Full text of "Madhva Logic by Susil Kumar Maitra"

MADHVA LOGIC

BEING AN ENGLISH TRANSLATION OF THE PRAMANA-

i^*'"'

CANDRIKA WITH AN INTRODUCTORY OUTLINE OF
MADHVA PHILOSOPHY AND THE TEXT IN

SANSKRIT

SUSIL KUMAR MAITRA, M.A., Ph.D.

Lecturer in Philosophy, University of Calcut^^£>, iVIJ&Y^
Author of "The Ethics of the HwdusJ^

PUBUSHED BY THE

CALCUTTA UNIVERSITY

1936

■ _••->

* f'r-U

-■•:■■■ ,*.'>«.■■..*■■.■; . ■ .

MUNSHI RAM MANOHAR LAL

Orieolei <u 1 oieign EooA-Sellers
M- 1165; Nai Sarak, DELH1-6

V#3

_■ s —

.■■I

•A'-'/tf-'-A' c : ' >."

PRINTED AND PUBLISHED BY BHTTJ'EKDRALAL BANEHJEE
AT THB CALCUTTA UNIVERSITY PBE88, SENAT8 HOUSB, CALCUTTA.

Beg. No. 686B.— October, 1936— B.

■:.■*"

3>-

— — *-.y

■■O*

'1;" *

.jS

-*" * - '*- *

■\

PEEFACE

The accredited authority on Madhva Logic
is Jayatlrtha, and his celebrated work, the
Pramanapaddhati, is the authoritative logical text
of " the Madhvas. The Pramanacandrika is a
shorter work and follows the Pramayapaddhati
closely, reproducing the language of the Paddhati
in many places and acknowledging the Paddhati
as its authority at the end of every section.
The Gandrika however has the merit of being
a clear presentation both of Madhva and other
rival views. The present translation, it is hoped,
will give a clear idea of Madhva logical theory
and its points of agreement and disagreement
with the theories of other schools. The Intro-
duction which gives an outline of Madhva Phi-
losophy will also be of use in understanding and
correctly appraising the Madhva viewpoint. ^

Jayatlrtha is supposed to have flourised to-
wards the middle of the fourteenth century.
According to one estimate he must be placed
between 1317 A.D. and 1380 A.D. Since the
author of the Gandrika refers throughout to
Jayatlrtha's Paddhati as his source-book, and
always with profound respect, he may be taken
to be one of Jayatlrtha's younger contemporaries.

MADHVA LOGIC

He must therefore have flourished either at the
latter half of the 14th or the beginning of the

15th -century.

My sincerest thanks are due to my colleague,
Dr. Satcowrie Mookerjee, for seeing the Sanskrit
text through the Press. He has however departed
from the original Madhva Vilasa edition (now
out of print) in two respects. In the first place,
he has divided the work into chapters-^an
evident improvement in form. Secondly, he has
changed the text itself in some places. As I am
unable to accept the correctness of all the changes
he has made, some of the passages as they occur
in the original Madhva Vilasa edition appearing
to me to be quite in order, I leave the whole
matter to the judgment of my readers.

-r -> -^f

"*- * "l»-t

*r ^*

^ f* >- ," s«--r? :

CONTENTS

Introduction

• » #

PAGE

ix-xxvi

The English Transxmtion

1-2

3-6

6-11

11-12

12-14

On Liberation and the Means to Liberation
On Uddeia or Statement
On Lak§ana or Definition
On Parik§d or Examination
Definition of Pramana or Valid Evidence
Definition of Samiaya or Uncertain Cognition 14-20
Definition of Viparyyaya or Error 20-21

Memory as Valid Knowing ... 21-22

Pramana as defined by other schools 22-25

Pramana as Kevalapramdw and Anupramma 25
Four kinds of Kevalapramana ... ■ 26-36

Anupramdna as Perception, Inference and .

Authoritative Communication

Definition of Perception

Seven Kinds of Perception

Definition of Inference

Factors of Inference

Invariable Concomitance as a Condition of

Inference 63 ^ 7

Different Kinds of Invariable Concomitance 57-61
Concomitance as known makes Inference
: Possible 61-62

36-37
37-39
39-51
51-52
52-53

MlDHVA logic

' " <

PAGE

Invariable Concomitance how known

Three Kinds of Inference

Nyftya View of the Different Kinds

Inference
Fallacies of Inference
Nyaya View of Hetu and the Fallacies
Authority or Sgama defined
Defects of Verbal Communications
Words and Sentences as Costituents of

Verbal Communications
Learning of Word-meanings
Primary and Secondary Functions of

Words

Agama as Personal and Impersonal

The Evidential Value of Sgama
Refutation of Arthupatti, Upamana, Anu-

palabdhi, Sambhava, Aitihya, etc.,

as Independent Pramanas
Intrinsic Validity and Extrinsic In-

validity

63-64
64-65

65-83

83-87

87-100

100-101

101-109

103-107
107-109
109-111

111-112

112-114
115-122

123-126

* " *

*. A

INTRODUCTION

Madhva philosophy is the conceptual formula-
tion of the religious attitude of devotion or
Bbakti and rests on the idea of an essential
distinction between the devotee and the object
of his worship. As the philosophic interpretation
of the Vedanta teachings it is therefore not merely
revolutionary but also heretical. Its dualistic
metaphysics and its conception of the Lord as
the efficient and not the material cause of the
world are a direct negation of the monism of the
Upanishadic teachings. It has thus been repu-
diated by Vedantists themselves as a gross cari-
cature of the Vedanta doctrines, particularly
by the Sankarite Advaitins who reject even
qualified non-dualism as inconsistent with Vedantic
Absolutism. Madhvaism thus stands to orthodox
Vedantism as Sufism does to Islamic Monotheism.
If pantheistic Sufism is the worst heresy of
Islamism, no less is Madhva Theism as an
interpretation of Vedanta monism.

The central conception in the Madhva meta-
physics is the idea of an eternal and unsur^
mountable gap between the Lord and the world
of inanimate objects and sentient souls. The
Lord is the highest reality and has independent
being.. The world and the individual souls axs

t MADHVA LOGIC

all dependent on the Lord, but are not existent-
ially one with Him. The Lord thus is the
efficient and not the material cause of the world
(c/. NyEya). The world depends on the Lord,
but also has being outside Him. So also have
the jivas or individual souls who are subservient
to the Lord and are his eternal servants. Thus
the distinctions between the Lord and the world
and between the Lord and sentient souls are
not merely essential but also eternal.

The main points of the Madhva Philosophy
are summarised in a Sanskrit sloka the purport
of which we give here in English : —

The Lord (Hari) is the highest reality (para-
tattw). The world is real. Difference is real.
Individual souls are the servants of the Lord
(Hareranucardh). They are distinguished by
superior and inferior excellences. Liberation is
the experience of untainted innate bliss. Bhakti
or devotion together with the Lord's grace is the
means to liberation. Perception, Inference and
Verbal Testimony are the sources of knowledge.
In regard to the Lord the Vedas are the sole
evidence. The Vedas are eternal and impersonal.
■:; The above clearly brings out. the wide diver-
gence of the Madhva and the Sankarite view&f
For the Sankarite the world is a false appearance
in the Absolute and is devoid of strict reality.
For the Madhvas the world has reality, though
po$ the self-dependent reality of the Lord, For

IM-RODUCIION xi

the Sankarites, distinction is an indescribable false
appearance in the undifferenced reality of the
Absolute. Hence the distinctive reality of the
world is an eternally cancelled appearance in
Brahman. For the Madhvas, distinction is not
only real but also eternal. Hence the five dis-
tinctions between the Lord and the inanimate
world, between the Lord and the individual souls,
between one individual soul and another, between
one inanimate object and another and between
an individual soul and an inanimate object are

both real and eternal.

This brings us to the Madhva view of the
nature and constitution of the world and its
scheme of the padarthas or knowables. Unlike
the Nyaya-Vaisesikas who recognise seven kinds
of knowables, the Bhattas who recognise five
(the seven of the Nyaya-vaisesikas minus visesa
and samavaya) and the Prabhakaras who recognise
eight [five of the Nyaya-vaisesikas barring
abhdva and viiesa, plus samkhya (number),
s&dr&ya (similarity) and Sahti (potency)], the
Madhvas recognise ten kinds of padarthas, viz.,
(1) substance, (2) quality, (3) action, (4) gene-
rality, (5) individuality, (6) the qualified sub-
stantive (visista), (7) the composite whole
(amsi), (8) Power or Sakti, (9) Similarity and
(10) Absence or Negation.

>i Of these, substances are of twenty different
kinds and comprise (1) The Supreme Soul or tb**

Xll MADHVA LOGIC

Lord, (2) Laksmi, the Lord's consort, (2) In-
dividual Souls, (4) Unchanging Akasa which is
the same thing as space, (5) Primordial Nature
{Prakrii), (6) The Three Gums, (I) Mahat or the
Great Principle, (8) Ahamkara or The Principle
of Egoity, (9) Understanding (Buddhi), (10)
Mind, (11) The Senses, (12) The Infra-sensibles
(matra), (13) The Elements, (11) The Universe,
(15) Nescience (Avidya)^ (16) The Alphabetical
Sounds, (17) Darkness, (18) Kesidual Traces
and Dispositions, (19) Time, (20) Reflection
(Pratibimba). *

Qualities again are of 41 different kinds
including the qualities of the Nyaya-Vai&sikas
as well as such other excellences and deficiencies
as serenity, steadfastness, gravity of mind, fear,
shame, strength, self-restraint, endurance, valour,
magnanimity, etc.

Actions again are either moral or non-moral.
Moral actions are objects of approval or disapproval.
Non-moraJ action is physical motion.

Generality is either eternal or non-eternal.

Individualities are innumerable and are the
bases of all differentiation and distinction.

By a qualified substance is meant a substantive
specified by an adjective.

Composite wholes are again either limited in
size or of unlimited extent.

Sakti or power is of four kinds, viz., inherent
power, adventitious power, the power that is

V5P

INTRODUCTION Xl«

unthinkable, and a word's power of meaning or
referring to an object.

Similarities are innumerable and are functions
of single objects as determined by their relations
to other objects. Thus though a similarity holds
between one object and another, it is a function
only of one and not of both.

Absence is of four kinds, viz., Antecedent
Absence, Emergent Absence, Absolute Absence
and Reciprocal Absence. Of these the first are
other than the locations they characterise. Not
so reciprocal negation. It is the same as its locus,
the negation being non-different from the entities
which negate each other. Further it is either
eternal or nou-eternal. As negation of eternal
entities it is eternal, as negation of non-eternal
entities it is non-eternal. This follows from such
negation being non-different from the entities
which so negate each other. Absolute negation
is the negation of what never, nowhere exists. It
is thus the negation of the unreal or the imaginary.
Though the entity which it negates is unreal, the
negation itself as the absolute absence of the unreal
is real. Thus absolute negation is the real absence
of the absolutely unreal.

The Madhva view of the Lord and the indivi-
dual soul presents many points of contact with,
as well as of divergence from, the Nyaya view. Atf
with the Naiy&yikas, the Lord, according to the
Madhvas, is the efficient and not the material cause

Xiv MADHVA LOGIC

of the world. The Lord further is independent,
all-pervading and is the subject of such qualities
as omniscience, etc. Laksral is the power of the
Lord. The jiva or individual soul, on the other
hand, is atomic (contrast Nyaya), is existentially
separate from, though dependent on, the Lord, is
ignorant, in bondage, etc. Further the jiva stands
to the Lord as reflection (pratibimba) to the
original. Through the knowledge of the Lord the
karmic potencies of the jiva wear away, and this
prepares the way to. liberation. Bhakti together
with the Lord's grace is the cause of Liberation
which consists in the experience of pure, inherent
bliss. Prakrti or primordial nature is the cause
of bondage and is the root of beginningless

nescience. Nescience itself is a positive category
and is the source of the two kinds of ignorance,
viz., ignorance as regards one's own nature and
ignorance as regards the nature of the Lord.

The Madhva view of the twenty kinds of
substance includes, it will be noted, not merely
the nine different kinds of the Nyaya- Vaise$ikas
but also .those of the Sankhya Philosophers.
Elementary Akasa of the M&dhvas, e.g., is the
same as the ikasa of the Nyaya-VaiSesikas, while
unchanging Aka3a is only the space or dvik of the
latter reintroduced under a different name. Thus
we have all the nine of the latter, viz., the five
elements, besides space, time, mind and self. But
in addition to these we have also some of the

INTRODUCTION XV

Sankhya metaphysics. For example, Prakrti, the
Gums, Buddhi, Ahamkdra, mind, the senses
(indriya), the infrasensibles {maira corresponding
to tanmatra), etc., are all Sankhya padarthas. To
these of the Nyaya-Vaisesika substances and the
Sankhya padarthas, the Madhvas add some of their
own such as Avidyd (Nescience), Pratibimba (Eeflec-
tion), the Alphabetical Sounds and Darkness.

As regards gums as qualities, it will be noted
that they are not the same as the three gunas
which are substances. The gunas as qualities are
attributes while the three gunas are substantive
reals. The gunas as qualities, it will be further
noted, include not only the Nyaya-VaiSe§ika
qualities but also many moral attributes of the soul
such as serenity, mental gravity, magnanimity, etc.
Tbl Nyaya-Vai^esikas will regard these latter as
compounds of certain primary qualities of the self
such as pleasure, pain, attraction, aversion, etc

The Madhva classification of actions

and the morally indifferent or neutral alsj
clear departure from the Nyaya-Vais*
For the Nyaya-Vaise§ikas willing is a qi
soul and not an action — a quality pj
attraction or aversion as its condition anl
being the object of moral judgment.
ism however willing is regarded as a kind of acting
and therefore as a species of the genus which
includes physical motion as well. ._....:

* . The Nyaya-VaiSe§ika view of generality is also

xvi msbhva logic

similarly modified by the Madhvas. The Naiya-
yikas consider nityatm or eternality to be part of
the definition of 'generality,' so that a 'generality'
which is not nitya is no generality. The Madhvas
reject this view and subscribe to the conception of
nitya and anitya generalities. Thus Brahminhood,
manhood, etc., are non-eternal generalities, since
their individual substrates are non-eternal. A man
may become a Brahmin through the practice of
penance and self-mortification just as contrariwise
one may lose Brahminhood through misdeeds, fio
also manhood may be lost in a subsequent birth,
it being possible for a man to be reborn as an
animal in a subsequent rebirth. Thus we must
suppose non-eternal generalities in such cases.
But a generality like that of individual self-hood
(fivatva) is eternal, for no jiva ever ceases to be£

And what is true of generality also holds of
particularity. Here also we must recognise,
according to Madhvas, both eternal and non-
eternal particularities. Thus the particularity of
an eternal spirit like the Lord is itself eternal -,
while the particularity of a non-eternal thing like
a jar is non-eternal. The Nyaya-Vaisesikas will
say that the particularity of a non-eternal thing
being due to the particularities of their eternal
constituents, no separate particularity for the.
whole as a compound need be assumed. But tbis
view does not appeal to the Madhvas. . g

In place of the samavaya relation of the Nyaya* -

-'4

INTRODUCTION XV11

Vaisesikas, again, the Madhvas will have the two
padarthas of the viiista or qualified substantive and
the amsl or composite whole. These two between
themselves comprise, according to the Madhvas,
every case of the so-called constitutive relation of
samavaya.

Sakti, power, and Sadrsya, similarity, are not
admitted as distinct padarthas by the Nyaya-
Vai&sikas. They are however recognised as such
by the Prabhakara Mimamsakas and the Madhvas
agree with the Prabhakaras in this respect. 3akti 9
however, according to Madhvas, includes, besides
the power in words to refer to their meanings or
objects, the unthinkable power which exists in
the Lord alone in its completeness and only
partially and in different degrees in other beings,
the adventitious power 'which is generated in an
idol or image through the inspiring influence-^
the worshipper's devotion and the inhe^nt <>r
innate powers of things. Similarity agaiji , |£
eternal or non-eternal like generality and part^
cularity- Thus the similarity of jivas or individual
souls and other eternal substances such as the Lord
is eternal, but the similarity of non-eternal things

VII

.like jars, cloths, etc., is itself non-eternal.
['- As regards Abhava or Absence, the Madhvas
hold that it has reality though the pratjuogi^
cpuntei>entity of the ahpwe in some eap^gr
■'f.Jwfc.. or unreal (e,flf., ?Dt absolute absence).^||g
A ibe NJ^ygyjkas hojsevfi* -Ahhaxa has

:■■: ■'•:? l V-~ '■■ ;;':-v:.v'- ■■■ : .,~, ■■■'■■ ■. ■■. '• . .'.

XViii MADHVA LOGIC

or objectivity but not saliva or reality. Abhava
presupposes reality (bhava) being adjectival to it,
but is not itself reality. Thus reality (bhavatva)
appertains to the six positives or bhava-paddrthas,
the first three (substance, quality and action) being
real through the universal of being inhering in
them (sattayogena sat) while the second three
(generality, particularity and inherence) being real
through relation to that in which reality inheres
(ekartha-samavaya) . For the Madhvas however,
'absence' or negation is a form of sattva or reality
just as is 'presence.' According to them, padarthas
include both the real and the unreal, the latter
being a paddrtlia or knowable without reality (e.g.,
sky-flower, hare's horn, etc.). Reality again
is either independent or dependent reality, the
former being the Lord Himself and the latter
including all positives (bhavah) and all negatives
(abhavah) other than the Lord. Thus negation,
according to the Madhvas, is a form of dependent
reality though the entity negated in the case of
absolute negation is the unreal or the imaginary.
The Madhva, the Sankara-Vedanta and the Naya-
Vai3e§ika views of absence thus present many in-
teresting points of agreement and difference. For
the Sankarites 'absence' has objectivity like its
opposite 'presence' and as such presupposes the
reality of the consciousness in which it appears. It
however does not affect the latter just as the snake-
appearance does not affect the nature of the rope

INTRODUCTION XIX

which is its substrate. Thus absence as objective
appearance presupposes a substrate of reality but
is not adjectival to the latter. For the Naiyayikas
however absence as objective not merely pre-
supposes reality but also determines or characterises
it. Hence absence though itself not a form of
positivity yet both presupposes and infects the
' latter. For the Madhvas however 'absence' is
itself a kind of dependent reality to be distinguished
from the kinds which positively fill experience.

We shall now close our survey of the Madhva
Philosophy with a brief statement of the distinctive

features of the Madhva Logic.

Pramam, according to the Madhvas, is either

kevalapramana or Anupmmana. Kevalapramana is

the knowledge which has pramanya through itself
as valid knowledge of objects. Anupmmana is
pramam as the conditioning process or activity
which gives rise to self-validating knowledge-
Thus kevalapramana is prama or valid knowledge
regarded as being its own pramdna or evidence
(cf. Eamanuja, Prabhakara), while Anupramdna is
evidence through conditioning or causing the result-
ing self-evidencing knowledge.

Anupramdnas are of three kinds, viz., Percep-
tion, Inference and Verbal Testimony.

Of these, Perception is of seven kinds, viz.,
the five kinds of external perception by the external
senses, internal perception by the mind and
perception through the Witnessing Intelligence

•A s

- o

XX MADHVA LOGIC

which is the seventh kind of Perception. Smrti, ■,
Recollection, is a form of internal perception : it
is an immediate presentation of the past through
the instrumentality of the mind. The impressions-:^
or residual traces of the past experience are the- j
connecting link between the past experience and; ;j
the present mental function. The recollection is- 1
the insertion of the past into the present (c/. ; :|
Bergson).

The Madhva view of witnessing knowledge
(gaksijridna) as a form of perception is peculiar.
The knower itself acting as an instrument of
knowledge is the Saksl or Witnessing Intelligence^
and the knowledge which results through the

instrumentality of the latter is perception. 'Tttei

objects of such perception include the intrinsi
nature of the self, the self's properties or attributes
such as pleasure, etc., avidya or nescience, the^
functions of the mind such as the cognitions of
the external sense (which are also cognised by%
the mind), pleasure, pain, etc. Thus what other <
schools will regard as objects of internal perception*!
are here regarded as being . perceptions of
witnessing subject. But as perceptions sucfc
witnessing cognitions will be generated events!
and will thus lack the timelessness involvedui$S|
the witnessing consciousness of temporal mental
events as temporal. This is why Sankarites
that the witnessing consciousness is a kind
pra tyak§a or perception . According to

i£*!

INTRODUCTION XXI

it is primary (anubhutirupa) but is not a generated
cognition being nitya or timeless.

As regards inference, the Madhvas bold that
the vyapii which mediates inferential reasoning
may be one of four kinds, viz., samavyapti (a
symmetrical invariable relation — corresponding to
Hamilton's U propositions) , vimnavyapti (an
asymmetrical invariable relation — corresponding
to A propositions), of the form of mutual exclusion
or parasparaparihara (corresponding to E proposi-
tions), or parasparasamavesa, mutual overlapping,
along with parasparaparihara, mutual exclusion
[answering at once to the three propositions,
(i) In some case at least where A is, B also is,

(it) In some case at least where A is, B is not, and
{Hi) In some case at least, where B is, A is not.
The relation, e.g., between 'being a male' and
? being a cook ' illustrates this form.] Further,
according to the Madhvas, co-presence of the
prolans (hetu) and the probandum (sadhya) either
temporally or spatially is not necessary for valid
inference. Hefcce the existence of the probans
in the subject of the inference need not be always
insisted on. When, e.g., one infers rain on the
top of the hill from the perception of the fullness
of the rivers at the base, the mark or hetu is the
"fuHness of rivers," and that which is inferred
hj< means of this, hm#, M., the sadhya ^
prob(mdmn r is "rain.-; But the place or
z&m m&k wfmdd^h the #afe ? av i^i

XX11 MADHVA LOGIC

of the inference is the hill-top, while the mark
or hetu, viz., 'fullness of rivers,' is observed not
on the hill-top but at the base of the hill. Hence
the Madhvas conclude, what is necessary for
producing valid inference is not the observed
copresence of probans and probandum, nor the
observed existence of the probans in the inferential
subject, but observation of the mark in any
suitable place and time (samucitadeiadivrtti) . The
mark in the above inference exists in the present
time while what is inferred therefrom, viz., 'rain'
belongs to the past. Similarly the mark is observed
at the base while the rain which is inferred belongs
to the hill-top.

It may be pointed out however that this is no
innovation of the Madhvas and cannot be regarded
as one of the Madhva contributions to logical
theory. The point was anticipated by the Mimam-
sakas long before the Madhvas. Parthasarathi
in the " Nyayaratnamala " discussing the nature
of vydpti rejects the view that as a condition of
inference it implies the copresence of the hetu
and the sadhya as an indispensable condition.
Smoke, e.g., which is rising up in the sky above
proves fire not in the sky above but on the ground
below. What is necessary therefore for inference
is not spatial or temporal copresence of hetu and
sadhya but simply fixed relation or niyama between
them. Thus the way in which a thing is cognised
in fixed relation to something else, in that way

INTRODUCTION

does it produce the cognition of its ct
when cognised again (c/. " Nyayaratnamala"
p. 57, Chowkhamba edition, 1900). The so-called
Madhva contribution in this respect is therefore no-
thing but a re-statement of the Miraamsaka view.

The Madhva rejection of vyaiirekivyapti as a
condition of inference is also no innovation of
the Madhvas. The same view is also taken both
by Mimamsakas and Sankarites long before the
Madhvas, and the Madhva view in this respect
is only a reproduction of earlier views. The
Madhvas however may legitimately claim their
classification of inference to be an improvement on
earlier logic. Thus, according to them, inference
is either from cause to effect or from effect to
cause or from one thing to another not related
to it as cause or effect. For the Buddhist such
non-causal relation is nothing but the relation
of co-essentiality between genus and species
(tadatmya). But Naiyayikas hold that there are
other such relations besides co-essentiality. The
Madhva view of non-causal inferences combines
in itself both the Buddhist and the Nyaya view-
points and has thus the merit of being a simplified
solution of the different issues.

