Note On Development Of Vaisesika Theory Of Anumana

Author(s): Masaaki Hattori
Publisher: Masaaki Hattori
Catalog link: https://jainqq.org/explore/269227/1

Page #1
PRASASTAPADA AND DIGNAGA? A NOTE ON THE DEVELOPMENT OF THE VAISESIKA THEORY OF ANUMANA By Masaaki Hattori, Kyoto In the third chapter of the Pramanasamuccaya, Dignaga quotes from a certain Vaisesika treatise a definition of the logical reason (hetu), which runs: tadvaddharmasya (abhidhanam) hetuh = A reason is (a state . In a recent issue of this journal, L. SCHMITHAUSEN published an article "Zur Lehre von der vorstellungsfreien Wahrnehmung bei Prasastapada" (WZKSA 14, 1970, pp. 125--129) in which he criticised my paper "Two types of non-qualificative perception" (Beitrage zur Geistesgeschichte Indiens, Festschrift fur Erich Frauwallner. Wien 1968, pp. 161-169) in some points. His arguments based on careful analysis of the relevant materials were very convincing, and obliged me to make emendations to my interpretation of Prasastapada's theory of perception. This paper is not intended as a rejoinder. I only discuss Dignaga's influence on Prasastapada's theory of anumana, which I could not treat in my former article. In this paper I use the following abbreviations: NS: Nyayasutra of Gautama, ed. by GANGANATHA JHA, Poona 1939 (Poona Oriental Series, No. 58) NV: Nyayavarttika of Uddyotakara, ed. by VINDHYESVARIPRASADA DVIVEDIN, Benares 1916 (Kashi Skt. Ser.). PDhS: Padarthadharmasamgraha of Prasastapada (= Prasastapada bhasya), ed. (1) together with Nyayakandali of Sridhara, by VINDHYESVARIPRASADA DVIVEDIN, Benares 1895 (Vizianagram Skt. Ser.), (2) together with Sukti, Setu and Vyomavati, by GOPINATH KAVIRAJ, Benares 1930 (Chowkhamba Skt. Ser.). The page and line numbers in ed. (2) are given in parentheses. Pramanasamuccaya and Vstti of Dignaga, Tibetan Version, The Tibetan Tripitaka, Peking ed., Reprint, Tokyo-Kyoto 1957. K = Kanakavarman's transl. (No. 5702), V = Vasudhararaksita's transl. (No. 5701). The reconstituted Sanskrit texts in this paper are taken from MUNI JAMBUVIJAYA's edition of the Vaisesikasutra, Appendix 7. VS: Vaisesikasutra of Kanada with the Commentary of Candrananda, ed. by MUNI JAMBUVIJAYA, Baroda 1961 (Gaekwad's Oriental Ser., No. 136). PS:

Page #2
170 MASAAKI HATTORI ment) of a property of tadvat2. From Dignaga's explanation, it is known that this definition was preceded by a definition of the proposition (pratijna): sadhyabhidhanam pratijna A proposition is a statement of that which is to be proved3. In reference to this latter, the rather ambiguous term tadvat is understood as meaning sadhyavat, that which possesses the sadhya (-dharma), i. e., the subject of a proposition (dharmin). Also in PS, chap. IV, a Vaisesika definition of the example (drstanta) is quoted, most probably from the same treatise, as: ubhayaprasiddho drstantah An example is that in which both (the sadhyadharma and the reason) are well known (to be present or absent) 5. Neither the title of the Vaisesika work from which Dignaga quoted, nor the name of the author of that work, is given by Dignaga or by his commentator Jinendrabuddhi. However, we know from these quotations that the Vaisesikas built up their theory of the syllogism before Dignaga's time, giving a definition to each member of the syllogism. Our sources for the study of the early Vaisesika system are very scanty, and we can hardly ascertain in which period the Vaisesikas came to form this theory. It is not traceable in the Vaisesikasutra. The definition of the proposition is virtually identical with that found in the Nyayasutra. The definitions of the reason and the example are simpler than those in the Nyayasutra. However, this does not necessarily mean that the Vaisesika theories as referred to by Dignaga are older than the Nyayasutra. The existence of a series of Vaisesika works attested by Mallavadin and his commentator