

( the notions of) remoteness and proximity which are contrary
to those created by space. It has the qualities number,
dimension, separateness, conjunction and disjunction ( besides
remoteness and proximity ). It is one, eternal and all-pervading. How it is to be inferred from the notions of remoteness
and proximity contrary to those of space, is being explained.
When an old man is standing near us we ought to get the
notion of his proximity; but actually the notion of remoteness
( in age ) as opposed to proximity ( in space ) is what arises
in our minds. Similarly a youth at a far-off place creates the
contrary notion of proximity ( in age ). These contrary
notions of remoteness and proximity should be caused by
something other than space, and thus lead us to infer the
substance time as its cause. Though time is one it is spoken
of as present, past and future due to the limitations ( or
circumstances ) caused by the actions in these three stages of
time, just as the same person is called a cook when he cooks,
a reader when he reads and some other when he is engaged in
other actions. That time is eternal and all-pervading is to
be understood as in the case of ether.

7:, f%\ [93] ^rafq-q-^q'^^Tsq'^^^T fas \

18

TARKABHASA ( a &

. TranslationSpace ( as a substance ) is tobe inferred from

the notions of remoteness and proximity contrary to those
caused by time. It is one, eternal and all-pervading. It has
the qualities number, dimension, separateness, conjunction and
disjunction ( besides remoteness and proximity ). It is inferred
from the notions of the directions, east, west, etc; because such
notions cannot be caused by any other substance. This is
gathered from the fact that a thing situated in any direction
eastj west etc. ( with respect to some fixed entity ) remains
in the same direction ( without change of direction ). Though
only one it is spoken of as east, west, etc, due to the limitation
caused by the contact of the sun with those spaces ( i.e. the
notion of east, west, etc. at any place is determined by the
apparent position of the sun at that place ).
NotesTime and space are defined in a more useful way in
the B. P. of f

snro* *rat surawTsnt TRT: I i45(b)

faun % * q a ii 46
f
47(a)
The inferences mentioned in the text for time and space
may be stated thus i^fe^lR^ qRc3W<3 3rcffraifreii?<7fa?$,
^^*T4?CT^ , f^i. This non-inherent cause can be found
by a process of elimination ( ^fk?^^^ ) as nothing but 35T35-f^F
iqa^ i^riT as only a quality can be a non-inherent cause.
Apart from the notions of old and young ox past, present etc.,
in the case of time, and of distance and nearness or of east,
west and other directions, in the case of space, these two
are not capable of producing anything, though they are causes
in general for all sorts of products. Another subtle, yet
important, difference between the two is stated as follows :
fwaqs^TO: 3>T3s: \ srfrtgTi^fr^T ^ J ^ means that
the relation of time with respect to an event or object is
always constant while that of space is varying. A thing of
the past or future time remains the same for all at present,
while the direction of a place or thing varies with the direction
from which it is observed. An example would make the idea
clearer. The Nirvana of Buddha happened 25 centuries ago.

189
is "event relates to past time and will remain so for ever.
That a solar eclipse will occur in February 1999 will remain a
future event till that date. But with regard to space it is not
so fixed. The Himalayas are to the north of India but south
of Tibet, east of Iran but west of Hong Kong; i.e. the direction
varies with the direction of the observer. A thing to one's
right or left can change to left or right when the thing or the
observer changes position. Another point of difference
between the two is that while divisions of time are determined
by the evolution or dissolution of things, those of space are
determined by the greater or smaller number of contacts with
objects that intervene between two spots. This is expressed
in Sanskrit as 3?afl7r 9T3frnfa:.l. iffitt ffrifa-' I This again
is due to the fact that tee divisions of time are based on the
movements of the sun while those of space have no such
connection. It will be of interest to note that '3ftT*ft??iftr
of Nadia tried to explain time and space as only certain
phases of the Omnipotence of *%* or snw^ ; but this has not
found favour with the later Naiyyikas.
.

8. mm^ [91] sn^W^q^HT^T

TranslationSoul is that which has the generality soul-ness.

Due to the differences in the experience of pleasure, pain etc.
it is different from body to body. It has been explained
earlier. It has five general qualities beginning with number
and nine special ones beginning with knowledge. Its eternality and omnipresence are similar to the previous substances, ( ether, time and space ).
Notes^^sTcf^ has already been explained as the first gifcr
under para 77. Why this and many other things are repeated
under different headings has beqn explained by the author
himself in the last para of the book.

9. R! [95]

190

TARKABHAS (

i arfa ^ 3 ^ * 1 ^ i gwrf|[ i
: i

fait

ft

Translation-Mind is that substance which has the generality mind-ness. It is atomic (in size ) and is always connected
with the soul ( except when in sleep ). It is an internal senseorgan and is the instrument for the cognition of pleasure,
pain, etc; also eternal. It has the eight qualities beginning
with number (dimension, separateness, conjunction, disjunction, remoteness, proximity and velocity ). Through contact
with it the external sense-organs produce the cognition of
objects and therefore mind i s a ( general ) cause for all cognitions. It is not perceptible but only inferrable. The syllogism
is like this. The cognitions of pleasure, pain, etc. are caused
hy an instrument other than the external sense-organs, eye,
etc., just like the act of cooking takes place without an axe
but, with other causes like fire etc. /This instrument is what is
called mind. Its dimension is sugser. Thus the substances
have been explained.
NotesMind has already been explained once under
para 84, and wiil again come to notice as the sixth
in para 121. The author's explanation at the end of the book
that extremely useful topics have been described over again
under different headings, is more a request to the reader to
put up with such repetitions than a convincing answer. In para
84 the emphasis is on mind's function as a sp*ii for the cognition of pleasure, pain and the other special qualities of srrar.
So it has been defined as ^raq^fsqrereraff^ *W- Because
Sf&ffi has been defined as that cognition arising from senseobject-contact and since the cognition of pleasure etc. comes
under zim^ mind is postulated as the internal organ for such
cognitions. Here the author wants to refute a section of the
Naiyykas who do not accept TR^cf as a STOT^ and so
defines mind through-tffiCKf, meaning thereby that it is as valid
a generality as ] any; and uses the abovermentioned causal

191
function as the reason for the inference to establish the
separate existence of mind as a substance.
Another important function of mind is that it acts as the
intermediary or connecting link between the external senses
and the soul in all external perceptions ( cf. accrotre etc. in
the text. ). Kanada uses this property for the proof of mind
when he says Wcf^TOefcrwr 5 W T *Trt5vn^t RSt f>f fl(
( V. S. III. 2-1 ). From experience it is seenfthat only when
mind acts as a link between the soul and the external senseorgan, cognitions arise and otherwise not. So even when
$oul-sense-object contact is there the presence or absence of
cognition is what accounts for the existence of mind.
This raises the question of dimension of the mind. Ordinarily cognition comes in only one after the other and not
^imultjaneously. This fact has led the Naiyyikas to stipulate
that mind is atomic in dimension ( arg ), as against the Bhtta
Mimmsaka view that it is all-pervading ( f%g ). Gautama
uses this fact as the proof for mind as he says ^Tqs^rrTgqfr^frtft fs^r ( N. S. 1-1-16 ). According to the Naiyyikas,
contact between two all-pervading substances is not admitted.
Moreover if the mind were also f%jj there would always be
the contact of the external senses with it'and all kinds of
perceptions would occur simultaneously, and also there would
l>e no sleep ( vide the notes below). 'The Bhttas contend
that frggTffcta is not impossible ( as in the case of time and
space ), that snororata operates only within the sphere of the
body, that several cognitions do not occur simultaneously is
not proved conclusively, and that several impeding reasons
can be advanced for that fact. The Prbhkaras accept the
Nyya view that mind is an eternal atomic 3*sr but do not
agree that saw^is perceived by the mind. The Advaitins hold
that mind is only a modification ( ff% ) of the w^n^ui which
is their inner instrument of knowledge. All these differences
are partly due to the conception of STTW^ ( and also of knowledge ) which vary widely with the different schools (vide
a otes under pars. 76 ).
,
It was mentioned in the previous para that if mind were
also to be f%g there would be no sleep. This;leads us to the

192
Nyya theory of sleep,which is - stated briefly as follows :;
l" ( Dinakar com :
on S. M;Choukhamba Edn.,,p 184 ). At first due to some motion
in the mind, it gets disjointed with the soul, by whiclj it loses
contact with the soul. Then it enters the place called gita^, a
narrow vein or entrai! near the heart, This entrance of the mind
into jccj;is sleep when all,its activity ceases. This notion is
perhaps borrowed by the, Naiyyikas from a statement in
j?>(IV-lf-19).This-g(^nfTsays that during sleepy leaves
the heart and enters gfte^ through the 72,000 arteries leading
to it. Being g% it has to be accepted; but in interpretation
there is wide variation,. The Advaitin says that it refers to
sftgr and the Naiyyika says //means icrto suit his doctrines.
However that be, the flaw in the Nyya theory-is that sou],
being f^r^g it cannot lose contact with the mind even in the
^y^-.yem.^' To overcome this later authors have proposed
"that contact of mind with the tactile-sense is necessary for all
cognitions/* V^ci^ Ka^ r np skin and so no cognitions arise
during sleep. This appear
as the previous
one of mind osing contact with the soul. It, would therefore
be better to explain! the phenomena of sleep and perception
by conjunction and disjunction of themind with the external
sense organs alone.

2. 301: [96]

TranslationNow the qualities are being described. Quality

is that which has generality, which serves as non-inherent
cause and which has no motion. It always abides in a
Vide S.fafc under ^if^tf 57.

193
substance. They are only twentyfour in number such as? colour,,
taste, etc. ( vide pp."27-28 ).
_, \
NotesThe three terms in the definition of gar are intended
to exclude overlapping ( sTfas^Tifr ) in ( 1 ) **n?r and the rest,
( 2 ) ^s*r and ( 3 ) ^respectively. This is also expressed in

L ^<rq [97] ^ r M

^fffafft
; I

TranslationQt
these 24, colour is that special quality
which is cognised by the eye alone. It abides in earth, water
and light. In earth it is of various kinds like white, black, e t c v
brought about by baking a n d so noo-eteroal. I n water atoms
and light atoms it is eternal a n d in their products non-eternal.
The colour in light is dazzling white a n d that in water n o o dazzling white. In both it is n o t due t o baking,
Y\
Notes-~ln the above definition the word TTi^f is necessary
to exclude fluidity ( ^ ^ ) which can be cognised by the tactile
sense also. The word Qn is intended to avoid sr*n which i s '
considered a ^-q, as also ^:qr^ which is a *TTfrr?3T, which t w o '
are grasped by eye alone. The word fgrta as attribute t o go*
is to avoid overlapping in H^T, Hfi, etc. which are only
^TO??gaT's. Even then t h e definition does not cover *wrigF;<79
and so the definition should be expanded in some such manner
as ^H37U5fH^. T h e colour in atoms, though n o t visible
being ' srgSsr, does have ^qegrsnRf. Still the definition would
o v e r l a p g^oT which comes under ^ as explained earlier. So
should be further expanded as siff^fiT^pT^ orgoift!T3^. T h e reader m a y note the ingenius methods employed by
Naiyyikas to perfect their definitions.
( 15th century ) in his commentary ^ q ^ T T on V. S.
states four conditions as necessary for the perception of
colour, vide Stra IV-1-8. ( 1 ) Hf^q^rror ( or &53rcraacR3>
i.e. the dimension n^gf in t h e substratum. T h e colour of
13

194

TARKABHS ( <n

atoms and dyads are not visible for want of this condition,
{ 2 ) 9?^r-or manifestedness. The colour of the visual senseorgan which comes under % ^ i s not visible as it is not manifest. (3) 3T?rfiT3Jpc3 = non-suppression by any sqifo. The natural
colour of %$f*un gold, gems etc.3WT^rg5f35is not grasped
as it is rfoqtTit vide para 85 j . ( 4 ) qKCsnfr i.e. the generality
colour-ness. If this is lacking 1W%> T ^ ^ , etc, will be liable
to be grasped by the eye. K. M. also mentions the first three
in para 100 below
Colour is of varioqs kinds. But later authors like sra^Hg
and fawira divide it into seven types :( 1 ) gpres = white; ( 2 )
blue or black; ( 3 ) qtef = yellow; ( 4 ) ^ = red; ( 5 )
green; ( 6) ^fq^r = grey; and ( 7 ) frer = variegated or
mixed. This division and the discussion on fi3^q is not
found in the V. S. or the N. S. nor in their w q ' s . Our
author has chosen to follow the *3TC in not mentioning
these divisions., All the same the discussion about the necessity
for a frfiq will be of interest. The T. D. of An. Bh.
States it briefly as follows i^

