Han silk reached Rome (~50 BCE), and Roman gold drained east

Rome before silk: a textile world without sericum

The Roman world of the late Republic — the world that would, within a century, develop a structural obsession with Chinese silk — clothed itself in wool and linen. Wool was the dominant fibre: Italian sheep grazed on the Apennine uplands and the Po Valley pastures supplied common cloth, while finer wool came from Taranto and Apulia in southern Italy, from Miletus in Asia Minor, and from northern Iberia. Linen came from Egyptian flax fields along the Nile and from the temperate flax-growing zones of Gaul and the Rhineland. Cotton was almost unknown in the Mediterranean; the cotton plant grew in India and was discussed by Greek writers — Theophrastus in the fourth century BCE describes "wool-bearing trees" — but as a textile in elite Roman use it was marginal until the early imperial period and never became dominant.

A Roman senator in 80 BCE wore a wool tunic and a wool toga; his wife wore a linen stola over a wool tunica. The colours were dye-limited and dye-expensive: the deep purple of the senatorial clavus came from murex shells processed at Tyre and a handful of other Phoenician dye-towns at a labour cost that made a single purple-bordered toga the equivalent of months of an artisan's earnings. Bright colours were status markers; undyed wool was for the poor. Within these limits, Roman elite fashion had its own internal hierarchy — the textures of fine-spun Apulian wool, the weave of Coan-island gauze, the cut of an embroidered linen — but the materials available to the elite were locally produced or short-haul imports from within the broader Mediterranean and Near Eastern world.¹

Roman attitudes toward textile spending were already tense before silk arrived. The lex Oppia of 215 BCE, passed during the Second Punic War, had attempted to restrict women's possession of more than half an ounce of gold, ban purple-bordered garments, and prohibit carriages drawn by horses within a Roman mile of the city. The law was repealed in 195 BCE after a famous demonstration of Roman matrons in the Forum, and the senatorial debate preserved by Livy — Cato the Elder against, Lucius Valerius for — established the genre of Roman moralist legislation against luxury textile consumption that would later be turned against silk. The vocabulary the moralists used in 16 CE was inherited from the lex Oppia debate two centuries earlier; the cultural anxiety about cloth as a corrosive of virtue was older than any Chinese fibre.

What Romans already knew about silk: Coa vestis from Cos

The Romans were not entirely innocent of silk before Chinese silk arrived. The Greek-speaking island of Cos, in the Aegean, produced a fabric known to Latin writers as Coa vestis — wild silk spun from the cocoons of Pachypasa otus, a Mediterranean moth, by a domesticated industry that Aristotle in the fourth century BCE attributed to a woman named Pamphile, daughter of Plates, who he says had first carded and spun the cocoons of these moths.² Coan silk was a luxury but not a foreign one. Its threads were coarser than the cultivated silk of the Chinese silkworm, and the woven cloth was more uneven and less lustrous; in elite Roman use it was a less expensive, more available substitute for what would later be called sericum. After the Augustan period the references to Coan silk dwindle in the Roman record, as Chinese silk replaced it almost entirely as the prestige textile.

The Romans also knew, vaguely, that the eastern peoples called Seres — a name derived from the Greek σῆρες, possibly a calque on a Han Chinese word — produced a finer fabric. They did not know how. They believed for a long time that silk was a kind of plant fibre, combed off trees in the eastern lands; Virgil in the Georgics II.121 calls it the vellera depectant Seres — "the Seres comb the fleece" — implying a vegetable origin. The Roman misconception that silk was carded from trees survived in some texts into late antiquity. The Chinese fact — that silk is the protein thread spun by the larvae of Bombyx mori moths fed on mulberry leaves — was a Han state secret backed by capital punishment for export, and it would not be broken until two Nestorian monks smuggled silkworm eggs to Constantinople in 552 CE.

