Constance S. Silver
The artistic patrimony of the Silk Road is all the more
Remarkable when one recalls that its monumental mural paintings and sculpture are derived primarily from Neolithic traditions that stem from simple clay-based technologies. For example, characteristic mud-brick construction, cut-stone masonry, and grotto surfaces were covered with mud renderings, upon which mural paintings and other decorative finishes were executed. Most of the pigments also contain clays. The paint media were obtained from plant gums, faunal glues, and other natural materials (Mora, Mora, and Philippot 1984:82-83).
Although separated by distance and culture, there are many mural paintings in the United States that share technical similarities with those of the Silk Road. The prehistoric and historic mural paintings of Native Americans of the Pueblo culture of the American Southwest are technically the closest. Classically Neolithic, they were painted with naturally occurring pigments in tempera media applied to mud and clay renderings. There are also tempera mural paintings in the Western tradition that date from the eighteenth, nineteenth, and twentieth centuries, although these are rarely executed on clay-based renderings. The following discussion focuses on minimum intervention—a key principle of modern conservation—for the treatment of monumental mural painting in tempera media, and reviews four recent case histories from the United States.
Minimum Intervention in the Treatment of Tempera-Based Murals
The principal objectives of minimum intervention in the treatment of tempera-based mural paintings are the retention of original materials and their original optical qualities and the retention of the mechanical equilibrium of rendering, preparation, and paint. Consequently, modern synthetic materials of conservation, which generally differ radically from traditional materials, are employed only when there are no other practical options.
However, there are additional technical and philosophical reasons for minimum intervention. Mural paintings are inherently problematic for
Conservators because they are a component of generally complex building systems that include the architectural/structural support, the materials of preparation and paint layers, the ever-changing interior and exterior environments in which the structure and murals exist, and the impacts from human use and misuse. All elements of the system must remain compatible if the mural painting is to survive in an undamaged state. Unfortunately, it is quite easy for any building system to become destabilized, and it is impossible to predict when instability may occur. It is also acknowledged that many conservation treatments have proved unsatisfactory or incompatible over time, in spite of the best modern climate-control systems. In this regard, the porous nature of the component materials of the murals of the Silk Road, and of similar tempera-based murals, makes complete removal of any poorly performing conservation materials virtually impossible.
Thus, minimum intervention is physically prudent. It is also philosophically attractive because the murals of the Silk Road—like those of the prehistoric Southwest—are the material record of unique moments in human history. When these murals are infused with modern synthetic materials, they are changed forever—even if the treatment accomplishes its goals. The following case histories from three regions of the United States—the Southwest, the Central Atlantic, and the Northeast—illustrate the viability of minimum intervention for the conservation of tempera mural paintings (Fig. 1).
Figure 1
Map showing the location of the sites discussed.
The prehistoric agrarian culture of the American Southwest culminated around 1250 c. e., in complex masonry villages constructed in huge caves, the world-famous cliff dwellings. By about 1350 c. e., the area had been abandoned by its prehistoric inhabitants, for reasons still unknown. Extensive paintings survive on the prehistoric architecture at such sites as Mesa Verde National Park, Colorado; and Canyon de Chelly, Arizona. Briefly summarized, the most complex architectural finishes are composed of a first rough coat of mud mortar applied directly to the masonry, and sometimes to the stone face of the caves. A second finish coat, of finer texture, was applied to the rough coat. This stratum is the surface for the painting. Often, a ground of white clay was applied in preparation for painting. The palette was simple, consisting of a small number of mineral pigments. The most outstanding mural paintings were executed in subterranean chambers called kivas, used for ceremonial purposes (Smith 1952).
Prehistoric Mural Paintings and Architectural Finishes in the American Southwest
In 1981, a field study was undertaken in Mug House Ruin, Mesa Verde National Park, a site administered by the National Park Service. The objective was to develop methods for the conservation of the architectural finishes and to test the conservation treatments in the field over several years. The research design was based on an observable fact: after eight hundred years in uncontrolled environments, in the absence of human contact, the murals had remained in remarkably good condition. Clearly, this “system” of architecture, architectural finish, and environment had proved itself compatible over many centuries. Where unaffected by human intervention, deterioration of the architectural finishes resulted largely from the fatigue and desiccation of the component materials over time. Therefore, modern materials of conservation were judged an option of last resort because they might introduce a destabilizing element into the generally stable system, in addition to changing the original materials' composition forever.
