Documente Academic
Documente Profesional
Documente Cultură
The main part of the book consists of a collection of ten articles which have
been written by various authors on different issues of what is commonly
called “Indian Logic”. Apart from a short preface (pp. vii–ix), the book
also contains an extensive introduction (pp. 1–25) and a bibliography of
primary and secondary literature on “Indian Logic” (pp. 216–221). The
ten chapters of the main part of the book represent reprints of earlier
published articles or excerpts from books: 1. Henry T. Colebrooke (1824),
‘The Philosophy of the Hindus: On the Nyāya and Vaiśes.ika Systems’
(pp. 26–58), 2. Max Müller (1853), ‘On Indian Logic’ (pp. 59–74), 3.
H.N. Randle (1924), ‘A Note on the Indian Syllogism’ (pp. 75–92), 4.
Stanisław Schayer (1932–1933), ‘Studies in Indian Logic’ (translated
by Joerg Tuske) (pp. 93–101), 5. Stanisław Schayer (1933), ‘On the
Method of Research into Nyāya’ (translated by Joerg Tuske) (pp. 102–
109), 6. Daniel H.H. Ingalls (1955), ‘Logic in India’ (pp. 110–116), 7.
I.M. Bochenski (1956), ‘The Indian Variety of Logic’ (pp. 117–150), 8.
J.F. Staal (1973), ‘The Concept of Paks.a in Indian Logic’ (pp. 151–161),
9. Sibajiban Bhattacharyya (1987) ‘Some Aspects of the Navya-Nyāya
Theory of Inference’ (pp. 162–182), 10. Bimal Krishna Matilal (1998),
‘Introducing Indian Logic’ (pp. 183–215).
On the one hand, the book is meant as an introduction into the area
of “Indian Logic” as well as a survey about Western research in this field
comprising a period of more than 150 years. In these respects the “Reader”
can be usefully employed by both specialists and non-specialists in order
to obtain a fairly accurate picture about the domain. In particular one must
welcome the fact that two important articles by St. Schayer have been
made accessible to English speaking readers. On the other hand, it appears
that the editor has also something else as his aim: To combat the cliché
of a fundamental disparity between the Eastern and the Western philo-
sophies inasmuch as, whereas philosophy in the West “looks outward”,
Review article on: Jonardon Ganeri (ed.), Indian Logic. A Reader. Richmond, Surrey:
Curzon Press 2001, pp. 221.
“is concerned with Logic and with the presuppositions of scientific know-
ledge” and considers an “analysis and clarification of the concepts which
ground scientific enquiry” as its job, Eastern, and in particular Indian,
philosophy is concerned with a “mysterious and fundamental sort of self-
knowledge” and therefore can hardly be regarded even as asking the same
questions. The introductory chapter, which is entitled “Introduction: Indian
Logic and the Colonization of Reason”, begins with citations from an
article written in 1955 by the Wykeham Professor of Logic in Oxford,
H.H. Price, in which such a belief of a “vast chasm” separating the two
philosophical traditions is expressed. An interesting point is that, as the
author demonstrates on the subsequent pages (2–4), representatives of the
“Neo-Hindu movement” contributed to the persistence of this cliché by
propagating a picture of the Indian tradition as a “Vedantic culture”, a
picture in which all elements which militated against this view had been
suppressed or relegated to the background, and this was probably due to
the “intellectual struggle for an indigenous, non-European, identity”, on
which Indian nationalists had embarked and which engendered a desire
to distil some allegedly unique “Indian essence” from India’s cultural
heritage. Accordingly the knowledge of all the ingredients and aspects
of Indian philosophy which reveal that the picture suggested by “neo-
Hindu” thinkers like Vivekananda, Radhakrishna and others as well as
the beliefs expressed in the article of H.H. Price are totally unrealistic
remained confined to a small circle of “specialists”. Since the present book
presents articles of members of this tiny group it is suited to counterbalance
a prevalent image.
