South America, with its geographic extremes—the Andes, the Amazon Basin, the deserts of Peru and northern Chile—elicited considerable ingenuity from its early inhabitants. Among their adaptive responses was the fabrication of appropriate clothing, blankets, and lightweight fiber containers. Handmade textiles created today in Andean cultures in Ecuador, Peru, and Bolivia, and in parts of Colombia, Chile, and Argentina, not only represent an ancient tradition, but remain as much a part of everyday life in many communities as are cars in North American cities.
Textiles made of cotton grown at low elevations on the coast and in the Amazon, or of fiber from the coats of the great camelid herds (llamas, alpacas, vicuñas,and guanacos) in the Andean highlands, can be traced back at least 5,000 years, to a time long before the advent of the Inca Empire (c. ad 1250–1532), which also prized finely woven pieces.
Because the desert coast of Peru and northern Chile is so dry, buried textiles—often in association with mummies—are often found in remarkable states of preservation. Such cloth has stories embedded in its warp and weft. Junius B. Bird (1907–1982), Curator of South American Archaeology at the American Museum of Natural History in New York City (AMNH), was one of the first to recognize the tales that even faded, soiled fragments of cloth could tell. He influenced several generations of archaeologists, anthropologists, art historians, and museum curators, and indirectly influenced many more, including me. Bird recognized the importance of preserving and exhibiting archaeological textiles and their value in documenting the lives and cultures of the people who made them. By making such research public, investigators across academic disciplines have been able to build on one another’s work.
In 1946 and 1947, Bird conducted archaeological excavations at Huaca Prieta, a preceramic site in the Chicama River valley on the north coast of Peru, documenting stratigraphic levels dating from about 3100 to 1300 bc. He and his crew excavated more than 9,000 ragged textile and cordage fragments, the kind of artifacts that were considered unimportant, mere detritus, by most archaeologists of that era.
More than 70 percent of the fabrics unearthed were created by the laborious hand technique of twining, in which two or more weft yarns are twisted around each other as one is wrapped over and one under each warp yarn. One small, twined textile that Skinner analyzed was found almost intact in an early burial. Nearly eleven inches long by more than eleven and a half inches wide, with stripes of diamonds, solid colors, and stylized condor heads in profile, it is the oldest known patterned Andean textile. Most of the blue-dyed selvages (self-finished edges) were preserved, while the body of the textile comprised natural brown, natural tan, and red-pigmented cotton yarn. A later piece from Huaca Prieta contains a front view of a spread-winged condor with a snake in its stomach [see images above ]. Skinner observed that other Huaca Prieta textile motifs depict birds, felines, crabs, double-headed snakes, and humans, which became mainstays of Andean textile iconography for the next 4,000 years.
Although the Huaca Prieta textiles were mainly fragments, many with no apparent identifiable motifs, Bird noted the information they contain, the tales they tell. For example, a textile’s fibers indicate whether wild plants, domesticated varieties, or both were used. If a cloth has realistic motifs, they can reveal not only that the artisans were careful observers of their environment, but which features they thought important. The abundant cottonseeds found at Huaca Prieta suggest that cotton was processed on the spot and that the inhabitants spun their own yarn. Fishnets with stone sinkers illustrate extensive use of the abundant marine resources available to coast dwellers. The kinds of dyes used tell us about a culture’s technological sophistication. If yarn is lumpy, it might mean that it was spun by a child just learning to spin or by an elder with arthritic fingers.
Here the tale takes a twist. Spinners convert raw fibers into yarn by cleaning them, then drawing them out and twisting them together, usually with a stick (spindle) and a small cylinder (whorl) to serve as a flywheel. A spindle can be made on the spot with a pampas grass stem, with a tiny potato for the whorl. The archaeological record is replete with spindles and their whorls made of more durable materials.
It’s a rare travel article about Peru that does not have a photograph of a woman spinning as she walks, her spindle dangling as it twists the fiber. There is truth in this stereotype, but in some communities males also spin, or females spin and males ply (twist together lengths of spun yarn, also using a spindle). This kind of spinning, called drop spindle, is done with the spindle held vertically. Sheep’s wool (introduced by the Spaniards) or camelid fiber is wrapped around the spinner’s wrist or attached to a distaff, frequently a sturdy stick about a yard long. Because wool and camelid hair have microscopic scales, the fibers hold together as they are twisted. The spinner pulls the fibers from her bundle with her left hand (called drafting), while snapping the spindle with the thumb and fingers of her right hand. Viewed from above, a spindle held vertically whirls clockwise (the most natural direction when using the thumb and fingers), and the yarn that is spun on it has a twist that resembles the slope of the letter Z.
