Byzantium stole sericulture in 552 — then locked it in the palace

Before: an empire that wore silk it could not make

In the spring of 552 CE, Constantinople was the best-dressed city on earth and could not produce a single thread of the fabric that made it so. Silk was woven in the capital — in the imperial workshops behind the palace walls, in the private houses of Berytus and Tyre on the Syrian coast, in Alexandria and Antioch — but every filament of the raw material had been reeled from cocoons grown somewhere east of the Sasanian frontier and carried west through Iranian hands. Byzantine weavers were world-class. Byzantine sericulture did not exist.⁷

This was not a minor gap in a luxury market. Silk in the sixth-century Eastern Roman world was constitutional material. Rank at court was legible in cloth: which grades of dyed silk a man could wear, what colour of border his tunic carried, whether his garment used murex purple at all, encoded his standing in the imperial hierarchy as precisely as a modern insignia.⁷²⁰ Silk was also the empire's principal instrument of diplomacy. When Constantinople needed to buy a frontier tribe, secure a marriage, or thank a Venetian fleet, it paid in cloth. Anna Muthesius's survey of the evidence establishes silk as an object of state at every level of Byzantine political practice — gift, bribe, salary, subsidy, relic-wrapping, shroud.²⁰

An empire whose diplomatic currency, court hierarchy, and liturgical furnishing all depended on a commodity monopolised by its principal military enemy had a structural problem. In the 540s that problem became acute.

The textile world of sixth-century Constantinople

The Roman Mediterranean had always been a wool-and-linen world. Wool was the ordinary cloth of the empire, spun in households and in the state's own weaving establishments; linen came out of Egypt in quantity; cotton was known but marginal. Silk was the exception in every sense — imported, weightless, expensive by a factor that put it outside any ordinary household's reach.¹⁹

Byzantine weavers had, by the sixth century, mastered the compound-weave techniques that make figured silk possible: weft-faced compound twill, the structure that lets a weaver repeat a large-scale design across a bolt of cloth. The looms existed. The dye industry existed, at an extraordinary level of technical sophistication — the murex fisheries and dye-houses that produced the graded purples were a state industry with its own hereditary workforce.⁵⁷ What did not exist was the front end of the supply chain: mulberry orchards, egg-rearing rooms, the annual cycle of feeding larvae through five instars, the reeling of filament off the cocoon.

So Constantinople bought raw silk. It bought it in bulk, through officials of the fisc called commerciarii, and it bought it at prices set on the far side of a hostile border.⁹¹⁰

Where the silk came from, and who took a cut

The route was long, and every stage of it was a toll booth. Between the Chinese and Central Asian rearing-grounds and the looms of Syria, raw silk passed through:

• Chinese and Tarim Basin producers, who reared the worms and reeled the filament, and whose techniques were the most advanced in the world
• Sogdian merchant houses, the trading diaspora based on Samarkand and Bukhara whose colonies ran from Gansu to Crimea and who dominated the overland silk business between the sixth and eighth centuries¹¹
• Sasanian Iranian customs and middlemen, who controlled every land approach to Roman territory and who could — and did — set the price, tax the transit, or close the route entirely
• Roman entry points at Nisibis, Callinicum, Artaxata, and later Dara, where the goods were assessed, taxed again, and searched

The Sasanian position was the decisive one. Iran did not merely tax the silk trade; it sat astride it. Sixth-century commerce between the two empires was legally confined to designated frontier towns, and after the Fifty-Year Peace of 562 that confinement was written into treaty: trade was to pass through Nisibis, Callinicum, and Dvin, with foreign traders restricted to Dara on the Roman side and Nisibis on the Iranian, where customs were collected and travellers searched on the explicit grounds that merchants made convenient spies.⁴ The system was designed by both parties to make trade legible and taxable. Its practical effect was that Iran held a valve on Roman silk.

