This was the last week of the suburban survey carried out by Hannelore Vanhaverbeke and her team. The westernmost reaches of the Düzen plateau were covered (some 2.5 hectares), as well as the fields located in the valley separating Düzen from a promontory to its west (1.5 hectares). The latter area yielded only few sherds, but a number of structures were observed, which are built in the same "primitive" style as those recorded at Düzen (no mortar, rough fieldstone socle, no tiles or brick).

We also made a half-day exploration of the promontory west of Düzen, and were rewarded with some exceptional discoveries: several "primitive" structures occupy this area as well, whereas the whole east side of the promontory facing the west side of Düzen, is surrounded by a wall built in the same technique and fashion as the western fortification wall at Düzen: large boulders were piled on top of each other, forming a massive wall on the crest of the promontory, with transverse walls running down slope. One stretch of wall was very well preserved up to a height of ca. 1.5 m and clearly shows the massive stones that were used in its construction. On the western side of the promontory, the remains of an ancient partly rock-cut path can be seen. In the same area, but higher on the bedrock face, our temsilci Orhan Bey discovered an intriguing inscription consisted of two large carved letters, which although using the Greek alphabet (EK), are not necessarily Greek (perhaps Pisidian?).

Looking back at our 2.5 weeks of survey, it is clear that the site at Düzen is of an exceptional nature. Not only is it absolutely not Roman (see also below), but its scale (ca. 20 ha at Düzen, and spreading into the valley and promontory to its west) and its massive defenses are baffling and without parallels in any other part of the territory of Sagalassos. The whole fortified area thus covers almost ca. 21.5 ha not including the fortified top of the Zencirlik Tepe (see Suburban Survey, July 24-28, 2005). Even if the fortified area was not as densely occupied as Hellenistic Sagalassos, the latter's fortifications only surrounded an area of 12.8 ha. Therefore, Tepe Düzen clearly is must be the Archaic predecessor of Sagalassos, which seems to have been occupied again after the latter site was abandoned in the seventh century A.D. As this region is of a karstic nature, meaning that its limestone subsurface is full of holes and tunnels, into which water sometimes disappears over dozens of kilometers, it cannot be excluded that during the initial occupation, the plateau had at least some natural springs, which later disappeared possibly because of seismic activity.

On Wednesday and Thursday, part of the survey team accompanied our geomorphologist Véronique De Laet on her exploration of the Basköy area, while the rest of the team spent two days classifying the pottery collected at Tepe Düzen. Eleven fabrics could be distinguished, most of which seem to be variants of a single large ware group. A fabric reference collection was established, and under the supervision of Jeroen Poblome, our chief ceramologist, we started to quantify all sherds (number of sherds and weight per fabric). Diagnostic pieces (rims, handles, bases, decorated body sherds) were sampled per fabric. These will be further studied for typo-chronological dating and functional categorization. A preliminary glance at these diagnostics indicates two "major" phases of occupation: a period between ca. eighth-fourth centuries B.C., and a probably important post-Roman component. The Hellenistic, Roman, and Early Byzantine period, well represented at Sagalassos, are here represented only by a limited number of sherds. Possibly the population at Düzen moved (because of water shortages?) to Sagalassos around the fourth century B.C., and seems to have returned at some later stage (from the twelfth/thirteenth century A.D. onwards or even later?). The presence of some cisterns may go back to that second occupation.

Left, a sherd, probably Geometric to Archaic, decorated with circles and a horseshoe-shaped motif near the upper edge. Right, a black-gloss sherd of Postclassical date. Bottom, an Early Iron Age sherd with incised leaf (or arrow head) decoration.