BUKHARA i. In Pre-Islamic Times



i. In Pre-Islamic Times

The site or town of Bukhara was one of many settlements in the large oasis formed by the mouths of the Zarafshan (Zarafšān) river in ancient Sogdiana. Since there is no evidence that the river reached the Oxus in historic times, it is a reasonable assumption that in the first millennium B.C. irrigation, using the water of the river, enabled an ever-growing population to expand the arable land of the oasis. At the time of Alexander the Great no city is reported to have existed in this area, and the history of Bukhara cannot be traced before the 4th or 5th century of our era, which is the probable date of the first coins with indigenous Bu­kharan Sogdian writing on them. The alphabet used is one derived from Aramaic.

The name Bukhara may be derived either from a Sogdian word *βuxārak, whence Old Turkish Buqaraq, meaning “fortunate place” (cf. Christian So. fwxʾr) or, less likely, from a local form of vihāra, a Buddhist monastery (see buddhism ii). Naršaḵī seems to favor the former, citing an Arabic word fāḵera with the same meaning, whereas Jovaynī (I, p. 76; tr. p. 98) supports the derivation from vihāra. The name is spelled pwxʾr in a Sogdian manuscript in Sogdian script of uncertain date (Henning, 1940, pp. 8-9).

On the obverse of the coins from Bukhara appears the bust of a ruler facing right and wearing a crown copied from the crown of the Sasanian Bahrām V (r. 420-­38). This gives the earliest date for the coinage, but it is unknown how much later than the time of Bahrām that the coinage actually began (see Frye, 1949, p. 26). The earliest coins have the legend βwγʾr γwβ ʾšδʾδʾ “King Ašδāδ of Bukhara”? (Smirnova, 1970, p. 56). Later kings have a legend reading βwγʾr γwβ kʾwʾ (or kʾnʾ) “king of Bukhara, the hero” (or: “Kā¦nā¦,” a personal name). On still later coins the third word of the legend is shortened to kʾw (So. “giant”) or kʾy, which Henning (apud Frye, 1949, p. 28) suggested was a Sogdian calque on the Middle Persian Kay (written kdy), a title first found on legends of the coins of Pērōz (r. 459-84). After the Arab conquest Arabic words were added to the coins, and gradually the Bukharan legend, no longer understood, degenerated to illegibility. Finally only Arabic legends appear, which for the most part are only pious formulae. The data of the coins with Arabic legends is from early ʿAbbasid times, for standard Islamic coins with only Arabic legends ousted the Bukharan coins by the time of the Samanids, although local issues of the Bukharan coins continued for several centuries. The long series of coins, however, reveals the conservatism of the people of the Bukharan oasis, and perhaps a longer usage of a local written form of Sogdian than hitherto assumed.

Although the coins reveal the existence of a pre-­Islamic government in the oasis, undoubtedly the area was settled before the beginning of the coinage. Naršaḵī’s assertion (pp. 7-8; tr. p. 6) that the site of Bukhara had been a swamp in ancient times but that the river brought silt that filled the lowlands and enabled people to live there probably is correct. There may even have been an Oxian lake there in very early times according to Ptolemy (4.12.3).

The Tārīḵ-e Boḵārā mentions several pre-Islamic rulers, but their names are uncertain, and we know nothing about them. The first ruler of Bukhara men­tioned by Naršaḵī (p. 8; tr. p. 7) is Abrūʾī or Abarzī. He became tyrannical and was overthrown by a Turkish ruler called Qarā Jūrjīn. Unfortunately neither person can be identified from other sources. Another ruler mentioned by Naršaḵī (p. 49; tr. p. 35) is Kānā, who is credited with introducing coinage into Bukhara of the time of Abū Bakr, the first caliph. This is hardly acceptable, but whether this is a misreading of the word kʾwʾ on the coins (see above) is uncertain. Another ruler is called Māḵ (p. 29; tr. p. 19), who is said to have built the bāzār in Bukhara called after his name, and still another king of Bukhara called Dīzoʾī is mentioned on a silver vessel (see Frye, 1950, p. 110). Again nothing is known about these rulers.

