The curator-conservator collaboration: remembering Roy Sieber.

At his memorial service held at the National Museum of African Art (NMAfA) in October 2001, it was noted that the conservation lab was second only to collection storage as Roy Sieber's favorite place to spend time in the museum (Fig. 1). Dr: Sieber shared our focus as conservators in defining authenticity and understanding the technical aspects of African art. His interests, which arose from his art historical perspective, are well known and well documented. He was intrigued by our interest in the application of technical analysis to the complexities of indigenous surfaces, construction methods, and postcollection enhancements of African objects. Dr. Sieber recognized the benefits of the relationship between curator and conservator: conservation expertise, while not necessarily providing definitive answers to his questions, could contribute to understanding the story that an object has to tell. We met these interactions in the lab with enthusiasm and knew that in the exchange we would gain valuable cultural insight to add to our own examination repertoire.

[ILLUSTRATION OMITTED]

Twenty years ago, when conservation science was beginning to be routinely applied to art objects, The New York Times printed an article focusing on the sometimes antagonistic relationship between curators and conservators when considering questions of authenticity (Rostron 1980). Many curators mistrusted the use of science in their realm of connoisseurship, which they believed could only be developed by vast exposure and experience, and which required "an eye." Conservators who were eager to offer useful technical information inadvertently reinforced this mistrust, as their science could not always provide the definitive answers sought by curators. It is refreshing, then, that Roy Sieber, the founding father of African art connoisseurship in the United States, appreciated the contribution science could make to the discipline of art history.

Examination in the conservation lab routinely describes technology, identifies materials, reveals restorations, and analyzes deterioration. More comprehensive findings, however, come from directed inquiries that refer to African cultural activities and incorporate a curatorial knowledge of style, manufacturing techniques, materials, and contexts of use. One such inquiry arose when the National Museum of African Art acquired a group of Tuareg hair knives whose handles were constructed in the traditional way by laminating together discs of different materials (Fig. 2). Typically the materials would be metals, woods, and coconut shell. The curators, however, suspected that this group of objects was contemporary and asked us to identify the media. One art historian cites the use of plastic flooring as a wood substitute (Loughran 1995), and indeed, we found that a ureaformaldehyde plastic was a component of the NMAfA knives. Developed in 1928, this material is sometimes referred to generically as linoleum. Our analyses helped to establish the earliest possible date of manufacture and to identify the increasingly diverse trade network that enabled Tuareg craftsmen to use new materials in traditional ways. Thus, the dialogue between curator and conservator produced more than a simple identification of materials.

In light of Dr. Sieber's interest in the collaboration between curator and conservator, it seems appropriate, in this volume dedicated to his memory, to include a brief overview of some analytical methods applicable to the study of surface, restoration, manufacture, and materials specific to African art. When we investigate specific questions, the application of analytical techniques should flow logically from microscopic examination to imaging procedures to techniques that identify specific chemical elements or compounds in a material. One should remember that any of these techniques have limitations as well as advantages. The least destructive methods--those which do not require samples--should be employed initially.

Microscopy

Microscopy is helpful in examining both materials and object surfaces. Tools range from stereomicroscopes with high-quality optics for low to mid-range magnification (10-500x) to sophisticated electron microscopes for high magnification ([greater than or equal to]100,000x). Specially mounted stereomicroscopes can be easily maneuvered to view surfaces at difficult angles and allow for three-dimensional, detailed visual examination. Other stereomicroscope systems, equipped with fixed viewing stages, permit detailed observation of minute samples removed from objects. If equipped with polarizing filters, these microscopes can also facilitate the identification of fibers, pigments, and particles. Rather than using visible light, electron microscopes form detailed images by directing a beam of high-energy electrons onto the surface of a sample. The depth of focus and high-resolution magnification yield data about technology and composition. Electron microscopes are usually equipped with an analytical system that also allows for simultaneous elemental analysis of the sample's surface.

For African art, microscopy is especially applicable to the examination of metals, painted surfaces, wood, textiles, and ceramics. Looking at the microstructure of metals can lead to the identification of techniques and postcasting working methods, and help evaluate alloys. In addition, through systematic studies of metallographic structure, the researcher can determine general manufacturing techniques to establish a baseline and thus identify variations. Corroded metal surfaces signal the effects of environment, postcollection coatings, and other treatments. For example, NMAfA's Benin collection includes many works whose surfaces were treated after the objects were removed from their original contexts. An unaltered surface can be seen on one plaque (Fig. 3), where indigenous red earth is detectable in the interstices of the design. On the other hand, the various metals expertly combined in a Benin pendant (Fig. 4) would originally have been distinguished by the differences in coloration, including that of the inlaid iron eyes and textured brass strip at the nose. However, the natural coloration is hidden under an coating of pigment (cadmium sulfide), applied to imitate clay investment. In another plaque (Fig.5), Prussian blue pigment, meant to replicate copper corrosion, obscures much of the surface.

