A Preliminary Report on the Sanskrit Manuscript of the Uttaragrantha of the Mūlasarvāstivāda Vinaya

Abstract

The discovery of the Schøyen–Virginia manuscript of the Uttaragrantha provides significant insights into the Mūlasarvāstivāda Vinaya. This newly identified Sanskrit manuscript offers a fresh perspective on monastic law codes, contributing original Sanskrit terms previously known only through Tibetan and Chinese translations, thereby enhancing our knowledge of Sanskrit–Tibetan–Chinese Vinaya terminologies. Also, by adding itself as a new textual witness to the Mūlasarvāstivāda Vinaya, it demonstrates the complex textual history and underscores the potential multiplicity in the Mūlasarvāstivāda Vinaya traditions or even the “Greater Sarvāstivāda” Vinaya traditions. Variations in chapter sequencing across extant versions of the Uttaragrantha suggest the possibility of the chapters originally existing as independent texts rather than as a collective, the Uttaragrantha. This article presents the latest findings on the Sanskrit manuscript fragments of the Uttaragrantha in the Schøyen Collection and the private collection, Virginia. Furthermore, it attempts to show the role of the S-V manuscript of the Uttaragrantha in improving our textual understanding of the Uttaragrantha and examining the potential multiplicity in the Mūlasarvāstivāda Vinaya traditions.

1. Introduction

The Mūlasarvāstivāda Vinaya, the monastic law code of the Mūlasarvāstivāda, consists of four parts: (1) the two Vinayavibhaṅgas, (2) the Vinayavastu, (3) the Vinayakṣudraka (also known as Kṣudrakavastu), and (4) the Vinayottaragrantha.1 Of these, the Uttaragrantha has been less extensively explored by scholars for various reasons. The primary reason is the limited availability of Sanskrit and Chinese materials related to the Uttaragrantha. Moreover, it was previously considered merely an appendix of the Mūlasarvāstivāda Vinaya, akin to the Parivāra in the Pāli Vinaya. However, this perspective began to shift following Schopen’s 2004 study (Schopen 2004, pp. 124–27, 270; Kishino 2013), which suggested that the Uttaragrantha should not be viewed merely as an appendix but as a text intrinsically integrated with and essential to the Mūlasarvāstivāda Vinaya. Since becoming known to scholars at the end of the last century, Sanskrit manuscripts of the Uttaragrantha in the Schøyen Collection and a private collection, Virginia, have provided significant insights into their historical significance in the Mūlasarvāstivāda Vinaya. This paper aims to present the latest findings on the Sanskrit manuscript fragments of the Uttaragrantha from these two collections. The new findings shed new light on texts previously accessible only through Tibetan and Chinese translations, suggesting their original forms. This enhances research on the Vinaya corpus with original Sanskrit terms and adds a new textual witness to the Mūlasarvāstivāda Vinaya.

2. The Extant Sanskrit Fragments of the Uttaragrantha

2.1. The Uttaragrantha

The Uttaragrantha is the final part of the Mūlasarvāstivāda Vinaya. The Uttaragrantha, viewed by modern scholars as an “appendix” to the Mūlasarvāstivāda Vinaya and likened to the Parivāra of the Pāli Vinaya, has been re-evaluated in recent scholarship. To be specific, the Uttaragrantha is estimated to have been composed around the same time as, or possibly even earlier than, other parts of the Mūlasarvāstivāda Vinaya (Schopen 2004, pp. 124–25).2 This contrasts with the Parivāra of the Theravāda Vinaya, which is considered to have been composed later than the other parts of the Vinaya.3 This re-evaluation is supported by references to the Uttaragrantha in significant Vinaya works by Viśeṣamitra and Guṇaprabha, attesting to its established scriptural authority by the 6th to 7th centuries. Moreover, the existence of translations into Tibetan and Chinese suggests that this text was accepted as an important scriptural source in India, suggesting a reassessment of its role. Despite its historical importance, the Uttaragrantha has received comparatively less scholarly attention, partly due to limited access to the original Sanskrit materials.

While no Indic version of the Uttaragrantha has survived in its entirety, the Tibetan Kanjur includes two translations: the complete ’Dul ba gźuṅ dam pa (D 7[b]/P 1037, 53 bam pos) and the incomplete ’Dul ba gźuṅ bla ma (D 7[a]/P 1036, 11 bam pos).4 It is unknown why these two versions were produced.5 The complete Uttaragrantha, the ’Dul ba gźuṅ dam pa, is thought to have been translated by the early 9th century; unfortunately, the translator’s name is not mentioned in its colophon (Clarke 2015, p. 77).6 The ’Dul ba gźuṅ dam pa consists of ten or eleven chapters depending on whether the Upāliparipṛcchā, which is located at the end of the Uttaragrantha, is counted as one chapter.7 The chapters are Upāli-paripṛcchā, the Vinītakā, the *Ekottarikā, the *Pañcaka, the *Ṣoḍaśaka, the Nidāna, the Muktaka, the Kathāvastu, the Māṇavikā, and the Mātṛkā. Furthermore, Yijing’s Chinese translations of the Uttaragrantha, while not complete, include two chapters: the Nidāna (Nituona 尼陀那) and Muktaka (Mudejia 目得迦) in one text titled Genben shuo yiqie youbu nituona mudejia (根本說一切有部尼陀那目得迦, T. 1452), produced in 703 CE.8

