Along and Across the Nahr el-Quweiq: EB I-IVA Ceramic Horizons and Interregional Connections

The Quweiq and Matkh plains, in north inland Syria, were densely settled areas during the Early Bronze Age, crossed by the River Quweiq, flowing north-south from the plateau of Gaziantep to disappear into the Matkh paleolake. The northern sector of the Nahr el-Quweiq was surveyed in 1970s, providing information on the 3rd millennium BC occupation of the Aleppo’s hinterland. However, to date, only little evidence is available from archaeological excavations carried out in the area, hampering to crosscheck the ceramic periodisation derived from the Nahr el-Quweiq survey, which has therefore been used, or referred to, for general comparisons only. The article aims at revising the chronology and connections of the Aleppo region during the EB I-IVA period in the light of recent excavations and surveys carried out in neighbouring areas (the Ebla region, the Middle Euphrates, the Sajur and Jabbul plains), ultimately exploring the role of the Aleppo region as a ‘bridge area’, which provided access to important commercial routes towards the Euphrates Valley, the ‘Amuq plain and the area of Gaziantep. The north-south axis – linking the Ebla region with Aleppo – had to be particularly important during EB IVA, when ḫalab appears to belong to the Ebla kingdom and was the seat of the cult of the Storm God as we learn from cuneiform texts from Palace G archives.


INTRODUCTION
The basin of the Quweiq River extends for ca. 130 km in a north-south direction, and ca. 45 km east-west, originating in the plateau of Gaziantep and crossing the like-named plain from north to south (Fig. 1). The Quweiq-basin can be divided into an upper and lower sector. The upper one is comprised between the Nahr el-Quweiq catchment and Aleppo, within the triangle formed by the sites of Aleppo, Bab and Aazaz, while the lower sector corresponds to the area comprised between Aleppo and the Matkh depression. The Matkh plain, measuring ca. 30 km N-S and 15 km E-W (ca. 450 km2), is a vast and irregular marshy swamp once occupied by a lake, in which the perennial water of the Nahr el-Quweiq and of several seasonal wadis flowed (Cantelli et al. 2013). Nowadays, the lake has almost completely dried up, except for some residual swamps visible especially during the rainy fall season. Ancient environmental conditions were favourable for dry-farming agriculture, with annual precipitation attested around the values of ca. 250-300 mm (Mantellini et al. 2013: 163-164;Cantelli et al. 2013, fig. 17.6), allowing the cultivation of cereals (mostly barley and wheat), as well as of lentils, olive groves, and fruits to be carried out (Arnoldus-Huyzendveld 2013;Wachter-Sarkady 2013).
During the Early Bronze Age (henceforth EBA), the Quweiq basin was a fertile and densely settled area, with small to medium sites clustering along the river and wadis system and around the Matkh lake. Until the last decades, the majority of information regarding the 3 rd millennium BC occupation derived from surface explorations carried out in the 1970s in the Matkh plain and along the northern sector of the Nahr el-Quweiq. The only exception was represented by the excavation of a sounding at Tell Rifa'at in years 1956-1960 that yielded little information about the EB IV period.
Since 2006, renewed excavations at the site of Tell Tuqan, located along the south western fringes of the Matkh paleolake, provided new data about the EBA period, revealing a long uninterrupted sequence spanning the EB II/III and the EB IVB periods (Peyronel 2011;Baffi and Peyronel 2013;Vacca 2014;2020: 173-180). Although the earliest EB III levels have been investigated by means of a deep sounding carried out in the northern sector of the Lower Town, the evidence brought to light thus far is noteworthy since it fits well with the sequence excavated at Tell Mardikh/Ebla (Mardikh IIA), enabling a refined chronological sequence based on these two key-sites to be outlined (Vacca 2014;. Moreover, in the past decades excavations and surveys have been conducted in neighbouring areas, along the Middle Euphrates, the River Sajur and in the Jabbul plain, making new materials available to foster the discussion about cultural connections, also in the framework of international research projects (e.g., the Arcane Project, the Ebla Chora Project, the Land of Karkemish Project). Thus, based on fine-tuned chrono-typological discourses and in the light of new hypothesis it is possible to look once again at the materials from the Nahr el-Quweiq area to exploit its informative potential. This work will consider the EB I-IVA ceramic horizon of the Aleppo/Nahr el-Quweiq area within a regional context in order to explore its cultural connections with the Jabbul plain and the Rivers Euphrates and Sajur eastwards and northwards, and with the Matkh plain and the Ebla region southwards. A brief overview of past works in the Nahr el-Quweiq and Matkh plains will be addressed to evaluate the available dataset. Subsequently, the EB I-IV ceramic assemblage from the Quweiq plain will be discussed focusing, in particular, on ceramic commonalities with the regions mentioned above (Table 1).
Based on data discussed in this article, 1 it appears that during EB I-III -and especially in EB IVA -the Quweiq area was characterised by the overlap of diverse ceramic traditions, as demonstrated by the co-presence of typical productions of both the Ebla and the Euphrates areas (especially the Karkemish, Tishrin and Tabqa sec- tors), which converge in the Aleppo/Nahr el-Quweiq region. The 'bridging' character of the Quweiq plain was probably favoured by the frequent movement of goods and people, as also hinted by Palace G cuneiform texts mentioning markets, merchants and traded goods. It seems, indeed, that besides the west-east route linking the Ebla region with the Euphrates through the Khanaser corridor, two other, northernmost axes -that intersected at Aleppo -were important roads, connecting eastwards and northwards the Ebla kingdom with the Euphrates Valley and the Sajur/Gaziantep area, where other important polities were located (Fig. 1).

