Nyayavatara and Buddhist Logical Works

Author(s): Alex Wayman
Publisher: Z_Nirgrantha_1_022701.pdf and Nirgrantha_2_022702.pdf and Nirgrantha_3_022703.pdf
Catalog link: https://jainqq.org/explore/269011/1

Page #1
THE NYAYAVATARA AND BUDDHIST LOGICAL WORKS BY DIGNAGA AND RATNAKARASANTI Alex Wayman The Nyayavatara (N. A. hereafter) is a Jaina work whose authorship had been attributed to Siddhasena Divakara. M. A. Dhaky has written an essay claiming with good reasons that Siddhasena Divakara is not the author of N. A., and that the author is rather a different Siddhasena known as Siddharsi'. The present essay is not concerned with authorship of N. A. It is instead concerned with the position that this Jaina work accepts certain teachings, opposing certain others, of the Buddhist logician Dignaga. And is also concerned with the N. A.'s theory on antarvyapti, and with whether this Jaina position agrees with what is set forth in Ratnakarasanti's antarvyapti treatise? Therefore, the present essay is necessarily concerned with the date of the N. A. Dhaky has decided that the N. A. is subsequent to circa A. D. 550, and hence is subsequent to Dignaga, born circa A. D. 4803 - hence around the year of death of Vasubandhu (brother, probably half-brother of Asanga), circa A. D. 400-480. Prof. Dhaky kindly supplied me with a copy of the N. A. text and translation by S. C. Vidyabhusana. Therefore, may I speak briefly about this translation and annotation. Since this translation by Vidyabhusana rendered the verses throughout with Western logical terminology (e.g. 'middle term'), as a translation it serves mainly to call attention to the various verses as having such and such content, while not necessarily exposing the intentions of the individual verses. Indeed, it is an unproven contention that Indian logic agrees with the conceptual divisions of Western logic, and there are some who do not concede thiss; while in my own translations of such logical literature I do not follow this Western terminology. But then the translator Vidyabhusana added extensive comments in English that do not refer to any commentaries on the N. A. These ample comments agree with his erudite work, A History of Indian logic -a work which reveals his excellent control over these materials. However, as his comments on the N. A. in particular only barely or vaguely contact its verses, while demonstrating that he knows a lot, I decided it would be justified to completely disregard these comments on N.A. A. Relation with Dignaga It has been known since the beginning of modern Dignaga studies that his Inference (anumana) is of two kinds--for oneself (svartha) and for others (parartha)? In fact, my own researches in Buddhist logic indicate a chief difference to be over the term sadhya. In 'inference for oneself' the sadhya is to be determined by the evidence (linga); and while it is in this condition I render the term as the 'evidential'. Once one determines the sadhya as the logical conclusion of the evidence, it can then be called the thesis and means this during the 'inference

Page #2
Alex Wayman Nirgrantha for others'. That is why in the Dignaga-Dharmakirti system inference for oneself' must precede 'inference for others'. Dignaga was probably not the first one to make such a division and in that order, but he should be given credit for making this division and with the necessary logical qualifications. 2 Now the Nyayavatara, k. 10, indicates its way of discriminating the two kinds of inference : svaniscayavad-anyesam niscayotpadanam budhaih | parartham manam adhyatam vakyam tad-upacaratah || Like the conviction in oneself, the arousal of conviction in others, is called by the wise the standard for others, and that assertion, in the sense of a metaphorical transfer. We notice the agreement with Dignaga that 'inference for oneself' should precede inference for others'. However, the verse's metaphorical attribution (upacara) to 'inference for others' seems to be a failed attempt to appear 'sparkling' and so at least the composition of this verse agrees with Dhaky's somewhat negative evaluation of this N. A. as a work of Jaina literature. We pass the N. A. k. 13: Sadhyavinabhuvo hetor-vaco yat pratipadakam | parartham-anumanam tat paksadivacanatmakam || A demonstrative statement of a reason that applies to the thesis and which mentions first the locus (paksa)-is Inference for others. Here the N. A. does better. This minimal statement-by avoiding explicit mention of an external example-agrees with inner pervasion'. And it specifically mentions the paksa, going on in N. A. 14 and 15 to insist upon it as essential. So far the N. A. does not disagree with Dignaga. Let us go to N. A. k. 28, which, as a literal statement, disagrees with Dignaga (here, my own translation) : Pramanasya phalam saksad-ajnanavinivartanam | kevalasya sukhopekse sesasyadanahanadhih il The result of pramana in the direct vision sense is the warding off of ignorance. Bliss and equanimity are the (result of the) kevala (absolute kind). The cognition of accepting and rejecting is the (result of the) remainder (conventional kind). In Dhaky's article this verse is held to refute k. 1.23 of Dignaga's Pramanasamuccaya. I take this number k. 1.23 to be a printer's mistake for k. 1.2-38. If we take the

