Post by ALTAR on Mar 2, 2010 13:30:34 GMT 3
www.transoxiana.org/Eran/Articles/benjamin.html
The Yuezhi Migration and Sogdia
Craig Benjamin
Introduction
Following the defeat of the formerly powerful Yuezhi confederation by the Xiongnu near Dunhuang in c.162 BCE, the Yuezhi dynasty and those tribes that remained loyal to it commenced a migration away from the Gansu that was destined to completely reshape the geopolitics of ancient Inner Asia. Both the Han Shu and Shi Ji provide evidence of their departure: "the Yuezhi had fled furious with the Xiongnu"1, the 'Yuezhi had fled and bore a constant grudge against the Xiongnu'2. The decision to migrate, despite still having a force of perhaps 100,000 armed archer warriors at their disposal is indicative of the severity of the defeat, and also of the steadily increasing power of the Xiongnu under Maodun and Jizhu during the preceding decades. The Yuezhi dynasty may in fact have considered such a move several times during the fourteen years between Maodun's initial raid against them in 176, and their ultimate defeat in 162.
Indeed, the fact that the migration seems to have been conducted in an orderly fashion suggests something of a planned strategic relocation rather than a rout. The Yuezhi's original intention was to move some 2000 kilometres to the northwest and resettle in the valley of the Ili River, a region occupied by a group of Sakas (or Scythians). They had no intention, nor any idea, that this journey would only be the first stage of a migration that ultimately would take them half away across Central Asia, until thirty years later they would find themselves in secure occupation of the fertile river valleys north of the Amu Darya, and masters of the former Greek kingdom of Bactria.
Leaving the Gansu in 162 the Yuezhi headed northwest towards the Ili Valley, settling near Ysyk Kul in present-day Kazakhstan. Corroborative evidence for this new location is provided by the Greek geographer Ptolemy who mentions a group called the Tagouraioi (clearly a variation on Tocharian, the Indo-European language spoken by the core Yuezhi) dwelling near Ysyk Kul3. Russian archaeologist Yu Zadneprovsky has noted a substantial number of podboy sites in the region, which he has tentatively identified as Yuezhi on the basis of their similarity to other podboy tombs discovered at the Haladun site near Minqin in the Central Gansu, which he also argues are Yuezhi. The Ysyk Kul region is rich in nomadic burial sites and some 370 tombs had been noted by as early as 1960. Of these, 80% were in pits, and attributed to the autonomous Sakas, and 17% in podboys, tentatively attributable to the Yuezhi4.
The Chinese sources show that the Ili/Ysyk Kul region was already populated by the Sai people, an eastern concentration of Sakas or Scythians who probably spoke an Indo-Iranian language. Upon arriving at the Ili, the Yuezhi quickly displaced them: (The Yuezhi) 'attacked the king of the Sai (who) moved a considerable distance to the south and the Yuezhi then occupied his lands'5. Most of the displaced Sakas subsequently undertook their own substantial migration, moving west and then south through the western Tarim Basin, crossing the so-called 'Suspended Crossing' (probably the Khunjerab Pass or similar, between present-day Xinjiang and northern Pakistan) before ultimately settling in Kashmir6.
The Yuezhi confederation occupied the former Sakan lands in the hope of permanently resettling there, and remained in residence for almost three decades. They no doubt felt they had successfully relocated, having escaped the Xiongnu menace and reestablished themselves in the fertile Ili Valley. They returned to their previous semi-nomadic, semi-sedentary lifeway and probably began to lose interest in Chinese/nomadic politics. But the Kunmo of the Wusun, the former neighbours of the Yuezhi in the Gansu, could not forget the ill treatment his people had suffered as a result of a Yuezhi attack upon them in 1737. Eventually he sought permission from his Xiongnu overlord (the new ShanyuJunchen, successor to Jizhu who had died in 158) to pursue the Yuezhi into the Ili and 'avenge his father's wrongs'8. In 132 BCE the Kunmo led a powerful force of mounted Wusun archers into the region which attacked and routed the no doubt surprised and dismayed Yuezhi, forcing them to once again uproot and resume their long march to the west.
The sources indicate that within a short space of time the Yuezhi passed through a region called Dayuan 'The Yuezhi thereupon went far away, passing Dayuan and proceeding west'9 and then through a land to the southwest called Kangju. Although the exact route remains a matter of some conjecture, the evidence of both the Chinese annals, and of Russian and Central Asian archaeology, leaves little doubt that the Dayuan through which the migrating horde passed can only be identified with the Ferghana Valley10. The Yuezhi apparently met with little resistance from the urbanised Dayuans/Ferghanese, and after perhaps spending some months (the winter of 132/1?) in southwestern Ferghana, they passed on unmolested. Zadneprovsky has also noted several single podboy burials that have been unearthed in the southwestern, northern and eastern parts of the Ferghana Valley, most concentrated in the Lyailyaka-Isfara-Sokha interfluve in southern Kyrgyzstan where over 300 podboy burials have been located. Although originally attributed to a separate culture by Baruzdin in 1960, Zadneprovsky argues for their re-attribution to the migrating Yuezhi, on the basis of their similarity to podboy sites also tentatively attributable to the Yuezhi in both the Gansu and Ysyk Kul region11.
Perhaps in the spring of 131 BCE then, the Yuezhi most probably moved from Ferghana into the 'state' of Kangju, probably the Zeravshan Valley in the heart of Sogdia. Some four or five years later they were followed through the region by Han envoy Zhang Qian, who was led there by guides and interpreters provided for him by the king of Dayuan. It is references to Kangju in the Han Shu and Shi Ji (and by Ptolemy in his Geographica), as well as the discoveries of Soviet and Russian archaeologists, that has provided evidence identifying Kangju with Sogdia, and thus of the role of Sogdia in both the migration of the Yuezhi and the mission of Zhang Qian. The intention of this paper is to consider the origins of the relationship that developed between the Kangju and Yuezhi dynasties, a relationship that subsequently evolved to provide vital political and military stability in the region throughout the Kushan Era. The initial task is to consider evidence that allows for the conclusive identification of Kangju as Sogdia.
References to Kangju and the Yuezhi in the Han Annals
Kangju Sogdia: Location and Lifeway
Location
'The seat of the king's government in winter is in Leyuenidi to the town of Beitian. It is distant by 12,300 li from Xian. One reaches (Le)yuenidi after a journey of seven days on horseback, and it is a distance of 9,104 li within the realm to the king's summer residence. To the east it is a distance of 5,500 li to the Seat of the Protector General'12.
'It is said: Some 2000 li to the northwest from Kangju is the state of Yancai. The trained bowmen number 100,000. It has the same way of life as Kangju. It is situated on the Great Marsh, which has no further shore and which is presumably the Northern Sea'13.
'Kangju is situated some 2,000 li northwest of Dayuan. The country is small and borders Dayuan. It acknowledges nominal sovereignty to the Yuezhi people in the south and the Xiongnu in the east'14
'(Wusun) adjoins Kangju in the northwest'15
'(The State of Wusun) and 5000 li to the west, to land within the realm of Kangju'16
Attempts by scholars over several centuries to geographically locate and delineate Kangju have not been helped by textual corruption in both the Han Shu and Shi Ji. And yet, although several words and even whole sentences are missing, the information provided is still in the same order as that for the other 'western states', so that any gaps cannot be substantial. Certainly the distances between Xian and Beitian are not quite reconciled, and the distance from Beitian and the king's summer capital (9104 li or 3641 kms) is surely corrupt. Hulsewe and Loewe suggest that the text may originally have read 'ninety one li' (or 36 kilometres, although this seems too low),17 while Pelliot noted Wang Kuowei's suggestion of 1104 li (441 kms)18 which is a more viable figure within a country described as 'small'.
The identities of both Beitian and (Le)yueni(di) are almost impossible to determine, however, Wang Kuowei identified the former (impossibly) with Ysyk Kul19 while Pulleyblank has argued that the latter might 'represent some form of the name Jaxartes'20. The distance between Beitian and (Le)yueni(di) is described as 'seven days on horseback' in the Han Shu, which Hulsewe and Loewe suggest equals about 500 li i.e. marches of seventy lior 28 kilometres per day through the mountainous country of the region21. The identification of these two principal settlements with Samarkand and Bukhara is one obvious possibility, although the distance between the two cities by road is about 200 or so kilometres which does not reconcile with any of the given statistics.
Pulleyblank discusses the possible Tokharian philological origin of the name 'Kangju', in his reconstruction of 'Old Chinese' *khan-kiah. In the Tokharian vocabulary (Tokharian 1A) there is the word kank, which means 'stone'. Thus Kangju could mean the 'Stone Country', i.e. Samarkand (or equally Tashkend as 'Stone City')22. A.K. Narain offers a precise geographical location for Kangju: 'the northeastern wedge of modern Uzbekistan into Kirghiziya and Kazakhstan; the eastern part of this wedge formed part of Dayuan'23. This description, however, does not allow for the inclusion of any lands south of the Syr Darya, thus excluding the entire Zeravshan Valley, the cultural heart and population centre of Sogdia.
The information provided by the texts is hardly ambiguous, however, and clearly suggests the identification of the 'state' of Kangju with ancient Sogdia. Kangju is to the north of the Amu Darya and the Yuezhi's principal city of Jianshi (Khalchayan in the Surkhan Darya valley?); to the west and northwest of the Ferghana Valley (where it also apparently adjoined the clearly very substantial, post-132 realm of the Wusun); and southeast of the western realms of the Xiongnu (which must therefore have included the steppes of present-day Kazakhstan). Kangju incorporated lands on either side of the middle Syr Darya, particularly the densely occupied Zeravshan Valley south of the Syr Darya, and must surely have included Samarkand and Bukhara (as Shishkina also argues below). Hence, according to the textual evidence at least, Kangju can only convincingly be located within the general geographical region of ancient Sogdia.
Population/Size
Households: 120,000 Individuals: 600,00024
'The country is small'25
The physical dimensions of the Kangju realm may not have been vast, but the population was substantial, which allowed the ruling dynasty to maintain a formidable military force.
Military Strength
Persons able to bear arms: 120,00026
'They have 80,000 or 90,000 skilled archer warriors'27
'(it) is not subject to the Protector General'28
'In the east (the inhabitants) were constrained to serve the Xiongnu'29
'It acknowledges nominal sovereignty to (Zurcher translates as 'it is subservient to')30 the Yuezhi people in the south and the Xiongnu in the east'31
'However, Kangju felt that it was separated (from Han) by a long distance, and alone in its arrogance it was not willing to be considered on the same terms as the various other states'32
'(He Wudi - heard that) to the north, there were (people or places) such as the Da Yuezhi and Kangju, whose forces were strong; it would be possible to present them with gifts and hold out advantages with which to bring them to court'33
The Chinese were clearly impressed by the strength of Kangju, finding them arrogant and militarily self-confident. The military resources of Kangju (120,000 armed men, 80,000 - 90,000 of which were skilled and presumably mounted archers) were substantial, and would not easily be defeated by the Han. Presumably the ruling Kangju dynasty and its pastoralist allies provided the bulk of the mounted archer warriors, while the sedentised agriculturists of the river valleys could be relied upon to provide the remainder. Eschewing any military option then, Zhang Qian argued instead (in his report to Wudi) that the Kangjuans could be persuaded by Han gifts and favours to consider becoming subjects (or at least allies) of the Chinese. In short, Kangju was powerful and remote enough to resist Han attempts to join their tributary confederacy by military means, but was clearly under some sort of sovereignty obligation to both the Yuezhi and the Xiongnu.
Environment/Lifeways
'The way of life is identical with that of the Da Yuezhi'34
'Its people likewise are nomads and resemble the Yuezhi in their customs'35
'In Kangju there are five lesser kings all the five kings are subject to Kangju'36
The last reference clearly indicates that 'Kangju' should be considered both as the Han name of the 'state' (that is the realm or region) of Kangju/Sogdia, but also of the dominant faction or dynasty which was controlling that realm at the time (i.e. the Kangju dynasty). Shishkina agrees, and argues that geo-political changes in Sogdia that became apparent towards the end of the second century BCE must have been as a result of Kangju hegemony:
'The historical situation of the first century BC suggests that these changes were related to the spread of the power of the Kangju, when this dynasty controlled Samarkand and Bukhara'37
The five lesser kings noted in the Han Shu were probably subordinate tribal groups within the realm of 'greater Kangju', and given that all are listed as having specific 'seats of government' (different to the two principal settlements named as belonging to the Kangju proper), may represent sedentised, agrarian-based 'peoples' living under Kangju hegemony38
The way of life of the dominant Kangju faction was probably that of semi-nomadic militarised pastoral nomadism, similar to the assessment of the lifeway of the Yuezhi soon after their arrival north of the Amu Darya that Zhang Qian provided to the Han court. If the Kangju state is thus to be identified with ancient Sogdia under Kangju dynastic hegemony, then a brief account of Sogdian history prior to the arrival of the Yuezhi is required to identify the probable date of the establishment of Kangju power, and also to clarify the archaeological and textual evidence of subsequent Yuezhi/Kangju interaction.
Sogdian Historical Framework Prior to the Arrival of the Yuezhi
Between 553 and 550 BCE, Cyrus II (r. 559-529), a leader of the Persian Achaemenid family, overthrew Astyages, King of the Medes, and brought Mesopotamia, Parthia and Anatolia under his control. By 539 he had conquered Bactria and much of Sogdia as well, where he established a line of fortresses on the Syr Darya. Sogdia was made the thirteenth satrapy of the Achaemenids, and paid tribute to Cyrus' successors. The oldest layers of Afrasiab, the ancient site of Samarkand, date from this Achaemenid period. But, whilst the city states of Sogdia and Bactria gained considerably through their incorporation into the Achaemenid Empire, they remained intent upon regaining their independence, which parts of Sogdia may have done by c. 400 BCE39
Some two centuries after Cyrus' death, Alexander of Macedon reconquered much of Central Asia, following his arrival in Bactria in 329 BCE. Alexander's principal opponent in the region was the Achaemenid ruler Darius' former satrap, Bessus, who had Darius murdered in modern-day Shahr-I Qumis before proclaiming himself as his successor. Bessus' troops consisted of armoured cavalry from Bactria and Sogdia, which, following their defeat at Gaugemela, he took back across the Amu Darya after destroying its bridges. Alexander led his troops on forced marches through the desert, crossed the Amu Darya on inflated hide rafts, and confronted his opponents who immediately sued for peace.