What Iran had that Rome did not

It is worth being precise about the asymmetry, because it was not simply that Iran sat on the road. Sasanian Iran had its own silk-weaving industry, and a good one. Iranian workshops produced the pearl-roundel compound twills whose design grammar — confronted animals in beaded medallions, senmurvs, winged horses, hunting kings — became the single most influential visual vocabulary in early medieval luxury textiles, and which Byzantine and later Islamic weavers absorbed wholesale.¹²

Some of that capacity had itself been acquired by force. Shapur I's third-century campaigns against the Roman east, culminating in the capture of Antioch and the capture of the emperor Valerian in 260, brought deported Roman craftsmen into Iranian cities as a matter of deliberate policy; skilled labour was among the assets a Sasanian king counted as spoil. The technical traffic between the two empires was old, two-directional, and frequently involuntary well before 552.¹²

So the sixth-century position was this: Iran could weave silk and could obtain raw silk. Rome could weave silk and could not obtain raw silk except through Iran. Neither empire could rear the worm. The one that solved that problem first would change the terms permanently — and both courts appear to have understood it.

The 540s: war, embargo, and a price the empire could not control

The Byzantine–Sasanian war that opened in 540 with Khosrow I's sack of Antioch turned that valve into a weapon. Silk prices rose. Procopius, writing his hostile Anecdota in the early 550s, records the consequence in detail that no defender of Justinian ever managed to refute: the emperor ordered the commerciarii not to pay more than fifteen gold pieces a pound for raw silk, and the Iranian sellers, who could get more elsewhere or simply wait, preferred not to sell at all.²

What followed was a policy disaster with named victims. Justinian then fixed the maximum price of finished silk stuffs at eight gold pieces a pound. The silk houses of Berytus and Tyre — old, private, technically excellent, and utterly dependent on a raw material whose import price the state had capped below its replacement cost — were destroyed. Procopius says the trade was ruined, that its masters were beggared, and that the workmen scattered. Some went into the imperial workshops. Others crossed into Iran and took their skills with them.²¹²

Peter Barsymes, Justinian's Syrian-born finance minister — comes sacrarum largitionum around 540 and again from 547, praetorian prefect of the East 543–546 and again from about 555 — is the official Procopius holds responsible for converting the wreckage into a state monopoly.² Whether Barsymes engineered the crisis or exploited it is not recoverable from a source this partisan. What is recoverable is the shape of the outcome: after the 540s, the Byzantine state was in the silk business, and its private competitors in Syria were not.

What the Romans did and did not know about the worm

The Roman world had been trying to understand where silk came from for five hundred years. Pliny the Elder had reported that the Seres combed it from the leaves of trees. The Romans knew Coa vestis, the wild-silk cloth of Cos woven from Pachypasa otus cocoons, which is a genuine silk but a different animal and a vastly inferior thread. What nobody in Constantinople had was the operational knowledge: that the fibre came from the cocoon of a single domesticated moth, Bombyx mori, which eats mulberry leaves and nothing else worth having, and which can be moved from one continent to another as a handful of eggs.

That is the specific ignorance that ended in 552. The transmission was not the discovery that silk came from a worm — Procopius's monks had to explain that too, but it was becoming known.¹³ The transmission was the arrival of the package.

The transmission: two monks, a hollow cane, and a state that wanted a fabric

Photograph by Daderot. Bombyx mori exhibit, Naturhistorisches Museum, Braunschweig. CC0 via Wikimedia Commons. · CC0

552: the audience with Justinian

Procopius tells the story in the eighth and last book of the Wars, written within a few years of the event he describes. Monks came to Constantinople from the east and asked for an audience:

"About the same time there came from India certain monks; and when they had satisfied Justinian Augustus that the Romans no longer should buy silk from the Persians, they promised the emperor in an interview that they would provide the materials for making silk so that never should the Romans seek business of this kind from their enemy the Persians, or from any other people whatsoever."¹

They explained their credentials: "They said that they were formerly in Serinda, which they call the region frequented by the people of the Indies, and there they learned perfectly the art of making silk."¹ And they explained the biology, including the detail that gives the account its ring of operational truth — that the eggs of a single hatching are innumerable, and that once laid, they are covered with dung and warmed until they hatch.¹

That last sentence is why historians take Procopius seriously here. A propagandist inventing a heist does not invent manure. The dung-warming detail is a specific, verifiable technique of egg incubation, and Gang Wu's 2023 reconstruction uses precisely such details — mulberry feeding, dung incubation, and the processing of naturally pierced cocoons rather than boiled ones — as a technical fingerprint that can be matched against the sericultural practice of candidate source regions.⁶

