Printer Friendly


Many scholars of Inner Asian history have long believed that after the destruction of the Uighur empire in A.D. 840, the victorious Kirghiz followed Inner Asian political tradition and established an empire that included the Mongolian plateau, particularly the Orkhon River valley--the traditional "heartland" of many earlier pastoral empires, including that of the defeated Uighurs. These scholars further held that the Kirghiz were expelled from Mongolia some eighty years later when Khitan forces marched into the region in 924. This line of reasoning, based largely on faulty assumptions and little evidence, is untenable. An examination of Western scholarship on the subject shows how this misconception came into being; further, a close look at the available evidence--documentary, archaeological, and geographical--reveals that after their defeat of the Uighurs the Kirghiz enjoyed at best brief and inconsequential control of the Orkhon valley. For a number of reasons, the Kirghiz remained in their homeland in the region of the upper Yenisei River of south Siberia. When the Khitans extended their power into the Orkhon region in the tenth century, they did not encounter the Kirghiz there.

RELYING AS HEAVILY AS IT DOES on records kept by outsiders, the history of Inner Asia often suffers from a lack of continuity, particularly in early periods. One of the many troubling gaps in our knowledge of the history of eastern Inner Asia, i.e., the geographical and cultural region north of modem China that centers on the Mongolian plateau, spans a relatively brief period from the middle of the ninth century to the early tenth century A.D.--The period immediately following the collapse of the Uighur (Ch. Hui-ho [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Hui-hu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], etc.) empire in A.D. 840 and preceding the expansion of Khitan (Ch'itan [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) power into the Mongolian steppe ca. 924. This gap has proved particularly distressing to historians because it is rather a late period to be so exceedingly arid, and because it follows a lengthy period about which far more is known--a period of a reasonably (for In ner Asia, at least) well-documented series of nomadic empires that followed one after the other, beginning in the late fourth century and ending in the mid-ninth: Jou-jan [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] [1] Turk (T'u-chueh [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]), and Uighur. These nomadic empires were essentially tribal confederations of pastoral nomadic peoples; each was multi-ethnic in nature but took its name from the dominant tribe. [2]

Several of the major confederated polities of eastern Inner Asia were centered around the same general area: the valley of the Orkhon River in what is now the north-central region of modem Mongolia. The Orkhon flows northward and joins the Selenga; the river then continues northward to flow into Lake Baikal in Siberia. The Turks in the period from the mid-sixth to the mid-eighth century, the Uighurs in the period from the mid-eighth to mid-ninth century, and the Mongols in the thirteenth century are all known to have made the Orkhon-Selenga region the political focus of their states. Ruins of cities such as the Uighur capital of Qarabalghasun and the Mongol capital of Qaraqorum, as well as important inscribed commemorative stelae from the Turk and Uighur periods, have been found there. It is also very likely that some earlier groups, such as the peoples known in Chinese sources as Hsiung-nu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (fl. ca. 200 B.C. to ca. A.D. 155) and Jou-jan (fl. ca. A.D. 380 to A.D. 555), viewed the general region of the Orkhon valley as their political nucleus as well.

The Turks, and possibly other peoples, also regarded this region as a national center in a spiritual or religious sense, as can be seen from Old Turkic references to the Otuken yis. The meaning of Otuken is uncertain, but yis means a mountain forest--"the upper parts of a mountain covered with forest, but also containing treeless grassy valleys. [3] It seems that the Otuken Mountains correspond to some part of the modern Khangay Mountains, from which the Orkhon and Selenga flow; some sources indicate that the term referred to the eastern foothills of the Khangay, near the Orkhon, while others suggest that it could mean the entire Khangay range. [4] It is clear from the eighth-century Old Turkic inscriptions that the Otuken Mountains--sometimes referred to as Otuken yer/yir, "the land of Otuken"--had a particular, spiritual significance for the Turks. Indeed, the region is referred to in the inscriptions as "sacred" (iduq). [5] Note the following statements from those inscriptions:

If the Turk qaghan rules from the Otuken Mountains, there will be no trouble in the realm.... A land better than the Otuken Mountains does not exist at all! The place from which the tribes can be [best] controlled is the Otuken Mountains. [6]

If you stay in the land of Otuken, and send caravans from there, you Will have no trouble. If you stay at the Otuken Mountains, you will live forever dominating the tribes! [7]

It was I myself, Bilge Tonuquq, who [had led] the Turk qaghan and the Turk people to the Otuken land. Having heard the news that [the Turks] settled themselves in the Otuken land, there came all the peoples who were living in the south, in the West, in the north and in the east [and submitted to us]. [8] Clearly, the Otuken Mountains had symbolic and, one supposes, strategic importance as the national "refugium" of the Turks. [9] The significance of the Otuken Mountains is reinforced by Chinese sources, which name them as the residence of various Turk qaghans. [10] Furthermore, the Otuken region retained its significance after the collapse of the Turk empire in 744, as can be seen from the stone inscriptions created by their successors, the Uighurs. [11]

In 840, the continuity of the Orkhon tradition, which had been established at least by the middle of the sixth century with the founding of the Turk empire in A.D. 555 and which continued through the history of the succeeding Uighur steppe empire, was threatened. In that year the Turkic Kirghiz (Hsia-chia-ssu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], etc.), a people then living in the region of the upper (i.e., southern) Yenisei River, in the general area of present-day Minusinsk and Abakan in south Siberia, northwest of the Mongolian plateau, successfully attacked the capital of the Uighur steppe empire that had held sway over much of eastern Inner Asia for nearly a century. The Uighurs, already weakened by internal strife and court intrigue, as well as by epizootics and famine, could not withstand the Kirghiz onslaught, and were quickly routed. Groups of Uighurs fled in virtually every direction. [12]

What happened in the Orkhon valley after the Uighur collapse? Many modern scholars, particularly in the West, have assumed that the Kirghiz followed steppe tradition and established themselves for some eighty years as masters of the Mongolian plateau and inheritors of the Orkhon political tradition. This interpretation was based largely on an argument of silence. There are no native Inner Asian sources that provide any information on this matter. [13] Chinese sources, particularly the standard historical works, in fact reveal very little about the Orkhon region in the period immediately following the Uighur collapse. It was not long after this that China's T'ang dynasty (618-907) began its own final downward spiral; it is not surprising, therefore, that Chinese records were more concerned with problems within China itself than with what was happening in Inner Asia and elsewhere.

The same scholars who presumed Kirghiz domination of the Mongolian plateau after 840 also supposed that the Kirghiz were pushed out of the region by the expanding power of the Khitans when their dynamic ruler Yeh-lu A-pao-chi entered the Orkhon region with his troops in 924. This misconception seems to have originated, in part at least, in the works of the French scholar Edouard Chavannes and the Russian scholar V. V. Bartol'd (W. W. Barthold), both of whom were generally cautious in their analyses. Chavannes wrote in 1897:

Vers le milieu du [IX.sup.e] siecle cependant, les Ouigours furent rompus et disperses; les Kirghiz, qui les avaient defaits, furent incapables d'etablir un gouvernement stable. Les Khitan en profiterent; n'etant plus contenus par leurs redoutables antagonistes, ils ne tarderent pas Ii devenir conquerants a leur tour. Dans les premieres annees du [X.sup.e] siecle, leur chef Apaoki [i.e., A-pao-chi] triompha des Hi et se les assimila si bien, qu'a partir de cette epoque, les territoires des Hi et des Khitan ne sont plus distingues par les historiens chinois; poussant plus au Sud encore, il depassa la Grande-Muraille et ravagea le Tche-Ii. Au Nord-Quest, il alla jusqu'a l'ancicnne residence des Ouigours sur les bords de l'Orkhon et, en l'an 924, ii y erigea une stale pour commemorer ses exploits. [14]

Chavannes did not refer to a Kirghiz empire that was threatened or diminished by the Khitans; indeed, he indicated that the Kirghiz had been unable to establish a stable government in Mongolia. Writing of the Khitans as conquerors, he nevertheless did not state that they had conquered the Kirghiz; indeed, he simply skirted the issue of whom, if anyone, the Khitans encountered in the Orkhon valley in 924. Some years later, Barthold displayed a bit less caution when he wrote on the subject:

Das letzte turkische Volk, das in der Mongolei herrschte, waren, soweit man nach den chinesischen Quellen urteilen kann, die Kirgizen, die im Jahre 840 die Uighuren besiegt hatten. Ihre Verdriingung aus der Mongolei war anseheinend verbunden mit dem Erstarken des mongolischen Volkes der Qytal zu Beginn des 10 Jahrhunderts, die im ndrdlichen China ein maichtiges Reich grundeten und diesem Reiche ihren Namen gaben. [15]

Barthold thus suggested that the departure of the Kirghiz from Mongolia was "apparently connected with" the rise of Khitan power in Manchuria and its subsequent expansion, although he gave no source for this information. In an earlier article in Russian, Barthold had also suggested that Khitan expansion into the Orkhon region must have been preceded by a victory over the Kirghiz, but admitted that there was no real evidence for such a victory. [16] This uncertainty was reflected in a statement by Karl Wittfogel and Feng Chia-sheng, who cited Barthold as their source.

