PRE-DINNAGA BUDDHIST TEXTS ON LOGIC FROM CHINESE SOURCES TRANSLATED WITH AN INTRODUCTION, NOTES AND INDICES BY GIUSEPPE TUCCI Member of the Academy of Italy 1929 ORIENTAL INSTITUTE BARODAPe OL ay vies) DRUBL pate. a arse ae egecwa cena ; eae fall No... Sef Te cee Printed by P. Knight at the Baptist Mission Press, 41, Lower Circular Road, Calcutta and Published on behalf of the Government of His Highness the Maharaja Gaekwad of Baroda by Benoytosh Bhattacharyya, at the Oriental Institute, Baroda. Price Rs. 9-0-0.TABLE OF CONTENTS. Preface o Abbreviations Introduction ‘Tarkadastra Upayahrdaya Vigrahavyavartani Satasastra .. Notes on TS Notes on UH. Notes on VV. Notes on 88. PART Il. Norzs. Inviczs. Index of Chinese Nyaya terms .. Index of works—a) Indian ie b) Chinese Index of Indian authors . S Index of Chinese authors a ‘Index of Modern authors a Pace. 1-1 V-ViL IX-XxXxX . take re 177 1-89 1-14 15-22 23-43, 44-58 61-82 83 84 84 86PREFACE. To my friend Vidhugekhara Sastri 0888 xpetooor # $idos cops. In this volume I have collected, translated, and commented upon the texts on logic or bearing upon the history and develop- ment of Indian logic, preserved in Chinese sources and anterior to Ditnaga. Two of these texts, viz., the Upayahrdaya and the Tarkagastra, have been translated not into English but into Sanserit, for the following reasons. First of all it has been proved that the more a text is translated into different lan- guages, the more it is subject to alterations. Secondly these works are very technical, since they are concerned more with rules of dialectical debates, vivada, than with epistemological theories. Although their bearing upon the first develop- ment of Indian logical theories is great, their theoretical interest is not so wide as that of some of later treatises like those of Dinnaga and Dharmakirti. Moreover, the contents of these two texts are such as to appeal very much to many of the naiyayi- kas of modern India who are masters of Sanscrit, but very often possess a scanty knowledge of English. When we write on Indian subjects we must try, if possible, not to forget the large mass of the Indian pandits, whose deep learning I had very often the opportunity to admire, and from whose collaboration our studies might be largely benefited. As regards my transla tion I must say that it has not been my purpose to give any hypothetical restoration of our texts into Sanserit. Even though the Chinese seems very often to adhere quite strictly to the original, we cannot state that it is literal in every passage. As a rule, the Chinese translators do not reproduce their text, verbatim, as the Tibetans do; they try to render the original in * such a way that it becomes intelligible to Chinese readers. It is evident therefore that my Sanscrit text is nothing else than a retranslation into Sanscrit, which is perhaps often very near to the original, but which, in no way deviating from the ex- tant Chinese texts, cannof claim to be arestoration. It is there-ii PREFACE. fore needless for me to say that my Sansorit does not pretend either to be elegant or perfectly idiomatic. My only purpose was to render as faithfully as possible the Chinese text and to be correct. The greatest difficulty has been found in identifying the various logical terms that occur in our treatises ; but a careful comparison with the extant Sanserit sources and the Tibetan translation of some analogous passages, as in the case of the Jati-section of the Tarkasastra, has helped me very much in finding with certainty the original Sanscrit form that was the basis of the Chinese. Since these logical terms represent some of the greatest difficulties that one meets in the Buddhist philo- sophical texts preserved in Chinese, they have been collected and arranged in the Index I. Considering that they may be of a wider interest, inasmuch as they are not concerned with mere logic, but also with some fundamental aspects of Mahayana Buddhism, the other two texts, viz., the Vigrahavyavartani and the Satasistra have been translated into English. Tf the Vigrahavyavartani has a great bearing upon the history of Indian logic, since it embodies the oriticism of Nagar- juna relating to the theory of the pramanas, the Satagastra, though it cannot be considered as a logical work, contains some references to the Nyayasitras, the importance of which is self-evident. Therefore, I have thought it to be useful to add here the complete translation of this work, which, though al- ready translated by me in an Italian review, has been accessible only to a few scholars. Its importance cannot be sufficiently emphasized. It is useless to say that these treatises also have been translated as literally as possible, so that the scholars who are interested with this same department of research, but do not know Chinese, may have an exact idea of the original. In defect of the Sansorit which is lost, the Chinese is the only text at our disposal. The “Notes” were composed when the four texts had already been printed : so, they embody also the corrections and emendations to them. They should therefore be always con- sulted while reading the texts. I must apologize for my English, but I hope that it willPREFACE. iii not be forgotten that this is not my own language, while owing to my continuous travelling it has been impossible for me to request some of my English friends to undergo the painstaking task of reading my writings. Before concluding these introductory remarks I must ex- press my gratitude to my friend Vidhuéekhara Sastri, Principal of Visvabharati, whose advice was often required and to whom I am indebted for many valuable suggestions; to the Pandit Herambhandtha Tarkatirtha, with whom I discussed some points of the Upayahrdaya and the Tarkagastra ; to Doctor Benoytosh Bhattacharya, Director of the Gackwad Oriental Series, and Baroda Oriental Institute, who undertook the publication of this volume in the Series that he directs with the acknowledged competence and who made all possible arrangements for its printing. Nor can I forget the greatest care shown by the Baptist Mission Press, the only one in India which can under- take publications of this sort, in carrying my book through the press. To all my krtajfata. Dacoa. Grusseen Tucot. January, 1929,As. ASS. BOAT. BEFEO. .. os. HIL. .. TA, THQ. JA. JAOS. .. JASB. .. JRAS. .. MMV. MS. on MSB. MV. He NB. “ NBI. NSB... ABBREVIATIONS. Asangas works, Abhidharmasangitisastra by As., Nanjid 1199, it Wk 52 HB aa Abhidharmasamyuktasangitigastra by Sthiramati, Nanjio 1178 Bal At 32 BE WE EB Bodhicaryavatara by Sintideva—with tika ed. by L. De La Vallée Poussin. Bibl. Indica. Bulletin de I’Ecole frangaise de 1’Extréme Orient. Carakasamhita, ed., Jivananda Vidyasigara, Cal- cutta, 1896. History of Indian logic by Vidyabhigana, Cal- cutta, 1921. Indian Antiquary. Indian Historical Quarterly. Journal Asiatique. Journal of the American Oriental Society. Journal of the Asiatic Society of Bengal. Journal of the Royal Asiatic Society. Khandanakhandakhadya. © Chowkh. Sansor. Series, Benares. Molamadhyamikakarikas, edited by L. De La Vallée Poussin Bibl., Ind-Buddhica. IV. Commentary thereon by Candrakirti (Prasanna- pada), ibid. Mimamsasttras (Kashi Sanser. Series, 42, Bonares 1910). Comm. thereon by Sabarasvamin. ibid. Mahavyutpatti. ed. by Mironov, Bibl. Buddhica, XII. Nyayabindu, by Dharmakirti (Kashi Ser. Series, 22, 1924), Tika thereon by Dharmottara. Vatsyayana comm. on the Nydyasittras, ed. by G. N. Jha (Chowkh, Sor. Series, 1925).vi ABBREVIATIONS. NK. Nyayakandali, Vizianagam Sor. Series. IV, 1895. Nyayamukba by Ditmaga, Nanjid, 1223, 1224 and my translation (Heidelberg 1929). NMe. .. comm. thereon by Shen-t’ai FR] BA iE ZH Pa ait a 2 NMJ. .. Nyayamafijari by Jayanta, Viz. Ser. Series VIIL. i 1895. NP. .. Ny&yapravesa Gaekwad’s Oriental Series, XXXIX, Part II. NPe. .. Comm. thereon by K’uei-chi FA] BA AL JE Ei Dit. NR. .. Nyayaratnakara