ArticlePDF Available

Egyptian-Canaanite Interaction at Nahal Tillah, Israel (ca. 4500-3000 B. C. E.): An Interim Report on the 1994-1995 Excavations


Abstract and Figures

The nature of the Dynasty 0-Dynasty I Egyptian presence in southern Israel has been a source of debate since the excavations of Tel Erani on the fringe of the northern Negev in the 1950s when numerous Egyptian artifacts were discovered. Since then, a wide range of models have been used to characterize the nature of Egyptian-Canaanite relations. These models include warfare/conquest, commercialization, distance parity, and world systems. The recent Nahal Tillah Regional Archaeology Project has provided a wealth of new empirical data to help clarify the nature of this interaction. Large-scale exposures on the Halif Terrace have revealed a wide range of Egyptian artifacts including epigraphic finds, evidence of Egyptian food consumption patterns, amulets, ceramics, an Egyptian-style mortuary structure, and more. The data from these new excavations require a reexamination of previously held assumptions concerning the dynamics of Egyptian-Canaanite interaction ca. 3500 to 3000 B. C. E.
Content may be subject to copyright.
Egyptian-Canaanite Interaction at Nahal Tillah, Israel (ca. 4500-3000 B. C. E.): An Interim
Report on the 1994-1995 Excavations
Author(s): Thomas E. Levy, David Alon, Yorke Rowan, Edwin C. M. van den Brink, Caroline
Grigson, Augustin Holl, Patricia Smith, Paul Goldberg, Alan J. Witten, Eric Kansa, John
Moreno, Yuval Yekutieli, Naomi Porat, Jonathan Golden, Leslie Dawson and Morag Kersel
Bulletin of the American Schools of Oriental Research
, No. 307 (Aug., 1997), pp. 1-
Published by: The American Schools of Oriental Research
Stable URL:
Accessed: 09-05-2016 12:37 UTC
Linked references are available on JSTOR for this article:
You may need to log in to JSTOR to access the linked references.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact
The American Schools of Oriental Research
is collaborating with JSTOR to digitize, preserve and
extend access to
Bulletin of the American Schools of Oriental Research
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Egyptian-Canaanite Interaction at Nahal
Tillah, Israel (ca. 4500-3000 B.C.E.):
An Interim Report on the
1994-1995 Excavations
Department of Anthropology
University of California, San Diego
La Jolla, CA 92093-0532, U.S.A.
Joe Alon Regional Museum
Kibbutz Lahav 85335, Israel
Department of Anthropology
University of Texas at Austin, TX
Israel Antiquities Authority
Ramat Aviv 61653, Israel
Odontological Museum
Royal College of Surgeons of England
London WC2A 3PN, England
Department of Ethnology and Prehistory
University of Paris X
Nanterre 92001, France
Department of Anatomy and Embryology
The Hebrew University
Jerusalem, Israel
Department of Archaeology
Boston University, Boston, MA 02215
Department of Geology and Geophysics
University of Oklahoma, Norman OK 73019
Department of Anthropology
Harvard University
Cambridge, MA 02138
Department of Anthropology
University of California, San Diego
La Jolla, CA 92093-0532
Ben Gurion University of the Negev
Beersheva, Israel
Israel Geological Survey
Jerusalem, Israel
Department of Anthropology
University of Pennsylvania
Philadelphia, PA 19104
Department of Archaeology
University of Alberta
Edmonton, Alberta, Canada
Nelson Glueck School of Biblical
Hebrew Union College-
Jewish Institute of Religion
Jerusalem 94101, Israel
The nature of the Dynasty O-Dynasty I Egyptian presence in southern Israel has
been a source of debate since the excavations of Tel Erani on the fringe of the north-
ern Negev in the 1950s when numerous Egyptian artifacts were discovered. Since then,
a wide range of models have been used to characterize the nature of Egyptian-Canaan-
ite relations. These models include warfare/conquest, commercialization, distance par-
ity, and world systems. The recent Nahal Tillah Regional Archaeology Project has
provided a wealth of new empirical data to help clarify the nature of this interaction.
Large-scale exposures on the Halif Terrace have revealed a wide range of Egyptian
artifacts including epigraphic finds, evidence of Egyptian food consumption patterns,
amulets, ceramics, an Egyptian-style mortuary structure, and more. The data from these
new excavations require a reexamination of previously held assumptions concerning
the dynamics of Egyptian-Canaanite interaction ca. 3500 to 3000 B.C.E.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
The Nahal Tillah project was designed to obtain
new archaeological data from southern Israel
to examine the role of early Egyptian civili-
zation in the rise of urban communities in the less ad-
vanced southern Levant during the late fourth to third
millennia B.C.E. While the centers of ancient Near
Eastern civilization were always in Egypt and Meso-
potamia, Syro-Palestine and its location on the land-
bridge between Africa and Southwest Asia provide
a unique opportunity for examining some of the pro-
cesses that led to the rise of the earliest secondary
states in this key periphery (Price 1978). Recent ex-
cavations in the Nahal Tillah region of Israel's north-
ern Negev desert, by our team, have revealed a wealth
of new Late Protodynastic/Early Dynastic Egyptian
archaeological (including some epigraphic) data that
will enable scholars to reexamine issues associated
with secondary state formation in this part of the
ancient Near East. The discovery in 1994 and sub-
sequent excavations in 1995 of the first Dynasty 1
Egyptian-style tomb in Israel opens up a unique
opportunity for analyzing the nature of early Egyp-
tian contact with the southern Levant. This interim
study presents a synthetic picture of the results of the
1994-1995 excavation seasons in light of a series of
core-periphery models being used to test our results.
The southern Levant never witnessed "pristine"
state formation of the type and scale usually associ-
ated with centers of state formation such as central
Mexico (Adams 1966; Drennan, Fitzgibbons, and
Dehn 1990), the Indus Valley (Kenoyer 1991), south-
ern Mesopotamia (Adams 1981; Algaze 1993; Weiss
et al. 1993; Zagarell 1986), and Egypt (Hoffman
1979; Hassan 1988; Wenke 1991). The degree to
which Egyptian state formation was an endogenous
process is controversial and scholars such as Bard
(1994) and Kantor (1992) highlight the evidence for
possible Mesopotamian influence. However, for our
purposes, the Nahal Tillah project focuses on the
social evolutionary impact of core-periphery rela-
tionships that involved an early state colossus (Egypt;
cf. Hassan 1988; Wenke 1991) and one of its periph-
eries (the southern Levant). Secondary state forma-
tion has been discussed for Syro-Palestine by scholars
such as Esse (1989), Falconer (1994), and Joffe
The issue of secondary state formation can most
profitably be examined in the context of core-
periphery interaction. From the 1960s through the
early 1980s, archaeologists concentrated their re-
search on social change by studying the internal
factors that promoted social reorganization. Some of
these endogenous factors include changes in local
exchange networks, technology, population structure,
and subsistence economies. For the most part, the
analysis of these issues was spurred on by an in-
terest in the identification of social "types," espe-
cially chiefdoms and states, which evolved out of an
assumed unilineal "social evolutionary stepladder."
While studies that focus on the adaptive success of
these generative social formations have been ex-
tremely useful for identifying local processes of
change, they have tended to pigeonhole social for-
mations into one social evolutionary category or an-
other without explaining how change occurs (Yoffee
1995). In the search for broad evolutionary models,
these studies have failed to identify the rich diver-
sity of social formations that make up the tapestry
of ancient societies in world history. In terms of
world archaeology, this problem is thrown into relief
when social transformations occur along the inter-
face between the prehistoric and historic periods. This
is when asymmetric interactions between different
socioeconomic organizations become markedly clear
in the archaeological record.
An essential framework for examining asymmet-
ric social interaction is the study of center-periphery
relations. Center-periphery studies are rooted in geo-
graphical studies of human spatial organization as
early as von Thunen (ca. 1826; von Thunen 1966)
and are commonly represented in the diffusionist and
hyperdiffusionist studies of the early part of this cen-
tury when all culture change throughout the world
was thought to come from Egypt (Smith 1923) or the
other centers of ancient Near Eastern civilization
(Childe 1934). In the 1980s, archaeologists reconsti-
tuted center-periphery studies by concentrating on
changes in power relationships between social for-
mations of distinct unequal levels of organization
(Cherry 1987; Renfrew and Cherry 1986; Champion
1989). The central issue in secondary state for-
mation is the dynamics of core-periphery relations
and how core civilizations influence culture change
in their less developed neighbors. This issue has
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Taur Ikhbeineh
Gerar * Halif Terrace ISRAEL
* * *Abu Hof
, z Gil'at
* En Besor
Ommt "00 O ARAD
100 Abu Matar ,
Ze'elim SHIQM 200 4E.SH..Safadi
Osnat 400
10 km 665
Fig. 1. Regional map of Northern Negev and the Nahal Tillah Project.
been recently highlighted through the application
and debate surrounding E. Wallerstein's (1974) world
systems model, which examines the economic asym-
metries of these ties and is discussed below (cf. Al-
gaze 1993; Chase-Dunn and Hall 1991; Frank 1993;
Kohl 1987; Stein 1993). This has led to alternative
models, based on the notion of distance-parity which
assumes that the "tyranny of distance" works toward
symmetrical relations of power between center and
periphery (Bairoch 1988: 11; Stein 1993). Interest
in testing center-periphery models with regard to
ancient Egypt and Canaan provided the primary cat-
alyst for initiating the Nahal Tillah Regional Ar-
chaeology Project in southern Israel.
The Study Area: Nahal Tillah
Nahal (Hebrew; Arabic, wadi) Tillah is a small
secondary seasonal drainage that debouches into the
larger Nahal Grar and is located near the interface
between the Irano-Turanian semiarid and Mediterra-
nean environmental zones of Israel's northern Negev
and southern Shephelah (fig. 1). This project focuses
on the excavation and geophysical survey of cave
sites adjacent to Abu Hof, a large Chalcolithic set-
tlement at the beginning of the Nahal Tillah drain-
age, and excavations at the Silo site on the nearby
eastern slopes of the Tel Halif Terrace, which covers
a total area of ca. 16 ha (fig. 2). This is some 6 ha
larger than the average size of Negev Chalcolithic
settlement centers (Levy 1995). The area is charac-
terized by Eocene chalk hills dissected by small sec-
ondary drainages with many small valleys in-filled
with loessial sediments. Rainfall averages ca. 400
mm annually. The Halif Terrace rises to ca. 490 m
above sea level and marks the watershed between
the Nahal Grar in the west and the secondary drain-
ages that flow into the BiqaDat Yaval (Yaval Valley)
in the east. The area dominates an important an-
cient transportation/trade route east-west from the
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
?:.:: .;::i::...:llii::::: :i:?: ??? ' ?~.::: .? ?? ?
:?: :::
: ::??I::.:::::: .::;??
'' :::'
i i
.?: :.: ; ? ? ? ? .:'. ' ?
::li:?~:~~i~~diiiii:idii' ii::l;;;i:ldi ?'? ? ? ? ? ?? ?? ::: ?iii: ' " .".?..?:..:'.??? ' ?
:~iil?::'::'ii;: . ?'?"':ll,[:nu!~'~??:TT::' ;;tiii::?i!i~iiiii6~l~~iiiii~Llii;8tii~ :I:[:1;; ::?~:i:i:
'iiiiirii''iiiiiiii'i :~6~~~ .I. '.i::' '"::' ' :ii:...:::::
?'?'?'~-??.i .
r,~"LaE~~ i: .: :':. .... .?;?;::.j:~~~ ?.?::~:: ~':l.:ili'ijiiiiii:ilili?:: :
:?.:r? .
'i :
r?rF" ~Z
* 'F~~:r*? I:
t I ;5 ?:;L:
?*~" ryr7,Z .~ ;?~~a]bl~, ?JI ~gFSL. ?*??
'I rP?''*"
-rr? - - Is s -E I %~
.;;::?::i:::1: :!:iPL:??:i::1::::: :??I:??' ..:':~?I:
i..... ??.i???::ll??ji?L---?*-r'ii~Lli~Liliii:" I::"tlllT~?~?l~"~lslII~IC~: i???::?:'':.::??-?r:.:PYL?i:P??:~I~*I~U '::::;I(1I::.:... .::?I???:m:11;::::::Ir:???:rrr~:::l-rrru
n: I?
I ,-' SasL--.l-sw~-
?????? ? ?
---~E~n~e~ ~lsntaPg
''::' :i
I ..~.
FIg. 2. Aerial views of the Halif Terrace site: a. in relation to the Yaval Valley, southern Israel; b. detail (photos by
Albatross, Inc.).
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
/\ ~
0 50
Fig. 3. Topographic map of Halif Terrace and excavation areas.
Mediterranean coast, and north-south from the north-
ern Negev through the southern Shephelah and north-
ward through the Judaean mountain system.
Previous Excavations in the Nahal Tillah
Region and Recent Tomb Discovery
The Nahal Tillah project grows out of earlier pi-
oneering research in this environmental contact zone
carried out under the direction of J. D. Seger of
Mississippi State University (Seger et al. 1990) and
D. Alon of the Israel Antiquities Authority on the
Halif Terrace (Alon and Yekutieli 1995). It is located
in the eastern side of Tel Halif (Arabic, Tell Khirbet
Khuweilifeh) near Nahal Tillah on the grounds of
Kibbutz Lahav (fig. 3; Alon 1974; 1977a; 1977b;
Seger 1987a; 1987b; 1987c; 1988; 1989; 1990; 1991;
Seger et al. 1990). Seger's precise work (Seger et al.
1990) provides an essential stratigraphic cornerstone
that demonstrates the rich evidence for the elusive
Chalcolithic through EB I sequence in southern Is-
rael. In an in-depth ceramic analysis, Dessel (1991)
made the first systematic attempt to understand the
changing nature of Egyptian/south Levantine inter-
action in southern Israel using pottery from Seger's
excavations on the Halif Terrace (Site 101 and 301).
Recently, Alon and Yekutieli (1995) made a similar
study using data from Alon's eight probes on the
Halif Terrace carried out in the early 1970s. As a
total of ten probes have been made on the Halif
Terrace, researchers have given separate names such
as Site 101, Site 301 (Seger et al. 1990), the "Villa
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Site" (Alon and Yekutieli 1995), etc., to the different
excavation operations at the site. Together, the early
excavations cover an area of ca. 1703 m2. The large
quantity of Protodynastic material found in these
excavations led to the formulation of a number of
testable models to help clarify the nature of Egyp-
tian-Canaanite interaction.
Modeling Egyptian Interaction
in Southern Palestine
The discovery of Protodynastic and Dynasty 1
Egyptian material culture in southern Palestine sug-
gests four primary models, not necessarily mutually
exclusive, which need to be tested to explain Early
Egyptian-South Levantine interaction and its impact
on local social change. These models include com-
mercialism, military conquest colony, colonial/world
system, and distance-parity. The most efficient way
to test these models is by taking a diachronic per-
spective of center-periphery interaction to monitor
changing relations between the Egyptian core civi-
lization and the less developed Levantine periphery.
As will be shown below, the Nahal Tillah region pro-
vides a unique opportunity for testing these models.
Commercialization Model. Stager (1992) has
presented the best documented arguments for a com-
mercialization model. He views Egyptian/Canaanite
interaction from the Chalcolithic through the EB II as
a series of increasingly complex exchange and trade
relations. During the Chalcolithic period (= Amratian
and Naqada I in Egypt; Kaiser and Dreyer 1982), the
recently domesticated donkey enabled sporadic trade
to be practiced along the overland north Sinai route.
