Tibetan Evidence for the Sources of Chapters of the Synoptic Suvarṇa-prabhāsottama-sūtra T 664 Ascribed to Paramārtha

Tibetan Evidence for the Sources of Chapters of the Synoptic Suvarṇa-prabhāsottama-sūtra T 664 Ascribed to Paramārtha

Michael Radich VictoRia UniVeRsity of Wellington, neW Zealand michael.radich@vuw.ac.nz

abstRact

Four chapters survive of a supposed translation of the Suvarṇaprabhāsottama-sūtra by Paramārtha (499–569). Versions of these chapters are also found in a later Chinese version of the sūtra by Yijing. In earlier work, I have argued that these chapters were most likely composed in China, basing my argument upon extensive verbatim correspondences between these chapters and a number of earlier Chinese texts. However, a significant obstacle still stands in the way of this thesis. A Tibetan version of the sūtra (here called ‘Tib II’) also includes the same chapters, and Tibetan tradition holds that this version is a translation from Sanskrit. Here, I examine evidence that suggests that these portions of Tib II might in fact be translations from Chinese, despite the reports of Tibetan bibliographers. In closing, I consider some broader implications of my findings.

Keywords

Tibetan translations, sūtras of Chinese composition, Tibetan catalogues, Trikāya

Introduction

In the synoptic version of the Sūtra of Golden Light (Suvarṇa-[pra]bhāsottama-sūtrendra-rāja; hereafter ‘Suv’), viz. the Hebu jin guangming jing 合部金光明經 T 664 compiled in 597 by Baogui 寶貴 (d.u.), four chapters are ascribed to Paramārtha真諦 (499–569; hereafter ‘P’). In an earlier study (Radich 2014), I showed that P’s Chinese versions of these chapters betray heavy debts to a range of previously unnoticed sources in the Chinese tradition. I further argued that in light of the overall pattern of these debts, the most economical conclusion would ordinarily be that P’s chapters were in some sense ‘composed in China’; that is to say, that they are texts ‘between translation and composition’ (Funayama 2006a), rather than entirely straight translations. However, I ended that study by noting that one major impediment stands in the way of such a conclusion: the Tibetan version of the text that Nobel called ‘Tib II’ is normally taken as a direct translation from an Indic original, and contains versions of these same chapters. In the present study, therefore, I will examine Tibetan evidence pertaining to whether or not P-Suv, or parts thereof, could have been composed in China.

Before considering this Tibetan evidence, I will first briefly review other evidence against my hypothesis, and in favour of the traditional understanding that the chapters in question were translated by P from an Indic source text.

On the basis of external evidence alone, the ascription to P of the chapters at issue would appear to be among the most secure ascriptions in all of P’s extant corpus.2 We have already seen that the ascription is seemingly supported by Baogui’s synoptic version of 597. In addition, the claim that P authored a translation of Suv is also supported by the earliest bibliographers after his era — Fajing 法經 (d.u., catalogue of 594), Fei Changfang 費長房 (d.u., catalogue of 597), and Yancong 彥琮 (557–610, catalogue of 602). The agreement among these sources is all the more striking given that, in other cases, these bibliographers sometimes represent two alternate traditions, which were only reconciled and unified (often artificially, and at cost) by their successors. In addition, interlinear notes in Fei Changfang and the biography of P in the Xu gao seng zhuan 續高僧傳 give highly specific details about the place of P’s translation, the circumstances of patronage under which it was produced, and persons who participated in the translation process. Moreover, two prefaces to the work have been preserved: one by Sengyin 僧隱 (d.u.), preserved in a manuscript version of the text in the Shōgozō 聖語蔵 collection; and another ascribed to Yancong.3 These prefaces, especially that of Sengyin, give considerable further detail. Moreover, Sengyin’s preface is cited by Yancong, which shows that it, too, should predate 597. Sengyin, moreover, is supposed himself to have been present when the text was translated (according to him, in 553). Bibliographers from Fei Changfang onwards also report that P authored a commentary on Suv. Such ‘commentaries’ seem often to have been produced alongside translations, and these reports might therefore also be taken as additional evidence that P translated a version of Suv (Radich 2012, 83). Indeed, fragments of a commentary on Suv ascribed to P have survived, and they provide quite firm evidence that all four of the chapters here called ‘P-Suv’ were indeed known to someone identified by the tradition as ‘Paramārtha’. To briefly list this evidence: The Trikāya chapter is mentioned in a comment of P’s reported by Huizhao

of abbreviations appears at the end of this paper. For details of the various witnesses to the texts under study, see Radich 2014. Chinese is transcribed throughout using the Hanyu Pinyin system.

2. The following paragraphs repeat information given in Radich 2014. For the evidence upon which the following summary is based, and other details about P-Suv, see Radich 2012, 47–50.

3. This ascription is carried in the Yuan and Ming editions only; T XVI 359 fn. 2; see also Ono 1934, 17. 慧沼 (648–714). Wǒnch’uk 圓測 (613–696) also reports a comment of P’s on the exposition of the three bodies in the text. Quotes in later works by Huizhao, Wǒnch’uk, Fazang 法藏 (643–712), Daosui 道邃 (d.u., fl. 8 c.), Hyōbi 平備 (fl. 8–9c), and Gangyō 願曉 also preserve P’s comments on specific loci and doctrines in all four chapters of P-Suv (and some other chapters of Suv).

Citations in other works also furnish quite extensive evidence that all of these chapters except P-Suv-yikong must have been in circulation by the Sui 隋 dynasty (581–618). These three chapters are already discussed and quoted, sometimes quite extensively, in the works of Huijun (慧均, d.u., fl. 574–580s?), Jingying Huiyuan 淨影慧遠 (523–592), Jizang 吉藏 (549–623) and Zhiyi 智顗 (538–597). P-Suv-trikāya is particularly well attested in these early witnesses. Thus, in sum, the external evidence that P did indeed translate portions of Suv is nearly as strong as such evidence gets for any text in his corpus. Against this evidence, however, we must weigh the extensive internal evidence surveyed in my previous work (Radich 2014) showing heavy debts to a range of earlier Chinese sources, in a distribution characteristic of works composed in China. Analysis of the Tibetan evidence Several extant versions of Suv contain versions of the chapters here called ‘P-Suv’, including the version by Yijing (‘YJ’) and several Tibetan translations, some of which are fragmentary. If those chapters were originally composed in China, it should be possible to trace all known versions of those chapters to a Chinese parent text. I will first briefly discuss several other versions of the text, before homing in on Tib II, which seems to pose a particular obstacle for the hypothesis that parts of P-Suv could have been composed in China.

Given that Yijing certainly had access to the text of P, we can provisionally regard it as possible in principle that YJ could derive from P. Further, Nobel’s Tib III is known to be a translation from YJ, and for that reason, we will also not regard it as likely to constitute an independent witness. The same goes for the texts shown by Oetke to represent two further (largely fragmentary) independent translations from YJ into Tibetan. It is also noteworthy that no Skt is known for any part of the chapters comprising P-Suv. This is particularly striking when we consider the Central Asian Sanskrit manuscript fragments. More than 80 fragments of portions of Suv have so far been identified, representing most chapters of the text as known from the fuller Nepalese Skt and the Khotanese; but no fragment has yet been identified as corresponding to any part of P-Suv.12 The Khotanese version of the text also contains nothing corresponding to any part of P-Suv.13 These striking gaps might lead us to question whether or not P-Suv, including corresponding portions of Tib II, are necessarily based upon Skt originals at all.

By a process of elimination, this means that the text we would consider most likely to constitute an independent witness is Nobel’s ‘Tib II’, which includes versions of all four of the chapters comprising P-Suv. All scholars to date have regarded Tib II as a translation based solely upon a Skt original. If this is correct, then Tib II would obviously represent an independent witness. On this basis, we would have to eliminate the possibility that any part of P-Suv was composed directly in Chinese.

However, close examination of the details of Tib II itself, the Tibetan bibliographic record concerning translations of Suv, and other circumstances surrounding the chapters of P-Suv show that we cannot be so certain that a Skt original existed for all of P-Suv as represented by Tib II. Indeed, I will try to demonstrate that especially in the case of P-Suv-trikāya, we cannot rule out the possibility that the version found in Tib II ultimately derives from P. The circumstances surrounding Tib II itself are somewhat convoluted and murky. According to the catalogues, Tib II was translated by Jinamitra, Śilendrabodhi and Ye shes sde under King Khri gtsug lde btsan (806–?838), alias Ral pa can (Nobel 1944, xiii). Nobel also states categorically that Tib II is based upon a Skt no longer extant.14 However, against this apparently simple picture, we should note a number of complications.

First, it is already clear that in some senses at least, Tib II is a hybrid. In particular, Nobel notes that in portions matching the shorter and earlier Tib I, which was indeed translated from Skt and corresponds to our extant Skt, Tib II is largely verbatim identical with Tib I.15 We can therefore suspect that for these portions, at least, Ye shes sde and his collaborators borrowed from Tib I rather than translating afresh from a Skt original. This shows that in principle, they were happy to borrow from other existing translations.16 In fact, according to Nobel, Tib II was only ‘revised and published’17 by the team of Jinamitra, Śīlendrabodhi and

lects, edits and studies all known fragments. For these details about the mss., see xxvi, xxxi (following Matsuda Kazunobu). The absence of Skt evidence for DhKṣ’s Mahā-parinirvāṇamahāsūtra is one of the main reasons scholars have doubted its Indic provenance. 12. Yuyama 2004, Skjaervo 2004, 1:xxxiii–xxxvii, Skjaervo 2009. Cf. Nobel 1944, xxii–xxiii. 13. This is so even though the Khotanese manuscript Q, according to Skjaervo, belongs to an early stage of ‘Recension D’, which is otherwise also represented by Tib II; Skjaervo 2004, 1:lvii. 14. Nobel 1958, xxvii; further on Tib II, see Nobel 1944, xx. 15. Nobel 1944, xiii. So much is this so that in some printings of the Tibetan canon, the editors seem to have regarded it as superfluous to print Tib I separately. See also Nobel xxiii.

