BUKHARA ii. From the Arab Invasions to the Mongols



ii. From the Arab Invasions to the Mongols

At the time of the Arab conquests in Transoxania, Bukhara was ruled by the indigenous Sogdian dynasty of the Boḵār-ḵodāts (see i, above). The first appearance of Arab armies there is traditionally placed in Moʿāwīa’s caliphate when, according to Naršaḵī (pp. 52-53, tr. Frye, pp. 37ff.), ʿObayd-Allāh b. Zīād b. Abīhe (q.v.) crossed the Oxus and appeared at Bukhara (end of 53/673-beginning of 54/673-74). Bukhara was at this time being ruled by a woman, Ḵātūn, as regent for her infant son the Boḵār-ḵodāt Ṭoḡšāda, but, despite an appeal to the nearby Turks (the Western Tiu-kiu or Türgeš), she had to submit and pay the Arabs a tribute of a million dirhams and 4,000 slaves. This was renewed two years later to the new governor Saʿīd b. ʿOṯmān, but it is quite clear that no permanent Arab control was established in the city (H. A. R. Gibb, The Arab Conquests in Central Asia, London, 1923, pp. 17-20).

The achievement of this was to be the task of Qotayba b. Moslem Bāhelī (q.v.), governor of Khorasan from 85/704 or 86/705 to 96/715. During arduous campaigns in Sogdia (87-90/706-09), Qotayba in the end overcame the resistance of the Bukharans and their Turkish allies, imposed a tribute of 200,000 dirhams to the caliph and 10,000 to the governor of Khorasan, and placed an Arab garrison in the city, forcing every home owner to share his residence with Arabs. In 94/712-13 he erected within the citadel the first mosque in Bukhara, on the site of a former (Buddhist, Zoroastrian?) temple (bot-­ḵāna), but had at the outset actually to pay the local people two dirhams a time to attend the Muslim worship there on Fridays (Naršaḵī, pp. 67-68, tr. p. 49; Gibb, op. cit., pp. 33-41). The Boḵār-ḵodāt Ṭoḡšāda remained as co-ruler with the Arabs until his murder in 121/73, and was latterly on terms of friendship with the Omayyad governor Naṣr b. Sayyār, but the city nevertheless continued to be a focus for local Sogdian and Turkish discontent against the Arabs (Naršaḵī, pp. 83-85, tr. pp. 60-62); thus in 110/728-29, the city had been lost to the Arabs after a local revolt for a whole year (Ṭabarī, II, pp. 1514, 1529; Gibb, pp. 69-70). During these years, both the Qaghan of the Eastern Tiu­-kiu, Mo-č’o, and the Qaghan of the Türgeš, Su-lu (called by the Arabs Abū Mozāḥem “the one who charges [like a bull or elephant],” d. 738; cf. Gibb, p. 85), remained ever-prepared to support Sogdian discontent against the Arabs; as late as 166/782, the governor of Khorasan Fażl b. Solaymān Ṭūsī built walls to protect Bukhara and Sogdia from Turkish incursions (Gardīzī, ed. Ḥabībī, p. 127; Naršaḵī, pp. 47-­48, tr. pp. 33-34).

Although the ʿAbbasid daʿwa had begun in Khora­san, the Arab colony in Bukhara speedily showed its disillusionment with the new regime. In 133/750-51 the revolt of Šarīk or Šorayk b. Shaikh Mahrī had to be suppressed by the governor of Bukhara and Samarqand Zīād b. ṣāleḥ, but two years later Zīād himself adopted a rebellious attitude—perhaps with the connivance of the caliph al-Saffāḥ—towards Abū Moslem, probably with the backing of the same discontented elements who had supported Šarīk, and had to be killed by Abū Moslem (Barthold, Turkestan2, pp. 194-96; E. L. Daniel, The Political and Social History of Khurasan under Abbasid Rule 747-820, Minneapolis and Chicago, 1979, pp. 87-88, 111-12). Together with the nearby town of Naršaḵ and the rest of Sogdia, the villages of the Bukhara oasis came out strongly for Hāšem b. Ḥakīm known as al-Moqannaʿ and his “wearers of white” (mobayyeża, sapīd-jāmagān). Prolonged fight­ing took place in the region around Bukhara from 159/776 onwards until Moqannaʿ was killed in 163/779-­80 by the Arab generals Moʿāḏ b. Moslem and Saʿīd Ḥarašī (Barthold, pp. 198-200; Daniel, pp. 139-43). During these same years, Bukhara had also participated in the revolt in 160/776-77 of the local Iranian mawlā of T¨aqīf, Yūsof Barm (Barthold, p. 198; Daniel, p. 166). In Hārūn al-Rašīd’s caliphate, a reform of the local Bukharan coinage was made by the governor of Khora­san Ḡeṭrīf b. ʿAṭāʾ Jorašī (governor 175-77/791-93), with alloy coins, the so-called “black” or “Ḡeṭrīfī” dirhams, replacing the old, largely disappeared silver coinage of the Boḵār-ḵodāts (Barthold, pp. 204-06; C. E. Bosworth, “Ghiṭrīf b. ʿAṭāʾ,” in EI2, Suppl.). Bukhara now rose again against the caliphal governor ʿAlī b. ʿĪsā b. Māhān, whose tyrannies and exactions had made him universally execrated in Khorasan, as part of the prolonged rebellion led by Rāfeʿ b. Layṯ b. Naṣr b. Sayyār; Hārūn’s general Harṯama b. Aʿyan had to march against Bukhara in 193/808-09 and reoccupy it, at the same time capturing Rāfeʿ’s brother Bašīr, whom Hārūn had killed (Bart­hold, pp. 200-01; Daniel, pp. 174-75).