As regards Agama, ct authoritative verbal testi-
mony, the Madhvas hold that it is both personal
and impersonal. Thus the Vedas are authorita-
tive evidence though devoid of a personal source.
But so also are the personal communications

xx i v MADHVA LOGIC

recorded in the Mahabharata and other sacred
works. In connection with Igama, the Madhvas
discuss the question whether words mean common
characters or denote individuals, and the Madhvas
decide for the dual character ot the objective
reference with the reservation however that in
the case of nouns or substantives the primary
reference is to an individual or individuals, while
in the case of adjectives, verbs, etc., it is some
attribute or character that is primarily meant.
The psychology of learning word-meanings is also
discussed by the Madhvas in this connection,
and the view which they advocate in this respect
is that the process of learning word-meanings
consists in a course of parental guidance by means
of uttered words accompanied by gesture-indic%yi
tions of the objects meant. The Madhvas reject
the Nyaya view of naming as a process o|:
upamuna based on the instruction of elders. ...^

In regard to validity and its opposite invalidity!
the Madhvas hold independent views though
apparently agreeing with the Mimamsaka theory
of intrinsic validity and extrinsic invalidity. Thus
intrinsic validity, the Madhvas argue, is intru^|
sicality in respect of utpatti (origination) or intrjag
sicality in respect of subjective acceptance 1S|
recognition (jnapti). Intrinsicality in respect
origin means that the validity arises from
same conditions as the cognition itself which;*
characterises. And intrinsicality in respect

:tt

. ?ft

INTRODUCTION . XXV

subjective recognition means that the agency that
cognises the cognition is also the agency that
cognises the validity of the cognition. Now as
regards intrinsicaUty in respect of origin, the
Madhvas agree with the Mimamsakas and reject
the NySya view of an additional efficiency in the
causal conditions as a condition of the validity
of the valid cognition. As regadrs subjective re-
cognition of the validity, the Madhvas hold how-
ever that intrinsicaUty here arises from the fact
that the witnessing Intelligence that cognises the
'..cognition is also the agency that cognises the
validity. This is a clear departure from the
Mimamsfc view, according to the Mimamsakas
(BhSfctas) neither the cognition nor its validity
being cogqised by any witnessing Intelligence,
both being cognised inferentiatly by the self from
the mark of knowoness in the object- As regards
invalidity, again, the QXfirinsicality m.y.r^spect
of origin consists, according to Madras, inits
wising foocn the presence of certain defects in
addition to the conditions, of cpgnition, while
^trittsicaUty in respect of invalidation or sub-
jective rejection consists in the cognition itself
being cognised by one agency, viz., the Witnessing
Intelligence and its invalidity being cognised
otherwise, i.e., inferentiaUy from the mark of
its practical failure, This also is an evident
departure from the ordinary MimamsS view ac-
onr<\\ n « to whir.h iasalidation comes either through

XXVi MADHVA LOGIC

the perception of defects in the causal conditions
or through the consciousness of discrepancy with

other experiences. '' J

It may be added here that the Madhva theory of /
falsity comes nearer the Buddhist than the Nyaya.
or Sankara-Vedanta views. Thus the Madhvas >•
reduce the false to the level of the imaginary and ,
the unreal so that what the illusory experience
apprehends is an absolute nought and not any
elsewhere, elsewhen reality (as Naiyayikas say),
nor any indescribable positivity without reality
(as Sankarites say). Further, correction as ab- i
solute negation is the cancellation or rejection of ?

this absolute unreality. Thus absolute negation " ; k

is the negation of a sheer nothing and not that 7.3
of an elsewhere, elsewhen real something as |
Naiyayikas say. For the Naiyayikas negation :A
is always the exclusion of a real something from
some real locus so that a negation of the unreal^
is sheer non-sense. The Judgment : '* The square- *
circle is not" is, according to Naiyayikas, equiva-
lent to the Judgment : "The square is not a circle/* q
though expressed differently. For the Madhvas
however the object of absolute negation is the g
unreal or the imaginary so that the Judgment does
not assert the exclusion of circle from square (as |
Naiyayikas say) but expresses the absolute unreal- ■
ity of a square which is a circle as well. 1

- , : . - ■■ • M

' • • - 1 -. f v- , . . ;

MADHVA logic

ENGLISH TRANSLATION

OF THE

PRAMANACANDRIKA

m
a . j *

- - pakt i ;

Reverence to the God with the Horse's neck,
the God who has LaksmI as his consort and who
incarnated Himself in Rama— the God of Hanu-
mana, in Krsija — the God of Bhima, and
in Vedavyasa— the God of Madhva. Om HarL v
,. Having touched the lotus-feet of the Lord of
Lak§mi and also those of my Guru or Preceptor,
I. proceed to write this 'Pramanacandrika 1 for
the easy comprehension (even) of young, immature

learners.

Everybody on this earth desires that happiness
alone shall be his lot and that not even the smallest
unhappiness shall ever mar his life. This is the
moksa or liberation that is sought by all. Since
this freedom or liberation comes only from the
knowledge of the absoluteness and independence

MADHVA LOGIC

-**v:

'■A'

:-u

of the Lord and the dependence or subservience
of everything else, it behoves every seeker of this J
freedom (moksa) to understand all things in this
way as being essentially subservient to or
dependent on the power of the Lord who is in-
dependent and absolute. Thus the commentator
observes, he that realises all these that are
dependent as being subject to the control of the
Lord becomes liberated from the bonds of the
empirical life. The knowledge of the dependent y
and the independent however comes from valid '""1
cognition and this is the reason why this
particular treatise has been undertaken with the
object of ascertaining the nature of valid cogni-
tion. Even though the master Jayatlrtha has
elaborately expounded the distinguishing marks of j
valid cognition and the rest in such works as the^l
Paddhati, etc. (Pramawpaddhati), yet, inasmuch ;|
as these works are not easily intelligible to ;
persons of feeble intelligence on account of theif
deep and thoughtful language in which they are ;
expressed, this treatise has been undertaken with|
a view to make the doctrines intelligible to thestf]
readers of average intelligence. And thus this
undertaking is not superfluous even though it
discusses most of the topics already discussed i%:
these other works and gives besides a brief account
of some of the objects of valid cognition as well.
Since the (scientific) knowledge of Praman
and other allied things presupposes the tria

",'

PRAMAtfACANDRIKA 3

statement (uddeia), definition (laksana) and
examination (pariksa), the definition of statement
and the rest is therefore first of all set forth.

UddeSa is the statement or verbal indication
of the subject-matter by means of its name only.

i In this definition the word 'udde&a' stands for
what is defined and the rest constitutes its defini-
tion, viz., the words ' verbal indication of the
subject-matter by means of its name only.' This
procedure (in regard to the thing defined and the
definition thereof) will also be observed in all
other cases (of definition) that will come up later
on. If we say that a sound as such is a verbal
indication, the babbling sound of the Ganges will
rank as a verbal indication and thus our definition
will be too wide. To exclude such cases we
include the word indication in our definition.
A verbal indication implies indication by
alphabetical sounds (and not by sounds as such

'" which may include non-alphabetical sounds such
as the babble of a river) . But if we stop here

:- and say that an alphabetical sound as such is a
verbal indication, the alphabetical sounds 'the
son of a barren woman' should pass as a verbal
indication. Hence to exclude such nonsensical
combinations of alphabetical sounds the word ;
* subject-matter ' (in ' indication of the subject-

i? matter') has been included in the definition.

% (The words 'son of a barren woman' ar^ift^^
iT,^««« M «f anvthitisr and therefore indicate

i : MIDHVA LOGIC

subject-matter.) But if we stop here and accept %
' verbal indication of the subject-matter ' as a ^
logically complete definition (of uddeso), the |
* caw ' ' caw ' of the crow will pass as a statement
or Uddeia. (The 'caw* 'caw' of the crow is a
combination of alphabetical sounds and it also
indicates something that really exists, viz*» the
crow's voice.) Hence the words 'by the name*
in the definition which mean * by the words of
the sacred language.' Even so, the definition ,
is too wide applying as it does to a sentence like
'The earth has the character of smell' which
amounts to a definition (and not to a verbal indica-
tion of the subject-matter only). To exclude such
cases, the word ' only * has been added (in ' by the
name only '). The meaning is : the name which
indicates the subject-matter in a statement is not "|
used with a view to bring out the distinguishing %
marks of the subject-matter. (The name is used
only to indicate the subject-matter and not to de-
fine it.) In the case of the sentence 'The earth
has the quality of smell ' which amounts to a defi-
' .nition, though the subject-matter is indicated
by means of the words (names) of the sacred i
language, yet, since the words have not been
selected without reference to the marks which
distinguish or define the subject-matter, the
sentence cannot rank as a mere statement.
Hence our definition of statement (as givefi|
above) is not too wide. Hence we conclude : a

-S

.11 \

"as

'J'u

PRAMA^ACANDRIKA 5

statement consists in the indication of a subject-
matter by means of such words (names) of the
Sanskrit language as do not refer to the marks
which distinguish or define the subject-matter in
question .

Some hold that the word 'only' has been
added (in the above definition of Uddeia) with a
view to exclude only such definitions as * The earth
has the quality of smell.' For (according to them)
an Udde£a is a statement of the subject-matter
without reference to its specifying or uncom-
mon properties. Thus in the sentence (which
is, in effect, a definition) .' The earth has
the quality of smell,' since the subject-matter
is indicated by reference to an uncommon
property, viz., the possession of odour, the
condition of indication of the subject,
without reference to any uncommon
of it is wanting, and thus our defi]
not applying to it) is not too wide,
be said that in the case of the earth

things, statements in this sense are impol
so far as such statements will have to
their respective subject-matters through the un«
common properties of earthiness and the rest that
distinguish them. For what is meant (by a verbal
indication without reference to the uncommon
properties of the subject-matter) is merely that the
statement should not contain any reference to dis-
tinctive or uncommon properties other than those

6 MADHVA LOGIC

that constitute the bare property of being the sub-
ject-matter in question.

This however is not a correct interpretation
of our definition, for the definition as thus inter-
preted will apply to sentences which are in the
nature of definitions such as ' The earth has the
property of earthiness ' and thus will be too wide.
('The earth has the property of earthiness ' is a
verbal definition and not a statement of a subject-
matter.) The sentence ' the earth has the property
of earthiness ' indicates a subject-matter (viz., the
earth) without reference to any uncommon proper-
ties other than those that constitute the bare i
property of being the subject-matter in question f
(i.e., the property of earthiness).

(Having explained the nature of Uddeia, :i
we now proceed to define Laksana or definition.) > .:'"-

A defining mark (laksaya) is an attribute
that exists only in the thing defined (and not in \
anything else), This means that a defining mark v
is an attribute that exists in every instance of
the thing defined and does not exist in anything . ■
else. Thus in the case of the cow, the ' posses- \£
sion of a dewlap' serves as a defining mark ■-.-]
as it exists only in (all) animals that are v
cows and does not exist in animals that are not

COWS. .-•". ;■*■£

* llYf,ll

If we say that an attribute as such is al
defining mark, the ■ possession of undivided hoofs *
will pass as a definition (of the cow) and thus

prama^tacandrikI 7

our definition of a defining mark will be too wide.
(Possession of undivided hoofs is an attribute,
but it is not an attribute of the cow which has
cloven hoofs.) Hence we insist on the existence
of the attribute in the thing defined.

If we stop here and rest content with saying
that a defining mark is an attribute that exists
in the thing defined, then the ' possession of
mixed colour ' will pass as a definition of the cow
and thus our definition of a defining mark will
be too wide. (Mixed colour exists in some cows
but not in all cows and therefore cannot be a
definition of the cow.) To exclude such attri-
butes (as do not exist in every instance of the
thing defined) we say the attribute (that is, a
t defining mark) must exist in every instance of

the thing defined. -

But this also does not suffice, for 'possession
of horns' may pass as a defining mark (of the
cow) as thus interpreted and thus our definition-:
becomes too wide. (Possession of horns can*:
not be a defining mark of the cow, for though
this attribute may exist in every instance
-of a cow, yet it exists also in other animals
such as the goat, the dear, etc) To exclude
such attributes (as exist both in the thing;

: : defined as well as other things) we say the attri-
bute (that is, a defining mark) must exist only

: ^:ii<all) the instances of the thing defined (and itf
nothin!

r — / »

?■;:■

■ ■■;

■1

3 mIdhva logic

What, then, is the purpose or end (prayojana)
subserved by the knowledge of a defining mark ? ,|
The purpose or end subserved by the knowledge i
of a defining mark is the differentiation of -■
the thing defined from all other things of a
homogeneous or heterogeneous nature as also the
correct use of terms (without a too wide or too
narrow meaning) . A thing is said to be homo-
geneous (sajatlya) with the thing defined
when it is specified by the next higher class that
subsumes the defined thing under itself. (This
means that the homogeneous is a species co- v
ordinate with the thing defined and subsumed. |
under the same immediately higher genus.) \|
A thing is said to be heterogeneous (vijatiya)
with the thing defined when it is not specified^
by the immediately higher genus that subsumes
under itself the character of the thing defined.,!
Thus in the definition of the * cow,* 'the essence J
of being a cow ' or ' cow-ness ' constitutes th^f|
character of the thing defined. The immediately |
higher genus comprehending this character (ofy
'cowness') is .' animality. ' Therefore the horse
and other animals which are characterised by
this generic character of animality are homo-
geneous (sajathja) with the 'cow.' (Contrary-
wise,) the jar and other things which are non*g
characterised by this generic character (o&
'animality') are heterogeneous (vijatiya) with
the 'cow.' By 'an immediately higher genra^T

-i*i

i-^

PRAMA^ACANDRIKA. 9

or ' next higher genus * is meant a genus which
while not including any higher class inclusive of
the thing defined is yet inclusive of the thing
defined. Thus in the example of 'animality/
since the class of animals is inclusive of the class
of cows without including the class of material
objects which also includes the class of cows, the
animal-class must be understood as the im-
mediately higher genus in relation to the cow-
class. These considerations (concerning the
nature of the next higher genus) leave no room
for the objection that all things being included
under one all-inclusive class, viz., the class of
knowables and the like, are all homogeneous with
one another and that therefore there is no real
heterogeneity anywhere. (Since all things come
under a common all-inclusive class, viz., the
class of knowables, they must all be said to be
of the same class or genus, i.e., homogeneous.
How then can you sensibly talk of the hetero-
geneous and of one thing being heterogeneous
with another? The answer to this objection is
furnished by the definitions we have given above
of the homogeneous and the heterogeneous as
based on the concept of the next higher genus.
When a thing is included under the next higher
genus or class of the thing defined, it is said to
be homogeneous with the thing defined. SVhen
a thing is iiot so included, it is said to be
heterogeneous.) In tins way in all other defini-

10 . MADHVA LOGIC J

tions the homogeneous and the rest are to be .■
understood.

(Having explained the nature of a valid
logical definition, we shall now proceed to explain
what constitutes the opposite, i.e., an invalid
definition or non-definition.)

An invalid definition or non-definition is
the negation of a logically valid definition.
Hence it is the negation of "that which,
existing in every instance of the thing
defined, is non-existent in things other than
the instances of the thing defined." As a .';
negation, it is of the nature of a negation ; ;
of a qualified thing. . A qualified negation ..;;
or a negation negating a qualified thing may
be of three kinds, viz., (a) a qualified nega-
tion negating the thing qualified, (&) a'.|
qualified negation negating the qualification ;|
of the thing, (c) a qualified negation negating
both the thing qualified and the quali-
fication of it. (An invalid definition is a qualified
negation, because it is the negation of a valid %t
definition which consists of a substantive and an
adjectival part, the substantive portion being
' that which is non-existent in other things ' and 28
the adjective qualifying it being ' while existing 'k
in every instance of the thing defined.' Thus
a non-definition, as being the negation of this
substantive qualified by the adjectival portion,
is a qualified negation.) Thus the non-definition

fcRAMAtfACANDRIKA. 11

"The cow is an animal possessing horns " is a
qualified negation negating the substantive
part of a valid logical definition. (It negates the
condition that 'the defining mark must not
exist in anything other than an instance of the
thing defined.' ' Possession of horns ' is existent
also in animals which are not cows.) Again the
non-definition * The cow is an animal possessing
mixed colour ' is a qualified negation negating the
qualification (that characterises a valid defini-
tion). (The qualification 'existing in every
instance of the thing defined ' is negated in this
case as every cow does not possess mixed colour.)
Lastly, the non-definition ' The cow is an animal
possessing undivided hoofs ' illustrates a qualified
negation negating the substantive as well as
the adjectival part of a valid definition. ('Posses-
sion of undivided hoofs' is present in animals
that are not cows and absent in cows. Thus it
negates the substantive 'absence from other
things ' and also the adjective ' existing in every

case of the thing defined.')

(Having explained the nature of a logical
definition and its opposite, we shall now proceed
to explain the nature of pariksa or examination.)

Examination or sifting of evidence (pariksa)
is mentally reflecting on the cogency or
otherwise (of the evidence that has been
adduced). Too wide use (ativyapU) consists in
the use of a defining mark that exists in things

■■■'.•

12 madhVa LOGlO

other than the thing defined. Too narrow use
(avyapti) consists in the use of a defining mark
that doe£ not exist in a part of the extent (i.e.,
in some instances) of the thing defined. Ab-
surdity (asambhava) consists in the use of a
property as a defining mark that does not exist
in any instance of the thing defined. That
which is defined by the defining mark is called
the laksya or thing defined.

(Let us now proceed to define the
subject-matter of this work, viz., Pramana.)
Pramana as such (i.e., pramana in general as
distinguished from any particular kind of
pramana) may therefore be first of all defined
here. Pramana (we hold) is that which agrees,
with the nature of the object (known). This
means that the essence of pramana consists in |
making the cognitum an object (of cognition) in
the form in which it actually exists. There are M
many things to be said here, but as this treatise 4S
is meant for young learners and as (more) in- |
telligent people may know all these from the
work called the Paddhati, therefore they are not
mentioned here.* The same observations hold

* The Pramdnapaddhati observes that pramana is here so defined

as to apply both to tbe cognitive process and the knowledge that results

therefrom. The cognitive process conduces to the apprehension of the

cognitum as it actually exists and is called Anupramapa. The resulting

knowledge also apprehends the cognitum as it actually exists and is,

called kevatopramana.

{Pramdpapaddhati t Madhva ViWsa Edition* p.

■'/}■

■Wk ■■■-

1L^ V V

also in respect of other matters that will come up
later on (in this treatise).

In the above definition of Pramana, since
the property of apprehending the object is common
to valid as well as doubtful and erroneous
cognitions, the qualification ' in the form in
which it actually exists* has been included. If
we had defined Pramana merely as that which
abides in the form in which it actually exists,
then our definition would apply also to the
cogniser and the cognitum (since these also
abide in the form in which they actually exist), :., ::
and thus will be too wide. Therefore we say : it
must apprehend the cognitum as well (and not
merely that it should abide in the form in which
it actually exists) . Here ' apprehending the
cognitum* means 'apprehending the cognitum! * x
immediately as well as mediately* (through the ;■'..£■■+}
help of a cognitive process). Hence our definition '* .;.;|j
as applying also to the cognitive processes of ,j||
perception and the rest (as also to the knowledge v;J

that results therefrom) cannot be said to be too .. : v
narrow. Nor can this be said to be an illegiti-
mate extension of the meaning of the term
Pramana, for such extension is quite unexception- ■;\
able (inasmuch as the term Pramam is, as a
matter of fact, used in both the above two senses
of the cognitive processes and the resulting
: knowledge)'. " " J* ■■' '' J

The subject who cognises a valid cognition

MADHVA LOGIC

■A

is called the cogniser or knower. The object that
is cognised by a valid cognition is called the
cognitum or object known. Knowledge of the
object as it actually exists is called valid knowledge
or valid cbgnition. If valid knowledge were not ■
defined as a form of knowledge, our definition
would be too wide as applying also to the
processes which are only instrumental in the
production of valid knowledge.* Similarly, if .
valid knowledge were not defined as apprehension
of the object as it actually exists, our definition "j
would also be too wide as applying to doubtful

is?

cognitions and the rest. .^1

What, then, is the essence of a doubtful;^!

cognition {saw&aya) ? It may be said that it cannak||
be said to consist merely in an indefinite appre£!|
hension for this amounts to a mutual dependence '::::{
(a circular definition). Thus (one may argue) io||j
so far as a certain cognition is a definite apprehe«?J||
sion the definition of a doubtful cognition as mergg^
indefinite apprehension amounts to a circular^
definition in asmuch as a certain cognition is the
other of a doubtful cognition and a doubtful cog-

* ~*5

* A distinction is drawn betweeu Pramd, valid knowledge, and
Pramana, the instrument of valid knowledge. Pramana signifies both
the source of the knowledge and the knowledge itself (as correct appre-
hension of the object). But Pramd means the knowledge only and
its instrumental means. Thas PramS must be distinguished from
processes of intellection that bring about the result of valid knowli
but Pramana may be used alike for the cognitive processes and
knowledge that results therefrom.

prama^acandrikI

nitioii (defined as indefinite apprehension) is like-
wise the other of a certain cognition. This
argument, we hold, is not a sound one. An
indefinite apprehension is in reality the correct
definition of a doubtful cognition. There is no
circle involved in this definition (as is contended
by the opponent). For by the term 'indefinite
apprehension ' is here meant a cognition that
appears clothed (bathed) in the numerous
mutually incompatible forms that manifest them-
selves in (float on the surface of) one single thing.
If we had defined doubtful cognition as that
which is a cognition our definition would be too

wide as applying also to the case of the (certain)

cognition 'There is a jar here.' Therefore
we say *it must be clothed in many forms.'
Even so however our definition would be too
wide as applying also to composite or collective
cognitions such as the simultaneous cognition
of a man and a post, or of a jar, a piece of cloth,
a pillar and a pitcher, etc. To exclude such
composite (certain) cognitions we say ' there must
be one single thing' (in which the different
forms are apprehended). But even then our
definition remains too wide as applying to cog-
nitions like ' This tree is of the Sim^upa species,'
'The jar is a substance,' etc. (In these also
there is cognition of different forms in a single
thing.) Therefore (to exclude these), we say 'the
different forms must be mutually incompatible. 1

16 MADHVA LOGIC

But even so, our definition remains too wide
as applying to erroneous cognitions like "This
is silver/ (In the case of the cognition
of silver in the locus of a mother of pearl, there
are two incompatible forms, viz.* the form of ,
mother-of-pearl and the form of silver, and these
forms refer to one single thing, viz., the mother-
of-pearl, and yet the illusion of silver in the i
mother-of-pearl is not a doubtful cognition.) J
To exclude such erroneous cognitions we say (not |
merely that there should be numerous incom-
patible forms but also) that, - the numerous incom-
patible forms should also manifest themselves^
as incompatible. 1 (In the case of the illusion*^
the form of silver manifests itself while that of
mother-of-pearl remains non-manifest to the
cogniser.) Thus (since in the doubtful cog]
the different incompatible forms are held also tte
present themselves as incompatible forms (refett|f
ring to one single thing), our definition is noij-
open to the aforesaid objection (of being too^

wide). -

Some hold that this doubtful cognition
arises, with the absence of its solvent as an
auxiliary condition, from five different cau
viz., (1) (cognition of) a common character,

(2) (cognition of) an uncommon chara

(3) (cognition of) contradictory characters (in
and the same thing), (4) positive cognition
certain objects) , and (5) non-cognition

:■,.

■M

PRAMAJjTACANDRIKA 17

certain objects). A doubt arising from the
perception of a common character is illustrated
in the case of the perception of a certain tall
stature which is common to a man and a post.
The perception of this common character calls
forth a simultaneous recollection of the two
forms of 'man' and 'post' in the perceiver's
mind as the result of which there arises in the
latter a desire to ascertain the true nature (of the
thing perceived, i. e., a desire to ascertain
whether the object perceived is 'a man* or 'a
post.') In the absence, however, of perception
of the crucial test, viz., the curved hollow which
is peculiar to the trunk of a tree or the possession
of a head, hands, etc., which is peculiar to a man,
there arises a doubt in the form of a mental
oscillation (between the two alternatives in the
form) ' Is that a man ? ' ' Or, Is that a post?.', ; A
doubt arising from the cognition of an uncom-
mon character is illustrated in the following case.
The cognition that sound is a quality that
belongs exclusively to Ether (Ikaia) awakens
a doubt, in the absence of a perception of the
solvent, as to whether it is an eternal or a non-
eternal quality. A doubt arising from the clash
of contradictory views is illustrated in the fofe;,
lowing case.. The man who discovers that the
.yaise§ikas teach that the sensibilities are consti-
■/•*gted by the.; elements t$: that the 8aiik%*
■ -teaches that they are aofc -and at the :■&&&&»»

.,*>

18 MADHVA LOGIC

cannot find the means of deciding between the
rival views is thrown into doubt as to whether the
sensibilities are constituted by the elements or
not. A doubt arising from a positive cognition
is illustrated in the case of a man who first
discovers the underground water in the act of
sinking a well. The discovery of the water
throws him into a speculative doubt (as to the
real source of the water), and he thus enquires,
for want of a solvent, as follows : 'Has the action
of digging only made manifest the water which
has been existent all along but remained non-
manifest? Or, has it made the non-existent
water start into existence? The following,
lastly, illustrates the case of a doubt arising from
non-apprehension. The man who learns from
hearsay that there lives a ghost in the banyan
tree yonder and yet finds none when he comes If?
near the tree is thrown into a mental uncertainty
as he does not cognise the solvent. He thus
enquires : ' Is the demon not perceived because :^
of its power of making itself invisible? Or, is it
unperceived because it does not exist ? ' * jl

Others hold that positive cognition and non- ■
apprehension (as causes of doubt) being only'J
modalities of the 'common character' (as a
cause of doubt), doubtful cognition must be
said to have three causes only (and not five
as stated above). How is- -'positive- cognition?:
to be regarded as a modality of the 'common!

■ # m g

■ft!

'■■-

PRAMAVACANDRIKA 19

character* that is supposed to be a cause of doubt?
In this way. There is positive cognition
of a jar that exists in a dark place when
a lamp is lighted and the surrounding darkness
is dispelled thereby. There is also positive
cognition of a jar that did not exist previously
till the potter has operated on the lump of clay
and brought a jar into being. (The positive
cognition is thus a character that is common to
an existent and a non-existent thing and thus
raises doubt as to the existence or non-existence of
the object in the mind of the cogniser). How
is non-apprehension a modality of the 'common
character?' In this way. There is non-appre-
hension of the existent such as the non-perception
of God as well as of the non-existent such as the
non -perception of the hair's horn. (Thus non-
perception as appertaining alike to the existent
and the non-existent raises doubt as to the
existence or the opposite of the object noa-
perccived.)