Simhasuri, who is supposed to have lived not much later than Dignaga, seems to show that the school tradition of the Vaisesikas was kept without interruption up to the time of Dignaga". If the more advanced theory of the syllogism had been formulated, prior to Dignaga, by the Vaisesikas, Dignaga ought to have referred to it. = 2 PS, K 140b.2: bye brag pa rnams kyan "de ldan chos ni gtan tshigs so" brjod pa ses bya ba cjug go. V 56b.5: bye brag pa rnams kyi "de dan ldan paci chos gtan tshigs so" ses brjod par byed do. 3 Ibid., K 140b.3: cdir yan gal te deci sgra "bsgrub bya brjod pa" dan cbrel par byed na... V 56b.6: cdir yan gal te de ses paci sgras "bsgrub bya brjod pa" dan mnon par cbrel par gyur te... 4 Dignaga gives various possible interpretations to the word tadvat in his examination of the Vaisesika definition of hetu, Cf. PS, K 140b.3 ff., V 56b.6ff. 5 Ibid., K 155b.7: bye brag pa rnams kyi "gni ga rab tu grub pa ni dpeco" ses bya ba... V 70a.3: bye brag pa rnams ni "gnis ka rab tu grub pa dpeco" ses zer ro. Cf. also K 152b.3, V 67a.5. Cf. NS, 1.1.33-37. 7 A. THAKUR, Introduction to JAMBUVIJAYA's edition of the Vaisesikasutra, pp. 10-14.

Page #3
Prasastapada and Dignaga 171 Thus, it may be reasonable to consider that the theories which Dignaga takes up for criticism were maintained by the contemporary Vaisesikas. When we compare the Vaisesika definition of the proposition, etc. as referred to by Dignaga with the same as explained in the Padarthadharmasamgraha of Prasastapada, we cannot but acknowledge a distinct gap between the two. Prasastapada is found to have introduced new ideas to the Vaisesika system with an entirely new phraseology, but, while doing so, he does not explain at all the reason for introducing new concepts. Take for example his definition of the proposition: anumeyoddeso 'virodhi pratijna = A proposition is a statement of the object of inference which does not involve contradiction. The new word anumeya is found employed in place of sadhya, and the meaning of anumeyoddesah is deliberately explained as follows: pratipipadayisitadharmavisistasya dharmino 'padesavisayam apadayitum uddesamatram pratijna = A proposition is a mere statement, with a view to showing that to which the reason (apadesa) applies, of a subject as qualified by the property, which it is desired to prove. The word avirodhin is also new to the Vaisesikas, and it is intended to mean that the proposition should be free from the five kinds of fallacy, that is to say, it should not be contradicted by perception (pratyaksa), inference (anumana), scripture (agama, or accepted theory: abhyupagata), one's own treatise (svasastra), and one's own words (svavacana). Prasastapada is known to have composed, besides the Padarthadharmasamgraha, an extensive subcommentary (tika) on a certain Bhasya, which had been written on the Vaisesikasutra accompanied by a Vakya'. It may be assumed that he fully explained in that lost work the 'new concepts that he introduced to the Vaisesika system. However, even on this assumption there remains a problem to be cleared up. The views propounded by Prasastapada are not of such kind as could be derived from the pre-Dignaga Vaisesika doctrines with a slight modification. There must have been an incentive which induced Prasastapada to make some essential changes in the Vaisesika theories. On this point Dignaga's criticism of the Vaisesika and the Naiyayika theories of inference and syllogism seems to throw some fresh light. Dignaga takes up for criticism the Naiyayika definition of the proposition: sadhyanirdesah pratijna (NS, 1.1.33), which is substantially the same as the older (i. e., pre-Dignaga) Vaisesika definition. According to him the word sadhya simply means that which is contrary to siddha, 8 PDhs, p. 233ff. (p. 599 ff.). 9 THAKUR, op. cit., p. 14; G. CHEMPARATHY, "Prasastapada and his other names," IIJ, XII (1970), p. 252, n. 45.

