The Reliance on Scripture and Vicissitudes of Textual Practices in Madhyamaka Thought

Shenghai Li Fudan University

What texts did Buddhists of South Asia and beyond read? How did they read, interpret, and use these texts? This essay focuses primarily on the first of the two questions and examines in this connection instances of citation found in the early Mūlamadhyamakakārikā commentaries and in a related Tibetan work as evidence of the uses of Buddhist texts. The collected samples indicate two major shifts in Indian and Tibetan Buddhist textual practices. The first transition occurred in the sixth and seventh centuries when Indian commentaries on the Mūlamadhyamakakārikā began to cite scriptural passages with greater frequency, especially from the Mahāyāna sūtras. The example of Tsong kha pa’s Madhyamaka work represents a later trend in which Tibetan writers repeated sūtra passages previously cited in the Indian texts that were the main objects of their study and attention. What emerges here is the pivotal role played by the middle-period Indian Mādhyamikas.

Writers such as Bhāviveka and Candrakīrti contributed very substantially to the collection of core scriptural citations that were deployed in the Madhyamaka texts. While some of the sūtra passages these Indian Madhyamaka authors used were circulated outside the circle of their own philosophical tradition, others appear to have been newly collected through their private reading experience. The familiar principle that characterizes the basic method of the Buddhist

scholastic enterprise—reliance on both scripture and reasoning—has been continuously employed in much of the history of Buddhist thought. Just as their Indian predecessors did in the first millennium, Tibetan writers of the second millennium also invoked this principle frequently and even included it in the titles of texts. 1 Thus, Go rams pa (1429–1489) started his criticism of Tsong kha pa’s view in the Lta ba’i shan ’byed by stating that he would offer a brief examination of his rival’s system “by using scripture and reasoning.” 2 Tsong kha pa (1357–1419), on his part, regarded his Lam rim chen mo as “having been drawn from the path of proper analysis using scripture and reasoning.” 3 While these two writers disagreed on doctrinal and philosophical points, the general guideline of using scripture and reasoning as a means of scholastic deliberation was held by both. Going back to the Madhyamaka tradition in India, in the sixth century Bhāviveka stated in Madhyamakahṛdaya that he had described reality that is “endowed with reasoning

An earlier version of this paper was presented at the 222nd annual meeting of the American Oriental Society, where it received stimulating responses from the audience. I would like to thank Donald R. Davis for his helpful comments. The revision of this paper benefitted from the suggestions provided by anonymous reviewers and the editorial advice from Stephanie Jamison. This research was assisted by a New Faculty Fellows award from the American Council of Learned Societies, funded by the Andrew W. Mellon Foundation. 1. Cabezón and Dargyay (2007: 13) mentions one such title: Rig ’dzin Chos kyi grags pa’s (1595–1657) A Response to a Refutation: A Necklace for Those Who Preach Scripture and Reasoning (Dgag lan lung rigs smra ba’i mgul rgyan). 2. Ibid., 114: da ni lugs gnyis pa la lung rigs kyis dpyad pa cung zad brjod par bya ste/. 3. LRChM 811.3–4: lung rigs kyis/ /tshul bzhin dpyod pa’i lam nas bdag gis drangs/. 544 Journal of the American Oriental Society 137.3 (2017)

and scripture,” which, “being examined by reasoning, remains unharmed.” 4 In the early seventh century, Candrakīrti asserted in the Prasannapadā, his commentary on the Mūlamadhyamakakārikā, that even Nāgārjuna relied on the same method: “Using reasoning and scripture, the ācārya [[[Nāgārjuna]]] composed this text for the purpose of removing doubts and misunderstanding.” 5 In a work that has been generally accepted to be authored by Nāgārjuna, 6 Ratnāvalī attests to the customary practice of demonstrating a point by recourse to both scriptural sources and reasoning: “This has been spoken by the Bhagavat, while the reason is also observed in this case.” 7 Indeed, in the Abhidharma texts of comparable antiquity, there was already evidence for the application of this established norm, 8 which both Mahāyāna and early Buddhist authors observed. What we have seen in these few instances are merely varied ways of expressing the basic underlying principle. In

this specific case Buddhist writers have rarely provided self-conscious and detailed accounts of the different manners in which they applied reasoning and— even less frequently—scripture. Moreover, we can hardly expect that the types of scripture and forms of reasoning employed and the understanding of what counts as reasoning could remain unchanged over such a vast span of time. To understand specific ways in which scripture and reasoning are employed in Buddhist texts and how these uses vary in the course of history, scholars of Buddhism must examine concrete cases of application of scripture and reasoning found in the literature. Between the two basic sources of Buddhist scholastic writing, reasoning has consistently received more scholarly attention, especially in the field of Buddhist philosophy. 9 However, Buddhist writers’ deployment of scriptural sources also holds special interest as it contains information about reading cultures of

different historical periods. Although citations have often been relegated to footnotes and other peripheral spaces, they are inscriptions in texts that can reveal histories of books in religious communities if they are examined with the kind of vigor to which inscriptions have been subjected in the study of Indian history. We cannot locate many Indian Buddhist authors geographically in the same way that inscriptions can be localized, but it is often possible to place a number of writers in the same text tradition in which later writers are aware of, and influenced by, the range of textual sources that their predecessors referred to. In other words, intertextuality plays a much greater role in the citation of texts than in the inscriptional records and has to be accounted for accordingly. For the purpose of such a study of citation and intertextuality, I have selected some interconnected portions of the texts composed by several Buddhist writers who belong to the Madhyamaka tradition. The pivotal figure of this exercise is Candrakīrti, a writer who continued the earlier Indian tradition of Madhyamaka interpretation and whose work also became

4. Lindtner (2001: 70): ato yuktyāgamopetaṃ tattvaṃ yat prāgudāhṛtam/ parīkṣyamāṇaṃ yuktyaivaṃ tad evāvyāhataṃ sthitam//. 5. MacDonald (2015: 1:205.4–5): ācāryo yuktyāgamābhyāṃ saṃśayamithyājñānayor apākaraṇārtham idam ārabdhavān/. 6. Joseph Walser, Nāgārjuna in Context: Mahāyāna Buddhism & Early Indian Culture (New York: Columbia Univ. Press, 2005), 271–78. 7. Hahn (1982: 162): uktaṃ etad bhagavatā hetur apy atra dṛśyate/. 8. On the use of this principle in Abhidharma texts, see Collett Cox, “The Unbroken Treatise: Scripture and Argument in Early Buddhist Scholasticism,” in Innovation in Religious Traditions: Essays in the Interpretation of Religious Change, ed. Michael A. Williams, Collett Cox, and Martin S. Jaffee (Berlin: Mouton de Gruyter, 1992), esp. 168–73. 9. Scholarly works that address the element of reason in Buddhist philosophy are too numerous to list. On the specific concept of yukti or reasoning, see Richard Nance, “On What Do We Rely When We Rely on Reasoning?” Journal of Indian Philosophy 35.2 (2007): 149–67, which also provides references to previous research. 545L i: Scripture and Vicissitudes of Textual Practices in Madhyamaka Thought

particularly influential in late Indian Buddhism and much of second-millennium Tibetan Buddhist thought. The comparative angle is supplied here by a consideration of the citations found in related Tibetan works composed by Tsong kha pa and in the earlier Madhyamaka commentaries, especially those written by Buddhapālita and Bhāviveka. Among these four Buddhist philosophers, each author was acquainted with the writings of every remaining author that predated him, if he was preceded by any. Therefore, they self-consciously regarded themselves as a part of the same philosophical tradition. In his recent work on the seventeenth chapter of the Prasannapadā (see n. 38 below), Ulrich Timme Kragh has demonstrated that Candrakīrti incorporated a very substantial amount of material from the earlier Madhyamaka commentaries. While my reading of the eighteenth chapter of the Prasannapadā against the earlier Madhyamaka commentaries confirms Kragh’s conclusion, the current study will bring Candrakīrti’s influence on Tibetan Madhyamaka thought into view while also emphasizing a form of intertextuality that is peculiar to the use of scriptural citations. With my narrower focus, I will demonstrate that we can uncover a gradual process of Madhyamaka writers’ collection of scriptural citations for the purpose of building a hermeneutic apparatus of its scholastic discipline. Buddhapālita already referred to many scriptural sources, but it is in Bhāviveka’s and Candrakīrti’s treatises and commentaries that the process gained momentum. These two authors also demonstrated a clear attention to the Mahāyāna sūtras. An underlying interest of this study is to discover the changing orientations in Buddhist scholastic practices, especially in regard to the use of the various categories of sūtra and śāstra literature.

the distinction between sūtra and śāstra Madhyamaka treatises and commentaries weave into their philosophical analyses an extensive number of citations. The majority of the texts cited come under the generic categories of śāstra and sūtra. The use of the term śāstra designates here commentaries on śāstras as well, since śāstras in the narrow sense and their commentaries share a great deal in content and method. Before we examine specific instances, it will be useful to discuss these designations very briefly to understand the significance of these textual categories. For the Madhyamaka writers discussed in this article, the term śāstra refers first and foremost to the foundational treatises of their tradition. Thus, in reference to the Mūlamadhyamakakārikā Candrakīrti says that Nāgārjuna, “the ācārya, has written this Madhyamaka Śāstra for the purpose of teaching the distinction between the provisional and definitive sūtrāntas.” 10 Statements of this kind offer a typical articulation of the relation between śāstras and sūtras, in which śāstras are seen as second-order formulations of the contents of sūtras, revealing what the uninitiated cannot learn by reading sūtras directly. In addition to the Mūlamadhyamakakārikā, influential Madhyamaka śāstras include Āryadeva’s Catuḥśataka, whose verses had been frequently cited at least since the time of Buddhapālita, and Ratnāvalī, which received particular attention from Candrakīrti. From Vasubandhu’s description of his own Abhidharmakośa as a śāstra, 11 Tarkajvālā’s reference to the author of Madhyamakahṛdaya as the author of the śāstra, 12 and Candrakīrti’s

