Reasons for comparing Sthiramati and Xuanzang

ON DOCTRINAL SIMILARITIES BETWEEN STHIRAMATI AND XUANZANG

HIDENORI SAKUMA

0. Introduction: Reasons for comparing Sthiramati and Xuanzang

The foundations of the system of doctrinal theories in the Chinese Faxiang 法相 school lie in the Cheng weishi lun 成唯識論, translated into Chinese by Xuanzang 玄奘 (with the wishes of his disciple Kuiji 窺基 being said to have been also strongly reflected in this translation). In the Hossō 法相 school of Japan too, efforts were made to preserve this tradition as faithfully as possible. Furthermore, in the traditions of the Faxiang school it is believed that the Cheng weishi lun was compiled on the basis of several

commentaries on Vasubandhu’s Triṃśikā, with Dharmapāla’s interpretations being deemed to represent the legitimate interpretation. The Cheng weishi lun is not, in other words, a translation of a single scholar’s commentary, and the legitimate view was determined by picking and choosing among several diverging views. The same method had been used when Xuanzang translated the Buddhabhūmiśāstra ten years earlier, with the interpretations of Bandhuprabha being deemed to represent the legitimate interpretation among those of several other scholars.

In the Faxiang school, the views of various Indian scholars were assessed in accordance with their treatment in the Cheng weishi lun. With the views of Dharmapāla being deemed to represent the legitimate view, the ideas of other scholars were recorded and judged to be not legitimate, and one gains the impression that Sthiramati in particular was an important target of criticism. This can also be inferred from Fukaura Shōbun’s detailed study of the Cheng weishi lun, in which he remarks more than once that Sthiramati was the scholar who stood on a par with Dharmapāla.1

1 Fukaura gives, for instance, the following explanation (1954, vol. 1: 341): “Were one to seek a great figure comparable with Dharmapāla among the ten great Yogācāra

Here we need to consider the question of how to deal with Dharmapāla. While we can accept that the Cheng weishi lun was compiled from a position that regarded Dharmapāla’s views as legitimate, there survives no commentary on the Triṃśikā by Dharmapāla himself in either the original Sanskrit or a Tibetan translation. In addition, to the best of my knowledge the only other work attributed to Dharmapāla survives only in

Chinese translation. If translations can be said to reflect the aims of the translator, this means that there exist no sources by which we can ascertain Dharmapāla’s true intent. If, as the traditional view would have us believe, Dharmapāla died at the early age of twenty-nine and Śīlabhadra, thought to have been one year older, succeeded him as head of Nālandā University and met Xuanzang when he was over one hundred years old, it must be assumed that Dharmapāla’s all-important system of doctrinal theories to be seen in the Cheng weishi lun

scholars, one would indeed have to point to Sthiramati. But his style of scholarship, as has already been mentioned, differs completely from that of Dharmapāla and adopts the stance of the merging of essential nature and external characteristics, recognizing the identity of phenomena and thusness.” It has been recognized in Japan too that Sthiramati has traditionally been understood as standing in opposition to Dharmapāla. But the assertion that Sthiramati’s philosophical tendencies are the same as those of Paramārtha, who translated the Mahāyānasaṃgraha and Mahāyānaśraddhotpādaśāstra, is no more than speculation on the part of Fukaura, and

there is a strong possibility that Sthiramati’s works were modified when being translated into Chinese as a result of the Chinese predilection for Tathāgatagarbha thought. It is questionable whether it is valid to go beyond the reflection of this predilection in the Shelun 攝論 school and link it to Sthiramati in India. In my experience, it is difficult to find any proof in extant commentaries by

Sthiramati of Fukaura’s assertion that Sthiramati’s style of scholarship, characterized by the merging of essential nature and external characteristics, was taken over by Paramārtha and developed into a doctrine asserting that all beings have one and the same nature. It should be noted that Sthiramati is not mentioned in the main text of the Cheng weishi lun and appears only in the afterword in a reference to the ten great bodhisattvas “Dharmapāla, Sthiramati, and so on.”

had been transmitted by Śīlabhadra. But in the doctrinal theories that I have studied to date this has not been the case. For example, whereas the correspondences between the eight consciousnesses (vijñāna) and four knowledges (jñāna), regarded as a matter of common knowledge in Faxiang doctrine, are firmly entrenched in the Cheng weishi lun, they are still in a transitional stage in Śīlabhadra’s Buddhabhūmivyākhyāna, which survives in Tibetan translation. How is one to comprehend the fact that something which ought to have been finalized at the time of Dharmapāla is still in a transitional stage in Śīlabhadra’s

writings? We may have to conclude that Dharmapāla did not give any thought to the correspondences between the eight consciousnesses and four knowledges.

Sthiramati (Anhui 安慧), meanwhile, is said to have been based at Valabhī and to have been a contemporary of Dharmapāla. But the scholar mentioned by Xuanzang alongside Guṇamati (Dehui 德慧) in the Datang xiyu ji in his accounts of Nālandā (9.3.5) and Valabhī (11.8.4) is not Anhui but Jianhui 堅慧. In the Datang Daciensi sanzang fashi zhuan 大唐大慈恩寺三藏法師傳 his name is given as An-hui. Among works included in the Taishō edition, the author of the Dacheng fajie wuchabie lun 大乘法界無差別論 (T. 31, nos. 1626 & 1627; neither translated by Xuanzang) is given as Jianhui, while the author of the Dacheng apidamo zaji lun 大

乘阿毘達磨雜集論 (T. 31, no. 1606; translated by Xuanzang) and Dacheng guang wuyun lun 大乘廣五蘊論 (T. 31, no. 1613; translated by Divākara) is given as Anhui. While a detailed examination of this state of affairs will be omitted here, the original Sanskrit equivalent of

both Jianhui and Anhui may be considered to have been Sthiramati.5 Further, a possible point of contact between Sthiramati and Xuanzang would have been Shengjun 勝軍 (*Jayasena), a contemporary of Dharmapāla who is mentioned in the Datang Daciensi sanzang fashi zhuan together with Guṇamati in connection with both Nālandā and Valabhī. It is recorded that Shengjun, under whom Xuanzang studied or with whom he spent time, had studied under Sthiramati and Śīlabhadra, and since it is not stated that Xuanzang

actually met Sthiramati, it would be reasonable to assume that there was no direct contact between Xuanzang and Sthiramati. A point worth noting is that nowhere in these works is it stated that the ideas of Xuanzang were in conflict with those of Sthiramati. Judging from the inscriptions at Valabhī, there was not just one scholar named Sthiramati. But in order to simplify the following discussion, I shall proceed on the assumption that the Sthiramati who wrote a commentary on the Mahāyānasūtrālaṃkāra, the Sthiramati referred to in inscriptions at Valabhī, and the Sthiramati under whom Jayasena studied were all the same person, and that he was, moreover, an important figure at Valabhī, which

5 Tsukamoto Keishō (1996: 526, Waḷā 1) already equates Dehui and Jianhui with Guṇamati and Sthiramati respectively. In Li Rongxi’s translation of the Datang xiyu ji included in the BDK English Tripiṭaka, Jianhui is rendered as “Sthiramati” (Li 1996: 284,3 & 343,4). Likewise, in Li’s translation of the Datang Daciensi sanzang fashi zhuan Anhui is also rendered as “Sthiramati” (Li 1995: 126,21). Previously, Hirakawa Akira (1979: 14) had already suggested that Jianhui might be the same person as Anhui. On the assumption that this view has become established in academic circles, I have therefore decided to regard both Jianhui and Anhui as Chinese equivalents of Sthiramati.

ranked with Nālandā as a centre of Buddhist learning. On this basis, I shall set about ascertaining the fact that the theories presented in Sthiramati’s commentary on the Mahāyānasūtrālaṃkāra and the theories deemed to be legitimate in the Cheng weishi lun bear a close resemblance to each other.