But the true view is that the so-called 'un-
common character' as well as 'the clash of
contradictory views * being really modalities
of the 'common character,' there is only one
cause of doubtful cognitions, viz., the cognition
of a common character.' The way in which these
(ti2., 'the uncommon character,' 'the clash of
doctrines,' etc.) are to be regarded as comprised
in the ' common character * should be understood

MADHVA LOGIC

* •

in the manner they are shown to be so comprised
in the work called the 'Paddhati.'

(Having explained the nature of doubt we
now proceed to explain the nature of erroneous

cognition.)

An error (viparyaya) is a cognition con-
sisting in the conscious certitude that a thing
exists just where as a matter of fact it does
not exist. If we define an error simply as a
cognition, our definition will apply also to
doubtful cognitions (which are also cognitions
and thus will be too wide* Hence we define*
error as a cognition which amounts to a certain^
knowledge or conviction. This excludes the|
cases of doubtful and uncertain cognitions. Bti|
since valid cognition is also as much self-eonfidenl-
as invalid cognition or error, we define error as i
cognition that apprehends a thing where
reality the thing does not exist. This di
tinguishes an error or invalid cognition f
valid cognition and thus our definition is fi
too wide. But even thus our definition fails
exclude the cases of doubtful cognitions,
doubtful cognition (though lacking in certitu
or decisiveness) also apprehends a thing where
does not exist. Hence we say, an error is;-|
certain cognition. Even thus however our de
tion remains faulty as applying to the case
valid cognitions as well, e.g., the cogni
'the tree is in contact with the

pramAsacandrikI 21

The contact does not exist in all parts
of the tree. (Therefore contact is asserted in
respect of an object which is devoid of contact
in some of its parts). Hence we insist on the
word just in our definition, i.e., we say an error
cognises a thing just where the thing does not
exist. Such errors arise from faulty perceptions,
fallacious reasonings and defects of verbal com-
munications. The illusion of silver in the locus
of a mother-of-pearl is an illustration of error
arising from faulty perception. An error of
reasoning is illustrated in the case of the man
who under the influence of blinding dust imagines
he perceives smoke and on the basis of the
illusory smoke infers the existence of fire in a ,
place where fire does not exist. Similarly,
when on the strength of the lying report of an
untrustworthy man one believes that there are ,
five different fruits lying on the banks of *■_..,:
neighbouring river, we have a case of an errO*^
arising from a faulty verbal communication. :-;.;.-,:
An objection however may be raised here), .^
We have defined Pramana as consisting in making f
the object of cognition to be cognised in the
form in which it exists (actually). Butthe defini-
tion may be objected to as being too wide applying
as it does to the case of 'memory' also (*Hjg$
•is not usually recognised as a Pramana)- Our
reply is : this is not so, for scripture testifies to
jgftf^flf that 'men^,f ; >rception,* : ;iaF^^r-;

22 mIdhva logic

*._«

and * inference ' are to be regarded as the 1
pramanas or valid sources of knowledge in regard
to such things as (dharma) merit and the like,
by all those who desire liberation. Thus we
have the testimony of scripture showing that
memory is a form of valid knowing.

Some philosophers define pramana as the
karana or instrumental cause of prama or valid
cognition. An instrument as such (according to
them) is not a pramana, otherwise any instrument
such as an axe would rank as pramana. Hence
the definition of pramana as an instrument of
valid cognition (and not as a mere instrument).
Similarly the epithet ' valid * is also necessary as >!
without it the definition would apply to non- |
valid and erroneous cognition and thus be
too wide. Lastly, the word instrument is also!
necessary as without it the definition would I
tantamount to an absurdity and also be tofr|
wide as applying to the consequence or result J
that follows from the instrumentality of valid
knowing.

We however do not accept the above view as^|
we consider the definition too narrow as not-|
applying to the result of valid knowing. (Our*'
view is that the word pramana signifies the
knowing act as well as the result of knowledge
that arises therefrom.)

Others define pramana as that which is;
pervaded by prama or valid knowledge.

PRAMAtfACANDRIKl &3

This also is not a tenable position. Every
knowable object is validly cognised by the Lord.
Hence every knowable such as the jar and the
like is pervaded by the Lord's valid knowledge.
Thus every knowable answers to the above defini-
tion of pramana or valid knowing as that which
is pervaded by valid knowledge. Thus the defini-
tion is too wide as applying to knowables (and
not merely to knowing acts).

With a view to escape from the above
difficulty, others modify the above definition as
follows : — Pramana is that which being either a
substrate or an instrument is at the same time
pervaded by valid knowledge.

Even thus however the definition is not fault-
less. The mention of the word 'substrate' in
the definition is without rhyme or reason Even
granting that the Lord is regarded as a standard
of validity, it cannot be said that this in itself
is a sufficient reason for introducing the word
1 substrate ' in the definition. For the word
pramana is derived by means of the suffix lyut,
and, according to the rules of grammar, the suffix
lyut applies only to the instrumental, the locative
and the nominative absolute. There is no rule
for its application to a nominative as such, i.e.
(as in the present case), to the agent, nominative
or subject of valid knowing. (The Lord is
regarded as the standard of valid knowing only
as the absolute knower, ie., as the agent or

-:*\

MADHVA LOGIC

subject of absolute knowledge and not as its^
substrate or instrument.)

It may be argued that even though the
as knower is the subject of knowledge, yet
also is the substrate or locus of such knowledge,
and thus may very well be the meaning of the
word pramana. But even this argument does not
bear examination. The Lord as knower is aO^g
agent of the knowing act and not its substrate Cji
locus in the strict sense. For what is a locus
adhikarana? A locus is that which is the adh
or container of the agent acting or the obj
acted on and is at the same time the a&raya-a
substrate of the action itself. (The Lord c
be the substrate of the knowing act of w
He is the subject.)

Others (the Prabhakaras) define PramanI f
anubhuti, i.e., as the apprehension of a.-fi
By anubhuti they mean cognition other than n
collection or memory. According to . t
anubhuti cannot be defined simply as ' other
recollection/ for in this case the definition
apply to objects of cognition like the jar and
rest (which are other than recollection). \
ean it be defined simply as e cognition/ for
this case the definition will apply to ' recollection^
(which is not an independent source of knowledge

according to Prabhakaras). _ : ,.,V.:

v But the Prabhakara definition of Pramapft H
open to the following objections,; .yte^bg$

■im

.i-'-S

. ".'»

^£i&H

' ' \

PRAMi^ACANDRIKA 25

place, it is too wide as applying to doubtful
cognitions (which are also forms of apprehension
or anubhuti other than recollection). Secondly
it is also too narrow as not applying to memory
(smrti) and the social codes derived from the
Vedas. (Smrti means recollection as well as the
social codes based on the Vedas and their teach-
ings. The latter are analogous to recollection
which is based on an original primary presenta-
tion. The Madhvas accept smrti as pramana in

both these senses.). -

(This closes our discussion of the definition of
Pramaqa. We now proceed to discuss its different

varieties.)! - - - '-■ ■"«■'
Pramana is of two kinds, viz., (1) Kevda?
pramana, i.e., self-contained, absolute knowing,
and (2) Anupramdna, i.e.,. valid knowing
as the instrumental cause of self-contained,
absolute knowing. This enumeration is baaed!
on the order of importance. (Kevalapramafa
being of superior importance is first mea*>:
tioned.) ". ,..;.^,;

- **

Kevalapramam means knowledge that agrees
with the nature of the object known. The words
'■ agrees with the nature of the object ' dispose
of uncertain and doubtful cognitions, while the
word 'knowledge' disposes of percerang (and
other intervening processes). (Keratapraro^ii*^
the resulting knowledge as distinguished from the
processes leading thereto.) ■ <■'-..■•■&£$§$$&"■

"4 ' ■ '

26 MADHVA LOGIC

There are four kinds of Kevalapramana (i.e.>
four kinds of self-sufficient, self-contained
knowledge) : — (1)* the knowledge of the Lord or
Ihara, (2) the knowledge of the Lord's Consort,
Atfcj Laksml, (3) the knowledge of the Sage or
Yogin A (4) the knowledge of the Non-Sage or
Ayogin.

The Lord's knowledge is the knowledge that
rests on, i.e., presupposes, itself only. The fact
of its l depending on itself only ' distinguishes the
Lord's knowledge from that of the Lord's Consort.
(The Consort's knowledge, while depending on
itself, also depends on the Lord's knowledge.)
The fact of its being i knowledge ' distinguishes
it from the Lord Himself whose knowledge it is.
(This is aimed at the Shankarite theory according
to which Brahman is nothing but pure self-
revealing Intelligence. The Madhvas distinguish
between the Lord Himself and the knowledge
which the Lord has of Himself and all other
things.) Or, we may say, the Lord's knowledge is
knowledge that embraces all that appertains either
to the Lord Himself or to that which is other
than the Lord. The Lord's knowledge cannot be
defined simply as knowledge, for in this case the :||
definition will be too wide as applying to the J
Yogin's knowledge as well (which is alsp : .J§
knowledge of a sort). To exclude the latter, the;?
definition stresses the words ' that embraces attgj
etc.' (The Yogin's knowledge is not all-embracing

•*»

prama#ac"andri£a 2?

as is the Lord's). Further, the word ' knowledge ' :
prevents a too wide application of the definition
to the perception of the Lord. The Lord's
knowledge is invariably in agreement with the
nature of the object known, constitutes the
essence or svarupa of the Lord, is without
beginning and without end.

The Consort's knowledge is knowledge that
depends on, i.e., presupposes, the Lord's knowledge
only. If the Consort's knowledge had been
defined simply as ' knowledge,* the definition would
have been too wide and applicable to the Lord's
knowledge as well. To exclude the latter, the
words * depending on the Lord's knowledge* have

been added. Since dependence implies a distinc-
tion (between the dependent and that on which
it depends), the possibility of confusion (between
the Consort's knowledge and the Lord's knowledge)
is precluded. But mere dependence on the Lord's
knowledge does not fully define the Consort's
knowledge. The knowledge of Brahma and
others is also characterised by this dependence
on the Lord's knowledge. To preclude such
extension of the definition (to the knowledge of
Brahma and others), the word 'only* has been
added. (The Consort's knowledge depends on the
Lord's knowledge only, but the knowledge of
Brahma and others depends both on the Lord's
knowledge and the Consort's knowledge.) Further,;
the word c knowledge r in the definition serves to

2& MADHVA LOGIC

distinguish the Consort's knowledge from the J
Consort herself. Or, we may say, the Consort's
knowledge is the non-reflective knowledge of all
objects other than the Lord Himself. Here the '§
word * knowledge' by itself fails to distinguish
the Consort's knowledge from knowledge like -
that of ourselves. Hence to prevent such too
wide application, the words ( of all objects ' have . ^
been added. Even then the definition has a too, -1
wide application to the Rju login's knowledge.
(The $ju Yog in also has a reflective knowledge,
of all objects). Hence the word 'non-reflective.*^
The Rju-Yogin has only a reflective knowledge of
objects. Again to say that 'the Consort's
knowledge is the non-reflective knowledge of all
objects ' does not distinguish it from the Lord's :i
knowledge (and thus the definition as so worded
remains too wide). Heoce the further qualifier
tion 'excepting the Lord Himself.' But the
Words * non-reflective knowledge of all object£|
and no 3uch knowledge of the Lord Himself
would be absurd and self -confuting. Henflg
the words 'other than 5 {%. e., non-reflecti
knowledge of all things other than the
Himself). Further, the word r knowledge * iij|
the definition distinguishes it from the Consort*
perception (t. e.» the process of percei
which leads to A or results in, knowl
The Consort's knowledge is also inv
in agreement with reality^ is the esse

■ ■ ..

** J .

PRAMAtfACANDRIKl 29

of the Consort herself and is beginningless and

eternal.

The Yogin' s knowledge is the knowledge that
has attained to special perfection or excellence
through the power born of the practice of yoga
or mental concentration. It is of three kinds,
viz., (1) the Rju Yogin's knowledge, (2) the
Tattvika Yogin's knowledge, (3) the Atattvika

Yogin's knowledge.

By Rju Yogin is meant a jiva or individual
soul who is capable of the spiritual excellence of
Brahman. The Rju Yogin's knowledge is the
reflective knowledge of all objects other than the
Lord Himself. As the mere words 'the Rju
Yogin s knowledge is knowledge ' will not prevent,
a too wide application of the definition to our;
knowledge as well, the words * of all objects * h$v
"been added. As even theo them. is a

^application to the Lord's knowledge,
* reflective * has been incorporated. As th
the definition entails an absurdity, the
being a reflective knowledge of all
and yet not a knowledge of the Lord
the words 'other than the Lord Himself ' BS*R=^
been added. This knowledge is of two kinds &^*$j&
(o) knowledge which is the essence or nature of '^M&
the , Yogin himself, and (&) knowledge which is
only a mental state of the Yogin. Of these*
knowledge constituting the Yogin's essence jft!

^begraningless and etaajfeJifcUe knowledge ]|I»

fe:*

>:**■*'■■ y *" m ■'**■

*..

30 MADHVA LOGIC

Yogin's mental state is beginningless only as
being comprised in a (beginningless) flow
stream (of states). Both these however are inH
variably in agreement with the nature of thej

objects (known).

The Tattvika Yogins are the supernal beings
(with godly qualities) other than the Rju Yogins
and full of the conceit of true knowledge of
reality. Merely saying that Y the Tdttvika$j>
are the gods with self-conscious knowledge of
reality ' would have involved a too wide applica-
tion to the Rju Yogins (who also have sdfH
conscious knowledge of reality). Hence the
words 'other than the Rju Yogins. 9 Merely
saying again that * the Tattvikas are other tharf
the Rju Yogins would have involved a-
wide application to the gods or spirits who are
non-Tattvikas. Hence the words 'with self
conscious knowledge of reality.' Since there
also ungodly beings (other than Rju Yogins) wh#j
also have self-conscious knowledge of reality^
therefore the words ' with godly qualities * have
been added. The Tattvika Yogin's knowledge $
that which being beginningless does not, even
by way of reflection, embrace all things other
than the Lord. Merely saying that ' the Tdttvi
Yogin's knowledge is knowledge ' would ha
entailed a too wide application to the Lord'
knowledge. Hence the words c does not embra
all things.' Even then there would have been

yfi.i

PRAMA^ACANDRIKA 31

too wide application to the Consort's knowledge.
The Consort's knowledge is non-all-embracing in
so far as it does not extend to the Lord Himself.
Hence the words ' other than the Lord.' Even
then, there would have been a too wide application
to the Rju Yogin's knowledge, for the Rju Yogin's
knowledge does not embrace all things other than
the Lord in the absence of reflection. Hence
the words 'even by way of reflection/ Even
then, however,, the definition would have a too
wide application to the non-Tattvika Yogin's
knowledge. Hence the words 'being beginning-
less/ It also is of two kinds, viz., (a) knowledge
which constitutes the svarupa or essence of the
Tattvika Yogin, and (b) knowledge which is
external (i.e., relates to external objects). - Of
these, 'essential 5 knowledge is in agreement with
reality, but * external ' knowledge is occasionally
false, i.e., not in agreement with the nature of
things. ...

The non-Tattvikas or 'Atdttvikas are the Gods
and Sages practising Yoga who are other than the
Rju and the Tattvika Yogin$. The non-Tattvika
Yogin's knowledge is that which, having a
beginning in time, is characterised by slight or
partial ignorance in respect of objects other than
the Lord. Merely saying that the non-Tattvika
Yogin's knowledge is knowledge would have
entailed a too wide application to the Lord's
irT^TxriafW rrWAfnrp .thii words * characterised

32 MADHVA LOGIC

by ignorance' have been incorporated. Even!
then there would have been a too wide applicant
to the non- Yogin's knowledge. Hence the word
' slight or partial/ Even then the definition would|
have entailed an absurdity, for the non-TattmW4 t t
knowledge in respect of the Lord is tainted by
immense (and not slight) ignorance. Hence the;-*
Words ' other than the Lord/ Even then there
would have been a too wide application to the
Tattvika Yogin's knowledge. Hence the words
'having a beginning in time/ This also is
two kinds, viz., essential knowledge, and exl
knowledge. The rules as to their truth or untrutl
are the same as in the previous case. The begin-:
ninglessness and the beginning-in4ime of
Tattvika and the wm-Tattvika Yogin's knowli
respectively should be understood in the mann<
explained in the f Paddhati/

The non-Yogins or Ayogins are the individi
souls other than the Yogins or sages. The wo]
* Jivas or individual souls ? differentiates
Ayogins from the Lord and the Lord's C<
and the words 'other than the Yogins*
tinguish the Ayogins from the Yogins.
■. The Ayogin's knowledge is knowledge chi
tensed by immense ignorance in regard to obj<
other than the Lord. Merely saying that
Ayogin's knowledge is knowledge chai
by immense ignorance' would have entailed
too wide application to the Yarn's

T"«.

*tfCi

-#**

PRAMA^ACANDRIKA 33

also. The Yogin's knowledge is also accompanied
by immense ignorance, viz., in regard to the
Lord. Hence the words 'other than the Lord/
The Ayogin's knowledge is also of two kinds as
in the previous cases. Besides, it is also
generated in time and perishable in time. This
generation or cessation means generation or
cessation with respect to the manifestation of the
svarupa or essence. The svarUpa or essence here
means the svarupa or essence of the knowing as
a cognitive process or state.

The Ayogins or non-Yogins are also of three
kinds : (1) those that are fit for liberation, (2)
those that are perpetual participators in (the
storm and stress of) life, (3) those that are fit
only to live the stupefied life of inertia or Tamas.
Of these, ' essential ' knowledge in the case of
' those fit to be liberated ' is in agreement with
the nature of reality, while the same in * the
perpetual participators in life* is of a mixed
character (partly true and partly false). Of
others, however, such knowledge is false as being
in non-agreement with reality- As regards
' external ' knowledge, it is both (true and false,
».e., in some cases true and in other cases false)
in all the three kinds of the non-Yogins..

The Vaisesikas accept sense-perception, ^
inference from a mark, memory and the intuitions
of the Sages as the four kinds of valid evidene&o:
Thio ah* Voieofiiw via™} hnwe.ver is untenable

'-■*ii

34 MADHVA LOGIC

as this enumeration does not include the Lord's
knowledge which is timeless and eternal as also
knowledge derived from authoritative communica-
tion (Agama). Further, recollection being the
effect of the action of the mind which is a
sense-organ, and the intuitions of sages being
only a species of Yogik knowledge, and Yogik
knowledge being itself a variety of sense-knowledge
aided by the power of Yoga (according to their
own admission), there is no reason for a separate
enumeration of memory or recollection and of.
the intuitions of sages as independent sources of;

knowledge.

[Having explained the nature of direct self^

contained knowledge (kevalapramaw), we sh
.now discuss the nature of pramfma as Ann*
pramdiia or mediating processes.]

Anupramana is the means or instrumental can
of valid knowledge. Merely saying * Anuprama
is knowledge ' would entail a too wide applicati
to knowledge itself (which is the result or effe
of Anupramana) and to uncertain and doubtf
cognition (which also is a species of knowledge
though not valid knowledge). Again ■, merel;
saying 'Anupramana is an instrument or effec-
tuating means ' will entail a too wide applicati
to the axe and other like instruments. Ag
merely saying 'Anupramana is valid knowled
will entail a too wide application to Kevalt
pramana (which is direct, self-contained v

■%i

PHAMA^ACANbRlKA \> V^5-^" - W

knowledge). Simply saying again that l Anu-
pramafta is the instrument of that which is valid '
will entail a too wide application to the instru-
ment of perception (which also is valid). Like-
wise saying merely that 'Anupramaiia is the
means to knowledge or cognition ' will entail a
too wide application to that which is a means to
doubtful cognition, etc. Lastly saying that 'Anu-
pramana is the cause of valid cognition ' will
entail a too wide application to the knower.
Thus our definition of Anupramiina is proved to
be necessary in all the points.

(What, then is a sadhana, instrumental cause
or means?) A Sadhana or instrument is that
which being absent, the effect does not arise, even
though other conditions like the knower, etc., are
present, and which being present unobstructed,
the effect necessarily arises. E.g., the axe in the
process of striking (the tree to be felled) . Hence
there is no too wide application to any and
every concomitant condition, e.g., no too wide
application to cases of erroneously cognised
marks or signs. Where error is involved, the
presence of defects of sensibilities, etc., is the bar
(to the cognition of the really effective means).

(The sadhana is thus the cause par excellence,
i.e., the most effective of the causal conditions.
And this brings us to the question of the. nature
of the cause.) ..."

The cause may be defined as the unconditional,

36 MADHVA LOGIC

invariable antecedent of the effect. And t
effect may be defined as that which is nega
by the negation * consisting in antecedent!
non-existence. (This means that the effect
is that which does not exist before it i
caused to be.) The cause again is of two kinds/
viz.. the material cause and the instrumental
cause. The cause which suffers transformation
(into the effect) is the material cause, for
example, Prakrti or primal matter as the cause
the universe, the lump of clay as the cause of t
jar. The cause which produces the effect witho
being itself transformed thereinto is the inst
mental cause, for example, the unseen m
forces (Adrsta) as the cause of the universe, t.
potter's stick as the cause of the jar. Hence
follows that among the assemblage of caus;
conditions that which is the cause par excellen&jjp
(i.e., most effective) is the sadhana, means -A
instrument. '^

. Anupramaqa is of three kinds, viz., percepti
inference and authoritative communication. ('
these) perception is the means to the appreh
of that which is comparatively proximate, is
mediated and present here and now. H
perception is limited in range, being restricted
a small number of objects (i.e., objects whic|
are near and present and not separated by w
barrier). Inference however is a means i&
knowledge of the remote, the mediated

■ L ■,

&"■■

PRAMAtfACANDRIKA 37

the past and the future. Lastly, Agama or
authoritative communication is an independent
source of the knowledge of all sorts of objects
of which the number is endless. The above
enumerative statement is based on the order of
the kind and number of objects made known
(respectively by the three kinds of knowing, viz. 9
perception, inference and authority). There being
three kinds of Anupramdna or effective means to
valid cognition, the cognition resulting therefrom
is also of three kinds, viz., perceptual, inferential
and authoritative knowledge. Of these, percep- ,
tual knowledge is the cognition that is produced
by the process of perceiving. This (cognition) is
what is meant by immediate knowledge, direct .
knowledge or witnessing knowledge (saksatkara) .
Similarly, the knowledge reached through in-
ference is called inferential or mediate knowledge.
Lastly, the knowledge derived from authoritati^.^
communication is called authoritative knowledge

or revealed knowledge.

(What, then, is perception as a source rf-^
knowledge?) Perception is the process of *||
sense-organ that is free from defects. (In other^S
words, perception is the stimulation, by the^
object, of an organ of sense not subject to anjf ;J>
sensory defects.) Here the word 'sense-organ'
prevents a too wide application to inference -iriti^

the rest. -'■ - ■ - '■ V •-*"'? ;W'w%

Rnt it mav be said, perception is a species o

:i!

38 MADHVA LOGIC

an instrumental cause, and an instrumental cause
is possible only through an intervening action or
operation. For example, the instrument of the
act of cutting is the axe, and the axe is an instru- ,^
mental cause of the cutting or felling of the tree
through the intervening operation of coming into
contact with the tree. The intervening operation
is that which being effected by the instrumental
cause serves to effect that which results from
the action of the instrumental cause, e.g., the
contact with the tree which is effected by the axe.
(What, then, is the intervening action in the
case of perception as an instrumental cause or
agency.')

We grant the force of the question which is
thus raised. And our answer is this. The sense-
organ is the instrument, and the contact of the
sense-organ with the object (perceived) is the
intervening operation (through which the sense-; ^|
organ produces its effect as an instrumental cause)
The consequence or effect which results from 'i
the action of the sense-organ as an instrument
is direct or immediate knowledge (saksatkara) M
Here the process or operation (i.e., the perceptive^
operation) is described as the operation of the|
eye (i.e., of the sense-organ concerned) with
a view to stress the fact that the organ (a*J
the seat of the operation) is primary while the J
operation inhering therein is secondary. Anjll
further the organs are said to be free froflb

PRAMAJJACANDRIKA 39

defects and to be in contact with their respective
objects with a view to ensure the truth or
validity of the knowledge resulting therefrom.
Hence there is no discrepancy in our definition.

Perception is of seven kinds, viz., the six
different kinds of perception consisting of the
operations of the senses and the perception of the
sdksi or witnessing consciousness. Of these, the
perception of the witnessing consciousness is
perception consisting in the operation of the
Witnessing Intelligence as organ or instrument.
The objects of such immediate knowledge or
presentation are : the essence of the Self or
Atman as well as the properties thereof, the
nescience underlying absence or privation, the
mind, the functions of the mind such as know-
ledge, pleasure and the rest, time, Aka$a in its
original unmodified condition, etc. It also
reveals its own nature.