10. MacDonald (2015: 1:201): ata evedaṃ madhyamakaśāstraṃ praṇītam ācāryeṇa neyanītārthasūtrāntavibh āgopadeśanārthaṃ/. 11. Śāstrī (1998: 1:3): śāstraṃ pravakṣyāmy abhidharmakośam. 12. For a recent appraisal of the problem of Tarkajvālā’s authorship in favor of its attribution to Bhāviveka and an instance of the phrase bstan bcos byed pa/śāstrakāra, see Malcolm David Eckel, Bhāviveka and His Buddhist Opponents (Cambridge, Mass.: Harvard Univ. Press, 2008), 21–23. 546 Journal of the American Oriental Society 137.3 (2017)

citation of the verses of his own Madhyamakāvatāra, 13 we learn that Buddhist writers recognize as śāstras not just canonized works of their tradition but their further expositions as well, including systematic treatises that they themselves have composed. Moreover, Mādhyamikas are also engaged in conversation with the śāstras of competing Indian text traditions. In the Madhymakāvatāra, Candrakīrti speaks of various śāstras of Indian philosophical traditions outside Buddhism, which expound notions of self that he endeavors to disprove. 14 Even within the Buddhist fold, Candrakīrti singles out the Yogācāra scholars Vasubandhu and Dharmapāla and the Buddhist epistemologist Dignāga as teachers who have turned their back on the unique tradition of Madhyamaka thought, although he recognizes them as authors of śāstras. 15 Bhāviveka also devoted several chapters of his Madhyamakahṛdaya to the criticism of rival śāstra disciplines ranging from Sāṃkhya, Vaiśeṣika, Vedānta, Mīmāṃsā to the Buddhist tradition of Yogācāra. It is evident from these statements that there is a clear awareness of one’s own śāstra

discipline as one among many and that the disparate śāstra traditions are often in conflict with one another. As is well known, śāstras are characterized by the deployment of numerous scholastic devices, while sūtras, presented as the teachings of the Buddha or the sayings of his disciples, are more varied in content and style. Within śāstras and their commentaries, the citation of passages from sūtras fulfills, above all, the purpose of justifying the views put forward in the śāstras by appealing to the scriptural status of the passages. The hierarchy of religious authority at work here, which assigns higher scriptural authority to

sūtras while delegating to śāstras the status of authorized interpretations of scriptures, would incline us to think that their respective functions in textual practice bear proportionate relation to the degrees of authority that they enjoy. However, the roles that sūtras and śāstras play in actual scholastic practice do not follow from their nominal status and even change over time, as the following pages will demonstrate. The śāstra-sūtra distinction considered so far will serve as a framework and a starting point for the examination of how their relationship is formed and transformed in the Ma dhyamaka tradition. The following analyses will demonstrate that despite the higher scriptural authority that sūtras appear to hold, evidence in the Buddhist literary history in India and Tibet points to a gradual rise of śāstras’ importance in regard to the roles that they have played in Buddhist scholasticism, such that they effectively replaced sūtras as the primary object of study, commentary, and debate. One practical means to measure the relative popularity of the two groups of texts is to compare the uses of śāstras and those of sūtras on the basis of the intertextual references that are found in the literature. By examining how certain writers refer to other texts, it is possible to obtain information about their reading habits.

citations in tsong kha pa’s madhyamaka works To carry out the kind of analysis that will lead to an outcome relevant to the question of the use of texts in scholastic practices, this and the following sections will gather data from 13. See List 1 below for the citations of the Madhyamakāvatāra in the eighteenth chapter of the Prasannapadā. At PPMV 214.3, Madhyamakāvatāra is referred to as a prakaraṇa. 14. MA 6.123: śāstre śāstre ye ’sya tīrthyair viśeṣā nirdiśyante tān ajātatvahetur/ yasmāt sarvān bādhate svaprasiddhaḥ santy asyāto nāpi sarve viśeṣāḥ//. In Li Xuezhu, “Madhyamakāvatāra-kārikā Chapter 6,” published with open access at Springerlink.com, DOI: 10.1007/s10781-014-9227-6. See the Tibetan translation in MA 241. Candrakīrti discusses the attributes of ātman, or self, described in the Sāṃkhya and Vaiśeṣika traditions in MA 6.121–23 and MABh thereto. See also James Duerlinger, The Refutation of the Self in Indian Buddhism: Candrakīrti on the Selflessness of Persons (London: Routledge, 2013), 56–59 and 93–98. 15. MA 407.

547L i: Scripture and Vicissitudes of Textual Practices in Madhyamaka Thought

several texts that belong to different phases of the Madhyamaka text tradition. The first instance is a sample taken from an early-fifteenth-century comprehensive Buddhist manual titled Lam rim chen mo, which was composed by the Tibetan writer Tsong kha pa. This text has enjoyed the reputation of being “one of the most renowned works of Buddhist thought and practice to have been composed in Tibet.” 16 Tsong kha pa reserved about a third of his treatise for the exposition of wisdom, which in turn contains a philosophical interpretation of the Mahāyāna Buddhist teaching of emptiness based on the Madhyamaka tradition and a treatment of related meditative practices. The portion on Madhyamaka philosophy occupies about two hundred pages in a standard modern edition, a fact that indicates the importance that the author attached to a śāstra discipline. We will focus here on the citations used in this philosophical section since it bears

the most direct relation to the Madhyamaka tradition that preceded him. 17 This rather lengthy section of the work contains many hundreds of citations, although the vast majority of them originate from śāstra sources. In contrast, we find only twenty-one quotations from the sūtras, which constitute less than five percent of the total textual references that the section contains. Table 1 furnishes a list of all the sūtra passages cited in Lam rim chen mo’s presentation of

Madhyamaka philosophy. A further detail that emerges as we scrutinize this relatively short list of sūtra citations is that the vast majority of the twenty-one passages were already directly or indirectly used in the Madhyamaka writings of the Indian author Candrakīrti, who is clearly declared by Tsong kha pa to be one of the primary Indian authorities on whom he relied. Only three of Tsong kha pa’s sūtra passages are not known to have been used by Candrakīrti. These three passages are two half stanzas from the Ratnaguṇasaṃcayagāthā (nos. 7 and 9) and a line from the Prajñāpāramitāhṛdaya (no. 8). Both texts are well-known Mahāyāna scriptures belonging to the Prajñāpāramitā class. Although Candrakīrti also cited at least two different passages from the former text, Tsong kha pa was

apparently independently familiar with it. 18 The second sūtra, known to the modern readers as the Heart Sūtra, is an extremely short text widely known to the Tibetans. 19 Candrakīrti’s frequent citation of the Samādhirājasūtra apparently also left an impact on Tsong kha pa. The citation of the stanzas 11–17 and 19–22 from the ninth chapter of this sūtra in the Lam rim chen mo (no. 16) is likely to impress the reader with its literary quality. It is a rare case where I have not been able to locate all these verses in Candrakīrti’s works. 20 Perhaps this is an instance where Tsong kha pa was motivated to locate the passage in the sūtra itself. Among the sūtra passages quoted both by Tsong kha pa and Candrakīrti, a few cases further strengthen the case of the latter’s influence on the former. In one instance (no. 18) Tsong kha pa cited two stanzas from the Samādhirājasūtra in the following sequence: stanza 7 of chapter 12 followed by stanza 16 of chapter 11. These two stanzas had been extracted in the same order in the Prasannapadā, which shows that Candrakīrti is to be credited with the initial perception of the two stanzas’ connection. Moreover, two sūtra passages listed in Table 1 do not exist in Tibetan translations that were available to Tsong kha pa independently

16. Tsong kha pa blo bzang grags pa, The Great Treatise on the Stages of the Path to Enlightenment (Ithaca, NY: Snow Lion Publications, 2000–2004), 1:17. 17. LRChM 567.13–769.7. 18. Candrakīrti’s two citations from the Ratnaguṇasaṃcayagāthā are in PPMV 166.11–167.4 and 353.8–354.2. Tsong kha pa also cites this text elsewhere in the Great Treatise, e.g., LRChM 450.12–14 and 454.10–12. 19. Georges Dreyfus reports that the Heart Sūtra is among a few short sūtras that are still used in present-day Tibetan liturgies. Dreyfus, The Sound of Two Hands Clapping: The Education of a Tibetan Buddhist Monk (Berkeley and Los Angeles: Univ. of California Press, 2003), 89. 20. Stanzas 13, 14, 15, and 19 do not appear to have been cited by Candrakīrti. 548 Journal of the American Oriental Society 137.3 (2017)

Table 1. Sūtra Citations in Tsong kha pa’s Lam rim chen mo, 567.13–769.7

Citations in LRChM Sūtra Sources Earlier Citations by Candrakīrti 1 568.18–569.3 and 569.6–12 Akṣayamatinirdeśasūtra (Braarvig 1993: 1:117–18) PPMV 43 2 569.16–18 Samādhirājasūtra VII 5 (Vaidya 1961: 36) PPMV 44 and 276 3 581.4–5 Samādhirājasūtra IX 23 (Vaidya 1961: 47) Yuktiṣaṣṭhikāvṛtti, D 3864, Dbu ma, vol. ’a, 5a7–b1; source of MA VI 30 and 31a (p. 112) 4 614.11 Samādhirājasūtra IX 23a (see no. 3) Yuktiṣaṣṭhikāvṛtti (see no. 3) 5 636.8–10 Anavataptanāgarājaparipṛcchā, D 156, Mdo sde, vol. pha, 230b2–3 PPMV 239, 491, 500, 504 6 636.15–16 Laṅkāvatārasūtra (Nanjio 1923: 76) PPMV 504 7 641.15–16 Prajñāpāramitāratnaguṇasaṃcayagāthā I 9cd (Vaidya 2003: 353) 8 642.1–2 Prajñāpāramitāhṛdaya (Vaidya 2003: 98) 9 642.3–4 Prajñāpāramitāratnaguṇasaṃcayagāthā I 28cd (Vaidya 2003: 355) 10 646.6–7 Śālistambasūtra (Reat 1993: 33), Aṅguttaranikāya (Morris et al. 1955–1961: 1:286), Saṃyuttanikāya (Feer 1884–1898: 2:25) PPMV 40 11 666.3–9

Pañcaviṃśatisāhasrikāprajñāpāramitā (Dutt 1934: 260–61) MABh 295 12 720.2–5 Saṃyuttanikāya (Feer 1884–1898: 1:135) MABh 257–8; cited also in Abhidharmakośabhāṣya IX, T 1558 XXIX 154b18–21; second stanza cited in Tarkajvālā, D 3856, Dbu ma, vol. dza, 80b3 13 732.8 The story of King Māndhātṛ; see, e.g., T 40 I 825a13 MABh 248; PPMV 574 14 732.19–20 See no. 13 PPMV 574 15 745.15–746.3 Samādhirājasūtra XXIX 13–16 (Vaidya 1961: 174) PPMV 109–110, PPMV 200, PPMV 549–550 (only XXIV 13cd, 15cd, and 16), MABh 144 (XXIX 13cd–14ab); see also PPMV 427 16 749.6–750.8 Samādhirājasūtra IX 11–17 and 19–22 (Vaidya 1961: 46–47) PPMV 178 (IX 17, 11), PPMV 550 (IX 17), Catuḥśatakaṭīkā (IX 11, 12, 16, 17, 20, 21, 22);a cf. PPMV 346 with IX 20

a. See Suzuki Kōshin, ed., Sanskrit Fragments and Tibetan Translation of Candrakīrti’s Bodhisattvayogācāracatuḥśatakaṭīkā (Tokyo: The Sankibo Press, 1994), 413; Karen Christina Lang, “On the Middle Indic Forms Found in Candrakīrti’s Quotations from Chapter Nine of the Samādhirājasūtra,” in Aspects of Buddhist Sanskrit: Proceedings of the International Symposium on the Language of the Sanskrit Buddhist Texts (Oct. 1–5, 1991), ed. K. N. Mishra (Sarnath: Central Institute of Higher Tibetan Studies, 1993), 445–46.

