The Göktürk khans ruled a belt of grassland and oases that reached from Mongolia to the Oxus. Across it the roads ran from spring to spring, and around it from oasis to oasis. The sources describe these roads in three ways. Some name the stops, some count the days, and most give a sum of li.

The names of the roads

The name Silk Road is modern. It seems to have been used first by the geographer Ferdinand von Richthofen, for the roads of Central Asia between 114 BC and AD 127.[1] The sources of the khaganate period name a road by its direction or by the people who used it. The Han Shu counts two roads out of the Jade Gate and Yangguan, a southern and a northern one.[2] Pei Ju counts three roads from Dunhuang.[3] In 647 the Tang court gave one road the name Road for Paying Court to the Heavenly Khan.[4] In 843 Li Deyu spoke of the Uyghur road, meaning the way to the far west through Uyghur land.[5] Menander names one road in the Caucasus, the road called Dareine.[6]

The steppe road north of the Tian Shan

In 607 Pei Ju described the northern road in the preface of his Illustrated Record of the Western Regions.[3] It started at Yiwu and ran by the Pulei Sea. It passed through the Tiele tribes and the court of the Göktürk khan. After the court it crossed the Beiliu river and reached the land of Fulin and the western sea.[3] Yiwu is identified as Hami, and the Pulei Sea as the lake of Barkol.[7]

Yiwu was a town of mixed peoples. The Xin Tang Shu says that merchants of the hu lived there in the mixed population. It had a thousand soldiers and was tied to the Tiele.[8] At the end of the Sui the town submitted to China. When the empire fell into disorder, it served the Göktürks again.[8] In 630 its lord came to the Tang court.[8] After the fall of Ilig Khan he gave up his seven towns, and the Tang made them the prefecture of West Yi.[8]

Fulin lay at the far end of the road. The Xin Tang Shu calls it the old Daqin on the western sea and puts it 40,000 li from the capital.[8]

Pei Ju’s preface is a plea for war. He wrote that the Göktürks and the Tuyuhun divided the lands of the Qiang and the hu between them and blocked the way, so that the tribute of the western states could not come through.[3] The complaint shows who held the road. The khan’s court stood on it, and the Göktürks were able to close it.

For the western khans the Xin Tang Shu and the Jiu Tang Shu give a count of days. From Yanqi, seven days’ march to the north-west reaches the Southern Court, and eight days’ march due north reaches the Northern Court.[9] The two courts were camps of the western khans. The name Northern Court was later carried also by the Tang protectorate at Beiting.[10]

Over the mountains to Suyab

In 629 or 630 the pilgrim Xuanzang took another way west.[11] He went from the oases of the Tarim over the Tian Shan into the Göktürk lands beyond, and he gave every stage in li.[11]

From Agni, which the older books call Yanqi, he went more than 200 li south-west over a small hill and across two great rivers.[11] Then he went more than 700 li west over a plain and came to Kucha.[11] From Kucha it was more than 600 li west across a small sand desert to Baluka, the old Gumo.[12] From Baluka it was more than 300 li north-west over a stony desert to Ling Shan, the northern plateau of the Pamirs.[11]

Snow lay in the valleys of Ling Shan in spring and summer. Xuanzang reports that travellers on this road were forbidden to wear red-brown clothes, to carry gourds or to shout, because violent storms struck anyone who broke the rule.[11] After 400 li of mountain road he came to the Great Clear Lake, which he also calls the Hot Sea and the Salt Sea.[11] It is the lake of Issyk-Kul. Travellers prayed at the lake for good fortune, and nobody fished in it.[11] Which pass he crossed to reach it is disputed.

