Hephthalites

Rouran epoch Göktürk epoch Uyghur epoch Kirghiz epoch

Rouran First Göktürk Tang rule Second Göktürk Uyghur Kirghiz Huns Western Göktürk Türgesh Karluk Khazar

Rouran 400–554
First Göktürk 552–630
Tang rule 630–682
Second Göktürk 682–744
Uyghur 744–840
Kirghiz 840–924
Huns 370–469
Western Göktürk 552–659
Türgesh 699–766
Karluk 766–800
Khazar 650–969

The Hephthalites were the last power of Hunnic descent to rule Central Asia. They were also the first great conquest of the Göktürks in the west. At their height in the early sixth century they held Sogdia and the country from Afghanistan to the oasis cities of the Tarim. They took tribute from north-western India. They had twice decided who sat on the Sasanian throne. Within a few years of the founding of the khaganate they were gone. Istemi and Khosrow I destroyed them, acting from opposite sides.

Name and composition

White HunsGreek
嚈噠Chinese
滑ChineseLiang Shu 54
挹怛ChineseSui Shu 83
HayāṭilaArabic
EbodaloBactrian

The name obscures one point. The Hephthalites should not be regarded as a distinct ethnic group at all. They were a confederation of nomadic and semi-nomadic groups, Altaic, Iranian and probably others. The Heftal dynasty and a core of War-Hun tribes held the confederation together.^[1] Their name in the sources is the name of a ruling house that gathered them. It is not the name of a nation that produced one. The idea goes back a century.^[2] Under the name Hephthal only a ruling political entity is to be understood. The mass of the state consisted of Kidarites, Kushans, Chionites and Huns.^[1] The Liang Shu calls the people Hua. It uses the name from which the later Yeda is read only for their king.^[3]^[4] The few data on their language show Turkic and Iranian elements and elements that are neither. This mixture points to an extremely mixed speech.^[1] One modern survey closes its review of the theories with the remark that many opinions compete and no general agreement is possible at the moment.^[1]

The prehistory runs through two earlier waves. In the middle of the fourth century a people called Xiyōn in Middle Persian and Chionitae in Latin began to press on the Sasanian frontier. The name is in all likelihood a rendering of the same ethnonym that the Chinese wrote as Xiongnu. Around 360 their chieftain Grumbates fought in Shapur II’s Roman war as a Persian ally. A second grouping formed under a chief named Kidāra. The Byzantine sources call this grouping the Kidarite Huns. Towards the end of the century new bands, probably War-Huns, entered the region. They pushed the Kidarites west towards Balkh and northern India. There Indian inscriptions know them as the Śveta Hūṇas, the White Huns, under Toramāṇa and Mihirakula. Out of this movement, and probably in connection with Rouran pressure, the Heftal dynasty emerged by the middle of the fifth century. The founder was Khingila. He united the Chionites and the Uar after they had fled westward before the Wusun, who were themselves being driven by the rising Rouran.^[5] The later closeness of Hephthalites and Rouran would then be a family matter before it is a political one. Whether the Rouran were in fact their overlords is not settled. The Liang Shu is explicit for the early period. While the Wei still dwelt at Sanggan, the Hua were a small state subject to the Rouran. Only later did they grow strong and take the lands round them.^[3]^[6]

The portraits on the silver coins of the Hephthalite kings of northern India in the fifth century wear a high cap and show no artificially deformed skull. Two skulls from Chionite graves of about 500 at Begram, north of Kabul, are undeformed as well.^[7] On that evidence an older view holds that the Chionites, whom Procopius calls White Huns, had nothing to do with the Huns of Europe.^[7]

Language and origins

The disagreement is old. The Liang Shu reports that their language could be understood only by speakers of Tuyuhun. Other accounts say it differed from that of the Rouran, who are believed to have spoken a form of the same Xianbei stock. The Liang Shu says that the language of the Hua was understood only after translation by people of Henan, the name it gives the Tuyuhun. The Wei Shu says that it differs from that of the Rouran, the Gaoche and the other Hu.^[8] The Liang Shu itself derives the Hua from the Cheshi of Turfan. The Eight Hua of the Cheshi fought for the Han general Ban Yong against the northern barbarians in 126.^[3] Once established, the Hephthalites used Bactrian, the language of the population and the bureaucracy they had conquered. This shows what they ruled, not what they were.

A single word may point the other way. The element čamūk stands in the names of several Sogdian rulers and in the name of a city in the History of Bukhara. The Tang transcribed it as the surname Zhaowu that they gave to Sogdians. Some scholars regard it as a Hephthalite word rather than a Sogdian one, on the ground of where and when it is attested.^[9] This is a proposal. The Hephthalite language is too little known to test it.