It must be noticed, however, that not everything which is topical in
the ten main chapters of the anthology directly relates to “Indian Logic”,
even if one takes the term in a broad sense. In particular the first two
articles by Colebrooke and Müller comprise a wider scope by giving
a presentation of entire philosophical systems and dealing with topics
which could be labelled as ‘metaphysics’, ‘ontology’ and ‘philosophy of
nature’. In contradistinction to Colebrooke’s essay, which for the most
part describes the system of Nyāya according to a relatively late state of
development, Müller’s article presents an outline of the Vaiśes.ika-system
which is more closely related to earlier textual sources, in particular
the Praśastapādabhās. ya. If one wished to apply the word ‘logic’ to the
entirety of the subject matter of those articles one would have to take it
in the sense of ‘systematic philosophy’ – and ‘logical’ as equivalent to
‘systematical’. On the other hand, the topic which represents a core of
what is commonly called ‘Indian Logic’ and which is connected with
the term ‘syllogism’ is brought to the fore only at the end of the articles,
INDIAN LOGIC AND INDIAN SYLLOGISM 55
pp. 47–48 (in particular § VII) and pp. 68–74. In contradistinction to this
the so-called “Indian Syllogism” is a main theme in the subsequent essays
of Randle, Schayer and Ingalls. Apart from the fact that Schayer brings
“Modern Logic” (the tradition which has been founded by Frege and
Russell and is represented, among others, by the “Polish logicians” like
J. Łukasiewicz) into play, an important difference between Randle’s and
Schayer’s treatments of the topic is that, Schayer explicitly correlates two
members of the (five-membered variety of the) Indian syllogism with two
rules of inference, namely the upanaya with the “rule of substitution”,
i.e. the rule which allows the replacement of any expression of the form
‘(x)Fx’ (‘everything is F’) by an expression of the form ‘Fa’ (‘a is F’) – or
rather the derivation of a proposition of the latter form from a proposition
of the former form – and the nigamana with the “rule of separation”
which allows the derivation of ‘Ga’ from the combination of ‘Fa’ and ‘Fa
→ Ga’ (‘If a is F, then a is G’), and which represents merely a special
implementation of the old rule of modus ponendo ponens. This has the
consequence that the entire “syllogism” can be viewed as a combination
of two main parts, an assertion (pratijñā) and a proof of the assertion
– It seems that Schayer does not explicitly state this consequence,
but it is at least a most plausible upshot of his analysis. – The whole
argument-scheme represented by the “five-membered” syllogism could
thus be depicted as follows:
Thesis:
1. Assertion (pratijñā): Ga – “The Scafell Pike is a place on which fire
occurs”
Proof:
2. Reason (hetu): (Because of the fact that) Fa – “(Because of the fact
that) the Scafell Pike is a place which is endowed with smoke”
3. Example (dr.s..tānta) = Statement of vyāpti (pervasion): (x)(Fx → Gx)
– “Everything which is a place on which fire occurs is also a place
which is endowed with smoke”
4. Application (upanaya) = Statement of the paks.adharmatā (the fact that
the object of the assertion is encompassed by the pervasion): Fa → Ga
– “The Scafell Pike is such that, if it is a place on which fire occurs it
is also a place which is endowed with smoke”
5. Conclusion (nigamana) = Statement of the sādhya (the probandum):
Ga – “The Scafell Pike is a place on which fire occurs”
56 CLAUS OETKE
tradition. This must also be borne in mind with regard to the topics which
are brought to the fore in Staal’s essay ‘The Concept of Paks.a in Indian
Logic’. It is noteworthy that Ingall’s article contains a brief section dealing
with Navya-Nyāya.