The Nahr el-Quweiq
The River Quweiq was investigated in years 1977-1979 by a team from the Institute of Archaeology, London University (Matthers 1978;Matthers ed. 1981). Intensive surveys carried out over an area of approximately 1300 km 2 , from the Nahr-el Quweiq catchments to the northern fringes of the Matkh depression, allowed identifying a total number of 88 sites dating from the Pre-Pottery Neolithic to the Mameluke period (Fig. 2). The survey also included an eastern area between the Nahr el-Quweiq and the Nahr edh-Dhahab (the latter river flowing into the Sabkhat al-Jabbul), close to the modern city of Al Bab; here further 8 sites were surveyed (Matthers 1981a: 18). The southernmost tells explored by the British archaeologists corresponded to Tell es-Is, located along the eastern bank of the Quweiq river, and Tell Hader, on the opposite side, both situated at the northern fringes of the Matkh, where the Nahr el-Quweiq disappears into the paleolake. Hence, the Matkh plain was not included in the survey since it was the objective of an intensive surface exploration carried out by the Italian Expedition to Syria (MAIS) in previous years, between 1970-1974(de Maigret 1978see below).
In the Quweiq River survey almost all the sites visited yielded evidence of Neolithic to Early Bronze Age occupation, a situation that seems favoured by the geology of the Quweiq-basin since most of the sites were not buried in the alluvium; conversely, the virgin soil was still visible along the valley (Dorrell 1981).
As for the prehistoric and Early Bronze Age phases (pre-EB IV) a total amount of 1000 sherds was selected and studied in order to assess the chronological sequence. J. Mellaart (1981) distinguished 8 main periods spanning from the Pre-pottery Neolithic to EB III, which he named Quweiq A-H (Table 1). The periodisation was based on parallels with the chronological sequences of extensively published sites, such as Hama, and those in the 'Amuq. The latter, in particular, was used as the main reference, and the prehistoric phases of the Quweiq settlements were paralleled with the 'Amuq sequence established by Braidwood and Braidwood (1960). Thus, for instance, Phase Quweiq F corresponds to 'Amuq F, Phase Quweiq G to 'Amuq G, and Phase Quweiq H to 'Amuq H, dated by Mellaart to EB I (4000-3250? 3300? BC), EB II (3300? 3250?-2900 BC) and EB III (2900-2500 BC) respectively (see Table 1 infra for a chronological discussion of the Quweiq phases).
During the first half of the 3 rd millennium BC (Phases G-H) the Quweiq plain is densely settled, with a total number of ca. 34 to 37 settlements (Fig. 2). In the following EB IV period, the number of sites rises to ca. 43, showing a continuous trend and a further expansion of human occupation throughout the plain. The chronology for this period, elaborated by Matthers (1981b), is mainly based on diagnostic types attested at Tell Mardikh/Ebla in the pottery horizon of Mardikh IIB, especially Caliciform Ware pottery, as well as at Hama J, 'Amuq I-J and at other sites along the Middle Euphrates River Valley.
However, with the exception of some types exclusively attested either in EB IVA or in EB IVB, for the majority of sites a neat chronological subdivision could not be made, based on published materials only, since the latter encompass several long-lasting vessel types (Matthers 1981b: figs 210-211). Nonetheless, the bulk of surface materials and, especially, those collected at the pottery dump relative to an EB IVA kiln from Tell Kadrich (Matthers 1981: 327-330, 342-345) convey an image of how the local EBA ceramic horizon should look like. Additional archaeometric analyses carried out on selected samples from the Kadrich furnace and from 23 sites in the region are provided as a complement to typological information (Riley 1981).
Overall, apart from the British survey of the Nahr el-Quweiq, the area around Aleppo and along the northern sketch of the river is part of a poorly studied ceramic province as far as the EBA period is concerned. In fact, the lack of large exposures of archaeological sites hampers the possibility of crosschecking the chronological sequence established for the Nahr el-Quweiq survey with stratified materials from secure contexts, and to better characterise the local ceramic horizon.
The only exception is represented by the sites of Tell Rifa'at and Tell Qaramel and by the shaft grave discovered at Ansari, in the southern periphery of Aleppo. While at the latter two sites only EB IVB levels have been investigated thus far, 2 Tell Rifa'at provided evidence of an earlier EB IVA occupation. The site, located ca. 35 km north of Aleppo, was briefly investigated by V. Seton Williams (Institute of Archaeology of London University) in two campaigns, in 1956and 1960(Seton Williams 19611967;Matthers 1981b: 327-341). It consists of a high tell -rising ca. 36 m above the surrounding plain -and probably extending over a surface of, at least, 5.8 ha, as its slopes are completely surrounded by the modern town. The excavation of a trial trench in the lower town of Tell Rifa'at (squares F I-II) allowed identifying EB IVA levels (Level IV) with child burials and badly preserved mudbrick architecture.
Recently, the northern Quweiq area has been the focus of a research project, headed by K. Kohlmeyer and J. Klinger, aimed at understanding the role of Aleppo as central place in a diachronic perspective. These studies have raised new interest about the region, although key archaeological information concerning the 3 rd millennium BC are still missing (Knitter et al. 2014;Del Fabbro 2012). In particular, data from the important ancient tell of Aleppo are virtually absent for the EBA, while spectacular remains for later Late Bronze and Iron Age periods are known (Gonnella et al. 2005;Kohlmeyer 2009;. The only evidence dating to EB IVA consist of spotted remains of an earlier structure below the later Storm-God temple and a cache of metal weapons associated with the earlier structure (Kohlmeyer 2016;. Interesting data concerning the role of Aleppo as crossroads have been extensively discussed in a recent article by R. Del Fabbro (2012), who reconstructs the importance of trade for the Aleppo region in a diachronic perspective (see infra).

The Matkh Plain
The Matkh plain was firstly explored in 1964 by the Italian Expedition at Tell Mardikh/Ebla in connection with excavations launched at the site (Liverani 1965). In the 1970s, a geoarchaeological survey was carried out by A. de Maigret, who documented a total number of 54 sites spread around the Matkh paleolake (Fig. 1). The systematic collection and study of diagnostic pottery allowed de Maigret to assess the chronological developments of a substantial number of tell-sites from the Late Chalcolithic (Phase I) to the Persian period (Phase VII), and to reconstruct the major phases of occupation of the Matkh region in relation to the fluctuation of the paleolake (de Maigret 1978;. Recently, works in the Matkh plain have been resumed by the Ebla Chora Project with the aim of characterizing land-use, environmental conditions, and settlement patterns during the EBA in the area defined as the chora of Ebla (Mantellini et al. 2013;Ascalone and D'Andrea 2013;Peyronel 2014;Vacca 2019;. 3 The results of these works allowed to finetune the regional chronology and to re-evaluate the dating of surface materials in the light of a broadened set of data deriving from long-term excavations at several sites, notably Tell Mardikh/Ebla, Tell Afis and Tell Tuqan. 4 During the EBA period the Matkh plain appears densely settled, with a total amount of 22 sites identified, 11 of which occupied from EB I-III. A distinct increase of settlement is observable in the following EB IVA (2550-2300 BC), when the number of sites almost doubled (from 11 to 19). Larger sites situated in strategic positions controlling access to the Matkh plain could have been Tell Tuqan (> 5 ha?) on the western side of the lake, Tell Dlamah (> 5 ha?) to the north-east of the Matkh depression, and Tell Berne (14 ha?), in the southern sector of the Nahr el-Quweiq, also surveyed by the British team.
The only excavated site in the Matkh plain is Tell Tuqan (Fig. 1, Table 1). Archaeological investigations, launched in 1978 with four short excavation campaigns and resumed since 2006 by the Italian Expedition of the University of Salento headed by F. Baffi, provided new evidence on earliest 3 rd millennium BC phases, as well as on the EB IVB period. In particular, the excavation of a deep sounding in Area P South (Lower Town North) revealed a long EB III architectural sequence, including a workshop for pottery manufacture and a later phase with large storage and crop-processing facilities, and a subsequent EB IVB occupation characterised by domestic buildings and storage structures. The EB IVA is not documented, with the exception of two infant burials dating to an ancient phase of the EB IVA, probably EB IVA1 or Initial EB IVA2. 5 Although the Matkh plain is part of the Nahr el-Quweiq drainage system from a geographic point of view, it has stronger link with the Ebla region, located only 15 km westwards, in terms of material culture and ceramic production throughout the EBA period. In fact, the ceramic horizons of the Idlib and Matkh plains are largely comparable and can be ascribed to the same northern inland Syrian ceramic region, which is characterised by a relatively homogeneous pottery assemblage in terms of vessel shapes, manufacturing techniques and decorative styles. A second degree of similarity can be recognised with the ceramic assemblage of the northern sector of the Quweiq river, suggesting interconnections with the latter area during EBA.