apart from their fragmentary quotation. The first passage is a stanza that both Candrakīrti and Tsong kha pa cited (no. 21), whose source both writers named as the Hastikakṣyasūtra. However, this passage is not found anywhere in the available Tibetan translation. 21 Another instance is a passage (no. 17) that originates from an early Buddhist sūtra that exists today in the forms of the Puppha Sutta in the Pāli Saṃyuttanikāya and a corresponding Chinese translation in the Saṃyuktāgama. This striking sūtra passage reads as follows in its Pāli version: “O monks! I do not argue with the world; but the world argues with me. The speaker of dhamma, monks, does not argue with anyone in the world. What is accepted by the wise people in the world to be not existing, monks, I also say that it does not exist. What is accepted by the wise people in the world to be existing, monks, I also say that it exists.” 22 The sūtra is not available independently in Tibetan. Consequently, Tsong kha pa identified a related passage in the Mahāyāna sūtra Trisaṃvaranirdeśa, which makes a reference to the statement in the Saṃyuttanikāya/Saṃyuktāgama but only records a partial version of that passage: “Thus I said, ‘The world argues with me, but I do not argue with the world.’” 23 The shorter version appears once in the Madhyamakāvatārabhāṣya, but on two separate occasions Candrakīrti cites the longer version that is not found in its entirety in the

21. This stanza is also not found in the two Chinese translations of the sūtra: T 813 and T 814. The language of Tsong kha pa’s citation (LRChM 763.10–13) differs slightly from the stanza in the Tibetan translation of PPMV (D 3860 Dbu ma, vol. ’a, 126a3 and 171a4–5). Minor departure of Tsong kha pa’s citations from the canonical versions is not uncommon. 22. Feer (1884–1898: 3:138): nāham bhikkhave lokena vivadāmi loko ca mayā vivadati// na bhikkhave dhammavādī kenaci lokasmiṃ vivadati// yam bhikkhave natthi sammataṃ loke paṇḍitānam aham pi tam natthī ti vadāmi// yam bhikkhave atthi sammataṃ loke paṇḍitānam aham pi tam atthīti vadāmi//. For the parallel in the Chinese Saṃyuktāgama, see T 99 II 8b16–26. 23. D 45, Dkon brtsegs, vol. ka, 9b5: des na ngas ’di skad du ’jig rten ni nga la rgol gyi/ nga ni ’jig rten dang mi rtsod do zhes gsungs so /. See the Chinese translations at T 310 XI 5a7–8 and T 311 XI 689b19. 17 751.19–20 Saṃyuttanikāya (Feer 1884– 1898: 3:138), Saṃyuktāgama T 99 II 8b16–26; source identified in LRChM as Trisaṃvaranirdeśaparivarta, D 45, Dkon brtsegs, vol. ka, 9b5

PPMV 370.6–8, MABh 179 and 289; source of MA VI 82 and VI 166 18 753.8–12 Samādhirājasūtra XII 7 and XI 16 (Vaidya 1961: 77, 70); cited in the same sequence as in PPMV PPMV 128

19 762.19–763.1 Anavataptanāgarājaparipṛcchā; same as no. 5 PPMV 239, 491, 500, 504 20 763.5–6 Anavataptanāgarājaparipṛcchā, D 156, Mdo sde, vol. pha, 230b2 PPMV 505

21 763.10–12 Hastikakṣya, not found in the extant Tibetan and Chinese translations PPMV 388, 514

550 Journal of the American Oriental Society 137.3 (2017)

Trisaṃvaranirdeśa, which makes it clear that what the Indian writer had in mind was an early Buddhist sūtra. 24 This influential fifteenth-century Tibetan work on Madhyamaka philosophy therefore reveals a state of textual practices in which the use of sūtras is almost completely mediated by the study of śāstras and commentaries. The citations in Tsong kha pa’s four remaining major Madhyamaka works appear to follow a similar pattern. Data is available for a text written as the condensed version of the Lam rim chen mo thirteen years after its composition. 25 The Madhyamaka section of Tsong kha pa’s Lam rim chung ngu shows the evidence of substantial reworking of the materials presented in the earlier work, an aspect of which is the incorporation into its one hundred and forty-six

citations of seventy-six passages that had not been used in the Lam rim chen mo. However, the seventy-six new citations are still dominated by the same śāstra sources and include only nine passages from sūtras and one from the Guhyasamāja Tantra. 26 The larger question that Tsong kha pa’s citation practices raises is whether there was any vigorous sūtra reading culture at this time in the history of Tibetan Buddhism. Ethnographic data have indeed revealed that present-day Tibetans are rarely involved in the study of the Buddhist sūtras. 27 Given that the amount of evidence considered here is extremely limited, to refrain from making disproportionate generalization, the case of Tsong kha pa suggests that, if sūtras were read separately, in the area of Madhyamaka studies they did not play an independent role in stimulating new ideas and reformulating the theoretical system. A reused sūtra passage offers no less persuasive force or emotional impact. For those who are aware of it, the intertextual connection in fact strengthens the sense of tradition, which is one of the important functions that citations perform.

śāstra and sūtra citations in candrakīrti’s prasannapadā The literary evidence from the beginning of the seventh century in India reflects a very different textual landscape with regard to the different levels of importance that were attached to sūtras, on the one hand, and śāstras and commentaries, on the other. To gauge the state of Buddhist textual practices in this period, I have gathered information about the citations used the eighteenth chapter of Candrakīrti’s Prasannapadā. Building on the earlier Madhyamaka commentarial tradition, the chapter discusses a range of topics in its rich interpretation of Nāgārjuna’s twelve verses in the Mūlamadhyamakakārikā, which proceed from an examination of the notion of self (ātman) to a description of the characteristics of reality

24. For the citation of the shorter version, see MABh 289.1–2. The longer version is cited at MABh 179.16–20 and PPMV 370.6–8. Moreover, in their references to the same source, neither Buddhapālita nor Bhāviveka cited the part of the sūtra passage that is found in the Trisaṃvaranirdeśa, nor did Candrakīrti in MA VI 82. See Lindtner 1981: 197 and 208; Bhāviveka, Dbu ma rtsa ba’i ’grel pa shes rab sgron ma (Beijing: krung go’i bod rig pa dpe skrun khang, 1994–2005), 57: 1274; MA 180.3–4. When Tsong kha pa uses this sūtra quotation (LRChM 751.19– 20, Rigs pa’i rgya mtsho ad Mūlamadhyamakakārikā XVIII 8, and Dgongs pa rab gsal ad MA VI 81), he generally cites the longer passage found in the Indian Madhyamaka texts. 25. For a list of Tsong kha pa’s five major works on Madhyamaka philosophy, see Jeffrey Hopkins, Tsong-khapa’s Final Exposition of Wisdom (Boston: Snow Lion, 2008), 16–18. 26. For a summary of the seventy passages that appear in the Madhyamaka sections of both Lam rim chen mo and Lam rim chung ngu and the seventy-six citations that are used only in the latter work, see the two tables provided in ibid., 18–22. Most of the nine sūtras from which Tsong kha pa cited new passages had been referenced in Candrakīrti’s works. 27. See, for instance, Dreyfus, Sound of Two Hands Clapping, 109.

(tattvasya lakṣaṇam). 28 Candrakīrti supports the presentation in the chapter with abundant scriptural citations that, when analyzed, also illustrate the degree of intertextuality involved in a growing commentarial tradition’s citation practices.

List 1: Citations in the Eighteenth Chapter of Candrakīrti’s Prasannapadā (PPMV 340–81) 1. Citations of Buddhist śāstras (22 passages) (A) Citations of the Mūlamadhyamakakārikā of Nāgārjuna (5 passages), excluding the stanzas from the eighteenth chapter that the Prasannapadā immediately comments on: 29 (1) Mūlamadhyamakakārikā XXVII 12 (PPMV 341.11–12); (2) XXVII 6 (PPMV 342.2–3); (3) XXIII 1 (PPMV 350.8–9); (4) XXV 24 (PPMV 364.15–16); (5) XIV 6ab (PPMV 376.8) (B) Citations of the Ratnāvalī, attributed to Nāgārjuna by Candrakīrti (5 passages): (6) Ratnāvalī I 31–34 (PPMV 345.5–12), Hahn 1982: 14 and 15; (7) I 29–30 (PPMV 346.5– 8), Hahn 1982: 12 and 13; (8) I 52–54 (PPMV 347.5–10), Hahn 1982: 22 and 23; (9) II 3–4 (PPMV 359.1–4), Hahn 1982: 40 and 41; (10) IV 94–96 (PPMV 359.11–360.2), Hahn 1982: 128–131 (C) Citations of the Catuḥśataka of Āryadeva (6 passages): (11) Catuḥśataka XII 23 (PPMV 351.13–14), Lang 1986: 116; (12) VIII 15 (PPMV 359.8– 9), Lang 1986: 82; (13) VIII 19 (PPMV 370.4–5), Lang 1986: 84; (14) VIII 20 (PPMV 372.5–6), Lang 1986: 84; (15) X 25 (PPMV 376.14–15), Lang 1986: 102; (16) VIII 22 (PPMV 378.4–5), Lang 1986: 86 (D) Citation of Bhāviveka’s work (2 passages): (17) Prajñāpradīpa (PPMV 351.16–352.6), D 3853, Dbu ma, vol. tsha, 183b4–7; (18) Prajñāpradīpa (PPMV 369.4–7), D 3853, Dbu ma, vol. tsha, 188b1–3 (E) References to Candrakīrti’s Madhyamakāvatāra (4 passages): (19) VI 120 (PPMV 340.8–11), MA 233; (20) VI 127–128 (PPMV 342.5–12), MA 245 and 247; (21) VI 121 (PPMV 344.5–8), MA 235; (22) I 8d (PPMV 353.1), MA 19 2. Citation of a non-Buddhist śāstra (1 passage): (23) A verse associated with the Lokāyata tradition (PPMV 360.6–7); see Lokatattvanirṇaya 113, in Suali 1887: 290 3. Citations of early Buddhist texts (6 passages): (24) Kṣudrakāgama (PPMV 348.11–12), cited in Vasubandhu’s Abhidharmakośabhāṣya IX, where the source is identified as Kṣudrakāgama (Śāstrī 1998: 2:933); (25) Ekottarāgama (PPMV 350.11–12, also in PPMV 451.12–13), T 125 II 687b22–23; (26) (PPMV 355.4) Kṣudrakāgama (cited in Abhidharmakośabhāṣya, Śāstrī 1998: 2:933); (27) Saṃyuktāgama

28. The continuity of the Madhyamaka commentarial tradition is visible in the following four Indian commentaries on the chapter: Akutobhayā, Buddhapālita’s vṛtti, Bhāviveka’s Prajñāpradīpa, and Candrakīrti’s Prasannapadā. For translations of the latter three commentaries on the eighteenth chapter of MMK into Western languages, see Lindtner 1981: 187–217; Malcolm David Eckel, “A Question of Nihilism: Bhāvaviveka’s Response to the Fundamental Problems of Mādhyamika Philosophy” (Ph.D. diss., Harvard Univ., 1980), 192–264; and J. W. de Jong, Cinq chapitres de la Prasannapadā (Paris: Geuthner, 1949), 1–36. 29. Stanzas being commented on are less relevant for our purpose. They have been excluded so that we can focus on the external sources and occasional references to the other chapters of MMK. The same principle will be followed in Table 2 and Lists 2 and 3.