A more extensive exposition of Navya-Nyāya is given in Bhat-
tacharyya’s article, and it seems that among the essays of the “Reader”
this treatise is apt to convey most insights about that area which is
widely still unexplored. A point which deserves to be mentioned here is
that, if Bhattacharyya is right, the Navya-Nyāya propounded the doctrine
that knowledge pertaining to some particular property which is de facto
pervaded by some other property – and presumably known to be pervaded
in this way – is not essential for inference: The decisive point for attribu-
tions of validity of inference is that the inferring subject knows that the
relevant substratum of inference (= the paks.a) possesses some property
which is pervaded by the property which constitutes the probandum. Since
the that-clause has to be taken in an opaque reading the requirement does
not entail that the inferring subject must know which individual property
fulfils the demand. This stands in contrast to the conception of the older
Nyāya, and according to the exposition of Bhattacharyya to the view of
the later Mı̄māṁsakas, which can be described by saying that an inferring
subject must know of some property that it occurs in some substratum (e.g.
some individual mountain) and – presumably – also know that this prop-
erty is pervaded by the property which constitutes the probandum. Another
information is noteworthy, even if it might not be immediately relevant for
the question of validity of inferences: If the author is right, the employment
of abstractors of higher order in Navya-Nyāya is rooted in the doctrine
that every expression relates to the range of its denotata under a certain
mode (prakāra), e.g. ‘cow’ under the mode of being a cow = cowness, in
contradistinction to, one might presume, a presentation of some group or
genus by an enumeration of all its members or by a non-synonymous term
with identical extension. Since it is assumed that the link with a particular
mode of denotation (= presentation?) pertains to all expressions (= all
nouns?)1 expressions like ‘cowness-ness’ which allegedly denote modes
of (first order) denotation denote this first order mode under a mode of
second order. Bhattacharyya’s reference to “Frege’s problem of infinite
regress of sense and denotation of higher orders” (p. 176) does not appear
far-fetched, but it might be a task for future research to explore whether
and to which extent the parallel reaches below the level of appearance.
1 Bhattacharya does not make clear whether the Navya-Nyāya theoreticians were
willing to consider this theorem as valid even with respect to words like ‘if’, ‘and’, ‘all’,
‘some’ etc.
58 CLAUS OETKE
world is conceived in such a way that there is something above pots etc.
Such a theory would be significant on account of the circumstance that it
would entail that identical spatio-temporal regions of the world are inhab-
ited by different types of entities. In this manner we are able to extract an
ontology which is much more interesting and which does not deserve to
be discarded by the dogma that only concrete particulars can be located
in space and time. This is nothing but a dogma, since e.g. genera can be
spatio-temporally located. – We say that the maple possesses a certain
area of occurrence or that the mammoth existed in the period of, say
between five million and a hundred thousand years ago etc. Why should
not entities denoted by terms like ‘pot-possessing-ness’ be akin to genera
at least in some respects? – Again, deeper investigations seem to be called
for in order to settle the question whether there are ways like the one
depicted above which would save the pertinent theories from the charge
of either absurdity or vacuousness. Unfortunately, Matilal does not only
not make explicit which particular authors and texts he has in mind but
leaves even obscure the fact that his account should be specifically related
to the Navya-Nyāya.2
2 Patrik Nyman has pointed out to me that T. Wada in his book Invariable Concomitance
in Navya-Nyāya, Delhi 1990, also ascribes the theorem that X-vattva is X to Navya-Nyāya
and that he regards this as an acceptance of a view allegedly held by Kātyāyana, the author
of the Vārttika on Pān.ini’s As..tādhyāyı̄. Wada – on p. 43, footnote 32 – refers to a section of
the Vārttika on As..tādhyāyı̄ 5.1.119 which reads siddham . tu yasya gun.asya bhāvād dravye
śabdaniveśas tadabhidhāne tvatalau. It remains mysterious, however, how this remark
could support a theorem of general identity or equivalence between any expression ‘A’
and a corresponding expression of the form ‘A-vattva’ or its denotata. First of all, such
a derivation appears to lead to the absurd consequence that – according to Kātyāyana
and the Navya-Nyāya (?) – all expressions denote qualities (gun.as). For if it holds true
that – in accordance with Kātyāyana’s statement – expressions of the form ‘X-tva’ and
‘X-tā’ denote qualities, an expression of the form ghat.avat-tva = ‘pot-possessing-ness’,
which results by substituting ghat.avat for ‘X’ must denote a quality = gun.a. So far, so
good. But because of the equivalence between X and X-vattva it must equally hold good
that ghat.a = ‘pot’ denotes a quality. The same holds good mutatis mutandis for all other
expressions. But if this were so, why did Kātyāyana suggest that (at least some) expressions
denote substances? The most natural interpretation of the phrase yasya gun.asya bhāvād
dravye śabdaniveśas tadabhidhāne tvatalau is that e.g in the case of the expression ‘pot’
the correlating expression resulting by the addition of an abstract-suffix to this expression
denotes precisely the quality on account of which pots are signified or characteriezed by
(an employment of) the expression ‘pot’. This appears to harmonize with Wada’s views
because he himself says that “suffixes ‘tva’ and ‘tal’ (i.e., tā) are employed to express a
quality of a ‘substance denoted by a word which is expressive of the substance’ as what
possesses the quality”. But from this fact it does not follow that that which a linguistic
expression denotes or signifies must be identical with the quality which the expression
attributes to its denotatum. It is true that if one refers to Aristotle by the term ‘the founder
of Western Logic’ one characterizes him in a certain manner, namely as somebody who
60 CLAUS OETKE
II
each single member which have been attested by different textual sources
into account.
