The Updated Data on Sanskrit Manuscripts of the Serindia Collection (IOM, RAS): Perspectives of the Study

Abstract

This article presents the preliminary results of the study on the Sanskrit manuscripts of the Serindia Collection of the Institute of Oriental Manuscripts, RAS. Basing on the previous researches, as well as on the results of the efforts of the ‘Sanskrit Group’ within Serindica Laboratory, the authors outline the structure and repertoire of the Sanskrit part of the Serindia Collection, supplementing it with the description of paleographic and codicological aspects of the Sanskrit manuscripts.

Full Text

The penetration of Buddhism into Central Asia dates back to the first centuries of the 1st millennium AD and is associated with India. However, it is not clear yet how Buddhism spread and developed in this region and what features this Indian religion had while its expanded inside the Tarim Basin (these territories are also known as Serindia and East [Chinese] Turkestan; at present Xinjiang Uyghur Autonomous Region of PRC). Basing on the paleographical research of manuscripts that have been preserved to nowadays, we can assume that Sanskrit was used as the main language of the transmission of the Buddhist tradition in the first centuries AD. Later, when Buddhist texts were translated into the local Central Asian languages, Sanskrit was used as a language of the Central Asian Buddhist written tradition.

Provenance and History of Study

The Serindia manuscripts Collection perhaps is the most multilingual and ‘multi-scriptual’ among all the manuscript collections of the St. Petersburg Institute of Oriental Manuscripts of the Russian Academy Sciences (IOM, RAS). It is very heterogeneous in terms of genres and functional purposes of the texts. A significant part of the Serindian Manuscripts Collection are Buddhist manuscripts in Sanskrit.

The St. Petersburg collection of Sanskrit manuscripts from the Tarim Basin oases consists mainly of the items that were obtained thanks to the efforts of the Russian consuls and diplomats in East Turkestan: the Consul-General in Kashgar Nikolai F. Petrovskii (1837–1908), Consul General in Urumqi Nikolai N. Krotkov (1869–1919), secretary of the consulate in Kashgar Mikhail I. Lavrov (1877–1934), doctor of the Russian consulate in Urumqi Aleksandr I. Kokhanovskii and others. The key role in this process belongs to Nikolai F. Petrovskii. Since 1867 he collected manuscripts and objects of Buddhist art and in every possible way contributed to the comprehensive research of the Serindia region. Through his activities he initiated a wide study of the Indian written heritage in Central Asia.

The first publisher of the Central Asian Sanskrit manuscripts kept in the Asiatic Museum was the academician Sergei F. Oldenburg (1863–1934), who conducted the identification of the Central Asian scripts and miscellaneous texts contained in the discovered manuscripts. He and Andrei I. Vostrikov (1902–1937) prepared an inventory of the collection, where 240 items were listed.

Manuscripts sent by Nikolai F. Petrovskii to Sergei F. Oldenburg during 1893–1903 were published in ZVORAO. In addition to the facsimiles and the transliteration these publications contain the academically founded identification of the texts, preliminary paleographic and codicological information. Sergei F. Oldenburg also planned to publish all Sanskrit manuscripts and fragments from East Turkestan in a separate special series. Thus, Irina V. Tunkina had found the Oldenburg transliterations of the Sanskrit manuscripts in the collection of acad. Fiodor I. Scherbatskoi in the St. Petersburg Branch of the Archive of the RAS. The typographic proof of the 4th part and the draft plan of the 5th part of the series of articles entitled “The Fragments of Kashgar Sanskrit Manuscripts from the Collection of Nikolai F. Petrovskii” were found in the same collection.

A. Staёl von Holstein (1877–1937) and Nikolai D. Mironov (1880–1936), the colleagues of Sergei F. Oldenburg, continued the research of Sanskrit manuscripts and fragments from the St. Petersburg Collection. Staёl von Holstein had published the transliteration of Kāśyapaparivarta-sūtra (SI 1905), as well as one Tibetan and four Chinese translations of this text. Mironov had published the bilingual fragment of Dharmapada in Sanskrit and Tocharian B.

