Renewed Excavations

* Report of the 2015 Excavation Season

Duration: March 18– April 19 2015

Directors:       Noor Mulder-Hymans, Jeannette Boertien, Margreet L. Steiner

(all Groningen University, The Netherlands)

Representative: Rami Frehat

 Abstract

The Renewed Excavations at Tell Abu Sarbut (2012 – 2015) have greatly enhanced insight in the occupation of the Middle Jordan Valley. The site seems to have been occupied for the first time in Late Hellenistic or Early Roman times. A long sequence of walls and floor layers was excavated, signifying occupation from the first / second century BC till the second or third century AD. Kitchens with cooking installations, courtyards with bread ovens and small rooms show that these were domestic buildings.

The site had been abandoned possible as the result of an earthquake or large fire as the floors of the uppermost Roman buildings were heavily charred and a thick layer of burnt debris covered the remains.

The next phase of occupation can be dated to the Abassid period (8th-9th centuries), when a series of buildings was erected on top of the earlier debris. These buildings seem to be domestic in nature, with courtyards with bread ovens and small alleyways between the buildings. The main northern wall of the building still stood up to three meters high. It is as yet unclear how long this occupation continued and why it ended.

In the Mamluk period (1250 -1550) a new phase of occupation started at the west side of the tell, and gradually spread over the earlier ruins at the east side. This occupation was connected with the cultivation of sugar cane in the Valley. After 1550 the tell was never occupied again.

Chapter 1: introduction

Tell Abu Sarbut is located in the central Jordan Valley, approximately three kilometers west of Tell Dayr `Alla.

Grid reference 35° 36´ E & 32° 12´ N

Jadis 2017021

MegaJordan nr. 9494

Previous work

Between 1988 and 1992 a team from Leiden University, the Netherlands, conducted four seasons of excavations at Abu Sarbut. The aim of the project was to excavate a rural site from the Islamic period. Therefore, only the upper layers of the tell were exposed. The final report of this projects was published in 2008; seven more articles were published over the years, and two Ph.D. dissertations were written on the Abu Sarbut material – see list of publications below.

Earlier remains, excavated in one of the test trenches, consisted of a heavily burnt mud brick building, of which parts of three rooms were exposed. This building, situated in a rural area and surrounded by agricultural fields, could well be a farm house, estate or villa rustica. The state of conservation of the building was exceptionally good. The walls still stood to a height of one meter and the rooms are filled with objects (see photograph).

The pottery consisted of ordinary household wares (bowls, cooking pots, storage jars), which could provisionally be dated to the Late Roman period (2nd-3rd centuries AD). Some sherds of

Aim of the renewed excavations 2012-2015

  • Establishing the lay-out en context of the Roman occupation layers
  • Spatial and functional analyses of the finds
  • Excavation of the eastern top of the tell where a longer sequence of occupation was to be expected
  • Research into the economic context of the finds on various levels: buildings, site, region.

Significance of the project

So far archaeological research into the Romans period has, for obvious reasons, focused on military and urban complexes. Small sites, be they isolated farms or hamlets, have been noted in surveys, but have not been researched in any systematic way. Philip Freeman calls this: “the greatest gap in our knowledge” (2001: 439).

The Jordan valley was quite densely populated in Roman times, as indicated by the East Jordan Valley Survey (Yassine et al. 1988): in the southern part of the Valley 29 sites were settled, against 13 in the preceding Hellenistic period. Eva Kaptijn discovered several new sites with Roman pottery in her Survey of the Zerqa Triangle, but concluded that most were the remains of very slight occupation (2009: 259). Only Tell ad-Addliyeh and Tell Amatha, where several test trenches were excavated, yielded (stone) architecture, possible villa’s, from the Roman period (Petit 2009). A Byzantine cemetery was excavated by Diane Kirkbride in the 1960’s, but the results were never published (but see Kaptijn 2009: 252-6). A Roman road passed through the Jordan Valley, as is borne out by several mile stone discovered there (Mittmann 1970).

The renewed excavations of Tell Abu Sarbut could provide insight in the agricultural settlements in the Jordan Valley, thereby filling in “the gap” in our knowledge of the period. Recently an agricultural complex has been excavated at Khirbat al-Mudayna in the Wadi ath-Thamad, by one of the directors (Daviau et al. 2000). The still unpublished details of this building will be available to the current expedition for comparative purposes.

Chapter 2: techniques and methodologies

Several squares measuring 10 x 10 m have been opened at the site. These squares were excavated by hand, using the Kenyon-Wheeler stratigraphic method of excavation. Every deposit that could be distinguished in a square on the basis of colour, texture or composition, regardless of the origin of that deposit, was called a `layer’. In each excavation square the layers were numbered consecutively. So in square H the first layer excavated was called H 1, the next H2 and so forth. Walls, ovens and other installations were also numbered according to this system. Later the layers were grouped into phases and subphases, on the basis of their stratigraphic position and the finds in the layers. A phase includes all architectural elements and deposits in use or resulting of activities in a certain period. When the architecture or the lay-out of the site underwent a major change, a new phase started.

During the first excavation season in 2012 squares F on the eastern summit and H in the depression were excavated. In 2014 two new squares were opened: Square X west of square F, and square G west of square H. This season square N was laid out, east of square H. South of square F a small extension was excavated to search for the south side of an important wall.

Map 2015

Chapter 3: Results

1) Squares G and H-west (Eveline van der Steen)

In square G and the western part of square H work continued from last season. Several phases could be distinguished.