There are also other Chinese texts which are related to the Uttaragrantha. First, the Sapoduobu pini modeleqie 薩婆多部毘尼摩得勒伽 (T.1441 [23]), translated by Saṅghavarman (僧伽跋摩) in 435 CE, is considered relevant to the Uttaragrantha, although it had been misunderstood as a commentary on the Shisong lü. Current scholars (Clarke 2004, 2015; Kishino 2008, 2013) suggest that this text is closely related to the Mūlasarvāstivāda Vinaya, particularly the Uttaragrantha, based on the similarities in structure and content. Second, a certain part of the Shisong lü 十誦律 (T.1435 [23]) corresponds to sections of the Uttaragrantha, even though the Shisong lü itself is not a translation of the Mūlasarvāstivāda Vinaya but a Vinaya belonging to the “Greater Sarvāstivāda”, as termed by Dhammadinnā.9 The Shisong lü, translated between 404 and 409 CE, was the result of a complex process that involved various translators, including Kumārajīva, *Puṇyatāra (Furuoduoluo 弗若多羅), *Dharmaruci (Tanmoliuzhi 曇摩流支), and Vimalākṣa (Beimoluocha 卑摩羅叉). Specifically, Recitations 8–10 of the Shisong lü correspond to the eight chapters, namely the Nidāna, the *Ekottarikā, the Kathāvastu, the Upāliparipṛcchā, the Māṇavikā, the Mātṛkā, the Vinītakā, and the Muktaka.10

The Uttaragrantha deals with a wide range of topics addressed in the Mūlasarvāstivāda Vinaya, which can be categorized as follows: (1) regulations or explanations identical or similar to those found in other parts of the Mūlasarvāstivāda Vinaya; (2) regulations or explanations that are not identical but related to the Vinaya rules previously established in the other parts of the Mūlasarvāstivāda Vinaya; and (3) regulations or definitions of certain items introduced for the first time, although they are not entirely new and they are treated or referenced in other parts of the Mūlasarvāstivāda Vinaya. In terms of structure, the Uttaragrantha predominantly utilizes piṇḍoddānas and uddānas to organize its topics.11 While some chapters, like the Upāliparipṛcchā, *Ekottarikā, *Pañcaka, Nidāna, Muktaka, and Māṇavikā, utilize both systems for thematic division, others, such as the Vinītakā and *Ṣoḍaśaka, employ only the uddāna system. The Kathāvastu and Mātṛkā chapters do not use these systems. The contents of the Uttaragrantha can be broadly categorized into three types: (1) a series of question-and-answer sessions between Upāli and the Buddha, (2) narrative stories in which the Buddha establishes regulations, and (3) chapters that incorporate both elements.

2.2. Sanskrit Uttaragrantha Fragments in the British Library Hoernle Collection (London) and the Turfan Collection (Berlin)

Until the discovery of a substantial portion of the Sanskrit manuscript of the Uttaragrantha in the Schøyen collection and the private collection, Virginia, only fragments, which are detailed in the following, were known to modern scholars, along with citations in later commentaries, such as Guṇaprabha’s Vinayasūtravṛttyabhidhāna-svavyākhyāna.12

The Hoernle Collection includes a total of eight fragments from the Mātṛkā chapter. Three of these fragments belong to the Pratisaṃyuktakhaṇḍaka (Or.15003/178, 15009/195, and 15009/271),13 while the rest belong to another section, the Vṛttakhaṇḍaka (Or.15007/504, 15009/48, 15009/57, 15009/443, and 15011/1).14 The Turfan Collection contains several fragments belonging to the Upāliparipṛcchā, Kathāvastu, Mātṛkā, and Muktaka chapters of the Uttaragrantha, all of which are accessible through the series Sanskrithandschriften aus den Turfanfunden (SHT III 937, SHT V 1068, SHT VIII 1943, and SHT XII 7185).15

2.3. Uttaragrantha Manuscripts in the Schøyen Collection and the Private Collection, Virginia

Academic attention was first drawn to the Uttaragrantha manuscript in the Schøyen Collection around the turn of the century. In 2000, Gregory Schopen mentioned its existence, referencing transcriptions by Klaus Wille (Schopen 2000, pp. 99–100, 140, I. 9). A subsequent publication in 2006 by Richard Salomon also mentioned the presence of the Vinaya and other texts within this intriguing find (Salomon 2006, p. 356). The academic discussion on the Uttaragrantha was enriched in 2009 when Wille detailed the manuscripts in the private collection, Virginia, at a conference held at Stanford University. This presentation was refined and published in 2014, spotlighting fragments from eight different manuscripts, including five Vinaya texts, that is, the Vinayavibhaṅga; Prātimokṣasūtra; Vinayavastu; and two manuscripts of the Uttaragrantha, tentatively called manuscript A and manuscript B (cf. Hartmann and Wille 2014, pp. 145–53 and Yao, forthcoming, 1.1.2. A Brief History of Research on the Bhaiṣajyavastu Fragments).

