Goals
 |
Area R, the ER-LR village in the lower right of the
map, is also known as el-Herri.
|
During the 1994 Umm el-Jimal season excavation resumed at
area R, el-Herri, a 300
m diameter basalt rubble-strewn area immediately southeast
of the Umm el-Jimal town wall. The surface covering of random
basalt stones give the impression of a natural deposit, but
on closer examination numerous aligned rows are a clear indication
of building activities. In 1984 test excavation involved the
sinking of five small sounding trenches (fig. 1; de Vries
1993: 436-8, 443-4, 450-1) that provided the preliminary information
for the 1994 excavation strategy. Thus the 1994 work focused
on one of eight roughly circular and distinctly darker mounds,
which consist of an ashy black soil, basalt chips, and thousands
of early Roman and later pottery sherds. The goals were to
distinguish and date the architectural and occupational phases
of this mound, and to test the hypothesis that already existed
for these mounds, viz. that they were areas where masons finished
the basalt blocks robbed from early Roman buildings surrounding
them, so that the wall lines visible at ground level were
all that remained at al-Herri. (The assumption was that these
redressed blocks were the reused masonry evident in the construction
of the castellum
and later buildings at the neighboring town site.)
1984 probes
The five trenches dug in the UJ84 season were scattered throughout
el-Herri in an effort to gain a rough understanding of the
region, and each trench was placed in relation to an architectural
feature visible above ground. R.1 transected a wall in the
southwest sector of Area R.
 |
Pre-excavation photo showing the ashy soil of R.8
in the foreground and the rubble-strewn expanse of al-Herri
in teh background (photo by G. Hammink).
|
The probe uncovered two walls with a well-dressed threshold
block bonded into a wall. R.2 transected what was thought
to be a perimeter wall along the southeastern border. where
the tumbled debris of el-Herri ended at this point and an
area of terraced fields began. The lack of pottery on the
southern (outer) side of this wall seemed to support the theory
that this was the village enclosure wall. R.3 sought to identify
a village entry gate in the northeast of area R. Three walls
were found in the probe before bedrock was reached, but no
gate was uncovered. R.5, in the northwest quadrant, identified
a roadway edged by two walls built directly on top of each
other with the upper wall's much poorer manufacture. R.4 was
laid out on top of one of the black mounds in the northwest
quarter of the area and intersected two walls showing on the
surface. The probe uncovered three bonded walls and a tabun
in the area they enclosed in a stratified context of four
possible phases of occupation ranging from early Roman/late
Roman occupations to a Late Roman III-IV/Early Byzantine I-II
abandonment. Because of R.4's accessible location and the
evident domestic architecture under the mound, the probe served
as the starting point for the 1994 excavation.
 |
[above]Area R 1994 excavation grid showing extent
of architecture exposed (Drawing by Herve' Irion and
John Soh).
|
[below] South balks of R.4 and R.8 (Drawing by A.
Momani and B. de Vries).
|
 |
The 1994 excavation
A. The excavation grid
Initially a 10 x 10 m area was laid out and divided into
four 5 x 5 m squares: R.4, R.6, R.7 and R.8. One meter on
the north and east sides of each square was left as a balk.
Three weeks into the excavation it became necessary to expand
the operation by adding four more squares in hopes of uncovering
the entire structure. The squares R.9 through R.12 were added
along the edges of the excavation area, with a 1 m balk separating
each of the new trenches from the old.
B. Architecture exposed
At the end of the 1994 season virgin soil was reached in
only one square, R.4, while the others remained in stratified
context. During the excavation the walls discovered in 1984
were fully exposed and numerous other features were uncovered
in the other squares. By the end of the season it became clear
that these new walls combined together to form a building
complex of three rooms, possibly with a courtyard. The first
room, located in the southern half of R.6, proved to be a
complex feature made up of numerous additions from different
building phases (R.6:031, 002 and 007). Starting at this point,
one of the major walls in the complex runs northeastward into
square R.4, where it turns to the northwest and forms the
corner of the second room of the complex (R.4:002, 004, 006
and R.6:003, 032) . This same wall continues into R.8, where
it finally ends. The second room consists of three different
walls and an entrance at the southwest corner. Two of the
walls of this second room had either partially collapsed or
had been robbed to a much lower level than the large, northeastward
running wall starting in R.6. In R.8 two walls were partially
uncovered that probably form the third room along the northern
half of that square (R.8:015 and 002, which may extend in
the east balk of R.9 as R.9:003 wall collapse). The northern
face of a small section of a final wall was uncovered at the
end of the season in the western quarter of the southern balk
of R.7. Because it parallels the east-west running wall in
R.8 and seems to form an external wall of the complex, it
is possible that a central courtyard for the building complex
was contained by these two walls (the open area of R.6, 7
and 8).