[ILLUSTRATIONS OMITTED]

Under magnification, these altered surfaces do not exhibit the naturally occurring corrosion layers, such as red cuprous oxide and green, basic copper carbonate, that might be expected. Solvent tests as well as instrumental analysis (see below) can confirm their inappropriate nature. In addition, it can be seen that "protective" coatings applied in the early twentieth century to Benin materials have caused the surfaces to appear darker and produced corrosion while leaching out alloying components of the castings. High magnification supplemented by instrumental analysis can identify these Western modifications (Fig. 6).

[ILLUSTRATION OMITTED]

One can also use microscopy to examine the painted surfaces of African artworks. These studies might involve looking at the stratigraphy of paint layers to understand the history of indigenous refurbishment or Western embellishment of the surface (Figs. 7a, b). Sometimes this can be accomplished under low magnification, but often it is useful to make a cross-section slide from a minute sample of pigment to view the layers more closely. In addition, dealers and middlemen have been known to strip objects of their colorful paint to suit the Western aesthetic. Detected under high magnification, evidence of residual paint might indicate a stripped surface. Polarized light microscopy (PLM) allows one to identify pigments as well as fibers (e.g., wool, cotton, coconut) and particles of substances such as plant starch and kaolin (Fig. 8). Using PLM to interpret the morphology and refractive indices of these materials can provide another piece of information for the researcher wishing to evaluate an object's authenticity.

[ILLUSTRATION OMITTED]

Although microscopy is useful in the examination of wood types, it cannot be relied on to determine provenance or confirm authenticity. A common belief is that the macroscopic features of wood--the color, texture, and grain pattern--are sufficient for accurate identification. For woods customarily used in European, American, and Asian furniture, it may be true that connoisseurs are adept at identifying wood, using its common name, like oak or mahogany. However, these features can be variable. Potential identification of the wood type used in an ethnographic artifact requires microscopic examination of the cell structure, and the results can only be interpreted within the parameters and, therefore, limitations inherent in the technique. The assignment of provenance based solely on microscopic wood identification is empirically virtually impossible.

The binomial system (genus and species) used to name trees and other flora originally was formulated using each plant's distinguishing reproductive characteristics, such as flowers, fruits, or cones, as well as vegetative characteristics, such as leaves and bark. Wooden artifacts usually incorporate only the internal components of the tree. Microscopic analysis of these internal components can often assign genus (in common terms, pine, oak, etc.) very accurately.

However, a genus may comprise hundreds of species, and these usually cannot be differentiated under the microscope because of their similarities in anatomy. Furthermore, many genera have species in more than one country or continent. If a wooden object was made from a genus that occurs in Africa as well as the Americas, it would be difficult to pinpoint the origin of the wood. Clearly, it may be possible to identify the general geographic region for a wood found in Africa, but not necessarily cultural provenance. On the other hand, when the absolute provenance of an object is already known, as might be the case in established ethnographic collections such as those of the Royal Museum of Central Africa in Tervuren, Belgium, or the Rijksmuseum voor Volkenkunde in Leiden, it is legitimate to deduce species once the genus has been determined microscopically.

It is significant to note that wood reference databases that could provide useful comparisons for ethnographic objects are not readily accessible; available references focus solely on wood used commercially. Finally, we must mention that wood analysis requires destructive sampling, thus limiting its potential usefulness to museums and private collectors. The process requires a fragment measuring 3cm x 3cm x 8cm in order to prepare slides featuring three sections of the wood (the tangential, radial, and cross-section) and accurately identify them (Fig. 9).

[ILLUSTRATION OMITTED]

Infrared and Ultraviolet Light 

Photographic techniques that take advantage of invisible wavelengths may reveal information about artifacts that the eye cannot discern in visible light. The electromagnetic spectrum is composed of a range of radiation frequencies from radio waves to X-ray and gamma waves. A very small portion of the spectrum (400-760nm) produces waves in the visible light range that are discernible to the eye. Infrared light (IR, 1-750nm), present in ordinary light sources, may be reflected or transmitted by paint, inks, and dyes. When recorded on infrared film using appropriate camera lens filters and light filters, the resulting image can yield information about damage, restorations, and faded or worn inscriptions. Infrared photography, which is widely used to view underdrawings in painting conservation, has been applied to the Museum's six Ethiopian icons (Fig. 10) (James unpublished).