Square G:

Mamluk layers

G50, 59, 60

In 2014 a small room was excavated in square G. Adjoining this room to the East, in the East baulk there was an installation, G41, possibly a large furn or oven, with heavy burning and ash layers. This installation was excavated when the E-baulk was removed. The firepit was bounded by low, narrow mudbrick walls, two of which were found, partly burnt through. Inside were heavy layers of white ash and heavily burnt material. There were no finds giving a clue as to its function.

Abbasid layers

G52, 53, 58, 62, 64, 65, 67

Abbasid material was only found in wash layers. Some evidence of possible squatter occupation or food preparation (hearths or fires), and in the E-baulk remains of a burnt layer belonging to a building in square H (G62). In the baulk was a row of stones, N-S, belonging to this period (G67), but with no related features.

Early Roman phase I

Walls / installations G33, 40, 54, 75, 80
layers G55, 56, 57, 61, 63, 66, 68, 69, 70, 71, 72, 74, 76, 78

In 2014 a room was found (walls G33, 40, 54) with heavily burnt material. The fourth wall was found this year in square H (H161). Walls H161, G40 and G54 had low benches. Inside the bench adjoining wall G54 was a built-in installation consisting of cylindrical plastered ‘pipes’. It has been interpreted as a possible water conduit (G75). The burnt material in the room was further excavated (G56, 70, 72). North of wall G33 and partly in what has been interpreted as a doorway was also burnt material (G55, 57, 63, 71, 76, 78), with much broken but complete in situ pottery, mostly large water jars (G72, 78) – see photo. Below the burnt layer were several earlier occupation layers belonging to the building (G61, 66, 68, 69). The floors in square H seem on average ca 40 cm higher than those in square G, which may be the consequence of the tell sloping to the W.

Early Roman phase II

Walls and installations G73, 77, 81, 87, 90, 91, 92
Layers G79, 82, 84, 85, 86, 88

Two walls, one E-W (G81) and a connecting wall to the S (G87) were found in the E-W trench. To the S and W of these walls was a heavily burnt layer (G88). To the N and E was a sequence of wash and rubble layers (G79, 82, 84, 85, 86) on top of a surface with three bread ovens on it (G90, 91, 92). Two small installations, consisting of a circular depression with dense white ash in it, were found to the north of this wall (G73), and to the east (G77) but may date from the time after the walls went out of use. Walls G81 and G87 were leveled and Roman phase I built directly on top of it (level ca -5.00 m).

Early Roman phase III

Walls and installations G94, 96
Layers G89, 93

In a small 2×2 m trench in the NE corner of the square another N-S wall was found, G96, with possibly a doorway in it. A heavy layer of surfaces and possible floors ran against it (G93) and an installation was found in this layer, next to the wall, consisting of a circular depression, ca 1.50 m diameter, filled with a layer of large stones and ash (G94). On top of this was a layer of mudbrick rubble and wash (G89), which was leveled for the construction of the next occupation (Roman phase II).

Earlier occupation levels in the small trench

Below the surface on which wall G96 stood (Roman phase III) was a heavy layer of wash and surfaces, with possible floors and much burnt brick, but no excavated structures (G95). This layer covered a N-S wall, G98, which was only preserved two mudbricks high, and a layer of wash and rubble (G97). It stood on a thin surface of heavy grey clay. Below this was another layer of wash layers, with little pottery (G99).

Square H-west:

In 2012 the west trench in square H was cleared down to level -3.20 m. In 2015 work continued there and two occupation phases were excavated.

Abbasid phase.

Walls H25, H157, H164.
layers H150, 151, 152, 153, 154, 156, 163.

Wall H25 (East-west), H157 and H164 are three walls of a niche which is open to the north. Bottom level ca -4 m. Inside the niche, and extending to the north are several layers of wash and rubble, on top of an irregular (tell?) surface: H154, 153, 156. H152 is a burnt layer or surface, possibly related to a bread oven (several fragments were found). Whether the walls were still in use at the time is unclear.

Early Roman phase.

Walls H161, 162, 165, 171, 85.
Layers H155, 158, 159, 166, 167, 168, 169, 170, 172.

Parts of a relatively large domestic (?) building were found, connecting to the uppermost building found in Square G in 2014 and 2015. Wall H161 is the east wall of a room which straddles both squares, and which was filled with heavily burnt rubble (H155). The south wall of this room is H171 (equals G54). To the east of this room is another space, bounded by H161 on the west, H162 to the north, H82 to the east and H165 to the south. In this small room large quantities of broken pottery were found in a heavy rubble layer (H169).

On top of this was a layer of heavy brick rubble (H166), and on top of that a layer of rubble and wash with burnt material in it (H158). This was directly underneath the Abbasid walls, and may have been a leveling layer.

To the north of this room was another space, bounded by H162 to the south, and H161 and H85 to the west and east, respectively. This space was filled with burnt rubble and wash layers, H159, 168.

H160 is a pit, which was dug through the Abbasid and partly through the Roman layers. It may date to the Mamluk period.

2) Squares H-east and N (Noor Mulder Hymans)

Square H

Square H was reopened in the East part of the North Trench, where the former Oven H9 was excavated. The Bottom of the H9 Oven stays at -3.01 on an Abbasid Surface (Phase 1).

Underneath and in front of H9 three more Oven body rings (H43, 44 and 45 at -3.51m) from the early Roman period (Phase 3) were found, documented and excavated.

