Doxographical Appropriation of Nāgārjuna’s Catuṣkoṭi in Chinese Sanlun and Tiantai Thought

Abstract

This article reconstructs the Chinese “practice qua exegesis” which evolved out of the doxographical appropriation of the Indian Buddhist catuṣkoṭi (four edges), a heuristic device for conceptual analysis and a method of assorting linguistic forms to which adherents of Madhyamaka ascribed ambiguous potential. It could conceptually clarify Buddhist doctrine, but also produce deceptive speech. According to the Chinese interpreters, conceptual and linguistic forms continue to be deceptive until the mind realizes that all it holds on or distinguishes itself from is its own fabrication. Liberation from such self-induced deceptions requires awareness of the paradox that the desire to leave them behind is itself a way of clinging to them. Chinese Sanlun and Tiantai masters tried to uncover this paradox and to disclose to practitioners how the application of the catuṣkoṭi, on the basis of such awareness, enables proper conceptual analysis in exegesis. From this approach followed the Chinese habit of construing doxographies in which hermeneutical and soteriological intent coincide. Understanding the inner unity of doctrinal manifoldness in the translated sūtra and śāstra literature from India via exegesis also made it possible to apprehend the ineffable sense of liberation.

1. Introduction: Deconstruction of the Conceivable and Inconceivable in Chinese Buddhism

The first indigenous Buddhist schools in East Asia—Tiantai 天台宗, Sanlun 三論宗, and Huayan 華嚴宗—emerged in the era of the Sui (581–618) and Tang (618–907) dynasties, after the transmission of Indian sūtra and śāstra literature to China had triggered a process of translation and exegetical activity. Supervised by Central Asian and Indian Buddhist masters, whose presence in China is attested since the end of the Han Dynasty (202 BC–220 AD), translation projects were usually promoted by local aristocrats or supported by the imperial court. The later schools descended from the exegetical traditions which arose due to some influential scholar monks who attempted to integrate Buddhist doctrine from India into the discourses of traditional Chinese thought.

Tiantai, Sanlun, and Huayan masters inherited from their predecessors certain methods of interpreting the translated sūtras and śāstras. The most common exegetical tool at that time was the means of doctrinal classification (doxography, taxonomy of doctrine, panjiao 判教). The hermeneutical purpose of this was to disclose to the Chinese practitioners the inner coherence between Indian Sarvāstivāda, Prajñāpāramitā, Madhyamaka, Yogācāra, and Tathāgatagarbha elaborations on the Buddhadharma (fofa 佛法, law of the Buddha). Irrespective of their differences in interpretation, all three schools resorted to the same conceptual blueprint, which was the tetralemma (catuṣkoṭi, siju 四句), frequently occurring in Kumārajīva’s (344–413) Chinese translations of Indian Madhyamaka texts.

Moreover, the hermeneutical motif was inextricably linked with the soteriological intent of showing that the various textual representations of the Buddhadharma are congruent with the meaning of awakening—the mind’s liberation from all self-induced deceptions. Thus, the prospected yet hidden sense of coherence was considered to be the highest meaning of the Dharma, ultimately independent from speech and concept. The Chinese practitioners, authors, and interpreters developed an ambiguous stance toward the textual transmission of Buddhist doctrine: the independence of its ultimate meaning from speech, on the one hand, and the indispensability of the canonical word in the understanding of that meaning, on the other (Kantor 2019c, pp. 8–11).

Such ambivalence of Chinese Buddhist hermeneutics was inspired by the Indian doctrine of the “four reliances” (catvāri-pratisaraṇāni, siyi 四依) mentioned in both the Mahāparinirvāṇa-sūtra (T12, no. 375, p. 642, a21–24) as well as the Da zhi du lun 大智度論 (*Mahāprajñāpāramitopadeśa), which Tiantai master Zhiyi 智顗 (538–597), Sanlun master Jizang 吉藏 (549–623), and Huayan master Fazang 法藏 (643–712) quoted in their own treatises and commentarial works on the sūtras and śāstras. Two of those four state: “Rely on the dharma but not on the person (yifa buyi ren 依法不依人); rely on the meaning but not on the words (yiyi buyiyu 依義不依語)” (Lamotte 1988).

Chinese interpreters derived from this their conviction that all textual manifestations of Buddhist doctrine had been designed for the multitude of non-awakened persons, while its ultimate meaning going beyond word and thought was instantiated only by the awakened—the Buddha. According to this vantage point, the canonical word and its ultimate meaning were neither identical nor separate, and the specific relationship between text and meaning implied both congruity and incongruity (D’Ambrosio et al. 2018, pp. 305–30).

For the Chinese, incongruity characterized such a relationship insofar as all linguistic expressions were considered to be false/provisional names (Jiaming 假名) which at best denote nominal existence but not real entities1, while congruity was seen to be true in the specific sense that the speech of the awakened, though consisting of false/provisional names, is not misguiding, as it can unveil falsehood of which the non-awakened is not aware.2

Hence, two types of speech had to be distinguished: (1) the word of the Buddha (Buddha-vacana, foyan 佛言), doctrinal discourse in sūtra and śāstra (yanjiao 言教), which uses falsehood for heuristic purposes and thus is instructive3, whereas (2) speech of the non-awakened unaware of its blindspots would lead listeners astray, and therefore was deemed as deceptive and called “verbal/conceptual proliferation” (prapañca, xilun 戲論, Salvini 2019, pp. 663–65; Siderits 2019, pp. 645–61). The ambivalent stance toward the transmission of the canonical word and the correspondent view about opposite types of conventional speech are concatenated characteristics of Chinese Buddhist hermeneutics, deeply rooted in the tradition of Prajñāpāramitā and Madhyamaka thought, and expressed by means of doxographies (Kantor 2019c, pp. 8–11).

For instance, the adherents of Tiantai and Sanlun believed that the inconceivable sense of liberation (acintya-vimokṣa, busiyi jietuo 不思議解脫) gives rise to the Buddha’s thoughts, speeches, and physical activities, transforming the world of non-awakened beings. The inconceivable was referred to as the hidden root (ben 本) from which conceptual differences in terms of doctrine were believed to descend, while, conversely, the diversity of doctrine in sūtra and śāstra was regarded as the visible traces (ji 跡) that lead the practitioners back to the hidden root (Kantor 2019a, pp. 851–915). The Chinese constructions of doxographies aimed to outline that type of soteriological circularity. Tiantai and Sanlun interpreted liberation—the hidden meaning (xuanyi 玄義), ineffable ultimate, or root (ben)—in terms of both the source and the goal of all the differing teachings—the traces (ji) visible as sūtra and śāstra texts.

To enact such circularity between text and meaning, Tiantai and Sanlun emphasized the non-duality of the visible and hidden in the Buddhadharma, called the“non-duality of root and traces”.4 Reconciliation of conceptualization with the inconceivable was a major concern of the hermeneutical project in all Chinese denominations. Particularly, the dialectical approaches of Tiantai’s, Sanlun’s, and Huayan’s doxographies focused on deconstructing the perceived duality of doctrine and liberation. The hermeneutics of deconstructive non-duality concurs with soteriology that intends to embody the ineffable via the practice of exegesis and doctrinal discourse.

The Chinese habit of construing doxographies exemplified the specific idea of “practice qua exegesis”: Buddhist wisdom, the mind’s liberation from self-induced deception, had to be accomplished by specializing in exegesis of text and doctrine.5 Wisdom-practice implicated awareness of the paradox that liberation from conceptual and linguistic hypostatization does not distance the mind from speech and thought (Kantor 2021, forthcoming).

The model which inspired the Chinese construction of doxographies was the distinction between Hīnayāna (Small Vehicle) and Mahāyāna (Great Vehicle), made by those Indian sūtras and śāstras which claimed that they themselves were promulgating the Great Vehicle. However, there was no accurate standard for such subdivision of doctrine, as it was differently defined depending on the respective denominations to which the authors or compilers of those texts were committed.

The view which was particularly influential in the Tiantai and Sanlun interpretations was that of the Prajñāpāramitā-sūtras, considering two major aspects of the Great Vehicle: (1) the insight that things arising and perishing in the worldly realm of sentient beings (saṃsāra) are compoundphenomena whose constituents are empty of an intrinsic nature, or, empty of inherent existence, and that this emptiness (śūnyatā, kong 空) is the true nature of all things which arise due to mutual dependency, called “pratītyasamutpāda” (yuanqi 緣起, conditioned arising); (2) the ultimate liberation from one’s own delusions, inspiring the liberation of other sentient beings.