(PPMV 355.5–6), Feer 1884–1898: 3:44 and 4:287, T 99 II 7c22–24, etc.; (28) Saṃyuktāgama (PPMV 370.6–8, also in MABh 179 and 289; source of MA VI 82 and VI 166), Feer 1884– 1898: 3:138, T 99 II 8b16–26; (29) Dharmapada (PPMV 354.5–6), stanza 160, in Hinüber and Norman 1994: 45; last two pādas cited in Abhidharmakośabhāṣya (Śāstrī 1998: 1:84); cited later in Prajñākaramati’s Bodhicaryāvatārapañjikā on IX 73, in Vaidya 1960b: 232 4. Citations of Mahāyāna sūtras (16 passages): (A) The Prajñāpāramitā class (5 passages): (30) Aṣṭasāhasrikāprajñāpāramitā (PPMV 353.3–6), Vaidya 1960a: 3–4; (31) Ratnaguṇasaṃcayagāthā II 4 (PPMV 353.8–354.2), Vaidya 2003: 356; (32) Aṣṭasāhasrikāprajñāpāramitā (PPMV 379.4–380.2), Vaidya 1960a: 238; (33) Aṣṭasāhasrikāprajñāpāramitā (PPMV 380.3–10), Vaidya 1960a: 257–58; (34) Aṣṭasāhasrikāprajñāpāramitā (PPMV 380.11–381.11), Vaidya 1960a: 259 (B) The Ratnakūṭa class (4 passages): (35) Kāśyapaparivarta (PPMV 358.10–12, source named in PPMV as Āryaratnakūṭa), StaëlHolstein 1926: 87; (36) Tathāgataguhyasūtra (PPMV 361.1–363.12), D 47, Dkon brtsegs, vol. ka, 161a2–162a4; (37) Tathāgataguhyasūtra (PPMV 366.1–7), D 47, Dkon brtsegs, vol. ka, 132b6–133a1; (38) Tathāgataguhyasūtra (PPMV 366.9–367.4), T 312 XI 722b23 (C) Samādhirājasūtra (3 passages): (39) Samādhirājasūtra XXXVII 35 (PPMV 354.10–355.02), Vaidya 1961: 268; (40) VIII 4, 5 (PPMV 367.13–16, also in PPMV 278.5–12), Vaidya 1961: 42; (41) XIV 87 (PPMV 368.2–3), Vaidya 1961: 93 (D) Other Mahāyāna sūtras (4 passages): (42) Avataṃsakasūtra (PPMV 367.6–10), T 279 X 79a23–b3; (43) Akṣayamatinirdeśasūtra (PPMV 374.2–3), D 175 Dpe bsdur ma ed. 60.311, T 397 XIII 197b8–10, T 403 XIII 597a5 (identified as Akṣayamatinirdeśa by Bhāviveka, also agrees with Bodhisattvapiṭakasūtra, T 310(12) XI 300c26–27, T 316 XI 872b4–5); (44) Satyadvayāvatārasūtra (PPMV 374.5– 375.6), D 179

(Saṃvṛtiparamārthasatyanirdeśa), Mdo sde, vol. ma, 148a5–149a4; (45) Lalitavistarasūtra XIII 102 (PPMV 377.1–2), Vaidya 1958: 126 5. Citations of sūtra passages of unknown identity (3 passages): (46) Said to originate from a sūtra (PPMV 349.11–12); (47) follows the same refrain as the preceding stanza in PPMV (see no. 31 of this table) (PPMV 354.7–8) 6. Unidentified sources (3 passages): (48) (PPMV 348.14–349.2, also in PPMV 133.14–134.4 and 429.12–430.4); (49) (PPMV 349.4–7); (50) (PPMV 370.2–3); cited also in Subhāṣitasaṃgraha, Bendall 1903: 385 7. A ubiquitous Buddhist statement: (51) (PPMV 355.7) From List 1’s classification of citations found in the eighteenth chapter of the Prasannapadā, it becomes immediately clear that Candrakīrti cites from Buddhist sūtras as least as frequently as he does from śāstra sources. The contrast between Table 1 and List 1 indicates that Candrakīrti lived in a Buddhist community where there was a stronger interest in the reading of Buddhist sūtras. Within the category of śāstras, it is not surprising that the quotations used by Candrakīrti, the Mādhyamika, were extracted mostly from the treatises composed by the founding members of his own tradition. Among these texts he cites 553L i: Scripture and Vicissitudes of Textual Practices in Madhyamaka Thought

five passages each from the Mūlamadhyamakakārikā (nos. 1–5) and Ratnāvalī (nos. 9–13) and six stanzas from Āryadeva’s Catuḥśataka (nos. 11–16) to relate relevant Madhyamaka śāstra passages to the points being discussed. On a few occasions, Candrakīrti also refers to Madhyamakāvatāra, a versified summary of Madhyamaka thought that he had composed earlier (nos. 19–22). Three additional śāstra citations in the chapter, on the other hand, reflect contemporary debates that took place among the Mādhyamikas and in the inter-sectarian context. All three citations are linked to the work of Bhāviveka, a Madhyamaka writer who predates Candrakīrti. One of them is a verse associated with the Cārvāka tradition of Indian philosophy (no. 23), which Bhāviveka has cited in the Prajñāpradīpa to represent the opinion of the rival group. Another citation from the Prajñāpradīpa (no. 17) presents Bhāviveka’s view that hearers (śrāvakas) and lone Buddhas (pratyekabuddhas), in contrast to the bodhisattvas, do not have the understanding of the Mahāyāna Buddhist doctrine of emptiness. Candrakīrti disagrees with his predecessor on this point and refers his reader to his argument in the Madhyamakāvatārabhāṣya for the position that an enlightened being following the early

Buddhist path must have realized emptiness. 30 William Ames has shown that Bhāviveka expresses his own view on this issue consistently in a number of places and that he criticizes his predecessor Buddhapālita for maintaining that early Buddhist scriptures teach emptiness, 31 a position that Candrakīrti also holds. That Candrakīrti sides with Buddhapālita is widely known from his critique of Bhāviveka’s method of logical argument in the first chapter of the Prasannapadā (PPMV 14.1–36.2) and his defense of Buddhapālita in that context. In the eighteenth chapter, aside from the debate about the relationship between early Buddhism and the doctrine of emptiness just discussed, there is a second point on which Candrakīrti displays his affinity with Buddhapālita’s interpretation. The discussion is about how the Mādhyamikas should respond to the criticism that their philosophy is a form of nihilism. After paraphrasing what appears to be Buddhapālita’s response to the criticism, Bhāviveka remarks that it is not effective and proceeds to provide his own reply. 32 Candrakīrti cites Bhāviveka’s paraphrase (no. 18), which he attributes to the teachers of the past (pūrvācāryā[ḥ]), without reproducing Bhāviveka’s negative assessment. In fact, Candrakīrti’s own response follows and expands what Buddhapālita wrote. This debate may indicate that Candrakīrti and Buddhapālita represent a small community of Madhyamaka scholars who hold certain views that are distinct from Bhāviveka’s more established interpretation. 33 The eighteenth chapter of the Prasannapadā also furnishes evidence for the active use of a wide variety of sūtras in the Indian Buddhist scholastic culture at the time of Candrakīrti.

30. MA I 8d (no. 22) is cited in the eighteenth chapter of PPMV, although the argument is found in MABh on the pāda. A part of Candrakīrti’s argument is to show that emptiness is already taught in the early Buddhist scriptures. 31. William L. Ames, “Bhāvaviveka’s Own View of His Differences with Buddhapālita,” in The SvātantrikaPrasaṅgika Distinction: What Difference Does a Difference Make, ed. Georges B. J. Dreyfus and Sara L. McClintock (Boston: Wisdom Publications, 2003), 54–56. 32. The paraphrase is found at D Dbu ma, vol. tsha, 188b1–3. Cf. Buddhapālita’s text in Lindtner 1981: 206–7. 33. The supposition that Candrakīrti was writing his commentary on MMK constantly consulting Buddhapālita’s and Bhāviveka’s texts does not work well for this particular case. It is difficult to explain why Candrakīrti cites Bhāviveka’s paraphrase of Buddhapālita, whose reply Bhāviveka is critical of, while giving his own reply in large part following Buddhapālita. Why does he not cite or paraphrase Buddhapālita himself? Why does he not respond to Bhāviveka’s negative assessment of Buddhapālita? One possible answer to this question is that Bhāviveka was referring to several teachers of the past, hence the plural pūrvācāryā[ḥ], of whose views Bhāviveka’s paraphrase might be a better representation.