on SV. NS. .. Nyayasitras in NSB. NSN. .. Nyayasicinibandha—in NSB. NV. .. Nyayavarttika ed. by Dvivedin, Benares 1915. NVIT. .. Nyayavarttikatatparyatika, Kashi Sanser. Series, 24, 1925. NVTTP... Nydyavarttikatatparyatikaparisuddhi. Bibl. In- dica. PAV. .. Prakaranaryavacasastra, Nanjié 1177, 1178 a EB A it. PB. _.._Prasastapadabhasya, Benares Sor. 8. PKm. .. Prameyakamalamirtends, Nirnayasigara Press ed. PM. _—.. Pram&namimamsa, Poona, (Arhatamataprabha- kara I). PS. :. Pramanasamuceaya, Mdo, ce (Narthang edition, University of Caloutta copy). PSV... vrtti., Ibid. ‘ SDS. .. Sarvadarganasafgraha, Poona, 1924. SDSm. .. Saddarganasamuccaya, ed. Suali, Calcutta, Bibl. Indica. Sk.G. .. Satkhyakarika, with Gaudapada’s comm., Chowk. Sor. 8. Sk.M. .. Id. with Matharavrtti, Ibid. 8s. +. Satasastra—by Aryadeva, in this vol. SSo. .. Comm. thereon by Ki-tsang Ff a& iif. .sv. TS. TSP. UH. VS. vv. vi. ZI YBS. ABBREVIATIONS. vii Slokavarttike, Chowkh. Ser. 8, Tarkikaraksé. From ‘the Pandit’ 1903. . Tattvasangraha by Santiraksita, Gaekwad’s O.S. Commentary thereon by Kamalaéila. Text of the Tarkaéastra as translated by me in this vol. Upayabrdaya as translated by me in this vol. Vaisesikasitras, in PB. Vigrahavyavartani. Vadavidhi. Zeitschrift fiir Indologie and Iranistik. Yogicaryabhimisastra by Maitreya and Asanga sy S00 (a SL iaINTRODUCTION. LL Tue Texts. I, Nanjid 1252. Tarnasasrra, 4p] cay Bac Tt is a mere fragment containing a pteliminary chapter on the wrong discussion and two other sections on the jatis and on the nigra- hasthanas. The work was translated into Chinese by Para- martha during the Liang Dynasty (552-557 A.D.). Paramartha also wrote a commentary upon it, called Yr EF 2A if, (in three sections) which is now lost. The book was widely known not only in India but also in Central Asia where Dharmagupta, if we are to follow the Chinese tradition, explained it. This text has been attributed by some authors of the Chinese catalogues of the canon to Vasubandhu. But there is no definite’argument to support this view. There are of course many similarities between this text and Vadavidhi, so far as the jatis are concerned, but this fact can be well explained, since before Diinaga there was very little disagreement in regard to this particular point. But there are more important discre- paneies, such as with the theory of syllogism. Our text follows the pastedvayava-doctrine, while Vasubandhu reduced the mem- bers of the syllogism only to three—as stated by K’uei-chi in his commentary upon NP. Chp. 1°. 1 Cfr, Bacout. Le canon Bouddhique en Chine, I, p. 423. (6). 2 Ofr. Siu Kao Seng Chuan chpt. IT. (Taishd Bd. Vol. 50. p. 435) Cir. 8. Lavr in JA, 1913, X1s., Vol. Il, p. 349. Ux, Vaisesika Philosophy, p. 84, n. 4. 8 Taishd ed. Vol. a4 p. 94 f+ tit ES) Se ie # iL a Rau ASO-ROIARE + Bodhisattva ‘Vasubandhu in Vadavidhana, ete. says that the syllogism threefold: pratijna, hetu, dretanta.” But, strange enough, Hui-chao = YZ. inhi OW Ea Ey attributes to Vasubandhu the theory of a fivefold syllogism. Taishd ed. Vol. 44, p. 144, 155, and 157. Though no express statement is found in this connection in his work, it seems that he does not consider the Tarkeéastra as a work of Vasubandhu. Cf. ibid. p. 144 and 157.x INTRODUCTION. One could also quote NV. pag. 136: avayavatraya evam aksanenopapadite tesam trayo durvihitah (cfr. my note on Vada- vidhi in THQ, vol. IV, p. 636). But this passage is not conclu- sive, since it may mean either that Vadavidhi did not take into consideration the upanaya and the nigamana, because it did not regard them as members of the syllogism, or that it accepted, without objection, their definitions as given by NS. The attri- bution of “Tarkaéistra” to Vasubandhu has been recently accepted by Mr. Lii Ch’eng ft Jif in his Chinese translation of the Pramanasamuccaya FE FS of Je WS +1: published in [AY 2, IV, 4. p. 58. (ofr. also Ui’s Studies in Indian Phil. I. p. 233). His chief argumentis that TS shows the same classifica- tion of the jatis as VVi, according to the fragments quoted by PSV., and that the three hetvabhasas, viz., asiddha, anaikantika, viruddha of TS, are the same as those alluded to by PS. as being expounded in VVi. Moreover VVi considers the viruddha-hetw (in the pararthanumana) as hetu and prat But these arguments cannot be considered as definitive. In fact we do not know very much about the activity of the logicians before Diinaga, and from the mere similarity between the two texts as regards the catalogue of the jatis we cannot deduce that TS., and VVi. are one and the same treatise, because it is quite possible that other manuals on Vada or Vivada, in which a similar classification was followed, were in circulation at that time. The examples given of the three hetvabhasas, as taken from VVi., do not correspond to those of TS., which are, on the other hand, similar to those of VS., and PB. Moreover, it seems to me rather strange that K’uei-chi does not say a single word about our text and its authorship in his works on logic, nor do we find any mention of it in the life of Vasubandhu written by Paramartha, the translator of TS. I know that this fact is not definitive, since he does not speak of any other logical work written by the acirya and because he might have thought, with Diinaga, that the authorship of VVi. was doubtful. Anyhow his silence is worth mentioning. For all these reasons I think that we are not as yet entitled to identify the two treatises. I must add that the title ‘ Tarkaéastra’ is a hypotheticalINTRODUCTION. xi one. But we must not forget that Tarkagastras are alluded to twice in PSV. (III a, 44a anumeyanirdedah pratijaa and IVa 66,6, paksadharmamatram hetuprayogah). The reference which is made here to particular theories on some special points suggests that Diimaga had definite treatises in his mind and did not merely think of NS. or VS., which he included at the end of PSV. among the tarkaéastras, while in the course of the book he always refers to them under their proper names, when criticising their doctrine. Tl. Nanjio1247. Upavamgpava, Fy i upp Sansorit original of the title is doubtful; 77 {§f may corres- pond, in the logical treatises, also to prayoga. However, the restoration Upayakaugalyahrdaya pioposed by Nanjid (cfr. Bagchi, p. 245) cannot be accepted. This work is a very ancient one. It was translated into Chinese twice. The first translation, made by Buddhabhadra of the Eastern Tsin (Bag- chi, p. 346) is now lost. The second one, still extant was made by Ki Kia Ye* at the time of the Northern Wei. The name of the author is lost, but it has sometimes been attributed to Nagarjuna. We have no grounds for accepting such a view. First of all we cannot forget that many works, of unknown authorship, were attributed to this great dialectician. This is the reason why in the Chinese Canon as well as in the bsTan-agyur we find so many treatises ascribed to Nagarjuna which are certainly works of a later date and written by another Nagarjuna (Siddha Nagarjuna). It is true that’ we find in our text several allusions to the Sinyavada ; but, first of all, Sinya is not peculiar to Mahayana only ; éinya, Santa, nivodha are technical and general terms of Buddhist lore, Moreover the passage at p. 12, where mention of Sunya is made, looks like an interpolation, since the pratipatti (or saksatkara or anubhava) was not enunciated before and comes abruptly after the praptakala. On the other hand the sentence: dvadasa nidanani, dubkham, samudayah, nirodhah, margah, saptatrimsat paksah, catvari sramanyaphalany ityadayo 1 Possibly from Central Asia, Serindis, Ofr. Diurevirz, BEFEO, XXIV, 1924. p. 65-66.n, Bacont, p. 245.sii INTRODUCTION. dharma buddhasya sanyagarthah. (p. 7), which expounds the Buddhist siddhanta, as opposed to that of the heretics, em- bodies the fundamental tenets and categories of the Hinayana. The same thing may be deduced from the discussion