Bessus was executed; the Macedonians installed themselves in the satrapal palace at Maracanda (Samarkand) and Sogdia, following some seventy years of independence, found itself incorporated into the new Macedonian Empire. But while Alexander campaigned further north along the Syr Darya, the Sogdians, under the leadership of Spitamenes, rose in his rear and massacred a garrison of Macedonians, inflicting arguably the worst defeat of Alexander's career40. Over the course of the ensuing eighteen months Alexander gained his revenge by reducing the fortified towns of Sogdia one by one, starting in the Hissar Mountains and moving along the Zaravshan Valley41.
At the heart of ancient Soghd were the valleys of the Zeravshan and Kashkadarya, and in his vengeful campaign along these densely occupied valleys the Macedonians may have killed up to 120,000 Sogdians42.
Amongst the many prisoners captured during the Sogdian campaign was the Princess Roxanne, daughter of another Sogdian opponent, Oxyartes. Alexander's subsequent decision to marry Roxanne was due partly to her exceptional beauty, but was also intended as a gesture to appease the rebellious Sogdians. After Alexander's death in Babylon in 323, Bactria and Sogdia immediately rebelled but were reconquered in c. 305 by his successor Seleucus Nicator (r. 311-281 BCE). However under Seleucus' son Antiochus I (r. 281-261), Bactria and (probably) Sogdia broke away again from Seleucid hegemony. None the less, Sogdia, along with much of Central Asia, was brought into the orbit of Hellenistic influence during its brief period of Macedonian conquest.
Antiochus I minted an extensive local coinage in the region, probably at Balkh (the 'capital' of Bactria). These were coins of large denominations: staters, tetradrachms and drachms. During the last two centuries before the Common Era, several series of diverse denominations and types were struck at Sogdian mints, and coins were widely used in Sogdia and Bactria, although perhaps only by the Greek population43
None the less the native population of Sogdia became used to Greek coinage during the Seleucid period, and when the inflow of Greek coins stopped following their independence from Antiochus I, local rulers began to mint their own. As Nymark has pointed out, however, these local issues were highly debased, and in fact were 'mere imitations of the most widespread Greek coins'44.
Yet these imitations remain as crucial (and often the only) evidence of political, economic and social developments in Sogdia during the first century BCE. Furthermore, both Sogdian and Bactrian imitation issues also constitute potential evidence for the Yuezhi during the 'five-yabghu period'.
In the mid-third century (c. 250) the Seleucid Governor of Bactria, Diodotus, established an independent Graeco-Bactrian kingdom, which may also have exercised a degree of control over Sogdia. In c. 230 Diodotus' son was overthrown by one of his satraps, a Greek settler called Euthydemus, who then ruled Graeco-Bactria for about forty years until c. 190 BCE. If Sogdia was indeed part of the incipient Graeco-Bactrian state, then the evidence of the Euthydemus imitation coinage indicates that some time late in the third century, during the lifetime of Euthydemus, Sogdia became an independent entity once more45.
Euthydemus concluded a peace treaty with the Seleucid ruler Antiochus III in 206, but did not attempt to reconquer Sogdia. Instead the Graeco-Bactrians expanded south into India, establishing the Indo-Greek kingdoms. If the 'state' of Kangju is indeed to be identified as Sogdia, then it was during this period of post-Seleucid independence, i.e. from c. 210 BCE, that the region came under the hegemony of the Kangju dynasty, which then continued to rule an independent Sogdia until it came under Kushan political influence early in the first century CE. As Bopearachchi concludes, under the Kangju dynasty 'Sogdiana probably remained free at least until the arrival of the Yuezhi in c. 130 BC'46.
The Passage of the Yuezhi through Kangju/Sogdia
Although neither of the Chinese sources categorically states that the Yuezhi horde passed through Kangju, the only logical inference to be drawn from the texts is that they did. In addition, Ptolemy continued to unknowingly chart the course of the Yuezhi migration by noting a group he this time called the Tachoroi (surely another variant of Tocharian) dwelling in Sogdia47.
The conclusion that the Yuezhi must have passed through the region is further strengthened by the fact that the Han sources do unequivocally show that Zhang Qian passed through Kangju during his search for the Yuezhi. The Han envoy was obviously well informed by the rulers of Dayuan as to the route followed by the Yuezhi who provided him with guides to lead him to the Yuezhi, unless they knew the migrants' route and probable whereabouts? And thus is likely to have followed closely in the original footsteps of his quarry. The chronology is straightforward enough.
C. 132/1 BCE: The Yuezhi depart Dayuan and continue their migration to the west
That the Yuezhi continued westwards in their migration following their passage through (and possibly winter residency in) Dayuan is implicit in the key Han Shu passage: 'passing Dayuan (and) proceeding west to subjugate Daxia'48.
There are three feasible route options west from Ferghana, whether starting from present-day Kukon in the centre of the valley, or Isfara in the southwest. The first is due north and then west, across the 2267 metre Kamcik Pass and through Angren into Tashkent, thence southwest to Samarkand. A second and more direct route is due west through present-day Chugand and Zizzach, thence southwest into Samarkand. However, given that the Zeravshan Valley was the agricultural and population heartland of Sogdia/Kangju (information probably given to the Yuezhi by the rulers of Dayuan who were no doubt anxious to encourage the Yuezhi to move on and seek suitable settlement lands elsewhere), the migrating horde may have chosen to follow a third route option from Chugand, south over the 3378 metre Sahristan Pass, then down into the upper Zeravshan Valley. If the Yuezhi leadership decided upon this latter route, then it would indeed have been necessary for them to winter in southern Ferghana before attempting the crossing of this high pass in the spring.
The Shi Ji also implies this in noting (in Watson's translation) that the Yuezhi 'moved far to the west, beyond Dayuan'49, which Zurcher reads as: 'They passed through Dayuan and to the west of thatcountry'50. Between Dayuan and Daxia (Bactria) lay only Kangju/Sogdia; anyone moving to the west, beyond Dayuan (or to the west of that country) and heading for northern Bactria would have to have passed through Sogdia. This probability is then strengthened by the unambiguous statement that Zhang Qian was taken to Kangju by his Dayuan guides and interpreters, and from there proceeded directly to the realm of the Da Yuezhi in northern Bactria.
129/128 BCE: Zhang Qian also passes through Kangju
The Han Shu notes that:
'(the king of Dayuan) sent off (Zhang) Qian, providing him with interpreters and guides. He reached Kangju who passed him on to the Da Yuezhi'51.
Or as Sima Qian puts it:
'The king of Dayuan trusted his words and sent him on his way, giving him guides and interpreters to take him to the state of Kangju. From there he was able to make his way to the land of the Great Yuezhi'52.
Despite its obvious military strength, Kangju (like Dayuan) also apparently facilitated (or at least did not hinder) the passage through its territory of both the migrating Yuezhi horde in c. 131 and the Han envoy in c. 128 BCE. Given the size of its military resources, Kangju was powerful enough to be not 'easily defeated by Han forces',53 although it was 'constrained to serve the Xiongnu' in the east,54 and (later) would acknowledge 'nominal sovereignty' (or even become 'subservient to') the Yuezhi in the south'55.
Does this acknowledgement suggest that parts of Sogdia (and the most populous parts at the Zeravshan valley and Samarkand) were actually invaded and defeated by the migrating Yuezhi, and then forced into a subordinate relationship thereafter? Certainly Torday is prepared to argue that not only did the Yuezhi defeat the Kangju dynasty in Sogdia, but in northern Bactria as well where he suggests the Kangju were also ruling:
'We must accept that they took the region from the Kangju by force, in a war. His (the Yuezhi king's) subsequent conquest of Sogdiana was probably a by-product of his ambitions further south-east where he had brought Daxia under his sway'56.
If Torday is correct, this says much about the military capabilities of the migrants that a displaced, previously fragmented and essentially homeless nomadic tribal confederation, soundly defeated twice during the previous three decades by the Xiongnu and the Wusun, was none the less able to invade and defeat the well-defended state of independent Kangju Sogdia. But there is simply no evidence to support Torday's reconstruction, and indeed his argument is unconvincing. He has the Yuezhi settling in the Jaxartes (Syr Darya) delta rather than northern Bactria at the end of their migration, which is clearly incorrect as textual and archaeological evidence indicates. He then suggests that the Yuezhi defeated the Kangju dynasty to gain control of Bactria, where the evidence shows it was a group of Sakas (who had established hegemony over those regions of Bactria to both the north and south of the Amu Darya) that the Yuezhi were forced to defeat and evict to gain control of the region.
Ultimately there is no suggestion (other than the mention of 'nominal sovereignty') that the Yuezhi were ever forced to confront the Kangju militarily. If there had been any need for conflict, then given the subsequent history of the region (including the invasion of Sogdia by the Karakhanid Dynasty in the 11th Century, for example), the acceptance of Yuezhi suzerainty by Kangju should more accurately be seen as an example of an all-too-familiar phenomenon in Central Asian history, and the greater military power of pastoral nomadic states over semi-sedentary, irrigation-based city states, whatever the relative size of their forces. Despite the fact that the ruling dynasty of Kangju was probably of a militarised, semi-nomadic ancestry, the bulk of the population was surely sedentised agriculturists.
The evidence might alternatively be interpreted as suggesting that Kangju (both the state and the dynasty) adopted a conciliatory and diplomatic position towards most of its neighbours, even including the migrating Yuezhi horde. Kangju acknowledged nominal sovereignty (or was subservient) to both the Xiongnu and the Yuezhi; they also assisted the Chinese by helping (or at least not impeding) Zhang Qian in his attempt to locate the Yuezhi and subsequently even sent a 'royal' son as envoy to the Han Court during the reign of Emperor Zheng. Undoubtedly, like the Wusun, Kangju felt that the Han court was remote and refused to become subject to the Protector General, but the dynasty was careful to offend no one Han, Xiongnu or Yuezhi. It is therefore more likely that the Kangju rulers of Sogdia agreed to allow the Yuezhi unmolested passage through their territory, and accepted some form of subservient relationship thereafter to avoid military conflict.
It is not at all necessary to envisage violent military confrontation between the two dynasties, but rather a situation where local rulers recognised that the Yuezhi were powerful enough to be worth accepting as symbolic overlords. Certainly the later Mongol invasions would show that where cities made peace with the migrating, invading force without offering resistance, they could survive and prosper under the rule of pastoralist conquerors, whereas those who refused were destroyed utterly. The apparent ease with which the Yuezhi subjugated Bactria (just to the south of and contiguous to Sogdia) a year or so later may also have been a salutary reminder to the Kangju dynasty. And, just as the Dayuans might have suggested the Zeravshan Valley in Sogdia as a possible resettlement location for the migrants, the rulers of Sogdia could in turn have nominated the fertile valleys north of the Amu Darya in northern Bactria for the Yuezhi to relocate, thus discouraging any thoughts the latter might have entertained of remaining in Sogdia itself.
The careful relationship established between the two dynasties in c.131 BCE became an enduring one. Kangju Sogdia would be drawn further and further into the Yuezhi/Kushan sphere of influence over the following centuries until substantial portions of its territory may have been incorporated into the Kushan Empire (although this is by no means certain). Certainly, according to the Hou Han Shu, by 83 CE the Kushans were happy to further cement this relationship through an alliance based on a 'bond of royal marriage' with the ruling family of Kangju57. Even as early as the last three decades of the second century BCE Kangju proved to be an ideal buffer for the Yuezhi between their new homeland north of the Amu Darya, and the Wusun and Xiongnu to the north and east.
The Kangjuans (or at least some elements of the Sogdian populace) also apparently acknowledged sovereignty to the Xiongnu in the east (presumably northeast),58 indicating the astonishing reach of Xiongnu influence. This is confirmed by the Han Shu inclusion of a statement in Zhang Qian's description of the 'state' of Dayuan that '(the area) west of Wusun as far as Anxi is close to the Xiongnu'59. That is, Xiongnu influence apparently reached from the steppes of eastern Mongolia to as far west as Parthia. Even allowing for possible Han exaggeration or misinformation, this means that the Xiongnu were in seasonal occupation of (or at least exercised some form of nominal hegemony over) a very substantial western realm indeed, including the steppelands of present-day Kazakhstan and Uzbekistan, from Lake Balkash to the Aral Sea.
It is also possible that this 'nominal sovereignty' took the ambiguous form of gift-giving, which the Xiongnu and Chinese sources interpreted as subordination, but which the locals (including the ruling dynasties of Dayuan and Kangju) understood merely as diplomatic courtesies. There is little or no evidence that the Xiongnu actively exercised military power further to the west, although they may have occupied temporary sites along the Middle Syr Darya, if the archaeological evidence of a burial mound at Zhaman-Togai attributed to the Xiongnu by the tomb's Soviet discoverers in 1968 is accepted60. Torday is prepared to read this acknowledgment of nominal sovereignty as proof of Xiongnu military subjugation of Kangju, and even attempts to date the beginning of that subjugation from early in the reign of Maodun. He finds the Han Shu passage evidence of 'Kangju's humiliation by Xiongnu in the east', and links this humiliating defeat to a steppe battle conducted by Maodun at Zaysan Nor in 203 BCE61. Torday's reconstruction is imaginative but there is simply no evidence to support it. In the hope of corroborating the textual suggestions of Yuezhi/Kangju interaction, one must turn to a consideration of archaeological evidence.
