Sardis, 2016

In 2016 the Sardis Expedition conducted a season of excavation, conservation, research, and site enhancement with a staff of 55 scholars, professionals, and students from 30 different universities and institutions. We are again grateful to the General Directorate for Cultural Heritage and Museums of the Ministry of Culture and Tourism for permission to carry out archaeological research in Turkey, and particularly to Excavations Division Director Melik Ayaz and Excavations Branch Director Umut Görgülü for their continual support. This year’s representative of the Ministry of Culture and Tourism was Necla Okan of the Izmir Archaeological Museum, whose good will, careful advice, constant help and expertise were greatly appreciated by all and helped make the season both productive and enjoyable.

Excavation was carried out in three areas: on a high hill in central Sardis known as Field 49, where we believe the Lydian palace was located; in a lower terrace known as Field 55, which was a sanctuary of the Imperial Cult in the early Roman period, and was converted to other uses in late antiquity; and at the western entrance to the city where we have uncovered the remains of a large Roman arch.

Excavations in Field 49 were directed at understanding complex Roman and Hellenistic strata in order to expose more of the deeply buried Lydian remains of the palace (fig. 1). On the north slope of the hill, the terrace is retained by a massive Lydian boulder wall 3 m thick and exposed for about 41 m (fig. 2). It must turn a corner in the northwest, and has been uncovered on the west slope of the hill, about 90 m to the south. The wall has usually been dated to the seventh century BC, but it may be significantly earlier.

Excavation to deeper levels within the terrace was made possible by the documentation and removal of one foundation of a Hellenistic room dating to the third or early second century BC. This allowed us to expose and excavate more of a layer of Achaemenid fill which has been encountered in previous years. Although the fill contains relatively small amounts of pottery of the later sixth and fifth centuries BC, the majority of the material seems to be rather earlier. In this fill were found beautifully worked marble wall blocks from an older building now destroyed, and fragments of two enormous bowls or perirrhanteria.

Among the most interesting artifacts from this trench, however, are clumps and individual scales from a suit of iron scale armor (M13.011; figs. 3, 4, 5). Parts of this same corselet were found in this stratum in 2012, 2013, and 2015; this year more than a thousand additional fragments were found. The armor scales come in two basic shapes: fish-scale types with curved bottom edges, and elongated rectangles. These probably belong to different parts of the same corselet, the smaller scales used for the shoulders and arms, the larger ones for the lower torso. In all, about 1350 scales and more than 200 scale fragments have been recovered between 2012 and 2016. This might represent 30-40% of the original corselet, which might have had about 3200 scales and might have weighed about 9.4 kg. Such scale armor was used from the Late Bronze Age into the Roman period, and examples are known from a number of Near Eastern sites including Nuzi, Nimrud, Persepolis, Gordion, Kamid el-Loz, Idalion on Cyprus, and Egyptian tombs.¹

We exposed more of a substantial mudbrick wall with gaps for timber supports, which is the earliest feature in this trench (figs. 2, 6). This wall probably predates the boulder terrace wall, but its date too is uncertain. The discovery of just a corner of a burned stratum on the floor within this wall, however, offers hope for a better date in the future.

A nearby trench uncovered a series of poorly preserved Roman and Hellenistic foundations and robbers trenches, associated with waterworks and two ovens (figs. 1, 7). A foundation of large boulders is similar in construction to the Lydian boulder wall and may well belong to the same phase, but independent dating evidence was lacking.

In a trench in the western part of the hill, work also focused on understanding the complex Roman and Hellenistic stratigraphy (figs. 1, 8). Within the Roman building explored in previous seasons, at least four phases of Hellenistic occupation were distinguished. The Lydian terrace wall may have been reconstructed in the Hellenistic period reusing the original blocks, and we seem to have the top of the earlier, Lydian phase belonging with the palace (fig. 9).² We could not uncover more of this because of the complexity of later remains.

A small sondage into the terrace fill in the corner of the trench, however, revealed unexpected results. The fill consisted of layers of earth, coarse gravel, and mudbrick debris, all sloping steeply down to the west (fig. 10). This fill is more than 5 m deep, reflecting an enormous terracing project. The pottery from this terrace fill was very sparse, but that from the lower levels was unexpectedly early, including both early Iron Age material such as cups with pendant semicircles, and Bronze Age ceramics such as carinated bowls with basket handles, similar to Bronze Age examples from Kaymakçı, Beycesultan and elsewhere (fig. 11). Two C14 samples from this fill were analyzed by TÜBİTAK Marmara Araştırma Merkezi. One sample dated to 1051-902 BC, the other 1747-1617 BC. The discrepancy between the dates may reflect the mixture of Bronze Age and Early Iron Age material in this fill, but I emphasize that this is based on a very small sample and much more work needs to be done. However, this is the first time we have recovered such early material here in central Sardis, and we hope that future work will illuminate the date of the terrace wall and the nature of early occupation here.