Perhaps the most useful imaging technique available makes use of ultraviolet radiation (UV, 400-10nm); objects are placed before an ultraviolet light source and viewed in the dark. This accessible, inexpensive technique forms part of the basic repertoire for curators and conservators examining works of art. Materials exposed to UV light, which is also invisible to the eye, may react in two ways. The radiation may be completely absorbed, resulting in a matte or darkened appearance; or the UV radiation may be absorbed and then re emitted as fluorescence (visible light). Absorption and fluorescence under UV light can indicate areas of restoration, such as fills in areas of loss, overpaint, or Western coatings (Fig. 11). With the correct filters and lighting, both/UV absorption and UV fluorescence can be captured on film. Neither UV nor IR imaging requires sampling, and when used properly, they do not damage the object in any way.

[ILLUSTRATION OMITTED]

X-Radiography

Methods that enhance visual examination, most notably X-radiography, were among Dr. Sieber's favorite tools (Sieber & Celenko 1977). A familiar, fundamental technique, X radiography uses high energy radiation to penetrate materials and record an image digitally or on photographic film. The image captured shows variations in density within a single material and the relative densities of materials in objects fabricated from multiple materials. X-radiography is particularly useful to the art historian for understanding the technology of manufacture (Figs. 12a, b), for detecting repairs and restorations (Figs. 13a, b), for viewing obscured interior areas such as cavities in Kongo or Bembe nkisi figures, and for aiding in attributions (Fig. 14a, b).

[ILLUSTRATIONS OMITTED]

The method can be applied to a broad range of materials, including ceramics, metals, wood, ivory, and composites. Low-voltage imaging ([less than]40kV) has been shown to be useful for looking at low-contrast, low-density features such as watermarks on paper. High-voltage equipment (200 300kV) is necessary to penetrate very dense materials like some metal alloys. An X-ray generating tube with a wide range capability (15kV-325kV) is the most flexible and can provide the largest scope of results. Application may also be limited by the size of the X ray facility, which can range from an oven-sized cabinet to a large room.

Different densities of materials present a potential obstacle when one is trying to obtain information about the interior contents or features of objects, such as the cavity in an nkisi. If the density of the container's material (e.g., wood) is significantly greater than that of the contents (e.g., feathers, cloth, plant substances), the image of the contents will be "burned out" by the higher voltage required to penetrate the exterior. It will be impossible to see what is concealed within the object. One caveat pertains to works that may subsequently be dated using thermoluminescence (TL): ceramics and metals with residual clay investment material must be sampled for TL dating prior to exposure to X-rays, as the radiation interferes with the results.

Chemical Tests

Relatively simple chemical tests can provide information relating to date of manufacture (terminus post quem/ante quem), assist in identifying indigenous materials, and determine the presence of restoration. Following visual examination, solubility tests, for example, can be helpful in distinguishing original from restored surfaces. However, complex mixtures on surfaces and changes to surfaces over time due to degradation preclude generalized statements about which solvents will differentiate indigenous patinas from Western finishes. Generally, one can readily remove the paint or pigmented wax of restorations with organic solvents, taking care not to damage culturally applied surfaces (Fig. 15). Artificial metal patinas and artificial corrosion are often achieved by mixing pigment with adhesive; these may also be solvent-sensitive. Examples from the NMAfA Benin collection were cited above. However, naturally occurring corrosion might also be soluble in certain organic solvents, particularly those formed from the interaction of oils with metals (Fig. 16).

[ILLUSTRATIONS OMITTED]

Microchemical tests are a form of qualitative analysis used to identify the nature of a variety of materials including metals, pigments, proteins, plastics, and salts. These tests are simple and relatively inexpensive, making them accessible to smaller conservation labs. Microchemical (or spot) tests involve the diagnostic reaction of minute samples of the unknown substance with specific chemical reagents (Fig. 17). The main limitation is the potential for contamination or interference from other compounds that may be present in the sample. In addition, if the test is carried out in situ, which is sometimes the case, the object's surface may be slightly discolored or marred.

Instrumental Analyses

Instrumental analyses provide more data on composition and may be used to corroborate results from microchemical tests. Many large museums have analytical equipment on-site, and university science departments are often willing and eager to facilitate art historical research. In addition, private companies offer technical analyses on a contractual basis. Analytical techniques most commonly applied to ethnographic material are discussed below.

