Anumana PDF

Part - 1
Chapters
(If validity is denied to inference then all worldly transactions cannot be
conducted with the mere help of perception. All the people of the world will
become motionless as if they are painted in a picture).
Chapter - 1
A brief discussion of the process of inference
General background
Inference, a historical retrospection
Ground of establishing inference
Anumana as in Nyaya
Chapter at a glance
Inference as to Samkhya
Vedantic concept of inference
Concept of inference in Jaina
Buddhistic view regarding inference
A brief discussion of Inference 11
First chapter
A brief discussion of the process of inference
The cause of all human behavior is knowledge . Knowledge of the objects of the
Universe. Without this the human being won't be able to behave consistently with
the world. Again, this knowledge is of two kinds viz. true or valid and false or
invalid. False or invalid knowledge gives birth to a false experience, which goes
against human behavior. Which does not represent the real nature of its object.
What remains is a valid knowledge. Now, a problem arises
General
Background that how to distinguish a valid knowledge from an invalid
one? What are the means over there through which we will
come to know the real nature of an object? This problem creates a room for the
intellectual traditions of Indian Philosophy to say something. It is accepted by all
the traditional systems of Indian Philosophy that the knowledge, which is a very
cause of human behavior, can be obtained either directly or indirectly. But these
philosophical systems have a disagreement regarding the means or proofs of
obtaining the knowledge. The number of proofs has varied greatly with different
schools from one to eight.
This idea is as old as Veda. The Vedic seers also have realized the same. A
historical study of inference provides some adequate information about the
consciousness of the distinction of direct and indirect knowledge of Rgvedic
period.
In Prof. Jwala Prasad's language "the earliest text which appears to be significant
for any inquiry are some verses in Rgveda2.
1
' H ^ c i g K f k ^ s l l ' H TSA
2
Important verses are 1.164.4 & 37, X.82.3 &4, X.129.6 & 7
A study of these, as will be shown presently, indicates on the part of their authors
a consciousness of the following facts: (i) those that are immediately present to the
senses; and (ii) those are not immediately present to the
Anumana,
a historical senses, that is to say, those that are remote3 ". In short, it
retrospection can be concluded that, the seers understood the cause of the
Universe by means of reflection, and have been inferred through another means of
knowing about the things, which are remote. Upanisadas (Jajnavalkya, 3-7) and
Aranyakas , in this regard, are in no way exceptional.
Now let us discuss that how the Indian philosophical systems have developed this
idea in the later period. Nyaya-Vaisesika school of Indian Philosophy accepts four
means of obtaining knowledge. In philosophical terminology
Anumana as in . „ .
Nyaya these ways of knowing are called pramanas. It is called
pramana because it maps the reality of the knowledge.
Therefore, these are considered as the instruments of obtaining a valid
knowledge5. Pramana has a real correspondence with objects, in the sense that the
nature and attributes of the objects, as revealed by Pramanas, are not
contradictory, despite of all variations in time, place and other conditions6.
Those four proofs or means are perception, inference, comparison and verbal
testimony7, which generate different kinds of valid knowledge. Anumana or
inference is an important source of knowledge. This is the instrument of inferential
3
HIE, p. 10
^frWciis^fcrî sryrpT ^ * e w l - CÂIK"^*. I.?
5
TSA
6
N V T T l . l . l a s i n N T K . P. 52
7
NS-1.1.3
cognition . The word Anumana is a kind of knowledge (mdna), which we get after
(ami) some other knowledge9. So, the after knowledge of an object or the
probandum (lingi-artha) through a probans (lingo) is called
is
inference. The word 'ami1 conveys that this knowledge is
Inference?
followed by some other knowledge. In other words, it is a

dependent knowledge. It is dependent on perception, which is an immediate
knowledge. Vdtsyayana holds that "no inference can follow in the absence of
perception"10.
The definition of anumana has been discussed in an ascending manner starting

from Gotama to Gangesa. According to Gotama that is called inference, which is
preceded by it, is of three kinds, namely, purvavat, sesavat and sdmanyatodrsta".
Here the paraphrase 'preceded by it' is not so clear.
Definition Immediately after that Vdtsdyana, the commentator of Nydya-
of
inference sutra, explains the word saying 'the perception of the
(invariable) relation between the probans and the probandum
as well as 'the perception of the probans'. The perception of the invariably related
probans and probandum means the recollection of the probans. Through this
recollection and perception of the probans the object will be inferred which is not
a direct knowledge12.
'What is the necessity of accepting inference, only one proof or means of

obtaining valid knowledge i.e. perception can fulfill all the requirements' is the
8
TSA
9
N B h . 1.1.3 focfr f c T ^ f a fcHf^ J Mlss5>KNiMMHH^HHH I
10
NBh. 1.1.5
" N S . 1.1.5
objection raised by Charvak philosophers13. This is the only school, which does
not accept inference and at the same time asks for the ground of its establishment.
Vdcaspati Misra, a prominent scholar of Nyaya tradition with other philosophers,

prepares a ground to establish the validity of inference14. The whole thing is in a
dialogue from such as:
Charvakas: Only perception is enough. We don't accept
another pramana called Inference?
Ground Naiyayikas: Well, it's up to you.

of Charvakas: But why you people accept it? What is the
establishing
inference ground there?
Naiyayikas: By the way, why do you ask this question to the
people like us? Why not to anybody else?
Charvakas: Because you people accept it.
Naiyayikas: You mean to say that we have either doubt or
illusion in our mind and you want to remove it.
Charvakas: Right.
Naiyayikas: But how did you come to know that? What is the instrument with
you, through which you measure the ignorance, and doubt or illusion lies in
others? Can you perceive that doubt etc. directly?
Charvakas: No, this is the speech of a person, which proves the entire thing.
Naiyayikas: Right, this the ground. Through the speech only we come to know
the mental condition of a person. This works as a mark or sign and this is nothing
but inference.
Thus, they establish the validity of inference.
13
Chan'ak philosophers do not accept inference as a means of valid knowledge. According to them
perception only can fulfill all the requirements.
14
Tarkabk/M, P-138, ed. by Surendradev Sastri.
The Nyaya-Vaisesika system of Indian philosophy has developed a systematic

process of inference. This process consists of a number of components, which are
related altogether in a particular way. The Nydya system analyses the process of
inference like this: A person, bearing the knowledge of
How does invariable concomitance (vyapti) ' of smoke and fire in his
it
occur? mind, sees the smoke on the mountain. Just after the
perception of the smoke on the mountain he recollects that
smoke is invariably associated with fire. Immediately after that the knowledge like
'the mountain is possessed of smoke which is pervaded by fire' arises. This is
called Paramarsa. After the knowledge of Paramarsa there arises the knowledge
like 'the mountain possess fire'. This is called inferential cognition. The above
process of inference can be explained like this:
1) hetu darshan (perception of hetu or sign on the paksa)
2) vyapti smarana - (remembrance of invariable
The
concomitance between probans and the probandum)
process
so far 3) paramarsa — ( that the paksa possesses hetu which is
pervaded by sddhya)
4) anumiti - (inferential cognition or the result)
This is the process of inference in short.
In Nydya system the classifications of inference are seen in various forms. It is

classified on the basis of both psychological and logical grounds. All and above
there are three main classifications of inference .
Psychologically, it is of two kinds, viz. Svartha (for one's sake) and parartha (for
the sake of others). And on the basis of logic it is said to be of three kinds.
According to PrdcTna Nydya, there are three types of inference like: purvavat,
,J
TSA
l6
IL,p-162
Sesavat and samdnyatodrsta whereas; Navya-Nyaya

Varieties classifies inference like kevalanvayi, kevala-vyatireki
and anvaya-\yatireki. Let us discuss these three classifications of inference
elaborately.
The first type of classification puts inference into two, namely, svartha and
Parartha. This division of inference is not available either in Gotama sutra or in
the Vaisesika sutra of Kanada. It is Prasastapada'7,
First type
of the who mentioned it for the first time in the commentary
classifications
of Vaisesikasiltra. This same classification is also seen
in the later period of Nyaya School. It is Udayandcdrya and Jayanta Bhatta18, who
have classified inference in this style. Acarya Udyotakara19, the sub-commentator
of Nyayasutra, mentions a word like 'parapratipddika', which proves the
superiority of the tradition of accepting the division.
If a person wants to convince to himself and for that if he follows the process to
infer something then that will be called as svdrthdnumana20. For example, a person
makes out a universal proposition (vydpti) that 'where there is smoke there is fire'
from repeated observations of smoke and fire in kitchen etc.
This universal proposition is nothing else than a relation between two things and is
called invariable concomitance. Bearing this knowledge in his mind, one day he
17
PBh, gune anumana prakarana cTW^RPltjHH ^ W t W ^ T ^ f ^ T T 3)c|i|c|^rt>l ^fct effl^ I
18
NM, Ch. 2
19
NV 1.l.32, myfcNK<fci $ cjicKJi^r^jcTr w^rer ^rf^nafa^fr cjicKfwdiîRi ^s^nraiTI
20
TSA
happens to see a string of smoke somewhere on the

Inference
For mountain (hetu darsan). All of sudden he will
one's sake
remember the universal proposition of smoke and fire
(vydpti smarana). Then he will realize that this smoke, like that of kitchen, is also
invariably associated with the fire {paramarsa). Then he will have the knowledge
of inferential cognition of fire (anumiti). This, in short, is called the inference for
one's sake.
The inference for others is a demonstrative one. The demonstration with the help
of five membered syllogisms. A person, being convinced for himself, attempts to
convince the same to others21. This is based on a
Inference pure psychology. For that he needs some formal
for
other's sake statements. In short, the later one is totally
dependent on the former one. Those syllogistic
expressionsor avayavas are: 1) proposition {pratijha), 2) the reason (hetu), 3) the
example (udahdrana), 4) the application (upanaya) and the conclusion
(nigamana). Such as:
Somebody proposes like 'the hill has fire' (pratijha)
Because it has 'smoke' (hetu)
'Where there is smoke there is fire'
for example, in a kitchen (udaharana)
This 'smoke is also associated with fire' (upanaya)
Therefore, 'it has fire' (nigamana)
Here, pratijha or assertion is a statement or proposition, which is to be

inferred. This lets know to the hearer what exactly is going to be inferred. For
example, the mountain is fiery. This statement is a proposition, because later it
is going to be proved.
After having such statement, now the hearer proceeds to know that on what
ground this statement is true. Then he is told that there is smoke. From the
perception of smoke he can conclude that it is a valid assertion. This is called
hetu or reason. This supplies a ground for the establishment of the above
assertion.
An example uddharana is a reminding factor of the relation of invariable
concomitance between probans and the probandum. After seeing the smoke on
the hill he recollects the similar instances, which he had experienced in the
kitchen etc. This is called example, the third member of the syllogism.
Upanaya is the fourth member of the group. This consists of two conditions
like: there is smoke on the mountain {paksadharmatdjhdna) and that smoke is
invariable associated with fire (vydptijdna). Now the hearer combines these
two into a complete one. And what he gets is pardmarsa.
The last member of that group is nigamana or conclusion. It proves what
had been asserted to be such as: there is fire on the hill.
According to second type of classification the inference is divided into three kinds
Second type such as: purvavat, sesavat and sdmdnyatodrsta22. It is
of the
classifications Gotama, who mentions these three forms of inference.
After that Vdtsdyana, the commentator of Nyayasutra,
comes forward to explain it in a systematic way. He has derived the meanings of
the classification in two ways .
21
TSA
22
NS 1.1.5
NBh 1.1.5, arsrar tj^crftfr...

According to Vatsayana that is called purvavat, where the effect is inferred from
its cause" . That means, a purx'avat inference is that in
Purvavat which we infer the unperceived effect from a
perceived cause. For example, we see the rising clouds
in the sky and inferred that the rain is going to fall.
It is an inference of some unknown thing (future rainfall) from a known thing
(rising clouds). In short, pur\>at is known as karana-lingaka.
A sesavat type of inference is the just opposite to purx'avat25. It is inferred

• through an unperceived cause from a perceived effect.
For example, a past rainfall of a river is inferred from
its fullness or muddiness of water, swiftness of current, which are perceived.
This type of inference is known as kdrya-lingaka.
Samanyatodrsta type of inference is the perception of some object at some time

which was somewhere else due to its movement26.
Samanyatodrsta
This inference is based on generalized relationship.
For example, the movement of the sun. The motion of the sun cannot be perceived
directly.
97
The third type of classification is specially designed by new logicians . But we

find such division in Nydyavdrtika also. Here it explains the types with relevant
Third type of examples. This division is based on different methods
the
classifications of establishing vyapti or invariable concomitance.
These are 1) kevaldnvayi, 2) kevalavyatireki and 3) anvayivyatirekin.
24
ibid
25
.bid
26
ibid
27
AC, kevalanvayiprakarana
This inference occurs when there is the relation of invariable concomitance

between the subject and the object. Here the knowledge of invariable
concomitance between probans and probandum is seen only through the methord
of agreement in presence (anvaya), since there is no negative instance of their
agreement in absence .
This will be applied there where the subject and object go together. For example,
Kevalanvayi 'All knowable things are nameable
°.^.y x pot is a knowable thine
r
positive) °
Therefore, it is nameable.'
In the above example, there is an affirmative concomitance between knowabihty
and nameability. All things are knowable and nameable. We can't just find its
negative concomitance.
This inference is obtained only through the negative invariable concomitance.

It is dependent on a vyapti or invariable concomitance
Kevalavyatireki
(only negative) between major term with that of the middle term. In this case
9Q
vyapti is a purely negative proposition . For instance, 'what is not different from
other element has no smell
The Earth has smell
Therefore, the Earth is different from other elements.'
This type of inference contains both positive and negative kind of invariable
concomitance30. In other words, there is vyapti or invariable concomitance
between the presence of the middle term with the presence of major term as well
as between the absence of the major term and absence of the middle term. For
example, if somebody proposes a statement like
Anvaya vyatireki
"where there is smoke there is fire" will be true and the
(both positive
statement like "where there is no smoke there is no
and negative):
fire" is also true. The former is the example of positive
concomitance whereas, the later one is the example of the negative one. Therefore,
the statement like 'mountain possesses fire because of smoke' is the instance of
both positive and negative invariable concomitance. In short, we will have a look
over the varieties of inference.
Inference
Classification on the Classification on the Classification on

basis of the purpose basis of perceptibility the basis of vyapti
Svartha *• Kevalanvayi
Kevalavyatireki
Pararth Punwat Sesavat Samanyatodrsta Anvayivyatireki
Though different opinions are observed between the two schools (Navya nyaya
and praclna nyaya) regarding the structure of inference still the function of both
the divisions are same. The components, which are involved in the process of
inference, will be discussed in the next chapter.
After discussing the Nyaya theory of inference now I proceed to sketch a brief
discussion of the process of inference as occurs in other philosophical systems.
Samkhya system considers inference as the knowledge of the pervader (vyapya)

arises from the knowledge of invariable concomitance3'. And the result is the
understanding of the concerned person32. Samkhya
, , karika explains inference as based upon the cognition
Inference
as in
of middle term and the major term33. According to this
Samkhya
system inference is of three kinds such as: purvavat,

sesavat and samanyatodrsta34. A further division of the inference is also available
in this system like vita and avita. This system requires a ten-membered syllogism
in the context of inference.
Samkaracarya, the founder of avaita vedanta, holds that there is no necessity of

inference with regard to transcandal knowledge (paramdrthikajndna) but it can be
applied into the empirical world (vyavaharika jnana) ' This school accepts
., ,_ . Inference as instrument of inferential cognition. And
&
Vedantic
concept of inferential cognition is that which is generated by knowledge
inference
of pervasion, functioning as cognition of pervasion. .

Medghanadasuri, a great figure of Visistadvaitavedanta, characterizes as two
terms i.e. pervader and the perdaded having a positive invariable concomitance
between them and from the sight of the pervaded the remote perveder's
remembrance occurs . Madhdcdrya, the founder of Dvaitaveddnta, defines
inference as ' nirdosopapattiranumdnam' . Here, upapatti means liriga of mark .
31
SS - 1/100, «R(«|«£j^|: MfcH^WM^MH
32
Bhasya on ibid, 3 ^ f a R l ^ iflÂJ ejtar I
33
SK - 5
34
ibid
35
BSSB- 2.1.11
36
Vedantaparibhasa, ch. 2, ^fedMAjMujciîd ^Rta^mcbicit^cbslH^Hl-TH I
37
Nayadyumani, p 195 as in BDB
"' Dasaprakaraiidni, p - 34
39
.bid.
So, the faultless mark, qualified by invariable concomitance, is called inference.

Here, there is a division of inference as karananumana, karyanumana and
samanyanumana40.
Prasthdnaratndkara, a suddhddvaita Vedantins text, defines inference as
the instrument of inferential cognition. And inferential cognition is a product of an
instrument namely invariable concomitance. However, this system accepts three
membered-syllogism.
Pramdna, according to Jaina logicians, is 'svapardbhdsi'. That means it is

the knowledge, which reveals itself as well as to others without any hindrance.
Basically, this system has a difference of opinions regarding the numbers of
pramdnas. Some Jaina logicians say it is two whereas, some
The concept other accepts it as four40. The definition of inference, as has
of inference i • i i . . .
een lven D
in Jaina " g Y the jaina logicians, is 'ami' means after and
i
p i osop y mana' means the instrument through which the knower
obtains the knowledge of place, time and nature41. 'Anu' here indicates the
remembrance of the knowledge of invariable concomitance between probans and
probandum.
So, after perceiving smoke from a distance, the person remembers the
invariable concomitance of probans and probandum and immediately after that the
inferential cognition arises in him. The old Jaina logicians accept three types of
inference whereas, the new logicians, accept two namely inference for one's self
and inference for others sake .
40
Anumana pramdna, p. 235
41
SyaJbadamanjri - 20, 3TJ M ^ l f w l ^ t W*FV TT^JTR^R ifati MRfe^ci
" Anumana pramdna, p. 235

Jaina logicians also accept vyapti as a main part of the process of inference. It,
according to them, is an invariable relationship of two objects. Further, they accept
vyapti, as of two kinds , like anvayavydpti and vyatirekavydpti. The components,
which are engaged to produce inference, are paksa, sadhya, vyapti, hetu and
hetvabhdsa. The number of syllogisms varies here. Some say it is five whereas ten
others. The faults of inference are paksdbhdsa, hetvabhdsa and drstdntdbhdsa.
Paksdbhdsa is of five kinds, drstdntdbhdsas are two in number and hetvabhdsa,
according to Siddhasena Divakara, is of three kinds such as: asiddha, viruddha
and anaikdntika.
This is the process of inference according to Jaina logicians.
The Buddhist School of Indian philosophy accepts pramana as infallible

knowledge44 (is the right cognition), of two kinds because of two types of objects
(mdnam dvividham visayadvaividhyat). Logicians of this system hold that the
objects can be perceived either directly or indirectly. No third alternative can be
Inference entertained. But we find the reference regarding four types of
according to pramdnas in VrQ-Dignaga school of Buddhisim45. In Prof.
Buddhist Stcherbatsky's language46, the direct knowledge as has been
defined by Buddhist logicians, is "a synthesis between a
sensation and a conception, the indirect one is a synthesis
between a sensation and two concepts".Buddhist logicians hold pramd as
practically useful knowledge and pramana is its source. Hence, pramana has a
capacity to reveal the real nature of the objects.
43
ibid, p. 214
44
N B , p . 4, arfaHcTTCcf) sTFT W W F P T
45
Upayahrudaya, p. 23 as in BDB, p. 236, ^fiw WW{ JJc^l-jHM^HHHH^fcl I
46
BL, p. 231
This revelation can be obtained either directly or indirectly. Therefore, this system
holds only two pramdnas viz. Perception and inference as two different sources of
acquiring the real nature of the objects.
Inference, according to this school, is cognition of an object through its mark.