Now, even if Dharmapāla was a contemporary of Sthiramati, there would still seem to be difficulties in immediately equating the views deemed to be legitimate in the Cheng weishi lun with those of Dharmapāla when one considers that none of his writings have survived in the original and one also takes into account the passage of time within the confines of Nālandā from Dharmapāla to Śīlabhadra and then to Xuanzang. But if one posits a picture pitting Sthiramati, based at Valabhī, against Dharmapāla, based at Nālandā, it seems strange that Sthiramati should have already completed the systemization of the correspondences between the eight consciousnesses and four knowledges.

At any rate, a scheme of correspondences between the eight consciousnesses and four knowledges cannot be found in Śīlabhadra’s writings but does exist in Xuanzang’s translations. It thus seems that this theory was either formulated by Xuanzang, who had an extraordinary enthusiasm for Abhidharmic systemization, during the course of translation or else he knew of the interpretation given in Sthiramati’s commentary on the Mahāyānasūtrālaṃkāra and adopted this as the legitimate interpretation. Xuanzang would naturally have known of Prabhākaramitra’s

translation of the Mahāyānasūtrālaṃkāra. Here I shall assume that since comparatively few typically Chinese interpretations seem to have found their way into the Chinese translation of the Mahāyānasūtrālaṃkāra, unlike in the case of the Mahāyānasaṃgraha, Xuanzang did not produce a new translation. That being so, it is probably safe to suppose that although Xuanzang translated neither the verses of the Mahāyānasūtrālaṃkāra together with Vasubandhu’s commentary, which together constituted an important work, nor the commentaries by Asvabhāva and Sthiramati, he was fully cognizant of their content.

Similar evidence can in fact also be found in connection with several other doctrinal theories. In order to make clear the thread of my arguments, I shall therefore in the following proceed on the assumption that the views deemed to be legitimate in the Cheng weishi lun were not advocated by Dharmapāla, but were actually propounded by Xuanzang at the instance of Kuiji. As for Sthiramati, I shall put to one side his connections with Valabhī and Dharmapāla and focus solely on the content of the commentaries attributed to him, which I shall consider to represent Sthiramati’s theories. 1. Focal points for a comparative examination of the doctrinal theories of Sthiramati and Xuanzang

When undertaking a comparative study of the doctrinal theories of Sthiramati and Xuanzang, it is necessary to indicate the criteria on which such a study is based. In the case of Sthiramati, I consider the Sanskrit originals and Tibetan translations of works attributed to him (the latter of which may be regarded as word-for-word translations) to represent his views. In the case of Xuanzang, on the other hand, I consider the Mahāyānasaṃgraha, Buddhabhūmiśāstra, and other works translated by him to represent his views. To regard Tibetan translations as the

equivalent of originals while viewing Chinese translations as expressions of the thought of their translator Xuanzang may seem to indicate a lack of consistency in my criteria. But grounds for equating Xuanzang’s translations with his own thought can be found in the findings of several researchers, and this ensures the validity of my criteria. Nonetheless, it is true that there is a difference between the criteria, and scrupulous care will be taken in the treatment of all material.

Next, I wish to mention the doctrinal theories I shall use as indices in my comparative examination. I shall focus on the following three topics, regarding which I have already achieved some results in past investigations.

1. Correspondences between the four knowledges and eight consciousnesses 2. Correspondences between the four knowledges and three bodies 3. The formation of the five-gotra system

1.1. Similarities between Sthiramati and Xuanzang as seen from correspondences between the four knowledges and eight consciousnesses The correspondences between the four knowledges and eight consciousnesses are not mentioned in either the verses of the Mahāyānasūtrālaṃkāra or the prose commentary attributed to Vasubandhu, both preserved in Sanskrit,14 nor are they mentioned in the Tibetan translation of Asvabhāva’s commentary on the Mahāyānasaṃgraha.15 In Śīlabhadra’s Buddhabhūmivyākhyāna (preserved in Tibetan) we find evidence of a transitional stage in the formulation of these correspondences.16 The finalized scheme of correspondences is

14 MSA(Bh) IX.67 (F: 38,18-23; L: 46,15–19): buddhajñānavibhāge daśa ślokāḥ / ādarśajñānam acalaṃ trayajñānaṃ tadāśritam / samatāpratyavekṣāyāṃ kṛtyānuṣṭhāna eva ca // 67 // caturvidhaṃ buddhānāṃ jñānam ādarśajñānaṃ samatājñānaṃ pratyavekṣājñānaṃ kṛtyānuṣṭhānajñānaṃ ca / ādarśajñānam acalaṃ trīṇi jñānāni tadāśritāni calāni /

15 See n. 10. The relevant passage in the Tibetan translation begins as follows (Haka-maya 2001: 496): rnam par shes pa’i phung po gyur pas ni me long lta bu dang / mnyam pa nyid dang / so sor rtog pa dang / bya bas grub pa’i ye shes la dbang ’byor pa thob ste /. Correspondences with the eight consciousnesses are not mentioned in any subsequent passages either. It is obvious from the material cited by Hakamaya that the correspondences between the four knowledges and eight consciousnesses are given in the corresponding Chinese translation by Xuanzang (see n. 17).