Page #4
172 MASAAKI HATTORI i. e., that which has not yet been proved (asiddha). Thus, a reason and an example, which have not yet been proved to be true, should also be recognized as sadhya. Such being the case, if the Naiyayika definition of proposition were to be accepted, there would follow the absurdity that the statement of the reason and the example of this sort would also be regarded as a proposition. For example, when one states, in order to prove that sound is evanescent, the reason because it is visible (caksusatvat)', the statement of this reason would be considered as a proposition, inasmuch as the visibleness of sound has not yet been proved (asiddha = sadhya). When one states the proposition and the reason as 'sound is eternal', 'because it is not tangible, (asparsatvat)', and then gives the example 'like buddhi', the statement of this example is also to be admitted as a proposition, since whether buddhi is eternal or evanescent has not yet been proved 10. Thus the defining of proposition as sadhyanirdesah was found to be defective by Dignaga. After thus refuting the Naiyayika definition of the proposition, Dignaga states that the same criticism applies to the definition given in the Vadavidhi: sadhyabhidhanam pratijnall. This is exactly identical with the one formulated by the pre-Dignaga Vaisesikas. The Naiyayikas and the Vaisesikas had to answer this criticism either by defending their definition through finding some justification for it or by making an alteration to their definition. Uddyotakara chose the first means: he emphatically repeated that the word sadhya was explained by Vatsyayana as signifying prajnapaniyadharmavisistadharmin (a subject as qualified by the property, which is to be made known)12. On the other hand, Prasastapada decided for the second means: he adopted a new expression anumeya in place of sadhya. " The term anumeya is used by Dignaga with a purpose. In his criticism of the Naiyayika definition of the proposition, Dignaga points out the ambiguity of the meaning of the word sadhya13. It could mean, firstly, the subject of the proposition (dharmin), secondly, a property possessed by the subject (dharma), and thirdly, the combination of the subject and the property. Dignaga then proceeds to demonstrate that any one of these three cannot be recognized as sadhya. Take for example the statement 'sound is evanescent'. As a matter of fact, the dharmin 'sound' is known to exist (siddha), and it is unnecessary to establish it 10 PS, chap. III, K 125b.2-4, V 43 b.6-44a.1. Cf. NV, p. 110.14ff. 11 PS, chap. III, K 126 a.3: ji ltar rigs pa can rnams la skyon brjod pa de ltar rtsod pa bsgrub par bsad pa la yan, bsgrub bya gtan tshigs dan dpe Itar snan ba bead pa dam bcac bar egyur ro. V 44 a.7. Cf. NV, p. 117.20. 12 NV, p. 110.14--20. 13 PS, chap. III, K 125b.4--7, V 44a.1-3.

Page #5
Prasastapada and Dignaga 173 by means of a reason. Thus the first alternative would violate the statement in NS, 1.1.34: ... sadhyasadhanam hetuh (The reason is a means to establish the sadhya). If sudhya were to mean the dharma 'evanescence', then there would be no example that possesses a similarity to the sadhya. There are things which are similar to sound in their evanescent nature, but nothing is similar to 'evanescence since this possesses no property. Thus the second alternative would contradict the definition of the example in NS, 1.1.36: sadhyasadharmyat ... (Through similarity with the sadhya ...). The third alternative has the defects of the first and the second. The same type of argument is made by Dignaga in his refutation of the Vaisesika definition of the reason 14. Dignaga himself holds that what is to be proved is the subject qualified by the property (dharmavisistadharmin), and termed it anumeya (object of inference) 15. It seems likely that Prasastapada followed Dignaga when he used the term anumeya in his definition of the proposition and explained it by the words dharmavisustadharmin. It has been noticed by STCHERBATSKY and other scholars that the fallacies of the proposition removed by Prasastapada with the word avirodhin are similar to those mentioned in the texts of Dignaga's school16. Dignaga's definition of paksa (proposition) is given in the Nyayamukha as follows: svayam sadhyatvenepsitah pakso viruddharthanirakrtah17. The last word is meant for removing the fallacies of the proposition, and it exactly corresponds to avirodhin in Prasastapada's definition. Attention is to be drawn to the similarity of sadhyatvenepsita with Prasastapada's expression pratipipadayisita. Prasastapada had no precedent to follow in the use of this desiderative form, but Dignaga evidently derived it from the definition of paksa in the Vadavidhana of Vasubandhu, which runs: pakso yah sadhayitum istah = A proposition is that which it is desired to prove18. Uddyotakara, who takes up this definition for criticism, understands that the word ista is intended for distinguishing the proposition from the reason and example which are 14 Ibid., K 140b.3ff., V 56b.6 ff. 15 Ibid., K 111a.6: rjes su dpag pasr bya ba) ni chos khyad par can gyi chos can yin te. V 30a.1: rjes su dpag par bya ba ni chos kyi khyad par du byas pa'i chos can no. Cf. Pramanavarttikabhasya, ed. by R. SANKETYAYANA, Patna 1953, p. 580.14. 16 STCHERBATSKY, ,,Rapports entre la theorie bouddhique de la connaissance et l'enseignement des autres ecoles philosophiques de l'Inde," le Museon, V, p. 129ff. (Cf. Buddhist Logic, vol. I, p. 346, n. 2); RANDLE, Indian Logic in the Early Schools, London 1930, p. 216. 17 Nyayamukha, k. 1, cited in NV, p. 116.7, 9, 17. 18 NV, p. 113.6.