_. They hold mark as a possessor of three qualities viz. its
The
definition presence in the subject, its presence in similar instances and
its presence in dissimilar instances. So, the inference is a knowledge of an object
through a mark, which (mark) possesses the above three aspects47. Buddhist
logicians classify inference as 5var//za(inference for one self) and parartha
(inference for other's sake).
Inference for one self is a knowledge of an object through its mark. In other
words, the cognition ofsadhya arises in a person who happens to see the reason51.
The inference for other's sake, according to DignJga, has
Svarthanumana three elements viz. Paksa, hetu and drstanta. He, otherwise,
calls it sadhana48. If any element of these three happens to be
defective then the total inference will be defective. According to him, there are
three types of sadhandbhdsas (fallacious grounds) namely, paksabhdsa (fallacious
grounds), hetvdbhasa (fallacious reasons) and drstantdbhasa or fallacious
examples49.
Thus, the process of inference for others occurs when a reason is set to produce a
conviction in others. Here, Buddhist logicians have employed three syllogisms or
47
N P , p . 7, 3T3HH fcT^TTsfc^TfT I f c ^ l ^ l F & 1 * > M ^ t H I
48
NB, Ch. 2
49
NP, p. 7, ?r?f trsTTfc CJTHIPI ? n w T I
nyayavayavas like an inferential subject, a probandum and an example . This

syllogism is a cause, which produces an inference in mind of the hero. Nyayabindu
explains syllogism concedes in communicating the three aspects of the logical
mark for others. Out of these three components they form a reasoning.
That is like this;

The proposition: This mountain is fiery
The reason: Because it has smoke
The example: All that has smoke is fiery like kitchen and
Whatever is not fiery has no smoke like lake etc51.
Components of inference
There are two main components of inference as it occurs in Budhhist logic. The
first one is Vyapti or invariable concomitance and the second is Paksadharmata.
The former is the ground of establishing inference whereas the latter is a supporter
of that ground. Invariable concomitance as discussed in Pramdnavartika is like
Vyapti this. The vyapti is not the mere observation of some positive
instance of a particular relation, for example, between smoke
and fire52. This case, according to Buddhist logicians, is not enough to establish
the relationship between smoke and fire because the observation proves that they
were connected in the past. This cannot satisfy the entire case. On the other hand,
it cannot be dependant on some negative instances of fire and smoke because we
cannot assure that there will be no instances in future of the presence of smoke in
the absence of fire.
According to Buddhist logicians hetu is of two kinds sadhamya and vaidhannya.

cbnIebKunncrrq- err WWm 3T P|A|H*ld I
N. N V
Therefore, the vyapti is proved neither by perception nor by non-perception. But

by the knowledge of causality or identity. Instead of saying "where there is smoke
there is fire' we can say that "smoke is produced when fire is produced. And it is
not there where fire is not there. So also in the case of identity. When it is known
that "A" possesses the nature of 'B' then we can define such types of relationship
there. Because in this case, without one the other cannot exist. This Vyapti is of
two kinds such as: Anvaya and Vyatireka.
Paksadharmata is another main component of the process of inference. A mere

hetu having an invariable relationship with sddhya cannot produce a valid
inference. Paksadharmata is one of three forms of hetu which
Paksadharmata
means paksesattvam. Another two types are like
paksyeasavtam (the presence of probans in similar instances)
and vipaksaeasavttam (the absence of probans in dissimilar instances). According
to Budhhist logicians, the hetu is of three kinds such as Karya, Svabhdv and
Anupalambha. The first one shows the cause and effect relationship like where
there is smoke there is fire whereas the second one proclaims the identity.
Faults of inference
In the above example the mountain is an inferential subject. If in any case it
happens to be proved as false then the inference will be false. In the similar way, a
false probandum and example will draw a wrong conclusion. Therefore, Buddhist
logicians have identified three kinds of defective grounds (sddhanabhasas), of
which a fallacious subject (paksdbhdsa) comes first.
Acdrya Dignaga has mentioned four types of fallacious subjects viz.

1. A fallacious ground against perception, (Pratyaksaviruddha)
2, An inferential ground against inference (Annumdnaviruddha)
3. An inferential ground denying the synonyms (Pratitiviruddha)

a sa asa ^ ^n j n f e r e n ti a l ground against one's own belief
(Svavacanaviruddha)
The first one, for instance, "sound is inaudible" goes against our perception.
And if such fallacious ground happens to be an inferential subject then it

will gave birth to a false inference. The second type goes against our experience.
For example, " a pot is eternal". In fact, it is not so because it is a product and
products are non-eternal.
The third type of fallacious ground denies the similarities. Like when
somebody says, "the moon is not candra". And the fourth one goes against one's
own statement like "Inference is not a source of knowledge". However, Dignaga
in his treatise called Nydpravesa has mentioned nine types of fallacious grounds .
Fallacious reasons, according to Dignaga55, is of three kinds such as: asiddha

(the unproved), aneikdntika (the inconclusive) and bddhita
(the contradictory one). Again, he subdivides asiddha into
four, aneikdntika into six and bddhita into four. So, the total
number is fourteen whereas, Acarya DharmakTrti56 gives the total number as eight.
1. Asiddha:
Dignaga does not give any definition of asiddha. He only mentions its categories.
But as its name suggests this reason is not proved i.e. familiar. There are four
divisions such as:
53
NB, Ch. 2
54
NP, Ch. 3
55
NP, p. 2
56
.bid.
a) Ubhaydsiddha: this kind of fallacious reason is unacceptable as a predicate to

both the parties i.e. two systems. For example,
Sound is non-eternal, because it is visible.
b) Anyatardsiddha:
It is acceptable by one party only i.e. by either system. For example,
Sound is evolved
Because it is a product.
Here, according to Mimdmdsakas and Samkhya philosophers, Sabda or word
cannot be manifested but it can be created.
c) Samdigdhasiddha: this type of fallacious reason has a doubtful existence. For

example,
The hill is fiery,
Because there is vapour,
Here mist can be mistaken as smoke.
d) Ashraydsiddha: here its very existence is doubtful. For example,

Ether is a substance
Because it has qualities.
2) Anaikdntika: Again, anaikdntika is of six kinds:

a. Sddhdrana Anaikdntika:
Aneikdntika : here the middle term is too common, it resides in the major term as
well as in the opposite of it. For example,
Sound is eternal
Because, it is knowable (i.e. knowable can be eternal things as well
as non-eternal things).
b. Asddhdrana Anaikdntika:
Sound is eternal
because it is audible
Here, Sravanatva is an uncommon thing. It exists neither in the major term nor in
its opposite.
c. SapaksaikadesavrittinnpaksvydpT aneikantika: It resides in some similar

instances but in all the dissimilar instances. For example,
Sound is not a product of effort
because it is non-eternal
d. Vipaksaikadesavrittih sapaksvyapi aneikantika: It resides in some unlike

instances but in all like instances. For example,
Sound is a product of effort
Because it is non-eternal.
e. Ubhayapaksaikadesavritti aneikantika: this type of fallacious reason resides

only in a part of similar instance and that of dissimilar instances as well.
Sound is eternal
Because it is incorporeal
f. Viruddhdbhicari anaikdntika: that which has a contradictory invariable

concomitance57.
Sound is non-eternal
Because it is a product
Here, the invariable concomitance is 'whatever is non-eternal is a product, like
pot' (Vaisesika opines) and
Sound is eternal
Because it is always audible
Here, the invariable concomitance is 'whatever is eternal is audible, like sound
etc' {Mimamasakas opine like this)
Here, both the reasonings are correct, but this leads to a contradictory conclusion.
3) Biruddha: it is again of four kinds such as:

a) Dharmasvarupaviparitasadhana: that which proves the very opposite nature of
the thing that which is going to be proved.
Sound is eternal
b) Dharmavisesa vipantasadhana: here, the sadhana or hetu itself is contradictory

to the major term. For example,
The eye etc. are in service of somebody else
Because, these are made out of particles, like bed etc.
c) Dharmisvarupavipantasadhana: It occurs when the middle term is inconsistant

with the minor term. For example58,
Samdnya is neither a substance nor a quality, nor an action
Because, it depends upon one substance and possessed qualities and action, like
generality-particularity. But in fact, generality does not depend upon particularity.
d) Dharmivisesavipantasadhana: That which proves the opposite of the particular

kind of Dharmin that means probandum is called Dharmivisesavipantasadhana.
In the above example the bhava may not only a dravya or guna or karma but it
may be asatpratyayakartu like abhciva.
57
ibid.
The number of fallacious examples varies among the Buddhist logicians. Dignaga
mentions the number as ten whereas, Acharya DharmakTrti extends the number up
to eighteen. However, they have divided the fallacious
_ _ -,, - examples into two categories like heterogeneous fallacies
r
Drstantabhasas: ° °
(Vaidharmaya Drstantabhasas) and homogeneous fallacies

(Sddharmaya Drstantabhasas). Both the types have equal numbers i.e. each class
has some sub types. Here, my discussion is based upon Dinaga's Nyayapravesa59.
a) Sadhandsiddha: An example non-cooperating with the thing that which helps to

establish probandum. For example,
Sound is eternal
Because it is incorporeal (amurta)
That which is incorporeal is eternal, like an atom.
Here the example i.e. atom cannot serve the purpose because they are not
incorporeal by nature.
b) Sadhyadharmasiddha: An example non-cooperating with that, which is going

to be proved (the probandum). For example,
Sound is eternal
That which is incorporeal is eternal like intelligence.
The intelligence is the example here, which is not eternal. So, this also is not able
to serve the purpose.
ibid.
59
NP,p.5
c) Ubhayasiddha: An example which is neither co-operative with middle term nor

with the major term. For example,
Sound is eternal
That which is incorporeal is eternal, as a pot
d) Ananvaya: An example showing lack of universal connection between probans

and probandum. For example,
This pot is eternal
That which is a product is eternal like a cloth.
e) Viparitdnvaya: Here, this example shows a reverse invariable concomitance

between middle term and major term. For example, "here it is seen that which is a
product is eternal". The invariable concomitance will be like whatever is non-
eternal is a product, like pot etc. It goes against our experience.
Similarly, heterogeneous examples also do not serve the purpose. He has

mentioned five kinds of heterogeneous examples with their respective examples,
such as:
f) Sadhyavyavrtta (non-existent of probandum)
g) Sadhandvyavrtta (non-existent of probans)
h) Ubhaydvyavrtta (non-existent of both i.e. probans and probandum)
(i) Avyatireka (a heterogeneous example showing an absence of disconnection
between probans and probandum)
j) Vipantavyatireka (The case where the absence of probans and probandum is

reversed).
Selected Readings for First chapter
Six ways of knowing

-Dr. D.M. Datta, University of Calcutta, 1972
Encyclopedia of Indian Philosophies
-Karl H. Potter, Matilal Banarasidas, New Delhi, 1977
A history of Indian logic
-Dr. Satis Chandra Vidyabhushan, M. Banarasidas, Delhi, 1978
Buddhist logic
-F. Stcherbatsky, M. M. publishers Pvt. Ltd., New Delhi, 1984
Anumcma pramana
-Dr. Baliram Shukla, Eastern Book linkers, Delhi, 1986
Indian logic
-Dr. B.N. Singh Asha publications, Varanasi, 1988
History of Indian Epistemology
-Dr. Jwala Prasad, M. M. publishers Pvt. Ltd., New Delhi, 1987
BhatTya Darsan Brhatkosa
-Bachulal Avasthi, Sharada publishing house, Delhi, 1997
List of abbreviations
BDB BhdratTyadarsana Brhatkosa BL Buddhist logic

SK Samkhyaka rika
BSSB Brahmasutrasdnkarabhdsya NP Nyapravesa
STK SamkhyatattvakaumudT HIL History of Indian logic
SBD Sydd\ 'ddamanja rT Prasastapddabhdsya
PBh
PV Pramdnavdrttika
NB Nydyabindu AC Anumdnacintamani
NS Nydyasutra TM Tarkdmrta
NBh Nydyabhdsya
SS Sdmkhya-sutra
TSA Tarkasamgraha, Anumdnakhanda
IL Indian Logic HIE History of Indian Epistemology
(Why does smoke arise from fire but not from the water? The only possible
answer is that smoke and fire are so related because we learn their relation from
experience).
Chapter - II
Components involved in the process of inference.
• Paksa
• Paksatd
• Paksadharmata
Chapter at a glance "> <?-//,
• Hetu
• Vydpti
• Paramarsa
• Hetvabhasa
Components involved in the process of Inference 34
Second chapter
Components involved in the process of Inference
Components means any of the parts of which something is made. These

components when they will be arranged in a systematic manner will build a large
object. In other words, we can define a large object as a set of components, which
are interrelated, which explain it fully, accurately and with a great care. Inference
is a full process of knowing and a component means an element of that process.
Now, let us assume inference as a large object to know its set of components.
Inference is defined as the knowledge arising through paramarsa1. What is
paramarsa then? Paramarsa or confirmatory cognition is the
General knowledge of paksadharamata qualified by the knowledge of
background
invariable concomitance or Vyapti2. Again, what is this

paksadharamata? Paksadharamata is the presence of the
probans in the inferential subject or paksa3. The invariable concomitance is the
invariable association between probans and probandum4. So, we have components
like paksa, sadhya, vyapti, paramarsa, paksadharamata, etc., which are involved
in the process of inference. On the other hand, we can define or analyze inference
through this set of components. Unless we know the nature and function of these
components we won't be able to explain inference. Let us now discuss these terms
one by one at length.
Paksa:
Paksa or subject is a substratum or locus, where the probandum is supposed to be
established. This substratum may be a place, a substance, an individual or any
other thing where an inferable thing can be predicted. The paksa is defined as the
subject, in which the presence of probandum is in doubt 5 . Let us have the familiar
example of the system: 'The mountain is fiery, because of smoke'. Here, fire is
probandum, because it is going to be established. Smoke is
presence of me reason, because it helps probandum in its establishment.
l
probandum is in
Somebody may start doubting regarding the existence of fire

only after seeing smoke on the mountain. Hence, in this case the mountain
performs the role of paksa. The word paksa is found, for the first time, in
Gotama's Nydya-sutra 6 . But he means paksa as 'side' because in later sutras he
mentions terms like svapaksa, parapaksa etc. Next comes Vdtsydyana, he
interprets 'paksa'' as 'establishing one's own thesis'. He also keeps the meaning
'side' unchanged. But it is Udyotakara, the sub-commentator of the Nydya-sutra,
develops its meaning as 'the subject of proposition' . Finally, that is called paksa,
o
which possesses the property called paksatd or subjectness . What is this paksatd
then?
Paksatd:
The word paksatd is derived from paksa with 'tal' suffix, which represents the
meaning as 'the nature of paksa'. Every mountain cannot be a paksa. For that
some necessary conditions are there. Paksatd, as it is defined by Visvandtha, is the
absence of certainty devoid of the desire to infer .
4
Vide: TSA >HI^4PWHl <4lfk1: I
5
TTSTT W- W f f t c j P r e ?fcf - N B h . 1.1.33, TSA.
6
NS. andNBh. 1.1.41
7
NVTT. 1.1.41
8
TSA
The mountain possessing the property called paksata is called

Paksata, is the
absence of paksa'0'. Logicians define paksata as where exists the absence
certainty devoid of
the desire to infer 0f the knowledge of the probandum qualified by the absence
of the desire to infer11. And that which possesses such quality is called paksa. The
following four possibilities can occur in this case:
+X -> +Y or both the entities i.e. probans and probandum are
there in the paksa.
+X -> -Y or probans is there but no probandum is there in the

paksa.
-X -> +Y or provans is not there but probandum is there in the

paksa.
-X -> - Y or both probans and probandum are not in the paksa.

Among these possibilities in the last case there will be no inference, because both
the components are absent there though there is a desire to establish the
probandum.
Paksadharmata:
The presence of vyapya on the paksa or subject is called Paksadharmata12. If this
is the definition of Paksadharmata then question may arise like 'why smoke alone
is called Paksadharm ? Why not tree, stone etc. which are also there in paksa?
these are also property of mountain. The only possible answer for it is this that we
do not infer fire on the basis of anything other than smoke. This is the only
effective means. In other words, we can say that all smokes in the world, though
they are pervaded by fire, are not the property of the inferential subject. So, in a
particular context a particular thing can become the property paksadharma.
" NSM. Verse-70

12
ciiiuq^Li L|4diR<^Rid Êrfcn, TSA
Again, vyapya means that which is qualified by vyapti or invariable concomitance.

In other words, the presence of the reason on the subject, which (reason) is
Paksadharmata, qualified by invariable concomitance, is known as paksata.
is the presence of
vyapya on the As mountain is the paksa in the inference, the cognition of
paksa or subject , , . , .• . . , ,
smoke as present in the paksa is the knowledge of
'vyapya 'as a property or dharma of that mountain. That's why it has been
described as paksadharmata. This with the help of some other things constitutes
paramarsa. That is to say that pa is greatly dependent on paksadharma.
Sadhya:
When a person sees smoke on the mountain, what he infers is 'either the hill is
related to the fire' or 'fire is related to the hill'. From both the sentences he'll
sadhya predict that 'some object' is on the mountain, which is
character of the invariably related to the smoke. And that is nothing but fire,
înference ^ because of his previous experience. Sadhya is the object of
inference. It is said to be an inferable character of the subject
of the inference. This sadhya is also called Uingin', because it is related to
'linga ''3 or mark. Again, it is known as sadhya because it is sought to be proved.
It cannot be perceived directly but through its significant mark it is supposed to be
inferred. For example, 'the mountain possesses fire, because of smoke'. Here,
smoke is the indicator and what it indicates is fire, which is otherwise called
sadhya or probandum.
Hetu:
Hetu, as defined by Gotama, is the means for establishing what is to be
established through the affirmative character of the example14. It is called 'hetu'
H %cj, NS. 1.1.34

because it causes the knowledge of the probandum. It is the means of proving of

the probandum, therefore, it is called instrument or sddhana. This also has termed
as gamaka or indicator because it serves the puipose of
Hetu indicating something (i.e. probandum). Kesava Misra'5
is the means of
establishing derives the meaning of linga or mark as it is the perception of
inference
sign or mark, which leads to the inference of probandum. For

example, 'this mountain possesses fire, because it has smoke'.
In this case, we assume to have fire on it, which can not be perceived directly. But
it is smoke, which provides an adequate information for the same. It is stated
before that paramarsa or confirmatory cognition is a necessary condition for
inference. And the definition of paramarsa highlights the role played by hetu or
probans to constitute it. Therefore, the classification of different kinds of hetu
helps to understand different kinds of vyaptijhana and anumana as well. Keeping
this view in mind, Naiyayikas have categorized hetu in to three clases16. Let us
discuss the nature and function of these categories.
The varieties:
If there is '+x' then there is '+y' —anvaya (yad satve yad sattvam)
And if there is '-x' then there is '-y' vyatireka
a) Anvaya- (yadabhdve yadabhavam)
vyatireki
Unga Let us fit this in to the stock example of the system 'mountain
possesses fire because of smoke'.
It can be shown like this:
there is 'smoke' there is 'fire' - positive invariable concomitance and
where there is no 'fire' there is no 'smoke' negative invariable concomitance.
So, the positive invariable concomitance occurs between presence of

probans and probandum and the negative invariable concomitance occurs between
15
c4|fac|<^H|2fwP lel^-JIH I- TB-2
16
TSA
the absence of probans and probandum. In short, the reason, which possesses both
the type of invariable concomitance i.e. positive and negative, is called anvayi-
vyatireki hetu.
The mark, which possesses only positive invariable concomitance between the
1 -j
probans and the probandum, is called kevalanvayi linga . For example, 'the pot is
nameable because it is knowable, like cloth e t c ' Here, the
reason is knowability and the probandum is nameability and
b) kevalanvayi
they pervade each other. In this case, only positive invariable

concomitance i.e. where there is knowability there is
nemability or vice-versa can be shown. So, the invariable concomitance is called
kevalanvayi and since this mark possesses this type of invariable concomitance, it
is called only positive mark.
To clarify the term kevalvyatireki we can have an example like 'the earth is
different from other than substances, because it has smell'.
._ . ... We cannot produce the negative invariable concomitance like
r
c) Kevala-vyatireki °
'where there is smell there is from other than substances'.

Thus, the mark, which possesses the negative type of invariable concomitance, is
called kevala-vyatireki linga18.
Sad-hetu:
Again, hetu can be either a sad-hetu or an asad-hetu. That is called a good reason
or sad-hetu, which produces the true inferential cognition. Otherwise it will be
called as a bad reason or an asad-hutu. A good reason has five characteristics like
1) It must exist in the inferential subject (pahe sattvam)
2) It should not in similar instances (sapakse sattvam)
17
ibid.
18
ibid.
3) It should not exist in the dissimilar instance (vipakse asattvam)

Characteristics
of a
4) It should not contradict (abadhitatvam)
sad-hetu
5) It should not have another reason to prove the opposite

(asatpratipaksatvam)
If a reason is having such characteristics then it will be called a valid reason. It is a
valid reason because it is capable of producing inference.
And that is called a bad reason, which appears as a hetu but is not a hetu in
fact. This is technically called hetvabhdsa, which is going to be discussed later.
Vyapti:
Like pardmarsa vyapti or invariable concomitance essentially depends on two
facts viz. the vydpya (pervaded) and the vydpaka (pervader). Thus, vyapti is a
correlation of these two facts. The pervaded one covers a narrow area whereas; the
pervader one covers a wider area. In simple language, the
Formal
definition vydpaka or pervader is present in all the places in which the
of
Vyapti vyapya or the pervaded one is present. This theory generates a
universal rule like 'where there is vydpya there is vydpaka'.
The instance of smoke and fire can be considered here. Two notions can be drawn
from the above instance like 'where there is smoke there is fire' and 'whatever
possesses smoke possesses fire'. It suggests that smoke is pervaded by fire. In
other words, fire is the pervader. But the reverse order is not true. In Prof.
Wada's19 language "The invariable concomitance of smoke with fire has been
regarded as relation so far, but it can be considered to be the state of invariably co-
existing with fire. The state of invariably co-existing with entity, smoke or fire,
exists in smoke and not in fire. This is because fire does not possess the state of
invariably co-existing with smoke, as in the case of the fire of a red-hot iron ball."
19
ICNN, P-35
Invariable concomitance plays a very important role in the process of inference.