In the 1950s, after a long ‘time-out’ caused by the period of repressions (1930s) and the World War II, the new stage of the research of the Serindia manuscripts heritage was began by a young talented scholar Vladimir S. Vorobiov-Desiatovskii (1927–1956). He investigated Sanskrit, Khotanese and Tibetan manuscripts, published folios of the Kāśyapaparivarta sūtra manuscript (SI 1905), syllabic tables of the South Turkestan and North Turkestan Brāhmī (SI 1039, SI 1040, SI 1909, SI 3120), fragments of the Sanskrit explanatory dictionary and Sanskrit-Tocharian bilingual text (SI 2042, SI 2042). Unfortunately, Vorobiov-Desiatovskii had an ill heart and passed away when he was just 29.

The new academic team, acad. Grigorii M. Bongard-Levin (1933–2008), Eduard N. Tiomkin (1928–2019), Margarita I. Vorobiova-Desiatovskaia (born in 1933), continued the pursue of Vladimir S. Vorobiov-Desiatovskii. They had introduced into academic circulation the numerous Sanskrit manuscripts of Serindia Collection (former Central Asian Collection). The result of their efforts was the publishing of three volumes of “The Monuments of Indian Writing from Central Asia” (1985, 1990 and 2004). The research of this academic group has reached a new level: the publications were provided with facsimiles, transliterations, translations and comments on the Buddhist texts that these manuscripts contain. Many ZVORAO articles of Sergei F. Oldenburg were republished by Bongard-Levin, Tiomkin and Vorobiova-Desiatovskaia in the above-mentioned editions. The innovations included the addition of new call numbers of the manuscripts, new codicological data, comments and the transliteration according to international standards.

In addition to the above-mentioned works, Bongard-Levin and Tiomkin and Vorobiova-Desiatovskaia published articles devoted to previously unknown Sanskrit manuscripts from East Turkestan. Some manuscripts were published in collaboration with the Japanese and German scholars.

Since 2010s Safarali H. Shomakhmadov has continued the research of Sanskrit manuscripts of Serindia Collection. Besides the articles written in collaboration with Margarita I. Vorobiova-Desiatovskaia, he had published some papers devoted to two Tangut block prints (SI 6563, SI 6564) containing Sanskrit dhāraṇī written in Siddhaṃ script.

Foreign scholars also did extensive research of the Sanskrit manuscripts from the Serindia Collection, both in cooperation with IOM colleagues and independently. Japanese researcher Hirofumi Toda wrote a comprehensive monograph devoted to the study of Lotus Sūtra that contains transliteration of the Central Asian version of the sutra accompanied with parallel passages of the all known Lotus Sūtra manuscripts from the European and Asian manuscripts collections (including the ones kept in Russia). In particular, Hirofumi Toda’s monograph contains transliterations of two manuscripts from the Serindia collection: the fragment of Lotus Sūtra (SI 2077) and so-called ‘Nikolai F. Petrovsky Kashgar manuscript’ (SI 1925, SI 1927). The last one is the most complete Central Asian version of Saddharmapuṇḍarīka-sūtra.

The German scholar Oskar von Hinüber (born in 1939) also had researched the Lotus Sūtra manuscripts from Serindia collection: he had published the transliteration and the detailed research of the colophons of some Lotus Sūtra manuscripts from the Serindia Collection including ‘Nikolai F. Petrovskii Kashgar manuscript’ (SI 1925, 1927).

Japanese scholars Seishi Karashima, Tatsushi Tamai from the International Research Institute of Advanced Buddhology and their colleagues made an important contribution in the methodology of the Sanskrit Buddhist manuscripts research. The result of close collaboration of the international research group headed by Karashima and St. Petersburg scholars resulted in the publication of the first volume of ‘The St. Petersburg Sanskrit Fragments’. This collective monograph contains re-editions of the follows manuscripts from the Serindia Collection: Śārdūlakarṇāvadāna (SI 1942, SI 3431, SI 5145), Ajitasenavyākaraṇa (SI 2085), Mahāparinirvāṇa-sūtra (SI 3038, SI 3042) and the part of Merv manuscript (SI 6580).