This is explained below :ObjectionInstead of postulating a separate colour as

f^r5Uq why should it not be taken as a mixture
of various colours ?
AnswerThat cannot be for colour is a
( one which covers the whole substratum ) and
you cannot say that all the various colours in
the mixture cover the whole cloth at the same
time.
defined as 5Esre?TT^lf^^'ni^'cITHmT5r%^nt, i.e.,
an attribute which does not co-exist with its own negation.
When a book is on the table only that side of the book which
is in contact with the table has chfr. The other parts of the
book and the table are not in contact with each other. Here)
therefore #t*T and sttirmTcc both co-exist together in the table

195

and the book. So gcHKKrta is an affOTorfftanf. Similarly

^sftteht is another example. Colour is not like this as it must
-occupy the whole of its substratum.
Objection Even if colour is s ^ q f f o what harm is there
if we say that in a variegated piece of cloth
all the colours in its component parts ( threads )
are seen ?
In that case there is no colour as such in that
Answer
cloth and so it will become imperceptible ( as
manifest colour is a cause for visual perception ).
Objection That can be overcome by postulating
*rcK3 (the fact of being inherent in a coloured
object) as the condition for perception
instead of ^ W H ( the fact of possessing
colour )
Answer This is certainly more elaborate and cumbersome than accepting ft^q for the perceptibility of a variegated cloth.
Here the Naiyyika doctrine that, while a product
^
is distinct from its component parts ( srerq^'s ), a collection
( ^g^ra ) is not different from the parts that make it up, is
alluded to. The grouping of qualities into fitaand W has
already been mentioned under para 16.

2. *e$ [98]
i ra1
F, TRATST I 3RJ
TranslationTaste is that special quality which is cognised
by the tongue. It abides in earth and water. In earth there
are all the six varieties, sweet, sour, salt, pungent, astringent
and bitter, which are caused by baking. In water there is only
sweet taste, which is not due to baking; but which is eternal
in water atoms aod non-sternal in products.

196
3. I5ST5

TARKABHASA (
[99]

TranslationSmell is the special quality cognised by the

npse. It abides only in earth and is non-eternal. It is of two
kinds, fragrant and non-fragrant. The smell that appears to come
from water, air, etc. is only through the relation inherence in
the contacted thing5, i.e. inherent in the earthy matter with
which water etc. are mixed up.

TranslationTouch is the special quality cognised by the

tactile sense alone. It subsists in all the four substances
earth, water, light and air- It is of three kinds, cold? hot, and
tepid ( neither hot nor cold ). Cold touch abides in water*
hot in light, and tepid in earth and air. That in earth is
caused by baking, but not so in air. It is non-eternal in earth,
but eternal in the atoms of the other three and non-eternal in
their products. These four qualities, colour, taste, smell and
touch, become perceptible only when they are manifested and
when they inhere in substances with ^ ^ dimension.
NotesThe definitions of the above three are to be understood in the same manner as explained under colour. In the
case of touch the word m^ is necessary as there are o|her
qualities cognisable by the tactile sense such as fuidity, contact,
number etc, On the analogy of f ^ ^ q there is no necessity to
accepta* feftTO or fp^w in objects having different TO'S
and q^T's as there is no harm in attributing them to their
sfRi's. ^Tfftfar ( quoted under ^<j ) however is of the
opinion that a f%^^RT is necessary on the analogy of f%^^qs.
( Cf c ' i w i ^ ^ 3 ^ q ^ ^ ^ 3 ^ T ^sn^^oii^r ^ \ ); while he

197t
denies a ffsRtf, Cf. "* ^ f ft<mi TStsfa fa*
fflffm* *TO t ^ r a i W * " (under V. S. VIM-6). This is
because of the view that air is only inferrable and not perceptible even by the tactile sense ( v*de para 86 ).
5. retR [101] 3PTT ^ ^ T f 3 ^ ^ R | 3 : ,

TranslationNumber is a common quality which serves as

the cause for expressing ( the notions of )wone, two, etc. It
extends from one to q*ri ( a hundred thousand billion = 1017 ).
Unity is of two kinds, eternal and non-eternal; that in eternal
things is eternal and that in products non-eternal. The latter
is produced by the number unity in its substratum which is its
inherent cause. Numbers two and above are always noneternal. The notion of duality is caused by a particular conception of the form 'this is one, this is one,' ( called r^T^fO
that arises with reference to any two objects. In this case the
two objects constitute the inherent cause, the unities in the
two objects ( separately ) the non-inherent cause and the
particular conception ( ar^Tf fi? ) itself is the general cause.
The loss of the notion of duality is caused by the loss of this
sf^Tf%. The origin of the notion of numbers three and
above is to be understood in the same way ( as in the case of
number two ).
NotesThe word | t j in the definition of number is to be
^understood as a*nWofH^?<, so as to exclude such
common ( ^nsTRiur ) causes as time, space, etc. q*TW means
half of Brahmas' after a ageHe is termed q*p=^5ftf^
which works up to the figure 432 x 1015. In Hindu astronomy
q*m is used for a number of 1$ digits which was considered
to be the highest number amenable for their manipulations.

It should be noted that in the N. V. system the words q^cf,.

fr, etc, denote the numbers one, two, three, etc. and thato

198

TARKABHS ( w&

refer to the substances they qualify as

m^: etc. In the case of sntogf, t^s* denotes
the relation of being the object of that cognitionat^Tgf^$*3 in certain cases as in expressions like
denotes a negative idea as 'not having a second
thing of the same kind'^escn^cn^ft^. Th notion of
unity is restricted to the substratum to which it relates, whereas
the notions of duality and of higher numbers refer partially
to the individual entities in the group and completely to only
the groups of two etc. This is stated in Sanskrit as
a

special type of conception that occurs on

seeing two or more things, as 'this is one, this is one; so there
are two'Cf. '&esf fprf s r ^ s f ^ I ( B. P. V. 108 ).
The notion of numbers two and above are caused by this
conception which may therefore be considered a notion of
addition. The origin of the notion of duality is a complicated process and takes seven moments of time as explained
by 3if *fosr in his-3<re3rc under V. S. VII-2-8. This is summarised below. ( 1 ) fc?^Twfe$*i: = The sense-organ comes
into contact with the objects; ( 2 ) 3K q^KTOfflFrac: = Then
arises the notion of the generality qsp^cT ( i.e. the
cognition ) in the things separately; ( 3 ) 3tT q^egg
Igf^sThen the notion of unity in each along with
( 4 ) %?5fgotq% s Next the notion of fjp^ as a quality comes
into being; ( 5 ) SRt f5^grgm = Then arises the notion of f&mm
in fiff^^q form; ( 6 ) ^cr ^ i t f g ^ ^ R = Then arises the notion
that the number two subsists in both the things; ( 7 ) 3"a fg^ffr23r = Then the final notion that the objects are two in
number. This lengthy process is considered necessary because
of the general rule that all cognitions last only for two
moments and get lost in the third moment after leaving an
impression ( fcqsrc ). But sr^Tgf% does nofc conform to this
rule, for it would come to an end at the fifth moment ( vide
the summary above ) when fgsr, its effect, has not come intobeing, and the cause can get destroyed only after the effect is
produced. To avoid this difficulty s f o n f e here is allowed
an; extra moment of life. The above process of the .origin of th

notion of duality is found summarised in a memorable verse

in the H^TO of i n o r a r a under the Vaiesika system

This theory of the origin of %<qr, as also that of

( explained earlier ) and that of disjunction caused by another
disjunction (fwT*lfo*FT,vide fw below ) were at one time
considered the most difficult and knotty topics in the Vaiesika
system. Ability to expound these satisfactorily through the
bewildering labyrinth of verbal subtlety was considered the
hall-mark of a Vaiesika. ( vide foot-note on p. 202 ).

6. qfbn<^ [102]

TranslationDimension is the cause for expressing the

notion of measurement. It is of four kinds, atomic or small,
great, short and long. The dimension of products is caused
by ( I ) number or ( 2 ) by dimension, or ( 3 ) by the aggregation of the component parts ( sra? ). For instance the
dimension of a dyad is caused by the number 'two' in its
component parts, the atoms. This duality in the atoms is
caused by the sr^Tg% of the Almighty. Here the word sftfr
means cause. The dimension in the triad is also caused by
the large number in its component parts which constitute its
inherent cause ( the dyads ). The dimension of tertiaries and
higher products is caused by the dimension in their component
parts which constitute their inherent cause. The dimension
of a mass of cotton ( and similar things ) is caused by th<

200

TARKABHS ( gr

loose combination of its component parts which form its

inherent cause- The smallest dimension in the atom ( qfWSW ) and the greatest one ( tR?rc3 ) existing in ether etc.
are eternal.
NotesHere and. in the definitions below the word ^s\m or
%g is to be understood as ^TTT^T^oifTfT^^T^^ as explained
under *N^T. [A. editions read the text as QTTT^ROI^H'^:
omitting fftm. ] The dimensions aropqr and Sf^sr refer to
both area and volume while gwr and ^^m relate to only
linear dimension. Regarding dimensions of dyads, triads, etc
and how number comes in to make up dimension, vide notes
tinder paras 90 and 9.

7. g w [ 1 0 3 ] ^ ^

^ ^ R I ^ i

TranslationSeparateness is the cause for expressing one

thing as separate from another. It is of two kinds, separateness from one thing, and separateness from two or more
things. The first kind is eternal in eternal things and noneter.al in non-eternal things (products ). The second is
always non-eternal.
NotesCSTW is justified as a quality diffrent from i ^ or
^Pt^x^T^r on the subtle ground that it is the object of a
positive cognition while ^ is the object of a negative
cognition. The difference lies in the subtle difference in
meaning betweeri the sentenceis 'a pot is separate from a cloth'
and 'a pot is different from a cloth'. Modern Naiyyikas do
not admit this subtle distinction which, they say, is due to the
context in which such sentences are used and so include
under ih[.

8. frfar: [104]

* The portion i T^5%t""' # ' lt "

margin of the original of my transcipt.

9frf^c^g[' is found added in the

201

TranslationConjunction is the cause for expressing the

notion that two things are in contact. It subsists in two
things and does not cover completely both the substance. It
is of three kinds; ( 1 ) conjunction caused by action in either
of the substrata, ( 2 ) that caused by action in both the
substrata and ( 3 ) that caused by another conjunction. An
example of the first kind is the conjunction of a moving kite
with a stationary pillar, or the conjunction of a tangible substance ( sjcf^sq- ) with any of the all-pervading ones. The
second variety occurs in the conjunction of two wrestlers when
both move to meet each other. The third type occurs when
there is conjunction between an effect and a non-effecfc due to
the conjunction between its cause and a non-cause thing as
for example the conjunction of the body with the tree when
there is conjunction of the hand with the tree.

