KARABALGASUN ii. The Inscription



ii. The Inscription

The trilingual inscription at Karabalgasun, in Old Turkic, Sogdian, and Chinese, of the eighth Uighur qaghan (r. 808-21 CE) in Mongolia commemorates the qaghan’s (Old Turkic ḵaḡan, qaḡan) own military achievements and those of his predecessors, as well as their adoption and support of the Manichean religion and church. It is one of the most important sources for the history of the Uighur Steppe Empire (744-840 CE) and the study of Manicheism in China and Central Asia, in particular because many of the historical events recorded in the inscription are only known from it.

The fragments of the granite stele that bore the inscription are found scattered on the ground ca. 500 meters to the south of the palace remains. In the light of the old map made by W. Radloff (plate XXVII), the spot where the fragments were discovered appears to be a courtyard of a large building complex. In the late 19th and early 20th centuries, when several European and Japanese expeditions visited the site, they discovered some forty fragments of the stele, of which about twenty carried the inscription. Curiously, when the Japanese team surveyed the stones in the 1990s, several stones bearing Chinese characters were missing, and their present whereabouts is not known (Moriyasu, Yoshida, and Katayama, pp. 209-10).

The Chinese and Sogdian versions of the inscription occupy three sides—one wide and two narrow—of the stele, and the other wide side is engraved in Old Turkic in Runic script. The latter, since it carries only one language, has been regarded as the front and main face. The Chinese text occupies the left half, the Sogdian, the right half. Columns of the Chinese text run from right to left, those of the Sogdian from left to right. Column XX of the Chinese version is inscribed on a narrow space made by removing the corner of the stone; and columns XXI-XXIV are on the narrow side. If the Chinese text occupied the entire narrow side, the number of the columns could have been about 34. Both the Chinese and the Sogdian texts are inscribed vertically, whereas the Old Turkic is written horizontally from right to left.

Among the three versions the Chinese is the best preserved; of the Old Turkic text only a few small fragments have survived, and their original placement on the stele cannot be determined. The Sogdian version is preserved relatively well. In the latest edition published by Moriyasu and Yoshida (Moriyasu, 2003, figures 1 and 2), each Chinese column is estimated to comprise 90 characters and be ca. 400 cm in height. The wide sides measure ca. 180 cm, the narrow sides ca. 70 cm. The stele was placed on the back of a stone tortoise (Moriyasu, Yoshida, and Katayama, p. 214; not on a crouched lion, as assumed by Radloff and later scholars) and was capped by a dragon surrounding a shield-shaped tablet. A round object like a ball was attached on the top. For the reconstruction see A. Heikel (plate III, facing p. X).

The inscription was first discovered in 1889 by N. M. Yadrintsev; he brought two fragments back to St. Petersburg, of which the Chinese texts were published in 1891 by E. Koch. In 1890 A. Heikel visited the site, and he published the photographs in 1892. In the same year Radloff also reproduced rubbings of the inscription in his Atlas, plates XXXI-XXXV. Subsequent studies of the inscription have been based mainly on these two works, in particular on the latter. The most studied version of the inscription is the Chinese. After preliminary editions by G. Devéria and Wassiljew (apud Radloff, 1895, pp. 286-91), G. Schlegel edited the text and adduced a translation and commentary in 1896. Columns VIII-X were re-edited and extensively commented upon by E. Chavannes and P. Pelliot (1913), who were also able to make use of the molds prepared by B. de Lacoste in 1909. The fact that the new religion mentioned in the text is Manicheism was first discovered by J. Marquart (1898).

In Japan and China several rubbings of the inscription have been available, and the Chinese text has been edited several times—most recently by Moriyasu and Yoshida (for the others, see Lin Meicun et al.). These editions all are without translation. Thus, at least among Western non-Sinologists, Schlegel’s now totally obsolete edition is still consulted.