In the following early EB I (Naqada IIb), small
caravans transported ores, bitumen, tree resins, and
other products from Palestine to Egypt. The domes-
tication of the grape at this time, and the already
domestic olive, were introduced into Egypt from Pal-
estine during this period. These products, along with
small quantities of wood, were traded along the over-
land route. By the late EB I (ca. Naqada IIIa2-IIIb-
cl; table 1), with the political unification of Egypt
during Dynasty 0, the commercial structure was
solidified through the establishment of trading posts
such as CEn Besor (Gophna 1995a). In this model,
Tell Erani (Stratum V) is viewed as a true mercantile
colony. The Egyptian unification at home and the
commercial system abroad are seen as providing the
primary impetus toward city-state formation in Pal-
estine in the following EB II (Gophna 1995a; Esse
Military Conquest Model. Developed by Y. Ya-
din (1955) through his analysis of the famous Nar-
mer Palette, the military colony model suggests that
southern Palestine was conquered and occupied by
Egyptian troops for a short period under King Nar-
mer. The early excavations at Erani were used by
Yeivin (1963; 1975) to support this model. A number
of scholars have categorically rejected an Egyptian
conquest and proposed a process based on "peaceful
Egyptian settlement" (Brandl 1989; Rosen 1988;
Porat 1992).
World System Colonial Model. Although only
recently considered for Early Bronze Age Palestine
(Algaze 1993; Dessel 1991; Joffe 1993; Oren and
Yekutieli 1992), this model has great potential for
explaining the impact of core-periphery interaction
in the study area. One of the underlying assump-
tions of this application has been the peaceful nature
of Egyptian activity in southern Palestine (Brandl
1992). The world system model was originally de-
veloped by Wallerstein to explain the processes of
interaction and unequal exchange which gave birth
to the modern European world capitalist system.
According to Wallerstein, world systems reflect the
growth of modern European states, which can only
be understood by the development of geographi-
cally extensive interregional trading systems made
up of different competing political systems. Among
other issues, Wallerstein's model (1974) has been
criticized for its chronological focus on the modern
world, its failure to recognize early external eco-
nomic ties that existed between non-Western po-
litical empires, and the rejection of the importance
of long-distance exchange in luxury goods in ce-
menting the social fabric of these early empires
(Schneider 1977; Kohl 1987). However, by adapting
these criticisms into Wallerstein's original model, a
growing number of archaeologists have suggested
that world systems also existed in the ancient world
(e.g., Algaze 1993; Edens 1992; Kohl 1987). Thus,
ancient world systems are characterized by asym-
metric power between the core and periphery based
on indirect modes of economic exploitation; core
dominance over the periphery through control of ex-
change; the development of long-distance exchange
relations that structure the economy of societies in
the periphery; and the technological and organiza-
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
TABLE 1. Correlation of Southern Canaan and Egyptian Stratigraphy.
Lower Egypt Upper Egypt Southern Canaan Period
Umm el-Qaab, Abydos:
Dynasty 1 Buto, Tomb Z King Djet Tel Arad, Stratum III EB II
(Naqada IIIc2) Stratum V Tomb O King Djer el Maghar
Tomb B10/15/19 King Aha CEn Besor, Stratum II
Tel Halif Terrace,
Dynasty 0 Buto, Tomb B17/18 King Nar(mer) Silo site, Stratum IIa-b
(Naqada IIIbl-IIIcl) Stratum IV Tomb B7/9 King Ka Tel Erani, Stratum V EB 1B (late)
Tomb B 1/2 King Iry-Hor Tel Arad, Stratum IV
CEn Besor, Stratum III
Dynasty 00 Buto, Tombs U/f; U/g; U/h Tel Erani, Stratum C EB 1B
(Naqada IIIal-2) Stratum III Tomb U/i Azor, cave tombs (middle)
Tombs U/k; U/j 'Scorpion'
Transitional Buto, Hierakonpolis: Tour Ikhbeineh
(Naqada IIc-d2) Stratum II Tombs 100 ("painted tomb"), 101 Tel Erani, Stratum D EB 1B (early)
Locality 29A Lachish (NW)
Silo site, Stratum IIIa
Site H
Late Chalcolithic Buto, Matmar: Tour Ikhbeineh EB LA
(Naqada IIb) Stratum I Tomb 3131 Tel Halif Terrace,
Maadi Silo site, Stratum IIIb
Chalcolithic Maadi Beersheva sites Chalcolithic
Tel Halif Terrace,
Silo site, Stratum IV
tional skills to extend their power over long dis-
tances. Applications of the world system model in
archaeology have been criticized (Schortman and
Urban 1994) primarily on the assumption that dis-
tance from the core has the effect of "leveling" out
power relations with the periphery, what Stein (n.d.)
has referred to as a "Distance-Parity" model (cf.
Champion 1989; Stein 1993).
Distance-Parity Model. Recently utilized by
Stein (1993) for modeling Mesopotamian/Anatolian
interaction, this model characterizes center-periphery
relations as primarily symmetric. Distance-parity
suggests that a core civilization's ability to exercise
hegemonic power deteriorates with distance, leading
to symmetry in economic and political relations with
societies on the periphery. This model is especially
interesting when considering the southern Levant as
one geographical locale. For example, while W. Rast
and T. Schaub (Schaub 1993) have investigated some
of the largest EB I sites in the southern Dead Sea
Plain region, such as Bab edh-Dhrac, Numeira, Safi,
Feifeh, and Khanazir, a marked paucity of Egyptian
material is present in the archaeological record. This
is remarkable considering the close proximity of this
region to the northern Negev. Thus, while the dis-
tance-parity model does not work for the northern
Negev, it may be appropriate for Jordan. This prob-
lem should certainly be investigated in greater detail
in the future.
Excavations on the Halif Terrace
To test these models with new data, a program of
extensive open-air excavation was initiated on the
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
o 5M
1 . 1 . . .......
1 1 1 171
,* ,. @s l
_2 I..0|1 4 0,. IIII
L - - - -~" [l
i ji
0 1 0. is~ `f5"; d ?
1 - I 1.82 1.96 1
1RO1 3 1.10
I,. 180 W 4 L9.
"(1 110 L8
L" II i.. II
ff ~ ~ ~ ~ ~ ,., . .|
Ir L. Io, ,' r .
_J _ _ _ .. /r , 0
L ~L -~ ~ ~ 87 103MMK
19 ! 20 I 1 2 3
Fig. 4. Plan of superimposed stratigraphy in Area A, Halif Terrace, 1994-1995 seasons.
Halif Terrace by the Nahal Tillah Project team. In
1994 and 1995, the main excavation efforts of the
expedition took place at the Silo site on the Halif
Terrace (fig. 3). Situated on the western flank of the
Yaval Valley, which winds its way north past Tell
Beit Mirsim, along the eastern Shephelah and up
toward Jerusalem, the Halif Terrace commands an
important position along the main artery that links
the northern Negev with the hill country. By the
end of the 1995 excavation season, our total expo-
sure on the terrace exceeded 1300 mn2. We antici-
pated that by the end of the 1996 excavation season,
our exposure would exceed the total area excavated
by previous archaeologists who have worked on the
Halif Terrace. Our broad horizontal excavation is
divided into four fields or areas (A, B, C, and D)
and is summarized below.
Area A. Area A (figs. 4-5) has provided the
best stratigraphic sequence at the site spanning both
the Chalcolithic and EB I periods. In 1995, seven
strata were delineated at the site (Strata I, IIa, IIb,
IIc, IIIa, IIIb, and IV). Stratum I at the site is the
most recent level; it is mixed and contains evidence
of small MB IIB tombs (Area C) and other late ma-
terial. Stratum IIa/IIc represents the late EB IB and
has yielded the largest number of Egyptian artifacts
of any strata on the terrace. During the 1995 season,
for the first time, we isolated an early EB IB phase
in Stratum IIIa with very little evidence of Egyptian
material. In general, the early EB IB represents a
pre-"Egyptian" phase in southwestern Canaan and
has been recognized by Y. Yekutieli at Erani, CEn
Besor, and now on the Halif Terrace (Y. Yekutieli,
personal communication). Stratum IIIb represents the
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
~*r?? :2::.*
i I:::
?*:.. .?:. :::
....: *~.
??:.;:? ..* i:.
8:~????; ?:?? t"'b:~i;:'i~?B~lt~~~I' ~C~B~ga~.:::l:*:';
::: ?
IL:: :8:
:,h.Z~1'"~1*?i"~ttk~I?~BJJI&YBllllll;~f~ ' b deB~-?1- ~B~(l~e~Iyy*l~d~(~j~t ::iS
Fig. 5. Area A excavations, Halif Terrace. Stratum IIIb is represented by the square-shaped structure.
post-Chalcolithic collapse in the northern Negev or
EB IA. The EB IA is also referred to as early EB I.
Stratum IV contains Chalcolithic remains.
The identification of the early EB IB phase is
very important because it will help clarify the local
evolution in settlement and its later response to the
Dynasty 0 Egyptian influx into southern Canaan.
The early EB IB is characterized by distinctive
rectilinear architecture, such as Room 3, seen here,
which is 7m x 5m large and is found in Stratum IIIa
(figs. 4-5). Some of the characteristic ceramic forms
of this phase are storage jars with diagonal incisions
on their necks, special double loop handles, and red
slip. In general, the assemblage is characterized by
"dolomitic" tempers.
Area A has also produced the richest assemblage
of epigraphic data from the Terrace. In 1995 two
more clay bulla were discovered. On the back of one
example, the impression of cloth can be seen, possi-
bly indicating its original function of sealing a sack
rather than a jar like the previous bulla found in
1994. Whereas the bulla found in 1994 was rooted in
Egyptian parallels, the new ones are more similar to
those found at CEn Besor Stratum III. All of the seal
impressions from our recent excavations are dis-
cussed in detail below.
In 1994, a well-preserved serekh sign with the
name Narmer was found in Area A (fig. 6; Levy et al.
1995). The following season we found a small, but
distinctive fragment of a serekh that may contain the
name of Ka or Narmer in the so-called name com-
partment. The serekh signs are also discussed in de-
tail below.
Area D. Area D consists of ca. 8 squares, each
5 x 5 m, and was established in the northernmost
area of our excavation. It borders J. Seger's excava-
tion at Site 301 on the terrace (figs. 7-8). In archi-
tectural terms, all the remains from this area date to
the late EB IB. A series of small rooms found here
are associated with the large-scale production of
bread. Here in the late EB IB stratum (IIb), a huge
number of Egyptian-style bread molds was found
associated with a large tabun attached to a small
room (Room 4, Loci 906, 840; fig. 8); they totaled
over 500 kilograms. The significance of bread-mold
production during the late EB IB on the Halif Ter-
race is discussed below.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Irl lliliiC::::: ?:?? ,
?::::.? ?' i
?: ?:...
???" '
j . 1??""
?;, :j:?i?
::?I? :r:;e:a::::::::. i.';l..:. :'.
"i ?:::' ::: i
Fig. 6. The Nar(mer) serekh, found in Area A, Halif Terrace.
Area C. Area C is located in the center of our
excavation area and has produced the most extensive
architectural evidence at the site including some re-
mains that may be public (figs. 9-10). Three sepa-
rate building phases were identified here spanning
Strata IIa/IIc and all date to the late EB IB. Stra-
tum I, however, has yielded two intrusive cist-shaped
graves dating to the MB IIB. The cists are ca. 1.2 m
long, 0.80 m wide, and 0.50 m deep. Each was asso-
ciated with MB IIB pottery vessels including small
button-base juglets (fig. 11). The discovery of these
small graves is of great interest considering the MB II
gap observed at the nearby Tel Halif mound excava-
tions (Seger et al. 1990: 3).
Most of the features in Area C date to the late
EB I and are connected to a large terrace wall (Wall
29; figs. 9-10), which is over 30 m long and was first
described by Seger in his excavations at Site 301
(Seger et al. 1990). This wall served as a kind of
S2 L.857
N . . . .. . 7. 8"
L 8610 9 ,-O G !L . 86 9
L. 8L_.. 5 7
L 8
t'.l I, t
L 1I. 82I
I 28 1- S l. 857
0 L. 82805
J? ~ 6 I. 9 ,. 901
. -cL857
1,.s4 __Scale 5
17 18 19 20
Fig. 7. Plan of Area D excavations, Halif Terrace, 1994-1995 seasons.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
?t "' ?:
e?~i r:ii~af"d
j?, -, t ~4~] a u i)?")~!
S:?Y a i : i'
? ? ??,(~P~LI~1:.'? ':z~
I ..?i?? :f.? Ir k' P 1 .ruc?
i n:;-?6 "i ~?: .~3? L:?ilij'
Y * ~? :.::
* ?I,
1: j. ???-5'f:I ?~
r.ti. *?:
:'d~+? ~S~EFlil~F~il
*? I?
?' ~ iiFI~~~1 ~h~il~ii' ??-?
iii: ??:?:
B: YikAir..i :r: i?r? ?
?.1.. ;C??,.: :,i .i;iii
??1. .
..i* ?::i' ::
,r ;??.???
ic':;it ~I
~: r.
i P; ??: :??1?. . i
I ?' : if:: ~tf
i' ??i .I
:::3 'f
:i:?c:':. '??
.~ iiti:.t:t~~i~
r t ii"
* -;
9~a ?,.
c~iL~4~ ~~~ I.-it~T E ra~b'
!if 9- 1, I
?, c:, ?
I :'lk?~~ c '~i;~?:
i ?$ ?
* r...? .ji.
i?a t:
i %~. n?~ e~:
r. ~i;?
.k? ~E
!~llll&%~4d.r::abfd~'~I~,~~~I?~ar~ u? :?~r..rrr?~? .? -?~
:.::? ?:?
"' 9
Ihi , ~la~?~~*t'~W~. Jl~..'liC:i:? .:: *I .Z
4 .r.
.Cr ii
* ?*
??i? :d:
'"' ;1~? i k ~' ..
:i. rS'?
I '* .* ,f; ~,?I:i"la;?.?;s: t
:u~ I.. .Li J
i i ;r ?L
r, bC
I .
?ri ii:
~iP ~:r~li~Fde~6
~? yl?
Fig. 8. Tabun found in Area D, Halif Terrace: a. area view; b. detail.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
0 31 606 L. 621 04 L 622
J W 636 W~53 w333
w29d I
L 636_ _
L 629 52 L. 586
L 630
L 615L61
H eScale s w2c6, 0
L0 588
L. 625. 629 2 L6
HI itP cnsE ~ ~ CL
S L. 565
LL. 60
TOMB 1 19
SL.543- L, 555 8
-t c2 ~OL
0--I- --- -0 0~
',,,,1.)L. 597
L. 601
E L. 611r
17 18 19 20
Fig. 9. Plan of Area C excavations, Halif Terrace, 1994-1995 seasons.
"foundation trunk" in which architectural branches
were added to it in both Areas C and D, and Site 301.
A large, trapezoidal building was excavated which
contains a beautiful stone-lined silo or basin (fig. 10).
This basin, more than 2 m in diameter and 1 m deep,
was sunk into the floor of this room. One of the large
flat slabs on the floor of the basin measures over
65 x 95 cm. Although this Halif Terrace example is
larger, it resembles the stone-lined basin found in
the "twin-temple" at Arad from the later EB II. From
a construction point of view, the most remarkable
observation is that a flurry of major reconstruction
activities took place in Area C, all during the late EB
IB when the trapezoidal building was constructed.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
:???? :
:;::? .:.:?:????::::.::::::.r!r:?:..
