The Berlin Turfan-Collection

by Simone-Christiane Raschmann

Among the archaeological expeditions from numerous countries to explore the ruins in the deserts of Central Asia, then terra incognita, there were also four German expeditions to the oasis of Turfan. Between 1902 and 1914 German scholars under the leadership of Albert Grünwedel and Albert von Le Coq made excavations in the area of Turfan, Hami, Kucha and Karashahr.

Among the finds were many manuscripts and blockprints, written in about 15 various languages with about 30 scripts. Most of these texts are written on paper, only few of them on other materials like silk, leather, birch bark, palm leaves or wooden tablets. Immediately after the return of the first expedition the scientific research on these texts began. Friedrich Wilhelm Karl Müller deciphered some fragments written in a "small script", the Manichaean Estrangelo. His article was published in 1904 and was the beginning of a new stage in the investigation of Manichaeism. In 1904, other scholars like Richard Pischel, Karl F. Geldner and Karl Foy also published first results of their investigations of the manuscripts.

After these first sensational publications, the Orientalische Kommission at the Berlin Academy of Sciences was founded in March 1912. It was to coordinate research on the more than 30,000 fragments of the so called "Turfan Collection".

Among the manuscripts and blockprints there were some texts in then unknown or little known scripts. Albert von Le Coq himself, besides being an expert on Central Asian Art, became a good specialist in interpreting some of these languages. He made many transcriptions from these texts, which were used also by other scholars like F.W.K. Müller, Willy Bang and Annemarie von Gabain for their studies and are still of great interest today.

Other manuscripts written in Brahmi script in an until then unknown language, which included some Indo-European elements, were discovered in the Kucha region. F.W.K. Müller found the word Toxri as a name of a language in a epilogue of a large Uigur text, of which a parallel version in this unknown language was also discovered. Thus the language became known as Toxri or Tokharian. Fragments of Tokharian texts were also found in the Turfan oasis. Later Emil Sieg and Wilhelm Siegling wrote a Tokharian Grammar on the basis of this material.

Before the discoveries of the Old Turkish manuscripts in Central Asia there were known only the old Turkish inscriptions from Mongolia and the Ottoman manuscripts from the West as written sources for Turkish history. The work on this text group was begun by K. Foy. The first important studies of this material were published by F.W.K. Müller in his Uigurica (1908), later growing to three volumes. His work was to be continued by Willy Bang and Annemarie von Gabain. Besides preparing further editions of texts, both of them began to use the texts as sources for investigations on old Turkish history.

The Turkish part of the collection is a good example for the great variety of the Turfan texts. It consists of about 8,000 fragments of various size. These fragments are parts of scrolls, folded books, Pothi-books and blockprints. Most of them are written in the so called Uigur script, which was developed from the Sogdian, some fragments are in Sogdian script. The Turkish Manichaeans used the Manichaean and the Uigur script for writing down their texts. A small number of the fragments was written in Turkish Runic writing, but also in Syrian, Tibetan and Brahmi script. In a seal of a Mongolian document there is an Uigur signature written in 'Phags-pa script.

Sometimes the script used already can give us a hint of the contents of the text. So it is certain that the Uigur Buddhists at no time used the Manichaean script and that on the other hand there are no Manichaean texts in Brahmi script. Also formal signs like punctuation can help to characterize a text. A great problem is the dating of the texts: most of them are undated, but if there is a date in the text (for instance in the colophon), it is one of the twelve animals cycle type which allows no absolute dating. But it is possible to say that most of the Manichaean Turkish texts and the oldest Buddhist texts were written between the 8th and the 10th century. Most of the Buddhist and non-religious texts and documents are from the 10th - 13th century. The Buddhist texts form the greatest part of the Turkish part of the Berlin Turfan collection. There are carefully written as well as carelessly written cursive texts and a number of blockprints. Sometimes in a carefully composed text the writers used red ink for special words like Buddha, Bodhisattva etc or for the indication of the beginning of Gathas. Some manuscripts and blockprints are illustrated, these pictures having been made by Uigur artists. The Buddhist texts are mostly written in Uigur script, only a small number is written in Tibetan or Brahmi script. Most of the Uigur Buddhist texts are translations from Chinese, Tibetan, Sanskrit, Tokharian or Sogdian languages. But there are also some examples of original Buddhist literature among the Uigur texts, for instance fragments of the Buddhist alliterative poetry and the Buddhist lay literature.

Besides these numerous Uigur Buddhist texts there are some fragments of Christian-Nestorian literature in Syrian script. Among the non-religious texts we find medical texts, astronomical texts and a group of economic documents. While the study of these economic texts is of a great importance for the history of the Uigur kingdoms, the edition of this text group poses great difficulties because of the cursive script and the very fragmentary size of these texts.

The discoveries of the four German Turfan expeditions were sent home in wooden boxes. Already at the site of discovery the manuscripts were wrapped in paper by the members of the expedition. On the top of every package a notice was placed of the contents together with a abbreviation signifying the number of the expedition and the site.

When these packages were opened in Berlin, this abbreviation was noted on every fragment. Later the fragments were put between glass plates which were closed by a special kind of adhesive strip. On these glass plates a second more detailed label was glued. This is the way in which the fragments are preserved to this day.

