Part 8

The expedition therefore was obviously against unbelievers. That the whole of _S_ughd was lost to the Arabs is clear from the fact that Asad found it necessary to take provisions for the army with him from Bukhārā. He was not successful in recapturing the city, however, and attempted no more than the damming of the canal sluices at Waraghsar.

The fate of the garrison of Samarqand has thus been passed over in silence, unless, perhaps, it is hinted at in Asad’s reference to the capture of Muslim women. Whether Ghūrak recaptured it with his own troops or with the aid of the Türgesh, it can scarcely be doubted that he had taken advantage of the dissensions in Khurāsān to realise his ambition and at last drive the Arabs out of his capital. Of all the conquests of Qutayba beyond the Oxus, Bukhārā, Chaghāniān, and perhaps Kish alone remained to the Arabs. A confirmatory detail is the cessation of _S_ughdian embassies to China between 731 and 740: now that independence (even if under Türgesh suzerainty) had been won again, there was no need to invoke Chinese support. Negative evidence of the same kind is afforded by the absence of any Arab embassy during the same period. Had the Arabs been in possession of _S_ughd, it is practically certain that Asad, as he had done before, would have renewed relations with the Chinese court. Against this view may be set the statement in _T_ab. 1613. 5 that Khāqān was preparing an army to invest Samarqand at the time of his assassination. This report is, however, from its nature untrustworthy, and is contradicted by the presence of the king of _S_ughd with _S_ughdian troops in the Türgesh army in 119/737 as well as by Na_s_r b. Sayyār’s expedition to Samarqand two years later. _S_ughd thus enjoyed once more a brief period of independence. In 737 or 738 Ghūrak died and his kingdom was divided amongst his heirs. He was succeeded at Samarqand by his son Tu-ho (? _T_arkhūn), formerly prince of Kabudhān. Another son Me-chuʾo (? Mukhtār) was already king of Māyamurgh, while the king of Ishtīkhan in 742 was a certain Ko-lo-pu-lo who may perhaps be identified with Ghūrak’s brother Afarūn[91].

The year after the campaigns against _H_ārith, 118/736, was devoted by Asad to the re-organisation of his province, including a measure which, it seems, he had already projected in his first term of office. This was the removal of the provincial capital from Merv to Balkh[92]. Since no other governor of Khurāsān followed his example we must seek the motive for the innovation either in the contemporary situation in Khurāsān and Transoxania or in Asad’s personal views. Explanations based on the former are not hard to find. Asad, on taking office, had been faced with a serious situation both in Lower _T_ukhāristān and across the river. He had obviously to establish a strong point _d’appui_. The loyalty of the garrison at Merv was not above suspicion but the garrison at Balkh was composed of Syrian troops, who could be trusted to the uttermost[93]. Merv was also less convenient for reaching _T_ukhāristān, which was at the moment the main area of operations. More important still, perhaps, Balkh was the centre from which all disturbances spread in Eastern Khurāsān, as in the revolt of Nēzak and the recent attempt of _H_ārith. As the holding of Balkh had enabled Qutayba to forestall Nēzak, it is possible that Asad felt that in Balkh he would be in a position to check all similar movements at the beginning. Other considerations may also have disposed him to take this view. Balkh was the traditional capital and on it, as we have seen, was focussed the local sentiment of Eastern Khurāsān. Merv, on the other hand, had always been the capital of the foreigners, of the Sāsānians before the Arabs. Asad’s personal friendship with the dihqāns may have given him some insight into the moral effect which would follow from the transference of the administration to the centre of the national life. Still greater would this effect be when the rebuilding was carried out not by the Arabs themselves but by their own people under the supervision of the Barmak, the hereditary priest-ruler of the ancient shrine. Quite apart from this, however, the rebuilding of Balkh was an event of the greatest significance, and once restored it soon equalled, if it did not eclipse, its rival Merv in size and importance. While the new city was being built, the army was employed in expeditions into _T_ukhāristān, for the most part under the command of Judayʿ al-Karmānī, who achieved some successes against the followers of _H_ārith and even succeeded in capturing their fortress in Badakhshān. Other raids were undertaken by the governor himself, but without results of military importance.