.::.?????. ...???::::.::r?:::::???r::
;;;;- : :
::i ::. :?~? ??:? ?'?':????"':::::E:::::?ii":::::?l?!:::
":r:? :i:::::.??:?:::i*L::::?i:??::i::i"???:: :::?:::':Pi::?
;;; :??;;L-??? ::..::'.:::;;;i.:::::i:;;:::;:
.JC ?,a i~:?
'':"?? :r~
?i~? d' :;::??
:. .:r..
...::;:r;?:... ?:???::?-::?rr?
::I::;:. "[bCti "'.~?~?:?: 'l""":
''''" .z~ir' .; ,.....?.,?..
aiP E" .:~':r .. .I?:::??
:W:?:I~'"':I~: "'* ill::''~i:: .......I::? :w. ,li;:'~~~,.
.;???? ?1:
*: *~::: ?????.?1:.::?
.i???...~ sr:?.?~ s -~lsauYl~~~n~kl*~asr~:a~?r;~!!l?r~L~?iy ;ii!;
?* "': ?- ?llr????
Fig. 10. Overview of Area C, Halif Terrace.
:.? ::i::
.?L!:::::. ::::
i?i i:
15:;;;;;; ; "; .;;;-
::lli::::::::. ~,:??:~
:.??r? :?i:+:
::: ::':::r;
:Iljii:iiiiii"i:? ?:;:::::
111 :
?:::?" ' :?i9rl~?illl~lilllsll*aurr~sP????~??lI:::::iii::::*l~*:::::::
Fig. 11. MB liB juglet found with burial (Locus 576), Area C, Halif Terrace.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
.d~r?:*~~~ r????-??? ~:%?~::.? ?:%,i:~r:v;a;~::18i!..r:i:i:.~fl'''...? :.?'"'::r? '?::'
?,,?~: ..i:?.
.?;? s~*.;:??.
9 ~j" ?? ? gi..
.?~ ?;:"
... *r,?
,??? i??:?
~ ~;??? 'c' ???3 ''
1.; . .:ii
?:??? ''i ':*
??;. i?J ::
U?: . ' ' ?:""~ifi':''iilf~t~
??i i:i
I? :i
?:i ;~B(jjis? ?I::::iili? i:? ~
*'it?::r::i.?(giiBI~B~llr~~ ~?i~s~Ei~Fl~fF~~~~:"~:"~?~itl":~l':
i iB;1
.J?' ' eestlP~;E;ICiic~~j~i~; 'iii.:"
:t' :4?
Fig. 12a. Tomb structure, late EB IB: overview.
This coincides with the building activities associated
with a large Egyptian-style tomb structure built some
10 m to the south (below).
Egyptian-Style Tomb. In 1994, a large stone-
lined dromos, over 10 m long, was discovered dat-
ing to the late EB IB period in Area C (fig. 12a,b).
The walls of the dromos were preserved to a height
of more than 2.5 m at their maximum. The dromos
was cut through earlier occupation strata, including
a storage silo dating to EB IA. A well-preserved
plastered floor was found with numerous ceramic
sherds dating to the late EB IB on the outside; it was
contemporary with the dromos walls. Between the
dromos walls is a passage ca. 1.2 m wide associated
with a steep, hard-packed earthen surface that func-
tioned as a "stairway" leading to the entrance of the
tomb. The tomb entrance consists of a stone-lined
passage, or ante-dromos, attached to the western end
of the dromos situated some 2 m below the site sur-
face. It was built with capstone blocks as large as
110 x 55 x 30 cm. The ante-dromos is approximately
2.8 m long, 1.6 m wide, and 1.2 m high (Levy et al.
in press). The 1996 excavations have revealed a
series of steps cut into the bedrock below the ante-
dromos. Thus, the ceiling of the ante-dromos is ap-
proximately 3 m above the bottom step which leads
into the burial chamber/cave (figs. 5, 12a,b).
'Yi: sire Bj: .?i~i?~ ~ "jg~Wf:%l? ~?J
f s
:?' . ?~
~F~??. 'P 8.?'?
* ~. f? B~?* ' "'. ??r~ ?:?21.
:;sf ,,
*r *ilb
.?? ~ ?U ~* ::.?
~e~. ?,I .?.~ .
?:~7~~~:~'r~ ? ,Izj~e~idi~s
- PL"4?L~p
1: ??;;
;::?:: ::i:i;:?l"i::::: :IsiiP::?l:? *"'
c~ ???"''
*:i?:!?:' .:n
:?.. ? I ? ;uu:.i:rru?l~r~r*lrrrmrs~ls~il~ ?rr -L?IL~ ''1U? -~-r-- aaa~--? .' .I-1~.?
;?li(?~i%.d~Si~i;i~8?re~?Bii+~g~EL~sVLBI r ~P~I~Y~i .. .?rv~?LPI*r
:??~arr.:: I??:-?-:--??(?
r ,
:.hl ?*4. :*
"' i*e~F?
'"'' *
~u*. ~bl
.. ???????
?::::;??L:~ ???1P1--? :?-~lj~p~~
it'?'? r
Fig. 12b. Tomb structure, late EB IB: detail of roof to ante-dromos.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
?:ii .
: ??; ??;::?'
.?I?:. ::i:
i. .?;ii
"':? .i?
r? i: I?iP
??:::r?; :iY:?'
?? :'
r ~:.:
: ::i
??:.i: ?.:
?::? ?'::: t P
:' ::lii3~~'i.:.lfli~~~il8~%~~ii;jiiiiiTli Eill~~L'i~i~a~s~L~:..::i?:
Fig. 13. Closing wall in back of the cave.
To safely excavate inside the burial cave attached
to the ante-dromos, in 1995 we removed all the cap-
stones to secure the area. Excavation inside the cave
was extremely difficult, and excavators were obliged
to work around a wood scaffold built by the Israel
Antiquities Authority to prevent a collapse. A section
was excavated through the cave deposits, ca. 1 m in
width through the length of the entire cave. Excava-
tions inside the tomb/cave were extremely difficult,
and approximately 120 m3 of sediment was removed
from the cave. After accumulating approximately
250 buckets of sediment, a "chain gang" removed
the buckets from the cave and brought it top-side,
where every bucket was sieved through a /4 inch
A beautifully built closing wall, some 5 m long
and preserved to a height of 2.2 m was found ap-
proximately 8 m from the west end of the ante-
dromos inside the burial cave (fig. 13). It is now clear
that there were two phases of use in the cave, the
earlier during EB IA and the later during the late EB
IB when the cave was reconfigured through quarry-
ing and used as part of the burial monument. The
cave itself has a very smooth, aesthetically pleasing
dome shape. During the late EB IB the massive wall
(fig. 13) was built to seal the cave in half, creating a
burial chamber. We believe the wall was built up to
the ceiling of the cave over 3.5 m high. When the
wall collapsed to its present height (most likely a
result of grave robbing shortly after EB I), the cave
was filled with sediment.
Thus far, the only human skeleton we have found
dates to the late EB IB when the cave was part of the
burial monument (fig. 14). It is an adult female ap-
proximately 25 years of age; she was buried in a
flexed position, on her left side, facing southeast to-
ward the opening of the tomb. No grave goods were
found with this individual. A full discussion of the
physical anthropology is given below.
Approximately 50 cm below the EB IB floor and
burial, we found evidence of the earliest occupation
of the cave dating to EB IA. Crude hemispherical
bowls with straw temper, and other EB IA pottery,
were found here along with a beautifully crafted
standing stone that was found erect in the cave. It
seems the cave was much larger during EB IA before
most of the ceiling collapsed. When the late EB IB
people decided to build the burial monument, they
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
: ?;: ? . .i: ~::: ??
i :I:~P~i~g~:liii~~~~iiii;~aj~4l~i~~iBE~ ;iiBi!BI~~~?B~:iiiiiiiii~?iiiiiPii!i::i i:f.
?II?? ..:.?.:
'" ??:r ?
.?:? %?"?:
;~;? ?; ?;l;ilts!~i?-;
.i. ?i:
i??.. r-:
''?:''? ? ::
Fig. 14. Female skeleton found in situ in the tomb.
simply divided the cave in half by building the mas-
sive wall we saw earlier.
We did not finish excavation of the burial cave
associated with the dromos during the 1995 season.
Although a group of excavators stayed in the field an
extra 3 weeks to complete this work, we simply ran
out of time. Approximately 20% of the cave re-
mained to be excavated, all of which is located near
the entrance to tomb under the wood scaffold area
that was foisted on us. Thus, we believe the final an-
swer concerning the tomb will be solved at the end
of the next excavation season.
The Egyptian artifacts are central to understand-
ing the nature of the Protodynastic presence in south-
ern Israel and consequently are described in some
detail here. A wide range of Egyptian-related finds
have come to light in the excavations at the Nahal
Tillah project. These data provide new information
for tracing the nature of Egyptian-Canaanite inter-
action, especially during the EB IB period. The fol-
lowing discussion highlights the most important finds
from the Halif Terrace (Silo site) excavations. The
majority of the small Egyptian finds were discovered
in Area A. These include clay bullae/cylinder seal
impressions, cylinder seal, incised serekh signs, other
ceramic small finds, and amulet(s).
Cylinder Seal Impressions
During the first two seasons a small number of
clay sealings or fragments thereof was found (fig.
15). Three of these (a-c) bear cylinder seal impres-
sions. Petrographic analysis by Y. Goren shows that
the clay used for shaping these sealings is of local
(e.g., non-Egyptian) loessy clay, and therefore con-
forms with the findings concerning various cylinder
seal impressions found at CEn Besor III. Thus far,
only one specimen (clay sealing reg. no. 649/94, Lo-
cus 19, B. 266, illustrated here as fig. 15a) has been
tested. Seal impressions b and c (below) are similar
to some of the CEn Besor examples (Gophna 1995a),
strongly suggesting a near-contemporaneity of the
two sites; they both yielded seal impressions depict-
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
0 5cm
Fig. 15. Cylinder seal impressions: a. G49/94, Locus 19, B. 266; b. G67/95, Locus 105, B. 1325; c. G67/95, Locus 102,
B. 1308.
ing a standing or striding figure facing to the left, a
not uncommon feature among the cEn Besor II cyl-
inder seal impressions (cf. Schulman 1976: nos. 2, 6,
8-9, 14-15; 1980: nos. 16-18, 36, 43).
The depiction of standing/striding figures is not
very common in the corpus of early cylinder seals or
cylinder seal impressions known from Egypt proper.
Assuming a date for both the cEn Besor III and Tel
Halif Terrace Silo site Stratum II around the reign of
King Narmer, an explanation for this dichotomy may
be seen in the facts that in Egypt proper, the total
number of clay sealings bearing cylinder seal im-
pressions dating from the reign of Narmer is still
rather limited (Kaplony 1963) and, moreover, that
such pieces stem from funerary contexts (mainly
from Cemetery B at Umm el-Qaab, Abydos); this is
in contrast to the finds in Israel, which all derive
from settlement contexts. Alternatively, perhaps one
should assume that the frequent occurrence of the
standing or striding figure reflects a local, i.e., Ca-
naanite theme so far unknown to scholars, and that
perhaps at least some of these impressions-although
undoubtedly of strong Egyptian affinity-reflect a
certain degree of local Canaanean thematic/ideo-
logical influence. The cylinder seal impressions are
as follows:
a. Clay sealing, reg. no. 649/94, Locus 19, B. 266,
found in Area A, in a fill dating to Stratum Hla (late
EB IB). Described in extenso in Levy et al. in press
(fig. 15a; see also Levy et al. 1995).
The cylinder seal impressions are preserved on
an oval-shaped lump of (accidentally-fired) reddish-
brown, local loessial clay. The same cylinder seal has
been rolled over the clay sealing three times, once on
each of its three faces. The inscription preserved
consists of a single, narrow line made up of three
elements: a face (hr), two flagpoles (ntr.wy), and two
identical signs read by Kaplony (personal commu-
nication, 24 June 1995) as mr.wy. It thus reads I.r
ntr.wy mr.wy, which can be translated as "My face is
(the face of) the Two Loving/Beloved Gods." For
the interpretation of its full meaning see Kaplony
and van den Brink (unpublished).
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
b. Clay sealing, reg. no. G-67.95, Locus 105, B. 1325,
6/1/95, found in Area A (Square V20) in an ashy pit
fill dating to Stratum lib (late EB IB).
The cylinder seal impressions are preserved on an
oval-shaped lump of fired, reddish-brown local, loes-
sial clay, ca. 5.5 cm long and 3.0 cm wide (fig. 15b).
A small part of the sealing is broken off and missing,
and hence the seal impressions are not completely
preserved. The same cylinder sealing has been rolled
at least twice over two adjoining faces of the clay
seal. Impressions of cloth and strings can be seen on
the otherwise plain back side; the cloth impressions
extend to the lower part of the impressed faces of the
sealing. The surface of the sealing is "marked" by
several narrow, oblong imprints: these are the neg-
ative imprints of burnt straw particles, incidentally
present in the clay of this particular sealing; they are
not a part of the cylinder seal impressions proper.
The incompletely preserved impressions show a
standing or striding figure (head not preserved), fac-
ing left and with uplifted arms. A schematic sign,
perhaps representing a tree or shrub with two sets
of curved branches protruding from the stem, can be
seen in front of this figure. Based on the few traces
remaining immediately to the left, it would seem that
this particular sign had been depicted/repeated twice.
No interpretation or reading is offered yet.
c. Clay bulla reg. no. G-67.95, Locus 102, B. 1308,
found in Area A (Square V20), in the same context
as clay sealing, shown in fig. 15b.
The cylinder seal impressions are preserved on a
lump of unfired, blackish-gray clay, ca. 3.1 cm long
and 3.0 cm wide (fig. 15c). Like the previous exam-
ple (fig. 15b), this impression is partly broken and
thus incompletely preserved. The same cylinder seal,
consisting of a single band, had been applied twice
on adjoining faces of the clay sealing, each starting
from a slightly different point on the cylinder seal,
something that made it possible to reconstruct the
inscription to its full extent. On its otherwise plain
back side are the impressions of two crossed strings.
The inscription, contained in a single register over
the full height of the cylinder's surface, shows two
pairs of superimposed triangles, pointing downward,
preceded by a standing/striding figure, head missing
and facing left, with right arm outstretched, left arm
bent; behind the back of this figure is a city/fortress
(?) oval (Kahl 1994: 886, sign no. aal7) with per-
haps two pairs of opposite watchtowers (?), below
which is a group of three adjoining triangles point-
ing upward; the latter can perhaps be read as i3s.t
("the mountainous or foreign country") or ("he
who belongs to the mountainous or foreign coun-
try"). The oval and the three triangles are followed
by three identical vertical signs that we have not yet
been able to identify (perhaps knotted strings [?], cf.
Gardiner 1969: sign list, V28). Until it becomes pos-
sible to read and understand each individual element
in the inscription, we can only speculate that it might
contain the following information: "the chief/ruler
(?) of the city/fortress (?) of the foreign country X."
Cylinder Seal
Cylinder seal, reg. no. G-67/95, Locus 92, B. 1226,
found in Area A (Square V20), in the same context
as seal impressions b and c, above. Dates to Stratum
lib (late EB IB).
The cylinder seal, made of fired clay and slightly
broken at one end measures 2.9 cm in maximum pre-
served height and 1.2 cm in diameter; it is not per-
forated. A shallow groove is impressed at one end of
the seal (fig. 16a); it is, however, so shallow that it
cannot be functional and it is questionable whether
the groove was made intentionally.
