King David’s Head from Gaza Synagogue Restored

058

Archaeologists uncover the past; museums make it accessible to the public. The restorer’s job lies in between, turning fragmentary remains into pieces that modern viewers can appreciate and understand. In the case of a sixth-century A.D. synagogue mosaic from Gaza, restoration work was aided by advanced computer technology to reproduce an image of King David that was thought to be forever lost.

The mosaic first became known to archaeologists in 1966, when the Italian journal Orientalia reported the discovery of an ancient church on the Gaza coast by the Egyptian Department of Antiquities. Within the church was a mosaic floor, described as depicting a figure of a female saint playing the harp and charming wild animals.¹ Two pictures of the mosaic were included in the article.

When Hebrew University Professor Michael Avi-Yonah read this report, he immediately saw in the photos of the mosaic a Hebrew inscription that clearly read “David” (dywd). He correctly identified the figure as King David playing the harp and the building as a synagogue rather than a church.²

After the 1967 Six-Day War, when Israel took control of Gaza, Israeli archaeologists rushed to the site to inspect the synagogue and mosaic they had read about but had never seen.

The site had lain unprotected since its discovery in 1965. When the Israelis arrived, they found that part of the mosaic had been destroyed: David’s head, one of his hands and several other small areas were gone.

The Israel Department of Antiquities promptly sent a team headed by Professor Asher Ovadiah to excavate the site.³ The ancient synagogue turned out to be the largest ever discovered in the Holy Land.⁴ The exact plan is unknown; Ovadiah suggests that it was almost square (85 by nearly 100 feet) and probably consisted of a wide nave and two aisles on either side created by four rows of columns, two on either side of the nave.

One entered the building through one of three entrances on the westernmost wall (at the bottom of the plan above). On the opposite wall was a niche, presumably for the Torah ark, and, in front of it, a podium (bema) from which the Torah scroll was read. This area was apparently divided from the rest of the prayer hall by chancel screens; lovely fragments of 060four intricately-carved marble chancel screens were found in the excavation.

The floor of the synagogue had been paved with beautiful mosaics. In the southernmost aisle, at right on the plan, was a well-preserved mosaic with the popular pattern of vines forming circles, or medallions, that contained animal figures. At the foot of this mosaic, flanked by two peacocks, is a Greek inscription by which the mosaic (and the synagogue) can be dated. It reads:

“(We) Menahem and Yeshua, sons of the late Isai (Jesse), wood traders, as a sign of respect for a most holy place, donated this mosaic in the month of Louos (the year of) 569.”

The date, given in the so-called Gaza Calendar, corresponds to 508/9 A.D., according to Avi-Yonah.

Mosaics with a similar composition, with vines creating medallions, were used in both churches and synagogues at this time. We know of a nearby church at Shellal and a synagogue at Nirim, from later in the sixth century, that contained mosaics very similar to the Gaza mosaic.⁵ The synagogue mosaic at Nirim depicts a menorah; the church mosaic at Shellal does not. The Gaza mosaic has no menorah, but the donors’ names, in the mosaic inscription, are typically Jewish. Further, at Gaza, the name David appears beside David’s head in Hebrew characters, rather than in Greek. It seems clear that this structure was a synagogue, rather than a church.

The figure of David wears a robe and strums a lyre. One hand plucks the strings from behind, while the other holds a plectrum—in this case, a short rod with a ball on each end—to strike the lyre. With a halo, dark eyes and a crown, David appears as a typical Byzantine emperor seated on his throne. Incidentally, royalty is often portrayed with a halo in the Byzantine period. In 061the inscription identifying the figure, the name is spelled somewhat differently than we usually see it. It is the so-called plene or full form—DWYD instead of DWD. The additional Y (yod) is used as a vowel, although Hebrew is written essentially only with consonants. David is spelled with the additional yod in the book of Chronicles, as well as in Ruth, Ezekiel, Hosea, Amos, Zechariah, Song of Songs, Ezra and Nehemiah, so there is Biblical precedent for this spelling.

Around the king are a lion cub, a snake and a deer (once thought to be a giraffe). The content and composition of the scene resemble the well-known motif of Orpheus charming wild animals with his music, and this was noted, in fact, immediately following its discovery.⁶ Orpheus was a popular subject in the art of the eastern Mediterranean in late antiquity; depictions of him surrounded by animals as he plays the lyre are found in Cyprus, Jerusalem, Palermo and Oudna, to name just a few sites. In all these depictions, Orpheus sits in the center of the field while each of the various animals around him turns or raises its head toward him. Orpheus’s sweet singing and lyre-playing were said to tame wild beasts and cause even trees and plants to bow down to him. In the most famous myth about Orpheus, he used his music to charm monsters of the underworld when he attempted to retrieve his wife from Hades.