Asad now planned a more ambitious expedition against Khuttal, partly in retaliation for the assistance given to _H_ārith, partly, it may be, to wipe off an old score. The chronology presents some difficulties at this point. _T_abarī relates two expeditions into Khuttal in the same year 119/737, both from the same source, but that which is undoubtedly the earlier is dated towards the close of the year (Rama_d_ān = September). Wellhausen avoids the difficulty by referring this expedition to 118, reckoning back from the appointment of Na_s_r b. Sayyār, the data for which are full and unimpeachable. This would seem the obvious solution were it not that the date given in the Chinese records for the assassination of Su-Lu, 738[94], agrees perfectly with _T_abarī’s dating of the Battle of Kharīstān in Dec. 737. The presence of Asad on the second expedition would then hang together with the “somewhat legendary” narrative of the Mihrjān feast. There seems reason, therefore, for dating this expedition in 120/738 and regarding it as having been despatched by Asad, though not actually accompanied by him. _T_abarī fortunately preserves also a short notice of the situation in Khuttal. The heir of as-Sabal, whose name is to be read as Al-Hanash, from the Chinese transcription Lo-kin-tsie[95], had fled to China, possibly on account of factional disturbances. On his deathbed as-Sabal appointed a regent, Ibn As-Sāʿijī, to govern the country until Al-Hanash could be restored. The moment was certainly opportune for making an expedition and Asad at first carried all before him. On his first appearance, however, Ibn As-Sāʿijī had appealed for aid to Su-Lu, who was at his capital Nawākath (on the Chu). The Khāqān, with a small mounted force including the _S_ughdian refugees, marched from Sūyāb (near Tokmak, on the Chu) to Khuttal in seventeen days, only to find Asad, warned of his approach by the regent, who was endeavouring to play both sides off against each other, in precipitate retreat. The baggage train had been despatched in advance under Ibrāhīm b. ʿĀ_s_im with a guard of Arabs and native troops from Chaghāniān but the main body was overtaken by the Turks as it was crossing the river and suffered severe losses. Asad, considering himself safe with the river between his army and the enemy, encamped and sent orders to Ibrāhīm to halt and entrench his position. The Turks, however, were able to effect a crossing; after an unsuccessful assault on Asad’s camp, they hastened to overtake the richer prize while the governor’s troops were too worn out to protect it. By sending a party under cover to fall on the troops of Chaghāniān from the rear while he himself attacked in front, the Khāqān forced an entrance into Ibrāhīm’s camp. Chāghān Khudāh, faithful to the last, himself fell with the greater part of his forces but the remainder of the garrison were saved by the timely arrival of Asad. According to the main account, the Arabs were allowed to withdraw to Balkh without further serious fighting. A variant account given by _T_abarī relates an unsuccessful assault by the Türgesh on Asad’s camp on the morning following the “Battle of the Baggage,” which happened to be the feast of Fi_t_r (1st October 737). On the retiral of the Arabs, the Khāqān, instead of returning to his capital with the honours of the day, remained in _T_ukhāristān.