In the part of its entry that goes back to Wei Shou, the Wei Shu calls them a separate branch of the Gaoche. They came out from beyond the northern frontier and moved south from the Golden Mountain, the Altai.^[10] The chapter as it stands is a Song restoration from the Bei Shi. It took in sentences of later histories, so that the assignment of any one sentence to Wei Shou’s own text is an inference.^[11] The entry adds that they intermarried with the Rouran and sent embassies to the Wei from the Tai’an era, 455 to 459, onward.^[12] This northern origin has been taken at its word. It makes the Hephthalites an East Asian people from the Altai who became Iranian only in the country they conquered.^[13]

The Zhou Shu derives them from the Great Yuezhi. It puts their king’s seat in a walled city west of Khotan. It remarks that their penal law and customs are about the same as those of the Göktürks.^[14] It records that brothers share one wife. The wife wears one horn on her hat if her husband has no brothers and adds a horn for each brother he has.^[14] Khotan, Parthia and more than twenty states great and small were subject to them. After the Göktürks broke them, their people scattered and their tribute ceased.^[14] Procopius calls them a people of Hunnic stock. He adds that they had long lived apart from the Huns, in a good land, and are the only Huns with white bodies and faces. He also describes retinues of personal servitors around their aristocrats, the comitatus familiar from the Iranian and Turkic world.

The emperor Yang of the Sui sent Wei Jie with a colleague to the western countries.^[15] The Tongdian quotes Wei Jie for one sentence. He asked the people of the country himself, and they all called themselves Yitian.^[16] The verdict that follows is Du You’s own, after a comparison with the Han Shu. What comes from distant lands and foreign tongues is corrupted and concerns very old times, so that the origin of the Hephthalites cannot be determined.^[16] The Sui Shu records no meeting with Hephthalites. The mission reached Kipin. It received Buddhist scriptures in a city called Wangshecheng, “the city of the royal residence”. It also took ten dancing girls, lion skins and asbestos from Shi.^[15] Wangshecheng has been identified with the Hephthalite capital, whose Chinese name the Wei Shu glosses in the same words. Another identification is Rajagriha in India.^[17]^[18] Fifteen centuries later the question is still open. The difficulty is not only the state of the sources: a confederation does not have one origin.

Rule over Sogdiana

The Hephthalite conquest of Sogdiana falls in 509.^[19] No written record says so. The date comes from two observations. In that year Hephthalite embassies abruptly replaced Sogdian ones at the Chinese court. At Panjikent the walls and a temple were partly destroyed and at once repaired, and a three-storey barracks for a permanent garrison was added to the outer wall.^[19] The Kidarite dynasty that they displaced had probably been waning since the 470s. The Sasanians and the Hephthalites themselves had beaten it, and the Hephthalites had driven it from its Bactrian bases from 456. The conquerors were fabulously rich. They had killed Peroz in 484. They had put his successor Kavad on the throne of Iran, and back on it more than once. For decades, at least to the beginning of Khosrow’s reign, the Sasanians paid them a colossal tribute. On one occasion Peroz sent as many silver coins as thirty mules could carry. Those coins are still found by the thousand in the markets of Afghanistan. A great part of the Sasanian money supply went that way to Central Asia. The Sasanian apocalyptic texts count Sogdian troops among the pillagers of Iran. The Sogdians merely changed masters and went on trading in China. They spread the silver through an empire that at its height in 520 nominally ran from Merv to Turfan and Gandhara. Sasanian silver became the currency of all Central Asia to Gansu for close on two centuries. The Sogdian merchants found in it the capital to enlarge their trade. The one Hephthalite of the years after the fall who is named in a western source is Katulph. At the Persian court he advised Khosrow to buy the Göktürk silk and burn it. The Maniakh article treats the scene.^[20]

The royal seat

The Chinese histories name the capital of the Hephthalite king Badiyan. They gloss the name as the city of the royal residence.^[21] The Zhou Shu puts it west of Khotan. The Wei Shu and the Sui Shu put it more than 200 li south of the river Wuxu, which is identified with the Oxus.^[22]^[23] The king did not stay in one place. The Wei Shu says that he kept his wives in separate camps two or three hundred li apart and went round them, a month at each. In the cold of winter he did not move for three months.^[24]