Bird considered cotton the initial fiber for the Andean textile tradition; the domestication of camelids (llamas and alpacas) or the use of the hair of wild guanacos and vicuñas came later. The wild species had to be captured and released or killed. Cotton, wool, and hair fibers all require processing to remove debris and align the fibers. Family members pick twigs, burrs, and dirt from alpaca hair or sheep’s wool, then pull the fibers more or less parallel into a thin sheet that is joined with other sheets and wrapped around the wrist or tied on a distaff.
An indigenous cotton (Gossypium barbadense), the main fiber used at Huaca Prieta, has a short “staple” (length) and a smooth surface, so the weight of a drop spindle would tend to pull the yarn apart before the fibers have had a chance to twist and hold together. One solution is to rest the bottom of the spindle on the concave surface of a potsherd, gourd, or something similar. Another solution is to support the spindle horizontally, twirling it as before but with the yarn now spun from the other end of the device. That results in a yarn with a so-called S-twist, the mirror image of a Z-twist [see illustration below].
Two lengths of yarn are usually twisted together (plied) in the opposite direction of the spinning, which produces a stronger, more even yarn. The majority of Huaca Prieta cotton textiles are S-spun and Z-plied. An alternative found at the site are warp (lengthwise) strands that were grouped in pairs for weaving, but not plied.
The next step is warping, setting up the foundation fibers for the textile, which is most frequently done by setting two stakes in the ground at a distance that matches the desired length of the textile and then wrapping yarn around them over and over in a figure-eight pattern. Each end of the warp is then transferred to a sturdy wooden loom bar, either by lashing the warp yarns to it or just slipping the looped ends of the figure eights over the bar. The warp needs to be under tension for weaving to be successful. Andeans have various solutions: one loom bar can be attached to a tree or beam and the other to a strap (backstrap) that goes around the weaver’s back. Another solution, particularly in such treeless areas as the Peruvian and Bolivian altiplano, is to lash the loom bars to four stakes in the ground. A third solution is to make a vertical frame.
In much of the northern Andes, males are the warpers and weavers, while in the southern Andes females predominate at these tasks, but there are many exceptions. All traditional Andean textiles are woven to size or made from two or more pieces joined side by side. The concept of tailoring—cutting garments to shape from woven cloth—is a European introduction.
Anyone doing the simplest weaving by hand in first grade, perhaps with six warps and using a blunt needle to interlace the weft yarn, would start weaving from left to right or right to left, usually going over the first warp and under the next, ending with going under the sixth warp. The weft yarn is then pulled through and beaten (pushed) tight with a straight, flat piece of wood (the beater or sword). The second pass then returns in the opposite direction. The result is a very simple, plain weave band. But if a larger textile is desired, going over-under by hand is not feasible, so strings are looped around selected warps (for example, every other one) and then tied to a stick. By lifting the stick, the strings, called heddles, in turn lift all the selected warps together, making a space (called the “shed”) through which it easy to pass the weft yarn beneath them, greatly speeding the weaving. For plain weaves, the alternate warps can be manipulated by a round wooden rod (shed stick) instead of looped into heddles. Archaeologists have excavated intact pre-Hispanic looms with this basic setup, even with partially woven textiles.
The most common Andean style is warp-faced weaving, where the warps predominate on the surface. To make motifs, such as diamond shapes, at least one pair of differently colored yarns is needed. The weaver uses her or his fingers to “handpick” certain warp yarns of one color, while dropping their “duplicates”—the paired strands of the second color—down. The result is a double-faced textile, with the same motif on both sides but in contrasting colors. Children start with simple bands and move on to larger, more complex pieces.
Textiles, however, are mute with respect to some techniques and processes, uses, gender roles, and meanings. What were the condor cloths used for? Did the Huaca Prieta artisans include condors in their textiles because the bird was sacred to them, or did they just like depicting the bird? The Incas considered condors to be messengers to the gods, but that culture came several thousand years later, and there is no way we can extrapolate backward from the Incas to Huaca Prieta. And what was the division of labor?
As Bird’s analysis of the Huaca Prieta finds progressed, he realized the importance of research on contemporary spinners and weavers and sent textile scholars off to remote Andean villages, engendering a flowering of research in the 1970s and 1980s. He valued the data gathered not only for the insights it offered into pre-Hispanic cloth, but for what we could learn about contemporary Andean textiles and cultures in their own right.