This season our goal was to excavate further layers under and around the oven rings to see if there are more Ovens in lower layers. The Ovens, so far, are from the Tannur type with a diameter of 0.70-0.80m. To the north of these oven rings a surface (H78) came up with small Roman sherds.

This is the surface on which the oven body rings of H43 and H45 were standing. Samples were taken from the inside (H76=H42) of H43: soil with ash and pieces of the Oven body ring. It will be used for further research of the clay type, the inclusions and flotation will tell us if organic material is in it. On the surface H78 (-3.51m)a clear imprint was visible from the oven ring (H43) and the burned material inside. The oven was still 0.35m high. The body ring H45 probably belonged to Oven H43 and was a piece which slided down from the oven.

The body rings were documented, samples were taken.

Another small oven (H77) with a diameter of 0.32m was found in the S-E corner of the North Trench with the lower part almost complete. It was still standing at a height of 0.27m. An extra layer of clay was put against the outside of the oven body ring to strengthen it. Part of the Oven body ring was still in the east baulk, but later excavated. The Early Roman Oven (Phase 3) was built under the destruction layer (Phase 2) and stood at -3.48m on the surface H78. A few more layers were excavated and the last locus reached was H95 at -3.86, a soil locus with a lot of burnt material in the N-E corner.

Square H, the North Trench was extended 2 meters to the south along the east Balk and 4m to the west. The top soil has been taken off and at -2.49 a half circle of an Oven came out, H83. The Oven is still 0.30m high, stays on 9 stones, which are placed at-2.80m on a Mud Brick Floor (H89) and is from the Abassid period (Phase 1). The other half circle stood in N20.

At 1.70 from the west an Abbasid Mud Brick Wall, H94, runs 2m from N-S at -2.68m. East of this wall a Mud Brick Floor, H89+N29, is bonding with the wall and was spreading out towards the Oven. Below the wall in two small trenches a burnt Mud Brick layer was reached at -2.99 to -3.05m. This burnt layer(Phase2) is also the layer on which the Oven H9 stood.

Square N

Our second goal this season was to open the adjacent Square N, east of Square H in order to find an enclosing wall for the Oven area or a “courtyard”. For this reason the 5x5m square N was opened east of Square H. In the 5x5m square N the Abbasid Phase (1) has been excavated until the Roman Phase (3) was reached.

The east Balk of 1.00m wide at the North Trench of H was already excavated in the previous season so Square H and N were already connected. A 1m E-W and 3,5m N-S Trench N5,8,15,19 adjacent to the east balk of Square H was excavated. In the Abbasid locus N15 an unknown clay formation was found. The last locus was a hard packed surface/floor N19 at a depth of -3.29 with Early Roman Pottery. This Floor is connected with N25 and N33.

Under the topsoil almost in the middle of the square two complete Mamluk pots were found, but no habitation level. Later it was clear that there was a Pit underneath, N17 with lots of Mamluk and Abbasid Pottery. The Pit had a diameter of 1.85m which took a big part of square.

In the S-W part of N the other half of the Abbasid Oven H83=N20 (top: -2.49, bottom: -2.78m) and the Mud Brick Floor, built against the Oven was excavated. The diameter of the Oven is 0.70m.

Another Oven N22 (Early Roman, phase 3) was standing against and in the East balk near the S-E corner of square N. The oven is still 0.20m high and the bottom is at -3.25. The Oven is partly broken of and measures 0.50mx0.50-0.60m. It stays on a very hard packed surface/Floor, N25at -3.18m=N33 which covers the rest of square N. A burnt layer runs over the Oven (Phase 2).

More to the north of the square a column (diameter=0.30m) with two drums stood on the same hard packed Floor, N33=N25 at -3.17m. Small fragments of Roman Pottery came out.

East of Oven 20 a very broken fragment of another Roman Oven appeared at -3.33m with a diameter of 0.50m. The Oven was cut by the Mamluk Pit N17.

Except for the Abbasid wall in H94 another wall showed up in the E-W baulk. It starts at c. 0.50m from the south-east corner at a depth of -3.13m and runs up to -2.48 at 4m to the west where it stops and is visible as a 0.60m wide wall in the east section around the corner where an extension has been set of 2x2m.The Mud Brick Floor N89 is at -2.83 adjacent to this wall. On the other side of this extension adjacent to the Mud Brick Floor N29 a Mud Brick is visible in the surface but not excavated. Another Pit N28was found in the extension, but not excavated.

Field phasing of squares H-east and N

Field Phase 1

In all trenches post-Roman wash and debris layers were encountered. From the Abbasid period some mud brick walls, surfaces of hard beaten earth, a Mud Brick floor, bread Ovens and two drums of a slim Pillar were excavated in Squares H and N. In square N several pieces of Abbasid Pottery with the “swirl” design were found. In Pit N17 many fragments of the Mamluk sugar pots and painted geometric ware were found.

Field Phase 2

A layer of heavily burnt debris was encountered in all trenches of Square H and N. The layer consisted of burnt mud bricks, ash pockets, fragments of charcoal, very red burnt debris and lighter red soil. It is interpreted as burnt roof material, burnt walls, and general debris, originating from the destruction of the Early Roman building.

Field Phase 3

Below the surface of Field Phase 2 a series of earth and debris layers, each deposited on top of a beaten earth surface, was found. It belongs to the Early Roman period.

