From criticism to approval: A reconsideration of Ji’s Yogācāra position on Madhyamaka

From criticism to approval: A reconsideration of Ji’s Yogācāra position on Madhyamaka

Sumi Lee

ABSTRACT

Madhyamaka and Yogācāra are two Mahāyāna schools which have distinct systems. In the seventh century East Asia, the doctrinal distinction between the two schools was received as doctrinal contrast in the polemic circumstance of Emptiness-Existence (C. kongyou 空有) controversy. In this context, Ji 基 (632–682), the putative founder of East Asian Yogācāra school, has been normally considered by scholars to have advocated ‘Existence’ (viz., Yogācāra) in opposition to ‘Emptiness’ (viz., Madhyamaka). It is problematic, however, to brand Ji’s Yogācāra position simply as anti-Madhyamaka. Although Ji evidently expresses evident criticism on such a Madhyamaka exegete as Bhāvaviveka (ca. 500–570) in some of his works, he also describes Bhāvaviveka in an amicable or even respective way in other works. By analyzing Ji’s extant works, this article argues that Ji’s scholastic attitude toward Madhyamaka changed from criticism to approval.

KEYWORDS

East Asian Buddhism; Madhyamaka; Yogācāra; Ji 基 (632-682); Emptiness-

Existence (kongyou 空有) controversy; Bhāvaviveka (ca. 500-570); Dharmapāla (ca. 6th century); the Mahāprajñāpāramitā-sūtra

Introduction

Madhyamaka and Yogācāra are distinct from each other in their doctrinal as well as soteriological systems, although they have much in common as the two representative schools of Mahāyāna Buddhism. At a certain historical point of East Asian Buddhist tradition, the distinction was received as doctrinal contrast between the two systems. It was in the seventh century, during which the famous pilgrim Xuanzang 玄奘 (602–664) introduced a new corpus of Buddhist canonical texts, that Madhyamaka and Yogācāra were considered to be doctrinally contradictory to each other. Ji 基 (632–682), Xuanzang’s major disciple as well as the putative founder of Dharma Characteristics school (C. Faxiang zong 法相宗), is one of the most representative figures in the polemic situation, which is known as Emptiness-Existence (C. kongyou 空有) controversy. It has been generally considered that Ji advocated the position of ‘Existence’ (viz., Yogācāra) in opposition to that of ‘Emptiness’ (viz., Madhyamaka), criticizing Madhyamaka exegete Bhāvaviveka (ca. 500–570; C. Qingbian 清辨/淸辯) on one hand and defending Yogācāra exegete Dharmapāla (ca. 6th century; C. Hufa 護法) on the other.

It seems too simplistic, however, to reduce Ji’s Yogācāra position into antiMadhyamaka within the dichotomized frame of Emptiness and Existence. Although Ji

evidently expresses criticism on Bhāvaviveka in some of his works, he describes Bhāvaviveka, along with Dharmapāla, in an amicable or even respective way in other works. In this respect, Ji’s inconsistent attitude toward Bhāvaviveka represented in his works requests further study. By analyzing Ji’s comments on Bhāvaviveka in his extant works, this article seeks to demonstrate that Ji’s scholastic view on Madhyamaka changed before and after Xuanzang’s translation of the Mahāprajñāpāramitā-sūtra, and explores doctrinal significance of the Mahāprajñāpāramitā-sūtra for Ji in connection with the two contemporary controversies, that is, the Emptiness-Existence controversy and the Buddha Nature controversy. In this article, I will first overview the polemic environment that led Ji to be engaged in the contemporary controversies. Then I will examine the change of Ji’s attitude on Bhāvaviveka by comparing the ways in which Ji describes Bhāvaviveka between his earlier and later works. Based on these, I will finally discuss theoretical significance of Ji’s intellectual shift on Madhyamaka in association with the two controversies.

Emergence of Buddha Nature controversy and Emptiness-Existence controversy

Ji’s criticism on Bhāvaviveka is usually understood in association with the doctrinal contrast between Madhyamaka and Yogācāra of East Asian tradition since he often puts Bhāvaviveka’s Madhyamaka view in opposition to Dharmapāla’s Yogācāra view. The contrast between Madhyamaka and Yogācāra in East Asia, namely, the EmptinessExistence controversy, is known to have been related to Xuanzang’s translation work that began in 645. As new canonical information was introduced through the translation, disagreements arose among the Buddhist exegetes over how to interpret various Buddhist doctrines. The Emptiness-Existence controversy was one of the two major controversies that then happened, along with the Buddha Nature controversy.

It was in this polemic environment that Ji entered priesthood. He became a novice monk under Xuanzang in 649 at the age of 17,2 the year in which Xuanzang translated Bhāvaviveka’s *Mahāyāna-hastaratna-śāstra (the Dasheng zhangzhen lun 大乘掌珍論; hereafter, Zhangshen lun). Which later became one of the key texts in the EmptinessExistence controversy. Just prior to this, in 648 and 649, respectively, Xuanzang translated two major Yogācāra texts, the Yogācārabhūmiśāstra (C. Yuqieshidi lun 瑜伽師地論) and the *Buddhabhūmi-śāstra (C. Fodijing lun 佛地經論). The doctrine of ‘Five Distinct Lineages’ (C. wozhong xing 五種性, S. pañcagotra) addressed in these two texts3 aroused resistance by the advocates of the notion of the universal Buddha Nature, and the tension between these two groups evolved into the Buddha Nature controversy. Ji began his career during this contentious time, in which the controversial issues emerged surrounding the doctrinal distinctions between the ‘old’ and ‘new’ translations of Buddhist texts. He must have thus witnessed the ongoing disputes right at the center of the controversies. This polemic situation then may well have influenced Ji throughout the process in which Ji formed his scholastic views and position as a Yogācāra exegete. It would be worthy, in this respect, to examine the contemporary situation of the two controversies before moving to analyze Ji’s works.

The theory of Five Distinct Lineages described in such texts as the

Yogācārabhūmiśāstra and the Fodijing lun was regarded as doctrinally opposed to the notion of universal Buddha Nature, which was represented in the previously translated scriptures, such as the Nirvān:ṇa Sūtra and the Lotus Sūtra. In particular, the notion of ‘Lineage devoid of the Nature’ (S. agotra, C. wu zhongxing 無種性), the fifth among the five lineages, appeared to diametrically conflict to the theory of universal Buddha Nature of the Nirvān:ṇa Sūtra. Furthermore, the idea of discriminative lineages was taken as contradictory to the One Vehicle theory of the Lotus Sūtra, because it was doctrinally connected to the Three Vehicles theory.4 The task given in this polemic milieu to the contemporary Buddhist thinkers, whether they were on the side of the group who agreed for the Five Distinct Lineages or the universal Buddha Nature, was probably to explain and resolve the seeming contradiction represented in the Buddhist literature. However, as easily expected, the efforts to resolve the doctrinal contradiction were not always peaceful. The exegetes of each group tended to interpret the other group’s position from their own perspectives. The Japanese Tendai monk Saichō 最澄 (767–822) describes in the Hokkeshūku 法華秀句, dated around 650, that Lingrun 靈潤 (fl. 650) and Shentai 神泰 (fl. 645, 657) had a dispute,5 presumably at some time between 648 and 650.6 Lingrun argues that the doctrine of ‘Lineage devoid of the Nature’ is just an incomplete teaching for ordinary beings or the Inferior Vehicle (S. hīnayāna, C. xiaosheng 小乘). He also says that the realm of sentient beings is equal to that of tathāgatagarbha on basis of the canonical authority of the Nirvān:ṇa Sūtra and other tathāgatagarbha texts such as the Ratnagotravibhāga. On the same line, he also criticizes the Yogācāra exegetes’ division of Buddha Nature into two types, that is, ‘Buddha Nature in Principle’ (C. li foxing 理佛性), which is inherent in all sentient beings, and ‘Buddha Nature in Practice’ (C. xing foxing 行佛性), which is owned by only some sentient beings, by arguing that the Buddha Nature cannot be divided.7 On the contrary, Shentai contends that not all sentient beings can become Buddhas relying on the Yogācārabhūmiśāstra as the canonical basis; he distinguishes ‘Buddha Nature in Principle’ former ‘Buddha Nature in Practice’ saying that although the first Buddha Nature, namely, the True Thusness (S. tathatā, C. zhenru 眞如), is inherent in all sentient beings, the second, that is, ‘Seeds of the Great Vehicle’ (C. dasheng zhongzi 大乘種子), is not intrinsic in all sentient beings.8 The controversy between Lingrun and Shentai may be said a typical example of the Buddha Nature controversy.

While the Buddha Nature controversy was ongoing, Xuanzang translated Bhāvaviveka’s *Mahāyāna-hastaratna-śāstra into the Zhangzhen lun (649), a verse of which worked as a trigger of the Emptiness-Existence controversy. Bhāvaviveka’s verse at the beginning of this work drew the contemporary Buddhist scholiasts’ attentions especially when Xuanzang translated in the very next year the Dasheng guangbailun shilun 大乘廣百論釋論 (hereafter, Guangbai shilun), Dharmapāla’s commentary on Āryadeva’s Catuhḥ: śataka-śāstra-kārikā (C. Guangbai lun 廣百論). According to the verse of the Zhangzhen lun, all dharmas, whether the Conditioned (C. youwei 有爲, S. samṃ: skrṛ: ta) or the Unconditioned (C. wuwei 無爲, S. asamṃ: skrṛ: ta), are empty in terms of True Nature (C. zhenxing 眞性, S. tattva), viz., the Ultimate Level of Reality (C. shengyi di 勝義諦, S. paramārtha-satya).9 When the Guangbai shilun was translated, Bhāvaviveka’s verse was considered to imply that all Three Natures (C. sanxing 三性, S. tri-svabhāva) are empty, taking the opposed position to Dharmapāla’s statement that only Imaginary Nature (C. pianji suozhi xing 遍計所執性, S. parikalpita-svabhāva) among the Three Natures is empty.10 Further, the Zhangzhen lun turned out to contain criticism on the ‘Exegete of Association’ (C. xiangying lunshi 相應論師), who appears in the Zhangzhen lun to refute such exegetes as Bhāvaviveka since he claims that those who dismiss not only the Imaginary Nature but also the Dependent Nature (C. yita qi xing 依他起性, S. paratantra-svabhāva) would be fallen into evil destinies (C. equ 惡趣).11 The Exegete of Association thus came to be regarded as Dharmapāla and its followers.12 The Bodi jing lun, already translated in 649, also mentions the controversy between Emptiness and Existence, which occurred a thousand years after the Buddha had taught the Dharma.13 Thus, there must have been pervasive acknowledgement on the controversy between Madhyamaka and Yogācāra even before the translation of the Zhangzhen lun and the Guangbai shilun. Although whether or not the controversy between the two Indian exegetes really happened as a historical fact was continuously questioned afterwards,14 such early acknowledgement of the controversy probably led the East Asian Buddhist thinkers to conceive the contemporary disputes as fundamental and inescapable doctrinal issues between Madhyamaka and Yogācāra.

It was in this polemic milieu that Ji spent time as a novice monk, and this may explain at least partly, if not fully, Ji’s strong criticism on Bhāvaviveka. For Ji, an enthusiastic young Yogācāra monk who was confident about doctrinal authenticity of the newly imported Yogācāra texts, any position that seemed opposed to the Yogācāra doctrines was probably his target of criticism. Ji had to counter Bhāvaviveka’s view of the Zhangzhen lun as well as the doctrine of universal Buddha Nature to defend his Yogācāra position. However, as I mentioned at the beginning, Ji’s critical attitude on Bhāvaviveka cannot be seen in his later works. In other words, the extant works that exhibit Ji’s strong criticism on Bhāvaviveka are presumed to have been composed in his earlier years, during which Ji worked as a committed young Yogācāra monk in the controversial situation. In the following section, I will explore Ji’s earlier works, such as the commentaries on the Cheng weishi lun 成唯識論, which represent Ji’s antagonistic sentiments to his opponents in the two controversies. Ji’s response to the two controversies in the earlier times

After entering the priesthood, Ji was first called on by Xuanzang to serve for the translation work at his age of 25.15 According to Zhisheng’s 智昇 (669–740) Kaiyuan shijiao lu 開元釋教錄, Buddhist catalogue of Buddhist scriptures and treatises, Ji worked as Xuanzang’s amanuensis (C. bishou 筆受) during the translation of the six texts,16 the Cheng weishi lun being the first text that Ji worked on. The fact that Ji worked on the Cheng weishi lun first presumably has more implication than it seems. There is a wellknown record in the Chengweishilun shuyao 成唯識論樞要 (hereafter, Weishilun shuyao) that implies Ji’s special role in translating the Chen weishi lun.17 Xuanzang originally planned to translate all 10 Indian exegetes’ commentaries but changed his mind on Ji’s suggestion that the translation should be compiled centering on Dharmapāla’s views. Moreover, Xuanzang discharged three other monks, Singbang 神昉 (fl. 645–651, 659), Jiashang 嘉尙 (fl. 659), and Puguang 普光 (645–664), who had been supposed to participate in the translation, and proceeded the translation only with Ji. In addition, it has been noted that the time when Xuanzang translated the Chen weishi lun seems odd for the general timeline of his translation work; Xaunzang translated Yogācāra texts during the first 5 years since he had returned in 645, but the Cheng weishi lun, the most representative work in East Asian Yogācāra school, was translated in 659 out of this pattern.18 This suggests that the translation of the Cheng weishi lun might not have originally planned by Xuanzang himself. Based on these aspects, it seems very likely that Ji’s role in the translation of the Cheng weishi lun was at least more than that of a mere assistant of Xuanzang.