The Vinayavibhaṅga and the Upāliparipṛcchā of the Uttaragrantha (manuscript B) of the Mūlasarvāstivāda Vinaya were examined by Masanori Shōno (2020a, 2020b, 2021), whereas the study of Uttaragrantha (manuscript A) had been undertaken by various scholars. Wille first transliterated the Uttaragrantha (manuscript A) around 2000 and produced a revised and improved version in 2011.16 After that, scholars including Shayne Clarke, Fumi Yao, Gudrun Melzer, and Jens Borgland continued the study until 2018. Despite only having access to limited materials, which were preserved as bundles with several folios stuck together, scholars produced foundational, yet unpublished works.17 Their works included the transliteration and identification of some accessible fragments at that time. Following these endeavors, several scholars, including myself, have further engaged in the study of the Uttaragrantha fragments found in the Schøyen Collection, with meetings held at the Norwegian Institute of Philology (hereafter PHI) in 2019, 2020, and 2023.18 In addition to the outcomes of the meetings at PHI, I have conducted detailed research on the Uttaragrantha (manuscript A) of the Māṇavikā chapter, along with other related materials in Tibetan and Chinese (Lee, forthcoming).

The Uttaragrantha (manuscript B) identified and researched by Shōno, to date, consists of two fragments found in the private collection, Virginia.19 These Sanskrit fragments, labeled F.15.1 and F.15.3, comprise one folio with both front and back (recto and verso) available, corresponding to Prātideśanikā 2–4 of the Upāliparipṛcchā in the Uttaragrantha, and are written in Gilgit/Bamiyan Type II script.20 Although only two fragments of the Uttaragrantha (manuscript B) have been identified and are exclusively found in the private collection, Virginia, a voluminous amount of the Uttaragrantha (manuscript A) is present in the two collections, including, as of March 2024, six chapters of the Uttaragrantha: the Vīnitakā, the Māṇavikā, the Ekottarikā, the Pañcaka, the Ṣoḍaśaka, and the Nidāna. The Uttaragrantha (manuscript A), which is the primary focus of this article, will be discussed in detail in the following section.

3. The Schøyen–Virginia Manuscript of the Uttaragrantha (Manuscript A)

3.1. General Description

The Uttaragrantha (manuscript A) is a birch-bark manuscript and is written in Gilgit/Bāmiyān Type II. Specifically, it seems close to Type B of two sub-groups of Gilgit/Bāmiyān Type II according to Melzer’s classification.21 Paleographic studies and radiocarbon analysis have agreed in suggesting that the Uttaragrantha (manuscript A) dates to between 671 AD and 770 AD.22 It is currently found in two private collections: the major part in the Schøyen Collection in Oslo, Norway; and the rest in the private collection in Virginia, USA. A preliminary investigation by Yao revealed that the Uttaragrantha (manuscript A) was combined with the Bhaiṣajyavastu manuscript, forming a continuous text in a single manuscript.23 This combined manuscript has been named the “Bhaiṣajyavastu-Uttaragrantha manuscript”, a name coined by Yao (forthcoming, 1.1.2. A Brief History of Research on the Bhaiṣajyavastu Fragments). The Bhaiṣajyavastu part of the manuscript is mostly written with eight lines per side, except for several folios of the final part of the Bhaiṣajyavastu, which are written with nine lines (Yao, forthcoming, 1.2.2. General Format and Quality of the Text). This nine-line format appears to be continuously used in the Uttaragrantha part.24 The Bhaiṣajyavastu-Uttaragrantha manuscript has one string hole at 1/3 of the length from the left margin of the folio.25 Folio numbers are found on the left margin of several folios. Where folio numbers are missing due to physical damage, we can often assume them by aligning the surviving Sanskrit text with the Tibetan version. Notably, except for the folios of the Nidāna chapter, the identified folios from the Vinītakā to *Ṣoḍaśaka chapters show a continuous sequence, a point that will be detailed in Section 3.2. Extant Chapters in the S-V Manuscript of the Uttaragrantha. We now have approximately 52 different folios, each of which consists of several fragments of various sizes. The Uttaragrantha part (hereafter, the S-V manuscript of the Uttaragrantha, or simply the Uttaragrantha manuscript) includes six chapters sequenced as follows: the Vinītakā26, the Māṇavikā, the *Ekottarikā, the *Pañcaka, the *Ṣoḍaśaka, and the Nidāna.

3.2. Extant Chapters in the S-V Manuscript of the Uttaragrantha

The folios detailed below were identified and organized as of March 2024. Future research will involve more identified fragments in the S-V manuscript of the Uttaragrantha.

3.2.1. The Vinītakā Chapter

According to Clarke (2015, pp. 77–78), the Vinītakā chapter deals with case law related to the first nine offenses in the Bhikṣuprātimokṣa, including the four pārājikas and five saṅghāvaśeṣas.

Fragments of the Vinītakā chapter are found in both collections. The majority is found in the private collection, Virginia, and the last folio of the Vinītakā is included in the Schøyen collection. Many fragments in the private collection, Virginia, remain in bundles with several folios stuck together, and consequently many folios inside are inaccessible. Accordingly, only 16 folios have been identified although the Vinītakā chapter consists of 23 folios, 206–228.

As aforementioned, the Uttaragrantha manuscript is combined with the Bhaiṣajyavastu, and the first chapter of the Uttaragrantha, followed by the Bhaiṣajyavastu, is the Vinītakā chapter in the S-V manuscript. The first folio of the Vinītakā chapter starts at 206r in the Bhaiṣajyavastu-Uttaragrantha manuscript, as the Bhaiṣajyavastu seems to end at 205v.27 However, the fragments of this first folio are notably partial, preserving only a small portion of the left margin in which the digit 6 from 206 survives. The final folio of the Vinītakā contains a colophon in which the Sanskrit title is attested.28

The identified folios of the Vinītakā chapter in the S-V manuscript of the Uttaragrantha are arranged with the Tibetan version in the following

Table 1. Concordance of the folios of the Vinītakā chapter and the Tibetan material.

