MÜLLER, Friedrich W. K.


MÜLLER, Friedrich W. K. (b. January 21, 1863 in Neudamm, present day Debno in Poland; d. April 18, 1930 in Berlin), a scholar of oriental cultures and languages who, confronted with the texts discovered by the four German expeditions to Central Asia (1902-1914), was able to make groundbreaking discoveries about many languages hitherto almost unknown and to make a major contribution to the establishment of the philological and historical study of texts in Middle Iranian and Old Turkish and the study of original Manichaean texts. He also contributed significantly to the discussion on Tocharian as well as the study of Central Asian Buddhist and Christian literatures.

F. W. K. Müller spent most of his life in Berlin where he studied from 1883 to 1889 and where he worked in the Ethnological Museum from 1887 to 1928. In the year 1901 he undertook a journey to China, Korea, and Japan for the Museum but the success of the first German Turfan expedition (1902-1903), organized by the Museum, in terms of texts sent back to Berlin, took his attention away from the ethnological and linguistic studies mostly on cultures and languages of the Indo-Chinese, Malay, Tibetan, Chinese and Japanese areas to which he had devoted himself up to then. His background in the Semitic languages allowed him to recognize the Manichaean script of many of the fragments as a close relative of Estrangelo. In these texts he recognized remnants of original Manichaean literature in Middle Iranian (Middle Persian, Parthian, and Sogdian), Early New Persian, and Old Turkish. From this starting-point he embarked on research into these and some other hitherto mostly unknown languages. The results appeared in publications of the Prussian Academy in Berlin over nearly thirty years. He, together with Albert von Le Coq, Karl Foy, Vilhelm Thomsen, Gabdul Rašid Rachmati, Willy Bang, Emil Sieg in Göttingen, Hubert Jansen and followed at a distance by Friedrich Carl Andreas and Salemann in St. Petersburg, formed the first generation of scholars working on the Turfan Collection in Berlin; scholars of the second generation such as Wolfgang Lentz and Annemarie von Gabain studied and worked with him.

Müller became assistant director in the Museum in 1896, and its director in 1907, the same year in which he also became a member of the Prussian Academy of Sciences. From 1904 on, most of his publications were in the proceedings (Sitzungsberichte and Abhandlungen) of the Academy. Though some of these studies were also published separately as short monographs, Müller published no books or surveys (Lüders, 1931, p. 1). All his publications are in German.

Müller’s work on Iranian. Müller is the founder of the study of the Parthian and Sogdian languages as well as of Manichaean Middle Persian. Until then, Middle Persian was known only through its Zoroastrian literature. Though he only distinguished Sogdian on the one hand from the Middle Persian and Parthian material on the other, without being able to divide the latter into Middle Persian and Parthian, he was able to provide reasonable translations of the texts he edited in 1904, after an initial short note in the same year. His discoveries had fundamental implications. They included his identification of the language of most of the fragments as Western Middle Iranian and the demonstration that the literature was Manichaean by comparing one fragment with a passage in Ebn al-Nadim’s Fehrest (See FEHREST iii. THE REPRESENTATION OF MANICHEISM IN THE FEHREST) about the Manichaeans and the identification of some fragments as belonging to the Šābuhragān, an original work by Mani in Middle Persian (Durkin-Meisterernst, 2004). Subsequently, in 1905, he published a Middle-Persian fragment of ‘the Shepherd of Hermas’ (See HERMAS) and then devoted his energies in large part to texts that had arisen in Central Asia in the languages of the area as opposed to the texts in the church languages Middle Persian and Parthian. The only exceptions are the Middle Persian texts such as T II D 135 (now MIK III 36) (“Hofstaat,” 1912) and M 1 (Mahrnāmag, 1913) that shed light on local social history through a large number of names and titles of Sogdian and Uigur Manichaeans. In concentrating on Central Asia he broadened his scope beyond Manichaean literature to Christian and Buddhist texts not least because of the help this translation literature provided for the decipherment of Sogdian and Old Turkish languages. He also turned to original texts for the information they provided and established that an Uigur stake inscription gave the date for the conversion of a Manichaean site in Khocho into a Buddhist temple, therefore dating the demise of Manichaeism there to the 8th century (“Zwei Pfahlinschriften,” 1915, pp. 4-5 and 26). However, this date has turned out to be much too early: Hamilton and Moriyasu have since dated the inscription to 1008 (Moryiasu, 2001, p.152 and n.14).

