Nattier Glorious Cities 2021

ISSN 1343-8980 Off-print [PDF version] 創価大学 国際仏教学高等研究所 年 報 令和二年度 （第２4号） Annual Report of The International Research Institute for Advanced Buddhology at Soka University for the Academic Year 2020 Volume XXIV 創価大学・国際仏教学高等研究所 東 京・2021・八王子 The International Research Institute for Advanced Buddhology Soka University Tokyo・2021 ERRATUM Nattier, “Glorious Cities and Eminent Clans” (ARIRIAB 2021) p. 139, n. 28: for: The reconstructed Han pronunciation of zia likely reflects the entry in the Digital Dictionary of Buddhism cited in the following note). read: The reconstructed Han pronunciation of zia likely reflects the voicing of the intervocalic sibilant (a feature that is typical of Gāndhārī; cf. the brief discussion in my entry for this term in the Digital Dictionary of Buddhism, cited in the following note). Glorious Cities and Eminent Clans: On Some Proper Names in Zhi Qian’s Mahāparinirvāṇa Sūtra (Part 1)* Jan NATTIER this paper is dedicated to the memory of my dear friend and colleague Stefano Zacchetti (1968-2020), with whom I spent many happy hours discussing Chinese Buddhist translations in Hachioji Abstract: In the third and final volume of his monumental comparative study of the biography of the Buddha, André Bareau examined the passage in which Ānanda pleads with the Buddha not to die in Kuśinagara, but rather in a more illustrious town. Comparing the names of such towns presented in seven different versions of the text, Bareau considered the list of names found in the translation of the Mahāparinirvāṇa Sūtra by Zhi Qian 支謙 (般泥洹經, T6) to be both unusually short and aberrant in content. This paper presents a detailed examination of each of the names given in Zhi Qian’s translation, showing that there was a consistent method governing his renditions (both transcriptions and translations) and that two of the names not recognized by Bareau can now be identified. This places Zhi Qian’s version of the text within the mainstream of parinirvāṇa narratives, while casting new light on Zhi Qian’s translation techniques. Keywords: André Bareau, Zhi Qian 支 謙 , Mahāparinirvāṇa Sūtra, Mahāparinibbāna Sutta, Śrāvastī, Vaiśālī, Bārāṇasī, Buddhist transcriptions, Buddhist translations, Chinese historical phonology Introduction One of the most memorable scenes in canonical accounts of the Buddha’s final days is the moment when Ānanda, realizing that the Buddha is about to pass away in the small town of Kuśinagara (Pāli Kusinārā), entreats him to choose a more illustrious locale for his demise. In the colorful translation of the Pāli version by Maurice Walshe, Ānanda pleads, “May the Blessed Lord not pass away in this miserable little town of wattle-and-daub, right in the jungle in the back of beyond!”1 He then recommends a variety of more suitable locales for the Buddha’s parinirvāṇa, including the great cities of Rājagṛha (P. Rājagaha), Śrāvastī (P. Sāvatthi), and Bārāṇasī (Skt. also Vārāṇasī; P. Bārāṇasī). The Buddha refuses Ānanda’s request, however, saying that Kuśinagara is in fact a quite appropriate place, since in the * I would like to thank Paul Harrison, Antonello Palumbo, and Michael Radich for valuable comments and suggestions on an earlier draft of this paper. I would also like to thank Stephen R. Bokenkamp, Li Jiangnan 黎 江南 and Zhao Luying 趙鹿影 for helpful observations made during a reading group based at Arizona State University on January 25, 2021, where we examined Zhi Qian’s lists of city and clan names (some of which will be dealt with in Part 2 of this paper). Any errors or infelicities that remain, of course, are my own. 1. See Walshe 1987, p. 266. ARIRIAB Vol. XXIV (March 2021): 131–146 2021 IRIAB, Soka University, JAPAN distant past it was the site of the glorious city of Kusāvatī, the royal seat of a world-ruling emperor (Skt. cakravartin, P. cakkavatti). An account of this scene is preserved in a wide range of sources, and in his detailed comparative study of versions of the biography of the Buddha André Bareau analyzed this passage as it appears in seven different texts (Bareau 1971, pp. 72–76). Of these, two are works preserved in Indian languages: • the Pāli Mahāparinibbāna Sutta (in the Dīgha Nikāya, sutta no. 16) • the Sanskrit Mahāparinirvāṇa Sūtra, ed. Ernst Waldschmidt (1950–51) Another two are contained in large Āgama collections translated into Chinese: • the Chang ahan jing 長阿含經 (Dīrgha Āgama; T1, Sūtra no. 2), referred to by Bareau as “Chinese A”; translated in 413 CE by Buddhayaśas 佛陀耶舍 and Zhu Fonian 竺佛念 • the Zhong ahan Jing 中 阿 含 經 (Madhyama Āgama; T26, Sūtra no. 68), translated in 397-398 by Gautama Sanghadeva 瞿曇僧伽提婆 (Bareau does not assign a letter to this source, simply referring to it in his discussion as “Āgama”) The final three are individual Chinese translations of various recensions of the Mahāparinirvāṇa Sūtra (i.e., works translated independently and not as part of an Āgama collection): • the Fo bannihuan jing 佛般泥洹經 (T5, Bareau’s “Chinese B”), an anonymous translation whose language and style points to its production in the 3rd century CE in the southern Wu kingdom;2 • the Bannihuan jing 般泥洹經 (T6, “Chinese C”), translated by Zhi Qian 支謙 in the first half of the 3rd century CE;3 and • the Da banniepan jing 大般涅槃經 (T7, “Chinese D”), translated in the 5th c. CE4 2. On the translator attribution of T5 see Nattier 2008, pp. 127–128. In a study published in 2010 Jungnok Park argued that T5 was one of the earliest translations produced by Zhi Qian, and that T6 is “a work by a successor in his circle rather than by Zhi Qian himself” (Park 2008[2010], p. 363). Park’s argument deserves close examination, but it is based on a number of problematic assumptions, such as the idea that where T5 contains references to filial piety that are absent in T6 “the writer of T6 … remove[d] them” (p. 364). Park’s table of “archaic renderings” and “ later ‘standardized’ ones” (p. 365) also contains a number of incorrect assertions (the rendering of bhikṣu as biqiu 比丘, for example—which is already well attested in the corpus of An Shigao— does not postdate An Xuan and Yan Fotiao’s translation of the term as chujin 除饉 “one who gets rid of hunger” as Park contends). The relationship between these two Chinese versions of the Mahāparinirvāṇa Sūtra is complex, and an in-depth comparative study of the two would be most welcome. One can hazard a guess, however, that such a study would yield at least two results: first, that neither of these two translations is (just) a revision of the other, but on the contrary, that one of them is a retranslation, based on a substantially different Indic-language source; and second, that the translator of the later of the two Chinese versions consulted the earlier one. This is very different from the case of the revision of Lokakṣema’s translation of a version of the text now known on the basis of its Sanskrit title as the Aṣṭasāhasrikā Prajñāpāramitā (T224) by Zhi Qian (T225B), where the revised version follows the content and much of the non-technical wording of the original very closely even while condensing its long lists and revising the rendition of proper names and Buddhist terms (see Nattier T2008[2010]). (On the late date of the introduction of these numerically based titles—8,000 Lines, 25,000 Lines, and so on—for the Prajñāpāramitā texts see Zacchetti 2005, pp. 37–40.) It should be noted, incidentally, that Park’s characterization of the corpus of authentic translations by Zhi Qian (p. 342) is based on the list given in Nattier 2003 (pp. 208–209 and 241–242), which is superseded by the more recent and much more detailed assessments given in Nattier 2008 (pp. 116–148). 3. See Nattier 2008, pp. 126-128. 