Chinese Tiantai, Sanlun, and Huayan doxographies assorted differing levels of Hīnayāna and Mahāyāna to interpret the variety of Buddhist concepts in a way that the supreme meaning of the Dharma can be comprehended in a gradual progress. The highest level, referred to in Tiantai and Huayan as the “round/perfect teaching” (yuanjiao 圓教), retained a sense of circularity: it constitutes and sublates Hīnayāna and Mahāyāna, that is, generates and suspends all doctrinal distinctions. Tiantai, Sanlun, and Huayan offered varying accounts of that thought, although it was the same conceptual figure of the catuṣkoṭi which had inspired them.

Western studies use “tetralemma” (four alternative assumptions) for the Sanskrit “catuṣkoṭi” (four edges, four points) and for its Chinese translations “four [alternative] phrases” (siju 四句) or “four [alternative] gates” (simen 四門)—briefly: “four alternatives”. As for the origin, use, and philosophical meaning of the catuṣkoṭi, Ruegg’s (1977, pp. 1–71) detailed article mentions that this scheme of classifying conceptual forms is already attested in the early philosophical literature of Indian Buddhism. The discussion in which its four alternative positions (koṭi) are raised concerns questions such as whether a tathāgata (Buddha) exists after death, whether the world has an end, and whether the world is eternal. Seyfort Ruegg explains: “In each of these cases the nature of a postulated entity and its relation to its predicate is investigated in such a way that all conceptually imaginable positions are exhausted; for an entity and its predicate can be conceptually related only in terms of these four limiting positions,” (Ruegg 1977, p. 2).

In other words, the tetralemma (catuṣkoṭi, siju) is a formal scheme of four mutually related yet distinct ways of referencing a specific doctrinal topic; it often (but not always) was considered to provide an exhaustive set of four mutually exclusive conceptual possibilities (Priest 2018, p. 23; Westerhoff 2005, pp. 367–95): (1) the application of a certain concept (2) the application of its negation (3) the application of both the concept and its negation, and (4) the application of neither the concept nor its negation.

In Indian Prajñāpāramitā and Madhyamaka literature, particularly in Nāgārjuna’s Mūlamadhyamaka-kārikā (Zhong lun 中論), there is both (a) the positive (cataphatic) and (b) the negative (apophatic) application of the four alternatives: (a) all the four are asserted; (b) all of them are rejected (Westerhoff 2005, pp. 367–95). In the Chinese interpretation of Sanlun master Jizang, the apophatic and cataphatic versions of the catuṣkoṭi indicate a strong tendency to merge into one another. In his commentary on Nāgārjuna’s Kārikā—the Zhongguan lun shu 中觀論疏, Jizang’s dialectical notion of “suspension (sublation) of the four alternatives” (jue siju 絕四句) highlights the thought that it is the same principle which constitutes and deconstructs all four.

Moreover, Tiantai, Sanlun, and Huayan developed interpretations in which all positions of the cataphatic catuṣkoṭi appear to be equally valid. Zhang’s recent article explains that the catuṣkoṭi can be described as a Buddhist form of dialectics which “consists of four arguments or perspectives, only one of which can be true at a time” (Zhang 2016, pp. 25–71). According to the Chinese interpretations, however, this cannot be mistaken in the way that only one perspective apart from the other three is valid, as none of the four represents an independent logical parameter. The position that excludes the other three is a construction leading the mind astray. Sanlun, Tiantai, and Huayan texts emphasize that such misconception is a symptom of the clinging and non-awakened mind.6

Therefore, Chinese Buddhist traditions regarded the linguistic–conceptual figure of the catuṣkoṭi as a symbol for both the awakened understanding of the textual transmission of Buddhist doctrine and its exact opposite—the non-awakened mind inconsonant with, or unaware of the Buddhadharma. The non-awakened mind was considered to be impeded by or entangled in its own conceptual constructions.

In his classic on mind-contemplation (guanxin 觀心), the Mohe zhiguan 摩訶止觀, Tiantai master Zhiyi, like Sanlun master Jizang in his commentary on the Vimalakīrti-nirdeśa-sūtra, the Jingming xuanlun 淨名玄論, developed the list of “ten types of catuṣkoṭi” (shizhong siju 十種四句) to demonstrate that our understanding of the four alternatives must not be confined to only one valid reading. The proper understanding must learn to see that it is the same catuṣkoṭi which can be either (1) the source of misguiding conceptualization or (2) the heuristic and analytic device for those who seek liberation from such deception.

Hence, the catuṣkoṭi in Tiantai, Sanlun, and also Huayan figures as a linguistic symbol which mirrors the ambivalence of conventional speech that could be divided, on the one side, into doctrinal discourse (yanjiao) as transmitted in sūtra and śāstra and, on the other, into verbal or conceptual proliferation (prapañca, xilun) of those who are unaware of the Buddhadharma. Those opposites were together epitomized through the same symbol of the catuṣkoṭi.7

Seen from this vantage point, the ambiguity of the catuṣkoṭi can be compared to the notion of “pharmaca”. In spite of their therapeutic purpose, pharmaca can be toxic. Their functioning as medicine does not eradicate their potential for poison, just as their poisonous effect retains potential for being medicine. Each has the potential to be its own reverse. The two are opposites yet inseparably one. The catuṣkoṭi is a symbol for such ambivalence in conventional speech. For the Chinese interpreters, the non-duality of the salutary and harmful, symbolized by the catuṣkoṭi, provides the condition that allows the deluded mind to transform itself into the opposite, liberation.

On account of such ambiguity, Buddhist masters in China were able to develop their respective doxographical frameworks (panjiao). In what follows, the article examines (1) Nāgārjuna’s catuṣkoṭi, as it figures in the Chinese transmission of Madhyamaka, and (2) the specific influence of this analytic method on the formation of doxography and the associated practice of hermeneutical deconstruction in Chinese Tiantai and Sanlun exegesis.

2. Nāgārjuna’s Application of the Catuṣkoṭi According to Kumārajīva’s Transmission

Nāgārjuna’s main work, the Mūlamadhyamaka-kārikā (Root Verses on the Middle Way), was translated by Kumārajīva in 409. Its Chinese title is called Zhong lun 中論 (Treatise on the Middle), and besides Nāgārjuna’s verses, the translation also contains one of the oldest existent commentaries ascribed to a person called Qingmu 青目 (*Vimalākṣa) or Binjialuo 賓伽羅 (*Piṅgala). The compiled text of the Zhong lun—Nāgārjuna’s verses and Piṅgala’s commentary, neither existent in Sanskrit nor in Tibetan—was highly influential throughout the history of Chinese Buddhist thoughts. Jizang’s and Zhiyi’s works borrowed a lot from this early Indian Madhyamaka treatise; particularly, the use of the catuṣkoṭi was relevant to the development of Tiantai and Sanlun. The Chinese masters usually did not differentiate between Nāgārjuna and Piṅgala when they quoted from Kumārajīva’s Zhong lun.8

All Chinese schools took Nāgārjuna’s positive (cataphatic) and negative (apophatic) applications of the four alternatives in account and construed their doxographies on that basis. As for the apophatic or negative sense, the first chapter of Nāgārjuna’s Mūlamadhyamaka-kārikā (Zhong lun) provides an example which has been very influential in Tiantai and Sanlun thoughts. In that chapter, Nāgārjuna investigates the notion of “condition” (pratyaya, yuan 緣), defined as a certain thing which has the capacity to give rise to another thing. However, in his discussion, he rejects the four possibilities of “arising” (utpāda, sheng 生): arising on account of either (1) itself, or (2) something else, or (3) both, or (4) neither (Zhonglun, T30, no. 1564, p. 2, b6–17).

Assuming that, in this way, all conceptual possibilities are exhausted, he concludes that—even in reference to “pratītyasamutpāda” (yuanqi, conditioned arising)—no real arising can ultimately be asserted. There is no real entity which arises and perishes. In a paradoxical fashion, he describes the nature of conditioned arising as “non-arising” (anutpāda, bushing 不生); another term for this is “emptiness” (śūnyatā, kong 空). The commentator Piṅgala explains that such sense of negation—the exclusion of all four conceptual possibilities—equals the “ultimate meaning” (diyi yi 第一義) which is emptiness of inherent existence, the ultimate sense of true emptiness (Zhong lun, T30, no. 1564, p. 1, c12).9

However, what is said about “arising” also applies in reverse, or is true of the opposite “non-arising,” that is, “emptiness”. Hence, another well-known example of the tetralemma’s negative application exactly expresses this in chapter 22: (1) emptiness (non-arising) should not be asserted, (2) nor non-emptiness, (3) nor both, (4) nor neither (Zhonglun, T30, no. 1564, p. 30, b22–23). Nāgārjuna explains: “It is only said for the purpose of provisional designation (dan yi jiaming shu 但以假名說),” (Zhonglun, T30, no. 1564, p. 30, b23). These two examples of denying all four alternatives are exclusive negations of an exhaustive set of conceptual possibilities concerning a topic that has been considered as crucial throughout all doctrinal discourse in the history of Buddhism, namely, “arising” and the opposite “non-arising” (emptiness).10

Nonetheless, as Nāgārjuna’s last remark stresses, the four positions might alternately and provisionally be employed in changing contexts of doctrinal discourse. What legitimizes the provisional use of any of these four in conventional speech is the rejection of misunderstanding all in the ultimate sense. No form of linguistic signification applies ultimately. The negative application of the tetralemma is a linguistic device for the exclusion of ultimate meaning in the domain of speech. Negation in the sense of what asserts the opposite is to be negated, too.

