Remote Sensing and Site Topography

Qach Rresh is remarkably well-suited for geophysical remote sensing investigation. Magnetic gradiometry survey was led by Creamer at Qach Rresh during 2021 and 2023, resulting in 8.1 hectares of land covered in total at the site. The east-west extent of the survey in these years was limited by harvesting schedules in adjacent fields, while the southern extent was limited by the seasonal drainage area which still held water during our fieldwork in 2023 (Fig. 2). Of the 203 20 x 20m squares surveyed in total at the site, 142 revealed architectural or otherwise anthropogenic features (a return rate of 70%).Footnote ²

Fig. 2. The results of magnetic gradiometry survey, 2021-2023. Data processed with destagger, Zero Mean Traverse, and despike transformations and reclassified to 0.125 x 0.125m resolution by nearest neighbor interpolation (basemap courtesy of ESRI)

Building A, identified in the magnetogram, is an apparently symmetrical structure measuring an estimated 41 x 38m built around a central courtyard with a possible single-axis entry in the southern end of the structure (Fig. 3). This entry opens into a subdivided chamber estimated to measure 19 x 12.5m, which then leads north into the courtyard (15 x 18m; see examples of similar architecture at Ziyaret Tepe [Matney et al. Reference Matney, Greenfield, Hartenberger, Jalbrzikowski, Köroğlu, MacGinnis, Marsh, Monroe, Rosenzweig, Sauer and Wicke2011]). Sixteen small rooms surround the courtyard area, possibly used for storage. To the northwest, another set of rooms is seen as part of a similar structure, Building B, where at least six rooms with estimated dimensions similar to those in Building A can be identified. While only a single arcade of these rooms (once again, possibly for storage) is observed in the magnetogram, it is possible that they formed part of a larger building complex, similar to that of Building A. The higher magnetic signature of Building B’s room fill is due to over a meter of refuse fill in these rooms, discussed further in the Excavation Results section below. Additional identified structures located between these two monumental buildings appear to be built on the same orientation as Buildings A and B, possibly indicating a coherent complex. Fainter signatures of walls alongside the east and west areas of Building A could indicate secondary sets of rooms which are more poorly preserved than those bordering the courtyard. Longer linear features running alongside building walls may be streets.

Fig. 3. Buildings A and B, with possible other architectural features marked. Yellow denotes potential architecture; light green denotes possible later circular enclosure features superimposed over earlier structures

We have detected a large complex of rooms further south, consisting of at least 12 rooms in a roughly linear arrangement, with additional architectural features present around them. The rooms making up this structure (which we are calling “Building C”) appear brightly in the magnetometry, which indicates that they are likely filled with the same dumping debris that Building B contains (Fig. 4). At least ten rooms have similarly high magnetic signatures, suggesting that they are filled with ashy debris and ceramics. Other similar highly magnetic anomalies in the south likely indicate more rooms filled with debris, although this is more difficult to tell from magnetometry alone.

Fig. 4. Building C and associated structures, along with Features 1 and 2 marked. Yellow denotes potential architecture, red denotes enclosure features, and light green denotes possible later enclosure features superimposed over earlier structures

Of importance to our investigations is the presence of at least two large structures (Features 1 and 2, measuring 120 x 50m and 60 x 30m, respectively) just north of Building C and south of Building A. These are the magnetic signatures of anthropogenic enclosures, possibly large animal pens or open-air walled areas where iron-rich dung accumulated along the fence line and created a higher magnetic signature than the surrounding soil. Another possible enclosure is located directly east of these two, but the magnetic signature is faint. Additionally, between Building C and Feature 1, there is a possible gate or entrance measuring roughly four meters wide. Thus, this complex is evidently associated with the features. The assumed association is further observed by the presence of a faint architectural structure directly to the west of Feature 1 and to the south of Feature 2.

Other non-architectural features were noted roughly 70m to the north of Building B. A cluster of these structures, each measuring roughly 20–40 m across, can be seen in Fig 5.Footnote ³ When compared to CORONA imagery of the area (Fig. 5c), it becomes evident that these are almost certainly the remains of qanat/karez construction. This identification aligns with the ceramics collected from the surface by RLIIM in this part of the site during seasons 2022–2023, which are largely Islamic in date.

Fig. 5. Identifiable collapsed qanat features in the northern portion of the magnetogram (B, in orange), with CORONA satellite imagery underlaid showing the relict path of this system (Mission 1039; Feb 1967). C shows regional qanat systems surrounding Qach Rresh, with intersecting systems identified with arrows

The site of Qach Rresh is largely flat, with the only notable topography occurring in relation to the area of Building A. This area contains a shallow sloping mound of roughly 1.5 m above the surrounding flat landscape. Any notable features which may have once been prominent in the area (such as the qanats) have likely been plowed under by the modern agricultural industry pervading in the region (Fig. 6).

Fig. 6. UAV oblique image of Qach Rresh during the 2023 season, looking south. (Op. C was not yet opened at the time of photograph)

Excavation Results

Excavations at Qach Rresh began in September 2022, with Operations A and B targeting Buildings A and B, respectively, to test their relationship to the magnetometry data and their preservation. In the second season, beginning in May 2023, Operation C was added, to investigate Building C following its identification in the extended magnetometry coverage of that season (see Fig. 7 for trench placement).

