^{page 1013 note 1} Zumu in Chinese translation.

^{page 1013 note 2} P'o-su-pan-tou in Chinese transcription. Tien-chin or Shih-chin in Chinesetranslation; cf. Vasubandhu's, biography: BEFEO., vol. iv, 1904, p. 40.Google Scholar

^{page 1013 note 3} Cf. B.N., App. i, pp. 371–2.Google Scholar

^{page 1014 note 1} According to Tucci: Lun-shih = Vāda-vidhi, Lun-kwei = Vāda-ridhāna, Lunhsin= Vāda-hrdaya. But Prof. A. Vostrikov has proved that the first is the Vādavi-dhāna and the second the Vādavidi. In European literature we have a detailed analysis of the identification of these Chinese titles in Professor Tucci's, article “ Buddhist Logic before ” (see JRAS., July, 1929, p. 482).Google Scholar He had already touched onthe subject in his article “ The Vadavidhi ” (IHQ., 1928, v. iv, p. 635).Google Scholar See also the work of Vostrikov, A., Logical Works of Vasubandhu, ch. i.Google Scholar Vidyabhusana, in his History of Indian Logic, p. 267, translates the third title, Lun-hsin, as Vāda-kauśalya. We have a mention of these titles in Chinese literature in Dignaga's text: Nyāya-dvāra (or mukha)- “ Cheng-li men lun ” and in the commentaries to Nyāya-praveśa- “ yin-ming ju cheng-li men lun ”, written by Hsüen-tsang's disciples. See below.Google Scholar

^{page 1014 note 2} Julien, St.. Journ. As., iv série, vol. xiv, p. 359, No. 245. Jou-chi-lun. 1 livre (Tarka-çcāstra) traduit sous les Liang par Parāmartha.Google Scholar

^{page 1014 note 3} Nanjio, Bunyiu. A catalogue of the Chinese translation of the Buddhist Tripitaka, 1883, London, No. 1252. Zu-shih-lun. “ Tarka-śāstra.” Composed by the Boddhisattva Vasubandhu. Translated by Paramārtha, A.D. 550, of the Khan dynasty, A.D. 557–589; 1 fasciculus, 3 chapters. It agrees with Tibetan.Google Scholar

^{page 1014 note 4} Cf. Appendix, No. 11.

^{page 1014 note 5} Dai Nihon Kōtei dai zō Kyō, vol. xix, bk, 5, pp. 67–73.Google Scholar

^{page 1014 note 6} p. 212.

^{page 1015 note 1} We find similar statements in the Buddhist cyclopædia, Fo-hsio ta zû tian, p. 1098, edited in Shanghai.Google Scholar

^{page 1015 note 2} We find some data, unfortunately incomplete, concerning Ju-shib lun, and the works connected with it in Prabodh Chandra Bagchi's Le Canon Bouddhique en Chine Paris, 1927, v. i. See Appendix 3 (f.n.).Google Scholar

^{page 1015 note 3} Paramārtha's stay in China coincided with two periods of Chinese political history connected with the reign of two dynasties, Leang and Ch'en. Thus the period from 548 to 557 falls upon the Leang and that from 557 to 569 upon Ch'en, therefore he is called sometimes Paramārtha of Ch“en, sometimes Paramārtha of Leang. The Chihyuan fa-pao k'au-tung tsung-lu Catalogue asserts that the translation of Ju-shih lun belongs to the Leang period, when about ten works had been translated by Paramārtha. Bunyiu Nanjio bases his statements concerning chronology on the same Chinese catalogue. See Appendix. See B.N., App. ii, pp. 423–5Google Scholar. About Paramārtha's biography see BEFEO., v. iv, 1904, pp. 3 and 60, and Bagchi, p. 418.Google Scholar