Hence, Westerhoff (2005, pp. 367–95, 2009, pp. 69–92) and also Zhang (2016, pp. 25–71) correctly observe that such excluding sense is not the type of negation used to construe correlative opposites such as “emptiness and non-emptiness,” or “non-arising and arising”. Nāgārjuna’s “non-arising” in the first chapter excludes not only arising but also non-arising; in his commentary to Nāgārjuna’s Kārikā, the Zhongguan lun shu 中觀論疏, Jizang expands on this thought (T42, no. 1824, p. 11, a28–b20; p. 28, c8–19).11

However, Westerhoff’s observation is inspired by Ruegg’s (1977, pp. 43–47) hint at Candrakīrti’s commentary to the Kārikā, the Prasannapadā, in which the sixth-century Indian commentator says that the exclusive sense of negation represents the case that traditional Sanskrit grammarians call “prasajya-pratiṣedha” (non-affirmative negation); in later Chinese translations it appears as “zheqian 遮遣”. Westerhoff suggests for the Sanskrit term “exclusion negation”.

Different from “prasajya”, negation used to construe and assert correlative opposites, or negation which, for instance, implicitly validates the opposite of what is negated, such as non-being in contradistinction to being, or non-arising contrasted with arising, might correspond to the grammatical term “paryudāsa-pratiṣedha” (implicative negation), in Chinese known as “zhebi 遮蔽”. Westerhoff understands this differently, translates “paryudāsa” as “presupposition negation,” and calls “prasajya” sometimes also “cancelling presuppositions”.12

Although Nāgārjuna’s Kārikā does not use either of the grammatical terms (prasajya or paryudāsa), Westerhoff’s attempt to understand the negative application of the catuṣkoṭi in terms of this distinction seems to be helpful. The external negation of all four koṭis can be considered as prasajya negation, differing from the internal negations of the second and fourth alternatives within the catuṣkoṭi itself. Those two (the second and the fourth) could, perhaps, be subsumed under paryudāsa negation. Following this distinction between internal and external negation, or perhaps prasajya and paryudāsa, we might say that the first and the second of the four koṭis (just like the third and the fourth), relate to one another as correlative opposites.13

This is to say, the negation in the second koṭi asserts the correlative opposite of the first, just as the first is the correlative opposite of the second and therefore could be regarded as the negation of the second: emptiness in the first koṭi is the negation of non-emptiness in the second and vice versa; self-arising is the opposite of arising from something else (not from itself) and vice versa. Hence, negation in the second koṭi (denying the first) seems to function as a paryudāsa negation referencing what pertains to the domain of speech and thought.

This is different from prasajya negation which rejects the four alternatives altogether and does not allow for any affirmative form of linguistic signification. The same sense of paryudāsa negation seems to apply to the relationship of the third and fourth koṭi, as the fourth is meant to deny the third and vice versa (see Piṅgala’s and Bhāviveka’s discussion in Section 3). What seems to constitute each of these four as distinct elements pertaining to the same set is the relationship of mutually dependent or correlative opposites (xiangdai 相待).14

As to the conceptual structure of the two mentioned instances of the catuṣkoṭi in the Zhong lun, paryudāsa negation could be considered a constitutive factor for such types of interdependency, that is, the two opposites (the first and the second koṭi) negate each other, and in such mutual negation they mutually assert. To deny one side is to assert the other, which means without its opposite neither side can be held. In denying its opposite, each of the two relies on the other.

This precisely is what generates the third koṭi. As neither the first nor the second can be held apart from the other, the third comes to the fore, embracing the entire relation of those two.However, again, it is the same principle of constituting correlative opposites that also applies to this third case. What one side asserts is the opposite of the other—to assert one side is to be the opposite of its own negation, which yields the fourth as that which is the implicit opposite of the third.

Consequently, the fourth excludes what the third includes; the fourth is neither the first nor the second. This could continue infinitely, as has been pointed out by the Chinese Tiantai and Sanlun interpretations as well as by the Da zhi du lun (T25, no. 1509, p. 259, c29-p. 260, a2), which even talks about five alternatives (wuju 五句)—the abandoning of all four (she shi siju fa 捨是四句法) is the fifth. However, again, this would trigger the abandoning of abandoning, etc., and thus elicit the endless repetition of what has been displayed by the preceding positions, that is, what has already become evident with the fourth koṭi. In such a sense, the catuṣkoṭi may represent an exhaustive set of four conceptual possibilities which are mutually exclusive (or distinct).

For Nāgārjuna, all this means that, although there is no real entity to which we could attribute the property of arising, “non-arising” or “emptiness” (only denying a real entity which arises) must not exclude the meaning of unreal arising. This is so because unreality (emptiness of inherent existence) is not equivalent to nonexistence or nothingness. What is unreal might either be deceptive or instructive and therefore is existentially relevant, and yet it is not what can really be appropriated or dismissed. Hence, in a limited, provisional sense, or salutary context, unreal arising functions as a heuristic means that should not and cannot be denied.

Ultimately, the true sense of non-arising, the emptiness of inherent existence, means to be empty of an irreducible or real core; this even applies to emptiness. However, linguistic signification, which tends to confuse its referents with independently existent entities, may evoke an image of reality, falsely attributed to “emptiness,” if “emptiness of emptiness” (kongkong 空空) is not seen in conditioned arising. Because of this paradoxical sense of “emptying emptiness again” (kong yifu kong 空亦復空) there is no fixed linguistic form which can truly predicate the ultimate sense (Zhong lun T30, no. 1564, p. 33, a20–21). The true meaning must be comprehended by paradoxically denying what this expression refers to.

The paradoxical differentiation of ultimate truth from conventional truth, expounded in chapter 24, embodies this sense and must be applied to all doctrinal conceptualization.15 Objects of our epistemic–propositional references might be considered as conventionally true in the sense that they are things that arise due to causes and conditions, and yet all this is empty of inherent existence and unreal. Confined to conventional truth, such (unreal) “arising” evades ontological determinacy in terms of either being, or non-being, or both, or neither. At the level of speech and thought, the distinction between conventional and ultimate truth is therefore necessary, and yet such differentiation does not really reach beyond the linguistic level. The distinction is self-inclusive, sublating itself due to the emptiness of what is differentiated. Differentiation entails non-duality.16

Linguistic expression, unaware of this paradox, is prone to confuse unreality with reality. Madhyamaka’s paradox of non-duality qua distinction could be rephrased asthe following: To separate the conventional from the ultimate means to mistake the unreal for the real, whereas not differentiating between the two undermines the true sense of emptiness in conditioned arising. In order to see non-duality, incessant distinction is necessary. That the two truths are neither identical nor separate is a paradox but not a contradiction, because contradiction would mean the opposite: identity and separation of the two at the same time, which contradicts the true sense of emptiness.17

Nāgārjuna’s negation of both real arising and real non-arising implies the paradoxical distinction of the two truths. The two negative forms of the catuṣkoṭi are applied to validate a certain type of conventional speech—one which advances awareness of the unreality of its own referents. Critical validation, rooted in invalidation, shows the limits of thought and speech, thereby revealing the heuristic value of language and analysis. This is believed to be the type of speech—“the supreme among all types of speech” (vadatām varam, zhushuozhong diyi 諸中說第一) leading to the pacification of conceptual proliferation (prapañcopaśama, mie xilun 滅戲論). The first verse in the Kārikā (Zhong lun) literally expresses this (Siderits and Katsura 2013, p.13; Zhong lun T30, no. 1564, p. 1, c11).

On that basis, the positive application of the tetralemma can fulfill the same purpose that is ascribed to the negative application. The cataphatic use of the third instance of the catuṣkoṭi in chapter 18 explicates this: “All is real, unreal, both real and unreal, neither real nor unreal. This is called dharma of all Buddhas. 一切實非實, 亦實亦非實, 非實非非實, 是名諸佛法”. (Zhong lun T30, no. 1564, p. 24, a5–7).18

If we look at the pre-modern commentarial literature, authored by Piṅgala, Bhāviveka, Candrakīrti, as well as the Chinese Tiantai and Sanlun interpreters, Zhiyi and Jizang, the positive version of this specific catuṣkoṭi has been interpreted in various ways, all of which can nevertheless be subsumed under three associated aspects: (1) the differentiation of the two truths with respect to the first and second koṭis which appear as opposites, (2) the relationship between the two in terms of correlative dependency, and (3) the gradation of the Buddha’s teaching, which, in the case of the Chinese interpreters, led to their classifications of Buddhist doctrine (doxography).