Among the twenty-two identified sūtra passages that appear in the chapter, sixteen are found in the Mahāyāna sūtras. Two popular Mahāyāna sūtra classes, the Prajñāpāramitā and what the Chinese and Tibetan catalogs call the Ratnakūṭa, 34 are represented by five (nos. 30–34) and four (nos. 35–38) citations respectively. The chapter also contains three passages from the Samādhirājasūtra (nos. 39–41), from which dozens of stanzas and a prose passage are cited in the Prasannapadā alone. 35 Candrakīrti also cites one passage from each of the following four Mahāyāna sūtras (nos. 42–45): Lalitavistara, Avataṃsaka, Akṣayamatinirdeśa, and Satyadvayāvatāra. The citations from the early Buddhist sūtras found in Candrakīrti’s chapter (nos. 24–29) mark the intersection between Madhyamaka thought and the early Buddhist theory of no self (anātman), which is treated in the earlier part of Nāgārjuna’s chapter and in the commentaries. They also illustrate the Madhyamaka tradition’s creative uses of early Buddhist texts. The citations represent major bodies of early Buddhist scriptures, including Kṣudrakāgama, Ekottarāgama, Saṃyuktāgama, and Dharmapada. Among them, a passage originating from the Saṃyuktāgama (no. 28) later attracted the attention of Tsong kha pa, who located a citation of it in the Mahāyāna sūtra Trisaṃvaranirdeśaparivarta (Table 1, no. 17), as we saw earlier. The statement from the early Buddhist sūtra, the Pāli parallel of which is the Puppha Sutta in the Saṃyuttanikāya, that “I accept as existent what is accepted to be existent in the world; I accept as not existent what is accepted as not existent in the world” was used by Buddhapālita, Bhāviveka, and Candrakīrti in their commentaries on the same verse of the Mūlamadhyamakakārikā. It was therefore initially incorporated into the Madhyamaka textual tradition to lend its weight to the general idea that the Buddha described certain things as true or as not true for the pragmatic purpose of helping those who will, in their present circumstances, benefit from accepting such views. 36 However, in Candrakīrti’s own system this passage assumes a more significant role. In his independent work it contributes to the important idea that conventional truth (saṃvṛtisatya) is just what is agreed upon by the ordinary people in the world. 37

the formation of a shared reserve of sūtra passages in the mid-first-miLLennium madhyamaka community The most significant differences between Tsong kha pa’s and Candrakīrti’s citation practices in their Madhyamaka works are the amount of the sūtra sources used and the extent to which the use of sūtra sources is conditioned by the prior tradition of śāstras and commentaries. Based on Tsong kha pa’s relatively sparing use of sūtra citations and his dependence on Candrakīrti’s works for such materials, it is difficult to avoid the conclusion that

34. In the Prasannapadā only Kāśyapaparivarta is referred to by the title of Ratnakūṭa, which is used in the Chinese and Tibetan canonical collections as the name for a class of sūtras. On the Ratnakūṭa class as an idea that originated in the Chinese Buddhist tradition, see Jan Nattier, A Few Good Men: The Bodhisattva Path according to The Inquiry of Ugra (Ugraparipṛcchā) (Honolulu: Univ. of Hawai’i Press, 2003), 31–36. The term is used here only as a convenient label for a group of sūtras as we know them now, rather than a class from Candrakīrti’s perspective. 35. See Collegiate Institute for the Study of Buddhist Literature, “The Sūtra of the King of Samādhis, Chapters I–IV,” in Studies in the Literature of the Great Vehicle: Three Mahāyāna Buddhist Texts (Ann Arbor: Collegiate Institute for the Study of Buddhist Literature and Center for South and Southeast Asian Studies, Univ. of Michigan, 1989), 32–34; Mitsukawa Toyoki, “Prasannapadā ni mirareru Gattō-zanmai kyō: sono inyōmen ni okeru naiyōkentei,” Indogaku Bukkyōgaku Kenkyū 15.2 (1967): 716–17. 36. The MMK verse in question is XVIII 8. For the Sanskrit of the sūtra passage, see PPMV 370.6–8. The interpretations provided by Buddhapālita, Bhāviveka, and Candrakīrti vary slightly. 37. Candrakīrti incorporated the sūtra passage’s idea into MA VI 82 and VI 166 and provided the citation on both occasions in MABh. See MA(Bh) 179.16–20, 180.3–4, and 288.20–289.2.

the prominence of śāstra grew tremendously in certain areas of Buddhist learning between the seventh and fifteenth centuries. This exercise in comparison provides a specific angle as well as further questions for the study of textual practices in the Indian Madhyamaka tradition. If the Tibetan Madhyamaka work of Tsong kha pa was highly influenced by his Indian predecessor Candrakīrti, were the scholastic works of Indian authors like Candrakīrti also highly mediated by the śāstras and commentaries of their own discipline? How did the Madhyamaka tradition in India gradually gather a body of scriptural passages as a part of its own hermeneutic apparatus? What kind of inference can we draw about the culture of sūtra reading around the time of Candrakīrti based on the analysis of citations found in the Madhyamaka texts? On the question of the mediation of a prior scholastic tradition, Kragh’s work on the seventeenth chapter of the Prasannapadā has already established that about one third of Candrakīrti’s sentences contain phrases, examples, quotations, and even complete lines that are also found in the earlier commentaries on the Mūlamadhyamakakārikā. 38 In particular, Kragh has used his data to argue that while Candrakīrti was unlikely to have direct knowledge of the two earliest extant Madhyamaka commentaries—the Akutobhayā and the work of Qingmu—he was consciously dependent on Buddhapālita’s and Bhāviveka’s Madhyamaka exegeses. 39 The data of citations obtained from the eighteenth chapter of the five early Madhyamaka commentaries on the

Mūlamadhyamakakārikā support the pattern that Kragh has discovered. Qingmu’s commentary on the eighteenth chapter has just two conspicuous scriptural quotations, 40 while the large amount of independent material that it contains indicates that it is not in the same line of successive commentaries to which the other texts appear to belong. Among the other four texts, Akutobhayā is the simplest and does not cite any text explicitly, although from this text a significant proportion of material found its way into the three later commentaries. In the eighteenth chapter, the close connection between Buddhapālita, Bhāviveka, and Candrakīrti is again confirmed by the shared materials in general and the passages that are cited by two or all three of these writers in particular. Table 2 below shows that eleven out of Buddhapālita’s fifteen citations were used by one or both of his successors. In the case of Bhāviveka, the citations that he shared with Buddhapālita, Candrakīrti, or both make up about a half of the quotations that he used in the chapter. Kragh’s discovery of significant bonds between Indian Madhyamaka commentators is based on his careful study of the so-called “exegetical parallels,” which are “words, phrases, clauses, or whole sentences” that are “used verbatim” in the context of commentators’ exegesis of the same verse. 41 A substantial proportion of the citations that I have collected from early Madhyamaka commentaries on the eighteenth chapter of the Mūlamadhyamakakārikā exhibit exegetical parallels, although this finding should be reassessed. The least significant point to be made here is that shared citations are not all used in the exegeses of the same verses. In some cases, a verse or a line that a commentator uses in the exegesis of one verse

38. See Ulrich Timme Kragh, Early Buddhist Theories of Action and Result: A Study of Karmaphalasambandha. Candrakīrti’s “Prasannapadā,” Verses 17.1–20 (Vienna: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien, 2006), 25–27. 39. Ulrich Timme Kragh, “Classicism in Commentarial Writing: Exegetical Parallel in the Indian Mūlamadhyamakakārikā Commentaries,” Journal of the International Association of Tibetan Studies 5 (2009): 41–42, 44, and 49–52. 40. T 1564 XXX 24c24 and 25a7–8. 41. Kragh, “Classicism in Commentarial Writing,” 17.

Table 2. Shared Citations in Buddhapālitamūlamadhyamakavṛtti, Prajñāpradīpa, and Prasannapadāa Buddhapālita Bhāviveka Candrakīrti no. 5 = no. 9 no. 2 no. 6 no. 3 = no. 10 not in chapter XVIII, PPMV 180.4–5 and 558.8–9 no. 4 no. 8 no. 11 no. 9 no. 29 no. 6 (on MMK 18.6) no. 14 (on MMK 18.8) no. 7 not in chapter XVIII, cf. PPMV 9.7 no. 11 no. 26 no. 12 no. 23 no. 15 (on MMK 18.7) no. 43 (on MMK 18.9) no. 8 no. 19 Cited in Candrakīrti’s Śūnyatāsaptativṛtti and by Avalokitavratab no. 10 = no. 3 See above no. 19a no. 18 no. 12 no. 20 no. 28 no. 13 not in chapter XVIII, PPMV 269.7–8 no. 14 no. 21 no. 15 no. 15 no. 22 no. 16 no. 25 = no. 11 a. The citations used in Buddhapālita’s and Bhāviveka’s MMK commentaries will be given in Lists 2 and 3 below. The citation numbers provided here refer to those given in Lists 2, 3, and 1. b. Lindtner 1981: 216 n. 91.

appears in another commentator’s interpretation of a different verse. 42 Additionally, different authors sometimes use familiar passages in different circumstances. Table 2 supplies a few instances in which Buddhapālita’s quotations appear in other chapters of Candrakīrti’s text or even in commentaries on different texts. 43 Such evidence inclines us towards the view that through their own reading and citation habits individual writers bring specific scriptural passage into the consciousness of their own community of interpreters. As a body of familiar citations becomes relatively stable within the community, individual passages get used in whatever contexts that the writers see fit. Although a significant proportion of citations do appear in the context of exegetical parallelism, usually functioning as proof-texts, scriptural passage can also be lifted out of such context to lend their ideas to the development of new philosophical positions, as we saw in Candrakīrti’s use of the Puppha Sutta in his independent compositions. 44

42. See, for instance, in Table 2, the parallel between Buddhapālita’s no. 6 and Candrakīrti’s no. 14 and that between Bhāviveka’s no. 15 and Candrakīrti’s no. 43. 43. See Buddhapālita’s nos. 3, 7, 8, 10, and 13 and Candrakīrti’s corresponding citations. 44. As shown earlier, the use of this specific source appears to have originally arisen out of the exegetical context. See Buddhapālita’s no. 12 (List 2), Bhāviveka’s no. 20 (List 3), and Candrakīrti’s no. 28 (List 1).

