Sardis, 2009

The Sardis Expedition conducted a ten-week program of excavation, conservation, restoration, and research during the summer of 2009, continuing into the fall with further restoration and study. We gratefully acknowledge the continuing support of Minister Ertuğrul Günay, Director General Orhan Düzgün, Deputy Director Nermin Beşbaş, Excavations Division Director Melik Ayaz and Excavations Division Officers Serap Kocaman and Serkan Atalay for permission to excavate and do research in Turkey, and for their support and good will. Uğur Terzioğlu, of the Samsun Museum, was the government representative this season, and we greatly appreciate all his generous help and advice during the season.

Excavation was undertaken in four sectors of the city: at the western fortification, at two locations in central Sardis, and on the Acropolis (fig. 1). At sector MMS, work focused on an earlier Lydian structure, probably a defense, underneath the monumental Lydian fortification. A trench begun in 2008 had revealed a corner of the structure, which had also been glimpsed in 1977. This corner is not a gate or opening in the wall, since the foundations continue on to the south; it may rather be the corner of a recess like those found in the later fortification. In 2009, the trench was expanded to the north to reveal more of this wall. The wall is interrupted by a stone feature, with a very rough “face” at right angles to the line of the fortification (figs. 2, 3). This line of stones is probably the back edge of a retaining or terrace wall; the deposit south of this face was an artificial fill consisting of lenses of different types of earth, consistent with this interpretation. The north side of this wall could not be exposed, as it is under the modern road; the retaining wall, if this is what it is, is at least three meters thick (fig. 4). This earlier phase of the fortification, therefore, seems to have had some of the elements of design of the later wall, such as a recess in the original construction, and a later reinforcement, with an earthwork supported by a massive retaining wall. The date of this earlier phase is probably in the second half of the seventh century BC.

In central Sardis, excavation within the Hellenistic and Roman theater concentrated on earlier levels predating the theater (fig. 1). More fragments of terracotta figurines associated with the goddess Cybele were found in the earliest layer of fill of the theater, suggesting that a sanctuary of the Mother goddess was located somewhere nearby in the third century BC.

Underneath the Hellenistic fill of the theater was a well-preserved Lydian building, probably a house. One space of this house, presumably an open space with a stone paved floor and limestone column, had been excavated in previous seasons. In 2009 a second room, measuring about 3.5 x 3.9 m, was cleared almost to floor level (figs. 5, 6). Entered through a door from the previously excavated space to its east, this was probably roofed; it may also have had a door to the west. Its floor area was largely taken up by two mudbrick benches or platforms, whose function was not completely clear. The corridors between these benches were littered with artifacts, including at least nine cooking pots, fragments of at least one Myrina amphora, at least one stemmed dish, at least one omphalos phiale, a lid, eleven loomweights and ten spindle whorls, and other pottery which remains to be cleared, mended and studied (figs. 7, 8, 9). Among the complete finds were a lamp, a lydion, a juglet, and a Corinthian warrior aryballos (fig. 10). Fragments of other imported vases of this same date were discovered as well, including two sherds of an Attic black-figured cup, perhaps a proto-Cassel cup, and a fragment of a Chiot chalice. Datable to the second quarter of the sixth century BC, these are the first solid evidence that the destruction of this house, like the houses previously excavated near the fortification, is to be attributed to Cyrus the Great in about 547 BC. One unusual local find is a conventional Lydian streaky-glazed skyphos; its accidental burning in the destruction of the house revealed part of a design that had apparently been painted over, including part of a lion and filling ornaments (fig. 11).

The theater was built between a series of natural spurs of the Acropolis. Two of these spurs were investigated in the 1980s and 1990s: the so-called Byzantine Fortress (sector ByzFort), and Field 49 (sector F49) (fig. 12). The ByzFort spur was revetted in the sixth century BC with a terrace wall of limestone ashlar blocks (fig. 13). The north slope of Field 49 was terraced with a wall of more roughly worked polygonal boulders, perhaps earlier, in the seventh century BC (fig. 14).

A trench on the west slope of Field 49 (fig. 15) revealed a sequence of Roman walls and foundations, dating probably to the first through fifth centuries AD. Most occupation levels within the excavated area associated with these structures had eroded away, leaving only subterranean foundations and pipes.

These Roman foundations were built into a 2-meter deep fill of earth, containing a rich deposit of early Roman pottery, lamps, terracottas, coins, glass, wall plaster, and other artifacts dating to the first half of the first century AD. Among these finds was a fine bronze patera handle with dragon’s heads and dolphins (fig. 16). The assemblage resembles fills with similar artifact assemblages from elsewhere at Sardis, and may be associated with the cleanup after the earthquake that destroyed Sardis in 17 AD. A stone pavement beneath this fill may represent a floor in use when the earthquake struck. Below this floor were a series of earlier Roman and Hellenistic layers, including at least one oven.