Page #6
174 MASAAKI HATTORI not yet proved 19. A disputant does not set forth a reason and an example with the desire to prove, but he mentions them as a known property of the subject of the proposition and as a known fact in which both the reason and the sadhyadharma are present or absent. Thus the word ista is effective to characterize the proposition as distinct from the yet unproved reason and example. As we have seen above, this idea underlies Dignaga's criticism of the Naiyayika definition of the proposition, and we may assume that Dignaga inherited it from Vasubandhu. In his Tika on the Vadavidhana, Dignaga slightly modified Vasubandhu's definition by adding the word svayam to it, as he thought it necessary to make clear that it is the disputant himself, and not an opponent, who desires to prove his proposition in a debate 20. The definition found in the Nyayamukha was formulated through this process. In the Pramanasamuccaya 21 we find the expression svarupena in place of sadhyatvena. But this does not make an essential change, because Dignaga himself explains in the commentary that svarupena means 'in its own form as a sadhya' but not 'as a yet unproved sadhana'. The essential term ista is of course not omitted in the Pramanasamuccaya. Thus in the case of Dignaga, the process through which he came to formulate his definition of the proposition is clearly traceable. On the contrary, in the Vaisesika system, there is no internal evidence to explain a considerable change from the older theory to Prasastapada's. It is highly probable that Prasastapada was conscious of the implication of the word ista (or ipsita) in Dignaga's definition when he expressed his thought with the desiderative form pratipipadayisita. It has already been remarked by scholars that the theory of the three necessary conditions that an inferential mark should satisfy (linga-trairupya), which Prasastapada sets forth in a verse and ascribes to Kasyapa, is exactly identical with the theory generally acknowledged as being expounded by Dignaga. It has also been pointed out that there 19 Cf. NV, p. 113.5-7: sadhyayor hetudrstantayoh prasanga iti manyamanaih kaiscid anyatha paksalaksanani kriyante tadyatha pakso yah sadhayitum ista iti. 20 Cf. NV, p. 117.1-2: yad api vadavidhanatikayam sadhayatiti sabdasya svayam parena ca tulyatvat svayam iti visesanam. For the authorship of this tika, see FRAUWALLNER, "Zu den Fragmenten buddhistischer Logiker im Nyayavarttikam," WZKM 40 (1933), pp. 294-296. 21 PS, K 124b.7-8 (V 43a.4-5): de yan ran gi no bo kho na batan bdag cdod (cf. Nyayabindu, ed. by D. MALVANIA, Patna 1955, III.38: svarupenaiva svayam isto 'nirakrtah paksa iti). ran gi no bo ses bya ba ni bsgrub bya yi no bos yin gyis sgrub byed ma grub paci no bos ni ma yin no (cf. Nyayabindu, III.39-40: svarupeneti sadhyatvenestah, svarupenaiveti sadhyatvenaivesto na sadhanatvenapi).

