Page #1
--------------------------------------------------------------------------
________________
Bāhyārthasiddhikārikā of Subhagupta
Masaaki Hattori
In his History of Indian Logic, S. C. Vidyabhusana gave the Sanskệt name for Dge-sruns, the author of the Sarvajña-siddhi-kārika and four other treatises preserved in the Tibetan Tripitaka, as Kalyānagupta ; while he identfied Subhagupta, who is called a follower of Dharmakirti (värttikānusärin) by Haribhadrasūri, with śubhākaragupta who flourished late in the 11th century. This theory, however, was criticised by B. Bhattacharya in the the foreword of G. O. S. edition of the Tattva-samgraha. He set forth the opinion that Haribhadrasūri must have referred to Bhadanta śubhagupta, whose views were mentioned several times in the Bahir-artha-parikșa of the Tattvasamgraha, and doubted whether śubhākaragupta, who was a Tāntrika, ever wrote any work on logic wherein he closely followed Dharmakirti. Y. Miyasaka has recently disclosed, in his article on Subhagupta's theory of the word, the view that Dge-sruns must be the Tibetan name for śubhagupta. His view is grounded on 1) that “dge' is considered as an equivalent for 'śubha', 2) that śubhagupta's theory of conventional designation' (sanketa) alluded to by Haribhadrasūri is similar to Dge-sruńs' theory expounded in his Śruti-parikșa-kārika and 3) that śubhākaragupta cannot be referred to by Haribhadrasūri, who, according to the recognized theory set up by H. Jacobi, is assigned to ca. 750 A. D. The writer of the present paper agrees with Y. Miyaşaka in regarding Dge-sruns and Subhagupta as being the same person, for this view can be verified by identifying quotations from Bhadanta śubhagupta in the Bahir-artha-parikșa of the Tattva-sangraha (-pañjika) with verses in the Bahyârtha-siddhi-kārika of Dge-sruns. The writer has prepared the following list of identification. (Skt.: Tattvasamgraha, Tib.:
(1)
(1) śubhagupta no Kotoba-ron, Chizan Gakuho, No. 22, pp. 54-55.
- 400 -
Page #2
--------------------------------------------------------------------------
________________
(1)
(2)
(3)
.
(4)
(10) Bāhyārthasiddhikārikā of Subhagupta (M. Hattori)
Bahyarthasiddhi-k., Pek. Ed., Mdo-hgrel, CXII): 1) dhi-mātratvene saṁsādhaye...... (p. 582, 1-2) =k. 25 (201 a, 3): blo tsam ñid du...... 2) tulyâpara-kşaņotpādād...... (p. 552, 6–7), cf. k. 31 (201 a, 8): hdra bahi skad cig...... 3) pratyeka-paramâņūnām..... (p. 551, 21–22) =k. 40 (201 b, 6): rdul phran rnams...... 4) Bhadanta-Subhaguptas tv āha......kevalam anava eva......na savayavatvam.
(p. 556, 16 ft), cf. k. 42 (201 b, 7): de phyir phyogs chaḥi...... . 5) nânyo 'sti grāhako...... (p. 569, 14—15) =k. 62 (202 b, 4–5): śes pa las gșan...... 6) punaḥ sa ev'āha-atha saha-sabda eka-kāla vivakṣayā....... (p. 567, 25ft), cf. k. 64
(202 b, 6): dus gcig tu ni...... 7) saha-sabdaś ca loke......(p. 567, 17–18) =k. 67 (202 b, 8): þjig rten ḥdi na...... 8) yadā ca sarva-prāṇabhștāṁ...... (p. 568, 1–2), cf. k. 69 (203 a, 1–2): gal te thams
cad...... 9) kiṁ cânyôpalambha...... (p. 568, 2–7), cf. k. 70 (203 a, 2–3) gșan kyis dmigs pa...... 10) pūrvikaiva tu...... (p. 569, 15—16) =k. 77 (203 a, 6–7): gan gis lhan cig....... 11) syād etat-na hi mukhyato yādịśam...... (p. 570,18 f), cf. k. 80–81 (203 a, 8-b, 2):
ses pa myon bar...... 12) sākāra-jñāna-pakse...... (p. 570, 23–24) =k. 83 (203 b, 3): ses pa rnam bcas...... 13) katham tad-grāhakam...... (p. 561, 11–12) =k. 85 (203 b, 4—5): de ḥdzin byed de...... 14) sākāram tan nirākāram...... (p. 573, 7–8) =k. 91 (203 b, 8): rnam pa bcas sam...... 15) sa hy āha-yathaiya bhavatām vijñāna-vādināṁ...... (p. 572, 10—12), cf. k. 97
(204 a, 4—5): ji ltar khyod kyi...... 16) atra Bhadanta-Subhaguptaḥ pramāņayati-yo jñāna-ākāraḥ sa samvāditve sati ...... (p. 574, 20-22), cf. k. 7 (200 a, 5) ses pa mi bşlu......