Here he was joined by _H_ārith, who advised him to undertake a winter raid into Lower _T_ukhāristān while the Arab troops were disbanded, undoubtedly in the expectation that the local princes would again unite with him against Asad. The governor retained his army at Balkh until the winter had set in, and in the meantime the Khāqān summoned forces to join him from _S_ughd and the territories subject to _T_ukhāristān. The enumeration which _T_abarī gives of the troops accompanying the Khāqān on this expedition shows very clearly how completely Arab rule in Transoxania and the Oxus basin had been supplanted by that of the Turks. We are told that besides the Khāqān’s own Turkish troops and _H_ārith with his followers there were present the Jabghu, the king of _S_ughd, the prince of Usrūshana, and the rulers of Shāsh and Khuttal. It is fairly certain, of course, that the list is exaggerated in so far as the actual presence of the princes is concerned (it is in fact partially contradicted in other parts of the narrative), but it can scarcely be doubted that forces from some, if not all, of these principalities were engaged. On the evening of the 9th Dhuʾl-_H_ijja (7th Dec.) news reached Balkh that the Türgesh with their auxiliaries, numbering some 30,000, were at Jazza. Asad ordered signal fires to be lit and with the Syrian garrison of Balkh and what other troops he could muster from the district marched out against them. The governor of Khulm sent in a second report that the Khāqān, having been repulsed in an attack on the town, had marched on towards Pērōz Nakhshēr, in the neighbourhood of Balkh. From this point the enemy, avoiding Balkh, moved on Jūzjān and occupied the capital[96]. Instead of continuing his advance immediately, the Khāqān halted here and sent out raiding parties of cavalry in all directions, an action which put it beyond doubt that the immediate object of the expedition was not the capture of Merv but the rousing of Lower _T_ukhāristān against the Arabs. Contrary to _H_ārith’s expectations, however, the king of Jūzjān joined with the Arabs, who marched towards Shubūrqān by way of Sidra and Kharīstān. From the conflicting narratives in _T_abarī, it seems that Asad surprised the Khāqān in the neighbourhood of Kharīstān (or Sān) at a moment when his available forces amounted only to 4,000. A furious struggle ensued, which was decided in favour of the Arabs by an assault on the Khāqān from the rear, on the initiative of the king of Jūzjān. It is in connection with the battle, which he describes as if it were a set engagement in which the whole of the opposing forces were engaged, that _T_abarī gives his list of the combatants. But as only 4,000 out of the total of 30,000 troops with the Khāqān were involved, the list is obviously out of place and the whole narrative shows the marks of rehandling. The Muslims gained an overwhelming success: the Khāqān and _H_ārith, having narrowly escaped capture in the confusion, were closely followed by Asad as far as Jazza, when a storm of rain and snow prevented further pursuit. They were thus able to regain the Jabghu in _T_ukhāristān, with happier fortune than the raiding parties, whose retreat was cut off by the vigilance of Al-Karmānī, and of whom only a single band of _S_ughdians made good their escape.

On this skirmish at Kharīstān, for it was little more, hung the fate of Arab rule, not only in Transoxania, but possibly even in Khurāsān, at least for the immediate future. Though the princes of Lower _T_ukhāristān fought for Asad in the first place, there can be little doubt that a victory for Su-Lu would have swung them back to the side of _H_ārith and the Turks, who would then have been in a position to follow up their attacks with the advantage of a base at Balkh, solidly supported by the Oxus provinces. From such a danger the Arabs were saved only by Asad’s resolution and fortunate selection of Balkh as his residence. The account given of Hishām’s incredulity on hearing the report shows how very serious the outlook had been and the extent to which the name of the Khāqān had become an omen of disaster. Kharīstān was not only the turning point in the fortunes of the Arabs in Central Asia, but gave the signal for the downfall of the Türgesh power, which was bound up with the personal prestige of Su-Lu. The princes of _T_ukhāristān and Transoxania found it expedient to treat him with respect as he was returning to Nawākath, but in his own country the dissensions long fomented in secret by the Chinese broke out. Su-Lu was assassinated by the Baga Tarkhan (Kūr_s_ūl); the kingdom fell to pieces. “The Turks split up and began to raid one another,” and the _coup de grâce_ of the Khanate was delivered at Sūyāb in 739 by the faction of Kūr_s_ūl, supported by the Chinese and with the assistance of Al-Ishkand and contingents from Shāsh and Farghāna[97][98]. With the collapse of the Türgesh kingdom disappeared the last great Turkish confederation in Western Asia for more than two centuries to come. The battle of Kharīstān assured the supremacy of the Muslim civilisation in Sogdiana, but it could not have attained the richness of its full development there unless all danger from the steppes had been removed. That this security was attained was due not to the Arabs, but to the Chinese diplomacy, which, by breaking down the greatest external obstacle to the Muhammadan penetration of Central Asia, brought itself face to face with the Arabs. This could scarcely have been realised at once, however, by the Arab government, whose immediate task was to restore its lost authority in Transoxania.

NOTES

[74] As the history of this and the following period has been given in considerable detail by Wellhausen (Arab. Reich 280 ff.) from the Arab point of view, it is intended in these chapters to follow only the situation in Transoxania and the course of the Türgesh conquests, avoiding as far as possible a simple recapitulation of familiar matter. Thus little reference is made to the factional strife among the Arabs, though it naturally played a very important part in limiting their power to deal with the insurgents.

[75] See Chavannes, Documents 285, n. 3.

[76] _Cf._ _T_ab. II. 1718. 3 ff.

[77] _T_ab. 1462. 11; _cf._ 1688. 10, 1481 f.

