Jiang on Ri, '『究竟一乗宝性論』と東アジア仏教: 五─七世紀の如来蔵・真如・種姓説の研究'

Shisho Ri
Tianren Jiang

Shisho Ri. 『究竟一乗宝性論』と東アジア仏教: 五─七世紀の如来蔵・真如・種姓説の研究. Tokyo-to Itabashi-ku: Kokusho Kankokai, 2020. iv + 662 pp. ISBN 978-4-336-06454-7

Reviewed by Tianren Jiang (Heidelberger Centrum für Transkulturelle Studien, Ruprecht-Karls-Universität Heidelberg) Published on H-Buddhism (October, 2021) Commissioned by Jessica Zu (USC Dornsife, School of Religion)

Printable Version: https://www.h-net.org/reviews/showpdf.php?id=56967

Is Gotra Ontologically Different from Tathatā in the Ratnagotravibhāga?

This monograph by Zijie Li is a study of the Jiujing yisheng baoxing lun 究竟一乗宝性論 (T1611, henceforth BXL), the only Chinese translation of the Ratnagotravibhāga (henceforth RGV), produced by *Ratnamati 勒那摩提 in 511 CE.[1] Li’s study focuses on the history of translation and acceptance of the ideas of “Tathāgatagarbha-containing” (Ch: rulaizang 如來藏, Skt: tathāgatagarbha), "Buddha Nature" (Skt: buddhadhātu), “Thusness” (Ch: zhenru 真如, Skt: tathatā), and the theory of “lineage” (Ch: zhongxing 種姓/種性, Skt: gotra) in East Asia from the fifth to the seventh century CE. In this study, Li tries to take a historical-philological approach to elucidate how an exegetical mode of translation of Sanskrit texts into Chinese resulted in doctrinal deviation from the Sanskrit original, which, Li contends, opens a horizon on the specific relationship between the Chinese translation of Buddhist texts and the Sinification of Buddhist thought.

Before summarizing the contents of this book, we should note that the presumption of this “historical” study is a philosophical hypothesis: that is, Shirō Matsumoto’s 松本史朗 dhātuvāda hypothesis regarding Tathāgatagarbha and Yogācāra thought. In a nutshell, the dhātuvāda hypothesis argues that gotra (or lineage) should be ontologically different from buddhadhātu (or Buddha Nature) in the Indic tradition. This hypothesis is used as a presumption by Matsumoto to interpret RGV. Following Matsumoto, Li also applies this presumption to the study of BXL, and thereby when he sees that gotra and buddhadhātu are conflated as one ontological entity in BXL, he argues that this is a “Sinified” doctrine that deviates from the Indic tradition. However, I think that whether the said conflation of gotra and buddhadhātu is actually absent in the Indic tradition requires further discussion, and it is imprudent to base his historical study on such a presumption. In the following, I first present Matsumoto’s dhātuvāda hypothesis. Then I summarize Li’s main arguments, which are based on dhātuvāda. Lastly, I explain why I see dhātuvāda as a problematic presumption, and therefore Li’s arguments as problematic.

Matsumoto believes that the Tathāgatagarbha and Yogācāra traditions share an ontological model of dhātuvāda (theory of locus). They argue for a non-empty basis (locus) underlying all dharmas, and so constitute a “generative monism” or “foundational realism” similar to non-Buddhist ātman theory. Matsumoto opposes this model to the Śūnyavāda, which posits that “all is empty, so the locus is also empty,” which he considers to be the authentic Buddhist thought.[2]

In the case of the Tathāgatagarbha tradition, according to Matsumoto, all beings are said to have buddhadhātu, or synonymously tathāgatagarbha, as the monistic ontological locus. However, only some of them have the gotra required as the cause of/for awakening (bodhihetu), which is based on the locus.[3] Even though they share the monistic locus of buddhadhātu or tathāgatagarbha, those who have no gotra can never achieve awakening, and are pejoratively named icchantika. Therefore, Matsumoto sees a dhātuvāda structure in Tathāgatagarbha thought. The seemingly egalitarian concept of Buddha Nature is belied by an ineradicable inequality in the theory of gotra, which, Matsumoto holds, resembles the caste system in India.[4] Following Matsumoto, Li believes that gotra is a different ontological entity from buddhadhātu or tathatā in Tathāgatagarbha thought. This idea constitutes the premise for a series of arguments in his book.