Fig. 7. Location of excavation trenches at Qach Rresh, 2022-2023

Operation A

In 2022, Operation A was opened over the northwestern corner of Building A, to investigate the potential storerooms and courtyard in relation to the slight topographic mounding in this area. Excavations were carried out in both 2022 and 2023, with Op. A. measuring 10 x 6m by the end of the 2023 season. Topsoil removal showed heavy modern plow scars affecting the remaining features directly below ground, to a depth of at least 30 cm. Two distinct occupation phases have been identified in Operation A: the building’s construction and operation during the late Neo-Assyrian/ Iron Age II period (Phase 2) and a later occupation uncovered in the courtyard, possibly dating to the post-Assyrian/ Iron Age III period and following an apparent abandonment of Building A (Phase 1).

Phase 1:

Phase 1 consists of an ephemeral occupation phase roughly 90 cm below the surface within Building A, within the northwest corner of the courtyard and postdating the building’s main use. This phase of use followed the abandonment of the building, apparently leveling and then settling upon the mudbrick collapse from the end of the previous phase, which formed 20–35 cm of rubble above the floor of Phase 2 and below the living surface of Phase 1. While the walls of Building A were difficult to identify due to mudbrick degradation,Footnote ⁴ we were able to trace the northern wall of the courtyard (A-020) running roughly NW-SE, based on a pit belonging to Phase 1 that had been cut into the mudbrick collapse fill of the courtyard’s northwestern corner (Fig. 8). The pit itself (A-006) was filled with ash, charcoal and animal bone. The northern edge of the pit ran parallel to Wall A-020’s south face, confirming the wall’s southern extent.

Fig. 8. Phase 1 (A-006) in Operation A. Walls A-020, 021, 025, and 026 (estimated extents) were constructed during Phase 2 but were still standing in some state of preservation during the ephemeral occupation during Phase 1

Phase 2:

Phase 2 is the main use phase of Building A, which has been dated to the later Neo-Assyrian period (Iron Age II), immediately following the building’s construction. Op. A extends over the northwestern corner of the courtyard (Room A1) and at least one adjacent storage room (A2), with another two possible storage rooms identified in the northeastern corner of the trench (Fig. 9). The floor level of this phase (A-024) was identified and traced in Room A2 at a depth of 114–118 cm below the surface. The floor itself was smooth but irregular, marked by a bluish-white smoothed mud plaster layer roughly 2 cm thick. A similar type of irregular clay flooring was noted by the excavators in multiple phases within the northwest corner (Area S6) of the external fortified enclosure surrounding Fort Shalmaneser at Nimrud (ancient Kalḫu; Lombardi Reference Lombardi2015). Sediment samples were collected from this context (A-024) and from the fill of Room A2 (A-022) for both macro and micro-botanical analysis (see Preliminary Botanical Analysis section below).

Fig. 9. Phase 2 in Building A

Although the mudbricks of wall A-020 were uniformly between 33 x 33 x 15cm to 35 x 35 x 15cm, the original builders of this wall did not neatly align each wall course. Between the three exposed courses of brick at various points of wall A-020, each course was oriented at a slightly different angle. Moreover, the makers of these mudbricks included a range of additional materials, such as small pebbles and medium-sized rocks, ceramic sherds, fragments of baked brick, and faunal remains, to create individual mud bricks. From the generally poor structural integrity of the bricks, we theorize that the walls in both buildings were constructed from mudbricks that had not yet fully dried. Furthermore, the many large inclusions (some measuring over 10 x 10cm) in the bricks made them of poor structural quality.

As preserved, wall A-020 is roughly seven rows wide (roughly 2.7m) and four courses high. The thickness of this wall (and the assumed comparable thickness of other walls when compared in the magnetogram) leads us to believe that they supported a second level. At about 80 cm below the surface, we were able to confidently identify wall A-021 running SW-NE, bonded at its northern end with wall A-020. This wall, although only currently preserved about 1.5 courses high, was visible in the W baulk of the trench to a height of six courses.

In general, the southern baulk of the trench was heavily disturbed by root activity. At a depth of about 45 cm, we identified the upper preserved extent of both the southern wall (A-025) and the eastern wall (A-026) of storeroom A2.Footnote ⁵

Compared to the diversity of room fill in neighboring Building B (see below), storeroom A2 largely contained a mix of animal bones, pottery sherds, and ceramic slag. However, in the upper layer of this fill we found what appears to be a clay administrative token akin to those found at other provincial Neo-Assyrian sites, e.g., Ziyaret Tepe (ancient Tušhan; Matney et al. Reference Matney, Greenfield, Hartenberger, Jalbrzikowski, Köroğlu, MacGinnis, Marsh, Monroe, Rosenzweig, Sauer and Wicke2011) and Tell Ta’yinat (Snow & Batiuk forthcoming). Measuring 27 x 20mm, this token resembles the ‘conical’ type found at Ziyaret Tepe (Fig. 10). Tokens may have functioned as computing aids to tally large sums and keep track of evolving amounts of commodities such as grain or livestock, acting as an informal ‘accounting ledger’ for local administration (MacGinnis et al. Reference MacGinnis, Monroe, Wicke and Matney2014; Monroe Reference Monroe, MacGinnis, Wicke and Greenfield2016). Although this object remains a single find, it is hoped that further excavation in neighboring storerooms will clarify its relationship to Building A.

Fig. 10. On left, possible conical token A016.157. On right, the range of token types found at Ziyaret Tepe, after Matney et al. Reference Matney, Greenfield, Hartenberger, Jalbrzikowski, Köroğlu, MacGinnis, Marsh, Monroe, Rosenzweig, Sauer and Wicke2011: fig. 13c

In the NE corner of Op. A, we found the beginning of two other rooms north of wall A-020 that are possibly bounded on their eastern/western sides by a wall abutting A-020 and running SW-NE. We were not able to establish the extent of these rooms, and this area will be a priority for the following field seasons.