Study of the Sogdian version was begun already in 1891 by Radloff, who mistook the language for Uighur (Radloff, 1891 p. LXXXV). F. W. K. Müller (1909) recognized the Sogdian language and transcribed and translated several expressions. In 1930 O. Hansen edited the entire text, and his edition had not been revised until 1988, when Yoshida published the text and translation afresh (cf. also Yoshida, 1990 for the English summary). Since the 1950s J. Hamilton had been revising the three versions, and his interim report was made public in 1990. Yoshida (apud Moriyasu et al., 1998, pp. 215-19) revised the texts of those fragments which he examined in situ.

The Old Turkic version was studied first by Radloff (1885, pp. 291-95), who published the text without translation. Radloff’s edition was only slightly revised by H. N. Orkun (pp. 37-39). Much later Moriyasu (apud Moriyasu et al., 1998, pp. 219-24) published his reading and translation basing on the rubbings made by himself. On this occasion he discovered that Fragments 7 and 9 of Hansen’s edition come from the same stone as his Runic text 7c, which is reproduced on Atlas plate XXXV-6 (Radloff, 1895, p. 293, Ui, d).

The first line of the stele, which is copied in the shield-shaped tablet on top, names the qaghan to whom it was dedicated: Ui. bu tängrikän [ay] tängridä bulmïs alp bilgä qaγan [. . . bitidimiz] “[We have written] this [inscription in praise of] the god-like qaghan (by the name of) Ay Tängridä Qut Bulmïš Alp Bilgä”/Chin. Jiu xing hui gu ai deng li luo gu mo mi shi he pi jia ke han sheng wen shen wu bei bing xu 九姓迴鶻愛登里囉汨没蜜施合毗伽可汗聖文神武碑幷序 “Inscription accompanied by preface (in praise) of Ay Tängridä Qut Bulmïš Alp Bilgä Qaghan of the Uighurs consisting of nine tribes for his sacred scholarship and divine martial virtue”; Sogd. ʾyny ʾʾy tnkry-δʾ xwtpwl-mys ʾl-p pyl-kʾ βγy ʾwyγwr xʾγ-ʾn γwβty-ʾkh ptsʾk np’x(š)[tw δʾrym] “[We have] written this monument(?) in praise of the lord, the Uighur Qaghan Ay Tängridä Qut Bulmïs (sic) Alp Bilgä.” This ruler was the eighth qaghan of the empire, and the inscription was established either during his reign (Chavannes and Pelliot, p. 180; Hamilton, p. 125) or shortly after his death in 821 CE (Haloun apud Minorsky, pp. 286, 300). The wording of the first lines seem to indicate close parallel between the Sogdian and the Old Turkic versions, from which the Chinese is independent.

This opening is followed by the titles and names of high officials of the court, headed by an inner prime minister (Chin. I-II; Sogd. 1-3). Then there is a brief mention of the remote past and the achievements of the first and the second qaghans (Chin. III-VI; Sogd. 4-7). Usurpation of the second Tujue Qaghanate by the joint force of the Uighurs, Basmils, and Qarluqs in 742 CE is also mentioned here. The Sogdian version attests the indigenous name of the Tujue royal clan so far only known from the Chinese transcription a shi na 阿史那: ʾʾšnʾs knty twr(k) ʾxšʾwnδʾr ʾ(ʾ)st(ʾn)t xwty ʾxšʾwnδ’rt ʾkrtʾnt “they (i.e., Uighurs, Basmils, and Qarluqs) arrested the ruler of the Turks of Ashinas tribe, and they themselves became rulers” (line 6).