Moreover, shared scriptural passages are arguably distinct from common interpretations and philosophical arguments in that their use is even less confined to specific scholastic traditions. Scriptures, after all, are shared textual resources of the Buddhist communities at large. In Bhāviveka’s and Candrakīrti’s commentaries on the eighteenth chapter of the Mūlamadhyamakakārikā, a number of citations are known to have been used in the scholastic texts outside the Madhyamaka tradition. A line quoted by Candrakīrti to show that the Buddha taught no self was used at least twice in Bhāviveka’s commentary. The line occurred earlier in Vasubandhu’s own commentary on his Viṃśatikā, 45 a foundational treatise on Yogācāra philosophy, which was attacked by Bhāviveka and Candrakīrti as a rival tradition. This textual source must have gained some currency especially in the Yogācāra circle, as Bhāviveka’s contemporary Dharmapāla cites it in his own commentary on the Viṃśatikā. 46 Kuiji also states in his seventh-century Chinese commentary on the same treatise that he is able to find the stanza in three

Sanskrit manuscripts. 47 This line could have been introduced to the Madhyamaka tradition via its active use by a competing tradition, but it is likely that the citation of it from the Kṣudrakāgama in the Abhidharmakośabhāṣya was a common reference point for both the Madhyamaka and Yogācāra writers. In the Madhyamaka texts it functions merely as a scriptural witness to the Buddhist theory of no self, 48 rather than serving any unique interpretive agenda. The citations shared by the Madhyamaka and Abhidharma traditions point to an earlier history of the uses of certain scriptural sources. Indeed, some citations used by Bhāviveka and Candrakīrti appeared in the earlier work of Abhidharmakośabhāṣya. 49 A source used in the Prasannapadā, which also appears in a series of earlier Buddhist texts, demonstrates a prolonged interest in a Buddhist verse and the productiveness of that preoccupation. The verse in question reads: “O desire, I know your root. Surely, you arise from imagination. I will not fancy you. You will then not be mine.” 50 In the early strata of Buddhist texts, it generally occurs in the discussion of sexual desire’s disruptive effect on the ascetic life, and it is commonly accompanied by an illustrative narrative. The Ekottarāgama, one of the main scriptural collections of early Buddhism, contains a short sūtra that concludes its discourse with this verse. 51 As it became a part of various Dharmapada collections, the Dharmapada commentarial tradition followed its established interpretive technique and related simple but dramatic stories to serve as the narrative context of the verse. 52 In the early Abhidharma work Dharmaskandha, the narrative framework of the verse is still retained, and it occurs in the context of discussing the nature of desire. 53 In the Nyāyānusāra of Saṅghabhadra, who was Vasubandhu’s contemporary, the reference to the verse no longer concerns its content.

45. See Candrakīrti’s citation in List 1, no. 26 and Bhāviveka’s citations in List 3, nos. 11 and 25. On Vasubandhu’s use of this source in his vṛtti on Viṃśatikā 8, see Sylvain Lévi 1925: 5. 46. T 1591 XXXI 88c3–4. 47. T 1834 XLIII 990a14–18. 48. PPMV 355.4: nāstīha sattva ātmā vā dharmās tv ete sahetukāḥ/. 49. In the case of Candrakīrti, see, for instance, nos. 24, 26, and 29 in List 1. In the case of Bhāviveka, see, for instance, List 3, nos. 5, 9, 11, and 25. 50. List 1, no. 25. PPMV 350.11–12: kāma jānāmi te mūlaṃ saṃkalpāt kila jāyase/ na tvāṃ saṃkalpayiṣyāmi tato me na bhaviṣyasi//. 51. T 125 II 687b22–23. Cf. references given in Collett Cox, Disputed Dharma: Early Buddhist Theories of Existence. An Annotated Translation of the Section on Factors Dissociated from Thought from Saṅghabhadra’s Nyāyānusāra (Tokyo: International Institute for Buddhist Studies, 1995), 400 n. 16. 52. The verse is in T 210 IV 571b20–21; Bernhard 1965: 112. The commentaries on the verse are found in T 211 IV 603a23–b29 and T 212 IV 626c27–627a20. 53. T 1537 XXVI 482b17–c11.

The opinion of an opponent, who cites the first pāda of the verse merely as an example of the Abhidharma discussion of what counts as a name, a syllable, and a phrase, is reported in this work, 54 perhaps suggesting that the verse was still well known at the time. In the Madhyamaka work Tarkajvālā, the verse is invoked in the voice of someone who follows the so-called śrāvaka path to illustrate an opinion that is thought to be associated with early Buddhism. 55 In the Prasannapadā this verse with a clear early Buddhist connection (List 1, no. 25) is used to corroborate Nāgārjuna’s statement that defilements such as desire originate from conceptualization, which will be eliminated in the direct experience of emptiness. 56 This is another instance where Candrakīrti finds harmony

between early Buddhist texts and Madhyamaka positions. The trajectories of the two verses from the Kṣudrakāgama and Ekottarāgama examined here reveal two routes through which scriptural sources were absorbed into the Madhyamaka textual tradition. Both examples indicate that certain scriptural passages were previously used in scholastic context outside the Madhyamaka community before they were incorporated into the Madhyamaka texts. Some passages that belong to this group were familiar to most Buddhist scholastics through their use in the writings of several Buddhist textual traditions, and their roots might even go back to their use in the earlier Abhidharma texts. The other source of the scriptural passages, however, appears to be popular Buddhist culture. The traces of the verse from the Ekottarāgama betray its circulation in wider Buddhist circles. Its close association with narratives points to its role in didactic and oral discourses. Whatever their sources might be, after the passages were incorporated into the Madhyamaka texts some of them might even have become a part of the more stable reserve of scriptural citations through collective and repeated uses and the influence of the tradition’s pivotal members.

the increase in number and variety of sūtra citations: the disjunction between buddhapāLita and bhāviveka It is true that Madhyamaka commentators cite common scriptural passages, but what is of particular interest to historians of Buddhism is individual writers’ unique patterns of citation, which may tell us something specific about what texts were available to them and how they used them. In fact, the citation practices of the five early Madhyamaka commentators vary significantly. Focusing again on the commentaries on the eighteenth chapter of the Mūlamadhyamakakārikā as an example, 57 the most apparent fact is that later texts employ increasingly more citations. Between the two earliest commentaries on the eighteenth chapter, the Akutobhayā does not refer to any text other than Mūlamadhyamakakārikā, while Qingmu’s commentary uses just two explicit citations, as we have seen above.

54. T 1562 XXIX 413b5–8. See Cox, Disputed Dharma, 379. 55. Eckel, Bhāviveka and Opponents, 306 and 109. 56. MMK (XVIII 5) 302, karmakleśakṣayān mokṣaḥ karmakleśā vikalpataḥ/ te prapañcāt prapañcas tu śūnyatāyāṃ nirudhyate//. In PPMV Candrakīrti cites the verse from the Ekottarāgama again (451.12–13) in his commentary on MMK XXIII 1. 57. We know the titles, or the names of the authors, of twelve Indian commentaries on the MMK. Among the extant commentaries on the eighteenth chapter, Yogācāra scholar Sthiramati’s commentary and Avalokitavrata’s seventh to eighth century sub-commentary have been excluded from consideration. On the basic facts of the twelve commentaries, see Kragh, “Classism in Commentarial Writing,” 7–10.

List 2: Citations in the Eighteenth Chapter of the Buddhapālitamūlamadhyamakavṛtti 1. Lindtner 1981: 192.1–3; cf. Saṃyuttanikāya 3:25 2. Lindtner 1981: 192.9–12; Catuḥśataka X 20 3. Lindtner 1981: 192.37–193.2; Saṃyuttanikāya 2:82 (cf. Saṃyuktāgama, at T 99 II 83b6–8) 4. Lindtner 1981: 193.14–17; Catuḥśataka XII 23 5. Lindtner 1981: 193.22–25; Mūlamadhyamakakārikā XXVII 8 6. Lindtner 1981: 194.16–19; Catuḥśataka VIII 20 7. Lindtner 1981: 194.31; Śālistambasūtra, in Reat 1993: 28 58 8. Lindtner 1981: 195.17–20; Catuḥśataka XIV 25 9. Lindtner 1981: 196.17–20; Prajñāpāramitāstotra 15, in Hikata 1958: 2 10. Lindtner 1981: 197.5–6; = no. 3 11. Lindtner 1981: 197.8–11; Catuḥśataka VIII 9 12. Lindtner 1981: 197.17–18; Saṃyuttanikāya 3:138 13. Lindtner 1981: 198.17–18; Saṃyuttanikāya 2:17 14. Lindtner 1981: 199.21–24; Catuḥśataka X 25 15. Lindtner 1981: 200.5–8; Catuḥśataka VIII 22 The citations that Buddhapālita uses in his eighteenth chapter, given in List 2, are more numerous than in the two earlier commentaries. When we take a closer look at Buddhapālita’s specific sources, we find that the verses from Āryadeva’s Catuḥśataka account for seven of the fifteen citations. He also cites a verse from the Prajñāpāramitāstotra (no. 9), a hymn written by the earlier

Madhyamaka author Rāhulabhadra, apart from a verse from another chapter of Nāgārjuna’s śāstra that he is commenting on. The frequent citation of Āryadeva’s verses to echo Nāgārjuna’s, a precedent that both Bhāviveka and Candrakīrti followed, indicates that there might have been some form of canonization of certain śāstras within the Madhyamaka tradition. Indeed, Candrakīrti’s works show comparable attention to established Madhyamaka śāstras, as we have already seen in his citation of the Ratnāvalī. Compared with his two successors, the number of Buddhapālita’s sūtra citations is relatively modest, as two of his six sūtra quotations were used in Nāgārjuna’s Mūlamadhyamakakārikā and were thus familiar to the Madhyamaka tradition prior to him. 59 It may seem unexpected that the sūtra sources he quotes in the chapter are mostly early Buddhist scriptures rather than Mahāyāna sūtras. 60 However, this appears less unusual in view of the fact that even Nāgārjuna’s Mūlamadhyamakakārikā refers mostly to early Buddhist scriptures, although the main ideas that Madhyamaka texts communicate are

apparently Mahāyānist. It is not until the time of Bhāviveka that Mahāyāna sūtra quotations began to appear with some frequency in the Madhyamaka commentaries. 58. Cf. PPMV 9.7–8. The last part of the citation, ’di med na ’di mi ’byung ngo, corresponds with the text of Mahāvastu: imasya asato idaṃ na bhavati. See PPMV 9 n. 7. 59. The sūtra citations are nos. 1, 3, 7, 10, 12, and 13. Nos. 7 and 13 were used in MMK I 10 and XV 7. Akira Saito has made similar comments on Buddhapālita’s citation patterns, noting that “(a)part from several brief quotations from sūtra-s Buddhapālita’s main authorities are Nāgārjuna and Āryadeva.” See his “A Study of the Buddhapālita-mūlamadhyamaka-vṛtti” (Ph.D. dissertation, Australian National Univ., 1984), xxviii–xxix. 60. The Śālistambasūtra (no. 7) has been described as a Mahāyāna sūtra not on the basis of its contents. See Reat 1993: 3–5. As noted above, this source had been used by Nāgārjuna already.