9. fkm*T- [105] fawft r f^RB5Tc2P%: I

?rr*r: i Refait
TranslationDisjunction is the cause for expressing the
notion that two things are , disjointed. Tt is: also-related to
two substrata and is ( always ) preceded by conjunction. It is
of three kinds : ( I ) that brought about by action in either of
the substrata; ( 2 ) that caused by action, in both; and ( 3 )
that caused by another disjunction. Examples of the three
kinds are respectively ( 1 ) disjunction between the kite and
the hill due to motion in the kite; ( 2 ) that between two
wrestlers; and ( 3 ) that between the body and the tree when
the hand is removed from the tree.
Notes^sft*! though actually a temporary relation, is
considered a quality in the N. V. system and is the fact of two
things being in contact and not the act of joining them
together. K. M. has mentioned three methods of its origin
following the Strakra while other authors like An : Bh :

202

TARKABHAA (

would like to give only two classes as e&ifer and ^stsr and
then classify the former into two heads. In the third example
there is only one ft*T, of the hand with the tree, and
actually no second Hfrl is produced. It is only the same
contact being expressed in another form. When one says *a
man is climbing the hill', everyone knows that it is his feet that
are in contact with the hill and there is no question of this
contact producing another contact of the whole body of the
man with the hill. For this reason the more modern Naiyyikas do not accept crotaTSStfsifa aQ d on the same principle
is not the absence of crsrtn for then it would only be
an srsnq1 and not a gar. It denotes the stage of separateness
thus causing the destruction of a previous conjunction. This
is what the author means when he says that it is OTtaj- The
process of fwr and fcti is supposed to take plac in the
following order. First fsOTT, an effort say, to lift the pot;
by this there is }%vim, the pot gets disjointed from its place;
through this igfftTT^r, th pot loses contact with the original
spot;; then another fan\, effort to place the pot at another place;
then ^ T R ^ ^ T J contact of the pot at another spot. From this
it will tie seen that {%*TFT causes tfifaRrer and is not ^RfrRrer
itself But it is not explained why the second and third steps
should not be taken as one and the same and what harm is
there if fww becomes an ^T^T^T ( H'itn^fT^r ) as the moderners
would like to classify it. Whatever has been mentioned about
^Nt*F3H*t*T applies also in the case of ftwnsifeOTT though
it is considered the third knot in the Vaiesika system as^
mentioned in p. 199*. These five qualities from ^s^T constitute
l
as they abide in all the nine

10 & 11. TC3lSTC [106] (a)

A. editions have this versp as part of the text at the end of

without any connection
..
;

203

( b)

Translation( a ) Remoteness and proximity are the ( specific = 3r3TOT?0T ) causes for expressing the notions of objects
being distant or near by. They are of two kinds; those brought
about by ( relation to ) space, and those brought about
by ( relation to ) time. The 'former is being explained ( first )
When two things are situated at a place we gain an
impression that one of them is nearer to us than the other.
Aided by this impression the notion of proximity in the nearer
thing is created by the contact of that thing with less number
of space points ( between it and the observer ) Similarly the
notion of remoteness is created by the contact of the other
thing with a larger number of space points aided by the
impression that that thing is farther away than the other.^fw
here means ( in contrast to its use in ^f?^Teft^ under
c*r^r ) the fact that an object has less number of contacts with
space points intervening between the object and the observer's
body and fifrqr means the fact of the object having greater
number of such space contacts.
( b ) Remoteness and proximity due to time are being
explained ( now ). When a youth and an old man are situated
at any place an impression arises that one is younger in age
then the other. Aided by this impression the contact between
the younger one and the time ( factor ) causes the notion of
proximity (i.e. younger age) in the youth, and a similar
contact of the elder with the time factor creates the notion of
remoteness (i.e. elder age) in the older man aided by the
impression that he is elder of the two.
Notes-w^ and sr<R*3.are only relative notions with respect
to space and time. As already mentioned ( vide p. 28 ) modern

24

TAREABHS (

aiylyikas d not include isjfgr, frvrm, q'Wand srqrTesi in the

list of qualities. The impression that one is nearer or older
lhan the other is also a kind of sr^ig % (vide under $*U ).
For <?^9i and wmm the three causes are the object itself, its
contact with space or time, and the above-mentioned sf^Tf fi?
and other things ( general causes ). A editions have the word
ara*wnf$re*ta after fiifwJsMtto- and 3N3>f<FScNt?R i n the
abpve text.

12.

(VrS.:V. 1-7. ).
TranlationWeight ( or gravity ) is the quality which is the
non-inherent cause for the first step in the process of falling.
It subsists in earth and water. It is said ( in V. S. : V. 1*7 ) that

the falling of a substance is caused by weight in the absence

of conjunction ( with another body ), velocity or effort*.
NotesThe word sr'ra in the definition of g^f ( as also
in the next s[%m ) is to avoid overlapping in the second and
subsequent moments of falling ( or flowing ) which are caused
by the velocity generated by the immediately preceding
moment of falling ( or flowing ). The V. S. quoted actually
reads only frfcwrr" P^n^ <TcR*$' which means that in the
absence of any contact ( i.e. any support ) a body falls by its
own weight. According to K. M's. reading frtn, i.e. contact
with a heavier body, RT, i.e. retardation and tmm i.e. effort or
counteracting force, are obstructions to felling and so only in
their absence^a body falls due to its weight. K. M, might have
had a different reading of the Sutra.

13. ^m^

[108]

TranslationFluidity is the quality that is the non-inherent

cause for the first step in flowing. It subsists in earth, light and
water. That in earth and light, eg. in ghee etc. and in gold
etc. respectively, is caused by fire and therefore artificial. In
water it is natural.

1.4. ****? [109] i^^fw^^T ^f^rT^fe i

TranslationViscosity is oiliness. It subsists in water alone.
It is caused ( in products ) by similar quality in their causes
( i.e. their component parts ). Like weight it is coeval with
th substance which constitutes its substratum.
NotesAnother useful Vd eflnition-^-the one in the text is
only a synonymis ^ a j i ^ R r o ^ ^ l g : ( T. S. ). ^ f is the
quality that is the specific cause for agglutination of powders
etc. ( like tta ). Why this agglutination could not be ascribed
to ^3x3" is because the 53?^ of gold and other metals is not
capable of causing it. The ^ p n t ^ f t gloss on the T. S. has
the following interesting remarks on ^"3
1
^

15. 5rs^: [110] ( a ) ^3^;. zfmn^ft gw: I

if ^

(b) ^

( V. S. IL
2-31 ) i

206

TARKABHAS ( cT
I

( c ) fnfc ^ ^ ^ ^ T ^ r r r ^ i

?
i

cPTrf|; i

crsrrfq- ^ R F q r f ^ T sq-pff^TT: i

Translation( a. ) Sound is the quality that is cognised by

the ear. It is the specific quality of ether. How can it be
cognised by the ear ? For, sound is produced at the place
where the drum is while the ear is attached to the hearer ( at a
different place). True. Sound ( a t first) arising from the
place of the drum produces another sound either in the form
of water-ripples or in the form of filaments of the Kadamba
flower, This second sound produces another ( and this in
turn another ). In this manner the last sound that raps the
ear is heard by it; neither the first nor the middle ones.
Similarly when a bamboo is clept into two the sound arising
at the place of cleaving produces another and this another and
so on, till the last sound reaching the ear region is cognised
by it; neither the first nor the middle ones* The experience
that one gets as 'I heard the sound of the drum', is only a misconception ( if it is related to the place of origin of the sound ).
( b ) In the production of sound from a drum, contact, of
the drum with ether is the non-inherent cause and contact of
the drum with the stick is a general cause. For the ^ S ^ T sound
produced when a bamboo is being clept, the disjunction

207

between the bamboo pats and ether is the non-inherent cause

and the disjunction between the two clept parts is the general
cause. Thus the origin of the first sound is caused either by
conjunction or by disjunction, while the intermediate and last
sounds are caused by the immediately preceding sound itself
as the non-inherent cause and favourable air currents etc. as
general causes. This is stated by the Strakra : "The,origin
of sound is caused by conjunction, disjunction, and by sound
also" ( V. S. 11-2-31 ). The inherent cause for all sounds including the first is ether itself. Like, action and cognition
sound also exists for three moments of time. The first and
intermediate sounds are destroyed by the sound succeeding
each (their effects ). Some say that the last and last but one
sounds destroy mutually as is the case with Sunda and Upasunda-thesetwp demons, brothers, fought and killed each other.
But this is not correct. For the penultimate sound which ( by
'hypothesis ) should last for three moments, gets lost at the
second moment of the last one; and so it does not exist in
the third moment of the latter and therefore not in a position
to destroy the last sound. For this reason it is held that the
destruction of the last sound is brought about by the absence of
the penultimate one ( it non-inherent cause ).
( c ) Sound is known to be destructible ( i.e. non-eternal )
through inference which can be stated as follows. Sound is
non-eternal, because, besides having a generality, it is cognised
by our external sense-organ, like ajar. In this syllogism the
fact of being non-eternal is the TTS^T. Non-eternality should be
ttaken to mean the fact of having a/orm liable to destruction
and not the fact of having an existence liable to destruction,
la the latter case there would be no non-eternality in antecedent
negation which has no existence as such. The ^g in the above
syllogism is 'besides having a generality sound is cognised by
our external sense-organ'. If the | g is stated as 'cognised by
sense-organ' only, it would stray into STF^ ( as it is cognised
by the mind and also has a generality ). So the attribute
*external' is added to sensesDrgan. Even then the | g would
stray into atoms etc. which are cognisable by the external
senses of Yogins. To avoid this the word 'our' is used. [ The
opponent objects* ift the middle ]; What is th authority for

208

TARKABH3A (

the existence of YoginS; ? ( Answer ) It is inference like this :

The atoms are perceptible by some one because they are
knowables like jars. If the first part 'having a generality' is
not included in the | g it would stray into generality etc. As
generality, particularity and inherence have no generality,,
this part ' c n W 3 i ^ rff is also included in the | g , :
NotesWhen a stone is thrown on still water we observe
that endless circular ripples are produced one outside the
other till the last one is lost on striking the bank or becomes
too dim t o b e visible. Similarly sound at the place of origin
produces sound waves which move farther and farther in all
directions till one such wave peaches the ears of the hearer.
Here last sound is to be understood as the one that is heard
by the hearer for it is not known actually wttn and where
such sound waves get lost* Moreover the saine sound is heard
by different persons standing at different places. The analogy
of water ripples t explain how sound travels is called eft^3?$ftrT. In this there is only one series of several circular
!i
sound waves.
Others however apply the analogy of the filaments of a
Kadamba bud, known as ^^gfssssnqr. In this flower, the
filament that shoots up first causes several other filaments to
shoot up iri all parts of if] simultaneously. In the same way
the first sound at the place of origin produces other sounds in
all the ten directions simultaneously. Each one of these causes
other sounds again in all directions'till at last one such sound
reaches the hearer's ear. In this explanation th sound waves
increase in geometric progression. The analogy in either case
is to be extended only to the manner in which sound waves
travel from tfie place of origin to the ears of the hearer.
In para 110 ( b ) the author explains all the three causes
for the origin of sound, as also the causes for its destruction.
For all sound ether is the inherent cause as sound inheres only
in ether. For the first sounds ( Sx ) the conjunction of the
drum with ether, or the disjunction of the bamboo with ether
forms the non-inherent cause, and the conjunction of the stick
with the drum or the disjunction between the two bamboo
parts, forms the general cause. For subsequent sounds begin-

?7

209

ing from the second, the immediately preceding sound itself

is the non-inherent cause and a favourable current of air,
proper distance, etc. would Tormgeneral causes. As soon as
S2originates Sx gets lost, and when S3 ^originates, S2 gets lost.
This is what the author ; means by ^R^?TR^i:. Figuratively
it can be written as follows :
Origin Sl5 S2, S3, S4
LOSS

.S^i, S*,

: S l r S 2 9 , :

.:_

......

Sn~2,Sn-l,

x
Sn

In this wafy when the last sound Sn originates tlje penultimate one Sn-i dies away at the. next moment. But Sn produces
no further one and so it would appear that it is not destroyed
at all. But since the Naiyyika is firm in saying that sound is
destructible, he has either to say that Sn-i and Sn mutually
cause the loss of the other, or Sn gets lost by the 4ss f its
inherent cause Sa-i. In the former case "the general iiictum
that any sound is lost in the third second after producing a
succeeding one" will not be satisfied and so the author Bas put
forwarithe second reason for the loss of Sn . [Here all the
printed texts and my transcript also read wf|^^E?r fwra^nfr^'S. But ep^is supposed to exist for four momenta ( vide
p. 34). Ch.Bh.,merely glosses over it saying*%f ^%farof T gff%%
^ ^ . In the B dition it is mentioned that one of the Mss omits
^rfff?^?^' In the Ms. from which my transcript wasl made
there is a remark in the margin in a dijfferent hand %ifo
It has not been possible to get other Mss. or commentaries to
clear the discrepancy. ]
In para 110 ( c ) K. M. wants to emphasize the N^iyyika
stand that sound is non-eternal by a pedantic syllogism, the
words of which he explains in great detail. In fact, the previous para 110 ( b ) itself was a prelude to this, to show that
sound is always a product and so is destructible in contrast
to the Mmmsa doctrine that sound is eternal. In this
connection the notes under rs^TOFT may be recalled here also*

'

..

