El-Khirba: Food processing and other ground stone tools from a Roman, Abbasid and Mamluk period site near Nes Ziyyona, Israel

Full text

345
E. Adama et al.: El-Khirba: Food processing and
other ground stone tools
1� Introduction
The archeological research on ground stone tools has
grown significantly in the past three decades; however,
studies on periods ground stone tool assemblages of the
southern Levant are still rather few (see however e.g.
Frankel 2003a, 2003b; Frankel and Syon 2016; Rosenberg,
in press; Sidi 2007; Squitieri 2017). The site of el-Khirba,
situated in the southern fringes of the modern town of
Nes Ziyyona, Israel (map ref. 180350/647475), located in
the southern coastal plain, offers a great opportunity
to delve into some of the aspects of ground stone tools
of the late periods in the archaeology of the southern
Levant, following large-scale salvage excavations (2007,
2012-13) that were carried out by the Israel Antiquities
Authority.
Seven excavation areas were opened (overall area ca.
8000 sqm): in Area A, at the centre of the site, tombs
from the Roman period, a building from the Abbasid
period and irrigation channels from the Abbasid and
late Ottoman-Mandate periods were found. In Area
B a large refuse pit from the Late Roman period with
Abbasid-period tombs dug into it was found; to its south
and west, irrigation channels from the Abbasid and
late Ottoman-Mandate periods were also discovered.
In Area C, located on the highest point of the site, a
pottery kiln from the Hellenistic period, a tomb from
the Roman period, pottery kilns and structure from the
Byzantine period, structures from the Abbasid period,
and an irrigation channel from the late Ottoman-
Mandate were unearthed.
In Area D a refuse pit, pottery kilns and pottery
production dumps from the Roman period, a pottery
kiln from the Umayyad period, and irrigation channels
from the late Ottoman-Mandate period were excavated.
In Area E a structure and winepress from the Roman
period, structures from the Byzantine period, cesspits
and irrigation channel from the Abbasid period, and
a tomb from an undetermined period were found. In
Area F, located at the eastern-most edge of the site,
a structure, cesspit and refuse pit from the Mamluk
period were excavated, in addition to a building that
was built in the late Ottoman period and continued to
function until the second half of the twentieth century.
In Area G irrigation channels and a building from the
late Ottoman-Mandate period and, as well as tombs
from an undetermined period were unearthed.
2� Methodology
All the ground stone tools in the assemblage (Tables
1-7) underwent a comprehensive attribute analysis,
which included the recording of each tool’s typology,
morphology, degree of preservation, raw material,
metrics (e.g. length, width and thickness), and weight.
Macro and microscopic observation were applied in
some cases to document production and use-wear


classifications followed mainly Frankel (2003b), Frankel
and Syon (2016), and Sidi (2000, 2007). Most of the tools
in the assemblage (ca. 80%) were found in well-defined
loci and will be presented according to their periods.
3� The assemblage
The ground stone tools assemblage includes 128 tools
and tool fragments made of various rock types (Tables
1-2). These were found in the Roman (n=49, 38.3% of the
assemblage), Byzantine (n=2, 1.6% of the assemblage),
Abbasid (n=43, 33.6% of the assemblage) and Mamluk
(n=6, 4.7% of the assemblage) strata. The remaining
were found in contexts that were not securely classified
to any of the previously mentioned periods (n=28, 21.9%
of the assemblage).
3.1 The assemblage of the Roman period (n=49)
The assemblage (Table 3) from this period encompasses
primarily lower grinding stones (n=5), upper grinding
stones (n=3), unidentified grinding stone fragments
(n=2) and Olynthus mills (i.e. ‘hopper-rubbers’, ‘pushing
mills’, ‘frame millstones’ or mola trusatilis), including
the lower (n=9) and the upper elements (n=6). Vessels
(n=12), a pestle (n=1), hammerstones (n=4), varia (n=5),
and various tool fragments (n=2) appear as well.
El-Khirba: Food processing and other ground stone tools
from a Roman, Abbasid and Mamluk period site
near Nes Ziyyona, Israel
Erez Adama, Uzi ‘Ad and Danny Rosenberg
Keywords: el-Khirba, Nes Ziyyona, Ground stone tools, Roman period, Abbasid period, Mamluk period


346
Tools Types Period Total %
Roman Byzantine Abbasid Mamluk Indistinct
Lower grinding stone 5 2 1 8 6.3
Upper grinding stone 3 2 1 6 4.7
Unidentified grinding stone fragment 2 5 2 4 13 10.2
Lower Olynthian mill 9 6 2 17 13.3
Upper Olynthian mill 6 3 2 11 8.6
Lower rotary mill 1 1 0.8
Upper rotary mill 12 1 13 10.2
Vessel 12 2 3 5 22 17.2
Pestle 1 2 3 2.3
Hammerstone 4 1 1 7 13 10.2
Varia 5 4 3 1 13 10.2
Various tool fragment 2 4 2 8 6.3
Total 49 2 43 6 28 128 100�0
% 38�3 1�6 33�6 4�7 21�9 100�0
Table 1. Distribution of tools types and periods in the el-Khirba
ground stone tools assemblage.
Periods Raw material Total %
Vesicular basalt Compact basalt Scoria Limestone Flint Indistinct
Roman 12 23 6 2 6 49 38.3
Byzantine 2 2 1.6
Abbasid 28 9 2 4 43 33.6
Mamluk 2 1 1 2 6 4.7
Indistinct 5 14 1 3 4 1 28 21.9
Total 47 48 2 11 7 13 128 100�0
% 36�7 37�5 1�6 8�6 5�5 10�2 100�0
Table 2. Distribution of tool types and raw materials periods in the el-Khirba
ground stone tools assemblage.
Tools Types Raw Materials Total %
Vesicular basalt Compact basalt Limestone Flint Indistinct
Lower grinding stone 2 2 1 5 10.2
Upper grinding stone 2 1 3 6.1
Unidentified grinding stone
fragment 1 1 2 4.1
Lower Olynthian mill 3 6 9 18.4
Upper Olynthian mill 1 5 6 12.2
Vessel 1 8 3 12 24.5
Pestle 1 1 2.0
Hammerstone 2 2 4 8.2
Varia 1 4 5 10.2
Various tool fragment 1 1 2 4.1
Total 12 23 6 2 6 49 100�0
% 24�5 46�9 12�2 4�1 12�2 100�0
Table 3. Distribution of tool types and raw materials in the Stratum VII
(Roman period) assemblage.

