Characterizing Colorants on the Menil Collection’s Prisoner Textile

Kari Dodson

Both in sheer size and in the density of its imagery, the Menil’s Prisoner Textile dazzles, especially when one realizes that it is only a fragment of what was once a much longer composition. Its vibrant colors and remarkable condition belie the centuries it has endured. While scholars have devoted attention to sequencing the 10 known fragments and determining the textile’s original function and significance, only two of the extant fragments have been scientifically analyzed, and those explorations remain unpublished. Better-studied textiles from the same time period and geographic area can shed some light on the Prisoner Textile’s fabrication, but extrapolation is only valid to a point—to truly know the Prisoner Textile we must examine the Prisoner Textile itself.

Reconsideration of the textile has already begun, as new radiocarbon dates indicate that the textile may have been produced 50–100 years later than previously thought.(1) The complete study and treatment of the Menil’s Prisoner Textile fragment will unfold over many months and encompass several facets, including an exploration into the coloration of the textile, which offers a glimpse into the types of insights technical art history can offer. Working together, conservators, conservation scientists, imaging specialists, and curators are collaborating with the twin goals of adding to the body of art historical knowledge and calibrating storage and exhibition practices.

Color in textiles can be generated by pigments or dyes. Generally speaking, pigments (used with binders) are insoluble and sit on the surface of the textile, while dyes are soluble and penetrate into the weave structure, sometimes chemically bonding with the fiber. Many dyes are susceptible to light-induced fading, a process known as photodegradation, whereas pigments, particularly those made from inorganic substances, are often more lightfast. Therefore, an artwork’s preservation depends in part on knowing how it is colored, and using knowledge about the lightfastness of the colors employed to provide an appropriate environment (Crews 1987).

To begin our exploration, we mined existing literature, drawing on the work of scholars in varied disciplines. A small and well-documented group of colorants from plant, mineral, and insect sources are commonly encountered on Andean textiles of this period (Phipps 2010; Roquero 2008; de Mayolo 1989; Cajias 1987; Seefelder 1994; Kashiwagi 1976; Boucherie 2009). A few of interest for this textile fragment are:

  • RED DYE – relbunium root (for example, chapi chapi), brazilwood, cochineal insects
  • RED PIGMENTS – the minerals lead red, cinnabar (vermilion), hematite (red ochre)
  • BLUE DYES – indigo, tara bean pods + colpa (iron sulfate), shellfish exudations
  • BLUE PIGMENTS – the mineral azurite
  • YELLOW DYES – logwood, q’olle flower, berberis, molle, yanali bark, ch’illca leaves
  • YELLOW PIGMENTS – yellow ochre (mineral in origin)

We began from the premise that the fragment would contain a combination of colorants from this select palette to create its imagery. This assumption enabled a process of elimination starting with technical analyses designed to distinguish between those colorants, capitalizing on their known chemical compositions and optical properties.

Many of the known pigments from our literature search are mineral-based and contain characteristic elements; for instance, cinnabar contains mercury, and hematite contains iron. X-ray fluorescence (XRF) spectroscopy, a nondestructive technique, can detect these elements and help determine if an inorganic pigment is present. In XRF spectrometry, the instrument directs a narrow beam of X-rays at an object, elements present absorb those X-rays, and then fluoresce unique X-rays of their own. The spectra of the X-rays produced by the object identify elements present, as long as they are heavier than carbon and present in more than trace quantities. Corina Rogge, the Menil’s research scientist, took multiple measurements of each color on the textile and was unable to find any elements that would indicate the presence of mineral pigments (fig. 2). Given the possible options, this would suggest that it is more likely that the colorants come from organic sources; that is to say, they are dyes.

Optical microscopy supports this hypothesis. In photomicrographs taken by James Craven, the Menil’s conservation imaging specialist, the interaction between the colorants and fibers is revealed. If a mineral pigment is present, discrete particles should be visible sitting on top of the threads when viewed under magnification. Dyes are molecular, not particulate, and many thousands of times smaller, so the color they create appears to saturate or stain fibers rather than forming a distinct layer. Indeed, based on these criteria, when viewed under magnification, colorants across the Prisoner Textile appear to derive their color from a dye-type process. In a photomicrograph taken of a blue area, there is no layer of pigment atop the weave structure (figs. 3, 4). Instead, we see that the individual fibers have taken up the blue color in the manner expected of a dye.