This book consists of six chapters with an introduction and a conclusion. Chapter 1 purports to identify the factors that influenced the choice of Chinese terminology used in BXL to translate the Sanskrit counterparts in RGV. The most important point in the first chapter is that it was a convention in the Chinese translation of Buddhist texts to translate gotra as *buddhadhātu (foxing 佛性) since *Dharmakṣema 曇無讖 (385–433 CE); Li holds that this likely had an impact on BXL, which also renders gotra as foxing (佛性) or zhenru foxing (真如佛性). Li infers that this shared emphasis on foxing/buddhadhātu, and neglect of gotra, betrays a subjective understanding of the Sanskrit text among *Dharmakṣema’s Chinese collaborators, who might have then exerted an influence on *Ratnamati’s translation team. In this sense, he regards the conflation of gotra and buddhadhātu as an instance of “Sinification.”

Chapter 2 focuses on the history of the reception of the Lakāvatārasūtra (LAS) in China. This chapter seems to intrude somewhat abruptly in the argument, as its association with BXL is not clearly discussed by the author. Chapter 3 continues to develop the argument of the first chapter. It argues that BXL’s rendering of gotra as *buddhadhātu/tathatā (foxing or zhenru foxing) led to the Sinitic version of Buddha Nature and Thusness theory, in which the idea of gotra that, according to Li and Matsumoto, must be distinctly differentiated from buddhadhātu is missing. The author contends that gotra should be a conditioned dharma, whereas buddhadhātu/tathatā is an unconditioned dharma. In his view, then, this problematic translation mixed up those two categories of dharmas, resulting in the emergence of the idea that the ultimate unconditioned reality (= buddhadhātu/tathatā) can act as a cause of worldly phenomena. This doctrine is not seen in any Indic sources. Chapters 4-6 then investigate further how this Sinitic conception of BXL affected the construction of Buddhist metaphysics in East Asia down to the seventh century CE.

Thus, in a nutshell, the primary argument of this study is that gotra, as a conditioned dharma, is ontologically different from buddhadhātu as an unconditioned dharma; but BXL erroneously erased or elided this distinction, and thus gave rise to the Sinitic version of tathāgatagarbha, which was hugely influential in East Asia. As this summary shows, the stakes in Li’s book are high, and his argument aims at issues with far-reaching historical significance.

However, in my view, Matsumoto and Li can hardly justify the dhātuvāda reading of RGV, which argues that “gotrabuddhadhātu.” In my view, even in RGV itself, before the Chinese translation, the referents of those two terms are ontologically the same thing, even though they are different in a soteriological sense. The reasoning by which I arrive at this conclusion requires several steps.

First, we need to note a set of equivalences that include buddhadhātu. A passage in RGV regarding “tathatā with defilement” and “tathatā without defilement” clearly spells out the ontological equivalence between tathatā, [buddha-]dhātu/tathāgatagarbha and dharmakāya: “Here, ‘tathatā with defilement’ is the [buddha-]dhātu not liberated from the sheath of defilements, which is called tathāgatagarbha. ‘Tathatā without defilement’ is the very same thing when it is characterized by the Transformed Basis in the Stage of Buddha, which is called the dharmakāya of the Tathāgata.”[5] Thus, while buddhadhātu could be designated by both of the terms tathāgatagarbha and dharmakāya, it is identical to tathatā (in its defiled aspect) as the same ontological entity.

Whether tathatā is covered by defilement or not has nothing to do with whether tathatā, [buddha-]dhātu, tathāgatagarbha, and dharmakāya have the same ontological referent. The difference between those terms is merely soteriological, not ontological. In other words, all those terms denote the same existent, the same ontological entity as the ultimate reality, despite that this entity is differently named according to its soteriological status, namely, whether it is covered by defilements or not, or whether it is realized or not from a practitioner’s perspective. Specifically, tathāgatagarbha denotes the tathatā, which is concealed by defilements, and therefore, tathāgatagarbha and tathatā mean the same ontological entity though this entity is called tathāgatagarbha in soteriological terms when it is in the “causal stage” (*hetu-avasthā), namely, when one has not become a buddha; dharmakāya is a term for the tathatā that has been disclosed from defilements, and is soteriologically in the “fruit stage” (*phala-avasthā), namely, when one has become a buddha, but again dharmakāya and tathatā are not ontologically different as they denote the same thing.[6] Also, both [buddha-]dhātu and tathatā refer to the ontological reality in spite of their soteriological status, so they can be used interchangeably. To put it metaphorically, buddhadhātu or tathatā is like the sun, and defilements are like darkness. The sun is called tathāgatagarbha to those whose eyes are obscured by the darkness and do not see the sun directly, and the sun is called dharmakāya when darkness has been completely removed and it can be seen directly. There is only one sun, like the unconditioned ultimate truth, which itself is not affected by darkness. In this passage, tathāgatagarbha, dharmakāya, buddhadhātu, and tathatā have the same ontological referent.