Orpheus images found their way from purely pagan contexts to both Christian and Jewish sites. Many depictions of Orpheus charming wild animals, with this story’s theme of serenity and peacefulness and its reference to resurrection, adorn early Christian catacombs in Rome and early Christian sarcophagi. The central wall of the famous Dura-Europos synagogue contains a scene generally believed to represent David, using the well-known model of Orpheus scenes.

The depiction of David in the Gaza synagogue is a brilliant combination of David as royalty and David as Orpheus, with the characteristics of each. He is Orpheus seated not on a rock, but on a throne; not with a cap, but with a crown. He is at once the ideal king and a harbinger of peace and redemption, taming even wild animals.

More’s the pity that the archaeologists found him with his face destroyed. In its damaged condition, the mosaic was inappropriate for exhibition. To preserve what had survived, the Department of Antiquities’ restorers lifted the mosaic from the site and brought it to Jerusalem. A new backing of reinforced concrete, about four centimeters thick, was cast onto its reverse side to prevent further deterioration. In that condition, the mosaic was placed in storage in the Rockefeller Museum, but not forgotten. Descriptions and an occasional photograph of the mosaic’s unique combination of features found their way into many a scholarly book or article. In 1992, archaeology curators Yael Israeli and David Mevorakh at the Israel Museum, where I work as a conservator, considered displaying the mosaic on an empty wall; first, however, it would need to be mounted and restored. They asked our restoration laboratory at the Israel Museum whether the destroyed parts could be reconstructed, based on pictures that had been taken before the loss. The Objects Restoration Laboratory, directed by Ruth Yekutiel, rose to the challenge.

Preparing a mosaic for exhibition requires exacting work, involving both technical skills and imagination. In this case, after a thorough examination and cleaning of the tesserae (mosaic stones), a specially designed metal construction was fitted to the back of the mosaic for display on the wall. Most of the stones were still in good condition. Only the black limestone tesserae had been tarnished by a whitish encrustation—the result of the oxidation of organic material present in this stone. Such encrustation must be removed by gentle scraping or grinding, either by hand—with fine-grade sandpaper or a surgeon’s scalpel—or with a power tool. The results 062in our case were striking. The cleaning of the black tesserae—used in most Byzantine mosaics to delineate the images—suddenly brought the depictions to life.

The salvaged mosaic scene consisted of three sections: the central image of King David, adjoined by an undulating snake and a lion cub, both straining to hear the lyre music; the head and slender neck of a deer or gazelle, to the right, surmounted by a geometric border; and a scrap of this border extending below into a wide guilloche- and wave-patterned frame. The three sections were aligned and then attached to the metal wall support. This phase presented a new set of technical snags. Because of the irregular thickness of the concrete backing, much “juggling” was necessary to ensure that the surfaces of all three slabs would be on a single plane. Finally, the sections were attached to the support with large nuts and bolts. This required drilling holes through the triple layer of tesserae, concrete and metal. Wherever possible, the holes were located in areas where tesserae were already missing; elsewhere the tesserae had to be carefully removed and then reinstalled. The three slabs then formed one unit.

Set off by the decorative borders, the fragments now merged into a coherent, if incomplete, scene. Although David’s face was still missing, his majestic pose and the captivated stance of the animals created an impact. Moreover, the David inscription caught one’s eye and seemed to act as a title—resembling a label in a museum. At this point, though technically ready for museum display, the David mosaic still awaited the most critical stage of the restoration process: the reconstruction of the royal head.

Unfortunately, only black-and-white pictures of David’s face had survived. Worse, they had been taken at an angle, making the head appear slightly distorted. The photographer had simply stood at the bottom of the mosaic, instead of using a ladder or a tripod and shooting straight down. Since the focal point of the picture was its center, David’s head was not only distorted, it was also somewhat out of focus.

To recreate the actual appearance of the head from the distorted photographs, the museum contacted the Computer Vision Laboratory at the Hebrew University, which researches new ways of processing information culled from images. For the first time, the technique known as “computer warping,” often used to adjust the angles of aerial photographs taken for military purposes, was applied by the laboratory to archaeology.

