Unraveling the (Production) Secrets of an Egyptian Textile

By Julie H. Wertz
May 28, 2020
A large rectangular off-white linen textile lies flat on a large table. Four people stand around the table looking at the textile, a set of large windows behind them. The textile has four purple squares in the middle and pairs of purple bands at either end. The central area contains loops of pile fabric with stripes of red, green, and blue loops bordering the purple bands. A man in the background at far right is standing at a table working.
A large Egyptian cover with inscriptions is examined in the Straus Center for Conservation and Technical Studies. Early Byzantine Egypt, 5th–6th century CE. Linen and wool: tapestry weave and supplementary weft pile. Harvard Art Museums/Arthur M. Sackler Museum, Gift of The Hagop Kevorkian Foundation in memory of Hagop Kevorkian, 1975.41.28.

Textiles have been an integral part of our lives for millennia and serve as objects of both function and beauty. How would this colorful, textured textile have been made in Egypt 1,500 years ago?

For most of human history, producing textiles was done manually, and often partly at home. Modern industry has largely removed this task from our daily lives, and it is now more common for clothing and domestic textiles to be purchased and replaced rather than made or mended. Even when we do make a garment by hand today, it’s significantly different from historical practice because the necessary supplies—such as thread, needles, buttons, or fabric—would likely be purchased outside the home.

Textiles served important functions in Roman and early Byzantine homes: to soften surfaces, separate rooms, decorate, and create comfort. The object pictured above was made in Roman Egypt and has been dated to the fifth or sixth century CE. It is exceptionally large (about 7 feet by 5 feet) and is noteworthy for being so intact. The thick, textured weave would have made it a pleasing object to have at home. While it ultimately served as a burial shroud, its original purpose is uncertain: it may have been used as a hanging or a covering. Because the presence of pile would have added both padding and insulation, it probably served as a cover for furniture.

The arid climate of Egypt helped preserve countless textiles like this one. Made with organic materials and buried in tombs, they were exposed to little moisture and protected from light. Many textiles come from undocumented excavations, and it was common for the decorative parts to be cut from the relatively plain surrounding fabric and sold as art objects. The cuff band pictured below, for instance, was removed from its garment.

A small rectangular textile the size of a sleeve cuff shows two bands of animals running to the right, including rabbits, goats, gazelles, and felines woven in purple yarn with red tongues on an off-white ground. The animals are set into roundels created by vine tendrils.
2004.204 Cuff band: Animals in roundels, Early Byzantine Egypt, 4th–5th century CE. Wool and linen, tapestry weave. Harvard Art Museums/Arthur M. Sackler Museum, Gift of Benjamin and Lilian Hertzberg, 2004.204.

Analytical tests revealed information about the materials used to make the large cover, which will be on display in the upcoming Harvard Art Museums exhibition Social Fabrics: Inscribed Textiles from Medieval Egyptian Tombs, curated by Mary McWilliams, the Norma Jean Calderwood Curator of Islamic and Later Indian Art.

Textile Fibers

Acquiring and processing the raw materials needed to make textiles has always required many resources and extensive labor. Textiles provide shelter and warmth and are necessary for such utilitarian objects as bags and sails. Frequently decorative as well as functional, textiles with colored elements and increasingly complicated designs called for additional raw materials such as dyes, as well as technical advances in weaving, to successfully meld form and function.

Production of this large cover involved a significant amount of raw materials, human capital, and technical know-how. The choice to include colored elements, elaborate designs, and features like pile indicates more labor was needed than usual. We can only guess at the original use of this piece, but we have the privilege to study it today because it was preserved for more than a millennium in a desert tomb. 

Wool and linen, which are found in this object, were the two major textile fibers available to people living in Egypt in the late Roman and Early Byzantine periods (third to seventh century CE). To obtain undyed wool yarn, a shepherd had to maintain a flock of sheep and shear wool at the appropriate time. The wool then had to be carded (combed) to remove debris and untangle the fibers before it was spun (twisted) into yarn. Linen required the cultivation and harvesting of flax, which had to be dried and threshed to remove seeds and debris. The stalks were then retted (soaked in water) to make them soft and scutched (crushed) to remove the woody material before they were heckled (combed) in preparation for spinning. This video from the Victoria and Albert Museum shows how these steps are carried out today using industrial machinery.

Spinning could be done in workshops or at home in the evenings, while waiting for other tasks to be completed, or by younger members of the house. The spinning wheel, familiar to us from fairy tales like Rumpelstiltskin, is a much later invention that was unavailable when this tapestry cover was made. Instead, yarn was spun on drop spindles, a wooden tool that could be carved from a small piece of wood. The fibers could be twisted clockwise or counterclockwise, producing yarn that is described as S- or Z-twist, respectively, for the way the slant aligns with either letter. The yarn in this object is all S-twist.

A seated woman wearing a colorful embroidered vest and white scarf on her head works with a tuft of unspun wool using both hands. A strand of yarn hangs from her right hand connected to an egg-sized ball of yarn wrapped around a stick. Photo: Susanne Ebbinghaus.
A woman in Kyrgyzstan demonstrates the use of a traditional drop spindle after giving a lesson in cleaning and carding the wool. Photo: Susanne Ebbinghaus.

