Pre-Pottery Neolithic B Settlement Patterns along the Balikh and Euphrates – Fact or Fiction?

PRE-POTTERY NEOLITHIC B SETTLEMENT PATTERNS ALONG THE BALIKH AND THE EUPHRATES - FACT OR FICTION? Peter M.M.G. Akkermans (National Museum of Antiquities, Leiden) The upper Euphrates valley as well as the basin of the river Balikh, a small perennial tributary of the Euphrates, have been the focus of much archaeological research in the past decade. The early prehistory of both regions has been recently investigated in considerable detail at sites such as Jerf ai- Ahmar, Haloula and Dja'de on the Euphrates (see e.g. the various contributions in this volume), and Sabi Abyad I and Sabi Abyad H on the Balikh (e.g. Akkermans 1993; Akkermans, ed., 1996; Verhoeven 1997). Generally speaking, both regions are well comparable in terms of material culture and the general sequence of occupation. However, the distribution and the nature of Neolithic occupation give evidence of some remarkable and as yet unexplained differences, as will be shown in more detail below. These differences are felt not to be merely due to constraints in the physical landscape, annual fluctuations in water level (prohibiting permanent settlement), or the like. Indeed, the size of the Euphrates valley exceeds that of die Balikh many times: whereas the Euphrates valley is (in Syria) almost 700 km long and covers an area of thousands of square kilometres, the Balikh basin is (in Syria) only about 100 km long, covering ca. 400 to 500 square kilometres. However, the conditions for living in the broad plains are largely the same: both the Balikh and the Euphrates are located in a more or less similar arid to semi-arid environment, and patterns of annual rainfall, climatic conditions and landuse are comparable. Nowadays, the upper part of both valleys is suited for dry-farming, whereas the lower part requires irrigation. Sites like Mureybet, Abu Hureyra or the more recently excavated Jerf al-Ahmar show that sedentary occupation along the Euphrates may have begun as early as the late 10* millennium B.C. Subsequently there is evidence of a more or less continuous sequence of early Neolithic settlement (Pre- Pottery Neolithic B, henceforth PPNB) along the Euphrates: here we may refer to sites like Mureybet IVA and Dja'de for the early PPNB (ca. 8700-8200 B.C.), to Mureybet IVB, Abu Hureyra 2A and the base of Haloula for the middle PPNB (ca. 8200-7500 B.C.), and to Abu Hureyra 2B, Haloula and the lower levels at Bouqras for the late PPNB (ca. 7500-7000 B.C.; see e.g. Cauvin and Cauvin 1993; Cauvin 1994)1. When compared with the Euphrates valley, it seems that settlement in the Balikh region started at a very modest scale and at a much later date, i.e. in the late 9th millennium B.C. or middle PPNB (or, perhaps, early PPNB). So far, only one Balikh site can be attributed to this period, i.e. BS 397 in the lower part of the valley2. This site BS 397 covers an area of about 0.4 ha and recent small plunder holes indicate that occupation deposits of ash and bone are between 50-75 cm in depth. The site, characterised by short-tanged Byblos Points together with long, unretouched sickle blades resembling those found at e.g. Dja'de on the Euphrates (cf. Coqueugniot 1994), may basically represent a hunting camp overlooking 1. The dates are calibrated dates B.C., following Cauvin 1994. The final PPNB or late Neolithic is not included in the present discussion. 2. The site was found during recent survey work by Tony Wilkinson of the Oriental Institute. University of Chicago. 523 P.M.M.G. AKKERMANS an early course of the Balikh. The lack of evidence for any other site of this period does not mean that the presence of other groups in the valley is necessarily excluded; the small group occupying site BS 397 cannot have operated in a social or cultural vacuum, if it were only to avoid genetic deformatioa Perhaps the valley was mainly inhabited or intermittently visited by mobile groups of hunter-gatherers, but it may also be the case that additional small, one-phase settlements have been buried below alluvial sediments or underneath later occupation deposits at certain mounds. However, it can hardly be doubted that this earliest Neolithic occupation of the Balikh valley was very restricted. In contrast with the Euphrates valley, where we have evidence of a definite trend towards long-term sedentarism at