at p. 18 on the existence of nirvana and of the arhatphala. Moreover, in the present case we must not forget that there is another work, the title of which shows some similarity with that of ours, I mean the 34 j[) Vadahrdaya of Vasubandhu, which is referred to by Shen-t’ai in his commentary upon the Nyaya- mukha (Taishé ed. Vol. 44, p. 77). Whatever the case may be, there is no doubt that the work was considered as an authoritative one, as is evidenced by the fact that it was trans- lated twice. Ki-tsang quotes from it in his commentary upon the SS. Moreover, as we shall see later on, it reflects ancient theories, which are akin much more to the classification of the heuristical categories, as contained in Carakasamhita, than to the present redaction of NS. This treatise has been critically edited,’ translated into Japanese, commented upon and compared with the Caraha~ samhita by Prof. Ui in his most important “ Studies in Indian Philosophy”? (in Japanese) E[] BE Pr BL Hf FE, vol. 1. p. 427 fol. (cfr. also I, p. 204). III. Nanjié 1251. Vieranavvavarrani, fli] Ht : There is no doubt that this was the title of the work and not Vivadasamana as suggested by Nanjid, since there are some stanzas of this work which are quoted with the title itself of the original treatise in the commentary of Candrakirti upon the Malamadhyamikakarikas. This is also the title found in the Tibetan translation of our text included in the bsTan-agyur. (do, Tsa, Corpmmr, Cat. 291, 293.) The Chinese translation was made by Vimokgasena and Gautama (Prajiia)-ruci in the year 541 A.D. (Bacout, pages 263 and 268) during the rule of the Eastern Wei. This translation is far from being perfect. First of all it is not literal; the translators tried to give the general sense of the text, rather than a literal rendering into Chinese. 1 And in fact the text bas not been well preserved and quite evidently it is in several places corrapt.INTRODUCTION. xiii Sometimes they succeeded, but sometimes they failed to repro- duce the concise expressiveness which is the peculiarity of the style of Nagarjuna. Moreover the Chinese style itself is not always perfect ; very often the sentence is modelled upon the Sanscrit original, quite against the rules of the Chinese syntax. So that here and there the meaning is obscure; and it is difficult to understand it correctly without the help of the Tibetan translation. For this reason I have added the edition of the Tib. version to my English rendering, while in the notes I have also explained, according to the Tibetan text, all the passages that in Chinese are obscure or do not reproduce exactly the sense of the Sanserit original. The Vigrahavya- vartani may be considered as a short treatise in which Nagar- juna explains the fundamental tenets of the Sinyavada refuting the objections raised against him by his opponents, heretics as well as Hinayanists.* The text is divided into two parts.” In the first part the opponent is supposed to expound his chief objections; in the second Nagarjuna replies. Moreover, we must distinguish in our text a karika portion and a yrtti, both by Nagarjuna. But they seem to have circulated quite independently, since in the bsTan-agyur they have been translated separately, though in the translation of the vytti the karikas have been regularly incorporated. It embodies a long refutation of the theory of the pramanas which, as far as our knowledge goes, is the first of its kind that has come down to us and which is strictly related to NS. IV. Nanjid 1189. Satagastra, Ff Zfy-* It is so named be- cause it formerly contained 100 élokas, that is sentences of 32 syl- lables each. This also is a fragment. In fact we know from the preface written by Seng-chao { 3 and from the commenta- 1 It was certainly written after the Madhyamika-karikis since a stanza from this work is quoted in the vrtti (p. 36). It was, therefore, a reply to the criticism with which his magnum opus met. 2 Or four if we consider the karika section (also divided into two parts) as indissoluble from the oytti, as it is in Chinese. 2 Bagour: I, p. 198 (93) Cir. Ux. Studies in Indian Philosophy I, p. 282 fol. Purr, in BEFEO. 1911, p. 361.xiv INTRODUCTION. tor Ki-tsang Gy fg that only the first part was translated by Kumarajiva (in the year 404 A.D.) and the second was left out because it was not considered to be very useful in China. Since each Chapter is said to have contained five slokas and the first ten Chapters only have been translated we have here the first 50 sittras or élokas of the entire book. i The sitra portion was written by Aryadeva, the pupil of Nagarjuna, the commentary was written by Vasu who has been identified by Peri and others with Vasubandhu,? It is a polemical work, the scope of which is to establish the exact doctrine of the éinya after refuting other views. The criticism is not only directed against the Vaisesika and the Sankhyas but also against the various Hinayana schools. Together with the works of Nagarjuna this is one of the most ‘important texts for the study of the madhyamika doctrine, And in fact in China and in Japan it was considered as one of the fundamental authorities by the school of the “three Sastras” so called because it based its teachings upon the Malamadhyamikakarika, the Dvadasanikayasastra and our text. There is a great analogy between SS. and Catuhsataka, though the arrangement of the objects is different. (Catuhéataka IX., X.=1X. of 88. XL, XIL=V., VI, XVIL=VIIL)2 The commentary by Vasu follows very often almost verbatim, the big work called Ta-che-tu-lun, Mahaprajiaparamita-sastra attributed to Nagarjuna, which in its refutation of the Hinayana as expounded by Katyayaniputra seems to have been written with the purpose of opposing a Mahayana Abhi- dharma to that of the Sarvastivadins. There is a diffuse commentary on SS. by Ki-tsang (649-623 A.D.): the treatise — SE FX BE by She 1 I also am doubtful about this identification nor do I think that the ‘arguments brought forward by Peri are definitive. ‘That there were two Vasu- bandhus is known from Abhidharmakosanyakhya; L. Dm La Varuin Poussis, Boudahiome, Btudes et materiauz, Vasubandhu et Yasomitra, p. VIII, n. 2. Any- how this question must be taken up again after a comparative study of all the works that have been attributed by the Chinese and the Tibetan tradition to ‘Vesubendhn. 2 Cir, my Studt Mahayanici, I, in Rivista di Stadi Orientali, X, p. 621. 4somaag = INTRODUCTION. xv Fa-she though not a commentary, contains also much useful information about our text; so also the = #& Zr Be by Ki-tsang. V. The chapters of YBS by Maitreya and of the works by Asafga, which are concerned with logic have not been included here, since they have already been discussed by me in a separate paper. I. Tun Contents. In my Paper “« Buddhist logic before Diiinaga” I have tried to give a synthetical and historical survey of the logical theories accepted or elaborated by the greatest Masters of Buddhism from Nagarjuna down to Vasubandhu and to show the relation and interdependence among the different views. I must therefore refer the reader to that paper for a discussion of the material which is collected in this volume. In this Chapter I shall only arrange the contents of our text in synoptical tables, which may enable the reader to be easily acquainted with the theories peculiar to each of them and to put them in a comparative relation with other sources at our disposal. 