Without a universal relation between two facts (i.e. vyapya and vyapaka) no valid
conclusion can be drawn from the premises . Now, there is a disagreement
between the logicians regarding the function of vydpti. Some
sa tnat
procesTof * ^ ^ funcuons as an
instrument of the process of the
inference21 whereas, some hold it as a component involved in
the process of inference. Again, vydpti , based on unequal extention, such as
smoke and fire, is called vishama-vydpti or asama-vydpti. It is called so because
there is a relation of non-equipollent between two terms like vyapya and vyapaka,
from one of which we infer the other but not vice-versa . For example, smoke and
fire. But a vydpti, which possesses equal extension between two terms, is called
samavyapti or equipollent concomitance. It happens to see between cause and
effect, substance and attribute, etc. and also in vice-versa. For instance, 'whatever
is produced is non-eternal' and 'whatever is non-eternal is produced'.
Paramarsa:
Pardmarsa or confirmatory cognition is produced with the help of two
technical terms viz. Vyaptivisista and paksadharmatajnana. Independently these
two terms have their respective meanings but when they come altogether they
generate a third kind of knowledge, which is nothing except pardmarsa. So, the
nature of pardmarsa will be clarified when the nature of its components will be
Paramarsa, clearly understood. The compound Vydptivisista-
the knowledge of
Vyaptivisista-hetu paksadharmatajnana can be dissolved like Vyaptivisiste
in the paksa
paksadharmatajnana (locative tatpurusa). It means, 'that

paksadharmatajnana which is Vyaptivisista (the knowledge of the property of the
subject which is qualified by invariable concomitance). Again, Didhitikara23
20
NTK, p-252
21
Mtmamsakas and Buddhists hold vyapti as an instrument in the process of inference.
2
~ Navya-Nayiyayikas say so.
23
Vide. NBn-2
dissolves the compound like as 'Vydptivisistum iyyaptiprakarakam) vat

paksudharmatajnanam iti'. In other words, lyat paksudharmatajnanam tat
Vyaptivisistam' (that the knowledge of the property of the subject is qualified by
invariable concomitance). Here also the idea has been kept unchanged.
Nyayahodhini24 considers the compound as karmadharaya keeping the meaning
unchanged.
All the above definitions, given by various logicians, convey the same meaning.
Hence, the conclusion can be drawn like 'vyupti is the visesana of the hetu'.
Therefore, paramarsa can be defined as 'the knowledge of Vydptivisistu-hetu' in
the paksa.
In the of process inference, paramarsa functions as an immediate cause. Of

course, there is a controversy over this issue, which will be discussed later.
According to Nuiyayikas this confirmatory knowledge functions as an immediate
cause in the process of inference. This can be explained as follows:
A person, who has the experience of invariable concomitance of smoke and fire,
sees a string of smoke on the mountain. And all of sudden he
u Ion
o-. recalls the relationship which has been experienced in the
paramarsa kitchen etc. previously. Then he realizes like 'this (mountain)
smoke is invariable associated with fire like kitchen'. This is called paramarsa or
confirmatory cognition. Immediately after that the inferential cognition like 'this
mountain possesses fire' arises.
Finally, whether something is an instrument or function of an instrument

depends on the notion of an instrument (karana).
ibid, as in Nyayukosa, p- 475

Hetvabhasa:
Paramarsa generates inference and hetu is the ultimate cause of paramarsa. A bad
reason produces an invalid paramarsa and invalid inference ultimately, which is
not desired. Therefore, logicians have tried to find out the characteristics of a bad
reason. This is, in logical terminology, is called hetvabhasa. The compound
hetvabhasa has two terms, hetu and abhdsa. Hetu in context
t is not a Q f m f erence m e a n s 'ground or probans of inference' and
r
real probans °
but appears abhdsa literally means 'appearance or fault'. Thus, the term
hetvabhasa signifies either a 'seeming ground' or 'fault of a
proposed ground'. These are the possible interpretations can be made out of the
term hetvabhasa. Maharsi Gotama25 has not mentioned any definition of
hetvabhasa. But it is Vatsydyana26 who has defined hetvabhasa as 'those which are
not real probans because of not possessing the real characteristics of a probans, but
which appears as a probans because of their similarities.
This definition of hetvabhasa matches with the first definition given above. There
is a difference of opinion regarding the types of hetvabhasa among the old
97
logicians. Gotama mentions the number as five. Vatsydyana also accepts it. But
it
is Udyotakara, who extends the meaning of the sutra little
Its
further. Vdcaspati Misra, the commentator of Nydya-sutra-
varieties
bhdsya-vdrttika, also agrees with him. On the other hand,

neo-logicians also accept five types of heh'dbhdsa. Acarya Gangesa, the author of
tattx'acintdmani, classified hetvabhasa in to five and later logicians also have
accepted it. Finally, five types of hetvdbhdsas have been accepted in the system.
Those are:
1) Savyabhicdra hetvabhasa:
Gotama mentions the name only in ~Nyaya-sutra - 1.1.1
2) Birudha hetvdbhdsa:
3) Satpratipaksa hetvdbhdsa:
4) Asiddha hetvdbhdsa:
5) Bddhita hetvdbhdsa:
1) Savyabhicdra hetvdbhdsa:
Vyabhicara means 'non-residing only in one place' or 'inconstancy' and
savyabhicara means 'that which exists with vyabhicara'. The very nature of this
fallacy or hetvdbhdsa is it involves contradiction of probans with that of
prabandum. For example,
proposition: sound is eternal
Savyabhicara reason: because it is intangible
hetvabhasa: example: whatever is intangible is eternal, as atoms
application: so the sound (so the sound is intansible)
conclusion: therefore, sound is eternal.
Again
proposition: sound is non-internal
reason: because it is intangible
example: whatever is intangible, is non-eternal, as intellect
application: so the sound (so the sound is intangible)
conclusion: therefore, sound is non-eternal
Here, intangibility is present in both the subject. There is no invariable
concomitance between 'intangible' and 'eternal' or 'non-eternal'.
a) Sddharana hetvdbhdsa
It is called sadhdrana or common because it is found to be present in both the
27
NS- 1.2.4
places i.e. where the probandum is present and where

a) General definition the probandum is absent. For example, 'the mountain
has fire, because it possesses smoke'.
Here, fire is the faulty probans because it is found in both the cases i.e. where
there is smoke, kitchen etc. and in the hot iron rod where smoke is not present.
b) Asddhdrana hetvabhasa:
The collocatedness of probans and probandum with certain rules and regulations
formulates the knowledge of invariable concomitance. The very nature
as asddhdrana-heh'dbhdsa is it is neither present in
Definition similar instance (sapaksa) nor in dissimilar instance
(vipaksa). It resides only in the subject . This is just
opposite to the Sadharana hetvabhasa.
For example, 'sound is eternal, because it possesses the property called
soundness'. The soundness is the reason here, which exists only in the similar
instance (sapaksa) but not in the contrary instances (vipaksa).
c) Anupasamhdri hetvabhasa:
It is defined as the 'hetu' devoid of any supporting instances, both positive and
negative, is called anupasamhdri.
'where there is '+x' there is '+y' (The positive instance ) and

'where there is '-x' there is '-y' (the negative instance)
Here, subject includes all the objects of the reason and the relation of probans and
probandum can not be found out either in positive instance or in negative instance.
Anupasamhari For example, 'all objects are non-eternal, because they are
hetvabhasa-. knowable'. Here, 'everything' is the subject. The sadhya is
non-eternal and knowability is the reason.
NSM, anumdnakhanda
Here, we cannot find any positive or negative invariable relation between probans
and probandum, because the subject 'everything' all the subjects. Nothing remains
there outside of the subject to be taken as an example.
2) Biruddha hetvdbhasa:
The contradictory fallacy has been defined as 'the reason, which opposes what is
to oe
Biruddha established'. For example, 'sound is eternal, because it
hetvdbhasa: i s produced'. Here, probans does not support to the
probandum rather it goes against it.
In other words, it obstructs to the knowledge of correct inference.
3) Satpratipaksa hetvdbhasa:
'Sat' means 'being presented (somewhere)' andpratipaksa means 'counter
thesis. So, that which is present in both the theses i.e. in the similar thesis as well
as in the counter thesis is called satpratipaksa. Let us have an example: -
Similar thesis:
Satprariaksa ,_, . ,,
s rout
hewabhasa- The P etc. are produced by an conscious agent,
because they are of the nature of effect, like etc'
Counter thesis:
'The sprout etc. are not produced by an conscious agent, because they are not
produced by any embodied being, like sky etc.'
In these examples, neither of two probans is capable of establishing the
proposed probandum.
4) Asiddha hetvdbhasa'.
The mark, which is being yet to be properly established ins not different
from the characteristics of the probandum. It occurs when the middle term is
wrongly assumed in any of the premises and so cannot be taken to prove the
conclusion.
For example,
Asiddha to, , • , .
S h a d o W 1S a
hetvdbhasa: Substance,
because it is characterized by the movement.'

This again is of three kinds such as:
Asraya means the resting place but in the context of inference it is the
a) Asayasiddha inferential subject Asrayasiddhi means the
hetvdbhasat < • , • . • . . .
subject is not known or non-existent.

For example, 'Sky flower is fragrant'. Here sky flower (the subject) does not exist.
Svarupa means nature. The very nature of this hetu is that it is non-existent in the
..... inferential subject. For example, 'sound is a
J r
b) Svarupasiddna '
hetvabhdsa' quality, because it has visibility'. Sound

possesses the property of hearing but not visibility. Therefore, visibility does not
exist in the subject.
The compound means 'that in which Vya/7yatva' is non-existent'. Again
'vyapyatva' means that which is qualified by vycipti or invariable concomitance.
In this case, there is no such type of relationship
c) Vydpyatvasiddha
heh-abhasa: between the probans and the probandum. The
peculiarity of this fallacy is that it requires a condition {upadhi). That which is
pervaded b by the probandum but not the pervader of the probans is called a
condition. For example, 'where there is contact of wet-fuel with fire there is
smoke'. Here, the contact of wet fuel is a condition, which does not permit vydpti
to arise.
5) Bddhita hetvdbhdsa.
'Fire is cold, because it is a substance'. Here, two pramanas are arising
simultaneously i.e. perception and inference.
Difinition
And it is proved that perceptual cognition is

stronger than inferential cognition because the later one is dependent on the
former. Therefore, the inference becomes weak. So, the absence of the probandum
is pervaded by another strong means of cognition is called contradicted fallacy.
* * * -6S- *p * *T*
Selected Readings for Second Chapter.
The Nyaya theory of knowledge
-Prof. Satischandra Chatarjee, University of Calcutta, 1950
Materials for the study of Navya-Nydya
-Prof. H.H. Ingalls, London, 1951
Anumana pramana
-Dr. Baliram Shukla, Eastern Books Linkers, Delhi, 1986
Indian Logic
-Dr. B. N. Singh. Asha prakashans, Varanashi, 1988
Invariable concomitance in Navya-Nyaya

-Prof. Toshihiro Wada, Sri satguru publications, Delhi, 1990
NS Nyayasutra
NBh Nyayabhdsya (Vdtsdyana)
TSA Tarkasagraha, anumanakhanda
NV Nyayavarttika
NVTT Nyayavarttika tdtparyatlkd
NSM Nyayasiddhdntamuktdvall
NSV Nyayasutra vrtti
NBn Nydyabodhirii, anumanakhanda
Di.A DidhitT, Anumanakhanda
NTK Nyaya theory of knowledge
ICNN Invariable concomitance in Navya-Nyaya
TBA Tarkabhasd, anumanakhanda
TCsnp Tattvacintamani, samdnyaniruktiprakarana
(Because due to the knowledge of particular smoke, there will not arise that
type of doubt because inferential cognition is already present there.
Otherwise, if all the factors required for definite knowledge are present
there then why paramarsa should not take place?)
Chapter - III
Theory of inference according to the Mimamsakas
• General background
• The nature of inference
• Kumarila theories of inference
• Prabhakara theories of inference
• Kinds of inference
• Constituents of a inference
• Conditions of a valid inference
• Observations
Theory of inference according to the Mimamsakas 48
Third chapter
Theory of inference according to the Mimamsakas
Every human behavior has two aspects such as: he thinks and he argues. He thinks
because he observes. Again there are two means of human observation; he closely
observes whatever he gets before him and he employs same knowledge, which is
obtained from close observation, for the remote objects. So,
direct perception and indirect perception are two means of
General knowing. Again, he argues because he tries to convince
background
others about his own knowledge of truths. He tries to justify

his own feelings in others. To do so, he develops a particular
method or system of reasoning, which is otherwise called logic. To do so, he
develops his ability to reason correctly. He seeks help of a systematic method of
obtaining knowledge. This systematic method, through which he obtains a valid
knowledge, is called pramana and the science of its application is called
pramanasdstra or Epistemology.
The Indian philosophical systems have a serious dealing with this part. Mimamsa
system, in this connection, is in no way exceptional. The very existence of things,
according to Mimamsakas, is two fold such as: means of knowledge (mdna) and
objects of knowledge (meya)1. Here, this chapter deals with the means of
knowledge according to Mimamsa system. Before doing so let us discuss what
knowledge is? And how it functions? The function of knowledge, according to
Mimamsa school of Indian philosophy, is to illuminate the things other than itself2.
Knowledge, as it is discussed in Mimdmsd system, is of two kinds valid and

invalid. Valid knowledge is the knowledge of an unknown real knowledge3. The
means of valid knowledge or pramanas, according to Bhatta school of purva
Mimdmsd, is a knowledge which does not fail to recognize4 whereas, Prabhakaras
hold pramdna as 'valid knowledge' (but not the means of valid knowledge). He
opines that pramdna or valid knowledge as 'experience, which is totally different
from remembrance'5. However, the numbers of pramdna, according to Kumdrila
Bhatta, is six whereas, according to Prabhdkara it is only five6. Those six
pramanas are Perception, Inference, Similarity, Analogy, Presumption and
Negation.
Among these pramanas, inference or anumana as defined in Sahara Bhasya is 'a

cognition of an object not in direct contact and forming one of the two sides to a
D - . f known relation, one of which is directly apprehended'7.
Let us analyze the above definition of inference given
by Sabarasvdmin. This definition contains three necessary conditions. Those are
asannikrste buddhih, jhatasambadha and ekadesadarsana.
Asannikrste buddhih means, 'the knowledge of an object, which is not in direct

contact with sense organs'. The term ''jhdtasambadhasmarand' indicates that there
should be a known relation between two objects, which are invariably related with
each other.
Ekadesadarsana means 'of which one should be directly apprehended'. It means

out of those two objects, which are invariably connected with each other, one
3
MM - verse no: 1
4
NR on SV, verse. 44 3jR>Hcllte ftsTR WHI"IH I
5
PP, p. 42 MHIuHjTjfrTT I
6
MM - verse no: 15
7
SBh. 1.1.5 3TyTpf ^ I d W - ^ r ^ ^ ^ K ^ l ^ ^ i M ^ H f a c ^ ffer |
should be seen from a distance. In simple language, ekadesadarsana refers to

vyapya darsana.
Now, let us organize this into a sequence.
First: ekadesadarsana (vyapya darsana)
Then: jhatasambadhasmarana (vyapti smarana)
Finally: asannikruste buddhih (anumiti)
An example like 'this mountain possesses fire because of smoke' can present a
clear picture of this format. According to Mimamsa system, smoke is ekadesa
(vyapya), because it is pervaded by fire. The word jhatasambadha indicates to a
known relation (say, relation of smoke and fire) i.e. 'where there is fire there is
smoke'. So the process of inference in this system appears like this:
Dhuma darsana
Vyapti smarana
Anumiti
There is no scope for 'application' on paramarsa here as Nyaya holds.

Mimamsakas hold 'buddhi' or knowledge as inference. From the above process of
inference it is clear that Mimamsa system does not require the third component
called paramarsa at all. There will be a discussion at length over this issue in
seventh chapter.
This definition of inference as given by Sabarasvamin, contains two
important terms like jhatasambadha and asannikrstabuddhi. These two terms have
been interpreted by later Mimamsakas in various ways. Let us discuss them one by
one.
Four alternative interpretations of jhatasamhadha given by Kumdrila:

Kumarila Bhatta, the author of Slokavdrttika, derives four possible alternative
meanings of the term jhatasamhadha. The first interpretation, as he suggests, is
pramata jhatasambadhah . Jhatasamhadha means the knower, who knows the
invariable relationship of probans and probandum i.e. between smoke and fire.
The compound has been dissolved by him in two ways such as: bahuvrlhi and
karmadharaya. When it is bahuvrlhi it expresses the meaning like jhdta
sambhadhah yena sah (the knower, who already has experienced the
sambandha). When it is karmadharaya it expresses the
meaning like jhdtah sambadhah yatra10 (where the
interpretation sambadha has been experienced i.e. paksaekadesa).
Thus, the form of inference, according to his first
interpretation, follows like this: when such a person {pramata, who knows the
invariable concomitance of smoke and fire) happens to see smoke from a distance
(somewhere in the mountain), then he remembers the invariable concomitance
and immediately the inferential concomitance i.e. this mountain possesses fire
arises in him.
His second interpretation suggests that the compound may be explained as

referring to the substratum (ekadesin), where the relationship of smoke and fire is
11 19
apprehended . Here, jhatasamhadha stands for similar instances (sapaksa) .

There are many places where smoke and fire are seen together like kitchen etc.,
And the most common place like mountain etc, where after observing smoke the
fire is inferred. The former one is called paksa whereas the later one is called
sapaksa.
8
SV-2, verse-3
9
Ibid, verse 3-4
10
This is also Bahubrihi
The paksa and sapaksa both are ekadesin (the instance) whereas smoke and fire
both are ekadesa (the relata). To fit the present context we may quote
Parthasdrathi Misra's definition of inference. Inference, according to him, is an
s . ekadesa (for example, smoke), of a sapaksaikadesin
interpretation ^ Q r e x a m p i e 5 kitchen), where invariable concomitance
of smoke and fire has already been recognized. And when it will be perceived in
the paksaeikadesin (i.e. mountain) there arises the cognition of another ekadesa
and this later cognition is called inference'13. Therefore, the form of inference on
the basis of second interpretation is 3rg*TFT ^msr^j ^^KÎ^CÎM^ arof^f^ ^fer.
The third alternative interpretation of Kumarila Bhatta takes a new turn. Here, he
derives the meaning of the compound jhasambadha as a 'known relationship'. He
dissolves the term with the help of karmadharaya samasa . And the word
ekadesa refers to a member of this relationship. The
Third
interpretation relationship only can occur between two terms and
members. Here, the members are smoke and fire. So, the inference, according to
his third interpretation, is the cognition of fire (say, second member) of a known
relationship holding from the perception of smoke i.e. first member.
The fourth possible interpretation of the compound 'jnatambadha' refers both

linga and lingin, whose relationship is known15. That which includes another thing
is called linga and lingin is that whose presence is inferred by linga.
Fourth Smoke always indicates the presence of fire therefore
it is linga and fire is being indicated by smoke
13
Ibid.
SV-2, verse 3,
Theory of inference according to the Mimathsakas 53
thereforeit is called lirigin. Hence, inference occurs when on the perception of one
part (smoke) of a logical whole, the cognition of the second part (fire) arises from
a known relationship i.e. invariable concomitance between two parts16.
Prabhdkara theories of inference

On the other hand, Prabhdkara Mimdmsaka, another commentator of Sahara
Bhdsya, interprets the compound jndtasambadha as a 'known relation'. So, the
inference, in the words of Prabhdkara, is 'a knowledge which arises of that which
is not in contact with sense organs and which arises after the perception of one
17
part' . Smoke is such an object whose relationship is established with fire and
which takes place somewhere in the mountain. The knowledge of fire, which
arises from the perception of the same, is not in contact with the sense organs.
The interpretation of the word asannikrsta

Further, both the Mimdrhsakas have come forward to define the word
Kumarila's asannikrsta. According to Kumdrilabhatta, the word
interpretation . .
asannikrsta, which literally means not in contact with

the sense organs', is non-determined. By giving the meaning 'non-determined' he
means to exclude two things such as: 'the detennination of the inferable object in
that form' and 'the determination of that in the contrary form'18. For example, the
inferential cause like fire etc. According to him, if the very existence of inferable
object i.e. fire will be previously determined through sense perception etc. then it
would be a restatement here (i.e. in the case of inference). Similarly, it would be
16
SV-2, verse 3, ~5§ cJT s I l d ^ - t ^ M d d c j qWFW I
17
Brhati, 1.1.5 3 J ^ H H ^ ^ I ^ i H I ^ ^ I M ^ 3JxHfacâf ffer
18
SV-2 verse 55 3^fa<£t^|rj| ^ £W5I fa^lfadH I
d l < ^ I J ^ d d d^Mîldlsft ^ I I
non-valid if it is determined as non-existing. Therefore, he has explained the term

as non-determined.
On the other hand, Prabhdkar Mimdmsaka, explains the term in a different way.
According to him, the word \sannikrsta' means
interpretation remembrance , and asannikrsta means that which is
different from remembrance. Because this can be a
cause to dust etc. As a matter fact, Sdlikandtha, the author of Prakaranapahcikd,
explains the word keeping the meaning unchanged .
A comparative study of both the schools provides some notable observations. The
way Kumarila explains is little complicated one. He gives all possible
interpretations, from various angles. On the other hand, Prabhdkara 's definitions
are seem to be simpler and to the point. Hence, the later one seems to be more
approaching.
Kinds of inference:
First type
The tradition of classifying inference into svartha and parartha, perhaps, has been
started from Dignaga. Pre-Dignaga school of Indian philosophy has nothing to
say in this regard . However, there is no reference of such division either in
Jaimimya sutra or in Sahara's commentary. But in Manameyodaya the inference
has been divided into two such as: svartha (inference for oneself) and parartha
(Inference for others).
l9
Brhati, 1.1.5, ^ c p ^ m ^ % w V f a W^\
20
srcrfacp^ ûnf^HHf^mi^Pk^TfrH I - f ? m ^ elites "ft on PP, C ^ H U M R ^ , p. 196
21
EBPM, p. 254
Theory of inference according to the Munaiiisukas 55
Inference for oneself (svartha) occurs when fire etc. is inferred after seeing smoke
etc. through the remembrance of invariable concomitance22. So, this process needs
three steps.
Hetn darsana (perception of smoke somewhere in the
mountain)
Vyaptismarana (remembrance of invariable
Svarthanumana
concomitance of probans and
probandum)
Anumiti (the knowledge 'that there is fire on the
mountain).
A close observation can direct that the definition given by Sabarasvamin and that
of present one is in no way different.
The inference for others, as it is defined in Manameyodaya, is 'some object (some

inferable object i.e. fire etc.) are known through the words of another'. This is an
inference of statements. Here, somebody after experiencing for himself tries to
convince to somebody else through some statements. Those statements, which are
known as syllogism, are like this:
Proposition (pratijna) 'mountain is fiery'
Pararthanumana Reason (hetu) 'because there is smoke'
Example (udaharana) 'where there is smoke there is fire
as in the kitchen'
Here, a further discussion continues regarding the numbers of avayavas.
Narayana, the author of Manameyodaya, criticizes Naiyayikas saying 'we, unable
to put up with repetition, speak of three members, either ending with the example
or beginning with the example' . In either way they accept only three-membered
MM, p. 63
MM, p. 63
syllogisms. We have seen the statement ending with example. Now let us see the
proposition beginning with example.
Example (uddharana): where there is smoke there is fire