On the other hand, gotra, together with tathatā and dharmakāya, is elsewhere defined as the threefold nature of tathāgatagarbha.[7] Since, as we just saw, tathatā, dharmakāya, and tathāgatagarbha are described as ontologically identical in the text, it strongly indicates that gotra is also a synonym with those terms in an ontological sense. The threefold nature of [buddha-]dhātu is also explicated as comprising dharmakāya, tathatā, and gotra.[8] Furthermore, in the discussion about the cause of Three Bodies of the Tathāgata, the text clearly defines gotra, or tathāgatagarbha, or buddhadhātu, as the same and single cause of all Three Bodies (including the dharmakāya).[9] Therefore, we can ascertain the equality of “gotra = tathāgatagarbha = buddhadhātu” on that basis. Thus, all those terms, tathatā, dharmakāya, tathāgatagarbha, buddhadhātu, and gotra, share one intrinsic nature, the difference among them is not ontological, and they are one unconditioned ultimate reality. In fact, RGV frequently suggests that gotra is identical with tathatā, dharmakāya, tathāgatagarbha, and buddhadhātu, which is demonstratively proved by Ching Keng’s comprehensive research on the pertinent passages in the text.[10]

In fact, although Matsumoto proposes to interpret RGV through the lens of dhātuvāda, he also admits that the author of RGV is inclined to show that gotra and dhātu are ontologically the same.[11] However, if gotra and dhātu are ontologically the same in RGV, then there is no reason to interpret RGV as dhātuvāda which differentiates gotra from the ultimate reality in the first place. Li seems to be unaware of this self-contradiction in Matsumoto’s reading of RGV. Li claims that he agrees with Matsumoto that RGV tends to equate gotra with dhātu; meanwhile, he insists on the presumption that gotra and dhātu should be different things in the Indic Buddhist thought, and it is the Chinese translation that breaks the distinction between gotra with dhātu. I would like to see textual evidence by which Li and Matsumoto can argue that gotra and dhātu should be different, while at the same time admitting gotra and dhātu are identical in RGV; as far as I have seen, however, they present none in their publications.

Noticeably, the gotra theory in RGV teaches only buddhagotra, which is very different from the gotra theory of Yogācāra school which teaches several kinds of gotra. Therefore, even if, for the sake of argument, Matsumoto and Li were successful in showing that the gotra theory in the Yogācārabhūmi is dhātuvāda, we should not assume a doctrinal coherence between the Tathāgatagarbha thought of RGV and Yogācāra thought of Yogācārabhūmi. In fact, however, their theories are problematic on this point, too, because whether the gotra theory in Yogācārabhūmi is dhātuvāda is still under contestation.[12]

In sum, the central argument of this study is that BXL, the Chinese translation of RGV, mixes up gotra, a conditioned dharma, and dhātu or buddhadhātu, which is equivalent to tathatā, an unconditioned dharma, resulting in the doctrine of the causation of the unconditioned dharma, which is unique to China. In my opinion, this thesis is not convincing, because the underlying assumption that this position represents dhātuvāda is problematic, and presumes a questionable ontological interpretation to RGV. Matsumoto's ontological theory has a lasting impact on the studies of Chinese Buddhist philosophy. Li's monograph is one of the most recent examples. However, it would be better if Li had taken the new development of the study of RGV, such as Keng's article, into account in his monograph.[13] Indeed, it seems to me that historical studies of the positions in primary texts should not set out from this kind of presupposition but should rather investigate the texts, as closely as possible and as far as possible, on their own terms.


[1]. I am much indebted to Michael Radich (Heidelberg University) and Ching Keng (National Taiwan University) for a careful reading and insightful comments on this reivew.

[2]. Shirō Matsumoto 松本史朗, Bukkyō Shisōron/Jyō 仏教思想論・上(Tokyo: Daizoshuppan, 2004), 55-56.

[3]. The Sanskrit term gotra is rendered into two Chinese terms with the same pronunciation as zhongxing 種性/種姓. Although there may be debates regarding whether two different xing, 性 and 姓, may cause different understandings of the idea of gotra, from my perspective, the point in question now is how to understand the idea of gotra in RGV. Without a solid understanding of gotra in its original meaning in Sanskrit, it will be impossible to consider to what extent those two different xing can cause ambiguity.

[4]. Shirō Matsumoto 松本史朗, Bukkyō Shisōron/Ge 仏教思想論・下 (Tokyo: Daizoshuppan, 2013), 63-66.

