Acknowledgments
We would like to express our deepest gratitude to our friends and colleagues Michael Lewis, Steve Renette, Stephen Batiuk, and Clemens Reichel, who took the time to review and comment on early iterations of this paper. We would also like to thank the three anonymous reviewers for providing us with insightful comments that helped improve this paper.
- 1 These include primarily, but not exclusively, charcoal samples, soil samples, some of the faunal sa (...)
1The aim of this paper is to present the Late Chalcolithic finds of the salvage operations conducted at the site of Tell er-Ramadi (henceforth Ramadi) from 2004-2006 and the subsequent study seasons carried out at the Deir ez-Zor Museum in 2008-2009. Due to the political situation in Syria access to some materials has not been possible.1 With no end in sight to the Syrian conflict, we decided to publish the data we have in a preliminary fashion until it becomes feasible to revisit other materials that remain in Syria.
2With the construction of a new rail line, the site was chosen as an ideal location for a number of piers that would support the rail bridge crossing the Euphrates River. This new project necessitated salvage excavations to be carried out along the southern part of the site in order to document archaeological remains before construction activities. As such, three seasons were conducted at the site carried out by the Department of Antiquities in Deir ez-Zor (2004), and the General Department of Antiquities and Museums in Damascus (2005-2006).
- 2 We use the Santa Fe chronological scheme here (LC1-LC5) as slices of fixed time (Rothman 2001). The (...)
3The focus for this paper will be on the better preserved and understood occupations that date to the 4th millennium BC (the Late Chalcolithic or LC period), where the site is represented by roughly 4 m of deposits belonging to the later part of the LC2 into LC4. The significance of the occupation sequence at Ramadi is that the site was initially occupied by northern Mesopotamian groups during phase 5 (LC2-3), while in the succeeding phase 4 (LC3-4)2 we see evidence of northern and southern Mesopotamian groups living side-by-side.
4The only attempt at contextualizing the Chalcolithic occupation at Ramadi was carried out in the context of the Uruk Expansion (Algaze 1993). Ramadi was considered a ‘station’ in the highly connected Uruk network (Algaze 1993: 48), Algaze defines stations as “smaller isolated Uruk settlements serving as links alongside overland routes between enclaves and the alluvium and also along important routes into the northern plains” (1989: 579). Alongside the sites of Rawa and Qraya, Ramadi occupied an important geographical position in the network as part the first cluster of Uruk sites north of the southern Mesopotamian heartland (ibid.). Stations in Algaze’s model tend to be underexplored, and their function within the system was determined based on their size and location.
5This paper seeks to expand the debate by addressing the micro developments that took place at the site and the nuances of cultural interactions between northern and southern Mesopotamian groups. We see that while Algaze’s notion of ‘station’ may work to describe the economic function of the site in the broader Uruk world, Ramadi has potential to expand our understanding of the events that took place during the 4th millennium, and provide insight into an otherwise poorly understood Chalcolithic of the lower Syrian Euphrates.
- 3 Middle Bronze Age and Ubaid levels and materials were not encountered during our excavations in the (...)
6Ramadi is a multiperiod tell located along the west bank of the Euphrates River in Syria roughly 12 km north of Mari (Tell al Hariri) (fig. 1a). The mound rises 10 m above the surrounding plain and measures 250 m north/south and 300 m east/west (Geyer and Monchambert 2003: 77) covering roughly an area of 8 ha (Geyer and Monchambert 2003: 245) (fig. 1b). The tell is composed of a high mounded area measuring 2 ha, and lower town roughly 6 ha in extent. The modern village of Ramadi covers the lower part of the site while the upper town is covered by a dense 20th century cemetery. The sequence of the site is composed of traces of Middle Bronze Age, latter part of Early Bronze Age (2400-2000 BC), Late Chalcolithic 3-4, Late Chalcolithic 2-3, and Ubaid period occupation.3
Fig. 1 –A. Map of the region with some of the sites mentioned in the text; B. Map of the site with excavation trenches; C. Plan of trenches (2004-2006) with modern graves indicated.
S. Razzaz and K. Abu Jayyab.
- 4 Cultivation was possible only in basins with deeper soil profiles that collected sediment and water (...)
7The Euphrates River Valley near the site is well defined having cut into the surrounding plateaus, the Shamiya to the west and the Jazira to the east. The site is located in an area that receives between 159-133 mm of annual rainfall, most of which is concentrated in the winter months. Precipitation is short, intense, and infrequent with high moisture deficit rendering the annual variability in rainfall up to 50% (Geyer and Besançon 2002: 62). Due to the short and intense rains, water infiltration is very limited and subject to runoff, causing high levels of soil erosion and shortage in water reserves in the soil (Geyer and Besançon 2002). This process rendered the high plateaus surrounding the river valley rarely suitable for crop cultivation.4 Occupation in the valley bottom also represents a conundrum; the only suitable source of water in the valley is the Euphrates, and because of the annual cycle of flooding, the closer settlement was to the water the greater the threat of destructive flooding. Accordingly, 4th millennium sites (e.g. Ramadi and Qraya) were founded on Pleistocene terraces (QII) that were composed of cemented solid pebbles that were resistant to the river’s erosive action (Geyer and Besançon 2002: 66). This allowed the inhabitants to take advantage of the proximity to the main river channel, while avoiding the threat of flooding.
8The location of Ramadi allows for certain advantages in terms of the diversification of subsistence pursuits. The site was perfectly positioned for easy access to surrounding deep, well-watered and fertile alluvial soils ideal for cultivation. Moreover, this location near the river provided access to a wide variety of riverine resources. Ramadi’s close proximity to the Shamiya steppe plateau provided access to grazing grounds for herds of sheep and goats. Proximity to the steppe also provides potential for hunting wild game. This is a documented practice for Chalcolithic sites located along the steppe margins such as Kuran (Zeder et al. 2013), Umm Qseir (Zeder 1994), and El-Kowm (Vila 1998b). Finally, the Euphrates River was a major artery for movement across the region especially in terms of facilitating south bound traffic.
9In 2004 the Department of Antiquities in Deir ez-Zor excavated a 12 × 10 m trench (fig. 1c) that paralleled the path of the train tracks. This operation recorded three levels of occupation and substantial architecture primarily from the Late Chalcolithic (henceforth LC). In 2005, a north/south, east/west grid was laid out across the site (fig. 1c) with the excavations focused on squares IV3 and V3. In 2006, excavations continued in square V3 in addition to opening square V2 and the southern sounding.
- 5 The decision to use a backhoe was made due to the limited time available to the team for excavation (...)
10The tell was covered by a large number of 20th century burials, densely packed together. The excavations in 2004 revealed that the top 2 m of the tell either consisted of topsoil or were heavily damaged by grave excavations. Accordingly, a decision was made to remove about 1-1.5 m of deposition from the top of squares IV3-V3 with a backhoe.5
11The occupation sequence uncovered (table 1) complemented Beyer’s excavations (Beyer 1991). Whereas in the 1990’s excavations along the north portion of the mound revealed a sequence consisting of Middle Bronze Age (phase 2), southern Uruk (LC3-4) (phase 4), and Ubaid strata (phase 6), excavations along the south of the mound showed southern Uruk layers sandwiched by the later part of the Early Bronze age (phase 3) and northern Mesopotamian LC2/3 levels (phase 5).