On the lower terrace known as Field 55, our goals were to explore the late antique features on the east side of the terrace (fig. 12,13, 14). It seems that the Roman neocorate temple and its sanctuary were dismantled sometime in the late Roman period, and the fragments reused in a “Spolia Wall” built almost entirely of statue bases, inscribed pedestals, and architectural fragments from the temple of the imperial cult.

One major question has been the nature of this spolia wall: does it belong to a building, to some other sort of structure? This was tentatively answered in 2016 with the discovery of a marble-paved road under the gravel with wheel ruts running north-south (fig. 15). The spolia wall ends neatly with reused inscriptions forming a thickened end; a deep socket is carved into the end of the wall. On the other side of the road, we uncovered the edges of blocks which seem to belong to a corresponding wall end. Other enormous reused inscription blocks found here were originally set upright into cuttings in the road, and further rough cuttings in the blocks supported wooden timbers. The ensemble may be reconstructed as a gate spanning the road, thus closing off the area to the south from that on the north. A gate in a massive wall suggests that the spolia wall might belong to a previously unsuspected late Roman fortification, perhaps meant to take advantage of the natural topography to enclose the terrace, stadium, and theater. Erosion and stone robbing have removed all traces of this wall over most of its course, but here in this low point, collapse and subsequent silting has preserved the wall. The situation might be similar to the “Gotenmauer” and the Byzantine fortification at Miletus. A small probe showed that it was a later addition to the terrace, and suggested a date for this fortification in the late 5th or early 6th c AD.

The conservators continued to clean and stabilize the assemblage of artifacts found in 2015 in a room adjoining the spolia wall, including bronze vessels, two two marble sigma tables and two platters, a six-lamped polykandelon, and a lamp stand (fig. 16).

Work on what is probably a late Roman house on top of the terrace has been slow due to the very fragile plaster on the walls. The continued efforts of a team of conservators allowed excavation of a second room, however, and revealed that the wall was decorated in what seems to be fabric draped along the walls in broad swaths, printed or woven with floral patterns (figs. 17, 18). Against the west wall of this room was a marble monopod table on a griffin foot; next to it was an iron sword—making five from these two rooms—and a spearhead, a hoe, a glass jar and other objects (fig. 19).³

West of the two painted rooms lay an architecturally complex group of spaces. One space was perhaps a courtyard with a round reused column shaft in its center. At the south side of the courtyard was a collapsed brick structure, perhaps a stairway. A small adjacent space was probably a latrine and is probably a later addition.

This area collapsed in one or more earthquakes in the early seventh century; as reported last year, the latest coin from the area is a follis of Heraclius dating to 611-612 AD. The stratigraphy suggests, however, that there were at least two events, with a short phase of rebuilding in between. This is generally consistent with evidence from elsewhere in the site for a major earthquake in the early 7th century — elsewhere the coins are a few years later, of 616 AD or so. An important question is how these seismic events are related to earthquakes also at Hierapolis, Laodikeia, Aphrodisias, Miletus, and other sites in western Anatolia, all of which seem to have suffered catastrophic collapses at about this time.

The third excavation area was aimed at opening more of a broad Roman colonnaded avenue at the western edge of the city to visitors to the Synagogue and Bath-Gymnasium complex (figs. 20, 21). Two years ago we unexpectedly discovered that its pavement was littered with fallen marble blocks, belonging to a huge arch that spanned the avenue. In 2016 we expanded the excavation to the west to uncover the fallen west face of the arch. This includes blocks from the outer façade, most of them again reused column drums from the temple of Artemis; and a few column shafts, Corinthian capitals, and other architectural elements whose position on the arch remains uncertain. The distribution of the blocks shows they fell more or less in situ (fig. 22). So it was particularly satisfying to find the inscribed keystone of the west face of the arch partly embedded in the marble pavement opposite the keystone found in 2014. This new inscription records the dedication of a statue of Dionysos by someone named Aurelius M[---, a an overseer of something not yet readable, who paid for it from his own funds. With a central span of 13 m, a width of over 33 m, and a height estimated at about 29 m, this arch seems to be the largest in the Roman world, significantly larger than the arches of Constantine or Septimius Severus, or those at Gerasa or Anazarbos (fig. 23).