[78] _T_ab. 1690. 16.

[79] Chav. Doc. 206 f., 293 f.

[80] Van Vloten, La Domination Arabe 28.

[81] _T_ab. 1533. 15.

[82] _T_ab. 1501. 2.

[83] Wellhausen 284 f.: van Vloten 22 f.: _T_ab. 1507 f.: Bal. 428 f.

[84] See Wellhausen 218.

[85] The variant readings in _T_ab. 1509. 11. (_cf._ Ibn al-Athīr) make it doubtful whether the taxes were reimposed on them or not.

[86] _T_ab. 1514. 11.

[87] See Yāqūt s.v.: Barthold, Turkestan 127: and _cf._ _T_ab. 1523. 3. The chief difficulty in _T_abarī’s text is the abrupt change at the last word of l. 14 on p. 1516: thumma ta_h_awwala (ashrashu) ilā marjin yuqālu lahu bawādaratun _faʿatāhum_ sabābatun ... wahum nuzūlun bikamarjata. The context shows that it was not to Ashras that Sabāba came but to the garrison of Kamarja with the news that the Khāqān was retiring past them (mārrun bikum).

[88] The chronological difficulties are explained by Wellhausen 285 ff. They are of small importance however, and it seems preferable to follow his dates for these campaigns.

[89] _Cf._ _T_ab. 1528. 9. with 1529. 5 f. 14 f.

[90] Van Vloten, _op. cit._ 29 ff.: Wellhausen 289 ff. (_cf._ 302 f.). Another account of _H_ārith is given by Gardīzī ap. Barthold Turkestan, Texts pp. 1-2.

[91] Chav. Doc. 210, 136, 140; Barthold, Arab. Quellen 21. n. 8.

[92] _T_ab. 1490, 1591. 18: Wellhausen 292 and 284 n.: Barthold in Zeitschrift für Assyriologie XXVI (1911) 261.

[93] _T_ab. 1590. 5. There does not seem to be any record of when these Syrians were settled at Balkh.

[94] Wieger 1643: Chav. Doc. 284 f.

[95] Chav. Doc. 168.

[96] As Jūzjān is distinguished from Shubūrqān in _T_ab. 1608. 17, it is probable that this was the town Kundurm or Qurzumān mentioned in Yaʿqūbī’s Geog. 287.

[97] _T_ab. 1613: Chav. Doc. 83 f., 122 n. As regards the adjective Kharlukhī applied to the Jabghu in 1612. 16, the most satisfactory explanation is that given by Marquart, Hist. Glossen 183 f.

[98] The frequent references in the Chinese annals to the association of Se-kin-tʾi, king of Kish, with the Türgesh raise an interesting problem. There can be no doubt that he is the same prince as Al-Ishkand, ruler of Nasaf, in the Arabic records. The name is Iranian and personal, not dynastic. (See Justi’s Iranisches Namenbuch.) Al-Ishkand is first mentioned in the account of the Battle of the Pass, (_T_ab. 1542. 8) where he appears in command of a cavalry force on the side of the Khāqān, though Kish and Nasaf were both in the hands of the Arabs (1545. 1). The forces which he commanded were therefore not the ordinary local troops. During _H_ārith’s siege of Tirmidh he received reinforcements from Al-Ishkand, but no statement is made on the composition of his forces. He is mentioned again as accompanying the Khāqān and the _S_ughdians in the attack on Asad before the “Battle of the Baggage” (1597. 17-18,) where the reading ‘I_s_pahbadh of Nasā’ is probably an error in the tradition. Again there can be no question here of local troops from Nasaf or Kish. In the Chinese records Se-kin-tʾi appears as the commander of an independent force, not merely a detachment of Turks or levies from Shāsh or Farghāna. The most reasonable conclusion is that Al-Ishkand was the commander of the corps of _S_ughdian refugees. This would explain the title “King of the Warriors” by which he is sometimes mentioned in the Chinese records (Chav. Doc. 147 n. 1 and 313). The actual term (Chākar) from which the title was derived does not appear in the Arabic histories in this connection, but it is perhaps possible that a variant of the name (derived from _razm_) is to be read in _T_ab. 1614. 2 for the meaningless “razābin al-Kissī.” In 1609. 15 a force of “Bābīya” is mentioned along with the _S_ughdians, and the name, though unrecognisable, probably refers to some forces connected with _S_ughd. Wellhausen’s conclusion that the _S_ughdians and “Bābīya” formed part of the personal following of _H_ārith b. Surayj seems to force the connection in the text too far (_h_amala ʾl-_h_ārithu waman maʿahu min ahliʾs-sughdi wal-bābīyati). On the other hand, since al-Ishkand appears as the ally of _H_ārith, we may conclude that some understanding existed between the latter and the _S_ughdians (and therefore the Turks) at the time of his revolt. It is probable that the _S_ughdian corps assisted in the recovery of Samarqand from the Arabs.