It would seem that there were several local lords in the oasis of Bukhara, especially in the towns of Paykand, Vardana, and Varaḵša. Both Paykand and Varaḵša are mentioned as residences of the rulers by Naršaḵī, but it is unknown whether they were local rulers or rulers of the entire oasis. Some kind of unity in the oasis is implied by the coinage, by the extensive irrigation system, and by the long walls around the settled and cultivated areas. The wall, called kampīrak or kampīr dovāl “old lady’s wall,” probably existed in pre-Islamic times although it may not have been completed (or extended) until the early ʿAbbasid period. In spite of an apparent unity of the oasis the success of the Arab conquest suggests there was little more unity in the oasis than between oases.

With a ruler of Bukhara called Bīdūn (or Bandūn) we reach the time just before the Arab conquest, for he is mentioned by a number of Arabic sources, although with several variant readings of his name. It is uncertain whether he was killed in battle with Salm b. Zīād, the first Arab commander to cross the Oxus in 681, or whether he was already dead and his widow, called Ḵātūn in the sources, was regent for their son Ṭoḡšāda. Under Ṭoḡšāda the Arab conquest of Bukhara was accomplished. It should be noted that in the Arabic sources the rulers of Bukhara were called Boḵār-ḵodāt, where the last word is Sogdian γwtʾw, used for the nobility or aristocracy of the Sogdian oases.

The boundaries of the oasis of Bukhara on the whole have remained constant during the last millennium, but from pre-Islamic times mounds or remains of buildings are found in the desert to the west, outside the present-­day oasis, attesting a larger area of settlement in more ancient times (see Shishkin, p. 22). There were many canals in the oasis that utilized the water of the Zarafshan river, and three of the major canals men­tioned in Arabic or Persian sources can be identified today: Šāpūrkām (today Shafrikan/Šāfrekān), Ḵarḡ/qānrūd (Kalkan), and Ḵetfar or ʿĀv/Ḡāw-Ḵetfar (Babkent Darya/Bābkand Daryā), which divided into the Andāna and the Rāmīṯan-Sāmjan canals (Naršaḵī, pp. 44-45, tr. Frye, p. 32; Eṣṭaḵrī, pp. 310-11; Ebn Ḥawqal, pp. 486-87, tr. Kramers, II, pp. 466-67; Barthold, Turkestan, pp. 113-16).

The only extensive historical pre-Islamic excavations in the oasis were those of a palace complex in Varaḵša in 1938-39 and 1949-54, revealing traces of wall paintings as well as clay statuettes. In the city of Bukhara the site of the mosque of Magoki Attar was investigated by V. A. Shishkin in the 1950s, and pottery and other small objects from the earliest layer suggested a date as early as the beginning of our era. Other sites, such as that of Paykand, have only been surveyed (Shishkin, p. 16).



R. N. Frye, Notes on the Early Coinage of Transoxania, New York, 1949.

Idem, “Additional Notes on the Coinage of Transoxiana,” American Numismatic Society. Museum Notes (New York) 4, 1950, pp. 105-14.

W. B. Henning, Sogdica, James O. Forlong Fund 21, London, 1940.

Jovaynī, Tārīḵ-e jahāngošā, ed. Qazvīnī; tr. Boyle. Naršaḵī, Tārīḵ-e Boḵārā, ed. Rażawī; tr. Frye. O. I. Smirnova, Ocherki iz istorii Sogda, Moscow, 1970.

V. A. Shish­kin, Varakhsha, Moscow, 1963.

O. A. Sukhareva, K istorii gorodov Bukharskogo khanstva, Tashkent, 1958.

Search terms:

 بخارا، دوره قبل از اسلام bokhara bokhara ghabl az islam  

(Richard N. Frye)

This article is available in print.
Vol. IV, Fasc. 5, pp. 511-513