Welcome to our fourth instalment of ‘Early Medieval (mostly) Textiles’. This month we are travelling through the Mediterranean, North Africa and Asia during Late Antiquity and the early medieval period, where we discover how the invention of steel needles changed weaving processes and the production of embroidery.
Karen Carr was an Associate Professor of History at Portland State University until she resigned her post in order to focus on Quatr.us Study Guides, her feminist, anti-racist history website, in 2011. She is also a Roman pottery specialist for the University of Manitoba excavations at Leptiminus in Tunisia. Her books include Vandals to Visigoths: Rural Settlement Patterns in Early Medieval Spain (2002), which examined the ways in which rural small farmers were tied into the Roman economy, and Shifting Currents: A World History of Swimming (forthcoming, 2021), which will appear next year, now that her children are mostly grown. She is now working on Women, Clothing, and Money from the Stone Age to the late Middle Ages. Karen has long been interested in women and textiles and she has presented papers and spoken on the subject throughout her career. Here she focuses on how the invention of steel needles changed the weaving process and embroidery production across Mediterranean, North Africa and Asia.
When Sewing Needles were high tech: Women and Work in the Early Middle Ages
In the last years of Roman rule in Egypt, in the seventh century AD, a group of women sit in a courtyard embroidering colorful silk thread onto white linen tunics. Their steel sewing needles glitter in the sunlight. These African women might be making clothes for themselves, or for the toddlers playing at their feet. But are they? The babies are naked. The women’s own tunics are worn and faded. Who will wear the fancy tunics they are decorating?
The mystery clears up when a better-dressed man strides into the courtyard without knocking. His assistant carries a basket of brightly dyed silk thread, with a stack of clean new white linen folded on top. The men distribute a few bronze coins from a leather pouch, hand over the new materials, heft the basket of finished work, and leave for the next house. The women are doing piecework.
Much though we’d like to imagine women embroidering little dresses for their daughters, in the Early Middle Ages most people bought most of their clothes in stores. Rich people bought new, in shops. Poorer people often bought secondhand, like Jiao Xian in China, who, when his clothes wore out, would “sell some firewood in order to buy old garments to replace them” (Ge Hong), or the customers of Jerome’s peddler, “who dealt in old clothes.” Instead of making clothes for their families, women worked long boring hours to produce cloth for their employers or enslavers, much of it expensive brocades, damasks, and embroidery. The early medieval textile industry was not a home industry, or a local industry, but a competitive inter-regional behemoth requiring expensive new technology and a cheap, largely female workforce.
Exploitation of women’s work has a long history in Egypt, where wooden models in Old Kingdom tombs already depict workshops where women spin and weave linen. And not only in Egypt; the textile industry exploited women everywhere. Scribes in Bronze Age Greece tracked the output of women enslaved in weaving and spinning sheds. By the Iron Age, Homer’s Iliad mentions piecework, as a poor woman “weighs her wool … working to win a pitiful wage for her children.” Roman epitaphs mention quasillariae, spinners, and lanipendi, wool-weighers — all apparently women. On the other side of Asia, a Chinese Han Dynasty poem describes how “Hungry, she still weaves/ Numbed with cold, she still weaves./ Shuttle after shuttle after shuttle. /The days are short, The weather chill,/ Each length hard to finish.” In Byzantine Egypt, the mulieres lanificae, or wool-working women, still worked under the same difficult conditions.
Steel Sewing Needles
But although these women’s work was tedious, it was neither simple nor static. The competitive fashion industry forced major textile producers to invest in ongoing research and development programs, often with government support. Textile workers constantly upgraded their skills. They had to use the absolute newest technologies: the highest quality steel and the most complex looms. Sewing needles seem simple, but for the embroiderers in their sunny courtyard, the steel sewing needles they were using were a major new technology, recently arrived in the Mediterranean from Central Asia.