Table 1 – Periodization at the site of Ramadi.
12The earliest occupation uncovered in this portion of the mound, phase 5, may fall in the late LC2/early LC3 period based on material parallels. The phase 5 occupation here was established on top of virgin soil which leads us to believe that the Ubaid occupation (phase 6) was concentrated along the northern part of the site (Beyer 1991). Phase 5 was the best preserved and most substantial of all the occupation phases excavated. This phase was divided into three subphases based on architectural remains.
- 6 We refer to this building as the public building. Here we do not see that this building acted as an (...)
13Our widest exposure comes from subphase 5a where traces of architecture were found across all the operations (fig. 2). This subphase also had the most damage with a combination of modern graves, and EB and Uruk pits cutting into it. Two significant complexes were uncovered here. In the northwest, a portion of a large complex was found in the 2004 trench and square IV3. The complex had thick walls (0.5-0.8 m in thickness) and consisted of two large halls (D1-D4), a smaller room (D5) and what seems to be a kitchen area with two tanurs embedded in the wall (D3). The thick walls, wide halls, and unique assemblage recovered (discussed in more detail below) led us to believe that this complex represented a public gathering space for the inhabitants of the town and was not purely a residential unit.6
Fig. 2 – Plan of Ramadi subphase 5a
14K. Abu Jayyab.
- 7 The destruction of this part of the complex was because of the foundations of a 20th century large (...)
15Another complex was excavated to the east of the public building across what seemed to be a stone paved alleyway. This complex extended across three squares (IV3-V3-V2) and seemed to have extended further to the east. The complex formed an L-shape wrapped around a paved courtyard in the northeast. The courtyard produced evidence of lithic production with three cores recovered in addition to fragments of debitage and two hammerstones interspersed between the stones of the pavement. A jar buried with the remains of an adolescent was also recovered from under the floor of the courtyard (fig. 2). The burial jar was a large triangular section rim jar, commonly found during the phase 5 occupation (fig. 2). The southern wing may have been composed of a tripartite building; however, we cannot say for sure since we do not know its complete shape. The central room of the southern wing seemed to have been an open-air space judging by the paving and the presence of two small hearths. The western wing was composed of three adjacent rooms but does not seem to fall into the general tripartite building plan. There was no clear access between the two wings of the complex and, if it was present, it was through the northern courtyard at the point where the western wing was damaged.7
- 8 This condition may suggest that EB and southern Uruk remains were ephemeral in this part of the sit (...)
16In square V2, phase 5a remains consisted of a few wall fragments.8 An interesting find was a cluster of bottles, pedestaled bowls (fig. 3), a miniature jar and cowry shell left in situ adjacent to a wall in square V2. These finds were perhaps in the corner of a room otherwise destroyed by later pitting. Two similar bottles and a similar miniature jar were found in the public building.
Fig. 3 – Collection of miniature vessels from square V2 (1-11) and the Public Building (12-14).
K. Abu Jayyab.
17Subphase 5b remains were found mainly in square V2, the southern sounding, and the northern part of square V3. The sparse distribution was mainly due to targeting areas where subphase 5a architecture was damaged. The main components of this level were portions of two large buildings in square V2 and the southern sounding (fig. 4).
Fig. 4 – Plan of Ramadi subphases 5b and 5c.
K. Abu Jayyab.
18Subphase 5c remains were even more ephemeral as they were only uncovered in the deepest portion of the southern sounding immediately above virgin soil. An unusual feature about the architecture recovered from this phase is that the wall had stone foundations, which has not been documented elsewhere at the site.
19The ceramic assemblage from Phase 5 remains rather homogenous throughout the sequence, albeit with changes in frequency of certain types and the appearance or disappearance of a few others. In total, Phase 5 yielded 756 diagnostic sherds with the majority coming from the latestphase, 5a (485). The majority of vessels were chaff-faced wares (CFW), made from clays tempered with plant or a combination of plant and mineral inclusions (62.2%). The second most common group were fine wares (24.2%) with either no visible inclusions (19.2%) or with very fine mica and plant remains (5%). Mineral tempered wares formed a small group (8.8%), while shell tempering was mainly reserved for cooking wares (4.3%) (fig. 5).
Fig. 5 – Ceramic fabric frequencies for Phase 5, Phase 4 northern, and Phase 4 southern Uruk.
K. Abu Jayyab.
20The assemblage could be divided into five functional categories: coarse serving bowls, storage jars, fine ware serving vessels, cooking vessels, and other highly specialized miscellaneous vessels. Coarse serving bowls represent the highest proportion of vessels observed (table 2).
Table 2 – Frequency of different vessel categories at Ramadi.
- 9 Some of the larger Coarse bowls could have been used for cooking or as lids, though with the absenc (...)
21There are eight main types of coarse serving bowls (table 3) ranging in size from what seems to be individual serving to group sharing bowls.9 The two most common forms of coarse bowls are inwardly beveled rim bowls (fig. 6: 1-5) and club-head bowls (fig. 6: 6-9). These types were used during the three sub-phases of Period 5 occupation. Inwardly beveled rim bowls are large bowls, ranging between 25-36 cm in diameter. They were primarily constructed through coiling with the majority wet wiped and slightly smoothed before firing.
Table 3 – Detailed vessel frequency at Ramadi.
22Two examples had scraped bottoms, which could have been a more common practice that is underrepresented due to the absence of a complete profile in most cases. The majority of the samples were made with vegetal inclusions or a mix of vegetal and mineral inclusions.
23Club-head bowls were less frequent. They range in diameter between 25-36 cm and appear to be made using thick coils. In terms of surface treatment, they were generally wet smoothed with some examples burnished. Some of these vessels were fired in a reducing atmosphere resulting in surface colours that are primarily grey, grey/brown and exhibiting a dark core. Most of these vessels contain vegetal inclusions with some having a mixed vegetal/mineral fabric. Club-head bowls are found across Northern Mesopotamia, mainly from Late LC2 and early LC3 levels, where they tend to follow the distribution of inwardly beveled rim bowls.
24Three types of serving bowls remain constant throughout the sequence, albeit less common: bowls with flat rims and an extended internal lip (fig. 6: 11), bowls with flat rims and an extended external lip (fig. 6: 10), and squat bowls with in-turned rims (fig. 6: 12). In the first two types, scraping along the bottom of the vessel on some examples can be noted. Additionally, a number of examples had a thick red slip applied to them. These vessels were primarily produced from a vegetal fabric with some containing additional minerals.
Fig. 6 – Phase 5 ceramics: coarse bowls, jars and basins
I. Schwartz.
25The Phase 5 assemblage surprisingly contains only two examples of hammerhead bowls that come from the latest subphase, 5a. Moreover, the examples found at Ramadi are not the classic hammerheads that are common during the LC3 (Pollock and Coursey 1995; Balossi-Restelli 2006) but rather a version with the rim extending towards the interior (Pollock and Coursey 1995: 105).