16. Nobel also says that the authors of Tib II did exercise some criticism towards Tib I, and did not merely take it over holus-bolus; Nobel 1944, xx–xxi. 17. … durchgesehen und herausgegeben; perhaps paraphrasing skad gsar bcad kyis bcos nas gtan la phab pa’o of the Tibetan colophon. I had access to the sTog version of the colophon only, via the U. Wien website, http://www.istb.univie.ac.at/kanjur/xml3/xml/ (accessed 2 November 2013). Ye shes sde, and he suggests that ‘the actual translation, which was however perhaps considerably modified, must be somewhat older’.18 If this is the case, it means that we know little about the time and circumstances of the original translation behind Tib II, and this gives us additional grounds for uncertainty about the actual source material(s) behind it. In addition, it seems that there may indeed have once existed at least one other translation (if not plural translations) from Ch, from which the translators or compilers of Tib II could in fact have borrowed. Nobel states that under King Mes ag Tshom (Khri lde gstug btsan, r. 704–755), Suv was translated from Ch — this being a different translation than Tib III, which is known to have been from Yijing (Nobel 1958, xvi).19 In addition, Oetke’s painstaking work has shown that there probably existed, at one time or another, at least three separate translations of Suv from Ch into Tib (Oetke 1977).20 Old Tibetan catalogues also contain confusing and conflicting information, suggesting that more translations once existed than are now extant. Indeed, Herrmann-Pfandt has even suggested that one of

I am indebted to Jan Nattier for the suggestion that Nobel might have had this colophon in mind. 18. ... die eigentliche, vielleicht aber stark veränderte, Übersetzung muss demnach etwas älter sein; Nobel 1958, xxvii–xxviii. The basis upon which Nobel makes this statement is unclear to me. Nobel cites Rockhill 1884, 224, who merely makes a general statement that these scholars and their contemporaries ‘corrected all the translations made previously’ etc.

19. Nobel cites Rockhill 1884, 218, but I cannot see the evidential basis for Rockhill’s statement. He also refers to Hoffmann 1956, 28 ff. 20. Much of Oetke’s book is devoted to demonstrating this claim, but see in particular the loci already cited above, n. 10. Oetke does not discuss in any detail the problem of the likely dates of the translations underlying the various witnesses now extant for each; nor does he discuss the problem of the possible identity or difference of any of the additional two versions his research uncovered with the texts in the various catalogues (for which see following n. 21). Cf. also remarks in Ueyama 1990, 122–124.

the texts recorded in those catalogues, lHan kar ma no. 87, was a translation from Baogui. If, then, as we have seen above was the case for Tib I, the translators of Tib II were happy to incorporate material from earlier versions, it seems possible that parts of some alternate translation from Chinese could also have been incorporated into Tib II, and that may have been a route by which some or all chapters of P-Suv could have made their way into Tib II from Chinese, and thus ultimately derive from P, rather than from a putative Skt original.

Much hinges, then, on whether portions of Tib II corresponding to P-Suv could plausibly be derived either directly or at some remove from P. This is in turn a somewhat complex question, for several reasons. First, Oetke’s remarkable work has amply shown that assessing the relationship between Tibetan translations and their Chinese source texts can be a very complex matter (Oetke 1977, passim). In addition, the question of features of style (including terminology and translation lexicon) that might distinguish Tib translations from Ch from other translations made from Skt is still an underdeveloped topic of investigation, and prior work affords few handholds by which to grapple with such problems. Any treatment of this problem within the scope of the present study must therefore be tentative, but with that caveat, I believe that a first survey of relevant portions of Tib II indicates that at least one chapter (‘Tib II-trikāya’) and possibly another (‘Tib II-yikong’), could in fact be translated from Chinese.

The (much later) Bu ston catalogue lists three translations, viz. nos. 208, 209, and 210. Nishioka is unsure that no. 208 matches to an extant text, and suggests only tentatively Tib II = D556/Q175; Nishioka 1980, 90, 1983, 179. It is also puzzling that the catalogue itself gives a different translator for this text, namely, rNam par mi rtog pa (*Nirvikalpa); Nishioka 1980, 72. (Demiéville has suggested that the translator rNam par mi rtog pa might be identified with 南撥特計波, a monk who accompanied a diplomatic mission to the Chinese court in 804; Demiéville 1952, 226–230 and 228 n. 1; cf. Tucci 1958, 47/351 n. 1.) Herrmann-Pfandt disagrees with Nishioka that Bu ston no. 208 corresponds to D556/Q175; she also notes that the translator differs; she regards Bu ston no. 208 as instead probably matching lHan kar ma no. 87, which she thinks is probably from Ch (from B; see above); Herrmann-Pfandt 140. In addition, a list of texts directly after no. 1427, said to have been ‘written ... by someone ... in error’ (kha cig gis ... bris pa ni nor ba yin) also includes a gSer ’od dam pa, of the identity of which Nishioka is also unsure; Nishioka 1983, 64–65.

The evidence for a Chinese original is, I believe, strongest for Tib II-trikāya. First, even just in Chinese, the correspondence between YJ and P is much closer than in the case of P-Suv-dhāraṇī 25 for example, to a degree that might lead us to suspect that YJ is only tweaking the wording of P, and does not himself necessarily have an independent (Skt) source. Second, in Tibetan, also, Nobel was moved to comment on the unusually close degree of overlap between Tib II and Tib III.26 When we read Tib II, P and YJ against one another, moreover, in passage after passage, the match between Tib and Ch is extraordinarily close. Such discrepancies as can be found amount to little more than minor variation in wording here and there;27 larger discrepancies such as those described below for P-Suv-dhāraṇī and -yezhang do not occur. Thanks to this extraordinarily close relation between the Ch and Tib texts, it is difficult to find any evidence that categorically disproves the possibility that Tib II was translated from Chinese. In itself, however, this negative evidence does not give us sufficient grounds to overrule the information of the Tibetan tradition, which holds that Tib II is indeed from an Indic original. We therefore need to look for additional evidence positively supporting a relation between Ch and Tib II.

For this chapter (-trikāya), discrepancies between Tib II and YJ are even fewer than those between Tib II and P.28 This means that most discrepancies between Tib II and P are reducible to questions of why YJ might have differed from P, but do

25. Cf. Nobel 1958, 41 n. 1. Takasaki also observes that YJ is much closer to his predecessors for the chapters of P-Suv in general than in other parts of the text, and notes instances in which YJ follows P in translation terminology, in preference to his own usual translations (e.g. for tathatā, śūraṃgamasamādhi); Takasaki 1974, 332. 26. Nobel says that although Tib II ‘has nothing to do with’ Tib III (he does not give his reason for this judgement), it nonetheless overlaps extensively with Tib III: ‘In some places, Tib II even matches better with Yijing’s version than Tib III’ (An einigen Stellen stimmt Tib. II sogar genauer zu I-tsing’s Version als Tib. III). He concludes from these facts that Yijing must have been very closely following a presumed Skt (so that Tib II would have been translated with equal precision from a virtually identical text) (Daraus geht hervor, dass I-tsing’s Version sich sehr eng an den zugrunde liegenden Skr.-Text anschliesst, ja sogar noch genauer als die Übertragung des P[aramārtha]); Nobel 1958, 41 n. 1. However, this conclusion clearly proceeds from the assumption that the translators of Tib II (and, indeed, YJ) must have had a Skt original before them, a possibility which, in light of the evidence presented below, I would suggest is not certain.

27. In this connection, we should bear in mind that the modern editions of the Ch texts as we have them are not certain guides to the exact wording of any hypothetical Ch sources that Tib translators may have used, so that minor differences in wording in Tib II could reflect variant readings in older Ch versions of a source text. For full consideration of the questions before us here, then, critical editions of the chapters in P-Suv, based upon consultation of all extant versions, would certainly be a desideratum. However, in view of the popularity of Suv among the Dunhuang manuscripts and even in the ‘stone canon’ at Fangshan 房山, and manuscripts preserved in Japan (cf. once more Radich 2014, and Ono 1929, 1934; Takeda Kagaku 2010, 218–225), preparation of such an edition could also be a daunting task.

28. To give only a few examples (references to Tib II refer to page and line number in Nobel 1958): • Tib II 201.16–17, YJ 408b12–15, missing at P 362c21; • Tib II mya ngan las ’das pa la gnas pa ma yin pas etc. 204.17–18, YJ 不住涅槃 409a9, P 不住於般涅槃 363b14 (cf. Nobel 52 n. 2); • Tib II mdzad pa chen po phun sum tshogs pa mngon du mi ston pa 205.29–30, YJ 具足大用不顯現故 409b6–7, P 具足之用不顯現故 363c8 (here, we should perhaps consider the possibility that 大 is a scribal error for 之, or vice versa; I am grateful to Jonathan Silk for this suggestion); • P 何者正修行而得清淨離於不淨 364b3, not matched at Tib II 209.4–5, nor at YJ 410a6 (however, for this same passage, see also p. 256 below);

not pose any challenge to the possibility that Tib II of this chapter was prepared without reference to any Skt.29 Our task is thus simplified, and for the most part, it will be sufficient for us to compare YJ and Tib II. In fact, comparison does reveal some grounds for suspicion that Tib II was translated from YJ (or a Ch text very much like it). In a number of passages, unusual wording in Tib II can best be explained by the supposition that the Tibetan translators were misreading the wording we see in YJ’s Chinese.30 Passage 1

YJ 409a2: 如是受化諸弟子等，是法身影，以願力故，於二種身現種種相，於法身地無有異相. Tib II 203.32–204.2: de bzhin du longs spyod rdzogs pa’i sku dang sprul pa’i skus ni ’khor rnams la chos kyi sku’i gzugs brnyan ston par smon lam gyi dbang gis sku de gnyis snang ste | mtshan ma sna tshogs su snang yang chos kyi sku’i gnas la ni mtshan nyid tha dad pa med do. The phrase 於二種身現種種相 in YJ could naturally be translated, ‘[the dharmakāya] manifests various features/characteristics (相 xiang, *lakṣaṇāni) in the [other] two kinds of body ...’. Here, it looks as if the translators of Tib II have misconstrued the four-four-four prosody of the Chinese, and read Ch as if is punctuated 於二種身現、種種相 ..., ‘... these two bodies are manifested, and manifested (sic) as various features/characteristics ...’. This hypothesis still cannot account for the redoubling of the verb snang, but we might speculate that the second instance of snang is the trace of an attempt by the Tib translators to recuperate some sense from the text after their initial misreading.