Among the articles devoted to the St. Petersburg manuscripts and published by international researchers it is necessary to mention the publication by Shin’ichirō Hori, dedicated to the Kalpanāmaṇḍitikā fragments (SI 2041/5, SI 3695) written in the so-called Gilgit-Bamiyan Brāhmī (Type II) script as well as the paper of German scholar Klaus Wille, who published the Buddhanāma-sūtra fragments (SI 3457, SI 3458, SI 3467).

Despite of the great number of articles of Russian and foreign researchers devoted to the study of Central Asian written heritage from the Serindia Collection kept in IOM, RAS, the main part of these manuscripts needs the more detailed research.

A General Survey of the Sanskrit Manuscripts from the Serindia Collection

The Sanskrit collection stands out among all the other holdings of the Serindia Collection due to its abundance and multeity. All kinds of material for the manuscripts, different script types and styles are presented in this collection. It is necessary to stress out the variety of contents of the Buddhist Sanskrit manuscripts: many texts belong to different historical periods and originate in the different Buddhist sects and Buddhist literature genres.

The number of all the inventoried Sanskrit manuscripts in the Serindia Collection kept within various sub-collections is about 700 items. Most of the Sanskrit manuscripts belong to the Petrovskii collection (464 items). Berezovskii collection is represented by 73 storage items, Oldenburg collection — 56, Krotkov collection — 42, Lavrov collection — 38, Malov collection — 10, Klementz collection — 3, Kolokolov collection — 3, Kozlov collection — 2. The Kokhanovskii and Likhachev collections have one item each. 7 items do not belong to any collection.

The identified Sanskrit manuscripts from the Serindia collection, perhaps, present all the main Buddhist canonical literature genres of both Hinayāna and Mahāyāna: Prajñāpāramitā texts, Vinaya fragments, Abhidharma treatises, Jātakas, Dhāraṇī. The latter was very popular among Serindian Buddhists. Up to date the overwhelming majority of the Buddhist Sanskrit texts (508 items) have not been identified, some of the fragments do not correspond to any known Buddhist Sanskrit text.

Materials

The analysis of the materials of the Sanskrit manuscripts allows to research the manuscript culture in Serindia particularly and in Central Asia in whole. The paper production in the Tarim oases had caused the active growth of the recording of Buddhist texts in local paper and neutralized the need to import Buddhist texts from India and China directly. In its turn, it had caused the formation of own Central Asian manuscript culture. It’s notable that the Buddhism reception within Serindia was reflected in the change of the manuscripts materials, the scripts development and the genre transformation of Buddhist literature.

Palm leaf is the earliest material of Buddhist manuscripts in Central Asia. Apparently, this is the case of Buddhist manuscripts import directly from India to the Tarim oases. This assumption is confirmed by the fact that the script of the Serindia Sanskrit texts written on palm leaf is either the Kuṣāṇa Brāhmī (2th–3th cc.), or the North-Western (Indian) Gupta (4th–6th cc.). The Sanskrit manuscripts fragments on palm leaves in Brāhmī script kept in the Turfan collection in Berlin and in the Serindia Collection in St. Petersburg are most ancient.

The ‘Sanskrit division’ of the Serindia Collection includes 20 manuscripts on palm leaf. The fragments of canonical Abhidharma — the philosophical level of Buddhist doctrine — have an undoubted interest for the detailed research. Two of them (2th–3th cc.) written in the Kuṣāṇa Brāhmī (‘Spitzer Manuscript’ SI 1424–1425 are the oldest Sanskrit fragments from the Serindia Collection. Another two fragments (4th–5th cc.) on palm leaf (SI 1426–1427) containing Abhidharma texts were written in the Indian Gupta script. Other palm-leaf manuscripts are small fragments containing few akṣaras only. They can be used to identify the script type and an approximate dating, but it is not possible to identify the texts.