3. Piṅgala’s and Bhāviveka’s Interpretations of Nāgārjuna’s Catuṣkoṭi in the Chinese Sources

Although Piṅgala (4th AD) and Bhāviveka (500–560) equally hold that the first two koṭis in Nāgārjuna’s positive application account for the differentiation of the two truths, they expound their views in opposite ways. For Piṅgala, the first koṭi represents ultimate truth (ultimate meaning, diyi yi 第一義), because everything is real (yiqie shi 一切實) in the specific sense that it is ultimate and true emptiness which is the real nature (shixing 實性) wherein all distinct and inexhaustible forms of conditioned arising are established; he explains:

Question: If the Buddha does not talk about the self and non-self, causing all fabrications of the mind to extinguish and bringing all speech to an end, how can he make persons become aware of the real mark of all dharmas? Answer: All Buddhas resort to the skill in devising infinite expedient means, and all dharmas are [ultimately] empty of a determinate mark. On that basis, and for the sake of saving the multitude of sentient beings, they sometimes teach that everything is real, or at another time explain that everything is unreal, or at a different moment say that everything is both real and unreal, or at some other occasion hold that everything is neither real nor unreal. [If they say that] everything is real, they set forth the real nature of all dharmas, so that all realize the ultimate meaning and the single mark of equality, called emptiness of any mark. This is like all the streams with varying colors and flavors [of water] flowing into the great ocean in which they become the single color and the single flavor. 問曰: 若佛不說我, 非我, 諸心行滅, 言語道斷者, 云何令人知諸法實相? 答曰: 諸佛無量方便力, 諸法無決定相, 為度眾生或說一切實, 或說一切不實, 或說一切實不實, 或說一切非實非不實. 一切實者, 推求諸法實性, 皆入第一義平等一相, 所謂無相, 如諸流異色異味入於大海則一色一味.

(Zhong lun T30, no. 1564, p. 25, a14–20).

What the first koṭi predicates of ultimate truth must be seen in the context of conventional speech. The manner in which it represents ultimate truth depends on the other three; the same mutual dependency of all four applies to each (see the succeeding part of this quote below). This is so, because the emptiness of an intrinsic nature is the real nature of all, which is to be empty of a determinate mark—in this sense, “everything is real”.

To mistake this first predication for the fixed form of the ultimate, excluding the other positions, contradicts its very meaning. Empty of a determinate mark, the ultimate is to be predicated in a way which reveals that there is no fixed predication of it. As this is true of all, the text illustrates the real nature with the single flavor, into which all differing flavors turn.

In such sense of mutual dependency, the second koṭi accounts for the opposite of the first, which is the perspective of the worldly realm before having attained the highest sense of the ultimate. This is unreality (feishi 非實) in the shape of the differentiated and multifaceted conventional truths that might lead to the ultimate in the same way as all flavors of water in the differing streams are capable of becoming the single flavor of the ocean. Using a similar image, the Da zhi du lun explains the relationship between the conventional and the ultimate in exactly this fashion.19 Piṅgala goes on:

[If the Buddhas say that] everything is unreal, [they mean that], before all dharmas dissolve into the real mark, each distinct view [of them] is empty of any real core, as all there is, only exists due to the manifold conditions which come together. [If they say that] everything is both real and unreal, [they observe] that the multitude of sentient beings can be divided into three types: those with the higher, middle, and lower [faculties]. Those with the higher [faculties] see that the marks of all dharmas are neither real nor unreal. Those with the middle [faculties] see that the marks of all dharmas are both real and unreal. Because of the shallowness of their wisdom, those with the lowest [faculties] see that the marks of all dharmas contain a small proportion of what is real and also a small proportion of what is unreal. When they observe that nirvāṇa is an unconditioned dharma and indestructible, they perceive what is real; when they observe that saṃsāra is a conditioned dharma and deceptive, they perceive what is unreal. [When the Buddhas say that] everything is neither real nor unreal, they do so because they deconstruct [such dual views about] what is real and what is unreal. Question: At another occasion the Buddha has said that one should distance oneself from [the view of] neither being nor non-being, why is it now said that what the Buddha has taught is neither being nor non-being [neither real nor unreal]? Answer: It was a different context when [the Buddha taught] the deconstruction of all the four types of clinging, [the apophatic catuṣkoṭi], but now in [Nāgārjuna’s exposition of] the four alternatives, [the cataphatic catuṣkoṭi], there is no entanglement in deceptive speech and thought. Once he heard the Buddha’s teaching, he attained the path of awakening, therefore he said “neither real nor unreal”. 一切不實者, 諸法未入實相時, 各各分別觀皆無有實, 但眾緣合故有. 一切實不實者, 眾生有三品, 有上, 中, 下. 上者觀諸法相非實非不實. 中者觀諸法相一切實一切不實. 下者智力淺故, 觀諸法相少實少不實; 觀涅槃無為法不壞故實; 觀生死有為法虛偽故不實. 非實非不實者, 為破實不實故, 說非實非不實. 問曰: 佛於餘處, 說離非有非無; 此中何以言非有非無是佛所說? 答曰: 餘處為破四種貪著故說, 而此中於四句無戲論. 聞佛說則得道, 是故言非實非不實.

(Zhong lun T30, no. 1564, p. 25, a20–b2).

The passage as a whole illustrates that clinging to either of the four, while dismissing any of the other three, is what would lead practitioners of Buddhist doctrine astray. Therefore, Piṅgala explains that the Buddha, at another occasion, applied the apophatic catuṣkoṭi, denying also the fourth position, and yet what the Buddhameant is the same that Nāgārjuna presents in his cataphatic exposition. If the third koṭi were misunderstood so that the other positions were not taken in account, it would misrepresent the two truths as duality, contravening the meaning of true emptiness. However, integrated into the entire catuṣkoṭi, the third might correctly point at the necessity to differentiate between two truths, that is, between saṃsāra and nirvāṇa.

Furthermore, Piṅgala clarifies that, in order to undermine any mistaken sense of duality, the fourth only fulfills the purpose to deconstruct the third. Although opposed to the third, the fourth is not meant to completely separate from the preceding three. If this view arises in the practitioners’ minds, they are advised to distance themselves from the fourth. What helps in this case is the “the deconstruction of all four types of clinging”—Nāgārjuna’s apophatic application of the catuṣkoṭi.

Overall, Piṅgala’s commentary shows that the apophatic and cataphatic versions of the catuṣkoṭi are not in contradiction with one another; the two represent the same. Yet, their relationship is paradoxical: although the catuṣkoṭi may heuristically function as a means of doctrinal analysis, it simultaneously bears the risk of becoming the source of entanglement in deceptive speech and thought. It undeniably is a medicine, but one whose toxic potential must not be ignored; to reveal its ineradicably harmful potential is to bring about its therapeutic effect.

In the Chinese Boredeng lun shi (*Prajñāpradīpa) translated by Prabhākaramitra (565–633), Bhāviveka appears to choose the opposite approach: the first position in the same catuṣkoṭi seems to represent a view reminding us of Sarvāstivāda: the eighteen dhātus (shiba jie 十八界) are the irreducible elements which constitute all compound phenomena that arise in the realm of saṃsāra.20 In contrast to those unreal phenomena, these elements are considered as real. However, for the Mādhyamikas, such a concept of saṃsāra is incomplete, as it lacks a thorough understanding of emptiness. Thus, it does not reach beyond the realm of conventional truth; “everything is real” (yiqie shi 一切實) in the first koṭi implies confinement to the domain of the conventional truth, in contrast to what the Zhong lun represents as Piṅgala’s view of the ultimate truth.

Developing a deeper understanding, the second koṭi points at emptiness, revealing unreality (feishi 非實) in all conditioned arising. This also applies to the eighteen elements. Emptiness is ultimately true in the sense that all arising is unreal. Unreality is seen from the higher vantage point of ultimate truth. Although Piṅgala and Bhāviveka interpret the first and the second koṭi in a reverse manner, they agree that the two positions represent the two truths as opposites. Bhāviveka’s interpretation seems to introduce a view of progression: first, the lower conventional and, second, the higher ultimate. The third koṭi reveals that the two truths are correlatively dependent. This is referred to as “both real and unreal”. The last step is the fourth koṭi, adumbrating full awakening as a state of mind devoid of clinging and conceptual distinction; this is called “neither real nor unreal”.21

Overall, Bhāviveka’s interpretation of the catuṣkoṭi seems to disclose to practitioners the ascending and graded levels of the Buddha’s teaching. He proceeds with this project by gradually deconstructing the stratified forms of clinging in their conceptual attitudes until full awareness of all fabrications in thought and speech is accomplished. Perhaps, it is the distinction between Mahāyāna and Hīnayāna which has inspired his view: the first koṭi may represent Sarvāstivāda as the lower understanding of Hīnayāna, while the succeeding levels describe the ascension to the higher stages of Mahāyāna until awakening is completed, which culminates in transcending all (deceptive) distinctions. Candrakīrti seems also to stress a similar reading of doctrinal gradation, yet without mentioning the two truths (Ruegg 1977, p. 23; Westerhoff 2005, p. 34).