• Tib II chos ’di thos pa’i ’bras bu rnam par smin pa yang zad par mi ’gyur ro 212.16–17, YJ 由聞法故，果報無盡 ... 410c13, P 於佛起業果報無邊 365a10; • Tib II btsun mo dang | slas dang | sras dang | sras mo dang blon po rnams kyang ’dum zhing rtsod pa med pa dang ... 213.22–23, YJ 二者、中宮妃后、王子諸臣和悅無諍 411a6–7 (match between Tib II and YJ is not perfect here either), P 二者輔相大臣和悅無諍 ... 365b1–2;

• Tib II nad med pa dang bde bas glo bur du chi ba med pa dang 213.26–27, YJ 無病安樂，無抂死者 ... 411a9, P 年命長遠富逸安樂 365b3. For some reason, such instances seem to increase in number towards the end of the chapter. Reverse examples of a better match between Tib II and P can be found, but are fewer and slighter (for example, Tib II dper na ’ga zhig gnyid log pa’i rmi lam na chu chen pos khyer nas lus dang rkang lag bskyod de 210.15–17, YJ 譬如有人於睡夢中，見大河水漂泛其身 410a29–b1, P 譬如有人於臥寐中夢見大水流泛其身 364c1–2; cf. Nobel 72 n. 4). 29. Why Yijing might have differed from prior translators in any of his texts is another complex question, potentially bearing on Yijing’s entire working method, which is another matter that scholarship to date does not equip us to address easily. I have been unable to discover any concerted studies of Yijing’s translation idiom. The most detailed relevant studies of which I am aware are Matsumoto 2007, 2008, but unfortunately, these are studies of transcription terms only. For the present, then, we should set aside the ancillary question of YJ’s reasons for deviating from P. I am grateful to Prof. Funayama Tōru for drawing my attention to Matsumoto’s work. 30. The method here resembles to some extent that of Oetke 1977, 37–40, 48–50, though my aims and findings differ; cf. also Takasaki 1976, esp. examples discussed 463–464; cf. also Wilkens’s attempt to ascertain whether the ‘Trikāya’ chapter of Mongolian III is based upon Tib rather than Ch; Wilkens 2001, 10–

Passage 2 YJ 409a18–19: 一者、起事心，二者、依根本心，三者、根本心. Tib II 204.30–32: [[[gsum]] gang zhe na] | dngos po la ’jug pa’i rnam par shes pa dang kun gzhi la gnas pa’i yid dang kung [sic!, for kun, Nobel 1958, 55 n. 1] gzhi rnam par shes pa’o. This passage is quite obviously discussing some kind of Yogācāra model, whereupon mind is divided into various types of consciousness (usually eight), which are then grouped into three main groups, corresponding to 1) the six main sense consciousnesses, 2) manas and 3) ālayavijñāna. The terms used for the three categories here raise a number of interesting problems, but for our present purposes, the most important is the middle category. As Nobel points out, it is ‘noticeable’ (or ‘odd’, auffallend) that Tib II here features kun gzhi la gnas pa’i yid (yid usu. = manas) in this term, instead of rnam par shes pa, which corresponds to the same Ch xin 心 in the two terms on either side of it (Nobel 1958, 54 n. 3). This could be another tell-tale sign that the translators of Tib II had before them the Chinese yigenbenxin 依根本心 (which is equally unusual in Ch).31

31. These rare terms further complicate the notion that P-Suv-trikāya was produced only by ‘P’. 1. Even benxin 本心 (without considering whether or not it is found in any larger compounds) almost never appears in any work ascribed to P: the sole instance I can find is a use to translate buddhi, 四本心於塵馳動今住一塵, Abhidharma-kośa-bhāṣya T XXIX 1559 213c23–24, Skt. capalātmano ’py ekatra viṣaye buddhir avatiṣṭhate, Pradhan 1967, 157.7, La Vallée Poussin 1980, 2:136.

2. The threefold set of terms found here seems to be unique, and is not found elsewhere in P: a) As already follows from the above, genbenxin 根本心 appears nowhere in P’s works. In fact, this compound is very rare altogether. One possible instance (which is difficult to understand) is found in the Akṣayamati-nirdeśa in the Saṃnipāta T XIII 397(12) 200c27. We also find scattered instances in Zhu Fonian, T XVI 656 2b15; in Bodhiruci, T XXVI 1525 258c1, 269c18; in Xuanzang’s 玄奘 (!!) MSgBh T XXXI 1597 354a24–29; the Mahāyāna-sūtrālaṃkāra T XXXI 1604 624a13–16; and in apparently later texts (T XVI 681 734a26–27; T XVIII 867 258b29–c15, 269b9; etc.). Only in Bodhiruci and Xuanzang does the term seem to correspond to *ālayavijñāna/*mūlavijñāna (and then rather loosely). b) Yigenbenxin 依根本心 never appears in any other translation text (nor does 依根本識). c) Qishixin 起事心 never appears in any other translation text (nor does 起事識). We should bear in mind here that it was P’s central mission, or at least his accomplishment, to transmit the Yogācāra system to China; and it is difficult to think of concepts more central to that system than the schema of manifold consciousness, or ālayavijñāna as the keystone of that system. Translators naturally sometimes use multiple equivalents for a single Indic term, but it would be peculiar if P deviated from his usual translation terms for concepts as technical and central as these. Against these facts, however, we might also note that Huizhao’s Suv commentary preserves a comment ascribed to P, identifying these terms with more ordinary Yogācāra concepts: 言「三心」者、依真諦三藏釋云：一、起事心、是六識皮；二、依根本心、是第七肉；三、根本心、是阿梨耶識骨, T XXXIX 1788 218b22–24 (I am grateful to Ching

Keng for pointing out this passage to me). Another similar instance may be found when P-Suv-trikāya invokes the (equally central) Yogācāra doctrine of the ‘three marks’ (三相; *trilakṣaṇa, *lakṣaṇa-traya). The terms used are: 1. *parikalpita-lakṣaṇa, 思惟分別相; 2. *paratantra-lakṣaṇa, 依他起相; 3. pariniṣpanna-lakṣaṇa, 成就相; 363b15–17. This set of three cannot be found anywhere else in the translation literature; indeed, the term for parikalpita, alone, is unique, as was noted by the author of T 2809 (T LXXXV 2809 1044a12–16). This pattern suggests a situation in which the translator was improvising equivalents for technical terms for which no settled translation had yet emerged. This may perhaps provide However, this example is certainly not decisive evidence for a Ch source for Tib II on its own, because we could also suppose that the Tib translators specified yid on the basis of their knowledge from other sources that manas was at issue; or that an underlying Skt term (if there was an Indic original) contained *manas, but Ch translators disregarded the distinction between that and *vijñāna; or that variant readings may have existed of an Indic original term.32

some circumstantial support for the traditional ascription of P-Suv (or, perhaps we might now say, the earliest layer of P-Suv) to the earliest period of P’s translation activity. However, this possibility is also complicated by relations between P-Suv and P’s other (later) works, particularly the Saṃdhi-nirmocana-sūtra (解節經 T 677) and (possibly) Wushang.

32. Indeed, it is somewhat difficult in any case to imagine what Indic term might underlie yigenbenxin/kun gzhi la gnas pa’i yid. One is tempted to imagine something like *[[[mūla]]-]āśrita-manas, with a background in something like the passage discussed by Schmithausen 1987, 2:325–326 n. 357, where he reconstructs ‘something like *tatrâlayavijñānam āśrayaḥ, pravṛttivijñānam āśritaṃ — tac ca saptavidhaṃ: cakṣurvijñānaṃ yāvan mano manovijñānaṃ (ca?)’. However, as Schmithausen himself discusses, manas only appears sporadically among the witnesses to this passage, so that it is difficult to be sure it gives support for the identification of manas as one of the parts of mind ‘supported’ (*āśrita) by the ‘basis’ (āśraya); in addition, in Schmithausen’s passage, manas is counted among the *pravṛttivijñānas, which are all said to be *āśrita, whereas here, by contrast, manas seems to be in a separate category of its own. It is also somewhat puzzling that, as Oetke shows, 依根本心 corresponds in Th 509 to len pa’i rnam par shes pa = Skt *ādānavijñāna (cf. Mahāvyutpatti 2018; Ishihama and Fukuda 1989, s.v.); Oetke 1977, 148, 169; cf. also Wilkens 2001, 133, 183. Context here (the placement of the consciousness in question between *mūlavijñāna and some kind of *pravṛttivijñāna), and (circumstantially) the use of yid = manas in Tib, seems clearly to show that the term in question corresponds to the seventh consciousness in the usual eightfold schema. By contrast, in Indic sources, ādānavijñāna is classically an alternate term for ālayavijñāna, viz., the eighth consciousness; Schmithausen 1987, 1:12–13, 23, 49–50, 56, 71–75, 89. If it were indeed ādānavijñāna = ālayavijñāna at issue, however, it would be equally unusual that Tib II has kun gzhi; la gnas pa’i yid (which corresponds perfectly well with the Ch, kun gzhi = genben 根本; gnas = yi 依).

We might, then, consider the possibility that we are dealing with an alternate schema to those surveyed by Schmithausen. In this light, it may be significant that the identification of seventh consciousness with ādānavijñāna was indeed known in China. For example, Jing-ying Huiyuan writes, 八名是何?一者，眼識。二者,耳識。三者，鼻識。四者，舌識。五者，身識。六者，意識。七者 [var. — ]，阿[var. 他]陀那識。八，阿梨耶識, T XLIV 1851 524b29– c2; cf. also the eightfold gloss on ādāna at 524c7–18. This would also seem to be consistent with some passages in the works of ‘P’ (though the authorship of these passages is itself a question that certainly warrants further investigation). For example, in Zhuanshi lun 轉識論: 次明能緣有三種：一，果報識，即是阿梨耶識。二，執識，即阿陀那識。三，塵識，即是六識 … [是名第一識，]依緣此識，有第二執識，此識以執著為體，與四惑相應：一無明、二我 見、三我慢、四我愛 etc., T XXXI 1587 61c8–9, 62a13–20; a similar view seems to be expressed in Foxing lun: 心者即六識心。意者阿陀那識。識者阿梨耶識, T 1610 XXXI 801c14–15; cf. also P MSgBh 180c8–13, 188a29–b2, 207a17–18. However, we should also note that this view seems to be in contradiction with passages in MSg, including P’s translation, which clearly identify ādānavijñāna with the eighth consciousness = ālayavijñāna, and not with the seventh consciousness = kliṣṭamanas: MSg T XXXI 1593 114a13–18, 114b20–22, 114c8–10; and also with other passages in P’s MSgBh T XXXI 1595 157b20–22, 157c11–158a1. This is not the place to fully investigate these issues, but this seems like an important item for future investigation.

