An empire with a nomad problem
The overland routes took shape in the second and first centuries BC, when two great powers sat at either end of Eurasia without knowing much about each other: the Han dynasty in China (206 BC to AD 220) and the expanding Roman state in the Mediterranean. Neither set out to build a trade route. The Han, in particular, were driven by war.
On China's northern frontier lived the Xiongnu, a confederation of horse riding nomads who raided farming districts and extracted tribute from the Han court. Emperor Wu, who reigned from 141 to 87 BC, decided to fight rather than pay. He sent the envoy Zhang Qian westward around 138 BC to look for partners. Zhang Qian spent about ten years as a captive of the Xiongnu, escaped, and returned some thirteen years after he set out with reports of kingdoms no Chinese official had seen: Ferghana, Sogdiana, Bactria and, by hearsay, India and Parthia.
Heavenly horses
Among Zhang Qian's reports, one detail caught the emperor's attention. The Ferghana Valley, which the Chinese called Dayuan, bred tall, fast horses, far better than the small steppe ponies China could obtain. Emperor Wu wanted these "heavenly horses" for his cavalry, since fighting nomads required horses that could match theirs. When Dayuan refused to trade and killed a Han envoy, Wu sent an army under the general Li Guangli. The first expedition failed in the desert; the second, in a war lasting from 104 to 101 BC, besieged Dayuan's capital until the locals handed over thousands of horses. The episode shows that the Silk Roads began less as a merchant's dream than as an arms race.
Holding the corridor
To protect the route, the Han pushed walls and watchtowers west through the Hexi Corridor and founded military commanderies, including Dunhuang. Two frontier gates, the Jade Gate and the Yang Pass, marked the edge of Chinese control. In 60 BC the court created the Protectorate of the Western Regions, a Han office that oversaw the oasis states of the Tarim Basin, collected hostages and tribute, and kept garrisons on farmland of their own.
The bureaucracy left a paper trail, or rather a wood trail. At Xuanquan, a postal station near Dunhuang, archaeologists found tens of thousands of inscribed wooden slips. They record meals served to passing envoys, horses lent and returned, and disputes. In one, envoys from Kangju, a state in the Sogdian region, protested that local officials had undervalued the camels they brought, judging fat animals to be lean and paying less. It is a small, human complaint that shows foreign visitors arriving under official escort and bargaining with Han clerks.
Rome's taste for silk
At the other end of the network, Romans knew silk long before they knew where it came from. They called the distant producers the Seres, meaning roughly "silk people", and some writers believed silk was combed from the leaves of trees. By the first century AD silk was a fashionable luxury, and moralists attacked it as immodest and wasteful. The naturalist Pliny the Elder, writing around AD 77, complained that luxury imports drained Rome's money to the east. In a nineteenth century translation he says that "India, the Seres, and the Arabian Peninsula, withdraw from our empire one hundred millions of sesterces every year: so dearly do we pay for our luxury and our women."
Between Rome and China lay powers that profited from the gap. The Parthians, who ruled Iran and Mesopotamia, controlled the overland routes in the middle and acted as middlemen. In AD 97 a Chinese officer named Gan Ying was sent west by the general Ban Chao to reach Da Qin, the Chinese name for Rome. He reached a great sea, probably the Persian Gulf, where Parthian sailors warned him the voyage could take years, and he turned back. Historians often suspect the Parthians had good commercial reasons to keep the two ends apart. Further east, the Kushan empire, which ruled Bactria and northern India from about the first to the third century AD, linked the routes running north to Central Asia and south to Indian ports.
Did Rome and China ever meet?
Barely. Chinese annals record that in AD 166 a party arrived by sea in what is now Vietnam claiming to represent the king of Da Qin, whose name resembles that of the emperor Marcus Aurelius Antoninus. Their gifts were ivory and tortoiseshell, which looked to Chinese eyes like southern goods, so historians suspect they were enterprising merchants rather than official envoys.
Worked example: testing Pliny's complaint
Should we accept Pliny's figure of one hundred million sesterces a year? Start by asking what kind of source it is. Pliny was a moralist as well as a collector of facts, and complaints about luxury and women's spending were a familiar Roman theme, so his number may be rhetoric rather than an audit. He also lumps India and Arabia together with the Seres, so even if the figure is right, most of the money may not have gone to China.
Next, look for independent evidence. Large hoards of Roman gold and silver coins have been found in southern India, which supports heavy trade with India. Very few Roman coins turn up in China, which fits the picture of relay trade through Parthian and Kushan hands. The lesson is to treat a vivid number as a clue to what people worried about, then test it against objects in the ground.