(Sensuous perception as arising from the
operation of the six different organs presupposes
the six organs of sense.) The six sense-organs
are the olfactory sense, the gustatory sense, the
visual sense, the tactual sense, the auditory sense
and the mind. Of these, the objects of the
olfactory sense are odour and the different kinds of
fragrance which are the specific modes of odour;
the objects of the gustatory sense are tastes and
its specific forms ; the objects of the visual sense
and touch are substances having magnitude and

40 "*' MADHVA LOGIC

perceptible (generated) colour and form, certaift^
qualities such as number and the rest, the
different kinds of motion, and the generic;J
characters. The feel of air is also an object of
touch. Air itself is (not directly perceived by §
touch, but only) inferred from its feel (which is %
revealed to touch). (The inference is as follows.) ;:
The wind that blows is felt as being neither hot _f_
nor cold. Since the feel is a quality like colour
and cannot go wandering, there must be some sub-
stance in which the feeling in question inheres or
abides. What, then, is the substance which is the|
substrate of the feeling in question ? It cannot
be earth, for earth, though characterised by per-3

ceptible touch, is also characterised by perceptibte|
form and colour, whereas the feeling which i
experienced (in the case of the blowing wind)
evidently has no substrate of a perceptible fo
and colour. Nor can the substrate in question b
either water or fire, for the feel which it indu
is not experienced as cold (as in the case
water) or hot (as in the case of fire). Nor again
can it be the four ubiquitous substances si
they are all devoid of the quality of such feeling.
The assumption of tactual properties in the
quitous substances would entail a perception
touch everywhere and at all times. Nor
can the mind be the substrate in question,
the mind is atomic or infinitesimal and the
of the infinitesimal or atomic is impe

PBAMA^fACANDRlKl Vj* \ 41

V ks ^

i.e., beyond sense-perception. Hence we conw*- - • v

that which constitutes the substrate of the quail
that is felt by touch in the case in question is
the (specific) substance we call air.

But it may be said air is perceptible (and not
inferred as above shown), because it is the
substrate of a perceptible touch just as is the
jar. This objection is not tenable, for air is the
- substrate of perceptible touch only because of an
adventitious condition, viz., its association with
an object of perceptible form and colour. (An
adventitious condition is anything that is in-
variably correlated with the probandum but is }
not an invariable correlate of the probans.) In ■■■ Q
the case in question, the condition is an in- .- ■&
variable correlate of the probandum as determined :%
by the property of the subject in which it , : (^
is inferred. (The subject of the inference is |§
: air which is an external substance and the pro^/j||
bandum is perceptibility) . That the adventitious^
condition is perceptible colour and form is proved ,;/^|
by the invariable correlation of external sob^gg
stances that are perceptible and the presenpgj
of perceptible colour and form together .*&*$$
the absence of any such correlation between tbe;|;g|

■ ground of the inference, viz, ' being the BpbBfe^g|
of perceptible touch ' and the adventitious condprf^

, tion, viz., the 'presence of perceptible colour^;;, J

form.' Thus we find that whatever is an external
■y Substance and is gerceived is also char^erised v;-

■ .-"■ 6

42 MiDHVA LOGIC

by perceptible colour and form, but we cannot
say that whatever is a substrate of a perceptible
touch has also perceptible colour and form,
(Hence being the substrate of a perceptible touch
is a valid ground for perceptibility only on the
condition that the substrate has perceptible colour
and form and not otherwise. In other words, the
ground is a ground only under conditions and
not unconditionally and therfore the inference in
question is not flawless.) Hence though you
affirm the presence of the ground (viz., being
the substrate of a perceptible touch), in the subject
of the inference (viz., air), yet perceptible form and
colour which always accompany the probandum^fj
i.e., perceptibility, is no invariable correlate of J
the ground (and so the ground fails to prove )
the presence of the probandim being shown-l
to be no invariable correlate of an invariable .

«**

correlate of the probandum.) Hence it follows^
that air is inferred from its special touch (and
not perceived as contended).

The objects perceived by the auditory sense |
are sounds and their varieties in the form of
the alphabetical or verbal sounds.

Inadvertence of the mind (partial or complete))!
colour-blindness, jaundice, etc., are the defected
of the five sensibilities. ,. /j

The objects (of the five different senses) arfrl
also objects of the mind (the sixth sense) and?
the mind makes them its objects by presiding

■ ■ j
"A

PRAMAtfACANDBIKi 43

over and guiding and directing the operations of
the five external senses.

But the mind also perceives past events and
in such perception it does not require the aid
of the external senses. The result of such per-
ception (of the past independently of the external
senses) is memory or recollection. Memory is
thus said to be an effect of internal perception
(by the mind as the sixth sense). The contact
of sense and object in this case (i.e., in the case
of a direct perception of the past by means of
the mind as the sixth sense) is furnished by
what we call the Samshara, trace or disposition
left behind by the past experience. (The mind
as the sixth sense has a direct vision of the past
and what we call the disposition or trace of the
past experience is nothing but the contact of the
past with the present, or as Bergson would say
perhaps, the insertion or prolongation of the
past into the present). The process (of the
mind in recollection) is thus analogous to that
of the specially gifted senses of the yogis which
possess extraordinary supernal powers due to
the practice of yogik concentration (and thereby
cross the gulf between the past and the present).

The defects of the mind (which vitiate the
mental or internal perceptions) are desires,

predispositions, etc. ,-

There are also defects of the objects perceived
(which interfere with their correct perception) and

44 - MABHVA LOGIC

■#

■■£

these are long intervals (of time or space), closed
proximity (in time or space), subtile or potentiaJJ
condition of the objects, intervening barriers :
(media) separating the objects from the perceiver,4
non-manifestation of the objects (e.g., of the stars |
in daylight), intermixture with objects of a like|g
or similar nature, etc. : /ft§

These defects being present, in some cases the^|
cognition itself is not generated and in some cases n
doubt or uncertainty arises (as to the evidence ofc|
; the cognition that is generated). -

In the case of cognition due to the operation
the senses, the cause consists of the contacts of the
four beginning with the self. Thus the self mu
come into contact with the mind, the mind wi
the sensibilities, the sensibilities with the objects;
The sensibilities have no immediate intuition of
v objects at a distance and must actually dart fo
to the place of the objects and come into contact
with them in order to reveal them. The se
same contacts which enable the sensibilities
reveal their respective objects are also the imine*
diate cause of the perception of their respecti^
absence. No intervening relations mediate
tween the contacts and the absence in the case
perception of absence (as Naiyayikas say).

One school of philosophers (the Naiyayika
": holds that the contact which brings on cog 1 ^'^
|L tion in the form of immediate perception
g of six different kinds. Thus (according^

■*■;»'-<*

"5>"»

PRAMASACANDRIKA 45 :

this scbool) it is either conjunction, or inner*
ence in the conjoined, or inherence in that
which is inherent in the conjoined, or inher-
ence, or inherence in the inherent, or the
relation of qualifying and qualified. Of these,
conjunction is the contact or relation that
is effected between the eye and substances like
jars, etc. (i.e., the relation which brings on the
perception of substances like jars, etc.). The
relation of inherence in the conjoined similarly
exists in the case of colour (which is a quality),
actions and generic characters. (These inhere in
substances like jars, dishes, etc., and these latter
are in conjunction with the eye. Therefore the
visual perception of colour, action and generic
character takes place through the relation of *i**'\
herence in that which is in conjunction with the
eye.) In the same way in tactual perception, the
relation which brings on perception by the tactual <U
sense is actual conjunction with the tactual sense
in the case of perception of substances such- W
as jars, etc., and the relation of inherence in that
which is in conjunction in the case of the quality
of touch, the actions and generic characters oT- f;A
these substances. So also in the case of the
internal perception of the self' by the mind, the
relation is conjunction of mind and the &&?%.
substance in the case of the perception of the latter
while it is inherence in that which is in conjun|K
tion in the case of the perception of the pleasure

46 MADHVA LOGIC

and pain that exist in the self. So also in the
case of perception of smell and taste by the
olfactory and gustatory senses, respectively, the
relation is inherence in that which is in conjunc-
tion, the smell and the taste being inherent in|
substances which are in conjunction respectively
with the olfactory and gustatory senses. Similar- •
ly, in the perception of the generic characters of-
qualities and actions, the relation which mediates^
is inherence in that which is inherent in sub-
stances in conjunction (the generic characters;
being inherent in the qualities and actions which^
are themselves inherent in the substances of wbic$?
they are qualities and actions). In the perception^
of sound By the auditory sense, however, the]
relation is simple inherence, for the auditory
sense is nothing but ether or AkaSa as limited^
by the tympanum of the ear (sound being a quality
of Akaia and so inherent in Akaia). (Th^
auditory sense being Kka§a itself as limited bjr
the tympanum and sound being inherent in
Akaia the relation which holds between sound
and the auditory sense in the case of perception <
sound is a relation of pure inherence.) But
perception of the generic characters, etc., of the
sound (inherent in particular sounds) is medial
by the relation of inherence in the inhen
(sounds being inherent in Akaia and tberefoi
in the limited AkUa which is the auditory sense,
and the generic characters, etc., of sounds bein,

prama^acandrika 47

inherent in sounds) . The perception of the absence
of the jar by the eye takes place through the relation
of qualification and qualified. In the case of the
perception in the form i In this place here, there
is no jar,' the absence of the jar is the qualified
and the qualification which specifies the absence
is the particular place or locality with reference
to which the absence is perceived. (The localisation
thus acts as the specifying attribute of the absence
in question.) In the case again of the perception
of the absence being in the form "This place
is characterised by the absence of the jar/ 9 the
place itself acts as the qualified substrate and
the absence of the jar is regarded as its qualifica-
tion. Similarly in the perception of Inherence
itself the mediating relation is that of qualified
and qualification (Inherence does not inhere. Nor
is it in conjunction with the objects between which
it holds. Hence the relation of inherence to the
objects between which it holds is said to be a
unique relation which is that of qualification**
and qualified. The inherence is a qualific3*£ces
of the objects between which the inherence haesent
Thus in the case of the inherence of the yhjoga)
cloth in its parts, viz., the threads, the inhevmen-
is related to the cloth and the threads by the jition)
tion of qualification to qualified (and not iom a
second relation of inherence nor by conjuncti^ Non-
All this however is fallacious and unteicausal
Qualities, actions* etc., being nothing apart

48 MADHVA LOGIC

■KK

(i.e., being non-different from) the things qualifi
or the things acting, etc., a relation of inheren
between qualities, etc., and their substrates J
impossible (for inherence presupposes a different
between the inherent and that in which t"
inherent inheres). Besides, there is no valifg
ground for the acceptance of inherence as reafcjj:
Though the relation of conjunction between /ftp
self and the mind is required in the case of t" "
perception of other objects, it is not so requi
in the perception of the self itself or its propert
and states, for the self and its properties beii
the objects presented to the witnessing Intellige
are not objects of perception by the mind. Ag
alphabetical sounds being themselves substan
are, not qualities at all. It cannot be said t
they cannot be substances as besides ha
generic characters they are perceptible to
one of our external senses. For this rule
according to our view, in the case of dar
~~>, (Darkness has generic character, is revealed
„ >>pne of our external senses, and yet is a subs
^ , not a quality.) And it fails also m
t of the light of the lamp according to

j ; of those who hold it to be substantive. <$
sound ' t . *., -r^gS

h th i of the lamp has generic character,

/ j 3nted to one external sense and yet ugp
{sounds . - r">^

: *ii„ i,ie Naiyayika to consist of nothing but

in tne i\ f J _ . _ ■<&-■■ ;*^

A a 3 which are substances.) Further tbo

and tne m , _ ._•-' -y -■■m$

Iphabetical sounds are qualities of

PRAMA^ACANDRIKi 4:9

Ikafa, yet since as qualities they axe non-
different from their substrate of AMia, a
relation of inherence between non-alphabetical
sounds and Ahtea is impossible. As regards
the relation of qualification and qualified, it
being only another name for the relation that
consists in nothing but the essence of a rela-
tion itself (svarupasambandha), no separate
relation of qualification and qualified really
exists. (Svarupsambandha is the name of
the relation that consists in the essence or
svarupa of a relation. Thus inherence is related
by svarupasambandha to the objects between
which it holds. This means that the relation
which relates inherence to its relata is no separate
relation but the svarupa or essence of the in-
herence itself.) ■ f ' ' '

• But some however say that memory is the
consciousness that is caused only by traces of past
experience. The adverb 'only' precludes recogni-
tion (which arises not simply from traces but also
require other factors) . Kecognition is the cogni-
tion that arises from the joint operation of traces
of past experience and sense-contact with present
objects. The word coincidence (samprayoga)
means contact (of sense aod object) . Non-men-
tion of traces (as a causal condition of recognition)
will make recognition indistinguishable from a
simple cognition like 'Here is a jar.' Non-
mention of coincidence or contact (as a caasal

50 MADHVA LOGIC [thU^k

condition) will again make recognition indis*
tinguishable from memory. . \>

; The above view is not tenable. For all
valid knowledge is due to a valid ground or
source of knowledge and therefore since memory
is one kind of valid knowledge, the cause of
memory, viz. z the traces of past experience, will
have to be admitted as an independent pramana
or ground of true knowledge. But in this case
the number of pramanas or valid sources of

knowledge will be four and this is unestablished
(as we have seen that the number is three and
neither more nor less). It cannot be said that
the objection applies also to our view (of recollec-
tion or memory). We hold memory to be due
to internal perception by the mind with con
tration of attention (bhavana) as an auxiliary
condition. (Hence in our view memory is M
form of perception, a kind of intellectual intuition
where the trace of the past experiences serves
function of contact of the present mind with t
past experience. Thus according to our view*
though memory is admitted to be a form of
knowledge, it is regarded only as a variety
perception, and so the cause of memory is not
separate source of knowledge, though no doubt |
is a ground of true knowledge.) „ ..;

There are four kinds of perception (as a
of valid knowledge), viz., the Lord's percepti©
the Consort's perception, the perception ,th«» ;^

. ■-

. •?*

v *.jk

PRAMA^ACANDRIKA 5l

belongs to the Yogin or Seer, and the perception
that belongs to an ordinary mortal (ayogin).
The objects of these different kinds of perception
are the same as the objects of the corresponding
cognitions or experiences which each kind induces
in its respective perceiver. For fuller details the
reader is referred to Jayatirtha's Pramana-
paddhati (which we think unnecessary to go into
over again here). Thus everything (relating to
perception) has been set forth and therefore we
close our chapter on perception as expounded in
this Pramanaeandrika on the lines chalked out
by the revered Jayatirtha. We bend our heads
in respect to the sage Vyasadeva. ^

Let us now proceed to the next Pramaija, ©&.>>
Inference. Inference is flawless reasoning, flawless
establishment or proof of a conclusion (by means
of a reason or ground). The synonyms of proof
are reasoning, arguing from a mark to the thing
marked, concluding on the basis of something
which is pervaded by an invariable relation to

something else.

Inference cannot be defined simply as reason-
ing or arguing from a mark, for the definition
would then apply to fallacious reasonings such as
those where the subject of the inference is
fictitious or unreal and where the mark is known
by a valid source of knowledge to exdwlr
'(instead of being invariably related to) the thing
marked. In these cases the reasonings are

52 MADHVA LOGIC

grounded on an invariable correlation and yet the
reasonings are fallacious. (In the former, though
the invariable relation holds, the subject of the
inference is non-existent and thus the locus in
which the relation is to prove the existence of the
probandum does not exist. In the latter, the
invariable relation is asserted in the reasoning
but in actual fact no such relation holds, and
thus the conclusion lacks material truth.) Hence
inference is defined not as reasoning merely, but
as flawless reasoning. Nor again can inference be.
defined simply as that which is flawless for in,
this case the definition will apply equally to'
Perception as a source of knowledge. Hence it is ,
defined as reasoning (which is flawless).

The instrumental cause of inference is the
sign or mark (by means of which we infer the
'•■;- probandum). The operation or process (of the
instrumental cause) which leads to or establishes
th$ collusion is reasoning or argumentation (i.e.,;
reasoning by means of the sign through which the ■■
"subject of the inference is brought in relation to* j
the probandum). The inferred conclusion is the J
result which emerges out of the process. Keason4 : : l
ing (Paramaria) consists in the cognition of the
mark in the form of its invariable relation to ther;?
probandum as a property of the subject of thet^
inference. For example in the inference of. fire
in yonder mountain from the perception of smoke :

therein, the reasoning consists in cognition ot

-.-■:■

-'■■■■■

& -;

PRAMAljrACANDRIKA

the smoke as an invariable concomitant of fire and
as being, as so invariably related, a property of the
mountain yonder. The cognition which results
from the process, viz. f yonder mountain is on fire,
is the resulting inference. Invariable relation
means invariable concomitance (or sequence) as in
the case * wherever there is smoke, there is fire/
By concomitance or co-existence is here meant a
relation merely between the ground of the infer-
ence and its probandunu By the invariableness
of the concomitance is not meant then co-inher-
ence in the same substrate. The invariableness
of the concomitance means simply that the con-
comitance is fixed and unfailing. The purport of
the whole is that vyapti or invariable
comitance consists in an unfailing
between the ground of the inference
probandum. It follows therefore t
essential character (the defining mark)
consists in the unfailing regularity of
tance (between two or more phenomena),
when we observe that wherever smoke is per-
ceived, there fire also is perceived, we are said to
cognise the relation of vyapti or invariable con-
comitance between 'smoke' and 'fire.' Here
* smoke f is the pervaded and fire the pervading
property. The locus or abode of the invariable
relation is called the pervaded wlnle that which
defines or marks off the relation is called the
pervading. [Thus if A is invariably related

54 MADHVA LOGIC

to B, A as the locus of the invariable relation
is the pervaded, while B as defining or mark-
ing off this (particular) invariable relation
from other invariable relations is the pervading'
or pervader.] By the pervaded being a property i|
of the subject of the inference is meant its;
existence in a suitable place (so as to make
its invariable relation with the pervader pos-- A
sible. It does not mean that the pervaded |
property should be spatially or temporally includ-
ed within the pervader). And thus our theory i
is free from flaws even of a trivial character. '

Some however give the following account olM
invariable concomitance. Invariable concomi^j||
tance (as an element or factor of inference) means; ||
the coinherence of the ground and the probandum
in one and the same locus so that the probandumM
can never be that which is negated by theg|
absolute negation that coinheres in the locus of thep
ground of the inference and also does not coinhenfe
in the locus of that which it negates. (In othe^
words, if the probandum is not that which; *8|
negated by the absolute negation that occupies thfl|
place where the ground exists and also does not
occupy the place where the object negated by th^
said absolute negation exists, then the relation
coexistence in the same locus between the grot;
and the probandum is an invariable relation.)

Mere coexistence with the probandum inl-
and the same locus does not fully bring out t

PRAMAtfACANDRIKA 55.

nature of vyapti or invariable relation. Take the
inference, f. i., "Yonder mountain is on fire,
because it has the character of knowableness."
Here coexistence in the same locus holds
(between the ground, * knowableness/ and the
probandum, 'fire,', i.e., the locus of ** fire/
e.g., the 'oven/ is also the locus of knowable
ness, i.e., the oven is a knowable object) . And
yet the inference is evidently fallacious. To
exclude such cases, the probandum (coexistence
wherewith will constitute vyapti) is qualified
as being one which is not that which is

; negated by the absolute negation which occupies
the place where the ground or H etu exists.
(This qualification of the probandum excludes the
case of the above fallacious inference and other
like cases- For -fire/ the probandum of the

; above inference, is that which is negated by the
absolute negation occupying the place where the
ground, viz., ' knowableness,' exists; e.g., fire:
never exists in the great lake and yet ' knowable-
ness * exists in> the great lake in so far as the lake
is a knowable object. Therefore the probandum*
'fire/ is not that which is absolutely non-existing

I . where the ground, ' knowableness/ exists.) Even
this qualification of the probandum, however, fails.

-to exclude the case of (the evidently fallacious),
inference, "The tree is in contact (with the;
monkey) because it has the generic character ol£ ;
substances." (Here 'contact ' is the probmdfym^;

56 MADHVA LOGIC

and it is not that which is absolutely non-existent
where the hetu, or ground, viz., 'the generic |
character of substances,' exists. 'The generic
character of substances * exists in substances.
* Contact * as a quality also exists in substances.
Hence ' contact ' is not that which is absolutely
non-existent in substances. And yet the inference
is fallacious.) To exclude such cases the proban-
dum is further qualified as being one which also
does not coexist in the same substrate with that
which is absolutely non-existent where the ground
or hetu exists. This excludes the case of c contact l
and the like. ['Contact' abides in substances.
It is thus not that which is absolutely non-
existent in substances. But despite this, contact
is also not that which does not coexist with that
which is absolutely non-existent in substances. ||
For 'contact* coexists with the absence of con*
tact in the same substance. 'Contact' of thft
tree and monkey coexists in the tree with th% ^
negation or absence of such contact in another
part of the tree. 'Contact' thus coexists with ; .gs
non-contact (with the absolute negation of
contact) in one and the same substrate.] Iff im
other words, non-contact or negation of cotti-^
tact being coexistent in the same substratft||
or substance with contact which is the object
of the negation, the rule, that the probandum
should not exist in the same substrate with
that which . is absolutely non-existent when*

■ -;■;

■■-3"*

t- i

■fe

'*>*"&§&

PRAMAtfACANDBlKl IJ

the ground exists, fails, and the case cannot
be regarded as one of invariable relation or

vyapti. ■'■''■

But all this, we hold, is fallacious and un-
sound. Take the case of the inference ' There is
rain on the hills higher up because the rivers
below are full.' Here from the fullness of the
rivers at the base we infer the occurrence of rain
at the top. In such inferences, where the
probandum occupies a different place from that
occupied by the prolans or ground, the above
definition of invariable concomitance as co-
existence, in the same substrate, of the ground
and the probandum altogether fails. It cannot be
said that our view is open to the self-same objec-
tion as the above view, for unfailing relation of
effect and cause holds equally, in our view, in the
ease of inferences where the prolans and the

probandum occupy different places. ^ .... ,

.. (We have so far discussed the meaning of
invariable relation. We now proceed to explairi
the different forms of invariable relation that
constitute the grounds of inference.) Dharmas
or properties of things may be related in four
different ways. Thus two Dharmas or properties
may be related by a positive symmetrical invariable
relation so that each is invariably concomitant
with the other. Two properties again may
be so related that one of them is an invariable
concomitant of the other, but not vice versa.

:.:-./.,■ 8; •.."'■'• ■■:..• . • .'. ,' m-

-„*i

58 MADHVA LOGIC

Two properties again may be invariably related
by the relation of mutual negation and exclu-
sion. Lastly, two properties may be so^jj
related that at least in one case where one is, the
other is, as also at least in one case where either ^
one is, the other is not. (This last relation is i
equal to the following three propositions taken
together where A and B express the two properties,
viz., ' At least in one case where A is, B is ' ' At
least in one case where A is, B is not/ ' At least ■*
in one case where B is, A is not.' The first
form of concomitance is similarly equal to the;|
two propositions — 'Wherever A is, Bis,' and
Wherever B is, A is/— taken together. The second
and the third will correspond respectively to the
propositions € In all cases where A is, B is ' and
'In no case where A is, B is.') We have av|
concrete illustration of the first form of invariable
concomitance in the unfailing relation that holds
between scriptural (Vedic) prohibition and con- f\
duciveness to demerit and sin and between scrip- J
tural injunction and conduciveness to merit and
righteousness . Thus whatever is scripturally
prohibited is productive of sin and demerit and
whatever tends to demerit and sin is scripturafly:.;!
prohibited. Here each of the two (related
properties)' is at once pervaded by, and pervader |
of, the other. Similarly it is] also observed thsk|
whatever is scripturally enjoined is also conducive
to merit and righteousness and whatever is

:-:m

PRAMAtfACANDRIKi 59

conducive to merit and righteousness is also laid
down by scripture. The second form of invari-
able concomitance is again illustrated in the case
of the relation between smoke and fire, as also
between the product of will and non-eternity.
Thus ( wherever there is smoke, there is fire * but
no invariable relation holds in the form ' Where-
ever there is fire, there is smoke,' for in the case
of the heated iron-ball (where fire is, but smoke
is not) the invariability fails. Here ' smokiness p
is- the pervaded and occupies a smaller area while
' fireness * is the pervader and occupies a wider
area. Similarly, 'whatever is a product of
will-causality is also non-eternal,' but no in-
variable relation holds in the form of the converse
of this, viz., 'whatever is non-eternial, is also a
product of will-causality,' for it is seen to fail in
the case of antecedent non-existence (which is
non-eternal and yet is no product of will-
causality). The third form of invariable con-
comitance is illustrated in the relation which
holds between the generic character of the cow and
the generic character of the horse as also between
that of the elephant and that of the lion. Thus
wherever there is the generic character of the cow,
there is nowise the generic character of the horse
and wherever there is the generic character of the
horse there is nowise the generic character of the
cow. No relation of pervader and pervaded holds
betwAPn *\thaT mio and thfl other, all relation

*•

MADHVA LOGIC

: being non-existent between the two. Simila;
~ T whatever is an elephant, is nowise a lion' ar
' whatever is a lion, is nowise an elephant'
well-known mutually exclusive relations to
noted in this connection. The fourth kind
invariable concomitance is illustrated in the
case of the relation which exists between tbe
the property of being a cook and the property of;
a man as also between the property of being
one of the five elements and the property
moving. Thus though in one particular ins
the property of being a cook and the prop
.of being a man may co-exist, yet in another
the property of being a cook may co-exist
that of being a woman to the exclusion of
of being a man as also in a third instance
property of being a man may co-exist with t
property of being a non-cook to the exclusion
the property of being a cook. In this case
-no relation of pervader and pervaded
between either one and the other, for inspite
a relation existing between the two, there
instances in which the relation fails. In the
way, though the property of being an ele
is co-existent in some instances with the pro
of moving {viz., in earth, water, air and
yet in the case of Skaia or ether (which is
element but does not move) the property of being;
an element exists to the exclusion of the pro
of movingi and in the case of the mind

■::"":-H9

property of moving exists to the

of the property of being an element.

is not an element and yet it moves.) , . ■ . ■ >

In all these when the property which is per-
vaded produces the cognition of the property that
is the pervader, we have what is called an in-
ference or anumana* The pervading property (of
which the inference produces the knowledge) is
called the inferred character, anumeya, or object

of inference.