From the lake it was more than 500 li north-west to Suyab.[11] Merchants of many countries lived mixed there.[11] West of Suyab stood several dozen lone towns, each with its own head. The heads did not obey one another, but all the towns served the Göktürks.[11]

Another 400 li west lay the Thousand Springs, where the Göktürk khan came every summer to escape the heat.[11] The deer there wore bells and let men come near. The khan had ordered that anyone who killed one be put to death.[11] A further 140 or 150 li to the west lay Talas, where merchants of many countries lived mixed as well.[11] Ten-odd li south of Talas stood a small town of about three hundred households. Its people were Chinese whom the Göktürks had carried off. They had banded together there, and in dwellings and dress they followed the Göktürks, but they kept their own speech and bearing.[11]

From Talas the road turned south-west. It was more than 200 li to White Water, the same again to Gongyu, and then 40 or 50 li south to Nujikan.[11] From Nujikan Zheshi lay 200 li to the west.[11] Zheshi is the country that the Tang called Shi, and Shi is Chach.

The Xin Tang Shu repeats these stages with small changes. It gives a little over 100 li from the Thousand Springs to Talas, and it takes Ling Shan to Suyab in one stage of 500 li, without the lake.[8] Since the Xin Tang Shu follows Xuanzang, the two count as one witness.

A li is not a fixed length. It is closer to a walking hour, longer on a plain and shorter in the mountains.[13] The li of Xuanzang vary between about 270 and 470 metres, so his sums do not fix a place.[14]

The oasis roads of the Tarim

The Han Shu says that two roads left the Jade Gate and Yangguan for the Western Regions.[2] The southern road went from Shanshan along the north side of the southern mountains and followed the river west to Suoju. Beyond the Pamirs it led to the Great Yuezhi and Parthia.[2] Suoju is the oasis of Yarkand.[15] The northern road went from the court of the Former Jushi along the northern mountains and followed the river west to Shule. Beyond the Pamirs it led to Dayuan, Kangju and Yancai.[2] Shule is the oasis of Kashgar.[15]

Modern writers add a third road. It crossed the desert past Loulan to Kucha and was the shortest of the three. They call it the middle road, after the Weilüe.[16] By modern reckoning it measured 1,700 kilometres from Yangguan to Kashgar, seven weeks of travel. The southern road with the stretch from Yarkand to Kashgar measured 1,900 kilometres, eight weeks. The northern road measured 2,100 kilometres, nine weeks.[17]

Pei Ju drew the same roads for the Sui, with other names.[3] His middle road ran from Gaochang through Yanqi, Kucha and Shule, over the Pamirs, and on to Persia.[3] His southern road ran from Shanshan through Khotan and over the Pamirs to Tokharistan and beyond.[3] His middle road therefore follows the northern road of the Han Shu, and his southern road follows its southern road. Pei Ju called Yiwu, Gaochang and Shanshan the gates of the Western Regions, and Dunhuang, where all the roads met, the throat.[3]

The oasis kings sat on these roads under Göktürk overlordship, which the khans had won in the expansion along the Silk Road. The king of Shule, for example, had been given a Göktürk wife.[8] In 640 the Tang took Gaochang and laid the Anxi protectorate over the oasis roads.[18] After the conquest Wei Zheng argued against making the country a prefecture. He said that the king of Gaochang had plundered the merchants and blocked tribute, and that the king had been punished for it.[8] In 719 the court ordered Yanqi, Kucha, Shule and Khotan to levy duties on the merchants of the Western Regions and to live on them.[8] Merchants who came by the northern road were taxed at Luntai.[8]

The road of Zemarchos

The road of Zemarchos shows the western end of the steppe in 569 to 571.[19] Zemarchos left Constantinople with Maniakh and reached the land of the Sogdians after a journey of many days.[20] From there the Göktürks led him to the camp of Istemi on Ektag.[20]

On the way home he travelled with a Göktürk embassy under Tagma Tarkhan.[20] The party crossed the river Oekh and came, after a long way, to an enormous lake.[21] The lake has been taken both for the Aral and for the Caspian.[22] Zemarchos followed its sandy shore for twelve days. He then came to the Ikh, the Daikh and the Attila.[21] The three rivers are identified as the Emba, the Ural and the Volga.[23]

Next the embassy came to the Ugurs, who held their country for Istemi, and then to the Alans.[24] The Alan ruler warned that the Persians lay in wait on the road through Miusimia, and he advised the road called Dareine.[6] Zemarchos sent ten porters with silk along the road through Miusimia, so that the Persians would think the silk had gone ahead.[6] He took Dareine himself, came to the land of the Apsilii and reached the Black Sea.[6] From there he went by ship to the Phasis and to Trapezus, and he rode the public post to Constantinople.[6] The embassy of Zemarchos tells the rest of the journey.