16 Nishio 1940, vol.1: 120,17–121,15: rnam pa gcig tu na dngos po shes pa dang / de dmigs pa zhes bya ba zlas dbye ba yin te / gnas ngan len mtha’ dag gi gnas kun gzhi rnam par shes pa gnyen po’i stobs kyis gnas ngan len ma lus pa dang bral bas yongs su gyur pa me long lta bu’i ye shes zhes bya ba gzhan gyi dbang dag pa zhes tha snyad gdags pa sems kyi rnam par rtog pa thams cad med pa’i ngo bo la ’di ni dngos po tsam mo zhes spyi’i rnam par sgro btags nas dngos po’i sgra brjod do // me long lta bu’i ye shes dmigs par bya ba dang / dmigs pa mnyam pa’i rnam pa yang gnas ngan len gyi gnas yongs su gyur na / de ltar rnam par bzhag go //

dngos po de shes pa ni dngos po shes pa ste / de la dmigs pa zhes bya ba’i tha tshig go / de yang mnyam pa nyid kyi ye shes yin no // de’i rjes la thob pa dag la ʼjig rten pa rang gi rtog pa yongs su gcod pa’i rnam pa gang yin pa de’i spyod yul yang gzhan gyi dbang gi ngo bo nyid yongs su gyur pa yin no // rjes las thob pa’i ye shes de ni so sor rtog pa’i ye shes kho na yin te / de rang gis rtogs pa la so sor rtog pa’i tshe / de gnyis yul yin pa’i phyir ro // de la ʼam de gnyis la dbang ba ste / ’di de la zad mi shes pa’i mtshan nyid yod pas zhes bya bar tshig rnam par sbyar ro // ’dis mtshon par byed shes par byed pas na zhes byed pa’i byed pa por byas pa’i phyir ro // dngos pa shes pa de dmigs pa de la ʼam dngos po shes pa dang / de dmigs pa de gnyis la dbang zad mi shes pa’i mtshan nyid ces bya ba’i tha tshig ste / ’dis ni lam gnas yongs su

found in Xuanzang’s translations of the Buddhabhūmiśāstra17 and Cheng weishi lun, and the same theory appears in Sthiramati’s commentary on the Mahāyānasūtrālaṃkāra. These points were touched on briefly in the previous section.

gyur yang bstan pa yin no // nyon mongs pa can gyi yid yongs su gyur na / me long lta bu’i ye shes chos kyi dbyings la dmigs nas mnyam pa nyid kyi ye shes skye ste / de bzhin yongs su dag na de la yang dbang zad mi shes pa ’thob po // The correspondences between ālayavijñāna and ādarśajñāna and between kliṣṭamanas and samatājñāna are clearly defined here, but it is not clear which consciousness is transformed into pratyavekṣājñāna, and there is no explanation here or elsewhere regarding kṛtyānuṣṭhānajñāna. The corresponding passage in Xuanzang’s translation of the Buddhabhūmiśāstra (T. 26: 324b4ff.) reads as follows: 有義此顯自性一

分。佛果四智即六相中自性一分。有爲功徳法者即是大圓鏡智。由對治力轉去一切麁重所依阿頼耶識。轉得清淨依他起性。遠離一切心慮分別。所縁能縁平等平等不可宣説。縁生法性不増不減。内證行相能現一切諸法影像。於一切境普能照了無分別故。總説名法智者。即是平等性智。由對治力轉去執著衆生及法第七末那。轉得清淨依他起性縁鏡智等及淨法界平等平等。内證行相故名爲智。彼所縁者。

即餘二智。由對治力轉去世間分別六識。轉得清淨依他起性。或出世間。或世出世。彼後所得縁上眞如及法智等。依他起性以爲境界。無執分別似所縁現。分別自内所證能證。用彼上説眞如法智。爲所縁故名彼所縁 。 Here the original would seem to have been been modified, and it is stated that the other two knowledges are connected to the six consciousnesses, although the translator did not go so far as to state which knowledge is connected to which consciousness.

17 An explicit indication of the relationship between the four knowledges and eight consciousnesses is found in the following passage (T. 26: 302b29ff.): 轉識蘊依得四無漏智相應心。謂大圓鏡心廣説乃至成所作心。轉第八識得大圓鏡智相應心。能持一切功徳種子能現能生一切身土智影像故。轉第七識得平等性智相應心。遠離二執自他差別證得一切平等性故。轉第六識得妙觀察智相應心。能觀一切皆無礙故。轉五現識得成所作智相應心。能現成辨外所作故。

No variants have been reported for this passage. It is thus evident that it presents the legitimate view of the Faxiang school in an unadulterated form. The corresponding section is, moreover, completely missing in the Tibetan translation of Śīlabhadra’s Buddhabhūmivyākhyāna. One is thus compelled to accept that this passage was added by Xuanzang.

In addition, it is also clear that the correspondences found in Prabhākaramitra’s Chinese translation of the Mahāyānasūtrālaṃkāra20 and Xuanzang’s Chinese translation of Asvabhāva’s commentary on the Mahāyānasaṃgraha differ from those given by Sthiramati and in Xuanzang’s translations of the Buddhabhūmiśāstra and Cheng weishi lun. In the former group the correspondences are ālayavijñāna – ādarśajñāna, manas – samatājñāna, manovijñāna – kṛtyānuṣṭhānajñāna, and five active consciousnesses – pratyavekṣājñāna, while in the latter group the

correspondences are ālayavijñāna – ādarśajñāna, manas – samatājñāna, manovijñāna – pratyavekṣājñāna, and five active consciousnesses – kṛtyānuṣṭhānajñāna. Since I have already demonstrated elsewhere that originally the former set of correspondences would have been the more natural interpretation, I shall not go into any further detail here.

Important in this regard is the fact that Prabhākaramitra’s translation of the Mahāyānasūtrālaṃkāra and Xuanzang’s translation of Asvabhāva’s commentary on the Mahāyānasaṃgraha were trans-

brgyad las kun gzhi dag na me long lta bu’i ye shes su gyur ro // nyon mongs pa’i yid dag na mnyam pa nyid kyi ye shes su ’gyur ro // yid kyi rnam par shes pa dag na so sor kun du rtog pa’i ye shes su ’gyur ro // mig nas lus kyi bar du rnam par shes pa lnga dag na bya ba grub pa’i ye shes su ’gyur te / ye shes bzhi dang chos kyi dbyings rnam par dag pa lnga thob pa la gnas gzhan du gyur pa lnga zhes ba’o // Similar explanations can also be found elsewhere in the same work.

20 T. 31: 606c23ff. — 四智鏡不動 三智之所依 八七五六識 次第轉得故 釋曰。四智鏡不動三智之所依者。一切諸佛有四種智。一者鏡智。二者平等智。三者觀智。四者作事智。彼鏡智以不動爲相。恒爲餘三智之所依止。何以故。三智動故。八七五六識次第轉得故者。轉第八識得鏡智。轉第七識得平等智。轉五識得觀智。轉第六識得作事 智。是義應知。

This represents the reading of the old Song edition, the oldest manuscript used by the editors of the Taishō edition when editing this text. lated earlier than Xuanzang’s translations of the Buddhabhūmiśāstra and Cheng weishi lun.23 The legitimacy of the correspondences is explained in the Buddhabhūmiśāstra, but the only reason given is the weak reason that the former set of correspondences is irrational because it does not follow the

regular order, while the latter set represents the legitimate view because it follows the regular order. Why would Xuanzang have been compelled to give such a reason? If Sthiramati had prior to this set forth this latter set of correspondences, it would mean that Xuanzang rejected the views of Śīlabhadra and Prabhākaramitra, the latter of whom is thought to have studied at Nālandā, and emended it on the basis of Sthiramati’s view. If Sthiramati’s view should prove to have been formulated around the same time as Xuanzang advanced this view, it would become necessary to rethink Sthiramati’s dates. Such is the positional relationship between Sthiramati and Xuanzang as deduced from our first index.