Page #3
Vol. 11-1996 Nyayavatara.... verse nos. to exclude the mangala verse, then the nos. 2-3 can be rendered as follows, as in my own rendition of the initial verse section by Dignaga from the Tibetan version": Nor are authorised (pramana) those like memory, and so forth, that repeatedly cognize and reduce to an endless series (anavastha-prasanga). Direct perception is free from constructive thought (kalpana) which adds a name, class, etc. (1,2). Its name (namely 'pratyaksa' = direct[ed] perception') is called by a sense organ (aksa) because of the unshared causes (namely prati = 'each one'). A factual base (dharmin) for multiple natures can in no case be understood from a (single) sense organ. (1.3). Assuming that I have above the right verses from Dignaga's PS, I have reservations about whether N. A. 28 is really refuting Dignaga. On the surface, it seems so. When this verse states, "The result of pramana in the direct vision sense [= pratyaksa) is the warding off of ignorance", this is an other worldly aim that differs from the Dignaga system which argues in a mundane manner. Indeed, the N. A. 28 employs the term pramana in an instrumental sense, as does certain Brahmanical systems. Dignaga employs the term pramana in a resultative sense-the result of prameya (which I render 'sanction'). Now, when scholars or pandits employ the basic words in different, even opposite meanings, the real argument is over the use of those words. The N. A. verse 28 and Dignaga's verses I, 2-3 are really talking at 'cross-purposes as though to different audiences. Indeed, each of the systems of Indian philosophy has certain important terms which are employed in senses different from how those words are used in other systems or in general Indian literature. The Jainas are no exception to this practice of employing certain terms in special meanings. One could fault Dignaga if he deliberately used the term pramana in a resultative way in defiance of the standard usage; but, as I myself pointed out, the medieval lexicons when defining the head word Pramana, have a number of definitions of instrumental value, thus agreeing with the instrumental use of the term, as well as definitions of resultative value, thus agreeing with the resultative use of pramanal. For example, one of the instrumental definitions is sastra, a technical treatise-an instrument of information; and one of the resultative definitions is maryada, a boundary, or limitation, evidently related to pramana's root to measure'; and a boundary is clearly the result of measuring. Hence, Dignaga does not have to argue for his use of the term pramana, any more than his rivals have to argue for their use of it. But clearly, when authors compose verses in the field of logic, employing the term pramana in an instrumental fashion, the verse comes out looking quite different than when

Page #4
Alex Wayman Nirgrantha Dignaga writes a verse with contrasting meaning for pramana. I could say for N. A. 28 that this author's way of writing introduces elements of a sermon nature, with doctrinal declarations, whereas Dignaga's verses are closely reasoned. Now I shall consider the implications of N. A. 21, referring to 'fallacies of the locus' (paksa-abhasa) and N. A. 22, referring to 'fallacies of the reason' (hetuabhasa). First, N. A. 21 : Pratipadyasya yah siddhah paksabhaso, sti (var, aksa) lingatah lokasvavacanabhyam ca badhito, 'nekadha matah || Vidyabhusana's translation was a valiant effort to make sense out of this verse". The verse begins, "What is proved (siddha) of 'that to be explained' (pratipadyasa = the sadhya, as Vidyabhusana correctly interprets)". Vidyabhusana interpreted the term siddha as affirmed', but in logic siddha means 'proved'; sadhana is the 'means of proof'; and sadhya is that to be proved. Since the verse begins with granting something has been proved of the sadhya-without qualifying what was proved--it is inconsequential to go on and talk about 'fallacies of the locus.' Now we turn to N. A. 22, for 'fallacies of the reasons : Anyatha'nupapannatvam hetor-laksanam-iritam tad-apratit'samdehaiviparyasais-tad-abhata (read : abhasah] || Besides, a nonsuitable characteristic of the reason has been announced, which is a fallacy by noncognitive dawning, doubt, or contrairity. The next verse, N. A. 23, attempts to clarify these types. The 'unproved' (asiddha) is the 'noncognitive dawning' (apparently not in the scope of normal cognition); the 'contradictory' (viruddha) is the 'contrairity'; the 'uncertain' (anaikantika) kind of reason being the one labelled 'doubt'. There is a sort of clumsiness in calling these 'fallacies of the reason'. Usually, it is the 'locus' (the paksa) which is fallacious (if indeed it be so). For example, if the thesis be "There is a fire on the lake tonight" should the reason be "Because we see smoke there," the reason is fallacious but not in the statement of the reason itself. If the thesis has been "There is a fire on the mountain tonight," that same reason would have been appropriate. The fallacy was actually in the paksa (the locus), since if it is indeed a lake, we do not see smoke, rather see mist. Thus, a wrong attribution to the paksa is involved. Granted though that in ordinary parlance one may speak of a fallacious reason. However, the N. A. itself devoted several verses to insisting upon the paksa (locus). In short, if the N. A. had presented the fallacies in terms of the paksa rather than in terms of the reason, there would have been consistency with the prior verses on the paksa, and then this text should be praised for its cogency and consistency. Considering the N. A.'s standard in verses 21 and 22, we begin to wonder about the qualifications of the author to write on the topic of logic.