Identified Folios of the Vinītakā Chapter in the S-V Manuscript	Tibetan (Derge)
206r	Na 288b5–289a7
206v	Na 289b3–290a6
207r	Na 290b6–291a2
207v	Na 291a7–291b4
208v	Na 293a3–293b2
209r	Na 293b3–294a6
210r	Na 295a6
213v	Na 300b1–301a4
214r	Na 301b3–301b7
215r	Pa 2b2–3
218r	Pa 6b6–7a3
220v	Pa 10b1–5
221v	Pa 12a1–4
222r	Pa 12b6–13b2
223r	Pa 15a3–4
226v	Pa 20b4–7
227v	Pa 21b4–22a5
228r	Pa 22a6–22b2

.29 Due to the fragmentary condition of the folios, the location of the Tibetan version is provided from where the first surviving word of the Sanskrit version appears to where the last word does in the folios.30

3.2.2. The Māṇavikā Chapter31

The Māṇavikā chapter explores a wide range of topics through the structure of questions and answers between Upāli and the Buddha. Although the topics are summarized by the piṇḍoddānas and uddānas, they do not necessarily agree on a singular, consistent theme. These topics are related to those found in various vastus of the Vinayavastu and the Kṣudrakavastu, as well as to regulations (Skt. śikṣāpada, Chi. xuechu 學處) outlined in the Prātimokṣasūtra.32

Following the Vinītakā chapter in the S-V manuscript, a total of nine folios pertain to the Māṇavikā chapter, specifically those numbered 228–236. The majority of the Māṇavikā fragments, totaling 116, is found in the Schøyen collection, while six fragments are included in the private collection, Virginia.33 Regrettably, a portion featuring the title, colophon, or any final sentence that would typically signify the end of a chapter is lost. We cannot directly discern the original Sanskrit title in this manuscript, but the Sanskrit title is attested in Guṇaprabha’s Vinayasūtra (VS 14.40) and its autocommentary to VS 1.98.34 The remaining portions represent approximately 40 percent of the entire text.

Table 2. Concordance of the folios of the Māṇavikā chapter and other materials in Tibetan and Chinese.

Identified Folios of the Māṇavikā Chapter in the S-V Manuscript	Tibetan (Derge)	T. 1441 [23]	T. 1435 [23]
228r	Pa 221b5–222a5	605a11–20	405a21–b2
228v	Pa 222a5–223a1	605a21–b6	405b2–15
229r	Pa 223a1–b1	605b6–15	405b15–c10
229v	Pa 223b4–224a5	605b17–c3	405c13–25
230r	Pa 224a5–225a2	605c4–c13	405c26–406b3
230v	Pa 225a4–b5	605c16–23	406b4–14
231r	Pa 225b5–226a6	605c24–606a13	406b14–406c14
231v	Pa 226b3–227a2	606a18–24	406c19–407a8
232r	Pa 227a2–b4	606a24–b11	407a8–27
232v	Pa 227b5–228b1	606b11–15	407a27–b23
233r	Pa 228b1–229a4	606b15–29	407b23–c22
233v	Pa 229a5–230a1	606b29–c11	407c22–408a10
234r	Pa 230a2–b7	606c11–22	408a10–c736
234v	Pa 231a1–b5	606c23–607a6	408c7–409a1
235r	Pa 231b6–232b4	607a7–13	409a2–b12
235v	Pa 232b5–233b2	607a14–24	409b12–c13
236r	Pa 233b3	-	409c16–17

below presents a comparative table of the Māṇavikā materials, including two Chinese texts, the Sapoduobu pini modeleqie and the Shisong lü, which are related to the Uttaragrantha, as previously mentioned.35

3.2.3. The *Ekottarikā Chapter

Among the Uttaragrantha’s chapters, three chapters bear specific textual characteristics, particularly in their systematic text arrangement: the *Ekottarikā, the *Pañcaka, and the *Ṣoḍaśaka. The title *Ekottarikā means “Increasing by One” or “Single Increments” (MW, s.v. ekottara), reflecting its structured arrangement of topics in an ascending numerical order. This chapter organizes its contents numerically, as established in 10 uddānas within each of the three piṇḍoddānas. It includes discussions on monastic roles, expertise in monastic discipline, significant teachings, and dialogues on understanding the dharma, structured around a unique numerical format for easy navigation and memorization (Clarke 2015, p. 78).

The *Ekottarikā chapter comprises 13 folios, numbered 236 to 248. All identified fragments of the 13 folios are preserved in the Schøyen collection. Each folio is preserved in varying states of fragmentary condition, yet the entire range of the text, from beginning to end, is currently available. However, a portion containing the title, colophon, or any final sentence is missing, leaving the Sanskrit title unconfirmed in this manuscript. While the Mūlasarvāstivāda Vinaya also does not confirm the Sanskrit title, a variation, “vinayottarikā”, appears in the Merv manuscript.37 The identified folios of the *Ekottarikā chapter in the S-V manuscript of the Uttaragrantha are arranged with the Tibetan version in

Table 3. Concordance of the folios of the *Ekottarikā chapter and the Tibetan material.