That this is extremely important ensues from the circumstance that in
view of certain, especially older, textual sources, e.g. the Carakasaṁhitā,
it cannot be maintained that the third member of the five-membered syllo-
gism always expressed a universal proposition. It follows from this that
we must consider the possibility that the five-membered scheme did not
contain any general statement, and we can go a little further and claim
that at certain stages of the historical development it might not even
include any allusion to a universal proposition at all. This opens the way
for a more radical revision of the common picture and one can discard
even Schayer’s supposition that the syllogism “essentially” possesses a
deductive inference as an ingredient. Since the link between Indian syllo-
gism and deduction is thus completely severed, the problem arises of
finding any intuitive idea which is suited to make the phenomenon intel-
ligible. This is, however, not impossible and, although the possibility of
alternative accounts should not be categorically denied, it seems to me at
present that the best idea is provided by the metaphor of a loan. It appears
worthwhile to elaborate that notion a bit more because this is apt to stimu-
late a critical assessment of the results which the “Reader” offers regarding
its most essential topic.
If one recalls the facts that the five-membered syllogism is intimately
connected with inference and inferential proof and that the typical task of
inference is to extend the realm of that which is knowable – or which can
be treated in a similar manner as knowledge that has been obtained by other
means – beyond the narrow limits of what can be perceived – either genera-
lly or with respect to individual persons and individual times –, then it is
not difficult to grasp the point of the association of the ideas of inference
and “epistemic loan” and the connection between the latter idea and the
“Indian syllogism”: The syllogism accounts for a situation in which some
rational subject does not possess the means of verifying some proposition
in certain alternative ways, in particular by perception. But in order to
make practice more successful it appears desirable to overcome the restric-
tions imposed by this deficiency. Accordingly, the possibility is provided
that a cognising subject uses inference in a way which is comparable to
taking a “cognitive credit”. This means that in situations in which someone
is unable to verify some fact in a more direct manner, he can employ
inference as a cognitive device “by default” with the aim of provisionally
accepting the truth of the pertinent proposition and acting in accordance
with the supposition of its truth as a “working hypothesis”, although he
accepts the risk that he might be unable to “pay off his loan” in the long run
62 CLAUS OETKE
the utterance of (2) raises the question: “But why is this – that there is
smoke on the hill which is the subject of our discussion – an argument for
your thesis?”. Accordingly the subsequent response, represented by (3)–
(5) constitutes an elucidation on the part of the proponent why he considers
(his utterance of) (2) as a justification for his assertion. However, even
if we grant this, it remains unintelligible why the proponent should have
started his justification with the utterance of (2) at all. If the justification
of the thesis that there is fire on a mountain relies on the facts (a) that
wherever there is smoke there is fire and (b) that on the pertinent moun-
tain there is smoke, the utterance of (2) – or of (4) – is pretty redundant.