- 10 Their absence could be attributed to the limited exposure in subphase 5c. Nevertheless, if they wer (...)
26Bevelled rim bowls (henceforth BRBs), are found at the phase 5 occupation at Ramadi prior to the appearance of other southern Uruk style vessels at the site. They are absent during the earliest subphase 5c,10 and appear initially during 5b (11.7% of diagnostic sherds) and increase during subphase 5a (16.5%).
- 11 Renette S., Abu Jayyab K., Gibbon E., Lewis M., Qadir Z., Cabral R. and Tomé A., Late Chalcolithic (...)
27The early appearance of BRBs at a site usually preceded the arrival of the full Uruk repertoire (Butterlin 2018: 412). This phenomenon could be seen at the sites of Hacinebi (Stein 2002: 150), Nineveh (Butterlin 2018: 412), Hamoukar during LC3 in area B (Reichel personal communication), and Kani Shaie during late LC2/early LC3.11 At Grai Resh and Tell Leilan, on the other hand, a number of BRBs were found embedded in the local northern Mesopotamian assemblage (Schwartz 1988; Kepinski 2011: 38) without a subsequent Uruk phase (Potts 2009).
28Storage jars account for 22.9% of the phase 5 assemblage (n: 173). The vast majority of the larger storage jars are rather homogenous in shape, fabric and construction technique. They are vegetal or mixed vegetal and mineral tempered with brown to pale orange surfaces and dark reduced cores. Judging by their bumpy surfaces they were primarily coil built with the rim attached as one piece to the body of the vessel. The finishing of these Jars was carried out by wet smoothing the surface by hand. The rim on the other hand seems to have been finished on a slow rotary devise, perhaps using a cloth, judging by the striations along the rim. The most common rim form (11%) is an extended triangular section rim (fig. 6: 15-20) and a variant of it that has an exterior bevelled rim with a groove along it (fig. 6: 16).
29Grooved rim jars and internally hollowed jars are two categories of storage vessel found during phase 5 occupation, and tend to increase during phase 4. Grooved rim jars are a larger form that follows the ware description above (fig. 6: 24-26). Internally hollowed jars are finer and smaller variation that has a slightly different rim shape with an internally bevelled lip (fig. 6: 22-23). They appear during subphase 5b and continue in the sequence.
30Two examples of flange rim jars (fig. 6: 27-28) were recovered from the phase 5a occupation. Both examples feature primarily vegetal inclusions, while one example was red slipped and burnished.
31Basins (fig. 6: 29-30) form a small but consistent portion of the phase 5 assemblage (3%). They are coarse, large, deep bowls which were mainly vegetal tempered. Similar basins were recorded at Grai Resh Level BII (Kepinski 2011: plate 7: 1-2), but do not seem to have a large-scale distribution.
32Fine ware vessels represent a third (29.8%) of all diagnostic vessels recovered from phase 5 (n: 226). Fine wares are made from well levigated clays with the majority not showing traces of any inclusions. Some fine wares have traces of very fine mineral and vegetal inclusions. Overall these vessels are thin walled and well fired ranging in colour from buff and yellow to pale green. They could be divided into three general forms: bowls, cups, and small beaded globular pots (fig. 7: 23-31).
Fig. 7 – Phase 5 ceramics: fine ware vessels, and cooking pots
I. Schwartz.
- 12 The observations on the surface of these vessels is compatible with Roux and Courtey’s (1998: 748-4 (...)
33The bowls (fig. 7: 1-14) and cups (fig. 7: 15-20) were primarily carinated and seem to have been produced in a number of steps. The first step was forming the rounded vessel base, perhaps in a mold, the second step consisted of adding coils or a slab on top of the base (judging by joint smoothing in the area of carination). Finally, the body and rim of the vessel were finished on a rotary device judging by the striations on these parts of the vessel.12 Two fine ware carinated bowl wasters suggests that pottery was being produced locally at the site. This is supported by the recovery of a ceramic scraper.
34Carinated bowls and cups were intended as drinking vessels, as they seem too small for food consumption and their round bases necessitates that they should be held at all time. The quality of these vessels suggests that they could have been reserved for drinking on special occasions.
35Two types of cooking vessels were found at Ramadi. The first type was the ubiquitous casseroles (fig. 7: 32-38), and the second was a coarse crudely made hole mouth pot (fig. 7: 39-44). Initially, during the earliest occupation, phase 5c, only hole mouth pots were recovered. From 5b onward both vessels were found at a comparable frequency.
- 13 Determining the volume of each sub-type of hole mouth pot is not possible since we do not have any (...)
36Hole mouth pots were crudely made coil-built vessels tempered with what appears to be fossiliferous limestone or crushed shell. In most cases the surfaces of these vessels were left untreated with only two examples from the three subphases burnished (n: 21). In terms of morphology they were restricted vessels that seem to fall into two size classes.13 The smaller examples had a simple rounded rim and ranged in diameter between 16 and 23 cm (fig. 7: 41). The larger variety had a slightly thickened rim and ranged in rim diameter between 25-30 cm (fig. 7: 39-40).
- 14 Although the more common LC3/LC4 variety with an out turned rim similar to those seen at Zeytinli B (...)
37Casseroles had two main sub-types: a finer variety with a beveled or beaded rim (fig. 7: 35-38) and a coarser variety with a heavier rounded rim (fig. 7: 32-34).14 The examples of casseroles found at Ramadi were primarily vegetal or mixed vegetal/mineral tempered. A few examples, however, were shell tempered (n: 3) while one vessel exhibited two different recipes between the base and upper body. Technically, casseroles were built in a similar fashion to carinated bowls with the bottom probably produced in a mould and the upper body coil-built and later smoothed. These vessels were mostly well smoothed and sometimes burnished and slipped, perhaps to reduce the porosity of the vessel (Skibo 2013).
- 15 With the exception of one pedestalled bowl, from 5c, this collection of unique vessels was recovere (...)
38Other less frequently recorded vessels are pedestalled bowls, bottles, pot stands, and miniature vessels (fig. 3). The majority of these vessels were found during the subphase 5a occupation.15 The large public building produced two bottles and a miniature jar, while the rest of these vessels were recovered from a damaged area south of the eastern complex (see above). A somewhat similar assemblage of bottles, pedestal bowls, and miniature vessels comes from Grai Resh level IIA (Kepinski 2011: fig. 14: plate 3). There they found these objects clustered together in a pit dating to the LC3 (Kepinski 2011: 36).