Another aspect that may be noted as regards to Ji’s role in the translation of the Cheng weishi lun is that Ji suggested that the Cheng weishi lun should be an edited compilation centering on Dharmapāla’s view because he was concerned about the spread of controversies in case all 10 exegeses were translated.19 Given that the controversies surrounding the new translations were the ongoing issue, Ji’s suggestion that the Cheng weishi lun should be the compiled translation was probably based on his idea that that way of translation would be the most appropriate for the current situation. For Ji, Dharmapāla’s exegesis must have seemed most proper and effective in defending the

Yogācāra views as well as disproving his rival exegetes’ views in the two controversies.20

In fact, the Cheng weishi lun contains two significant doctrines that Ji adopted to respond to the rival views in the two controversies. First, the doctrine of ‘Uncontaminated Seeds’ (C. wulou zhongzi 無漏種子), that is, the spiritual base of sentient beings to become a Buddha, is usually regarded as providing a theoretical basis for the doctrine of Five Distinct Lineages of the Yogācāra school, because the Uncotaminated Seeds are explained in five types in the Cheng weishi lun, thereby conforming to the five types of Lineage.21 In this way, the doctrine worked for Ji as the canonical evidence for the Yogācāra doctrine of Five Distinct Lineages on one hand and as counterevidence to the universal Buddha Nature theory on the other. Second, the Cheng weishi lun contains criticism on ‘those who take the Mahāyāna teaching of emptiness of no characteristics as the ultimate [[[truth]]]’ (C. youzhi dasheng qianxiang kongli wei jiujing zhe 有執大乘遣相空理為究竟者).22 In his commentary on the Cheng weishi lun, Ji interprets this phrase as aimed at ‘Madhyamaka of No Characteristics’ (C. wuxiang dasheng 無相大乘), that is, Bhāvaviveka’s doctrinal position.23 However, of course, there is no mentioning on such a particular figure as Bhāvaviveka in the Cheng weishi lun.

Ji’s adversarial attitude toward his opponents of the controversies clearly appears in his commentaries on the Cheng weishi lun. The Cheng weishi lun shouji 成唯識論述記 (hereafter, Weishilun shouji) and the Weishilun shuyao, or at least the first drafts of them, were presumably written almost simultaneously with the translation of the Cheng weishi lun.24 Then it may be said that these two works reflect Ji’s mindset when he was engaged in the translation work of the Cheng weishi lun. In fact, in these works Ji not only defends the Yogācāra doctrine of Five Distinct Lineages but also refutes the opponents, as follows. In the Weishilun shuji, for instance, Ji seeks to prove the doctrine of Five Distinct Lineages based on several sources, such as the Yogâcārabhūmi, the Lanṅ: kâvatāra-sūtra (C. Ru lengqie jing 入楞伽經), the *Anuttarâśraya-sūtra (C. Wushangyi jing 無上依經), the Mahāprajñāpāramitā-sūtra, and the Sūtrâlamṃ: kāra-śāstra (C. Da zhuangyan lun大莊嚴論).25 In the Weishilun shuyao, Ji particularly articulates the fifth group of the Five Lineages, ‘sentient beings devoid of spiritual lineage’ (S. agotra, C. wu zhongxing 無種性), i.e., ‘Incorrigibles’ (S. icchantika, C. yichanti 一闡提), by dividing it into three categories, that is, yichandijia 一闡底迦 (S. icchantika), achandijia 阿闡底迦 (S. ācchantika/anicchantika), and adiandijia 阿顛底迦 (S. ātyantika). The first group of sentient beings refers to those who enjoy samṃ: sāra and thus cut off all wholesome roots (C. duanshan chati 斷善闡提); the second to the bodhisattvas who choose not to enter nirvān:ṇa to save sentient beings and are also known as ‘greatly compassionate icchantika’ (C. dabei chanti 大悲闡提); and the third to those who do not have any lineage and never become a buddha (C. bijingwuxing chanti 畢竟無性闡提).26 Here, Ji clearly addresses the existence of the sentient beings who never attain Buddhahood as the third type of icchantica, thereby advocating the doctrine of Five Distinct Lineages.

Meanwhile, Ji refutes Bhāvaviveka’s view by drawing upon some passages of the Cheng weishi lun. Although the Cheng weishi lun contains no direct criticism on Bhāvaviveka, as mentioned above, Ji interprets some passages of the Cheng weishi lun as aimed at Bhāvaviveka. The Cheng weishi lun expresses criticism on ‘those who take the Mahāyāna teaching of emptiness of no characteristics as the ultimate [[[truth]]]’ for their ‘negating the [fundamental] consciousness and all dharmas relying on False Inference (S. anumānābhāsa, C. si biliang 似比量)’27; in commenting this passage, Ji regards this position as corresponding to Bhāvaviveka’s.28 In another passage, the Cheng weishi lun states that the Middle Way of Yogācāra is free from ‘two extreme views of increasing and decreasing’ (C. zengjian erbian 增減二邊),29 and Ji explains the view of ‘decreasing’ as Bhāvaviveka’s view of emptiness.30

The Middle Way of Yogācāra teaching, which Ji argues is the orthodox teaching of the Cheng weishi lun, is explained in the Cheng weishi lun through Maitreya’s verse of the Madhyānta-vibhāga.31 In explaining this verse, Ji distinguishes the Middle Way from the one-sided view of Bhāvaviveka or the hīnayāna; he criticizes Bhāvaviveka and the hīnayāna for adhering respectively to the side of Emptiness and the hīnayāna as only to the side of Existence.32 In a later passage, the Cheng weishi lun states that those who negate both ‘Ultimate Truth’ (S. paramārtha-satya, C. zhendi 眞諦) and ‘Conventional Truth’ (S. saṃvṛti-satya, C. sudi 俗諦) are incurable advocates of ‘Wrongly Attached Emptiness’ (C. equ kong 惡取空)33; Ji explains this position as refering to Bhāvaviveka’s. Then Ji continues to criticize Bhāvaviveka for taking all Three Natures as empty by saying that in Yogācāra position Imaginary Nature does not exist and yet Dependent Nature and Perfected Nature exist.34 Given that Ji’s hostile attitude in the commentaries reflects his doctrinal standpoint around the time when the Cheng weishi lun was translated, we may say that Ji had the intention of somehow using the Cheng weishi lun to respond to the two ongoing controversies when he suggested Xuanzang the compiled version of the translation.

Besides the commentaries of the Cheng weishi lun, some other works of Ji’s also contain strong criticism on Bhāvaviveka. In 661, that is, 2 years after the translation of the Cheng weishi lun, Xuanzang translated the Madhyānta-vibhāga (C. Bian zhongbian lun 辨中邊論) and the Vimṃ: śatikā-vijñaptimātratāsiddhi-śāstra (C. Weishi ershi lun 唯識二十論),35 and Ji’s commentaries on these works, the Bian zhongbianlun shuji 辨中邊論述記 and the Weishi ershilun shuji 唯識二十論述記, also represent the criticism. In the Bian zhongbianlun shuji, Ji refers to Bhāvaviveka as a heretic with a [wrong] view of emptiness that dharmas have no true reality.36 Likewise, in the Weishi ershilun shuji Ji criticizes Bhāvaviveka’s verse of the Zhangzhen lun by describing it as False Inference.37 I have discussed above that Ji assumed antagonistic attitude in the controversial environment around 659. In fact, the Bian zhongbianlun shuji is presumed to have been composed in 661, right after Xuanzang’s translation of the Madhyānta-vibhāga38. The Weishi ershilun shuji also appears to be drafted, if not completed in the current form, around this time.39 If we accept that these two works were composed around 661, then Ji’s criticism represented in these works may well be regarded on the same line with his hostile attitude displayed in the commentaries of the Cheng weishi lun.

In the Zajilun shuji 雜集論述記,40 Ji’s commentary on Sthiramati’s Commentary on Asanṅ: ga’s Abhidharmasamuccaya, Ji also shows critical attitude toward Bhāvaviveka by making a sharp contrast between Nāgārjuna/Bhāvaviveka and Maitreya/Vasubandhu. At the beginning of the text, for instance, Ji says that Bhāvaviveka formed a faction relying on Nāgārjuna and that Bhāvaviveka’s views, despite being a Mahāyāna teaching, does not represent the true meaning.41 The Zajilun shuji also appears to have been composed not long after the translation of the Cheng weishi lun. It seems that it was composed at the latest before 663, the year when the Mahāprajñāpāramitā-sūtra 大般若波羅蜜經 was translated, because in this work Ji makes a mistake in citing the Mahāprajñāpāramitāsūtra; Ji cites in this work a passage from the Liqufen 理趣分 (‘Section on Maxim of Prajñā’), the 578th fascicle of the Mahāprajñāpāramitā-sūtra, but this passage in fact does not appear in this fascicle, but in others.42 This strongly suggests that the Mahāprajñāpāramitā-sūtra had not been translated yet until Ji was composing the Zajilun shuji, although he may have been aware of the passage of the Liqufen.

To sum, Ji’s works composed during the time between 659 and 663 appear to contain his antagonism against Bhāvaviveka. This antagonism may be said as part of Ji’s response to the opponents of the two ongoing controversies. Since around the time when Ji took part in the translation of the Cheng weishi lun, he began to express his sentiments through his commentaries of the Cheng weishi lun and other works, such as the Bian zhongbianlun shuji, the Weishi ershilun shuji, and the Zajilun shuji. It is in this polemic context that Ji’s Yogācāra view is often interpreted in opposition to Madhyamaka, particularly Bhāvaviveka’s, view, and this scholastic image of Ji has been established as the standard model of East Asian Yogācāra. Against the backdrop of the sharp confrontation between the two groups in the Emptiness-Existence controversy, Ji’s direct criticism on Bhāvaviveka must have been evident enough to conceive him as the most representative Yogācāra exegete. However, this typical characterization on Ji appears to need more consideration, because Ji’s criticism, as mentioned before, remarkably diminishes in his later works, particularly in his works composed after the translation of the Mahāprajñāpāramitā-sūtra. The following sections will discuss why and in what way Ji changed his attitude to his doctrinal rivals.

Ji’s response to the two controversies in the later times

In the above section, I have discussed that among Ji’s works the commentaries of the Cheng weishi lun, the Bian zhongbianlun shuji, the Weishi ershilun shuji, and the Zajilun shuji, all dated before 663, contain his strong criticism on Bhāvaviveka and that this criticism has been generally taken as his Yogācāra position on Madhyamaka. In this section, I will discuss on Ji’s comments on Bhāvaviveka in his later works, presumably composed after 663, such as the Dabore boluomiduojing bore liqufen shuzan 大般若波羅蜜多經般若理趣分述讚 (hereafter, Liqufen shuzan), the Bore boluomi duoxin jing youzan 般若波羅蜜多心經幽贊 (hereafter, Xinjing youzan), the Dasheng fayuan yilin zhang 大乘法苑義林章 (hereafter, Yilin zhang), and the Yinming ru zhengli lun shu 因明入正理論疏 (hereafter, Yinminglun shu). Unlike the works discussed in the last section, these works do not contain any hostile expressions toward Bhāvaviveka, such as the False Inference or the Wrongly Attached Emptiness; rather Ji juxtaposes Bhāvaviveka and Dharmapāla, or Madhyamaka and Yogācāra without describing any doctrinal contradiction between them. In these works, Ji takes a moderate, if not completely sympathetic, attitude to Bhāvaviveka, sometimes even paying respect to him. The distinction in Ji’s attitude on Bhāvaviveka among Ji’s extant works needs further research, especially when considered that Ji’s Yogācāra position has been typically interpreted as opposed to Madhyamaka on the basis of his criticism on Bhāvaviveka. Let us first examine the way in which Ji addresses Bhāvaviveka in these texts.