Application (upanaya): mountain possesses smoke (which is pervaded
by fire)
Conclusion (niganama): therefore, the mountain is fiery.
Mimamsakas hold that by premise the purpose of the conclusion and by the reason
that all the application are served. Then why unnecessarily five statements will be
accepted?
Second type:
A further division of inference is available in the commentary of Sahara svdmin
on Jaiminlya MTmarhsa sutra . He divides inference into two like
pratyaksatodrsta sambdha and samanyatodrsta sambadha. The definition of these
terms has not been defined but has been illustrated by
pratyakshodrsta
him. The commentator explains pratyaksatodrsta

sambdha as 'from the perception of smoke the form of fire is inferred' . Because
it is based on the relationship, that which is directly perceived. The relationship,
which has been experienced previously somewhere in the kitchen etc. After
remembering invariable relationship between smoke and fire, whenever,
somebody sees smoke somewhere on the mountain, a certain knowledge arises in
him like 'this mountain is fiery'. This is what is calledpratyakshodrstam.
The inference that based upon a generalized relationship has been illustrated by
Sahara svdmin like this: one can infer that there is motion in the sun because it
:J
SBh, 1.1.5
changes its place like other movable objects like Devadatta etc' 26 . Here, the
movement of the sun cannot be perceived directly but can be
inferred from its motion (i.e. changing place from time to
Samanyatodrsta
time), bo, irom the above instance we can

formulate a general statement like: 'whenever an object changes its place it has
motion. This instance is based upon a known fact. Therefore, it is called inference
with generalized relationship or samanyatodrsta sambadha.
Kumarila Bhatta's interpretation on second type:

The above division of inference created a room for Kumarila Bhatta to raise an
objection. He opposes it saying 'there is mutually exclusiveness between these two
classes of inference'27. He finds no difference. The relation of
smoke with fire is not directly seen. So, all so the case of
The above
division carries changing position and movement. On the other hand, if
no value
somebody claims saying that though the relationship is not

directly seen but it has been experienced directly in kitchen etc. we reply them like
the relation between change of position and movement is also cannot be seen
directly but it can be assumed from the activities of Devadatta. Hence, both the
illustrations can be applied in samanyatodrsta .
He, therefore, rejects the term pratyaksatodrsta and puts visesyatodrsta in its
place. According to Kumarila, a division must possess some unique
characteristics. But it does not happen in case of pratyaksatodrsta and
samanyatodrsta, because visesa is opposite to samanya but not to pratyaksa.
26
Ibid.
27
SV, verse 138
28
Ibid, verse 140
Theory of inference according to the Mimarhsakas 58
Visesyatodrsta or seen by particulars type of inference, according to Kumarila

Bhatta, is based on an invariable relation between two
Visesyatodrsta particulars. It has been illustrated like this. A person observes
a particular fire produced out of some fuel like cow dung etc.
with its effect that is smoke. And one day he happens to see a particular string of
smoke somewhere in the mountain then he remembers the relation between the
two particulars and then he concludes that a particular fire which exists on
mountain, which is not in contact with our sense organs. It is called so because it
is based upon a particular relation of a particular perception29.
On the other hand, Samanyatodrsta is knowledge of the relation by perception of

two general objects. Kumarila keeps the idea unchanged with
Samanyatodrsta Sahara in this connection30. However,
samdnyatodrstanumana deals with the class of objects.
The example is the familiar inference of the fire with smoke in the hill.
Prabhakara Misra's interpretation on second type:

Prabhakara Mimarhsakas explain Sahara's division of inference in a different
way. He classifies inference on the basis of the self-property of the object of
inference (sddhya). Sometimes the self-property of the probandum is perceptible
and sometimes it is imperceptible.
When it is perceptible it is called drstasvalaksana and when it is not perceptible it

is called adrstasvalaksana. So according to him, inference is
Drstasvalaksana '" ' "
and
of two kinds that is drstasvalaksana and adrstasvalaksana .
Adrstasvalaksana
The self-property of the object of inference that is fire-ness is
Ibid, verse 140-142

Ibid, verse 145
Theory of inference according to the MTmathsakas 59
perceptible. That is why this is the example of the first category32. But in the case
of second one the self-property of the prabandum is imperceptible. For example,
action or movement and potency (sakti)31'.
Third type:
Third type of classification of inference also has been accepted by the system. Of
course, Sahara Bhasya has nothing to say about it. Prabhakaras do not say
anything about it also. Kumarila Bhatta also does not mention this classification.
But it is Narayana, the follower of Bhatta school of Mimarhsa, who classifies
inference into three like anvaya-vyatireki, kevalanvayi and kevela-vyatireki34. This
classification is based upon the invariable concomitance.
The first one is kevalanvayi linga or purely positively concomitant. The existence
of Prabandum where there is the existence of probans is called kevalanvayi vyapti
„ .. or positive pervasion. And inference, which possesses
r r
Kevalanvayi
anumana anavayivydpti is called kevalanvayi or only positive
35
inference . For example, the knowledge manifested by another knowledge,
because it's knowledge like jar etc. Here one cannot produce a negative example
because all the objects can be illustrated by knowledge. As it is impossible to
produce a negative example of the inference hence it is called Kevalanvayi
anumana.
A kevalavyatireki vyapti or only negative perversion occurs there where the non-
existence of probans is seen in case of non-existence of the probandum.
32
ibid.
33
ibid.
34
MM, p. 53 dx^l^MM fofasj 3Mi|ot|£Afo> ^ c i W f t ^ T ^ Z l f ^ f % ^ |
35
MM, p. 53 Wr-T^T ^ W \ ^TTOWft K T O SMiMlfkl: I
Theory of inference according to the Munamsakas 60
And the inference, which possesses only negative invariable

kevala-vyatireki . , , ,. _ ... .
concomitance or kevalavyatireki vyapti, is called only
negative inference . For example, every cognition is self-
manifesting because it is cognition. Where there is no self-manifestedness there is
absence of cognition-ness. For example, say a pot. Here it is not possible to
produce a positive example i.e. what is self-manifesting is also a cognition.
Therefore, it is called only negative inference.
That is called an inference of both positive and negative or anvayavyatireka

anumana, which contains both the type of invariable concomitance37. In other
words, the inference containing both the positive and negative
Anvayavyatireka pervasion is called anvayavyatireka. For example, the
anumana
mountain possesses fire because of smoke. We can have it in

both the ways i.e. 'where there is smoke there is fire' - positive pervasion and
'where there is no smoke there is no fire' - negative pervasion. And the
combination of both the pervasions generates a third one called positive and
negative pervasion as well.
Among these three types of inferences, the followers of Kumarila, however, do not
accept the only negative inference38. And in its place they put the fifth variety of
valid knowledge called anupalabdhi.
The constituents of inference

The key component of the process of inference is vydpti because it brings together
two things i.e. vyapya and vydpyaka in a certain rule. Again, vydpya and vyapaka
exist on a locus called paksa. The validity of inference is

Nature and totally dependent on a good hetu. So, in this context,
function of the
components hetvabhasa will be discussed. So, we have components called
paksa, sadhya, hetu, vydpti and hetvabhasa. On the other
hand, 'avayavas' which are engaged by the system to generate inference in others,
are also supposed to be discussed here. Now let us define these components first
as it is discussed in Mimamsa system.
Paksa, otherwise called ekadesin, has been defined by Mimamsakas as 'that which
is qualified by the property (namely sadhya), whose very existence is doubtful39. It
is called ekadesin because it possesses a property called
paksa
'ekadesi' ('ekadesah asti hi ekadesin'). This is a substratum
or locus of both vydpya and vyapaka. Vydpya, which is already known, and
vyapaka, which is to be known through vydpya. When such situation will arise
there that will be considered as paksa. But when the attribute (say, fire) does not
make one to be desired to know then what is qualified by such attribute will not be
considered as paksa40. In their terminology it is called paksdbhdsa or a fallacious
subject (which is going to be discussed later).
Hetu:
Mimamsakas define hetu or probans as 'that which is pervaded and having either
instrumental attribute' . So, which is pervaded by the probandum and having an
instrumental attribute is called a probans or reason.
It is also called 'gamaka' because it leads to knowledge of something, which is not

directly seen. It is a sign or indicator of the thing that which is going to be
39
MM, p. 66 fasllRid^RlRl^^ *TST ^ M I
40
Ibid. cZTRTT W T W f f %?£ |
41
MM, p. 59
inferred. Hence, it is called linga. It is called, on the other

hand, a hetu because this is the only cause of the
of hetu establishment of the probandum. Because it is the only means
of proving the probandum so it is called as sadhana. It has
another epithet called niyamya also. In Mimamsa system the invariable association
of vyapya and vyapaka is called niyama. In this case, vyapya (say, smoke) is
niyamya and vyapaka (say, fire) is lniyamaka'. The former is being controlled by
the later . However, on the basis of reason or hetu the invariable object is
supposed to be established.
Its variety:
Mimamsakas also have classified reason into three varieties like: anvaya-vyatireka
linga or positive negative mark, kevaldnvayi linga or only positive mark and
kevalavyatireki linga or only negative mark43. Let us discuss the nature and
function of these marks.
Mimarhsakas, in this regard, are in no way different from that of Naiyayikas. The
reason or mark, which possesses both the type of invariable concomitance, is
Anvaya-vyatireka called anvyaya-vyatireki-Uhga. This reason or mark h a s
llA a:
& named after 'anvyaya-vyatireka-vyapti' because the nature of
hetu is dependent on vyapti. Here, anvyaya-vydpti or positive invariable
concomitance is the invariable concomitance between probans and probandum. In
other words, where there is presence of probans there is presence of probandum.
Again, vyatireka-vydpti or negative invariable concomitance occurs between the
absence of probans and the absence of probandum. We have an example like
'mountain possesses fire, because of smoke'. Here, smoke has both the invariable
4i
EBPM,p. 214
MM, amimanaparichheda
concomitances i.e. positive and negative. Therefore, it is called positive and

negative mark.
The mark, which possesses the symptoms of a positive invariable concomitance, is

called a positive invariable concomitance mark. Here,
Kevalanvayi Unga: the opposite site of positive invariable concomitance
cannot be observed. For example, 'the pot is namable
because it is knowable, like cloth etc' The only positive invariable concomitance
like 'where there is knowability there is namebility'44 can be produced out of this
example. But the other part cannot be experienced. Hence, it is called only
negative mark.
The mark or Unga possessing only negative invariable concomitance is called only
negative mark. Here, we can't produce any positive invariable
lnga:
concomitance. For example, 'the Earth is different from other
substances, because it has smell'. Its opposite example cannot be experienced.
Here, smoke possessing only negative invariable concomitance; hence it is named
after it.
Characteristics of a sad-hetu:
It has been mentioned that the validity of inference is totally dependent on the
validity of the reason. In other words, a good reason or sad-hetu has the
potentiality to produce a valid inference. Therefore, logicians are very conscious
to determine whether the hetu is true or fallacious. Mimamsakas have determined
characteristics of a good reason as well as that of a bad one.
44
Ibid.
Theory of inference according to the MTmathsakas 64
It is mentioned in Mdnameyodaya that the positive and negative concomitance

has five aspects such as: being the attribute of the subject (paksadharmatva),
existence in the co-subject (sapaksesattvam), non-existence in the counter-subject
(vipaksadvaydvrttit\'am), un-contradictory {abddhittvam) and having no counter
probans (asatpratipaksattvam).
The first one suggests that the reason should be present on the inferential
ground only. For example, the hill is fiery, because of smoke'. Here, the reason i.e.
smoke has to be present otherwise a valid inference cannot be drawn at all.
The second aspect of a good reason is that it is supposed to be present only

in co-subject or similar instances46. For example, kitchen etc., where the smoke
exists along with fire.
The third aspect of a good reason is its absence in counter subject47. For
example, a big lake etc., where the non-existence of probandum is not known.
That is to say that in a lake etc. we cannot find the presence of probandum and
probans as well.
Its fourth aspect is that it must be un-contradicted by some other strong

proofs. For example, if it is said that fire is cold, because it is a substance', then
there will be a contradiction. Because, the coldness of fire is not experienced.
Hence, there should not arise any contradiction.
Non-existence of an opposite probandum is having no counter probans. In

other words, there should not be two equal strong reason i.e. the absence of
Ibid.
Theory of inference according to the Mimdriisakas 65
probandum, which is related to the reason should not be contradicted, called a

positive invariable concomitance mark.
However, a mark or reason possessing both the concomitance i.e. positive

and negative, has the above five aspects. The only positive mark has only four
aspects, because it has non-existence in the counter subject. The only negative
mark has also four aspects, because it has non-existence in the similar instance.
Therefore, only positive and negative mark can satisfy all above mentioned five
conditions.
Characteristics of a bad reason have been discussed in later part of this
chapter.
Vyapti:
The term of vyapti is not found either in Mimamsa sutra or in Sahara bhasya. He
only mentions the term 'sambandha' in the commentary, which, in later period,
has been explained by the sub-commentator as vyapti48.
Kumarila Bhatta says isambandha' is nothing but an
invariable concomitance between linga and lihgin. Liriga or
in the language of yyapya is one which appears either in all cases of the vyapaka
Kumarila Bhatta
or in some cases, but vyapaka or linga exists in all cases of

vyapya49. In other words, where there is probans there is probandum but not the
vice-versa. However, he holds invariable concomitance as a necessary condition of
the process50. According to him, the means of invariable concomitance is the
experience of co-instances in different times and places51. It is the repeated
observation of the existence of smoke where there is fire, which formulated a
universal relation. This relation is nothing but vyapti. He further tells that vyapti is
48
SV, verse 31/2, ^F^TGTTT^rf^T^ f e l ^ w f e l RHR?'HI
49
Ibid, verse 6, ^R ^ZTRPjf^S^ srrqcj^FS ^J^ti I
50
Ibid.
Theory of inference according to the Mtmarhsakas 66
a necessary condition generally between two universals, but sometimes it is a

relation of two particulars. The example for the first case is smoke and fire,
whereas, Krttika and Rohini is the example of the second case.
Now what Prabhakara school of Pun'a Minmamsa holds is that a single

observation (sakrtdarsana) is enough for the knowledge of vyapti. He entertains a
condition (udddhi) saying 'repeated observation {bhuyodarsand' or togetherness
of hetu and sadhya is useful in showing it i.e. upadhi. Prabhakara holds vyapti as
a niyama, which is a natural relation. "Niyana", says Govinda P. Bhatta, "is the
empirical law of invariable association of two things on the basis of which the
knowledge of one of the things leads to the knowledge of another. This empirical
law may imply any kind of relation between two things and the inference based on
it is strictly in accordance with the past of the relation" . However, this natural
relation means the relation without any condition (upadhi).
Prji h h 3 Ic 3 v 3' s
interpretation Condition or upadhi is that which is said to be another reason

to lead probandum directly53. In short, the relation between
fire and smoke is a conditional (aupadhika), whereas, the relation between smoke
and fire is natural (anaupadhika). The former one needs another factor (say, wet
fuel), which is the pervader of the probandum. But the later one does not require
such things. An example can make clear this point. One can state that 'there is
smoke because there is wet fuel connection. But one need not say that 'where
there is wet fuel contact there is fire'. The first one is the example of conditional
relation whereas the second one is that of natural.
51
ibid.
52
EPSE
EPSB, p. 227
MM, p. 26, WTrfsjRfcf W >HIU|^ TTrairiJifHci, ^ - v N ' } ^ |
Theory of inference according to the Mimarhsakas 67
Conditions of a valid inference:

Inference for others' sake, as it occurs in Mlmdmsd system, is greatly dependent
on three constituent propositions such as: pratijnd, hetu and uddharana. The
validity of inference depends on these three factors. If either of these is proved to
be fallacious then the whole inference will be proved as invalid. Therefore, like
Buddhist logicians54, Mimarhsakas, more particularly Kumdrila Bhatta, has
indicated the fallacious grounds corresponding three factors such as: pratijnbhasa,
het\'dbhdsa and drstdntdbhdsa55. Theses are the key components of the process
that's why it is necessary to determinate their fallacious aspects.
Pratijndbhdsa
Sucarita Mishra, while commentating Slokavdrttika, defines pratijndbhdsa as
sadhyanirdesa56. Proposition or pratijnd is a statement of establishing the object of
inference (sadhya). Sddhaya should not be previously known (m^W J ^ d d ) and
it should not be in a contradicted form (cifeqîidlsft). If it is
The definition previously known then there is no need of re-establishing it.
And if it is contradicted by another means of valid knowledge
then it cannot be established. In other words, an unknown and un-contradicted
statement is called a proposition. If it is proved to be otherwise then it will known
as pratijndbhdsa or the fallacies of inferential subject. Ndrdyana, classifies
fallacious proposition into three such as: (1) siddhavisesana, (2) bddhitavisesana
and (3) aprasiddhavisesana57. Let us discuss them briefly.
If a proposition is already known by perception and still it is going to be inferred

then it will be called as siddhavisesana pratijndbhdsa. For example, "fire is hot".
54
see first chpater
55
MM, p. 59, 3TSJMlcl5ll!oO^ < J « N M ^ ^ W T I
(FMI cfTszrcts^nf^skTfgt mf^t y r 11
56
Comm. on verse no-54,
Theory of inference according to the Mimdrhsakas 68
The second one occurs when a proposition is contradicted by another stronger

pramana then it is called as badhitavisesana pratijndbhdsa.
For example, "fire is not hot". The third one is
aprasiddhavisesana pratijndbhdsa, which is not an
established fact. From example, "Earth etc. has an omniscient creator". Here, it is
not a known fact, because we don't find such things in ajar etc.
Hetvabhasa
There is no reference of hetvabhasa or fallacies of inferential subject either in
Jaiminlya sutra or in Sahara bhasya. The later commentators like Kumarila
Bhatta. and Prabhakara Mishra have determined it. Of
Classification of course, different opinions can be notified between
hetvabhasas . ,
these two over this issue. Kumarila s classification

seems to be little controversial . But Mdnameyodaya,
a treatise ofBahtta school, presents a clear picture of this class of fallacies. On the
other hand, Parthasarathi Misra also brings into notice of an improved
classification of fallacies of probans in his treatise called SatradTpika59.
Hetvabhasa, according to Mimamsa system, generally is of three kinds viz.