4. The text is attributed to Faxian 法顯 in the Taishō canon but this attribution has been questioned in recent 132 Of the above sources relating this event discussed by Bareau, all are various recensions of the Mahāparinirvāṇa Sūtra, with the exception of the account found in the Chinese Madhyama Āgama, which occurs at the beginning of a version of the Mahāsudarśana Sūtra (T26[68], Da Shanjian wang jing 大善見王經). This scene also occurs at the beginning of its Pāli counterpart, the Mahāsudassana Sutta (DN 17), where as in the Chinese it recounts the glorious past of the city of Kusinārā when it was ruled by a cakkavatti named Mahāsudassana. Bareau did not discuss the Mahāsudassana Sutta in his analysis of this passage (though he did include the version found in the Chinese Madhyama Āgama), presumably because its list of city names (DN ii.169, lines 11–12) is precisely the same as that given in the Pāli Mahāparinibbāna Sutta, and thus sheds no new light on Indian traditions concerning the names of these locales. One other account of this scene, however, might have been expected to appear in Bareau’s discussion, despite the fact that it occurs not in a sūtra but rather in a vinaya text. In his monumental study of the Sanskrit Mahāparinirvāṇa Sūtra Ernst Waldschmidt included the passage describing this event found in the Chinese and Tibetan translations of the Mūlasarvāstivāda Vinaya (Waldschmidt 1951, p. 305). Bareau might perhaps be excused for not drawing on these two vinaya texts here (though he refers to Waldschmidt’s edition and translation of the texts elsewhere) for, as we shall see, they present exactly the same list of six city names, in the same order, as in the Sanskrit Mahāparinirvāṇa text. Nonetheless, since my concern in this paper is not merely with the content—that is, with the Indian terms underlying the Chinese and Tibetan translations—but also with the Chinese terminology itself, I will include this text in the discussion below. Great Cities in the Mahāparinirvāṇa Sūtra: An Overview In his discussion of this passage Bareau pointed out that these accounts differ not only in the names of the towns given but also in their number, with six names given in some versions (Pāli, Sanskrit, Chinese B, and the Madhyama Āgama), seven in another (Chinese A), and eight in yet another (Chinese D). Zhi Qian’s version alone (Bareau’s Chinese C), according to his reading, contains only three names (Bareau 1971, p. 73). He also emphasized the fact that the order of the names in the various lists—aside from the fact that the first one is always either Campā, or Śrāvastī, or Rājagṛha—is chaotic in the extreme (in Bareau’s words, it appears to be “le fruit d’un hasard capricieux,” p. 74). In analyzing the various accounts of Ānanda’s plea Bareau did not provide complete lists of the city names found in each of his seven sources, but simply summarized their content as follows: Only Rājagṛha is cited in all of them. Campā (except Chinese C), Śrāvastī (except Chinese C), Bārāṇasī (except Chinese C), and Vaiśālī (except Pāli) are mentioned just six times; Sāketa scholarship. See Radich 2018 and more recently Radich 2019, which offers a rich stylistic comparison (performed using the sophisticated search capabilities made possible by TACL, “Text Analysis for Corpus Linguistics”) of T7 with the other works traditionally ascribed to Faxian. Radich demonstrates convincingly that T7 stands apart from the rest of the Faxian corpus, while bearing a closer resemblance to certain texts ascribed to Guṇabhadra 求 那 跋 陀 羅 , as well as certain additional texts outside that corpus (for a summary of his findings see pp. 266–268). For earlier scholarship on the attribution of T7 see the sources summarized on CBC@ at https://dazangthings.nz/cbc/text/819/. 133 appears in four lists (Pāli, Sanskrit, Chinese A and B). Kauśāmbī is only named twice (Pāli, Chinese D) as is Kapilavastu (Chinese A, [Madhyama] Āgama). Ayodhyā and Taksaśilā [sic]5 are cited only by Chinese D, and the Mallas by Chinese C.6 Five decades have now passed since the publication of the final volume of Bareau’s thoroughgoing study, and significant progress has been made in our understanding of both the translation terminology and the translation techniques used by Zhi Qian and his predecessors. We also now have access to a dizzying array of technological resources, including two fully digitized versions of the Chinese canon (CBETA and SAT) and a variety of search utilities, tools that scholars of Bareau’s generation could not even have imagined. Using all of these resources it is now possible to cast new light both on the content of Zhi Qian’s source-text and on the derivation of several of his translation and transcription terms. The Great Cities in Detail: Lists Found in Bareau’s Sources For purposes of the present study it is necessary to go beyond Bareau’s general description to give the full lists of the names found in this passage in each of his sources, as these will provide an important point of reference for the analysis presented below. In each case the Pāli or Sanskrit name will be given in its citation form (rather than with any case endings found in the actual texts), and Sanskrit equivalents will be given for each of the Chinese place names for convenience of reference (without implying that all of these works were actually translated from that language, which was certainly not the case).7 I have also eliminated the epithets that follow some but not all of the names (國, 大國, or 城) in order to focus on the transcriptions or translations of the names themselves. Excluding for the moment the version produced by Zhi Qian (Chinese C)—which abounds in problems to be discussed below—but tabulating the names given in all of the other sources adduced by Bareau, we obtain the following lists: Pāli (PTS ed., DN ii.146, l. 14–15):8 Campā, Rājagaha, Sāvatthi, Sāketa, Kosambī, Bārāṇasī Sanskrit (Waldschmidt ed., p. 304): Śrāvastī, Sāketa, Campā, Bārāṇasī, Vaiśālī, Rājagṛha Chinese Madhyama Āgama (T26, 1.515b17–19): 瞻 波 (Campā), 舍 衛 (Śrāvastī), 鞞 舍 離 (Vaiśālī), 王舍 (Rājagṛha), 波羅奈 (Bārāṇasī), 加維羅衛 (Kapilavastu) Chinese A = Dīrgha Āgama (T1, 1.21b10–12): 瞻婆 (Campā), 毗舍離 (Vaiśālī), 王舍(Rājagṛha), 婆祇 (Sāketa), 舍衛 (Śrāvastī), 迦維羅衛 (Kapilavastu), 波羅奈 (Bārāṇasī) 5. The name is typically written Takṣaśilā in Sanskrit and Takkasilā in Pāli (where, however, it is very rare in the canon, occurring only seven times in the Vinaya—six of which are found in a single story—as well as twice in the Jātaka section and twice in the Mahāniddesa, with no occurrences at all in the Sutta section, a pattern which seems certain to be an indication of the lateness of the appearance of this locale on the horizon of Buddhist geographical awareness). On other forms of the name see Marshall 1975 (reprint of the 1951 edition), vol. 1, p. 1, n. 1. 6. Bareau, op. cit., p. 73 (my translation). 7. All Sanskrit equivalents given are my own, as many of these Chinese names are not yet registered in existing dictionaries. 8. As mentioned above. the list given at the beginning of the following text, the Mahāsudassana Sutta (DN ii.169, lines 11–12), is identical. 134 Chinese B (T5, 1.169c14–15): 舍衛 (Śrāvastī), 沙枝 (Sāketa), 栴波 (Campā), 王舍 (Rājagṛha), 波羅奈 (Bārāṇasī), 維耶梨 (Vaiśālī) Chinese D (T7, 1.200c25–27): 王舍 (Rājagṛha), 毘 耶 離 (Vaiśālī), 舍 衛 國 (Śrāvastī), 婆 羅奈 (Bārāṇasī), 阿踰闍 (Ayodhyā), 瞻波 (Campā), 俱睒彌 (Kauśāmbī), 德叉尸羅 (Takṣaśilā) One additional source included in Waldschmidt’s edition (and, as noted above, not dealt with by Bareau in this part of his study) should be added to these lists for completeness. In providing parallels to the the versions of this passage found in the Sanskrit and Pāli versions of the Mahāparinirvāṇa Sūtra, Waldschmidt also collated the list of city names given in the discussion of this event in the Mūlasarvāstivāda Vinaya, providing an edition (but not a translation) of the Tibetan, and a translation (but not an edition) of the Chinese: Mūlasarvāstivāda Vinaya (Waldschmidt ed., 1951, p. 305):9 室羅伐 / Mnyan yod (Śrāvastī), 娑 (var. 