Sewing needles themselves were not new. They had been around for forty thousand years, since the Ice Age. Early needles were carved from slivers of bone, bored through (using wet sand as an abrasive) to make an eye. By the Iron Age, sewing kits included bronze and iron needles. But all of these early needles were brittle: if they were thin enough to go through fine fabric without poking big holes in it, they were always snapping in half. Then in the 300s BC, the Indian steel industry developed enough to make high quality steel sewing needles, flexible and thin enough for fine embroidery.
Crucible steel wasn’t developed in order to make sewing needles. Expensive new technologies are often military, and most of the new steel went for swords. When Charlemagne invaded Spain in 788 AD, as celebrated in the Song of Roland, he was seeking to plunder “Indian swords”. The new crucible steel technology involved Indian steel-workers who wrapped clay around a mixture of brittle, high-carbon pig iron and low-carbon wrought iron, charcoal, rice hulls, and quartz, and baked their package — the crucible – slowly for several days (Figure 1). The process forced carbon into the wrought iron, yielding the black-and-white watered silk pattern of Damascus steel (Figure 2). Each clay crucible made enough steel for one sword. Viking traders bought that steel in Iran, on the shores of the Caspian Sea, and carried the ingots back to the Baltic, where blacksmiths forged them into swords. But that same steel ingot could yield thousands of sharp, flexible sewing needles. Needle-making was a lot of work, even after the laborious process of making crucible steel. The medieval Iraqi philosopher al-Ghazali claims (perhaps rhetorically) that shaping steel ingots into sewing needles required twenty-five specialized stages. Steel itself was expensive, and then metalsmiths had to extrude the wire, possibly through a draw plate (Figure 3), cut the wire into the length of two needles, sharpen both ends, pierce the eyes in the middle, smooth and polish both the length and the eye, cut the two needles apart, anneal them, and polish them again. But the high prices commanded by the resulting embroidery made up for the cost of this lengthy process.
Steel needles were probably originally developed in India, like the steel from which they were made. India’s fabled surgeons published two major surgical textbooks (the Charaka Samhita and Sushruta Samhita) before 100 AD, perhaps inspired by the new needles. Traveling doctors probably brought steel needles to China, where in the second century BC, the doctor Chunyu Yi stood trial for malpractice for introducing acupuncture. A century later, the Neijing, a Chinese medical text, accepted acupuncture as standard medical procedure.
But in the competitive atmosphere of the fashion business, Chinese textile manufacturers quickly adopted the new needles. Already in the 300s BC, a tomb in Baoshan contained a steel sewing needle, together with an iron thimble-ring. Thimbles were another new invention, made to protect embroiderers’ thumbs from their new steel needles. About 150 BC, another steel needle appears in the lavish tomb of Xin Zhui, a wealthy woman of the Han Dynasty, at Jianling, Mawangdui. The eye of this needle still has silk threaded through it. Archaeologists have found no more needles — the needles were expensive! – but plenty of bronze thimble-rings survive from the Han Dynasty (ca. 200 BC – AD 200) (Figure 4). With their new needles and thimbles, Chinese embroiderers produced elaborate embroidered silks for merchants to sell on the Silk Road. We find these embroideries as far west as the Altai Mountains in Kazakhstan (Figures 5,6,7). But just as the invention of steel needles inspired the invention of thimbles, the needles also seem to have inspired innovations in wooden looms.
In addition to the steel needle, Xin Zhui’s tomb complex held a new kind of silk textile woven with repeating patterns, manufactured on a new kind of loom: the pattern loom (Figure 8). We know exactly what this new loom looked like, because another Han Dynasty tomb, at Laoguanshan in Chengdu, contained a tiny weaving workshop with four such looms, along with figurines depicting two male weavers and a number of women spinners and bobbin-winders. The new pattern looms have foot-treadles to change the shed, and an elaborate superstructure with multiple heddles (Figure 9). Two people run them, a weaver down below to throw the shuttle, and a drawboy or drawgirl perched on top of the superstructure to change the heddle and form the pattern. By lifting the various heddles in a set rotation, the drawboy or drawgirl produced a repeating pattern in the warp threads that would appear as an image in the cloth. In a pattern loom, the pattern is in the warp threads that are on the loom before you start weaving, not the weft thread on the shuttle, which remains the same throughout (Figure 10). Imperial funding through the government silk workshop at nearby Chengdu probably supported the development of this elaborate “jin silk” loom, which, in addition to being able to produce patterns, was about three times faster than earlier looms.