asiddha, anaikantika and badhita60. Of course, there is a controversy regarding on
it like when Pundit Narayana adds one more (i.e. asadharana hetvabhasa) to the
list then Kumarila considers it as a subdivision61.
58
EBPM,p. 273
59
SastraJTpika heftabhasa prakaruna.
60
SD, p. 120 a r t e 3i$*iRi*rin ^fteR 3ft WJ\ &g£m
61
it will be discussed later.
Theory of inference according to the Mlmdmsakas 69
Asiddha hetvdbhasa:
Asiddha or un-established reason is the absence of the knowledge of the probans,
which is qualified by invariable concomitance, is related to the subject6". It of five
kinds such as: svarupdsiddha, sambandhasiddha, vyatirekdsiddha, Asrayasiddha
vydpyah'dsiddha63. In the case of svarupdsiddha, the very nature of the reason is
un-established. For example, "Buddha is the knower of both
.rv v N dharma and adharma". Here, the reason omniscience is an
imaginary character. Hence a valid conclusion cannot be
drawn. Sambandhasiddha denies the relation. For example,
"the fire does not burn, because it is cold". Hence, it is not a valid knowledge.
Where the identity of both paksa and hetu becomes one and same the
vyatirekdsiddha fallacy arises there. For example, "Cow is an animal having
dewlap etc., because it is denoted by the word 'cow'. Here, the inferential subject
and the reason are one and same. Asrayasiddha mainly denies the asraya or
substratum of the reason. For example, "the sky lotus is fragrant, because it is a
flower". In the case of vydpyaftdsiddha the reason resides only in a part of the
inferential subject. For example, "air and space are non-eternal, because they are
tangible".
When the reason exists in both the subjects (i.e. inferential subject and its counter
subject) then that will be called as non-conclusive fallacy64. It is again of two
kinds like: savyabhicara and sapratisadhana . Kumarila adds the third variety
introducing the term called asddhdrana66. When a fallacy entertains a
contradiction of probans with probandum then that will be called as savyabhicdrin
, c v s hetvdbhasa. For example, "sound is eternal, because it
is intangible". Here, intangibility (probans) does not
12
SD, p. 120
'3 Ibid.
4
Ibid.
5
Ibid, p. 121
* EBPM, p. 275
Theory of inference according to the Mtmarhsakas 70
regularly accompany eternity (probandum). Sapratisddhan hetvabhasa occurs

there where two contradictory reason emphasize the same object (i.e. object of
inference) are both are equally strong. For example, "air is perceptible, because it
is a substance and it is formless". Again "air is imperceptible, because it is a
substance without color". Here, both the arguments are equally strong.
The very nature of this fallacy is that it contradicts and does not allow probandum
to be proved. It proves the absence of the probandum
instead of proving its existence. For example, "sound
is eternal, because it is produced". This is also called Badita or obstructer because
it does not allow inference to be proved as a valid inference.
Drstantdbhdsa:
Drstantdbhasa means fallcaious example. And an example, in definition of the
author of Mdnameyodaya, is a place where the pervasion of probans and
its division Probandum is established. This is the place where the object
classification of the inference and its cause are present with a certain rule.
This is again of two kinds such as: sddharmya (positively similar) and vaidhaymya
(negatively similar). "What has smoke has fire, like a kitchen etc.," is the example
of the former one and "what has not fire does not fire does not have smoke, like
water." If, in any case, the example happens to be fallacies then the whole
inference will be invalid". Therefore, it is necessary to know the nature and
function of the fallacious examples. However, Ndrayana, has classified the similar
instance into four like: devoid of the object of the instance (sadhyahina), devoid of
probans (sddhanahma), devoid of both i.e. probans and probandum (ubhaydhlna)
and devoid of substratum {Asrahina). Similarly dissimilar example also has been
divided into four kinds such as: sddhyavydvrtta (not absent from the probandum),
sddhanavyavrtta (not absent from the probans), ubhayavyavrtta (not absent from
both) and AsrayahTna (devoid of base) .
However, the entire discussion of Nardyana is based upon Slokavdrtika .
Observation:
A comparative study of the process of inference of both the systems provides
some notable observations. Though Mimamsa system has a different interpretation
Similarities of inference still the process remains unchanged. The
and
dissimilarities invariable association of vyapya and vyapaka is called
niyama. Nayiyayikas accept five membered syllogisms
whereas, Mimamsakas reduce it into three. Nyaya system accepts five heh'dbhdsas
whereas Mimamsa accepts only three. In some points they are in one track. For
example, dsraydsiddha hetvabhasa, savyabhicara hetvdbhdsa etc., whereas in
some points they in a different track. For example, regarding avayavas, numbers
hetvabhasa etc. However, in spite of lots of similarities and dissimilarities both the
systems have accepted inference as a means of valid knowledge.
MM, p. 89
SV, verse 108-137
Selected readings for third chapter
Epistemology of the Bhatta School of Purva Mimamsa

-Prof. Govardhan P. Bhatta, Chowkhamba Sanskrit Series,
Varanasi, 1962.
The Prabhakara school of purva Mimamsa

-Prof. Ganganatha Jha, Motilal Banarasi Dass, Delhi, 1978.

Bhatiya Darsan Brhatkosa

-Bachulal Avasthi, Sharada publishing house, Delhi, 1997
Paramarsa in Navya-nydya (Ph.D. thesis)

-Krishna H. Kulakarni, CASS, University of Pune, 1999.
BDB Bharatlyadarsana Brhatkosa
PP Prakarana panjnikd
MM Mdnameyodaya
*
SV-2
Slaokavarttika
EBPM Epistemology of the
Bhatta school of purva Mimamsa
MS MTmamsdsutra
SBh Sabarabhasya
NR Nyaya ratndkara
W% ^ ^ i f a s w m ^ T t x W ^frTT SFT WcTT, ^ ^TTSaf nt^ftfcT c i O T M , ^Ttssj j p f f a ^
CRT 5Pf4, ^ 3TsIRl5lM' ^TT 3#rfcfT( W ^ ^ f ^ T T ^ f ^ T x ^ MR>HHM^) |
I cn^RTFPf TTT^FT I
(Pramatr means one who, being guided by the desire to seek the object, is led to
activity. Pramana is the instrument by which the pramata rightly knows the
object. Prameya is the object rightly known. Pramiti is the right knowledge of the
object. With these four, tattva reaches its fulfillment).
Chapter - IV
Theory of inference according to PracTna Nyaya
General background
Process of inference according to Gotama
Process of inference according to Vatsyayana
Process of inference according to Udyotakara
Chapter at a glance Process of inference according to Vacaspati Misra
Process of inference according to Udayana
Process of inference according to Jayanta Bhatta
Observation
Theory of inference according to Pracma Nyaya 12
Fourth chapter
Theory of inference according to PrdcTna Nyaya
Philosophy deals with concepts. Concepts like soul, liberation etc. It characterizes
the concepts and finally establishes it on the basis of a set of reasoning. The
application of this set of reasoning gives birth to various intellectual traditions,
which are otherwise called hetuvidyd or hetusastra. This hetusastra changed into
tarkasdstra later on when the reasons were indebted by a group of intellectuals.
General ^ m s ^ a c t briefs me
history of origin and development of
ackgroun Indian philosophical traditions. Now let us turn our look into
the internal affair of the systems. Systems are divided into subsystems, because of
the method of application of its logicians, because of the method of clarification
(pariskdrapaddhati). Another possible cause of its division could be the debate
among themselves. They examine the statement and take forward to discourse.
This helped in the emergence of the system called "Navya-Nyaya". However, this
chapter deals with origin and development of inference starting from Gotama to
Udayana (i.e. from 200 B.C. to 1050-1100 A.D.).
Theory of inference according to Gotama:

Nyayasutra is the first systematic writing of this tradition. The authorship of this
magnum opus is attributed to Gotama1, who seems to have flourished around 200
B.C. Nyayasutra mainly deals with the nature of objects (prameya). Therefore, it
is called the science of objects or prameyasdstra. However, the author has given a
primary importance to the instruments of obtaining valid knowledge {pramana) by
placing it at the very outset of the book .
1
He is also known as Gautama, Aksapada, Dirghatapas etc. as it is ascribed in various scriptures.
2
Vide. N . S . 1.1.1 W W ^ P F R F J cMs1Mlr?|:^Hlf£| J IH: I
Theory of inference according to PracTna Nyaya 73
According to him, the correct judgment of the nature of the sixteen categories
(among which pramana has been placed first) leads to salvation. It seems that
pramana is a member of the class of sixteen categories
Pramana in °
Pracma Nyaya formulated by Gotama for application of anumana. He does

not give any definition of pramana anywhere but he classifies
it into four viz. pratyaksa (perception), anumana (inference), upamdna
(comparison) and sabda (verbal understanding)3.
After defining pratyaksa Gotama proceeds to explain anumana. He defines

anumana as "preceded by it"4. It is understood that it refers to pratyaksa (which
has been pointed out by the commentators later), which has
The theory
«n far been discussed in the previous sutra. So, he means to say that
inference is greatly dependent on perception. In simple
language, inference requires a previous perceptual knowledge like if a river is
banked then it can be inferred that there was a rainfall etc. Now opponents raise an
objection saying, "inference cannot be a means of valid knowledge, because it
suffers from vyabhicdra" (he raises this question in the sutra no 2.1.37). He
immediately replies saying that our inference based on something else other than
the part, fear etc5. In this way, he has established the authority of inference but has
not mentioned a systematic process of it. He only mentions name of the members
of syllogism (avayavas), the name of components like hetu, sddhya, het\'dbhasa
etc., in the siitra. It implies that he is referring to pararthanumana. There is no
reference about vydpti as well. However, later logicians have arranged these
components into a sequence.
3
Vide. N.S. 1.1.3 ^rc^m^Tr#RFRT^TT HHl"llPi I
4
N.S. 1.1.5am d<^4cb f%fc(wpTFR...I
Theory of inference according to Pracina Nyaya 74
Gutama divides inference into three types such as: purvavat, sesavat and
samanyatodrsta , but he does not give any definition of these divisions. However,
these terms have been explained by later commentators.
Sutrakara mentions nothing about pararhtanumana, but defines the members of

syllogism. These are of five types such as: pratijna, hetu udaharana, uapanaya
and nigamana1. A proposition or pratijna is the declaration of what is going to be
established (sadhyanirdesah). This is a statement, which is
Members
of
supposed to be ascertained with the help of other members of
syllogism
syllogism. Hetu is an establishing factor. He defines hetu or

reason as the means for establishing the object of inference (sddhya) through the
homogeneous (sadharmya) and heterogeneous (vaidharmya) character of the
example . He also uses synonyms like liriga, sadhana etc. Udaharana or example
is of two kinds such as: sadharmya or homogeneous and vaidharmya or
heterogeneous . A sadharmya udaharana possesses the thing that which is going
to be established (sadhyasddharmydt) and that which implies its opposite i.e.
which does not possesses the object of inference is called vaidharmya udaharana.
Upanaya or application, as he defines, is dependent upon example
(uddharanasdpeksa). Though he has not mentioned any division of upanaya still
its two division is understood because of its dependency on the udaharana. The
sadhrmya upanaya can be concluded with the statement like "it is so" (tatha) and
the vaidharmya upanaya can be concluded with the statement like "not like so"
(na tatha) . Nigamana is the affirmation of the proposition (pratijnayah
6
N . S . 1.1.5 . . . T j f e & W W T T ^ c f r ^ xf |
7
Vide, N.S. 1.1.32 wfas1l^lû|)iH.UpHHHMcii|c||: |
8
Ibid. 1.1.34/35
9
Vide, N.S. 1.1.36 îgWumm^l^M+iltR ^ 1
'1
10
Vide N.S. 1.1.38
punarvacanam) . It is the confirmation of the proposed statement after the reason

(hetvapadesdt).
Except these, the term like sadhya also occurs for several times in the sutra but
there is no definition of it. Similarly, the concept of vyapti is also not found in the
Gotama sutra. Though he mentions the term like paksa but not in sense of
'inferential ground' but in the sense of 'side' (i.e. svapaksa, parapaksa).
According to Maharsi Gotama, the proper knowledge of hetvdbhdsa is one of the

means of attaining the ultimate peace . He does not define it but classifies it into
five such as savyabhicara, viruddha, prakaranasama, sadhyasama and kdldtita13.
Savyabhicara hetvdbhdsa is anaikdntika. It means not one
pointed i.e. the reason, which leads to more than one
Hetvabhasa
conclusion . He further defines viruddha hetvdbhdsa or

contradictory fallacy as reason, which opposes what is
supposed to be established15. A reason is supposed to be an indicator of the object
of inference but in spite of being so it contradicts it. Hence it is called a
contradictory fallacy. Prakaranasama hetvdbhdsa is such a fallacious reason,
which intends to establish something but produces doubt regarding the same
(prakarana cintd)16. The fallacious reason that which is not different from sadhya
(sadhya avisista) because it requires to be established (sddhyatvdt) is called
sadhyasama hetvdbhdsa. A proposition always stands to be proved but in this case
the reason itself needs another proof to be proved17. And Kdldtita or mistimed is a
11
VideN.S. 1.1.39 SotM&tlkHRlslWI: ^4^H faWFFl
12
VideN.S. 1.1.1.
13
Vide N.S. 1.2.4. ^c^rJKRi^WchxiuitlH^TÎ^HcbMlclldl i^JFfmF |
14
Vide N.S. 1.2.5. ar%rf^fr ^rfraRT I
15
Vide N.S. 1.2.6. fa^RFTR^f ^ M fa^gr |
16
Vide N.S. 1.2.7.
17
Vide N.S. 1.2.8.
fallacious reason which is assigned after lapse of time {kdlatyaydpadista).

However, the later logicians of this school will discuss these fallacious reasons
elaborately with their respective examples.
This much infonnation we get from Gotama on the theory of inference. Though
examples are not provided still he provides the basic framework of inference,
which grows and bears fruit in the later period after a due care of a number of
logicians in the form of gardeners.
Theory of inference according to Vdtsyayana:

Vdtsyayana is supposed to be the earliest commentator of Nyaya-sutra of Gotama.
Though, a controversy has been entertained by various scholars like 'he is not the
first commentator of Nydya-sutra,ig still they are in agreement that the earliest
work on Nyaya-sutra available to us is the commentary of Vdtsyayana. He is also
known as Paksilaswdmi, Drdmila . However, his time, as has been fixed
71
approximately, is 450 A.D .
Pramdna, according to Vdtsyayana, is arthavat (i.e. invariably connected with the

object) because successful activity (samartha-pravrtti) results when the object
{artha) is verified by pramdna . He holds pramdna as instrument of valid
knowledge. However inference, as he defines, is 'the after knowledge {pascdt
mana) of an object {artha) through a reason {linga), which is rightly ascertained
{rnita)23. That is to say that probans and probandum are invariably connected.
18
Vide N.S. 1.2.9. *Mlc4|AlNftki: *ldldld: I
19
EIP,p. 129
» Ibid.
NSG, p. xiii
22
NBh. Opening statement WHluidlS^fclMdl M^frWWÎd I
The inference of effect from its cause is called piin'avat . This effect should be
immediately prior to it (i.e. cause). For example, the inference of a future rainfall
from a heavy dark cloud. Sesavat inference is the
First
inference of a cause from its effect28. For example,
explanation
from the fullness and swift current of the river, we

infer that there was a rainfall on the mountain. The former one is called kdrana-
lingaka because from the cause only the effect is being inferred whereas the later
one is called karya-lingaka because from the effect the cause is going to be
inferred. Samanyatodrsta, as its name suggests, is a notion of generality. Of
course, there is no definition of Samdnyatodrstdnumdna in Nydyabhdsya, but its
example characterizes it as an inference of general co-relation. For example, the
perception of an object at some places which was previously somewhere else .
Because of its movement it changed the place. It means, we can draw an invariable
relation like whatever changes its place is preceded by movement. Such general
perception of invariable relation leads one to infer that the sun has movement
because it changes its place.
Perhaps Vdtsydyana was not satisfied with the first explanation so he stated
explaining it further saying "athavd purvavaditi"i0. Or it could be the case that he
has quoted some other commentator's view (as he is not supposed to be the
immediate commentator of Nydyasutra)31. However, Vdtsydyana has given second
alternative explanation of the division of inference. Piin'avat, as he defines
alternatively, is the inference when an object, which is not in direct contact with
the sense organs (apratyaksya), is inferred through the perception of two objects
27
NBh. 1.1.5 TJcfacT - TF3 cbKiiH e M ^ ^ c ) I
28
ibid. ? t W - ^ ^T#W cjjRWpftac) I
29
ibid. ^ H M c D ^ M - sl^l^JchH^H ^ P T T S ^ T 3?faftRl I
30
NBh. 1.1.5
31
AP, p. 69
{yathdpun'apratyaksobhiita-anyataradarsan) as they were previously perceived32.

For example, perception of smoke etc. somewhere in the mountain helps to infer
Second fire, which is not in direct contact. The word Sesavat
ema lve means residual (parisesa). Hence, the inference of
explanation \t . J >
residual is a definite knowledge {sampratyaya) resting

on the remainder after the elimination of certain possible objects
(prasaktapratisedha) because of the irrelevance in the case of other objects
(anyatra aprasamgdt)33. For example, sound has been characterized as an existent
and non-eternal, which are the common characteristics of substance, quality and
action. So it is being differentiated from sdmdnya, visesya and samavdya because
these are eternal. But when doubt arises whether sound is a substance, quality or
action, it can be eliminated like this: sound cannot be a substance because it
inheres in a singular substance (eka-dravyah'dt). It cannot be an action because it
is the cause of another sound (Sabddntarahetuftdt). What remains is quality.
Therefore, sound can be regarded as a quality34. Sdmdnyotadrsta inference is that
in which the relation between the probans and probandum (lingalinginosambadha)
is being imperceptible (apratyaksya) and the imperceptible probandum
(apratyaksolingi) is inferred from similarity (sdmdnydt) of the probans to
something else35. Let's have an example, the self (dtmd) is inferred from desire
etc., because desire etc. are qualities and qualities always inhere in substance.
Therefore, the self is the substratum of qualities like desire etc.
However, these are clear indications of a form of pa rdrthdnumdna.
32
NBh. 1.1.5 ^ f ^2TT^4 3^âT^cRtS^d^ci^^HlS^dVîl5'R^T^^]I^HH
33
Ibid. ^ rT y^Tby(cl^S^T5lTSyRl«dlf^^HI"l +IHk^: I
34
Ibid.
Members of syllogism
Vatsyayana means that the four inferential components like pratijha etc., are
ascertained by the four instruments of valid knowledge mentioned by Gotama. As
a matter of fact, the inference components, without being based on instruments of
valid knowledge, cannot ascertain any truth. Therefore, he holds that the
inference-components like pratijha etc. are ascertained by four pramdnas, like
perception etc, such as: pratijha or declaration of statements is verbal
understanding (dgama) because 'what is going to be
Avayavas established has been declared' i.e. 'the mountain is
are
ascertained fiery'. The probans or hetu for the establishment of the
by pramunas
proposition is inference (anumdna) i.e. 'because there

is smoke'. The uddharana or example is perception like 'where there is smoke
there is fire, like kitchen'. Upanaya or application is comparison i.e. in the kitchen
the smoke was pervaded by fire and here there is same smoke (i.e. it is also
pervaded by fire). And nigamana or conclusion is a demonstration of the result of
all these pratijha . Vatsyayana calls it paramo nydyah or Nyaya par excellence .
Pratijha, as sutrakdra defined, is a declaration of what is going to be proved

(sddhyanirdesah). But it does not satisfy all necessary conditions. Therefore,
Vatsyayana clarifies it extending the definition little further. Pratijha, as he
defines, is a proposition stating a specified subject as qualified by the
TO
characteristics sought to be proved .
36
Ibid. ^ 5 MH|U|*Hc|M: 3TTW yfasTT, ^ ^ H H H , \ ^ | ^ u j JR^KPT I v i M H ^ H H H H ,
37
NBh.
The nuclear sense to be noted is that Vatsyayana has used the word sadhya in the
sense ofpaksa29. He also keeps the meaning of hetu unchanged. However, from
his explanation of hetu it can be easily concluded that there
Avayavas as per
Vatsyayana are of two divisions of probans such as anvayin and
\yatirekin (both are dependent on the nature of vydpti).
Vatsyayana has no second voice regarding the third, fourth and fifth member of
syllogism i.e. udaharana, upanaya and nigamana respectively. Vatsyayana also
accepts five kinds of hetvdbhdsas without any objection. He then just expands the
definition of those fallacious reasons with their respective examples.
However, both the philosophers i.e. Gotama and Vatsyayana have not
mentioned anything about pardmarsa. but they certainly have put the idea that has
been discovered in the later period.
Theory of inference according to Nydyavdrttika:

The process of inference has been developed gradually in the course of time. In
between the Buddhist logicians interacted and criticized, which helped Naiydyikas
a lot to develop their system in a proper manner.
Appeared
ma
Udyotakara41, a man of logic, wrote a sub-commentary on
challenging
period Nydyasutra. He appeared by the time when the Buddhist
logicians such as: Digndga and DharmakirtTv/ere challenging the system.
Udyotakara takes each and every word of Nydyasutra 1.1.5 and defines it. The
first word is 'atha', which has been defined by him, as sequence or continuity . It
means after defining pratyaksa sutrakara proceeds to define anumana. The very
idea of doing so is that the word 'atha' carries more than one meaning such as:
39
Nyaya Philosophy by Debiprasad and Mrnalakanti, p. 107
40
AP, p. 87
41
He is also known as Bharadvaja, Pasupatacarya. HIL, p. 124
Theory of inference according to Pracmu Nydya 82
mangala, anantara, arambha, prasna and kartsnya . Here, he adopts the

meaning 'anantara'. Then he takes the word 'tat-purvakam' where he gives three
alternative explanations44. The word 'tat-pun'akam', as per his first explanation,
stands for all forms of valid knowledge (tdni-piin'akam). This explanation of
Udyotakara declares that all means of valid knowledge
Udyotakara (including that of inference and verbal understanging as well)
on
Nyayasatra [s preceeded by it. This also does not go against Bhasya's

declaration that 'inference proceeds from perception',
because all means of valid knowledge are indirectly dependent on perception.
Second explanation is tat-purvakam and the meaning that it signifies is only
'perception'. And the third explanation is ite-pur\>akam, means perceptions, which
represents two perceptions by two premises such as lingalingisamambdha darsana
(perception of the relation between the probans and the probandum) and
lingadarsanam (i.e. perception of probans).
This second perception awakens the impression left by the former
perception, which leads to remember the relation between the probans and the
probandum. And after this remembrance, when the probans will be perceived the
third perception of the probans with former two perceptions along with
remembrance gives birth to the inferential cognition, which is known as
paramarsa45. This process of inference follows four steps such as:
42
NV 1.1.5 3Ta|^H*tf4
43
Amarakosa - H*HldH»*N|WW!McMcc-i|fc|2it 3 M
44
N V 1.1.5 ?TTfa ct efcT ^ ^ T cTfot dc^4cbH
45
Ibid. fJJcHlcldl fecikfel^lciSlHkl ^^Kl(^c^c^x1>ichld ^jfcfT, ^ e U H - v K ^T
the probans? To which several answers have been given such as: Some people
explain that it (inference) is the remembrance of the relation between the probans
and probandum48. Some people hold that this remembrance as the means, but only
in so far as it is aided by the perception of the relation of the probans and the
probandum and such other functions49. A third party
r J
Lingaparamarsa
is
_ considers that lingaparamarsa (the probans as invariably
anumana
concomitant with the probandum) is the inference50. Keeping

all these discussions in mind Udyotakara gives his own statement. He holds that in
case of in the case of inference all these factors are necessary, because they all
play an equal role in the process. But among these factors the last one i.e.
lingaparamarsa seems to be most important, because we find that the inferential
cognition of the probandum follows immediately after paramarsa5J'.
Thus, Udyotakara emphasizes on paramarsa, more precisely, on lingaparamarsa.
The division:
Udyotakara now turns to the part of the sutra, which says that inference is of three
kinds. First of all he divides inference on the basis of the nature of hetu. This type
of division is neither seen in Nyayasutra nor in Nyayabhdsya. Those inferences are
anvayi anumana (positive inference), vyatireki anumana

lyision ase (negative inference) and anvayivyatireki anumana (positive
vr
on the nature of
hetu
and negative inference)52. The positive inference occurs there
where the probans exists in association with probandum in the inferential subject
{paksa) but not in vipaksa5 . For example, sound is non-eternal because it is a
product. Here, it has no dissimilar example. In the case of negative inference the
probans exists in the subject in which case there is nothing apart from the subject
in which the probandum is known to be exist. And which (hetu) does not exist
where the probandum is known to be absent54. For example, the living body is not
soul-less because, it is associated with life. The third one i.e. positive-negative
inference suggests that where there is probans there is probandum and where there
is absence of probans there is absence of probandum55. For example, sound is non-
eternal, because while belonging to a generality and being possessed of specified
individuality, it is perceived by the external organs of ordinary human life like us.
After defining three-fold inference of his own now he proceeds to define

the three-fold division of inference given by Vatsyayana. He defines it in a
different manner. The word 'pun'a' means probandum and purvavat means which
invariably have the probandum for its probans56. Similarly, the word 'sesa' stands
for the probandum and other things related to it. So, that which has the probandum
Theory of inference according to Prucina Nyaya 87
which is concomitant with the subject and in the things

