The Archaeology of Midas and the Phrygians: Recent Work At Gordion

7 A Decorated Roof at Gordion: What Tiles Are Revealing about the Phrygian Past

A half-century of digging at Gordion has brought to light a dazzling array of Phrygian architectural terracottas known to laymen as roof tiles. After pottery, clay tiles are probably the most abundant artifacts at the site, and their recovery and analysis has been a persistent theme in the exploration of Gordion. Brushing aside the dust of ages to reveal a tile molded in relief and still brilliantly painted can be exciting and rewarding. At once architectural and artistic, the tiles offer a glimpse of Phrygian building styles and decorative tastes and provide clues about the site’s history. Those who have dug at Gordion also know that the terracottas can be confusing. They are more likely to be found discarded in late trash dumps than they are associated with their original buildings. Fortunately, even scattered parts of a tiled roof share certain characteristics, and these allow some sense to be made of a large, disparate body of material. As an illustration of recent approaches to studying the Gordion tiles, this chapter brings together six previously unrelated types of tile that once formed a single, highly decorated roof system. RECOVERY AND PREVIOUS RESEARCH Clay roof tiles were probably invented in the Greek city of Corinth, where they were first used to cover the Temple of Apollo early in the 7th century BC. The new type of roof improved on traditional thatch and mud techniques by being both water- and fireproof. Tile technology soon spread to other Greek sanctuaries and had reached Sicily, south Italy and Etruria in the west, and Ionia in the east, by the last quarter of the 7th century BC. The Phrygians of central Anatolia were using tiles by the first half of the 6th century (Glendinning 1996a, 1996b; Le Roy 1967; Robinson 1984; Roebuck 1990; Wikander 1988, 1990; Winter 1993). Archaeologists first became familiar with ancient Phrygian tiles thanks to the work of Alfred and Gustav Körte. Excavating in 1900 along the southwest edge of Gordion’s Citadel Mound (see Fig. 3-1), the brothers recovered a few fragmentary tiles. Although their reconstruction of a tiled building (Fig. 7-1) is untenable—mistakenly combining tiles of different fabric, size and design—the illustration nevertheless conveys the exuberant decoration typical of Phrygian architecture (Körte and Körte 1904:153–70). The six types of tiles presented here come from the excavations of Rodney S. Young between 1950 and 1973. His teams recovered massive numbers 7 A DECORATED ROOF AT GORDION What Tiles are Revealing about the Phrygian Past MATT GLENDINNING 83 of terracottas, especially in early seasons along the southeast edge of the mound. Young’s assemblage included many well-preserved examples, and while repetitive pieces were often discarded, several thousand tiles are still preserved. The quantity of recurring types suggests that the tiles were massproduced for use on numerous buildings within the fortified Middle Phrygian settlement (Fig. 7-2). The date, design, and location of the tiled roofs have always been a matter of debate. In what is still a major reference work, for example, Swedish scholar Åke Åkerström (1966) dated some of the Gordion tiles to the late 6th century and others to the late 5th, while more recent work suggests a significantly earlier Figure 7-1 Körte and Körte (1904) tiled building reconstruction. A DECORATED ROOF AT GORDION THE ARCHAEOLOGY OF MIDAS AND THE PHRYGIANS 84 chronology (Isık 1991). Uncertainty surrounds the Gordion tiles due to the way they are often studied. Earlier research often treated the tiles as isolated works of art, an approach that allowed unrelated stylistic dates to be assigned to pieces that may once have belonged to the same building. A more holistic method is adopted here, in which the tiles are first identified as interrelated components of actual roofs, based on shared characteristics such as fabric, form, size, decoration, and archaeological context. Serving as a set of cross-referencing variables, these features provide the most reliable information about how the tiles were originally used. Current research thus approaches the terracottas with new questions. Which ones worked together to form actual roofs? What do the tiles’ stratigraphic contexts suggest about date of manufacture and use? How do tiles elsewhere in Anatolia compare to those at Gordion? And what do the tiles tell us about developments at Gordion? Figure 7-2 Plan of Middle Phrygian settlement, Gordion. 85 THE TILES...