Archaeological Evidence of the Yuezhi in Kangju/Sogdia
Introduction and History
The textual references thus strongly imply that during the latter stages of their migration the Yuezhi horde probably passed through some regions of the 'state' of Kangju Sogdia. However, as is the case with most aspects of Yuezhi history, archaeological confirmation of this apparent course of textually-attested events is hardly overwhelming. The most likely location for the discovery of material evidence would surely be somewhere in the valley of the Zeravshan River, which has its source between the Turkestan and Zeravshan ranges in Tajikistan, and then flows west through Samarkand and out into the deserts of Uzbekistan. In ancient times the valley was fertile and sheltered, providing excellent conditions for early Bronze Age agrarian settlements. The discovery in 1976 of the Bronze Age settlement of Sarazm on the upper reaches of the river showed that the valley had been settled for thousands of years, perhaps since as early as the mid-4th Millennium BCE62.
Archaeological investigation of the middle and lower Zeravshan Valley commenced in 1940 with a survey carried out on the construction site of the Katta-Kurgan reservoir, halfway between Samarkand and Bukhara, which resulted in the discovery of Sogdian burial mounds of the 'Kushan period'63. Between 1945 and 1949 Terenozhkin conducted systematic research in the immense town site of ancient Samarkand, Afrasiab64. This was followed by expeditions from the Uzbek Branch of the Soviet Academy of Sciences in the late 50s and early 60s, which thoroughly investigated a series of sites around Samarkand65. In 1946, meanwhile, the Academy of Sciences of the Tajik Soviet Socialist Republic began investigating the upper reaches of the Zeravshan, east of Samarkand, which, under the direction of Yakubovsky, laid the ground work for a systematic archaeological survey of the entire length of the upper regions of the valley66. Yakubovsky's preliminary work was carried on through the 1960s by Mandleshtam in particular, who discovered numerous burial mounds and settlements, many of them apparently left by migrating nomadic groups during the early Kushan era (as Soviet archaeologists tended to label the period between the second century BCE and the late second/early third century CE)67. However, within this broad and general chronology the attribution of particular types of burial structures and fortified settlements to specific groups of pastoral nomads was again very difficult. Mandleshtam was the first to attempt to narrow the search down somewhat in his address to the 1968 conference on the Kushans in Dushanbe:
'The dating of the investigated monuments (which is substantiated by the results of recent excavations of town-sites) enables us to state with certainty that they belonged to the nomads, who in the last third of the second century BC destroyed the Graeco-Bactrian Empire. Archaeology reveals four groups of nomads, which probably relate to four different tribes'68.
Obelchenko, however, who from 1952 onwards carried out large-scale excavations of nomadic burial mounds in the middle and lower Zeravshan, particularly around Bukhara, linked all of the funerary monuments discovered in the Zeravshan to an invasion of 'Sarmatian' tribes in the second to first centuries BCE69. Gorbunova has summarised Obelchenko's mid-50s theory thus:
'The discovery and investigation of Sogdian pastoralists' sites were begun by O.V. Obelchenko. He isolated among the cemeteries an earlier group dating from the second century BC to the first century AD, and a later group dating from the second to the fourth centuries AD. He regards both sites as relics of the Sarmatian tribes whose attacks contributed to the final defeat of Graeco-Bactria'70.
Obelchenko's conclusions were questioned by Mandleshtam, who argued that the attribution of all of the mounds as Sarmatian was too generalised and superficial, particularly as various groups of burials displayed quite distinctively different characteristics.
'The culture of all of them (the burial mounds) exhibits traits resembling the culture of the Sarmat tribes, but this is mainly a 'temporal' resemblance, which is observed over a vast area. A more concrete comparison points to links with areas to the N and NE of Central Asia'71.
Zadneprovsky also disagrees with Obolchenko,72 and has argued that similarities between possible Yuezhi tombs in northern Bactria and those discovered in the Bukharan oasis by Obelchenko suggest that the latter are also relics of the passage of the Yuezhi through Sogdia: 'The coincidence of the design, funeral ceremony and the accompanying inventory definitely point to their similarity'73.
Artifacts from Nomadic Tombs in Ferghana and Sogdia
However, even a detailed analysis of the 'accompanying inventory' of grave goods discovered at the various nomadic tomb sites does not clarify the matter74. Pottery vessels (including censers) occur in all types of graves, as would be expected. Gorbunova argues that 'they have a strictly local character, originating in the pottery centres nearest to the cattle-breeding people's area of habitation, or being diffused only over the local area'75. Thus jugs (with and without handles) are common to the whole region, although they vary in shape from one location to the next. Bactrian and Sogdian jugs and goblets were almost all fashioned on a potter's wheel, which differentiates them from Khorezmian ceramics, for example.
Weapons have also been discovered in most nomadic cemeteries throughout former Soviet Central Asia, including swords, daggers, iron arrowheads and fragments from composite bows, all weapons generally associated with militarised nomads. Swords and daggers with a cross hilt are characteristic of both the Sogdian and Bactrian sites, whereas the majority of swords from other areas (i.e. areas probably not visited by the migrating Yuezhi) do not have cross-hilts, thus allowing for possible identification of the unique Sogdian and Bactrian examples as Yuezhi. Also characteristic of Bactrian and early Sogdian sites are arrowheads with barbed triangular flanges, which differ markedly from the triangular-flanged, straight-based arrowheads found in Khorezmia, Turkmenia and at Kaunchi culture sites (near Tashkent). The graves of Ferghana, emphasising that region's role throughout the Bronze and Iron Ages as an obvious funnel for all manner of Eurasian nomads, contain arrowheads of almost every type76. The arrowheads discovered at early Kushan sites in Sogdia are similar to those found at Begram in Bactria, thus dating them to the 'late-Yuezhi/early Kushan period'77.
Household objects, including knives, spindle-whorls and a variety of toilet-articles also demonstrate considerable local variation. Bronze mirrors have been found at sites all over former Soviet Central Asia, with Khorezmia, Sogdia and Bactria yielding mirrors of a similar so-called 'Sarmatian' type, and Ferghana again yielding the largest number and greatest variety of examples78.
Belt buckles have also been found in extensive numbers, indicating that most pastoral nomads probably wore belts with buckles. Differentiation of buckle types is also a potential source of tomb identification. The link between Sogdian and Bactrian sites is again apparent Gorbunova notes that 'belt buckles from Bactrian and early Sogdian cemeteries are obviously of the same kind, and differ from the buckles of other places'79. The nomadic cultures associated with the different tomb types thus display a series of distinctive individual traits through their material possessions, but also a range of other features commonly shared between all nomadic 'peoples', which increases the difficulties of definitive attribution. Undeniably there is substantial evidence of nomadic passage along (and even temporary occupation of) various sites in the Zeravshan Valley, and it remains a matter of ongoing interpretation as to which tombs and funerary objects (if any) might be convincingly identifiable as Yuezhi.
Shishkina argues, perhaps too dismissively, that there is no evidence whatsoever of Yuezhi influence, although his conclusion that the densely populated Zeravshan valley proved ultimately unsuitable as a possible relocation site for the nomads is self-evidently correct:
'In the second century BC the area between the Amu Darya and the Syr Darya was the object of a massive invasion by peoples of a foreign culture, of which tribes of the Great Yuezhi made up a considerable part. Nevertheless the character of the Hellenized culture was preserved and there are no traces of new influences. One may assume that (the Yuezhi) passed through the Zeravshan Valley peacefully and did not remain there for long. There certainly would not have been enough room on the densely populated and cultivated lands of Soghd for masses of migratory peoples'80.
Shishkina's claim that there is 'no trace' of the passage of the nomads obviously takes no account of Zadneprovsky's argument that significant numbers of podboy tomb types, similar in style and design to those found in other areas of probable Yuezhi occupation (the Gansu, Ysyk Kul and Ferghana), have also been unearthed along the Zeravshan. Furthermore, in addition to the similarity of tomb-type, many of the artifacts discovered in podboy tombs also display a range of common features, particularly those found in tombs in Sogdia and Bactria; Gorbunova notes similarities between 'Bactrian and early Sogdian cemeteries with their distinctive types of pottery, weapons, belt buckles and women's clothing'81. Whilst it is clearly impossible to prove that these tombs and their artifacts were left along the Zeravshan by the migrating Yuezhi, there is certainly sufficient evidence to at least entertain the possibility.
'Sogdian' Statuettes
Further archaeological evidence of Yuezhi activity in Sogdia might be provided by the attribution of a particular sub-group of Sogdian statuettes unearthed at several sites in and around Samarkand, particularly at Tali Barzu south of the city, and also at Afrasiab. A form type of female statuette discovered at these (and other nearby) sites, characterised by the position of the hands under the breasts or resting on the upper stomach, and with distinctive facial features and dress, has been analysed by Fiona Kidd of Sydney University82. The primary focus group of figurines is distinct and stable in form a female figure with small protruding breasts and very slim arms bent sharply at the elbows so that her hands are resting on her upper stomach. Related groups include statuettes with stumps instead of arms, or with arms resting on the lower rather than upper stomach. The three groups are united by their similar style of facial features. As a stylistic form, similar examples are known from Mesopotamia at least as early as the 2nd Millennium BCE, although it has been suggested that the form probably dates much further back into antiquity83.
Where there might be some (admittedly very tenuous) link between a particular sub-type of these figurines and the Yuezhi is through a comparative coroplastic analysis of costume styles. Examples from the Samarkand area show a variety of clothing styles pleats on the lower part of the dress; a looser dress worn over long trousers with folds falling from the waist; a long pleated dress with pleats falling from the waist; or with pleats falling all the way from the shoulders84. Abdullaev has tentatively attributed one particular group of figurines clothed in a long dress with a flared lower half to the Yuezhi during their interim residency of the Zeravshan Valley85. The identification is somewhat strengthened by the discovery of similarly attired figurines at other Yuezhi and early Kushan sites, notably Tillya-tepe and Dalverzin-tepe86.
Sogdia lay at the heart of a network of ancient migration and trade routes, and it is only to be expected that Sogdian art would be subjected to a range of stylistic influences, including those of pastoral nomads, be they steppe- or oasis-dwellers. Kidd argues that 'the nomadic legacy of the peoples living in this region must also be recognised as an important factor in the formation of style in Sogdian art generally'87. That the Yuezhi occupied, if only for a relatively brief period, parts of the Zeravshan valley including the greater Samarkand oasis is more than likely, given the cumulative textual and archaeological evidence considered above. The possibility that further evidence of their residency might be found through a coroplastic analysis of the costume of figurines unearthed at probable Yuezhi sites in the region is thus also worthy of consideration.
Archaeological Evidence: Conclusion
Archaeological evidence for the migrating Yuezhi confederation in both Sogdia and Ferghana is at best inconclusive. Mass migrations of nomadic 'peoples' would hardly be expected to leave a substantial archaeological record anyway, given that they generally did not construct settlements or 'townships' meant to last longer than a winter season. During the long, cold Central Asian winters the dead would be buried beside the encampments, and in the spring the tribes would move on in search of new dwelling places. This would surely have been the lifeway of the Yuezhi during the short period of forced migration following their eviction from the Ili Basin in c. 133/2 and preceding their arrival in northern Bactria in c. 130 BCE. Thus the only record of their passage through the Ferghana and Zeravshan valleys would be their funerary monuments and the grave goods they contained. The incidence of podboy tombs in both regions (mostly with a common north-south orientation), the similarity of household objects, weapons, arrowheads and belt buckles in those tombs, and the possible evidence of Sogdian statuettes, provides clearly circumstantial but arguably cumulatively convincing evidence that there does indeed exist a material record of the passage of the Yuezhi. Textual and archaeological evidence thus allows for the following tentative narrative reconstruction of events.
Conclusion and Summary
As a result of their defeat by the Wusun in 133/2 BCE, and after almost three decades of residency in the Ili Valley, the confederation of the Da Yuezhi was forced to resume its migration westwards, moving initially into the Ferghana Valley. Here they may have briefly occupied sites in the Isfara region before continuing their journey (in the spring of 132/1?), travelling west and south into the Zeravshan valley of Sogdia, before perhaps following it into the Samarkand and Bukharan oases. This itinerary is implied by several references in the Han Shu and Shi Ji, and also tentatively reinforced by Russian and Central Asian archaeology.
In the penultimate stage of their journey the Yuezhi horde entered the realm of Kangju, a dynasty in control of a region most probably identifiable as Sogdia and the Zeravshan Valley. 'Kangju' was the Han name for both the 'state' and the powerful ruling dynasty, which may have been in control of Sogdia since it gained independence from the Graeco-Bactrians late in the third century BCE, during the reign of Euthydemus. Ptolemy provides incidental evidence of the identification of Kangju with Sogdia, and once again continued to unknowingly chart the course of the Yuezhi migration by noting a group he this time called the Tachoroi (Tocharians) dwelling in Sogdia88.
Archaeologists have unearthed podboy tombs potentially attributable to the Yuezhi at the periphery of the Bukharan and Samarkand oases, as well as in the upper valley of the Zeravshan. All are similar in design, funeral ceremony and 'accompanying inventory' of artifacts to other possible Yuezhi podboy monuments found in northern Bactria89. In addition, the coroplastic analysis of a group of Sogdian statuettes, might eventually provide supporting evidence of brief Yuezhi occupation.
Both the Han Shu and Shi Ji suggest some level of diplomatic interaction between the Yuezhi and Kangju dynasties prior to the Yuezhi's arrival in northern Bactria90. Although Kangju is described as a 'small' country in the Chinese sources, its military strength was substantial, with some 120,000 persons able to bear arms, according to Ban Gu,91 of which 90,000 were skilled archer warriors according to Sima Qian92. Unlike Dayuan, it was described as powerful enough to be 'not easily defeated by Han forces',93 but none the less was 'constrained to serve the Xiongnu' in the east,94 and acknowledged 'nominal sovereignty' (or was 'subservient to') the Yuezhi in the south'95.