The Sanskrit version of Bhāviveka’s Prajñāpradīpa is not extant; only the Chinese and Tibetan translations have been transmitted. Jizang and Zhiyi were not aware of Bhāviveka’s thought, and later Tiantai tradition did not pay attention to the translations from the Tang dynasty. By contrast, Piṅgala’s commentary exerted significant influence on both Zhiyi and Jizang, although the two Chinese interpreters critically remarked that his annotations are one-sided. Interestingly enough, Bhāviveka’s outline of the gradated structure of Buddhist doctrine yet demonstrates a remarkable degree of conceptual affinity with the Chinese patterns of doxography. Besides Bhāviveka’s taxonomic approach to the catuṣkoṭi, his discussion also offers another reading from a soteriological point view, similar to Piṅgala’s commentary, which shows that, for him, the concept of catuṣkoṭi must be treated with a sense of ambiguity (Boredeng lun shi T30, no. 1566, p. 108, a4–b2).

4. Chinese Sanlun and Tiantai Interpretations of Nāgārjuna’s Catuṣkoṭi

In the Chinese transmission of Indian sūtra and śāstra literature, this third instance of the catuṣkoṭi appears not only in Kumārajīva’s translation of Nāgārjuna’s Kārikā, the Zhong lun, and the Boredeng lun shi, but also in the the Da zhi du lun (not existent in Sanskrit or Tibetan), which expands on it in tandem with the two truths and another associated concept, called the “four siddhāntas” (si xitan 四悉亶) (T25, no. 1509, p. 61, b6–18).

The Sanskrit term “siddhānta” means established doctrine, and the Chinese created a transliteration for this. The four siddhāntas are an extension of the two truths. In terms of doctrinal debate, “four siddhāntas” only played a significant role in Tiantai tradition. The last of the four accounts for the supreme or ultimate meaning (diyi yi 第一義) which instantiates “shixiang 實相”, rendered as real mark, that is, the nature of reality. One of the Sanskrit terms for which ancient translators construed those Chinese expressions was “dharmatā”—literally: dharma nature (nature of all things). Chinese Buddhist texts used for dharma nature also the literal translation “faxing 法性”.

In Tiantai and Sanlun, “shixiang”, “faxing”, and “diyi yi” were often treated as synonyms. According to the Chinese transmission and interpretation of Prajñāpāramitā and Madhyamaka, the nature of reality equals the mind’s liberation from its self-induced deceptions, as the undistorted mind sees all things that it encounters in the way they actually are. This also includes its own nature which many Tiantai masters interpreted as the nature of reality (real mark, dharma nature, ultimate meaning)—if what is seen by the unobstructed mind is identical to what things truly and really are, “true and real” (zhenshi 真實) can be predicated on both what is seen and what is seeing, that is, such differentiation must dissolve in the real mark. This is ultimate meaning.

Like other Prajñāpāramitā and Madhyamaka texts, the Da zhi du lun stresses that ultimate meaning or real mark (shixiang) evades linguistic representation. However, mentioning the two truths in association with the catuṣkoṭi, the text seems not to imply that ultimate meaning or real mark must completely separate from speech and thought. Instead, the proper way of approaching it is to neither separate from all four, nor to cling to any apart from the others. To consider all four together is to enter the state of mind which realizes that all the objects it clings to or separates from are its own fabrications—the liberated mind sees that separating entails clinging! In this paradoxical sense, ultimate meaning or real mark instantiated by the liberation of the mind from deception does not separate from speech and thought, although there must be no clinging to a fixed form of linguistic expression; (Da zhi du lun T 25, no. 1509, p. 61, b6–18).

Jizang’s Sanlun works and Zhiyi’s Tiantai texts quote this catuṣkoṭi and comment on it several times, using the version from the Da zhi du lun, which they regard as identical to that of the Zhong lun (Kārikā). In his commentary on the Kārikā, the aforementioned Zhongguan lun shu, Jizang observes ontological indeterminacy when examining the question of what the nature of reality (shixiang—real mark) means. In his discussion, he employs a binary which frequently occurs in Chinese exegesis of sūtra and śāstra. This binary, most probably resulting from a fusion of indigenous thought and Buddhist concepts, started to appear in the sixth century and, in the present context, might best be translated as “cohesive body and functions” (tiyong 體用).22 According to Jizang, the Chinese term “real mark” (“shixiang” probably used for the Sanskrit “dharmatā”) must be understood through these two aspects, that is, the ultimate meaning of Buddhadharma, the sense of liberated mind, becomes accessible to practitioners by perceiving the cohesive body in the functioning of all the four alternatives. He explains:

These four gates [catuṣkoṭi] are the expedient means [kauśala-upāya] bound for the real mark [dharmatā]. To let one’s mind, [in an unimpeded manner], wander through all four gates is to enter the real mark [dharmatā]. Hence, the four gates display the functions [of the real mark], and what is not divided into four [within all four] is the cohesive body [of the real mark]. 此之四門, 皆是實相方便. 遊心四門便入實相. 故以四門為用, 不四為體;

(Zhongguan lun shu T42, no. 1824, p. 124, a12–23).

Awareness of ontological indeterminacy involves the previously mentioned paradox of distinguishing two truths (non-duality qua/yet differentiation); “cohesive body and functions” is analogous to “ultimate truth and conventional truth”. In a preceding passage, Jizang explains that the catuṣkoṭi (four gates, simen) accounts for the conceivable, explicable, and visible “functions of the real mark” (shixiang yong 實相用) (Zhongguan lun shu T42, no. 1824, p. 124, a14–20). The Chinese term “busi 不四,” literally “not-four” and here translated as “what is not divided into four [within all four],” describes the indivisible which is the “cohesive body of the real mark” (shixiang ti 實相體). Neither of the four can be understood as an independent ontological predication of what reality ultimately is, yet each of the four must be seen in its specific heuristic value and significance.

The cohesive body of the real mark, understood as an indivisible whole, is invisible or hidden in the sense that meaning conveyed by awakened speech evades any fixed or determinate form of predication. If the catuṣkoṭi is to be seen as a linguistic symbol for an exhaustive set of alternative predications of reality, the indivisible and invisible body of the real mark—cohesion of the ultimate meaning of awakened speech—is not confined to just one apart from the other three predicative forms, nor is it what goes beyond or separates from all four. Cohesion is what sustains the four distinct alternatives of predication, each of which is the visible form of the body’s functioning.

This is analogous to the emptiness that constitutes all differing phenomena which are incessantly changing instances of conditioned arising, each of which points back at the very nature of all, emptiness of an intrinsic nature; or it is like ultimate truth that constitutes all conventional truths; each is capable of guiding the practitioner back to the very source it descends from. The two aspects of the catuṣkoṭi, that is, the hidden cohesion of ultimate meaning and its differing functions of predication, imply a sense of circularity which, as will be shown below, accounts for the hermeneutical circle that the mind must enter to accomplish liberation from its own deceptions.

Again, the hidden cohesion of the ultimate meaning can be apprehended in each of the four distinct forms (the visible)—like emptiness in conditioned arising. This can be illustrated on the basis of the previous explanation that the four alternatives in the catuṣkoṭi are interdependent or correlative opposites: the first asserts what the second denies and vice versa, which brings about the third, embracing the entire opposition of the two; the opposite of this is the fourth excluding the first and second; like the first and the second, the third and the fourth are interdependent opposites. As each of the four entails all four, we could say that the invisible and indivisible which constitutes all—the hidden cohesion of all four—is equally embodied by each of the distinct four.

The interlinked relationships make up the “cohesive body of the real mark” (shixiang ti) which is not really divisible into four (busi), although its functions can and must be classified into four distinct forms. The “real mark” has both a hidden (indistinct) and a visible (distinct) side, inseparably bound up with one another. Its cohesive body cannot be represented by any of its distinct functions alone, and yet the cohesion of “not-four” (busi), the indivisibility of the real mark, or the ultimate meaning is wherein all four functions emerge as mutually distinct and visible forms.

The practitioners’ awareness of the differentiation and non-duality of the distinct and indistinct sides of the real mark—the “four functions” and the indivisible “not-four”—features the understanding of the ultimate meaning, which can be described in terms of the hermeneutical circle. Distinguishing the two (the distinct and the indivisible) entails realizing non-duality; the same applies in reverse: seeing their non-duality requires differentiating the two. According to Jizang’s exposition, entering the hermeneutical circle in such examination of the catuṣkoṭi triggers the understanding of ultimate meaning, which requires awareness of the paradox that the mind’s intent of separating from its own fabrications entails its clinging to them. This is so, because what is unreal cannot be appropriated, nor can it be dismissed, and yet its existential relevance cannot be denied either. Such awareness culminates in seeing that the ultimate meaning instantiating the nature of reality (real mark or liberation from deception) evades ontological determination.