347

3�1�1 Lower grinding stones (n=5)
The lower grinding stones from the Roman context
are made of vesicular (n=2) and compact (n=2) basalt;
a single lower grinding stone was made of an indistinct
material. Four of these were found in trash pits (Loci
172, 572, 246) in Area B, and one was found in a refuse
pit (L528) in Area D. All lower grinding stones are
broken, with both longitudinal and crosswise fractures.
Out of the five lower grinding stones, four are lower
grinding slabs; of these, two have a flat-convex cross-
section and were produced from vesicular and compact
basalt; the other two, classified as large worn (by
intensive usage) grinding slabs with an offset angle
edge, were also produced from vesicular and compact
basalt. The above grinding slab made of vesicular basalt
has two active surfaces bearing a few ochre stains. The
fifth lower grinding stone (Fig. 1) is a shallow and thin
palette made of an unidentified raw material. This
tool has a cluster of seven small cavities (each about
0.5 cm in diameter) on the side. All lower grinding
stones demonstrate abrasion marks.
are broken, one with a crosswise fracture and the
second with both longitudinal and crosswise fractures.
All three upper grinding stones are oval in plain view.
One of the tools made of vesicular basalt has a plano-
convex cross-section. The second tool is an elongated
loaf-shaped tool with a trapezoidal cross-section (Fig.
2.1), produced from compact basalt. The third upper
grinding stone (Fig. 2.2) is whole and made of vesicular
basalt (weighing 1090 gr) with a flat cross-section
resulting from bifacial grinding usage.
All upper grinding stones demonstrate distinct grinding
use-wear. In particular, one also shows slight abrasion
marks on the edge of the active surface. All three of
the upper grinding stones also bear production-wear,
indicated by several pecking marks on the surrounding
faces.
3�1�3 Unidentified grinding stone fragments (n=2)
Two unidentified grinding stone fragments were found
in Roman contexts. Both are made of basalt (vesicular
and compact). The first tool was found in a trash pit
(L573) in Area B, while the second tool was found in a
refuse pit (L528) in Area D. Both fragments are broken
with crosswise fractures and demonstrate overall
amorphous shapes, with two opposed active surfaces.
3�1�4 Lower Olynthus mills (n=9)
The lower Olynthus mills from the Roman contexts are
made of vesicular (n=3) and compact (n=6) basalt. Of
these, eight were found in trash pits (Loci 183, 187, 241,
246) in Area B, half of which were found in L246. The
ninth tool (Fig. 3.1) came from a fill (L537) in Area D.
All tools are broken, with both longitudinal and cross-
width fractures.
Five have two parallel flat active surfaces. In addition,
five tools have groups of 2-8 preserved ridges (e.g. Fig.
3.2) on one of their active surfaces, and a sixth tool
has ten preserved ridges on the active surface. The
poor preservation of the rest of the mills prevented
determining the provenance based on
ridge patterns (e.g. Frankel 2003b: 9). All
the lower Olynthus mills demonstrate
distinct abrasion and polish marks.
Further, all tools show clear evidence of
production and shaping of the periphery,
including chiseling (55% of the lower
Olynthus mills), pecking (44% of the lower
Olynthus mills) and battering (11% of the
lower Olynthus mills) marks.
3�1�5 Upper Olynthus mills (n=6)
The upper Olynthus mills are made of
vesicular (n=1) and compact (n=5) basalt.
Figure 1. Lower grinding
stone from stratum VII.
3�1�2 Upper grinding stones (n=3)
The upper grinding stones from the Roman context
are made of both vesicular (n=2) and compact (n=1)
basalt. All were found in trash pits (Loci 183, 187, 190)
in Area B. Out of the three upper grinding stones, two
Figure 2. Upper grinding stones from stratum VII.


348
Figure 3. Lower Olynthian mills from stratum VII.
Four tools were found in trash pits
(Loci 172, 183, 573) in Area B; the
other two mill fragments were found
        
(L528) and the second in a fill (L537).
Five of the mills are classified as
‘rectangular’ or ‘square hoppers’
(see Frankel 2003b: 7-8, fig. 4),
while the original form of the last is
undetermined.
All upper Olynthus mills were found
broken, with both longitudinal and
cross-width fractures. Five of the
mills have 3-8 preserved ridges on
their grinding surfaces, and the sixth
tool (EK18) has 14 preserved ridges
on the grinding surface (Fig. 4.1). The
Figure 4. Upper Olynthian mills from stratum VII.

349

low level of preservation interfered somewhat with
determining the ridges patterns, although the pattern
of tool EK18 matches the pattern of upper Olynthus
mills found at sites in the southern Levant, such as
Ramat Hanadiv, Gal’ed and Tel Mevorakh (see Frankel

maximum distance of ca. 70 km from one another. The
angle between the upper plane and the surrounding
side plane (henceforth ‘rim’, see Frankel 2003b: 11) is
preserved on five upper Olynthus mills. The formed
angles range between 96°-107° (average of 102°). The
rims also show production marks such as abrading,
polishing and chiselling.
Four upper Olynthus mills feature one preserved broken
‘handle slot’ (see Frankel 2003b: 12; Rosenberg, in
press), with a rectangular cross-section (average length
of 8.4 cm). Three slots display use-wear (polish), and
one slot bears possible chisel marks. Underneath the
handle slot, a special cavity used to secure the handle
was preserved and noted on only two of the millstones,
both of these bearing an oval cross-section. The upper
Olynthus mills have a shaped funnel, which was used to
direct the grains into the central slit and, from there,
to the grinding surface (see also Frankel 2003b: 5). The
preservation level of three mills allows the angles of
the funnels to be measured (relative to the rim); two
slopes from the rim are in a 142° and 143° angle, while
the third (Fig. 4.2) slopes at a 123° angle. The latter is
made of vesicular basalt.
3�1�6 Vessels (n=12)
The vessels from the Roman contexts are made of
vesicular basalt (n=1), compact basalt (n=8) and
limestone (n=3). All the vessels were found in trash pits
(Loci 183, 189, 241, 242, 246, 573) in Area B; however, one
vessel (EK82) was found in L241, which also included a
  
