archeological site of a capital of the Uighur Khaghanate (second half of the 8th century to first half of the 9th century). Karabalgasun is located in the Orkhon valley, 320 km west of Ulan Bator (Ulaanbaatar), 30 km north of Karakorum.



i. The Site

Karabalgasun is located in the Orkhon valley, 320 km west of Ulan Bator (Ulaanbaatar), 30 km north of Karakorum. This capital was named variously in the Chinese sources: Huihu cheng “Uighur city” (Jiu Tangshu, chap. 145, p. 5213); Woluduo cheng “Ordu city” (Liaoshi, chap. 93, p. 1378); Buguhan cheng “Bögü Khan City” (Liaoshi, chap. 30, p. 356). It was mentioned also in the Islamic sources: madīnat of ḵāqān, malek of the Toghuzghuz “city of Khaghan (Old Turkic ḵaḡan, qaḡan), king of the Uighurs” in Tamīm b. Baḥr’s 9th-century report (Minorsky, pp. 279, 283); Ordū-bālīḡ “Ordu city” in Jovayni’s 13th-century History (part 1, pp. 40, 42, 192). Ordu means “royal camp” and Balïq means “town” in Old Turkic (Clauson, pp. 203, 335); this capital must have been called Ordu Balïq by the Uighurs during their period of rule there. It was the first large-scale city constructed by the nomadic states in Mongolia, although walled cities had been constructed by nomadic rulers in the Eurasian steppes since the 8th century. All of them had the same role as Ordu Balïq and the same characteristics described below.

The archeological site was surveyed by V. V. Radlov and D. A. Klements in 1891, by V. L. Kotvich in 1912, by S. V. Kiselev in 1949, and by a joint expedition of Japan and Mongolia in 1996. The remains of the city consist of a walled, rectangular fortress and a vast urban area outside its boundaries. The northern and the southern fortress walls are about 420 m in length, and the eastern and the western fortress walls are about 340 m (Hayashi and Moriyasu, p. 199). A well-preserved part of the wall is 7 m in height even now. The wall was constructed by erecting packed earth ramparts—the traditional Chinese method (see Chang).

At the southeastern corner of the fortress there stands a high platform, 12 m in height, which might have been a palace or a donjon. In the western part of a fortress there stands a pyramid-like tower, 13-14 m in height, which Kiselev considered a watchtower (Kiselev, 1957, pp. 93-95). On top of the tower there could have been set up “a golden tent” that is mentioned in Tamīm’s report and in the Xin Tangshu (Minorsky, p. 283; Xin Tangshu, chap. 217-b, p. 6149).

The traces of gates are distinguishable on the eastern and the western walls. The western gate was a main one, because it was strongly guarded by inner and outer square, attached walls. Each wall had several barbicans. The wall was surrounded by a ditch, which is now filled in by soil but easily traceable by its deep green grass. Along the outside of the northern and the southern walls, there line up 6-8 small towers, which were considered as suburgan or stupas by Radlov and Klements on insufficient grounds (Radlov, p. 6; Klements, pp. 49, 52). The fragments of a large stela with Uighur-Turkic, Sogdian, and Chinese inscriptions (see below, ii) are scattered 500 m away from the southeastern corner of the fortress.

The urban area is spread out south of the fortress; Kiselev estimated its extent as 25 sq km. Radlov thought that the fortress was constructed by the Mongol ruler Möngke Khan (r. 1251-59), while the urban area was built during the Uighur Khaghanate (Radlov, pp. 4, 6; Klements, p. 51). Against this view, Kiselev judged that both belonged to the Uighur Khaghanate, because he discovered the same ceramics from the donjon, the fortress, and the urban area (Kiselev, 1957, p. 94). Furthermore, from the donjon and the fortress he found several roof end tiles (antefixa) with lotus pattern (Kiselev, 1959, p. 167), which is a typical ornamentation of Tang China. But the tiles were not brought from China. The paste analysis of them shows that the tiles were made within close range of Karabalgasun (Mitsuji and Muraoka, p. 109). These tiles might have covered the roofs of a royal residence and administrative and ceremonial buildings.

The fortress might have been built by the second Khaghan, Gele (r. 747-59), and the urban area might have been developed from the third Khaghan, Bögü (r. 759-79). In the east and the north of Karabalgasun there are several ruined fortresses that must have belonged to the Uighur period. One of them is an archeological site, Tsagaan Baishin, 10 km east of Karabalgasun; roof tiles with the same paste as Karabalgasun’s tiles were collected there (Mitsuji and Muraoka, p. 108).