In the Liqufen shuzan, the commentary on the Liqufen of the Mahāprajñāpāramitāsūtra, Ji shows respect to Bhāvaviveka by not only referring to Bhāvaviveka as the ‘true worthy one’ (viz. arhat; C. yingzhen dashi 應真大士) whose mind stays within the Buddha’s teaching of principle (C. lizong 理宗) but also noting that he is extolled as Bodhisattva Mañjuśrī.43 Further, this passage is the equivalent part of that of the Zajilun shuji, which I have discussed above,44 and we see that Ji’s tone in describing Bhāvaviveka in this text is very different from that of the Zajilun shuji. In the Zajilun shuji, as discussed above, Ji takes a critical attitude to Bhāvaviveka, depicting him as having formed a faction relying on Nāgārjuna,45 but here in the Liqufen shuzan he says that Bhāvaviveka had no attachment to a faction.46

Moreover, in another passage of the Liqufen shuzan, in which Bhāvaviveka and Dharmapāla’s views on emptiness are explained, Ji describes that the two exegetes both explain all things as empty47; In Bhāvaviveka’s view, all things are empty in terms of Ultimate Truth (C. shengyi 勝義諦, S. paramārtha-satya), while in Dharmapāla’s position all Three Natures may be said as empty as well. In another place, Ji says that Bhāvaviveka and Dharmapāla’s interpretations even accord to each other in terms of Superior Purity (C. zuisheng qingjing 最勝淸淨) of the wisdom.48 I have examined in the last section that in such works as the commentaries of the Cheng weishi lun, Ji argues for Dharmapāla’s view that only the Imaginary Nature, not all Three Natures, is empty, while refuting Bhāvaviveka’s view that all the Three Natures are empty.49 Ji’s statements here in the Liqufen shuzan then obviously does not conform to the position represented in the commentaries of the Cheng weishi lun; rather the two positions seem contradictory to each other.

One way to explain the inconsistency in Ji’s works in his attitude to Bhāvaviveka may be to prove that Ji transformed his doctrinal position over his career. Evidently, the Liqufen shuzan was written in 663 at the earliest, as this work is the commentary on a fascicle of the Mahāprajñāpāramitā-sūtra, while the works that exhibit the criticism were presumably all written before 663, as discussed in the last section. In fact, other Ji’s works, in which Ji mentions Bhāvaviveka and yet does not express any harsh criticism on him turn out to have been written, or at least first drafted, after 663, as I will discuss below. In the Xinjing youzan, the commentary on the Heart sutra, Ji mentions Bhāvaviveka, but with no criticism. Like the Liqufen shuzan, the Xinjing youzan appears to have been composed after the translation of the Mahāprajñāpāramitā-sūtra in 663 because it cites the Liqufen.50 Throughout this work, Madhyamaka and Yogācāra, respectively, referred to as ‘those who are excellent in emptiness’ (C. shengkong zhe 勝空者) and ‘those who conform to [circumstances]’ (C. ruying zhe 如應者), are paralleled to each other with no mutual contradiction addressed. Ji says, for instance, that the ‘Saint Nāgārjuna’ (C. sheng Lengmeng 聖龍猛) attempted to eliminate the attachment to Existence and promoted

the teaching of Emptiness (C. kongzong 空宗), while Asanṅga and Maitreya established the teaching of the Middle Way (C. zhongdaojiao 中道敎) to remove both attachments to Existence and Non-existence; it is in this context that Ji places Bhāvaviveka’s verse of the Zhangzhen lun in juxtaposition with Maitreya’s verse of the Bian zhongbian lun,51 the two key verses in the Emptiness-Existence controversy.

In another passage of this work, Ji explains the emptiness from the perspectives of Madhyamaka and Yogācāra. Ji explains both Madhyamaka and Yogācāra as accepting the emptiness; Ji says, ‘those who are excellent in emptiness,’ viz., Madhyamaka, say that if one realizes the principle of Ultimate Truth, then he ultimately illuminates emptiness of natures, while ‘those who conform to [circumstances],’ viz., Yogācāra, say that all Three Natures may be said empty as the Buddha’s concealed intention (C. miyi 密意) to destroy the attachment to Existence, although they are in fact neither empty nor nonempty.52 This statement by Ji reminds us of the passage of the Liquefen shuzan examined above,53 in which Bhāvaviveka and Dharmapāla are said to accord to each other in that they both explain all things as empty. In this passage of the Xinjing youzan, both Madhyamaka and Yogācāra are similarly said to accept emptiness, although Yogācāra’s acceptance of all Three Natures as empty are explained as the Buddha’s concealed intention. Ji’s attitude to compromise Bhāvaviveka and Dharmapāla, or Madhyamaka and Yogācāra, represented in these two works, both dated after 663, is a remarkable change from that of his earlier works composed before 663. This change, again, suggests that Ji went through an intellectual shift around the year of 663.

In the Yilin zhang, the encyclopedic work on Mahāyāna doctrines, Ji also explains Madhyamaka and Yogācāra in a compromising way.54 He describes the two Mahāyāna traditions both as ‘[Teachings of] own master’ (C. zishu 自主), and, on the contrary, refers to non-Buddhists (C. waidao 外道) and hīnayāna as ‘different [viz., non-Buddhist] teachings’ (C. yizong 異宗). Then Ji depicts Bhāvaviveka, who considers all dharmas as empty, as ‘Master of the Border’ (C. bianzhu 邊主) and Dharmapāla, who finds the Middle Way, as ‘Master of the Center’ (C. zhongzhu中主). Here, again, we cannot see any apparent criticism on Bhāvaviveka, even if Bhāvaviveka and Dharmapāla are placed in different doctrinal states, namely, the Master of the Border and the Center. Rather Ji evidently considers Bhāvaviveka as a teacher of Buddhism along with Dharmapāla. Like the Liqufen shuzan and the Xinjing youzan, the Yilin zhang is presumed to have been drafted after 663 since it contains a citation of a passage from the Liqufen as well.55 Besides the three works above, the Yinminglun shu, Ji’s commentary on the Yinming ruzhengli lun 因明入正理論 (S. Nyāyapraveśa), mentions Bhāvaviveka once with no criticism to him; in this passage Bhāvaviveka is referred to as ‘bodhisattva’ in a respective way.56 Since a phrase from the Mahāprajñāpāramitā-sūtra is also cited,57 this work also seems to have been written after 663.

One thing that should be noted in the current discussion is that although Ji’s attitude toward Bhāvaviveka changed from harsh criticism to moderate approval, this does not mean that Ji came to conceive Bhāvaviveka’s doctrinal views as equally significant as Dharmapāla’s. Although Ji compromised Bhāvaviveka’s Madhyamaka position with Dharmapāla’s Yogācāra view in his later years, Ji never treated Madhyamaka with the same level of the teaching as Yogācāra throughout his career. Such hierarchical perspective on Madhyamaka and Yogācāra is represented not only in his criticism on Bhāvaviveka in his works dated before 663 but also in his doctrinal taxonomies in the works after 663. In the Yilin zhang, for instance, Ji places Madhyamaka and Yogācāra, respectively, on the second and the third/highest level of teaching in the taxonomy of Three-period Teachings (C. sanshi jiao 三時敎); Madhyamaka and Yogācāra, referred to, respectively, as ‘Implicit Meaning’ (C. buliao yi 不了義, S. neyārtha) and ‘Definitive Meaning’ (C. liaoyi 了義, S. nītārtha), have the implication that the former is the lower level than the latter.58 Likewise, in the Shuo wugoucheng jing shu 說無垢稱經疏 (hereafter, Wugou shu) and the Miaofa lianhuajing xuanzan 妙法蓮華經玄贊 (hereafter, Fahua xuanzan), the respective commentary of the Vimalakīrtinirdeśa-sūtra and the Lotus sūtra, written after 663,59 Ji locates Madhyamaka and Yogācāra, respectively, at the seventh and eighth of the eight levels of taxonomy of Buddhist teaching.60 Yet, just as in other works after 663, Ji expresses no doctrinal contrast or contradiction between Madhyamaka and Yogācāra in these hierarchical taxonomies. This shows that in later works Ji embraces not only Dharmapāla’s Yogācāra but also Bhāvaviveka’s Madhyamaka in the one system of Buddhist teaching, while considering them as two distinct levels of the teaching. However, again, Ji never denied the higher doctrinal significance that Yogācāra had over Madhyamaka.

So far, I have discussed how Ji’s attitude toward Bhāvaviveka changed by examining Ji’s comments on Bhāvaviveka in his extant works from the perspective of the Emptiness-Existence controversy. The change of Ji’s attitude may be noticed from the perspective of the Buddha Nature controversy as well. In the works after 663, Ji relied on the so-called tathāgatagarbha scriptures, such as the Śrīmālādevī-sūtra; he had not relied on the tathāgatagarbha scriptures before 663 even when commenting the same content of the same context. For instance, ‘essence of all characteristics returning to the nature’ (C. shexiang guixingti 攝相歸性體)61 is one of the notions that appear in Ji’s works throughout his career, but the scriptures that Ji cites as its canonical evidence are not identical in the works. In such works as the Weishilun shuji and the Zajilun shuji, dated before 663, Ji quotes a phrase as canonical support for this notion only from the Vimalakīrtinirdeśa (C. Shuo wugoucheng jing 說無垢稱經), that is, ‘all sentient beings are all Thusness; all dharmas are also Thusness.’62 However, in the Liqufen shuzan, the Yilin zhang, and the Wuguo shu, written after 663, Ji includes more scriptural evidence in commenting the equivalent passage. All three of these works quote a phrase ‘all sentient beings are tathāgatagarbha (/Thusness)’ from the Liqufen as well as the phrase from the Vimalakīrtinirdeśa63; the Yilin zhang and the Wuguo shu quote another phrase, ‘those who experience birth and death are tathāgatagarbha,’ from the Śrīmālādevī-sūtra in addition to those of the Vimalakīrtinirdeśa and the Liqifen.64 In other words, Ji introduces the Śrīmālādevī-sūtra as the canonical support for the notion ‘essence of all characteristics returning to the nature’ only in his later works, such as the Yilin zhang and the Wuguo shu.

One thing that is noteworthy in this regard is that Xuanzang never translated the Śrīmālādevī-sūtra; rather, the concept of tathāgatagarbha of this sūtra was used as a canonical evidence for the doctrine of universal Buddha Nature in the Buddha Nature controversy. We may imagine that Ji was probably reluctant to use the sūtra before 663 when the controversies were intense, and probably it is probably the reason why the works dated before 663 do not contain the quotation from the sūtra. Ji’s introduction of the Śrīmālādevī-sūtra, however, suggests that Ji’s intellectual shift led him to accept such a so-called tathāgatagarbha text as the Śrīmālādevī-sūtra as the canonical source.65

Further, it was after 663 that Ji noted the Lotus sūtra and composed its commentary, that is, the Fahua xuanzan.66 In the polemic situation, the Lotus sūtra was one of the canonical basis for Ji’s opponents in the Buddha Nature controversy; One Vehicle theory of the Lotus sūtra was regarded as doctrinally contradictory to the discriminative idea of Five Distinct Lineages.67 Just like the Śrīmālādevī-sūtra, this scripture was never translated by Xuanzang. Nevertheless, Ji composed the commentary of this sūtra, and this strongly suggests that Ji began to embrace the One Vehicle position in his later years. Now the question would be why Ji changed his attitude on Bhāvaviveka’s Madhyamaka view and the notion of universal Buddha Nature, or the tathāgatagarbha theory, around the time of the translation of the Mahāprajñāpāramitā-sūtra. Was Ji influenced by the Mahāprajñāpāramitā-sūtra, and, if then, in what way would it be? In the following section, I will explore the connection between the change of Ji’s doctrinal position and the Mahāprajñāpāramitā-sūtra and then discuss the doctrinal significance of Ji’s intellectual transition in association with the two contemporary controversies. Translation of the Mahāprajñāpāramitā-sūtra and its influence on Ji

The fact that Ji changed his attitude to Bhāvaviveka before and after the translation of the Mahāprajñāpāramitā-sūtra suggests that this shift of Ji’s doctrinal standpoint was somehow associated to the translation of the Mahāprajñāpāramitā-sūtra. What should be noted in this respect is that all the later works, with the only exception of the Yinminglun shu,68 quote the particular phrase ‘all sentient beings are tathāgatagarbha’ from the Liqufen of the Mahāprajñāpāramitā-sūtra.69 This implies that the later works share a certain doctrinal point in connection to the phrase of the Liqufen. In fact, among the 500 fascicles of the Mahāprajñāpāramitā-sūtra, the Liqufen is the only fascicle on which Ji wrote a commentary. Then it is very likely that the phrase of the Liqufen exerted a crucial influence on Ji’s intellectual transformation.