width fracture and four with both longitudinal and
cross-width fractures.
Out of the twelve vessels, six are shallow curved bowls
with short thick legs (Figs. 5.1-5.6), also known as
‘footed mortars’ (Sidi 2007: 550), ‘footed bowls’ and
mortaria (Hovers 1996: 177; and also Rosenberg, in
press), three are shallow bowls (Figs. 5.7-9) and three
are mortars (Figs. 6.1-6.3, see also Sidi 2007: 549, Pl.
6.1-6.5). The shallow and footed bowls (Figs. 5.1-8)
have an estimated rim diameter range of 22 to 45 cm.
The smallest of those (Fig. 5.5) is also the sole vessel
made of vesicular basalt. One footed bowl (Fig. 5.6) did
not preserve a full profile, so the rim diameter was
not measured. Of the three mortars (Figs. 6.1-6.3), two
have a preserved rim; one of these has an estimated
rim diameter of 9.5 cm (Fig. 6.1), and the other has
an estimated rim diameter of 16 cm (Fig. 6.2). 75%
of all examined rim profiles demonstrate a distinct
flattening (e.g. Masarwa and Barshad 2005); the other
(25%) rims are rounded.
The shallow footed bowl fragments preserved only
a single leg each. The leg cross-sections are either
square (n=3, Figs. 5.1, 5.3, 5.6), oval (n=2, Figs. 5.2, 5.4)
or circular (n=1, Fig. 5.5). Three of these footed bowls
(Figs. 5.1-5.3) exhibit thin incisions along the rim,
parallel to the preserved handle or ‘knob’; the knobs
are either an elongated rectangular shape (n=1, 3.6 cm
thick, Fig. 5.1), leaf-shaped with five petals (n=1, 4.5 cm
thick, Fig. 5.2), clover leaf-shaped (n=1, 3.3 cm thick, Fig.
5.3) or an elongated semi-hemispherical shape (n=1, 4.5
cm thick, Fig. 5.6), also classified as a ‘rounded knob’
(e.g., Masarwa and Barshad 2005) in most of the cases,
the handles hovered above the leg.
All vessels, including the three limestone mortars,
demonstrate distinct use-wear on the interior surface
such as abrasion (n=10), polish (n=2), and possible
pounding (n=4). The vessels demonstrate a variety of
production marks on their exteriors such as delicate
pecking and possible chiseling marks. One of the
mortars shows traces of drilling (see Ilan 2016). Most
vessels have crosswise body fractures (n=7), while the
remainder have a longitudinal body fracture (n=1) or a
damaged leg (n=1), rim (n=2), or knob (n=2).
3�1�7 Pestle (n=1)
The pestle from the Roman context is made of compact
basalt and was found in a trash pit (L246) located in
Area B. The pestle (Fig. 7) has one active end that was
used for grinding rather than pounding. The pestle
is conical, with both flat and slightly convex ends.
Production marks include smoothing on the body.
3�1�8 Hammerstones (n=4)
The hammerstones are made of limestone (n=2) and
flint (n=2). Three were found in a trash pit (L246) in
Area B. The fourth tool, made of limestone, is slightly
smaller than the others; it was found in a trash pit
(L573) in Area D. All tools show clear battering marks.
3�1�9 Varia (n=5)
Items included here could not be classified into a
formal tool type. These are mostly made of an indistinct
material, although one is made of vesicular basalt. Three
tools were found in trash pits (Loci 241, 246) in Area B;
one of these was found in L241 that also contained a
burial. An additional tool was found in a possible trash
pit (L9506) in Area C, and one was found in a refuse pit
(L528) in Area D.
Most of these finds (n=4) feature at least one broken
face. One object, smaller than the rest (10.4 cm x 8.1


350
Figure 5. Vessels from stratum VII.

351

Figure 6. Mortars from stratum VII.
cm across and weighing 375 gr), is flaked. Another
object has three smoothed surfaces with one face that
also features battering marks. Another item has five
modified surfaces, with an amorphous rounded shape
and opposed convex and concave planes. The fourth
object has three flat planes, with two parallel to each
other; the worked side is flat with several polished
areas. The last find is ovular, showing many battering
scars on its sides.
3�1�10 Various tool fragments (n=2)
Various tool fragments that could not be classified
typologically due to severe fragmentation were
included here. These are made of limestone and
undetermined material. Both were found in trash pits
(Loci 246, 573) in Area B. The limestone fragment is part
of a broken elongated tool that was later secondarily
used for grinding/smoothing and hammering. The
Figure 7. Pestle from stratum VII.


352
second tool is amorphous; on one of the tool’s edges
there are several battering marks.
3.2 The Byzantine assemblage (n=2)
The ground stone tools assemblage (Table 4) from
Byzantine contexts encompasses only two vessels made
of compact basalt. These vessels were found in Area C
  