The following objection may however be
raised to the view of inference expounded above,
viz., that it is not possible for the smoke that
exists in the mountain far away to produce a
valid cognition of fire in the man who exists
here in his house. The reply to this objection is
as follows. In the case of inference the instru- ||
mental cause is a known agency and not an im- ■•* '
known condition as in the case of perceptioa^
(The relation of smoke to fire is known to the man|^^
who makes the inference, but in perception the
action of the sensibilities with reference to the

*W*

object perceived is not known before the P«^|M
ception.) - ■ ■ "'~;*&ea#'

The reply however does not seem to be coff- -
vincing, for there are people (e.g., the savage**!^
the Cocoanut Island) in whom the perception-of
smoke at a distance does not call forth the
cognition of fire. The answer is that m $Hs
case though they have a cognition of the form of

■ -,vl

■V ■

■ *i
' "

62 mIdHVA LOGIC

the smoke, yet they have no knowledge of it as an
invariable concomitant of fire. For even when
an object like smoke might previously be cognised
as an invariable concomitant of something else
such as fire, it may fail to be cognised on account
of failure of memory as an invariable concomitant
of the latter in a fresh instance and thus fail to
produce the cogniton of fire. (Therefore in the
case of savages where the smoke was never cog-
nised as invariably related to fire, a cognition of
smoke in the first instance cannot possibly
produce the knowledge of fire.)

Therefore we conclude : when an accurately
and correctly cognised mark or sign is accom-
panied by a recollection of its invariable con- ;|
comitance with the thing marked or signified M
and is thereby able to produce the cognition of
the thing marked or signified in a fit place or
locality, we have what is called an inference or ,|
*anumana. Hence even though the form of the^jj
marked or signified thing may be already known : |
yet since the inference makes it known further in
relation to a particular place or locality, the||
inferential process is not useless or superfluous.
(The inference, in other words, entails a real
march of thought conducing as it does to a:|
new synthesis of the already known thing with a :|
place or situation to which it was not previously
known to be related.) Hence inference consists
of two factors: (1) invariable concomitance .(of

£Z

3ft

srsE

PRAMAtfACANDRIKA 63

■

the mark with the thing marked) ; (2) the
presence of the mark in a suitable place such
as will make possible the inference of the
thing marked either in the same place or some
other (causally or otherwise connected) place.
There is no rule that the mark should also
be cognised as a property existing in the subject
of the inference (for the mark may exist in
one place, e.g., the fullness of the rivers at
the base, and the thing marked, viz., rain, may
be proved to exist in some other place, e.g., at the
top of the hill).

A question here arises : how does the cognition
of the invariable concomitance arise? How in
other words, do we arrive at the knowledge of an
invariable concomitance between different objects
or events ? The answer is, by means of the
corresponding perceptions, inferences and testi-
mony. Thus in the case of the invariable con-
comitance of smoke and fire, we arrive at the
knowledge of the concomitance of smoke with fire
by the perception of the one together with the
other in the domestic oven and other places.
Here repeated observation and non -observation of
the contrary are the auxiliary conditions. But
how can perception which apprehends only that
which is present and is in contact with the
sensibilities, apprehend an invariable concomitance
that extends not merely to all cases (actual
and possible) but also to the past (and the

MADHVA LOGIC

: T"' ■

future and the remote) ? It would be possib
if you admitted a transcendental cont
(pratyasatti) of past and present and of h
and far (as Naiyayikas do), but according
to you, no such transcendental contact exists
(between the present sense organ and the past
and distant objects). The answer to this is:.
though there is no transcendental contact between J
a particular instance and its samanya or classg
yet since the past and the distant are capable
being drawn into relation to the present instan ?
by means of similarity or resemblance, the cognj|
tion of an invariable concomitance as extendi '
to all instances (actual and possible) is ful||
established. Invariable concomitance as m
known mediately by means of inference will
illustrated later on. The following are inst
of invariable concomitance known from authj»«
tative testimony. 'Whoever is a Brahmin,
person who must not be put to death,' '
animal that is a cow is one that must not 1
touched with the feet,' ' Whatever is enjoined j

. the Vedas, ought to be accomplished as a duty/ |
Inference is of three kinds: inference *^
effects, inference from causes, and inferenceJroi|J
phenomena that are neither causes
When an effect is the ground of our inference
the cause, we have an inference from an effi
e.g., when from the presence of smoke we J|

p the existence of fire. When the cause

'-** »\ -'

PRAMAtfACANDWKl 65

the ground of an inference of the effeet, we have
an inference from a cause, e.g., when we say,
'yonder mass of clouds which owes its special
character to its own cause proves an impending
rain-fall.' When a particular mark proves the
existence of a probandum without being either
the cause or the effect of it, we have an inference
from something which is neither a cause nor an
effect, e.g., when the presence of taste proves the,,
existence of colour.

Inference may be divided again into two
classes from another standpoint, viz., into infer-
ence of what is specifically observed and inference
of what is generically observed. Thus whei*|*he
object inferred is perceptible we have an inference
of the specifically observed, e.g., when &»»
inferred from smoke. Where the object inferred
is not perceptible, we have an inference of the,
generically observed, e.g., when the visual,

; sensibility is inferred from the cognition of colour.

Some (the Naiyayikas) hold that inference is

of three kinds, viz., Kevalanvayl inference,

i Kevalavyatireki inference and AnvayavyatireMM

r m

inference. ;,:

(According to the Naiyayikas) the pakja or the
;*; subject of an inference is that substrate which is

to be proved to own the, probandum as its property.

To exclude the sapaksa or the co-ordinate o*H^
Isttbject and other like substrates, the subject is

defined as that which is to own the proband*** as a

tfer,- 9 - ., • •■ .■■■■'■

:■?-- t

'C-

*■*

66 = 'MADHVX LOGIC

property. Since the property here means that the $|
cognition whereof is to be produced by the sign
or mark (in the inference), the definition is not too
wide. (The sapaksa is also a substrate of a pro-
perty, but it is not the substrate of the property
the cognition whereof is to be produced by the mark
or sign. It is the substrate of a property which
is homogeneous with the property that constitutes
the probandum, but it is not the substrate of the
property which is the probandum itself.). But it
may be said that the hetu or ground has also this
character of being characterised by the probwndum^^
as being related to it by conjunction (and thus the |
definition is too wide as applying to the hetu or ^
ground as well). To meet this objection the J
paksa has been defined as a substrate. (The hetu : r$
is not the substrate of the property that constitutes J|
the probandum, but is related to it only by J .^|
concomitance or conjunction, but the paksa
is the substrate in which the probandum is to
be proved to abide as a property.) The
sapaksa or co-ordinate of the inferential subject is
that substrate which owns a property which is
homogeneous with that which constitutes the
probandum. To preclude a too wide application^ 1
to the case of the smoke in the oven, the definition
includes the word 'substrate.' [The smoke in the
oven is the familiar instance or drstanta and n

• * #

the sapaksa. It is distinguished from the la
by the fact that it is not the substrate, while

PRAMAtfACANDBIKA 67

sapaksa is the substrate of a property (in this
case the kitchen fire) homogeneous with the
property constituting the probandum (i.e., the fire
in the mountain)]. To have defined the sapaksa
simply as the substrate of the probandum (and not
of a property homogeneous with the probandum)
would have been absurd. (For the sapaksa and
the paksa would in that case have been identical,
and a sapaksa other than the pak§a would have
been an impossibility.) Therefore in the defini-
tion the word ' homogeneous ' has been included.
The meaning is that the sapaksa is the substrate
of a property homogeneous with the probandum
and as such is devoid of the character of
uncertainty (i.e., the uncertainty that characterizes
the paksa as the substrate of the probandum). V
The vipaksa or contra-ordinate of the inferential '^
subject is a substrate devoid alike of the pro-
bandum and every property homogeneous with
the probandum. The words ' devoid of the pro*
bandum' distinguishes the vipaksa from the
paksa while the words ' devoid of every property,
homogeneous with the probandum' distinguishes
it from the sapaksa. (The sapaksa is a similar
instance in which the existence of a pror^iy like
the probandum is known for certain, and the
vipaksa is a dissimilar instance in which the
non-existence of the probandum and of all
properties similar to the probandum is known
for certain.) ;■-," ;.'*-.■:>;'■ ; ' '-'■ '- -

68 MADHVA LOGIC

[These definitions of the inferential subject,
the co-ordinate of the inferential subject and the
contra-ordinate to the inferential subject, prepare
the way for the definitions of Kevalanvayi and
other forms of inference.]

Thus the Kevalanvayi inference is one which
is based on a ground that pervades the subject
and also exists in its co-ordinates but which has J
no contra-ordinate to its subject actually existing.
[In other words, a Kevalanvayi inference is one |
that is based on numerous instances of agreement-
in presence but is without any instance of agree-
ment in absence.] - ^

Kevalanvayi inference cannot be defined

simply as an inference in which there existed
no co-ordinate to the subject, for in this case if ;;
will be indistinguishable from the inconclusive
reasoning 'All things are nameable, because theyg
are knowable.' (Here the subject of the inference
being 'all things' or 'everything/ no eontraN^
ordinate to the subject exists, but since there is
here also no co-ordinate to it, the ground of the
inference, viz., the invariable relation between
* knowableness ' and ' nameableness/ is without a
corroborative familiar instance and thus lacks
material certitude.) Hence the words *in whi

the ground is existent in the co-ordinate.*
even thus the definition remains imperfect for
inference which has no <x>ntra-ordinate to its
ject and in which the ground is existent a3m

Y«jp : iki

Swh

Prama^acandbikI 6$

f*w- ■

Ithe co-ordinate, is not distinguishable from the

I fallacious reasoning in which the ground does not

Sexist in the subject of the inference, e.g., the
inference 'sound is nameable because it is visible.'
To exclude such fallacious reasonings Kevalanvayi
inference is further specified by the qualification
that it is an inference in which the ground must
exist in the subject. Even this, however,
is not sufficient, for this by itself does not
distinguish it from such fallacious reasonings
as ' The jar and sound are nameable, for they
are characterised by shape' where the ground,

''?■ viz., 'shape,' exists only in one part of the subject,
viz., the 'jar,' and not in the other part, viz.?

ft 'sound.' To exclude such fallacious reasonings

land prevent these being confounded with
Kevalanvayi inference, the ground is stated as

{ being not merely existent in the subject but also
as pervading it. The pervasion of the subject

£ may again be of two forms, viz., (1) Where the
ground pervades both the subject and its eo-'

* . ordinates, and (2) where the ground pervades the
subject but exists only in a portion (i.e., in some)

i-0f the co-ordinates. > Sound is namable, because*
is knowable, just as is the jar' illustrates the case
where the ground pervades not merely the subject
but also its co-ordinates. In the same reasoning
the words 'because it is a quality just as is colour
On place of the words 'because it is knowaWe

IP* as is the jar ') illustrates the case where the

MADHVA LOGIC

ground exists only hrsome (i.e. a portion of 1
entire extent) of the co-ordinates. (The gro\&
viz., 'quality' is predicable of some nameabl
things, but not of all narneable things.)

The invariable concomitance which constitu
the ground of inference is of two kinds, viz., in-
variable concomitance as agreement in presence
and invariable concomitance as agreement fe
absence. Invariable concomitance as agreement
in presence consists in the invariable confj
comitance of the ground of the inference with
the probandum. Invariable concomitance
agreement in absence consists in the invariabli
concomitance of the absence of the probandm
with the absence of the ground. In the case
agreement in presence, the ground is the pervad
and fhe probandum is the pervader. In the
of agreement in absence, the absence of the pro--
bandum is the pervaded and the absence of t
ground the pervader. In every case an invariable
concomitance is understood as following in t
wake of the pervaded. In the case of the abov^
inference ' sound is narneable, because it is know-
able, just as is the jar,' the invariable

' knowable ' with ' narneable *
agreement in presence only. T]
presence here is ' whatever
narneable, just as is the jai
no agreement in absence
1 what is not narneable is

cam

comitance of
based on an
agreement in
knowable, is
But we have
the form

.. M + +;

PBAMi^TACANDRIKA. 71

knowable,' it being impossible to get any
case of an object which is not nameable as
illustrative of such absence, since all things are
nameable and there is nothing that is not name-
able. This is why inference based on such con-
comitance consisting of agreement in presence
only is called inference of the Kevalanvayi type.
Where the ground pervades the subject, where no
co-ordinate of the subject exists and where further
the ground is excluded from every instance of a
contra-ordinate to the subject, we have an inference
of the Kevalamjatirekl type (according to Nyaya).
If Kevalavyatirekl inference had been defined as
one in which the ground is excluded from (some
instances of) the contra-ordinate to the subject,
the definition would have been too wide and
would have applied to the fallacious reasoning
based on a non-invariable ground, 'yonder moun-
tain is on fire, because it is a mountain.' To ex-
clude such reasonings the definition lays down that
the ground must be excluded from every instance
of a contra-ordinate to the subject. In the present
case, the ground, viz., 'being a mountain,' though
excluded from such contra-ordinates as 'the great
lake,' 'the sheet of water,' etc., is yet not excluded
from such other contra-ordinates as 'a fireless
mountain/ 'a hill without fire.' Hence the ground
is not excluded 'from every case of a contra-ordinate
to the subject' and thus does not come up to - the
requirements of the definition (of a Kevalavyatireki

:V«

MADHVA LOGIC

inference as set forth above). Thus the defi
does not apply to such cases and is not too wi
(as applying to such fallacious reasonings as wel
It must be noted however that the mere fact of
ground being excluded from every instance of
contra-ordinate or dissimilar instance does no
suffice as a complete definition of the Kevatid.
vyatirekl inference, for as such it remains
distinguished from Anvayavyatirekl inferen
based both on agreement in presence and agre
ment in absence. To exclude such inferen
the definition stresses the fact that ' no
ordinate of the subject should exist,' i.e.,
similar instance where the existence of the pro
bandum should be known for certain should e
(In Anvayavyatireki inference, the existence of t|
co-ordinate or similar instance is a sine qu& n
while , in Kevalavyatireki the non-existence

. $hs> co-ordinate is a sine qud non.) Bute
this added qualification does not suffice aig
definition of the Kevalavyatireki inference,
as such it has a too wide application to
fallacious inference based on a Svarupasi
ground (i.e., on a ground that does not exu

*in the inferential subject), viz., 'The
of the finite individual has a soul aceo$
panyingit, since this body is conscious. '....
the ground, viz., 'consciousness' is non-existent
the ' body ' which is the inferential subject,
dead body, e.g., is devoid of consciousness.

S.1' •

PRA&f AtfACANDBIKl 73'

exclude such cases, the definition insists on the
existence of the ground in the subject. Even
this, however, does not suffice, for as so qualified
the definition applies to the fallacious inference
based on a ground that exists in one part of the
subject (and not in the whole of it), viz., "The
finite individual and the Lord are omniscient,
because they are all-creating." (Here the ground
'all-creating' is true only of the Lord and not of
the finite individual, i.e., it holds good of one part
of the subject and not of the whole of it,) To
exclude such cases the definition says, 'The
ground must pervade the subject.' The following
is an instance of a Kevalavyatireki in£
answering to all the above requirement
Lord is all-knowing, because He is alft^^ting/ „v^Jr
It is based on the invariable agreement/if £bsen<^ j>
viz., 'Whatever is not all-knowing, is ^n^ 1 ^ J£
creating, just as is Devadatta.'
innumerable instances illustrating this a
in absence but none illustrating the positive
ment in presence between ' what is all-creating ' and
'what is all-knowing,' for Kamkr§na and other
Incarnations of the Lord are comprised in the
subject of the inference (and therefore cannot serve
as corroborative illustrations) while other finite
individuals are non-omniscient (and therefore
cannot be cited as illustrations of the agreement
in presence). For these reasons (©«., that it is
based on an invariable relation which can be

MADHVA LOGIC

:&S :

actually observed only as agreement in absence!
and not as agreement in presence) such inference |
is called kevalavyatireki inference.

An inference in which the ground pervades the
subject, exists in the co-ordinates or known similar
instances wherein the probandum exists, and is ex- .|
eluded from every instance of a contra-ordinate to
the subject is an anvayavyatireki inference. An.'-$
anvayavyatireki inference is not completely defined fl
as one in which the ground is found to be non-
existent in the contra-ordinates, for as such the
definition fails to exclude the fallacious inference
based on a non-invariable ground, viz., ' The body I
of the finite individual is non-eternal, because it
has the character of the element of earth.' In this
inference the ground, viz., 'character of the
element of earth,' though non-existent in such|f§
eternal entities as the ether, etc., is yet existent ii
such other eternal entities as the atoms of earth^j
etc. (Thus though excluded from some objects
which are not non-eternal, it is yet not excluded
from some other objects which are also not non-1
eternal.) To exclude such cases, the definitioi||
says, ' the ground must be excluded from ever^
instance of a contra-ordinate.' But this also by
itself does not suffice as a complete definitioi
for as such it remains indistinguishable from
kevalavyatireki inference based on agreement I
in absence only. Hence the definition adds il
words, ' the ground must exist in the co-ordinates!

V-

tRAMAtfACAND&IKA 16

ffi- (The kevalavyatireki is devoid of co-ordinates.}

Even with this added qualification, the definition
if" has a too wide application to the fallacious
;, : " inference based on a svar&pasiddha ground (i.e., a
| ground that does not exist in the subject), viz.,
;•• 'Devadatta is all-knowing, because he is all-
V. creating.' (Here the ground 'all-creating' is

excluded from all cases of 'not all-knowing' and
:V also exists in 'what is all-knowing,' viz., the
J:.-' Lord. But it does not exist in Devadatta, the
|- subject of the (inference). Hence the defini-
| tion further adds, 'the ground must exist in ■■

the subject.' Even now however the definition ; ;
\y. applies to the fallacious inference based on y
&,a bhagasiddha ground (i.e. , a ground that exists; .;;?

only in one part of the subject and not the whole
I of it), viz., ' The mountain and the lake are oafW"
%' fire, because they smoke.' (Here the ground,^
J' smoke' exists in one part of the subject, tiz.^y
J •: 'mountain,' and not in the other part, viz., the £|

'lake.') To exclude such cases the definition
jfc* says 'the ground must pervade (i.e., exist in the,
f§ whole of) the subject.' As 'smokiness' is

absent in the 'lake' (being true only of the
h 'mountain ') it lacks the character of pervading
H the subject (in the above case) . . ''"-^M

Such anvayavyatireki inference based bo^jg

on agreement in presence and agreement in
H absence may again be of two kinds. 33«^ v
i.«tf .<• « inference with a ground that .

-*;

^:-

l 9m

MADHVA LOGIC

■A

exists in every case of a co-ordinate, or aga
it may be an inference with a ground t
exists only in some (not all) instances of
co-ordinates. For example, the inference, 'The
finite individual is eternal, because there is no
cause that can put an end to it' is a case;
of an anvayavyatireki inference with a ground
existing in all cases of the co-ordinates. W\
respect of all eternal things such as the ether,
etc., the ground, viz., ' absence of a cause of|
an end or destruction ' holds good. Again, the
inference ' The mountain is on fire, because it-
smokes ' is an instance of an anvayavyatireki
inference with a ground that exists only in some
(and not all) its co-ordinates, for in some ' fiery V
things (e.g., the red-hot iron ball) ' smoke 1
(which is the ground of the inference) does not
exist. This latter example is a typical anvaya-
vyatireki inference and takes its name from the
following two invariable relations (of presence
and absence) on which it is based, viz., 'What-
ever smokes, is on fire, just as is the oven ' anig
' Whatever is not on fire, does not smoke, just

as is the great lake.' ' ■ ?|jj

Inferences based on agreement in presence
and agreement in absence take their character
from the corresponding invariable concomi
Thus we have invariable concomitance based o
agreement in presence in * Wherever there
smoke, there is fire.' And we have invari

fc PBAMA^ACANDBIKA

concomitance based on agreement in absence in
'Wherever fire is not, smoke is not.' And
inferences that resemble these two forms of
concomitances considered together are called
Anvayavyatireki Inferences or inferences based on
agreement in presence and agreement in absence.

All this however (i.e., this division of inference
into kevalanvayi, kevalavyatireki and anvaya-
vyatireki) we (the Madhvas) reject as untenable.
For we consider an agreement in absence as being
unsuitable for proving the presence of the sadhya
or probandum. In proving the presence of a
positive entity by means of the presence of
another positive (entity), an invariable relation
between the absence of one and the absence of
the other has no logical scope. (An m-'
variable relation between the negation of one
thing and the negation of another does not
justify any positive step from the presence of one
to the presence of the other.) For in this case
the positive ground exists in the subject of the
inference (and thus falls within the domain of
affirmation) while the invariable relation as an
agreement in absence (the absence of the proban-
dum and the absence of the ground) belongs to the
domain of negation and thus occupies a different
place. Thus the invariable relation occupies one
place (the domain of negation) and the ground as a
property of the subjeot occupies a different place
(the domain of affirmation). (Hence there 18 no

78 MADHVA LOGIC

relation between the two to justify an inferenti
step.) How then, it may be asked, are kevala4
vyatireki inferences in vogue? In this way.^
Here also the real ground is an invariable relation
in presence, e.g., the invariable relation between
'omniscience ' and ' all-creativeness ' in the above
inference. But it is impossible to cite positive
similar instances of this agreement in presence in
response to the demand for an indication of the ■%
actual places where this invariable relation holds.
Hence for accomplishing this end by means of
inference, an invariable relation of absence ac-
quires relevancy (for our purpose). For example,
if in the above instance it is asked: — "What
proof have we of an invariable relation between
' all-creativeness ' and ' omniscience ? " w6g
can say at once that 'all-creativeness' must
be pervaded by 'omniscience,* for it ift||
that which is negated by the negation which
pervades the negation of 'omniscience.' When
one thing is so related to another thing that the
negation of the former pervades the negation of
the latter, the former thing is invariably relatafij
to the latter. (Thus if A is so related to B that
the negation of A is pervasive of the negation of
B, i.e., if 'All not-B is not-A,' then A is invaria
related to B, i.e., 'All A is B.') This relation is
admitted, e.g., by the person seeking 'fire,'
holding between ' smokiness ' and ' fireness^
In the case of the so-called dtwayavyatire^^

ft*?--'

PRAMAtf ACANDRIKA 79

inferences supposed to be based on agreements in

% presence as well as absence, the agreement in
absence is in reality purposeless and out of place.
The invariable relation in such cases is sufficiently

:■ established by perception, etc., (of the instances of
agreement in presence). (Thus the agreement
in absence is without real usefulness.) Notwith-

■ : standing this it may be conceded that the agree-
ment in absence serves some sort of purpose as

. indicating in a way that the positive relation of
agreement in presence is not negatived by any
instances of the failure of the agreement (i.e., by

; any instances to the contrary.)

According to another (Nyaya) classification,
inference is of two kinds, viz., (1) inference for
oneself, and (2) inference for convincing others.

t Of these, inference for oneself is the cause of self-

| conviction and the knowledge one gathers for one's
g-: own self. ...'., J

An inference for oneself takes place in the

|foUowing way. A person in the first place makes\

repeated personal observation of the togetherness

of ' smoke ' and ' fire ' in the oven and other

^places. From such observation he gathers that'

there is an invariable relation between ' smoke '

and 'fire/ Having gathered the invariable

relation, when he draws near a 'mountain' and

^thrown into doubt as to the existence of ' fire-

the mountain, he notes the trail of smoke

from the mountain and recollects the

^v.