婆 ) 雞 多 / Gnas bcas (Sāketa), 占 波 / Tsam pa (Campā), 婆 羅 痆 斯 / Bā rā ṇa sī (Bārāṇasī), 廣嚴 / Yangs pa can (Vaiśālī), 王舍 / Rgyal po’i khab (Rājagṛha) As mentioned above, in his analysis of these lists Bareau emphasized the many differences among them: first, in the total number of names they contain (ranging, in his view, from three to eight); second, in some of the names themselves; and third in the wild fluctuations in their sequence. There are, however, what we might call certain “affinity groups” that can be discerned in the above lists, the first consisting of the Dīrgha Āgama (Bareau’s Chinese A) and the Madhyama Āgama, which have virtually the same list of names (the sole exception being the addition of Sāketa in DĀ), albeit in a different order. Even closer is the match between Chinese B (T5) and the Sanskrit, where the six names given are exactly the same, with just one minor difference in sequence (Rājagṛha occurring sixth and last in the Sanskrit but fourth in T5). An additional member of this second group is the Mūlasarvāstivāda Vinaya, which in both its Chinese and Tibetan translations exhibits a list identical with the Sanskrit Mahāparinirvāṇa Sūtra in both content and sequence.10 The fact that the anonymous Fo bannihuan jing 佛般泥洹經 (T5, Chinese B) bears such an unexpectedly strong resemblance to the texts examined in Waldschmidt’s study suggests that a thorough comparative analysis of other portions of this Chinese translation and the Sanskrit Mahāparinirvāṇa Sūtra might yield illuminating results.11 Great Cities in Zhi Qian’s Version of the Text With the above information in hand we may now turn to Zhi Qian’s version of the sūtra (T6, the Bannihuan jing 般 泥 洹 經 , listed by Bareau as “Chinese C”), which presents a host of 9. I have supplied the Chinese names from Waldschmidt’s source-text, viz., the Mūlasarvāstivāda Vinayakṣudrakavastu, Genben shuo yiqie you bu binaye za shi 根 本 說 一 切 有 部 毘 奈 耶 雜 事 translated by Yijing 義淨 in the early 8th century (T1451, 24.392c26–27). I have also converted all of the Tibetan names given in Waldschmidt’s edition to the Wylie system and have given them in their standard spellings (for variants see Waldschmidt’s notes on p. 305). 10. As mentioned above, it may be for this reason that Bareau did not discuss these vinaya texts here. 11. A general sense of the content of T5 can be obtained from the lightly annotated Italian translation produced more than a century ago by Carlo Puini (Puini 1919). I would like to thank Peter Skilling for calling my attention to this publication and for providing me with a photocopy of Puini’s book. 135 intriguing problems and thus must be dealt with separately. Bareau refers to it as “aberrant,” stating first that its list of names is reduced to just three elements, and going on to say that while it fails to mention either Śrāvastī or Bārāṇasī, it is the sole version to include the Mallas in the list, a fact that is quite surprising given that they were the inhabitants of Kuśinagara, which is precisely the town that Ānanda is criticizing as an unsuitable venue for the Buddha’s demise (Bareau 1971, p. 73). As we shall see, there are other problems with the idea that this list includes the Mallas in addition to those mentioned by Bareau. Because of the many difficulties in interpreting Zhi Qian’s version of this short passage it will be useful to examine his list of great cities in the full context in which it occurs. Zhi Qian’s translation reads as follows: 賢者阿難問佛言： “近此左右有聞物大國、王舍大國、滿羅大國、維耶大國。佛不於彼 般泥洹？何止(<–正)於此褊陋小城？” （T6, 1.185b13–15) * reading with n. 5 to the Taishō edition (正＝止＜三＞) Leaving all of the problematic (or even slightly problematic) names in Chinese for the moment, we may provisionally translate this passage as follows: The venerable Ānanda asked the Buddha, “Nearby, close to here, are the great city of Wenwu (聞 物), the great city of the Royal Residence (王舍, Rājagṛha), the great city of Manluo (滿羅), and the great city of Weixie (維耶).12 Won’t the Buddha enter parinirvāṇa in one of these? Why stay in this cramped and crude little town?”13 Judging by the clear parallelism in this passage, what we have here is surely a list of the names of four great cities (大國)14 rather than just three as Bareau supposed. What seems to have happened is that, not recognizing wenwu 聞物 (which could be understood as meaning “hear things” or “things heard,” and thus perhaps interpreted as “illustrious” or “much talked about”) as the name of a particular place, Bareau appears to have taken this word as an epithet of the list as a whole (though he does not say so explicitly). If this is the case, it would explain why he considered “Chinese C” to contain a total of only three place-names, viz., 12. An explanation of the rationale for using the pinyin reading xié (rather than the more common yé or yē) for the character 耶 will be given below (p. 139, n. 28). 13. Or perhaps better, “this narrow (褊) and out-of-the-way (陋) little town” (for lòu 陋 in this sense see Kroll 281b). The Pāli has three epithets here: kuḍḍa “wattle-and-daub” (var. khuddaka “small,” but Waldschmidt’s Sanskrit text supports the lectio difficilior of kuḍḍa), ujjaṅgala “barren, deserted,” and sākhā “branch” (in the sense of “outside the center, peripheral,” or in very colloquial English, “in the middle of nowhere”). 14. It is important to note, incidentally, that 國 should be translated as “city” and not “country” here. There is a tendency in Buddhist Studies to reflexively translate 國 as “country” (in the sense of “nation” or “state”), but this often distorts the meaning of what is meant by the word in Buddhist translations. Indian texts distinguish between janapada “country, realm” (i.e., an area inhabited by a particular group of people) and nagara “city, town” (with other terms, such as gāma, used for smaller centers of habitation such as villages). The capital city (or the principal town) of a janapada was generally referred to as a mahānagara “great city.” A standard list of sixteen great states (mahājanapadas)—though one that certainly postdates the time of the Buddha—can be found in Pāli literature, viz., Aṅga, Magadha, Kāsī, Kosala, Vajji, Mallā, Cetī, Vamsā, Kuru, Pañcāla, Macchā, Sūrasena, Assaka, Avantī, Gandhāra, and Kamboja (see Malalasekera, DPPN under these individual names, and cf. PED s.v. janapada). The names we find in the texts with which we are concerned here, however—in all seven of the texts studied by Bareau, including Zhi Qian’s translation—are the names of great cities (mahānagara) rather than countries or states (janapada). The word 國 should thus be translated here not as “country” but as “city,” which indeed is one of its meanings in Chinese of this period (especially in the sense of “capital city of a state”; see Kroll 149a). 