When conservators Michael Duffy and Emily Mulvihill began the process of restoring Andy Warhol’s Gold Marilyn Monroe (1962), they discovered that all that glitters isn’t gold. The title is a bit of a misnomer. In fact, the commercial paint the artist used to cover the canvas of his modern-day icon was a blend that included a less precious copper-based metal. And, further complicating things, early restoration projects introduced pigments into the canvas that contained other metal elements, like copper.
This seems fitting for an artist who once famously said that if you wanted to know about him, you should just look at the surface of his work. Warhol’s choice to create the illusion of gold rather than use the real thing feels deliberate—a way of underscoring his celebrity-as-sacred take on Byzantine icons whose careful, expensive gold leaf highlighted devotional portraits of saints and madonnas.
We sat down with Duffy and Mulvihill to discuss their new approach to cleaning and conserving Gold Marilyn Monroe, the surprising prevalence of gold in other artists’ works at MoMA, and the material that may, in the end, be the best solution for conserving Warhol’s painting. (Yes, that would be real gold.)
This interview has been edited for length and clarity.
Emily Mulvihill prepares to work on Gold Marilyn Monroe.
How did you begin to approach this project?
Michael Duffy: Gold Marilyn was on our conservation to-do list for a while, because some of the older restorations from the ’60s and ’90s have started to darken. The main focus of the treatment was to reduce or remove those old restorations and then tackle the main area below the figure of Marilyn where there is discoloration. We had to understand the painting in order to know what could be done to treat it, so we wanted to do a more in-depth investigation into the actual materials of the background and the “gold” used. The main portrait in the middle is in beautiful condition. It’s really the gold canvas background that we’re trying to decipher and find out the best way to treat.
How did you begin your research into the gold background?
Emily Mulvihill: Chris McGlinchey, our conservation scientist, did analysis of the paint. We found that the pigment is a brass alloy and the binder is a vinyl polystyrene. Together, those combine to create a unique paint that’s very susceptible to changes in the environment. Figuring out a way to work with that material has been interesting, because it’s so sensitive to all of our solvents. Because the pigment is brass, it corrodes easily and is susceptible to water. It’s tricky because nothing we have in our toolbox is ideal. So one of the goals was to try to replicate the paint. We were in contact with Ulysses Jackson from Golden Paints. He’s a scientist too, so he and Chris tried to work out a replication.
MD: He actually made us a paint that was similar in characteristics to what we think was used for Marilyn, but instead of a commercial paint that would have been used in the ’60s, this is anacrylic paint that’s available now. Just by looking at Marilyn, you can see the paint was applied in different ways. Some strokes suggest a brush was used, while areas next to her head suggest a spray can was used. Those were commonly available even in a hardware store. Really, Warhol’s goal was just to get a gold background down. There are all kinds of imperfections in the way it’s applied. It’s not a classic gold leaf technique, where [the application] would be perfect because you’re working with expensive materials.
This work has a long history of restoration, including shortly after it was painted. What happened with that original restoration that needed to be either replaced or reversed?
MD: Something happened in the ’60s to create this mark underneath the portrait. It seems like it was addressed at the time but we don’t have any documentation. In 1969, conservators added treasure gold, which is typically used on frames and is a wax-based gold material containing copper alloy and a combination of copper pigments that give you a gold look. Unfortunately, to remove the treasure gold also would disturb the original paint, so we’re limited in how much we can do. In the ’90s, I actually painted over the old restoration using watercolor, but the watercolor darkened with time. It was a 20-year disguise that now needs to be redone. Since 1998, new materials have been formulated to work more safely using solvents. Solvents we typically use are petroleum based or alcohol based to dissolve varnish, old restorations etc. Similar to turpentine or “mineral spirits” or rubbing alcohol. Sometimes they are used in gel formulations or in emulsions with water-based solvents to be more effective and safer for the art and the conservator. There’s a new system being developed in Italy using micro-emulsions, so not one single solvent but a combination. Those are either applied just as a solvent, or they can be used in a gel material that allows them to be safely applied to the surface.
EM: These gel systems limit the interaction of the solvent with the surface, so there’s less chance of interacting with the original paint layer.
Andy Warhol. Gold Marilyn Monroe. 1962
Had you ever used these solvents before?
MD: I haven’t personally.