The roads of the Uyghurs

The first road of the Uyghurs to Chang’an was a road they asked for. At the banquet of the Wine Fountain in 647 the Tiele chiefs asked the emperor for a way to his court.[25] The Zizhi tongjian puts the road south of the Uyghurs and north of the Göktürks.[4] The Xin Tang Shu puts it through the lands of the Uyghurs and the Göktürks.[26] The emperor ordered sixty-eight post stations south of the desert, near the Pelican Spring. They were stocked with horses, drink and meat for the envoys.[25] The tribes were to pay a yearly tax in sable skins.[25]

The Xin Tang Shu puts the Uyghur homeland 7,000 li from the capital. The Tang Huiyao says 6,900.[27]

The Uyghurs met the Chinese at springs on the northern frontier. In 757 the court granted them 20,000 bolts of silk a year, which an Uyghur embassy was to collect at the army of Shuofang.[28] Their horses went south to the capital, up to tens of thousands at a time, in the horse-silk trade.[29] In 813 several thousand Uyghur horsemen rode up to the Pelican Spring, and the frontier army was put on guard.[29] In 821 Princess Taihe went north to marry the Uyghur khan.[29] The frontier command reported that 760 Uyghurs with camels, horses and carts were arriving one after another at the Huanglu spring to meet her.[29]

The same road served the oases of the west. In the Kaiyuan years the Uyghurs killed Wang Junchuo, the governor-general of Liangzhou. They thereby cut the road by which the states of Anxi sent their tribute to Chang’an.[28]

Later the Hexi corridor passed to the Tibetans, and the way to the western protectorates led through Uyghur land.[5] The Jiu Tang Shu says that Beiting and Anxi sent their reports to court by way of the Uyghurs, and so came to depend on them.[29] The Uyghurs demanded without limit. Beiting lay nearer, and they took from it whatever it needed to live.[29] In 790 the Tibetans took the Beiting protectorate.[29] After that Anxi was cut off, and nobody knew whether it still stood. Only the people of Xizhou held out.[29]

In the second month of 843 Zhao Fan reported that the Kirghiz were attacking the protectorates of Anxi and Beiting and that troops should be sent.[5] Li Deyu answered. He wrote that Anxi lay 7,100 li from the capital and Beiting 5,200.[5] In peacetime, he said, the road west ran from Hexi and Longyou out through the Yumen pass, past one prefecture after another, each with a strong garrison.[5] After the troubles Hexi and Longyou had fallen to the Tibetans. Anyone who wished to reach Anxi and Beiting then had to take the Uyghur road.[5] The Uyghurs were now broken, and nobody knew whether the road belonged to the Kirghiz.[5] Even a successful relief, he added, would leave a garrison of ten thousand men at each place, and no road to supply them.[5]

On the way to the Uyghur capital the Arab envoy Tamim ibn Bahr counted twenty days through steppe with springs and grass but no villages. Only the men of the relay service lived there, in tents. He then travelled twenty days among villages lying close together before he reached the king’s town.[30]

The road to the Volga Bulgars

Ibn Faḍlān travelled in 921 and 922 from Baghdad to the Volga Bulgars and recorded the stages of his journey.[31] His road is more than three centuries younger than the first Göktürk khaganate. It crosses the steppe of the Oghuz and the Pechenegs, not that of the Göktürks.

