Book Reviews

Author(s): J W De Jong
Publisher: J W De Jong
Catalog link: https://jainqq.org/explore/269702/1

Page #1

BOOK REVIEWS
THE ESSENCE OF METAPHYSICS. Abhidharmahsdaya. Translated and annotated by Charles Willemen. Publications de l'Institut Belge des Hautes Études Bouddhiques, Série "Études et textes" No. 4. Bruxelles, 1975. XXVII, 366 pp. et 25 pp. de texte chinois. LE COEUR DE LA LOI SUPRÊME. Traité de Fa-cheng. Abhidharmahydayaśāstra de Dharmasri. Traduit et annoté par I, Armelin. Paris, Librairie Orientaliste Paul Geuthner, 1978. 388 pp.
THE PUBLICATION of two translations of the Abhidharmahşdaya (T. no. 1550) testifies to the growing interest in Abhidharma studies in recent years. In 1977 José van den Broeck published a French translation of Ghoşaka's Amộtarasa (T. no. 1553) and Marcel van Velthem a French translation of Skandhila's Abhidharmavatāraśāstra (T. no. 1554).1 Mrs. Armelin's translation was already finished in 1963. Since then she has added a long introduction (pp. 3-48) in which she makes much use of Yamada Ryūjō's Daijā bukkyo seiritsuron josetsu (Kyoto, 1959). In her preface she announces a translation of Dharmatrāta's Samyukta-abhidharmasāra (T. no. 1552) and a book on the importance of the Abhidharma in Indian thought (L'importance de l'Abhidharma dans la pensée indienne).
Willemen's introduction (pp. V-XXVII) is divided into two parts. The first deals with the three texts called (Samyukta) abhidharmahsdaya (T. nos. 1550, 1551 and 1552), their authors and the Chinese translations (pp. V-XXVII). The name of the author of I (T. no. 1550) is Fa-shêng. Western scholars assume that his Sanskrit name is Dharmaśrī but several Japanese scholars prefer Dharmaśreșthin. In his L'aide-mémoire de la vraie loi (Paris, 1949, p. 51, n. I) Lin Li-kouang pointed out that Dharmatrāta in his Ch'u-yao ching (T. no. 212, p. 643a) quotes the opinion of Tan-mo-shih-li on the importance of the
1 La saveur de l'immortel (A-p'i-t'an Kan Lu Wei Lun). La version chinoise de l'Amộtarasa de Ghosaka (T.1553). Traduite et annotée par José van den Broeck. Louvain-la-Neuve, 1977. Le traité de la descente dans la profonde loi (Abhidharmävatāraśāstra) de l'Arhat Skandhila. Traduit et annoté par Marcel van Velthem. Louvain-la-Neuve, 1977. Reviews of both works will be published in the Toung Pao.
151