particular places, associated with the construction of increasingly elaborate buildings and the development of agriculture and animal husbandry, the present data from the Balikh valley basically suggest a continuation of the earlier, Epi- Palaeolithic hunter-gatherer mode of subsistence and community organisation for a long period of time. To judge from the material culture found at the surface of BS 397, the Neolithic trends being in the process of development along the Euphrates certainly had an appeal to the small and probably largely mobile bands inhabiting the Balikh region, but it took some 1000-1500 years before the Neolithic way of life was fully implemented in this area. At present, both the surveys (cf. Akkermans 1993) and the excavations at sites like Tell Assouad, Tell Damishliyya and Tell Sabi Abyad II (Cauvin 1972; Akkermans 1986/87, 1988; Verhoeven 1997) suggest that structural changes in Neolithic society and its associated partem of settlement in the Balikh valley first occurred in the late PPNB, starting at about 7500 B.C. Most likely, these changes do not reflect population movements or some 'colonisation' but are part of a profound transformation in local prehistoric society from a mainly (semi-)nomadic mode of socio-economic organisation based on hunting and gathering during the middle PPNB to a largely sedentary and agricultural way of life in the late PPNB. A detailed account of late PPNB settlement in the Balikh valley has been given elsewhere (Akkermans 1993); here some of the main outcomes will be summarised. First, the number of sites in the Balikh region dramatically increases during the late PPNB: at least 24 sites can be dated to the second half of the 8"1 millennium B.C., the majority of which is found in the upper part of the valley (i.e. the area nowadays suited for dry-farming). Relatively few sites are found in the more arid southern half of the valley. Second, most late PPNB sites are small, varying in size between 0.2 and 1 ha, with only very few sites being larger. The main exception is Tell Mounbatah in the central part of the valley, covering an area of about 7 ha and probably representing a regional centre of production and distribution of goods and services to other sites nearby. Third, the various Balikh sites are distributed in a more or less linear pattern along the river and its tributaries (wadis), at close distances from each other. The easy availability of water in the otherwise dry valley must have been a main determinant in the choice of an area for settlement. Fourth, the sites tend to appear in small clusters of two or more. Partially this clustering of small sites into larger agglomerations is related to environmental circumstances in the valley (e.g. the presence of water or other basic resources) but socio-economic variables may have played a role as well (e.g. kin-related groups working closely together and occupying a specific area passed on from generation to generation). When considering each of these clusters as coherent socio-economic units interacting at all levels, it appears that over half of the late PPNB population lived in no more than four larger agglomerations (Akkermans 1993:194-96). The excavations at Tell Assouad, Tell Damishliyya and Tell Sabi Abyad II suggest that most of these late PPNB sites in the Balikh valley were permanently occupied for rather long periods of time, i.e. up to about 500 years. However, it should be taken into account that at least some sites present evidence for short-term occupation and perhaps served as seasonal hunter-gatherer stations or as look-outs for hunting. In addition, it appears that the southern part of the Balikh valley seems to have been deserted towards the end of the 8th millennium B.C. (the final PPNB settlements are all found in the northern half 124 Pre-Pottery Neolithic B Settlement Patterns of the valley). From an analytical point of view, it should also be kept in mind that the precise chronological position of one site vis-a-vis another site is usually difficult to establish. For example, the excavations at Sabi Abyad II have made it clear that at least three chronologically distinct assemblages or phases can be recognised on the basis of local developments in the lithic industry (Copeland, in press). However, these phases are hard to identify with certainty at other sites if one has to work merely with lithic stray