The Serindia Manuscripts Collection contains about 90 Sanskrit fragments on birch bark written in Indian script types, as well as in Central Asian ones — the Turkestan Gupta (4th–5th cc.) and the southern variety of the Early Turkestan Brāhmī (5th–6th cc.). Regarding the birch bark manuscripts from the Serindia Collection we can see that, as a rule, the upper thin translucent layers were used for making a leaf of the manuscript which were then glued or pressed. The manuscripts containing the fragments of Vinaya (SI 1943) and Abhidharma (SI 6583) are rare examples of identified birch bark manuscripts. Most of the items are pieces of poorly preserved birch bark manuscript fragments containing only separate words without any chance to identify the text content.

Buddhist manuscripts fragments on paper are largest manuscripts collection in the ‘Sanskrit division’ of the Serindia Collection. The vast majority of Dhāraṇī texts, all versions of Mahāyāna texts: Saddharmapuṇḍarīka sūtra, Prajñāparamitā-sūtra, Mahāparinirvāṇa-sūtra and Kāśyapaparivarta-sūtra are written on paper. This group includes two large avadānas — Śārdūlakarṇāvadāna and Ajitasenavyākaraṇa — and many other manuscripts of the Serindia Collection. The paper spread within Serindia had coincided with the appearance of first translations of Buddhist texts into local languages. The Central Asian Brāhmī based on Indian scripts was adapted for the writing Buddhist texts in local East Turkestan languages. The process of Brāhmī scripts adapting in the Northern and Southern Tarim oases had led to certain changes in the Central Asian manuscript culture. As a result, since 4th till 10th cc. some manuscript traditions were formed as well as many local scripts had appeared. Those ‘new scripts’ were used for write Buddhist texts both in Sanskrit and in local languages.

Scripts

Only a few little fragments on palm leaves and birch bark from the ‘Sanskrit division’ of the Serindia Collection were written in Kuṣāṇa Brāhmī script. The codicological description of these manuscripts is complicated by the severe damage of the items. Only two paginated leaves have survived — the ‘Spitzer manuscript’ fragments (SI 1424–25) that have a pagination on the left side of the obverse. The margins are narrow, not graphically delineated. The binding hole on the left side of the leaf doesn’t mark decoratively. There are no guidelines. The approximate size of the sheet is 4×20 cm.

The manuscripts written in Indian Gupta script (4th–6th cc.) are more numerous. This script type was used in manuscripts on palm leaves (Abhidharma fragments, SI 1426–27), on birch bark (Vinaya fragments SI 1943; Abhidharma fragments SI 6583) and on paper (Śārdūlakarṇāvadāna, SI 1942). Pagination occurs both on recto (SI 1942) and verso (SI 2038, SI 6580). The fields are narrow (0.6 cm) and not marked graphically. There is a hole for binding on the left side of a leaf that is not graphically marked any way. Guidelines are not visible, nevertheless, a fairly high level of calligraphy is observed, the lines are even (in SI 1942 and SI 2038 the distance between lines is 0.8–0.9 cm). Woven paper, good quality (paper pulp is even, equally shredded), there are traces of a primer. Manuscripts on palm leaf and paper are single-layer while birch bark manuscripts are glued together from two layers. The average size of this type of manuscript is 5×22 cm (on a palm leaf), 5×19 cm (on birch bark) and 7×18 cm (on paper).