However, the murals of the subterranean kivas were actively deteriorating from human intervention: they had been excavated in the early 1960s and left exposed, the surface of the murals acting as the site of evaporation for groundwater moving through the masonry. Several conservation problems had developed and become very serious: detachment of rendering from the wall; delamination between strata of the rendering; friable rendering; flaking paint; efflorescence; root penetration; burrowing insects and rodents; and surface dirt.
It was posited that water would be an effective conservation material because of its capacity to relax and replasticize the brittle and deformed clay-based rendering and paint. This would permit compaction and cohesion of the delaminated strata of the rendering and their reattachment as a cohesive unit on the masonry. Water also enhances the bond between paint and rendering, perhaps also by reactivating any remains of desiccated original media that might still be present.
Pilot conservation treatments were carried out on one kiva and on one standing wall of Mug House in 1981 (Fig. 2) (Silver 1991). Briefly summarized, the treatment to stabilize rendering and paint entailed temporary attachment of supportive wet-strength tissue facings to the surface,
Figure 2
Mug House Ruin, Mesa Verde National Park, Colorado.
Using water applied by brush. This was followed by spraying of water, until the rendering became malleable. A mixture of 50% water and 50% isopropyl alcohol was injected between the detached rendering and the masonry, followed by injections of diluted polyvinyl acetate emulsion.1 The treated area, now a cohesive unit, was pressed back onto the plane on the wall and allowed to dry under pressure for twenty-four hours. The adhesive was the only synthetic material used in this treatment because no other practicable category of material was available. However, it was used as a thin stratum in discrete areas of detachment between the surface of the masonry and the rendering. No synthetic materials were infused into the fabric of the mural nor applied to its surface.
After nine years, the treated areas had remained stable. Based on these results, two comprehensive projects to conserve very unstable mural paintings were carried out at Aztec Ruins National Monument, New Mexico (Silver, Snodgrass, and Wolbers 1993). Aztec Ruins National Monument is a prehistoric site administered by the National Park Service. The first project was undertaken in 1990 to stabilize the mural paintings in Room 156, a roofed masonry structure (2.8 x 3.1 x 3.4 m high). Although roofed, there had been leaks—causing serious detachment of the mural paintings from the masonry walls—erosion, and considerable deposition of mud on the painted surfaces.
The murals were stabilized using the methods developed in 1981 during the pilot treatments at Mesa Verde National Park. However,
Rhoplex AC-33, an acrylic emulsion adhesive, was substituted for polyvinyl acetate emulsion because there is some evidence that it will age better in the field. The mud was removed with various mechanical methods. The murals have remained stable.
The mural painting in Room 117 of Aztec Ruins National Monument was treated in 1992 (Fig. 3). Measuring about 1 x 1.3 m, it is located on a fragmentary wall that had been partially exposed to the elements since excavation around 1920. The mural, composed of horizontal fields of white and red paint, is unusual and important because many anthropomorphic, zoomorphic, and geometric forms were etched into the upper white field.
The mural had deteriorated and become very unstable due to its semiexposed position and to artificial and adverse changes in drainage patterns. The instability was exacerbated by the effects of heavy coatings of shellac and cellulose nitrate, which had been applied as consolidants in the past. These coatings were unsightly because they had consolidated dirt on the surface and had blanched (i. e., turned milky and opaque). More disturbingly, however, the coatings were contracting, causing the painted stratum to curl and detach from the mud rendering.
A treatment to remove the coatings was developed by application of solvent mixtures through Gore-Tex2 and wet-strength tissue. The Gore-Tex appears to exert a poultice effect, drawing the dissolved coatings from the mural and depositing them on the tissue. However, laboratory examination showed that some adhesive remains fixed below the painted surface, again indicating that it is impossible to remove all consolidants from this type of porous material.3 The murals were then stabilized with the treatment used in Room 156, mentioned above. The final step of the project entailed controlled backfilling as the only viable and cost-effective way to protect the mural from the elements.