Interpretation

In square H and N the remains of an Abbasid and Early Roman-period “Courtyard” with a series of surfaces, a few wall remains, pits and ovens were located. Over the Roman remains a layer of heavily burnt debris was found in square H, North Trench, but that tapered away in square N. The destruction could have been caused by a fire or a possible earthquake.

Dating: Based on the finds the Roman layers may be dated from the 1st century BC – beginning 2nd century AD. The Abbasid layers must be further studied to date.

3) Squares X and F (Jeannette Boertien)

On the 21st of March 2015 the excavations in Square F and X were restarted. Originally square F was opened in 2012 and excavations continued in the southwestern trench in 2014. In 2015 this trench was reopened measuring 5m x4.50m.

In square X the excavations started in 2014, with a 5×5 m trench in the Southeastern part of the square, and on March 21st 2015this trench was reopened. Because of four graves that could not be excavated the trench measured only 3×5 m.GravesX43 and X40 were found in 2014. This season another grave (X60) was found directly east to tomb X43. The small grave measuring 0.80×1.00m was not excavated. An additional grave or the continuation of grave X40 could be seen in the South section of X.

The different graves and the pits in the northern part of square X made it impossible to dig larger parts of square X. In order to be able to reconstruct the situation it was decided to concentrate the work on the Eastern part of square X combined with digging deeper in the 2014 Southwestern trench of Square F.

The eastern baulk of square X was excavated unto -1.50m, that isthe height on which the excavations in 2014 stopped and at this level the first Abbasid phase was found in both squares X and F. In the baulk a continuation of the Mamluk stone structure X54, X56, X58 and pits X57, X62, X63 were found. These layers are part of the Mamluk plastered room and the courtyard with ovens found in square X in 2014.

At the Southeastern edge of square X the top of a mudbrick wall became visible; this is the continuation of wall F33 running E-W all over square F. Parts of wall F33 were already found in 2012 and 2014(originally labeled as F111 in 2012 in the Eastern trench of square F). In square X the wall ends after 0.90m.

Wall F33 is 0.60m wide, running E-W all over Square F and it continues 0.90m into square X. Wall F33 could be associated with floor layers F31/F34 at -1.44m onwards to floor F200 and the soil layer F205 underneath this floor at -3.25m.

The wall stood 2.25m high (-1.00m to -3.25m) and 20 layers of mudbricks could be distinguished at the East side,the wall was built on a layer of dark soil at -3.25 to -3.47m.

There was another wall of the same size and depth wall Fex11this wall abuts wall F33 in the south and runs N-S. It was 0.60m wide and it stood 2.47m high (-0.75 m to -3.22m).

Four Abbasid phases could be connected with both huge walls F33 and Fex11.

Abbasid Field phase I

In square X the first Abbasid phase consisted of a courtyard with two ovens.In 2014 the first Abbasid floor (X30) was found in the middle of the trench. Floor X30 was made of hard yellow clay and was running up to the Northeastern corner of the trench. In 2015 more floor layers could be traced and two ovens were found. Oven X65 situated to the North section. The oven was oval in shape and measured 0.90×0.70m, it was dug into the yellow hard floor X67 from -1.48 to -1.69 m. To the South of oven X65 was cement floor X51 (part of which was already found in 2014) and at the Southern edgeof the courtyard was mudbrick wall X71 running E-W all over the square. Wall X71 abuts Wall F33 in the east and runs 5.25m to the West where it was also visible in the West section. Wall X71 was made of fine yellow clay, the mudbricks measuring 0.36×0.36x018m. The wall was 0.54m wide and the top of the wall was visible from -1.64m onwards ending at -3.13m. To the South of wall X71 was a small patch of burnt debris X76, the burnt layer was around oven X 72, this oven was 0.60m wide and only 0.17 m of the construction could be seen in the South section. Exactly on the edge of Square X to square F East of oven X65 was mudbrick wall X79 running N-S, 0.40 m wide and 1.70m long.

Floors X51 and X67 and X64 were built on a yellow layer of soilX66 containing many sherds.Floors X67 and X51 were constructed as the floors in square F. These floors are made of hard cement or hard yellow earth on a layer of sherds and soil. This makeup layer of is made of sherds from the first century AD, apparently retrieved from old layers situated in other parts of the tell. Underneath the sherds was a stabilization layer of soft yellowish/grey soil. X66 issuch a layer situated in the eastern part of the trench where a 0.40m wide wall X79 ran N-S between. The wall was 1.70 m long running N-S dividing squares X and F.

In Square F the Abbasid I layer was found in 2012 consisting of two yellow mud floors (F5/F11, F14) and two grey cement floors (F17 and F25) on top of each other, with Abbasid pottery on top of them. Floors F17 and F25 were made of hard cement-like material with small pottery sherds inlaid in the cement. Under these cement floors a fill or make-up was laid, consisting of layers of sherds. The sherds included Early Roman wares, taken from elsewhere on the tell. At the south side of the square mudbrick wall F33/F111 runs E-W and Wall Fex11 abuts it at the South.

Oven X65 in floor X67

Abbasid Field phase II

In square F the second Abbasid phase consisted ofWall F33 in the South and to the north were two rooms divided by a wall (X145) with a doorway at the Western side. The floors were made of hard cement and constructed as described above. To the East in square X was a courtyard between two walls. In the north stood wall X80 it was 0.60m wide, running E-W all over the square. To the north of wall X80 against the north section was a small patch of soil (X82). In the southern part of the courtyard was wall X71 also running E-W all over the square. Wall X71 abuts wall F33. At the south side of the square mudbrick wall F33/F111 runs E-W and Wall Fex 11 abuts it at the South.