Identified Folios of the *Ekottarikā Chapter In the S-V Manuscript	Tibetan (Derge)
236r	Pa 22b2–23a6
236v	Pa 23a7–24a6
237r	Pa 24a6–25a6
237v	Pa 25a7–26a5
238r	Pa 26a6–27a5
238v	Pa 27a7–28a7
239r	Pa 28b1–29a6
239v	Pa 29a7–30a5
240r	Pa 30a5–31a6
240v	Pa 31b4–32a2
241r	Pa 32a2–7
241v	Pa 33b1–4
242r	Pa 33b5–34b2
242v	Pa 34b5–35b3
243r	Pa 35b3–36b6
243v	Pa 36b7–38a3
244r	Pa 38a4–39a5
244v	Pa 39a7–40b1
245r	Pa 40b1–41b3
245v	Pa 41b5–42b3
246r	Pa 42b4–43a1
246v	Pa 44a1–44b1
247r	Pa 44b1–45b1
247v	Pa 45b2–46b2
248r	Pa 46b2–47a2
248v	Pa 47a5–7

below.38

3.2.4. The *Pañcaka Chapter

As in the *Ekottarikā chapter, the *Pañcaka chapter features a systematic arrangement of text, listing terms in groups of five. Clarke (2015, p. 78) notes that this chapter deals with accusations, explaining how monks should report the offenses of their fellow monks and how the accused monk should respond, categorizing different types of offenses into groups of five.

The *Pañcaka chapter in the S-V manuscript of the Uttaragrantha comprises nine folios, numbered 248 to 256. Each folio is preserved in varying conditions. The full sequence of folios, from beginning to end, is available. However, a portion containing the title, colophon, or any final sentence of this chapter is also missing, leaving the Sanskrit title unconfirmed. While no other materials of the Mūlasarvāstivāda Vinaya confirm the Sanskrit title either, a variation, “vinayapaṃcika”, is found in the Merv manuscript.39 The identified folios of the *Pañcaka chapter in the S-V manuscript of the Uttaragrantha are arranged with the Tibetan version in the following

Table 4. Concordance of the folios of the *Pañcaka chapter and the Tibetan material.

Identified Folios of the *Pañcaka Chapter in the S-V Manuscript	Tibetan (Derge)
248v	Pa 47b2–48a1
249r	Pa 48a3–48b5
249v	Pa 48b7–49b3
250r	Pa 49b4–50a7
250v	Pa 50b2–51a4
251r	Pa 51b4–52a2
251v	Pa 52a7–52b2
252r	Pa 52b7–53b3
252v	Pa 53b7–54b4
253r	Pa 54b5–55b3
253v	Pa 55b4–56b2
254r	Pa 56b4–57b1
254v	Pa 57b3–58b1
255r	Pa 58b4–59b6
255v	Pa 59b7–60a6
256r	Pa 60a7–61a4

.

3.2.5. The *Ṣoḍaśaka Chapter

The *Ṣoḍaśaka chapter is a relatively short chapter consisting of a list of 16 topics followed by two further lists of five attributes (Clarke 2015, p. 78). The section primarily deals with the settlement of legal cases, focusing on the sthalastha monk as in charge of the settlement procedure (Borgland 2014a, pp. 54–56).

Five folios, numbered 256–260 of the *Ṣoḍaśaka chapter, are found in the S-V manuscript of the Uttaragrantha. The text survives from the beginning to approximately 80% toward the end, which means that the part corresponding to Derge Pa 69a–70a is currently missing. Accordingly, a portion typically containing the title, colophon, or any final sentence is missing, leaving the Sanskrit title unconfirmed in this manuscript. This Sanskrit title is not confirmed in other Sanskrit sources of the Mūlasarvāstivāda Vinaya either, but the term “vinayaṣoḍaśika” is attested in the Merv manuscript.40 The identified folios of the *Ṣoḍaśaka chapter in the S-V manuscript of the Uttaragrantha are arranged with the Tibetan version in the following

Table 5. Concordance of the folios of the *Ṣoḍaśaka chapter and the Tibetan material.

Identified Folios of the *́Ṣoḍaśaka Chapter in the S-V Manuscript	Tibetan (Derge)
256v	Pa 61a5–62a3
257r	Pa 62a6–63a2
257v	Pa 63a5–63b7
258r	Pa 64a2–64b5
258v	Pa 64b5–65b1
259r	Pa 65b2–66a7
259v	Pa 66b1–67a5
260r	Pa 67a6–68a3
260v	Pa 68a5–68b7

.

3.2.6. The Nidāna Chapter41

The term “nidāna” encompasses various meanings such as “cause”, “motive”, “motivation”, “beginning”, and “introduction”. According to Kishino (2013, p. 50), the Nidāna chapter covers a wide range of diverse topics either in a series of questions and answers between Upāli and the Buddha or through narrative stories in which the Buddha establishes regulations. These topics are associated with those found in several vastus and the Kṣudrakavastu, and they appear to be organized into a group as outlined in piṇḍoddānas and uddanas. This structuring bears similarities with that of the Māṇavikā chapter.