Kangju displayed obvious ability at inter-state diplomacy by balancing these competing (and essentially opposed) interests recognising the nominal sovereignty over parts of its country by both the Xiongnu and the Yuezhi, aiding Zhang Qian and subsequent Han envoys, and later sending a Kangju prince as envoy to China. It is therefore probable that the rulers of Kangju agreed to give the Yuezhi safe passage through their territory, and accepted some form of subservient relationship to avoid military conflict. The Sogdian dynasty may even have encouraged the Yuezhi to settle at the southern extremities of their territory, and suggested the fertile river valleys north of the Amu Darya as an ideal homeland for the migrants.
Eventually Kangju would be drawn further and further into the Yuezhi/Kushan realm over the following decades and centuries until the southern portions of its territory were probably incorporated into the Kushan Empire, or at least into its sphere of influence96. By 83 CE the Kushans would further cement this relationship through an alliance based on a 'bond of royal marriage' with the ruling family of Kangju97. Kangju/Sogdia thus proved to be an ideal buffer for the Yuezhi between their new northern Bactria homeland and the Wusun and Xiongnu realms in the steppes to the north and east. Ultimately, as a result of the generally cooperative relationship apparently established almost immediately between the Kangju and Yuezhi dynasties in c. 131 BCE, Sogdia went on to become an integral and stable member (through direct incorporation or alliance) of the greater Kushan sphere of influence until the dissolution of the Kushan Empire in the mid-third century of the Common Era.
Craig Benjamin
March 2003
Dr. Craig Benjamin is a Lecturer in History at Macquarie University in Sydney, Australia. He can be contacted at: craigbenjamin[@]optusnet.com.au
1 HS 61 1B.
2 SJ 123. See B. Watson, Records of the Grand Historian by Sima Qian Han Dynasty II (Revised edtn., Columbia University Press 1993) p. 231.
3 Ptol. Geog. vi, 14, 7-14. Hermann located Ptolemy's Tagouraioi to the north of the 'Alexandrovski Range', north of Ysyk Kul, in A. Hermann, Sakai col. 1770-1806', in Pauly Wissowa, Real-Encyclopadedie Der Classischen Alterumewissenschaft, 2e Reihe, Vol I (Stuttgart 1914).
4 Zadneprovsky (1999) op. cit., pp 4-5. See also J.P. Mallory and V.H. Mair The Tarim Mummies (London 2000) pp. 156 and 158.
5 HS 61 4B.
6 On the Hanging Pass see HS 96A 12A/12B; HS 96B 1B; H. Tsuchiya, 'Tracing Ancient Routes in Northern Pakistan', in M. Alram and D.E. Klimburg-Salter, Coins, Art and Chronology (Vienna 1999) pp. 353 ff.; A. Stein, Ancient Khotan (Oxford 1907) pp. 1-46.
7 See HS 61 4B.
8 HS 61 4B.
9 HS 61 1A.
10 See A.K. Narain, The Tokharians (Shillong 2000) p. 36, n. 11 for example.
11 Y.A. Zadneprovsky, 'Migration Paths of the Yueh-chih based on Archaeological Evidence', Circle of Inner Asian Art Newsletter No. 8 (April 1999) p. 5.
12 HS 96A 15B.
13 HS 96 A 17A.
14 SJ 123, Watson p. 234.
15 HS 96B 1B.
16 Ibid.
17 A.F.P. Hulsewe and M.A.N. Loewe, China in Central Asia. The Early Stage: 125 B.C. A.D. 23. An Annotated Translation of Chapters 61 and 96 of the History of the Former Han Dynasty (Leiden 1979) p. 125, n. 299, 2.
18 P. Pelliot, 'L'Edition collective des oeuvres de Wang Kouo-wei', in T'oung Pao 26 (1929) p. 151.
19 See Ibid. p. 150.
20 E.G. Pulleyblank, 'The Consonontal System of Old Chinese', in Asia Major 9 (1962) p. 94; see also L. Torday, Mounted Archers: The Beginnings of Central Asian History (Durham 1997) p. 319 n. 34 for a detailed discussion of the philological link between Kangju and the Jaxartes (Syr Darya).
21 Hulsewe and Loewe op. cit., p. 125, n. 299, 2.
22 E.G. Pulleyblank, 'Chinese and Indo-Europeans', Journal of the Royal Asiatic Society 2 (London 1966) p. 28.
23 Narain (2000) op. cit., pp. 29-30.
24 HS 96A 15B.
25 SJ 123, Watson p. 234.
26 HS 96A 15B.
27 SJ 123, Watson p. 234.
28 HS 96A 15B.
29 Ibid.
30 E. Zurcher, 'The Yueh-chih and Kaniska in the Chinese Sources', in A.L. Basham, ed., Papers on the Date of Kanishka (Leiden 1968) p. 360.
31 SJ 123, Watson p. 234.
32 HS 96A 16A.
33 HS 61 3A.
34 HS 96A 15B.
35 SJ 123, Watson p. 234.
36 HS 96A 17A.
37 G.V. Shishkina, 'Ancient Samarkand: Capital of Soghd', Bulletin of the Asia Institute 8 (1994) p. 90.
38 Hulsewe and Loewe, op. cit., pp. 130-1, ns. 320, 321, 322, 323 and 324 provide comments on attempts to identify the five principal towns of the lesser kings.
39 See R. Frye, The History of Ancient Iran (Munich 1984) p. 141.
40 Arrian, Book Four, 5-7 (1971) op. cit., pp. 208-211 for a description of this defeat.
41 See for example, A.B. Bosworth, Conquest and Empire: The Reign of Alexander the Great (Melbourne 1988) pp. 117, and 109-10.
42 See for example A. Nymark, Culture and Art of Ancient Uzbekistan vol.1 (Moscow 1991) p. 155.
43 See for example E. Rtveladze, The Ancient Coins of Central Asia (Tashkent 1987) p. 46.
44 Nymark op. cit., p. 156.
45 See O. Bopearachchi, 'The Euthydemus Imitation and the Date of Sogdian Independence', Silk Roads Art and Archaeology 2 (1991/2) pp. 11-12; also Rtveladze op. cit., p. 46.
46 Ibid., p. 12.
47 Ptol. Geog. vii, 2, 15; vi, 14, 7-14. See W.W. Tarn, The Greeks in Bactria and India (Cambridge 1938, 2nd edtn. 1951) p. 517; and J.W. McCrindle, Ancient India as Described by Ptolemy (London 1885) p. 281
48 HS 96A 15A.
49 SJ 123, Watson p. 234.
50 SJ 123.3b, trans. E. Zurcher (1968) op. cit., p. 360.
51 HS 61 2A.
52 SJ 123, Watson p. 232.
53 HS 61 3A.
54 HS 96A 15B.
55 HS 96A 15B; and Zurcher (1968) op. cit., p.360.
56 Torday op. cit., p. 301.
57 HHS 77.4a.
58 HS 96A 16A.
59 HS 96A 19A.
60 On the Zhaman-Togai Xiongnu tomb see A.G. Maximova, M.S. Merschiev, B.I. Vainberg, L.M. Levina, Drevnosti Chardary (Antiquities of Chardara) (Alma-Ata 1968) pp. 175-190.
61 Torday op. cit., pp. 301-2.
62 On Sarazm see A.I. Isakov, 'Sarazm: An Agricultural Centre of Ancient Sogdiana', Bulletin of the Asia Institute 8 (1994) pp. 1 ff.
63 See V.A. Shishkina, 'Arkheologicheskie nablyudeniya na stroitel' stve Katta-Kurganskogo vodokhranilishcha' (Archaeological Observations on the Building Site of the Katta-Kurgan Reservoir) Izvestiya Uzbfil AN SSSR Bulletin of the Uzbek Branch of the USSR Academy of Sciences (1940) No. 10, pp. 19-24.
64 See A.I. Terenozhkin, 'Sogd I Chach' (Soghd and Chach) KSIIMK 33 (1950) pp. 152-169; A.I. Terenozhkin, 'Roskopki na gorodishche Afrasiabe' (Excavations on the Site of Afrasiab) KSIIMK 36 (1951) pp. 136-140.
65 See V.I. Shishkina, 'Uzbekistanskaya arkheologicheskaya ekspeditsiya AN UzSSR' (The Uzbekistan Archaeological Expedition of the Uzbek Academy of Sciences. Field Work in 1956-1959), IMKUz (Tashkent 1961) pp. 36-43; V.I. Shishkina, Afrasiab sokrovishchnitsa drevnei kultury (Afrasiab Treasury of Ancient Culture) (Tashkent 1966) pp. 8-10.
66 See A.Y. Yakubovsky, 'Itogi rabot Sogdiiko-Tadzhikskoi arkheologicheskoi ekspeditsii v 1946-1947 gg' (Results of the Work of the Sogdian-Tajik Archaeological Expedition of 1946-1947) MIA SSR No. 15 (Moscow-Leningrad 1950) pp. 13-28.
67 See for example, A.M.Mandleshtam, 'Mogilnik v s. Zosun verkhovya r. Zeravshan (The Burial Site in Zosun Village, Upper Zeravshan) Izvestia Otdelniya Obshchestvennykh nauk AN Tadzh. SSR No. 40 (1965) pp. 29-44.
68 A.M. Mandleshtam, 'Archaeological Data on the Origin and Early History of the Kushans', in Central Asia in the Kushan Period (1970) op. cit., p. 166.
69 See for example O.V. Obelchenko, 'Kuyu-Mazarsky mogil'nik' (The Necropolis of Kuyu-Mazar) Trudy IIA AN Uz. SSR VIII (Tashkent 1956) pp. 205-227; and O.V. Obelchenko, 'Kurgany okolo sel. Khazara' (The Kurgans Near the Village of Khazara) IMKUz (Tashkent 1963) pp. 57-65.
70 N.G. Gorbunova, 'Early Nomadic Pastoral Tribes in Soviet Central Asia during the First Half of the First Millennium A.D.', in G. Seamen, ed., Foundations of Empire: Archaeology and Art of the Eurasian Steppes (Los Angeles 1992) p. 34; and see O.V. Obelchenko, 'Kurgannye mogilniki epokhii Kushan v Bukharskom oazise' (Kurgan Barrows of the Kushan Epoch in the Bukhara Oasis), in Central Asia in the Kushan Period, vol I (Moscow 1974) for a summary of Obelchenko's conclusions.
71 Mandleshtam (1970) op. cit., p. 166.
72 Y.A. Zadneprovsky, History of Central Asian Nomads in the Kushan Period', in Kushan Studies in the U.S.S.R. (Moscow 1970) pp. 148-9.
73 Zadneprovsky (1999) op. cit., p. 5.
74 See Gorbunova (1992) op. cit., pp. 36 ff for a general outline of material artifacts discovered at a range of Central Asian sites.
75 Ibid., p. 36.
76 Ibid., p. 39; and see G. Frumkin, Archaeology in Soviet Central Asia (Leiden 1970) pp. 42 ff, for a detailed summary of Ferghana Valley archaeological discoveries, including grave artifacts.
77 See R. Ghirshman, Begram Recherches Archeologiques et Historiques sur les Kouchans (Paris 1946), Plate XXXVI, Nos. BG 290 a and b for excellent photographs of the Begram arrowheads with barbed triangular flanges. At the British Museum in January 2001 Elizabeth Errington was good enough to show me a colour transparency of a similar arrowhead from the Charles Masson collection. See also E. Errington, 'Rediscovering the Collections of Charles Masson', in M. Alram and D.E. Klimburg-Salter, eds., Coins, Art and Chronology. Essays on the pre-Islamic History of the Indo-Iranian Borderlands (Vienna 1999) pp. 207ff.
78 See Gorbunova (1992) op. cit. p. 44, Fig. 5 for a map illustrating the 'Distribution of Mirror Types'.
79 Ibid., p. 41.
80 Shishkina op. cit., pp. 89-90.
81 Gorbunova op. cit., p. 42.
82 Fiona Kidd, 'The Chronology and Style of a Group of Sogdian Statuettes', in C. Benjamin and S. Lieu, eds., Walls and Frontiers in Inner Asian History, Silk Roads Studies VI (Turnhout 2002) pp. 197 ff.
83 P. Ucko, Anthropomorphic Figurines of Predynastic Egypt and Neolithic Crete with Comparative Material from the Prehistoric Near East and Mainland Greece (London 1968). See also M. Gimbutas, The Goddesses and Gods of Old Europe (London 1982) for early examples of similar figurines.
84 Kidd op. cit., pp 197 ff.
85 K. Abdullaev, in R.K. Suleimanov, Drevnii Nakshab: Problemi' Tzivilizatzii Uzbekistana VII v. do N.E. (Tashkent 2000) p. 203.
86 On the Tillya-tepe figurines see V.I. Sarianidi, Khram I Nekropol' Tillyatepe (1989) p. 53 fig. 15 and p. 57 figs. 17, 25 and 70; on the Dalverzin-tepe figurines see K. Abdulaev, 'Portrayal of musicians in Bactrian teracotta figurines', in Information Bulletin 7 (1984) pp. 52-56 figs. 4-5.
87 Kidd op. cit., p. 211.
88 Ptol. Geog. vii, 2, 15; vi, 14, 7-14. See Tarn (1938/1951) op. cit, p. 517; and McCrindle (1885) op. cit., p. 281.
89 Zadneprovsky (1999) op. cit.,p. 5.
90 HS 96A 15B; SJ 123, Watson p. 232.
91 HS 96A 15B.
92 SJ 123, Watson p. 234.
93 HS 61 3A.
94 HS 96A 15B.
95 HS 96A 15B, and Zurcher op. cit.,p. 360.
96 Whether Sogdia/Kangju or Khorezmia were ever actually part of the Kushan Empire is a matter of some conjecture. See, for example, Svend Helms, 'Ancient Chorasmia: The Northern Edge of Central Asia from the 6th Century B.C. to the mid-4th Century A.D., in David Christian and Craig Benjamin (eds.), Worlds of the Silk Roads: Ancient and Modern (Turnhout 1998), pp. 85-6; and S. Helms (in private correspondence) who has argued that Khorezmia at least remained nominally independent.