Abbasid Field phase III

Phase III showed a different layout. In square F Wall F33 still existed but a two layers high stone wall (F184) was built unto it. The stones stood 0.50m high and a yellow mudbrick wall (F167/F173) was built unto it. This wall was as high as the stones and the yellow clay used for the wall covered the front and upper part of the stones. Apparently forming a bench all over the length of the wall. At the western edge 0.20m to Square X stonewallF184 stopped as did the mudbrick construction/wall F167/ F173, while the mudbrick wall F33 continued 1.00m into square X.

In F the dividing wall F145 did not exist anymore, and three floor layers (F172, F175 and F177 could be distinguished and associated to Wall F167/F173 in the South. A 0.80m long wall F214 was visible in the East section. The wall was situated in the Northeastern corner of the square. Because wall F214 could not be excavated and the section did not reveal what was to the north of it, this part of the square could not be interpreted.

To the east in square X a new wall X89 was built to the north of Wall X80 that still existed. Wall X89 was situated Southwest-Northeast forming a sharp 45° angle with wall X80. To the North of wall X89 was a small patch of soil in which many small tools were found such as rubbing stones and a spindle. Wall X100 ran E-W at the southern end of the courtyard it was only visible in the West section of square F. Another spindle was revealed in the southern part of the courtyard to wall X100 and the Eastern edge of Wall F33. At the south side of the square mudbrick wall F33/F111 runs E-W and Wall Fex11 abuts it at the South.

Abbasid Field phase IV

In this phase wall F33 was the southern edge of the squares. Wall X80 did not exist anymore making the courtyard broader(the floor layers ran up to wall X89 and wall X100. These layers showed heavy mudbrick tumble and soil ending on burnt layer X91.

In square F wall F179 divided the Northern and the Southern rooms. It seems that the Southern room was a corridor between walls F33 in the South and Wall F179 in the North. The small street or corridor was 2.00 m wideit was sequently paved with hard cement or hard yellow clay (floors F176,F181,F183 and F196 F204). The floor of the northern room (F192) was paved with a white and hard material. In the later part of this phase the floor (F194) of the Northern room was made of yellow clay and a pit (F193) was situated in the northeastern corner of the room. The pit revealed many sherds of (cooking)pots and bones.At the South side of the square floor F204 and the soil layers underneath being floor F200 and soil layer F205 could be associated with mudbrick wall F33 at -3.10 m to -3.25 m. In the extension floor Fex13 and soil layer Fex14 could be associated with wall Fex11 and Wall F33 at -3.11m to 3.22m. Wall Fex11 abuts wall F33 in the south and runs to the South.

New phase

Down -3.22 a new phase started in which wall F33 and wall Fex11 did not exist and some hard floor layers were found in the Southern part of square F (F207, F210) and inthe extension(Fex13 and Fex15).A yellow soil layer followed each of these floors with many sherds from the Early Roman period. The floors were not associated with wall F33 because it did not exist yet. This phase could only be traced in the southern part of square F and in Fex. The final height reached in the southern part of square F was -3.57m and in the extension it was -3.30m.The pottery from the soil layers and the construction of the floors in this phase is comparable to the Abbasid layers. The excavations stopped on April 19th 2015.

Wall F33 and wall Fex11.

Chapter 4: Discussion

Tell Abu Sarbut has shown a very fragmented occupational history. So far three main periods of occupation were attested at the site. The uppermost layers at the east side of the tell show several phases of Mamluk habitation, dating to 1250 – 1550 AD. In general this layer was not deeper than 50 cm. The earlier excavation (1988 – 1990) which had concentrated the the west side of the tell, dug more than three meters of Mamluk buildings. The uppermost remains visible at the surface of the tell in the form of stone walls are all Mamluk in date. In this period many large and small pits have dug in the surface of the tell, seriously disturbing the older layers.

Underneath the Mamluk occupation walls and floors of the Abassid period (8th-9th centuries) were found. At the eastern top layers from this period were at least three meters deep. West of the top the occupation was thinner, and consisted of some walls and floors only. The older excavation had not encountered this period, as the new Mamluk inhabitants had dug away the older layers.

Immediately underneath the Abbasid occupation a thick layer of Early Roman buildings was encountered. This layer has not been reached (or was not present) at the east side of the tell. In square G digging continued to -6,50 m. and still Early Roman or Hellenistic sherds were found.

No earlier or later material was found during the excavation.

In 2012 we reported that we had found remains (floors and a bread oven) dating to the Late Roman period. After extending the area of this occupation it turned out that these layers were Abassid in date, not Late Roman.

Chapter 5: Site evaluation

Tell Abu Sarbut is a large tell (ca. 250 x 125 m) which lies amidst agricultural fields, but is not suitable for agriculture itself because of the heavy clayey topsoil. The tell is generally undisturbed, except for the surface which is littered with plastic from the fields. The tell is private property and as far as we can see not in any immediate danger, except for the looting that takes place occasionally (see chapter six). No restoration or conservation is needed.

Chapter 6: Challenges

Ownership

Tell Abu Sarbut is private property. We have secured the cooperation of the owners of the tell and signed an agreement that allows us to use the tell in 2012, 2014 and 2015. According to this agreement we have filled in our excavation squares completely after the final season in 2015.