The seal is engraved with a geometric pattern
composed of 11 vertical lines extending over the full
length of the seal's surface and crossing six smaller,
horizontal lines. A series of three horizontal and much
deeper engraved lines going all around the seal's sur-
face give the definite impression of having been
applied secondarily. A possible explanation for this
could be that it was done to invalidate the seal.
Stylistically, this rather irregular pattern does not
seem to fit into the usual patterns of cylinder seals
with lattice pattern found at various sites of the
EB I in Israel, Jordan, and beyond (cf. van den Brink
1995: 203). Although certainly anachronistic, it some-
how reflects a much earlier theme, viz. Nagada IIc/d
cylinder seals found in Egypt (e.g., Boehmer 1974:
figs. 13-14). Its closest parallel, however, is found
once more in the corpus of cEn Besor seal impres-
sions, viz. no. 51 interpreted by Schulman (1992:180,
fig. 1, no. 51) as a "wickerwork shrine."
Incised Serekh-signs
a. Pottery fragment with prefiring incised serekh,
showing the name of (Horus) Nar(mer) (Reg. no.
G49-94, Locus 14, B. 259)found in Area A, Stratum
Ib, in a fill above a large circular platform (silo?;
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
.... ;; ::i r: i :o _ "
?r? C:
? " " " .... I
" .. .i ! ...
Fig. 16. Cylinder seal (a) and incised serekh sign (b).
fig. 6); it dates to the late EB IB. The sherd is made
of Egyptian Nile silt and is described in detail in
Levy et al. (1995: 31-32).
b. Pottery fragment with incised (prior to firing)
serekh, probably with the name of (Horus) Nar(mer)
(Reg. no. G67-95, Locus 58, B. 1028, 7/6/95), found
in Area A (Square V3), in a fill context dating from
Stratum IHa (figs. 16b, 17:3a-3b).
Only a small, yet distinctive part of the left side,
middle section of the serekh has been preserved. The
sherd measures ca. 3.2 cm x 2.1 cm in length, has a
thickness of ca. 8-9 mm, and comes from a hand-
made, burnished, well-known Egyptian storage ves-
sel, e.g., the so-called wine jar (van den Brink 1996).
The interior surface of the fragment is light reddish
brown, the exterior surface is light grayish brown.
The break is light reddish brown all through, turning
slightly grayish brown close to the exterior surface.
The fragment shows part of the left (vertical) bor-
der line of the serekh, and part of the (horizontal)
line dividing the serekh into two parts; the lower part
represents what is generally accepted to be a styl-
ized, niched palace-facade (indicated usually by three
to five vertical parallel lines, of which in the exam-
ple here only the top of one is preserved). In the
upper part of the serekh, the so-called name com-
partment (since it usually contains the Horus name
of the king), the end of a horizontal line, perpen-
dicular to, and in effect cut by a vertical one is
preserved. This is all that is left of this king's name.
Although initially thought to represent the name
of (Horus) Ka (we thank C. Kohler for this observa-
tion), the predecessor of (Horus) Narmer (cf. recon-
struction drawing fig. 17:3b), for two reasons we
now believe that this sample is another example of
a Nar(mer) serekh.
First, the angular joint of the left arm (figs. 16b,
17:3a), though certainly attested in First Dynasty
potmarks representing a pair of open arms (van den
Brink 1992a: fig. 7, Group III), is atypical for writ-
ing the name of this particular king. With one pos-
sible exception, the main element in all surviving
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
I 5
! ,? 4.
ii 3.: .
?. a v:?:?
'4 .'C
m~: ~~
1:?1?,?,?;i1 ?t?',
I I ?r .?-a???;?i -:-- .??..
i, *~?'li
?: ~I:??
~9 1 \ ::? ??n: .::~ :"?:??
.?i ?.
?? :-, :
c i.
?; "t rry
?: C-.
..I? ?2:
:x: ii
,;",f: '~? .e
r. ?- ?~ ???
??, :::'?~s:L~
I~?:JhT~?' I'
?:: ?????;:,
?" '
?. ?':
0 5cm
Fig. 17. Miscellaneous small finds: 1. cylinder seal
(G-67/95, Locus 92, B. 1226); 2. pendant (G-67/95,
Locus 104, B. 1302); 3a. serekh reconstruction of
"Nar(mer)"; 3b. serekh reconstruction of "Ka"; 4. Egyp-
tian potter's mark (G67/95, Locus 821, B. 3069);
5. modeled pottery w/giraffe (?) legs (G67/95, Locus
812, B. 3034); 6. modeled pottery w/human (?) legs
(G67/95, Locus 861, B. 3174).
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
examples of pottery incised (Horus) Ka serekhs, con-
sist of a single, more or less fluid curved line (Kaiser
and Dreyer 1982: fig. 14, nos. 23-33). For this reason
Helck (1987), for instance, denies that the sign de-
picts a pair of open arms; according to him, this rep-
resents a kind of cloth with fringes at both ends.
The second reason is that, contrary to most of the
incised serekhs with the name of (Horus) Ka, the
(reconstructed) hands would be pointing away from
the stylized palace facade. In almost all cases, irre-
spective of whether the open arms are upright or
down, the arms point toward the palace facade. For
this very reason Kaplony (1958: 54-57) suggests
reading the pair of arms not k3 but sbin ("he who
embraces [the palace]").
The exceptional status of this fragment regarding
reading here the name of (Horus) Ka should make
one cautious. In view of the other incised serekh
recently found at the Silo site on the Halif Terrace,
it would seem more prudent to opt for a reading of
the name (Horus) Nar(mer). The fragment would fit
very well, for example, the serekh fragment of (Ho-
rus) Nar(mer) found at Tell Ibrahim Awad (van den
Brink 1992b: 52, fig. 8, no. 3; Levy et al. 1995: 31),
constituting the rear part of the very schematically
rendered catfish (read as "[Horus] Nar[mer]").
Egyptian Ceramic Small Finds
Small, red slipped and pebble-burnished sherd with
potmark and punctate decoration (Reg. no. G67-95,
Locus 821, B. 3069, Area D (Square K/18), found in
a fill from Stratum lib (fig. 17.4). It measures ca.
3.4 cm x 2 cm and is 6 mm thick; the interior is
grayish black, exterior: reddish brown and the break
reddish brown all through with no sign of oxidation
Similar fragments with identical surface treatment
and punctured decoration have been found in strati-
graphic contexts in late Protodynastic/Early Dynas-
tic deposits in at least two settlement sites in the
Delta, e.g., Tell el-FaraCin/Buto (with ca. 15 frag-
ments recovered from Strata IV-III (C. Kohler, per-
sonal communication 1996; Kohler in press: pl. 57,
nos. 3-4, 12) and Tell el-Farkha (with at least one
fragment, also incised before firing with a partially
preserved potmark (M. Chlodnicki, personal com-
munication 1996), pointing once more to the close
relations between Egypt and Canaan at this time.
The potmark itself (applied before firing of the
vessel) consists of a combination of two signs, a bird
(probably an ibis) and above it a branch or hand. Al-
though not common, the potmark fits well into the
corpus of Early Dynastic potmarks in Egypt (van
den Brink 1992a: fig. 16, group LV, nos. 1-4); bas-
ing ourselves on the only close parallel potmark (van
den Brink 1992a: fig. 16, group LV, no. 2), we opt for
interpreting this sign as a branch rather than a hand.
As a potmark incised before the firing of the ves-
sel, the ibis (bird) can appear either alone (attested
twice) or in combination with one or two other signs
(attested 12 times and 1 time respectively). To what
these, and other Early Dynastic potmarks refer is
still much debated.
Egyptian Modeled Pottery
a. Pottery fragment showing modeled parts of the
two legs of a giraffe (fig. 17.5). Reg. no. G67-95,
Locus 812, B. 3034, Area D (Square L/19), found in
a rocky fill context, dating from Stratum lib (late
The fragment is ca. 6.9 cm x 4.4 cm x 2.1 cm
thick. Both the exterior and the interior are light
reddish brown, with a gray core associated with light
reddish brown oxidation zones. The surface has been
smoothed with no slip application. The legs of the
giraffe (?), facing left, have been impressed with
some kind of stylus, to render the flecked pattern of
the animal's pelt.
Although we know of no ceramics parallels, gi-
raffes have been depicted on some of the early Na-
gada III commemorative stone pallets (Asselberghs
1961: pls. 73: 130, 91: 161; the latter also shows a
(geometric) pattern indicating the pelt pattern). Gi-
raffes also appear on an Early Dynastic stone seal
found in the cemetery at Helwan (1988: 44, note 35).
b. Pottery fragment showing modeled parts of a leg
and two (human/animal?) feet (fig. 17.6). Reg. no.
G67-1995, Locus 861, B. 3174, Area D (Square L/
19), found in the same context as the giraffe sample
noted above, dating to Stratum lib (late EB IB).
This fragment is ca. 9.3 cm x 7.3 cm, and is 1.9-
2.2 cm thick; less well fired than the previous exam-
ple (no. 1 above); and much closer to the common
bread mold materials. On the exterior, traces of a
white slip have been preserved. On the basis of this
fragment, it is difficult to determine whether the leg
and two feet are human or animal-like.
Published parallels for modeled pottery are very
difficult to find, although B. Adam's (1986; 1994)
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
: .i:;;z .....
::: . .
Fig. 18. Modeled ceramic fragment from Coptos (courtesy
B. Adams).
makes it clear that uncoated, straw tempered mod-
eled clay objects were produced in Egypt already
during the Early Dynastic period. Compare, for in-
stance, a fragment from Coptos (fig. 18), showing a
human figure in relief (UC 34808). Although not
found in a stratigraphic context, it is usually taken to
be Early Dynastic.
Another instance of modeled pottery, this time
of "hand-supported, bulbous breasts beneath a crude
handle copying a Palestinian type made of straw-
tempered Nile silt" is published in Adams and Fried-
man (1992: 334-35, fig. 16). It was excavated at
Hierakonpolis, Upper Egypt, Locality 6, in Tomb
11, which is of Protodynastic date. Associated arti-
facts also include an imported, late EB I juglet.
Egyptian Amulet
Beetle-amulet pendant made of black stone (stea-
tite?), Reg. no. G67-95, Locus 104, B. 1302, Area A
(Square V1-2), found in a fill from Stratum lib/Illa
and dating to the late EB IB (fig. 17.2). The beetle
is 2.6 cm in length, the back 7 mm wide, triangular
in section, with a perforation (ca. 2-5 mm in diam-
eter) carefully drilled (and today partially broken)
through the upper part of the object.
The head of the beetle has been very schematically
indicated by a single short, horizontal incised line, the
prothorax by two short, horizontal incised lines and
the elytra by a single, incised line (Keimer 1931).
Although more elaborate examples of this beetle
amulet exist (Keimer 1931; Ward 1978: 43), it is
clear that this specimen represents Agrypnus noto-
donta Latr. The closest parallel to our example de-
rives from Nagada (Petrie 1914: pl. 43, 261a). Beetle
amulets are common in late Protodynastic/Early Dy-
nastic burial contexts in both Upper and Lower Egypt.
The type survives well into the Old Kingdom period.
On the basis of a Protodynastic palette fragment,
perhaps deriving from Abydos, that shows (twice) in
high relief this very beetle in association with the em-
blem of the Egyptian goddess Neith, a close connec-
tion between this insect and the deity may be assumed.
This is underscored by the discovery of a similar
beetle made of gold leaf, found in a Protodynastic
context by Reisner at Naga ed-Der in Upper Egypt.
The latter has the emblem of the goddess Neith
carved on its back (quoted in Keimer 1931: pl. 3, c l).
This would imply that the beetle most probably was
dedicated in Protodynastic times to this deity, who
had her main place of adoration in Sais in the north
Delta. The amulet was clearly meant to attract the
help and protection of Neith. To carry such an amulet
in antiquity implies that the bearer was associated
with the Halif Terrace site and accepted or was
aware of the religious concept behind it. The follow-
ing section summarizes the assemblage of human
remains found in the Nahal Tillah excavations.
The human population that utilized the artifacts
found in the Nahal Tillah excavations is represented
by a small, but important, sample of physical re-
mains. Consequently, a detailed study of the human
remains is presented here. The 1994-1995 excava-
tion seasons at Nahal Tillah produced three intact
burials and numerous random finds of human skele-
tal elements extending from the Chalcolithic through
MB II periods. One of the intact burials, NT95-3
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
was found in the Egyptian tomb, Locus 625, dated to
the Early Bronze Age. The other two burials, desig-
nated NT95-1 and NT95-2, were found adjacent to
walls in Loci 576 and 587 and have been dated to the
MB II. Preliminary descriptions of the three intact
burials are given below. Age and sex estimates were
based on standards given in Bass (1987) and Ube-
laker (1978). Measurements were taken according to
the definitions in Bass (1987).
NT95-1 (Locus 576) MB IIB
NT95-1 is the remains of a child aged 3-4 years,
found together with a jug and a juglet in fill adjacent
to a wall (fig. 11). Only cranial fragments, teeth, and
a few postcranial bones were recovered; but the po-
sition of the bones suggests that this was a disturbed
primary burial rather than a secondary interment.
Osteological Inventory. Bones present included
cranial fragments (frontal, parietal, occipital, and left
petrous), right ramus of the mandible, rib fragments,
and tibia and femur shaft fragments.
Dental Inventory. Only mandibular teeth were
found. They included central and lateral deciduous
incisors, one deciduous canine, the right deciduous
first and second molars, and the developing perma-
nent central incisors.
Pathological Conditions and Abnormalities. No
abnormalities were noted on the bones, but some
of the teeth were stained brown. This staining was
probably a postmortem occurrence.
NT95-2 (Locus 587)
NT95-2 is the complete skeleton of an infant ap-
proximately 18 months of age found in fill next to
two walls. This individual was lying on the right
side facing east and in a flexed position. No artifacts
were found with the burial, but the similarity of this
burial context to that of NT95-1 (i.e., in fill next
to wall features, at similar elevations, and in close
proximity), suggests that NT95-2 is also an MB II
Osteological Inventory. Bones found included
cranial fragments (frontal, parietal, occipital, right
orbital, right and left petrous), mandible (right and
left horizontal rami), rib fragments, unfused spinous
process of a vertebra, right and left scapulae, right
and left humeri, right and left radii, left femur, right
femur shaft, left tibia, and right tibia shaft.
Dental Inventory. Teeth included a mandible
with both deciduous and permanent teeth present;
deciduous central incisors, right deciduous lateral
incisor, deciduous right and left canines, permanent
right central incisor, permanent right lateral incisor,
and permanent right canine. Maxillary teeth included
deciduous right and left central and lateral incisors,
deciduous right and left canines, deciduous right and
left first and second molars, permanent central inci-
sor, permanent right and left lateral incisors, and a
permanent right canine.
Pathological Conditions and Abnormalities.
The teeth show the same brown staining seen in
NT95-1. There is some postmortem damage to the
bones, most likely from roots.
NT95-3 (Locus 625)
Excavation of this large Egyptian tomb is still
unfinished, but to date only one skeleton has been
found. It was identified as a young woman aged ap-
proximately 25-30 years. The body was lying on the
left side in a flexed position facing southeast, toward
the entrance of the tomb. Some potsherds and faunal
remains were recovered from the area around the
skeleton and one large potsherd was recovered from
beneath the head, but no other artifacts were directly
associated with the body. The potsherds recovered
date this burial to the EB IB.
This burial was found in the fill of the tomb,
above the later of the two surfaces defined, along the
north and east walls. Within this burial locus, sev-
eral medium-sized rocks (+/-20 cm in diameter)
were recovered in, on, and around the skeleton. The
skeleton was in good condition, however, which
would not have been the case if the rocks had fallen
on it from the roof. It is therefore likely that these
rocks were part of the original burial context. The
preservation of the skeleton was good, and all bones
were in anatomical articulation except for the cra-
nium. This was found in several pieces above the
lower limbs.