"

16 & 21.
[111]( a ) 3T%^T3ft ffe: = Cognition is that ( quality )
r-

which objects are made known.

210

TA^KABHS (

( b ) sftifo: qp, ^^W^t^VQ

= Pleasure

is

satisfaction and desirable for all.

( c ) qter:?, - ' R C ^ r ^ ^ ^ ^ f t W - P a i n
ing and is undesirable for all.
( d ) TT !^r Desire is attachment.
( e ) ^pt^ft -gq": := Hatred is anger.

is suffer-

(' f ) ^TTf * ^T?ToT: = Effort is exertion.

( g ) g^sqtfir: W ^ T W ^ S T T : = T l l e s i * qualities beginning
with cognition are perceptible by this mind ( only ).

22 & 23 ^ q O T f n ^ ] ; - ^ ^ : ^ ^ ! ^ ^ 5 ! ^ ^ 1
i ; eft ^ r s r ^ r ^ q - R ^ ^ i ^ T T ^ ^ ^ I

TranslationMerit and demerit are the special causes for

pleaure and pain. Though they are not perceptible hey are
known ( to exist ) from the Vedas, as well as through inference.
( THe inference is as follows ). Devadatta's body and other
things are produced by some special qualities of his soul,
because they, while being products are the causes for his
enjoyment, like any other object obtained by his efforts. These
special qualities of the soul which produce (his ) body etc. are
( his ) merit and demerit, as effort and th other special
qualities are not capable f producing the body etc.
24. cBErc; [113]

t f ^ ' i ffrfarsr^*3 :^*?Nrf^i"<sr^f%T: i

trMtf ^fsNf^lr^TT: I

211
Translation--Faculty is of three kinds, velocity* mental
impression and elasticity Of these velocity subsists in the
five substances, earth, water, light, air and mind; and it is the
caused by motion. Impression abides only in the soul. It is
produced by experience and is the cause for recollection. It
causes recollection only when aroused. This arousing comes
in through auxiliary aids such as perception of similar things
etc, as stated ( in an old verse ); "Similar things, unseen entities
{ such as merit and demerit )y constant thinking and so forth
arouse the seeds of recollection". Elasticity subsists in certain
tangible substances ( which have touch ). It is that quality
which, when its substratum such as a bow is bent or distorted,
helps in bringing it to* its original shape.
The eight qualities beginning with cognition and ending with
demerit together .with impression (the second variety of
#E3PT* ) constitute the special qualities of the soul.
Thus the 24 qualities have been explained.
NotesIt will be observed that fspT as such cannot be
. defined properly because its three varieties are so incongruous.
In fact the second variety *rraTT is the only real fasER and
that word is often used as a synonym for N$i?. *n^n is a
special quality while the other two come under common ones.
Ballantyne in his translation of the f^r^FcrgfBT^ supposes
that "the similarity of the three varieties of sfepft consists in
the thing acting by itself without an impulse from an external
agent". It is doubtful whether any of the three varieties ( or
aoy of their substrata ) acts by iise//whatever that may mean.
Here kn has been translated as velocity though its common
meaning is speed. Actually speed or velocity is initially
caused by a motion; but in the text it is mentioned as the cause
for motionfpTT^g. One way of getting over this difficulty is
to take fifistttg as a ^gsftff compound when it can mean that
ifaf has fifrsu ( motion ) as its cause. Ch : Bh.'s comments here
are .'33331!' $Pr3ftOTTinfopngtrefft'; this indicates that he had
a different reading here ( some such reading as ^ n r g ^ or
^l^TFfTST ). The *n. q. of ff^TO is explicit on this point
when it says Sip^r g km **n^, tmt Isjsr: fatf" (V. 158 ). In
its commentary, g^T^sfr it is exjplained further : c M

, wf i

5^
I'. THs meansVelocity is of two kinds,
that produced by motion and the other by velocity itself. In
an object like an arrow initially k*[ is caused by the push or
pull that is given to it. This initial ihr destroys the previous
motion and causes a second motion and so on ( till the arrow
coms to a stop ). Since one motion is an obstacle to another
( the preceding one) without this ifar the destruction of the
previous motion and the origin of the succeeding one cannot
happen. When the ijir in the two halves of a jar sets in
motion the jar itself it is an example of velocity created by
another velocity. If the reading fs^llg is correct, K. M. must
have meant this ^r^^i^fr; bun then it s vague and farfetched, which is unlike the author's method which is explicit
and straightforward.
.
Elasticity is found only in certain earthy substances and
not in water, light and air which also have the quality touch.
Here also the expression ^q^^f|[?t^f f% appears to be somewhat vague. Modern Niyyikas would like to exclude %T and'
fqfjTCBH'W from qualities and explain them as varieties of action
or motion ( ^ 4 ^ )
3. ^ ^

[114] ?^H FTTN"^' I . ^ r f T ^ 3v# I

^h i
5#pft'tlt
3Tf

^ T - ^

TranslationNow activity is being explained. It is of the

form of motion and like qualities subsists only in substances.
It co-inheres along with dimensions that do not subsist in allpervading substances. The fact of having such limited dimensions is called ajtf^f ( = corporeality ). Activity is the cause
of conjunction at a second place after destroyig the previous
conjunction. It is of five kinds( 1 ) moving upwards; ( 2 )
moving downwards, ('3 ) contracting; ( 4 ) expaniJing and
( 5 ) 'going*. Whirling and other kinds f activity come under
going'.

213
NotesThe words $ir,fithr aiid ^f^r, togh commonly
used as synonyms, have some difference in meaning, sph^ is
motion in general; f&?T is actual performance of an action,
and cfk is volitional effort. The word s p ^ used here should
be distinguished from the syntactic object of verb. The
Naiyyikas distinguish between voluntary activity ( iift<E$ )
and involuntary activity ( STOCTTJCTO ). The essential function
of every activity is firstly to produce ( 1 ) disjunction, ( 2 )
then the destruction of the previous contact and ( 3 ) finally
conjunction with a second place. The origin of an activity
occupies one moment and thus an activity after fulfilling its
function comes to an end in the fifth movement. It is because
of this.that 5 p ^ is conceived as the direct cause of a disjunction and, as leading to a further conjunction ( Cf.
vide p. 34, ).
The division of W ^ into five kinds as in the text, though
unsatisfactory is being adopted by all writers on the subject in
deference to the Strakra ( <OT['). Moreover it does not
make any difference by dividing spH^in any other form, or by
not adopting any kind of division.

5, r w p ^ [115] (a ) arfff^SRq^l^WFn^

( b ) 3T^ T f e r ^ I " sq-f^rsrfg'f^Rf \*TWRi' fef I

i
: i

Translation{ a ) Generality is the cause for comprehensive

class notions. It abides in substances, qualities and activities.

214

TARKABHS ( * |

It is eternal, only one and inheres in many things. It is of two

kindsthe /highest5 and 'lower5. The highest form is 'existence*
( or being )CRTas it inheres in all the three ( ^sq, got and
3W3[ ) This constitutes pure generality as it causes the notion
of comprehensiveness only; The 'lower' generality consists of
substance-ness, quality-ness, etc. and inheres in less number of
things (as compared with the former). It is exclusive
generality in as much as it causes both comprehensive and
exclusive notions.
( b ) In this connection someone says that there.is no such
thing as generality apart from the individual entity ( ssffe ).
To this we reply as follows. What is the basis for that comprehensive class notion which arises on seeing many individuals
of the same class such as cows, each having separate distinguish
ing features, if there is no unique thing that unifies all the
individuals ? That unique thing is what is called generality
(The opponent again objects ). That comprehensive notion
is due to the difference which each individual has from other
things; (e.g. each individual cow is different from all now
cows such as horses etc. ). It is only this difference that is
presented as the object of that comprehensive class notion
and not a generality in the positive form. ( Answer ). It is
not so for the unique comprehensive class notion is always
presented only in the positive form ( and not in the form of a
negation ).
NotesPara 115 ( b ) is to refute the Buddhists who do not
accept cernid ( vide p. 18 ). Their contention is that a word
like nfc denotes primarily the difference that that object, the'
cow, has from things other than itself. They call this
( vide p. 130 )

5. ftt*n [116] fra^fr

fa^s^fe:

I5: I
TranslationParticularity subsists in eternal substances and
is the sole cause of their exclusive notions. The eternal subStances are the five beginning with ether together with the atoms
of the other four ( earth, water, light, air ).

,VH

215

6. ^nwFir [117] 3T5gr%sir: r^^f: psr^r: i

T f fe: I ^T ^ *TT
i

TranslationInherence is the relation between two inseparable things. This has been explained earlier ( vide p. 13 ).
( Objection by the Buddhist ). You say that the 'whole* and
its 'parts' are inseparable and so the relation between them is
inherence. But this is not correct as there is nothing like a
'whole' apart from its 'parts'. For instance, an object called jar
is nothing but an aggregation of many atoms arranging therselves in somp particular form. This is answered as follows :?
That the jar is a single gross entity' is the cognition arising
out of perception. This cognition does not objectify the
atoms which are many, which are not gross and which are not
perceptible by the sense-organs. Nor can this /cognition be
said to be invalid as there is no sublating judgement ( t o
contradict it').

[118]
Translatin-Thus the six categories beginning with substance
have been explained. All these six are positive categories a&
they are presented in our cognitions as positive entities.

[119] ( a )

f: i

216

C:

TARKABHSS '% q

i sr

Translation("a. ) Now a seventh category which is of the

form of negation and which is cognised in the form of nonexistence is being described. This, in short, is of two kinds
relational negation pind mutual negation. The former again
is ofvthreQ typesantecedent negation, subsequent negation or
destruction, and absolute negation.
( b ) Antecedent negation is that which exists in the cause
before the product comes into existence, e.g. the negation of
cloth in the threads (before cloth is produced ). It has no
beginning as it has no origin, but it has an end as the coming
out of the effect itself constitutes its end.
( c ) The absence of a thing in its cause after the thing itself
is destroyed is subsequent negation, and hence called destruction; e.g. the absence of the jar in its two halves after it is
broken, which is brought about by striking it with a bar or
some such thing. It has an origin but no end as a thing
destroyed has no -second origin.
( d ) When the non-existence of a thing can be predicated
in respect of all times, past, present and future, it is called
absolute negation, e.g. the non-existence of colour in air.
( e ) Mutual egation ( or difference ) is that which denies
the identity of two things, e. g. a jar is not a cloth.
( f ) Thus the qifo called si ( objects of knowledge ) has
been described. ( An objection ). These objects ( which have

.;V;-..; , .

...