The embassy left Baghdad on Thursday 11 Safar 309, which is 21 June 921.[31] It passed Hulwan and Hamadan and waited eleven days at Rayy for Ahmad ibn Ali.[32] From Rayy it went east through Simnan and Damaghan to Nishapur, and on by Sarakhs and Merv to the edge of the desert of Amul.[33] It rested three days there to let the camels recover. Then it crossed the desert to Amul and the Jayhun, and came by Baikand to Bukhara.[34] It stayed twenty-eight days in Bukhara.[35]

The envoys feared that winter would close the road. They went back to the river and hired a boat to Khwarazm, more than 200 farsakhs away.[36] The Khwarazmshah first refused to let them enter the land of the Turks. In the end he gave leave.[37] The embassy went on by water to Jurjaniya, fifty farsakhs, and wintered there.[38] The river froze for three months, and the ice was seventeen spans thick.[39] In mid-Shawwal, 16 February 922, the thaw began.[40] The embassy bought Turkic camels and had folding boats made of camel skin for the rivers ahead. It laid in bread, millet and salted dried meat for three months.[40]

It left Jurjaniya on Monday 2 Dhu al-Qa’da, 4 March 922.[41] The first stage was the ribat of Zamjan, which Ibn Faḍlān calls the Gate of the Turks.[41] After ten days across a steppe without mountains, the caravan marched each night from the middle of the night to midday or the afternoon. After fifteen such nights it reached a great stony mountain. Beyond the mountain lay the Oghuz.[42]

A Muslim could cross the Oghuz country only with a Turk as his friend. He brought the friend a garment, a veil for his wife, pepper, millet, raisins and nuts. The friend pitched a tent for him and gave him sheep as far as his means allowed.[43] A traveller whose camels or horses were worn out could leave them with his friend, take fresh animals and money, and settle the account on his return.[44] Ibn Faḍlān names the king of the Oghuz Yabghu and one of their commanders, Etrek.[45]

Beyond the Oghuz the embassy crossed the Yaghandi in the folding boats. It crossed the Jam, the Jakhch, the Uzil, the Erden and the Wabna as well. Then it came to the Pechenegs, who camped by a lake, and stayed one day.[46] The next river, the Jaikh, was the greatest the travellers had seen. Several men and animals drowned in it.[47] Seven more rivers lay beyond it before the land of the Bashgirds, against whom Ibn Faḍlān took the greatest precautions.[47] After the Bashgirds the road crossed six more rivers, each roughly two to four days from the last.[48]

The embassy reached the king whom Ibn Faḍlān calls the king of the Bulgars on Sunday 12 Muharram 310, which is 12 May 922.[49] Ibn Faḍlān counts seventy days of travel from Jurjaniya to the king’s country.[49] The kingdom is the Volga Bulgaria of the later sources.

Carriers and cargo

The roads were walked by merchants. The Xin Tang Shu says of Samarkand that its men are skilled at trade and fond of profit. At twenty they leave for neighbouring countries and go wherever there is gain.[8] Pei Ju’s informants were rich merchants who had travelled everywhere.[3] The Xin Tang Shu says that Sogdians, whom it calls the hu of nine names, went with every Uyghur embassy to China.[26] In the early years of Zhenguan the Tang emperor told an envoy from Bukhara that the Western Turks had already surrendered and that merchants and travellers could now go on their way. The various hu were delighted.[8]

The cargo that moved north on these roads was silk. The Göktürks first came to the frontier markets for it, and the Uyghurs later drew it for their horses.[50] Iron, salt, furs, jade and horses also moved on these roads.

Gaps in the evidence

No source used here lists the stops of the steppe road between Yiwu and the khan’s court. The itineraries count in li, in days and in farsakhs, and the units cannot be added together. What the sources give are fragments of six roads. Each fragment was seen by one kind of traveller: an official, a pilgrim, an envoy or a minister.