Another excavation related to this Touristic Enhancement Project was to clear a drain under the Synagogue. This proved to be larger than expected, 1.85 m wide and 2.35 m high, and paved with schist slabs which bear wheel ruts from a previous use (fig. 24). It does not connect with any of the waterworks in the Bath, but in its final stage it was fed from a smaller drain to the south. The openings of the tunnel were built with cut marble blocks and voussoirs, suggesting that it was originally a cryptoporticus opening to a street at a lower level than that preserved today, and only later turned into a drain.

The five-year project to remove disfiguring biofilm from the temple of Artemis continued with the cleaning of the south cella wall, bases of the north and south peristyle, and four standing columns of the east porch (figs. 25, 26). Work from tall scaffolding goes slowly, but the crew of a dozen women have also taken more and more responsibility for the work (figs. 29, 30, 31). The cleaning continues to reveal new aspects of the temple, for instance a Lydian inscription on one of the pedestalled columns, published by Butler but long hidden by lichen (figs. 32, 33).

The Touristic Enhancement Project aims to build protective shelter roofs over the Lydian Fortification and Synagogue, and connect and explain these sectors through visitor paths and signage. A preliminary design for the Synagogue roof was submitted to the Ministry of Culture and the Röleve, and that for the fortification is essentially complete. A model built from modern mudbrick allows us to test different methods of protecting the original mudbrick section of the Lydian fortification. We also continued our program of informational signage with eleven new and renewed informational signs on the Roman avenue, Byzantine Shops, and elsewhere (fig. 34).

In the conservation laboratory, work focused particularly on the assemblage of metal objects and coins found this year and in previous seasons. The team of seven conservators cleaned more than 300 coins and almost 200 other metal artifacts (some of which were very time-consuming, such as thousands of iron armor scales). They examined more than 5,000 metal artifacts in the depots, and began a new project to rehouse unstable artifacts in Escal with water and oxygen scavengers to slow deterioration. In addition, the team stabilized and cleaned the wall paintings and artifacts in the field.

Study of the Byzantine fortifications and soft conglomerate cliffs of the Acropolis was greatly aided by the use of drones, which have allowed us to document these features adequately for the first time (fig. 35).

Progress in publication includes the submission of monographs on the 8,000 coins found at Sardis since 1972, by Jane Evans; more than 450 inscriptions found since 1958, by Georg Petzl, and the temple of Artemis, by Fikret Yegül, and continuing work on sectors HoB and PC, by Andrew and Nancy Ramage and Gül Gürtekin-Demir, the Synagogue, by Andrew Seager, and other projects.

Gjerstad, E., J. Lindros, Sjöqvist, and A. Westholm. 1935. The Swedish Cyprus Expedition. Finds and Results of the Excavations in Cyprus 1927-1931. 2. Stockholm: Swedish Cyprus Expedition.

Greenewalt, C.H., jr. 2007. “Sardis: Archaeological Research and Conservation Projects in 2005.” Kazı Sonuçları Toplantısı 28: 743-756.

Mallowan, M.E.L. 1966. Nimrud and its Remains. New York: Dodds.

Schmidt, E.F. 1957. Persepolis II: Contents of the Treasury and Other Discoveries. Oriental Institute Publications 69. Chicago: Oriental Institute Press.

Verčík, M. 2016. “Nichts als Schrott? Nahöstliche Panzerschuppen aus dem Apollon- Heiligtum in Didyma.” Distant Worlds Journal 1: 11-26.

Young, R.S. 1956. “The Campaign of 1955 at Gordion.” American Journal of Archaeology 60: 249-266.

• 1For example, from Nimrud (Mallowan 1966, 2, 409ff); Persepolis (Schmidt 1957, 97-101); Gordion (Young 1956, 257); and many other sites; the example from Idalion is among the most completely preserved (Gjerstad et al. 1935, 536). A new find from Didyma might therefore be Lydian rather than Assyrian (Verčík 2016).
• 2In 2017 this was shown to belong to an earlier Hellenistic phase.
• 3Four swords were discovered in 2005 at the base of the south wall of room 1: Greenewalt 2007, 745.