1 Buddhist logic before Dinnaga (here abbreviated: BLBD.), JRAS. July 1929, p. 451, Op. additions in same Journal October, 1929. The sanscrit names of the seven heads alluded to in YBS. have been proserved in MV 245, 1180: viida, vadadhikarana, vidadhisthana, vaddlanikara, vadanigraha, vadanih- sarang, vide bahukara dharmah.xvi NS. I, Praména | IL. prameya IIL. samsaya IV. prayojana V. destanta VI. siddbanta VII. avayava VILL. tarka IX. nimaya X. vada XI. jalpa XIL. vitands XII. hetvabhasa XIV. chala XVI. nigrahasthana INTRODUCTION. Tue CaTecoriss. os. UB. l vada= | jalpa 1 drstanta, x-x J vitenda (=V, 12) 2siddhanta (=VI, 16) 2-7 dravya, guna, karma, sémanya, visesa, sama- $ vakkyaprasemsi vaya (=II) 34) 8 pratijiia 4 vakyadosa 9 sthapana =XVI, 33) 10 pratisthapand Spramina (=I, 11-14 hetu, drstanta, upa- 18-21) naya, nigamane 6 praptakale 7 hetvabhasa: (=XIIL, 36, 37) 8 chala (=XIV, 15 uttara 16 siddhanta (=VI) 17 Sabdab 35) 18-21 praména (=I) 22 samsaya (=III) IL 28 prayojana (=1V) 24 savyabhicira) Lr 26 vyavasiya (=IX) 9 nigrahasthana (XVI, 39-44, 58) 27 arthapatti 28 sambhava Tt. 29 anujojya 10 jati (XV) 30 ananuyojya 32 anuyoge 32 pratyanuyoge 38 vakyadoga (= XVI) 34 vakyapragams& 36 chala (=XIV) 36 ahetu 37 stapakata} a 38 upalambha (=XIII) 39 parihara 40 pretijaanani ) 41 abhyanujaia, 42 hetvantara 3 arthantara 44 nigrahasthina | xvi YBS.; PAV.) 2 vada ® vidadhikarana ® vadadhisthina (a) sadhya (dharmin and dharma) (6) sadhana (5 avayavas and 3 pra- manas) 8 vadalankara 4 vadanigraha! (@) samnyasa (®) abbibhava (c) dose 4 vadanibsaraga © vade bahukara dharmisTT ap Pe be ee io eee INTRODUCTION. xvii PRAMANA. NS. pratyaksa. anumana. upamana. sdbda. vs. id. id. . . os: id. id. id. id. (aitihya), vy. id. id. id. id. UH. id. id. id. id. YBS., AS. id. id. id? TS. id. id. Booed Vvi. id. id. 2 a PS., NM. id. id. dg i PB, id, id. : oe (SvanrHa]-aNUMANA. (a) parvavat. UH. Ifa man who has seen a child possessing some character- istic marks, some years later comes across an indivi- dual having the same marks as that boy, he infers that this man is that same boy. (p. 18.) : ng-mu? (comm. on the MMK. XVIII, 1-12). Being aware by previous experience that wherever there is smoke ‘Ts’ there is fire, a man who sees smoke knows that in 1 It is usually believed that the Buddhists accepted only two pramanas ; but, as it is evidenced by our texts, this is not true. The Yogaciras, seem to have adopted the theory of the three pramanas as expounded in YBS. and AS. even after the reform of Difnaga. So Sthiramati commenting upon the Madhy- Gntavibhaga vriti by Vasubandhu, (a copy of which has been brought by me from Nepal and will shortly be published) says: ata aha: pram@yatrayam nigrityeti pramanatrayavirodhena: pramanatrayam punah pratyaksam anu: minam agamas'ca. Cf. also Vijndptimatratasiddhi by S. Luvi, p. 26, That this classification was peculiar to the sect, which did not accept the reform of new logic, is proved by'the fact that Haribhadra in his Abhisamayalankaraloka (ist chapter in my forthcoming edition) expounds the same theory : pratyak- sGnumandgamapramanaih (more than once.) 2 Ts'ting-mu a B (=Blu-eye; according to Seng-jui, a disciple of Kumsrajiva, this name is a translation of Sanser., Pifigala) must not be identi- fied with Candrakirti nor with Aryadeva—Cfr. upon him Pzar in BEFEO. 1Ol1 p. 865 n,1 Ut Page, Phil. p. 45-48. Wattmsme Die Mittlere Lehre des Nagarjuna, nach der chin. Version abertr. 1912. p. X which embodies the ‘transl. of his comm.xviii INTRODUCTION. this case also there must be fire. (WALLESER’s Transl. p. 116, Ur, PV. Phil. p. 88 fll.). Gaudapada on SK. 5. Inference of rain on account of the sight of clouds. MSB. As Ts’ing-ma. NSB. (a) Same as Gaudapada. (6) Same as Ts’ing-mu. () Sebavat. UH. Having felt the salty taste of one drop of sea-water, one infers that other drops of sea-water must have the same taste (p. 13). Ts'ing-mu. Perceiving that one grain has been cooked, one infers that all other grains of rice have been cooked. Gaudapada. Same as UH. NSB. (a) On seeing that rivers are swollen one infers that there was rain. (6) When one says that sound is real, but non-eternal, then it may be either dravya, or guna or karma. Since it is neither dravya nor karma, it must be guna. (c) samanyato drsta. UH. When a man knows that by movement he can go from one place to another, and sees that the sun and the moon occupy different places in the sky at different times, he infers that they also move. (p. 14 cfr. SS. p. 28). Ts'ing-mu. SameasUH. Or inference of the atman from pain, joy, ete., (theory of the Vaigesika and Nyaya. ofr. notes on $$.) Gaudapada. do. MSB. As UH. NSB. (a) do. (0) Inference of déman from pain, joy, etc.INTRODUCTION. xix PAaRSRTHANUMANA. Dasavaikalika-niryukti by Bhadrabahu and NB, p. 137: ten members.? NS. .. pratijiia, hetu, dystanta, wpanaya, nigamana. cs. st id ids id. id. id. vv. a id id id. a t UH. . id ids id. [id.] [id.]* AS. o fae td cid: id. id. YBS. ; PAV. id. id. id. sapaksa vipaksa.® ASS... id. id. id. upanaya —nigamana. TS. c id, id, id id. id. wi. id id i se ne PS.NM. .. id id id. _ wo PB. «+ pratijia, apadesa, nidaréana, anusan- — pratya- dhana, — mnaya. TRAIRUPYA OF THE HETU. TS. .. paksadharma, sapakse satta, — vipaksavyituytti. VVi. —.. [probably, though no specific allusion is found in PSV.] PS.; NM. id. id. vipaksasattoa, PaxsABuasa. TS. +. svavacanavir., pratyaksavir., anumanavir., lokavi j ruddha. PS. ; NM. id. id. id. id. agamavir. 1 The two classifications are independent, That alluded to in NB. is mentioned by Diiinaga in NM. comm. on last karika. 2 Though they are not enunciated, we may infer that upanaya and niga- mana were admitted by the author of UH. as it may be gathered from the examples given there. cfr. p. 10. 3 i ¢ i, ar mal A These, according to K’uei-chi, are equal to fia] py and BB tn sapakea and vipakea (dretanta).xx INTRODUCTION. HetrvaByasa. NS. .. savyabhi- viruddha, prakarana- sadhya- kalatita. cara, . sama, sama, ahetu cs. .. ‘saméaya- id. varnya- (id) sama, sama. UBy | canabhia = ids id. id. id. cara § chala yuktiv. drstantav. AS. -. viruddha—d) anigcita, 8) sadhyasama TS. :. asiddha, — anaikantika, viruddha. Wi. id, id. ia. PS.; NM. id. id. id. PB. ted id. sandigdha, anadhyavasita id. Smwpwayta NS... sarvatantra, pratitantra, adhikarana, abhywpagama. OSes id. id. id. id. TS. Se id. id. id. id. UH... id. id. : es (sarvasama), (sarva- —adau samah adaw bhinnak Bhinna), pasead bhin- paécat samah. nah, Cuan. NS. ++ vakehald simanyachala upacéirachala, os. ec id. We UH. id. id. [id] Varvarrasausa (the Vakyadosa is the opposite of it), CS... anyiina, anadhika, arthavat, anaparthaka, aviruddha. UB: i ta id, & id, plus: ananuyo (adigata- vacana- —_yuktiv. jya, padartha), dharmani- drgtantav. baddha,NS. I. sidharmyasama II. vaidharmyasama III. utkarsasama IV: apakarsasama V. varnyasama VI. avarnyasama ® VIL. vikalpasama VIII. sidhyasame praptisama, . apriptisama . prasaigasama . pratidrstantasama . anutpattisame . Samséyasama . prakeranaseama . ahetusama, |. arthapattisama, . avigesasama . upapattisama . upalabdhisama XXI. anupalabdhisama XXII. nityasama XXIII. anityasama XXIV. karyasama INTRODUCTION. Jatt. UE. TS id. f 1 sédharmya id. 2 vaidharmya id. 3 vikalpa 3 | 4avisesa- ia. 3 5 praptyaprapti S| 6 ahetu 7 upalabdhi 8 saméaya ) 9 anukti [10 katyabheda ( 11 prasotiga 2 arthépatti 13 pratidrstanta 8 & i 14 anutpatti 15 nitya, ia. bhedabheda praénabahulyam, uttardlpata praéndlpaté, uttarababulyam xxi VVi. P.S. NM.