Page #5
Vol. II-1996 Nyayavatara.... This is not to deny that there is some ambiguity in the term paksa itself, as has been well discussed by Staal'. B. The Antarvyapti theory The N. A. has its basic statement about "inner pervasion" in its verse 20 : Antarvyaptyaiva sadhyasya siddher-bahirudahrtih vyartha syat tad-asadbhave 'py-evam nyayavido viduh || Logicians have noticed that an external example would be useless (vyartha), since even in its absence there is proof of a thesis (sadhya) just by internal pervasion (antarvyapti). The author seems to have adopted an awkward expression 'useless' for what would be more accurate--a term signifying unnecessary', since external examples are often useful for teaching purposes. Indeed, this author himself presented an external example in his verse 19, which is the best composed of any I have noticed in this text: Sadhye nivartamane tu sadhanasyapy-asambhavah khyapyate yatra drstante vaidharmyeneti sa smotah || The clue to the meaning of this verse is his use of the locative absolutes, which lend a conditional "When such and such is the case." And the explanation is suggested in Stcherbatsky's well-known work on logic. So the verse is now rendered : For which place it is mentioned that given a thesis which is being denied, also the impossibility of a means of proof (= a reason), it is held that the example is in the sense of vaidharmya (discordance). In illustration, it might be said : "Suppose there were no fire on the mountain, and also suppose there is no reason of seeing smoke there, like the water of a lake (the 'discordant example')." And then the speaker continues : "But there is a fire on the mountain, for now we do see some smoke there !" Also, this author had presented the 'concordant example' in his N. A. verse 18 : Sadhyasadhanayor vyaptir-yatra nisciyatetaram sadharmyena sa drstantah sambandhasmaranan matah 1 Where the pervasion of the thesis and the proving method is definitely established, the example is claimed as with concordance due to remembering the connection. Here the author shows that the external example is unnecessary for establishing the required pervasion; but since people at large should remember the connection,

Page #6
Alex Wayman a concordant example is used for teaching purposes. So, even though the author had labelled as 'useless' (vyartha) an external example, he himself found it useful to mention the 'concordant example' (sadharmya-drstanta) (in N. A. 18) and the 'discordant example' (vaidharmya-drstanta) (in N. A. 19). Before comparing Ratnakarasanti's position on Antarvyapti with that of the N. A. it is well to refer to his approximate dating. Because he is well-established in the field of Buddhist Tantra, I included him in my Sarnath, Varanasi, seminar contribution on "An Historical Review of Buddhist Tantras"." Here I explained why his dating is tied in with that of the great Tibetan translator Rin-chen Bzang-po, born A. D. 958 and that he would be a junior by about 20 years, hence born about A. D. 978. Hence he lived several decades after the author of N. A. It follows that Santi-pa (as the Tibetans call him), in defending the theory of Antarvyapti, does not directly depend on the Jaina work N. A., even though this latter work may contain the first use of the term antarvyapti. Santi-pa's treatise is subtle and difficult. When I translated it, using the text edited by Haraprasad Shastri1s compared with the Tibetan translation, I found editing mistakes both in the Sanskrit edition as well as in the Tibetan rendition. 6 I believe that Shastri's introductory comments are quite appropriate and insightful. To put it in my own words, the usual examples given for the theory of pervasion (vyapti), or invariable concomitance, are illustrated by the 'syllogism' of claiming that there is fire on that mountain, giving as reason that we notice smoke there, as in the case of the kitchen. Hence, the generality of inferences is based on prior experiences, and there is a constant appeal to previous experiences in a given culture, which justifies the example that has been popularly experienced. Hence, one could not argue against the claim that the pervasion in such cases requires inclusion of the example. Nirgrantha But then Santi-pa claims that in some cases of a thesis and reason, which he terms sattvahetu (reason for a real or existent thing), it is not possible for an example to be part of the pervasion. In illustration, if the Buddhist argues, "What exists is momentary," and the various sects of India have their own theses (on Atman, Brahman, etc.) of similar declaration-it does not help to say "like a pot", since being momentary does not add a quality to the sattva as does the color 'red' to a pot, nor is a pot a paragon for what exists. Therefore, Buddhism has doctrinal reasons for the assertion "What exists is momentary," which presumably appeal to the followers of the lineage which so asserts. The various philosophical and religious schools of India all have their own declarations, giving reasons that appeal to the followers of such schools, and usually not providing any example that is popular and accepted by people at large. Now, presuming that various of those "reasons for an existent thing" are reasonable and acceptable, it must be that the pervasion was just between the thesis and the reason and does not extend to an example. Hence, even if an example is provided