finds from the surface of these other sites. Consequently, the present settlement pattern may not (or, more likely, does not) reflect a correct picture of occupation for the late PPNB period in its entirety; the actual density of settlement at a single point in time may have been much more restricted3. Even if we assume that all sites were simultaneously occupied, it appears that the population size of the Balikh valley as a whole in the later 8"1 millennium B.C. must have been very small, assumedly comprising an average of about 900-1000 persons only (cf. Akkermans 1993:170,186ff). Doubtlessly, the region's carrying capacity was far from reached (ibid.: 165). Now, when comparing the present picture of PPNB settlement distribution of the Balikh area with that of the Euphrates valley, it appears that occupation in the latter region is strikingly limited and far below expectation (fig. 1). In view of the size of the Euphrates valley, many times exceeding that of the Balikh region, one would actually expect dozens, if not hundreds, of PPNB sites in this area4. In fact, only some 15 PPNB sites (early, middle and late together) are known along the entire Syrian Euphrates so far. Most sites are found in what is now the Lake Assad area, whereas the valley south of Tabqa seems to contain a handful of sites only, located at considerable distances from each other. The upper Syrian Euphrates or Tishrin region has yielded only two PPNB settlements5, i.e. Haloula and Dja'de. Both sites give evidence of chronological differences: whereas both were occupied in the middle PPNB, only occupation at Haloula seems to have continued into the late PPNB. This Tishrin evidence is strikingly different from that of the upper Balikh region some 100 km further east: whereas we find only one, large late PPNB site (Haloula) along the upper Euphrates, up to 18 small and large, late PPNB sites are found along the upper Balikh! Apart from the Lake Assad area, PPNB occupation along the Euphrates gives the impression of consisting largely of rather isolated, autonomous 'islands' in a vast and empty area, with limited intersite relationships. If correct, this picture is wholly unlike that of the Balikh valley outlined above. What reasons may account for these obvious distinctions in settlement between the Balikh and the Euphrates? Before turning to this question, I first wish to stress that the paucity of prehistoric settlement along the Euphrates is not restricted to the PPNB era only but holds for the later prehistoric period as well. For example, only 10 early Pottery Neolithic (i.e. pre-Halaf) sites are known from the Euphrates, with (as in the PPNB) hardly any sites known south of Tabqa (fig. 2). The differences between the Balikh and the Euphrates are even more pronounced in the Halaf period: whereas almost 40 Halaf sites are found in the Balikh region, only 10 sites are known from the much larger Euphrates basin (fig. 3). These sites are found either along the upper or along the lower Euphrates, with, except for Tell Zaidan at the confluence of Balikh and Euphrates, no sites known from the region between Tabqa and Deir ez-Zor. 3. Several scenarios can he envisaged, such as a pattern of settlement evenly distributed in time and space; settlement shifting from one site to another; or periods of alternating high and low settlement density. 4. This holds even more if we take into account the general environmental overlap between both valleys and the similar basic conditions for settlement. 5. Or three, if we include Mullah Assad along the neighbouring Sajour (cf. the various contributions in Sanlaville, ed., 1985). 525 P.M.M.G. Akkermans Survey procedures Our present picture of prehistoric settlement along the Baiikh and the Euphrates is certainly partly a reflection of the intensity of archaeological survey work in the areas, and of the survey methods used and the survey aims. The Baiikh valley has been intensively surveyed in recent years. Following some brief work up to the late 1970s (cf. Copeland 1979; Akkermans 1993 and references therein), the extensive 1983 University of Amsterdam field reconnaissance has produced hundreds of sites of all periods, many of which were very small and hardly visible in the field (some previous knowledge on the location of the sites was obtained through the use of satellite images, aerial photographs, various kinds of maps and