136 Wangshe 王舍 (an unproblematic rendition of Rājagṛha), Manluo 滿羅 (which he took to be a transcription of the clan-name Malla, though as noted above he considered its presence quite unexpected here, an issue to which we will return below), and Weixie 維耶 (which he seems to have accepted without comment as a rendition of a form of Vaiśālī).15 Each of these names offers much that is of interest, and it will be worthwhile to examine them individually here. Rather than doing so in the order in which they appear in the text, however, I will examine them in increasing order of difficulty—beginning with the most transparent case and proceeding on to more complicated ones—as it is the relatively straightforward terms in this list that can help us to understand how the more elusive ones came about. Wángshè 王舍 “Rājagṛha” Bareau had no difficulty whatsoever in understanding this name, as it has long been recognized as a translation of the city name Rājagṛha. This term, which I translated above as “Royal Residence,” is a simple calque (rāja “king” = 王, gṛha “house” = 舍), and it appears already in the biography of the Buddha translated by Kang Mengxiang in the latter part of the Han dynasty (T196, 4.156a7).16 It is amply documented in standard reference works, and thus posed no problems in his interpretation of the contents of this list. One point should be made, however, before passing on to the next of the names of great cities to be discussed here: this translation conveniently conforms to the standard Chinese format for place-names, as it consists of just two characters (cf. Luoyang 洛陽, Chang’an 長 安 , and the vast majority of pre-modern Chinese geographical names). The significance of this model will become evident as we examine the other city names below. Wénwù 聞物 “Śrāvastī” The term Wenwu, by contrast, seems to have eluded Bareau’s grasp, for he did not include it in his list of place-names but rather, as noted above, apparently took it to be an epithet of the list as a whole. It is understandable that Bareau would have had difficulty in recognizing Wenwu as a rendition of Śrāvastī, since this form occurs in only a handful of translated texts. Already in its earliest occurrence, however, it is clear that it is the name of a place, for it appears in the opening nidāna of the oldest Chinese translation of the Ugraparipṛcchā, the Fa jing jing 法鏡經 (T322) produced by An Xuan 安玄 and Yan Fotiao 嚴佛調 in the late 2nd century, as the locale where the sūtra was preached. Fortunately we have several parallel versions of the sūtra available in Chinese and Tibetan, making it a straightforward matter to compare the wording they contain.17 The two Chinese parallels both have Shewei 舍 衛 ,18 which is well documented as a transcription of a form of Śrāvastī.19 The Tibetan, for its part, has the standard translation of Śrāvastī as Mnyan yod.20 15. For a discussion of this unusual rendition see below, pp. 139–141. The handful of occurrences in two anthologies newly attributed to the community of An Shigao are all in portions of these texts that Paul Harrison has shown are later intrusions, and thus they should not be used to argue for an earlier appearance of the word (see Harrison 1997 and 2002). 17. A synoptic table providing page and line numbers for parallel passages in the Chinese and Tibetan editions of the text is given in Nattier 2003, Appendix 1. 18. See Schuessler 2009, 舍 śaB/śaC (1-48), 衛 was (28-5). 19. See for example Akanuma 1931, pp. 607a and 608a, s.v. Sāvatthi. . 20. As registered in the Mahāvyutpatti, no. 4110. 16. 137 With this information in hand we are then in a position to locate the entry documenting Wenwu 聞物 as the equivalent of Śrāvastī (or rather, Pāli Sāvatthi) in Akanuma’s dictionary of Indian proper names, citing the very passage with which we are concerned.21 (The difficulty for sinologists in using Akanuma’s dictionary, of course, is that it is based on Indian rather than Chinese proper names, and thus one must already know its Pāli or Sanskrit equivalent before one can look up a Chinese term.) Some widely used dictionaries of Buddhist Chinese, including those published much more recently—Hirakawa’s Buddhist Chinese-Sanskrit Dictionary, for example, or the Fo Guang Dictionary of Buddhism—do not register the term at all.22 At first this would seem to be true of Nakamura’s venerable Bukkyōgo daijiten 佛 教 語 大 辞 典 as well, but though the term does not appear under the perhaps expected pronunciation of *bunbutsu (ぶんぶつ), a very brief entry—not citing any translated texts at all, but referring only to the Yiqiejing yin yi 一 切 經 音 譯 (T2128), a medieval Chinese lexicographic work—is tucked away under the reading monmotsu (もんも つ) instead.23 Neither Akanuma nor Nakamura discussed the derivation of the term, but in light of the Tibetan parallel cited above it quickly becomes clear that what we have here is an etymological translation of the name. In Tibetan, Śrāvastī is interpreted as consisting of śrāv- (from the verbal root śru “hear,” Tib. mnyan) plus the third-person singular verb asti “exists” (Tib. yod).24 An Xuan and Yan Fotiao, on the other hand—taking a similar approach but working some six centuries earlier—divided the word into śrāv- (again understood as “hear,” Ch. wen 聞) + vastī, apparently interpreting the latter as if it were a form of vastu “object, thing” (Ch. wu 物). The double use of the letter v in this translation—first taken as part of the component śrāv-, then used again as part of –vastī (interpreted as vastu)—places Wenwu 聞物 within a specific sub-category of etymological translations: what we might call “overlapping translations,” in which a certain part of the source-term (ranging from a single consonant or vowel to a number of syllables) is used twice. Other examples of this phenomenon are easily adduced, such as fanzhi 梵志 for brāhmaṇa (where the word is interpreted as consisting of the name of the god Brahmā + manas “mind”) and, perhaps most notably of all, du wuji 度無 極 for pāramitā, treated as if it consisted of the components pāram + ita, interpreted as “gone (ita) to the other shore (pāra)” and translated as 度 , with the latter part of the word then interpreted again as amita “limitless” and translated as 無 極 .25 This phenomenon is not 21. See Akanuma 1931, p. 607a, s.v. Sāvatthi. See Hirakawa’s Bukkyō kanbon daijiten 佛教漢梵大辭典 (Tokyo 1997), not registered on p. 959 (no. 2972, s.v. 聞 ). and the online version of the Foguang da cidian 佛 光 大 辭 典 (version of 2011) at https://www.fgs.org.tw/fgs_book/fgs_drser.aspx (accessed on 30 January 2021), no results returned for 聞物. 23. See Nakamura p. 1372d (s.v. 聞物國 もんもつこく). 24. Cf. the entry in the Madhyavyutpatti (Ishikawa ed., 1990, #362, pp. 113–114): śrāvasti (var. śrabasti, śrābastī) zhes bya ba drang srong mnyan pa (var. po) zhes bya ba (var: om. ba) dang | drang srong yod pa zhes bya gnyis kyis (var. kyi) grong khyer 'di dang po btsugs pa yin pas na (var: om. na) mnyan yod ces bya |. I am grateful to Michael Radich for calling my attention to this reference. A different explanation is given in the commentary to the Arthaviniścaya Sūtra, where the city is said to be named for just one person, the ṛṣi Śravasta, who had a hermitage there (see Samtani 2002, p. 49 and for the Sanskrit text Samtani 1971, p. 77; I would like to thank Peter Skilling for providing this reference). 25. See Zürcher 2007, p. 336, n. 140. Zürcher referred to the term du wuji as a “double translation,” but it might be better to reserve this category for translation terms in which a single component of the source-term is translated in two different ways, e.g., the translation of pratyekabuddha as yuan yi jue 緣一覺 “causally and singly awakened,” where the first part of the word is interpreted first as pratyaya, translated as 緣 “cause,” and 22. 