Page #2

THE EASTERN BUDDHIST
practice of kāyānusmrti. According to Lin this transcription confirms the reconstruction Dharmaśrī proposed by Pelliot in 1930 (JA, 1930, II, p. 267 ff.). Willemen adds that it is most probable that this T'an-mo-shih-li is the author of the Abhidharmahşdaya. It is necessary to point out that the kāyānusmyti is not mentioned in the Abhidharmahụdaya (cf. Lin, p. 330). Dharmatrāta quotes also two verses by Dharmasri (Tan-mo-shih-li) which obviously belong to a different work (T. no. 212, p. 626a 17-21). In both instances Dharmasri is called (sthavira, ārya or bhadanta). It is therefore doubtful that the Dharmasri quoted by Dharmatrāta is the author of the Abhidharmahşdaya. Willemen accepts Lin Li-kouang's theory according to which the author of the Ch'u-yaoching is identical with the Dharmatrāta or Bhadanta mentioned in the Mahāvibhāṣā. If Dharmatrāta in his Ch'u-yao ching quotes Dharmaśrī, then he must have lived before the masters of the Mahāvibhāşā (second century A.D. according to Willemen). Willemen believes that it is possible to go back further because he accepts Frauwallner's theory that the Abhidharmahsdaya is older than the Jñanaprasthāna. Frauwallner mentions in this connection Tao-yen's preface to Buddhavarman's translation of the Abhidharmavibhāṣā (T. no. 1546, p. Ibu-12). However, it does not seem likely that Tao-yen's testimony is more reliable than the other often conflicting statements on the date of Dharmaśrī in Chinese sources (cf. Lin, p. 51 and Willemen, pp. VII-VIII). As to the relation of the Abhidharmahsdaya to the Jñānaprasthāna and the Mahāvibhāşā the opinions of scholars vary greatly. Willemen, who knows well the works of Japanese scholars, quotes those of Fukuhara (cf. p. VIII), Yamada and Sakurabe (p. XXII; read Sakurabe H., instead of Sakurabe T.). It is interesting to note that Willemen and Mrs. Armelin attribute different opinions. to Yamada. Willemen writes: "R. Yamada thinks that the Abhidharmahsdaya is a little earlier than the Jñānaprasthāna, but that there is no direct relation between the two texts, and he goes on to say that the Abhidharmahsdaya was probably written at the same time as the Mahāvibhāṣā.” Mrs. Armelin writes: “Selon le Professeur Yamada, le Hşdayaśāstra composé de dix sections (daśavarga) est postérieur au Jñānaprasthāna parce que son style est plus prolixe que celui des Aşțaskandha" (p. 12). Willemen refers to p. 113 of Yamada's book. Yamada indicates here as his provisional conclusion that the Abhidharmahrdaya is later than the Jñānaprasthāna and roughly contemporary with the Mahavibhāṣā. However, on p. 428 of his book Yamada declares positively that the Abhidharmahsdaya is earlier than the Mahāvibhāṣā.
In his introduction Willemen carefully examines the problems connected with the Chinese translations of the three Abhidharmahsdaya texts. He has also consulted a Tun-huang manuscript, probably written in the beginning of the fifth century, of Samghadeva's translation of Dharmasri's Abhidharmahsdaya, and has been able to show that this text contains exactly 250
152

Page #3

BOOK REVIEWS
stanzas. In the Taishō text several stanzas have not been distinguished as such. Willemen reproduces the Taisho text adding a continuous numbering of the 250 stanzas. Moreover, in a concordance he indicates the corresponding stanzas in the works by Upaśānta and Dharmatrāta (pp. 304–313). According to him Upaśānta was a bahirdeśaka, a master of Gāndhāra, and lived probably in the third century. Dharmatrāta lived also in Gāndhāra but he relied heavily on the vaibhāșika views as expressed in the Mahāvibhāṣā. Willemen lists a number of doctrinal differences between Dharmaśrī and Dharmatrāta (pp. XXIII-XXIV). He concludes that Dharmasri probably represents the early bahirdeśakas. Upaśānta almost always agrees with Dharmaśrī but his work shows traces of vaibhāșika influence. Finally Willemen shows that Dharmasri's Abhidharmahsdaya and Dharmatrāta's Samyuktābhidharmahsdaya were very important in China until the publication of Hsüan-tsang's translations of the Kośa and the Mahāvibhāṣā.
Mrs. Armelin's introduction deals with the following topics: 1. L'Abhidharma des Sarvāstivādin, p. 3. 11. La date de composition de l'Abhidharmahỉdayaśāstra, . p. 7. III. Plan du Hşdayaśāstra, p. 12. Iv. Comparaison des chapitres des deux
traités, p. 15. v. Le titre de l'ouvrage de Dharmaśrī, p. 19. VI. Positions doctrinales du Hşdayaśāstra, p. 20. VII. Dharmaśrī et l'école des Sarvāstivādin, p. 34. VIII. Dharmaśrī, arhat ou bodhisattva? p. 45. Résumé, p. 47.
Whereas Willemen in his introduction pays much attention to the history of the Chinese translations, Mrs. Armelin is more interested in the doctrinal problems. As mentioned above, she has made much use of Yamada's work but does not mention at all the important abhidharma studies by Erich Frauwallner (WZKS 7, 8, 15, 16, 17). As to the relation between the Jñānaprasthāna and Dharmasri's Abhidharmahşdaya she disagrees with Yamada and believes that Dharmasri was probably a contemporary of the author of the Jñānaprasthāna. Both Willemen and Mrs. Armelin prefer the title Abhidharmahsdaya to that Abhidharmasāra used by de La Vallée Poussin and other scholars. Willemen does not explain his preference. According to Mrs. Armelin hrdaya is to be preferred because Dharmasri wanted to reveal the hşdaya, the most important part of Buddha's doctrine. He did not want to extract the sāra, the essence, from the abhidharma doctrines. Mrs. Armelin adds that the word abhidharmasāra would imply that Dharmasri tried to extract the essence from the Abhidharma and especially from the, Şațpādaśāstra which, from a doctrinal point, is not very likely. Her arguments are not entirely convincing. It is quite well possible that Dharmasri in his relatively short work wanted to present the essence of the abhidharma doctrines known to him.
Mrs. Armelin compares the way of salvation as presented by Dharmaśrī with the brahmanical way of salvation as found in the Bșhadāraṇyakopanişad. She writes: "Dharmasri ne cite pas les théories des Upanişad. Toutefois, son
153