The Bahyârtha-siddhi-karika consists of 183 couplets. Judging from the fact that prose fragments commencing with “Bhadanta Subhaguptas tv aha...” etc. are often quoted by Kamalaśīla, it appears that there was originally its prose commentary, which, however, was not translated into Tibetan. It is quite obvious that this treatise is intended for refuting the Vijñānavāda theory and proving the objective reality of external object. It
(1) Text reads tvena (2) For jñanāc cakṣuşair read jnanän näkşa-dhir. (3) For loke 'nyon ('syan) naivanena (-anyena) vina read loke 'smin naivanyena
vina. (4) Text reads sattvar prāna-, perhaps incorrectly. (5) For Prajñānam read sajñānam.
- 399
Page #3
--------------------------------------------------------------------------
________________
(6)
Bāhyārthasiddhikārikā of Subhagupta (M. Hattori)
(11) starts with stating the Vijñānavāda doctrine that the object of eyes and other senses does not exist in the external world independently of the consciousness, just as the object seen in a dream or the second moon seen by those who are afflicted with eye-disease are not really existent, and this doctrine is criticized through various arguments. It is noteworthy that Subhagupta evidently directs his attack toward the views of Vasubandhu and Dignāga. In the Vimśatika-vijñaptimātrata-siddhi, realists assail the Vijñānavādins, arguing that if the object were not real, then our purposive action tending toward it would not be fulfiled, as for instance we cannot be satisfied with meals and drinks seen in a dream. In vindication of the Vijñānavāda theory, Vasubandhu, citing an instance of nocturnal emission which is caused in spite of the absence of real woman, asserts that an action toward the object is possible even though the object is unreal. This thought of
The latter Vasubandhu is retorted by Subhagupta. He is of the opinion that the reality of an external object is proved not through the fulfilment of an action tending toward it, but through the ascertainment of the object (arthaniścaya). Man can act upon an illusory cognition, but he can not get then the ascertainment of the object. It might be argued that a man would not discharge semen during a dream if he had not ascertained that he is with a woman. In anticipation of this argument, Subhagupta states that the seminal emisson is the result of the carnal desire which may be aroused whether there is a real object or not. If it should be maintained that an object is unreal, then the object seen in a dream would not be distinguished from the object seen through the actual sense, both being existent only in the consciousness. But the fact is that a man does not find the child whom he got.in a dream, while, if he really got a child, he can ascertain its objective reality. The above arguments are recognized as being levelled against Vasubandhu's theory.
śubhagupta holds the view that the external object is the union of atoms.
(7)
(6) Cf. Vimšatika, k. 4 & Comm. (7) Cf. Bahyârtha-siddhi-k., kk. 124-133 (205 a, 6-b, 4).
- 398 -
Page #4
--------------------------------------------------------------------------
________________
(8)
(9)
(10)
(11)
(12) Bāhyārthasiddhikārikā of Subhagupta (M. Hattori) As an atom does not exist independently by itself, a single atom can never be reflected in the image. We recognize the union of many homogeneous atoms as a single substance, just as we form the idea of continuance on really a series of similar moments. In the course of his discussion on atoms, śubhagupta refers to the following view :- the presentation of a certain form is not produced from that (=external object), because (the external object which is supposed to have a visible form) is not a real substance. This is exactly the Dignāga's view set forth in his criticism of the Vasubandhu's definition of perception as being produced by that external object (tato’rthad vijñānam pratyakşam). A visible object must be the aggregation of many atoms, but, according to Dignāga, the aggregation is able to be apprehended by unifying through the conceptual construction a multiplicity of perception produced by each of its constituents. Hence, being a product of the conceptual construction, the aggregation is not a real substance and is incapable of producing the mental reflex of its own form. śubhagupta contradicts the above-stated view of Dignāga, asserting that the aggregation of atoms is to be regarded as a simple substance existing objectively. His theory may well be identified with the realist doctrine ascribed to Vāgbhata in the Alambana-parikșa-țika of Vinītadeva.