Finally one must consider the fact that it is on the one hand difficult to
see why rhetorical aspects should possess a determining influence on the
elaboration of a canonical form of presentation of an argument – after all,
does not rhetorical efficiency rely on novelty and originality rather than on
conformity to stereotypes? – and on the other hand it is hard to understand
why someone should have believed that a rather clumsy and repetitious
way of presentation should possess extraordinary rhetorical merits.
Let us see how the situation is if we drop the supposition that any
universal proposition is presented in order to “derive” a conclusion and
assume that the relevant situation could be characterized by the metaphor
of someone who applies for the grant of an “epistemic loan”. It is true
that one could depict the point by imagining a dialogue between two
interlocutors, but the conversation exhibits a completely different form:
1) Thesis: “Fp” (“The Scafell Pike is a place on which fire occurs”)
Sceptic: “Why should one believe that”/“Why should one grant that
the thesis is true?”
2) Answer: “Because Gp” (“Because the Scafell Pike is a place which is
endowed with smoke”)
Sceptic: “Sure, there is some subjective support for the thesis. But is
this also an objective support? Why should the mere circumstance that
(it is known that) ‘G’ is true of p support the proposition that ‘F’ is
true of p?”
3) Answer: “This is not merely a subjective support, because the
examples E1 , E2 . . . En show that ‘G’ and ‘F’ are connected in other
instances so that one can definitely say that if one would argue in an
analogous manner with respect to E1 etc. as it is argued here with
respect to p one would argue for a true thesis.”
Sceptic: “Why are those considerations regarding E1 etc. relevant to
the present issue which concerns p?”
64 CLAUS OETKE
4) Answer: “E1 etc. and p are alike in all respects which are relevant to
any argument which pertains to ‘F’ and ‘G’ and accordingly to the
present argument.
5) Therefore it is reasonable to grant the truth of the thesis.”
The point of the metaphor of an epistemic loan becomes more perspicuous
in view of the following analogy between a person who asks for a loan and
a creditor:
1*) Businessman: “I ask for a loan of X ”
Bank-manager: “I do not see any reason to grant you the loan.”
2*) Businessman: “I use the money in order to invest it for a venture in
telecommunications.”
Bank-manager: “The risk is too high.”
3*) Businessman: “A, B, C . . . have also invested in telecommunications
and made so much profit that they could have paid off a comparable
loan.”
Bank-manager: “But these are different cases.”
4*) Businessman: “Yes, but they are alike in the relevant respects.
5*) Therefore my request for a loan of X is justified.”
It is easy to see that under the present hypothesis none of the steps of the
syllogism is superfluous or redundant. To be sure, the first and the last
step allude to one and the same fact, but this does not entail a redunda-
ncy because the function of referring to an identical fact is different in
both cases: The first step expresses that a certain claim is being made,
whereas the last step conveys that that which is claimed here is based on
objective considerations and that anybody who makes exactly the same
(kind of) claim is in principle able to present a justification. The charge
of redundancy on account of the identity of steps 2) and 4) simply evapo-
rates because 4), according to the presented analysis, expresses something
entirely different and none of the remaining members nor any conjunction
of them entails that which is expressed here. However, it is precisely this
ingredient which represents the crucial point of the argument: It raises the
problem regarding the conditions under which an analogy can be trans-
ferred to a pertinent instance. It is extremely difficult to specify such
conditions in a general manner. Even in this regard the analogy between
the epistemic and the financial case holds good. Part of the difficulty
results from the fact that the principles which could underlie decisions
as to whether or not one might treat the pertinent and the sample cases
as alike in all relevant respects are sensitive to contextual circumstances:
In certain situations, e.g. if the (financial or epistemic) loan is small, one
INDIAN LOGIC AND INDIAN SYLLOGISM 65
but tend to be used in such a way that individual occurrences comply with
the use of other occurrences of the same type. Without such extension-
justifying principles inferences of this kind are completely invalid, and
therefore analogical reasoning can be “unscientific”. But this is precisely
the point on account of which I myself have, sometimes quite severely,
criticised some colleagues because they employ analogical cases, e.g.