Looting

The tell has been severely looted in the years between the excavation seasons of 2012 and 2014, and again before the final season in 2015. Several deep to very deep holes were dug in the surface of the tell and in one of the excavation squares. We have at our own costs filled in most of these holes for safety reasons.

Two stone pillars were encountered on the top of the tell. The first one, a broke pillar of whitish stone, was lying on the surface of the tell in 2012, and was still present in 2015, although its location was different every year we came back. This pillar was not excavated by us and we doubt very much it belongs to Abu Sarbut, as hardly any stone architecture has been excavated so far. In 2013 a pinkish coloured pillar was found lying on the surface of the tell, but it had gone in 2014.

It is our hypothesis that these pillars were looted from other sites and temporarily `stored’ at Tell Abu Sarbut. We don’t know from which site the pillars came.

Recommendations

As the site is private property we have no recommendations as to its further development.

Information on the results of the excavations, both past and current, may be made available to the general public in the museums in Salt, Ajlun and the newly to built regional museum in Dayr Alla. We would appreciate it very much if our finds were exhibited there and are very willing to cooperate with the museum directors to provide background information on the tell and descriptions of the objects.

Conclusions

The Renewed Excavations at Tell Abu Sarbut (2012 – 2015) have greatly enhanced our knowledge of the occupation of the Middle Jordan Valley. The site seems to have been occupied for the first time in Late Hellenistic or Early Roman times, when the Jordan Valley was developed to provide food for the growing populations of the cities of the Decapolis, in this case Pella. A long sequence of Early Roman walls and floor layers was excavated, signifying occupation, possibly from the first century BC till the second century AD. More detailed study of the pottery is needed to refine the dating of the occupation and the function of the buildings.

The site had been abandoned in the second century, possible as the result of an earthquake or large fire as the floors of the uppermost Roman buildings were heavily charred and a thick layer of burnt debris covered the remains.

The next phase of occupation can be dated to the Abassid period (8th-9th centuries), when a series of buildings was erected on top of the earlier debris. These buildings seem to be domestic in nature, with courtyards with bread ovens and small alleyways between the buildings. It is as yet unclear how long this occupation continued and why it ended.

In the Mamluk period (1250 -1550) a new phase of occupation started at the west side of the tell, gradually spreading over the earlier ruins at the east side. This occupation was connected with the cultivation of sugar cane in the Valley. After 1550 the tell was never occupied again.

 

* Report of the 2014 Excavation Season

Project name: The Renewed Excavations of Tall Abu Sarbut

Directors: Noor Mulder-Hymans, Jeannette Boertien, Margreet L. Steiner (Groningen University, The Netherlands)

Duration: 16/3 – 24/4 2014

Representative: Rami Frehat

Goal

The goals of the 2014 season were:

  • Continue work on the eastern part of the tell (squares F and X) where remains from the Early Islamic and Roman periods have been found
  • Establishing the lay-out en context of the 3rd/4th century burnt building in square H and square G to the west of it
  • Extent the area of Early Roman remains underneath the burnt building
  • Start analysis of the pottery in cooperation with the Univ. of Groningen

From Wednesday March 19 till Monday April 22 work on the tell took place. The excavations continued in the squares that had been excavated in the season 2012: F at the east side of the tell and in Square H 20 m west of it. Two new squares were opened: squares X and G, both adjacent to the old squares. This season of excavations brought many new finds, the stratigraphy became much clearer, and the area of excavation was enlarged.

Plan Tall Abu Sarbut-ed  Plan of the tell

In the squares F and X, under some Late Islamic (Mamluk) occupation layers, more walls and floors of the Early Islamic period (mainly Abbasid) were uncovered, yielding many finds from that period as well as sherds and small finds from earlier (Roman) occupation layers that were taken from elsewhere on the tell and used as filling and make-up layers for the early Islamic floors.

Floor F155   Floor F155

All in all a ca. 1.80 m thick layer was excavated in which seven floor layers on top of each other could be distinguished: two yellow hard mud floors (F 5/11, F14) and five cement-like floors (F17, F25, F142, F149 and F155) inlaid with pottery sherds and each stabilized with two makeup layers: a layer of rather small fragments of earlier sherds laid over a layer of soil with some sherds and mudbrick pieces. The floors all ran up to two E-W mudbrick walls F33 and F37/145. Abbasid relief (Kerbschnitt), molded and combed wares were found on and under the floors as well as white pottery, jars with swirl decorations and painted wares.

Squares H and G yielded more material from the Early and Late Roman periods (walls, floors, heavily burnt roof debris) as well as some walls and surfaces from the Early and Late Islamic periods. In square G a room with several floor layers from the Mamluk period was uncovered, while in square H two mudbrick walls, one running E-W, the other N-S, could be assigned to the Early Islamic period. No floors could be discerned belonging to this structure, but an almost complete decorated Abbasid jar was found nearby.

Abbasidische kruik  Abbasid jar, partly restored

Under a heavily burnt layer running over both squares, remains from the Roman period were found. In square H a series of courtyards with bread ovens and some complete pottery vessels was excavated at the east side of the square, while at the west side burnt roof debris covered several floors. In between these two areas two subsequent mudbrick walls could be distinguished, signifying several distinct phases in the occupation, all belonging to the Early Roman period.