138 limited to translated terms, but is found in transcriptions as well. We will encounter an example of such an “overlapping transcription” in one of the remaining city names to be discussed below. The interpretation of Śrāvastī as if it were derived from śrāv- + vastu is peculiar to say the least, and it goes without saying that this does not represent the actual etymological origin of the word. But this is far from an isolated example in Buddhist literature. One need only consult the discussion of the meanings (decidedly plural) of the word arahant in the classic Pāli compendium by the fifth-century scholar-monk Buddhaghosa, the Visuddhimagga, to find a variety of interpretations of this sort.26 These include, we might note, the parsing of the word as consisting of ari “enemy” + hata, understood as derived from the verbal root han “strike, vanquish, kill.” This historically impossible but pedagogically quite useful interpretation is the same as the one that yielded the standard Tibetan translation of the word as dgra bcom pa “one who vanquishes the enemy.”27 Finally, we should note that the translation of Śrāvastī as Wenwu 聞物, like Wangshe 王舍 for Rājagṛha, again conforms to the standard two-character format of Chinese place names. The significance of this feature will become evident as we move to a discussion of the two transcribed city names on Zhi Qian’s list. Wéixié 維耶 “Vaiśālī” We now come to a term that is not a translation but a transcription, and only a partial transcription at that. The use of Weixie (usually written Weiye)28 alone for Vaiśālī is quite rare, and the earliest occurrences of this rendition are in the text we are examining here.29 The same text, however, includes a number of occurrences of the three-character transcription Weixieli 維耶離. The two are obviously related, and it seems likely that the shorter form is an abbreviation of the longer one. It is normal for early Chinese transcriptions of Buddhist names and terms to drop a final short -a (e.g., chan 禪 for a Middle Indic form of dhyāna) or even a final long -ā (e.g., boluomi 波羅蜜 for pāramitā).30 Other final vowels are often omitted as well, as are the final then again as pratyeka, translated as 一 “single” (on this translation see Boucher 1998, pp. 490–491 and Karashima 1998, p. 566). 26. See Warren and Kosambi 1950, Chapter VII, §3ff. (p. 162ff.) and cf. Ñāṇamoli 1999, p. 192ff. Cf. also the remarks on this multiplicity of interpretations in Nattier 2003, pp. 218-219. 27. See Mahāvyutpatti no. 5138. 28. See Schuessler’s Later Han (LH) reconstruction at 1-47, where Mandarin 耶 yé corresponds to LH ja, 耶 xú to LH ziɑ, and 耶 xié to LH zia; for the pinyin reading of 耶 (which alternates with 邪) as xié cf. HD 8.654b. The reconstructed Han pronunciation of zia likely reflects the entry in the Digital Dictionary of Buddhism cited in the following note). 29. Aside from the use of Weixie 維耶 as a transcription of the name of Viśākhā, the Buddha’s paradigmatic female lay disciple (in Zhi Qian’s T87, Zhai jing 齋經; cf. my entry in the Digital Dictionary of Buddhism at http://www.buddhism-dict.net/cgi-bin/xpr-ddb.pl?q=%E7%B6%AD%E8%80%B6), I have only been able to locate a handful of other cases where 維 耶 is used alone to transcribe Vaiśālī. There is one occurrence in Dharmarakṣa’s Xianjie jing 賢劫經 (T425, 14.11a17), though some editions have the additional (and expected) character 離 (see n. 6 to the Taishō apparatus). Another (with the variant character 邪 for 耶) is registered in the Yiqie jing yin yi 一 切 經 音 義 (T2128, 54.518b10; cf. Akanuma 759a) citing fascicle 12 of the apocryphal Guanding jing 灌頂經, but the received text has the longer form 維耶離 (T1331, 21.532b11). The occurrences in T44 (1.830a3) and T2045 (50.175a16) are probably metri causa, while the occurrence in the Shijia pu 釋迦譜 (T2040, 50.75b11) is in a citation from Zhi Qian’s T6, to be discussed in Part 2 of this study. 30. The character 蜜 contains a final -t (LH mit, 29-41), so it is only the vowel -ā (and not the entire syllable -tā) that has been dropped. 139 syllables -ya and -ka (and sometimes -kha). But the omission of final -lī would be unusual. Indeed, all of the transcriptions of Vaiśālī attested during Zhi Qian’s time and before—aside from the two-character form we are concerned with here—represent all three syllables of the Indian term, each with a separate character for final -li: • Wéixiélí 維耶離 (LH wi zia liai/liaiC; 28-11, 1-47, 18-11) • Huīshèlì 墮舍利 (LH hyai/hwai śaB/śaC liC; 19-9, 1-48, 26-24) • Wéishèlì 惟舍利 (LH wi śaB/śaC liC; 28-11, 1-48, 26-24) • Wéixiélí 維耶梨 (LH wi zia li; 28-11, 1-47, 26-24) Of these the first appears for the first time in the biography of the Buddha translated by Kang Mengxiang, the Zhong benqi jing 中 本 起 經 (T196, 4.161b23) and subsequently became standard in Zhi Qian’s own work,31 while the second—sometimes written with the variant character suí 隨 (LH zyai, 19-9) for huī 墮 —is attested in the corpus of their predecessor Lokakṣema.32 The third is found only in one text, an archaic translation of the story of the parinirvāṇa of Mahāprajāpatī by an unknown but surely Han-period translator (T144, Da’aidao bannihuan jing 大愛道般泥洹經).33 More specifically, it occurs only in the latter part of the text (3.868a19–20ff.), while in earlier passages the standard Lokakṣema-school transcription 墮舍利 appears.34 The fourth and last transcription given above appears in just a small handful of translations (including T5, Bareau’s “Chinese B”), at least two of which bear the hallmarks of a third-century Wu kingdom style.35 Given that both a longer form (維耶離) and this shorter one (維耶) occur in the text with which we are concerned (the former occurring seven times, the latter with only four occurrences), the use of the short form found in this passage clearly requires an explanation. And here context provides the key, for as we have seen the two translated names discussed above both consist of just two characters, as does the final transcribed name to be discussed below. In this passage, therefore, Zhi Qian may have been attempting to provide symmetrical renderings by using two-character forms of both translated and transcribed names. The fact that each name is paired with the epithet “great city” (大 國 ), thus creating four-character units, may have been an additional factor as well. In Zhi Qian’s work Weixieli 維耶離 is found in T474 (14.519a9 and passim), T557 (14.909c6), and T1011 (19.680b6), as well as seven times in the text with which we are concerned here (T6, 1.178c23 and passim). It occurs sporadically in later translations as well, including several texts produced by Dharmarakṣa. 32. On this term see the entry in Karashima 2010, p. 156 s.v. duò shè lì, but note that the character 墮 should be written in pinyin as huī (LH hyai, hwai, 19-9) rather than the commonly used duò (LH duɑB < duɑiB, 19-16). 33. The attribution of the text to Bo Fazu 白法祖 in the Taishō canon is not reliable. For a synopsis of the work of several Japanese scholars on this topic see the entry for T144 on Michael Radich’s CBC@ site at https://dazangthings.nz/cbc/text/2235/. 34. The same is true of several other proper names in this text (including those of Śāriputra, Maudgalyāyana, and an otherwise unknown gṛhapati named *Yaśodha), which likewise shift from one form at the beginning the text to another in the latter part. It is intriguing that the forms found in the early parts of the text are those typical of the works of Lokakṣema, while those used in later sections are unique to this text. This may indicate that the first part of the text (but not the remaining portion) has been revised by members of Lokakṣema’s school. 35. In addition to T5 (Fo bannihuan jing 佛般泥洹經), this transcription is found in T145 (Fomu bannihuan jing 佛母般泥洹經), which shares some idiosyncratic vocabulary with T5 and seems certain to be related to it in some way; in T805 (Zhantan shu jing 栴檀樹經), an anonymous scripture that has some Wu kingdom features and should be investigated further; and T553 (Nainū qiyu yinyuan jing 㮈女祇域因縁經), a text attributed to An Shigao but certainly not his work (see Nattier 2008, p. 15 n. 6, and cf. now the entry for this text on CBC@ at https://dazangthings.nz/cbc/text/755/). 31. 