Meanwhile, the line of Boḵār-ḵodāts continued in the city (by the middle of the 2nd/8th century, at least nominally Muslim, though prone to apostatizing), at times giving help to the Arabs (as did Qotayba b. Ṭoḡšāda initially against Šarīk), but in general adopting an ambiguous attitude towards the Arabs, so that Qotayba b. Ṭoḡšāda and also his brothers S.kān and Bonyāt were all killed by the Arab commanders, the last-named for complicity in Moqannaʿ’s revolt (Naršaḵī, pp. 11, 13-16, 24, tr. pp. 8-11, 17, 110 n. 36). Although the political importance of the Boḵār-ḵodāts was reduced by the Arabs, their social influence must have been still significant, for they continued to hold their estates up to the time of the Samanid Amir Esmāʿīl b. Aḥmad (279-95/892-907), when these lands were confiscated to the state but the equivalent income paid out in cash to Bonyāt’s descendant Abū Esḥāq Ebrāhīm (d. 301/913-14). Naršaḵī adds, moreover, that the latter’s family was still to be found in the villages of Safna and Sīavonj in the Bukhara oasis (pp. 15-16, tr. pp. 11-12).

In the 3rd/9th century, Bukhara had closer connec­tions with the capital of the governors of Khorasan, Marv, and then with the capital of the Taherid gover­nors, Nīšāpūr, than with the more easterly parts of Sogdia and what was initially the capital of the Samanids at Samarqand. After the fall of Nīšāpūr to Yaʿqūb b. Layṯ in 259/873, the authority of the Saffarids was briefly recognized in the ḵoṭba of Bukhara, but a certain Ḥosayn b. Ṭāher Ṭāʾī (it is uncertain whether he was in fact a scion of the dispossessed Taherids, cf. R. Vasmer, in Numismatische Zeitschrift 23, 1930, p. 148) tried to seize control of the city but had to leave Bukhara in defeat; in the midst of the chaos and strife following the fall of the Taherids in Khorasan, the notables and ʿolamāʾ sent to the Samanid ruler in Samarqand and Farḡāna, Naṣr b. Aḥmad, who in Ramażān, 260/June-July, 874 sent his younger brother Esmāʿīl, the future amir, as governor of Bukhara, and order was eventually restored in the city (Naršaḵī, pp. 106-09, tr. pp. 77-82). Bukhara now began a period of nearly a century-and-a-half’s prosperity, and func­tioned as a cultural center both for Arabic learning and for the efflorescence of New Persian literature under the patronage of the Samanid amirs, who soon made the city their administrative capital. A famous passage by the scholar of Nīšāpūr T¨aʿālebī, prefixed to the section of his literary anthology the Yatīmat al-dahr on the literary luminaries of Bukhara, praises the city in the era of the Samanids as “the focus of splendour, the Kaʿba of empire, the meeting-place of the unique figures of the age, the rising-place of the stars of the literary men of the world, and the forum for the outstanding per­sonages of the time” (ed. Cairo, 1375-77/1956-58, IV, p. 101, tr. in Browne, Lit. Hist. Persia I, pp. 365-66).

It is from the Samanid period also that we possess detailed information on the topography and urban structure of the city in early medieval times, and one notes from this, when comparing it with the infor­mation of later times, that, unlike other eastern Iranian cities where displacements of urban sites have occurred, Bukhara has remained essentially on the same spot since the Arab conquest. The 4th/10th-century geographers mention the usual tripartite division of a citadel (ko­handez); town proper (šahrestān); and a suburb (rabaż) between the original town and the wall of early Islamic times, the latter having seven gates of iron. The citadel contained the Boḵār-ḵodāts’ palace and the original mosque of Qotayba b. Moslem. To its east, and dividing it from the completely separate šahrestān, was an open, sandy space called the Rīgestān, where in Samanid times the Amir Naṣr b. Aḥmad (301-33/914-43) built a palace and where the dīvāns of the administration was situated (detailed in Naršaḵī, pp. 36-38, tr. pp. 25-26). In this same century, too, a further, outer wall was built, having eleven gates; the city had clearly expanded, though geographers and travelers still condemned it as an unsanitary and crowded place (the “anus of the world,” in the words of one poet of the period; see T¨aʿālebī, Laṭāʾef al-maʿāref, ed. E. Abyārī and Ḥ. K. ṣayrafī, Cairo, 1960, p. 216, tr. Bosworth, The Book of Curious and Entertaining Information, Edinburgh, 1968, pp. 139-40). See for descriptions of the city, with references to the geographers, Barthold, pp. 100-17; Le Strange, Lands, pp. 460-62; and also Naršaḵī, pp. 16-­42, tr. pp. 12-29; Ḥodūd al-ʿālam, tr. Minorsky, pp. 112-­13, comm. p. 352.