With this plan in mind, it would seem that the inhabitants
of Umm el-Jimal at el-Herri built this building without an
advance plan. This can be seen from the numerous unbonded
walls spread throughout the complex. In R.8, where the corner
of the third room is formed by the two main walls, the third
wall seems to be built against the corner in order to form
the room. R.6 has a similar third wall, though it is unclear
at this time if it is built against or bonded with the main
wall. This piecemeal construction could indicate that the
building was constructed in separate units and expanded as
the peoples' needs changed.
 |
5:00 AM in Area R -- the beginning of the work day
|
All walls are constructed of the basalt
stone blocks typical of Umm el-Jimal using techniques
common to the area with two rows of aligned basalt field stones
with rubble fill between. This rubble serves as a core to
the wall while small chips and cobbles are used as chinking
stones in the wall facing itself to keep the courses even.
There is also evidence for use of a very pale brown mudbrick
material as a mortar in R.4 and R.8, and as fill between the
rocks in R.9.
By the end of the season five occupational phases could
be distinguished in area R. 1) The original early Roman structure
built above virgin soil; 2) late Roman reuse and construction
of new walls with the addition of the tabun in R.4; 3) late
Roman reuse and a second tabun in R.4; 4) early Byzantine
reuse of the earlier building and the construction of a new
wall, in addition to a period of abandonment with robbing
out or collapse; 5) dump activities and the deposition of
the topsoil layer.
C. Stratigraphy
- Phase I (stratum 7, early Roman): This phase represents
the earliest building activities conducted by the people
of Umm el-Jimal in area R. The two bonded walls forming
the northeast corner of the complex are clearly related
to this phase. The eastern wall, R.4:004, is preserved seven
courses high to a height of approx. 1.45 m while the northern
wall, R.4:002, varies from four to five courses in height,
approximately 0.95-1.20 m high. Wall R.4.002 extends to
the west as R.4:056 and R.8:017; wall R.4:004 extends south
as R.6:003.
Also, as appears in the R.4/R.8 south balk drawing,
the walls to the east (R.4:004) and west (R.4:006) of the
second room are built above a very compact coarse brownish
soil, R.4:057, devoid of artifacts and considered to be
virgin soil. The lowest soil above virgin soil, R.4:054=026,
a thin brown, ashy layer, included a few scattered pottery
sherds (early Roman/second century A.D.), and appears to
be associated with the eastern wall as the sherds were mostly
laid against the wall. On the basis of this wall R.4:004
and the wall sections bonded with it can be dated to phase
I. Thus the L-shaped wall complex running E-W through R.8
and R.4 and N-S through R.4 and R.6 belong to the basic
early second century structure that was embellished through
remodeling in later phases.
- Phase II (Stratum 6, late Roman, phase 1): Phase
II represents the reuse of the phase I architectural features
(the eastern and northern walls) along with the addition
of two new walls, which lie in R.6 and R.4. The first, R.4:006,
was added to construct the second room, and the second,
R.6.032, adds an entrance to the first room where tabun
activities were concentrated during this phase. The phase
I eastern and northern walls, R.4:002 and 004, served as
the walls for this new tabun room. R.4:006, the western
wall of the tabun room, is constructed against the northern
wall in R.4 without any bonding between the two, and runs
southwestward for about 2.25 m, paralleling the eastern
wall. R.6:032, bonded at a right angle with R.4:006, runs
southeasterly for about 1.00 m, so that a space of about
.70 m for a doorway is left between it and wall R.6:003=R.4.004
for access to the tabun room.
The western wall of the tabun room, R.4:006, is architecturally
later because it is butted against R.4:002, and because
it is much more poorly constructed than the phase I walls.
The wall belongs to phase II because of the association
with R.4:053, an occupation layer deposited on phase I layer
R.4:054 and consisting of slightly compact ash with small
cobbles, bone, glass, and charcoal pieces and pottery dated
to early and late Roman as well as early Byzantine period.
Thus the addition could be dated somewhere in the late Roman
and possibly even into the early Byzantine period.
The clearest representative of this period comes inside
the tabun room, a 2.25 m by 1.35 m space where two successive
tabuns were built during the building's life. The first
(lower) tabun consists of a ring of fired clay (0.85 m diameter)
protected and/or supported by a ring of medium to large
stones encased in unfired brick. Surrounding the tabun structure
are numerous layers of ashy soil, some with straw and decayed
brick inclusions. These layers could be considered contemporary
with the tabun, or they could be post occupational debris
layers related to this later tabun. Pottery finds date this
first tabun to the late Roman period. The second tabun was
constructed in the northeast corner of the room, directly
above the first, and also dated to the late Roman period.
Its 0.85 m dia. 0.04-5 m thick clay wall was preserved to
0.63 m high and also had a supporting structure of basalt
stones and clay built around the first tabun.
- Phase III (stratum 6, late Roman, phase 2): This
phase is only represented in the second (tabun) room. It
is nothing more than an accumulation of layers around the
second tabun of phase II. It consists of a clay rich orange
layer with some ash that extends over the southern half
of R.4. A fine, windblown loess layer that contains many
tabun fragments also covers the same area in R.4. It would
seem that this phase represents collapse of the tabun and
an accumulation of abandonment layers. All layers from this
phase relate to the Late Roman IV period.