140 Be that as it may, the fact that Weixie 維 耶 is intended as a rendition of the city name Vaiśālī in this passage is beyond dispute, and Bareau was certainly correct in interpreting the term in this way.36 As we shall see, the fact that it is abbreviated in such an atypical fashion will have great significance for interpreting the remaining—and most enigmatic—city name on Zhi Qian’s list. Mǎnluó 滿羅 [sic] “Bārāṇasī” Most elusive of all the names on Zhi Qian’s list is Manluo, a term that seems certain to be a transcription but whose referent, at first glance, appears completely opaque. The reconstructed pronunciation manla (LH mɑnB lɑ<lɑi, Schuessler 24-57, 18-10) does seem to lend itself to Bareau’s interpretation of the term as representing “Malla,” though as noted above he thought it strange that the Mallas should be included on this list, as they were the inhabitants of Kuśinagara, which is precisely the locale that Ānanda is criticizing here. But there are two other (and arguably more significant) obstacles to interpreting Manluo as “Malla” here. First, the other three terms in Zhi Qian’s list (as indeed, in all of the lists found in the other versions of the text given above) are the names of cities, not of countries or particular peoples or clans. Most telling of all, however, is the fact that the Mallas are mentioned at a number of points elsewhere in the text, and there the name is not transcribed as Manluo but rather translated as Huashi 華氏 “Flower Clan.”37 But if Manluo is not a transcription of “Malla,” then what name could it represent? Here we will need to employ a multi-pronged approach, including all of the following sources of information: (1) considering the form of the other names in the passage where it occurs; (2) examining the names given in all of the available parallels to this passage to identify possible candidates for its antecedent; and (3) having identified one or more candidates, locating the translations or transcriptions used in other texts produced by Zhi Qian to represent the relevant name(s). Once these steps have been carried out, it will quickly become clear that in the present case a fourth procedure will be required as well: (4) to consider the possibility that Manluo was not in fact Zhi Qian’s original rendition, and to attempt to identify the underlying term for which the received reading is a scribal error. Beginning with the immediate context in Zhi Qian’s own work, we have seen that all of the city names in this passage consist of just two characters. One of these, Weixie for Vaiśālī, was rather unexpected. But if Weixie can represent Vaiśā[lī], might Manluo might be a transcription of just the first two syllables of a longer name as well? Turning to the parallels found in other versions of this scripture, aside from Rājagṛha, Śrāvastī, and Vaiśālī (which are already accounted for in Zhi Qian’s translation by the three terms discussed above), the two city names that occur with the greatest frequency, as noted by Bareau in the summary cited above, are Campā and Bārāṇasī. Clearly Campā is not a good candidate to be the source of Manluo, nor do any of the other names that appear on the above lists appear to offer good possibilities. But Manluo does at least contain the character luo 羅, 36. Indeed this equivalence was recognized almost a century ago by Akanuma Chizen, who registers this form—citing the very passage with which we are concerned—in his dictionary (see Akanuma 1931, pp. 757b and 759a, s.v. Vesāli). 37. This term will be discussed in Part 2 of this study. For a brief discussion of its derivation—Zhi Qian has interpreted the word “Malla” as if it were a form of mālā “flower garland”—see my entry for 華氏 in the DDB (http://www.buddhism-dict.net/cgi-bin/xpr-ddb.pl?q=%E8%8F%AF%E6%B0%8F). 141 a standard transcription of the Indic syllable ra or rā. Thus—despite the strangeness of the first character—we should consider the possibility that Manluo might somehow have originated as a transcription of Bārā[ṇasī]. To evaluate this seemingly far-fetched possibility we must now move to the third step, to try to identify the ways in which Bārāṇasī is represented elsewhere in Zhi Qian’s work, if indeed it occurs there at all. This is not at all a straightforward affair, since much of Zhi Qian’s vocabulary is not yet documented in existing dictionaries (a situation that was, indeed, one of the major factors that stimulated my initial interest in studying his work). Two occurrences of Bārāṇasī can quickly be located in his corpus, however: first the widely used form Boluonai 波羅奈 , which appears to have been introduced by Kang Mengxiang in his biography of the Buddha produced toward the end of the Han period (T196, 4.147c29 and passim; in Zhi Qian’s work see T556, 14.908a20), and second a rare but easily recognizable variant, the four-character form Boluonaiyi 波羅奈夷, which occurs once in Zhi Qian’s own biography of the Buddha (T185, 3.483a9) and was subsequently borrowed in Dharmarakṣa’s version, which draws heavily on Zhi Qian’s earlier work (T186, 3.532b2–3). To find other, as yet undocumented renditions of Bārāṇasī in Zhi Qian’s corpus is a much more difficult task. But one place to begin is by noting that this city name is often followed by the word guo 國 (or less commonly, cheng 城), and that in many (but not all) cases it is transcribed in a form ending in the character nai 奈. One approach, then, would be to search in Zhi Qian’s translations for the combinations -奈 國 and -奈 城 . The latter combination elicits no results, but a search for 奈 國 in Zhi Qian’s corpus yields a vital clue: the term Pulinnai guo 蒱 隣 奈 國 occurs as the name of a city in Zhi Qian’s Bo jing chao 孛 經 抄 (T790, 17.729c2), where it is used at the beginning of a jātaka tale introduced to explain the events that have taken place earlier in the text.38 Like so much of Zhi Qian’s idiosyncratic vocabulary, the term Pulinnai 蒱 隣 奈 is not registered in existing dictionaries, but the context in which it occurs—since jātaka tales often begin, as is well known, with a reference to Bārāṇasī—makes it plausible that this is a transcription of that name. And when we consult the reconstructed Later Han pronunciation of the word (LH bɑ lin nɑs, Schuessler 1-67, 32-26, 21-27), it begins to appear quite likely that what we have here is a transcription of a (presumably Middle Indic) form of Bārāṇasī. The first and third characters, pu 蒱 (LH bɑ) and nai 奈 (LH nɑs), with the final vowel (-ī) dropped as is so often the case, are completely unproblematic. But the second character, lin (隣 in SAT, appearing as 鄰 in CBETA), might seem unexpected as a transcription of an Indic -rā- or -rāṇ-. In fact, however, it is well attested as corresponding to Middle Indic -raṇin the word dhāraṇī (tuolinni 陀隣尼) in translations produced by Lokakṣema and others.39 Thus there is no obstacle to accepting lin 隣 as a representation of the syllable -rāṇ- in Zhi Qian’s transcription of Bārāṇasī. It is also here that we meet with an example of an “overlapping transcription,” for in this case the letter ṇ is transcribed twice, both at the end of the syllable represented by the second character (lin 隣 for rāṇ) and at the beginning of the third (nai 奈 [LH nas] for ṇas). The 38. The text as a whole has no known parallel, but the first portion relates the well-known story of the murder of a female renunciant named Sundarī, instigated by members of a non-Buddhist religious group and blamed on the Buddha and his followers (see Malalasekera’s Dictionary of Pali Proper Names, vol. 2, pp. 1216-1217). 