The fiscal context: an empire that paid in metal

The Roman empire of the early imperial period had an economy that ran on precious metal. The denarius — the standard silver coin — circulated across the empire; the aureus — the standard gold coin — funded long-distance trade. The empire mined silver in Iberia at Rio Tinto and at Cartagena, gold in the western Balkans and northwestern Iberia, and accumulated bullion through tribute, taxation, and military extraction across the provinces. The Roman economy was monetised at a high level by ancient-world standards — taxes were paid in coin, soldiers were paid in coin, long-distance trade settled in coin — and this monetisation made the empire structurally vulnerable to any outward flow of bullion that was not returned by an equivalent inflow of value.

When silk began to reach Roman markets in volume in the late first century BCE, it arrived in this context. The Romans had no fabric to send back east that the Han Chinese, Sogdians, or Parthians wanted; they had glass, coral, fine pottery, some bronze and silver objects, and amber, but these were specialty exports of small volume. What they sent east in bulk was money — gold and silver coin, ingots, and worked metalware that could be melted down. This is the structural setup that would, by the time of Pliny the Elder in the 70s CE, produce the famous denunciation of the eastward bullion drain and the moralist literature blaming Roman women's silk dresses for the precious-metal hemorrhage of an empire.

The transmission: a four-stage relay across Eurasia

The Han-Roman silk trade was never a single road and never a continuous caravan. It was a four-stage relay across roughly five thousand miles of mountains, deserts, river valleys, and steppe, with Chinese silk handed off from one set of intermediaries to another at successive transshipment points. No Roman merchant traveled to Chang'an; no Han merchant traveled to Rome. The silk did, and it accumulated transit costs at every stage.³

The Han state's western expansion under Wudi

The corridor existed because the Han state built it for military reasons. From 138 BCE, the emperor Wu (r. 141–87 BCE) — known to later historians as Han Wudi — dispatched the diplomat Zhang Qian on a mission to the Yuezhi, an Indo-European-speaking people who had been pushed westward into Bactria by the Xiongnu nomadic confederation of the Mongolian steppe. Zhang Qian was captured by the Xiongnu and held for ten years; he eventually reached the Yuezhi in Bactria and returned to Chang'an after thirteen years, in 126 BCE, with detailed reports on Central Asian polities — Fergana, Sogdiana, Bactria, Parthia — and on the existence of an organised trading economy stretching west toward the Mediterranean.⁴ Zhang Qian's report did not open the Silk Road in the popular sense; what it did was supply the Han court with strategic intelligence about the lands beyond the Xiongnu, and that intelligence was used to plan the wars that followed.

The Han–Xiongnu wars (133 BCE – 89 CE intermittently, with the most intense phase under Wu) pushed Han military control across the Hexi Corridor — the narrow strip running west between the Tibetan plateau and the Mongolian steppes — and into the Tarim Basin oasis system. By 60 BCE the Han had established the Protectorate of the Western Regions at Wulei, garrisoning the oasis cities of Loulan, Niya, Khotan, Kucha, and Turfan with Han troops and converting them into agricultural and military bases.⁵ The Tarim corridor was an imperial extraction project, not a commercial one — its purpose was to deny the Xiongnu the southern flank of the steppe and to project Han power toward the Pamirs — but once garrisoned and policed, it became the western terminus of a corridor along which Sogdian and Bactrian merchants could move goods further west.

Sogdian and Bactrian intermediaries

The Sogdians — Iranian-speaking inhabitants of the city-states of Samarkand, Bukhara, Panjikent, and the broader Zarafshan valley — were the principal long-distance carriers of silk between the Tarim oases and the Iranian plateau. By the early Han period they had established merchant colonies along the eastern Silk Road; by the late Han period Sogdian was the lingua franca of the trans-Eurasian corridor. Sogdian letters preserved in the Tarim sands — the so-called "Ancient Sogdian Letters" found at Dunhuang, dating to c. 313 CE — document Sogdian merchant networks that ran from Chang'an to the Black Sea.⁶ Bactrians, similarly, operated the segment from the Pamirs south into northern India and west into the Iranian plateau; the Kushan empire (c. 30 CE – 375 CE) that consolidated Bactrian political power in the first three centuries CE was the intermediate political authority over much of this segment.