Where Badiyan lay is argued, and the proposals are far apart. One puts it at Badghis near Herat. The Arabic geographers call Badghis the old capital of the Hephthalites, and the Chinese name answers to it.^[18] A second rejects Badghis. The later struggles of the Hephthalite prince Nezak aimed at Balkh and upper Tokharistan. It therefore puts the city at Huo, the later Kunduz, which the Arabic geographers know as War-waliz.^[25] A third puts it in Badakhshan, at Faizabad. It argues from the route of the Chinese embassy of 519 and from the country that Xuanzang calls Himatala.^[26]

The temples and stupas that the later histories give the city may be a late layer. They may belong to the report that Pei Ju made in 607, after the realm had fallen. The argument is that the earliest account, of the embassy of 519, found a people that did not believe in the Buddhist law and served many divinities that the pilgrim called heretical.^[27] The Liang Shu says that they worshipped the spirits of heaven and fire. The Liang Shu and the Wei Shu say that they lived in felt tents.^[8]

Embassies to China

The Hephthalites sent embassies to the Northern Wei from the Tai’an era onward. At the end of the Zhengguang era they sent a lion. The rebellion of Moqi Chounu held the lion back at Gaoping until his defeat, when it was forwarded to the capital. After the Yongxi era their tribute stopped.^[24] The Liang court received their envoys in 516, 520 and 526. In 520 they brought a yellow lion, white sable furs and Persian brocade.^[3] After the Göktürks had broken the realm, envoys still came to the Sui court in the Daye era, between 605 and 618.^[15]^[28]

Traces in art

Their rule over Sogdiana is read in part from images, and two crowns are tied to them. The first is the winged crown that the Sasanian king Peroz took as his third crown in 476/477.^[29] The Sasanians paid Peroz’s drachms to the Hephthalites in large quantities. The Hephthalites struck coins of their own in imitation of them.^[29] After Peroz no Sasanian king wore the winged crown for more than a hundred years, until Khosrow II in 590. On that ground Peroz’s crown is taken as the model of the crown of the Hephthalite kings.^[29] The second is a crown with three crescents. The coins show it first under Khingila, about 460 to 490. His successors varied it.^[30]

Terracotta figurines of the sixth and seventh centuries from the Surkhandarya valley wear both crowns, a caftan with a triangular collar and a particular hairstyle. Figures of the same kind are painted at Bamiyan, at Kizil near Kucha and in Sogdiana.^[29] This dress may be the sign of a cultural unity that Hephthalite rule created across the lands from Sogdiana to the Tarim basin. That argument denies that the dress came from Sasanian Iran.^[29]

Sogdians who lived in China had the crowns drawn on their tombs. On the north wall of the sarcophagus of Shi Jun, buried at Xi’an in 580, Shi Jun comes with his caravan to a king who sits in a yurt and wears a winged crown.^[29] The yurt marks a nomad ruler. The long hair down the back that marks a Göktürk in the same scene on the couch of An Jia is missing. A Sasanian king would not be shown in a yurt.^[29] The king has therefore been identified as a Hephthalite ruler. The winged crowns on the coins of the Nezakshah dynasty of Kapisa, struck soon after the fall of the Hephthalite realm, are taken to show that the Hephthalites had set the fashion. Shi Jun’s life, which ended in 579, would then lie in the Hephthalite period and not in the Göktürk one.^[29]

The crown does not mark a Hephthalite alone. A second figure with a winged crown on the same sarcophagus appears three times as the chief figure of the banquets and the hunt. His crown has a small circle and a crescent between the wings. He may be another Hephthalite ruler. He has also been identified as Shi Jun himself.^[31] Yu Hong, who was buried in 592 at Taiyuan, wears the winged crown on the panels of his sarcophagus. A headdress of gold foil in the shape of a winged crown lay on the head of a dead man of the Sui or Tang period in a tomb at Jiulongshan near Guyuan in Ningxia.^[30] The triple-crescent crown appears as well. Two non-Chinese guardians wear it on the gates of the tomb of Di Caoming at Jingbian in northern Shaanxi. Di Caoming was buried in 579, and his title Tianzhu of Xia is understood as that of a Zoroastrian priest.^[30] On the view that joins these finds, Sogdian leaders in China took up the crowns of the conquerors of Central Asia. The triple-crescent crown, which had been counted a Sasanian type since the 1950s, is a Hephthalite one.^[32]

Relations with Sasanian Persia

Their relations with Sasanian Iran are the clearest case of what a nomadic power beside a settled empire could do. They raided the borderlands. They sold their cavalry to claimants for the throne. They killed a king who came looking for their treasure. Peroz took the crown from his brother Hormizd III with Hephthalite help. He died in battle against those same allies in 484. Kavad regained his throne with Hephthalite assistance and may have had to acknowledge their overlordship to do it. At that point they were, in effect, the arbiters of the Sasanian succession.