Having overthrown the Uighurs in 840, they [i.e., the Kirghiz] probably held the Orkhon region for a while but were driven back again at the beginning of the tenth century when the Ch'i-tan became the masters of Central Asia. [17]

The work of Chavannes apparently was the source for stronger statements by Rene Grousset in the latter's L'Empire des steppes:

Les Kirghiz s'insstallerent a Ia place des Quigour dans la "Mongolie imperiale sur le haut Orkhon, autour de l'actuelle Qara-balgassoun et de l'actuelle Qaraqoroum. Mais ces tribus siberienees firent regresser Ia Mongolie vers Ia barbarie. Les Kirghiz resterent maitres du pays jusque vers 920, epoque ou ils devaient etre vaincus par le peuple mongol des K'i-tan et rejetes vers les steppes de I'Ienissei.... En 924, il EA-pao-ki, i.e., A-pao-chi] penetra en Mongolie, poussa jusqu'au haut Qrkhon, entra a Qara-balgassoun, en chassa les Turcs Kirghiz qui occupaient cette region depuis 840, et les refoula vers le haut Ienissei et les steppes de I'Ouest., [18]

This account by Grousset, in the only truly successful--and hence still much consulted--synthetic history of Inner Asia in a Western language, apparently led to the general acceptance of the notion that the Kirghiz remained in Mongolia, masters of the Orkhon region, until 924, when they were driven out by the invading Khitans. Later scholars, lulled into a comfortable and unquestioning conviction by Grousset's unequivocal statements and his references to such learned scholars as Chavannes and E. Bretschneider (as well as Barthold's writings), wrote with even less caution, confidently asserting the historicity of the Khitan expulsion of Kirghiz power from central Mongolia in 924. [19]

Chinese sources, however, can help us put to rest the notion of a Kirghiz empire in Mongolia, and of a tenth-century Khitan-Kirghiz conflict there. During the reign of the T'ang emperor Wu-tsung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (r. 840-46), at which time the Uighur steppe empire was destroyed, at least four official Kirghiz envoys were sent to China in the years following the Uighur collapse. More than once these envoys expressed the Kirghiz intention to move into the Uighurs' former territory. [20] They also set forth a plan to seize the cities of An-hsi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] and Pei-t'ing [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Besbaliq) in the northern Tarim Basin, and revealed fears concerning a possible Uighur alliance with Tibet. [21] Although Wu-tsung was intrigued by the prospect of allying with the Kirghiz and possibly recovering parts of the western regions for the T'ang empire, he was quickly talked out of the idea by Chief Minister Li Teyu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] who forcefully explained the impracticality of such efforts. [22] Although some scholars have interpreted the Chinese sources as asserting that the Kirghiz attacked and subjugated An-hsi and Pei-t'ing without Chinese assistance, there is no clear evidence that such was the case. [23]

Furthermore, the Chinese seem to have expected that the Kirghiz would establish themselves in the Orkhon region; a letter from Wu-tsung's court at Ch'ang-an [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] to the Kirghiz ruler stated: "Now that the Uighur camp has been levelled, the mountains and rivers of our [two] nations are no longer separated. Now that you have become a neighbor on our border, [We are again able] to examine your tribute and petitions." [24] The Chinese were in fact interested in establishing good relations with the Kirghiz--no doubt in the expectation that they would be the new power in Mongolia--even going so far as to accept the fictive kinship ties between the two ruling houses that the Kirghiz themselves seem to have promoted. The T'ang imperial house of Li claimed as one of its eminent ancestors the Han general Li Kuang [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (d. 119 B.C.), while the Kirghiz believed that at least some of their people (including, apparently, their ru ler) were the descendants of Li Kuang's grandson, Li Ling [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (d. 74 B.C.), a famous Han general in his own right who was captured by the Hsiung-nu in 99 B.C. and lived among them for the remainder of his life. [25] Li Ling enjoyed a position of prominence at their court, and may have served as an official with some responsibility in regard to the people then known as Chien-k'un [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]. who are commonly identified with the Kirghiz. [26] On the basis of this Kirghiz claim, and his desire to maintain cordial relations with the Kirghiz, T'ang Wu-tsung went so far as to have the Kirghiz ruler's name entered into the register of the imperial family. [27]

Despite the expressed intention of the Kirghiz to occupy the Orkhon region, and the expectation of the Chinese that they would do so, letters carried back to the Kirghiz ruler by subsequent envoys show that this did not occur. These letters, written by Li Te-yu in the emperor's name, mention the great distance between the T'ang realm and that of the Kirghiz, and the difficulty of communication between them. A letter of 843 states that the Kirghiz were separated from China by "the various foreign peoples. [28] This is not to be taken literally, but does imply a distance greater than that between Chinese territory and the Orkhon valley.

Another Kirghiz envoy, who arrived in the spring of 844, complained to the Chinese of communication problems between the two nations, mentioning blocked roads and the failure of Chinese envoys to come to the Kirghiz court. He also reiterated the Kirghiz wish to move into Uighur territory, clearly indicating that they had not yet done so. Internal evidence from the letter written in reply to this envoy's message shows that there was indeed a general failure to communicate effectively; the envoy's letter from the Kirghiz ruler asked about the fate of the Chinese T'ai-ho [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Princess, daughter of the T'ang emperor Hsien-tsung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (r. 805-20) and therefore Wutsung's paternal aunt, who had been in Uighur hands since her marriage to an Uighur qaghan in 821, even though a previous Chinese letter to the Kirghiz qaghan, sent a year earlier in April or May 843, had already informed him of her return to the T'ang court. The Chinese letter once again speaks of the great distances which hindered communication between the two nations. Of even greater interest is its statement: "We have also heard that this autumn you wish to move to the Uighurs' [former] camp (ya-chang [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]). By destroying their great nation and then occupying their former place, you will thereby cause all the foreign peoples to be in awe of your majesty. The Uighurs would then lose hope and draw near our border. This can be called quite a good plan." [29] It is evident from this letter that by the time of the envoy's arrival in the spring of 844, and quite possibly by the time of the letter's composition in the spring of 845, [30] the Kirghiz still had not occupied the old Uighur heartland--the Orkhon valley--or territory near it.

Subsequent T'ang-Kirghiz relations were even more muddled. In May 845, Wu-tsung appointed an official to grant the Kirghiz qaghan a Chinese title, but this official still had not been sent in April 846 when the emperor died--a further indication of the distance and relative unimportance to China of the Kirghiz. [31] Wu-tsung's successor Hsuan-tsung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (r. 846-59) initially planned to carry out the appointment, but then heeded the advice of his officials, who argued against this course of action by claiming that the Kirghiz were a "secluded, distant, and small" people, and did not send the envoy. [32] Later, however, in 847, Hsuan-tsung reversed his decision once again and did send an official to grant a title to the Kirghiz qaghan. [33] What prompted the emperor's change of heart, described as "sudden" (ts'u [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]), [34] is unknown, but it may have been the result of an attack on T'ang territory in the region of Ho -hsi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] by a combined Uighur-Tangut force, apparently instigated by the Tibetans, the month before the appointment was made. [35]

From 847 to 874, only three more Kirghiz envoys are known to have come to China; subsequent records were lost. [36] No single source names all three envoys; two are mentioned in Tzu-chih t'ung-chien: an Alp Inanc (Ho I-nan-chih [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]), who came to China in 863, and a General Icreki (?) (I-chih-lien-chi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]), who came in 867. [37] As for the third envoy, he is mentioned only in an edict from the brush of emperor Hsuan-tsung himself, dated to the spring of 856. This edict, which addressed the question of granting an appointment to the ruler of the Uighurs in An-hsi, mentions "the Kirghiz Li Chien[CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] who came to the T'ang court. Nothing more is known of his mission, save that he was accompanied by some Uighurs. The emperor's edict is of further interest, however, in that it refers to the Uighurs' former land as "empty"--but with walls and towns remaining--and speaks of th eir desire to recover it. In fact, the edict mentions that many groups of Uighurs had indicated their wish to retake their homeland; in this discussion, there is no mention of a Kirghiz occupation. [38]

These last three known Kirghiz envoys seem to have been interested in trade with China and a regular annual exchange of envoys, as well as an alliance. At least one of them (Alp Inanc) once again brought up the idea of taking An-hsi and other unspecified territories (presumably Pei-t'ing) from the Uighurs, which strongly suggests that the Kirghiz still had not done so. The T'ang government refused this request. [39] In addition to mentioning these envoys, Chinese records state that in 848 a Kirghiz force, said to number 70,000, attacked the Shih-wei [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] people of eastern Mongolia--who were at that time harboring some Uighur refugees--and, after defeating them, took the Uighurs they had captured back "north of the desert" with them. [40] Some of these may have been the Uighurs who later accompanied Li Chien. Hsin T'ang shu, the source that mentions that there were three Kirghiz envoys between 847 and 874, adds that the Kirghiz did "unexpectedly could not seize t he Uighurs.' [41] Since the sources make clear that the Kirghiz did engage in successful raids to capture groups of Uighur refugees, this statement most probably refers either to the large group of Uighurs who fled southeastward and were in the process of establishing a power base in KansU and the northern Tarim Basin, or to the Uighur homeland in Mongolia itself.