The merchants moved by camel caravan in convoys of fifty to several hundred animals, with armed guards, taking three to six months for any one leg of the journey. They did not move silk as bulk cargo; they moved it as a high-value-per-weight commodity wrapped in protective coverings, often interlayered with bolts of cheaper local cloth that could be sold along the way. A Han silk bolt weighing perhaps two kilograms might pass through a dozen hands between Chang'an and Antioch, with each intermediary adding a markup of fifty to several hundred percent depending on the local risk and tariff conditions.

Étienne de la Vaissière, in his Histoire des marchands sogdiens (2002), reconstructs the operational structure of Sogdian commerce from the Ancient Sogdian Letters, the Mount Mug archive, and Chinese-source references to Sogdian merchants resident in Han and Tang capitals. Sogdian merchant houses operated as multi-generational lineages with stable partners at each transshipment point — a Bukharan family might have a brother resident in Samarkand, a cousin in Khotan, a son-in-law in Dunhuang, a nephew in Chang'an. Capital flowed through these family networks in the form of credit instruments — promissory notes, partnership contracts, joint-ownership shares of a caravan's load — that allowed a merchant to commit goods to a route without travelling the entire distance himself. By the second century CE the Sogdian network was the most sophisticated trans-Eurasian commercial infrastructure outside the Mediterranean and the Indian Ocean basins. Han Chinese silk moved through it at every stage; so did Indian cotton, Bactrian gold, Tarim jade, Iranian carpets, and the slaves who were also moving along the corridor in both directions.

Parthian middlemen and the information embargo

The largest single intermediary was the Parthian empire (247 BCE – 224 CE), which controlled the Iranian plateau between the Sogdian east and the Roman west. The Parthians charged tariff and tolls on silk passing through their territory and made considerable revenue from the trade. They also took active steps to prevent direct Han–Roman contact. In 97 CE the Han general Ban Chao sent his envoy Gan Ying westward with instructions to reach Daqin — the Han name for the Roman empire. Gan Ying reached the Persian Gulf coast, possibly at Charax Spasinou, where Parthian sailors told him the sea voyage to Daqin took three months with favourable winds, two years with adverse winds, and that "the vast ocean urges men to think of their country, and get homesick, and some of them die." Gan Ying turned back. The Han chronicle the Hou Hanshu, composed in the early fifth century CE from earlier Han records, comments directly: "The king of Anxi [Parthia] wished to control the trade in multi-coloured Chinese silks, and so blocked Ying from reaching Rome."⁷

The Parthians' information embargo was effective. The Romans of the high empire knew of "the Seres" — they had a name for the eastern silk-producers — but they had no accurate geography, no political knowledge of the Han state, and no understanding of how silk was actually made. The Han, conversely, had a name for Rome — Daqin, "Great Qin" — and a generally favourable but vague picture of it as a peer empire ruled by elected officials whose dignity was protected against the wishes of the people. Each empire knew the other existed and knew the other commanded a vast wealth; neither could reach the other without going through middlemen whose business model depended on keeping them apart.