About forty Sanskrit manuscripts written in Turkestan Gupta script (4th–5th cc.) in the period before the division into Northern (Tocharian) and Southern (Khotanese) Brāhmī varieties are the special paleographic interest: Turkestan Gupta script has the characteristic features of both scripts. Manuscripts of this type are presented mainly on paper (except of small pieces of birch bark and poorly preserved wooden tablets). Fragments of Vajrapāṇihṛdaya (SI 2034), Mahāsāhasrapramardinī (SI 2040/1–2), Aniruddha-sūtra (SI 3031), and a manuscript that content is defined as an ‘spell’ (SI 2025) are examples of manuscripts written in this script. Pagination, like in the previous script case, occurs both on recto (SI 3031) and verso (SI 2025). Margins and hole for binding are not graphically marked. The laid paper of good quality. Guidelines are not visible, however, equal distances between the lines (0.9–1 cm) are observed. Leaves have different sizes, there are both medium (7×20 cm, SI 2024) and large (8×30 cm, SI 3031).

It’s necessary to mention separately the manuscript written in Turkestan Gupta in a scroll format (SI 2040/1–2) that is unique for the Sanskrit division of the Serindia Collection. The size of the one manuscript fragment is 12.5×7.5 cm, the second — 57×7.5 cm. The woven paper of good quality. There are no guidelines, the distance between the lines is 1–1.2 cm. The margins are narrow, about 0.5 cm.

Examples of Northern Brāhmī varieties in the Serindia Collection are extremely few in comparison with Southern Brāhmī scripts caused by the history of expeditions to East Turkestan organized by Russian and foreign researchers in the late 19th — early 20th cc. However, the available manuscripts from the Northern oases of Tarim basin allow us to trace the evolution of the Brāhmī script in the north of Serindia. Thus, some items in the collection represent the Early Turkestan Brāhmī (type I, 5th–6th cc.) — the script of Sanskrit manuscripts compiled in the Northern oases of the Tarim basin (SI 2027, SI 2035, SI 2045, SI 2069–70). There are no guidelines in these manuscripts; the binding hole is not specially marked. The primed paper, sometimes woven (SI 2027) and laid (SI 2070). The approximate leaf size is 6×21 cm.

The manuscripts written in Northern Turkestan Brāhmī developed from Early Turkestan Brāhmī (type 1) and spread in northern Serindia oases in 7th–9th cc. are presented more widely in the ‘Sanskrit division’ of the Serindia Collection. Prātimokṣa-sūtra (SI 964–971), the syllable tables (SI 1038–1040), Śārdūlakarṇāvadāna (SI 3431, SI 5145), Māṇibhadra-dhāraṇī (SI 2036, SI 2041/1), Dhāraṇī (SI 2037), Nagaropama-sūtra (SI 2041/3, SI 2041/6), Pārājikadharma fragment (SI 2041/7), etc. are among the manuscripts written by this script type. In contrast with mentioned above script types with unstable pagination the pagination in the manuscripts written in Northern Turkestan Brāhmī is placed strictly on verso. The manuscripts written in this script type have guidelines made by red ink (SI 2041/3, SI 2041/6–7, SI 5145) and brown one (SI 1040). The one more peculiarity inherent in the manuscripts written in Northern Turkestan Brāhmī is the graphical marking both left and right margins: the leaves of the manuscript SI 5145 have traces of marking — both margins are outlined in red ink. The paper of high quality, some places with perfectly shredded paper pulp without any visible fiber fragments, laid (SI 964–971) or without laid traces (SI 1040), in some cases primed (SI 964–971, SI 2041/6, SI 5145). There is usually a rectangular space marked by guidelines around the binding hole. The standard leaf size is 5×20 cm (SI 964–965, SI 2041/3). However, there is also a large format (8×55 cm, SI 5145). Thus, the peculiar tradition of the manuscripts decoration developed in the northern Tarim oases: the pagination strictly on verso, the required guidelines, the marking of the margins and the rectangular area for binding hold are specific features of the Northern Turkestan manuscripts.