39It is important to note here that the assemblage recovered from the fill of the public building was distinct from the rest of the subphase 5a assemblage (table 4). Overall, the public building had a higher concentration (42% compared to 20%) of fine ware carinated bowls and cups, and a slightly higher concentration of storage jars. Coarser bowls were not as common in the public building as they were at the rest of the site (19% compared to 50%), and BRBs were completely absent from the public building. In terms of cooking pots, the public building only had open casseroles and did not produce any evidence of the crude home-made hole mouth pots. Another factor that distinguished the public building was that the only two sealings found at the site came from this structure. Very few complete fine ware carinated bowls were found directly on the floor. The majority of the assemblage from the public building came from the fill of the rooms, however a large number of the vessels from the fill of the building were near complete, which could imply that they may have originally been part of the artifacts from within the structure.
Table 4 – Distribution of vessel categories across the various phases at Ramadi.
40Within the public structure itself the majority of fine ware (51/55) vessels were recovered from the large halls (D1 and D4). Hall D1 seemed to have been the focal gathering or dining locus of the building with the majority of fine wares, storage jar, and coarse bowl remains recovered from there. The kitchen area contained one casserole open cooking pot and two course bowls.
41The combination of architectural character and overall artifact signature may point to a special status associated with the western ‘public’ building. This building may have acted as a gathering place for the inhabitants of the site and/or the heads of households from the site and nearby areas, where essential negotiations, transactions, and festivities were carried out. Judging by the number of cups and fine bowls recovered, these gatherings would have been accompanied by eating and drinking.
42Without radiocarbon dates we can only build the chronology of the site based on relative dates (table 1). The strong parallels that Ramadi 5 has with the dated sites of Hacinebi level A (dated to roughly 3800 BC) and Grai Resh (Kepinski 2011), in addition to Leilatape sites from the Caucasus suggests an occupation somewhere during late LC2 and LC3 (3800-3600 BC).
- 16 Almost all the lithics come from phase 5 at Tell Ramadi, which corresponds with the LLC occupation (...)
- 17 As a result of the limited information conveyed in lithic illustrations, Canaanean blades were iden (...)
43Almost all the lithics come from phase 5 of the occupation of Ramadi.16 The most common tool form present in the assemblage are blade fragments (n: 33), with two technologies attested at the site (fig. 8). The first are large, regular Canaanean blades17 (n: 16) that would have been produced through pressure flaking with a lever and were made from a beige, fine-grained flint. The vast majority of Canaanean blades are represented by segmented medial fragments that range between 35 mm and 59 mm in length and tend to have trapezoidal cross sections (fig. 8: 3). Without more information about the bulbs and platforms it is impossible to determine the type of tip that was used on the lever, although copper tips are attested among contemporary assemblages (Altınbilek-Algül et al. 2012; Manclossi, Rosen and Miroschedji 2016).
Fig. 8 – Lithics from Ramadi.
A. Glasser.
44Simple blades (n: 16) comprise the second blade technology represented at Ramadi (fig. 8: 2). These blades are generally less than 20 mm wide and around 5 mm thick and have triangular and trapezoidal cross sections. These blades were likely produced through a combination of regular hand pressure flaking and through reinforced pressure flaking using a standing crutch. Larger, irregular blades are probably indicative of indirect percussion as well (Altınbilek-Algül et al. 2012). The recovery of seven simple blade cores, the higher proportion of complete blades and proximal segments, and some possible debitage material, suggests that simple blades were manufactured on-site. These blades were made from homogenous, fine-grained flint that varied in colour from beige to dark brown. These were likely produced from transported river nodules accessible from the Euphrates River.
45The cores from Ramadi (fig. 8: 1) share strong affinities with simple blade cores recovered from Phase A at Hacinebi (Edens 1999). They are bullet shaped, with a single platform and unidirectional blade removals. Some show evidence of edge grinding and a number appear to have been fully expended. Their lengths range from 40 to 66 mm, with diameters ranging between 11 to 33 mm. It is worth noting that most of the negative blade scars found on recovered cores are smaller than any of the illustrated simple blades, with widths less than 8 mm. Cortex is found covering almost an entire face on at least two of cores recovered, which may relate to how these cores were immobilized during knapping (Manclossi, Rosen and Miroschedji 2016). The presence of some debitage, as well as flake removals on one of the cores, suggests that manufacturing blade cores involved some degree of shaping, however the low proportion of debitage, as well as one blade with cortex along the distal end suggest that shaping may have been minimal (Manclossi, Rosen and Miroschedji 2016). Appropriately shaped nodules may have been preferentially selected to minimize the amount of shaping.
46It is difficult to ascertain if the illustrations reflect use-wear, but it is worth noting that the Canaanean blades have a much higher proportion of working edges that appear to have use-wear. The same is true even when only comparing medial segments. It is not clear if the retouch, gloss, or bitumen were absent like the illustrations would imply, but it seems more likely these attributes were just not recorded. Gloss was recorded along the ventral surface of a working edge on one Canaanean medial fragment.
47The co-occurrence of these two blade technologies at Ramadi is consistent with lithic assemblages from a number of contemporary sites (Braidwood and Braidwood 1960; Edens 1999; Thomalsky 2012, 2019). Canaanean blades have a widespread distribution in the Near East between 4000 and 1950 cal. BC and are found at a number of sites including Norsuntepe, Tell al-‘Abr, Arslantepe VII, Tell Yarmouth, Grai Resh, Tell Afis, and Tell al-Judaidah phase F, Hama phase K, and maybe Ras Shamra IIIB (Philip 2002; Thomalsky 2012; Chabot and Pelegrin 2012; Manclossi, Rosen and Miroschedji 2016; Braidwood and Braidwood 2016). It is believed that Canaanean blades were mass produced at a few key centres and were circulated across the Near East as exchange networks expanded (Thomalsky 2012). Workshops have been identified at Hacinebi, Titris Höyük, Grai Resh, and Tell Brak (Edens 1999; Hartenberger, Rosen, and Matney 2000; Thomalsky 2012). Use-wear studies suggest that Canaanean blades were most commonly hafted as segments in sickles to harvest cereals, while some evidence also points to their use for threshing. Their appearance is associated with the intensification of agriculture and craft specialization (Hartenberger, Rosen and Matney 2000; Thomalsky 2012; Khalidi 2014). The lack of cores and debitage—characteristic of Canaanean blade production—at Tell Ramadi suggests that these blades were either imports or produced in a part of the site yet to be excavated. Alternatively, simple blade technology across LC sites is associated with household level production, which is consistent with the evidence from Ramadi (Edens 1999; Thomalsky 2012; Chabot and Pelegrin 2012). The recovery of four simple blade cores, along with six hammerstones and debitage, from the courtyard of the eastern complex may suggest that this area was used for flint knapping. It is also worth noting that the largest concentration of lithics were recovered from the public building, which includes a disproportionate number of complete blades and proximal segments, and one proximal fragment from an obsidian blade (the only piece of obsidian recovered from the site).