How or in what way was Ji influenced by the Liqufen especially by the phrase? At first glance, this phrase ‘all sentient beings are tathāgatagarbha’ does not seem consonant to the typical Yogācāra doctrine of Five Distinct Lineages; rather it appears to defend the doctrine of universal Buddha Nature, which Ji’s opponents advocated in the Buddha Nature controversy. It was not the case, however, that the group who advocated the doctrine of Five Distinct Lineages completely negated the notion of universal Buddha Nature. As discussed before, the seeming contradiction between the universal Buddha Nature and the Five Distinct Lineages was generally resolved in the Yogācāra by this group using the theory of two types of Buddha Nature, that is, Buddha Nature in Principle and Buddha Nature in Practice. In accordance with the theory of two types of Buddha Nature, Ji also explains the Buddha Nature in two types in the Fahua xuanzan; the Buddha Nature in Principle is inherent in all sentient beings, while the Buddha Nature in Practice is owned by only some sentient beings.70 Just as such a Yogācāra exegete as Shentai did during the Buddha Nature controversy, Ji also resolved the contradiction between the universality and particularity of the Buddha Nature by attributing the former to the principal level and the latter to the practical level. Then Ji continues to say that although tathāgatagarbha is said to exist [in all sentient beings], there is no teaching that all sentient beings attain Buddhahood and that the discrimination of the lineages is based on the particularity of Buddha Nature in Practice.71 Here, we clearly see that Ji places the universal Buddha Nature, that is, tathāgatagarbha, just in the level of Principle.

Ji’s understanding of tathāgatagarbha as corresponding to in terms of Principle may be also examined in his comment in the Liqufen shuzan on the relationship of tathāgatagarbha and Thusness. Explaining the phrase ‘all sentient beings are tathāgatagarbha’ of the Liqufen, Ji says that tathāgatagarbha is the name of Thusness in fettered state (C. zaichan 在纏) and Thusness in fettered state is named as such because all sentient beings have Principle of Thusness (C. zhen[ru]li 真[如]理).72 This statement reveals that Ji identifies tathāgatagarbha owned by all sentient beings with Thusness in the level of Principle. Moreover, in the Wuguo shu, Ji quotes not only the above phrase of the Liqufen but also other similar phrases from such scriptures as the Śrīmālādevī-sūtra and the Vimalakīrtinirdeśa while relating them to Thusness; Ji quotes ‘those who experience birth and death are tathāgatagarbha’ from the Śrīmālādevī-sūtra and ‘all sentient beings are Thusness; all dharmas are also Thusness’ from the Vimalakīrtinirdeśa to explain that the essence of all dharmas are Thusness.73 Just as Ji does in the Liqufen shuzan above, in the Wuguo shu Ji explains the similar phrases, in which all sentient beings and/or dharmas are referred to as tathāgatagarbha and/or Thusness, in terms of the Principle.

I discussed above that in the later works Ji began to accept tathāgatagarbha scriptures, such as the Śrīmālādevī-sūtra, as canonical support to explain the doctrinal notion of ‘essence of all characteristics returning to the nature.’ Ji’s embracement of the tathāgatagarbha texts also may be related to his explanation of these phrases, in which all dharmas/sentient beings are said Thusness/tathāgatagarbha, in terms of Thusness. Ji was able to embrace the notion of universal Buddha Nature (viz., tathāgatagarbha) because he placed it in the sphere of Principle of Thusness. In other words, Ji reconciled the otherwise inconsistent two doctrines of the universal Buddha Nature and the Five Distinct Lineages by assigning them to the two separate spheres, Principle (C. li 理) and Practical (C. xing 行).

Although the notion of two types of Buddha Nature was noted by some Yogācāra exegetes before Ji, it does not appear that Ji paid his attention to the notion in his earlier period. He realized the need to draw on the notion in his later years.74 Ji must have been aware of this notion during his earlier period in the polemic situation of the Buddha Nature controversy, but it does not seem that he had adopted it as his major scholastic doctrine until he felt the need to reconcile the conflictions of the controversies. Such need appears to have emerged along with the translation of the Mahāprajñāpāramitāsūtra, because the Mahāprajñāpāramitā-sūtra, one of the ‘authentic’ Buddhist scriptures brought by Xuanzang from India, turned out to contain the phrase, ‘all sentient beings are tathāgatagarbha. Now, given that the ‘authentic’ scripture clearly stated that all sentient beings had the Buddha Nature, and Ji probably had to accept and explain it with no doctrinal contradiction to the Yogācāra doctrine of Five Distinct Lineages within one comprehensive Buddhist system. The notion of two types of Buddha Nature may be also associated to Ji’s change of attitude toward Bhāvaviveka since there is a doctrinal similarity between Bhāvaviveka’s view of emptiness and the doctrine of universal Buddha Nature. Just as the universal Buddha Nature is owned by all sentient beings, the Bhāvaviveka’s concept of emptiness, is applied to all dharmas. Ji criticized the universal Buddha Nature theory because he considered that this notion could not explain the particularity of Buddha Nature; likewise, Ji refuted Bhāvaviveka’s universal view that all dharmas, whether the Conditioned or the Unconditioned, was empty, because he regarded this view as contradicting to Dharmapāla’s view that only Imaginary Nature was empty. Yet, Ji attempted to resolve the contradiction between the universality and the particularity of the Buddha Nature by separating it into two distinct spheres of Principle and Practice; likewise, he tried to release the conflict between the universality and the particularity of the emptiness by attributing them to two distinct spheres of Principle and Practice, as I shall discuss below.

In the Xinjing youzan, Ji explains the concept of emptiness in the sphere of Principle; in commenting the phrase ‘[[[Bodhisattva Avalokiteśvara]]] clearly perceives that five aggregates of being are all empty’ (C. zhaojian wuyun deng jie kong 照見五蘊等皆空), Ji states that the emptiness refers to Principle of Thusness (C. zhenru li 真如理).75 According to Ji’s doctrinal taxonomy, the Prajñāpāramitā-sūtra corresponds to the seventh teaching of ‘ultimate truth that all are empty’ (C. shengyi jiekong zong 七勝義皆空宗), that is, Madhyamaka teaching including Bhāvaviveka’s.76 Thus, it may be said that Ji interprets the notion of emptiness of Madhyamaka, such as Bhāvaviveka, in the sphere of Principle. At the very next phrase, Ji quotes the phrase of the Liqufen, ‘all sentient beings are tathāgatagarbha; all dharmas are Thusness,’ and also he explains it in terms of Principle of Thusness, by saying that the reason why the Prajñāpāramitā-sūtra says this phrase is because phenomena do not have separate nature apart from Principle, that is, Thusness, the original essence.77 Here, again, Ji associates the phrase of the Prajñāpāramitā-sūtra to Principle of Thusness.

Besides, in other places of his later works, Ji associates Bhāvaviveka’s notion of emptiness to Thusness. In the Wuguo shu, for instance, Ji says that in Bhāvaviveka’s view the correct practice of the emptiness relies on the realization of Principle of Thusness.78 In the Liqufen shuzan, Ji says that Principle of Thusness is represented in all dharmas and it accords to Absolute Emptiness (C. bijie kong 畢竟空).79 In the Yilin zhang as well, it is said that Principle of Thusness manifests due to emptiness because it is the nature of emptiness, and thus is named as emptiness.80 To sum, in the later works, Ji interprets Bhāvaviveka’s notion of emptiness in terms of Principle, along with Thusness and tathāgatagarbha. When the emptiness is considered in the level of Principle, the doctrinal contradiction between Bhāvaviveka and Dharmapāla’s views on the emptiness can be resolved, because Bhāvaviveka’s notion of emptiness is categorized in a sphere that is separate from Dharmapāla’s. The emptiness that Bhāvaviveka applies to all dharmas (or, Three Natures) belongs to the sphere of Principle, while the emptiness that Dharmapāla attributes only to Imaginary Nature, not all Three Natures (or dharmas), is on the level of Practice.

For Ji, the Buddha Nature and the Emptiness-Existence controversies were both derived from the tension between the notions of universality and particularity, and he attempted to resolve this conflict by categorizing them into two separate spheres of Principle and Practice, respectively. In Ji’s doctrinal system, the contradiction between the doctrines of the universal Buddha Nature and the Five Distinct Lineages in the Buddha Nature controversy is solved by dividing the Buddha Nature into two types, one in Principle and the other in Practice; the equivalent contradiction in Emptiness-Existence controversy between Bhāvaviveka’s universal emptiness and Dharmapāla’s discriminative emptiness is released by dividing the emptiness into two types, that is, one on the level of Principle and the oner on the level of Practice. On the level of Principle, all dharmas are equal or universally ‘Such’ or empty; likewise, all sentient beings are universally tathāgatagarbha. On the contrary, on the level of Practice, all dharmas are not said equally empty; sentient beings are divided into five groups according to their spiritual Lineages. The seemingly contradictory doctrines of both the Buddha Nature controversy and the Emptiness-Existence controversy were explained by Ji as two distinct yet compatible levels of teaching within one comprehensive system of Buddhism. Concluding remarks

Against the background of the Buddha Nature controversy and the Emptiness-Existence controversy, it has been generally considered that Ji advocated one side of the controversies. However, the analysis of Ji’s works has disclosed that the doctrinal dichotomy of the controversies cannot be always interpreted as the contrast between the two doctrinally opposed groups. It can be said that Ji was a dedicated Yogācāra exegete throughout his career. But this does not mean that Ji hold an antagonistic attitude to Madhyamaka. In fact, Ji’s attitude toward his opponents in the controversies were not consistently hostile; in the later years, he changed his attitude to embrace the opponents’ views. In other words, Ji’s earlier position represents the typical antagonistic scheme of dichotomy; however, in his later years Ji responded to his opponents by embracing their views as a part of the broad system of Buddhist teaching, although they were still leveled lower than Yogācāra teaching. It may be then said that the point of the controversies lies on the way how the compatibility of two seemingly contradictory doctrines was explained, not on which doctrinal position was taken between the two opposed views. This also leads us to reflect on the way in which East Asian Buddhism is explained in term of the dichotomized frame, such as ‘old’ Yogācāra and ‘new’ Yogācāra, One Vehicle and Three Vehicles, Dharma Nature School and Dharma Characteristics School, and so on. Evidently, not all aspects of East Asian Buddhism are to be explained in the simplistic dichotomies; rather East Asian Buddhism may be said to be a continuation of attempts to reconcile variety of doctrinal contradictions. Ji’s doctrinal position in the later period, in this respect, represents the reconciling feature of East Asian Buddhist tradition.

Notes

1. The Dharma Characteristics School has been generally considered as the representative Yogācāra school of East Asia after Xuanzang, and thus the virtual founder Ji’s scholastic position is identified as that of the school. Therefore, Ji’s views on his opponents in the contemporary controversies are also regarded as representing the position of the Dharma Characteristics School. There are historical and theoretical problems, however, in placing the entire tradition of East Asian Yogācāra that inherited Xuanzang’s Yogācāra teaching under the umbrella category of the Dharma Characteristics School. For the discussion on the problems surrounding the establishment and acceptance of the ‘Dharma Characteristics School’ in East Asia, see Lee (Forthcoming 2015).

2. 至年十七遂預緇林。及乎入法。奉勅為奘師弟子。始住廣福寺 (宋高僧傳 T2061:50.725c09-

11). For more information on Ji’s biography, see Weinstein (1959) and Sponberg (1979).