217

been described so-far ) have po; separate, existence from their

cognitions ( according to the Vijnnavdin Buddhist ) or apart
from &W(\( according to the Advaitin ). [ So how do you say
that these are different from sjre (their cognitions ) or s r ^
(ssraq^.) ? ]. ( Answer ) Ttiat these objects have separate
existence cannot be denied, as their existence is established by
valid instruments of cognition such as perception etc. [So
their separate existence apart from p
or 3 ^ should be
accepted ].
Notes--All the 3$tf3> categories have been explained earlier
under para 16 and may be recalled have. In the last para ( f )
above the author wants to refute both the Vijnavdin
Buddhist and the Advaitin in the same breath, as there is
some sort of similarity when they say that objects have not
real separate existence apart fromtheir cognitions, but have
only an illusory ( mfTOfW ) existence.
The Buddhist philosophers are divided mainly into four
schools of thought. Gautama Buddha, it may be noted, never
preached any philosophy as such, but only a conduct of life.
The differences in thought among the various schools arose
from the interpretations which' various Buddhist scholars put
into his teachings long after his Nirvana. Though a large
number of schools and sub-schoolsas many as over twentysix-are found mentioned in ancient Buddhist texts, they were
all classed later under the four broad schools( 1 ) nbuf^cp,
( 2 ) -3*ff ( 3 > WWfire (or ^ f ^ T ^ ) and ( 4 ) qim^K ( or
f^fRcnfFO. Who made this classification or how and when
it was made is not definitely known. But they are found
mentioned as four distinct schools in the 3. ?. & of wrar^isr
( circa 14th century ), perhaps for the first time. All these
schools have certain notions in common among them such as
the momentariness of all things, non-existence of a separate
s n w ^ , non-acceptance of categories like HTR'T and the rest,
acceptance of only two 5WT*Ts ( R^^ and agWT ), etc, etc.
The ^f?^ holds that objects as perceived by the senseorgans are real only partially i.e. so long as they produce
their cognitions which again are; ^finsp. The ecmf;^ thinks
that in a perception, only the form of an object is presented as

218

TARKABHSA (

T, and thfough this the object is tobe inferred, thus reducing perception itself to lwer order. The other two schools
do not accept the distinction between frw and fcpsifr^ ( the
object and its cognition ) Th w?TfTcf5 is the real nihilist and
contends that both object and its knowledge have no ; reality
but are only fleeting appearancesrf aT3R. The trtaT^R-firWTT*
^ 1 ^ maintains that ^ R te the only real thing and that objects
reflected in knowledge are just like things in a dream without
any real separate existence. It is with this school that the
Advaitin has some sort of similarity in certain respects. For
the word SJR the Advaitin would substitute snn^and for ^W%
( = non-existence ) fm?3[ or Qm&L ( = unreality ). Hence the
oft-quoted verse
U

The pluralism of the N. V. system does not agree with

either of these views on the ground that it is imposible to
controvert valid judgements and experiences gained through
perception etc. and hence fe^sr and f W ^ should have real
separate existence.

( 1 2 ) 1 % ('the fifth ^
[12J ( )

m. ffe: i ^a'^T^ft ^"T | f e : i m =#

(d)

219

e ) *pf 3TFT

faf

n^

>

Translation( a ) Cognition ( gf? ) is that which is denoted

by such synonymous words as ( f%; - cognition ) - ^ ^ ^
( = understanding ), ^ R ( = knowledge ), u&im( = apprehension );
etc. Or it can be defined as that which manifests objects (to
the mind). It is, in brief, of two kindsexperience and
recollection. Experience is also of two kindsvalid and
invalid one.
,
(
( b ) Valid experience is that which conforms to the real
nature of the object apprehended and it is caused by valid
instruments of cognition such as perception, etc; for example
the perception of jar etc. by the faultless visual organ; the
inference of fire through the knowledge of the reason, smoke,
the apprehension of the potency of a Word like *mq by seeing
similarity to the cow; the knowledge that the sacrifice ^ % eta is the means to attain heaven by hearing or reading such
Vedic sentences as 'one desirous of heaven should perform tbe
sacrifice.'
( c ) Invalid experience is that which does not conform to
the real nature of the object apprehended and it is caused by
invalid instruments of cognition. It is of three kindsdoubt,
ratiocination and misapprehension. The former two are being
explained further on ( vide paras 128 and 133), Misapprehension is knowing a thing as what it is not and hence an
error; for instance on seeing a piece of shell etc. wl\ich is not
real silver, one apprehends it as 'this is silver'.
( d ) Recollection is of two kinds, valid and invalid. Both
these kinds occur during wakeful stage. Alltognitions arising
in dreams are invalid recollections as all that is apprehended
as *that' (in the cognition ) appears as 'this* ( in the dream )..
due to certain defects.

220

,.

TARKABHS ( z

( e ) All cognitions are devoid of form and no object

reflects its own form in the cognition, because, the theory that
cognition arises with the form of the object reflected in it (ei^TIft^FTWi^ ) has beep refuted. For the same reason the inference
of the object from its form reflected in the cognition is rejected
as the existence of objects like a jar etc. is established by
perception. All cognitions are fully defined by the objects
presented in them and without being related to the objects no
cognition can be apprehended by the inind; because the knowledge ( or judgement ) that arises from any such apprehension
is of the form 'I have the cognition of the jar* and not merely
of the form 'I have a cognition'.
NotesThe division of gf^ or 5jra has been shown on p.
39 and for the explanation of valid and un valid cognitions see
the end of RTOT section in Part I. The author has not directly
refuted the *n$rcfcTTOT5 b u t has just pointed it in para 119 (f).
A fuller discussion on the topic would be out of place in a
primer like thisvide notes under thfvt para.

(13 ) TR^f ( The sixth

[121] 3%fofo*f TR": I
TranslationMj.ndis the internal sense organ and it has
been explained already ( vide para 80 f. ).

( The seventh
[122]

^^rt^rsft

^TTffpTT i

consists in performing such acts as

sacrifices etc. which ultimately lead to merit and dement
This constitutes the basis for all things going on in the world
( i.e. without activity life does not function )

( 1 5 ) #ra: (The eighth

[123] cffT TTtwt^T: I

221
TranslationDefects are attachment, aversion and ignorance. Attachment constitutes desire, aversion anger and ignorance false knowledge or misapprehension.

( 16 ) $PT"*TT^ ( The ninth

[124] J^^qfrT: t^^TR: I
TranslationRebirthis having a fresh origin and consists
in the soul getting into another body after leaving the present
one.

( 17 ) r$^ ( The tenth

[125] iwf %r:,
TranslationResult is enjoyment; i.e. experience of pleasure
or pain.
( 18 ) igimil

( T h e eleventh

. \ . [126] cfYlT | : f T I c f ^ ^ c T ^ I
TranslationPain is trouble and has already been explained
(vid para l i t e ).

( 19 ) 3H3*: ( The twelfth

[127] ( a ) Tfr^sq^: I

222

TARKABHS (

Trnslation( a ) Salvation is release from birth. This consists in complete cessation of all the twentyone forms of
suffering. These twentyone are those arising from ( i ) the
body, ( 2 to 7 ) the six sense-organs, ( 8 to 13 ) the six objects
of these six, ( 14 to 19 ) the six kinds of cognition ( from the
six senses ), ( 20 ) pleasure, and ( 21 ) pain. These include
those that arise directly as well as indirectly ( from all these
21 ). Pleasure is also considered suffering as it is always
mixed with pain, s r g ^ here means 'always mixed up with
or accompanied by (pain)'. This implied statement (that
pleasure is pain ) is ( to be understood ) in the same manner
as honey mixed with poison becomes itself poison.
( b ) How does this salvation come about? This is being
described now. When a man has understood the real nature
of all things from the astras and has cognised the defects in
the objects of enjoyment, he loses attachment to these and
becomes desirous of release and ( to gain that ) he takes to
meditation. On attaining perfection in meditation he realises
the true nature of the soul and thus gets rid of defects and
sufferings. Then by performing actions without any attachment to results he ceases acquiring further merit or demerit. By
his yogic powers (acquired by meditation ) he comes to know
his past merits and demerits ( which caused his present birth ).
He collects them together and ends them by enjoying their
effects. By this he exhausts all his previous Karma ( ie. their
effects ) and when the present body dies off, his soul has no
new body to enter into and thus loses all contact with the
twentyone forms of suffering for want of their cause ( the
body). This release from these 21 forms of suffering is
salvation.
Thus the 12 rafsthe second category in N. S. I-1-1 has
been explained. Next the author begins to describe briefly
he other categories, *?fat, etc.

223

( 20 ) ?cfeflTn ( The third category )

[128] ( a ) f ztfi CfSFWTf, I
: mm-, i g- fwfasr: i

Translation( & ) Now doubt is being explained. When in

one and the same object there arises the apprehension of two
or more conflicting notions, that :beeojries doubt. It is of
three kinds. The first variety is that caused by the observation
of certain common attfibutes of two things while not noticing
any differentiating features between the two; e.g. whether the
object ahead is a tree stump or a person. In this case on
seeing an object at a distance the observer does not notice any
curvedness or crevices that would enable the object being
recognised as a tree stump; nor does he see any clear definition of head, hands, etc. which would enable the object being
recognised as a person. At the same time he observes some
features common to both the stump and a person such as
tallness, shortness, etc. Then he gets a doubt in the form
4
Is this a stump or a person V.
(ib^ The second variety of doubt is that arising from the
difference foft opinion aboufon, and the same thing when there

224

TAR&ABMS (

is no Special reason to ignore er accept either; e.g. whether

sound is eternal or not. It happens like this. When one says
that sound is eternal and another that it is not, the hearer,
who sees no special grounds either way, doubts whether sound
is eternal 'or not.
( c ) The third variety of doubt is that caused by abserving
some peculiar attitude or characteristic, without noticing anything special, in an object. For example when one apprehends
smell which is the special quality of earth, but which does not
indicate whether it subsists in eternal or non-eternal things
and when he does not see any special reason ( to decide either
way ), the doubt arises whether earth is eternal or not Because
it will be non^eternal if smell is not present in eternal things
and it will be eternal if smell is not present in non-eternal
things.
NotesThe theory f doubt W as important and abstruse
as that of knowledge and the fact that Gautama has "mentioned
it as the third category just after SWTFT and srito shows that its
importance was recognised from very early times. He defines
as
fipTO* #!FB ( I--23 ) The vagueness of this Sutra has
given rise to various interprtations. The arr^T explains it
as five different types of doubt caused by
( 1 ) CFRTWR

=Hrauore? as in

( 2 ) ar^sm

s . * a i ^ r ? f as in qfiron

as in ^ ^ f^r

fa

( 3 ) fgurRrq^r
=f^^g[TW as in snwn
( 4 ) ^q^s^s^rsf^T An indecision caused by a positive
cognition such as water is cognised
as existing in a pond as also in a
mirage. From this a doubt arises
whether existing water is the object
of cognition or non-existing water.
( 5 ) gqswrww^T = An indecision caused by a negative cognition such as there is no
water in this pond, which is covered

225
up by grass etc This also gives rise,
to the same kind of doubt as ia
No. 4.
The 3Tf$35 is definite in saying that the first three are
the only three varieties of doubt and that the portion 3T735'
WRrar* in ttu Sutra is only a qualification for the first part.
K. M. has followed the ^ifl^^TT and has summarised his views
as in the text above. The qw&rc of vrttrftT follows the
WT^TC in giving five varieties; bit the examples he gives are
different except in the first and fourth cases, which fact is
indicative of another interpretation of the Sutra.
(
(
(
(

2)
3 ) fsrfrw t ^ ^fir^TftfhfinrarTf.-,
4 ) 3q^5:
ft5 ) 313395^: fk

The modern Naiyyikas would limit i&^m to two varieties^

the first arising from the grasp of ^ r w ^ H and the second
arising from that of $rgT9ft<W74. The examples are the same
as those given by sHshr ( quoted above ); ( Vide S. M. under
V. 130. ). '
Of these the first one, *sng^? g^it ^rr, is the real example
for doubt as it is understood ordinarily. Here also there is
some controversy between the ancient and modern schools.
The former say that it is only one doubt
^ - w h i c h is in conformity with their
definition
^^^(8^^^^^^^^^^^^
*faw! Th&
modern school hold that tTRi^rilsT^r?^ in the above is not
sufficient to avoid srfirs^rPcT in such valid cognitions as ^t
and so the definition should be stated as q^srf
^vrRTVT^s^^Si^. According to them the above
doubt would resolve itself into a mixture of two doubts as
fqFsRrftw^^igwr^TTSTspri and f^^fir^^^s^R^^T^^Tf 4 ;

( in other words as 'whether it is a stump or not' and 'whether

it is a man or not.' ) This depends mainly on one's experience^
whether he apprehends it as: one cognition or as two., All
15

226

TARKABHAS (

these above differences regarding varieties of doubt are due to

the causes of the doubt and not in the actual fact of it.