X-ray fluorescence spectroscopy (XRF) is a widely used, nondestructive analytical technique used to identify the presence of individual chemical elements that may be a component of metals, ceramics, glass, or pigments. A designated spot on an artifact is subjected to high energy produced by an X-ray tube. Excess energies emitted from the material are recorded on a spectrograph as peaks characteristic of the constituent elements in the material. The height of these characteristic peaks indicates the quantity of each element present. At NMAfA we applied this technique to determine the alloy composition of a metal Bamum bracelet (Fig. 18). We know that alloys with a high zinc content ([greater than]33%) are of more recent manufacture (Craddock 1985). The bracelet was determined to consist of 87.17% copper, 11.23% zinc, 0.56% lead, 0.53% iron and 0.41% tin. The low zinc composition, therefore, does not contradict the nineteenth-century provenance assigned to this piece. XRF cannot be used to detect elements with an atomic number below that of calcium. With the appropriate instrumentation, one can carry out XRF directly on the surface of the object itself, and thus avoid sample preparation.

X-ray diffraction (XRD) is a technique used to determine the composition of crystalline materials and simple crystalline mixtures. Materials that have crystalline structures are typically inorganic in nature (metal, ceramic, stone, some pigments). Unlike XRF, XRD is capable of determining precise compounds raffler than identifying the key elements present. When X-rays are introduced into a material, they are diffracted in a characteristic pattern, based on the regular, repeating crystalline structure. These patterns are then compared to known reference patterns for identification. For example, white accretions on Magdalene Odundo's Untitled 1 (Fig. 19) were identified by XRD as calcium hydroxide, an insoluble salt; these resulted from a combination of Odundo's raw material choice and her firing technique (Moffett et al. 2002). XRD requires only a very small sample (approximately 1.0mm[SUP.2]) or; with the appropriate equipment, the analysis can be performed in situ on the object's surface.

The analytical techniques called Fourier-transform infrared spectroscopy (FT-IR) and gas chromatography (GC) are used to characterize organic materials. FT-IR analysis records the characteristic energy level of individual bonds in these materials. For example, proteins are characterized by the presence of N-H bonds, which have a specific vibration frequency. A material's spectra of absorbed energies are compared with spectra of known reference materials. We employed this technique when a Kabyle vessel arrived at the Museum with a white bloom across parts of the surface. FT-IR confirmed that the bloom was more likely a part of the original natural resin coating than a Western addition. The coating had been visually altered by the separation of a minor, perhaps oily component from the resin (Fig. 20). Complex mixtures and degradation of the material can affect the success of FT-IR analysis.

[ILLUSTRATION OMITTED]

In gas chromatography (GC) complex mixtures are separated into individual components for identification. It is not uncommon to characterize a material by FT-IR and then further refine the identification with GC analysis.

FT-IR and GC are applicable to organic materials, both naturally occurring and synthetic, that have been used to con struct ethnographic cultural objects. They are also applicable to organic materials that may have been applied to these objects after collection, such as adhesives, coatings, waxes, and plastics. These techniques, requiring minute samples, can distinguish between broad classes of materials such as oils, waxes, carbohydrates, proteins, resin, and synthetics. For example, if the accretion on an object is suspected to be blood, analysis can only determine if it is or is not a protein. It is not always possible to differentiate material within these classes.

Artifact Dating

The dating of African objects using instrumental method s can be difficult. Techniques that have been developed for archaeological objects with associated contextual material are not easily applied to ethnographic or archaeological objects removed from their original contexts. Perhaps one of the best-known techniques is thermoluminescence (TL), which is used on ceramic works. It is a technique for relative dating, and the art historian may find it helpful in distinguishing between objects of some antiquity and modern replicas or fakes. Obtaining reliable absolute dates, however, may be complicated by a lack of archaeological contextual material, such as soil samples, or by the need for multiple samples from an object, because old sherds are sometimes incorporated into a ceramic object of recent manufacture. TL dating for African material is widely discussed in the art historical literature. Ethical issues concern not only the need for multiple samples, which damage the work, but also the forger's ability to falsify the radioactive content of a ceramic body (Mcintosh 1989; McIntosh & McIntosh 1986).

To note, the radiocarbon dating (C14) of organic ethnographic material is restricted to objects that are more than 200 years old. Organic objects typically do not survive that long in aggressive tropical climates. Consequently, the best dating techniques for African artifacts remain stylistic analysis and terminus post/ante quem, where the earliest or latest possible manufacturing date is derived from materials analysis or technological studies.