It is interesting to note that KTAM 83–84 features something like *mūlacitta and *mūlacittāśrita /*mūlacittāśritya, though in a context that does not feature *pravṛttivijñāna, and which otherwise cannot be straightforwardly equated with the present Suv-trikāya pasasage: ye shes dang ni gnas gyur pa | de gnyis de tshe phan tshun du | gnas kyi dngos por ’dod pa ste | rtsa ba’i sems las rnam rtog bzhin || rtsa ba’i sems la brten nas su | rnam par rtog pa ’byung ba ltar | de bzhin byung tshe de yang ni | de brtas byed par ’dod pa yin || Perhaps we can at least say, in light of the difficulty of identifying an underlying Indic term, that the conjecture of an Indic original here introduces more problems than it solves.

YJ 409a24–26: 一切諸佛 ... 於第三身，與諸佛同體. Tib II 205.11–12: sangs rgyas thams cad kyi ... chos kyi sku ni sangs rgyas thams cad dang sku ’thun pa’o. In this example, it seems most likely that we should understand YJ thus: ‘All the Buddhas ... in respect to the third [[[dharma]]-]body, are identical in essence with all Buddhas’. In other words, as very often in Buddhist Chinese, ti 體 is being used to mean ‘substance, essence’, not ‘body’. (If the authors or translators of the Chinese text had wished to say ‘body’ in this context, we would expect them rather to have used shen 身.) However, Tib II reads ‘... in respect of the dharma body, are equal to all the Buddhas [in] body’ (cf. Nobel 1958, 57 n. 4). The simplest explanation for these textual facts is that the translators of Tib II had a Ch source text, and misread ti to mean ‘body’. The same confusion also seems to underlie the following two examples (4 and 5):

YJ 410b4–6: 如是法界一切妄想不復生故，說為清淨，非是諸佛無其實體. Tib II 210.22–24: de bzhin du chos kyi dbyings la rnam par rtog pa thams cad mi skye ba’i phyir rnam par dag pa zhes bya ste | sangs rgyas rnams kyi yang dag pa’i sku ni med pa ma yin no.

YJ 410b12–14: 此三清淨，是法如如，不異如如，一味如如，解脫如如，究竟如如，是故諸佛，體無有異. Tib II 210.35–211.4: de gsum rnam par dag pa ni chos kyi de bzhin nyid dang | gzhan ma yin pa’i de bzhin nyid dang | ro gcig pa’i de bzhin nyid dang | rnam par grol ba’i de bzhin nyid dang | mthas klas pa’i de bzhin nyid kyi phyir sangs rgyas thams cad ni sku tha mi dad pa’o. Passage 6 YJ 409c13–16: 如來 ... 依 ... 一切自在、一切法平等攝受，如是佛法，悉皆出現. Tib II 207.28–208.2: de bzhin gshegs pa ... dbang thams cad dang mnyam pa nyid kyi chos thams cad ’dzin pa la sogs pa sangs rgyas kyi chos thams cad mngon par ’byung ngo. In this passage, it is not entirely clear where Tib II la sogs pa (‘etcetera’) comes from, but it seems possible that it could come from a misunderstanding of Ch deng 等 in 一切法平等, *sarvadharmasamatā, reading it instead in its alternate meaning of ‘and so on’. However, this example is also not decisive, since it also seems possible that la sogs pa could be an attempt to render (an equivalent to) Ch 如是佛法, ‘Buddha-dharmas such as these ...’.

By contrast, if we imagine a scenario in which there only ever were two texts, Ch and Tib, the relation between the two is extremely straightforward.

YJ 409c17–18: 一百八十不共之法 ... Tib II 208.4: ma ’dres pa’i chos brgya brgyad cu ... In this example, both Ch and Tib II give the very unusual number of ‘one hundred and eighty āveṇikadhārmas’. It is possible, of course, that both derive from an Indic text that featured the same anomaly. However, all other things being equal, it is more economical to assume that they derive from a known common source, namely P (which also features the same peculiar number, 364a16; cf. also the connnection to the Wushangyi jing 無上依經 T 669 examined in Radich 2014, n. 116). Passage 8 P 364b2–3: 世尊！何者為善? 何者不善? 何者正修行而得清淨離於不淨? YJ 410a5–6: 世尊！何者為善? 何者不善? 何者正修，得清淨行?

Tib II 209.3–5: bcom ldan ’das dge ba ni gang mi dge ba ni gang yang dag pa’i bsgrub pa gang bgyis pas rnam par dag pa’i spyod pa ’thob par ’gyur. This is the only one of our examples that might be better explained on the basis of a direct relation between P and Tib II, rather than between YJ and Tib II. (Note that whatever puzzles this may raise, it does not affect the ultimate question of whether Tib II could be a translation from some Ch original, either P itself, or a text derived from it at some remove.) The heart of the problem is Tib bgyis pas. We might read yang dag pa’i bsgrub pa gang bgyis pas rnam par dag pa’i spyod pa ’thob par ’gyur as follows: ‘... by means of effecting (bgyis pa) what right practice might the practice of purification be achieved?’ However, in both Ch and in Tib, it is also natural to read the sentence up to that point as a series of three simple questions: ‘What is the good? What is the bad? What is right practice?’ (In YJ, this is more or less the only viable way to read.) This leaves bgyis pas dangling. We might alternatively explain Tib II by supposing that the translators had some Ch like that of P before them, and read, continuing the four-four-four prosody with which the sentence begins, 何者為善、何者不善、何者正修、行而得清淨, so that bgyis pas is their attempt to represent xing er 行而. Passage 9 YJ 409c12–13: 是故如來常住自在安樂清淨.

Tib II 207.27–29: de’i phyir de bzhin gshegs pa rtag par bde ba dang rnam par dag pa la mnga[’] brnyes shing bzhugs pa’o. Context here (for which, and further discussion, see below p. 261) clearly shows that what is at issue in this passage is a somewhat modified version of the ‘four inversions’, in the particular sense proper to tathāgatagarbha texts, that is: the Tathāgata is characterized by permanence (nitya), bliss (sukha), self (ātman) and purity (śuddhi). In the variant on this rubric seen here, instead of ‘self’ we encounter zizai 自在/mnga’ brnyes (something like *vaśita, aiśvarya). We might therefore roughly translate YJ, ‘Therefore, the Tathāgata is eternal, autonomous, blissful and pure’. Once more, however, Tib varies, and the most obvious explanation for the Tib reading is that the translators had this Ch before them, but misread it; parsing: 如來 [[[subject]]] 常 [adverb] 住 [verb] 自在 [[[verbal]] noun, ‘mastery’] 安樂清淨 [[[locative]]], they translated: ‘Therefore, the Tathāgata constantly dwells/ remains in mastery over bliss and purity’. In this case, it is difficult to imagine a single underlying Indic prototype from which both Ch and Tib could be derived, and context seems clearly to show that YJ’s more straightforward reading is the right one. The most economical explanation for Tib II, therefore, is that it was based upon a misreading of a Ch source text.

YJ 408c20–22: 復次，菩薩摩訶薩入無心定，依前願力，從禪定起，作眾事業，(如是二法無有分別，)自在事成. Tib II 203:16–19: gzhan yang byang chub sems dpa sems dpa chen po ting nge ’dzin du snyoms par zhugs pa las sngon gyi smon lam gyi dbang gis | bsam gtan de las bzhengs te | don dang phrin las thams cad mdzad pa ltar chos rnam pa gnyis kyang rnam par rtog pa med par | don thams cad mdzad pa la mnga[’] brnyes pa’o. [Translating from YJ:] ‘When the bodhisatva-mahāsattva enters into the samādhi free of mentation, in virtue of the power of his prior vow, when he arises from [that] samādhi, he accomplishes all the acts [of a Buddha]; (in this manner, the two dharmas are not distinct from one another, and thereby) mastery over action is realized’. It seems fairly clear from context here that YJ’s shiye 事業 is a single term, representing Skt karma, ‘action’; comparison to other texts suggests that the type of ‘mastery’ in question is thus karmavaśitā, ‘mastery over action [for the benefit of all sentient beings]’. Tib II, however, reads ‘mastery over ends (*artha) and action (don dang phrin las)’, and then ‘mastery over the achievement of all ends (don thams cad mdzad pa)’. Particularly in the former instance, I propose that we might explain Tib II by supposing that the translators had the Chinese before them, and incorrectly parsed shiye as two separate terms, reading shi 事 in an alternate meaning of ‘thing’ (= *artha = don) and ye 業 alone as corresponding to *karma/*kriyā. However, again, the evidence of this passage is not entirely unambiguous, because it is also true that the idea of ‘achieving all ends’ (sarvārthakaraṇa) can be part of discussion of the type of ‘mastery’ in question.37

In sum, in these various passages, Tib II-trikāya evinces pecularities which might be explained by the supposition that it was translated from Chinese. I do not claim that any of these passages alone, nor even all in combination, incontrovertibly prove that the source text of Tib II-trikāya must certainly have been Chinese. However, in combination with the unusually close match between Tib and Ch in this chapter, the unusually close match between P and YJ, the absence of any Skt fragments, and the fact that P is thought in other instances to have possibly been associated with the production of new sūtras (or parts thereof), I suggest that this evidence should certainly persuade us to entertain the hypothesis that Tib II-trikāya might have been produced from a Chinese source text, derived at some remove from P, rather than from Skt. This means that ‘P-Suv-trikāya’ could be a ‘sole exemplar’, from which all other known witnesses are derived. This, in turn, means that we should consider certain possibilities about the provenance of the chapter, the contexts in which it might have been produced, the nature of its association with the rest of Suv, and the nature of its relation to certain other texts. I will return to these further implications below. The case of Tib II-yikong is more difficult to evaluate.38 As with Suv-trikāya, YJ depends very closely upon P, and Tib II usually matches YJ better than P. The match between YJ and Tib II sentence-by-sentence is overall very good. However, my general impression is that there are also somewhat more minor discrepancies in details of wording between YJ and Tib II than are found in Suv-trikāya. On a cursory examination, with one possible exception,39 I have also found no places, like

37. See once more the Madhyānta-vibhāga-bhāṣya passage cited in the note preceding (n. 35). 38. So far as I can ascertain, Nobel 1958, for instance, notes no discrepancies that would help us decide whether this chapter could be from Ch.