Destruction by the Göktürks and Persia

Then the Göktürks came. Istemi, his brother Bumin’s viceroy for the west, moved against the Hephthalites in alliance with Khosrow I. The Göktürks attacked from the north-east and the Persians from the south-west. The Hephthalite state was destroyed and its ruler killed. The victors divided the territory along the Oxus. This arrangement would sour into the Göktürk–Persian war within a generation. The Sui Shu adds that when the Hephthalite country fell into disorder, the Göktürks sent a Tong Shad to take it by force.^[15] The text adds the two characters 字詰, which some scholars take together as his name, Zijie.^[33] In the same chapter, however, the word zi 字 introduces the name of a foreign ruler. By that usage his name would be Jie alone.^[15]

The realm fell about 560.^[5] The decisive engagement, the battle of Bukhara, belongs to those years. Istemi concluded the alliance with Khosrow only shortly before the invasion.^[5] Two other datings are held. One dating puts the destruction in 557.^[34] A second reviews the proposals in the field, where 557, 558, 563 and 565 have all been argued. It sets the end between 563 and 567. Its ground is that Menander’s envoys of 568 could answer the emperor’s question whether they had subjected all the Hephthalites with the one word “all”.^[1] The difference of a decade sits at the junction of Chinese, Persian, Arabic and Greek reckonings.

The fullest reconstruction of the campaign is assembled from Firdausi, Tabari, Dinawari and the Greeks. It runs as follows.^[1] The Göktürks reached the Aral Sea and the Hephthalite border in 555.^[1] The Sasanian–Göktürk alliance was sealed by the marriage of Khosrow to a daughter of Istemi. One dating puts the marriage in 554, another after the victory.^[1] The first collision, under Mukan, fell in 555. The second, under Istemi, fell in 558.^[1] The Hephthalite king Gatfar had a counsellor named Katulph who advised against war. Katulph was insulted and went over to Khosrow. The Hephthalites killed a Göktürk embassy passing through Sogd, all but one man, and the war became inevitable.^[1] The Göktürks took Chach and crossed the Chirchik. Gatfar gathered troops from Balkh, Shughnan, Termez and the rest of the realm near Bukhara. He declined battle on the plain and fought in the hills at Nesef. There, after eight days, the Göktürks won.^[1] The survivors chose the ruler of Chaghaniyan, Faganish, who submitted to the Sasanians.^[1] Khosrow took the country south of the Amu Darya. The Göktürks took Sogd, Chach, Ferghana and eastern Turkistan. An agreement dated to 566 or 571 fixed this division. The Amu Darya became the frontier, with northern Tokharistan a Hephthalite buffer paying tribute to the Göktürks.^[1] Every element of that account is reconstruction from sources at second hand.^[1]

The name of the last king is not secure either. Tabari gives him as Wrz. Al-Masʿudi gives Axsunwāz, which recalls Axsunwār, the title or name borne by the Hephthalite ruler in Peroz’s day.^[35] Firdausi calls him Ghatfar.^[36] A line of older scholarship made Katulph the last king. He was in fact a noble at the court and never held the throne.^[37]

Map of Asia around 560: the Göktürk Khaganate from Manchuria to the Caspian, Istemi advancing through Chach to the battle of Bukhara, Khosrow I advancing from the south to Balkh, beside Sasanian Persia, the Northern Zhou, the Northern Qi and the Chen — Around 560 Istemi advances through Chach into Sogdiana and destroys the Hephthalites in the battle of Bukhara, while Khosrow I takes their lands south of the Oxus as far as Balkh. The Oxus becomes the frontier between the Göktürks and Persia. Modern map; Istemi's route is drawn.

Aftermath and legacy

The collapse ended the last of the War-Hun confederations. It opened the western steppe to the Göktürks. Remnants stayed in Bactria as Göktürk vassals and fought beside them against the Sasanians. Part of the Hephthalite nobility later passed under Persian rule instead. After the Göktürk victory no Hephthalite king ruled the whole country again. In the south, Herat, Talikan and Tokharistan lay now in Persian hands, now in those of local princes, until the Arabs came. Among these princes the Persian sources single out the Kushans and a Hephthalite king.^[38] Petty dynasts of Hephthalite origin may have continued in the region for generations. Nezak Tarkhan, lord of Badghis, is called a king of the Hephthalites in one Arabic source. The Arabs took his castle in 703 or 704. He made terms with Qutayba ibn Muslim in 705 or 706. In 708 or 709 he tried once more to drive out the Arabs and restore the Hephthalite realm, which ended his career.^[39] The Arabic notices for the sixty years from his first appearance to his end may not all concern one man.^[40] In 644 Xuanzang passed through a country east of Kishm that he calls Himatala. It is understood as a Hephthalite remnant. Its people lived in felt tents, its borders touched those of the Göktürks, and it had taken over their customs.^[41]^[42]