The manuscripts from the Southern oases of Tarim basin (Khotan, Niya, Endere) are the most representative and numerous among all Sanskrit manuscripts of the Serindia Collection. According to L. Sander classification, the southern group of Brāhmī varieties is represented by four successive script types: Early Turkestan Brāhmī (type 2, 5th–6th cc.), Early Southern Turkestan Brāhmī (7th–8th cc.), Southern Turkestan Brāhmī (8th–9th cc.), Later Southern Turkestan Brāhmī (9th–10th cc.). The manuscripts written in Early Turkestan Brāhmī (type 2) and Southern Turkestan Brāhmī are most numerous among the manuscripts discovered in the southern Tarim oases because of that, obviously, they correspond to the periods of Buddhism flower in Khotan. The examples of Later Southern Turkestan Brāhmī are very few and characterized by somewhat artsy calligraphy.

The fragments of Mahāparinirvāṇa-sūtra (SI 3038, SI 3042, SI 3043), numerous fragments of Prajñāpāramitā texts (SI 2016–2019, SI 3032–3033, SI 3685, SI 3687), Saddharmapuṇḍarīka-sūtra (SI 3030, SI 3330, SI 3332/3), Suvarṇabhāsottama-sūtra (SI 3329/1–3), Kāśyapaparivarta-sūtra (SI 3037), Prātimokṣa-sūtra (SI 3332/1), etc. are examples of the manuscripts written in Early Turkestan Brāhmī (type 2). The Serindia Collection contains the Sanskrit manuscripts (in poṭhī format) written in this script type that have a huge size. There are manuscripts fragments of non-standard size (SI 2017, SI 2019). Thus, the size of a whole leaf of the manuscript was approximately 25×60 cm that obviously exceeds the standard size. To prepare leaves of such large size they were glued together in two parts. This is confirmed by the gluing stripes preserved in the central part of the fragment SI 2017 and on the right side of the fragment SI 2019.

The colored miniature (the Buddha image inside double round frame, SI 2019) as a binding hole decoration appears at first time in the manuscript from the southern Tarim oases. There are guidelines not only for lines but also for left margin designation (SI 3030). The pagination is located on recto strictly. The paper is laid; traces of primer are not always visible; the leaves surface is smooth, possibly polished. There is paper of high as well as low quality; there are fiber traces in paper pulp of the manuscripts SI 3030, SI 3330. There are manuscripts of absolutely different sizes: there are medium (8×35 сm, SI 3037) and small (6×18 сm, SI 3332/1) manuscripts in addition to the large format poṭhī leaves. We can say that 5th–6th cc. is the period of the beginning of the Southern Turkestan manuscript tradition.

The next stage of the development of the manuscript tradition in the southern Serindia oases is reflected in the manuscripts written in Early Southern Turkestan Brāhmī (7th–8th cc.) — such as Kāśyapaparivarta-sūtra (SI 1905, SI 2014), Saddharmapuṇḍarīka-sūtra (SI 1941), Ratnarāśi-sūtra (SI 2013) and others. The paper of these manuscripts has characteristics similar to the manuscripts of the previous tradition: it’s a laid, with a fiber fragments. The margins and the binding hole are not marked (except of the manuscript SI 1905 where the binding hole is marked by a circle and the left margin is outlined by guidelines). The pagination is located on the recto in all manuscripts. The manuscript SI 1905 has wooden covers. Most likely, the manuscripts with a large number of leaves had wooden covers for better preservation. The wooden cover has a binding hole. These covers are not decorated in any way additionally, the text traces can be found on the inside. The one more manuscript has a similar wooden covers (Saddharmapuṇḍarīka-sūtra, SI 1925). The manuscripts sizes are 6×40 cm (SI 1905), 10×26 cm (SI 1941) and 6×28 cm (SI 2013–2014).