As of March 2024, only four folios of the Nidāna chapter are identified in the S-V manuscript of the Uttaragrantha, which are numbered 283, 284, 285, and 289. These identified folios correspond to 92b–98a and 103b–105b in the Tibetan Derge edition, indicating that the remaining part covers from approximately the 2/7 point to the 3/7 point. The following

Table 6. Concordance of the folios of the Nidāna chapter and the Tibetan material.

Identified Folios of the Nidāna Chapter in the S-V Manuscript	Tibetan (Derge)	Chinese (T. 1452 [24]) 根本說一切有部尼陀那目得迦
283r	Pa 92b4–93b1	423c23–424a21
283v	Pa 93b3–94b1	424a26–424b16
284r	Pa 94b1–95a5	424b16–424c12
284v	Pa 95a7–96a6	424c16–425a4
285r	Pa 96a7–97a5	425a9–425a29
285v	Pa 97a7–98a4	425b5–11
289r	Pa 103b6–104b6	427a18–1942
289v	Pa 104b7–105b4	-43

shows the concordance of the four folios with the Tibetan version:

3.3. The Structure of the Uttaragrantha

In summary, the identified folios of the six chapters in the S-V manuscript of the Uttaragrantha are sequenced as follows: Vinītakā (206r–228r3)–Māṇavikā (228r3–236r1)–*Ekottarikā (236r2–248v3)–*Pañcaka (248v3–256r9)–*Ṣoḍaśaka (256v1–260v9)–Nidāna (283r1–289v9). Of particular interest is why the Uttaragrantha is combined with the Bhaiṣajyavastu for reasons not yet clear. Also, the sequence that the Vinītakā chapter is positioned as the first chapter in the S-V manuscript of the Uttaragrantha, is an intriguing aspect of the Uttaragrantha’s structure.

Considering the length of the Nidāna chapter in the Tibetan version, among the currently missing folios numbered 261–282, approximately one folio is expected to be the final part of the *Ṣoḍaśaka, and about 21 folios are expected to be part of the Nidāna. The sequence of the chapters in the S-V manuscript of the Uttaragrantha largely aligns with that of the complete Tibetan version of Uttaragrantha. However, the Māṇavikā chapter is positioned as the second section of the Uttaragrantha part of the S-V manuscript, whereas in the Tibetan version, it is the ninth out of ten or eleven sections. The variations in sequence are found in the two Chinese texts, the Sapoduobu pini modeleqie and the Shisong lü, differing from both the Tibetan and Sanskrit versions of the Uttaragrantha. The Sapoduobu pini modeleqie does not include the Nidāna and Muktaka chapters, following the order: Kathāvastu–Upāliparipṛcchā(a)–Vinītakā–Mātṛkā–Māṇavikā–*Ekottarikā–Upāliparipṛcchā(b). The Shisong lü does not include the *Pañcaka and *Ṣoḍaśaka chapters, and the chapter order also varies from the other versions: Nidāna–*Ekottarikā–Kathāvastu–Upaliparipṛcchā–Māṇavikā–Matṛkā–Vinītakā–Muktaka.44 A comparison of the structures of the Uttaragrantha of the Tibetan and Sanskrit materials and Chinese versions related to the Uttaragrantha, the Sapo-duobu pini modeleqie, and the Shisong lü, is presented in

Table 7. Extant chapters of the Uttaragrantha from the four sources and a comparison of chapter order.

Tibetan Version	The S-V Manuscript of the Uttaragrantha	T. 1441 [23] The Sapoduobu pini modeleqie		T. 1435 [23] The Shisong lü (Recitation 8–10)
Upāliparipṛcchā	Vinītakā	Kathāvastu	毘尼衆分事 (565a14–569b29)	Nidāna45 (346a7–352b26)	Recitation 8 八誦 (346a5–378c7): 毘尼增一法 (346a10–369b24) 增一後 (369b25–378c7)
Vinītakā	Māṇavikā	Upāliparipṛcchā	優波離問分別波羅提木叉 (569c1–579b25); 優波離問事 (579b26–582b12)	*Ekottarikā (352b27–354c8; 355c10–373c6)
*Ekottarikā	*Ekottarikā	Vinītakā	毘尼摩得勒伽雜事46 (582b13–593b20)	Kathāvastu (373c7–378c6)
*Pañcaka	*Pañcaka	Mātṛkā	毘尼摩得勒伽雜事 (593b21–605a5)	Upāliparipṛcchā (379a3–405a20)	Recitation 9 九誦 優波離問部 (379a1–409c19)
*Ṣoḍaśaka	*Ṣoḍaśaka	Māṇavikā	毘尼摩得勒伽雜事 (605a9–607a25)	Māṇavikā (405a21–409c18)47
Nidāna	Nidāna	*Ekottarikā	毘尼摩得勒伽雜事 (607a25–611b11)	Mātṛkā (410a5–423b9)	Recitation 10 十誦 (410a1–470b20): 比尼誦 (410a5–445c6) 五百比丘結集三藏法品 (445c13–450a26) 七百比丘集滅惡法品 (450a27–456b8) 毘尼中雜品 (456b9–461b29) 因縁品 (461c1–470b19)
Muktaka		Upāliparipṛcchā	優波離問波羅夷; 問十三僧伽婆尸沙; 問三十事; 問波夜提; 問波羅提提舍尼事 (611b15–626b9)	Vinītakā (424b16–445a12)
Kathāvastu				Muktaka (456b9–470b19)
Māṇavikā
Mātṛkā

below. This table displays the extant chapters of the Uttaragrantha from these four sources, highlighting differences in the chapter order among them.