MAJOR FEATURES AND DIACHRONIC TRENDS IN POTTERY PRODUCTION
In the following analysis selected wares and morpho-functional types are discussed, 6 focusing on their chronotypological variation and geographical distribution in the Nahr el-Quweiq plain. Comparisons with neighbouring areas, such as the Ebla region, the Jabbul plain and the Euphrates River Valley, will be explored. In order to allow comparisons with the different regions to be made only widely shared diagnostic shapes and wares have been selected, thus leaving aside vessels types that can be considered exclusively local productions. The discussion follows a chronological order, starting from EB I to EB IVA, thus from the beginning to the third quarter of the 3 rd millennium BC. It must be noted that several EB IVA types discussed in this section are long-lasting shapes documented throughout EB IVA-B/ENL 4-5/EME 4-5 and thus the exact dating of surface materials cannot be pointed out. In order to have firm points, EB IV types selected for the discussion include at least one specimen coming from Tell Rifa'at and/or Tell Kadrich. The type's nomenclature follows that adopted in typological studies on the EBA period of norther-inland Syria and the Euphrates Valley.

Reserved Slip Ware
Reserved Slip Ware (RSW) is a typical pottery decoration of the Late Chalcolithic (LC 5, Early Reserved Slip Ware) and the beginning of EBA (EB I-II, Late Reserved Slip Ware), characterised by a painted light-coloured slip which is wiped-off in oblique radial lines, or sometimes in alternating oblique and horizontal lines, to expose the darker clay underneath. RSW is widespread in northern Syria and south-eastern Anatolia, particularly along the Syrian-Anatolian Euphrates (in the Sajur plain, 7 at Birecik-Karkemish, Tishrin, Karababa and Elaziğ). According to the new Arcane periodisation the RSW complex covers the first half of the 3 rd millennium BC (3100/2900-2700/2600 BC) and corresponds to Periods EME 1-2, EUE 1, EJZ 0-1 and ENL 1-2 (Jamieson 2014). This ware is still attested, albeit less frequently, in the second half of the 3 rd millennium BC, in a form with reserved horizontal bands (Horizontal Reserved Slip Ware;Mazzoni 2002: 75).
Along the Nahr el-Quweiq RSW is particularly abundant among surface materials assigned by Mellaart to Phase Quweiq G. It occurs at 19 sites (Fig. 3) and the most recurrent shapes are preservation jars -some of which identical to those documented in the Jabbul plain survey 8 -and shallow or carinated bowls (Mellaart 1981: figs 157-158). The latter type is particularly well-represented at Tell Berne, which yielded several exemplars of bowls manufactured with a fine ware and with spiral reserved slip decoration on the inner surface (Mellaart 1981: fig. 158: 873, 875, 877, 879). Similar decorations on bowls are attested at Tell Tuqan, 23 km further south from Tell Berne in the Matkh basin. A bowl from the pottery workshop of Phase 10 (EB III) is decorated with the reserved slip technique, with the yellowish-buff slip wiped off in horizontal lines on the outer surface below the rim, while on the inner side the paint is reserved in a spiral pattern (Vacca 2014: fig. 6:1). Reserved slip decoration on bowls is quite common from period EME 2b in the Euphrates Valley, while in earlier phases it appears restricted to closed vessels and applied especially on the jar's shoulder (Sconzo 2015: 94).
While in the Quweiq plain RSW seems quite widespread, following the trend documented along the Upper and Middle Euphrates, in the Matkh plain and in western inland Syria this production is less frequently attested, similarly to what has been observed for the Tabqa Dam area (Sconzo 2015: 94; see also Jamieson 2014: 98). In the 'Amuq plain, RSW occurs in small percentage in Phase G (3-8%) and H (1-6%) contexts (Welton 2020: 59). At Ebla findings of EB I-II beakers with reserved slip decoration in oblique radial bands and reserved slip jars come from pre-Palace G levels (Mazzoni 2002: pl. XXX: 11-12; Vacca 2020: pl. XII: 1-2). In later EB III phases RSW is still attested, although occurring in small percentage at both Ebla and Tuqan (ca. 1%; Vacca 2020: 97-98). RSW is also found as surface materials at Tell Suffane in the Jazr plain (Mazzoni 2006: 384, fig. 4: a-f) and in LC-EB I levels at Tell Afis (Mazzoni 2002: pl. XXX: 7-8, 13).

Painted Simple Ware
During the first half of the 3 rd millennium BC, different regional painted styles developed across Northern Mesopotamia and the Levant. In the Upper Turkish Euphrates, several types of painted ware are attested, such as the Elaziğ Ware of Keban and Karakaya regions, the Gelinciktepe Painted Ware of the Malatya region, and the Karababa Painted Ware (Rova 2014: 11-16). Along the Middle Euphrates, from the Tabqa Dam area up to Qara Quzaq, another typical production is documented -the Euphrates Monochrome Painted Ware -characterised by a red dark or purple painted geometric decoration applied by means of a brush-like tool and reminiscent of the western Multiple-Brush Painted Ware (Sconzo 2015: 95, EME 2;Russo et al. 2018). In western Syria peculiar local productions are documented in the Ebla region during EB II-III (local Painted Simple Ware, PSW) and in the 'Amuq Plain (Multiple-Brush Painted Ware; Braidwood and Braidwood 1960). The production of the Ebla region (documented at Tuqan, Ebla, Tell Mastuma) is characterised by an opaque black or red-brown painting with geometric motifs (lines, zigzags, latticework, undulated or horizontal lines) resembling the Multiple-Brush style of the 'Amuq and the PSW of the Orontes Valley (Vacca 2020: 98-99).
Along the Nahr el-Quweiq, only few wall fragments of painted sherds were collected among surface materials and classified by Mellaart (1981: 186) as Multiple-Brush Painted Ware. These are characterised by a matt brown painted decoration consisting of wavy horizontal lines. Similar undulating motifs occur also on jars from Umm el-Marra VI (Schwartz et al. 2003: fig. 3: 1, 3-4). The findings from the latter two areas, considered by the authors as reminiscent of Multiple-Brush Painted Ware of the 'Amuq, can be understood as local productions stylistically comparable with the painted tradition of the Matkh and Idlib plains, where jars with wavy lines are quite common, besides other stylised and simplified patterns occurring on both closed and open shapes.