Page #7
Vol. II-1996 Nyayavatara.... for such a doctrinal assertion, this example could be explained as intended for teaching purposes, and was outside the inferential process that 'proved the thesis. In short, this antarvyapti theory does not displace external pervasion, but rather asserts a class of theses for which external pervasion does not apply. The N. A. discussed above may well have had a similar theory to justify the antarvyapti position, but we should grant that this text extends its antarvyapti coverage to all cases of syllogisms, not just to those where it is difficult to apply a popular example, except one only for teaching purposes. In the following centuries, with much arguing between the Buddhist logicians and other groups such as the followers of Nyaya and certain Jaina authors, these theories became much more subtle, with many more qualifications than we find in the N. A. That we find this theory defended by Ratnakarasanti-a well-known Buddhist pandit--shows that the rival schools of India are not always disagreeing. They have more agreements than is usually admitted in the survey books on Indian philosophy. REFERENCES : 1. M. A. Dhaky, "The Date and Authorship of Nyayavatara," Nirgrantha 1, Ahmedabad 1995. 2. Cf. Alex Wayman, "Ratnakarasanti's Antarvyaptisamarthana," Journal of the Asiatic Society (Calcutta), XXVII, No.2, pp. 31-44; Note: the printers of this issue omitted a page between pages 32 and 33; and an insert of the omitted material is available from the author (A. Wayman). 3. Dhaky, "The date..." I also use this date, accepting E. Frauwallner, "Landmarks in the History of Indian Logic" (WZKSO, Band V, 1961). 4. Satis Chandra Vidyabhusana, "Nyayavatara by Siddha Sena Divakara," included by Jaina Sahirya Vikasa Mandala in Siddeasena's Nyayavatara and Other works, ed. A.N. Upadhye, Bombay 1971. 5. Cf. Hidenori Kitagawa, "A Note on the Methodology in the Study of Indian Logic," Journal of Indian and Buddhist Studies, Vol. VIII, No.1, 1960, pp. (19) to (29), beginning, "To interpret Indian logic using the terminology of Aristotelean logic, according to my opinion, is not to represent Indian logic as it is, but merely to review Aristotelean logic as applied to Indian logic." Kitagawa put out a large book on Dignaga in Japanese, publ. 1965, including much Tibetan from the Pramanasamuccaya and the English translation of Dharmakirti's Sastanantarasiddhi as "A Refutation of Solipsism." Then, the essay by J.F. Staal (unaware of Kitagawa's essay), "The Concept of Paksa in Indian Logic, Journal of Indian Philosophy 2 (1973), pp. 156-166, beginning, "In studying a civilization different from our own we are prone to impose the conceptual framework and prejudices of our own tradition...S. C. Vidyabhusana, the first historian of Indian logic, looked at his subject through eyes so colored by what he regarded as Aristotelian logic, that he talked of the Indian syllogism' and saw in it traces of the influence of Aristotle-a historical claim no serious student of Indian logic would nowadays wish to make his

Page #8
Alex Wayman Nirgrantha own." 6. M. M. Satis Chandra Vidyabhusana, A History of Indian Logic, Motilal Banarsidass, (Delhi 1971 edn). 7. For example, H. N. Randle, Fragments from Dinnaga (Motilal Banarsidass, 1981 reprint) Index, p. 85, has references to the two kinds of inference. 8. Referring to Masaaki Hattori, Dignaga, On Perception, tr. and annotated (Cambridge, Mass., Harvard University Press, 1968), the verse no. 23 would take the reader out of Dignaga's own system to a section where he attempts to refute one or other Brahmanica! philosophical school in this matter of pratyaksa; whereas the topic of the N. A. verse under discussion goes with Dignaga's own position. 9. From my MS A Millennium of Buddhist Logic, Voi. I, section "Introduction to the Dignaga-Dharmakirti system by the `elevens' ". 10. Several years ago, at an annual meeting of the American Oriental Society, I delivered a paper, "A lexical approach to the philosophical notion of Pramana." 11. Vidyabhusana, n. 4, above, p.14 : "If that of which the major term or predicate (sadhya) is affirmed is opposed by evidence (linga), the public understanding, one's own statement, etc., we have that which is known as the fallacy of the minor term or thesis (paksabhasa), of which there are many varieties." 12. As referred to in n. 5, above. 13. Th. Sicherbatsky, Buddhist Logic, Vol. Two (New York : Dover Publications, 1962), p. 142. 14. This seminar volume is being published by the Tibetan Institute, Sarnath, probably in 1996. 15. Mahamahopadhyaya Haraprasad Shastri, Six Buddhist Nyaya Tracts (Calcutta : Asiatic Society, 1910), pp. 103-114.