This ceiling was part of the so-called Martin commission from 1937. René Martin supervised an artisan workshop in Rabat, Morocco, and Doris Duke and her then-husband James Cromwell contracted him to design multiple rooms and bespoke pieces for their Honolulu estate. Although Duke played with the designs and made significant changes, the foyer ceiling, the living room ceiling, and the living room doors into the Mihrab room are still central elements within the interior.

I am here at Shangri La for a total of six weeks as a graduate intern in art conservation. Within our profession, we specialize in different types of materials and collections. My particular interest is the historic interior, architectural paints, and decorative surfaces. Therefore, my internship was designed around these three painted wooden architectural elements from the Martin commission.

It truly is “another world” up close to the ceilings because the surfaces and condition of the objects are magnified. From ground level, these objects appear well-preserved, albeit dark. The sheer size and intricacy are mesmerizing and distracting. However, inches away from the surface, it is a conservator’s nightmare. Paint is flaking, and these are not one or two small loose chips; the design looks like it is exploding from within. Large flakes of paint have peeled away from the wooden surface and curled into themselves.

You may be wondering: Is this something that happens often with painted surfaces? The answer is occasionally, especially more so when the paint is exterior and exposed to weathering, but in general the ceiling’s poor condition is fairly extreme. So the next questions that come to mind are: Why is this happening? Can we fix it? How can we prevent it from happening again?

These questions drove my internship. My investigation actually began prior to my arrival at Shangri La, back on the mainland at my home institutions the University of Delaware and the Winterthur Museum. The full-time conservator at Shangri La, Kent Severson, mailed samples from the foyer ceiling to Delaware. Using these samples, I developed a treatment protocol with advice from my Master’s supervisor, Professor Richard Wolbers.

You might look at details of the ceiling and think, How do you even begin to salvage it? The answer is one paint flake at a time. Patience is a key trait of being an art conservator. Each day my work consisted of gently applying a dilute adhesive (in conservator jargon, a consolidant) with a small brush behind and sometimes on top of each paint flake. The adhesive Richard and I chose to use is a very stable resin often used for exterior coating application. It also allows water vapor (but not liquid water) to pass through the coating. This is a key property because Shangri La is exposed to many exterior elements, including salty sea spray. The solvent that the adhesive is dissolved in also helps gently relax the paint so that it can be uncurled and manipulated back into place. Occasionally, minimal heat was applied through a thin silicone barrier to further encourage relaxation and adhesion. In select areas, especially the flower motifs, the paint was extremely brittle. Therefore, a water-based thermoplastic adhesive was used instead. The original paint, being water-sensitive, was further softened with this adhesive system.

Even though there is not much visual improvement, the treatment has been very successful. In these four weeks, I was able to treat a little more than a third of the foyer ceiling. I also stabilized the paint on the living room doors, which were in much better condition than the foyer ceiling.

So the remaining question is, Why did the paint deteriorate in this manner? It is important to elaborate that the flaking paint is mostly concentrated in areas of gilded decoration. It is also important to discuss a second issue: globs of resin on the surface. This phenomenon is known as “extractive bleed” and refers to resin present in the wood that begins to exude and migrate to the surface. The migration can be motivated by heat and humidity, of which Hawai‘i has plenty. There are certain types of wood that are more prone to “extractive bleed,” and cedar is one of them. The wooden objects of the Martin commission were thought to all be made of cedar, but being diligent conservators, Kent and I both wanted to confirm this with some hard evidence. I took small samples of various components of both ceilings and one sample from the doors. Dr. Harry Alden, a wood specialist in Maryland, analyzed the samples under the microscope and all were found to be cedar.

It is ironic that a book on Moroccan homes and gardens from 1926, which Martin used for design inspiration, clearly warns against using cedar as the base for painted decoration because of the excess amount of resin and peeling paint (Jean Galloti, Le jardin et la maison arabes au maroc [Paris: Editions Albert Levy Librairie Centrale de Beaux-Arts, 1926], 1:78. Translation provided by Ashley Miller.). Perhaps Martin had had success with painted cedar in the past (after all, the Moroccan climate is much more arid than the Hawaiian). We will probably never know why he used cedar, but his choice of wood is the beginning of understanding the current condition.