48Four small tabular scrapers, or fan scrapers, were also recovered from Ramadi (fig. 8: 5). These are characteristic flakes with cortex that covers the dorsal surface and steep unifacial dorsal retouch that extends along 40-60% of the circumference. These are consistent with types found across the Levant and the Near East from the Pottery Neolithic to the Early Bronze Age including Norsuntepe, Hacinebi, Habuba Kabira, Grai Resh, Tepe Gawra XI-IX, Tell Brak, Jebel Aruda, Sheikh Hassan, Umm Qseir, El-Kowm 2 Caracol, Hassek Höyük, Amuq G levels, Al-Judaidah and Tell Qraya (Shimabaku 1984; Edens 1991; Philip 2002). Interestingly, the tabular scrapers from Ramadi were made from the same fine-grained, beige chert as the Canaanean blades. The four tabular flakes found at the site were recovered from different contexts. These include the eastern courtyard, the outside space by the public building, phase 4 pit II, and sub-phase 5c. Use-wear studies from other sites suggest that tabular scrapers were primarily used for skinning and animal butchering activities, although other suggestions for their use include hide-processing, wood working, reed scraping, and sheep shearing (Rowan and Levy 1991; Henry 1995; Barket and Bell 2011; Yerkes et al. 2016).
- 18 The full assemblage has not been studied as of yet, the material presented here is a sample from th (...)
49A total of 274 faunal specimens were recovered from domestic contexts of phase 5 at Ramadi,18 of which 242 could be identified to the species level. Domestic species dominate the assemblage (table 5), with wild mammal remains significantly less represented. Sheep and goat represent the majority of the domestic species exploited during phase 5, constituting 87.2% of the NISP, with the ratio of sheep to goat almost equal (1:1). Among the goat remains, four specimens are significantly larger than the rest of the assemblage, and could be indicative of the presence of wild goat (capra aegagrus) during this occupational phase.
Table 5 – Faunal remains at Ramadi: distribution of taxa at Ramadi and age profile of caprine slaughter at Ramadi phase 5.
50In order to obtain a preliminary understanding of the caprine exploitation system during phase 5, their slaughter profile was produced from isolated teeth (39 lower and upper teeth; Ducos 1968; Payne 1973; Vila 1998; Helmer and Vigne 2004). Analysis of this profile suggests most goats were slaughtered during early adulthood (class D, 1 to 2 years: 36.8% of NISP) in order to obtain maximum meat yields.
51Although meat is typically the first product to be exploited from domestic caprine species, the presence of older animals (class EF, G, HI) suggests that secondary products, such as milk and wool, were also being exploited. The secondary peak in the slaughter of age class G caprines (between 4 and 6 years old), indicates that some caprines were being maintained for longer periods of time, likely for the exploitation of their fleece. This type of exploitation pattern was similarly identified at El-Kowm 2, however from a later period (3500-3300 BC) (Vila 1994, Helmer et al. 2007). The discovery at Tell Ramadi of many objects related to the processing of wool, including spindle whorls and loom weights (fig. 9), seems to confirm this hypothesis.
Fig. 9 – Objects from Ramadi Phases 5 and 4.
A. Glasser.
52Equines represent the second most exploited animal during the phase 5 occupation of Ramadi. However, the poor preservation of these remains makes their specific determination difficult and it has not been possible to identify the specific species these remains represent (wild donkey, hemione etc.).
53Other species are also present in smaller proportions. This includes gazelle, hare, birds and bovids. Bovids are represented by only three specimens, with their size correlating with domestic cattle species.
54The archaeozoological data from phase 5 of Tell Ramadi, suggests the animal economy is largely based on caprines (sheep and goats) and their various products, before (milk, fleece) and after their slaughter (meat, fat, skin). Cattle, represented in very small quantities, are probably exploited for their strength in agricultural activity or as pack animals or beast of burden. Finally, wild animal exploitation during this phase remains a secondary subsistence strategy, likely as a food supplement.
55The most common objects found at the site during phase 5 were spindle whorls (fig. 9: 6-9, 13). In total, 26 were recovered from this phase, most of which were found in subphase 5a (n: 19). Spindle whorls from 5a were primarily recovered from the eastern complex (n: 16), with only three coming from the public building. Two loom weights were found one in each of the public building and one in the eastern complex (fig. 9: 14). It seems that several steps of the weaving chaîne opératoire were carried out in the eastern complex. The presence of tabular flakes, spindle whorls, and loom weights in this complex point to activities such as sheering, spinning, and weaving taking place here.
56Grinding stones were extremely rare from this part of the site, the only two fragments found were from the open area to the south of the public building. A stone door socket was also recovered from this area (fig. 9: 11). A number of hammerstones were found in different parts of the site, however the main concentration was in the eastern complex courtyard in association with flint cores and debitage (fig. 8: 4).
57Two clay sealings with rope impressions (fig. 9: 3,5), and scrunched up sealing clay (fig. 9: 4) were found in hall D4 in the public complex. One of the sealings had a seal impression which could not be deciphered (fig. 9: 3), while the other one was blank on its dorsal side (fig. 9: 5). Other objects recovered include a worked cowrie shell (fig. 9: 2), three clay sickles (fig. 9: 12), ten clay sling bullets (fig. 9: 15), a miniature limestone vessel from the public building (fig. 9: 1) and a ceramic scraper.
58The material from phase 4 was found in two discard pits, and was mainly composed of southern Mesopotamian Uruk materials (fig. 10). Pit I, the northern smaller pit, was found in the 2004 trench. The pit was oval in shape, oriented northeast/southwest roughly 4 × 2 m in diameter at its opening and narrowed towards its bottom. Its fill consisted of reddish/brown silt interspersed with ash lenses. The pit contained southern Mesopotamian ceramics including BRBs and a fragment of a clay sickle blade.
Fig. 10 – Plan of Ramadi Phase 4 and the western section of the deep sounding detailing the deposition layers of Pit II.
K. Abu Jayyab.
59A larger pit, pit II, was excavated in the southern sounding (fig. 10). The extent that was excavated in the southern sounding was 8 m long and 3 m wide. The bottom of the pit near the northern section of the sounding reached a depth of 2.3 m. While the majority of the pit was contained in the southern sounding, a small extension of it remains outside of this trench and unexplored. Pit II contained a similar fill profile to pit I with silty deposit interspersed with ash lenses. The material recovered from pit II was a combination of southern Uruk—including over 1,500 complete and fragmented BRBs—and northern Mesopotamian ceramics, interspersed with spindle whorls, clay sickles and faunal remains.
60During the phase 4 occupation it appears that this portion of the site was used as an open-air space designated perhaps for waste disposal. The deposition points to the disposal of waste products primarily resulting from large gatherings and feasting events (cooking vessels, serving bowls and animal bones) in addition to day-to-day dumping (broken sickles and spindle whorls). Despite the absence of clear sediment layering, the deposition of material in pit II seemed to have been carried out in discrete episodes. That is materials were discarded in one group, buried with silty sediments, followed by another discard episode. A total of seven such episodes were documented in the pit. Judging by the clustering of material within the pit, we consider an episode to be a single or clustered event in time, not protracted.
- 19 This practice has been attested at Qraya (Reimer 1989), Tell Humeida (Fenollós and Al-Abedallah 201 (...)