Merchants sold Han Dynasty jin silks and the new Chinese embroideries all along the Silk Road, as far away as India and the Mediterranean Sea, where they must have caused a sensation. In India, these woven images were matched by images block-printed onto cotton using indigo or madder dyes. India exported these printed cottons in the first and second centuries AD, as we know from excavations at the entrepot of Quseir al-Qadim in the Arabian Peninsula.
Roman textile manufacturers in Syria and Egypt could not block print on linen (too hard to dye) or wool (too rough), and they had neither pattern looms nor steel sewing needles. The Periplus of the Erythraean Sea tells us that “Indian steel” was imported to the Roman Empire along with Indian printed cotton by the mid-first century AD, but if the famous Syrian doctor Galen used steel needles for surgery, he doesn’t say so. He refers only to steel scalpels, made from European steel from Noricum (now Austria). No Roman steel needles have survived, even in the arid Egyptian desert; those that have been found are of bronze. Lawrence Bliquez suggests that Roman steel needles have all rusted away, but the total absence of bronze thimbles indicates that steel needles were simply not in use. The Romans didn’t produce silk, either. They could not duplicate Chinese embroideries, nor Chinese pattern-weaving. Perhaps that is why Pliny, writing his encyclopedia in the first century AD, worries about millions of sesterces leaving the Empire eastward every year (NH 12.41), and Seneca asserts that “We can dress ourselves without any trade with China.” (Epistle 91). Yet Julius Caesar had imported silk curtains on his chariot, and Caligula, Titus, and Vespasian all wore imported silk clothing. What textiles could Roman manufacturers make to send back east, to compete with these new silks? They found an answer in the colorful wool weaving we know as Coptic tapestry.
Mediterranean weavers had been weaving complex colored images for centuries. They wove in wool, more available around the Mediterranean than further east, and wool took dye much better than either linen or silk. Already in the Iliad, Helen weaves “a great double-width purple cloth, showing the many battles on her behalf.” Greek black figure vases from the 500s BC show women in patterned cloth. By the 100s BC, we know, such tapestries were being exported, like the polychrome Sampula wool tapestry that was exported, perhaps from Palmyra or Bactria, to western China. On these wool tapestries, the pattern is in the weft, not the warp.
To the embroidered silk dragons of China, Mediterranean weavers now returned detailed pictures in brightly colored wool wefts on linen backing (Figure 11). These images were not embroidered, because the Mediterranean did not yet have steel needles. They were woven by hand with short lengths of wool wefts, as Mediterranean people had been doing for some time. But now the weavers used new, much faster, horizontal tapestry looms: more expensive to build, like the Chinese pattern looms, but like pattern looms also more efficient. Around the second or third century AD, Egyptian weavers started to produce quantities of these colorful “Coptic” textiles (probably the name is short for “Egyptian”).
As Pliny had railed against the new embroidered silks, conservatives now opposed the new Coptic textiles. Around 400 AD, the Syrian bishop Asterius complained that “They have invented some kind of vain and curious warp and ‘broidery which, by means of the interweaving of warp and weft, imitates the quality of painting and represents upon garments the forms of all kinds of living beings, and so they devise for themselves, their wives and children gay-colored dresses decorated with thousands of figures.” What Asterius is railing against, perhaps without knowing it, must be the shift from vertical to horizontal looms, which made figured tapestry cheaper, and so more common, than it had been before.