1.2. Similarities between Sthiramati and Xuanzang as seen from correspondences between the four knowledges and three bodies In order to simplify things, I first wish to confirm the following facts. The purity of the Dharma-realm (dharmadhātuviśuddhi) was added to the four knowledges as a distinguishing feature of the state

23 According to Kuwayama Shōshin (Kuwayama and Hakamaya 1981: 49ff.), Xuanzang would have met Prabhākaramitra shortly before his departure for Central Asia and India and would have obtained from him information about these regions and about Śīlabhadra at Nālandā. I too believe that this is highly likely to have been the case. As is noted by Hakamaya (ibid.: 195), it may be safely assumed that Prabhākaramitra translated the Mahāyānasūtrālaṃkāra during Xuanzang’s twenty-year absence from China. A list of works translated by Xuanzang with their dates can be found in the same work (252ff.).

of Buddhahood, and together these are referred to as the five dharmas (or five elements). The purity of the Dharma-realm is characterized as principle and the four knowledges as wisdom. Originally the five dharmas and three bodies represented different schemata, and the process of their development also differed. The four knowledges of the Buddha appear in their finalized form already in the Sanskrit verses of the Mahāyānasūtrālaṃkāra (IX.6776). Since there are no earlier passages indicative of the development of this concept, the circumstances of

its establishment are not known. As for Buddha-bodies, the basic theory until then had posited two bodies, namely, the physical body (rūpakāya) and the Dharmabody (dharmakāya). With the emergence of the Yogācāra school, a three-body theory consisting of the dharmakāya or svabhāvakāya, the saṃbhogakāya (enjoyment-body), and the nirmāṇakāya (transformation-body) came to be advanced from the

standpoint of Mahāyāna Buddhism. Because of the use of the two different terms dharmakāya and svabhāvakāya, these came to be treated as two separate bodies, resulting in effect in a four-body theory. In later times, the four-body theory developed into a five-body theory and other theories of multiple Buddha-bodies.

The four knowledges and three bodies are mentioned in chapter 9 of the Mahāyānasūtrālaṃkāra, with the purity of the Dharma-realm being discussed in verses 56-59, Buddha-bodies in verses 60-66, and the four knowledges in verses 67-76. But there is no mention of any correspondences between them in either the verses or Vasubandhu’s and Asvabhāva’s commentaries, and they appear only in Prabhākaramitra’s Chinese translation of the Mahāyānasūtrālaṃkāra and in the Tibetan translation of Sthiramati’s commentary (SAVbh). Apart from this, the correspondences between the four knowledges and three bodies are also described in the Tibetan translation of Śīlabhadra’s Buddhabhūmivyākhyāna.

The correspondences between the five dharmas and three bodies are clearly described in Sthiramati’s SAVbh. In SAVbh IX.60 they are explained with reference to āśrayaparāvṛtti: ālayavijñāna turns into ādarśajñāna and is associated with the dharmakāya, which also corresponds to the svabhāvakāya; kliṣṭamanas turns into samatājñāna and manovijñāna into pratyavekṣājñāna, and these are associated with the saṃbhogakāya; and the five active consciousnesses turn into kṛtyānuṣṭhānajñāna, which is associated with the nirmāṇakāya.

In Prabhākaramitra’s Chinese translation of the Mahāyānasūtrālaṃkāra these correspondences are indicated in X.53ff., corresponding to IX.59ff. in the Sanskrit text. Prabhākaramitra presents the relationship between the eight consciousnesses and four knowledges in a form different from

that of Sthiramati and Xuanzang, and it may be summarized in the following manner: ālayavijñāna turns into ādarśajñāna and kliṣṭamanas into samatājñāna, and these are associated with the dharmakāya; the five active consciousnesses turn into pratyavekṣājñāna, which is associated with the saṃbhogakāya; and manovijñāna turns into kṛtyānuṣṭhānajñāna, which is associated with the nirmāṇakāya.

In the case of Śīlabhadra’s Buddhabhūmivyākhyāna, on the other hand, in which the correspondences between the four knowledges and eight consciousnesses have not been finalized, one must posit the following relationships. First, it is stated that ālayavijñāna turns into ādarśajñāna and kliṣṭamanas into samatājñāna, but no relationships are posited between the other consciousnesses and knowledges. Under these circumstances, the correspondences between the five dharmas and three bodies are as follows: the purity of the Dharmarealm and ādarśajñāna are associated with the svabhāvakāya (= dharmakāya), samatājñāna and pratyavekṣājñāna are associated with the saṃbhogakāya, and kṛtyānuṣṭhānajñāna is associated with the nirmāṇakāya.

be quoted here. Reference should be made to my earlier studies on this subject (Sakuma 1982, 1987, 1989). Thus, the correspondences between the five dharmas and three bodies differ from one text to another, and in content they are even more complicated than has been indicated in the above.

That being so, how are these correspondences treated in Xuanzang’s translations of the Buddhabhūmiśāstra and Cheng weishi lun? First, in the case of the Buddhabhūmiśāstra it is difficult to comprehend even the gist of the relationship between the five dharmas and three bodies. If one assumes that Xuanzang considered the connections between the two with reference to Śīlabhadra’s Buddhabhūmivyākhyāna, it is to be surmised that he decided that it would

be difficult to deal with the relationship between the five dharmas and three bodies, and also the eight consciousnesses, with the consistency of Abhidharmic categories. It is obvious, in other words, that Xuanzang was rather perplexed about the relationship between the five dharmas, three bodies, and eight consciousnesses when he translated the Buddhabhūmiśāstra.

How much clearer, then, had the relationship between these three become when Xuanzang translated the Cheng weishi lun ten years later? In the Cheng weishi lun, the relationship between pratyavekṣājñāna and the three bodies is by no means clear, but it can be generally inferred that the purity of the Dharma-realm is associated with the svabhāvakāya, ādarśajñāna with the self-enjoyment body, samatājñāna with the other-enjoyment body, and kṛtyānuṣṭhānajñāna with the nirmāṇa-kāya. Pratyavekṣājñāna is subtly related to both the other-enjoyment body and the nirmāṇa-kāya, but I cannot go into details here.

The concepts of self-enjoyment body and other-enjoyment body had in fact already appeared in Xuanzang’s translation of the Buddhabhūmiśāstra, but it was only in the Cheng weishi lun that they were to some extent clearly utilized in explaining the relationship between the five dharmas and three bodies, and they result in effect in a four-body theory. The four-body theory is discussed at great length in the Abhisamayālaṃkāra, and possible connections with this work raise some interesting questions. But the Abhisamayālaṃkāra was not translated into Chinese, and I shall not delve any further into this subject since it would lead us away from the question at hand.