Page #7
Prasastapada and Dignaga is a close affinity between Prasastapada and Dignaga in their theories of fallacious reasons. And the dependence of Prasastapada on Dignaga was accepted by STCHERBATSKY and other scholars on the basis of the fact that these theories are ascribed by the later Bauddhas and their opponents always to Dignaga and never to Prasastapada 22. However, some other scholars like FADDEGON and RANDLE did not admit the chronological priority of Dignaga to Prasastapada and assumed that both of them had a common source from which each derived his theory independently of the other 23. When we examine Dignaga's criticism of the Vaisesika view of fallacious reasons, we cannot but believe that Prasastapada changed the older Vaisesika theory under the influence of Dignaga. I will show some evidence below. Dignaga quotes VS, 3.1.10-11, which he understood as mentioning three kinds of anapadesa (fallacious reason), i. e., aprasiddha, asat and samdigdha. According to Dignaga's interpretation as explained by Jinendrabuddhi, a reason is to be recognized as aprasiddha when its relation to what is to be proved is 'not well known'. Suppose, in order to prove the proposition 'there is fire on the mountain', one states the reason because of smoke'. If the causal relation between smoke and fire is not well known to either or both of the disputants through previous experiences, this reason is called aprasiddha. On the basis of this interpretation, Dignaga points out that aprasiddha is not a fallacy of the reason. The relation of smoke with fire is not well known when one has never seen or does not remember a hearth and the like, in which smoke is co-present with fire, or when he, though experiencing the co-presence of smoke with fire, does not realize that smoke is the effect of fire. Thus the fallacy of aprasiddha is to be found not in the reason but in the example: the fallacy lies either in the fact that there is no example to be cited (drstantabhava) or in the fact that smoke is not well known in the example as a cause for proving fire (hetvarthenaprasiddhah). Another possible interpretation may be that aprasiddha refers only to the object to be proved: if fire is not well known, the reason 'because of smoke', which is stated to prove the existence of fire, is characterized as aprasiddha. However, this interpretation would violate VS, 3.1.9: prasiddhipurvakatvad apadesasya, which is intended to say that one should have perception of fire as related with smoke before he mentions smoke in the reason 24. 175 22 STCHERBATSKY, op. cit.; FRAUWALLNER, "Candramati und sein Dasapadarthasastram," Studia Indologica, Festschrift fur W. Kirfel, Bonn 1955, pp. 71-73. 23 FADDEGON, The Vaisesika System, Amsterdam 1918, pp. 319-320; RANDLE, op. cit., p. 188. 24 PS, chap. III, K 1478.4-7, V 62b.8-63a.3.

Page #8
176 MASAAKI HATTORI Prasastapada also quotes VS, 3.1.10-11, and interprets them, like Dignaga, as mentioning three kinds of anapadesa. However, he differs from Dignaga in his understanding of the meaning of aprasiddha, etc. He bases his explanation of fallacious reasons on the theory of the triple-conditioned inferential mark (trirupalinga), and adopts the terms asiddha (unreal reason), viruddha (contradictory reason) and samdigdha (dubious reason) for aprasiddha, etc. in the Vaisesikasutra 25. It is not clear which one of these three was recognized by Prasastapada himself as corresponding to aprasiddha. Since samdigdha'is common to the Vaisesikasutra, he must have understood by the word aprasiddha either asiddha or viruddha. He distinguished from samdigdha the fourth type of fallacious reason, i. e., anadhyavasita (inconclusive reason) and states that the word aprasiddha in the sutra is to be applied also to this type of fallacious reason 26. His explanation that anadhyavasita