In a rather odd postscript to this relatively meager information regarding Sino-Kirghiz relations, Chinese sources indicate that both Kirghiz and Tibetan troops were allied with the Lu-lung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] governor Li K'uang wei [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] and the T'u-yti-hun [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] leader Ho-lien T'o [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (who was based in the region of Ta-t'ung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) in 890 when the latter two drove into the Sha-t'o [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Turk leader Li K'o-yung's [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] territory of Hotung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] and attacked the Che-lu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] army in northwestern Ho-tung. Li K'o-yung was able to defeat Li K'uang-wei and Ho-lien T'o within a short time; the sources do not indicate i f Kirghiz and/or Tibetan troops were present at this defeat. [42] Our sources do not allow us to know if the Kirghiz and Tibetan troops involved in this event were acting on behalf of a particular ruler or were independent mercenaries.

It should be remembered that by this time T'ang China was in a state of severe decline. In 875 began the great Huang Ch'ao [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] rebellion that spelled the beginning of the end for the dynasty. During the last thirty years of T'ang rule, internal affairs were so pressing that the court's range of vision naturally contracted, and there was less interest in--and knowledge of--foreign peoples. The same is generally true of the turbulent Five Dynasties period (907-60). Although Inner Asia was not completely ignored in records of the Five Dynasties, there is almost no mention of the Kirghiz. [43]

All this evidence suggests that the Kirghiz did not expand into the Orkhon region in the mid-ninth century beyond a few raids or sorties. Their great distance from the Chinese border resulted in haphazard and frustrating contacts with the T'ang government. During the 840s the Kirghiz were able to campaign throughout parts of eastern Inner Asia, as their attack on the Shih-wei reveals, but there is no evidence to suggest that such forays continued past 848. We have already seen that their supposed attack on An-hsi and Pei-t'ing most likely is another chimera, and Kirghiz involvement in the 890 raid on Ho-tung tells us nothing about their power on the Mongolian plateau. [44] Thus, while the Kirghiz may have exercised some brief control over the Mongolian plateau immediately following the Uighur collapse of 840, this must have been ephemeral, and no real evidence for it has been found.

A Chinese source does state that after the Kirghiz attack on the Uighurs and the sacking of the Uighur capital, the Kirghiz ruler moved his standard south of the Lao Mountains, which were also known as Tu-man. [45] This was a fifteen-day ride by horse from the Uighur capital. [46] (The Kirghiz ruler's camp in the Yenisei region was said to be a forty-day ride by camel from the Uighur capital. [47] The identity of the Lao Mountains, or Tu-man, is open to some debate. Basing his argument on Chinese information, Chavannes suggested that they were either the Tannu-ola Range or the Sayan Range. [48] Barthold believed them to be probably ([CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) in the Tannu-ola Range; [49] this idea was expanded upon by L. R. Kyzlasov, who believed that after 840 the Kirghiz established their general headquarters ([CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) south of the Tannu-ola Mountains, probably ([CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) in the Tes-Khem Valley near the lake called Ubsu-Nur. [50] This information was repeated by Peter Golden, who presented it without qualification and referred to the e stablishment of a Kirghiz "capital" rather than a general headquarters. [51] It does seem that the Kirghiz enjoyed some control over what is today part of northwestern Mongolia: the region around Ubsu-Nur, just south of the upper Yenisei River and the Tannu-ola Mountains. This would not have been much of an extension of power; indeed, we cannot be certain that this territory was not already under Kirghiz control prior to 840. Neither the duration nor the extent of this control can be determined. [52] It is possible that the movement of the Kirghiz ruler's headquarters was of a temporary nature, only intended for the length of the campaign against the Uighurs.

Although Kirghiz letters to the T'ang court indicated their wish to take over the Uighurs' former lands, there is no evidence that they successfully did so. It must also be remembered that at least some of the displaced Uighurs also communicated to the T'ang court their intention to regain their former state in Mongolia--something they never appear to have actually attempted. Hsuan-tsung's edict of 856, mentioned above, indicated that both the Uighurs who had followed the Kirghiz envoy Li Chien and other Uighurs, who had recently submitted to T'ang officials at the northern border, had expressed their desire to recover their native land, and had also stated that the Uighurs in An-hsi fervently wished for this as well. [53] The T'ang court had consistently exhorted the Kirghiz to attack the Uighurs to eliminate those who remained, [54] but the Kirghiz raids did not succeed in destroying the strongest group of Uighurs at the northern frontier of China. That group, led by Oge (Wu-chieh [CHINESE CHARACTERS NOT R EPRODUCIBLE IN ASCII TEXT] Qaghan, was ultimately defeated by the Chinese themselves. [55] Given the problems of diplomatic exchange between the Kirghiz and China, it maybe that the former considered it either too difficult or insufficiently consequential to send an army all the way to the T'ang border in order to exterminate the remnants of their enemies.

And what do the Chinese sources--the only pertinent ones available for this particular topic--tell us concerning the Khitan (Liao) expansion into Mongolia? In the Liao shih account of A-pao-chi's march to the Orkhon, there is no mention whatsoever of the Kirghiz. Instead, the passage mentions the Khitans attacking a tribe known in Chinese as Tsu-pu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], who have been identified with the Tatars (Ta-ta [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]). [56] The passage further states that A-pao-chi defeated "all the foreign tribes in the Hu-mu-ssu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Mountains." [57] These also could not be construed as identical with the Kirghiz. Although the account in Liao shih consistently mentions Khitan campaigns against even minor tribes, it contains no mention at all of a campaign against the Kirghiz. In fact, Liao shih mentions the Kirghiz only twelve times, in connection with six events or lists:

a "tribute mission" to the Liao court in October/November 952; [58]

another "tribute mission" in December 976/January 977; [59]

a mention that the general(s) of the southwest border caused people of the Kirghiz nation "who admired culture" to come [presumably to the Liao court] in 931; [60]

as a place of exile for A-pao-chi's seditious nephew Yeh-lu P'en-tu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], who was sent to the Kirghiz as an "envoy" in 948; [61]

in the list of "Armies of Dependent States"; [62] and

in the list of "Officials of Dependent States." [63]

As for the two "tribute missions," this is a standard term used for the arrival of all foreign envoys at the Liao court, and cannot be considered as evidence that the Kirghiz were politically subordinate to the Khitans. As for the lists of "dependencies," these too offer no proof, since such nations as Persia (Po-ssu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] and the [CHARACTRS NOT REPRODUCIBLE IN ASCII]Abbasid Caliphate (Ta-shih [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] were also included in them. The most striking information comes from the exile of A-paochi's nephew to the Kirghiz; this would indicate that the Kirghiz realm was indeed far from that of the Khitans.

If there was no Khitan-Kirghiz war, and the Kirghiz nation appears by 924 by have been focused once again in the upper Yenisei region, then what of the idea of a "Kirghiz empire" that was established on the ruins of the Uighur steppe empire? This is also a fiction, based on no solid data.

It would seem, then, that instead of being the site of a "Kirghiz empire," Mongolia from 840 until 924--and indeed until the rise of Cinggis Qan in the twelfth century--was politically fractured among many contending tribes. While existing records do not allow for a clear picture, all available evidence suggests that the Kirghiz never established themselves as a power on the Mongolian steppe, and that the great "Khitan-Kirghiz war" of 924 is nothing more than a figment of scholars' imaginations. In fact, Kirghiz control over the Mongolian plateau seems to have been both weak and exceedingly brief, lasting at most only a few years. The Kirghiz do not appear to have been anywhere near the Orkhon valley when the Khitans entered the region in 924. Although some scholars have avoided the fantasy of the Kirghiz empire and suggested that the Kirghiz did not establish a state centered on the Orkhon valley, [64] none has adequately explained what did happen after 840, and why the Kirghiz remained in the Yenisei basin and did not move southeastward into the Orkhon region.

The question remains: if the Kirghiz did not establish themselves in the Orkhon valley, why did they not? One theory holds that the "sacred" Orkhon region held little appeal for the Kirghiz because they were not originally a Turkic people. Much has been made of Chinese statements concerning the physical appearance of the Yenisei Kirghiz:

Their people are all large with red hair, white faces, and green eyes. They consider black hair to be unlucky. As for those with dark eyes, they say these must be the descendants of [Li] Ling. [65]

On the basis of this statement and also the survival in Chinese texts of at least one apparently non-Turkic word (and possibly others) used by the Kirghiz, it has been thought that the Kirghiz were originally a non-Turkic people who were later Turkicized. Based on references to the Kirghiz as tall and fair-skinned, with light-colored hair and eyes, and on the remnants of a few words in Chinese texts, as reconstructed in T'ang-period pronunciation, some scholars have argued that the Kirghiz may be Samoyedic or Paleo-Siberian in origin. [66] Given the prevalence of linguistic borrowing in Inner Asia, this evidence does not seem sufficiently persuasive to claim that the Kirghiz were originally a non-Turkic people, and it may never be possible to solve the question of the "original" Kirghiz ethnicity or language.