YJ 425a24–b1: 善女天！云何五蘊能現法界? 如是五蘊不從因緣生。何以故? 若從因緣生者，為已生故生? 為未生故生? 若已生生者，何用因緣? 若未生生者，不可得生。何以故? 未生諸法即是非有，無名無相 ... Tib II 266.15–22: rigs kyi lha mo phung po lnga la chos kyi dbyings su ji ltar gdags zhe na | phung po lnga de ni rgyu dang rkyen gyis ma bskyed pa ste | de yang gal te rgyu dang rkyen las skye ba zhig yin na ni skyes zin pa skye ’am mi skye ba skye zhes bya | gal te skyes zin pa skye na ni rgyu dang rkyen kyang mi dgos pa’i rigs so | gal te ma skyes pa skye na ni skye ba dmigs su med par ’gyur ro | de ci’i phyir

zhe na | gang ma skyes pa’i chos rnams ni yod pa ma yin pa nyid pas ming med pa mtshan med pa ... As I read it, this passage is explaining that the five skandhas cannot be the products of causes and conditions, because things have to either have come into being already, or be yet to come into being. The text then considers the very moment of their arising. After it dismisses the former alternative, YJ says, ‘If [on the other hand] it is [a matter of] something that has not yet come into being coming into being, [then] it cannot come into being 若未生生者，不可得生’, because dharmas that have not yet come into being do not exist (etc.; in other words, the thing in question would be both present and future, which is an impossible contradiction).

The key, for our present purposes, is 不可得生, in which I read bu ke de 不可得 as an ordinary Ch phrase meaning ‘it is not possible [for it to]’. However, Ch bukede is also a quasitechnical term in Buddhist Chinese (especially common in Prajñāpāramitā-like contexts like the present chapter), corresponding to Skt anupalabdhi. This term can mean that something is impossible, but it usually has the implication that something cannot be known or apprehended, no matter how hard one searches for it; that is to say, that it is not available as an object of cognition (upalambha, upa/labh). The usual Tib equivalent for this term is some variant on ma dmigs pa (cf. Mahāvyutpatti 948, 4461, 971, Ishihama and Fukuda 1989, s.v.). The equivalent sentence in Tib here could be translated as something like ‘If it is [a matter of] those observed above in Suv-trikāya, where the Tibetan translators might have been misreading a Chinese original. Closer investigation of this chapter might be a worthwhile task for future research.

The cases of Tib II-dhāraṇī and Tib II-yezhang are much more complicated. At first glance, it seems most unlikely that either chapter is a translation from either Chinese version (P or YJ). For example, in the case of Tib II-dhāraṇī, the ten bodhisatva bhūmis expounded by the chapter are linked to the practice of the ten pāramitās (‘perfections’). As Nobel already observed, there is a major and systematic discrepancy between Tib II and both Chinese versions in the exposition of these perfections — in both Chinese versions, each perfection is said to be based upon five dharmas,40 but in Tib II, each is said to be based upon ten, and the exposition differs accordingly in further details (Nobel 1958, 130 n. 1). Such a large systematic difference would on its own be enough to render an extant Chinese source for Tib II extremely unlikely; in addition, however, the exposition in Tib II differs in numerous other details from both P and YJ.41 For similar reasons, it also seems initially unlikely that Tib II-yezhang is from any of our extant Chinese versions.42

However, in light of various pieces of new evidence examined in my previous study (Radich 2014), it will probably be necessary to thoroughly re-examine these chapters of Tib II for possible relations not only to Chinese versions of Suv, but also for relations to the other texts upon which these two chapters are based, viz., BDBh and the *Karmāvaraṇa-pratisrabdhi-sūtra (KAP). First, it is striking that these are precisely the two chapters with the most extensive debts to entire previous texts — indeed, they could justifiably be called alternate versions of BDBh and KAP. Moreover, in both cases, independent translations of BDBh and KAP exist in Tibetan, which could possibly have been used as a basis upon which to modify or even ‘correct’ Suv.43 In addition, however, we have also noted above that discrepances between the Shōgozō 聖語蔵 manuscript and Baogui versions of the text demonstrate that parts of P-Suv may have undergone revision, and I argued on the basis of indicators of a broadly ‘Sui’ style that this could also be true of other parts of our extant P-Suv in B. We should therefore also consider the possibility that more Chinese versions of the text may once have existed than we now

something that has not yet come into being coming into being, [then] that would be [an] inapprehensible coming-into-being (skye ba dmigs su med par ’gyur)’. In other words, it looks to me possible that the Tib translators, with Ch 不可得生 in front of them, misread it as a modifier-headword nominal construction: 不可得 + 生.

40. This is also one of the unusual features of the path structure expounded by the chapter pointed out by Ulzii 2011. Nobel notes that the manner in which the ten bhūmis are discussed in this chapter and in the ‘Trikāya’ chapter is very different, and even contradictory; Nobel 1958, 67–68 n. 7, 68–69 n. 4. 41. See, for example, Nobel 1958, 124 n. 2, 127 n. 4, n. 7, 128 n. n. 1, n. 5, n. 6, n. 7, 129 n. 1, n. 2, n. 4, n. 5. 42. For example, a paragraph in P/YJ is missing from Tib II, Nobel 1958, 102 n. 3; a paragraph in Tib II is shorter than P/YJ and differs from them in details, 103 n. 5. Cf. also the long Suvyezhang passage interpolated into Uyghur Suv-trikāya, used by Wilkens as evidence that the translators must have used Tib II as well as YJ as their source text (Tib II 232.6–235.19). This passage is found only in Tib II, without any parallel in YJ; Wilkens 2001, 30. 43. Kimura 2008, 2 states that there is no Tib (or Skt) corresponding to BDBh. However, the Tib Ratnakūṭa does in fact contain an *Akṣayamatiparipṛcchā, Q760(44)/D88. According to the Derge colophon, this translation was the work of Ye shes sde and Surendrabodhi; KAP is found in Tib at Q885/D219, and attributed to Jinamitra, Dānaśīla, and Ye shes sde. This places both texts in the same milieu as Tib II. possess, and that discrepancies between Tib II and P or YJ may reflect a different underlying Ch source text, rather than necessarily indicate that Tib II’s source was indeed Skt. Finally, it is also striking that in both chapters, the Buddha’s main interlocutor differs from P-Suv (and YJ), but in each case, agrees with the source text behind the chapter. In Ch Suv-yezhang, it is Śakra who poses the main question of the chapter to the Buddha, but in Tib II, it is Śāriputra, exactly as in KAP; similarly, in Ch Suv-dhāraṇī, the Buddha’s main interlocutor is something like *Siṃhadhvajāpratihataprabhaṃkara (Nobel’s Skt; 師子相無礙光焰菩薩); but in Tib

II it is Akṣayamati (bLo gros mi zad pa), exactly as in BDBh (Nobel 1958, 124 n. 2). The complexity of the relations between Tib II-yezhang and –dhāraṇī and P-Suv, YJ-Suv, and other possible source texts in Chinese and Tibetan would carry me too far beyond the task I have assigned myself in the present study. However, in light of the new sources of the chapters identified in my previous study (Radich 2014), and these other telling details, it seems best for the interim to conclude that the sources of these four chapters in Tib II should be thoroughly re-examined; and for the present, to remain open to the possibility that they, too, might ultimately be derived from Chinese source texts, even if by processes more complex than those we might imagine for Suv-trikāya and –yikong. More generally, this requires us also to consider the possibility that Tib II as a whole is a hybrid work, drawing upon multiple sources, even if it is true that chapters aside from the four considered here were indeed translated from Sanskrit originals.

Thus, in sum, a range of evidence suggests that Tib II-trikāya might be translated from a Chinese source text, or at least, that this possibility cannot be ruled out. Although there is less positive evidence, the same may also be true of Tib II-yikong; and we cannot even rule the possibility out for Tib II-yezhang and -dhāraṇī. It is therefore possible that P-Suv is a ‘unique exemplar’, in the sense that all other extant versions might derive from it at some remove. The Tibetan evidence therefore does not preclude the possibility that P-Suv may have been composed in China, and indeed, in the case of P-Suv-trikāya, at least, even reinforces that likelihood — strangely enough, given that the dependence of P-Suv-trikāya on the *Kāya-trayāvatāra-mukha-śāstra (KTAM) constitutes the strongest evidence for a direct connection between any part of P-Suv and Indic source materials. Implications In closing, I will attempt to lay out some possible implications of the findings of this study as a whole. Implications for the origins and nature of Suv–trikāya

Suv-trikāya exhibits a number of features that may be significant, when viewed in light of the possibility that the chapter is a ‘sole exemplar’ text, and possibly composed (or composed in part) by P. The most significant fact about the chapter is that some of its content is reminiscent of themes also seen elsewhere in the P corpus. As Hamano has already noted, whereas KTAM expounds trikāya doctrine in a straightforwardly Yogācāra manner, P-Suv-trikāya blends that exposition with elements of tathāgatagarbha doctrine. First, Suv-trikāya loosely identifies the dharmakāya with tathāgatagarbha: P 364a2–3: 若了義說，是身即是大乘、是如來性、是如來藏 ... YJ 409c4–5: 若了此義，是身即是大乘、是如來性、是如來藏 ...