is asiddha in either direction seems to show that he identified aprasiddha with asiddha27. However, his commentators Vyomasiva and Sridhara take aprasiddha as standing for viruddha, for the reason that viruddha is not known to exist (aprasiddha) in the instances similar to what is to be proved (sapaksa) 28. This interpretation is also acceptable as representing the idea of Prasastapada himself, because he explains the second condition of the inferential mark that a viruddha does not satisfy with the words: yad ... anumeyadharmanvite ... prasiddham ... (that ... which is known to exist ... in things possessing the property of what is to be proved ...) 29. Whichever the case may have been, it is noticed that the interpretation of aprasiddha proposed by Prasastapada or his commentators is entirely different from that given by Dignaga. 25 PDhs, p. 204.24-26 (p. 562.14--17), p. 238.9--10 (p. 604.13-14). 26 Ibid., p. 239.13 (p. 605, 15--16): ayam aprasiddho 'napadesa iti vacanad avaruddhah. 27 Ibid., p. 239.11--12 (p. 605.14-15): so 'nyatarasiddho 'nadhyavasayahetutvad anadhyavasitah. For the meaning of the term anyatarasiddha in this passage, see RANDLE, op. cit., p. 213, n. 1. anadhyavasita is illustrated by Prasastapada as follows: satkaryam utpatteh -- Proposition: An Effect exists in its cause), Reason: Because it originates (from its cause), PDhS, p. 239.12 (p. 605.15). This reason is mentioned in the Vadavidhi as an example of viruddha. In his criticism of the Vadavidhi theory of fallacious reason, Dignaga points out that this reason is either asiddha or asadharananaikantika. Cf. PS, K 145a.6, 145b.1-2, V 61 a.4, 61 a.5-6. 28 Vyomavati, p. 569.25--27: aprasiddho 'napadesa iti viruddhavaro. dhah, tasya sapakse 'prasiddhatvat. Nyayakandali, p. 205.2-3: aprasiddha iti viruddhasadharanayoh parigrahah tayoh sadhyadharmena saha prasiddhyabhavad ahetutvam. Cf. Candrananda on VS, 3.1.10. 29 PDhs, p. 201.18-20 (p. 562.10-12). Cf. also ibid., p. 237.16-17 (p. 603.4--5): yad ... tatsamanajatiye sarvatra samanyena prasiddham ...

Page #9
Prasastapada and Dignaga 177 It seems that Dignaga understood the meaning of aprasiddha in accordance with the traditional interpretation of the same term by the Vaisesikas. In Jinendrabuddhi's commentary on PS, chap. II, there is a passage cited from a certain Vaisesika work, which explains anapadesa with the following words: "Smoke, for instance, is an anapadesa for a person to whom the relation of smoke with fire is aprasiddha"30. Apparently the thought that is expressed in this passage is reflected in Dignaga's criticism of the aprasiddha fallacy. The work is named Bhasya. It may be the Bhasya of Ravana, whose theory of perception is referred to in PS, chap. I, or the Bhasya of an anonymous author, on which Prasastapada is said to have written an extensive subcommentary. Certainly it is not the Padarthadharmasamgraha which is called Prasastapadabhasya, nor is it a post-Prasastapada work, because the interpretation of the word aprasiddha given in it is acknowledged, in contrast with that in the Padarthadharmasamgraha, to be more faithful to the original idea of the Vaisesikasutra. Originally the Vaisesikas recognized only two kinds of fallacious reason, i. e., asat and samdigdha mentioned and illustrated in VS, 3.1.11--1231 The sutra 3.1.10: aprasiddho 'napadesah immediately follows the sutra which states that a valid reason is preceded by the universal knowledge (prasiddhi) of the relation of an inferential mark (linga) to its possessor (lingin)32. Therefore it seems clear that the sutra 3.1.10 is meant for giving a definition of fallacious reason in general with the word aprasiddha: = A mark whose relation to its possessor is aprasiddha is a fallacious reason. Perhaps the same sutra came to be interpreted by some of the Vaisesikas before Dignaga's