The biological package: what actually had to move

Understanding why this transmission was possible at all requires understanding how little had to travel. Sericulture is not a machine. It is a biological system that can be reduced, for transport purposes, to three things:

1. Eggs of Bombyx mori, which are dormant, minute, and survive weeks of cool transit — a matchbox holds tens of thousands
2. Mulberry, either as seed or cuttings, or the local knowledge that a Morus species already growing in the destination will serve
3. Procedural knowledge — when to warm the eggs, how many times the larvae moult, when to give them straw to spin on, how to get filament off a cocoon

Everything else — looms, dyes, weaving skill — Constantinople already had. This is the reason the theft worked, and the reason it could be described honestly as espionage rather than conquest: the entire strategic value transferred inside a walking stick.

Theophanes's Persian, and the problem of two stories

Procopius is not the only witness. Theophanes of Byzantium, writing later in the sixth century, told a different version, preserved because Photius summarised his lost history in the ninth century. In Photius's Bibliotheca, codex 64, the carrier is not a pair of monks but a single Iranian:

"A certain Persian... had taken with him the eggs of these worms (concealed) in a walking-stick... succeeded in bringing them safely to Byzantium... In the beginning of spring he put out the eggs upon the mulberry leaves which form their food."³

The two accounts cannot both be literally true, and the historiography has generally not tried to make them so. What they agree on is more interesting than what they dispute: that the transfer happened under Justinian, that it was clandestine, that the vector was eggs rather than adult insects, that mulberry came with it, and that a hollow cane or similar container was involved. Wu argues that the whole framing of the event as a single dramatic transfer is the problem — that the evidence is better explained as a process of appropriation spread across regions and generations than as a heist with a date:

"The transfer of sericulture into Byzantium is a critical episode in the global dissemination of silk production technology. However, it is now widely accepted that the explanatory model portraying the transfer as a one-off event is at odds with the historical facts."⁶

This atlas keeps the date 552 as an anchor because it is the date the sources give and the date around which Byzantine state policy visibly changes. But the honest version is Wu's: something arrived in the middle of the sixth century, and it took the empire centuries to make it work.

Where "Serinda" was — Khotan, Sogdiana, or neither

"Serinda" is a compound that appears to fuse Seres (the silk people) with India, and it has been a scholarly puzzle for three centuries. Candidates proposed at various times include China proper, the Tarim Basin oasis kingdoms — Khotan above all — Sogdiana, Ceylon, northern India, Cochinchina, the south-east Caspian coast, and even Syria. None of the etymological arguments has proved decisive.

Wu's contribution is to abandon etymology and reason from technique instead. Comparing the practices Procopius and Theophanes describe against what is known of sericulture in each candidate region, he identifies Khotan and the surrounding Tarim Basin as the immediate provenance of Byzantine sericulture — not China directly.⁶ The reasoning turns on the non-killing method: Chinese sericulture killed the pupa by boiling or salting to preserve an unbroken filament for reeling, whereas the practice that reached Byzantium let the moth emerge and worked the pierced cocoon into floss. That is a Khotanese modification, not a Chinese one.

The brief for this record was written on the assumption of a Sogdian origin, and Sogdiana remains where the commercial network ran. But the technical evidence points one oasis further east.

The Sogdian network that carried it — and kept trading anyway

Whatever the eggs' provenance, the road they travelled belonged to the Sogdians. Étienne de la Vaissière's Histoire des marchands sogdiens reconstructs a trading diaspora out of Samarkand, Bukhara, and Panjikent whose colonies dotted the routes from Gansu to the Crimea between the fourth and eighth centuries, and whose members were the actual carriers of most silk moving west.¹¹

Here is the part of the story that the "industrial espionage" framing obscures: the theft did not put them out of business. Sogdian merchants continued to dominate the overland silk trade for two more centuries after Byzantium had worms of its own. The reason is the one Wu documents — Byzantine sericulture stayed small and technically inferior, and the empire went on importing raw silk in quantity from the east well past the tenth century.⁶¹⁴ Byzantium stole the capacity to make silk. It did not steal the capacity to make enough.