were found, and the second (Fig. 8.2) in a fill (L5017).
The vessels are shallow footed bowls (e.g. Sparks 2007:
130), found broken. The original rim diameters are 36
cm and 50 cm. The rim profiles narrow towards the top
and have rounded edges. Both vessels preserve a single
leg, indicating that the complete vessels likely had
three legs. The legs each have a rectangular and circular
cross-section. All vessels demonstrate distinct use-wear
such as abrasion and polish as well as pounding on the
inner surfaces of the bowls. The vessels also exhibit
production marks like delicate pecking.
parts, vessels (n=3), a hammerstone (n=1), varia (n=4)
and various tool fragments (n=4). Most of the tools are
made of vesicular (65.1% of the Abbasid assemblage)
or compact (20.9% of the Abbasid assemblage) basalt.
The remaining tools are made of limestone and various
indistinct raw materials.
3�3�1 Lower grinding stones (n=2)
The lower grinding stones are made of vesicular basalt.
One was found in a structure (L730) in Area C, and the
other was discovered inside a cesspit (L9029) in Area
E. Both are broken, with longitudinal and cross-width
fractures. The tools are lower grinding slabs, one of
these a worn large grinding slab. Both lower grinding
stones have a single grinding surface. Use-wear includes
smoothing, polish and striation marks. Evidence of
production includes leveling of the passive surface of
one lower grinding stone and battering marks on the
exterior of the second tool.
3�3�2 Upper grinding stones (n=2)
The upper grinding stones are oval and made of
vesicular basalt. Both tools were found in structures
(Loci 730, 842) in Area C. They are broken across their
lengths and widths. Both upper grinding stones have a
plano-convex cross-section; one tool is elongated and
loaf-shaped. Use-wear includes intensive polish on one
of the tool’s grinding surface and a few striation marks
on the grinding surface of the other.
3�3�3 Unidentified grinding stone fragments (n=5)
The unidentified grinding stone fragments from
Abbasid contexts are made of vesicular (n=4) and
compact (n=1) basalt. Four of these were found inside
structures (Loci 730, 789, 816, 842) in Area C. The fifth
fragment was found in Area D. Four of the tools have
two active surfaces, and only one has a single grinding
surface.
Tool Type Raw material Total %
Compact basalt
Vessel 2 2 100.0
Total 2 2 100�0
% 100�0 100�0
Figure 8. Vessels from stratum VI.
Table 4. Distribution of tool types and raw materials in the
Stratum VI (Byzantine period) assemblage.
3.3 The Abbasid assemblage (n=43)
The assemblage (Table 5) from this context includes
lower grinding stones (n=2), upper grinding stones
(n=2), unidentified grinding stone fragments (n=5),
lower Olynthus mills (n=6), upper Olynthus mills (n=3),
rotary hand mills (also called ‘rotary hand querns’,
‘rotary mills’, ‘hand mills’, ‘rotary grinders’ and molae
manuariae) that include lower (n=1) and upper (n=12)

353

Tools Types Raw material Total %
Vesicular basalt Compact basalt Limestone Indistinct
Lower grinding stone 2 2 4.7
Upper grinding stone 2 2 4.7
Unidentified grinding stone fragment 4 1 5 11.6
Lower Olynthian mill 2 4 6 14.0
Upper Olynthian mill 2 1 3 7.0
Lower rotary hand mill 1 1 2.3
Upper rotary hand mill 12 12 27.9
Vessel 3 3 7.0
Hammerstone 1 1 2.3
Varia 1 3 4 9.3
Various tool fragment 3 1 4 9.3
Total 28 9 2 4 43 100�0
% 65�1 20�9 4�7 9�3 100�0
Table 5. Distribution of tool types and raw materials in the Stratum IV
(Abbasid Period) assemblage
3�3�4 Lower Olynthus mills (n=6)
The six lower Olynthus mills found in Abbasid contexts
are made of vesicular (n=2) and compact (n=4) basalt.
All were found in structures (Loci 709, 805, 842, 910,
971) in Area C, and one of these was, more specifically,
found between an installation and a wall (W13).
All of the tools are highly broken, the fragments
representing only a small fraction of the original tool.
Five of these seem to have two active surfaces (thus
were reused after fracture). Four of these secondarily
used fragments have between two and six preserved
ridges (on only one of their active surfaces, suggesting
that there was no attempt to form ridges on the second
later grinding surface). The level of preservation
inhibited determining the specific ridge patterns (e.g.
Frankel 2003b: 9). Abrasion and polish on the active
surfaces were noted in all examples. All the mills show
evidence of production marks, including smoothed and
worked faces and, less frequently, chisel marks on the
tools’ perimetres.
3�3�5 Upper Olynthus mills (n=3)
The upper Olynthus mills are made of vesicular (n=2)
and compact (n=1) basalt. All three fragments were
found in structures (Loci 842, 930, 959) in Area C; one
of these was found between an installation and a wall
(W13), and the other two were found close to a wall
(W945) and an installation (L771). The fragmentation
prevented sub-type identification, although all mills
include features that indicate they are ‘standard
upper Olynthus’ mills (Frankel 2003b: 8, 12, fig. 10:I.1).
One mill fragment presents a deep peripheral ‘rim’,
suggesting that this fragment lacks an ‘attachment
device’ (Frankel 2003b: 12, fig. 10:II.2, defined as a mill
without any handles). Another fragment is classified as
an ‘almost lacking hopper’, which according to Frankel
is a rare and unusual form (Frankel 2003b: 11-12, fig.
10:I.5).
All tools are broken, with both longitudinal and
crosswise fractures. The tools have a single grinding
surface. One tool has a concave grinding surface, while
the two tools have flat grinding surfaces. The grinding
surfaces do not have ridges.
All the mills have a preserved measurable angle
between the upper plane and the surrounding sides
(Frankel 2003b: 11), measuring 99°, 119° and 132° (with
an average of about 116°). Rims of two mills show
production marks (chiseling). One fragment has a
preserved damaged ‘handle slot’ (see Frankel 2003b: 12)
that is 3.2 cm deep. The standard upper Olynthus mills
have a shaped funnel used to lead the ground material
into the central ‘slit’ (Frankel 2003b: 5).
3�3�6 Lower rotary hand mill (n=1)
The lower rotary hand mill from the Abbasid context is
made of vesicular basalt. It was found between the floor
fill and the northern wall (W85) of a structure (L825) in
Area C. The tool (Fig. 9) was found broken, with both
longitudinal and cross-width fractures. The mill has
one flat active surface with another levelled parallel
plane. The tool shows distinct use-wear in the form of
abrasion and concentric striation marks, reflecting the
rotational movement of the upper mill stone (Fig. 10).
Based on the striations marks, the tool’s active surface
is at least 26 cm in diameter. Production marks include
battering marks and smoothed areas.