time as mentioning aprasiddha as a type of fallacious reason distinct from asat and samdigdha. However, the import of the word aprasiddha as referring to the relation of an inferential mark to its possessor was not changed by them. Dignaga states that three kinds of fallacious reason are recognized by the Vaisesikas, but evidently he does not know any new interpretation of the word aprasiddha. In his criticism 30 Visalamalavati 'Pramanasamuccayatika, Tibetan version, Peking ed., Mdo-hgrel CXV Re, 132 b.6--133 a.1: de ltar ni bsad Ogrel las"... gan sig gi du ba la soge pa rnams me la sogs pa rnams dar brel pa rab tu grub pa ma yin pa deci nor du ba la sogs pa rnams dmige kyis bstan pa ma yin paoo" ses pa ... 31 Cf. STCHERBATSKY, op. cit., p. 168; FADDEGON, op. cit., p. 302; RANDLE, op. cit., pp. 190 191. 32 It is stated in VS, 3.1.8 that a linga is samyogin, samavayin, ekarthasamavayin, or virodhin of the lingin. In reference to this sutra, the word prasiddhi in VS, 3.1.9 is understood as signifying the universal knowledge concerning the relation of a linga to the lingin. 12 WZKSA XVI

Page #10
178 MASAAKI HATTORI of the Vaisesika theory, there is no trace to show that aprasiddha was taken to imply asiddha or viruddha. The asiddha fallacy is not acknowledged to have been known to the Vaisesikas or to the Naiyayikas before a systematic classification of fallacious reasons based on the theory of linga-trairupya was made by Prasastapada. The explanation given by Vyomasiva and Sridhara for identifying aprasiddha with viruddha is obviously foroed. That which is a prasiddha in similar instances is not always a viruddha. What constitutes the characteristic feature of viruddha is that it proves the opposite of what is to be proved. Thus it seems quite likely that Dignaga's criticism impelled Prasastapada to abandon the older Vaisesika theory of the fallacious reason. Instead of introducing modifications to the older theory, Prasastapada found it more convenient to adopt wholly Dignaga's classification of fallacious reasons based on the theory of lingatrairupya. Apparently aprasiddha was no longer recognized by him as a type of fallacious reason. He cited the sutra with a pretence of following the tradition, but his commentators were compelled to a forced interpretation of the older terminology. The other two fallacious reasons, asat and samdigdha, are illustrated in VS, 3.1.12: visani tasmad asvo visani tasmad gauh (It is a horse, because it has horns; It is a cow because it has horns). Dignaga says first that the reason itself, i. e., hornedness (visanitva) in both illustrations, is neither absent (asat) nor doubtful (samdigdha): it is a well-known property of the subject of the proposition. This is no more than a quibble: it is clear that asat and samdigdha in the sutra are meant for expressing the reasons which prove respectively that which is absent and that which is doubtful. But, Dignaga continues, that which is absent (asat) is not proved by this type of reason: what is proved by the reason 'hornedness' is not the absence of a horse, but it is that which is contrary. (viparita) to a horse. "Therefore, this reason) is called a contradictory reason (viruddha) as it proves the opposite (of what is to be proved] (viparitasadhanat) 33." Both Vyomasiva and Sridhara identify asat with asiddha for the reason that asiddha is absent in the subject of the proposition 34. But, as stated above, the asiddha fallacy was not known to the older Vaisesikas. Evidently asat in the Vaisesikasutra corresponds, as Dignaga 33 PS, K 147 a.7-147b.2 (V 63a.3-6): ... deci phyir edi ni Ogal ba yin te bzlog pa sgrub par byed paci phyir ro. 34 Nyayakandali, p. 205.4: ... dharmini vrttyabhavat... Vyomavati, p. 569.26-27: asann ity asiddhah, tasya hi paksadharmatvenasattvat. Cf. Candrananda on VS, 3.1.11.