As for the two concepts of self-enjoyment body and other-enjoyment body, it is in fact possible to detect intimations of the former in Sthiramati’s SAVbh. Unfortunately the procedures necessary for demonstrating this are somewhat involved, and limited space does not allow me to reproduce them here. Reference can be made to my previously published study on this subject.

If my above analysis is correct, it is possible to infer the following process. The five dharmas and three bodies initially developed as two separate theories, but by the time of Śīlabhadra and Prabhākaramitra correspondences between the two had been established. Xuanzang had doubts about his teacher Śīlabhadra’s views in terms of Abhidharmic categories, and his solution could be easily explained were one to assume that he

borrowed the notion of the selfenjoyment body and the schema of correspondences between the four knowledges and eight consciousnesses from Sthiramati. This is, of course, no more than a possibility, but in light of the investigations I have conducted until now, it would seem natural to me to view the situation in this fashion. Such is the positional relationship between Sthiramati and Xuanzang as deduced from our second index.

他地上菩薩所宜變現淨土。或小或大或劣或勝。與他受用身作所依止處。謂隨初地菩薩所宜現小現劣。如是展轉乃至十地最大最勝。於地地中初中後等亦復如是。Also 294b14ff.: 自受用土雖遍法界一一自變各自爲主不相障礙。他受用土雖諸佛變然一合相亦一相身攝受爲主不相障礙。There are no corresponding passages in the Tibetan translation of Śīlabhadra’s Buddhabhūmivyākhyāna.

1.3. Similarities between Sthiramati and Xuanzang as seen from the formation of the five-gotra system One theory propounded by the Faxiang school in China and Japan that became the cause of much debate with other schools was the five-gotra system presented in the Buddhabhūmiśāstra translated by Xuanzang. When one traces its content back to India, one can certainly detect a process whereby the part of the five-gotra system relating to beings without any possibility of attaining Buddhahood and the part relating to the

three vehicles gradually merged. It is to be surmised that Mahāyāna Buddhism advanced the idea of the three vehicles of the śrāvaka, pratyekabuddha, and bodhisattva out of a need to assert its legitimacy vis-à-vis Mainstream Buddhism, but initially beings without any possibility of attaining Buddhahood were not juxtaposed to the three vehicles. The question of gotra (lineage) was simply discussed quite separately from the idea of three vehicles in terms of beings with the possibility of attaining Buddhahood (gotra) and beings without any possibility of attaining Buddhahood (agotra). It would appear that these two separate groupings were first brought together as five categories in Sthiramati’s SAVbh.

The overall current of thought leading to the five-gotra system can be understood in the following manner. Initially, the vehicle among the three vehicles to which a practitioner belongs is not determined, and if one supposes that his association with one of the vehicles is gradually determined in the course of his practice, then the initial stage corresponds to the indeterminate lineage and the stage when his lineage has been determined corresponds to one of the three vehicles. Therefore, the indeterminate lineage and the three vehicles are not parallel categories. The question of gotra and agotra had already been raised from the time of the Yogācārabhūmi, and it can also be readily inferred that there was some

connection between the state of having the possibility of attaining Buddhahood (gotra) and the three vehicles. But it was in Sthiramati’s SAVbh that agotra is first presented alongside the indeterminate lineage and the three vehicles. The Laṅkāvatārasūtra is often considered to provide a theoretical basis for the five-gotra system, but as is indicated in the Yuqielun ji 瑜伽論記, it was known from an early stage that the Laṅkāvatārasūtra was unsuitable as the theoretical basis of the fivegotra system. In light of the above, I wish to show the process leading to the five-gotra system with reference to the “Gotra Chapter” in the Mahāyānasūtrālaṃkāra.

In the Sanskrit text of the Mahāyānasūtrālaṃkāra, the first ten verses of the “Gotra Chapter” explain the existence of gotra and the eleventh verse explains the absence of gotra, or agotra. Within this overall framework, the verses necessary for the establishment of the five-gotra system were verses 6 and 11. It would presumably be safe to assume that originally there was no intention in either the verse section or Vasubandhu’s commentary to forge a direct link between these two verses.

A verse on the distinction between the kinds [of lineages]: The lineage may be determinate or indeterminate, shaken or unshaken By conditions. This distinction between lineages is, in brief, fourfold. (v. 6) In brief, lineages are fourfold. They are determinate and indeterminate, and these are in [that] order unshaken and shaken by conditions. (Mahāyānasūtrālaṃkāra III.6)

As can be seen in this verse, there are lineages that are determinate and others that are indeterminate. Since the term “three vehicles” is used in Vasubandhu’s commentary on verse 2, “determinate” means belonging to one of the three vehicles. “Indeterminate,” on the other hand, means that the practitioner, under the guidance of a teacher, is still in a state of vacillation

regarding his lineage. This later became the independent category of “indeterminate lineage,” but it is unlikely to have been regarded as an independent category at this stage. This verse simply gives expression to the process of practice, that is, to the fact that there are both those who, under the guidance of a teacher and so on, are no longer in a state of vacillation and those who are still vacillating in the midst of their practice.

In contrast, it is evident that in Asvabhāva’s commentary (MSAṬ)35 and Sthiramati’s SAVbh36 these four categories have clearly come to be treated as the three vehicles of the śrāvaka, pratyekabuddha, and bodhisattva and, independent of these, an indeterminate lineage. Next, Asvabhāva’s MSAṬ and Sthiramati’s SAVbh begin their explanations of Mahāyānasūtrālaṃkāra III.11 in the following manner. Asvabhāva’s commentary on verse 11 begins by commenting directly on Vasubandhu’s commentary with the statement “Where it says, ‘In this sense “he who does not have the quality for parinir-

35 MSAṬ III.6 (D. 51b6-52a3; P. 58b6-59a2): rigs nges pa ni nyan thos dang / rang sangs rgyas dang / sangs rgyas kyi rigs su nges par gnas pa gang yin pa ste / nyan thos nyid thob (D: ’thob P) par nges pa’i rigs gang yin pa de ni nam yang rang sangs rgyas dang sangs rgyas nyid ’thob pa’i rgyur mi ’gyur ro // de bzhin du rang sangs rgyas dang / sangs rgyas kyi rigs dag kyang sbyar bar bya’o // ma nges pa ni (em.: pa’i DP) rkyen gyi dbang gyis nyan thos dang rang sangs rgyas dang / sangs rgyas kyi (D: kyis P) rigs rnams kyi rgyur ’gyur te / dper na ri’i phyogs gang dag la (P: las D) gdon mi za bar gser ’ba’ zhig ’byung gi / dngul ’ba’ zhig kyang ma yin la / zangs ’ba’ zhig kyang ma yin pa de lta bu yang yod la / phyogs gang zhig ’jim gong dril ba la sogs pa’i bcos legs (P: lags D) bya ba’i dbang gyis (D: gyi P) gdon mi za bar res ’ga’ gser ’byung la / res ’ga’ dngul la sogs pa ’byung bar yang yod pa de lta bu’o // de nyid kyi phyir rigs nges pa ni / rkyen rnams kyis mi ’phrogs la ma nges pa ni ’phrogs pa yin no //