Identification of the individual as female was based
on the pointed chin, the rounded skull with small
mastoid processes, the obtuse subpubic angle of the
pubic bone, and wide sciatic notch of the ilium.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Osteological Inventory. All bones were present.
Dental Inventory. Maxillary teeth included the
left incisor; right canine; right and left first pre-
molars; right and left second premolars; right and left
first, second, and third molars. All mandibular teeth
were present except from the lower central incisors,
which appear to have been congenitally absent.
Pathological Conditions and Abnormalities.
Analysis to date has noted several pathological con-
ditions in this individual. There is a thickening of
the cranial bones in the region of the lambdoidal su-
ture, together with pitting on the internal surface of
the occipital bone and on the external surface of the
posterior parietal bones indicative of porotic hyper-
ostosis. The mandibular central incisors are most
likely congenitally absent since there is no evidence
of postmortem or antemortem removal of these
teeth, and there are no spaces for them in the dental
arch. There is a marked overjet and overbite of the
upper teeth. This has resulted in grooving along the
occluso-lingual surface of the maxillary posterior
teeth. The attrition pattern is fan shaped in the lower
incisal region with a steep incline (lingual to buccal)
on the premolars and molars. Calculus is evident on
all teeth, especially in the incisal region. There is
moderate to severe alveolar resorption in the max-
illa; the roots of the right second premolar, the right
second molar, and both first molars are exposed
lingually, and the root of the left canine is exposed
buccally. Such extensive alveolar resorption is un-
usual in an individual of this age.
The skull and other bones are still being cleaned
and restored, but some long bone measurements were
taken for stature and robusticity estimations and are
presented in table 2. At this stage of analysis it is not
possible to determine whether or not this individual
was of Egyptian origin, or indeed if the tomb was
designed especially for her.
The faunal remains collected during our excava-
tions provide an important indicator of changes in
the palaeoeconomy at the Halif Terrace site that can
be linked to greater issues of Egyptian-Canaanite
interaction. Of the 1,125 bone finds identified to
mammalian taxon in the 1994 assemblage, only 739
were from well-stratified deposits, 6 from the Chal-
colithic level, 344 from EB IA, and 389 from EB IB.
The identifications are summarized in table 3. For
EB IA and EB IB, these sample sizes are adequate to
indicate the range of species present and to give an
idea of their relative numbers; but more detailed
analyses of aging and sex patterns await the retrieval
of larger samples in future seasons of excavation.
At 6-10% the bones of wild animals make up
a substantial proportion of the assemblage (fig. 19).
They are all ungulates that could have been hunted
in the vicinity of the site: gazelles (Gazella sp.), wild
cattle (Bos primigenius), hartebeest (Alcelaphus sp.),
and wild pigs (Sus scrofa).
Sheep and goat bones make up over 70% of the
domestic assemblage; pigs comprise 4%. The remain-
ing 25% comprises cattle and equid bones (fig. 20).
The most surprising feature of the assemblage is the
large number of equid remains, some of which are
from domestic horses (Equus caballus). Some of the
smaller equid bones are definitely of domestic don-
keys (E. asinus); it is possible, though not at all likely,
that a few of the bones identified merely as equid
derive from wild onagers (E. hemionus) or asses
(E. africanus), rather than horses or donkeys.
In terms of meat weights (fig. 21) cattle comprise
50-61% of the ungulates (excluding equids), which
is normal for an Early Bronze Age site away from
the desert edge, where seven other Early Bronze Age
TABLE 2. Long Bone Measurements,
Skeleton NT95-3, Locus 625.
Bone element Measurement
Left femur Maximum length = 419 mm
Bicondylar length = 417 mm
A-P diameter mid-shaft = 25 mm
M-L diameter mid-shaft = 23 mm
Maximum diameter head = 39 mm
Left tibia Maximum length = 346 mm
A-P at n. foramen = 29 mm
M-L at n. foramen = 20 mm
Left fibula maximum length = 341 mm
Robusticity index A-P + M-L diam. X 100 = 11.04
(Bass 1987):
Stature estimation (Bass 1987):a
Caucasoid (2.32 femur + 65.53 ?3.94) 162.74 cm ? 3.94
Negroid (2.10 femur + 72.22 +3.94) 160.21 cm ? 3.94
aThese measurements are presented for comparative
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
TABLE 3. Mammal Bones by Strata, 1994 Excavations, Halif Terrace.
IV III lla/b I
Chalcolithic EB IA EB IB mixed
Taxon No. No. % No. % No. % Total no.
Goat 0 16 4.7 16 4.1 17 4.4 49
Sheep 0 27 7.8 36 9.3 27 7.0 90
Sheep/goat 2 169 49.1 212 54.5 219 56.7 602
Domestic cattle 0 26 7.6 52 13.4 51 13.2 129
Bos/equid 1 8 2.3 5 1.3 6 1.6 20
Donkey 0 4 1.2 8 2.1 5 1.3 17
Equid 1 26 7.6 10 2.6 20 5.2 57
Horse 0 7 2.0 4 1.0 4 1.0 15
Domestic pig 0 11 3.2 15 3.9 6 1.6 32
Dog 0 15 4.4 6 1.5 8 2.1 29
Goat (wild?) 0 0 0.0 1 0.3 0 0.0 1
Gazelle 2 28 8.1 14 3.6 11 2.8 55
Wild cattle 0 5 1.5 5 1.3 3 0.8 13
Hartebeest 0 2 0.6 3 0.8 7 1.8 12
Wild pig 0 0 0.0 2 0.5 2 0.5 4
Totals 6 344 100.0 389 100.0 386 100.0 1125
assemblages had a mean of 59.2% of cattle; seven
Chalcolithic assemblages had a mean of 72.6% (Grig-
son 1995).
The differences between the two EB I levels, EB
IA and EB IB, are intriguing, although more material
will be needed before we can be sure that they are
genuine. In fig. 22, which shows the relative propor-
tions of domestic ungulate remains, there appears to
be a decrease in the proportion of sheep and goat
bones from EB IA to EB IB, with a corresponding
rise in cattle. However, when equid remains are in-
cluded (fig. 20) it can be seen that the proportion
of sheep and goat bones is constant at about 73%,
whereas cattle increase at the expense of equids. In
both levels the number of pigs is constant.
A few donkey bones have been found in almost
all sites of the Chalcolithic and Early Bronze Age in
the Levant; these animals were probably used for
transport. Later texts describe highly organized don-
key trains carrying goods across country and this
must have been the backbone of long distance trade
on land. There was a general supposition that do-
mestic horses were not introduced into the Levant
and Egypt until the second millennium, but Davis
(1976) found horse remains at Arad from the third
millennium and small domestic horses seem to have
been present in the fourth millennium in the Chal-
colithic period in the northern Negev (Grigson 1993).
Heavy loads may have been pulled by draft cattle,
possibly in vehicles or on sledges.
The change from hoe to plow agriculture, which
enabled intensive and extensive cultivation of heavier
soils may already have taken place in the Chal-
colithic period. Domestic donkeys were used for
plowing by the third millennium in Mesopotamia
(Postgate 1986), but it seems likely that cattle would
have been used more frequently than donkeys. As
will be shown below, ceramic analyses provide one
of the best ways to identify the pace of change in the
palaeoeconomy and other dimensions of culture in
the Nahal Tillah region.
Ceramics provide one of the best indicators for
monitoring the rapidity of Protodynastic Egyptian
contacts with Israel. During the 1995 Nahal Tillah
excavations at the Halif Terrace and caves of Abu
Hoff, more than 3100 kg of pottery sherds were re-
covered, analyzed, and stored. In an effort to com-
plete the most comprehensive analysis possible, we
have expanded the number of parameters recorded in
the analysis of the 1995 assemblage. Most important,
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
WILD 10%
N =344 N= 389
Fig. 19. Proportions of wild and domestic animal bones, EB IA, EB IB.
-- i!iiii:::::: PI GS 4%
N =294 N =358
Fig. 20. Ungulate proportions, including equids, EB IA, EB IB.
given the flexibility of our computerized recording
system, this year's expansion in data collection is
readily and easily comparable to the results of the
1994 ceramic analysis. This year's new methods are
best seen as an enhancement and expansion of one
seamless and unified analysis that includes both the
1994 and 1995 assemblages.
As with the 1994 ceramics, the analysis of the
1995 assemblage was done in two parts. The first
part, which was begun in the field, was the iden-
tification of diagnostic sherds (rims, handles, bases,
and some other types) on a locus-by-locus basis.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
PIGS 3% t~~iii~iiiiiiiiiiIi~iiiiiiiii~iiiiiiii IS 3
CATTLE 50% /"0_ !!!!!!!!! CATTLE 61%
N= 252 N= 335
Fig. 21. Ungulate meat weights, excluding equids, EB IA, EB IB.
PICGS 4% ! jij;.. PIGS 4%
CATTLE 12% ii CATTLE 17%
N = 252 N = 335
Fig. 22. Ungulate proportions, excluding equids, EB IA, EB IB.
Parameters of vessel form, rim form, handle form,
decoration application technique and motif, and clay
type were recorded. In addition, a minimum number
estimate was made for quantitative purposes.
The second part of the analysis involved a much
expanded and enhanced analysis of these diagnostic
sherds (which make up about 10% of the assemblage)
and the nondiagnostic bodysherds. On a basket-by-
basket basis, the field school students separated all
sherds into the ten major clay types (as well as an
"other" category for nonidentifiable clays) by visual
inspection. In addition, the sherds were further sub-
divided on the basis of decoration, then counted and
weighed. Thus, we have a count of the number and
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
weight of the different clay types and decoration tech-
niques for each basket excavated in 1995 (excluding
the heavily disturbed Stratum I and probe loci).
A third approach was developed for the study
of the diagnostic pottery fragments associated with
Tomb 1. In addition to recording the same parame-
ters used for 1994 and 1995 diagnostic sherd analy-
sis, counts of the number of rim, handle, base, and
other diagnostic sherds were made. In this way, trends
(both chronological and spatial) can better be stud-
ied, without solely relying on the somewhat prob-
lematic minimum number estimate.
Stratigraphic Characterizations
The pottery assemblage of the Halif Terrace is
very diverse, widely varying in form, decoration,
clay type, and ethnic tradition. Much of this variance
can be seen in the chronological dimension, because
the different stratigraphic phases hold different ce-
ramics. In the final occupation phases in Stratum IIa
and IIb, two ceramic traditions are present, one Egyp-
tian and the other Canaanite.
Stratum IV Stratum IV, the deepest and oldest
occupation phase identified at the Halif Terrace has
a Chalcolithic ceramic assemblage (paralleling sites
such as Shiqmim, Bir-es Safadi, and Abu Matar). As
seen in fig. 23a, the primary clay type used in this
stratum is large-grained crushed wadi grits. This clay
type is used for all types of forms, ranging from
small open vessels (such as V-shape bowls) to large
closed vessels (such as pithoi) and specialized vessels
(such as churns and pedestaled bowls). Sherds of
vessels made with this clay type tend to be between
2 mm and 1 cm in thickness, lacking a firing core,
and have a brown-orange color. Large grits (about
0.5 mm-1 mm) usually black, are visible in both the
sherd's cross section and its surface. The shapes tend
to be fairly regular, with somewhat smooth exterior
surfaces. The most common types of decoration are
the "scrapped slip" (a beige slip, applied then re-
moved in horizontal bands, leaving many horizontal
beige bands around the body of the vessel), a red
painted horizontal band, usually at the rim, or shoul-
der or a vessel, and more rarely a combination of
these two techniques.
Small fine-grained grits also peak in Stratum IV,
although this peak is very small in figs. 23a and 24a.
This clay type is used exclusively for small open
vessels, such as V-shaped bowls and cornets. Body
sherds of this type vary in thickness between about
1 mm and 4 mm. (Because of their small size, this
clay type only has a small peak in figs. 23a and 24a.)
Sherds of this type tend to vary between a creamy
orange color and cream, and the grit inclusions are
barely visible without magnification. Red painted rim
bands were the only type of decoration seen on this
clay type.
Stratum IIIb. A sizable sample of the pottery
from the underground room Locus 272 and pit
Locus 271 was reanalyzed, to better compare it to
this year's sample. We found that the underground
rooms and pits of Area B in the 1994 excavation
held a peculiar pottery assemblage, which although
designated as "Stratum III," differs greatly from the
Stratum IIIa unearthed in Area A in 1995. Unfortu-
nately, we have no direct stratigraphic information
on the sequencing of the semisubterranean rooms
and pits of Area B to the clear sequence in Area A.
Thus, based solely on their ceramics, the under-
ground rooms and pits of Area B, designated as
Stratum III in 1994, will be termed Stratum IIIb in
this analysis.
As seen in figs 23b and 24b, the "grog and grit"
clay type peaks dramatically in Stratum IIIb, as does
the "brittle straw" clay type. These two types of clay
characterize the EB IA period. Grog and grit sherds
tend to be very weak, crumbly, and porous with very
rough, cracked, and pitted surfaces. This clay type
was used in forms varying from large bowls to large
closed jars. The composition of grog and grit ap-
pears to be miscellaneous household debris varying
from mostly grog to grog and straw, to grow, straw,
and large crushed grits. The brittle straw clay type
seems to have had a more specialized use than the
grog and grit. The vast majority of the sherds of this
clay type are of the finger-pinched manufactured
hemispherical bowls. The brittle straw clay contains
varying amounts of different grits (limestone, sand,
and grog) and tends to have a brown color.
The ceramics of Stratum IIIb were relatively
crude compared to the rest of the site's assemblage.
The potters who produced these wares seem to have
invested very little effort in creating regular forms
and decoration. These ceramics seem to convey little
aesthetic or symbolic information and may not have
been a very important medium for such expression.
Possibly the Stratum IIIb ceramics reflect results of
the collapse of the Chalcolithic societies in southern
Palestine. With demographic decline, shift in set-
tlement, and economic changes in a possible shift
towards a more "commercial" commodity-oriented
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Chalcolothic clay types, by stratum
50% -- Small grain
E 40% grit
--- Large grain
S30% grit
~ 20%
2a 2b 3a 3b 4
Stratum a
EB IA clay types, by stratum
m 60%
E 40% -----Grog and grit
-U----Brittle straw
o 20%
2a 2b 3a 3b 4
b Stratum
EB IB clay types, by stratum c
o -- Calcite
S40% -- --Dolomite
m -- Grog
-, 30%
2a 2b 3a 3b 4
d Egyptian clay types, by stratum
? 60%
8 50%
- --Hard straw
: 40% --fl-Bread mold
30% -Egyptian
-5 30%-
2a 2b 3a 3b 4
Stratum Calcite Dolomite Grog Grog and grit Small grain grit Large grain grit Brittle straw Hard straw Bread mold Egyptian Total Kg.
2a 5% 2% 22% 4% 2% 10% 1% 10% 19% 20% 452
2b 7% 1% 16% 4% 1% 7% 1% 11% 34% 15% 2,009
3a 75% 5% 57% 18% 4% 24% 2% 2% 2% 5% 132
3b 14% 0% 3% 52% 4% 12% 8% 2% 1% 0% 104
4 10% 0% 8% 7% 7% 63% 2% 1% 0% 0% 108
Fig. 23. Ceramic clay types, by stratum (% of stratum total kg).
economy, the social relations involved in pottery pro-
duction may have also changed (Levy 1995; Gophna
1995a). Political and social boundaries may have be-
come more fluid in the post-collapse EB IA. In this
situation, people may have required less regular and
consistently patterned pottery to communicate social
boundaries that were less discrete than Chalcolithic
social structures. Additionally, the relative crudeness
of these Stratum IIIb ceramics may reflect a loss of
some somewhat specialized knowledge of ceramic
production techniques.