V. THE RECONQUEST OF TRANSOXANIA.

The reaction produced by the downfall of the Türgesh power was manifested in Transoxania in the first place by an increased regard for China. The princes had found the Türgesh yoke no less galling in the end than that of the Arabs[99]; the country was as wasted and impoverished by their continual raids as it had been under the latter. The profitable native and transit trade, the source of the entire wealth of the cities, must have shrunk to negligible proportions if it had not wholly ceased. All classes of the people therefore were weary of war and sought only a peace consonant with their self-respect. For the attainment of these aims it was vain to look to China; the granting of bombastic titles to a few princes brought neither comfort nor aid. A final opportunity was thus offered to wise statesmanship to swing the whole country round to the Arabs almost without a blow. For two years, however, the situation seemed to remain much as it was, except for an expedition into Khuttal, probably on the pretext of assisting the ruling house against a usurper from Bamiyān. Nevertheless some progress had been made by the administration in regaining the prestige it had lost. This was due not merely to the effect of the victories over _H_ārith and the Türgesh, but even more to Asad’s personal relations with the dihqāns. He had, as we have seen, gratified the national pride of the people of _T_ukhāristān by transferring the seat of power from Merv, the capital of the foreigners, to Balkh, the centre of their national life. As had been the case even in his first term of office, he was able to attract to his side many of the more influential elements in Lower _T_ukhāristān and the Ephthalite lands—to this, in fact, was largely due his success in the struggle with the Turks. More striking evidence still is afforded by the conversion of the dihqāns at this period, amongst them the minor chief Sāmān-Khudāh and probably also the Barmak. By this means Asad laid the foundations for a true reconciliation and Narshakhī’s work amply attests the honour which later generations attached to his name. His work was of course incomplete in that it was practically confined to the ruling classes and naturally did not extend to the now independent dihqāns of _S_ughd.

Early in 120/738 Asad died, and after a lapse of some months the governorship was conferred by Hishām on Na_s_r b. Sayyār. For the subject peoples no choice could have been more opportunely made. Na_s_r was one of the few men who had come with honour and reputation through the external and internal conflicts of the last thirty years. Belonging to the small and almost neutral tribe of Kināna, his position bore a strong similarity to that of Qutayba in that both were more dependent on the support of a powerful patron than on their tribal connexions, and therefore, though favouring Qays, less frantically partisan. In contrast to Qutayba, however, Na_s_r, after thirty years of active leadership, knew the situation in Khurāsān, Transoxania, and Central Asia as no Arab governor had ever done. He had seen the futility of trying to hold the country by mere brute force, and the equal futility of trying to dispense with force. While he held the support of Hishām, therefore, he set himself to restore Arab authority in Transoxania. The appointment of Qa_t_an b. Qutayba, who had inherited much of his father’s ability, to command the forces beyond the river gave earnest of an aggressive policy. The appointment was not to Samarqand, as Wellhausen says, but “over _S_ughd,” _i.e._, the garrisons in Bukhārā and probably Kish, who were responsible in the first place for keeping the surrounding districts in subjection. The governor himself then carried out a brief expedition, intended apparently to punish some rebels in the neighbourhood of the Iron Gate, possibly in Shūmān. Having thus vindicated the authority of the administration, Na_s_r returned to Merv and delivered the famous Khu_t_ba in which the system of taxation and conditions of amnesty were at last laid down in a form satisfactory to the mawālī and the subject peoples[100]. The results were as he had foreseen. The princes and people of Transoxania submitted, as far as we can judge, without opposition when Na_s_r with his army marched through _S_ughd to re-establish the Arab garrison and administration in Samarqand.