It is quite clear, however, that by the mid-ninth century, the Kirghiz were indeed a Turkic people. Chinese sources reveal that they employed Turkic words such as bas ("beginning"), ay ("moon, month"), and qam ("shaman") as well as some of the Turkic (and pre-Turkic) titulature that was used by the Turks and Uighurs. [67] Furthermore, inscriptions that are apparently Kirghiz have been found in the Yenisei basin that employ the Old Turkic runiform script and use a completely Turkic vocabulary. Chinese sources indicate that the Kirghiz language during the T'ang period was identical to that of the Uighurs [68]-whose language was virtually identical to that of the Turks. It seems ultimately impossible to determine if the Kirghiz were a non-Turkic people who were Turkicized, or were indeed "originally" a Turkic people. It must not be forgotten that Indo-European peoples were present in Inner Asia, including Siberia, in early times, but gradually disappeared, presumably absorbed by other elements of the population .

Still, because some scholars consider the Kirghiz to have been of non-Turkic origin, it has been suggested by Golden that:

The Qirgiz, as might be expected in view of their varied ethnic background, were little concerned with Turkic traditions, the symbolism of the Ottukan yts, its place in Old Turkic imperial thinking. [69]

In another article, the same author expanded on this idea.

[The Qirgizi] were a Turkicized people of an as yet undetermined palaeo-Siberian origin (perhaps Samoyedic) who, undoubtedly as a result of their non-Turkic origins (the process of Turkicization may not even have been completed by this time) and being less attuned to the Turk tradition, broke with many Turk customs. They did not attempt, in the classical manner, to assert their hegemony over the larger Turkic world. They did not, as had been customary, establish their capital in holy, Turk ground around the Orxon and Selenga rivers. Indeed, their hold on Mongolia appears to have been confined to the northwest. Thus, when Apaoki, the ruler of the Mongol-speaking Qitan', campaigned in Mongolia in 924 and seized the old Tuck grounds on the Orxon-Selenga, he had no opposition. [70]

While one can agree with Golden's assertion that the Kirghiz did not establish an empire based on the Orkhon region, it is difficult to ascribe this to "non-Turkic origins." We have seen that the Kirghiz employed traditional Inner Asian titles, suggesting a connection with a very ancient Inner Asian (and, originally, not necessarily Turkic--although many of these titles contain elements that are undeniably Turkic) tradition, of which the centrality of the Orkhon region may well have been a part. As shown in the letters of Li Te-yu the Kirghiz did indeed wish to continue several Turkic-Uighur imperial traditions. Not only did they request imperial appointment from the T'ang emperor, as had earlier Turk and Uighur rulers as well as others, but they specifically wished the name Tengri (Teng-li [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], etc.)--the Inner Asian term for the sky or heavens, as well as for the supreme god associated with the sky, and a term which had been associated with earlier Turk qaghan s and nearly every Uighur qaghan [71]--to be employed in the appointed title. This request was refused by the T'ang court. The Chinese letter insisted that Tengri, as a component of Uighur titles, was not suitable for the titles of the Kirghiz qaghan. [72] The reality was that the T'ang court was not eager to increase Kirghiz prestige; the Chinese must have understood the power of the name Tengri among Inner Asian peoples. Golden himself asserts that the Uighurs "undoubtedly to strengthen their authority after their takeover of the Turk lands, especially emphasized the use of the word Tengri in their titulature.' [73] Just as the term Tengri was important to the Uighurs in their titulature, particularly in an ideological sense, so was it important to the Kirghiz. In addition, we have seen that the Kirghiz did in fact express an intention to move into Mongolia, as well as to take over parts of eastern Turkestan which had at times been under the control of the TUrks and Uighurs. Again: why did they not?

Part of the solution may be found in the ecology of the Kirghiz Yenisei homeland, in a combination of interrelated geographic and economic factors. [74] The upper Yenisei basin is a transition zone that combines a forest-meadow mountain ecosystem with some steppe and wooded steppe zones; it is the southern edge of the Siberian taiga. Such an area of mixed vegetation zones would naturally give rise to a mixed economy, rather than to the more specialized economy of the pastoral peoples of the Mongolian steppe. Clear and significant differences in elevation and climate between the upper Yenisei and Orkhon regions should also be noted; the former, although further north, is at a notably lower elevation. [75] As a result of its geographical situation, the upper Yenisei basin is quite suitable for cereal-crop agriculture, which archaeological evidence shows to have been practiced there since ancient times. Complex irrigation systems date from at least the first millennium B.C. [76]

It is well known that agriculture was important to the economy of the Yenisei Kirghiz. They raised millet, barley, wheat, Himalayan barley, and possibly rye. According to Hsin Tang shu, they sowed in the third month, approximately April, and harvested in the ninth month, approximately October. [77] Today, the upper Yenisei is still known for the growing of grains, including wheat, oats, barley, and rye. [78] Archaeological excavations in the upper Yenisei region have unearthed both locally made iron ploughshares and larger imported iron ploughshares from China. These were from wooden ploughs with iron ploughshares requiring draught animals to pull them. Irrigation was extensive, and remained so until the Mongol period. [79] Grains were harvested with iron sickles and ground with paired rotating millstones. [80]

Available evidence, both archaeological and textual, paints the Kirghiz as being far more involved with agriculture than their Turkic cousins in Mongolia. Although the ninth-century traveller Tamim ibn Bahr mentioned agriculture among the steppe Uighurs, his information is insufficient to determine its extent. [81] It is not my intention to suggest that the Turks and Uighurs practiced no agriculture whatsoever, for they assuredly did, but rather to suggest that the type and scope of Kirghiz agriculture was better suited to the upper Yenisei region than to the Orkhon valley, and this made it difficult for the Kirghiz to move.

The upper Yenisei area is also suitable for the rearing of livestock, which has ancient roots there as well. In order better to exploit the steppe areas of the upper Yenisei, the peoples there apparently practiced not only semi-settled, open-pasture animal husbandry, but also semi-nomadic grassland pastoralism. Among their domestic animals, the most important were cattle, of which a rich farmer could own several thousand head. The Yenisei Kirghiz also raised camels, sheep, and horses. Their horses, described in Chinese records as "quite large," [82] were apparently not the typical Mongolian ponies so well suited to the steppe, but larger horses, presumably of more westerly origin. The raising of cereal grains by the Kirghiz allowed for the storage of grains and hay for fodder, so that winter foraging under the snow would have been unnecessary for these horses. [83]

Hunting and fishing also were important to the Kirghiz economy in the eighth and ninth centuries. The forests and rivers of south Siberia were rich in fur-bearing aaimals and fish. Furs, woolens, and other fabrics were important in Kirghiz trade with eastern Thrkestan, the Arabs, and possibly Tibet. Pelts were also used for internal taxation within the Kirghiz realm. [84]

Chinese sources also mention the excellence of Kirghiz metallurgy at this time. [85] Along the river valleys of the Yenisei basin, remains of ancient blast furnaces have been found, along with slag piles and supplies of coal. [86] Iron forging apparently was the predominant form of metallurgy, although the sources also mention gold and tin. Today, the region in and around Minusinsk and Abakan has significant coal deposits, and iron and gold deposits are quite nearby. [87]

Hsin Wang shu states that in the winter the Kirghiz dwelt in houses with bark roofs. [88] Wang hui yao indicates that the Kirghiz lived in groups along the rivers of their land and adds, "they make houses from wood, and cover them with bark." [89] Log houses are known to have been used in the upper Yenisei region since very early times, as shown by ancient petroglyphs. Tents, however, are also depicted. [90] Chinese sources speak of the Kirghiz ruler's tent, which was surrounded by a wooden palisade "instead of walls." Other Kirghiz leaders are mentioned as living in tents smaller than that of the qaghan. [91]

The picture that emerges from this evidence is one of a semi-sedentary society with well-developed pastoral elements, a society which carried on such settled activities as grain agriculture and advanced metalworking with blast furnaces, as well as animal husbandry that involved some movement of herds. It would seem that at least a portion of the Kirghiz population left the settled (permanent) winter dwellings in the summer to pasture their animals. What remains unknown is just how such tasks were divided among the population, nor do we know the extent of Kirghiz transhumance and/or sedentarization. But by the eighth and ninth centuries A.D., there was clearly a highly mixed economy within the Kirghiz state. [92]

This mixed economy that was well suited to the mixed forest-meadow-steppe environment of the upper Yenisei would not have been easily transplanted to the Mongolian steppe, and perhaps would not have flourished so effectively there. The Kirghiz may have experimented with some sort of political presence in that region, but these experiments would have proved difficult, given the different ecology of the Mongolian plateau and the problems of such a significant relocation/expansion. Since they were under no pressure from other peoples to leave their Yenisei habitat, and since they received no support or encouragement from the Chinese, the Kirghiz apparently chose to retain their established lifestyle in the Yenisei valley and not to make the dramatic change required by a move to the Mongolian plateau. The general difficulty of timing in regard to migrations should also be mentioned; changing pastures was in fact a very risky business for pastoral peoples, and could result in significant loss of livestock. [93] I t is also possible that the large Kirghiz horse, vital to their military campaigns, was unsuited for life on the Mongolian steppe, particularly in winter when it would have been less able than the typical Mongolian pony to forage for grass under the snow, and so would have required fodder. Thus, while the Kirghiz could make military forays into the Mongolian steppe, it is quite possible that they could not easily establish a permanent military or political presence there.