39. For references to the occurrence of the term in Lokakṣema’s Daoxing jing (T224) see Karashima 2010, p.508, s.v. tuó lín ní. 142 same is true of the analogous term tuolinni 陀隣尼, used to transcribe dhāraṇī (with lin 隣 for raṇ followed by ni 尼 for ṇī) mentioned just above, which first appeared in Lokakṣema’s work. In sum, in Pulinnai 蒱隣奈 we have what is clearly a transcription of Bārāṇasī, produced by parsing the word (in very non-Indic fashion) into the syllables bā + rāṇ + ṇas and omitting the final vowel -ī. And with this additional—and quite different—transcription in hand,40 we should now make one further observation concerning the text in which it occurs. The critical apparatus to the Taishō edition of Zhi Qian’s Bo jing chao 孛 經 抄 (T790) contains an enormous number of variant readings, a situation that is highly unusual and which brings to mind the case of Lokakṣema’s Banzhou sanmei jing 般 舟 三 昧 經 (T418), which likewise displays a tremendous number of variations. As Paul Harrison (building on earlier work by Sakurabe Hajime) has pointed out, this text has been transmitted in two different recensions: one containing the original (unrevised) portion of the first part, and the other containing the revised recension throughout.41 Could it be that here, as in the case of the Banzhou sanmei jing, the manuscript tradition has preserved both revised and unrevised versions of the text? In this connection it should be noted that in addition to the Bo jing chao attributed to Zhi Qian, Sengyou credited Lokakṣema with a translation entitled Bo ben jing 孛 本 經 (“The Original Bo Scripture”? or perhaps better, if 本 is being used here in the sense of 本起, “The Bo Former-Events Scripture”?).42 It is possible, therefore, that the text as we have it is actually a revision by Zhi Qian of Lokakṣema’s work; it is also possible that the large number of variant readings contain traces of that earlier version.43 To work through this problem in all its details would take us far afield from the task at hand, but what is important for our purposes is the fact that not one but two variants are given in the critical apparatus to the Taishō edition for the initial character pu 蒱 of Pulinnai 蒱隣 奈: (1) the homophonous character pu 蒲, whose pronunciation is the same as that of 蒱 both in pinyin (pú) and in its reconstructed Later Han pronunciation (LH bɑ); and (2) man 滿, the very character that we find in the name Manluo that appears in Zhi Qian’s list of great cities. In light of these variant readings it is now quite clear what has happened: the character man 滿 in Manluo is the result of a copying error—that is, a misreading by a scribe who had no idea what the pronunciation of the underlying Indic-language term might have been, and substituted a visually similar but phonologically completely different word. Zhi Qian’s original translation, we can safely conclude, contained Puluo 蒲羅 (or far less likely, Puluo 蒱羅) and not Manluo 滿羅, intended as an abbreviated transcription of Bārā[ṇasī]. As it turns out, the scribal emendation of pu 蒲 to man 滿 seen here is far from an isolated case. In comments on an earlier draft of this paper Michael Radich and Antonello Palumbo sent numerous additional examples, drawn from Buddhist translations and secular 40. Just as this paper was going to press I received word from Antonello Palumbo that W. South Coblin had also identified Pulinnai 蒱 隣 奈 as a transcription of Bārāṇasī in a discussion of the use of yu 魚 -group characters in Eastern Han Buddhist transcriptions. Coblin does not provide any details on the derivation of the term, however, nor does he offer an explanation for the unexpected use of the character 隣 for -rāṇ- (Coblin 1993, p. 898). 41. See Sakurabe 1975 and Harrison 1990, pp. 221–235 and 248–249; for a summary of their findings together with some additional remarks on the “rhetorical community” to which the revised text belongs see Nattier 2008, pp. 81–83. 42. See the Chu sanzang ji ji, T2145, 55.6b22. 43. For a brief discussion of this issue see Nattier 2008, p. 132. 143 sources respectively, of this same interchange.44 Indeed, wrote Palumbo, “the likelihood is that, whenever Han transcriptions of foreign terms are involved, man 滿 must be a scribal error for pu 蒲 .”45 He also observed that “pu 蒲 is regularly used as a phonogram in late Western Han transcriptions of foreign terms in the Han shu – Xiyu zhuan [漢書 – 西域傳] . . . This is not the case with man 滿, which never occurs in W[estern] Han secular transcriptions, and only in one case in E[astern] Han, where, however, it is almost certainly again a typo for pu 蒲 (the name of the Parthian king Manqu 滿 屈 , which must be 蒲 屈 for Pakur < Pacores).” There is a clear directionality, in other words, in transcriptions from this period and their later transmitted forms: it is pu 蒲 that represents the original version, while man 滿 is the result of scribal emendation.46 In Zhi Qian’s (original) term Puluo 蒲羅, in sum, we have a case parallel to that of Weixie 維耶 for Vaiśā[lī], where the translator has opted for transcription rather than translation but has done even greater violence to the name by transcribing only two of its four syllables. It is thus hardly surprising that the presence of the city of Bārāṇasī on Zhi Qian’s list has gone unrecognized for so long. Conclusions By examining each of the four city names contained in the passage from Zhi Qian’s Mahāparinirvāṇa Sūtra discussed above, we have been able to show that the list of names found in his translation is not nearly as abberrant as Bareau had supposed. The four names (not just three) that it contains—translations of Śrāvastī and Rājagṛha, and abbreviated transcriptions of Vaiśālī and Bārāṇasī—are found in virtually all other versions of the text, with only the Pāli lacking a reference to Vaiśālī. Indeed, Zhi Qian’s list bears a particularly close relationship to that found in T5 (Bareau’s “Chinese B”), which has the same four names in the same order, but with the addition of Sāketa and Campā between the first and second names found on Zhi Qian’s list.47 We have also been able to observe at close range the translator’s methodology, used at least in this part of the text: all proper names should consist of just two characters, and in transcriptions of Indic-language names this sometimes resulted in radical amputation of the Texts where Radich found the alternation between pu 蒲 and man 滿 in the critical apparatus to the Taishō edition were T125 (2.767c7 and passim), T221 (8.l108c24), T643 (15.677c18), T721 (17.405a3), T2122 (53.565a16), T2128 (54.789b12), and T2129 (54.955c4). He also located an interchange between pu 蒲 and pu 蒱 in T21 (1.265a2–3) and T23 (1.294b26). In secular sources Palumbo cited examples of the replacement of an original pu 蒲 by man 滿 from the Shiji 史記 (the name of an ethnic group in Sichuan, the Baoman 苞滿, for which the Tang commentary Shiji suoyin 史記索隱 reports that some manuscripts have pu instead of man), the Hou Han shu 後漢書 (a walled city in Turfan named Jinpu cheng 後漢書, later replaced by Jinman 金滿 in written sources), and the Sanguo zhi 三國志. quoting the lost Wei lüe 魏略 (a Central Asian kingdom called Manli 滿犁, mentioned several times in the Han shu by the name Puli 滿犁). 45. Antonello Palumbo, e-mail of 8 February 2021 (quoted with permission). 46. Palumbo also offered the very interesting observation that the transcription of the two syllables bā and rā as puluo 蒲羅 is a hybrid form, drawing the first character from the yu 魚 group and the second is from the ge 歌 category (op. cit., quoted with permission). This would harmonize, in my view, with a hypothetical scenario in which Zhi Qian took the transcription of the name as Pulinnai 蒱隣奈 (borrowed from Lokakṣema during his revision of the text now preserved as T790) as his point of departure, but modified it—perhaps now dissatisfied with the use of the character lin 隣 to transcribe rāṇ—in light of his familiarity with the later form Boluonai 波 羅奈 introduced by Kang Mengxiang and found in other post-Lokakṣema translations. 47. Chinese B (T5, 1.169c14–15): 舍衛 (Śrāvastī), 沙枝 (Sāketa), 栴波 (Campā), 王舍 (Rājagṛha), 波羅奈 (Bārāṇasī), 維耶梨 (Vaiśālī). 44. 