His first publication of a Sogdian text was in 1904 when he presented the bilingual Middle Persian and Sogdian text in M 172 I (“Handschriften-Reste II,” pp. 100-103 with some initial observations there and on p. 110 with reference to Andreas; on p. 111, ‘Sogdian’ is used to characterize this language). In 1905 Edward Sachau published a page from a Christian Sogdian text written in Syriac script with some additions and established (pp. 976-77) a similarity or possible identity with the language of the Sogdian text published by Müller. The page belonged to a find of Christian Syriac, Sogdian, Turkish, and even some Middle Persian texts in Bulayïq. In these texts Müller identified some pages as translations from the New Testament on the basis of Syriac captions in the Sogdian texts and published these in 1907, making use of the known Christian text to elucidate the Sogdian version. In 1909 Müller, as if by chance, as he describes, was able to establish that the trilingual Karabalgasun inscription which described the adoption of Manichaeism by the Uigurs contained a Sogdian version which Radloff had thought was Uigur. With access to genuine Sogdian material in well-differentiated scripts, he was able to recognize forms unclear to others. This discovery confirmed and extended the range and role of Sogdian and Sogdian-speakers in Central Asia. In 1912 he published a series of Christian Sogdian texts; the Buddhist texts that he had intended including were postponed and were finally published posthumously (in Müller-Lentz, 1934).

In his work on the Iranian texts, Müller had the cooperation of Andreas who was given a set of photographs of a considerable part of the collection. Andreas’ contribution to Müller’s work is hard to evaluate since much of Andreas’ work remained unpublished during his lifetime but he and Müller corresponded regularly. In Berlin H. Jansen worked for some years on the Iranian materials and compiled word-lists of the Western Middle Iranian and Sogdian fragments (“Sogdische Texte I,” p. 91 and footnote where only Middle Persian and Sogdian are mentioned). One of Andreas’ students, W. Lentz, became a close collaborator on Sogdian and published “Sogdische Texte II” (1934) from Müller’s papers. Part of the material in “Sogdische Texte II” had already been included for publication in “Sogdische Texte I” but was taken out at a late stage (“Sogdische Texte I,” p. 89 and “Sogdische Texte II,” p. 505).

Individual Middle Persian and Parthian texts and fragments published by Müller have over the years been re-worked by others but no systematic review of “Handschriften-Reste II” has been made since it contains for the most part only some of the fragments belonging to a particular text. All the Christian Sogdian material published in “Sogdische Texte I” was thoroughly revised by Werner Sundermann in three articles (1974-1981). In these articles, Sundermann supplied more material belonging to this group of texts and, in contrast to Müller, reconstructs the book to which the fragments belonged. This was a pericope that contained New Testament texts in a different order to that chosen by Müller. Thanks to the great advances made in the study of Sogdian since Müller’s publications (and also, of course, because of them) Sundermann was able to add a great deal of precision to the editions and pay particular attention to the Syriac original. Nicholas Sims-Williams (1991) gave a survey of the Christian Sogdian material, including Müller’s work on it (p. 120-21).