Another important factor is trade. Chinese sources inform us that the Kirghiz engaged in trade with other peoples, including the regions of An-hsi and Pei-t'ing (Besbaliq) and the [CHARACTERS NOT REPRODUCIBLE IN ASCII]Abbasid Caliphate. [94] They also maintained contact with T'ang China--which must have included trade in connection with the Kirghiz "tribute" missions--from 648 (near the end of the reign of T'ang T'ai-tsung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] until 758, at which time they suffered a serious defeat by the then rising power of the Uighurs and so were unable to continue these contacts. [96] But despite their inability to continue trade with China, the Kirghiz managed to maintain lively trade contacts with the [CHARACTERS NOT REPRODUCIBLE IN ASCII]Abbasid Caliphate, the Qarluqs (Ko-lo-lu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] and Tibet. [97] The importance of this southwesterly commercial and political axis can be seen by the fact that the Kirghiz ruler who began hostilities with the Uighurs around 820 was married to the daughter of the Q arluq ruler (yabghu), and his mother was a Turgesh (T'u-chishih [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) woman. [98] The Qarluqs and Turgesh were both western Turkic peoples who lived in the region of Lake Balkhash, north of the T'ian-shan and southwest of the Kirghiz.

It may well be that the Kirghiz did not wish to move into the Orkhon region for fear of being too far removed from their established trade network. Renewed trade with China would certainly have been attractive, and the sources indicate that the Kirghiz wished to trade horses with China, much as the Uighurs before them had done--to the latter's great profit. But it is also clear from our sources that, in the period immediately following the Kirghiz victory, communications between the two were poor; after Wu-tsung's death, it would have been quite apparent to the Kirghiz that the Chinese had no burning desire to reestablish contact. Further Kirghiz embassies to China met with little success. As the T'ang dynasty was becoming increasingly weak, the Chinese were delighted with the collapse of Uighur power, and no doubt pleased that the onerous horse-silk trade which the Uighurs had promoted was finally at an end (although the effect of its cessation on Chinese horse supplies is not known, and may have presented a problem of supply for the T'ang court).. No one in Ch'ang-an was eager to set up the Kirghiz in their place.[99] As for the Kirghiz, their well-established commercial and political ties to the southwest (which linked them to trade with the Tarim Basin, Tibet, and the Middle East) may have been too important and successful to risk in an uncertain--and apparently unwelcome--venture with a weakening China.

Yet another factor in the Kirghiz adherence to their traditional Yenisei homeland may have been problems among the Kirghiz leadership. According to Chinese sources, the Kirghiz ruler who brought down the Uighur steppe empire was himself murdered by one of his officials at some point during the Hui-ch'ang reign period (i.e., 4 February 841-6 February 847), but his death was not reported to the Chinese court at that time.[100] It may have been that the Kirghiz leadership, already discouraged by the slowness of their communication with the T'ang court, did not want to reveal the ruler's death for fear that it would cause further delays. Furthermore, we cannot ignore the general problems of political cohesion typical of Inner Asian states. It may be that local Kirghiz leaders (chieftains) were willing to come together in order to throw off Uighur rule--and its concomitant taxes--and to plunder the Uighur realm. Once that had been accomplished, and the booty distributed, these leaders may well have realized that a continued effort to exercise control over the Mongolian plateau would yield diminishing returns, and so would have been less willing to contribute to the efforts necessary for such control.

In conclusion, the historical gap for the Mongolian Plateau from about 840 to about 924 remains troublesome, but at least there is some evidence to indicate what did not happen. It seems that the Kirghiz presence there was neither long nor significant, save for its destructive power. It can be safely asserted that no Kirghiz empire was established in the Orkhon valley or near it. We hear little of the Kirghiz after the brief flurry of contacts with China in the early 840s, save for a few diplomatic missions and other isolated references. There was no great battle for the Orkhon between Kirghiz and Khitans.

As for the Uighurs, many of them had fled, although it is likely that at least some remained in the Orkhon region, but they too were unable to play an important role. It would seem plausible that the creation of a political vacuum in the Orkhon region allowed other peoples to move into the area--or, if they were already there as former subordinates of the Uighurs, to enjoy an increased level of freedom and, possibly, local power. One such people were the Tatars (Tsu-pu), who were encountered by the Khitan ruler A-pao-chi when he campaigned into Mongolia in 924 and defeated them. The Orkhon region then became--for a time--a remote outpost of the Khitan Liao empire.

(1.) Only the Chinese form of this name, with its variants Ju-ju [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Juan-juan [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], etc., is known.

(2.) Anatoly Khazanov makes an excellent point when he argues that the term "confederation" can be misleading when applied to the multi-ethnic polities created by nomadic peoples, since "nomadic associations are not always formed on a voluntary basis"; see A. M. Khazanov, Nomads and the Outside World, tr. Julia Crookenden (Cambridge: Cambridge Univ. Press, 1984), 152. Keeping this in mind, the term retains some utility in that it suggests a polity in which the various sub-units retain a high degree of autonomy.

(3.) Gerard Clauson, An Etymological Dictionary of PreThirteenth-Century Turkish (Oxford: Oxford Univ. Press, 1972), 976. See also Drevnetiurkskii slovar' (Leningrad: Izdatel'stvo "Nauka," 1969), 268.

(4.) See K. Czegledy, "Coyay-quzi, Qara-qum, Kok-ong," Acta Orientalia Academiae Scientiarum Hungaricae 15 (1962): 62, n. 5.

(5.) Talat Tekin, A Grammar of Orkhon Turkish (Bloomington: Indiana Univ. Publications, 1968), 234, 267; see also Alessio Bombaci, "Qutluy Bolsun! (Part two)," Ural-altaische Jahrbilcher 38 (1966): 17--18.

(6.) Tekin, A Grammar of Orkhon Turkish, 261. I have modified this and other of Tekin's translations to conform to the styles and forms (spellings) I have used throughout this paper.

(7.) Tekin, A Grammar of Orkhon Turkish, 262.

(8.) Tekin, A Grammar of Orkhon Turkish, 285.

(9.) On the religious connotations of the Otuken, see Jean-Paul Roux, "La Religion des Turcs de l'Orkhon des [VII.sup.e] et [VIII.sup.e] siecles (deuxieme article)," Revue de l'Histoire des Religions 161.2 (1962): 200-201; and Alessio Bombaci, "Qutluy Bolsun! (Part two)," 15-19. See also Annemarie von Gabain, "Steppe und Stadt im Leben der altesten Turken," Der islam 29.1 (1949): 35-37.

(10.) See Paul Pelliot, "Neuf notes sur des questions d'Asie centrale," T'oung Pao, ser. 2, 26 (1929): 212-19.

(11.) For the Sine-usu inscription, see G. J. Ramstedt, "Zwei uigurische Runeninschriften in der Nord-Mongolei," Journal de la Societe Finno-Ougrienne 30 (1913): 22-27. For the Terkhin inscription, see S. G. Klyashtorny, "The Terkhin Inscription," Act Orientalia Academiae Scientiarum Hungaricae 36 (1982): 341-46; and Talat Tekin, "The Tariat (Terkhin) Inscription," Acta Orientalia Academiae Scientiarum Hungaricae 37 (1983): 46-51. These two scholars present readings that are often highly divergent.

(12.) See Michael R. Drompp, "The Writings of Li Te-yu as Sources for the History of T'ang-Inner Asian Relations" (Ph.D. diss., Indiana Univ., 1986), for a study of those Uighurs who fled to the Chinese border.

(13.) Some have argued that the brief Suci Inscription (written in Turkic language and Turkic runiform script) provides evidence for Kirghiz control over Mongolia; see, for example, L. R. Kyzlasov, Istorila Thvy v srednie veka (Moscow: Izdatel'stvo Moskovskogo Universiteta, 1969), 95. But there is nothing in the inscription that really provides such evidence. For the Turkic text and a German translation of the inscription, see Ramstedt, "Zwei uigurische Runeninschriften in der NordMongolei," 1-9.