Despite Asterius’s opposition, Coptic tapestry weaving succeeded in creating demand both west and east. Excavation of the Tar Caves in western Iraq found many examples of Egyptian-style wool-faced tapestry moving east. In Yingpan, in western China, a man was buried about 300 AD in a gorgeous red and gold wool caftan, locally woven, but with a pattern of Roman cherubs and bulls. In Europe, where vertical looms and bronze needles remained in use until about 1000 AD,imported Coptic tapestries were also popular; one was found buried with a Frankish prince under Cologne Cathedral. West Africa took a different path, with Fulbe-speakers quickly adopting both the horizontal tapestry loom and sheep-herding for wool. And further east, weavers in the Sasanian empire invented the drawloom.
Around 200 AD, Sasanian textile workers living around the Caspian Sea in what is now Iran started up local production of both Chinese silk and Indian cotton. Like Indian textile workers further south, they produced cheap images on cotton fabrics by painting the cotton with wax and then dyeing it indigo blue, or by stamping it. And they exported their work, judging from a fragment of Central Asian cloth combining Greek and Chinese motifs found in a late third century burial at Niya in western China (Figure 12). Not long afterward the Sasanians started to build drawlooms that could create silk weft-facing tapestries called “samite”. Samite typically involved a complex pattern of roundels, or circles, holding two figures or animals, often facing each other symmetrically (Figure 13). The new Sasanian drawlooms could produce more complex images than a pattern loom. It was faster than the Coptic tapestry looms, and also faster than embroidery. Sasanian drawlooms were huge, and complicated, and very expensive. Manufacturers had to make a much bigger capital investment. But it paid off: Sasanian silk samite cloth quickly became popular among the wealthy both west and east of Iran.
By about 500 AD, Chinese Six Dynasties weavers imitated Sasanian designs on their own expensive drawlooms, shifting from their old warp-faced patterns to the newer weft-faced ones (Figures 14, 15). As they had earlier, they made up the cost of the expensive looms by underpaying the weavers. A Tang Dynasty poem still bemoans the fate of the overworked weaver: “The wind is cruel. Her clothes are worn and thin. /The weaver girl blows on her fingers. /Beside the dark window, back and forth, /she throws a shuttle like a lump of ice./ During the short Winter day, /she can scarcely weave one foot of the pattern./ And you expect me to make a folk song of this, /for your silken girls to sing?”
Mediterranean textile manufacturers had by now learned how to breed silkworms for their silk, and they also constructed the expensive drawlooms. The Roman government was funding both the silk research (or industrial espionage) and the construction of these huge drawlooms: the Roman emperor Justinian was devoted to improving the Roman textile industry because of tax issues with the Sassanians at this time. The first Mediterranean drawlooms were built in Egypt and Constantinople about this time, and Egyptian drawlooms soon produced good imitations of Sasanian samites. The big Imperial weaving factories held a monopoly on production, and presumably could oppress their workers accordingly.
Both Sasanian samites and their Byzantine imitations found eager customers in Europe, where they were considered the most luxurious possible cloth. Samite appears regularly in the rich burials of European rulers and saints in the seventh and eighth centuries AD, and is regularly mentioned by early medieval sources such as Charlemagne’s biographer Einhard. Charlemagne himself, after he was done hunting for “Indian steel” swords in Spain, was buried at Aachen in the early 800s AD shrouded in a samite cloth imported from Constantinople. As befitted an emperor, his shroud was purple, woven to depict a Roman quadriga, with a charioteer driving four horses (Figure 16).
Despite the political upheavals of the 600s, where Islamic empires took over both Iran and Egypt, and the Tang Dynasty took over China, these same Sasanian roundels continued to be produced throughout Afro-Eurasia well into the ninth century. But as in Iran, so in China and along the Mediterranean coast these big drawlooms remained very expensive. And, perhaps because embroidery with steel sewing needles had come to seem a cheaper alternative to the huge capital investment required to construct a drawloom, Mediterranean textile manufacturers finally started to provide steel embroidery needles to their workers.