^{page 1015 note 4} Cf. B.N., App. i, p. 372.Google Scholar

^{page 1015 note 5} Cf. O.T.F., No. 24. Ui, Yaiśeṣika Philosophy, p. 84 Google Scholar, f.n. 4. “ Dharmagupta, a Buddhist of Southern India, came to China in A.D. 590 and died in 619 (B.N., App. ii, p. 131)Google Scholar. According to his life (B.N., No. 1,493, p. 92 b, No. 1,485, p. 66) he read the work in Sha-lö (a province of Chinese Turkestan), when on the way to China. What heread consisted of 2,000 slokas. But the extant work consists of about 330 slokas.Some of the oldest catalogues, Nos. 1,604, 1,609, mention that the book is in two vols., but the extant one is in one vol. And the beginning of the work clearly shows that it is only a concluding part of the original. It has only three sections in a chapter, apparently the last, called Paripṛcchā (1) wrong refutation, (2) true refutation (of jāti), and (3) nigraha-sthāna. The work is sometimes ascribed to Vasubandhu, but this is doubtful. Paramārtha commented on it, but the commentary (3 vols.) has been lost.He also translated the Nigraha-sthāna śāstra (1 vol.). the Paripṛcchā-śāstra (1 vol.), and the Cheng-shwo (or lun) tao-li-lun (1 vol.), and he wrote a commentary (5 vols.), the last.”Google Scholar

^{page 1016 note 1} Kai-yan shih-chiao lu, p. 666 Google Scholar, see Appendix. Hsü kao-seng chwan, p. 92 b. B.N., No. 1,493. The texts of both the works are identical.Google Scholar

^{page 1016 note 2} Cf. Appendix.

^{page 1016 note 3} Ui, , Vaiśeṣika Philosophy, p. 84.Google Scholar

^{page 1016 note 4} Fasciculus = chüan, “ a roll ” may not mean a completed work.

^{page 1016 note 5} Kokuyaku dai zōkyō, Ron bu. Dai jū go kan. Inmyō nyūshō ri ron, Kaitai, pp. 52–6.Google Scholar

^{page 1016 note 6} O.F.S., 24, p. 84, f.n. 4. See also the above-mentioned Ui's article “ Introductionto Hetuvidyā ”, p. 56.

^{page 1021 note 1} Professor Ui defines its Sanskrit equivalent in his book, Vaiśeṣika Philosophy, p. 84.Google Scholar

^{page 1021 note 2} Here we note a reference of the catalogue No. 5 (see Appendix, p. 29, f.n. 1 and 2)in the seventh century concerning the dating of Paramārtha's translations of Ju-shih lun, on one side, and Fan-shih lun and To-fu lun on the other, as this is connected with the question of the dating of the translation of Tarka-śāstra into Chinese given by Bunyiu Nanjio. The materials offered in this catalogue are derived from Paramārtha's biography, which does not even mention our treatise. But the catalogue points out that Ju-shih lun and some other treatises were translated between A.D. 548–554 in Cheng-kwang-ssö temple in the reign of the Emperor Wu-ti of Leang. On his wayback to India, after the completion of his work, Paramārtha stopped in the above-mentioned temple in 556 and lived then till 569. It is during this period that hetranslated Fan-chih lun and To-fu lun.

^{page 1022 note 1} Professor Tucci, , in his article, “ Buddhist Logic before ” (JRAS.July, 1929), concludes that Tarka-śāstra denotes it as a generic idea, as some logical work in general, but at the same time he considers that particular Tarka-śāstra (i.e. Ju-shih lun) to be a single text which he had fully transposed into Sanskrit. Moreover he does consider it to be one whole, but does not ascribe its authorship to Vasubandhu. He thinks it was written by some Buddhist author before Dignāga. It seems to me that an analysis of the history of the text, as well asanalysis of its contents, does not confirm Professor Tucci's categorical statement, whodid not pay attention to the philological materials of the catalogues, expounded inthe present work.CrossRefGoogle Scholar

^{page 1022 note 2} According to Professor Ui's calculations (Vaiśeṣika Philosophy, p. 84, f.n. 4), who departed from the total amount of Chinese characters, taking 32 characters forone sloka.Google Scholar

^{page 1023 note 1} Cheng-li men lun-Nyāya mukha, according to Tucci.Cf. G., Tucci, The Nyāya-mukha of Dignāga, 1930, Heidelberg. Otherwise Nyāyadvāra, B.N. 1223, 1224.Google Scholar

^{page 1023 note 2} Jin-ming ju cheng-li lun, B.N. 1216.