The failed alternatives: Aksum, and the Turkic embassy of 568

The smuggling was not Justinian's first attempt to route around Iran, and not his last. Two others are documented:

• The Aksumite scheme (c. 531). Justinian sent an embassy to the Aksumite kingdom in the Ethiopian highlands proposing that Aksumite merchants buy raw silk in India and resell it to Rome, bypassing the Persian Gulf via the Red Sea. It failed on simple logistics: Iranian buyers, closer to the Indian ports, reached the ships first and bought out the cargoes.
• The Turkic–Sogdian embassy (568). After Justinian's death, the Western Türk khagan Sizabulos sent an embassy to Constantinople at the urging of Maniakh, described in Menander Protector's history as chief of the Sogdians. Maniakh had first tried to get Iranian permission for Sogdian merchants to sell silk directly into Iran; rebuffed, he proposed a Roman alliance instead, and came himself to negotiate it. Menander preserves the exchange.⁴

The 568 embassy is the clearest evidence that the 552 smuggling had not solved the problem. Sixteen years after the monks, the Byzantine state was still negotiating for access to eastern raw silk — and the Sogdians were still the people to negotiate with.

What changed and what was replaced

The imperial monopoly: from purchase to production

The most consequential change was institutional, not agricultural. Before the 540s the Byzantine state was the largest customer in a silk market. After the crisis of the 540s and the arrival of the worm in the 550s, it was the market.

Robert Lopez's 1945 study in Speculum remains the founding account of what this meant in practice: manufacture and sale of the highest grades became an imperial monopoly, processed in imperial establishments and sold only to authorised buyers, and the emperor used control of silk, purple, and gold thread as a lever of foreign policy — granting or withholding export as a diplomatic instrument.⁹ Anna Muthesius's later work revised Lopez substantially, showing the picture was more mixed than a pure state monopoly and that a regulated private industry persisted alongside the imperial workshops.⁸ George Maniatis's reconstruction of the tenth-century private silk industry shows how elaborate that regulated sector became.¹³

The regulatory architecture is best documented for the tenth century, in the Book of the Eparch, promulgated under Leo VI around 911–912. It divides silk work into separately licensed guilds:

Each guild had a defined transaction perimeter, public declaration requirements, fixed sales channels, and penalties for hoarding or for selling outside its licence.¹³¹⁷ This is not a free market that grew up around a new technology. It is a fabric industry organised as a branch of the fisc.

Kekolymena: the grades that could not leave the palace

At the top of the system sat the kekolymena — literally "forbidden" goods. The category covered strategic commodities generally (grain, salt, precious metal, iron, arms) and, within textiles, the highest grades of murex-purple silk. Specific named purples — oxeon, porphyraeron, and the diblatta grades — were reserved to imperial manufacture, and dealers in finished garments were forbidden to sell them to intermediaries who would resell them to foreigners.⁷¹⁷

The prohibition was political as much as economic. A grade of cloth that only the emperor could produce and only the emperor could give away is a diplomatic instrument of enormous efficiency: its value cannot be counterfeited by any rival, because the rival cannot make it. That is what Byzantium built out of the smuggled eggs.

It leaked, of course. Liutprand of Cremona, sent to Constantinople in 968 and searched on his way out, complained that the forbidden purples he had been forbidden to export were freely available in western markets through Venetian intermediaries.⁷

The fiscal logic, and why the state wanted it

The reason a late-Roman state should want to own a textile industry outright is not obvious to a modern reader, and it is worth spelling out. The Byzantine fisc had a structural problem that the Economic History of Byzantium documents across its chapters on exchange and state finance: the empire's revenue was largely agricultural and cash-based, its principal expenditures were military and diplomatic, and it had chronic difficulty converting tax receipts into the specific instruments foreign policy required.¹⁵

Silk solved that conversion problem elegantly. A bolt of imperial-grade figured silk was:

• Compact and durable, so it could be carried to a Caucasian court or a steppe camp without a baggage train
• Unambiguously valuable to every party in Eurasia, without needing to be assayed like coin
• Impossible to counterfeit at the highest grades, because the dyes and looms were state-controlled
• Manufacturable from domestic inputs once the worms arrived, which meant it cost the treasury labour and mulberry rather than gold

Nicolas Oikonomides's study of the seals of the kommerkiarioi — the officials who handled the state's commodity transactions between the sixth and ninth centuries — reconstructs the administrative machinery through which this worked, and shows the silk trade running through the same fiscal apparatus that managed the state's other strategic commodities.¹⁰ The monopoly was not primarily a rentier's greed. It was a treasury converting cloth into geopolitics.