:.:'■■■

- -A'-a-

MADHVA LOGIC

invariable relation between 'smoke' and 'fire
(which he gathered from repeated previous obser
vation). "When the recollection takes place (t
'where smoke is, fire also is'), and the saitj
person draws near the mountain (with the trailin
smoke), the knowledge at once flashes forth that
'the mountain yonder has smoke which is an
invariable concomitant of fire.' Thus does hej
make an inference for himself. This last step
(i.e., the knowledge that the mountain possesses
' smoke ' which is invariably related to ' fire ') is
called paramars'a or inferential reasoning. From
this (reasoning process) arises the knowled
that 'the mountain is on fire.' (The abov$
illustrates inference for oneself.) As regard^
inference for others, it is a fully-expressed reason-
ing consisting of five steps which are employed
to convince others as to the way of inferring

* fire ' from the (observed) presence of ' smoke^*
The five steps are :— (1) ' Yonder mountain is
fire,' (2) 'because it smokes,' (3) 'Whate
has smoke, is also on fire, just as is the oven*
(4) 'So is it with this (mountain yonder),' (

* therefore it (the mountain) is so (on fire).' B
all this even a second or third person is ass
of 'fire' from the knowledge of the pres
of the established mark or sign thereof ({•«»»
the sign of 'smoke'). The above five steps i
called respectively (1) Pratijiia, (2) Hetu,
Udaharanaffl, (4) Upanayah and (5) Nigamm

PRAMltfACANDBIKA 81

Of these pratijna '(the statement of the proposition
to be proved) consists in the statement of the
subject of the inference as possessing the proban-
dum as a property. (In the above inference),
e.g., the statement ' Yonder mountain is on fire '
is the pratijna. (2) The hetu (the ground of the
inference) is the statement of the mark or sign
" with a suffix indicative of its instrumentality
(towards the conclusion), e.g. t the statement
'because it smokes.' '(3} The drstantah is the
concrete case in which the invariable relation or
vyapti is apprehended. It is of two kinds, viz.,
(a) sadharmyadrstantah, and (6) midharmya-
drstantah. A concrete example in which an
invariable relation of presence is apprehended
is called a sadharmyadrstantah, e.g., in the in-
ference (of fire)_ from smoke, the case of the
oven. A concrete case in which an agreement
in absence is apprehended is a vaidharmya-
dfstantah x e.g., in the same inference from
'smoke,' the case of the great lake. The
udaharana is the statement of the concrete case
or example as exemplifying or illustrating the
invariable relation of which it is a case in point.
(The udaharana is thus a statement of a case in
: point while the drstantah is just the concrete case
?'■!' and no statement of it as illustrative of the
; invariable relation.) It is of two kinds, viss,^
|; (a) sadharmyodaharanam, and (b) vaidharmyo-
daharanam. A statement of a concrete example

m li

mIdha logic

*",^a

illustrating an agreement in presence is if
sadharmyodaharanam > e.g., the statement 'Whafc|
ever is smoky, is fiery, just as is the oven.'
statement of a concrete example illustrating an
agreement in absence is a vaidharmyodaharamnitfi
e.g., the statement 'What is not fiery, is not
smoky, just a&is the great lake/ (4) Upanayah^
is the statement of the mark, the invariable!
relation whereof has been well-established in the
Concrete example, as existing in the subject of the
inference. It is also of two kinds according to
the nature of the concrete example (which

establishes its invariable relation)* ' Th^t

mountain yonder has a trailing smoke just as the
oven' is a case of a sadharmyopanayah. 'The
mountain is not devoid of smoke like the lake:
is a- case of a vaidharmyopanayah. (5) Nigt$
manam (the conclusion) is the statement of thfe>
subject '(as characterised by the probandum) as
proved or demonstrated, e.g., the statement
* Therefore, yonder mountain is om fire-'

All this, we hold, is unsound and untenable. 4f|

there is no scope for the two kinds of vydpti
invariable relation (in inference), so also there
no scope for the two kinds of udaharana (illustratl
ing such relation). Besides, the alleged rule as
the necessity of five steps in inference is an
proved assumption. The way in which the 8
posed necessity of the five steps may be refu
has been set forth in the Paddhati and the

■ ■-■

' *',

■■■''.

§£;' PRAMll?ACANDBIKA : 83

is referred thereto for an understanding of the
method of the refutation. This closes our disser-
tation on the nature of inference.

We shall now discuss the fallacies of reason-
ing. The fallacies are of two kinds, viz., (I)'
fallacies arising from discrepancy or contradiction
! (virodha), and (2) fallacies of inappropriateness
(asangati). Of these, the fallacies of contradiction .;
are of three kinds, viz., contradiction in the
; pratijM or proposition to be proved, contradiction
.. in the hetu 01 ground, and contradiction in the ;
)■- dr§tantah or illustration. Contradiction in the
*.': pratijM again may be of two kinds, viz., contra-
: diction of the pratijM or proposition to be proved ;.
J: with what is established by the recognised source^ J
I^Vof knowledge, and internal self-contradiction ifrjM
the proposition in question. Of these again, --con^#
' tradiction with the evidence of the accepted sources ] :
% of knowledge may be of two kinds, viz., contradict ^|

tion with the deliverance of a stronger evidence
¥■ or proof, and contradiction with the deliverance of
an evidence of equal strength or force. The follow-
H -ing is an example of a proposition in contradictioig

with the deliverance of stronger evidence :— 'The
j^subject-matter of controversy (i.e., the world)
is false; because it is an object of perception j
Whatever is an object of perception, is false, pfc
as is the silver that is (falsely) perceived iaJWO;
^shining mother-of-pearl.' . This conch^ /$.•;:
ieotradicted by the ,e*idence <t £*£!£ MM

.KiSr>

=*>

MADHA LOGIC

which objects like the jar, etc., are presented
real, by the evidence of inference also as it proves
the opposite, viz., * The subject under discussion^
(the world) is real, because it yields expected
results, just as admittedly real things do,' and
lastly by the evidence of scriptural testimony^
which declares the world to be real. It thus J
runs counter to the combined evidence of percept
tion, inference and authoritative testimony.
Hence it is in contradiction with evidence of
stronger force or strength. As an example of|
contradiction with evidence of equal force <H?:;
strength we have the following pair of inferences :
—(1) ' The disputed subject (i.e., the world) i»:|
false ; because it is perceptible ; just as is the silver
perceived in the locus of the mother-of-pearLVJ
(2) 'The disputed subject (i.e., the world) is real;
because it is the object of valid knowledge ; josfeS
as is the self.' In these two inferences, the
corresponding invariable relations as also the
presence of the respective grounds in the corres-
ponding subjects being exactly of the same order,
we have here a contradiction between evidences
of equal strength and force. An internally discre-
pant or self-contradictory statement may, agafliJJ
be of two kinds. It may be an apasiddhanta
a jatL An apasiddhanta is an asserted pro]
tion that contradicts the accepted beliefs of one'
own school of thought. Since one has subscril
to the tenets of the school to which one i

ra.

PRAMA^ACANDRIKA 85

elected oneself, in making a statement contradict-
ing such tenets one is really contradicting oneself.
This is why an apasiddhdnta is regarded as a
form of self-contradiction. A statement of the
existence of God by one belonging to the atheistic
Sankhya School of thought is an example of this
kind of self-contradiction. When one refutes
oneself by the very assertion one makes we have
that form of self-contradiction which is called
jati. For example, one who says 'My mother
is childless ' commits this form of self-contradic-
tion. Virodha or contradiction in the hetu or
ground is also of two kinds, viz., svarupdsiddkih
and avyaptih. The following is an example of
svarupasiddhih.: — ' Sound is non-eternal, because
it is visible. ' Here visibility is non-existent in
sound, sound being audible (and not visible).
Avyaptih again is of three kinds. We have
avyaptih when the mark or sign (the ground of
the inference) is related to the probandum as well
as the absence of the probandum. iWe have also
avyaptih where the mark or sign is related to the
absence of the probandum without being related
to the probandum. Lastly, we have avyaptih
where the mark or sign is unrelated both to the
probandum and the absence of it As an example
of the first (of these three) we have the follow-
ing :— f Sound is non-eternal, because it is know-
able/ The following is an example of the
second :^-' Sound is eternal, because it is a

MADHVA LOGIC

product of will/ As an example of the third
have : 'All that is, is non-eternal, because
exists.' In this last example, the subject of tbli
inference being ? whatsoever that is' the hetu^
or ground is unrelated both to the probandu
and the absence thereof. (Since the subject h
this case is 'all,' i.e., everything actual
possible, there is nothing outside the subject
serve the purpose of a probandum or the abse:
thereof. Therefore the hetu or ground as
property of the subject is without relation to
probandum as well as the negation of it. T
being no probandum, there is also no absence
probandum and thus the hetu or ground is wi
out relation to either.) Contradiction in
example is of two kinds, viz., contradicti
arising from the example being without relati
to the probandum, and contradiction arisi
from the example being without relation to
ground. The former is illustrated in the follow
ing: — 'The mind is non-eternal, because it ha|
shape, just as the atom has. 5 The second*;*}
illustrated by the same inference if in place of.
words 'the atom' we substitute the
r action.' (We shall now deal with the f
of inappropriateness.) An example of the £
of inappropriateness is addressing to an admitted!
theist the traditional theistic argument: — *Tfce§
earth and the rest have an intelligent
because they are effects, just as is a piecey

. L ... - : \\

PBAMAtfACANDRIKl -'W- '

cloth.' The inappropriateness consists here in
the absence of any demand for such an inference
(the addressee being a theist and therefore not
standing in need of being convinced). The in-
appropriate is just that for which there is no
real need. This is the definition of the inappro*

priate. -

Others - (the Naiyayikas) enumerate the fol-
lowing five as the essential characters of the hetu
or ground of a valid inference :— (1) existence in
the subject of the inference as its property or
Hharma, '(2) existence in the co-ordinates or
' similar instances, (3) exclusion from the contra-* ;
ordinates to the inferential subject, i.e., from :
; . dissimilar instances, (4) non-sublation of its visaya
§ or object* (5) absence of a counter-feef u or : : ,|
f' : counter-ground leading to a contradictory conclu-
sion. Of these, all the five characters are attri-
buted to the ground of an amayavyatirete-j
(^inference. The ground of a kevalanvayi inference; :|:
§ however should possess only four of these^-j

there being no contra-ordinate or dissimilar
§ instance in such inference and so the exclusion of -^
Ij the ground from the contra-ordinate or dissimilar
instance being impossible in this case. The
„ ground of the kevalavyatireki is likewise required
feto possess only four characters, there being no
co-ordinate or similar instance in such inferen^g^
|bkI therefore existence of the ground in the:
lidmate.or- similar, instance being <*M4jg

■A *i -

•4-2

"**

MlDHVA logic

-.1-

■:&

: -Ki

- .1 ■ *

■ r t*

yf't I

question in this case. The fallacious hetu
ground (according to these Naiydyikas) is
ground that possesses only some of the abo
characters and does not possess the rest. T
fallacious ground is either the asiddha, or t
viruddha, or the anaikantika, or the kalutyay
padista, or the satpratipak$a ground or hetu. An
asiddha or unestablished hetu or ground -is one th
is devoid either of the character of invariable rela-
tion (to the probandum) or of the character afg
being a property of the subject of the inference
There are three kinds of an asiddha or uflg
established hetu, viz., an diraydsiddhah hetu,
smrUpasiddhah hetu and a vtjdpyatvdsiddhah hetu>
An d&raydsiddhah hetu, i.e., a ground with i
diraya or substrate unestablished, is of two kin
mz. r z, ground with an asserted substrate 'that d
not actually exist, and a ground with an assertftjp
substrate in which the existence of the
bandum is admitted as an established fact,
former is illustrated in the following inference It,
' The sky-lotus is fragrant, because it is a lot
just as is the lotus in the lake/ Here
substrate of the ground is the sky-lotus,
a sky-lotus nowhere exists. The second
is illustrated in the case where the
inference 'The earthy etc., have an intelligent!
author, because they are effects, Just as r jfej
piece of cloth ' is employed for the benefit
convinced theist. Here the probandunbs:

prama^acandrikI 89

admitted (before the inference) to be a character
of the subject, there is no subject in which the
existence of the probandum is doubted. Thus
there being no subject to which the probandum
may be (hypothetical^) attributed for purpose of
demonstration and proof by means of the in-
ference, the substrate of the hetu or ground is
non-existent for logical purposes. (Here the
subject of the inference is ' the earth, etc,/ and
this, according to the theist, being admittedly the
handiwork of God, the proving of the same b
means of the presence of the ground
superfluous. Thus the hetu has no logic
tion with reference to the asserted su
this is the same as saying that the
subject is no logical subject, i.e., does
for logical purposes. The svarupasiddhah
ground is illustrated in the following : * Sou
non-eternal, because it is visible.' Here the
ground is 'visibility/ and this is non-existent
is sound, sound being audible (and not visible).
The vyapyatvdsiddhah hetu or ground is of twflK^
kinds, viz., a hetu or ground devoid of all relatio&g|
to the probandum, and a hetu or ground rdateSg*
to the probandum only through an extraneous
|c©»dition. The former ia illustrated . in -th^|gg
^following :— ' All that is, is momentary, becao*#p|||
^exists/ -Here as the subject of the infei^»^gfe||"
S»B4 (and thus comprises everything), there WW? "' "
similar instance or co-ordinate to the subject outside

12 , : v m

■£

':■,'

MADHVA LOGIC

the subject, and therefore the hetu or ground has
relation to anything outside the subject, i.e., haal
no relation to the probandum (there being n#
probandum in reality). The latter is illustrated
in the following : — ' Vedic sacrifice (of animals)
is a cause of sin, because it entails destruction of
life, just as is the killing of a Brahmin.' Efere^
' scriptural prohibition * is the extraneous condi^j
tion through which ' destruction of life ' become*^
* productive of sin.' (On condition that the;
1 destruction of life ' is also one that is prohibited^
by scriptures, is it a source of sin. Thus not^
' destruction of life * as such, but such ' des
tion of life ' as is scripturally prohibited, is &|

source of sin.) (What, then, is an extraneous|
condition or upadhi ?) An upadhi or extraneous!
condition is defined as one which pervades thp|
probandum but does not pervade the ground.
' Scriptural prohibition ' (in the above inference)^
is an extraneous condition in this sense. Tlnl^
wherever there is productivity of sin, therfc
scriptural prohibition may or may not be. Iff
the subject of the above inference, e.g., (&«*,.!$
sacrifice sanctioned by Vedic prescription) there
is destruction of life but no scriptural prohibition:
But how, it may be asked, is the presence of
extraneous condition a defect in reasoning?
this way, we reply.) The presence of an extra-
neous condition is a defect as revealing the faita
of the concomitance (on which the inference

*::*>.
a

»*

■ ■■'■■

m->:-*

; fRAMA?TACANl)RIKi 8f '

based) or as showing the inference to lead to a
contradictory conclusion, i.e., as showing forth
the subject as related not to the probandum
but to the contradictory of it. For the extraneous
condition, e.g., scriptural prohibition, being non-
pervasive of the ground (viz., animal sacrifice),
may also be excluded from the subject of the
inference (viz., Vedic sacrifice), and being
so excluded may effect the exclusion of the
probandum which it pervades, viz., productivity
of sin. And so it may show forth the ground*
viz., 'animal sacrifice/ as being related, not to
the probandum, but to the contradictory of the

! probandum. (In other words, it proves the
failure of the concomitance between the ground
and the probandum, for as non-pervasive of the
ground, it shows forth the ground as capable of /■■;■
falling outside its range and thereby as failing to
be related to the probandum which it pervades.)
For the pervading (i.e., the extraneous condition :,'
pervading the probandum) being itself non-

■ pervasive of, i.e., capable of being unrelated to, ••: ,
the ground, the pervaded (i.e., the probandum '%
which is pervaded by such extraneous condition) ^||
must also be similarly capable of being unrelated
to the ground. And so the subject of the infer-
ence being shown to be reduced to the position
of the contraordinate to the subject, the ground

/asserted to be existing in the subject is shown ta ;

:be, reduced to the position of a fallacious non- ,..,.

■■?m

92 MADHVA LOGIC M®

&£]

';i^a»

^ ■

■W

invariable ground. Further, the upadhi
shows forth the subject as related to the cont#i||
dictory of the probandum, i.e., it gives rise to an||
inference proving a contradictory conclusion,
the following way. The upadhi or extrane
condition, being itself excluded from the subj
of the inference, also effects the exclusion the
from of the probandum which it pervades,
thus the absence of the upadhi establishes t
absence of the probandum. Hence we get t
counter-inference establishing a contradicto
conclusion: — 'Vedic sacrifice is not productft&f
of sin, because it is not prohibited, just as i
the daily meal.' In this way ' sacrifice of life '
is a vyapyatmsiddah hetu or ground, being subj
to an extraneous condition (in the matter of i
invariable relation to the probandum). u$)ad
may be of four different kinds. In the first p
an upadhi may be an extraneous condition
vading the probandum without qualification <
restriction. Secondly, an upadhi may pervade t
probandum as qualified by a property of
subject of the inference. Thirdly, an upadhi
be a condition pervading the probandum as quaJk f
fied by a property of the ground or sddhana. La
an upadhi may be a condition pervading the : f|!jfl
bandum as qualified by a neutral property (**&>:$■
property which is neither a property of the subje<S||
nor a property of the ground of the inference)*!
The first of these is illustrated in the i

■"R*;

ssff*

W;i^y

PRAMA^ACANDKIKA

V*'.

* Yonder mountain is smoky, because it is fiery,' ,
where the presence of greenwood is the upadhi or
extraneous condition. (Provided the fire is green-
wood fire, it is a sign of smoke. Thus ' greenwood
is the condition of the fire being a sign of smoke.
Now this ' greenwood ' is pervasive of the proban-
dum 'smoke' without any qualification.) The
second is illustrated in the following :—' Air is
perceptible, because it is the substrate of percep-
tible touch.' Here the upadhi is ' generated colour '
which is pervasive of perceptibility in external
substances. Here ' being an external substance',
is a property of the subject of the inference, viz.,. ■
air. Now 'generated colour' (which is the,:...
upadhi) does not pervade the probandujn ^
(viz., perceptibility) simply without qualification.
In quality, etc., e.g., there is 'perceptibility/
inspite of the absence of 'generated colour.'
Hence we have to say 'perceptibility in sub-
stances.' (' Generated colour ' does not pery*dfcp|
•.perceptibility ' as such, but ' perceptibility itk^
external substances.' This excludes ' perceptibility
of quality, etc' which is not pervaded by 'generated^
colour.') But even this does not suffice. The
Atman or self, e.g., is a perceptible substance,
perceptibility of the Atman as substance is ^|g|
pervaded by generated colour. And
!?have to say, 'perceptibility in external,
l^wices.' (The Atman ,4f ; not an ^"J*?
bstance. It is an internal substance revealed

so we

?<t:

. * ~'*

*'/ ' v \

■ : s*t:

.■.-mm

94 MADHVA LOGIC •..#■

to internal perception.) The third form .^
upadhi is illustrated in the following:—'!
ehild in the mother's womb ought to be a dark-
coloured one, because it is the child of Maitri.|.
Here the upadhi is ' vegetable diet/ and it
pervades 'darkness of colour with respect to
child of Maim: In this case the tipaihi f \
* vegetable diet,' does not pervade the proband^
without 'qualification/ In the (unbaked) ja^
e.g., there is darkness of colour, but there is no
'vegetable diet/ Therefore we have to say ' the,
upadhi pervades darkness of colour as relating fe|
a child of MaitrV The fourth form is illustrate^
in the following :— ' The colour of the atom is
perceptible, because it is an object of knowledge
just as is the jar/ Here the upadhi is ' genera
colour/ and it pervades * perceptibility relating ifl|
external substances/ Now 'being an ext
substance ' is a neutral property in this case,
is not a property of the subject (' the colour *
the atom'). It is also not a property of the
ground (* knowability '). Further in this cast|
we cannot say that ' where perceptibility ;1S^
there also generated colour is/ since the con-
comitance fails in the case of e quality ' and tb^
like. Hence 'generated colour' is not pervasi*6;
of the probandum (i.e., perceptibility) simplj
without qualification. Hence we say
ceptibility relating to external substances 1 .
the meaning is ' where perceptibility relating

■■-■■

».•>■

V-*

prama^acandrik! 95

external substances is, there generated colour
also is,' and in this respect the upadhi, ' generated
colour/ pervades the probandum, ' perceptibility \
But this upadhi, 'generated colour/ does not
pervade the ground, ' knowability \ We cannot
say ' whatever is knowable, is also characterised
by generated colour/ the concomitance being seen
to fail in the case of the subject of the inference
(the colour of the atom). (The colour of the
atom is knowable, but it is non-generated colour.)
If we define upadhi merely as a condition that
does not pervade the ground (and leave out the
other part of the definition, i.e., that such condition
must also be pervasive of the probandum), then
in the inference ' sound is non-eternal, because
it is a product of will' the property of ' being a
jar ' will be. an upadhi, for in a sound which is
a product of will, there is absence of the property
of a jar. (But the above inference is free from an
upadhi, and therefore our mutilated definition of
upadhi is the cause of the illegitimate assumption
of an upadhi in this case.) Again if we define
upadhi simply as a condition that pervades the
probandum (omitting the other part of the defini-
tion that 'it must be non-pervasive of theground'),
then in the inference of - fire ' from * smoke/ ' the
nature of being a substance ' will be an upadhi.
(Fire is a substance and therefore pervaded
by 'the nature of being St substance'.) There-
tore (to avoid the absurdity of supposing an

^v*

MADHA LOGIC

«padfei where there is none) we say * the tfpadhi:^
must also be non-pervasive of the ground,;
(The nature of being a substance perv
* smoke ' also and thus is not non-pervasive
the ground.) The viruddhah heta, i.e., t
contradictory ground, is one that is pervaded
the negation of the probandum. E.g., the grou0j|
in the inference 'sound is eternal, because it is if
product of will/ is a contradictory or viruddha
ground. The ground in this inference is 'being
product of will/ and 'being a product of will'
pervaded by 'non-eternality' which is the negation
of 'eternality/ The anaikantikah hetu or gro
is one that is non-invariable (i.e., one which
not invariably related to the probandum).
anaikdntika ground is of three kinds, viz., the g
sadharana or common, the asadharana or
common and the anupasathhdri or inconclusive
Of these the sddhdranah anaikantikah is a gro
that exists also in that which is a negation of
probandum. (It is common to the proba
and its negation and therefore is called sad
anaikdntika or common non-invariable.)
ground in the inference 'the mountain is on
because it is knowable' is an example of this
of a non-invariable ground. The ground in
case is 'being a knowable* and this holds
also of the lake which is devoid of fire.
uncommon anaikdntika is a ground^ that
excluded from all co-ordinates and contra-ordi

Hal

?*>?.; A

pramI^tacandrika

to the subject and exists only in the subject.

?': For example, the ground in the inference 'the
element of earth is eternal, because it has odour *

f is a case of an uncommon non-invariable ground.
Here 'possession of odour' which is the ground or
hetu exists only in the subject of the inference,
viz., in the element of earth, and is excluded alike
from all other eternal and non-eternal objects.
The inconclusive anaikmtika is a ground which
'is bereft alike of instances showing its agreement
in presence with the probandnm. and instances .,
showing its agreement in absence therewith. For
example, the ground in the inference 'All that
m, is non-external, because it is knowable ' is a
-case of an inconclusive non-invariable ground.
Here the subject of the inference being 'all *l*gi

* is* there is nothing outside the subject to serve as

-an illustration of the invariable relation between

the ground and the probandnm. The kalMyaya-

t ; padista hetu or ground is one that seeks ^YjMg
a probandnm the negation of which is established
by valid evidence to be the property of the subject.

mblated

ground

following inference is a case in point :— Fire is
devoid of heat, because it is knowable/ Here
the probandnm is 'absence of heat but tfie
Rogation of this probandnm, viz., 'heat, is proved
by tactual perception to be the property of fire
which is the subject of the inference. The

v- -V'

plfccY.13

98 MADHVA LOGIC

satpratipaksa hetn or ground is one which has a
counter-ground opposed to it proving the negation 0|
of the probandum (in the subject of the inference). i$
For example, the ground of the inference 'sound
is eternal, because it is audible, just as is the
class-character or generic nature of sounds' is
countered and stopped from functioning by the
ground of the inference 'sound is non-eternal,
because it is an effect, just as is the jar/ A ;
countered or hindered ground is also called a , ■
prakaranasamafy hetu. t K

Just as we have the fallacious ground or . : >
pseudo-ground (in various forms) so also we have ^
the fallacies of the example or ud&harana, i.e., ..."
fallacious examples or pseudo-examples (udaharana-
bhasah). There are many different kinds of the
fallacious example. For example, with reference
to examples illustrating relations of agreement in
presence, we have first the case of a fallacious .-
example which is bereft of relation to the * ■•
probandum. Thus in the inference 'the mind is ■
non-eternal, because it has shape; whatever has ,
shape, is non-eternal; just as is the atom,' the ^
'atom* which is cited as illustrating the agree- ^
ment in presence is a pseudo-example in this sense.
For the atom i3 devoid of 'non-eternality' and so
is devoid of relation to the probandum. Secondly,
we may also have pseudo-examples that are bereft
of relation to the ground. In the same inference
if we say 'just as is action ' (in place of * just

:-,r

..»

it.

PRAMAtfACANDBlfcA 99

■ V

as is the atom') we shall have a pseudo-example

bereft of relation to the ground, for action, though

non-eternal, is yet devoid of * shape.' And

thirdly, we may also have fallacious examples

which are bereft of relation both to the ground

and the probandum, e.g., in the same inference,

if we say ' just as is ether ' (in place of ' just

as is the atom') we shall have a pseudo-example

devoid of relation to the ground as well as the

probandum. (Ether is devoid of shape and is

devoid of non-eternality.) In the case of examples

illustrating relations of agreement in absence, we

may also have pseudo-examples devoid of relation

to the absence of the probandum. For example, in

.the same inference if we say ' just as is action (by

agreement in absence).' (Action is non-etemal

and therefore is devoid of relation to the negation

or absence of non-eternality.) Secondly, we may

also have in such cases fallacious examples in the

form of examples bereft of relation to the absence of

the ground. For example in the same inference, if

we say ' what is not non-eternal, is not an object

with a shape, just as is the atom.' And

lastly, we may have also fallacious examples

; bereft of relation both to the absence of the

ground and the absence of the probandum. For

example, in the same inference, if we say

• 'just as is the jar.' (The jar is non-eternal

if and therefore bereft of relation to the negation

feflf non-eternality. The jar further has a shape

:>5 V

*-£»

:?*

:--

100 MlDHVA LOGIC

and is therefore bereft of relation to the negation
of shape.)

All these (i.e., the Nyaya fallacies so far set
forth) we reject as untenable. Why ? Because^
some of these so-called fallacies are not fallacies?
strictly speaking, while the rest are comprehended
in 'conflicting evidence * and the other fallacies weg|

have explained above.