Pigments and dyes absorb visible light, each at a different wavelength, which is why we perceive them as different colors. The wavelengths of light that are not absorbed are instead reflected. Normal photographs are reproductions of what we see made when cameras capture the visible light reflected from a surface. Documenting how colorants reflect light from regions of the spectrum we can’t see, including the infrared and ultraviolet ranges, is often useful as well, particularly because not everything that appears the same color to our eye absorbs and reflects different wavelengths of light equally. That is to say, even pigments and dyes that are visibly similar to the human eye will have different absorption and reflectance spectra that can be used to distinguish them.

Multiband imaging is a technique in which a series of images is captured with different filters in front of the camera lens, each designed to allow only specific, narrow ranges of wavelengths of light to pass through and be detected. This series of images can reveal unique material properties that the eye alone is unable to detect (figs. 5, 6). In 2014, a team of imaging specialists, conservators, and conservation scientists at the Smithsonian Museum Conservation Institute developed a multiband imaging technique that can identify indigo (Webb, Summerour, and Giaccai 2014). Indigo most strongly absorbs visible light at ~660 nm and reflects the most light at ~800 nm, which is in the infrared range (Leona and Winter 2016). By capturing band pass images in these ranges and then subtracting one from the other, an image is created in which areas of the textile colored with indigo appear bright white.

Since both the XRF and the microscopy already performed had indicated the use of dyes, and knowing that indigo was a commonly encountered blue dye on textiles from the Andes, Craven applied this nondestructive method to the Prisoner Textile. In the subtraction image of the Menil’s fragment (fig. 7), we can infer that every area that appears white is almost certainly colored using indigo dye. Interestingly, that not only occurs in both dark- and light-blue areas (including preparatory sketches and details such as eyes, fingers, toes, and ropes) but also in areas we would describe as green, suggesting that green was achieved by overpainting yellow areas with indigo dye. Now in possession of a “map” of indigo, future plans include confirmation of this identification (eliminating the possibility of a blue with a similar reflectance curve) using fiber optic reflectance spectroscopy (FORS), as well as microfade testing of all colorants to measure their lightfastness.

At this point in our explorations, we have only just begun to understand the Prisoner Textile’s colorants. Indigo is a relatively lightfast dye with a linear fading rate, so it is unlikely to be the most sensitive of the colorants used on the textile (Crews 1987). The more we are able to divine about the susceptibility of the other colorants to light-induced damage, the easier it will be to plan for the textile’s preservation. For the time being, exhibition parameters such as light levels and display duration will be conservative, and will adjusted as we uncover more. Every bit of information, no matter how small, fills lacunae in our knowledge about this piece in particular, about its sister panels by extension, and about Andean artistic traditions in general. Not only will this knowledge continue to inform the preservation of the textile, but the artists’ material choices may have deeper meaning to those studying the art, language, science, technology, and economy of this period, as the objects they affect are nothing if not products of their time, bearing fingerprints of their culture. By undertaking technical art historical studies such as these, scholars are better poised to improve our understanding of these captivating objects, and the people who made them.


Antúnez de Mayolo, Kay Ketchum. 1989. “Peruvian Natural Dye Plants.” Economic Botany 43 (2): 181–91.

Boucherie, Nathalie. 2009. “Telas Pintadas Nasca: Pigmentos y Tecnica Pictural. Primeros Resultados.” Actas IV Jornadas Internacionales Sobre Textiles Precolombinos, 79–92.

Cajias, Martha. 1987. Manual de Tintes Naturales. La Paz: SEMTA.

Crews, Patricia Cox. 1987. “The Fading Rates of Some Natural Dyes.” Studies in Conservation 32 (2): 65–72.

Kashiwagi, K. Maresuke. 1976. “An Analytical Study of Pre-Inca Pigments, Dyes, and Fibers.” Bulletin of the Chemical Society of Japan.

Leona, Marco, and John Winter. 2016. “Fiber Optics Reflectance Spectroscopy: A Unique Tool for the Investigation of Japanese Paintings.” Studies in Conservation 46 (3): 153–62.

Odegaard, Nancy, Scott Carroll, and Werner S. Zimmt. 2000. Material Characterization Tests for Objects of Art and Archaeology. London: Archetype Publications.

Phipps, Elena. 2010. Cochineal Red: The Art History of a Color. The Metropolitan Museum of Art Bulletin. Vol. 67. New York: The Metropolitan Museum of Art.

Roquero, Ana. 2008. “Identification of Red Dyes in Textiles from the Andean Region.” In Textile Society of America Symposium Proceedings.