Every day up on the scaffolding, I became more acquainted with nuances of the decoration and its condition. More and more, the deterioration seemed characteristic of excessive water damage—I mean more than just the daily 75% relative humidity. I started digging in the Shangri La records with the help of archivist Dawn Sueoka. I could not have asked for a better partner; Dawn is so enthusiastic and always eager to go the extra mile to solve a mystery. I was able to find that there were reoccurring issues with the foyer and living room roofs, including a major leak in the foyer corresponding to one of the most deteriorated areas right in front of the main door. This was an exciting find, but it still didn’t answer why the gilded areas were the most problematic.

Conservation is an interdisciplinary profession and we conservators really enjoy working with allied professionals. I was very fortunate to have PhD candidate in art history, University of Michigan, Ashley Miller, onsite, completing a residency at Shangri La. Ashley found a French reference from 1917 (Jean Gallotti, “Les Métiers d’Art au Maroc,” France-Marco: Revue mensuelle illustrée, No. 5 [May 15, 1917]: 8–19) that discusses traditional gilding materials used in Morocco. It appears that common oil sizes (the stuff that makes the metal leaf stick to the surface) were from “sibar resin” (which refers to either aloe resin or myrrh oil). We don’t know for sure if either of these was used on the Martin commission, but myrrh oil is very similar chemically to cedar oil (the resin coming out of the wood). It is likely that as the resin exuded out of the wood, it could have softened and mixed with the oil size, completely destabilizing the gilding system.

Analysis is still ongoing and I sent additional samples of the metal leaf and resin back to the Winterthur Museum Scientific Research and Analytical Laboratory. There is always more than one factor contributing to the condition of an object. There are some obvious areas of retouching on both the foyer ceiling and the doors. Doris Duke was known to repair her own ceramics and touch up decoration with acrylics. Sometimes, introducing new materials can disrupt a paint system or accelerate deterioration. As conservators, we are continuously researching the stability and longevity of the materials we use. We also try to perform treatments that could be easily reversed or retreated in the future, and we never paint over or remove original material.

Unfortunately, the decorative surfaces of the Martin commission are no longer representative of their original design and appearance. Most of the gilding is completely gone or so tarnished that it appears black. The varnish has severely yellowed and darkened, but it may be the original varnish (Gallotti also discusses the preferred appearance of a varnished surface to counter the inherent matte-ness of distemper paints) and should remain with the object. The exuding resin could be removed, but more will likely migrate to the surface given the tropical climate. Furthermore, some of the exuding resin has fused with the varnish, and so removal would create a discordant appearance.

In order to help unify the design, I have been testing some selective retouching overtop of the exuding resin. This way, the resin is toned down and blends in more with the surrounding colors. The paints I am using for retouching are Golden Qor paint, which are based on an Aquazol resin. Aquazol is a stable conservation material that has been used extensively at Shangri La to stabilize other painted decoration. It holds up well against heat and humidity and does not pick up dirt like Acrylic-based paints because there is no surfactant used.

I am currently working on a digital reconstruction of the ceilings and doors using Photoshop, so people can have a better idea of what they looked like when they were originally installed in 1938. The colors are not mathematically accurate (that would require colorimetry), but they are a close approximation based on revealed areas of original paint under the varnish layer. Look for this image to be posted with my final treatment report.

Phew!! I’ve certainly been busy at work! But after all, I’m also in Hawai‘i, and I have been able to experience some additional culture and entertainment. I learned how to stand-up paddle on the North Shore and am now totally hooked! I also climbed the stairs at Koko Crater. I was very impressed by the collection at the Honolulu Museum of Art; if you want to see what Waikīkī looked like before all the hotels there are some great 19th-century canvases on view. For a somber history lesson and beautiful koa wood furniture, I definitely recommend ‘Iolani Palace.