pedological reports; cf. Akkermans 1993). Additional intensive coverage on foot of most parts of the valley has yielded dozens more of (often very) small sites (e.g. Wilkinson in press b). At present, there can be little doubt that the Baiikh valley is one of the most thoroughly surveyed regions of northern Syria. The situation is different in the case of the Euphrates, where most regional field reconnaissances have been selective and, apart from being restricted to specific parts of the valley, usually limited to sites immediately visible and/or located within easily accessible lands (e.g. Van Loon 1967; Kohlmeyer 1984; Copeland 1985; Moore 1985)6. These surveys mainly emphasise the larger sites and pay little attention to the recovery of small sites. Others have already rightly pointed out that more work on smaller sites is required, if we are to understand the full range of socio-economic organisation of ancient society (Schwartz and Falconer 1994). Actually, more detailed survey work in the vicinity of Euphrates sites under excavation has repeatedly led to the discovery of (usually small) sites hitherto unknown. For example, the presence of several very small aceramic Neolithic sites each represented by scatters of flints is reported in the immediate vicinity of Abu Hureyra (Moore 1975:56). Likewise, the close investigation of the fields surrounding the large Bronze Age site of Tell Amarna gave evidence of a Halaf settlement (Tunca, this volume), and intensive survey around Sweyhat led to the discovery of several small sites of mainly Bronze Age date (Wilkinson in press c). These few examples already indicate that our present inventory of settlements along the Euphrates as a whole is incomplete and that, in particular, small sites are under- represented. Erosion and sedimentation The varying degree of survey intensity partly explains the differences in the pattern of prehistoric settlement along the Baiikh and Euphrates, but other variables may be at work as well. In the case of the Euphrates, various researchers working in the region have noted the restricted occurrence of prehistoric sites in the vast basin, and have usually attributed it to the erosional effects of the meandering river and the steady reworking of the alluvial plain (e.g. P.A. Akkermans el alii 1981:495; Wilkinson 1978, in press a). Certainly it cannot be denied that the meandering Euphrates has taken away parts of (or perhaps even entire) prehistoric sites in the course of time but one may wonder whether this is the only reason for the paucity of prehistoric settlement. Recently, Geyer and Besancon (1997) attributed the absence of prehistoric (i.e. pre-Halaf) sites within the actual flood plain of the lower Syrian Euphrates to the fact that this plain represented an unpredictable and hostile environment in early Holocene times, with regular turbulent floodings. As a consequence, Neolithic settlement avoided the valley bottom and instead concentrated on the higher situated Pleistocene terraces bordering the valley. Changing climatic conditions subsequently led to an alluvial formation (termed QO) gradually filling in the valley bottom and thus 6. Moore 1985:46 points out that "The survey was earned out using a Volkswagen minibus. Thus the survey was restricted to those areas of the valley which could be reached by a road or passable track". 526 Pre-Pottery Neolithic B Settlement Patterns expanding the opportunities for human settlement and land exploitation in Halaf and Ubaid times. Geyer and Besancon's hypothesis sounds attractive at first sight but does not solve our basic problem: where are the sites to provide us with a reasonably complete system of settlement, conforming to, for example, the Balikh region? The terraces bordering the valley, assumedly the main area of occupation, do not come up to our expectations concerning the density of local PPNB occupation. Are these sites all eroded or washed- away? Perhaps we should assume a less black-and-white picture and accept, next to the inhabitation of the terraces, the possible presence of some short-term, seasonal occupation at selective places within the valley floor (cf. Geyer and Besancon 1997:8); if so, the Euphrates pattern of PPNB settlement must have been considerably different from that of the Balikh, where PPNB occupation is mainly characterised by sites permanently occupied for rather long periods