Operation B

Operation B was opened in 2022 as a trench of 10 x 10m, then expanded in 2023 to 16 x 11m. Like Op. A, the area was affected by plow scars up to 30 cm deep underneath the topsoil and mudbrick architecture was difficult to distinguish from the surrounding soil matrix. Immediately, we noticed a proliferation of broken baked bricks within the topsoil, intermixed with river cobbles and sherds. To preserve stratigraphic levels in the baulk for reference and sediment sampling, we maintained a modified Wheeler’s Box Grid trench organization, splitting Op. B into four separate quadrants (the SE and SW quadrants were eventually combined; see Fig. 11).

Fig. 11. The layout of Operation B in 2022-2023

Late wall remains

Roughly ten cm below the surface we discovered two single-course linear arrangements of river cobbles, both arranged in right-angles and adjacent to one another, covering an area of nearly 1.5 x 1.7m (B-004, Fig. 12). The cobbles measured roughly 10–30cm each in length.Footnote ⁶ We theorize that these may be the remnants of agropastoral activity in a later period, such as pens for livestock or field boundaries. Surface ceramic scatters, along with the evidence for a relict qanat system through this area (see Remote Sensing and Topography section above), indicates that an Islamic date of construction and use may be likely.

Fig. 12. The remains of feature B-004 in Operation B – likely Islamic period walls. North indicated by trowel in inset photograph

Post-Assyrian/ Iron Age III refuse deposits within Building B

Building B, while evidently built at the same time as Building A according to its orientation and construction, exhibited unexpected differences in use. In the eastern half of the SW quadrant during 2022, we encountered large amounts of sherds and animal bones beginning at a depth of 35 cm in what we identify (consulting the magnetic gradiometry data) as room fill of Room B1 (Fig. 13). Based on soil color and texture, we were able to estimate the orientation and thickness of the mudbrick wall (B-080) bounding the room on its western side. We identified the extent of this wall before continuing to excavate the room’s fill in a deep sounding in the northeastern corner of the SW quadrant, where we reached what we believe to be Building B’s original floor at a depth of 1.6m. Below the floor may be a possible foundation trench filled with mudbrick constructed to add further support to the building, though this identification remains tentative (Fig. 14). The eastern (B-082) and northern (B-081) walls of Room B1 were then identified and traced in the extensions of the trench in the 2023 season (see Fig. 13). The southern wall (B-083) was more difficult to discern, which we later realized was due to the wall’s poorer preservation; while walls B-080, B-081, and B-082 were all preserved up to a height of 1.3 meters or more, southern wall B-083 was only preserved to a height of 80cm and was obscured under fill. It is unclear at this time why there is such a difference in preservation. Once these walls were identified, we were able to measure the width of the walls as between 2.5–2.7m thick, comparable to the walls in Op. A. Likewise, mudbrick size varied between 32–35cm in width and length. The final size of Room B1 measures roughly 5.5 x 4.8m. We expect that Building B, for the same reasons as Building A, likely had at least two stories.

Fig. 13. The extent of Room B1 in Operation B, with partial Room B2 shown

Fig. 14. The northern baulk of Op. B’s deep sounding, giving the most comprehensive stratigraphy of deposits in Room B1

While we initially suspected that the rooms of Building B may have experienced burning in antiquity due to their higher relative magnetism expressed in the magnetic gradiometry data, it became clear during the excavation of Room B1 that this signature was instead caused by the deposition within the rooms of multiple strata of refuse consisting of ceramics, faunal remains, baked brick and mud brick, miscellaneous artifacts, and ash. This was further clarified during the excavation of the deep sounding in the northeastern corner of Op B.’s SW quadrant, in which we were able to identify several discrete deposition events of such refuse (Fig. 14).

Sediment samples from each stratigraphic unit of the fill of Room B1 were collected for flotation. Sediment samples were taken from the northern baulk of the sounding to further analyze the refuse layers. In total, a maximum depth of 1.1 m of refuse was deposited in Room B1. This trash was apparently dumped there after the collapse of Building B’s upper floor and roof, as indicated by the 55 cm of mudbrick collapse (at its thickest point) below the refuse strata and above the floor. The makeup of the refuse deposits varied, with some comprised of an ashy matrix with few artifacts, while others were composed nearly entirely of sherds, faunal remains, and broken household implements (such as various spindle whorls, bronze and iron pins, a bronze bracelet, part of a figurine, and possibly implements used for textile manufacture, discussed below in Small Finds). Preliminary analysis of ceramics recovered from these trash deposit strata indicates a clear post-Assyrian/ Iron Age III date (c. 600–350 B.C.E.; see Preliminary Ceramic Analysis section below). The general types of artifacts present in the refuse strata seem to indicate a domestic source, rather than industrial production.

While it is difficult to date Building B’s abandonment relative to its secondary use as a dumping ground, we propose that the preservation of Room B1’s walls up to a height of 1.3m might indicate that it became a dump shortly after the collapse of its upper floor and roof. If Building B’s abandonment was concurrent with Building A’s, which seems likely, then some of the refuse deposited in Building B could be related to Building A’s Phase 1 occupation in the post-Assyrian (Iron Age III) period. It is unlikely, however, that the massive amount of refuse which fills Room B1, along with the refuse which we now know to be present in the other rooms of Building B, would be associated with the ephemeral occupation of Op. A’s Phase 1 alone, and instead likely spans beyond.