( 21 ) ^ T ^ r R ^ ( T h e

fourth

category )

[129] 3FT gxi: SRT ^ SPTfa^ I

TranslationThat by which one is urged to act is purpose
{ or motive ). It consists in gaining pleasure and avoiding
pain, for actions of all men in sound health are directed to
the attainment of these two.

( 22 ) TOT?W ( The fifth category )

[130] ^Tf^^%^Tf^^srfa-

TranslationExample is that thing about which there is

good agreement between the two parties in a debate. It is of
two kinds. The first is example through similarity as the
kitchen for the probans smoke in an inference of fire. The
second one is example through dissimilarity, as the pond for
the same probans.

( 23 ) % ^ F r : ( The sixth category )

[131] a r o i ^ ^ n ^ ^ s i : ftrsrer: I *r

ff TOT

TranslationAn established conclusion is that which is

accepted as authoritative. It is of four kinds( 1 ) that which
is accepted by all astras; ( 2 ) that which Is accepted as
proved in another Sastra; ( 3 ) that which follows as a corollary to an established conclusion; and ( 4 ) that which is

227

acceptable as such only on the basis of another thesis. Instances of the first kind are such statements as 'everything
possessing qualities exists.9 Those of the second are such as
the fact that the mind is a sense-organ as proved in the sister
system Vaiesika. Those of the third kind occur in such facts
as *God is Omniscient' which follows as a corollary to the
conclusion 'Earth etc. are created by a maker'. Examples of
the fourth variety are such statements as sound is accepted as
a quality by the followers of Jaimini ( i.e. the Mmamsakas )
only as the basis for discussing whether it is eternal or not.
NotesRegarding f^Fcf also there are wide differences
among the commentators in the interpretation of the otras of
Gautama on the subject ( vide N. S. I. 1-26 to 31 ); and these
are mainly verbal rather than material. H ^ C means an
established and accepted conclusion on any topic. This may
vary from system to system and from school to school within
a system. One system or school may accept or reject, fully or
partly, the feR of another. Any division offr^T?ras of so
many varieties on the basis of this fact will not elucidate the
subject. Consider the example *mind is a sense organ' given
for the 2nd variety of fe^!*3 io the text. The same example
has been given for the 3rd and 4th varieties by other authors,
showing thereby how widely they differ in the interpretation
of the Stras. In the fourth example there is no question of
any fr^ar. The Mmmsaka argues that sound is eternal
within the framework of his system and the Naiyyika rejoins
with equal vehemence that it is not. Neither side yields to
the other and no conclusion emerges as both appear equally
reasonable. In his eagerness to follow the Sutrakara K. M.
has briefly paraphrased the words of the Stras, as in most
other Jopics.

( 24 ) 3pP73[U ( The seventh category )

[132] S T ^ H W W ^ T 3T^T5TT: I r
{I. 1-32. )
fcffTT,

228

T A R K A B H S ( <r&

f, mi xft tft

:I

TranslaiionThe several steps used in stating an inference

for others are called members of syllogism. They are five in
number beginning with nfcfgi as stated in N. S. I. 1-32( 1 )
srfr^n = general enunciation of the minor term as possessing,,
the major term, as *the mountain has fire% ( 2 ) |g=sthe statement of the probans in the ablative or instrumental case, as
'because of the presence of smoke'; ( 3 ) g^TCT*! =* the statement of an example along with the s^TH, as *wherever there
is smoke there is fire as in the kitchen'; ( 4 ) Hq^r^s? statement
that the subject in question possesses the reason ( which is concomitant with the probandum), as 'the mountain has smoke
or like that'; and ( 5 ) faiTTR = the statement emphasising the
existence of the probandum in the subject, as 'therefore the
mountain has fire' or simply 'it is also like that.' These five
statements serve as parts or limbs of the body of an inference h
and so are called STWS'S. They are not to be understood as
inherent causes as parts are to the whole; because sound inheres only in ether ( and these statements are only sound in
the form of particular words ).
( N. B. All these have been explained earlier" under
and are only repetitions ).

( 25 ) rmi ( The eighth category )

[i33] gr^sf^^r^: i
4
T: I

229

f
cf SfcT
: I

Translationfis a kind of hypothetical reasoning leading to an undesirable result (hence reductio ad absurdwn ).
It consists in arguing that if, out of two concomitant things,
the concomitant ( ssfp^r ) one is present, the presence of the
other? its correlate s?TTq^ should also be present; e.g. if a jar
were to exist here it should be perceivable like the spot ( where
it stands ). This kind of hypothetical reasoning helps the
instruments of valid cognition. It is in this way. Suppose
someone concludes that there is no fire in the hill after doubting about its existence ( from seeing smoke issuing from it ).
Then if another were to tell him fiIf there is no fire in the hill
there would be no smoke also', this kind of argument is what
is called ft* It helps in establishing the presence of the
probandum in the subject which is the object of the inference.
Because it points out the absurdity of the hill being smoky in
the absence of fire ( its cause ), it is an aid to inference. Someone says that 3P is included under doubt; but this is not
correct as w has only one thing as its object and that too a
definite one ( while doubt objectifies two contradictory things
in one object ).

( 26 ) f ? n w , ( The ninth category )

[134] f^ftS^qTT^Jra'T^ , T&f *TT*TT
Translationfk^ ( or decisive knowledge ) is firm conviction and it comes as the result of the instruments of valid
cognition.

230

TARKABHS (

( 27 ) ^ - ^ 7 ( The 10th, 11th and 12th categories )

[135] ( a )

( c)

Translaiion( a ) Discussion ( sn^ ) is argument between

two persons or parties desirous of arriving at the truth ( of the
topic under discussion ). It is in this that the eight clinchers
or points of defeat are employed ( to win over the opponent ).
These eight are ( 1 ) deficiency ( in the reasoning ), ( 2 )
superfluity ( by bringing in extraneous or irrelevant points ),
( 3 ) stating something contrary to established conclusions, and
( 4 to 8 ) the five fallacious reasons.
( b ) Disputation (^?^) is also argument but with the desire
of winning over the opponent and it accomplishes two things,
It usually ends in establishing one's view-point by demolishing
the opponent's.
( c ) Wrangling ( fk^^i ) is ( destructive ) argument which
does not establish one's own position, but which is directed
only to find fault with the opponent's position, in fact the
wrangler has no position of his own to establish, mm or
discussion ( under which broad head come all these three )
constitutes the arguments put forth by a number of persons
assembled together, some raising objections and others trying
to answer them ( regarding any topic proposed for decision ).
NotesThe word ^I is used here for discussion or argumentation and Gautama has divided it under three heads. The
first,^^, is carried on by two parties in a healthy atomsphere
with the sole object of arriving at the truth and understanding
the topic under discussion in all its aspects. This is seen
even now in Sanskrit Tols where the students after taking

23 i
lessons from the Guru, gather together and discuss throughly
what they have learnt. This is also called fpre. The second
variety, H^, is that in which both parties know where they
stand, but in which each wants to defeat the other by establishing his superiority, This dual purpose is what the author has
in mind when he says aTOTCrraTO^h This can be compared
more or less to the discussions that take place in tniret^..
Such straic meetings used to be held till recently under
royal patronage in the princely States in the South. The
third, f^af^, is arguing without any purpose except to find
fault with the opponent Examples of this type can be found
here and there in the controversial works of reputed authors
in all the systems.

( 28 ) f r ^ m r a n ( The 13th category )

[136] ( a ) 3W
i rsfqr

I cf *t ^ f e f ^ - S T ^ i f ^ ^

:I

Translation( a ) A probans becomes its opposite ( 3f|g ) if

it fails to satisfy any one out of the five conditions, such as
existence in the subject etc ( explained under ^ H ). Because
it appears to be like a valid reason by virtue of conforming to
only some of those conditions, it is called tscnsTTH one resem-

232

TARKABHS ( xwm )

bling like a valid reason. They are only five in number( 1 )

th unfounded or unproved reason; ( 2 ) the contradictory
one; ( 3 ) the inconstant or straying one; ( 4 ) the opposing
or counterbalanced one; and ( 5 ) the stultified or belated one.
( b) In this connection 3?3R says, "A reason is said to be
proven ( f^ ) if its concomitance with the probandum is established and if it is found to exist in the subject. If it fails in
this respect it is said to be unproven ( srfn^). This definition
applies to the 'contradictory' and other faulty reasons and
therefore suffers from the defect of afssnfg. We shall however
explain it in such a manner as to avoid this overlapping. With
respect to faulty reasons that condition which is first detected
as making it faulty is to be regarded as 'fallacy'. When one
fallacy of the reason is indicated by this faulty condition alone,
that fact itself would be sufficient to close the discussion and
there is no need to point out other fallacies that may be detected
later. In view of this, only that reason which is concomitant
with the contrary of the probandum, can be called 'contradictory'. Similarly in cases where the fallacy first detected is
inconstancy etc. those reasons become inconstant etc. In like
manner where it is first detected that the reason is not invariably* concomitant with the probandum or that it does not exist
in the subject, that reason is to be taken as the 'unproven*
ooe ( sifs?^[ ) etc.

1. srfrc^ [137] { a )

(b)

i Mv i

233

Translation{ & ) The unproved reason as explained above

is of three forms: ( i ) that which has an unestablished
subject; ( ii ) that in which the actual form of the reason is not
-proved; and ( iii ) that which has its invariable concomitance
with the probandum not proved. Of these the first variety is
that in which the subject ( w ) of the probans is not known or
established, as in the syllogism; the sky-lotus is fragrant,
because it is a lotus, similar to one in the pond. Here the
subject, sky-lotus, is a non-existing ( i.e. imaginary ) thing ( and
so unknown ). [ In this the q^f as it is, is unknown. So the
author gives another example where the q^f, though well-known
is lacking in its attribute q^T. ] Here is another example of
^Tsnsnfe^-*A jar is non-eternal, because it is a product, ,Hke
a cloth.' (An objection) Here the subject sjar' is not an unknown
or unproved thing. So how can you classify this as arwrifr^?
At best it can be said to be a case of superfluity ( proving a
proved thing ) as it only proves the non-eternality in ajar
which is already established. (Answer ). It is not so, because
a thing by itself does not become the subject of an inference.
It becomes a subject only when the existence of the probandum
in it is in doubt as has been laid down (in the Nyya Bhsya):
"A syllogism operates neither with reference to a thing
totally unknown, nor with reference to a thing definitely
known, but with reference t a thing the presence of
which is in doubt.'1 There is no doubt about the existence of
non-eternality in a jar as it is a definitely known fact. Even
though the jar as such is a known object, it cannot be considered as q^ in this inference as there is no doubt about the
existence of the probandum (non-eternality ) in it. [ In other
words it lacks the attribute q^m which has been defined by
K. M. as e f ^ g i ^ w s r , vide para 57 ]. So the probans (in
the above syllogism ) is

[138] ( a ) ^ ^ T f e g ; *r 3 ^ eft t ^

Elfter,

234

TARKABHS ( aq|v

( b ). TOT

f^Wrfe^fa^
i

( c)
i
(d)

( e ) sra^f^NrFfr^V m\ t^ r^^^q^ qt: i

Translation138 (a) A reason is said to be ^^qfr^ when

its presence in the subject is not proved; e.g. 'Generality is noneternal because it is a product.' Here the reason 'being a
product' is absent in the subject generality. 'A partially
unknown reason' ( *n*nf&3g; ) also comes under 3H$<TTf(5( I for
instance in an inference like'the atoms of the four substances
beginning with earth are non-eternal because they have smell,'
the reason 'having smell' is not found in the subject as a whole
but only in a part of it, the earth-atoms. Therefore, being
'partially unknown* it becomes r^qf?.
( b ) Similarly, ( i ) when the qualification is not proved;
( ii ) when the qualified is not proved; ( iii ) when the qualification is impotent; or ( iv ) when the qualified is impotent ( and
similar instances ) all come under ^r^qf^. An example of
the first form is'sound is eternal because being a substance it