^{page 1023 note 3} K'wei-chi (Jap. Ki-ki) (632–682). Cf. Bukkyō jiten, p. 199 Google Scholar; Ui's Vaiśeṣika Philosophy, p. 1, f.n. 4, and Ui's, On the Author of Mah¯āyāna sutrālaṃkāra, p. 220.Google Scholar Shên-t'ai (seventh century); cf. Bukkyo jiten, p. 624.Google Scholar Wen-kwei (seventh century). Hwei-chao (? 714); cf. Ui's, On the Author of Māhāyāna sutrālaṃkāra, p. 221.Google Scholar

^{page 1023 note 4} Shen-t'ai, “ Jin-ming ju cheng-li-men lun shu chi ”; Wen-kwei, “ Jin-ming ju cheng-li lun su; K'wei-chi, ibid.: Supplement to Tripiṭaka, i, vol. lxxxvi, fasc. 4. Hwei-chao, ” Jin-ming ju cheng-li lun i tuan “; Hwei-chao, ” Jin-ming ju cheng-lilun suan-yao ”; ibid., vol. lxxxvi, fasc. 5.

^{page 1024 note 1} B.N. 1224, translated by Hsüen-tsang and B.N. 1223, translated by I-tsing. G. Tucci translates this passage: “ The proposition and the other terms are called the proof (sadhana). Here is called ” proposition ” only that particular argument that wewant to prove in accordance with our own opinion. It must be such as no argumentcontradictory (to it) can exclude (it). ”

^{page 1024 note 2} Cf. G. Tucci: ” ” The proposition, etc“ This means that through the formulation of a proposition, a reason, and an example, an argument, which has not yet been understood by author (man), is made evident to him. That many terms represent the sādhana, syllogism, was already asserted by Vasubandhu in his Vādavidhi, etc.”

^{page 1024 note 3} Tucci translates this passage: ” Moreover this erroneous formulation of the jatis has been already settled in the main by myself in my commentary upon the Vādavidhi, etc.”, and adds that for him “ the translation is doubtful ”. I think Professor Tucci is right in his confession and the translation of this passage is wrong.First of all the text has no indication to “ my commentary ”. Japanese edition of Nyāyamukha gives us instead of the character “ shu ” the character “ tang ”, which isquite possible. Thus the word “ p“eng tang ” as well as “ p“eng shu ” can be translatedas “ associate in philosophy ” pointing out Vasubandhu and not Dignāga andthe word “ wo ” is possessive pronoun—“ my ”. The same interpretation I have found in the Japanese article of Professor Ui on the Nyāya-mukha in vol. v, p. 472, and p. 692 of his Indo tetsugaku kenkyū.Google Scholar

^{page 1024 note 4} Suppl. to Tripiṭaka, i, vol. lxxxvi, fasc. 4, p. 314 b.Google Scholar

^{page 1025 note 1} Sugiura, in his Hindu Logic as preserved in China and Japan, p. 32, s a y s: “ … when Hiuen-tsang was in India he saw throe books on logic ascribed to Seish (Vasubandhu), namely Ronki, Ronshiki, and Ronshin” (f.n. 1, Murakami's Immyōjensho, 129;Dinna also speaks of this). If so, it must have been possible for Hsüen-tsang to passa certain information regarding these treatises to his pupils. Besides they must have known Pramāṇasamuccaya, and could derive some facts from this latter.Google Scholar

^{page 1026 note 1} Suppl. to Tripiṭaka, i, vol. lxxxvi, fasc. 4, p. 337 b. on the true relation between Vāda-vidhi and Vāda-vidhāna, cf. Prof. Vostrikov's work quoted above. With regard to these two treatises an interesting passage from Fen-i ming-i chi (B.N. 1640), composed in the twelfth century, should be mentioned. It runs that Vasubandhu, convinced of Manoratha's wrong deductions, wrote Lun-kwei and Lun-shih which are said to have been called Paramārthasaptati and were directed against the Sānkhya system. But, as we know, Paramārthasaptati is Vasubandhu's special work, written during his old age and devoted to a refutation of Sānkhya theories. The confusion of facts in this Chinese version is evident.Google Scholar

^{page 1026 note 2} Suppl. to Tripiṭaka, i. vol. lxxxvi, fasc. 4, p. 335 a.Google Scholar

^{page 1026 note 3} ibid., p. 350a.