The look of the thing: what Byzantine silk became

The aesthetic consequence was not a Byzantine style displacing an Iranian one. It was the opposite: with production now domestic, Byzantine workshops scaled up their output of designs largely inherited from the Sasanian repertoire they had been importing for a century. Otto von Falke's Kunstgeschichte der Seidenweberei, still the foundational art-historical survey of silk weaving, laid out the pattern genealogies that make this visible — the pearl roundel, the confronted beasts, the heraldic symmetry — as a continuous Iranian-into-Byzantine-into-Italian succession rather than a sequence of independent inventions.¹⁸

The surviving objects bear it out. The great Byzantine silks in western church treasuries — the elephant cloth at Aachen, the Gunthertuch at Bamberg, the lion and griffin samites scattered through European reliquaries — are Constantinopolitan products carrying a compositional logic that came from Ctesiphon.⁷¹⁸ The empire stole the worm from the east and went on borrowing the pictures from the east too.

Mulberries in Syria, Greece, and the Peloponnese

The agricultural change was slower and much harder to see. Sericulture requires mulberry plantations, and mulberry plantations take years; it requires a rearing workforce with acquired judgement; and it leaves almost no archaeological trace. Wu emphasises the resulting evidentiary gap: there is no solid documentation of domestic Byzantine sericulture between the sixth century and the eleventh, which makes the intervening five hundred years a matter of inference.⁶

What can be established is where the industry surfaced when the sources resume. Kostis Kourelis, surveying the medieval Peloponnese for Dumbarton Oaks, sets out the sequence plainly: "Silk production is said to have been introduced into Constantinople in the sixth century when two legendary Byzantine monks smuggled cocoons," and "Byzantine silk was initially under tight imperial control, but sericulture spread to provincial cities in the ninth century." By then, he notes, "Patras, along with Corinth and Thebes, became famous for its silk manufacturing."¹⁶

By the twelfth century Thebes in Boeotia was the empire's pre-eminent silk town, producing figured cloth that contemporaries rated at or above the capital's, with Corinth serving as its principal entrepôt.¹⁴ David Jacoby's work on silk in western Byzantium before 1204 is the standard account of how a capital-centred imperial industry became a provincial Greek one.¹⁴

What the smuggling displaced

Three things were displaced, and they were not the things one would expect.

The Syrian private silk houses. The industries of Berytus and Tyre were not outcompeted by Byzantine sericulture; they were killed by their own government's price controls a decade before the worms arrived, and the state absorbed what was left.²¹² The transmission's institutional consequence — production as a fiscal monopoly — was built on that wreckage.

Iran's transit rent. Sasanian Iran's leverage over Roman silk was a revenue stream and a strategic asset. Its erosion was gradual rather than sudden, since Byzantium kept importing, but the direction of travel after 552 was one way.

Wild silk and the substitutes. Coa vestis and the other pre-Bombyx silks of the Mediterranean disappear from the record as commercial products. Once real silk could be grown in Roman territory, the imitations had no market.

What did not transmit

This is the part of the story the heist narrative gets wrong, and it is worth stating flatly: China's advantage survived the theft by roughly a millennium.

Wu's reconstruction identifies three technical practices that were lost or never acquired on the road west:⁶

1. Fire-heated cocooneries. Chinese rearers warmed eggs and larvae with managed fire. The Khotanese practice substituted dung; later Byzantine descriptions substitute human body heat and wrappings of linen and leather. Each substitution is a downgrade in control.
2. Killing the chrysalis. Chinese sericulture killed the pupa — boiling or salting the cocoon — precisely so that the filament could be reeled off unbroken, in lengths of hundreds of metres. The practice that reached Byzantium let the moth emerge and pierce the cocoon, which destroys the continuous filament and leaves only short-staple floss to be spun. This is the single most consequential loss in the whole transmission.
3. Scale. Byzantine sericulture never industrialised. Surviving Mediterranean textiles from roughly 600 to 1300 contain very little spun silk relative to reeled imported silk — which is to say that the empire kept buying eastern raw silk for seven centuries after it had learned to grow its own.