Platters and Platter-bowls
A characteristic production of the EB II-III period in the Levantine area (coastal and inner Syria, Lebanon, and southern Levant) is represented by plain or slipped and burnished large shallow platter-bowls or platters, generally made of red-orange or dark-red clays, displaying a variety of inwardly protruding to upright rims (Vacca, D'Andrea 2020 with relevant bibliography).
In the survey along the Quweiq River several specimens of platter-bowls were collected from a consistent number of sites and assigned to Phase H (Fig. 4). These vessels are manufactured with buff, red or orange clays and can be either plain or slipped and burnished (Phase H;Mellaart 1981: 186-187, figs 164-167). A similar surface treatment is applied to jars dated to the same period and characterised by a thin red slip then burnished (8 exemplars; Mellaart 1981: 188, nos 973-980).
In the Ebla region, several platters and platter-bowls are documented in stratified contexts dating to EB III from Ebla, Tell Tuqan and Tell Mastuma (Fig. 7:2,5,7). Similarly to the Quweiq river exemplars, the specimens from the Ebla region can be manufactured either with buff clays or with red-orange pastes (Fig. 7: 1,3,6). The surface is plain, or it is covered with a white or red slip and burnished horizontally. The latter surface treatment is rare, and it occurs almost exclusively on platter-bowls accounting to ca. 1% of the EB III assemblage of both Ebla and Tell Tuqan (Vacca 2020: 97, fig. 3.7).
Platters and platter-bowls appear to be a production exclusively documented in the Ebla region and in the Quweiq and Jabbul plains, being instead virtually absent in the early 3 rd millennium BC ceramic assemblage of the Middle Euphrates. In the latter area, red slip and burnished wares are very rare and mainly confined to the Turkish sector of the Euphrates, occurring on different vessels categories (high-stemmed bowls; stemmed carinated bowls; small jars with four pierced lugs; Sconzo 2015: 95; EME 2). Moreover, few exemplars of hand-made bowls with inverted bent rims documented in the Karkemish and Tabqa areas are restricted to the very beginning of EBA (EME 1), without any further development in the following period (Sconzo 2015: 113). Thus, based on published data the Jabbul plain together with the Quweiq river basin, appear to be the easternmost areas where platter-bowls are documented thus far.

Pots with Triangular Lugs
A long-lived type very common along the Euphrates River Valley, especially in the Karababa and Karkemish sector, from period EME 3 to EME 5 is represented by Cooking Ware globular or ovoid-shaped pots with triangular lugs protruding from the rim (Figs 5, 6: 8-13). The fabric is generally coarse and quartz-tempered, while the outer surface of the pots is usually burnished (Cooper 2006: 15, fig. 1.4:1;Sconzo 2015: 125, type 75, here Fig. 7: 12). This morphological type, close to the eastern EJZ 3a tradition, extends further west from the Euphrates Valley into the Quweiq river basin and the Jabbul plain starting from EB III, while it seems completely absent in the Matkh plain and, more generally, in the Ebla region. 9 Later on, during EB IVA, some reminiscent forms of lugged pots occur in western contexts, such as Phase I of the 'Amuq (EB IVA, Braidwood, Braidwood 1960, figs 308: 3-6, 309: 2) and Building P4 at Ebla (Marchetti 2013: fig. 7.36: 48). In the Quweiq river survey CW pots with triangular lugs have been documented at 19 sites and dated to both Phases G and H (Fig. 5). Mellaart (1981: 186-187, nos 890-955) recognises two varieties: one earlier type (Type A, Quweiq G), hand-made and manufactured with buff wares with the outer surface burnished, and a later type (Type B, Quweiq H), wheel-made and manufactured with a grey or brown paste, with the outer surface left coarse or poorly burnished.

Sugar-loaf Beakers and Ovoid Corrugated Goblets
From around the mid-3 rd millennium BC (EME 3, ENL 3), a trend toward specialisation of the ceramic repertoire, with the introduction of new morphological types for serving and consuming liquids, is documented over a large area spanning the Levant, Anatolia, and Mesopotamia. 10 Different local traditions of drinking sets develop approximately at the same time with distinct morphological types reflecting different drinking behaviours and socio-cultural preferences.
In the Quweiq basin, in addition to the goblets found at Tell Rifa'at, several exemplars were retrieved as surface materials at 32 sites across the plain, as well as in the kiln dump at Tell Kadrich (Matthers 1981b: 329, 347, figs 208, 210; here Fig. 8: 3).

Bowls with inturned moulded rim and bowls with ribbed band
Two types of bowls have been selected by Matthers (1981: 329) as diagnostic elements of EB IVA among surface materials collected during the Quweiq river survey. The first type encompasses 'bowls with a crescentic rim and a cordon below the rim on the outside' (Matthers 1981b: 329;Fig. 8: 8-10). These bowls were found in good number in the kiln dump at Tell Kadrich, in stratified contexts at Tell Rifa'at, as well as among surface materials being attested at 25 sites across the plain (Matthers 1981b: figs 206: 29, 208: 10-14, 210). Matthers (1981: 329) compares the type with similar bowls from Palace G at Ebla. It seems reasonable to assume a western connection for this type; in fact, similar vessels are widely attested in the Ebla region and in stratified contexts at Tell Mardikh (Palace G and Building P4) with different varieties, from shallow to deep bowls with vertical or curving sides and triangular to everted ledge rims, in some cases with tripod bases (Fig. 8: 11-12;Mazzoni 1982: figs XXVII: 3, 5, XXXI: 5, 24;Marchetti 2013: figs 7.30: 15, 7.35: 21). Although the comparison with the western types is the most likely, especially for bowls with ribbed bands from Tell Kadrich and Tell Rifa'at, a similarity of some bowls with ribbed bands collected during the survey with that of EME 3-4 fruit-stands cannot completely excluded, based also on the occurrence of some stems and trumpet-like bases among surface materials (see Mellaart 1981: fig. 169: 964, 967-970). However, bowls on fruit-stand are generally carinated and have band rims.
The second type is a Simple Ware 'small bowl with upright rim' (Matthers 1981b: figs 208: 6-7, 210), which was found at 26 sites across the Quweiq plain, as well as among discarded materials in the kiln dump at Tell Kadrich ( Fig. 8: 13-15). The author compares the bowl type with specimens retrieved in Palace G at Ebla (Mardikh IIB1, Matthiae 1980: fig. 16: 3 rd row, fig. 17: top row). However, according to published drawings, the bowls from the Quweiq survey are rather comparable with similar types from the Euphrates Valley, which are characterised by a more or less inturned and modelled rim, marked by an exterior groove (Fig. 8: 13-18). This kind of bowls, produced in Simple or Euphrates Banded Ware, is documented during period EME 4-5, and distributed in the Karkemish, Tishrin and Tabqa sectors (Sconzo 2015: type 82, Pl. 15: 22-23).