Stratum lila. Stratum IIIa was reached only in
a fairly limited exposure in Area A. Judging by the
pottery recovered from that phase, this limited area
held some fairly specialized activities. As seen in
figs. 23c and 24c, calcite is the most common clay
type recovered from this stratum. Because of calcite's
thermal properties, it is widely assumed that it was
used for cooking pots; the fact that calcitic tempered
sherds are often burnt helps to confirm this assump-
tion (Dessel 1991; Y. Yekutieli personal communica-
tion 1996). Calcite seems to come into use in EB IA
(Stratum IIIb) and is very common in EB IB (Strata
IIa-IIIa). The large proportions of calcitic tempered
sherds found in this stratum probably reflect some
sort of cooking activity and related refuse in the
Two other clay types, dolomitic temper and grog
temper, also become common in Stratum IIIa. These
clay types (together with calcite temper) are the
most common clays used in the EB IB for forms.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Chalcolothic clay types, by stratum EB IA clay types, by stratum
S60% 60%
o 0
- 50% - 50%-
S40% - 40%
S-4- Small grain o
E 30% grit E 30% - --Grog and grit
S-- Large grain - --- Brittle straw
S20% - 20%
10% 1 10%
0% q 0%; ? , ,
2a 2b 3a 3b 4 2a 2b 3a 3b 4
Stratum a b Stratum
EB IB clay types, by stratum c d Egyptian clay types, by stratum
S60% 60%
S50%- 50%-
fi) U,
S40%- 40%
O -4- Calcite C -4--Hard straw
E 30% ' ooie E 30%-
30% Dolomite -- Bread mold
-s- \ ,Grog ---~ - Egyptian
L 20%- - 20%
10% 10%
0% 0 I %
2a 2b 3a 3b 4 2a 2b 3a 3b 4
Stratum Stratum
Stratum Calcite Dolomite Grog Grog and grit Small grain grit Large grain grit Brittle straw Hard straw Bread mold Egyptian Total
2a 6% 2% 31% 4% 3% 10% 1% 6% 4% 22% 25,355
2b 10% 1% 24% 5% 1% 9% 1% 10% 8% 19% 82,915
3a 14.6% 3.7% 37.3% 10% 1.8% 13.4% 2% 1.1% 0.5% 3.3% 10,062
3b 7% 0% 2% 42% 7% 13% 12% 0% 0% 0% 5,844
4 6% 0% 12% 5% 27% 45% 1% 1% 0% 0% 9,670
Fig. 24. Ceramic clay types by stratum (% of stratum total sherd count).
Dolomitic-tempered vessels are more often decorated
with whitewash and red painted lines than grog-
tempered vessels (fig. 25a), and this difference is
especially noticeable in Stratum IIIa. Perhaps tied
in with higher incidence of decoration, dolomitic-
tempered sherds tend to be harder, more regular, and
smoother than grog tempered sherds. Grog-tempered
vessels tend to be fairly regular, with somewhat
smoothed surface although usually less so than do-
lomitic-tempered sherds. The overall quality of these
sherds is noticeably higher than that of the grog and
grit sherds of Stratum IIIb.
One potentially important change from the previ-
ous two strata is the relative increase in the number
of storage vessels and the decrease in the number of
serving vessels (fig. 25b). While the sample is quite
small in this stratum (and may reflect specialized
cooking activities), in all following strata this in-
crease in storage vessels and decrease in serving ves-
sels continues. Ethnographic evidence suggested that
serving wares tend to have higher breakage rates than
storage vessels (Orton 1993), so the observation of
large quantities of storage vessel sherds is somewhat
unexpected. Several issues surrounding the problems
of minimum numbers, especially with regard to bowl
counts, need to be further addressed in looking at
this problem. But this may still be an interesting ob-
servation and may reflect the changing nature of the
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
% of grog and dolomite sherds with wash and painted line decoration
o 30%
2 25%
" 20%
15 -- % dolomite with
15%- wash and paint
---- % grog with
-'10%- wash and paint
V 5%
0% 1 I
2a 2b 3a 3b 4
% dolomite % grog Total Total
with wash with wash dolomite grog
Stratum and paint and paint sherds sherds
2a 6% 2% 574 7,745
2b 5% 3% 1,160 20,262
3a 32% 6% 375 3,750
3b 0% 0% 4 128
4 0% 0% 32 1,201
Whitewash and paint decoration were found only on
sherds of dolomite or grog clay types
Fig. 25a. Percentage of grog and dolomite with wash and painted line decoration.
settlement's economy, which required more storage
in later periods.
Stratum lc. Stratum IIc was reached in very
limited exposures in Area C. A preliminary look at
the pottery of this stratum shows that it most closely
parallels that of Stratum IIIa. Since it is doubtful that
direct stratigraphic sequencing can be used to con-
nect Stratum IIc with the clear sequence in Area A,
some sort of relative method is needed. Although
one possibility is to use some form of seriation, at
this point it is probably premature, since only 4 kg of
pottery was recovered from Stratum IIc, making for
a very small sample size. Perhaps the 1996 season
will increase the sample from this phase, making fur-
ther study possible.
Stratum lib. The exposure of Stratum IIb was
the largest of any single phase at the Halif Terrace.
Generally, the character of the assemblage changes
dramatically, with a huge jump in the prevalence of
Egyptian pottery (figs. 23d, 24d). Because of the
difficulty in visually distinguishing among marl clays,
Nile alluvial clays, and locally obtained clays, these
separate clay types were lumped together under the
category "Egyptian." Sherds of this category tended
to have smaller inclusions than the local Canaanite
wares, were more regularly formed, and usually were
harder. Petrographic analysis outlined in this report
indicates that all of the sampled "Egyptian" sherds
were locally produced. The sherd bearing the Narmer
serekh, recovered in the 1994 season, stands alone as
a petrographically observed Egyptian import (Levy
et al. 1995). Open and closed vessel fragments of the
"Egyptian" clay type were recovered. The closed
vessels tended to have very standardized and similar
rim shapes and were usually undecorated, while the
vast majority of the open vessels had slip or burnish
We have identified another Egyptian clay type,
termed "hard-straw." Sherds of this type were char-
acterized by visually observable vegetable inclusions,
greater thickness than other clay types, and greater
strength than the other straw-tempered clay type,
"brittle straw." Potters locally worked the "hard-
straw" clay type to fashion mostly large basins (prob-
ably with storage and dough and beer production
uses; Chazan and Lehner 1990) and lotus bowls. The
bread mold clay type, used exclusively for bread
mold baking vessels, is similar to the hard-straw clay
type but includes miscellaneous large inclusions in
addition to straw. Bread molds were also less care-
fully formed. Petrographic evidence and common
sense (since they are so bulky) indicate that bread
molds were also a result of local Egyptian produc-
tion. Both of these clay types tended not to be dec-
orated, with the exception of some pictorial reliefs
on hard-straw sherds and the rare use of the Canaan-
ite whitewash decorative technique. Meanwhile, the
Canaanite EB IB clay types of grog, dolomite, and
calcite persist and were recovered in this stratum
(figs. 23c, 24c).
Stratum Hla. The pottery assemblage from Stra-
tum IIa differs little from that of Stratum IIb (figs.
23d, 24d). The most important difference is the
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Stratum Storage Serving Total Min
2a 49% 32% 621
2b 40% 36% 2,099
3a 48% 32% 113
3b 20% 77% 329
4 19% 66% 125
Trends in numbers of storage and serving vessels;
Egyptian and Canaanite types Included
50% 1
40 - -- Storage
, 30% !
. 20% - -
; 10% - -
o% I I I
2a 2b 3a 3b 4
Fig. 25b. Trends in numbers of storage and serving vessels, including Egyptian and Canaanite forms.
smaller quantity of bread mold fragments recovered
from this stratum. This probably results from the
lack of a comparable structure in Stratum IIa to
Area D's Stratum IIb tabun (discussed below). As
seen in fig. 26b,d, the quantity of bread molds in
Stratum IIa dropped off in Area D from the previous
stratum. This may indicate that the tabun had gone
out of use in the later phase. Otherwise, the Stratum
IIa pottery assemblage is indistinguishable from that
of Stratum IIb.
Ceramic Spatial Patterns
Because the exposure of Stratum IIb is so large,
some spatial characterizations can be made. The char-
acterizations made here are very general, and are be-
tween Areas A, C, and D. No attempt has been made
in this preliminary analysis to sort through the dif-
ferent types of context, so all loci in these areas
(including very mixed fills) were used in this com-
parison. Later examination will focus more attention
on taking contextual data into account. Nevertheless,
some very interesting trends are evident.
Area A, Stratum lib. In Area A, the Egyptian
clay types ("Egyptian," bread molds, "hard straws")
make up most of the assemblage (fig. 26d). Of this
Egyptian assemblage, a slight majority are bread
molds. A similar amount of the pottery (by weight) is
imported Egyptian ware. A mix of mostly EB IB clay
types makes up the Canaanite assemblage (fig. 26c).
Area C, Stratum lib. In Area C, we see a differ-
ent story. Canaanite pottery is far more common here
than the Egyptian pottery, suggesting either some
chronological difference or a spatial-functional dif-
ference with other areas (fig. 26c,d). Also common
in this area are Chalcolithic clay types and forms.
This suggests some interesting site formation pro-
cesses at work, in which much earlier material has
been mixed into this assemblage. Further investiga-
tion is needed with particular attention to contexts
that seem less disturbed if we are to understand the
meaning of the Area C's differing ceramic assem-
blage. In the next season, it may prove very worth-
while to tie in the stratigraphic relationship of this
area to the rest of the site.
Area D, Stratum lib. One of the most dramatic
findings of the analysis of the pottery from this area
is the huge number of bread mold sherds associated
with the tabun structure (fig. 26d). Over 500 kg of
bread molds were recovered from this area. A pre-
liminary investigation of the weights of individual
bread molds showed that they vary between 2.5 and
5 kg. Thus, we have recovered a minimum of 100 to
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
Stratum 2a: distribution of Canaanite functional groups Stratum 2a: distribution of Egyptian functional groups
E 40% E Storage
, I' Storage . " Serving
E 30% Serving - E
SICook Cook
ac d a b a c d
Area AreaPreparation
Stratum 2b: distribution of Canaanite functional groups Stratum 2b: distribution of Egyptian functional groups
0% c d10%
20"/ I!' Flour
0% 60%
a b c d a b c d
Stratum 2b: distribution of Canaanite functional groups Stratum 2b: distribution of Egyptian functional groups
60% C d
50% 60%
50% U rStorage
E 40%~ E
r S 40% U Serving
SIStoragel c
30% "' Cook E
o0 Serving -30% Cook
2E 20% 0 Flour
510o% tj 10%
a b c d a b c d
Area Area
Canaanite Egyptian
Total Min. Flour
Stratum Area # Storage Serving Cook Storage Serving Cook Preparation
2a a 267 22% 11% 13% 24% 12% 10% 8%
2a c 289 39% 34% 14% 3% 6% 0% 3%
2a d 65 17% 11% 11% 22% 22% 8% 11%
Canaanlte Egyptian
Total Min. Flour
Stratum Area # Storage Serving Cook Storage Serving Cook Prepration
2b a 754 18% 14% 14% 19% 16% 11% 9%
2b b 56 25% 45% 11% 4% 9% 2% 5%
2b c 514 26% 58% 6% 3% 4% 0% 3%
2b d 775 12% 6% 11% 19% 18% 26% 9%
Canaanite Egyptian
Total Min Flour
Stratum # Storage Serving Cook Storage Serving Cook Preparation
2a 621 29% 22% 13% 14% 10% 5% 6%
2b 2099 18% 22% 11% 15% 13% 14% 7%
3a 113 47% 32% 20% 1% 0% 0% 0%
3c 329 20% 77% 3% 0% 0% 0% 0%
4 125 19% 66% 14% 0% 0% 0% 0%
Notes: "Serving" vessels are all bowl forms.
"Storage" vessels are closed forms (pithoi, jars, jugs, juglets, non-calcite tempered holemouths) and Canaanite basins.
"Cooking" vessels are calcite tempered holemouth jars (Canaanite) and bread-molds (Egyptian).
"Flour Prep." vessels are basins (Egyptian case only, Canaanite are arbitrarily assumed to have a storage function).
Functional interpretations are very preliminary and are intended only for gross spatial comparisons.
Fig. 26. Distribution of Canaanite and Egyptian pottery by functional groups and Egyptian vs. Canaanite pottery.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
200 individual bread molds in Area D. The bread
molds associated with this tabun structure may re-
flect some centralized food redistribution, a smaller
parallel to an Old Kingdom baking facility at Giza
(Chazan and Lehner 1990). The other two Egyptian
clay types were also significantly represented in this
area, as were small amounts of Canaanite EB IB pot-
tery (fig. 26c). This small amount of Canaanite pot-
tery and the huge predominance of Egyptian pottery
contrasts dramatically with the mostly Canaanite as-
semblage in the adjacent Area C.
Establishing ethnic identity based on material cul-
ture found in the archaeological record is extremely
difficult and subject to much debate. Nevertheless,
some patterning of the ceramic assemblage of Stra-
tum IIb probably relates to the assertion of different
social identities. In evaluating identity relationships,
recent work has shown the value of considering spa-
tial patterns and stylistic elements, and evaluating
possible social distributions of information (Shennan
1989). Consideration of these factors was applied to
the Nahal Tillah ceramic assemblage.
The petrographic analysis, based on a small num-
ber of sherds, indicates that the majority of the
ceramics of both the Egyptian and the Canaanite tra-
ditions were probably manufactured in the vicinity
of Nahal Tillah. Thus it appears that potters working
in these two distinct manufacturing traditions pro-
duced differing ceramics at the same site (or at least
in the immediate vicinity of each other). This social
distinction was expressed not only in the production
of pottery, but also in the spatial segregation of pot-
tery use and disposal. As shown, the Egyptian pot-
tery concentrates in Areas A and D, while Areas B
and C contained mostly Canaanite ceramics. This
suggests different spatial patterns in the disposal of
pottery along lines that may tentatively be called
As expressed in the different ceramic traditions,
the people who occupied the Halif Terrace main-
tained at least two separate social identities, one
"Egyptian" and the other local "Canaanite." The
terms "Egyptian" and "Canaanite" may be problem-
atic, since it is difficult to know how the actors in the
past conceived of their social differentiation. Sup-
port for a possible ethnic distinction may be seen in
the two distinct cooking traditions present at the site:
an Egyptian cuisine, which appears to have mostly
used bread molds, and a Canaanite cuisine, which
used calcitic tempered holemouth jars as primary
food preparation vessels.
In addition, there appears to have been a relative
lack of pottery "hybrid" combinations of Egyptian
and Canaanite elements. The few hybrid examples
seen appear to be superficial Egyptian copies of
Canaanite forms or elements of Canaanite forms.