The room with burnt roof debris continued into square G, where two mudbrick walls were found. West of this room another courtyard with some bread ovens was dug. It seems we are dealing with two separate domestic units here, both with a courtyard with bread ovens, and that both units were in use for a long time in the ER period. In square H a deep probe showed that occupation continued to at least 5 m below the top of the tell.


* Report of the 2012 Excavation Season

Project name: The Renewed Excavations of Tall Abu Sarbut

Directors: Noor Mulder-Hymans, Jeannette Boertien, Margreet L. Steiner (Groningen University, The Netherlands)

Duration: 21/3 – 18/4 2012

Representative: Badr Aladwan                                                                                                                                                                                                                                                                                                                                                                                                                                

Goal

During the earlier campaigns (1988 -1992) the excavation focussed on the upper, Islamic, layers of the tell. Earlier remains, excavated in one of the test trenches, consisted of a heavily burnt mud brick building, of which parts of three rooms were exposed. The pottery dated the last phase of this building firmly in the Late Roman period (3rd-4th centuries AD).

The goal of the renewed excavations consist of establishing the lay-out and context of the burnt building, of spatial and functional analyses of the finds, as well as excavating the eastern top of the tell where a longer sequence of occupation from the Roman-Byzantine period is to be expected.

In this first season several trenches were excavated in two squares of 10 x 10 meter: squares F and H.

Square F

Square F is located on the eastern top of the tell. At the north side of the square (on the slope of the tell) several wash layers were found over some walls and a stone pavement, containing a mixture of Mamluk and Roman/Byzantine pottery. A trench at the south side of square F revealed several walls with three consecutive Roman floor layers in between.These floors were heavily disturbed by at least six later (Mamluk) pits that had been cut through them till a depth of 2.10 m.

At the west side of the square the situation was different. Here a layer with Roman and Byzantine pottery was encountered only 10-20 cm. below the surface of the tell. At a depth of ca. 1.50 m two mud brick walls, 80 cm wide, were running east-west and were associated with two floor layers made of cemented sherds and packed with Roman pottery as well as glass fragments, fragments of limestone vessels, a roof tile fragment and animal bones.

Square F – west trench

Square H

Square H is located in the lower area of the tell. Just to the west of this square the remains of the late Roman building had been excavated. The renewed excavations uncovered parts of the courtyard belonging to this building. Some five superimposed surfaces of packed mud could be excavated, all containing much pottery from the Early to Late Roman period, mainly cooking pots, jars, bowls and juglets. At the east side of the square an almost complete bread oven was found, underneath which several earlier ovens were visible at the end of the excavation season. Other finds include glass fragments, pieces of stone vessels, some coins and several iron nails.

Pottery

Bread oven

* General information Tell Abu Sarbut 

Tell Abu Sarbut is located in the central Jordan Valley, approximately three kilometers west of Tell Dayr `Alla.

Grid reference 35° 36´ E & 32° 12´ N

Jadis 2017021

MegaJordan nr. 9494

The tell measures about 250 m east-west and 125 m north-south. From the highest point,  -248 m, it gently slopes to -252 m at the east and south sides and to -255 m at the north and west sides.

Excavations 1988 – 1992

Between 1988 and 1992 a team from Leiden University, the Netherlands, conducted four seasons of excavations at Abu Sarbut. The aim of the project was to excavate a rural site from the Islamic period. Therefore, only the upper layers of the tell were exposed. The final report of this projects was published in 2008; seven more articles were published over the years, and two Ph.D. dissertations were written on the Abu Sarbut material – see List of publications

Earlier remains, excavated in one of the test trenches, consisted of a heavily burnt mud brick building, of which parts of three rooms were exposed. This building, situated in a rural area and surrounded by agricultural fields, could well be a farm house, estate or villa rustica. The state of conservation of the building is exceptionally good. The walls still stand to a height of one meter and the rooms are filled with objects (see photographs).  If the rest of the building is in the same state of preservation, the excavations could provide an unique possibility to reveal architecture as well as many objects in situ – an ideal starting point for stratigraphical, architectural and contextual studies

The pottery consisted of ordinary household wares (bowls, cooking pots, storage jars), which could provisionally be dated to the Late Roman period (2nd-3rd centuries AD). Some sherds of Roman luxury wares – red slipped wares and bowls with applied reliefs – came from the east side of the tell.

Aim of the renewed excavations

  • Establishing the lay-out en context of the burnt building
  • Spatial and functional analyses of the finds in the burnt building
  • Excavation of the eastern top of the tell where a longer sequence of occupation from the Roman-Byzantine period is to be expected
  • Research into the economic context of the finds on various levels: buildings, site, region.

 Strategy and planning

Tell Abu Sarbut is a tell with two “summits”, at the eastern and the western sides, with a “valley” in between – as the contour map shows (see below).  Roman remains have been encountered all over the site.

At the eastern side Roman/Byzantine layers were encountered at ca. 60 cm. below the surface of the tell, at – 248,70 m (below sea level). Here an occupation sequence of several meters of Roman/ Byzantine layers is to be expected.

At the western side Roman remains were found below several meters of Mamluk occupation layers, starting at -253,60 m. In the valley between the eastern and the western tops the remains of the burnt Roman building have been excavated, at ca. -252 m, underneath ca. 1 m of Mamluk wash layers.

The main fixed point on the tell, secured in concrete, is still in place. A contour map of the tell was made in 1988 and will be used again (see below). In 2012 we will concentrate on two areas: square F on the eastern summit and square H in the “valley”.