354
3�3�7 Upper rotary hand mill (n=12)
The upper rotary hand mills (also known as a ‘runner’
or catillus) are made of vesicular basalt. The majority
(n=11, Figs. 11.1-11.3) were found in structures (Loci
721, 731, 776, 842, 900, 902, 930, 938, 959, 998) in Area
C. Five upper rotary hand mills were found in the
surface matrix, two were found near walls (W945 and
W95), two were found near a furnace and one was
found between a wall (W13) and an installation (L771);
only one mill was found in a cesspit (L9056) in Area E
(Fig. 11.4). All tools were found broken, representing
a small fraction of the original mills. The majority of
the upper rotary hand mills (n=10) have one flat active
surface and a parallel flat upper plane. Two differ:
while both have a flat active surface, one has a slightly
convex upper plane, and the other has a slightly
concave upper plain.
Only eight of the mills have sufficient preservation,
enabling metric analysis. A group of six mill stones
are similar in size, with a diameter that ranges
between 33 cm and 39 cm, averaging at 34.5 cm (Fig.
11.4); the other two mills represent the diameter
range extremities, one 20 cm in diameter and the
other 55 cm in diameter. The average thickness
(measured near the edges) of all upper mills is 4.2
cm. Three mills preserved segments of the central
upright ring; these mills are 5.6-6.9 cm thick (average
of 6 cm).
Four upper mills preserved an indicative funnel-
shaped central perforation or ‘eye’ (Hauken and
Anderson 2014: 37-38). In three examples these are
conical, while in one example it is circular. Three
      
broken square shape and the other two an oval shape
(Fig. 11.2). All sockets are polished. One also has a
partially preserved drill mark, likely representing the
‘vertical handle socket’ (Hauken and Anderson 2014,
fig. 32:A). This socket is 2.4 cm in diameter and located
on the upper plane (Fig. 11.3), 9 cm from the edge of
the central ring (1.2 cm in depth). Rotational-wear
was evidenced inside the socket, suggesting rotational
the movement of a loose handle rather than a fixed
one. All the tools demonstrate distinct abrasion on the
active surface. The upper rotary hand mills also show
evidence of production, including mostly smoothed
surfaces but also some battering marks on the tool
peripheries.
3�3�8 Vessels (n=3)
The vessels are made of compact basalt and all are
broken crosswise, retaining small fractions of the
original tools. Two were found in Area C structures
(Loci 748, 940); one was found in the surface matrix in
Area C; the third vessel was also found in the surface
matrix (L9041) in Area E. Two vessels are shallow bowls
with flat bases (Fig. 12.1-12.2); the rims are convex to
flat-convex. Each bowl is semi-hemispherical with
one preserved ‘rounded knob’ each, like similar basalt

Masarwa and Barshad 2005: fig. 4). The third vessel is an
upright mortar, with slightly in-curving walls and a flat
base. The vessel has a concave internal base (Fig. 12.3)
and a rounded rim.
The vessels display use-wear such as polish on the
inner surface; soot marks were documented on one of
the shallow bowls. Production marks include delicate
pecking on the vessels’ exteriors; the inner surface of
the upright mortar also shows evidence of drilling.
3�3�9 Hammerstone (n=1)
The hammerstone is made of limestone and was found
above a plastered floor of a structure (L797) in Area C.
The hammerstone is oval, with battering marks on two
opposite edges.
3�3�10 Varia (n=4)
Items included here could not be classified to a formal
tool type and are made of indistinct materials (n=3)
and limestone (n=1). All were found in Area C, three in
structures (Loci 776, 735, 959). These include a pebble
bearing two opposite flat planes; on one of these, a
small shallow depression ca� 2 cm in diameter and 0.2
Figure 9. Lower rotary hand mill from stratum IV.
Figure 10. Striation marks on lower rotary hand mill


355

Figure 11. Upper rotary hand mills from stratum IV.
cm in depth was hewn (Fig. 13). Another item is oval
with one flat modified plane, showing signs of polish.
The remaining are quite amorphous, one with two
modified faces and the other with one modified surface,
both of these bear battering scars and levelled surfaces,
and, on one, several smoothed areas.
3�3�11 Various tool fragments (n=4)
Various tool fragments that could not be classified to
a formal type are made of vesicular basalt (n=3) and an
undetermined material (n=1). All were found in Area C,
three in structures (Loci 747, 794, 940). One was found


356
Figure 13. Varia object from stratum IV.
nearby a wall (W941), and another was found close to
a furnace; the remainders were found in the surface
matrix. One fragment has a 10 cm long modified
concave surface, with few striation marks and polished
areas. This fragment could represent part of a lower
grinding stone or Olynthus mill. Another fragment is
amorphous, slightly curvy and elongated, with few
smoothed areas. The third fragment is amorphous,
bearing a 3.5 cm long groove that ends near one of
the edges (possibly a piece of a rotary hand mill or an
upper plane of an upper grinding stone). The forth
fragment is rectangular and weighs nearly five kg.
This is possibly a broken anvil (e.g. Smith 1875: 534,
610).
3.4 The Mamluk assemblage (n=6)
The Mamluk assemblage (Table 6) encompasses
unidentified grinding stone fragments (n=2), a
hammerstone (n=1), and items classified as varia (n=3).
Figure 12. Vessels from stratum IV.

357

Tools Types Raw Materials Total %
Vesicular basalt Scoria Flint Indistinct
Unidentified grinding stone fragment 2 2 33.3
Hammerstone 1 1 16.7
Varia 1 2 3 50.0
Total 2 1 1 2 6 100�0
% 33�3 16�7 16�7 33�3 100�0
Table 7. Distribution of tool types and raw materials in the
indistinct contexts assemblage.
Table 6. Distribution of tool types and raw materials in the Stratum III
(Mamluk period) assemblage.
Tools Types
Raw material
Total %
Vesicular
basalt
Compact
basalt Scoria Limestone Flint Indistinct
Lower grinding stone 1 1 3.6
Upper grinding stone 1 1 3.6
Unidentified grinding stone fragment 2 2 4 14.3
Lower Olynthian mill 2 2 7.1
Upper Olynthian mill 2 2 7.1
Upper rotary hand mill 1 1 3.6
Vessel 5 5 17.9
Pestle 1 1 2 7.1
Hammerstone 1 2 4 7 25.0
Varia 1 1 3.6
Various tool fragment 1 1 2 7.1
Total 5 14 1 3 4 1 28 100�0
% 17�9 50�0 3�6 10�7 14�3 3�6 100�0
3�4�1 Unidentified grinding stone fragments (n=2)
The unclassified grinding stone fragments are made
of vesicular basalt. Both the fragments were found in
structures (Loci 609, 624) in Area F. Each fragment bears
two parallel worked and levelled grinding surfaces. A
few polished areas were noted on these, resulting from
grinding.
3�4�2 Hammerstone (n=1)
The hammerstone is made of flint and was found in
a structure (L610) in Area F. It is complete, though it
exhibits many flake scars.
3�4�3 Varia (n=3)
Items included here could not be assigned to a formal
tool type. These are made of scoria (n=1) and indistinct
rocks (n=2). All were found in structures (Loci 632, 638)
in Area F. Two items are broken and one is whole. One
of the broken tools has two worked and levelled parallel
planes. Both planes show small flake scars on the edges,
and on one of the planes there are six battering scars. In
the centre of that tool, there are two dents that are linked
(3.2 cm and 2.2 cm in diameter and 0.3 cm in depth).
The second broken tool has two parallel levelled planes.
The tool has a flaked and battered edge. The third tool
included here is a complete amorphous pebble, with an
elongated shallow channel-like depression.
3.5 Tools from indistinct context (n=28)
The assemblage from indistinct contexts (Table 7)
includes a lower grinding stone (n=1), an upper grinding
stone (n=1), unidentified grinding stone fragments
(n=4), vessels (n=5), pestles (n=2), hammerstones
(n=7), lower Olynthus mills (n=2), upper Olynthus mills
(n=2), an upper rotary hand mill (n=1), varia (n=1), and
various tool fragments (n=2). These are made primarily
of basalt.
4� Discussion
The ground stone tools of el-Khirba were found in
Roman, Byzantine, Abbasid and Mamluk contexts, with
additional tools found in indistinct loci and the surface.
Most of the ground stone tools were from Roman (38.3%