References

  1. ^ Herrmann 1910, p. 10
  2. ^ a b c d Han Shu, juan 96
  3. ^ a b c d e f g h i Sui Shu (Book of Sui), juan 67
  4. ^ a b Zizhi tongjian (Comprehensive Mirror in Aid of Governance), juan 198
  5. ^ a b c d e f g h Cefu yuangui, juan 994
  6. ^ a b c d e Menander Protector; Blockley 1985, fr. 10.5
  7. ^ Herrmann 1910, p. 75, 81
  8. ^ a b c d e f g h i j k l Xin Tangshu (New Book of Tang), juan 221
  9. ^ Xin Tangshu (New Book of Tang), juan 215; Jiu Tangshu (Old Book of Tang), juan 194
  10. ^ Xin Tangshu (New Book of Tang), juan 215
  11. ^ a b c d e f g h i j k l m n o p q Record of the Western Regions, juan 1
  12. ^ Record of the Western Regions, juan 1; Xin Tangshu (New Book of Tang), juan 221
  13. ^ Herrmann 1910, p. 37
  14. ^ Herrmann 1910, p. 52–53
  15. ^ a b Herrmann 1910, p. 38
  16. ^ Herrmann 1910, p. 106, 115
  17. ^ Herrmann 1910, p. 115
  18. ^ Jiu Tangshu (Old Book of Tang), juan 3; Jiu Tangshu (Old Book of Tang), juan 198
  19. ^ Blockley 1985, p. 263, 267
  20. ^ a b c Menander Protector; Blockley 1985, fr. 10.3
  21. ^ a b Menander Protector; Blockley 1985, fr. 10.4
  22. ^ Blockley 1985, p. 265–266
  23. ^ Blockley 1985, p. 266
  24. ^ Menander Protector; Blockley 1985, fr. 10.4, fr. 10.5
  25. ^ a b c Xin Tangshu (New Book of Tang), juan 217; Zizhi tongjian (Comprehensive Mirror in Aid of Governance), juan 198
  26. ^ a b Xin Tangshu (New Book of Tang), juan 217
  27. ^ Xin Tangshu (New Book of Tang), juan 217; Tang Huiyao (Institutional History of the Tang), juan 98
  28. ^ a b Jiu Tangshu (Old Book of Tang), juan 195; Xin Tangshu (New Book of Tang), juan 217
  29. ^ a b c d e f g h Jiu Tangshu (Old Book of Tang), juan 195
  30. ^ Sinor 1990, p. 337
  31. ^ a b Ibn Fadlan; Canard 1958, p. 51
  32. ^ Ibn Fadlan; Canard 1958, p. 52
  33. ^ Ibn Fadlan; Canard 1958, p. 52–53
  34. ^ Ibn Fadlan; Canard 1958, p. 53–54
  35. ^ Ibn Fadlan; Canard 1958, p. 57
  36. ^ Ibn Fadlan; Canard 1958, p. 58
  37. ^ Ibn Fadlan; Canard 1958, p. 58–59
  38. ^ Ibn Fadlan; Canard 1958, p. 59, 62
  39. ^ Ibn Fadlan; Canard 1958, p. 61
  40. ^ a b Ibn Fadlan; Canard 1958, p. 63
  41. ^ a b Ibn Fadlan; Canard 1958, p. 66
  42. ^ Ibn Fadlan; Canard 1958, p. 66–67
  43. ^ Ibn Fadlan; Canard 1958, p. 70–71
  44. ^ Ibn Fadlan; Canard 1958, p. 71–72
  45. ^ Ibn Fadlan; Canard 1958, p. 76–77
  46. ^ Ibn Fadlan; Canard 1958, p. 81–83
  47. ^ a b Ibn Fadlan; Canard 1958, p. 84
  48. ^ Ibn Fadlan; Canard 1958, p. 86
  49. ^ a b Ibn Fadlan; Canard 1958, p. 87, 89
  50. ^ Zhou Shu (Book of Zhou), juan 50; Jiu Tangshu (Old Book of Tang), juan 195

Sources

Literature

  • Albert Herrmann, Die alten Seidenstrassen zwischen China und Syrien. Beiträge zur alten Geographie Asiens, Bd. 1 (1910)
  • R. C. Blockley, The History of Menander the Guardsman: Introductory Essay, Text, Translation, and Historiographical Notes (1985)
  • M. Canard, La relation du voyage d'Ibn Fadlān chez les Bulgares de la Volga (1958)
  • Denis Sinor (ed.), The Cambridge History of Early Inner Asia (1990)

Related

What links here1

Places
Anxi Protectorate

Cite this article

Yavuz, Emre-E. “Trade Routes of the Steppe.” Turkic Atlas, revised 8 November 2026. https://turkicatlas.org/institutions/trade-routes-of-the-steppe/

Last revised: 8 November 2026