The term "superficial" is used since, while the fin-
ished product of these hybrids may appear similar
to Canaanite prototypes, manufacturing details of the
Egyptian copies differ from the local Canaanite pro-
duction techniques. One such example is seen in an
intact vessel recovered from Area A (fig. 27). The
clay type of this vessel closely matches that usually
seen in standard Egyptian forms, as does the careful
surface treatment. Similarly, while the vessel has a
Canaanite-like everted rim and ledge handles, its
overall form is broader at the shoulder and narrower
at the base than Canaanite prototypes. So, while the
vessel displays many Canaanite elements, its pro-
duction details more closely match standard Egyp-
tian techniques. Paralleling this vessel, some ledge
handles appear to have been constructed with a
characteristic "Egyptian" paste, were more carefully
shaped, and were produced with a separate interior
component. Again, such hybrid examples are rare
and superficial.
The limited number of ceramic hybrids suggests
that the Egyptian and Canaanite potters did not
freely share information of pottery manufacture with
each other. Even if information were shared, these
potters seem to have rarely put it to use. This lack
of pottery hybrids and information sharing probably
indicates some social distance between the Egyp-
tian potters and the local Canaanites. Possible fac-
tors leading to such social distance range widely and
can include linguistic differences and status differ-
entials. Additionally, information sharing may have
been limited by formalized relations between these
two groups with contact mediated through a subset
of individuals.
The Tomb
Because of time constraints, the pottery (all in
sherd form) from the tomb has only been partially
analyzed. However, some broad generalizations are
possible based on the results so far. Fill loci 625 and
655 contain both Canaanite and Egyptian EB IB
pottery, and the surprising presence of what is typi-
cally called Canaanite EB II pottery. This EB II pot-
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
.~.:. .?-?;i
ia~s~a~g~a~lllll ?? ??:?;::~
r! ??:
;?; :?
?: ?-:1.?~
:i.:: r
::i::?:;?::::: i
"?'lli'""il: ??:.:::
'I . ~u::i. ;: ? . i
*; "":'::?I :i
a ?ik,"::9I~I
?'?:'' :? .n.
??:~ :???
........ .:. n.:
:; ..: ? :i?? ?. ?..? ?il::B:
::..~~....: .?::'r-:~.?'? :i
:a? ~;:
?' ~ .r: ?:;?:?:
i I. . :::::I.:.lii:i:.i i.!:P.:i'(i?l:",,,,,;::;
:.......:.; :B
Fig. 27a,b. Egyptian copy of Canaanite jar with ledge handles, Silo site, Area A, Halif Terrace.
tery is characterized by different forms (high-necked
jars) and decoration (red slips and burnishes on grog
clays). The presence of these "EB II" sherds in EB
IB contexts will require a detailed analysis of the de-
velopmental history of these wares in southern Pal-
estine. It is still unknown if any Egyptian pottery
contemporaneous with the "EB II" has been found in
the tomb loci. The loci below the latest of the tomb
floor deposits is characterized by EB IA pottery very
similar to that found in Stratum IIIb. This suggests
a very long, multiphased use of the cave which was
later transformed into the tomb chamber.
The transformation of the uses of this cave/tomb
are important. The earliest use of the Tomb 1 cave
had elements we have interpreted as having cultic
functions, at a time prior to signs of an Egyptian
presence at the site. Later use of Tomb 1 included
construction of the long east-west dromos and mas-
sive interior wall. These later additions appear to have
transformed the cave into architecture that closely
parallels Egyptian tomb design. This architectural
"Egyptianization" of the cave with possible prior
Canaanite cultic significance may have been an ex-
plicit display of power relations between the newly
arrived Egyptians and local Canaanites.
To supplement the ceramic analysis, some 215
sherds from the Nahal Tillah excavations were
selected for petrographic analyses, representing all
forms, loci, and levels at the site. For the preliminary
report, 44 sherds were randomly chosen for analy-
ses. The final report will present a complete study
of the entire petrographic assemblage. For the study
presented here, thin sections were made from each
sherd for examination under the petrographic micro-
scope, using magnifications of x25 to x400. The
temper and matrix of each sample were identified
and described. The firing temperature was estimated
on the basis of mineralogical changes found in the
sherds and was assigned a value of low, medium, or
high (L, M, or H).
The samples were divided into petrographic groups
according to criteria established by Porat (1989).
Both the composition of the temper and the matrix
were important for the classification into groups. The
samples from Nahal Tillah were compared to a large
collection of thin sections of Early Bronze ceramics
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
from all over Israel. The sources of the raw materials
were determined on the basis of geological maps of
Israel and the comparative collection.
The petrographic data of each thin section are
listed in table 4. Some highly fired samples could not
be assigned to a group because many of the features
were obliterated by the high temperature. Below is a
description of each petrographic group and which
forms belong to it. The codes used in table 4 to
describe the temper components and clay types are
given in the detailed description of each group.
Group 1 is well defined whereas groups 2 and 3 are
more general and may be divided into subgroups in
the future, when more data are required.
Group 1, Loess
The sherds are lightly tempered with rounded
limestone grains. The grains have a micritic texture
and some show evidence of originating in pedogenic
carbonate nodules, such as a concentric or zoned
structure and inclusions of silty quartz. In several
samples rare land snail fragments are found (bl).
Other rare temper components in this group are si-
licified limestone, mudballs, shells, sand size quartz
(qz) or chert (ct) grains.
The matrix contains abundant, well-sorted coarse
silt-sized quartz and silt with finely dispersed carbon-
ates (clay type 2). Heavy minerals such as hornblende,
epidote, iron oxides, sphene, and mica accompany
the quartz. Clay coatings are occasionally found on
the silty quartz. Very rarely, poorly preserved fora-
minifera and ostracoda (microfauna) fragments are
A subgroup of Group 1 (Group la) contains straw
molds (vg) and large vesicles and is generally more
coarsely made than the main group. The firing tem-
perature of this group varies, from < 700 to 900 C.
The source of the raw materials used for manufac-
turing this group is loess. The association of silty
quartz and heavy minerals together with finely dis-
persed carbonate is typical of loess in the Negev.
The presence of the pedogenic carbonate nodules in-
dicates that the loess was modified by soil forming
processes characteristic to loess in an arid region such
as the Negev. In some samples, small amounts of
weathered limestone, chalk, or marl are mixed with
the loess, either naturally as a fluvially redeposited
loess or intentionally added by the potter. Loess with
such pedogenic features is found in Nahal Tillah site.
Therefore the vessels belonging to this petrographic
group were most likely manufactured on site.
All the Egyptian vessels analyzed so far belong to
this group (table 4). Several of the special vessels,
including a sieve (No. 193) or a sherd with bamboo
relief (No. 198) are also made of loess. From the ves-
sels not defined as Egyptian, several bowls belong to
this group (Nos. 219, 276).
Group 2: Dolomite
All samples containing dolomite, either as sand-
sized temper or as silt-sized in the matrix, were in-
cluded in this group. It is not as uniform as Group 1;
however dolomite is the common feature.
The temper contains one or more of the following
components, in decreasing order of occurrence: do-
lomite, usually rhomb-shaped, occasionally with a
fibrous texture; limestone of varying sizes, shapes,
and textures; calcite, crushed, often zoned and with
large crystals; shale fragments, often containing silty
dolomite; grog, usually dark and opaque, occasion-
ally with dolomite remains; and quartz, rounded, rare.
In most samples belonging to this group, the
matrixes contain silt-sized dolomite crystals, often
rhomb shaped and zoned (clay type 4). Other sam-
ples have a calcareous clay without microfauna (clay
type 6) or the matrix is rich in clay minerals (clay
type 5). The firing temperature of this group varies
from low, for which all petrographic features can be
observed, to high, for which only some dolomite
crystals remain visible.
The source of the dolomite group is in the Judaea-
Hebron mountain region. Dolomitic clays and rocks,
suitable for pottery production, are found in rocks of
Cenomanian age, exposed - 15 km northeast of Nahal
Tillah. The vessels could have been produced either
locally using clays brought from some distance or im-
ported from sites in the Judaea-Hebron mountains.
Jars of varying sizes and one bowl belong to this
petrographic group. A few holemouth jars, typically
tempered with calcite, are also made of dolomitic clay.
Group 3: Calcite
The temper components of the sherds in this
group vary as much as those of Group 2 and have
similar textures and properties. The most common
temper is crushed calcite, with coarse, zoned crys-
tals. Other components, in addition to those described
in Group 2 are marl and chalk, with microfauna;
various iron oxides, various components with pedo-
genic features; and mudballs. The matrix varies. It is
usually calcareous (clay type 6) and often contains
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
TABLE 4. Description of Petrographic Samples.
Sample Temper Clay Firing Pet.
no. Locus Basket Description 1 2 3 4 5 type temp. group
188 625 2227 EB II lug handle mr qz fe is 1 L 3
189 874 3118 Ledge handle gr ls 4 M 2
190 874 3128 Lug handle dl 4 L 2
191 829 3066 Red slip and burnish gr sf qz oo 1 L 3
192 64 1062 Brown painted net, small Is qz sf gr 3 H ?
193 885 3177 Sieve, red slip and burnish Is pd bl 2 L 1
196 847 3170 Red slip/burnish, Eg? Is pd ck qz ct 2 L 1
198 92 1242 Bamboo relief, whitewash? Is 2 L 1
219 79 1201 Bowl base ls bl 2 L 1
220 15 287 Holemouth bowl cc ls qz ck 1 L 3
221 294 1177 Holemouth bowl cc 6 L 3
268 86 1289 Holemouth bowl cc ls dl sf 4 L 2
268 86 1251 Holemouth bowl cc 5 L 3
270 86 1288 Holemouth bowl cc ls dl 5/4 M 2/3
271 86 1181 Holemouth bowl cc ls 4 M-L 2
273 86 1289 Storage jar, body sherd dl 4 H 2
274 86 1288 Storage jar, body sherd dl 6 M
275 86 1288 Bowl dl 5/6 M-H 2
276 86 1168 Hemispheric bowl Is vg 2 L 1
278 86 1288 Storage jar, body sherd gr del cc qz 4 M-H 2
280 86 1251 Storage jar, body sherd sf gr ls 5/6 H 2/3
284 86 1251 Storage jar, body sherd, gr cc ls qz sf 4 M-L 2
red wash and paint
286 86 1168 Storage jar, body sherd ls pd cc qz 4 L-M 2
310 10 168 Eg. storage jar, body sherd ls bl 2 L 1
315 10 168 Storage jar, body sherd ck pd ox 1 L 3
319 102 1275 Eg. bowl rim ls bl mb pd 2 L 1
321 102 1275 Eg. bowl rim Is bl ct 2/1 L-M 1
322 102 1275 Cylindrical jar ?? 2? VH ?
325 102 1275 Eg. storage jar, body sherd ls bl 2 L 1
327 102 1275 Eg. bowl Is bl 2 M 1
329 102 1275 Eg. bowl Is 2 L 1
331 885 3276 Eg. bowl lis qz 2 M 1
332 884 3276 Breadmold vg ls mb bl 2 L la
336 884 3273 Eg. storage jar, body sherd ls pd 2 M 1
337 625 2330 Eg. bowl rim Is pd 2 L-M 1
338 625 2330 Lotus bowl rim ls bl 2 L la
339 625 2330 Lotus bowl rim Is 2? M la
341 884 3273 Storage jar, body sherd, Is? mb 2/6 H ?
348 870 3288 Holemouth bowl Is 4 VH 2
344 864 3273 Holemouth bowl cc ck mb 1 L 3
345 864 3273 Amphoriskos dl sf gr 5 L 2
349 861 3169 Eg. storage jar ls pd 2 M-H 1
354 102 1343 Eg. bowl ck ls 1 L la
385 625 2333 Holemouth bowl bs vl Is 6 L ?
The temper components are listed in decreasing amount; i.e., 1 is the most common component listed in column 1 of the section on temper.
Abbreviations: bl-biogenic limestone, bs-basalt, cc-calcite, ck-chalk, ct-chert, dl-dolomite, fe-iron oxides, gr-grog, Is-limestone, mb-
mudballs, mr-marl, oo-oolites, ox-pedogenic iron concretions, pd-pedogenic carbonate, qz-quartz, sf-shale fragments, vg-vegetal material,
vl-volcanic. Clay types: 1-calcareous with microfauna, 2-calcareous with silty quartz, 3-noncalcareous with silty quartz, 4-dolomitic,
5-clay-rich, 6-calcareous. Firing temperatures: L-low (< 7500C), M-medium (750-8500C), H-high (850-9500C), VH-very high (> 9500C).
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
microfauna such as foraminifera and ostracoda (clay
type 1). Only rarely is it rich in clay minerals (clay
type 5). The firing temperature of the vessels is usu-
ally low, less than the decomposition temperature of
calcite (- 750-8000C).
The source of the raw materials is in marl and
chalk rocks, with some components contributed by
soils developed on these rocks. Formations made of
chalk and marl, of Senonian and Turonian age, out-
crop 2-3 km east of Nahal Tillah. The vessels were
most likely produced on the site, the raw materials
collected from nearby. Holemouth jars are the most
common in this petrographic group. Less common
are store jars and jugs of various sizes.
Samples that do not belong to any of the above
groups, or that are worth special discussion, are dis-
cussed below. No. 322 and No. 341 are too highly
fired to be assigned to a petrographic group; how-
ever, both were made of calcareous clays and are
most likely local. No. 322 contains abundant silty
quartz and thus could be a highly fired version of
Group 1. No. 341 contains decomposed limestone
and could also belong to Group 1. Holemouth jar
No. 385 contains volcanic rock fragments as temper.
The volcanic rock has a trachytic texture with pla-
gioclase laths, vesicles, and glassy fragments. Pheno-
crysts of plagioclase and pyroxene occur, as well as
olivine altered to iddingsite or to iron oxides. Very
rare carbonate grains are found. The matrix is rich in
clay minerals, with silty carbonate and rare quartz.
The firing temperature was low.
In general, the rock fragments found in the sherd
are derived from basalt. However, comparing this
sherd to samples from northern Israel, where volca-
nic rocks are exposed, reveals many differences be-
tween it and common pottery from, for example, the
Golan Heights The source of this vessel has yet to be
determined. Sample no. 192 is made of a clay-rich
matrix, unlike any other vessel, but the temper com-
ponents resemble both those of Group 2 and Group 3,
so this vessel may be local.
Petrographic Summary
The results so far show that all Egyptian vessels
are made of loess, a local clay. The use of this clay
for local production of Egyptian pottery mirrors find-
ings from other EB I sties in the Negev such as Tel
Erani, cEn Besor, and others (Porat 1989). However,
in Nahal Tillah even storage jars are locally made,
whereas at other sites they were imported from Egypt.
All holemouth jars are made with calcite temper.
The tradition of using calcite as temper for cooking
utensils is well known in Early Bronze Age in Israel,
but also during other periods and other regions.
Lithic studies can also be used to monitor Egyp-
tian-Canaanite interaction. Unlike Tel Erani, where
the lithic assemblage seems to indicate the presence
of Egyptian flint knappers, no such evidence has
been found to date on the Halif Terrace. Systematic
studies of EB I chipped stone assemblages are rela-
tively recent, and the potential information concern-
ing ancient production and exchange that may be
derived from their analysis remains to be fully ex-
ploited. Although the assemblage reported here rep-
resents only one season of excavation, the chipped
stone assemblage from Nahal Tillah (Silo site, Halif
Terrace) is a relatively large collection and repre-
sents a northern Negev EB I lithic repertoire. The
EB I chipped stone assemblages from southern Is-
rael are of particular interest because of the continu-
ity with earlier Chalcolithic traditions but with the
added introduction of Egyptian influences, which are
substantiated by the pronounced presence of Egyp-
tian ceramic imports, local ceramic imitations, and
other Egyptian features. The ground stone assem-
blage from the Halif Terrace will be discussed in a
special publication concerning that subject, as yet
The 1994 lithic assemblage includes all observed
pieces, debitage or tools; nothing was discarded. Pits,
surfaces, and most ashy deposits were fully sieved
through /4" mesh. Thus, the lithic assemblage repre-
sents a good sample of debitage and tool types found
at the site. For brevity, typological categories are
based primarily on Rosen's synthesis (1989). Sum-
marized data on the debitage are presented by stra-
tum designations in table 5; flint tools are similarly
portrayed in table 6.