According to the Shine-Usu inscription (W5), the Khaghan Gele built another city, Bay Balïq “Rich city” on the Selenge river for (or by) the Sogdians and the Chinese in 757-58 (Moriyasu, 1999, pp. 181, 185). This city is mentioned as Fugui cheng “Rich city” in the Xin Tangshu (chap. 43-b, p. 1148). The archeological site of Bay Balïq can be identified with a complex of three ruined walled fortresses, Bii Bulak, on the northern bank of the Selenge river, 11 km west of Khutag Öndör sum, 220 km north of Karabalgasun.

Kiselev excavated one dwelling site in the urban area of Karabalgasun and discovered wax, copper plate, and bronze ingot. So he considered it as the site of a metalcaster’s atelier (Kiselev, 1957, p. 94). He also found millstones in the urban area and confirmed the existence of cultivated fields around Karabalgasun (Kiselev, 1957, pp. 94-95). Tamīm stated that the capital was “a great town, rich in agriculture and surrounded by rustāqs full of cultivation and villages lying close together” (Minorsky, pp. 279, 283).

Conclusions. (1) Ordu Balïq was constructed by Chinese workers. (2) The inhabitants of this city were the upper class of the Uighurs (royal family, aristocracy, and guardsmen), the Chinese, and the Sogdians. (3) The Uighurs engaged in politics, diplomacy, and military affairs there. (4) The Chinese and the Sogdians engaged mainly in trade, commerce, handicraft industries, and agriculture. According to the Chinese historical sources, some of them also participated in politics, diplomacy, and military affairs,.



Sen-dou Chang, “Some Observations on the Morphology of Chinese Walled Cities,” Annals of the Association of American Geographers 60/1, 1970, pp. 63-91.

G. Clauson, An Etymological Dictionary of Pre-Thirteenth-Century Turkish, Oxford, 1972.

T. Hayashi and T. Moriyasu, “Palace and City of Qara-Balgasun,” in T. Moriyasu and A. Ochir, eds., Provisional Report of Researches on Historical Sites and Inscriptions in Mongolia from 1996 to 1998, Osaka, 1999, pp. 199-200; electronic edition: http://ir.library.osaka-u.ac.jp/metadb/up/LIBSIALK01/Mongol.

ʿAlāʾ-al-Din ʿAṭā-Malek Jovayni, Tāriḵ-e jahāngošāy, ed. Moḥammad Qazvini, 3 vols., E. J. W. Gibb Memorial Series XVI/1-3, London and Leiden, 1912-37; tr. John A. Boyle as The History of the World Conqueror, 2 vols., Manchester, 1958; repr. as Genghis Khan: The History of the World Conqueror, Manchester, 1997.

Jiu Tangshu (Old history of the Tang Dynasty), ed. Xu Liu et al., Beijing, 1975.

S. V. Kiselev, “Drevnie goroda Mongolii,” Sovetskaya Arkheologiya 1957/2, pp. 91-101.

Idem, “Iz istorii kitaĭskoĭ cherepitsy,” Sovetskaya Arkheologiya 1959/3, pp. 159-78.

D. A. Klements, “Arkheologicheskiĭ dnevnik poezdki v Srednyuyu Mongoliyu v 1891 g.,” in Sbornik trudov Orkhonskoĭ ekspeditsii II, Sankt-Peterburg, 1895, pp. 48-59.

Liao Shi (History of the Liao Dynasty), ed. Tuotuo et al., Beijing, 1974.

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

T. Mitsuji and H. Muraoka, “Tiles and Bricks from the Sites in Mongolia,” in Moriyasu and Ochir, pp. 106-10.

T. Moriyasu, “Site and Inscription of Šine-Usu,” in Moriyasu and Ochir, pp. 177-95.

V. V. Radlov, “Predvaritel’nyĭ otchet o rezul’tatakh ekspeditsii dlya arkheologicheskogo issledovaniya basseina r. Orkhona,” in Sbornik trudov Orkhonskoĭ ekspeditsii I, Sankt-Peterburg, 1892, pp.1-54.

Xin Tangshu (New history of the Tang Dynasty), ed. Xiu Ouyang et al., Beijing, 1975.

(Toshio Hayashi)

Originally Published: December 15, 2010

Last Updated: April 24, 2012

This article is available in print.
Vol. XV, Fasc. 5, pp. 529-530