Jars with Ovoid Body and Swollen Rim
At Tell Mardikh/Ebla, jars with ovoid body, swollen rim, and tripod, pointed or rounded bottom are welldocumented in destruction levels of the EB IVA city. Fragmentary and complete exemplars have been found in Palace G (Mazzoni 2013: fig. 5.25, 5.32) and in Building P4 (Marchetti 2013: fig. 7.33:40). Specimens from Palace G were mainly retrieved in a storeroom located to the back of the Court of Audience (L.2617), where 19 jars were found empty and stacked, on three horizontal rows, against the wall in the back of the room . These large containers are characterised by a highly fired hard fabric, with pinkish brown or red colour pastes, and by a fine matrix with quartz, gehlenite, and calcite aggregates (D'Andrea, Vacca 2013). They were manufactured through wheel-coiling technique, with highly smoothed outer surfaces; some of these jars also bear cylinder seals and potter's marks on the outer surfaces of the vessels (Mazzoni 1992). Their capacity clusters around 40-50 litres, with few larger exemplars containing ca. 100 litres (D'Andrea, Vacca 2013: fig. 6.12). The manufacturing technique, the hard metallic-fired fabric and the limited number of exemplars found in sealed contexts at Ebla indicate a specialised function of these vessels, probably to contain prized liquids, such as wine (D'Andrea, Vacca 2013). S. Mazzoni (1992; suggested a local manufacture of ovoid and tripod jars in the area of Ebla and a distribution of these containers towards a northern circuit including the Quweiq and the Euphrates areas. Jars with thickened rim are documented at different sites in the Quweiq river survey and from stratified context excavated at Tell Rifa'at, although no one is complete (Matthers 1981b: fig. 211, hole-mouth jar types 2, 4; here Fig. 9: 1). These jars are made of hard light brown ware, with pale grey core and white grits among the inclusions; diameters range from ca. 18 cm to ca. 36 cm. One jar from Tell Chair bears an impression of a cylinder seal applied below the rim with a herring-bone or vegetal motif (Collon 1981: fig. 259; Fig. 9: 2).
Further north jars with swollen rim occur along the Sajur river at Tilbeshar Höyuk (Kepinski 2007: 156, fig. 10.5), while in the Middle and Upper Euphrates this type is virtually absent. The exemplar from Tilbeshar, characterised by a cordon or ribbed band at the junction between the neck and the body, is comparable with a complete specimen from Ebla ( Fig. 9: 3, 5).

Necked jars with moulded rim
Together with the previous jar type, 'small fine jars with thickened rim' have been considered diagnostic shapes to recognise an EB IV occupation at several sites along the Quweiq river (Matthers 1981b: fig. 211, types 1-3). At least two different types of jars are comprised within this category, including jars with vertical rim and an inner concavity (type 1) and necked jars with moulded rim marked by a groove and in some cases characterised by an inner step (types 2-3). All these different types were found in stratified layers at Tell Rifa'at (Matthers 1981b: figs 204: 15, 205: 31, 206: 9-10, 13), as well in the kiln dump at Tell Kadrich (Matthers 1981b: fig. 208: 27-28, 30-32), and can be compared with similar vessels from the Euphrates are. Overall, during EB IV a limited set of similar rim types occur on different kind of vessels, from small and medium size jars to spouted vessels and chalices ( Fig. 10: 1-12).
The first type, although represented only by fragmentary exemplars, is similar to small low-footed chalices from Gre Virike Period II, as well as to contemporary spouted vessels from Tell Hadidi (Fig. 10: 1-3).
The second type, encompassing necked jars with moulded rim, is the most frequent, occurring at 20 sites in the Quweiq plain, and finds comparisons with similar types produced in the Middle Euphrates area (Fig. 10: 4-7), between the Karkemish and Tabqa sectors during period EME 4-5 (Cooper 2006: 18, fig. 1.6: a-b; Sconzo 2015: type 103). Thus, the Quweiq area together with the Jabbul plain, where the latter type occurs at Umm el-Marra (Schwartz et al. 2006: fig. 21: 6-7), represent the western area of diffusion of necked jars with moulded rim.

Early Bronze I-III
As argued above, the pottery assemblage of the first half of the 3 rd millennium BC is documented only by surveys carried out in the northern sector of the Nahr el-Quweiq basin, while stratified sequences that can be used as chronological reference are documented in nearby areas all around the Quweiq plain (e.g., Tell Tuqan, Tell Mardikh/Ebla, Umm el-Marra, Oylum and Tilbeshar Höyük, as well as at different sites excavated along the Euphrates river). Based on these data surface materials from the Quweiq survey can be framed within a broader context and the chronological attribution of each phase could be slightly revised, while the overall considerations on single period assemblages can be maintained. For instance, Phase Quweiq F can be assigned to a Late Chalcolithic horizon (LC 1-3/4?) as suggested by the occurrence of Coba bowls, deep cups with rounded profile, everted rim jars, and internal-hollowed rim jars, as well as by the ubiquitous presence of Chaff-Faced Ware. 12 Phases Quweiq G and H assemblages can be instead assigned to the late LC-first half of the 3 rd millennium BC. According to Mellaart, Phase G is characterised by Reserved Slip Ware, Multiple-Brush Painted Ware, and pots with triangular lugs (Mellaart 1981: 154-157) and can be tentatively dated to the LC-EB I/II or EME 1-2 with respect to the Middle Euphrates sequence (see Sconzo 2015: 91-92 for the beginning of period EME 1). However, most of the wares and vessel types of Phase G is said to continue also in Phase H (Mellaart 1981: 158-159), possibly dating to EB III/EME 3. These include pots with triangular lugs (more properly assigned to this phase rather than Phase G), Reserved Slip and Multiple-Brush Painted Wares, while other ware categories, such as Brittle Orange and Red Burnished Wares, and vessel types, such as platter-bowls, are considered typical of Phase H.
Looking at the distribution of the discussed wares and types it appears that during the first half of the 3 rd millennium BC some major trends, characterising also the following period, start to emerge. The close formal affinities of the pottery from the surveyed sites allow to define the Quweiq region as a regional ceramic area, characterised by the occurrence of local traits (e.g., a fair number of orange and red burnished vessels) and other features that it shares, instead, with the nearby Jabbul and Sajur areas, and more broadly with neighbouring regions of Ebla, the 'Amuq and the Middle Euphrates (e.g., Reserved Slip and Painted Simple Wares).
Besides this, it is possible to also notice the occurrence of some features that characterise the Quweiq plain as a 'buffer' area where different traditions converge. This aspect is recognizable in the spatial patterning of some types, such as in the case of platter-bowls and pots with triangular lugs, the former linked to a western Levantine production and the latter related to an eastern tradition of the Euphrates and Jazirah areas (Figs 4-5). Based on published evidence, each of these types is only sporadically documented beyond the River Quweiq basin. Direct and frequent connections with the Sajur plain and the Karkemish/Karababa sectors are indicated by the high incidence of RSW and by the presence of fruit-stands, 13 pots with triangular lugs (Fig. 7), and trays with triangular lugs and notch decoration, which are very similar to vessels from Gre Virike, Period II (Engin 2007: fig. 18.8.10-14). 14 Likewise, significant connections can be traced with the Matkh plain, noticeable in the occurrence of reserved slip bowls at Tell Berne and Tell Tuqan, and in the presence of plain and burnished platter and platter-bowls (Fig. 7).