Involvement in the An Lushan (q.v.) rebellion and the subsequent introduction of Manicheism as the Uighurs’ state religion in 763 CE during the third qaghan’s reign (r. 759-79 CE) are recorded in Chin. VI-X (Chavannes and Pelliot, pp. 177-199; Clark, pp. 87-88), Sogd. 8-13. In column VIII of the Chinese version, after the description of the third qaghan’s sojourn in Luoyang, Manichean monks are reported to have entered the territory of Uighurs; this passage is assumed to be the account that marks the qaghan’s conversion. On the dating see Clark, who prefers to date the conversion earlier than 763. M. Ishida improved upon Chavannes and Pelliot’s text by recognizing mo xi xi de [黙]傒悉徳 = M.Ir. mahistag “elder, presbyter” in column VIII (Ishida, p. 171). A Manichean Uighur text describing the dispute among the Uighurs about the adoption of the religion is also known (W. Bang and A. von Gabain; Clark, pp. 101-4). On the historical background see also Clark’s detailed discussion. This is followed by the very short reference to the fourth, fifth, and sixth qaghans’ enthronement (Chin. XI; Sogd. 13-14). In line 12 of the Sogdian version the contemporary možak “teacher, the leader of the Eastern diocese” is referred to as βγy mʾr nywrwʾn “god-like lord New-Ruwān.” In the Chinese version he appears as fawang 法王 (column X) “king of law.” He also seems to be mentioned in a Manichean Uighur document of the church history (Moriyasu, 2002, pp. 134-57). On another late Uighur text expounding the introduction of Manicheism among the Uighurs, see P. Zieme.

The rest of the inscription mainly records military achievements of the seventh qaghan (r. 795-808) and of the eighth qaghan before he mounted the throne. The seventh qaghan did not originate from the Uighur royal clan by the name of Yaghlaqar. The circumstances in which he came to succeed the throne are described in Chin. XIII and Sogd. 16. The latter reads as follows: pr sʾt pw(yrw)xty xwy-štr ʾyl ʾwkʾsy ʾl-pw xwtl-wγ t(yk)ʾyn nʾm δʾβr ZY xwt(y)[ M](N) ʾʾz-y mrtsʾr MN sʾt ʾyδʾyty yxwstʾy ZY ʾnyʾz-ʾnk (w)mʾt “He (the sixth qaghan?) gave the name (and title) ‘Prince Alp Qutluγ’ to the state councilor who was a chief of all the ministers. Since his birth he himself was distinguished and different from all the people.” In the text the seventh qaghan differs from the others in that only he is referred to as Tian Ke Han “heavenly qaghan” in Chinese and prnpδy/prnxwntʾk ʾxšywnʾk “fortunate sovereign” in Sogdian. As the supporter of Manicheism he is compared with Angel Jacob (Sogd. 17: yʾkwβ βrʾy-štʾk), and his patronage of the Manichean churches and monks is also mentioned (Chin. XXII, Sogd. 22-23).

The seventh qaghan’s campaign was first directed against Qïrγïz in the north of the Mongolian plateau (Chin. XIII-XIV, Sogd. 18) and later toward the southwest against Tibetans and Qarluqs; after the siege of Beiting or Bešbalïq, present-day Jimsa, he proceeded to Kucha and destroyed the Tibetan army in Yushu, located some 30 km to the west of Karashar (Chin. XV-XVI, Sogd. 18?-19). The qaghan pursued the Qarluqs and reached the Syr Darya. On another occasion he attacked the Qarluqs and Tibetans and chased them westward to Ferghana (Chin. XVII and XX, Sogd. ?). The Sogdian version (line 19) seems to refer to the area around Karashahr as ctβʾr twγrʾystn “four Tughri countries” (not twγrʾkcʾny, as generally assumed). On twγry see Henning, who cites and discusses this passage in connection with the nomenclature of “Tocharian” (Henning, p. 550). The Chinese chronicles date the battle for the control of Beiting to the years around 790 CE (Moriyasu, 1981). In the light of contemporary Khotanese documents one may date the battle in Kucha to 798 CE (Yoshida, 2009).