Whether the non-killing method reached Khotan through Buddhist objection to killing the pupa or through simple technical loss is not settled. Either way its consequences travelled: higher cost per unit of usable fibre, lower quality, and a permanent ceiling on Byzantine output.

Onward: Sicily 1147, and the lineage of stolen cuttings

The final displacement was of Byzantium itself, and it happened the same way the original acquisition had — by taking the thing rather than buying it, except with soldiers instead of monks.

In 1147, during the Byzantine–Norman war, a Sicilian fleet under George of Antioch raided the Greek mainland. At Thebes the Normans sacked the silk workshops and carried off the weavers — from Corinth and Athens as well — to Palermo, where they were set to train Sicilian workers.¹⁴ Within a generation Sicily was producing figured silks closely comparable to the Byzantine lion silks. The Italian silk industry that would dominate late-medieval Europe begins, in significant part, with a shipload of abducted Theban weavers.

The pattern outlasts antiquity. Sericulture and its associated crops moved repeatedly by theft, smuggling, and coerced transplantation: into the Islamic Mediterranean, into Norman Sicily, into Lucca and Florence and Lyon, and — in the same idiom eleven centuries later — tea seedlings out of China and rubber seeds out of Brazil. The two monks are the earliest well-documented instance of a manoeuvre that industrial states have never stopped performing.

What the cost was

Photograph by Steven Zucker. Justinian and his retinue, mosaic, Basilica of San Vitale, Ravenna, consecrated 547. CC BY 2.0 via Wikimedia Commons. · CC BY 2.0

The bill was not paid in blood

Set the accounting honestly. Nobody was massacred for the silkworm. There is no siege in this record, no sack, no documented mortality figure attributable to the transmission. The eggs moved in a stick; the empire that received them fought no war to get them. Compared with most transmissions in this atlas, the direct human cost of the acquisition itself is close to nil.

What the record does contain is a well-documented transfer of coercion — an empire that broke a monopoly it resented and reconstituted the same extraction inside its own walls, with its own people in the bound positions.

The weavers who were already property

Byzantine imperial textile production inherited a legal regime that was explicitly hereditary and explicitly unfree. The late Roman gynaecea — state weaving establishments, fourteen of them listed for the western empire alone in the Notitia Dignitatum — were staffed by workers bound to their trade for life and by birth. The Codex Theodosianus devotes a whole title to them: X.20, De murilegulis et gynaeciariis et monetariis et bastagariis, "on the purple-fishers and the weaving-workers and the mint-workers and the carriers."⁵

Its provisions are not ambiguous:

• Workers in these trades could not leave their post or transfer to another service; a constitution of 384 forbids desertion outright and fines officers a pound of gold for each fugitive accepted.⁵
• Anyone concealing a workshop worker paid five pounds of gold (372).⁵
• Children of murileguli, including children of mixed parentage, inherited the workshop obligation through the maternal line (425).⁵
• The related regime for arms-factory workers mandated stigmata — public marks branded on the arm — so that absconders could be identified (398).⁵

This is the labour system into which the smuggled worms were delivered. The dye-houses that made the kekolymena purples were staffed by hereditary murex-fishers; the imperial weaving establishments by hereditary weavers. The transmission's institutional beneficiary was a monopoly whose skilled workforce could not legally quit.

Tyre and Berytus: an industry destroyed by its own government

The clearest named damage in the record is Roman-on-Roman. Procopius's account of the 540s silk crisis names the cities and describes the outcome: the silk manufacturers of Berytus and Tyre — private, ancient, and among the most skilled in the Mediterranean — were beggared when the state fixed the price of finished silk at eight gold pieces a pound while raw silk could not be had at fifteen.² The workmen dispersed. Some were absorbed into the imperial workshops. Others emigrated to Sasanian Iran and taught Iranian workshops what they knew.²¹²

Procopius is a hostile witness and his figures should be held loosely. But the direction of the story is corroborated by everything that follows: after the 540s, the great private Syrian silk houses are gone from the record, and the industry's centre of gravity is the palace.

The irony is exact and worth naming. The stated purpose of the whole enterprise — the monks' pitch to Justinian, in Procopius's words, was that "the Romans no longer should buy silk from the Persians"¹ — was to end an extraction by a foreign power. The immediate victims of the policy that produced it were Roman craftsmen, and the beneficiaries of the workers' flight were Iranian workshops.