- Phase IV (stratum 5, early Byzantine): This phase
includes the reuse of the phase I walls in R.8 (002 and
015) to form room three in the NW corner of the square.
A surface associated with the two walls, a hard packed ashy
layer, contains late Roman and early Byzantine sherds. In
addition, a high probability of association between this
layer and the hard plaster-like material found in R.9, just
to the north, exists, and it would also seem that the eastern
wall of room three extends to the north into R.9, where
it appears as a collapsed segment of the wall.
In R.6 a new wall segment, R.6:031, seems quite strange
compared to the rest of the walls uncovered in area R. It
is roughly built with large field stones in the upper courses
and smaller ones in the lower, though much of this part
is missing. This could be due to animal presence, since
the occupational layers associated with this phase seem
to be badly disturbed. As many as three uneven, multicolored
layers containing ash and charcoal, mixed with early/late
Roman and early Byzantine sherds are associated with the
wall. These layers also seem to correlate with other, similar
layers from R.7, R.8, and R.4. Thus phase IV represents
a period of early Byzantine occupation, most likely for
only a short time. This was followed by a time of abandonment
and robbing out or collapse of some of the walls.
- Phase V (stratum 5, early Byzantine [?]): This
phase consists of a complex and confusing series of soil
layers that are stratigraphically consistent throughout
most of area R. In addition to these soil layers, a section
of collapse from the east-west running phase I wall between
R.4 and R.8 can be included. In the shared balk of R.4 and
R.8, a section of large, collapsed basalt stones with a
loose, fine ashy soil throughout was uncovered. This tumble
is associated with the E-W phase I wall. Throughout the
soil and tumble, a mix of late Roman sherds along with a
handful of green-glazed sherds of unknown (Mamluk?) date
were found.
The multiple soil layers from this phase include varying
amounts of ash and charcoal, with large amounts of mud-brick
and tabun fragment inclusions. Altogether, more than four
distinct layers of loose, ashy soil lie across the whole
of area R. The layers also contain huge amounts of pottery
sherds, most dating to late Roman and early Byzantine, with
the occasional early Roman and rare occurrences of green-glazed
sherds.
This phase is composed of large scale dumping activities
of somewhat undetermined date due to the presence of the
occasional green-glazed sherds. That these sherds should
be dated Mamluk is unlikely, for they occur without the
more common Mamluk painted sherds, and they occur in an
otherwise exclusively LRom/EByz context.
 |
R.9 looking east; The balk on the left shows the
ca. 0.75 m accumulation of phase V ashy dump that
overlay the earlier occupation loci (bottom of the
trench and collapsed wall in far balk). (Photo by
G. Hammink).
|
The topsoil layer at area R consisted of a very fine, loose,
ashy soil with vegetation roots, small stones, a few rocks,
mudbrick chunks, tabun fragments, bone, glass, and large
amounts of pottery sherds from late Roman and Byzantine
periods. However, in the east of R.6 the soil does not contain
such large amounts of pottery sherds but instead contains
basalt blocks of all sizes that are either tumble from the
walls in R.6, or the remnants of dumping during a later
period. The few pottery sherds from this area also date
to Late Roman and Early Byzantine times. The topsoil in
R.11 and R.12 also shows a difference from the rest of area
R. The soil in these regions is much looser with small rocks
and only a few weathered sherds.
Conclusions
The 1994 excavation season in area R at el-Herri uncovered
a large building complex of three rooms and a possible courtyard.
In the first phase (ERom) of the complex the first walls were
constructed, as can be seen in the northeast corner of the
complex. The second phase (LRom) saw the reuse of the phase
I walls and the addition of two new walls found in R.6 and
R.4. Two successive tabuns were also constructed in the northeast
corner of the second room during this period. The third phase
(LRom) is an abandonment period seen only in the second room.
Phase four (EByz), the final period of occupation, again saw
reuse of the original phase I walls and, again, added another
wall as seen in R.6. The third room, in R.8, also saw some
use during this phase. The fifth and final phase is a period
of area-wide abandonment and dumping across the complex. The
date of this phase is unclear, but could be early Byzantine
unless the green-glazed sherds prove otherwise.
Though the season was an ultimate success, further excavations
are needed in area R in order to clarify the occupational
and stratigraphic sequence of the el-Herri area. More trenches
are also needed to solidify the dating and identify the source
of the phase V dump layers. This can be done by further excavations
of area R. More information must also be gathered on the Early
Roman/Nabataean occupation of el-Herri by completing the excavation
of all squares to the earliest stratum and possibly, excavating
another complex not covered by the thick layers of ashy dump.
Bibliography
de Vries, Bert
1993 The Umm el-Jimal
Project, 1981-1992. Annual of the Department of Antiquities
of Jordan 37: 433-60.
|