^{page 1026 note 4} ibid., p. 347a. See also fasc. 5, p. 426.

^{page 1026 note 5} Suppl. to Tripiṭaka, i, fasc. 4, p. 352 a.Google Scholar

^{page 1026 note 6} ibid., p. 349b. Cf. also p. 370a.

^{page 1026 note 6} Cf. also p. 350a.

^{page 1027 note 1} ibid., p. 352b. See also p. 353a.

^{page 1027 note 2} Suppl. to Tripitaka, i, v. Ixxxvi, fasc. 5, p. 404a, ibid., p. 409a.

^{page 1027 note 3} Cf. Suppl, . to Tripitaka, i, vol. Ixxxvi, fasc. 5, p. 404 a.Google Scholar

^{page 1027 note 4} ibid., p. 409a; ibid., p. 418a.

^{page 1027 note 5} Suppl, . to Tripiṭaka, i, vol. Ixxxvi, fasc. 5, p. 405 a.Google Scholar

^{page 1027 note 6} ibid., fasc. 4, p. 331a.

^{page 1027 note 7} Suppl, . to Tripiṭaka, i, vol. Ixxxvi, fasc. 4, p. 335 a.Google Scholar

^{page 1028 note 1} ibid., p. 418b.

^{page 1028 note 2} Tucci, G. in his article, “ Buddhist Logic before ” p. 479 Google Scholar, only summarized the data given by K'wei-chi and Shen-ti'ai when saying: “ … We gatherboth from K“wei chi (ch. iii) and Shên T“ai (ch. ii) that the theory of the vipakṣa was known to the ancient masters, who held two different opinions about it, which were not accepted by Some thought that the vipakṣa is that which excludes the sapakṣa, as well as the pakṣa, so in the syllogism “ sound is non-eternal, because it isa product, like a pot,” the vipaksa “ ether ” excludes the contrary of the non-eternalas well as of the pot. On the other hand other logicians said that the vipakṣa is everything except the non-eternal while for , as is known, vipakṣa is yatra pakṣo na vidyate ”.

^{page 1028 note 3} Cf. p. 15, f.n. 1.

^{page 1028 note 4} Suppl, . to Tripitaka, vol. lxxxvi, fasc. 5, p. 337 b.Google Scholar

^{page 1028 note 5} Suppl, . to Tripipaka, vol. lxxxvi, fasc. 5, pp. 423 a, 406a.Google Scholar

^{page 1029 note 1} p. 32.

^{page 1029 note 2} Tucci, G. in his article “ Buddhist Logic before ” p. 451, points outthat the only source of Suali, Vidyabhusana, Keith, etc. was Sugiura, and characterizeshim as follows: “ But being himself absolutely without knowledge of orthodox nyāya and of Sanscrit, he is in his statements and in his translations very often misleading… ”Google Scholar

^{page 1029 note 3} p. 129.

^{page 1029 note 4} We find the same statement in Hsie Meng's book, Outline of Buddhism (Fo-hsio ta kang, p. 33).Google Scholar

^{page 1030 note 1} JRAS., 1928, p. 383.Google Scholar

^{page 1030 note 2} Cf. also the Japanese article of Professor Ui, H., Indo tetsugaku kenkyú, vol. v, p. 547.Google Scholar

^{page 1030 note 3} IHQ., vol. iv, p. 636.Google Scholar

^{page 1031 note 1} JRAS., 1929, p. 451.Google Scholar

^{page 1032 note 1} Kokuyaku dai zōkyō. Ron bu. Daijū go kan. Inmyō nyū shō ri ron; pp. 52–6.Google Scholar