The 1994 excavations on the Halif Terrace (Silo
site) recovered 20,722 pieces of chipped stone, in-
cluding 28 hammerstones. Nearly 90% (n = 18,500
or 89.28%) of this assemblage is debitage, a typical
proportion of virtually all well-collected, nonspe-
cialized sites (Rosen 1989: 212). More than half of
the debitage recovered from the 1994 excavations
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
TABLE 5. Nahal Tillah Silo site debitage assemblage frequencies.
Debitage I I/II Hla lla/b Ib III III/IV IV Total Relative %
Chips and chunks 2838 1 247 72 911 1217 6 11 5303 28.66
Primary flakes 220 0 30 11 48 55 1 3 368 1.99
Flakes 5868 10 646 197 1629 2192 14 33 10589 57.24
Canaanean blades 34 0 5 1 3 12 0 0 55 .30
Blades 317 1 24 10 48 78 2 0 480 2.60
Bladelets 51 0 3 0 5 10 0 0 69 .37
Flake cores 173 0 14 15 30 33 0 0 265 1.43
Blade cores 4 0 0 1 2 1 0 0 8 .04
Bladelet cores 42 0 0 0 2 7 0 0 51 .28
Flake core fragments 774 4 87 44 170 158 3 3 1243 6.72
Blade core fragments 62 0 2 1 2 0 0 0 67 .36
Bladelet core fragments 2 0 0 0 0 0 0 0 2 .01
Total debitage 10386 16 1058 352 2850 3763 26 50 18500
Relative % 56.14 0.09 5.72 1.90 15.40 20.34 0.14 0.27 100.00
Core weights (kg) 13 0 2.2 .55 4.87 2.87 0 0 23.49
Core wt. (%) 55.34 - 9.37 2.34 20.73 12.22 - - 100.00
Waste weights (kg) 112.12 0.45 13.63 13.70 33.98 33.78 .80 .85 209.31
Waste wt. (%) 53.37 0.21 6.51 6.55 16.23 16.14 .38 .41 100.00
was collected from Stratum I contexts. As Stratum I
is disturbed overburden with mixed materials span-
ning the Chalcolithic to modern periods, meaningful
interpretations cannot be made on this stratum.
The debitage reflects a typical Early Bronze Age
lithic inventory, with nearly 95% of the assemblage
the result of flake based reduction (cores, core frag-
ments, and flakes; table 5). Next in order of predom-
inance are blades, blade cores, and core fragments,
which contribute just over 4% of the debitage as-
semblage. With a blades-to-core ratio of 60:1, it is
unlikely that all were produced in the excavation
area exposed in 1994. Based on one excavation sea-
son, it is still premature to interpret this pattern;
however, the question will be fully investigated as a
broader horizontal exposure is made at the site.
Bladelet production is also represented, albeit at
a relatively insignificant proportion of the total as-
semblage. However, the relatively large number of
bladelet cores and fragments suggests that these were
probably produced on-site as part of the domestic
tool production. As noted by Rosen (1989: 204),
bladelet tools are uncommon from EB I contexts and
are primarily used for the manufacture of transverse
arrowheads and microlithic lunates. The existence of
both tool types is so limited from the Silo site, and
from such poor contexts (all are from Stratum I; see
table 6), that the numerous bladelets cannot cate-
gorically be connected to production of these tools.
However, the majority of evidence for bladelet pro-
duction was retrieved from either Stratum I or Stra-
tum III contexts, so the possibility that these are
EB IA or Chalcolithic intrusions cannot be categor-
ically dismissed. A larger sample, particularly from
Stratum II contexts, will help resolve whether these
are part of the EB IB assemblage or an early stratum
at the site.
As reflected in the debitage, the predominant flint-
knapping technique on the Halif Terrace, as at most
EB I sites in the northern Negev, represents an expe-
dient flake industry, termed an ad hoc production by
Rosen (1989: 200). This flake tool industry (table 8)
accounts for almost 80% of the tool assemblage.
Table 8 groups tool classes together by general re-
duction technique. Tables 7 and 8 both indicate that
the predominant technique is flake reduction and that
the second most common type is that of backed and
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
TABLE 6. Absolute counts for 1994 Nahal Tillah Silo site tool assemblage.
Stratum % of Total
Tool class I Hla lla/b Ib III III/IV IV Total assemblage
Tabular scrapers 34 2 7 11 4 0 0 58 7.31%
End scrapers 19 3 1 6 3 0 0 32 4.03%
Side scrapers 51 3 2 8 6 0 0 70 8.83%
Generic scrapers 13 0 1 0 1 0 0 15 1.89%
Notches 83 11 6 16 1 0 1 118 14.88%
Denticulates 17 1 1 1 1 0 0 21 2.65%
Borers 35 5 1 10 4 0 0 55 6.94%
Drills 8 0 0 3 1 0 0 12 1.51%
Canaanean sickle blades 29 2 2 6 13 2 0 54 6.81%
Rtchd Canaanean blades 55 4 1 9 8 0 1 78 9.83%
Backed blades 22 3 2 5 3 0 0 35 4.41%
Sickle blades 42 8 2 12 11 0 0 75 9.46%
Notched blades 6 0 0 0 1 0 0 7 0.88%
Curved blades 1 0 0 0 0 0 0 1 0.13%
Retouched blades 76 7 1 18 10 0 0 112 14.12%
Lunates 3 0 0 0 0 0 0 3 0.38%
Transverse points 1 0 0 0 0 0 0 1 0.13%
Micro End scrapers 2 0 0 1 0 0 0 3 0.38%
Undetached microliths 1 0 0 0 0 0 0 1 0.13%
Retouched bladelets 3 1 0 2 0 0 0 6 0.76%
Core tools 6 0 1 0 0 0 0 7 0.88%
Hammer stones 18 1 2 1 6 0 0 28 3.53%
Egyptian knife 1 0 0 0 0 0 0 1 0.13%
Total per stratum 526 51 30 109 73 2 2 793 100.00%
% By strata 66.33% 6.43% 3.78% 13.74% 9.21% 0.25% 0.25% 100%
Misc. trim#* 997 82 36 181 128 2 3 1429 64.31%
# Not included in above counts
* Includes retouched flakes
TABLE 7. Debitage Types, 1994 Silo Site
Assemblage, Halif Terrace.
Type No. Totals Assemblage (%)
Flakes 10,957
Flake cores/fragments 1,508
Total 12,465 94.45
Blades 480
Blade core/fragments 75
Total 555 4.21
Bladelets 69
Bladelet core/fragments 53
Total 122 0.92
Canaanean blades 55 0.42
Grand total 13,197 100.00
retouched blade production (ca. 10%). This proba-
bly reflects the local nonspecialized domestic sphere
of lithic production and includes the various scrap-
ers, notches, denticulates, and retouched and utilized
flakes. We have used the term "generic scrapers" for
items that have clear scraper retouch but do not fall
into the "side" or "end" scraper categories. We do not
include more complex tools like tabular scrapers and
drills/borers in the ad hoc category. Drills are more
elongated than simple borers and show more care in
the production of their long point. As at other EB I
sites, such as cEn Shadud in the Galilee (Rosen
1985), notches and denticulates are the most com-
mon flake tool type (table 6). Blade tools, the second
most common type of tools, are primarily used in
the manufacture of sickle blades, a trend established
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
TABLE 8. Tool Types, Silo Site,
Halif Terrace (1994).
Tool type No. Assemblage (%)
Tabular scrapers 58 2.64
Flake tools, miscellany 1752 79.85
Canaanean blades 132 6.02
Sickles and retouched blades 230 10.48
Bladelet tools 14 .64
Core tools 7 .32
Egyptian 1 .05
Total* 2194 100.00
* Excludes hammerstones
in the Chalcolithic period. As noted (above), arrow-
heads virtually disappear in EB I settlement assem-
blages. The distribution of the major classes of tools
found at the Silo site on the Halif Terrace are shown
in fig. 28.
Although sometimes recovered from earlier Chal-
colithic contexts (Rowan and Levy 1994; Perrot
1992; Milstein and Ronen 1985), "Canaanean" blades
(large prismatic blades) do not become a significant
part of Palestinian lithic assemblages until EB I. The
technological contrast between the production of
Canaanean blades and the flake industry, noted pre-
viously (Rosen 1983; Schick 1978), underscores the
possible dichotomy between the local and nonlocal
production spheres. Despite the potential superior-
ity of the prismatic Canaanean blades (Rosen 1983;
1989) by reversing them in the haft, their usage con-
tinues in tandem with regular sickle blades, the latter
a continuity with the preceding Chalcolithic tradi-
tion. The important contrast is the paucity of blade
cores, and particularly Canaanean blade cores, though
more recent (1995) excavations indicate that some
reduction of Canaanean blades may have occurred
on site during EB I. These continued excavations
have raised the possibility that tabular scraper and
Canaanean blade production may have used the same
1994 Nahal Tillah Silo Site Tool Assemblage
I Ila
- lia/b
i lb
Number O Ill/IV
30 __
o oj ?2. : c5 ei o i i ri i ed -J j 05 Cj o j Cls
sor- IV
T.S. = Tabular scrapers N = 58 C.S.B. = Canaanean sickle blades N = 54 T.P. = Transverse points N = 1
E.S. = End scrapers N = 32 R.C.B. = Retouched Canaanean blades N = 78 M.E.S. = Micro end scrape N = 3
S.S. = Side scrapers N = 70 B.B. = Backed blades N = 35 U.M. = Undet. microliths N = 1
G.S. = Generic scrapers N = 15 S.B. = Sickle blades N = 75 R.BLTS. = Retouched bladelets N = 6
N. = Notches N = 118 N.B. = Notched blades N = 7 C.T. = Core tools N = 7
D. = Denticulates N = 21 C.B. = Curved blades N = 1 H.S. = Hammerstones N = 28
B. = Bores N = 55 R.B. = Retouched blades N = 112 E.K. = Egyptian knife N = 1
Dr. = Drills N = 12 L. = Lunates N = 3
Fig. 28. Distribution of flint tools by stratum, silo site, 1994.
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
nonlocal, fine brown flint as previously suggested
for Tel Halif in the EB III strata (Seger et al. 1990),
where some Canaanean blade cores were found. Fu-
ture analysis will focus on correlation of debitage
flint types with tools. What remain unclear are the
reasons for the continued use of the smaller sickle
blades alongside the new Canaanean technology.
There are a number of explanations for the coexist-
ence of both sickle blade types; one is the operation
of different specialists creating the two types inde-
pendently, trading from different production areas;
another may be that the Canaanean and other sickle
blades represent a by-product of socioeconomic dif-
ferentiation between groups at the site. The possible
existence of two spheres of blade production lend
support to Rosen's model (1983: 28) of localized
trade. There are other possibilities, and the atypical
overlap of these two production types (Rosen 1989:
218) will be studied in the future.
Tabular scrapers are large cortical flakes struck
from large flint nodules or veins of flint. The use of
"tabular" is retained here for recognition and does
not necessarily imply procurement from tabular veins
of flint (Rosen 1989: 200). Retouch is typically dor-
sal and usually limited to the edge. Bulbs of per-
cussion are often removed during manufacture, or
thinned with ventral retouch. The flint used at Nahal
Tillah is typically a dark to light brown, very homog-
enous and fine grained. Tabular scrapers may rep-
resent another specialized tool production, and may
have been manufactured for trade (Rosen 1989: 202-
3). Shapes of these scrapers vary from very elon-
gated, ovoid, to virtually rectangular. None have the
engraving found on some specimens seen from Arad.
In contrast with lithic assemblages from the pre-
ceding Chalcolithic period, there is a lack of core
tools (i.e., axes, adzes, chisels, and picks) in the
EB I period (fig. 28; table 6). The virtual disappear-
ance of core tools is common to EB I sites and was
noted previously at Tel Halif (Seger et al. 1990: 23)
and Palestine in general (Rosen 1989: 216, table 1).
The reason for the disappearance of these core tools
is unclear. The suggestion that copper axes replaced
those of flint seems unlikely (Rosen 1984), given
the high value and subsequent recycling of copper
Finally, there is limited evidence for Egyptian flint
imports. A poorly crafted bifacial Egyptian knife
found in 1994 is clearly of nonlocal origin based on
both technique and the amber-colored flint. This knife
is invasively retouched, probably using pressure-
flaking retouch, and most closely resembles a knife
from Tarkhan (Petrie 1914: pl. 7:5). In 1995, several
fragments of beautifully pressure flaked "ripple-
knives" came to light (fig. 29a). In addition, a
number of curved sickle blades may represent an
Egyptian production technique utilizing local flints.
Future research, with a much larger sample size, will
concentrate on delineating more possible Egyptian
components in the lithic industry at Nahal Tillah.
Comparison with EB I sites with known Egyptian
production techniques such as Tel Erani (Rosen 1988),
cEn Besor (Gophna and Friedmann 1993; Yeivin
1976) and Site H in the Wadi Gaza (Roshwalb 1981)
will provide additional clues to the problem of eth-
nicity and exchange mechanisms in southern Canaan.
In addition to the wealth of EB I lithic remains, to
our surprise during the 1995 season we discovered
two pressure flaked arrowheads that seem to date to
the Pottery Neolithic period (fig. 29b). According to
Gopher's (1994: 43) in-depth study of Neolithic ar-
rowheads, one of these is clearly a Type A8 or small
Nizzanim point (length = 2.68 cm). The other seems
to fall within the A7 Givat ha-Parsa point tradition,
although its rather long length (ca. 4.22 cm; the tip
is snapped off) might argue for its being a Jericho
point (Gopher 1994: 43, Type A4) from the PPNB.
The fact that this example exhibits complete pres-
sure flaking on both ventral and dorsal sides argues
in favor of its being a Givat ha-Parsa point, mak-
ing both samples Pottery Neolithic in date. Sites dat-
ing to the Pottery Neolithic period are extremely rare
in the ecotone that encompasses the very southern
Shephelah and northern Negev foothill zone. This
may be due to the burial of these sites through nat-
ural formation processes related to slope erosion in
the southern Judaean mountains. In any event, the
discovery of these arrowheads suggests that in future
research it may be possible to isolate a Pottery
Neolithic occupation level on the Halif Terrace.
Geophysical surveys were made on both the Halif
Terrace and the Abu Hof village. The presence of
numerous metal fenceposts associated with cattle
enclosures on the Halif Terrace made it impossible
to obtain reliable geophysical results from that site.
However, the open-air village at Abu Hof west of
Kibbutz Lahav was undisturbed, making it amenable
to geophysical research. Abu Hof is a Chalcolithic
village site comprised of two distinct areas: a low-
This content downloaded from on Mon, 09 May 2016 12:37:28 UTC
All use subject to
-"-- ii4ilii
.?i:jtxarars~itm::p~:=tr.ll~%~s~ ~L~ I e~ II li?I::::f::::: . ?1: :* .
::::::? ?
i: i
??? ..;
~::? :??:?
...: :
:..:: ::"
3 Cr"17
Fig. 29. a. Fragment of Egyptian ripple knife (length = 3.3 cm); b. Two Neolithic arrowheads from silo site, Halif Terrace;
left, Nizzanim point; right, ha-Parsa point (1995).
land area on a collu