Our fieldwork goals are derived from two larger, underlying questions of our work on the Halaf period in southeastern Turkey. The first question concerns degrees of mobility in Halaf society, the second the internal structure of Halaf households. From previous excavations of Transitional Halaf and Late Halaf levels at Kazane Höyük, just outside the modern city of Şanlıurfa, but also from interpretations of sites such as Khirbet esh-Shennef and the Burnt Village at Sabi Abyad, it has become apparent that Halaf communities were more mobile than previously imagined. Our research addresses mobility on two scales, the short-term, yearly round of semi-sedentary transhumance and the longer term, periodic shift of whole settlements. We suspect that Halaf groups may have been partly sedentary and partly nomadic, with some people remaining year round in their villages and others moving out of the settlements with herds during certain times of the year. We sampled intensively well-stratified contexts at Fistikli Höyük for plant and faunal material in order to be able to assess evidence for seasonal occupation.
Another scale at which mobility operates is the long term. Many Halaf sites have short stratigraphic sequences, indicating that settlements were abandoned and re-established frequently. There is no indication that this was caused by major disruptions due to raids, warfare, or forced movement resulting from shortages in natural resources. Rather, we think that this was part and parcel of Halaf cultural practices. One of our research objectives was therefore to date excavated strata as precisely as possible. Present knowledge about the relative chronology of Halaf ceramics and other artifacts is insufficient for this purpose. Instead, absolute dating of well-stratified contexts is necessary for the establishment of a reliable Halaf chronology. We therefore took a large number of radiocarbon samples from different contexts at Fistikli Höyük with the goal of assessing the occupation span of the site as precisely as possible. Accelerator mass spectrometry (AMS) radiocarbon dates have given very good results, supporting our presumption of a short occupation of Fistikli Höyük.
Our second major research goal was to assess the size and composition of households in a small Halafian community. Households are best conceived of as economic units, not as kin (family) or architectural (house or co-residential) units. We define households as units that aim at producing what they consume, but that very rarely reach that goal. From such a definition it follows (1) that households strive towards independence from each other, and (2) that households are defined by sets of productive and consumptive activities. In the Halaf case, we started with the assumption that in all households some basic productive and consumptive activities occurred, but that they were not necessarily carried out in all such units with the same intensity. Already at this preliminary stage of evaluation of our excavation, we realize that this initial model needs reformulation.
To the north of Fistikli Höyük, near Zeugma/Belkis, the Euphrates breaks through several major Miocene formations. Below Zeugma, a sharp southward bend of the river created the high eastern limestone cliff on which the Crusader castle of Birecik is located. Just south of Birecik, the river valley widens and river terraces are visible. The Euphrates changes its course frequently in a meandering pattern in the floodplain. Today, the main channel is in the western half of the floodplain, at some distance from Fistikli Höyük, but it is probable that the settlement was located much nearer to its banks in Halaf times. Excavations have shown that the site is located on a natural mound, so that the cultural deposits are much shallower than originally expected.
Today, peasants around Fistikli Höyük plant grain, cotton, and various vegetables such as peppers, cucumbers, tomatoes, eggplants, parsley and mint. Ruins of old bridges and mills in the area indicate that the river valley and lower terraces were agriculturally highly productive in Ottoman times. As one moves away from the well-watered soils of the Euphrates River plain up into the hills, soil depth decreases dramatically. On these barren landscapes, recently introduced tractor-drawn plows make the establishment of pistachio plantations possible. A consequent wide-ranging change in the landscape is part of a rapid establishment of agribusiness in southeastern Turkey. At the core of the development are hydraulic works such as the Carchemish and Karababa dams. The goal of diverting Euphrates water into otherwise dry plains has led to absentee landlordism, stark increases in cash crop cultivation and a concomitant decrease in locally consumed products, the impoverishment of a proletarianized peasantry and rural-urban migration on a massive scale. In the Carchemish dam area, such social changes are only beginning to be felt.
The Birecik region is at the intersection of three major linguistic groups: Turkish speakers in the west, Kurds in the east and north, and Arabs in the south. However, divisions occur most often along age groups and a city - countryside divide. Younger people tend to use Turkish more often, as do people who live in or near Birecik. In this area, the formation of distinct socioeconomic classes seems to underlie present social change.
In two seasons of fieldwork, we excavated an estimated 14% of the total area of the Halafian occupation at Fistikli Höyük. An overview of the exposed architecture, features and outside surfaces from the excavation units reveals a clear structure of the hamlet. The main part of the site is occupied by several, regularly spaced large tholoi and smaller cell-plan structures. There could have been other similar buildings further south, underneath the modern house on the southern part of the site. Nowhere are walls of a tholos superimposed over those of an earlier one. Even if not all tholoi were constructed at the same point in time, it is likely that they were built in a brief succession, and the latest had been erected before any of them had completely decayed. Each tholos, with the possible exception of Tholos II in Unit A, had an adjacent cell-plan building. This stands in contrast to the structure of the Burnt Village at Sabi Abyad, where the central part is dominated by large cell-plan buildings, and tholoi are irregularly placed in more peripheral areas of the site.
Open areas occupied a lot of space at Fistikli Höyük. This accords well with the findings from approximately contemporaneous Early Halaf Sabi Abyad. The peripheral parts of the site display variable uses. The western edge was a garbage dump on a steep slope towards the Euphrates. In the two northeastern Units I and K, two massive, lengthy earthworks were found. They do not seem to be part of the same building effort, and their function is not entirely clear. They may have served as protective devices against flooding from the Euphrates or the adjacent hills at the eastern edge of the river valley. It seems that the site was especially vulnerable to such damage on its northeastern side. Two small channels in different parts of the site with sequences of remodeled floors and fills of silt and sand may have had a related function: water removal.
The northern part of the site contained sequences of outside surfaces that were densely littered with many different kinds of artifacts. The objects found on such outside surfaces could partly be trash dumped at the edge of the site; however, we expect at least some of them to be debris related to work which had occurred there. According to a spatial analysis of finds from the Late Halaf site of Kazane, many household activities such as food consumption may have taken place in outside spaces. The end of the Halaf use of the site is marked by intensive dumping events, involving pit digging, in the eastern peripheral area of the site. At the center of the site, just below the present surface, a vessel, probably smashed intentionally, was found between the walls of a tholos and a cell-plan building.
For information on the nearby site of Kazane Höyük, please visit the website at http://bingweb.binghamton.edu/~rbernbec/kazane.html.