144 original term. As we will see in Part 2 of this study, the same is true of the proper names— both the names of cities and of the clans that inhabited them—in the portion of Zhi Qian’s Mahāparinirvāṇa Sūtra that relates the arrival of various groups to claim their share of the Buddha’s relics after his cremation. Finally, we have seen that the best chance for success in identifying the prototype of a given Chinese transcription or translation is to examine the term in question from as many angles as possible. This includes paying close attention to the stylistic features of the passage in which the term occurs, to parallels in texts by other translators, and to any available instances of other translations or transcriptions of the term in question found in works by the same translator. Finally, it is always necessary to bear in mind that the received text as we have it may not be identical with the one that left the original translator’s desk. Scribal errors, especially in the case of transcriptions, are abundant in the texts that have come down to us, and it is essential not to restrict our inquiry to the forms of a given term that appear in even the best critical editions of canonical texts. Though the passage explored above is quite brief, the number of steps that were required to clarify the identity of the proper names it contains has made it clear that such work requires a multi-faceted methodological approach. In Part 2 of this study we will examine a second cluster of proper names from Zhi Qian’s Mahāparinirvāṇa Sūtra, which will require similarly rigorous techniques and which will yield some similarly surprising results. Bibliography Akanuma Chizen 赤沼智善, 1931. Indo-bukkyō Koyū-meishi Jiten 印度佛教固有名詞辞典, Kyoto (21967). Bareau, André, 1971. Recherches sur la biographie du Buddha dans les Sūtrapiṭaka et les Vinayapiṭaka anciens. Vol. 2, Part 2. Paris: École Française d'Extrême-Orient. Boucher, Daniel J., 1998. “Gāndhārī and the Early Chinese Buddhist Translations Reconsidered: The Case of the Saddharmapuṇḍarīkasūtra.” Journal of the American Oriental Society, vol. 118, no. 4, pp. 471–506. Coblin, W. South, 1993. “BTD Revisited: A Reconsideration of the Han Buddhist Transcriptional Dialect.” Bulletin of the Institute of History and Philology, Academia Sinica, Vo. LXIII, Part IV (Taipei, Taiwan), pp. 867–943. Harrison, Paul, 1990. The Samādhi of Direct Encounter with the Buddhas of the Present. Tokyo: The International Institute of Buddhist Studies. _____, 1997. “The Ekottarikāgama Translations of An Shigao.” In Petra Kieffer-Pülz and Jens-Uwe Hartmann, eds., Bauddhavidyāsudhākaraḥ: Studies in Honour of Heinz Bechert on the Occasion of His 65th Birthday (Indica et Tibetica 30) (Swisttal-Odendorf: Indica et Tibetica), pp. 261–283. _____, 2002. “Another Addition to the An Shigao Corpus? Preliminary Notes on an Early Chinese Saṃyuktāgama Translation.” In The Sakurabe Ronshu Committee, ed., Early Buddhism and Abhidharma Thought in Honour of Dr. Hajime Sakurabe on His Seventy-seventh Birthday (Kyoto, Japan: Heirakuji shoten, 2002), pp. 1–32 [horizontal section]. Hirakawa Akira 平川彰, ed., 1997. Buddhist Chinese-Sanskrit Dictionary (Bukkyō kanbon daijiten 佛教漢梵大 辭典). Tokyo: The Reiyukai. Karashima Seishi 辛嶋静志, 1998. A Glossary of Dharmarakṣa's Translation of the Lotus Sutra. Bibliotheca Philologica et Philosophica Buddhica, I. Tokyo: The International Research Institute for Advanced Buddhology, Soka University. _____, 2010. A Glossary of Lokakṣema’s Translation of the Aṣṭasāhasrikā Prajñāpāramitā. Bibliotheca Philologica et Philosophica Buddhica XI. Tokyo: Soka University, The International Research Institute for Advanced Buddhology. Kroll, Paul W., 2015. A Student’s Dictionary of Classical and Medieval Chinese. Leiden/Boston: Brill. Marshall, Sir John, 1975. Taxila: An Illustrated Account of Archaeological Excavations Carried Out at Taxila under the Orders of the Government of India, between the years 1934 and 1934, Vol. 1. Delhi: Motilal Banarsidass. Reprint of the 1951 edition (Cambridge University Press). Nakamura Hajime 中村元, 1981. Bukkyōgo daijiten 佛教語大辞典. Tokyo: Tōkyō Shoseki. 145 Nattier, Jan, 2003a. “The Ten Epithets of the Buddha in the Translations of Zhi Qian 支謙.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University for the Academic Year 2002 (Tokyo, 2003), vol. VI, pp. 207–250. _____, 2003b. A Few Good Men: The Bodhisattva Path according to The Inquiry of Ugra (Ugraparipṛcchā). Honolulu: University of Hawaii Press. _____, 2008. A Guide to the Earliest Chinese Buddhist Translations: Texts from the Eastern Han 東 漢 and Three Kingdoms 三 國 Periods. Bibliotheca Philologica et Philosophica Buddhica, X. Tokyo: The International Research Institute for Advanced Buddhology, Soka University. _____, 2008[2010]. “Who Produced the Damingdu jing 大明度經 (T225)? A Reassessment of the Evidence.” Journal of the International Association of Buddhist Studies, vol. 31, nos. 1-2, pp. 295–337. Ñāṇamoli, Bhikkhu, trans., 1999. The Path of Purification (Visuddhimagga) by Bhadantācariya Buddhaghosa. Seattle: Buddhist Publication Society. Reprint of the 1975 edition. Park, Jungnok, 2008[2010]. “A New Attribution of the Authorship of T5 and T6 Mahāparinirvāṇa-sūtra.” Journal of the International Association of Buddhist Studies, vol. 31, nos. 1-2, pp. 339–367. Puini, Carlo, 1919. Mahaparinirvana-sutra, ovvero il libro della totale estinzione del Buddha nella redazione cinese di Pe-fa-tsu. Lanciano: R. Carabba. Radich, Michael, 2018. “A Triad of Texts from Fifth-Century Southern China: The *Mahāmāyā-sūtra, the Guoqu xianzai yinguo jing, and a Mahāparinirvāṇa-sūtra ascribed to Faxian.” Journal of Chinese Religions vol. 46, no. 1, pp. 1–41. DOI: 10.1080/0737769X.2018.1435370. _____, 2019. “Was the Mahāparinirvāṇa-sūtra 大 般 涅 槃 經 T7 Translated by ‘Faxian’? An Exercise in the Computer-Assisted Assessment of Attributions in the Chinese Buddhist Canon.” Hualin International Journal of Buddhist Studies: E-journal vol. 2, no. 1, pp. 229–279. Sakurabe Hajime 桜部建, 1975. “Hanjuzanmaikyō kanken” 般舟三昧経管見. Hashimoto Hakushi taikan kinen Bukkyō kenkyū ronshū 橋本博士退官記念佛教研究論集 (Osaka), pp. 173-180. Samtani, N. H., ed., 1971. The Arthaviniścaya-Sūtra & Its Commentary (Nibandhana) (Written by Bhikṣu Vīryaśrīdatta of Śrī-Nālandāvihāra). Critically edited and annotated for the first time with Introduction and several Indices. Patna: K. P. Jayaswal Research Institute. _____, trans., 2002. Gathering the Meanings: The Compendium of Categories, the Arthaviniścaya Sūtra and its Commentary Nibandhana. Translated from the Sanskrit with an introduction and notes by N. H. Samtani. Berkeley: Dharma Publishing. Schuessler, Axel, 2009. Minimal Old Chinese and Later Han Chinese: A Companion to Grammata Serica Recensa. Honolulu: University of Hawaii Press. Waldschmidt, Ernst, 1950-51. Das Mahāparinirvāṇasūtra : Text in Sanskrit und Tibetisch, verglichen mit dem Pāli nebst einer Übersetzung der chinesischen Entsprechung im Vinaya der Mūlasarvāstivādins / auf Grund von Turfan-Handschriften herausgegeben und bearbeitet von Ernst Waldschmidt. 3 vols. Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin, Philosophisch-historische KlasseBerlin: Akademie-Verlag. Walshe, Maurice, trans., 1987. Thus Have I Heard: The Long Discourses of the Buddha. London: Wisdom. Warren, Henrry Clarke, and Dharmananda Kosambi, eds., 1950. Visuddhimagga of Buddhaghosâcariya. Cambridge, Massachussetts: Harvard University Press. Zacchetti, Stefano, 2005. In Praise of the Light: A Critical Synoptic Edition with an Annotated Translation of Chapters 1-3 of Dharmarakṣa’s Guang zan jing 光讚經, Being the Earliest Chinese Translation of the Larger Prajñāpāramitā. Bibliotheca Philologica et Philosophica Buddhica VIII. Tokyo: The International Research Institute for Advanced Buddhology, Soka University. Zürcher, Erik, 2007 [1959]. The Buddhist Conquest of China. 3rd edition. Leiden: E. J. Brill. 146