Page #4

THE EASTERN BUDDHIST
hmanicalfficult to discoverpose Buddhist Loadāranyaka-upa
oeuvre trahit ses connaissances, et peut-être même ses origines brahmaniques" (p. 27). On p. 29 she remarks: “Dharmaśri confronte l'enseignement du Buddha à celui des penseurs brahmaniques partisans de la Bșhadāranyaka-upanişad." It is of course always possible to oppose Buddhist doctrines to Brahmanical doctrines but it is difficult to discover in Dharmasri's work such a specific reaction to Brahmanical doctrines as maintained by Mrs. Armelin. However, the comparisons made by Mrs. Armelin are instructive insofar as they help to bring out more clearly the nature of the Buddhist doctrines taught by Dharmasri.
In a chapter on Dharmasri and the school of the Sarvāstivādins (pp. 34-46) Mrs. Armelin examines the doctrinal systems of Dharmaśrī, Upaśānta, and Dharmatrāta. She remarks that although Vasubandhu has been influenced by the Sautrāntikas and the Pāścātyas, his doctrinal position is closer to that of the Kashmirians than to that of other Abhidharmikas. On several points he rectifies opinions expressed by Dharmasri and defends vaibhāșika doctrines. As example, she mentions that according to Dharmaśrī the kāyavijñapti is a movement (gati) but according to the Vaibhāşikas a figure (samsthāna). She adds that this is a doctrine of the Vätsiputriyas but without drawing attention to the fact that this is not said in the Kośabhāşya but in Yaśomitra's Vyākhyā (p. 345. 16).
With regard to the different Abhidharma schools of the Sarvāstivādins much is still obscure. Of the great mass of texts only very few have yet been translated into Western languages and it is at present not possible to arrive at any definitive results. In a long and learned introduction to his translation of Ghoşaka's Amptarasa José van den Broeck has pointed out the contradictions found in the sources. It will be a very delicate task to distinguish the different Sarvāstivāda schools to which the texts refer. The three works by Dharmaśrī, Upaśānta and Dharmatrāta and Vasubandhu's Kośa constitute an important stream in the development of abhidharma doctrines. The translation of Dharmasri's work by Willemen and Mrs. Armelin makes it possible to compare in detail his work with that of Vasubandhu. Mrs. Armelin has already announced a translation of Dharmatrāta's work. It is to be hoped that Upaśānta's work will also be translated in order to make it possible to compare in detail the three Abhidharmahsdayas and Vasubandhu's Kosa
The Abhidharmahsdaya is a difficult text to translate. The verses especially create problems and a correct interpretation is often only possible with the help of the prose commentary and by comparing the renderings of the verses in the works by Upaśānta and Dharmatrāta. Willemen stresses his great debt to the Japanese version by Watanabe, Mizuno and Oishi in the Kokuyaku Issaikyo, Bidon-bu XXXI (Tokyo, 1932). It is a pity that Mrs. Armelin does not seem to have made any use of this excellent translation which contains also many helpful notes. Willemen seems to have made more use also of the other two Abhidharmahsdaya texts than Mrs. Armelin. He translates technical terms,
154