In his treatise on the Lotus Sutra (Fahua xuanlun 法華玄論), Jizang explains ontological indeterminacy through the term “indeterminate mark” (wuding xiang 無定相), which is just another term for real mark (shixiang 實相) and, in this discussion, is directly linked to the catuṣkoṭi. According to that treatise, the indeterminate mark or real mark is dynamic, non-excluding, and infinite (wuliang 無量), alternately taking shape in any of the four alternatives. Jizang explains:

Hence, one should know that the [Buddha-]dharma has no determinate mark; it is just due to [the act of] awakened understanding that there is such following [of any of the four]. … The one fully aware of this realizes in respect of all four cases that they are [the ultimate meaning of] the Buddhadharma, whereas the one unaware of this, in all four cases, falls prey to the dharma of deception. … The one, who accomplishes wisdom in terms of prajñā and upāya, while studying this catuṣkoṭi, is not likely to fall prey to any of the four [false] views [by clinging to them]. 故知: 法無定相, 唯悟是從. … 了者, 於四句皆是佛法. 不了者, 四句皆是魔法. …得般若, 方便, 學此四句, 不墮四見.

(Fahua xuanlun T34, no. 1720, p. 381, a22–b14).

Jizang clearly stresses the ambiguous potential of the catuṣkoṭi, as adopting such a method of classifying conceptual forms either manifests the ultimate meaning, or may generate confusion. Most importantly, there is no fifth or even sixth koṭi (edge) that goes beyond all four or is within all four, and yet the way in which all four predicate the indeterminate mark—the real mark or ultimate meaning—is to not assert any of the four, nor to deny any of them—not asserting amounts to not denying and vice versa.

Again, to observe this paradox is to see that the ultimate meaning instantiating the nature of reality (real mark) evades ontological determination. Even the view that contradiction is the definite form that truly predicates ultimate reality contradicts Jizang’s sense of ontological indeterminacy. Such interpretation entails entanglement in deceptive distinction, clinging to one of the four—the third koṭi. The paradox which Jizang reveals is not a contradiction expressing ontological significance. Metaphysics of dialetheism should not be superimposed on Jizang’s practice of observing paradoxes in his soteriological approach to exegesis.

Zhiyi’s Tiantai view of Nāgārjuna’s positive version of the catuṣkoṭi is very similar to the Sanlun interpretation. The Tiantai master also emphasizes that the catuṣkoṭi represents the last of the four siddhāntas, and, similarly, the discussion must take two aspects into account. In his treatise on the Lotus Sutra (Miaofa lianhua jing xuanyi 妙法蓮華經玄義) and his commentary on the Vimalakīrti-nirdeśa-sūtra (Weimo jing xuanshu 維摩經玄疏), Zhiyi quotes the catuṣkoṭi from the Da zhi du lun and expounds the ultimate meaning (diyi yi) or dharmatā in reference to (1) the explainable (keshuo 可說) and (2) the unexplainable or ineffable (bukeshuo 不可說; T33, no. 1716, p. 687, a12–22; T38, no. 1777, p. 520, c15–28).

For him, the four alternatives in the catuṣkoṭi represent the explainable side (keshuo) of the ultimate meaning (diyi yi), which comes close to Jizang’s “functions of the real mark” (yong 用), while the ineffable or unexplainable resembles Jizang’s “cohesive body of the real mark” (ti 體). Zhiyi further states that those who do not see non-duality in such distinction between the explainable and unexplainable fall prey to their clinging to either one of the four, generating nothing but “prapañca” (xilun)—entanglement in deceptive speech, while the awakened who integrates the four sees one in four and four in one, without taking either one, or four, or both, or neither as the ultimate meaning. The awakened practitioner understands the “real mark of all dharmas” (zhufa shixiang 諸法實相), by observing non-duality while differentiating the explainable and ineffable sides.

Where Jizang’s and Zhiyi’s interpretations converge is the point that all the alternatives in the catuṣkoṭi are interdependent links apart from which the ultimate meaning cannot be realized—there is no fifth, sixth, or any further point; also, any of the four separated from the other three would contradict that meaning. According to both Zhiyi and Jizang, the catuṣkoṭi is the conceptual blueprint for doctrinal classification. However, unlike Jizang, Zhiyi seems to relate this specific catuṣkoṭi only to the second level of his fourfold classification of doctrines (huafa sijia 化法四教). In order to understand the difference between Jizang’s and Zhiyi’s interpretation, the aspect of doxography (doctrinal classification) must be discussed.

5. The Catuṣkoṭi as a Conceptual Blueprint for Tiantai’s Doxography

In the Mohe zhiguan, Zhiyi’s Tiantai teaching explicitly hints at the ambiguous potential of the catuṣkoṭi, distinguishing opposite methods of its application: (1) the catuṣkoṭi as a proper linguistic means which can convey the ultimate meaning realized by the awakened or liberated mind, or (2) the same linguistic forms in deceptive speech arising from the non-awakened state of mind (T46, no. 1911, p. 68, a29–b57).

According to Tiantai, the first type embraces the perceived form of proper doctrinal discourse in sūtra and śāstra, and, as a consequence of this perception, Zhiyi and his eminent disciple, Guanding 灌頂 (561–632), constantly resort to the catuṣkoṭi as an analytical technique of examining all the exegetical issues which they discuss in their own works. In this regard, the catuṣkoṭi is dealt with in a twofold way: (1) it is an instrument of conceptual analysis by means of which Tiantai construes and interprets doctrinal contents of Indian Buddhist texts; (2) it is itself an object or topic of examination.

In the Mohe zhiguan, Zhiyi first analyzes the catuṣkoṭi as it figures in deceptive speech and then expounds how it functions in proper doctrinal discourse; the two steps are connected with one another: he first uncovers the paradox that the intent of separating from the four alternatives entails the reverse—the mind’s clinging to them (see Section 6). He further holds that revealing this paradox is the prerequisite for disclosing to the deluded practitioners the salutary side of the catuṣkoṭi, namely, its application as a means of conceptual analysis in doctrinal discourse and exegetical discussion. This culminates in establishing Tiantai doxography modeled after the four alternatives. Its taxonomy consists of four levels, classifying and assorting the perceived variety of doctrinal content in sūtra and śāstra:

(1) The first level, called “Tripiṭaka teaching” (sanzang jiao 三藏教), holds the position of being (you 有) equivalent to the first koṭi “everything is real” (yiqie shi), and represents Hīnayāna as it appears to Tiantai in the shape of Sarvāstivāda in the Āgama-sūtras and Abhidharma-śāstras. Although this view comes very close to Bhāviveka’s interpretation of the first koṭi, Tiantai stresses that such a form of Hīnayāna does not develop any point of intersection with Mahāyāna.

(2) The second level, called “common teaching” (tongjiao 通教), represents the reverse position of non-being (wu 無), equivalent to the second koṭi “everything is unreal” (yiqie feishi), and reveals what Hīnayāna and Mahāyāna share in common; this encompasses some of the doctrinal contents expounded by the Prajñāpāramitā- and Vaipulya-sūtras and also indicates similarities with Bhāviveka’s view of the second koṭi.

(3) The third level, called “distinct teaching” (biejiao 別教), is the position of both being and non-being (feiyou feiwu 亦有亦無), equivalent to the third koṭi of “both real and unreal” (yi shi yi feishi). “Distinct teaching” designates the sublime sense of Mahāyāna, excluding Hīnayāna; Tiantai perceives this in the Yogācāra texts and even more obviously in the Tathāgatagarbha texts, as the latter tend to stress the wisdom of both emptiness (śūnya) and non-emptiness (aśūnya, bukong 不空).

(4) The fourth level, called “round/perfect teaching” (biejiao 圓教), takes the stance of neither being nor non-being (feiyou feiwu 非有非無), equivalent to the fourth koṭi of “neither real nor unreal” (yi shi yi feishi). The round/perfect teaching embodies circularity, as it is both the source and the goal of the preceding levels. With varying degrees of explicitness, it appears in all Mahāyāna sūtras, but it is said to be most distinctively manifested by the “one [Buddha-]vehicle” (ekayāna, yisheng 一乘) in the Lotus Sutra.