Early Bronze IVA
With respect to the second half of the 3 rd millennium BC, besides survey data, additional information is available and come from a small sounding carried out at Tell Rifa'at and from the fortuitous discovery of a kiln dump at Tell Kadrich. Overall, the EB IVA pottery repertoire shows a trend towards homogenization in terms of manufacturing techniques and wares, with finer pastes fired at high temperatures (McGrath, Grabrovaz 1981).
During EB IVA ties with neighbouring areas are even more evident than before. From this period onward, the Quweiq region appears highly integrated in a large network of contacts in all directions, and especially towards the south-west with the Ebla area and towards the east with the Middle Euphrates, with which the Quweiq region share similar modes of ceramic production and consumption, as well as common cultural and culinary preferences, reflected in the use of comparable cooking and table ware ceramic assemblages (Fig. 8).
According to S. Mazzoni (1985: 10), during EB IV the Quweiq basin forms part of the north-central 'caliciform' culture -extending from Ebla to the Jabbul and along the whole course of River Quweiq -characterised by the prevalence of pedestal and tripod pots, multiple-grooved rim bowls, spouted jars, and teapots. Similarly, C. Kepinski (2007: 155), discussing the evidence from Tilbeshar Höyük, maintains that during the second half of the 3 rd millennium BC a same ceramic horizon is recognizable along the Euphrates sector -from Lidar to Emar -and in the Sajur, Nahr el-Quweiq and Jabbul plains. Differently, C. Marro (2007: 229), analysing the pottery assemblage of Oylum Höyük considers the site as part of the same ceramic province as Tilbeshar and Gaziantep (at least in EB III-IV), which is different from that of the Quweiq area, the latter showing stronger links with Ebla.
Looking from the perspective of materials collected in the seventies in this region, and analysed in this article, it seems that the remarks put forward by the different authors hold true for the Quweiq valley, where we can find the presence of two different, but mutually influenced cultural traditions. The Quweiq area is, in fact, characterised by strong interconnections with both the Ebla and the Euphrates areas -the same trend has been also noticed for the later MBA period (Nigro 1998) -as testified to by the convergence of peculiar regional productions in the Quweiq plain. This is the case, for instance, of bowls with inturned and moulded rim, typical of period EME 4-5 along the Euphrates, which are quite widespread also in the Quweiq plain (but not documented further south in the Matkh and Idlib plains), or of bowls with ribbed band, a diagnostic type of EB IVA2 in the Ebla region that is quite common also in the Quweiq basin (but not documented further east in the Euphrates area) . The contemporaneous local production of vessel typologies of both 'eastern' and 'western' tradition is further supported by the discovery of a kiln dump at Tell Kadrich. Among overfired, and in some cases melted, sherds -mixed with collapsed pieces pertaining to the kiln structure -we can find corrugated goblets and bowls with ribbed band, as well as necked jars and bowls with moulded rims (Fig. 8). The same association can be seen in EB IVA burials and stratified levels excavated at Tell Rifa'at. At the latter site we can also find ovoid goblets together with 'sugar-loaf' beakers ( Fig. 8: 1, 5). It thus seems that the Quweiq valley was located at the intersection of both 'eastern' and 'western' ceramic industries.
Due to its geographic position at the crossroads between different areas, the Quweiq plain probably played an important role also in the diffusion of some ceramic types and stylistic features between the Euphrates and inland Syria area by means of imports or stylistic and technological transfer.
Conversely, vessels types that can be ascribed to a local Middle Euphrates tradition that spread in the western area encompass, for instance, tripod bowls and the so-called Syrian bottles. Along the Euphrates river, tripod hemispherical bowls are documented since Period EME 3 in different domestic and funerary contexts (Sconzo 2015: types 53, 85) and appear in the western inland Syria ceramic assemblage slightly later during EB IV, as documented by the findings from Palace G and Building P4 at Ebla (Mazzoni 2013: fig. 5.15;Marchetti 2013: fig. 7.26: 8; here Fig. 8: 12).
Finally, spouted jars, teapots, and bowls with beaded rim, as well as grooved-rim jars, are common types shared by the Middle Euphrates and northern inland Syria apparently developing at the same time in the local ceramic tradition of both regions (Fig. 10: 3, 12).
However, it should also be noted that some ceramic productions remain confined in the different geographical areas and are not documented, at least based on published evidence, in the Quweiq plain. This is the case of Euphrates Banded Ware, a typical production of the Euphrates area during EME 3-4 (Falb et al. 2014;Sconzo 2015: 99-100, 104-105), that is virtually absent along the Quweiq river (while it is quite common in the Jabbul plain) 15 ; the presence of sporadic findings of this ware in western Syria attests for imports and long-distance connections between the two areas. Imported Euphrates Banded Ware vessels are thus far documented in stratified context at Ebla in pre-palace phases (Building G5, EB IVA1) and at Hama in houses of Phase J8, dating to the same period (Vacca 2015;2020: 94-95, 273-274, fig. 6.15: 1). Similarly, Painted Simple Ware jars and deep bowls (e.g., Mazzoni 1982: fig. XXVIII: 3-8;Marchetti 2013: fig. 7.22: 20, 23, 7.35: 33, 36), characteristic of western inland Syria and the Ebla region, are virtually absent along the Euphrates river, as well as in the Quweiq plain, while some specimens can be found at Umm el-Marra, together with painted trefoil-mouthed jars (Schwartz 2016: figs 9:4, 17, 18:2). In the latter two cases, the spatial patterning of types highlights a west-east connection between the Ebla region and the Jabbul plain, which was equally important as the north-south axis in the EBA network (see Steinkeller 2021).