Müller’s work on Old Turkish. In “Uigurica [I],”(p. 3) Müller says that he first turned, together with Foy, to Manichaean-Turkish texts for help with Manichaean terminology. After Foy’s early death Le Coq continued to work on the Turkish texts but Müller soon joined him in a series of publications that were to concentrate on the Buddhist material for which the Chinese originals were available. He started, though, with a Christian text found together with the Sogdian Christian texts that had allowed him to make progress in that language (“Uigurica [I],” first text ‘The Adoration of the Magi’) before turning in the same publication to fragments of the Buddhist Sutra of Golden Light, the Sanskrit Suvarṇaprabhāsasūtra, the Turkish title of which, Altun yaruq, he was able to identify (p. 10-11) and then to find the corresponding parts of the text in the Chinese version (in Yijing’s translation). In additional notes, amongst other things, he turns to Mongolian to establish successfully the meaning of individual words, thereby (p. 47) showing the indebtedness of Mongolian to Turkish and Sogdian for its Buddhist vocabulary, and presents his interpretation of part of the stake-inscription he was later to publish in full in 1915. In “Uigurica II” he continued with other Buddhist texts for which he was often able to identify the Chinese originals; on occasion he refers to Tibetan versions (e.g. pp. 51-52). Apart from translating the texts Müller was able, partly in public discussion with Radloff, to introduce points of Turkish grammar which he illustrated from the new material and, with the help of the texts, to supply the facts that made sense of previously known but only tentatively understood references, particulary in Chinese sources. In “Uigurica III” he published fragments of a text he identified as the Daśakarmabuddhāvadānamālā. The Uigur text in transcription and the German translation are placed in parallel columns. Here no Chinese text is given because a direct original of the Uigur version was not found; as the colophon says, the Uigur text was translated from Tocharian (Müller and von Gabain, 1931, p. 678). The last edition prepared by Müller was published from his papers by von Gabain as “Uigurica IV.” It continues the publication of avadāna fragments begun in “Uigurica III,” but von Gabain was able to establish that the correct title of the work is Daśakarmapathāvadānamālā (p. 679). The text is presented in transcription and translation on facing pages.

Von Gabain (1991) and Zieme (1991, p. 134-5) gave overviews on Müller’s work on Uigur and his methododology. As von Gabain points out (1991, p. 102), Müller was very well versed in the Chinese material, was able to identify the Chinese originals and, indeed, often suggested that a retranslation of a difficult Uigur text into the Chinese was the only way to make the sense clear. It is worth mentioning that Müller’s work on The Sutra of Golden Light has retained its importance to the present day because some of the fragments that he worked on were lost in the Second World War and therefore his transcriptions in his papers are the fullest available record of these lost pages (Raschmann 1998, p. 297).

When the text fragments recovered by the first expedition arrived in Berlin in 1903, Müller developed a simple system of signatures composed of M (for Manichaean) and consecutive numbers from 1 to 919 for the fragments he recognized as Manichaean on the basis of the script or some formal aspects. Each signature was written on a label placed on the small glass sheets between which the fragments were fixed. As Boyce (1960, p. xxi) describes, Müller numbered the fragments as he worked on them, naturally taking the better preserved and more legible ones first. He discarded the indicators of the find-sites that had been written in the field on the small packages into each of which a number of fragments had been placed. This has since proved to have been an unwise decision since the precise location of some of the fragments from the first expedition are now no longer known whereas this information is generally available for the finds of the other three expeditions and can be relevant for deciding if separate pages could have belonged to the same book. Von Le Coq was responsible for putting the Turkish fragments in order and giving them signatures. In the case of the fragments recovered in the other three expeditions, the find-site was simply transferred to each fragment of a group. Lists were compiled but these were lost. A system of unique signatures for each fragment was not developed until after the Second World War.

Formal aspects of Müller’s editions. In publishing the fragments in Manichaean script, Müller adopted a mixed transliterative and transcriptional system in Latin script, supplying vowels not indicated by the Manichaean script and marking with a cicumflex those that are indicated: For example, what in the original is written xwmr mstyft appears as khvamar mastêf (“Handschriften-Reste II,” p. 53, second line). Though necessarily subjective and perhaps slightly confusing in the choice of the circumflex to mark a written vowel (which need not always be long), Müller’s careful system was fully convertible. Even when he unnecessarily distinguished two forms of the letter as č and ǰ (“Handschriften-Reste II,” p. 5) his consistency ensured that his texts were fully convertible. Salemann 1908, preferring a non-interpretative transliteration, converted Müller’s mixed system into a Hebrew transliteration, to Müller’s annoyance. Salemann’s system rapidly gained acceptance, though in 1934 Walter Bruno Henning (q.v.) transferred it from Hebrew to Latin script. For the slightly modified Syriac script of the Christian Sogdian texts, Müller chose in “Sogdische Texte I” (1913) to present the text in two versions: in the original Syriac font as well as in a mixed Latin transliterative and transcriptional system similar to that he which he had used for the Manichaean script. In 1926 in editing a Buddhist Sogdian text, Müller used a Latin system only, this time for the most part based on Gauthiot’s transliteration. In the Müller-Lentz edition of 1934, Lentz abandoned Müller’s modifications to Gauthiot’s system but added some refinements (p. 506ff).