* id. id. id. id. id. id. id. id. id. id. id, id. [2] id. id. 16 svarthaviruddha [?] id. ia. ia. id. id. ia. id. id. id. id. id. id. id. id. 1 It is interesting to note that if we are to follow Hui-chao op. cit. Vol. 44, p. L , there was another work in which logical theories were dis- cussed and which contained a list of Jatis absolutely similar to that of Tarka- Sastre. This work was the Kk FE aR $8 Bfgj composed by Nagar- juna and translated by Kumarojiva, perfectly with the list of jatie of TS. 2 They are only ditganabhasas (=hetvabhisas, NP. 8). dogodbhivantné jatyuttaraniti, But I could not find anywhere 'mention of such work: anyhow, the fact remains that the table given by Hui-chao agrees CE. NB 183 abhita-xxii INTRODUCTION, hetusama vyapti avyaptisama viruddhe, aviruddha asaméaya érutisame érutibhinna NIGRAHASTHANAS, NS. os. UHe = ‘TS. I. pratijiahani id. id. id. _ IL. pratijfiantara oe (id.) id. IIL. pratijiavirodha -viruddha- (id) id. IV. pratijfiasamnyasa . (id.) id. V. hetvantara id. oe id. YI. arthintara id. oo id. VII. nirarthaka id. as id. (vyartha) VIII. avijfiatartha oe id. id. TX. aparthaka id. (id.) id. (anarthaka) X. apraptakala id, id. id. (atitakala) XI. nyfna id. id. id. (vakyadosa) XII. adhika id. id. id. (vakyadosa) XIII. punarukta id. id. id. (vakyadosa) XIV. ananubhagana a id. id. XV. ajfiana id, id. id. XVI. apratibha ae id, id. XVII. viksepa ae id. id. XVIII. matanujfia id. id. id. XIX. paryanuyojyopeksana id. id. id. XX. niranuyojyanuyoga id. id. id. XXI. apasiddhanta +e at id. XXII. hetvabhisa id. (id.) id.INTRODUCTION, xxiii This is the schematic summary of the logical doctrines that we find in our texts. It is evident that they are specially con- fined to vada or vivéda rules to be followed in the course of debates among the representatives of different schools, a kind of elenchoi* consisting of a definition and of some examples. ‘There were some arguments which seem to have been discussed with special interest by the followers of the various sects, since they were strictly connected with the fundamental tenets accepted by them. The disputes are regarding the eternity or non-eternity of sound, the eternity or non-eternity of atman, the existence of akdsa as an eternal element, the existence or non-existence of nirvana, the relation between the whole and the parts, the existence or non-existence of jati, and samanya, that is all those arguments which are discussed at length in later polemical works and which prove the antiquity and the continu- ity of these views. Except Upiyahrdaya which perhaps reflects some Hinayana views, the other texts belong to the Mahayana and therefore we should not be astonished, when we see that the theories expounded there are those of the momentariness ‘of everything, the impossibility of admitting the absolute and the fundamental voidness of every notion. The systems that are alluded to and the theories which are refuted are those of the Saikhyas, the Mimamsakas, but chiefly the Vaisesikas. This quite agrees with what we know from other sources in which the Vaisesika is always the opponent to whom a greater importance is given. Ill. Our TExTs anv NS. Now the question may arise: can the texts collected and studied here throw some light upon the vexed problem of the date of NS.? In fact it is known that the opinion of scholars differs as regards the time in which the ultimate redaction of the present NS. took place. Jacobi® maintained that NS. and 1 There is no doubt that catalogues of the hetvabhdsas were in circulation since old times ; ef, besides the hetucakra of Didmaiga the list contained in NV. 164 fil. 2 Jacowt, Die Indische Logik (Nach. v.d. K. G. d. W. zu Gttingen, phil. hist. KI. 1901. p. 460. fll.; The dates of the Philosophical siitras (J.A.0.S.' Vol.iv INTRODUCTION. VS. were composed between 200 and 450 A.D. Suali proposed for VS. the date 250-300 and for NS. -the date 300-350.* Stherbatsky had first maintained that NS. assumed its definite shape after the 5th century ; then, he changed his view and held that no certain conclusion could be derived from the fact that there are in NS. references to Buddhist Idealism, since this Idealism is not that of Vasubandhu, but an older one. Ui pro- posed the date 150-250. Prof. Dasgupta is rather sceptical as regards the possibility of fixing the exact date when the Satras were codified.! I think that he is quite right, because we are not in a condition to determine the various strata that were superimposed in the actual redaction of the sitras. First of all we must distinguish at least two different parts that must be- long to two quite different periods, I mean the section dedicated to the pure edda rules and the polemical one.* It is evident that while these may have been inserted in various times, when the system had began to develop and acquire a definite shape, so that it became necessary to defend its tenets against the theories of rival systems, the other section may have a quite different origin as it seems to be evidenced by our texts and other comparative materials at our disposal, such as Caraka- samhita. Adhyaya I and. adhyaya V deal with the vada- rules and have but little connection with the polemicals adhyayas II, ITI, and 1V. We must confess that we do not XXXI, p. 1 fl.) 4 contribution towards the early history of Indian Philosophy TA, 1918, p. 101. 1 Suaur, Introduzione allo studio della jilosofia indiana. p. 14. 2 Srouenparsxy Erkenntnistheorie und Logik nach der Lehre der spateren. Buddhisten, Manchen, Kapitel I, and Anhang II. Nachtrag zu Kapitel I. 3. Ux Vaisesika Phil. p.86. In his important essay upon The Chronology of the systematisation of the Nydya and the redaction of NS. iE 3H a MR eeu ~ ae ‘ ? QA mk awHtke i MK wR K (Studies in Indian philosophy p. 213) he proposes the date 200-250 for the systematisation of the doctrine and 300-350 for the redaction of NS. The essay of Prof. Ui is thé best and the most exhaustive on the subject. Many of my conclusions agree with his views. 43.N. Das Gurra, History of Indian Phil. p. 2 £. 9. 5 That the NS, result from various strata was already stated by Hara- prasida Sastri, J.B.A.S. 1905.INTRODUCTION. xxv know anything about their origin and the fact that they are of Buddhist origin can neither be affirmed nor denied by the sources at our disposal. 16 seems rather possible that when the systems had begun to assume a definite individuality and dis- putes among the various schools of thought increased, the practical importance of the vada-rules came to be recognized by the various sects, so that we may logically assume that different Vada-Sastras, a kind of vademecum for philosophical discussions. always in vogue in India and well attested by Brahminical as well as Buddhist sources,’ were in circulation quite indepen- dently, among the various schools. In all these treatises the heuristical element and purely formal logic are predominant ; in fact the Pramanasamuccaya is perhaps the first extant work in which that epistemological element which was afterwards to be elaborated by Dharmakirti and was to exercise a great influence upon the future development of Indian Nyaya had first come to the forefront. Various specimens of those vademecums of Vadas have come down tous. Besides NS. we have the chapters on vada written by Maitreya and Asanga, the vada sections of Cara- kasamhita, Upayahrdaya and Tarkasastra translated here and Vadavidhi attributed to Vasubandhu. Moreover we should not forget the Tantrayukti in which we find the counterpart of some of the categories of CS. or UH. or NS.* If we compare the lists given in the preceding chapter we shall easily recognize that some- classifications of NS. do not appear to be so old as those of Caraka, Maitreya®, Asaiga or Upayahrdaya. It is sufficient to compare e.g., the lists of th» Nigrahasthanas or of the Jatis; neither in Maitreya nor in Asanga nor in U. hrdaya nor in Caraka is there 1 References to vada and pargat have been collected by VipvApnigaya in his HIL. We may add that according to the traditions preserved in the Chinese Canon the greatest Buddhist masters such as Aévaghosa, Nagarjuna, Asaiga, Vasubandhu, succeeded in establishing their views on account of their dialectical ability shown in public debates, against the most famous opponents; one must remember the vadadwndubhis which were kept at the court of all priness. 2On the Tantrayukti of. Rosen, Zur Friihgeschichte der indichen Philo- sophie (in Festgabe Jacobi p. 346). 