Sardis, 2018

Archaeological research at Sardis in 2018 included excavation, site conservation and restoration, geophysics, publication and study. We are grateful to the Ministry of Culture and Tourism, and particularly to Excavations Office Director Melik Ayaz, and Umut Görgülü. Seval Konak of the Izmir Museum was the representative of the Ministry of Culture; we enjoyed and benefitted from her professional expertise and good will throughout the season (Fig. 2).

The first of the three areas of excavation is Field 49, a prominent hill in the center of the Lydian city (figs. 1, 3, 4, 5). In the 6th c BC and earlier, this hill and the adjacent hill, ByzFort, were terraced with monumental terrace walls, transforming the landscape, creating broad, flat areas that were centers of elite occupation in the Lydian period and long afterwards. The architecture and finds, such as architectural terracottas, an ivory furniture inlay, metal vessels, chunks of jasper, armor and clumps of arrowheads, suggest that in the Lydian period this was a palatial district. The hill was extensively occupied in the Hellenistic and Roman periods; during these periods, Lydian buildings were often robbed and reused in later buildings. The area was converted into a cemetery in the early Byzantine era.

Our goals in excavation were to understand the history of terracing and of buildings on top of these terrace. The north side of the hill was terraced with a massive wall built of boulders, 3 m thick and exposed for more than 41 meters, and probably dating to the 8th c BC (figs. 4, 5). In 2017, however, we discovered that the terrace was expanded in the late 7th or early 6th c BC with the construction of an even more massive boulder wall 5 m thick, on a slightly different orientation, which was subsequently reinforced at yet a different orientation (figs. 6, 7).

In 2018 we excavated further to the west to expose more of this wall, and discovered that it had been almost entirely robbed out, probably in the Persian period. This adds weight to the observation made in earlier years, that the Lydian buildings go out of use after the capture of Sardis in 547 BC, and the hill is not occupied again until the Hellenistic period.¹ The only traces of use and occupation on the hill probably relate to robbing. The Achaemenid Persians, therefore, do not seem to have taken over the Lydian palace and institutions here, but left the city within the walls abandoned.

A trench on the western slope of the hill produced the longest sequence of occupation at Sardis (figs. 4, 8, 9). A surprising discovery in 2018 was the discovery of a room dating to the 14th or 15th c AD, one of the only traces of Medieval settlement in central Sardis (fig. 10). A complex sequence of monumental Hellenistic buildings, mostly built from reused Lydian limestone blocks, plausibly belongs with the establishment of Sardis as a Seleucid capital in the 3d century AD (fig. 9).

This trench produced evidence for Achaemenid activity. A large pit cut into the Lydian terrace fill contained large quantities of animal bones, and luxury objects such as ivory inlays, iron lamellar armor, and 28 bronze arrowheads (figs. 11, 12). The latest pottery from the pit dates to the fifth century, as does a tiny silver coin of Teos (fig. 13). However, the ivories, armor, and other small finds may well be detritus from looting the contents of the Lydian palace, dating to the first half of the sixth century BC. We discovered somewhat similar detritus in 2011-2013 in a trench to the south, where disturbed, mixed strata produced items such as the ivory furniture inlay, sealstone, metal vessels, and also human bones from at least two individuals, presumably casualties of the sack of 547 BC.²

Robbing in the Achaemenid and Hellenistic periods seems to have removed most walls of the Lydian palace, leaving deep robbers’ trenches. One face of a massive limestone wall 3.7 m thick was discovered in this central trench in 2017. The back face of this wall, however, had been robbed out to a depth of about 3 m (figs. 7, 8). A single lamp fragment of Howland Type 21 suggests a date in the Persian period for the robbing.

Another important goal of excavation here was to investigate the terrace fills, to learn more about the chronology of the terrace. About 5 m of Lydian terrace fill consisting of almost sterile gravel and sand, sloping down to the west, dates to the first half of the 6th c BC and so probably belongs with the latest Lydian phase of the terrace, built of limestone ashlars (although that phase of the terrace wall has not been discovered yet in this trench; fig. 14).