(14.) Edouard Chavannes, "Voyageurs chinois chez les Khitan et les Joutchen," Journal asiatique, neuvieme serie, 9 (1897): 381-82. He was a bit less cautious a few pages later (p. 407, n. 1), writing: "Les Kirghiz, qui habitaient primitivement sur les bords du Kem ou haut Ienissei, avaient etendu leur domination jusqua I'Orkhon apres leurs victoires sur les Ouigours, au milieu du [IX.sup.e] siecle." Still, there is no reference to a Kirghiz empire or anything of the sort, nor to the duration of that domination.

(15.) W. W Barthold [= V. V. Bartol'd], Zwolf Vorlesungen Uber die Geschichte der Tiirken Mittelasiens (1935; rpt. Hildesheim: Georg Olms Verlagsbuchhandlung, 1962), 97. In his article on the Kirghiz for Encyclopedia of Islam, Barthold was once again cautious, indicating that the Kirghiz, after conquering the lands of the Uighurs in Mongolia, were probably driven from Mongolia in connection with the founding of the Khitan empire "and the advance of the Mongol peoples." See W. Barthold, "Kirgiz," Encyclopedia of Islam, first ed. (Leiden: E. J. Brill, 1913-1936), 2:1025.

(16.) V.V. Bartol'd (= W. W. Barthold), "Kirgizy," in his Sochineniia [(Collected) Works], vol. 2 (Moscow: Akademiia Nauk SSR, 1963), 498-99. The article was originally published in 1927 in Frunze (Pishpek). In the same article (p. 489), Barthold noted that after the victory over the Uighurs, the Kirghiz ruler did not establish a capital on the Orkhon, but did move (according to Chinese sources) to the south side of the Lao [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Mountain(s), or Tu-man [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Barthold believed this to refer to the Tannu-Ola Mountains.

(17.) Karl Wittfogel and Feng Chia-sheng, History of Chinese Society: Liao (Philadelphia: American Philosophical Society, 1949), 106.

(18.) Rene Grousset, L'Empire des steppes (Paris: Payot, 1948), 176, 181. Grousset also indicated (p. 181, n. 1) that he had obtained information regarding this topic from E. Bretschneider, Medieval Researches from Eastern Asiatic Sources: Fragments towards the Knowledge of the Geography and History of Central and Western Asia from the 13th to the 17th Century (London: Kegan Paul, Trench, Trubner & Co., Ltd., 1910), 1: 265 (apparently an error for 241), but there is nothing there to suggest that Bretschneider's work did indeed contribute to Grousset's information concerning a war between the Khitan and the Kirghiz. Bretschneider simply gives a brief description of the destruction of the Uighur empire by the Kirghiz.

(19.) There are many examples. See Karl H. Menges, The Turkic Languages and Peoples: An Introduction to Turkic Studies, UralAltaisehe Bibliothek, vol. 15 (Wiesbaden: Otto Harrassowitz, 1968), 24; Denis Sinor, Inner Asia: A Syllabus (Bloomington: Indiana Univ. Publications, 1971), 129; V. Minorsky, tr. ed., Hudud al-[CHARACTERS NOT REPRODUCIBLE IN ASCII]Alam lam, 'The Regions of the World'. A Persian Geography 372 A.H.-982 A.D., E. J. W. Gibb Memorial Series, n.s., vol. 11 (London: Luzac & Co., 1937), 282; Sechin Jagchid and Van Jay Symons, Peace, War, and Trade along the Great Wall: Nomadic-Chinese Interaction through Two Millennia (Bloomington: Indiana Univ. Press, 1989), 43; Svat Soucek, "Kirghizia," in Encyclopedia of Asian History (New York: Charles Scribner's Sons, 1988), 2: 318-19; L. P. Potapov, "The Khakasy," in The Peoples of Siberia, ed. M. G. Levin and L. P. Potapov, tr. Stephen Dunn (Chicago: Univ. of Chicago Press, 1964), 345; and Berthold Spuler, "Gesehichte Mittelasiens sell dem Aufcreten der Turken' in Geschichte Mittelasiens, H andbuch der Orientalistik, vol. 5, part 5 (Leiden: E. J. Brill, 1966), 188. Spuler does indicate a few pages earlier (pp. 164-65) that after 840 the Kirghiz maintained their political focus on the upper Yenisei region, and did not continue the traditions of the Turks and Uighurs; perhaps he believed that the Kirghiz held some sort of sway over Mongolia, which was ended by the Khitans, but never chose the Mongolian plateau as the political center of their nation. This opinion seems reflected in Thomas J. Barfield, The Perilous Frontier: Nomadic Empires and China (Oxford: Basil Blackwell, 1989), 164-65; Barfield states that the Kirghiz did not establish an empire, but "appeared content with the loot from Karabalghasun and went home." Nevertheless, he does go on to indicate his belief in a Khitan-Kirghiz conflict. The comments in Annemarie von Gabain ("Steppe und Stadt im Leben der tiltesten Ttirken," 49) reflect not only this chimera but other confusions as well. The Cambridge History of Early Inner Asia, ed. Denis Sinor (Cambridge: Cambridge Univ. Press, 1990), ignores the question, while The Cambridge History of China, vol. 6: Alien Regimes and Border States, 907-1368, ed. Herbert Franke and Denis Twitchett (Cambridge: Cambridge Univ. Press, 1994), 65, simply states that "in 924-5 A-pao-chi led a great expedition into the steppe, which conquered the tribes of northern Mongolia and reached as far as the old Uighur capital city Ordu Baliq [sic], on the Orkhon River." There is no suggestion of a war with the Kirghiz, or a Kirghiz presence in Mongolia, but neither are the "tribes of northern Mongolia" identified.

(20.) A letter from the T'ang court to an Uighur minister indicates that the Kirghiz expressed the desire to move into the Uighurs' former territory and then extend their sway over the region of the northern Tarim Basin. This particular passage in Li Te-yu's works is problematic in that there are possible lacunae; see Ts'en Chung-mien [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] "Li Te-yii Hui-ch'ang fa p'an chi pien cheng shang" [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Shihhsiieh chuan-k'an [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] 2.1 (1937): 203, n. 7; see also Drompp, "The Writings of Li Te-yu," 224 and the pertinent notes, pp. 256-63. Another letter, written to the Kirghiz qaghan in response to a letter he had sent to the T'ang court, again stated the Kirghiz plan to take the Uighurs' former territory; see Drompp, "The Writings of Li Te-yu," 323.

(21.) Li Te-yu, (Li Wei-kung) Hui-ch'ang i p'in chi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Shanghai: Shang-wu yin-shu-kuan, 1936), 8.64; for a translation, see Drompp, "The Writings of Li Te-yu," 224.

(22.) Liu Hsu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] et al., Chiu T'ang shu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Peking: Chung-hua shu-chu, 1975), 174.4522-23, Ou-yang Hsiu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] et al., Hsin T'ang shu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Peking: Chung-hua shu-chu, 1975), 180.5337; see also Ssu-ma Kuang [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Tzu-chih t'ung-chien [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Peking: Ku-chi ch'u-pan-she, 1956), 247.7973-74. The account presented in Chiu T'ang shu is translated in Drompp, "The Writings of Li Te-yu," 284-86.

(23.) See, for example, S. G. Kljaltornyj, "Das Reich der Tataren in der Zeit vor Cinggis Khan," Central Asiatic Journal 36.1-2 (1992): 76-77. The Chinese texts, however, are ambiguous. Tzu-chih t'ung-chien 246.7968 suggests that the Kirghiz had already taken An-hsi and Pei-t'ing, but the text from which this account is clearly derived--a letter from the brush of Li Te-yu (see n. 20 above)--indicates that i [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] "already" is an error for i [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] "in order to"; see Li Te-yu, Hui-ch'ang i p'in chi 8.64, translated in Drompp, "The Writings of Li Te-yu," 224. Li Te-yu's letter suggests that the Kirghiz were planning such an attack, but had not actually carried it out; this would certainly make far better sense given the context, since the Kirghiz clearly were asking for Chinese assistance--something they would hardly do had they already subjugated these regions. Tsai Wen-shen (on whose Turkish translation of the relevant passage in Tzu-chih rung-chien Kljaltornyj based his interpretation) was familiar with Li Te-yu's writings and also preferred this interpretation of the text, rendering the Chinese verbs into Turkish future tense and thereby asserting that the Kirghiz actions against An-hsi and Pei-t'ing were at that point planned, not accomplished. See Tsai Wen-shen, "Li Te-yu'nun mektuplarsna gore Uygurlar (840-900)" (Ph.D. diss., Taipei, 1967), 148.

(24.) Li Te-yu, Hui-ch'ang i p'in chi (addendum), 285. For a translation of this letter, see Drompp, "The Writings of Li Teyu," 289-92.

(25.) Li Kuang's biography may be found in Ssu-ma Ch'ien [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] Shih chi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Peking: Chung-hua shu-chu, 1962), 109.2867-76; Pan Ku [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] et al., Han shu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Peking: Chung-hua shu-chu, 1964), 54.2439-49. The former has been translated by Burton Watson, Records of the Grand Historian of China: Translated from the Shih chi of Ssu-ma Ch'ien, 2 vols. (New York: Columbia Univ. Press, 1961), 2:141-52; the latter by Watson, Courtier and Commoner in Ancient China: Selections from the History of the Former Han by Pan Ku (New York: Columbia Univ. Press, 1974), 12-23. Li Ling's biography may be found in Shih chi, 109.2877-78 and in Han shu, 54.2450-59. The latter, more detailed, has been translated by Watson, Courtier and Commoner in Ancient China, 24-33. Note that this kinship arrangement places the T'ang ruling house, as descendants of Li Kuang, in the senior position and the Kirghiz ruling house in the junior.