We know from the archaeological record and from Spanish chronicles that the Incas were as textile-oriented as their antecedents. Ethnic groups conquered by the Incas were required to wear their traditional hairstyles and dress, as a way to identify and control them. All Andeans paid a tax to the state in labor or in kind (the Incas had no currency), some serving as soldiers, builders, stone masons, couriers, farmers, camelid herders, spinners, dyers, weavers, collectors of dyestuffs, potters, metallurgists—whatever the empire needed. Most worked in their home communities and sent goods to Cuzco, the Inca capital. Others served labor rotations in Cuzco or in the army. Reciprocity is a core Andean value; the families of men serving rotations away from home were fed from communal stores. Tawantinsuyu was a stratified society; the Sapa (supreme) Inca and nobility wore sumptuous garments made from the finest vicuña and alpaca, and ornaments of silver and gold.
The code remained unbroken for nearly 400 years until French scholar Sophie Desrosiers, of the School of Higher Studies in Social Sciences in Paris, deciphered Murúa’s instructions. Her findings were presented at the 1984 Junius B. Bird Conference on Andean Textiles in Washington, D.C., and published in the 1986 conference proceedings. Desrosiers had researched weaving in present-day Potolo, Bolivia, and used that knowledge to make a sample belt from the coded instructions. She decided that Murúa’s a, c, e, and v represented four colors, and after much research concluded that they stood for the Spanish amarillo (yellow), colorado (red), encarnado (purple), and verde (green). The numbers (with x representing 10) she took to indicate the order and number of warps visible on the surface of the belt, forming the motifs.
The twelve heddle numbers were a mystery, however. Desrosiers did not think that they referred to twelve distinct heddles, as no known Inca looms with multiple heddles were known to exist. However, the rainy Andean highlands are not conducive to textile preservation; most pre-Hispanic textiles are found on the coast. Subsequently, at the AMNH, Desrosiers found a pre-Hispanic example of a belt with the same motifs, but woven by a slightly different technique. That belt, identified as Inca, had been excavated in Chimbote on the Peruvian coast by Adolph Bandelier in an 1894 expedition sponsored by the museum. It measures 131 inches long by 1 1/2 inches wide and is made from natural-colored camelid yarn in black, light brown, dark brown, and (now somewhat yellowed) white [see photograph below]. After examining this belt, Desrosiers concluded that Murúa’s numbers referred to multiple heddles after all. Evidently by Inca times, weavers had figured out how to semi-mechanize the process of weaving repeated motifs, eliminating the handpicking of each row through the use of multiple heddles. Murúa’s instructions specify how to create a repeated motif composed of four differently colored triangles arranged to form a diamond. Each line of code specifies how to handle all 104 yarns and their duplicates, since when one warp is picked up, its duplicate drops to the other side of the belt, giving a double-faced fabric considered by the Incas to be the most prestigious textile. The four warps and their duplicates on each edge of the belt are not included in Murúa’s picking sequence because they simply form a plain-weave border in a fifth color.
The weaving tradition documented by Murúa was thought lost until, in 2002, when on a research trip to Huamachuco District, in northern Peru’s La Libertad Region, I was stunned to see that a colleague, Joseph Fabish, had wrapped his sleeping bag with a contemporary belt that was almost identical to Murúa’s. In 2004, we found female weavers around La Yeguada who still weave what they call sara or sarita belts with multiple heddles on the backstrap loom, just as Murúa wrote. The weavers did not know the meaning of sara (corn) or sarita (little corn), however, because Quechua has not been spoken in the region for more than a century. For them, these were just names for belts with this motif.
I asked weaver Maria Polo the names of the colors in the diamond-shaped motif (each comprising four triangles), hoping to learn if Desrosiers had correctly identified the colors symbolized by a, c, e, v. Doña Maria gave me a funny look and said the colors weren’t the issue: the diamond shape is a coco (seed or fruit pit), and the four triangles that made it up were puntas de maíz, four growing tips of corn, or tiny corncobs. In San Ignacio the women did not know what sara meant, either, but this reply provided the first documented instance of contemporary local weavers associating the belts with corn.
Unraveling the mystery of the sara belts that Murúa described has involved multiple researchers, library archives, a pre-Hispanic belt in the AMNH collection, a Textile Museum publication, and anthropological fieldwork. The result is evidence of a continuous Inca weaving tradition going back at least 400 years, a remarkable story of cultural survival. Junius Bird would have been pleased.