Tib II 207.12–14: don de mngon par shes na sku de nyid theg pa chen po’o | de nyid de bzhin gshegs pa’i ngo bo nyid do | de bzhin gshegs pa’i snying po ste ... ‘If someone understands the intent [of this doctrine] correctly, this very body [viz. dharmakāya] is the greater vehicle (mahāyāna); [this body] is the essence of the Tathāgata (*tathāgatasvabhāva); [this body] is tathāgatagarbha …’ In addition, the chapter also features the ‘four inverted’ virtues of the Tathāgata (slightly modified in P only ), viz., eternity, bliss, [true] self, and purity. These ‘four inversions’ are a well-known feature of tathāgatagarbha doctrine. We have already encountered this passage above (Passage 9, p. 256): Passage 9

P 364a9–11: 如此法身，依於自體說常、說實，依大三昧故說於樂，依於大智故說清淨；是故如來常住、自在、安樂、清淨. YJ 409c11–13: 如此法身，依於自體說常、說我，依大三昧故說於樂，依於大智故說清淨，是故如來常住、自在、安樂、清淨. Tib II 207.24–29: chos kyi sku ’di ni rang gi ngo bo nyid la brten pas rtag pa zhes bya’o | bdag ces bya’o | ting nge ’dzin chen po la brten pas na bde ba zhes bya’o | ye shes chen po la brten pas na rnam par dag pa zhes bya ste | de’i phyir de bzhin gshegs pa rtag par bde ba dang rnam par dag pa la mnga[’] brnyes shing bzhugs pa’o.

[Translating from YJ:] ‘With reference to its essence (*svabhāva), it is taught that this dharmakāya is eternal, it is taught [that it is] self; with reference to [its] great samādhi(s) it is taught that it is blissful; with reference to [its] great wisdom (*jñāna) it is taught that it is pure. Therefore is the Tathāgata eternal, masterful, blissful and pure’. If, then, as I have argued, there is a possibility that Suv-trikāya was composed by P himself, these details, while relatively few and imprecise in nature, further support that possibility, by suggesting that the content of the chapter echoes doctrinal developments to which P is also known to have been connected. Further investigation of this possibility is an important task for future research. Implications for the textual history of Suv The possibility that the four P-Suv chapters were composed in China obviously might also require revision of some aspects of prior discussions of the textual history of Suv itself.

Implications for the history of Buddha-body doctrine In light of the possibility that P may actually have aut hored Suv-trikāya, and introduced his own characteristic blend of Yogācāra and tathāgatagarbha thought in the process, it seems somewhat ironic that the study of Suv-trikāya has an unusually long history in Western Buddhology, and was the target of possibly the first attempt in Western scholarship to characterise Buddha-body doctrine, namely, Schmidt’s pioneering attempt to translate the chapter in 1832. However, since then, the only serious scholars I know of who have paid close attention to the content of the chapter are Takasaki, Hamano and (indirectly, in a sense, because focusing on the much more elaborate Uyghur version) Wilkens.

In any case, even though the present paper cannot conclusively prove that Suv-trikāya was authored by P or in China, the mere possibility that this is so urges caution in using the contents of the chapter as evidence for any development of Buddha-body doctrine in India. The obvious exception to this caution is the portions that overlap with KTAM, but in that case, we may as well ‘cut out the middle man’ and use KTAM itself as our evidence. Conversely, however, the possibility that Suv-trikāya may have been authored by P and/or in China may increase its interest for the history of Buddha-body doctrine in China, especially as it may well have been the first text to introduce a systematic exposition of trikāya doctrine to a Chinese readership, and may thereby have been particularly influential in forming Chinese understandings of the doctrine. Implications for the study of Tibetan translation texts

I argued above that Tib II, especially Tib II-trikāya, may have been translated from a Chinese source. If this suggestion is correct, this would be, to my knowledge, the first known instance in which a translation turned out to be based upon a Chinese and not a Sanskrit source, against the information of the Tibetan bibliographic tradition. By the same token, it is easy to imagine reasons that various parties in the construction of the Tibetan tradition (and modern scholarship) might want to insist upon the Indic provenance of as many canonical texts as possible — as part of the larger project to align Tibetan Buddhism with Indian Buddhism and secure its ‘authenticity’. Thus, the present instance might suggest that further investigation should be directed towards the identification of other possible similar cases.

Implications for the Paramārtha corpus

As I mentioned at the outset of this paper, the external evidence in support of the ascription of P-Suv to P is extraordinarily strong. However, we can now see that the information given to us by the tradition about the text is unlikely to be entirely accurate. According to the tradition, P was motivated to translate these chapters because they were missing from DhKṣ’s earlier Suv (T 663). However, in this study and the prior study it complements (Radich 2014), I have suggested that given the extensive pattern of tangled and close connections to prior Chinese texts, it is more likely that someone in fact composed the chapters, or significant portions thereof, in China, largely on the basis of pre-existing Chinese translations of other texts; and, further, that whether or not this original author was P (as still seems likely), the texts were further modified by someone slightly later. We should consider what this might suggest more generally about the information in the catalogues on P’s corpus as a whole. We already have another famous example of a text in P’s corpus, for which the ascription to P is no longer regarded as authentic, which nonetheless bears a highly detailed preface ascribing it to P: namely, the Mahāyāna Awakening of Faith (AF). Yet, as Chen Yinque has shown, the AF preface, supposedly by Zhikai 智愷 (Huikai 慧愷, d.u.), contains some detail that is historically accurate (Chen 1948). In this, it resembles the present Sengyin preface to some extent, which also contains very concrete-sounding details (though I cannot vouch for the historical veracity of those details).

Notably, both of these texts are ascribed by the bibliographic tradition — especially Fei Changfang — to the early period of P’s activity, under the Liang 梁 (that is, down to about 555 or 556). When we expand our view to the full set of (extant) texts ascribed to this early period, they make a striking group: P-Suv, AF, the Niepan jing ben you jin wu jie lun (涅槃經本有今無偈論 T 1528), the Rushi lun fan zhinan pin (如實論反質難品 T 1633), the Samaya-bhedoparacana-cakra (十八部論 T 2032), and possibly the Wushangyi jing 無上依經 T 669 (hereafter ‘Wushang’)

(according to Zhisheng 智昇, who bases himself upon a postface).

Thus, if the present analysis shows that P-Suv was not exactly a pure ‘translation’, the texts ascribed to this early period of P’s ‘translation’ activity are all problematic in one way or another (with the possible exception of the Rushi lun fan zhinan pin). We should therefore consider the somewhat radical possibility that P may not actually have translated any texts at all — at least not ‘from scratch’ — during this early period. Indeed, to some extent, this early period might have represented a convenient hole in P’s vita, into which problematic texts could expediently be plugged, and thereby lent legitimacy. Another important implication also follows for our understanding of the P corpus. Before I stumbled upon the first hints that led me to the evidence I study here, I was inclined to regard P-Suv as one of a handful of texts in the P corpus for which the ascription was most secure. In addition to the unusually rich external evidence, this is a text that appeared (thanks mainly to the existence of an almost identical supposed independent witness in Tib II) to be a clear translation, with little or no interpolated material. In this, it contrasted with texts like MSgBh, Jueding zang lun (決定藏論 T 1584), Foxing lun (佛性論 T 1610) or Zhuanshi lun (轉識論T 1587), which may well incorporate ‘lecture’ material, that is, records of oral expositions that were given alongside the translation process (Funayama

2006b, 50–55). Most importantly, there was no known prior translation in China (by contrast to texts like the Vajra-cchedikā prajñā-pāramitā 金剛般若波羅蜜經 T 237), which should have meant that the style of the text was unadulterated P. On this basis, I was in the habit of including P-Suv among a handful of core texts in the P corpus that are similarly fairly ‘straight’ translations, for which there exists similarly strong external evidence for P’s authorship (Radich 2014, n. 142). In attempts to analyse questions of authorship on the basis of P’s stylistic traits, these texts comprised my benchmark. If a possibly spurious ascription of a composite text could in this manner lie concealed in the heart of the P corpus, as it were, then this suggests that more than ever, we must carefully reassess problems of authorship in that corpus from the ground up.

Finally, it is also worth noting that the present study urges that we reassess the nature and contents of Wushang in particular. Wushang is a composite text, which scholars have long suspected might have been composed in China (Takasaki 1966, 49–52). Shimoda has shown that Wushang uses material originally referring to relics from the Adbhuta-dharma-pariyāya to book-end material about tathāgatagarbha from the Ratna-gotra-vibhāga (Shimoda 1997, 85–86). Takasaki has also studied in detail some intricate and fascinating relationships between the structure and content of Wushang and that of the Suvikrānta-vikrāmi-paripṛcchā prajñā-pāramitā 勝天王般若波羅蜜經 T 231 ascribed to *Ūrdhvaśūnya 月婆首那, as well as the Ratna-gotra-vibhāga (Takasaki, 1988). In the present study, we have also seen, at numerous points, that Wushang shares many interesting and unusual features with P-Suv, including some concrete ideas and passages, the unusual number of 180 āveṇikadharmas, diction apparently more characterstic of the Sui, and a colophon dating it to the same period of P’s activity.59 We should thus consider the possibility that the two texts may have originated in similar circumstances, including the possibility that Wushang, too, is somehow connected to a milieu closer to the Sui.

Acknowledgements

I would like to express my gratitude for help with this paper to Jamie Norrish, Prof. Funayama Tōru 船山徹, Ching Keng, Prof. Jan Nattier, Prof. Jonathan Silk, Prof. Leonard van der Kuijp, Arthur McKeown and Ryan Overbey. I also thank the two anonymous reviewers. Abbreviations

AF The Mahāyāna Awakening of Faith 大乘起信論 T 1666 BDBh *Bodhisatva-daśa-bhūmika-sūtra, esp. T 307 (Kumārajīva), T 308 (Jijiaye 吉 迦夜) CBETA Chinese Buddhist Electronic Text Association

59. See Table 8 in Radich 2014, various phrases and terms in Table 9 in Radich 2014 (esp. 於意云 何); and Radich 2014, ns. 102, 109, 112, 116, 117, 146 (phrases marked ‘cf. Wushang’); above n. 31; Passage 7; and p. 264. Note also this parallel: 百分千萬億分，乃至算數譬喻所不能及, T XXIV 1491 1089b5–6, 1089b17–18 (twice); Wushang T XVI 669 469a23–b1: 造塔如阿摩羅子大。戴剎如針大。露槃如棗葉大。造佛如麥子大。此功德於前所說。百分不及一。千萬億分不及一。乃至算數譬喻所不能及。阿難。若此功德不迴向阿耨多羅三藐三菩提 (cf. Radich 2014, n. 146)。此功德聚所獲福報。盡娑訶世界微塵數。作他化自在天王化樂天王兜率陀天王夜摩天王三十三天王 (cf. Radich 2014, n. 146)。況復轉輪聖王.

