Sardis, 2021

Archaeological research at Sardis in 2021 included a three-month season of excavation, site conservation and restoration, followed by another three months of site conservation, research, and construction. As always, work at the site was supported by the permanent staff in Cambridge, Mass. Due to the pandemic the staff was smaller than usual, but after more than a year of isolation, we were all happy to be back together. Vaccines, masks, weekly PCR testing, and other measures kept the entire team free of covid-19 for the summer.

As always, we are deeply grateful to the Ministry of Culture and Tourism, and particularly to Excavations Office Director Umut Görgülü, Nihal Metin, and Özge Yurdakul, for their continual support. Okan Cinemre of the Ankara Museum of Anatolian Civilizations was the representative of the Ministry of Culture for the first three months, and Eyüp Can Kale, of the Manisa Museum, was representative during the last part of the season; we enjoyed and benefitted from their professional expertise and good will throughout the season.

Excavation again took place in two regions of the ancient city (figs. 1, 2). Field 49 is a hill in the center of Sardis, which was terraced with monumental terrace walls in the Lydian period and earlier. We believe this area was part of the Lydian palace, but in most of the hill, Lydian structures have been entirely robbed out in the Persian and Hellenistic periods (fig. 3).

In one spot, however, Lydian walls survive. One wall is a terrace or platform built of finely worked limestone ashlar masonry, similar to the terrace walls on the neighboring hill, ByzFort, and on the Acropolis (figs. 4, 5). In front of this terrace wall was a poorly preserved mudbrick structure, perhaps a small platform or base; this may explain the change of construction from limestone to sandstone in the terrace wall behind. The other wall is narrower, built of rough schist stones. The passage between these two walls may lead into up to the top of the terrace.

The earth floor around these walls was covered by a layer of destruction debris including burned brick, charcoal, ash, and limestone flakes spalled from the wall above by intense heat (fig. 6). Pottery and other finds from this destruction debris date to the middle of the sixth century BC, and we identify the destruction as the famous Persian capture of Sardis in 547 BC, when Cyrus defeated Croesus (fig. 7).

We have excavated this destruction level elsewhere in the city, and it often covers well-preserved domestic assemblages, weapons, and casualties of battle. One of these casualties, a soldier excavated in 1987 near the western fortification, was found with his helmet and a Lydian silver coin, a twenty-fourth stater, near his head. However, the floor excavated in 2021 was covered with only a thin layer of destruction debris, and preservation was not as good.

Among the finds were more than 24 bronze arrowheads in the passageway, evidence of the violence of the battle (fig. 8). In front of the terrace was a scatter of fragments of human bones including parts of one arm and the skull were scattered on the floor, many of them very burned. Left unprotected after the battle, the remains were probably disturbed by both erosion and animals when the city was abandoned (fig. 6).

Among the arm bones was an iron knife, and pocket of loose earth containing nine silver coins (figs. 9, 10). These have not been cleaned yet, but the lion and bull are visible on some, and they seem to be two Lydian croeseid staters, four twelfth-staters, and three twenty-fourth staters. This is the first such find (not a hoard, but a find) of Lydian coins found at Sardis since 1922, and its excellent and closely datable context, together with their owner, makes them of particular significance in the study of this earliest silver currency in the world.

Near the skeleton was a bronze object consisting of about five sheets riveted or corroded together (fig. 11). One sheet depicts a wing with very carefully worked feathers. Two other sheets are riveted together, with decorative false rivets alternating with real ones below a scalloped pattern. It may be a fragment of a large-scale bronze sculpture, damaged in the looting of the palace; perhaps a guardian griffin or sphinx set up near this entrance. This is one of the only pieces of large-scale Lydian bronze sculpture known, and further documents the elite nature of this region. These discoveries help to confirm the elite nature of this part of the city, and our identification as the palace of Croesus.