Anonymous Han craftsman. Painted silk funerary banner from Mawangdui Tomb 1, c. 168 BCE. Hunan Provincial Museum, Changsha. Public domain via Wikimedia Commons. · Public domain

Roman entry points: Palmyra, Alexandria, Antioch

Silk entered the Roman empire at four principal points. Palmyra, the Aramaic-speaking caravan city in the Syrian desert, controlled the overland route between the Euphrates valley and the Roman provinces of Syria and Judaea; the famous Palmyra Tariff inscription of 137 CE, a bilingual Greek and Palmyrene Aramaic limestone slab now held at the Hermitage Museum, lists silk among the goods subject to customs duties at the city gates.⁸ Palmyrene merchant colonies operated at Vologesias on the Euphrates and at the head of the Persian Gulf, securing the supply at points outside Roman territory and bringing it into the empire on Palmyrene authority. Antioch, the Roman provincial capital of Syria, was the main inland entrepôt for the Palmyran trade. Alexandria, on the Egyptian coast, received silk arriving by sea through the Red Sea–Indian Ocean route, transshipped from Indian and Arabian ports onto Roman vessels at Berenike and Myos Hormos. The maritime route ran in parallel to the overland route from at least the first century CE; the Periplus Maris Erythraei, a Greek mariner's manual of c. 50 CE, lists "Chinese silk yarn, cloth, and floss" among the goods available at the port of Barbarikon at the mouth of the Indus.⁹ Finally, the Black Sea ports — Phasis, Trapezus, and the Crimean cities — received silk through the Caucasian routes.

The cumulative effect of the four-stage relay was that a bolt of Han silk that left Chang'an worth one Han unit of value sold in Rome for between fifty and one hundred times that. The intermediaries — Sogdian, Bactrian, Parthian, Palmyran — captured the great bulk of the markup. Han merchants received Han prices in Chang'an; Roman buyers paid Roman prices in Rome; the middle of the chain absorbed most of the gold.

Photograph by Lord Pheasant. Roman fresco from the Villa of the Mysteries, Pompeii, c. 60 BCE. Pompeii archaeological site, Italy. Public domain via Wikimedia Commons. · Public domain

What silk replaced: an elite consumption revolution

Silk arrived in Roman elite consumption in volume sometime in the late first century BCE. The earliest secure archaeological evidence of Chinese silk in the Roman world is a small set of fibres found at Pompeii, Herculaneum, and other Vesuvian sites destroyed in 79 CE; chemical analysis has identified some of these as Bombyx mori mulberry silk rather than the local Pachypasa otus wild silk of Coa vestis. By the Augustan period, silk garments appear in the literary record as a known elite consumption item; by the reign of Tiberius (14–37 CE) silk had become enough of a public moral question to provoke senatorial legislation.

Silk in the Augustan and Julio-Claudian courts

The earliest named consumer in the literary record is Augustus's only daughter Julia, who according to a tradition reported by Suetonius wore silk garments at the court of her father; Augustus is supposed to have rebuked her for the impropriety. Caligula (r. 37–41 CE) appeared in public wearing silk; Nero (r. 54–68 CE) made silk a feature of his court ceremonial. By Domitian's reign (81–96 CE), silk was being worn by elite women across the senatorial order and the broader equestrian class, and the imperial domus reportedly imported silk for ceremonial occasions in quantities that contemporaries found alarming.¹⁰

The garments produced from imported silk were not made up in China. Han Chinese silk arrived in the Roman world primarily as raw silk yarn or as woven bolts of plain silk; Roman weavers, particularly in Syria (Tyre, Berytus) and in the Greek east (Smyrna, Antioch), re-wove the silk into Roman-style garments, often unraveling the heavier Han weaves and re-spinning the thread to produce the gossamer-light transparent fabrics that became the signature of imperial-period silk consumption. This re-working was where much of the price was added on the Roman side, and where the most morally contentious garments were produced: the "silken air" that scandalised the senators was not Chinese textile work but Roman.

Failed sumptuary law: Tiberius 16 CE

In 16 CE the Roman senate, with Tiberius's support, passed a senatusconsultum that, among other restrictions on luxury, forbade men from wearing silk garments. The law's specific target, recorded by Tacitus in Annals II.33, was ne vestis serica viros foedaret — "that silken cloth should not dishonour men" — and the legislation reflected an ideological position that silk garments, by clinging to the body and revealing its contours, were appropriate at best for women and at worst for no one.¹¹ The senators forbade themselves and the equestrian order from wearing silk on civic occasions.