of time. However, when taking into account that we have at present little precise insight in the development of the Euphrates QO formation, the speed of alluvial sedimentation, and the topography of the original valley floor, we may also propose a different perspective: the Euphrates basin was originally littered with small and low PPNB sites, conform the Balikh valley, but these have been buried below the later, QO sediments. Wilkinson (in press c) already points out that in the flood plain near Sweyhat small sites may have been buried owing to the steady aggradation of fine sediments washed from the slopes of the plateau. Elsewhere, too, we have some evidence pointing in the direction of the burying of small sites below (alluvial) sediments. For example, a Halafian site has recently been found at Tell Masaikh near Terqa, immediately next to the river and buried to a depth of almost one metre below the present plain level (Maria Grazia Masetti-Rouault, pers. comm.). Another Halaf site has been found near Tell Amarna, deeply buried below alluvial sediments but partly exposed by a shallow wadi cutting through the site (Tunca, this volume). The excavations at neighbouring Jerablus Tahtani gave evidence of the burying of 4"1 millennium, Uruk occupation levels deeply below the present water table. Jerablus Tahtani presents data on enormous Euphrates discharges in the late 3rd millennium B.C. which seem to have led to devastating inundations not only in the region around Jerablus and Carchemish but in other parts of the river valley as well (Peltenburg el alii 1996:15, 21-23). Investigations near Tell Bi'a, at the confluence of the Balikh and Euphrates, showed that a surface dating from the late 3rd to early 2nd millennium B.C. was buried at a depth of about one metre below the present field level, whereas an earlier surface of as yet unknown date was found at a depth of 1,5 metre (Schirmer 1987:64). Similar evidence comes from the Balikh valley, where recent research pointed towards considerable post-early 6Ih millennium sedimention, varying from area to area, even on a micro-scale (cf. Akkermans 1993:144-45; Wilkinson 1996, in press b). If, for example, settlement at the PPNB site of Sabi Abyad II had been interrupted and confined to the lower levels only, this site would have been buried up to two metres below the modem field level, invisible to the present-day researcher. Likewise, the earliest levels of occupation hitherto exposed at Pottery Neolithic Sabi Abyad I are found some four metres below the present field level, whereas at nearby Tell Hammam et-Turkman the earliest Ubaid deposits occur at a depth of some three metres. If these profound effects of sedimentation upon the visibility of prehistoric settlement already hold for a relatively minor valley like the Balikh one, one may wonder what the effects are in a truly large basin such as that of the Euphrates. In this respect, the vast alluvium of southern Mesopotamia, covering innumerable sites, immediately comes to mind (cf. Adams 1981). Some concluding remarks So far, the argumentation has been based on the assumption that the presently known partem of prehistoric settlement along the Euphrates, with its very limited number of sites distributed in a dispersed manner, cannot be complete and, consequently, cannot reflect reality. In its turn, this assumption is derived from the prehistoric pattern of occupation recognised in the Balikh valley, thereby taking this Balikh P.M.M.G. akkermans evidence as a norm. But what evidence do we have that this norm holds some validity on a wider scale, i.e. beyond the Balikh valley proper, and can. indeed, be widely extrapolated? In order to answer this question, one would like to have more detailed information on PPNB settlement in northern Syria in general, but unfortunately the actual number of sites to work with has been very limited until now. Very few PPNB sites are known from the Khabur region so far (Nishiaki 1992; Hole 1994), and the same holds for example for the Qoueiq area north of Aleppo (see the various contributions in Matthers, ed., 1981) or the Menbij region (various contributions in Sanlaville, ed., 1985). The density of PPNB settlement in these regions seems to be even less than that of the Euphrates valley but it is stressed that some of our earlier remarks concerning survey procedures certainly hold for the regions just