The second area of excavation was the plaza belonging to the Roman temple of the Imperial Cult. In late antiquity this area was converted to elite housing, and work last year focused on two rooms of one of these houses (figs. 12, 13, 14).

A small room on the east side of the terrace had been used as a lime slaking pit, in one of the many renovations of the house (fig. 15). To the west is a marble-paved courtyard. In 2021 we exposed the western part of the room, including a colonnade with at least five columns. Like the other houses here, this room was destroyed in an earthquake in the early seventh century, leaving the columns toppled on the floor, with one Ionic and one Corinthian capital still in situ (figs. 16, 17). The walls of the room were painted with faux marble, very similar to a room to the south.

The pavement of the eastern part of the court was well preserved, but marble from the western half of the room was salvaged after the earthquake by a large robber’s pit. A coin of Constans II (642-3) was found in a layer of fallen wall plaster which had been cut by this robber’s pit, dating the pit to the second half of the seventh century or later.

A number of other early Byzantine artifacts postdating the earthquake were found in the western part of the court. Near the bottom of the pit was a heart-shaped buckle probably dating to the seventh-tenth centuries. A lead seal of Martyrios Metropolitan, dated to the seventh century AD, was found in an upper layer (fig. 18). And finally, two Byzantine lead amulets were also found in this area (figs. 19, 20). They both show, on one side, a face within a circle surrounded by eight serpent’s heads. Inscriptions on similar amulets identify this as womb or uterus. On the other side is a mounted Holy Rider led by an angel, and spearing a fallen she-demon. Similar amulets have been variously dated from the seventh to the twelfth century. These artifacts are among the relatively scarce evidence for occupation in central Sardis after the early seventh century AD.

Geophysical survey focused on the central part of the city, and particularly on the identification of the seismic fault that runs through the center of Sardis and was responsible for at least two catastrophic earthquakes that levelled the city, one in 17 AD and one in the early seventh century. Prof. Dr. Mahmut Drahor and his team reported on this work in a separate talk at the 2022 symposium.

Work on the monumental arch at the western entrance of Sardis continued, documenting the fallen blocks with a high-resolution photogrammetric model. Other topics of study included human skeletal remains, neutron activation of ceramics, architectural terracottas, sculpture, and conservation of ceramics and other artifacts.

New Carbon-14 dates for Bronze Age and Early Iron Age features on Field 49 refined the chronology of the early phases of Lydian occupation on this hill. An Early Bronze Age terrace fill dates to the later third millennium BC, compatible with ceramics from this fill. A sequence of samples from occupation deposits on this terrace dates to between 2147 BC - 1605 BC (fig. 21). The Early Iron Age mudbrick building in the northern trench was constructed in the ninth century BC, and destroyed in the early eighth century, our earliest example of Lydian monumental architecture (figs. 22, 23).

The Synagogue of Sardis is the largest in the ancient world, and one of the most lavishly decorated. It was restored in the 1960s and 1970s, but rain and weather in the past 50 years have caused the concrete beddings of the mosaics to deteriorate. If left unprotected, this deterioration would continue causing irreparable damage to the mosaics (figs. 24, 25).

Over the past 12 years, therefore, we designed a protective roof over this building, and we completed its construction in 2021 (figs. 26, 27). The roof is supported on the walls reconstructed in the 1960s, not on ancient remains. The concrete beam was covered with masonry to match the Roman walls beneath. Curving steel trusses support a series of alternating “dancing” fabric panels which allow light and air into the building.

In addition to the cracks in the concrete beddings of the mosaic, lichen and bacteria had caused some areas of mosaic to turn almost completely black. While we constructed the roof, therefore, the team of women trained in stone cleaning and conservation removed half a century of biofilm from the mosaics, restoring the original brilliant colors (figs. 28, 29).

The roof covers more than 2,000 m² and will keep the building safe from weather in the winter, and cool for visitors in the summer. We are now considering a more extensive restoration of the mosaics, and also planning a similar roof to protect the nearby Lydian fortification, and open that to the public.