Trividha
stands for the similar to that is called true probans or sadhetu62. And that
classification ,.
of hetu which is never found apart from the things of the same kind is
called sandigdha hetu . He also gives another alternative explanation of the word
'trividha'. The word 'trividha', as he suggests, may be taken to restrict the number
of inferences65. However, according to him, there are endless divisions of
inference66. Hence, to restrict these the sutra speaks of three kinds.
Members of syllogism:
Udyotakara gives a clear definition of avayavayas. He holds avayavayas or factors
of reasoning as a statement which forms part of inferential argument67. So, the
definition of avayavayas, as has been given by Udyotakara, follows like this.
'Those factors of reasoning, which serve to accomplish a purpose not
accomplished by other, and which, through these purposes combine together in the
focus of a single compound sentence, and thus collectively accomplish the one
desired purpose of proving the conclusion to the other party'68.
Udyotakara also accepts a syllogism containing five members. He then just

elaborates the idea of Vdtsydyana. Of course, there is a rigorous debate between
Nayiyayikas and Buddhist logicians regarding the definition of pratijha. The other
Theory of inference according to Pracma Nyaya 88
members of syllogism like hetu, udaharana, upanaya and nigamana have been
discussed at length by him. However, no difference of opinion with bhasya has
been seen.
Hetvabhasa:
Udyotakara gives a full-fledged definition of hetvabhasa. Those are fallacious
probans because they don't posses all the characteristics of a good hetu but are
sufficient enough to appear as such69. So, it doesn't satisfy all the necessary
characteristics of a good reason but it appears to be so. He then proceeds to define
„, , the numbers of heh>abhasas. It's number, as he suggests, is
OD
The number
of more than five. On one hand their number cannot be
hetvabhdsas
calculated (ananta), but in other hand, more precisely, their
number goes to 176. He suggests this number on the basis of the diversities of
time, person and the thing to be proved70. There are sixteen cases where the
sadhya pervades the hetu11. A similar list of sixteen arises where hetu pervades of
"77
only a part of sadhya , and another sixteen occurs where hetu is absent from
sadhya . Finally, the number of hetvabhasa, as has been accepted by the system,
is five.
69
NV 1.2.4 %êT§MT'HNK^c|1 ^ I H M I ^ c N m M H H I : " t^TTTOTT I
70
ibid. c f T M - ' q ^ - ^ ^ T j f c r t f T ^ A i q R * i ^ * \ • , xHHMd: ^TTtiicj^Mijûi ^p^rnrpter
HMJCJHH: tfcJîdfcWcW I
71
Ibid. cT^ W^ciimcbST^fcr ^?f?TT I
72
Ibid. ^ T l ^ ^ ^ l ^ R l E R f s ^ R |
73
Ibid. TJcf mZFffivA'fc |
Theory of inference according to Praclna Nyaya 89
Udyotakara gives an alternate definition of Viruddha hetvdbhdsa. It occurs,

according to him, when either the reason contradicts the hypothesis or hypothesis
Difference contradicts the reason74. For example, "sound is eternal,
w tn
^ because it is a product". He further proceeds to discuss
Bhdsya
sddhyasama or asiddha hetvdbhdsa. Here he classifies it into
three such as prajhapaniyadharrnasamdna (Vacaspati Misra calls it
svarupasiddha), dsraydsiddha and anyathdsiddha hetvdbhdsa 5. The first one
occurs where the hetu is as much as in need of proof as sddhya. Second one occurs
where the locus or substratum (dsrya) of hetu is not known and in the case of third
one the hetu can be explained in such way (anyathd) which would make it a proof.
A befitting example for it is "shadow is a substance because it has motion".
Thus, Udyotakara has developed the idea of inference in his commentary on
Nyayabhdsya.
Theory of inference according to Nydyavarttikatdtparyatika

The science (sdstra), which is the cause of liberation of mankind, was composed
by Aksapada and explained by Paksilasvdmi. Apart from this Udyotakara wrote
vdrttika because people were misled by the bad reasoings of logicians like
Dignaga etc . This is the opening statement of Vacaspati Misra, the writer of
Tdtparyatikd on Nyayavdrttika. His time, as has been fixed by the philosophers, is
a matter of controversial issue. However, it is finally accepted that he lived in
between 900-980 AD approximately77.
74
NV 1.2.6 ^KT W R T i^fafteft t ^ M afasiifcKleraî ^?T faWt ^RRi I
75
NV 1.2.8 #S?FT%£^%rT Wtfo - M s l N - t a ^ H H , s n ^ q i f ^ T ST^MlRHc^Ri I
76
NVTT, opening statement. 3T8T ^ R ^ S ^ N T ^ T f^fcfi?cft W^ Wrfti, c^cMlR^ ^ WJÛ
•qf^cH^Jif^ni,...^c|,KlTt)y^MHi^c||c;^4^ ciiRi^cbK^M^^H ^ k f a l
77
EIP,p. 10
Vclcaspati Misra flourished in a quiet developed stage of Nyaya system. His effort
on Nyayasutra is known as Nyayavarttikatatparyatikd. A valid cognition,
according to Vacaspati Misra, is invariably associated with its object, because it
produces activities that are successful. In other words, after knowing an object our
activity, with regard to that object, proceeds successfully78.
A question, regarding the validity of inference, was put forth by the Buddhist
logicians. "If the inference is not valid", replies Vacaspati Misra, "then either we
may invite a vicious circle by trying to prove the validity of inference by another
inference. Or we may fail to express a meaningful proposition unless we admit
verbal understanding as a separate source of cognition apart from perception"79.
Vacaspati Misra has not prepared any new definition for inference. Rather he
claims that the definition of inference, such as 'tat-pun'akam', suffers from the
narrow application as well as over application80. Because
Vacaspati ^rom m e wor(i <^> p erC eption is implied. It means inference
Misra on
Nyayasutra is preceded by perception. If it is so then verbal testimony etc.

can be regarded as inference because these are also preceded
by perception. This is the case of over application81. Secondly, if somebody first
infers fire and then he proceeds to infer the hotness of the inferred fire then in that
case this definition of inference will not work (because the later inference is not
preceded by perception). This case, in philosophical language, is called narrow
application of the definition82. Therefore, keeping all problems in view, he
dissolves the compound ltatpur\'akam'' in three alternative ways.
78
NVTT 1.1.1
79
Ibid.
80
NVTT 1.1.5 d^KciHi^facît^'UJHdSFTAdRc-UcT OT? I
81
Ibid.
82
Ibid.
First of all he derives the meaning as Hani' of the word Hat'*"'. By doing so he
includes not only perception but all pramanas also. His second alternative
explanation conveys meaning like 'that which is preceded by two perceptions.
Here, he explains the terms like Vastyayana. Similar thing happens in the case of
third explanation also.
Vacaspati Misra opines that the example of samanyatodrstanumdna, cited by the

commentator, is not fit to the context. According to him, this example comes
under sesavat inference because here also from the change of the place of sun the
motion of sun can be inferred . Samanyatodrstanumdna, as he explains, is the
inference of probandum (dharmi), which is seen to be invariably connected
O f
(avindbhava) with the probans . For example, the place known as with cranes is
the means of inference of the place as with water. However, he supports
Udyotakara in the context of endless division of inference.
He has explained the word Hrividha' of inference on the basis of five

characteristics of hetu, which are classed under following three heads such as
sesavat, samanyatodrsta and sesavatsamanyatodrsta86. Again the number of
hetvabhasa, as given by him, is 203287.
Vacaspati Misra, while defining the subdivisions of heft'dbhasa, says

savyabhicdra and aneikantika are synonymous. Therefore, one can be used as a
definition of the other . He divides sadhyasama hetvabhasa into four varieties

keeping three varieties of Udyotakara unchanged. The additional one is known as
ekadesasiddha. It occurs when the hetu is partially absent from the inferential
SQ
subject i.e. paksa . Vacaspati Misra is not satisfied with the explanation of
kalatyaydpadista hetvabhasa given by Vdtsydyana and Udyotakara90.
However, in this way Vacaspati Misra has depicted the process of inference.
Theory of inference according to Nyayavdrttikatdtparyatikdparisuddhi (NVTTP)

The main objective of Udayana in writing parisuddhi was the exposition of the
views of Vacaspati Misra, the author of Tatparyatika91. Parisuddhi is a
commentary of Tatparyatika. When Vacaspati Misra was criticized by a host of
Buddhist logicians, Udayana, better known as Acarya, flourished to protect him
from those kutdrkikas. In the language of Anantalal Thakur it is written as
"...Udayana is the last great master of ancient Nyaya and is, according to many,
the priest of that fertile marriage between Nyaya and Vaisesika which gave birth to
Navya Nyaya" . His time, as has been fixed by Frauwallner, falls in between 1050
to 1100 AD93.
However, as he declared to follow Vacaspati Misra strictly. So he did. Hence his

theory of inference carries no difference of opinion with comparison to the theory
of Vacaspati Misra. Therefore, there is no need of explaining it further.
88
AP, p. 108
89
Vide NVTT 1.2.8
90
AP, p. 108
91
NVTTP Mangalacaranam ^ T W ^ f f ^ ^T ^ d d l V$ti I
'2 NVTTP edited by Anantalal Thakur, ICPR, New Delhi. Introductory passage.
Theory of inference according to NyayamanjarT

Apart from the tradition of writing sub-commentaries on commentaries we have a
tradition of writing independent commentary on Nydyasutra. NyayamanjarT, a
mature Sanskrit work on Indian logic, is of this kind. Its
Nyayamanjari
a mature authorship is ascribed to a Kashmirian Brahman, Jayanta
Sanskrit work
of Bhatta by name. He seems to appear after Udyotakara by
Indian logic . . „ ,„,, „ ...
before Vacaspati Misra . Hence his time, as approximately
has been fixed by Steinkellener, is 840 to 900AD94. Needless to say that a good
deal of controversy has been entertained with this regard. However, this author is
famous for his independent views.
Pramana, according to Jayanta Bhatta, is a collection of two types of entities like
conscious and unconscious . This collection, otherwise called pramana, produces
the cognition of an object, which is other than illusion and doubt. This pramana,
which has been considered as an instrument of valid knowledge96, is of four kinds.
NyayamanjarT forms a chapter on inference where five subsections have been
presented. Those are as follows.
1) the fivefold nature of probans
2) the problem of invariable concomitance98
3) justification of the validity of inference99
4) the Nyayasutra definition of inference oo
93
EIP,p. 10
94
EIP, p. 9
95
NM on sutra 1.1.3 «teTRto^|c]| ^m$\ WH|U|H I
Theory of inference according to Praclna Nyaya 94
5) the cognition of time and space101
However, the way Jayanta Bhatta defines anumana is neither seen in the
commentary of Vatsyayana nor in the sub-commentary of Udyotakara. In his
language, inference is the cognition relating to the probandum (lihgin jnanam) not
in contact in sense organ (paroksa), which appears when the probans fulfilling a
fivefold nature (pancalaksakdlingdt) and there is a recollection of relevant relation
called invariable concomitance between probans and the probandum
{gruhitanniyamasmrteh) . In this case the knowledge of the probans along with
the meaning of invariable concomitance acts as pramana and the knowledge of
probandum acts as the result of that pramdnal0i. However, from his explanation of
inference following steps can be noted.
1) the perception of probans (which fulfills five essential characteristics)
2) recollection of the relation between probans and probandum
3) inferential cognition
Hence, it is clear that Jayanta Bhatta does not employ the factor called paramarsa
in the process of inference. He considers it saying, "the first step is the perception
of reason, then the remembrance of universal concomitance. The third step is,
according to some, the knowledge of confirmatory cognition"104. The above
discussion clearly state that he does not require any component called paramarsa.
102 J
NM 1.1.5 *T3^fflU|cMf^-J||d <lfedlRMH^c|:|
103
Ibid. 3T^f fa^T-f^T 5lHR^41^d ^T ( ^ J | y R l ^ W u l > H f e d MHIÎH I fa^lslH
104
Ibid. 3T? PfSTTf f c ^ K ? H , cTclT flfcTSRI+Hî, cRTT M r f ^ F ^ iKHSislH, ?TcTT
Without this component, as he records, the upanaya, the fourth member of the
syllogism, can be established105.
After defining anumdna in the above-mentioned way Jayanta Bhatta comes

forward to define the structural analysis of the Nydyasutra. According to him, the
word itat-pur\>akam'' constitutes the definition of inference. 'Tat' is a pronoun
here; it signifies the perception, which is implied from the context106. He holds that
this pronoun 'tat' is in dual number (two perceptions). It also does not create any
problem to include comparison etc. One perception discovers the relation of
universal concomitance while the perception discovers the presence of the reason
in the subject of inference .
Jayanta Bhatta now analyzes the division of inference. A sesavat anumdna, as

Vdtsydyana describes, is the case of inference of effect from its cause as in the
case of inferring rain from the current of the river. But it is not right, says Jayanta
Bhatta, because it may be the case that the rain (or effect) occurs somewhere in the
upper part of the country and swift current of the river is seen due to its connection
with the lower part of the same country108. Therefore, the cause is inferred from its
effect is merely a verbal testimony. But what is in reality is that the subject of the
inference plays the part of the consequence as well as that of reason as it has
105
Ibid. $WTft *Ri miHSfslH ^TTf^T ^^H*JcMH^c|î}c|IMRsi4H;J*M 5rfcPTxT^ I
106
N M 1.1.5 ^T zf 4>KUH * l 4 ^ l c ^ R ) w4cT I ^ T 4 ^ T cTTcfcT TOrtH^TtH I effetcf
107
Ibid. ^ c I M r ^ H I ^ PUMMHHKI d^4^b JPcT^ft ^ c|mic|cfci ifcf d<WJ|c|-d4
different aspects . And sdmdnyatodrsta inference occurs where we infer the

result from the reason, which is neither a cause nor an effect110. For example, the
list of a wood apple is inferred from its color etc. Here color and taste inheres in a
wood apple but not causally related to each other111.
Jayanta Bhatta takes the credit of introducing the twofold division of inference
such as svartha and pardrtha to the Nyaya system. This tradition of dividing
inference has been originated from Prasastapada, the commentator of
Vaisesikasutra. In the language of CD. Bijalwan "Jayanta has the credit of
introducing these two types in the Nyaya and thereby initializing a new tradition
so far not accepted in this school" . On the other hand, he does not accept the
three-fold division of inference such as Kevaldnvayi etc113.
Jayanta Bhatta follows Gautama in the context of describing the fallacious

reasons. He gives a detailed study of all the five types and sub types of
hetvdbhdsas with a nomenclature.
The conclusion:
The above discussion of inference helps us to find out a good deal of notable
points. Inference keeps on changing its form starting from Vdtsydyana to
Udayana. Udyotakara gives a finishing touch to it introducing a concept called
pardmarsa, which has been widely accepted by the new logicians. Jayanta
introduces the twofold division of inference i.e. svartha and pardrtha to the
system, on which the Navy a Nyaya tradition of Indian logic is greatly dependent.
*t* 1* 1* - H S 1**1**1*
109
Ibid. WTTsfa^J epfffc]4ck^ EFfapFJjfoffi ffcT f^jfcT: I
110
Ibid. ^TcT 31cbl4cbKU|^dlfcvH^lld d l ^ c | fcil^l-uS^HH I
111
Ibid. ^rewfTT ^HcJlRjc^KU^cb ^ftcailRs^M, 1 3 di]k-4M cj->i44>NU|'iWl
"2ITK,p.l42
113
Ibid., p.143
Selected Readings for Fourth chapter

Indian Theory of Knowledge
-Dr. CD. Bijalwan, Heritage publishers, New Delhi, 1977
Anumana pramdna
NS Nydyasutra
NBh Nydyabhdsya (Vdtsydyana)
NV Nydyavdrttika
NVTT Nydyavdrttikatdtparyatikd
NVTTP Nydyavdrttikatatparyatikdparisuddhi
NSV Nydyasiitravrtti
NTK Nydya theory of knowledge
NPs Nydyaprakdsa
NSG Nydyasutra ofGotama
HIL History of Indian logic
AP Anumana Pramdna
ITK Indian theory of Knowledge
NM NydyamahjarT
EIP Encyclopedia of Indian Philosophies
3Js|<HI«lleHfNMSl[elcbH^| 3TPt I
(A woman, a child, a cow-heard, a cultivator and such other persons know another
object, lying beyond the ken of their sense organs, by means of its sure mark with
absolute certainty.)
Chapter - V
Theory of inference according to Navya Nyaya
General background
The process so far
Garigesa on inference
Chapter at a glance
Sasadhara's findings
ViswanathcCs additions
Annambhatta's clarifications
97
Theory of inference according to Navya-nydya
Chapter - V
"A close dialogue", says Prof. V.N. Jha, "with the Buddhist logicians and
philosophers generated remarkable structural depths in classical Indian
philosophical thoughts. The experience of this dialogue generated the necessity of
formulating a language of precise communication. This gave
birth to the Navya-nydya language"1. The above statement
Background Qf p r o f jha briefs the necessity of developing a tradition,
which provides the language of precise communication. The
other alternative asset of this tradition is its method of clarification
(pariskarapaddhati). It leaves no ambiguity. This is the only cause for which all
the prominent schools of Indian philosophy adopted this language for the purpose
of communication.
Historically, this tradition starts from Gahgesa, the author of Tat\>acinamani
(1200AD). This tradition mainly deals with pramanas. That's why it is called
pramdna-sastra. However, this chapter aims to discuss the process of inference as
it occurs in the Navya-nydya system. The central focus of this chapter will be
Anumanacintamani of Gahgesa. While doing so, it will keep an eye for the
logicians like Viswanatha, Annambhatta and Sasadhara etc. In short, this chapter
is a chronological study of new-logicians on the concept of inference.
Navya-nydya follows a systematic process of knowledge. First of all, it defines
knowledge or buddhi as the cause of all our worldly behavior . The word
'vyavahara', lies in the definition, expresses the meaning as 'utter words for the
purpose of communicating ideas'3. This buddhi cannot be always true, because
1
Opening statement of a booklet on "Navya-nydya language and Methodology", published by SSUS,
Kalady.
2
TSA- >H4<^c|eiK^r4t5l1*lH I
98
Theory of inference according to Navya-nyaya
sometimes we may be misguided by false impressions. Therefore, the system has

divided it into two such as remembrance (smriti) and experience (anubhava) .
Again, smrti or remembrance cannot satisfy the purpose because this is a product
of mental impression5. It is just a recollection of a thing,
as
'" which had been experienced in the past. Now what is left is
r
Na\ya-nyaya
anubhava, which is again twofold Wk&prama or valid

knowledge and aprama or invalid knowledge. Aprama or invalid knowledge is the
cognition of a thing not having an attribute, as it seems to possess6. For example,
the false impression of a snake in a rope. The above definition of aprama can be
presented in a pictographic form such as:
Rope
Ayathdrtha anubhava
On the other hand, prama or valid knowledge discloses the real nature of the
object. The framework of a valid knowledge is greatly dependent on three terms
such as visesya, prakara and sambadha. These three are interdependent. Prakara
or adjective is the property of visesya. It resides in its locus i.e. visesya through a
relation (samambdha). Again, this relation is not an ordinary one it is eternal. For
example, 'cow'. It is called 'cow' because it has a property called 'cowness'. In
other words, 'cow' is qualified by 'cowness'. Here, 'cow' is visesya, 'cowness' is
prakara and samambdha brings them together. This samambdha is eternal because
99
without 'cowness' the 'cow' has no existence and vice-versa. However, the
present diagram can clear the picture.
<&-
The real nature of the object (yathartha anubhava) Frame work of a valid knowledge
Again, the system has divided prama into four types like pratyaksa, anumiti,
upamiti and Sdbdabodha. Theses are the ways of knowing. These pramanas have
their respective instruments (karana) and immediate operations (vydpdra) to
produce the result such as:
Knowledge Instrument Operation Result
Pratyaksa Indriya Sannikarsa Pratyaksya
Anumana Vydptijnana Paramarsa Anumiti
Upamdna Sddrsya Atidesa-vdkya Upamiti
Sabda Padajndna Padarthadhi Sdbdabodha
Among these pramanas inference stands first because it has a great utility in our
everyday life. We cannot perceive each and every object directly but that does not
mean that it has no existence. Therefore, for the objects, which are beyond our
perception, we employ this process of acquiring knowledge, which is called
anumana or inference.
100
Inference, according to Navya-nydya, is the instrument of inferential cognition or