The manner in which some of these so-called
fallacies may be proved to be logically flawless as : ^
also the manner in which the rest may be showa,1|
to be comprehended in our enumerations of the ^
fallacies are fully set forth in the PaddhaU ioj&
which the reader is referred for an understanding;!^
thereof. We refrain from going over the same ,
ground here for fear of prolixity. Thus every-||
thing (relating to inference) has been correctl||||
and intelligibly set forth. This closes the chapte£j|
on Anumana of the Pramanacandrika which
follows the track shown by the reverend feet-<$J)g
Sri Jayatlrtha. Let our heads bend in honour to|p
the sage Vedavyasa and let us close with an i*£ : l|
vocation of the name of Hari and the incantation
of Om.

We now proceed to discuss the nature of
Authority or Agama as a source of knowledge
Any verbal communication free from defects b
Sgama. The qualification *free from defects'
■ (in the above definition) distinguishes agama h
the mere (deceptive) appearance thereof,

W- : . : - ■ ■-■■■

PRAMA^ACAMDMKA 101

Hie words 'verbal communication' differentiate it

from perception and other sources of knowledge.
'&.'. What, then, are the defects of a verbal
^jporamunication? The defects of a verbal commu-
nication are :-(D unintelligibility, (2) conveying
of the opposite of the true or correct information,,
? (3) conveying of what is already known, (4) con-
veying of "useless information (for which nobody
cares), (5) conveying of information not derived,
:, or sought for by the person to whom it is con-
veyed, (6) conveying of a command or injunction,
to accomplish the impossible, (7) conveying of
advice of a more difficult means when easier means
I are well within reach, etc. Of these (1) umntelli-,

gibility is of two kinds, viz., (a) unintelligibly. ; ,
I due to want of significant words, and (6) un-r.^

intelligibility due to want of intelligible relation
: ; : (between the words of a verbal communication)..
Examples of the former are : -'Because ka^a-ta- .,..
ta-pc'B are ja-ba-ga-da-ha' (of. abracadabra).
Examples of the latter are :-'The cow M^ga
P f horse,"Manisan elephant,' 'The bowl.rr%g|
'i cumin seed,' ' The ten apples are five cakes. U)
Conveying the opposite of what is true is illus-
trated in the following : -'The world is unrea^|
I: ' The Sudras have the right of access to the,
IfeVedas,' 'The Brahmins have no such right, etc.
' £&. Examples of conveying what is already k«^||J
«e :-« The sun rises in the east, and sets gJSg?
-—*!-. < u.i AMUH .» » WR et * 'The Nimbajr^rt :»/.■■ -> s

w-^

■a

102 MADHVA LOGIC

bitter,' etc. The objection that 'conveying thfc
known cannot be a defect as it adds to the forced
of the evidence that is already available (and thus
strengthens our knowledge), misses the real point
at issue. Later evidence strengthens our know-
ledge only where earlier evidence has failed to
remove uncertainty and produce complete certitude
in regard to that about which we were in doubt.
(Hence where no uncertainty exists, the attempt at
further enlightenment by means of verbal commu-
nication is waste of energy.) (4) Examples of
stupid, pointless communications (for which
nobody cares) are :•— ' How many teeth has the
crow ? ' ' What is the weight of the sheep's
egg ? * ' How many threads of hair are there ,';|
in the blanket?' ' What is the news of the province
of Cola? ' (Gola being the name of the place where - : S
the questioner himself lives), etc. (5) Examples p
of communications which are of no use to the*.;/
persons to whom they are conveyed are : — c Advice v;I
of business and trade to one who has subdued
the desires of the world,' etc. (6) Examples of
communications enjoining the accomplishment of
the impossible are : — ' When alluding to a person
who is dead and gone, one proceeds to describe an
elixir that will bring the dead back to life and
that may be found in the north of a certain hiHl
called Mrtiharamahtdhara,* etc. (7) Examp
advising more difficult and less accessible remedi
where easier ones are at hand are : — * To ask

'£

PRAMA^ACANDRIKA

103

man to cut down something with an axe which
he can easily remove by his finger-nails,' 'To
advise a thirsty man on the Ganges banks to sink
a well for quelling his thirst,' etc.

A verbal communication is made up of words
and sentences. A word is a combination of
letters with a case-ending. The words 'with
a case-ending ' (in the above definition) exclude
non-sense combinations of letters like ja-ba-
ga-da-ia (which are devoid of case-endings). A
sentence is a combination of words characterised
by mutual expectation, suitability of relation, and
proximity (in time) . Expectation is that relation

" of one word of a sentence to another (or others)
of the same sentence without which the relation
expressed by the sentence will not be apprehended,
e.g., in the sentence ' Bring the jar,' the verb
('bring') without the object ('jar') will not
produce the apprehension of the relation of a
verb to its object, and so the verb ' bring' has the
relation of expectation to the object ' jar ' . Or we
may say, expectation is the fulfilling (on the part
of the subsequest word) of the intent or expecta-
tion which is generated by the word preceding it

. (in a sentence). Therefore, 'The cow is a
horse,' ' Man is an elephant,' etc, are not
sentences, for in these the words do not expect

: ; i.e., enter into relation to, one another. Though

/^'expectation' is, strictly speaking a property of
consciousness, yet objects .(denoted by words),

104 " MADHA LOGIC

as producing in the bearers of their respecti
names an expectation of these objects in mut
relation, are also said to * expect * one ano
And as words (the names of objects) denote su
objects (with mutual expectation), words also a
said to have ' expectation.'

By suitability or fitness of relation (bet

the words of a sentence) is meant the absence iif^
the cognised relation (between the said words) of?;
any clash with valid evidence. Thus in the words
'moistening by water/ the relation of effect and!
cause which is asserted between the act
* moistening ' and the agency of ' water ' remaiM
uncontradicted by valid evidence. Here the coBff^
patibility of the relation between 'moistening'
and the agency of * water * constitutes the s
lity or fitness of the relation between them. F
this reason the words, 'He is moistening
means of fire,' do not constitute a real sen
there being no suitability of relation in this
Between 'fire' and the act of 'moistening
there is no question of a mutual agreement^
compatibility of relation.

By proximity (in time) is meant the express- 1|
ing of the words (in a sentence) without -aBj;|j
long pause or interval of time between the
ferent words. Thus the words, '■ Bring the
uttered without any long break or interval'
time between them, have this character
proximity (in time) . For this reason* -tfeR

PRAMA^ACANDRIKA 105

■ ' Bring the cow/ uttered separately, each after an
hour or thereabout after the previous one, will
not constitute a sentence, for they lack the
character of sufficient (temporal) proximity (to
produce a unity of meaning).

In verbal communication as a source of
knowledge, the sentence is the instrumental
cause, the recollection of the meanings of
the constituent words, the intervening process,
and the knowledge of the meaning of the
sentence, the result (of the process).
Verbal communication also (like inference)
conveys knowledge only of such correctly Com-
municated objects as are accompanied by the
knowledge of the meanings of the constituent
words of the communication. In this respect it
is unlike perception (as an instrument of know-
ledge) which makes things known by its
bare existence (without being itself known
or apprehended), for an authoritative commu-
nication like inference depends on a known
instrumental cause (i.e., the knowledge of the
meanings of its constituent words). Otherwise
the absurdity will follow that an authoritative
v knowledge will have to be admitted where a
verbal declaration (of a truth), though existing
k in itself, has not been actually heard by a parti-
:{ : cular person, or, even though heard by him,
5 has not been understood because of lack of know-
ledge of the meanings of the constituent words.

§•'• 14.

106

MADHVA LOGIC

(We now proceed to discuss how words mi
tteir respective objects.) Some hold that woi
like 'the cow,' etc., mean only class-characters,
these being the first to be presented as attribul
(as soon as the words are heard). The individuals
are reached (mediately) through these class-
characters which drag them behind themselves.
Others hold that words mean the individuals
as specified by their corresponding class-charao^
ters. According to this view, a class-name is
subject to the conditions which regulate the use

^ok words. Thus all words such as ' ether/ etc.^
mean specified individuals, these alone being thftf
determining conditions of the use of words.
Otters' hold that words like £ the jar,* etc., meatt
class-characters; proper names like * Devadatta,'
etc., mean individuals ; words like ' possession

/. the "dewlap* mean shape or make; while words
like * the cow,' etc., mean all the three. The real

> fac£ however, is that the meaning of a wordil

' jusf "{fiat object which is immediately presented^
to consciousness as soon as the word is heara-i
Tn the apprehension of the meaning of a w<
similarity acts as connecting or mediating link-
Thus since the word 'jar' calls forth the i<k
both of c individual jars * and ' the , class-characte||J
of jars,' this word must be supposed to be cap)
of meaning both the ( class-character T and ' thi
individuals.' Again, since the word ' white ' *
forth the idea of the quality of white <$$%

^ * l n

-♦.-V'.-ry-:

i~ , :**

prama^taoandmka , '*■ ifft c

P well as the substrate of the quality (the white
pithing), it must be supposed to mean both ' the
quality ' and ' the substrate.' Similarly, the word
H 'gone' means both the action of ' having gone '
W and ' the agent of the action,' the word 'stick-.
# in-hand' means both the 'stick' and 'the
^person, Devadatta, who holds it in his hand,' etc.
i. Or, we may say, words like 'cow' etc., being
5 nouns or substantives, must denote individuals,
while words like ' bring,' etc., should mean attri-
butes or adjectives. .In combinations of words
| such as 'bring the cow,' the act of bringing
being made possible through an individual
'■: agent of the act, the meaning should be conceived ■'■.].
v.; as consisting in the individual (in the individual -i j ^

f agent of the act) . '"'"' , '■';■/%

T'- (The question now has to be discussed, how 3
SPwe acquire a knowledge of the meanings of
p words.) Our view is that we learn the meanings
foof words from the signs made with the fingers
T" (by our elders while uttering the words). Thus
!; the child sitting on the person of its father or
l-mother begins to learn the meanings of wmgg
I- when the said father or mother tries to rouse '$g|

from a state of inattention and to draw its atten-
tion towards himself or her^a^J «*gp
i words which he or she may uttef^ by' mabn|j|=

signs with the fingers, or by producing *
feoft sound by Striking one finger-tip against *

"another- In this waj^be «hild is taught the v

ri=v

■ 'w

■

♦.*

108 MADHVA LOGIC

meanings of such sentences as 'Child, that
is your mother/ 'that, your father, 1 ' that*
your brother,' 'the man is eating the plantain
fruit,' etc. Thus by the said signs the child
gradually learns that the said words are related
in a general way to the said objects pointed^;
out by these gestures and that the relational
is the relation of meaning or signifying these
objects. Later on when such words as ' This is

your sister/ 'That is your friend/ 'He is
eating a cake/ etc., are uttered in his presence,
and he begins to note the different contexts in. is

.-.■83

which these different words are uttered, he
learns, through the differences of the contexts

"•'MS

to distinguish the specific meanings that attach ||
to particular words, e.g., that the word 'mother'
attaches to 'the female parent,' etc. Otherof"
however opine that the meanings of words are<s
learnt from the behaviour of the seniors or elders.
Thus the inquisitive child, when he hears a senior
say to a junior, ' Bring the cow/ and notes that?
immediately afterwords the junior is prompted^
to the act of bringing the cow, concludes t$
agreement and difference that the action of the
junior is prompted by the knowledge produce^
by the words of the senior. And so assuring!
himself, when he hears other sentences spo
in other contexts such as 'Bring the hor
'Secure the cow with a rope/ etc., he ga
from the divergent contexts that the

<<#£*&

PRAMAJjTACANDRlKA i 109

' cow ' means one particular kind of animal, the
word ' horse,' animal of a different kind, etc. But
this view is not tenable. The quickly-forgetful
child cannot possibly retain the consciousness of
the word he hears till the actual bringing of the

object desired.

A word has two kinds of function or vrtti, viz.,
(1) primary (mukhya) and (2) non-primary or
secondary (amukhya). The direct or primary
function of a word is its iakti or power of
referring to, or meaning, a particular object.
Sakti is defined as that relation between a word
and an object which is conducive to the recollec-
tion of the object (as soon as the word is heard) .
Samaya, sangati, sanketa, vacaka, etc., are used
as synonyms of sakii. This iakti is of three
kinds, viz., yogah, rtMmA yogarudhih. Of
these, the power to refer to an object by virtue
of the powers of the constituent parts of the
. word is yogah. The power of meaning which
- belongs to a word as a whole irrespective of the
powers of its constituent parts is ru4hih. Lastly,
the power of meaning which is derived from
both (i.e., both from the word as a whole and
the meanings of the constituent parts), is yoga-
ruMh. Of these, some words mean their
respective objects through the meanings of its
H. "constituent parts only, such as the words pdthaka
■'< (reader), pacaka (cook), dan& (the man with a
I a stick in hand), kun4all {the coiling thing), etc.

■jM

110

MADHVA LOGIC

■ ■?::

Some words, again, mean their respective object*
by the powers inherent in the words as a whol
irrespective of the powers of the constituents,
Such are words like ghata (jar), pata (piece
cloth), etc. Lastly, some words mean objec
through their powers as a whole as reinforced by|
the powers of the constituents such as the word|
pankaja (the lotus), etc. In this manner att'g
other words with primary meanings, such as thef
mahayogafr, etc., should be understood (as signi-
fying their objects).

The non-primary or secondary function of a^
word is called Laksana or Implication. Implica-
tion is a kind of relation to the object of a word's!

&akt% or power of meaning. There are two kin
of Implication, viz., (1) Implication which is in*
dependent of relation to the direct meani
(Jahallaksana) , and (2) Implication in whioK|
the direct meaning also enters as a facte
(ajahallakfana). * The milkman lives in
Ganges ' illustrates the former. (Here ' the!
Ganges ' means not the river, but the banks
the river Ganges.) ' Men with umbrellas
going* illustrates the latter. (Here chatriti
i.e. 9 'men with umbrellas,' means * pedestrian®.
According to another classification, Implicat
is of the following two kinds, viz., (1) Impli
tion in which the implied meaning is ij
dependent of any special end or purpose to
subserved, and (2) Implication depending <&

-■-.■.

■I s

L#l :

.*■■■ *.

prama^tacandrikI 111

some purpose that is subserved. "' Travellers are

going' is an example of the former. (Here

I'm&rgaht literally 'roads,' means 'travellers'

by implication.) Here the traversing of the

• roads by travellers being observed to happen

without any special end or purpose to be served,

* ,: such implication is also called mjhalakma.

'The milkman lives in the Ganges' is an example

of the latter. (Here ' the Ganges ' means ' the

banks of the Ganges.') In this case ' living near

the Ganges' being prompted by considerations

of sanctity and the like, the implication is called

kevalalak?ana. The inapplicability or failure of

the primary meaning is the real cause of an

implication. In a similar way should be con-

f ceived other non-primary functions of words such

"/'as the gaunt (the deferred), etc. ,';.,;

" A~gama, i.e., an authoritative verbal communi-

: cation is of two kinds, viz., (1) communication

having a personal source, and (2) communication

IS devoid of a personal source. The Rg-veda and

:> other ; orthodox scriptures are the impersonal

k&gamas, or Agamas without a personal source.

^ Valid personal communications are those recorded

>in"'the Mahabharata and other sacred works.

> v Thus, it is said, " Scriptures are of two kinds,;

|V viz., (1) the eternal, and (2) the non-eternal.

Such, for example, are the Vedas beginning with

Rg-vefa the MahSbharata, the Paflcaratra, the

original Ramayana, and. the Puxanas. AH these

.Vi

i>3

,«

112 MADHVA LOGIC

as also all those that follow in • the wake of tliesii
are to be regarded as the (authoritative) sacre^l
scriptures. Those that are other than these. :
Janatdana, they are heretical and should not
be regarded as authoritative scriptures, etc." ; .1

But it may be said : * The Vedas have 1^
personal source, because they consist of collections
of sentences, just as are the verses of Kalidasa r
and the rest.' Our reply is, this conclusion does v
not follow, the inference in question being
vitiated by the presence of an extraneous condK :■■
tion. The extraneous condition is ' a personal
origin established by tradition/ In other words,
€ a collection of sentences ' (which is the ground
of the inference) is in itself no proof of a personal
origin. It is a sign thereof only on the condi-
tion that such personal origin is established by
tradition.

But it may be said : ' The Vedas are devoid of
evidential value and validity, because despite the
due accomplishment of the Vedic prescriptions
the promised fruits are not realised, just as
are the hopes generated by the utterances of;^|
deceitful people.' Our reply is, this is not the
case, because the above inference contradicts the
following valid reasoning: — 'The Vedas are
authoritative, because they consist of imperso
prescriptions, i.e., of sentences without a perso:
origin, just as are the lunatic's ravings by a
ment in absence.' (The lunatic's ravings

-■

PRAMA^ACANDRIKA 113

non-authoritative, and they are not without a
personal source. And thus they illustrate the
agreement in absence ' what is not authorita-
tive is also not without a personal source.')
It cannot be said that this inference entails
the fallacy of an asiddha hetu or unestablished
ground. For Sruti (authoritative scripture)
declares that the sentences of the Vedas
(i.e., the Vedic prescriptions) are eternal verities.
(Thus the eternity, i.e., lack of a personal
origin, in respect of the Vedic sentences,
is not asiddha or unestablished.) Smjii (i.e.,
the secondary scriptures derived from the
primary scriptures) also declares that the Vedic
sentences are without beginning and without
end, are eternal or timeless, have inherent
authority and are self-existent. Nor is non-
fruitfulness or non-efficacy of Vedic prescriptions
a proof of invalidity (of the Vedas), for the pres-
criptions being seen to bear the promised fruits
only in the properly qualified agents accomplish-
ing them, the non-perception of the fruits in
other cases must be attributed to the inherent
disqualifications of the agents. .

But it may be said : * Agama is not valid
u., evidence, bacause it does not prove anything, just
I as a deceitful utterance.' Thus the visaya or
I;; object to be proved by valid evidence is of two
fjdn&s, viz., (1) the immediate, and (2) the
emote nr mediate. Of these, the immediate is

remote or mediate, ui *«

m; . 15

:*-„■■

1X4

MADHVA LOGIC

f?ft

the object evidenced by perception. There is no
other object, besides these two, which can servel
as an object to be established by the evidence rf|
Kgama or authoritative communication. If you
say that Agama causes the knowledge of the
mediate, then as, in your view, nothing is, valid ":||
independent evidence which causes the knowledge
of the already known, and Agama has application
only to objects proved by other forms of evidence,
so all Kqamas will be shorn of evidential value aft|!|

""2*

wanting in any distinctive objects to be proved^
thereby. All this, we reply, is wrong. Just as
the evidential value of visual perception in regard
to objects distinct from the objects of hearing
cannot be denied in spite of the fact t
visual perception proves only immediate obje
. just as hearing does, so also the evidence
Kgama in regard to objects distinct from t
objects of inference, e.g., in regard to s
special objects as heavenly happiness, libera
etc. (which are not objects of inference), ; ii
unimpeachable despite the fact that Kgama ■
indistinguishable from inference in the matt~*
of the mediateness of the objects it proves. If|
cannot be said that * heavenly happiness,
'liberation/ etc., are objects of inferential pr^® 8
and so Agama is devoid of any special objects' "ttf
be established by its evidence. For if you s|§
this, we shall say contrary wise that Kg

&2fiM

being evidence of mediate : objects (sue

4v. PRAMA^ACANDRIKi .115

'heavenly happiness,' etc.) is a valid source
of knowledge, while inference, being devoid of
proper special objects (i.e., objects not charac-
terised by the mediateness that characterises the
objects proved by Igama) is devoid of evidential

authority.

The Vawesikas hold that Igama is included
in (i.e., is a variety of) inference. But this
view is wrong, for even in the absence of the
recollection of invariable concomitance and the
like, there is realisation of the meaning of a
verbal communication, this being a matter of
common experience.

Bhaskara and his followers hold that Agama
without a personal source is independent evidence,
but Agama having a personal origin is a variety .g-
of inference. This also is a wrong view. For the
sum of conditions for the comprehension of $*^
import of sentences, viz., expectancy, suitability
of mutual relation between the constituent words,
etc., being identical in both (i.e., both Personal
and Impersonal Igama), there is no valid ground^
for assuming any special character attaching to J g|

only one of these. ... _■....'-■■- '•'.■■ •■ •• .•'• : sS
Thus have we described all the three forms

of evidence or Pramana. Anything other than
these (three) is not a'prdmSm or valid source- /<4jj|

: knowledge. - •■-" ■■ ••-:-' '-'' .''*j& : ?i'-

But it may be said: there is another form
rf avid*™* distinct from the above three, tnz.,

01 evidence tnswuo-

■■■ m

116 mIdhVa logic

Arthapatti or Presumption. When a thing of%
event is seen to be inexplicable except on the &
assumption of something else and when on the ,
basis of our observation of this otherwise inexpH*yf|
cable thing we are led to presume that which is T|
necessary to account for it, we proceed according $
to the method of Arthapatti or Presumption as
a source of knowledge. Thus when we learn v
by perception or reliable testimony that Caitra j>
is alive and yet is not in the house, we at once ..-|
presume that he must be somewhere outside, for ;■;
absence inside of one who is alive is not ;■%
explicable except on the assumption of his i
existence outside. Hence the proof here of outside
existence consists in the Arthapatti or Presume ^

tion which is created therefor by the otherwise :|
inexplicable fact of inside non-existence of one ^
who is alive. This process is distinct from the -^
processes of perception and the rest, for outside
existence is not an object of perception and the

-test. • " ; :

This, we hold, is not the case, for the «v?|
called presumption is only an inference '&&$$
disguise). (The inference is as follows: %j.
* Caitra must exist outside, because, though alive,
he is absent inside ; whoever is alive, and
does not exist in a certain place, must exist in

— m +

some other place, just as I myself do.
inference being quite competent to produce
knowledge of the outside existence, whati#|

Til

«4 ■ I

" 41 "

cVl

PRAM A?A C AKDBIK A

11?

use of assuming a separate source of knowledge
such as Presumption? In this inference the
words 'though living' preclude the case of the
dead (who are non-existent because not alive).
Similarly, the substantive (qualified by the
adjective ' though alive '), i.e., 'absence inside,'
precludes the case of Devadatta and the like
who are existent within the house.

But it may be said : there is another separate
source of knowledge, viz. , comparison. Compari-
son ( Upamana) is the cognition of an object as
characterised by likeness to another, e.g., of likeness
to the cow as produced by the recollection of a
comparative statement as an auxiliary condition.
Thus a person ignorant of the meaning of the
word ' gavaya ' first learns from a forester that
'gavaya* means ' an animal resembling a cow.'
Thereafter when later on he comes across in a
jforest an animal looking like a cow, he recollects
the previous advice of the forester that ' gavaya
is an animal like a cow. ' Thereon the knowledge
dawns on him * That animal (resembling the cow)
must be what is meant by the word gavaya.'
As this knowledge is not caused by perception
and the rest, the process (which generates the
' knowledge) is regarded as an independent source
. of knowledge called Upamana or comparison.

This, we hold, is not the case, for (the so-
;• called) comparison is really comprised in m-
i^ Terence. The inference in such cases is as

118 MADHVA LOGIC

follows:— ' The subject of the enquiry is t&|
meaning of the word gavaya, because, not
being a cow, it possesses resemblance to a cow|
just as is the jar by agreement in absence.*
. (What is not the meaning of gavaya, is also noil
that which, not being a cow, bears resemblance

* to the cow, just as is the jar which is not the;||
meaning of gavaya and is also not that which,||
not being a cow, bears resemblance to tbe||
cow.) To preclude illegitimate application to a v|
second cow resembling a first cow, the words j

',.* *not being a cow' have been included, and to :
preclude illegitimate application to the jar an<|
the rest, the substantive 'that which bears
resemblance to the cow ' has been included.

But it may be said: there is yet anoth*
separate source of knowledge, viz., negation^
Abhava. This must be admitted in order
account for the cognition of negation. ^ln^
non-apprehension (i.e., negation or absence of
apprehension) of the jar and the rest assures;
us of the absence or negation of the jar, etc.;
* This non-apprehension is just the negation or
absence of apprehension. The non-apprehensioiiu
being apprehended or realised in consciousness*
the negation, i.e., the absence, of the jar, etc.,
is also cognised or apprehended.

This is not the case, we say; for this
called negation as a source of knowledge is
reality comprehended in one or other <>t.

.--:■

i **■*■■ to

»X t

ffe% ■-" PRAMAKACANDRIKA ..' ''''' 119

■y t ' '

v ■

Pramanas we have explained above. Thus,
according to our view, that which causes the
knowledge of negation (in a particular case) is
the Pramana or evidence of the negation in that
case. For example, the evidence in regard to the
present non-existence of the Kauravas is the
testimony of the Mahabharata. In the case of
Devadatta's lack of vision we have the following
inference as proof or evidence of the lack or
negation :— The subject of controversy (Deva-
datta) is devoid of vision, because he is ignorant
of the nature of colours.' Similarly, the absence
of pleasure, etc., is evidenced by the immediate
intuition of the witnessing intelligence. The
realisation of the non-existence of a jar before
oneself results from a quickly-produced percep-
tion. It is not an effect of non-apprehension
only, for it has the nature of an immediate
positive experience. No doubt, a proximate
non-apprehension is also an indispensable
condition. But the mere fact of a proximate
non-apprehension as an indispensable condition
does not constitute the latter an independent
source of knowledge, for in that case by a similar
line of reasoning one may say that in the cogni-
tion of positive reality the non-apprehension of
its negation is the real evidence or proof.
? Where in the midst of darkness we cognise the
absence of a jar and the like by means of ex-
ploring with the hands, the non-apprehension

120 MADHVA LOGIC

qua non-apprehension is not the real cause of*
our knowledge of the negation or absence. Thejl
non-apprehension is a cause here only as itia 1
treated as a sign or mark serving as the ground
of an inference. The inference in such cases is *:
as follows: — 'The jar does not exist here, for, f
though fit to be perceived, it is not actually ■
observed here, just as an elephant.' But it may
be said : a negation or absence being admittedly v
incapable of positively stimulating the sensibili- ■]
ties, we cannot sensibly talk of negation being
perceived by the -senses. Our reply is : this is •.,
not the case as there is no bar to a negation ;.:
being in contact with the sensibilites just as there j
is none with respect to a positive entity. : .^ \

But it may be said : we have mathematical -|
or quantitative reasonings (sambhava) and these ; !
should constitute a separate source of knowledge, i
Thus when the cognition of the greater leads to
the cognition of the less we have sambhava or
quantitative reasoning. , For example, the
knowledge that there is one hundred yields or
establishes the knowledge that there is fifty-five.