Small finds in Op. B

The small finds excavated thus far in Room B1 attest to domestic lifeways at Qach Rresh, including the local manufacture of textiles as well as human-animal interactions. Found within multiple levels of dumping strata were a series of clay spindle whorls (one complete, three partial; Fig. 15) and clay loom weights (two complete, three partial). Two of these spindle whorls were decorated along their outer edges with a series of incised lines. One (B067.351) with an internal perforation of 2.5 mm and weight of 8 g was likely used to spin finer threads (cf. Sauvage Reference Sauvage, Harlow, Michel and Nosch2014). However, the heavier weights of the remaining spindle whorls (16–38g) and loom weights (≥40 g) suggest that thicker threads were more typical of textiles produced at the site. The loom weights are of the ‘doughnut’-shaped type widely characteristic of Iron Age sites in Southwest Asia. Leather-working may have also been practiced at Qach Rresh. A ceramic disk (B062.324; Fig. 15, right) with edges that have been fashioned into ten points may have been used to mark holes in animal hide to guide the path of the awl, similar to modern methods of shoe and saddle-making using wheel prickers (Salaman Reference Salaman1986). The surfaces of the disk had been smoothed, perhaps by manual use over time.Footnote ⁷

Fig. 15. From left to right: decorated spindle whorls B072.337 and B003.427, and ceramic disk B062.324

The diversity and types of faunal remains (see below in Preliminary Faunal Analysis) indicate that the inhabitants of Qach Rresh were closely engaged in livestock maintenance. In particular, the presence of equid bones in post-Assyrian contexts may relate to the use of horses in animal husbandry and managing livestock, a practice continued on the Erbil Plain to this day. Ephemeral remains of iron rings and rods may have once constituted parts of horse trappings, albeit less complex in design than those found in elite buildings of the Neo-Assyrian period (Curtis Reference Curtis2013: pls. LXXX–LXXXI). Two fragmentary clay animal figurines found in Room B1, possibly of horses, resemble those found in domestic contexts and throughout at Assur (Klengel-Brandt and Onasch Reference Klengel-Brandt and Onasch2020). Conversely, the presence of iron arrowheads in both Room B1 and in Building C (see below) may attest to opportunistic hunting events; trace evidence of wild game was also found at the site.

Two finds bear further mention: one, a glass bead, (H: 8mm; Dm: 11mm) was likely originally translucent (B062.416). Its surface is heavily weathered and displays evidence of severe pitting, and the corrosion from the surrounding soils have given the bead an iridescent appearance (Schmidt Reference Schmidt2019:13). The second is an almost-complete ceramic pipe-lamp (B060.279; Fig. 16), its fragmentary spout stained black by sustained use. Its closed, globular form would have increased the overall burn time of a single fueling, as opposed to open form, ‘saucer’ types (Kelsall Reference Kelsall2018). Although it resembles closed-form types dated to the seventh century B.C.E. from Fort Shalmaneser (BM 1992,0302.181) and palaces on the Nimrud citadel (BM 92973), it does not have the rolled rim typical of these examples (Oates Reference Oates1959). Its flared rim instead may relate to Neo-Elamite examples, of which a double-spouted example is known from Susa (SB 5329; Louvre).

Fig. 16. Ceramic pipe-lamp B060.279

Operation C

Operation C was opened during the 2023 season with the intention to investigate a small area of one of the rooms of the structure designated as Building C to the south of Operations A and B. A test trench of 2 x 4m was excavated to identify the contents and preservation of one of the rooms identified in the magnetometry (see Fig. 4). A depth of 1.5 m was reached before the excavations of 2023 were closed at the end of the season.

The room into which Operation C extends, as its magnetic signature indicated, was used as a trash deposition area after its main use, like Building B. While we have not yet reached below the trash deposits, we predict that Building C, like Building A, was constructed during the height of the Neo-Assyrian imperial period. Operation C extends into the center of the room while uncovering part of a single wall that runs NNE-SSW and protrudes roughly one meter into the trench’s eastern side; several of its mudbricks were able to be defined (Fig. 17). This is likely a dividing wall between rooms, as can be seen in Fig 4. The bricks measured roughly 34 x 34 cm, which is equivalent to those of Buildings A and B.

Fig. 17. Operation C at the end of season 2023. Note the sloping ashy strata in north baulk, which contains the majority of refuse room fill. (Cut in top left of baulk for trench entry steps)

There were several layers of ash deposit and mudbrick collapse, followed by refuse deposition, which yielded a high number of sherds and faunal remains. Although the excavations only reached the upper limits of fill within this room, small finds suggest a general correlation with those present in the upper fill of Room B1 in Building B. Two spindle whorls (one complete, one fragmentary), one loom weight, a metal arrowhead (Fig. 18) and the spout of a metal vessel were found in the accumulated fill about 40 cm above the top of the preserved wall.

Fig. 18. Metal Arrowhead from Op. C fill contexts (C001.342)

Unlike the trash deposition strata in Operation B, which can be quite thin (possibly indicating individual household trash deposits), the ashy strata and refuse strata in Op. C were thicker (roughly 10–20cm thick on average) and might indicate that larger amounts of refuse were deposited in the rooms of Building C at one time. All refuse strata slope down from the preserved wall, beginning at 68 cm below the surface, which may indicate that this wall had already deteriorated to that height by the time the rooms were used as dumping sites. Preliminary ceramic analysis of the sherds recovered from Building C’s fill indicate a potentially later post-Assyrian date (c. 600–350 B.C.E.; see Preliminary Ceramic Analysis section below).

Preliminary Ceramic Analysis

During seasons 2022 and 2023, the RLIIM project excavated 25,473 sherds, with analysis led by Kaercher. In Operation A we recovered 2,245 sherds, Operation B had 18,473 sherds (almost entirely from refuse contexts), and Operation C had 1,719 sherds. The forms of the ceramics do not vary much between the operations, with most being classified as large jars. Below is a preliminary presentation of our results and associated dating.