With the fall of the Samanids at the end of the 4th/10th century, the Qarakhanids took over Trans­oxania; Bukhara was temporarily occupied by Hārūn or Ḥasan Boḡrā Khan in 382/992, and then in 389/999 it was definitively occupied by the Īlak (Ilig) Naṣr b. ʿAlī of Uzgand. For the next century-and-a-half, it was part of the western Qarakhanid khanate, ruled by descendants of the Īlak Naṣr, but under the loose, decentralized rule of the Turkish tribesmen from the steppes, Bukhara inevitably lost much of its political importance. The city was affected by an extremist Shiʿite, pro-Fatimid propaganda movement in 436/1044-45, until Boḡrā Khan Ebrāhīm b. Naṣr (the later Ṭamḡāč Khan) ordered a general massacre of Ismaʿili sympathizers there (Ebn al-Aṯīr, IX, p. 524; cf. Barthold, pp. 304-05, who here wrongly identifies this Boḡrā Khan). In the second half of the century, Ebrāhīm’s son Šams-al-Molk Naṣr built a new Friday mosque in the city, and laid out adjacent to it a new palace and an enclosure of gardens and hunting grounds (ḡūroq) called Šamsābād ([continuator of] Naršaḵī, pp. 40-41, 73, tr. pp. 29, 53; Barthold, pp. 109-­10, 315-16, 318). We also know of disturbances in Bukhara arising from hostility between the khans and the ambitious local ʿolamāʾ, beginning with Šams-al-­Molk’s execution of the Imam Abū Ebrāhīm Esmāʿīl ṣaffār in 461/1069 (Samʿānī, Ansāb, ed. Hyderabad, VIII, p. 318). However, the long reign of Arslān Khan Moḥammad b. Solaymān (495-524/1102-30) brought peace to the city and enabled the khan to embark on an extensive building program there; he rebuilt the citadel and city walls, and erected a new Friday mosque and two new palaces.

These years also saw the rise to prominence in Bukhara of the influential family of Hanafite scholars of the house of Borhān, holding the spiritual title of ṣadr (see aúl-e borhaún), and in the century or so from the decline of Qarakhanid power in Bukhara till after the first appearance of the Mongols, the Āl-e Borhān were frequently able to exercise political power in Bukhara, either on their own behalf or else as representatives in the city of outside, conquering powers. Thus, when the pagan Qara Khitay defeated the Saljuq sultan Sanjar at the battle of the Qaṭvān steppe in 536/1141, the victors used them as their governors in Bukhara for some seventy years, and the Āl-e Borhān collected tribute for them from the city. An expedition by the Ḵᵛārazmšāh Tekeš b. Īl Arslān against Bukhara took place in 578/1182, and a further one in 594/1198 is mentioned by Ebn al-Aṯīr (XII, pp. 137-38) alone, however, of the sources (hence the historicity of this last attack was doubted by Barthold, pp. 344-46, cf. ibid., p. 342, and Bosworth, in Camb. Hist. of Iran V, pp. 191-92). But in 604/1207, ʿAlāʾ-al-Dīn Moḥammad b. Tekeš Ḵᵛārazmšāh certain­ly came to Bukhara after negotiations with the last Qarakhanid ruler of Samarqand ʿOṯmān Khan and other Muslim elements discontented with the rule of the Qara Khitay; he rebuilt the citadel there before return­ing to Ḵᵛārazm, ([continuator of] Naršaḵī pp. 34-35, tr. p. 25; Barthold, pp. 359-60). Khwarazmian author­ity continued in some form in Bukhara, for in 614/1217-18 ʿAlāʾ-al-Dīn Moḥammad commanded that the ʿAb­basid caliph Nāṣer’s name be omitted from the ḵoṭba there, and he deposed the ṣadr Borhān-al-Dīn Mo­ḥammad b. Aḥmad (Barthold, pp. 375, 379); but then in D¨uʾl-ḥejja 616/February 1220, the city fell to the army of the Mongols of Jengiz Khan,



The local history of Naršaḵī is a prime source for this period, but the early 9th/15th century history of Moʿīn-al-Foqarāʾ Aḥmad b. Mo­ḥammad, the Ketāb-e Mollā-zāda, also contains much information on the pre-Mongol Islamic history of Bukhara; see on this, Barthold, p. 58, and R. N. Frye, “City Chronicles of Central Asia and Khurasan: the Kitāb-e Mullāzāde,” in Avicenna Commemoration Volume, Calcutta, 1956, pp. 89-92; also in his Islamic Iran and Central Asia (7th-12th centuries), London, 1979, no. XXXIII.

See also R. N. Frye, Bukhara, the Medieval Achievement, Norman, Oklahoma, 1965. Idem, “Bukhara,” in EI2.

(C. E. Bosworth)

Originally Published: January 1, 2000

Last Updated: January 1, 2000

This article is available in print.
Vol. IV, Fasc. 5, pp. 513-515