Moreover, “round/perfect” (yuan 圓) also indicates the sense of suspending differences, leveling out opposite forms of one-sidedness, and undermining the duality between, or exclusivity of, views, conceptual positions, and mental attitudes; therefore, it also epitomizes the meaning of the middle way (madhyamaka, zhongdao 中道), often defined as “neither separating nor adhering” (buli buji 不離不即). This is to say, the round/perfect level of “neither being nor non-being”—the middle way—does not exclude any of the other three positions; it does not go beyond the preceding three. The round/perfect encloses all of them and is capable of manifesting the sense of “neither being nor non-being” through these differing forms of “being”, “non-being”, and “both being and non-being”. In this fashion, Tiantai tries to show that it neither separates from nor clings to anything, realizing the sense of liberation. Moreover, this view of the catuṣkoṭi implies a reinterpretation of Madhyamaka’s two truths.

Apocryphal sūtras, composed in China during the fifth century, started to talk about a third truth, that is, the “highest meaning of truth in/as the middle path” (zhongdao diyi di 中道第一義諦). Tiantai and Sanlun adopted this notion and construed the concept of the “threefold truth” (sandi 三諦). According to Tiantai’s round/perfect teaching, this means that separation of the middle from the conventional and the ultimate undermines the very meaning of those two. At the fourth level of doctrine, conventional truth and ultimate truth, respectively, instantiate the highest meaning of truth in the middle path, and each equally displays the ultimate meaning of the Buddhadharma. Each of the three truths embodies all three. To understand this is to simultaneously see one in three and three in one, yet it does not imply the reality of one and that of three.23 Such an ambiguous and paradoxical structure of the threefold truth is called “inconceivable” (busiyi).

Again, the highest level of doctrine presenting the inconceivable sense of threefold truth is the round/perfect teaching (yuanjiao). For heuristic reasons, it engenders its own reverse—the lower three levels which can lead the non-awakened practitioners back to the highest. However, in contrast to the inconceivable round/perfect teaching, the lower three of all four levels in Tiantai doxography conceptualize the ultimate meaning of Buddhist doctrine, generating instructive yet incomplete notions of it. In this way, those lower teachings no longer maintain mutual integration of the three aspects that constitute truth in such cohesion. Only the manner in which the round/perfect teaching exhibits the middle path as the all-embracing Buddha nature (nature of awakening, foxing佛性) in both conventional truth and ultimate truth features the complete sense of the threefold truth—the Tiantai sense of inconceivable.

Although Tiantai master Zhiyi construes his taxonomy of Buddhist doctrine on the methodological basis of the catuṣkoṭi, at this point, he reverses the course of discussion: the catuṣkoṭi turns into an object of examination and therefore must be reinterpreted from the viewpoint of all four doctrinal levels (Mohe zhiguan T46, no. 1911, p. 73, b25–p. 75, b27). The catuṣkoṭi realized in its inconceivable sense due to the round/perfect teaching does not equal those meanings which itembodies when it is seen from the vantage points of each of the three conceptualizing levels. The same applies to all doctrines transmitted in the Indian Buddhist texts—according to Tiantai. All notions mentioned in sūtra and śāstra—such as the two truths, conditioned arising etc—must be examined from the vantage point of this fourfold classification, called the “four teachings of the transforming dharma” (huafa sijiao).

Hence, each of the four levels of teaching looks at the same linguistic form of the catuṣkoṭi in a different way. Depending on the doctrinal level of discussion, each of the four (“being”, “non-being”, “both being and non-being ”, “neither being nor non-being”) changes its meaning four times. Despite such content-related amplification of each of the four expressions, the inner relations between all 16 modifications together are evident. Hence, assorted in this pattern of interdependence and mutual constitution, all of them are unified as a systematic whole. In this fashion, they are believed to systematically evince the hidden and inner coherence of the doctrinal diversity in sūtra and śāstra. Zhiyi quotes extensively from all Indian Buddhist texts, to find canonical evidence for each of the 16 views which he considers as that whichsystematically summarizes, epitomizes, and encompasses the entire content of doctrinal discourse in the transmitted canon of sūtra and śāstra. Aside from that hermeneutical purpose, the intended soteriological effect of this classification and expansion is that the ascension of a certain doctrinal topic from the lower conceptualizing levels to the highest inconceivable, that is, the integration of its conceptualization into the inconceivable sense, becomes intelligible to practitioners.

Zhiyi’s treatment of the catuṣkoṭi pursues the hermeneutical goal of highlighting the inner unity of the numerous and various doctrines in the translated texts from India. This coincides with his soteriological intent of showing that the various textual representations of the Buddhadharma are congruent with the ultimate meaning of liberation. In this way, he demonstrates that the hidden sense of coherence—for him the highest meaning of awakening—ultimately is independent from linguistic signification, and yet does not really separate from speech and concept. For him, the textual corpus of the Buddhist canon embodies precisely this. Again, Tiantai’s doxographical appropriation of the catuṣkoṭi mirrors how the hermeneutical project concurs with the soteriological intent: to deconstruct the perceived duality of doctrinal conceptualization and inconceivable liberation is to integrate the ineffable—the ultimate meaning—with discourse in sūtra and śāstra. Nāgārjuna’s catuṣkoṭi is a major factor in this project.

6. The Catuṣkoṭi as a Symbol for Deception

However, how does the same catuṣkoṭi figure as a deceptive form of speech characterizing the deluded or non-awakened mind? To make the deceptive effects of the catuṣkoṭi transparent, Zhiyi develops the Tiantai practice of contemplating the mind (guanxin 觀心)—a practice of introspection that detects all the blindspots which the non-awakened mind itself creates and falls prey to. The contemplating mind comes to see that certain types of distinctions are the source of mistaking the referents of our speech for independently existing entities. Tiantai commences with contemplating emptiness (kongguan 空觀) which first deconstructs the assumption of an intrinsic nature immanent to things and then unveils the paradox that the intent of separating from the distinct forms of the catuṣkoṭi entails the reverse—the mind’s clinging to deceptive distinctions.

This contemplation is designed for the common or non-awakened persons (fanfu 凡夫) who not only fail to liberate their minds from deceptive construction, but also increase the same. Taking each of the four alternatives as a logically independent or exclusive predication, they end up construing even a catuṣkoṭi of the catuṣkoṭi. Zhiyi classifies such unrestrained proliferation of non-awakened speech (prapañca, xilu) into [1] the simple catuṣkoṭi (dan siju 單四句) which construes ontological views about being, non-being, both, neither, [2] the double catuṣkoṭi (fu siju 複四句), [3] the multifarious catuṣkoṭi (juzu siju 具足四句), and [4] the dissolution of speech (jueyan 絕言), also called absence of speech (wuyan 無言) (Mohe zhiguan T46, no. 1911, p. 62, b8–10). This fourth position represents the mind’s intent of realizing complete silence, separatefrom the preceding three types of the catuṣkoṭi. Compared to the simple catuṣkoṭi, the two types of the double and multifarious catuṣkoṭi increase the complexity of each of the four alternatives.

The double catuṣkoṭi (fu siju) might emerge, when the four alternatives of the simple catuṣkoṭi are rejected without realizing the sense of ontological indeterminacy. More complex forms of metaphysical confusion arise from this, embracing the following four alternatives: (1) being of being, being of non-being (有有, 有無); (2) non-being of being, non-being of non-being (無有, 無無); (3) both being of being/non-being and non-being of being/non-being (亦有有無, 亦無有無); (4) neither being of being/non-being, nor non-being of being/non-being (非有有無, 非無有無, T46, no. 1911, p. 62, c12–13). Seeing that the first and second koṭi in the simple catuṣkoṭi are correlative opposites mutually constituting each other, the deluded mind construes the double catuṣkoṭi and holds that “being” must represent being of both being and non-being, while the opposite “non-being” means non-being of both being and non-being; the same applies to the third (including the preceding two) and to the fourth (excluding them).

Zhiyi further explains that, if one intends to escape from the entanglements in the double catuṣkoṭi, but still continues to lack an awareness of ontological indeterminacy, one might fall prey to the multifarious catuṣkoṭi, which entails one’s involvement with even more complex forms of confusion. When holding the first position “being” in the multifarious catuṣkoṭi, the deluded mind observes that all four alternatives are mutually constitutive to one another, therefore “being” includes four aspects: being of being (you you 有有), being of non-being (you wu 有無), being of both (you yiyou yiwu 有亦有亦無), and being of neither (you feiyou feiwu 有非有非無); the same applies to the opposite “non-being”, to the third, and to the fourth; each of the four includes four (Mohe zhiguan T46, no. 1911, p. 62, c16–23).

Most importantly, these 16 deluded views are not equivalent to the 16-fold sense of the four levels of teaching in Zhiyi’s doxography of the Buddhadharma. Although examining the catuṣkoṭi from different perspectives, each of these four taxonomic levels helps achieve awareness of ontological indeterminacy which eventually dissolves all metaphysical dogmatism, whereas each of the four positions of the multifarious catuṣkoṭi, to the contrary, strengthens the grip on one specific ontological view excluding the others, which exacerbates metaphysical confusion.