The Uighurs’ persistent attack against the Qarluqs in the north of the Tianshan range brought them westward to the frontier of the Abbasid empire. Although there is damage to the Chinese version (column XIX), it does make reference to the realm of Tajiks (tʾz-ykʾnʾk ʾxšʾwʾn), the Khorasan Amir (xwrʾsʾn xmʾyr), and the caliph (mwmyn xmʾyr < Arabic amir al-muʾminin) in lines 20-22. The inscription claims that even the Abbasid rulers paid homage to the qaghan. It is difficult to substantiate this claim by referring to the Islamic sources, where the Uighurs are called Toghuzghuz. One might be able to consider it as connected with the revolt of Rāfiʿ b. Layṯ in Samarqand (ca. 806-10), during which he secured aid from many sides, including Toghuzghuz, Qarluqs, and the armies of Tibet (Minorsky, pp. 301-2). It is to be noted that in line 23, the qaghan is described as not having left the world (i.e., died), and the caliph in question is most likely to be Hārun al-Rašid (r. 786-809).

What Hansen calls Fragments 6 and 9 are in fact come from one and the same stone, and Fragment 9 belongs to the narrow side (Moriyasu et al., p. 210). Its text is estimated to correspond to lines *32-*43. Mention of Alp Bilgä Qaγan in line *41 is compatible with the assumption that considerable space is dedicated to the recording of his own achievements after the seventh qaghan’s death.



W. Bang and A. von Gabain, “Türkische Turfan-Texte, II. Manichaica,” SPAW, 1929, pp. 411-30.

E. Chavannes and P. Pelliot, “Un traité manichéen retrouvé en Chine, traduit et annoté (deuxième partie), Journal Asiatique, 10th ser., 20, 1913, pp. 99-199 (in particular pp. 177-99).

L. Clark, “The Conversion of Bügü Khan to Manichaeism,” in R. E. Emmerick et al., eds., Studia Manichaica. IV. Internationaler Kongreß zum Manichäismus, Berlin, 14.-18. Juli 1997, Berlin, 2000, pp. 83-123.

G. Devéria, “Transcription, analyse et traduction des fragments chinois du second et du troisième monument,” in Société Finno-Ougrienne, Inscription de l’Orkhon, Helsinki, 1892, pp. XXVII-XXXVIII. 

J. Hamilton, “L’inscription trilingue de Qara Balgasun d’après les estampages de Bouillane de Lacoste,” in A. Haneda, ed., Documents et archives provenant de l’Asie Centrale, Kyoto, 1990, pp. 125-33.

O. Hansen, “Zur soghdischen Inschrift auf dem dreisprachigen Denkmal von Karabalgasun,” in Journal de la Société Finno-Ougrienne 44/3, 1930.

A. Heikel, “Les monuments près de l’Orkhon,” in Société Finno-Ougrienne, Inscription de l’Orkhon, Helsinki, 1892, pp. VII-XIII, Tabellen 44-63.

W. B. Henning, “Argi and the ‘Tocharians’,” BSOS 9, 1938, pp. 545-71.

M. Ishida, “Tonkō hakken ‘Manikōbutsu kyōhō giryaku’ ni mietaru nisan no gengo ni tsuite” (Some foreign words found in the ‘Compendium of the teachings of Mani, Buddha of Light’ discovered in Dunhuang), in H. Ikeuchi, ed., Shiratori hakushi kanreki kinen tōyōshi ronsō (Papers on Oriental history in honor of Dr. Shiratori on the occasion of his 60th birthday), Tokyo, 1925, pp. 285-98.

E. Koch (tr. from Russian by P. Lemosof), “Deux pierres avec inscriptions chinoises,” T’oung Pao 2, 1891, pp. 113-124.

Lin Meicun et al., “Jiu xing huigu kehan bei yanjiu” (New studies on the Chinese version of the trilingual inscription of the Uighur Khanate), Eurasian Studies/Ouya yanjiu 1, 1999, pp. 151-71.