3 That Maitreya was an historical ac@rya and not a mythical being has been proved by Ux. On the author of the Mahayana Satralankara ZII. 1928, p. 215. This is also implicitly indicated by Haribhadra in the maigala~ earaya to bis Abhisamay@laik@rdloka,xxvi INTRODUCTION. such a systematical classification of these categories as may be compared to that of NS. Even Upayabrdaya shows a stage which is very near to that of Caraka, but more primitive than that of NS.; on the other hand there is a great analogy, I should rather say similarity, between NS. and Tarkasistra and Vadavidhi.” This fact shows that this section of NS. is closely related, if not as regards its sources, at least as regards the time of its compila- tion, to those Buddhistic treatises, which themselves have many similarity with the logical section of Pragastapada, We can- not in fact, deny that when any new development took place in the elaboration of the Vada-rules the change must have worked almost contemporaneously in all schools, because no discussion is possible without a fundamental agreement as re- gards the chief elements of the discussion itself. The same vada-rules must have been accepted and followed by the two disputants. As regards the second section, its polemical aspect is evi- dent, Sankhya and Mimamsaka, but chiefly Buddhist theories are alluded to and refuted; of course we cannot gather many elements which can help us in fixing the date of this section, because, even when it is granted, as I said before, that it is independent of the first, we have no grounds for assuming that all the parts of this section were entirely written at the same time. On the other hand interpolations are traceable. ‘The allusion to the Buddhist Idealism cannot be used as a definitive proof, as Jacobi thought.? It is wrong to state that Indian Ideal- ism was inaugurated by Vasubandhu, since the idealistic views had already been expounded in such works as Lankavatira- stra, Sandhinirmocana, and already systematised by Maitreya ete And NS. IV. ii. 31, 32 can be quite well interpreted in the light of these books. So also the famous sitra IV, 3 The treatment of the jatis, shows an attempt towards a classification and a reduction of the types which announces Difmaga’s theory. ‘There is in fact agreement between NM and TS. and VVI. 2 fr. Kerra Indian logic and atomism, p. 23, and Srouzrnatsky, Anhang I. But Ui compares the Viméika of Vasubandhu 16, 17 with NB. IV, IT, 29, al. 8 Cfr. my notes on the Laakivatéra, THQ. Vol. IV, n. 3. ges eae yINTRODUCTION. xxvit ii, 26: buddhya vivecanat tw bhavanam yathatmyanupalabdhis tanteapakarsane patasadbhavanupalabdhivat has its counter- part in Lankavatara-sitra p. 53: govisénam anuso vibhaj- yamanam punar apy anavo vibhajyamand anutvalaksanendva- tisthante and p. 207: pralayo nama & paramanoh pravicayapartkea vindéo, ete., (Cfr. NV. 529 paraménavo "pi bhagaéo vivicyamanas tavad yavat pralayak.) More precision is possible as regards the reply to the criticism against the pramanas. There is no doubt that in this case the compiler of this section of NS. meant to refute the attacks of Nagarjuna as contained in his Vigrahavyavartani and implicitly in MMK. As I have shown in the notes to # s treatise it appears from the comparison of the two texts that they are here interdependent. The two objec- tions raised by Nyayasitras against his criticism concerning the antinomy of the “reason ” viz., that if the “reason ” is self-con- tradictory, the refutation would be impossible and that the pra- mana, though an instrument of knowledge, can equally well be an object of knowledge, were known to Nagarjuna. He, in fact, replies to them and shows that the answer of the opponent, viz., dipavat, (he has: agnivat) was understood by him in a way different from that of Vatsyayana (who refers to the wrong interpretation of the Acaryadesiya NVIT. 371) but quite in accordance with NS. V, i. 10. All this would seem to indicate a contemporariness between Nagarjuna and the writer of those siitras, But this does not necessarily imply that the whole of NS. as we have it now, was definitely redacted at that time. In fact, Aryadeva in Satagastra quotes many siitras from NS. These sitras are: III, i, 125 TI, i, 1; INT, i, 18; IN, i, 7; IIL, i, 14. Other sitras seem to have been known to him IL,i, 31; II, i, 41 (ef. notes on $8). But these quotations are re- ferred to by the commentator not as being taken from the Nyaya system, but from the Vaisésikas. This fact seems to indicate that at the time of Satagastra and its commentator, the Nyaya system had not yet originated as a different school, but that the section II—IV was already existent as a part of a Vaisesika treatise. This is in accordance with the fact that Badarayana as pointed out by Bodas does not seem to know the Nyaya,’ 1 This seems also to agree with the fact that no mention of the nydya can be found in old Buddhist Texts.xxviii INTRODUCTION. though he is aware of the Vaisesika and that many Vaisesike- sitras are inserted in NS.1 So we can only state that at the time of Aryadeva and Vasu some of the stitras now belonging to the section II-IV were already existent though belonging to a Vaisesika text and not to the NS. A better deter- mination would be possible if we could use in this connection a passage from Tattvasiddhigastra (IV century A.D.) by Hari- varman. The passage, already signalised by Ui, rans thus: «The six things such as dravya, ete., ave real according to Ulaka, the 26 tattvas are real according to the Sankhyas. The sixteen varieties of arguments (artha) are real according to the Na-ya- siu-mo. FS HI} (4 JE.” (Tattvasiddhi Chp. 23 Taisho edit. Vol. XXXII p. 256 _f,) The chinese term may be restored into Naya-sauma? ; where naya could be taken as a synonym for nyaya; more- over the number 16 exactly corresponds to the number of the padarthas of NS., as they were known to the later Chinese commentators such as Ki-tsang who, for the first time, gives in his commentary on $$ (Chapter Vth.) the complete translation of the first satra of NS. XM A MMR OTK BR — kif R= ROB EER K 1 Ch Ur, Vais. Phil. and Studies, etc., I, p. 190 Al, But there is no doubt that many sltrAs of VS. were quoted by Vatsyayana and wore afterwards con- sidered as belonging to NS. Ctr. f. i, BUBEN Die Nyayasitras. P, 64, 67, ete. 2 The Sawmas are quoted as a particular heretical sect in a single passage as far as my knowledge goes viz., in Bhégyacandra by Raghiittama, (NSB. ed. ps 80) Carvdka-sauma-saugata-jind-'rhata-digambarah gad bahyth siddhantah > naiyayika-vaisesika-stakhya-patanjala-jaiminiya-badarayaniyah sad vaidikasid dhantah. Another reference to the Saumas has been found by Prof. Bagehi in the Akulaviratantra (revealed by Minanathe f, 1. 6) and kindly communicated to me: Samvadayanti ye kecin nyaya-vaisesikas tatha | Bauddhas tu arihanti ye somasiddhantavidinah ||. As suggested by His Holiness Hemaraja Sarma Gura of His Highness the Maharaja of Nepal, in a talk upon this subject that had with him, it is quite possible that in the Somasiddhanta there ia the earliest reference to the Tantric schools. It is known in fact the importance of eandra in mystical theories of the Tantras: while nayottara (naya) is the name of some very old Tantras or rather categories of tantras; as a matter of fact the categories expounded by Ki-tsang as characteristic of the school agree much more with Tantric tonets than with the doctrine of the Jainas.nema | INTRODUCTION. xxix S&H RS TWA am me tb — 6 #8 Se + Ae tO eR RP tA A. “The (followers) of MaheSvara (Pasupatas) expound the aixteen categories : pramaya etc., (=NS. L, I, 1): the category which does not seem to agree completely with NS. is 13= realization.” | a me “self- Moreover, in another part of his comm. (Chp. 7) he quotes among the various other vidyas the nyayavistare FQ, JS ft hi ZB 3a 1A HE TA FE.“ Nyayavistarasastra, which explains the logic of the dharmas.” This means that he was well informed. It is therefore curious that he identifies the Na-ya su-mo (of Tattvasiddhi) with the Nirgranthas or *Jnatiputras, Nataputtas and ascribes to them a series of sixteen topics which, have nothing in common with the 16 padarthas of NS. J EF OK AK AB WN 6 B+ «Nigran- thas called also Na-ya-su-mo.” This statement is surprising specially when we examine the list of the 15 topics attributed to them, which does not contain, any definite analogy with the well-known nirgrantha doctrines. The{Na-ya] su-mos are quoted also in a few other sources,“but unfortunately even from there we cannot gather any substantial information.