Page #11
Prasastapada and Dignaga understood it, to viruddha. Prasastapada cites 'visani tasmad asvah' as an example of viruddha 35. Noteworthy is the phraseology that Prasastapada uses in his explanation of viruddha. It is surprisingly close to Dignaga's expression. Prasastapada says: "The reason ... which is present in the opposite of the object of inference, is a contradictory reason (viruddha), because it proves the opposite of what is to be proved (viparitasadhanat) 36" It is highly improbable that Dignaga criticized the Vaisesika theory with the phraseology borrowed from Prasastapada or his Vaisesika predecessor. Dignaga fully discusses elsewhere the fallacies of reason on the basis of the hetucakra, which he invented to examine all the possible relations between a reason and a sadhyadharma 37. The characteristic feature of each type of fallacious reason is made clear by him through this procedure. Prasastapada also has a clear notion of each fallacious reason, but his description of it is rather concise. He seems merely to summarize the theories elaborated by someone. But Dignaga's criticism shows evidently that the distinctive feature of viruddha was not properly known to the pre-Dignaga Vaisesikas. It is almost certain that Prasastapada is indebted to Dignaga for his explanation of the viruddha fallacy. As for the reason that causes doubt (samdigdha), Dignaga criticizes the Vaisesikas for their mentioning only the one which is present both in things homogeneous with the subject of the proposition and in things heterogeneous to it, which Dignaga calls sadharananaikantika, and for their not mentioning asadharana and viruddhavyabhicarin 38. asadharana means that which is present exclusively in the subject of the proposition, as for example the audibility (sravanatva) of sound, while viruddhavyabhicarin signifies a pair of valid reasons which, when combined, proves two contradictory properties predicated of the same subject, thus causing doubt. These two types of doubtful reasons are classified by Prasastapada under the head of anadhyavasita, which is not found in Dignaga's system. It has already been shown by scholars that the view to which Prasastapada refers with 'tti kecit' when distinguishing anadhyavasita from samdigdha is the one held by Dignaga 39. In the light of Dignaga's 179 35 PDhS, p. 238 (p. 604.25). 36 Ibid., p. 238.17-19 (p. 604.23-24): yo hy anumeye 'vidyamano 'pi tatsamanajatitye sarvasmin nasti tadviparite casti sa viparitasadhanad viruddhah. 37 PS, chap. III, K 131 b.6ff., V 49a.6ff. 38 Ibid., K 147b.2 ff., V 63 a.6 ff. For the same reason Dignaga criticizes the Vadavidhi, cf. PS, K 145a.7, V 61 a.5-6. 39 The explanation of viruddhavyabhicarin in the Nyayamukha is cited by Prasastapada, cf. Tucci, The Nyayamukha of Dignaga, Heidelberg 1930, p. 31, n. 58.

Page #12
180 MASAAKI HATTORI criticism, we may know the incentive that urged Prasastapada to create a new category. He intended to defend the older Vaisesika theory against Dignaga's attack. The reason for the marked difference between Prasastapada and the Vaisesikas known to Dignaga in their theories of the example may also be explained if we place Dignaga between the two. Dignaga mentions four kinds of fallacious example, among which the last one is that which simply shows the simultaneous presence or simultaneous absence of the sadhyadharma and the reason without stating the invariable relation between them 40. In the passage explaining the fallacy of this type of example, the Vaisesika definition 'ubhayaprasiddho dTstantah' is cited by Dignaga. According to Dignaga, one has to show in the example that the reason is pervaded by the sadhyadharma 41: the reason should be accompanied by the sadhyadharma on the one hand, and it should be on the other hand absent where there is no sadhyadharma. The relation of pervader and pervaded between the sadhyadharma and the reason should not be reversed. If that which pervades the sadhyadharma were stated as the reason, it would not be sufficient to prove the proposition, since it allows the case in which there is the reason but not the sadhyadharma. Thus, the example which shows the mere simultaneous presence of the sadhyadharma and the reason is not a true example, because the simultaneous presence of the two does not prove that their relation is invariable; and moreover it allows the possibility of the relation of pervader and pervaded being reversed. Prasastapada must have been well aware of Dignaga's theory when he observed, with terms similar to Dignaga's, that the exemplification (nidarsana) is the showing of the invariable conformity (anuvidhana) of the inferential mark to the object of inference, or that it is the showing of the absence of the inferential mark in that which is opposite to the object of inference (anumeyaviparyaye lingasyabhavah) 42 The classification of an example into two, the one through similarity (sadharmya) and the other through dissimilarity (vaidharmya), is found in the Nyayasutra, but the relation of pervader and pervaded between the sadhyadharma and the reason was not properly understood even by Vatsyayana 43. Dignaga emphasized his explanation of this relation, while Prasastapada simply states it without giving any explanation. This, I believe, is evidence of Prasastapada's dependence on Dignaga. 40 PS, chap. IV, K 152b.2--4, V 63 a.6--7. 41 Nyayamukha, k. 11 = PS, chap. IV, k. 2, K 148 a.6, V 63 b.8. Cf. NV, p. 129.11ff. 42 PDhS, p. 246.15-18 (p. 611.16--19). 43 Cf. Nyayabhasya, ad sutra 1.1.35--37.