358
of the assemblage) and Abbasid (33.6%) contexts. The
assemblage is dominated by tools that can be linked
to food processing and preparation. Basalt dominates
the assemblage, and this mainly reflects the diffuse
preference for basalt when producing grinding tools.
The relative frequencies of vesicular and compact
basalt are similar (36.7% and 37.5%, respectively). The
majority of the tools in the assemblage were found
broken (94.5%), and thus we should be cautious when
discussing the significance of the tools’ find spots,
as this probably reflects discard rather than the use
patterns.
Most ground stone tools found in Roman loci were
associated with food production (over 70%); Olynthus
mills appear alongside ‘regular’ grinding tools that are
known from other Roman sites (e.g. Ayalon 2012: 126,
fig. 4:6-7; Sidi 2007: 554-555, Pl. 11.3-7). The Olynthus
mills show clear morphological and ridge pattern
similarities to mills found at other sites in Israel,
such as Horvat ‘Aqav-Ramat Hanadiv, Gal’ed and Tel
    
Kammon (Hartal 2012), Nahal Tut (Alexander 2006:
169, fig. 59:1-2), Tel Shunem (Covello-Paran 2010: 7, fig.
5), Kfar Otnai (Rosenberg, in press) and En-Gedi (Sidi
2000). The popularity of the Olynthus mills at these
sites and others during the Roman period suggests
an increasing need for intensive food processing. Of
the twelve vessels found in Roman contexts nine are
shallow bowls; six of the shallow bowls are footed.
Some of these have a preserved knob at the rim, which
was part of a larger set of knobs. Similar bowls were
found at many Roman sites such as Tel Ishqaf (Segal
         

and Kobrin 2016: Fig 18:4), Ar’ara (Massarwa 2007: fig.
4), Kafr Qari` (Masarwa and Barshad 2005: fig. 4) and
‘Ein el-Sha’ara (Kayesar 2013: 47-48, fig. 2:17). These
examples seem to share morphometric similarities,
suggesting that some kind of morphometric
conventions in footed basalt bowls existed during the
Roman period.
The Abbasid assemblage also shows a clear dominance
of food processing tools (91.5% of the assemblage);
these include grinding stones, Olynthus mills and
rotary mills. The nine Olynthus mills found in Abbasid
contexts seem to represent intrusions from the Roman
period as these are generally absent from early-Islamic
sites in the region, only used until the Byzantine period
in Israel (Frankel 2003a: 47) and up to the second half
of the third century in the eastern Mediterranean
(Frankel and Syon 2016: fig. 11.17). Notably, rotary mills
were used at el-Khirba only during the Abbasid period.
     
‘Opher Park’ (Kletter 2005: 92, fig. 26:4), Ras Abu Dahud
(Dayan et al. 2013: fig. 15), Ramla (Elisha 2010: fig. 8),

Kobrin 2016: fig. 9:4), Holot Yavne (Gorzalczany et al.

18), En-Gedi (Sidi 2007: 548-549, Pl. 5) and Elat (Rapuano
and Schick 2013: 157, fig. 25:1-2). This rather late
appearance of rotary hand mills is considered unusual,
since during the Byzantine period their popularity
grew and eventually they replaced the Olynthus mills
(Frankel 2003a: 46).
Three vessels were found in Abbasid contexts. Two are
shallow bowls, and the latter is an upright flat-based
mortar; the shallow bowls have a preserved knob at the
rim area, part of a set of knobs. Such vessels are found
at other early Islamic sites, including Holot Yavne
(Gorzalczany et al. 2010: 38), Horbat ‘Illin (Greenhot
 
fig. 25:6-7), Neve Ur (Shalem 2002: 170-171, fig. 17:1-6)
and Nahal Anim (Fraiberg 2017: fig. 25:12). Even though
the early Islamic vessels were found in low frequencies,
it is worth noting that their rounded rim profiles
and the absence of legs distinguishes these from the
majority of the Roman vessels.
5� Conclusions
Summing up, this report has offered a descriptive
account of the ground stone tools assemblage of el-
Khirba and insights into the diachronic trends of food
processing tools from the Roman into the Abbasid
period. It is clear that Olynthus mills were favored
during the Roman period, yet these were not fully
replaced during the Byzantine period as Frankel (2003a:
46) previously suggested. It is interesting to note that
while rotary hand mills were found at Roman sites and
coexisted with Olynthus mills during the first century
(e.g. Frankel 2003a: 46; Rosenberg in press), at el-Khirba
we see no evidence for rotary hand mills in Roman
contexts. The vessels show a continuous use of shallow
bowls from the Roman through the Abbasid period.
With the exception of three large mortars, the bowls
show a preference for compact basalt. Despite these
similarities, some interesting trends were noted: the
rim cross-sections change from flat to round; footed
bowls disappear during the Abbasid period and are
replaced by flat-based forms; decorative incisions
that were formed along the flat rim during the Roman
period disappear; the variety of curved knob shapes
decreases overtime, with only the semi-hemispherical
type used in later periods.
Acknowledgments
We wish to thank B. Tzin and R. Chasan for their fruitful
discussions, insights and help in refining this chapter
and the Israel Antiquities Authority for permitting the
preliminary publication of the data. Graphic assistant
by A. Regev-Gisis and drawings were made by M.
Shuisky.