To this our reply is : quantitative reasoning is
only a variety of inference. The inference here
is as follows : ' Devadatta must own fifty-five, ^
because he owns one-hundred, just as I myself

do/ .' "; y._;:\,

There is another kind of knowing, viz., knoWr,J
ing by the method of exhaustion. Thus whe^l

f'Zl

w +

43, 7

PRAMAtfACANDBIKA ,151

by eliminating other possible alternatives one
after another we arrive at the last or remaining
alternative, our knowledge of the truth of this
last alternative is gained by the method of exhaus-
tion. It is of two kinds, viz., (1) that which
proceeds by the method of affirmation, and (2)
that which proceeds by the method of negation
or exclusion. The former is illustrated in the
following : — When we know that the two persons
before us are Gaitra and Maitra, then the know-
ledge, ' this one of the two persons is Caitra,'
entails the knowledge, ' the other one is Maitra.'
The second is illustrated in the following :«*r
When we know that the two persons before us
are Caitra and Maitra, then the knowledge, ' this
one (of the two) is not Caitra ' entails the know-
ledge, ' this, then, must be Maitra.' ■_-_ . - j; -
This also, we hold, is a case of inferential
knowledge. The inference is as follows : The
disputed subject is Maitra, because, being either
Gaitra or Maitra, he is not Caitra (in fact), just
as is Gaitra (by agreement in absence). (In
other words he who is not Maitra, is also not he
who, being either Caitra or Maitra, is not Caitra,
just as is Caitra, who is not Maitra and is also
not he who, being either Caitra or Maitra, is not
Caitra) Here to preclude the illegitimate exten-
sion of the hetu to the cases of the jar, etc., the
h^ords ' being either Caitra or Maitra ' have been
included. _

-5- ■ v ** .-*■#>£ Ji-i •£■ *t- ; r^y* ^ - " " v ' *"

122 MADHVA LOGIC

Upakrama, etc., are also forms of inference?!
for they produce an inferential knowledge of the
purport of sentences. An unbroken continuum of
tradition without any known originator thereof is ^
Aitihya or Tradition- For example, the hearsay, g
1 In this fig-tree lives a demon' (is a case of aitihya). r|
This is nothing but verbal testimony (and should not
count as an independent source of knowledge). "So f t
also all the rest such as omens and signs (Sakuna) 9
written language (Upi)> gesture language (cesta) . ■:
—ail these as sources of knowledge are included
in the forms of evidence (i.e., the three we have '_*
explained) we have set forth above. We shall
now explain the nature of validity itself . The
intrinsic nature of validity is of two kinds? t>t0 M

(1) intrinsicality in respect of origin (utpattt),
and (2) intrinsicality in respect of verification ^
in consciousness (jiiapti). Of these, intrinsi
cality in respect of origin means that the
validity arises from the same conditions as
cause the cognition itself of which the validity
is a logical character. In other words, validity
being an intrinsic or essential character of -a
cognition in the matter of its origin means that
the causes which produce a cognition are also
the causes which produce the validity of the cogni-
tion. By intrinsicality in respect of verification
in knowledge is meant that the validity is apprefj
hended by the same agency a& is. the cbgnitic
itself of which it is a logical character, • Itt-^

••-.

■::■■;

* t

PRAMAtfACANDRlKA 128

words, validity being an intrinsic or inherent
character of a cognition in respect of its being
known as such means that that which apprehends
the cognition also apprehends the validity of the
cognition.

... The extrinsicality of invalidity is likewise
of two kinds, viz., (1) extrinsicality in respect
of the origin of the invalidity, and (2) ex-
trinsicality in respect of its confirmation in
consciousness as such. Of these, extrinsicality
or adventitiousness in respect of origin means
that the invalidity arises from conditions other
than those which cause the cognition itself*
Again, extrinsicality in respect of the invalidity
being known as such means that the invalidity
is cognised by an agency other than the agency
which cognises or apprehends the cognition itself.
Thus everything has been set forth in its proper
place.

- The Sankhya philosophers however hold that
both validity and invalidity are intrinsic alike
in respect of their origin and their being known
as such. The Naiyayikas, on the contrary,
hold that both are extrinsic. And they say
in this connection that a cognition itself is
caused by the sense-organs, etc., while the
validity of the cognition is caused by the presence
of certain specially efficacious qualities in the
il causes of a cognition. Similarly, the
didity of the cognition is due to the presence

1 ■ r" *~\> •+.

124 "MADHVA LOGIC *|,

V, * ■

mm -^

* '*»

of certain special defects or deficiencies in th#i
causes of a cognition. And so also (in tito£
matter of being known) a cognition itself i»|
cognised immediately by internal perception by tb#f|
mind. But the validity of a cognition is
cognised mediately by inference from the mark or j|
sign of successful or unsuccessful practical re^;^
action. Thus the general condition (according to ]■■
NySya) of valid cognition is an efficacious quality
(in the cause), and the general condition dfjjj
invalid cognition is a certain defect or deficiency
(in the cause). The Buddhists, again, say^T
invalidity is intrinsic while validity is extrinsic^
But the real fact is, both the cognition and its]
validity are caused by the sense-organs and the -
rest. But the invalidity is generated by /&^
sense-organs, etc., together with certain defect^
as auxiliary conditions. Similarly, both theg;
cognition and its validity are cognised immediate
ly by the witnessing intelligence. But
invalid cognition is cognised only as cogn
by the witnessing intelligence. The invalidity

of it is known mediately by inference from
mark of practical unfruitfulness or failure.

But it may be said : the position taken by
the author here is not tenable. The author's v
is : the agency which cognises the cognition a
cognises the validity thereof. Further the po
of the instruments of cognition to produce
Cognition becomes the power to produce

jPRAMi^ACANDRIKA l2§

invalidity when aided by the special causes of
invalidity. The power thus becomes as it were
a different power through the influence of the
defects in the matter of producing invalidity.
Moreover, the cognition of these instruments
takes place through a different agency, for the
sense-organs, etc., are themselves cognised each
by its own pramana or suitable knowing process
(e.g., the sense-organs are cognised by inference
from the results which their actions produce). At
the same time their conduciveness to the production
of valid knowledge is also cognised by inference
(i.e., by another inference from the mark of
practical f ruitf ulness) . But why should this be
so? (i.e., Why should you suppose that the
agency which cognises a cognition also cognises
its validity, but the fitness of the sensibilities to
'? produce a valid cognition is not cognised by the
, : same agency, i.e., the inference that cognises the
sensibilities themselves?) What is the bar to the
supposition that the validity is cognised by an
agency other than that which cognises the cognition
(just as the fitness of the sensibilities is cognised
t by an inference different from that which cognises
the sensibilities themselves) ? ' .

^ Our reply to this is : this cannot be admitted,

I ; .because any such admission will entail an infinite
regress. Thus we say : the validity of a valid
I cognition must itself be cognised as such by
I? „„™„ «.im virion, otherwise there will be no

126 _ w MADHVA LOGIC

valid cognition anywhere. And such validity
cannot possibly be cognised by any agency other
than that which cognises the cognition itself..
In the event of any other agency apprehending
the validity, the validity of this apprehension will
itself have to be apprehended by another (i.e., a
third) apprehension, thus leading to an
infinite regress. And if to avoid the infinite
regress we say that the second cognition
apprehending the validity of the primary cogni-
tion is self-validating or self-evident, then sq
may also be the primary cognition. There-?
fore by the method of elimination of other;;-
possible alternatives, the intrinsicality of validity
is established as the only position that remains
unshaken. It cannot be said that the objection
of an infinite regress holds equally in respect ':
of our view that the validity is imme*
diately cognised by the witnessing Intelli- ;
gence, for the witnessing Intelligence (in our
view) is- self-revealing and as such reveals ■
both itself and its validity. But why not
assume the same with regard to cognition
also? Because, we reply, cognition being
a state or function of the internal organ is non-
intelligent and as such is incapable of self-
illumination or self -revelation. The validity of
the instruments should be understood on the
lines of the Paddhati. Thus is everything beauti-i
fully explained.

■** Ui

y\r:

PRAllAtfACANDBIKA

This brings the chapter on Iga
Pramanacandrika by Srimacchalarifefi

close.

May the sage Sri Vedavyasa be pleas

M^I*lNfe*l

ire*r. ^rlr^:

frft urannk 1*1 «**» *#^ w J*""

130

MADHVA LOGIC

4IT4K3 T?W I ^flt 'TTOfa I irt^T'WTt^src : I rfSTfo
«!el<tfd I BW «WUd*II ^lRpffet 4ld)«B& I WWt

^rn^ff^ *fa ^4*ddiMi ^JufdUKctiifti*;; *3*

3T3W I H'*WdM'nJ<chlfd^l?BWiyK^y^iqMi.l3W'

rftrr^r^«^«^i

r: i mi

e«

*T- «

PRAMA^ACANDRIKA 131

m^fai * foc&rd « mr« : I a«ift?iTra stow* fa «
fifrramfri wraroi^i ^a i a* *fffS*9 ****

srafcs^ ^P^T^'wsnfir. i «* www

-V * "

132

MADHVA LOGIC

q*arq" W§ STW3TW Srffalfi I *S*«jTH*lfq W3T*l \

^wm^i^n f^ctft5H«e(i^ 1 *rt «sra *^# *ni
qz *t> vrsis^rrfH*. 1 < nft#Bi^nn9»n#i^ im 1

ni'wg* <NHHy*ft gjqz^WT^TT *?3TT> **I?T-

ar^ara yg: fsjapn, sir $^fadi i n^Hfcfdmirn : 1 *ra1

isr^juj 1 ?^ ^to^ MqqiTHTqT-resiaT: ^jiw-
^T 1 i* tok^n&mi: #jrat tot 1 sn*-

--"■*!

PRAMAtfACANDRIKi 133

xpjqt wst fMfafsreTOrat wugwfaflig* *i»*U<ff?«f
*toAnjM*j«^ fam wrw- jprat ^ » ^rerarr^-
Wfrawn'wi fdm!ij*^aiMd: «'swt *refo firatf

TO Ut ^ ww ftft » ^^rf^w: ^>rat ami

*rfafl «w2»Tt *?n*fa wh«1«iu<iw»wwH set i
*»mt gznft*i ^5n«T^nqHH^<r*<rd i *pjq*w:

134 MADHVA LOGIC

<t.qfafd 1 CTflBV M*H*\y\^i>Tf\oH\VH: I *?«$<< WT3T-

wroniirert *rai i ^^t *m ^ifa =^n^tf?r irans-
g# *mra: i xstsrri«mrH$% I ?nr, ^fifsarre: i

"■Jfl

* For the use of the word vfflpf in masculine, compare—

<anr*ftsfa sr j'finrt ««*m«<: s^ffir a

SarasvatiA:antJia6fearano, eh., 61. 70.

PRAMA^ACANDRIKA

135

* ►

Wt Jl^l^ t^ H«ltrt ff^T

to. nHT^p^TOt^Jtra?T \fh %j, i i ?rerw«3sfa

ijfoM m$l warrtmifkfow*: I to «'w-
toto ftfro t^wgroro %fa i to *reT*i-

I wt tm i ^faft » to>kto ifcHfof-

: i

s*wiR{«wir<i"«ifH;

/♦ *

136 MADHVA LOGIC

tow? ^rata^rawf? fa<* ^ « tfonmrw-

aftnn* g Jprrcatsril^ i **mfa f^wlf <icn«*i i
sfa tundra ^t^wssrcrafaiw wt mfe*-

-■■

PRAMA^ACANDMKi 137

trefwitf^fti wsst^i ynt^ T^n^ -
W&wnfc <j**u Hih^wifrwaftiig-

^ * r

<#'

'^"^^^"^^r

f*^" -' -

■^sii

.vijrf^Be

138

MADHVA LOGIC

praMasacandrika 139

?nrafc ?^T^t wrssmfon iw

*3*ffignT «n«iMt ^ft?ro ftwn i ■«■*
**« «M iW ^i**"™ *** **<**;

140

MADHVA LOGIC)

/- >S

*i, * '^

PRAMAtfACANDRIKA 141

*pi tut *w**rat: wrararcj. i rm ^w«m«ww

cnra«ii «h«m<wW * < " l " ,ll "- Jll4J '...! 1 ?"
iEww^a* ««M ***** wii^ fo*** *

-Ta.

142

V r

MA.DHVA LOGIC

;* i

I ?ra I TOT *afn wwi-

1 -

^gf*N lfSTO( I §TGTOT3W «P9ita<3TO #mwi

^: wiitu 1

y—

.•.^r—

FreK

i *rat forbid i Tipnaw

farot *nfm i to «wto tat: wfr ww-

anting i aa wt «m: *f»a*W. i amrafe-
aWfer aro tmPwfaiBftkJ faafwwj *ft *

■;■*•*■«

144 mIdhva logic •' " ; ;£|

qrcqrarosnfafwnT i ?m*lf w$& &m

- y.

5Bsi*f v fentar tot ^*iwit*i«iw-»»w^|
Sweeny i ** fi^i ?rar^«rsf«^ «fl»M|

sb&^to^i *» w^sfa *****
fo mafafa sura: i «i«nfo<*w ?w sRwafafa

* gfeas*r* ^fasmfvi ftft to: i

PRAMA.NAGANDRIKA. 145

3iTBi3TTirawT9: i **ra^wn*ni i «* m wa *
qmaswaa: S**f^ **n^ m^ra 3^3-

14G MADHVA LOGIC

tm: i qa*n«wmi g ^TfinjTHdteufww: « ft
sitow: * i ^erai: i *n ^: *t =! q?T wsan j

41<Ud)fdf<!l$*r *Mfrd<4l * W[>sft w: Tift <- lfli " n

3ft

PkAMAtfACANDRIKA 147

^%nt wff fore: i ^sf^fonw* *n«rcf*?T-

smfa: i *rt. *nra§^tfii i *<re3fa *fa*wwfow

«wwN4,fafo i tfsfawn w^^qf-^^-
?ren»ta irf?rerat g^f^fcfa <m#*S*3fti i

*ifirftife:i mr * f^ 5 ^ «* *****
**,, *m*z ifin " 9 ■***< *****

i48 MADHVA LOGIC

i *refa * *nfo*?rr * w^fh tot t^m *fa arf?ft*-

qrf«hrtre*n^ §^: ^RfaqwsjTafTSTWF* *fa ihfljj
tot: «*i*dffiif<fd ^qiftrisfosrTfH: i *ra: vm-

3fltapm I ?H^§ qlriV^ ^fe«*dt OTH^Tf^fe

«■*

***<.

t-BAMA^ACANDEIKi 149

?TH«ns3: l *? fs 3* ^T^l %feWT^ TOT

WW I qtsfrWT^ T »T*f?l « W^l 1 «*fo TOT

g^ i ^TfirroH&RftTOfaww ^ TO «ta «t* i

W^fM « to t£* sntm i f*^ *nf*r?re*-
wpi# i firnffiqfti^raa^ st ^foa«wi$*gii
«at$g«r^i at st^itto ^nt^fNti§«TO i w»-
*s^rf^ sgar wr« <*iirHr«*i*i*iyi TO*<ta

?PH«1iTO^ar«TMi ft*«|i *PTO*af?tft%ftr
g *fr wi i fiH«maifo i tTOt%asrra: «wtr

Sfii i

S^rcgnM ftfro ^ra <wro * ifiii to *ro

q ianfaftfil *nt* *s*tar ?i^ffrf to?ttoW to: »
nrt wiro «Ptot* * f* ^ ** wfin
to fMMMMM ***** '***'** ^

150

MA.DHVA LOGIC

iTTO *TWcf ^fai "BT*<*JrtJ^ I nr^mV

wh<w<hhJ *$$& i treat ^%WT^ ^q-wic^ |

forssw fea- q- ^ fg: i tot ^<wifcfd i srrtw-

tot wJTTrjJna «i?H«: i 5^t?tfan*nfFrw , >ra^
twsBrer. tot fifewpra w&tf?. i *nffr-

*mrafe«iiriifoiM ^ra^i^iHuj i tot qt w^H J
^sfrwnr tot sirre ^fa i ^fa^STfinRsi'f-

«T5[ a wMa « spur, a wfa tot WW*

3RTctf ^r^gtrro: i ^arsta

aa «m«w mvi twaTsfa mrfttn iwMf.

--*

■JE?

-W

'-'V*

•■$?

PRAHA^ACANDRIKA

151

fattarwfaj^Tct i aa f^t^fefw. i ufatnfaita-
fgfafta-ssT'rTfaflwsTa, i aa wfrmfa^t ftrfar. i

aa iresnmT'JrfMtasiteT^if g faaa fa^n, ssasiq,
aisa afapan, am ssfarcsrafafa i *ra aa, az *arfi?-
a?a%v, fawa a<a»i, *rata*ir<di<i, a«*faqaaf?eaT-

gawarfefta t 5^: l wiawronfadawterf**

i aaraOTas^ aafaqwraaV.
anaia^ awa*?nmpirfa*Nt *Tasr. I *aaaafa*lat
fefaa: j srofaaTaaarfaw^i i aa aafara-

jaa at ro i ffteiflwKnw : i aa?-

aTaaaa^rrta ^aaiftaaaa

jt^ I aar (*Hlw<«u«yw«a'

iti ^k: i *iaaa *a .«*wfiraHfc i aw

fTTrTT 4%y<*i1< I

^ngrfini

152 MADHVA LOGIC

*na afa a?«Taaa a*F«: ; wpww^wrmfH i aroit
TOT si^tsfaa: n*\uvife fa I ftr^tat tot ais?t taar.
3faai3Tf%fa i aatat tot aswtaaf «**ifefd i *re
aaw a^caatwaawain ar% i

ggi a i fa ftaT t%t%^: i aTvsaaanai anHl**i ^ta i
*tt*i tot watsta<a ^aWi mwntf-ifo i tVata
tot "aaaTaiTTa tot awta i

aaraaf *W««ll<«ti4t<jfd | ^HTqfTfffWT^Tra

aspRsrt^f?! i aiMir^ifac-^tis-faftfd a§r^>nt?ta i

aa* guwra wwwh faqarcnwfa: aaTfaa-
faaaca *nwfdmM' afa tat a«3 wfa I aaTiaa-

cqfa3 :t a f <r . q^rfa faafearfa i %a3TTJataa^i ^wift

^mfir i aw fannvnaa faqispsrreMceroTTaT^ i

%rei *qftft£Mft sfo aiaTft ^sarfa, aw aqtSTMTaa w»

awgqq'a: i aa faafisa'sqa, arfaaaatVaT: «nf«i«M-
jfen faww. i a aTfaa-f2ra^Taant*r^irF^3raT-
qf^-aiataqarw. q^ ^amrer: i sjnfipra^araa*:-
faanai fgtfar. i a =a fafaa:, araaTtaa: uvufttV
arracaTfaaafa i aaTaaTtaat fefaa: i wrsroa:
faaaTaa i a ta i ^rrot tot i aaarctaaS *****
'a^^a^^TarttaFda'a^ i *ra aaareta'^araa:, * f
aTaaa i taatat aaT i ata^iRa afa fawife*

■■a

*is

PRAMltfACANDRIKA 153

vravfti *"* *** ** ***** ^ T ^ fl ^

<$ fan* <msfo Wfcwwmwm 13 ;™*z

ftmmor m fiwn* «**■*■*•*'
rftritairt fern i «wfc ** <mr*TT*fc™

. , .-. 3>& / ■ •>-

154 MADHVA LOGIC

^^^-HW^ira^rafH i *rrat m\— q&ft ««qi*i1*fl-
«rnwg5?re:q^ 1 *ra ^f%5^r^ trow: 1 ^atiw-

IK, —

otllM^dt «H% I ?piT^ «uf*WHJcl 1 *rat wfr

n<*rad erateriwn^fara ^rM«um<hm "shir i w

H*\>M d^5d*&X|fafd ^rT^*toqTq^iffT llw I

wftfa 1 w^mv^fi wifaft«*fsr *hi«j*iw1 -?WB"

»■..-»(:

■«.'.

----- " * ■ k-5>5

PRAMAljrACANDRiKA 155

m n«mffl Sgfzrw. I *raT aj^ ftw

*ra ^^nfq twani s^T^t *n% ' to *wwra:

wanwrere* *w*rf to tort « '^^ [
gm sp^t fSrar: ^st^^^I «^**i ^V™.

156

MADHVA LOGIC

,., - -i

M.i

■-03

>/?^

ft** ft**; «* **^^ *™T?

„- -v- --'' - "'• ""

158 MADHVA LOGIC

^3T ^STTf? I W*WP^lrfdMW*a^5re^ i&\-

*ll4Ul<i «KTO ^^i^m ^ffqfa ; <afc<<i irf?r «uq I Mi-
di!*^ dft ^Wddlq^SJ TWlt? |

^|ch|§-T^ydwfafa*rTt a?Rt *RJ^t 3T3BP( I *P3

«id«l*i3f9a f3R3TO^i tf^M^d f^fT fsfc<JI4iiHH4l«c|9-
^tWT5|^iaf*T^IT5Wtr^^J 3dM^d W^FTf^ I 5P5T3T

i^TT^r^ i mws ^k^: 3^t"

^M«h<3*l «Icftl§-T ^§=31^ I d'Hfdm^ctftq^TOfa

41«mi i ^ren si^h ftra^ga ; sra%^Tqt: ^iRraFW-

^a ^ 5TC^q5T^ fd<q i^ <* JI MK : «WM l JHM

-■?:*

, ;m

PBAMA^ACANDRIKA 159

^Pwarq , i 4»«*wi ■4iiiwa' *&&m: *rasfa 7i^nfiw.
f^fajK i sgwqt ^r^ir. i *ra wrfaw^: sj^jreftrf'iftrfr-
*rat^^ " ^d^j gtMi-^i^fH maririts^. unfair

q fa^qi^fe gg K^ifc^frffi l ^ rata i am ft

firero ^t «b «ikBm i to *gs*f *ra attf *nar **
fctra a «n?fw «*<«a » MHmwnft« Tfe i a*r 3*

160

MADHTA LOGIC

« n ftft» i * qf qi »w«*ici fate *ptoto « it arar^fa-
■vm°i<^ nfwftfsr *are?r wrf. I wi «w

sftrfif fro i gwsgw ^fo ' *a sifawsi
%fa i ?ra <ti<«iid«iM4ii«Tn: ?ra «w^rr

■ *'^-

, v--

'^1

,3&

PRAMA^TACANDRIKA 161

BZJtsRlWltsfa Kd\4WI«MI<l **«w«-*i3 I liflUi

frni i i^m*n *ffa qfrpwn t^it writ «*n: i
yprfll ftf*nit ^t fogtsfa^Hfa ^ i

3<.Himw*i S 9 1 iNh vn^ *^ ■
3i5 «te ^^^ sraw^raTq. *ifcKWifc*i+i-

IIWT^I rl^^sfq qfl3Tgq«PWT?l, fWSlWITW-

^f^fa %9 i 3<r. wrote ^rotafcraracrei 39fnt%-
**taWRn?ftfa f faftwiq , i * *^ : V*™

^irf^jt 'tr^rw^ ?R5^m^ *<|g*snft-

unmwnwni m ft» ft^t fcft«

*ntfra: qfhn«i i ww lismw *** '

Sjfe^iS -^ wi*r. < **nf*: *«^ "«*" ***
21

162

MADHVA LOGIC

1 fij ^CTTWilRj*!^

3*T fif«KiHii4i«i; TOW

#1 ^ifdftaf f*

* w*

■\::

:*\

■^i

.'■■m

* ^wts^Hiw Pro *S3 to: ^r^E^i fafriwiKict to*'*; t
t wm; (?)

■-.,.-:

pramai*acandjukI 163

WoM^gwR q^Rwfcnt i "tat ^%^ <*tas* «f?r

^^reaR^^rss^fsrefa'^ Monism i *rarft

■4

164 MADHVA LOGIC

^#rt%^ ^a^f^fa (rat%rcra: i ■erCTft atitaf-
Sfnwii wra it ftrarrfts#f?f ^twi i <^r»mipi

* ftnri^ *f?r g sfets'a* its: i

PRAMA^ACANDRIKA 165

sp*w| i 5N wf srra3 3^3 ffffl nroro strcra *far i

166

MADHVA LOGIC

<iw<inEtRwm i3 far *rW«r?i #sr i jraras*-

f '-**.'!

wwwaf St *rm$ <ict q* wra*?wi*«ra fl^terer ^insrw w^