(26.) Hsin T'ang shu, 217B.6146-47.

(27.) Hsin T'ang shu, 217B.6150. On the Chinese acceptance of the existence of kinship between the two ruling houses, see Li Te-yu, Hui-chang i pin chi, 6.38, translated in Drompp, "The Writings of Li Te-yu," 280. Note that this fictive kinship had been "recognized" at least as early as the Ching-lung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] reign period (5 October 707-4 July 710) of the T'ang emperor Chung-tsung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]; see Hsin T'ang shu, 217B.6149. The Kirghiz ruler is initially referred to in Chinese sources as a-jo [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], the original form of which is unknown, and later as qaghan.

(28.) Li Te-yu, Hui-chang i pin chi, 6.39. This letter has been translated in its entirety in Drompp, "The Writings of Li Te-yu," 303-9.

(29.) See Drompp, "The Writings of Li Te-yu," 318-25.

(30.) On the dating of this letter, see Drompp, "The Writings of Li Te-yu," 350, n. 111.

(31.) Tzu-chih t'ung-chien, 248.8015, 8023, 8026.

(32.) Tzu-chih t'ung-chien, 248.8026.

(33.) Tzu-chih t'ung-chien, 248.8030; Hsin T'ang shu, 217B.6150.

(34.) Hsin T'ang shu, 217B.6150.

(35.) Tzu-chih t'ung-chien, 248.8030.

(36.) Hsin T'ang shu, 217B.6150-51.

(37.) Tzu-chih t'ung-chien, 250.8107 and 8117. The Kirghiz (presumably Turkic) form of I-chih-lien-chih has not been firmly established. The T'ang pronunciation suggests something like Turkic icreki, "belonging to the royal court," perhaps "court chamberlain" (Clauson, An Etymological Dictionary of Pre-Thirteenth-Century Turkish, 31); this is a known element in Turkic titles, but the third syllable, with its final -n, is particularly problematic. See Todo Akiyasu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], Gakken Kan-Wadaijiten [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Tokyo: Gakushu Kenkyusha, 1978), 28 (i), 562 (chih), 1320 (lien), and 416 (chi).

(38.) Tung Kao [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] et al., comps., (Ch'in ting) Ch'uan T'ang wen [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Taipei: Ta-t'ung shu-chu, 1979), 80.1045-46; and Sung Min-ch'iu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], comp., T'ang ta chao-ling chi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Taipei: Ting-wen shu-chu, 1978), 128.692-93. Ch'uan T'ang wen gives no date for the edict, while T'ang ta chao-ling chi dates it to the second month of the tenth year of the Ta-chung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] reign period, that is, 10 March-8 April 856. Tzu-chih t'ung-chien, 249.8059, however, dates the edict to 16 April 856. The envoy's name, Li Chien, which is Chinese rather than Turkic, points to the connection between the Kirghiz, Li Ling, and the T'ang ruling house of Li.

(39.) Tzu-chih t'ung-chien, 250.2172.

(40.) Tzu-chih t'ung-chien, 248.8032.

(41.) Thin T'ang shu, 217B.6150.

(42.) Hsin T'ang shu, 218.6161; Tzu-chih t'ung-chien, 258.8404-5.

See also The Cambridge History of China, vol. 5: Sui and T'ang China, 589-906, part I,775-76, which does not mention the participation of the Kirghiz or Tibetans. For more information on the location of the Che-lu Army, see Tzu-chih t'ung-chien, 253.8206 and Yen Keng-wang [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] T'ang-tai chiao-t'ung t'u-k'ao [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], vol. 5 (Taipei: Chung-yang yan-chiu-yuan li-shih yuwen yen-chiu-suo, 1986), 1407. On the career of Ho-lien T'o and his animosity toward Li K'o-yung, see Gabriella Mole, The T'u-yu-hun from the Northern Wei to the Time of the Five Dynasties, Serie Orientale Roma, vol.41 (Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1970), xxii-xxiii, 195-206.

(43.) Hsin T'ang shu, 217B.6150-51 mentions the three subsequent Kirghiz envoys from 847 up through the Hsien-t'ung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] reign period (17 December 860-16 December 874), and suggests that there were some later envoys "inquiring about official appointments," the records for which have been lost. Tzu-chih t'ung-chien mentions Kirghiz envoys to China in 863 and 867; see Tzu-chih t'ung-chien, 250.8107, 8117.

(44.) Indeed, these Kirghiz may have been mercenaries; there is no evidence that they were acting on behalf of any Kirghiz ruler.

(45.) See n. 16 above.

(46.) Hsin T'ang shu, 217B.6150.

(47.) Hsin T'ang shu, 217B.6148.

(48.) Edouard Chavannes, Documents sur les Tou-kiue (Turcs) occidentaux, recueillis et commentes, suivi de notes addition-elles (1903; rpt. Paris: Librairie d'Amerique et d'Orient AdrienMaisonneuve, n.d.), 98, n.2. According to Hsin T'ang shu, 43B.1149, the Lao Mountains (and the Chien = Turkie Kem = Yenisei River) were to be found in the territory of the Kirghiz. On the identification of the Chien/Kem with the Yenisei, see Louis Hambis, "Notes sur Kam, nom de I'Yenissei superieur," Journal asiatique 244 (1956): 281-300.

(49.) Bartol'd, "Kirgyzy," 489.

(50.) L. R. Kyzlasov, Istoriia Tuvy v srednie veka, 94. Kyzlasov gives no evidence or source to support his location of the Kirghiz general headquarters after 840. Lake Ubsu-Nur is found in northwestern Mongolia, near the border with Tuva and the town of Ulangom.

(51.) Peter B. Golden, "The Migrations of the Oguz," Archivum Ottomanicum 4 (1972): 60. In discussing the Kirghiz attack on the Uighurs, Hsin T'ang shu, 217B.6149 states that the Kirghiz ruler personally acted as commander of his troops. After sacking and looting the Uighur capital (and obtaining the T'ai-ho Princess, whom he attempted to have escorted back to China), he then moved his camp to the south of the Lao Mountains. It seems that this was not a permanent move, but one involved with the Kirghiz campaign into Mongolia. The Kirghiz sent many raids into Mongolia for several more years as mop-up campaigns against the remnants of the Uighurs; this camp in the Tannu-ola or Sayan region might have served as a base headquarters for those raids. The Chinese word for "camp" or "standard," ya [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], was also used to refer to the Uighur capital, but in this case it seems better here to translate it as "camp, general headquarters," etc.

(52.) It is interesting to note that just to the south of Ubsu-Nur is Hyargas-Nur, or "Kirghiz Lake."

(53.) Ch'iian T'ang wen, 80.1046; T'ang ta chao-ling chi, 128.693.

(54.) See, in particular, Li Te-yu's letters to the Kirghiz ruler in Hui-chang i p'in chi, 6.37-38, (addendum) 285-86, 6.38-40, and 6.40- 42. All of these have been translated in Drompp, "The Writings of Li Te-yu," 276-82, 289-92, 303-9, and 320-24.

(55.) The story of the Chinese defeat of the Uighur refugees is told in detail in Drompp, "The Writings of Li-Te-yu."

(56.) Wittfogel and Feng, History of Chinese Society, 101-2.


(58.) Liao shih, 6.70; 70.1135.

(59.) Liao shih, 8.96; 70.1137.

(60.) Liao shih, 3.32; 70.1129-30.

(61.) Liao shih, 5.64; 61.937; 77.1257; 113.1508.

(62.) Liao shih, 36.431.

(63.) Liao shih, 46.758.

(64.) See, for example, Ts'en Chung-mien, Sui T'ang shih [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (rpt. Hong Kong: Perlen Book Co., n.d.), 2: 394, as well as Peter B. Golden, An Introduction to the History of the Turkic Peoples (Wiesbaden: Otto Harrassowitz, 1992), 180-83. Despite his remarks here, the chimera of a Khitan-Kirghiz war crept into at least one of Golden's later works, "Cumanica IV: The Tribes of the Cuman-Qipeaqs," Archivum Eurasiae Medii Aevi 9 (1995-97): 113.

(65.) Hsin T'ong shu, 217B.6147. For a similar statement, see Wang P'u [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] et al., T'ang hui yao [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Taipei: Shih-chieh shu-chu, 1974), 100.1785, which attributes the idea of darkhaired and dark-eyed Kirghiz as descendants of Li Ling to one Ko Chia-yun [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] who served as protector-general (normally tu-hu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] but here written tu-hu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) of An-hsi and wrote a work, no longer extant, entitled Record of the Western Regions (Hsi yu chi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]). Note that Ko Chia-yun served in a number of important posts in the western regions of the T'ang Empire from 734 to 741; see Wu T'ing-hsieh [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] T'ang fang-chen nien-piao [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (Peking: Chung-hua shu-ch u, 1980), 3:1203, 1221, 1232-33.