Müller also used Manichaean type in some of his publications, most notably in his edition of Mahrnāmag (1913). Similarly, he and von Le Coq used Uigur type in their publications of Uigur texts. In the Turkish texts translated from Chinese he included the Chinese in a column next to the text. Generally, however, he did not investigate the precise relation of the Turkish to the Chinese text; as with the material in other languages he used the Chinese text to gain access to the translation on which he focused his attention.

In most of his works Müller preferred interlinear translations and did not usually include a glossary (except in “Sogdische Texte I,” where H. Jansen provided a list of words with references but without meanings). In “Sogdische Texte II” Lentz included a glossary with meanings, references and additional information. Similarly, in his editions of Turkish texts Müller did not include a complete glossary but rather lists or indices of interesting words.

International cooperation. Though an initial Russian-German cooperation had been agreed on before the first German Turfan expedition, it did not result in the direct sharing of texts. It is unfortunate that Müller felt unable to enter into a real cooperation with Salemann, who converted the texts from Müller’s mixed transcription into a pure transliteration in order to arrive at a clearer representation of what was actually preserved. Without access to the originals and without the correction that only Müller or someone else in Berlin could have made, Salemann’s work was held back. Müller (1909, p. 730, n.1) expressed his impatience with Salemann’s retransliteration of fragments first published by Müller and complained that Salemann worked on a description of the grammar of the texts instead of publishing more texts; a complaint that sounds curious today since Salemann was the leading specialist in the grammar of Middle Persian and presented in 1908 not only a glossary of the fragments but also a reverse index, an indispensable tool for anyone working on the fragments. He cooperated with Radloff at a distance and often with a caustic tone. He was involved in the identification and publication of some texts in the Stein collection in London (Müller 1926; in “Sogdische Texte II” Lentz lists relevant items in Müller’s papers).

Müller’s work on other languages. Prior to the Turfan fragments Müller worked on a large variety of Asian languages, including Chinese, Korean, Japanese, Malay and dialects of Siamese. His broad university training in the ‘classical’ oriental languages and the ability to work his way quickly into any language when he needed it stood him in good stead when confronted with the at first bewildering variety of scripts and languages recovered from Central Asia. His competence in Chinese and in Buddhist literature proved invaluable.

Müller’s work on other languages preserved in the Turfan fragments did not usually lead to an edition of the text. In the case of the Early New Persian text M 106, he listed the Arabic loan-words therein in “Handschriften-Reste II,” (p. 106); in 1915 he published an edition of a double page of the Syriac and New Persian texts of a fragment of a Psalter. He identified Bactrian fragments (von Gabain 1991, p. 101). His work with E. Sieg on the Uigur and Tocharian versions of Maitrisimit was an early recognition of the need for such a cooperation and one which, given the dependency of Uigur Buddhist literature on Tocharian, is rightly continued today. He also introduced the term ‘Tocharian’ for this as yet unknown and unnamed language. Though the term turned out to be inaccurate, it has been retained by most scholars (Krause, 1971, 5-8).



M. Boyce, A Catalogue of the Iranian Manuscripts in Manichean Script in the German Turfan Collection, Berlin 1960.

D. Durkin-Meisterernst, “F. W. K. Müller’s work on Iranian Languages,” in D. Durkin-Meisterernst, S. Raschmann, J. Wilkens, M. Yaldiz, P. Zieme, eds., Turfan Revisited, Berlin 2004 (forthcoming).