Even if one’s mind rejects all three types of the simple, double, and multifarious catuṣkoṭi, adhering only to the fourth position of complete dissolution of speech, the mind locks itself into a deceptive form of silence. Silence as the result of the mind’s intention of eliminating speech has not really resolved the mind’s clinging to deceptive dualities; such silence does not differ from speech. The mind is still captured in the circular entanglement of deceptive distinctions which it has construed due to its misuse of the conceptual forms of the catuṣkoṭi. Such dualistic silence is not identical to the silence described in the Vimalakīrti-nirdeśa-sūtra, which is non-dual in the sense of neither excluding speech nor silence. Therefore, the exclusive sense of silence is classified as the fourth position in the deceptive version of the catuṣkoṭi. In the Mohe zhiguan, Zhiyi explains the whole structure:

[When contemplating the mind, one might realize that] the four alternatives are false/provisional [forms of designation], they are deceptive and unreal. The true principle [of non-arising] goes beyond word and speech, dissolving all four alternatives, which indeed is non-arising. Although one might then say [such insight] is what makes one overcome the four alternatives, in fact one has not overcome them at all. Briefly, there are three types of [intentions] of going beyond the four alternatives: first, [going beyond] the simple; second, [going beyond] the double; third, [going beyond] the multifarious [catuṣkoṭi]. If one says that by virtue of true principle one goes beyond speech, then this is only overcoming the simple catuṣkoṭi, but one has not overcome the second alternative of the double catuṣkoṭi [equivalent to non-being of both being and non-being], nor has one overcome the first of the multifarious catuṣkoṭi [equivalent to being of neither being nor non-being]. One should know: the net of deceptive views is closely knotted, it is difficult to escape from it. … One might say again: Stepping beyond the simple, double, and multifarious catuṣkoṭi and overcoming them, this is wherein speech and debate come to an end, circulating thoughts extinguish, and obliteration and purification persist, as it is the path of non-arising and dissolution of speech. The one who calculates in this way yet lapses into the [deluded] view about the ineffable and dissolution of speech. What has this to do with the proper path? This one only says ‘dissolution of speech’, but when one’s speech ends, one’s speech is not dissolved. Why is it so? One expounds dissolution depending on non-dissolution with the result that dissolution yet implies dependency. As long as the sense of interdependent opposites arises, one should not speak of dissolution. It is as if one tries to escape from empty space, but how is it possible to evade true principle [which embraces all]? 四句皆假, 虛妄不實. 理在言外, 絕於四句, 乃是無生. 謂出四句, 實不出也. 略有三種四句外：一單, 二複, 三具足. 若謂理在言外者, 乃是出單四句外, 不出複見第二句, 亦不出具足見初句. 故知: 見網蒙密, 難可得出. … 又復言: 出單, 複, 具足四句之外, 言語道斷, 心行處滅, 泯然清淨, 即是無生絕言之道. 如此計者, 還是不可說絕言之見. 何關正道？徒謂絕言, 言終不絕. 何以故? 待不絕而論絕, 絕還是待. 待對得起, 不應言絕. 如避虛空, 豈有免理!

(Mohe zhiguan T46, no. 1911, p. 66, c19–p. 67, a5).24

Overall, Zhiyi’s purpose of construing the fourfold classification of the deceptive catuṣkoṭi is to highlight all the blindspots which impede the deluded mind. Therefore, he uncovers the paradox that the intent of excluding the four distinct forms intensifies the mind’s involvement with deceptive distinction, producing new types of the catuṣkoṭi even more complex than those it tries to eliminate.

As long as the mind does not realize that its generation of delusive views functions like the circular operation of a closed system, it continues to confuse its own fabrications with real entities, erroneously holding that the objects of its epistemic–propositional references pertain to a world of independent things. The mind’s circular entanglement in its own metaphysical deceptions is like a self-contained system. Hence, enacting such features of the deluded mind, the scheme itself must be self-referential. It subsumes various types of the catuṣkoṭi in non-awakened speech and silence under the same pattern of the catuṣkoṭi, assorting the gradated levels of the mind’s entanglement in metaphysical constructions which are all deceptive. The fourfold catuṣkoṭi is a symbol for the mind’s unawareness of its circular entanglement in self-induced deception.

However, Tiantai also believes that it is the mind’s recurrent experience of falling prey to its own deceptions which makes it realize that there is nothing it can really hold on, neither within nor apart from its circulating in metaphysical dogmatism. Then, it really comes to see that the objects of its epistemic–propositional references are not independently existing entities, which amounts to seeing their real nature, yet being fully aware of ontological indeterminacy. Again, this is neither separating from anything, nor clinging to anything, nor asserting anything, nor denying anything, as what is unreal cannot be appropriated, nor dispensed with.

Yet, to see what unreality actually and really is means to apprehend its existential relevance—veiled and concealed in respect to what it is in actuality, unreality is deceptive and leads one astray, which brings suffering. However, to reveal and expose its real nature is, in effect, to explore the path to liberation from deception. Aware of what it truly is, the deceptiveness of unreality is instructive. Unreality retains its heuristic significance due to its deceptiveness. Therefore, the same catuṣkoṭi also functions as a symbol of Buddhist doctrine, opening up the path to liberation and awakening.

7. Concluding Remarks

The article aims to show that, in the Chinese appropriation of the Indian catuṣkoṭi, the ambiguous potential of this method of classifying conceptual–linguistic forms was a major factor which shaped the development of Buddhist debate in East Asia. The catuṣkoṭi was considered as that which might either inspire awareness of ontological indeterminacy or entail the reverse effect of generating confusion over metaphysical dogma. In China, the catuṣkoṭi served the purpose of classifying both doctrinal discourse in the transmission of the canonical word and non-Buddhist metaphysics considered to be deceptive.

Hence, Chinese masters hold that, in order to avoid mistaking the catuṣkoṭi in a way which entails further clinging to deceptive distinction, its ambiguity must be perceived when studying Indian sūtras and śāstras. Therefore, in the Mohe zhiguan, Zhiyi unfolds the inner connection of those opposite aspects, assorting ten inter-connected types of catuṣkoṭi (shizhong siju 十種四句). Jizang partially uses the same expressions that Zhiyi mentions; however, his list of the ten types of catuṣkoṭi (shizhong simen 十種四門) seems to follow a dialectical pattern, culminating in the sublation of the four alternatives, called “suspending (sublating) four” (juesi 絕四).25 Nonetheless, according to Jizang, “suspending four” is preciselywhat validates the use of the four alternatives in the sense of the Dharma. Despite the differences in Zhiyi’s and Jizang’s approaches, the sense of ambiguity remains the same.

Moreover, Jizang’s view also seems to come close to what Zhiyi, in the Mohe zhiguan, calls the “inseparability of deconstruction and construction” (jipo jili 即破即立)—the deconstruction of deception is the construction of doctrinal discourse or vice versa. Deconstruction of deception in the practice of contemplating emptiness engenders doctrinal discourse which in the shape of sūtra and śāstra provides the guideline that the practitioners must pursue. In such circularity, doctrine and contemplation are mutually complementary (jiaoguan xiangzi 教觀相資).

Whereas Zhiyi’s concern in the Mohe zhiguan consists of linking the structural pattern of the catuṣkoṭi to his method of mind-contemplation (guanxin) in tandem with his doxography (panjiao), Jizang applies the catuṣkoṭi as a strategem in linguistic pragmatics to make practitioners’ aware of certain paradoxes when examining the sense of non-duality (buer 不二), which debunks deceptive distinction via differentiation of the two truths. In his Dasheng xuanlun 大乘玄論, Shier men lun shu 十二門論疏, and the Erdi yi二諦義, Jizang presents his respective Sanlun teachings of three or four levels of two truths (sichong erdi 四重二諦) guiding that sense of non-duality. The Dasheng xuanlun further expands on the same scheme in terms of doxography (Chen 2013), but all three works discuss the catuṣkoṭi in combination with the two truths.

Overall, the Tiantai and the Sanlun versions can be subsumed under what I call hermeneutics of deconstruction, because the deconstructive goal of the two as well as their motif of taking the entire canonical transmission from India into account seems to be the same.

Furthermore, Tiantai and Huayan are similar in the specific respect that the two develop their doxographies on the basis of the catuṣkoṭi and, in turn, analyze the same catuṣkoṭi from the viewpoint of their doxographies. This entails a type of circularity which the two Chinese schools share in common: the catuṣkoṭi figures as both the tool and object of their exegetical examinations. They also discuss their doxography in the same circular manner: the taxonomy of doctrine appears to be both the method and the topic of their discussions. Such treatment of the catuṣkoṭi in the context of doxography slightly differs from Jizang’s Sanlun approach, although all three Chinese schools seem to agree that it must be a certain sense of circularity based on which soteriological theory in Buddhism can solve the question of how transformation from a non-awakened state of being into the opposite becomes possible.