97 HHS 77.4a.
The Yuezhi Migration and Sogdia
Craig Benjamin
Introduction
Following the defeat of the formerly powerful Yuezhi confederation by the Xiongnu near Dunhuang in c.162 BCE, the Yuezhi dynasty and those tribes that remained loyal to it commenced a migration away from the Gansu that was destined to completely reshape the geopolitics of ancient Inner Asia. Both the Han Shu and Shi Ji provide evidence of their departure: "the Yuezhi had fled furious with the Xiongnu"1, the 'Yuezhi had fled and bore a constant grudge against the Xiongnu'2. The decision to migrate, despite still having a force of perhaps 100,000 armed archer warriors at their disposal is indicative of the severity of the defeat, and also of the steadily increasing power of the Xiongnu under Maodun and Jizhu during the preceding decades. The Yuezhi dynasty may in fact have considered such a move several times during the fourteen years between Maodun's initial raid against them in 176, and their ultimate defeat in 162.
Indeed, the fact that the migration seems to have been conducted in an orderly fashion suggests something of a planned strategic relocation rather than a rout. The Yuezhi's original intention was to move some 2000 kilometres to the northwest and resettle in the valley of the Ili River, a region occupied by a group of Sakas (or Scythians). They had no intention, nor any idea, that this journey would only be the first stage of a migration that ultimately would take them half away across Central Asia, until thirty years later they would find themselves in secure occupation of the fertile river valleys north of the Amu Darya, and masters of the former Greek kingdom of Bactria.
Leaving the Gansu in 162 the Yuezhi headed northwest towards the Ili Valley, settling near Ysyk Kul in present-day Kazakhstan. Corroborative evidence for this new location is provided by the Greek geographer Ptolemy who mentions a group called the Tagouraioi (clearly a variation on Tocharian, the Indo-European language spoken by the core Yuezhi) dwelling near Ysyk Kul3. Russian archaeologist Yu Zadneprovsky has noted a substantial number of podboy sites in the region, which he has tentatively identified as Yuezhi on the basis of their similarity to other podboy tombs discovered at the Haladun site near Minqin in the Central Gansu, which he also argues are Yuezhi. The Ysyk Kul region is rich in nomadic burial sites and some 370 tombs had been noted by as early as 1960. Of these, 80% were in pits, and attributed to the autonomous Sakas, and 17% in podboys, tentatively attributable to the Yuezhi4.
The Chinese sources show that the Ili/Ysyk Kul region was already populated by the Sai people, an eastern concentration of Sakas or Scythians who probably spoke an Indo-Iranian language. Upon arriving at the Ili, the Yuezhi quickly displaced them: (The Yuezhi) 'attacked the king of the Sai (who) moved a considerable distance to the south and the Yuezhi then occupied his lands'5. Most of the displaced Sakas subsequently undertook their own substantial migration, moving west and then south through the western Tarim Basin, crossing the so-called 'Suspended Crossing' (probably the Khunjerab Pass or similar, between present-day Xinjiang and northern Pakistan) before ultimately settling in Kashmir6.
The Yuezhi confederation occupied the former Sakan lands in the hope of permanently resettling there, and remained in residence for almost three decades. They no doubt felt they had successfully relocated, having escaped the Xiongnu menace and reestablished themselves in the fertile Ili Valley. They returned to their previous semi-nomadic, semi-sedentary lifeway and probably began to lose interest in Chinese/nomadic politics. But the Kunmo of the Wusun, the former neighbours of the Yuezhi in the Gansu, could not forget the ill treatment his people had suffered as a result of a Yuezhi attack upon them in 1737. Eventually he sought permission from his Xiongnu overlord (the new ShanyuJunchen, successor to Jizhu who had died in 158) to pursue the Yuezhi into the Ili and 'avenge his father's wrongs'8. In 132 BCE the Kunmo led a powerful force of mounted Wusun archers into the region which attacked and routed the no doubt surprised and dismayed Yuezhi, forcing them to once again uproot and resume their long march to the west.
The sources indicate that within a short space of time the Yuezhi passed through a region called Dayuan 'The Yuezhi thereupon went far away, passing Dayuan and proceeding west'9 and then through a land to the southwest called Kangju. Although the exact route remains a matter of some conjecture, the evidence of both the Chinese annals, and of Russian and Central Asian archaeology, leaves little doubt that the Dayuan through which the migrating horde passed can only be identified with the Ferghana Valley10. The Yuezhi apparently met with little resistance from the urbanised Dayuans/Ferghanese, and after perhaps spending some months (the winter of 132/1?) in southwestern Ferghana, they passed on unmolested. Zadneprovsky has also noted several single podboy burials that have been unearthed in the southwestern, northern and eastern parts of the Ferghana Valley, most concentrated in the Lyailyaka-Isfara-Sokha interfluve in southern Kyrgyzstan where over 300 podboy burials have been located. Although originally attributed to a separate culture by Baruzdin in 1960, Zadneprovsky argues for their re-attribution to the migrating Yuezhi, on the basis of their similarity to podboy sites also tentatively attributable to the Yuezhi in both the Gansu and Ysyk Kul region11.
Perhaps in the spring of 131 BCE then, the Yuezhi most probably moved from Ferghana into the 'state' of Kangju, probably the Zeravshan Valley in the heart of Sogdia. Some four or five years later they were followed through the region by Han envoy Zhang Qian, who was led there by guides and interpreters provided for him by the king of Dayuan. It is references to Kangju in the Han Shu and Shi Ji (and by Ptolemy in his Geographica), as well as the discoveries of Soviet and Russian archaeologists, that has provided evidence identifying Kangju with Sogdia, and thus of the role of Sogdia in both the migration of the Yuezhi and the mission of Zhang Qian. The intention of this paper is to consider the origins of the relationship that developed between the Kangju and Yuezhi dynasties, a relationship that subsequently evolved to provide vital political and military stability in the region throughout the Kushan Era. The initial task is to consider evidence that allows for the conclusive identification of Kangju as Sogdia.
References to Kangju and the Yuezhi in the Han Annals
Kangju Sogdia: Location and Lifeway
Location
'The seat of the king's government in winter is in Leyuenidi to the town of Beitian. It is distant by 12,300 li from Xian. One reaches (Le)yuenidi after a journey of seven days on horseback, and it is a distance of 9,104 li within the realm to the king's summer residence. To the east it is a distance of 5,500 li to the Seat of the Protector General'12.
'It is said: Some 2000 li to the northwest from Kangju is the state of Yancai. The trained bowmen number 100,000. It has the same way of life as Kangju. It is situated on the Great Marsh, which has no further shore and which is presumably the Northern Sea'13.
'Kangju is situated some 2,000 li northwest of Dayuan. The country is small and borders Dayuan. It acknowledges nominal sovereignty to the Yuezhi people in the south and the Xiongnu in the east'14
'(Wusun) adjoins Kangju in the northwest'15
'(The State of Wusun) and 5000 li to the west, to land within the realm of Kangju'16
Attempts by scholars over several centuries to geographically locate and delineate Kangju have not been helped by textual corruption in both the Han Shu and Shi Ji. And yet, although several words and even whole sentences are missing, the information provided is still in the same order as that for the other 'western states', so that any gaps cannot be substantial. Certainly the distances between Xian and Beitian are not quite reconciled, and the distance from Beitian and the king's summer capital (9104 li or 3641 kms) is surely corrupt. Hulsewe and Loewe suggest that the text may originally have read 'ninety one li' (or 36 kilometres, although this seems too low),17 while Pelliot noted Wang Kuowei's suggestion of 1104 li (441 kms)18 which is a more viable figure within a country described as 'small'.
The identities of both Beitian and (Le)yueni(di) are almost impossible to determine, however, Wang Kuowei identified the former (impossibly) with Ysyk Kul19 while Pulleyblank has argued that the latter might 'represent some form of the name Jaxartes'20. The distance between Beitian and (Le)yueni(di) is described as 'seven days on horseback' in the Han Shu, which Hulsewe and Loewe suggest equals about 500 li i.e. marches of seventy lior 28 kilometres per day through the mountainous country of the region21. The identification of these two principal settlements with Samarkand and Bukhara is one obvious possibility, although the distance between the two cities by road is about 200 or so kilometres which does not reconcile with any of the given statistics.
Pulleyblank discusses the possible Tokharian philological origin of the name 'Kangju', in his reconstruction of 'Old Chinese' *khan-kiah. In the Tokharian vocabulary (Tokharian 1A) there is the word kank, which means 'stone'. Thus Kangju could mean the 'Stone Country', i.e. Samarkand (or equally Tashkend as 'Stone City')22. A.K. Narain offers a precise geographical location for Kangju: 'the northeastern wedge of modern Uzbekistan into Kirghiziya and Kazakhstan; the eastern part of this wedge formed part of Dayuan'23. This description, however, does not allow for the inclusion of any lands south of the Syr Darya, thus excluding the entire Zeravshan Valley, the cultural heart and population centre of Sogdia.
The information provided by the texts is hardly ambiguous, however, and clearly suggests the identification of the 'state' of Kangju with ancient Sogdia. Kangju is to the north of the Amu Darya and the Yuezhi's principal city of Jianshi (Khalchayan in the Surkhan Darya valley?); to the west and northwest of the Ferghana Valley (where it also apparently adjoined the clearly very substantial, post-132 realm of the Wusun); and southeast of the western realms of the Xiongnu (which must therefore have included the steppes of present-day Kazakhstan). Kangju incorporated lands on either side of the middle Syr Darya, particularly the densely occupied Zeravshan Valley south of the Syr Darya, and must surely have included Samarkand and Bukhara (as Shishkina also argues below). Hence, according to the textual evidence at least, Kangju can only convincingly be located within the general geographical region of ancient Sogdia.
Population/Size
Households: 120,000 Individuals: 600,00024
'The country is small'25
The physical dimensions of the Kangju realm may not have been vast, but the population was substantial, which allowed the ruling dynasty to maintain a formidable military force.
Military Strength
Persons able to bear arms: 120,00026
'They have 80,000 or 90,000 skilled archer warriors'27
'(it) is not subject to the Protector General'28
'In the east (the inhabitants) were constrained to serve the Xiongnu'29
'It acknowledges nominal sovereignty to (Zurcher translates as 'it is subservient to')30 the Yuezhi people in the south and the Xiongnu in the east'31
'However, Kangju felt that it was separated (from Han) by a long distance, and alone in its arrogance it was not willing to be considered on the same terms as the various other states'32
'(He Wudi - heard that) to the north, there were (people or places) such as the Da Yuezhi and Kangju, whose forces were strong; it would be possible to present them with gifts and hold out advantages with which to bring them to court'33
The Chinese were clearly impressed by the strength of Kangju, finding them arrogant and militarily self-confident. The military resources of Kangju (120,000 armed men, 80,000 - 90,000 of which were skilled and presumably mounted archers) were substantial, and would not easily be defeated by the Han. Presumably the ruling Kangju dynasty and its pastoralist allies provided the bulk of the mounted archer warriors, while the sedentised agriculturists of the river valleys could be relied upon to provide the remainder. Eschewing any military option then, Zhang Qian argued instead (in his report to Wudi) that the Kangjuans could be persuaded by Han gifts and favours to consider becoming subjects (or at least allies) of the Chinese. In short, Kangju was powerful and remote enough to resist Han attempts to join their tributary confederacy by military means, but was clearly under some sort of sovereignty obligation to both the Yuezhi and the Xiongnu.
Environment/Lifeways
'The way of life is identical with that of the Da Yuezhi'34
'Its people likewise are nomads and resemble the Yuezhi in their customs'35
'In Kangju there are five lesser kings all the five kings are subject to Kangju'36
The last reference clearly indicates that 'Kangju' should be considered both as the Han name of the 'state' (that is the realm or region) of Kangju/Sogdia, but also of the dominant faction or dynasty which was controlling that realm at the time (i.e. the Kangju dynasty). Shishkina agrees, and argues that geo-political changes in Sogdia that became apparent towards the end of the second century BCE must have been as a result of Kangju hegemony:
'The historical situation of the first century BC suggests that these changes were related to the spread of the power of the Kangju, when this dynasty controlled Samarkand and Bukhara'37
The five lesser kings noted in the Han Shu were probably subordinate tribal groups within the realm of 'greater Kangju', and given that all are listed as having specific 'seats of government' (different to the two principal settlements named as belonging to the Kangju proper), may represent sedentised, agrarian-based 'peoples' living under Kangju hegemony38
The way of life of the dominant Kangju faction was probably that of semi-nomadic militarised pastoral nomadism, similar to the assessment of the lifeway of the Yuezhi soon after their arrival north of the Amu Darya that Zhang Qian provided to the Han court. If the Kangju state is thus to be identified with ancient Sogdia under Kangju dynastic hegemony, then a brief account of Sogdian history prior to the arrival of the Yuezhi is required to identify the probable date of the establishment of Kangju power, and also to clarify the archaeological and textual evidence of subsequent Yuezhi/Kangju interaction.
Sogdian Historical Framework Prior to the Arrival of the Yuezhi
Between 553 and 550 BCE, Cyrus II (r. 559-529), a leader of the Persian Achaemenid family, overthrew Astyages, King of the Medes, and brought Mesopotamia, Parthia and Anatolia under his control. By 539 he had conquered Bactria and much of Sogdia as well, where he established a line of fortresses on the Syr Darya. Sogdia was made the thirteenth satrapy of the Achaemenids, and paid tribute to Cyrus' successors. The oldest layers of Afrasiab, the ancient site of Samarkand, date from this Achaemenid period. But, whilst the city states of Sogdia and Bactria gained considerably through their incorporation into the Achaemenid Empire, they remained intent upon regaining their independence, which parts of Sogdia may have done by c. 400 BCE39
Some two centuries after Cyrus' death, Alexander of Macedon reconquered much of Central Asia, following his arrival in Bactria in 329 BCE. Alexander's principal opponent in the region was the Achaemenid ruler Darius' former satrap, Bessus, who had Darius murdered in modern-day Shahr-I Qumis before proclaiming himself as his successor. Bessus' troops consisted of armoured cavalry from Bactria and Sogdia, which, following their defeat at Gaugemela, he took back across the Amu Darya after destroying its bridges. Alexander led his troops on forced marches through the desert, crossed the Amu Darya on inflated hide rafts, and confronted his opponents who immediately sued for peace.