Steel Needles reach the Mediterranean
Even now, the steel for these needles was still coming from India or perhaps Central Asia. As late as 1000 AD, the Islamic doctor Al-Zahrawi still recommended Indian steel for medical instruments. But Mediterranean blacksmiths had learned to draw out the Indian steel ingots into sewing needles. Although archaeologists have not yet found any steel needles from Constantinople, the first bronze thimble rings appear there, and in Sicily, about 600 AD. These thimbles imitate older Chinese thimble rings, and like them they were designed to keep steel needles from piercing your finger when you pushed the needle through the cloth, so the bronze thimbles imply the existence of steel sewing needles. Confirming this impression, embroidery immediately became much more common in Egypt, and all around the Mediterranean. In this example from Constantinople of embroidered silk on linen, the two figures confront each other inside their roundel in the Sasanian style (Figure 17).
Thus in about nine hundred years the steel sewing needle changed its position entirely. In 300 BC the new needles were incredibly expensive tools, requiring government research to create and significant capital investments to provide to embroiderers, but yielding high profits to the investor. By 600 AD, though steel needles had not become much cheaper, the new drawlooms represented an investment that was greater by orders of magnitude, and so by comparison steel needles seemed inexpensive, a cheaper substitute for the drawloom.
Throughout this time, the Afro-Eurasian textile industry remained inter-regional, competitive, and heavily invested in the latest technology, so that although we associate early medieval crucible steel mainly with men, swords, and battlefields, the same steel was also used by women embroidering at home in their courtyards.
I would like to thank Karen for sharing her work and insights into this interesting, yet under studied topic. It just shows that even the smallest object can have a huge impact on societies and their cultures, which makes them just as exciting and inspiring as the largest monuments. Next month Christina Petty discusses her experimental weaving project, which investigates whether it is possible to produce a 2/1 twill weave on a warp-weighted loom.
Contact: If you would like to contact Karen about her work, please use one of the links below:
Bibliography / Further Reading:
Bagnall, Roger and Raffaella Cribiore, Women’s Letters from Ancient Egypt (Ann Arbor 2015).
Bai Xinghua and R.B. Baron, “How Old is Acupuncture? Challenging the Neolithic Origins Theory”, from Bai Xinghua and RB Baron, Acupuncture: Visible Holism (Oxford 2001).
Barnes, Ruth, Indian Block-Printed Textiles in Egypt: The Newberry Collection in the Ashmolean Museum, Oxford (Oxford 1997).
Bliquez, Lawrence, The Tools of Asclepius: Surgical Instruments in Greek and Roman Times (Netherlands 2014).
Carroll, D. L., “Dating the Foot-Powered Loom: The Coptic Evidence,” American Journal of Archaeology 89 (1985), 168-173.
Feuerbach, Ann, ‘The glitter of the sword: the fabrication of the legendary Damascus steel blades,’ Minerva 13/4 (2002), 45–48.
Gasparini, Mariachiara, Transcending Patterns: Silk Road Cultural and Artistic Interactions Through Central Asian Textile Images (United States 2019).
Meister, Michael, ‘The Pearl Roundel in Chinese Textile Design,’ Ars Orientalis 8 (1970), 255-267.
Morgan, Faith Pennick, Dress and Personal Appearance in Late Antiquity: The Clothing of the Middle and Lower Classes (Netherlands 2018).
Schoeser, Mary, World Textiles: a concise history (United Kingdom 2003).
Wagner, Mayke et al., “The ornamental trousers from Sampula (Xinjiang, China): their origins and biography,” Antiquity 83 (2009), 1065-1075.
Williams, Alan, The Sword and the Crucible: A History of the Metallurgy of European Swords Up to the 16th Century (Leiden 2012).
Wilson, R.J.A., “On Early Thimbles: A Seventh-Century-AD Example from Punta Secca, Sicily, in Context,” Oxford Journal of Archaeology 35.4 (2016), 413-432.
Zhao, F. et al.,“The earliest evidence of pattern looms: Han Dynasty tomb models from Chengdu, China,” Antiquity 91 (2017), 360–374.