NMAfA conservators share Dr. Sieber's belief in the "primacy of the object" as a foundation for professional and intellectual inquiry (Ross 1992:41). To this inquiry; the conservator brings a wide repertoire of material technology studies that includes a variety of analytical tools to contribute to the understanding of African art. These tools can be used for identifying materials used in manufacture or for characterizing products of indigenous use, corrosion, or deterioration residues found on African objects. These tools can aid in research when studying either a body of similar objects or objects with similar cultural attributes. For example, NMAfA recently completed a technical study of Ethiopian icons (James, unpublished), and it is considering a project to study Yoruba beadwork. The latter proposal was precipitated by the acquisition of a royal vest (ewuileke) whose beads, threads, and craftsmanship indicate routine reworking and regular embellishment (Fig. 21).

[ILLUSTRATION OMITTED]

Such research is more comprehensive and contributes most significantly to understanding African art when developed in collaboration with the curator. Dr. Sieber taught that the responsibility of the art historian was to assess an object based on criteria of the culture that produced it--to ascertain that it is central to its style, that it is central to its type, that it is historically valid, and that it was not only accepted within the culture that produced it but had a place and use within that culture (Ross 1992). Simply stated, it is through a highly trained and discerning eye that a curator develops this skill. Similarly, it is significant to note that Dr. Charles Tumosa, Senior Research Chemist at the Smithsonian Center for Materials and Education, teaches conservators that analytical equipment, no matter how sophisticated, is an extension of the human senses (personal communication, 2003).

The development of collaborative research begins by formulating the questions that characterize the foundation of scientific inquiry: Why are we undertaking analysis, and what is it we are looking for? To this dialogue, the curator brings a vast experience in looking at objects, knowledge about cultural canons and context, and information about possible postcollection histories. The conservator brings a knowledge of the physical properties of materials, a familiarity with the chemistry, and training in the use and application of appropriate analytical methods. A collaborative and integrated interpretation of information results not only in a better understanding of collection objects but also in the emergence of new avenues of inquiry and the development of the most applicable conservation treatments.

[This article was accepted for publication in April 2003.]

The NMAfA conservation department remains indebted to the staff of the Smithsonian Center for Materials Research and Education (SCMRE)--particularly Waiter Hopwood Melanie Feather, Charles Tumosa, Harry Alden, Mary Ballard, and Melvin Wachowiak and to Janet Douglas of the Freer Gallery Department of Conservation and Scientific Research. We thank them for their continued technical and analytical support. We are also grateful to our colleague Stephanie Hornbeck, Associate Conservator, NMAfA, for her insightful comments on this paper

References cited

Craddock, P.T. 1985. "Medieval Copper Alloy Production and West African Bronze Analyses-Part 1," Archeometry 27, 1:17-41.

James, Erica. Unpublished "A Technical Study of Ethiopian Icons at the National Museum of African Arts, Smithsonian Institution."

McIntosh, Roderick J. 1989. "Middle Niger Terracottas before the Symplegades Gateway," African Arts 22, 2:74 83, 103-4.

McIntosh, R. I. and S. K McIntosh. 1986. "Dilettantism and Plunder Illicit Traffic in Ancient Malian Art," UNESCO Museum 149:49-57.

Loughran, Kristyne. 1995. Art of the Forge. Washington, DC: National Museum of African Art.

Moffett, Dana, Stephanie Hornbeck, and Stephen Mellor. 2002. "Common Problems in Archeological and Ethnographic Conservation Intersect with the Contemporary: Case Studies of Two African Objects." Preprints of the 13th Triennial Meeting, Rio de Janeiro, 2:679-84. International Committee for Conservation. London: James & James.

Ross, Doran H. 1992. "Interview with Roy Sieber," African Arts 25, 4: 36-51.

Rostron, Bryan 1980 "The Curator Versus the Scientist in Today's Art Museum," The New Hines, Aug.

Sieber, Roy and Theodore Celenko. 1977. "Rayons x et art africain," Arts d'Afrique Noirre" 21:16-28

Stephen Mellor is chief conservator at the National Museum of African Art, Smithsonian Institution. He holds an MS. degree in conservation from the Winterthur Museum, University of Delaware, and a B.A. in anthropology from George Washington University. He recently completed conservation needs assessment surveys for several museums in South Africa and Angola.

Dana Moffett is senior conservator at the National Museum of African Art, Smith-sonian Institution. She holds a B.Sc. degree in conservation from the Institute of Archaeology, University of London, and an M.A. in anthropology from the University of Denver. She recently completed a large conservation project on concrete screens from Nigeria, by the artist Adebisi Akanji, and has a particular expertise in the conservation of African beadwork.