Bessus was executed; the Macedonians installed themselves in the satrapal palace at Maracanda (Samarkand) and Sogdia, following some seventy years of independence, found itself incorporated into the new Macedonian Empire. But while Alexander campaigned further north along the Syr Darya, the Sogdians, under the leadership of Spitamenes, rose in his rear and massacred a garrison of Macedonians, inflicting arguably the worst defeat of Alexander's career40. Over the course of the ensuing eighteen months Alexander gained his revenge by reducing the fortified towns of Sogdia one by one, starting in the Hissar Mountains and moving along the Zaravshan Valley41.
At the heart of ancient Soghd were the valleys of the Zeravshan and Kashkadarya, and in his vengeful campaign along these densely occupied valleys the Macedonians may have killed up to 120,000 Sogdians42.
Amongst the many prisoners captured during the Sogdian campaign was the Princess Roxanne, daughter of another Sogdian opponent, Oxyartes. Alexander's subsequent decision to marry Roxanne was due partly to her exceptional beauty, but was also intended as a gesture to appease the rebellious Sogdians. After Alexander's death in Babylon in 323, Bactria and Sogdia immediately rebelled but were reconquered in c. 305 by his successor Seleucus Nicator (r. 311-281 BCE). However under Seleucus' son Antiochus I (r. 281-261), Bactria and (probably) Sogdia broke away again from Seleucid hegemony. None the less, Sogdia, along with much of Central Asia, was brought into the orbit of Hellenistic influence during its brief period of Macedonian conquest.
Antiochus I minted an extensive local coinage in the region, probably at Balkh (the 'capital' of Bactria). These were coins of large denominations: staters, tetradrachms and drachms. During the last two centuries before the Common Era, several series of diverse denominations and types were struck at Sogdian mints, and coins were widely used in Sogdia and Bactria, although perhaps only by the Greek population43
None the less the native population of Sogdia became used to Greek coinage during the Seleucid period, and when the inflow of Greek coins stopped following their independence from Antiochus I, local rulers began to mint their own. As Nymark has pointed out, however, these local issues were highly debased, and in fact were 'mere imitations of the most widespread Greek coins'44.
Yet these imitations remain as crucial (and often the only) evidence of political, economic and social developments in Sogdia during the first century BCE. Furthermore, both Sogdian and Bactrian imitation issues also constitute potential evidence for the Yuezhi during the 'five-yabghu period'.
In the mid-third century (c. 250) the Seleucid Governor of Bactria, Diodotus, established an independent Graeco-Bactrian kingdom, which may also have exercised a degree of control over Sogdia. In c. 230 Diodotus' son was overthrown by one of his satraps, a Greek settler called Euthydemus, who then ruled Graeco-Bactria for about forty years until c. 190 BCE. If Sogdia was indeed part of the incipient Graeco-Bactrian state, then the evidence of the Euthydemus imitation coinage indicates that some time late in the third century, during the lifetime of Euthydemus, Sogdia became an independent entity once more45.
Euthydemus concluded a peace treaty with the Seleucid ruler Antiochus III in 206, but did not attempt to reconquer Sogdia. Instead the Graeco-Bactrians expanded south into India, establishing the Indo-Greek kingdoms. If the 'state' of Kangju is indeed to be identified as Sogdia, then it was during this period of post-Seleucid independence, i.e. from c. 210 BCE, that the region came under the hegemony of the Kangju dynasty, which then continued to rule an independent Sogdia until it came under Kushan political influence early in the first century CE. As Bopearachchi concludes, under the Kangju dynasty 'Sogdiana probably remained free at least until the arrival of the Yuezhi in c. 130 BC'46.
The Passage of the Yuezhi through Kangju/Sogdia
Although neither of the Chinese sources categorically states that the Yuezhi horde passed through Kangju, the only logical inference to be drawn from the texts is that they did. In addition, Ptolemy continued to unknowingly chart the course of the Yuezhi migration by noting a group he this time called the Tachoroi (surely another variant of Tocharian) dwelling in Sogdia47.
The conclusion that the Yuezhi must have passed through the region is further strengthened by the fact that the Han sources do unequivocally show that Zhang Qian passed through Kangju during his search for the Yuezhi. The Han envoy was obviously well informed by the rulers of Dayuan as to the route followed by the Yuezhi who provided him with guides to lead him to the Yuezhi, unless they knew the migrants' route and probable whereabouts? And thus is likely to have followed closely in the original footsteps of his quarry. The chronology is straightforward enough.
C. 132/1 BCE: The Yuezhi depart Dayuan and continue their migration to the west
That the Yuezhi continued westwards in their migration following their passage through (and possibly winter residency in) Dayuan is implicit in the key Han Shu passage: 'passing Dayuan (and) proceeding west to subjugate Daxia'48.
There are three feasible route options west from Ferghana, whether starting from present-day Kukon in the centre of the valley, or Isfara in the southwest. The first is due north and then west, across the 2267 metre Kamcik Pass and through Angren into Tashkent, thence southwest to Samarkand. A second and more direct route is due west through present-day Chugand and Zizzach, thence southwest into Samarkand. However, given that the Zeravshan Valley was the agricultural and population heartland of Sogdia/Kangju (information probably given to the Yuezhi by the rulers of Dayuan who were no doubt anxious to encourage the Yuezhi to move on and seek suitable settlement lands elsewhere), the migrating horde may have chosen to follow a third route option from Chugand, south over the 3378 metre Sahristan Pass, then down into the upper Zeravshan Valley. If the Yuezhi leadership decided upon this latter route, then it would indeed have been necessary for them to winter in southern Ferghana before attempting the crossing of this high pass in the spring.
The Shi Ji also implies this in noting (in Watson's translation) that the Yuezhi 'moved far to the west, beyond Dayuan'49, which Zurcher reads as: 'They passed through Dayuan and to the west of thatcountry'50. Between Dayuan and Daxia (Bactria) lay only Kangju/Sogdia; anyone moving to the west, beyond Dayuan (or to the west of that country) and heading for northern Bactria would have to have passed through Sogdia. This probability is then strengthened by the unambiguous statement that Zhang Qian was taken to Kangju by his Dayuan guides and interpreters, and from there proceeded directly to the realm of the Da Yuezhi in northern Bactria.
129/128 BCE: Zhang Qian also passes through Kangju
The Han Shu notes that:
'(the king of Dayuan) sent off (Zhang) Qian, providing him with interpreters and guides. He reached Kangju who passed him on to the Da Yuezhi'51.
Or as Sima Qian puts it:
'The king of Dayuan trusted his words and sent him on his way, giving him guides and interpreters to take him to the state of Kangju. From there he was able to make his way to the land of the Great Yuezhi'52.
Despite its obvious military strength, Kangju (like Dayuan) also apparently facilitated (or at least did not hinder) the passage through its territory of both the migrating Yuezhi horde in c. 131 and the Han envoy in c. 128 BCE. Given the size of its military resources, Kangju was powerful enough to be not 'easily defeated by Han forces',53 although it was 'constrained to serve the Xiongnu' in the east,54 and (later) would acknowledge 'nominal sovereignty' (or even become 'subservient to') the Yuezhi in the south'55.
Does this acknowledgement suggest that parts of Sogdia (and the most populous parts at the Zeravshan valley and Samarkand) were actually invaded and defeated by the migrating Yuezhi, and then forced into a subordinate relationship thereafter? Certainly Torday is prepared to argue that not only did the Yuezhi defeat the Kangju dynasty in Sogdia, but in northern Bactria as well where he suggests the Kangju were also ruling:
'We must accept that they took the region from the Kangju by force, in a war. His (the Yuezhi king's) subsequent conquest of Sogdiana was probably a by-product of his ambitions further south-east where he had brought Daxia under his sway'56.
If Torday is correct, this says much about the military capabilities of the migrants that a displaced, previously fragmented and essentially homeless nomadic tribal confederation, soundly defeated twice during the previous three decades by the Xiongnu and the Wusun, was none the less able to invade and defeat the well-defended state of independent Kangju Sogdia. But there is simply no evidence to support Torday's reconstruction, and indeed his argument is unconvincing. He has the Yuezhi settling in the Jaxartes (Syr Darya) delta rather than northern Bactria at the end of their migration, which is clearly incorrect as textual and archaeological evidence indicates. He then suggests that the Yuezhi defeated the Kangju dynasty to gain control of Bactria, where the evidence shows it was a group of Sakas (who had established hegemony over those regions of Bactria to both the north and south of the Amu Darya) that the Yuezhi were forced to defeat and evict to gain control of the region.
Ultimately there is no suggestion (other than the mention of 'nominal sovereignty') that the Yuezhi were ever forced to confront the Kangju militarily. If there had been any need for conflict, then given the subsequent history of the region (including the invasion of Sogdia by the Karakhanid Dynasty in the 11th Century, for example), the acceptance of Yuezhi suzerainty by Kangju should more accurately be seen as an example of an all-too-familiar phenomenon in Central Asian history, and the greater military power of pastoral nomadic states over semi-sedentary, irrigation-based city states, whatever the relative size of their forces. Despite the fact that the ruling dynasty of Kangju was probably of a militarised, semi-nomadic ancestry, the bulk of the population was surely sedentised agriculturists.
The evidence might alternatively be interpreted as suggesting that Kangju (both the state and the dynasty) adopted a conciliatory and diplomatic position towards most of its neighbours, even including the migrating Yuezhi horde. Kangju acknowledged nominal sovereignty (or was subservient) to both the Xiongnu and the Yuezhi; they also assisted the Chinese by helping (or at least not impeding) Zhang Qian in his attempt to locate the Yuezhi and subsequently even sent a 'royal' son as envoy to the Han Court during the reign of Emperor Zheng. Undoubtedly, like the Wusun, Kangju felt that the Han court was remote and refused to become subject to the Protector General, but the dynasty was careful to offend no one Han, Xiongnu or Yuezhi. It is therefore more likely that the Kangju rulers of Sogdia agreed to allow the Yuezhi unmolested passage through their territory, and accepted some form of subservient relationship thereafter to avoid military conflict.
It is not at all necessary to envisage violent military confrontation between the two dynasties, but rather a situation where local rulers recognised that the Yuezhi were powerful enough to be worth accepting as symbolic overlords. Certainly the later Mongol invasions would show that where cities made peace with the migrating, invading force without offering resistance, they could survive and prosper under the rule of pastoralist conquerors, whereas those who refused were destroyed utterly. The apparent ease with which the Yuezhi subjugated Bactria (just to the south of and contiguous to Sogdia) a year or so later may also have been a salutary reminder to the Kangju dynasty. And, just as the Dayuans might have suggested the Zeravshan Valley in Sogdia as a possible resettlement location for the migrants, the rulers of Sogdia could in turn have nominated the fertile valleys north of the Amu Darya in northern Bactria for the Yuezhi to relocate, thus discouraging any thoughts the latter might have entertained of remaining in Sogdia itself.
The careful relationship established between the two dynasties in c.131 BCE became an enduring one. Kangju Sogdia would be drawn further and further into the Yuezhi/Kushan sphere of influence over the following centuries until substantial portions of its territory may have been incorporated into the Kushan Empire (although this is by no means certain). Certainly, according to the Hou Han Shu, by 83 CE the Kushans were happy to further cement this relationship through an alliance based on a 'bond of royal marriage' with the ruling family of Kangju57. Even as early as the last three decades of the second century BCE Kangju proved to be an ideal buffer for the Yuezhi between their new homeland north of the Amu Darya, and the Wusun and Xiongnu to the north and east.
The Kangjuans (or at least some elements of the Sogdian populace) also apparently acknowledged sovereignty to the Xiongnu in the east (presumably northeast),58 indicating the astonishing reach of Xiongnu influence. This is confirmed by the Han Shu inclusion of a statement in Zhang Qian's description of the 'state' of Dayuan that '(the area) west of Wusun as far as Anxi is close to the Xiongnu'59. That is, Xiongnu influence apparently reached from the steppes of eastern Mongolia to as far west as Parthia. Even allowing for possible Han exaggeration or misinformation, this means that the Xiongnu were in seasonal occupation of (or at least exercised some form of nominal hegemony over) a very substantial western realm indeed, including the steppelands of present-day Kazakhstan and Uzbekistan, from Lake Balkash to the Aral Sea.
It is also possible that this 'nominal sovereignty' took the ambiguous form of gift-giving, which the Xiongnu and Chinese sources interpreted as subordination, but which the locals (including the ruling dynasties of Dayuan and Kangju) understood merely as diplomatic courtesies. There is little or no evidence that the Xiongnu actively exercised military power further to the west, although they may have occupied temporary sites along the Middle Syr Darya, if the archaeological evidence of a burial mound at Zhaman-Togai attributed to the Xiongnu by the tomb's Soviet discoverers in 1968 is accepted60. Torday is prepared to read this acknowledgment of nominal sovereignty as proof of Xiongnu military subjugation of Kangju, and even attempts to date the beginning of that subjugation from early in the reign of Maodun. He finds the Han Shu passage evidence of 'Kangju's humiliation by Xiongnu in the east', and links this humiliating defeat to a steppe battle conducted by Maodun at Zaysan Nor in 203 BCE61. Torday's reconstruction is imaginative but there is simply no evidence to support it. In the hope of corroborating the textual suggestions of Yuezhi/Kangju interaction, one must turn to a consideration of archaeological evidence.