The law failed completely. Tacitus notes its passage and Cassius Dio mentions it; neither historian records any prosecution under it, and the literary record from the next century is full of male silk-wearing. By the time of Pliny the Elder in the 70s CE, the silk trade was substantially larger than it had been under Tiberius, and Pliny's denunciation in the Natural History assumes a Roman audience saturated with silk goods. Successive emperors made gestures against silk consumption — Aurelian in the third century reportedly refused his wife a silk garment on cost grounds — but no Roman regime succeeded in suppressing the trade. The structural drivers were too strong: elite status was bound up with conspicuous consumption, conspicuous consumption gravitated to the rarest available goods, and silk was the rarest of the textiles within reach.

The moralist denunciations

The literary denunciation of silk is the most articulate evidence we have for the trade's social impact. Pliny the Elder, writing in the 70s CE, opens Natural History XII.41 with the famous accounting: "India, the Seres, and that peninsula [Arabia] together drain our empire of one hundred million sesterces every year, by the most modest computation. That is the price our luxuries and our women cost us."¹²

Pliny's figure is debated; modern scholars treat 100 million sesterces as the upper end of a plausible range and note that he was a moralist as well as a natural historian. McLaughlin, reviewing the numismatic and archaeological evidence for the eastward bullion flow, argues that the order of magnitude is supportable: the Roman empire of the high imperial period had a total annual budget of around 800–1,000 million sesterces, of which perhaps 10–15% may have flowed east through luxury trade, with silk a major component.¹³ Whatever the precise figure, Pliny's denunciation was a political intervention as much as an economic measurement — he was arguing that the empire was bleeding gold for women's clothes, and he was naming silk as the centre of the problem.

Seneca, in passages of his philosophical and rhetorical works, denounced silk in even more pointed terms: "I see silken clothes, if one can call them clothes at all, in which nothing serves to defend either the body or, in the end, the modesty of the wearer." The transparency of the woven silk, the way the fabric clung to the body, the visibility of the female form through the cloth — this was the specific moral object of Seneca's denunciation, and the imagery would echo across Roman moralist writing into late antiquity.¹⁴

Tacitus's Annals preserves the senatorial debates over silk legislation in a voice that is itself half-moralist: he describes the silk legislation as part of a broader moral panic over imported luxury and notes its complete inefficacy. Cassius Dio, two generations later, repeats the same pattern: laws against silk, no enforcement, a continuously expanding trade. The Roman moralist denunciation of silk was a literary genre — a stable feature of imperial-period intellectual production — for at least three centuries.

What silk displaced

The arrival of Chinese silk displaced several existing prestige textiles in Roman elite consumption. Coan silk — the Mediterranean wild silk of the Greek island of Cos — disappears from the elite record after the Augustan period; references in literature dwindle, archaeological finds become rare, and the industry on Cos itself contracts. The fine wool of Taranto and Apulia retained its market but was displaced from the highest-status garments. Embroidered linen, dyed in purple or saffron, retained ceremonial uses but lost its position as the most desired prestige fabric. The murex-purple industries of Tyre and Sidon, which had supplied the most expensive non-silk textiles in the Mediterranean, continued operating — Tyrian purple was complementary to silk rather than displaced by it — but the highest-status garments of the imperial court were now silk dyed in Tyrian purple, a combination that priced ordinary purple wool out of the elite market.

The chain continued downstream. Roman tastes for silk shaped the products that Sasanian and then Byzantine workshops would produce after the Persian and Byzantine empires acquired sericulture in the late antique period. The Tang Chinese silk industry, the Byzantine silk industry of the sixth century onwards, the medieval Italian and Sicilian silk industries of the twelfth century, the early modern Lyon and Spitalfields silk industries — each was a downstream consequence of the appetite that Han silk arriving in Rome had created. The transmission's persistence is measured not only in centuries but in the structural geography of European luxury textile production for two millennia.