Ch Chinese D Derge edition of the Tibetan Tripiṭaka DhKṣ *Dharmakṣema 曇無讖 IBK Indogaku Bukkyōgaku kenkyū 印度学仏教学研究 KAP *Karmāvaraṇa-pratisrabdhi-sūtra, esp. T 1491 (Jijiaye 吉迦夜), T 1493 (*Jñānagupta) KTAM *Kāya-trayāvatāra-mukha-śāstra = sKu gsum la ’jug pa’i sgo zhes bya ba’i bstan bcos, D3980/Q5290. Asian Classics Input Project electronic text (Lhasa, bstan ’gyur, ha 1a–8a). MSg Mahāyāna-saṃgraha 攝大乘論 T 1593 MSgBh Mahāyāna-saṃgraha-bhāṣya 攝大乘論釋 T 1595 P Paramārtha 真諦 P-Suv Chapters of Suv ostensibly translated by P P-Suv-dhāraṇī Suv Ch. 6, ‘On the Utterly Pure Bhūmis of the Dhāraṇīs’ 陀羅尼最淨地品, ascribed to P P-Suv-trikāya Suv Ch. 3, ‘On Distinctions among the Three Bodies’ 三身分別品, ascribed to P P-Suv-yezhang Suv Ch. 5, ‘On the Extinction of Karmic Obstructions’ 業障滅品, ascribed to P P-Suv-yikong Suv Ch. 9, ‘On Fulfilling Vows in Reliance upon Emptiness’ 依空滿願品, ascribed to P Q Peking version of the Tibetan Tripiṭaka Skt Sanskrit Suv Suvarṇa-prabhāsottama-sūtra 金光明經: T 663 (DhKṣ); T 664 (Baogui), T 665 (YJ) Suv-dhāraṇī The ‘Dhāraṇīs’ chapter of Suv, irrespective of version Suv-trikāya The ‘Three Bodies’ chapter of Suv, irrespective of version Suv-yezhang The ‘Karmic Obstructions’ chapter of Suv, irrespective of version Suv-yikong The ‘Fulfilling Vows in Reliance upon Emptiness’ chapter of Suv, irrespective of version T Taishō shinshū daizōkyō 大正新脩大藏經, as accessed via CBETA (2011) References to the Taishō follow the order: volume number (roman numerals), text number (arabic numerals), page, column and line number. Thus e.g. T VIII 225 483b17 is Volume 8, Text no. 225, page 483, second register, line 17. Tib Tibetan

Tib II Tib trans. of Suv ascribed to Jinamitra, Śīlendrabodhi and Ye shes sde D556/Q175 (see Nobel 1944: xiii) Tib II-dhāraṇī The ‘Dhāraṇīs’ chapter in Tib II Tib II-trikāya The ‘Three Bodies’ chapter in Tib II Tib II-yezhang The ‘Karmic Obstructions’ chapter in Tib II Tib II-yikong The ‘Fulfilling Vows in Reliance upon Emptiness’ chapter in Tib II Tib III Tibetan translation of YJ Suv by Chos grub, D555/Q174 (see Nobel 1944: xiii) Tib IV Second, independent translation of YJ Suv into Tib, only extant in fragments, represented primarily by some variant canonical readings, and (Dunhuang manuscripts) Th 507, Th 508, Th 509, and Th 515 (see Oetke 1977) Tib V Third, independent translation of YJ Suv into Tib (see Oetke 1977) Wushang Wushangyi jing 無上依經 T 669 X Shinsan dai Nippon zokuzōkyō 卍新纂大日本續藏經. References formatted as for T

YJ Yijing 義淨, especially his translation of Suv, T 665 Bibliography CBETA (Chinese Buddhist Electronic Text Association). 2011. Taishō shinshū daizōkyō 大正新脩大藏經. Edited by Takakusu Junjirō 高楠順次郎 and Watanabe Kaigyoku 渡邊海 旭. Tokyo: Taishō shinshū daizōkyō kankōkai/Daizō shuppan, 1924–1932. CBReader v. 3.10. Chen, Yinque 陳寅恪. 1980. ‘Liang yi Dasheng qi xin lun wei Zhikai xu zhong zhi zhen shiliao 梁譯大乘起信論偽智愷序中之真史料 [1948]’. In Chen Yinque wenji zhi san: Jinming guan cong gao er bian陳寅恪文集之三　金明館叢稿二編, 147–152. Shanghai: Shanghai guji chubanshe. D’Amato, Mario. 2012. Maitreya’s Distinguishing the Middle from the Extremes (Madhyāntavibhāga), Along with Vasubandhu’s Commentary (Madhyāntavibhāga-bhāṣya): A Study and Translation. The American Institute of Buddhist Studies, Columbia University’s Center for Buddhist Studies and Tibet House US. New York: Columbia University Press. Demiéville, Paul. 1929. ‘Sur l’authenticité du Ta tch’eng k’i sin luoen’. Bulletin de la maison français-japonais 2(2): 1–78. Reprinted in Choix d’études bouddhiques 1929–1970, 1–79.

———. 1952. Le concile de Lhasa: Une controverse sur le quiétisme entre bouddhistes de l’Inde et de la Chine au VIII. siècle de l’ère chrétienne. Paris: Impr. nationale de France. Funayama, Tōru 船山徹. 2006a. ‘Kan’yaku to chūgoku senjutsu no aida: Kanbun Butten ni tokuyū na keitai wo megutte’ 「漢訳」と「中国撰述」の間──漢文仏典に特有 な形態をめぐって [Between Chinese Translation and Chinese Composition: On a Particular Configuration in Chinese Buddhist Scriptures]. Bukkyō shigaku kenkyū 仏教史学研究 45(1): 1–28. ———. 2006b. ‘Masquerading as Translation: Examples of Chinese Lectures by Indian Scholar-Monks in the Six Dynasties Period’. Asia Major 19(1–2): 39–55. Funayama seminar. Unpublished. ‘Shintai Kon kōmyō kyō so itsubun’ 眞諦『金光明疏』佚文. Materials prepared for research seminar led by Funayama Tōru, 2005-2011, entitled ‘Shintai sanzō to sono jidai’ 真諦三蔵とその時代, Kyoto University. Habata, Hiromi. 2007. Die zentralasiatischen Sanskrit-Fragmente des Mahāparinirvāṇamahāsūtra: Kritische Ausgabe des Sanskrittextes und seiner tibetischen Übertragung im Vergleich mit den chinesischen Übersetzungen. Marburg: Indica et Tibetica Verlag. Halkias, G. 2004. ‘Tibetan Buddhism Registered: A Catalogue from the Imperial Court of ’Phang Thang’. The Eastern Buddhist 36: 46–105. Hamano Tetsunori 浜野哲敬. 1984. ‘Sanjin nyūmon ron ni tsuite 『三身入門論』について’. IBK 32(2): 162–163.

———. 1985. ‘Kon kōmyō kyō ‘Sanjin funbetsu bon’ ni tsuite 『金光明経・三身分別品』について’. IBK 33(2): 296–299. Herrmann-Pfandt, Adelheid. 2008. Die lHan kar ma: Ein früher Katalog der ins Tibetische übersetzten buddhististchen Texte. Kritische Neuenausgabe mit Einleitung und Materialien. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Hoffmann, Helmut. 1956. Die Religionen Tibets: Bon und Lamaismus in ihrer geschichtlichen Entwicklung. Freiburg: K. Alber. Ishihama, Yumiko 石濱裕美子 and Fukuda Yōichi 福田洋一. 1989. A New Critical Edition of the Mahāvyutpatti: Sanskrit-Tibetan-Mongolian Dictionary of Buddhist Terminology/新訂翻訳名義大集. Tokyo: Tōyō bunko. de Jong, Jan Willem. 1979. Review of Oetke 1977. Indo-Iranian Journal 21: 300–304. Kanaoka, Shūyū 金岡秀友. 1980 Kon kōmyō kyō no kenkyū 金光明経の研究. Tokyo: Daitō shuppansha. Kawagoe, Eishin 川越英真. 2005. Dkar chag ʼPhang thang ma. Sendai: Tōhoku Indo-Chibetto kenkyūkai 東北インド・チベット研究会. Kimura, Kiyotaka 木村清孝. Jū jū kyō, hoka 十住経・他. 2007. Shin kokuyaku daizōkyo, Indo senjutsu bu, Kegon bu 4 新国訳大蔵経・インド撰述部・華厳部 4. Tokyo: Daizō shuppan.

———. 2008. ‘Kan’yaku Bosatsu jū jū kyō kō: dōitsu keitō no ni kyō to no hikaku kenkyū 漢訳『菩薩十住経』考―同系統の二経との比較研究’. Kokusai Bukkyōgaku daigaku daigakuin kenkyū kiyō 国際仏教学大学院大学研究紀要12: 1–25. Lalou, Marcelle. 1953. ‘Les textes bouddhiques au temps du roi Khri-sroṅ-lde-bcan’. Journal Asiatique 241: 313–353. La Vallée Poussin, Louis de. 1980. L’Abhidharmakośa de Vasubandhu. Brussels: Institut Belge des Hautes Études Chinoises. Ligeti, Lajos. 1942–1944. Catalogue du Kanǰur Mongol imprimé. Bibliotheca Orientalis Hungarica 3. Budapest: Société Kőrösi Csoma. Ludvik, Catherine. 2007. Sarasvatī, Riverine Goddess of Knowledge: From the Manuscript-Carrying Viṇā-Player to the Weapon-Wielding Defender of the Dharma. Leiden: Brill. http:// dx.doi.org/10.1163/ej.9789004158146.i-390 Matsumoto, Shōkei 松本照敬. 2007. ‘Bongo senji mon no gengo hitei 『梵語千字文』の原語比定’. Naritasan Bukkyō kenkyūjo kiyō 成田山仏教研究所紀要 30: 35–101. ———. 2008. ‘Bongo zatsumyō no gengo hitei 『梵語雑名』の原語比定’. Naritasan Bukkyō kenkyūjo kiyō 成田山仏教研究所紀要 31: 55–155. Nagao, Gadjin M. 長尾雅人. 1964. Madhyāntavibhāga-bhāṣya: A Buddhist Philosophical Treatise Edited for the First Time from a Sanskrit Manuscript. Tokyo: Suzuki Research Foundation.