^{page 1032 note 2} Ui analyses in detail the difference, which existed, according to Hsüen-tsang's school, between the old and the new logic. K'wei-chi says that “ Boddhisattva Asanga took all the five parts of the syllogism, i.e. Pratiĵñā, Hetu, Udāharaṇa, Upanaya, and Nigamana, for Sādhana the three parts of the syllogism. But, according to Dign¯aga's new logic, only Hetu and Udāharaṇa can be considered to be Sādhana, and Pratijñā is Sādhya. Thus it is wrong to take all the three or the five parts of the syllogism for Sādhana.” Wen-kwei and Shên-t'ai treat it in the same way. It is in this that Hsüen-tsang's school saw the difference between the old and the new logic and note d that the old logic was not sufficiently worked out. However, continues Ui, Kwei-chi is wrong in his treatment of this point. “ Sādhana ” has two different meanings in logic: first, Sadhana as opposed to Dūṣaṇa, in this case it wholly refers to the three and the five parts of the logical formula. But, if we speak of Sādhana as the opposite of Sādhya, it is Sādhya that figures as Pratijñā and Sādhana as Hetu and Udāharṇa. Therefore Asañga and Vasubhandu regarded Sādhana as an antithesis of Dṣaṇa and accepted all the three or five parts of the logical formula. Dignāga opposes Sādhana to Sādhya, therefore he took as Sādhana Hetu and Udāharaṇa only, and referred Pratijñā to Sādhya. K“wei-chi made a confusion of the two ideas and therefore gave them a wrong interpretation.“

^{page 1033 note 1} Yin san hsiang.

^{page 1033 note 2} Fan-chih nan p“in.

^{page 1033 note 3} Tao li.

^{page 1033 note 4} Ti wei nan.

^{page 1033 note 5} Ti fu ch“u.

^{page 1034 note 1} Fo hsing lun. B.N. 1220.

^{page 1034 note 2} (a) Shih kên bên fa; (b) t“ung-lei so shê; (c) i-lei, hsiang li.

^{page 1034 note 3} Chu ching lu.

^{page 1035 note 1} Cf. Logical Works of Vasubandhu.

^{page 1035 note 2} I have not mentioned two articles of great importance, included by H., Ui, in the vol. v of his Studies of Hindu Philosophy Google Scholar, in Japanese. One of these articles is dedicated to ” Pre-Dignāga's logic ” and the other one to the analysis of “ Nyāyamukha”. Besides we have a new work in Chinese dedicated to the same question and written by Professor Hsü Ti-shan (cf. Yenching, , Journal of Chinese Studies, No. 9,1931)Google Scholar, The Buddhist Logical Treatises Madhyamika and Yogacāra Schools before ,” pp. 1828–1868 of this work are specially dedicated to Vasubandhu and his logical works including Ju-shih lun, with re-edited Chinese text of our treatise.Google Scholar

^{page 1035 note 3} Bagchi, Prabodh Chandra in Le Canon Bouddhique en Chine (Paris, 1927)Google Scholar, gives a list of our treatises and the catalogues that contain them. Nevertheless, these data, notwithstanding the general value of the work as such, are incomplete and do not give us materials whence we could draw conclusions as to the history of our texts. Thus with regard to Ju-shih lun we read (p. 423): “ Jou-cheu louen un chapitre LK (73b, 6); NL (71b, 15); TK (85b, 15); KL (53b, 9) dit le colophon de l'ouvrage porte le titre: Jou-cheu louen fan tehe nan p“in. L'ouvrage existe. Nanjio 1252. Tarka-çāstra.“ We see that Prabodh Ch. Bagchi in his enumeration mentions our respective catalogues Nos. 2, 4, 5, and 7-the rest are not mentioned. The same may be said with regard to the three other works (p. 429): “ Fan tchen louen, en un chapitre. LK (65b, 8); NL (77b); TK (85b); KL (61a, 1) fait remarquer qu'il n'est pas certain si l'ouvrage forme une partie du Jou chen louen.” “ To fou louen, en unchapitre. LK (65b, 8); NL (77b); TK (85b); KL (61a, 1).” “ Tcheng chouo tao li louen, en un chapitre. LK (65b, 9); NL (77b); TK (86a); KL (61a, 1). The abovementionedcatalogues are nothing but our respective catalogues Nos. 2. 4, 5, and 7; the others are not mentioned by Prabodh Ch. Bagchi.