What the cost was

The cost of the Han-Roman silk transmission is not visible in any single named atrocity. There was no sack of silk-bearing cities, no enslavement of silk weavers, no campaign of religious extermination tied to the trade. The cost is in the structural pressures the trade created on both the Han and Roman ends of the corridor, and in the second-order consequences — fiscal, demographic, military — that flowed from those pressures across centuries.

Roman bullion east: the precious-metal drain

The most direct cost was the eastward flow of Roman precious metal. The empire produced silver and gold in significant volume — the Iberian mines alone may have produced 200 tonnes of silver per year at peak in the first century CE — but the silk trade, together with the parallel trades in pearls, spices, and other Eastern luxuries, consumed a substantial fraction of that production. Roman coin hoards in southern India and the Persian Gulf attest to the volume of the flow. The Hazara district of the Punjab has yielded denarii of Augustus and Tiberius; Tamil Nadu has yielded coin hoards numbering in the thousands; the south Indian and Sri Lankan archaeological record shows Roman coin circulation from the first century BCE through the third century CE.¹⁵ These coins were not lost to circulation in the sense of being hoarded by Indians for ornament — though some were — but were transferred east in payment for goods that were then consumed in Rome.

The structural drain compounded across the imperial period. By the third century CE the Roman empire was facing severe fiscal pressure, and the eastern luxury imports were one of multiple drains that contributed. The Crisis of the Third Century (235–284 CE) — a period of monetary debasement, civil war, and provincial breakaway — had multiple causes, but the cumulative effect of two and a half centuries of eastward bullion flow was one of them. The denarius, which had been roughly 95% silver under Augustus, fell to less than 5% silver by the reign of Gallienus (260–268). Some of that debasement was inflationary mismanagement; some was the inability of an empire that had bought silk for two and a half centuries to maintain the silver content of its standard coin.

The numismatic record from south India is the most physically direct evidence for the drain. Pattanam in Kerala, Karur in Tamil Nadu, Pudukkottai, Akkenpalle, Vellalur — the south Indian inland sites of Roman coin recovery — have produced hoards of denarii and aurei that in many cases were intentionally defaced before deposition, the imperial portrait slashed across the face to invalidate the coin for re-export. The defacement was a south Indian practice for handling Roman silver: the metal was kept as bullion or melted into local jewelry, but the political claim of the imperial portrait was struck off so that the coin no longer functioned as Roman political object. Tens of thousands of such defaced denarii have been recovered. They are the physical residue of the silk and pepper trade — the Roman half of the transaction, deposited in south Indian soil where the empire's gold and silver came to rest.

Han fiscal cost: the corridor as imperial extraction project

The corridor that delivered the silk was itself an extraction project the Han had built and maintained at great cost. The Tarim Basin garrisons cost the Han treasury enormously; the wars against the Xiongnu cost hundreds of thousands of lives across the campaigns of Wu and his successors. The Han imperial chronicles record the strain: agricultural colonies (tuntian) were established at the western oases to reduce supply-line costs, but the Han military presence in the Tarim required continuous reinforcement, and the political will to maintain it waxed and waned across the dynasty. The Han elites began questioning the Tarim project as early as the first century BCE; by the late Eastern Han it was being maintained largely by the personal authority of individual military governors like Ban Chao (32–102 CE) rather than by sustained court commitment. The corridor cost paid by the Han fell most heavily on the Xiongnu and Tarim oasis populations, on the conscripted soldiers, and on the convict labour that built and supplied the western garrisons.