Tempera Mural Paintings by Candido Portinari in the Library of Congress
Figure 3
Mural painting in Room 117, circa 1300 c. e., 1 X 1.3 m, Aztec Ruin National Monument, New Mexico.
The Library of Congress in Washington, D. C., maintained by the architect of the U. S. Capitol, has recently been renovated and restored. On 12 February 1990, a frozen pipe ruptured above the Hispanic Reading Room. Water flooded the area above the room and soaked into the mural Mining far Gold, painted in 1941 by the Brazilian artist Candido Portinari (1903-62) in a tempera technique on a lime-based plaster (Fig. 4). According to his son, Portinari painted with Totain glue, a form of rabbit-skin glue still used today. Laboratory analyses confirmed the presence of rabbit-skin glue, but also suggested that egg albumen might have been used, as well, in a very limited way.4
Following the flood, the ambient conditions of the room and the water content of the mural painting were monitored. The relative humidity in the room stabilized about three weeks after the flood. The mural was allowed to dry slowly for several months. The flooded cavity wall was opened, and fans maintained circulation of air in the room and in the wall. Although the mural had been soaked, it remained remarkably stable as it dried. However, the paint did powder and flake in many areas. Salts composed of sodium sulfate, sodium chloride, and sodium carbonate effloresced in some areas.
A conservation treatment was carried out in April 1991. Salts were removed on a cotton swab moistened with saliva. Powdering paint was consolidated with rabbit-skin glue diluted 1:18 in water, and applied as a fine mist. Areas of lost paint were inpainted with pure powdered pigments in water. The mural painting has remained stable since treatment.
The Murals of Reginald Marsh in the U. S. Custom House, New York City
Figure 4
Candido Portinari, Mining far Gold, 1941, 5 X 6 m, Jefferson Building, Library of Congress, Washington, D. C.
One of the most outstanding cycles of mural paintings in the United States was executed in 1937 in the rotunda of the U. S. Custom House, New York City, by modern American master Reginald Marsh (1898-1954) (Fig. 5).
The sixteen murals show alternate scenes of the port of New York as it appeared in 1937, with trompe l'oeil portraits of important historic explorers. Laboratory analyses were positive for protein, supporting archival documents that suggested that Marsh employed a tempera medium composed of pigments mixed in limewater and skim milk applied to a dampened lime plaster; however, Marsh's idiosyncratic technique has not been identified with complete certainty.
From 1937 until 1970, the rotunda was a busy office space. With the exception of heat in the winter, the environment remained uncontrolled. From 1970 until 1991, the building was abandoned and unheated, and the roof often leaked. Thus, when each mural was individually examined in 1990, preparatory to the renovation and reoccupation of the building, areas of powdering and flaking paint were anticipated and recorded. However, the overall stability of Marsh's rather odd and inherently fragile system of monumental painting within such poor ambient conditions militated against introduction of consolidants and fixatives that could prove destabilizing over time.5 It was decided that Ethulose (ethylhydroxyethyl-cellulose) would be employed as the fixative. Although Ethulose is a modern synthetic product, it shares many properties and is compatible with traditional materials. An aqueous solution ranging from 1% to 1.5% in deionized water and ethanol 1:1 was applied by spraying. Moreover, like the highly diluted rabbit-skin glue used to treat the Portinari mural, Ethulose did not alter the optical quality of the colors.
Conclusions
Figure 5
The murals of Reginald Marsh, painted in 1937 in the rotunda of the U. S. Custom House, New York City.
Tempera mural paintings that share many technical similarities with those of the Silk Road exist throughout the world. Two principal conclusions are drawn from the case histories discussed in this paper. First, although inherently fragile and sensitive to the environment, mural paintings executed in tempera media can remain stable over centuries, especially when protected from human intervention. Second, many problems of conservation can be treated effectively with minimal intervention, which avoids the introduction of potentially incompatible modern materials that change the character of the mural painting forever and may well prove deleterious over time.