In terms of surface color, buff wares range from white (5Y 8/1 white), to yellow (10YR 7/6 yellow), to brown (7.5YR 5/4 brown). Red wares range from pink (5YR 7/4 pink) to red (2.5YR 6/6 light red). Gray wares are usually gley 1 2/2 light greenish gray to 10YR 3/1 very dark gray. Green wares are from 2.5Y 7/2 pale yellow to 5Y 8/4 pale yellow. Vegetal inclusions could be chaff, straw, or other grasses and are visible as elongated pseudomorphs. Mineral inclusions are sand to grit sized particles in black and white, and in Op. C, rarely red. Most ceramics seem to be wheel made, some with wheel markings, especially visible on the bases. Interestingly, in Ops. B and C, blocky coils start to appear. These are relatively low fired and tend to crumble in the field, but some small square-shaped coil fragments remain. When discovered in situ, the blocky coil sherds are very large, possibly indicating pithoi for storing large quantities of materials.

In terms of forms, generally the divisions are as follows: shallow bowls (open vessels with a rim angle of less than 30°); deep bowls (open vessels with a rim angle of between 30° and 90°); cups/goblets (usually carinated with a flaring rim, 8–15 cm in diameter, typically palace or very fine ware); jars (rim diameter less than the shoulder diameter, with short or tall necks); holemouth jars (closed vessels).

The ceramics from Op. A (Fig. 19) range in surface color through buff (1,234; 56%), red (427; 17%), gray (460; 20%), and green (114; 8%). They mainly have vegetal inclusions (1,713; 76%) followed by mineral inclusions (256; 11%), vegetal/mineral (146; 7%), and no inclusions (131; 6%)Footnote ⁸ . Surface treatments range from smoothed (2,097; 93%) to slipped (white or buff slipped on red ware; 95; 4%), incised (28; 2%), and appliqué (5; 1%). Most were body sherds (2018; 90%), followed by rims (171; 8%), bases (51; 1%), handles (4), one spout, and one full profile of a bowl. The rims that were able to be drawn (51) were categorized as shallow bowls (1; 2%), deep bowls (20; 34%), jars (33; 56%), and holemouth jars (5; 8%). For this preliminary study, very few comparanda were identified for this operation, and these were dated to the Iron II–III period, as established by S. Anastasio (Reference Anastasio2010). Figure 19 no. 12 matches a vessel from Nimrud (Oates Reference Oates1959: XXXV:3). Figure 19 no. 21 matches vessels from Tell Rimah (Postgate et al. Reference Postgate, Oates and Oates1997: pl. 56:493) and Tell Shalgiyah (Green Reference Green, Hausleiter and Riche1999: fig. 6:9).

Fig. 19. A selection of ceramics from Op. A

The ceramics from Op. B (Fig. 20) range in surface color between buff (11,944; 64%), red (4,821; 26%), green (955; 6%), and gray (755; 4%). They mainly have mineral inclusions (7,774; 43%), vegetal inclusions (3,988; 23%), no inclusions (3,524; 20%), and vegetal/mineral (2,443; 14%). The surface treatments are mainly smoothed (16,490; 89%), slipped (1,599; 9%), burnished (223; 1%), incised (95), appliqué (65), and one painted sherd dating to the Halaf period (possibly from inclusion in mud bricks). Most sherds were body sherds (14,143; 77%), followed by coils (2,609; 14%), rims (1,292; 7%), bases (360; 2%), handles (42), spouts (9), lamps (5), worked sherds (4), full vessels (3), and a jar stopper. The forms that were recorded so far (as we are still processing some sherds) are divided into shallow bowls (5; 1%), deep bowls (69; 15%), carinated cups/goblets (31; 6%), jars (312; 63%), holemouth jars (68; 14%), and mini-jars (7; 1%). For this preliminary study, few comparanda have thus far been identified for this operation also. The few that were date to the Iron II–III chronology as established by Anastasio (Reference Anastasio2010), and into the fifth–fourth century B.C.E. (Rocco Palermo pers. comm.). Figure 20 no. 1 matches vessels from Nimrud (Oates Reference Oates1959: pl. XXXVI: 28) and Tall Knedig (Kulerman-Ossen and Martin Reference Kulermann-Ossen, Martin, Klengel-Brandt, Kulermann-Ossen and Martin2005: Taf. 113c). Figure 20 no. 3 matches a vessel from Sharqat (Miglus Reference Miglus2000: Abb. 29b). Figure 20 no. 4 matches vessels from Tall Sheikh Hamad (Kreppner Reference Kreppner2006: Taf. 11, 4) and the LoNAP project (Gavagnin et al. Reference Gavagnin, Iamoni and Palermo2016: fig. 18: 12-13). Figure 20 nos. 7 and 10 match vessels from Khirbet Khatuniyeh (Curtis and Green Reference Curtis and Green1997: fig. 35: 109 and fig. 37: 156). Figure 20 no. 8 matches a vessel from Khirbet Qasrij (Curtis Reference Curtis1989: fig. 31:140). Lastly, Figure 20 nos. 12 and 20 are identified as fifth–fourth century B.C.E. by R. Palermo (pers. comm.).