List 3: Citations in the Eighteenth Chapter of Bhāviveka’s Prajñāpradīpa 1. D 3853 Dbu ma, vol. tsha, 178b3: a sūtra passage, source not identified. Cited in MABh 244.15–18 and 254.14–16 2. D 179b7: MMK V 2 3. D 180a3: = no. 2 4. D 180b4: source identified as Prajñāpāramitāsūtra in the Chinese translation of Prajñāpradīpa (T 1566 XXX 105a5) 5. D 180b4–5: Dharmapada 160, source identified in the Chinese translation as a sūtra from the Āgamas; last two pādas cited in Abhidharmakośabhāṣya, in Śāstrī 1998: 1:84 6. D 184a4–5: Catuḥśataka X 20 7. D 184a5–6: Sāṃkhyakārikā 62cd 8. D 185b4: Catuḥśataka XII 23 9. D 185b5: same as no. 5 10. D 185b5: source not identified 11. D 186a7–b1: Kṣudrakāgama, cited in Abhidharmakośabhāṣya, in Śāstrī 1998: 2:933 12. D 186b4–5: two verses associated with the Lokāyata tradition. First verse cited in PPMV 360.6–7 (List 1, no. 23) 13. D 187a1–2: Suvikrāntavikrāmiparipṛcchā, in Hitaka 1958: 32 14. D 187b6:

Brahmaviśeṣacintiparipṛcchā, T 586 XV 48a12, T 587 XV 80a29–b1, Lhasa Bka’ ’gyur 161 (D 160), Mdo sde, vol. pa, 102b2 15. D 187b6: Akṣayamatinirdeṣa, Braarvig 1993: 1:73, T 403 XIII 597a4, T 397(12) XIII 197b8–10 16. D 187b7: common phrase appearing in numerous sūtras in the Mdo sde section as well in the Ratnakūṭa and Prajñāpāramitā sections of the Bka’ ’gyur 17. D 188a4: Kāśyapaparivarta (referred to as Ratnakūṭasūtra in Prajñāpradīpa), Staël- Holstein 1926: 94–95 18. D 188a5–6: Brahmaviśeṣacintiparipṛcchā, T 585 XV 7b8–10, T 586 XV 39b10–12, T 587 XV 69c9–11, Lhasa Bka’ ’gyur 161 (D 160), Mdo sde, vol. pa, 61b3–4 19. D 188a6–7: Catuḥśataka XIV 25 19a. D 188b1–3: not a direct quote but a reference to Buddhapālita’s response to the criti cism that Mādhyamikas are nihilists 61 20. 189a6: Saṃyuttanikāya 3:138 21. 190b7: Catuḥśataka X 25 22. 191a4–5: Catuḥśataka VIII 22 23. 191b2–3: Suvikrāntavikrāmiparipṛcchā = no. 13 24. 191b3: source not identified 25. 191b4: Kṣudrakāgama = no. 11 26. 191b4–5: Mañjuśrīvikrīḍitasūtra, T 817 XVII 818c18–20, T 818 XVII 827a23–24, Lhasa Bka’ ’gyur 97 (D 96), Mdo sde, vol. kha, 362b2–3 The quotations found in the eighteenth chapter of the Prajñāpradīpa, given here in List 3, show that the kinds of sources Bhāviveka used finally bear resemblance to those used by Candrakīrti. As anticipated, Bhāviveka cites from early Madhyamaka śāstras, a category in which we find five verses from the Catuḥśataka (nos. 6, 8, 19, 21, 22), all of which were also used by Buddhapālita in the commentary on the same chapter, as well as one verse from the 61. See the discussion above of Candrakīrti’s reference to this passage (List 1, no. 18).

Mūlamadhyamakakārikā, which was quoted twice (nos. 2 and 3). The three early Buddhist sources identified in the chapter (nos. 5, 11, 20) had been cited earlier at least partially either in Buddhapālita’s commentary or in the Abhidharmakośabhāṣya, 62 both of which would have been familiar to Bhāviveka. Although his predecessors also engaged in disputations with non-Buddhist opponents, one aspect of Bhāviveka’s writing that is absent in the earlier Madhyamaka commentaries is the display of his vast knowledge of the philosophical views of many rival traditions and the resourcefulness with which he critically engaged with these positions. The Sāṃkhya (no. 7) and Cārvāka (no. 12) verses cited in the chapter hardly reflect the frequency with which philosophical encounters are rehearsed in the Prajñāpradīpa. His knowledge of the state of Indian thought is perhaps better represented in the treatment of various Buddhist and non-Buddhist philosophical systems he presented in the later chapters of the Madhyamakahṛdaya. The eighteenth chapter of the Prajñāpradīpa also contains eight passages quoted from the Mahāyāna sūtras, which are to be reckoned statistically as the most significant category of sources. Most of these citations are traceable to specific Mahāyāna sūtras—Suvikrāntavikrāmiparipṛcchā (no. 13 = no. 23), Brahmaviśeṣacintiparipṛcchā (nos. 14 and 18), Akṣayamatinirdeṣa (no. 15), Kāśyapaparivarta (no. 17), and Mañjuśrīvikrīḍitasūtra (no. 26)—while another passage (no. 16) appears in numerous Mahāyāna sūtras. These Mahāyāna sūtra passages are newly introduced into the interpretation of the specific verses in the chapter, and they are not acquired from a parallel exegetical context from a prior Madhyamaka commentarial

tradition. The fresh use of these passages, therefore, contrasts with other Buddhist sources adduced in the chapter—the early Buddhist scriptural passages that were familiar to the Mādhyamikas or the scholastic Buddhist communities at large. The impression that in this period not all of the Madhyamaka authors’ Mahāyāna sūtra sources come from a well-defined body of shared texts and passages is further strengthened by the fact that in this chapter only a very small proportion of Bhāviveka’s and Candrakīrti’s citations from this textual category overlap. We can identify no more than a shared passage from the Akṣayamatinirdeśa (no. 15 of List 3 and no. 43 of List 1) and one shared text, Kāśyapaparivarta (no. 17 of List 3 and no. 35 of List 1), in the two commentaries on the chapter. For instance, when these two commentators felt compelled to provide textual evidence for Nāgārjuna’s statement that “neither any self nor any no self was taught by the buddhas,” 63 Bhāviveka produced one passage from the Suvikrāntavikrāmiparipṛcchā, 64 while Candrakīrti cited another from the Kāśyapaparivarta. 65 There are certain Mahāyāna texts that were commonly used by Indian Buddhist writers. The Kāśyapaparivarta and Akṣayamatinirdeśa, the two Mahāyāna sūtras that both Bhāviveka and Candrakīrti cited from, indeed appear to be such texts, as they were referred to by many Buddhist authors since the early stage of Mahāyāna Buddhism. 66 On the other hand, it is possible that Bhāviveka’s use of Suvikrāntavikrāmiparipṛcchā and Candrakīrti’s use of Samādhirāja came from individual initiative, judging from the evidence that I am aware of. Ryusho Hikata has argued for a date of the Suvikrāntavikrāmiparipṛcchā based on the fact that no reference to this Prajñāpāramitā sūtra earlier than Bhāviveka has been

62. No. 9 = no. 5; no. 25 = no. 11. 63. MMK (XVIII 6cd) 302: buddhair nātmā na cānātmā kaścid ity api deśitam/. 64. List 3, no. 13. 65. List 1, no. 35. 66. See Staël-Holstein 1926: v, xiv n. 2, and xvi; Braarvig 1993: 2: lii–lvii.

found. 67 While the Samādhirājasūtra was referred to prior to the work of Candrakīrti, no one before Candrakīrti is known to have made use of it nearly as extensively as he did. 68 Based on our current state of knowledge, we can tentatively conclude that passages from these two Mahāyāna sūtras were brought into the tradition of Madhyamaka commentaries not because they were a part of a common curriculum or routinely recited scriptures. Bhāviveka’s and Candrakīrti’s references to these specific sūtras could perhaps provide a glimpse into the private readings of these two Madhyamaka writers.

concLusion Citations allow a glimpse into the world of texts that an author inhabits, while the social and religious environments as well as the prior textual traditions constitute the external context. Both contribute to our knowledge of the intellectual milieu from which philosophical ideas emerge. This short study has highlighted two major transitions in the citation practices of Madhyamaka writers. The first transition occurred around the sixth and seventh centuries in India, when the Mahāyāna sūtras became a clear category of textual reference in the Ma dhyamaka commentaries. A related trend also occurred at this time, when citations began to appear with greater frequency in these commentaries. The second transition is a very large process through which śāstras became the main

textual category to occupy the attention of Buddhist scholastics. While more study will be necessary to examine the gradual change of interest in the intervening centuries between Candrakīrti and Tsong kha pa, there are already some clues that might help explain why the authority of the śāstras grew in relation to that of sūtras. In the works of Buddhapālita, Bhāviveka, and Candrakīrti we have seen certain Madhyamaka texts, such as Catuḥśataka and Ratnāvalī, being referred to with great frequency, indicating the formation of core Ma dhyamaka texts. When major Madhyamaka śāstras like the Mūlamadhyamakakārikā became the primary texts of

interest and objects of exegetical effort, sūtra passages essentially had to assume a supporting role. Although they have a more exalted symbolic status, sūtras were often invoked to lend weight, give depth, or provide justification for the positions that were taken in the śāstras. In practice, śāstras provided the guidelines for the interpretation of the sūtras. Thus, in the Madhyamakāvatārabhāṣya, when an interlocutor suggests that scripture (āgama) itself be relied on for the ascertainment of reality, Candrakīrti rejected the proposal: “This is not so. Since the intention of the scripture is difficult to ascertain, those like us are not able to give instructions on reality even through scripture. I say so from the perspective of [giving instructions] independently. However, the intention of the scripture is ascertained by seeing the correct interpretations of the scripture, which are the śāstras composed by the trustworthy beings.” 69 The explanation that Candrakīrti supplies here suggests that śāstras have secured a very special place between canonical texts and the readers, assuming the role of an indispensable interpretive authority. The citation patterns of the early Indian Madhyamaka commentators on the eighteenth chapter of the Mūlamadhyamakakārikā also suggest a sudden emergence of inter

67. Hikata 1958: lxxxii. On Bhāviveka’s quotations from this sūtra, see ibid., lxxvi–lxxvii n. 1. 68. For citations from the Samādhirājasūtra, including Candrakīrti’s, see Collegiate Institute, “King of Samādhis,” 32–38. Sūtrasamuccaya, a work that predates Candrakīrti, cites from the sūtra only four times. 69. MABh 75: ’di yang yod pa ma yin te/ lung gi dgongs pa nges par dka’ ba’i phyir bdag cag ’bra bas lung las kyang de kho na nyid bstan par mi nus so // rang dbang nyid kyi dbang du byas nas de skad du brjod kyi/ bstan bcos tshad mar gyur pa’i skyes bus byas shing lung phyin ci ma log par ’chad pa mthong ba las lung gi dgongs pa nges pas ni/. I have emended chad pa to ’chad pa on the basis of D (To. 3862) Dbu ma, vol. ’a, 245a1.