, "7
anwniti. Anumiti is produced out of paramarsa , which has two components like
vydptijhdna and paksadharmatdjhdna i.e. paramarsa is the
The process
so far knowledge of the property of the inferential subject (linga),

which is qualified by invariable concomitance8. Here, linga or
hetu is a thing from which the existence of another thing will be inferred and
vydpti is the invariable relationship between the two i.e. the linga and the sadhya
(the thing to be inferred). The process goes like this:
Step - 1: perception of the linga (hetudarsana)

Step - 2 : remembrance of the invariable relationship of hetu and sadhya.
(vydptismarana)
Step - 3: confirmatory cognition (paramarsa)
Step - 4: inferential cognition (anumiti)
Let us fit the above framework of inference into the stock example of the system
i.e. 'parvato vahnimdn dhumdf (the mountain has fire, because it has smoke). A
person, who has experienced the presence of smoke and fire together in the
kitchen, finds a whiff of smoke somewhere in the mountain.
Step - 1: dhuma darsana
Then, he recollects the universal relationship of smoke and fire (which he had
acquired in the past).
Step - 2: vydptismarana
After that he combines step-1 and step-2 into one like 'the smoke, which exists in
the mountain, is qualified by a relationship called invariable concomitance'. This
combined knowledge produces a third one called paramarsa.
Step - 1 + step - 2 = step - 3
101
In the above process of inference the intermediate operation {vyapara) is

paramarsa. The knowledge of invariable concomitance acts as instrument9. Now,
here is a controversy with those Old-logicians. They (Old logicians) hold that the
knowledge of invariable concomitance is the immediate operation (vyapara)10,
whereas, New-logicians substitute paramarsa in its place. Again the old school
holds that the liriga, being known11 as a concomitant, is the instrument of the
inference, where as, the new logicians hold the knowledge of invariable
concomitance as the instrument of inference12. Now let us first define what is
vyapara and karana, then only we will be able to analyze the whole thing.
The vyapara or intermediate operation is defined as the knowledge that which is

produced by the cause and which produces the effect, which effect was produced
Definition of ^ytne cause
• Here, tad means instrument. Hence the whole
vyapa definition follows like this: 'Karanajanyatve sati
karanajanyaphalajanakatvam'. Let us apply the above
definition into the present context. Here, karana is vyaptijnana, so
Vydptijnanajanyatve sati vyaptijnanajanyaphala (anumiti) janakaWam.
Next comes karana. The logicians have interpreted it in two different ways.
According to the first interpretation, karana should be that cause which is
extraordinary and it should possess vyapara (vyaparavat asadharana karanam
karanam). This is the standpoint of Old logicians. And karana, according to new
9
BP, verse - 66 <4NK>kj WR^fr, cf^Tjf 5grf$$r#<T I
10
The fourth chapter carries the whole discussion.
1
' Linga being known means it must be present at the time of inference,
12
BP, verse 67 5llilHH fcT^t ^ tRuf ^ t% I
dv^^<^) Trfcf d^wjuHcfrl ^nxrr^ri

102
logicians, is the cause, which should not disassociate from its effect
{phaldyogavyavachinna karanm karanam). That means it has to be an immediate
Karana, connection with the object. But so far the process of
Its
definition inference, carried by new-logicians, the knowledge of vyapti
and
function does not lead immediately lead to a conclusion. In between a
component called paramarsa comes to the stage. So the problem, therefore, can be
solved, drawing the solution like through such paramarsa or lingaparamarsa the
knowledge of vyapti leads to the conclusion14.
Thus, the process of inference, in general, has been explained in Navya-nyaya
system of Indian logic.
Inferential cognition, as defined by Gangesa, is the knowledge produced by the

knowledge of a Middle term, invariable concomitance with the Major term, exists
in the subject15. And the instrument of this inferential cognition is inference16.
This instrumental cause is lingaparamarsa . Let us take an example like
The hill has fire,
because it has smoke.
Here, pakshadharmatdjhana (the knowledge that the reason exists in the subject)
is the third perception of smoke. And this smoke is not an ordinary smoke. It is
qualified by invariable concomitance {vyaptivisista). So this hetu is invariably
associated with sadhya. This knowledge is called the instrument {karana) of
inference. And the knowledge of a middle term pervaded by a major tenn existing
ID §
in the subject is the immediate operation {vydpdra) . Here, we find a clear picture
14
NTK, p. 262
103
of the process of inference. He also brings a clear distinction of karana and

vydpdra.
Thus, is the process of inference given by Gangesa. Viswandtha Pancdnana, the

author of NydyasiddhdntamuktavalT, has followed the same process. He holds that
in the process of inference the invariable concomitance acts as instrument and
paramarsa acts as the intermediate operation19. Annambhatta, on the other hand,
clarifies the whole thing in his text called Tarkasamgraha following the method of
Gangesa.
We can compare the above process of Gangesa, Viswandtha and
Annambhattahatta with that of Sasadhara, who appeared after Acarya Udayana
but before Gangesa. This study will continue in the second part of this work.
Elements of inference:
In the second chapter the general definitions of the components, which are
involved in the process of inference, have been discussed. Now I proceed to define
these components on the basis of Navya Naiydyikas, especially Gangesa.
After establishing the authority of inference, now Gangesa raises the question that
'what is that vydpti, which is the cause of the inferential cognition?' . To define
vydpti he has given twenty-one tentative definitions, which
as to are traditionally known as provisional definitions
Gangesa
(pun'apaksalaksana) of vydpti and finally rejects all. Thus, at

the end he gives twenty-second definition of vydpti, which is called conclusive
91
definition (siddhdntalaksana) . However, Navya-naiydyikas discuss not only the
19
BP, verse 66 c i | | i | N ^ mm\: ^ M aqrf^pfr T^cF I
20
TCA - \yaptivada H ^ f t R ^ ^ ^ ^ ^5T cZnf^TT I
21
ICNN,p.l26
104
logical character of vydpti but also the manner of its establishment

(vydptigrahopdya). It is established by the frequent observation of the co-existence
of the probans with that of probandum .
The most vital element of inference, as this system regards, is paramarsa. It is

defined by Gangesa as the cognition that the subject has probans, which is
qualified by the invariable concomitance23. In other words, the subject has not the
probans merely, but it is invariably related to the probandum.
Paramarsa
Viswanatha defines paramarsa as the cognition mat this

mountain has smoke, which is invariably related to the probandum24. Annambhatta
identifies it {paramarsa) with inference25. However, this concept has covered a
wider area in this system.
Hetvdbhasa:
Logicians have identified some obstructing factors of inference, for which a valid
inference cannot be produced. Navya-naiyayikas have defined two types of
obstructing factors of inference such as Hetvdbhasa and Uddharandbhdsa26.
Gangesa, at the very outset of Hetvdbhasa prakarana, mentions that by exposing
0*7
fallacious in the argument of his opponent, one can achieve truth and victory . He
gives three alternative definitions of Hetvdbhasa , express that fallacy is an object
22
TD, anumanakhaiula tjHIîilcij'ilk^ ^TTÎ-^TTEpfjft^lT W ^ I R < ^ H . . .
23
TCA, paramarsaprakarana cyifklRlRlbd t ^ w f c W F f WTH^T |
24
NSM, verse 68 <ôij|U>jtjHc|Hls4 ^^- ^fcT §TR WT^fr |
25
TSA - d^NxMHSfls^HH I
26
AP, p. 362
" TCA, Hetvabhdsaprakarana
28
Ibid. cp. c r a i ^ f c l c h K u ^ d m N M f c l i J l R m y n s J §lHfo*Wc<lH I
105
of knowledge, which obstructs an inference. Gangesa divides Hetvabhasa into

five such as Savyabhicdra, Viruddha, Satpratipaksa, Asiddha and Badhita .
Savyabhicdra is a probans in which abides a character, the possessor of which

causes that presence of two alternatives, which produces doubt in the
probandum30. This has been sub-divided into three like Sadharana, Asddharana
and Anupsamharii. Sadharana or too-general fallacy abides in the locus of
probandum as well as in its absence31. For example,
This hill has smoke,
because it has fire.
The Asddharana fallacy abides neither in the locus nor in its absence. For
example,
This hill is full of smoke,
because it is full of ether. (Here the reason is ether, which has no locus).
A non-conclusive fallacy Anupsamharii Hetvabhasa has no either affirmative or
negative examples. For example:
All things are impermanent,
because they are knowable.
The fallacious reason, which is the counterpart of that non-existent, which

constantly accompanies the major term, is called contradictory fallacy Viruddha
Hetvabhasa. For example,
This hill is full of fire,
9
Ibid, cl ^ ^Tcqf^^KR^^ÎciMaTTfeg^Tf^TTT V5Q |
0
Ibid. mH >H^M^Hcbc^(^îllM>r«rnîT^4cii5iHRfc|<q^ ^rfcT t ^ F M T ^ (WrfÎRT)
1
The whole discussion on hetvabhasa is based on hetvdbhasa-samanya-nirukti oj'Tatvacintamani
106
because it is full of water.

Satpratipaksa fallacy seeks to establish the existence of probandum on one hand
and provides the reason, which seeks to establish the non-existence of it. For
example,
One reason - Sound is eternal, because it is audible.
Opposite reason - Sound is non-eternal, because it is produced.
Asiddha Hetvabhasa is of three kinds viz. srayasiddha, svarupasiddha and

vydpyatv asiddha. The fallacy, whose locus is not known, is called as
Asraydsiddha hetvabhasa. That is called svarupasiddha Hetvabhasa, whose
nature is not known to exist in the locus. And if the nature of a fallacious reason is
not useful in establishing its invariable concomitance with the major term then that
will be called as vydpyat\'dsiddha heU'dbhdsa, which has been subdivided into
three like: sdsiddhi, hetx'dsiddhi and vydptyasiddhi.
Badhita fallacy occurs when the major term, which is assigned to minor term, does
not exist in it. For example,
Fire is cold,
because it is a substance
Gangesa has divided it into ten such as paksa pratyaksa badhita, paksa anumdna
badhita, paksa sabdabddhita etc. Viswandtha follows Gangesa. He changes
Badhita into kdldtyapdista, which has no further divisions. Annambhatta holds the
views of Gangesa.
107
The division:
Gangesa has divided inference into three on the basis of the nature of vyapti, such
as kevalanvayi, kevalavyatireki and anvaya-vyatirekf2. This classification of
inference, in fact, is not new to the system. It has its origination in Parisuddhi, the
text of Udayndedrya. Gangesa does not follow the tradition of classifying
inference into three such as purvavat, sesavat and sdmdnyatodrsta (which is given
by Vdtsyayana). The kevalanvayi anumana, as Gangesa defines, is an inference in
which there is no negative example3 . For example,
It is namable,
because it is knowable.
An exclusively negative inference or kevalavyatireki anumana has no affirmative
example. Such as:
The Earth is different from the things,
because it possesses smell.
The inference, which has both affirmative and negative examples, is called
anvaya-vyatireki anumana . For example,
The hill has fire,
because it has smoke (affirmative example)
as a kitchen and not as a lake (negative example)
We do not find such type of classification in Viswndtha's text. According to
Annambhatta, above classification is exclusively meant for hetuM.
108
Another division
We also find another type of division of inference such as: svartha (inference for
one's self) and pardrtha (inference for others sake) the former is based on
psychology whereas the later one is totally based on logic.
A person, who repeatedly observes the togetherness (invariably connected) of

Svarthanumana smoke and fire in the kitchen etc., sees smoke from a
distance. After seeing it he recollects that 'where there is smoke there is fire'.
Then he realizes that the hill has smoke, which is invariably connected with fire.
Finally, he reaches in the conclusion that the hill has fire. This process is called the
process of inference for one's self.
After inferring for himself same person proceeds to convince to others by

employing five-membered syllogism. This is called an inference for the sake of
others .the process goes on like this:
1. first of all he states that 'this hill has fire' - Pratijnd
Pararthan
-umana 2. because it has smoke - Hetu

3. where there is smoke there is fire like kitchen - Uddharana
4. the smoke, possessed by the hill, is fire-pervaded - Upanaya
5. therefore, the hill has fire - Nigamana
The above five-membered syllogism generates the inferential cognition of fire in

other's mind. Hence it is called pardrthdnumdna. Viswandtha remains silent on
this point. Annambhatta follows Gangesa35.
Members of syllogism:
Gangesa strongly opposes to the definition of avayavas given by Old logicians .
Avayavas, according to him, is a sentence that gives rise to the knowledge, which
Vide TSA.
109
contributes to the production of an entire knowledge, which produces

consideration . He also rejects the definition of pratijhd saying that 'the
declaration of probandum cannot be a definition of pratijhd, because it suffers
from over-application . He defines pratijhd as 'is a sentence, which causes
knowledge, whose object is the same as that of the conclusion and which
contributes to the production of entire knowledge, which produces
consideration'39. Gangesa has prepared three alternative definitions of hetu40. All
his definitions convey the meanings like 'hetu produces the knowledge, that which
gives rise to the consideration'. He has divided hetu into three classes like
anvayihetu, vyatirekihetu and anvayavyatirekahetu41.
There is no difference of opinion with old-logicians with regard to the definition

of udaharana. The only difference is its division. Old-logicians divides it under
the name sddharmya and vaidharmya, whereas, Gangesa names it as anvaya and
vyatireka42. Gangesa defines Upanaya as the cause of consideration and it is of
two kinds such as anvaya-upnaya and vyatireka-upanaya43. Nigamana produces
the knowledge of major term as indicated by that of middle term through its
invariable concomitance. Hence, much difference is not seen on this point.
* * * $ • * * *
7
Ibid. ai^Rlxi^HcbKUI-fer^T-^fPRt W$M?P !JI«{5lH-v>HcMlcKi ^TRJT |
8
Ibid. cT^ ^T Vfom W2T f ^ F - îrarq^sfaoA||i^: I
19
Ibid.
10
Ibid. cf>. %^m xf &<tfi\$\4>Hv\\'tg\fc\&mWi\ WfiiWfi ^ s l l - M ^ c b x H I U l R q ^
subtilv*H* tl^R^Rb Htw«j<rtH I
1
Ibid.
2
AP,p. 356
3
Ibid, p. 459
Selected Readings for Fifth chapter
The Navya-Nydya Theory of Knowledge
-Dr. L.C. Mullati, Karnataka University, Dharwad, 1977
Studies in Language, Logic and Epistemology
-Prof. V.N. Jha, Pratibha Prakashan, New Delhi, 1984
Anumdna pramdna
TSA Tarkasamgraha, anumdnakhanda

BP Bhasdpariccheda
NTK Nyaya theory of knowledge
ICNN Invariable concomitance in
Navya-Nyaya
TD TarkadTpikd
NSM Ndyasiddhdntamuktdvali
AP Anumdna pramdna
-Î|A|C1|(T|CF> 1.1.H
(The first perception of the probans occurs from the knowledge of the relation of
probans and probandum. Perception of probans, somewhere in the mountain, is the
second perception of probans. From frequent perception of probans leaves
impression like "smoke is pervaded by fire", which is followed by another
perception such as: 'the smoke, which is invariably related to fire, exists in the
mountain'. This is called the third perception of the probans.)
Chapter VI
Concept of Trtiyalingapardmarsa and its necessity
Is it paramarsa, lingapardmarsa or
trtiyalihgaparamarsal
Nature and function of paramarsa
Chapter at a glance Is it karana or vydpara?
Its necessity
Concept of Trtiyalingaparamarsa and its necessity ' '0
Sixth chapter
Concept of Trtiyalingaparamarsa and its necessity
Naiydyikas have used the term paramarsa with a nomenclature. In some texts it is
used as lirigapardmarsa,1 somewhere it is trtiyalingaparamarsa2. Somewhere it is
trtiyalingadarsana3 whereas, somewhere it is only paramarsa4'. Though paramarsa
_ , has been used in different names still its nature and function has
Though
paramarsa been kept unchanged. It is called lirigapardmarsa because it is
has
different names such a knowledge of probans, which establishes the inferring
still its -
nature and object through a relation called invariable concomitance . This is
. mn called trtiyalingaparamarsa (third time perception of the
is Siinic
probans) because it is the last of the cognition of smoke, which is

essential for the inference of fire. Let us see how it happens. Someone, for the first
time, perceives smoke invariably connected with fire in a place like kitchen etc. A
frequent perception of such case generates knowledge in his mind like "vahnivydpya
dhuma"" (smoke is pervaded by fire). That which covers a wider area is called
vyapaka. For example, fire. And that which covers a less area is called vyapya. For
example, smoke. So, fire is the pervader and smoke is the pervaded one. This can be
presented in the following way:
l
N V 1.1.5 clfcJ^HRlH WQti ^ W ScWITKft ^ ? T T ^ H l ^ i m i H ^ M H ^ H M wfo I
2
TCA. w t f cîR^RlbdslH ff^Ftf? RRl^Rlk^JsIM V& cfT ^ C T I ^ R H ^ N X M H ^ : -d^ftMH^T,
3
cHTTRpFffel ftcMcT^K^TT^n^ff^ . ^ A | [ ^ l ^ < - | * c > M H P l f t f I --M | i) q R tj i|
Sidddnta candrodaya a commentary on Tarkasamgraha by Srikrrsna Dhurjati as it is

quoted in Athale and Bodas p . 254 cZTTf^faoto ^ P T ^ JIHilcflfd fcTff cf^ £|HlRw>W W^RT?
Concept of Trtiyalingaparamarsa and its necessity 11!
Fire
Smoke
This knowledge of invariably connected fire and smoke in the kitchen etc. is called
first perception of the probans or M^HRH^M wr?f. This knowledge leaves an impression
like 'smoke is pervaded by fire'. After that another smoke happens to be seen by the
same person somewhere in mountain etc. This is called the second perception of
smoke or fscfkfeiî vwvft. Here the recollection of the impression, left by the first
perception of the probans, will occur. Again this recollection does not produce the
result i.e. inferential cognition immediately. There will be a further recollection of
smoke. This smoke is not like an ordinary one. It is that smoke, which exists in the
inferential subject, is invariably connected with fire. This is called third perception of
the probans or cidkfc-i-s-'i w*ref, which is known as an immediately preceding cause
(carama karana) of inferential cognition. In short, the probans i.e. smoke is known
for the first time when we observe its invariable relation with the probandum i.e. fire
in the kitchen etc. It is known for second time when it is related to the inferential
subject. It is considered for the third time when we know it as a property of the
inferential subject, which is invariably related to the probandum. Immediately after
that the inferential cognition arises6. It is also called trtiyalingadarsana7, because the
linga or smoke is being perceived third time in the whole process. Finally, it is called
pardrnarsa or confirmatory cognition because it confirms that the probans, exists in
the inferential subject, is qualified by invariable concomitance ( W J M ^ 3T^T ifcf
! 12
Concept of Trtiyulingaparamarsu and its necessity
qTPT?f:).
Nature of paramarsa:
Paramarsa, according to Gangesa, is the cause of anumiti i.e. inferential cognition.
He defines it as the knowledge of the property of the subject, which is qualified by
o
invariable concomitance . This paramarsa produces anumiti. In other words, anumiti

is ultimately dependent on paramarsa, which is again dependent on the probans. If it
is a sad-hetu, it produces a valid paramarsa and a valid paramarsa generates a valid
inference. Mathuranatha Tarkavdgisa, a celebrated commentator of Tah'acintdmani,
clarifies Garigesa's statement. He holds paksadharma as paksasambadha and
vyaptivisista means vyaptivisiste jnanarn (saptamT tatpurusa) keeping its meaning
„„. . , „ unchanged . The same idea reflects in
What is paramarsa/
NydyasiddhantamuktdvalT. Visvanatha defines paramarsa
]0
as the the
means knowledge
knowledge,
that which
the concomitant
confirms the(vyapya)
relation exists
of whatin isthe
possessed
subject of
(paksa)
invariable
. It
concomitanc is called paramarsa . Thus, paramarsa has been defined by Navy a
concomitance
Naiyayikas.
Naiyayikas have followed a systematic process of inference in which paramarsa is