* 2 ‘The sixteen topies are divided as follows: (a) 8 derived from érutajnana: ii ax A en) 5-8 four vedas, (b) 8 derived from the “ cultivation-mind (E the worship e:-3) of the stars, planets and gods ( 48 TR): 8 the cultivation of the practice of rsis. It is evident that no connection can be found between these doctrines and those accepted by the Jainas 2 These references are to be found in the so called Sii#ralankara attributed to ASvagho3a, but which is probably Drstantapanktt of Kuméralata (not Kal- pandmanditika as given by Opens. Brilchstiioke der Kalp. des Kumdralata. Cte. S. Levi. La Drstantapankti et son auteur. J. A. Tome CCXI 1927 p. 95). This text (translated into French by Hse), contains an allusion to a sect called ope fet Wado as EE RE EE Joiauei-mo which Haber, wes incline to consider viz, 1. astronomy and geography, 2. arithmetic, 3. medicine, 1-6 the cultivation of the six divine practices; 7 the cultivation ofNXX INTRODUCTION. Therefore I think that until new evidence is brought forward as regards this school and its tenets, the passage of Tattvasiddhi cannot be of great help in solving the present question. Anyhow from what has been said above it seems quite probable that the actual NS. is the result of the combination of one section containing Vada-rules posterior to those preserved in CS. or UH. but older than TS. and another, chiefly polemical, the existence of which or rather of some of its sitras can be traced since the time of Nagarjuna and Aryadeva. This codification must therefore have taken place after the appearance of this master and his commentator Vasu—(IIT A.D.); any further determination cannot be reached at present specially when we consider that we do not know anything about this Vasu, and we cannot even assert or deny that he is the same as Vasubandhu. As regards Vatsyayana there is no doubt that he knew TS, since he quotes from it referring to its criticism of the artha- patti (cf. p. 25). The fact that he cites the doctrine of the ten members of the syllogism but says nothing about the reforms of Vasubandhu seems to indicate that the works in which the buddhist doctor had expounded his theories of the syllogism in three members were unknown to him. a transcription of Jyotig¢oma; but this seems to me impossible. ‘The other refer- ence is in the IE FA ZA) Madhytintanugamasistra (apon which seo may Studi Mahayanici, p. 3, n. 1) by Asonga, where the Na-y-su-mo are quoted in @ passage in which logical theories are discussed (it has first been signalised by Uiin Vais. Phil. and in Studies in Ind. Phil I. p. 215).PART I.. ARTA | [Retranslation into Sanskrit.|a urARUAA | Wa Fa FATA | (RATS) WaT AaTasMevanarafafa agaist aware afe was, aereayat aa ate [wadrea] waeaet ara, WHeeTAATa aa ya fae ae- wrat agaadt warai, were aeeatat varia afe qi sara azacaenasa, TATeeAATA- Laval ] sf Fgadiere acqar | aaa | wa qaaanrafata Fadia, adi waren sft wey) Fa eta Aq aeearai afacr- We ated TaTTatigaais aT? | aha agvaata areitfa aa avaarafafa waar qaqa | wa fad, ait awd wrafafa aa qraaaafata at Waa | TTT TATETS- ATS | WawATTAMeTA wAATeTAEA- area ar? waarelal FeeTAeIWSASTAa, TAT Taser aT Wt ausfag A wat | aA- arate | seamed, wan: quel ya- Feats TAT | SATMSTATTSTTUT, fi waAAT Ba) | Tareas: weaafesy | seaTHEws8 AAA | Wa, Wat: WIed G TAS, Ts TwMETAfTERTT ae wed vfseafa? | aaa, TaisaeTet waa Gwaun weaias weala- fafa fata a ar gar adtegqgaricdatrrares fasraTaara: | SUS | Wad BISA ACISATAASAISTATH aT? WMATA, aa aa ata WAH away freq! seus, after Fa sfa 3a, ara: wuserfad: | fassa, ara- erfa: wrefafes | fraeraaaatafa Fe, aeAaa a WaT | qTaSaarrad | srurafafa aq, afeeea | wat FAT yaaatfa 1 att afe gard ya- aatfa a aaa | afe qaadl, afe ta FATT garda yaaalfa Grd faeay) AAA TaA- wrrafata Seater ey | Taya: Ware faeay | HaTAeTa ada Beara Aaa B71, anfe anaraaafeay | weed fe aaant awereteeterst: | cet afeaia sirafasraa: weg 9 sfaa: 9 weadioefar yerafasia, aetna TaaTIMSaeET hs: | aquiataedg taq | Aa aAAqaTAd MTA Saat arafafe weq aera fe awd Fa fara) afe aweafer cat wreafafa area) er afeaaa wedifaan qr! we qadciaa, «ARMA | a marae weqaareiam: | ate Sqaieanstaa weal, ufe faa Sa Sqar | cafaaqarar- fast | aqaraerfaaarerfa: | reafafa, aes ware | TaTaferqareafad, aearafata ofa- farsa | Srafaed Faq agai aA awaH- wrafata catafaesa cantata, ar fe safaa- saree [aat fe] Squeera: artrecea: AAT ATTAT TTA | Sawa: | sar TAAT|TS | artaara: | wat Ae ge, frat Fal BTTa- FE | AIT WaT ATTA | AAATAIT- fafa | aetsatacensfaeatarat = facts waa Tl xe WA Tat we, aM Tafeta xe BR ser wet ee, aa MSG AAT | FET aaaqraal, 7e1 UT AIA | aA ATT fafa varia awafaean | serait avatar ATSIAATT | az (waa)yta wa aaaaafedarfearfefa afeefaera? saat ard faa fata) afe afaeuaeuar ca ae aa feat wae: nfaesmMataRte: | We aeaT Ala | aeragay | aaa seam | ate Haga IEAM WAT TI AY wal afe areq, afe aamquaa, FA aTeread | seitara warafeta afee 1 sea- |é SATAA | STAT aera | TARTS, aaa sera HA] TAT Aqa AA a ahiafa, FTV, TAT ASH HAL Asa AAT ata, afaqaa aarafefa | saaiaereafata ae areas waa fase | TTS TaTha | aaa wa ara aar waar feared g wasn aerfe | waqaaafeia 3, Hanif fafaqafata waTamareatfa St wad wal afe wadt avai Arad, afe warfa sai arafeia ag va ASQAT | AT warez faeia | We yaa | ag (Wadia AA TeTafes- fafa afearat (waat ard) frard ferutda ate awaated, Tet aes aqaafesy | ate aaae- fares, wer asaaa a OTA se aaa A OTH, aa wadrad aa aga aefeafaia | ar aat mi cfeats era afe aganfasfais wa- drat nega | afe wi sarafed, ate wage Ha wiser | afe wage weaafas A aia, afe wa aaaafa Aarfeea) afe Gratijaa agaafed aegaq | aefad aeaa wa freq aarareafasy | cafad a aa: fad, aerfag a wid afe fad, atecfed, vat az (radia aganfestata afaccarey | afe waded AA SWwearrarqurauaa aet a wa waeefars ©aaUTAA | 2 aa aaa wa ad, adi AA wed AG a wea | afeerat (want avs) fara faatsa afe waa Argied ATAUTUA, TSI HATA BVA ag a waife | fi gaara wary GEAAAPTE BWws Ne, WIT TRAMPTG GEA WAH! afe arag (wate) eaATTT ( wUSse Fai WHT fet ), FRAAAGIE Geet Ta WHITA| ATA GISA- RANT Wied Fa WH, AGT Wea GEATT- ATI BL FE? Cea ww AT | Ae VW | Aw a ae faa aa a WeaTquraag | [aa] wa wed ay ware a TST aTfeT | sua ausafaia ausarata | afe aaaus- araTauTa Geafafa aa TA, ag- are 9 SWwafaa sa A A | Aa YaeUEAE aT wareqarad ona) wafiewsd aa a oaracqurad ooraifa, adteeq facta Bawa aTTATEG WAR! we ada BWA AATANTAG WTA | sa AarqUTAAT- SATE | AAT Wet ATH aq wWa- fafa 3, saqurenta aq: ced quad Ta aa Wa | Wee MWe TATRA GUE ATA aq wafafa aq, fadteaft qed aTaTaqTa wed aa aq waa | fected eed aTaTAg- we awaits aa wad cfr Iq, Wat GEA| Ee = ARTA | AMTSARTS wea ATA aq waa fq Wee Qled AAAMAGUTAA | THIS ATA aq waa sa fed quest avaroamad- wei, Tet WA AA aq Wat aa FRAY | BS YA MEATS Tq, GSAT farserarafa: | afe wary eed ATaATTA, amt wagaawra fareerarata:| ate wary SAAT Ted ae, Ged ay frae- MIA A wad, acrenhe weAAqUTY ees aga fare: | faa ger waewd aa [ad] auwafa, ceiseaqure | yer we waa GIE- arfa aat waTaaruraa | ST UTE aT, a aal Bwanfrwraay | afe UTATTHTTM, TT werrderie: | Tar Taraaitaread WaRTET- way TTT TAY TATAATAT | SUT | Wa Wet Bel Wetfaiareer AST wafa ae HeTaTTAMA? | wa wea faTTTAT | ag (Wadia Heed WS, [ad feed g TaT- fefa afeerat (waa ava) fara fautaa af Heed ya, eure warfeta wee | AA shaEng Bape aaa | € aq, wewd YS waead g varfefa) camETe wafiaad wate, adiaeat fafa waes- weg elit) fare aqara waar watfe aearia qaadtf wugary weatife, wat WaTahe yea aad qedifa warmawed wad) afe wee qaaiiaste wareusd af, afe aaara atest, waa A UI! ae wed Te waddle erat: wa qed a wa eft afe aura wa qa wed wa wld, aa oe warfega a ara | wars- wa Ta wed wera aeqaq afe Surat: yae Buea A TAT | Bea | ART aafe qed worfe a wai asada aaa Wa, Gusd g TofS afreT | aefafed waa, aa