359

Bibliography
        
Storage Depot from the Late Fourth Century BCE.
‘Atiqot 52:131-189.
Ayalon, E. 2012. Agricultural and Industrial Installations
and Burial Caves from the Roman and Byzantine
    ‘Atiqot 70:123-144. (Hebrew;
English summary 90*-91*).
Covello-Paran, K. 2010. Sulam. Hadashot Arkheologiyot
122, available online at http://www.hadashot-esi.
org.il/report_detail_eng.aspx?id=1339&mag_id=117
(accessed August 25, 2017).
Dayan, A., Shiff, M., Arbel, Y. and Gendelman, P. 2013.
Ras Abu Dahud. Hadashot Arkheologiyot 125, available
online at http://www.hadashot-esi.org.il/report_
detail_eng.aspx?id=2224&mag_id=120 (accessed
August 25, 2017).
Eisenberg-Degen, D. and Kobrin F. 2016. Nahal ‘Ashan,
Neve Menahem B. Hadashot Arkheologiyot 128,
available online at http://www.hadashot-esi.org.
il/report_detail_eng.aspx?id=25081&mag_id=124
(accessed August 25, 2017).
Elisha, Y. 2010. Ramla. Hadashot Arkheologiyot 122,
available online at http://www.hadashot-esi.org.
il/report_detail_eng.aspx?id=1501&mag_id=117
(accessed August 25, 2017).
     Hadashot Arkheologiyot
129, available online at http://www.hadashot-esi.
org.il/report_detail_eng.aspx?id=25257&mag_
id=125 (accessed August 25, 2017).
Frankel, R. 2003a. Mills and Querns in Talmudic
Literature — A Reappraisal in Light of Archaeological
Evidence. Cathedra 110: 43-60. (Hebrew).
Frankel, R. 2003b. The Olynthus mill, its origin, and
diffusion: Typology and distribution. American
Journal of Archaeology 107: 1-21.
Frankel, R. and Syon, D. 2016. Mills and Querns, in D.
Syon (ed.) GAMLA III The Shmarya Gutmann Excavations
1976-1989 – Finds and Studies Part 2 (IAA Reports 59):
85-96. Jerusalem: Israel Antiquities Authority.
Gorzalczany, A., Barkan, D., Iechie, L., Nagar, Y.,
Finkielsztejn, G., Gorin-Rosen, Y. and Ariel, D. 2010.
A Site from the Persian, Hellenistic and Early Islamic
 Atiqot 62: 21-46. (Hebrew;
English summary pp. 171*-172*).
Greenhut, Z. 2004. Early Islamic Remains at Horbat Illin
(Upper). ‘Atiqot 47: 15-32. (Hebrew; English summary,
pp. 210*-211*).
Haddad, E. 2013. Mishmar David. Hadashot Arkheologiyot
125, available online at http://www.hadashot-esi.
org.il/report_detail_eng.aspx?id=5434&mag_id=120
(accessed August 25, 2017).
Hartal, M. 2012. A Settlement from the Roman Period
‘Atiqot 70: 39-49 (Hebrew; English
summary, pp. 83*).
Hauken, A. D. and Anderson, T. J. 2014. Collection
Report: Rotary Querns in the Museum of Archaeology
University of Stavanger. Stavanger: University of
Stavanger.
Hovers, E. 1996. The Groundstone Industry, in D. T.
Ariel and A. De Groot (eds) Excavations at the City of
David 1978-1985 Directed by Yigal Shiloh – Excavation
Final Report Vol� IV (Qedem 35): 171-203. Jerusalem:
Hebrew University.
Ilan, D. 2016. The ground stone components of drills in
the Ancient Near East: Sockets, flywheels, cobble
weights, and drill bits. Journal of Lithic Studies 3(3):
261-277.
Kayesar, T. 2013. A Roman Tomb at ‘Ein el-Sha’ara.
‘Atiqot 73: 45-52.
Kletter, R. 2005. Early Islamic Remains at ‘Opher Park,
Ramla. ‘Atiqot 49: 57-99.
Masarwa, M. and Barshad, D. 2005. Kafr Qari. Hadashot
Arkheologiyot 117 available online at http://
www.hadashot-esi.org.il/report_detail_eng.
aspx?id=218&mag_id=110 (accessed August 25,
2017).
Massarwa, A. 2007. Ar’ara. Hadashot Arkheologiyot 119,
available online at http://www.hadashot-esi.org.
il/report_detail_eng.aspx?id=684&mag_id=112
(accessed August 25, 2017).
Netzer, E. 1991. Masada III: The Yigael Yadin Excavations
1963-1965 Final Reports; The Buildings, Stratigraphy and
Architecture. Jerusalem: Hebrew University.
Peacock, D. P. S. 1989. The Mills of Pompeii. Antiquity 63:
205-214.
Rapuano, Y. and Schick, T. 2013. An Early Islamic
Settlement and a Possible Open-Air Mosque at Eilat.
‘Atiqot 75: 129-165.
Rosenberg, D. Rural stone tools in the Jeezrael Valley:
The Late Roman/Early Byzantine stone tools of
Kfar Otnai, Israel, in Y. Tepper (ed.) IAA Reports (in
press).
Shalem, D. 2002. Nevé Ur—An Early Islamic Period
Village in the Bet She’an Valley. ‘Atiqot 43: 149-
176.
Segal, I. 2006. Chemical study of basaltic Hopper-
rubbers (Olynthus) mills from Nahal Tut and ‘En
Hofez. ‘Atiqot 52: 197-200.
Sidi, N. 2000. Roman and Byzantine small objects, in
Y. Hirschfeld (ed.) Ramat Hanadiv Excavations – Final
Report of the 1984-1998 Seasons: 177-186. Jerusalem:
The Israel Exploration Society.
Sidi, N. 2007. Stone utensils, in Y. Hirschfeld (ed.) En-
Gedi Excavations II – Final Report (1996-2002): 544-573.
Jerusalem: The Israel Exploration Society.
Smith, W. 1875. A Dictionary of Greek and Roman Antiquities.
New York: Franklin Square.
Squitieri, A. 2017. Stone Vessels in the Near East during the
Iron Age and the Persian Period: c� 1200-330 BCE� Oxford:
Archaeopress Archaeology.
Storck, J. and Teague, W. D. 1952. A history of milling: Flour
for man’s bread. London: Geoffrey Cumberlege.
Thurmond, D. L. 2006. Handbook of food processing in
classical Rome. London: Brill.