For the first years after the expeditions, all materials were kept at the Indian Department of the Museum für Völkerkunde in Berlin. Later on, in 1926, the texts were separated and the responsibility for the preservation, conservation and edition was given to the Prussian Academy of Sciences until the such time when the publication of these materials would be finished. A systematical arrangement of the Turfan manuscripts, especially the Sanskrit texts was attempted, but never successfully completed. The first important publications of Turfan texts apart from those mentioned above are for instance: Central Asian Sanskrit texts in Brahmi script from Idikutshahri, Chinese Turkestan by Heinrich Stönner (1904), Fragments of the Sanskrit canon of the Buddhists from Idykutshari, Chinese Turkestan (1904); Fragments of a Sanskrit grammar from Chinese Turkestan by Emil Sieg (1907 and 1908); Fragments of buddhist dramas (1911) and Medical Sanskrit texts (1927) by Heinrich Lüders; Türkische Manichaica from Chotscho by Albert von Le Coq published in three volumes between 1911-1922; Türkische Turfan-Texte (I-VII) published by Willi Bang, Annemarie von Gabain and Gabdul Reshid Rachmati between 1929-1936.

During the last years of World War II (in autumn 1943) the Turfan collection was sent to various places in Germany for safekeeping, for instance to the mines at Winthershall, Solvayhall and Schönebeck/Elbe. Thus most of it was saved from destruction.

As result of the division of Germany after World War II, the Turfan collection was split as well. Those parts, which had been sent to places in what was to become the Soviet Zone of Occupation, were returned by the Soviet forces in August 1945 to the rooms of the Orientalische Kommission, Berlin, Unter den Linden. This institution was newly constituted as Institut für Orientforschung at the newly founded Deutsche Akademie der Wissenschaften zu Berlin. Other parts of the collection, from places in the western occupation zones, for instance Ansbach, were handed over to the Mainzer Akademie der Wissenschaften und Literatur only in 1947. Because of the absence of specialists for several groups of the manuscripts this collection was split again. A number of manuscripts (especially Iranica) were sent to the Hamburg University where Wolfgang Lentz was working. Another part, mostly Sanskrit texts, was sent to Göttingen.

About one third of the fragments had to be restored at this time because the glass plates were broken during the various transports. A further task was restoration of order and the cataloguing of the various parts of the collection.

In Mainz Sinasi Tekin wrote a provisional catalogue of all fragments preserved there. In the meantime those fragments had been given a supplementary new shelf number "Mainz" with a running number for each single fragment. Although this catalogue is unpublished, copies of it are still used today.

Further work on the large group of Sanskrit texts was done for the first years after World War II in cooperation between Göttingen under the direction of Ernst Waldschmidt and Berlin, where his pupil, Dieter Schlingloff, was working up to 1961 and where the main part of fragments of this text group is preserved still today. Also under the direction of Ernst Waldschmidt and with the support of the Akademie der Wissenschaften zu Göttingen several publication series concerning the Sanskrit Turfan texts were founded, for instance Sanskrittexte aus den Turfanfunden and Sanskrit Wörterbuch der buddhistischen Texte aus den Turfan-Funden. Soon after the foundation of the project of a union catalogue of oriental manuscripts in Germany (Katalogisierung der orientalischen Handschriften in Deutschland), Ernst Waldschmidt and his colleagues began to prepare a catalogue of Sanskrit fragments. This work is still continuing today, now under the direction of Heinz Bechert. Up to now 7 volumes of the catalogue Sanskrithandschriften aus den Turfanfunden have been published. A supplementary volume by Lore Sander deals with palaeographic aspects of the Sanskrit manuscripts of the Berlin Turfan collection.

In 1960 Mary Boyce published her Catalogue of the Iranian Manuscripts in Manichean Script in the German Turfan Collection.

More than sixty years ago, F.W.K. Müller, the founding father of the Turfan studies in Berlin, used to give a piece of advice to students who were privileged to attend his lectures at Berlin University. He admitted with some resignation that it is impossible for an ordinary European scholar to understand a Buddhist text thoroughly without the help of a Japanese buddhologist or a Buddhist priest from East Asia. So it was a great luck for the Turfan research team of the former Academy of Sciences of the GDR, that they were given for more than 20 years the opportunity of cooperation with Japanese colleagues, especially of the Dunhuang-Turfan Research Staff in the Institute of Buddhist Cultural Studies of Ryukoku University. One of the results was the publication of two volumes of the Katalog chinesischer buddhistischer Textfragmente.

In 1971 in continuation of the Türkische Turfantexte, a new series Berliner Turfantexte was initiated at the former Institute of Ancient History and Archaeology of the Academy of Sciences of the GDR. 17 volumes of this series including editions of Iranian, Turkish, Tibetan, Chinese and Mongolian Turfan texts have been published up to now. Smaller editions and articles were published in the research journal Altorientalische Forschungen.

In 1956 the Iranian texts from Hamburg were returned to Mainz. Later the so called "Mainz fragments" were handed over to the Staatsbibliothek - Preußischer Kulturbesitz Marburg. In the seventies the State Library including the Turfan collection moved to Berlin.

In 1987 the first volume of a catalogue of Turkish Turfan texts including 269 fragments of the Mainz collection was published by Gerhard Ehlers (VOHD 13,10). After the reunification of Germany in 1989 the Turfan collection was reunited as well and is now being cared for by the Oriental Department of the State Library at Berlin. Two research groups are working now on the Berlin Turfan texts, one editing Turkish and Iranian texts, one cataloguing these two text groups.

An English colleague, Nicholas Sims-Williams, is preparing a catalogue of the Sogdian fragments in Nestorian script. Further catalogues of Turkish Turfan texts are being prepared.