terncqaiata afeert (waet arg) faard fautia) afe qaeg uf tanufasrenfawerarraa, aeT war fareeraaros: | aafafa Be, waa ya- Sq afters tenfaerara | afe Baa. gfasaraadt a fareeracrafaeer ware Ta fara ARTA wAAedS: | Tat Bq aalfa aaa aa | Taal Bq a aaa, aa Waa aay | adt a ofamaca fates arate awaaa | af wen wa Squares ufseriua:, aal WaTaaag Wwadiia Seqaa wwaie | 2re ARTE | afa = ate watfy aeerfa Partie geet wagatia aearafe Pantie way acy watt fareerauie | aa aT Taare a ware fermefe agadia Bq ufreroaad faara sia wegen | agadianaiant aerate afeeray (waar ard) feared fautaa | sar ® ufam, sear a waa efa ameraaa) sere wae: ofamear facrteatat satfa | afe aamacat wagat- zaMMesMeal vaesufaaeaca a faests afe- waTawaea wala, afe wan: aria wea waq | weatat A aaATiATa | | aa areata | af aatat wararatat, a agai | Aaa | sentifa a aq) Wa aga wa agalacane}r waganmaiat Haq afe wagaatat a wader aga ufaerd awaaaterctafa aegay | ag (Wadiaad Ydarateraesad serif afeerat (waat ard) fasta fautaa | sreqra- fam waafasra: ofacetaa fragt) aca- afest waafaare: ofaceata faest, a- am qa sfa Aq adare wacdeueay aarfa, waar | afe waratacga, afe waafamsmesican | waeMatat terfa aar ”aaa I Ww vaca seria | waaraaré wafeedt) aaa ate wary Fi usata, afe arias qwwaq | wae! sat war ysqaafae: we afar wafa awarfa fermeraia areata 4 seralaeraqeaad | dae aRadta ATA yaaa seat aferear | WU | aRadia AgaAAaafefa | AA TTE- FARA AST ASA | TITERS AeA | wae aadera arated waif, cer agadtaa- vafeia wegen ae (watt) a Aer wae ATA aaa af afeeray (waat ard) fasta farstaa aa arqeraa sft aga waar waeed wd- feraniafa a at?) afe araadfaeaniafa wel war Se Sa aTasratfa ate we aTaaTaeaed: wa wa fad wierawaa g «fat ami aa wifeareafa cer aa waeea aA wal afe waaiea A Wad, aT waar aeaqard aera | ale waaeaaTA TT sweat waar Farge: | He: fast wa wae wfaee: | afa war WATTA |wa fatty TACT | (ara@ate) ween fafaudrarafa faacta- wieanargied freaawasia | afe qsa- Rafafararaied ar farearay | [aa] fartrewaa | afe ufrerfed que Fase + Uga ATA afeadaawafrayaea | fardaaued cuir! (9) eTeR ISA (2) PIRI way (3) framewwaq (2) afatrawaq (y) are freusag (2) aequasay (9) saefar- wu (F) HaTEIEA (¢) TafATITAA (¢°) ATAATAUSAT | a. [Fa] eTaeewway | TUTTE area fase | ( WTaATE ) “wedisfaar: TARTAN Fz: naire: | vorg fare weenie aafa meal sf sfa emnfaa afar ure1 afz Waa rarereTaTMTT ESR! | faa: RT| ART TITER Ferar efer | [=a] areas AA i Ree oes a, were sfaay sft wrfaa afaaret are) afe (mega) faararn- Suatseastaa sta, vet fai ara? | TaTATH- Seay a ctacictinicaineontnsate | aR area at fa wa wee xf) ard sade maria: | (TATE) TA BUSA fATAA sfas a (@USA) | qa ofa 1 tarfatencarterd fe @q- fefa aaa | wauatut caataradariaaaradt- AT AisfrareraMans tari taceeare faa ata | erase frerafaeaar FATTER aT | ufaaal maariaacerineaeraa Tet waa wa af waa annae weerfaarat erafa oacreafe aaa wea faa ofasraarfa | afe 4 weed: fast ware: fast wafefa | (RTaATE) Hat WweaAqaa Fa sia Al waa Vaart farcaenta- ated | afar Sq: warfare: | aa wera: woe frowarefas | caregfafs- Tarawa | wasge A ae | TaTRaaUTa fandiaq | afe waar ufserfadt tqumeqgqeit waver waqews faq aafseamfas faarat eusata ar [faaa-] auwafafe: | Fa sfa aq faerarercaegarfaaareaasg: aia aq, aerframnfaratee attag 4 waa ef afte) fra edictariataacaee- + famartaitarhatarcecanrs |AB TAMIA | a. FuReIAR | aesuReTed Fee wauaagqaa | (wrearS ) wafer: Wee | ATL? BATHE | aq aaa aaa) aarart fara | WIS AIAN RSS ST: | ufearat ure) afe weet fararntrate faa afa, wer fai ora | afe azauet were, vet a faq sfa ceed | edt adereredi- sf: wezeq faar | (ware) wef: ATH! Tat TT sf: aaaaTa,, TAT Wegisha | ufaarat ure afe azarateecerfaaat wat wife, vet fe area?) weet fa wa aatyaeraAAe | Ae AC (mnears) fratta Ta awa | Fa ef Bq) sere tuartaaareantaeaaaca | | Aa- afar fuafacran Sareea fanaa Paar fannie | carefafantgtartaaed wugfag a uanfa | aisariit: writ Sq: acra- areistaa sft va (Bat) vert: arene faueraretter | Tara AisTTATMTS: | a yataafreriad Bq wet faatsedearfefa ae vat: aaa feud a ada xeafeal Sq: | 3. fame | wee snes faRAISA_AA |aaa | ve (areas) weanfaa: caadteraearezateta weeeaifaaar | ufratal UTS | RAPE Neate ara erate) safe gameeeiq i vaca wad, WeMaTETaNataane | wd yemETat: yaa faire) we caadiaaeniae: az waataaisafaar: | vase faa 1 fautintacansaa | Fa sf Sq sarefaa- squfhansarent fawararen: | wequras- APRarATaaTATA | TARTRATE YA: Waser | ql wen | caTewerfoaeg: feat STTATM TS | waat weeraa sta faired masa) aa faa sfa| aeErIguqerarianaia a qatafa faanfa | marquee froareqaa a afasrafaray | A areTauaata faeareqaa ufawrafacam | Fa cia Aq) TererEeET- aaa a yafsenia | wasg: Te faud a ada carearsa: | waaqaa (Fat) qa wafasrara aueatand efa Aq) wHeTTad fafautqaatiad aareqem | aeqel qufafa vada TaTSae: wes faatiaq | a. afeeIssy | THETA ATT faite yaaeraatairewwaqaa | wast: awaea weg Feretvadae aaUaA | afe after weal safateq: afe after gar sifacfy aga, cf wrfta) ufaaret ore afe aataar varafae waist aat araedr wae waurfatanas: | Fa sia Iq we aaq ayaa! ara fi yaaa awaad vaqafaad weqea | aratafi waifa aafa fafreaqen faardtfa Saif qari) selfeauaiste seat faa vewfaer) Fa afa aq werfeersashe aarafateaarane | war Stu: weet agaise gare cit arew- warfsaaqaraatasg | fau@atacwt Fa sfa Faq waat a war warfzaraa afafagata g Aut faire warzuTsa | aaufaateaiaaat wroafa | aarerfa- caaararararat Ag aTeATaquretaa | a aeafer aT asta gaat fa frarta aarfeta aq Aaafa fare avet aeagat a wed a a fates ara aerate aramafer cet aaa wafaraereny | afe aguretta Bqurce, feat atu amare Bquivered fray merfartiar | qaaiiae: weg Baeza | | AT SRERT sfaaq: | ufaarat ore! Sq: visa ware fea at araral Squqaiea | Bada weerqare - fecip etniasaay | x aa | sqaraarenra Wai: | ae ATA wef fa) aq we vafaquena saat frafa feature wanes | fa Sq: ufast wnaTaa | faudavaaawwaa | Fa sft Ae safes saa: edarararaa: | eearfaarsaiara: WERT: | ara: wae feracerrferaisfar- aareqaa erfaa 1 mera: aterentag- wrefas: | sat fe fda Sqr area | afe fear ufaaturataa sfeaq tqpiaa wera faraidistrand wa waat aeRt yar wareera sft wate) aT EET yaedaq wareaeq) afe yeast wai sagard atagaa fereraa) afe ya wee BAS A Safe MAMTA SHTSATAT ATTTA | maT arraeafefa area afe afe- fafwegfectinaad arafad = watts, seared @ a afaarad aisat facet: | y. oTeorfiteway | Sq areal oratfa a aT? ‘afe avaq arat uraifa aerate) aa Bq ara a ouratfa aera | waearefa- aWwaqaa | ufaarel ure afe Bq eral oTAifa caer aria darters a arvafa) azar aetarfic : waxafufad a gadehaft, Bqeecara! 3as ARMA | afe araafed, cer SquaTaa | ate g ATIATTET fa fasalrte] Bqat) warsqefts: | sa a wraifa war aisaaqqaeares, careqcfas: | ae Vquraawesaas: | safTa eeaTaaat sfeararar SeaTaae: | faudatarawd | feet Bq wafrea- apratga afe waa: quwaqufatgqeatad war fad waa afe aeaatqeatiad cer faadaq | aarfefa Aq) va AAT aq: Tai ararzaafa gq wufaareart are aaa- arena | araarasft aera sar a saat gat sf Aq Are aMTETAeg- frqat an euedsta mehr separa a 9 ageeiq) aarepareagqaineesd &. ages | sara Sierra wereg- BwaTAA | afrarat ure, fi Bq: wrerqa ware. watt oafe araaq: ore arese care arafa ara, Sq aa aH? | TT waa, arasa org, der fad and fa Saat?1 wa BTTATSG: | war yTNSTATAY: ae feai are aT acerca ea TIGA | THT VAATARTA: | fardetransry | ge ef Aq) aad earsaaa | re aga Sqat asva ger dtu: fagararcia wef aafa a afaaartia aeqqareafa | wary qaeafitga & apaaeq euwafs aerfe- qdavarqusa, a q fasH aa awe ade Bqersaeglafa ararat aa Bel aeepraegiies cfs at aA, wet ufafearsaart qatar aoc faar- ware g Bqaiwirasa | ager apaafeat cea Sqaitaaaa | aga vata ys Bgaia drow wad Vqalttavaa | sarfafed Sq: wry fre g warfefa azdtaq | aad afa, fer Sat squfafcia Staq wusaafaay | warfefa 31 waanta [az] aq waragqaiatavaa | aegis fare cara faatafatcia aa, ata feats a dager, ya vraste [ Satie] ararauafa: Tard aTataata: | caren Ta Tara: | ©, wefarwway | fafretqarfearattar- Bieafeia wrufarewaqawa | afearat ore ate meetin erase Sfaareret aa waal a fasa as faaarner | am faqeraratenaal trates Rataay aa daria aarefaae ari a waa aar- We wearers | BTA Va WaT atfea aa aa fare: war) seri frat "oa yet: wis wajanaafatara |