The processing of the newly excavated pottery assemblages is in full swing, and will be reported on in a later stage. Processing thousands of sherds a week involves washing most of them (except for those to be sampled for residue analysis), drawing them, reassembling them where possible, taking pictures, and finally studying them. The last process is carried out under the supervision of Jeroen Poblome by Philip Bes (KULeuven) and Senem Özdem and Yaprak Özkünü (both Istanbul University). Meanwhile, the process of critical evaluation of the final occupational/abandonment deposits in the urban mansion is reaching preliminary conclusions. The way most of these deposits originated is fairly complicated, if not problematic.
In many of the rooms, the permanent floors, possibly consisting of a combination of tile paving, stone slabs, opus sectile, mosaic or even wood panelling, was no longer present. This absence can be combined with the high variability and in general more than average amounts of residual material, i.e. material of late Hellenistic to Roman Imperial date, pre-dating the final occupational stages of the house, in the lower strata. This material was even identified in rooms where the permanent floors were still present. These results most probably imply that the permanent floors were removed at some stage(s) and the floor substrate deposits at least partially reworked. When, why, and how this was done are still open questions. Most, if not all of the residual material cannot be associated with the occupation of the house and came in as constructional backfill or levelling material. An important quantity pre-dates the earliest attested phase of construction (the first half of the second century A.D., based on the use of brick as a construction material). Another observation is that the major fourth-century A.D. (re)construction phase of the reception hall wing, the bathing complex and the southern dining suite is poorly represented in this material. The association between this type of material and the various documented building phases thus seems problematic.
As far as the occupation of the house is concerned, it is now clear that this continued into Phase 9 (A.D. 550/75-640/50) of the relative chronology of Sagalassos red slip ware, but that this phase is represented unevenly in the material record. Each room contains the odd piece, while only a restricted set of rooms contained fairly complete vessels. The best-preserved set of vessels was found in the cellar of Room XXVI. This cellar was backfilled, however, and this stratigraphy contained the remains of a purse with a set of coins, of which the most recent ones were struck in A.D. 574/5. Allowing for an unknown period of circulation of these coins, this probably means that the closing of the cellar happened fairly early in Phase 9, at the latest during the early seventh century A.D. Most probably the rest of the house was abandoned around the same period, considering the small quantities of Phase 9 pottery, which, as an assemblage, seems slightly earlier in sequence compared to, for instance, the assemblages excavated in other years in the Early Byzantine re-arranged North-East Building on the Upper Agora, or the West Portico of the Lower Agora.
In general, the small quantities of Phase 9 pottery can indicate that most rooms were kept tidy and clean of rubbish, or were cleaned thoroughly upon abandonment, before the roof and ceilings came down. The fact that a restricted set of rooms contained fairly complete pottery could imply that the final occupation was only organized in a nucleus of the house, or that waste was allowed to accumulate only in these rooms. Most of this pottery was only relatively complete, however, implying that these were the remains of the contents of the rooms upon abandonment. On the whole, we consider a restricted final phase of occupation as the most likely option.
Whether the permanent floors were removed before, during or after this final stage of occupation is not clear, but most probably we have to allow most of these processes happening more or less together. Fact is that most of this process was completed by the time the structure started to fail and parts of it to collapse. As the latter process was probably gradual, while the house was most probably abandoned before the rest of the town of Sagalassos, there must in any case have been sufficient time to strip the structure of whatever could be recycled and reused.
The best-represented phase in the pottery assemblages is actually Phase 8 (A.D. 450/75-550/75). The taphonomy of this material does not allow an association with occupation, however. The way the sherds are broken clearly attests material in secondary context, i.e. most probably related to the last reworking of the rooms, and mostly floors. In this way, the material of Phase 8 is comparable to the residual pottery. Considering the length of this Phase 8 and the amount of rooms, we most probably have to allow for a certain sequence of such structural accommodations, rather than one general phase of rebuilding. Obviously, the relative chronology of the structural changes made to the walls of the rooms should shed light on this aspect.
In general, the methodology of interpreting the various activity patterns in one set of layers is very interesting. The potential for thorough contextual analysis of the final, let alone the other phases of occupation of the house, based on the ceramic remains, is fairly low, however.