J. Marquart, “Historische Glossen zu den alttürkischen Inschriften,” Wiener Zeitschrift für die Kunde des Morgenlandes 12, 1898, pp. 157-200.

V. Minorsky, “Tamim ibn Baḥr’s journey to the Uyghurs,” BSOAS 12/2, 1948, pp. 275-305.

T. Moriyasu, “Qui des Ouïgours ou des Tibétains ont gagné en 789-792 à Beš-balïq?” Journal Asiatique 269, 1981, pp. 193-205.

Idem, “Uiguru kara mita Anshi no ran” (The rebellion of An Lu-shan [755-763] from the Uighurs’ viewpoint [with special reference to an Uighur document Mainz 345]), Studies on the Inner Asian languages 17, 2002, pp. 117-70.

Idem, “Four Lectures at College de France in May 2003.

History of Manicheism among the Uighurs from 8th to 11th centuries in Central Asia”/“Korëju do furansu kōenroku. Uiguru-manikyōshi tokubetsukōgi),” in World History Reconsidered through the Silk Road/Sirukurōdo to sekaishi, Osaka, 2003, pp. 23-111, maps 1-8, figures 1-3.

T. Moriyasu, Y. Yoshida, and A. Katayama, “Kara-Barugasun hibun” (Qara Balgasun Inscription), in Moriyasu and Ochir, pp. 209-27, plates 14a-14q.

T. Moriyasu and A. Ochir, Mongorukoku genzon iseki, hibun chōsa kenkyū hōkoku (Provisional report of researches on historical sites and inscriptions in Mongolia from 1996 to 1998), Osaka, 1999.

F. W. K. Müller, “Ein iranisches Sprachdenkmal aus der nördlichen Mongolei,” Sitzungsberichte der Preußischen Akademie der Wissenschaften, Phil.-hist. Klasse, 1909, pp. 726-30.

H. N. Orkun, Eski Türk Yazïtlarï II, Istanbul, 1939.

W. W. Radloff, Das Kudatku Bilik des Jusuf Chass-Hadschib aus Bälasagun, Teil 1., St. Petersburg, 1891.

Idem, Atlas der Althertümer der Mongolei. Arbeiten der Orchon-Expedition, 1. Lieferung, St. Petersburg, 1892.

Idem, Die alttürkischen Inschriften der Mongolei, 3. Lieferung, St. Petersburg, 1895.

G. Schlegel, Die chinesische Inschrift auf dem uigurischen Denkmal in Kara Balgassun, übersetzt u. erläutert, Mémoires de la Société Finno-Ougrienne IX, Helsinki, 1896.

Y. Yoshida, “Karabarugasun hibun no sogudogoban ni tsuite” (On the Sogdian version of the Karabalgasun inscription), Bulletin of the Society for Western and Southern Asiatic Studies, Kyoto University/Seinan ajia kenkyū 28, 1988, pp. 24-52.

Idem, “Some New Readings of the Sogdian Version of the Karabalgasun Inscription,” in A. Haneda, ed., Documents et archives provenant de l’Asie Centrale, Kyoto, 1990, pp. 117-23.

Idem, “Karabalgasun Inscription and the Khotanese Documents,” in D. Durkin-Meisterernst, Ch. Reck, and D. Weber, eds., Literarische Stoffe und ihre Gestaltung in mitteliranischer Zeit, Wiesbaden, 2009, pp. 349-60.

P. Zieme, “You guan monijiao kaijiao huigude ijian xinshilio” (A new Manichean Uighur document concerning the adoption of Manicheism by the Uighurs), Journal of Dunhuang studies/Dunhuang xue jikan, no. 65, 2009, pp. 1-7.

In this entry the Sogdian text is cited from Y. Yoshida’s edition, in preparation.

(Y. Yoshida)

Originally Published: December 15, 2010

Last Updated: April 24, 2012

This article is available in print.
Vol. XV, Fasc. 5, pp. 530-533