61Massive disposal pits filled with BRBs and other materials seemed to have been a common feature at other nearby Uruk sites.19 Judging by the ceramics recovered (both northern and southern) the pits seem to date to the LC3/ LC4 periods of the Santa Fe chronological scheme (Rothman 2001). Such discard pits can provide valuable insight into past human activity (Skuldbøl and Colantoni 2018: 641; Wright et al. 1980).
- 20 This count excludes BRBs, which amount to 1,574 examples from pit II. A similarly high frequency of (...)
62In total, 1,760 diagnostic sherds were recovered from the two phase 4 pits. Only 13 diagnostic sherds were recovered from pit I, while the majority of the ceramics came from pit II (1,747). The vast majority of vessels were BRBs (n: 1575, 89.5%), many of which were complete (n: 103) or near complete. Pit I contained exclusively southern Uruk material with six fragments of medium storage jars, two scraped cooking pots, a carinated bowl, four complete BRBs, and a fragment of a clay sickle blade. Pit II, on the other hand, contained a mix of southern Mesopotamian (n: 125)20 and northern Mesopotamian Late Chalcolithic materials (n: 61). The contrast between the northern and the southern assemblages could be seen not only in forms (fig. 11), but also in tempering practices (fig. 5b and c), and manufacturing techniques. In southern vessels, the use of mineral inclusions (limestone, calcite) is more commonly attested than in the northern Mesopotamian assemblage, which consists primarily of chaff-faced wares. In terms of manufacturing techniques, the use of the wheel is attested on southern Uruk storage jars. It was not clear if the wheel was used in the manufacturing process or in the finishing of the jars, with striation lines observed on vessel interiors. Such marks were only observed on finer carinated bowls of northern Mesopotamian LC materials.
Fig. 11 – Phase 4 ceramics: Northern Mesopotamian wares (top), Southern Mesopotamian Uruk wares (bottom).
E. Gibbon.
63The most common southern Mesopotamian forms recovered (besides BRBs) were combed cooking wares (fig. 11: 18-20). These vessels have flaring rims, either tapered or rounded, and a round body. Straight spouts (fig. 11: 21) and flat handles (fig. 11: 20) were commonly applied to these vessels. Similar examples were recorded at Qraya (Simpson 1983: fig. 82), Hacinebi (Pierce 2000: 117: fig. 14), Kurban VI A (Algaze 1990: plate 25: H), Chogha Mish (Delougaz and Kantor 1996: plate 96), Nineveh upper level 3 (Gut 2002: plate 4c), and Sheikh Hassan (Boese 1995: 20: Fig. 7: a-b). Excavations at Šarafabad uncovered a pit that contained a large number of similar cooking vessels (Wright et al. 1980: 272-273).
64Flaring rim storage jars exhibit a wide variety of rim morphologies and body shapes. The most common form is a short triangle section rim (fig. 11: 23, 31). Storage jars are commonly decorated with nail impressions (fig. 11: 25), punctate (fig. 11: 26-27), incised cross-hatching, and reserve slip (fig. 11: 28-33). The most commonly attested decoration was a combination of a line of nail impression/punctate along the shoulder of a vessel with diagonal lines of reserved slip on the body (fig. 11: 31, 33). Fragments of nose lug jars (fig. 11: 24) were also found in pit II though not as common as cooking wars and reserved slip wares. One example each of a band rim bowl and a jug with an extended handle (fig. 11: 16) were also recovered from pit II. Similar examples of these jugs were attested at Chogha Mish (Delougaz and Kantor 1996: plate 82: A-G), and from Uruk in the sounding levels XI-XII (Sürenhagen 1987: fig. 24). What is conspicuously absent from the southern assemblage are the ubiquitous string cut bowls, and coarse trays, which were commonly found at the nearby Uruk sites of Qraya (Simpson 1983), and Hacinebi (Pollock and Coursey 1995: 113).
65The northern Mesopotamian assemblage from phase 4 showed close parallels to the phase 5 assemblage. Inwardly beveled rim bowls (fig. 11: 6), fine ware carinated bowls (fig. 11: 9) and cups, triangular rim jars, and in-turned squat bowls (fig. 11: 7) were all present. The major differences were the absence of cooking pots (with the exception of one casserole), the appearance of jars with multiple grooves along the interior rim (fig. 11: 12) (Balossi-Restelli 2006: fig. 11: K), and the appearance of classic hammerhead bowls (fig. 11: 1-5).
66Ramadi 4 seems to date to sometime around the later part of LC3 and LC4. The northern Mesopotamian assemblage recovered from this level parallels those of Hacinebi B1 and 2. The southern Uruk assemblage shows close parallels to the assemblage found at the nearby site of Qraya, where the most recent levels were dated to roughly 3600 BC (Wright and Rupley 2001: 100-101).
- 21 Only the seventh (locus 5) and the first (locus 58) depositional episodes from the pit have been ex (...)
67The faunal assemblage from phase 4 includes 352 specimens of which 275 have been identified (table 5). The entire phase 4 assemblage was recovered from a single pit (Pit II),21 with similar distribution of species in each fill layer.
68As in the previous phase, the remains of caprines (sheep and goat), are the majority with 60.7% of the NISP (table 5). However, the ratio of sheep to goat has significantly increased from the previous phase, with three sheep to a single goat (3:1). This ratio could be linked to a privileged exploitation of sheep for its wool, in addition to its meat and milk. Unfortunately, it is not possible to confirm this hypothesis due to the insufficient number of caprine teeth (n: 4) present in the assemblage to build a reliable slaughter age profile. The four teeth that were recovered from the phase 4 levels belong to caprines between 2 and 4 years old. Although caprine remains were predominant during both phases 4 and 5, there is a significant change in herd composition between these phases. We can see that by phase 4 there is a sharp increase in the sheep to goat ratio from approximately 1:1 in phase 5, to 3:1 during phase 4.
69After caprines, the second most common taxon are suids, with 21.5% of the NISP. These mainly correspond to domestic pigs, however nine of the specimens have been attributed to wild boars. Cattle are the third most common domestic taxon. There are nine identified cattle specimens, two of which have been identified as aurochs due to their large dimensions. Other identified wild animals from this phase include equines, fallow deer of Mesopotamia, gazelles, hares, avifauna, fish and carnivores (table 5). Unfortunately, since the faunal data available for this phase comes from a single pit, it provides only a partial picture of consumption during this period, only associated with events related to pit deposition.
70Comparison of herd ratios between Ramadi 4 and contemporary Tell Qraya (Simpson 1983: 490-499), located a few kilometers to the northwest, show significant differences in herd composition. At Tell Qraya the sheep to goat ratio is 1:1, while suids are completely absent. At Ramadi 4 the sheep to goat ratio is 3:1 and suids are the second most common taxon. It is difficult to specify the reasons for these variations in breeding choices or herd composition, both between the phases 4 and 5 at Ramadi itself, or between two neighboring contemporary sites (Ramadi 4 and Qraya). This leads us to suggest that the deposit within the pit was likely associated with a special event or occasion, that necessitated or inspired a substantially different subsistence or consumption practice.