3. The Yogācārabhūmiśāstra mentions this doctrine in a scattered way, while the Fodijing lun lists the Five Lineages in a passage. The Five Lineages in the Fodi jing lun are as follows: (1) Śrāvaka Lineage (S. śrāvaka-gotra, C. shengwen zhongxing 聲聞種性), for those who will become arhats via the śrāvaka vehicle, (2) Pratyekabuddha Lineage (S. pratyekabuddhagotra, C. dujue zhongxing 獨覺種性), for those destined to become Solitary Buddhas via the pratyekabuddha path, (3) Tathāgata Lineage (S. tathāgata-gotra, C. rulai zhongxing 如來種性), for those destined to become Buddhas, (4) Indeterminate Lineage (S. aniyata-gotra, C. buding zhongxing 不定種性), who may follow any of three vehicles, and (5) Lineage devoid of Supramundane Merits (S. *agotra, C. wuyou chushi gongde zhongxing無有出世功德種性), who are ineligible for the liberation, or who are devoid of the potential to become enlightened by being icchantikas (無始時來一切有情有五種性。一聲聞種性。二獨覺種性。三如來種性。四不定種性。五無有出世功德種性。如餘經論廣說其相。分別建立前四種性。雖無時限然有畢竟得滅度期。諸佛慈悲巧方便故。第五種性無有出世功德因故。畢竟無有得滅度期。諸佛但可為彼方便示現神通。說離惡趣生善趣法。彼雖依教勤修善因得生人趣。乃至非想非非想處。必還退下墮諸惡趣。諸佛方便復為現通說法教化。彼復修善得生善趣。後還退墮受諸苦惱。諸佛方便復更拔濟。如是展轉窮未來際。 不能令其畢竟滅度 (佛地經論 T1530:26.298a12-24)). The fifth is generally known as the Lineage devoid of the Nature (C. wuxing zhongxing 無性種性). Also see Yoshimura (2004, 237–240).

4. The famous line of the Lotus Sūtra may be noted in this regard: ‘only the One Vehicle teaching exists; there is neither the second (alt., Two) [Vehicle(s)], nor the third (alt., Three) [Vehicle(s)]’; 唯有一乘法，無二亦無三 (妙法蓮華經 T262:09.08a17-18). Those who advocated the notion of universal Buddha Nature interpreted this line as the Lotus Sūtra’s enhancement of One Vehicle over Three Vehicles by translating the numbers, respectively, as ‘Two’ and ‘Three’; on the contrary, Ji translated them as ‘second’ and ‘third,’ thereby claiming that the Lotus Sūtra did not dismiss the Three Vehicles.

5. Tendaishū Shūten Kankōkai (1912, 172–187; 193–194).

6. Yoshimura presumes that this controversy happened around between 648 and 650 on the basis of the fact that the translation of the Yogācārabhūmi and the Bodijing lun, which contain the doctrine of ‘Five Distinct Lineages,’ such as the Yogācārabhūmi and the Bodijing lun, was completed in 648 and 649, respectively, and the Hokkeshūku was composed around 650; see Yoshimura (2009). In this article, Yoshimura also discusses on another controversy between Fabao 法寶 (ca. 627–705) and Huizhao 慧沼 (648–714) happened around at the late seventh or early eighth century. Yoshimura says that Fabao, who worked for Xuanzang in his translation project, argues in the Yisheng foxing juijing lun 一乘佛性究竟論, dated between 695 and 699, that all sentient beings equally have Buddha Nature in both Principle and Practice, but Huizhao, one of Ji’s disciples, asserts in the Nengxian zhongbian huiri lun 能顯中邊慧日論, dated about 714, that there are the distinct spiritual levels of sentient beings, which are derived from the ‘Original Seeds’ (C. faer zhongzi 法爾種子). 7. Tendaishū Shūten Kankōkai (1912, 193–194). Also see Yoshimura (2009, 307–306).

8. Tendaishū Shūten Kankōkai (1912, 172–187). Also see Yoshimura (2009, 306–305). Shentai mentions another type of Buddha Nature in addition to the two types of Buddha Natures mentioned here, that is, ‘Buddha Nature of Ordinary Beings’ (C. fanfu foxing 凡夫佛性). 9. The verse reads: ‘In the level of True Nature, the Conditioned [[[dharmas]]] are empty, since they dependently arise like an illusion. The Unconditioned [[[dharmas]]] do not have Reality (C. shi 實, S. bhūta); they do not arise, like illusory flowers in the sky’; 真性有為空如幻緣生故無為無有實不起似空華 (大乘掌珍論 T1578:30.268b21-22). Bhāvaviveka explains later in the passage the True Nature as Ultimate Level of Reality; 真義自體說名真性。即勝義諦 (大乘掌珍論 T1578:30.268c12).

10. 由無遍計所執性故。亦說為空。非自性空。無生滅等。如來處處說三自性。皆言遍計所 執性空。依他圓成二性是有 (大乘廣百論釋論 T1571:30.248a28-b02); also see 此有密意。密意如何。謂此諸經唯破遍計所執自性。非一切無。若一切無便成邪見 (大乘廣百論釋論 T1571:30.247c27-28). 11. 相應論師有作是說。… 便成無見。不應與言。不應共住。自墮惡趣亦令他墮。如是成立遍計所執 。自性為空。及依他起自性為有。契當正理。若此義言。依他起性亦無所有故立為空。汝便墮落如上所說。過失深坑亦復成就誹謗世尊聖教過失 … 言有為法從眾緣生非自然有。就生無性說彼為空。 此有何義。若此義言。眼等有為依他起上不從因生常無滅壞。眼等自性畢竟無故。說名為空。便立已成。… 無性故空。不應說言就生無性說彼為空。… 若依他起自然生性。空無有故說之為空。是則還有立已成過。既許依他眾緣而生。實不空故應不名空。我則不爾。云何迷成相應師義 (大乘掌珍論 T1578:30.271c22- 272a26).

12. Although the controversy surrounding Bhāvaviveka and Dharmapāla’s different perspectives seems to have been pervasive and intensified, the doctrinal conflict does not seem so polemic at this beginning stage. As Moro Shigeki indicates, Japanese Hossō monk Zenju 善珠 (727–797) reports in the Yuishiki bunryō ketsu 唯識分量決 that the issues regarding these two works, which were raised by several Silla monks of the seventh century, were just on whether or not the two verses conform to each other. In other words, the Silla scholarmonks just noted the similarity between the verses, not the difference or contrast between them. Dharmapāla’s verse of the Gaungbai shilun, which is similar to Bhāvaviveka’s, reads as follows: ‘Since all Conditioned dharmas arise dependent on conditions, they are like illusions, which are not real substance. All Unconditioned dharmas also are not truly existent, since they do not arise, just like the hair on a tortoise’; 又所執境略有二種。一者有為。二者無為。諸有無法從緣生故。猶如幻事非實有體。諸無為法亦非實有。以無生故。譬似龜毛 (大乘廣百論釋論 T1571:30.225a05-08). See Shigeki (2004).

13. 說佛正法但經千載。非佛教法但住千歲。又聲聞藏雖佛去世百年已後。即分多部。而菩薩藏千載已前。清淨一味無有乖諍。千載已後乃興空有二種異論。是故說言。如來正法但經千載 (佛地經論 T1530:26.307a05-09). 14. When mentioning this dispute in the Bore boluomi duoxin jingzan 般若波羅蜜多心經贊, Korean Yogācāra exegete Wŏnch’ŭk 圓測 (613–696) cites Bandhuprabha (C. Qinquang 親光), the author of the Fodi jing lun; according to Wŏnch’ŭk, Bandhuprabha said that the dispute happened between Bhāvaviveka and Dharmapāla in South India; 親光釋曰千年已前佛法一味過千年後空有乖諍。佛滅沒已一千年後南印度界健至國中有二菩薩一時出世。一者清辨二者護法。為令有情悟入佛法立空有宗共成佛意。清辨菩薩執空撥有令除有執。護法菩薩立有撥空令除空執。然則空不違有即空之理非無不違空即色之說自成。亦空亦有順成二諦非空非有契會中道 。佛法大宗豈不斯矣 (佛說般若波羅蜜多心經贊 T1711:33.544a16-24). Such scholars as Fukaura Seibun, however, argue that there was not a virtual dispute between the two exegetes, by referring to Xuanzang’s record in the Da Tang xiyuji that although Bhāvaviveka visited Dharmapāla in order to discuss with him, Dharmapāla refused to meet him with the reason that he was fully occupied with practice; 城南不遠有大山巖，婆毘吠伽(唐言清辯)論師住阿素洛宮待見慈氏菩薩成佛之處。論師雅量弘遠，至德深邃，外示僧佉之服，內弘龍猛之學。聞摩揭陀國護法菩薩宣揚法教，學徒數千，有懷談議，杖錫而往。至波吒釐城，知護法菩薩在菩提樹，論師乃命門人曰：「汝行詣菩提樹護法菩薩所，如我辭曰：『菩薩宣揚遺教，導誘迷徒，仰德虛心，為日已久。然以宿願未果，遂乖禮謁。菩提樹者，誓不空見，見當有證，稱天人師。』」護法菩薩謂其使曰：「人世如幻，身命若浮，渴日勤誠，未遑談議」人信往

復，竟不會見。論師既還本土 (大唐西域記T2087: 51.930c25-931a08). See Seibun (1954, 147–148). Fukaura also claims that the dispute between Śīlabhadra and Jñānaprabha reported by Huayan exegete Fazang 法藏 (643–712) based on Divākara’s testimony is also suspected, or at least exaggerated; see Fukaura (1954, 149–173). Although the dispute itself is not a historical fact, the doctrinal contradiction between Bhāvaviveka and Dharmapāla was repeatedly discussed in East Asian Buddhist tradition. 15. 年二十五應詔譯經。講通大小乘教三十餘本。創意留心勤勤著述。蓋切問而近思。其則不遠矣 (宋高僧傳 T2061:50.725c14-16). 16. The six texts which Ji worked on are the Cheng weishi lun (659), the Madhyânta-vibhāgakārikā (C. Bian zhongbian lun song 辯中邊論頌) (661), the Madhyânta-vibhāga (C. Bian zhongbian lun 辯中邊論) (661), the Vimṃ:śatikā-vijñapti-mātratā-siddhi (C. Weishi ershi lun 唯識二十論) (661), the Samaya-bhedōparacana-cakra (C. Yibuzong lun lun 異部宗輪論) (662), and the Abhidharma-dhātu-kāya-pāda-śāstra (C. Apidamo jieshenzu lun 阿毘達磨界身足論) (663); 辯中邊論頌一卷(見內典錄彌勒菩薩造龍朔元年五月一日於玉華寺嘉壽殿譯沙門大乘基筆受) 辯中邊論三卷(見內典錄世親菩薩造第二出與中邊分別論同本龍朔元年五月十日於玉華寺嘉壽殿譯至三十日畢沙門大乘基筆受) … 唯識二十論一卷(見翻經圖世親菩薩造第三出與元魏智希陳真諦出者同本龍朔元年六月一日於玉華寺慶福殿譯沙門大乘基筆 受) … 成唯識論十卷(見內典錄護法等菩薩造顯慶四年閏十月於玉華寺雲光殿譯沙門大乘基筆受) … 阿毘達磨界身足論三卷(見翻經圖筏蘇蜜多羅造龍朔三年六月四日於玉華寺八桂亭譯畢沙門大乘基筆受) … 異部宗輪論一卷(見翻經圖世友造第三出與十八部論部異執論並同本龍朔二年七月十四日於玉華寺慶福殿譯沙門大乘基筆受) (開元釋教錄 T2154:55.556c04-557b06).

17. 初功之際十釋別翻。昉．尚．光．基四人同受。潤飾．執筆．撿文．纂義。既為令範務各有司。數朝之後基求退迹。大師固問。基慇請曰。自夕夢金容晨趨白馬。英髦間出。靈智肩隨。聞五分以心祈。攬八蘊而遐望。雖得法門之糟粕。然失玄源之淳粹。今東出策賚。並目擊玄宗。幸復獨秀萬方穎超千古。不立功於參糅。可謂失時者也。況群聖製作。各馳譽於五天。雖文具傳於貝葉。而義不備於一本。情見各異稟者無依。況時漸人澆。命促惠舛。討支離而頗究。攬初旨而難宣。請錯綜群言以為一本。揩定真謬權衡盛則。久而遂許。故得此論行焉。大師理遣三賢獨授庸拙　此論也 (成唯識論掌中樞要

T1831:43.608b29-c14).

18. See Yoshimura (1999).

19. See note 17 above.

20. Although Ji may have thought that the translation of the Cheng weishi lun focused on Dharmapāla’s position would be proper in the controversial situation, it seems that the perspective that Bhāvaviveka’s doctrinal position was fundamentally antagonistic to Dharmapāla’s became pervasive and dominant after the translation. Since the Cheng weishi lun was focused on Dharmapāla’s doctrinal views, it must have worked as a source material of Dharmapāla’s view, thereby making it easier to compare it to Bhāvaviveka’s. Further, Ji’s commentaries on the Cheng weishi lun, which contains harsh criticism on Bhāvaviveka, as discussed soon below, must have worked as a sort of catalyst to intensify the controversy.

21. 依障建立種姓別者意顯無漏種子有無。謂若全無無漏種者彼二障種永不可害即立彼為非涅槃法。若唯有二乘無漏種者 彼所知障種永不可害。一分立為聲聞種姓 一分立為獨覺種姓。若亦有佛無漏種者彼二障種俱可永害。即立彼為如來種姓。故由無漏種子有無障有可斷不可斷義 (成唯識論 T1585:31.09a21-28). 22. 有執大乘遣相空理為究竟者。依似比量撥無此識及一切法 (成唯識論 T1585:31.16a06-08).