A. von Gabain, “Die erste Generation der Forscher an den Turfan-Handschriften,” in H. Klengel and W. Sundermann, eds., Ägypten Vorderasien Turfan, pp. 98-105.

Horst Klengel and Werner Sundermann, eds., Ägypten Vorderasien Turfan, Schriften zur Geschichte und Kultur des Alten Orients 23, Berlin, 1991.

W. Krause, “Tocharisch,” in Handbuch der Orientalistik. Abt.1, Band IV Iranistik, 3. Abschnitt, Leiden and Cologne, 1971.

H. Lüders, “Gedächtnisrede auf Friedrich Wilhelm Karl Müller,” SPAW 1931, Sonderabdruck.

T. Moriyasu, “Uighur Buddhist stake inscriptions from Turfan,” in L. Bazin and P. Zieme eds., De Dunhuang à Istanbul. Hommage à J. R. Hamilton, Silk Road Studies 5, Turnhout, 2001, pp. 149-223.

S.-C. Raschmann, “Aus den Vorarbeiten F.W.K. Müllers zum Altun Yaruk Sudur,” in J. P. Laut and M. Ölmez, eds., Bahşï Ögdisi. Festschrift für K. Röhrborn, Freiburg and Istanbul 1998, pp. 295-304.

E. Sachau, “Literatur-Bruchstücke aus Chinesisch-Turkestan,” SPAW, 1905, pp. 964-78.

C. Salemann, “Manichäische Studien I. Die mittelpersischen Texte in revidierter transcription, mit glossar und grammatischen bemerkungen,” Mém. de l’Acad. imp. des Sc. de St. Petersburg. VIII, ser. Cl. hist.-phil. VIII, No. 10, 1908.

N. Sims-Williams, “Die christlich-sogdischen Handschriften von Bulayïq,” in H. Klengel and W. Sundermann, eds., Ägypten Vorderasien Turfan, pp. 119-25.

W. Sundermann, “Nachlese zu F. W. K. Müllers ‘Soghdischen Texten,’” AoF 1, 1974, pp. 217-55; 3, 1975, pp. 55-90; 8, 1981, pp. 169-225.

F. M. Trautz, “Professor Dr. F. W. K. Müllers Veröffentlichungen von 1889-1924.

Zeitlich geordnet,” Asia Major 2, 1925, pp. xi-xvi.

Fr. Weller, B. Schindler, “F.W.K. Müller,” Asia Major 2, 1925, pp. vii-x.

P. Zieme, “Gedanken zur Bearbeitung der alttürkischen buddhistischen Texte,” in H. Klengel and W. Sundermann, eds., Ägypten Vorderasien Turfan, pp. 134-40.