est in the Mahāyāna sūtras around the time of Bhāviveka. It is tempting to cite the work of Gregory Schopen, who finds virtually no inscriptional evidence for traces of Mahāyāna Buddhism in India before the sixth century, 70 and to conclude that the evidence from the Mūlamadhyamakakārikā commentarial tradition corroborates his findings. However, one can hardly maintain this position when scholastic texts’ citations of Mahāyāna sūtras in general are taken into account. Indeed, references to Mahāyāna sūtras in the earlier works such as Daśabhūmikavibhāṣā (T 1521), Prajñāpāramitopadeśa (T 1509), and Mahāyānasūtrālaṅkāra are very extensive. Thus, religious activities affiliated with Mahāyāna are visible in the texts from earlier times, although not represented in the expression of institutional identity in the medium of inscription. In the case of the commentaries on the Mūlamadhyamakakārikā, the focus on the eighteenth chapter alone could have amplified the disjunction between Bhāviveka and his predecessors. In his work on the Akutobhayā, C. W. Huntington notes that this commentary has made eleven explicit references to other texts. Among its nine sūtra references, the sources of two are named as the Prajñāpāramitasūtra(s), while the other two traceable quotations are cited from the Anavarāgrasūtra, which corresponds to the Anamataggo-saṃyutta in the Saṃyuttanikāya. 71 Qingmu’s commentary embedded in

Kumārajīva’s translation of the Mūlamadhyamakakārikā, which refers to sūtras more frequently than the Akutobhayā does, also identifies the source of two citations as the Prajñāpāramitā. 72 Therefore, what distinguishes Bhāviveka and Candrakīrti from the earlier commentators is only the frequency with which they cited from the Mahāyāna sūtras. A probable explanation for the fact that earlier commentators infrequently or rarely referred to Mahāyāna sūtras is that certain Madhyamaka interpreters lived among followers of early Buddhism, an environment in which shared scriptures had greater power of purchase. That the

Mūlamadhyamakakārikā could be explained with sūtra references drawn almost completely from sources other than Mahāyāna sūtras, as is the case with Buddhapālita, says something about certain communities of its early interpreters as well as the religious and social milieu of Nāgārjuna himself. Previous scholarship on Buddhist śāstras has often privileged the role of reason. Placing śāstras in the context of their relation with scriptures highlights instead the hermeneutical dimension of Buddhist scholasticism. In many Buddhist scholastic writings, scriptural citations constitute a significant proportion of the texts, which demonstrates that constant engagement with scripture was an important aspect of the writers’ thought process. Buddhist writers from the Abhidharma era to the contemporary period have characterized the use of both scripture and reason as their basic scholastic method. Much work still lies ahead in the investigation of śāstras’ uses of scripture. Future research needs to study the range of textual strategies involved in the deployment of scripture, the manners in which scholastic communities and traditions maintained and transmitted selected contents of the earlier texts, and the extent to which the Buddhist scholastic enterprise was a hermeneutical process that often produced fresh ideas through reading received texts in new contexts. Such investigations will open a new window onto Buddhist scholastic cultures of the past.

70. Gregory Schopen, “The Mahāyāna and the Middle Period of Indian Buddhism: Through a Chinese Looking Glass,” in Figments and Fragments of Mahāyāna Buddhism in India: More Collected Papers (Honolulu: Univ. of Hawai’i Press, 2005), 3–24. 71. Clair W. Huntington, “The Akutobhayā and Early Indian Madhyamaka” (Ph.D. dissertation, University of Michigan, 1986), 1: 202–5. 72. T 1564 XXX 24c24, which appears in the eighteenth chapter, and T 1564 XXX 1b28. 564 Journal of the American Oriental Society 137.3 (2017) abbreviations D Sde dge edition of the Tibetan Bka’ ’gyur and Bstan ’gyur LRChM Lam rim chen mo; in Tsong kha pa Blo bzang grags pa 1985 MA and MABh Madhyamakāvatāra and Madhyamakāvatārabhāṣya; in Louis de La Vallée Poussin 1907–12 MMK Mūlamadhyamakakārikā; in Shaoyong Ye 2011 PPMV Prasannapadā and the edition of Louis de La Vallée Poussin 1903–13 T The Taishō edition of the Chinese Buddhist scriptural collection; in Junjirō Takakusu and Kaigyoku Watanabe 1924–32

bibliography of text editions Bendall, Cecil. 1903. “Subhāṣita-Saṃgraha: An Anthology of Extracts from Buddhist Works Compiled by an Unknown Author, to Illustrate the Doctrine of Scholastic and of Mystic (tāntrik) Buddhism.” Le Muséon 22: 375–402. Bernhard, Franz. 1965. Udānavarga, vol. 1: Einleitung, Beschreibung der Handschriften, Textausgabe, Bibliographie. Göttingen: Vandenhoeck & Ruprecht. Bhāviveka, Jñānagarbha, and Klu’i rgyal mtshan. 1994–2008. Dbu ma rtsa ba’i ’grel pa shes rab sgron ma (Prajñāpradīpa). In Bstan ’gyur (dpe bsdur ma) 57: 935–1516. Beijing: Krung go’i bod kyi shes rig dpe skrun khang. Braarvig, Jens, ed. and tr. 1993 Akṣayamatinirdeśasūtra. 2 vols. Oslo: Solum. Cabezón, José Ignacio, and Geshe Lobsang Dargyay. 2007. Freedom from Extremes: Gorampa’s “Distinguishing the Views” and the Polemics of Emptiness. Boston: Wisdom Publications. Dutt, Nalinaksha. 1934. The Pañcaviṃśatisāhasrikā Prajñāpāramitā. London: Luzac & Co. Feer, Léon, ed. 1884–1898. The Saṃyutta-nikāya of the Sutta-piṭaka. 5 vols. London: published for the Pali Text Society by Henry Frowde. Hahn, Michael. 1982. Nāgārjuna’s Ratnāvalī, vol. 1: The Basic Text (Sanskrit, Tibetan, Chinese). Bonn: Indica et Tibetica Verlag. Hikata, Ryusho. 1958.

Suvikrāntavikrāmi-paripṛcchā Prajñāpāramitā-sūtra. Fukuoka: Kyushu University. Hinüber, Oskar von, and K. R. Norman, eds. 1994. Dhammapada. Oxford: Pali Text Society. Lang, Karen. 1986. Āryadeva’s Catuḥśataka: On Bodhisattva’s Cultivation of Merit and Knowledge. Indiske Studier 7. Copenhagen: Akademisk Forlag. La Vallée Poussin, Louis de, ed. 1903–13. Madhyamakavṛttiḥ: Mūlamadhyamakakārikā (Mādhyamikasūtras) de Nāgārjuna avec la Prasannapadā, commentaire de Candrakīrti. St. Petersburg: Aca démie Impériale des Sciences. , ed. 1907–12. Madhyamakāvatāra par Candrakīrti: Traduction tibétaine. St. Petersburg: Académie Impériale des Sciences. Lévi, Sylvain, ed. 1925. Vijñaptimātratāsiddhi: Deux traités de Vasubandhu. Viṃśatikā (La vingtaine) accompagnée d’une explication en prose, et Triṃśikā (La trentaine) avec le commentaire de Sthiramati. Paris: Librairie ancienne Honoré Champion. Li, Xuezhu. Madhyamakāvatāra-kārikā Chapter 6. Published with open access at Springerlink.com. DOI: 10.1007/s10781–014–9227–6. Lindtner, Christian. 1981. Buddhapālita on Emptiness [Buddhapālita-mūla-madhyamakavṛtti XVIII]. Indo-Iranian Journal 23: 187–217. , ed. 2001. Madhyamakahṛdaya of Bhavya. Adyar: Adyar Library and Research Centre. MacDonald, Anne. 2015. In Clear Words: The Prasannapadā, Chapter One. 2 vols. Vienna: Verlag der Österreichischen Akademie der Wissenschaften Morris, Richard, Edmund Hardy, Mabel Hunt, and Caroline A. F. Rhys Davids, eds. 1955–1961. The Aṅguttara-nikāya. London: Published for the Pali Text Society by Luzac. Nanjio, Bunyiu. 1923. The Laṅkāvatāra Sūtra. Kyoto: Otani University Press.

Reat, N. Ross. 1993. The Śālistamba Sūtra: Tibetan Original, Sanskrit Reconstruction, English Translation, Critical Notes (Including Pāli Parallels, Chinese Versions and Ancient Tibetan Fragments). Delhi: Motilal Banarsidass. Śāstrī, Swāmī Dwārikādās, ed. 1998. Ācāryayaśomitrakṛtasphuṭārthavyākhyopetam Ācāryavasubandhuviracitam svopajñabhāṣyasahitañ ca Abhidharmakośam. Varanasi: Bauddha Bharati. Staël-Holstein, Baron A. von, ed. 1926. The Kāśyapaparivarta: A Mahāyānasūtra of the Ratnakūṭa Class. Shanghai: Commercial Press. Suali, Luigi. 1887. Lokatattvanirṇaya di Haribhadra. Giornale della società asiatica italiana 18: 263– 319. Suzuki, Kōshin, ed. 1994. Sanskrit Fragments and Tibetan Translation of Candrakīrti’s Bodhisattvayo gācāracatuḥśatakaṭīkā. Tokyo: The Sankibo Press. Takakusu, Junjirō, and Kaigyoku Watanabe, eds. 1924–32. Taishō shinshū Daizōkyō. 85 vols. Tokyo: Taishō Issaikyō Kankōkai. Tsong kha pa Blo bzang grags pa. 1985. Mnyam med tsong kha pa chen pos mdzad pa’i byang chub lam rim che ba. Xining, China: Mtsho sngon mi rigs dpe skrun khang. Vaidya, Paraśurama Laksmana, ed. 1958. Lalita-Vistara. Darbhanga: Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning. , ed. 1960a. Aṣṭasāhasrikā Prajñāpāramitā: With Haribhadra’s Commentary Called Āloka. Darbhanga: Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning , ed. 1960b. Bodhicaryāvatāra: With the Commentary Pañjikā of Prajñākaramati. Darbhanga: Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning. , ed. 1961. Samādhirājasūtra. Darbhanga: Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning, , ed. 2003. Mahāyānasūtrasaṃgraha: Part I. 2nd ed. Darbhanga: Mithila Institute of PostGraduate Studies and Research in Sanskrit Learning. Ye, Shaoyong. 2011. Mūlamadhyamakakārikā: Dbu ma rtsa ba’i tshig le’ur byas pa shes rab ces bya ba; Zhonglun song; Fanzanghan hejiao, daodu, yizhu. Shanghai: Zhongxi shuju.