As stated above, śubhagupta is called by Haribhadrasūri a follower of Dharmakīrti. However, the following is mentioned evidently in opposition to Dharmakīrti's theory: - there is no apprehender other than the cognition, and the cognition through senses is not produced without an object; it is on this ground that the blue (object) and its cognition are considered as being apprehended together (saha-samvitti) and not on the ground of their
(12)
(13)
(8) ibid., k. 40 (201 b, 6). (9) ibid., kk. 31-32 (201 a, 8-b, 1). (10) ibid., k. 34 (201 b, 2). (11) Pramanasamuccaya, I, k. 16 a: yad-ābhāsā na sā tasmāt. cf. M. Hattori,
Dignāga no Chishiki-ron, Tetsugaku Kenkgū, No. 463, Chap. IV, note 12. (12) Bahyartha-siddhi-k., kk. 35-36 (201 b, 2-4). (13) Cf. S. Yamaguchi, Seshin Yuishiki no Genten Kaimei p. 451.
- 397
Page #5
--------------------------------------------------------------------------
________________
(14)
(15)
(16)
(17)
Bāhyārthasiddhikārikā of Subhagupta (M. Hattori)
(13) non-difference. In the Pramaņa-viniscaya, Dharmakirti propounds the theory that (proposition :) the blue (object) and its cognition are not different, (reason :) because they are invariably found together (sahộpalambha-niyama). Subhagupta examines the meaning of the term 'together' (saha). If it is taken as implying simultaneity (eka-kala), then the reason is inconclusive, for the Buddha's intellect is not without a difference from the mind of living beings, although the former is operative simultaneously with the latter, or the mind is considered as being different from other subordinate mental activities which operate simultaneously with it. If, on the other hand, the term 'together' is to mean the sameness, then the reason is contradictory, because it is generally admitted that the term 'together' presupposes difference' (anya). Moreover, we should recognize the fact that the object which is cognized by one person is cognized also by another. Accordingly, if the term 'together' is interpreted as 'one' (eka) in the sense of 'by one person', then the reason is inadmissible (asiddha). Even when 'one' is taken in the sense of 'as one', the reason contradicts the fact that the omniscience who has insight into the minds of all living beings does not regard them as one'.
We have to omit, in this short paper, to enter further into details of the arguments developed in the Bāhyârtha-siddhi-karika. It, however, should not be overlooked that śubhagupta's interpretation of sahộpalambha-niyama' as the necessary relation of the cognition with the external object is undoubtedly realistic, while the above-mentioned theory of Dharmakīrti is probative of his being a Vijñānavādin. We learn from this fact that Subhagupta was not a faithful successor of Dharmakīrti's thought. It
(18) seem's likely that śubhagupta was a teacher of Dharmottara, for the latter
(14) Bahyârtha-siddhi-k., k. 62 (202 b, 4–5). (15) Pek. Ed. Mdo-hgrel XCV, 263 b, 4–5: Ihan cig dmigs pa nes pahi phyir H sno
(=snon) dan de blo gşan ma yin/ cf. Sarvadarśanasangraha, Govt. Or. Ser.,
p. 32: sahô palambha.niyamad abhedo nila tad-dhiyoh/ (16) Bahyartha-siddhi-k., k. 64 (202 b, 6). (17) ibid., kk. 67-69 (202 b, 8-203 a, 2). (18) Cf. Vidyabhusana, Hist. of Ind. Logic, p. 328.
- 396 -
Page #6
--------------------------------------------------------------------------
________________ (14) Bahyarthasiddhikarika of Subhagupta (M. Hattori) and his successors are known to have interpreted Dharmakirti's theory from the Sautrantika viewpoint; but no conclusive evidence in support of the guru-sisya relation between the two has been obtained so far. All we can say quite safely is that Subhagupta flourished after Dharmakirti and before Santaraksita and Haribhadrasuri: that is to say, his date can be fixed at ca. 650-750 A. D. While, this paper was being printed, I was presented dy Dr. E. Frauwallner with an article published in WZKSOA, Bd. I, a footnote of which urged me to refer to his another article "Dignaga und anderes" published in the Festschrift Moriz Winternitz. He proved therein already the identy of Subhagupta with Dge-sruns, basing upon the same sources that I made reference to. I regret for my inattention to this article. However, the list of identification and brief accounts of Subhagupta's arguments, which were left unpublished by Dr. Frauwallner, may not be of no value. -395