It is unclear which chapter sequence is the earliest. Currently, it is not possible to determine whether the sequence found in our extant Tibetan version reflects the order in the Indic originals used by the translators, or if it was later adjusted by them. Despite these uncertainties, two key observations can be underscored: First, the variation in chapter order across all versions may provide evidence of the multiplicity of the Vinayas of the Mūlasarvāstivāda. Second, the difference in chapter sequence across the four sources suggests that each chapter of the Uttaragrantha may have initially existed as an independent Vinaya text. The texts of the Uttaragrantha are frequently cited in Viśeṣamitra’s Vinaya-saṃgraha and the commentaries of the Guṇaprabha’s Vinayasūtra, but this is done by referring to their specific titles rather than using a collective title Uttaragrantha. Also, even though Yijing translated two chapters of the Uttaragrantha, no Chinese term for Uttaragrantha is found in his works. Furthermore, the title Uttaragrantha is not found in any Vinaya text written in Sanskrit or other Indic languages, nor is it attested in the S-V manuscript of the Uttaragrantha. This suggests that the recognition of the title Uttaragrantha might not have been widespread or utilized in India during the time of the Vinaya scholars around the 6th to 7th century CE. The collective reference to these texts as Uttaragrantha could have emerged later, reflecting a development in how these texts were grouped and conceptualized within the tradition (cf. Kishino 2013, pp. 30–35). It is also supported by the fact that the estimated production of the S-V manuscript falls within the 7th to 8th centuries, overlapping with this timeframe.

4. New Findings from the S-V Manuscript of the Uttaragrantha

The S-V manuscript of the Uttaragrantha leads to a significant advancement in the study of the Mūlasarvāstivāda Vinaya, introducing a new textual witness and enhancing research in various aspects. The understanding of the Uttaragrantha, accessible through Tibetan and Chinese translations, has been challenging due to the considerable difference in translation styles from those of the Vinayavastu. The S-V manuscript of the Uttaragrantha contributes to the expansion of the Vinaya corpus with original Sanskrit terms, offering a better understanding of the text. These original Sanskrit terms lead to a more accurate understanding of ambiguous variations of Vinaya technical terms in Tibetan and Chinese. An example examined in

Table 8. Tibetan variations of a Vinaya technical term, Skt. chanda.

Sanskrit in the Māṇavikā	Tibetan in the Māṇavikā	Sanskrit in the Vinyavastu in the MSV	Tibetan in Other Parts, the Vinayavibhaṅgas and Vinayavastu in the MSV
chanda	dad pa	chanda	’dun pa

below is the variation of a Vinaya technical term in Tibetan compared to the original Sanskrit term in the Māṇavikā chapter of the S-V manuscript of the Uttaragrantha.

The concept of “chanda” (consent or approval) plays a crucial role in formal ecclesiastical acts (Skt. karman; Tib. las; Chi. jiemo 羯磨). When the Community performs a formal ecclesiastical act, it is a fundamental principle that all monks within the boundary of that Community participate in the formal act. Participation in the community’s karman was, according to Hirakawa (1982, p. 20), one of the most important aspects of the group’s monastic life. There are two modes of participation. The first is direct personal attendance, and the second is by granting chanda “approval, consent (BHSD, s.v. chanda)” to another monk, thereby signifying agreement with the results of the karman. This indirect mode of participation, through chanda, is permitted only in exceptional circumstances, such as when a monk is ill.

It is noteworthy that, in the Māṇavikā chapter, the Tibetan rendering of chanda differs from that in the Vinayavastu of the Mūlasarvāstivāda Vinaya. Specifically, in the Māṇavikā, chanda is translated as dad pa, which literally means “faith” or “trust”, often corresponding to the Sanskrit term śraddhā (Negi, s.v. dad pa).48 In contrast, vastus of the Mūlasarvāstivāda Vinaya, such as the Karmavastu and Adhikaraṇavastu, use the term ’dun pa, generally meaning “intention” or “will” as an equivalent of chanda.49 Given the same context in both the Māṇavikā and Adhikaraṇavastu, or in other vastus where chanda appears, the employment of different Tibetan translations prompts the question of why. A plausible explanation might be the involvement of different translators.50 However, the original Sanskrit term in the S-V manuscript of the Uttaragrantha helps us to understand the original content of the Māṇavikā and confirms that there is a variant translation for chanda, which differs from the one more commonly known to modern scholars in the Tibetan translation.

Second, the S-V manuscript of the Uttaragrantha serves as an invaluable source in providing new textual evidence for verifying the multiplicity of the Mūlasarvāstivāda Vinaya. In the following example, the S-V manuscript of the Uttaragrantha is closer to the Shisong lü rather than the Tibetan translation of the Uttaragrantha. This not only suggests that the S-V manuscript might originate from a different lineage than the one used for the Tibetan translation among the Vinayas of the Mūlasarvāstivādins, but it also makes us consider that the Mūlasarvāstivāda Vinaya should be understood within the broader context of the Vinayas of the “Greater Sarvāstivāda”.

