Battle of Talas

In 751 an army of Tang China, commanded by the general Gao Xianzhi, fought an army of the Abbasid Caliphate near the Talas River in Central Asia. After several days of fighting, the Karluks, a Turkic people who had been fighting on the Tang side, went over to the Abbasids or attacked the Tang from the rear. The Tang army collapsed; Gao Xianzhi escaped with a small part of his force.

The battle is famous because it is the only large battle in history between the Chinese empire and an Arab–Islamic empire. It took place at the meeting point of two expanding worlds, on the routes of the Silk Road.

Its importance is still argued about. Some writers have called it a turning point that decided whether Central Asia would become Chinese or Muslim; others see it as a frontier battle with limited consequences. The two sides did not continue to fight, and within a year they were exchanging ambassadors again.

In the first half of the eighth century Tang China was at the height of its power. From its Protectorate General to Pacify the West (Anxi), based at Kucha in the Tarim Basin, it controlled or influenced oasis states far to the west. Its main rival in the region was the Tibetan Empire, which was competing with Tang for control of the Pamir mountains and the routes to India and Central Asia.

Meanwhile, Arab armies had conquered Iran and, in the early eighth century, the lands between the Amu Darya and Syr Darya rivers, which the Arabs called Ma wara’ al-nahr (Transoxiana). This was the home of the Sogdians, famous merchants of cities such as Samarkand and Bukhara. Local rulers sometimes appealed to China for help against the Arabs.

In 750 the Umayyad Caliphate was overthrown by the Abbasid revolution, which had begun in Khurasan in north-east Iran. Its military leader there, Abu Muslim, controlled the eastern provinces. The new dynasty needed to secure its eastern frontier, just as Tang commanders were pushing further west.

The Tang commander, Gao Xianzhi (Korean Go Seonji), was the son of a Goguryeo family that had moved to China after Tang and Silla destroyed the Korean kingdom of Goguryeo in 668. His father also served as a Tang officer in the west. Gao rose through the army of the western protectorate.

In 747 he led an expedition across the Pamirs, among the highest mountains in the world, and defeated Lesser Bolü (in the Gilgit region of modern Pakistan), a state allied with Tibet. The crossing of high passes with an army was an astonishing feat; the explorer Aurel Stein, who studied his route, praised it highly. After this, many small states in the region accepted Tang overlordship.

Gao later became military governor at Kucha. His career ended tragically: during the An Lushan Rebellion, after a retreat, he and fellow general Feng Changqing were accused by a eunuch overseer and executed in early 756.

The immediate cause lay in a conflict between two Central Asian states. Ferghana, a Tang ally, was in dispute with Chach (Chinese Shi, the region of modern Tashkent). In 750 Gao Xianzhi marched against Chach. According to Chinese sources, the king of Chach submitted, but Gao broke his promise of safety, plundered the city and took the king to Chang’an, where he was executed. Even Chinese historians condemned this as greed and bad faith.

The king’s son escaped and turned for help to the neighbouring states and to the Arabs. The Abbasid authorities in Khurasan, under Abu Muslim, sent an army under Ziyad ibn Salih. Chinese sources say that when Gao heard that the Arabs were planning to attack the Tang’s western garrisons, he decided to strike first and marched far to the west with Tang soldiers and allied troops.

Seen from each side, the war therefore looked different. To Tang, it was a campaign to punish disloyal vassals and pre-empt an Arab threat. To the Abbasids and many Sogdian and Turkic rulers, it was a response to Chinese aggression and treachery in Chach, and a chance to secure the new dynasty’s eastern border.

Gao’s army advanced about 700 Chinese li (roughly 300–400 km) beyond the Tang frontier and met the Abbasid army near Talas, probably in July 751. Chinese sources speak of about 30,000 Tang and non-Chinese troops; Arabic sources give much larger figures for the Chinese side. Modern estimates for both armies vary widely, and none can be confirmed.