23. 論。有執大乘至及一切法。述曰。第五清辨無相大乘。於俗諦中亦說依他．圓成有故。真諦皆空故。今言空者遣遍計所執。彼執此文為正解故。彼依掌珍真性有為空等似比量。撥無此識及一切法皆言無體 (成唯識論述記 T1830:43.359a01-05).

24. Hayashi presumes that the Weishilu shuji was composed along with the translation of Cheng weishi lun on the basis of Ji’s statement in the Yibuzong lun lun shuji 異部宗輪論述記, the commentary of the Samaya-bhedōparacana-cakra; Ji says in this text that he titled the commentary ‘Record of what has been described’ (C. shuji述記) because he compiled it by following [[[Xuanzang’s]]] teaching while he was translating the text (今我親教三藏法師玄奘以大唐龍朔二年七月十四日於玉華宮慶福殿重譯斯本基虗簉譯僚謬參資列隨翻受旨編為述記 (異部宗輪論述記 X844:53.568b06-09)). Hayashi also suggests that the two commentaries were composed almost simultaneously because the two texts cite to each other in significant numbers. Hayashi also indicates that these two works were consistently revised thereafter because Ji’s later works, such as the Fahua xuanzan 法華玄贊 and the

Weishi ershi lun shuji 唯識二十論述記are cited in them. For the details, see Hayashi (2012). 25. 辨教所被機者。依瑜伽等有五種姓。一菩薩。二獨覺。三聲聞。四不定。五無姓。此論第三云。入見 菩薩皆名勝者。證阿賴耶故正為說 (成唯識論述記 T1830:43.230a14-15); 論。又諸有情至不由熏生。　述曰。此第二引經論通證。即是十卷入楞伽第二卷。無上依經上卷。善勇猛般若第一卷。大般若經第五百九十三卷。說前種姓。大莊嚴論第一卷末種姓品。及此瑜伽第二十一聲聞地。皆說有五種姓別。故應定有法爾無漏種子。不由熏生 (成唯識論述記 T1830:43.304c13-19).

26. 有五種性證法。一聲聞乘性。二辟支佛乘性。三如來乘性。四不定乘性。五者無性謂一闡提。… 楞伽所說二種闡提。初是斷善根具邪見者。後是菩薩具大悲者。初者有入涅槃之時。後必不爾。以眾生界無盡時故。無性有情不成佛故。大慈菩薩無成佛期。然第五性合有三種。一名一闡底迦。二名阿闡底迦。三名阿顛底迦。一闡底迦是樂欲義。樂生死故。阿闡底迦是不樂欲義。不樂涅槃故。此二通不斷善根人。不信．愚癡所覆蔽故。亦通大悲菩薩。大智大悲所熏習故。阿顛底迦名為畢竟。畢竟無涅槃性故。此無性人亦得前二名也。前二久久當會成佛。後必不成 (成唯識論掌中樞要T1831:43.610b29-611a01).

27. See note 22 above.

28. See note 23 above.

29. 由斯遠離增減二邊。唯識義成契會中道 (成唯識論 T1585:31.39a03-04)

30. 論。由斯遠離至契會中道。述曰。總結。由此二文三師理故遠離二邊。無心外法故除增益邊。有虛妄心等故離損減邊。離損減邊故。除撥無如空花清辨等說。離增益邊故。除心外有法諸小乘執。唯識義成契會中道。無偏執故。言中道者正智也。理順正智名契會中道 (成唯識論述記 T1830:43.488a08-14).

31. The Madhyānta-vibhāga had not been translated yet by Xuanzang when the Cheng weishi lun was translated, but the Zhongbian fenbie lun 中邊分別論, Paramārtha’s translation of the Madhyânta-vibhāga dated 558, conceivably was available to Ji. Xuanzang translated the Madhyânta-vibhāga to the Bian zhongbian lun 辨中邊論 later in 661. The verse reads as follows: ‘The mind that falsely discriminates exists; the duality (viz. Atman and dharmas) which manifests itself in it is absolutely non-existent; in this false discrimination is Emptiness only; in Emptiness, too, there is this false discrimination. I therefore say that all dharmas are neither empty nor non-empty. There is existence of false discrimination, nonexistence of the duality of Atman and dharmas, existence of Emptiness in false discrimination, and existence of false discrimination in Emptiness. That is the Middle Way.’ I have used here Wei Tat’s translation (Wei 1973, 511); 虛妄分別有於此二都無此中唯有空　於彼亦有此故說一切法非空非不空有無及有故是則契中道 (成唯識論 T1585:31.39b04-07; 辨中邊論 T1600:31.464b16-17; b25-26). This verse is also often quoted in contrast to Bhāvaviveka’s verse by later exegetes, such as Huizhao, as well.

32. 此中應言三故字。謂有故即妄分別。無故即能．所取。及有故即俗．空互有。…　是則契中道者。謂非一向空如清辨。非一向有如小乘故。名處中道。謂二諦有不同清辨。二取無不同小部。故處中道 (成唯識論述記 T1830:43.490b11-18).

33. 此識若無便無俗諦。俗諦無故真諦亦無。真俗相依而建立故。撥無二諦是惡取空。諸佛說為不可治者應知諸法有空不空。由此慈尊說前二頌 (成唯識論 T1585:31.39b17-20). 34. 論。撥無二諦至說前二頌。　述曰。若撥無識及性。即撥無二諦。佛說為不可治者。沈淪生死病根深故。即清辨等。應知諸法遍計所執無故有空。依他．圓成有故有不空也。故彌勒說前二頌。即前中邊頌。二十唯識義與此同 (成唯識論述記 T1830:43.492b25-c01).

35. 辯中邊論三卷(見內典錄世親菩薩造第二出與中邊分別論同本龍朔元年五月十日於玉華寺嘉壽殿譯至三十日畢沙門大乘基筆受) … 唯識二十論一卷(見翻經圖世親菩薩造第三出與元魏智希陳真諦出者同本龍朔元年六月一日於玉華寺慶福殿譯沙門大乘基筆受) (開元釋教錄 T2154:55.556c06-15). 36. 執法有實種類甚多。執法無實如空見外道清辯等計。然如所執法即無實。如依他性法即有實。故不可言。彼亦撥無假法性故。依他性中實我則無。故不同法。須置假似我之言 (辨中邊論述記 T1835:44.40a28-b03).

37. 依清辨等。破有為空。真性有為空。緣生故。如幻。彼似比量。非真比量。若我真性。離心言故。有為非空。若汝真性。非極成有。唯是空故 (唯識二十論述記 T1834:43.983c02-05). 38. Hayashi argues this on the basis of the fact that in the Bian zhongbianlun shuji Ji cites such a work as the Weishilun shuji although he does not cite his later works. See Hayashi (2012).

39. The Weishi ershilun shuji may be presumed to have been first drafted while Xuanzang was translating the Vimṃ:śatikā-vijñaptimātratāsiddhi-śāstra, just like the Weishilun shuji (See n. 24 above). Yet, as Hayashi indicates, this work also appears to have been revised later by Ji. For the details, see Hayashi (2012, 199–201).

40. The extant Zajilun shuji consists of 10 fascicles, but it has been suspected that the latter part from the fifth fascicle might not have been created originally by Ji. See Hayashi (2009). Ji’s comment on Bhāvaviveka in the Zajilun shuji is included in the first fascicle. 41. At the beginning part of the Zajilun shuji, Ji says that the work consists of six sections; 初發論端略以六門解釋一教起所因二論興所為三彰體性四顯宗旨五釋題目六解本文 (雜集論述記 X796:48.1c02-03). Among these, the third section on the ‘essential nature’ (C. tixing 體性) and the forth on the ‘cardinal meaning’ (C. zongzhi宗旨) [of the text] contain Ji’s descriptions of Bhāvaviveka: 彰體性者略有二種一依清辯朋龍猛言勝義諦中一切無相諸法皆空 教無所教 體亦非體 世俗諦中可亦說有 句言章論聲為教體 (雜集論述記

X796:48.2a20-22); 顯宗旨者略有二說一依清辯朋輔龍猛 … 雖大乘宗然非此意 (雜集論述記 X796:48.4c04-14). These two sections appear in the Liqufen shuzan as well, although Ji’s tone on Bhāvaviveka in the Liqufen shuzan is different from that of the Zajilun shuji. I will discuss about this soon below. 42. Ji cites 10 stages that are shared by all Three Vehicles from the Liqufen, as follows: 理趣等經說一淨觀地二種姓地三第八地四具見地五薄地六離欲地七已辦地八獨覺地九菩薩地十如來地此說三乘共行十地攝法 (雜集論述記 X796:48.67c09-11). But this passage on the 10 stages are stated in such fascicles as 56th; 416th; 417th and so on, not in the Liqufen. Therefore, Ji does not seem to have read the Liqufen yet, but just is aware that these 10 stages are stated in a certain fascicle of the Mahāprajñāpāramitā-sūtra. Based on this, I do not attribute this work to Ji’s later works.

43. 敘經宗旨者 佛滅度後 九百年間 有應真大士。厥名清辨。身同數論之儀。示無朋黨之執。心處釋迦之理宗無偏滯之情。時人號為妙吉祥菩薩。神異聖德廣如別記 (大般若波羅蜜多經般若理趣分述讚 T1695:33.25c19-23).

44. Just as Ji says at the beginning of the Zajilun shuji that this work consists of six sections (See note 41 above), Ji states at the beginning of the Liqufen shuzan that the Liqufen shuzan is composed of four sections; 將讚經文略以四門解釋。一敘經宗旨。二顯經體性。三彰經勝德。四釋經本文。(般若理趣分述讚 T1695:33.25c18-19). Among these, ‘the cardinal teaching’ and ‘the essential nature’ [of the text] are duplicated in the two works.

45. See note 41 above.

46. See note 43 above.

47. 若依清辨世俗可有勝義皆空今說正義空寂。若依護法應依下釋。… 且如凡夫遍計所執色蘊等法。本來空寂空寂即清淨。但有妄心境都無故。依他門有本性亦空。… 故知依他亦性空寂。幻法上無自然性故名之為空。圓成之體實理而說非空不空。由依二空門方可顯證亦可名空。故三種法皆名為空。上護法解。清辨釋言。又世俗妄說色等可有勝義入真色等皆無。故本空寂如第二月妄有即真無。故本皆空寂下諸空寂皆準此知。並應二解 (般若理趣分述讚 T1695:33.40b19-42b12).

48. 經曰。由遠離故自性寂靜由寂靜故自性清淨由清淨故甚深般若波羅蜜多最勝甚深。　讚曰。生死囂煩真如寂靜故由遠離。自性亦靜煩惱有染空性清淨故由寂靜顯體清淨。此意總顯。由一切法自性空故空寂句義即是菩薩所有句義。諸法既空故離虛妄離虛妄故體無囂動故性寂靜。既無囂動自性空寂故非染污自性清淨。自性清淨者顯般若最勝清淨。觀照悟此本性清淨真相自體本來清淨。故文字等甚深般若最勝清淨。清辨護法二釋隨應

(般若理趣分述讚 T1695:33.48c22-a03).

49. See note 34 above.

50. The Xinjing youzan cites the following phrase from the Liqufen; 又此空者即真如理。性非空有因空所顯。遮執為有故假名空。愚夫不知執五蘊等定離真有。起相分別。今推歸本體即真如。事離於理無別性故。由此經言。一切有情皆如來藏。一切法等皆即真如 (般若波羅蜜多心經幽贊 T1710:33.535c01-05). Although ‘Liqufen’ or ‘Liqu’ is not particularly mentioned here, the only place of the Mahāprajñāpāramitā-sūtra, in which this passage appears, is in the Liqufen. Furthermore, there is another passage that cites from the Liqufen in the Xinjing youzan; 理趣分說。信學此經速能滿足諸菩薩行。疾證無上正等菩提 (般若波羅蜜多心經幽贊 T1710:33.542a10-12). The original text of this passage reads: 佛說如是灌頂法門般若理趣智藏法已，告金剛手菩薩等言若有得聞如是灌頂甚深理趣智藏法門信解、受持、讀誦、修習，速能滿足諸菩薩行，疾證無上正等菩提 (大般若經第十會般

若理趣分 T220:07.988a08-12).