Fig. 20. A selection of ceramics from Op. B

The ceramics from Op. C (Fig. 21) range in surface color between buff (1,138; 63%), red (468; 28%), gray (86; 5%), and green (69; 4%). They mainly have mineral inclusions (945; 55%), followed by no inclusions (354; 18%), vegetal/mineral inclusions (336; 20%), and vegetal (126; 7%). Surface treatments range between smoothed (1,610; 90%), slipped (125; 8%), incised (12; 1%) burnished (8; 1%), appliqué (3), and painted (3). Most sherds were body sherds (1,359; 77%), followed by coiled sherds (269; 15%), rims (95; 6%), bases (29; 2%), handles (7), and one spout. The rims that were drawn are divided into deep bowls (3; 5%), carinated cups/goblets (4; 6%), jars (52; 79%), and holemouth jars (7; 10%). For this preliminary study, few comparanda have been identified for this operation either; these date to the Iron II–III chronology as established by Anastasio (Reference Anastasio2010). Figure 21 nos. 3, 6, 10, and 12 match those from Khirbet Qasrij (Curtis Reference Curtis1989: fig. 43: 327, fig. 25: 50, fig. 32: 162, fig. 37: 228, respectively). Figure 21 nos. 14 and 19 match Khirbet Khatuniyeh (Curtis and Green Reference Curtis and Green1997: fig. 57: 379 and fig. 64: 409, respectively).

Fig. 21. A selection of ceramics from Op. C

The ceramics from Op. A match the Iron II to Iron II–III chronology (800–500 B.C.E.) or Neo-Assyrian to post-Assyrian periods. The ceramics from Op. B and Op. C fall mainly into the Iron III chronology (60–400 B.C.E.) with limited Neo-Assyrian forms. In 2023, further diagnostic ceramics dating from the fifth–fourth century B.C.E. were recovered. With the brief analysis conducted in the field, there are confirmed differences in the occupation of Op. A and Ops. B/C. This could be due to different activities occurring in these operations, a slight difference in dating (the majority of Op. A’s contexts being earlier than B, hence the different tempers), or just variations in the rural nature of this site. Ops. B and C have been confirmed as refuse dumps for later period habitation.

Preliminary Faunal Analysis

The 2023 RLIIM expedition at Qach Rresh included a zooarchaeological program overseen by Poolman, which included the identification and analysis of animal bones from archaeological contexts excavated in both 2022 and 2023. The analysis was carried out with the intention of understanding consumption and economic practices and the ways these shaped the communal lives of humans and animals during the site’s post-Assyrian period.Footnote ⁹ A total of 1,707 specimens were identified to any level of taxonomic specificity, with a further 896 unidentified specimens counted and weighed, from 31 unique contexts of 2022 and 2023’s excavations of Operations A, B, and C. The vast majority of remains identified (1,592 NSP) came from Op. B, followed by Op. A (37 NSP), and Op. C (77 NSP). This analysis will only concern those remains from contexts with secure chronological assignments. Because the sample sizes of faunal specimens from chronologically secure contexts in Op A. (n=20) and Op. C (n=4) were very small, they are excluded from the present analyses and will be explored in future reports.

The taxonomic composition of identified faunal materials from post-Assyrian fill levels in Operation B can be found in Table 1 and Figure 22. Livestock taxa predominate our sample with caprine (Ovis/Capra) specimens most numerous (38% NISP, n=123), followed by pigs (Sus scrofa domesticus, 32% NISP, n=104) and then cattle (Bos taurus, 15% NISP, n=50). These animals formed most of the dietary meat sources at post-Assyrian Qach Rresh. The relatively robust representation of pig remains is notable, as cattle (our third most numerous taxon) often follow caprine in number in the faunal collections of many Iron Age Southwest Asian contexts (Poolman et al. forthcoming).

Table 1: Taxonomic composition of identified faunal materials from post-Assyrian fill levels in Operation B

Fig. 22. Relative frequency of major species from post-Assyrian fill-levels in Operation B (SG = Sheep/Goats; O_Domest = Other Domesticated animals)

Post-mortem utilization of Qach Rresh’s three major livestock taxa can be assessed from the relative representation of different skeletal elements, presented in Figure 23. Caprine and cattle specimen pools are similar in that they are dominated by head elements (primarily horncores and mandibles), with a relative dearth of post-cranial remains that are most often associated with consumption waste. This suggests that the meatiest cuts of caprine and cattle remains were not consumed and disposed of in Op. B, but elsewhere, either on- or off-site. In contrast, pig remains show a relatively more even representation of body parts; while head elements also predominate, they are followed closely by axial and upper limb elements that bear the most meat of the animal’s body, as well as the lower-limb butchery byproducts of these animals. This suggests that pigs were likely kept alive, butchered, and consumed near Op. B’s premises before being disposed of in the fill layers of this building.

Fig. 23. Relative representation of different skeletal elements

Other domestic/commensal taxa identified include equids (Equus spp., n=31, 10% NISP), canids (Canis spp., n=2, <1% NISP), snails (Mollusca spp., n=2, <1% NISP) and rats (Rattus spp., n=2, <1% NISP). Equids, which include donkeys (Equus asinus), mules (Equus mulus), horses (Equus caballus), and wild onagers (Equus hemionus), were most likely kept on-site for labour, travel, and animal management purposes in the case of domestic taxa, while onagers, widespread across steppe and valley lands of Iron Age Upper Mesopotamia, would have constituted a wild prey animal. Domesticated canids perhaps aided in the herding of livestock alongside humans and their equids. Two equid limb elements bore cutmarks, suggesting that these animals were butchered for their hide or meat following their deaths. The presence of rats and snails, meanwhile, are common agents of decomposition for anthropogenic rubbish containing large amounts of moist, organic debris.