36 SAVbh III.6 (D. 45a4-45b1; P. 49a3-49b1): de la rigs nges pa ni gang nyan thos su rigs nges par gnas pa dang / rang sangs rgyas su rigs nges par gnas pa dang / sangs rgyas su rigs nges par gnas pa ste / nyan thos su rigs nges par gnas pa yang rigs des nyan thos kyi byang chub nyid ’thob kyi ji ltar byas kyang nams kyang rang sangs rgyas kyi byang chub dang / sangs rgyas su ’thob pa’i rgyur mi ’gyur ro // rang sangs rgyas kyi rigs nges pa yang rigs des rang sangs rgyas kyi byang chub nyid thob kyi ji ltar byas kyang nams kyang nyan thos dang sangs rgyas kyi byang chub ’thob pa’i rgyur mi ’gyur ro // [[sangs rgyas]] kyi rigs can yang rigs des sangs rgyas kyi byang chub nyid ’thob (D: thob P) kyi ji ltar byas kyang nams kyang nyan thos dang rang sangs rgyas kyi byang chub tu mi ’gyur ba’o // rigs ma nges pa ni rkyen gyi dbang gis nyan thos dang rang sangs rgyas dang sangs rgyas (D: dang sangs rgyas, missing in P) kyi rigs gang yang rung ba cig gi (D: gyis P) rgyur ’gyur te / nyan thos kyi dge ba’i bshes gnyen dag gis bsgral na ni nyan thos kyi rigs can du yang ’gyur / rang sangs rgyas kyi dge ba’i bshes gnyen gyis bsgral na ni / rang sangs rgyas kyi rigs can du yang ’gyur / byang chub sems dpa’i dge ba’i bshes gnyen gyis bsgral na ni sangs rgyas kyi rigs can du yang ’gyur ro (D: gyur ba’o P) //

vāṇa” is meant by “he who dwells in no lineage”’,”37 but Sthiramati adds: “Where it says ‘a verse on the distinction of the lineage-less,’ having earlier explained the lineage of the śrāvaka, the lineage of the pratyekabuddha, the lineage of the bodhisattva, and the indeterminate lineage, it now explains the lineage-less.” Whereas Asvabhāva clearly refers to the lineages of the three vehicles and the indeterminate lineage in his commentary on verse 6, but does not link them directly to the verse on agotra, there is clear evidence in Sthiramati’s commentary of an intent to create a scheme of five gotras. Here one can discern the manner in which the scheme of five gotras gradually evolved.

In Xuanzang’s translation of the Buddhabhūmiśāstra this scheme developed into five categories consisting of the lineages of the three vehicles, the indeterminate lineage, and the lineage-less, the last of which was simplified to mean those without any possibility of attaining Buddhahood. In the Tibetan translation of Śīlabhadra’s Buddhabhūmivyākhyāna there is no mention whatsoever of these ideas. Originally, Mahāyānasūtrālaṃkāra III.11 referred to two kinds of agotra, namely, those who are unable to attain Buddhahood at the present point in time but will be able to after a certain period of time, and those who will never attain Buddhahood. It is to be surmised, therefore, that in order to simplify the five gotras, Xuanzang

37 MSAṬ Ⅲ.11 (D. 52b1f.; P. 59a8f.): don ’di la ni rigs med pa la gnas pa yongs su mya ngan las mi ’da’ ba’i chos can yin par bshad do zhes bya ba na. restricted the meaning of agotra to those without any possibility of attaining Buddhahood.

To sum up, the five categories of the five-gotra system do not appear in the verses of the Mahāyānasūtrālaṃkāra or Vasubandhu’s commentary; in Asvabhāva’s commentary the content of III.6 was clarified in the form of the lineages of the three vehicles and an indeterminate lineage, and Sthiramati further linked this verse to agotra mentioned in III.11; Xuanzang, it is to be surmised, simplified the content of agotra by restricting it to those without any possibility of attaining Buddhahood and thus brought to completion the five-gotra system, regarded as one of the distinguishing features of Faxiang doctrine. Here too one senses greater similarities between the doctrinal theories of Sthiramati and Xuanzang than between those of other scholars. 2. A comprehensive assessment: by way of conclusion

On the basis of the data on the three doctrinal theories summarized above, I wish to focus here in particular on the doctrinal similarities to be observed between Sthiramati and Xuanzang. The doctrinal theories selected here for the purpose of comparison represent of course just one part of the theories of the Yogācāra school, and therefore it is not my intention to apply the conclusions reached below to the entire body of Yogācāra theories. The correspondences between the four knowledges, the eight consciousnesses and the fivegotra system taken up in the above are doctrinal theories that in the Faxiang school of

China and Japan are treated as if they are self-explanatory, but they were not necessarily clearly defined in India, and therefore they are unlikely to have been central theories of the Yogācāra school. In point of fact, the correspondences between the four knowledges and eight consciousnesses as clarified in the Cheng weishi lun and the correspondences between the five dharmas and three bodies, clarified to a certain degree in the Cheng weishi lun, are not mentioned at all in the Sanskrit text of Sthiramati’s commentary on the Triṃśikā, on which the Cheng weishi lun would naturally have been based. Since it is to be surmised that Xuanzang would have been motivated by different aims in the case of the five-gotra system, it is not surprising that this is not mentioned in Sthiramati’s

commentary on the Triṃśikā. But in verses 29 and 30, which discuss āśrayaparāvṛtti, the body of emancipation (vimuktikāya), and the dharmakāya, Sthiramati neither mentions the four knowledges nor touches on the three bodies. In addition, there is no mention of the four knowledges in Xuanzang’s translation of the Abhidharmasamuccayabhāṣya, attributed to Sthiramati, nor do they of course appear in the Sanskrit Abhidharmasamuccayabhāṣya, said to be the work of Jinaputra, although the question of its authorship has not yet been resolved.43 This means that there is a need to consider why Xuanzang should have attributed it to Sthiramati. Likewise, there are no references to any such ideas in Sthiramati’s commentary on the Madhyāntavibhāga. At any rate, when one considers that in

his commentary on the Mahāyānasūtrālaṃkāra Sthiramati would seem to display an enthusiasm for using his encyclopaedic knowledge to systematize the doctrinal theories of the Yogācāra school in line with Abhidharmic categories, it seems strange that he makes no mention whatsoever of the four knowledges in his commentary on the Triṃśikā. Assuming that, as is currently estimated, he lived during the period between 510 and 570, would he have mentioned the four knowledges and discussed their relationship with the eight consciousnesses only in his commentary on the Mahāyānasūtrālaṃkāra among the voluminous commentaries he composed during his lifetime simply because the four knowledges happened to be mentioned in the verse section of the Mahāyānasūtrālaṃkāra? If that were the case, then Sthiramati’s failure to mention the four knowledges and three bodies in his commentary on the Triṃśikā could be explained by the fact that they do not figure in the verses of the Triṃśikā.