Beneath the 6th century deposit, however, we did not find terrace fills dating to the 7th or 8th centuries BC. Instead, below the Lydian terrace fill was about 2 meters of stratified occupation deposits dating to the Late Bronze Age. This consisted of layers of earth with lenses of ash and burning, laid more or less horizontally. A number of partly restorable pots such as “drums” are similar to those found at Beycesultan and elsewhere (fig. 15).

Beneath the Late Bronze Age occupation strata, the consistency of fill again changed, to a series of sloping, almost sterile fills of gravel and sand, remarkably similar to the Lydian terrace fills above. Pottery from these fills dates to the Early Bronze 3B period, about 2,000 BC (fig. 16). This Early Bronze Age fill rests on bedrock, sloping down steeply to the west, to a depth of 14 meters below surface (fig. 17).

This small, deep sondage produced a number of surprising results. First, we have never discovered strata of the Early Bronze Age at Sardis before; this is the earliest level ever reached, by some 800 years. Occasional earlier objects have been found, such as a fine mace head and a number of celts; but always in later contexts, and we had always imagined that they were collected by Lydian and later inhabitants from other sites.³ Although the Sardians prided themselves on being “protochthonous” and “autochthonous,” Strabo (13.6) tells us that the city was only founded after the Trojan War, relatively late compared with cities like Miletus and Ephesus. Apparently, Strabo was wrong.

Second, the nature of occupation is interesting. Rather than occupation levels like the Late Bronze Age strata, the four-meter-thick layer of sloping, almost sterile sand and gravel fill is so similar to later Lydian terrace fills, that it is hard to believe that it is anything but another massive terrace fill. And the sloping bedrock beneath would not have been habitable without a terrace to level the surface. We have not yet uncovered the terrace wall that would have retained this fill, of course, and since it is buried at least 15 meters deep, it is unlikely that we will do so. But our current hypothesis is that this four-meter-thick sloping stratum of gravel belongs to a terrace probably dating to EB3B, and certainly earlier than the Late Bronze Age occupation levels above. These early Sardians were apparently already transforming the natural landscape through the construction of artificial terraces, an approach to urban development which we had always considered characteristic of the Iron Age Lydians.

Below the Lydian palace is another artificial terrace known as Field 55 (figs. 1, 3, 18, 19, 20). In the early Roman period this belonged to a sanctuary of the Imperial cult, focused on an octastyle pseudodipteral temple. In the late Roman period, it had been transformed to elite housing with painted rooms and well-preserved domestic assemblages. The excellent preservation is due to the sudden destruction by earthquake in the early 7th century AD, which left houses and nearby fortification walls toppled.⁴

Excavation in 2018 opened a new area north of the trench dug in 2013-2017. One space was probably an open marble-paved courtyard, with traces of what might have been a basin or other hydraulic feature (fig. 21). The room was much less well preserved than the spaces excavated in previous years, however, probably because the area had been dug out after the destruction to recover the marble pavement blocks.

One wall of this court was painted to resemble colored marble revetment (fig. 22). In the earthquake that destroyed the area, this wall was split and offset by a few centimeters, and fell nearly intact into the space to the east. Three arched windows were well preserved in the collapse, measuring about 0.7 x at least 0.9 m (figs. 23, 24). We hoped to discover further well-preserved domestic assemblages under this wall, like those found in the rooms to the south. However, the room had apparently been abandoned before the earthquake, leaving only a bare earth surface. Among the spolia incorporated in the late Roman wall, however, were part of a monumental Latin inscription, and the lower part of a draped female figure, both probably spolia from the early Roman sanctuary (fig. 25). Deeper excavation intended to clarify the earlier history of the terrace were inconclusive.