P. Zieme et al., Turfanforschung, Berlin-Brandenburgische Akademie der Wissenschaften, Berlin 2002.

Bibliography of Müller’s works from 1904 on. (A bibliography of Müller’s works up to 1924 and including works published before 1904 and works on topics not covered here is to be found in Trautz, 1925.) “Handschriften-Reste in Estrangelo-Schrift aus Turfan, Chinesisch-Turkistan” [I], SPAW, 1904, 348-52; Handschriften-Reste in Estrangelo-Schrift aus Turfan, Chinesisch-Turkistan. II Teil, Anhang zu APAW, 1904; “Über die kultur- und sprachgeschichtliche Bedeutung eines Teils der Turfan-Handschriften,” Zeitschrift für Ethnologie 37, 1905, pp. 414-15; “Eine Hermas-Stelle in manichäischer Version,” SPAW, 1905, pp. 1077-83; “Beitrag zur genaueren Bestimmung der unbekannten Sprachen Mittelasiens,” SPAW, 1907, pp. 958-60; “Antrittsrede bei Aufnahme in die Königlich Preußische Akademie der Wissenschaften,” SPAW, 1907; “Die persischen Kalenderausdrücke im chinesischen Tripitaka,” SPAW, 1907, 458-65; “Neutestamentliche Bruchstücke in Soghdischer Sprache,” SPAW, 1907, pp. 260-70; “Uigurica (I). 1. Die Anbetung der Magier, ein christliches Bruchstück. 2. Die Reste des buddhistischen Goldglanz-Sūtra. Ein vorläufiger Bericht,” APAW, 1908; “Ein iranisches Sprachdenkmal aus der nördlichen Mongolei,” SPAW XXVII, 1909, pp. 726-30; “Uigurica II,” APAW 1911; “Der Hofstaat eines Uiguren-Königs,” in Festschrift V.Thomsen zur Vollendung des 70. Lebensjahres, Leipzig, 1912, pp. 207-13; “Soghdische Texte. I,” APAW, 1912, Berlin 1913; “Ein Doppelblatt aus einem manichäischen Hymnenbuch (Maḥrnâmag),” APAW 1912, Berlin 1913 (Figure 1); "Zwei Pfahlinschriften aus den Turfanfunden. 1. Die uigurische Inschrift auf dem Pfahle I B 4672. 2. Die chinesische Pfahlinschrift,” APAW 1915, Nr.3, 1915; “Ein syrisch-neupersisches Psalmenbruchstück aus Chinesisch-Turkistan,” in Festschrift E. Sachau, ed. G. Weil, Berlin, 1915, pp. 215-22; F. W. K. Müller and E. Sieg, “Maitrisimit und Tocharisch,” SPAW, 1916, pp. 395-417; “Toxri und Kuišan (Küšän),” SPAW, 1918, pp. 566-86; “Uigurische Glossen,” in Festschrift für F. Hirth (Ostasiatische Zeitschrift 8 (1919-20), Berlin 1920, pp. 310-24 ;"Uigurica III. Uigurische Avadāna-Bruchstücke (I-VIII),” APAW, 1922; “Reste einer soghdischen Übersetzung des Padmacintāmaṇi-dhāraṇī-sūtra,” SPAW,1926, pp. 2-8; “Eine soghdische Inschrift in Ladakh,” SPAW, 1925, Berlin 1926, pp. 371-2; “Ein uigurisch-lamaistisches Zauberritual aus den Turfanfunden,” SPAW, 1928, pp. 381-86; F. W. K. Müller and A. Von Gabain, “Uigurica IV,” SPAW, 1931, pp. 675-727; F. W. K. Müller and W. Lentz, “Soghdische Texte. II.” Aus dem Nachlaß herausgegeben von Wolfgang Lentz, SPAW 1934, pp. 504-607.

Müller’s works on Turkish were collected in Sprachwissenschaftliche Ergebnisse der deutschen Turfan-Forschung, Text-Editionen und Interpretationen von A.A. von Le Coq, F. W. K. Müller, W. Bang, A. von Gabain, G.R. Rachmati, W. Thomsen, Gesammelte Berliner Akademieschriften, 1908-1938; Mit Vorwort von G. Hazai. Band I, Leipzig 1972 (Opuscula III, 1, I). Nearly all of Müller’s works on Iranian were collected in: Sprachwissenschaftliche Ergebnisse der deutschen Turfan-Forschung. Text-Editionen und Interpretationen von F.W.K. Müller, A. A. von Le Coq, K. Foy, G. R. Rachmati. Gesammelte Berliner Akademieschriften 1904-1932. Mit 27 Tafeln. Teil 3, Leipzig 1985 (Opuscula III, 3).

Facsimiles of the texts edited by Müller are to be found in: G. Hazai, P. Zieme eds., Sprachwissenschaftliche Ergebnisse der deutschen Turfan-Forschung. Uigurica I - IV. Faksimiles zu den Text-Editionen von A. von Le Coq, F.W.K. Müller, W. Bang, A. von Gabain, G. R. Rachmati, V. Thomsen. Band 4, Leipzig 1983 (Opuscula III.4).

(Desmond Durkin-Meisterernst)

Originally Published: July 20, 2004

Last Updated: July 20, 2004