mentioned as well (for example, in the Qoueiq only those sites indicated on the French 1:50,000 maps were visited and sampled). Somewhat disappointingly, it seems that at present we can hardly go beyond the conclusion that PPNB society had a wide distribution in Syria. In the discussion above, the emphasis was largely on site number and site distribution but size and layout of Neolithic settlement and aspects of settlement hierarchy may also be of relevance in defining regional variability. In the case of the Balikh and the Euphrates, it seems at first sight that there are, indeed, some indications that PPNB settlement along the Euphrates vis-a-vis the Balikh valley was organised along widely divergent lines. A more detailed look, however, gives reasons for doubt in this respect. For example, the exceptional size of some settlements along the Euphrates, such as Haloula, covering an area of about 7 ha, or Abu Hureyra, with its 12 ha one of the largest PPNB settlements in the Near East, is striking but it is not a unique feature of the Euphrates region. Tell Mounbatah on the Balikh. too, may easily have had a size of at least 7 ha in the late 8* millennium B.C. Another feature immediately attracting our attention is the regular pattern of house construction, the consistency of the internal structure of the buildings through time, and the rather unified, clustered layout of the Neolithic settlements along the Euphrates (with Bouqras as the most outstanding example; cf. P.A. Akkermans et alii 1981, 1983), all indicating a careful planning and organisation of habitation. The monumental stone constructions uncovered at Haloula (Molist in Weiss 1994:105-106. and in this volume) point towards joint efforts and an organisation of labour far beyond the reach of an individual or a small group of people; undoubtedly, the local community as a whole must have been involved in construction works of this scale. In contrast, the available evidence from the Balikh valley points towards the presence of very small and rather haphazardly built villages or hamlets, very irregular in layout and lacking any planning (Akkermans 1993; Verhoeven 1997). In this respect, we may perhaps recognise a more general distinction in the nature of occupation, with residence along the Euphrates being characterised by a considerable aggregation requiring close cooperation at all levels between the persons involved, and with settlement in the Balikh being defined by dispersion and restricted socio-economic organisation and planning. However, it is more likely that this supposed division results from differences in the nature of the excavated sites: whereas excavation in the Balikh region has been restricted to small sites only7, the opposite holds for the Euphrates valley. Large sites like Bouqras, Abu Hureyra, Haloula and Mounbatah may represent regional centres acting as foci of production, exchange and social interaction. If so, settlement at these centres may have demanded (or may have led to) an internal architectural organisation and residence layout wholly different from that of small rural sites. The presence of more than one regional centre along the Euphrates may simply be a reflection of site hierarchy in relation to the vast extent of the Euphrates basin, with each centre controlling a specific part of the valley. At present, it seems that our set of data is too incomplete to draw any far-ranging conclusions supporting a division in Neolithic settlement patterning between Balikh and Euphrates; I strongly feel that 7. For example. Damishhyya and Sabi Abyad II, each covering an area of less than 0.5 ha 528 PRE-POTTERY NEOLfTHIC B SETTLEMENT PATTERNS the present differences are mainly a reflection of the present stale of research, not of reality in prehistory. We are in urgent need of additional, large-scale excavation but even more of intensive survey work and detailed geomorphological research if we wish to come to a better understanding of PPNB settlement and society in Syria. Bibliography Adams R. (McC.) 1981 Heartland of Cities - Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates, Chicago/London. Akkermans P.A. et alii 1981 "Stratigraphy, Architecture and Layout of Bouqras", in J. Cauvin, P. Sanlaville, eds., Prehistoire du Levant. Paris, pp. 485-501. 