Across and Beyond 3 rd Millennium BC Borders
Overall, while the geographic distribution of types and styles highlights patterns of connectivity between different areas, it is however difficult to determine whether these patterns might be associated with relevant socio-economic activities, political interests, and more broadly historical events. In this respect, information gathered from textual evidence, especially from the Ebla archives, can shed light on the late EB IVA socio-political and socio-economic context of the northern Levant. However, several interpretative constrains in reconstructing the geographic horizon of the Ebla texts derive from the uncertainties in identifying ancient toponyms with archaeological sites (Bonechi 1993;Biga 2015).
For instance, any of the tells of the Quweiq plain has been so far identified; however, some proposals have been advanced, such as the possible identification of Tell Tuqan or Tell Hader with the city of NIrar, an allied kingdom near to Ebla, frequently quoted in the texts and located in close proximity to a water source tentatively identified with the Matkh lake ('at the waters of Mašat'; Biga 2008; Biga, Karbotly 2020).
The only site along the River Quweiq that has been identified to date with an ancient toponym is Aleppo (Halab x (LAM) ki , Ḫalab), quoted in the Mari and Ebla texts (Lambert 1990). The site is mainly mentioned in the texts with reference to the cult of the Storm God Hadda, who regularly received offerings and official visits of the ruling family of Ebla during annual celebrations (Bonechi 1990: 31-33;Archi 2010;Biga 2013: 264;. The city is also the seat of fairs (KI:LAM 7 ), where precious timber 16 among other commodities was sold (Catagnoti 2016: 32;Biga 2003). Aleppo was not a capital of a kingdom during the EB IVA, as also inferred by some documents mentioning an overseer, ugula, who administrated the cultic centre on behalf of Ebla (Archi 2010). In this view it has been argued that the city was probably part of the Ebla's reign long before the period covered by the archives, at least since the reign of the fifth but last king  The cultural affinity of the centres located in the Aleppo area with the Ebla region has been suggested also based on the analysis of personal names related to different locations quoted in the Ebla texts (Bonechi 1990;Catagnoti 2010). A broader Semitic region corresponding to Syria could be recognised with further onomastic sub-regions, such as that of Ebla (including Aleppo, Emar and Karkemish), the Middle Euphrates (from Mari to Tuttul), the Mediterranean coast, the Cilician area, the Khabur triangle and the area encompassing the foothills north of Aleppo (Gaziantep) as far as Harran (Bonechi 1991: fig.  1; Catagnoti 2010).
So far, only few archaeological information is available for the EBA occupation of Aleppo. Remains an earlier structure dating to EBA were discovered through limited soundings carried out in the inner cella of the LBA temple of the Storm God (Kohlmayer 2016: 301). The identification of limestone slabs laid over the bedrock, as well as of remains of wall foundations and of an EBA floor led the author to assume the existence of an EBA predecessor of the later temple with a similar layout and orientation. Moreover, a foundation deposit containing several bronze and precious stone objects was discovered in connection with this early building (Kohlmayer 2016: fig. 20;2020: 15-18, figs 14-15). It consists of a lapis lazuli pendant, a gold sheet, one toggle pin, a ring, three spearheads and a dagger, plus other small bronze tools. Interestingly, the bent-tanged spearheads, characterised by a pair of parallel slots on the blade, pertain to a typology widespread in northern Syria and Anatolia (e.g., Tarsus, Tell Ahmar/Til Barsip hypogeum, 'Amuq I;Stronach 1959: 107-111;Philip 1989: 337-339, fig. 16: 32, type 15).
The relationships of the Ebla kingdom with the area of the Euphrates have been reconstructed based on Ebla texts and the renowned 'Enna-Dagan Letter' (ARET 13, 4) sent by the king of Mari to the king of Ebla Irkabdamu and recalling the history of relationships between the two kingdoms throughout several generations (Fronzaroli 2003).
During the last years of Igriš-Ḫalab reign, the king of Mari Iblul-il defeated Abarsal and attacked two fortresses in the land of Ḫassuwan. The Mari's military campaigns along the Euphrates led Ebla to pay heavy tribute to avoid armed conflict, a situation that would lasted several years, from the end of Igriš-Ḫalab reign and for the most part of the reign of Irkab-damu (Archi 2015a). The Mari's attempt to control the Euphrates Valley through a series of military expeditions was probably aimed at defining areas of influence, rather than establishing actual territorial control (Archi 2015a). This situation seems to have reversed a few years later during the reign of Irkab-damu, when the territory under the control of Ebla extended up to Karkemish and down to Emar along the Euphrates River (Archi et al. 1993: 238-239;Archi and Biga 2003;Archi 2015a;2015c: 172).
In the 'Treaty between Ebla and Abarsal' (ARET 13, 5) several centres and 'castles' are listed among the Ebla's and Abarsal's domains, all located in an area encompassing the eastern and western banks of the Euphrates and probably the Khabur river (Fronzaroli 2003). Leaving aside the identification of archaeological sites with ancient cities, which still remains hypothetical, the fact that the sphere of Ebla political influence extended up to include both Aleppo and Karkemish, as well as Emar to the east, might let us assume that the Quweiq plain, in the Aleppo's hinterland, was comprised within the territory controlled by Ebla, being geographically located in-between those cities.
Other independent city-states formed an alliance with Ebla, enjoying favourable conditions for trade as it follows from the 'Treaty between Ebla and Abarsal' (ARET 13, 5, Fronzaroli 2003: 43-82;Archi 2011). Among these Ḫ aššum/Ḫassuwan, Ursa'um, and Kakmi'um were important cities, capital of local kingdoms. Regular relationships between Ḫassuwan and Ebla are testified to by the frequent mention in the texts of merchants, scribes, and workers arriving at Ebla from Ḫassuwan (Archi 2015b: 421). This city had a king (malkum) and was one of the tributary kingdoms of Ebla until the fourth year of Ish'ar-damu, when it is supposed to have been ultimately conquered by the latter king (Archi 2015b: 424). Other texts mention the allied city-states of Ursa'um and Kakmi'um, sending man to Ebla or receiving spearheads probably in relation to an imminent military campaign (Archi 2015b). Ursa'um and Ḫassuwan were probably neighbouring states and their location north/north-east of Ebla, in the area of Gaziantep, has been proposed also based on their mention in later 2 nd millennium BC texts (Bonechi 1991;Archi 2015b;Biga 2015). Thus, the northern periphery of the area under Ebla's influence, probably extended close or beyond the Syrian-Turkish border.
The Sajur plain, situated in a strategic position at the crossroads between northern Syria and the Euphrates Valley, could have been home of one of these city-states. The Sajur Valley has been surveyed by Archi, Pecorella andSalvini in the 1970 (Archi et al. 1971) and, more recently, by an international team of the Durham university (Wilkinson et al. 2016). A high number of EBA sites is documented, with some large tells such as Tilbeshar Höyük, reaching 56 ha of extension in the second half of the 3 rd millennium BC (Kepinski 2005). This large mound has been tentatively identified with the cities of Ḫ aššum/Ḫassuwan or Ursa'um (Archi et al. 1971: 44-45;Archi 2015b).