51. 聖龍猛等為除有執採集真教究暢空宗。如別頌曰。真性有為空如幻緣生故無為無有實不起似空華彼言世俗可說法有。依勝義諦一切皆空。雖此真空性非空有。寄詮勝義理皆性空。有情由是次生空見。無著菩薩 復請慈尊 說中道教。雙除二執。而說頌言。虛妄分別有於此二都無此中唯有空於彼亦有此故說一切法非空非不空有無及有故是則契中道彼言世俗說我法有依勝義諦唯此二空。雖佛為破執空執有總相宣說諸法有空。或說諸法非空非有。名字性離空有雙非。勝義寄詮有空有有 (般若波羅蜜多心經幽贊 T1710:33.523c13-28).

52. 經曰。照見五蘊等皆空。　贊曰。此顯由行甚深般若得正慧眼。達空名照。謂色受等諸有為法。皆有三世．內外．麁細．劣勝．近遠。積聚名蘊。此五。謂色．受．想．行．識。等言等取處等諸法。勝空者言。前破能觀執顯能觀空。今破所觀執顯所觀空。若癡所蔽迷勝義理。於蘊等中妄執為有。如處夢者見境現前。若正了知勝義諦理。不生執著。如夢覺位了境非有。故行般若便照性空。如應者言。雖修一切皆行般若。證真遣妄由慧照空。故此偏說。此中空言即三無性。謂計所執本體非有相無自性。所以稱空。諸依他起。色如聚沫。受喻浮泡。想同陽焰。行類芭蕉。識猶幻事。無如所執自然生性。故亦名空。圓成實性因觀所執空無方證。或無如彼所執真性。故真勝義亦名為空。據實三性非空非不空。對破有執總密說空。非後二性都無名空。說一切空是佛密意。於有及無總說空故。如世尊說。相生勝義無自性　如是我皆已顯示若不知佛此密意　失壞正道不能往 (般若波羅蜜多心經幽贊 T1710:33.535b10-29). 53. See note 47 above.

54. 然古大德。總立四宗。一立性宗。雜心等是。二破性宗。成實等是。三破相宗。中百等是。四顯實宗。涅槃等是。今即不爾。於中分二。初陳異宗。後列自主。陳異宗中。復分為二。初陳外道。後陳小乘 … 列自主中復分為二。初列邊主後列中主。列邊主者。謂清辨等朋輔龍猛。般若經意說諸法空。… 此由所說勝義諦中皆唯空故。名為邊主。列中主者。謂天親等輔從慈氏。深密等經。依真俗諦說一切法有空不空。… 非空非有中道義立。即以所明說一切法非空非有中道之義。以為宗也 (大乘法苑義林章 T1861:45.249c04- 251a14).In another passage, Ji says in a similar way that there are two ways to manifest the Great Vehicle: one is the manifestation of the ‘essence of extreme’ (C. bianti 邊體) and the other is the manifestation of the ‘Middle Way’ (C. zhongdao 中道); 顯大乘中。復分為二。初顯邊體。後顯中道。顯邊體者。龍猛．清辨咸作是言。勝義諦中一切無相諸法皆空。何教何為體。世俗諦中可亦說有。… 顯中道者。瑜伽論攝決擇分八十一卷說 (大乘法苑 義林章T1861:45.251c21-252a07).

55. The citation in the Yilin zhang is as follows: 凡論出體總有四重。一攝相歸性體。即一切法皆性真如。故大波若經理趣分說。一切有情皆如來藏 (大乘法苑義林章 T1861:45.252c09-

11). 56. 清辯菩薩。般若燈論。釋有四義。謂發端標舉簡持指斥 (因明入正理論疏 T1840:44.96c10-11).

57. 大般若云。是世間出。故名世間。造世間故由世間故。為世間故。因世間故屬世間故。依世間故名為世間。廣如第五百卷說此有二種。一非學世間。除諸學者。所餘世間所共 許法。二學者世間。即諸聖者所知麁法。若深妙法便非世間。(因明入正理論疏 T1840:44.115a28-b05). However, this passage is in fact in the 498th fascicle; 舍利子如是六種波羅蜜多，何因緣故名為世間 復何因緣名為出世 舍利子 世間者，謂彼六種波羅蜜多，是世間故，名為世間；造世間故，名為世間；由世間故，名為世間；為世間故，名為世間；因世間故，名為世間；屬世間故，名為世間；依世間故，名為世間。 (大般若波羅蜜多經卷第四百九十八 T220:07.535a26-b03).

58. 世尊初於一時。在婆羅泥斯仙人墮處施鹿林中。唯為發趣聲聞乘者。以四諦相轉正法輪。雖是甚奇甚為希有。一切世間諸天人等。先無有能如法轉者。而於彼時所轉法輪。有上有容是未了義。是諸諍論安足處所。世尊在昔第二時中。唯為發趣修大乘者。依一切法皆無自性無生無滅本來寂靜自性涅槃。以隱密相轉正法輪。雖更甚奇甚為希有。而於彼時所轉法輪。亦是有上有所容受。猶未了義。是諸諍論安足處所。世尊於今第三時中。普為發趣一切乘者。依一切法皆無自性。無生無滅。本來寂靜自性涅槃。無自性性。以顯了相轉正法輪。第一甚奇最為希有。于今。世尊所轉法輪。無上無容是真了義。非諸諍論安足處所 (大乘法苑義林章 T1861:45.251c21-252a07). 59. It is known that the Wugou shu was written in 672 through 674, based on the record at the end of the work, that is, a passage saying that Ji began the work in 672 (咸享三年) and finished it in 674

(五年); 基以咸享三年十二月二十七日。曾不披讀古德章疏。遂被并州大原縣平等寺諸德迫講舊經。乃同講次。制作此文。以贊玄旨。夜制朝講。隨時遂怠。曾未覆問。又以五年七月。遊至幽明蘇地。更講舊經。方得重覽。文雖疏而義蜜。詞雖淺而理深。但以時序怱迫。不果周委言。今經文不同之處。略并敘之。諸德幸留心而覽也 (說無垢稱經疏 T1782:38.1114a20-27). The Wugou shu also quotes the phrase of ‘all sentient beings are tathāgatagarbha’ from the Liqufen; 故大般若言。一切有情皆如來藏 (說無垢稱經疏

T1782:38.1001a14-15). Even if this passage does not directly mention Liqufen or Liqu, the only place of the Mahāprajñāpāramitā-sūtra in which this passage appears is in the Liqufen (Also see n.50). The Fahua xuanzan also cites the phrase from the Liqufen; 般若理趣云。一切有情皆如來藏。普賢菩薩遍自體故。由證普遍賢善之理能證之道名為普賢 (妙法蓮華經玄贊 T1723:34.852a27-29). Hayashi presumes that this text was composed in late 670’s and thus this explains the doctrinal similarity between this work and the Shuo Wugoucheng jing shu, which was written in 672 through 674; see Kana (2012).

60. 七勝義皆空宗。謂清辨等。明說空經。以為了義。說一切法。世俗可有。勝義皆空。八應理圓實宗。謂護法等。弘暢花嚴深密等經。雖說二諦。隨其所應。具有空理。圓妙無闕。實殊勝故。今此經非前六宗。後二唯是大乘所說。故知通是二宗所攝 (說無垢稱經疏 T1782:38.999b04-09); 宗有八者。一我法俱有。犢子部等。二有法無我。薩婆多等。三法無去來。大眾部等。四現通假實。說假部等。五俗妄真實。說出世部等。六諸法但名。說一部等。七勝義皆空。般若等經。龍樹等說中百論等。八應理圓實。此法華等．無著 等說中道教也 (妙法蓮華經玄贊 T1723:34.657a29-b06). In this passage of the Wuguo shu, Ji clearly mentions Bhāvaviveka’s teaching as the seventh. 61. This is one of the four kinds of the essence (C. ti 體), which Ji describes in his several works, along with ‘the essence of all objects returning to consciousness’ (C. shexiing guishi 攝境歸識), ‘all that are provisional return to reality’ (C. shejia guishi 攝假隨實), ‘nature and functions are distinctly explained’ (C. xingyong bielun 性用別論).

62. 攝相歸性。皆如為體。故經說言。一切法亦如也。至於彌勒亦如也 (成唯識論述記 T1830:43.230b28-29); 二者護法勝子親光等說凡論出體總有四重一攝相歸性體即一切法 皆性真如 無垢稱云 一切眾生皆如也 一切法亦如也 (大乘阿毗達磨雜集論述記 X796:48.3a08-10). Xuanzang translated the Vimalakīrtinirdeśa into the Shuo wugoucheng jing 說無垢稱經 in 650. Besides, the Chengweishilun liaojian 成唯識論料簡 and the Weishi ershilun shuji 唯識二十論述記 also mention this concept of ‘essence of all characteristics returning to nature’; the former quotes only the phrase from the Vimalakīrtinirdeśa just as the Weishilun shuji and the Zajilun shuji, while the latter just mentions the concept with no quotation. 此中第一攝相歸性門者相謂相狀即依他起諸有為法相狀故別性謂實性即真如理是諸有為真實性故由此諸法性即真如故聲名等真如為體亦可說名攝妄歸真門諸師隨自所宗故淨名云一切眾生皆如也一切法亦如也 (成唯識論料簡 X806:48.371b09-13); 第二明論宗體造論所由者。於中亦三。一辨論宗。二辨論體。三辨造論所由。 初辨宗者。 所明唯識唯識為宗。離自所明更無宗故。次辨體者。體有四種。如餘處說。攝相歸性。真如為體。攝境歸識。以心為體。攝假歸實。以聲為體。性相別論。即有二種。一增上緣。許佛說法。以佛無漏聲名句等。為其教體。佛不說法。大定智悲。為其教體。二親因緣。隨佛說不說。皆於能聽者。耳意識上所變聲等。為其教體 (唯識二十論述記 T1834:43.979b03-11). 63. The quotations from the Liqufen and the Vimalakīrtinirdeśa in the Liqufen shuzan are as follows: 護法釋言。教體有四。一攝相歸性體。般若論說。應化非真佛亦非說法者。說法不二取無說離言相。此經下言。一切有情及法皆即真如故。甚深般若波羅蜜多亦即真

如。無垢稱言。文字性離。無有文字。是即解脫。解脫相者即諸法也。又一切法亦如 也。故知教體性即真如 (般若理趣分述讚 T1695:33.26a25-b02); 第三句云遍金剛智大觀自在者。即下十四段中第十二一切有情住持遍滿勝藏法門。謂佛具證一切有情皆如來藏。普賢菩薩自體遍故。能證如來藏遍諸有情。此能證智名遍金剛智。摧滅生死一切障故 (般若理趣分述讚 T1695:31b16-21); 經曰。謂一切有情皆如來藏普賢菩薩自體遍故。讚曰。自下第二正陳所說於中有四。此即第一。一切有情皆如來藏者如來藏者即是真如在纏之名。出纏之時名法身故 … 故說普賢菩薩自體遍有情體故。有情者皆如來藏。廢用顯體名普賢遍 (般若理趣分述讚 T1695:33.58a11-26). Here, Ji explains tathāgatagarbha and Thusness in the same level. I will explain this soon below. 64. The Yilin zhang has the quotations from all three texts of the Liqifen, the Śrīmālādevī-sūtra, and the Vimalakīrtinirdeśa as follows: 凡論出體總有四重。一攝相歸性體。即一切法皆性真如。故大波若經理趣分說。一切有情皆如來藏。勝鬘經說。夫生死者是如來藏。無垢稱言。一切眾生皆如也。一切法亦如也。眾賢聖亦如也。至於彌勒亦如也。諸經論說如 是非一。一切有為諸無為等有別體法是如之相 (大乘法苑義林章T1861:45.252c09-15). The Wuguo shu also quotes from all the three works: 若依勝義。通論諸法體性有四。一攝相歸性體。二攝餘歸識體。三攝假隨實體。四假實別論體。攝相歸性體者。一切有為無為等法。體即真如。真如為本。故大般若言。一切有情。皆如來藏。勝鬘亦言。夫生死者。是如來藏。若無如來藏者。不能厭苦樂求涅槃。是故如來藏。是依。是持。是建立。無垢稱言。一切有情皆如也。一切法亦如也 (說無垢稱經疏 T1782:38.1001a11-18). 65. The Shengmanjing shuji 勝鬘經述記, a commentary of the Śrīmālādevī-sūtra, is generally attributed to Ji, since at the beginning of this work it is recorded that Ji’s disciple Yiling 義令 (d.u.) wrote down Ji’s words (大乘慈恩 基法師說 門人 義令 記 (勝鬘經述記