To obtain colored yarn, the spun fibers had to be scoured in a hot alkaline bath to remove natural oils, waxes, and dirt that would interfere with the dyeing process. Dye material from plant sources would have been cultivated locally or imported, and each source required its own preparatory treatments before it could be applied to the yarn. Most dyes of good quality required a pre-treatment after scouring called mordanting, where aluminum or iron salts are applied to the fibers to facilitate dye uptake. Although many plants can impart some color to textiles, the majority are not considered useful for dyeing because the result is not particularly attractive and doesn’t hold fast (that is, resist fading). Dyeing, especially with natural materials of variable quality, was a specialist practice that required a large quantity of water and knowledge to achieve an even, fast color.


Weaving, like dyeing, was a trade that called for specific tools and skills. Using a loom, the weaver would string the warp upon the loom’s frame before adding the perpendicular weft over and under the warp to create the fabric. The width of the loom and number of warp yarns would determine the fabric width and thread count. Archaeological evidence gives us some insight into workshop practice and textile production in ancient Egypt; more information about tools from the time can be found here.

We know from examining damaged areas of this textile that the warp yarns are made of wool and paired to strengthen the structure. At the workshop where it was made, the loom was strung with the warp. An even, adequate tension would have been necessary to ensure that the weave lay flat. A plain weave, such as that seen in most bedsheets, is formed from single wefts that are passed over and under single warps. In a tapestry, the wefts also pass over and under the warps, but they are densely packed together to create a “weft-faced” textile where the warps are largely concealed. Maintaining even tension was important here, too. The dense structure enabled weavers to introduce design elements, like the purple and blue inscriptions on the ends of this piece, by doubling back wefts to create shapes rather than running the full width of the fabric. This created gaps between colors, which either could be closed by interweaving the two sides or left as a slit. The rigid structure of tapestry makes it less suitable for whole garments, but it is ideal for decorative elements like cuffs, sleeves, and tunic bands set into a fabric with a looser and more flexible weave, or for use as a domestic textile like this cover. Here, many areas of purple tapestry bear a linear interlace design created with fine, off-white wool yarns in a technique known as supplementary weft wrapping.

A close-up view of the textile shows, to the left in the off-white section of the work the exposed underlying tapestry weave, annotated with the words “Lost pile exposes underlying tapestry” with a white arrow. In the central part of the textile, another note reads “Exposed paired warps,” with an arrow pointing to the warp yarns. In the middle of the image is a line of red and blue pile loops, labeled “Supplementary weft-loop pile.” On the other side of the loops is a band of purple tapestry weave bearing an off-white geometric design. The note reads “Tapestry (with supplementary weft wrap).”
Several weave structures have been applied in the large cover.

The central area of the cover and the colored stripes of loops are woven with a supplementary weft-loop pile, additional weft yarns that extend in twists from the textile surface. These create a pile fabric with vertical structure, like terrycloth or velvet, which helps retain heat. The looped weft rows alternate with regular tapestry rows, adding texture and dimension to the textile. The red, green, blue, and purple elements of the cover are all woven with wool yarn, while the areas of undyed wefts and loops are linen. Wool fibers are mostly keratin, a protein that also forms the basis of human hair. Linen fibers are primarily cellulose, a carbohydrate that is also the major component of wood and other plant material. The chemistry of protein fibers like wool and silk is better at absorbing mordants for the dyeing process than that of cellulosic fibers like linen and cotton, yielding deeper and brighter colors. For this reason, linen is frequently left undyed.

A young woman leans over a textile that extends over part of a table. The off-white textile bears red, green, and purple decoration and, along the edge on the left, a purple inscription; the woman holds one of the red pile loops with a pair of tweezers. A pair of hands holding a mobile phone extends into the picture, aiming at the hand with the tweezer.
Julie Wertz, with assistance from Katherine Eremin, the Patricia Cornwell Senior Conservation Scientist in the Straus Center for Conservation and Technical Studies, documents the sample location of a red fiber for dye analysis in the paper lab of the Straus Center.

Dye Analysis

To identify the materials used, we carried out analysis of the dyes extracted from microsamples of loose fiber ends by a separation and detection technique called high-performance liquid chromatography with mass spectrometry (HPLC-MS). Two red samples of different hues were taken from separate areas, and one each of green, blue, and purple. Prior to extraction, the samples were examined under a microscope to remove any debris or extraneous fibers. They were also photographed to document the original color, which when seen alongside the color of the extract can be valuable data to interpret the results.

Five images show very close-up views of small different-colored spun fibers samples arranged horizontally. The top three images, from left to right, show blue, green, and red fibers, respectively, while the two at the bottom red and purple fibers.
These images show the colored fiber samples documented at 50x total magnification for the purpose of dye analysis. Photo: Julie H. Wertz.