Also in this case the identification remains hypothetical; however, the area of the Sajur valley shows ceramic links with both the Euphrates and the Quweiq plain and appears to be the northernmost area of distribution of ovoid jars with swollen rim, testifying to the existence of models and technological features that were commonly adopted across the Ebla, Aleppo and Sajur areas.
The Aleppo region, as well as the Sajur plain, were areas specialised in olive oil and wine production. 18 Excavations at Tilbeshar Höyük revealed the widespread presence in houses of level IIIC (EB IVA, ca. 2500-2300 BC) of plastered basins with concentration of tartaric acid, as well as high number of olive stones and grape seeds among botanical remains (Kepinski 2007). According to the Ebla texts the most of wine seems to have been received from different centres located to the north of Ebla (Archi 1993: 18;Milano 1994: 435-437).
Overall, the emergence and consolidation of the Ebla kingdom likely fostered a new set of economic and political relations and a more integrated economic system, reflected in terms of material culture in a trend towards homogenization of the pottery repertoire. Besides this, common characters, and similarities with the ceramic tradition of the Ebla region appear over a larger area, and especially at sites on the right bank of the Euphrates river, directly communicating with the areas of Rivers Sajur and Quweiq, and with the Jabbul plain. The introduction of locally manufactured vessels inspired by western Syrian prototypes in the Euphrates Valley might suggest frequent economic relations as a consequence in the change of the balance of political powers, with Ebla extending its political authority as far as the city of Karkemish towards the north and the independent city-state of Emar to the east (an inter-dynastic marriage between the latter two cities is documented during the reign of Irkab-damu; Archi et al. 1993: 290). This expansionist policy of Ebla, however, did not affect local productive systems, and despite the good parallels between the Euphrates and inland western Syria during EB IVA the Euphrates region 'continued to follow its own developmental trajectory' (Cooper 2019: 191). Looking specifically at the archaeological evidence of the Ebla influence along the Euphrates little can be said and, apart from the existence of heavily fortified centres that would lead one think to 'castles' (bàd ki -bàd ki ) mentioned in the Ebla texts -but which are actually much older in date than EB IVA -, there is no proof of Eblaite direct administrative control (official seals and sealings) in the Valley (Cooper 2010).
Ebla as one of the most important trading powers of northern Levant established a number of political and economic agreements with neighbouring kingdoms, ensuring through the enlargement of its sphere of political influence the control over major trading routes, as also suggested by the lively activity of merchants and trading agents coming from different cities and kingdoms mentioned in Ebla texts (for merchants see recently Benati, Bonechi 2020: 56-57). Via Aleppo, Ebla had direct access to key commercial roads leading to Cilicia through the 'Amanus Gate' (probably to be identified with the powerful kingdom of Armi; Bonechi 2016: 86; Steinkeller 2021; but see also Archi 2011), to the 'Amuq plain through the River Afrin (probably Kakmiʾum; Bonechi 2016: 59), to the Sajur plain and the area of Gaziantep (probably Ḫ aššum/Ḫassuwan or Ursa'um) by way of A'azz, and to the Euphrates basin crossing through the Nahr edh-Dhahab or the Jabbul plain via Tell Abu Danne and Umm el-Marra (Del Fabbro 2012: fig. 6; here Fig. 1). Evidence for Aleppo's status as a hub for the trade is also hinted by the mention of a market (KI:LAM 7 ), where timber -one of the more valuable marketed commodities -was sold (Catagnoti 2016: 32;Benati, Bonechi 2020;Steinkeller in press).
The Quweiq plain seems to have acted in mediating relationships between the Ebla core area and other polities that fell under the sphere of political influence of Ebla. Its role as a 'buffer' zone (or 'land bridge'; Del Fabbro 2012: 203) seems reflected in the composition of the ceramic assemblage, which displays some south-western Syrian elements typical of the Ebla region besides characteristic features of the Middle Euphrates Valley, especially of the Karkemish/Tishrin/Tabqa sectors, and the Sajur area. This liminal character can be detected also during the first half of the 3 rd millennium BC, when a similar 'duality' in the material culture is noticeable and reflected in the production and circulation of types inspired by both a 'western' and an 'eastern' tradition. Substantial ceramic connections between Ebla, the Quweiq-Jabbul plains and the Euphrates can be envisaged also for the EB IVB period, following the disruption of the city of Ebla (D'Andrea pers. comm).
Looking at the distribution of EBA sites across the Nahr el-Quweiq plain (Fig. 2), compared with a chart of soil exploitation, a correlation between settlement location and rivers system has been noted (Del Fabbro 2012: 207-208, figs 2-3). Moreover, according to Del Fabbro (2012: 209) the presence of rivers and wadis along transit routes 'ensured many rest stops and easy access to water'. In addition to these roads that cross the plain, another pattern that stands out is the distribution of several sites along a north-south alignment bordering the western fringe of the Nahr el-Quweiq plain, close to the Jebel Sim'an foothills and overlapped to the modern road that leads to the Syrian-Turkish border passing through A'zaz and Kilis (Figs 1-2).

CONCLUSION
Due to the exiguous documentation the Quweiq valley has always been used, or referred to, for general comparisons only, without exploiting its informative potential. A recent research project focused on Aleppo and its hinterland, aimed at understanding the role of Aleppo as central place in a diachronic perspective, has raised new interest on the region (Del Fabbro 2012; Knitter et al. 2014), although key archaeological data about the 3 rd millennium BC occupation are still meagre.
The route Ebla-Aleppo had to be particularly important during the EBA (Del Fabbro 2012), not only for north-south connections linking Ebla, Aleppo, and the Sajur/Euphrates areas, but also for west-east connections with the city of Emar and centres located along the Tabqa Dam basin. This could be an alternative path -likewise modern motorways (Fig. 1) -to another important west-east road that passed through the corridor of Khanaser (Peyronel 2014: 116;Steinkeller 2021). The analysis of the pottery repertoire collected during the 1970s survey in the Nahr el-Quweiq plain -and its comparison with better-known stratified sequences from the nearby Ebla region, Euphrates Valley, and Jabbul and Sajur plains -allowed to detect some significant patterns of connectivity in a diachronic perspective and to further elaborate on the possible key-role of the Aleppo region with respect to the Ebla's socio-political and socio-economic interests during the EB IVA period.