The intermediary fattening

Between the senders who paid in conscripted labour and the receivers who paid in bullion, the intermediaries grew rich. Sogdian merchant cities — Samarkand, Bukhara, Panjikent — became wealthy urban civilisations on the back of the silk trade; the lavish wall paintings of Panjikent and the splendour of pre-Islamic Sogdian elite tombs were paid for by silk transit. Palmyra in the second and third centuries CE was one of the richest cities of the Roman east, built on the Eurasian trade — its temple of Bel, its colonnaded streets, its commercial inscriptions document a city whose prosperity was tied to the goods that passed through. The Parthian empire's revenue from silk tariffs was a structural component of its fiscal base.

This was the silk trade's third cost stratum. The intermediary cities prospered for as long as the trade ran through them. When the trade routes shifted — as the maritime route from the Red Sea expanded relative to the overland route in the late imperial period, as Palmyra was sacked by Aurelian in 273 CE, as Sasanian intermediation began to replace Parthian, as the Byzantine empire's acquisition of sericulture in 552 CE cut the long-distance silk-yarn trade — the intermediary cities decayed. Palmyra never recovered after Aurelian; the Sogdian commercial network collapsed under the Arab conquests of the seventh and eighth centuries and the subsequent Turkic-Mongol displacements; the overland Silk Road that had run continuously from the second century BCE was a shadow of its former volume by the year 1000.

The Tarim oasis towns whose surviving silk fragments now sit in the British Museum, the Hermitage, the Indian Museum at Calcutta, and the Otani Collection in Japan provide the third stratum's physical record. Aurel Stein's expeditions to Niya, Loulan, Khotan, and Miran between 1900 and 1930 recovered silk fragments, wooden documents, Buddhist manuscripts, and the residue of a thousand years of Silk Road urban life from sites that had been buried by sand after the trade routes shifted. The silks Stein extracted — Han period brocades with animal patterns and auspicious tokens, Tang period figured silks woven for the Central Asian market, Sogdian-style hunting-scene textiles — document the upstream end of what the Roman empire was buying, and the decline of the corridor when the trade ceased to run.

The secret kept until 552 CE

The technical secret of sericulture remained Han Chinese for nearly seven hundred years after the Roman silk trade began. The Chinese state had executed export of silkworms or mulberry-cultivation knowledge as a capital offence from at least the Han period; the secret leaked first to the Korean and Japanese courts in the Three Kingdoms period, then to the Iranian plateau under Sasanian state sponsorship in the fifth or sixth century, and finally to the Byzantine empire in 552 CE when two Nestorian Christian monks — the names given in Procopius's Wars are unrecorded but their mission is well attested — smuggled silkworm eggs and mulberry-cultivation knowledge from Khotan or Sogdiana to Constantinople, hidden inside hollow bamboo canes. Justinian's establishment of imperial silk workshops in Constantinople, Berytus, and Tyre broke the Eastern monopoly and ended the structural Roman dependence on imported silk yarn.¹⁶

The half-millennium of dependence had reshaped the imperial economy, the Mediterranean fashion order, the urban geography of the Levant and Central Asia, and the diplomatic relationship between successive Mediterranean and East Asian polities. Han Chinese silk did not destroy the Roman empire — many things destroyed the Roman empire — but it was a structural element of the bill the empire paid across centuries, and the moralists who denounced it were not, in their accounting, wrong.

The cocoon of a worm fed on mulberry leaves in the Yellow River valley became, by the time it reached Rome, the most expensive cloth in the empire, and the empire paid for it in coin that did not come back.

What Pliny the Elder saw and named in 77 CE was the structural problem of a monetised consumer economy buying a luxury that only one foreign producer in the world could supply, through a chain of middlemen whose entire wealth depended on keeping the producer and the consumer apart. The Roman moralist tradition called it a moral failure of Roman women. The economic ledger called it the price of conspicuous consumption settled in gold. The Han imperial archives called it the western market for the empire's most prestigious craft. The Sogdian merchant houses, the Bactrian caravan brokers, the Parthian customs officers, and the Palmyran caravan syndics each called it the route they grew rich on. None of them was wrong; the transmission was simply large enough to accommodate every one of those framings at once.