There is one further moot point, namely, the fact that up until the time of Xuanzang’s translations of the Buddhabhūmiśāstra and Cheng weishi lun one can trace in the Tibetan translation of the Buddhabhūmivyākhyāna (thought to be the work of Śīlabhadra), Prabhākaramitra’s Chinese translation of the Mahāyānasūtrālaṃkāra, and Xuanzang’s Chinese translation of Asvabhāva’s commentary on the Mahāyānasaṃgraha a process whereby the doctrinal theories of the four knowledges and eight consciousnesses gradually merged and their correspondences were developed. The same process can be seen in the correspondences between the five dharmas and three bodies, and if one recognizes a similar process with regard to the five-gotra system too, the theories presented in Sthiramati’s

commentary on the Mahāyānasūtrālaṃkāra turn out, as is evident from our above investigations, to have overly close similarities to the theories ultimately formulated by Xuanzang. It might be suggested that one should consider Sthiramati’s commentary on the Mahāyānasūtrālaṃkāra separately from all his other works and regard it as the work of someone else with the same name, but it is not such a simple matter. When one also takes into account developments in the idea of āśrayaparāvṛtti and questions relating to the treatment of the trisvabhāva theory, it becomes exceedingly complicated. Therefore, it is desirable to reach a conclusion here with the qualification that it applies only to the topics dealt with in the above. With such a qualification, it may be assumed that the relationship between Sthiramati and Xuanzang in the realm of philosophical thought was closer than we have until now imagined. With this as my conclusion for the time being, I wish to bring this paper to a close.

Abbreviations and references

Dey, Nundo Lal. 1927. The Geographical Dictionary of Ancient and Mediaeval India. London: Luzac. Fukaura Shōbun 深浦正文. 1954. Yuishikigaku kenkyū 唯識学研究 [Studies in Yogācāra doctrine], 2 vols. Kyoto: Nagata Bunshōdō 永田文昌堂. Hakamaya Noriaki 袴谷憲昭. 1969. “Genjō-yaku Shōdaijōron shaku ni tsuite – Chibettoyaku to no hikaku ni yoru ichi kōsatsu – ” 玄奘訳『摂大乗論釈』について – チベット訳との比較による一考察 – [[[Xuanzang’s]] translation of the Mahāyānasaṃgrahopanibandhana: Compared with its Tibetan translation]. Indogaku Bukkyōgaku Kenkyū インド学仏教学研究 18-1, pp. 140-141. Repr. in Hakamaya 2001, pp. 490-503.

____ 2001. Yuishiki shisō ronkō 唯識思想論考 [Studies in Yogācāra thought]. Tokyo: Daizō Shuppan 大蔵出版. Hirakawa Akira 平川彰. 1979. Indo Bukkyōshi インド仏教史 [History of Indian Buddhism], vol. 2. Tokyo: Shunjūsha 春秋社. Kuwayama Shōshin 桑山正進 and Hakamaya Noriaki 袴谷憲昭. 1981. Genjō 玄奘 [[[Xuanzang]]]. Tokyo: Daizō Shuppan 大蔵出版. Law, Bimala Churn. 1954. Historical Geography of Ancient India. Paris: Société Asiatique de Paris. Li Rongxi, tr. 1995. A Biography of the Tripiṭaka Master of the Great Ci’en Monastery of the Great Tang Dynasty. BDK English Tripiṭaka 77. Berkeley: Numata Center for Buddhist Translation and Research.

____ tr. 1996. The Great Tang Dynasty Record of the Western Regions. BDK English Tripiṭaka 79. Berkeley: Numata Center for Buddhist Translation and Research. Mizutani Shinjō 水谷真成, tr. 1999. Daitō saiiki ki 大唐西域記 [[[Datang xiyu ji]]], vol. 3. Tokyo: Heibonsha 平凡社. MSA(Bh): Mahāyānasūtrālaṃkāra(bhāṣya). F: Funahashi Naoya 舟橋尚哉, ed., Nepāru shahon taishō ni yoru Daijō shōgonkyōron no kenkyū ネパール写本対照による大乗荘厳経論の研究 [[[Mahāyānasūtrālaṃkāra]] (Chapters I, II,

III, IX, X) revised on the basis of Nepalese manuscripts], Tokyo: Kokusho Kankōkai 国書刊行会, 1985; L: Sylvain Lévi, ed. and tr., Mahāyāna-sūtrālaṃkāra: Exposé de la doctrine du Grand Véhicule selon le système Yogācāra, Paris, 1907. MSAṬ: *Mahāyānasūtrālaṃkāraṭīkā by Asvabhāva. Peking ed., no. 5530; Derge ed., no. 4029.

Nishio Kyōo 西尾京雄. 1940. Butchikyōron no kenkyū 佛地経論之研究 [A study of the Buddhabhūmi-vyākhyāna], 2 vols. Tokyo: Kokusho Kankōkai 国書刊行会. Sakuma Hidenori 佐久間秀範. 1982. “Gohō to sanshin no musubitsuki” 五法と三身の結びつき [On the connection between the five elements and the three bodies, with special reference to the Buddhabhūmivyākhyāna]. Indogaku Bukkyōgaku Kenkyū 印度学仏教学研究 31-1, pp. 124-125. ____ 1983. “Shichi to hasshiki no ketsugō kankei – sono seiritsu katei – ” 四智と八識の結合関係 – その成立過程 – [On the development of the connection between the four jñāna and eight vijñāna]. Indogaku Bukkyōgaku Kenkyū 32-1, pp. 178-179.