Iran's lost rent and the wars that followed

The Sasanian side of the ledger requires care, because the temptation to overclaim is strong. Iran's control of the silk routes was worth a great deal — customs at the treaty towns, the middleman's margin, and the strategic ability to squeeze Constantinople at will. Byzantine sericulture, plus the Turkic alliance the Sogdians brokered in 568, eroded that position over the following half-century.

It does not follow that the silkworm caused the catastrophic Byzantine–Sasanian war of 602–628 that exhausted both empires and left them open to the Arab conquests. That war had its own causes: dynastic murder in Constantinople, Khosrow II's claim to avenge Maurice, and a century of unresolved frontier disputes. Loss of transit revenue belongs on the list of pressures, as a contributing factor among several — not as the cause. The atlas records it as such, and declines to draw a straighter line than the evidence supports.

The lineage of coerced labour

The strongest cost claim in this record is not about 552 at all. It is about what the package did every subsequent time it moved.

• Byzantium (from the sixth century): production concentrated in imperial establishments with a hereditary, legally bound workforce, and in a licensed guild system that dictated what its members could buy, make, and sell.⁵¹³¹⁷
• The medieval Peloponnese and Boeotia (ninth to twelfth centuries): provincial silk towns organised around great estates and their dependent labour. The Patras magnate Danielis, who supplied Basil I with luxury textiles in the later ninth century, ran her textile enterprises with a household of thousands of slaves and sent consignments of skilled female weavers to the emperor as gifts.¹⁶
• Norman Sicily (1147): an industry founded on weavers abducted at spear-point from Thebes, Corinth, and Athens and resettled in Palermo to train replacements.¹⁴
• Later transplantations: the movement of sericulture into the Islamic Mediterranean, Italy, and France repeatedly rested on regulated, indentured, or coerced labour — a pattern this record flags as a lineage rather than a causal chain, because each transplantation had its own local politics.

What this record does not claim

The confidence rating on this record is deliberately middling, and the reasons should be visible to the reader rather than buried in a metadata field.

• The date is a convention, not a measurement. Procopius places the episode around 552; Theophanes tells a version with a different carrier; the two cannot be reconciled. Wu's argument that the whole one-off-transfer model is wrong has not been refuted.⁶
• The provenance is inferred, not documented. Khotan is the best-argued source on technical grounds. Sogdiana is where the commercial network ran. No source says either.⁶¹¹
• The five-century gap is real. Between the sixth century and the eleventh there is no solid documentation of Byzantine domestic sericulture. Anyone describing an unbroken industry from Justinian to the Komnenoi is describing a reconstruction.⁶
• Procopius on the 540s is a hostile source. The price figures and the ruin of the Syrian houses come from a writer whose purpose was to indict Justinian. The pattern is corroborated by the industry's disappearance from the record; the numbers are not independently attested.²¹²

None of this softens the cost claims, which rest on Roman law and on the 1147 abduction rather than on Procopius. It does mean the transmission's mechanism is the least secure part of a well-secured story.

The bill

What the transmission of sericulture to Byzantium cost, stated plainly:

1. The private silk industries of Berytus and Tyre, destroyed by Justinian's price controls in the 540s, their masters ruined and their workers dispersed — some into state workshops, some into exile in Iran.²
2. The consolidation of silk work under a hereditary, legally unfree labour regime already codified in Roman law and carried forward into the Byzantine imperial workshops.⁵
3. Sasanian transit revenue, eroded from 552 onward — a real strategic loss, a contributing pressure on the seventh-century wars, not their cause.
4. The abduction of the weavers of Thebes, Corinth, and Athens in 1147, the downstream cost of having concentrated an industry into a small number of seizable towns.¹⁴
5. A template for technology theft between rival states that has been reused, with escalating stakes, for fifteen centuries.

Against that, the transmission's achievement: a fabric industry in Europe. Every European silk thereafter — Lucchese, Florentine, Genoese, Lyonnais, Spitalfields — descends from worms that reached the Mediterranean in the middle of the sixth century in something the size of a walking stick, carried by people whose names nobody wrote down.

The monks got a paragraph in Procopius. The weavers got the Codex.