Archaeological Evidence of the Yuezhi in Kangju/Sogdia
Introduction and History
The textual references thus strongly imply that during the latter stages of their migration the Yuezhi horde probably passed through some regions of the 'state' of Kangju Sogdia. However, as is the case with most aspects of Yuezhi history, archaeological confirmation of this apparent course of textually-attested events is hardly overwhelming. The most likely location for the discovery of material evidence would surely be somewhere in the valley of the Zeravshan River, which has its source between the Turkestan and Zeravshan ranges in Tajikistan, and then flows west through Samarkand and out into the deserts of Uzbekistan. In ancient times the valley was fertile and sheltered, providing excellent conditions for early Bronze Age agrarian settlements. The discovery in 1976 of the Bronze Age settlement of Sarazm on the upper reaches of the river showed that the valley had been settled for thousands of years, perhaps since as early as the mid-4th Millennium BCE62.
Archaeological investigation of the middle and lower Zeravshan Valley commenced in 1940 with a survey carried out on the construction site of the Katta-Kurgan reservoir, halfway between Samarkand and Bukhara, which resulted in the discovery of Sogdian burial mounds of the 'Kushan period'63. Between 1945 and 1949 Terenozhkin conducted systematic research in the immense town site of ancient Samarkand, Afrasiab64. This was followed by expeditions from the Uzbek Branch of the Soviet Academy of Sciences in the late 50s and early 60s, which thoroughly investigated a series of sites around Samarkand65. In 1946, meanwhile, the Academy of Sciences of the Tajik Soviet Socialist Republic began investigating the upper reaches of the Zeravshan, east of Samarkand, which, under the direction of Yakubovsky, laid the ground work for a systematic archaeological survey of the entire length of the upper regions of the valley66. Yakubovsky's preliminary work was carried on through the 1960s by Mandleshtam in particular, who discovered numerous burial mounds and settlements, many of them apparently left by migrating nomadic groups during the early Kushan era (as Soviet archaeologists tended to label the period between the second century BCE and the late second/early third century CE)67. However, within this broad and general chronology the attribution of particular types of burial structures and fortified settlements to specific groups of pastoral nomads was again very difficult. Mandleshtam was the first to attempt to narrow the search down somewhat in his address to the 1968 conference on the Kushans in Dushanbe:
'The dating of the investigated monuments (which is substantiated by the results of recent excavations of town-sites) enables us to state with certainty that they belonged to the nomads, who in the last third of the second century BC destroyed the Graeco-Bactrian Empire. Archaeology reveals four groups of nomads, which probably relate to four different tribes'68.
Obelchenko, however, who from 1952 onwards carried out large-scale excavations of nomadic burial mounds in the middle and lower Zeravshan, particularly around Bukhara, linked all of the funerary monuments discovered in the Zeravshan to an invasion of 'Sarmatian' tribes in the second to first centuries BCE69. Gorbunova has summarised Obelchenko's mid-50s theory thus:
'The discovery and investigation of Sogdian pastoralists' sites were begun by O.V. Obelchenko. He isolated among the cemeteries an earlier group dating from the second century BC to the first century AD, and a later group dating from the second to the fourth centuries AD. He regards both sites as relics of the Sarmatian tribes whose attacks contributed to the final defeat of Graeco-Bactria'70.
Obelchenko's conclusions were questioned by Mandleshtam, who argued that the attribution of all of the mounds as Sarmatian was too generalised and superficial, particularly as various groups of burials displayed quite distinctively different characteristics.
'The culture of all of them (the burial mounds) exhibits traits resembling the culture of the Sarmat tribes, but this is mainly a 'temporal' resemblance, which is observed over a vast area. A more concrete comparison points to links with areas to the N and NE of Central Asia'71.
Zadneprovsky also disagrees with Obolchenko,72 and has argued that similarities between possible Yuezhi tombs in northern Bactria and those discovered in the Bukharan oasis by Obelchenko suggest that the latter are also relics of the passage of the Yuezhi through Sogdia: 'The coincidence of the design, funeral ceremony and the accompanying inventory definitely point to their similarity'73.
Artifacts from Nomadic Tombs in Ferghana and Sogdia
However, even a detailed analysis of the 'accompanying inventory' of grave goods discovered at the various nomadic tomb sites does not clarify the matter74. Pottery vessels (including censers) occur in all types of graves, as would be expected. Gorbunova argues that 'they have a strictly local character, originating in the pottery centres nearest to the cattle-breeding people's area of habitation, or being diffused only over the local area'75. Thus jugs (with and without handles) are common to the whole region, although they vary in shape from one location to the next. Bactrian and Sogdian jugs and goblets were almost all fashioned on a potter's wheel, which differentiates them from Khorezmian ceramics, for example.
Weapons have also been discovered in most nomadic cemeteries throughout former Soviet Central Asia, including swords, daggers, iron arrowheads and fragments from composite bows, all weapons generally associated with militarised nomads. Swords and daggers with a cross hilt are characteristic of both the Sogdian and Bactrian sites, whereas the majority of swords from other areas (i.e. areas probably not visited by the migrating Yuezhi) do not have cross-hilts, thus allowing for possible identification of the unique Sogdian and Bactrian examples as Yuezhi. Also characteristic of Bactrian and early Sogdian sites are arrowheads with barbed triangular flanges, which differ markedly from the triangular-flanged, straight-based arrowheads found in Khorezmia, Turkmenia and at Kaunchi culture sites (near Tashkent). The graves of Ferghana, emphasising that region's role throughout the Bronze and Iron Ages as an obvious funnel for all manner of Eurasian nomads, contain arrowheads of almost every type76. The arrowheads discovered at early Kushan sites in Sogdia are similar to those found at Begram in Bactria, thus dating them to the 'late-Yuezhi/early Kushan period'77.
Household objects, including knives, spindle-whorls and a variety of toilet-articles also demonstrate considerable local variation. Bronze mirrors have been found at sites all over former Soviet Central Asia, with Khorezmia, Sogdia and Bactria yielding mirrors of a similar so-called 'Sarmatian' type, and Ferghana again yielding the largest number and greatest variety of examples78.
Belt buckles have also been found in extensive numbers, indicating that most pastoral nomads probably wore belts with buckles. Differentiation of buckle types is also a potential source of tomb identification. The link between Sogdian and Bactrian sites is again apparent Gorbunova notes that 'belt buckles from Bactrian and early Sogdian cemeteries are obviously of the same kind, and differ from the buckles of other places'79. The nomadic cultures associated with the different tomb types thus display a series of distinctive individual traits through their material possessions, but also a range of other features commonly shared between all nomadic 'peoples', which increases the difficulties of definitive attribution. Undeniably there is substantial evidence of nomadic passage along (and even temporary occupation of) various sites in the Zeravshan Valley, and it remains a matter of ongoing interpretation as to which tombs and funerary objects (if any) might be convincingly identifiable as Yuezhi.
Shishkina argues, perhaps too dismissively, that there is no evidence whatsoever of Yuezhi influence, although his conclusion that the densely populated Zeravshan valley proved ultimately unsuitable as a possible relocation site for the nomads is self-evidently correct:
'In the second century BC the area between the Amu Darya and the Syr Darya was the object of a massive invasion by peoples of a foreign culture, of which tribes of the Great Yuezhi made up a considerable part. Nevertheless the character of the Hellenized culture was preserved and there are no traces of new influences. One may assume that (the Yuezhi) passed through the Zeravshan Valley peacefully and did not remain there for long. There certainly would not have been enough room on the densely populated and cultivated lands of Soghd for masses of migratory peoples'80.
Shishkina's claim that there is 'no trace' of the passage of the nomads obviously takes no account of Zadneprovsky's argument that significant numbers of podboy tomb types, similar in style and design to those found in other areas of probable Yuezhi occupation (the Gansu, Ysyk Kul and Ferghana), have also been unearthed along the Zeravshan. Furthermore, in addition to the similarity of tomb-type, many of the artifacts discovered in podboy tombs also display a range of common features, particularly those found in tombs in Sogdia and Bactria; Gorbunova notes similarities between 'Bactrian and early Sogdian cemeteries with their distinctive types of pottery, weapons, belt buckles and women's clothing'81. Whilst it is clearly impossible to prove that these tombs and their artifacts were left along the Zeravshan by the migrating Yuezhi, there is certainly sufficient evidence to at least entertain the possibility.
'Sogdian' Statuettes
Further archaeological evidence of Yuezhi activity in Sogdia might be provided by the attribution of a particular sub-group of Sogdian statuettes unearthed at several sites in and around Samarkand, particularly at Tali Barzu south of the city, and also at Afrasiab. A form type of female statuette discovered at these (and other nearby) sites, characterised by the position of the hands under the breasts or resting on the upper stomach, and with distinctive facial features and dress, has been analysed by Fiona Kidd of Sydney University82. The primary focus group of figurines is distinct and stable in form a female figure with small protruding breasts and very slim arms bent sharply at the elbows so that her hands are resting on her upper stomach. Related groups include statuettes with stumps instead of arms, or with arms resting on the lower rather than upper stomach. The three groups are united by their similar style of facial features. As a stylistic form, similar examples are known from Mesopotamia at least as early as the 2nd Millennium BCE, although it has been suggested that the form probably dates much further back into antiquity83.
Where there might be some (admittedly very tenuous) link between a particular sub-type of these figurines and the Yuezhi is through a comparative coroplastic analysis of costume styles. Examples from the Samarkand area show a variety of clothing styles pleats on the lower part of the dress; a looser dress worn over long trousers with folds falling from the waist; a long pleated dress with pleats falling from the waist; or with pleats falling all the way from the shoulders84. Abdullaev has tentatively attributed one particular group of figurines clothed in a long dress with a flared lower half to the Yuezhi during their interim residency of the Zeravshan Valley85. The identification is somewhat strengthened by the discovery of similarly attired figurines at other Yuezhi and early Kushan sites, notably Tillya-tepe and Dalverzin-tepe86.
Sogdia lay at the heart of a network of ancient migration and trade routes, and it is only to be expected that Sogdian art would be subjected to a range of stylistic influences, including those of pastoral nomads, be they steppe- or oasis-dwellers. Kidd argues that 'the nomadic legacy of the peoples living in this region must also be recognised as an important factor in the formation of style in Sogdian art generally'87. That the Yuezhi occupied, if only for a relatively brief period, parts of the Zeravshan valley including the greater Samarkand oasis is more than likely, given the cumulative textual and archaeological evidence considered above. The possibility that further evidence of their residency might be found through a coroplastic analysis of the costume of figurines unearthed at probable Yuezhi sites in the region is thus also worthy of consideration.
Archaeological Evidence: Conclusion
Archaeological evidence for the migrating Yuezhi confederation in both Sogdia and Ferghana is at best inconclusive. Mass migrations of nomadic 'peoples' would hardly be expected to leave a substantial archaeological record anyway, given that they generally did not construct settlements or 'townships' meant to last longer than a winter season. During the long, cold Central Asian winters the dead would be buried beside the encampments, and in the spring the tribes would move on in search of new dwelling places. This would surely have been the lifeway of the Yuezhi during the short period of forced migration following their eviction from the Ili Basin in c. 133/2 and preceding their arrival in northern Bactria in c. 130 BCE. Thus the only record of their passage through the Ferghana and Zeravshan valleys would be their funerary monuments and the grave goods they contained. The incidence of podboy tombs in both regions (mostly with a common north-south orientation), the similarity of household objects, weapons, arrowheads and belt buckles in those tombs, and the possible evidence of Sogdian statuettes, provides clearly circumstantial but arguably cumulatively convincing evidence that there does indeed exist a material record of the passage of the Yuezhi. Textual and archaeological evidence thus allows for the following tentative narrative reconstruction of events.
Conclusion and Summary
As a result of their defeat by the Wusun in 133/2 BCE, and after almost three decades of residency in the Ili Valley, the confederation of the Da Yuezhi was forced to resume its migration westwards, moving initially into the Ferghana Valley. Here they may have briefly occupied sites in the Isfara region before continuing their journey (in the spring of 132/1?), travelling west and south into the Zeravshan valley of Sogdia, before perhaps following it into the Samarkand and Bukharan oases. This itinerary is implied by several references in the Han Shu and Shi Ji, and also tentatively reinforced by Russian and Central Asian archaeology.
In the penultimate stage of their journey the Yuezhi horde entered the realm of Kangju, a dynasty in control of a region most probably identifiable as Sogdia and the Zeravshan Valley. 'Kangju' was the Han name for both the 'state' and the powerful ruling dynasty, which may have been in control of Sogdia since it gained independence from the Graeco-Bactrians late in the third century BCE, during the reign of Euthydemus. Ptolemy provides incidental evidence of the identification of Kangju with Sogdia, and once again continued to unknowingly chart the course of the Yuezhi migration by noting a group he this time called the Tachoroi (Tocharians) dwelling in Sogdia88.
Archaeologists have unearthed podboy tombs potentially attributable to the Yuezhi at the periphery of the Bukharan and Samarkand oases, as well as in the upper valley of the Zeravshan. All are similar in design, funeral ceremony and 'accompanying inventory' of artifacts to other possible Yuezhi podboy monuments found in northern Bactria89. In addition, the coroplastic analysis of a group of Sogdian statuettes, might eventually provide supporting evidence of brief Yuezhi occupation.
Both the Han Shu and Shi Ji suggest some level of diplomatic interaction between the Yuezhi and Kangju dynasties prior to the Yuezhi's arrival in northern Bactria90. Although Kangju is described as a 'small' country in the Chinese sources, its military strength was substantial, with some 120,000 persons able to bear arms, according to Ban Gu,91 of which 90,000 were skilled archer warriors according to Sima Qian92. Unlike Dayuan, it was described as powerful enough to be 'not easily defeated by Han forces',93 but none the less was 'constrained to serve the Xiongnu' in the east,94 and acknowledged 'nominal sovereignty' (or was 'subservient to') the Yuezhi in the south'95.