____ 1984. “‘Chi’ to ‘shiki’ – ryōsha no ketsugō kankei to sono seiritsu katei – ” ＜智＞と＜識＞ – 両者の結合関係とその成立過程 – [[[Jñāna]] and vijñāna: On the development of their interconnection]. Buzan Gakuhō 豊山学報 28/29, pp. 125-141. ____ 1987. “‘Sanshin’ to ‘gohō’ – ryōsha no ketsugō kankei to sono seiritsu katei” ＜三身＞と＜五法＞ – 両者の結合関係とその成立過程 [The five ele- ments and the three bodies: On the development of their interconnection]. In Takasaki Jikidō hakushi kanreki kinen ronshū: Indogaku Bukkyōgaku ronshū 高崎直道博士還暦記念論集・インド学仏教学論集 [Felicitation volume for Professor Jikidō Takasaki on the occasion of his sixtieth birthday], pp. 387411. Tokyo: Shunjūsha 春秋社. ____ 1989. “Genjō ni okeru ‘busshin’ no atsukaikata” 玄奘における＜仏身＞の扱い方 [Some aspects of Xuanzang’s treatment of Buddha-bodies]. Bukkyō Bunka 仏教文化 25, pp. 94-108. ____ 1992a. “Genkan shōgon ron daihasshō o meguru Indo shochūshakuka no bunrui – sanshinsetsu to shishinsetsu – ”『現観荘厳論』第八章をめぐるインド諸註釈家の分類 – 三身説と四身説 – [The classification of the “Dharmakāya Chapter” of the Abhisamayālaṃkāra by Indian commentators: The threefold and fourfold buddhakāya theories]. Shitennōji Kokusai Bukkyō Daigaku Bungakubu Kiyō 四天王寺国際仏教大学文学部紀要 24, pp. (1)- (30). ____ 1992b. “Genkan shōgon ron o meguru sanshinsetsu gurūpu ni yoru daiisshō daijūshichige kaihen no keii”『現観荘厳論』をめぐる三身説グループによる第一章第十七偈改変の経緯 [The detailed process of the modifica- tion of the 17th verse in the first chapter of the Abhisamayālaṃkāra by the “threefold buddhakāya theory” group]. In Mano Ryūkai hakushi shōju kinen ronshū: Hannyaharamitta shisō ronshū 真野龍海博士頌壽記念論集・般若波羅蜜多思想論集 [Felicitation volume for Professor Ryūkai Mano on the occasion of his seventieth birthday], pp. 183-194. Tokyo: Sankibō Busshorin 山喜房仏書林.

____ 1994. “The Classification of the Dharmakāya Chapter of the Abhisamayālaṃkāra by Indian Commentators: The Threefold and the Fourfold Buddhakāya Theories.” Journal of Indian Philosophy 22, pp. 259-297. ____ 1996. Sanskṛt Word-Index to the Abhidharmasamuccayabhāṣyam edited by N. Tatia with the Corrigenda. Tokyo: The Sankibo Press. ____ 2002. “Chūgoku, Nihon Hossō kyōgaku ni okeru shiki to chi no ketsugō kankei – fūin sareta dairokushiki → jōshosachi, gogenshiki → myōkanzatchi no seitōsei” 中国・日本法相教学における識と智の結合関係 – 封印された第六識→成所作智、五現識→妙観察智の正当性 [The connection be- tween the jñāna and vijñāna in the Fa-hsiang school in China and Japan: On the rightness of the sealed-up connections – sixth vijñāna → kṛtyānuṣṭhānajñāna and first five vijñāna → pratyavekṣājñāna]. In Kimura Kiyotaka hakushi kanreki kinen ronshū: Higashi Ajia Bukkyō to sono shūhen 木村清孝博士還歴記念論集・東アジア仏教とその周辺 [Felicitation volume for

Professor Kiyotaka Kimura on the occasion of his sixtieth birthday], pp. 6586. Tokyo: Shunjūsha 春秋社. ____ 2007a. “In Search of the Origins of the Five-Gotra System.” Journal of Indian and Buddhist Studies 54-3, pp. 1112-1120. ____ 2007b. “Yugashijiron ni mirareru jōbutsu no kanōsei no nai shujō”『瑜伽師地論』に見られる成仏の可能性のない衆生 [On sentient beings with no possibility of attaining Buddhahood mentioned in the Yogācārabhūmi]. Tetsugaku Shisō Ronshū 哲学・思想論集 (Institute of Philosophy, University of Tsukuba) 32, pp. 130-156. ____ 2007c. “Goshō kakubetsu no genryū o tazunete” 五姓各別の源流を訪ねて [In search of the origins of the five-gotra system]. In Katō Seiichi hakushi koki kinen ronbunshū: Shingon mikkyō to Nihon bunka 加藤精一博士古稀記念論文集 真言密教と日本文化 [[[Shingon Esoteric Buddhism]] and Japanese culture: In honour of Dr. Katō Seiichi on his seventieth birthday], vol. 2, pp. 265-305. Tokyo: Nonburusha ノンブル社.

SAVbh: *Sūtrālaṃkāra-vṛtti-bhāṣya by Sthiramati. Peking ed., no. 5531; Derge ed., no. 4034. Schmithausen, Lambert. 1969. Der Nirvāṇa-Abschnitt in der Viniścayasaṃgrahaṇī der Yogācārabhūmiḥ. Österreichische Akademie der Wissenschaften, Philos.-hist. Klasse, Sitzungsberichte, 264. Bd., 2. Abh. Wien: Österreichische Akademie der Wissenschaften.

Seizō Bunten Kenkyūkai 西蔵文典研究会. 1979. Seizō bunken ni yoru Bukkyō shisō kenkyū, daiichigō: Anne-zō “Daijō shōgonkyōron shakuso” – Bodaibon (I) 西蔵文献による仏教思想研究 第１号 安慧蔵『大乗荘厳経論釈疏』 – 菩提品 (I) [Studies in Buddhist thought based on Tibetan texts, 1: Sthiramati’s Sūtrālaṃkāra-vṛtti-bhāṣya, “Bodhi Chapter” (I)]. Tokyo: Sankibō Busshorin 山喜房佛書林.

Shizutani Masao 静谷正雄. 1979. Indo himei mokuroku インド碑銘目録 [A catalogue of Indian inscriptions]. Kyoto: Heirakuji Shoten 平楽寺書店. Tsukamoto Keishō 塚本啓祥. 1996. Indo Bukkyō himei no kenkyū インド仏教碑銘の研究 [A study of Indian Buddhist inscriptions], vol. I. Kyoto: Heirakuji Shoten 平楽寺書店. Tsukamoto Keishō 塚本啓祥, Matsunaga Yūkei 松長友慶 and Isoda Hirofumi 磯田煕文, eds. 1990. Bongo butten no kenkyū III: Ronsho hen 梵語仏典の研究 III 論書篇 [A descriptive catalogue of Sanskrit Buddhist literature, III: Treatises]. Kyoto: Heirakuji Shoten 平楽寺書店. TV: Sylvain Lévi, Vijñaptimātratāsiddhi, deux traités de Vasubandhu, Viṃśatikā accompagnée d’une explication en prose et Triṃśikā avec le commentaire de Sthiramati, Paris, 1925. Yamaguchi Susumu 山口益. 1966. Kanzō taishō Benchūbenron, fu Chūbenfunbetsuron shakuso bonpon sakuin 漢蔵対照 弁中辺論 附 中辺分別論釈 疏梵本索引 [The Madhyāntavibhāga in Chinese and Tibetan, with an index to the Sanskrit text of the Madhyāntavibhāgaṭīkā]. Tokyo: Suzuki Gakujutsu Zaidan 鈴木学術財団.