Page #5

BOOK REVIEWS
and gives the Sanskrit equivalents only in his notes. Both in Mrs. Armelin's book and in that of Willemen the notes are printed after the translation, which makes the study of their works more difficult. Mrs. Armelin often refers to de La Vallée Poussin's translation of the Kośa. This work is rarely mentioned in Willemen's notes although it is obvious that he has studied it very carefully. De La Vallée Poussin's translation of the Kośa is accompanied by a detailed commentary in which he quotes extensively from Yasomitra's Vyäkhyä, the only Sanskrit text available to him. Since the publication of the Sanskrit text of the Abhidharmakośabhāṣya in 1967 it has become even more necessary to refer to the Kosa in translating abhidharma texts, and especially a text such as Dharmaśrī's Abhidharmahṛdaya which has been an important source for Vasubandhu, if not directly, at least indirectly through the later adaptations by Upaśānta and Dharmaträta. If for technical reasons it is not possible to add extensive notes at the bottom of the page, the best solution would probably be to publish the translation on the left-hand page and to reserve the opposite page for a detailed commentary.
Two brief passages of Dharmasri's Abhidharmahṛdaya were translated by de La Vallée Poussin in the introductory volume of his translation of the Kosa (Paris-Louvain, 1931), pp. LXV and LXVI-LXVII. The first deals with the three obstacles (dvarana): Bhagavat dit qu'il y a trois avaramas, karman, klesa et vipäka. Quelle est leur définition?
"Les actes d'anantarya qui sont sans remède, les passions développées, l'acte mauvais senti dans les mauvaises destinées, ce sont là les avaraṇas." Ces trois font obstacle au Dharma; ils empêchent de prendre les dharmas d'Arya (T. 1550, p. 815a23-27). Willemen translates this passage as follows:
Question: What about the characteristics of the three obstructions, as explained by the World-Honoured One: the obstruction of action, the obstruction of affliction, and the obstruction of retribution?
Answer: (63) Let it also be known that the immediate and irredeemable actions, the afflictions one produces on a large scale, and the unwholesome retributions experienced in the woeful courses, are obstructions.
With these three dharmas, the so-called obstructions, one surely will not experience the dharmas of the noble (Willemen, p. 46).
Willemen's translation is more literal than that of de La Vallée Poussin. However, de La Vallée Poussin's interpretation of E. ZIE is to be preferred: "These three obstacles to the dharma make it impossible to receive the dharmas of the noble."2 It would have been helpful if Willemen
2 The Kokuyaku issaikyō translation has: "This obstacle to the three dharmas. . ."(!): kono sanbō no shōge wa kanarazu shōbō o jusezu.
155

Page #6

THE EASTERN BUDDHIST
had quoted the first half of Kośa iv. 96: ānantaryāṇi karmāņi tivrakleso 'tha durgatiḥ, and had referred to de La Vallée Poussin, Kośa, iv, p. 201 ff.
Mrs. Armelin's translation is as follows: Question: "Le Bienheureux a parlé de trois empêchements (nivaraņa): l'empêchement des actes (karmāvaraņa), l'empêchement des passions (kleśāvarana) et l'empêchement de la rétribution (vipākāvarana). Quels en sont les caractères ?” Réponse : (111-31) "Les actes à rétribution immédiate (anantarya), irrémissibles, peuvent produire des passions. Dans les mauvaises destinées (durgati) on ressent la mauvaise rétribution. Tels sont les empêchements, sachez-le.” Ces trois essences (dharma) sont empêchements: parce qu'elles rendent impossible la réception des essences saintes (āryadharma), on les appelle empêchements (p. 95).
Probably nivarana is an error for āvarana. Mrs. Armelin translates HET one in the same way as Willemen. Her translation of the first half of the stanza does not bring out the fact that it deals with the first two obstacles and has to be rectified accordingly.
The second passage translated by de La Vallée Poussin explains vijñapti and avijñapti: Pour l'avijñapti: quand on fait un acte d'une manière ferme, la : pensée peut changer, le germe demeure. Si, par exemple, un homme prend les engagements de moralité, sa pensée peut ensuite être mauvaise ou non-définie: la moralité continue cependant (T. 1550, p. 81203-4: * #EX .
Pitill F. # . T* C*. 58. Willemen has: "Noninformation: when the actions one performs are firm. This element arises even though it takes place among thoughts which are different, e.g. even among unwholesome and indeterminate thoughts someone who is well ordained follows (the precepts)” (p. 28). Mrs. Armelin translates: “La 'non-information, c'est, lorsque l'acte accompli est ferme, ce qui continue à fonctionner dans d'autres pensées où naissent des germes [d'essences mentales]. Ainsi, dans les pensées mauvaises et indéfinies d'un homme vertueux, qui pratique les régles de discipline (śīla), celles-ci continuent, les unes et les autres, à guider sa conduite" (p. 75). Both translations are not entirely satisfactory. It seems to me that this passage has to be rendered as follows: "As to non-information, if one has firmly performed an act, a seed of it is produced (and continues to exist] while his thought changes. Just as when a man has well taken upon himself the discipline, it continues to exist while his thought is bad or non-defined.” De La Vallée Poussin's translation is excellent although it is somewhat free.
Stanza 151 (T. 1550, p. 813b20-21) describes the first trance. Willemen's translation of the stanza is as follows:
The one with five members, with adjusted and discursive thinking, and also with three feelings, different classes and four thoughts, they call it the first trance.
156