The manuscripts written in Southern Turkestan Brāhmī are most numerous within ‘Sanskrit division’ of the Serindia Collection. Buddhanāma-sūtra (SI 2075–2076, SI 3015–3016, SI 3027), Ajitasenavyākaraṇa (SI 2085), Dharmaśarīra-sūtra (SI 2086, SI 3014), Samādhirāja-sūtra (SI 3001–3002), Sumukhanāma-dhāraṇī (SI 3026), Prajñāparamitā (SI 3682) as well as almost all versions of the Lotus Sutra that will be discussed below are excellent examples of this script type. The codicological features of these manuscripts are very similar: the binding hole is marked by a circle; the paper is laid, with an occasional fiber fragments; the pagination is located on recto only. Some manuscripts have guidelines (SI 3001–3002, SI 3015). The left margin is marked with guidelines very rare (SI 2085), the right margin is present sometimes but it hasn’t graphical guidelines. The margins in the manuscripts are both wide (2.5 cm, SI 3002) and narrow (0.8 cm, SI 3027). The leaves have absolutely different sizes: there are manuscripts of both small — 4×20 cm (SI 2086), and large — 20×58 cm (SI 3682) formats. It’s remarkable that the leaf size didn’t determine the lines number. For example, manuscripts SI 3014 (8×37 cm) and SI 3026 (4×25 cm) contain the same number of lines: 4 lines on both sides of the leaf. In the first case, lines are located at a great distance from each other (2.0 cm), in the second — at a distance of 0.8 cm.

As was mentioned above the manuscripts written in Southern scripts are most numerous in the Serindia Collection. In particular, all Saddharmapuṇḍarīka fragments are written in the scripts spread in the southern oases of Serindia. The main part of the Lotus Sutra fragments was written in Southern Turkestan Brāhmī as well as the Kashgar manuscript of Petrovskii that is the longest version of the sūtra (about 400 leaves) and the core of the Sanskrit manuscripts containing the Saddharmapuṇḍarīka text. The group of the Lotus Sutra manuscripts perhaps is one of the largest in the Serindia Collection (28 items). As a rule, these manuscripts belong to the Petrovskii Collection: SI 1925; SI 1927; SI 1933–1941; SI 2077; SI 2092; SI 2093; SI 2098; SI 3000; SI 3013; SI 3025; SI 3030; SI 3044; SI 3631; SI 3693; SI 3694. Four items (two ones in each collection) are presented in the collections of Mikhail I. Lavrov (SI 3330; SI 3332/3) and Sergei E. Malov (SI 4517; SI 4519). One more item belongs to the Sergei F. Oldenburg Collection (SI 4645). The codicological features of these manuscripts are similar to other fragments written in Southern Turkestan Brāhmī. The paper of good enough quality, the no primer traces are not visible. In some manuscripts, mainly in large and decorated ones (for example, SI 3025 size: 14×54 cm), margins are outlined and there are laid traces made by a lead stick. In addition, the manuscript SI 1933 has an image (Śākyamuni Buddha with a donator), and there is a double decorative circle for a miniature on some leaves of SI 1925 and SI 1927.

It’s necessary to mention separately the Sanskrit manuscripts written in the so-called Gilgit-Bamiyan Brāhmī (type2; 6th c.). Manuscripts SI 2041/5 and SI 3695 are the fragments of the Kalpanāmaṇḍitikā (“Decorated Poem”) attributed to the Buddhist poet Kumāralāta. The two other fragments (SI 2998, fragment No. 5 and SI 5521) written in the same script are not identified. This Brāhmī script type has another name — proto-Śāradā — and, as the name suggests, it precedes the western version of the Gupta script — Śāradā that arose approx. the 8th c. The codicological features of the fragments SI 2041/5, SI 2998 and SI 3695: the paper is one-layered, laid, of very high quality, with perfectly shredded paper pulp. SI 5521, in contrast, has a two-layered paper, of poor quality, the ink is slightly blurry. It can indicate a poor quality of material or a poorly processed paper surface. The pagination occurs only in one fragment (2041/5) — on verso. The guidelines are not visible. Obviously, this manuscripts were compiled on the territory of modern Afghanistan and Pakistan but the circumstances of their acquisition and existence in the Serindia manuscript Сollection remain unclear.