supposed to be a part. It is a part because it helps in forming the whole i.e. inference.
Then the basic question arises like 'How does that occur,
How does it occur? ,. , , , ,. , n, - , •
which helps something else in occunng? paramarsa is a
Concept of Trtiyalingaparamarsa and its necessity 113
combination of two knowledges. The first one is a recollection (of vyapti) and the
second one is an experience (of that recollection i.e. that vydptivisista hetu exists in
the mountain) i.e. hetu is vydptivisista.
The process continues like this: first of all a person perceives an invariably connected
smoke in the kitchen etc. From this perception he comes to know that 'smoke is
pervaded by fire'. Then he proceeds towards mountain where he happens to see
smoke from a distance (hetudarsana). After seeing it he recollects the relationship of
smoke and fire like 'where there is smoke there is fire' (vyaptismarana). After that he
joins the previous knowledge (dhumavahnivyapya) with the current one {yatra dhuma
tatra vahni) and forms a third kind of knowledge i.e. the smoke, which exists in the
mountain, is invariably connected by fire. This knowledge is known as paramarsa or
confirmatory cognition. In the very next moment the knowledge of fire i.e. 'mountain
possesses fire' arises there. This is called anumiti and the component, which produces
this knowledge, is called paramarsa.
Paramarsa is vydptivisista-paksadharmatdjhdna i.e. the knowledge of

paksadharmata, which is vydptivisista. This shows that paramarsa has two
components. The first component is vydptivisista, it means 'that which is qualified
by vyaptV. Here, the word 'that' means 'something', which is
Constituents characterized by vyapti. The second component is paksadharmata. It is
of
paramarsa defined as "vydpyasya hetoh paksavrttitvatrC i.e. the existence of hetu
on paksa. A question can be raised here like 'are all the objects, existing
on mountain, called paksadharmatdV The answer is obviously 'no'. Because though
all smokes are pervaded by fire still we cannot say that all smokes are paksadharma
because there should be a desire to infer. So, a particular smoke, which is present in
the inferential subject at the time of inference, will be called as paksadharma and
paksadharmata is the property, which exists in the paksadharma. These are two
Concept of Trhyalihgaparamarsa and its necessity 114
separate components. When they come together they constitute a whole and that is
called paramarsa. So, paramarsa is a knowledge of the property of the mountain
qualified by invariable concomitance.
Paramarsa functions as vyapara in the process of inference, because the inferential

cognition takes place through it. There is a controversy regarding the function of
paramarsa. Old logicians hold it as an instrument, because it produces
paramarsa an i mm ediate result whereas, new logicians oppose it. If it is a karana,
argue new logicians, then there will not arise inferential cognition from the smoke
which was in the past time and which will be in the future time13.
We have discussed the definition of karana and vyapara in the last chapter. Now we
will see whether this paramarsa acts as karana or vyapara in the process of
Paramarsa is the inference. In this connection Gangesa opines thus:
vyapara of the process
of paramarsa is not an instrument because it has absence of
inference
intermediate function14. He strongly holds paramarsa as

an ultimate reason (carama karana) of the process of inference15. The later logicians
have accepted Gangesa. But it is Annambhatta, who holds paramarsa as an
instrument of the process of inference16.
1
Bhasapariccheda-68
13
NSM-67 3T5f HlrSVll^ ol||U|<^H 5llilHM fcH^Hjfafcl <^f%....Wjftcfl feT^t ^T°t XWITKHHICH
feTg#T fcH^H fa^H ^ f a f a ' } W ^ f t f c T ^RW^ f c ^ W T d^l^lH'^Nlfcifrl I
14
TCA,paramarfaprakaranam^ cijmW^NI^ IKIH*1: cWÎ*^ cT^fr, t ^ J cJUf^TsTFT WW M^H^
Concept of Trtiyalihgaparamarsa and its necessity 115
Mimamsakas argue like 'why anumiti does not arise from the second perception of the
probans (f§cfkfere*i <Ki*f?f). Here also the doubt about the existence of fire is certainly
there. What is the need of accepting a third time perception of
Its necessity . . , .,17 . „ . • , „ , ,
the probans then? . Yes you are right , reply Naiydyikas, you
just imagine of a person who has not ascertained vyapti or has
forgotten it, will he be able to draw inference? Therefore, the third time perception of
probans is very much essential, because after observing smoke in the subject and after
recollecting vyapti a third cognition is necessary to cause the inferential cognition.
This third cognition is called paramarsa or lihgaparamarsa or
trtiyalihgaparamarsa''' .
In TarkadTpika Annambhatta shows the necessity of accepting trtiyalihgaparamarsa

saying 'there is necessity of accepting trtiyalihgaparamarsa otherwise the expression
like 'hill has smoke, which is invariably connected by fire' cannot generate verbal
understanding. Therefore, for uniformity in logic the concept should be accepted.' 19
*UVMcMi£|<fcHHIull'H|cH ^M^KI ^WHlkMId I Tarkabhdsa, anumanakhanda.
^fcT ^ ^ fcT^TWFRhl Ibid.

(First of all there will arise the remembrance or experience of invariable
concomitance and thereafter will arise the cognition of smoke, which is located in
the subject, in which there is a property called "smokeness', which is delimitor of
the state of being pervaded (by the probamdum) appears as the qualifier.
Immediately after that the inferential cognition arises without any hindrance.
When this can be explained in the above manner, it is cumbrousome to postulate a
qualified cognition (i.e. Pardrnarsa) in the middle.)
Chapter VII
Objection raised against Trtiyalihgapardmarsa
• Objections
Objection raised against Trtiyalingaparamarsa 116
Seventh chapter
Objection raised against Trtiyalingaparamarsa
"Your conviction that the third perception is the ultimate cause of the inferential
cognition is totally baseless." This is just a challenge of MTmamsakas put to the
Nayiyayikas. The Nayiyayikas, on the other hand, entertain such type of
challenging discussions in an enthusiastic manner. It is Raghunatha Siromani, who
invited others saying "dusyam me vacah param nipunam vibhavya" (I welcome
you to criticize me, but think hundred times before doing so). By tracing the
history of Nydya-Vaisesika system of Indian intellectual tradition it can be
observed that the entire Nyaya system has been developed in a dialogue form.
Dialogue between two systems, viz. Nyaya and Buddhist, Nyaya and MTmamsa
etc.
In the same way, Lingaparamarsavada, a chapter of the text NyayasiddhantadvTpa

of Sasadhara, is written in a dialogue form. Here, there is a controversy between
Nyaya and Mlmamsd regarding paramarsa.
Both the participants have strong arguments over the issue called paramarsa like it
should be regarded as a special component in the process of inference or not. But
both have the agreement over the statement that the knowledge of the existence of
the probans on the subject is the cause to the inferential cognition.
However, this chapter deals with the objections raised by the opponents. The
entire dialogue has been quoted from the chapter called Lingaparamarsavada of
Sasadhara.
Objection raised against Trtiyalirigupararnarsa 117
Objection no-1: - Is it the case that a qualified cognition is the cause of an

inferential cognition on account of positive or negative concomitance? Or, in some
other way?
Here the author presents the position of Mimamsakas. Naiydyikas
standpoint is that the qualified cognition is the cause of an inferential cognition.
'But in what form?' asks Mimdmsaka. Is that the cause of an inferential cognition
on account of either positive or negative concomitance? Or in some other way?
These are two alternatives left with Naiydyikas. The first alternative can be
explained in this way: positive concomitance means the presence of something is
dependent upon the presence of another thing. For example,
+X = +Y (positive concomitance)
Negative concomitance means the absence of something occurs due to the absence
of another thing. For example,
-X = -Y (negative concomitance)
Now, Mimamsakas are asking to the Naiydyikas that is it the case that the presence
of that qualified knowledge causes the inference and its absence causes the
absence of it? This can be presented in this way:
+qualified knowledge = +inference
-qualified knowledge = -inference
Or do you have any other way out to prove this?
Objection no-2: - You are {Naiydyikas) postulating the causal relationship between
paramarsa and anumiti, but it is not acceptable, because the force of the principle
of elimination gets exhausted over there where we find a form 'the state of being a
smoke', which is different from the form 'the state of being something'.
Objection raised against Trtiyulinguparumarsa 118
Again, it is not possible to postulate that the paramarsa is the cause of inferential
cognition, because when that paramarsa itself is not there then how will it be the
cause of anumitP.
Therefore, the paramarsa , as postulated by you (Naiydyikas), cannot be
established.
Objection no-3: - Well, if you claim saying that the cause of direct perception is
stronger than the cause of inferential cognition then this is also not acceptable,
because in that case it will lead to a continuous flow of the third perception and
also because it will lead to contingency of awareness.
Objection no-4: - Again, your saying that 'the powerfulness of the factors of an
inferential cognition will lead to a continuous flow of inferential cognition and in
that case, the knowledge that T am inferring' will not arise' is also not acceptable,
because the earlier factors have already produced the result, so other factors will
not arise since the result has not already come into existence.
Therefore, there is no possibility of the third perception due to the absence
of the cause.
Objection no-5: - Again, if you say that 'well, that does not matter if paramarsa
is not possible there (i.e. paramarsa is not the cause of inferential cognition) but
let the senses be the cause there (i.e. in the process of inference)'. This is also not
being accepted, because the absence of the contact with the object.
Objection no-6: - Again, if you say that when the smoke is presented by another
knowledge (i.e. remembrance) then there will arise a mental perception of
paramarsa . This is also not right, because mind has no capacity to cause a
cognition of an external object.
Objection raised against Trtiyalingaparamarsa 119
Objection no-7: - You cannot say that 'let us assume that the impression of that
object will generate the mental perception of the pardmarsa, because impression
of what? Since there was no experience of mountain possessing smoke, which is
pervaded by fire.
Objection no-8: - Again, how do you Naiydyikas support third perception?

Because it is not possible to know the state of being pervaded by the difference of
other than the Earth in the probans, namely, the Prithivitva.
Objection no-9: - Well, if you say that whatever is pervaded in reality, the
knowledge of that qualified by 'some-thing-ness' is the cause of the inferential
cognition then what type of inferential cognition is going to be established over
there? The inferential cognition presented by a particular probans or a probans in
general?
Objection no-10: - Again, if you say that whatever is not pervaded by fire in reality
such as a beam of the dust, even from the qualified knowledge of that, there arises
the inferential cognition of fire. This is also not accepted, because I (Mlmdmsaka)
don't accept that such an inferential cognition arises there. Assuming your
(Naiydyikas) standpoint to be true if we analyze it then such an inferential
cognition can be explained to have arisen from the cognition of that which is, in
reality, pervaded by fire.
Therefore, your conviction that the third perception is the ultimate cause of
inferential cognition is totally baseless.
Hence, there is no cause and effect relation between pardmarsa and anumiti
* * * • $ & * * *
-fei^ J IM^HîcJK
(Still, if you (Mimamsakas) are asking in a friendly manner 'what is the cause
there?' then listen: - Mind is the only cause there through the relationship of
remembrance of smoke, which will arise through relationship of inherence with
what is connected with it.)
Chapter VIII
Sasadhara^s reply to the objections
• Replies
120
Sasadhara's reply to the objections
Eighth chapter
This chapter presents the replies made by Sasadhara
Your (Mimdmsakas) objection:

Is it the case that a qualified cognition is the cause of an inferential cognition on
account of positive or negative concomitance? Or, in some other way?
Our (Naiydyikas) reply (no - 1): - In that case the second alternative holds true.
Because even if, after the remembrance of the invariable concomitance, there can
not arise an inferential cognition though there arises a cognition of existence of the
smoke 'in some form' in the subject. And this 'other form' is nothing but 'the state
of being pervaded by the probandum by the thing of elimination'.
Thus, the qualified cognition i.e. paramarsa gets established.
Your objection:
You are postulating that paramarsa cannot be a causal factor to anumiti, because
the force of the principal of elimination gets exhausted there where we find a form
of 'the state of being smoke', which is different from the form 'the state of being
something'.
Again, paramarsa is not the cause, because when it {paramarsa ) itself is

not there how it will be the cause of anumiti.
Our reply (no - 2): - The process of inference has not been terminated there,
because all the necessary factors have not been taken into consideration.
Sasadhara's reply to the objections 121
Again, to your second objection that paramarsa itself is not there, we have this to
say that paramarsa is there but you are not accepting it, because you follow a
mechanical process of inference whereas, we are interested in the epistemology of
inferential cognition.
Hence, to generate a particular variety of anumiti we need a confirmatory
perception.
Thus, the paramarsa gets established.
Your objection:
Your conviction that the cause of direct perception is not stronger than the cause
of inferential cognition, because it will lead to a continuous flow of the third
perception and also it will lead to contingency of awareness.
Our reply (no - 3): -On answer to this we say that there is no such chance of chain
since after paramarsa the anumiti will be produced and there will end one
process. If someone wants to infer again then a fresh process will begin.
Therefore, there is no chance of any continuity of paramarsa .
Your objection:
You are not accepting that the powerfulness of the factors of an inferential
cognition will lead to a continuous flow of inferential cognition and in that case
the knowledge that T am inferring' will not arise. And in the reply you are telling
that 'if you accept it then the result will not come into being, because the earlier
factors have produced the result'.
Our reply (no - 4): - This is also not tenable, because once the pardmarsa-stagQ is
reached it will definitely produce the result namely, the anumiti.
122
Your objection:
You are telling that senses cannot be the cause there, because of the absence of the
contact with the object.
Our reply (no - 5): - This is absurd, because this is not the process of perception
and hence we cannot expect the contact with object.
Your objection:
You are telling that when the smoke is presented by another knowledge
(remembrance) then there will not arise a mental perception of paramarsa,
because mind has no capacity to cause a cognition of an internal object.
Our reply (no - 6): - This is also not tenable. Because mind is the common cause
of any cognition and over and above this it can reveal external objects too.
Moreover, paramarsa is not a mental perception.
Your objection:
You are also telling that the impression of that object cannot generate the mental
perception of paramarsa , because there was no experience of mountain
possessing smoke, which is pervaded by fire.
Our reply (no - 7): - This is also not reasonable because the inferential cognition
of fire is not a remembrance.
Your objection:
Again, you are telling that the third perception should not be entertained, because
it is not possible to know the state of being pervaded by the difference of other
than the Earth in the probans, namely, Prthivitva.
s
123
Our reply (no u 8): - This is not a case of kevala-vyatireki-anumdna. Therefore,

your objection does not hold good here.
Your objection:
You don't accept that whatever is pervaded in reality, the knowledge of that
qualified by 'some-thing-ness' is the cause of inferential cognition. And you want
to know whether such inferential cognition is presented by a particular probans or
a probans in general.
Our reply (no - 9): - The invariable concomitance is always between two
universals and hence the objection, raised by you, holds no ground.
Your objection:
You don't accept that whatever is not pervaded by fire in reality such as a beam of
dust, even from the qualified knowledge of that there arises the inferential
cognition of fire.
Our reply (no - 10): - A false cognition of vydpti will result into false anumiti. This
does not rule out the possibility of a true cognition of vydpti resulting into a valid
inferential cognition.
There in the process of a true anumiti, the knowledge of paramarsa is absolutely

essential.
Thus, there is a causal relationship between paramarsa and anumiti.
***.$& * * *
(There cannot arise an inferential cognition of fire even if there arises a cognition
of the existence of the smoke 'in some form' in the subject after the remembrance
of the invariable concomitance. Therefore, it is necessary to accept that the
cognition of the smoke on the subject must be cause in 'some other form' before
the arising of the inferential cognition. And 'other form' is nothing but 'the state of
being pervaded by the probandum by the theory of elimination and thus the
qualified cognition (i.e. the Pardmarsa) gets established.)
Conclusion
Conclusion 124
Conclusion
Form whatever we have discussed so far; the following conclusions can be drawn:
1) Mimdmsd system of Indian philosophy holds that from the knowledge of

invariable concomitance (vyaptijnana) and the knowledge of the existence of
the hetu on the inferential subject {paksadharmatajnana) the inferential
cognition can arise. If it is asked that 'what is the cause of an inferential
cognition?' they will answer that 'the knowledge of the probans, which is an
attribute of the inferential subject, is the cause there'. So, in this analysis there
is no scope of accepting a qualified cognition in between.
2) Secondly, for the inference of others' sake {pararthdnumdna) they require only
three steps starting from either pratijhd or uddharana. According to them,
upanaya cannot be a factor of generating inferential cognition in others.
If this is accepted then the following problem may arise.
1) If from vyaptijnana and paksadharmatajnana the inferential cognition arises

then the following problem may occur. Suppose we will imagine of a person
(say, X), who has the knowledge of vydpti and another person (say, Y), who
has no knowledge of vydpti, happens to see smoke in the mountain then in this
case anumiti may arise in 'Y', which is illogical. But if we accept pardmarsa
this problem won't arise, because pardmarsa is a combination of these two
knowledges.
Conclusion 125
2) Secondly, unless it is confirmed that the mark or probans, which is invariably

related to the probandum, exists on the subject, the conclusion namely, 'the
subject possesses probandum' cannot be drawn.
3) Moreover, there will arise no verbal understanding from the sentence called
upanaya if paramarsa is not accepted as a step in the process of inferential
cognition.
The form of Inference with only three members comes closer to the form of
Inference of Western tradition. In Western tradition the form is as follows:
'p' implies 'q'
there is 'p'
.'. there is q
The first step may be compared with the concept of vyapti in the Indian tradition,
the second step may be compared with dvitiya-lingadarsana and the third step
may be compared with nigamana.
The Naiyayikas process of Inference is not a mechanical form. They are interested
in the epistemology of an inferential cognition. Hence, it is necessary for them to
postulate a state of paramarsa or a confirmatory perception or cognition to
account for the entire process of generating a particular variety of cognition called
anumiti. Therefore, in the Nyaya-Vaisesika form of Inference the fourth avayava
or sentence is absolutely necessary.
***$$.***

Anumana PDF

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Anumana PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Anumana PDF

Uploaded by

Copyright:

Available Formats

Part - 1

followed by some other knowledge. In other words, it is a

The definition of anumana has been discussed in an ascending manner starting

'What is the necessity of accepting inference, only one proof or means of

Vdcaspati Misra, a prominent scholar of Nyaya tradition with other philosophers,

Ground Naiyayikas: Well, it's up to you.

The Nyaya-Vaisesika system of Indian philosophy has developed a systematic

In Nydya system the classifications of inference are seen in various forms. It is

Sesavat and samdnyatodrsta whereas; Navya-Nyaya

happens to see a string of smoke somewhere on the

Here, pratijha or assertion is a statement or proposition, which is to be

NBh 1.1.5, arsrar tj^crftfr...

A sesavat type of inference is the just opposite to purx'avat25. It is inferred

Samanyatodrsta type of inference is the perception of some object at some time

The third type of classification is specially designed by new logicians . But we

This inference occurs when there is the relation of invariable concomitance

This inference is obtained only through the negative invariable concomitance.

Classification on the Classification on the Classification on

Pararth Punwat Sesavat Samanyatodrsta Anvayivyatireki

Samkhya system considers inference as the knowledge of the pervader (vyapya)

system inference is of three kinds such as: purvavat,

Samkaracarya, the founder of avaita vedanta, holds that there is no necessity of

of pervasion, functioning as cognition of pervasion. .

So, the faultless mark, qualified by invariable concomitance, is called inference.

Pramdna, according to Jaina logicians, is 'svapardbhdsi'. That means it is

" Anumana pramdna, p. 235

The Buddhist School of Indian philosophy accepts pramana as infallible

Inference, according to this school, is cognition of an object through its mark.

nyayavayavas like an inferential subject, a probandum and an example . This

That is like this;

According to Buddhist logicians hetu is of two kinds sadhamya and vaidhannya.

Therefore, the vyapti is proved neither by perception nor by non-perception. But

Paksadharmata is another main component of the process of inference. A mere

Acdrya Dignaga has mentioned four types of fallacious subjects viz.

3. An inferential ground denying the synonyms (Pratitiviruddha)

And if such fallacious ground happens to be an inferential subject then it

Fallacious reasons, according to Dignaga55, is of three kinds such as: asiddha

a) Ubhaydsiddha: this kind of fallacious reason is unacceptable as a predicate to

c) Samdigdhasiddha: this type of fallacious reason has a doubtful existence. For

d) Ashraydsiddha: here its very existence is doubtful. For example,

2) Anaikdntika: Again, anaikdntika is of six kinds:

c. SapaksaikadesavrittinnpaksvydpT aneikantika: It resides in some similar

d. Vipaksaikadesavrittih sapaksvyapi aneikantika: It resides in some unlike

e. Ubhayapaksaikadesavritti aneikantika: this type of fallacious reason resides

f. Viruddhdbhicari anaikdntika: that which has a contradictory invariable

3) Biruddha: it is again of four kinds such as:

b) Dharmavisesa vipantasadhana: here, the sadhana or hetu itself is contradictory

c) Dharmisvarupavipantasadhana: It occurs when the middle term is inconsistant

d) Dharmivisesavipantasadhana: That which proves the opposite of the particular

(Vaidharmaya Drstantabhasas) and homogeneous fallacies

a) Sadhandsiddha: An example non-cooperating with the thing that which helps to

b) Sadhyadharmasiddha: An example non-cooperating with that, which is going

c) Ubhayasiddha: An example which is neither co-operative with middle term nor

d) Ananvaya: An example showing lack of universal connection between probans

e) Viparitdnvaya: Here, this example shows a reverse invariable concomitance

Similarly, heterogeneous examples also do not serve the purpose. He has

j) Vipantavyatireka (The case where the absence of probans and probandum is

Six ways of knowing