According to the fullest accounts, the armies fought for about five days without a decision. Tang crossbowmen are said to have been effective at first. Then the Karluks, a Turkic confederation serving with the Tang, changed sides, attacking the Tang army while the Abbasids pressed from the front. The Tang army broke.

In the retreat the road was blocked by fleeing Ferghanese troops and baggage. The Tang officer Li Siye cleared a path with a club so that Gao could escape; according to the Zizhi Tongjian, another officer, Duan Xiushi, rebuked him for saving himself and abandoning the troops, and Li then helped gather the survivors. Only a few thousand men returned. The rest were killed or captured.

Large numbers of Tang soldiers were taken prisoner; Arabic sources give figures in the tens of thousands, which are probably exaggerated. Many were taken west into the Abbasid lands. Some were craftsmen, and later accounts say they worked as weavers, painters and papermakers.

The defeat did not lead to a long war. The Abbasids did not invade the Tarim Basin, and Tang remained active in the region for a few more years. Chinese records note Arab embassies to the Tang court from 752 onward. Indeed, during the An Lushan Rebellion a few years later, Arab troops were among the foreign soldiers who helped Emperor Suzong recapture the capitals.

The real turning point for Tang power in Central Asia came in 755, when the An Lushan Rebellion forced the court to recall its best troops from the frontier. Tibet then seized the corridor of the Hexi region and cut the Tang off from its western garrisons, which held out in isolation until the late eighth century. It is this, more than Talas itself, that ended Tang dominance in the west.

Talas is often described as a contest between two empires, but most of the people who fought and suffered there were neither Han Chinese nor Arabs. The Tang army in the west relied heavily on non-Chinese soldiers: Turkic cavalry, troops from oasis states and allies such as Ferghana. The Abbasid forces in Khurasan included Persians, Arabs settled in the east and local Central Asian allies. The Karluks, whose change of sides decided the battle, were pursuing their own interests as a rising Turkic power between the empires.

For ordinary soldiers the campaign meant marching hundreds of kilometres across deserts and mountains, far from home. The poet Du Fu, writing in Chang’an around this time, described in “Song of the Wagons” the grief of families whose sons were sent to endless frontier wars, and the bones lying unburied beside Kokonor lake. Talas was one of several heavy defeats in 751, and Chinese sources record the losses in tens of thousands.

The prisoners paid a lasting price. Taken thousands of kilometres from home, most disappear from the record. Only a few, like Du Huan and the craftsmen he mentions, are known by name, and only Du Huan is known to have returned. For the people of Chach, whose king had been executed and whose city had been plundered in 750, the war began as a story of betrayal by the Tang; their suffering is also part of the history of Talas.

One of the prisoners was Du Huan, a young man from Chang’an and a kinsman of the scholar-official Du You. He spent more than ten years in the Abbasid lands, travelling as far as Kufa in Iraq, then the Abbasid capital, and possibly further. In 762 he returned to China by sea, landing at Guangzhou.

He wrote an account of his travels, the Jingxingji (“Record of Travels”). The full book is lost, but Du You quoted parts of it in his encyclopedia, the Tongdian (801). The surviving fragments describe the customs of the Arabs, their dress, food and prayers, the Islamic ban on wine and on eating pork, the Friday gatherings and the caliph’s preaching. They also name Chinese painters and silk weavers working in Kufa, prisoners taken in 751.

Du Huan’s text is one of the earliest Chinese eyewitness descriptions of Islam and of the Abbasid world, and his generally admiring tone has led scholars such as Angela Schottenhammer to suggest that he enjoyed considerable freedom during his years there.

A famous story links the battle to one of the great technology transfers in history. The eleventh-century Arabic writer al-Tha’alibi, quoting an earlier source, said that Chinese prisoners taken by Ziyad ibn Salih introduced papermaking to Samarkand, which then became famous for its paper. Many popular histories repeat that the battle brought paper to the Islamic world and eventually to Europe.