The dyes found on the blue sample were indigotin and indirubin, similarly structured molecules found in a variety of plants around the world. Significant sources include indigo (Indigofera tinctoria) and woad (Isatis tinctoria); these indigoid dyes were the only reliable way to produce a sufficiently high-quality blue on textiles for most of human history. (In fact, they are still used to create the characteristic blue of denim.) Geographically, Egyptian dyers probably had better access to indigo, which was produced in India on a large scale at the time this textile was made. Since the late 1890s, it has been possible to manufacture synthetic indigo through chemical processes. Although we can identify indigoid dyes in fiber extracts of the sample, there is not enough distinction between natural sources like indigo and woad, or even synthetic indigo, for us to know definitively which one was used.

Indigoids also appear in the green fiber. Natural dyes yield primarily shades of red, yellow, and blue, which can be combined for additional colors. Unlike mixing paint, each dye material has its own optimal conditions, and the baths must be done sequentially. To dye green, the wool fibers were colored in yellow and blue baths. Yellow dyes are the most fugitive of natural dyes, meaning they tend to fade first. This is why greenery in European tapestry with a long display history frequently appears bluish, because the indigoids in the original green have outlasted the yellow tones (see this tapestry). Many yellow dyes are flavonoids, a class of molecules that contributes to plant and flower coloration and performs biological functions, among other roles. Flavonoids appear in a variety of plants, though some can act as chemical markers for a distinct source. Further analysis of the yellows in this green sample and a comparison to known reference material is necessary to confirm which yellow material was used.

An off-white, textured shell with a conic pattern of ridges lies on its side on a grayish-white surface.
This murex shell once housed the shellfish from which Tyrian purple was harvested. Forbes Pigment Collection, Straus Center for Conservation and Technical Studies, Harvard Art Museums, Straus.730.

Purple can be dyed in a variety of ways, and our modern use of the term includes a broader range of hues than was available in the ancient world. There is one natural dye that yields a true purple: the famous Tyrian purple harvested from the glands of a few species of shellfish of the genus Murex, found in the eastern Mediterranean (see above). The name comes from the ancient Phoenician city of Tyre, now in modern-day Lebanon. Thousands of shellfish were required to produce a relatively small amount of dye—in the early 20th century, a German chemist extracted 1.4 grams of dye from about 12,000 mollusks. Because of the intense labor involved, the dye’s use was reserved for high-ranking members of society and was subject to sumptuary laws, initiating the continued association of purple with royalty. Overharvesting of the shellfish in ancient times drove the practice into obsolescence by the Middle Ages, meaning Tyrian purple is a good indicator of a truly ancient object because knowledge of its use declined significantly. Its chemistry is similar to that of indigo, and both dyes must undergo a chemical reduction process, historically achieved through fermentation, before being applied to fibers. (It’s worth noting that Tyrian purple was desirable enough for someone to tolerate smelly fermenting vats of shellfish glands.) This talk about Tyrian purple gives more information about its history and chemistry.

A large glass jar with a flip-top lid and a yellowed paper label bearing a pencil inscription “Madder root grown in E.W.F.’s garden in 1920’s” is filled with cut segments of woody, red-brown madder root, which look like very small sticks.
This jar of madder roots was grown by former Fogg Museum director Edward W. Forbes in his garden in the 1920s. Forbes Pigment Collection, Straus Center for Conservation and Technical Studies, Harvard Art Museums, Straus.3545.

More common among objects that were not of high status or religious significance is purple yarn dyed by subsequent blue and red baths, in a similar fashion to green. Dyes in the purple ground of the cuff band illustrated above include blue indigoids and components such as alizarin and purpurin that correspond to a red dye called madder (see the jar of madder roots above). The purple fiber in the large cover, however, contained no indigoids. The purple dyed without indigoids was probably made using madder alone. Madder is polygenetic, meaning it dyes a variety of colors based on its variable composition and the mordant used. With aluminum it typically produces a bright, true red; with iron it yields a deeper, more purple hue. This assumption is also consistent with the more reddish purple of the large cover compared with the bluish purple of the cuff band.

Analysis of the two red samples of the large cover showed that they were also dyed with madder. There are variations in minor components, which explains the slight difference in shade and indicates that two different madder sources were probably used along with indigo and an as-yet unidentified yellow plant source.

Radiocarbon Dating

Measuring the amount of isotopic carbon-14 (14C) in a sample, known as radiocarbon dating, can provide more information about when an object was made. Living organisms, like plants and animals, accumulate the unstable isotope throughout their lifetimes, a process that stops upon death. We know how long it takes for carbon-14 to break down, which can be compared to the mass of the sample to estimate how many years ago the plants that provided the flax and the sheep that provided the wool died. The result is calibrated against the amount of environmental isotope throughout time, based on previous analyses of tree ring and ice core samples. Analysis at the University of Arizona Accelerator Mass Spectrometry Laboratory determined the material used in this captivating cover likely dates from the early fifth to mid-sixth century CE. This coincides roughly with the date proposed on the basis of art historical comparisons of Byzantine objects, an ongoing project at the museums.


Julie H. Wertz is the Beal Family Postgraduate Fellow in Conservation Science in the Straus Center for Conservation and Technical Studies at the Harvard Art Museums.