1983 "Bouqras Revisited: Preliminary Report on a Project in Eastern Syria", Proceedings of the Prehistoric Society 49:335-372. Akkermans P.M.M.G. 1986/7 "Excavations at Tell Damishliyya, a Neolithic Settlement in the Balikh Valley, Northern Syria". Annales Archeologiques Arabes Syriennes 36/37:40-66. 1988 "The Soundings at Tell Damishliyya". in M.N. van Loon, ed., Hammam et-Turkman, I, Istanbul, pp. 19-67. 1993 Villages in the Steppe - Later Neolithic Settlement and Subsistence in the Balikh Valley, Northern Syria, Ann Arbor. Akkermans P.M.M.G., ed. 1996 Tell Sabi Abyad - The Late Neolithic Settlement, Istanbul. Cauvin J. 1972 "Sondage a Tell Assouad (Djdzireh, Syrie", Annales Archeologiques Arabes Syriennes 36/37:85-88. 1994 Naissance des divinites, naissance de Vagriculture, Paris. Cauvin M.-C, Cauvin i. 1993 "La sequence neolithique PPNB au Levant nord", Paleorient 19:23-27. Copeland L. 1979 "Observations on the Prehistory of the Balikh Valley, Syria, during the 71" to 4lh Millennia B.C.", PaUorient 5:251-275. 1985 "The Survey of RCP 438 in 1979". in P. Sanlaville. ed.. Holocene Settlement in North Syria (BAR IS 238), Oxford, pp. 67-98. in press "The Flint and Obsidian Artefacts", in M. Verhoeven, P.M.M.G. Akkermans, eds., Excavations at Tell Sabi Abyad II - The Pre-Pottery Neolithic B Settlement. Istanbul. Coqueugniot E. 1994 "L'industrie lithique de Dja'de el Mughara et le dfibut du PPNB sur 1'Euphrate syrien (sondages 1991 et 1992)". in H.G. Gebel, S.K. Kozlowski. eds.. Neolithic Chipped Stone Industries of the Fertile Crescent, Berlin pp. 313-330. Geyer B., Besancon J. 1997 "Environnement et occupation du sol dans la vall6e de 1'Euphrate syrien durant le N6olithique et le Chalcolithique", Paleorient 22:5-15. 529 P.M.M.G. Akkermans Hole F. 1994 "Khabur Basin PPN and Early PN Industries", in H.G. Gebel, S.K. Kozlowski, eds., Neolithic Chipped Stone Industries of the Fertile Crescent, Berlin, pp. 331-347. Kohlmeyer K. 1984 "Euphrat-Survey - Die mit Mitteln der Gerda Henkel Stiftung durchgefiihrte archaologische Gelandebegehung im syrischen Euphrattal", Mitteilungen der Deutschen Orient-Gesellschafl 116:95-118. Manners J., ed. 1981 The River Qoueiq, Northern Syria, and its Catchment (BAR-IS 98), Oxford. Moore A.M.T. 1975 "The Excavation of Tell Abu Hureyra in Syria: A Preliminary Report". Proceedings of the Prehistoric Society 41:50-77. 1985 "The Archaeological Survey of 1977", in P. Sanlaville. ed., Holocene Settlement in North Syria (BAR-IS 238), Oxford, pp. 45-66. Nishiaki Y. 1992 "Preliminary Results of the Prehistoric Survey in the Khabur Basin: 1990-91 Seasons", Paleorient 18:97-102. Peltenburg E. et alii 1996 "Jerablus-Tahtani. Syria, 1995: Preliminary Report", Levant 28:1-25. Sanlaville P., ed. 1985 Holocene Settlement in North Syria (BAR-IS 238), Oxford. Schirmer W. 1987 "Landschaftsgeschichte um Tall Bi'a am syrischen Euphrat", Mitteilungen der Deutschen Orient-Gesellschaft 119:57-71. Schwartz G.M., Falconer S.E. 1994 Archaeological Views from the Countryside - Village Communities in Early Complex Societies, Washington/London. Van Loon M.N. 1967 The Tabqa Reservoir Survey, Damascus. Verhoeven M. 1997 "The 1996 Excavations at Tell Sabi Abyad II, a Later PPNB Site in the Balikh Valley, N-Syria", Neo-Lithics 1:1-3. Weiss H. 1994 "Archaeology in Syria", American Journal of Archaeology, 98:101-158. Wilkinson T.J. 1978 "Erosion and Sedimentation along the Euphrates Valley in Northern Syria", in W.E. Brice, ed.. The Environmental History of the Near and Middle East since the Last Ice Age, London, pp. 215-226. 1996 "Sabi Abyad: the Geoarchaeology of a Complex Landscape", in P.M.M.G. Akkermans, ed., Tell Sabi Abyad - The Late Neolithic Settlement, Istanbul, pp. 1-24. in press a "Holocene Valley Fills of Southern Turkey and Northwestern Syria - Recent Geoarchaeological Contributions", Quarternary Science Reviews. In press b "Water and Human Settlement in the Balikh Valley, Syria - Investigations from 1992- 1995", Journal of Field Archaeology. in press c Settlement and Landuse at Tell es-Sweyhat and in the Upper Lake Tabqa, Syria, Chicago. 530 Pre-Pottery Neolithic B Settlement Patterns Figure l> Distribution of Pre-Pottery Neolithic B settlements along the Balikh and the Euphrates. P.M.M.G. Akkermans Figure 2.- Distribution of Pottery Neolithic (pre-Halaf) settlements along the Balikh and the Euphrates. 532 pre-pottery neolithic b settlement patterns Figure 3.- Distribution of Halaf settlements along the Balikh and the Euphrates.