71The great diversity of wild species represented at Ramadi reflects the diversity of the surrounding environment. The presence of the Mesopotamian fallow deer, wild boar and aurochs attests to the existence of wet and wooded areas near Tell Ramadi, while gazelle and hare are indicative of desert and steppe environments. Indeed, located on the edge of the Euphrates, Tell Ramadi is optimally positioned to bring together these different environmental contexts.
- 22 The excavators at Tell Humeida use the presence of clay sickles as an indication of Ubaid occupatio (...)
72Fragments of clay sickle blades (fig. 9: 17) were recovered in large numbers from pit II (n: 133). Their presence stood in sharp contrast to the Canaanean blades and prismatic core blades found in phase 5 contexts. Clay sickles are generally seen as a southern Mesopotamian adaptation to the scarcity of flint resources (Algaze 2001b: 207; 2018: 35), and stood as an isocrestic style indicative of the presence of groups from the south at a given locale (see Rothman 2002: 57; Stein 2002: 151). They have been recovered at neighbouring sites, such as Qraya (Reimer 1989: 284) and Tell Humeida22 (Fenollós passim 2011: 208; Fenollós and Al-Abedallah 2016: 248), within exclusively southern Uruk sequences. Despite their widespread distribution, little has been published about clay sickles. Use-wear studies from Tell el-‘Oueili suggest that clay sickles were used for harvesting cereals and processing tough stemmed plants like reeds, rushes and sedges, fronds and fruits (Benco 1992; Pollock et al. 1996). The existence of clay sickles alongside flint sickle blades in phase 5 at the site suggest that they were reserved for different tasks, however in phase 4 they seem to appropriate many of the tasks once reserved for flint sickles.
73In addition to the clay sickles, 21 spindle whorls were also recovered from pit II, 5 of these were fashioned out of ceramic sherds (fig. 9: 7). Additionally, fragments of four different ladles (fig. 9: 16) were also found. Parallels of these ladles could be seen at Uruk sites such as Chogha Mish (Delougaz and Kantor 1996: plate 82: H-L), and Qraya (Reimer 1988). Finally, two jar stoppers (fig. 9: 15), and a flint scraper were also found within the pit.
- 23 The discard signature attested in pit II seems to be the opposite of what Stein (2012b) recovered f (...)
- 24 According to Pollock (2003, 2012), and D’Anna (2012) the two activities were not mutually exclusive (...)
74When looking at the materials from the discrete depositional episodes in pit II an interesting pattern emerges. Each of the seven depositional episodes had a somewhat similar signature with fairly consistent frequencies in vessel categories (table 6). Each deposit had a high frequency of southern Mesopotamian vessels (92-97%), mainly BRBs (between 82-92%). This was consistently followed by southern Uruk storage jars (4-8%), and either northern Mesopotamian coarse bowls or southern Mesopotamian cooking pots (1-5%).23 The close clustering of vessels and the relative consistency in vessel frequencies in each deposit points to a similar short-lived event taking place during each depositional episode. Considering the overwhelming number of bowls found in the deposit (BRBs and CFW coarse bowls) the events all involved eating, and/or receiving a ration.24 In addition to the bowls, the consistent presence of animal bones and cooking pots also supports the notion that deposits were formed from discarded vessels from a relatively large feasting event. In this context the storage jars could have been brought containing ingredients and beverages (judging by the spouted vessels), and discarded along with the other items after the feast was complete.
Table 6 – Vessel frequencies of each depositional episode in pit II.
75Assuming that each deposit represented a coherent and contained event, the assemblage recovered from each of these episodes likely points to an independent feasting event. In terms of relative proportions, the first event (earliest) represented the smallest, while the last event was the largest, with almost twice the number of vessels recovered from the other events.
76The question that remains is who may have been involved in these feasting events? Judging by the ceramic assemblage composition, the majority of foods were brought, cooked, and subsequently consumed, in southern Mesopotamian vessels and only partially supplemented with northern Mesopotamian serving vessels. This could indicate that the sponsorship of the feasting events was under the patronage of a person or persons who, if not southern Mesopotamians, were highly integrated into a southern Mesopotamian habitus (Bourdieu 1977).
- 25 The lack of variation in depositional units within the pit, and the absence of thick ash deposits, (...)
77The materials recovered from the pits could hint at the seasonality of the depositional events. Wright et al. (1980: 274) argued that the presence of items such as spindle whorls and sickle blades in a discard pit likely indicated a late winter/early summer deposition event. Their assertion was made based on the fact that spinning and harvesting were activities that took place across this time frame when sheep were sheered and crops harvested. The use of items pertaining to such activities (spindle whorls for spinning and sickles for harvesting) would generate higher rates of discard as a result of damage. The even distribution of spindle whorls, and clay sickle blades across the seven episodes from pit II at Ramadi could indicate that these festivities took place in the spring in conjunction with harvesting of crops and wool.25 The clay sickles and the spindle whorls were perhaps part of a ritual deposition indicating the conclusion of labour as some of the spindle whorls were in pristine shape. It is difficult to gauge if each depositional episode was carried out as part of a seven-day celebration (more or less), or if they were deposited seasonally, representing a seven-year deposit.
78The faunal assemblage from pit II marks the appearance of a new domestic species at the site: the pig. Pigs represent a significant part of the assemblage during phase 4, while they were completely absent from phase 5. The high frequency of pigs, as well as the diversity of wild animals such as Mesopotamian deer, gazelle, hare, birds, and fish, and especially of aurochs, leads us to think that these remains are the result of a feast or a special event. Indeed, it has been suggested that aurochs held special status as symbols of strength and prestige during this period (Vila 1994).
79Based on ethnographic work (see Hayden 2014), we know that feasts tend to have a transformative effect intimately linked with social change (Dietler and Hayden 2001: 16-17). In essence, they provided a means to integrate people as they act on different levels to mediate relations within a community or a region (Dietler 2001: 69). At the same time, feasts could also fundamentally reinforce exclusivity (Dietler 2001: 85-86; Bray 2003: 9; Pollock 2003, 2013). The feasting event/s represented by the pit deposit seemed to have acted as an integrative action that may have facilitated various exchanges (Dietler 2001: 69). In this case the exchanges were perhaps a means of integrating labour (perhaps both northern and southern) into the southern Mesopotamian world. Judging by the labour tools and the cooking and serving vessels, the patrons of this feast seemed to have supplied the labour tools (clay sickle blades, and perhaps spindle whorls), rations, and perhaps a share of the produce to the inhabitants of Ramadi in exchange for either a portion of the harvest (and the wool) or for communally tending to their lands and flocks. This process may have begun during phase 5, when services were elicited from a few individuals, indicated by the appearance of BRBs at the site, and culminated in a full integration of the two worlds.
- 26 The appearance of pigs for the first time in pit II, in addition to remains of aurochs, was perhaps (...)