Contour plan of Tell Abu Sarbut with proposed excavation squares in red.

Square H

The Roman burnt building (farmhouse) is located in Square  H.  We will open up several trenches in this 10 x 10 m square to connect with the earlier remains and to outline the plan and size of the building. Then the Roman layers will be minutely excavated with attention to the find spot of every object to allow spatial analysis of the finds.

Square F

This square (also 10 x 10 meters) is located on the eastern summit. Most of the Mamluk layers have been excavated in 1988, so we can start to excavate the earlier remains without delay. Here some Roman luxury wares were found: a red slipped bowl and rim sherd with an application of a flute player.

Future seasons

Two or three subsequent excavation campaigns are foreseen for 2013 and 2014, depending on the result of the first campaign. Publications are scheduled for ADAJ (preliminary reports), while the final report will be published by Peeters Publishers as a follow-up of the final report of the earlier excavations.

Heritage management

The excavations of Tell Abu Sarbut are not only important to archaeologist but also to the local communities, because it concerns their history and their relationship with the past. Therefore we want to involve the inhabitants of the region in the project, not only as paid workers but also as concerned citizens and targeted audience.  Information on the results of the excavations, both past and current, will be made available to the public in different forms, such as information boards and show cases with excavated materials. Restoration of the excavated architecture will not be possible as the tell is private property and the excavation trenches will be backfilled.

Significance of the project

So far archaeological research into the Romans period has, for obvious reasons, focused on military and urban complexes. Small sites, be they isolated farms or hamlets, have been noted in surveys, but  have not been researched in any systematic way. Philip Freeman calls this: “the greatest gap in our knowledge” (2001: 439).

The Jordan valley was quite densely populated in Roman times, as indicated by the East Jordan Valley Survey (Yassine et al. 1988): in the southern part of the Valley 29 sites were settled, against 13 in the preceding Hellenistic period. Eva Kaptijn discovered several new sites with Roman pottery in her Survey of the Zerqa Triangle, but concluded that most were the remains of very slight occupation (2009: 259). Only Tell ad-Addliyeh and Tell Amatha, where several test trenches were excavated, yielded (stone) architecture, possible villa’s, from the Roman period (Petit 2009). A Byzantine cemetery was excavated by Diane Kirkbride in the 1960’s, but the results were never published (but see Kaptijn 2009: 252-6). A Roman road passed through the Jordan Valley, as is borne out by several mile stone discovered there (Mittmann 1970).

The renewed excavations of Tell Abu Sarbut could provide the first completely excavated rural Roman building in the Jordan Valley, thereby filling in “the gap” in our knowledge of the period. Recently an agricultural complex has been excavated at Khirbat al-Mudayna in the Wadi ath-Thamad, by one of the directors (Daviau et al. 2000). The still unpublished details of this building will be available to the current expedition for comparative purposes.

At the University of Groningen there is a strong collaboration between the different departments of Ancient History, Classics, Archaeology and Theology and Religious Studies. This collaboration is organized in CRASIS: a research and education institute dedicated to the interdisciplinary study of Graeco-Roman Antiquity.

Renewed excavations at Tell Abu Sarbut that aim to uncover the possible remains of a farmhouse from the 2nd-3rd century have the potential to contribute significantly to the interdisciplinary research on Late Antiquity conducted at the University of Groningen.

Staff

The project will be directed by Dr. Jeannette H. Boertien, Drs. Noor Mulder-Hymans and Dr. Margreet L. Steiner, associate researchers of the Faculty of Theology of the University of Groningen, The Netherlands.

Advisors include: 

Dr. Mladen Popovic, Qumran Institute, University of Groningen

Prof. Zeidan Kafafi, Fac. Archaeology and Anthropology, Yarmouk University

Dr. Christopher Tuttle, ACOR, Amman

Dr. Bram van As, Institute of Pottery Technology, Leiden University

Prof. Nabil Bader, Fac. Archaeology and Anthropology, Yarmouk University

Dr. Paul Belien, Numismatic Museum, Utrecht

Drs. Diklah Zohar, Leiden University

Publications referred to:

Daviau, P. M. M., Mulder-Hymans, N. and L. Foley, 2000. Preliminary Report of Excavations at Khirbat al-Mudayna on the Wadi ath-Thamad (1996–1999): The Nabataean Buildings, with a contribution by C. J. Simpson. Annual of the Department of Antiquities of Jordan 44:271-282.

Freeman, Philip, 2001. Roman Jordan. In: MacDonald, B., Adams, A and Bienkowski, P., The Archaeology of Jordan. Sheffield Academic Press, Sheffield: 427-60.

Mittmann, M., 1970. Beiträge zur Siedlungs- und Territorialgeschichte des Nordlichen Ostjordanlandes. ADPV, Wiesbaden.

Kaptijn, E. 2009. Life on the Watershed. Reconstructing Subsistence in a Steppe Region Using Archaeological Survey: a Diachronic Perspective on Habitation in the Jordan Valley. Sidestone Press, Leiden.

Petit, L. 2009. Settlement Dynamics in the Middle Jordan Valley during Iron Age II. BAR S2033, Archaeopress, Oxford.

Yassine, Kh., Ibrahim, M., Sauer, J., 1988. The East Jordan Valley Survey, 1976 (second part). In: Yassine, Khair, 1988. Archaeology of Jordan, Essays and Reports. Dept.of Archaeology, Univ. of Jordan, Amman: 187-206.