360
Williams-Thorpe, O. and Thorpe, R. S. 1991. Archaeology,
geochemistry and trade of igneous rock millstones
in Cyprus during the Late Bronze Age to Roman
periods. Geoarchaeology 6: 27-60.
Williams-Thorpe, O. and Thorpe, R. S. 1993.
Geochemistry and trade of eastern Mediterranean
millstone from Neolithic to Roman periods. Journal
of Archaeological Science 20: 263-320.

Stone Tools in the Ancient
Near East and Egypt
Ground stone tools, rock-cut installations and
stone vessels from the Prehistory to
Late Antiquity
edited by
Andrea Squitieri and David Eitam
Archaeopress Archaeology

Archaeopress Publishing Ltd
Summertown Pavilion
18-24 Middle Way
Summertown
Oxford OX2 7LG
www.archaeopress.com
ISBN 978-1-78969-060-6
ISBN 978-1-78969-061-3 (e-Pdf)
© Authors and Archaeopress 2019
Cover illustration: Threshing oor with many rock-cut cupmarks and 4 shallow basins. Tel Bareqet
(Israel). Photo by David Eitam.
All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means,
electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners.
Printed in England by Oxuniprint, Oxford
This book is available direct from Archaeopress or from our website www.archaeopress.com

To Karen Wright,
for establishing a new approach to
the study of ground stone tools

i
Contents
List of Figures and Tables �����������������������������������������������������������������������������������������������������������������������������������iii
List of Authors ������������������������������������������������������������������������������������������������������������������������������������������������� viii
Introduction ������������������������������������������������������������������������������������������������������������������������������������������������������ 1
David Eitam and Andrea Squitieri
Methodology and Classification
The archaeology of discard and abandonment: presence and absence in the ground stone
assemblage from Early Neolithic Bestansur, Iraqi Kurdistan �������������������������������������������������������������������������� 9
David Mudd
Survey of Rock-Cut Installations at Tel Bareqet (Israel): Food Processor devices in Epipaleolithic,
PPNA and the Early Bronze �����������������������������������������������������������������������������������������������������������������������������27
David Eitam
Ayn Asil and Elephantine (Egypt): remarks on classification and function of
ground stone implements ��������������������������������������������������������������������������������������������������������������������������������52
Clara Jeuthe
Documentation: Non-Archaeological and
Archaeological Sources in Comparison
Mill-songs� The soundscape of collective grinding in the Bronze and Iron Age Near East
and eastern Mediterranean �����������������������������������������������������������������������������������������������������������������������������71
Luca Bombardieri
Rotary Querns and the Presentation of the Past ���������������������������������������������������������������������������������������������81
Jennie Ebeling
Bourgul in Talmudic and Classical Literature, and Today ������������������������������������������������������������������������������93
Rafael Frankel
Wine and oil presses in the Roman to Late Antique Near East and Mediterranean:
Balancing textual and archaeological evidence ����������������������������������������������������������������������������������������������97
Tamara Lewit and Paul Burton
Raw Material and Manufacture
Tool marks on Old Kingdom limestone vessels from Abusir – production of canopic jars
and model vessels ������������������������������������������������������������������������������������������������������������������������������������������113
Lucie Jirásková
Raw material variety and acquisition of the EB III ground stone assemblage of
Tell es-Safi/Gath (Israel) ��������������������������������������������������������������������������������������������������������������������������������121
Jeremy A. Beller, Haskel J. Greenfield, Mostafa Fayek, Itzhaq Shai, and Aren M. Maeir
Function and Uses
The ground stone assemblage from the Early Bronze Age I site Wadi Fidan 4: Gender aspects ������������������153
Yael Abadi-Reiss, Mohammad Najjar and Thomas E. Levy

ii
Cereal processing in stone agri-technological system at late Natufian Huzuq Musa
in the Jordan Valley ���������������������������������������������������������������������������������������������������������������������������������������162
David Eitam
Cuboid-Spheroid Stone Object – an Archaic Scale Weight – Public Weighting-Systems
in Iron Age Israel �������������������������������������������������������������������������������������������������������������������������������������������179
David Eitam
Groundstone Tools from Site 35 – an Early Iron Age Copper Smelting Site in
the Timna Valley (Israel) �������������������������������������������������������������������������������������������������������������������������������189
Aaron Greener and Erez Ben-Yosef
The Iron Age stone tool assemblage of Gird-i Bazar, in the Kurdish Autonomous Region of Iraq���������������210
Andrea Squitieri
Sites and Tools
Macrolithics and the on-going use of stone tools in Qantir-Piramesse and Tell el-Dabʿa-Avaris,
Eastern Delta/Egypt ���������������������������������������������������������������������������������������������������������������������������������������223
Silvia Prell
Millstones, Mortars, and Stone Bowls from Tel Dover and the Southern Levant ����������������������������������������234
Refael Frankel
Stone Tools of the Iron Age Ein Gev and their Implication� The Japanese Excavation Project �������������������278
David Eitam
Selenite (gypsum) from the North Sinai collection: likely function and technology of production ������������299
Joan S. Schneider, David Valentine, Avraham Gabay, and Eliezer D. Oren
The stone tools and vessels from Tel Miqne-Ekron: a report on the Bronze and Iron Ages ������������������������305
Ianir Milevski
El-Khirba: Food processing and other ground stone tools from a Roman, Abbasid and
Mamluk period site near Nes Ziyyona, Israel ������������������������������������������������������������������������������������������������345
Erez Adama, Uzi ‘Ad and Danny Rosenberg