Only a small number of faunal specimens could be reliably assigned to wild taxa (n=11, 3% NISP), primarily consisting of deer (Cervid spp.), gazelle (Gazella gazella), and avian specimens. Curiously, two elements were assigned to red deer (Cervus elaphus), including the distal portion of a tibia and a section of mandible (Lister Reference Lister1996). Red deer are typically relegated to high-altitude and heavily wooded environments of Southwest Asia (Geist Reference Geist1998; Straus Reference Straus1981), so the appearance of this large-bodied game taxon at a relatively small site on the Erbil Plain is admittedly unexpected and may reflect seasonal migratory patterns of ancient red deer populations and opportunistic hunting on the part of Qach Rresh’s Iron Age inhabitants (Zhang et al. Reference Zhang, Liu and Teng2013).

Overall, Op. B’s fill layers provide the following impressions of the faunal landscape at post-Assyrian Qach Rresh. Caprines and pigs were most often consumed on site, with caprine meaty elements often transferred elsewhere for consumption, while pigs were kept, consumed, and disposed whole by Qach Rresh’s inhabitants. Cattle appear less frequently and are also primarily represented by head and lower limb elements, suggesting that their consumption was incidental among ancient occupants. The residents of post-Assyrian Qach Rresh likely lived alongside a handful of equids and dogs who helped with some of the agropastoral duties of this rural site, while also occasionally hunting itinerant groups of artiodactyl game like gazelle, red deer, and/or onager that would seasonally make their way through the Erbil Plain. Our assemblage paints the picture of a relatively self-sustaining faunal community, with humans, equids, and dogs participating in the production of caprine and cattle meat for off-site provisioning (e.g., market exchange, institutional provisioning, etc.), as well as more localized human consumption of caprine byproducts, occasional game meat, and pork before disposal of material by local biotic agents (i.e., dogs, rats, and snails).

Preliminary Botanical Analysis

A combined programme of micro- and macro-botanical analysis was initiated at Qach Rresh during the 2022–2023 field campaigns by Laugier and Proctor. This programme has the twin goals of reconstructing the local paleoecology of the Erbil Plain and local agricultural decision-making strategies during the first millennium B.C.E. Analyses of macro-botanical remains of crops, processing debris, and wild/weedy seeds can be used to examine a wide range of questions regarding agropastoralism, including agricultural decision-making, labour organization, social (in)equality, fuel use, and grazing/foddering strategies, among others (Marston and Castellano Reference Marston, Castellano, Steadman and McMahon2021; Miller Reference Miller1997; Spengler Reference Spengler2019). To that end, flotation samples were collected from primary and secondary occupational deposits, including floors and ash lenses, in Op. A and from secondary refuse deposits in Op. B. Flotation samples were processed via bucket flotation for the initial campaigns, while a water-recycling flotation system will be implemented in future seasons. Macro-botanical analysis is currently in progress, and the results will be presented in a future report.

Discussion: Assyrian Administration and Post-Collapse Practice

With the data from the 2022-2023 seasons, the site of Qach Rresh has already contributed significantly to the discussion surrounding Assyrian presence in the imperial heartland beyond its urban and mega-urban centers. Current debates revolve around the nature of Assyrian urbanism and its sudden expansion of settlement into the interfluvial countryside – notable, as it is the first time such areas were occupied to any significant degree. Yet, little archaeological evidence for this phenomenon exists from current excavations, with the only studies of rural communities in the Assyrian heartland stemming from large-scale surveys (Morandi Bonacossi Reference Morandi Bonacossi, Düring and Stek2018; Ur et al. Reference Ur, Babakr, Palermo, Creamer, Souroush, Ramand and Nováček2021).Footnote ¹¹ These survey projects show an explosion in the number of small-scale sites first appearing during the early Iron Age. In addition to settlements of the Bronze Age remaining largely occupied on the Erbil Plain, there is the new appearance of evenly spaced small villages characterizing the Neo-Assyrian landscape, with the overall number of sites increasing by 36% from the Late Bronze Age to the Iron Age, representing an increase greater than any other period transition yet seen (Ur et al. Reference Ur, Babakr, Palermo, Creamer, Souroush, Ramand and Nováček2021).

What remains to be answered is the degree to which this settlement system of the Assyrian Empire was planned – that is, imposed by top-down administration efforts. In the first millennium B.C.E., the average size of previously-occupied Neo-Assyrian sites remained relatively stable between regions, while the number of settlements expanded but the total settled area did not. This indicates a growing rural population overall, without an accompanying depopulation of urban centers (Altaweel Reference Altaweel2006, Reference Altaweel2007; Ur et al. Reference Ur, Babakr, Palermo, Creamer, Souroush, Ramand and Nováček2021; Ur and Osborne Reference Ur, Osborne, MacGinnis, Wicke and Greenfield2016; Wilkinson et al. Reference Wilkinson, Wilkinson, Ur and Altaweel2005). If increased agricultural production was the end goal of this new settlement system, then this can be convincingly aligned with Assyrian efforts to organize the landscape both on a physical and an ideological level, where a king’s control over the “four corners” or, the known world, depended on producing civilization from wilderness, order from chaos, or developing non-Assyrian areas into Assyrian-governed ones (Liverani Reference Liverani2017: 41-54).

Qach Rresh’s place within this shifting settlement system remains only preliminarily understood. While we had known initially that this site was home to what were apparently large administrative buildings, from the magnetic gradiometry investigations conducted in 2021, the results of the 2022 and 2023 seasons have revealed that Qach Rresh was far more expansive than initially thought, with the discovery of Building C and Features 1 and 2 revealing a higher capacity for storage and production than Buildings A and B alone could provide. Potentially, noting its adjacency to the larger Assyrian center of Trpa Spiyan 2 km to the south (under the modern-day village)Footnote ¹² and the relatively short distance to Erbil, it is not a stretch to imagine that Qach Rresh could have operated as a strategic depot for agricultural and animal products, possibly as a site for staging such products ultimately destined to be consumed by Erbil’s institutions and inhabitants. Resource extraction formed the core of Assyrian imperial concerns; perpetual expansion and maintenance of the empire fostered an ever-increasing reliance upon both resources for sustenance (agriculture and pastoralism) and resources for exchange (precious goods and specialized services) (Parker Reference Parker2003; Rosenzweig Reference Rosenzweig2018; Rosenzweig and Marston Reference Rosenzweig and Marston2018; Thareani Reference Thareani2016). This was true as well for Assyria’s imperial core, where growing cities far outstripped the resources available nearby to sustain their rising populations and thus relied upon importing goods and labour (Oded Reference Oded1979; Postgate Reference Postgate1992; Radner Reference Radner and Dercksen1999). In addition to Erbil, the Assyrian center Kilizu (modern Qasr Shemamok) might have also been a local destination for goods, though it is located further from Qach Rresh (20km on the most direct path) (Masetti-Rouault et al. Reference Masetti-Rouault, Rouault, Calini, Cremonini, Gasparotto, Herr, Mahmoud, Picotti, Poli, Rossi, Marti, Rouault, Tenu, Ali Mohamed Amen, Déroche, Masetti-Rouault and Nicolle2022). Based on the proximity of the site to Trpa Spiyan, it is almost certain that a road connected the larger site and Erbil (running nearly directly north-south) and as such may have served Qach Rresh as well. Between Kilizu and Trpa Spiyan in the southwestern part of the Erbil Plain lies the site of Aliawa, which was also occupied in the Neo-Assyrian period (Peyronel et al. Reference Peyronel, Minniti, Moscone, Naime, Oselini, Perego and Vacca2019). This might imply a road which linked these sites, and by extension Qach Rresh, as well.

The location of these larger sites within a day’s walk of Qach Rresh possibly situates the site within the wider framework of purposeful settlement planning mentioned above. It is possible that the sudden increase in newly established settlements may have related to the Assyrian practice of deportation. Begun under Assur-dan II in the tenth century B.C.E., deportation was one of the empire’s main tactics of administration and oppression (Oded Reference Oded1979). Deportations and resettlements of conquered populations drastically altered the makeup of the empire’s core. It reached its peak in the eighth and seventh centuries B.C.E. under the Sargonid dynasty, with the highest estimates from Oded’s foundational study of deportation concluding that as many as four and a half million people were relocated over a 250-year period, with 85% of them settled in the core of the Assyrian Empire (Oded Reference Oded1979). Deportation strategies not only destabilized the new territories brought under Assyrian hegemony, to prevent uprisings, but also provided a new source of direly needed labour for the Assyrians to exploit. While most written records describe the resettlement of these uprooted peoples within cities (‘“[I add]ed (them) in great numbers to the massive fo[rc]es of the god Aššur… I filled Assyria in its entirety like a quiver. I distributed the re[st of them] like sheep and goats among my palaces, my nobles, the entourage of my palace, and [the citizens of Ninev]eh, Calah, Kalzu, (and) Arbela,” [Leichty Reference Leichty2011: Esarhaddon 033 RINAP, rii 18-22]), many were settled outside of these urban spaces as well (Oded Reference Oded1979: 47, 67–74). This is some evidence for this phenomenon seen in the Jazira, where the stele of Adad-nirari III from Tell al-Rimah recounts the purposeful foundation of over three hundred villages, settled by the ‘subjugated peoples’ from Hatti, Amurru, Damascus, and Na’iri which Adad-nirari III boasts of conquering (Morandi Bonacossi Reference Morandi Bonacossi1996: 164–165; Page Reference Page1968: 143). It is unclear how Qach Rresh was situated within Assyria’s increasing demand for labour and sustenance and how these demands affected its residents. However, its existence as a newly established site with facilities for the production/storage of goods, paired with the monumental scale of those facilities, almost certainly points to centralized governmental or elite involvement in its construction and operation.

Just as interesting is Qach Rresh’s apparent abandonment upon (or just before) the collapse of the Assyrian Empire. The site’s use in the post-Assyrian/ Iron Age III period thus far remains based on the deposits of refuse within Buildings B and C, which have yielded more ceramic data from the sixth–fourth centuries B.C.E. than any other site on or around the Erbil Plain. Archaeological knowledge of post-Assyrian and Achaemenid occupation within and surrounding the Assyrian heartland is largely limited to urban contexts, where discussions of ‘squatter’s settlements’ dominate (Curtis Reference Curtis, Lanfranchi, Roaf and Rollinger2003; Kuhrt Reference Kuhrt1995). It has largely been assumed, from short contemporary descriptions of the Assyrian heartland (such as that preserved in Xenophon’s Anabasis) and the dearth of archaeological evidence (Hauser Reference Hauser and Frahm2017; Palermo Reference Palermo, Genito and Maresca2019a), that the countryside depopulated rapidly upon the collapse of the Assyrian Empire and largely remained that way until an upswing of settlement numbers in the Hellenistic and Parthian periods (Palermo Reference Palermo2019b). Recent discussions have framed the weakening of the Assyrian Empire and the following abandonment of the landscape within discussions of environmental and climatic fluctuation (Schneider and Adalı Reference Schneider and Adalı2014; Sinha et al. Reference Sinha, Kathayat, Weiss, Li, Cheng, Reuter, Schneider, Berkelhammer, Adalı, Stott and Edwards2019). The general presence of a community at or near Qach Rresh apparently engaging in robust agropastoral lifeways and maintaining community production of textiles, along with other products, may complicate this picture as RLIIM continues.