Sardis, 2019 and 2020

Archaeological work at Sardis in the summer of 2019 included excavation in three regions of the ancient city, conservation and site preservation, geophysical survey, research, and publication, with a team of 70 professionals and students. Due to the pandemic, the 2020 season was greatly curtailed, but with the support of the Ministry of Culture and Tourism and Harvard University, a team of 11 was able to hold a 5-week field season of site conservation and research, while the rest of the team worked remotely on research and publications. As always, we are deeply grateful to Director General Gökhan Yazgı, to Kazılar Dairesi Başkanı Melik Ayaz and Umut Görgülü, and Kazılar Şübesi Müdürü Y. Nihal Metin of the Ministry of Culture and Tourism, for their constant help and support in good times and in difficult times. Cengiz Aslantaş of the Izmir Museum, representative of the Ministry of Culture and Tourism in 2019, and Sedrettin Öğünç, also of the Izmir Museum and representative in 2020, greatly contributed in many ways to the success of both seasons, and we are grateful to them for their kind care and good will.

A natural hill at the foot of the acropolis called Field 49 has been a focus of excavation since 2009 (figs. 1, 2, 3). Archaeological remains here span the entire history of the ancient city. In the Lydian period we believe that this was part of the palace complex, with elite buildings set on imposing terraces above the lower city. Monumental Hellenistic remains suggest that it might have been a royal complex in this period as well. In the intervening Persian period, however, we so far have found no signs of occupation, only robbing and spoliation of blocks from earlier buildings. Occupation levels predating the Lydian period include an important Early Iron Age level, and beneath that, Late Bronze Age and Early Bronze Age strata, exposed so far only in small sondages. Dense Roman phases of the 1st - 6th century AD were largely abandoned in the 7th century AD, but occupation in the later 7th to 9th centuries AD is attested by a cemetery on the hill, and reoccupation in the 14th-15th centuries AD was recently discovered.

Excavation in 2019 focused on four trenches (fig. 2). A trench on the north end of the hill clarified the long history of Lydian terracing. One important question addressed this season was the phasing of the 47-meter-long boulder terrace wall that now dominates the hill (figs. 3, 4, 5). Analysis of the stratigraphy and construction suggests that the wall may have two major building phases, an earlier phase in the 8th century BC, and an extensive rebuilding in the 6th century BC. This hypothesis awaits confirmation through excavation of further sealed deposits. A lower terrace wall, 5 meters thick, was probably constructed in the late 7th or 6th century BC and expanded thereafter, but was largely robbed out in the Persian period. The course and construction of this lower wall were also clarified in 2019.

An important discovery of the 2019 season was a corner of an early Lydian mudbrick building predating the earliest phases of the stone terrace walls. The new corner suggests that the building includes a room about 5 m wide, perhaps with adjoining rooms on the south and east. Its mudbrick walls are 80 cm thick, and were reinforced by thick wooden posts at 1.1 m intervals; its roof was supported by a larger post in the center of the space (figs. 6, 7, 8). The substantial walls and wooden structure illustrate the large scale of the building, which seems hardly domestic. As in previous years, ceramic finds were sparse and not closely datable, but suggest a date in the 8th or 9th century BC; and a C14 date of charcoal from one of the posts offered a range from 996-837 BC (95.4% probability).¹ Such a date makes this among the earliest monumental buildings known at Sardis. Other strata and architectural features of the Early Iron Age were discovered in previous years downslope from this building, and may be recognized elsewhere on the hill and in the vicinity.

A trench on the west slope of the hill preserves a series of important Lydian, Persian, Hellenistic, and Roman structures, as well as strata dating to the Early Bronze and Late Bronze Ages (figs. 2, 9). Work in previous years had removed selected Roman and Hellenistic features in order to allow access to earlier strata; work in 2019 focused again on Hellenistic features including walls, drains, and earth floors, mostly of the 2nd century BC (figs. 10, 11). These were apparently all cut by the monumental Hellenistic terrace or platform built of reused Lydian limestone and marble blocks, largely exposed in previous seasons.

A large pit of the 5th century BC in this trench is one of the few coherent deposits of the Persian period from this hill and indeed from central Sardis. The pit contained large quantities of pottery and animal bones, as well as arrowheads, fragments of scale armor, and ivory inlays. The arrowheads, armor, and ivory may be refuse from earlier Lydian buildings which were systematically looted at this time. The purpose of this pit is unclear. It is not associated with any known occupation level on this hill, but if it is a robber’s trench, it is not clear what earlier feature(s) it was robbing.

Further exploration of the Bronze Age layers was not possible since the sondages where these strata have been exposed had both been excavated to bedrock, and later remains constrain the area where we can explore these early levels. But a C14 sample from the lower levels of the Early Bronze Age terrace fill dates to 2346-2140 BC (95.0% probability).

Excavation further down the west slope of the hill exposed more structures built against the terrace at a lower level (figs. 9, 12). A room exposed in 2018 was built from unmortared fieldstone walls. Pottery from the room suggested a date in the 14th or 15th century AD, the latest occupation known in central Sardis. The adjacent room to the south, however, uncovered in 2019, is quite differently constructed of banded brick masonry and spolia, with a plastered interior. All the doors were blocked, and the room was filled with 2 m of loose mortar, rubble, tiles, and fragments of fallen masonry, much of which seems to be debris from an earthquake. Floor level was not reached, and the date of the potential earthquake, and of the architecture of the room, remain undetermined.