All these Hellenistic and Roman features were built against a monumental terrace wall built of well-cut limestone ashlar blocks (fig. 17). Much of this wall had been destroyed and robbed out, and its blocks reused in the Roman walls; but further cleaning to the south proved that limestone construction rises up to 4.3 m high (fig. 18). It is unclear whether the upper portion of the wall is original or rebuilt. A layer of limestone chips in front of the wall probably represents construction debris; the small amount of pottery associated with this fill seems to date to the first half of the sixth century BC, approximately contemporary with the similar limestone terrace wall on ByzFort. The workmanship of the two walls is somewhat different, however, as the new wall is built from smaller blocks, and does not seem to have the distinctive drafted margins characteristic of the foundation courses of the ByzFort terrace and other Lydian ashlar construction. This new terrace wall is also different in construction and orientation from the rougher boulder terrace wall excavated in 1982 on the north face of the Field 49 spur; it may belong to a different phase of Lydian terracing. Multiple phases of terracing are also attested on the ByzFort hill.

The upper region of central Sardis in the Lydian period, therefore, seems to have been built up in a series of monumental terraces, encompassing these two natural spurs; or, perhaps, a single monumental terrace, wrapping around both spurs. Moreover, this new terrace wall is exactly aligned with the house in the theater, suggesting that a single scheme of urban organization encompasses this whole region.

Finally, excavation continued on the acropolis, where illicit digging in 2007 had revealed a fill containing fine pottery of the Lydian and Persian periods, partly excavated in 2008 (figs. 19, 20). As previously, we found no trace of architecture in situ, but the discovery of many fragments of rooftiles and architectural terracottas, including a fragment of a sima with a Pegasus and small fragments of a disk akroterion in the shape of a gorgoneion, suggests that high-status building(s) stood somewhere nearby. Other finds from this fill include fragments of at least two Attic black-figured cups of the mid-sixth century BC, Achaemenid bowls and late column kraters of the later sixth or fifth centuries BC, and a fragment of a faience vase. The ceramics and other finds from this fill range over a period of a century or more, from the early sixth century (?) into the second quarter of the fifth century BC.

Two Lydian coins were discovered in this fill: one electrum third-stater, similar to one found in 2008, and one silver croeseid stater, bringing the number of Lydian coins found in this small area to three (figs. 21, 22, 23, 24, 25). Like an electrum coin found in this same trench in 2008, the new electrum coin has a golden-colored surface, but cracks and pits reveal more silver-colored core. This difference in composition between surface and core is usually understood as the result of natural depletion due to prolonged burial. The two electrum coins were analyzed by Prof. Dr. Bulent Önay of 9. September University in Izmir, with an SEM with EDX attachment. The cracks and pits reveal the core of the coins, which consists of about 46-49% gold and 44-47% silver. The surface composition varies widely, from 57% gold / 35% silver to 89% gold / 6% silver. Moreover, the composition varies in closely adjacent areas of the coins, apparently the result of wear and damage to the surface. Raised areas such as the snout and eyebrow of the lion are visibly more silver-colored, apparently because they were more abraded and worn; and analysis showed that they had a lower gold content. The metal inside a scratch on this year’s coin had a lower gold content than did the surface around it, suggesting that the scratch had cut through an enriched layer into the depleted core. The surface enrichment of the coins therefore seems to have occurred before the coins were circulated, rather than being the result of natural processes; otherwise it would not show patterns of wear from the circulation of the coin. It has long been hypothesized that the Lydians deliberately added silver to “dilute” the alloy of natural electrum from about 70-80% gold, the proportion thought to be found naturally in the Pactolus river, to about 55% gold, thus reaping a significant profit. I tentatively suggest that in addition to diluting the alloy, the Lydians artificially enriched the surfaces of these coins after they were struck, to make them appear as if they contained the original amount of gold, reflecting the value at which they circulated.¹

Site conservation and restoration included cleaning and stabilizing Late Roman wall paintings in sector MMS, and lifting, stabilizing, and resetting mosaics of the Late Roman Synagogue, which had been damaged by freezing during a particularly cold winter. As part of an ongoing program to replace aging temporary shelters over excavation sectors, a new temporary roof was built over an area of the Lydian fortification and Roman colonnaded street and houses in sector MMS/S. A program of informational signage was begun throughout the site: seven signs were designed, printed, and installed, at the entrances to the site, at the Artemis Temple, Bath-Gymnasium Complex, Synagogue, and at sector PN.

The project begun in 2007 to restore a corner of the marble revetment in the Marble Court, was brought to completion. Panels of Afyon marble were fixed to a steel frame, set in modern restoration to avoid damage to the original structure. The pattern was based as far as possible on existing remains, but where the design is uncertain — as it is for most of the superstructure — the reconstruction was kept deliberately simple (figs. 26, 27, 28).

(This article is adapted from the Kazı Sonuçları Toplantısı 32: 358-367)

• 1See Cahill et al. forthcoming.