Nobody can say when they ceased to exist as a cultural unit, because their population was mixed from the start. Traces have been claimed in the ethnogenesis of the Pashtuns, of the Turkic Karluk and Khalaj in Afghanistan, and of the Äynu of eastern Turkistan. Al-Khwārazmī has been understood to connect the Khalaj and the Kanjina with them.^[43] The Kamiji of Chaghaniyan, a Turkic tribe still named in the eleventh century, have been taken as a remnant of the Hephthalites as well.^[43] These are proposals, not findings.

References

^ a b c d e f g h i j k l m n Kurbanov 2010
^ Markwart 1938, p. 45
^ a b c d Liang Shu, juan 54
^ Herrmann 1925, p. 565,571
^ a b c Yavuz 2025
^ Herrmann 1925, p. 569
^ a b Werner 1956, p. 11
^ a b Liang Shu, juan 54; Wei Shu, juan 102
^ Kageyama 2007, p. 11
^ Wei Shu, juan 102; 内田吟風 1972, p. 375
^ Herrmann 1925, p. 566
^ Wei Shu, juan 102; 内田吟風 1972, p. 377
^ 内田吟風 1972, p. 376
^ a b c Zhou Shu (Book of Zhou), juan 50
^ a b c d e Sui Shu (Book of Sui), juan 83
^ a b Tongdian, juan 193
^ Herrmann 1925, p. 579
^ a b Chavannes 1903, p. 224
^ a b de la Vaissière 2005a
^ Menander Protector
^ Zhou Shu (Book of Zhou), juan 50; Wei Shu, juan 102
^ Zhou Shu (Book of Zhou), juan 50; Wei Shu, juan 102; Sui Shu (Book of Sui), juan 83
^ Markwart 1938, p. 47–48
^ a b Wei Shu, juan 102
^ Markwart 1938, p. 43–45
^ Herrmann 1925, p. 572–576
^ Herrmann 1925, p. 568–571,577
^ Herrmann 1925, p. 578–579
^ a b c d e f g h Kageyama 2007, p. 12
^ a b c Kageyama 2007, p. 13
^ Kageyama 2007, p. 12-13
^ Kageyama 2007, p. 13-14
^ Markwart 1938, p. 148
^ Golden 1992
^ Tabari
^ Felföldi 2001, p. 197
^ Felföldi 2001, p. 199
^ Herrmann 1925, p. 565
^ Markwart 1938, p. 41–43
^ Markwart 1938, p. 41
^ Record of the Western Regions
^ Herrmann 1925, p. 575–576
^ a b Markwart 1938, p. 93

Sources

Zhou Shu (Book of Zhou), juan 50
Sui Shu (Book of Sui), juan 83, 84
Tabari
Menander Protector
Wei Shu, juan 102
Liang Shu, juan 54
Record of the Western Regions
Tongdian, juan 193 (after Sui Shu and Zhou Shu)

Literature

Peter B. Golden, An Introduction to the History of the Turkic Peoples (1992)
Emre-E. Yavuz, The Gokturk Empire: 2nd Edition (2025)
Aydogdy Kurbanov, The Hephthalites: Archaeological and Historical Analysis (2010)
Étienne de la Vaissière, Sogdian Traders: A History (2005)
Szabolcs Felföldi, A prominent Hephthalite: Katulph and the fall of the Hephthalite Empire (2001)
Uchida Ginpū 内田吟風, 魏書西域傳原文考釋（中） (1971)
Uchida Ginpū 内田吟風, 魏書西域傳原文考釋（下） (1972)
Etsuko Kageyama, The Winged Crown and the Triple-crescent Crown in the Sogdian Funerary Monuments from China: Their Relation to the Hephthalite Occupation of Central Asia (2007)
Josef Markwart, Wehrot und Arang. Untersuchungen zur mythischen und geschichtlichen Landeskunde von Ostiran (1938)
Albert Herrmann, Die Hephthaliten und ihre Beziehungen zu China (1925)
Joachim Werner, Beiträge zur Archäologie des Attila-Reiches (1956)
Édouard Chavannes, Documents sur les Tou-Kiue (Turcs) occidentaux (1903)