Mark Rothko’s Harvard Murals

By Charles Kessler

This exhibition (on view until July 26, 2015) is not only about a group of Mark Rothko paintings done at the peak of his career, but it’s also about an ingenious restoration technique – a way to restore the color of these faded murals via non-invasive digital projections.

Mark Rothko's tryptic from his Harvard Murals, illuminated with digital projections that restore the original color (Stephanie Mitchell/Harvard Gazette). 

I should note that on the website of the Harvard Museums, there is very little photo documentation of the exhibition or information about the restoration process. This is inexcusable, especially in an educational institution. To make matters worse, photography of the work, even of the wall labels, was not allowed. Fortunately I happen to have taken some photos before I was told it was verboten.

Verboten photograph of some studies for Mark Rothko's Harvard Murals. 

The exhibition features a much-damaged five-panel mural that Rothko was commissioned to make in 1961-62 for the penthouse dining room of Harvard University’s Holyoke Center (now the Richard A. and Susan F. Smith Campus Center).

Rothko's Panel Five inside the Holyoke Center, January 1968 (from the Harvard University Archives).

In addition, 38 studies for the murals are in the exhibition, including this little beauty:

Mark Rothko, Untitled (study for Harvard Murals), 1961, tempera on purple construction paper, 7 x 12 inches (Harvard Art Museums, Gift of the Mark Rothko Foundation).

And, presented here for the first time, a sixth mural which was painted for the commission and held in reserve until Rothko decided which five paintings he wanted for the final installation. (I was not able to find a photograph of this painting.) Since it was ultimately not included in the final installation, it was rolled up and stored until now, so fortunately it never got damaged.

Rothko had made paintings for a particular site before – the Seagrams murals (which he decided not to install), the Rothko Chapel, and the Phillips Collection Rothko Room (see comment below). But the Harvard Murals were the only ones where he personally supervised the installation.  He also chose the color of the walls –  a rather obnoxious greenish ochre that I assume he wanted in order to contrast with the canvases.

The murals are installed in a space that approximates the penthouse dining room’s original dimensions; but without the penthouse's furnishings, windows and drapes, the room feels too big for the work – it's not the immersive environment that I imagine the Holyoke Center penthouse conference room must have been. And intended or not, they don't have the theatrical religiosity of the Rothko Chapel or the mysterious poetry of his earlier work – what Clement Greenberg disdainfully referred to as “moody paintings.”

The murals are more structured than Rothko's typical earlier work with its soft-edged floating rectangles, but not structured enough to deal with their much greater size. Rothko referred to them as “portal paintings,” and they are usually composed of two "columns" framing an interior void. To succeed, the columns and voids would need to be experienced as somewhere between a solid and a colored cloud, or field of color. If the columns are too wide, they become color fields and lose their solidity, too narrow they become a figure in a ground and don't breathe and glow. The same with the voids. If they are too big relative to the size and shape of the canvas, or if there is too much space between the columns and the edge of the paintings, they become just a colored space for the columns to inhabit. The voids lose the solidity needed to compress the space and create a taut, dynamic composition.

Unfortunately, the mural panels don't achieve this balance. Panel Four for example:

Mark Rothko’s Panel Four (photo from WBUR Radio website).

But a successful balance is achieved in many of the smaller studies, like this one:

Mark Rothko, Untitled (Harvard Mural Sketch), 1962, oil on canvas, about 6 feet high (Collection of Kate Rothko Prizel).

The main problem with the Harvard Murals, though, is that they faded badly – so much so that by 1979 the university took them down and put them in storage. Rothko used light-fast pigments, but the binder he used made the pigments highly unstable and prone to fading. This was made even worse because even though Rothko insisted that special light-blocking shades be drawn during the day, people kept the shades up most of the time – apparently the view of the Charles River was just too tempting.

The penthouse of the Holyoke Center in January 1963 (Elizabeth H. Jones/President and Fellows of Harvard College). Rothko is in the middle wearing a sport coat. 

In addition, people ate in the room and had parties there; as a result, food and drink were sometimes spilled on the paintings, and furniture was bumped into them. There was even graffiti on some of the paintings, which shows a shocking lack of respect for what was then contemporary art.

Ordinarily conservators would paint a protective layer of varnish over a painting that needed restoring in order to isolate and protect the original painting from the touched up restoration. Harvard conservators could repair the damage from food and drink, and even the graffiti, but they could not restore the faded color without the isolation varnish masking the subtle brushwork and matte/gloss contrasts in the works.

So, expanding on the ideas of Canadian conservator Raymond Lafontaine, who used slide projectors in the 1980s to partly recover the colors of some art, the Harvard conservators, together with scientists from the MIT Media Lab, came up with what they called “compensation images” – colored light digitally projected on the faded murals, that would restore the original rich colors in a non-invasive way.

The digital imaging software and tools used to calibrate the color corrections onto the damaged Rothko paintings
 (photo: Jesse Costa/WBUR Radio).

To achieve this was no small task. First they had to determine the murals’ original colors. Fortunately there were some good Ectachrome transparencies (large slides) taken of the work in 1964; even though they too faded, the conservators were able to correct their color according to a common formula that has become standard in the field. Then, to verify they had the colors right, they compared the color of the corrected transparencies with the unfaded sixth canvas that had been rolled up and stored the whole time. Next they photographed the existing faded canvases, creating an amazing 2 million pixel scan for each panel. And finally, they made a map of the "compensation color image" by essentially subtracting the faded color from the original color to discover what color is missing from each pixel. These colors are projected onto the paintings, making the color look like it did fifty years ago.

Conservation scientist Jens Stenger holds a white board demonstrating the noninvasive digital projection of one of the damaged Mark Rothko paintings (photo: David Ryan, The Boston Globe).

Of course the experience isn’t the same as viewing paint on canvas. It is after all projected light. Nevertheless, as can be seen at 4:00 every day when they shut the digital projection off so people can see what the faded work looks like, the digital projection definitely helps to bring these faded paintings back to life.