063

In addition to using one of the two oblique photos, the lab used a photo of the mosaic in situ taken “straight down,” but after the head had been destroyed. These photos were first scanned into a computer. Having initially reconstructed the angle from which the oblique photo had been taken, doctoral student Eyal Piassetzky then selected some 30 “control points” from each photograph, very specific points such as a corner of the lyre, and the intersection of the two lines of the letter daleth in David’s name. Using these reference points, the computer connected the two pictures and then superimposed the oblique image onto the vertical one. To visualize this process, you may imagine the oblique image pivoting in space, and the straight-down image lying flat. The computer affixed the distorted image to the frontal one by rotating it until it could be pinned down, point by point, with the “control points” of the two layers overlapping exactly. The original head from the oblique black-and-white photo now occupied its place on the headless straight-down image, adjusted to its correct shape and dimensions.

Even this meticulous process did not suffice for restoration purposes, however: The new image provided us with the correct outline of David’s head and halo, but almost all the details of the face, hair and crown had been irretrievably lost. At this crucial moment, the oblique photographs proved valuable. Because these photos had been shot from the side, the play of light and shadow on the mosaic made most of the individual stones discernible. We now proceeded to combine information supplied by the new and the old oblique images.

With only black-and-white images, we could never precisely identify the color of individual tesserae. Because of this limitation and the absence of certain details, we decided to restore the missing parts not with tesserae, but by etching the restored image into plaster, a technique never before employed.

First, a perfectly smooth plaster surface was prepared for the areas to be restored. The next task was to copy the now-straightened outline into place. To do this, we first traced the part of the mosaic that bordered the missing area onto a sheet of paper and hung that on a wall. Using an overhead projector, we then screened the computer-generated image onto the wall in such a way that it overlapped the tracing. We then knew that we had the correct dimensions for the restoration, and we traced the projected image of the head onto another sheet of paper. Then, carrying the paper over to the destroyed area of the mosaic, we penciled over the picture again, with carbon paper between it and the smooth plaster surface. As we painstakingly examined the old photographs through a magnifying glass, more of the individual tesserae became distinguishable and could be copied onto the paper and from there to the plaster.

Several details, particularly David’s mouth and the hand behind the harpstrings, stubbornly eluded our decipherment. The lack of information about color was especially frustrating. Without a doubt, King David’s crown would have been embellished with colored stones, to denote jewels. Was the king portrayed as a redhead, as he is described in Samuel (1 Samuel 17:42)? And how did his halo look? We had to leave such questions open, rather than complete the picture from our imagination.

Having finished the drawings, we etched them into the plaster foundation. The etched lines represented the spaces between the tesserae; we painted them a grayish color to resemble the mortar between the tesserae in the original mosaic.

Choosing the appropriate colors for the plaster “tesserae” was more difficult. For the lyre, the left hand and part of the plectrum held in the right hand, 094the colors could be deduced from the surviving parts of the mosaic. For the right shoulder and its garment, the face, the hair, the crown and the halo, we chose various shades of gray-brown to convey the only information we had, that of the black-and-white photographs.

As a result, the restored parts of the mosaic appear schematic in comparison with the original sections. In a review of the restoration, Jerusalem Post art critic Meir Ronen called it “cartoonlike” and “mechanistic.” He has a point, though he obviously was unaware of the limitations we faced, as well as our professional considerations. We too would have liked to do more, had we had the information. But no restoration can truly restore the glory of the original. All in all, however, technical skills joined with updated technology to create a successful exhibition piece. We tried to give the public some impression of how the original looked—to the extent we can now know it. An essential principle of museum restoration is that everything we do must be reversible. We have adhered to this principle in the restoration of the David mosaic; should more information become available in the future, changes can readily be made.

Perhaps some BAR reader has a color picture of the mosaic before David’s face was destroyed, a picture that we have never seen before? We will be grateful for pictures of any quality.

(For further details, see Connie Kestenbaum Green, “The Elusive Face of King David: Restoring a Mosaic from the Ancient Synagogue at Gaza,” Israel Museum Journal 11 [Spring 1993], p. 65.)

Archaeologists uncover the past; museums make it accessible to the public. The restorer’s job lies in between, turning fragmentary remains into pieces that modern viewers can appreciate and understand. In the case of a sixth-century A.D. synagogue mosaic from Gaza, restoration work was aided by advanced computer technology to reproduce an image of King David that was thought to be forever lost. The mosaic first became known to archaeologists in 1966, when the Italian journal Orientalia reported the discovery of an ancient church on the Gaza coast by the Egyptian Department of Antiquities. Within the church was a mosaic […]