(66.) See, for example, W. Schott, "Uber die achten Kirgisen," Abhondlungen der koniglichen Akademie der Wissenschaften, phil.-hist. KI. (1864), 441-47 (but note that many of Schott's ideas have since been proven to be incorrect); Louis Ligeti, "Mots de civilisation de Haute Asie en transcription chinoise," Acta Orientalia Academiae Scientiarum Hungaricae 1 (1950- 51): 150-68; Kyzlasov, Istorija Tuvy v srednie veka, 88-90; and Golden, An Introduction to the History of the Turkic Peoples, 176-78.

(67.) Hsin T'ang shu, 217B.6147, 6148.

(68.) Hsin T'ong shu, 217B.6 148 states that the language and script of the Kirghiz were exactly the same (cheng t'ung [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]) as that of the Uighurs. See also Michael R. Drompp, "A Note on Interpreters of Turkic Languages in Late T'ang China," in Altaic Religious Beliefs and Practices: Proceedings of the 33rd Meeting of the Permanent International Altaistic Conference, Budapest, June 24-29, 1990, ed. Geza Bethlenfalvy et al. (Budapest: Research Group for Altaic Studies, Hungarian Academy of Sciences, and Department of Inner Asiatic Studies, Eotvos Lorand University, 1992), 103-9.

(69.) Golden, "The Migrations of the Oguz," 60. Golden states that the Kirghiz "did not, as had been customary, establish their capital in Mongolia on the Orxon or Selenga but rather on the south side of the Tannu-ola range in the Tes-Xem valley near lake Ubsu-Nur." His reference to the Tanno-ola comes from Barthold, although Golden does not indicate that this identification of Lao Mountains/Tu-man with Tannu-ola might be questioned.

(70.) Peter B. Golden, "Imperial Ideology and the Sources of Political Unity amongst the Pre-Cinggisid Nomads of Western Eurasia," Archivum Eurasiae Medii Aevi 2 (1982): 54. Again, note the apparent contradiction between the final sentence here and the statement found in the same author's "Cumanica IV," 113, that the Khitans drove the Kirghiz out of Mongolia in the tenth century.

(71.) In the runiform Orkhon inscriptions, the Turk ruler Bilge Qaghan referred to himself as "Heaven-like and Heaven-created" (Tengriteg Tengride bolmis Turk Bilge Qaghan); see Tekin, A Grammar of Orkhon Turkish, 231, 275. In addition, the penultimate ruler of the Turks was called Tengri Qaghan; see Chiu T'ang shu, 194A.5177-78. On Uighur titles from the steppe empire period, see Colin Mackerras, The Uighur Empire According to the T'ang Dynastic Histories: A Study in Sino-Uighur Relations (Columbia: Univ. of South Carolina Press, 1972), 192-93.

(72.) Li Te-Yu Hui-chang i pin chi, 6.39; for a translation, see Drompp, "The Writings of Li Te-yu," 305-6.

(73.) Golden, "Imperial Ideology," 45.

(74.) Anatoly Khazanov has written: "Conditionally, all forms of pastoralism may be regarded as different methods of economic adaptation, the parameters of which are determined, in the final analysis, by ecology and level of technological development"; see Nomads and the Outside World, 69.

(75.) See, for example, The Times Atlas of the World, 8th ed. (New York: Times Books, 1990), plates 3, 6, and 42.

(76.) Sevyan Vainshtein, Nomads of South Siberia: The Pastoral Economies of Tuva, ed. Caroline Humphrey, tr. Michael Colenso, Cambridge Studies in Social Anthropology, vol. 25 (Cambridge: Cambridge Univ. Press, 1980), 145.

(77.) Hsin T'ang shu, 217B.6147. It should be noted that this information regarding Kirghiz agriculture, along with much other information on the Kirghiz that is to be found in Hsin T'ang shu and other sources (but is not in Chiu T'ang shu, which has no separate account of the Kirghiz) probably derives from accounts that were written down after T'ang officials interviewed the Kirghiz envoy Chu-wu Alp Sol (Chu-wu Ho Su [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]), who arrived at the Chinese court on 16 March 843. (On Chu-wu Alp Sot and his mission, see Drompp, "The Writings of Li Te-yu," 282-94.) Although these accounts have not survived, documents from Li Te-yu's brush indicate that the Kirghiz envoys were questioned about their land, geography, etc.; this seems the most likely source for the information recorded in Hsin Tang shu. See Li Te-yu Hui-ch'ang i pin chi, 2.11-12 and Hsin T'ang shu, 217B.6150.

(78.) See J. C. Dewdney, USSR in Maps (New York: Holmes & Meier, 1982), 41 and 99; see also Oxford Regional Economic Atlas: The U.S.S.R. and Eastern Europe (Oxford: Oxford Univ. Press, 1956), 34-35 and 38-39.

(79.) Kyzlasov, Istorija Tuvy v srednie veka, 116. See also Alexander Mongait (tr. Yevgeny Ganushkin), Archaeology in the U.S.S.R. (Moscow: Foreign Languages Publishing House, 1959), 306-7, as well as A. P. Okladnikov, "Ancient Population of Siberia and Its Cultures," Russian Translation Series of the Peabody Museum of Archaeology and Ethnology, Harvard University, vol. 1, no. i (New York: AMS Press, 1959), 57.

(80.) Okladnikov, "Ancient Population of Siberia," 57; Kyzlasoy, Istoriia Tuvy v srednie veka, 116.

(81.) See V. Minorsky, "Tamim ibn Bahr's Journey to the Uyghurs' Bulletin of the School of Oriental and African Studies 12 (1948): 283 and 296. Minorsky believed that Tamim's journey should be dated to 821.

(82.) Hsin T'ang shu, 217B.6147.

(83.) Kyzlasov, Istoriia Tuvy v srednie veka, 118.

(84.) Hsin T'ang shu, 217B.6147-48.

(85.) Hsin T'ang shu, 217B.6147. 147.

(86.) Potapov, "The Khakasy." 346.

(87.) See Oxford Regional Economic Atlas: The U.S.S.R. and Eastern Europe, 54-55, 62-63, and 70-71; see also George Kish, Economic Atlas of the Soviet Union, 2nd ed. (Ann Arbor: Univ. of Michigan Press, 1971), 54 and 59, as well as Dewdney, USSR in Maps, 49, 51, and 99.

(88.) Hsin Fang shu, 217B.6148.

(89.) Fang hui yao, 100.1784.

(90.) Okladnikov, "Ancient Population of Siberia," 34.

(91.) Hsin T'ang shu, 217B.6148.

(92.) Vainshtein, quoting the work of S. S. Sorokin, states that "agriculture was a common feature of all forms of pastoralism which are tied to alluvial valleys of large rivers and to mountain regions." See Nomads of South Siberia, 164. On the question of semi-nomadism and semi-sedentarism, see Khazanov, Nomads and the Outside World, 19-22.

(93) Anatoly Khazanov, "Ecological Limitations of Nomadism in the Eurasian Steppes and Their Social and Cultural Implications," Asian and African Studies 24.1 (1990):2.

(94) Hsin T'ang shu, 217B.6147-48.

(95) Hsin T'ang shu, 217B.6149; Chiu T'ang shu, 3.60-61.

(96) Chiu T'ang shu, 195.5201. For a translation, see Mackerras, The Uighur Empire According to the T'ang Dynastic Histories, 66.

(97) Hsin T'ang shu, 217B.6149. This source indicates that the route to and from Tibet went through Qarluq territory, and that the Kirghiz served as escorts at some stages of this route, so as to prevent the Uighurs from robbing the travellers. It also states that the [CHARACTERS NOT REPRODUCIBLE IN ASCII]Abbasids would trade cloth to the Kirghiz once every three years. Most likely this trade involved other products as well. For a further discussion of this trade route, see Christopher I. Beckwith, The Tibetan Empire in Central Asia: A History of the Struggle for Great Power among Tibetans, Turks, Arabs, and Chinese during the Early Middle Ages (Princeton: Princeton Univ. Press, 1987), 147 and 147-48, n. 21.

(98) Hsin T'ang shu, 217B.6149.

(99.) Hsin T'ang shu, 217B.6150.

(100.) Hsin Fang shu, 217B.6150. The text suggests that the murder of the

Kirghiz ruler occurred before the Kirghiz envoy Chu-wu Alp Sol, who arrived at Ch'ang-an on 16 March 843, was Sent on his mission. Kyzlasov suggests a date of 847 for the Kirghiz ruler's death, but gives no evidence to support this; see Istorija Tuvy v srednie veka, [95].
COPYRIGHT 1999 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:The Journal of the American Oriental Society
Geographic Code:90ASI
Date:Jul 1, 1999

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters