A Wayside Shrine in Northern Moab: Excavations in the Wadi ath-Thamad

Part 1

The site and its excavation

1

Introduction and geographical setting

P. M. Michèle Daviau with an excursus by Carlos E. Cordova

During the first season of regional survey in 1996, the Wadi ath-Thamad Survey team discovered site WT-13, south of the heavily-built town of al-Rumayl (WT-18; Fig. 1.1). Attention was drawn to this isolated hill in central Jordan because of its location, overlooking the Wadi Wala (the continuation of the Wadi ath-Thamad) and the discovery on its surface of two ceramic female figurines by Rommel Ghrayib (Department of Antiquities representative); these were accompanied by fragments from ceramic statues, two beads and a scarab. Along with these artefacts, the team collected 370 ceramic sherds which provided evidence for an Iron Age date.¹ In view of the unusual nature of the finds and the date of the ceramic remains, some of which appear to be contemporary with the Iron Age pottery from Khirbat al-Mudayna on the Wadi ath-Thamad (WT-1, henceforth, Mudayna Thamad),² WT-13 was deemed worthy of further study.

The objectives of excavation were to determine the exact character of the site, its date and its relationship to the nearby site of al-Rumayl, the cultural sphere reflected in the material culture, and the sequence of phases of construction and repair. Given its small size, it was assumed to be a specialized site, possibly associated with cultic activity.

The identification of a cultic site is always problematic, especially when there are no known parallels within the same cultural sphere and no relevant textual evidence (Renfrew 1985:3; Webb 1999:11). Such was the case with the discovery of a bronze bull figurine, and the subsequent excavation and publication of the temenos in the hill country of Samaria where it was apparently found (Mazar 1982:27, n. 2). The interpretation of this walled enclosure as an open air cult site generated considerable scholarly discussion (Coogan 1987; Mazar 1988; Wenning and Zenger 1986; Zevit 2001:178–180). At the same time, the Bull Site yielded only a small amount of pottery and a handful of unusual artefacts, one fragment resembling a ceramic plaque or architectural model (Mazar 1982: fig. 10). Surprisingly, no other explicitly cultic objects were reported; instead the remaining finds (16 diagnostic pot sherds, 24 lithic tools, and one piece of folded bronze (Mazar 1982: figs. 9, 11, 13) appear to be domestic or craft-related in nature. In spite of this small assemblage and the fact that evidence for repeated cultic performance is lacking, there are few who challenge the cultic identification of the site.³

The situation at certain town sites is also problematic; for example, the presence of fenestrated stands at Dhībân and Ai strongly suggests cultic activity, especially when associated with female figurines. However, the current case for a recognizable communal cult centre at Dhībân is very tenuous, because it consists only of a pair of stone walls entering a balk on the east side of a monumental building (Morton 1989:245).⁴ This hardly constitutes the material correlates for a temple and Routledge (2004:168–173) has now convincingly demonstrated that the building in Area L was a palace. In contrast, Room 65 at Ai was not originally identified as a shrine by its excavator, although the benches along two walls are strong architectural evidence that this room may have been a cultic area (Zevit 2001:153–156). This interpretation is supported by the presence of a cultic stand and several chalices, items usually associated with religious activities (compare Lachish Room 49; Aharoni 1975: fig. 6). However, the small number of finds at all of these sites is in contrast to sites found more recently in the Negev and Arabah, such as Ḥorvat Qiṭmit (Beit Arieh 1995) and En Ḥaṣeva (Cohen and Yisraeli 1995), where a large number of ceramic figurines, statues, cultic stands, lamps, specialized ceramic vessels and artefacts constitutes the paraphernalia of repeated cultic activity at these sites.⁵

To date, similar Iron Age sites with large numbers of cultic items have not been found in Ammonite territory on the Transjordanian plateau. Prior to the rise of the Ammonite kingdom, there were two Late Bronze Age sites with high status artefacts, the Amman Airport temple (Hankey 1974a; 1974b), and the partially excavated building at Umm ad-Dananir (McGovern 1989:35). In each case, these were heavily-walled structures with only limited outdoor activity areas. In view of their artefact assemblages, these sites may not have been temples or shrines in the ordinary sense, but served instead a special purpose (Herr 1983:23–28). At Tall al- Umayri, south of Amman (Bramlett 2008) and in the Jordan Valley at Tall Dayr Alla (Franken 1992) and Pella (Bourke 2012; Ward 2014), there were Late Bronze Age temples which did serve as major religious centres in their respective settlements. These impressive temple buildings yielded rich cultic assemblages. Dating to late Iron Age I, an open court building located outside of Tall al-Mazar proper has also been identified as a sanctuary (Yassine 1984), although the number of figurines and explicitly cultic material recovered from this site is limited (see Chapter 15).

Figure 1.1. Map of Iron Age Palestine and Transjordan (prepared by C. Ramsoomair)

The paucity of figurines from Transjordan dating to the early Iron Age is not surprising. Only a handful of town sites were recently excavated with occupation which dates to Iron Age I; most notably, Tall al- Umayri (Clark 1996) and Tall Jawa in Ammon (Daviau 2003); Khirbat al-Mudayna al- Aliya (Routledge 2000; 2004); Khirbat al-Mudayna al-Mu arradja (Olávarri 1983), Ara ir (Olávarri 1965) and Khirbat al-Lahun (Homès-Fredericq 1997) in central Moab, with few, if any, ceramic figurines assigned to Iron I contexts.⁶ The same phenomenon, an apparent decline in coroplastic art, was seen on Cyprus during the same period. It was only in early Iron II, with the increasing contact between Cyprus and Phoenicia, that figurine production increased significantly on the island with the introduction of Levantine cultic drama and myths (Karageorghis 1991:13). This same Phoenician influence may help to explain the rich repertoire that developed concurrently in the small kingdoms of Transjordan (see Stern 2003).

Prior to excavation at Site WT-13, Ammonite and Moabite figurines known from various contexts in settlement sites were most often associated with the domestic cult (Dabrowski 1997; Daviau 1996: fig. 4:5; Daviau and Dion 1994: fig. 1; Daviau 2014a) or with funerary customs (Saller 1966). Other cultic assemblages are without provenience and come from private collections. A case in point is an assemblage of artefacts in the Museum of Art and Archaeology (University of Missouri-Columbia) published by Weinberg (1978) which includes female figurines, a model shrine, perforated cups, animal figurines and lamps. There was no known provenience for these artefacts except Weinberg’s suggestion that they originated in the area of Mount Nebo. Although several Iron Age tombs have been excavated at Mount Nebo and in Madaba itself (Dornemann 1983:31–35, 48–49), such an artefact group is usually associated with town sites (Clark, Herr and Trenchard 2000), possibly with domestic cult activity areas (Holladay 1987:276; table 3; Daviau 2001b; 2014a), or with cultic centres, rather than with graves. To date, only two town sites with temples where such an intact assemblage could have been recovered have been excavated among the known sites with Iron Age occupation on the Moabite plateau, Mudayna Thamad (Daviau and Steiner 2000) and Atarus (Ji 2012; Daviau 2012). Both of these sites are likely candidates for such an unprovenienced assemblage, although Mudayna Thamad was undisturbed prior to excavation and most of its finds were broken during its final destruction, while relatively well preserved finds have been recovered from Atarus. Although the discovery of dozens of broken figurines, ceramic statues, and tripod cups in the shallow deposition layers at WT-13 reflects the nature of this small site, it does not offer a likely find spot for the nearly intact assemblage studied by Weinberg.⁷

Chronological setting

The rise of identifiable polities in Transjordan is dated to the Iron Age II, with little or no continuity with earlier Iron Age I sites in some regions, such as in Moab, both north and south of the Wadi Mujib. Various attempts to assign dates to Iron Age horizons in Jordan are based on the better known chronological links between occupation at tall sites in Palestine and Phoenicia and evidence for influence from Egypt and Mesopotamia. The beginning of the 20th dynasty (Egypt) and the disruptive movements of Sea Peoples throughout the Levant marks the transition from Late Bronze to Iron Age I in Cisjordan and the northern Levant. For central Jordan, both the beginning and the end of Iron I does not appear to be related to broader historical events, with Iron II occupation being recognized primarily in the difference in material culture, which itself varies considerably between Gilead, Ammon, Moab and Edom. For the purpose of his temple study, Mierse (2012:60) adopted the generally accepted Iron I sequence utilized by Bloch-Smith and Nakhai (1999) and the sequence for Iron II delineated by Herr (1997), although the beginning of Iron IIA in Moab is not fixed by any securely dated series of events or by C14 dates.

For WT-13, the operating assumption is that the principal phases date to late Iron Age I–IIA and to Iron IIB. Comparison with pottery and objects from surrounding sites is also plagued with the same chronological uncertainty, although a few clear links for each of the phases with Iron Age II occupation at Mudayna Thamad can be demonstrated. So too, imported pottery and objects make an earlier and later date unlikely (Chapters 3, 5–7; see also Daviau and Klassen 2014).

The site in its geographical setting

The site of WT-13 is located on a rounded hilltop just south of a large bend in the Wadi ath-Thamad/Wadi Rumayl (Fig. 1.2).⁸ The wadi then turns south and winds around the west side of the hill through a valley with steep slopes falling 284 m to the valley bottom.⁹ South of WT-13, the valley flattens out, giving a broad view of the hills to the west and south. On the east and north, the site is separated by a shallow gully from the fortified settlement of al-Rumayl (WT-18; Chapter 14) and the field of cisterns and graves (WT-82–WT-86) to its south. WT-13 is thus located on an isolated promontory in view of a major Iron Age settlement, which itself watches over a natural ford on Wadi Rumayl to the north and over the farm lands along the Wadi ath-Thamad to the east.

Figure 1.2. Topographic map of WT-13 with working grid (C. M. Foley)

In 1996, the site of al-Rumayl South,¹⁰ now MEGA Jordan Site 6540, was thought to be WT-13. However, WT-13 is now identified as a separate site, number 2999 in the MEGA Jordan database. Coordinates established by the Wadi ath-Thamad Survey in 2001 are UTME 772771, UTMN 3496672-Zone 36. WT-13 is located 295 m south¹¹ of the centre of the better known site of Khirbat al-Rumayl,¹² a large Iron Age fortress and town site on the south side of Wadi Rumayl.¹³ From al-Rumayl, WT-13 is at a bearing of 190 degrees and is west of the modern paved road which runs from al-Rumayl to Umm al-Rasas. On the north, south, and west sides of WT-13, there is evidence of mechanical grading (tractor or bulldozer) that resulted from road work around the hill, and from terracing to prevent erosion of the hillside. Cultivated areas are located on the saddle of the hill to the east, as well as in the gully separating WT-13 and the modern road. Another extensive area of cultivated fields is seen in the wadi bed to the southwest.¹⁴

Looking north from the top of WT-13, one can see the heights south of Madaba and, to the south, the beginning of the Dhībân plateau. At present, the modern road running between al-Rumayl and WT-13 begins at Nitl in the north and winds its way to Umm al-Rasas in the southeast. Whether this is the trajectory of an ancient road cannot now be determined, although the position of al-Rumayl is prominent in the landscape. The location of WT-13 was chosen with care, being the westernmost point on the heights, before the wadi bends to the west toward the Early Bronze Age site of Khirbat Iskandar. On this hilltop, the site could serve as a landmark or beacon for travellers coming from the south, although it is blocked from view on the north by the hill of al-Rumayl. The importance of such a setting for sacred space has been recognized by various scholars (Alcock and Osborne 1994:23; Steinsapir 1999:182), even though we may not be able to recover evidence for the natural surroundings in antiquity or appreciate all the implications of that landscape for the ancient inhabitants (see Chapter 14).

Excursus: Geography of the Wadi ath-Thamad system

Carlos E. Cordova

The Wadi ath-Thamad originates in the Jordanian plateau a few kilometers east of the Desert Highway which runs from Amman to Aqaba. The wadi enters the Thamad Graben and meanders along a relatively shallow valley, where at least three alluvial terraces were formed. The Thamad Terrace is the oldest, containing Epipaleolithic and Early Neolithic occupations. The Tur al-Abyad Terrace contains Chalcolithic material, and the youngest terrace, named Historic, contains Nabataean and Roman pottery fragments (Cordova 2000).¹⁵

At the confluence with Wadi Za faran, Wadi ath-Thamad becomes Wadi al-Rumayl. From this point on, the wadi runs at the foot of the Rumayl escarpment, formed by the Rumayl Fault. It is on the edge of this escarpment that Khirbat al-Rumayl is located, occupying a vantage point for the entire valley to the north and east. Wadi Za faran merges into Wadi al-Rumayl on a wide flood plain at about 270 m to the north northwest (ca. 340 m Ε) of al-Rumayl. From this point the wadi turns south and enters a deep canyon (ca. 200 m) bisecting the Dhībân Plateau. South of WT-13, the wadi turns west, from which point it is known as Wadi al-Wala.¹⁶

The streams of these wadis have no permanent flow. Major fluvial activity occurs during the winter and early spring, when rains are relatively frequent. For the rest of the year, water remains in pools, often used by local Bedouin to provide water to sheep and goats and for irrigation using pumps.

Site WT-13 is located on the eastern or left side of the wadi, close to the edge of the plateau. The site is located on the al-Hisa Phosphorite formation of Cretaceous age. This formation presents strata of phosphate, limestone and other marine sediments poor in chert (Al-Hunjul 1995). Intense caving has occurred in the soft strata of al-Hisa Phosphorite, creating small caves and natural cavities, often used for the storage of food and water. This formation is underlain by the Amman Silicified Limestone, which is more resistant to erosion and provides more sources of chert.

Biogeographically, the site is located on the Irano-Turanian Steppe, which in this part of the country is characterized by anabasis (Anabasis articulata), spiny saltwort (Noaea mucronata), wormwood (Artemisia herba-alba), Syrian rue (Peganum harmala), and oleander (Nerium oleander). All these are mostly antipastoral plants, meaning they are unpalatable or poisonous to sheep and goats. The impact of pastoralism on the slopes of the wadi has been tremendous, resulting in intense erosion and frequent landslides. In some cases, landslides are associated with active faulting.

References for the Excursus

Cordova, C. E.

2000 Geomorphological evidence of intense prehistoric soil erosion in the highlands of central Jordan. Physical Geography 21:538–567.

2007 Millennial landscape in Jordan: Geoarchaeology and cultural ecology. Tucson, AZ: University of Arizona Press.

Cordova, C. E.; Foley, C.; Nowell, A. and M. Bisson

2005 Landforms, Sediments, Soil Development, and Prehistoric Site Settings on the Madaba-Dhiban Plateau, Jordan. Geoarchaeology: An International Journal 20:29–56.

Al-Hunjul, N. Gh.

1995 The Geology of Madaba Area: Map Sheet (3153-II). Amman: Geological Map Division, Natural Resources Authority.

Exploration history

The site of WT-13 does not appear to have been noticed by the explorers of the late 19th and early 20th century, such as Tristram, who rode along the Wadi ath-Thamad in 1873, or by Brünnow and von Domaszevski (1904–5). These explorers, along with Musil who followed them (1907), documented a large number of sites between Madaba and Qasṭal, including Mudayna Thamad. Nevertheless, it was only in 1996 that WT-13 was located by the Wadi ath-Thamad Project Regional Survey under the direction of J. Andrew Dearman (Fig. 1.3a).¹⁷ Exploration in the immediate vicinity of al-Rumayl, and the documentation of its 42 cisterns and caves, led the team to the west side of the road and to a hill overlooking Wadi Rumayl. At the time, WT-13 appeared as an area of disturbed soil with two fallen stone monoliths (0.40 × 0.40 × 1.40 m in size), collapsed wall stones, and surface vegetation. Extending along the oblong pile of ruined walls to the southeast, was a 7.00 m long wall line (one row of stones wide), possibly indicating the remains of an enclosure or courtyard. It was in the midst of the collapsed debris that pottery, figurines and artefacts were first recovered (Figs. 1.4, 1.5). The initial assessment of the pottery was that it dated to the Iron II and late Iron II periods, well represented in the carinated bowls (Fig. 1.4:1–3), Moabite-style square rim cooking pots (Fig. 1.4:19), and the rib-sided ceramic mortar bowl (Fig. 1.5:15). A unique jug rim (Fig. 1.5:6) with a formal parallel in a red slipped sherd from the 10th–9th century Stratum V at Tel Malḥata (Freud 2015: fig. 4.174:10) was also an indicator of the date of the site. Another feature, situated on the northeast slope of the hill, is a bell-shaped cistern capable of serving as a source of water for activities carried out on the top of the hill (Chapter 14). Additional collapsed stone features were also noticed during the first two seasons of survey (1996, 1997), some of which may be natural depressions in the bedrock.

Figure 1.4. Unstratified Survey pottery

1. (A10/101.20). Carinated Bowl; D 13.0 cm; very pale brown (10YR 7/3) fabric, very pale brown (10YR 8/2) exterior.

2. (A10/101.21). Carinated Bowl; D 15.0 cm; light red (2.5YR 6/6) fabric, light reddish brown (2.5YR 6/4) exterior.

3. (A10/100.30). Carinated Bowl; D 25.0 cm; Leiden. ¹⁸

4. (A10/15.2). Carinated Bowl; D 14.0 cm; light brownish gray (10YR 6/2) fabric and exterior.

5. (13R/16.1). Inverted rim bowl; D 20.0 cm; pink (7.5YR 7/4) fabric, pink (5YR 7/4) exterior.

6. (A10/16.1). Inverted rim; D 23.0 cm; pink (5YR 7/3) fabric, very pale brown (10YR 8/3) exterior.

7. (13R/16.4). Hemispherical; D 24.0 cm; pink (7.5YR 8/4) fabric, pinkish white (7.5YR 8/2) exterior.

8. (A10/16.2). Hemispherical; D 26.0 cm; light red (2.5YR 6/6) fabric, very pale brown (10YR 7/4) exterior.

9. (13R/46.12). Incurving rim; D 21.0 cm; very pale brown (10YR 7/3) throughout.

10. (A10/100.34). Incurving rim; D 32.0 cm; Leiden

11. (13R/30.8). Cup; D 13.0 cm; very pale brown (10YR 7/3) throughout.

12. (A10/12.3). Cup; D 11.0 cm; very pale brown (10YR 7/3) fabric and exterior.

13. (A10/3.1). Krater; D 28.0 cm; reddish yellow (7.5YR 7/6) fabric and exterior.

14. (13R/71.20). Krater; D ca . 22.0 cm; Leiden.

15. (13R/38.9). Krater; D 29.0 cm; light brown (7.5YR 6/4) fabric, grayish brown (10YR 5/2) exterior.

16. (13R/46.20). Cooking pot; D 16.0 cm; red (2.5YR 4/6) fabric and exterior.

17. (13R/18.3). Cooking pot; D 17.0 cm; red (2.5YR 5/6) fabric throughout.

18. (A10/32.7). Cooking pot; D 13.0 cm; light reddish brown (2.5YR 6/4) fabric, pale red (10R 6/4) exterior

19. (13R/30.13). Cooking pot; D 15.0 cm; reddish brown (2.5YR 5/4) fabric, reddish brown (5YR 5/4) exterior.

20. (13R/30.14). Cooking pot; D 26.0 cm; reddish brown (5YR 5/3) fabric, yellowish red (5YR 5/6) exterior.

Figure 1.5. Unstratified Survey pottery and figurines

1. (A10/100.33). Storejar; D 9.0 cm; Leiden.

2. (13R/15.4). Storejar; D 8.0 cm; light reddish brown (2.5YR 7/4) fabric, pink (5YR 7/3) exterior.

3. (13R/15.10). Storejar; D 10.0 cm; pink (7.5YR 7/3) fabric and exterior.

4. (13R/46.13). Jar/Jug; D 9.0 cm; light gray (10YR 7/2) fabric, very pale brown (10YR 7/3) exterior.

5. (13R/15.7). Jug; D 8.0 cm; very pale brown (10YR 8/3–8/2) fabric and exterior.

6. (13R/15.3). Jug; D 8.0 cm; Leiden.

7. (A10/.100.31). Jug; D 9.5 cm; Leiden.

8. (A10/2.4). Jug; D 9.0 cm; very pale brown (10YR 7/3) fabric and exterior.

9. (13R/15.16). Lamp; D 8.5 cm; very pale brown (10YR 8/4) fabric, light red (2.5YR 6/8) exterior, mottled.

10. (13R/16.6). Base; D 5.5 cm; pink (5YR 7/4) fabric and exterior.

11. (13R/18.15). Base; D 14.5 cm; light brown (7.5YR 6/4) fabric, red (2.5YR 5/6) exterior.

12. (A10/100.6). Base; D ca . 4.0 cm; pink (7.5YR 8/3) fabric, pink (5YR 7/4) exterior.

13. (A10/100.2). Base; D 6.5 cm; very pale brown (10YR 7/4) fabric and exterior.

14. (A10/1.1). Basin; D 36.0 cm; light gray (5Y 7/2) fabric and exterior.

15. (13R/15.11). Mortar; D 25.0 cm; reddish yellow (5YR 7/6) fabric and exterior.

16. (13R/30.15). Painted body sherd; reddish yellow (5YR 7/6) fabric, very pale brown (10YR 8/4) exterior, yellowish red (5YR 5/6) paint.

17. (13R/18.7). Painted body sherd; reddish yellow (7.5YR 7/6) fabric, reddish yellow (7.5YR 6/6) exterior, red (2.5YR 5/6) paint.

18. (13R/16.5). Painted body sherd; very pale brown (10YR 8/3) fabric, very pale brown (10YR 8/4) exterior, red (2.5YR 5/6) paint.

19. (13R/16.2). Painted body sherd; very pale brown (10YR 8/3) fabric, reddish yellow (7.5YR 7/6) exterior, dark gray (10YR 4/1) and red (2.5YR 4/8) paint.

20. WT 15. Naked female with disc ( Chapter 4 ).

21. WT 21. Naked female with disc ( Chapter 4 ).

Excavation history

Visited and disturbed by local folk following our initial discovery, WT-13 was subsequently excavated in an intensive salvage excavation in 1997 (June 2–12).¹⁹ It was soon evident that human activity at the site extended over an area of at least 14.00 × 16.00 m. The heaviest concentration consisted of a ring of piled up boulders and loose soil which formed a rough oval that partially covered two parallel walls visible at ground level. Disturbed areas provided evidence that modern looters had reached bedrock at several places within the oval, although they did not expose it completely.

During the first excavation season (1997), two squares (A1, A2), each 6.00 × 6.00 m and aligned north-south, were opened to expose the features in the centre of the site. Due to the presence of wall lines along the north and south balks, these two squares were extended to 8.00 m north-south and excavation resumed in these two squares in 1999 (Fig. 1.3b). The enlarged area included most of the disturbed remains between the two visible walls (W1001, W1002) which ran east-west along the north and south sides of the concentration of collapsed boulders and topsoil. Documentation also continued around the site itself, locating several cisterns,²⁰ areas of exposed bedrock, and recent bulldozer activity.

Figure 1.3. a) View of collapsed wall stones prior to excavation, looking west; b) Working grid showing distribution of squares (redrawn by P. M. M. Daviau, 2014)

Figure 1.4. Unstratified survey pottery: 1–10) bowls; 11–12) cups; 13–15) kraters; 16–20) cooking pots

Figure 1.5. Unstratified survey pottery and objects: 1–3) jars; 4–7) jugs; 8) juglet; 9) lamp; 10–13) base forms; 14) basin; 15) ceramic mortar; 16–19) painted sherds; 20–21) figurines

In 1998, members of the Wadi ath-Thamad Project excavating at Mudayna Thamad opened two new half squares (3.00 × 6.00 m in size), A3 to the north and A5 to the south of Square A2. Bedrock was soon reached in A3, indicating that the excavations in 1997 had come close to the northern edge of the original deposition. In A5, broken vessels were strewn throughout the uppermost soil layers (A5:1, 2). Additional soil from inter-seasonal disturbance extended across the bedrock in Square A2, in the same area which had been dug to bedrock in 1997. This square was again cleaned to recover ceramic material strewn over the bedrock. What was learned from this brief season was that the site was larger than previously imagined, and that additional work was needed to understand its character.

In 1999, a two week season was undertaken under the supervision of Laura Foley. All four squares were re-opened (A1, A2, A3, and A5) and one new square (A4), west of Square A2, was opened in an attempt to uncover the perimeter of the site on the northwest. In Square A4, the remains of the western wall (W1004) were exposed on the crest of the hill, overlooking the wadi. A topographical map of the site and its slopes extending down into the wadi bed was prepared by C. M. Foley, director of the regional survey at that time.

The working grid, aligned to magnetic north, was re-established in 2001 reducing the size of Squares A1 and A2 to their original size and a topographical map of the larger area, including both WT-13 and al-Rumayl (WT-18), was prepared by Robert Force. With the discovery of the west wall in Square A4, it became imperative to clarify the easternmost wall (W1003) in Square A1. Thus, a new square (A6) was opened to the south of A1 and east of A5,²¹ while excavation in A1 was resumed. Indeed, a complex association of walls was exposed, and the southeast corner of the structure was revealed.

In a final five week season (2003), all remaining loci in Squares A1, A2, A4, and A6 were completed and one new square (A9) was opened on the west. This square was located on the west slope south of A4 and both were excavated in order to determine the full extent of the archaeological remains. Additional 6.00 × 6.00 m squares (A7 and A8) were laid out and opened on the east to complete the exposure of the east wall (W1003). Square A8 was extended a further two meters to the east in order to determine the trajectory of the bedrock scarp exposed in A7. The eastern extremity of the site is marked by exposed bedrock and thin soil layers that form a saddle, linking WT-13 with the ridge that extends southeast as far as the modern road. Additional excavation in Square A11 to the north of North Wall 1001 completed our work at the site.

Introduction to the stratigraphy

Based on changes in the architecture and artefactual assemblages, excavation identified two principal Iron Age use phases, each with a distinct configuration and function (Table 1A).²² Within Strata III and II, there is evidence for intensive use of the site, but not of occupation as such. Included in Stratum I are post-Iron Age remains as well as a stone inscribed in Thamudic script and random finds from the Nabataean–early Roman period, which were widely scattered in topsoil layers and were not related to any securely dated architectural construction or identifiable activity area.

Table 1A. Stratigraphic sequence

Documentation and recording methods

Dearman compiled the initial documentation of Site WT-13 in 1996 using the Regional Survey recording sheets designed by the Hesban Survey and modified according to the needs of the Wadi ath-Thamad project.²³ At that time, the architecture was not exposed, but lay buried under the collapsed debris and a thick layer of wind-blown soil. Initial documentation of the size of the site and of its features was revised in later seasons. Eight features were documented on the west side of the modern road, although some of these features may have been related more closely to the site of al-Rumayl than to the activities which took place on the hilltop at WT-13.

During each excavation season, the locus sheets in use at Mudayna Thamad, the principal site under excavation by the Wadi ath-Thamad Project, were used at WT-13 to record the details of soil layers, architecture, and installations.²⁴ Additional forms consisting of running lists for locus numbers, pottery pails, objects, and photographs, along with daily journal sheets, weekly summaries and field phase sheets were also included in the field notebooks.

Object registration

All fragments of organic, geological and ceramic materials (except pottery) were saved and registered as objects (for example, WT-13•A2/31.160=Site [WT-13] Field+Square [A2] / Pottery Pail number [31].Object number [160]); a shortened form includes only the siglum ‘WT’ and the object number (WT 160). Following their cleaning and identification, partially preserved objects with cultural significance, such as figurines, were assigned a season and artefact number (WT 35-2/535=Site [WT] + Object number [35] - Season [2] / Artefact number [535]).²⁵ In the discussions that follow (Chapters 4, 5), all objects are grouped according to types and sub-types within each class, and are identified by their registration number (WT + Object number), the figure number within the text, and by their Field+Square : Locus / Pottery Pail number; artefacts chosen by the Department of Antiquities are identified by both their object and artefact number. This is followed by material, Munsell Soil Color Codes (where appropriate), technology, decoration (if any), and measurements. The maximum dimensions in centimetres are indicated for artefacts that stand upright as H(eight), W(idth) or D(iameter), and T(hickness), while items which lie flat are described with their L(ength), W(idth) or D(iameter), T(hickness).²⁶ Objects with depressions and perforated objects include the interior diameter (int D) and, where relevant, the depth of the depression (int Dpt). The condition (complete, broken, chipped, partially restored, or fragment) of each object appears last. Full details of measurements and weight in kilograms, initials of artists and explanatory comments are recorded last in the database. Parallels from other sites are listed in the text where possible, although an exhaustive study of fragmentary items is not included in this presentation.

Curating the artefacts

At the end of each season, members of the Department of Antiquities of Jordan reviewed the objects and chose those suitable for museum display. These objects are now housed in the Madaba Museum.²⁷ Over six seasons the Department retained a total of sixty-nine artefacts from WT-13. These items (Table 1B) are designated in the database as located at the DoA. The remaining objects and samples are either at Wilfrid Laurier University (WLU) or in Jordan.

Ceramic registration

The pottery from WT 13 is almost exclusively Iron Age, although a number of sherds represent vessels that date to the Roman period. Since the shrine was not in use at that time, these sherds are considered random finds.²⁸ However, like all diagnostic ceramic remains from the site, the Roman sherds were registered. During excavation, pottery pail numbers are listed on the reverse of their respective locus sheets and in a running list for each square. Individual diagnostic sherds from a given pail are registered according to their find spot, pottery pail and item number; this code includes the Site [WT13] +Field [A]+Square [2] Pail number [35] Sherd number [1] (for example, WT13 A2•35•1). The full registration found in the text may include the locus number, Field+Square : Locus number / Pail number. Sherd number (A2:23/35.1). Because several loci may constitute a single feature, this number is given only in certain instances. More common is a streamlined number including only Field+Square / Pail number. Sherd number (A2/35.1). Due to the condition of the ceramic material, only a small number of vessels could be partially restored and were given vessel numbers in the range P1300–P1399. An individual P number represents a group of sherds that constitute a discrete vessel whose formal type could be identified. Unfortunately, there were no intact vessels apart from two miniature juglets (P1301, P1302), which were too thick and too small to break easily. However, both are missing their rim. Altogether, nearly 35,000 sherds were recovered and, of these, 4,182 are diagnostic. Of the diagnostics, more than 2500 were colour coded using Munsell Soil Color Charts and more than 1600 were drawn and now serve as the basis for the discussion of the ceramic assemblage and activity areas.²⁹

Table 1B. Artefacts in the Madaba Museum

Control of the artefact and ceramic corpus is complicated by the division of the assemblages and the slow recognition of possible mends, not to mention the re-identification of certain fragments from pottery to artefacts and vice versa. In spite of extensive study, the identity of certain fragments remains uncertain. This is reflected in the diversity of cultural traditions and the complexity of the ritual behaviours which were carried out at WT-13 at various times during the Iron Age.

Notes

1Although a handful of sherds are Roman in date, there were no architectural remains from this period at WT-13 and no other evidence for later reuse of the site. This mixture of Iron Age and Roman remains is common at numerous sites in northern Moab, where Iron Age forts and town sites were reused in the Nabataean–early Roman period. Nearby sites with early Roman period habitation that may account for the presence of these sherds include Qasr az-Za faran (WT-34) and Tauga (WT-6).

2Site WT-13 is a separate and distinct site from Mudayna Thamad ( pace Mierse 2012:138, who identifies WT-13 (3 km to the west) as the second shrine at Mudayna (instead of the shrine with two standing stones in front of the North Gate; Daviau et al . 2006: fig. 4). Within the territory of ancient Moab, there are six sites known as Khirbat al-Mudayna (Miller 1989b). The site on the Wadi ath-Thamad with this name is the northernmost settlement, assigned the number WT-1 by the Wadi ath-Thamad survey team.

3Zevit (2001:180) compares this site in its simplicity to Middle Minoan peak sanctuaries; however, these sanctuaries are characterized by craggy mountain tops riddled by fissures, one or more trees, religious iconography (rock drawings), a temenos wall and, at certain sites, structures consisting of several rooms, along with horns of consecration, figurines or statues, and a wide variety of cultic paraphernalia and pottery, including rhyta and kernoi (Rutkowski 1986:73–98).

4Although I agree with the interpretation of these finds as cultic, Coogan’s criterion of isolated does not seem to apply to a building inside a town site, for example, the earlier interpretation of the structure in the city of Dhībân ( pace Mattingly 1989:229).

5At Ḥorvat Qiṭmit, 792 objects were registered, not including ceramic vessels, while at En Ḥaṣeva, there were 67 objects of clay and 7 of stone (Cohen and Yisraeli 1995:224); these numbers are in stark contrast to the situation at the Bull Site with its handful of possible cultic artefacts.

6The Iron Age I architecture at Tall Jawa in Ammon was exposed in a very limited area; the majority of the 900+ pottery sherds were in secondary contexts (Daviau 2003:323–42).). Ji and Attiyat (1997) and Ji and Lee (2003) conducted survey on the Dhībân Plateau in which they identified sites with Iron Age I pottery and studied them in association with those which were previously excavated, such as Khirbat al-Lahun, Ara ir, and Khirbat al-Mudayna al-Mu arradja. No evidence for cultic sites was recognized.

7The only anomaly in the Missouri collection is the presence of animal figurines. No zoomorphic figures from among the more than 179 figurines at Mudayna Thamad were located in Temple 149 (Daviau and Steiner 2000), while the figurines and statues present in the temple at Atarus are quite different in style from the equid and bovine figurines from other town sites. In the WT-13 corpus, there is a meagre representation of zoomorphic figures consisting of two ceramic fragments, one head and one rear quarters, and, possibly, one stone paw(?); Chapter 5 below.

8As the wadi continues to flow southwest, it becomes known as Wadi Wala.

9This distance is line of sight, measured with a Bushnell Compact 800 Yardage Pro, which reads in both metric and standard measurements and in this case corresponds to total station readings. The topographical map was prepared by Robert Force and Christopher M. Foley using a Leica total station and UTM coordinates were assigned to both sites.

10 Based on the 1996 report of the Wadi ath-Thamad Regional Survey, Site WT-13 was identified as Rumayl South as noted in the updated on-line version of JADIS (prior to 2009) and cited as such in our early publications (Daviau 2006:24, n. 18); the site was not listed in the 1994 printed edition; now see MEGA Jordan.

11 The distance from a datum set up at al-Rumayl to WT-13 reads 270 m at a bearing of 190°; this reading was taken by C. M. Foley in 2003 for the Wadi ath-Thamad Survey.

12 The ancient name of Khirbat al-Rumayl remains in dispute; for a discussion of the location of sites in this region that are mentioned in biblical texts and in Eusebius, see Dearman (1989b), MacDonald (2000), and Vieweger (1995). The location of al-Rumayl is given in JADIS (1994) as PGE 233000, PGN 109400; JADIS # 2310.002. The readings taken by Foley are UTME 772801, UTMN 3496932.

13 It is at the point where Wadi Za faran flows south and meets Wadi ath-Thamad that the name Wadi Rumayl is used, although it is just the continuation of Wadi ath-Thamad.

14 This description reflects the situation in 1996 (field report of J. Andrew Dearman); little change has occurred in subsequent years, except for continued erosion. Further downstream, a dam has been constructed on Wadi Wala, but this does not have an impact upstream in the valley below WT-13.

15 Carlos Cordova, then at Oklahoma State University, served as a member of the Wadi ath-Thamad Survey team in 1998 and 1999. For a report on land forms and prehistoric sites, see Cordova (2000; 2007) and Cordova et al . (2005).

16 Also known in earlier literature as Wadi Heidan (Eds.).

17 The Wadi ath-Thamad Survey is designed to investigate the Wadi ath-Thamad drainage system, including the Wadis Shabik, Rawiq, and Za faran. Such a survey will help to put the Iron Age and Nabataean settlements at Mudayna Thamad in a broader landscape, revealing the water and land management strategies in the drainage basin and the changing settlement patterns over time.

18 Sherds sent to Leiden for study and analysis are indicated as such.

19 Team members for all seasons are listed in the Preface.

20 Eight features had been identified during the 1996 season.

21 Square A6 included the southern 2.00 m of Square A1, which had not been dug below topsoil loci in previous seasons. The same was true for A5 in the southern sector of A2. These squares had been expanded to a length of 8.00 m in the first season, in order to document the location of Wall 1002 running east–west just south of the limits of A1 and A2.

22 An earlier version of the strata utilized by Dolan in her dissertation (2007:106) served as the basis for analysis of artefacts and ecofacts by various scholars, even though it later became apparent that only two strata (IIB, IIA) represent recognizable cultic activity areas at the site.

23 The Wadi ath-Thamad Project is grateful to Gary Christopherson, who granted permission to use the 1996 Hesban Survey Manual and its recording sheets. The decision to use an existing system was made in order to record survey data in a standard format that could be compared with other sites in Jordan using the same or a similar recording system.

24 The locus sheets are a slightly revised version of those used at Tall Jawa with permission of the Madaba Plains Project. Soil, architecture, and installation sheets are included in the Wadi ath-Thamad Excavation Manual.

25 In order to distinguish finds from WT-13 from those excavated at Mudayna Thamad during the same season, the artefact siglum includes the site code (WT)+object # - season / artefact #. Although object numbers are continuous, beginning in 1999, artefact numbers for finds from WT-13 and from Wadi ath-Thamad survey sites, begin each season at 500 (WT 67-2/567). Objects from Mudayna Thamad begin with the site code (MT) and include an object #; artefacts of museum quality have the site code (MT)+object # - season / artefact #. At Mudayna, all artefact numbers are in the range of 1–499 in a given season (for example, MT 390-4/11).

In 1999, objects continued to receive an object number in the running list, but only well-preserved artefacts were assigned an artefact number, since only a certain number of these objects were chosen for the Madaba Museum at the division of finds. Beginning in 2001, all artefacts were presented to the Department of Antiquities, Madaba Museum. Following the change of policy by the Department of Antiquities of Jordan regarding objects, the Season+Artefact number was added in the database to each object assigned to the Madaba Museum from the 1996–1998 seasons.

26 Only a few object entries include the weight (Wgt), which is indicated in grams (g) for those artefacts that weigh less than 1,000 grams. Larger objects are weighed in kilograms (kg).

27 With the completion of the new Jordan Museum in Amman, a small number of artefacts from the Wadi ath-Thamad Project have been requested by the Museum and are now on display.

28 Along with the handful of Early Roman-Nabataean sherds, a stone with three Thamudic letters inscribed on one surface is assigned to the same period (WT 559; see Chapter 13 ).

29 In an initial ceramic study, Dolan (2007) studied 92 diagnostic sherds from secure loci. More recent study has noted that during excavation hundreds of bases were recovered, some of which are ambiguous in their assignment to a particular vessel and thus are less informative than rim sherds in determining the class and function of the vessels of which they were a part. Only a selection of these base sherds has now been drawn and tentatively assigned to their class in the online database.

2

Approaches to religious rituals and cultic behaviour

P. M. Michèle Daviau

Introduction

As Colin Renfrew realized when he excavated the sanctuary of Phylakopi on Melos (1985:1), the discovery of a site with cultic artefacts presents a challenge for both the excavators and the interpreters of the data. The excavation by the Wadi ath-Thamad Project of two cultic sites within 3.0 km of each other (Temple 149 at Mudayna Thamad and WT-13), each with a distinct assemblage, was an even greater challenge; first to properly identify the cultic structures at these sites and then to design strategies for their excavation, recording, mapping and interpretation. In our identification of these sites and their cultic assemblages, we benefited greatly from the theoretical approach developed by Renfrew, consisting of criteria suitable for the recognition and explanation of data from sites for which there are no contemporary texts to illuminate the beliefs and rituals of the ancient worshippers. This is not to say that there are no ancient texts which might prove helpful, only that there are no specific references in the literature to cultic buildings at these sites in Jordan or to their cultic paraphernalia. The archaeological remains at both sites were recognized as cultic on the basis of the presence of explicitly religious objects and, in the case of Mudayna Thamad, of a building (Temple 149) with features similar to other securely identified temples, such as the temple at Arad (Herzog 2001, 2002) and the more elaborate Temple 650 at Tel Miqne-Ekron (Gitin 2012: fig. 6). Secondly, parallels for certain figurines found at WT-13 with those from known cult sites in the Levant facilitated the designation of a number of objects as elements of cultic equipment.

At the same time, WT-13 has very little in common with Temple 149 or the temple discovered at Atarus, southwest of Madaba,¹ leading to the hypothesis that there is considerable variability in religious expression across central Jordan in the Iron Age. An even greater diversity in religious practice at sites in ancient Palestine west of the Jordan has been clearly demonstrated by Ziony Zevit in his 2001 study of the religions of ancient Israel. Zevit was able to review not only the full corpus of published sites, but also the results of recent, unpublished excavations which produced material culture remains related directly to religious activity. Based on the features and artefacts preserved in the archaeological record, he assigned to each site various criteria which he adapted from those of