EM: I haven’t, but I’ve experimented with it before. It’s usually just used as an aqueous solution to remove dirt. We are trying to test it and get as comfortable as we could with it before we apply it to the painting.
What kinds of tests are you running?
EM: This goes back to why we needed to make a replica of the original paint, because we wanted to make mock-ups to thoroughly get comfortable with the new technique. So we’ve been doing some tests on our mock-up. We haven’t quite found the right combination of solvents yet.
MD: We were also collaborating with a conservator at the American Museum of Natural History who’s using these gels actively on conservation of their totem poles. You might think, how would a totem pole relate to this? But in fact, it’s really trying to safely remove a synthetic coating from another synthetic paint. So that’s our dilemma here: to reverse the treasure gold.
How has this project differed from other Warhol projects that you might have done?
MD: The way it’s trimmed on the back seems a little different than other Warhols. If you look at the top edge, you can see this is the manufacturing edge of the canvas, which typically has a red line. The bottom edge is the same. It’s very faint, but we know he used the whole roll. There’s a really great little section on the right edge where someone stapled it, and then decided to unstaple it and then restaple it. It also gives you the idea that the canvas maybe had a different life. I think, like a lot of Warhol paintings this size, you do see these little surface imperfections that could have been already there before it was even painted on.
Emily Mulvihill repairs a dark mark on Gold Marilyn Monroe.
What does this reveal about the Factory, Andy Warhol’s New York City studio?
MD: I can picture this roll of canvas lying around, and Warhol directing someone to go get it and paint it gold, and then putting it on the floor to add the silkscreen image. On the back of the canvas you can even see some ink, suggesting it either was faced down at some point or lying on top of another canvas that was just inked. I think it tells you it wasn’t pristine art making.
Have you ever treated works in the collection that include real gold?
MD: A few years ago, we treated a work by Agnes Martin called Friendship. She used actual, very thin gold leaf applied directly to a gesso surface, which is a pretty traditional method. It’s not traditional to put it on canvas but, in any case, she did. You don’t see an iridescence like you do on Marilyn, which was a clue that it’s not real gold from the start. Gold leaf is super stable. It’s not sensitive to light because it’s an inert metal. It doesn’t react to humidity and corrode like copper does, for example. Friendship was made around ’63, similar to Marilyn, so I think it makes an interesting comparison. It’s pretty impressive that Martin was able to cover the whole canvas with gold leaf because it’s tricky to use. Trickier than just taking spray paint or brushing gold on.
Agnes Martin. Friendship. 1963
Have other gold paints been as challenging as this one?
MD: There’s an untitled [Robert] Rauschenberg, which isn’t really technically gold paint but gold leaf painting. That’s a conservation challenge because the gold leaf is not firmly attached to the canvas underneath. It’s floating around inside this glass box, so when it’s moved there’s risk of detaching the gold leaf. Although, you could argue with that because apparently he wanted that effect of the gold leaf moving to give the idea that it was not a static object.
Did you discuss this work with other conservators who have worked on Warhol?
MD: We talked to a couple different, highly respected conservators in New York who’ve worked on many Warhols. They came here and looked at the painting with us and gave their advice.
EM: It’s really nice to be able to reach out and not work in a vacuum, to collaborate, get everyone’s opinion, and then build a consensus and move forward. That’s been a great part of the project.
Robert Rauschenberg. Untitled. c. 1953
What’s next for Gold Marilyn?
EM: I’ve already removed the restoration from the ’90s, so it’s already improved. We still want to see what the treasure gold from the ’60s is doing—if we think we can leave it, or if we should go forward with more of these cleaning systems. The treasure gold appears to be restricted to one area at the top of that drip black mark, so we might not have to do as much in painting as we initially thought. The goal is to minimize it, not make it invisible. Ultimately, we’ll look at the work in the galleries and probably have to do some minor inpainting there because that’s where it’s being displayed. It looks very different up here in natural light than it will in the galleries.
MD: We’ve cleaned the surface to the extent we can safely, but it’s aged. We want to respect that and some of the imperfections that show how it was created. There are things that have happened though, like the dents and impression from the stretcher bar, that we’d like to improve if possible. For the inpainting, we will probably use a water-based, actual gold, and maybe tint it to match the iridescent gold.