The Tibetan version of the Māṇavikā (Derge ’dul ba pa 223a6–7)

de ñid la sṅa dro ba daṅ | de ñid bdun pa daṅ | de ñid la nam ‘tsho’i bar du bar ruṅ ṅam | ruṅ ste śa ni sṅa dro ba’o || de’i ‘og tu btsos pa’i źag ni bdun pa’o || bdun pa de’i ‘og tu bsregs pa’i thal ba ni nam ‘tsho’i bar du ba’o ||

[Upāli asked:] “Is it proper that one particular substance can be [the medicine to be consumed] in the morning, within seven days, and throughout life?”

[The Buddha said:] “It is proper. Meat is [the medicine to be consumed] in the morning. After that, fat of boiled [meat] is [the medicine to be consumed] within seven days. After seven days, the burnt ash is [the medicine to be consumed] throughout life”.

The uddāna, which encompasses the dialogue above, of the Tibetan version (Derge ’dul ba pa 223a4)

mdor na |

ruṅ ba dag daṅ ‘dres pa daṅ ||

źag bdun du yaṅ bza’ mi bya ||

tshil bu daṅ yaṅ de bźin te ||

bu ram śiṅ daṅ sṅa dror sbyar ||

Summary of Contents:

Things appropriate, mixed

For seven days, should not be consumed;

Also, fat is likewise,

Sugarcane, in the morning, suitable.

The S-V manuscript of the Māṇavikā 229r6

… m atyanta(rakāle) māṃsaṃ pakvaṃ vasā s(āptā) + + + + + + + + + + + + + + + + + … kāḥ || evaṃ …

[Upāli asked:] …

[The Buddha said:] “… then, fat [from] boiled meat is [the medicine to be consumed within] seven [days]. … Likewise, …

The uddāna, which encompasses the dialogue above, of the Sanskrit version (229r3)

uddānam* kalpikena sa(ṃ)sṛṣṭaṃ saptāhaṃ c(a) abhakṣikaṃ vasād api tilā ikṣu‹ḥ› kālikaṃ cāpi yojayet* || ❁ (||) …

Summary of Contents:

[§1.6.1] By appropriate, Mixed,

Seven days, Not to be eaten;

[§1.6.3] From fat as well as Sesame seeds;

[§1.6.2] Sugarcane is suitable in the morning.

The Shisong lü (T. 1435 [23] 405c7–9)51

胡麻是時藥，壓作油是七日藥，燒作灰是盡形藥。肉是時藥，煮取脂是七日藥，燒作灰是盡形藥。

[The Buddha said:] “Sesame seeds are the medicine to be consumed in the morning; when pressed into oil, they become the medicine to be consumed for seven days; when burned to ash, they are the medicine to be consumed throughout life. [Also,] meat is the medicine to be consumed in the morning; the fat extracted from boiling it becomes the medicine to be consumed for seven days; when burned to ash, it is the medicine to be consumed throughout life”.

T. 1441 [23]

No counterpart

This example addresses whether a single substance can be processed into three different types of medicine. While the Tibetan version only includes meat that can become the three different kinds of medicine through specific processes, both the Sanskrit version and T. 1435 provide two examples: meat and sesame seed. Even if the Sanskrit word tila (sesame seed) is not confirmed in the question-and-answer between Upāli and the Buddha in the Sanskrit version due to the damage to the manuscript, its appearance might be inferred from the term Skt. tilā (sesame seed) in the uddāna that encompasses this dialogue. Also, the use of Skt. evaṃ (likewise) in the Buddha’s answer implies the presence of another example following meat.52

This example suggests that the S-V manuscript of the Uttaragrantha may serve as a crucial instrument for scholars who aim to verify and comprehend the multiplicity of the Mūlasarvāstivāda Vinaya. A more detailed investigation of the textual relationship among the four sources, namely the Sanskrit and Tibetan versions of the Uttaragrantha, the Shisong lü, and the Sapoduobu pini modeleqie, drawing on a broader range of textual evidence, will be conducted in my future research.

5. Concluding Remarks

This article examines the rediscovery and re-evaluation of the Uttaragrantha of the Mūlasarvāstivāda Vinaya by the new Sanskrit manuscript fragments, namely the S-V manuscript of the Uttaragrantha. The research on the S-V manuscript of the Uttaragrantha has marked a significant milestone in the study of the Mūlasarvāstivāda Vinaya, providing a new insight into our comprehension of the monastic law codes. As of March 2024, the newly identified Sanskrit text of the Uttaragrantha, comprising 52 folios, corresponds to 86 folios in the Derge edition of the Tibetan Uttaragrantha. This material not only presents the monastic law code in its original language but also helps to overcome numerous challenges that have arisen due to the reliance on the limited materials existing only as translations in Tibetan and Chinese. Furthermore, this article attempts to show the role of the S-V manuscript of the Uttaragrantha in shedding light on the complex textual history and the potential multiplicity in the Mūlasarvāstivāda Vinaya traditions. It does so by introducing additional evidence that highlights the differences in sequence in the chapters of the Uttaragrantha and by exploring various relationships with related materials, contradicting the expectation that it would be close to the Tibetan translation of the Uttaragrantha. The folios that have been identified thus far constitute merely the initial phase, with the expectation of more identification in the near future. Further identification and analysis of the S-V manuscript of the Uttaragrantha will significantly expand the canonical corpus of the Mūlasarvāstivāda Vinaya. With the invaluable support provided by the new Sanskrit textual evidence, a more comprehensive examination of the Uttaragrantha with its related materials will be conducted by the present author.