80Understanding the developments that took place between phases 5 and 4 at Ramadi is complicated. This is primarily due to the fact that we are comparing a fairly wide exposure from phase 5 with pit deposits from phase 4. This always raises the question whether the variation was related to actual changes across the site or whether they were simply governed by the nature of the contexts. Nevertheless, discrepancies between the two phases are too great to be attributed purely to the nature of the contexts alone, and must take into account changes in the ethnic makeup of the site’s inhabitants. Initially, during phase 5, the site was occupied by a group that used traditional northern Mesopotamian chaff faced wares. In the subsequent phase 4, the site shows two distinct assemblages that existed side-by-side; northern Mesopotamian chaff faced wares, and southern Mesopotamian Uruk wares. The differences extend far beyond that to other aspects of the material assemblage; the use of clay sickles instead of the commonly used flint sickle blades and the choices in dietary preferences.26
81Changes in feasting practices between phase 5 and 4 at Ramadi further support this notion. Initially during phase 5 we see the gradual introduction of BRBs into the otherwise northern Mesopotamian assemblage. In southern Mesopotamia, BRBs were usually associated with redistributive central institutions (Johnson 1973; Nissen 1988; Pollock 2012), functioning as disposable receptacles for either dry rations, ready to consume foods (Pollock 2012: 156), or leavened bread (Chazan and Lehner 1990; Goulder 2010; Potts 2009). At Ramadi, with the absence of a clear redistributive institution there is no reason to believe, so far, that they were used for redistribution by local centralized institutions at the site. In fact, in this part of the site, BRBs were absent from the public building. The presence of these vessels could imply that a number of individuals at Ramadi were in contact with the Uruk world, probably through nearby Uruk sites (e.g. Qraya). They could have found the expedient nature of producing these bowls attractive for their own purposes. Alternatively, a few members of the community may have engaged in economic relations by providing services to nearby southern Uruk institutions (herding flocks, farming, spinning, weaving, construction), thereby entering into the economic system of the south. This gradual coupling would have facilitated the increased presence of southern Mesopotamians at the site, observed during the phase 4 occupation.
- 27 The containers and the serving dishes were left in the complex and were not ritually discarded or d (...)
82This event signified the juxtaposition of a new, southern, system of patronage (rationing) on an already established local system, which was manifest in the feasting activities sponsored by actors affiliated with the public building. Feasting practices in the public building were in essence non-destructive,27 enclosed and exclusive, perhaps only involving heads of lineages, and their external exchange partners. The purpose of these gatherings was possibly to create alliances between equals who shared this limited space. Events at the public building perhaps included the larger community, yet maintained a clear spatial hierarchical separation (D’Anna 2012).
- 28 Gender task attribution is somewhat difficult and may present a risk, however, if spinning was a ge (...)
83This was followed in phase 4 by an apparent shift in the power relations. Here we see a multi-episode sponsored work feast. The multi-episode feasting event attested from pit II seemed to have been taking place within the context of crop harvest, wool collection, and spinning,28 evident from the discarded tools within the pit, perhaps as a symbolic gesture and a nod to labour. Under these circumstances the celebration of the labour force involved would have appeared as the focal point of the feast, while concealing true power relations. Judging by the make-up of the ceramic assemblage from the pit, the feast appeared to have been sponsored by either southern Mesopotamians or individuals strongly affiliated with southern Mesopotamian groups. If this were the case then the power dynamics of patronage, and by extension power at the site, was being shared if not usurped. Such a power shift without clear violence, may suggest that this shift happened gradually and did not lead to clear fractures at the site. We may argue here that this was achieved through the development of a hybrid practice, from the intersection of meal rationing and feast sponsorship, that was flexible enough, and mutually intelligible to all participants. The practices of feast sponsorship and ration redistribution blended into a hybrid institution that remained intelligible to the participants through the structures of patronage and hospitality.
- 29 Based on surveys (Simpson 1983; Pfälzner 1984; Bernbeck 1993; Geyer and Monchambert 2003; Lönnqvist (...)
- 30 As there seems to be more regional variation in ceramic assemblages during the post-Ubaid (Stein 20 (...)
- 31 The site of Anaiat al-Sharqi III (and to a lesser extent, Gilib al-Abud) along the Wadi Ajij produc (...)
84The traditional local/foreign interactions that characterize Algaze’s model of the Uruk expansion are not as clear here as in other regions. Based on surveys (Simpson 1983; Geyer and Monchambert 2003), and limited excavations, this stretch of the Euphrates from Jebel Abdul Aziz to Rawa does not seem to have a clear transition from the Ubaid to the Late Chalcolithic 2/3.29 Both Ramadi and Qraya, the two major sites that have been excavated, produced sequences that were comprised of Ubaid levels (Simpson 1983; Beyer 1991), followed by a hiatus before occupation resumed roughly around 3800-3700 BC.30 An explanation is that the region in general may have been nearly abandoned after the Ubaid period (roughly 4500 BC), and subsequently reoccupied by settlers arriving from both northern31 and southern Mesopotamia. Evidence of deteriorating climatic conditions starting roughly around 4500 and lasting until 4000 BC supports this notion (Clarke et al. 2016). The effects of dryer conditions would have been compounded in this region by low rainfall levels and high interannual variability (Geyer and Besançon 2002: 62), making the areas outside of the river valley even less habitable. If that were the case, then the inhabitants of Ramadi phase 5c would have been part of a new wave of settlers who were perhaps as foreign to the area as the Uruk settlers coming from southern Mesopotamia and settling at Qraya. This condition would have made this region a true frontier zone for both groups and an area where various practices could have mixed.
- 32 So far, we do not have any evidence of exotic goods being transported through Ramadi in bulk (one p (...)
- 33 In this stretch of the Euphrates we see both colonization (Qraya) and acculturation (Ramadi) discus (...)
- 34 In a similar manner Vallet et al. (Vallet et al. 2019: 183) state that the Uruk sites of Logardan a (...)
85Finally, stations were characterized as such based on their small size, relative isolation, and location along transit routes (Algaze 1989: 579). The initial work at Ramadi does not justify categorizing the site as a station. Beyond the locational affordance to act as a conduit for transporting goods there does not seem to be much evidence of the inhabitants engaging in regular long-distance exchanges. The inhabitants of Ramadi carefully exploited their diverse surroundings (steppe and river valley), and do not seem to be involved, as far as the evidence shows, in movement of exotic objects. That being said evidence from the site shows that there was a gradual shift from relative isolation to increased engagement with the Uruk world perhaps through contacts with the site of Qraya or other nearby sites. Qraya32 may have acted as link within the Uruk network, and may have played a role in pulling the inhabitants of Ramadi into the Uruk world, in a process of acculturation (Butterlin 2018: 410).33 In that regard, both Qraya and Ramadi were embedded within the social dynamics of the region,34 and played a more visible role in the transformation of their immediate surroundings. Accordingly, ‘stations’, as a category may suffice at the macroregional scale, but at the medial and micro scales the circumstances that guaranteed their continued operation was historically contingent and differed dramatically based on the surrounding landscape, and the group dynamics between host and colonial communities (Stein 1999; Vallet et al. 2019: 183).
Flux RSS