Historians now treat this story with caution. Paper was already in use in Central Asia long before 751; the Sogdian “Ancient Letters” from the fourth century, found near Dunhuang, were written on paper, and there is evidence that paper was used, and probably made, in Samarkand decades before the battle. Jonathan Bloom, in Paper Before Print (2001), argues that the Talas story is best understood as a convention crediting Chinese craftsmen, rather than an exact record of how the technique arrived.

What is clear is that papermaking spread through the Islamic world in the late eighth century: paper mills are recorded in Baghdad in the 790s, and paper transformed Islamic administration, scholarship and book culture. Whether some Talas prisoners contributed is possible, but the battle was not the single cause.

The Karluks were a confederation of Turkic tribes whose early homeland lay around the Black Irtysh river and the Tarbagatai mountains, in the borderlands of modern Kazakhstan, China and Mongolia. In the 740s they joined the Uyghurs and Basmyls in overthrowing the Second Turkic Khaganate, and they were becoming a powerful force in the western steppe.

Why they turned against the Tang at Talas is unclear. Some Arabic writers suggest they had already been in contact with the Abbasids; modern historians often see their action as an opportunistic choice by a rising power that did not want either empire to dominate the region. The Chinese sources simply record their defection as the decisive moment.

After 751 the Karluks expanded westward; by 766 they had taken Suyab in the Chu valley, near the area where the battle was fought, and made it their capital. The Arabic historian al-Ya’qubi reports that a Karluk ruler accepted Islam under the caliph al-Mahdi (775–785), but the conversion of the Karluk people was gradual. They are often linked to the rise of the Karakhanid state, the first Turkic dynasty to adopt Islam, in the tenth century, although the Karakhanids’ origins are debated. This long-term process, rather than the battle itself, is what made Islam the dominant religion of the region.

The Russian orientalist V. V. Barthold argued that the battle decided which of the two civilizations, Chinese or Muslim, would predominate in Central Asia. This view was influential in the twentieth century and is still repeated in popular works, which often add the claim that the battle led to the Islamization of the Turks.

Many later historians have been more cautious. They point out that the Abbasids made no attempt to exploit the victory eastward; that Tang activity in the western regions continued until the An Lushan Rebellion; that Tibetan power, not Arab power, filled the vacuum left by Tang; and that the conversion of the Turkic peoples of Central Asia to Islam was a slow process that took place mainly in the tenth century and later, through trade, missionaries and the Samanid and Karakhanid states. On this reading, Talas was a significant frontier defeat, but its long-term effects have been exaggerated.

Perspectives also differ by region. In Chinese historiography it has usually been treated as one of several frontier defeats of 751, alongside losses to Nanzhao in the south-west and to the Khitan in the north-east, and as a sign of over-extension before the rebellion. Arabic chroniclers gave it little attention at the time. In Kazakhstan and Kyrgyzstan it is sometimes presented as part of national history on local soil, while in Korea interest has focused on Gao Xianzhi as a general of Goguryeo descent.

The Chinese sources are the fullest. The Tongdian of Du You (801) preserves Du Huan’s eyewitness fragments; the biographies of Gao Xianzhi and Li Siye in the Old Book of Tang (945) and New Book of Tang (1060) and the narrative of Sima Guang’s Zizhi Tongjian (1084) describe the campaign, the Chach affair and the retreat. They are written from the Tang court’s viewpoint, and they sometimes blame Gao’s greed for the defeat.

On the Arabic side, the battle is barely mentioned in the surviving early chronicles. The fuller Arabic accounts come from much later writers such as Ibn al-Athir (d. 1233) and al-Dhahabi (d. 1348), who give large figures for the Chinese dead and captured. Al-Tha’alibi’s eleventh-century paper story is a separate tradition. These later texts depend on sources that are now lost and must be used with caution.

Because the two traditions rarely mention the same details, historians have to combine them carefully, and numbers of troops and casualties remain very uncertain. The exact site of the battle has not been identified archaeologically; it is usually placed near modern Taraz in Kazakhstan or Talas in Kyrgyzstan. Future archaeological work in the Talas valley may help, but no battlefield remains have so far been securely linked to 751.