Page #7

BOOK REVIEWS
Mrs. Armelin's translation of the first pāda (Celle qui possède cinq membres comprend le "raisonnement" et la "réflexion," p. 158) is not correct because the commentary distinguishes between a trance with five members and a trance with vitarka and vicāra. The commentary explains first the five members (anga). Then the following question is put: "Since we know the five members, what need is there for adjusted and discursive thoughts now?" (Willemen, p. 107). The answer is: #
EAE. AT THE AC.
F
T TEX. (p. 823b29-ci). Willemen translates: "As for the members, the so-called wholesome ones are among the five members. They say that a defiled (trance) and an indeterminate one also have adjusted thoughts and discursive thoughts, but not the wholesome ones.” Willemen explains in a note that the five argas are kusala in the kuśala first dhyāna. When klişta or avyāksta, vitarka and vicāra are not among the argas. Willemen's explanation is correct but I believe that his translation has to be rectified as follows: "The members are said to be wholesome. They [vitarka and vicāra] are included in the five members. [With vitarka and vicāra] refers to a defiled and an indeterminate trance. They also comprise vitarka and vicāra but are not wholesome.” Mrs. Armelin states in a note (547): "Toute cette réponse à la question ainsi posée est fort obscure. Vasubandhu n'est guère plus explicite.” Indeed Vasubandhu does not deal with this problem but in the Mahāvibhāṣā there is a long discussion on the problem of constituent elements of defiled dhyānas which do not possess the characteristics of argas (T. 1545, ch. 161, p. 814ab).
In the Chinese translation of Dharmasri's Abhidharmahsdaya Dk is used to render vyañjana as explained in Dharmatrāta's Abhidharmahşdaya, T. 1552, p. 943a (cf. Willemen, p. VII, note 225). This has been overlooked by Mrs. Armelin who translates DF (p. 825016 and c24) with 'saveur' (cf. n. 612). Vyañjana is defined as follows: Dk
. 1872. (p. 831a3). Willemen translates: "Vyañjanas: the elements of a combination of sentences, etc. such as gāthās and scriptural texts” (p. 158). Mrs. Armelin translates: “Les syllabes (vyañjanakāya): l'ensemble des phonèmes dont la réunion forme un mot qui désigne un objet; comme, par exemple, 'stance,' 'aphorisme, etc.” (p. 216). This definition of vyañjana has been studied also by José van den Broeck who writes: “Dharmasri et Ghosaka interprètent le terme vyañjana comme signifiant "un discours": le vyañjanakāya est "une substance (dravya), consistant dans l'ensemble des phrases (pada)” (Abhidharmasära, T. 1550, ch. 4, p. 831a3); c'est "l'ensemble des énoncés développés” (Amrta)” (op. cit., p. 61; cf. also pp. 232 and 260). The Kokuyaku issaikyo translation does not follow the punctuation of the Taisho edition and puts a dot after : "Vyañjana, i.e., a detailed statement of a fact by a combination of padas, such as a gāthā and a sūtra.” In a note the translators explain that of here has the meaning
(bunsho) "literary composition.” It seems to me that the interpretation of
157

Page #8
THE EASTERN BUDDHIST this definition in the Kokuyaku issaikyo translation is the correct one but it is difficult to understand how vyanjana came to be interpreted in this way by Dharmasri and Ghosaka. There are many more problems in Dharmasri's Abhidharmahsdaya which ought to be examined but this would far exceed the limits of a review. Thanks to the conscientious efforts of Willemen and Mrs. Armelin this text has become now much more accessible. Their translations as well as the Kokuyaku issaikyo translation will be of great help in the study of this important and difficult text. Very welcome also are the glossaries compiled by the translators. Mrs. Armelin's book contains a Sanskrit-Chinese glossary (pp. 327-352) and a Chinese-Sanskrit glossary (pp. 353-371). Willemen has added an index of Sanskrit terms (pp. 314-331), a Chinese-Sanskrit-English glossary (pp. 332-351) and an EnglishSanskrit glossary (pp. 352-366). J. W. DE JONG 158