textual passages of other sources which look similar, without reflecting
on the existence or non-existence of a corresponding extension-warranting
principle. The circumstance that in the realm of the attribution of word-
meanings such principles (normally) exist, does not prove that they exist
equally in other areas, and to derive the latter from the former is extremely
naïve. But all this highlights the tremendous importance of the step of
reasoning represented by the fourth member of the syllogism. It is diffi-
cult to decide whether or not the syllogism embodies an awareness of
the mentioned methodological requirement. The textual sources present
various formulations of the upanaya, some of which do and some of which
do not harmonize with such an assumption. Therefore the question must
be left open at present. – The circumstance that various formulations of
some members of the Indian syllogism have been handed down does by
no means decide the issue. The supposition that the varieties reveal lack of
clarity in the original conception is at best one possible hypothesis. Isn’t it
equally conceivable that the five-membered syllogism was the product of
one particular person endowed with extraordinary subtlety of thought and
that the deviant formulations result from the fact that his creation had never
been appropriately understood in the subsequent tradition? Again one can
only answer: We do not know.
What is the principle by which the above presented account deviates
from those which interpreted the syllogism as the embodiment of a process
of deductive reasoning? Once more, I would like to employ a metaphor: If
the rival account were correct, we should compare the corresponding syllo-
gistic reasoning with investment of capital. A correlating theory would
be an account as to how one can use capital on the supposition that one
possesses it. But, if my proposed interpretation is correct, this is not the
original aim of the theoretical account of the Indian syllogism. It pertains
rather to the situation in which somebody does not possess the money
which he needs or desires and decides to overcome this situation by
taking a loan. The theory of the Indian Syllogism is – originally – neither
concerned with the question of certain knowledge or knowledge of neces-
sary truths nor with the question of making deductions from premises.
If concern with questions of the latter type is regarded as an essential
68 CLAUS OETKE
ingredient of the concept of ‘logic’, then we should say that the theory
of the Indian Syllogism is not a logical theory. But it would be a grave
mistake to conclude from this alone that it must be inferior. It amounts by
no means to a denial of the great importance of logic, if one asserts that
the question as to whether ‘logic’ or ‘methodology’, as it is embodied by
the theory of syllogism, should be ranked higher is ideological. After all,
should one say that the problem of how one invests one’s capital is more
important than anything else in the field of economy? This can hardly be
true, because issues pertaining to ways and methods of obtaining capital
are in certain respects even more fundamental.
All this should not be misunderstood as a denial of the possibility to
criticize the account embodied by the Indian Syllogism. The theory is far
from perfect even if one assesses it on the background of its own concerns.
The parallel with the bank-loan helps one to discern that the account
represented by the five-membered syllogism exhibits features of restriction
and conservatism which are problematical. May be it is an acceptable prac-
tice for a bank to give a loan only in cases in which analogous investments
have been demonstrably successful. But such a policy would be disastrous
for an economy. Presumably the same must be said concerning methodo-
logy of knowledge. A doctrine which demands that every claim must be
supported by strictly analogous examples cannot even cope with inferences
which rely on quantitative theorems, e.g. equations between pressure and
temperature. But the charge that the Indian Syllogism was inferior before it
introduced universal propositions is completely erroneous and the picture
of a slow progress from pre-logical towards logical thinking in classical
India is fundamentally mistaken. Instead of interpreting the introduction
of universal propositions together with the adoption of the notions of
“concomitance” and vyāpti as a break-through on the way to “genuine
logic”, one comes probably much nearer to the truth if one understands the
corresponding developments as a change of policy. This must not preclude
that the changes resulted from insights.
The problem of the understanding of the Indian Syllogism might
be regarded as exemplary for the problems of understanding alien
cultures. Although one is immediately able to recognize affinities between
“Western” and “Indian logic” due to the fact that both are related to
reasoning, inference and argumentation in a similar manner as credit
and investment have something in common, the identification of the
appropriate sub-categories is by no means easy. Problems of this sort
become extremely thorny if in the milieu of the “cultural interpreter” a
corresponding sub-department does not exist. The only feasible way of
achieving understanding in such situations might consist in the “creative
INDIAN LOGIC AND INDIAN SYLLOGISM 69