The third area of work was the western entrance to the city (figs. 1, 3, 26). In the Lydian period a road led through a gate in the Lydian fortification. In the Roman period, the descendent of this road, now a broad, colonnaded avenue, led through a monumental three-bayed arch. The central bay of this arch was 13 m wide, the widest known in the Roman world. This arch, and the neighboring Synagogue and other structures, collapsed in the same 7th century earthquake that destroyed the houses in Field 55, and the main road through Sardis was never cleared.

Work in 2018 focused on further excavation and stabilization of a mosaic in the north portico of this road, which bears a dedicatory inscription by its patron, Flavius Maionios, “magnificent count and governor,” probably dating to the late 6th century (figs. 27, 28, 29). Cleaning was made much easier by the generous donation of a Lynton Phoenix conservation laser, which removed a thick layer of concretion that covered parts of the mosaic without damaging the fragile tesserae (fig. 30). This project was funded by a generous grant from the American Embassy in Ankara, and we are now constructing a glass floor over the mosaic to display it to visitors.

Two deep sondages were intended recover stratigraphic information to date the arch and associated structures. Sealed pottery associated with the arch was very scarce, and its date is still uncertain. One sondage did reveal earlier road surfaces, however, and also the “Brick Fall” destruction debris associated with the Persian capture of the city in 547 BC.

The sector was transformed by stabilizing and restoring the Lydian and Roman structures excavated in the 1970s, 1980s, and 1990s, including the Lydian Gate, Persian blockage walls, and Roman colonnades. Scarps were straightened and reinforced with drystone walls (figs. 31, 32, 33).

New 3d reconstructions of the fortification, gate, Roman arch, and other features of this area enable us to better visualize and explain its complicated history (figs. 34, 35, 36).

2018 was the fifth and final year of the program to clean the Temple of Artemis by “disinfecting” the marble blocks, treating them with a gentle biocide that kills the damaging bacteria and lichen without harming the stone. We focused on the southernmost standing column of the east peristyle, and on fallen blocks in the sanctuary; with the completion of these areas, the project was brought to a successful conclusion (figs. 37, 38, 39, 40, 41, 42, 43, 44). This has been one of the most satisfying projects at Sardis, and it has been important to the team of 14 or so women and men who have worked so hard, sometimes under difficult conditions on vertiginous scaffolding, and ensured that every aspect of the project was a success. It is especially rewarding as we bring Fikret Yegül’s 30-year publication project towards completion this year.

A geophysical survey led by Prof. Dr. Mahmut Drahor produced excellent results despite the very complicated stratigraphy and topography of Sardis. The GPR survey focused on the region around Field 49, and revealed more of the city plan and buildings in the city center (fig. 45).

Anthropologist Prof. Dr. Yılmaz Erdal and his team studied a late Roman multiple burial from a Lydian tumulus excavated in 1966 at Duman Tepe (BT66.1).⁵ They identified 92 individuals and point out that there are no signs of trauma suggestive of violent death, while the age distribution of the deceased does not suggest that they died of a plague or other epidemic.

Monograph publications of the last year include Prof. Jane DeRose Evans, Sardis Monograph 13: Coins from the Excavations at Sardis: Their Archaeological and Economic Contexts: Coins from the 1973 to 2013 Excavations (2018). Like all our publications, they are also available for download from our web site (http://sardisexpedition.org), and we are incorporating the information into our on-line searchable database.

• 1Cahill 2019.
• 2See Cahill, “Recent Fieldwork at Sardis” 123; Dusinberre, “A New Seal from Sardis,” and preliminary reports for the 2011, 2012, and 2013 seasons.
• 3Mace head: S05.007:11959 (Greenewalt, “Sardis, 2005” fig. 12); celts such as S10.010:12610 from a mid-6th c BC Lydian house southwest of the Temple of Artemis: Cahill 2012, fig. 31.
• 4See preliminary reports for 2013, 2014, 2015, 2016, 2017.
• 5Hanfmann, “Ninth Campaign at Sardis,” 47-52.