[5]. Edward Hamilton Johnston, ed., Ratnagotravibhāga Mahāyānottaratantraśāstra by Asaṅga (Patna: Bihar Research Society, 1950), 21.8-9: tatra samalā tathatā yo dhātur avinirmuktakleśakośas tathāgatagarbha ity ucyate | nirmalā tathatā sa eva buddhabhūmāv āśrayaparivṛttilakṣaṇo yas tathāgatadharmakāya ity ucyate||

[6]. Jikidō Takasaki 高崎直道, Hōshō ron/Hokkai musabetsu ron 宝性論・法界無差別論 (Tokyo: Daizoshuppan, 1999), 31.

[7]. Johnston, ed. Ratnagotravibhāga Mahāyānottaratantraśāstra, 69.18-20: trividha svabhāvam adhikrya cittavyavadānahetos tathāgatagarbhasya navadhā buddhabimbaādisādharmyam anugantavyam | trividha svabhāva katama | svabhāvo dharmakāyo 'sya tathatā gotram ity api | tribhir ekena sa jñeya pañcabhiś ca nidarśanai|| For English translation, see Jikidō Takasaki, A Study on the Ratnagotravibhāga (Uttaratantra): Being a Treatise on the Tathāgatagarbha Theory of Mahāyāna Buddhism, Serie Orientale Roma 33 (Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1966), 284. In personal contact, Ching Keng advices that RGV I.149 describes gotra as twofold, which makes it difficult to consider that gotra, tathatā, and dharmakāya are the same ontological entity. Ratnagotravibhāga, 71.18-19: gotraṃ tad dvividhaṃ jñeyaṃ nidhānaphalavṛkṣavat | anādiprakṛtisthaṃ ca samudānītam uttaram|| However, following Takasaki's translation, I think this means "gotra [of the Buddha] is known to be twofold," so it seems to be a description of gotra from an epistemological angle but may not be read as an ontological definition of gotra. It may be reasonable to conisder "twofold gotra" as one ontological entity that can be epistemologically known as twofold.

[8]. Johnston, ed. Ratnagotravibhāga Mahāyānottaratantraśāstra, 69.21-70.1: tribhir buddhabimbamadhusāradṛṣṭāntair dharmakāyasvabhāva sa dhātur avagantavya | ekena suvaradṛṣṭāntena tathāgatasvabhāva | pañcabhir nidhitaruratnavigrahacakravartikanak abimbadṛṣṭāntais trividhabuddhakāya utpattigotrasvabhāva iti|| For English translation, see Takasaki, Study on the Ratnagotravibhāga, 284.

[9]. Johnston, ed. Ratnagotravibhāga Mahāyānottaratantraśāstra, 72.7-12: ity evam ebhir avaśiṣṭai pañcabhir nidhitaruratnavigrahacakravartikanak abimbadṛṣṭāntais trividhabuddhakāyotpattigotrasvab hāvārtham adhiktya tathāgatadhātur eā garbha sarvasattvānām iti paridīpitam | trividhabuddhakāyaprabhāvitatva hi tathāgatatvam | atas tatprāptaye hetus tathāgatadhātur iti | hetvartho 'tra dhātvartha | yata āha | tatra ca sattve sattve tathāgatadhātur utpanno garbhagata savidyate na ca te sattvā budhyanta iti|| For English translation, see Takasaki, Study on the Ratnagotravibhāga, 289-90.

[10]. Ching Keng, “Lun Rulaizang sixiang zai jietuoxue shang de genben kunnan — yi Bao xin lun wei zhongxin de tantao 論如來藏思想在解脫學上的根本困難—以《寶性論》為中心的探討,” Hanyu Foxue pinglun 漢語佛學評論 3 (2013): 201-31.

[11]. Matsumoto, Bukkyō Shisōron/Ge, 77.

[12]. Nobuyoshi Yamabe, “Once Again on ‘Dhātu-vāda,’” Critical Review for Buddhist Studies 21 (2017): 9-43.

[13]. Keng, “Lun Rulaizang sixiang zai jietuoxue shang de genben kunnan.”

Editor’s note: A previous version of this review incorrectly glossed “Buddha Nature” as “dhātuvāda.” The correct term in Sanskrit, “buddhadhātu,” has been used instead in the current version.

Citation: Tianren Jiang. Review of Ri, Shisho, 『究竟一乗宝性論』と東アジア仏教: 五─七世紀の如来蔵・真如・種姓説の研究. H-Buddhism, H-Net Reviews. October, 2021. URL: https://www.h-net.org/reviews/showrev.php?id=56967

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.