X352:19.898b08-09)), but the authorship of this work has been suspected. There is still strong possibility, however, that Ji himself lectured on the Śrīmālādevī-sūtra in the later period of his life, given Yiling’s testimony on this at the end of the work as follows; 如有人寫者。好用心力。仰於感享歲。影從法師。首尾中間。曾無蹔捨。謹於界部寶國伽藍。令自春初。聽勝鬘經及二十論。隨所採撮。諮聽未聞。既受指麾。編為述記。但文約義廣。披閱稍難又於儀鳳二年夏洛都東太原寺。重更普搆。舒文展義。疎缺更[甦生 ± 益]。廣引文證。庶後學者。願莫嗤焉。義令記也 (勝鬘經述記 X352:19.924b19-c01). 66. See note 59 above. 67. See note 4 above. 68. This is probably due to the specific genre of the Yinminglun shu, that is, Buddhist logic. However, it is still very likely that the Yinminglun shu was composed after 663, given that Bhāvaviveka is referred to as ‘bodhisattva’ in this work. See note 56 above. 69. For the quotation of the phrase in the Liqufen shuzan, see note 63 above; for that in the Xinjing youzan, see note 50 above; for the quotation in the Yilin zhang, see note 55 and note 64 above; for the quotation in the Wuguo shu and the Fahua xuanzan, see note 59 and note 64 above. The original texts of ‘all sentient beings are tathāgatagarbha’ or ‘all sentient beings are Thusness’ in the Liqufen are as follows; 爾時，世尊復依一切住持藏法如來之相，為諸菩薩宣說般若波羅蜜多 一切有情住持遍滿 甚深理趣 勝藏法門，謂一切有情皆如來藏，普賢菩薩自體遍故；一切有情皆金剛藏，以金剛藏所灌灑故；一切有情皆正法藏，一切皆隨正語轉故；一 切有情皆妙業藏，一切事業加行依故 (大般若經第十會般若理趣分 T220:07.990b01-07); 一切有情即真如故，甚深般若波羅蜜多亦即真如; 一切法即真如故，甚深般若波羅蜜多亦即真如 (大般若經第十會般若理趣分 T220:07.989c19-21). There are also similar phrases at T220:07.751a29-b01; T220:07.414b14; T220:06.1038b21-22. 70. 然性有二。一理性勝鬘所說如來藏是。二行性楞伽所說如來藏是。前皆有之 後性或無 (妙法蓮華經玄贊 T1723:34.656a25-27). 71. 談有藏無說皆作佛。依善戒經地持論中唯說有二。一有種姓二無種姓。彼經論云性種姓者無始法爾六處殊勝展轉相續。此依行性有種姓也。無種姓人無種性故。雖復發心懃行精進終不能得無上菩提。但以人天善根而成就之。即無性也 (妙法蓮華經玄贊 T1723:34.656a27-b04). 72. 經曰。謂一切有情皆如來藏普賢菩薩自體遍故。讚曰。自下第二正陳所說於中有四。此即第一。一切有情皆如來藏者如來藏者即是真如在纏之名。出纏之時名法身故。藏謂庫藏諸佛所有一切功德皆在其中名如來藏。現行功德未能起故不名法身。又此真性正實如來藏在纏中名如來藏。一切眾生皆有真理故 (般若理趣分述讚 T1695:33.58a11-18). It appears that Ji uses 真理 and 真如理 with the same meaning, since, in other places of his works, Ji uses 真理 in the place of 真如理. For instance, in the Wuguo shu Ji uses the two terms without distinction in listing four types of nirvān:ṇa as follows: 涅槃有四。一自性清淨涅槃。謂一切法相真如理。二有餘依涅槃。謂苦因盡所顯真理。三無餘依涅槃。謂苦果盡所顯真理。四無住處涅槃。謂所知障盡。大悲般若所轉真理 (說無垢稱經疏 T1782:38.1059a01-05). Also see T1782:38.1042a02; 1091b01-02; 1109b03, etc. In some places, Ji uses 真如理 and 真如 mixed as well. In the Chengweishilun liaojian, for instance, Ji refers to the nature of all dharmas as Principle of Thusness, or Thusness; 攝相歸性門者相謂相狀即依他起諸有為法相狀故別性謂實性即真如理是諸有為真實性故由此諸法性即真如故聲名等真如為體 (成唯識論料簡 X806:48.371b09-11). 73. 若依勝義。通論諸法體性有四。一攝相歸性體。二攝餘歸識體。三攝假隨實體。四假實別論體。攝相歸性體者。一切有為無為等法。體即真如。真如為本。故大般若言。一切有情。皆如來藏。勝鬘亦言。夫生死者。是如來藏。若無如來藏者。不能厭苦樂求涅槃。是故如來藏。是依。是持。是建立。無垢稱言。一切有情皆如也。一切法亦如也 (說無垢稱經疏 T1782:38.1001a11-18). I have introduced this passage above when explaining the notion of ‘essence of all characteristics returning to the nature’ (see note 62 above). In a similar way that the Wuguo shu does, the Yilin zhang also quotes the phrases from the Liqufen, the Śrīmālādevī-sūtra, and the Vimalakīrtinirdeśa to explain that the nature of all dharmas is Thusness. See note 61 and note 64 above. 74. In fact the Fahua xuanzan, in which Ji explicates this notion, is one of Ji’s latest works. See note 59 above. 75. 經曰。照見五蘊等皆空。贊曰。… 又此空者即真如理。性非空有因空所顯。遮執為有故假名空。愚夫不知 執五蘊等 定離真有 起相分別 (般若波羅蜜多心經幽贊 T1710:33.535b10-c03).

76. See note 60 above. 77. 今推歸本體即真如。事離於理無別性故。由此經言。一切有情皆如來藏。一切法等皆即真如 (般若波羅蜜多心經幽贊 T1710:33.535c03-04). 78. 經。牟尼如是(至)無所住。贊曰。此有一歸。歎證真空。清辨解云。善修空者。證真理故。如空無罣礙。如蓮不著水。諸相已遣。更無所遣。生死盡故。出世願滿。更無所願。故今稽首。如空無住。得真諦者。護法解云。善修空者。空有二種。一云舜若。此但名空。空者無也。即遍計所執。二云舜若多。此云空性。空之性故。體即真如。性是有也。善修所執空故。一切相遣。善修圓成空性故。一切願滿。由此歸佛。如空無住。 舊經似有文中仍闕 (說無垢稱經疏 T1782:38.1022c29-a09). Here, the fact that Ji interprets Bhāvaviveka’s concept of emptiness in terms of universally applied Principle becomes clear when compared to Ji’s explanation of Dharmapāla’s notion of emptiness. Dharmapāla, Ji says, divides the emptiness into the two types, that is, first, the mere absence (C. kong 空), and second, the nature of emptiness (C. kongxing 空性), the essence of which is Thusness. In other words, Ji makes a contrast between two exegetes: Dharmapāla discriminates the emptiness as Thusness from the mere absence, whereas Bhāvaviveka only deals with the former. This By this, Ji implies that Bhāvaviveka, who does not divide emptiness into the types, considers all Natures (viz., all dharmas) in terms of Thusness.

79. 此真如理印於諸法。契畢竟空。方能證會佛皆圓證 (般若理趣分述讚 T1695:33.33b20-21). 80. 其真如理因空所顯空之性故。名之為空 (大乘法苑義林章 T1861:45.365c21-22). Disclosure statement No potential conflict of interest was reported by the author. Funding This work was supported by the Academy of Korean Studies (KSPS) Grant funded by the Korean Government (MOE) [Grant Number: AKS-2012-AAZ-104]. Abbreviations C Chinese S Sanskrit T Taishō shinshū daizōkyō 大正新脩大藏經. editors, Takakusu Junjirō 高楠順次郎, Watanabe Kaikyoku 渡辺海旭, et al. 1924–1934. 100 vols. Tokyo: Taishō Issaikyō Kankōkai. X Shinsan dainippon zoku zōkyō 新纂大日本續藏經 (Wan xinzuan xuzang 卍新纂續藏). editors, Kawamura Kōshō 河村孝照 et al. 1975–1989. 90 vols. Tokyo: Kokusho Kankōkai.

References

Primary Sources

Dabore boluomiduo jing 大般若波羅蜜多經 T220
Miaofa lianhua jing 妙法蓮華經 T262
Fodijing lun 佛地經論 T1530
Dasheng guangbailun shilun 大乘廣百論釋論 T1571
Dasheng zhangzhen lun 大乘掌珍論 T1578
Cheng weishi lun 成唯識論 T1585
Bian zhongbian lun 辨中邊論 T1600
Dabore boluomiduojing bore liqufen shuzan 大般若波羅蜜多經般若理趣分述讚 T1695
Bore boluomi duoxinjing youzan 般若波羅蜜多心經幽贊 T1710
Foshuo bore boluomi duoxinjing zan 佛說般若波羅蜜多心經贊 T1711
Miaofa lianhuajing zuanzan 妙法蓮華經玄贊 T1723
Shuo wugouchengjing shu 說無垢稱經疏 T1782 Cheng weishilun shuji 成唯識論述記 T1830
Cheng weishilun zhangzhong shuyao 成唯識論掌中樞要 T1831
Weishi ershi lun shuji 唯識二十論述記 T1834
Bian zhongbianlun shuji 辨中邊論述記 T1835
Yinming ru zhengli lun shu 因明入正理論疏 T1840
Dasheng fayuan yilin zhang 大乘法苑義林章 T1861
Song gaoseng zhuan 宋高僧傳 T2061
Da Tang xiyu ji 大唐西域記 T2087
Kaiyuan shijiao lu 開元釋教錄 T2154
Shengmanjing shuji 勝鬘經述記 X352
Dasheng apidamo zajilun shuji 大乗阿毘達磨雑集論述記 X796
Chengweishilun liaojian 成唯識論料簡 X806
Yibuzong lun lun shuji 異部宗輪論述記 X844

Secondary Sources

Fukaura Seibun 深浦正文. 1954. Yuishikigaku kenkyū 唯識學研究. 2 vols. Vol. 1. Kyōto: Nagata Bunshōdō. Hayashi Kana 林香奈. 2009. Ki sen Daijō abidatsuma zōshū ron jukki no seiritsu to shisō 基撰『大乗阿毘達磨雑集論述記』の成立と思想. Indogaku Bukkyōgaku kenkyū 印度學佛教學研究. 58:252–255.

Hayashi Kana 林香奈. 2012. Ki sen to sareru ronsho oyobi Daijō hōen girin shō no seiritsu katei ni tsuite 基撰とされる論疏および『大乘法苑義林章』の成立過程について. Pulgyo hakpo. 61:189–212. Lee, S. Forthcoming 2015. Redefining the “Dharma Characteristics School” in East Asian Yogācāra Buddhism. The Eastern Buddhist. 46. Forthcoming. Tendaishū Shūten Kankōkai 天台宗宗典刋行會, editor. 1912. Dengyō Daishi zenshū 傳敎大師全集. 5 vols. Vol. 3. Tōkyō: Tendaishū Shūten Kankōkai 天台宗宗典刋行會. Shigeki Moro 師茂樹. 2004, October. Hyŏnjang ŭi Yusik piryang kwa Silla Pulgyo: Ilbon ŭi munhŏn ŭl chungsim ŭro현장(玄奘)의유식비량(唯識比量)과신라불교: 일본의문헌을중심으로. Paper presented at: First International Conference on Buddhist Studies of Kŭmgang University 제1차 금강대학교국제불교학술회의; Korea. Sponberg, A. W. 1979. Vijñaptimātratā Buddhism of the Chinese monk K’uei-Chi (A.D. 632-682) [Ph. D. dissertation]. University of British Columbia; p. 2–20. Wei Tat 韋達, trans. 1973. Ch’eng Wei-shih Lun: the doctrine of mere-consciousness. Hong Kong: Ch’eng Wei-shih Lun Publication Committee. Weinstein, S. 1959. A biographical study of Tz’u-En. Monum Nihon. 15:119–149. Yoshimura Makoto 吉村誠. 1999. Genjō no Daijōkan to santen bōrin setsu 玄奘の大乗観と三転法輪説. Dōyō no shisō to shūkyō 東洋の思想と宗教. 16:58–62. Yoshimura Makoto 吉村誠. 2004. Yuishiki gakuha no goshō kakubetsu setsu ni tsuite 唯識学派の五姓各別説について. Komazawa Daigaku Bukkyōgakubu kenkyū kiyō 駒沢大学仏教学部研究紀要. 62:223–258. Yoshimura Makoto 吉村誠. 2009. Tō shoki no Yuishiki gakuha to busshō ronshō 唐初期の唯識学派と仏性論争. Komazawa Daigaku Bukkyōgakubu kenkyū kiyō 駒沢大学仏教学部研究紀要. 67:310–296.