Conclusions

The first stage of the Buddhism institutionalization in Serindia (2th–4th cc.) was marked by the intensive ‘export’ of Buddhist teaching, as a rule, Hinayāna (Sarvāstivāda), from the northern regions of India and Kashmir. It can be testified, first of all, by the repertoire of the discovered Sanskrit fragments and, secondly, by the not typical for Serindia manuscripts material — palm leaves and birch bark. During this period, Buddhist Dhāraṇī are spread in Serindia. These spell texts contain besides of Buddhist deities the names of Vedic gods. It’s very typical for the the religious syncretism of the Kuṣāṇa Empire.

The second stage — 4th–6th cc. — is characterized by the keeping of Sanskrit as the main language of Buddhist manuscript tradition and the change of the manuscripts material from birch bark and palm leaves to locally produced paper. Ideologically, a turn towards Mahāyāna Buddhism is outlined. Such Mahāyāna texts as Mahāparinirvāṇa-sūtra, Prajñāpāramitā-sūtra, Saddharmapuṇḍarīka-sūtra (the fundamental text for East Asian Buddhism) appear. At the same time, the Dhāraṇī texts retain their popularity.

The peculiarity of the third stage — 6th–9th cc. — when both the Lotus Sūtra and another fundamental texts of Buddhist doctrine (Dharmaśarīra-sūtra, Prātimokṣa-sūtra) as well as Dhāraṇī remain their actuality, is an active appearance of the Buddhist texts in local languages (such as Khotanese Jñānolka-dhāraṇī) that marks the completion of the Buddhism institutionalization in Serindia.

In terms of codicology of the analyzed manuscripts it should be noted that, except of rare for Serindia the scroll manuscript format, poṭhī is a typical manuscript format for all Tarim oases regardless of the material. There were decorating changes that became more ornamental. Particularly, the margins and the area around a binding hole as well as guidelines needed for calligraphic writing started to be outlined. The pagination and features of leaves design began to vary depending on the region. It was in a same time (the middle of 1st millennium AD) when Brāhmī script was dividing into two separate scripts traditions — Northern and Southern. The earlier manuscripts were characterized by unstable pagination — the leaves could be paginated both on the recto and verso. Later, the recto pagination had become typical for the manuscripts from Southern Tarim oases but verso — for Northern ones. The outline of the area around a binding hole is another peculiarity of this stage: for Southern Tarim manuscripts it was a circle but for Northern ones – quadrangle. Moreover, Southern Tarim manuscripts had leaves with empty circles prepared for colored miniatures; two manuscripts have such images. The Northern Tarim manuscripts have neither decorated circles nor miniatures. The one more difference of the large deluxe manuscripts written in Southern Turkestan Brāhmī is the colorless guidelines in contrast of the Northern Turkestan Brāhmī manuscripts that guidelines are in black or red ink.

At present, the main goal of the Sanskrit group of the Laboratory Serindica is to compile the catalogue that would provide comprehensive codicological description of the Sanskrit manuscripts from the Serindia Collection. The group’s tasks also include the research of Buddhism Sanskrit manuscripts in terms of their historical, cultural and source study value: the preliminary texts identification, general classification of the Central Asian Buddhist manuscript heritage and the clarification of its regionally specific features. In addition, the work on the compilation of an electronic dictionary of Buddhist texts from the Serindia Collection has recently begun.

Abbreviation

ARIRIAB: Annual Report of the International Research Institute for Advanced Buddhology at Soka University

GRVL: Glavnaia redaktsiia vostochnoi literatury

MO: Manuscripta Orientalia

PPV: Pis’mennye Pamiatniki Vostoka [Written monuments of the Orient]

WMO: Written Monuments of the Orient

ZVORAO: Zapiski Vostochnogo Otdeleniia Imperatorskogo Russkogo Arkheologicheskogo Obshchestva [Proceedings of the Oriental Department of the Imperial Russian Archaeological Society]