Two other trenches, at the northwest and southern parts of the hill, exposed multiple phases of Roman building documenting the long and intense occupation of this part of Sardis (figs. 2, 5, 13). The intersecting and often poorly preserved architectural phases did not allow a coherent reconstruction.

Geophysical survey of this hill in 2019 used GPR and deep-penetrating ERT, IPT and other techniques to map the bedrock and man-made alterations of the hill, in order to understand better how human activity has transformed the natural topography of this region of the city.

This lower terrace in central Sardis was the site of a Julio-Claudian temple and sanctuary of the imperial cult, and subsequent late Roman elite housing (figs. 1, 14). Work in 2019 again focused on buildings on the east side of the terrace, exposing more of a marble-paved courtyard and rooms to its east by joining the trench dug in 2013-2017 with that dug in 2018 (figs. 15, 16). Like the rooms to the south, these areas seem to have been destroyed by an earthquake in the early 7th century, but unlike the southern rooms, they do not seem to have been occupied at the time of destruction.

The full dimensions of the court (Room 9) have not been established, but it is at least 6.7 x 13 m (figs. 17, 18). It was paved with spolia, and was bordered on the east and south by raised platforms which supported a marble basin, draining into an open channel to the south. An opening had been roughly cut through the south wall of the court into the adjacent space, perhaps to accommodate this drain; this may be a late feature of the complex. Doors on the east, however, were all blocked, raising questions of access to those spaces. Painted wall plaster on the east wall, in a faux-marble pattern very similar to that of Room 1 to the south, suggests that the perimeter of the court was roofed, but no column bases or other firm evidence of a colonnade have been exposed. Few artifacts were discovered in this space and little articulated fallen masonry, in contrast to the adjacent rooms, suggesting that it was abandoned at the time of the earthquake and also subject to extensive salvaging after the earthquake.

When partly excavated in 2018, the adjacent room to the east (Room 10) preserved the collapsed remains of the dividing wall, complete with three arched windows. This did not continue to the southern part of this space excavated in 2019, however. Instead, a small roughly-built structure measuring about 3 x 3.7 m was inserted into the fallen debris (fig. 19). This postdates the earthquake(s) that brought an end to much of this residential structure in the early 7th century AD. A trapezoidal room (Room 11) to the south was partly excavated, but floor level was not reached.

A deep sondage within Room 10 was dug to levels of the 1st century AD, probably the terrace fill associated with the construction of the Julio-Claudian sanctuary in the decades after the earthquake of 17 AD.

A salvage excavation in 2019 of a tomb about 3 km SW of Sardis, under the direction of the Manisa Archaeological and Ethnographic Museum, uncovered a small cist grave, probably of the Persian period and probably originally covered by a tumulus. The hillslopes nearby are dotted with a number of other Lydian tumuli. The cist was built from marble slabs clamped together with lead and iron pi-clamps, with a marble floor and flat ceiling. The tomb had been looted in antiquity, but much of the skeleton remained, as did five small gold beads and a gold finger ring, missed by the looters (figs. 20, 21). The skeleton, analyzed by Prof. Yılmaz Erdal, belonged to an elderly woman of about 35-60 years. The aryballos-shaped gold beads are similar to beads discovered by the Butler Expedition, probably dating to the 5th century BC.2

Conservation, preservation, and presentation of ancient structures and artifacts of all periods remained a priority for the 2019 and 2020 seasons. In 2019 conservators completed the conservation of a Roman mosaic in sector RT / MMS/N, whose inscription records the donation of this mosaic portico by one Flavius Maionios (see Petzl, Sardis: Greek and Latin Inscriptions Part 2 no. 424; figs. 22, 23, 24). A steel and glass floor was constructed over the mosaic to present it to visitors, while informational signage explains the archaeological remains in this area (fig. 25).

A protective mudbrick “skin” was constructed against the cut face of the Lydian fortification where it had been destroyed in the 1950s during the construction of the Izmir-Ankara highway (fig. 26). This also served as a trial run for a similar skin planned to protect the original east face of the fortification. The team also repaired blocks of the limestone gate in the fortification in sector MMS/N, and consolidated and restored other features in this sector.

A team of women and men cleaned the Marble Court and other remains in the area of the Bath-Gymnasium complex in 2019 and 2020, using the same techniques developed to clean the marble of the Temple of Artemis (fig. 27).

Working throughout the year, and working from home during the 2020 pandemic when the Sardis research center at Harvard University was closed, the editorial team brought out two final reports in 2019 and 2020. Prof. Georg Petzl’s Sardis: Greek and Latin Inscriptions Part II: Finds from 1958 to 2017 (Sardis Monograph 14, 2019) presents 488 inscriptions with text and commentary, including important collections of texts from the Roman sanctuary at Field 55, the Synagogue, and other parts of the ancient city and its surroundings. Prof. Fikret Yegül’s The Temple of Artemis at Sardis (Sardis Report 7, 2020) publishes his documentation and research on this magnificent building, with stone-by-stone plans and elevations of the building drawn at 1:20. Like all our publications, they are available on our web site https://sardisexpedition.org.

• 1These dates were further refined in 2021: see KST report forthcoming.
• 2For instance Curtis, Sardis XIII: Jewelry and Gold Work (1925) nos. 24, 29, 44, 57.