Nattier, Jan. 2010. ‘Re-Evaluating Zhu Fonian’s Shizhu duanjie jing (T309): Translation or Forgery?’ Annual Report of The International Research Institute for Advanced Buddhology at Soka University 13: 256. Nishioka, Soshū 西岡祖秀. 1980, 1981, 1983. ‘Putun Bukkyō-shi mokurokubu sakuin 「プトゥン仏教史」目録部索引 (I);’ ‘~ (II);’ ‘~ (III)’. Tōkyō daigaku bungakubu bunka kōryū kenkyū shisetsu kenkyū kiyō 東京大学文学部文化交流研究施設研究紀要 4: 61–92; 5: 43–94; 6: 47–201. Nobel, Johannes, ed. 1937. Suvarṇabhāsottamasūtra: Das Goldglanz-sūtra, Ein Sanskrittext des Mahāyāna-Buddhismus, Nach den Handschriften und mit Hilfe der tibetischen und chinesischen Übertragungen. Leipzig: Otto Harassowitz. ———. 1944. Suvarṇaprabhāsottamasūtra: Das Goldglanz-sūtra, Ein Sanskrittext des MahāyānaBuddhismus, die tibetischen Übersetzung mit einem Wörterbuch. Leiden: E. J. Brill/ Stuttgart: W. Kohlhammer. ———. 1958. Suvarṇaprabhāsottamasūtra: Das Goldglanz-sūtra, Ein Sanskrittext des MahāyānaBuddhismus, I-Tsing’s chinesische Version und ihre tibetische Übersetzung. Leiden: E. J. Brill.

Oetke, Claus. 1977. Die aus dem Chinesischen übersetzten tibetischen Versionen des Suvarṇaprabhāsasūtra: Philologische und linguistische Beiträge zur klassifierenden Charakterisierung übersetzter Texte. Alt- und Neu-Indische Studien herausgegeben vom Seminar für Kultur und Geschichte Indiens an der Universität Hamburg, 18. Wiesbaden: Franz Steiner Verlag GmbH.

Ono, Genmyo 小野玄妙. 1929. ‘Ryō Shintai yaku Kon kōmyō kyō jobun 梁真諦訳金光明経序文’. Butten kenkyū仏典研究 1, no. 2: 5. (Ono Genmyō 小野玄妙 =) Niryō Sei 二楞生. 1934. ‘Taizō bunko koitsu zenpon mokuroku (1) 大藏文庫古逸善本目録(一)’. Pitaka ピタカ 5: 15–20. Pradhan, P., ed. 1967. Abhidharmakośabhāṣya of Vasubandhu. Patna: K. P. Jayaswal Research Institute. Radich, Michael. 2008. ‘The Doctrine of *Amalavijñāna in Paramārtha (499-569), and Later Authors to Approximately 800 C.E’. Zinbun 41: 45-174. ———. 2012. ‘External Evidence Relating to Works Ascribed to Paramārtha, with a Focus on Traditional Chinese Catalogues’. In Shintai sanzō kenkyū ronshū 真諦三藏研究論集 [Studies of the Works and Influence of Paramārtha], edited by Funayama Tōru 船山徹, 39–102[L]. Kyoto: Kyōto daigaku jinbun kagaku kenkyūjo/Institute for Research in Humanities, Kyoto University.

———. 2014. ‘On the Sources, Style and Authorship of Chapters of the Synoptic Suvarṇaprabhāsottama-sūtra T664 Ascribed to Paramārtha (Part 1)’. Annual Report of The International Research Institute for Advanced Buddhology 17: 207–244. ———. Forthcoming. ‘Problems of Attribution, Style, and Dating Relating to the ‘Great Cloud Sūtras’ in the Chinese Buddhist Canon (T 387, T 388/S.6916)’. In Buddhist Transformations and Interactions: Essays in Honor of Antonino Forte, edited by Victor Mair, Tansen Sen and Chen Jinhua. Sino-Platonic Papers. Rockhill, W. Woodville. 1884. The Life of the Buddha and the Early History of His Order, Derived from Tibetan Works in the Bkah-Hgyur and Bstan-Hgyur, Followed by Notices on the Early History of Tibet and Khoten. London: Trübner. Schmidt, Isaak Jakob. 1832. ‘Über einige Grundlehren des Buddhaismus. I.–2. Abhandlung’. In Mémoires de l’Académie Impériale des Sciences de Saint-Pétersbourg. Sixième Série. Sciences politiques, histoire et philologie. Tome I, 221–262. St-Pétersbourg. [Not seen.]

Schmithausen, Lambert. 1987. Ālayavijñāna: On the Origin and Development of a Central Concept in Yogācāra Philosophy. 2 vols. Studia Philologica Buddhica Monograph Series IVb. Tokyo: The International Institute for Buddhist Studies. Shimoda, Masahiro 下田正弘. 1997. Nehan gyō no kenkyū: Daijōkyōten no kenkyū hōhō shiron 涅槃経の研究―大乗経典の研究方法試論. Tokyo: Shunjūsha. Silk, Jonathan A. Unpublished. ‘Two Chinese Sūtras in Tibetan Translation: The Gaṅgottaraparipṛcchā and the Amituo jing’. Skjaervo, Prods Oktor. 2004. This Most Excellent Shine of Gold, King of Kings of Sutras: The Khotanese Suvarṇabhāsottamasūtra. 2 vols. Central Asian Sources V. Cambridge: Harvard University Department of Near Eastern Languages and Civilizations. ———. 2009. ‘Fragments of the Suvarṇabhāsottamasūtra in Or.15009 and Or.15010’. In The British Library Sanskrit Fragments: Buddhist Manuscripts from Central Asia, edited by Karashima Seishi and Klaus Wille, Volume II.1, 597–635. Tokyo: The International Research Institute for Advanced Buddhology, Soka University.

Stein, Rolf A. 1983. ‘Les deux vocabulaires des traductions indo-tibétaine et sino-tibétaine dans les manuscrits de Touen-houang’. Bulletin de l’École Française d’Extrême-Orient 72: 149–326. Translated into English as: ‘The Two Vocabularies of Indo-Tibetan and Sino-Tibetan Translations in the Dunhuang Manuscripts. Tibeta Antiqua I’. In Rolf Stein’s Tibeta Antiqua, with Additional Materials, translated and edited by Arthur P. McKeown, 1–96. Leiden, Boston: Brill, 2010. Suzuki, D. T. 1999. Studies in the Laṅkāvatāra Sūtra. Delhi: Motilal Banarsidass. [First edition: London: 1932.] Takasaki, Jikidō 高崎直道. 1966. A Study on the Ratnagotravibhāga (Uttaratantra), Being a Treatise on the Tathāgatagarbha Theory of Mahāyāna Buddhism. Rome: Serie Orientale Roma.

———. 1974. Nyoraizō shisō no keisei: Indo daijō Bukkyō shisō kenkyū 如来蔵思想の形成―インド大乗仏教思想研究. Tokyo: Shunjūsha. ———. 1976. ‘Some Problems of the Tibetan Translations from Chinese Materials’. In Proceedings of the Csoma de Kőrös Memorial Symposium, Held at Mátrafüred, Hungary, 24–30 September 1976, edited by Louis Ligeti, 459–467. Bibliotheca Orientalis Hungarica 23. Budapest: Akademiai Kiado. ———. 1988. ‘Mujōe kyō to Shōtennō hannya 『無上依經』と『勝天王般若』’. Naritasan Bukkyō kenkyūjo kiyō 成田山仏教研究所紀要 11: 171–194. Takeda, Kagaku Shinkō Zaidan Kyōu Shooku 武田科学振興財団杏雨書屋, ed. 2010. Tonkō hikyū eihensatsu: Kyōu shooku zō 3 敦煌秘笈　影片冊　杏雨書屋蔵 3. Osaka: Takeda Kagaku Shinkō Zaidan 武田科学振興財団.

Tucci, Guiseppe. 1986. Minor Buddhist Texts, Parts One and Two. Delhi: Motilal Banarsidass. [[[Wikipedia:Rome|Rome]]: IsMAO, 1956, 1958]. Ueyama, Daishun 上山大峻. 1967-1968. ‘Daibankoku daitoku sanzō hosshi shamon Hōjō no kenkyū 大蕃国大徳三蔵法師沙門法成の研究’. Tōhō gakuhō 東方学報 38: 133–198; 39: 119–222.

Ueyama, Daishun 上山大峻. 1990. ‘Daibankoku daitoku sanzō hosshi Hōjō no hito to gyōseki 大蕃国大徳三蔵法師法成の人と業績’. In Ueyama, Tonkō Bukkyō no kenkyū敦煌佛教の研究, 88–246. Kyoto: Hōzōkan. Ulzii, Jargal ウルジ・ジャルガル. 2011. ‘Kon kōmyō kyō no kenkyū: “Saijō ji darani bon” no kōzō ni tsuite『金光明経』の研究―「最浄地陀羅尼品」の構造について’. IBK 59(2):

124–127[L]. Wilkens, Jens. 2001. Die drei Körper des Buddha (Trikāya): das dritte Kapitel der uigurischen Fassung des Goldglanz-Sūtras (Altun Yaruk Sudur): eingeleitet, nach den Handschriften aus Berlin und St. Petersburg herausgegeben, übersetzt und kommentiert. Berliner Turfantexte 21. Berlin: Kommission Turfanforschung der Berlin-Brandenburgischen Akademie der Wissenschaften/Brepols. Yuyama, Akira. 2004. ‘The Golden Light in Central Asia’. Annual Report of The International Research Institute for Advanced Buddhology at Soka University 7: 3–