235
has no touch.' Here the absence of touch as qualified by
'being a substance' is the reason and not mere 'absence of
touch'. Sound is not a substance but a quality and therefore
\he qualification of the reason is un-proved. When the qualification part is absent, the whole qualified thing is also absent,
just as when either the stick or the person is absent, the qualified whole, 'man-with-stick' is also absent. Thus follows the
( c ) An example of the second form ( fifasnfei;)
in an inference like 'sound is eternal because having no
it is a substance*. Here the qualified part (frfaq)
reason is absent in the subject and hence the full reason
absent as explained above.

occurs
touch
of the
is also

( d ) An instance where the qualification of the reason is

impotent ( as not serving any purpose ), is 'sound is eternal
because being a quality it has no cause' -Here the qualification
'JJOT^' serves no purpose as the qualified portion ( STHPT^H^ )
alone is sufficient to prove 'eternality'. So the reason has the
fault of 'impotent' qualification. [ An impotent qualification
is as good as a non-existing one ]. For want of the qualification
the whole qualified thing becomes unproved and thus the
fallacy ^^qif% occurs. ( An objection ), Now sound is a
gor and so how can you say that gongr is absent in it? (Answer).
True, sound is a SJOT, but it does not serve any purpose here
by being used as a qualification of the reason. Only that
thing can be considered a (useful) qualification if it serves
the purpose of making the reason distinct from other things.
Here the qualification rpre^ serves no such useful purpose and
is therefore termed as srwi ( impotent ) [ and so becomes
as good as non-existent ].
( e ) The defect of 'the qualified being impotent9 occurs
when in the above inference the reason is stated with the
attribute and the substantive interchanged, as 'sound is eternal
because being without a cause it is a quality9. [ Here also
Tgnm serves no useful purpose even though it is mentioned as
the substantive; and so the defect remains the same with a
slight change in name. ]

236

TARKABHAS (

[139] ( a )

(b)

1 sr^sr^s^trf^ssr1 i

q; i
i
i

f?rf^g?f,

Translation^ a ) The third variety, swaKmfo^;, occurs

when the invariable concomitance of the reason is not ascertained. This happens in two ways( 1 ) When the reason does
not coexist with the probandum and ( 2 ) When it is related to
the probandum through an adventitious circumstance ( sqfr ).
An instance of the first kind is 'whatever exists is momentary
like clouds; sound and other things do exist ( and so are
momentary ). Here 'sound and other things' constitute the
subject, 'momentariness' is the probandum and 'existence' is the
probans* There is no proof to ascertain the invariable con-

23?

comitance between Existence' and 'momentariness'. [ The

syllogism being taken from a Buddhist author is expressed in
their form as already explained under para 53 (c) ].
( b ) Now the second form of s^T^^fr^, that caused by
the presence of sqfr, is being explained. In an inference like
'this son of Maitri is black because he is Maitri's son', 'blackness' is sought to be proved through the reason 'the fact of
being Maitri's son'. But this reason is not a causal agent for
'blackness', and the real causal agent is 'feeding on leafy
vegetables etc'. This causal agent is called sqfr. So in the
above inference 'feeding on leafy vegetables' acts as sqifw*
just as in the inference of smoke through fire as reason
'contact with wet fuel' is sqf. Because of this the reason
'being Maitri's son' is related to srqfr and so becomes
( c ) Another instance of nsKn%9 ( with 3<Tfw ) occurs,
in an inference like "killing in sacrifices produces demerit
because it is 'killing' just like 'killing' outside sacrifices". Here
'killing' is not a causal agent for 'demerit', but 'prohibited-ness'
which thus becomes sqfr. This has been explained earlier
( vide para 53d. ), and so here also the reason is sso^^Tf^r,
( Objection ). The definition of sqifr is that it is co-existent
with the probandum, while at the same time it is not co-existent,
with the probans. This does not apply to 'prohibitedness' and so
how can it become wfr h e r e ? (Answer). Not so. The definition
does apply to 'prohibitedness'. We have the s^f^cT Vherever
there is demerit there is prohibitedness' and so it is co-existent
with the probandum. At the same time we do not have the
ssnfcT 'wherever there is killing there is prohibitedness' as the
latter is not observed in 'killing in sacrifices' ( which is not
prohibited ). So it is not co-existent with the probans, bil
( and s becomes sqf )'. Thus the three forms of ss
have been explained;
s-All these points have been dealt with in some detail
under para 53 of Part I and the notes thereunder may be
referred to. In fact the whole of fNHr section here is a
repetition as in so many other cases.

2 38

TARKABHAS (

2. firent [140] fresr rr^^rf^rqrifq-q-Tcefr |g;:

TranslationA contradictory reason is that which has

invariable concomitance with the opposite of the probandum
in question; for instance when 'eternality' is to be established
through the reason 'fact of being a product'. This reason has
invariab'e concomitance with the opposite of the probandum
as it is well-known that 'whatever is a product is non-eternal'
( i.e. perishable ). Therefore the reason $35*3 here is a Contradictory5 one.

3. s R ^ l f ^ i [141] ( a ) ^rTO^|5T#^Tf^:,

^^

( c)

Translation( a ) That reason which carries doubt about

the presence of the probandum ( in the subject ) becomes
snr^TSR ( =having no fixity ) or ^S^TTT? ( =the inconstant
one ). It is of two kinds, the common one or too wide strayer
and the uncommon one or too narrow strayer. The first one
is that which exists in all the three, q^ ( the subject ), mi$
( positive instance ), andfre^T( negative instance ); as in the
inference 'sound is eternal because it is knowable*. The reason
( knowability here ) exists in the subject sound, in positive
instances like eternal ether etc, and also in negative instances
like non-eternal jar etc, as all things are knowable. Hence
^knowability' here is a wide strayer ( H T W U ^ P P - T ).

239
( b ) The second form ( srflPROT ) is that reason which does
not exist in both positive and negative instances, but exists only
in the subject; as in the inference 'earth is eternal because it
has smell.' Here the reason ( having smell ) does not exist in
positive instances like eternal ether etc., as also in negative
instances like non-eternal water etc., but exists only in earth
as it is its-exclusive quality.
( c ) The inconstancy ( s^fr^R ) of these reasons comes
about by their not fulfilling the necessary condition. This
condition here is that when there is both SW and fq^r, the
reason should exist only in the Cfq^f and never in the f%qpf.
;[ Then only the infallibility of the s?n%can be grasped ]. When
the reason is such that its opposite is what is concomitant with
the probandum, (as when it strays into fq^), it fails to fulfil the
above condition. This failure is the fallacy iqrftrarc and this
can occur in two ways either when the reason exists in both
3fq^r and fw( then it is mrcoi ) or when it does not exist in
both ( then it becomes efareiOT ).
NotesIn para ( c ), perhaps the author means to say
frerfqfcrw and not ^roforftaranccrer, which is the reading
in all the editions and in the MSS seen by me. The latter
reading would make sqrfpsR as only another form of
( vide para 140 X

4. rorcorcra: [142]
t : > f frerr, w.

f :

240

TARKABHS (

TranslationA reason becomes 'opposing or counterbalanced' ( sj^^R ) when it is opposed by an equally strong:
reason to prove the opposite of the probandum. Hence it isalso termed RSifcq^r, one having an equal opponent. For
instance the inference 4sound is non-eternal, because it has no
attributes of eternal things', has an equally strong one proving
the opposite of the probandum in 'sound is eternal because it has
no attributes of non-eternal things'. So the latter is an opponent,
of the former* If the second inference is not of equat
strength it does not become the opponent of the first.
It is like this. An inference which tries to prove the
opposite of the probandum in the first falls under any of the
three heads ( 1 ) that on which the first one is dependent;:
( 2 ) that which depends on the first; and ( 3 ) that which is
neither of the two ( Le, independent ). The first variety is
only a hindrance being stronger than the primary inference;,
e.g. an inference like'the atom is non-eternal because it has
touch like a jar, is not opposed by the inference establishing
the existence of the atom even though the same one proves the
eternality of the atom also ( the opposite of the first probandum,,
vide para 89 ). This is because the former is dependent on
the latter as it is the latter one which establishes the existence
of the subject itself ( ifting ). If the subject itself is not established by valid instruments of cognition, the inference
proving non-eternality in it does not arise for want of the
subject, and so it presupposes the inference establishing the
existence of the atom and thus becomes dependent on it and
consequently hindered ( i.e. gets rejected ) by it. The second
variety ( sqsfisf^ ) being weaker than the original inference
gets automatically rejected by the latter, as will happen when
the two inferences of the first kind are reversed in their order
( i.e. the one establishing the existence of the atom as the first
and then the one proving noneternality in it). The third
variety ( of opposing inference ) is the real 3*stRrq? a s

241
two are of equal strength and so get counter-balanced ( thus
preventing any valid conclusion ).
5. <CT3R*IOTf^gt or mw> [143] ( a )

I5:

( b ) ^ ^fer^q-: q^r
Translation( a ) The stultified reason is that where the
presence of the opposite of the probandum in the subject is
ascertained by other means like perception etc. It is the same
as what is called cnfrer ( = the annulled reason). Take for
instance the inference fire is not hot because it is a product
like water'. Here the opposite of the probandum ( i.e. hot-ness)
is ascertained by actual perception ( and so the inference
stands stultified ). Another example occurs when 'momentariness' ( ^ ^ 3 ) is sought to be established through the
reason 'existence' ( w% )vide para 139 (a). Here also the
opposite of the probandum (i.e. non-momentariness ) is established by perception in the form of recognition ( ^ ^ ) ,
which is of the form 'this is the same jar that I had seen earlier'.
This recognition comes from the impression of a past experience while seeing the jar again. It connotes two stages of time,
the past (reflected by n = the previously experienced ) and
the present ( reflected by sr4=the one seen in front now ).
This therefore proves that the jar exists for some time ( longer
than a moment ).
( b ) These five kinds of fallacious reasons, srfe^; etc.,
when they are lacking in one or the other out of the five conditions for a valid reason such as subsisting in the subject etc.,
do not serve as valid reasons as they do not prove the existence of the probandum in the subject.
16

242

TARKABHS (

Ts [144] ^sfqr
: 1 ergriff 1

TranslationAll definitions constitute reasons which have

only purely negative concomitance ( as they serve to separate
the thing defined from other things not coming under that
definition). Their defects are three in number:g
partial applicability; arfasmfccr = over-applicability, and
= total inapplicability. All these defects come under these
fallacies ( t^TWR" ) and are not additional to these five. For
instance srffs^Tf^ is a form of the fallacy 3tn<BR3ifofE [ vide
para 139 (a) ] as such a definition will not be excluded from
all negative instances ( fq^r ) ( vide para 46 ) and will also be
related to 3P7TW, as when a cow is defined as *an animal' (qg).
Here the 'fact of having dewlap etc'. ( rrcaifpm ) is the causal
agent for 'being a cow' ( 3rt?3f ) and not canimal-ness' ( n^m )~
[ So 9rnfE?rt3 becomes gqif ] Similarly STOTTPS is a form
ofwnff% (vide para 138a) as when the cow is defined as
an animal having variegated colours ( as this definition would
apply only to a few cows ). The third defect srtfmr comes
under *^RTfrf% as when the cow is defined as 'an animal
with uncloven hoofs' ( as this definition will not apply to
any cow. ).
NotesThe three defects of a definition ( &^m ) have been
explained on p. 6 and how these definitions are %ff353fcf^3~
|g's is explained on p. 81, para 46. 'How these three defective
definitions come under $*3in's can be easily seen by using
them as reasons in syllogisms; e.g, (1) in ar ifh
fpr*3 becomes ^rqff, for we have
^ ^ s ^ q ^ a n d we do not have
W$ as it fails in m>% etce^RT3*nq^. For the other two
eases, the inference, afH^fasft fvr^ s ^ R ^ ^ or
would provide the example.