Kangju displayed obvious ability at inter-state diplomacy by balancing these competing (and essentially opposed) interests recognising the nominal sovereignty over parts of its country by both the Xiongnu and the Yuezhi, aiding Zhang Qian and subsequent Han envoys, and later sending a Kangju prince as envoy to China. It is therefore probable that the rulers of Kangju agreed to give the Yuezhi safe passage through their territory, and accepted some form of subservient relationship to avoid military conflict. The Sogdian dynasty may even have encouraged the Yuezhi to settle at the southern extremities of their territory, and suggested the fertile river valleys north of the Amu Darya as an ideal homeland for the migrants.
Eventually Kangju would be drawn further and further into the Yuezhi/Kushan realm over the following decades and centuries until the southern portions of its territory were probably incorporated into the Kushan Empire, or at least into its sphere of influence96. By 83 CE the Kushans would further cement this relationship through an alliance based on a 'bond of royal marriage' with the ruling family of Kangju97. Kangju/Sogdia thus proved to be an ideal buffer for the Yuezhi between their new northern Bactria homeland and the Wusun and Xiongnu realms in the steppes to the north and east. Ultimately, as a result of the generally cooperative relationship apparently established almost immediately between the Kangju and Yuezhi dynasties in c. 131 BCE, Sogdia went on to become an integral and stable member (through direct incorporation or alliance) of the greater Kushan sphere of influence until the dissolution of the Kushan Empire in the mid-third century of the Common Era.
Craig Benjamin
March 2003
Dr. Craig Benjamin is a Lecturer in History at Macquarie University in Sydney, Australia. He can be contacted at: craigbenjamin[@]optusnet.com.au
1 HS 61 1B.
2 SJ 123. See B. Watson, Records of the Grand Historian by Sima Qian Han Dynasty II (Revised edtn., Columbia University Press 1993) p. 231.
3 Ptol. Geog. vi, 14, 7-14. Hermann located Ptolemy's Tagouraioi to the north of the 'Alexandrovski Range', north of Ysyk Kul, in A. Hermann, Sakai col. 1770-1806', in Pauly Wissowa, Real-Encyclopadedie Der Classischen Alterumewissenschaft, 2e Reihe, Vol I (Stuttgart 1914).
4 Zadneprovsky (1999) op. cit., pp 4-5. See also J.P. Mallory and V.H. Mair The Tarim Mummies (London 2000) pp. 156 and 158.
5 HS 61 4B.
6 On the Hanging Pass see HS 96A 12A/12B; HS 96B 1B; H. Tsuchiya, 'Tracing Ancient Routes in Northern Pakistan', in M. Alram and D.E. Klimburg-Salter, Coins, Art and Chronology (Vienna 1999) pp. 353 ff.; A. Stein, Ancient Khotan (Oxford 1907) pp. 1-46.
7 See HS 61 4B.
8 HS 61 4B.
9 HS 61 1A.
10 See A.K. Narain, The Tokharians (Shillong 2000) p. 36, n. 11 for example.
11 Y.A. Zadneprovsky, 'Migration Paths of the Yueh-chih based on Archaeological Evidence', Circle of Inner Asian Art Newsletter No. 8 (April 1999) p. 5.
12 HS 96A 15B.
13 HS 96 A 17A.
14 SJ 123, Watson p. 234.
15 HS 96B 1B.
16 Ibid.
17 A.F.P. Hulsewe and M.A.N. Loewe, China in Central Asia. The Early Stage: 125 B.C. A.D. 23. An Annotated Translation of Chapters 61 and 96 of the History of the Former Han Dynasty (Leiden 1979) p. 125, n. 299, 2.
18 P. Pelliot, 'L'Edition collective des oeuvres de Wang Kouo-wei', in T'oung Pao 26 (1929) p. 151.
19 See Ibid. p. 150.
20 E.G. Pulleyblank, 'The Consonontal System of Old Chinese', in Asia Major 9 (1962) p. 94; see also L. Torday, Mounted Archers: The Beginnings of Central Asian History (Durham 1997) p. 319 n. 34 for a detailed discussion of the philological link between Kangju and the Jaxartes (Syr Darya).
21 Hulsewe and Loewe op. cit., p. 125, n. 299, 2.
22 E.G. Pulleyblank, 'Chinese and Indo-Europeans', Journal of the Royal Asiatic Society 2 (London 1966) p. 28.
23 Narain (2000) op. cit., pp. 29-30.
24 HS 96A 15B.
25 SJ 123, Watson p. 234.
26 HS 96A 15B.
27 SJ 123, Watson p. 234.
28 HS 96A 15B.
29 Ibid.
30 E. Zurcher, 'The Yueh-chih and Kaniska in the Chinese Sources', in A.L. Basham, ed., Papers on the Date of Kanishka (Leiden 1968) p. 360.
31 SJ 123, Watson p. 234.
32 HS 96A 16A.
33 HS 61 3A.
34 HS 96A 15B.
35 SJ 123, Watson p. 234.
36 HS 96A 17A.
37 G.V. Shishkina, 'Ancient Samarkand: Capital of Soghd', Bulletin of the Asia Institute 8 (1994) p. 90.
38 Hulsewe and Loewe, op. cit., pp. 130-1, ns. 320, 321, 322, 323 and 324 provide comments on attempts to identify the five principal towns of the lesser kings.
39 See R. Frye, The History of Ancient Iran (Munich 1984) p. 141.
40 Arrian, Book Four, 5-7 (1971) op. cit., pp. 208-211 for a description of this defeat.
41 See for example, A.B. Bosworth, Conquest and Empire: The Reign of Alexander the Great (Melbourne 1988) pp. 117, and 109-10.
42 See for example A. Nymark, Culture and Art of Ancient Uzbekistan vol.1 (Moscow 1991) p. 155.
43 See for example E. Rtveladze, The Ancient Coins of Central Asia (Tashkent 1987) p. 46.
44 Nymark op. cit., p. 156.
45 See O. Bopearachchi, 'The Euthydemus Imitation and the Date of Sogdian Independence', Silk Roads Art and Archaeology 2 (1991/2) pp. 11-12; also Rtveladze op. cit., p. 46.
46 Ibid., p. 12.
47 Ptol. Geog. vii, 2, 15; vi, 14, 7-14. See W.W. Tarn, The Greeks in Bactria and India (Cambridge 1938, 2nd edtn. 1951) p. 517; and J.W. McCrindle, Ancient India as Described by Ptolemy (London 1885) p. 281
48 HS 96A 15A.
49 SJ 123, Watson p. 234.
50 SJ 123.3b, trans. E. Zurcher (1968) op. cit., p. 360.
51 HS 61 2A.
52 SJ 123, Watson p. 232.
53 HS 61 3A.
54 HS 96A 15B.
55 HS 96A 15B; and Zurcher (1968) op. cit., p.360.
56 Torday op. cit., p. 301.
57 HHS 77.4a.
58 HS 96A 16A.
59 HS 96A 19A.
60 On the Zhaman-Togai Xiongnu tomb see A.G. Maximova, M.S. Merschiev, B.I. Vainberg, L.M. Levina, Drevnosti Chardary (Antiquities of Chardara) (Alma-Ata 1968) pp. 175-190.
61 Torday op. cit., pp. 301-2.
62 On Sarazm see A.I. Isakov, 'Sarazm: An Agricultural Centre of Ancient Sogdiana', Bulletin of the Asia Institute 8 (1994) pp. 1 ff.
63 See V.A. Shishkina, 'Arkheologicheskie nablyudeniya na stroitel' stve Katta-Kurganskogo vodokhranilishcha' (Archaeological Observations on the Building Site of the Katta-Kurgan Reservoir) Izvestiya Uzbfil AN SSSR Bulletin of the Uzbek Branch of the USSR Academy of Sciences (1940) No. 10, pp. 19-24.
64 See A.I. Terenozhkin, 'Sogd I Chach' (Soghd and Chach) KSIIMK 33 (1950) pp. 152-169; A.I. Terenozhkin, 'Roskopki na gorodishche Afrasiabe' (Excavations on the Site of Afrasiab) KSIIMK 36 (1951) pp. 136-140.
65 See V.I. Shishkina, 'Uzbekistanskaya arkheologicheskaya ekspeditsiya AN UzSSR' (The Uzbekistan Archaeological Expedition of the Uzbek Academy of Sciences. Field Work in 1956-1959), IMKUz (Tashkent 1961) pp. 36-43; V.I. Shishkina, Afrasiab sokrovishchnitsa drevnei kultury (Afrasiab Treasury of Ancient Culture) (Tashkent 1966) pp. 8-10.
66 See A.Y. Yakubovsky, 'Itogi rabot Sogdiiko-Tadzhikskoi arkheologicheskoi ekspeditsii v 1946-1947 gg' (Results of the Work of the Sogdian-Tajik Archaeological Expedition of 1946-1947) MIA SSR No. 15 (Moscow-Leningrad 1950) pp. 13-28.
67 See for example, A.M.Mandleshtam, 'Mogilnik v s. Zosun verkhovya r. Zeravshan (The Burial Site in Zosun Village, Upper Zeravshan) Izvestia Otdelniya Obshchestvennykh nauk AN Tadzh. SSR No. 40 (1965) pp. 29-44.
68 A.M. Mandleshtam, 'Archaeological Data on the Origin and Early History of the Kushans', in Central Asia in the Kushan Period (1970) op. cit., p. 166.
69 See for example O.V. Obelchenko, 'Kuyu-Mazarsky mogil'nik' (The Necropolis of Kuyu-Mazar) Trudy IIA AN Uz. SSR VIII (Tashkent 1956) pp. 205-227; and O.V. Obelchenko, 'Kurgany okolo sel. Khazara' (The Kurgans Near the Village of Khazara) IMKUz (Tashkent 1963) pp. 57-65.
70 N.G. Gorbunova, 'Early Nomadic Pastoral Tribes in Soviet Central Asia during the First Half of the First Millennium A.D.', in G. Seamen, ed., Foundations of Empire: Archaeology and Art of the Eurasian Steppes (Los Angeles 1992) p. 34; and see O.V. Obelchenko, 'Kurgannye mogilniki epokhii Kushan v Bukharskom oazise' (Kurgan Barrows of the Kushan Epoch in the Bukhara Oasis), in Central Asia in the Kushan Period, vol I (Moscow 1974) for a summary of Obelchenko's conclusions.
71 Mandleshtam (1970) op. cit., p. 166.
72 Y.A. Zadneprovsky, History of Central Asian Nomads in the Kushan Period', in Kushan Studies in the U.S.S.R. (Moscow 1970) pp. 148-9.
73 Zadneprovsky (1999) op. cit., p. 5.
74 See Gorbunova (1992) op. cit., pp. 36 ff for a general outline of material artifacts discovered at a range of Central Asian sites.
75 Ibid., p. 36.
76 Ibid., p. 39; and see G. Frumkin, Archaeology in Soviet Central Asia (Leiden 1970) pp. 42 ff, for a detailed summary of Ferghana Valley archaeological discoveries, including grave artifacts.
77 See R. Ghirshman, Begram Recherches Archeologiques et Historiques sur les Kouchans (Paris 1946), Plate XXXVI, Nos. BG 290 a and b for excellent photographs of the Begram arrowheads with barbed triangular flanges. At the British Museum in January 2001 Elizabeth Errington was good enough to show me a colour transparency of a similar arrowhead from the Charles Masson collection. See also E. Errington, 'Rediscovering the Collections of Charles Masson', in M. Alram and D.E. Klimburg-Salter, eds., Coins, Art and Chronology. Essays on the pre-Islamic History of the Indo-Iranian Borderlands (Vienna 1999) pp. 207ff.
78 See Gorbunova (1992) op. cit. p. 44, Fig. 5 for a map illustrating the 'Distribution of Mirror Types'.
79 Ibid., p. 41.
80 Shishkina op. cit., pp. 89-90.
81 Gorbunova op. cit., p. 42.
82 Fiona Kidd, 'The Chronology and Style of a Group of Sogdian Statuettes', in C. Benjamin and S. Lieu, eds., Walls and Frontiers in Inner Asian History, Silk Roads Studies VI (Turnhout 2002) pp. 197 ff.
83 P. Ucko, Anthropomorphic Figurines of Predynastic Egypt and Neolithic Crete with Comparative Material from the Prehistoric Near East and Mainland Greece (London 1968). See also M. Gimbutas, The Goddesses and Gods of Old Europe (London 1982) for early examples of similar figurines.
84 Kidd op. cit., pp 197 ff.
85 K. Abdullaev, in R.K. Suleimanov, Drevnii Nakshab: Problemi' Tzivilizatzii Uzbekistana VII v. do N.E. (Tashkent 2000) p. 203.
86 On the Tillya-tepe figurines see V.I. Sarianidi, Khram I Nekropol' Tillyatepe (1989) p. 53 fig. 15 and p. 57 figs. 17, 25 and 70; on the Dalverzin-tepe figurines see K. Abdulaev, 'Portrayal of musicians in Bactrian teracotta figurines', in Information Bulletin 7 (1984) pp. 52-56 figs. 4-5.
87 Kidd op. cit., p. 211.
88 Ptol. Geog. vii, 2, 15; vi, 14, 7-14. See Tarn (1938/1951) op. cit, p. 517; and McCrindle (1885) op. cit., p. 281.
89 Zadneprovsky (1999) op. cit.,p. 5.
90 HS 96A 15B; SJ 123, Watson p. 232.
91 HS 96A 15B.
92 SJ 123, Watson p. 234.
93 HS 61 3A.
94 HS 96A 15B.
95 HS 96A 15B, and Zurcher op. cit.,p. 360.
96 Whether Sogdia/Kangju or Khorezmia were ever actually part of the Kushan Empire is a matter of some conjecture. See, for example, Svend Helms, 'Ancient Chorasmia: The Northern Edge of Central Asia from the 6th Century B.C. to the mid-4th Century A.D., in David Christian and Craig Benjamin (eds.), Worlds of the Silk Roads: Ancient and Modern (Turnhout 1998), pp. 85-6; and S. Helms (in private correspondence) who has argued that Khorezmia at least remained nominally independent.
97 HHS 77.4a.