243

( 29 ) ^ ( The 14th category )

14. ^ [ 1 4 5 ]

graft H^rrit, fft %f%

TranslationWhen someone uses a certain word (or words)
in one sense and if the hearer were to find fault with it by
construing it in a different sense, this practice is called qss or
quibbling. For instance when one says 're^s^sq' meaning
thereby that 'this man has a new blanket', the hearer scoffs at
the speaker; "He is too poor to have nine blankets. He has
not even two. What to speak of nine !" One who argues
thus is called a ^ ^ i f f \ or a quibbler.
NotesHere the word ?T3 is used by the speaker in the
sense of cnew' but taken by the hearer as meaning 'nine'
( usually wantonly to find fault with the speaker ). This
happens usually with double entendre words. Instances of
other kinds are '^fift m{\\ c^mt m*w*9 etc., if the jf& is taken
mistakenly.
( 30 ) STTfen ( The 15th category )
15. 3Tfn [146]

I qf?

i rr^^ftsfqr
TranslationAn incorrect rejoinder constitutes n% or
futile objection. It is of many kinds such as TaBtfafflT,. SN^ra, etc. [ N. S. V. 1-1 enumerates 24 varieties ]. For fear of

244

TARKABHASA (

unnecessary elaboration all these are not being explained

here ( but only two as models ). When ( on hearing a syllogism ) one argues that the subject should possess another property also which exists in the corroborative example, but which
is not concomitant with the probandum, it becomes ^^OT 5t%s
For instance when one says 'sound is non-eternal because it
is a product like a jar'., the opponent argues 'if so, like a jar,
sound should also be a composite thing ( n^cf ) for the same
reason ( ff^ft! ). When, through some property which is not
concomitant with the reason but is found in the corroborative
example, one asserts the 'opposite of that property, with which
the reason is not concomitant, this amounts to $rq3qFHTr snf.
For example in the same inference as above ( proving noneternality in sound ) the opponent begins to argue ; 'if sound
is non-eternal being a product like a jar, then for the same
reason it should be non-audible like a jar as the jar is certainly
not audible'.
NotesIn the first case H R ^ ^ W is not a property of sound
and merely because it co-exists ( and is not concomitant ) with
3>W3T in the jar, its presence in sound is wantonly ascribed
hoping to discredit the first reason. Similarly STTOTO^TOTO is
found to co-exist ( and not concomitant ) with 3ft3ft3 in a jar
and so its negation, i.e. $TT30TC3 is being denied in the subject
in question ( sound ) and hence the incorrectness or futility in
bothxases

( 31 ) forOTTOt ( The 16th category )

16.

TranslationA vulnerable point ( or clincher ) is that by

which the opponent is defeated. It is of many kinds such as

245
^ = Deficiency, ( 2 ) wfm = Superfluity; ( 3 )
deviation from established conclusions; ( 4 ) srofc? = Irrelevancy; ( 5 ) STsr%*n = Inability to find the (correct) answer;
( 6 ) i7T3^ = Accepting the opponents view; ( 7 ) fRte =
contradiction; and so forth. f T he N. S. enumerates 22 such
varieties including t^nro also as the last in V. 2-1 ] All
these are not being explained here for fear of unnecessary
elaboration. [ The author has selected at random seven of
the important ones for illustration ]. ( 1 ) When what is said
falls short of what is intended it is ?^p. ( 2 ) When it is
more than what is necessary it becomes srf^. ( 3 ) When it
does not conform to established conclusions the fault becomes
STqff^T. ( 4 ) When something is said irrelevant to the topic
under discussion it is SRJPR. ( 5 ) When one is not able to
find the correct answer (within a reasonable time) it is sfsfew.
( 6 ) When one acquisces in the opponent's view which is
contrary to his own it is called imjpp. ( 7 ) When one contradicts his own previous statement it becomes fgffas.
Notes^5, nfr andfstqrra1have been briefly referred to
under |<WTO in Pt. I, vide p. 109-110. The importance attached
to these defects even during the Sutra period shows that the
Nyya system was developed by Gautama more as a science of
argumentation and that even minor slips of the tongue or
memory were not tolerated in the least. All these are minor
defects which everyone should avoid even in ordinary talking
and writing, ^ T O O T ' S comments while defending the explanation of ^ q and frer^ in his w q under Stra IV. 2. 50. are
applicable here also,
It is of interest to note that while the Buddhists and Jainas
who were the toughest opponents at that time, were vehement
in their denunciation of sp, Sf and fcsit^qrR ( as also of
^ q and fffng ), they were in no way backward in using such
means in their works and in explaining a greater number of
varieties of *nf% and fro*HR ( vide wfS and s r ^ f to
mention only a representative from those two schools ). It was
only after the Gtqftfrurfar of tffter took roots in the minds of
the learned did the attention on these unimportant topics
wane. In fact the study of the N. V. Sutras and their Bhsyas

246

TARKABHAS ( ra

also was thrown into oblivion since then, from where an inquisitive student will have to dig out these classics at present.
The SJSRO'S in logic written under the influence of rfssfrsnq do
not even make a passing reference to these and K. M. also has
spared the reader by showing only a few typical examples
under ^ach and that too perhaps out of his eagerness to follow
the ^33>F? a s much as possible.

( 32 ) vM4J^K* ( Conclusion )
[148]

TranlationIn this book those topics which are extremely

useful have been explained over and over again^[e.g. au^, f%,
$Rk, ^^T!j f ^ r , ^ I W H etc ], and those that are not so useful
are left out ( without detailed explanation ]. This will not be
considered a defect as this much would suffice for the enlightenment of a beginner in Nyya.
Thus ends the work Tarka Bhs written by Keava Misra.

INDEX
to Authors and Works referred to in the Text and Notes
( f^origfft, $g?$rf, wrunftsjf )P- 18, 30, 105,
121,231,232.
P. 5.
P. 10, 28, 29, 38, 42, 43, 61, 102, 103, 106,
118, 122, 131, 132, 188, 192, 198, 211.
P. 130, 135.
: ( TOWtteTfSC. )P. 43, 170.
: ( ^fr*3TOfo95R: )P. 65, 68, 77, 95, 173.
: ( ^^TOsnBWffqL)-^P. 3,23,33,35,43, 80, 82, 92, 99,
142, 143, 147, 164, 169, 177, 185,
186,209,211.
P- 99.
P. 105.
ftrer ^ (of mivvrs)P. 3, 24, 38, 60, 62, 86,.
160, 168, 194,201,205.
P. 77.
( j ^ ^ p t r w s f t w n v l )P. 192.
^^^fq^T55pT^^^ JP. 62.
( ci^ihwwin )P. 205.
( of ^^5CTV^ )P.

40.

( of ffcw or ^qT?, also referred to as SQtrakra in

many places )P. 3, 4, 36, 59, 76, 77, 85, 93, 95,
108, 109, 110, 111, 112, 120, 131
132,166, 168, 169, 172, 173,191,
201, 222, 224, 227, 228,243, 245.
--P. 5S 9, 33.
: )P. 157 (Prbhkaras are referred to at
very many places )

: ( f toqp*i*N&f )P. 41, 71, 77, 95.

248

TARKABHS (
3P. 192.
P. 181.
( ^FKrq^tar )P. 17, 41.
( **T*rcnT-^*wrc: )P. 225.

( g&Hfcratif3, )P. 199, 217.

(tftofeRB)P. 157.
tafa: ( tfftfrow )-P. 17, 177, 189.
: ( ^RTT^rerc: )P. 4, 5, 59, 67, 85, 86, 120, 173, 174.
233.
V ( ^#CT*R: )P.

225.

-P. 74, 75.

; ( of *<Wf )P. 14, 26, 28, 29, 30, 71, 95, 173, 191,
198, 204,205,207,213,225.
sq^TSfsOP. 193, 965 197.
; ( of f T n f ^ g )P. 78; 79, 157 ( The Bhttas are
referred to at many places )
f(of%mf^ifT)--P. 20.
v ( 3f^f37ii[<wn;c'n )P. 120.
AristotleP. 74, 77, 83, 111, 112.
A Primer of Indian Logic by Prof. S. Kuppuswamy SastryP.
76, 113, 138, 164.
Ballantyne ( as quoted in Bodas edition of TSFf in B. 8. P. S)
P. 211.
De Morgan ( as quoted in Sidgwick's Fallacies )P. 112.
Indian Logic in its early stages by RndleP. 77.
Indian Logic and Atomism by A. B. KeithP. 77.
Logic as the Science of Pure Concept by Benedetto Croce, translated into English by Douglas AinslieP. 78
System of Logic by J. S. Mill ( People's Edn. )P. 66, 111.
Welton ( Manual of logic, Vol. I. )P. 112.
Whately ( Elements of Logic, Book II. )P. 111.

CORRIGENDA
( b. denotes that lines are to be counted from bottom )
KitroducioB

Read
Page No. Line
Instead of
mehod
28
method
14
33
29
17
.5 ... .. H a n d 18 .,,.'
18 and 19
. 34
44
6(b).
Put a full stop after improvements, omit
the words r'and here' and begin with
capital *W'
44
Insert the word 'Sastri' between Lai
4(b)
bahadur and Kendriyii .
3(b)
44
an
and
Text -.
thouh
i*-', .
3
though
Kb)
;
8(b)
haso
mitted

r
3
has omitted
4
19
5
s-7(b)
7
differentio
,8
differenti7
5TTJM
ll(b)
8
7
9
9
accidently ;
accidentally
22
11
^nicx
3(b)
11
12
cherefore
therefore
9
19
1 3 ,..-,
13r'
subsists
7(b)
subsist
13
l(b)
14
1?
above.
above).
14
.< I2(b)
1
15
SerafsR
17
22
21
23 :. 7(b)
25
4
25
19
26
,6(b)
(1-1-2)
( 1-1-4 )

250
Page No.
27

Line
17

28
29
30
34

4(b)
16
4
19

35
37
37
37
41
41
42
43
43
43
43
43
47
49
57

15
16
17
18
4
21
10
8
22
27
28

60
64
66
68
78
85
85
85
87
87
87
88
88

TARKABHASA (
Instead of

' -'' 4(b)

4(b)
20
15
5(b)
14(b)
2
3(b)
7(b)
10
21
9(b)
22
23
3(b)
17

Kb)

*&m )

Read
After give add!
here itself
varities
varieties
rigour
rigor
II
11
Between the and Buddhists add Bhtta
Mimrbsakas and some
i
srgr
iff

mm
m

w%.

iftc

srte

stre

( iCffTT O.f

TO5T, ( o f

vition; and so on,"

vide p.
system of logic

nition;" and
so on.
aCTTfsw
vide p. 96
system of Logic*

AGC
HGK
AKC
AKC

AFC
HFK
AJC
AJC

ire

Page No.
89
91
94

Line
6
9
2

95

21

99

ll(b)
6
12
16
1Kb)
3(b)

108
109
114
122
126
129

10(b)
10
5(b);
21
12(b)

130
130
131
139

251

CORRIGENDA
Instead of

Read

if

of

non-eternal
syllogisms
gold

close bracket}
after moment.

, eternal
sqllogism
Gold

views

view

beecome

become
After Answer add
True.
its
rect

t AA

144

6
9
8(6)
5

146
146
147

10(b)
3
2
7(b)
ll(b)
17:;
3(b)
;S(b)

150
151
152
153
155
157
157
160
161

165
166

x,-

reet

aST

which

which

ll

th

this
?^
add the heading
(10) ^snfifT^i
cognitions

3TO

:..

At the lop below

l(b)

161
163
164

it

cognition

16
>

Before S.K.S. add

Prof:
(1)

in

(ID

252
Page No.

TARKABHASA
Line
Instead of

170

175
177
177
179

8(b)
11
16
9

After ftffTO add

Bhfs

14
17

199
200
201
205
207
209
213
218
219
221
222
226
226
228
232
233

16
6
6
17
22
21
I2(b)
5(b)
1
9(b)

235

16

241
?43
245

3
10(b)
16

Bh?s

In this and following lines for sgsp

( in next page also ) wherever it occurs
read

182
185
192
194
197

Read

i?iro35E

liket his

like this

^3

Kb)
5
6(b)

Kb)

TOT

Brahmas'

Brahma's
and
delete after a

OTf
substance
m: !
it

substances .
errs |
its

HT 1 ^

^T ^T

he

the

Kb)
14
8
12

established
established
Omit comma after possessing
After (aff^) close the quotation,"
Before 'Hete' mark (b) and begin fresh
paragraph.
After cause pjjt a full stop ( . ) and omit
the dash.

acquisces

acquiesces